Extracellular Vesicles Derived from Gram-Negative Bacteria, such as Escherichia coli, Induce Emphysema Mainly via IL-17A–Mediated Neutrophilic Inflammation

This article is featured in In This Issue, p.2957

Biological contaminants in indoor air can induce immune dysregulation in the lung, resulting in inflammatory pulmonary disorders, such as asthma and chronic obstructive pulmonary disease (COPD) (1). In terms of the immunopathogenesis of airway inflammation, the eosinophilic subtype represents inflammation induced by IL-4– and IL-13–secreting Th2 cells, whereas the neutrophilic subtype is related to both IFN-γ– and IL-17–dependent Th1 and Th17 inflammation (2–4). Neutrophilic inflammation is known to be important in the COPD pathogenesis (5). Emphysema (a key phenotype of COPD) is an important cause of irreversible airflow limitation, because the destruction of lung tissue around small airways seen in emphysema patients prevents the airways from holding their functional shape upon exhalation (6, 7).

Extracellular vesicles (EVs) are spherical bilayered phospholipids ranging in size from 20 to 200 nm in diameter, so-called nanovesicles, that are produced ubiquitously from all Gram-negative bacteria and some Gram-positive bacteria investigated to date (8, 9). Previous biochemical and proteomic studies have revealed that Gram-negative bacteria–derived EVs are composed of outer membrane proteins, LPS, outer membrane lipids, periplasmic proteins, DNA, RNA, and other factors associated with virulence (8). The fact that bacteria-derived EVs harbor pathogen-associated molecular patterns (PAMPs), such as LPS, peptidoglycan (PG), lipoteichoic acid (LTA), that induce innate immunity, and also harbor proteins that induce T cell response, led to us the notion that the inhalation of bacterial EVs can evoke immune dysregulation and inflammation in the lung.

Escherichia coli is one of the most important model organisms of Enterobacteriaceae. Our previous work has shown that E. coli–derived EVs induce systemic inflammation mimicking sepsis after entering the bloodstream (10) and that EVs from indoor dust and also from Staphylococcus aureus can induce neutrophilic pulmonary inflammation (11, 12). In the current study, we hypothesized that EVs from Gram-negative bacteria, especially E. coli, are an important cause of neutrophilic inflammation and thereby emphysema in the lung. To test this, we aimed to evaluate whether E. coli EVs are causally related to the pathogenesis of emphysema. We also tried to determine the immunologic mechanisms of emphysema induced by E. coli EVs.

Wild-type (WT), IFN-γ–deficient, and TNF-α–deficient mice (12, 13) (all C57BL/6 background) were purchased from The Jackson Laboratory (Bar Harbor, ME). IL-17A–deficient mice (C57BL/6 background) were gifted from Y.-C. Sung (Pohang University of Science and Technology [POSTECH], Pohang, Republic of Korea) (12). TLR2-deficient and TLR4-deficient mice (12) (both of C57BL/6 background) were purchased from Oriental BioService (Kyoto, Japan). These mice were bred in pathogen-free facilities at POSTECH. The animal experimental protocols were approved by the Institutional Animal Care and Use Committee at POSTECH (permit number 2011-01-0021). MH-S cells (mouse alveolar macrophages originated from BALB/c; American Type Culture Collection) were cultured in RPMI 1640 medium supplemented with 10% FBS and grown in the presence of 100 U/ml penicillin and 100 μg/ml streptomycin.

Indoor dust was collected from a bed mattress in an apartment using a vacuum cleaner (Samsung, Seoul, Republic of Korea). The dust was stored at room temperature until further use (11).

EVs were isolated from indoor dust and from the culture supernatants of E. coli isolated from the peritoneal lavage fluid of mice with cecal ligation and puncture, as described previously (10, 11). Briefly, the dust samples were incubated in PBS for 12 h at 4°C with stirring and centrifuged twice at 10,000 × g for 15 min. E. coli in lysogeny broth was cultured at 37°C and centrifuged twice at 10,000 × g for 15 min. Supernatants were filtered with a 0.22-μm vacuum filter to remove any remaining cells. EVs were prepared by centrifugation in a 45 Ti rotor (Beckman Instruments, Fullerton, CA) at 150,000 × g for 3 h at 4°C. EVs were diluted in PBS and stored at −80°C.

To confirm the role of LPS in E. coli–derived EV-induced cytokine production, MH-S cells were incubated with E. coli–derived EVs (100 ng/ml) in the presence or the absence of polymyxin B (PMB; 1 μg/ml). To examine in vitro innate immune responses induced by E. coli–derived EVs in WT, TLR2-deficient, and TLR4-deficient mice, peritoneal macrophages were isolated as described previously (12). Briefly, thioglycollate-elicited peritoneal macrophages were seeded in 24-well plates, incubated with E. coli–derived EVs, TLR2 agonist (heat-killed Listeria monocytogenes, 10⁷ cells/ml), or TLR4 agonist (LPS from E. coli K12), and the culture supernatants were collected 16 h after the application. The levels of TNF-α and IL-6 in culture supernatants were measured by ELISA.

To evaluate the effects of E. coli–derived EVs on the development of lung inflammation, E. coli–derived EVs were administered intranasally to WT C57BL/6, IFN-γ–deficient, IL-17A–deficient, or TNF-α–deficient mice. Lung inflammation was evaluated 24 h after the final application. PBS was applied as a negative control. To inhibit LPS activity, PMB (1 μg) was instilled intranasally with 100 ng E. coli–derived EVs.

Cellularity in bronchoalveolar lavage (BAL) fluid was analyzed, as described previously (1). Briefly, after counting the total number of cells in the BAL samples, the cell were stained by Diff-Quik solution and were classified as macrophages, lymphocytes, neutrophils, or eosinophils. To detect emphysema, lungs were fixed in buffered 4% formaldehyde and embedded in paraffin. Sections were stained with H&E. Chord length was measured using the Computer Assisted Stereological Toolbox system (Olympus, Tokyo, Japan). In H&E-stained lung sections, the average interalveolar septal wall distance (mean linear intercept) was measured separately by two people in a blinded fashion (14).

One microgram of E. coli EVs was administered intranasally to C57BL/6 mice. After 1 h, mice were sacrificed, and lungs were removed. Lungs were fixed with 1% paraformaldehyde in PBS containing 1% sucrose, blocked with 5% horse serum in TBST (0.3% Triton X-100 in TBS), and incubated with anti-E. coli EV polyclonal Ab, which had been developed in our laboratory by immunizing rabbits with the E. coli EVs, as previously described with minor modifications (15), and anti–SP-C Ab (Santa Cruz Biotechnology, Santa Cruz, CA). After treatment with Alexa-Fluor 488–conjugated donkey anti-rabbit IgG and Alexa-Fluor 594–conjugated donkey anti-goat IgG (Invitrogen, Carlsbad, CA), lungs were counterstained with Hoechst (Sigma-Aldrich, St. Louis, MO) and analyzed using an FV1000 laser scanning confocal microscope (Olympus).

MH-S cells were incubated with 1 μg/ml DiO (Invitrogen)-labeled E. coli EVs in the presence or the absence of 10 μg/ml PMB for 6 h. Cells were then washed with PBS, fixed with 4% paraformaldehyde in PBS containing 4% sucrose, permeabilized with 0.2% Triton X-100 in PBS, stained with Hoechst (Sigma-Aldrich), and analyzed using an FV1000 laser scanning confocal microscope (Olympus). To examine the role of TLRs in the uptake of E. coli–derived EVs, the above protocol was applied to peritoneal macrophages isolated from WT, TLR2-deficient, and TLR4-deficient mice.

The elastase activity from lung lysates without detergent was measured using an Elastase assay kit according to the manufacturer’s instructions (Invitrogen).

Proteins from soluble and EV fractions isolated from indoor dust were separated using SDS–PAGE and transferred to polyvinylidene difluoride membranes. Blocked membranes were incubated with an E. coli EV polyclonal Ab.

A method of RT-PCR using dust EVs to detect E. coli EV 16S rRNA has been described previously (16). The primers for E. coli EV 16S rRNA were 5′-GGAAGAAGCTTGCTTCTTTGCTG-3′ and 5′-AGCCCGGGGATTTCACATCTGACTTA-3′. The measurement of matrix metalloproteinases (MMPs) mRNA expression were performed by RT-PCR described previously (16, 17). The real-time quantitative analysis of MMPs was performed using the LightCycler 480 system (Roche Diagnostics, Indianapolis, IN) with LightCycler 480 SYBR Green I master (Roche Diagnostics).

Cells isolated from regional lymph nodes were restimulated with plate-bound anti-CD3 and CD28 Abs for 6 h. The culture supernatant was collected and examined by ELISA. Cytokine levels from BAL fluid were measured by ELISA, according to the manufacturer’s instructions (R&D Systems, Minneapolis, MN).

For multiple comparisons among groups, ANOVA was initially used followed by unpaired t or nonparametric Mann–Whitney U test between two groups. To test for trends, the ANOVA linearity test was used. For statistical analysis, we used GraphPad Prism 5, and the statistical significance was set a priori at p < 0.05.

E. coli is a Gram-negative bacterium that is not confined to the intestine, and its ability to survive for brief periods outside the body makes it an ideal indicator organism to test indoor dust samples for fecal contamination (18). In our previous report, indoor dust harbors lots of EVs (11). Thus, we evaluated whether indoor dust contains E. coli–derived EVs. The presence of E. coli EVs in indoor dust was assessed using an anti-E. coli EV polyclonal Ab. The present data showed that the Ab reacted with EVs in indoor dust but not soluble factors derived from indoor dust (Fig. 1A). In addition, genotyping using an E. coli–specific 16S rRNA primer showed that E. coli EVs were present in indoor dust (Fig. 1B). Furthermore, it was determined that E. coli EVs are taken up by entering the airways via resident lung cells. After the airway administration of E. coli–derived EVs, whole-mount imaging of lung tissue labeled with an E. coli EV Ab indicated the uptake of E. coli EVs into resident lung cells, including alveolar type II cells (Fig. 1C). To evaluate innate immune response induced by E. coli EVs, EVs were administered intranasally into the mouse airways in differing doses of 1, 10, and 100 ng. Following the administration of E. coli EVs, total cell number in BAL fluid was increased by EVs in a dose-dependent manner (Fig. 1D). In mice receiving 10 and 100 ng E. coli EVs, the production of TNF-α and IL-6 in BAL fluid also increased within 8 h after the administration (Fig. 1E).

To evaluate whether the repeated inhalation of E. coli EVs induces neutrophilic inflammation and thereby structural changes in the lung, E. coli EVs (1, 10, and 100 ng) were administered into the mouse airways twice per week for 4 wk. The lung infiltration of neutrophils was enhanced by the inhalation of E. coli EVs in a dose-dependent manner (Fig. 2A). As for structural changes, the inhalation of E. coli EVs induced emphysema in a dose-dependent manner (Fig. 2B). In addition, the production of IFN-γ–inducible protein-10 and IL-6 (downstream mediators of IFN-γ and IL-17, respectively) was also significantly increased by airway exposure to E. coli EVs in a dose-dependent manner (Fig. 2C). The airway administration of E. coli EVs also dose-dependently enhanced the production of both IFN-γ and IL-17A by lung T cells (Fig. 2D). Moreover, E. coli EVs enhanced the expression of MMP-2 and MMP-9 mRNA in lung tissues (Fig 2E). Taken together, these observations indicate that E. coli EVs can induce neutrophilic inflammation and emphysema in a dose-dependent manner.

We evaluated the role of IFN-γ and IL-17A in the development of emphysema after airway exposure to E. coli EVs (100 ng) twice per week for 4 wk in IFN-γ–deficient and IL-17A–deficient mice, and WT control mice. The present study showed that the lung infiltration of neutrophils induced by E. coli EVs was significantly inhibited by the absence of IL-17A but not by the absence of IFN-γ (Fig. 3A). In contrast, emphysematous changes induced by E. coli EVs were partially abolished by the absence of IFN-γ or IL-17 (Fig. 3B). Remarkably, the protein amount of IFN-γ in BAL fluid were absent in IL-17–deficient mice, whereas the BAL IL-17A levels were not inhibited by the absence of IFN-γ (Fig. 3C). In addition, the production of IFN-γ from regional lymph node cells restimulated with anti-CD3 and CD28 Abs was also inhibited by the absence of IL-17A, but IL-17A production was not diminished by the IFN-γ absence (Fig. 4D). These findings suggest that IL-17A is a key mediator in the development of emphysema induced by exposure to E. coli–derived EVs and that the production of IFN-γ is dependent on IL-17A.

To evaluate the early-phase effects of E. coli EVs on the emphysema pathogenesis, 100 ng E. coli EVs were administered into the mouse airways and then evaluated 18 h after this application. This experiment showed that neutrophilic inflammation and emphysema were induced by E. coli EVs (Fig. 4A, 4B). Elastase activity was enhanced in E. coli EV–treated mice in comparison with PBS-treated mice (Fig. 4C). In addition, the mRNA expression of MMP-9, but not MMP-2, in the lung was increased by the application of E. coli EVs (Fig. 4D). Taken together, these observations indicate that E. coli EVs can induce emphysema, possibly via protease and elastase activity of neutrophils recruited by E. coli EVs.

To evaluate the role of proinflammatory cytokines (IFN-γ, IL-17A, and TNF-α) on the emphysema pathogenesis causally related to E. coli EVs, 100 ng EVs was administered into the airways of mice in which the candidate genes had been disrupted, and then phenotypes were evaluated 18 h after the application. This study showed that lung inflammation induced by E. coli EVs was reversed by the absence of IL-17A or TNF-α (Fig. 5A). The elastase activity enhanced by E. coli EVs was reversed by the absence of IFN-γ, IL-17A, or TNF-α (Fig. 5B). The mRNA expression of MMP-9 enhanced by E. coli EVs was reversed by the absence of IL-17A or TNF-α (Fig. 5C). Collectively, these findings suggest that early immune responses, such as elaboration of proinflammatory cytokines, are important in the development of emphysema via protein degradation by inflammatory cells.

E. coli EVs harbor various PAMPs, including LPS and outer membrane proteins, which can be recognized by pattern recognition receptors, such as TLRs. We tried to elucidate the role of TLR signaling in the development of lung inflammation induced by E. coli EVs. The production of cytokine by E. coli EVs was evaluated in TLR2-deficient and TLR4-deficient mice. Peritoneal macrophages were isolated from WT, TLR2-deficient, and TLR4-deficient mice and stimulated with E. coli EVs. The production of TNF-α and IL-6 was abolished completely by the absence of TLR4 but only partially by the absence of TLR2 (Fig. 6A). In addition, we evaluated the role of TLRs on the uptake of E. coli EVs by inflammatory cells. E. coli EVs were taken up by peritoneal macrophages isolated from WT mice but not from TLR4-deficient mice (Fig. 6B). Taken together, these findings suggest that TLR4 signaling is important in the uptake of E. coli EVs and the production of proinflammatory cytokines induced by E. coli EVs.

To examine the role of LPS in E. coli EVs in the development of emphysema, PMB, LPS antagonist, was treated with E. coli EVs. The uptake of E. coli EVs by MH-S cells (mouse alveolar macrophages) was inhibited by PMB treatment compared with PBS treatment (Fig. 6C). The production of TNF-α and IL-6 induced by E. coli EVs was partly abolished by PMB treatment (Fig. 6D). To assess the in vivo effect of LPS on phenotypes induced by E. coli EVs, EVs were administered into the mouse airways with or without PMB. This experiment showed that the number of inflammatory cells in BAL fluid was partly decreased by PMB treatment compared with PBS treatment (Fig. 6E). In addition, the production of IL-6 and TNF-α induced by E. coli EVs was also partly inhibited by PMB treatment (Fig. 6F). Taken together, these results suggest that LPS in E. coli EVs plays an important role in the uptake of EVs by host cells.

The role of infectious agents in the etiology of diseases once believed to be noninfectious is increasingly being recognized (19). Noxious or biological contaminants in indoor air can induce chronic inflammatory pulmonary disorders, such as asthma and COPD (20). Common biological contaminants in indoor dust include viruses, bacteria, fungi, dust mites, and pet dander (21, 22). Previous reports indicate that organic dust can induce airway inflammation related with Th1/Th17 cell responses or TLR2 signaling (23, 24). Our previous findings showed that indoor dust collected from bed mattress induces neutrophilic pulmonary inflammation and that EVs in indoor dust also induce neutrophilic pulmonary inflammation, which is related with Th1 and Th17 cell responses (25). To our knowledge, this is the first report demonstrating that EVs derived from Gram-negative bacteria, especially E. coli, can induce neutrophilic inflammation and thereby emphysema mainly in an IL-17A–dependent manner. We also demonstrate that the TLR4-signaling pathway is critical in the development of E. coli EV–induced phenotypes.

Gram-negative bacteria, including E. coli, produce EVs during normal growth (26). Gram-negative bacteria–derived EVs contain a wide variety of PAMPs, which can be capable of inducing innate immune responses and include LPS, outer membrane lipids, outer membrane proteins, periplasmic proteins, cytoplasmic proteins, DNA, RNA, and other virulence-associated factors (8, 27). Our previous study demonstrated that the entry of E. coli EVs into the bloodstream induces systemic inflammation mimicking sepsis (10). In addition, our previous work indicates that EVs in indoor dust are important in the development of neutrophilic inflammation in the lung (11). Previously, we also found that the Gram-positive bacterium S. aureus produces EVs (9), which may be causally related to atopic dermatitis in the skin (28) and also neutrophilic inflammation in the lung (12). The present study showed that the repeated inhalation of E. coli EVs induced neutrophilic inflammation and thereby emphysema mainly via an IL-17A–dependent mechanism. These findings suggest that bacteria-derived EVs may be a novel causative agent of inflammatory diseases whose agent(s) are unknown.

Our previous data demonstrated that the repeated inhalation of indoor dust can induce neutrophilic inflammation associated with enhanced infiltration of IFN-γ–positive and IL-17A–positive T cells in the lung (11). Much clinical evidence indicates that neutrophilic inflammation in the airways is related to asthma severity (1). Neutrophilic inflammation is also important in the pathogenesis of COPD (5). Most animal models of emphysema are provoked by cigarette smoke, and IFN-γ is also a key mediator of cigarette-induced emphysema (29, 30). In addition, transgenic studies have shown that high levels of IFN-γ in the airways induce not only noneosinophilic asthma but also emphysema (31, 32). The present study showed that IFN-γ is important in the development of emphysema, but not neutrophilic inflammation, induced by airway exposure to E. coli EVs and also in the elastase activity enhanced by E. coli EVs. To sum up, these findings suggest that IFN-γ is involved in the development of emphysema, possibly via elastin degradation by inflammatory cells.

Although recent evidence indicates that IL-17A contributes to the recruitment of neutrophils in the lungs (33) and to the development of asthma (3, 33, 34), the role of IL-17A in the emphysema pathogenesis remains to be determined. The present data show that both neutrophilic inflammation and emphysema induced by E. coli EVs were abolished by the absence of IL-17A. In addition, elastase activity enhanced by E. coli EVs was found to be absent in IL-17A–deficient mice. Recent data have shown that Th17 cells can also produce IFN-γ, in addition to IL-17A (35). Interestingly, the current study showed that IFN-γ production was completely abolished by the absence of IL-17A, whereas IL-17A production was unaffected by the absence of IFN-γ. Taken together, these findings imply that IL-17A is a key mediator in the development of neutrophilic inflammation and emphysema induced by E. coli EVs, in which IFN-γ appears to be produced by Th17 cells rather than Th1 cells.

LPS, a component of the outer membrane of Gram-negative bacteria, is ubiquitously present in the indoor environment and induces the production of proinflammatory and immune modulating mediators via TLR4 (36, 37). The present study showed that E. coli EV uptake by macrophages is dependent on both LPS and TLR4 signaling. In addition, lung inflammation and the production of proinflammatory mediators induced by E. coli EVs are markedly inhibited by the absence of TLR-4 and also partly inhibited by the absence of TLR-2. In addition, the uptake of E. coli EVs by macrophages was completely abolished by the absence of TLR-4 or by PMB treatment. Collectively, these findings suggest that signals other than LPS are also involved in the development of emphysema phenotypes induced by E. coli EVs; however, the interaction of LPS in E. coli EVs and the TLR4 receptor is critical in the uptake of EVs by inflammatory cells.

In summary, our present data indicate that airway exposure to EVs derived from E. coli, the most important model organism of enteric Gram-negative bacteria, can induce neutrophilic inflammation and emphysema mainly in an IL-17A–dependent manner. Moreover, the present in vitro and in vivo data show that the TLR4 receptor is important in the development of immune and pathologic phenomena induced by E. coli EVs mainly via initial interaction between LPS in E. coli EVs and the TLR4 receptor. Thus, EVs derived from Gram-negative bacteria, including E. coli, represent a novel target for the control of COPD.

The authors have no financial conflicts of interest.