Icaritin improves pregnancy outcomes in endometriosis by inhibiting ovarian granulosa cell senescence

To identify specific DEGs and potential signalling pathways associated with EMS-associated infertility, RNA sequencing and functional enrichment analyses were conducted. Using the criteria of P  < 0.05 and |log₂FC| ≥ 1, significant DEGs were screened from the transcriptomes of human GCs in CON group and EMS group. Principal component analysis (PCA) and clustering analysis of DEGs (Fig.  1 A and B) revealed distinct transcriptional differences between the two groups, with clear clustering patterns. Compared with those in CON group, 2,872 upregulated genes and 1,883 downregulated genes were identified in EMS group (Fig.  1 C). Then GO and KEGG analyses were conducted. GO annotation analysis focused on cellular and metabolic process (Fig.  1 D), whereas KEGG annotation analysis pointed to energy metabolism, cell growth and death (Fig.  1 E). GO enrichment analysis associated DEGs with NADH dehydrogenase complex and mitochondrial respiratory chain complex I (Fig.  1 F), whereas KEGG enrichment analysis revealed differential pathways involving ROS, OXPHOS, cellular senescence, and P53 signalling pathway (Fig.  1 G). Thus, we hypothesized that ovarian GCs in patients with EMS-associated infertility may exhibit OXPHOS dysfunction and cellular senescence. Fig. 1 Analysis of RNA sequencing of human GCs in CON group ( n  = 10) and EMS group ( n  = 8). ( A ) PCA plot of transcriptomes from the two groups. ( B ) Clustered heatmap of significant DEGs. ( C ) Volcano plot of significant DEGs. ( D ) GO annotation analysis plot of significant DEGs. ( E ) KEGG annotation analysis plot of significant DEGs. ( F ) GO enrichment analysis plot of significant DEGs. ( G ) KEGG enrichment analysis plot of significant DEGs Analysis of RNA sequencing of human GCs in CON group ( n  = 10) and EMS group ( n  = 8). ( A ) PCA plot of transcriptomes from the two groups. ( B ) Clustered heatmap of significant DEGs. ( C ) Volcano plot of significant DEGs. ( D ) GO annotation analysis plot of significant DEGs. ( E ) KEGG annotation analysis plot of significant DEGs. ( F ) GO enrichment analysis plot of significant DEGs. ( G ) KEGG enrichment analysis plot of significant DEGs Next, we treated KGN cells with FF to validate sequencing results. Firstly, CCK-8 assays were performed to determine optimal concentration and duration of FF. After treating KGN cells with FF from normal control females (CON-FF) or EMS patients (EMS-FF) at varying volume concentrations for 24 and 48 h, a significant decreasing trend in optical density (OD) values was observed with increasing FF concentration. Significant intergroup differences in OD values were detected at 10% volume concentration, which became more pronounced after 48 h (Fig. S1 ). Therefore, concentration of 10% and duration of 48 h were selected. OXPHOS-associated assays revealed that ATP concentration and OCR in EMS-FF-treated KGN cells were significantly reduced (Fig.  2 A and B). Further analysis revealed that basal respiration, maximal respiration, and ATP-coupled respiration levels were significantly reduced in EMS-FF group (Fig.  2 C). Additionally, ROS levels, percentages of SA-β-gal-positive cells and proportions of cells in G1 phase in EMS-FF group were significantly elevated (Fig.  2 D - I). Similarly, mRNA and protein expression levels of P21 and P53 were significantly greater in EMS-FF group (Fig.  2 J - O). Collectively, these findings confirmed that EMS-FF may induce both OXPHOS dysfunction and cellular senescence in KGN cells, supporting our hypothesis that ovarian GCs in patients with EMS-associated infertility may exhibit OXPHOS dysfunction and cellular senescence. Fig. 2 EMS-FF may induce OXPHOS dysfunction and senescence in KGN cells. ( A ) Comparison of ATP concentrations between CON-FF group and EMS-FF group ( n  = 3/group). ( B ) Comparison of OCR between CON-FF group and EMS-FF group ( n  = 3/group). ( C ) Comparison of basal respiration, maximal respiration, proton leakage, and ATP-coupled respiration levels between CON-FF group and EMS-FF group ( n  = 3/group). ( D ) Comparison of ROS levels between CON-FF group and EMS-FF group ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells between CON-FF group and EMS-FF group ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions between CON-FF group and EMS-FF group ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( L ) Comparison of P21 protein expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( M ) Comparison of P53 protein expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (L). ( O ) Quantification of P53 protein expression levels in (M). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant EMS-FF may induce OXPHOS dysfunction and senescence in KGN cells. ( A ) Comparison of ATP concentrations between CON-FF group and EMS-FF group ( n  = 3/group). ( B ) Comparison of OCR between CON-FF group and EMS-FF group ( n  = 3/group). ( C ) Comparison of basal respiration, maximal respiration, proton leakage, and ATP-coupled respiration levels between CON-FF group and EMS-FF group ( n  = 3/group). ( D ) Comparison of ROS levels between CON-FF group and EMS-FF group ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells between CON-FF group and EMS-FF group ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions between CON-FF group and EMS-FF group ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( L ) Comparison of P21 protein expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( M ) Comparison of P53 protein expression levels between CON-FF group and EMS-FF group ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (L). ( O ) Quantification of P53 protein expression levels in (M). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Energy metabolism defects, such as those in glycolysis and OXPHOS, may occur prior to cellular senescence [ 26 ]. In this study, we hypothesized that OXPHOS dysfunction in ovarian GCs from EMS patients may be a critical factor driving cellular senescence. To validate this hypothesis, we focused on oocytes—another cell type closely associated with GCs. GCs play a vital role in oocyte quality and ovarian function [ 27 ]. A previous study published human oocytes single-cell RNA sequencing study [ 28 ]. In the present study, we intersected significant DEGs identified by oocytes single-cell RNA sequencing with those identified by GCs RNA sequencing, and screened 595 core genes for further analysis (Fig. 3 A). GO enrichment analysis focused on mitochondrial respiratory chain complex I and NADH dehydrogenase activity (Fig. 3 B). KEGG enrichment analysis highlighted ROS, OXPHOS, and cellular senescence (Fig. 3 C). These findings suggest that GCs and oocytes from EMS patients may exhibit concurrent OXPHOS dysfunction and cellular senescence. Fig. 3 NDUFS8 may play a critical role in the progression of OXPHOS dysfunction and cellular senescence in ovarian GCs from EMS patients. ( A ) Venn diagram of intersecting DEGs between the human oocyte single-cell sequencing dataset (CON: EMS = 16:16) and the human GC RNA sequencing dataset (CON: EMS = 10:8). ( B ) GO enrichment analysis plot of intersecting DEGs. ( C ) KEGG enrichment analysis plot of intersecting DEGs. ( D ) OXPHOS-associated genes among intersecting DEGs. ( E ) Comparison of NDUFS8 mRNA expression levels in KGN cells after intervention with CON-FF or EMS-FF ( n  = 3/group). ( F ) Comparison of NDUFS8 protein expression levels in KGN cells after intervention with CON-FF or EMS-FF ( n  = 3/group). ( G ) Quantification of NDUFS8 protein expression levels in (F). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant NDUFS8 may play a critical role in the progression of OXPHOS dysfunction and cellular senescence in ovarian GCs from EMS patients. ( A ) Venn diagram of intersecting DEGs between the human oocyte single-cell sequencing dataset (CON: EMS = 16:16) and the human GC RNA sequencing dataset (CON: EMS = 10:8). ( B ) GO enrichment analysis plot of intersecting DEGs. ( C ) KEGG enrichment analysis plot of intersecting DEGs. ( D ) OXPHOS-associated genes among intersecting DEGs. ( E ) Comparison of NDUFS8 mRNA expression levels in KGN cells after intervention with CON-FF or EMS-FF ( n  = 3/group). ( F ) Comparison of NDUFS8 protein expression levels in KGN cells after intervention with CON-FF or EMS-FF ( n  = 3/group). ( G ) Quantification of NDUFS8 protein expression levels in (F). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test. * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Given that OXPHOS dysfunction may precede cellular senescence [ 26 , 29 ], we focused on OXPHOS related genes among 595 intersecting genes (Fig. 3 D). Comparative analysis of log 2 FC values revealed that these genes were significantly downregulated in ovarian GCs but upregulated in oocytes (Tables S4 and S5). Notably, NDUFS8 was the only gene with top two highest absolute log 2 FC values in both datasets. Then we treated KGN cells with 10% CON-FF or EMS-FF for 48 h. NDUFS8 mRNA and protein levels were significantly reduced in EMS-FF group (Fig. 3 E - G). These results suggest that NDUFS8 downregulation may be critical in OXPHOS dysfunction and cellular senescence in ovarian GCs from EMS patients. To investigate the role of NDUFS8 in OXPHOS dysfunction and cellular senescence, this study employed lentiviral shRNA transfection to establish stable negative control (NC) and NDUFS8-knockdown (Sh-NDUFS8) KGN cells. First, optimal MOI of 10 was determined in preliminary experiments (Fig. S2). Subsequent PCR and WB analyses confirmed successful knockdown (Fig. S3). In formal experiments, OXPHOS assays revealed that ATP concentration and OCR were markedly reduced in Sh-NDUFS8 group (Fig.  4 A and B). Further analysis revealed decreased basal respiration, maximal respiration, and ATP-linked respiration in Sh-NDUFS8 group (Fig.  4 C), indicating that NDUFS8 knockdown impairs OXPHOS in KGN cells. Fig. 4 NDUFS8 knockdown may induce senescence in KGN cells via ROS-dependent pathway. ( A ) Comparison of ATP concentration between NC and Sh-NDUFS8 groups ( n  = 3/group). ( B ) Comparison of OCR between NC and Sh-NDUFS8 groups ( n  = 3/group). ( C ) Comparison of basal respiration, maximal respiration, proton leakage, and ATP-linked respiration between NC and Sh-NDUFS8 groups ( n  = 3/group). ( D ) Comparison of ROS levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells in NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( L ) Comparison of P21 protein expression levels in among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( M ) Comparison of P53 protein expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (L). ( O ) Quantification of P53 protein expression levels in (M). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA(between four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant NDUFS8 knockdown may induce senescence in KGN cells via ROS-dependent pathway. ( A ) Comparison of ATP concentration between NC and Sh-NDUFS8 groups ( n  = 3/group). ( B ) Comparison of OCR between NC and Sh-NDUFS8 groups ( n  = 3/group). ( C ) Comparison of basal respiration, maximal respiration, proton leakage, and ATP-linked respiration between NC and Sh-NDUFS8 groups ( n  = 3/group). ( D ) Comparison of ROS levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells in NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( L ) Comparison of P21 protein expression levels in among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( M ) Comparison of P53 protein expression levels among NC, Sh-NDUFS8, NC+CoCl₂, and Sh-NDUFS8 + NAC groups ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (L). ( O ) Quantification of P53 protein expression levels in (M). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA(between four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Previous studies have shown that OXPHOS inhibition leads to ROS accumulation [ 30 ]. Notably, mitochondrial Lon protease facilitates ROS production by modulating NDUFS8 expression [ 31 ]. Thus, we hypothesized that NDUFS8 knockdown may induce cellular senescence via ROS overproduction. To test this, NC group and Sh-NDUFS8 group were treated with pro-oxidant cobalt chloride hexahydrate (CoCl₂·6 H₂O) and ROS scavenger N-acetylcysteine (NAC), respectively, followed by analysis of cellular senescence markers. Compared with those in NC group, ROS levels, percentages of SA-β-gal-positive cells and proportions of cells in G1 phase were elevated in Sh-NDUFS8 group and NC+CoCl₂ group, which were reversed by NAC (Fig. 4 D - I). Similarly, mRNA and protein expression levels of P21 and P53 were significantly increased in Sh-NDUFS8 group and NC+CoCl₂ group, and NAC reversed these increases (Fig. 4 J - O). These findings suggest that NDUFS8 knockdown may induce senescence in KGN cells via ROS-dependent pathway. Given anti-inflammatory, anti-oxidation and anti-aging effects of Ica, we hypothesized that Ica has great therapeutic potential in EMS. EMS mice were established and treated with Ica or Die via gavage. Posttreatment comparisons of ectopic lesions were exhibited in Fig.  5 A. Compared with CON group, EMS group presented significantly greater abdominal adhesion scores (Fig.  5 B). However, abdominal adhesion scores, volumes and quantities of ectopic lesions were significantly lower in Ica group and Die group than EMS group, with comparable efficacy between these two treatments (Fig.  5 B - E). These results indicated that Ica can alleviate EMS progression. Fig. 5 Therapeutic effect analysis of Ica on EMS mice and network pharmacology analysis of the mechanism of Ica in treating EMS. ( A ) Abdominal dissection images of mice in CON, EMS, Ica, and Die groups ( n  = 6/group). Red arrows indicate ectopic lesions. ( B ) Comparison of degrees of abdominal adhesion among CON, EMS, Ica, and Die groups ( n  = 6/group). ( C ) Histomorphological observation of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( D ) Quantitative comparison of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( E ) Comparison of volumes of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( F ) Comparison of foetal counts among CON, EMS, Ica, and Die groups ( n  = 6/group). ( G ) Quantification of foetal counts in (F). ( H ) Venn diagram of intersection targets between Ica and EMS. ( I ) “Ica-EMS-targets” network diagram. ( J ) PPI network of intersection targets. ( K ) Network diagram of core targets of Ica in treating EMS. ( L ) GO enrichment analysis of intersection targets. ( M ) KEGG enrichment analysis of intersection targets. All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA(between three and four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Therapeutic effect analysis of Ica on EMS mice and network pharmacology analysis of the mechanism of Ica in treating EMS. ( A ) Abdominal dissection images of mice in CON, EMS, Ica, and Die groups ( n  = 6/group). Red arrows indicate ectopic lesions. ( B ) Comparison of degrees of abdominal adhesion among CON, EMS, Ica, and Die groups ( n  = 6/group). ( C ) Histomorphological observation of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( D ) Quantitative comparison of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( E ) Comparison of volumes of ectopic lesions among CON, EMS, Ica, and Die groups ( n  = 6/group). ( F ) Comparison of foetal counts among CON, EMS, Ica, and Die groups ( n  = 6/group). ( G ) Quantification of foetal counts in (F). ( H ) Venn diagram of intersection targets between Ica and EMS. ( I ) “Ica-EMS-targets” network diagram. ( J ) PPI network of intersection targets. ( K ) Network diagram of core targets of Ica in treating EMS. ( L ) GO enrichment analysis of intersection targets. ( M ) KEGG enrichment analysis of intersection targets. All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA(between three and four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant To assess reproductive capacity, vaginal smears were employed to choose female mice in oestrus to conduct mating trials (Fig. S4A) and to confirm successful mating (Fig. S4B). At GD14, pregnant uteri were dissected, and foetal counts were recorded (Fig.  5 F and G). Compared with CON, EMS mice presented significantly lower foetal counts, which were increased significantly after Ica treatment. However, Die did not exhibit improvement effects. Given significant therapeutic effects of Ica in EMS mice, we employed network pharmacology to explore specific mechanism. Targets of Ica and EMS were identified through database predictions, and intersection targets were screened by Venn diagram (Fig.  5 H). These intersection targets were imported into Cytoscape and STRING database to generate “Ica-EMS-targets” network diagram (Fig.  5 I) and PPI network with 50 nodes and 304 edges (Fig.  5 J), respectively. After PPI network was imported into Cytoscape, core targets of Ica in treating EMS were identified (Fig.  5 K). To clarify primary mechanisms underlying therapeutic effects of Ica on EMS, GO and KEGG enrichment analyses were performed on intersection targets. GO enrichment analysis enriched in “ATP binding”, “oxidoreductase activity” and “oestradiol 17-β-dehydrogenase activity” (Fig.  5 L). Furthermore, KEGG enrichment analysis enriched in “ROS”, “oestrogen signalling pathway”, “ovarian steroid hormone biosynthesis” and “progesterone-mediated oocyte maturation” (Fig.  5 M). Thus, therapeutic effects of Ica in EMS may be associated with regulation of ROS levels, energy metabolism, oestrogen metabolism and ovarian function. Based on results of network pharmacology analysis, we hypothesized that Ica may inhibit GC senescence in EMS by regulating ROS levels and energy metabolism. Furthermore, NDUFS8 may induce GC senescence via ROS-dependent pathway, and ROS/P53/P21 axis is pivotal in promoting cellular senescence [ 32 ]. Thus, we hypothesized that Ica may target NDUFS8/ROS/P53/P21 axis to inhibit GC senescence in EMS. To validate these hypotheses, KGN cells were treated with 10% CON-FF or EMS-FF for 48 h, followed by Ica intervention in EMS-FF group. To determine optimal concentration and duration of Ica, CCK-8 assays were performed. The results revealed that concentration of 16 µM and duration of 48 h were optimal (Fig. S5). Compared with CON-FF group, NDUFS8 mRNA and protein expression levels in EMS-FF group were significantly decreased, and Ica intervention reversed this situation (Fig.  6 A - C). ROS levels, percentages of SA-β-gal-positive cells and proportions of G1-phase cells in EMS-FF group were significantly increased, which were reversed by Ica (Fig.  6 D - I). Similarly, P21 and P53 mRNA and protein expression levels in EMS-FF group were significantly upregulated, which were significantly reduced after Ica intervention (Fig.  6 J - O). In summary, Ica may inhibit KGN cell senescence by regulating NDUFS8/ROS/P53/P21 axis. Fig. 6 Ica may inhibit KGN cell senescence by regulating NDUFS8/ROS/P53/P21 axis. ( A ) Comparison of NDUFS8 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( B ) Comparison of NDUFS8 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( C ) Quantification of NDUFS8 protein expression levels in (B). ( D ) Comparison of ROS levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( L ) Comparison of P53 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( M ) Comparison of P21 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (M). ( O ) Quantification of P53 protein expression levels in (L). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA (between three groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Ica may inhibit KGN cell senescence by regulating NDUFS8/ROS/P53/P21 axis. ( A ) Comparison of NDUFS8 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( B ) Comparison of NDUFS8 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( C ) Quantification of NDUFS8 protein expression levels in (B). ( D ) Comparison of ROS levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). Scale bar = 100 μm. ( E ) Quantification of ROS levels in (D). ( F ) Quantitative comparison of SA-β-gal-positive cells among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). Scale bar = 100 μm. ( G ) Quantification of SA-β-gal-positive cells in (F). ( H ) Comparison of cell cycle distributions among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( I ) Quantification of cell cycle distributions in (H). ( J ) Comparison of P21 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( K ) Comparison of P53 mRNA expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( L ) Comparison of P53 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( M ) Comparison of P21 protein expression levels among CON-FF, EMS-FF and EMS-FF + Ica groups ( n  = 3/group). ( N ) Quantification of P21 protein expression levels in (M). ( O ) Quantification of P53 protein expression levels in (L). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA (between three groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant To validate whether Ica can inhibit GC senescence and improve ovarian function in EMS, we established EMS mice and treated with Ica or Die. Compared with EMS group, atretic follicle rates in CON group and Ica group were significantly lower (Fig.  7 A and B). Besides, CON group and Ica group presented significantly higher AMH levels and lower FSH and LH levels than EMS group (Fig.  7 C - E). IHC analysis of ovarian tissues revealed that NDUFS8 expression levels were significantly greater in CON group and Ica group than EMS group, whereas P21 and P53 expression levels were significantly lower (Fig.  7 F - K). Moreover, there were no significant differences in between Die group and EMS group (Fig.  7 A - K). The detected results of mouse primary GCs revealed significantly greater ROS levels, percentages of SA-β-gal-positive cells and proportions of G1-phase cells in EMS group than CON group and Ica group (Fig.  7 L -Q). Notably, therapeutic effects of Ica were obviously better than Die. In summary, Ica may inhibit GC senescence and improve ovarian function in EMS mice by regulating NDUFS8/ROS/P53/P21 axis. Fig. 7 Ica may inhibit GC senescence and improve ovarian function in EMS mice by regulating NDUFS8/ROS/P53/P21 axis. ( A ) H&E staining of ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Red arrows indicate atretic follicles. Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( B ) Quantification of atretic follicles in (A). ( C ) Comparison of serum AMH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( D ) Comparison of serum FSH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( E ) Comparison of serum LH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( F ) IHC NDUFS8 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( G ) Quantification of NDUFS8 expression levels in (F). ( H ) IHC P21 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( I ) Quantification of P21 expression levels in (H). ( J ) IHC P53 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( K ) Quantification of P53 expression levels in (J). ( L ) Comparison of ROS levels in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). Scale bar = 100 μm. ( M ) Quantification of ROS levels in (L). ( N ) Quantitative comparison of SA-β-gal-positive cells in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). Scale bar = 100 μm. ( O ) Quantification of SA-β-gal-positive cells in (N). ( P ) Comparison of cell cycle distributions in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). ( Q ) Quantification of cell cycle distributions in (P). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA (between four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant Ica may inhibit GC senescence and improve ovarian function in EMS mice by regulating NDUFS8/ROS/P53/P21 axis. ( A ) H&E staining of ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Red arrows indicate atretic follicles. Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( B ) Quantification of atretic follicles in (A). ( C ) Comparison of serum AMH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( D ) Comparison of serum FSH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( E ) Comparison of serum LH levels among CON, EMS, Ica, and Die groups ( n  = 6/group). ( F ) IHC NDUFS8 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( G ) Quantification of NDUFS8 expression levels in (F). ( H ) IHC P21 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( I ) Quantification of P21 expression levels in (H). ( J ) IHC P53 expression levels in ovarian tissues from CON, EMS, Ica, and Die groups ( n  = 6/group). Original image scale bar = 200 μm. Enlarged image scale bar = 50 μm. ( K ) Quantification of P53 expression levels in (J). ( L ) Comparison of ROS levels in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). Scale bar = 100 μm. ( M ) Quantification of ROS levels in (L). ( N ) Quantitative comparison of SA-β-gal-positive cells in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). Scale bar = 100 μm. ( O ) Quantification of SA-β-gal-positive cells in (N). ( P ) Comparison of cell cycle distributions in mouse primary GCs among CON, EMS, Ica, and Die groups ( n  = 3/group). ( Q ) Quantification of cell cycle distributions in (P). All data are expressed as Mean ± SD and were analysed by Unpaired Student’s t test (between two groups) or One-way ANOVA (between four groups). * P  < 0.05, ** P  < 0.01, *** P  < 0.001, **** P  < 0.0001, and ns, not significant

FF samples were obtained from the Reproductive Medicine Center of the First Affiliated Hospital of Naval Medical University between February 1, 2023 and July 30, 2023. The study cohort included 8 patients with ovarian EMS-associated infertility who were receiving assisted reproductive technology (ART) therapy and 10 normal control (CON) females who were receiving ART therapy due to male spouse infertility. Baseline characteristics of study participants are in Table S1 . FF was collected during the oocyte retrieval procedures. FF and GCs were isolated at the hospital’s central laboratory. The detailed inclusion and exclusion criteria for both groups are provided in the Supplementary Material. This study was approved by the Ethics Committee of the First Affiliated Hospital of Naval Medical University (approval no. CHEC2019-100) and conducted in accordance with the Declaration of Helsinki. FF was collected in sterile 50-ml centrifuge tubes and centrifuged at 2000 rpm for 10 min at 4 °C. The supernatant was stored at -80 °C. The pellet was resuspended in 5 ml of phosphate-buffered saline (PBS) (Servicebio, Wuhan, China), transferred to a 15-ml tube, and gently layered with an equal volume of human peripheral blood lymphocyte separation solution (Biosharp, Beijing, China). After centrifugation at 2000 rpm for 20 min at 4 °C, the white cloudy interphase layer was transferred to a new 15-ml tube. An equal volume of 1× red blood cell lysis buffer (Yuanye, Shanghai, China) was added, followed by gentle pipetting. Following centrifugation at 2000 rpm for 10 min at 4 °C, the pelleted GCs were resuspended in 1 ml of TRIzol reagent (Invitrogen, USA) and stored at -80 °C prior to RNA sequencing. Total RNA was extracted using TRIzol reagent according to the manufacturer’s protocol. mRNA libraries were constructed and subjected to quality-control procedures using a 2100 Bioanalyzer (Agilent, Beijing, China). Qualified libraries were sequenced on the Illumina HiSeq 2500 platform (Illumina, San Diego, CA, USA). Significantly differentially expressed genes (DEGs) were identified and subjected to Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. KGN, an immortalized cell line derived from human ovarian GC tumors, is currently the most commonly used in vitro model in the field of reproductive endocrinology to study human ovarian GCs. Compared with other common ovarian tumor cell lines, its core advantage is that it retains the hormone responsiveness of GCs, such as high basal expression of functional follicle-stimulating hormone (FSH) receptors and aromatase [ 25 ]. The KGN cell line used in this study (Qui Cell, Shanghai, China) was cultured in DMEM/F12 (Qui Cell) supplemented with 10% foetal bovine serum (FBS) (Procell, Wuhan, China). FF pools were created by homogenizing equal volumes from six individual frozen FF samples. FF was diluted in DMEM/F12 to final concentrations. Ica powder (MCE, USA) was dissolved in dimethyl sulfoxide (DMSO) (Sigma, USA) and diluted in DMEM/F12 to final concentrations. Cell proliferation was assessed using a Cell Counting Kit-8 (CCK-8) (Qui Cell) according to the manufacturer’s instructions. Intracellular ATP levels were quantified using an ATP assay kit (Beyotime, Shanghai, China) following the manufacturer’s protocol. KGN cells were seeded (1 × 10⁴ cells/well) into Seahorse XFe96 cell culture microplates (Agilent, USA) and incubated overnight at 37 °C for adhesion. Before analysis, the cells were washed and incubated in Seahorse XF Base Medium (Agilent, USA) for 1 h at 37 °C (non-CO₂). The real-time oxygen consumption rate (OCR) was measured using the Seahorse XFe96 Analyser (Agilent) and Seahorse XF Cell Mito Stress Test Kit (Agilent) according to the manufacturer’s protocol. Cell viability was assessed using PrestoBlue reagent (Thermo Fisher, USA), and OCR values were normalized to viable cell counts. Intracellular ROS levels were measured using an ROS assay kit (Solarbio, Beijing, China). Senescence-associated β-galactosidase (SA-β-gal) activity was assessed with an SA-β-gal staining kit (Beyotime, Shanghai, China). The cell cycle distribution was analysed using a cell cycle detection kit (Beyotime). All procedures followed manufacturers’ protocols. The mRNA expression levels of NADH: ubiquinone oxidoreductase core subunit S8 (NDUFS8), CDKN1A, and TP53 were quantified by RT-qPCR. The primer sequences are provided in Table S2. RNA extraction, reverse transcription, and qPCR were performed using the SteadyPure Quick RNA Extraction Kit (Accurate Biology, Hunan, China), Evo M-MLV RT Premix Kit (Accurate Biology), and SYBR Green Pro Taq HS Premix Kit (Accurate Biology), respectively, according to manufacturers’ instructions. Cells were lysed in RIPA buffer (Epizyme, Shanghai, China) containing 1% phosphatase inhibitor (Epizyme) and 1% protease inhibitor (Epizyme). Protein concentrations were determined using a BCA assay kit (Epizyme). Equal protein amounts were separated by SDS-PAGE and transferred to PVDF membranes (Merck, Germany). After blocking with 5% nonfat milk for 2 h, membranes were incubated overnight at 4 °C with the following primary antibodies: anti-NDUFS8 (Abcam, UK; Cat# ab170936), anti-P21 (ABclonal, Wuhan, China; Cat# A19094), anti-P53 (Proteintech, Wuhan, China; Cat# 60283-2-lg), and anti-β-actin (Huabio, Hangzhou, China; Cat# JF53-10). HRP-conjugated secondary antibodies (goat anti-rabbit IgG-HRP, Abmart Cat# M21002 ; goat anti-mouse IgG-HRP, Abmart Cat# M21001 ) were applied for 1 h at room temperature. Finally, protein bands were visualized using an enhanced chemiluminescence detection system. NDUFS8-knockdown KGN cells were generated by lentiviral shRNA transduction (shRNA sequences in Table S3; virus synthesized by Genomeditech, Shanghai, China). The lentivirus was diluted in DMEM/F12 to achieve final multiplicity of infection (MOI). The transduction efficiency was evaluated via fluorescence microscopy (Olympus, Japan). KGN cells were transduced at optimal MOI and knockdown efficiency was validated by RT-qPCR and WB. The canonical SMILES of Ica was retrieved from PubChem and input into SwissTargetPrediction (species: Homo sapiens ) to identify Ica targets. EMS-associated targets were obtained from GeneCards using “endometriosis” as keyword. Intersection targets were identified by Venn analysis and uploaded to STRING for protein-protein interaction (PPI) analysis (confidence score > 0.4; species: Homo sapiens ). PPI network was analysed using CytoScape 3.7.2 with Centiscape 2.2 plugin to identify core targets based on degree, closeness, and betweenness centrality. Intersection targets were analysed via DAVID for GO and KEGG pathway enrichment (species: Homo sapiens , identifier: official gene symbol). Six-week-old female C57BL/6 mice (Vital River, Beijing, China) were housed under specific pathogen-free conditions with ad libitum access to food and water. Donor mice received subcutaneous injections of β-oestradiol benzoate (2 µg/mouse in sesame oil; Yuanye, Shanghai, China) three times at 48-h intervals. After one week, uterine and adipose tissues were harvested, minced into fragments 1 mm³ in size in PBS (Servicebio, Wuhan, China), and intraperitoneally injected into recipient mice (donor-to-recipient ratio, 1:2) 0.5 cm above the urethral orifice. CON mice received adipose tissue fragments only. Treatments were as follows: CON group and EMS group: sesame oil (0.2 ml/day, gavage); Ica group: Ica in sesame oil (40 mg/kg/day, 0.2 ml/day, gavage; MCE, USA); Dienogest (Die) group: Die in sesame oil (300 µg/kg/day, 0.2 ml/day, gavage; MCE, USA). Treatments were administered for 21 consecutive days starting one week postmodelling. All animal experiments were approved by the Experimental Animal Ethics Review Committee of the First Affiliated Hospital of Naval Medical University (approval no. CHEC(AE) 2022-002). Ten-week-old male C57BL/6 mice (Vital River) were acclimated for one week. After treatment completion and 1 week washout, females were paired with males (2♀:1♂) during oestrus. Vaginal smears of sperm positive indicated gestational day (GD) 1. At GD14, dissect abdomens of mice to obtain ectopic lesions, ovaries and uteruses. Obtained ovaries were minced in DMEM/F12 and digested with 0.1% collagenase IV (Nordmark, Germany) in TESCA buffer (Solarbio, Beijing) at 37 °C and 60 rpm for 30 min. After digestion was terminated with DMEM/F12, the mixture was filtered through a 70-µm strainer (Corning, USA) and centrifuged at 4 °C and 2000 rpm for 10 min. Obtained pellet was mouse primary GCs. Serum was isolated from orbital blood (centrifugation: 3000 rpm, 15 min, 4 °C). Hormone levels were quantified using following ELISA kits according to protocols of Jiangsu Meimian Industrial Co., Ltd: anti-Müllerian hormone (AMH) (Meimian, MM-0470M1), FSH (Meimian, MM-0163M1), and luteinizing hormone (LH) (Meimian, MM-0636M1). Ovaries were fixed in 4% paraformaldehyde (Biolight, Shanghai), embedded in paraffin and sectioned. Sections were stained with H&E standard protocols. Morphology was analysed under an Olympus BX53 microscope (Olympus, Japan). Ovaries were fixed in 4% paraformaldehyde (Biolight, Shanghai), embedded in paraffin and sectioned. After deparaffinization and rehydration, endogenous peroxidase was blocked with 3% H₂O₂ (Anjet High-Tech, Shandong, China) for 30 min. Sections were incubated overnight at 4°C with the following primary antibodies: anti-NDUFS8 (Abcam, UK; Cat# ab170936), anti-p21 (ABclonal, China; Cat# A19094) and anti-p53 (Proteintech, China; Cat# 60283-2-lg). After being washed with PBS, sections were incubated with the following HRP-conjugated secondary antibodies at 37°C for 45 min: goat anti-rabbit IgG-HRP (Abmart, Cat# M21002 ) and goat anti-mouse IgG-HRP (Abmart, Cat# M21001 ). Sections were stained with 3,3’-diaminobenzidine tetrahydrochloride (Caiyou, Shanghai) for 6 min and counterstained with haematoxylin (Merck, Germany). Slides were dehydrated in ethanol, mounted with neutral resin (Aladdin, Shanghai), and imaged under an Olympus BX53 microscope (Olympus, Japan). All statistical analyses were performed using SPSS 21.0. Data visualization was generated with GraphPad Prism 9. Two-group comparisons of measurement data that conformed to normal distribution and homogeneous variances were performed with Unpaired Student’s t test. Otherwise, Mann-Whitney U test was employed. Multigroup comparisons of measurement data that conformed to normal distribution and homogeneous variances were performed with One-way ANOVA test. Otherwise, Kruskal-Wallis H test was employed. Count data were statistically analyzed by Fisher’s exact test. Statistical significance was defined as P  < 0.05.

In summary, we conclude the following (Fig.  8 ): ovarian GCs in patients with EMS-associated infertility may exhibit OXPHOS dysfunction and cellular senescence; NDUFS8 may induce senescence in KGN cells via ROS-dependent pathway; Ica may inhibit KGN cell senescence induced by EMS-FF, inhibit GC senescence and improve ovarian function and pregnancy outcomes in EMS mice by regulating NDUFS8/ROS/P53/P21 axis. Fig. 8 Mechanistic diagram of how Ica improves pregnancy outcomes in EMS by inhibiting ovarian GC senescence Mechanistic diagram of how Ica improves pregnancy outcomes in EMS by inhibiting ovarian GC senescence

EMS is an oestrogen-dependent chronic inflammatory disease characterized primarily by chronic pelvic pain and infertility [ 1 , 2 ]. Although the strong association between EMS and infertility has been well documented, the causal relationship between them remains incompletely defined. Growing evidence suggests that EMS-associated infertility is linked to diminished ovarian function and reduced oocyte quality, where dysfunction of GCs, the primary nutrient support system for oocytes, plays a critical role [ 5 ]. Ovarian GCs constitute the largest cell population and main functional cells within follicles, and follicular development is characterized by rapid growth and proliferation of GCs [ 33 ]. Oocyte activation and sustained growth depend on the nutritional and paracrine functions of surrounding GCs. Therefore, the functional integrity of GCs is a key factor in determining oocyte quality, ovarian function, and reproductive capacity in EMS patients. After determining research objectives, we performed RNA sequencing on GCs extracted from FF of normal women and patients with EMS-associated infertility. Enrichment analysis revealed that significant DEGs between the two groups were enriched in pathways related to ROS, OXPHOS, cellular senescence, and P53 signalling pathway. Notably, ROS accumulation is not only a key hallmark of cellular senescence, but can also further exacerbate the process of cellular senescence [ 34 ]. Furthermore, P53, a critical regulator of cell cycle and cellular senescence, can drive cellular senescence through P53/P21 tumour suppressor pathway [ 35 ]. Given that significant inhibition of OXPHOS pathway can lead to ROS accumulation [ 30 ], we propose that OXPHOS dysfunction and subsequent cellular senescence may constitute the pathological basis in GC dysfunction in EMS-associated infertility. We subsequently treated KGN cells in vitro with FF from the two groups for indirect validation. The results exhibited abnormal OXPHOS and aggravated cellular senescence phenotypes in EMS-FF-treated KGN cells. In summary, these findings revealed a novel possible mechanism underlying the decline in reproductive capacity of EMS-associated infertility. To explore the mechanism of OXPHOS dysfunction and cellular senescence, we intersected the published single-cell RNA sequencing dataset of human oocytes with our human GC RNA sequencing dataset. Notably, OXPHOS-related genes exhibited opposite expression patterns: they were significantly downregulated in GCs of EMS patients, but significantly upregulated in oocytes. We speculated that impaired OXPHOS function in GCs may lead to insufficient energy supply to oocytes, thereby triggering compensatory increases in OXPHOS levels within oocytes. Among these OXPHOS-related genes, NDUFS8, a core subunit of the iron-sulfur fragment in mitochondrial complex I [ 36 ], was the most significantly downregulated gene in GCs. Considering that both inhibition of the OXPHOS pathway and reduced NDUFS8 expression can induce ROS accumulation [ 30 , 31 ], we mechanistically linked NDUFS8 knockdown to ROS-driven cellular senescence in KGN cells by introducing shRNA lentivirus, pro-oxidants and antioxidants. Pro-oxidants exacerbated this phenotype, which was reversed by antioxidants, establishing an NDUFS8-induced ROS-dependent KGN cell senescence pathway. These results indicate that targeting NDUFS8/ROS axis may serve as a therapeutic strategy for EMS-associated infertility. Considering the remarkable effects of Ica on anti-tumour, anti-inflammatory, anti-oxidation and anti-aging, we hypothesized that Ica represents a promising therapeutic candidate for EMS and EMS-associated infertility. Network pharmacology predicted that the mechanism of Ica in treating EMS likely involves modulating energy metabolism, oestrogen metabolism, ovarian function, and ROS levels. In vivo, Ica treatment significantly suppressed ectopic lesions and improved reproductive capacity in EMS mice. Most importantly, Ica significantly improved ovarian function and inhibited GC senescence in EMS mice. Furthermore, in vivo experiments of mice and in vitro experiments of KGN cells verified the mechanism by which Ica may inhibit GC senescence through NDUFS8/ROS/P53/P21 axis. These results revealed great potential of Ica in treating EMS and EMS-associated infertility. Currently, peritoneal inflammation and ectopic lesions are main topics in EMS-related basic researches [ 37 ], which rarely targets pregnancy outcomes in EMS-associated infertility relatively. Existing studies of EMS-associated reproduction focus mainly on ovarian function and endometrial receptivity [ 38 , 39 ]. In ovarian function related researches, the emphases are on oocytes, GCs, and their intercellular communication [ 40 , 41 ]. Owing to ethical restrictions and stringent requirements, it is difficult to obtain human oocytes directly for research. Within researches related to GCs, researchers focus primarily on mitochondrial function and oxidative damage [ 42 – 44 ]. Therefore, building upon published findings and focusing on current research hotspots, we aimed to uncover novel mechanisms and therapeutic approaches for EMS-associated infertility. To the best of our knowledge, this is the first study to reveal the critical mechanisms of OXPHOS dysfunction and cellular senescence in ovarian GCs of EMS-associated infertility and to demonstrate the central role of NDUFS8/ROS axis in this process. Furthermore, this study is the first to confirm that Ica may inhibit GC senescence by regulating NDUFS8/ROS/P53/P21 axis, thereby improving ovarian function and pregnancy outcomes in EMS. These findings provide new theoretical bases and potential therapeutic strategies for EMS-associated infertility. However, certain limitations of this study need to be addressed. Firstly, mechanistic investigations in this study lacked sufficient support from a large number of clinical samples. Secondly, toxicological characteristics of Ica have not been studied to confirm whether it has toxic side effects on mice embryos, nor have its toxic doses and treatment durations been established. Finally, animal experiments lacked data on long term reproductive outcomes. Therefore, future research should involve more clinical samples for in-depth studies to confirm the pivotal role of NDUFS8/ROS/P53/P21 axis in the pathogenesis of EMS-associated infertility. Additionally, attention must be paid to biosafety characteristics of Ica to lay solid foundations for subsequent clinical trials. Undoubtedly, inhibiting GC senescence is expected to become a crucial therapeutic strategy for EMS-associated infertility.

Endometriosis (EMS), an oestrogen-driven chronic inflammatory disorder [ 1 ], which typically manifests as chronic pelvic pain and infertility, posing significant threats to the reproductive health of childbearing women [ 2 ]. A significant association exists between EMS and infertility, with approximately 40–50% of EMS patients exhibiting reproductive dysfunction [ 3 ]. However, the precise mechanisms linking EMS and infertility remain incompletely elucidated [ 4 ]. Studies indicate that EMS-associated infertility is closely linked to diminished ovarian function and compromised oocyte quality [ 5 ], potentially attributable to the aberrant function of granulosa cells (GCs), which constitute the primary nutrient supply system surrounding oocytes. Owing to their specialized structure and function, oocytes possess a limited capacity for glucose uptake and direct utilization, relying almost entirely on surrounding GCs for energy provision [ 6 ]. In addition, GCs establish essential microenvironment for oocyte maturation through autocrine and paracrine signalling involving steroid hormones and growth factors [ 7 ]. Consequently, the functional state of GCs is a critical determinant of oocyte quality, ovarian function, and reproductive capacity. Cellular senescence refers to a state of stable cell cycle arrest accompanied by gradual loss of proliferative capacity [ 8 ]. Recent researches increasingly implicate cellular senescence in the pathogenesis of various diseases, including osteoarthritis, atherosclerosis, and neoplasms [ 9 – 11 ]. Furthermore, its impact on female fertility warrants significant attention, particularly in infertility-associated conditions such as EMS, polycystic ovary syndrome, and ovarian insufficiency [ 12 , 13 ]. Cellular senescence may contribute to infertility by reducing oocyte quantity and quality and by inducing uterine and placental dysfunction [ 14 – 16 ]. Our previous studies revealed iron overload in the follicular fluid (FF) of patients with EMS-associated infertility [ 17 ], which may trigger excessive oxidative stress in ovarian GCs, leading to elevated reactive oxygen species (ROS) levels concomitant with decreased mitochondrial membrane potential, reduced adenosine triphosphate (ATP) production [ 18 ]. Notably, ROS is a well-established hallmark of cellular senescence [ 19 ]. Therefore, it is imperative to investigate whether excessive ROS levels induce GC senescence, thereby impairing nutritional support to oocytes, compromising oocyte quality and ovarian function, and ultimately contributing to EMS-associated infertility. Icaritin (Ica), a bioactive isoprenylated flavonoid derivative, is produced by the hydrolysis of icariin derived from Epimedium species [ 20 ]. Pharmacological studies have demonstrated that Ica has diverse biological activities, including anti-tumour, anti-inflammatory, anti-oxidation and anti-aging effects [ 21 , 22 ]. The main mechanisms underlying its anti-inflammatory effects include reducing the release of inflammatory cytokines, suppressing the activation of the NF-κB signaling pathway, inhibiting STAT and MAPK-mediated signaling pathways [ 21 ]. Besides, Ica contains a polyphenolic hydroxyl group, which can scavenge free radicals, enhance superoxide dismutase activity and eliminate excessive ROS by regulating the Nrf2/ARE/HO-1 signaling pathway, thereby reducing lipid peroxidation and mitochondrial dysfunction, ultimately exerting effects including anti-oxidative stress, delaying aging and protecting nerves and cells [ 23 ]. Although traditionally considered a benign condition, EMS displays multiple features resembling invasive cancer and intimate associations with chronic inflammation [ 24 ]. Meanwhile, excessive oxidative stress and senescence in ovarian GCs may lead to decreased ovarian function, which is closely related to EMS-associated infertility. Considering remarkable effects of Ica on anti-tumour, anti-inflammatory, anti-oxidation and anti-aging, we hypothesized that Ica represents a promising therapeutic candidate for EMS and EMS-associated infertility. This study aimed to investigate pathogenic mechanisms of EMS-associated infertility from the perspective of GC senescence and to explore therapeutic effects and underlying mechanisms of Ica on ovarian function and reproductive capacity in EMS.

Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1