SIRT3 attenuates AGEs-induced senescence in human granulosa cells through enhancing mitophagy.

AGEs have been shown to be responsible for a number of ovarian aging-related female subfertility pathogeneses. This work explored the role of AGEs in the senescence of hGCs. hGCs were clinically collected and subsequently treated with AGEs in vitro. First, we found that treating hGCs for 24 and 48 h at different doses (50, 100, and 200 μg/ml) markedly increased SA-β-gal activity in a way that was dependent on both concentration and time (Fig.  1 A and S1 A). AGEs also increased the expression of γ-H2AX, a well-established marker of DNA damage in aging cells (Fig.  1 B, C and S1 B). Moreover, AGEs increased the P16 INK4a , P21, P53, and apoptosis-related activated caspase 3 and Bax protein expressions, both being hallmarks of senescence (Fig.  1 D and S1 C-H). Finally, we investigated how AGEs affected hGC proliferation and hormone production. AGEs were found to suppress the proliferation of hGCs after treatment for 24 h ( P  < 0.01) and 48 h ( P  < 0.001) (Fig.  1 E). After 48 h of exposure to AGEs, the estradiol-17β (Fig.  1 F) and progesterone (Fig.  1 G) productions, the major steroid hormones synthesized by granulosa cells were markedly reduced ( P  < 0.05). Collectively, these data exhibited that AGEs promoted hGC senescence. Fig. 1 Concentration- and time-dependent induction of cellular senescence in hGCs by AGEs. hGCs underwent in vitro treatment with various concentrations of AGEs (50, 100, or 200 μg/ml) for durations of 24 h or 48 h. A  Representative pictures of the stained hGCs were used to evaluate senescence using SA-β-gal staining. Scale bar = 50 μm. B  Immunofluorescence microscopy revealing γ-H2AX expression in hGCs, with representative images displayed. Scale bar = 10 μm. C , D  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through Western blot analysis. E  CCK-8 assay was employed to assess cell proliferation ( n  = 5). F , G  Following 48 h AGEs treatment, ELISA quantification of estradiol-17β and progesterone concentrations was performed ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05, ** P  < 0.01, *** P  < 0.001 versus the control group Concentration- and time-dependent induction of cellular senescence in hGCs by AGEs. hGCs underwent in vitro treatment with various concentrations of AGEs (50, 100, or 200 μg/ml) for durations of 24 h or 48 h. A  Representative pictures of the stained hGCs were used to evaluate senescence using SA-β-gal staining. Scale bar = 50 μm. B  Immunofluorescence microscopy revealing γ-H2AX expression in hGCs, with representative images displayed. Scale bar = 10 μm. C , D  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through Western blot analysis. E  CCK-8 assay was employed to assess cell proliferation ( n  = 5). F , G  Following 48 h AGEs treatment, ELISA quantification of estradiol-17β and progesterone concentrations was performed ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05, ** P  < 0.01, *** P  < 0.001 versus the control group Increasing studies have proved that compromised mitochondrial function and mitophagy play key roles in driving cellular senescence. Consequently, we next investigated whether AGEs-induced hGC senescence was related to changes in mitochondrial function or mitophagy. First, using JC-1 staining, we found that the MMP exhibited a dose-dependent increase in green fluorescence and a corresponding decrease in red fluorescence after AGE treatment, indicating a decrease ( P  < 0.05) in the MMP (Fig.  2 A-C). The MMP continues to be maintained as the prerequisite for ATP production, and its collapse can compromise ATP generation. Consistent with these findings, AGEs treatment significantly reduced the ATP content of hGCs in a way that depends on concentration ( P  < 0.05) (Fig.  2 D). Moreover, a reduction in the MMP results in excessive ROS production. As expected, AGEs treatment markedly enhanced ROS production in hGCs ( P  < 0.01) (Fig.  2 E-G and S2 A). Mitochondrial outer membrane permeabilization is a feature of cellular senescence that facilitates the release of mitochondrial DNA (mtDNA) into the cytosol. Here, we detected specific mtDNA (D-loop region) by quantitative PCR and detected increased mtDNA in the cytosolic fractions of hGCs after AGEs treatment ( P  < 0.05) (Fig.  2 H). Together, AGEs primarily impaired mitochondrial function via lowering MMP, increasing intracellular ROS production, and promoting mitochondrial membrane permeabilization. Fig. 2 Impairment of mitochondrial function and mitophagy in hGCs following AGEs exposure. hGCs underwent in vitro treatment with AGEs at 200 μg/ml for a 48 h duration. JC-1 fluorescent probe was utilized to assess MMP. A  Representative fluorescence intensity data were captured using laser confocal microscopy, scale bar = 20 μm and ( B , C ) Flow cytometric analysis ( n  = 4). D  Quantification of ATP levels in hGCs subjected to AGEs treatment ( n  = 4). Detection of intracellular ROS was performed through laser confocal microscopy, scale bar = 50 μm ( E ) and flow cytometric methods ( F , G ) (n = 4). H  Quantification of mtDNA in cytosolic fractions ( n  = 3). I  Western blot examination revealed protein levels of LC3A/B, P62, PINK1, and Parkin within mitochondrial fractions. The data are presented as the average plus or minus the SEM. Compared to the control group, * P  < 0.05, ** P  < 0.01 Impairment of mitochondrial function and mitophagy in hGCs following AGEs exposure. hGCs underwent in vitro treatment with AGEs at 200 μg/ml for a 48 h duration. JC-1 fluorescent probe was utilized to assess MMP. A  Representative fluorescence intensity data were captured using laser confocal microscopy, scale bar = 20 μm and ( B , C ) Flow cytometric analysis ( n  = 4). D  Quantification of ATP levels in hGCs subjected to AGEs treatment ( n  = 4). Detection of intracellular ROS was performed through laser confocal microscopy, scale bar = 50 μm ( E ) and flow cytometric methods ( F , G ) (n = 4). H  Quantification of mtDNA in cytosolic fractions ( n  = 3). I  Western blot examination revealed protein levels of LC3A/B, P62, PINK1, and Parkin within mitochondrial fractions. The data are presented as the average plus or minus the SEM. Compared to the control group, * P  < 0.05, ** P  < 0.01 Damaged mitochondria can typically be degraded via mitophagy to maintain a healthy mitochondrial population. Mitophagy impairment blocks the turnover of dysfunctional mitochondria and accelerates premature senescence. To look at how AGEs affect mitophagy, we examined the protein levels of key autophagy markers LC3A/B and P62, as well as those of the critical mitophagy-related regulators PINK1 and Parkin. AGEs raised the amount of the autophagic substrate P62 in the mitochondrial fraction while suppressing the expression of Parkin, PINK1, and LC3A/B (Fig.  2 I and S2 B-E). We further examined mitochondrial ultrastructure using transmission electron microscopy (TEM). Our results revealed that mitochondria exhibited fragmentation or cristae disruption upon AGE treatment. Moreover, the fusion of mitochondria with autophagosomes and autolysosomes was significantly inhibited, indicating an impairment of mitophagy (Fig. S2 F). These results collectively indicated that AGEs impaired mitochondrial function and mitophagy in hGCs. To further explore the intrinsic connections between mitophagy and the senescence of AGEs-treated hGCs, we modulated mitophagy using the potent mitophagy activator UA and the commonly used mitophagy inhibitor CsA on the basis of AGEs. Our results revealed that UA attenuated the AGEs-induced reduction in the MMP ( P  < 0.01) (Fig.  3 A-C) and ATP ( P  < 0.05) (Fig.  3 D) content, whereas CsA further exacerbated these effects. We then looked at how UA and CsA affected the amount of intracellular ROS and mtDNA released. Our results revealed that UA decreased the high levels of intracellular ROS and cytosolic mtDNA caused by AGEs, whereas CsA further increased the intracellular ROS levels ( P  < 0.01) (Fig.  3 E-G and S3 A) and mtDNA release ( P  < 0.05) (Fig.  3 H). Moreover, UA reversed the AGE-induced rise in P62 expression and the reduction in LC3 II, PINK1, and Parkin protein expression in the mitochondrial fraction, whereas CsA had the opposite effect (Fig.  3 I and S3 B-E). TEM analysis further confirmed that UA treatment improved mitochondrial morphology and restored the fusion of mitochondria with autophagosomes (Fig. S3 F). Fig. 3 Impact of UA and CsA on mitochondrial functionality and mitophagy in hGCs. hGCs underwent treatment with AGEs at 200 μg/ml for a 48 h period, either by itself or in conjunction with CsA (1 μM) or UA (20 μM). A  Representative JC-1 staining fluorescence patterns were assessed via laser confocal microscopy, scale bar = 20 μm and B , C  Flow cytometric analysis ( n  = 4). D  Quantification of ATP levels in AGEs-exposed hGCs co-treated with UA or CsA ( n  = 4). Intracellular ROS production was evaluated using laser confocal microscopy, scale bar = 50 μm ( E ) and flow cytometric detection ( F , G ) ( n  = 4). H  Cytosolic mtDNA quantification ( n  = 3). I  Western blot evaluation demonstrated expression levels of LC3A/B, P62, PINK1, and Parkin proteins in mitochondrial fractions. The data are presented as the average plus or minus the SEM. * P  < 0.05, ** P  < 0.01 Impact of UA and CsA on mitochondrial functionality and mitophagy in hGCs. hGCs underwent treatment with AGEs at 200 μg/ml for a 48 h period, either by itself or in conjunction with CsA (1 μM) or UA (20 μM). A  Representative JC-1 staining fluorescence patterns were assessed via laser confocal microscopy, scale bar = 20 μm and B , C  Flow cytometric analysis ( n  = 4). D  Quantification of ATP levels in AGEs-exposed hGCs co-treated with UA or CsA ( n  = 4). Intracellular ROS production was evaluated using laser confocal microscopy, scale bar = 50 μm ( E ) and flow cytometric detection ( F , G ) ( n  = 4). H  Cytosolic mtDNA quantification ( n  = 3). I  Western blot evaluation demonstrated expression levels of LC3A/B, P62, PINK1, and Parkin proteins in mitochondrial fractions. The data are presented as the average plus or minus the SEM. * P  < 0.05, ** P  < 0.01 Considering that the most popular mitophagy mechanism in mammalian cells is PINK1-Parkin-mediated, we next specifically silenced PINK1 gene via the shRNA method (Fig.  4 A) to investigate the role of PINK1-Parkin signalling in AGEs-treated hGCs. Like CsA, PINK1 knockdown exacerbated AGEs-induced mitochondrial dysfunction, as evidenced by a further reduction in the MMP ( P  < 0.05) (Fig.  4 B-D) and ATP content ( P  < 0.05) (Fig.  4 E). PINK1 knockdown also increased intracellular ROS levels ( P  < 0.05) (Fig.  4 F-I) and increased cytosolic mtDNA release ( P  < 0.05) (Fig.  4 J). All of these findings pointed to the critical role that mitophagy plays in preserving mitochondrial functioning in hGCs treated with AGEs. Fig. 4 PINK1 knockdown exacerbates mitochondrial dysfunction and impairs mitophagy in AGEs-exposed hGCs. Following PINK1 silencing, hGCs underwent incubation for 48 h with or without AGEs at a concentration of 200 μg/ml. A  Western blot validation confirmed the efficiency of PINK1 suppression via shRNA. Representative JC-1 staining fluorescence images were captured through laser confocal microscopy, scale bar = 20 μm ( B ) and flow cytometric assessment ( C , D ) ( n  = 3). E  Quantification of ATP production in AGEs-exposed hGCs ( n  = 3). Intracellular ROS generation was measured using laser confocal microscopy ( F , G ) and flow cytometric analysis ( H , I ) ( n  = 3). Scale bar = 50 μm. J  Cytosolic mtDNA content determination ( n  = 3). The data are presented as the average plus or minus the SEM. * P  < 0.05, *** P  < 0.001 PINK1 knockdown exacerbates mitochondrial dysfunction and impairs mitophagy in AGEs-exposed hGCs. Following PINK1 silencing, hGCs underwent incubation for 48 h with or without AGEs at a concentration of 200 μg/ml. A  Western blot validation confirmed the efficiency of PINK1 suppression via shRNA. Representative JC-1 staining fluorescence images were captured through laser confocal microscopy, scale bar = 20 μm ( B ) and flow cytometric assessment ( C , D ) ( n  = 3). E  Quantification of ATP production in AGEs-exposed hGCs ( n  = 3). Intracellular ROS generation was measured using laser confocal microscopy ( F , G ) and flow cytometric analysis ( H , I ) ( n  = 3). Scale bar = 50 μm. J  Cytosolic mtDNA content determination ( n  = 3). The data are presented as the average plus or minus the SEM. * P  < 0.05, *** P  < 0.001 We next dissected the impacts of mitophagy modulation by UA and CsA on AGEs-induced hGC senescence. Similarly, reduced SA-β-gal staining indicated that UA reduced AGE-induced hGC senescence (Fig.  5 A, B), decreased γ-H2AX fluorescence intensity and expression (Fig.  5 C, D, and S3 G), and decreased P21, P16 INK4a , and P53 protein levels (Fig.  5 E and S3 H-J). In contrast, CsA exacerbated the expression of these senescence markers. In addition, UA ameliorated AGEs-induced hGC apoptosis, whereas CsA exacerbated it upon AGEs treatment (Fig. S3 K-M). Finally, the decrease of progesterone ( P  < 0.05) and estradiol-17β ( P  < 0.05) production caused by AGEs was dramatically restored by UA, whereas CsA had a synergistic inhibitory effect on steroidogenesis (Fig.  5 F, G). We next examined the effects of PINK1 knockdown on AGEs-triggered senescence in hGCs. PINK1 gene silencing also increased the expression of SA-β-gal staining (Fig.  6 A and B) and γ-H2AX (Fig.  6 C-E) induced by AGEs. Additionally, P21, P16 INK4a , and P53 protein expression was increased by PINK1 silencing (Fig.  6 F-I). In conclusion, mitophagy had a negative correlation with the AGE-induced senescence of hGCs. This senescence process can be mitigated by activating mitophagy or aggravated by inhibiting it, underscoring the pivotal role of mitophagy in regulating cellular senescence. Fig. 5 Impact of UA and CsA treatment on hGC senescence. For 48 h, hGCs were incubated with or without UA (20 μM) or CsA (1 μM) supplementation, either in the presence or absence of AGEs at 200 μg/ml. A  Representative SA-β-gal staining micrographs demonstrating hGC senescence detection. Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs showing γ-H2AX expression in hGCs. Scale bar = 10 μm. D , E  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. F , G  Estradiol-17β and progesterone levels were quantified using ELISA methodology ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05 Fig. 6 PINK1 knockdown exacerbated senescence in AGEs-exposed hGCs. Following PINK1 silencing, hGCs were subjected to treatment with or without AGEs (200 μg/ml) for a 48 h duration. A  Representative SA-β-gal staining micrographs for hGC senescence assessment ( n  = 3). Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs depicting γ-H2AX in hGCs. Scale bar = 10 μm. D , F  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. E , G - I  Densitometric quantification of γ-H2AX, P16 INK4a , P21, and P53 relative expression with β-actin normalization ( n  = 3). The data are presented as the average plus or minus the SEM. * P  < 0.05 Impact of UA and CsA treatment on hGC senescence. For 48 h, hGCs were incubated with or without UA (20 μM) or CsA (1 μM) supplementation, either in the presence or absence of AGEs at 200 μg/ml. A  Representative SA-β-gal staining micrographs demonstrating hGC senescence detection. Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs showing γ-H2AX expression in hGCs. Scale bar = 10 μm. D , E  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. F , G  Estradiol-17β and progesterone levels were quantified using ELISA methodology ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05 PINK1 knockdown exacerbated senescence in AGEs-exposed hGCs. Following PINK1 silencing, hGCs were subjected to treatment with or without AGEs (200 μg/ml) for a 48 h duration. A  Representative SA-β-gal staining micrographs for hGC senescence assessment ( n  = 3). Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs depicting γ-H2AX in hGCs. Scale bar = 10 μm. D , F  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. E , G - I  Densitometric quantification of γ-H2AX, P16 INK4a , P21, and P53 relative expression with β-actin normalization ( n  = 3). The data are presented as the average plus or minus the SEM. * P  < 0.05 The dynamics and function of mitochondria are crucially regulated by SIRT3, one of the most significant deacetylases. We next explored whether SIRT3 was involved in AGEs-induced mitochondrial impairment and cellular senescence in hGCs. First, we observed significant downregulation of SIRT3 protein expression in both granulosa cells from aged patients and AGEs-treated hGCs (Fig.  7 A, B and S4 A, B). To investigate SIRT3's function in more detail, we next silenced SIRT3 via the shRNA method, and the silencing efficiency was confirmed by immunoblotting (Fig. S4C). Our results demonstrated that SIRT3 silencing further reduced the MMP ( P  < 0.05) (Fig.  7 C-E) and the intracellular ATP content ( P  < 0.05) (Fig.  7 F). However, SIRT3 silencing increased ROS production ( P  < 0.05) (Fig.  7 G-I and S4 D) and cytosolic mtDNA levels in AGEs-treated hGCs ( P  < 0.05) (Fig.  7 J). To evaluate the effect of SIRT3 silencing on mitophagy, we also measured the LC3 II, P62, Parkin, and PINK1 protein levels in the mitochondrial fraction. Efficaciously, in the presence of AGEs, SIRT3 silencing increased the expression of P62 while further suppressing the LC3 II, Parkin, and PINK1 protein levels (Fig.  7 K and S4 E-H), indicating that SIRT3 controls mitophagy by acting upstream of the PINK1-Parkin pathway. It was also shown by the TEM data that SIRT3 knockdown further exacerbated the disruption of mitochondrial ultrastructure upon AGEs treatment, which presented as mitochondrial swelling and cristae damage, along with a diminished fusion between mitochondria and autolysosomes (Fig. S4 I). Taken together, these findings showed that SIRT3 silencing worsened the mitochondrial damage and mitophagy suppression brought on by AGEs in hGCs. These results highlighted the protective roles of SIRT3 in maintaining mitochondrial function and integrity under AGEs-induced stress. Fig. 7 SIRT3 knockdown further compromised mitochondrial function and mitophagy in AGEs-exposed hGCs. A  SIRT3 protein levels in hGCs derived from young patients (Y, ≤ 35 years, n  = 4) and aged patients (A, ≥ 38 years, n  = 4). B  hGCs underwent treatment with various concentrations AGEs (50, 100, or 200 μg/ml) for a 48 h period. SIRT3 protein levels were determined via western blot. Representative JC-1 fluorescence intensity was assessed using laser confocal microscopy, scale bar = 20 μm ( C ) and flow cytometry ( D , E ) ( n  = 3). ( F ) ATP production in AGEs-exposed hGCs following SIRT3 silencing ( n  = 3). Cellular ROS accumulation was evaluated by laser confocal microscopy, scale bar = 50 μm ( G ) and flow cytometry H , I  ( n  = 3). J  mtDNA abundance in the cytosolic compartment ( n  = 3). K  Western blot determination of LC3A/B, P62, PINK1 and Parkin protein levels in the mitochondrial compartments. The data are presented as the average plus or minus the SEM. * P  < 0.05 SIRT3 knockdown further compromised mitochondrial function and mitophagy in AGEs-exposed hGCs. A  SIRT3 protein levels in hGCs derived from young patients (Y, ≤ 35 years, n  = 4) and aged patients (A, ≥ 38 years, n  = 4). B  hGCs underwent treatment with various concentrations AGEs (50, 100, or 200 μg/ml) for a 48 h period. SIRT3 protein levels were determined via western blot. Representative JC-1 fluorescence intensity was assessed using laser confocal microscopy, scale bar = 20 μm ( C ) and flow cytometry ( D , E ) ( n  = 3). ( F ) ATP production in AGEs-exposed hGCs following SIRT3 silencing ( n  = 3). Cellular ROS accumulation was evaluated by laser confocal microscopy, scale bar = 50 μm ( G ) and flow cytometry H , I  ( n  = 3). J  mtDNA abundance in the cytosolic compartment ( n  = 3). K  Western blot determination of LC3A/B, P62, PINK1 and Parkin protein levels in the mitochondrial compartments. The data are presented as the average plus or minus the SEM. * P  < 0.05 Our above results highlight the crucial roles of SIRT3 in maintaining mitochondrial function and mitophagy. Considering the central role of mitochondria in controlling cellular senescence, next, in order to investigate the possible impact of SIRT3 on AGEs-induced hGC senescence, we inhibited the expression of the SIRT3 gene. SIRT3 knockdown enhanced SA-β-gal staining (Fig.  8 A, B) and γ-H2AX expression (Fig.  8 C, D and S5 A). In addition, SIRT3 knockdown further raised P21, P16 INK4a , and P53 protein levels (Fig.  8 E and S5 B-D). SIRT3 knockdown further promoted hGC apoptosis induced by AGEs (Fig. S5 E-G). Moreover, contrasted with the sh-scrambled comparative group, the production of progesterone ( P  < 0.05) and estradiol-17β ( P  < 0.05) was further lowered after SIRT3 silencing (Fig.  8 F and G). Overall, these results confirmed that SIRT3 silencing exacerbated AGEs-induced senescence in hGCs. Fig. 8 SIRT3 knockdown exacerbated the senescence phenotype of AGEs-exposed hGCs. hGCs underwent treatment with or without AGEs (200 μg/ml) for a 48 h duration following SIRT3 silencing. A  Images of SA-β-gal staining that are representative for the evaluation of hGC senescence. Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs of γ-H2AX in hGCs. Scale bar = 10 μm. D , E  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. F , G  The levels of estradiol-17β and progesterone were quantified by ELISA ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05 SIRT3 knockdown exacerbated the senescence phenotype of AGEs-exposed hGCs. hGCs underwent treatment with or without AGEs (200 μg/ml) for a 48 h duration following SIRT3 silencing. A  Images of SA-β-gal staining that are representative for the evaluation of hGC senescence. Scale bar = 50 μm. B  Assessing the level of SA-β-gal positive in cells quantitatively ( n  = 3). C  Representative immunofluorescence micrographs of γ-H2AX in hGCs. Scale bar = 10 μm. D , E  γ-H2AX, P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. F , G  The levels of estradiol-17β and progesterone were quantified by ELISA ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05 Our above results revealed that SIRT3 could alleviate AGEs-triggered hGC senescence by regulating mitophagy and mitochondrial function. To further evaluate the protective roles of SIRT3 in mitigating hGC senescence, we generated stable SIRT3-overexpressing hGC lines. Successful SIRT3 overexpression was confirmed by western blot (Fig.  9 A). We first investigated the effects of SIRT3 overexpression on mitochondrial function and mitophagy under AGEs treatment. As expected, SIRT3 overexpression improved mitochondrial function, as demonstrated by increased mitochondrial membrane potential (Fig. S6 A, B), reduced ROS production (Fig. S6 C, D), elevated ATP levels (Fig. S6 E), and a decrease in mtDNA released into the cytosol (Fig. S6 F). Furthermore, SIRT3 overexpression concurrently activated mitophagy, as seen by the mitochondrial fractions' decreased P62 expression and elevated LC3 II, PINK1, and Parkin protein levels (Fig.  9 B and S7 A-D). Consistently, TEM revealed that SIRT3 overexpression ameliorated mitochondrial ultrastructure and enhanced the fusion of mitochondria with autophagosomes (Fig. S7 E). In light of these results, we then investigated how SIRT3 overexpression affected AGE-induced senescence in hGCs. Cellular senescence indicators, such as SA-β-gal staining (Fig.  9 C, D) and γ-H2AX intensity in the nucleus (Fig.  9 E), were suppressed by SIRT3 overexpression. Furthermore, in AGE-treated hGCs, SIRT3 overexpression inhibited the protein levels of P16 INK4a , P21, and P53 (Fig.  9 F and S7 F-H). Additionally, SIRT3 overexpression inhibited AGE-induced apoptosis (Fig. S7 I-K). Significantly, AGE-inhibited production of progesterone ( P  < 0.05) and estradiol-17β ( P  < 0.05) (Fig.  9 G and H) was recovered by overexpression of SIRT3. In conclusion, SIRT3 activated mitophagy to reduce AGE-induced hGC senescence. Fig. 9 SIRT3 upregulation attenuated the senescence phenotype of AGEs-exposed hGCs. hGCs underwent treatment with or without AGEs (200 μg/ml) for a 48 h duration following SIRT3 overexpression. A  The upregulation of SIRT3 was confirmed by western blot analysis. B  Western blot determination of LC3A/B, P62, PINK1, and Parkin protein levels in the mitochondrial fractions. C  Images of SA-β-gal staining that are representative for the purpose of evaluating hGC senescence. Scale bar = 50 μm. D  Evaluation of SA-β-gal positive cells after treatment with AGEs using quantitative methods ( n  = 3). E  Representative immunofluorescence micrographs of γ-H2AX in hGCs. Scale bar = 10 μm. F  P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. G , H  The levels of estradiol-17β and progesterone were quantified by ELISA ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05, n.s., not significant SIRT3 upregulation attenuated the senescence phenotype of AGEs-exposed hGCs. hGCs underwent treatment with or without AGEs (200 μg/ml) for a 48 h duration following SIRT3 overexpression. A  The upregulation of SIRT3 was confirmed by western blot analysis. B  Western blot determination of LC3A/B, P62, PINK1, and Parkin protein levels in the mitochondrial fractions. C  Images of SA-β-gal staining that are representative for the purpose of evaluating hGC senescence. Scale bar = 50 μm. D  Evaluation of SA-β-gal positive cells after treatment with AGEs using quantitative methods ( n  = 3). E  Representative immunofluorescence micrographs of γ-H2AX in hGCs. Scale bar = 10 μm. F  P16 INK4a , P21, and P53 protein expression were evaluated through western blot analysis. G , H  The levels of estradiol-17β and progesterone were quantified by ELISA ( n  = 4). The data are presented as the average plus or minus the SEM. * P  < 0.05, n.s., not significant

AGEs solution was purchased from Abcam (ab51995, USA). Urolithin A (UA, HY-100599) and cyclosporine A (CsA, HY-B0579) were purchased from MedChemExpress (USA). To create stock solutions, UA and CsA were first dissolved in dimethyl sulfoxide. The KGN cell line was obtained from the NHC Key Laboratory of Study on Abnormal Gametes and Reproductive Tract. China's Cobioer Biosciences Co., Ltd. is the supplier of the HEK293T cell line (CBP60439). hGCs were obtained from 477 patients (≤ 35 years) who were undergoing IVF treatment for infertility due to tubal or male factors at our reproductive medicine center. Individuals with infertility issues due to endometriosis, PCOS, or endocrinopathies were excluded. The Ethical Review Board of the First Affiliated Hospital of Anhui Medical University examined and authorized this research (No. 2022097), and all recruited patients gave their informed consent. hGCs were purified using the controlled ovarian stimulation protocol and method as previously reported (Xue et al. 2021 ). To summarize, women were given recombinant human follicle-stimulating hormone (Gonal-F, Merck KGaA, Germany) and gonadotropin-releasing hormone (GnRH) (Dephereline, Ipsen Pharma Biotech, France) for superovulation. Patients received 10,000 IU of hCG (AESCA Pharma, Austria) for ovulation induction once three or more follicles had grown to a size of 18 mm in diameter. 36 h after hCG injection, transvaginal ultrasound-guided follicle retrieval was carried out. Density centrifugation was then used to separate the hGCs from follicular aspirates collected from individuals undergoing oocyte extraction. After combining patient follicular fluid samples, they were centrifuged for 10 min at 400 × g. After that, Hank's buffer supplemented with 0.2% hyaluronidase and 50 μg/ml deoxyribonuclease I was used to resuspend the pellet. Then, it was incubated for 30 min at 37 °C. Ficoll-Paque Plus (GE Healthcare, Buckinghamshire, UK) was lightly coated with the suspension before being centrifuged for 20 min at 600 × g. After being extracted from the interphase, the hGC layer was rinsed three times with phosphate-buffered saline (PBS). The isolated primary hGCs were used for the next experiments. For assessing SIRT3 protein expression, hGCs from both young (≤ 35 years) and aged (≥ 38 years) patients were derived from individual donors. The cell lines of hGCs, KGN, and HEK293T were grown in a humidified incubator (37 °C, 5% CO 2 ). The medium included a penicillin–streptomycin combination (100 IU/ml and 100 μg/ml, respectively) (15,140,122, Gibco, USA) and a fetal bovine serum (10%, 10,100,147, Gibco, USA). Based on earlier research, the AGE concentrations (50–200 μg/ml) employed to treat hGCs were chosen (Chen et al. 2017 ; Merhi et al. 2018 ; Yao et al. 2018 ). For some experiments, purified hGCs were incubated with UA (20 μM) or CsA (1 μM) for one hour before treating AGEs. For the detection of the concentrations of estradiol-17β and progesterone, hGCs were treated with AGEs and FSH (50 ng/ml) for 48 h before ELISA test. MMP was evaluated by choosing the lipophilic cationic JC-1 Dye ( M34152 , Thermo Fisher, USA). Intracellular ROS production was quantified employing the DCFH-DA (D399, Invitrogen, USA). In brief, 0.1% poly-L-lysine (P6282, Sigma-Aldrich, USA) was used to precoat glass-bottom plates (D35-10–1.5-N, Cellvis, USA) before hGCs were seeded. After testing, the cells were stained with either DCFH-DA (10 μM) or JC-1 dye (1 μM), and they were then incubated at 37℃ for 20 min in the dark. PBS-Tween-20 (0.1%) was used twice to rinse the cells, the cellular fluorescence was quantified using a confocal microscopy system (LSM800, Carl Zeiss, Germany), and the results were analyzed via ImageJ software. For the test using flow cytometry, the cells in each group were retrieved and incubated with JC-1 (1 μM) or DCFH-DA (10 μM) at 37℃ for 20 min in the dark. The cells underwent three PBS washes before being subjected to flow cytometry using a FACSVerse device (BD Biosciences, USA). Finally, the fluorescence intensity was analyzed via FlowJo software (Version 7.6, Tree Star). Cells exhibiting JC-1 red-negative fluorescence were identified as having collapsed mitochondrial membrane potential and were included in the analysis for quantitative analysis. The average DCF fluorescence intensity within the cells was quantified to represent the intracellular ROS level. In accordance with the supplier's instructions, cell viability was evaluated using a CCK-8 kit (C0038, Beyotime Biotechnology, China). After being seeded (1 × 10 4 ) in 96-well plates, hGCs were cultivated for 24 or 48 h under the specified conditions. After that, each well containing 100 μl of culture media was pipetted with 10 μl of CCK-8 reagent, and the mixture was incubated (37 °C, 2 h). We measured the absorbance at 450 nm using a microplate reader (800 TS, BioTek Instruments, USA). hGCs were seeded onto glass-bottom culture dishes (D35-10–1.5-N, Cellvis, USA), previously coated with 0.1% poly-L-lysine (P6282, Sigma-Aldrich). Following the experimental procedure, before being fixed with 4% paraformaldehyde for 15 min at room temperature, the cells were rinsed with PBS. The cell membrane was then permeabilized by treatment with Triton-X 100 (0.1%, 15 min), followed by blocking with a bovine serum albumin solution (1%, 4 °C, 1 h). A primary antibody against γ-H2AX (C2036S, Beyotime Biotechnology) was added to the cells after they were rinsed with 0.1% PBS-Tween-20. The cells were then incubated overnight at 4 °C. Subsequently, cells were stained using a Cy3-labeled secondary antibody. Following two washes with 0.1% PBS-Tween-20, the nuclei were stained with DAPI Fluor mount-G (0100–20, SouthernBiotech, USA). Fluorescence was detected using a confocal microscope (LSM800, Carl Zeiss, Germany), and the data were analyzed with ImageJ software (Version 1.42q). As directed by the manufacturer, the intracellular ATP levels were measured using a firefly luciferase. Following the aforementioned procedures, 2 × 10 5 cells were broken down, and the supernatant was collected by centrifugation (5 min, 12,000 × g). Next, a 96-well plate was filled with 100 μl of ATP detection reagent and a 20 μl portion of the supernatant. Detection of absorbance was performed with an enzyme marker (BioTek Instruments, USA). The standard curve was also generated at the same time, and the total ATP levels were normalized to those of the control group. Following the manufacturer's instructions, a commercial β-galactosidase staining kit (C0602, Beyotime Biotechnology, China) was used to detect the presence of SA-β-gal activity. In summary, hGC cells were fixed using paraformaldehyde (4%) after being washed three times with PBS (15 min, room temperature). The cells were cultured with a freshly made SA-β-gal staining working solution at 37 °C in the dark overnight in a CO 2 -free incubator after three PBS washes. Before being examined under a microscope, the cells were twice washed with PBS after incubation. Senescent cells were identified using bright-field microscopy when they showed blue cytoplasmic staining. ImageJ software was used to quantify SA-β-gal-positive cells (Version 1.42q). In the end, the findings were given in the form of the number of positive cells found in each optical field of view. After being seeded into 96-well plates with 1 × 10 5 hGCs per well, the cells were stimulated with FSH (50 ng/ml) for 48 h under the specified treatment conditions. Following that, the various groups' cultural media were collected. Following the kit protocol's directions, the amounts of progesterone (JEN-02, Joyee Biotechnics, China) and estradiol-17β (JEN-01, Joyee Biotechnics, China) in the supernatants were measured. Briefly, all reagents and samples were equilibrated to room temperature prior to use. The standards provided in the kit and samples were added in duplicate (90 μl per well) to the antibody-coated microplate. Subsequently, 10 μl of reconstituted horseradish peroxidase (HRP)-conjugated estradiol-17β or progesterone was pipetted into each well. The plate was incubated for two hours at room temperature in the dark after being gently swirled for ten seconds. Following incubation, each well was filled with 100 μl of tetramethylbenzidine substrate solution, and after that, the combination was left overnight at room temperature and in the dark for twenty minutes. A microplate reader was used to rapidly detect absorbance at 450 nm after 50 μl of stop solution was added to each well to halt the process. The minimum detectable concentration of progesterone with this kit was 0.01 ng/ml, with intra- and interassay variability maintained below 10% and 15%, respectively. Similarly, for estradiol-17β detection, the analytical sensitivity reached 12 pg/ml, and for both intra-assay and inter-assay measurements, the coefficients of variation were maintained below 10% and 15%, respectively. The knockdown of SIRT3 and PINK1 was performed with a lentivirus-mediated delivery system. Briefly, HEK293T cells were transfected with a lentiviral vector encoding SIRT3 shRNA (6 μg) or PINK1 shRNA (6 μg) utilizing Lipofectamine 3000 (L3000015, Invitrogen) in accordance with the manufacturer's instructions, together with PMD2.G (0.5 μg) and PsPAX2 (1.5 μg). After 48 h of transfection, the supernatant containing the lentivirus was collected and passed through a 0.45 μm filter. To inhibit SIRT3 or PINK1 expression, hGCs were seeded the day before transfection in 6-well plates at a density of 3 × 10 5 cells/well. The cells were then transduced with lentivirus-containing media for 48 h. To create SIRT3 -knockdown or PINK1 -knockdown cell lines, the culture medium was changed after infection to a new complete medium with puromycin (2 μg/ml) for continuous selection over a period of seven days. A non-targeting control shRNA was subjected to the same procedures to establish a control cell line. The efficiency of SIRT3 and PINK1 knockdown was verified by a western blot experiment. The sequence of sh SIRT3 is 5′-CCGGCCCAACGTCACTCACTACTTTCTCGAGAAAGTAGTGAGTGACGTTGGGTTTTTG-3′. The sequence of sh PINK1 is 5′-CCGGGCCGCAAATGTGCTTCATCTACTCGAGTAGATGAAGCACATTTGCGGCTTTTT-3′. A scrambled shRNA (5′-TTCTCCGAACGTGTCACGT-3′) was used as the negative control. For SIRT3 overexpression in hGCs, the pcDNA3.1-SIRT3 plasmid was constructed following established methods (Yang et al. 2010 ). Total RNA was isolated from KGN cells, and RT-PCR was used to create the full-length cDNA of SIRT3. The primers specific for human SIRT3 are as follows: forward, 5′-CCGCGGTACCATGGCGTTCTGGGGTTG-3′; reverse, 5′-CCGCTCTAGACTATTTGTCTGGTCCATCAAGC-3′. On the day prior to transfection, purified hGCs were plated at a density of 3 × 10 5 cells per well in 6-well plates. The cells were then transfected with 6 μg of pcDNA3.1-SIRT3 plasmid using Lipofectamine 3000 (L3000075, Invitrogen), following the manufacturer’s protocol. An empty pcDNA3.1 plasmid served as a negative control. Following transfection, 800 μg/ml G418 (A1720, Sigma-Aldrich, USA)-containing media was used to select the positively transfected cells. Successful SIRT3 expression in the transfected hGCs was verified through western blot analysis. Isolations of the cytosolic, mitochondrial and nuclear fractions from hGCs were performed as previously described (Dias et al. 2020 ). Then, the cytosolic DNA was purified via the phenol–chloroform extraction method. In brief, the cytosolic pellet was resuspended in TRIzol reagent (15596018CN, Invitrogen, USA) and incubated (5 min, room temperature). Chloroform was then added in a 5:1 ratio with TRIzol, followed by an additional 5-min incubation. Following a centrifugation (12,000 × g, 4 °C, 15 min), the supernatant containing RNA was carefully extracted. After that, the solution was gently mixed by inversion with pure ethanol (TRIzol: EtOH 3.3:1) and incubated for 5 min. To remove the DNA, the samples were spun at 2000 × g and 4 °C for 5 min, and then the liquid above was removed. The next step was to resuspend the DNA pellet in a sodium citrate buffer with a concentration of 0.1 M and a pH of 8.5. It was then let to settle for half an hour. After being cleaned with 75% ethanol, the samples were centrifuged twice for 5 min at 4 °C and 2,000 × g. The DNA pellet was then resuspended in PCR buffer after being allowed to air dry. The GoTaq qPCR Master Mix kit (A6001, Promega, USA) was used to perform quantitative RT-PCR. The 2 −ΔΔCt technique was then used to assess the relative amount of mtDNA, and the results were normalized to the GAPDH level. The following are the particular primer sequences that were utilized: human mtDNA forward, 5′-CGAAAGGACAAGAGAAATAAGG-3′; reverse, 5′-CTGTAAAGTTTTAAGTTTTATGCG-3′; and human GAPDH forward, 5′-CCACCATGGAGAAGGCTGGGGC-3′; reverse, 5′-AGTGATGGCATGGACTGTGGTC-3′. TEM was performed to analyze mitochondrial ultrastructure and the formation of mitophagosomes. After being collected and washed with PBS, hGCs exposed to experimental treatments were set at 4 °C in half a milliliter of glutaraldehyde buffer (2.5%, pH 7.4) overnight in order to preserve fragile intracellular structures. After fixation, post-fixation was then carried out using 2% osmium acid at 4 °C for 2 h. The samples underwent a series of graded ethanol dehydration procedures after being cleaned with PBS. They were then implanted in epoxy resin and polymerized for two days at 60 °C to form stiff blocks. Uranyl acetate and lead citrate were used to stain the ultrathin sections, which were 45 nm thick, that were gathered on copper grids. Lastly, transmission electron microscopy was used to examine the intracellular ultrastructural features (JEOL, Tokyo, Japan). Proteins derived from hGCs or mitochondrial fractions were isolated using RIPA lysis buffer (89,901, Thermo Scientific, USA) supplemented with a protease inhibitor cocktail. The isolated proteins were then put onto polyvinylidene difluoride membranes after being separated using 10% SDS-PAGE. Afterward, the membranes were incubated with primary antibodies diluted in TBST (Tris-buffered saline containing 0.1% Tween-20) overnight at 4 °C after being blocked for two hours at room temperature in 5% nonfat milk. The primary antibodies used included: 1:1000 for anti-P16 INK4a (18,769, CST, USA), anti-P21 (2947, CST, USA), anti-P53 (2524, CST, USA), anti-γ-H2AX (2577, CST, USA), anti-PINK1 (6946, CST, USA), anti-LC3A/B (12,741, CST, USA), anti-P62 (5114, CST, USA), anti-Caspase-3 (14,220, CST, USA), anti-Cleaved Caspase 3 (9664, CST, USA), anti-Bax (5023, CST, USA), anti-SIRT3 (sc-365175, Santa Cruz, USA), and anti-Parkin (1:1000, 14,060–1-AP, Proteintech, China), and 1:2000 for anti-Tom20 (42,406, CST, USA) and 1:5000 for anti-β-actin (3700, CST, USA). After three washes with TBST, the membranes were left at room temperature for one hour to incubate with a secondary antibody that was conjugated with HRP. A chemiluminescence reagent with increased signal detecting capabilities was used ( P90720 , Millipore, USA) and an imaging system with quantified bands was used (28,955,810, GE Healthcare, USA) for band visualization. β-actin and Tom20 were utilized as internal loading controls for cytosolic and mitochondrial protein samples, respectively. All statistical analyses were performed using GraphPad Prism software (Version 8.01). Mean plus or minus the standard error of the mean (SEM) is how the data are shown. For comparisons between two groups, the one-tailed Mann–Whitney test was used, and for differences among multiple groups, one-way ANOVA followed by Tukey's multiple-comparison post hoc test was used. Three separate experiments were conducted for each experiment to guarantee repeatability. Statistical significance was defined as P  < 0.05.

Ovarian aging manifests earlier than in most other organ systems in humans and plays a pivotal role in determining overall health and longevity (Babayev, & Duncan 2022 ; Ruth et al. 2021 ). Granulosa cells are essential for supporting oocyte growth and developmental competence by producing steroid hormones, secreting autocrine and paracrine factors (Hennet, & Combelles 2012 ; Uyar et al. 2013 ). Notably, mounting data suggest that granulosa cell senescence contributes significantly to the onset of ovarian aging. However, the factors and precise mechanisms that drive hGC senescence have not yet been fully illuminated. Our current study demonstrated that AGEs promoted hGC senescence by impairing mitochondrial function and inhibiting mitophagy. Moreover, we identified SIRT3 as a key regulator that attenuates hGC senescence and dysfunction by enhancing mitophagy and maintaining mitochondrial homeostasis. Our findings highlight SIRT3 as a potential therapeutic target for mitigating AGEs-induced cellular senescence and improving age-related female infertility. AGEs are polymers formed through Maillard reactions between the carbonyl groups of carbohydrates and the primary amino groups of proteins; these polymers exhibit enzymatic stability and resistance to degradation (Unoki, & Yamagishi 2008 ). Extensive studies have revealed that age-related tissue accumulation of AGEs is correlated with various pathophysiological diseases, including female reproductive dysfunction (Tatone, & Amicarelli 2013 ). The proportion of high-quality embryos and the amount of ovarian reserve are negatively correlated with the concentration of AGEs in follicular fluid (Jinno et al. 2011 ; Yao et al. 2018 ). AGEs exert their detrimental effects by promoting oxidative stress, protein damage, and inflammation, all of which contribute to cellular senescence. In the present study, AGEs treatment resulted in mitochondrial dysfunction and impaired mitophagy, ultimately driving cellular senescence in vitro. Our findings align with those of an earlier investigation showing that AGEs induce bone marrow mesenchymal stem cell (BMSC) senescence by leading to mitochondrial dysfunction (Guo et al. 2021 ). Our present research provides novel insights into the detrimental effects of AGEs accumulation on female fertility. The main way that AGEs cause harm is by attaching themselves to certain receptors for AGEs, which are also expressed on granulosa cells (Pertynska-Marczewska, & Diamanti-Kandarakis 2017 ). However, the molecular mechanisms by which the AGEs-RAGE axis promotes ROS production in hGCs remain to be further studied, although several different pathways have been investigated in other cell types (Zhou et al. 2024 ). Furthermore, as ROS is required for the final stage of glycation, increased ROS levels brought on by AGEs might further promote the development of further AGEs, resulting in a vicious cycle of mitochondrial damage and oxidative stress (Giacco, & Brownlee 2010 ). This cycle may accelerate hGC senescence and disrupt cellular functions. Mitochondrial dysfunction increases during aging, which in turn contributes to aggravated cellular senescence and age-related diseases (Miwa et al.  2022 ). Our results that AGEs compromise MMP, ATP production, and elevated intracellular ROS and cytosolic mtDNA are consistent with previous studies of age-related mitochondrial decay (Zhang et al. 2025a , b ). More importantly, we establish a clear link between this mitochondrial dysfunction and a specific defect in mitophagy-associated quality control. Impaired mitophagy leads to less mitochondrial turnover and accumulation of dysfunctional mitochondria, which further results in cellular senescence and age-related disorders. For example, reduced mitophagy in granulosa cells and oocytes has been linked to female reproductive dysfunction during aging (Huang et al. 2022 ; Shen et al. 2021 ). In this study, mitophagy was significantly inhibited in AGEs-treated hGCs, as seen by elevated P62 in the mitochondrial fraction, reduced PINK1, Parkin, and LC3 II protein levels, and compromised mitochondria-autophagosome fusion. This is significant because while general mitochondrial decline has been noted in aging ovaries, the specific impairment of mitophagy in hGCs under AGE stress has been less clear. The activation of mitophagy with UA improved mitochondrial function, reduced hGC senescence, and restored hormone production, whereas CsA exerted the opposite effects by inhibiting the mitophagy process. These results highlighted the importance of mitophagy in preventing hGC senescence, which aligns with several previous studies (Guo et al. 2021 ; Kataura et al. 2024 ; Kelly et al. 2024 ; Ma et al. 2023 ). On the other hand, by upsetting mitochondrial homeostasis, excessive mitophagy may cause cell death (Ma et al. 2020 ; Yang et al. 2024 ). Thus, the process of mitophagy should be precisely regulated in clinical applications. Moreover, AGEs are known to activate inflammatory signaling, endoplasmic reticulum stress, and other forms of autophagy or cell death pathways (Zhang et al. 2025a , b ), which may concurrently influence cellular senescence. Future studies employing more comprehensive pathway analysis or combinatorial interventions are needed to fully delineate the relative contributions of mitophagy versus other mechanisms. The best-understood signaling route for mitophagy in animals is PINK1-Parkin. Following the collapse of the mitochondrial potential, PINK1 stabilizes on the mitochondrial surface and then attracts Parkin to the mitochondria. Parkin then initiates selective mitophagy by ubiquitinating the outer membrane proteins of malfunctioning mitochondria. Mutations in PINK1 or Parkin that cause loss of function damage mitochondria and are directly linked to aging-related disorders and cellular senescence (Kelly et al. 2024 ; Lu et al. 2023 ). In this investigation, AGEs caused hGCs to lose their mitochondrial membrane potential. PINK1 and Parkin protein levels in mitochondria declined in tandem, suggesting that PINK1-mediated mitophagy initiation was compromised. Moreover, inhibition of mitophagy via PINK1 knockdown exacerbated AGEs-induced mitochondrial dysfunction and hGC senescence. These results further emphasized the protective effects of mitophagy and the essential roles of PINK1-Parkin signalling in AGEs-treated hGCs. In addition, mitochondrial fragmentation is a prerequisite for mitophagy initiation, which allows the segregation of damaged mitochondria via fission followed by lysosomal degradation (Twig et al. 2008 ). Importantly, PINK1 can modulate the mitochondrial fission/fusion machinery (Yang et al. 2008 ). Thus, PINK1-regulated mitochondrial dynamics may additionally contribute to the regulation of hGC senescence in addition to recruiting Parkin. Furthermore, several protein receptors, including BNIP3, FUNDC1 and NIX/BNIP3L, have also been demonstrated to mediate mitophagy in mammalian cells (Liu et al. 2014 ). Therefore, more studies are needed to investigate whether these receptors that mediate mitophagy are involved in protecting hGCs against senescence induced by AGEs or other toxic metabolites. The most significant novel contribution of our work lies in identifying SIRT3 as a key upstream regulator of this protective mitophagy in hGCs. While SIRT3′s roles in controlling mitochondrial metabolism (Hirschey et al. 2010 ; Wang et al. 2019 ) and regulating metabolic stress (Diao et al. 2021 ; Guo et al. 2021 ; Marcus, & Andrabi 2018 ; Zhu et al. 2022 ) have been recognized, its function in hGCs, particularly under AGEs challenge, was unexplored. In this investigation, AGE-treated hGCs and hGCs from elderly people showed a substantial reduction in SIRT3 expression. The excessive impairment of mitochondrial function and mitophagy might have been due to decreased expression of SIRT3 in these hGCs, as SIRT3 silencing further deteriorated mitochondrial function and mitophagy. In addition, SIRT3 deficiency further promoted hGC senescence. In contrast, SIRT3 overexpression markedly enhanced mitophagy and suppressed hGC senescence in vitro, indicating that SIRT3 could modulate mitochondrial homeostasis and hGC function by regulating the mitophagy process. Moreover, the primary locations for steroid hormone production are mitochondria (Miller 2013 ). SIRT3 overexpression also corrected the AGE-induced reduction in progesterone and estradiol-17β synthesis. These results further confirmed the crucial roles of SIRT3 in maintaining hGC function, mainly by regulating mitochondrial homeostasis. However, we investigated only SIRT3, which is an important sirtuin that regulates mitochondrial function and mitophagy. In addition to SIRT3, other sirtuin family members that have been implicated in protecting cells from accelerated senescence, such as SIRT1 and SIRT6 (Chen et al.  2020 ; Guo et al.  2022 ), should also be studied in the future. Despite these revelations, our research has a number of shortcomings. First, the research was conducted exclusively in vitro. Thus, the physiological relevance of AGEs and the therapeutic potential of targeting SIRT3 and mitophagy remain to be validated in vivo. Second, while this study demonstrated that SIRT3 overexpression enhances PINK1-Parkin-mediated mitophagy, the direct regulatory relationships within the SIRT3-PINK1-Parkin axis have not been completely clarified. By altering the acetylation state of PINK1 and Parkin, SIRT3 has been shown to directly control their expression (Liu et al. 2025 ). Alternatively, other studies have found that SIRT3 may first deacetylate transcription factors FOXO1 and FOXO3a, which in turn upstream regulate the PINK1 signalling pathway (Xi et al. 2024 ). The underlying mechanisms in our study remain to be further studied. Finally, besides the mitophagy activator UA, we did not explore other specific pharmacological interventions targeting SIRT3, which could have important clinical implications for treating age-related ovarian dysfunction. To verify the function of the SIRT3-mitophagy axis in preventing ovarian aging, further research should fill these gaps. In summary, the mechanisms by which SIRT3 attenuates AGEs-induced hGC senescence have been elucidated at least in part. Specifically, SIRT3 restored mitochondrial function and homeostasis by enhancing mitophagy, which is associated with the upregulation of PINK1 and Parkin. The activated mitophagy likely removed damaged mitochondria in hGCs to prevent cellular senescence (Fig.  10 ). Our current study not only confirmed the formerly unrecognized role of SIRT3 in protecting hGCs from senescence but also clarified that pharmacological regulation of SIRT3 or mitophagy could be an effective strategy for therapeutic intervention in aging-related ovarian dysfunction. Fig. 10 By controlling mitophagy and mitochondrial activity, SIRT3 mitigates the senescence of hGCs caused by AGEs By controlling mitophagy and mitochondrial activity, SIRT3 mitigates the senescence of hGCs caused by AGEs

The ovary undergoes early age-associated dysfunction in humans, with significant decreases in follicle numbers and oocyte quality, which results in an increased incidence of female infertility, spontaneous miscarriages, and congenital anomalies (Wang et al. 2020 ; Wen et al. 2020 ). Human granulosa cells (hGCs) in the follicular region are located in the immediate vicinity of oocytes and play key roles in oocyte maturation, competency, and function (Hsueh et al. 2015 ; Matsuda et al. 2012 ; Uyar et al. 2013 ). Importantly, there is mounting evidence that granulosa cell senescence hinders follicle growth, resulting in follicular atresia and ultimately ovarian dysfunction (Dumesic et al. 2015 ; Tatone, & Amicarelli 2013 ). Cellular senescence is a stable and irreversible cessation of the cell cycle in response to various stresses. However, current studies on the factors involved in hGC senescence are indirect and inadequate. The stable substances that result from non-enzymatic glycation processes are known as advanced glycation end products (AGEs). Previous evidence has shown that AGEs accumulate with advancing age and contribute to the development of age-related diseases (Tessier 2010 ). According to recent research, AGEs accumulate in the follicular microenvironment during normal aging and compromise oocyte developmental competence by inducing excessive ROS production and protein damage (Merhi 2014 ; Pertynska-Marczewska, & Diamanti-Kandarakis 2017 ). Furthermore, elevated AGEs play significant roles in impairing various functions of hGCs by inhibiting steroidogenesis synthesis and reducing glucose intake (Diamanti-Kandarakis et al. 2016 ; Merhi et al. 2018 ). However, a direct association between AGEs accumulation and hGC senescence, as well as the potential underlying mechanisms, has not been demonstrated. Mitophagy, a selective process of degrading superfluous or damaged mitochondria via autophagy, governs mitochondrial recycling and turnover to maintain cellular homeostasis (Ajoolabady et al. 2020 ). Growing evidence has shown that mitophagy deficiency contributes to various types of cellular senescence and aging-related diseases by disrupting mitochondrial function and quality (Guo et al. 2021 ; Picca et al. 2023 ). Notably, several studies have indicated that increased mitophagy can ameliorate female reproductive disorders associated with aging. For example, mitophagy can be enhanced by nicotinamide mononucleotide treatment in mouse granulosa cells, thereby ameliorating age-related diminished ovarian reserve (Huang et al. 2022 ). Similarly, mitophagy can also be activated by spermidine to rejuvenate oocyte quality in aged mice (Zhang et al. 2023 ). These findings highlight mitophagy as a potential therapeutic target for protecting granulosa cells against senescence in aging females. However, the specific roles of mitophagy in AGEs-induced hGC senescence are still elusive. The seven sirtuins (SIRT1–7) are a family of highly conserved histone deacetylases that are reliant on NAD + in mammals, with different subcellular locations, targets, and regulatory mechanisms (Kupis et al. 2016 ). Among these seven sirtuins, SIRT3 is predominantly localized in mitochondria, where it regulates the deacetylation of mitochondrial proteins to play important roles in modulating mitochondrial homeostasis (Marcus, & Andrabi 2018 ). Moreover, numerous age-related illnesses, including osteoporosis, osteoarthritis, neurodegenerative disorders, and cardiovascular diseases, are linked to SIRT3 dysregulation (Grabowska et al. 2017 ; McDonnell et al. 2015 ). For example, senescent human mesenchymal stem cells and mouse kidney tubular epithelial cells showed reduced SIRT3 expression. The restoration of SIRT3 could counteract senescence phenotypes and aging-associated degeneration (Diao et al. 2021 ; Yang et al. 2022 ). Genetic variations in the SIRT3 gene have also been linked to human longevity (Yang et al. 2007 ). In addition, SIRT3 deficiency accelerates mouse ovarian senescence by disrupting the ovarian follicle reserve and the quality of oocytes (Zhu et al. 2022 ). Nevertheless, the potential effects of SIRT3 in regulating mitochondrial function and cellular senescence in hGCs remain largely unexplored. In summary, AGEs act as a potential upstream stressor that may impair mitochondrial quality control, a process in which mitophagy serves as a critical regulatory step. However, it remains unknown whether and how AGEs directly induce hGC senescence, and whether this process involves the disruption of mitophagy. Furthermore, although SIRT3 is recognized for its protective roles in aging and mitochondrial function, its specific function in regulating mitophagy and cellular senescence in hGCs under AGEs exposure has not been established. The goal of the current research was to look at how SIRT3 affects mitochondrial homeostasis and the fate of hGCs under AGEs treatment in vitro. Apart from emphasizing the potential benefits of SIRT3 or mitophagy as therapeutic targets to decrease granulosa cell senescence and enhance ovarian function in clinical practice, we anticipate that our study will provide fresh insights into the processes behind ovarian aging.

Below is the link to the electronic supplementary material. ESM 1 (DOCX 3.54 MB) (DOCX 3.54 MB)