TNF-α and Macrophages Are Critical for Respiratory Syncytial Virus–Induced Exacerbations in a Mouse Model of Allergic Airways Disease

Asthma is a complex inflammatory disease of the lung characterized by recurrent airflow obstruction, wheezing, airway inflammation, and airway hyperresponsiveness (AHR) (1). Asthma exacerbations result in a significant worsening of asthma symptoms (2) and occur in patients with asthma regardless of disease severity, but are most frequent in patients with severe disease (3). Although those with severe asthma represent a relatively small proportion of all asthma patients (≤10%) (4), disease exacerbation in this subset negatively impacts quality of life and accounts for significant healthcare costs (5, 6). In stable disease, current treatments include inhaled corticosteroids, long-acting inhaled β₂-agonists, and oral corticosteroids for severe exacerbation. These approaches often fail to adequately control symptoms in patients with severe disease (7–9). Therefore, an improved understanding of the mechanisms driving severe asthma, and disease exacerbation in particular, is required for the development of novel therapeutic approaches.

Viral respiratory infections are the most common triggers of exacerbation in adults (76%) (10) and children (80–85%) (11, 12). Respiratory syncytial virus (RSV), rhinovirus, and parainfluenza viruses are the most common viruses detected during an asthma exacerbation (10–14). RSV is a negative-stranded RNA virus of the paramyxoviridae family that causes bronchiolitis, wheezing, and impaired lung function in children (15) and accounts for 7.2% of hospitalizations in elderly asthmatic patients (16). Furthermore, childhood RSV infection is associated with an increased risk of developing asthma in adult life (17, 18).

RSV infection in animal models significantly increases neutrophil, eosinophil, and lymphocyte infiltration into the airways and induces high levels of proinflammatory cytokines including IL-1, IL-6, IL-12, IL-13, IFN-γ, TNF-α, MCP 1 (MCP-1/CCL2), CCL3 (MIP-1α), CCL5 (RANTES), CXCL10 (IP-10), and keratinocyte-derived protein chemokine (KC/CXCL1; the mouse homolog of IL-8) (19–22). The specific patterns of cytokine induction (and their impact on AHR) varies depending on the RSV strain being assessed.

Several mouse studies have also investigated links between RSV infection and allergic airways disease. RSV infection before OVA challenge induces increased AHR, associated with increased eosinophil and neutrophil numbers and IL-4 expression, which were abolished following anti–IL-5 Ab treatment (23). Further, RSV infection during the OVA challenge phase resulted in increased AHR, eosinophil numbers, and enhanced expression of IL-4 and IL-5 in the lung (24). By contrast, RSV infection with A2 strain induced AHR in mice with allergic airways disease without inducing Th2 cytokines, but rather through increased IL-17A levels (25, 26). Finally, RSV infection after OVA challenge in the absence of adjuvant sensitization also increased AHR and induced high levels of IL-4 and IL-5 and the accumulation of eosinophils and neutrophils into the airways (20). These studies demonstrate that RSV infection can increase a number of features of allergic airways disease, when infection occurs before or during allergen challenge, or in the absence of previous allergic sensitization. In these models, increased AHR occurs predominantly through enhanced IL-4 and IL-5 production and increased accumulation of eosinophils and neutrophils. However, these studies have not assessed the impacts of RSV infection in the context of pre-existing allergic airways disease akin to chronic asthma, in which viral infection may drive disease exacerbation. Further, these studies have not addressed the potential roles of inflammatory cytokines and macrophages in persistent AHR or viral-induced exacerbation.

In clinical studies, high levels of the neutrophil chemokine IL-8 and high percentages of neutrophils are commonly observed during an acute exacerbation of asthma, whereas eosinophils are generally associated with allergic asthma (10, 27). Neutrophils are also predominant in sputum from patients with severe asthma, who often respond poorly to standard corticosteroid treatments (28). Furthermore, increasing evidence suggests important roles for pulmonary macrophages in the pathogenesis of severe asthma, particularly in steroid-resistant asthma (29–34). Simpson et al. (29) has reported impaired macrophage function in patients with noneosinophilic asthma and macrophage activation is associated with impaired lung function (35). Our group recently demonstrated that macrophages are a crucial link in the development of steroid-resistant airway inflammation and AHR via impaired glucocorticoid receptor translocation (30, 33, 34). In an acute rhinovirus-induced exacerbation model, induced MCP-1 increases AHR and airway inflammation in mice with allergic airways disease, with bronchoalveolar lavage fluid (BALF) macrophages being a major producer of MCP-1 (36). In an extension of these findings, rhinovirus-induced exacerbation also increased macrophages in the lung, which produced high levels of IL-13 (37). Macrophages contributed to increased AHR and airway inflammation through a CCR2-dependent mechanism (37). These studies provide evidence for the critical role of macrophages in infection-induced exacerbation. However, little work has focused on the role of pulmonary macrophages during RSV-induced exacerbation, particularly the impact on prolonged increases in AHR beyond the acute viral infection phase.

Several clinical studies have identified increased TNF-α and MCP-1 levels in serum from virus-infected patients with asthma compared with noninfected patients (38, 39). TNF-α is a key proinflammatory cytokine produced primarily by monocytes/macrophages, but also by eosinophils, T cells, mast cells, and lung epithelial cells. In airways disease, TNF-α plays important roles in the induction of AHR (40, 41) and the recruitment of eosinophils and neutrophils into the airways (42, 43). MCP-1 is a key chemokine regulating macrophage recruitment and is also produced by activated macrophages (19, 44). However, roles for TNF-α and MCP-1 in the pathogenesis of viral-induced exacerbation of AHR and inflammation remain poorly understood, and the immunological mechanisms underlying RSV-induced exacerbation of asthma are still largely unknown.

The aim of the current study was to establish a mouse model of RSV-induced exacerbation of AHR and airway inflammation to study the underlying mechanisms and investigate potential therapeutic approaches. We observed that RSV infection in the context of underlying allergic airways disease exacerbated prolonged AHR responses. Exacerbations were accompanied by increased infiltration of macrophages and neutrophils into the lung and increased proinflammatory cytokine and chemokine production (including TNF-α, MCP-1, and KC). Increased AHR during viral-induced exacerbation was significantly resistant to steroid therapy; however, inhibition of TNF-α or MCP-1 or the depletion of macrophages inhibited enhanced airway reactivity. We also observed increased TNF-α levels in sputum samples from patients with neutrophilic asthma, a patient population that is often poorly controlled by steroid therapy. Our study provides a novel mouse model of RSV-induced exacerbation of AHR and airway inflammation and highlights key functional roles for macrophages and the innate immune response in the pathogenesis of disease. Our findings suggest novel therapeutic approaches for the treatment of acute exacerbation by targeting macrophage function and proinflammatory factors, including TNF-α and MCP-1.

BALB/c male mice (5 to 6 wk old) were obtained from the University of Newcastle Animal Services Unit, and experiments were performed in the Hunter Medical Research Institute animal facility under specific pathogen-free conditions, following review and approval from the local animal care and ethics committee (A-2013-337).

As previously described (45), human RSV (long strain, type A) was propagated in Hep-2 cells for 4 d. Supernatant and cell lysate were collected, and virus was purified using two-step sucrose gradients (33 and 77%) by ultracentrifugation. RSV titer was quantified by plaque assay using Hep-2 cells. Where indicated, RSV was UV-inactivated for 40 min prior to inoculation into mice.

Mice were initially sensitized to OVA (50 μg/mouse) or PBS in 1 mg alhydrogel (aluminum hydroxide [alum]) in 200 μl PBS on day 0 by i.p. injection. Mice were then challenged with 1% OVA aerosol in saline daily for 30 min/day from days 13–16. Where indicated mice were exposed to RSV or UV-inactivated RSV (0.5 × 10⁶ PFUs/mouse) by intranasal administration on day 19. All endpoints were assessed at day 24 of the model (5 d postinfection).

Where indicated, mice were treated with dexamethasone (DEX) (Sigma-Aldrich) (1 mg/kg) in 200 μl PBS by i.p. injection on days 18 and 20. For intervention studies, mice were treated by i.p. administration of 0.2 mg anti–TNF-α (clone XT3.11; BioXCell) or anti–MCP-1 (clone 2H5; BioXCell) in 200 μl PBS or the corresponding isotype controls (rat IgG2a or hamster IgG; BioXCell, respectively) on days 20 and 22. To deplete macrophages, mice were treated with 2-chloroadenosine (2-CA) (10 μl 10 mmol 2-CA or PBS control) by intratracheal administration on days 20 and 22 (32, 34).

Airway resistance in response to increasing concentrations of methacholine (MCh; Sigma-Aldrich) was measured using a Flexivent apparatus (FX1 system; Scireq, Montreal, Canada), as previously described (32, 33). Briefly, mice were anesthetized with a mixture containing xylazine (10 mg/kg) and ketamine (100 mg/kg) by i.p. injection. A cannula was then inserted into the trachea, and mice were ventilated with a tidal volume of 8 ml/kg at a rate of 450 breaths/min. Mice were initially challenged with aerosol saline followed by increasing concentrations of MCh (0.3, 1 3, 10, and 30 mg/ml in saline) for 10 s at each dose. Airway reactivity was recorded and presented as percentage increase over baseline (saline).

BALF was collected by repeated washes of the right lung lobes with sterile HBSS. BALF was then centrifuged to collect cellular infiltrate, and supernatants were stored at −80°C for subsequent cytokine quantification. RBCs were removed by using hypotonic RBC lysis buffer, as previously described (34). Total cell counts were then determined by hemocytometer, and remaining cells were cytospun onto glass slides. Differential leukocyte counts were determined by using morphological criteria by light microscopy (×100) on May-Grunwald– and Giemsa-stained slides, counting a minimum of 300 cells/slide/sample.

Pulmonary macrophages were isolated from mouse lungs as previously described (46). Briefly, macrophages were mechanically extracted from mouse lung tissues through 70-μm cell strainers. Macrophages then were purified by gradient centrifugation (Histopaque 1083; Sigma-Aldrich) and seeded into a six-well plate at a concentration of 6 × 10⁶ cells/ml in DMEM containing 20% FCS. After 3 h, >95% of adherent cells were macrophages, collected into TRIzol, and kept at −80°C for further use.

Patients with asthma, defined by clinical diagnosis with evidence of AHR to hypertonic saline and/or bronchodilator response (47), were recruited and categorized via induced sputum inflammatory cell counts. Participants with a sputum eosinophil count of ≥2% and a sputum neutrophil count <61% were classified as eosinophilic asthma. Participants with a sputum neutrophil count of ≥61 and <2% eosinophils were classified as neutrophilic asthmatic. Participant clinical characteristics are summarized in Supplemental Table I. All participants gave written informed consent prior to their inclusion in the study, which was approved by the Hunter New England Area Health Service and the University of Newcastle Research Ethics Committees.

Induced sputum samples from 40 patients (20/group) were obtained using nebulized hypertonic (4.5%) saline (48). For the assessment of inflammatory cells, DTT was used to disperse cells from mucus, as previous described (49). Total cell counts and cell viability (trypan blue exclusion) were performed with a hemocytometer, followed by preparation of cytospins for differential cell counts using May-Grunwald-Giemsa. Selected sputum plugs (100 μl) were stored for RNA analysis in RLT buffer (Qiagen, Valencia, CA) at −80°C. RNA was extracted using RNeasy Mini kits (Qiagen) and quantified using the Quant-iT RiboGreen assay (Invitrogen, Carlsbad, CA). RNA (200 ng) was reverse-transcribed to cDNA using high-capacity cDNA reverse transcription kits according to the manufacturer’s instructions (Applied Biosystems, Foster City, CA). TaqMan qPCR primer and probes for TNF-α (Hs00174128_m1) and MCP-1 (Hs00234140_m1) were purchased in kit form (Applied Biosystems) and combined in duplex real-time PCRs using an ABI Viia7 real-time PCR machine (Life Technologies). Expression levels of TNF-α and MCP-1 were calculated using 2^−ΔΔCt relative to the reference gene eukaryotic 18S rRNA and an internal calibrator.

As previously described (50), RNA was isolated from lung tissues using TRIzol reagent (Invitrogen), purified by phenol-chloroform extraction, and quantified on a Nanodrop 1000 spectrophotometer (Nanodrop). cDNA was synthesized by RT-PCR using random hexamer primers (Invitrogen) from 1 μg RNA on a T100 thermal cycler (Bio-Rad).

Quantitative PCR (qPCR) was performed on a Viia7 real-time PCR machine (Life Technologies) using SYBR reagents with primers listed in Supplemental Table II. Thermal cycling conditions consisted of an initial denaturing step (95°C: 3 min) followed by 40 cycles of denaturing (95°C, 5 s) and annealing (60°C, 30 s). The mRNA levels were normalized to hypoxanthine phosphoribosyltransferase (internal control) and expressed as a fold-change relative to control samples (PBS group). Absolute quantification of RSV levels was determined using plasmid copy number standards for RSV N gene and hypoxanthine phosphoribosyltransferase.

Lung tissue was homogenized using RIPA buffer (Sigma-Aldrich) plus protease/phosphatase inhibitor mixture (Cell Signaling Technology; Danvers, MA) on a Tissuelyser LT tissue disruptor (Qiagen) at 50 Hz for 8 min and stored at −80°C for further cytokine measurement.

TNF-α and MCP-1 concentrations in lung homogenates and BALF samples were determined using a cytometric bead array (CBA) kit according to the manufacturer’s instructions (BD Biosciences), with a limit of detection for TNF-α (7.3 pg/ml) and MCP-1 (52.7 pg/ml). Data were collected on a BD FACS Canto II flow cytometer and analyzed using FCAP Array software (version 3.0; BD Biosciences). An ELISA kit was used to quantify KC according to the manufacturer’s instructions (R&D Systems) with a limit of detection of 15.6–1000 pg/ml. All lung protein data have been normalized to total protein levels (picograms per milligram of total protein).

Lung tissues were fixed in 10% neutral buffered formalin for 24 h and then transferred to 70% ethanol. Lungs were paraffin-embedded and sections stained using chromotrope for eosinophil quantification or periodic acid-Schiff for mucus-secreting cells (MSCs). Numbers of MSCs and eosinophils were determined by morphological criteria under the light microscopy at ×100 original magnification and quantified as described previously (32).

All data were analyzed using GraphPad Prism Software (Version 6; GraphPad, San Diego, CA) and presented as means ± SEM. Lung function data were analyzed by two-way ANOVA followed by a Bonferroni post hoc test for comparison between groups. Other data including different cell counts, RNA and protein levels, and histological data were analyzed using unpaired two-sided Student t test. Differences were considered statistically significant if p < 0.05.

To establish a mouse model of RSV-induced exacerbation of AHR and airway inflammation, we employed our well-established murine model of prolonged AHR (32). Initially, BALB/c mice were sensitized with OVA (or PBS control) in alum and later exposed to OVA aerosol to establish allergic airways disease. Three days after the final OVA aerosol challenge, some groups were inoculated with either RSV or UV-inactivated RSV, and lung function was assessed on day 5 postinfection.

OVA-sensitized/challenged (OVA-treated) mice exhibited increased airways resistance in response to increasing doses of MCh, compared with control PBS-treated mice (PBS sensitized/OVA challenged) (Fig. 1A, 1B), as we have demonstrated previously (32). In PBS control mice infected with RSV (PBS/RSV), no increase in airways resistance was observed above baseline levels (Fig. 1A, 1B). However, RSV infection dramatically increased airway resistance in OVA-treated mice (OVA/RSV) (Fig. 1A, 1B). By contrast, UV-inactivated RSV treatment had no effect on AHR in OVA-treated mice (OVA/UV-RSV) (Fig. 1A, 1B). We quantified RSV levels in the lungs by qPCR, and no significant differences were observed between the OVA/RSV and PBS/RSV group (Fig. 1C). RSV was not detected in lungs from the OVA/UV-RSV group (Fig. 1C).

Total and differential immune cell numbers were quantified in BALF to characterize the extent and type of inflammatory infiltration. OVA-treated mice exhibited increased total BALF cell numbers, predominantly consisting of eosinophils and macrophages, but neutrophil numbers were unaltered compared with PBS-treated mice (Fig. 1D–H). RSV infection alone (PBS/RSV) had no effect on eosinophil numbers, but did increase macrophage, neutrophil, and lymphocyte numbers (Fig. 1E–H). OVA/RSV-treated mice exhibited a marked increase in total cell infiltrate, compared with OVA and PBS/RSV-treated mice (Fig. 1D). Importantly, OVA/RSV mice had marked increases in macrophage and neutrophil numbers in the BALF, whereas eosinophil and lymphocyte numbers remained unchanged compared with the OVA group (Fig. 1E–H). These findings demonstrate that RSV infection in the context of pre-existing allergic airways disease exaggerates AHR, which is accompanied with increased macrophage infiltration of the airways.

Further assessment of eosinophil numbers in lung tissues by histology revealed similar findings, with increased eosinophil numbers in OVA-treated animals compared with PBS-treated controls (Supplemental Fig. 1A). PBS/RSV treatment did not increase eosinophil numbers, and no difference was observed among the OVA, OVA/UV-RSV, or OVA/RSV groups (Supplemental Fig. 1A). We also quantified the number of MSCs by histology, as mucus hypersecretion is a key feature of allergic airways disease. RSV infection alone (PBS/RSV) induced a slight increase in MSC numbers (Supplemental Fig. 1B). OVA treatment led to marked increases in MSC cell numbers, which were unchanged following additional exposure to either UV-RSV or RSV (Supplemental Fig. 1B). Thus, although RSV infection exacerbates airway resistance in the context of pre-existing allergic airways disease, it does so without altering eosinophil infiltration and mucus hypersecretion within the lung.

We next investigated the impact of RSV-induced exacerbation on local cytokine and chemokine responses within the lung, to gain insight into potential mechanisms driving the development of viral-induced exacerbation. RSV infection in both OVA/RSV and PBS/RSV-treated mice induced high levels of TNF-α and MCP-1 in the lung, which were not observed in OVA/UV-RSV or OVA-treated mice (Fig. 2A, 2B). Further, MCP-1 protein levels were significantly increased in OVA/RSV-treated mice, above levels observed in PBS/RSV-treated mice (Fig. 2B). We also observed increased expression of KC mRNA following RSV infection (in both PBS/RSV and OVA/RSV groups) by qPCR, although increased KC protein was only detectable in lung homogenates from OVA/RSV-treated mice (Fig. 2A, 2B).

At the endpoint assessed, the key Th2 cytokines IL-5 and IL-13 were unchanged in OVA-treated mice, following inoculation with RSV (OVA/RSV) or UV-inactivated RSV (OVA/UV-RSV) (Supplemental Fig. 1C, 1D). Thus, RSV infection induces increased levels of TNF-α, MCP-1, and KC in the lung, but does not amplify the underlying production of Th2 type cytokines.

To determine the relevance of our observations to human disease, we assessed the levels of TNF-α and MCP-1 in sputum samples from patients by qPCR. Levels of TNF-α were significantly increased in participants with neutrophilic asthma compared eosinophilic asthma (Fig. 3A). MCP-1 levels were also higher in neutrophilic asthma (although not significant; p = 0.132) (Fig. 3B). These data are similar to observations in our mouse model and may indicate a key role for TNF-α in neutrophilic, but not eosinophilic, asthma.

The main treatment for patients with severe asthma is inhaled corticosteroids, although patients with viral-induced exacerbation and severe asthma often fail to respond (51). Therefore, we next assessed the effect of DEX treatment on RSV-induced exacerbation of AHR and airway inflammation to determine whether our model replicated this feature of clinical disease. OVA/RSV and OVA-treated mice were treated with DEX 24 h before and postinfection/noninfection with RSV (days 18 and 20), subsequent to the establishment of allergic airways disease.

DEX treatment of OVA-treated mice (OVA/DEX) effectively suppressed allergic inflammation and decreased airway resistance to baseline (PBS) levels (Fig. 4A, 4B). Further, DEX decreased total inflammatory cell numbers in BALF, in particular eosinophils and lymphocytes (Fig. 4D, 4E, 4H). By contrast, DEX treatment only partially suppressed the RSV-induced AHR exacerbation (OVA/RSV/DEX) (Fig. 4A, 4B). Further, in OVA/RSV-treated mice, DEX treatment had no effect on total BALF cell numbers or numbers of infiltrating macrophages, neutrophils, or lymphocytes (Fig. 4D, 4F, 4H). Interestingly, DEX treatment in OVA/RSV-treated mice increased BALF eosinophil numbers, compared with OVA/RSV controls (Fig. 4E). Similar findings were seen in histological assessments, in which DEX treatment effectively suppressed eosinophil and MSC numbers in OVA-treated mice (OVA/DEX), but not following viral-induced exacerbation (OVA/RSV/DEX) (Supplemental Fig. 2). DEX treatment had no effect on RSV viral loads within the lungs (Fig. 4C). In lung tissue, levels of TNF-α and MCP-1 but not KC mRNA were slightly decreased by DEX treatment in OVA/RSV-treated mice (Fig. 5A). However, protein levels remained unchanged following DEX treatment, with TNF-α, MCP-1, and KC remaining elevated (OVA/RSV/DEX) (Fig. 5B).

Based on previous evidence of a role for pulmonary macrophages in steroid-resistant disease, we also assessed inflammatory cytokine expression in ex vivo–isolated lung macrophages. Macrophages from OVA/RSV-treated mice produced high levels of TNF-α and MCP-1 (Fig. 5C), compared with macrophages isolated from OVA or PBS-treated mice. Interestingly, DEX treatment had no effect on TNF-α or MCP-1 expression levels, compared with OVA/RSV-treated controls (Fig. 5C). We did not observe any change in KC expression in isolated pulmonary macrophages between our groups (data not shown). Overall, these data indicate that RSV-induced exacerbation of AHR and airway inflammation in allergic airways disease is significantly resistant to glucocorticoid treatment, similar to clinical findings during viral-induced exacerbation, and identify pulmonary macrophages as a major source of TNF-α and MCP-1.

To investigate the functional role of TNF-α in the RSV-induced exacerbation of AHR and inflammation, mice were systemically treated with anti–TNF-α Ab. Anti–TNF-α treatment effectively suppressed AHR compared with isotype control treatment (Fig. 6A) and had no effect on RSV levels in the lung (Fig. 6B). In addition, anti–TNF-α effectively decreased total BALF cell numbers, decreasing eosinophils, macrophages, neutrophils, and lymphocytes, compared with isotype control (Fig. 6D–G). In addition, TNF-α neutralization decreased the number of eosinophils and MSCs in lung tissue (Fig. 6H, 6I).

Neutralizing TNF-α also suppressed TNF-α, MCP-1, and KC expression in lung tissues from OVA/RSV-treated mice, compared with isotype control-treated mice (Fig. 7A). Further, anti–TNF-α decreased TNF-α and MCP-1 protein levels, but not KC, in lung homogenates compared with isotype-treated controls (Fig. 7B). However, KC protein levels were decreased in BALF supernatants (Fig. 7C) following anti–TNF-α administration. These findings demonstrate a key functional role for TNF-α in RSV-induced exacerbation of AHR and airway inflammation.

MCP-1 has previously been reported to play important roles in the pathogenesis of asthma (36, 52, 53) and is increased in our model. Furthermore, increased MCP-1 levels potentially were downstream of TNF-α production. We next examined the effects of administering neutralizing Abs against MCP-1 in our RSV-induced exacerbation model. Neutralization of MCP-1 completely abolished RSV-induced exacerbation of AHR, although total viral levels in the lung were significantly increased, suggesting impairment of antiviral host-defense responses (Fig. 8A, 8B). Furthermore, inhibiting MCP-1 reduced total inflammatory cell numbers in the BALF (Fig. 8C), with decreased numbers of both eosinophils and macrophages (Fig. 8D, 8E). However, neutrophil and lymphocyte numbers were not significantly altered by MCP-1 neutralization (Fig. 8F, 8G). In lung tissue, TNF-α and KC expression levels were unaffected by anti–MCP-1 Ab administration (Fig. 9A). However, levels of MCP-1 protein were markedly decreased following Ab administration (Fig. 9B). These findings demonstrate that MCP-1 plays a critical role in promoting macrophage recruitment during viral-induced exacerbation and in the induction of AHR.

Next, we investigated the impact of macrophage depletion in our model using 2-CA administration [a purine analogue that specifically depletes pulmonary macrophages (32, 54, 55)]. Following 2-CA administration, RSV-induced AHR was reduced compared with vehicle-treated controls (Fig. 10A), with no effect on RSV levels in the lung (Fig. 10B). 2-CA treatment dramatically decreased BALF macrophage numbers, but had no impact on other inflammatory cell counts (Fig. 10C–F). Further, 2-CA treatment significantly decreased expression of TNF-α and MCP-1 (but not KC) in lung tissues of OVA/RSV-treated mice compared with vehicle-treated mice (Fig. 10G). Although 2-CA treatment had no effect on inflammatory cytokine protein levels in lung homogenates (data not shown), it suppressed levels of TNF-α and MCP-1 (but not KC) protein in BALF (Fig. 10H). Together, these findings indicate a key role of macrophages in RSV-induced exacerbation of AHR and airway inflammation and suggest that targeting pulmonary macrophage function may provide therapeutical potential.

Viral respiratory infections are common triggers for disease exacerbation in patients with asthma (11, 13, 14). In this study, we established a model of RSV-induced exacerbation of AHR and inflammation during pre-existing allergic airways disease to investigate mechanisms underlying infection-induced exacerbation. Although several previous studies have assessed the ability of RSV infections to alter airways function, they only assessed the impacts of infection on the establishment of allergic airways disease (20, 23, 24). By contrast, we examined the role of RSV infection in exacerbation of pre-existing allergic airways disease, assessing changes in the prolonged AHR response (32).

Our well-established allergic airways disease model results in rapid AHR induction 1 d after the final OVA challenge and prolonged AHR that persists for >1 wk (32). When mice were infected with RSV during the postchallenge phase, RSV infection led to significantly exaggerated AHR. In a previous study, RSV infection alone could induce AHR, which we did not observe in our model (23). This difference may be due to differences in lung measurement techniques (forced oscillation technique, in our study, versus unrestrained whole-body plethysmography, used previously) (23). In other studies assessing interactions between RSV infection and development of allergic airways disease, viral infection increased eosinophil and Th2 cytokines, including IL-4, IL-5, and IL-13 (20, 21). By contrast, our RSV-induced asthma exacerbations do not occur through enhanced eosinophil numbers or IL-5 and IL-13 expression. More importantly, we did observe increased macrophage and neutrophil numbers in the BALF and high levels of TNF-α, MCP-1, and KC in lung homogenates. Our data are similar to observations of RSV strain A2 infection during OVA challenge, which augmented AHR through increased IL-17A and neutrophil accumulation, without elevated Th2 cytokines (25, 26). Significant increases of IL-17A protein in the total lung homogenates of PBS/RSV- (14.28 ± 1.37 pg/mg) and OVA/RSV-treated (13.11 ± 0.907 pg/mg) mice were detected (data expressed as mean ± SEM; n = 6; p < 0.001), as compared with groups treated with PBS (5.522 ± 0.635 pg/mg), OVA (4.589 ± 0.967 pg/mg), or OVA/UV-RSV (4.703 ± 0.339 pg/mg). These findings suggest that RSV infection alters lung function via different mechanisms, depending on the timing of infection in relation to allergen sensitization/challenge. Similar to our findings, rhinovirus-induced AHR exacerbation of pre-existing allergic airways disease also increases MCP-1 expression from macrophages (36); however, exacerbation of AHR occurs through increased IL-13 expression (37). Although allergic airways disease is commonly characterized by activated Th2 cells, increased expression of Th2 cytokines (e.g., IL-5 and IL-13), eosinophil infiltration, and mucus hypersecretion (56, 57), these parameters were not induced by RSV infection in OVA/RSV-treated mice.

Although we observed increased TNF-α and neutrophil infiltration in both PBS/RSV- and OVA/RSV-treated mice, only OVA/RSV-treated mice exhibited exaggerated AHR. This demonstrates that induction of TNF-α and accumulation of macrophage and neutrophils alone are not sufficient to induce AHR. We propose that acute inflammation acts together with underlying Th2 inflammation to exacerbate AHR. Similarly, although LPS administration can stimulate neutrophilia and increase inflammatory mediators (such as TNF-α, IL-1, and IL-6), LPS stimulation alone does not induce AHR unless there is a pre-existing Th2 background (34, 58, 59). Collectively, these findings indicate that RSV-induced exacerbation of AHR and airway inflammation are not driven by increased Th2 responses and eosinophil infiltration, but rather may be induced by increased expression of TNF-α and MCP-1 and innate immune cell activation, in particular macrophages.

Our observations are consistent with recent clinical findings from patients suffering viral-induced asthma exacerbation (27, 38, 39, 60, 61). In particular, high levels of IL-8 and predominant neutrophils, rather than eosinophils, are observed in patients with an acute exacerbation who fail to respond to corticosteroid treatment (10, 27, 28). Furthermore, high levels of TNF-α were detected in the nasal fluid of patients with asthma following exacerbation, and TNF-α levels predicted subsequent exacerbation frequency (60). TNF-α levels in serum were also associated with severity of RSV infection and bronchial asthma (39). Furthermore, TNF-α levels are increased in induced sputum samples from patients with asthma (62, 63), in particular patients with severe steroid-resistant disease (64). An association among sputum neutrophils, disease severity, and TNF-α levels has also been demonstrated, with sputum TNF-α levels significantly associated with neutrophil numbers in patients with severe asthma, but not healthy controls or patients with mild asthma (65). High levels of serum MCP-1 during an asthma exacerbation were also related to viral infection (38, 61, 66), and MCP-1 levels were increased in RSV-infected monocytes from patients with asthma compared with monocytes from healthy individuals (38, 61). Importantly, we also observed increased TNF-α and a trend toward increased MCP-1 in sputum samples from patients with stable neutrophilic asthma compared with eosinophilic asthma.

Corticosteroid therapy often fails to control exacerbation in patients with asthma, in particular those with severe disease (7–9, 28). These patients often present with neutrophilic, rather than eosinophilic, asthma, and few effective treatment options are available (28, 67). In our study, RSV-induced exacerbation of AHR and airway inflammation increased macrophage infiltration, whereas numbers of eosinophils are unchanged compared with OVA-treated mice. DEX treatment suppressed all features of allergic asthma including AHR, eosinophil numbers (BALF and lung tissues), lymphocytes, and MSCs in mice with allergic airways disease. However, DEX treatment only partially suppressed RSV-induced exacerbation of AHR and had no effect on inflammatory cell numbers or the levels of the innate inflammatory cytokines MCP-1, TNF-α, or KC. In fact, eosinophil numbers were paradoxically increased slightly following steroid treatment. These findings are similar to a recent report of RSV-induced exacerbation of house dust mite–driven allergic airways disease, in which disease exacerbations were also resistant to steroid treatment (68). Several recent clinical trials have also demonstrated no benefit from DEX treatments following RSV-induced lung inflammation in children (69–71). Further, although high doses of corticosteroids could partially prevent recurrent wheezing in RSV-induced bronchiolitis, it failed to improve long-term respiratory outcomes (72). Taken together, our data clearly indicate that RSV-induced exacerbation of AHR and airway inflammation are driven by a steroid-resistant pathway that is linked to innate immune responses, mimicking many features of clinical disease.

TNF-α has been reported to play important roles in severe steroid-resistant asthma through induction of AHR, recruitment of neutrophils (73), and the induction of glucocorticoid resistance (74). In the current study, we showed that the levels of TNF-α increased in OVA/RSV-treated mice and were not inhibited by DEX treatment. Interestingly, neutralization of TNF-α inhibited RSV-induced AHR and airway inflammation reducing the level of all inflammatory cells in BALF. Further, decreased expression of inflammatory cytokines TNF-α, MCP-1, and KC were observed in the lung after anti–TNF-α treatment. TNF-α has also been reported to play an important role in the development of allergic airways disease; in particular, blocking TNF-α could suppress hallmark clinical features of allergic airway inflammation including AHR, eosinophil recruitment, Th2 cytokine (IL-4, IL-5, and IL-13) production, and mucus hypersecretion (75). We also found higher levels of TNF-α in the sputum of patients with neutrophilic asthma compared with eosinophilic asthma. This observation supports a previous study showing increased TNF-α levels in BALF and endobronchial biopsy samples from patients with severe asthma compared with patients with mild disease (64). Several clinical studies have assessed TNF-α as a potential therapeutic target for asthma and severe steroid-resistant airway inflammation (64, 76, 77). In an initial study of patients with moderate asthma, anti–TNF-α treatment failed to improve lung function, although it did reduce exacerbations (76). A further study, specifically targeting severe asthma, demonstrated a trend toward decreased exacerbations, although this study was discontinued due to increased infections following treatment (78). In an unrandomized, unblinded case series of seven patients with severe steroid-resistant asthma, anti–TNF-α treatment yielded promising results with improved asthma control and reduced exacerbations (77). Clinical trials have not been performed that specifically assess the impacts of anti–TNF-α treatments on viral infection-induced exacerbation. Taken together, our data indicate that TNF-α is a potential target for treatment of viral-induced asthma exacerbation.

MCP-1 is an important regulator in the pathogenesis of allergic asthma, and neutralizing MCP-1 has been shown to suppress AHR and inflammation in allergic airways disease (52, 53). Interestingly, we observed that RSV infection significantly increased MCP-1 levels in OVA-treated mice, compared with PBS-treated mice. We also observed a trend toward increased MCP-1 levels in sputum samples from patients with neutrophilic asthma compared with eosinophilic asthma. Previous studies have demonstrated increased serum MCP-1 levels in patients following acute viral respiratory infection (38). Nasal aspirate MCP-1 levels are also increased in children with asthma suffering from respiratory viral infections (79). In our model, administration of anti–MCP-1 Abs completely suppressed AHR induction and reduced macrophage and eosinophil numbers in the airways. This is similar to the protective effect of MCP-1 neutralization on rhinovirus infection–induced exacerbation of AHR mentioned previously (36). We also observed increased virus levels after anti–MCP-1 treatment (although not after 2-CA treatment), which highlights a role for MCP-1 in virus clearance. Whether this occurs directly through MCP-1 activation of macrophages or indirect mechanisms via other cell types was not determined. In this study, we identify a critical regulatory role for MCP-1 in the recruitment of macrophages and development of RSV-induced exacerbation of AHR and airway inflammation in mice with allergic airways disease.

Activated macrophages have also been shown to play important roles in the pathogenesis of severe asthma and steroid-resistant asthma (30, 33, 34, 80). In our previous work, we also demonstrated a key role for macrophages in prolonged AHR following OVA sensitization/challenge alone in the absence of viral infection (32). Although 2-CA treatment had no effect on eosinophil or MSC numbers, it effectively reduced BALF macrophage numbers by >90% (32). In the current study, we observed increased macrophage numbers in the lung following RSV-induced exacerbation, and macrophage numbers were decreased by both TNF-α and MCP-1 treatments. Further, our data show that pulmonary macrophages are a major source of TNF-α and MCP-1 following RSV infection. These data support previous findings that alveolar macrophages are a potent source of TNF-α during the late asthmatic reaction (81). Critically, our findings demonstrated that infiltration of the airways by macrophages and the expression of TNF-α and MCP-1 were not dampened by DEX treatment. Similarly, alveolar macrophages isolated from patients with severe asthma exhibit corticosteroid resistance, with more inflammatory cytokine production (including MCP-1) following LPS stimulation in the presence of DEX, compared with macrophages from patients with nonsevere asthma (80). Our findings suggest that TNF-α and MCP-1 promote macrophage recruitment to the lung as a critical component of steroid-resistant airways disease exacerbation. To test this hypothesis, we assessed the direct impacts of macrophage ablation following 2-CA treatment. Our findings demonstrate that macrophage depletion completely inhibits RSV-induced exacerbation of AHR in mice with allergic airways disease and specifically reduced BAL macrophage numbers. This is accompanied by reduced proinflammatory cytokine expression (TNF-α, MCP-1, and KC) in lung tissues and protein levels in BALF. Our current data suggest that macrophages also play an important role in the pathogenesis of RSV-induced exacerbation of AHR and airway inflammation in mice with allergic airways disease.

In summary, we have described a mouse model that mimics aspects of RSV-induced exacerbation of asthma in humans. Notably, we demonstrated a central role for macrophages, TNF-α, and MCP-1 in the pathogenesis of RSV-induced exacerbation of AHR and inflammation in mice with allergic airways disease. These findings suggest that targeting a macrophage/TNF-α/MCP-1 axis through single or multiple interventions may be a beneficial approach for the treatment of viral-induced asthma exacerbation.

We thank Dr. Gerard Kaiko and Dr. Nicole Hansbro for technical assistance in the generation of RSV. We also thank Jennifa Gosling and Brian Wong of Five Prime Therapeutics for support of this study.

This work was funded in part through a collaboration with Five Prime Therapeutics Inc.