2.1 Patient sample collection

The participants in this research were chosen from among surgical patients in our department. Those with POP underwent either vaginal anterior wall repair or anterior pelvic floor reconstruction surgery, following the Pelvic Organ Prolapse Quantification (POP-Q) system, as evaluated by at least two senior gynaecologists. The primary focus was on assessing the degree of uterine prolapse and varying degrees of anterior and posterior vaginal wall protrusion. The control group comprised individuals of the same age range who had undergone total hysterectomy for different gynaecological conditions during the same timeframe. Exclusion criteria for the study encompassed patients with cervical precancerous lesions, early-stage cancer or PFD.

All patients in each group had experienced natural menopause and had not used systemic hormones in the 3 months leading up to the surgery. They had no history of neurological or respiratory system diseases, functional ovarian tumours, metabolic diseases, connective tissue diseases, prior pelvic floor surgeries, genitourinary infections, pelvic deformities or related conditions.

Patients received preoperative information about the study and provided signed informed consent forms. Surgical procedures for the POP group involved either vaginal anterior wall repair or anterior pelvic floor reconstruction, while the control group underwent either vaginal or abdominal hysterectomy. During surgery, a cold knife was utilized to obtain a full-thickness tissue sample from the region near the vault of the vagina, measuring approximately 1.0 cm × 1.0 cm. The tissue was rinsed with physiological saline, placed in a sterile, enzyme-free EP tube labelled for identification, promptly immersed in liquid nitrogen and stored in a − 80°C freezer within 20 min for subsequent frozen preservation.

The study was approved by Shanghai First Maternity and Infant Hospital, School of Medicine, Tongji University, and written consent was obtained from the participant.

2.2 Primary rat BMSC isolation and culture

The research utilized 6-month-old female Sprague-Dawley rats and received approval from the ethics committee at Shanghai First Maternity and Infant Hospital. The femurs of the rats were dislocated under anaesthesia, and bone marrow cells were flushed using 10 mL syringes with 5–10 mL cold Iscove's modified Dulbecco medium (IMDM). After centrifugation at 150 × g for 5 min, the cell pellet was resuspended in IMDM, layered onto Percol separation solution (density 1.073 g/mL, Sigma-Aldrich, St. Louis, MO) and centrifuged at 400 × g for 30 min. The cotton-like cells at the interface were collected, rinsed and cultured with IMDM supplemented with 20% foetal bovine serum and 1% streptomycin/penicillin (Sigma-Aldrich). The medium was changed after 24 h, and cells were maintained until approximately 80% confluence. After trypsinization, replating and cultivation until the second passage, plastic-adherent, fibroblast-like cells were considered BMSCs. To verify the multidirectional differentiation potential of BMSCs, we induced the cells into osteoblastic, adipogenic and chondrogenic cells for 20 to 30 days. After the differentiation of the cells to a certain extent, they were treated with the corresponding stains and finally photographed with a microscope (Olympus, Japan). These cells were harvested for characterization. Before use, the harvested BMSCs were stained and verified with fluorescein isothiocyanate-conjugated mouse anti-rat CD44, CD34, CD105 and CD45 (BD Bioscience, Franklin Lakes, NJ).

2.3 Exosome extraction and identification

The BMSCs were verified by their capability of osteogenic and adipogenic differentiation, as well as the expression of surface molecular markers of CD44, CD105, CD34 and CD45 examined by the flow cytometry. BMSCs were cultured in a serum-free DMEM medium for 48 h at 37°C with 5% CO₂. Recombinant human TNF-α was obtained from Biolegend (San Diego, CA). BMSCs at the third passage were cultured with or without 20 ng/mL TNF-α for 48 h. The resulting supernatant was gathered and underwent a series of procedures, including centrifugation (300 g, 4°C, 10 min), subsequent centrifugation (16,500×g, 4°C, 20 min) and filtration using 0.22 μm filters for sterilization. Following this, the supernatant experienced ultra-high-speed centrifugation (120,000×g, 70 min, 4°C) to yield precipitates, which were then resuspended in phosphate-buffered saline (PBS) to form an exosomal suspension. Nanoparticle tracking analysis using NanoSight NS300 equipment (Malvern Instruments Ltd., Malvern, UK) was employed to disclose the size distribution of exosomes.

2.4 PFD rat model

The rat model of vaginal distention, previously documented,^{19, 20} was employed to simulate childbirth injury. In brief, a Foley balloon filled with 2.5–3.0 mL water was introduced into the rat vagina on Day 0, connected to a pressure transducer to exert pressure (approximately 0.15 kg) on the pelvic floor tissue. After 4 h, the Foley balloon was deflated and removed. Fourteen days post-vaginal distention, the experimental rats were categorized into four groups: normal control, pelvic floor dysfunction (PFD), PFD rats treated with exosomes from BMSCs and PFD rats treated with exosomes (Exo) and TNF-Exo. Specifically, exosomes (500 μL PBS) were administered via the caudal vein for seven consecutive days starting on Day 14. On Day 28, bladder baseline pressure, peak bladder pressure, leak point pressure (LPP) and void volume were measured to assess the severity of PFD.

2.5 Enzyme-linked immunosorbent assay (ELISA)

The quantification of rat IL-6, TNF-α and MMP2 involved the use of ELISA kits (Abbexa, Cambridge, UK) following the provided manufacturer's guidelines. In summary, the specific primary antibody captured IL-6, TNF-α and MMP2 in the samples, which were subsequently identified using a biotin-labelled secondary antibody. The assays were developed with avidin-peroxidase and its substrate, and the readings were taken at 450 nm using a microplate reader.

2.6 Western blot

The anterior vaginal walls or exosomes were subjected to lysis, and the resulting soluble supernatants (30 μg) were separated through 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis and subsequently transferred onto polyvinylidene fluoride (PVDF) membranes. Following blocking with 5% nonfat milk, these PVDF membranes underwent incubation with primary antibodies targeting Elastin, Collagen I, Collagen III and β-Actin (Santa Cruz Biotechnology Inc., Dallas, TX). Subsequently, peroxidase-conjugated secondary antibodies (Sigma-Aldrich) were applied, and the membranes were developed using Super Ecl Solution (GE Healthcare, Little Chalfont, Buckinghamshire, UK). β-Actin served as a normalization reference for relative expression, and NIH-Image J was utilized for analysis.

2.7 RT-qPCR

Total RNA was isolated from the anterior vaginal walls and BMSCs using TRIzol reagent (Invitrogen) following the manufacturer's instructions. Subsequently, cDNA was synthesized in a reverse transcription process using a High-Capacity cDNA Reverse Transcription Kit (Bio-Rad, Hercules, CA) and a One Step PrimeScript miRNA cDNA Synthesis Kit (Takara, Dalian, China). The resulting cDNA was then amplified using SYBR Green master mix (Roche, Penzberg, Upper Bavaria, Germany). The amplification protocol involved an initial step at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min on an ABI 7500 real-time PCR system (Applied Biosystems, Foster City, CA). Expression data were normalized to GAPDH expression and quantified using the comparative ΔCT method. The primer sequences are provided as following: MMP2: forward 5′- CCTGTCTCCTGCTCTGTAG−3′, reverse 5′-AGTAGCACCTGGGAGGGATA-3′; Elastin, forward 5′-CTTCCTGGTGGAGTTCCCGGTGGA-3′, reverse 5′-CCGATGCCACCAATACCACCGACA-3′; COL1A1, forward 5′-ATCAGCCCAAACCCCAAGGAGA-3′, reverse 5′-CGCAGGAAGGTCAGCTGGATAG-3′; COL3A1, forward 5′-TGATGGGATCCAATGAGGGAGA-3′, reverse 5′-GAGTCTCATGGCCTTGCGTGTTT-3′; IL-6, forward 5′-GACTGATGTTGACAGCCACTGC-3′, reverse 5′-AGCCACTCCTTCTGTGACTCTAACT-3′; TNF-α, forward 5′-CATGATCCGAGATGTGGAACTGGC-3′, reverse 5′-CTGGCTCAGCCACTCCAGC-3′; GAPDH: forward 5′-GACATGCCGCCTGGAGAAAC-3′, reverse 5′- GACATGCCGCCTGGAGAAAC-3′.

2.8 Statistical analysis

Statistical analyses were conducted utilizing SPSS software (SPSS Inc., Chicago, IL). The results are presented as the mean ± standard deviation (SD). Brown–Forsythe ANOVA test followed by Dunnett's T3 multiple comparisons test was employed for the comparison between groups, and statistical significance was determined at a p < 0.05.

3.1 IL-6, TNF-α and MMP2 were elevated in vaginal wall tissues from patients diagnosed with POP

To compare the levels of IL-6, TNF-α and MMP2 in vaginal wall tissues between patients diagnosed with pelvic organ prolapse and a control group, we gathered a total of 80 vaginal wall tissue samples. This comprised 40 samples from patients with POP and 40 samples from control patients without POP. The outcomes indicated a significant increase in IL-6 (Controls: 36.54 ± 9.84 pg/mg tissue; POP: 57.62 ± 15.01 pg/mg tissue; p < 0.001), TNF-α (Controls: 18.92 ± 5.58 pg/mg tissue; POP: 35.16 ± 8.35 pg/mg tissue; p < 0.001) and MMP2 levels (Controls: 17.31 ± 5.44 pg/mg tissue; POP: 26.15 ± 6.20 pg/mg tissue; p < 0.001) in the vaginal wall tissues of patients with POP when compared to the control group (Figure 1A–C). Moreover, the results of Pearson correlation analysis demonstrated a robust positive correlation between IL-6 level and TNF-α level (r = 0.57, p < 0.001) (Figure 1D), IL-6 level and MMP2 level (r = 0.48, p = 0.002) (Figure 1E), as well as TNF-α level and MMP2 level (r = 0.43, p = 0.005) (Figure 1F).

3.2 TNF-Exo successfully restored both void volume and bladder void pressure in PFD rats

To investigate the biological effects of Exo and TNF-Exo on rats with PFD, a rat PFD model was established. It is worth noting that TNF-Exo treatment did not induce any obvious side-effects in the normal healthy rats, including weight loss and abnormal behaviours (data not shown). As depicted in Figure 2A–C, the baseline pressure of the bladder showed no significant differences across all experimental groups (NC: 46.53 ± 3.74 cm water; Vehicle: 47.13 ± 4.09 cm water; Exo: 46.26 ± 4.02 cm water; TNF-Exo: 47.28 ± 4.14 cm water; all p > 0.05) (Figure 2A). In comparison with normal control rats, those with PFD displayed a notable decrease in void volume (NC: 1.19 ± 0.07 mL; Vehicle: 0.57 ± 0.08 mL; Exo: 0.78 ± 0.09 mL; TNF-Exo: 1.02 ± 0.11 mL; both p < 0.001 for Exo and TNF-Exo groups vs. Vehicle group) (Figure 2B) and bladder void pressure (NC: 91.35 ± 9.14 cm water; Vehicle: 46.85 ± 8.25 cm water; Exo: 63.84 ± 10.35 cm water; TNF-Exo: 82.35 ± 8.20 cm water; p = 0.0048 for Exo vs. Vehicle, and p < 0.001 for TNF-Exo vs. Vehicle) (Figure 2C), indicating the successful construction of the rat PFD model. Following daily treatment with exosomes for 7 days, both the Exo and TNF-Exo groups exhibited increased void volume and bladder void pressure. Notably, TNF-Exo treatment demonstrated more effective restoration of void volume and bladder void pressure compared with Exo treatment (p < 0.001 for void volume and p = 0.0021 for bladder void pressure).

3.3 TNF-Exo effectively improved peak bladder pressure and leak point pressure in PFD rats

PFD rats exhibited a noteworthy reduction in peak bladder pressure (NC: 132.80 ± 10.24 cm water; Vehicle: 69.45 ± 11.03 cm water; p < 0.001) and leak point pressure (NC: 85.98 ± 5.79 cm water; Vehicle: 46.47 ± 6.76 cm water; p < 0.001) compared with control rats (Figure 3A,B). As anticipated, the administration of Exo partially reinstated the diminished peak bladder pressure (Exo: 88.45 ± 8.97 cm water; p = 0.0033 vs. Vehicle) and leak point pressure (Exo: 60.88 ± 6.49 cm water; p < 0.001 vs. Vehicle) observed in PFD rats (Figure 3A,B). Moreover, the injection of TNF-Exo further reversed the decreased peak bladder pressure (TNF-Exo: 107.70 ± 14.43 cm water; p < 0.001 vs. Exo) and leak point pressure (TNF-Exo: 73.47 ± 7.00 cm water; p < 0.001 vs. Exo) (Figure 3A,B).

3.4 TNF-Exo enhanced the expression levels of Elastin, Collagen I and Collagen III in the anterior vaginal walls of PFD rats

The mRNA and protein levels of Elastin, Collagen I and Collagen III in the anterior vaginal walls of PFD rats were further assessed. Reduced mRNA levels of Elastin, Col1a1 and Col3a1 were observed in PFD rats when compared to NC group (p = 0.0026 for Elastin; p = 0.0044 for Col1a1; p = 0.0101 for Col3a1) (Figures 4A and 5C). Exo, especially TNF-Exo, treatment significantly enhanced their mRNA levels in PFD rats. Similarly, Exo treatment partially restored the expression of Elastin, Collagen I and Collagen III, while TNF-Exo treatment nearly completely restored their protein expression when compared to the NC group (Figures 4D and 5G).

3.5 TNF-Exo suppressed the expression levels of IL-6, TNF-α and MMP2 in the anterior vaginal walls of PFD rats

As anticipated, elevated levels of IL-6 (NC: 28.97 ± 5.34 pg/mg tissue; Vehicle: 72.26 ± 10.01 pg/mg tissue), TNF-α (NC: 19.83 ± 4.01 pg/mg tissue; Vehicle: 50.82 ± 5.85 pg/mg tissue) and MMP2 production (NC: 15.66 ± 3.02 pg/mg tissue; Vehicle: 53.26 ± 5.51 pg/mg tissue) were observed in the anterior vaginal walls of PFD rats (Figure 5A–C). Exo treatment partially restored the production of IL-6 (52.53 ± 7.39 pg/mg tissue), TNF-α (37.82 ± 6.03 pg/mg tissue) and MMP2 (37.65 ± 4.47 pg/mg tissue), whereas TNF-Exo treatment almost completely restored IL-6 (37.32 ± 6.63 pg/mg tissue), TNF-α (26.97 ± 5.34 pg/mg tissue) and MMP3 (24.32 ± 5.02 pg/mg tissue) production. Similar trends were found in mRNA expression of IL-6, TNF-α and MMP2 in these groups (Figure 5D–F). It was observed that TNF-Exo treatment could greatly enhance the mRNA expression of IL-6, TNF-α and MMP2 in PFD rats.