Prenatal Exposure Linked to 30% Lower Ovarian Reserve

Transgenerational inheritance of DOR phenotypes was observed after prenatal PrP exposure

To determine the long-term effects of PrP exposure on offspring reproductive systems, we intraperitoneally injected pregnant ICR mice with human-relevant dosages of PrP during the fetal sex determination period. The experimental procedure performed on animals was shown in Fig. 1a. Prenatal PrP exposure had no significant effect on the body weight of pregnant mice (Supplementary Fig. 1a), the number of birth litters per dam (Supplementary Fig. 1b–d), or the sex ratio of offspring mice (Supplementary Fig. 1e–g). A significant decrease in the male birth weight of F1–F3 offspring and the female birth weight of F3 offspring was found in the PrP-exposed groups (Supplementary Fig. 1h–m), which was consistent with previous findings in a cohort study22. PrP exposure did not affect ovarian weight or index of F1–F3 female offspring at 12 weeks (Supplementary Table 1).

Ovarian reserves of offspring mice were assessed by Anti-Müllerian hormone (AMH) levels and follicular quantity. AMH levels in F1–F3 offspring of the PrP-exposed group decreased significantly compared to those of the control group (Fig. 1b). The morphological features of ovarian follicles at various developmental stages in mice are presented in Supplementary Fig. 2. In the three generations, the PrP-exposed group experienced a considerable rise in the number of atretic follicles (Fig. 1c, d). In parallel, the number of primordial follicles significantly decreased in the PrP-exposed group of F1–F3 offspring (Fig. 1d). Additionally, PrP exposure led to a higher proportion of irregular estrous cycles in the F1–F3 offspring and a significant difference in the F3 generation with protracted metestrus and diestrus periods (Fig. 1e, f). The 17β-estradiol (E2) levels in F1 and F3 offspring in the PrP-exposed group displayed a decreasing trend (Fig. 1g) and progesterone (P4) levels did not differ between the control and the exposed groups in all three generations (Fig. 1h). Collectively, these findings suggest that decreased ovarian reserve induced by prenatal PrP exposure can be inherited transgenerationally.

Apoptosis of granulosa cells and deteriorated oocyte quality were seen in multiple generations

The apoptosis of granulosa cells (GCs) is the leading cause of follicular atresia and poor ovarian reserve23. Therefore, apoptotic GCs in F1–F3 offspring were detected by terminal deoxynucleotidyl transferase-mediated nick-end labeling (TUNEL) staining. Across all F1–F3 generations, GC apoptosis exhibited a marked elevation in the PrP-exposed groups compared to controls (Fig. 2a, b). Electron microscopy was performed to observe the structure of GCs in F1–F3 offspring mice and revealed significantly increased mitochondria abnormalities (Fig. 2c), mainly manifesting as a swollen morphology and broken cristae (Fig. 2d). Interestingly, the synapses between GCs and oocytes were also blurred in the F3 PrP-exposed group (Fig. 2d).

Since deteriorated oocyte quality is an important hallmark of DOR and oocytes serve as the medium for TEI, we further assessed the oocyte quality of offspring by ovulation induction (Fig. 2e) and observed the stages of oocyte maturation and meiosis (Fig. 2f). The number of MII oocytes drastically decreased in PrP-exposed group compared to the control group in F1 and F3 generations (Fig. 2g). This observation was further corroborated by a marked decrease in the protein expression levels of bone morphogenetic protein 15 (BMP15), a known protective factor for oocyte quality, in the F1–F3 offspring from the PrP-exposed group (Fig. 2h, i). Collectively, these data indicate that GC apoptosis is a mechanistical contributor to the observed reduction in ovarian reserve and the suppression of AMH. Importantly, the compromised oocytes may harbor transgenerational epigenetic signatures, potentially mediating intergenerational transmission of reproductive dysfunction.

The hypomethylation of Rhobtb1 was observed in successive generations

DNA methylation is the dominant mechanism underlying TEI caused by ancestral exposure, with the transmission of epigenetic information carried by germ cells across generations being the most crucial component6. To identify the DNA methylation alterations in oocytes at single-base resolution upon PrP exposure, we conducted single-cell whole-genome bisulfite sequencing (scWGBS) on 72 MII oocytes from F2 offspring after ovulation induction (Fig. 3a). Additionally, the whole-genome bisulfite sequencing (WGBS) was employed to determine the methylation levels in ovarian tissues from both F2 and F3 offspring mice (Fig. 3a).

Overall, cytosine-phosphate-guanine (CpG) sites methylation levels displayed a similar pattern in the ovarian tissues of F2 and F3 generations (Fig. 3b). However, an increase in hypomethylated CpG sites (~30%) and hypermethylated CpG sites (~70%) was observed in F2 MII oocytes upon PrP exposure (Fig. 3b). We identified 2114 and 2147 significantly hypermethylated and hypomethylated regions respectively in the ovarian tissues of PrP-exposed F2 offspring. These regions correlate to 1308 exclusively hypermethylated and 1251 hypomethylated genes. The oocytes of PrP-exposed F2 offspring exhibited 12,721 significantly hypermethylated regions and 16,631 hypomethylated regions, corresponding to 7253 genes that were exclusively hypermethylated and 10,117 hypomethylated genes. The ovaries of PrP-exposed F3 offspring had 1505 significantly hypermethylated regions and 1283 hypomethylated regions, corresponding to 963 exclusively hypermethylated genes and 766 hypomethylated genes, respectively (Fig. 3c). The percentage distribution of each functional annotation was shown in Fig. 3d. Notably, in contrast to the ovarian tissues, most hypomethylated regions in the oocytes were located within the promoter region (56.8%) (Fig. 3d).

To identify epigenetic molecules transmitted across generations, we performed an overlap analysis of methylation sequencing data from F2 ovaries, F3 ovaries, and F2 oocytes. A Sankey diagram was constructed to illustrates the concordant methylation changes in differentially methylated regions (DMRs) across these groups (Fig. 3e). Remarkably, DMRs associated with Rho-related BTB domain-containing protein 1 (Rhobtb1) and the gene 4930519F24Rik exhibited consistent hypomethylation patterns in F2 ovaries, F2 oocytes, and F3 ovaries (Fig. 3e). It is noteworthy that 4930519F24Rik has not been named or studied yet, and given the regulatory roles of Rhobtb1 in cell growth and apoptosis24, we prioritized the Rhobtb1 gene for further investigation. Surprisingly, methylation levels in Rhobtb1 promoter region were significantly reduced in both F2 and F3 ovarian tissues, as well as in F2 oocytes of PrP-exposed group, compared to the control group (Fig. 3f, p < 0.05; F2 ovary diff.Methyl = 0.45, F2 oocyte diff.Methyl = 0.33, F3 ovary diff.Methyl = 0.34). This suggests that aberrant methylation of Rhobtb1 may serve as an epigenetic factor transmitted from F2 to F3 generation. The heatmap displayed the methylation patterns in the Rhobtb1 promoter region across ovarian tissues from F2 and F3 offspring, and oocytes from F2 offspring (Fig. 3g). Furthermore, RNA sequencing of the F3 offspring ovaries revealed an increased trend in Rhobtb1 expression levels in the exposed group compared to the control group (Fig. 3h, p = 0.198). Correspondingly, the protein levels of RhoBTB1 in F2 and F3 ovaries (Fig. 3i, j) and F2 oocytes increased significantly in the PrP-exposed group relative to the control group (Fig. 3k, l). These results suggest that the hypomethylation in the promoter region of Rhobtb1 mediate the TEI of DOR induced by prenatal PrP exposure.

Hypomethylated Rhobtb1 contributed to granulosa cell apoptosis and reduced AMH secretion

As established previously, prenatal PrP exposure induced excessive GC apoptosis in F1–F3 offspring ovaries, identifying GC apoptosis as a primary mechanism underlying transgenerational DOR pathogenesis. To further investigate the impact of PrP exposure on Rhobtb1 in GCs, we employed the human ovarian granulosa cell line (KGN) for in vitro exposure experiments. KGN cells were treated with 50 μM and 100 μM PrP for 48 h, based on the results of the IC50 (Supplementary Fig. 3). This treatment resulted in a significant reduction in methylation levels at the CpG site −1555 to the transcription start site (TSS) within the promoter region of Rhobtb1 in the PrP-exposed groups (Fig. 4a). Correspondingly, there was a marked increase in RhoBTB1 proteins levels after PrP exposure (Fig. 4b, c). The concentrations of AMH and E2 were significantly decreased post-exposure (Fig. 4d, e). Moreover, a dose-dependent increase in apoptosis was observed (Fig. 4f, g). Subsequently, western blotting analysis was used to determine the protein levels of key genes involved in the mitochondrial apoptotic pathway in GCs. As expected, a noticeable decrease in B-cell lymphoma 2 (BCL-2), and a significant increase in Bcl-2 associated X protein (BAX) and cleaved Caspase-3 (active-Caspase3) were observed (Fig. 4h, i).

To further clarify the effects of RhoBTB1 overexpression on GC apoptosis, we employed an adenoviral vector encoding Rhobtb1 to up-regulate RhoBTB1 in KGN cells in the absence of PrP exposure. The infection efficiency exceeded 90% (Fig. 4j), resulting in a significant increase of RhoBTB1 proteins expression (Fig. 4k, l). Upon the overexpression of RhoBTB1, there was a notable decrease in AMH and E2 levels (Fig. 4m, n), accompanied by an elevated number of apoptotic GCs (Fig. 4o, p). Similarly, the mitochondrial apoptosis pathway-related genes including BAX, BCL-2, and active-Caspase3 exhibited significant alterations (Fig. 4q, r). Collectively, these findings suggest that Rhobtb1 hypomethylation may contribute to the excessive apoptosis of GCs.

Potential targets for RhoBTB1 identified through ovarian transcriptomic profiles and DNA methylation patterns

The abnormal epigenetic molecules carried by germ cells lead to the TEI of diseases by regulating downstream target genes or pathways. To explore the mechanisms underlying the TEI of DOR induced by hypomethylated Rhobtb1, we conducted RNA sequencing (RNA-seq) and WGBS on the ovarian tissues from the F3 generation (Fig. 5a). We discovered 544 differentially expressed genes (DEGs) in the ovaries of F3 progeny exposed to PrP, in comparison to control ovaries, with 288 down-regulated and 256 genes being up-regulated (Fig. 5b; Supplementary Fig. 4a). The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis revealed significant up-regulation of the p53 signaling pathway and oxidative phosphorylation (Fig. 5c), both of which are implicated in apoptosis in GCs25,26. Conversely, genes involved in steroid biosynthesis, which are integral to ovarian endocrine function27, were significantly down-regulated (Fig. 5d). Subsequently, the RNA levels of genes involved in these three pathways were validated by quantitative real-time polymerase chain reaction (qRT-PCR) (Fig. 5e–j). Protein levels of Catalase (CAT), superoxide dismutase 2 (SOD2), and hydroxysteroid 17-beta dehydrogenase 7 (HSD17B7) protein levels in the ovaries of the PrP-exposed F3 offspring were significantly lower than those of the control group, while BAX exhibited the opposite trend (Supplementary Fig. 4b–i). These results suggest that oxidative stress, increased apoptosis, and the disturbance of sex hormone synthesis are the main biological processes involved in the DOR-like phenotypes of F3 offspring.

Epigenetic changes in the ovaries of F3 PrP-exposed mice were uniformly distributed across all the chromosomes (Fig. 5k). The comparison of the ovaries of PrP-exposed F3 offspring to controls identified 1505 significantly hypermethylated and 1283 significantly hypomethylated regions, corresponding to 963 exclusively hypermethylated and 766 exclusively hypomethylated genes (Fig. 5l). To gain more insight into the functional implications of these epigenetic modifications after PrP exposure, we assessed the enrichment of DMRs within annotated regions relative to their background regions (Fig. 5m). KEGG analysis revealed the pathways that hypermethylated and hypomethylated genes enriched respectively (Supplementary Fig. 4j, k).

We searched for overlap between differentially methylated genes and DEGs in F3 offspring to determine whether variations in gene expression were related to changes in DNA methylation (Fig. 5n). 38 genes were identified through this cross-omics integration (Fig. 5o). Chymotrypsin-like elastase family, member 3B (Cela3b), claudin 18 (Cldn18), cathepsin K (Ctsk), and gap junction protein, gamma 3 (Gjc3) displayed methylation changes in the promoter region (Fig. 5p). Cela3b might serve as a novel potential prognostic biomarker for ovarian serous cystadenocarcinoma28. In ovarian tumors, Cldn18 has frequent ectopic activation29. Ctsk is regarded as a regulator of oocyte quality30. However, to date, no studies have explicitly reported the expression profile or functional regulatory roles of Gjc3 in ovarian tissues. Among the 38 common genes, fibroblast growth factor 18 (Fgf18) is closely related to GC apoptosis and atretic follicles31,32. However, the altered methylation region of Fgf18 was located in the distal intergenic region (Fig. 5p). As the hypothalamus is the primary regulator of reproductive function, we performed next-generation sequencing-based bisulfite sequencing PCR (BSP) in the hypothalamus of control and PrP-exposed mice of F3 generation to determine whether alterations in DNA methylation in the ovaries of PrP-exposed mice could occur in other tissues. Similar to the results observed for ovaries, the methylation levels of Cldn18, Ctsk, and Gjc3, in the promotor regions increased significantly in the PrP-exposed group compared to the control group (Supplementary Fig. 4l). RNA expression of the genes was validated in the hypothalamus of control and PrP-exposed mice by qRT-PCR (Supplementary Fig. 4m). Similar to the results observed for ovaries, the RNA expression levels of Cldn18 and Ctsk decreased significantly in the PrP-exposed group compared to the control group. These results suggest that the five genes mentioned above are potential targets for DOR.

RhoBTB1 activated the MAPK pathway and induced GC apoptosis by upregulating FGF18 via a ubiquitination-dependent pathway

RhoBTB1, an atypical member of the Rho-GTPase family, exhibits a unique structural architecture composed of an N-terminal GTPase domain, a proline-rich hinge region, tandem BTB domains (BTB1 and BTB2), and a conserved C-terminal region. Notably, the BTB domains facilitate protein complex assembly through homomeric or heteromeric oligomerization33. Classical mechanistic studies demonstrate that the BTB1 domain specifically recruits the Cullin 3-based E3 ubiquitin ligase complex, while substrate recognition is cooperatively mediated by the BTB2 domain, GTPase domain, and C-terminal region. This coordinated interaction enables substrate ubiquitination and subsequent degradation via the ubiquitin-proteasome system, thereby regulating protein homeostasis34.

To explore the interaction and potential regulatory patterns between Rhobtb1 and the five aforementioned genes, we utilized the STRING database. Our analysis revealed indirect interactions between RhoBTB1 and FGF18, as well as between RhoBTB1and CTSK (Fig. 6a; Supplementary Fig. 5a). Notably, CTSK protein levels exhibited an increased trend following GCs PrP exposure (Supplementary Fig. 5b), a finding that contrasts with the results of ovarian RNA-seq assay from the unexposed F3 generation (Fig. 5p). The observed reduction in CTSK expression in F3 ovaries implies that decreased CTSK expression may function as a protective factor, mitigating the detrimental effects of PrP exposure. It is noteworthy that the methylation level of Fgf18 in the F3 PrP-exposed group showed no significant difference in the promoter region, indicating that Fgf18 is not an epigenetic messenger molecule, but is likely a target regulated by abnormal epigenetic molecules, which inherited across generations. Additionally, as shown in Fig. 6a, mitogen-activated protein kinase 3 (MAPK3) had an interaction with FGF18. Further, an increase in FGF18 proteins and an activation of MAPK signaling pathway were observed after PrP exposure in KGN cells (Fig. 6b, c). This phenomenon was also observed in Rhobtb1 over-expression models (Fig. 6d, e), which was in accordance with previous studies that the MAPK signaling pathway can be activated by fibroblast growth factors (FGFs)35. Upon silencing Fgf18 using siRNA-F02 in GCs (Supplementary Fig. 5c), identified as the most effective siRNA, following PrP exposure, there was a significant reduction in the activation of the MAPK signaling pathway, whereas the expression of RhoBTB1 remained unchanged (Fig. 6f, g).