Editor’s Perspective What We Already Know about This Topic What This Article Tells Us That Is New Background Human umbilical cord mesenchymal stromal cells possess considerable therapeutic promise for acute respiratory distress syndrome. Umbilical cord mesenchymal stromal cells may exert therapeutic effects via extracellular vesicles, while priming umbilical cord mesenchymal stromal cells may further enhance their effect. The authors investigated whether interferon-γ–primed umbilical cord mesenchymal stromal cells would generate mesenchymal stromal cell–derived extracellular vesicles with enhanced effects in Escherichia coli (E. coli) pneumonia. Methods In a university laboratory, anesthetized adult male Sprague–Dawley rats (n = 8 to 18 per group) underwent intrapulmonary E. coli instillation (5 × 109 colony forming units per kilogram), and were randomized to receive (a) primed mesenchymal stromal cell–derived extracellular vesicles, (b) naïve mesenchymal stromal cell–derived extracellular vesicles (both 100 million mesenchymal stromal cell–derived extracellular vesicles per kilogram), or (c) vehicle. Injury severity and bacterial load were assessed at 48 h. In vitro studies assessed the potential for primed and naïve mesenchymal stromal cell–derived extracellular vesicles to enhance macrophage bacterial phagocytosis and killing. Results Survival increased with primed (10 of 11 [91%]) and naïve (8 of 8 [100%]) mesenchymal stromal cell–derived extracellular vesicles compared with vehicle (12 of 18 [66.7%], P = 0.038). Primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles reduced alveolar–arterial oxygen gradient (422 ± 104, 536 ± 58, 523 ± 68 mm Hg, respectively; P = 0.008), reduced alveolar protein leak (0.7 ± 0.3, 1.4 ± 0.4, 1.5 ± 0.7 mg/ml, respectively; P = 0.003), increased lung mononuclear phagocytes (23.2 ± 6.3, 21.7 ± 5, 16.7 ± 5 respectively; P = 0.025), and reduced alveolar tumor necrosis factor alpha concentrations (29 ± 14.5, 35 ± 12.3, 47.2 ± 6.3 pg/ml, respectively; P = 0.026) compared with vehicle. Primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles enhanced endothelial nitric oxide synthase production in the injured lung (endothelial nitric oxide synthase/β-actin = 0.77 ± 0.34, 0.25 ± 0.29, 0.21 ± 0.33, respectively; P = 0.005). Both primed and naïve mesenchymal stromal cell–derived extracellular vesicles enhanced E. coli phagocytosis and bacterial killing in human acute monocytic leukemia cell line (THP-1) in vitro (36.9 ± 4, 13.3 ± 8, 0.1 ± 0.01%, respectively; P = 0.0004) compared with vehicle. Conclusions Extracellular vesicles from interferon-γ–primed human umbilical cord mesenchymal stromal cells more effectively attenuated E. coli–induced lung injury compared with extracellular vesicles from naïve mesenchymal stromal cells, potentially via enhanced macrophage phagocytosis and killing of E. coli.

Human mesenchymal stem/stromal cells (MSCs) have been reported to produce an M2-like, alternatively activated phenotype in macrophages. In addition, MSCs mediate effective bacterial clearance in pre-clinical sepsis models. Thus, MSCs have a paradoxical antimicrobial and anti-inflammatory response that is not understood.Here, we studied the phenotypic and functional response of monocyte-derived human macrophages to MSC exposure MSCs induced two distinct, coexistent phenotypes: M2-like macrophages (generally elongated morphology, CD163, acute phagosomal acidification, low NOX2 expression and limited phagosomal superoxide production) and M1-like macrophages characterised by high levels of phagosomal superoxide production. Enhanced phagosomal reactive oxygen species production was also observed in alveolar macrophages from a rodent model of pneumonia-induced sepsis. The production of M1-like macrophages was dependent on prostaglandin E and phosphatidylinositol 3-kinase. MSCs enhanced human macrophage phagocytosis of unopsonised bacteria and enhanced bacterial killing compared with untreated macrophages. Bacterial killing was significantly reduced by blockade of NOX2 using diphenyleneiodonium, suggesting that M1-like cells are primarily responsible for this effect. MSCs also enhanced phagocytosis and polarisation of M1-like macrophages derived from patients with severe sepsis.The enhanced antimicrobial capacity (M1-like) and inflammation resolving phenotype (M2-like) may account for the paradoxical effect of these cells in sepsis .

Enhancing the immunomodulatory effects of mesenchymal stromal cells (MSCs) may increase their effects in sepsis. We tested the potential for overexpression of Interleukin-10 (IL-10) in human umbilical cord (UC) MSCs to increase MSC efficacy in Escherichia coli (E. coli) pneumosepsis and to enhance human macrophage function. Pneumonia was induced in rats by intratracheal instillation of E. coli ((2.0–3.0) × 109 Colony forming units (CFU)/kg). One hour later, animals were randomized to receive (a) vehicle; (b) naïve UC-MSCs; or (c) IL-10 overexpressing UC-MSCs (1 × 107 cells/kg). Lung injury severity, cellular infiltration, and E. coli colony counts were assessed after 48 h. The effects and mechanisms of action of IL-10 UC-MSCs on macrophage function in septic rodents and in humans were subsequently assessed. Survival increased with IL-10 (9/11 (82%)) and naïve (11/12 (91%)) UC-MSCs compared to vehicle (9/15 (60%, p = 0.03). IL-10 UC-MSCs—but not naïve UC-MSCs—significantly decreased the alveolar arterial gradient (455 ± 93 and 520 ± 81, mmHg, respectively) compared to that of vehicle animals (544 ± 52, p = 0.02). Lung tissue bacterial counts were significantly increased in vehicle- and naïve-UC-MSC-treated animals but were not different from sham animals in those treated with IL-10 overexpressing UC-MSCs. IL-10 (but not naïve) UC-MSCs decreased alveolar neutrophils and increased alveolar macrophage percentages compared to vehicle. IL-10 UC-MSCs decreased structural lung injury compared to naïve UC-MSC or vehicle therapy. Alveolar macrophages from IL-10-UC-MSC-treated rats and from human volunteers demonstrated enhanced phagocytic capacity. This was mediated via increased macrophage hemeoxygenase-1, an effect blocked by prostaglandin E2 and lipoxygenase A4 blockade. IL-10 overexpression in UC-MSCs enhanced their effects in E. coli pneumosepsis and increased macrophage function. IL-10 UC-MSCs similarly enhanced human macrophage function, illustrating their therapeutic potential for infection-induced acute respiratory distress syndrome (ARDS).