Human bone marrow-derived MSCs were generated using standard plastic adherence method from healthy donor bone marrow aspirates (surplus to hematopoietic stem cell transplantation, obtained from the Newcastle Cellular Therapy Facility, Newcastle upon Tyne, UK). In brief, bone marrow mononuclear cells (MNCs) were isolated by density gradient centrifugation using Lymphoprep™ (Axis-Shield). MNCs were then plated at a density of 2 × 10⁷ cells/flask in T-25 tissue culture flasks in basal medium containing Dulbecco's modified eagle medium, 100 IU/ml penicillin, 100 μg/ml streptomycin, 2 IU/ml heparin and 2 mM L-glutamine (all from Sigma-Aldrich), supplemented with 5% human platelet lysate (hPL; PLTMax, Mill Creek Lifesciences) (23). The cells were cultured for 3 days at 37°C in a 5% CO₂ incubator. The non-adherent cell fraction was discarded, and fresh medium was added to the adherent cells. Medium was refreshed every 3 days and cells were passaged when the culture reached 70–80% confluence. MSCs at passage 3 were characterized according to the criteria described by the International Society of Cellular Therapy (ISCT) (24) and used in all experiments throughout this study.

MSC-EVs were collected from MSC conditioned medium by differential ultracentrifugation, as previously described (25). EV-depleted medium was prepared by overnight ultracentrifugation at 100,000 × g of basal medium supplemented with 10% hPL. When passage 3 MSCs reached 90% confluence, cells were washed twice with phosphate buffered saline (PBS, Sigma-Aldrich) and cultured in EV-depleted medium, at a final concentration of 5% EV-depleted hPL, for a further 48 h prior to MSC-EV isolation. The conditioned medium was then centrifuged at 400 × g for 5 min at 4°C to exclude detached cells and debris. The resulting supernatant was centrifuged at 2,000 × g for 20 min at 4°C, transferred to ultracentrifuge tubes (Beckman Coulter) and centrifuged sequentially at 10,000 × g for 45 min and at 100,000 × g for 90 min at 4°C using a 45Ti rotor (Beckman Coulter) in a BECKMAN L8-80 ultracentrifuge (Beckman Coulter). The MSC-EV pellet was washed in 60 ml of PBS then re-suspended in at least 100 μl of sterile PBS and stored at −80°C.

Collected MSC-EVs were characterized based on their morphology, particle size and surface protein expression. EV morphology was visualized using transmission electron microscopy (TEM). Briefly, 5 μl of PBS suspended MSC-EVs were adsorbed for 30 s onto a carbon-coated, glow discharged grid. Excess liquid was removed with a filter paper (Whatmann no. 50, Sigma-Aldrich). Samples were stained with 1% uranyl acetate for 30 s. Excess uranyl acetate solution was removed and the MSC-EV loaded carbon-coated grids were dried under a lamp. Grids were examined using a Philips CM100 Compustage (FEI) transmission electron microscope and digital images were collected using an AMTCCD camera (Deben) housed in Electron Microscopy Research Services, Newcastle University. The particle size and concentration were measured by nanoparticle tracking analysis (NTA). Briefly, MSC-EV pellets were diluted at 1:100–1:500 ratios with sterile-filtered PBS and analyzed using a Nanosight LM10 (Malvern), as described by the manufacturer's protocol. Three 1-min measurements were taken for each sample. The acquired data was processed using NTA 2.3 software (Malvern). EV surface markers CD63, CD9 and CD81 were analyzed by flow cytometry. The volume of EVs suspension equivalent to 5 μg of protein [measured with the microBCA protein assay kit (ThermoFisher Scientific)], were coated onto 4-μm aldehyde/sulfate latex beads (LifeTechnologies) by overnight incubation on a rotary wheel at RT. The reaction was stopped by incubation with 1 M glycine (Sigma-Aldrich) for 30 min at RT. The MSC-EV-bead complex was washed twice with PBS with 0.5% fetal calf serum (FCS, Invitrogen), and incubated with CD63 PE (H5C6), CD9 PerCPCy5.5 (M-L13) and CD81-APC (JS-81) antibodies or corresponding isotype controls (all from BD Biosciences) for 20 min at 4°C. Following further washes in PBS with 0.5% fetal calf serum data was acquired using a BD FACS Canto II cytometer and analyzed with FlowJo software (Tree Star).

Peripheral blood mononuclear cells (PBMCs) were obtained from healthy donors by gradient centrifugation over Lymphoprep™. CD3⁺ T cells and CD14⁺ monocytes were immuno-magnetically isolated by depletion of non-T cells using the Pan T cell isolation kit (Miltenyi Biotec) and positive selection using human CD14 microbeads (Miltenyi Biotec), respectively. Highly purified (>90% purity) CD3⁺ T cells and CD14⁺ monocytes were used in the subsequent experiments.

Mature and immature monocyte-derived dendritic cells (mDC and iDC, respectively) were generated from magnetically isolated CD14⁺ monocytes (26). The cells were cultured in complete medium (RPMI 1640 with 10% FCS, 2 mM L-glutamine and 100 IU/ml penicillin, 100 μg/ml streptomycin) supplemented with IL-4 and GM-CSF (both at 50 ng/ml; Immunotools) for 7 days. MSC-EVs at a concentration of 20 μg/ml were added to the DC generation culture on day 3, when fresh medium containing the same concentrations of IL-4 and GM-CSF was replenished. Mature DCs were generated by adding LPS (100 ng/ml; Sigma) to the DC culture on day 6. The EV-treated immature and mature DCs (iDC-EV and mDC-EV) were harvested on day 7 and used in the subsequent experiments.

The influence of MSC-EVs on T cell proliferative response was evaluated in two experimental settings. Initially MSC-EVs were added directly into the co-cultures of CFSE labeled CD3⁺ T cells and allogeneic mDCs, with the co-cultures without added MSC-EVs as controls. Subsequently, MSC-EVs were added into the DC generation cultures as described above. Mature DCs, with or without previous exposure to MSC-EVs, were co-cultured with CFSE labeled allogeneic CD3⁺ T cells. The same T cell: DC ratio (10:1) was used in both experimental settings. The amount of MSC-EVs used in both settings was identical (20 μg/ml), which was pre-determined by dose titration (Supplementary Figure S1). This MSE-EV dose was used in all functional analysis throughout this study. The levels of T cell proliferation were determined by CFSE dilution using flow cytometry on day 5 of the co-culture, with the frequency of CFSE^low cells as readout for the levels of T cell proliferation.

MSC-EVs were labeled with PKH26 Red Fluorescence Cell Linker Kit (Sigma-Aldrich), following the manufacturer's protocol with minor modifications. Briefly, 20 μg/ml of MSC-EVs were pelleted at 100,000 × g for 90 min and re-suspended in 1 ml of Diluent C to which 2 μl of PKH26 was added. Diluent C containing PKH26 without MSC-EVs was used as a negative control. After a 20-min incubation at RT in the dark, the reaction was stopped with 1 ml of 1% bovine serum albumin (BSA; Sigma-Aldrich). Labeled MSC-EVs were washed twice with filtered PBS at 100,000 × g for 90 min then re-suspended in complete medium and added to the co-culture of allogeneic CD3⁺ T cells and mDCs. After 24 h of incubation at 37°C in a 5% CO₂ humidified atmosphere, the cells were washed in PBS, suspended in serum-free RPMI 1640 and seeded onto a poly-L-lysine coated coverslip (2 × 10⁵ cells/coverslip). Cells were incubated for 15 min at 37°C to allow attachment to the coverslip and fixed with 2% paraformaldehyde for 15 min at RT. Cells were washed 2 × with PBS and blocked with antibody diluent (PBS with 0.1% Triton X (Sigma-Aldrich) and 0.5% BSA), for 30 min at 4°C. The cells were then incubated overnight with mouse anti-human CD3 antibody (1:100, BD Biosciences) at 4°C. Coverslips were washed 2 × with PBS containing 0.1% Triton and 0.1% BSA and incubated with the secondary antibody (goat anti-mouse Dy649, 1:300; Jackson Immunoresearch) for 2 h at 4°C. The coverslips were washed 2 × and incubated with HLA-DR FITC (1:20, BD Biosciences) for an additional 4 h, washed 2 × and mounted with Vectashield mounting medium containing DAPI (Vector Laboratories). The samples were imaged using Axioimager Z2 fluorescence microscope with axiovision software V4.8 (Zeiss).

The staining for cell surface proteins were performed following standard protocols by incubation of cells with pre-optimized concentrations of antibodies at 4°C for 20 min in FACS buffer (PBS with 2% FCS and 1 mM EDTA). Flow cytometry data were acquired on BD FACS Canto II cytometer and analyzed with FlowJo software version X (Tree Star). Unless otherwise stated all antibodies were supplied by BD Biosciences: CD25 APC (M-A251) or CD25 PECy7 (2A3); CD8 APC (SK1); CD4 V500 (RPA-T4), CD3 PerCPCy5.5 (SK7), CD38 PE (HIT2); CD80 PE (L307.4); CD83 FITC (HB15e); CD86 PECy7 (FUN-1), HLA-DR PerCP (L243); CCR7 PECy7 (G043H7; Biolegends). Appropriate isotypes were used as controls.

The ability of immature DC to uptake antigens was assessed by incubation of immature DCs (5 × 10⁴) with FITC-dextran (1 mg/ml; Sigma-Aldrich) in basal media for 1 h at 37°C including a control incubated at 4°C to exclude extracellular binding of FITC-dextran. Cells were then washed three times in cold FACS buffer and analyzed by flow cytometry.

The level of cytokines was quantified in the culture supernatants collected from either DC generation cultures on day 7 or DC and T cell co-cultures on day 5 using cytometric bead array flex kit, according to the manufacturer's instructions (BD Biosciences). DC supernatants were assessed for IL-6, IL-10, IL-12p70, and TGF-ß1 whilst DC-T cell co-culture supernatants for IL-6, IFNγ, TNFα and IL-2. Data was acquired on a FACS Canto II and analyzed with FCAP software (BD Biosciences).

DC migration was assessed as described by Anderson et al. (27), with minor modifications (27). Briefly, migration was measured using a transwell system (pore size 8 μm, Corning Lifesciences), mDCs (1 × 10⁵) were added in the upper chamber, and medium, with or without CCL21 (250 ng/ml, R&D systems) in the lower chamber. After 24 h-incubation at 37°C migrated mDCs in the lower chamber were harvested and counted using a Neubauer chamber (Celeromics). Migration efficiency was calculated from the percentage of the input cells that had migrated. The percentage of mDCs which had migrated toward the medium without CCL21 served as a negative control. The migration conditions used in this study were pre-optimized (Supplementary Figure S2).

Total RNA in MSC-EVs was purified using the Total Exosome RNA and Protein isolation kit (Thermo Fisher Scientific). Total RNA from the parent MSCs was purified from cell lysates using the Qiagen RNAeasy mini kit, as per the manufacturer's instructions. The concentration and quality of total RNA from MSC and MSC-EVs was determined using Bioanalyzer 2100 with the RNA 6000 Nano and Pico kits, respectively (Agilent technologies). At least 2 ng of total RNA were used as input for the nCounter® Human miRNA Expression assay V3 kit (NanoString Technologies). The miRNA expression profiles were then analyzed using the nSolver software V2.5 (NanoString Technologies) according to the manufacturer's instructions. Samples were normalized to the geometric mean of the 100 miRNAs with the highest counts and a microRNA was considered as differentially expressed when its expression exhibited a fold increase >1.5 and a p-value < 0.05 as determined by the nSolver software using Student's t-tests.

Validation of MSC-EV and MSC miRNA expression profiles was performed by qRT-PCR. Total RNA extraction and quality was assessed as aforementioned. The miRNA specific copy DNA synthesis was carried out using the SensiFAST™ probe HI-ROX master mix (Bioline) and the Taqman® primer probe sets for: hsa-mir-21-5p (5′-UAGCUUAUCAGACUGAUGUUGA-3′), hsa-miR-223 (5′-UGUCAGUUUGUCAAAUACCCCA-3′), has-miR142-3p (5′-UGUAGUGUUUCCUACUUUAUGGA-3′), hsa-miR-126 (5′-UCGUACCGUGAGUAAUAAUGCG-3′) and U6 snRNA control (5′-GTGCTCGCTTCGGCAGCACATATACTAAAATTGGAACGATACAGAGAAGATTAGCATGGCCCCTGCGCAAGGATGACACGCAAATTCGTGAAGCGTTCCATATTTT-3′), (all from Thermo Fisher Scientific).

The miRIDIAN microRNA hsa-miR-21 mimic and the miRIDIAN microRNA transfection control with Dy547 were purchased from Dharmacon. The siRNA dried pellets were re-suspended in RNase free water to a stock solution of 20 μM, then stored at −80°C. The transfection of DCs was performed based on a lipid-siRNA transfection method as previously described (28). Briefly, 3 × 10⁵ CD14⁺ monocytes were seeded in 24-well plates in complete media supplemented with 50 ng/ml of GM-CSF and IL-4. On day 3 of the culture, the lipid-siRNA complexes were formed by combining 117.5 μls of warm serum-free RPMI with 3.75 μls of HiPerFect transfectant (Qiagen) and 3.75 μls of siRNA stock solutions (final concentration of 200 nM). The lipid-siRNA mixture was incubated for 15–20 min at RT with constant mixing to enhance the formation of the complexes. 125 μls of the lipid-complex were added to one side of a well of a tilted 24-well plate and day 3 DCs were added directly into the lipid-siRNA complexes and the plate was gently swirled to ensure uniform distribution of the lipid-siRNA complexes then incubated at 37°C and 5% CO₂ for 4 h. The transfection was then stopped with RPMI supplemented with GM-CSF and IL-4 and cells were kept in culture until day 7 as aforementioned. Transfection efficiency and viability of the cells was monitored on days 4, 5, 6, and 7 of DC generation and the effect of miR-21 overexpression on DC maturation and function was assessed.

Unless otherwise stated, all statistical analysis was carried out using GraphPad Prism version 6 (GraphPad Software). Mann-Whitney U tests or Student's t-tests were used to determine statistical significance. Differences were considered statistically significant at a p-value < 0.05.

MSCs were generated from healthy bone marrow aspirates and expanded in xeno-free conditions. The generated MSCs were confirmed as being compliant with the ISCT criteria (24) prior to be used for EV purification (Supplementary Figure S3). MSC-EV purification yielded an average of 1.16 × 10¹¹ ± 2.95 × 10¹⁰ particles/ml, or 432 ± 229 particles/cell (Figure 1A). Purified MSC-EVs showed a modal size of 152 ± 23 nm (Figure 1B), and exhibited the expected cup-shaped morphology, as assessed by TEM (Figure 1C). MSC-EVs positively expressed common transmembrane proteins enriched in EVs, including CD63, CD9, and CD81 (Figure 1D).

To identify whether MSC-EVs primarily target DCs or T cells, PKH26 labeled MSC-EVs were added into the co-culture of purified T cells and mDCs then visualized the spatial association of MSC-EVs with DCs and T cells. We found a predominant association between MSC-EVs and DCs detected by the fluorescence staining overlap of PKH26 labeled MSC-EVs (red) and HLA-DR labeled DCs (green). No association was detected between the labeled MSC-EVs and CD3⁺ T cells (Figure 2).

Immature DCs (iDCs) effectively capture and process antigens, which can be assessed by their uptake of FITC-dextran. Upon DC maturation following the stimulation with LPS (a TLR4 ligand), a function switch occurs from antigen capture and processing to predominantly presenting antigens to T cells (29). As expected, iDCs showed significantly higher levels of FITC dextran uptake than the mDCs (p = 0.003). MSC-EV treated iDCs (iDC-EV) showed a decreased ability to take up FITC dextran (p = 0.012; Figure 3), suggesting that MSC-EVs can exert a modulatory effect on DC function by blocking the first step of DC maturation.

We next examined the effect of MSC-EV treatment on DC maturation and activation by assessing their expression of CD83, CD38, co-stimulatory molecules CD80 and CD86, as well as HLA-DR. Mature and immature DCs showed expected high or low levels of all the aforementioned surface markers, respectively (p < 0.05; Figure 4A). MSC-EV treated mDCs expressed a markedly reduced expression of CD83, CD38 and co-stimulatory molecules CD80 compared to the untreated mDCs (p = 0.009, 0.05, and 0.03, respectively). No significant difference was detected in the expression of CD86 and HLA-DR between MSC-EV treated and untreated mDCs (Figure 4A).

Upon LPS stimulation mature DCs secrete high levels of pro-inflammatory cytokines, IL-6 and IL-12 (30). To verify the impact that MSC-EVs may have on mature DC cytokine secretion following LPS stimulation the levels of IL-6, IL-12-p70, IL-10, and TGF-β were measured in the culture supernatants of iDCs, mDCs, and MSC-EV treated mDCs. The levels of IL-6 production by MSC-EV treated mDCs was markedly lower than that secreted by untreated mDCs (p = 0.01) but higher than that secreted by iDCs (p = 0.03). MSC-EV treated mDCs produced the IL-12-p70 at a level that was over 100-fold lower than that produced by untreated mDCs (p = 0.036) but no significant difference to that produced by iDCs (p > 0.05). Importantly, MSC-EV treated mDCs showed an increased production of the anti-inflammatory cytokine TGF-β compared to untreated mDCs (p = 0.05). No significant difference was observed in IL-10 production across all DC groups (Figure 4B). To evaluate the functional effect of MSC-EV treated DCs on T cell responses we stimulated allogeneic T cells with MSC-EV treated DCs and measured T cell cytokine secretion and T cell proliferation. The results showed that T cells stimulated with MSC-EV treated DCs secreted lower levels of IL-6 and IFN-γ compared to those co-cultured with untreated DCs (p = 0.05 for both) while no significant difference was observed in TNFα and IL-2 production between the two groups (p > 0.05 for both; Figure 4C). However, MSC-EV treated DCs did not show a suppressive effect on allogeneic T cell proliferation compared to the T cells stimulated with untreated mDCs (Figure 4D).

CCR7 plays an essential role in DC homing to the lymph nodes. Upon the presence of inflammatory stimuli, DCs mature and gain CCR7 expression which, by binding to its ligand CCL21 or CCL19, direct DCs to the lymph nodes for antigen presentation (31). We evaluated CCR7 expression on mDCs, with or without MSC-EV treatment, and their ability to migrate toward CCL21 using the transwell system. We detected a 3-fold lower CCR7 expression on MSC-EV treated DCs compared to the untreated counterpart (p = 0.003, Figure 5A). In addition, MSC-EV treated DCs showed a significantly reduced migration efficiency toward CCR7 ligand CCL21 in comparison to untreated DCs (22.4 ± 4.0% and 73.2 ± 6.4%, respectively; p < 0.0001, Figure 5B).

To further scrutinize the molecular mechanisms underlying MSC-EV mediated modulation of DC maturation and function, we profiled the miRNA contents in MSC-EVs in comparison to their parent cells using the NanoString technology. High quality of RNA extracted from MSCs and MSC-EVs was confirmed by bioanalyzer analysis. MSC-derived RNA showed RNA integrity numbers (RIN) >7 and exhibited single narrow peaks at the 18 and 28S ribosomal RNA (Figure 6A). MSC-EVs enclosed only small RNA species with sizes from 25 up to ~200 nucleotides. No RIN was calculated for MSC-EV derived RNA due to the absence of the ribosomal RNA peaks (Figure 6B). The microRNAs in MSCs and MSC-EVs displayed a polarized distribution as shown by the principal component analysis plot (Figure 6C). The NanoString microRNA profiling identified 79 microRNAs that were differentially expressed between MSC-EVs and MSCs. These differentially expressed miRNAs were presented in a heatmap based on a hierarchical clustering analysis (Figure 6D). Amongst these 79 miRNAs, 49 were enriched in MSC-EVs while the remaining 30 were enriched in the parent cells. MiRNAs with known effect on DC maturation and function, including miR-21-5p, miR-142-3p, miR-223-3p, and miR-126-3p, were detected within the top 10 enriched miRNAs in MSC-EVs (Table 1). These microRNAs were validated using RT-qPCR in an independent cohort of samples and showed significantly higher expression levels in MSC-EVs compared to parent cells (Figure 6E), suggesting a potential selective packaging of these microRNAs into EVs, which may serve as a vehicle to deliver the miRNA contents to the target cells. A detailed list of the 79 differentially expressed miRNAs detected in MSC-EVs and respective parent cells are provided in Supplementary Table S1.

MicroRNA-21-5p is one of the most enriched miRNAs enclosed in MSC-EVs (Table 1). Target analysis revealed that the CCR7 gene can be targeted by miR-21-5p (Figure 7A), suggesting the enriched miRNA-21-5p in MSC-EV may function as a potential molecular mechanism underpinning MSC-EV mediated impairment of DC migration and function. To confirm this hypothesis, we examined whether overexpression of miR-21-5p in DCs could recapitulate the effect of MSC-EV treatment by transfecting DCs with a miR-21-5p mimic and a fluorescently labeled miRNA scramble on day 3 of DC generation culture using a lipid based transfection method as previously reported (28). Post-transfection assessment showed the highest transfection efficiency at 72 h with comparable cell viability between transfected and non-transfected cells (Supplementary Figures S4A–C). Overexpression of miR-21-5p was confirmed by RT-qPCR compared to the non-transfected and scramble miRNA controls (p < 0.05; Supplementary Figure S4D). Functionally, miR-21-5p mimic transfected DCs exhibited a significant decrease in their migratory capacity toward CCL21 in comparison to non-transfected and scramble miRNA transfected controls, with a migration efficiency of 18.25 ± 4.43%, 44.75 ± 10.33%, and 44.75 ± 10.76%, respectively (p < 0.05 for both; Figure 7B). MiR-21-5p mimic transfected DCs also showed a trend toward a reduction in CCR7 expression compared to the controls, although a statistical significance was not achieved (Figure 7C). In addition, a tendency of secreting lower levels of IL-6 and IL-12p70 was observed in MiR-21-5p mimic transfected DCs compared to both controls, although a statistical significance was not reached (p > 0.05 for all; Supplementary Figure S4E).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.