Microvesicles Derived From Human Bone Marrow Mesenchymal Stem Cells Restored Lung Protein Permeability And Reduced Inflammation In E.Coli Endotoxin-Induced Acute Lung Injury In Mice Partially Through The Transfer Of Keratinocyte Growth Factor MRNA

ObjectiveWe previously found that human mesenchymal stem cells (MSC) restored lung protein permeability and alveolar inflammation following E.coli endotoxin-induced acute lung injury (ALI) in part through the secretion of soluble factors. Recently, MSC were found to release microvesicles (MV) that were biologically active because of the presence of mRNA with reparative properties. MV are circular fragments of membrane released from the endosomal compartment as exosomes or shed from the surface membranes. The current studies were designed to determine if MV released by human bone marrow derived MSC would be effective in restoring lung protein permeability and reducing inflammation in E.coli endotoxin-induced ALI in mice.MethodsCulture of Human Bone Marrow Derived Mesenchymal Stem CellsHuman bone marrow derived mesenchymal stem cells (hMSC) were obtained from an NIH repository from Tulane Center for Gene Therapy. The adult stem cells met all of the criteria for the classification as MSC as defined by the International Society of Cellular Therapy. Briefly, hMSC were cultured with aMEM containing16.5%Fetal Bovine Serum (FBS),1%L-glutamine and1%penicillin-streptomycin in a humidified5%CO2incubator at37℃. Only MSC with the total passage number less than or equal to8were used in the experiments.Isolation of MSC MicrovesicleshMSC were cultured until confluent in P75flasks and then were serum starved for48hours in fresh conditioned medium (aMEM or FBM supplemented with only0.5%of Bovine Albumin Fraction). To isolate the MVs, conditioned medium was centrifuged at3,000rpm for20minutes to remove cellular debris, then at28,500rpm or100,000g for one hour at4℃, washed in PBS and submitted to a second ultracentrifugation. MSC or normal adult human lung fibroblasts (NHLF) MV were resuspended with PBS according to the cell count (10μl MV per1×106cells) and stored at-80℃for furtherMicrovesicles with or without Keratinocyte Growth Factor siRNA PretreatmentFor siRNA experiments, hMSC were cultured in6-well plates,2.25×105cells/well, pretreated with siPORT NeoFX containing100nM KGF siRNA for24h. Pretreatment with a scrabbled siRNA was used as control. The conditioned medium was then replaced, and the subsequent medium for6-well plate was collected48h later for MV isolation as previously described.RNA Isolation and RT-PCR for Keratinocyte Growth Factor and Angiopoietin-1Total RNA was isolated from MV using RNeasy Mini Kit. After isolation, RNA samples were treated with DNAse I for1h at room temperature to remove contaminating DNA and stored at-80℃. The purity and content of the RNA was analyzed and quantified by measuring the absorbance ratio at260and280nm using Nanodrop1000Spectrophotometer. Primers used for the RT-PCR of human angiopoietin-1(ANGPT1), human KGF/FGF7, and human GAPDH, were purchased from QIAGEN-SABiosciences. The RT-PCR assays were conducted following the One-Step RT-PCR protocol described by QIAGEN-SABiosciences.E.coli Endotoxin Induced Acute Lung Injury in MiceC57BL/6male mice (10-12week old, Jackson Laboratory) were used in all experiments. Animals were maintained in the animal facility at the University of California, San Francisco (UCSF). All experimental protocols were approved by the Institutional Animal Care and Use Committee at UCSF.In Vivo Experimental DesignMice were first anesthetized with Ketamine (90mg/kg) and Xylazine (10mg/kg) intra-peritoneal (IP). ALI was then induced by the instillation of LPS from Escherichia coli O111:B4at4mg/kg intra-tracheally (IT). Several groups were tested:1) Phosphate buffered saline (PBS) as carrier control,2)750,000MSC in PBS as a positive cellular control,3) lx MSC MV intravenously (Ⅳ) through the tail vein, lx MSC MV IT,2x dose of MSC MV IT, KGF siRNA-pretreated MSC MV, Negative Control siRNA-pretreated MSC MV, and NHLF MV. At the end of48h, both plasma and BAL samples were collected from each mouse for assessment of neutrophil counts, cytokine and protein level measurements and histology as described below.Assessment of Neutrophil Counts, Cytokine and Protein Levels in Broncho alveolar Lavage (BAL) and PlasmaBoth BAL and plasma samples were obtained from mice at48h after endotoxin-induced lung injury. Total cell count was determined with the use of a counter and a differential of the white blood cells were obtained using Hemavet HV950FS. MIP-2and KGF were measured in both BAL and plasma samples with ELISA kits. Total protein concentration was measured in the BAL fluid from all experimental groups, as a marker of endothelial and epithelial permeability, by Pierce BCA Protein Assay kit. Extravascular Lung Water/Lung Wet-to-Dry RatioGravimetric lung water determination was done as follows:the whole lung was excised, weighed, and then homogenized after the addition of1ml of double distilled H2O (ddH2O). The homogenate was centrifuged at12,000rpm for10min to obtain the supernatant, which was weighed; the hemoglobin of the supernatant was also measured. A blood sample was obtained by needle puncture of the right ventricle, and the wet weight, hemoglobin, and hematocrit of the blood sample was obtained. All samples were placed in a drying oven at55℃for48h, and the dry weights were subsequently determined. The final excess lung water and lung wet-to-dry ratio were calculated as described in previous publications.HistologyLungs from LPS and hMSC derived MV and NHLF derived MV treated groups were excised48h after endotoxin-induced lung injury and fixed with100%ethanol IT under20cm of H2O pressure. After fixation, lungs were embedded in paraffin, cut into5μm sections, and stained with H&E.Co-culture of LPS-stimulated RA W264.7cells and MicrovesiclesTo determine whether MV is able to regulate production of pro-inflammatory cytokines released by LPS-stimulated macrophages like MSC themselves, macrophages were co-cultured either with MSC MV in a standard24-well plate or with MSC in a Transwell plates in the presence of LPS. RAW264.7cells were cultured with DMEM with high glucose containing10%FBS,1%L-glutamine and1%penicillin-streptomycin and seeded at a concentration of5×10cells/well in a24-well plate after mechanically scrapped and resuspended in medium. MSC MV (30μl) or MSC (250,000cells in the upper chamber) were then co-incubated with RAW264.7cells in the presence of LPS (500ng/ml) in serum starved conditioned medium (DMEM without FBS) at37℃for different period of time (6h,12h,24h respectively). Control wells were prepared with only RAW264.7cells in conditioned medium without LPS stimulation. The cell culture supernatants were then collected at each time point after co-incubation to assay for the levels of IL-10, TNF-α and MIP-2with ELISA kits. Each independent experiment was performed four times.Statistical AnalysisComparisons between two groups were made using an unpaired t test. For comparisons between multiple groups, ANOVA with Bonferonni correction was used. A value of p <0.05was considered statistically significant. Analyses were done using SPSS software and GraphPad Prism software. Data are shown as mean±SD.ResultsIsolation and Characterization ofMesenchymal Stem Cell MicrovesiclesUsing scanning electron microscopy, microvesicles, approximately200nM in size, were released from MSC in culture and appeared homogeneous as circular fragments. By RT-PCR, MSC MV expressed the mRNA for several key MSC paracrine factors, specifically angiopoietin-1and keratinocyte growth factor (KGF).Effect of Mesenchymal Stem Cell Microvesicles on Inflammatory Cell Influx into the Endotoxin-Injured AlveolusIntra-tracheal (IT) endotoxin produced a robust inflammatory response in the alveolus. Histology and evaluation of the BAL fluid of the endotoxin-injured mice revealed an increase in influx of white blood cells, specifically neutrophils, an elevation of the inflammatory cytokine MIP-2and an increase in protein levels suggesting lung protein permeability and pulmonary edema at48h. There were very minimal systemic effects of endotoxin IT. Simultaneous administration of30μl of MSC MV reduced the influx of white blood cells by36%, of neutrophils by73%, the elevation of MIP-2by49%and the increase in protein levels by35%as compared to the endotoxin injured lung. Interestingly, the MSC MV had a very similar response as to MSC themselves. NHLF MV at the same dose had no therapeutic effect.Dose Response Effect of Mesenchymal Stem Cell Microvesicles on Endotoxin-Induced Acute Lung InjuryDoubling the dose of MSC MV given IT simultaneously with the endotoxin had a minimal dose response effect in reducing the influx of neutrophils and the elevation of MIP-2levels in the BAL fluid at48h.Effect of KGF siRNA Pretreatment of Mesenchymal Stem Cell on the Therapeutic Effect of MicrovesiclesTo determine the therapeutic importance of KGF mRNA expression in the MSC MV, we eliminated the mRNA level using siRNA pretreatment of the MSC prior to the collection of the conditioned medium. Administration of MV from KGF siRNA pretreated MSC partially eliminated the therapeutic effect of the MV. As compared to MV isolated from a negative control siRNA pretreated MSC, KGF siRNA MSC MV increased the influx of neutrophils by47%and the elevation of MIP-2levels by56%in the BAL fluid of the endotoxin-injured lungs. Therapeutic Effect of Mesenchymal Stem Cells Microvesicles on Extravascular Lung WaterThe simultaneous administration of MSC MV (30μl) IT reduced the extravascular lung water (EVLW) of mice injured with E.coli endotoxin induced ALI by45%at48h. This effect was similar to mice with endotoxin induced ALI treated with MSC. Pretreatment of MSC with KGF siRNA prior to the collection of the MV eliminated the therapeutic effect of MV on EVLW as compared to MV collected from MSC pretreated with a negative control siRNA.Co-culture of RAW264.7cells and Microvesicles Reduced Level of Macrophage Stimulation by LPS and Increased Anti-inflammatory CytokinesThe addition of MSC MV to LPS-stimulated RAW264.7cells reduced the levels of TNF-a and MIP-2at all the time points (6h,12h and24h) compared with only LPS-stimulated macrophages. The degree of reduction was similar in the presence of a Transwell that prevented cell contact between RAW264.7and MSC. Interestingly, IL-10was significantly higher in both MSC and MSC MV groups at6h and remained higher at12h. By24h, IL-10levels were not different between MSC or MSC MV-treated groups and only LPS-stimulated groups.Conclusions1. The intra-tracheal or intravenous administration of MV derived from human MSC decreased the influx of inflammatory cells, both total white blood cells and neutrophils, and MIP-2cytokine and total protein levels in the alveolus of mice injured with E.coli endotoxin at48h.2. The administration of MV were associated with a reduction in extravascular lung water, a measure of pulmonary edema, in endotoxin-injured mice at48h.3. The therapeutic effects of MSC MV were similar to MSC themselves.4. MV expressed the mRNA for KGF, a MSC secreted paracrine factor previously found to restore alveolar fluid clearance in both E.coli bacteria and endotoxin induced ALI in an ex vivo perfused human lung.5. In the alveolar fluid of endotoxin-injured mice treated with either intra-tracheal or intravenous MV, the KGF protein levels was elevated.6. KGF siRNA-pretreatment of MSC partially eliminated the therapeutic effect of MV released by MSC on both inflammation and extravascular lung water, suggesting that the therapeutic benefit with MV appears to be in part through the transfer of KGF mRNA to the injured alveolar epithelium and lung endothelium and the expression of the protein.