Patients

Thirty women with endometriosis and thirty women with tubal factor infertility were included in this study at Shanghai Ji Ai Genetics and IVF Institute, affiliated with Fudan University, from March 2018 to May 2019. All the women underwent laparoscopic surgical examination of the abdominal cavity and complete excision of endometriotic tissue. The disease was diagnosed clinically or by ultrasonography and verified by surgical findings and postoperative pathological examination. In addition, laparoscopic examination of the abdominal cavity excluded the presence of any other pelvic pathology that could potentially confound the data observed. There were 24 women with stage III and six women with stage IV disease (all women were diagnosed with tubal endometriosis via pathological biopsy). The women in the control group all underwent surgery for laparoscopic tubal sterilisation, and the absence of endometriosis was verified after a surgical examination of the abdominal cavity. Women who repeatedly had high baseline levels of serum follicle-stimulating hormone (FSH) (> 15–20 IU/l), a seriously deformed uterus, or any other active infections were excluded from this study.

Fig. 2

Relative expression levels of hsa-miR-145-5p in the endometriosis and control groups. Scatter plots depicting the relative expression levels of hsa-miR-145-5p between the two groups during the window of implantation (WOI), unpaired t-test. *P < 0.05; **P < 0.01. Mean ± SD shown by bars

Fig. 3

Characterization of the isolated extracellular vesicles (EVs) from human uterine luminal fluid. A Representative image of transmission electron microscopy (TEM) images showing morphology of isolated EVs purified from human uterine luminal fluid. B Nanoparticle tracking analysis by ZetaView PMX120 indicated that the mean diameter of the EVs was 93.6 nm. C Western blotting analysis suggested that the EVs from uterine luminal fluid expressed three positive markers (CD9, CD81, and CD63) but not one negative marker (calnexin) (Fig. 3C includes cropped blot images, and the original blot image can be obtained from Figure S2). D Crossing points (y-axis) for hsa-miR-145-5p in cells, EVs extracted from primary hEEC cultures, supernatants, and total supernatants without depletion of EVs. The box indicates the first, median, and third quartiles, and the error bars indicate the minimum and maximum values. hEECs, human endometrial epithelial cells

ULF sample preparation

The participants with regular menstrual cycles of 25–33 days were selected for collection of ULF samples. ULF samples were obtained from all the patients using a Frydman catheter (Prodimed; Neuilly-en-Thelle, France) during the patient’s implantation window, based on the date of the last menstruation. None of the women received hormonal treatment in the 3 months preceding the biopsy and ULF collection. Briefly, a speculum was inserted into the subject while in the lithotomy position. Subsequently, the cervix was cleansed, and the Frydman catheter was gently introduced to aspirate in 20–50 µL of endometrial secretion as described [5]. ULF samples were stored at − 80 °C until they were used for RNA extraction.

Fig. 4

Effects of miR-145-5p mimic-enriched EVs on in vitro development of mouse IVF embryos. Mouse IVF embryos co-incubated with miR-145-5p mimic-enriched EVs or scramble miRNA EVs on day 1.5 of culture. The continuing culture was conducted for 72 h, and the development rate of the blastocyst was calculated. The proportions of day-1.5 embryos that developed to the blastocyst stage in the untreated control (n = 205), scramble-EVs (n = 162), and miR-145-5p mimic groups (n = 180) were 46.31% ± 4.23%, 45.25% ± 3.79%, and 14.13% ± 1.79%, respectively. All the experiments were repeated at least three times. Chi-square test, *P < 0.05; **P < 0.01; ***P < 0.001. EVs, extracellular vesicles

Ovarian stimulation and human oocyte collection

This study involved 52 patients enrolled in the assisted reproduction programme at the Shanghai Ji Ai Genetics and IVF Institute, affiliated with Fudan University. In total, 106 MI oocytes were obtained from 52 consenting couples. And all 106 MI oocytes have matured. The patients were stimulated with GnRH agonists (Ferring Pharmaceuticals, Switzerland) and recombinant FSH (Gonal F, Merck-Serono, Geneva, Switzerland). Human chorionic gonadotropin (hCG; Profasi, Merck-Serono) was injected when there was at least one 18-mm follicle and three or more 16-mm follicles. Ultrasonography-directed oocyte retrieval was performed 36 h after hCG administration. After 2–4 h of incubation, the cumulus masses of the oocytes were removed with a sharp needle and treated with 0.1% hyaluronidase in Dulbecco’s phosphate-buffered saline (DPBS) (w/v) (Irvine Scientific, Santa Ana, CA, USA) in preparation for ICSI. Only MI oocytes without the first polar body (PB) were used in this study. MI oocytes were cultured in a fertilisation medium (Vitrolife, Sweden) supplemented with 10% HSA in an incubator at 37 °C in 6% CO2 in the air for 5–7 h until they became MII oocytes [11]. Oocytes were then fertilised using ICSI and incubated in the fertilisation medium. Normal fertilisation was assessed and confirmed by the presence of two pronuclei and a second PB at 16–18 h after insemination.

Fig. 5

Effects of miR-145-5p mimic-enriched EVs on the expression levels of Notch pathway components in mouse blastocyst-stage embryos. Mouse IVF embryos co-incubated with miR-145-5p mimic-enriched EVs or scramble miRNA EVs on day 1.5 of culture. The continuing culture was conducted for 72 h, and the mRNA expression profiles of 10 genes related to the Notch signalling pathway were measured by qRT-PCR. The histograms present four genes that encode proteins that are members or regulators of the Notch signalling pathway with markedly different expression levels (A) and six genes that encode proteins that are members or regulators of the Notch signalling pathway with nonsignificant differences in expression (B) between the miR-145-5p mimic (n = 173) and control (n = 119) groups. All the experiments were repeated at least three times. Unpaired t-test; *P < 0.05; **P < 0.01; ***P < 0.001. EVs, extracellular vesicles

Fig. 6

Effects of miR-145-5p mimic-enriched EVs on in vitro development of human IVF embryos. Human IVF embryos co-incubated with miR-145-5p mimic-enriched EVs or scramble miRNA EVs on day 2 of culture. The continuing culture was conducted for 96 h, and the development rate of the blastocyst was calculated. A Representative image of scramble miRNA EVs added to IVF embryos for 96 h. B Representative image of miR-145-5p mimic-enriched EVs added to IVF embryos for 96 h. C Proportions of embryos that developed into the blastocyst stage in the scramble-EVs (n = 45) and miR-145-5p mimic-EVs (n = 48) groups were 37.5% ± 3.95% and 18.75% ± 1.46%, respectively. All the experiments were repeated at least three times. Chi-square test, *P < 0.05; **P < 0.01; ***P < 0.001. EVs, extracellular vesicles

Fig. 7

Effects of miR-145-5p mimic-enriched EVs on the expression levels of NOTCH pathway components in human blastocyst-stage embryos. Human IVF embryos co-incubated with miR-145-5p mimic-enriched EVs or scramble miRNA EVs on day 2 of culture. The continuing culture was conducted for 96 h, and the mRNA expression profiles of 10 genes related to the NOTCH signalling pathway were measured by qRT-PCR. The histograms present three crucially differentially expressed genes that encode proteins that are members or regulators of the NOTCH signalling pathway (A) and seven genes that encode proteins that are members or regulators of the Notch signalling pathway with nonsignificant expression differences (B) between the miR-145-5p mimic (n = 48) and control (n = 45) groups. All the experiments were repeated at least three times. Unpaired t-test; *P < 0.05; **P < 0.01; ***P < 0.001. EVs, extracellular vesicles

Collection of mouse embryos

Female B6D2F1 mice, aged 6–8 weeks, were superovulated using 5 IU of pregnant mare serum gonadotropin (Ningbo Second Hormone Factory, Ningbo, China) followed by 5 IU of hCG after 48 h (Ningbo Second Hormone Factory). The vaginal plug was checked the day following mating. The day when a vaginal plug was observed was considered day 1 of pregnancy. On day 1.5 of pregnancy, the mice were euthanised by cervical dislocation, and embryos were flushed from the oviduct with phosphate buffered saline (PBS; Life Technologies, Grand Island, NY, USA) using a 30-gauge blunt needle (code procedure 2015/VSC/PEA/00048). The embryos were then flushed four times in M2 medium, incubated at 37 °C in 5% CO2 and 95% humidified air, and used for co-culture experiments and transcriptomic assays.

RNA isolation

RNA extraction was performed using a method described in an earlier study [12]. The miRNeasy Kit (QIAGEN, Hilden, Germany) was used to isolate and purify miRNAs according to the manufacturer’s protocol. Briefly, uterine fluid was gently centrifuged to remove cellular debris and blood. The supernatant was stored at 4 °C, and the mucus within the sample was retrieved and suspended in 1 mL of PBS (pH 7.6; Life Technologies). Subsequently, 500 µL of ULF supernatant from each patient was transferred to an Axygen™ centrifuge tube (Corning, Tewksbury, MA, USA) and thoroughly mixed with 700 µL of QIAzol Lysis Reagent (QIAGEN). After 5 min of incubation at 24 °C, 140 µL of chloroform was added to the mixture, and the mixture was vigorously vortexed. Subsequently, the RNA pellet was collected by centrifugation at 3865 ×g for 30 min at 4 °C. The aqueous phase was carefully transferred to a new tube, and a 1.5 volume of absolute ethyl alcohol was added. The RNA pellet was then placed in an RNA-binding column and washed twice. Finally, the pellet was dissolved in 30 µL of nuclease-free H2O.

miRNA analysis and profiling

The study design is presented in Fig. 1. Thirty ULF samples from patients with endometriosis were classified as the endometriosis group and 30 samples from control subjects were classified as the non-endometriosis group. Briefly, 30 ng of RNA was initially reverse-transcribed using Megaplex RT Primers Pools A and B and then preamplified with Megaplex Preamp Primers Pools A and B. Subsequently, 900 µL of the preamplified product was loaded on a TaqMan Array Human MicroRNA Card and run on an Applied Biosystems 7900HT thermocycler following the manufacturer’s protocol. The cards contained assays for 766 mature miRNAs in Sanger miRBase version 18.0. Subsequently, miRNA profiling was performed with TaqMan Array Human MicroRNA Cards A and B v3.0 (Applied Biosystems). The analysis was performed following a previous study [12]. Finally, detailed data analysis was performed using the Real-Time Statminer software package (Applied Biosystems).

MiRNA validation

To validate the miRNA arrays, we initially measured the expression levels of the candidate miRNAs in each uterine fluid sample in the two groups (30 samples from the endometriosis group and 30 samples from the control group) by qRT‒PCR with a TaqMan miRNA assay. Furthermore, to identify and verify differentially expressed miRNAs correlated with blastocyst development potential, we selected candidate miRNAs in the endometriosis group with Raw Ct (miRNA) < 30 to exclude miRNAs with low expression levels. These miRNAs were selected for subsequent verification analysis and had the highest relative expression quantities. Second, the expression levels were normalised based on an internal reference: U6 snRNA [13]. Finally, the relative expression levels were calculated as 2−ΔCt where ΔCT = Raw Ct (miRNA)-Raw Ct (U6).

Isolation of EVs from ULF

Briefly, ULF samples were diluted in PBS (Life Technologies), vigorously vortexed, and filtered using a 0.22-µm syringe filter (Pall). The filtrates were centrifuged at 300 × g for 10 min to remove whole cells. The supernatant was subjected to a second centrifugation at 2000 × g for 10 min to remove dead cells and then centrifuged again at 10,000 × g for 30 min to remove cell debris. The supernatants were refiltered with a 0.22-µm syringe filter (Pall) and ultracentrifuged at 120,000 × g for 70 min. The supernatants were discarded, and the EVs from ULF (ULF-EVs) pellets were gently washed once with 200 µL of PBS to remove residual extract buffer, resuspended in 20 µL of PBS, and stored at − 80 °C. The size and purity of the isolated ULF-EVs were determined using a nanoparticle tracking analyser (ZetaView PMX 120, Particle Metrix, Germany). Western blotting analysis of the expression of the EV-specific markers CD63 (Abcam, Cambridge, UK), CD81 (Abcam) and CD9 (Abcam) and the negative control marker calnexin (Abcam) was performed, as described [14].

hEEC cultures and EV isolation

Endometrial biopsy samples were obtained from healthy donors in the luteal phase who underwent endometrial biopsy (ages 23–39 years). The endometrial samples were minced into small pieces < 1 mm and then digested in trypsin (Difco, BD Biosciences, MD, USA). Human endometrial epithelial cells (hEECs) were grown from isolated endometrial glands and purified as previously described [15]. This cell type was cultured and grown to confluence in steroid-depleted medium: 75% DMEM (GIBCO, Grand Island, NY, USA) and 25% MCDB-105 (Sigma, St. Louis, MO, USA) containing antibiotics and supplemented with 10% charcoal-dextran-treated FBS (HyClone, Logan, UT) and 5 µg/mL insulin (Sigma). The homogeneity of cultures was determined by analysing the morphological characteristics and verified by immunocytochemical localisation of cytokeratin, vimentin, and CD68. When the cultures reached confluence, they were washed with DMEM (Sigma‒Aldrich) to remove FBS-contaminated EVs and cultured in DMEM (Sigma‒Aldrich). After 48 h, the conditioned medium was collected, and EVs were isolated, as described in the section above.

Western blotting analysis

Antibodies against human CD63, CD9, CD81 (System Biosciences), and calnexin (Enzo Life Sciences) were used for western blotting. First, proteins were extracted from the EVs secreted from primary hEEC cultures with radioimmunoprecipitation assay (RIPA) lysis buffer supplemented with protease and phosphatase inhibitors (Sigma‒Aldrich). The concentration of proteins in the extract was measured using a BCA Kit (Thermo Fisher Scientific, CA, USA). Subsequently, the protein extracts were separated by 10% SDS-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes (Merck Millipore, Germany). First, the membranes were incubated overnight with primary target antibodies diluted 1:1000 in 5% BSA. Second, the membranes were washed three times with 1% PBS-T and incubated with secondary anti-mouse and anti-rabbit antibodies (System Biosciences) at a concentration of 1:20,000. Third, specific proteins were measured using the SuperSignal West Femto Chemiluminescence kit (Thermo Fisher Scientific).

Transmission electron microscopy

Briefly, EVs in suspension, isolated from ULF, were mixed with equal volume of 4% Paraformaldehyde solution for fixation and absorbed by the discharged electron microscope grid. Then, the uranyl acetate was added to the copper grid to precipitate for 1 min, and the floating liquid was absorbed with filter paper. The sample was dried at room temperature for a few minutes before transmission electron microscopy (TEM) (Jeol Korea) imaging was performed at 100 kV.

Embryos co-cultured with miR-145-5p mimic-enriched EVsMouse embryos

After fertilisation, the early mouse embryos were successively washed with G-MOPSplus medium (Vitrolife) and G-1 medium (Vitrolife) three times and then cultured in humidified air with 6% CO2 at 37 °C for 48 h. The activated embryos were then randomly divided into two groups. hEECs were transfected with miR-145-5p mimic or pre-miR scramble using Lipofectamine 2000 (Thermo Fisher Scientific). After transfection, the transfection medium was replaced by fresh MEM supplemented with 1% penicillin/streptomycin, 1% l-glutamine, and 10% EV-depleted FBS (Thermo Fisher Scientific). After 48 h of culture, the spent medium was collected for EV extraction. miR-145-5p mimic-enriched EVs were extracted from the spent medium with the Total Exosome Isolation Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. The miR-145-5p mimic levels in the EVs were determined by reverse transcription qPCR (RT‒qPCR) (Thermo Fisher Scientific). Embryos can take up free and EV-associated miRNAs from uterine fluid [5, 10]. Therefore, to select the most suitable concentration of miR-145-5p mimic EVs, day-1.5 embryos were cultured for 72 h with 50, 100, and 200 µg/mL mimic hsa-miR-145-5p mimic-enriched EVs. To validate the success of the internalization process, a RT‒qPCR was performed in both miR-145-5p mimic-treated and control embryos. In this study, embryos were treated with EVs at a physiological dose of 100 µg/mL. Once the working concentration (100 µg/mL) was selected, the day-1.5 embryos (1–2 embryos per 50 µL droplet) were incubated continuously for 72 h, and their blastocyst rate was calculated.

Human embryos

The human embryos were also randomly divided into two groups. Day 2 human embryos (n = 9) were cultured for 96 h with 50, 100, and 200 µg/ml miR-145-5p mimic EVs to select the most suitable concentration of miR-145-5p mimic-enriched EVs. Once the working concentration (100 µg/mL) was selected, the embryos (1–2 embryos per 50 µL droplet) were continuously cultured for 96 h. Subsequently, embryonic development was evaluated at the blastocyst stages.

Expression levels of NOTCH signalling pathway genes in the miR-145-5p mimic co-cultured and control groups

We collected mouse and human embryos after 72–96 h of co-culture with miRNA-145 mimic-EVs (human embryos: n = 48; mouse embryos: n = 180) or NC mimic-EVs (human embryos: n = 45; mouse embryos: n = 98). The miRNeasy Micro Kit (QIAGEN) was used for the isolation and purification of RNA from the embryos according to the manufacturer’s instructions [16]. The expression levels of 10 target genes (NOTCH1, NOTCH2, NOTCH3, NOTCH4, DLL1, DLL4, JAGGED1, JAGGED2, HES1, and HES2) that encode proteins that are members or regulators of the NOTCH signalling pathway in human/mouse blastocyst-stage embryos were measured by qRT‒PCR. The measurements were then compared between the miR-145-5p mimic co-cultured and control groups. Before PCR, whole transcriptome amplification (TaKaRa, Dalian, China) was performed because the quantity of RNA was limited owing to the small number of embryos. We performed three biological triplicates for each sample.

Statistical analysis

All the experiments were repeated at least three times for each group. The data are presented as the mean ± SEM or mean ± SD. Student’s t-test and chi-square test was used to assess the differences. All the statistical analyses were performed using SPSS (version 19.0; SPSS Inc., Chicago, IL, USA). Differences were considered significant when P < 0.05.