Extracellular Vesicles from Interferon-γ–primed Human Umbilical Cord Mesenchymal Stromal Cells Reduce Escherichia coli–induced Acute Lung Injury in Rats

Acute respiratory distress syndrome (ARDS) is an inflammatory lung injury that causes life threatening acute hypoxic respiratory failure and 40% mortality.¹  No therapeutics have proven efficacious for ARDS. Mesenchymal stem/stromal cells are multipotent adult stem cells with the capacity to differentiate into multiple cell types such as osteoblasts, adipocytes, and chondroblasts. Mesenchymal stromal cells are present in multiple tissues, including the bone marrow and adipose tissue, and are found in the perivascular space in several tissues, including the umbilical cord, an easily accessible biologic waste product that generates high amounts of mesenchymal stromal cells. We have recently demonstrated that human umbilical cord–derived mesenchymal stromal cells attenuate Escherichia coli (E. coli)–induced acute lung injury.² 

Mesenchymal stromal cells possess immunomodulatory properties that have shown substantial therapeutic promise in preclinical models.^2–6  Key mechanisms of action of mesenchymal stromal cells include a decreased alveolar-capillary barrier permeability, enhanced alveolar fluid clearance, enhanced macrophage function, and enhanced bacterial clearance and mitochondrial transfer.^2–10  The immune modulatory effects of mesenchymal stromal cells are mediated via both cell-to-cell contact–dependent and independent (i.e., paracrine) mechanisms.⁹ 

Mesenchymal stromal cells release extracellular vesicles, which are small, spherical membrane fragments comprising distinct populations such as exosomes, microvesicle particles, and apoptotic bodies. The smaller exosomes (50 to 150 nm in size) arise from intracellular endosomes, whereas larger microvesicles (150 to 1,000 nm in size) originate directly from the plasma membrane.¹¹  Mesenchymal stromal cell–derived extracellular vesicles include both exosomes and microvesicles, and demonstrate therapeutic effects in several disease models including cardiovascular,¹²  renal,¹³  liver,¹⁴  and lung diseases.¹⁵ 

Key mesenchymal stromal cells effects are mediated via modulation of the host immune response to tissue injury and/or microbial infection.⁹  Mesenchymal stromal cells are activated or primed by the injury/infection microenvironment, which may be replicated by previous activation of the mesenchymal stromal cells.¹⁶  Interferon γ is a major proinflammatory cytokine secreted by activated T and natural killer cells. Interferon γ upregulates immune-related genes in mesenchymal stromal cells, including major histocompatibility complex and costimulatory molecules CD80 and CD86 and transforming growth factor β, hepatocyte growth factor, and indoleamine 2,3, dioxygenase.¹⁷  Accordingly, we postulated priming mesenchymal stromal cells in vitro with interferon γ might enhance the immunomodulatory function of mesenchymal stromal cell–derived extracellular vesicles.^18,19 

This study investigated the therapeutic effect of mesenchymal stromal cell–derived extracellular vesicles derived from interferon γ–primed or naïve human umbilical cord mesenchymal stromal cells in a rat model of E. coli–induced ARDS. We hypothesized that extracellular vesicles derived from mesenchymal stromal cells primed with interferon γ would enhance macrophage phagocytosis and killing of E. coli bacteria, and would more effectively attenuate E. coli–induced lung injury compared with extracellular vesicles from naïve mesenchymal stromal cells. We assessed the impact of E. coli pneumonia and mesenchymal stromal cell–derived extracellular vesicle therapy on pulmonary endothelial nitric oxide synthase,²⁰  nicotinamide adenine dinucleotide phosphate oxidase,²¹  and high-mobility group box 1 protein,²²  given their previously reported effect on capillary vascular permeability. Some of the results in this manuscript have been presented in abstract form.²³ 

Human umbilical cord–derived mesenchymal stromal cells were isolated from human umbilical cords obtained from full-term, consenting donors undergoing caesarean delivery at Mount Sinai Hospital, Toronto, Canada, according to a protocol approved by research ethics boards at both the University of Toronto and Mount Sinai Hospital’s Research Center for Women’s and Infants’ Health. Human acute monocytic leukemia cell line THP-1 (American Type Culture Collection, ATCC, Manassas, VA) was cultured in RPMI-1640 containing 2 mM l-glutamine and 10 mM N-2-Hydroxyethylpiperazine-N’-2 ethanesulfonic acid supplemented with 10% fetal bovine serum and 0.05 mM 2-mercaptoethanol. THP-1 cells were differentiated into macrophages with 100 nM phorbol 12-myristate 13-acetate for 24 h and allowed to rest in standard medium for 24 h.²⁴ 

Extracellular vesicles were isolated from the conditioned medium of interferon γ–primed or naïve human umbilical cord mesenchymal stromal cells (primed or naïve mesenchymal stromal cell–derived extracellular vesicles, respectively). Mesenchymal stromal cells were primed by incubation with 50 ng/ml interferon γ for 8 h; this approach was chosen based on pilot dose response studies in our laboratory and on previous reports.²⁵  Mesenchymal stromal cell–derived extracellular vesicles were isolated as previously described.²⁶  Briefly, conditioned medium from mesenchymal stromal cells was subjected to centrifugations at 300 and 2,000 g (10 min each) to remove floating cells and cell debris. The harvested supernatant was then centrifuged at 100,000 g (Beckman Coulter Optima L-100XP Ultracentrifuge) at 4°C for 90 min and the pellet was suspended in 100 μl of phosphate buffered saline. These centrifugation steps have been demonstrated to produce a pure extracellular vesicle fraction and to remove debris.²⁶ 

The therapeutic dose of mesenchymal stromal cell–derived extracellular vesicles of 100 million extracellular vesicles per kilogram, an amount derived from 35 to 40 million mesenchymal stromal cells, was chosen during pilot studies.²  The number of mesenchymal stromal cell–derived extracellular vesicles was first determined by flow cytometry using a FACSAria III SORP with small particle detection modifications.²⁷  Briefly, mesenchymal stromal cell–derived extracellular vesicles were reconstituted in phosphate buffered saline followed by applying a certain volume (100 μl) for flow cytometry. Extracellular vesicles were detected and quantified using Side Scatter Pulse Height based on a scatter gate defined by small particle beads (Apogee, United Kingdom). Subsequently, the protein content of aliquots of 3 to 4 × 10⁷ mesenchymal stromal cell–derived extracellular vesicles (determined by flow cytometry) from naïve and interferon γ–activated mesenchymal stromal cells were then assessed by Bio-Rad protein Assay (USA), and were found to be similar (fig. 1A). For subsequent mesenchymal stromal cell–derived extracellular vesicle doses, protein content was used as an index of mesenchymal stromal cell–derived extracellular vesicle content, to facilitate animal dosing. This allowed the real-time estimation of mesenchymal stromal cell–derived extracellular vesicle dose immediately before administration, an important factor in facilitating clinical studies in due course.

Mesenchymal stromal cell–derived extracellular vesicle samples were mixed with paraformaldehyde (final concentration: 2%) at room temperature. Transmission electron microscopy samples (5 μl) were mounted on a 400 mesh formvar/carbon-coated copper grid and fixed with 1% gluteraldehyde in cold Dulbecco’s phosphate buffered saline for 5 min. Grids were then washed with sterile water (greater than 15 Milli-Q.cm) and allowed to dry. Micrographs (64,000× or 92,000×) were obtained using a Philips CM10 transmission electron microscope with a side-mounted AMT HR 11 megapixel camera. Mesenchymal stromal cell–derived extracellular vesicles images were calibrated by their scale bars and characterized using ImageJ software (n > 100 each).²⁸ 

All work was approved by the Animal Care and Use Committee of the Keenan Research Center for Biomedical Science of St. Michael’s Hospital, Toronto, and conducted under license from Health Canada. Specific pathogen–free adult male Sprague–Dawley rats (Charles River Laboratories, Canada) were used in all experiments. Animals were anesthetized by inhalational induction with isoflurane and intraperitoneal 40 mg/kg ketamine (Pfizer, United Kingdom). After confirmation of depth of anesthesia by paw clamp, intravenous access was obtained via tail vein; animals were intubated and received approximately 5 × 10⁹E. coli bacteria (E5162, serotype O9 K30 H10) colony forming units per kilogram, intratracheally.^5,29 

Thirty minutes following E. coli instillation, animals were randomized into three treatment groups and underwent intravenous administration of: (1) interferon γ–primed umbilical cord–derived mesenchymal stromal cells (primed mesenchymal stromal cell–derived extracellular vesicles, n = 11 animals); (2) naïve umbilical cord–derived mesenchymal stromal cells (naïve mesenchymal stromal cell–derived extracellular vesicles, n = 8 animals); or (3) an equal volume of phosphate buffered saline (vehicle) (n = 18 animals).

At 48 h after E. coli instillation, animals were reanesthetized as described above; intravenous access was established via tail vein, and anesthesia was maintained with alfaxalone (alfaxadone 0.9% and alfadadolone acetate 0.3%). A tracheostomy tube was inserted and intra-arterial access was sited in the carotid artery. Muscle relaxation was induced with cisatracurium besylate, and the lungs were mechanically ventilated using a small animal ventilator (CWE SAR 830 AP; CWE Inc., USA) with 30% O₂ in 70% N₂ at a respiratory rate of 80 min^–1, tidal volume 6 ml, and positive end-expiratory pressure 2 cm H₂O. Systemic arterial blood pressure and peak airway pressure were continually measured. Lung static compliance and arterial blood gas analysis were measured after 20 min and were repeated on 100% O₂ after 15 min. Animals were then humanely euthanized by exsanguination under anesthesia. Bronchoalveolar lavage fluid differential, leukocyte counts, and lung bacterial colony counts were completed. Bronchoalveolar lavage concentrations of tumor necrosis factor α and monocyte chemotactic protein 1 were determined using enzyme-linked immunosorbent assay (R&D Systems, USA, and MyBioSource, USA), and bronchoalveolar lavage protein concentrations were measured (Bio-Rad protein Assay). The left lung was isolated, fixed, and used for histologic studies.

Western blots were performed on lung tissues as described.²  The primary antibodies for endothelial nitric oxide synthase, 1:1,000 (mouse immunoglobulin 2a), and Gp91phox (Nox-2), 1:1,000 (mouse immunoglobulin 1) from BD Biosciences (Canada) and high-mobility group box 1 protein 1:1,000 (mouse immunoglobulin 2b) from BioLegend (USA) were used. Signals were detected using an ECL-Plus kit (Amersham Biosciences, USA). Band intensities were quantified and expressed relative to that of β-actin.

Hydrogen peroxide levels were measured in lung tissue using an Amplex Red assay kit (Molecular Probes, USA) in accordance with the manufacturer’s instructions. Fluorescence was quantified using 544-nm excitation and 590-nm emission, and levels were normalized for protein content.

The left lung lobe was fixed in 10% formalin neutral buffer (Sigma-Aldrich, Canada), cut into five transversal sections, placed in the histo-processing cassettes, and processed overnight using Leica TP1020 Automatic Tissue Processor. The H&E staining of the slides was done using the Leica Autostainer. A Nikon Upright E800 microscope was used for microscopy and using Image J 1.47v software (Wayne Rasband, National Institutes of Health, USA), the percentage of airspace and acinar tissue were calculated in each slide in five randomly selected animals from each group and three sham rats. For each rat, 30 randomly selected high powered fields were assessed. The assessors were blinded to group allocation.

Activated THP1 cells (250,000 cells per well) were treated with mesenchymal stromal cell–derived extracellular vesicles (72 h) then exposed to ampicillin-resistant red fluorescent protein–expressing E. coli (DH5a) bacteria (5 × 10⁶).⁵  Plates were spun (1,500 rpm; 60 sec) then placed at 37°C for 20 min. Cells were then washed twice with phosphate buffered saline and incubated with Gentamycin (Sigma-Aldrich) for 30 min. Cells were then lysed with 1% Triton X-100 (Sigma-Aldrich) and Tris/NaCl/EDTA (TNE) lysis buffer (50 mM Tris HCl pH 7.4, 100 mM NaCl, and 1 mM EDTA) and lysates containing phagocytosed bacteria were cultured on ampicillin containing Petri dishes (Fisher Scientific, USA). To determine bacterial killing after phagocytosis, in a separate experiment, after exposure to gentamycin cells were washed (3×) and returned to the incubator with fresh medium for 1 h. As above, cells were then lysed and lysates cultured in the presence of ampicillin. The next day bacterial colonies were counted and the percentage killing was expressed by dividing the number of colonies in plate no. 2 by the number in plate no. 1 and expressing as a percentage.³⁰ 

No formal statistical power calculation was conducted. However, from our previous experience with this model, at least eight surviving animals were required in each group to determine between group differences in alveolar-arterial oxygen gradient in our three-group design.^2,4,5  We anticipated a mortality in the range of 20% to 30% in untreated animals based on pilot studies, but the mortality rate in the mesenchymal stromal cell–derived extracellular vesicle treated animals was unknown. To minimize the overall use of animals in these experiments, while ensuring sufficient animals for assessment of injury, we randomly allocated eight animals to each group, and decided a priori to replace nonsurviving animals on a 2:1 basis.

All data were analyzed using GraphPad Prism (GraphPad Software, USA). The distribution of all data was tested for normality using Kolmogorov–Smirnov tests. Underlying model assumptions were deemed appropriate on the basis of suitable residual plots. Survival was analyzed using the log-rank test. Data were analyzed by one-way ANOVA, with post hoc testing using Bonferroni multiple comparison test. A two-way ANOVA, with group and time as factors (nonrepeated measure), and post hoc testing using Bonferroni multiple comparison test was used to assess macrophage phagocytosis. A two-tailed P value of less than 0.05 was considered statistically significant. Following application of Chauvenet’s criterion, a single outlying data point, in the data for alveolar–arterial oxygen gradient was identified and removed (Supplemental Digital Content 1, https://links.lww.com/ALN/B882). Data are presented as mean ± SD except where otherwise indicated in the figure legend.

Thirty to 40 million mesenchymal stromal cell–derived extracellular vesicles were isolated from each batch of conditioned medium from 3 × 10⁷ primed and naïve mesenchymal stromal cells, respectively. Both flow cytometric (fig. 1B) and transmission electron microscopy (fig. 1C) analyses demonstrate a wide distribution of sizes of mesenchymal stromal cell–derived extracellular vesicles from naïve and primed mesenchymal stromal cells. The mean size of mesenchymal stromal cell–derived extracellular vesicles from interferon γ–primed mesenchymal stromal cells was larger than those from naïve mesenchymal stromal cells (71.8 nm ± 15.7 nm vs. 47.7 nm ± 25.2 nm; fig. 1, D and E).

In total, 37 animals underwent E. coli–induced lung injury, with 11 animals receiving interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles, eight animals receiving naïve mesenchymal stromal cell–derived extracellular vesicles, and 18 animals receiving vehicle therapy.

Treatment with primed (10 of 11 [91%]) and naïve (8 of 8 [100%]) mesenchymal stromal cell–derived extracellular vesicles enhanced survival in our E. coli–induced rat model of ARDS, as compared with animals treated with phosphate buffered saline (12 of 18 [66.7%]; P = 0.038; fig. 2A).

In surviving animals, primed mesenchymal stromal cell–derived extracellular vesicles (n = 10)—but not naïve mesenchymal stromal cell–derived extracellular vesicles (n = 8)—significantly reduced alveolar-arterial oxygen gradient (fig. 2B), and attenuated the increase in alveolar permeability, as evidenced by reduced alveolar protein concentrations (fig. 2C), compared with vehicle therapy (n = 12). There was no significant effect of primed or naïve mesenchymal stromal cell–derived extracellular vesicles on alveolar E. coli colony forming units (fig. 2D).

There was no significant effect of mesenchymal stromal cell–derived extracellular vesicles on the overall number of leukocytes in the alveolar space in response to E. coli (fig. 3A). In contrast, primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles increased the percentage of mononuclear phagocytes in alveolar fluid (fig. 3B). Mesenchymal stromal cell–derived extracellular vesicles did not alter the percentage of neutrophils in the bronchoalveolar lavage fluid (fig. 3C). Primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles significantly attenuated the E. coli induced proinflammatory cytokine response, as evidenced by reduced alveolar tumor necrosis factor α concentrations (fig. 3D). There was no significant effect of primed and naïve mesenchymal stromal cell–derived extracellular vesicles on bronchoalveolar lavage monocyte chemotactic protein 1 concentrations (Supplemental Digital Content 2, https://links.lww.com/ALN/B881).

Representative images demonstrate severe E. coli–induced lung injury evidenced by decrease in airspace, increase in alveolar tissue, and alveolar wall thickening, compared with sham animals (fig. 4, A and B). Mesenchymal stromal cell–derived extracellular vesicles—both naïve and primed—reduced structural damage (fig. 4, C and D). Stereologic analysis of lung sections demonstrates that primed mesenchymal stromal cell–derived extracellular vesicles were significantly more effective than naïve mesenchymal stromal cell–derived extracellular vesicles in restoring lung structure after E. coli–induced lung injury (fig. 4, E and F).

Primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles restored pulmonary endothelial nitric oxide synthase protein concentrations to levels comparable to that seen in sham animals (fig. 5, A and B). There were no significant differences in the expression of high-mobility group box 1 protein or nicotinamide adenine dinucleotide phosphate oxidase between the treatment groups (Supplemental Digital Content 3, https://links.lww.com/ALN/B883).

Interferon γ–primed—but not naïve—mesenchymal stromal cell–derived extracellular vesicles significantly enhanced bacterial phagocytosis in activated THP1 cells compared with both naïve mesenchymal stromal cell–derived extracellular vesicle or vehicle treated cells (fig. 6A). Although both naïve and primed mesenchymal stromal cell–derived extracellular vesicles enhanced macrophage killing of phagocytosed E. coli, primed mesenchymal stromal cell–derived extracellular vesicles significantly enhanced E. coli killing compared with naïve mesenchymal stromal cell–derived extracellular vesicles (fig. 6B). Representative confocal micrographs of THP1 cells incubated with fluorescently-labeled E. coli demonstrate enhanced phagocytosis of E. coli by THP1 cells treated with primed mesenchymal stromal cell–derived extracellular vesicles (fig. 6C).

These studies provide a number of novel and important insights. First, our findings demonstrate the therapeutic potential of mesenchymal stromal cell–derived extracellular vesicles derived from the umbilical cord, a more plentiful and easily accessed source of mesenchymal stromal cells, in a clinically relevant model of E. coli–induced pneumonia. Both naïve and interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles reduced mortality in severe sepsis attributable to pulmonary E. coli instillation. Second, we demonstrate the potential to further enhance the effect of mesenchymal stromal cell–derived extracellular vesicles by previous mesenchymal stromal cell priming with interferon γ. Interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles were more effective than naïve mesenchymal stromal cell–derived extracellular vesicles in reducing E. coli–induced lung injury, decreasing alveolar–arterial oxygen gradient and alveolar protein leak. Third, we provide mechanistic insights into the actions of mesenchymal stromal cell–derived extracellular vesicles, namely enhancement of macrophage phagocytosis and killing of bacteria and restoration of endothelial nitric oxide synthase, which may restore capillary endothelial barrier function. Taken together, our demonstration that interferon γ priming of human umbilical cord–derived mesenchymal stromal cells can produce extracellular vesicles with enhanced therapeutic effect in a clinically relevant infection model suggests that this approach deserves further exploration as a potential therapy for ARDS.

There is considerable interest in the potential for mesenchymal stromal cell–derived extracellular vesicles to reproduce the effects of mesenchymal stromal cells, while avoiding any disadvantages of whole cell administration. Extracellular vesicles express surface molecules, such as CD29, CD73, CD44, and CD105, which are characteristic of mesenchymal stromal cells.³¹  They contain bioactive molecules, such as proteins, mRNAs, microRNAs, and lipids, which may mediate the cross-talk between mesenchymal stromal cells and immune cells including macrophages, potentially reprogramming these cells to enhance their effects.

The extracellular vesicle population is composed of distinct fractions, which include the larger microvesicles, (150 to 1,000 nm in size) that originate directly from the plasma membrane,^11,32  and smaller exosomes (50 to 150 nm in size) that arise from intracellular endosomes. Some studies have demonstrated beneficial effects with these subpopulations in isolation.^33,34  However, should the full extracellular vesicle fraction, which combines both populations, prove efficacious, this would make it more feasible to manufacture and isolate this subcellular population, expediting translation to clinical testing. We observed that the mean size of mesenchymal stromal cell–derived extracellular vesicles from interferon γ–primed cells was larger than those from naïve mesenchymal stromal cells. The significance of this observation is unclear, but it necessitates further investigations to better elucidate the structure of primed versus naïve mesenchymal stromal cell–derived extracellular vesicles.

This study provides further insights into the therapeutic potential of mesenchymal stromal cell–derived extracellular vesicles. Umbilical cord–derived mesenchymal stromal cell–derived extracellular vesicles attenuated E. coli–induced lung injury in our in vivo rodent model. Both naïve and primed mesenchymal stromal cell–derived extracellular vesicles demonstrated benefits, enhancing animal survival after E. coli pneumonia. Naïve mesenchymal stromal cell–derived extracellular vesicles also reduced histologic lung injury. Although the effect of naïve mesenchymal stromal cell–derived extracellular vesicles on other physiologic indices of lung injury in surviving animals was less marked, this may reflect survival bias given that one-third of vehicle-treated animals did not survive the protocol. In contrast, interferon γ priming of the umbilical cord–derived mesenchymal stromal cells further enhanced the beneficial effects of their extracellular vesicles in this model, further decreasing physiologic and histologic indices of E. coli–induced lung injury.

Previous studies suggest benefit for bone marrow–derived mesenchymal stromal cell–derived extracellular vesicles in other disease models, including cardiovascular,¹²  renal,¹³  liver,¹⁴  and lung diseases.^15,35  Bone marrow mesenchymal stromal cell–derived microvesicles attenuated murine E. coli–induced pneumonia, potentially via a mechanism involving enhancement of monocyte/macrophage phagocytosis of bacteria.¹⁵  Bone marrow–derived microvesicles attenuated endotoxin-induced lung injury, decreasing alveolar white blood cell influx by 36% and neutrophil influx by 73%.³⁶  Both intrapulmonary and intravenous injection of microvesicles resulted in comparable effect in this model.³⁶ 

Most studies to date have used microvesicles or extracellular vesicles from bone marrow–derived mesenchymal stromal cells. However, the bone marrow requires an invasive biopsy to obtain tissue, and the yield is relatively low. These issues are compounded when one considers the high numbers of mesenchymal stromal cells that would be required in humans. The umbilical cord is a more clinically feasible mesenchymal stromal cell (and extracellular vesicle) source, given their ease of isolation, higher proliferation, and greater self-renewal capacity than bone marrow–derived mesenchymal stromal cells, which have been used in most studies to date. Demonstrating an effect of extracellular vesicles from umbilical cord–derived mesenchymal stromal cells is an important advance in this regard.

The current study provides further insights into the mechanisms of action of mesenchymal stromal cell–derived extracellular vesicles. Specifically, the beneficial effects of mesenchymal stromal cell–derived extracellular vesicles appear to be mediated via enhancement of macrophage E. coli phagocytosis and killing. These findings are supported by previous studies from our group and others, suggesting that the interaction between the mesenchymal stromal cell–derived extracellular vesicle and the macrophage is central to its mechanism of action. Mesenchymal stromal cell–derived microvesicles³⁷  and extracellular vesicles³⁸  appear to induce a switch in macrophages from M1- to M2-like phenotype in nonseptic preclinical rodent models of thioglycollate-induced peritonitis³⁷  and streptozotocin-induced diabetes.³⁹  The primary mechanism of action of mesenchymal stromal cell–derived extracellular vesicles may vary depending on the major injurious mechanism. In a severe burn rat model, mesenchymal stromal cell–derived extracellular vesicles decreased proinflammatory cytokines (tumor necrosis factor α and interleukin 1β) while increasing interleukin 10, mediated via exosomal microRNA 181-c inhibition of the toll-like receptor 4 signaling pathway.⁴⁰ 

We demonstrate that interferon γ priming enhances the effect of extracellular vesicles on macrophage bacterial phagocytosis, a key step in the host response to invading pathogens. Priming also further enhances macrophage killing of phagocytosed E. coli. The potential for interferon γ to prime mesenchymal stromal cells is underlined by its potential to upregulate immune-related genes, including major histocompatibility complex and costimulatory molecules CD80 and CD86 and transforming growth factor beta, hepatocyte growth factor, indoleamine 2,3, dioxygenase in mesenchymal stromal cells.⁴¹  These findings are consistent with an emerging body of evidence suggesting that priming of mesenchymal stromal cells can enhance their effect. Microvesicles derived from toll-like receptor 3 agonist (Poly I:C)–primed bone marrow mesenchymal stromal cell have also demonstrated enhanced bacterial killing both in vitro and in vivo.³⁶ 

Taken together, these effects of both naïve and interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles are consistent with recently reported findings from our group demonstrating that mesenchymal stromal cells also enhance M1-like macrophage functions, particularly macrophage phagocytosis and antimicrobial killing capacity,³⁰  in addition to its better recognized effects on M2-like macrophages. Further studies are needed to determine whether interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles may enhance macrophage bacterial phagocytosis and killing via enhanced induction of the M1-like phenotype.

Mesenchymal stromal cell–derived extracellular vesicles may also attenuate the proinflammatory cytokine response to bacterial infection. Tumor necrosis factor α may stimulate the generation of reactive oxygen species by immune cells, which additionally leads to tissue damage.⁴²  In this study, mesenchymal stromal cell–derived extracellular vesicles attenuated alveolar concentrations of tumor necrosis factor α produced in response to E. coli infection. These findings extend previous reports demonstrating that mesenchymal stromal cell–derived microvesicles can attenuate human monocyte and alveolar epithelial type 2 cell production of tumor necrosis factor α,^15,36  and reduce alveolar tumor necrosis factor α induced by E. coli instillation into the lung.²  In the current study, we observed that the concentration of tumor necrosis factor α was significantly lower in bronchoalveolar lavage of animals treated with primed mesenchymal stromal cell–derived extracellular vesicles in comparison with those treated with naïve mesenchymal stromal cell–derived extracellular vesicles or vehicle, indicating that primed extracellular vesicles more effectively attenuate the proinflammatory cytokine response. Primed mesenchymal stromal cell–derived extracellular vesicles also increased the percentage of macrophages in the lung leukocyte infiltrate.

Endothelial nitric oxide synthase is a key producer of nitric oxide,⁴³  a protective regulatory element of inflammatory-immune processes, which is also implicated in regulating lung permeability.²⁰  Primed human mesenchymal stromal cell–derived extracellular vesicles restored lung tissue endothelial nitric oxide synthase to levels comparable with that seen in uninjured animals, whereas unprimed mesenchymal stromal cell–derived extracellular vesicles did not have this effect. These findings are consistent with a previous demonstration that mesenchymal stromal cells increased expression of endothelial nitric oxide synthase in pulmonary microvascular endothelial cells in vitro, and in the rat lung after endotoxin induced injury.⁴⁴  Mesenchymal stromal cell–conditioned medium also restored vasoreactivity in human umbilical vein endothelial cells with impaired vasculogenesis attributable to pharmacologic nitric oxide synthase inhibition.⁴⁵  Although the mechanism by which mesenchymal stromal cell–derived extracellular vesicles enhance endothelial nitric oxide synthase expression remains to be elucidated, heat shock protein 90 is abundant in mesenchymal stromal cell–derived extracellular vesicles⁴⁶  and can bind to endothelial nitric oxide synthase and influence its calcium sensitivity and phosphorylation state.⁴⁷ 

The precise mechanisms by which mesenchymal stromal cell priming with interferon γ alters the structure or content of extracellular vesicles deserves further investigation. The protein content of extracellular vesicles derived from primed mesenchymal stromal cells was similar to those from naïve mesenchymal stromal cells. Although the mean size of mesenchymal stromal cell–derived extracellular vesicles from interferon γ–primed cells was larger than those derived from naïve mesenchymal stromal cells, the significance of this observation remains unclear.

These studies provide insights into the therapeutic potential of interferon γ–primed mesenchymal stromal cell–derived extracellular vesicles, and suggest that additional studies are warranted. However, there are some limitations to these studies that should be considered. First, although these studies were conducted in relevant preclinical models of E. coli–induced acute lung injury, caution is required in extrapolating to the clinical condition of ARDS. Second, we studied a single dose and administration route of mesenchymal stromal cell–derived extracellular vesicles. This was based on pilot dose–response studies demonstrating that this was an effective dose of mesenchymal stromal cell–derived extracellular vesicles in these models of E. coli lung injury. We used the intravenous route of administration, based on previous studies from our group and others demonstrating that this route is as at least as effective as other, more invasive routes, such as intrapulmonary or intraperitoneal administration.^4,6  Additional dose–response studies, and studies using differing routes of administration, would provide further insights regarding these mesenchymal stromal cell–derived extracellular vesicles. Furthermore, characterizing the features and content of extracellular vesicles after priming mesenchymal stromal cells with interferon γ will be essential to understand their underlying therapeutic mechanisms.

In conclusion, we report that interferon γ priming of human umbilical cord–derived mesenchymal stromal cells generated mesenchymal stromal cell–derived extracellular vesicles with enhanced capacity to attenuate E. coli–induced lung injury compared with extracellular vesicles from naïve mesenchymal stromal cells. A key mechanism of action appears to be mediated via enhancement of macrophage phagocytosis and macrophage killing of E. coli. Overall, the results of this study support the idea of application of mesenchymal stromal cell–derived extracellular vesicles as an alternative treatment for ARDS.

The cells used in these studies were provided free of charge by Tissue Regeneration Therapeutics Ltd, Toronto, Canada.

Support for this study was provided by an operating grant from the Canadian Institutes for Health Research (CIHR) and by the Keenan Research Center for Biomedical Sciences at St. Michael’s Hospital, Toronto, Canada (to J.G.L. and G.F.C.).

The authors declare no competing interests.