NF-κB activation within the epithelium has been implicated in the pathogenesis of asthma, yet the exact role of epithelial NF-κB in allergen-induced inflammation and airway remodeling remains unclear. In the current study, we used an intranasal house dust mite (HDM) extract exposure regimen time course in BALB/c mice to evaluate inflammation, NF-κB activation, airway hyperresponsiveness (AHR), and airway remodeling. We used CC10-IκBαSR transgenic mice to evaluate the functional importance of epithelial NF-κB in response to HDM. After a single exposure of HDM, mRNA expression of proinflammatory mediators was significantly elevated in lung tissue of wild-type (WT) mice, in association with increases in nuclear RelA and RelB, components of the classical and alternative NF-κB pathway, respectively, in the bronchiolar epithelium. In contrast, CC10-IκBαSR mice displayed marked decreases in nuclear RelA and RelB and mRNA expression of proinflammatory mediators compared with WT mice. After 15 challenges with HDM, WT mice exhibited increases in inflammation, AHR, mucus metaplasia, and peribronchiolar fibrosis. CC10-IκBαSR transgenic mice displayed marked decreases in neutrophilic infiltration, tissue damping, and elastance parameters, in association will less peribronchiolar fibrosis and decreases in nuclear RelB in lung tissue. However, central airway resistance and mucus metaplasia remained elevated in CC10-IκBαSR transgenic mice, in association with the continued presence of lymphocytes, and partial decreases in eosinophils and IL-13. The current study demonstrates that following airway exposure with an asthma-relevant allergen, activation of classical and alternative NF-κB pathways occurs within the airway epithelium and may coordinately contribute to allergic inflammation, AHR, and fibrotic airway remodeling.
The NF-κB pathway is a critical regulator of both innate and adaptive immune responses in a wide variety of cell types. Upon stimulation, the IκB kinase (IKK) signalsome, consisting of IKKβ, IKKα, and IKKγ, is activated, leading to IKKβ-mediated phosphorylation of IκBα. Phosphorylation of IκBα in turn leads to its subsequent ubiquitination and degradation by the 26S proteasome, thus allowing for the transcription factor RelA to translocate to the nucleus. This event results in RelA-dependent transcription of genes important in cell survival, proliferation, and inflammation (1, 2). A wide variety of agonists can activate the classical NF-κB pathway in lung epithelial cells and the resultant release of proinflammatory mediators crucial in the recruitment and activation of dendritic cells, lymphocytes, neutrophils, and many other cells in the lung (3). In addition, an alternative NF-κB pathway exists, which requires activation of NF-κB–inducing kinase (NIK) and subsequent phosphorylation of IKKα. IKKα in turn phosphorylates p100, leading to its partial processing to p52. This allows subsequent nuclear translocation of RelB/p52 and transcriptional activation of a subset of NF-κB–dependent genes (4, 5). It was originally thought that the alternative NF-κB pathway played a predominant role in lymphocyte activation and lymphoid organ development (6). However, recent work from our laboratory demonstrated that both classical and alternative NF-κB pathways are activated in lung epithelial cells in response to diverse proinflammatory stimuli and that both pathways coordinately regulate proinflammatory responses (7).
Activation of the classical NF-κB pathway within the airway epithelium has been demonstrated to play a critical role in acute inflammation and allergic airways disease. CC10-IκBαSR transgenic mice, which are refractory to IκBα degradation and NF-κB activation in the lung epithelium, were demonstrated to be strongly protected from airway inflammation induced by LPS (8). Following i.p. sensitization and challenge with OVA, CC10-IκBαSR transgenic mice showed a marked diminution of airway inflammation compared with WT littermate controls, although OVA-induced airway hyperresponsiveness (AHR) was unaffected in CC10-IκBαSR transgenic mice compared with controls (9). A similar protection against OVA-induced allergic inflammation and peribronchiolar fibrosis has been observed in mice following epithelial-specific ablation of IKKβ (10).
It remains unclear to date whether activation of NF-κB within epithelial cells plays a role in the orchestration of inflammatory responses in vivo to an asthma-relevant allergen following sensitization via the airways. It also remains unknown whether both NF-κB pathways are activated following exposure to an Ag. House dust mite (HDM) is a multifaceted allergen to which up to 85% of asthmatics are allergic (11). HDM has been shown to signal through the classical NF-κB pathway in human bronchial epithelial cells in vitro (12). Therefore, the goal of the current study was to determine the activation of classical and alternative NF-κB in epithelial cells in vivo in response to HDM, and to address its effect on HDM-triggered airway inflammation, remodeling, mucus, and AHR. Our results demonstrate the functional importance of epithelial NF-κB in HDM-induced acute inflammatory responses, AHR, and airway remodeling. We also demonstrate activation of both classical and alternative NF-κB pathways in response to HDM. These findings illustrate the complexity of activation of the NF-κB pathways in settings of allergic airways disease and suggest a broader role for epithelial NF-κB in lung disease pathogenesis.
Materials and Methods
CC10-IκBαSR mice were generated as previously described (8) and backcrossed onto the BALB/cJ (n = 10) (The Jackson Laboratory, Bar Harbor, ME) background. Transgene-negative littermates were used as a control. All experiments were approved by the University of Vermont Institutional Animal Care and Use Committee.
A human bronchial epithelial (HBE) cell line was kindly provided by Dr. Albert van der Vliet (University of Vermont) and cultured as described previously (13, 14), and primary human nasal epithelial cells were cultured as described previously (15). Human cell lines were exposed to either PBS or 25 μg HDM (Greer, Lenoir, NC). All protocols that use primary human nasal epithelial cells were approved by the Institutional Review Board.
BALB/cJ mice (The Jackson Laboratory) were subjected to daily intranasal instillation with 50 μg HDM (35 endotoxin units/mg, normalized to protein content) extract resuspended in PBS, or PBS alone as a vehicle control. Briefly, mice were either instilled with 1 dose of HDM and euthanized 2 h, 6 h, or 24 h later or instilled with 3 doses of HDM on 3 consecutive days and euthanized 24 h thereafter. In addition, mice were exposed to HDM 5 d/wk for 1, 2, or 3 wk (5, 10, or 15 instillations, respectively) and harvested 72 h after the final exposure (Fig. 1A).
Assessment of AHR
Mice subjected to 5, 10, or 15 administrations of HDM were anesthetized with an i.p. injection of pentobarbital sodium (90 mg/kg), tracheotomized, and mechanically ventilated at 200 breaths per minute and assessed in response to increasing doses of methacholine (0, 3.125 mg, 12.5 mg, and 25 mg). Respiratory mechanics were assessed with a forced oscillation technique on a computer-controlled small animal ventilator (Flexi-vent, SCIREQ, QC, Canada), as previously described (16, 17), and the parameters Newtonian resistance (Rn), tissue damping (G), and elastance (H) were calculated.
Serum IgG1 and IgE
Following euthanization, blood was collected by heart puncture and immediately spun through a microtainer, and serum was separated. Analysis of serum IgG1 and IgE was performed via ELISA methods, using 1 μg/ml HDM to coat the 96-well plates.
Following euthanization, bronchoalveolar lavage (BAL) was collected using 1 ml PBS. Cell counts were determined (Advia 120 Automated Hematology Analyzer), and differential cells were analyzed by the Hema3 kit (Fisher Scientific, Kalamazoo, MI) by counting a minimum of 300 cells per mouse, as previously described (17).
Histopathology/immunofluorescence/α-smooth muscle actin immunohistochemistry
Following euthanization, the left lobe was inflated with 4% paraformaldehyde and mounted in paraffin-embedded 5-μm sections. H&E and periodic acid–Schiff (PAS) imaging was all performed with three small bronchioles (×20 magnification) per animal. Mucus metaplasia was assessed with a blinded scoring system by two independent investigators with the following scale: 0, no reactivity; 1, minimal staining; 2, moderate staining; and 3, prominent staining. Scores were averaged according to treatment group. Immunofluorescence was performed as previously described to detect nuclear RelA and RelB in situ (18). RelA and RelB Abs were purchased from Santa Cruz Biotechnology. Staining for α-smooth muscle actin (α-SMA) was performed on lung sections following incubation of slides for 20 min in 0.01 M sodium citrate, pH 6.0 at 95°C. Slides were then blocked with 2% normal goat serum for 30 min, incubated with mouse mAb against α-SMA (1:5000 dilution; Sigma Aldrich) overnight at 4°C, and then incubated in biotinylated anti-mouse IgG for 30 min at room temperature. Subsequently, the slides were incubated in avidin-biotin complex–alkaline phosphatase (Vectastain ABC-AP, Vector Laboratories, Burlingame, CA) for another 30 min at room temperature. After rinsing the sections in PBS, the substrate Vector Red (Vector Laboratories) was added for 20 min; this substrate reacts with the bound alkaline phosphatase, thus producing an intense red color. The slides were then counterstained in hematoxylin and imaged (×20 magnification) for analysis.
Assessment of collagen content
Collagen was assessed from the upper right lobe of the lung after an overnight digestion with 10 mg pepsin in 0.5 M acetic acid. Quantification was performed by the Sircol Assay according to the manufacturer’s instructions (Accurate Chemical and Scientific, Westbury, NY). Masson’s trichrome reactivity was evaluated in three bronchioles (×20 magnification) per animal, and analyzed by two blinded investigators with the following scale: 0, no reactivity; 1, minimal staining; 2, moderate staining; and 3, prominent staining.
All data were evaluated using Graphpad Prism 6 Software (Graphpad, San Diego, CA). A one-way ANOVA was used with Bonferroni corrections to adjust for multiple comparisons, and all p values < 0.05 were considered statistically significant. Histopathological/α-SMA immunohistochemistry scores were analyzed using the Kruskal–Wallis test and Dunn multiple comparison post hoc tests.
Inflammatory response in BALB/c mice exposed to HDM
Because of the well-known role of NF-κB in the orchestration of inflammation, we first sought to assess the extent and kinetics of the HDM-induced inflammatory response through a time-course analysis illustrated schematically in Fig. 1A. Neutrophils in BAL were significantly increased 6 h following a single intranasal challenge with HDM, and remained elevated until 24 h post three challenges, compared with PBS controls (Fig. 1B). It is important to note that neutrophils were elevated 24 h following exposure to the vehicle PBS, potentially indicative of a nonspecific response to the intranasal instillation. Neutrophils were also detectable in BAL 72 h post 5, 10, or 15 challenges, with statistically significant increases occurring in response to HDM 72 h post 15 challenges. Although macrophages in BAL tended to increase in response to HDM, these trends were not statistically significant. No eosinophils or lymphocytes were detected in BAL up to 24 h post three challenges. In contrast, robust increases in eosinophils and lymphocytes occurred in BAL 72 h post 10 and 15 challenges with HDM, compared with PBS controls (Fig. 1B). Although some fluctuations in the total number of macrophages in BAL were observed throughout these time points, these were not statistically significant (Fig. 1B).
Increases in nuclear localization of RelA and RelB in airway epithelium in response to HDM
We next evaluated a potential role for epithelial NF-κB in response to HDM. We investigated kinetics of activation of both classical and alternative pathways in the bronchiolar epithelium via the assessment of nuclear presence of RelA and RelB. Nuclear presence of RelA and RelB (indicated by yellow staining) within the bronchiolar epithelium was increased 2 and 6 h following a single challenge with HDM, compared with PBS controls in the nuclei (Fig. 2). Nuclear RelA decreased to control reactivity by 24 h, but increased again 24 h post 3 challenges with HDM, indicative of a bimodal activation pattern (Fig. 2). In contrast, increases in nuclear RelB were apparent throughout the time course evaluated in this study (Fig. 2). No clear evidence for nuclear localization of RelA or RelB within the airway epithelium was apparent 72 h after 5, 10, or 15 challenges with HDM compared with PBS (data not shown).
Proinflammatory mediator expression following acute HDM exposures
To investigate the early inflammatory response after HDM exposure, we evaluated mRNA expression of IL-33, KC, GM-CSF, MIP-2, CCL20, IL-6, IL-25, and thymic stromal lymphopoietin, proinflammatory mediators shown to be produced by epithelial cells. Expression levels of all genes analyzed were significantly increased following 2-h exposure to HDM, with the exception of IL-25 and thymic stromal lymphopoietin, whose expression did not change (data not shown). mRNA expression of CCL20, GM-CSF, MIP-2, KC, and IL-6 remained elevated 6 h after a single administration of HDM and tended to decrease toward control levels thereafter (Supplemental Fig. 1).
Role of epithelial NF-κB in proinflammatory cytokine expression induced by HDM
We next sought to address the role of epithelial NF-κB in HDM-induced proinflammatory responses using CC10-IκBαSR transgenic mice. Although nuclear RelA and RelB content increased in the bronchiolar epithelium in WT mice in response to HDM, these increases were not observed in CC10-IκBαSR mice exposed to HDM (Fig. 3A). HDM-mediated increases in mRNA expression of CCL20, GM-CSF, MIP-2, KC, IL-33, and IL-6 in WT mice were strongly attenuated in CC10-IκBαSR mice (Fig. 3B), demonstrating the importance of NF-κB activation in the bronchiolar epithelium in the orchestration of the acute proinflammatory responses to HDM.
Effects of repeated HDM exposure on AHR, mucus metaplasia, and remodeling
We next subjected mice to 5, 10, and 15 challenges of HDM. Serum content of HDM-specific IgG1 and IgE showed no apparent increases 72 h following 5-d challenge. Marked increases in HDM-specific IgG1 and IgE were apparent following 10 and 15 challenges with HDM, compared with PBS controls, indicative of activation of adaptive immune responses (Supplemental Fig. 2A). AHR, a feature of allergic airways disease, was assessed via forced oscillation mechanics using ascending doses of methacholine. Rn increased significantly over PBS controls after 10 and 15 challenges with HDM, whereas no changes were apparent in mice subjected to 5 challenges with HDM (Supplemental Fig. 2B). G, which is indicative of tissue resistance and small airway dysfunction, was increased following 5 challenges with HDM, and remained increased throughout 10 and 15 challenges. H was also increased following 5, 10, and 15 challenges of HDM. Increases in H were most prominent after 10 challenges of HDM and tended to decrease after 15 challenges (Supplemental Fig. 2B). Histopathological evaluation revealed a robust inflammatory response to HDM, with prominent peribronchiolar and perivascular cellular infiltrates being apparent following 10 and 15 challenges with HDM (Supplemental Fig. 2C). In addition, mucus metaplasia was apparent following 5, 10, and 15 challenges of HDM, based upon staining with PAS reagent and Muc5ac expression (Supplemental Fig. 2C, 2D). Another hallmark of severe allergic airways disease is peribronchiolar fibrotic remodeling. Following 15 challenges with HDM, overall collagen content increased in the lung tissue (Supplemental Fig. 2E), consistent with increases in peribronchiolar collagen deposition evaluated via Masson’s trichrome staining (Supplemental Fig. 2C). These results collectively demonstrate that the HDM exposure regimen used in this study induces a number of the hallmark features of allergic airways disease.
Role of epithelial NF-κB in HDM-induced inflammation
We next determined the role of epithelial NF-κB in the inflammatory response induced following 15 challenges of HDM. HDM-mediated increases in total cell counts in BAL observed in WT mice were significantly decreased in CC10-IκBαSR mice (Fig. 4A). BAL eosinophils were significantly decreased in CC10-IκBαSR transgenic mice exposed to HDM, whereas BAL lymphocytes remained elevated in HDM-exposed CC10-IκBαSR transgenic mice compared with WT littermates (Fig. 4B). Total macrophage number in BAL remained unchanged in CC10-IκBαSR mice (Fig. 4B). In contrast, HDM-mediated increases in BAL neutrophils in WT mice were completely attenuated in CC10-IκBαSR mice. Consistent with these observations, increases in BAL content of the neutrophil chemoattractant KC in WT mice exposed to HDM were absent in CC10-IκBαSR mice (Fig. 4C). HDM-specific IgG1 and IgE levels in serum were decreased slightly in CC10-IκBαSR HDM-exposed mice, although not significantly, compared with littermate controls, but remained elevated over PBS controls (Fig. 4D). Levels of IFN-γ, IL-13, and IL-17A in homogenized lung tissue were significantly decreased in CC10–NF-κBSR mice exposed to HDM, compared with WT animals (Fig. 4E), collectively suggesting an attenuation of Th-mediated responses to HDM following epithelial-specific inhibition of NF-κB. We next sought to assess inhibition of nuclear RelA and RelB in lung tissue in CC10–NF-κBSR mice exposed to HDM. After 15 challenges with HDM, no consistent increases in nuclear RelA or RelB were detected in the bronchiolar epithelium (data not shown), in contrast to earlier time points at which increases in nuclear RelA and RelB were readily observed (Fig. 3). Assessment of homogenized lung tissue also showed variable and inconsistent patterns of nuclear RelA (data not shown). However, clear increases in nuclear RelB content were observed in WT mice, which were absent in the CC10-IκBαSR transgenics (Fig. 4F).
Role of epithelial NF-κB in HDM-induced AHR and remodeling
WT and CC10-IκBαSR mice subjected to 15 challenges with HDM demonstrated equivalent increases in Rn. In contrast, HDM-mediated increases in G and H were significantly decreased in CC10-IκBαSR mice compared with littermate controls (Fig. 5A). Consistent with attenuated inflammatory cell profiles in BAL, peribronchiolar infiltrates were attenuated in CC10-IκBαSR mice compared with controls (Fig. 5B). In contrast, mucus metaplasia was not observed to be decreased in HDM-exposed CC10-IκBαSR mice, compared with WT animals (Fig. 5B, 5D). These findings are consistent with a lack of attenuation of HDM-induced Muc5AC mRNA (Fig. 5C) and modest decreases in IL-13 content (Fig. 4E). Biochemical analysis of collagen demonstrated that HDM-mediated increases in WT mice were significantly decreased in CC10-IκBαSR transgenic mice (Fig. 5E), consistent with less peribronchiolar collagen detected via histopathological analysis (Fig. 5B, 5F). Furthermore, α-SMA, a known marker of airway thickening and remodeling, was assessed via immunohistochemistry. Peribronchiolar α-SMA staining was significantly increased following HDM administration in WT mice, as compared with PBS controls. In contrast, no increases in peribronchiolar α-SMA reactivity were detected in CC10-IκBαSR mice exposed to HDM, in comparison with PBS controls (Fig. 5G, 5H).
HDM-induced activation of the classical and alternative NF-κB pathways in human nasal and HBEs
Because we demonstrated increases in nuclear RelB in the bronchial epithelium in response to HDM administration, we sought to determine whether the alternative NF-κB pathway could be activated directly by HDM in human epithelial cells. Results in Fig. 6A demonstrate increased levels of NIK and p52 in response to HDM in primary human nasal epithelial cells obtained from two independent donors, and similar increases were observed (Fig. 6B) in HBE cells, albeit less robust, demonstrating activation of the alternative NF-κB pathway. In addition, phosphorylation of RelA at serine 536, a separate parameter of NF-κB activation, was also increased in response to HDM (Fig. 6A, 6B), demonstrating that in addition to the known ability of HDM to activate the classical NF-κB pathway, it also induces activation of the alternative NF-κB pathway in lung epithelial cells.
NF-κB is a regulator of inflammation and immunity, and its role in the pathogenesis in asthma has been suggested on the basis of evidence of its activation in the bronchiolar epithelium from asthmatics (19) and from studies in mouse models of allergic airways disease (18). Notably, studies previously performed in our laboratory have revealed a crucial role for lung epithelial NF-κB in the ALUM/OVA model of allergic airways disease (9). Moreover, transgenic expression of constitutively active IKKβ in lung epithelial cells was sufficient to cause neutrophilic inflammation and AHR, as well as enhanced sensitization to an inhaled Ag (17, 20). Despite these prior observations, the more generalized importance of epithelial NF-κB in allergic airways disease following sensitization/exposure to a relevant allergen via the airways has yet to be determined. Results presented in this article describe a critical role for nonciliated airway epithelial NF-κB in promoting inflammation, AHR, and fibrotic remodeling following extended challenges of HDM, whereas differences in mucus metaplasia did not appear affected.
Results of the current study somewhat contrast our previous work using the ALUM/OVA model of i.p. sensitization followed by challenges of aerosolized OVA. Notably, CC10-IκBαSR mice subjected to the ALUM/OVA protocol were strongly protected against the development of eosinophilic inflammation and mucus metaplasia, but were not protected against OVA-induced AHR (9). The reasons for these discrepant findings remain unclear. It is likely that the route of sensitization of an allergen dictates the nature of the subsequent immunological and pathophysiological response. In support of the latter is a previous report demonstrating that the i.p. sensitization regimen with ALUM/OVA triggers eosinophilic, Th2-driven inflammation (21). In contrast, sensitization with OVA via the airways, along with LPS as the adjuvant, led to neutrophilic-dependent AHR and Th17-associated inflammation (22).
In the current study, we demonstrated that CC10-IκBαSR mice did not display increases in airway neutrophils following 15 challenges with HDM compared with PBS controls, whereas neutrophils were robustly increased in BAL of WT mice. In contrast, HDM-induced eosinophilia was only partially attenuated in CC10-IκBαSR mice, whereas BAL lymphocytes were not significantly affected. IL-5, a potent eosinophil chemoattractant in mice and humans, was not detectable at any experimental time points (data not shown); however, we did demonstrate a strong decrease in KC (human IL-8) in CC10-IκBαSR mice in comparison with WT mice exposed to HDM. These findings demonstrate that following inhibition of epithelial NF-κB in the setting of HDM-induced disease, the inflammatory process is not uniformly inhibited but preferentially affects neutrophils. This potential differential effect of NF-κB inhibition on the inflammatory process could explain the impact of airway remodeling and AHR. Consistent with the continued presence of eosinophils, HDM-specific IgG1 and IgE, and IL-13 in CC10-IκBαSR mice exposed to HDM, mucus metaplasia and Muc5AC expression remained elevated in these animals. Previous studies demonstrated that IL-13 can activate epithelial cells to produce Muc5ac by an NF-κB–independent mechanism (23). In light of those observations, the lack of an impact of CC10-IκBαSR mice in HDM-induced mucus metaplasia is, therefore, perhaps, not surprising.
The role of neutrophils in the pathogenesis of asthma remains unclear. HDM exposure in mice has been associated with mixed neutrophilic/eosinophilic, Th2/Th17-linked inflammation, as well as production of IL-13 and IL-17A (24). In this study, the preferential diminution of neutrophils following inhibition of epithelial NF-κB was associated with normalization of G and H parameters toward values observed in controls; in contrast, Rn, reflective of the central airways, was unaffected. The disparate effects of the CC10-IκBαSR transgene on G and H, compared with Rn, are puzzling. Previous studies have suggested that inflammation contributed to enhanced leakage of fibrin to the airway surface, leading to decreased stability of surfactant proteins and increased surface tension, and facilitated contractility of smooth muscle. These mechanisms have been linked to increases in the closure of distal airways (25–27). It is plausible that decreases in inflammation, along with decreased α-SMA in bronchioles, account for decreases in G and H, which were observed in CC10-IκBαSR transgenic mice. This putative explanation would need to be addressed with additional studies. Our data also demonstrate that KC, a potent neutrophil chemokine, was not increased in CC10-IκBαSR HDM-exposed mice in comparison with PBS controls, indicating a potential mechanism whereby neutrophils are decreased in CC10-IκBαSR mice. Altogether, these findings are suggestive of a role for neutrophils in promoting increases in peribronchiolar collagen deposition and associated changes in G and H. Alternatively, it is also plausible that other mediators control neutrophil trafficking to the airways, promote peribronchiolar remodeling, and produce changes in AHR. In this regard, IL-17A, which we also demonstrate to be decreased in the tissue of CC10-IκBαSR mice, has been shown to stimulate production of chemokines important in neutrophil recruitment in a KC-dependent manner (28) and has been implicated in the pathogenesis of pulmonary fibrosis (29); its functional contribution, however, remains to be determined. In addition, eosinophils expressing TGF-β1 have been shown to be important in allergen-induced peribronchial fibrosis (30); hence it is possible that TGF-β1 also plays a role in peribronchiolar fibrosis downstream of activation of NF-κB. Additional studies will be required to formally address these putative scenarios.
In addition to demonstrating a role for epithelial NF-κB in promoting inflammation, AHR, and fibrosis following 15 challenges of HDM, we also established a contributing role for epithelial NF-κB in regulating proinflammatory gene expression following a single exposure of HDM. Numerous proinflammatory mediators have been implicated in the development of allergic airways disease (3), and our studies demonstrate a putative role for epithelial NF-κB in regulating expression of several of these molecules in response to HDM. GM-CSF and CCL20 have been shown to be important in the recruitment/activation of dendritic cells, which are required for T cell activation and recognition of Ags. Notably, exposure of human asthmatic bronchial epithelial cells to the HDM component, Derp1, was shown to be important in dendritic cell recruitment in a CCL20-dependent manner (31). In addition, IL-33 has been shown to be important in the activation of a variety of cell types crucial to the development of asthma, such as T helper type 2 cells, eosinophils, dendritic cells, and mast cells (32). As previously mentioned, the activation and infiltration of neutrophils is emerging as a potential phenotype in severe, steroid-resistant asthma. Cytokines KC and MIP-2 have both been shown to be important in the recruitment of neutrophils and production of HDM-specific IgE (33). Altogether, our data indicate a likely role for epithelial NF-κB in the recruitment/activation of several cell types important for the development of HDM-induced asthma. This observation suggests a mechanism whereby epithelial NF-κB activation is the crucial step between contact with an allergen and downstream manifestations of asthma.
In addition to the role of NF-κB in airway epithelium demonstrated in this study, it is plausible that activation NF-κB in other cell types contributes to the pathophysiology of allergic airways disease. Notably, our findings demonstrate increased immunofluorescence of nuclear RelA and RelB in parenchymal regions following exposure to HDM (Fig. 3). Unraveling the cell types in which NF-κB is activated and their functional contribution to allergic airways disease would require additional cell-specific labeling and targeting strategies, which were beyond the scope of the current study. For example, neutrophil elastase–induced secretion of TGFβ-1 from smooth muscle cells was shown to be dependent on NF-κB activation (34), and the smooth muscle contractile force in response to IL-17A was dependent on NF-κB activation (24), suggesting a putative role of NF-κB activation in smooth muscle cells in AHR and remodeling. Secretion of eotaxin, a potent eosinophil-activating factor, by fibroblasts was also shown to be dependent on NF-κB activation (35). One notable finding of the current study is that HDM activates both the classical and the alternative NF-κB pathway within the bronchiolar epithelium, as evidenced by the increases in the nuclear presence of both RelA and RelB in mice exposed to HDM. Similar to observations in this study, we recently demonstrated increases in nuclear RelA and RelB in the parenchymal regions following administration of LPS, a component of HDM (7), suggesting that TLR4 activation by HDM may be contributing to the observed increases in RelA and RelB observed. Increases in nuclear RelA and RelB within the epithelium occurred rapidly and were sustained for at least 24 h after three challenges. However, 72 h post 5, 10, or 15 challenges, there was no clear evidence of increased RelA or RelB in the bronchiolar epithelium, possibly owing to the timing of tissue analysis after the last administration of HDM, which was 72 h, a time when NF-κB activation may have resolved. However, we did observe sustained increases in nuclear RelB in lung tissue homogenates 72 h after 15 challenges with HDM in WT mice, suggestive of NF-κB activation in other cell types. Increases in nuclear RelB were attenuated in CC10-IκBαSR mice, suggesting a putative role for RelB in the orchestration of HDM-mediated AHR and fibrotic remodeling. In addition to the demonstration that RelB was increased in the bronchiolar epithelium following HDM exposure, we also demonstrated that HDM directly activated both NF-κB pathways in both human bronchial and nasal epithelial cells. Although extensive studies have been conducted to unravel the molecular regulation and pathophysiological relevance of the classical NF-κB pathway, far less information is available for the alternative pathway. The latter pathway was originally thought to play a role in adaptive immune responses and in development of lymphocytes and lymphoid organs. However, emerging studies have pointed to a coordinate function of both classical and alternative NF-κB in the orchestration of proinflammatory responses. For example, exposure of lung epithelial cells with TNF-α, polyinosinic acid, LPS, IL-17A, lipoteichoic acid, and CD40L, agonists that signal through distinct families of receptors, led to a coordinate activation of classical and alternative NF-κB pathways and subsequent proinflammatory responses (7). In contrast, adenovirus-mediated delivery of RelB afforded protection against cigarette smoke–induced neutrophilic inflammation (36), suggesting potentially complex roles for the alternative NF-κB pathway in the regulation of proinflammatory and immune responses. The classical NF-κB pathway previously has been shown to be important in HDM-mediated proinflammatory responses in HBE cells in vitro (12), and Derp1, a component of HDM, was shown to be important in the activation of NF-κB in human asthmatic bronchial epithelial cells (37). Patients with allergic asthma demonstrate increased classical NF-κB activation in nasal epithelial cells in response to HDM in comparison with healthy controls (38). Intriguingly, the HDM component, β-glucan, was shown to be crucial for activation of allergic rhinitis in nasal epithelial cells via TLR2, in contrast to another HDM component, LPS, which is important in promoting allergic airways disease via activation of TLR4 in bronchial epithelial cells (39). Despite these previous studies and data presented in this study, additional studies are needed to better understand the timing and locale of activation of classical and alternative NF-κB pathways, the components of HDM that are required to activate either pathway, as well as the relative contributions of the classical and alternative NF-κB pathways in eliciting HDM-triggered allergic airways disease.
In summary, we demonstrate in the current study the importance of NF-κB activation within the bronchiolar epithelium in HDM-induced inflammation, AHR, and fibrotic airway remodeling. We also demonstrate that both classical and alternative NF-κB pathways are activated by HDM. Data presented in this study showed that in the setting of HDM-induced allergic airways disease, inhibition of epithelial NF-κB plays a more prominent role in attenuating neutrophilia, AHR, and remodeling, in comparison with eosinophilia, IgE, and mucus metaplasia. Therefore, it is plausible that therapeutic approaches that target NF-κB via interference with degradation of IκBα may have a stronger impact on asthmatic patients with predominant neutrophila in contrast to patients with predominant eosinophilia. Corticosteroids, the most common therapy for asthma, inhibit NF-κB, but have many off-target effects. Current therapies that are being developed for asthma are aimed at inhibiting the proinflammatory effects of NF-κB signaling in the lung (40) and are focused on inhibition of IKKβ, the dominant kinase in the classical NF-κB pathway (40). The small-molecule IKKβ inhibitor, IMD-0354, attenuated HDM-induced eosinophilia, globlet cell hyperplasia, subepithelial fibrosis, smooth muscle cell hypertrophy, and lung resistance using an i.p. sensitization model (41). The disparate findings of the latter study relative to the present findings may relate to the different sensitization route, the mechanism of inhibition of NF-κB, and the cell types wherein NF-κB inhibition occurred. Furthermore, i.v. administration of RelA antisense oligonucleotides prior to OVA challenge resulted in dampened responses of inflammation, AHR, and TH2 responses in mice (42). On the basis of these collective findings, therapeutics designed to inhibit both facets of the NF-κB pathway may hold larger therapeutic potential for allergic asthma and other allergic diseases of the lung.
The authors have no financial conflicts of interest.
We thank the University of Vermont Microscopy Imaging Center and the Vermont Lung Center, Burlington, VT, for assistance with these studies.
This work was supported by National Institutes of Health Grants T32 HL076122, T32 ES07122, NIGMS P30, GM103532, and R01 HL060014, and by an American Thoracic Society unrestricted grant and a Parker B. Francis fellowship (to V.A.).
The online version of this article contains supplemental material.
Abbreviations used in this article:
- airway hyperresponsiveness
- bronchoalveolar lavage
- chemokine (C-C motif) ligand 20
- tissue damping (parameter)
- elastance (parameter)
- human bronchial epithelial (cell)
- house dust mite
- IκB kinase
- keratinocyte-derived chemokine
- NF-κB–inducing kinase
- periodic acid–Schiff
- Newtonian resistance (parameter)
- α-smooth muscle actin
- Received May 20, 2013.
- Accepted October 11, 2013.
- Copyright © 2013 by The American Association of Immunologists, Inc.