Mesenchymal stromal cells-derived extracellular vesicles reprogramme macrophages in ARDS models through the miR-181a-5p-PTEN-pSTAT5-SOCS1 axis

Rationale A better understanding of the mechanism of action of mesenchymal stromal cells (MSCs) and their extracellular vesicles (EVs) is needed to support their use as novel therapies for acute respiratory distress syndrome (ARDS). Macrophages are important mediators of ARDS inflammatory response. Suppressor of cytokine signalling (SOCS) proteins are key regulators of the macrophage phenotype switch. We therefore investigated whether SOCS proteins are involved in mediation of the MSC effect on human macrophage reprogramming. Methods Human monocyte-derived macrophages (MDMs) were stimulated with lipopolysaccharide (LPS) or plasma samples from patients with ARDS (these samples were previously classified into hypo-inflammatory and hyper-inflammatory phenotype) and treated with MSC conditioned medium (CM) or EVs. Protein expression was measured by Western blot. EV micro RNA (miRNA) content was determined by miRNA sequencing. In vivo: LPS-injured C57BL/6 mice were given EVs isolated from MSCs in which miR-181a had been silenced by miRNA inhibitor or overexpressed using miRNA mimic. Results EVs were the key component of MSC CM responsible for anti-inflammatory modulation of human macrophages. EVs significantly reduced secretion of tumour necrosis factor-α and interleukin-8 by LPS-stimulated or ARDS plasma-stimulated MDMs and this was dependent on SOCS1. Transfer of miR-181a in EVs downregulated phosphatase and tensin homolog (PTEN) and subsequently activated phosphorylated signal transducer and activator of transcription 5 (pSTAT5) leading to upregulation of SOCS1 in macrophages. In vivo, EVs alleviated lung injury and upregulated pSTAT5 and SOCS1 expression in alveolar macrophages in a miR181-dependent manner. Overexpression of miR-181a in MSCs significantly enhanced therapeutic efficacy of EVs in this model. Conclusion miR-181a-PTEN-pSTAT5-SOCS1 axis is a novel pathway responsible for immunomodulatory effect of MSC EVs in ARDS.


Characterization of the morphology, size and surface marker expression of MSC-EVs
MSC EVs were characterized according to the International Society for Extracellular Vesicles (ISEV) guidelines(2). Morphology was assessed using electron microscopy. MSC-EVs were isolated from 10 6 cells and the EV pellet resuspended in 500 µl of 4% paraformaldehyde (PFA, ThermoFisher) and fixed for 30 minutes at RT. After fixation and washing the pellet three times, the EVs were resuspended in 30 µl PBS before 10 µl of each sample was added into formvar/carbon-coated grids (Science Services, München) for 20 minutes at RT. The grids were dried on filter paper (Whatman, UK) and fixed with 2% glutaraldehyde for 10 minutes.
The grids were then washed in distilled water for 1 minute and subsequently dried. Next, a drop of 1% tannic acid ACS reagent (Sigma Aldrich, UK) was added for 40 minutes to stain mitochondrial membranes. Grids were then washed twice using PBS for 1 minute each, dried and stained with TAAB EM heavy metal stain 336 (TAAB, Laboratory Equipment LTD) for 30 minutes. Grids were then submerged in 50% ethanol (Sigma Aldrich) before being washed in distilled water. The morphology of MSC-EVs were visualized using transmission electron microscopy (TEM) (JEOL, JEM 1400Plus, Japan).
MSC-EV size distribution and concentration was assessed on a Nanoparticle tracking analysis (NTA) device, NanoSight NS300 (Malvern, UK). MSC-EVs were extracted from 10 6 cells and diluted in distilled water to a final volume of 1 mL. Particles present in EVs were measured according to the NanoSight NS300 user manual with the detection threshold set to measure as many particles as possible. For each test, five 1-minute videos were recorded at 25 o C and the NanoSight Software NTA was used for analysis.
Subsequently, 1 mL of 0.1% BSA (Sigma Aldrich) supplemented with 0.01% NaN3 (G-Biosciences) was added to the sample and left overnight. Bead-coupled MSC-EVs were centrifuged at 2000 x g for 10 minutes, washed and centrifuged. The pellet was then stained with FITC Mouse anti human CD44 (BD Biosciences, #555478) and PE-Cy7 Mouse antihuman CD63 (BD Biosciences, #561982) for 45 minutes at 4 o C. After the incubation, MSC-EVs were washed twice and resuspended in PBS for flow cytometry. Bead-coupled EVs and the respective isotype control antibody (negative control) were included in the flow cytometry experiments. Gating of EV-coupled beads were performed based on FCS/SSC parameters, with unbound EVs or possible antibody aggregates excluded from the analysis. Flow BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) cytometry was performed using a BD FACSCanto™ ll flow cytometer using the FACSDiva software and data analysis was performed using FlowJo software (FlowJo, Ashland, OR).

Treatment of MDMs with MSC-CM/MSC-EVs
Monocyte-derived macrophages (MDM) were treated with MSC-CM or MSC-EVs in the presence of E. Coli LPS O111:B4 (Millipore) at 10 ng/ml or 10% ARDS patients plasma from either a hypoinflammatory or hyperinflammatory subphenotype(4). Ethical approval for use of patient samples for laboratory research was obtained from the Office for Research Ethics Committees Northern Ireland.
Membranes were subsequently probed with secondary antibody, anti-rabbit IgG HRP-linked Antibody #7074 at 1:2500 dilution in TBS-T (Cell Signalling Technology). Bands were visualised using the Clarity Western ECL Substrate (Bio-Rad) or SuperSignal West Femto Maximum Sensitivity Substrate (ThemoFisher Scientific) on the Chemi-XX6 G-Box (Syngene) using the GeneSys software. Densitometric analysis of bands were carried out using ImageJ 1.47v software (NIH). The relative density of the target protein and a loading control (β-actin BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) Proinflammatory cytokine secretion of tumor necrosis factor-α (Human TNF-α Duoset -DY210) and interleukin-8 (Human IL-8 Duoset -DY208) in cell supernatants were analysed as per the manufacturer's instructions (R&D). ELISAs were developed using TMB-I solution (Biopanda Diagnostics), with absorbances measured at 450 nm (with subtraction of correction reference wavelength of 570 nm) and data analysed by 4-paramater logistical regression.

Membrane-based phospho-kinase antibody array
The Proteome Profiler Array human phosphor-kinase array (R&D, ARY003B) kit was used to detect the phosphorylation of protein kinases. MDMs were seeded at 300,000 cells per well in 24 well plates and co-cultured with MSCs (60,000 per insert) seeded on 0.4μm transwell inserts (ThermoFisher). MDMs were stimulated with ARDS patients BALF from the HARP2 clinical trial for 24 hrs (ISRCTN88244364) before lysis in the kit lysis buffer containing protease and phosphatase inhibitors. The detection of phosphorylated proteins was visualized using chemiluminescent detection and the relative levels of protein phosphorylation (pixel density in each spot of the array minus background) was assessed between the samples using densitometry.

Small RNA Next Generation Sequencing
MSCs were stimulated with BALF (30% Cf) from ARDS patients for 24 hrs, EVs isolated and RNA including small RNAs extracted (miRNeasy, Qiagen). RNA integrity was assessed on Qubit™ RNA HS Assay Kit (Invitrogen). Small RNAs were converted to cDNA libraries using the NEXTFLEX® Small RNA-Seq Kit v3 (Perkin-Elmer). QC of the libraries and NGS was performed by the genomics core facility at Queens University Belfast on a NextSeq 550 System (Illumina). FASTq files were uploaded onto CLC Genomics Workbench using the small RNA pipeline/analysis tool (Qiagen Digital Insights). This tool was used for trimming of sequencing reads, counting and annotating the results using miRBase v21.

Knockdown of SOCS1 expression and pharmaceutical targeting of STAT5 in MDMs
Small interfering RNA (siRNA) transfection was used to knock down expression of SOCS1 in human MDMs. siRNAs included ON-TARGET plus Human SOCS1 SMARTPool siRNA and ON-TARGET plus non-targeting control Pool (Dharmacon). MDMs were seeded at 300,000 BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) µM siRNA solution was then prepared in 1X siRNA buffer. In separate tubes, the siRNA Tube 1 (2.5 µl 5 µM siRNA plus 47.5ul serum-free RPMI1640) and the DharmaFECT transfection reagent Tube 2 (1 µl plus 49ul serum-free RPMI1640) were prepared and left for 5 min at RT.
These were then mixed together before incubation for a further 20min at RT. In each well, 100 µl of transfection complexes were added to 400 µl of 10% FBS antibiotic-free RPMI1640. The final concentration of siRNA was 25 nM. Cells were incubated at 37 o C in 5% CO2 for 48 hr.
Western Blot was used to assess the transfection efficacy.
AC-4-130 (AOB36422, AOBIOUS), a specific STAT5 inhibitor that targets the SH2 domain of STAT5 was used to block STAT5 signalling in macrophages. 1mg AC-4-130 was dissolved in dimethyl sulfoxide (DMSO) to generate 1000 µM stock. 5 µM AC-4-130 or the same volume of DMSO in serum-free RPMI1640 was used to treat MDMs for 24 hrs before assessment using western blot.

Overexpression of miRNA181a-5p in MDMs and MSCs
Cells were transfected with miRIDIAN microRNA human miR-181a-5p mimic (Dy547-labeled, Dharmacon) or miRIDIAN microRNA mimic negative control (Dharmacon). MDMs were seeded at 300,000 cells per well in 24-well plates or MSCs at 200,000 cells per well in 6-well plates. The following day the cells were washed and media replaced with 200 µl Opti-MEM (ThemoFisher). Transfection complexes were generated in separate tubes and left at room temperature for 5 minutes. Tube 1 (2 µl Oligofectamine (ThermoFisher) was added to 7.5 µl Opti-MEM for MDMs, or 6 µl Oligofectamine was added to 22.5 µl Opti-MEM for MSCs) and for Tube 2 (1.25 µl 20 µM miR-181a-5p / miRNA scramble (negative control) was added to 40 µl Opti-MEM for MDMs, or 3.75 µl 20 µM miR-181a-5p / miRNA scramble (negative control) was added to 120 µl Opti-MEM for MSCs). Tubes 1 and 2 were mixed and left for 20min at RT, subsequently, 50 µl of transfection complexes were added to appropriate wells. After a 4 hr incubation, 125 µl Opti-MEM with 30% FBS was added to each well for 48 hr. Fluorescence imaging of live MDMs or MSCs was performed using the EVOS FL Auto Imaging System (Life technologies) at 10X magnification usually at 24 hrs after transfection.

Transfection of MSCs with Locked nucleic acid (LNA) miR181a-5p
HSA-miR181a-5p miRCURY locked nucleic acid (LNA) miRNA inhibitor (sequence 5'-ACTCACCGACAGCGTTGAATG -3', QIAGEN) was transfected into MSCs to suppress BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)  (5).

In Vivo LPS-induced Acute Lung Injury Model
Anesthetised C57BL/6 male mice ( BALF was centrifuged at 400 x g at 4 o C for 10 minutes and the supernatant was collected for testing. The cell pellet was resuspended in 200μL eBioscience TM 1X RBC lysis buffer (Invitrogen, ThermoFisher) and incubated at room temperature for 3-4 minutes with occasional shaking, before centrifugation at 400 x g at 4 o C for 5 minutes. Cells were then resuspended in 200 μL DPBS (Gibco). The total number of leukocytes was estimated in each sample using trypan blue exclusion assay.
An SP8 inverted confocal microscope (Leica, TCS SP8, Germany) was used to image the fluorescence at 100X magnification using the LAS X software.

Statistical analysis
Statistical analyses were performed using Prism 7 software (GraphPad, USA). Experiments for MDMs were performed using at least three different donors, the average of three technical replicates was taken as a single data point for each donor, and the data points were pooled together for statistical analysis. Pooled data were scored for normality by Shapiro-Wilk test and presented as the mean with SD. For parametric data, two-tailed Student's t-test or oneway ANOVA followed by Bonferroni post hoc test was used. For non-parametric data, Mann-Whitney U test or one-way ANOVA followed by Dunn's selected comparisons was used.