Ginsenoside Rb1 alleviates uterine aging via AMPK-mediated mTOR inactivation.

Uterine aging leads to decreased endometrial receptivity, characterized by EnSC senescence, oxidative stress, and apoptosis. 23 Endometrial tissues were collected from young (<35 years) and aging patients (≥35 years) (Supplementary Tables S1 and S2 ). EnSCs were successfully isolated from fresh endometrial tissue obtained from women of reproductive age. To verify the purity of the isolated cells, we performed immunofluorescence staining using cytokeratin-19 (CK19) as an epithelial marker and vimentin as a stromal marker. The majority of the isolated cells exhibited strong vimentin expression, confirming their stromal cell identity. In contrast, epithelial cells isolated in parallel displayed intense CK19 staining with minimal vimentin expression, clearly distinguishing EnSCs from epithelial cells (Fig. 1a and Supplementary Fig. 1a ). These results demonstrated the successful separation and identification of EnSCs from human endometrial tissue samples. To optimize the treatment dose and time, we evaluated the safety and effectiveness of Rb1 under different conditions (Supplementary Fig. 1b–e ). EnSCs from aging donors were treated with different doses of Rb1 for different durations, and no obvious toxicity was observed (Supplementary Fig. 1c ). CDKN1A (p21), CDKN2A (p16), and APOE are known aging markers of multiple cell types, including stromal cells, 13 , 24 whose expression is significantly elevated in the EnSCs of aging women according to single-cell RNA sequencing (scRNA-seq) data (Supplementary Fig. 1d ). Among all the treatments, treatment with 20 μmol/L Rb1 for 12 h resulted in the most significant decrease in CDKN1A expression (Supplementary Fig. 1e ) and was selected for subsequent experiments. Fig. 1 Rb1 alleviates aging-related phenotypes in human EnSCs. a Representative immunofluorescence images of primary EnSCs with vimentin expression (scale bar, 100 µm). b The levels of the oxidative stress markers MDA, 8-OHdG, and LDH in EnSCs in the three groups. c The ROS levels in the EnSCs in the three groups were detected by flow cytometry, and the average ROS fluorescence intensity was quantitatively analyzed. d SA-β-gal activity in EnSCs among the three groups. e The expression of CDKN1A , CDKN2A , and APOE mRNA in EnSCs among the three groups was measured by qRT‒PCR. Human GAPDH mRNA expression levels were used as an endogenous reference. The gene expression level was normalized to that of GAPDH and is presented as the fold change (2 −∆∆Ct ) above that of the young group. f EnSC apoptosis in the three groups was detected by flow cytometry analysis and quantitative analysis of apoptosis ratios. g Apoptosis-related protein expression in the EnSCs in the three groups was analyzed by western blotting. β-actin was used as an endogenous control. h Quantification of cleaved caspase-3 (C-CAS3), cleaved caspase-9 (C-CAS9) and BCL-2/BAX protein expression. The data are expressed as the means ± SDs. ( n  = 3 independent biological replicates) * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, EnSCs in the young female group; Aging, EnSCs in the aging female group; Aging + Rb1, EnSCs in the aging female group treated with Rb1; ROS, reactive oxygen species) Rb1 alleviates aging-related phenotypes in human EnSCs. a Representative immunofluorescence images of primary EnSCs with vimentin expression (scale bar, 100 µm). b The levels of the oxidative stress markers MDA, 8-OHdG, and LDH in EnSCs in the three groups. c The ROS levels in the EnSCs in the three groups were detected by flow cytometry, and the average ROS fluorescence intensity was quantitatively analyzed. d SA-β-gal activity in EnSCs among the three groups. e The expression of CDKN1A , CDKN2A , and APOE mRNA in EnSCs among the three groups was measured by qRT‒PCR. Human GAPDH mRNA expression levels were used as an endogenous reference. The gene expression level was normalized to that of GAPDH and is presented as the fold change (2 −∆∆Ct ) above that of the young group. f EnSC apoptosis in the three groups was detected by flow cytometry analysis and quantitative analysis of apoptosis ratios. g Apoptosis-related protein expression in the EnSCs in the three groups was analyzed by western blotting. β-actin was used as an endogenous control. h Quantification of cleaved caspase-3 (C-CAS3), cleaved caspase-9 (C-CAS9) and BCL-2/BAX protein expression. The data are expressed as the means ± SDs. ( n  = 3 independent biological replicates) * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, EnSCs in the young female group; Aging, EnSCs in the aging female group; Aging + Rb1, EnSCs in the aging female group treated with Rb1; ROS, reactive oxygen species) To thoroughly evaluate the antiaging effects of Rb1, we first examined changes in oxidative stress, which is one of the hallmarks of aging. Malondialdehyde (MDA), 8-hydroxy-2’-deoxyguanosine (8-OHdG), and lactate dehydrogenase (LDH) have been widely used to quantify oxidative stress. 25 , 26 The increases in the levels of MDA, 8-OHdG, and LDH in aging EnSCs were significantly reversed by Rb1 (Fig. 1b ). The average fluorescence intensity of ROS was greater in the aging group than in the young group and was downregulated by Rb1, revealing the alleviation of aging-related oxidative injury (Fig. 1c ). We also observed senescent phenotypes in aging EnSCs, with increased levels of senescence-associated β-galactosidase (SA-β-gal) activity (Fig. 1d ) and higher gene expression levels of CDKN1A , CDKN2A , and APOE (Fig. 1e ). Rb1 could also partially reverse these senescence-associated features (Fig. 1d, e ). The percentage of apoptotic cells markedly increased in the aging group but decreased upon Rb1 treatment (Fig. 1f ). The changes in the expression of apoptosis-related proteins were similar (Fig. 1g, h ). These findings confirmed that Rb1 could alleviate aging-related damage in EnSCs. To evaluate the effects of Rb1 on reproductive outcomes and endometrial function in vivo, we compared young, aging, and aging + Rb1-treated mice. Compared with young control mice, reproductive aging mice, a stage characterized by an age-associated decline in uterine and reproductive function, presented a reduced litter size, whereas Rb1 treatment significantly increased the number of live-born pups (Fig. 2a–c ). Consistently, blue dye assays revealed fewer implantation sites in aging mice, whereas Rb1 administration increased the number of implantation sites (Fig. 2b–d ). Fig. 2 Rb1 improves reproductive outcomes and endometrial receptivity and reduces the expression of aging-associated senescence markers in mice. a Representative images of litters from the young, aged, and aged + Rb1 groups, showing an increased number of pups after Rb1 treatment. b Representative implantation sites visualized by blue dye staining in each group. c , d Quantification of the number of live-born pups and implantation sites in the three groups. e , f Immunofluorescence staining of HAND2 and LIF, two key markers of endometrial receptivity, in uterine tissues (scale bar, 50 µm). Red: target proteins; blue: DAPI. g , h Immunofluorescence staining of the senescence markers p16 and p21, which revealed increased expression in aging mice and decreased expression after Rb1 treatment (scale bar, 50 µm). i SA-β-gal staining of uterine tissues showing increased senescence in aging mice, which was attenuated by Rb1. j , k mRNA levels of the decidualization markers Prl and Igfbp1 in stromal cells from young, aging, and aging + Rb1 groups after in vitro decidualization induction. Data are shown as the means ± SDs ( n  = 3 independent biological replicates); * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, young murine uteri ( n  = 6); Aging, aging murine uteri ( n  = 6); Aging + Rb1, aging murine uteri treated with Rb1 ( n  = 6)) Rb1 improves reproductive outcomes and endometrial receptivity and reduces the expression of aging-associated senescence markers in mice. a Representative images of litters from the young, aged, and aged + Rb1 groups, showing an increased number of pups after Rb1 treatment. b Representative implantation sites visualized by blue dye staining in each group. c , d Quantification of the number of live-born pups and implantation sites in the three groups. e , f Immunofluorescence staining of HAND2 and LIF, two key markers of endometrial receptivity, in uterine tissues (scale bar, 50 µm). Red: target proteins; blue: DAPI. g , h Immunofluorescence staining of the senescence markers p16 and p21, which revealed increased expression in aging mice and decreased expression after Rb1 treatment (scale bar, 50 µm). i SA-β-gal staining of uterine tissues showing increased senescence in aging mice, which was attenuated by Rb1. j , k mRNA levels of the decidualization markers Prl and Igfbp1 in stromal cells from young, aging, and aging + Rb1 groups after in vitro decidualization induction. Data are shown as the means ± SDs ( n  = 3 independent biological replicates); * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, young murine uteri ( n  = 6); Aging, aging murine uteri ( n  = 6); Aging + Rb1, aging murine uteri treated with Rb1 ( n  = 6)) Immunofluorescence staining demonstrated markedly reduced expression of the receptivity markers HAND2 and LIF in aging uteri, which was restored by Rb1 (Fig. 2e, f ). Aging mice also displayed increased levels of the senescence markers p16 and p21 and increased SA-β-Gal activity, all of which were attenuated following Rb1 treatment (Fig. 2g–i ). Furthermore, in vitro decidualization assays revealed that the induction of Prl and Igfbp1 expression was significantly impaired in stromal cells from aging mice, whereas Rb1 treatment partially rescued their expression (Fig. 2j, k ). Together, these results indicate that Rb1 improves reproductive capacity, enhances endometrial receptivity, and alleviates uterine aging by reducing senescence and restoring decidualization capacity. Next, we explored the pathology of uterine aging and the antiaging effect of Rb1 using a mouse model. In vivo, compared with control mice, aged mice subjected to daily intraperitoneal injection of Rb1 for two weeks exhibited no significant changes in serum biochemical parameters, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), urea (UREA), and creatinine (CREA). Moreover, histological examination of liver and kidney tissues revealed no apparent pathological abnormalities (Supplementary Fig. 2 ). These findings collectively indicate that Rb1 treatment did not induce evident cytotoxicity or systemic toxicity in our experimental settings. The morphology and structure of murine endometria were not significantly altered by aging or Rb1 treatment (Fig. 3a ). However, we observed increased apoptosis in the endometrium of the murine uterus in the aging group compared with that in the young group, which could be eliminated by Rb1 (Fig. 3b ). Consistent with those in human EnSCs, the expression of aging markers such as Cdkn1a , Cdkn2a , and Apoe was upregulated in aging murine uteri (Fig. 3c and Supplementary Fig. 3a ). The gene expression of SASP factors such as Ccl2 , Il6 , and Hgf was also increased in aging murine uteri (Fig. 3d ), whose upregulation was confirmed using scRNA-seq data from human endometria (Supplementary Fig. 3b ). Upon Rb1 treatment, aging-related gene expression was effectively reduced, indicating a potential antiaging effect of Rb1 (Fig. 3c, d ). Fig. 3 Rb1 partially reverses aging-related phenotypes and proteomic changes in the murine uterus. a Representative H&E staining images showing the morphology of the endometrium in murine uteri among the three groups (scale bar, 50 µm). b Representative TUNEL assay images of apoptosis in endometrial cells of murine uteri (scale bar, 50 µm). c The expression levels of Cdkn1a , Cdkn2a , and Apoe mRNA in the murine uterus in the three groups were measured by qRT‒PCR. d mRNA expression of the proinflammatory SASP factors Ccl2 , Il6 , and Hgf in the murine uteri of the three groups was measured by qRT‒PCR. The mouse Gapdh mRNA expression levels were used as an endogenous reference. The gene expression level was normalized to that of Gapdh and is presented as the fold change (2 −∆∆Ct ) above that of the young group. e , f PCA plots of the distributions of the three groups based on the top 3000 and top 500 variably expressed proteins, respectively. g , h Heatmaps and clustering analysis results display the correlations between all fifteen samples based on the top 3000 and top 500 variably expressed proteins, respectively. i , j Volcano plots showing the differentially expressed proteins (DEPs) between the Aging group and the Young group and between the Aging + Rb1 group and the Aging group. Red represents the upregulated DEPs, and blue represents the downregulated DEPs (|log2FC| > 0.5 and BH adjusted p value < 0.25). k , l GSEA showing positive enrichment of the downregulated DEPs (NES = 3.06; FDR = 0) and negative enrichment of the upregulated DEPs (NES = -2.87; FDR = 0) in aging murine uteri compared with young murine uteri. The data are expressed as the means ± SDs. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, young murine uteri ( n  = 5); Aging, aging murine uteri ( n  = 5); Aging + Rb1, aging murine uteri treated with Rb1 ( n  = 5); FC fold change, NES normalized enrichment score, FDR false discovery rate, DEPs downregulated/upregulated proteins in aging murine uteri treated with Rb1 compared with those in aging murine uteri) Rb1 partially reverses aging-related phenotypes and proteomic changes in the murine uterus. a Representative H&E staining images showing the morphology of the endometrium in murine uteri among the three groups (scale bar, 50 µm). b Representative TUNEL assay images of apoptosis in endometrial cells of murine uteri (scale bar, 50 µm). c The expression levels of Cdkn1a , Cdkn2a , and Apoe mRNA in the murine uterus in the three groups were measured by qRT‒PCR. d mRNA expression of the proinflammatory SASP factors Ccl2 , Il6 , and Hgf in the murine uteri of the three groups was measured by qRT‒PCR. The mouse Gapdh mRNA expression levels were used as an endogenous reference. The gene expression level was normalized to that of Gapdh and is presented as the fold change (2 −∆∆Ct ) above that of the young group. e , f PCA plots of the distributions of the three groups based on the top 3000 and top 500 variably expressed proteins, respectively. g , h Heatmaps and clustering analysis results display the correlations between all fifteen samples based on the top 3000 and top 500 variably expressed proteins, respectively. i , j Volcano plots showing the differentially expressed proteins (DEPs) between the Aging group and the Young group and between the Aging + Rb1 group and the Aging group. Red represents the upregulated DEPs, and blue represents the downregulated DEPs (|log2FC| > 0.5 and BH adjusted p value < 0.25). k , l GSEA showing positive enrichment of the downregulated DEPs (NES = 3.06; FDR = 0) and negative enrichment of the upregulated DEPs (NES = -2.87; FDR = 0) in aging murine uteri compared with young murine uteri. The data are expressed as the means ± SDs. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1, ginsenoside Rb1; Young, young murine uteri ( n  = 5); Aging, aging murine uteri ( n  = 5); Aging + Rb1, aging murine uteri treated with Rb1 ( n  = 5); FC fold change, NES normalized enrichment score, FDR false discovery rate, DEPs downregulated/upregulated proteins in aging murine uteri treated with Rb1 compared with those in aging murine uteri) We also profiled the uterine proteomics of young mice, untreated aging mice, and aging mice treated with Rb1 ( n  = 5 for each group). Compared with those of young uteri, the proteomic landscape of aging uteri was distinct when different numbers of the most variable proteins were used (Fig. 3e–h and Supplementary Fig. 3c, d ). These aging-related changes were partially reversed by Rb1 when protein expression changes between young and aging uteri were compared with changes between treated and untreated aging uteri (Fig. 3i–l ). Notably, the proteome of treated aging uteri was more similar to that of young uteri than to that of untreated aging uteri (Fig. 3e–h and Supplementary Fig. 3c, d ), confirming the antiaging effect of Rb1. To investigate the underlying mechanism of the antiaging effect of Rb1, we performed an enrichment analysis of selected proteins whose expression was perturbed by aging and whose expression was reversed by Rb1 (Fig. 4a, b ). These proteins were significantly enriched in mTOR-related pathways (Fig. 4b and Supplementary Table S3 ), whereas proteins affected only by either aging or Rb1 were not (Supplementary Fig. 3e ). After the dysregulated proteins were converted into their homologs in humans, we observed consistent and specific enrichment in mTOR-related pathways (Supplementary Fig. 4 ). The upregulation of proteins in mTOR-related pathways was also confirmed in untreated aging uteri compared with young and Rb1-treated aging uteri (Supplementary Fig. 5 ). Therefore, we speculated that Rb1 could protect the uterus from aging by regulating mTOR activity. Fig. 4 Rb1 protects murine uteri from aging by regulating mTOR activity. a Venn diagram showing the intersection of the upregulated proteins from the aging group vs. the young group and the downregulated proteins from the aging + Rb1 group vs. the aging group. b KEGG and GO enrichment analyses of the 96 DEPs showing a close connection with mTOR signaling pathways. c Representative immunofluorescence images of p-mTOR in the EnSCs of murine uteri (scale bar, 100 µm). d The protein expression of mTOR and its downstream effectors in murine uteri was analyzed by western blotting. e The expression densities of mTOR and its downstream effector proteins relative to those of β-actin were assessed by densitometric analysis in murine uteri. f The expression levels of mTOR and its downstream effector proteins in HESCs were analyzed by western blotting. β-actin was used as an endogenous control. g The expression densities of mTOR and its downstream effector proteins relative to those of β-actin were assessed by densitometric analysis in HESCs. h Proposed model illustrating the role of mTOR activation in endometrial stromal cell senescence and uterine dysfunction. The schematic was created with BioRender. Each value is the mean ± standard deviation (SD) ( n  = 3 independent biological replicates) and was calculated by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 ( Rb1 ginsenoside Rb1, D-gal D-galactose, Young young murine uteri, Aging aging murine uteri, Aging + Rb1 aging murine uteri treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1 D-gal-induced HESCs treated with Rb1) Rb1 protects murine uteri from aging by regulating mTOR activity. a Venn diagram showing the intersection of the upregulated proteins from the aging group vs. the young group and the downregulated proteins from the aging + Rb1 group vs. the aging group. b KEGG and GO enrichment analyses of the 96 DEPs showing a close connection with mTOR signaling pathways. c Representative immunofluorescence images of p-mTOR in the EnSCs of murine uteri (scale bar, 100 µm). d The protein expression of mTOR and its downstream effectors in murine uteri was analyzed by western blotting. e The expression densities of mTOR and its downstream effector proteins relative to those of β-actin were assessed by densitometric analysis in murine uteri. f The expression levels of mTOR and its downstream effector proteins in HESCs were analyzed by western blotting. β-actin was used as an endogenous control. g The expression densities of mTOR and its downstream effector proteins relative to those of β-actin were assessed by densitometric analysis in HESCs. h Proposed model illustrating the role of mTOR activation in endometrial stromal cell senescence and uterine dysfunction. The schematic was created with BioRender. Each value is the mean ± standard deviation (SD) ( n  = 3 independent biological replicates) and was calculated by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 ( Rb1 ginsenoside Rb1, D-gal D-galactose, Young young murine uteri, Aging aging murine uteri, Aging + Rb1 aging murine uteri treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1 D-gal-induced HESCs treated with Rb1) Therefore, we quantified the activity of the mTOR signaling pathway through the phosphorylation status of mTOR, ribosomal protein S6, and S6 kinase (S6K). 27 , 28 We observed elevated phosphorylation levels of mTOR (p-mTOR), S6 (p-S6), and S6K (p-S6K) in the endometria of aging mice, with no obvious changes in the abundance of total proteins (Fig. 4c–e ), indicating aging-induced activation of mTOR. These observations were confirmed in human EnSCs treated with D-galactose (D-gal), a widely used aging model. 29 D-gal treatment for 72 h (from 10 to 50 g/L) successfully induced aging phenotypes, including elevated ROS levels and gene expression levels of CDKN1A (Supplementary Fig. 6a–c ), whereas treatment with 40 or 50 g/L D-gal resulted in nonnegligible cytotoxicity (Supplementary Fig. 6d ). Thus, EnSCs were treated with 30 g/L D-gal to mimic the aging process. Similarly, the D-gal-induced activation of mTOR signaling could be repressed by Rb1 (Fig. 4f, g ), indicating that Rb1 could exert an antiaging effect by inactivating mTOR signaling. AMPK-mediated suppression of mTOR signaling could reduce the SASP in cells and tissues in aging mice. 21 Rb1 was also reported to alleviate oxidative stress and aging by facilitating AMPKα phosphorylation. 30 , 31 To explore whether AMPK is a potential target of Rb1, we characterized the binding mode between Rb1 and AMPK through molecular docking. Docking analysis revealed that Rb1 formed a stable binding mode with AMPK, with a binding score of –10.7 kcal/mol (Supplementary Table S4 ), suggesting a favorable interaction (Fig. 5a ). To experimentally confirm this binding, we next performed surface plasmon resonance (SPR) analysis, which revealed a high-affinity interaction between Rb1 and AMPK α, with a dissociation constant (KD) of 5.40 × 10 −7  M (Fig. 5b ). Consistently, molecular dynamics (MD) simulations revealed that the Rb1-AMPK complex rapidly reached equilibrium and remained stable during a 10-μs trajectory, with Gibbs free energy landscape analysis identifying two well-defined low-energy conformational states, further supporting stable and energetically favorable binding (Fig. 5c, d ). We then assessed the functional phosphorylation sites of AMPK regulated by Rb1 using site-specific phospho-antibodies and Western blotting. The results revealed that Rb1 selectively increased phosphorylation at Thr172, the critical activation site of AMPK, whereas no significant changes were observed at Ser485 or Ser491 (Fig. 5e, f and Supplementary Fig. 7a ). In line with these findings, phosphorylation of Thr172 was markedly reduced in both aging murine uteri and D-galactose-induced senescent human EnSCs but was restored upon Rb1 treatment (Fig. 5g and Supplementary Fig. 7b ). Finally, we used Compound C, an AMPK inhibitor, to further verify the role of AMPK in the regulation of mTOR activity. The AMPK-mediated suppression of mTOR activity was reversed by Rb1, whereas the effect of Rb1 was abolished by AMPK inhibition, confirming that AMPK is an upstream regulator of the mTOR pathway and a key mediator of the antiaging activity of Rb1 (Fig. 5h, i and Supplementary Fig. 7d–f ). To further exclude potential off-target effects associated with pharmacological inhibition and to genetically validate AMPK dependency, we next performed siRNA-mediated knockdown of AMPKα1 in human endometrial stromal cells. Genetic depletion of AMPKα1 largely abolished the ability of Rb1 to suppress mTOR signaling, as evidenced by the restored phosphorylation of mTOR, S6K, and S6 (Supplementary Fig. 7g ). Consistent with these findings, the inhibitory effects of Rb1 on the expression of senescence-associated secretory phenotype (SASP) genes, including IL6 , CCL2 , and IL1B , were markedly attenuated upon AMPKα1 knockdown (Supplementary Fig. 7h ). Together, these pharmacological and genetic approaches establish AMPKα1 as a necessary upstream mediator of Rb1-induced mTOR inhibition and its senescence-modulating effects. Fig. 5 Rb1 inactivates mTOR signaling via AMPK phosphorylation. a Surface image, 3D cartoon, and binding model of Rb1 interacting with AMPK protein. Residues are shown as magenta sticks, with hydrogen bonds and hydrophobic interactions depicted as dashed lines. b Surface plasmon resonance (SPR) sensorgram showing the direct binding of Rb1 to AMPK at different concentrations (0.31–20 μM). c Gibbs free energy landscape of the Rb1-AMPK complex during MD simulation, plotted as a function of the RMSD and radius of gyration (Rg), indicating two stable low-energy conformational states. d Time evolution of the RMSD of the Rb1-AMPK complex over a 10-μs simulation, demonstrating overall structural stability throughout the trajectory. e Immunofluorescence staining of p-AMPKα (Thr172) in uterine sections from the young, aged, and aged + Rb1 groups (scale bar, 50 µm). f Western blot analysis of p-AMPKα (Thr172) and total AMPK in uterine tissues from the young, aged, and aged + Rb1 groups, with relative quantification shown below. g Western blot analysis of p-AMPKα (Thr172) in uterine tissues from the control, D-gal, and D-gal + Rb1 groups, with relative quantification shown below. h , i Western blot analysis of AMPK, mTOR and its downstream effectors (p-S6K, S6K, p-S6, S6, p-mTOR, mTOR) in HESCs. Optical densities relative to those of β-actin were assessed by densitometric analysis. β-actin was used as an endogenous control. j Schematic illustration of the proposed mechanism by which Rb1 attenuates stromal cell senescence. The schematic was created with BioRender. The data are presented as the means ± SDs ( n  = 3 independent biological replicates) and were analyzed by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01 (Rb1 ginsenoside Rb1, D-gal D-galactose, Com-C AMPK inhibitor, Young young murine uterus, Aging aging murine uterus, Aging + Rb1 aging murine uterus treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1D-gal-induced HESCs treated with Rb1, D-gal + Rb1 + Com-C D-gal-induced HESCs treated with Rb1 and an AMPK inhibitor) Rb1 inactivates mTOR signaling via AMPK phosphorylation. a Surface image, 3D cartoon, and binding model of Rb1 interacting with AMPK protein. Residues are shown as magenta sticks, with hydrogen bonds and hydrophobic interactions depicted as dashed lines. b Surface plasmon resonance (SPR) sensorgram showing the direct binding of Rb1 to AMPK at different concentrations (0.31–20 μM). c Gibbs free energy landscape of the Rb1-AMPK complex during MD simulation, plotted as a function of the RMSD and radius of gyration (Rg), indicating two stable low-energy conformational states. d Time evolution of the RMSD of the Rb1-AMPK complex over a 10-μs simulation, demonstrating overall structural stability throughout the trajectory. e Immunofluorescence staining of p-AMPKα (Thr172) in uterine sections from the young, aged, and aged + Rb1 groups (scale bar, 50 µm). f Western blot analysis of p-AMPKα (Thr172) and total AMPK in uterine tissues from the young, aged, and aged + Rb1 groups, with relative quantification shown below. g Western blot analysis of p-AMPKα (Thr172) in uterine tissues from the control, D-gal, and D-gal + Rb1 groups, with relative quantification shown below. h , i Western blot analysis of AMPK, mTOR and its downstream effectors (p-S6K, S6K, p-S6, S6, p-mTOR, mTOR) in HESCs. Optical densities relative to those of β-actin were assessed by densitometric analysis. β-actin was used as an endogenous control. j Schematic illustration of the proposed mechanism by which Rb1 attenuates stromal cell senescence. The schematic was created with BioRender. The data are presented as the means ± SDs ( n  = 3 independent biological replicates) and were analyzed by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01 (Rb1 ginsenoside Rb1, D-gal D-galactose, Com-C AMPK inhibitor, Young young murine uterus, Aging aging murine uterus, Aging + Rb1 aging murine uterus treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1D-gal-induced HESCs treated with Rb1, D-gal + Rb1 + Com-C D-gal-induced HESCs treated with Rb1 and an AMPK inhibitor) NF-κB is a transcription factor involved in many biological processes, including cellular senescence, aging-related inflammation, and oxidative stress injury. 32 Additionally, the NF-κB signaling pathway is downstream of mTOR. 33 , 34 We therefore hypothesized that the antiaging effect of Rb1 might be exerted through NF-κB. NF-κB is held in the cytoplasm in an inactive state by specific inhibitors and moves into the nucleus to activate gene transcription upon inhibitor degradation. 35 The nuclear translocation of the p65 subunit is essential for NF-κB signaling. 36 We observed abnormal accumulation of p65 in the nucleus in aging murine uteri, and Rb1 partially reduced the excess nuclear translocation of p65 (Fig. 6a ). Consistently, the mRNA expression of proinflammatory SASP factors, including Il6 , Ccl2 , and Il1b , was significantly upregulated in aging uteri, whereas Rb1 administration markedly attenuated their expression (Fig. 6e ). We observed similar results in D-gal-induced human EnSCs (Fig. 6b ). We further used the mTOR activator MHY1485 to confirm the role of the mTOR/NF-κB regulatory axis of Rb1. Rb1 failed to decrease the nuclear translocation of p65 upon mTOR activation (Fig. 6c ). Notably, the aging phenotypes of D-gal-induced senescent human EnSCs remained when Rb1 was used together with MHY, including elevated gene expression of aging markers, SASP factors, apoptotic cell rates, and ROS levels and disrupted mitochondrial membrane potential (Fig. 6d, f–i and Supplementary Fig. 7 ). Overall, we concluded that the antiaging effect of Rb1 is exerted via NF-κB and is highly dependent on the activation status of mTOR. To directly assess NF-κB transcriptional activity, we performed NF-κB-responsive luciferase reporter assays in human endometrial stromal cells. Compared with control cells, senescent cells exhibited significantly elevated NF-κB reporter activity. Treatment with Rb1 markedly suppressed NF-κB-driven luciferase activity, whereas reactivation of mTOR abolished this inhibitory effect. These results provide direct evidence that Rb1 attenuates NF-κB transcriptional activity downstream of AMPK/mTOR signaling, which is consistent with the observed reduction in nuclear p65 accumulation and SASP gene expression (Fig. 6j ). Fig. 6 Rb1 regulates NF-κB activity by inactivating mTOR. a Western blot analysis of nuclear NF-κB (Nu-NF-κB) protein levels in uterine tissues from the Young, Aging, and Aging + Rb1 groups, with relative densitometric quantification. b Western blot analysis of Nu-NF-κB in HESCs from the control, D-gal, and D-gal + Rb1 groups. c Western blot analysis of Nu-NF-κB in HESCs to investigate the interaction between NF-κB and mTOR under different treatments (Ctrl, D-gal, D-gal + Rb1, D-gal + Rb1 + MHY). d qRT‒PCR detection of the mRNA expression of the senescence-related markers CDKN2A, CDKN1A, and the proliferation marker PCNA in HESCs normalized to that of GAPDH. e Relative mRNA expression of proinflammatory SASP factors ( Il6 , Ccl2 , and Il1b ) in uterine tissues from the indicated groups. f Flow cytometry analysis of apoptosis in HESCs under the indicated treatments and quantification of apoptosis ratios. g Statistical analysis of apoptosis rates in HESCs. h Flow cytometry analysis of ROS levels in HESCs, with histograms showing the distribution of fluorescence intensity. i Quantification of ROS levels in HESCs from different groups. j Schematic illustration of the proposed mechanism: senescent endometrial stromal cells secrete proinflammatory SASP factors that contribute to uterine aging, which can be reversed by ginsenoside Rb1. Rb1 primarily exerts its effects by activating AMPK, thereby inactivating the mTOR pathway and subsequently reducing NF-κB activity. The schematic was created with BioRender. Each value is the mean ± standard deviation (SD) ( n  = 3 independent biological replicates) and was calculated by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1 ginsenoside Rb1, D-gal D-galactose, MHY an mTOR activator, Young young murine uteri, Aging aging murine uteri, Aging + Rb1 aging murine uteri treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1 D-gal-induced HESCs treated with Rb1, D-gal + Rb1 + MHY D-gal-induced HESCs treated with Rb1 and mTOR activator) Rb1 regulates NF-κB activity by inactivating mTOR. a Western blot analysis of nuclear NF-κB (Nu-NF-κB) protein levels in uterine tissues from the Young, Aging, and Aging + Rb1 groups, with relative densitometric quantification. b Western blot analysis of Nu-NF-κB in HESCs from the control, D-gal, and D-gal + Rb1 groups. c Western blot analysis of Nu-NF-κB in HESCs to investigate the interaction between NF-κB and mTOR under different treatments (Ctrl, D-gal, D-gal + Rb1, D-gal + Rb1 + MHY). d qRT‒PCR detection of the mRNA expression of the senescence-related markers CDKN2A, CDKN1A, and the proliferation marker PCNA in HESCs normalized to that of GAPDH. e Relative mRNA expression of proinflammatory SASP factors ( Il6 , Ccl2 , and Il1b ) in uterine tissues from the indicated groups. f Flow cytometry analysis of apoptosis in HESCs under the indicated treatments and quantification of apoptosis ratios. g Statistical analysis of apoptosis rates in HESCs. h Flow cytometry analysis of ROS levels in HESCs, with histograms showing the distribution of fluorescence intensity. i Quantification of ROS levels in HESCs from different groups. j Schematic illustration of the proposed mechanism: senescent endometrial stromal cells secrete proinflammatory SASP factors that contribute to uterine aging, which can be reversed by ginsenoside Rb1. Rb1 primarily exerts its effects by activating AMPK, thereby inactivating the mTOR pathway and subsequently reducing NF-κB activity. The schematic was created with BioRender. Each value is the mean ± standard deviation (SD) ( n  = 3 independent biological replicates) and was calculated by two-sided Student’s t test. * p  < 0.05; ** p  < 0.01; *** p  < 0.001 (Rb1 ginsenoside Rb1, D-gal D-galactose, MHY an mTOR activator, Young young murine uteri, Aging aging murine uteri, Aging + Rb1 aging murine uteri treated with Rb1, Ctrl HESCs without D-gal pretreatment, D-gal D-gal-induced HESCs, D-gal + Rb1 D-gal-induced HESCs treated with Rb1, D-gal + Rb1 + MHY D-gal-induced HESCs treated with Rb1 and mTOR activator)

Fertility declines dramatically with advanced maternal age and is accompanied by elevated risks of certain reproductive disorders. 37 , 38 In addition to aging-related abnormalities in oocytes and embryos, uterine aging has nonnegligible adverse effects on pregnancy. 4 Our current findings, together with recent clinical evidence, strongly support the notion that endometrial aging compromises receptivity and negatively influences pregnancy outcomes. In this study, compared with their younger counterparts, endometrial stromal cells from older women displayed elevated senescence markers, increased oxidative stress, and increased apoptosis, suggesting impaired decidualization capacity and reduced receptivity. Consistent with our recently published work in Nature Aging, 7 we performed a systematic analysis of 1149 patients who underwent PGT-A. After excluding embryonic aneuploidy, we found that biochemical pregnancy, clinical pregnancy, and live birth rates were significantly lower in middle-aged women, highlighting the adverse impact of endometrial aging on fertility. However, unlike comprehensive studies on the ovaries, oocytes, and embryos of aging pregnant women, 39 , 40 the underlying mechanism of uterine and endometrial aging remains elusive, strongly limiting the development of treatment strategies for uterine aging. Recent studies have further highlighted the pivotal role of stromal cell senescence in compromising endometrial receptivity and implantation. Stromal cell senescence has been shown to impair decidualization and disrupt the interaction between endometrial stromal cells and trophoblasts, thereby directly interfering with successful embryo implantation. 9 Moreover, inflammatory stimuli such as interleukin-1β were reported to induce and accelerate EnSC senescence through the c-Jun N-terminal kinase pathway, leading to impaired decidualization. 41 In line with these findings, a recent review emphasized that multiple molecular activities, including oxidative stress, inflammation, fibrosis, the DNA damage response, and cellular senescence, contribute to uterine aging and the deterioration of endometrial function. 42 Together, these studies provide compelling evidence that EnSC senescence is a mechanistic driver of impaired decidualization and defective maternal–fetal interactions, reinforcing its critical role in implantation failure and reproductive decline. Moreover, in the aged endometrial samples in our study, EnSCs displayed elevated expression of senescence markers (p21, p16, and APOE), increased secretion of SASP factors such as Il6 and Ccl2 , and impaired decidualization capacity. In addition, epithelial cells exhibit aging-associated alterations, including abnormal proliferation and compromised barrier function. These results indicate that stromal senescence and defective decidualization, together with epithelial dysfunction, constitute critical phenotypes of uterine aging. Panax ginseng C. A. Mey, a well-known traditional Chinese medicine with multiple pharmacological activities, plays a crucial role in modulating a variety of disorders associated with aging. Ginsenosides are the main bioactive components, and Rb1 is among the most abundant ginsenosides and has antitumoral, antiaging, antioxidative, and anti-inflammatory effects. 43 – 45 Importantly, Rb1 has been proven to attenuate aging-related oxidative stress in various tissues in different animal models. 46 , 47 For instance, Rb1 can restore aging-related oxidative damage in granulosa cells and improve the fertility of aging. 17 Therefore, Rb1 represents a promising senescence-modulating agent that was applied in this study to ameliorate uterine aging. In terms of safety, we observed no obvious cytotoxicity of Rb1 in primary human EnSCs even at relatively high concentrations (up to 100 μmol/L) or across different time points, supporting the suitability of 20 μmol/L Rb1 for 12 h as an optimal treatment condition. To further assess potential systemic toxicity in vivo, we performed toxicological analyses in aged mice subjected to daily intraperitoneal Rb1 administration. Serum biochemical parameters, including ALT, AST, UREA, and CREA, did not significantly differ between the Rb1-treated and control groups. Moreover, histological examination of liver and kidney tissues by H&E staining revealed no apparent pathological alterations. These findings indicate that Rb1 treatment did not induce hepatic or renal toxicity in our experimental settings. Although these results, together with those of previous reports, support the favorable safety profile of Rb1, we acknowledge that future studies should include long-term administration, comprehensive hematological and biochemical assessments, and pharmacokinetic profiling to fully establish the systemic safety of Rb1 for clinical application. To evaluate the antiaging effect of Rb1 on the uterus and endometrium, we collected aging human endometrial cells and murine uteri, which were treated with Rb1. Rb1 could partially reverse aging phenotypes in both human EnSCs and murine uteri, including elevated expression of aging markers and SASP factors, oxidative stress, and apoptosis. We also investigated the AMPK/mTOR/NF-κB regulatory axis (Fig. 6j ), through which Rb1 exerts its antiaging effect on senescent human EnSCs and aging murine uteri. Utilizing molecular docking, we speculated that AMPK is the direct target of Rb1. Through binding with AMPK, Rb1 reversed the aging-related decrease in p-AMPK levels and inactivation of the mTOR signaling pathway. The excess accumulation of p65 (an NF-κB subunit) was then eliminated because of mTOR inactivation. Notably, pharmacological inhibition of AMPK or enforced activation of mTOR efficiently abolished the antiaging effects of Rb1, further confirming that the senescence-modulating actions of Rb1 are critically dependent on AMPK activation and mTOR suppression. This AMPK/mTOR/NF-κB regulatory axis has been widely reported to be related to aging processes, including those in reproductive organs. For instance, mTOR signaling plays a key role in uterine aging and the timing of birth, 48 , 49 whose inhibitor, rapamycin, has been proven to prolong life and slow aging-dependent endometrial lesions. 50 Recent studies have demonstrated direct associations between the SASP and NF-κB-mediated inflammation. NF-κB, a major stress responder, is activated to induce inflammation during senescence and has been reported to be regulated by mTOR. 32 Here, we revealed that Rb1 prevented HESC senescence and the SASP by inhibiting NF-κB transcription. Moreover, the phosphorylation of AMPK causes the inactivation of mTOR and subsequently plays a role in cell protection. 51 In a recent study, AMPK was shown to be activated via the γ1 subunit to protect HepG2 cancer cells. 52 Here, we further revealed that Rb1 could activate AMPK through interaction with the α1 subunit of AMPK, and this mechanism has been demonstrated to play an important role in the prevention of cell senescence in the uterus. While mTOR-related pathways emerged as the most significantly enriched cluster in our proteomic analysis of aging and Rb1 intervention, several other functionally relevant pathways were also dysregulated by aging and subsequently reversed upon Rb1 treatment. Of particular interest is the reversal of genes associated with the “cellular response to amino acid stimulus” pathway. Key components of the lysosomal nutrient-sensing machinery—including HSF1, RRAGC, RRAGD, and LAMTOR2/3—which are critical for mTORC1 activation, 53 – 55 were restored toward low expression levels. These findings suggest that Rb1 may fundamentally restore cellular amino acid sensing, a process known to be impaired with aging. Furthermore, the reversal of genes involved in glucose import, such as TSC2 and RPS6KB1, 56 indicates that Rb1 ameliorates age-related insulin resistance and glucose metabolic dysfunction, thereby directly linking mTOR signaling to systemic glucose homeostasis. Additionally, we observed normalization of a procoagulant gene signature—including markers such as PROC and GP1BA 57 , 58 —implying that Rb1 may mitigate the increased risk of thrombosis and vascular complications associated with aging. This represents a novel and promising dimension of its potential geroprotective effects. It is important to acknowledge the physiological differences between the human endometrium and murine endometrium when our findings are interpreted. In humans, the endometrium undergoes cyclical proliferation, differentiation, and menstrual shedding, whereas mice exhibit estrous cycles without overt menstruation. Moreover, decidualization occurs spontaneously in human endometrial stromal cells under the influence of progesterone and cAMP, whereas in mice, decidualization is typically induced only upon embryo implantation. Despite these species-specific distinctions, fundamental cellular and molecular pathways governing oxidative stress responses, mitochondrial function, and cellular senescence are conserved across species. Therefore, murine models remain valuable for investigating uterine aging, but validation using human endometrial samples, as performed in this study, enhances the translational relevance of our findings. Notably, the AMPK–mTOR–NF-κB signaling axis described in this study does not imply a rigid or universal linear hierarchy across all biological contexts. Previous studies have reported bidirectional crosstalk between mTOR and NF-κB signaling, with NF-κB capable of modulating mTOR activity under certain inflammatory or oncogenic conditions. In the context of senescent endometrial stromal cells and aging uterus, however, our data support a functionally dominant pathway in which mTOR acts upstream of NF-κB to regulate senescence-associated inflammatory responses. Specifically, pharmacological reactivation of mTOR consistently restored NF-κB nuclear accumulation, transcriptional activity, and SASP gene expression, even in the presence of Rb1, indicating that mTOR activity is required for downstream NF-κB signaling in this model. Thus, while reciprocal regulation between these pathways may exist in other cellular systems, our findings define a context-dependent signaling hierarchy that is particularly relevant to uterine aging and stromal cell senescence. Future studies are needed to further elucidate the potential feedback loops between mTOR and NF-κB under distinct physiological and pathological conditions. From a clinical perspective, our findings suggest that Rb1 may represent a potential mechanistic candidate for targeting uterine aging in women of advanced maternal age. Given its natural origin, favorable safety profile, and ability to target the AMPK/mTOR/NF-κB axis, Rb1 could be considered an adjuvant intervention to improve endometrial receptivity, reduce implantation failure, and potentially extend reproductive longevity. Future clinical studies with larger patient cohorts are warranted to validate the therapeutic efficacy of Rb1 and determine the optimal dosing regimens. Moreover, the development of Rb1-based formulations with enhanced bioavailability and targeted delivery, such as intrauterine devices or nanocarrier systems, may further facilitate its translation into clinical practice. Beyond reproductive health, exploring the effects of Rb1 on other aging-related uterine disorders, such as endometrial atrophy or adenomyosis, will provide a broader understanding of its therapeutic potential. In summary, we characterized the crucial role of the AMPK/mTOR/NF-κB regulatory axis in uterine aging and identified Rb1 as a potential mechanistic candidate for modulating uterine aging-associated dysfunction. Although further studies are needed to elucidate the underlying mechanisms in greater detail and to evaluate the relevance of these findings in larger and well-defined human cohorts, our data provide preclinical and mechanistic evidence supporting the role of Rb1 in regulating endometrial stromal cell senescence and inflammatory signaling. Given its favorable safety profile as a natural compound, Rb1 may offer a valuable foundation for future mechanism-based strategies aimed at targeting uterine aging.

Pregnancy at advanced maternal age (over age 35) is among the leading causes of adverse maternal–perinatal outcomes and has become more prevalent worldwide in recent decades because of the advancement of modern societies. 1 – 3 Delayed childbearing, changing educational and professional priorities, and broader access to assisted reproductive technologies have all contributed to a continuous increase in the proportion of women attempting conception later in life. This demographic shift has made reproductive aging a major medical and societal challenge. Although adverse reproductive outcomes in aging women are closely associated with hypo-ovarianism, abnormalities in other reproductive organs, such as the uterus, also contribute to declining fertility. 4 In this context, the uterus should no longer be regarded as a passive recipient of embryo implantation, but rather as an active and highly dynamic organ whose age-dependent functional deterioration may critically influence reproductive success. The endometrium is dramatically altered in women older than 35 years, along with a dysregulation of crucial endometrial functions and a possible link to female infertility. 5 These alterations include impaired regenerative capacity, aberrant hormone responsiveness, disrupted stromal–epithelial communication, and a decline in the ability to establish a receptive microenvironment for implantation. Recent studies have highlighted the effects of aging uteri on reproductive failure and have validated their adverse effects on fertility using animal models. 6 In our recent study, we systematically analyzed the clinical outcomes of 1149 patients who underwent Preimplantation Genetic Testing for Aneuploidy (PGT-A) and demonstrated that, even after embryonic aneuploidy was excluded, biochemical pregnancy, clinical pregnancy, and live birth rates were significantly lower in middle-aged women. These findings provide robust clinical evidence that endometrial aging leads to impaired receptivity, thereby contributing to reproductive decline in older women. 7 Aging uteri fail to undergo normal decidualization upon hormone treatment, 8 which is likely the result of endometrial stromal cell (EnSC) senescence. 9 Decidualization is a highly coordinated process required for embryo implantation and early placental development, during which stromal cells undergo extensive morphological, transcriptional, metabolic, and secretory reprogramming in response to ovarian steroid hormones and intracellular cyclic AMP signaling. When this transformation is compromised, the endometrium becomes less receptive and less capable of supporting maternal–fetal communication. Senescent EnSCs are characterized by a senescence-associated secretory phenotype (SASP), excess production of reactive oxygen species (ROS), abnormal expression of oxidation-related enzymes, and apoptosis. 10 These age-associated changes not only impair stromal cell fitness, but also reshape the local microenvironment by promoting chronic low-grade inflammation, extracellular matrix remodeling, and paracrine dysfunction. EnSC senescence and changes in their secretome have been suggested to be partially responsible for decreased endometrial plasticity, implantation failure, and recurrent pregnancy loss, 11 , 12 which are prevalent in aging women. 9 , 13 Increasing evidence further suggests that stromal senescence may disrupt the reciprocal interactions among stromal cells, epithelial cells, immune cells, and invading trophoblasts, thereby amplifying implantation defects beyond a single cellular compartment. Therefore, EnSC senescence represents a key hallmark of uterine aging, and therapeutic targeting of senescence-associated dysfunction in EnSCs may provide an effective strategy to alleviate uterine aging. However, despite growing recognition of the importance of stromal aging, the molecular pathways that couple cellular senescence to global endometrial dysfunction remain incompletely defined, which has limited the development of mechanism-based therapeutic interventions. Panax ginseng C. A. Mey is a globally known natural product and is widely used to treat aging-related diseases such as cancers and neurodegenerative diseases. 14 , 15 As a classic medicinal herb with pleiotropic biological activities, ginseng has long attracted attention because of its ability to regulate oxidative stress, inflammation, metabolism, and tissue repair, all of which are intimately involved in the biology of aging. Ginsenoside Rb1 (Rb1) is the most abundant ginsenoside in ginseng and has anti-inflammatory, antioxidative, and antiaging effects. 16 Previous studies in multiple organ systems have shown that Rb1 can suppress oxidative injury, improve mitochondrial homeostasis, attenuate inflammatory signaling, and preserve cellular viability under stress conditions. These properties make Rb1 a particularly attractive candidate for the treatment of uterine aging, in which oxidative stress and chronic inflammatory activation are thought to drive progressive stromal dysfunction. Our previous study revealed that Rb1 is an effective intervention for ovarian aging because of its antioxidative effects on granulosa cells. 17 This finding further supports the possibility that Rb1 may exert beneficial effects across multiple compartments of the female reproductive system. However, whether Rb1 can attenuate aging-related EnSC senescence and endometrial dysfunction is still unknown. In particular, it remains unclear whether the protective action of Rb1 in the uterus is linked to canonical nutrient- and stress-sensing pathways that are known to regulate senescence, inflammation, and cellular homeostasis. To address this question, aging human endometrial cells and murine uteri were treated with Rb1, whose antiaging effects were characterized thoroughly by proteomic analyses and functional validation. By integrating human samples, aged mouse models, and mechanistic experiments, we aimed not only to determine whether Rb1 can ameliorate uterine aging phenotypes, but also to identify the signaling framework through which these effects are mediated. Rb1 effectively alleviated uterine aging in humans and mice. Proteomic analyses of murine uteri revealed that the antiaging effect of Rb1 likely occurred through regulation of the activity of mammalian target of rapamycin (mTOR). Because mTOR is a central metabolic hub linking nutrient sensing, inflammation, and senescence, this observation suggested a plausible mechanistic entry point for understanding how Rb1 restores endometrial homeostasis. Docking simulation studies have suggested that Rb1 can form a suitable steric complementarity with the binding site of AMP-activated protein kinase (AMPK), which is an upstream regulator of the mTOR signaling pathway. 18 Rb1 failed to suppress mTOR signaling upon AMPK inhibition, indicating that the effect of Rb1 on mTOR was mediated by AMPK. To further link mTOR to EnSC senescence, we focused on nuclear factor-kappa B (NF-κB), which is a major regulator of the SASP and is downstream of mTOR. 19 – 22 Rb1 could eliminate accumulated NF-κB in the nucleus by regulating AMPK/mTOR activity in both human EnSCs and murine uteri, thus ameliorating aging-related phenotypes. In summary, our findings provide new insights into the role of Rb1 in the protection against uterine aging and suggest that AMPK/mTOR/NF-κB signaling pathways are potential therapeutic targets for aging-related uterine dysfunctions. More broadly, this study establishes a mechanistic rationale for targeting stromal senescence and inflammatory signaling as a strategy to improve endometrial function in women of advanced maternal age.

All animal experiments were conducted in accordance with institutional guidelines for the care and use of laboratory animals and were approved by the Institutional Animal Care and Use Committee of Peking University Third Hospital (Approval No. A2022146). The collection and use of human samples were approved by the Medical Science Research Ethics Committee of Peking University Third Hospital (Approval No. M2022696). Written informed consent was obtained from all participants prior to their inclusion in the study. Endometrial samples were obtained from patients undergoing hysteroscopic examination at the Center for Reproductive Medicine, Department of Obstetrics and Gynecology, Peking University Third Hospital, Beijing, China. The study population included a younger group of patients aged 22–32 years ( n  = 15) and an aging group aged 40–45 years ( n  = 19). To minimize cycle-related variability, all samples included in this study were collected during the mid-secretory phase, corresponding to the window of implantation. Menstrual cycle dating was determined on the basis of a combination of menstrual history, serum hormone levels, and histological evaluation according to established criteria. Samples that did not meet these criteria or were collected outside the defined mid-secretory phase were excluded from further analyses. Patients with endometrial abnormalities or those who had received sex hormone treatment or hormone replacement therapy within the previous three months were excluded from the study. Patients with known gynecological or systemic conditions that could confound endometrial function were excluded. Specifically, women with a history or clinical suspicion of endometriosis (e.g., dysmenorrhea, chronic pelvic pain), prior laparoscopic diagnosis of endometriosis, or imaging findings suggestive of endometriosis on transvaginal ultrasound or, when indicated, MRI were not included. Patients with evidence of endometrial inflammatory disorders, including chronic endometritis, pelvic inflammatory disease, or intrauterine adhesions, were also excluded. Routine gynecological examination and hysteroscopy performed prior to the ART procedures confirmed the absence of intrauterine infection or pathology in all the included participants. In addition, patients with polycystic ovary syndrome, autoimmune diseases, diabetes mellitus, or other systemic inflammatory disorders were excluded to minimize confounding effects. The detailed exclusion criteria are summarized in Supplementary Table S2 . Primary human endometrial stromal cells (EnSCs) were isolated from endometrial tissues. The collected samples were placed into 15-ml centrifuge tubes containing sterile Dulbecco’s phosphate-buffered saline (DPBS; Gibco, USA) and transported to the laboratory at 4 °C within 2 h. The tissues were subsequently washed with DPBS containing 1% penicillin–streptomycin (Gibco, USA) until all blood residue was removed. The cleaned tissues were then diced into 1 mm³ blocks using a scalpel and digested with collagenase V (Sigma-Aldrich, USA) diluted in DMEM/F12 (Gibco, USA) for 20 min at 37 °C in an incubator with 5% CO 2 . Following enzymatic digestion, the cells were centrifuged at 1000 rpm for 5 min, and the resulting pellets were resuspended in fresh DMEM/F12 (Gibco, USA). The cell suspension was sequentially passed through 100-μm and 40-μm cell strainers stacked together with a tube placed underneath. Owing to the difference in cellular size, endometrial epithelial cells remained on the 40-μm strainer, while EnSCs passed through both strainers and were collected in the tube. The isolated EnSCs were then resuspended in DMEM/F12 (Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) before being transferred to 6-well plates for further culture. Primary human EnSCs and HESCs were cultured in DMEM/F12 (Gibco, USA) supplemented with 10% FBS (Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) at 37 °C in a 5% CO 2 incubator. EnSCs were treated with 20 μmol/L ginsenoside Rb1 for 12 h, while senescence was induced in HESCs using D-gal (Solarbio, CN). Senescent cells were detected using a senescence β-galactosidase staining kit (Cell Signaling, USA). Cells were seeded in 6-well plates and grown to 75% confluence before they were treated with 30 g/L D-gal for 24, 48, or 72 h. Following treatment, the cells were fixed for 15 min at room temperature, washed with PBS, and stained with β-galactosidase solution overnight at 37 °C. The next day, senescent cells were observed under a light microscope, and three different areas or at least 500 cells were counted. For long-term storage, the cells were covered with 70% glycerol. The experiment was repeated three times. Cell viability was assessed using a CCK8 kit (Apexbio, USA) according to the manufacturer’s instructions. Cells were seeded into a 96-well plate at a density of 7000 cells per well, followed by the addition of various concentrations of Rb1 (0, 10, 20, 40, 80, and 100 µmol/L) for different durations (6, 12, and 24 h) or D-gal (10, 20, 30, 40, and 50 g/L) for 72 h. Subsequently, 10 μL of CCK8 solution was added to each well, and the plate was incubated at 37 °C for 2 h. Cell viability was then determined by measuring the absorbance at 570 nm using a microplate reader (Multiskan GO, Thermo Fisher Scientific, USA) after 60 s of agitation. The results are expressed as a percentage of the control, representing cell viability. Reactive oxygen species (ROS) levels, apoptosis rates, and mitochondrial membrane potential were measured using flow cytometry. The analyses were performed on a CytoFLEX instrument (Beckman Coulter, USA), and the resulting data were processed with CytExpert software version 2.0. ROS levels were quantified using 2,7-dichlorofluorescein-diacetate cellular ROS assays (Abcam, UK) according to the manufacturer’s instructions. Apoptosis rates were determined with an Annexin V-FITC/PI apoptosis kit (Elabscience, Wuhan, China). The mitochondrial membrane potential (ΔΨm) was assessed using a JC-1 assay kit (Abcam, UK) following the manufacturer’s instructions. After drug treatment, the cells were incubated with 3 μmol/L JC-1 working solution at 37 °C in the dark for 20 min. The cells were then rinsed three times with DPBS and analyzed using flow cytometry (BD FACSAria III). The results are expressed as the proportion of cells with decreased mitochondrial membrane potential. All animal experiments were approved by the Animal Care and Treatment Committee of the Peking University Health Science Center and conducted in accordance with established ethical guidelines. Female C57BL/6 mice were divided into a young group (8 weeks) and a reproductive aging group (10 months). Reproductive aging mice were randomly assigned to receive daily intraperitoneal injections of Rb1 or vehicle control for two weeks, while young mice served as controls. The mice were housed in a controlled environment with consistent temperature and humidity and had unrestricted access to sterile food and water. Ginsenoside Rb1 (molecular weight = 1109.29; purity >98%; PCS0759) was freshly prepared in saline and administered intraperitoneally to Rb1-treated aged mice daily for two weeks. The levels of MDA and superoxide dismutase (SOD) in EnSCs across the three groups were measured using assay kits (Beyotime, Shanghai, China) according to the manufacturer’s protocols. The cell culture media were collected to determine extracellular lactate dehydrogenase (LDH) levels. LDH release was quantified using an LDH cytotoxicity kit (Roche, Switzerland) following the manufacturer’s instructions. This assay involves the conversion of lactate to pyruvate, which then reacts with 2,4-dinitrophenylhydrazine, producing a brownish-red color in a basic solution. The absorbance of the resulting solution was measured at 450 nm using a microplate reader (SpectraFluor, Tecan, Austria). Endometrial tissues for H&E staining were fixed in 4% paraformaldehyde, embedded in paraffin, and cut into 5-μm sections. The sections were dewaxed, hydrated, and stained with hematoxylin and eosin. For immunofluorescence, endometrial cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100, and blocked with 1% BSA. The cells were then incubated overnight at 4 °C with primary antibodies against cytokeratin-19 (ab52625; Abcam, UK; 1:50), vimentin (ab8978; Abcam, 1 µg/ml), phospho-mTOR (ab131538; Abcam, UK; 1:100), and phospho-AMPK (AF3423; Affinity Biosciences; 1:100). This was followed by a 1-h incubation at room temperature with the corresponding fluorescent secondary antibodies. The slides were counterstained with 1 μg/mL DAPI (Invitrogen, USA) to visualize the nuclei. Immunofluorescence images were captured using a confocal scanning microscope (LSM710; Zeiss, Germany). SPR measurements were performed on a Biacore T200 system (GE Healthcare). Recombinant AMPK α1 protein was immobilized on a CM5 sensor chip using amine coupling, and serial dilutions of Rb1 (0.31–20 μM) were injected in running buffer (0.05% Tween-20 in PBS). Sensorgrams were recorded and analyzed using a 1:1 binding model to calculate the kinetic parameters ( k a , k d ) and the equilibrium dissociation constant ( K D ). Docking poses of the Rb1-AMPK complex were subjected to 10000 ns molecular dynamics simulations using the GROMACS 2020.6 package. The amber14sb force field and TIP3P water model were applied. The system was neutralized and equilibrated prior to production runs at 300 K and 1 atm. Trajectory analyses included root mean square deviation (RMSD), root mean square fluctuation (RMSF), and Gibbs free energy landscape mapping. Serum samples were collected from experimental mice after Rb1 treatment. Biochemical indices, including alanine aminotransferase (ALT), aspartate aminotransferase (AST), urea (UREA), and creatinine (CREA), were measured using an automatic biochemical analyzer (Hitachi 7600, Japan). Major organs (liver, kidney, heart, spleen, and lung) were fixed in 4% paraformaldehyde, embedded in paraffin, and sectioned for hematoxylin–eosin (H&E) staining. Histological morphology was assessed by a pathologist who was blinded to the treatment groups. Primary endometrial stromal cells (EnSCs) were isolated from young and aging mice as described above. Cells were cultured in phenol red-free DMEM/F12 supplemented with 2% charcoal-stripped FBS, 0.5 mM 8-bromo-cAMP, and 1 μM medroxyprogesterone acetate (MPA) for 4 days to induce decidualization. Total RNA was extracted using TRIzol reagent, and cDNA was synthesized using a PrimeScript RT kit (Takara). Quantitative PCR was performed to determine the mRNA levels of Prl and Igfbp1 , which were normalized to those of Gapdh . Mouse uteri, primary human EnSCs, and HESCs were lysed on ice using tissue or cell protein extraction buffer (Sigma-Aldrich, USA) supplemented with 1% phosphatase and protease inhibitor cocktail (Sigma-Aldrich, USA). Equal amounts of protein from each sample were separated by gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto a polyvinylidene fluoride (PVDF) membrane. The membranes were blocked with 5% nonfat milk powder in buffer for 2 h at room temperature, followed by incubation with specific primary antibodies at 4 °C for 12 to 18 h. The primary antibodies used targeted the following proteins: β-actin (ab8226, Abcam, UK, 1 µg/ml), cleaved caspase-3 (ab2302, Abcam, UK, 1:50), cleaved caspase-9 (9509, Cell Signaling Technology, USA, 1:1000), BCL-2 (ab182858, UK, 1:1000), mTOR (ab134903, Abcam, UK, 1:10000), phospho-mTOR (ab109268, Abcam, UK, 1:1000), S6K (ab32359, Abcam, UK, 1:1000), phospho-S6K (ab59208, Abcam, UK, 1:1000), S6 (ab225676, Abcam, UK, 1:1000), phospho-S6 (ab215214, UK, 1:5000), AMPKα (2532, Cell Signaling Technology, 1:1000), phospho-AMPKα (2535, Cell Signaling Technology, USA, 1:1000), NF-κB (8242, Cell Signaling Technology, USA, 1:1000), and LAMIN B (12586, Cell Signaling Technology, USA, 1:1000). After primary antibody incubation, the membranes were washed three times with Tris-buffered saline supplemented with polysorbate 20 (TBST) and then incubated with the appropriate secondary antibodies for 1.5 h. Following another round of TBST washes, the membranes were developed using enhanced chemiluminescence (Pierce, USA) and imaged using a Bio-Rad system. Total RNA was extracted using TRIzol™ reagent (Invitrogen) according to the manufacturer’s instructions. Complementary DNA (cDNA) was synthesized from 1 μg of total RNA using the RevertAid First-Strand cDNA Synthesis Kit (Thermo Fisher). Quantitative real-time PCR (qRT-PCR) was performed on an ABI 7500 instrument (Applied Biosystems, NY) using SYBR Green Real-time PCR Master Mix (Applied Biosystems, NY). The primer sequences are listed in Supplementary Table S5 (Supplementary materials). The mRNA levels of target genes were normalized to the expression of the housekeeping gene GAPDH. AutoDock Vina 59 was used to perform molecular docking of Rb1 with AMPK α1. The structure of Rb1 was obtained from PubChem ( https://pubchem.ncbi.nlm.nih.gov/ ). The structure of AMPK α1 was obtained from the RCSB Protein Data Bank ( https://www.rcsb.org/ ), with a protein PDB ID of 4RER. 60 Rb1 was converted to a PDBQT format, with all the bonds set as rotatable. The receptor (AMPK α1) was kept fixed, while the ligand (Rb1) was allowed flexibility. The flexible docking simulation was conducted using the Lamarckian genetic algorithm. The binding pocket for the peptides was selected on the basis of the binding site of the natural ligand staurosporine in AMPK α1. The coordinates of the search space are provided in Supplementary Table S4 (Supplementary materials). Possible conformations of the ligand were explored within the grid box, and the conformation with the lowest binding free energy was identified as the most likely. Interactions between protein‒ligand complexes were visualized using PyMol. Uteri from all 15 mice were suspended in lysis buffer supplemented with 8 M urea, 1% sodium deoxycholate, and a protease inhibitor cocktail. The mixture was vortexed thoroughly and subjected to three rounds of high-throughput tissue grinding. The samples were then incubated at 4 °C for 30 min, with vortexing every 10 min. Following centrifugation at 12,000 ×  g for 20 min at 4 °C, the protein concentration in the supernatant was determined using a BCA protein assay kit. The sample preparation process included protein denaturation, reduction, alkylation, tryptic digestion, and peptide cleanup. A commercially available iST sample preparation kit (PreOmics, Germany) was used according to the manufacturer’s protocols. Briefly, after the samples were ground with liquid nitrogen, 50 µL of lysis buffer was added, and the mixture was heated at 95 °C for 10 min at 1000 rpm with agitation. The sample was then cooled to room temperature, followed by the addition of trypsin digestion buffer and incubation at 37 °C for 2 h at 500 rpm with shaking. The digestion process was halted using stop buffer. Peptide cleanup and desalting were performed using an iST cartridge with the recommended wash buffers. Peptides were eluted with elution buffer (2 × 100 µL) and lyophilized using a SpeedVac. The peptide mixture was redissolved in buffer A (20 mM ammonium formate in water, pH 10.0, adjusted with ammonium hydroxide) and then fractionated by high-pH reversed-phase separation using an Ultimate 3000 system (Thermo Fisher Scientific, USA) connected to a reversed-phase XBridge C18 column (4.6 mm × 250 mm, 5 μm; Waters Corporation, USA). The separation was performed with a linear gradient from 5% buffer B to 45% buffer B over 40 min (buffer B: 20 mM ammonium formate in 80% acetonitrile, pH 10.0, adjusted with ammonium hydroxide). The column was re-equilibrated to its initial condition for 15 min. The flow rate was maintained at 1 mL/min, and the column temperature was maintained at 30 °C. Six fractions were collected, and each was dried in a vacuum concentrator for subsequent analysis. Peptides were redissolved in 30 μL of solvent A (0.1% formic acid in water) and analyzed using an Orbitrap Fusion Tribrid mass spectrometer coupled to an EASY-nanoLC 1200 system (Thermo Fisher Scientific, USA). A 3-μL peptide sample was loaded onto an analytical column (Acclaim PepMap C18, 75 μm × 25 cm) with a 120-min gradient ranging from 5% to 35% solvent B (0.1% formic acid in acetonitrile). The column flow rate was set at 200 nL/min, and the column temperature was maintained at 40 °C. The electrospray voltage was set to 2 kV. The mass spectrometer was operated in data-independent acquisition (DIA) mode and automatically switched between MS and MS/MS modes. The parameters used were as follows: (1) MS: scan range ( m / z ), 350–1500; resolution, 120,000; AGC target, 1 × 10 6 ; maximum injection time, 50 ms; (2) HCD-MS/MS: resolution, 30,000; AGC target, 1 × 10 6 ; collision energy, 32; and stepped CE, 5%; and (3) DIA with a variable isolation window (each window overlapped 1  m / z , and the window number was 60). DIA proteomics data were analyzed and processed by Spectronaut Pulsar (Biognosys AG, Switzerland) with the default settings. The retention time prediction type was set to dynamic iRT. Data extraction was performed by Spectronaut Pulsar on the basis of extensive mass calibration. Spectronaut Pulsar 61 determined the ideal extraction window dynamically depending on iRT calibration and gradient stability. The false discovery rate (FDR) cutoff at the precursor and protein levels was 1.0%. Decoy generation was set to be mutated, similar to scrambled generation, but a random number of amino acid position swaps was applied (min = 2, max = length/2). The normalization strategy was set to local normalization. The average peak area of the top 3 peptides with an FDR less than 1.0% was used to quantify the major group of proteins. The results represent the final expression of a particular protein across samples, resulting in a 19,221 × 15 protein expression matrix. After the same expressed proteins from all the samples were merged and at least three samples were included, a total of 6257 proteins were identified and obtained. Log2 transformation was applied to the expression levels of the proteins. PCA was performed using the “prcomp” function, and factoextra v1.0.7 was used for visualization. Missing values in the proteomics data were imputed using the mean expression levels of the corresponding proteins in all the samples. The Pearson correlation matrix of all the samples was generated by the “cor” function in R v4.2.2, and the “pairwise.complete.obs” method was adopted. Correlation plots were generated by pheatmap v1.0.12. Genes and protein symbols between humans and mice were identified and converted using the “homologene” function. The significance of the DEPs was determined by Student’s t test, and a BH procedure was used to control the FDR. DEPs were defined as those whose absolute value of the log2-fold change was >0.5 and whose adjusted p value was <0.25. Here, we employed an adjusted p value threshold of <0.25 to identify DEPs, aiming to comprehensively characterize the aging phenotype and the Rb1-induced proteomic profile. While a more stringent cutoff could increase the confidence of the results, it may also exclude biologically relevant proteins, especially under complex conditions such as aging. This threshold is consistent with that of similar studies, 62 – 65 which have demonstrated its statistical validity in the field. To address concerns about false positives, we performed permutation tests by randomly shuffling sample labels five times. The results, shown in Supplementary Table S6 , confirmed that the DEPs identified (adjusted p value < 0.25) remained consistent across permutations and were not enriched in any specific pathways, further supporting the reliability of our findings. Enrichment analysis and visualization were performed via GSEA v4.3.2, clusterProfiler v4.6.2, and enrichplot v1.18.3. The function “cnetplot” was used to show the relationships between the key DEPs and pathways. A protein–protein interaction (PPI) network of the DEPs was constructed using STRING v11.5 and Cytoscape v3.8.3 66 to highlight the hub genes as well as their biological interactions. Human endometrial stromal cells were transfected with an NF-κB–responsive firefly luciferase reporter plasmid together with a Renilla luciferase plasmid as an internal control using Lipofectamine 3000. Twenty-four hours after transfection, the cells were subjected to the indicated treatments. Luciferase activity was measured using a dual-luciferase reporter assay system according to the manufacturer’s instructions. Firefly luciferase activity was normalized to Renilla luciferase activity. Each experiment was repeated three times, and the results are expressed as the means ± SDs. For parametric variables, statistical comparisons between two groups were performed using a two-tailed unpaired Student’s t -test, whereas differences among three or more groups were evaluated by one-way ANOVA followed by Tukey’s post-hoc test. P value < 0.05 was considered to indicate statistical significance.

Supplementary Materials Supplementary Materials