Introduction

Premature ovarian insufficiency (POI) is a condition that affects women of reproductive age and is characterized by irregular ovulation, premature amenorrhea, and the ovary becoming non-functional1,2. Cyclophosphamide is one widely used clinically alkylating agent antitumor drug, and prolonged use of the medication can cause amenorrhea and ovarian fibrosis in females, which is a frequent cause of POI in clinical practice3,4,5. Stem cells have been widely used in the field of regenerative medicine due to their ability to differentiate into multiple lineages, low immunogenicity, and lack of ethical concerns6,7. Though it has been reported that stem cell treatment can restore ovarian function in women with POI8,9, the underlying mechanism remains unclear.

Based on the cases that have been made public, the decline in ovarian function in POI is associated with a reduced supply and eventual depletion of the primordial follicle pool10. The mammalian ovary consists of a certain number of follicles, to create numerous layers of spindle-shaped cells encircling the follicle’s outer border, TCs are attracted and differentiated from ovarian stromal cells11,12. TCs synthesize and release androgens in response to enzymes such as steroidogenic acute regulatory protein (StAR), 17beta-hydroxysteroid dehydrogenases (17βHSD), and prohibitin (PHB)13,14,15. These androgens infiltrate granulosa cells through the basement membrane, leading to the production of estrogen. Finally, estrogen enters the blood circulation system and controls the function of target organs such as the endometrium16.

Mitochondria are the site of terminal catabolism of cellular energy substrates, providing the cell with ATP necessary for metabolism. Mitochondria fuse and divide in response to the changing demands of cellular energy17. Mitochondrial dynamics are regulated by mitochondrial fusion proteins mitofusin 1 (Mfn1), mitofusin 2 (Mfn2), and mitochondrial fission proteins dynamin-related protein 1 (Drp1) embedded in the mitochondrial membrane18,19. It is reported that specific knockdown of Drp1 in mouse oocytes suggests that mitochondrial fission affects follicle maturation and ovulation, regulating calcium signaling and cell-to-cell communication20. Whether changes in mitochondrial dynamics in the TCs in patients with POI are involved in impaired ovarian function requires further investigation.

The highly conserved serine/threonine kinase, GSK3β was first discovered as a regulator of glycogen synthesis and as a catabolic/anabolic pathway coordinator21. Recent studies have shown that GSK3β influences mitochondrial biological processes22. Li et al. thought that the curcumin-hUMSCs combination therapy reduced inflammation and oxidative stress and improved neurological function following acute ischemic stroke (AIS) through the AKT/GSK3β/-TrCP/Nrf2 axis mediated anti-inflammatory microglial polarization23. Although hUMSCs were capable of repairing structural and functional damage to the ovary in POI, according to our prior study24,25. Further research is required to determine its connection to GSK3β-induced aberrant mitochondrial metabolism.

The purpose of this study was to evaluate whether hUMSCs transplantation could restore ovarian function in POI mice. At the same time, to explore whether GSK3β is involved in the process of hUMSCs transplantation improving TCs mitochondrial dynamics imbalance and thus affecting hormone synthesis in POI mice.

Materials and methods

Animals

We purchased female C57BL/6J mice from Jinan Pengyue Experimental Animal Breeding, Ltd (Shandong, China). The mice were 6–7 weeks old and weighed between 16–20 g. All animals were kept in the animal facility, fed a typical pellet diet, and had unrestricted access to water. The investigation was carried out in compliance with the recommendations of the National Laboratory Animal Care and Use Research Committee. This study was performed following protocols approved by the Binzhou Medical University Institutional Animal Care and Use Committee (permit number: 2022-371).

Chemicals and reagents

The anti-cancer chemotherapy drug CTX was purchased from Meilun Biotechnology Co., LTD (MB1315-1, Meilunbio, Dalian, China). The GSK3β inhibitor SB26763 (HY-12012), Phosphoramide mustard (HY-137316A) was purchased from MCE corporation. The DAPI (C0065) and Dimethyl sulfoxide (DMSO) (D8371) were obtained from Solarbio Biotechnology (Beijing, China). GSK3β (22104-1-AP), StAR (12225-1-AP), PHB (60092-1-Ig), and GAPDH (60004-1-Ig) were obtained from Procell Life Science and Technology Co., Ltd (Wuhan, China). 17βHSD (A1983) and Mfn1 (A9880) were obtained from ABclonal corporation (Wuhan, China). Mfn2 (1:50, 9482S) and Drp1 (8570S) were obtained from CST corporation (Massachusetts, USA). The ELISA kit of estradiol (E2, mlE98523), follicle-stimulating hormone (FSH, ml001910), luteinizing hormone (LH, ml063366), testosterone (ml001948) were obtained from Shanghai Enzyme-linked Biotechnology Co., Ltd.

Establishment of mice POI model

Mice were randomly assigned to one of the following four groups (N = 12/group): (1) Control, (2) POI, (3) PBS, or (4) hUMSCs. In the Control group, 100 μL of PBS containing 0.01% DMSO was injected intraperitoneally on days 1–14. Based on and adapted from the previously reported protocol, CTX was dissolved in 0.01% DMSO and administered via an intraperitoneal injection to generate the in vivo POI mouse model26,27. The regimen for CTX involved a 75 mg/kg dose on day 1 followed by daily doses of 4 mg/kg on days 2–14 (Fig. 2A). On day 15, the PBS group received 150 μL of phosphate-buffered saline (PBS) injected into the tail vein while the hUMSCs group received 150 μL of PBS containing 5 × 105 hUMSCs injected into the tail vein28. For the hUMSCs injection, it was necessary to avoid an excessive injection speed which could lead to death from congestive heart failure. The weights were measured and recorded every 3 days during the establishment and treatment of the POI mouse model.

One week after these tail vein injections, mice were anesthetized with pentobarbital sodium (40 mg/kg, i.p.) and collected blood from the eye socket. The blood was allowed to stand at room temperature for 2 h to fully coagulate. The serum was collected by centrifugation at 3000×g for 15 min at 4 °C. The ovaries were harvested, some of them were fixed in 4% paraformaldehyde for histological evaluation, and the others were snap-frozen in liquid nitrogen and stored at − 80 °C for subsequent protein and mRNA analysis.

Ovarian index

At the time of mouse sampling, fresh ovaries on both sides of the mice were removed and blotted out on filter paper to absorb the remaining water. Then placed ovaries on an electronic balance to weigh the wet weight of the ovaries and calculate the ovarian index. Ovarian index = wet weight of bilateral ovaries (g)/body mass of mice before death (g) × 100%.

hUMSCs culture and identification

The hUMSCs were acquired from Procell Life Science and Technology Co., Ltd. (CP-CL11, Wuhan, China). When the third passage of hUMSCs were confluent to about 70%, the cell morphology was observed under a light microscope. We employed flow cytometry (FCM) to detect the surface marker molecules CD73, CD90, CD44, HLA-DR, CD34, and CD45 on the hUMSCs. FCM was performed using a FACSCanto II cytometer (BD Biosciences, Franklin Lakes, NJ, USA). FCM analyses were performed using the FlowJo software (Version 10.8). Differentiation studies were carried out as previously described29. When the hUMSCs reached 70 and 100% confluency, the medium was removed and then adipogenic and osteogenic differentiation medium was added, respectively. All differentiation processes were in strict accordance with the kit instructions (Wei Tong Biotechnology, Shenzhen, China). For adipogenic staining, the cells were 4% paraformaldehyde fixed, and lipid vacuoles were tested with Oil Red O after 14 days in culture. For osteocyte staining, 4% paraformaldehyde-fixed cells were tested for calcium with Alizarin Red after 28 days.

TCs culture

The TCs were incubated in a dedicated complete medium, which were purchased from Procell Life Science and Technology Co., Ltd (CP-M205, Wuhan, China). Cells were then cultured in a 5% CO2 incubator with saturated humidity at 37 °C. Cells growth was monitored daily, and cell passages were conducted every 3 days. To investigate the effect of CTX treatment on TCs, we established CTX injury models. In the CTX group, normal TCs were treated with CTX for 24 h. In the hUMSCs group, normal TCs received a 24-h treatment of CTX and hUMSCs supernatant. In the SB21673 group, normal TCs were pretreated with SB216763 for 2 h and then treated in the same way as CTX group (Fig. 5C).

Enzyme-linked immunosorbent assay (ELISA)

Serum samples from each mouse were collected, centrifuged and stored at − 80 °C. Serum levels of E2, FSH, LH and testosterone in TCs culture supernatant were measured using ELISA kits according to the instructions.

Hematoxylin and eosin (H&E) staining

Collected ovarian tissues were fixed with 4% formaldehyde for 24 h and embedded in paraffin, then cut into 5 μm sections and taken on slides. The tissues section was stained with hematoxylin and eosin for 30 min at room temperature. All photographs were taken with a Nikon Eclipse 80i microscope (Nikon Corporation, Tokyo, Japan). The ovarian morphology and the number of follicles were assessed with double-blind.

Ovarian follicle count method refers to previous literature reports30,31. Follicles with only flattened pregranulosa cells surrounding the oocyte were defined as primordial follicles. Follicles with a single layer of granulosa cells surrounding an oocyte and at least one cuboidal granulosa cell were classified as primary follicles. Follicles were classified as secondary follicles when there was at least one granulosa cell starting to form a second layer in the section analyzed. Atretic follicles were irregular in shape, with severe nuclear asymmetry and pyknosis in the oocytes.

Immunofluorescent (IF) staining

Expressions of GSK3β, StAR, 17βHSD, PHB, Mfn1, Mfn2, and Drp1 proteins were determined using immunofluorescent staining. Ovarian tissue and TCs samples were subjected to frozen sections and fixed, then incubated at 4 °C overnight with the following primary antibodies: anti-GSK3β (1:100), anti-StAR (1:100), anti-17βHSD (1:50), anti- PHB (1:200), anti-Mfn1 (1:100), anti-Mfn2 (1:50) or anti-Drp1 (1:100). On the following day, these ovarian tissue samples and TCs were first maintained at room temperature for 1 h and then incubated for 1 h with the following secondary antibodies: goat anti-Rabbit IgG, Alexa Fluor 488 (A23220, Abbkine, China), or goat anti-Rabbit IgG, Alexa Fluor 594 (A23420, Abbkine, Wuhan, China). Ovarian tissue samples and TCs were then incubated with DAPI at room temperature in the dark for 10 min. Stainings were then visualized using an inverted fluorescent microscope (Leica, Germany). Image J software was used to analyze the mean fluorescence intensities (MFI).

Quantitative real-time polymerase chain reaction (qRT-PCR)

mRNA was measured using a quantitative real-time PCR assay. The housekeeping gene GAPDH was used as an internal control. Data from a minimum of three replicates were calculated using the 2−ΔΔCt method. Primers are attached in Supplementary Table 1.

CCK-8 cell viability assay

The CCK-8 kit (BS350C, Biosharp, Anhui, China) was used to evaluate the effects of different concentrations of cyclohexanamine and SB216763 on TCs viability. Cells (5 × 103 cells/well) were seeded onto 96-well plates. After the cells were adherent and grown to 80%, the medium was replaced with one containing CTX (0–70 μM). 24 h later, the CCK-8 working solution was added to the cell culture medium for 1 h. Absorbance values were read at 450 nm, and cell viability was determined according to the kit instructions. In the intervention experiment, TCs cells were pretreated with SB216763 (0–40 μM) for 2 h and then incubated with CTX for 24 h. CTX and SB216763 were dissolved in DMSO which was 0.01% of the culture medium.

Mitochondrial membrane potential (MMP) measurement

The MMP of TCs was estimated using the Enhanced Mitochondrial Membrane Potential Assay Kit (C2006, JC-1, Beyotime, Shanghai, China) according to the instructions. The TCs in each group were incubated for 20 min at 37 °C in the dark using JC-1 staining buffer. Then, the TCs were washed and digested for MMP detection with the use of flow cytometry (FACSCanto™ II, BD, USA) or viewed using inverted fluorescent microscopy (Leica, Germany). Each data point represents a minimum of three replicates.

Western blot analysis

Ovarian tissue and cultured TCs were lysed using Radio Immunoprecipitation Assa (RIPA) buffer. Protein concentrations were determined using the bicinchoninic acid (BCA) assay (Solarbio, Beijing, China). Proteins were separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Then determine the range of the target strip according to the protein ladder (26616, Thermo Fischer Scientific; former Savant, MA, USA) and cut the gel. Subsequently, transferred proteins onto polyvinylidene fluoride (PVDF) membranes (22860, Thermo Fischer Scientific; former Savant, MA, USA). PVDF membranes with protein imprints and exposed images will no longer undergo any cropping, the original images are attached in Supplementary Documents 1. The expression of the target protein was calculated by the ratio between the target protein and GAPDH from the same gel and transferred to the single PVDF membrane. After blocking with 5% skim milk, membranes were incubated overnight at 4 °C with the following polyclonal antibodies: anti-GSK3β (1:3000), anti-StAR (1:1000), anti-17βHSD (1:500), anti-PHB (1:2000), anti-Mfn1 (1:1000,), anti-Mfn2 (1:2000), anti-Drp1 (1:2000) or anti-GAPDH (1:20,000). The membranes were then washed three times with TBS plus Tween 20 (TBST) and incubated with secondary antibodies (1:40,000, ZB-2301, ZSGB-BIO, China) at 37 °C for 1 h. Expressions of each protein were determined using the enhanced chemiluminescence reagent (ECL) kit (ED0015-C, Sparkjade Science Co., Ltd., Shandong, China). The density of protein bands was analyzed using Image J software.

Statistical analysis

Data were expressed as means ± SD and were analyzed with use of SPSS 26.0 software. Differences between two groups were determined using independent samples t-tests, while that with more than two groups using a one-way analysis of variance (ANOVA) with the LSD test for post-hoc comparisons. Experimental data plots were plotted using the GraphPad Prism 9.0 software program. The P < 0.05 was required for results to be considered as statistically different.

Ethics approval and consent to participate

In the experiment stage, animals were handled in compliance with the Basel Declaration. The Ethics Committee of Binzhou Medical University has approved the use of animals in research.

ARRIVE guidelines

All methods were carried out in accordance with ARRIVE guidelines.

Results

Characterizations of hUMSCs

Under the light microscope, the morphology of the isolated cells resembled that of fibroblasts (Fig. 1A). The presence of alizarin red S (Fig. 1B) and oil red O (Fig. 1C) staining demonstrated that the hUMSCs could develop into adipocytes and osteoblasts. The results of FCM showed that more than 96% of isolated expressed CD44, CD73 and CD90, but not CD45, CD34 and HLA-DR, which further verified the phenotype of hUMSCs (Fig. 1D).

Figure 1
figure 1

Characterizations of hUMSCs. (A) The representative pictures of hUMSCs cell morphology was observed under the light microscope, showing a typical fibroblastic shape (Scale bar = 200 µm). (B) The osteogenic differentiation ability of the induced differentiated hUMSCs was verified by alizarin red staining (Scale bar = 200 µm). (C) Oil red O staining was used to verify the adipogenic differentiation ability of the induced differentiated hUMSCs (Scale bar =  100 µm). (D) FCM detecting hUMSCs surface markers CD73, CD90, CD44, HLA-DR, CD34 and CD45. The data are presented as the means ± SD of at least three independent experiments.

Full size image

hUMSCs treatment repairs ovarian morphology and function in POI mice

As shown in Fig. 2B,C, compared with the control group, the body weight and ovarian index of were significantly reduced in the POI group mice (P < 0.01), while after hUMSCs treatment, the body weight and ovarian index were significantly increased compared to the PBS group (P < 0.01). H&E staining was used to assess the effects of hUMSCs on ovary morphology in POI mice. As shown in Fig. 2D,E, compared with that of the control and hUMSCs groups, ovarian tissue in the POI and PBS groups was markedly atrophied. In addition, the number of primitive follicles, primary follicles, and secondary follicles in the ovarian tissue were significantly decreased, and more atretic follicles were observed in the ovarian tissues of these POI mice (P < 0.05). After hUMSCs treatment, the number of normal follicles at each stage of development was greatly increased and the number of follicles showing atresia was greatly reduced (P < 0.05). Results from serum hormone assays revealed that lower levels of E2 and higher levels of FSH and LH were obtained in POI versus control mice (P < 0.001). In contrast, when assayed at 7 days after treatment with hUMSCs, serum E2 levels were significantly increased, while LH and FSH secretions were decreased in POI mice (P < 0.01, Fig. 2F–H). These data demonstrate that hUMSC treatment restored ovarian morphology and function within this CTX-treated in POI mice.

Figure 2
figure 2

hUMSCs treatment resortes ovarian morphology and function in POI mice. (A) The scheme of animal experimental design. (B) Body weight of mice (n = 12). (C) Ovarian index (n = 6). (D) H&E staining representative photos of the sectioned ovarian tissues (Scale bar = 220 µm). (E) Comparison of ovarian follicle counts at various phases of development (n = 6). (F–H) Levels of E2, LH and FSH hormones in serum (n = 3). The data are presented as the means ± SD. *P < 0.05 and ***P < 0.001, compared with the Control group, #P < 0.05, ##P < 0.01and ###P < 0.001, compared with the PBS group.

Full size image

hUMSCs treatment promotes ovarian hormone secretion in POI mice

Next, we assessed the expression levels of proteins involved in testosterone synthesis, StAR, and 17βHSD, in ovarian tissue samples of POI mice. From immunofluorescent staining, we found that StAR, PHB and 17βHSD fluorescence in the POI group was much less intense than that observed in the Control group. However, these levels exhibited an opposite trend in POI mice receiving hUCMSCs treatment (P < 0.05, Fig. 3A–D). Furthermore, increased expressions of StAR, 17βHSD, and PHB were found in the hUCMSCs group as determined with western blots, results which were consistent with that from immunofluorescent staining (P < 0.05, Fig. 3E–H). Therefore, these data indicate that the decreased ability to produce ovarian hormones in this CTX-induced model of POI was restored following hUCMSCs treatment.

Figure 3
figure 3

hUMSCs treatment promotes ovarian hormone secretion in POI mice. (A) Representative confocal images of Immunofluorescence staining. (Scale bar = 60 µm). (B–D) MFI of 17βHSD , StAR and PHB in ovarian tissues as analyzed by ImageJ software (n = 3). (E) Representative bands of protein expression detected by Western blot. (F–H) Expressions of 17βHSD, StAR and PHB protein as quantitated by ImageJ software (assigned a value of 1, n = 3). The data are presented as the means ± SD. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with the Control group, #P < 0.05 and ##P < 0.01, compared with the PBS group.

Full size image

hUMSCs treatment restored mitochondrial dynamics imbalance in the ovaries of POI mice

Subsequently, we investigated the effects of hUMSCs on ovarian mitochondrial dynamics in POI mice. The GSK3β positive cells were significantly increased in the POI versus Control group as based on our immunofluorescent staining results. In contrast, hUMSCs treatment markedly decreased the numbers of GSK3β positive cells as compared to that obtained in the PBS group (P < 0.001, Fig. 4A,B). We observed that expressions of the proteins associated with mitochondrial dynamics Mfn1 and Mfn2 were decreased, while that of Drp1 was increased following CTX treatment, effects which were restored by hUCMSCs as demonstrated with both western blot (P < 0.05, Fig. 4C–G) and qRT-PCR (P < 0.001, Fig. 4H–K) assays. These results suggest that an in vivo intervention with hUCMSCs significantly inhibits GSK3β expression in the ovarian tissue of POI mice, and that the impaired ovarian testosterone synthesis in POI mice may be related to imbalances in mitochondrial dynamics resulting from GSK3β.

Figure 4
figure 4

hUMSCs treatment restored mitochondrial dynamics imbalance in the ovaries of POI mice. (A) Representative confocal images of Immunofluorescence staining of in ovarian tissues (Scale bar = 60 µm). (B) MFI of GSK3β in ovarian tissues as analyzed by ImageJ software (n = 3). (C) Representative bands of protein expression detected by Western blot. (D–G) Expressions of Mfn1, Mfn2, Drp1 and GSK3β protein as quantitated using ImageJ software (assigned a value of 1, n = 3). (H–K) The mRNA expressions of Mfn1, Mfn2, Drp1 and GSK3β (assigned a value of 1, n = 3). The data are presented as the means ± SD. **P < 0.01 and ***P < 0.001, compared with the Control group, #P < 0.05 and ###P < 0.001, compared with the PBS group.

Full size image

hUMSCs supernatant increases the MMP in TCs after CTX treatment

To reveal the potential mechanism by which hUMSCs supernatant restores TCs function in POI mice, we selected a CTX concentration of 30 µM and a GSK3β inhibitor SB216763 concentration of 10 µM for subsequent experiments (P < 0.01, Fig. 5A,B). A schematic of the in vitro experimental design is shown in Fig. 5C. The mitochondria play an important role in the synthesis of testosterone, and its activity depends on normal MMP. The decrease in mitochondrial membrane potential, one of the hallmarks of impaired cell function, as shown in Fig. 5D,E, the ratio of red and green fluorescence intensity significantly decreased in the CTX group compared with the Control group. Conversely, compared with the CTX group, the rate of the ratio between red and green fluorescence intensity was significantly raised in the hUMSCs group and SB216763 group (P < 0.01). In the same way, according to flow cytometry, green fluorescence represents the population of cells with decreased MMP, and the results are expressed as percentages. In the CTX group, the ratio of the number of TCs with MMP decrease to the total number of cells was the largest, and MMP of some cells increased significantly after hUMSCs supernatant or SB216763 treatment (P < 0.05, Fig. 5F,G). These results suggested that hUMSCs supernatant or SB216763 treatment weakens the adverse effects of CTX on MMP.

Figure 5
figure 5

hUMSCs treatment recovered mitochondrial dynamics imbalance in ovarian tissue of POI mice. (A,B) Different concentrations of CTX and SB216763 effect on the cell viability of TCs (assigned a value of 1, n = 3). (C) The scheme of the experimental design. (D,E) The MMP of TCs by immunofluorescence analysis (Scale bar = 60 µm, n = 3). (F–H) The MMP of TCs by FCM (n = 3). The data are presented as the means ± SD. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with the CTX group.

Full size image

hUMSCs supernatant regulates mitochondrial fusion and fission imbalance by targeting GSK3β in TCs

To further establish whether GSK3β plays a vital role in the mitochondrial dynamics of TCs, the expressions of GSK3β and proteins associated with mitochondrial dynamics were examined in these in vitro preparations of TCs. As shown in Fig. 6A–E, CTX treatment markedly increased the protein expressions of GSK3β and Drp1 as compared with that in the Control group, while treatments with hUMSCs supernatant or the GSK3β inhibitor SB216763 significantly reduced these levels as compared with that obtained in the CTX group (P < 0.05). As compared with the Control group, expressions of the mitochondrial fusion proteins Mfn1 and Mfn2, were substantially reduced in the CTX group, effects which were notably reversed in groups receiving hUMSCs supernatant or SB216763. Similarly, immunofluorescent assay results provided further support for the above results (P < 0.01, Fig. 6F–J). We also found that mRNA levels of GSK3β and Drp1 were markedly increased, while mRNA levels of Mfn1 and Mfn2 were significantly decreased in the CTX group. Equally, effects were reversed in response to hUMSCs supernatant treatments. With the use of SB216763, the reductions in mitochondrial fusion-related proteins and the increases in fission proteins after CTX treatment may result from the enhanced activity of GSK3β (P < 0.001, Fig. 6K–N). These data provide evidence that hUMSCs can elevate mitochondrial dynamics in CTX-treated TCs.

Figure 6
figure 6

hUMSCs supernatant regulates mitochondrial fusion and fission imbalance by targeting GSK3β in TCs. (A) Representative bands of protein expression detected by Western blot. (B–E) Expressions of Mfn1, Mfn2, Drp1 and GSK3β protein as quantitated by ImageJ software (assigned a value of 1, n = 3). (F) Representative confocal images of Immunofluorescence staining (Scale bar = 60 µm). (G–J) MFI of Mfn1, Mfn2, Drp1 and GSK3β in TCs as analyzed by ImageJ software (n = 3). (K–N) The mRNA expressions of Mfn1, Mfn2, Drp1 and GSK3β (assigned a value of 1, n = 3). The data are presented as the means ± SD. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with the CTX group.

Full size image

hUMSCs supernatant increases the production of testosterone synthesized in TCs after CTX treatment

We evaluated the protective effects of hUMSCs or SB216763 upon the production of testosterone synthesized in CTX-treated TCs. Immunofluorescent staining was performed to determine whether any changes in the StAR enzymes involved in testosterone synthesis were present in TCs. As shown in Fig. 7A–B, the immunofluorescent levels of StAR were significantly decreased in the CTX versus the Control group. After treatment with either hUMSCs supernatant or SB216763, immunofluorescent levels of StAR were markedly increased (P < 0.001). Moreover, TCs treated with CTX showed decreasing expressions of 17βHSD and PHB, while an opposite trend appeared in the hUMSCs supernatant or SB216763 groups (P < 0.05, Fig. 7D–G). Furthermore, the testosterone content in the TCs supernatant of each group was measured. Our results showed that the testosterone content in the CTX group was significantly lower than that in the Control group, and after the treatment with hUMSCs or SB216763 treatment, the testosterone content in the supernatant increased significantly (P < 0.01, Fig. 7C). These aforementioned data suggest that the protective effects of hUMSCs supernatants on TCs relied on the inactivation of GSK3β to increase the production of testosterone within TCs exposed to CTX.

Figure 7
figure 7

hUMSCs supernatant increases the production of testosterone synthesized in TCs after CTX treatment. (A) Representative confocal images of Immunofluorescence staining (Scale bar = 60 µm). (B) MFI of StAR in TCs as analyzed by ImageJ software (n = 3). (C) Levels of testosterone in culture medium (n = 3). (D) Representative bands of protein expression detected by Western blot. (E–G) Expressions of 17βHSD, StAR and PHB protein as quantitated by ImageJ software (assigned a value of 1, n = 3). The data are presented as the means ± SD. *P < 0.05, **P < 0.01 and ***P < 0.001, compared with the CTX group.

Full size image

Discussion

Hormone replacement therapy (HRT) is now the most widely used approach of treating POI. It cannot, however, successfully restore ovarian fertility and function32. POI has gained new hope as a result of the widespread use of mesenchymal stem cells (MSCs), and a number of stem cell transplantation therapies that aid mammals in rebuilding function in their impaired ovaries have been identified33,34. The earlier studies revealed that MSCs might reduce chemotherapy-induced POI, proving that MSCs could restore ovarian endocrine function while repairing ovarian structure35. According to our prior studies, the POI model would cause variable degrees of harm to the ovarian structure and endocrine system, particularly the loss of functioning follicles at every stage of development36,37. Compared to the control group, the POI group mice showed significant weight loss and a significant decrease in ovarian index, while hUMSCs treatment significantly reversed the above indicators. Similarly, this study shows that these circumstances improved following hUMSCs treatment, as demonstrated by a considerable increase in blood E2 levels, a major decrease in gonadotropin levels, and a substantial rise in the number of healthy follicles.

Additionally, this study discovered that POI mice had aberrant TCs functioning, which showed up as a reduction in testosterone synthesis. TCs are a crucial site for the intracellular synthesis of testosterone in females, as are the adrenal glands and ovaries, which can both produce a certain quantity of testosterone38. The synthesis of testosterone requires the participation of some key enzymes, StAR is known to transport free fatty acids (FFA), the raw material for testosterone synthesis, to mitochondria39. One of the crucial enzymes for testosterone synthesis is 17HSD, and its reduction invariably results in a drop in testosterone secretion40. According to our results, it may be because CTX interferes with StAR and 17βHSD activities, so the synthesis of testosterone by TCs decreases. After the treatment of hUMSCs, both the amount of testosterone in the culture supernatant of TCs and the expression of the enzymes involved in testosterone production were elevated. PHB is a multifunctional protein associated with many cellular processes such as cell cycle, proliferation, apoptosis, senescence, cellular immortalization, adipogenesis and differentiation. In the ovary, PHB is widely expressed and its expressions are age and follicular stage regulated41. In addition, PHB has a multifaceted relationship with sex steroid hormones, estrogens and androgens, and their receptors42. On the one hand, PHB functions as a co-repressor of sex steroid receptors, while on the other, it has been identified as a target gene under sex hormone regulation43,44. In this study, we observed a significant decrease in PHB expression in TCs and ovarian tissue caused by CTX-induced damage, accompanied by abnormal hormone secretion levels of E2, FSH, and LH. However, hUCMSCs and their supernatant intervention significantly upregulate the PHB expression and repair abnormal hormone levels in TCs and ovarian tissue. Reportedly, alterations in the phosphorylation state of PHB can contribute to the survival and differentiation processes of granulosa cells (GCs) that are controlled by FSH45. Furthermore, there have been reports of a considerable rise in oxidative and phosphorylation levels in GCs during POI46. We speculate that MSCs may participate in the regulation of related endocrine functions by regulating PHB phosphorylation status in TCs, which requires further investigation.

The delicate balance between mitochondrial fission and fusion, and large rearrangements in the mitochondrial network can be seen in response to cellular insults and disease47. Dysfunction in the major components of the fission and fusion machineries including dynamin-related protein 1 (Drp1), mitofusins 1 and 2 (Mfn1 and Mfn2) and ensuing imbalance of mitochondrial dynamics can lead to many diseases48. Recently, it has been shown that the function of the ovary is closely related to mitochondrial dynamics. Feng et al. suggested that Leucyl-tRNA synthetase 2 (LARS2) regulated the mitochondrial dynamics and function in granulosa cells49. Researchers has shown that mitochondrial fusion and division are interdependent, when there is moderate to severe injury, cells tend to use fusion to replace damaged mitochondria with healthy ones in an attempt to repair the damage50,51. In situations of intense stress, normal or elongated mitochondria will divide52. Based on this information, we speculated that mitochondria divided into smaller compartments to protect themselves from the negative effects of CTX. The findings supported our ideas by revealing a considerable increase in the expression of the mitochondrial division protein Drp1 in the CTX group. Concurrently, we discovered that following CTX treatment, the mitochondrial fusion proteins Mfn1 and Mfn2 were also diminished to differing extents. From the standpoint of mitochondrial dynamics, our study confirmed that CTX leads to a shortage of mitochondrial division in TCs, which in turn causes abnormalities related to testosterone synthesis. To better perform the function of mitochondrial testosterone synthesis and energy metabolism, on the other hand, hUMSC treatment inhibits excessive fission of the mitochondria and enhances mitochondrial fusion.

In addition, numerous signaling mechanisms have been revealed from research to control mitochondrial dynamics. Evidence shows that GSK3β mediates the phosphorylation of Drp1. Increased GSK3β activity induces mitochondrial and synaptic dysfunction by regulating Drp1 in the diabetic model53. Blockage of GSK3β-mediated Drp1 phosphorylation will provide neuroprotection in neurons and Alzheimer's disease (AD) mouse models54. Liu et al. reported that sevoflurane induced neurotoxicity links to a GSK3β/Drp1 dependent mitochondrial fission and apoptosis55. As mentioned earlier, MSCs participate in the regulation of target cell oxidation and phosphorylation56,57. It is reported that Zinc (Zn) down-regulated GSK3β expression and protects against mitochondrial damage, which in turn prevents apoptosis58. Chen et al. found that formation in Diabetic cardiomyopathy (DCM) may occur because Mfn2 inactivation leads to impaired myocardial mitochondrial function. However, the symptoms were alleviated with GSK3β inhibitors59. Dong et al. found that in the ovary, through the PI3K/AKT/GSK3β signaling pathway, the apoptosis of GCs was reduced to restore the ovarian function of POI60. Thus, we infer that CTX elevates GSK3β activity, resulting in an imbalance in the dynamics of the mitochondria. To counteract the effects of GSK3β, TCs were pretreated with inhibitor SB216763. As predicted, CTX raised Drp1 expression while decreasing Mfn1 and Mfn2 expression. Drp1 expression dropped while Mfn1 and Mfn2 expression rose when GSK3β activity was inhibited. Moreover, pretreatment with a GSK3β inhibitor decreased the destruction of the mitochondrial membrane potential caused by CTX and the decrease in testosterone in TCs. But we also found that hUMSCs culture supernatant could similarly repair injured TCs and had a similar therapeutic effect to GSK3β inhibitors. The above data suggest that hUMSCs transplantation restored the ability of TCs to synthesize testosterone by decreasing GSK3β activity and increasing mitochondrial dynamics. As mentioned earlier, MSCs participate in the regulation of target cell oxidation and phosphorylation56,57. Therefore, we speculate that the regulation of GSK3β on the mitochondrial dynamics of TCs in the POI process, as well as the intervention mechanism of hUCMSCs, may be related to the phosphorylation regulation of GSK3β, and this event should be further studied.

Conclusion

In conclusion, our results indicate a novel role for hUMSCs in alleviating CTX-induced abnormal structure and function of ovarian. The mechanism could be related to the downregulating GSK3β expression and restoring mitochondrial imbalance in TCs on POI mice ovaries. The results of this investigation offer a fresh approach to the development of a therapeutic method for patients with POI-related metabolic abnormalities (Fig. 8).

Figure 8
figure 8

Schematic illustration of hUMSCs transplantation restoring TCs function in CTX-induced POI. hUMSCs transplantation suppressed GSK3β expression in TCs, restore the imbalance in mitochondrial dynamics imposed on by CTX, and restart testosterone synthesis in TCs.

Full size image