The eight herbs used in HSYC ( Table 1 ) were purchased from the Hangzhou Hospital of Traditional Chinese Medicine. After soaking, boiling with distilled water, concentrating to a thick slurry, and dried to powder format with a crude dose of 2.25 g/ml. The resulting HSYC powder was dissolved in water and used for murine intragastric administration.

The research was conducted in accordance with all relevant ethical regulations. Institutional Animal Care, the Animal Experiment Center of Zhejiang Chinese Medical University, approved all the experimental procedures used in this study (Approval number: IACUC-20180402-02).

Sixty C57/BL6 female mice aged 6–8 weeks were obtained from Shanghai Laboratory Animal Company (SLACCAS, license no. SCXK 2017-0005, Shanghai, China). Mice were raised under standard conditions for a week before the formal experiment in a barrier facility with stable specific-pathogen-free levels, temperatures, humidity, and light cycles in the Animal Experimental Research Center of Zhejiang Chinese Medical University (license no. SYXK 2021-0012, Zhejiang, China). Research was performed according to the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. The 60 mice were randomly distributed into six groups: CON, CTX, HSYC-L, HSYC-M, HSYC-H, and Q10 groups. Based on published literature and our pre-experiment result, the POI model was injected intraperitoneally with cyclophosphamide (CTX) at 90 mg/kg (7), and the success rate was determined by observing the estrous cycle for 7 days. Once the POI model was established, the Q10 group were intragastrically administered coenzyme Q10 (Eisai, Tokyo, Japan) at a dose of 1.25 mg/kg, while mice in the HSYC-L, HSYC-M, and HSYC-H groups were given HSYC by intragastric administration at doses of 6 g/kg, 12 g/kg, and 25 g/kg at 8 AM daily, according to the conversion of human clinical dosage.

Mice in the CON and CTX groups were administered saline solution by gavage. Duration of treatment was 3 weeks. Within 48 hours of PMSG injection, CO₂ inhalation was used to euthanize the mice and blood and ovaries were collected for subsequent use. Actually, the modeling success rate was about 70%, 30 mice being successfully established (n=6).

Fresh murine ovarian tissues were placed in a culture dish. F12 media was then added and the follicles were punctured with a sterile 1 mL syringe needle to release the GCs into the culture media. GCs were gently blown, filtered with a 200 mesh sieve, and centrifuged at room temperature for 3 min at 1000 ×g. The GCs were then transferred to a new sterile petri dish and incubated for 48 hours at 37°C with 5% CO₂. The nonadherent cells were removed and the media changed every 72 h.

Serum samples obtained before the mice were euthanized were centrifuged at room temperature for 20 min at 3000 ×g. Serum hormone levels, including anti-Müllerian hormone (AMH), estradiol (E2), progesterone (P), luteinizing hormone (LH), and FSH; and OS related products, SOD, GSH-Px, and malondialdehyde (MDA), were measured using ELISA kits (Meimian, Shanghai, China), according to the manufacturer’s instructions. These included Mouse E2 ELISA kit, Mouse LH kit, Mouse FSH kit, Mouse P kit, Mouse AMH kit, Mouse SOD kit, Mouse GSH-Px kit, and Mouse MDA kit.

The total antioxidant capacity (T-AOC) of serum samples was also assessed. Serum samples, standards, and distilled water blank controls were prepared according to the T-AOC assay kit (Aidisheng, Jiangsu, China). The absorbance value was then determined at 590 nm in a CMaxPlus microplate reader (Molecular Devices, San Jose, CA, USA).

ROS content was determined by flow cytometry using 2’-7’-dichlorodihydrofluorescein diacetate (DCFH-DA) (Beyotime, Shanghai, China). DCFH-DA was diluted with serum-free medium at 1:1000, with a final concentration of 10 μmol/L. Diluted DCFH-DA (1 mL) was added to the cells of each group after removing the cell culture media and incubated at 37 °C for 20 min. Cells were then washed three times with serum-free cell culture solution and prepared for flow cytometric detection. GC mitochondrial membrane potential (MMP) levels were detected using the JC-1 assay kit (Beyotime). Cell culture media were removed in each group and 1 mL incomplete culture medium and 1 mL JC-1 working solution was added after washing with PBS. Negative and positive control groups were set according to the instructions. All groups were incubated at 37 °C in the dark for 20 min. GCs were then digested, collected, centrifuged, washed twice with JC-1 staining buffer, and resuspended before flow cytometry (C6, BD, USA) was performed.

A 4% paraformaldehyde solution was used for fixation of the cells, 0.5% Triton X-100 for permeabilization, and 3% bovine serum albumin for blocking. Anti-LC3 primary antibody (Cell Signaling Technology, Danvers, MA, USA) was added, the solution was incubated overnight, and the secondary antibody was added and incubated in dark for 30 min. After staining the nucleus with 4’,6-diamidino-2-phenylindole, the sections were sealed with an anti-fluorescence quenching agent and visualized using a Ts2-FC fluorescence microscope (Nikon, Tokyo, Japan).

Mito Tracker Red CMXRos (Beyotime, Shanghai, China) solution (200 µM) was added to the cell culture media at a ratio of 1:2000 to pre-incubate at 37 °C and subsequently, 500 µL diluted Mito Tracker Red CMXRos working solution was added to each well. Cell culture media was then infused with Hoechst solution, and incubated for 10 minutes without light. A Ts2-FC inverted fluorescence microscopy was used to visualize the mitochondrial morphology (Nikon).

The cells were collected and fixed in glutaraldehyde and by osmic acid, rinsed, gradient ethanol dehydrated, and embedding agent penetrated and sectioned. After staining with uranium acetate and lead citrate, the sections were visualized and images captured under transmission electron microscope (TEM) (H-7650, HITACHI, Tokyo, Japan).

Following paraffin embedding and sectioning (4 μm), ovarian tissues were dehydrated with xylene and ethanol, stained with hematoxylin and eosin, sealed and visualized under an Eclipse E100 microscope (Nikon). The ovarian structure was observed and the number of follicles at all levels were counted.

Following RNA extraction by the Trizol standard method (Thermo Fisher Scientific, Waltham, MA, USA), RNA was reverse transcribed in a thermocycler (Mastercycler; Eppendorf, Hamburg, Germany) using the cDNA reverse transcriptase kit (CWBiotech, Beijing, China), according to the manufacturer’s instructions. The cDNA was then amplified using the fluorescence-quantitative PCR kit (TAKARA, Shiga, Japan), according to the manufacturer’s instructions. The cycling conditions were as follows: 95 °C, 10 min; 95 °C, 15 s; 60 °C, 60 s; for 40 cycles. The reactions were performed using the CFX Connect Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA), and the gene expression levels were calculated by the 2^-△△CT method. The primer sequences (5′ to 3′) of target genes are shown in Table 2 .

After addition of radioimmunoprecipitation assay lysis buffer (Beyotime), tissues were ground in a homogenizer and then lysed on ice for 1 h. The protein concentration was then determined using the BCA Kit (Solarbio, Beijing, China). The proteins were denatured with loading buffer at 100 °C for 5 min. Following protein gel electrophoresis, membrane transfer, blocking, and incubation of primary antibody and secondary antibody, photographic developer was added to the membrane and the band was developed in ECL chemiluminescence instrument with β-actin as internal reference. The antibodies as follows: anti-PINK1 (Affinity Biologicals, Ancaster, ON, Canada; DF7742; 1:1,000), anti-Parkin (Affinity Biologicals; AF0235; 1:1,000), anti-LC3 (Cell Signaling Technology; 4108s; 1:1,000), anti-p62 (Cell Signaling Technology; 8025s; 1:1,000), anti-Beclin1 (Affinity Biologicals; AF5128; 1:1,000), anti-NLRP3 (Affinity Biologicals; DF7438; 1:1,000), anti-GSDMD, anti-IL-18 (Affinity Biologicals; AF4012; 1:1,000), anti-caspase-1 (Proteintech, Rosemont, IL, USA; 22915-1-AP; 1:1,000), anti-IL-1β (Proteintech; 16806-1-AP; 1:1,000), and anti β-actin (Affinity Biologicals; AF7018; 1:2,000).

Study data are presented as means ± standard deviations (SD) and were analyzed statistically using one-way analysis of variance and Dunnett’s test. A statistically significant difference was defined as a P-value less than 0.05 (P<0.05), and P-value less than 0.01 (P<0.01) was considered highly significant.

Follicle development and sex hormone levels predict ovarian function. Compared with the CON group, the E2, P and AMH serum levels in the CTX group were significantly decreased (P<0.01); however, FSH and LH levels were higher than those the CON group. Following medium-dose HSYC, high-dose HSYC, or Q10 intervention; E2, P, and AMH were significantly increased (P<0.05 or P<0.01), while FSH and LH levels were decreased compared to the CTX group (P<0.05 or P<0.01) ( Figures 2A-E ).

The ovaries in the CON group were well-developed with visible follicles at all grades. In contrast, those of the CTX group showed atrophy, and the number of primary and secondary follicles was significantly reduced compared with the CON group (P<0.01). Furthermore, GC arrangement in mature follicles remained disturbed. Follicles of different stages and corpora lutea could be observed in all treatment groups, especially in HSYC-H and Q10 groups where the cumulus GCs in mature follicles were neatly arranged radially ( Figure 2F ). According to the follicle count results, the number of primary follicles was significantly higher in most of the treatment groups compared to the CTX group (P<0.05 or P<0.01); however, the number of primordial and mature follicles did not differ significantly between these groups (P≥0.05) ( Table 3 ).

As is shown in Figure S1A, B , the isolated GCs (mainly consists of GCs but not purified) grew and adhered the wall. After 24 hours of culture, GCs begin to adhere to the wall and proliferate. After 48h of culture, the cells proliferated significantly and grew vigorously. It can be seen under the microscope that the adherent cells are complete in shape, clear in edge, uniform in size, polygonal or spindle shaped, the nucleus is blue with oval shape and the cytoplasm is light pink by hematoxylin and eosin staining.

In the flow cytometry result, mean fluorescence intensity represented ROS expression. The ROS level was significantly increased in the CTX group compared to the CON group (P<0.01); while ROS production was inhibited in different treatment groups, including the HSYC-M, HSYC-H, and Q10 groups (P<0.05 or P<0.01) ( Figures 3A, B ).

To investigate the level of OS, the amount of SOD, GSH-Px, MDA, and T-AOC in serum were determined. There was a significant increase in SOD and GSH-Px levels in the CTX group compared to the CON group (P<0.01). Conversely, the MDA level was higher than that in the CON group (P<0.01). Compared to the CTX group, SOD and GSH-Px were increased in HSYC-H, HSYC-M and Q10 groups, while MDA decreased (P<0.05 or P<0.01). Additionally, T-AOC exhibited a significant decrease in the CTX group compared with that in the CON group (P<0.01), and was considerably elevated in the HSYC-M, HSYC-H, and Q10 groups compared to the CTX group (P<0.05 or P<0.01) ( Figures 3C-F ).

To explore the protective effect of HSYC on mitochondrial dysfunction, we investigated the MMP level in GCs by flow cytometry using JC-1 and mitochondrial staining. Q2 represented cells with normal mitochondrial membrane potential, while Q3 represented cells with decayed mitochondrial membrane potential. Compared with the CON group, the mitochondrial membrane potential reduction rate of GCs in the CTX group was significantly increased (P<0.01). However, in the HSYC-M, HSYC-H, and Q10 groups the mitochondrial membrane reduction rate was decreased (P<0.01), indicating that the cellular mitochondrial membrane potential was considerably elevated ( Figures 4A, B ). Similarly, mitochondria of GCs in the CON group exhibited generally tubular-reticular morphology, as confirmed by mitochondrial staining analyses, and the CTX group exhibited spherical mitochondria. Compared with the CTX group, mitochondria were longer and displayed a reticular distribution in each treatment group ( Figure 4C ).

The quantification data of LC3 fluorescent intensity indicated that expression levels in the GCs of the CTX group were significantly higher than those in the CON group (P<0.01). The LC3 expression levels in the HSYC-M and HSYC-H groups were lower than those in the CTX group (P<0.05 or P<0.01). A similar trend was observed in the Q10 group. ( Figures 5A, B )

To evaluate the autophagy level, we observed the ultrastructure of autophagosomes and autophagolysosomes of GCs under TEM. The GCs in the CON group had clear nuclei and abundant organelles, such as endoplasmic reticulum and mitochondria. Autophagosomes and autophagolysosomes were present in lesser numbers. Conversely, there was a sharp increase in the number of those in the CTX group (P<0.01). HSYC-M, HSYC-H, or Q10 treatment led to a considerable reduction in the number of autophagosomes and autophagolysosomes compared with the CTX group (P<0.05 or P<0.01), which indicated an improvement in the degree of autophagy ( Figures 5C, D ).

We assessed the mRNA and protein expression of mitophagy-related genes using qPCR and WB, respectively. In the CTX group, the mRNA and protein expression levels of PINK1, Parkin, LC3, and Beclin1 were higher, whereas p62 was lower compared with the CON group (P<0.01). With HSYC-M, HSYC-H, or Q10 intervention, the expression levels of PINK1, Parkin, LC3, and Beclin1 were significantly decreased, and p62 were significantly increased compared with the CTX groups (P<0.05 or P<0.01) ( Figure 6 ). The original bands are presented in Supplementary Figures S2–7 .

mRNA and protein expression levels of NLRP3 inflammasome were detected by qPCR and WB analyses, respectively. Results indicated that the NLRP3, caspase-1, GSDMD, IL-18, and IL-1β mRNA expression in the CTX group was significantly higher than that of the CON group (P<0.01). Compared with the CTX group, NLRP3, caspase-1, GSDMD, IL-18, and IL-1β mRNA expression was significantly lower in the HSYC-M, HSYC-H, and Q10 treatment groups (P<0.05 or P<0.01). WB analyses showed similar results ( Figure 7 ). The original bands are presented in Supplementary Figure S2, 8-12 .