The Animal Experiment Administration Committee of Tongji University approved all animal experimental procedures (TJAA06422201). Thirty-five female Sprague-Dawley (SD) rats, aged 10–12 weeks (weighing 220–270 g), were housed in the specific pathogen-free (SPF) facility of Tongji Hospital with free access to commercial chow. The rats were randomly assigned to control and hypothyroidism (hypo) groups. The hypo group (20 rats) received drinking water containing 0.02% (w/v) methimazole (MMI) for 4 weeks (Kent et al., 2022), while the control group (15 rats) received drinking water. During the modeling period, the estrus cycle of the rats was monitored daily using vaginal smears, and their body weight was measured weekly.

After 4 weeks, thyroid tissue and serum were collected. Hypothyroidism was mainly evaluated by serum levels of thyroid-stimulating hormone (TSH), free triiodothyronine (fT3), and free thyroxine (fT4). The serum levels of sex hormones were also measured. The bilateral ovaries were weighed and subjected to morphological evaluations. Superovulation was induced to assess ovulation capacity. Oocytes and cumulus granulosa cells were collected, and single-cell libraries were established using the Smart-seq2 method for high-throughput sequencing.

Blood was collected through cardiac puncture and left at room temperature for 1 h. It was then centrifuged at 1,000 g for 20 min at 4°C to obtain serum. According to the manufacturer’s instructions, serum levels of thyroid hormones and sex hormones were measured using an ELISA kit (Bim, USA). Sex hormones include follicle-stimulating hormone (FSH), luteinizing hormone (LH), estradiol (E2), progesterone (P), and prolactin (PRL). Each group included at least three biological samples, each with at least three technical replicates.

The thyroids and ovaries were quickly collected, with fat removed, and fixed in 4% paraformaldehyde. After dehydration with ethanol, the tissues were embedded in paraffin. Serial sections of the embedded tissue were made, each with a thickness of 5 μm (Chen M. et al., 2022). For ovarian tissue, 10 sections were randomly selected from the largest cross-section for H&E staining, and the number of follicles at different stages was counted. The criteria for determining the stages of follicles have been described previously (Fu et al., 2017). Ovarian follicles are classified into primordial follicles (PrF), primary follicles (PF), secondary follicles (SF), antral follicles (AF), corpora lutea (CL), and atretic follicles (AtF) (Zhou et al., 2023).

After completing the modeling, superovulation was performed in both rat groups. 40 IU of pregnant mare serum gonadotropin (PMSG) (Sansheng, China) was administered via intraperitoneal injection, followed by 40 IU of human chorionic gonadotropin (hCG) (Sansheng, China) 48 h later. After 14–16 h, cumulus-oocyte complexes (COCs) were collected from the ampulla of the fallopian tubes (Wang et al., 2020). COCs were then placed in a solution of hyaluronidase (Aibei, China) for 2–3 min to obtain denuded oocytes. The number of oocytes obtained and the proportion of abnormally morphologic oocytes were recorded.

Oocytes and cumulus granulosa cells were collected through superovulation, with the control group collecting 39 oocytes and their corresponding cumulus granulosa cells, and the hypo group collecting 35. The cDNA library was constructed using the Smart-seq2 method. In brief, cells were lysed to extract mRNA. The mRNA was then reverse transcribed into cDNA, followed by amplification and purification. The cDNA was fragmented, adapters were added and then amplified and purified to prepare the library. Sequencing was performed using the Illumina NovaSeq 6000 system.

We utilized the FPKM value as the index for measuring gene expression and analyzed gene expression in the two groups. Differentially expressed genes (DEGs) of the cumulus granulosa cells or oocytes between the hypo group and control group were identified based on | log₂ (fold change) | >1 and P value <0.05 using the R package. Gene ontology (GO) enrichment analysis was conducted using the R topGO package and the DAVID online tool (https://david.ncifcrf.gov/).

Ovaries were collected on the day of estrus and placed in pre-cooled M2 medium (Aibei, China). 26-gauge needles were used to puncture the antral follicles to release granulosa cells. The granulosa cells were collected from the medium by centrifugation. Following the manufacturer’s instructions, mRNA was extracted using TRIzol (Invitrogen, USA). Reverse transcription was performed using the RevertAid First Strand cDNA Synthesis Kit (Thermo Scientific, USA), and qPCR was completed using TB Green Premix Ex Taq II (Takara, Japan). mRNA levels were normalized to β-actin. Each experiment was repeated at least three times independently. The primer sequences are shown in Supplementary Table S1.

Ovarian paraffin sections were stained using the TUNEL staining kit (elabscience, China). In brief, sections were deparaffinized, hydrated, and permeabilized. The sections were then incubated with the Labeling Solution. DNA was subsequently stained with DAPI. After staining, the sections were observed using the fluorescence microscope (Nikon, Eclipse Ti).

Oocytes were incubated in M2 medium containing 10 μM H2DCFDA (MCE, USA), 5 μM MitoSOX Red (MCE, USA), and 10 μg/mL Hoechst 33342 (Beyotime, China) at 37°C in a 5% CO₂ atmosphere for 15 min. They were then washed three times in the M2 medium and observed under the fluorescence microscope.

All data were analyzed using GraphPad Prism 8 and presented as mean ± standard error of the mean (SEM). Each experiment was repeated at least three times. If the data conformed to the normal distribution, the t-test was used. If not, the Mann-Whitney test was used. P value <0.05 was considered a significant difference. Significance was defined as *P < 0.05, **P < 0.01, ***P < 0.001 and ****p < 0.0001.

To investigate the effect of adult-onset hypothyroidism on the ovarian follicle development, we first established a rat model by using methimazole (MMI). After for 4 weeks inducation, we found the thyroids were significantly enlarged in the hypothyroidism (hypo) group, while H&E staining showed that the volume of thyroids follicles were significantly smaller than that in the control group, and the colloid in the follicular cavities were reduced (Figure 1A), while the body weight of the hypo group were significantly lower than that of the control group and showed a decreasing trend starting from the third week (Figure 1B). The hormone assay showed the level of TSH in the hypo group was significantly higher compared with the control group, while the levels of fT3 and fT4 were remarkably lower (Figures 1C–E; Supplementary Table S2), which was consistent with the hormonal changes of hypothyroidism. These results suggested that the model was successfully established (Korkmaz et al., 2023).

To determine the effect of hypothyroidism on the estrus cycle, we detected the vaginal smears for each phase including proestrus (P), estrus (E), metestrus (M), and diestrus (D) (Salleh et al., 2022; Guillén-Ruiz et al., 2021) (Supplementary Figure S1A), the results showed that the estrus cycle of rats in the hypo group became irregular, primarily manifested as a significantly prolonged estrus cycle, extended from 4.37 ± 0.18 days to 6.48 ± 0.21 days, while the proportion of proestrus in the overall estrus cycle was decreased from 9.37% ± 1.88%–5.43% ± 1.69%, and the proportion of estrus tended to increase (Table 1; Supplementary Figures S1B,C). These data suggested that the estrus cycle was abnormal in rats with adult-onset hypothyroidism. The Figure 1F reflects the typical estrus pattern of the two groups of rats.

To evaluate the impact of hypothyroidism on follicular development, we first collected and weighed the ovaries from both groups. Although the ovarian weight in the hypo group was significantly lower (Figure 2A), there was no significant difference in the ovary-to-body weight ratio between the two groups (Figure 2B). Furthermore, H&E staining showed that the number of corpora lutea (CL) in the hypo group was significantly reduced, and the number of atretic follicles (AtF) was remarkably increased, with a trend towards a decrease in antral follicles (AF) (Figures 2C,D). Additionally, serum sex hormone levels showed significant abnormalities. Compared to the control group, levels of FSH, LH, and PRL significantly increased, while levels of E2 and P significantly decreased in the hypo group (Figures 2E–I; Supplementary Table S3). These results confirmed the effects of hypothyroidism on serum sex hormone levels and follicular development.

Furthermore, we assessed ovulation in both groups of rats (Figures 2J–L). Compared to the control group, the number of oocytes retrieved from the hypo group was significantly lower, while the proportion of abnormally morphologic oocytes increased (Figures 2K,L), indicating that hypothyroidism led to abnormal ovarian follicle development and negatively impacts ovulation.

Granulosa cells provide material support and facilitate signal transduction for oocytes during follicle development (Chen et al., 2021; Li et al., 2022). To further investigate how hypothyroidism impairs ovulation capacity, we focused on the potential changes in granulosa cells and oocytes. After collecting oocytes and corresponding cumulus granulosa cells from the control and hypo groups, the transcriptome was detected by using single-cell RNA sequencing. Compared to those in the control group, granulosa cells in the hypo group had a total of 1,583 differentially expressed genes (DEGs), with 941 upregulated and 642 downregulated (Figure 3A). GO enrichment analysis showed that the upregulated DEGs were mainly enriched in apoptosis, inflammation, and oxidative stress, while the downregulated DEGs were primarily involved in gap junctions, cell proliferation, and sex hormone response (Figure 3B), suggesting that the damage to granulosa cell function, including increased apoptosis and decreased proliferation ability, may be a critical cause of follicular atresia and oocyte maturation disorders.

Apoptosis is one of the common ways of cell death. Single-cell RNA sequencing data suggest that adult-onset can promote granulosa cell apoptosis. In addition, previous studies have shown that ferroptosis and pyroptosis are also associated with granulosa cell death in polycystic ovarian syndrome (PCOS) and diminished ovarian reserve (DOR) (Wang et al., 2025; Yang et al., 2025), therefore, we investigated three major cell death pathways: ferroptosis, pyroptosis, and apoptosis. The ferroptosis-related genes, including Glutathione Peroxidase 4 (Gpx4), Solute Carrier Family 7 Member 11 (Slc7a11), Acyl-CoA Synthetase Long Chain Family Member 4 (Acsl4), Lysophosphatidylcholine Acyltransferase 3 (Lpcat3), Apoptosis Inducing Factor Mitochondria Associated 2 (Aifm2), and GTP Cyclohydrolase 1 (Gch1), and the pyroptosis-related genes, including Caspase 1 (Casp1), Caspase 11 (Casp11), Gasdermin D (Gsdmd), and Gasdermin E (Gsdme), were analyzed by qPCR. The results showed no significant differences in the expression levels of these genes between the control and hypo groups (Figures 3C,D), indicating that the death of granulosa cells in the hypo group is not due to ferroptosis or pyroptosis.

We also detected the key genes involved in the apoptosis pathway, including BCL2 Apoptosis Regulator (Bcl2), BCL2 Like 1 (Bcl2l1), BCL2 Associated Agonist Of Cell Death (Bad), Caspase 3 (Casp3), and Caspase 9 (Casp9). The results indicated that the expression levels of Bcl2 and Bcl2l1 were significantly decreased in the hypo group, while the expression level of Bad was remarkably increased. No significant differences were observed in the expression levels of Casp3 and Casp9 between the groups (Figure 3E). These results are consistent with scRNA-seq data, suggesting an enhancement of the apoptotic pathway in granulosa cells. Furthermore, we conducted TUNEL staining on paraffin sections of ovaries from both groups, focusing on the apoptosis status of granulosa cells. The results showed no significant differences in the proportion of TUNEL-positive granulosa cells between the groups in primary and secondary follicles. However, in the hypo group, especially in the pre-ovulatory stage of antral follicles (AF), the proportion of TUNEL-positive granulosa cells was significantly increased (Figures 3F,G). These results suggested that adult-onset hypothyroidism has no effect on granulosa cells in the primary and secondary stages, but significantly induced apoptosis at the antral follicle stage, which may impact oocyte development.

To further investigate the effects of hypothyroidism and granulosa cell apoptosis on oocytes, we further analyzed single-cell transcriptome data. The analysis revealed a total of 956 DEGs in the hypo group, with 496 genes upregulated and 460 genes downregulated (Figure 4A). GO enrichment analysis indicated that the upregulated DEGs in the hypo group were mainly enriched in pathways related to oxidative stress, cell death, and aging, while the downregulated DEGs were enriched in cell adhesion and subsequent embryo development (Figure 4B). These findings suggest that excessive cell death and oxidative stress were common characteristics between granulosa cells and oocytes in hypothyroidism.

During the maturation of oocytes, the dynamic balance between intracellular oxidative and antioxidative systems must be maintained. Granulosa cells play a crucial role in maintaining the redox balance of oocytes, and an increase in granulosa cell apoptosis may lead to oxidative stress in oocytes (Shi et al., 2023). Sequencing results indicated that many upregulated DEGs in the hypo group were enriched in biological processes related to oxidative stress. Based on this, we conducted tests on relevant markers, which primarily included intracellular ROS levels (indicated by H2DCFDA staining, Green) and mitochondrial superoxide levels (indicated by MitoSOX Red staining, Red). Compared to the control group, the levels of intracellular ROS and mitochondrial superoxide were all significantly elevated in the hypo group (Figures 4C–E), indicating oxidative stress in oocytes of the hypo group. Notably, we found that the intracellular ROS levels of different oocytes in the control group were similar, but the levels of mitochondrial superoxide varied significantly, suggesting heterogeneity in mitochondrial superoxide levels among oocytes. However, mitochondrial superoxide levels in oocytes were increased and became similar in the hypo group (Figures 4C,D), likely due to the apoptosis of granulosa cells.