IFN Regulatory Factor 3 Potentiates Emphysematous Aggravation by Lipopolysaccharide

Chronic obstructive pulmonary disease (COPD) is a progressive disease that affects ∼10% of people worldwide (1). By 2030, COPD is predicted to be the third leading cause of mortality and fifth leading cause of being disabled (2, 3). Acute exacerbation of COPD is often induced by infection with Gram-negative bacteria, which occurs frequently during the clinical course of COPD and is a major cause of mortality and decreased quality of life (4). Despite intensive treatment with antibiotics, bronchodilators, and steroids during episodes of exacerbation, respiratory function often remains worse than before the exacerbation (5, 6).

Gram-negative bacteria contain LPS, also called endotoxin, which is recognized by host innate immune receptor TLR4 and subsequently triggers inflammation (7). Downstream signaling of TLR4 consists of two major pathways: the MyD88-dependent pathway and the Toll/IL-1R domain–containing adaptor inducing IFN-β (TRIF)-dependent pathway. The MyD88 pathway activates NF-κB, resulting in the production of inflammatory mediators, such as IL-6 and TNF-α, which are involved in COPD pathogenesis (8, 9). The TRIF pathway activates IFN regulatory factor-3 (IRF3) and produces type 1 IFN and IFN-γ–inducible protein-10 (IP-10) (10, 11). Although each pathway potentiates a distinct profile of cytokines and chemokines, NF-κB subunit p35 and IRF3 cooperate as an enhanceosome (12–14), suggesting that the inflammatory mediators mainly regulated by NF-κB may also be regulated by IRF3. Although the TLR4–MyD88–NF-κB pathway is thought to have a role in COPD exacerbation (15), the contribution of the TLR4–TRIF–IRF3 pathway remains to be elucidated.

In the present study, we hypothesized that IRF3 has an essential role in the pathogenesis of LPS-induced acute exacerbation of COPD. We used a mouse model of COPD acute exacerbation in which LPS was administered after emphysema was induced by using porcine pancreatic elastase (PPE). In PPE plus LPS–treated IRF3^−/− mice, emphysema formation was attenuated, and the expression of multiple cytokines and chemokines associated with COPD in the acute inflammatory phase was suppressed. Alveolar epithelial cells, lung macrophages, and neutrophils were partially involved in these differences of expressions. Although the number of neutrophils infiltrating into the lungs of IRF3^−/− mice was increased compared with similarly treated wild-type (WT) mice, the neutrophils isolated from the inflamed lungs exhibited decreased apoptosis, phagocytosis, and autophagy. Consequently, IRF3 contributes to the pathogenesis of LPS-induced COPD exacerbation based on different production of inflammatory mediators from multiple cell types that may lead to altered neutrophil function.

Specific pathogen-free male C57BL6/N mice (8–12 wk old) were purchased from Sankyo Laboratory (Tokyo, Japan). IRF3^−/− mice (C57BL/6J background) (16) were purchased from RIKEN BioResource Center (Tsukuba, Japan), with the approval of Dr. T. Taniguchi, who originally generated this strain. The mice were housed in a specific pathogen-free environment with free access to food and water, under approval of the Institutional Animal Experiments Committee of Musashino University. Care and use of the animals followed the Principles of Laboratory Animal Care formulated by the National Society for Medical Research of Japan.

In our COPD exacerbation model, WT and IRF3^−/− mice were anesthetized with ketamine (90 mg/kg i.p.) and medetomidine (1 mg/kg i.p.); 0.5 U of PPE (Sigma-Aldrich, St. Louis, MO) was administered intranasally on day 0; and then 25 μg of LPS from Escherichia coli 0111:B4 (Sigma-Aldrich) or vehicle (PBS) was administered intranasally on days 7, 10, and 14 (Fig. 1A). In our acute inflammation model, mice intranasally received 0.5 U of PPE or 25 μg of LPS or PBS (PPE group, LPS group, and PBS group, respectively); mice in the PPE plus LPS group intranasally received 25 μg of LPS 1 wk after PPE treatment (Fig. 2A).

At weeks 0, 1, 2, 4, and 6, mice were anesthetized with ketamine (90 mg/kg i.p.) and medetomidine (1 mg/kg i.p.) and fixed in a dorsal position in a chamber for microcomputed tomographic scanning of small animals (LaTheta LCT-200; Aloka, Tokyo, Japan) (17).

After the scanning range was determined, computed tomography (CT) of the chest was performed in respiratory-gated mode. Acquired data were analyzed according to CT number in Hounsfield units (HU). First, low-attenuation areas, the volume of which reflects emphysema, were quantified as those having a CT number of −871 to −610 HU, as previously described (18). Second, the total lung volume without vasculature volume was quantified as regions having a value in the range of −871 to −250 HU and was confirmed by calculating the lung area from actual images. Third, the total intratracheal air volume was determined with the same CT number range of −871 to −250 HU. The percentage of low-attenuation areas was calculated as the ratio of the volume of low-attenuation areas to total lung volume. Because it was included in both original values, intratracheal air volume was subtracted from the low-attenuation volume and the total lung volume.

At week 6, mice were euthanized, a tracheal cannula was inserted, and the lungs were inflated with 4% paraformaldehyde for fixation, paraffin embedded, and cut in 5-μm sections. To histopathologically assess emphysema formation, we stained the lung sections with H&E and calculated the mean linear intercept, as previously described (19, 20).

In our acute inflammation model, mice were euthanized on the day after their last treatment. To obtain bronchoalveolar lavage fluid (BALF), the trachea was catheterized, and the lungs were washed with 1 ml of sterile saline three times. The isolated BALF was centrifuged at 450 × g for 10 min at 4°C, pellets were dissolved in 1 ml of PBS, and the cells were counted. Differential cell counts were performed on cytospin preparations stained with Diff-Quik (Sysmex, Hyogo, Japan) according to the manufacturer’s instructions. The methods for total RNA extraction from harvested lungs, cDNA synthesis, and quantitative PCR have been previously described (20). The expression of genes of interest was calculated according to the comparative threshold cycle method, using β-actin as an internal control.

Naive macrophages were isolated from BALF as previously described (21), and alveolar epithelial cells were isolated as previously described (22, 23), with some modifications. Briefly, after lungs were perfused with PBS, 0.1% Dispase II (Wako Pure Chemical Industries, Osaka, Japan) in HBSS was instilled via a catheter inserted in the trachea. Then lungs were incubated in the Dispase solution for 45 min at room temperature and then cut into small pieces in DMEM containing 0.01% DNase (Sigma-Aldrich). The cell suspensions were filtered through cell strainers (Corning, Corning, NY), centrifuged at 150 × g for 10 min at 4°C, and washed with 0.01% DNase in DMEM. Cells were then purified using centrifugation on a discontinuous Percoll gradient, followed by incubation in Primaria (Corning) tissue culture plates at 37°C for 12 h. Nonattached cells were used for culture. The purity of alveolar epithelial cells was confirmed by immunostaining for pro-surfactant protein C.

Macrophages were cultured in DMEM with 5% FBS and 1% penicillin/streptomycin. The next day, cells were simultaneously treated with filtered 8 mU/ml PPE in PBS and 20 ng/ml LPS. After 24 h, RNA was harvested from these cells and used for quantitative PCR analysis. Alveolar epithelial cells were cultured in bronchial epithelial cell growth medium (Lonza, Walkersville, MD) supplemented with 5% FBS and 1% penicillin/streptomycin for 4–6 d to ∼90% confluence. After cells were treated for 24 h with 8 mU/ml PPE plus 20 ng/ml LPS (PPE plus LPS), RNA was harvested from these cells.

For the experiments using lung neutrophils and macrophages, lungs were perfused with 3 ml of PBS via the right ventricle, and both lobes were pooled for cell sorting. The lobes were cut into small pieces and digested in HBSS supplemented with 0.2% collagenase (Wako), 0.1% Dispase II, and 0.025% DNase for 60 min at 37°C. Cells were passed through a 40-μm cell strainer and washed with PBS. RBCs were lysed with lysis buffer (BD Biosciences, San Diego, CA), and the remaining cells were suspended in sorting buffer (PBS with 2% FBS and 1% penicillin/streptomycin). Then cells were incubated with biotinylated anti–Gr-1 Ab and allophycocyanin-conjugated anti-CD11c Ab (eBioscience, San Diego, CA) before incubation with streptavidin-PE or streptavidin-PerCP-Cy5.5 (eBioscience) for an apoptosis assay; Gr-1^highCD11c^low cells were sorted (SH800 cell sorter; Sony, Tokyo, Japan) as lung neutrophils, and Gr-1^intCD11c^high cells were sorted as lung macrophages (24). Dead cells were identified by propidium iodide (PI; Sigma-Aldrich) staining and excluded from our analyses except for an apoptosis assay. Cell purity was confirmed by cytospin preparations stained with Diff-Quik (Sysmex). For RNA extraction, ∼4 × 10⁵ neutrophils for each sample were used immediately after sorting.

Lung neutrophils (2.5 × 10⁵ cells per tube) that had not been affected by PI were incubated with FITC-labeled annexin V and PI in an apoptosis detection kit (Takara, Tokyo, Japan) or with MitoPT in a MitoPT apoptosis detection kit (ImmunoChemistry Technologies, Bloomington, MN) and analyzed by flow cytometry (SH800 cell sorter; Sony). Data were analyzed according to the manufacturer’s instructions. Data were expressed as the difference between test samples and controls without annexin V.

Sorted lung neutrophils or macrophages (2 × 10⁵ cells per tube) were washed with PBS, suspended in sorting buffer, and incubated with carboxylate-modified, fluorescent yellow-green latex beads (Sigma-Aldrich) at 37°C, under 5% CO₂ for 2 h. After cells were washed with PBS and suspended in sorting buffer, phagocytosing cells were detected as FITC⁺ cells by flow cytometry. Data were calculated as the difference in fluorescence intensity of the cells with or without beads. In some experiments, cytochalasin D (Sigma-Aldrich) at final concentration of 20 μg/ml was used as an inhibitor for phagocytosis. Additionally, cytospin preparations underwent nuclear staining by using mounting medium containing DAPI (Vectashield, Burlingame, CA) according to the manufacturer’s instructions, followed by microscopy (FSX100 Bio Imaging Navigator; Olympus, Tokyo, Japan) for analysis of phagocytosing cells.

Sorted lung neutrophils or macrophages (2 × 10⁶ cells/ml) were washed and suspended in HBSS. For 2 h at 37°C, under 5% CO₂, half of the cells were incubated in starving condition (in HBSS without supplement) alone, and the remaining cells were incubated in starving condition with chloroquine, an inhibitor of lysosomal acidification. To prepare whole-cell protein extracts for immunoblotting, cells were washed and lysed in cell lysis buffer (Medical and Biological Laboratories, Aichi, Japan). After incubation, partial cells were prepared for immunocytochemistry via cytospin. Autophagic flux was assessed by the difference in immunoblotting of autophagosome protein LC3-II expression in the presence and absence of chloroquine (25).

A DC protein assay kit (Bio-Rad Laboratories, Hercules, CA) was used to quantify neutrophil and macrophage proteins in LDS sample buffer (Life Technologies, Carlsbad, CA). A Multigel II mini 10/20 (Cosmo Bio, Tokyo, Japan) was used for electrophoresis. For immunoblotting, 10 μg of neutrophil or macrophage protein was used. After electrophoresis, proteins were transferred to a polyvinylidene difluoride membrane; membranes were blocked with with ECL prime blocking reagent (GE Healthcare, Buckinghamshire, U.K.) and incubated with anti-LC3b Ab (Abcam, Cambridge, U.K.) or anti–β-actin Ab (control; Cell Signaling Technology, Beverly, MA) as the primary Ab and HRP-conjugated anti-rabbit IgG Ab (GE Healthcare) as the secondary Ab. Bands were detected by chemiluminescence, and band intensities were measured by using ImageQuant TL software (GE Healthcare).

For immunocytochemistry, cytospin preparations of lung neutrophils treated with or without chloroquine were fixed with 4% paraformaldehyde. Cells were treated with permeabilization buffer (BD Biosciences), blocked with 3% FBS in PBS for 1 h, and incubated with anti-LC3 mAb (Medical and Biological Laboratories) as the primary Ab for 1 h and FITC-conjugated anti-mouse IgG1 (eBioscience) as the secondary Ab for 1 h. Nuclei were stained and cells were evaluated as for the phagocytosis assay.

Cigarette smoke extract (CSE) was prepared based on the method previously described (26). Briefly, smoke from four Marlboro cigarettes (Philip Morris, Victoria, Australia) was bubbled into 10 ml of PBS for 2 min per cigarette. Each cigarette contained 1.0 mg of nicotine, 12 mg of tar. Crude extract was filter sterilized using a 0.22-μm filter and frozen in aliquots until use. Alveolar macrophages and alveolar epithelial cells were incubated in medium containing 2% CSE for 24 h and then supernatants and mRNA were harvested. For induction of apoptosis on neutrophils, cells were incubated in medium containing 8 or 16% CSE for 1 h and then used for the measurement. For measurement of TNF-α in culture supernatants, a Quantikine mouse TNF-α ELISA kit (R&D Systems, Minneapolis, MN) was used according to the manufacturer’s instructions.

A Student t test was used for comparison of two groups, whereas ANOVA and a Tukey–Kramer honestly significant difference test were used for comparison among multiple groups (JMP Pro 11.2.0; SAS Institute, Cary, NC). The level of statistical significance was set at p < 0.05. Data were expressed as the mean ± SEM.

Using chest microcomputed tomography (micro-CT), we examined the morphologic changes of the lung. To quantify the volume of emphysematous changes, the ratio of the volume of low-attenuation regions to total lung volume was calculated as the percentage of low-attenuation regions according to acquired images (Fig. 1B). Compared with those in WT PPE plus LPS mice, emphysematous changes in the IRF3^−/− PPE plus LPS group were significantly (p < 0.05) suppressed at 4 wk (Fig. 1C). Next, we analyzed the change in percentage of low attenuation area after LPS administration. In the WT group, LPS administration induced a significant increase in PPE-induced emphysema formation only at 4 and 6 wk (Fig. 1D). Histopathological examination at 6 wk showed enlarged alveolar spaces with destruction of alveolar septae in both the WT and IRF3^−/− PPE plus LPS groups; however, according to our quantitative examination, the mean linear intercept of the IRF3^−/− PPE plus LPS group was significantly lower than that of the WT group (Fig. 1E, 1F). These data suggest that LPS administration exacerbates the PPE-induced emphysematous change in the lung in an IRF3-dependent manner.

To examine the mechanisms underlying the distinct morphologic changes in the lung, we assessed the cell, cytokine, and chemokine profiles during acute inflammation induced by PPE plus LPS. The stimulation protocol is shown in Fig. 2A. Compared with WT mice, the PPE plus LPS group of IRF3^−/− mice exhibited suppressed mRNA expression of MCP-1, TNF-α, MIP-1α, IP-10, and keratinocyte chemoattractant. Among these mediators, IP-10 expression in the PPE plus LPS group of WT mice was also higher than that in the PPE or LPS only WT group. In contrast, a significant difference of MIP-2 expression was not detected, and the concentration of IL-17A transcripts was higher in IRF3^−/− mice than in WT mice (Fig. 2B, 2C).

Alternatively, compared with those in the PPE plus LPS group of WT mice, total cell and neutrophil counts were significantly increased in the BALF of the PPE plus LPS group of IRF3^−/− mice (Fig. 3A). The numbers of total cells and neutrophils were greater in the PPE, LPS, and PPE plus LPS groups than in the PBS group. It is notable that LPS-induced augmentation of the number of total cells and neutrophils in BALF after PPE treatment was detected only in IRF3^−/− mice (Fig. 3B). However, the augmentation was observed in macrophages both in WT and IRF3^−/− mice. These data suggest that various mediators related to COPD were produced by PPE plus LPS treatment, and augmentation by LPS after PPE treatment was observed in IP-10 expression and cell number of macrophages. In IRF3^−/− mice, multiple mediators were suppressed, resulting in the attenuated morphologic changes in these mice. Controversially, neutrophil recruitment to lung was somewhat increased during acute inflammation.

To further investigate pathogenic roles of IRF3 in this exacerbation model, we isolated alveolar epithelial cells and alveolar or lung macrophages to examine their function. As for alveolar epithelial cells treated by PPE plus LPS in IRF3^−/− mice, the expression of MCP-1 was significantly suppressed, although IP-10 expression showed no significant difference (Fig. 4A). Naive alveolar macrophages expressed IP-10 and MIP-1α in response to PPE plus LPS in WT mice, but their expression was both suppressed in IRF3^−/− mice (Fig. 4B).

We next isolated lung macrophages (considered to include inflammatory monocytes) from lungs treated by PPE plus LPS and examined their function. IP-10 expression was diminished in IRF3^−/− lung macrophages; in contrast, MIP-1α expression showed no difference (Fig. 5A). Then, we assessed the phagocytic activity of harvested macrophages using fluorescent latex beads. FITC mean fluorescence intensity (MFI) and the proportion of FITC⁺ macrophages showed no significant change between the WT and the IRF3^−/− group (Fig. 5B). The extent of its activity inhibited by the phagocytic inhibitor cytochalasin D in WT or IRF3^−/− macrophages was 11 ± 1 and 9 ± 1%, respectively, which was also almost comparable. These data suggest the partial involvement of these cells in the distinct inflammatory mediator profiles induced by PPE plus LPS, and this effect was dependent on IRF3, although phagocytic activity appeared to be almost the same between WT and IRF3^−/− lung macrophages.

Given the apparently contradictory data of increased neutrophil counts but deceased morphologic changes in IRF3^−/− mice, we hypothesized that the function of the neutrophils recruited to the lung due to PPE plus LPS treatment was altered in these mice. First, we investigated the profiles of mediators produced by neutrophils that recruited to the lung after PPE plus LPS treatment. IP-10 mRNA expression was significantly increased in lung neutrophils derived from PPE plus LPS–treated WT mice compared with those from LPS only–treated WT mice (Fig. 6A). In contrast, this augmentation was not induced in IRF3^−/− mice (Fig. 6A). Regarding MIP-1α mRNA, the expression was significantly impaired in IRF3^−/− mice compared with WT mice (Fig. 6A).

Next, to investigate the survival ability of lung neutrophils, we examined apoptotic activity, as measured by using annexin V and the detection of the change in mitochondrial membrane potential. The apoptotic activity of neutrophils from mice treated with PPE plus LPS was significantly reduced in the IRF3^−/− group compared with the WT group in both assays (Fig. 6B).

Then we assessed the phagocytic activity of harvested neutrophils in the same way used for lung macrophages. FITC MFI and the proportion of FITC⁺ neutrophils indicated that neutrophils of the WT group exhibited elevated phagocytic activity of latex beads than did those of the IRF3^−/− group (Fig. 6C). The extent of their phagocytic activity inhibited by cytochalasin D was 17 ± 3% in WT neutrophils, whereas almost none of this effect was observed in IRF3^−/− neutrophils. Therefore, in the IRF3^−/− mice treated with PPE plus LPS, lung neutrophils exhibited suppressed phagocytic activity.

Finally, using the difference in the production of autophagy-related protein LC3-II in the presence or absence of chloroquine (ΔLC3-II), we investigated the autophagic activity of lung neutrophils and macrophages. Lung neutrophils from both WT and IRF3^−/− mice treated with PPE plus LPS showed increased LC3-II in the presence of chloroquine, whereas this difference was indistinct in the lung macrophages isolated from these two groups (Fig. 7A). As detected by Western blotting, ΔLC3-II of lung neutrophils was significantly greater in WT mice than in IRF3^−/− mice (Fig. 7B). Additionally, incubation with chloroquine increased LC3B expression in lung neutrophils from the WT group but not from the IRF3^−/− group (Fig. 7C). Therefore, treatment with PPE plus LPS activated autophagy in the lung neutrophils from WT mice, but that activity was suppressed in IRF3^−/− mice. These data, together with the production of chemokines, suggest that various functions of neutrophils were altered in neutrophils derived from PPE plus LPS–treated IRF3^−/− mice.

As cigarette smoke is profoundly involved in the pathogenesis of COPD, we examined the effect of CSE on alveolar macrophages, alveolar epithetilal cells, and lung neutrophils derived from PPE plus LPS–treated mice. Alveolar macrophages showed suppressed TNF-α production by 2% CSE administration in both WT and IRF3^−/− cell types (Fig. 8A). A similar result was detected in the expression of MCP-1 mRNA in alveolar epithetilal cells (Fig. 8B). As for neutrophils, we measured apoptotic activity using annexin, the change in mitochondrial membrane potential, and necrotic activity using PI staining. At 4% CSE, no significant change was detected in all assays compared with PBS-treated neutrophils (data not shown). In contrast, at 8% CSE, the apoptotic activity was significantly increased both in WT and IRF3^−/− group, and 16% CSE showed further increased apoptotic activity (Fig. 8C). In all conditions, WT neutrophils exhibited greater apoptotic activity compared with IRF3^−/− neutrophils. Additionally, necrotic activity detected by PI showed only in WT neutrophils at 16% CSE (Fig. 8C). These results indicate that the CSE effect on the apoptotic and necrotic change may be different between WT and IRF3^−/− neutrophils, although its effect on the production of chemokines showed similar change between both strains.

In this study, we demonstrated a novel role of IRF3 in a COPD exacerbation model. According to our morphologic studies using micro-CT and histopathology, the IRF3-associated pathway contributed to the additional emphysematous formation induced after LPS stimulation.

Micro-CT revealed that a single dose of LPS after PPE administration causes more severe emphysema at 12 wk after treatment than does PPE administration alone, suggesting that the sequential use of PPE followed by LPS in mice generates a useful model of COPD exacerbation (18) In our model, we observed additional changes of emphysema formation earlier after LPS administration than found in the previous study (Fig. 1).

The pathogenesis of COPD exacerbation is complicated and many types of the cells are involved, depending on the trigger of exacerbation. In a viral exacerbation model using dsRNA polyinosinic-polycytidylic acid or respiratory viruses, numbers of macrophages, neutrophils, and lymphocytes in BALF are elevated, and many mediators such as type 1 IFNs, IL-18, IP-10, and TNF-α are induced, leading to progression of emphysema or airway hyperresponsiveness (27, 28). This effect is partially dependent on TLR3 (27). Meanwhile, LPS or bacteria are used for bacterial exacerbation (29). LPS is recognized by TLR4, and subsequently many inflammatory mediators that are augmented in COPD patients are released by LPS stimulation (30–36).

In this LPS-triggered exacerbation model, a previous study demonstrated that CD8⁺ T cells infiltrated into the alveolar space and might be involved in pathogenesis (18). In the present study, neutrophils and macrophages in BALF are increased by PPE plus LPS treatment in WT and IRF3^−/− mice. Some inflammatory mediators in the mice treated with PPE plus LPS were dependent on IRF3. Among these mediators, IP-10 reportedly is IRF3-dependent (37). The chemokine IP-10 is known to be involved in a murine model of acute lung injury (38). Although the expression of IP-10 was increased in WT mice treated with PPE plus LPS compared with PPE or LPS only treated group, the neutralization of IP-10 by using anti–IP-10 Ab did not significantly modulate emphysematous formation after PPE plus LPS administration (T. Ishii and N. Yamashita, unpublished observations). In IRF3^−/− mice, other mediators were also suppressed; these include TNF-α, which is dependent on the MyD88–NF-κB pathway (39). This finding might be attributed to cooperation between NF-κB subunit p35 and IRF3 as an enhanceosome (12–14). In contrast, in IRF3^−/− mice, MIP-2 and IL-17A, which are associated with neutrophil recruitment, did not decrease compared with similarly treated WT mice. In a previous study, IRF3 interacted with retinoic acid receptor–related orphan receptor γt, a transcriptional regulator of differentiation for Th17 cells, and suppressed IL-17A production (40). This interaction may reflect overexpression of IL-17A and increased neutrophils in BALF.

As the origin of these mediators, many types of cells, including alveolar epithelial cells, macrophages, and neutrophils, were involved in our study. IRF3 is constitutively expressed both in immune cells and structural cells (11). These cells are known to be involved in COPD pathogenesis (41). In our observation, we presumed that alveolar epithelial cells and alveolar macrophages are involved in initial inflammation, and then activated macrophages, including inflammatory monocytes, and neutrophils augment the mediators, although other cell types, including lymphocytes and endothelial cells, may also be involved, and an analysis using bone marrow chimeric mice would provide a deeper insight. Elastolytic protease such as matrix metalloproteinases and cathepsins may also be involved in COPD pathogenesis, but in some models no apparent relationship is observed (42). We did not detect any difference between two strains (data not shown).

Neutrophils also have a pivotal role in COPD, and neutrophilic airway inflammation correlates with the decline in lung function (43). In the present study, despite less prominent morphologic changes of emphysema, neutrophil recruitment increased in PPE plus LPS–treated IRF3^−/− mice. Lung neutrophils isolated from IRF3^−/− mice exhibited significantly low levels of mediator production, apoptosis, phagocytosis, and autophagy. Apoptotic neutrophils are associated with COPD in two aspects (44): 1) reduced apoptotic activity, which potentiates prolonged neutrophil inflammation, and 2) a high proportion of apoptotic neutrophils, which results in secondary necrosis, releasing proinflammatory components into the air spaces (45). In the present study, we speculated that suppression of both the apoptotic and perhaps necrotic activities of IRF3^−/− neutrophils contribute to the increased neutrophil number in BALF, resulting in the decrease in the COPD morphologic changes despite the augmented neutrophil number in IRF3^−/− mice.

The influence of the phagocytic activity of neutrophils on COPD remains unclear; however, in a study using Haemophilus influenzae, the phagocytosis of bacteria by neutrophils led to cellular necrosis due to the release of toxic granules from cytotoxic neutrophils, thus augmenting COPD (46). We similarly used phagocytic activity as a marker of neutrophil function. The attenuated phagocytosis of bacteria by inhibition of IRF3 has also been reported in mouse peritoneal macrophages (47) and human monocytes (48). Our findings suggest that nonspecific phagocytosis by neutrophils differed between WT and IRF3^−/− mice and might be associated with necrosis and tissue damage via the release of the cytoplasmic contents of neutrophils.

Autophagy and apoptosis are related to diseases in both inductive and suppressive mechanisms (49). Autophagy may be associated with COPD pathogenesis in many aspects, including bronchial and alveolar epithelial cells and macrophages (50, 51). For example, elevated autophagy in lung epithelial cells promotes apoptosis through LC3b activation (51). Suppression of autophagy in neutrophils is suggested to be involved in autoimmune encephalomyelitis (52) and LPS-induced breakdown of the blood–brain barrier (52). To our knowledge, we are the first to describe alteration of autophagy in neutrophils in a COPD mouse model, suggesting that progressive autophagy and apoptosis in lung neutrophils may influence COPD exacerbation. The induced autophagy may be relevant to apoptosis and phagocytosis mutually, although it has been demonstrated that autophagy is also involved in differentiation and proliferation of neutrophils in lymphoid organs (53). Because autophagy is also activated by infection with dsDNA virus such as HSV1 through the cytoplasmic adaptor protein stimulator of IFN genes protein that is associated with IRF3 activation (54, 55), the observed difference in autophagic activity may be applicable in some virus-associated exacerbation model.

As cigarette smoking has an essential role in COPD pathogenesis, we examined the effect of CSE on various cell types. CSE suppressed expression of chemokines in alveolar epithelial cells and macrophages as previously described (56), and these suppressions were similar in IRF3^−/− counterparts. Cigarette smoke exposure can increase cellular stress, leading to apoptosis in many types of lung cells (57). In our experiments, CSE induced suppressed apoptosis and necrosis in IRF3^−/− neutrophils compared with WT neutrophils, indicating that IRF3 affects neutrophil function induced by CSE.

In summary, in this study we clarified the role of IRF3 in the pathogenesis of LPS-triggered COPD exacerbation. IRF3 knockout in mice decreases several inflammatory mediators that are involved in COPD in multiple types of lung cells, although impaired neutrophil function is observed in IRF3^−/− mice. Abs that target specific cytokines are expensive therapeutics and therefore have limited utility. In contrast, targeting a single transcriptional factor that influences multiple mediators important for the disease pathogenesis, such as IRF3, may be a promising, practical solution.

We thank Prof. T. Taniguchi (Department of Molecular Immunology, Institute of Industrial Science, University of Tokyo, Tokyo, Japan) and the RIKEN BioResource Center for providing IRF3^−/− mice. We thank Emi Kishimoto for excellent secretarial assistance.

The authors have no financial conflicts of interest.