Human mesenchymal stem cell microvesicles for treatment of Escherichia coli endotoxin-induced acute lung injury in mice

Abstract

We previously found that human mesenchymal stem cells (MSC) or its conditioned medium restored lung protein permeability and reduced alveolar inflammation following Escherichia coli endotoxin-induced acute lung injury (ALI) in an ex vivo perfused human lung in part through the secretion of soluble factors such as keratinocyte growth factor (KGF). Recently, MSC were found to release microvesicles (MVs) that were biologically active because of the presence of mRNA or miRNA with reparative properties. MVs are circular fragments of membrane released from the endosomal compartment as exosomes or shed from the surface membranes. These studies were designed to determine if MVs released by human bone marrow derived MSCs would be effective in restoring lung protein permeability and reducing inflammation in E. coli endotoxin-induced ALI in C57BL/6 mice. The intratracheal instillation of MVs improved several indices of ALI at 48 hours. Compared to endotoxin-injured mice, MVs reduced extravascular lung water by 43% and reduced total protein levels in the bronchoalveolar lavage (BAL) fluid by 35%, demonstrating a reduction in pulmonary edema and lung protein permeability. MVs also reduced the influx of neutrophils and macrophage inflammatory protein-2 levels in the BAL fluid by 73% and 49%, respectively, demonstrating a reduction in inflammation. KGF siRNA-pretreatment of MSC partially eliminated the therapeutic effects of MVs released by MSCs, suggesting that KGF protein expression was important for the underlying mechanism. In summary, human MSC-derived MVs were therapeutically effective following E. coli endotoxin-induced ALI in mice in part through the expression of KGF mRNA in the injured alveolus.

Keywords: Acute lung injury; Keratinocyte growth factor; Lipopolysaccharide; Mesenchymal stem cell; Microvesicles.

© 2013 AlphaMed Press.

Figures

FIGURE 1. Electron Microscopy Images of MVs Released by MSCs and MSC MV mRNA Content

Electron microscopy demonstrates that MVs are released by MSCs in vitro following stress such as serum starvation. (A) MVs released into the inter-cellular gap separating two MSCs; bar is 2 μm. Enclosed image shows purified MVs, which appears to be a collection of homogeneous spheroids; bar is 0.5 μm. (B) RT-PCR demonstrated that MSC MV expressed the mRNAs for KGF and Ang1, two secreted proteins previously found to be involved in the therapeutic effect of MSCs in ALI.

FIGURE 2. Effect of MSC MVs on Influx of Inflammatory Cells in Endotoxin-Induced ALI in Mice

The administration of MSC MVs reduced the influx of inflammatory cells in endotoxin-induced ALI in mice. (A) IT MSC MVs improved lung injury as assessed by histology. H&E staining of lung sections at 48 h demonstrated a reduction in inflammatory cell influx, edema, blood and thickening of the interstitium in endotoxin-injured lungs treated with MSC MVs. The MVs derived from NHLF showed no therapeutic benefit on lung injury. Scale bars, 200 μm. The IT administration of MSC MVs decreased the total white blood cells (WBC) in the BAL fluid of endotoxin-injured mice. Data is shown as mean ± SD, N = 4 for NHLF treated, N = 14 for PBS treated and N = 19-20 for endotoxin or endotoxin + MSC MV treated. *, P < 0.0003 vs. PBS treated mice by ANOVA (Bonferroni). (B) More significantly, IT administration of MSC MVs reduced the influx of neutrophils into the BAL fluid of endotoxin-injured mice. Absolute neutrophil counts are shown as mean ± SD. *, P < 0.002 vs. PBS treated and †, P < 0.002 vs. endotoxin treated mice by ANOVA (Bonferroni).

FIGURE 3. Effect of Intra-tracheal MSC MVs on Inflammation in Endotoxin-Induced ALI in Mice

IT administration of MSC MVs reduced the level of inflammation and protein permeability in the alveolus of endotoxin-injured mice. (A) IT MSC MVs decreased the level of MIP-2 in the BAL fluid of endotoxin-injured mice. Data are expressed as mean ± SD, N = 14-15. *, P < 0.006 vs. PBS treated and †, P < 0.002 vs. endotoxin treated mice by ANOVA (Bonferroni). (B) IT MSC MVs decreased the total protein level in the BAL fluid of endotoxin-injured mice. Data are shown as mean ± SD, N = 14-16. *, P < 0.003 vs. PBS treated by ANOVA (Bonferroni).

FIGURE 4. Dose Response of MSC MVs on Endotoxin-Induced ALI in Mice

Doubling the dose of MSC MVs had no additional anti-inflammatory effect in endotoxin-injured mice. (A) MSC MVs, 1× or 2×, had a similar response to MSCs, the cells themselves, in reducing the influx of neutrophils into endotoxin-injured mice at 48 h. Data are shown as mean ± SD, N = 13-14 per MSC or MSC MV (2×), N = 29 for MSC MV (1×) and N = 36 for endotoxin. *, P < 0.003 vs. endotoxin treated mice by ANOVA (Bonferroni). (B) Doubling the dose of MSC MVs had no additional effect in reducing the MIP-2 level in the BAL fluid of endotoxin-injured mice. Data are shown as mean ± SD, N = 9-13 per MSC or MSC MV (2×), N = 24 for MSC MV (1×) and N = 32 for endotoxin. *, P < 0.002 vs. endotoxin treated mice by ANOVA (Bonferroni).

FIGURE 5. Expression of KGF Protein from MSC MVs

The administration of MSC MVs increased the secretion of KGF protein in the BAL fluid of endotoxin-injured mice and in the conditioned medium of human alveolar epithelial type II cells injured with cytomix. (A) IT or IV administration of MSC MVs increased the levels of KGF protein in endotoxin-injured mice at 48 h, similar to the level of MSCs. Data are expressed as mean ± SD, N = 4-9 for treatment groups and N = 12-13 for PBS or endotoxin treated mice. *, P < 0.0002 vs. PBS treated and †, P < 0.002 vs. endotoxin treated mice by ANOVA (Bonferroni). (B) KGF protein levels was also elevated in the conditioned medium of primary cultures of human alveolar type II cells injured with cytomix and exposed to 100 μl of MSC MVs in the upper chamber. Data are expressed as mean ± SD, N = 3.

FIGURE 6. Effect of KGF siRNA Pretreatment of MSCs on the Therapeutic Effect of Secreted MSC MVs

KGF siRNA pretreatment of MSC eliminated much of therapeutic effects of MVs released by MSC. (A) RT-PCR confirmed that KGF mRNA in MSC MVs was eliminated by KGF siRNA pretreatment of MSCs for 24 h. Red arrow, location of KGF band by size. GAPDH was used as a housekeeping gene and was expressed by both KGF siRNA-pretreated MSC MVs and control MSC MVs. (B) The IT administration of MVs released from KGF siRNA pretreated MSCs significantly increased the influx of neutrophils and the elevation of MIP-2 level in the BAL fluid of endotoxin-injured lungs compared to administration of MVs from Neg Control siRNA pretreated MSCs, demonstrating a partial loss of therapeutic effect of the MVs. Data are expressed as mean ± SD, N = 14-15 per treatment group, *, P < 0.05 vs. MV isolated from a Neg control siRNA pretreated MSC for neutrophil count; *, P < 0.02 vs. MV isolated from a Neg control siRNA pretreated MSC for MIP-2 level. (C) The administration of MSC MVs had a similar effect as the cells themselves in reducing EVLW following endotoxin-induced ALI. Pretreatment of the MSCs with KGF siRNA eliminated the therapeutic effect of the MSC MVs on EVLW as compared to Neg control siRNA-pretreated MSC MVs. Data are expressed as mean ± SD, N = 9-10 per treatment groups, N = 21 for endotoxin. *, P < 0.002 and is significant by ANOVA (Bonferroni) vs. endotoxin injured mice; N = 10, *, P < 0.03 for KGF siRNA vs. Neg Control siRNA Pretreated MSC MV.

FIGURE 7. Effect of MSC MVs on RAW 264.7 Cells

Co-culture of RAW 264.7 cells with MSCs or MSC MVs following exposure to endotoxin decreased inflammatory cytokine/chemokine secretion and increased the anti-inflammatory cytokine IL-10 level. (A & B) The simultaneous addition of MSCs or MSC MVs significantly reduced the levels of both TNFα and MIP-2 in the conditioned medium at all the time points (6 h, 12 h and 24 h) compared with RAW 264.7 cells exposed to endotoxin. For TNFα: Data are expressed as mean ± SD, N = 4. *, P < 0.0001 at 6 h, *, P < 0.002 at 12 h and *, P < 0.0001 at 24 h vs. endotoxin by ANOVA (Bonferroni). For MIP-2: Data are expressed as mean ± SD, N = 4. *, P < 0.0002 at 6 h, *, P < 0.005 at 12 h and *, P < 0.008 at 24 h vs. endotoxin by ANOVA (Bonferroni). (C) The simultaneous addition of MSCs or MSC MVs with RAW 264.7 cells significantly increased the level of IL-10 in the conditioned medium at 6 h and remained higher at 12 h for MSC MVs treated cells compared to cells exposed to endotoxin alone. By 24 h, IL-10 levels returned to baseline for all groups. Data are expressed as mean ± SD, N = 4. *, P < 0.002 vs. endotoxin at 6 h and *, P < 0.002 at 12 h vs. endotoxin exposed cells by ANOVA (Bonferroni).