Abstract Background Premature ovarian insufficiency (POI) is a challenging condition with limited effective treatments. Adipose-derived stem cells (ADSCs) have demonstrated potential in tissue repair, and their extracellular vesicles (ADSC-EVs) show more safety in clinical translation. However, the role and mechanism of ADSC-EVs in the treatment of POI are not yet fully understood. This study aims to investigate the protective effects of ADSC-EVs on rat POI models induced by 4-vinylcyclohexene diepoxide (VCD) and to explore the potential therapeutic mechanisms. Methods Rat ADSCs and ADSC-EVs were isolated and characterized. The POI rat model was established via intraperitoneal injection of VCD for 15 consecutive days. ADSCs and ADSC-EVs were injected into the ovaries for treatment. Ovary function was assessed by monitoring estrous cycles, follicle counts, sexual hormone levels, and ovulation. Molecular mechanisms were investigated using TUNEL staining, immunohistochemistry, quantitative polymerase chain reaction, and western blotting. In vitro, primary rat granulosa cells were treated with VCD in the presence or absence of ADSC-EVs. Cell proliferative ability, hormone secretion, apoptosis rate, and relative molecular expression were measured. Whole-transcriptome sequencing and DIA proteomics of ADSC-EVs were performed to identify bioactive molecules. Results ADSC-EVs protected granulosa cells from VCD-induced apoptosis, promoted cell proliferation, enhanced hormone secretion, and upregulated KITL expression. Treatment with ADSCs and ADSC-EVs in POI rats significantly improved estrous cycles, follicle counts, and serum levels of estradiol (E2), follicle-stimulating hormone (FSH), and anti-Müllerian hormone (AMH). These treatments activated the KITL/KIT/PI3K/AKT signaling pathway, downregulated pro-apoptotic genes (Bax, Caspase3), and upregulated anti-apoptosis genes (Bcl2). ADSC-EVs are highly enriched in mRNA of Kit and PI3K, and both transcriptomic and proteomics enrichment analysis predominantly focused on PI3K/AKT pathway. Conclusions ADSCs and ADSC-EVs effectively protect and restore ovarian function in VCD-induced POI rats. The mechanism involves inhibiting apoptosis in granulosa cells and activating the KITL/KIT/PI3K/AKT pathway in ovary. ADSC-EVs, with advantages in clinical translation, hold significant potential for POI treatment. Graphical abstract [50]graphic file with name 13287_2025_4553_Figa_HTML.jpg Supplementary Information The online version contains supplementary material available at 10.1186/s13287-025-04553-6. Keywords: Premature ovarian insufficiency, 4-vinylcyclohexene diepoxide, Adipose-derived stem cells, Extracellular vesicles Introduction Premature ovarian insufficiency (POI) is a complex condition characterized by the premature depletion of follicles and the cessation of ovarian function before the age of 40 [[51]1]. Its prevalence is rising due to a range of social, psychological, and environmental factors. By 2024, POI affects 2−4% of women worldwide [[52]2]. Ovarian function decline progresses through three stages: diminished ovarian reserve (DOR), POI, and premature ovarian failure (POF) [[53]3]. Symptoms include infertility, hormonal imbalances, and other complications, all of which severely compromise quality of life affected individuals [[54]4]. The challenge of restoring fertility in POI patients is formidable [[55]5, [56]6], and they often exhibit poor ovarian response necessitating oocyte donation for assisted reproductive technologies [[57]7, [58]8]. Therefore, developing safe and effective treatments to replenish ovarian reserve and activate dormant follicles is crucial for POI management. In the rapidly evolving field of regenerative medicine, stem cell therapy is emerging as a promising approach for the treatment of POI. Mesenchymal stem cells (MSCs) are a type of multipotent non-hematopoietic stem cells noted for their low immunogenicity [[59]9]. MSCs are extensively studied for their potential to restore homeostasis in inflamed or injured organs and tissues, offering promising cell-based therapeutic strategies for various diseases [[60]10]. MSCs can be isolated from various tissues, including umbilical cord, placenta, bone marrow, and adipose tissue [[61]11, [62]12]. MSC transplantation is extensively utilized in POI research, exploring mechanisms such as homing, immunomodulation, and anti-apoptosis [[63]13, [64]14].Notably, extracellular vesicles (EVs) within MSCs’ secretome, carrying cytokines, growth factors, and RNAs, can be transferred to target cells in damaged tissues, thereby exerting long-term therapeutic effects [[65]15, [66]16]. Adipose-derived stem cells (ADSCs) are MSCs isolated from adipose tissue [[67]17–[68]19], considered promising therapeutic agents due to their abundant source and minimal extraction damage [[69]17, [70]19]. ADSCs have also demonstrated efficacy in POI treatment by promoting angiogenesis and reducing granulosa cell apoptosis and senescence [[71]20, [72]21]. However, concerns regarding potential risks of immunological rejection and malignant transformation limit their clinical application [[73]17]. ADSC-EVs, extracellular vesicles secreted by ADSCs, contain bioactive substances similar to ADSCs’ paracrine factors [[74]17, [75]22]. As for clinical application, ADSC-EVs have several advantages over ADSCs, including eliminating the risk of malignant transformation, being cell-free, and possessing low immunogenicity [[76]17, [77]23–[78]26]. Current studies have demonstrated that ADSC-EVs enhanced cyclophosphamide-induced ovarian function via the modulation of the SMAD and AMPK/mTOR pathway [[79]27, [80]28]. More extensive and thorough investigations are necessary for validation and clarification of ADSC-EVs therapy on POI. Various POI animal models have been utilized in stem cell research with MSCs derived from different sources, and the therapeutic mechanisms usually display variations. 4-vinylcyclohexene diepoxide (VCD) is a metabolic product of industrial chemical toxicants. Studies have confirmed that VCD selectively affects primordial and primary follicles, leading to follicle atresia via disrupting KIT/phosphatidylinositol 3-kinase (PI3K)/AKT pathway in oocyte and ultimately resulting in POI [[81]29–[82]31]. The KIT/PI3K/AKT pathway is implicated in controlling the dormancy and activation of primordial follicles [[83]32]. Reduced follicle recruitment elevates gradually serum follicle-stimulating hormone (FSH) and reduces anti-Müllerian hormone (AMH) and estradiol (E2) levels [[84]31, [85]33, [86]34]. This is a gradual process similar to the natural decline of human ovarian function [[87]35, [88]36]. Several therapeutic strategies have been introduced in VCD induced POI treatment such as platelet-rich plasma injection [[89]37]. For MSC strategies, placental MSCs had been reported ameliorating VCD induced POI by modulating macrophage M2 polarization [[90]38]. However, the impact of ADSCs and ADSC-EVs on VCD-induced POI models and the underlying mechanisms remain unclear. EVs contain various substances, including proteins and various types of RNAs. RNA delivery by EVs in mammalian cells and its applications was important and having promising application prospects [[91]39]. Due to the high content of microRNA (miRNA), many previous studies have detected miRNA in EVs [[92]40, [93]41]. miRNA sequence revealed that ADSC-EVs contained some proangiogenic miRNAs (miR-126, miR-130a and miR-132) and an antifibrotic miRNA family (miR-let7b and miR-let7c) to ameliorate erectile function in diabetic rats [[94]42]. Additionally, miR-126 and miR-130a can up-regulate expression of PIK3R/AKT pathway [[95]43, [96]44]. Therefore, miRNAs targeting the PI3K/AKT pathway may be one of the reasons for regulating ovarian function. Besides, it is worth exploring whether mRNA delivery of EVs, which could directly translate into functional proteins, has an impact on ovarian function [[97]45]. The mRNA in mouse mast cell-derived EVs can be delivered to human mast cells and translated [[98]46]. The RNA transfer of ADSC-EVs to neurons was observed by fluorescence labeling [[99]47]. In this study, we completed the full-length transcriptome and proteomics of ADSC-EVs to investigate bioactive factor involved in POI treatment. Therefore, this study aims to investigate the therapeutic effects and mechanisms of ADSCs and ADSC-EVs on VCD-induced POI. Our findings provide a robust foundation for the clinical translation of ADSC-EVs in regenerative therapies for POI. Materials and methods Ethical approval All animal experiments were performed in accordance with standard protocols, animal welfare regulations, and institutional guidelines of International Peace Maternity and Child Health Hospital, School of Medicine, Shanghai Jiao Tong University. We confirm that the study was carried out in compliance with the ARRIVE guidelines. All process for rat experiments were approved by the Ethics Committee on animal subjects of International Peace Maternity and Child Health Hospital (Project: Establishment of prediction model for premature ovarian insufficiency (POI) and application of precision diagnosis and treatment; Approval number: GKDW-A-2024-46; Date of approved: 2024-03-25). Preparation and culture of ADSCs Female Sprague-Dawley (SD) rats (aged 12 weeks, weighted 160–180 g) were acclimatized for one week and euthanized with intraperitoneal injection of 1.5% pentobarbital sodium solution. The inguinal fat pads were excised, washed, and digested with 0.125% type I collagenase solution at 37 °C, 200 rpm for 50 min. Then the digestion was filtered through a sterile 70 μm cell strainer and cultured in T25 flasks. The cells were passaged when cells reached 80–90% confluence. Identification of ADSCs' surface markers ADSCs (P3) were taken for surface marker antibody identification using the following procedure: The cells were trypsinized, harvested, and incubated with fluorescently labeled antibodies. The surface marker of rat ADSCs identification including: Anti-CD44 (Biogems, 06511), Anti-Rat CD73 antibody (BD Biosciences, 551123), Anti-Rat CD90 (Biogems, 876573), Anti-CD11b (Biogems, 06511), Anti-Rat CD34(Biogems, 06511), Anti-Rat CD45 (AbD Serotec: 010), labeled with PE, PE-CY7 or FITC. After incubating in the dark for 40 min, the cells were washed and then detected by flow cytometer (BD Biosciences, FACSCelesta). Assessment of multipotency To determine the multipotency of the isolated ADSCs, they were cultured with induction medium for adipogenesis, chondrogenesis and osteogenesis. ADSCs adipogenic differentiation medium consists of induction medium (A liquid): containing fetal bovine serum, isobutylmethylxanthine, glutamine, insulin, dexamethasone, rosiglitazone, and 1% penicillin-streptomycin. Maintenance medium (B liquid): containing serum, glutamine, double-antibiotic, and insulin. Cells were cultured with A liquid for 3 days and B liquid for 1 day in one cycle, for a total of 12–20 days. Finally, oil red O staining was used to observe the results. ADSCs osteogenic differentiation medium contained β-glycerophosphate, ascorbic acid, glutamine, double-antibiotic, and dexamethasone. Cells were cultured with this medium for 4 weeks and the cells were stained with alizarin red solution. ADSCs chondrogenic differentiation medium contains ascorbic acid, dexamethasone, proline, sodium pyruvate, ITS (containing insulin, human transferrin, and selenite), and TGF-β3. Cells were cultured as cell aggregates in chondrogenic differentiation medium for four weeks. At the end of the incubation, the cell aggregates were fixed and embedded in Tissue-Tek O.C.T. compound (Sakura, Japan). The sections were stained with Alcian blue kit (following manufacturer’s instructions). ADSCs-EVs extraction and identification EVs Isolation Kit purchased from Umibio Science and Technology Group (UR52121, China, Shanghai) was used to extract extracellular vesicles from ADSCs culture medium (P8 and earlier generations). When the cells reach 80−90% confluence, the complete medium was removed, and serum-free medium was added for continued culture. After 72 h, the culture supernatant was collected and centrifuged at 4 °C, 3000 g, for 10 min to remove cell debris. Added 1/4 volume of EVs Concentration Solution to the ADSCs supernatant and sealed and vortexed for one minute. Then, incubated the solution at 4 °C for 2 h and centrifuged again at 4 °C, 10,000 g, for 60 min. Resuspend the pellet in PBS buffer and centrifuge at 4 °C, 12,000 g, for 2 min. Loaded the EVs-containing supernatant onto an EVs Purification Filter column and centrifuged at 4 °C, 3000 g, for 10 min. Collected the purified ADSC-EVs from the bottom of the column and stored at -80 °C for future analysis. Nanoparticle tracking analysis (NTA) Calibrate the NTA instrument (Malvern’s NanoSight NS300) using a standard particle suspension with a known particle size and concentration. Adjust the instrument settings, including the laser power and detection threshold. Diluting the ADSC-EVs sample to an appropriate concentration. Vortexed gently to disperse the particles evenly throughout the solution. Load the diluted ADSC-EVs sample into a clean, disposable cuvette. Place the cuvette into the NTA instrument and initiate the analysis, allowing the instrument to record the Brownian motion of the particles within the sample. The NTA instrument captured video footage of the particles as they move and analyzed to determine particle size and concentration. The software tracked individual particles and calculate their diffusion coefficients. Transmission electron microscopy (TEM) Resuspend a suitable amount of ADSC-EVs in 100 µL of 2% paraformaldehyde solution on ice. Add 5 µL of the ADSC-EVs suspension to a Formvar-carbon coated grid. After gentle washing with PBS, the grid was placed on drop of 1% glutaraldehyde for 5 min. Then, the grid was washed with double-distilled water for 2 min, repeating eight times to thoroughly wash away the fixative. Next, the grid was placed on a drop of osmium tetroxide for 5 min, followed by a 10-minute incubation on methylcellulose. Air-dry the grid for 10 min and place the grid into the observation chamber of the electron microscope and capture TEM images. POI rat model and treatment Female SD rats (4 weeks old) were randomly divided into four groups (n = 8 per group): Control group, VCD group, ADSCs-treated group (VCD + ADSCs), and ADSC-EVs-treated group (VCD + EVs). Rats were administered intraperitoneal injections at a dose of 2 mL/kg body weight. The control group received sesame oil injections, while the other groups were injected with VCD solution (160 mg/kg) for 15 consecutive days to establish a POI model. The POI model was assessed by estrous cycle test, and without normal estrous cycle for at least ten days indicated a successful model for a rat. With successful phenotype of POI, the rats were administered in situ ovarian injections (20 µL per side) as follows: (1) Control group: bilateral in situ ovarian injection of PBS using control rats; (2) VCD group: bilateral in situ ovarian injection of PBS using POI rats; (3) VCD + ADSCs group: bilateral in situ ovarian injection of ADSCs suspension (total cell counts was 1 × 10^5 per ovary) using POI rats; (4) VCD + EVs group: bilateral in situ ovarian injection of ADSC-EVs suspension ( total quantity of EVs was 10 µg per ovary, equivalent to EVs produced by 1 × 10^5 cells) using POI rats. Procedure was as followed: Anesthetized with 1.5% pentobarbital sodium solution (30 mg/kg), the ovaries were exposed in the abdomen incision. A micro syringe was used to inject relative solution directly into the ovarian body. Once no bleeding was observed in the ovaries, the abdomen was closed with absorbable sutures, and 100,000 units of penicillin were administered postoperatively to prevent infection. Vaginal smears of rat Vaginal smears were collected daily between 8:00 and 9:00 AM. The procedure is as follows: A sterilized medical cotton swab was moistened with physiological saline and inserted into the rat’s vagina for approximately 1 cm. The swab was then rotated clockwise 3–4 times and gently withdrawn. The material collected on the swab was transferred to a glass slide by rotating the swab to spread the cells without overlapping. The slide was then fixed with 4% paraformaldehyde and stained with hematoxylin and eosin. The cell morphology was observed under a light microscope. The normal estrous cycle of rats typically lasts 4–5 days and is divided into four stages: proestrus; estrus; metestrus; and diestrus. Enzyme-linked immunosorbent assay (ELISA) of hormones AMH (SENBEIJIA, SBJ-R0231), FSH (SENBEIJIA, SBJ-R0712), E2 (SENBEIJIA, SBJ-R0136), and progesterone (SENBEIJIA, SBJ-R0697) levels in serum samples were determined by the ELISA kit. Blood samples were collected from the orbital sinus of SD rats. Medium of granulosa cells were collected directly. Experiments were performed according to the manufacturer’s instructions. In brief, the ELISA plate was prepared with blank, standard sample, sample wells, and incubated for 90 min. Then, the wells were incubated with biotinylated antibody for 1 h at 37 °C. After washing five times, the wells were incubated with HRP enzyme conjugate working solution for 30 min at 37 °C. Subsequently, incubated wells with TMB substrate solution for 15 min at 37 °C in the dark, followed by the stop solution. The absorbance was immediately measured using an ELISA reader. Standard curves for E2, FSH, and AMH were plotted and the concentrations of the samples were then calculated based on their measured OD values. Assessment of ovarian follicle counts Ovaries were collected and fixed by 4% formalin, and then embedded with paraffin. The ovarian tissue was serially sectioned at intervals of 20 μm (micrometers), mounted in order on glass microscope slides, stained with HE, and counted the follicles in each section. Four stages of follicles (primordial, primary, secondary, and antral) were identified and categorized. By examining three consecutive sections, the total number of follicles at each stage can be determined and summed to provide a comprehensive count of the follicles present in the ovary. The proportion of follicles in each group was calculated and contrasted. The experiment was replicated three times. Superovulation Rats from control, VCD, VCD + ADSCs, and VCD + EVs groups (n = 5) were used for ovulation tests. PMSG and hCG were diluted with 0.9% sodium chloride injection to a concentration to 40 IU/mL. At 6:00 PM on the day after estrus, the rats were intraperitoneally injected with PMSG (40 IU per rat). 48 h later, hCG (40 IU per rat) was intraperitoneally injected. At 14–16 h post-hCG injection, rats were euthanized by CO₂ asphyxiation. Oviducts were excised and flushed with pre-warmed M2 medium using a 1 mL syringe fitted with a blunted 27-gauge needle. Cumulus-oocyte complexes (COCs) were collected and treated with 0.1% hyaluronidase for 5 min at 37 °C to remove cumulus cells. Mature metaphase II (MII) oocytes with intact zona pellucida were counted under a stereomicroscope. Immunohistochemistry (IHC) Immunohistochemical analyses were performed as previously described [[100]48]. Briefly, rat ovaries were fixed, dehydrated, paraffin-embedded, and sectioned at 5 μm. The sections were deparaffinized, rehydrated, and incubated by H[2]O[2]. Then, the sections were subjected to sodium citrate buffer at high pressure for 30 min for antigen retrieval. Next, the sections were blocked with 10% donkey serum in PBS for 60 min at room temperature. Primary antibodies were diluted in blocking solution and added to the sections for overnight incubation at 4 °C. The samples were washed three times with PBST (PBS buffer with 0.05% tween-20) and incubated for 30 min at room temperature with biotinylated secondary antibodies. DAB substrate kit (Beyotime, China) was used for color developing. The samples were then counterstained with hematoxylin, dehydrated, and mounted with neutral balsam. Primary antibodies used were listed in Supplementary Table [101]1. Western blotting Tissue, EVs, and cell protein expression was analyzed by immunoblotting as previously described [[102]49]. In brief, RIPA buffer was used to lysis the tissues or cell on ice. Then, the total protein extracts were separated by electrophoresis on 10% (w/v) sodium dodecyl sulfate-polyacrylamide gels. Next, the gels were transferred onto polyvinylidene fluoride membranes, and probed with primary antibodies at 4 ℃ overnight. The second antibodies: goat-anti-rabbit (Millipore) or goat-anti-mouse (Millipore) conjugated to horseradish peroxidase were incubated with the membranes at a dilution of 1:5000 for one hour at room temperature. ECL Plus (Amersham) were used for chemiluminescence detection. Primary antibodies used were listed in Supplementary Table [103]1. RNA extraction and quantitative real-time polymerase chain reaction (RT-qPCR) Total RNA of ovaries, EVs, and cells were extracted using the Trizol and trichloromethane on the ice. The RNA was then reverse transcribed to cDNA using the RevertAid^Tm First Strand cDNA Synthesis Kit (Fermentas, Canada) according to the manufacturer’s methods. Quantitative real-time polymerase chain reaction was carried out with SYBR Premix Ex Taq (Takara, Japan) kit. qPCR reaction was carried out for 40 cycles of denaturation at 95 °C for 15 s, annealing at 58 °C for 15 s, and extension at 72 °C for 30 s (ABI PRISM 7900; Applied Biosystems, USA). The cycle time (Ct) values were determined through analysis using the Sequence Detection System and analysis software (Applied Biosystems). Data analysis was conducted using the 2–ΔΔCt method. Normalization of each sample was done based on its β-actin transcript content. The experiments were replicated three times. The primer sequences can be found in Supplementary Table [104]2. Ovarian granulosa cell extraction and treatment The extraction of ovarian granulosa cells from SD rats was performed using a trypsin digestion method. SD rat was injected subcutaneously with pregnant mare serum gonadotropin at a dose of 40 IU. Forty-eight hours after the injection, the ovaries were quickly removed and transferred into pre-cooled DMEM/F12 culture medium supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 U/mL streptomycin. Subsequently, the ovaries were cut into small fragments and incubated in DMEM/F12 culture medium with 0.25% trypsin containing 0.01% EDTA for 1 h at 37 ℃, with gentle agitation every 10 min. The digest was filtered through a 200-micron sieve and centrifuged. The pellet was resuspended in DMEM/F12 culture medium and placed in an incubator (37 °C, 5% CO[2], saturated humidity) for culture. Rat ovarian granulosa cells were cultured for 48 h and then the medium was replaced with different experimental media. The cells were divided into three groups for the experiment. A, Control group: the medium for normal ovarian granulosa cells. B, VCD group: the medium with VCD at a concentration of 30 µM. C, VCD + ADSC-EVs group: the medium with VCD at a concentration of 30 µM and with ADSC-EVs of 10 µg/mL. The culture medium and cells were collected at 24 h and 48 h after treatment and utilized in following experiments. TUNEL assay TUNEL cell apoptosis detection kit (Beyotime, China) was used to examine apoptosis in paraffin section of rat ovaries. The analysis was performed according to the manufacturer’s instructions. In brief, remove paraffin by immersing slides in xylene and rehydrate. Add 20 µg/mL DNase-free protease K dropwise onto slides and incubate for 15 min at 22℃. Then, the sample was incubated with the TUNEL detection solution. Finally, the samples were analyzed by fluorescence microscopy (Leica, Germany). Cell counting kit (CCK)-8 assay Primary rat ovarian granulosa cells were seeded into 96-well plates. The cells were monitored for proliferation every 24 h for a total of 9 days, with five replicate wells for each time point. At each time point, 10 µL of CCK-8 reagent (Beyotime, China) was added to each well, and the plates were returned to the incubator for additional 4 h. The absorbance at 450 nm was measured using a microplate reader (Bio-Tek, USA). The cell proliferation rate was calculated as follows: Cell proliferation inhibition rate (%) = (1– average OD value of each treatment group / average OD value of control group) × 100%. Apoptosis detection by Annexin V/PI staining The proportions of early and late apoptotic cells were assessed using FITC Annexin V Apoptosis Detection Kit I (BD Pharmingen™, USA). After 48 h of culture in medium with or without VCD and ADSC-EVs, the cells from each group were harvested and washed with PBS. One hundred thousand cells from each group were then incubated in Annexin V-FITC binding buffer with Annexin V-FITC and propidium iodide (PI) at room temperature in the dark for 15 min. The FITC and PI emission wavelengths of stained cells were analyzed by flow cytometry (BD FACSymphony™ A1). SMART-Pico mRNA sequence Total RNA was isolated from ADSC-EVs samples using the TRIzol (Invitrogen, CA, USA). RNA concentration and purity were measured using a NanoDrop ND-1000 spectrophotometer (NanoDrop, Wilmington, DE, USA). RNA integrity was evaluated using the Agilent Bioanalyzer 2100 (Agilent Technologies, CA, USA), with samples meeting thresholds of DV200 > 30% or RIN > 4.0. Integrity was additionally confirmed via denaturing agarose gel electrophoresis. Stranded RNA-seq libraries were constructed using the SMARTer^® Stranded Total RNA-Seq Kit v2–Pico Input Mammalian Kit (Takara Bio USA, Inc.). Briefly, 10–50 ng of RNA was denatured at 72 °C for 3 min with SMART Pico Oligos Mix, followed by first-strand cDNA synthesis at 42 °C for 90 min using Template Switching Oligo Mix (TSO) as reverse transcription primers. cDNA was amplified via 5 cycles of PCR with SeqAmp DNA Polymerase to incorporate sequencing adapters and index tags. Amplified DNA was purified using AMPure XP beads (Beckman Coulter) and resuspended in ZapR v2 Master Mix to generate preliminary libraries. Ribosomal RNA-derived cDNA was removed by hybridization with R-Probes v2 and digestion using ZapR™ v2 enzyme (37 °C, 60 min). The rRNA-depleted product was amplified with universal primers (12–16 cycles, adjusted based on input RNA quality) and purified with AMPure XP beads. Final libraries were resuspended in nuclease-free water and stored at − 20 °C for up to 2 months. Libraries were paired-end sequenced (PE150) on an Illumina NovaSeq™ 6000 platform (LC Bio Technology Co., Ltd., Hangzhou, China) following standard protocols. Reads were mapped to the rat genome using STAR version 2.7.3, and error-corrected UMI counts were calculated from Ensemble gene annotations (Rattus_norvegicus.mRatBN7.2.112). Fragments Per Kilobase of transcript per Million mapped reads (FPKM) were used for differential gene analysis. Astral-DIA proteomics To streamline the protein analysis workflow, we began by extracting proteins from weighed samples using a lysis solution with 8 M Urea/50mM Tris-HCl and Roche cocktail, followed by homogenization and centrifugation to isolate the supernatant. Protein quality was ensured through Bradford quantitation and SDS-PAGE analysis. For protein digestion, we used trypsin on 150 µg of protein, resulting in peptides that underwent desalting and drying. These peptides were then analyzed via Nano-LC-MS/MS across multiple platforms: Orbitrap Exploris™ 480 with specific DIA mode settings, Astral with both 8 min and 24 min gradient separations, and timsTOF pro2 in the diaPASEF mode. DIA-NN was employed for library-free DIA data analysis, searching MS/MS data against the Uniprot database with parameters set for trypsin specificity, allowing up to two missed cleavages, and accounting for fixed and variable modifications. Results were stringently filtered at a 1% false discovery rate (FDR) for inclusion in subsequent analyses, ensuring a robust and comprehensive protein identification and quantification process. Data analysis Values represent mean ± standard error of the mean (SEM), with the number of samples (n) being ≥ 3 in independent experiments. GraphPad Prism 8.3 (Prism, USA) was used for data analysis and plotting. Data between groups were analyzed using two-tailed Student’s t-test for two groups and one-way ANOVA with Tukey’s post hoc test for multiple groups to determine statistical significance. p < 0.05 indicated statistical significance. Results ADSCs were successfully enriched and exhibited multipotency Rat ADSCs were isolated from inguinal fat pads and exhibited typical spindle-shaped morphology (Fig. [105]1A). Differentiation assays confirmed that ADSCs exhibited multipotency by successfully differentiating into adipocytes, osteocytes, and chondrocytes (Fig. [106]1B–D). Flow cytometry analysis revealed positive expression of CD44 (90.2%), CD73 (98.4%), CD90 (96.9%), and negative expression of CD11b (0.19%), CD34 (0.67%), CD45 (0.71%), confirming their identity as ADSCs (Fig. [107]1E). Then, ADSC-EVs were extracted from the medium of ADSCs (Fig. [108]1F), exhibiting a median diameter of 125.6 nm, with a particle concentration of 1.2 × 10^9 particles/mL detected by NTA (Fig. [109]1G). Transmission electron microscopy and western blotting confirmed the presence of ADSC-EVs, characterized by their lipid bilayer membrane and expression of specific membrane proteins (CD9, CD63) and membrane generation-related complex TSG101 (Fig. [110]1H, I). Fig. 1. [111]Fig. 1 [112]Open in a new tab Preparation and identification of ADSCs’ and ADSC-EVs. A: Photo of the primary ADSCs isolated from rat after culturing for 4 days. Bar = 100 μm. B: Oil-red staining of ADSCs after induction into adipocytes. Bar = 100 μm. C: Alizarin-red staining of ADSCs after induction into osteoblast. Bar = 100 μm. D: Alcian blue staining of ADSCs after induction into chondrocyte. Bar:100 μm. E: The surface markers of ADSCs detected by flow cytometry. The markers included positive expression of CD44, CD73, CD90 and negative expression of CD11b, CD34, CD45. The control was the negative cocktail (NC). F: The schematic diagram of ADSC-EVs abstraction protocol. G: The NTA detection result of ADSC-EVs. x-axis was particle diameter and y-axis was particle count. H: The morphology of ADSC-EVs observed by transmission electron microscopy. Bar = 200 nm. I: Western blotting of markers (TSG101, CD9, CD63) of ADSC-EVs. Control panel was positive control sample. Full-length blots are presented in Supplementary Fig. [113]1 Rats treated with VCD successfully mimiced the majority of clinical manifestations associated with POI Continuous intraperitoneal injection of VCD (160 mg/kg) for 15 days established a POI model of rats. To assess this model of POI, estrous cycles, serum hormones, and follicle counts were analyzed. Two weeks following administration of VCD, the incidence of estrous cycle disorders in rats progressively increased, culminating at 8–10 weeks, by which time all rats exhibited estrous cycle disorders (Fig. [114]2A). The ELISA results of serum hormone revealed a gradual decline in AMH level and an increase in FSH level between 4 and 10 weeks after VCD administration, while E2 level showed a notably decrease until 10 weeks (Fig. [115]2B–D). H&E staining and quantification analysis of ovary sections revealed a decrease in primordial and primary follicles four weeks after VCD administration (Fig. [116]2E–G). Additionally, secondary and antral follicles showed a decline ten weeks post-administration (Fig. [117]2H, I). Above all, this model effectively mimicked the progressive characteristic of human POI and was used for subsequent experiments. Fig. 2. [118]Fig. 2 [119]Open in a new tab Rat POI model of long-term injection of VCD. A: Disordered estrous cycle percent of rats. B: The serum E2 levels assessed by ELISA after PBS or VCD administration for 4 weeks, 7 weeks, and 10 weeks. C: The serum FSH levels assessed by ELISA of rats after PBS or VCD administration for 4 weeks, 7 weeks, and 10 weeks. D: The serum AMH levels assessed by ELISA of rats after PBS or VCD administration for 4 weeks, 7 weeks, and 10 weeks. E: The HE staining of rat ovaries in control group and VCD group after PBS or VCD administration for 4 weeks, 7 weeks, and 10 weeks. Bar = 200 μm. F: The line graph of primordial follicle counts at 4 weeks, 7 weeks, and 10 weeks. G: The line graph of primary follicle counts at 4 weeks, 7 weeks, and 10 weeks. H: The line graph of secondary follicle counts at 4 weeks, 7 weeks, and 10 weeks. I: The line graph of antral follicle counts at 4 weeks, 7 weeks, and 10 weeks. Each group at each time point, n = 3, and values represent mean ± SEM. Statistical analyses were carried out by two-tailed student’s t-test. *p < 0.05; **p < 0.01, ***p < 0.001, as compared with the control group at same time point Therapeutic efficacy of ADSCs and ADSC-EVs on POI rats To assess the effectiveness of ADSCs and ADSC-EVs on POI in rats, we administered in situ injections of PBS (VCD group), ADSCs (VCD + ADSCs group), and ADSC-EVs (VCD + EVs group) into the ovaries of POI rats. And the control group received PBS injections. Four weeks after treatment, the estrous cycle and serum hormone were analyzed to assess the ovarian function of rats (Fig. [120]3A). Prolonged estrous cycles were observed in the VCD group, whereas the duration and proportion of regular cycles were restored in the VCD + ADSCs and VCD + EVs groups (Fig. [121]3B, C). Subsequently, the HE section showed that the VCD + ADSCs and VCD + EVs groups have significantly increased numbers of follicles, especially in the primordial and primary follicles comparing to the VCD group (p < 0.001, Fig. [122]3D, E). In order to evaluate the ovulation capacity after treatment, a PMSG/hCG ovulation induction experiment was conducted. Rats in VCD + ADSCs and VCD + EVs group exhibited significantly higher numbers of retrieved oocytes per animal (41.4 ± 1.2 in the VCD + ADSCs group, and 44 ± 1.5 in the VCD + EVs group ) compared to the VCD group (29.4 ± 3.3; p < 0.01)(Fig. [123]3F, G). Additionally, E2 and AMH levels were decreased significantly in the VCD group but were restored in the VCD + ADSCs and VCD + EVs groups (p < 0.001, Fig. [124]3H, J). Simultaneously, FSH levels were downregulated in both treatment groups (p < 0.001, Fig. [125]3I). The improvement of estrous cycle, hormone levels, and follicle count in the VCD + EVs group exhibited a superior condition compared to those in the VCD + ADSCs group, but the number of retrieved oocytes was similar in both groups. In summary, both ADSCs and ADSC-EVs had significant therapeutic effect on ovarian function in POI rats. Fig. 3. [126]Fig. 3 [127]Open in a new tab The therapeutic benefits of ADSCs and ADSC-EVs on POI rats. A: Scheme of establishment of VCD-induced POI model and treatment of ADSCs and ADSC-EVs. B: Representative estrous cycles of rat in control, VCD, VCD + ADSCs, and VCD + EVs groups. P, proestrus, E, estrus, M, metestrus, D, diestrus. C: Distribution of estrous cycle stages (%) in each group. D: Quantitative analysis of the primordial follicles, primary follicles, secondary follicles and antral follicles in each group. E: Representative images of H&E staining of ovaries from each group. Bar = 200 μm. F: Representative images of MII oocytes obtained by superovulation induction in each group. Bar = 100 μm. G: Bar graph showing the number of retrieved oocytes per rat in each group with PMSG/hCG-treatment. H: The serum E2 levels assessed by ELISA of rats in each group 4 weeks after treatment. I: The serum FSH levels assessed by ELISA of rats in each group 4 weeks after treatment. J: The serum AMH levels assessed by ELISA of rats in each group 4 weeks after treatment. n = 5 in each group. Values represent mean ± SEM. Statistical analyses were carried out by one-way ANOVA with Tukey’s post hoc test. *p < 0.05; **p < 0.01, ***p < 0.001. NS, no significance ADSCs and ADSC-EVs rescued POI by inhibition of apoptosis and activation of KITL/KIT/PI3K/AKT pathway The decline of ovarian function caused by VCD involved promoting atresia of primary and primary follicles and abnormal apoptosis of granulosa cells. TUNEL staining showed that the apoptosis was heightened in VCD-induced POI rats, which was mitigated by the treatment with ADSCs and ADSC-EVs (Fig. [128]4A, B). RT-qPCR analysis revealed that ADSCs and ADSC-EVs reduced the expression of Bax (pro-apoptosis gene) and Caspase3 (apoptosis marker) in ovary of POI rat, while increasing Bcl2 (anti-apoptosis gene) (p < 0.001, Fig. [129]4D). Meanwhile, the protein levels of BAX/BCL2, CASPASE3 declined after treatment of ADSCs and ADSC-EVs detected by IHC and western blotting (Fig. [130]4C, F). Furthermore, the pathway KITL/KIT/PI3K/AKT, which is involved in communication between granulosa cells and oocytes during follicle development, was upregulated in ADSCs and ADSC-EVs groups (Fig. [131]4E, F). Collectively, ADSCs and ADSC-EVs ameliorate ovarian function by inhibiting apoptosis and enhancing the KITL/KIT/PI3K/AKT pathway after VCD treatment in vivo. Fig. 4. [132]Fig. 4 [133]Open in a new tab ADSCs and ADSC-EVs rescue POI by inhibition of apoptosis and activation of KITL/KIT/PI3K/AKT pathway. A: Immunofluorescence images of ovaries stained with TUNEL in each group. Green: TUNEL, Blue: DAPI. Bar = 100 μm. B: The TUNELpositive cell percent of ovarian section in each group, n = 5. C: The IHC staining images of ovaries labeling BAX and CASPASE3 in each group. Bar = 100 μm. D: RT-qPCR results of Bax, Caspase3, and Bcl-2 in ovaries. n = 3. E: RT-qPCR results of Kit, Kitl, PI3K, and Akt in ovaries. n = 3. F: Western blotting image of BAX, CASPASE3, BCL2, KIT, KITL, PI3K, and AKT in ovaries. Panels represent control group (Control), VCD group (VCD), VCD + ADSCs group (ADSCs), VCD + ADSC-EVs group (EVs). β-ACTIN served as the loading control. Full-length blots are presented in Supplementary Fig. [134]2. The experiments have been replicated over three times. Values represent mean ± SEM. Statistical analyses were carried out by one-way ANOVA with Tukey’s post hoc test. *p < 0.05; **p < 0.01, ***p < 0.001 ADSC-EVs promoted VCD-damaged granulosa cell viability and function To investigate the potential mechanisms by which ADSC-EVs contribute to the treatment of POI, the primary granulosa cells from rats were isolated and treated in vitro. After exposure to 30 µM VCD, CCK-8 assay results demonstrated that ADSC-EVs mitigated VCD-induced inhibition of proliferation (2.25 ± 0.34% and 14.19 ± 1.69% for the VCD + EVs group compared to 10.69 ± 0.40% at 24 h and 40.41 ± 0.80% at 48 h for VCD group, p < 0.001, Fig. [135]5A). Annexin-V/PI double staining revealed that the ADSC-EVs increased the proportion of viable cells, and reduced the proportion of both early and late apoptotic cells (all p < 0.05, Fig. [136]5B). Additionally, E2 and progesterone secretion of granulosa cells was inhibited in the VCD group; however, the ADSC-EVs upregulated the secretions (p < 0.001) (Fig. [137]5C, D). Mechanistically, classical apoptosis regulation pathway Bax/Bcl2 could activate Caspase3 and initiate a cascade reaction of cell apoptosis, which was concerned with VCD induced POI. RT-qPCR analysis showed that Bax and Caspase3 were significantly decreased in the VCD + EVs group (p < 0.001); while the level of Bcl-2 were significantly increased (p < 0.001) comparing to the VCD group (Fig. [138]5E). Western blotting analysis confirmed the same trend (Fig. [139]5F). Additionally, Kit/Kitl pathway also involves in VCD induced POI. We observed that ADSC-EVs restored the expression of Kitl in granulosa cells (Fig. [140]5E, F). These findings demonstrate that ADSC-EVs exert a significant protective effect on granulosa cells damaged by VCD, enhancing survival, preserving cellular functionality, and upregulating the expression of Kitl. Fig. 5. [141]Fig. 5 [142]Open in a new tab ADSC-EVs promoted VCD-damaged granulosa cell viability and function. A: The proliferation inhibition rate detected by CCK8 assay of granulosa cell cultured with or without ADSC-EVs for 24 h and 48 h. n = 3. B: Annexin-V/PI staining and flow cytometry detection of granulosa cell cultured with or without VCD and ADSC-EVs for 24 h. n = 3. C: The medium E2 levels assessed by ELISA of granulosa cell cultured with or without VCD and ADSC-EVs for 24 h and 48 h. n = 3. D: The medium progesterone (P) levels assessed by ELISA of granulosa cell cultured with or without VCD and ADSC-EVs for 24 h and 48 h. n = 3. E: RT-qPCR results of Bax, Caspase3, Bcl-2, and Kitl in granulosa cell cultured with or without VCD and ADSC-EVs for 24 h. n = 3. F: Western blotting image of BAX, CASPASE3, BCL2, and KITL in granulosa cell cultured with or without VCD and ADSC-EVs for 24 h. Full-length blots are presented in Supplementary Fig. [143]3. The experiments have been replicated over three times. Values represent mean ± SEM. Statistical analyses were carried out by two-tailed student’s t-test and one way ANOVA with Tukey’s post hoc test. *p < 0.05; **p < 0.01, ***p < 0.001 Total RNA-seq of ADSC-EVs and proteomic analysis of ADSCs/ADSC-EVs revealed multiple bioactive molecules involved in the PI3K/AKT pathway To further investigate the bioactive substances in ADSC-EVs, we performed whole-transcriptome sequencing of ADSC-EVs. Given the low RNA content in EVs, we used the SMART cDNA synthesis method to obtain more comprehensive gene information (Fig. [144]6A). The mRNAs detected in the three EV samples exhibited high molecular correlation and similar numbers of expressed genes (Fig. [145]S1A, B), with some heterogeneous expression also observed (Fig. [146]6B). Due to the high sensitivity of this sequencing method, we detected over 20,000 genes in total. We focused on analyzing the top 1000 highly expressed genes common to all three samples. KEGG pathway enrichment analysis revealed that these highly expressed genes were enriched in the PI3K/AKT pathway, motor proteins, oxidative phosphorylation, and ribosome (Fig. [147]6C). GO enrichment analysis focused on translation, nucleus, and ribosome (Fig. [148]6D). We also performed enrichment analysis on all genes expressed in the three samples. GO enrichment analysis mainly identified genes related to G-protein-coupled receptor signaling pathways, DNA transcription and translation regulation processes, cell membrane components, and protein-binding genes (Fig. [149]S1C). KEGG enrichment analysis showed enrichment in endocytosis, neuroreceptor signal interaction, PI3K/AKT pathway, and oxidative phosphorylation (Fig. [150]S1D). Additionally, we identified the expression of Kitl/Kit/PI3K/Akt pathway-related genes in the transcriptome sequencing (Fig. [151]6E). We validated the key genes by qPCR (Fig. [152]6F) and found that Kit and PI3K were highly expressed in ADSC-EVs compared to ADSCs, while Kitl had a higher expression level in ADSCs. Furthermore, we conducted 4D-DIA proteomics to detect bioactive proteins potentially present in ADSCs and ADSC-EVs. Principal component analysis (PCA) demonstrated that ADSCs and ADSC-EVs can be divided into two distinct groups (Fig. [153]S1E). Compared to cells, EVs had 625 significantly upregulated proteins and 4278 significantly downregulated proteins (Fig. [154]7A-C), with EV-specific molecules being highly expressed (Fig. [155]7D). We performed KEGG enrichment analysis on the significantly upregulated molecules in EVs (Fig. [156]7E) and found enrichment in the PI3K/AKT signaling pathway and cell adhesion-related pathway. The heatmap of the positive regulatory molecules of the PI3K/AKT pathway (Fig. [157]7F) identified different highly expressed molecules in ADSCs and ADSC-EVs. These results indicate that ADSC-EVs specifically enrich KIT/PI3K/AKT-related mRNAs and proteins, which can regulate the downstream signaling pathways of target cells. Fig. 6. [158]Fig. 6 [159]Open in a new tab SMART-Pico mRNA sequence of ADSC-EVs. A: The flowchart of ADSC-EVs SMART-Pico mRNA sequencing. n = 3. B: Venn diagram showing the distribution of all detected genes in three EV samples. C: KEGG pathway enrichment analysis of the top 1000 highly expressed genes in EV samples. The y-axis represents the number of genes, and the x-axis represents the KEGG pathway names. D: GO pathway enrichment analysis of the top 1000 highly expressed genes in EVs samples. The y - axis represents the number of genes, and the x - axis represents the pathway names. E: The FPKM of PI3K/AKT pathway molecules. F: RT-qPCR analysis of Kit, Kitl, and PI3K expression in ADSCs-EVs and ADSCs (n = 3). Values represent mean ± SEM. Statistical analyses were carried out by two-tailed student’s t-test. *p < 0.05; **p < 0.01, ***p < 0.001. Fig. 7. [160]Fig. 7 [161]Open in a new tab Proteomics sequence of ADSCs and ADSC-EVs. A: Differential protein scatter plot of ADSCs and ADSC-EVs samples. With the log[2](Exp + 1) of the control group on the x-axis and the log[2](Exp + 1) of the experimental group on the y-axis. The color gradient from red to blue indicates the significance of expression differences. B: Bar chart of the number of proteins significantly increased and decreased in EVs compared to ADSCs. C: Volcano plot of differentially expressed proteins in ADSCs and ADSC-EVs samples. The x–axis represents log[2](FC), and the y–axis represents–log[10](pValue). Red dots represent significantly upregulated proteins, blue dots represent significantly downregulated proteins, and gray dots represent non–significantly different proteins. D: Heatmap of the expression levels of EV–specific proteins between the two groups. E: Heatmap of the expression levels of molecules with positive regulatory effects on PI3K/AKT pathway between the two groups. F: KEGG pathway enrichment analysis of upregulated differentially expressed proteins in ADSC-EVs Discussion Premature ovarian insufficiency (POI) continues to be a challenging condition with limited treatment options. Stem cell-based therapies, including adipose-derived stem cells (ADSCs) and ADSC-derived extracellular vesicles (ADSC-EVs), offer promising approaches for POI management. In this study, we demonstrated that both ADSCs and ADSC-EVs had significant therapeutic effects on ovarian function in POI rats. Mechanistically, ADSCs and ADSC-EVs reduced granulosa cell apoptosis and activated the KITL/KIT/PI3K/AKT pathway in the POI rat models. Our results provide strong evidence supporting the clinical use of ADSC-EVs as a promising treatment for POI. ADSC-EVs, with their cell-free characteristic, ease of use, and lower immunogenicity, represent a promising alternative to ADSCs. Our study simulated the extraction and purification processes of ADSCs and ADSC-EVs and validated their efficacy and mechanisms in the treatment of POI. It was of great clinical significance that allogeneic or autologous ADSCs and ADSC-EVs could be applied in POI patients and restore the ovarian function. Our findings show that ADSC-EVs show comparable effect in repairing and protecting ovarian function to ADSCs. In some indicators, such as estrous cycle, follicle count, hormone level detection, and mRNA improvement, EVs show slightly superior trends to ADSCs. We believe that the administration of ADSC-EVs will enhance the local content of some bioactive components, such as mRNA, miRNA, and protein. RT-qPCR assay revealed that ADSC-EVs contained much higher levels of Kit and PI3K mRNA compared to ADSCs (Fig. [162]6F), which may account for their enhanced efficacy in activating the KITL/KIT/PI3K/AKT pathway. In the superovulation model, an excessive number of follicles were mobilized. Both ADSCs and ADSC-EVs demonstrated the ability to counteract the VCD-induced decline, achieving levels close to the control group, hence no significant difference was found between them at this stage. Regarding the apoptosis pathway, Bcl2/Bax and Caspase3 serve as indicators of early apoptosis, whereas TUNEL marks mid-stage apoptosis [[163]51]. Both treatments effectively suppressed early apoptosis, with EVs exhibiting a slightly more pronounced modulatory effect at the mRNA level. The lack of intergroup differences in TUNEL-positive cells implies that both treatments provided comparable cytoprotective effects over time. In all, ADSC-EVs can achieve therapeutic effects comparable to those of ADSCs. Our study demonstrates the therapeutic potential of ADSC-EVs in treating POI in an animal model. However, several critical considerations must be addressed before translating these findings into clinical applications [[164]17]. First, a more systematic and comprehensive safety assessment is essential, particularly regarding reproductive toxicity. Although our results showed no immediate adverse effects in the animal model, the long-term impact of ADSC-EVs on ovarian function, fertility, and potential transgenerational effects remain unclear. Rigorous preclinical studies, including chronic toxicity evaluations and multi-generational reproductive assessments, are necessary to ensure the safety of ADSC-EVs in humans. Furthermore, the heterogeneity of ADSC-EVs poses a challenge for clinical standardization. Variations in EVs size, cargo composition, and bioactivity depending on donor characteristics, isolation methods, and culture conditions could influence therapeutic outcomes. Developing standardized protocols for EVs isolation, characterization, and quality control is crucial to ensure consistency and reproducibility in clinical settings. Additionally, the optimal dosage, route of administration, and frequency of ADSC-EVs treatment need to be carefully determined through well-designed pharmacokinetic and pharmacodynamic studies. Another important consideration is the potential immunogenicity of ADSC-EVs. While EVs are generally considered less immunogenic than their parent cells, the presence of donor-specific antigens or bioactive molecules could still elicit immune responses in recipients. Comprehensive immunogenicity profiling, including in vitro and in vivo immune response assessments, should be conducted to minimize the risk of adverse immune reactions. Lastly, the ethical and regulatory aspects of using MSC-EVs for POI treatment must be addressed. Clear guidelines for donor consent, EVs sourcing, and clinical trial design are necessary to ensure ethical compliance and patient safety. Collaborative efforts between researchers, clinicians, and regulatory agencies will be critical to advancing ADSC-EV-based therapies from bench to bedside. The latest research has proposed optimized extraction procedures for EVs, methods to increase yield, and has focused on engineered vesicles capable of targeted enrichment [[165]52]. The development of new technologies will provide further support for the clinical translation of MSC-EVs. The underlying mechanisms by which MSCs and MSC-EVs improve ovarian function have been inconsistent across various animal models and MSCs types. Previous studies reported that ADSCs administered intravenously significantly reduced cyclophosphamide-induced senescence and apoptosis in granulosa cells and increased the growth factors in rat ovaries [[166]20, [167]21]. ADSC-EVs targeted on SMAD and AMPK/mTOR pathway to improve ovarian function in mice [[168]27, [169]28]. Considering the advantages of VCD-induced POI model, we investigated the ADSC-EVs effect on this model. Preliminary biochemical and molecular investigations revealed that VCD activated the intracellular pro-apoptotic pathway and related to Bad, Bax, and Bcl-XL genes [[170]50, [171]53–[172]56]. Based on these researches, the cell apoptosis was assessed in this study. The TUNEL staining and cell apoptosis detection revealed that ADSCs and ADSC-EVs rescued VCD-induced apoptosis of granulosa cells, which involved the expression of apoptosis regulation genes Bax, Bcl2 and Caspase3. Consistent evidence from various studies indicates that mediators secreted by MSCs, including interleukin-6 [[173]57], prostaglandin E2 [[174]58], microRNA (miR) -29a-3p, miR-125b-5p, and miR-93 can modulate the expression of Bax/Bcl-2 in diseased cells [[175]59–[176]62]. Consequently, the therapeutic impact of ADSCs and ADSC-EVs is linked to the modulation of mRNA and protein expression associated with apoptosis and anti-apoptosis pathways. While granulosa cell survival supports ovarian function, further research is needed to assess the impact of ADSCs and ADSC-EVs on oocytes. Given that VCD targets the KIT receptor on the oocyte membrane [[177]55, [178]63–[179]65], we evaluated the expression of KIT/PI3K/AKT pathway and the KIT ligand, known as KITL. Surprisingly, the expression of the KITL was significantly increased in the granulosa cell treated with ADSC-EVs and the KITL/KIT/PI3K/AKT pathway was upregulated in ovaries with ADSCs and ADSC-EVs treatment as well. The interaction between granulosa cell-derived KITL and KIT produced by oocytes and theca cells plays a significant role in follicular development, including granulosa cell proliferation, differentiation, and migration, as well as the survival of oocytes [[180]66, [181]67]. Furthermore, the KIT/PI3K/AKT pathway is implicated in controlling the dormancy and activation of primordial follicles [[182]32]. We hypothesized that ADSCs and ADSC-EVs elevate KITL expression in granulosa cells, which then bind to the KIT receptors on the membrane of oocytes in primordial and primary follicles and activate the PI3K-AKT pathway. This interaction subsequently activates anti-apoptotic genes and follicle development-related genes downstream of AKT in oocytes. Enhancing the vitality of granulosa cells and oocytes improves the maintenance and development of primordial and primary follicles, thus boosting ovarian function in rats. Subsequently, it was of interest to investigate the approach by which ADSC-EVs stimulate the expression of KITL in granulosa cells. Studies revealed that MSCs expressed KIT and KIT-positive ADSCs promote the cell growth and angiogenesis [[183]68, [184]69]. Cell-associated or soluble KIT activated membrane-bound KITL, which then triggers the AKT/mTOR/CREB signaling pathway and promotes cell proliferation [[185]70]. Additionally, the secretome of human umbilical cord MSCs contained hepatocyte growth factor, which binding with receptor c-Met in granulosa cell and promoted the expression of KITL [[186]71]. Therefore, ADSCs may express and secret KIT, which are present in ADSC-EVs, with other bioactive factors that modulates the expression of KITL in granulosa cell. In our study, we found that ADSC-EVs expressed mRNAs and regulatory proteins related to the KITL/KIT/PI3K/AKT pathway. This suggests that their effect on target cells may be achieved through both mRNA and protein delivery. Additionally, molecules related to cell adhesion and ribosomes were enriched in EVs, which is consistent with previous sequencing results [[187]72, [188]73]. This may be related to their functions in cell fusion and regulation of protein synthesis in target cells. Combining previous studies, the effect of EVs on target cells is likely the result of the combined regulation of various substances and factors [[189]74]. In different disease models, the emphasized regulatory molecules contained in EVs exert their respective therapeutic effects. A key focus for future research is to delve into how to assess the quality of EVs products and identify key molecules to ensure that EVs from different batches and sources can be applied to therapy. Comprehensive validation studies need to be conducted and supplemented, which will also provide more research ideas for the development of engineered EVs [[190]52]. This study has several limitations. First, the ADSC-EVs used were derived from rats, and validation in human cells or tissues was not conducted, which may limit direct extrapolation to human conditions. Second, the VCD-induced POI model represents a relatively simplified approach that does not fully capture the complexity of clinical POI etiologies. Despite these limitations, we have made every effort to simulate the extraction of ADSC-EVs and the intra ovarian injection, which holds significant implications for clinical applications. ADSCs can be autologously supplied, thereby circumventing the risk of immune rejection. Additionally, in situ injection enables targeted delivery to the ovary, reducing losses due to circulation and minimizing the impact on other organs. Although some clinical trials have initiated in this area, there have been no reports on the outcomes of MSC-EVs for the treatment of POI [[191]75]. Our team is currently working on related projects to promote the clinical translation of MSCs and MSC-EVs, investigating the precise efficacy and mechanisms of MSC-EVs based therapy for POI, with the aim of benefiting a broad patient population. In summary, the restoration and protection of ovarian function in POI rats by ADSCs and ADSC-EVs may be mediated through the activation of the KITL/KIT/PI3K/AKT pathway and inhibition of apoptosis. Our findings highlight ADSC-EVs as a highly promising cell-free therapeutic strategy for POI. Conclusion Allogeneic treatment of ADSCs and ADSC-EVs on VCD-induced POI obtained great therapeutic effect, mirroring the human POI process and therapy. The therapeutic mechanism involves repressing the apoptosis in granulosa cells and triggering the KITL/KIT/PI3K/AKT signaling cascade, thereby improving follicle development and ovulation. This study suggests that ADSC-EVs have considerable potential for clinical translation and application, presenting a novel and effective strategy for managing POI. Supplementary Information Below is the link to the electronic supplementary material. [192]Supplementary Material 1^ (25.2KB, xlsx) [193]Supplementary Material 2^ (715.2KB, docx) Author contributions L.W, G.FH, and Z.Q conceived and designed the study. L.YQ, C.F, and Z.W completed most of the data collection and analysis. S.D, Z.L, Q.R, H.QH, L.SM, C.BY, D.HX, and Z.ML participated in sample collection, data analysis, and further explanation. L.W, G.FH, Z.Q, and L.YQ wrote the manuscript. All authors made substantial contributions in critically revising the manuscript and approved the final version of manuscript. Funding This work was supported by the National Key Research and Development Program of China (2022YFC2703002), National Natural Science Foundation of China (82371726, 82200541, 82401899), Joint Funds of the National Natural Science Foundation of China (U24A20658), Innovative Research Team of High-Level Local Universities in Shanghai (SHSMU-ZDCX20212200), Shanghai Hospital Development Center Foundation (SHDC22022303), Key project of Medical and Industrial intersection of Shanghai Jiao Tong University (YG2023ZD27), and the Postdoctoral Fellowship Program of CPSF (GZC20241036). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. Data availability The data underlying this article will be shared on reasonable request to the corresponding author. The transcriptomic datasets generated in this study are deposited in the Gene Expression Omnibus (GEO) repository under accession number [194]GSE301855 (accessible at [195]https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE301855). The proteomics datasets are available in the iProX database with Project ID: IPX00125260 (accessible at [196]https://www.iprox.cn//page/project.html?id=IPX0012526000). All supporting data not contained in these repositories are included in this article and its supplementary information files. Declarations Ethics approval The authors declare that they have not use AI-generated work in this manuscript. This study didn’t involve patient participation. Conflict of interest The authors declare no competing financial or non-financial interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yanquan Li, Feng Chen and Wei Zhao contributed equally to this work. Contributor Information Qing Zhang, Email: zhangqing081@sina.com. Fanghao Guo, Email: fhguo@sibs.ac.cn. Wen Li, Email: liwen@shsmu.edu.cn. References