Research Studies
Study Tracker
Gestational fluoride exposure compromises ovarian function via SIRT1-ATF4 axis in female offspring.
| Ecotoxicology and Environmental Safety | Gu Y, Shi Y, Wei B, Zhao Y, Huo M, Zeng L, Zhao Z, Song Y, Yu X, Chen Y, Sun M. | 311:119872.
Protective effect Resveratrol SIRT1 Apoptosis Ovary/Uterus Animal Study
View PDF

Abstract

Full-text study online at
https://www.sciencedirect.com/science/article/pii/S0147651326002010

Highlights

  • Gestational fluoride exposure induces ovarian dysfunction in female offspring.
  • Fluoride activates SIRT1-ATF4 axis, driving ovarian apoptosis via BAX-Caspase3.
  • ATF4 promoter hyperacetylation enhances transcription, implying epigenetic toxicity.
  • Resveratrol rescues fluoride-impaired ovarian function via SIRT1 activation.

Fluoride exposure is linked to impaired reproductive capacity in female mammals. Ovarian reserve is established during fetal development. Maternal exposure to adverse environmental factors during pregnancy can impair offspring ovarian function, with prenatal fluoride exposure potentially contributing to ovarian dysfunction, though underlying mechanisms remain unclear. We established a model with C57BL/6J mice that were given 100 mg/L sodium fluoride (NaF) from the 1st day of pregnancy to the 20th day. NaF exposure impaired ovarian function in female offspring, manifested as reduced ovarian reserve, disrupted follicular maturation, and increased follicular atresia. Molecular analyses demonstrated elevated ATF4 expression, decreased SIRT1 levels, upregulated CHOP, and activation of the BAX/BCL-2 apoptotic pathway in offspring ovaries, indicating ovarian apoptosis as the primary mechanism of ovarian dysfunction. Chromatin immunoprecipitation (ChIP) revealed that NaF exposure increased H3K9ac binding at the Atf4 promoter region, suggesting SIRT1-mediated regulation enhances –Atf4 acetylation and transcription. Consistently, NaF-treated KGN cells showed reduced SIRT1 and increased ATF4. Critically, resveratrol, a specific SIRT1 activator, effectively upregulated SIRT1 expression in both fluoride-exposed animal models and KGN cells. This intervention ameliorated ovarian dysfunction phenotypes and inhibited granulosa cell apoptosis. These results establish SIRT1-ATF4 signaling as the mechanistic core of fluoride-induced ovotoxicity and validate its targeting with Resveratrol as a therapeutic strategy against environmental ovarian dysfunction.

    Keywords: Fluoride exposure; Ovarian dysfunction; SIRT1-ATF4 signaling; Resveratrol

    1. Introduction

    Fluoride is present in environmental matrices such as soil, water bodies, industrial effluents, cosmetics, and food products. It is commonly added to toothpaste, mouth rinses, and drinking water for preventing tooth decay. However, excessive fluoride exposure may cause a range of adverse effects (Geng et al., 2014, Ommati et al., 2025). The main effects of fluoride exposure are manifested in dental and skeletal systems. In addition, recent reports have demonstrated that fluoride exposure may induce perturbations to hematological homeostasis, liver function, kidney physiology, and nervous system integrity. Recently, the toxicological profile of fluoride was extended to reveal its endocrine-disrupting properties, which have a particular impact on reproductive health. For example, fluoride has a dose-dependent effect on testicular dysfunction in males and ovarian follicular atresia in females (Ommati et al., 2025).

    It is well known that the fertility rate in China is decreasing dramatically (Wang, Kong, Fu, and Qiao, 2024), ovarian dysfunction is gradually attracting attention from women, which is mainly manifested in quantitative and qualitative alteration of the ovarian oocyte reserve (May-Panloup et al., 2016). Ovarian reserve, defined as the finite pool of primordial follicles established during prenatal development, serves as a critical determinant of female reproductive lifespan and shows marked vulnerability to perturbations in the intrauterine environment (Yao et al., 2022). Recent studies have shown that adverse intrauterine environment induces changes in epigenetic modifications and adult diseases. Adverse intrauterine environment which refers to a detrimental pregnancy environment jointly triggered by maternal, placental, and environmental factors such as maternal malnutrition, metabolic disorders during pregnancy (e.g., gestational diabetes, obesity), placental dysfunction, intrauterine hypoxia, stress, exposure to toxic substances (e.g., fluoride, heavy metals), and maternal inflammatory responses, can impair normal fetal development and developmental programming, induce stable epigenetic modifications, and subsequently increase an individual’s risk of developing multiple chronic adult diseases (Bölte et al., 2019, Yao et al., 2022). This view is exactly in line with the Fetal Origins of Adult Disease (FOAD) hypothesis and the Developmental Origins of Health and Disease (DOHaD) hypothesis proposed by David Barker (Barker, 1990). The hypothesis suggests that detrimental intrauterine stimulation correlates with an elevated risk of various adult diseases in the offspring.

    Numerous investigations have demonstrated that sodium fluoride exposure can compromise ovarian architecture in female mammals, diminish follicle counts, disrupt ovarian hormone secretions, and thus impact fertility (Dong et al., 2023, Geng et al., 2014, Zhou et al., 2013). As research in this domain advances, there is increasing concern about the impact of fluoride on general health, especially on the reproductive system. Nevertheless, not much study has documented the impact of fluoride exposure during gestation on ovarian function in next generations. Based on the FOAD and DOHaD hypothesis and previous studies, it is reasonable to hypothesise that sodium fluoride exposure may affect ovarian function in offspring. The objective of this study was to examine reproductive system impairment in offspring resulting from maternal fluoride exposure during gestation. Our research revealed for the first time that maternal fluoride exposure may hinder ovarian development and folliculogenesis in offspring.

    Sirtuin-1 (SIRT1) is a multifunctional enzyme, which is a nicotinamide adenine dinucleotide (NAD+)-dependent class III histone deacetylase and regulates the epigenetic landscape by dynamically regulating the level of histone acetylation. SIRT1 mediates bidirectional transcriptional regulation through site-specific deacetylation at histone loci, particularly H3K9 and H4K16 (Etchegaray and Mostoslavsky, 2016; Y. Li, Daniel, and Tollefsbol, 2011), consequently generating dose-dependent regulation of downstream gene expression, which is crucial for the modulation of various biological processes, including cellular survival/apoptosis, proliferation/senescence, inflammatory signaling, metabolic homeostasis, and adaptive stress responses (Grzeczka and Kordowitzki, 2022, Nishigaki et al., 2022). A great deal of studies have verified that SIRT1 is significant in regulating ovarian function, critically adjusting oocyte quality, follicular development, and the aging of ovarian granulosa cells (Nishigaki et al., 2022, Park et al., 2020).

    ATF4 (Activating Transcription Factor 4) is a constituent of the ATF/CREB (cAMP response element-binding protein) family of basic leucine zipper (bZIP) transcription factors (Di et al., 2019). ATF4 mRNA is extensively expressed in multiple tissues and is involved in numerous biological processes, including endoplasmic reticulum stress, apoptosis, metabolism, and stress adaptation. Alterations in its expression have been associated with various diseases. It also plays an crucial role in regulating ovarian function (J. Liu et al., 2019; Ma et al., 2024). We used bioinformatics software (Cistrome data and ChIP-Atlas) to predict the possible binding of H3K9ac and H3K27ac in the ATF4 promoter region, which led us to hypothesise that SIRT1 further regulates ATF4 transcription by modulating the acetylation of H3K9 or H3K27, thereby inducing changes in the downstream molecular pathways and consequently regulating ovarian function.

    Resveratrol (3,5,4′-trihydroxytrans stilbene, Res), a naturally occurring phytochemical, has been demonstrated to be a selective activator of SIRT1 (Borra et al., 2005, Pirola and Fröjdö, 2008). Res exhibits cytoprotective properties against various diseases, including anti-cancer, anti-inflammatory, antioxidant, and anti-aging actions, along with estrogen control. Numerous studies have shown that Res is intricately linked to follicular growth and development, extends ovarian lifetime, and mitigates age-related sterility in rodents (Chen et al., 2022, Nie et al., 2021, Wu et al., 2019).

    Based on the previously mentioned information, we proposed that fluoride exposure during pregnancy, as an adverse intrauterine stress, may affect ovarian function in offspring. Therefore, we applied a mouse model of prenatal fluoride exposure and cell culture assays to validate our hypothesis. With this model, we established the key role of the SIRT1-ATF4 pathway in ovarian hypoplasia in fluoride-exposed female offspring and revealed the dominant role of epigenetics (mainly histone acetylation) in it. This indicated that the SIRT1-ATF4 pathway could be a promising target for safeguarding against fluoride-induced ovarian damage. Furthermore, we have confirmed the protective efficacy of resveratrol both in vitro and in vivo, which can protect against sodium fluoride-induced impairment to ovarian function in offspring.

    2. Materials and methods

    2.1. Chemicals and reagents

    Sodium fluoride (#S111590, Fengxian, Shanghai, China) was purchased from Aladdin, and resveratrol (#SRT501, Monmouth Junction, NJ, USA) was obtained from Med Chem Express. Unless specified otherwise, all other components were analytical grade and obtained from local sources.

    2.2. Experimental subjects

    The Experimental Animal Center of Soochow University (Suzhou, China) provided the 8-week-old C57BL/6J mice used in this study. The First Affiliated Hospital of Soochow University’s Animal Experiment Ethics Committee approved all experimental protocols and procedures. All specific pathogen-free mice were housed in cages containing no more than five mice each, under controlled conditions: a constant temperature ranging from 22°C to 24°C, humidity levels between 40 % and 60 %, a 12 – hour light/dark cycle from 8 am to 8 pm, and continuous air exchange. Female mice were randomly divided into two groups, with 10 mice in each group: the control group and the sodium fluoride (NaF) treatment group. Researches have shown that mice excrete fluoride more efficiently than humans (Zhao et al., 2020). Based on population-level environmental exposure, the biological characteristics of mice, prior researches, and previous molding experience in our institute, a fluoride ion concentration of 100 mg/L was used (Cai et al., 2025; W. Li et al., 2022). Male and female mice in each group were randomly mated overnight at a ratio of 1:2. Upon the presentation of a vaginal plug in female mice throughout the morning, the day was designated as gestational day 0.5 (GD 0.5). Pregnant mice were housed individually in cages under the same environmental and exposure conditions. The pregnant female mice were then given drinking water containing 100 mg/L NaF by oral gavage. 21 days later, when the offspring were born, the 100 mg/L NaF drinking water was replaced with normal drinking water. After weaning at 21 days of age, female and male offspring were housed separately in different cages. Female offspring mice were used in this study.

    2.3. Resveratrol administration management

    Resveratrol was prepared following the manufacturer’s protocol (Med Chem Express). A stock solution of the chemical was dissolved in dimethyl sulfoxide (DMSO) to a concentration of 30 mg/mL and refrigerated at -20 C to preserve the chemical. The operational solution was formulated by diluting the stock solution to 10 % (v/v) with DMSO, followed by the sequential addition of 40 % polyethylene glycol 300 (PEG300), 5 % Tween 80, and 45 % saline in succession. The mixture was vortexed thoroughly until a homogeneous and clear solution was achieved. F1 female mice were divided into three experimental groups (n = 10 per group): a control (Con) group, a sodium fluoride (NaF) group, and a NaF with resveratrol rescue (NaF+Res) group. Starting at 4 weeks of age, the rescue group received daily intraperitoneal (i.p.) injections of the afore-mentioned working solution (30 mg/kg/day), whereas the Con and NaF groups were given an equivalent amount of vehicle solution (DMSO + PEG300 + saline) using the same method. This regimen would continue for 4 consecutive weeks, concluding when offspring reached 8 weeks of age.

    2.4. Estrous cycle determination

    The daily estrous cycle of the offspring mice was determined for 14 consecutive days, starting from 8 weeks of age. Saline lavage was employed to obtain vaginal epithelial cells, subsequently followed by smear preparation, fixation in anhydrous ethanol, and hematoxylin-eosin (H&E) staining. The four stages of the estrous cycle can be distinguished by the relative proportions of leukocytes, nucleated epithelial cells, and cornified squamous epithelial cells. The proestrus smear is characterized by a predominance of nucleated epithelial cells. The estrus smear is mainly composed of cornified squamous epithelial cells. The metestrus smear has a similar ratio of leukocytes, cornified squamous epithelial cells and nucleated epithelial cells. While the diestrus smear is dominated by leukocytes (Sano et al., 2020).

    2.5. Measurement of serum hormone levels

    Blood samples were obtained from F1 female offspring at 10 weeks of age via retro-orbital plexus puncture under anesthesia. Ensure blood collection occurs during the mouse’s diestrus phase between 9:00 AM and 10:00 AM (Wen et al., 2024). Following clot formation at room temperature (25°C, 30 min), specimens were centrifuged (4°C, 1000 × g, 10 min) to isolate serum fractions. Aliquoted supernatants were stored at -80°C until further examination. We used species-specific ELISA kits from Shanghai, China, to measure the levels of estradiol (E2; Cat# ml063198, Mlbio), follicle-stimulating hormone (FSH; Cat# ml001910, Mlbio), and anti-Müllerian hormone (AMH; Cat# ml037597, Mlbio) in serum. Standard curves were generated with serial dilutions (0–500 pg/mL) in parallel with biological samples, which were loaded in duplicate. After 60 min of incubation with the primary antibody at 37°C, the plates were examined with HRP-conjugated secondary antibodies (100 uL/well, 1:5000 dilution) followed by five washes with PBS-T (1 min/wash). Colorimetric development was initiated with TMB substrate (A/B mixture, 50 uL/well) under dark conditions. Lastly, the termination solution was added to each well and the optical density (OD value) was detected at 450 nm within 15 min.

    2.6. Ovarian morphology and HE staining

    Following blood collection, we surgically removed the ovaries of 10-week-old F1 female mice under anesthesia, ensuring that the ovaries were collected during the mice’s diestrus phase. Then they were fixed in 4 % paraformaldehyde at 4°C for 24 h, dehydrated using a graded ethanol series (70 % – 100 %), and embedded in paraffin. Serial cuts were performed on 5-um-thick tissue sections, and then they were stained by standard hematoxylin – eosin (HE). The three largest cross-sectional ovarian profiles from each animal were selected for quantitative histomorphometry analysis. We carried out brightfield imaging with a Nikon Eclipse Ci-L microscope integrated with NIS – Elements BR 5.30.00 software.Under the microscope, the continuous sections of all types of follicles (total, primordial, primary, and mature including secondary and atretic) were counted at 50 m intervals.

    2.7. TUNEL staining

    We used the Terminal deoxynucleotidyl transferase dUTP Nick End Labeling (TUNEL) test to find cells that were dying in ovarian tissue. The previously mentioned method was used to get paraffin sections of the ovaries. Tissue sections were incubated with TUNEL reaction mixture (Beyotime Biotechnology, C1088, Shanghai, China) for 1 h at 37°C in a humidified light-blocked chamber, followed by three 5-minute washes with PBS. DAPI (Sigma-Aldrich, D9542) was used to stain the nuclei. A Zeiss fluorescence microscope was used to look at and take pictures of the sections. ImageJ software was used to measure the relative fluorescence intensity.

    2.8. Cell culture and treatment

    We got hold of the KGN human granulosa cell line from Bena Culture Collection (#337610, BNCC, China) and cultured it in DMEM/F12 medium (#SH30023.01, Hyclone, Logan, UT, USA) supplemented with 10 % fetal bovine serum (FBS, #A5670701, Gibco, USA) and 1 % penicillin/streptomycin (#SV30010, Cytiva, USA). Cells were cultured in a humidified incubator at 37°C with 5 % CO2 and the medium was changed every 48 h. KGN cells were placed into 6-well plates at a density of 1 × 105 cells per well. Cells reaching 70–80 % confluency were designated as the control group, which received fresh maintenance medium without experimental interventions. Parallel wells were allocated to treatment cohorts. All experimental batches were set up with three replicate control wells to ensure consistency.

    Based on the results of CCK-8 viability assay, experimental groups were established as follows: (1) NaF-treated group: 3 mM NaF; (2) Rescue group: co-treatment with 3 mM NaF and 25 uM resveratrol. Following a 24-hour drug exposure period, cells were harvested for further experimentation.

    To functionally validate the regulatory interaction between SIRT1 and ATF4, we conducted siRNA-mediated gene silencing in KGN cell models following the manufacturer’s optimized protocol. The si-ATF4 oligonucleotide was chemically synthesized by Gene Pharma (Suzhou, China), and the siRNA sequence with the best silencing efficiency was screened by pre-experimentation : sense 5′-GCCUAGGUCUCUUAGAUGATT-3′ and antisense 5?-UCAUCUAAGAGACCUAGGCTT-3′. The si-SIRT1 oligonucleotide was synthesized by Sangon Biotech (Shanghai, China), with the maximal knockdown efficacy sequence being sense 5′-UUCAAUAUCAAACAUCGAU-3′ and antisense 5′-AGCGAUGUUUGAUAUUUGAA-3′. Cells were collected 48 h after transfection for subsequent experiments.

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

    We used TRIzol reagent (A33251, TaKaRa, Japan) to extract total RNA from ovarian tissues or KGN cells. Then, we used the RevertAid First Strand cDNA Synthesis Kit (K16225, Thermo Fisher Scientific, USA) to reverse transcribe the RNA according to the manufacturer’s instructions. The Bio-Rad CFX96 real-time fluorescence quantitative PCR system was used to measure gene transcription levels. The data were then normalised to the B-actin reference gene and analysed using the 2-??Ct method. All experiments were conducted in triplicate. The RT-qPCR primer sequences are listed in Table S1 (Supporting Information).

    2.10. Western blotting

    Mouse ovarian tissue or KGN cell samples were homogenized in RIPA lysis buffer (P0013B, Beyotime Biotech, China) with added phosphatase/protease inhibitor cocktail (P002, NCM Biotech, Shanghai, China). After sonication, samples underwent centrifugation at 12,000 × g for 30 min at 4°C. Protein content was determined using the bicinchoninic acid (BCA) Protein Quantification Kit (P0012S, Beyotime Biotech, China). For each individual sample, 25 ug of protein was separated by means of 7.5–12.5 % SDS PAGE (sodium dodecyl sulfate – polyacrylamide gel electrophoresis) according to molecular weight, and then electrotransferred onto 0.22 um or 0.45 m nitrocellulose membranes (HATF00010 Millipore, Germany).The membrane was incubated with 5 % nonfat milk for 2 h to eliminate nonspecific binding, membranes were rinsed three times (10 min per wash) and subsequently probed with primary antibodies under gentle agitation at 4°C for 16 h. Information on the primary antibodies used in our experiments is contained in Table S2. The membranes underwent three washes with TBST. Subsequently, they were treated with a horseradish peroxidase-conjugated secondary antibody (ThermoFisher, 1:5000 dilution) at ambient temperature for one hour. We used AlphaEase FC software (v3.1.2) for quantitative densitometric analysis. B-actin was used as an internal control to normalize protein expression levels (Table S2).

    2.11. Histone extraction

    Histone extractions were purified from ovary tissues or KGN cells using the EpiQuik Total Histone Extraction Kit (OP-0006, Epigentek, USA) with minor modifications while adhering to the manufacturer’s protocol. Purified histones were mixed with 5 × loading buffer (CW0027, Cowin Biotech, China) after concentration measurement and denatured at 95°C for 10 min prior to subsequent Western blot analysis.

    2.12. Chromatin immunoprecipitation (ChIP) assay

    ChIP assays were conducted by means of the Sample ChIP Plus Enzymatic Chromatin IP Kit. (9005, Cell Signaling Technology, USA). The brief steps are as follows: take 25 mg of ovarian tissue or 1 × 10^7 cells, crosslink them with 37 % formaldehyde at room temperature for 30 min, and then add glycine to quench the reaction. The tissue was washed three times with PBS buffer containing protease inhibitors and PMSF. A Dounce homogeniser was used to turn the tissue into a suspension of single cells. Chromatin fragments with a length of 150–900 bp were obtained by combined treatment with micrococcal nuclease and ultrasound. Chromatin was centrifuged at 16000 g for 10 min at 4 °C. Overnight precipitation was performed using H3K9ac antibody and IgG control to precipitate protein/DNA complexes. After purification, the immunoprecipitated DNA was quantified by RT-qPCR using specific primers and normalized to input. Primer sequences for gene promoter amplification are listed in Table S1 (Supporting Information). The IP efficiency was manually calculated using the percent input method according to the manufacturer’s protocol, with the formula shown below: Percent Input = 2 % × 2^(C[T] 2 %Input Sample – C[T] IP Sample). The immunoprecipitated signal from each experiment was quantified as a percentage of the total input chromatin using this methodology.

    2.13. Flow cytometry assay

    An Annexin V-FITC/PI Apoptosis Detection Kit (Cat# P-CA-201; Pricella Biotechnology, Wuhan, China) was employed to evaluate apoptosis in KGN cells. Cells were collected using EDTA-free trypsin, washed twice with PBS, and subsequently resuspended in 100 uL of 1 × binding buffer. Cell suspensions were successively labelled with Annexin V-FITC (5 uL) and PI (10 uL), followed by a 15-minute incubation at 25 °C in light-protected environment. The fluorescence signals from 1 × 10^4 cells per sample were acquired utilising a CytoFLEX LX flow cytometer (Beckman Coulter, Brea, CA, USA). The data were analysed using CytExpert software (v2.4), with live cells (Annexin V-/PI-) as the internal reference.

    2.14. CCK8 assay

    After drug treatment, the CCK-8 colorimetric assay (C0037, Beyotime, Shanghai, China) was used to measure how well KGN cells could grow. KGN cells were resuspended in 96-well flat-bottomed microplates at a density of 5 × 10³ cells per well in 200 uL of complete medium. Once the pretreatment with different drugs was carried out for 24 h, 20 uL of CCK-8 reagent was introduced to each well. A background control containing medium and CCK-8, but no cells, was also prepared for photometric calibration. Then the plates were put for a 2 – hour incubation at 37 °C in a chamber with 5 % CO2 and a damp atmosphere. Absorbance at 450 nm was measured by means of a microplate reader (Tecan, Männedorf, Switzerland). The formula was used to calculating cell viability (%): [(Absorbance of experimental wells – Absorbance of blank wells) / (Absorbance of control wells – Absorbance of blank wells)] × 100.

    2.15. Data and statistical analysis

    Quantitative results are presented as mean ± standard error of the mean (SEM) obtained from three independent triplicate measurements. Statistical analysis was conducted with GraphPad Prism (version 9.5.0). Two-tailed unpaired Student’s t-tests with Welch’s correction were used for two-factor comparisons. One-way or two-way ANOVA with Bonferroni post hoc tests were used for multigroup comparisons. Significance thresholds were defined as: *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, with non-significant (ns) demarcated for P >0.05. Unless otherwise specified, all inferential comparisons are based on the solvent-treated control group.

    3. Results

    3.1. Gestational fluoride exposure adversly affects ovarian function in female offspring

    To investigate the impact of maternal fluoride exposure during pregnancy on offspring ovarian function, we adopted the previously established animal model (Fig. 1A). Following continuous exposure to 100 mg/L NaF during gestation, the two pairs of incisors in pregnant mice from the NaF group exhibited a more pronounced whitening effect compared to the control group (Fig.S1 A). This finding further validates the rationality and success of establishing the experimental mouse model. Morphometric analysis of the ovaries at postnatal 10 weeks showed significant impairment of ovarian development in NaF-exposed offspring. Compared to the control group, the fluoride group exhibited markedly reduced ovarian size and ovarian index. (ovarian weight / body weight ratio) (Fig. 1B, C). We did a vaginal cytology analysis to find out more about murine estrous cyclicity (Fig. 1E). We recorded the length of time that the two groups of mice went through each cycle, calculated the average value and then plotted a line graph. Control animals exhibited consistent 5-day cycles, while NaF-exposed mice exhibited extended cyclicity (approximately 8 days), primarily due to prolonged diestrus and proestrus phases (Fig. 1D). Serum ELISA analyses indicated specific endocrine modifications in NaF-exposed mice, showing increased follicle-stimulating hormone (FSH) levels alongside decreased estradiol (E2) and anti-Müllerian hormone (AMH) concentrations (Fig. 1F-H). Histological analysis of ovarian sections demonstrated pronounced morphological alterations in the NaF group, characterized by diminished ovarian volume and substantial depletion of primordial follicles. While the ratio of primary to secondary follicles remained consistent across groups, a notable rise in the count of atretic follicles was recorded in NaF-treated mice (Fig. 1I, J). These experimental results collectively illustrate that gestational fluoride exposure causes ovarian developmental impairment in female progeny, presenting as ovarian dysfunction.

    Fig. 1

    1. Download: Download high-res image (1MB)
    2. Download: Download full-size image

    Fig. 1. Gestational fluoride exposure triggers ovarian dysfunction in female offspring. (A) Schematic diagram of animal model experiments; (B) Ovarian appearance dimensions in control and NaF group; (C) Ovarian index in 10-week-old mice in two groups, ovarian index = ovarian weight (mg)/mouse body weight (g), (n = 10 each group); (D) Comparison of the estrus cycle in two groups, (n = 10 each group); (E) Schematic diagram of HE staining microscopic examination of vaginal smears during the estrous cycle in mice; (F,G,H) Serum FSH, E2, and AMH levels in 10-week-old mice in two groups (n = 6 each group); (I) HE staining of ovary sections from two groups of mice, red mark: primordial follicle, green mark: secondary follicle, blue mark: antral follicle, yellow mark: atretic follicle; (J) Proportion of follicles of all stages in the ovaries of two groups (n = 6 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    3.2. Gestational fluoride exposure induces ovarian cell apoptosis in female offspring through ATF4-CHOP signaling pathway

    Data from our study suggested that fluoride-exposed offspring had weakened ovarian function. We also found a large number of atretic follicles in fluoride-exposed offspring. As follicular degeneration is considered to be caused by cell apoptosis and there is a mechanistic relationship between ovarian dysfunction and abnormal apoptosis regulation in ovarian tissues (Stringer et al., 2023, Yan et al., 2024), we subsequently performed TUNEL staining on ovarian tissue sections to corroborate this mechanism. Quantitative analysis indicated a substantial increase in TUNEL-positive apoptotic cells in the fluoride-treated group relative to controls (Fig. 2A, B). To further explore the molecular mechanisms, we validated through RT-qPCR that Atf4 and Chop were significantly upregulated in NaF group. Literature and KEGG pathway analysis suggested that this pathway may activate apoptosis-related cascades (Gao et al., 2023, Shi et al., 2025). The subsequent validation of apoptosis-associated genes demonstrated a significant downregulation in the anti-apoptotic gene Bcl2 within the NaF group, alongside an increase in the Bax/Bcl2 ratio, indicative of apoptosis activation (Fig. 2C). In addition, we extracted total protein from the ovarian tissues and performed western blotting analysis. The findings indicated that the protein levels of ATF4 and CHOP were markedly elevated in the NaF group. The concentration of the anti-apoptotic protein BCL-2 was significantly reduced in the NaF group. Concurrently, the expression levels of pro-apoptotic BAX and cleaved Caspase-3 increased, indicating activation of the apoptotic pathways (Fig. 2D, E).

    Fig. 2

    1. Download: Download high-res image (799KB)
    2. Download: Download full-size image

    Fig. 2. Gestational fluoride exposure induces ovarian cell apoptosis in female offspring through ATF4-CHOP signaling pathway. (A, B) TUNEL staining of ovary sections and tunel positivity level statistical graph of two groups of mice (n = 6 each group); (C) Atf4, Chop, Bcl2, Bax mRNA expression levels and Bax/Bcl2 ratio in control and NaF group (n = 5 each group); (D, E) ATF4, CHOP, BCL2, BAX, Cleaved-Caspase3 protein levels in control and NaF group (n = 6 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    3.3. ATF4-CHOP pathway mediates apoptosis in NaF-treated KGN cells

    To explore the mechanistic effect of ATF4-CHOP pathway on the ovarian dysfunction of offspring induced by fluoride exposure, we constructed a KGN cell model. KGN cells were cultured in the culture medium containing different concentrations of sodium fluoride (NaF) ranging from 0 to 4 mM. The cell viability was analyzed by CCK-8 assay, and Atf4 expression levels were detected by RT-qPCR. Based on these results, 3 mM NaF was selected as the working concentration (Fig. S1A-C). Afterwards, we explored whether NaF could induce apoptosis in KGN cells. The results showed that NaF could induce apoptosis in KGN cells as analyzed by flow cytometry (Fig. 3A, B). RNA was extracted from both control and NaF-treated groups, and the samples were further analyzed by RT-qPCR. The findings indicated elevated mRNA expression levels of ATF4 and CHOP in NaF-treated KGN cells. Correspondingly, gene expression in the apoptosis pathway also changed, and we observed an increase in the Bax/BCL-2 ratio. (Fig. 3C). To further verify the alterations in ATF4 and CHOP protein levels we extracted total cellular proteins. Western blot analysis showed significant increase of the two proteins after treatment with NaF. Concurrently, alterations in apoptotic pathway proteins were noted, featuring increased BAX expression, reduced BCL-2 levels, and heightened activation of cleaved Caspase-3 (Fig. 3D,E). To further confirm that ATF4 involved in the activation of the downstream apoptotic cascade, we transfected KGN cells with si-ATF4. The silencing efficiency of ATF4 mRNA after transfection was confirmed ( Fig. S1E). A parallel reduction of CHOP protein expression was seen at 48 h after transfection, and down-stream apoptotic regulators were altered accordingly, characterized by reduced expression of Bax, increased expression of BCL-2 and reduced expression of cleaved Caspase-3. (Fig. 3F, G). Thus, it can be concluded that NaF treatment of KGN cells modulates the ATF4–CHOP pathway, leading to activation of the apoptotic pathway and subsequent cell death. Next, we will further elucidate the mechanisms underlying the upregulation of ATF4 expression.

    Fig. 3

    1. Download: Download high-res image (818KB)
    2. Download: Download full-size image

    Fig. 3. ATF4-CHOP pathway mediates apoptosis in NaF-treated KGN cells. (A, B) Flow cytometric analysis revealed differential apoptosis rates in NaF-Treated vs. Control KGN Cells (n = 3 each group); (C) Comparative analysis of ATF4, CHOP, BCL-2, and BAX mRNA expression with BAX/BCL-2 Ratio in two groups of KGN cells (n = 3 each group); (D, E) ATF4, CHOP, BCL2, BAX, Cleaved-Caspase3 protein levels in two groups of KGN cells (n = 6 each group); (F, G) Alterations in ATF4, CHOP, BCL-2, BAX, and Cleaved Caspase-3 protein levels in KGN Cells following transfection with si-NC versus si-ATF4 (n = 6 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    3.4. SIRT1-mediated deacetylation of H3K9ac regulates Atf4 transcription and contributes to fluoride-induced ovarian dysfunction in female offspring

    Epigenetic modifications have emerged as critical mediators linking adverse pregnancy exposures to long-term offspring health outcomes. To explore whether the upregulated Atf4 expression in the ovaries of NaF exposed female offspring is epigenetically regulated, we first used the Cistrome Data Browser to predict potential histone H3 lysine 9 acetylation (H3K9ac) binding sites within the Atf4 promoter region (Fig. 4A). Based on these predictions, we designed specific primers for chromatin immunoprecipitation (ChIP) assays targeting the candidate region (Fig. 4B).

    Fig. 4

    1. Download: Download high-res image (844KB)
    2. Download: Download full-size image

    Fig. 4. SIRT1-mediated deacetylation contributes to ovarian dysfunction in fluoride-exposed female offspring via ATF4 promoter acetylation. (A) Prediction of potential H3K9ac modification sites in the ATF4 gene promoter region using the Cistrome Data Browser database; (B) Schematic representation of the ATF4 gene promoter region indicating positions of qRT-PCR primer pairs (p1-p6); (C) Levels of SIRT1 protein and H3K9ac histone in CON and NaF groups (n = 6 each group); (D) Levels of SIRT1 protein and H3K9ac histone in two groups of KGN cells (n = 6 each group); (E) Protein expression levels of SIRT1, ATF4, CHOP, BCL2, BAX, and Cleaved-Caspase3 after transfecting KGN cells with si-ATF4, si-SIRT1, co-transfection of both, or controls (n = 4 each group); (F) ChIP-qPCR detection of H3K9ac enrichment at Atf4 promoter regions in ovarian tissues: CON vs. NaF groups (n = 3 each group). (G) ChIP-qPCR detection of H3K9ac enrichment at ATF4 promoter regions in KGN cells in two groups (n = 3 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    SIRT1, a class IIIhistone deacetylase, exerts its regulatory effect on gene transcription primarily through deacetylating H3K9 on histone H3. To clarify the association between SIRT1-mediated H3K9 acetylation and elevated Atf4 expression in NaF-exposed ovarian tissues, we detected the protein levels of SIRT1 and H3K9ac in ovarian tissues from both control and NaF-exposed offspring. The results showed that compared with the control group, NaF-exposed offspring exhibited a significant decrease in SIRT1 protein expression and a marked increase in H3K9ac levels (Fig. 4C). Furthermore, ChIP-qPCR analysis confirmed that H3K9ac enrichment at the Atf4 promoter region was significantly higher in the NaF-exposed group than in the control group (Fig. 4F).

    Collectively, these in vivo findings suggest that NaF exposure downregulates SIRT1, which in turn enhances H3K9ac modification at the Atf4 promoter region. This epigenetic change promotes Atf4 transcriptional activation and subsequent overexpression, ultimately contributing to ovarian dysfunction in female offspring.

    3.5. SIRT1-mediated deacetylation of H3K9ac regulates ATF4 transcription and modulates apoptosis in KGN cells

    To further validate the SIRT1-H3K9ac-ATF4 regulatory axis and explore its functional role, we performed in vitro functional experiments using KGN cells. Consistent with the in vivo results, NaF treatment of KGN cells led to a similar expression pattern of SIRT1 and H3K9ac: a significant reduction in SIRT1 protein levels accompanied by a notable increase in H3K9ac (Fig. 4D), confirming the conservation of this regulatory response between in vivo and in vitro models.

    To demonstrate that SIRT1, functioning as a deacetylase, influences acetylation at the H3K9 site, we introduced si-SIRT1 and confirmed its silencing efficiency (Fig. S1G). Transfection of si-SIRT1 in KGN cells resulted in elevated H3K9ac levels.

    To dissect the hierarchical regulatory relationship between SIRT1 and ATF4, we conducted dual knockdown experiments using si-SIRT1 and si-ATF4. SIRT1 silencing alone resulted in increased ATF4 expression, which further activated the downstream apoptotic pathway. In contrast, ATF4 knockdown did not affect SIRT1 expression (Fig. 4E), demonstrating that SIRT1 acts upstream of ATF4 in this regulatory pathway. Additionally, ChIP assays in KGN cells showed that NaF treatment significantly enhanced the binding of H3K9ac to the ATF4 promoter region (Fig. 4G), consistent with the in vivo ChIP-qPCR results.

    Based on the findings above, we propose the molecular mechanism underlying NaF-induced ovarian cell apoptosis: NaF exposure downregulates SIRT1 expression, which reduces H3K9ac deacetylation and enhances H3K9ac enrichment at the ATF4 promoter region. This epigenetic modification promotes ATF4 transcription and overexpression, leading to increased CHOP levels and subsequent activation of the apoptotic pathway in ovarian cells.

    3.6. Resveratrol rescues fluoride-induced ovarian dysfunction in offspring by upregulating SIRT1

    Resveratrol, an established SIRT1 agonist with ovary-protective properties, was administered in our model to further investigate the mechanism of SIRT1 in NaF exposure-induced ovarian dysfunction in offspring and its role in modulating acetylation. In the NaF group, ovaries showed a significant reduction in volume and a decrease in the ovarian index (Fig. 5A, B). Histological analysis revealed improved ovarian morphology in resveratrol co-treated mice, featuring increased ovarian volume, elevated primordial follicle counts, enhanced antral follicles, and reduced atretic follicles (Fig. 5C,D). Vaginal cytology characterized murine estrous cycles, with recorded cycle lengths used to calculate means and generate line graphs. Control group maintained regular 5-day cycles, whereas NaF exposure prolonged cycles to 8 days, primarily due to extended metestrus and proestrus phases. Resveratrol co-treatment normalized cycle length to 6 days and shortened diestrus (Fig. 5E). TUNEL staining showed that the proportion of apoptotic cells in the ovary decreased significantly after the Resveratrol treatment compared with the NaF alone group (Fig. 5F, G). Subsequently, total protein was extracted from mouse ovarian tissues. We observed that in the NaF group with resveratrol co-treatment, the protein level of SIRT1 increased, with a concomitant decline in H3K9ac levels (Fig. 5H, I). Following the reduction in acetylation levels, concomitant decreases were observed in ATF4 expression and its downstream effector CHOP in the resveratrol-treated group. This was accompanied by consequential alterations in apoptosis-related proteins (BCL-2, BAX, cleaved Caspase-3/Caspase-3), ultimately leading to diminished activation of the apoptotic pathway (Fig. 5J,K). Collectively, these findings demonstrate that resveratrol ameliorates fluoride-induced ovarian dysfunction in offspring via SIRT1 upregulation.

    Fig. 5

    1. Download: Download high-res image (931KB)
    2. Download: Download full-size image

    Fig. 5. Resveratrol attenuates fluoride-induced ovarian dysfunction in offspring via SIRT1 upregulation. (A) Ovarian appearance dimensions in control, NaF, and Resveratrol groups; (B) Ovarian index in 10-week-old mice in three groups (n = 6 each group); (C) HE staining of ovary sections from three groups of mice; (D) Proportion of follicles of all stages in the ovaries of three groups (n = 6 each group); (E) Comparison of the estrus cycle in three groups, (n = 6 each group); (F, G) Tunel staining of ovary sections and tunel positivity level statistical graph of three groups of mice (n = 3 each group); (H, I) Protein expression levels of SIRT1 and histone H3K9ac in offspring ovaries from control, NaF-treated, and NaF+resveratrol co-treated groups (n = 6 each group); (J, K).

    Protein expression levels of ATF4, CHOP, BCL2, BAX and Cleaved Caspase-3 in offspring ovaries from control, NaF-treated, and NaF+resveratrol co-treated groups (n = 6 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    3.7. Resveratrol supplementation rescues fluoride-induced apoptosis in KGN cells

    Validation experiments in KGN cells were performed to further examine resveratrol’s targeting effects on SIRT1 and its subsequent impact on the ATF4-CHOP axis. Resveratrol was used at non-cytotoxic concentration of 25 uM with dose calculated using CCK-8 viability and did not have any adverse effect on cell viability, and the dose used was kept below the IC50 value and consistent with the established literature protocol (Zhang et al., 2023). The initial evaluation of resveratrol’s agonist action on SIRT1 demonstrated that a 48-hour treatment with 25 uM resveratrol in KGN cells dramatically increased both SIRT1 mRNA expression and protein levels (Fig. S1I, J), thereby validating its target-specific activation. Flow cytometry analysis demonstrated that resveratrol markedly reduced apoptosis in NaF-treated KGN cells (Fig. 6B), while simultaneously mirroring in vivo findings via SIRT1-mediated H3K9 deacetylation, which inhibited ATF4/CHOP expression (Fig. 6C), ultimately leading to a decrease in downstream apoptotic pathway activation (Fig. 6D).

    Fig. 6

    1. Download: Download high-res image (794KB)
    2. Download: Download full-size image

    Fig. 6. Resveratrol supplementation rescues fluoride-induced apoptosis in murine ovary and KGN cells. (A) Determination of cell viability and IC50 values by CCK-8 assay in KGN cells exposed to graded concentrations of Resveratrol (0–100 uM, 24 h, n = 9 each group); (B) Flow cytometric analysis revealed differential apoptosis rates in control, NaF-treated, and NaF+resveratrol co-treated groups of KGN cells(n = 3 each group); (C) Levels of SIRT1 protein and H3K9ac histone in three groups of KGN cells (n = 6 each group); (D) ATF4, CHOP, BCL2, BAX, Cleaved-Caspase3 protein levels in three groups of KGN cells (n = 6 each group). Data are mean±SEM; ns, no significance, p > 0.05; *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001.

    4. Discussion

    The world health problem of infertility against the backdrop of the increasing environmental degradation is a binding health issue that needs some urgent intervention. The prevalence of infertility is about 17.5 % among couples all over the world and has gained a serious concern over the past years (Han et al., 2024). One of the key causes in the decreasing fertility is ovarian dysfunction. Depletion of ovarian reserve which is one of the key symptoms of ovarian dysfunction, is mainly determined during prenatal stage. By incorporating the FOAD hypothesis, the DOHaD paradigm (Barker, 1990), and the existing evidence, it comes to be known that exposure of the developing fetus to adverse stimuli may impair the development of ovarian reserve in the offspring, which in the future results in reduced ovarian activity. Currently documented prenatal exposures compromising offspring ovarian function encompass: bisphenol A (BPA), perfluorooctanoic acid (PFOA), polycyclic aromatic hydrocarbons (PAHs), cannabinoids, di(2-ethylhexyl) phthalate (DEHP), caffeine, zearalenone, dexamethasone, among others (Castel et al., 2020, Gong et al., 2023, Jurisicova et al., 2007, Kong et al., 2021; Liu et al., 2021; Yao et al., 2022; Zhao et al., 2024).

    Fluoride, a ubiquitous environmental compound, is primarily ingested by humans through drinking water. Studies indicate that fluoride ion concentrations in drinking water from high-fluoride areas range from 0.01 to 48 mg/L (Zhao et al., 2020). Fluoride has been proven as a reproductive toxicant in several studies, which has negative impacts on the structural integrity and functional competency of the male and female reproductive systems. Nonetheless, it has comparatively limited research on the intergenerational impacts of it on the reproductive capacity of offspring. We have created a murine model of a prenatal exposure to fluoride in pregnant mice, as well as, we were the first to note that a prenatal exposure to fluoride created offspring with both low ovarian reserve and function. Considering the environmental fluoride content in drinking water and the fact that rodents excrete fluoride more efficiently than humans. In previous studies by Zhao et al. (2020) and Mo et al. (2023), it was confirmed that mice models consuming drinking water containing 25, 50, 100, and 150 mg/L of NaF corresponded to environmental fluoride ion levels of 11.3, 22.6, 45.2, and 67.8 mg/L, respectively. In the mouse pregnancy exposure modeling, we chose the dose of 100 mg/L sodium fluoride, which integrated the framework proposed by Y. Hu et al. (2024) and the experimental protocol verified by our laboratory. It was done because this concentration was selected to reflect the human exposure patterns and to consider interspecies variations in the pharmacokinetics and clearance of fluoride (Cai et al., 2025; Li et al., 2022).

    The observations in the fluoride exposed offspring were a decrease in both the ovarian volume, a reduced ovary to body weight index, and a break in the ovular estrous cyclicity. The disruption of the estrous cycle largely took the form of a long proestrus and diestrus cycle which is the evidence of a low ovarian reserve and the evidence of a rapid aging of the reproductive system. NaF group showed deterioration of ovarian functionality as further indicated by the change in the levels of sex hormones, namely, increased FSH and reduced E 2 and AMH, which denoted the lack of follicular development and loss of follicular pool. In a mechanistic manner, follicular depletion of quantity or quality results in a reduction in ovarian hormone secretion which in turn induces a significant increase in FSH by negative feedback (Jiao et al., 2021). The levels of AMH has a close relationship with antral follicle count (AFC) and primordial follicle pool size, which is a biomarker of primordial follicle reserve. In contrast to the diminishing AMH values that suggest exhaustion of primordial follicle pool, decreased E2 level demonstrate the dysfunction of the granulosa cells as well as the insufficiency of the maturation of the dominant follicles (Li et al., 2018). These modifications of serum hormone patterns proved to be exact concurrences of ovarian histomorphological modifications. It is important to mention that the NaF group had lower primordial and antral follicles with structural devastation of the granulosa cell layer. At the same time, the percentage of follicular atresia was much higher in this group, and more cells were TUNEL-positive, which all proved a high level of cell apoptosis and the overall deterioration of ovarian functions.

    The mechanisms responsible for ovarian dysfunction in fluoride-exposed offspring are intricate and multifaceted. The most significant factor is the dysregulation of the apoptotic cascade, marked by expedited follicular atresia and compromised oocyte quality due to granulosa cell apoptosis (Stringer et al., 2023). This was initially confirmed by histopathology and TUNEL staining of ovarian sections,laying the foundation for subsequent exploration of molecular mechanisms.

    The apoptosis of granulosa cells is the main process of the follicular atresia. Although the entire apoptotic machinery is not fully characterized, research indicates that the endoplasmic reticulum stress pathway may play a crucial role in the regulation of early follicular degeneration and granulosa cell death, as demonstrated in various models. In addition, sequencing of RNA under the condition of stress due to intracellular fluoride ingestion in ovine granulosa cells showed a significant increase in ATF4, which revealed the importance of the ATF4-mediated endoplasmic reticulum stress response (Ma et al., 2024). There are consolidated insights on the execution of the ATF4-CHOP axis in the ER stress pathway showing that this transcriptional re-programming of the apoptotic execution process by increasing the pro-apoptotic protein BAX and reducing the anti-apoptotic protein BCL-2 initiates the mitochondrial cascade of apoptosis (Cao et al., 2022). Complementing these findings, our experiments showed a substantial upregulation of ATF4 at both the transcriptional and translational levels. This upregulation was observed both in ovaries of F1 offspring after gestational fluoride exposure in vivo and in KGN cell cultures exposed to NaF in vitro. At the same time as ATF4 induction, there was a cascade response of ER stress-apoptotic activity, with CHOP (a direct ATF4 target) highly increased, and the BAX/BCL2 rate bursting, accompanied by an accumulation of Cleaved Caspase-3, all indicated the start of the apoptotic cascade. Silencing of ATF4 via siRNA significantly inhibited this ER stressapoptotic axis, which appeared as inhibited transactivation of CHOP, and recovered BAX/BCL2 apoptotic rheostat, and prevented Caspase-3 proteolytic activation in KGN cells. This consolidation confirms ATF4 as a crucial mediator of fluoride-induced intergenerational ovotoxicity, mediating maternal toxicant exposure into germline variations in ER stress commitment and consequent loss of follicles.

    SIRT1 is a NAD+ -dependent lysine deacetylase involved in the deacetylation of histone and non-histone proteins, coordinating a wide range of aging-related events such as the epigenetic modification, cell cycle, transcriptional regulatory, apoptotic signaling, and stress-resistance-related energy metabolism. Therefore, SIRT1 has been extensively researched for its possible significance in aging-related disorders and enhancing overall health (Narita, Weinert, and Choudhary, 2019). Recent studies have reinforced the ovarian protective effect of SIRT1, demonstrating its ability to mitigate oocyte aging by diminishing ROS buildup, correcting spindle apparatus abnormalities and mitochondrial dysfunction, and influencing epigenetic modifications. In our study, reduced SIRT1 expression was detected in ovarian tissues of fluoride exposed offspring and KGN cells treated with NaF, suggesting its potential protective role in ovarian function may be impaired. Notably, SIRT1 knockdown triggered ATF4 upregulation, which in turn activated the downstream CHOP protein, ultimately leading to the activation of the cell apoptosis pathway. However, ATF4 silencing failed to change the expression level of SIRT1, which could be an indication of the upstream regulatory effect of SIRT1 on ATF4. To further strengthen the regulatory role of SIRT on ATF4 we identified the presence of the binding site on the region of ATF4 promoter on the Cistrome DB site containing H3K9ac, which is a frequent acetylation regulatory site of SIRT1. ChIP supported occupancy at H3K9ac-greater chromatin loci upon ATF4. Improved ATF4 binding to H3K9ac-marked regions in NaF-exposed tissues in the ovary and in KGN cells was accompanied with hyperacetylation of H3K9ac. This epigenetic alteration increased the chromatin accessibility of ATF4 leading to transcriptional upregulation. Importantly, the knockdown of SIRT1 diminished deacetylation, resulting in increased H3K9ac levels and, subsequently, elevated ATF4 expression. This finding substantiates SIRT1’s function as a histone deacetylase regulating the ATF4 regulatory axis. Current literature suggests that SIRT1 maintains ovarian function mainly by regulating oxidative stress. However, our experimental results demonstrate a new mechanistic role for SIRT1 as a deacetylase in coordinating specific epigenetic mechanisms that regulate pro-apoptotic gene expression (Xing et al., 2021).

    Resveratrol, as a specific SIRT1 activator, increased both SIRT1 expression and its deacetylation activity in granulosa cells (Chen et al., 2022, Ortega and Duleba, 2015). In our study, resveratrol was utilized in rescue experiments employing both animal and cellular models to confirm the essential function of the SIRT1-ATF4 axis in fluoride-induced ovarian dysfunction. Our study found out that ovarian functioning of fluoride-treated offspring was restored by resveratrol intervention through the process of restoring normal estrous cycle, normal ovarian morphology, and decrease in apoptosis. Equally, with KGN cell models, the effect of resveratrol co-exposure was high in neutralising apoptosis of NaF-treated cells. We also validated the suppression role of the apoptotic pathway by reducing the expression of ATF4 and CHOP after the intervention of resveratrol. Besides, we ensured that resveratrol raised the level of deacetylation of H3K9ac on ATF4 promoter, which in turn suppressed the expression and transcription of ATF4. As has been reported in many studies, resveratrol is an effective agent that helps ovarian reserve and delay premature ovarian failure (Chen et al., 2022, Hu et al., 2024; Hu et al., 2024; Ortega and Duleba, 2015), and our study supports the mechanism. The obtained results can form a basis of a prospective clinical trial on resveratrol as a possible drug in the treatment of the dysfunctional ovary and premature ovarian insufficiency (POI).

    5. Limitations and prospects

    Although this study reports important findings regarding the impairment of ovarian function in female offspring caused by fluoride exposure during pregnancy via the SIRT1-ATF4 axis, it still has certain limitations. First, regarding the exposure model, although we verified the success of the model via maternal mouse characteristics based on previous successful modeling experience, we did not directly quantify fluoride concentrations in the serum or ovarian tissues of the F1 offspring. Instead, we relied on the well-established pharmacokinetic profile of fluoride, which is documented to effectively cross the placental barrier and accumulate in fetal tissues (Castiblanco-Rubio and Martinez-Mier, 2022, Dec et al., 2019). Consequently, our in vitro experiments were designed to mechanistically dissect the direct impact of fluoride on granulosa cells, yielding molecular signatures consistent with our in vivo observations. Future investigations incorporating precise tissue dosimetry would further strengthen the quantitative link between maternal intake and fetal ovarian burden. Second, regarding the mechanistic verification, while we demonstrated that activating SIRT1 with Resveratrol effectively reversed ATF4 upregulation and ameliorated ovarian damage, we did not employ genetic approaches (such as siRNA) to directly ablate ATF4 under fluoride exposure conditions. Although our upstream pharmacological rescue and ChIP assays strongly implicate the SIRT1-ATF4 axis, the inclusion of a specific ATF4 knockdown group would provide definitive evidence confirming the necessity of ATF4 as a downstream mediator of apoptosis. In future studies, we will utilize granulosa cell-specific gene editing tools to fully delineate this hierarchical signaling machinery.

    6. Conclusion

    Combined in vivo and in vitro evidence in this research reveals that prenatal exposure to fluoride causes ovarian dysfunction in offspring. Exposure to fluoride suppressed SIRT1 gene and activity and caused hyperacetylation of histones (H3K9ac) at the promoter of the expression element: ATF4. This epigenetic modification amplified the expression of ATF4 that then activated CHOP expression. The high CHOP changed the BAX/BCL2 ratio to cause apoptosis of the granulosa cell. Thus, follicular atresia increased, which eventually resulted in ovarian dysfunction. SIRT1 emerges as a promising therapeutic target, as supplementation with its activator resveratrol reduced H3K9ac acetylation, downregulated the ATF4-CHOP, attenuated apoptosis and follicular attrition, and rescued the diminished ovarian reserve phenotype. These findings elucidate a novel pathogenic cascade, the SIRT1-ATF4-CHOP apoptotic signaling axis, underlying fluoride-induced intergenerational ovarian dysfunction.

    CRediT authorship contribution statement

    Bin Wei: Formal analysis. Ming Huo: Software. Yan Zhao: Data curation. Zejun Zhao: Software. Li Zeng: Data curation. Xi Yu: Software. Yueyang Song: Investigation. Miao Sun: Project administration, Funding acquisition, Conceptualization. Youguo Chen: Supervision, Conceptualization. Yajun Shi: Writing – review & editing. Yannan Gu: Writing – original draft, Methodology, Investigation.

    Ethics approval and consent to participate

    All animal experiments were carried out in accordance with the Laboratory Animal Management Regulations with approval of the Research Ethics Committee of the First Afliated Hospital of Soochow University.

    Funding

    This research was funded by the National Key R&D Program of China (2022YFC2703700 and 2019YFA0802600), Nantong Municipal Science and Technology Bureau Social Welfare Project (MS2025027) and the Nantong University Special Research Fund for Clinical Medicine (2025LY043).

    Declaration of Competing Interest

    The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

    Acknowledgments

    The authors appreciatively acknowledge the National Key R&D Program of China, Nantong Municipal Science and Technology Bureau and Nantong University Special Research Fund for their support. We acknowledge Institut for Fetology, The First Affliated Hospital of Soochow University for their support.

    Appendix A. Supplementary material

    What’s this?

    Data availability

    Data will be made available on request.

    References

    Cited by (0)

    1
    Yannan Gu and Yajun Shi contributed equally to this work.