Myometrial extracellular vesicles promoted endometrial mesenchymal stem/stromal cells to self-renewal via jag1-mediated notch signaling.

Human endometrium (proliferative phase: n = 15, secretory phase: n = 10) were collected from women with regular menstrual cycles (age: 41–52 years) who underwent abdominal surgery for benign non-endometrial pathologies (Additional file: Table S2). The menstrual cycle phase of the endometrium was determined by an experienced histopathologist, who evaluated the hematoxylin and eosin stained endometrial sections of each sample. The women recruited were not on hormonal treatment for at least 3 months before surgery and an informed written consent was signed by each patient prior to participation in the study. Ethical approval was obtained from the Institutional Review Board of The University of Hong Kong/Hospital Authority Hong Kong West Cluster and The Institutional Review Board of the University of Hong Kong-Shenzhen Hospital. Single-cell suspensions of endometrial stromal cells were obtained using a reported method [ 6 ]. Briefly, endometrial tissues were minced into small pieces and digested with PBS containing collagenase type III (0.3 mg/ml, Worthington Biochemical Corporation, Freehold, NJ, USA) and deoxyribonuclease type I (40 µg/ml, Worthington Biochemical Corporation) in a shaking water bath for 1 h at 37 °C. After two rounds of digestion, the dispersed cells were loaded onto Ficoll-Paque (GE Healthcare, Uppsala, Sweden) followed by incubation with anti-CD45 antibody coated Dynabeads (Invitrogen, Waltham, MA, USA) to eliminate the red blood cells and the leukocytes, respectively. Stromal cells were then negatively selected from the endometrial cells using microbeads coated with antibody against the epithelial cell marker CD326 (EpCAM) (Miltenyi Biotech, Bergisch Gladbach, Germany). The purified stromal cells were plated onto 100 mm dishes coated with fibronectin (1 mg/ml, Invitrogen) and cultured in growth medium (GM) containing 10% FBS (Invitrogen), 1% L-glutamine (Invitrogen) and 1% penicillin–streptomycin (Invitrogen) in DMEM/F-12 (Sigma-Aldrich, St Louis, MA, USA) in a humidified carbon dioxide incubator at 37 °C for 7–14 days. The medium was refreshed every 7 days until the cells reached 90% confluence. Isolation of eMSC (CD140b + CD146 + cells) was performed using two sequential beadings with magnetic beads [ 6 ]. First, the stromal cells were incubated with phycoerythrin (PE)-conjugated anti-CD140b antibody at 4 °C for 45 min followed by another incubation with anti-mouse IgG1 magnetic microbeads (Miltenyi Biotech) at 4 °C for 15 min. The obtained cell suspension was applied to MS columns (Miltenyi Biotech) with a magnetic field to isolate the CD140b + cells. The selected CD140b + cells were expanded in GM for 7 to 10 days to degrade the bound microbeads before they were trypsinized and incubated with the anti-CD146 microbeads (Miltenyi Biotech) at 4 °C for 15 min. The CD140b + CD146 + cells were collected for subsequent experiments. Stromal cells at passage 1–3 were used in this study. The identity of the stromal cells was confirmed by expression of vimentin but not epithelial marker cytokeratin immunoreactivities (Additional file Fig S1 A). Fig. 1 Myometrial EV can stimulate the functions of eMSC. A Relative proportion of CD140b + CD146 + cells after indirect coculture with different preparation of myometrial cells by flow cytometry ( n = 6). B Relative clonogenic activity of eMSC after indirect coculture with different preparation of myometrial cells ( n = 6). C Relative proportion of CD140b + CD146 + cells after culture with myometrial CCM (Myo-CCM) or EV depleted myometrial CCM (Myo-DCM) by flow cytometry ( n = 7). D Relative clonogenic activity of eMSC after culture with myometrial CCM (Myo-CCM) or EV depleted myometrial CCM (Myo-DCM) ( n = 9). Results are presented as mean ± SD; * P  <.05; *** P  <.001. Abbreviations: eMSC, endometrial mesenchymal stem/stromal cells; EV, extracellular vesicles; Myo, myometrial cells; siRAB27A, siRNA to R AB 27 A ; siCtrl, scrambled control siRNA; CCM, concentrated conditioned medium Myometrial EV can stimulate the functions of eMSC. A Relative proportion of CD140b + CD146 + cells after indirect coculture with different preparation of myometrial cells by flow cytometry ( n = 6). B Relative clonogenic activity of eMSC after indirect coculture with different preparation of myometrial cells ( n = 6). C Relative proportion of CD140b + CD146 + cells after culture with myometrial CCM (Myo-CCM) or EV depleted myometrial CCM (Myo-DCM) by flow cytometry ( n = 7). D Relative clonogenic activity of eMSC after culture with myometrial CCM (Myo-CCM) or EV depleted myometrial CCM (Myo-DCM) ( n = 9). Results are presented as mean ± SD; * P  <.05; *** P  <.001. Abbreviations: eMSC, endometrial mesenchymal stem/stromal cells; EV, extracellular vesicles; Myo, myometrial cells; siRAB27A, siRNA to R AB 27 A ; siCtrl, scrambled control siRNA; CCM, concentrated conditioned medium Myometrial samples (proliferative phase n = 10, secretory phase n = 7) were collected from women aged 43–48 years old, who underwent abdominal hysterectomy for benign non-endometrial pathologies (Additional file Table S3). Isolation of myometrial cells was carried out as described [ 6 ]. Briefly, the myometrial tissue was digested in 5 ml PBS containing collagenase type III (300 μg/ml, Worthington) and deoxyribonuclease type I (40 μg/ml, Worthington) at 37 °C. After 3 h of digestion, single cell suspensions were filtered through 100 μm sieves (BD Bioscience, San Jose, CA, USA) and seeded onto 100 mm dishes with GM in a humidified carbon dioxide incubator at 37 °C. The myometrial cells were trypsinized and passaged when they reached ~ 80% confluence. Myometrial cells at passage 2 to 6 were used in this study. The identity of the myometrial cells was confirmed by expression of smooth muscle marker αSMA [ 6 ]. Myometrial cells were treated with mitomycin C (0.5 mg/ml, Sigma-Aldrich) for 3–4 h at 37 °C, washed and harvested. For coculture, freshly isolated eMSC were seeded onto 6-well plates (300 cells/well) coated with fibronectin, and mitomycin treated myometrial cells were seeded into transwell inserts (EMD Millipore, Billerica, MA, USA). The ratio of eMSC to myometrial cells is 1: 90, based on our previous study [ 21 ]. All the conditions were performed in triplicates and monoculture served as control. Myometrial cells (~ 1 × 10 6 cells) were seeded in T-75 flasks for 3 days in GM, washed with PBS and cultured with 5 ml of DMEM-F12 medium for 24 h. The myometrial conditioned medium was collected and filtered through 0.22 μm filters. Half of each conditioned medium preparation was used to isolate EV and generate the corresponding amount of EV depleted conditioned medium (described below). While the other half of the conditioned medium was concentrated using Amicon ultra-15 centrifugal filter devices (EMD Millipore) with a molecular weight cut-off of 10 kDa. The amount of proteins from 5 ml of myometrial CM was considered as one unit of myometrial CCM proteins and 1/6-unit were added into the GM for eMSC culture [ 6 ]. EV were isolated by differential centrifugation. In brief, myometrial or stromal cells were cultured in DMEM/F-12 medium for 24 h. The supernatants were collected, centrifuged at 2000 g for 30 min to remove dead cells and at 16,000 g for 30 min to eliminate cell debris, followed by filtering through a 0.22 μm filter. EV were pelleted by ultracentrifugation at 120,000 × g for 70 min at 4ºC (Beckman Coulter Optima L‐100 XP, Beckman Coulter, CA, USA). The supernatants were collected as myometrial EV depleted medium (DM) and concentrated to reach the same volume of the corresponding CCM. EV were then washed with PBS and centrifuged as described before. The resulting pellets were suspended in PBS and its protein content was quantified by the BCA Protein Assay Kit (Thermo Scientific Inc, Rockford, USA). The isolated EV were stored at −80 °C for further use. For functional assays, EV derived from myometrial or stromal cells were used at a concentration of 10 μg/ml. The morphology of EV was observed under transmission electron microscopy (TEM). Freshly isolated EV were pipetted onto a carbon film and incubated for 5 min. The sample was negatively stained with 2% uranyl acetate and air-dried for examination. Images were captured using a Philips CM100 transmission electron microscopy at the Electron Microscope Unit, The University of Hong Kong. The diameter of EV was measured by a Nanoparticle Tracking Analyzer (NTA, Nanosight NS300, Malvern Instruments). The EV pellets were diluted with PBS and injected into the sample chamber of the analyzer machine. A total of 11 positions were recorded and analyzed for NTA measurements. The expression of exosomal surface markers were detected by Western blot (described below) using the following antibodies: mouse anti-CD63 (1:1000, Abcam, Cambridge, UK), mouse anti-CD81 (1:1000, Abcam), mouse anti-ALIX (1:1000, Abcam) and mouse anti-GM130 (1:1000, Abcam) antibodies. Freshly isolated EV were labeled with PKH67 (Sigma-Aldrich) according to the manufacturer’s instruction. Briefly, 50 μg EV were resuspended in 0.5 ml 2 × Diluent C containing PKH67, incubated for 3 min with periodic mixing and addition of 1% BSA to stop the incubation. The PKH67 labeled EV were ultracentrifuge at 120,000 g for 70 min to remove the non-bound dye and added to the GM of eMSC at 37 °C for 18 h. Internalization of EV by eMSC was analyzed by a fluorescence microscope or flow cytometer. Myometrial EV samples were processed and analyzed at the Proteomics and Metabolomics Core Facility, Centre for PanorOmic Sciences, The University of Hong Kong. EV samples were resolved using SDS-PAGE and visualized with Coomassie Brilliant Blue staining. Each gel lane was excised into individual slices, which were then subjected to in-gel digestion. Proteins within the gel slices were reduced with 10 mM tris(2-carboxyethyl) phosphine (TCEP) and alkylated with 55 mM 2-chloroacetamide (CAA). Digestion was performed using trypsin (1 µg/µl) at 37 °C overnight. The resulting tryptic peptides were extracted sequentially using 50% acetonitrile (ACN) with 5% formic acid (FA), followed by 100% ACN. The extracted peptides were pooled, dried, and desalted using C18 StageTips before LC–MS/MS analysis. Raw mass spectrometry data were processed using MaxQuant 1.6.14.0. Raw data was searched against the Human Swissprot FASTA database containing 20,361 entries, using settings as below: oxidized methionine (M), acetylation (Protein N-term) were selected as dynamic modifications, and carbamidomethyl (C) as fixed modifications with a minimum peptide length of 7 amino acids was enabled. Confident proteins were identified using a target-decoy approach with a reversed database, strict false-discovery rate 1% at peptide and peptide spectrum matches (PSMs) level; minimum ≥ 1 unique peptide, ≥ 2 peptide spectral matches (PSMs). The functional enrichment analysis of proteins was conducted by the gene set analysis toolkit (WebGestaltR v0.4.6) [ 21 ]. To block Notch signal, eMSC were treated with N- [N-(3,5-difluorophenacetyl- l -alanyl)] -(S)-phenylglycine t-butyl ester (1.25 µM, DAPT, R&D Systems Minneapolis, MN, USA) and cultured in GM for 14 days. The culture medium was changed every 3 days. After 14 days, the eMSC were harvested for further analysis. To block the secretion of EV, myometrial cells were transfected with 10 pmol of siRNA directed against RAB27 A (ID s11693 and s11685; Ambion, Grand Island, NY, USA) or random siRNA with scrambled sequence (Ambion) using Lipofectamine RNAiMax transfection reagent (Invitrogen). The cells were harvested at 24 h after transfection and seeded into transwell inserts (EMD Millipore, Billerica, MA, USA) for coculture. The knockdown efficiency was measured by western blotting (Additional file Fig S1B) and the concentration of total EV protein was detected by the BCA Protein Assay Kit (Additional file Fig S1 C). Endometrial MSC (2 × 10 4 /well) were seeded onto 24-well plates coated with fibronectin. After overnight incubation, the cells were transfected with 10 pmol of siRNA directed against NOTCH1 (ID s9633 and s9635; Ambion) or random siRNA with scrambled sequence (Ambion) as described above. The next day after transfection, the cells were washed and cultured with myometrial EV for functional assays. Myometrial cells were cultured in T75 flasks until 70–80% confluent. The cells were then transfected with 10 pmol of siRNA against JAG1 (ID s1174 and s1175; Ambion) or random siRNA (Ambion). After transfection for 24 h, the cells were washed and cultured in DMEM/F12 for 24 h. The supernatant was collected from the JAG1 siRNA or control siRNA transfected myometrial cells to isolate si JAG1 EV and siCtrl EV, respectively. The knockdown efficiency of JAG1 in myometrial cell derived EV was measured by western blotting (Additional file Fig S1D). Assessment of eMSC markers, CD140b and CD146 on endometrial stromal cells was carried out using multi-color flow cytometry [ 6 ]. The cells were labeled with PE-conjugated anti-CD140b antibody (2.5 μg/ml, PR7212 clone, Mouse IgG1, R&D Systems) and FITC-conjugated anti-CD146 antibodies (5 μg/ml, OJ79c clone, mouse IgG1; ThermoFisher Scientific) in dark for 45 min at 4 °C. Isotype matched controls were included for each antibody. Following the final washing step, the labeled cells were analyzed by a CytoFlex™ flow cytometer (Beckman Coulter, CA, USA). The cells were gated according to the forward and the side scatter profiles. FlowJo Software (Tree Star Inc) was applied to analyze the data. Endometrial MSCs were seeded onto 6-well plates (300 cells/well) coated with fibronectin and cultured in different conditions for 14 days. Medium was changed every 4 days. Cloning efficiency was evaluated by the number of colonies forming units (CFUs) divided by the number of cells seeded multiplied by 100. Cultured eMSC were lysed in cell lysis buffer (Ambion) containing protease inhibitors. The denatured protein samples were subjected to 10% SDS-PAGE and transferred to 0.45 μm polyvinylidene difluoride membranes (Immobilon™-P, Milllipore). The membranes were blocked by 5% skim milk for 1 h at room temperature and incubated with specific primary antibodies (Additional file Table S4) overnight at 4 °C. The following day, the membranes were incubated with horseradish peroxidase conjugated secondary antibodies (Additional file Table S5) for 1 h at room temperature and visualized using the Western Bright ECL Kit (Advansta, CA, USA). The intensities of the western blot bands were evaluated by the Quantity One software and normalized to β-actin. Mice were provided by the Center of Comparative Medicine Research at The University of Hong Kong. All experimental procedures were approved by the Committee on Use of Live Animals in Teaching and Research, The University of Hong Kong, Hong Kong. The mice were kept under standard conditions with a light/dark cycle of 12 h/12 h and free access to food and water. The animal experimental setup is shown in Additional file Fig S5 A. Female NOD-SCID mice (6–8 weeks old) were used to establish the endometrial injury model by electrocoagulation [ 14 ]. Mice at diestrus were operated after anesthesia and a vertical incision in the abdominal wall was made to locate the left uterine horn. A small incision was made in the upper region of the uterine horn to insert a monopolar electrode into the lumen. Monopolar electrocoagulation was performed with 50 W of power. While electrifying, the electrode pen was gradually moved out toward the incision of the uterine horn at constant pace and was repeated to ensure the entire uterine horn was damaged. The whole process lasted for 3–4 s. The uterine horn was returned into the abdominal cavity and the abdominal wall and skin layer were sutured and disinfected. The right uterine horn of each mouse was untreated and served as self-control. The female NOD-SCID mice were randomly allocated into 4 groups: 1) endometrial injury group—uterine horn injected with 20 μl of PBS; 2) eMSC group – transplantation of 5 × 10 5 eMSC resuspended in 20 μl of PBS; 3) myo-EV group transplantation of 20 μg myo-EV resuspended in 20 μl of PBS and 4) eMSC/myo-EV group—combined transplantation of 5 × 10 5 eMSC and 20 μg myo-EV resuspended in 20 μl PBS. All the treatments were delivered via intrauterine injection immediately after injury was established. The uteri were collected at postoperative day 7. To trace eMSC/myo-EV in mouse endometrium, eMSC were labeled with CM-Dil dye (Thermo Fisher Scientific) [ 14 ]. Briefly, freshly isolated eMSC were incubated with the CM-Dil dye for 5 min at 37 °C and then a further 15 min at 4 °C. After labeling, the cells were washed and mixed with PKH67 labeled myo-EV and transplanted in the mouse uterus. The uterine horns were harvested at post-intrauterine transplantation day 7, frozen in optimal cutting temperature (OCT) compound (Sakura Finetek) in liquid nitrogen and stored at −80 °C. Frozen sections at 5 µm thickness were stained with DAPI (Thermo Scientific) for 1 min and examined under a Carl Zeiss LSM 800 inverted confocal microscope with the use of the Zeiss LSM ZEN 2019 software (Carl Zeiss) at The Imaging and Flow Cytometry Core, CPOS, The University of Hong Kong. Paraffin sections (5 µm) were stained with hematoxylin (Sigma-Aldrich) and eosin (Sigma-Aldrich) using standard protocols. Average endometrial thickness was measured from transverse section of the uterus—the vertical distance from the luminal epithelium to the endometrial–myometrial interface using the Image-Pro Plus software (version 6.0, Media Cybernetics) from 10 serial sections of the same animal [ 14 ]. The relative endometrial thickness and number of glands were determined as the injured uterine horn (left side) divided by its untreated intact uterine horn (right side). Paraffin sections were dewaxed and underwent antigen retrieval, followed by denaturation with 0.1 N HCl for 45 min. The sections were then quenched with 3% hydrogen peroxide for 10 min, blocked with 5% BSA/PBS for 1 h, and incubated with rabbit anti-Ki67 (1:500, Abcam) or rabbit anti-cleaved caspase 3 (1:500, CST) antibodies at 4 °C overnight. On the next day, the sections were incubated with biotinylated goat anti-rabbit secondary antibodies (1:200 dilution, Dako, Hamburg, Germany) for 1 h and then with the Vectastain ABC reagent (Vector Laboratories) for 30 min. Positive staining was revealed by DAB solution (Dako) under a Zeiss Axioskop II microscope (Carl Zeiss). Images were captured using a Photometrics CoolSNAP digital camera (Roper Scientific, Trenton, NJ, USA). The relative expression of Ki67 and cleaved caspase 3 were determined as the injured uterine horn (left side) divided by its untreated intact uterine horn (right side). Quantitative real-time polymerase chain reaction (qPCR) with Taqman probes were used (Additional file Table S6). The total RNA was isolated using the Absolutely RNA microprep kit (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer’s instructions. The concentration of total RNA was quantified by spectrophotometry. RNA was reversed transcribed to cDNA by the PrimeScript DNA Reverse Transcription kit (Takara Bio Inc., San Jose, CA, USA). PCR was conducted by a 7500 Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). The mixtures were incubated at 50 °C for 2 min and 95 °C for 10 min, followed by 40 cycles of 15 s at 95 °C and 1 min at 60 °C. Gene expression was measured in triplicate and presented as relative gene expression using the 2 −ΔΔCt method and normalized to 18S as internal control. Data were analyzed using the GraphPad PRISM software (version 8.00; GraphPad Software Inc., San Diego, CA, USA). Distribution normality was tested using the Shapiro–Wilk test. Differences between two groups were analyzed using the Mann–Whitney U test for non-parametric data and the two-tailed unpaired Student’s t test for parametric data. Kruskal–Wallis test followed by Dunn’s post-test were used for multiple group comparison. Data are represented as mean ± SD. A difference with P-value of < 0.05 is considered as significant.

To investigate whether the myometrial cell derived EV play a role in the stimulatory effect on eMSC functions, the RAB27 A gene was silenced using siRNA to inhibit the EV secretion from myometrial cells (Additional file Fig S1B, n = 4, P < 0.01). Knockdown of RAB27 A significantly reduced the total protein concentration of EV derived from the myometrial cells (Additional file Fig S1 C, n = 3, P < 0.001). Consistent with the previous published results, coculture with myometrial cells remarkably enhanced the relative proportion of cells coexpressing CD140b and CD146 (Fig.  1 A, n = 6, P < 0.05) and colony formation (Fig.  1 B, n = 6, P < 0.001). A decline in the eMSC phenotypic expression (Fig.  1 A, n = 6, P < 0.05) and clonogenicity (Fig.  1 B, n = 6, P < 0.05) was observed when eMSC were cocultured with myometrial cells with knockdown of RAB27 A . To confirm the role of myometrial cell-derived EV in the indirect coculture system, eMSC were treated with myometrial cell concentrated conditioned medium (Myo-CCM) or EV depleted Myo-CCM (Myo-DCM). As expected, Myo-CCM notably increased the proportion of CD140b + CD146 + expressing cells (Fig.  1 C, n = 7, P  < 0.05) and colony formation of eMSC (Fig.  1 D, n = 9, P  < 0.05). The phenotypic expression of CD140b and CD146 in eMSC slightly decreased in the Myo-DCM when EV were depleted from the Myo-CCM, though the decrease did not reach a statistically significant difference (Fig.  1 C, n = 7, P  = 0.07). The Myo-DCM treatment significantly reversed the effect of Myo-CCM on clonogenicity of eMSC (Fig.  1 D, n = 9, P  < 0.05). Taken together, these results suggest that the EV were involved in the stimulatory effect of the myometrial cells on eMSC activities. Next, the EV from the conditioned medium of myometrial cells were isolated by ultracentrifugation. Transmission electron microscopy revealed that the EV had a round or cup-shaped morphology (Fig.  2 A). Nanoparticle tracking analysis of the EV showed a narrow size distribution with a mean diameter around 114 nm (Fig.  2 B). The myometrial EV contained exosomal proteins including Alix, CD63 and CD81 and were negative for the cis-Golgi protein GM130 (Fig.  2 C). Fig. 2 Myometrial derived EV increased the phenotypic expression and clonogenic activity of eMSC. A Morphology of myometrial derived EV under transmission electron microscope (red arrow). Scale Bar = 100 nm. B Diameter analysis of myometrial derived EV by nanoparticle tracking microscope (n = 3). C Representative western blotting images of exosomal protein CD63, CD81, Alix and GM130 in myometrial derived EV (n = 3). D Representative immunofluorescence images show internalization of myometrial derived EV by eMSC (n = 3), insets are enlarged figures of the green signals expressed by the EV. Scale Bar = 50 µm. E Representative images and relative proportion of CD140b + CD146 + cells after treatment with myometrial derived EV by flow cytometry (n = 5). F Representative figures showing the total colony formation and relative clonogenic activity of eMSC after treatment with myometrial derived EV (n = 4). Results are presented as mean ± SD; ** P  <.01. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells Myometrial derived EV increased the phenotypic expression and clonogenic activity of eMSC. A Morphology of myometrial derived EV under transmission electron microscope (red arrow). Scale Bar = 100 nm. B Diameter analysis of myometrial derived EV by nanoparticle tracking microscope (n = 3). C Representative western blotting images of exosomal protein CD63, CD81, Alix and GM130 in myometrial derived EV (n = 3). D Representative immunofluorescence images show internalization of myometrial derived EV by eMSC (n = 3), insets are enlarged figures of the green signals expressed by the EV. Scale Bar = 50 µm. E Representative images and relative proportion of CD140b + CD146 + cells after treatment with myometrial derived EV by flow cytometry (n = 5). F Representative figures showing the total colony formation and relative clonogenic activity of eMSC after treatment with myometrial derived EV (n = 4). Results are presented as mean ± SD; ** P  <.01. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells To evaluate the proteins in the myometrial EV, we performed LC–MS/MS analysis. A total of 1,008 exosomal proteins were detected (Additional file Fig S2 A). These proteins were predicted to be involved in cellular components, biological processes, and molecular functions. Using the gene set analysis toolkit to generate a network of KEGG terms, the enriched pathways associated with myometrial EV proteins were identified (Additional file Fig S2B and Fig S2 C). These proteins were implicated in multiple functions, including focal adhesion, ECM-receptor interaction, gap junctions, the PI3 K-Akt signaling pathway, estrogen signaling pathway, and protein digestion and absorption (Additional file Fig S2 C). To determine whether eMSC could uptake myometrial cell derived EV, a fluorescent dye PKH67 was used to label the EV. After incubation with the labeled EV for 18 h, a strong green fluorescence signal was observed around the nucleus of eMSC (Fig.  2 D), demonstrating internalization of the myometrial EV by eMSC. Next, the role of myometrial EV on eMSC activities was determined by assessing the phenotypic expression of eMSC after treatment with myometrial EV. Figure  2 E showed that treatment with myometrial EV significantly increased the proportion of CD140b + CD146 + cells (n = 5, P  < 0.01). The colony formation of eMSC was also enhanced compared to the control group (Fig.  2 F, n = 4, P  < 0.01). The eMSC are supported by the endometrial stroma. Therefore, the role of stromal cell derived EV on the maintenance of eMSC was also evaluated. The morphology and size of the stromal EV were determined by transmission electron microscope (Additional file Fig S3 A) and nanoparticle transmission analysis (Additional file Fig S3B). These stromal EV displayed a round, cup-shaped morphological feature with an average diameter of 129 nm, expressed Alix, CD63 and CD81 and without GM-130 (Additional file Fig S3 C). Interestingly, when the stromal EV were cultured with eMSC, they did not change the phenotypic expression (Additional file Fig S3D, n = 5, P  = 0.46) or the clonogenic activity of eMSC (Additional file Fig S3E, n = 5, P  = 0.49) when compared to the control group. To determine whether the stimulatory effect of myometrial cells on eMSC is associated with Notch signaling, the indirect coculture system was used. Coculture with myometrial cells significantly increased the proportion of CD140b + CD146 + expressing cells (Additional file Fig S4 A, n = 6, P  < 0.05) and clonogenicity (Additional file Fig S4B, n = 5, P  < 0.001) of eMSC when compared with monoculture. Addition of DAPT to inhibit the γ-secretase complex nullified these effects (Additional file Fig S4 A and Fig S4B, n = 5 - 6, P  < 0.01). Western blotting confirmed that the coculture with myometrial cells enhanced the protein expression of NICD, HEY-2 and HES-1 (Additional file Fig S4 C, n = 4 -5, P  < 0.05) in eMSC, highlighting the involvement of Notch signaling in the observation. The expression of these Notch target proteins reduced after supplementation of DAPT into the coculture system (Additional file Fig S4 C, n = 4 - 5, P  = 0.09 for NICD, P  < 0.05 for HEY-2 and P  < 0.01 for HES-1). The activity of Notch signaling in eMSC after treatment with Myo-CCM was also confirmed. The presence of Myo-CCM remarkably increased the expression of NICD (n = 7, P  < 0.05), HEY-2 (n = 5, P  < 0.05) and HES-1 (n = 7, P  < 0.05) in eMSC when compared with the control group (Additional file Fig S4D). The use of Myo-DCM failed to show a positive effect (Fig S4D, n = 5- 7, P  < 0.05 for NICD and HEY-2, P  = 0.11 for HES-1). Overall, these results suggest that the Notch signaling is involved in the positive regulatory effect of myometrial cells on eMSC activities. LC–MS/MS analysis detected Notch ligand JAG1 in the myometrial EV (Additional file Table S1). The presence of Notch ligands in myometrial cells and their derived EV were confirmed by western blotting. As shown in Fig.  3 A, JAG1 and DLL4 were expressed in the myometrial cells, whereas only JAG1 was detected in their EV. Treatment with myometrial EV notably increased the expression of Notch related proteins, NICD, HEY-2 and HES-1 (Fig.  3 B, n = 5, P  < 0.05 for NICD and HES-1, P  < 0.01 for HEY-2) in eMSC. Stromal EV also contained JAG1 (Fig.  3 C). However, there was no change in the activity of Notch signaling in eMSC after culture with the stromal EV (Fig.  3 D, n = 5–7). Fig. 3 Cell type specific effect of EV on activation of Notch signaling in eMSC. A Representative western blotting images of Notch ligands protein expression in myometrial EV (n = 3). B Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with myometrial EV (n = 5). C Representative western blotting images of Notch ligands protein expression in stromal cells derived EV (n = 3). D Representative western blotting images and quantitative analysis of Notch signals protein expression in eMSC after culture with stromal EV (n = 5 - 7). E Representative flow cytometry images and internalization percentage of stromal EV or myometrial EV by eMSC (n = 3). F Representative western blotting images and quantitative analysis of JAG1 and ADAM17 expression in stromal cells and myometrial cells (n = 7 - 8). G Representative western blotting images and quantitative analysis of JAG1 expression in stromal EV or myometrial EV (n = 5). Results are presented as mean ± SD; * P  <.05; ** P  <.01 l; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; Stro, stromal cells Cell type specific effect of EV on activation of Notch signaling in eMSC. A Representative western blotting images of Notch ligands protein expression in myometrial EV (n = 3). B Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with myometrial EV (n = 5). C Representative western blotting images of Notch ligands protein expression in stromal cells derived EV (n = 3). D Representative western blotting images and quantitative analysis of Notch signals protein expression in eMSC after culture with stromal EV (n = 5 - 7). E Representative flow cytometry images and internalization percentage of stromal EV or myometrial EV by eMSC (n = 3). F Representative western blotting images and quantitative analysis of JAG1 and ADAM17 expression in stromal cells and myometrial cells (n = 7 - 8). G Representative western blotting images and quantitative analysis of JAG1 expression in stromal EV or myometrial EV (n = 5). Results are presented as mean ± SD; * P  <.05; ** P  <.01 l; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; Stro, stromal cells Next, internalization of EV from the two cell types by eMSC was assessed. Using flow cytometry, it was revealed that eMSC took up a higher percentage of myometrial EV than the stromal EV (Fig. 3 E, n = 3, P  < 0.05). Western blotting results demonstrated that the myometrial cells expressed more than ~ threefold of JAG1 protein than the stromal cells (Fig. 3 F, n = 7, P  < 0.001). The protease ADAM17 mediates JAG1 shedding from the cell membrane of endothelial cells [ 22 ]. Therefore, we proposed ADAM17 as a candidate protease that cleaved JAG1 from endometrial niche cells to trigger the extracellular release of JAG1. Western blotting confirmed that the myometrial cells expressed more ADAM17 than the stromal cells (Fig. 3 F, n = 8, P  < 0.01). Consequently, the level of JAG1 in the myometrial EV was significantly higher than that in the stromal EV (Fig. 3 G, n = 5, P  < 0.05). These results demonstrate the potential cleavage specificity of JAG1 in endometrial niche cells. To confirm whether the enhanced effect of myometrial EV on eMSC was Notch dependent, DAPT was used to suppress the Notch signaling. As expected, DAPT treatment abolished the positive regulating effect of myometrial EV on the phenotypic expression (Fig.  4 A, n = 6, P  < 0.01) and clonogenicity (Fig.  4 B, n = 4, P  < 0.01) of eMSC. The expression of Notch related proteins was reduced in eMSC after treatment with myometrial EV in the presence of DAPT (Fig.  4 C, n = 4 - 6, P  < 0.05 for NICD and HES-1, P < 0.01 for HEY-2). Fig. 4 Myometrial EV require Notch signaling to increase the self-renewal and clonogenic activity of eMSC. A Relative proportion of CD140b + CD146 + cells after culture with myometrial EV in the presence of Notch inhibitor DAPT (n = 6). B Relative clonogenic activity of eMSC after culture with myometrial EV in the presence of Notch inhibitor (n = 4). C Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with myometrial EV in the presence of Notch inhibitor (n = 4 - 6). D Relative proportion of CD140b + CD146 + cells when transfected with siCtrl or siNOTCH1 and cultured with myometrial EV (n = 5). E Representative western blotting images and quantitative analysis of Notch related proteins in eMSC transfected with siCtrl or siNOTCH1 in the presence of myometrial EV (n = 4 - 6). Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; siNOTCH1, siRNA to N OTCH1 ; siCtrl, scrambled control siRNA Myometrial EV require Notch signaling to increase the self-renewal and clonogenic activity of eMSC. A Relative proportion of CD140b + CD146 + cells after culture with myometrial EV in the presence of Notch inhibitor DAPT (n = 6). B Relative clonogenic activity of eMSC after culture with myometrial EV in the presence of Notch inhibitor (n = 4). C Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with myometrial EV in the presence of Notch inhibitor (n = 4 - 6). D Relative proportion of CD140b + CD146 + cells when transfected with siCtrl or siNOTCH1 and cultured with myometrial EV (n = 5). E Representative western blotting images and quantitative analysis of Notch related proteins in eMSC transfected with siCtrl or siNOTCH1 in the presence of myometrial EV (n = 4 - 6). Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; siNOTCH1, siRNA to N OTCH1 ; siCtrl, scrambled control siRNA The role of the Notch receptor was investigated since NOTCH1 was expressed in eMSC [ 17 ]. Knockdown of NOTCH1 on eMSC nullified the stimulatory effect of myometrial EV (Fig. 4 D). An inhibitory effect was observed in the percentage of CD140b + CD146 + cells (Fig. 4 D, n = 5, P  < 0.01) and in the expression of various Notch target proteins: NICD (Fig. 4 E, n = 6, P  < 0.01), HEY-2 (Fig. 4 E, n = 4, P  < 0.001) and HES-1 (Fig. 4 E, n = 5, P  < 0.01) after transfection of NOTCH1 siRNA into eMSC. To evaluate the functional activity of JAG1 in EV, the expression of JAG1 in myometrial cells was silenced using siRNA and the EV were collected from the conditioned medium. Indeed, the expression of JAG1 in the myometrial EV was reduced after siRNA transfection (Additional file Fig S1D, n = 3, P  < 0.001). The increased phenotypic expression of eMSC induced by treatment with myometrial EV was abrogated when JAG1 was knockdown in the myometrial EV (Fig.  5 A, n = 6, P  < 0.001). The clonogenicity of eMSC was also decreased when cultured with JAG1 knockdown EV (Fig.  5 B, n = 6, P  < 0.05). Western blotting demonstrated that the upregulation of Notch signaling in eMSC was lost when JAG1 was silenced in the myometrial EV (Fig.  5 C, 5D, n = 6, P  < 0.05). These findings reveal a novel mechanism of JAG1/Notch signaling in the regulation of eMSC through EV. Fig. 5 Knockdown of JAG1 in myometrial cells abolishes the stimulatory effect of their derived EV on eMSC activities. A Relative proportion of CD140b + CD146 + cells by flow cytometry (n = 6). B Relative clonogenic activity of eMSC after culture with different preparation of myometrial EV (n = 6). C-F Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with different preparation of myometrial EV (n = 6). Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; siJAG1 JAG1 , siRNA to JAG1 ; siCtrl, scrambled control siRNA Knockdown of JAG1 in myometrial cells abolishes the stimulatory effect of their derived EV on eMSC activities. A Relative proportion of CD140b + CD146 + cells by flow cytometry (n = 6). B Relative clonogenic activity of eMSC after culture with different preparation of myometrial EV (n = 6). C-F Representative western blotting images and quantitative analysis of Notch related proteins in eMSC after culture with different preparation of myometrial EV (n = 6). Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Myo, myometrial cells; siJAG1 JAG1 , siRNA to JAG1 ; siCtrl, scrambled control siRNA The functional role of myometrial EV on eMSC in vivo was investigated in a mouse endometrial injury model. On day 7 post-injury, CM-Dil labeled eMSC and PKH67 labeled myometrial EV were observed beneath the luminal epithelium of the mouse endometrium (Additional file S5B, upper panel). Some eMSC/myo-EV were localized to the lower stromal regions of the endometrial tissue (Additional file Fig S5B, lower panel). The regenerative capability of eMSC/myo-EV in repair of the injured endometrium was assessed by comparing the morphology of the endometrium in different treatment groups. As expected, endometrial thickness and number of glands were reduced in the injured uterine horn compared with the control horn (Additional file Fig S6 A - 6 C, n = 4, P  < 0.001). No differences were detected in the endometrial thickness between the injured group and the myometrial EV group (Fig.  6 A - 6 C). Transplantation with eMSC elevated the endometrial thickness, though it did not reach a significant difference (Fig.  6 A - 6 C). Remarkable endometrial regeneration was observed in the eMSC/myo-EV group, as evidenced by the increase in tissue thickness (Fig.  6 C, n = 5, P < 0.01) and number of glands (Fig.  6 D, n = 5, P  = 0.06). Taken together, these results show that the myometrial EV can facilitate eMSC in repairing the damaged mouse endometrium. Fig. 6 Proliferation and apoptosis of injured mouse endometrium in different treatment groups. A Representative H&E-stained images of the injured mouse endometrium in different treatment groups. Scale bar: 100 µm. B High magnification images of the region. Scale bar: 50 µm. Relative endometrium thickness (C) and number of endometrium glands (D) compared to its contralateral untreated control in different treatment groups at postoperative day 7. E Representative immunohistochemistry images of proliferative marker Ki67 and apoptotic marker cleaved caspase-3 in the injured mouse endometrium in different treatment groups. Positive staining indicated by the red arrows. Scale bar: 20 µm. Relative expression of Ki67 (F) and cleaved caspase-3 (G) in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups at postoperative day 7. Results are presented as mean ± SD; * P  <.05; ** P  <.01. n = 3–5 per group. Abbreviations: EV extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Str, stroma; GE, glandular epithelium; LE, luminal epithelium; Myo, myometrium Proliferation and apoptosis of injured mouse endometrium in different treatment groups. A Representative H&E-stained images of the injured mouse endometrium in different treatment groups. Scale bar: 100 µm. B High magnification images of the region. Scale bar: 50 µm. Relative endometrium thickness (C) and number of endometrium glands (D) compared to its contralateral untreated control in different treatment groups at postoperative day 7. E Representative immunohistochemistry images of proliferative marker Ki67 and apoptotic marker cleaved caspase-3 in the injured mouse endometrium in different treatment groups. Positive staining indicated by the red arrows. Scale bar: 20 µm. Relative expression of Ki67 (F) and cleaved caspase-3 (G) in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups at postoperative day 7. Results are presented as mean ± SD; * P  <.05; ** P  <.01. n = 3–5 per group. Abbreviations: EV extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells; Str, stroma; GE, glandular epithelium; LE, luminal epithelium; Myo, myometrium Furthermore, proliferation and apoptosis in different treatment groups were evaluated by immunohistochemistry. The number of proliferating Ki67 + cells in the injured mouse endometrium was reduced remarkably (Additional file Fig S6D and S6E) while the expression of cleaved caspase 3 was increased (Additional file Fig S6 F and S6G) when compared with the control group. The endometrial proliferation and apoptosis between the injury group and the myometrial EV alone group remained similar (Fig.  6 E - 6 F). The numbers of Ki67 + cells were significantly higher in the eMSC and the eMSC/myo-EV groups than the injury group (Fig.  6 E and Fig.  6 F, n = 4–5, P  < 0.01). Moreover, transplantation of eMSC or eMSC/myo-EV effectively reduced apoptosis of the injured mouse endometrium when compared with the injury group (Fig.  6 E and Fig.  6 G, n = 3–5, P  < 0.05). Overall, these results indicate that transplantation of eMSC or eMSC/myo-EV promotes proliferation while reduces apoptosis of the mouse endometrial cells. To investigate angiogenesis in the injured endometrium, we evaluated Vegfa mRNA expression using qPCR and confirmed its protein levels by Western blotting. The expression of Vegfa mRNA was significantly reduced in the injured uterine horn compared to the control horn (Additional file Fig S7 A, n = 4, P  < 0.01). No significant difference was observed between the injured group and the myometrial EV group (Fig.  7 A). Notably, transplantation of eMSC or eMSC/myo-EV remarkably enhanced angiogenesis in the injured endometrium, with the most pronounced therapeutic effect observed in the eMSC/myo-EV group (Fig.  7 A, n = 3–5, P  < 0.01 for the eMSC group, P  < 0.001 for the eMSC/myo-EV group). Consistent with these findings, Vegfa protein expression exhibited a similar trend (Fig.  7 B and Fig S7B, n = 5). Fig. 7 Angiogenesis and fibrosis of injured mouse endometrium in different treatment groups. A Relative mRNA expression of angiogenesis marker Vegfa , B Representative western blotting images and quantitative protein analysis of Vegfa in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups. C Relative mRNA expression of fibrotic markers: fibronectin ( Fn ), Collagen type 1a ( Col1a1 ) and alpha smooth muscle actin ( Acta2 ) in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups. Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. n = 3–5 per group. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells Angiogenesis and fibrosis of injured mouse endometrium in different treatment groups. A Relative mRNA expression of angiogenesis marker Vegfa , B Representative western blotting images and quantitative protein analysis of Vegfa in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups. C Relative mRNA expression of fibrotic markers: fibronectin ( Fn ), Collagen type 1a ( Col1a1 ) and alpha smooth muscle actin ( Acta2 ) in the injured mouse endometrium compared to its contralateral untreated control in different treatment groups. Results are presented as mean ± SD; * P  <.05; ** P  <.01; *** P  <.001. n = 3–5 per group. Abbreviations: EV, extracellular vesicles; eMSC, endometrial mesenchymal stem/stromal cells To assess the extent of endometrial fibrosis, the gene expression of several key fibrotic markers, including Fn1, Col1a1 and Acta2 were evaluated. Compared to the control, the injured endometrium exhibited significantly elevated expression of these fibrotic genes (Additional file Fig S7 C, n = 4, P  < 0.05 for Fn1 and Col1a1 , P  < 0.01 for Acta2 ). Transplantation of myo-EV alone had minimal impact on endometrial fibrosis (Fig.  7 C, n = 4). In contrast, both the eMSC and eMSC/myo-EV groups demonstrated a significant reduction in the expression of fibrotic genes compared to the injury group (Fig.  7 C, n = 4, P  < 0.05). These findings collectively suggest that transplantation of eMSC or combined eMSC and myo-EV following injury not only promotes angiogenesis but also attenuates fibrosis in the injured endometrium, highlighting their potential therapeutic value in endometrial repair.

Stem cell niche is a specialized microenvironment that regulates self-renewal of modulates functional properties of stem cells [ 23 ]. Here, we studied cross talk between eMSC and their niche components in human endometrium. The myometrial cells are candidate niche cells of eMSC modulating their biological functions [ 6 , 24 ]. In this study, we demonstrated that myometrial EV can enhance the clonogenicity and self-renewal activity of eMSC. Specifically, the JAG1 containing EV activated Notch signaling in eMSC to regulate their bioactivities. To our knowledge, this is the first report studying the role of EV as a novel intercellular communication method between eMSC and their niche cells. In human endometrium, studies have shown the importance of EV in mediating various physiological as well as pathological processes [ 25 , 26 ]. The present study revealed the involvement of EV in mediating the communication between the eMSC and their surrounding niche. Our findings showed that EV released from the myometrial cells can be internalized by eMSC, leading to a significant effect on the stemness of eMSC. Similar cell–cell communication via EV has been observed in other somatic stem cell niches. The tumor derived EV can reprogramme MSC and dramatically change their phenotypes and functions, enhancing tumor progression [ 27 ]. In addition, EV from human dental pulp cells can efficiently induce odontogenic differentiation of the human dental pulp stem cells and bone marrow derived stromal cells in vitro and in vivo [ 28 ]. Classical Notch signaling depends on direct cell–cell contact interactions. However it can also be activated at a distance by EV containing Notch ligands [ 19 , 29 ]. DLL4 in the endothelial EV can alter the phenotype of endothelial cells towards tip cells via Notch signaling [ 30 ]. JAG1 in EV of MSC in the dental pulp can activate Notch signaling and induce angiogenesis of endothelial cells [ 20 ]. Using the indirect coculture setup, we observed the activation of Notch signaling in eMSC and similar phenomenon occurred after treatment with factors from Myo-CCM. Therefore, we postulated that certain Notch ligand(s) might be packaged in the myometrial EV. Indeed, JAG1 was the only Notch ligand detected in the myometrial EV and activated Notch signaling, thereby promoting self-renewal and clonogenicity of eMSC. Pharmacological inhibition of Notch signaling with DAPT or silencing of NOTCH1 nullified these stimulatory effects. An interesting finding in this study was that EV derived from the myometrial cells but not the stromal cells elicited the stimulatory actions on eMSC. Several possibilities may lead to the difference observed. Firstly, the ability of eMSC to internalize myometrial EV was double that of stromal EV. Secondly, the amount of JAG1 in the myometrial EV was greater than that of the stromal EV. Thus, the myometrial EV induced a stronger Notch activity in eMSC, resulting in better stem cell maintenance. JAG1 plays important role in homeostasis of various stem cells [ 31 , 32 ]. In the postnatal subventricular zone, JAG1 signals are required for the self-renewal of neural stem cells [ 31 ]. For eMSC, when JAG1 was silenced in the myometrial EV, the self-renewal and clonogenic activity of eMSC was reduced. Hence, JAG1 mediated Notch signaling activation to induce the self-renewal of eMSC in vitro. It will also be insightful to analyze the myometrial EV proteome relative to the stromal EV proteome to understand the myometrial EV effects on eMSC in greater detail. Different types of MSC can be used to treat animal models with IUA [ 33 ]. Several recent studies examined the synergistic effect of combined transplantation of menstrual blood derived stem cell with other molecules in animal models of injured endometrium [ 34 , 35 ]. Zhang et al. demonstrated that platelet-rich plasma improved the therapeutic effect of menstrual blood derived stem cell on endometrium regeneration and fertility restoration in a rat model [ 35 ]. However, very few studies investigate the role of niche cells or niche cells derived EV on eMSC in mediating regeneration in vivo. Here, our results demonstrated that transplantation of eMSC together with myometrial derived EV can effectively repair the injured mouse endometrium. The observed therapeutic feature was potentially achieved by elevating cell proliferation and suppressing apoptosis of the endometrium. In addition, transplantation of eMSC with myometrial EV effectively enhanced angiogenesis and reduced fibrosis of the injured endometrium. The functional role of Notch signaling in the treatment of IUA has been described by several groups [ 36 , 37 ]. Findings in this study revealed that the myometrial EV stimulated the biological function of eMSC via Notch signaling in vitro. In the near future, it would be important to explore whether Notch signaling plays a role in the injured mouse endometrium transplanted with eMSC/myo-EV. This study presents several novel discoveries but with limitations. In the present study, we mainly focused on myometrial cells and stromal cells in the endometrial niche compartment. Endometrial epithelial niche cells at menses can also regulate eMSC function [ 4 ]. EV from endometrial epithelial cells actively support embryo growth, development and implantation [ 10 ]. Proteomic analysis revealed that the proteins within these EV were highly enriched with molecules for cell adhesion, differentiation, migration, extracellular matrix organization and vasculature development [ 38 ]. Since JAG1 is highly expressed in the endometrial epithelial cells [ 39 ], further studies should characterize their expression in EV derived from endometrial epithelial cells. Moreover, epithelial organoids derived from multiple tissues actively secrete EV to exhibit their regulatory effect [ 40 ]. Whether endometrial epithelial organoids produce EV that modulate the properties of eMSC needs further investigation. Second, eMSC-derived EV contain a wide range of proteins and microRNAs that are involved in immune processes, apoptosis and regeneration of the endometrium [ 41 , 42 ]. Future studies should examine the role of eMSC-derived EV and their action on the niche cells. Third, the biogenesis of EV, including generation, sorting, releasing, internalization and their function in recipient cells are complicated. The regulatory mechanisms involved are not well-defined. Given that microRNAs are highly enriched in EV, performing a miRNA microarray analysis on the isolated EV will provide novel insights of the regulatory networks within the eMSC niche. In conclusion, JAG1-containing EV can be up-taken by eMSC and subsequently enhance the maintenance of eMSC via upregulation of Notch signaling. These results suggest a new mechanism of JAG1/Notch1 signaling on the regulation of eMSC that does not require classical cell–cell contact.

Human endometrium is a highly regenerative tissue undergoing cycles of proliferation, differentiation and shedding in reproductive women [ 1 ]. In the past two decades, accumulating evidence have shown endometrial stem/progenitor cells are key players in endometrial remodeling [ 2 ]. Endometrial MSC can be isolated from endometrial biopsies or full thickness endometrial tissues based on the co-expression of CD140b and CD146 [ 3 ]. Efforts have been made to understand the regulatory role of niche cells on eMSC functions [ 4 , 5 ]. We demonstrated that the myometrial cells could stimulate the self-renewal activity of CD140b + CD146 + eMSC via paracrine mechanism [ 6 ]. Whether other forms of intercellular communication, such as extracellular vesicles (EV), may participate in stem cell maintenance in the human endometrium remains largely unknown. EV are vesicles of size 50–150 nm in diameter that are released from cells into surrounding environment [ 7 ]. As a player in cell–cell communication, EV can deliver a variety of bioactive substances including miRNAs, proteins and lipids to recipient cells and regulate their biological function [ 8 ]. In the female reproductive tract, EV have been identified in follicular fluid, endometrium, placenta and amniotic fluid [ 9 ]. For example, EV-based communication between blastocyst and endometrium during implantation is critical for a successful pregnancy outcome [ 10 ]. EV are also involved in the development of endometriosis [ 11 ]. These observations strongly suggest that EV serve as a mean of cell–cell communication in the endometrial microenvironment. However, no studies have investigated the cell–cell communication of EV between eMSC and their niche cells. Asherman’s syndrome or intrauterine adhesions (IUA) is a common gynecological disease, caused by endometrial infection or damage to the basal layer of endometrium [ 12 ]. Patients with IUA often suffer from infertility, amenorrhea, or recurrent pregnancy loss, which severely hamper their reproductive functions [ 13 ]. The potential use of eMSC in repairing damaged endometrium have been reported as these cells exhibit several therapeutic features including increased endometrial proliferation, promotion of angiogenesis and reduction of fibrosis [ 14 ]. In vitro studies demonstrate the importance of niche signals from the uterine microenvironment on eMSC self-renewal and proliferation [ 4 , 6 ]. However, little is known whether the endometrial niche cells derived EV can facilitate the therapeutic effect of eMSC in mediating endometrial regeneration in vivo. Notch signaling is a cell–cell signaling pathway activated by specific receptor-ligand interactions [ 15 ]. It is highly conserved during evolution and plays a critical role in various biologic processes, such as tissue homeostasis and maintenance of somatic stem cells [ 16 ]. Notch signaling is involved in regulating eMSC in vitro and in vivo [ 17 , 18 ]. Although the myometrium is anatomically adjacent to the basalis where the eMSC reside, there is no direct “cell–cell contact” between eMSC and myometrial cells under physiological condition. Besides cell–cell contact, Notch signaling can be activated by EV [ 19 , 20 ]. Our preliminary data showed that indirect coculture with myometrial cells activated Notch activity in eMSC. Therefore, we hypothesized that myometrial cells may secrete Notch ligands containing EV that modulate eMSC function in vitro. In the present study, we investigated the role of EV as a novel mode of intercellular communication between eMSC and their niche cells. We demonstrated that EV produced by myometrial cells, but not stromal cells, maintained the eMSC. The effect was mediated by JAG1, which was incorporated into the myometrial derived EV. We further showed that the JAG1 packaged in the myometrial EV could activate Notch signaling in eMSC and enhance their biological function, and that a combined transplantation of eMSC with myometrial derived EV enhanced the therapeutic effect of eMSC in endometrial regeneration in vivo.

Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Supplementary Material 6. Supplementary Material 7. Supplementary Material 8. Supplementary Material 9. Supplementary Material 10. Supplementary Material 11. Supplementary Material 12. Supplementary Material 13. Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Supplementary Material 4. Supplementary Material 5. Supplementary Material 6. Supplementary Material 7. Supplementary Material 8. Supplementary Material 9. Supplementary Material 10. Supplementary Material 11. Supplementary Material 12. Supplementary Material 13.