Stage-, dose- and course-dependent disruption of mice fetal ovarian development following prenatal dexamethasone exposure Authors: Lu Chen, Shan Jin, Zhaojun Wang, Abudukade Subinu, Fan Mo, Jing Huang, Hui Wang Published: 2026 ABSTRACT AbstractBackgroundDexamethasone, a synthetic glucocorticoid, is widely used during pregnancy to prevent preterm infant respiratory distress syndrome. While our prior work established that multiple courses of prenatal dexamethasone exposure (PDE) impair offspring ovarian function with transgenerational effects, the specific impacts of exposure stages, doses, and courses remain poorly defined.Materials and methodsPregnant Kunming mice were treated with dexamethasone during mid–late gestation (GD14–15) or late gestation (GD16–17) with low/moderate/high doses, either as a single course or as two consecutive courses. Fetal ovaries were collected at GD18 to assess morphology, cell proliferation/apoptosis, steroidogenic markers, and oocyte-development genes; network toxicology was used to infer candidate pathways.ResultsPDE disrupted fetal ovarian morphogenesis, characterized by premature germ-cell nest breakdown, reduced proliferation and increased apoptosis. PDE at different stages, doses and courses can increase ovarian steroid synthesis (SF1\StAR\CYP11A1\Hsd3b\CYP19A1) while suppressing oocyte and folliculogenesis genes (NOBOX\Figlα\Sohlh2\Bmp15\Gdf9). These effects are stage-, dose- and course-dependent, with the most obvious toxicity resulting from exposure in the mid–late stages of pregnancy, at high doses and in a single course. Network toxicology and qPCR revealed the NR3C1-MAPK14-IL6 axis: IL6 was positively correlated with proliferation and oocyte development markers, while NR3C1/MAPK14 was associated with apoptosis and steroidogenesis.ConclusionsThis study provides experimental evidence for guiding clinical “precision medicine” and effectively assessing the risk of fetal ovarian developmental toxicity.Keywords: Prenatal dexamethasone exposure, Ovarian developmental toxicity, Pregranulosa cells, Oocyte development, Medication safety Introduction The theory of Developmental Origins of Health and Disease (DOHaD) points out that environmental disturbances during the critical period of embryonic development may lead to abnormal organ development and long-term dysfunction [1]. This theory has been extended from the initial focus on metabolic diseases to the field of reproductive system development, in which the ovary is extremely sensitive to microenvironmental changes [2]. Ovary is the core organ of the female reproductive system, which has two key physiological functions, reproductive and endocrine. Ovary is not only a determinant of female fertility, but also a key regulatory organ for general health. A large number of studies have shown that adverse environmental exposures during pregnancy can affect the ovarian homeostasis of female offspring. For example, caffeine exposure during pregnancy can lead to ovarian dysfunction in female offspring [3]. Ethanol exposure during pregnancy increases the risk of premature ovarian failure in female adult offspring [4]. In addition, exposure to the environmental endocrine disruptant bisphenol A during pregnancy can also increase the proportion of ovarian cell apoptosis in female offspring [5]. Therefore, it is crucial to focus on the developmental programming of the ovary during the intrauterine period. As a new type of environmental pollutant, drugs play an irreplaceable role in diagnosis, treatment and protection of public health. However, medication during pregnancy has a “double-edged sword” effect [6]. Under the framework of DOHaD theory, the effects of drug exposure during pregnancy on the ovarian health of offspring have attracted increasing attention. For example, exposure to azithromycin during pregnancy can cause abnormal ovarian morphology and function in female offspring [7]. Moreover, some studies have suggested that acetaminophen exposure during pregnancy can reduce the number of oocytes in fetal ovary and inhibit cell proliferation [8]. Amoxicillin exposure during pregnancy can also lead to hyper steroidogenesis in ovarian pregranulosa cells by activating IGF1 [9]. Therefore, it is a key research topic related to human reproductive health to pay attention to the effect of drug exposure during the intrauterine period. As a synthetic glucocorticoid drug, dexamethasone has anti-inflammatory, immunosuppressive and metabolic effects. The use of dexamethasone during pregnancy can enter the fetal blood circulation system through the placenta and promote fetal lung maturation [10], thereby preventing the occurrence of respiratory distress syndrome after birth and reducing neonatal mortality [11, 12]. However, some studies have found that prenatal dexamethasone treatment can cause low birth weight, especially under multiple courses [13]. In terms of the female reproductive system, previous studies in our laboratory have found that multiple courses of dexamethasone use during pregnancy can lead to abnormal ovarian function development in offspring, and have multi-generation genetic effects [14, 15]. However, there is a lack of a systematic evaluation system for the effect of maternal dexamethasone exposure during pregnancy on ovarian damage in offspring. With the full implementation of “precision medicine”, the safety research of medication during pregnancy has also been deepened and refined [16]. The current standard prenatal dexamethasone regimen recommended by international guidelines is 6 mg intramuscular injection four times every 12 h (2017 guidelines). However, in the actual application process, maternal physiology and fetal development may produce different responses under this medication standard. A large number of animal experiments have shown that dexamethasone exposure during pregnancy has a quantitative-effect relationship on offspring development [17, 18]. In addition, the safe gestational age of dexamethasone use during pregnancy also varies in clinical practice [19]. These differences highlight the need to evaluate the efficacy and risks of medications used at different stages of pregnancy, which is a core issue in precision medicine research. Similarly, multiple dosing regimens are equally controversial. When a pregnant woman does not deliver within one week after initial treatment but still maintains the risk of preterm birth, an additional course of treatment is often recommended to improve neonatal survival. However, there is still a large gap in the research to analyze the ovarian damage of female offspring exposed to dexamethasone at different times, doses and courses during pregnancy. In summary, in this study, mice exposed to dexamethasone during pregnancy at different stages (gestational day (GD)14–15 or GD16-17), doses (0.2, 0.4 or 0.8 mg/kg, bid) and courses (single or double courses) were established according to the clinical guidelines for the use of dexamethasone. Fetal mice ovaries of all groups were collected at GD18 for further studies. To explore the potential adverse effects on fetal ovarian development through multiple dimensions such as ovarian morphology, cell proliferation/apoptosis, follicular development, and hormone synthesis function, and to explore the possible mechanisms. The results of this study will provide experimental evidence for guiding clinical rational drug use and effectively assessing the risk of developmental toxicity of fetal ovaries, so as to promote the application of precision medicine in guiding rational drug use during pregnancy. Materials and methods Chemicals Dexamethasone sodium phosphate injection was sourced from Huangzhong Pharmaceutical Co., Ltd. (Wuhan, China; No. H42021492). Anesthesia was maintained with isoflurane, obtained from RWD Life Science Co., Ltd. (Shenzhen, China; Cat. No. R510–22) and Baxter Healthcare (USA). For RNA extraction and quantification, we used TRIzol™ Reagent (Thermo Fisher Scientific, USA; No. 15596026) and commercial RT-qPCR kits (Vazyme Biotech, Nanjing, China). Antibodies targeting PCNA (Cat. No. A0264), Ki67 (Cat. No. A20018), Caspase-3 (Cat. No. A2156), and StAR (Cat. No. A22166) were supplied by ABclonal Technology Co., Ltd. (Wuhan, China). NOBOX (Cat. No. sc-390016) were supplied by Santa Cruz Biotechnology Co., Ltd. (Wuhan, China). All remaining reagents met analytical grade specifications. Animals and treatment The study protocol received ethical approval from the Animal Experiment Ethics Committee of Wuhan University Medical College (Approval No. ZN2021221) and complied with the regulations set by the Chinese Animal Welfare Committee. Animal experiments were conducted at the Center for Animal Experiments/Animal Biosafety Level–III Laboratory of Wuhan University (Wuhan, China), a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC International). Specific pathogen-free (SPF) Kunming mice were used for the experimental study [Certification No: 110324210106770357]. Purchased from SPF (Beijing) Biotechnology Co., Ltd. (Beijing, China). After one week of adaptive feeding, the treatment was performed according to the schematic diagram (Fig. 1). Specific experimental procedures and doses were referred to previous studies [20]. Fetal mice ovaries of all groups were collected at GD18 for further studies. Specific fetal ovarian collection methods were referred to previous studies [7]. To minimize the suffering of the animals, mice were euthanized by 2.5% isoflurane overdose followed by cervical dislocation prior to sample harvest (anesthetized and unconscious). The schematic illustration of animal treatment and experiment grouping. GD, gestational day; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Hematoxylin-eosin (HE) staining Fetal ovaries were fixed overnight in 4% paraformaldehyde and processed for standard HE staining. For each experimental group, paraffin-embedded ovarian sections were prepared, aiming to obtain the largest cross-sectional profile for analysis. All sections were examined and imaged under an Olympus AH-2 light microscope [21]. To ensure objectivity, a second researcher, blinded to the group assignments, performed all quantitative assessments. For each sample, systematic microphotography captured key morphological parameters, including the maximum cross-sectional area and diameter of the ovary. Furthermore, for oocyte classification and counting, three random interstitial fields per slide were selected and analyzed. Total RNA extraction, reverse transcription and RT-qPCR for ovary Ovarian tissues from individual fetal mice within a litter were homogenized in TRIzol™ reagent using zirconium beads. Total RNA was then extracted with chloroform and isopropanol according to the standard protocol. RNA concentration and purity were measured on a NanoDrop 2000 spectrophotometer, and samples were normalized to 1 µg/µL. One microgram of total RNA was reverse-transcribed into cDNA using a commercial kit, and the product was stored at -20 °C. All primers were designed with Primer Premier 5.0 (Premier Biosoft International, CA), and their specificity was confirmed by NCBI BLAST analysis (see Table 1). RT-qPCR was performed on a QuantStudio®5 system (Thermo Fisher Scientific) using ChamQ Universal SYBR qPCR Master Mix. Gene expression levels were quantified relative to the housekeeping gene Gapdh, which was amplified concurrently in each run [22]. The 2−ΔΔCt method was applied for data normalization. Table 1Oligonucleotide primers of mice in real-time quantitative PCRGenesForward primerReverse primer GAPDH GCAAGTTCAATGGCACAGGCCAGTAGACTCCACGACA PCNA CACTCCACTGTCTCCTACAGCCTAAGATGCTTCCTCATC Ki67 CTGGTCTCAAAGGACCAATCCTCTTCATCTGCCTCTACTTTC Caspase3 GATATTGTTAGCGGTTCCTGTGAACCAGGCTGTCGTCTAA SF1 CCAGTACGGCAAGGAAGAGAGGCTGAAGAGGATGAGGA StAR GGGAGATGCCTGAGCAAAGCGCTGGCGAACTCTATCTGGGT CYP11A1 CACTCCTCAAAGCCAGCATCAAGATTACGAAGCACCAGGTCATTCAC Hsd3b AGGTGACAGTGTTGGAAGGAGACTGACATCAATGACAGCAGCAGTG CYP17A1 ACTATCCGAGAAGTGCTGCGTATGCTCCGAAGGGCAAGTAACTC CYP19 ATGGGCCTCCTTCTCCTGATCAGGCACTTCCAATCCCCAT NOBOX GACATGGGACCTCAGGATTAGAGTCTTCTGGTGGTAGAAATG Figlα AGAGCGTGAGCGGATAAACCAGAACACAGCCAAGTATC Bmp15 GTGCTCAGGCTAAACTTCTTGGAGGGAACACTGGTTATTT Gdf9 GATGTGACCTCCCTCCTTCAGCCTGGGTACTCGTGTCATT Sohlh2 TCTCAGCCACATCACAGAGGGGGGACGCGAGTCTTATACA IL6 GCCTTCTTGGGACTGATGCTAGCCTCCGACTTGTGAAGTG NR3C1 TGAAGCTTCGGGATGCCATTGCTGTCCTTCCACTGCTCTT MAPK14 TCACGCCAAAAGGACCTACCATTCCTCCAGTGACCTTGCG Oligonucleotide primers of mice in real-time quantitative PCR Immunofluorescence staining (IF) and quantitative detection Paraffin-embedded ovarian tissues were sectioned at 5 μm thickness. Sections underwent deparaffinization in xylene and rehydration, then were subjected to antigen retrieval in 10 mM citrate buffer (pH 6.0) at 65 °C for 5 h. After washing three times with PBS, non-specific sites were blocked with 5% BSA for 30 min at room temperature. The sections were then probed with anti-PCNA, anti-Caspase3, anti-StAR and anti-NOBOX antibodies (1:100 each) overnight at 4 °C, followed by a 2-h incubation with a fluorescent secondary antibody (1:200) at room temperature. Nuclei were visualized with DAPI (1 µg/mL, 10 min) [23]. Image acquisition used an Olympus AH-2 light microscope, with subsequent analysis performed using the accompanying Olympus software. Network toxicology analysis The SMILES sequence of dexamethasone was obtained from the PubChem database. To identify dexamethasone -related target genes, data were retrieved from the Swiss Target Prediction and chEMBL, with duplicate entries removed. For precocious puberty and diminished ovarian reserve - related target genes, the GeneCards and NCBI databases were queried using “precocious puberty” and “diminished ovarian reserve” as the keyword, and duplicate genes were eliminated. Statistical analysis Statistical analysis was conducted using Image J (National Institutes of Health, Bethesda, USA) and Prism 8.0 (GraphPad Software, La Jolla, CA, USA). Quantitative results are presented as the mean ± standard error of the mean (SEM). Comparisons between the control and treatment groups were made using a two-tailed Student’s t-test. For experiments involving more than two groups, data were analyzed by one-way analysis of variance (ANOVA) followed by Tukey’s post-hoc test. The Pearson correlation coefficient (r) was calculated with Prism 8.0 to assess the linear relationship between two variables. The coefficient ranges from − 1 (perfect negative correlation) to + 1 (perfect positive correlation), with 0 indicating no correlation. Correlation strength was interpreted according to established guidelines. P < 0.05 was considered statistically significant. Results Effects of PDE at different stages, doses and courses on fetal ovarian morphology To observe the effect of different periods, doses and courses of PDE on the morphology of fetal ovaries, we used HE staining to observe the maximum diameter/body weight, cross-sectional area/body and the number of oocytes per unit area of fetal ovaries to evaluate the growth and development status of fetal ovaries (Fig. 2a). After careful observation of its morphology and structure, it was found that the number of empty follicles in fetal ovarian tissue increased to varying degrees, and the cystic structure of primordial germ cells was blurred under different stages, doses and courses of exposure. Compared with the control group, late gestation PDE (PDH) significantly increased the maximum ovarian diameter/body weight ratio and decreased the number of oocytes per unit area, while mid-late gestation PDE (PDML) had no significant changes. Low-dose PDE (PDL) had no significant changes, while medium-dose PDE (PDM) significantly increased the maximum diameter/body weight of fetal ovary, while high-dose PDE (PDH) further increased the maximum diameter /body weight of fetal ovary, and the number of oocytes per unit area decreased only in the high-dose PDE group. Compared with their respective control groups, single course of PDE (PDH) and double courses of PDE (PDD) increased the maximum ovarian diameter/body weight and decreased the number of oocytes per unit area (Fig. 2b, c). In conclusion, PDE can cause abnormal fetal ovary morphology and decreased oocyte number in female offspring, and the damage is more obvious at the end of pregnancy, high dose and single course of administration. Effects of PDE at different stages, doses and courses on the fetal ovarian morphology. a Ovarian morphology (HE staining, n = 3); (b) Ratio of maximum diameter to body weight (µm/g); (c) Number of primordial follicles per unit area (N/104 µm2). Mean ± S.E.M., n = 3. *P < 0.05, **P < 0.01 vs. C. HE, hematoxylin-eosin; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Effects of PDE at different stages, doses and courses on fetal ovarian cell proliferation and apoptosis Cell proliferation and apoptosis are important indicators for assessing developmental toxicity. To determine whether PDE at different time, dose and duration can affect cell proliferation and apoptosis in fetal ovary, we detected the mRNA and protein expression of PCNA, Ki67 and Caspase3. The results suggested that, in terms of stage, we observed that mid-late PDE significantly inhibited ovarian cell proliferation (Fig. 3b), while late PDE significantly inhibited PCNA expression and significantly increased apoptosis compared with their respective controls (Fig. 3c). PDE significantly inhibited the expression of PCNA and Ki67 in a dose-dependent manner, and high dose PDE also significantly increased the expression of Caspase3 (Fig. 3d-f). In terms of treatment courses, single-course PDE inhibited PCNA and increased Caspase3 expression, while there were not significantly different in double-course PDE (Fig. 3g-i). In addition, the effect of PDE on the protein expression of PCNA and Caspase3 was also observed by immunofluorescence analysis of fetal ovaries, which was consistent with the mRNA expression. A single course of high-dose PDE in late gestation showed the most significant inhibition of PCNA expression and activation of Caspase3 (Fig. 3j-m). In conclusion, PDE can inhibit cell proliferation and promote cell apoptosis, and there is a certain dose-effect relationship, among which single treatment and high dose have the most significant effects. Double courses have no significant difference in transcriptional level but show damage in protein expression. Effects of PDE at different stages, doses and courses on proliferation and apoptosis in fetal ovary. a-c Relative mRNA level of PCNA, Ki67, and Caspase3 in stages effect, n = 8; (d-f) Relative mRNA level of PCNA, Ki67, and Caspase3 in doses effect, n = 8; (g-i) Relative mRNA level of PCNA, Ki67, and Caspase3 in courses effect, n = 8; (j-m) Representative immunofluorescence images and mean fluorescence intensity of PCNA and Caspase3, n = 3. Mean ± S.E.M. *P < 0.05, **P < 0.01 vs. C or CON, #P < 0.05, ##P < 0.01 vs. C2. PCNA, proliferating cell nuclear antigen; Ki67, antigen identified by monoclonal antibody Ki 67; GAPDH, glyceraldehyde phosphate dehydrogenase; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Effects of PDE at different stages, doses and courses on fetal ovarian estrogen synthesis function The ovary is the main site of estrogen synthesis, so we observed the effect of PDE on the expression of genes involved in fetal ovarian endocrine function. The results showed that the mRNA expression of steroidogenic enzymes in mid-late PDE and late PDE were increased, especially SF1/Star/CYP19A1, and the proportion of changes in late PDE was greater than that in mid-late PDE (Fig. 4a-f). Low-dose PDE and medium-dose PDE had no significant effect on steroid synthetase expression, while high-dose PDE showed a significant increase (SF1, Star, CPY11A1, Hsd3b, CYP19A1) (Fig. 4g-l). In terms of treatment, we observed that single course PDE and double courses of PDE both promoted steroid synthesis to varying degrees, but single course PDE was more effective in promoting the expression of SF1/Star/CPY11A1 (Fig. 4m-o). However, double courses PDE was more significant in promoting Hsd3b/CYP19A1 (Fig. 4p-r). At the same time, we further confirmed at the protein level by immunofluorescence, results showed that the expression of StAR was significantly increased in different stages, doses and courses of treatment, which was consistent with the results of mRNA detection (Fig. 4s, t). In conclusion, PDE promotes fetal ovarian steroidogenic function, most significantly in the late gestation with a single course of high-dose PDE. Effects of PDE at different stages, doses and courses on estradiol synthesis function in fetal ovary. a-f Relative mRNA level of SF1, Star, CYP11A1, Hsd3b, CYP17A1 and CYP19A1 in stages effect, n = 8; (g-l) Relative mRNA level of SF1, Star, CYP11A1, Hsd3b, CYP17A1 and CYP19A1 in doses effect, n = 8; (m-r) Relative mRNA level of SF1, Star, CYP11A1, Hsd3b, CYP17A1 and CYP19A1 in courses effect, n = 8; (s, t) Representative immunofluorescence images and mean fluorescence intensity of StAR, n = 3. Mean ± S.E.M. *P < 0.05, **P < 0.01 vs. C or CON, #P < 0.05, ##P < 0.01 vs. C2 or C3. SF1, splicing factor 1; Star, steroidogenic acute regulatory protein; CYP11A1, cytochrome P450, family 11, subfamily a, polypeptide 1; Hsd3b, hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase cluster; CYP17A1, cytochrome P450, family 17, subfamily a, polypeptide 1; CYP19A1, cytochrome P450, family 19, subfamily a, polypeptide 1; GAPDH, glyceraldehyde phosphate dehydrogenase; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Effects of PDE at different stages, doses and courses on fetal ovarian oocyte development The reproductive function of the ovary depends on the normal development of the oocyte, which plays a crucial role in promoting follicular development by expressing various regulatory factors involved in follicle assembly and maturation. To evaluate the effect of PDE on oocyte development, the expression changes of oogenesis related genes (NOBOX, Figla, Sohlh2, Bmp15 and Gdf9) in serial ovaries were examined. The results suggested that PDE only reduced the expression levels of NOBOX and Bmp15 in the mid-late pregnancy (Fig. 5a, c), while late PDE had a significant inhibitory effect on the above five genes (Fig. 5a-e). In terms of doses, different doses of PDE reduced the mRNA expression of NOBOX, Figla, Sohlh2 and Bmp15 in fetal ovaries, and there was a good dose-effect relationship (Fig. 5f, h-j), while low and medium doses of PDE increased the expression of Gdf9 to a certain extent (Fig. 5g). The expressions of NOBOX, Figla, Bmp15 and Gdf9 genes were significantly inhibited by both single course and double courses PDE (Fig. 5k-o), and Sohlh2 expression was also significantly inhibited by single course PDE compared with double courses PDE (Fig. 5m). We also performed immunofluorescence experiments on the oocyte development marker protein NOBOX, and the results also suggested that late gestation single course of high-dose PDE had the most significant inhibitory effect on NOBOX expression (Fig. 5p, q). Finally, we performed a correlation analysis between steroidogenic genes and genes involved in oocyte development, and the results suggested that steroidogenic genes (SF1, Star, CYP11A1, Hsd3b, CYP19A1) were negatively correlated with genes related to oocyte development (Nobox, Figla, Sohlh2, Bmp15, Gdf9) (Fig. 5r, |r| ≥ 0.70). In conclusion, PDE can lead to a decrease in the expression of oocyte marker genes, in which, late pregnancy > mid-late pregnancy, high dose > low and medium dose, single course > double courses. Effects of PDE at different stages, doses and courses on oocyte function and follicle development in fetal ovary. a-e Relative mRNA level of NOBOX, Figla, Sohlh2, Bmp15 and Gdf9 in stages effect, n = 8; (f-j) Relative mRNA level of NOBOX, Figla, Sohlh2, Bmp15 and Gdf9 in doses effect, n = 8; (k-o) Relative mRNA level of NOBOX, Figla, Sohlh2, Bmp15 and Gdf9 in courses effect, n = 8; (p, q) Representative immunofluorescence images and mean fluorescence intensity of NOBOX, n = 3; (r) Correlation between steroidogenic genes (SF1, Star, CYP11A1, Hsd3b, CYP19A1) and oocyte development genes (Nobox, Figla, Sohlh2, Bmp15, Gdf9). Mean ± S.E.M. *P < 0.05, **P < 0.01 vs. C or CON, #P < 0.05, ##P < 0.01 vs. C2 or C3. NOBOX, NOBOX oogenesis homeobox; Gdf9, growth differentiation factor 9; Bmp15, bone morphogenetic protein 15; Figlα, folliculogenesis specific basic helix-loop-helix; Ddx4, DEAD box helicase 4; GAPDH, glyceraldehyde phosphate dehydrogenase; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Network toxicology conjectures the potential mechanism of PDE-induced ovary dysplasia at different stages, doses and courses In order to explore the underlying mechanisms of PDE induced ovarian morphological abnormalities and functional damage at different stages, doses and courses of treatment, we used network toxicology methods to explore the underlying mechanisms. Based on the previous results, we hypothesized that PDE could cause decreased ovarian reserve and increased susceptibility to precocious puberty in offspring. Combining with the drug target of dexamethasone, we found that there were 22 common genes (Fig. 6a). There was an interaction between the main functional genes (Fig. 6b, c). GO enrichment analysis of related genes showed that they mainly focused on development and growth regulation, positive regulation of apoptosis process and estrogen response element binding process (Fig. 6d). Furthermore, the top three functional genes were detected by RT-qPCR, and the results showed that PDE could significantly inhibit the expression of interleukin-6 (IL6) and promote the expression of nuclear receptor subfamily 3 C group member 1 (NR3C1) and mitogen-activated protein kinase 14 (MAPK14) (Fig. 6e, i,m). Furthermore, the correlation between the target genes detected above and the key genes of each cellular process described above was analyzed. The results suggested that IL-6 was significantly positively correlated with Ki67, PCNA, and multiple genes involved in oocyte development (Fig. 6f, h), but not with steroid synthesis (Fig. 6g). NR3C1 and MAPK14 showed a significant positive correlation with the key apoptosis gene Caspase3 and steroid-related synthesis genes (Fig. 6j, k,n, o). Taken together, IL6, NR3C1 and MAPK14 may be involved in PDE induced inhibition of ovarian cell proliferation, apoptosis, steroidogenesis and oocyte development. Specifically, IL6 may be associated with the inhibition of cell proliferation and oocyte development, while NR3C1 and MAPK14 are mainly involved in the increase of apoptosis and steroid synthesis, especially in the late pregnancy, single course and high dose of PDE. Network toxicology conjectures the potential mechanism of PDE-induced dysplasia of different stages, doses, and courses of PDE. a Intersection of diseases and drug target genes; (b, c) Diagram of Genetic Interactions; (d) Gene GO enrichment analysis diagram; (e) Relative mRNA level of IL6 at different stages, doses, and courses of PDE, n = 8; (f) Correlation between IL-6 and Ki67/PCNA/Caspase3, n = 8; (g) Correlation between IL-6 and SF1/Star/CYP11A1/Hsd3b/CYP17A1/CYP19A1, n = 8; (h) Correlation between IL-6 and NOBOX/Figlα/Sohlh2/Bmp15/Gdf9, n = 8; (i) Relative mRNA level of NR3C1 at different stages, doses, and courses of PDE, n = 8; (j) Correlation between NR3C1 and Ki67/PCNA/Caspase3, n = 8; (k) Correlation between NR3C1 and SF1/Star/CYP11A1/Hsd3b/CYP17A1/CYP19A1, n = 8; (l) Correlation between NNR3C1 and NOBOX/Figlα/Sohlh2/Bmp15/Gdf9, n = 8; (m) Relative mRNA level of MAPK14 at different stages, doses, and courses of PDE, n = 8; (n) Correlation between MAPK14 and Ki67/PCNA/Caspase3, n = 8; (o) Correlation between MAPK14 and SF1/Star/CYP11A1/Hsd3b/CYP17A1/CYP19A1, n = 8; (p) Correlation between MAPK14 and NOBOX/Figlα/Sohlh2/Bmp15/Gdf9, n = 8. Mean ± S.E.M. *P < 0.05, **P < 0.01 vs. C or CON, #P < 0.05, ##P < 0.01 vs. C2 or C3. PCNA, proliferating cell nuclear antigen; Ki67, antigen identified by monoclonal antibody Ki 67; SF1, splicing factor 1; Star, steroidogenic acute regulatory protein; CYP11A1, cytochrome P450, family 11, subfamily a, polypeptide 1; Hsd3b, hydroxy-delta-5-steroid dehydrogenase, 3 beta- and steroid delta-isomerase cluster; CYP17A1, cytochrome P450, family 17, subfamily a, polypeptide 1; CYP19A1, cytochrome P450, family 19, subfamily a, polypeptide 1; NOBOX, NOBOX oogenesis homeobox; Gdf9, growth differentiation factor 9; Bmp15, bone morphogenetic protein 15; Figlα, folliculogenesis specific basic helix-loop-helix; Ddx4, DEAD box helicase 4; GAPDH, glyceraldehyde phosphate dehydrogenase; C1, control of PDML; PDML, dexamethasone mid-late gestation high-dose exposure; C2, control of PDL, PDM, PDH; PDL, dexamethasone late gestation low-dose exposure; PDM, dexamethasone late gestation moderate-dose exposure; PDH, dexamethasone late gestation high-dose exposure; C3, control of PDD; PDD, high-dose dexamethasone exposure—double courses Discussion Rationale for modeling PDE across gestational stages, doses and courses The National Institutes of Health (NIH) recommends four intramuscular injections of dexamethasone at 6 mg/ h for pregnant women at risk of preterm birth [24, 25]. The dose of dexamethasone used by pregnant women in clinical prescriptions ranges from 0.05 to 0.2 mg/kg·d [26, 27]. Furthermore, interspecies dose conversion was performed by body surface area ratio [28], and the above range corresponds to a mouse dose of 0.62–2.47 mg/kg·d [24]. Given that certain cell types of mice are more sensitive to glucocorticoid-induced apoptosis [29], in this study, the dose range of 0.4–1.6 mg/kg·d was set to explore the effect of prenatal dexamethasone exposure (PDE) on the fetus at a dose lower than the clinical equivalent. In clinical practice, PDE treatment lasts from 24 weeks of gestation to term [30, 31], while in animal studies, it is usually administered 2–11 days before delivery [32, 33]. DOHaD theory points out that exposure to adverse environments during early development can change developmental trajectories and increase the risk of disease susceptibility in offspring [34]. As a critical period of growth and development, fetal period is particularly sensitive to external interference [35]. Ovarian development has several important time nodes, and the development status of fetal ovary during the intrauterine period can affect the adult reproductive ability after birth [36]. Although dexamethasone has a clear effect on promoting fetal lung maturation and reducing neonatal respiratory distress syndrome, its long-term safety is still controversial [37], and there is a large gap in the effect of ovarian development, suggesting that the “precision medicine” strategy needs to be improved urgently. In this study, the effects of PDE on fetal ovarian growth and development were systematically evaluated at three different time points and different doses: mid-late stage (GD14-15), late stage (GD16-17) and double treatment (GD14-17). In this study, it was found that PDE could cause abnormal morphological development of fetal ovary and affect the development and maturation of primordial germ cells in female fetal mice. However, different periods, doses and courses of PDE affect the fetal ovary to different extents. PDE can inhibit cell proliferation, increase apoptosis, enhance steroid synthesis function and inhibit oocyte development, but the effect is in the order of late pregnancy > middle-late pregnancy, high dose > medium and low dose, single course > multiple courses. In summary, this study confirmed the effects of PDE on fetal ovaries at different times, doses and courses of treatment, which will help to enrich the research on ovarian developmental toxicity of dexamethasone and its rules. PDE induces changes in ovarian morphology and cell proliferation/apoptosis in fetal mice During normal development, embryonic day 13.5 (ED13.5) marks the termination of mitotic activity of oogonia and its transition to the meiotic stage, resulting in differentiation into primary oocytes. Oocytes from this period are aggregated in the ovary in the form of “germ cell nests”, forming tight cell clusters through intercellular junctions [38, 39]. In the early postnatal period, programmed dissociation of the germ cell nest structure occurs, accompanied by the invasion and encapsulation of the surrounding somatic cells, and finally the process of wrapping a single oocyte by the somatic cell layer is completed, forming the basic structural unit of the primordial follicle [40]. In addition, we found that the number of oocytes per unit area decreased in the PDH group but not in the PDML group. This might be because exposure during mid-to-late gestation (GD14–15) may still occur before oocyte nests are fully formed, allowing for stronger compensatory or reparative capacity in offspring; whereas late-gestation exposure (GD16–17) coincides with the critical window for the transition of oocyte nests into primordial follicles, making damage more difficult to reverse [41]. In this study, we found that PDE could cause pathological changes such as premature lysis of fetal oocyte nests, protrusion of pregranule cells into the nest and encapsulation of oocytes, and decrease in the number of oocytes per unit. It is suggested that PDE may cause premature ovarian morphological development in offspring, but decrease primordial follicle reserve. During intrauterine development, ovarian cells proliferate actively, which is essential for establishing normal ovarian structure and function. Previous studies have shown that exposure to adverse environment during pregnancy can interfere with the balance of proliferation and apoptosis of fetal ovarian cells [42, 43]. Proliferating cell nuclear antigen (PCNA) is expressed in both oocytes and granulosa cells, which is often used as an early indicator of granulosa cell proliferatio [44], and participates in the assembly regulation of primordial follicles [45]. Ki67 also plays a role in the toxicity of offspring ovarian development induced by adverse environment, and is widely expressed in various ovarian somatic cells [46]. In addition, Caspase3 is considered to be a necessary protease to promote granulosa cell apoptosis, but its role in the regulation of germ cell apoptosis is relatively limited [47]. In the present study, we found that PDE could decrease the expression of PCNA and Ki67 and increase the expression of Caspase3 in fetal ovaries. It is suggested that the effect of PDE exposure in late pregnancy is stronger than that in late pregnancy, and the effect of single course of exposure is stronger than that of double course of exposure, and there is a certain dose-effect relationship. This imbalance of cell proliferation and apoptosis between pregranulosa cells and germ cells is not conducive to the formation of primordial follicle pool, and may affect the reproductive function of female offspring after birth. PDE induces changes in ovarian steroidogenic function and oocyte development in fetal mice The ovary’s core endocrine function is the synthesis and secretion of steroid hormones through a conserved enzymatic cascade. This system uses cholesterol as a substrate and ultimately generates progesterone, androgens, and estrogens through the action of a series of rate-limiting enzymes in the mitochondria and endoplasmic reticulum. Among them, SF1, StAR, CYP11A1, Hsd3b and CYP19A1 are the key regulators and core enzymes in this synthetic cascade, which together constitute the molecular basis of ovarian steroidogenesis. Studies have confirmed that adverse environmental exposure during pregnancy can lead to the activation of ovarian transcription factors in female fetal mice, thereby affecting the estrogen synthesis function [48]. In this study, we observed the PDE increased the mRNA levels of SF1/Star/CYP11A1/Hsd3b/CYP19A1 in fetal ovarian to varying degrees, suggesting premature activation of steroidogenic differentiation in pregranulosa cells. Compared with CON, the late gestation > mid-late gestation, high dose > medium and low dose, single course > multiple courses. The mechanism of germ cell fate determination and follicle formation during mammalian ovarian development is an important basis for reproductive biology. Primordial germ cells migrate to the undifferentiated genital ridge during early embryonic development and then initiate the meiotic program at specific spatiotemporal signals (e.g., around embryonic day 13.5 in mice) to differentiate from proliferating oogonia into primary oocytes arrested in meiotic prophase. Thereafter, the oocyte is gradually surrounded by somatic-derived pregranulosa cells, which assemble together to form primordial follicles. In rodents, this process of follicle assembly mainly occurs early after birth, and the primordial follicle reserve pool of the ovary is established at this stage [49, 50]. The normal formation of primordial follicles strictly depends on the precise coordination between the oocyte and its surrounding pregranulosa cells, which is the key to the complete assembly of primordial follicles and their subsequent activation potential [51]. This process is tightly controlled by a multi-level regulatory network, including the regulation of meiosis and follicle-related gene expression by oocyte-specific transcription factor (such as FIGLA [52], NOBOX [53], etc.), as well as the synergistic action of systemic and local signaling molecules such as steroid hormones (such as estrogen ). Meanwhile, egg-derived factors such as growth differentiation factor 9 (GDF9) and bone morphogenetic protein 15 (BMP15) secreted by oocytes are not only essential for the maintenance of primordial follicles, but also regulate the function of granulosa cells through paracrine signals, thus affecting the initial growth and activation process of follicles [54, 55]. In this study, we found that the expression of oocyte marker genes Figlα and Nobox was down-regulated in all groups, and the expression of follicle-stimulating gene GDF9 was also significantly reduced. These results suggest that PDE can inhibit oocyte development and decrease primordial follicle reserve. Therefore, we speculate that the hyperfunction of fetal ovarian pregranulosa cells and the inhibition of oocyte development caused by PDE may lead to the occurrence of peripheral precocious puberty after birth, and the shortening of reproductive life due to the reduction of primitive reserve, and the susceptibility to premature ovarian failure in adulthood. NR3C1-MAPK14-IL6 may be involved in PDE induced abnormal ovarian development in fetal mice IL-6 is a pleiotropic cytokine that plays a central role in the cross-network of immune regulation, inflammatory response and cell homeostasis [56]. IL-6 is continuously produced by ovarian somatic cells, and its expression fluctuates dynamically during the menstrual or estrous cycle, suggesting that its function is regulated in a cycle-dependent manner. IL-6 is not a simple inflammatory marker, but as an important paracrine/autocrine factor in early ovarian development, it supports cell survival, promotes proliferation and participates in the remodeling of immune microenvironment through JAK/STAT pathway [57]. NR3C1 is a core member of the ligand-activated transcription factor superfamily and a key effector of the body’s neuroendocrine response to internal and external environmental challenges. It can bind endogenous glucocorticoids (such as cortisol) or exogenous synthetic analogues (such as dexamethasone) with high affinity, widely regulating gene expression networks [58, 59]. The transcriptional inhibition of IL-6 gene by NR3C1 is one of the classic anti-inflammatory effects of glucocorticoids [60]. MAPK14, a core member of the p38 mitogen-activated protein kinase family, is a key hub for intracellular signal transduction. Some studies have pointed out that MAPK14 is the core molecular switch for the ovary to sense and integrate stress signals, regulate cell fate and endocrine function, and regulate the functional differentiation of granulosa cells during follicular development [61]. In this study, we found that PDE could up-regulate the expression of NR3C1 and MAPK14 and down-regulate the expression of IL-6 in fetal ovaries. We also observed that different doses of PDE could inhibit cell proliferation, promote apoptosis and lead to inhibition of oocyte development. Therefore, it is speculated that the intrinsic mechanism is that the abnormal activation of NR3C1 signal in fetal ovary caused by PDE exposure during pregnancy. On the one hand, NR3C1 directly inhibits IL-6 transcription and impairs local ovarian pro-survival and homeostasis signaling; On the other hand, NR3C1 activation up-regulates MAPK14 expression and activity by inducing cell stress, thereby initiating a strong pro-apoptotic program. In conclusion, the up-regulation of NR3C1/MAPK14 and the down-regulation of IL-6 have a synergistic effect, and PDE can disrupt the balance between germ cell survival and early follicle formation through the NR3C1-MAPK14-IL6 axis, eventually leading to abnormal ovarian development in fetal mice. Limitations of the study It should be noted that this study aims to clarify the effects and differences of stage, dose, and course on the ovarian function of fetuses exposed to dexamethasone during pregnancy. Therefore, the experimental design did not involve in vivo and in vitro experiments to elucidate the molecular mechanism of ovarian dysfunction and the susceptibility to diseases after birth. This is the limitation of this study and will be the focus of our future work. Conclusion In conclusion, in this study, we first simulated the clinical application characteristics of dexamethasone, and established PDE mice models with different stages, doses and courses of treatment to explore the effects of PDE on ovarian morphology and function in fetal mice. We found that PDE could lead to premature ovarian morphological development, enhanced steroidogenesis of pregranulosa cells, and inhibition of oocyte development in the offspring. The effect of the late gestation was greater than mid-late gestation, and single course was greater than double courses, and there was a quantitative-effect relationship. This study comprehensively examined the changes in the toxicity of PDE exposure at different stages, dose and courses on ovarian development. Through network toxicology, possible mechanisms were identified, providing new theoretical basis for guiding safe medication during pregnancy. It also provided an experimental foundation for analyzing the ovarian developmental toxicity caused by dexamethasone, the susceptibility to adult diseases, and the intrauterine programming mechanism. Acknowledgements The authors extend their sincere thanks to the members of the Hubei Provincial Key Laboratory of Developmentally Originated Disease. Authors’ contributions L.C.: Writing-original draft, Visualization, Project administration, Investigation. S.J.: Software, Methodology, Resources, Investigation, Writing-Review & Editing. ZJ.W.: Methodology, Formal analysis. A.S.: Validation, Supervision. F.M.: Visualization. J.H.: Writing-original draft, Methodology, Funding acquisition, Formal analysis. H.W.: Writing-review & editing, Supervision, Funding acquisition, Conceptualization. Funding This work was supported by grants from the National Natural Science Foundation of China (No. 82030111, U23A20407, 82504938), and the Natural Science Foundation of Hubei Province (No. 2024AFB781). Data availability Data will be made available on request. Declarations Competing interests The authors declare no competing interests. Conflict of interest The authors declare that they have no conflict of interest. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Lu Chen and Shan Jin contributed equally to this work. Contributor Information Jing Huang, Email: ZN001360@whu.edu.cn. Hui Wang, Email: wanghui19@whu.edu.cn. References Associated Data Data Availability Statement Data will be made available on request.