The Lung Microbiome, Immunity, and the Pathogenesis of Chronic Lung Disease

The microbiome is defined as the “ecological community of commensal, symbiotic and pathogenic organisms that share our body space” (1). Most studies of host and microbiome interaction in the human have focused almost exclusively on bacteria, biotic factors, and the host. These complex communities of microbiota that inhabit environments such as the lung, skin, or gut are now appreciated for their role in maintaining organ, tissue, and immune homeostasis. One striking example is the early observation that germ-free mice have absent/impaired secondary lymphoid architecture with resulting loss of lymphoid cells (2). Additionally, commensal microbiota can have both systemic and site-specific autonomous immune effects. For example, Staphylococcus epidermidis colonization of the skin promotes CD4 cell IFN-γ production, which protects against infection with the parasite Leishmania major. In contrast, colonization of the gut with S. epidermidis had no effect (3). In other situations, it is well established that alterations of the gut microbiome can influence immune responses at distal sites. Antibiotic treatment, which disrupts gut microbiota accompanied by increases in fungal colonization, can greatly exaggerate the allergic response to intranasal challenge with the mold spore Aspergillus fumigatus. With antibiotic treatment, mice showed increased levels of eosinophils, mast cells, IL-5, IL-13, IFN-γ, IgE, and mucus-secreting cells (4). More recently, modulation of gut microbiota through the use of probiotics has been shown to increase the frequency of B cells expressing IgA in the colon and lymph nodes, likely secondary to increased lymph node T follicular helper cells and IL-23–expressing dendritic cells (5), all changes that are likely to improve host defense at mucosal sites or response to vaccination. Conversely, antibiotic treatment can limit development of T follicular helper cells (6). The response of pathogen recognition receptors (PRRs) of the innate immune response in the lung is well described. We know that defective components of the innate response can predispose to overwhelming infection and mortality and in some cases reduce injury from pathogens (7–9). The relationship between resident microbiota and this flagship innate response and the subsequent adaptive immune response in the lung is poorly understood. With this backdrop, we have chosen to explore what is known about the potential role of the microbiota (both gut and lung) and host interaction in regulating the pathogenesis of several important lung diseases.

The lung is an organ constantly exposed to microbiota either through inhalation or subclinical microaspiration from birth. Historically, medical texts allude to a sterile lung environment, and this dogma has persisted in contemporary medicine. In the last decade, a revolution of sorts has taken place in our understanding of how the lung and microbiota interact and exist. This revolution stems from new knowledge that the lung is not sterile (10) and, in fact, harbors an abundance of diverse interacting microbiota. As mentioned above, the gut microbiome modulates host mucosal defense (11, 12); however, there is a paucity of information regarding the potential role of lung microbiota to regulate immunity and homeostasis.

The lung is not sterile, contrary to centuries of dogma asserting the same. Throughout the 1900s this inference was reinforced by respiratory culture-based protocols that sought only to identify clinically significant pathogens and by a spurious conclusion that upper respiratory tract microbes cultured from the lung represented contamination (10). The lung is a warm environment exposed to 7000 l of diverse microbe-ridden air every day (13). Microbes adapt and exist in hostile environments similar to polar ice sheets of Antarctica and hot sulfur springs of Japan, yet the idea the human lung existed in a sterile state was “cultivated” in the medical literature for decades (14, 15). The upper respiratory tract and oropharynx, where microbes are found in abundance (16), is in direct communication with the lung, and subclinical aspiration of oropharyngeal contents occurs universally in humans (17, 18). Thus, the lung is subject to a constant level of microbe immigration and elimination through host mucosal defense and mucociliary clearance. Modern studies employing sequence-based bacterial identification techniques support the presence of a consistently detected diverse bacterial signal in the lung of healthy humans (19–22). Therefore, the notion of lung sterility has been debunked.

This new understanding was animated by the discovery of culture-independent techniques for bacterial identification. The most commonly used approach involves high-throughput sequencing of amplicons of the 16S rRNA gene, a highly conserved locus in the bacterial genome. Sequences are then categorized and classified according to publically available prior knowledge taxonomic databases to allow for measures of total and relative abundance and diversity. The first application of culture-independent techniques was undertaken in a cohort of healthy controls and asthma patients. Hilty et al. (20) reported that healthy airways contain bacteria similar to, but distinct from, the upper respiratory tract, and airways of asthma patients were enriched with the phylum Proteobacteria.

Sampling the lung for microbiome sequencing is technically challenging given the relatively low biomass. Furthermore, sampling lower airways by bronchoscopy requires passage of the instrument through the oral or nasal route. This course allows for a theoretical risk of pharyngeal contamination of samples. Importantly, the mouth and nose microbiota are markedly different and studies have not identified any detectable influence on the reported microbiota based on scope insertion site (22, 23). Additionally, if pharyngeal carryover from the bronchoscope insertion site was heavily influencing the reported microbiota, then serial dilutions of samples should result in dilution of bacterial communities and signal. Several studies have established that this is not the case (19, 24). The evidence thus supports minimal contamination from pharyngeal sources acquired by bronchoscopic methods.

Accurate and relevant studies of the microbiome require consideration of several principles. Representative samples must be acquired from the distal airways, and, as discussed above, the possibility of contamination from other niche microbiota must be addressed. Furthermore, contamination can occur during sample processing; use of “no template” controls is essential in low-biomass studies to assess the effects of reagent contamination. The extraction of nucleic acid requires lysis of species, some of which are more susceptible to cell disruption than others. This discrepancy can lead to overrepresentation of some species over others. Further steps that may alter the accuracy and reliability of acquired data include the generation of appropriate PCR primers, data normalization, the choice of reference database, and divergent measures of diversity (25). The use of 16S rRNA sequencing remains a pillar step in the sample processing; however, 16S rRNA sequencing may not be able to differentiate between species with varying immunogenicity and pathogenicity (26). The microbiome is subject to a number of factors that are known to change its composition, including age, diet, ethnicity, and study design, and in humans requires careful management of these potential influences (27–29). However, in tandem, researchers are developing novel and ingenious methods to limit any possible error in microbiome studies. For example, as an alternate process to 16S rRNA sequencing and primer choice, metagenomic data are being generated through the sequencing of all DNA from a sample (shotgun sequencing), which may be even more informative (30). Culture-based techniques remain highly relevant as complements to culture-independent techniques in determination of viability, in speciation, and in microbial phenotyping.

The human microbiota inhabits several organs and is primarily colonized by six phyla: Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, Fusobacteria, and Cyanobacteria (31–34). Murine studies have demonstrated that bacterial load in the lungs increases during the first 2 wk of life, and the phyla of organisms found in the lung shifted from Gammaproteobacteria and Firmicutes toward Bacteriodetes (35). Such developmental changes in the microbiota were associated with accumulation of a PD-L1–dependent T regulatory cell population that could promote tolerance to allergen challenge (35). These data suggest that acquisition of a lung microbiome is an important early life event necessary to protect the lung from injurious responses to inhaled Ags. In humans, studies have largely focused on gut microbiota and shown that newborns acquire microbiota that resembles their mother quickly and in a manner specific to the method of delivery. Dominguez-Bello et al. (36) reported that infants born by vaginal delivery acquired bacterial communities resembling their mother’s vaginal microbiota, dominated by Lactobacillus, Prevotella, or Sneathia species. Infants born by cesarean section acquired skin-predominant Staphylococcus, Corynebacterium, and Propionibacterium species. These communities were undifferentiated across multiple body habitats in the infants, in contrast to the diverse communities evident in the mothers. No studies to date have examined the dynamic changes that may occur in the lower respiratory tract microbiota as childhood progresses. However, it is likely from studies of the upper respiratory tract and intestinal microbiota that these bacterial communities are dynamic (37, 38). There is relatively low bacterial biomass in the human lung. Bacterial loads from bronchoalveolar lavage have reported ranges from 4.5 to 8.25 log copies/ml (39, 40). Further analysis of lung tissue samples demonstrates some 10–100 bacterial cells per 1000 human cells (41). The healthy lung has been studied using culture-independent techniques, and the predominant phyla are Bacteroidetes and Firmicutes (21, 24). Although individuals exhibit some spatial variation in the microbiota of their respiratory tract, intrasubject variation is significantly less than that of intersubject variation (39).

The lung is a relatively low nutrient resource compared with the intestinal tract for supporting microbiome development. Furthermore, physiological conditions are regionally variable even in healthy lungs. Conditions that affect bacterial proliferation include oxygen tension, blood flow, local pH, temperature, effector inflammatory cell disposition, and epithelial cell architecture (42, 43). Coupled to this variable biogeography of the lung microbiome are the factors that influence microbe immigration and elimination from the lower respiratory tract. Taken together, these factors determine the dynamic change of the microbial ecosystem of the lung.

Lung disease alters the population dynamics through effects on immigration/elimination and the local conditions of the microbial ecosystem of the lung (Table I). Chronic lung disease in many forms results in heterogeneous architectural distortion of the lung, including upper lobe predominant destruction of the terminal bronchioles in emphysema and lower lobe predominant distortion of the parenchyma by honeycombing in idiopathic pulmonary fibrosis (IPF) (44). Changes in the viscosity of the mucus and pH occur in cystic fibrosis (CF) (45). The resultant changes in oxygen tension, ventilation, perfusion, inflammatory cells, and other local factors exert pressure on population dynamics. Immigration of microbiota from the upper respiratory to the lower respiratory tract is primarily promoted by subclinical aspiration, which occurs in both health and disease, and overt clinical infection occurs when local defense is blunted or overwhelmed (17, 18). Chronic lung disease is commonly associated with gastroesophageal reflux, which may result in elevated volumes of microaspiration (46, 47). Elimination is achieved by cough and mucociliary clearance. Host inflammatory cells are responsible for eradication of pathogens, and the type and number of effector cells are associated with certain features of the microbiome. In a comparison of inflammatory cells and microbiota detected in bronchoalveolar lavage fluid, Segal et al. (24) demonstrated increased community abundance of Prevotella and Veillonella species (common anaerobic oral commensals) associated with higher levels of lymphocyte and neutrophil inflammation. Therefore, the lung microbiome has a potential role to play in the pathogenesis of chronic lung disease through both the ability of lung microbiota to modulate local inflammatory responses and the influence of chronic lung diseases on the lung microbiome in turn.

Chronic lung diseases include asthma, chronic obstructive pulmonary disease (COPD), CF, and IPF. Interestingly, these diseases are all characterized by natural histories that are punctuated by periods of acute exacerbations. Exacerbations are characterized by acute worsening of pulmonary symptoms and a decline in pulmonary function. Such exacerbations are responsible for significant mortality and morbidity. The onset of exacerbation may herald accelerated disease progression, and many patients fail to return to their previous functional and physiological baselines (48). Studies of these events may reveal key data that reanimate our current understanding of the pathogenesis of chronic lung disease, and it is likely that these exacerbations are accompanied or induced by microbiota dysbiosis.

The relationship between COPD exacerbation and acute bacterial infection of the airway remains disputed. Potential pathogens cultured from sputum during COPD exacerbations are less frequently cultured during periods of clinical stability (49). Sethi et al. (50) identified similar culture densities of Haemophilus influenzae and lower densities of Moraxella catarrhalis and Streptococcus pneumoniae in sputum collected during acute exacerbations compared with samples during clinical stability. The use of antibiotic therapy in COPD exacerbations also lacks clarity. Recent work reported a clear role in reducing the rate of treatment failure for severe disease in hospitalized patients, but the role is unclear for mild to moderate disease (51). Culture-independent techniques have identified a diverse and abundant pulmonary microbiota in exacerbations from a variety of sampling types (41, 52–54), and exacerbations are definitively associated with changes in respiratory microbiota and airway inflammation. Millares et al. (52) analyzed sputum from COPD patients during exacerbations with paired sampling from periods of clinical stability and found increases in the relative abundance of bacteria associated with exacerbations, namely Haemophilus, Pseudomonas, and Moraxella. Huang et al. (53) reported an alteration in community content toward the Proteobacteria phylum during COPD exacerbations, including nontypical COPD pathogens. Additionally, the influence of viral exposure may trigger COPD exacerbations (55) but the relationship between viral infection, microbiome composition, and host defense is poorly understood. Patients experimentally infected with rhinovirus develop clinical features of COPD exacerbation, and culprit viruses have been isolated in respiratory samples from 36 to 56% of patients with exacerbations versus 6–19% of patients at clinical baseline (56–59). Interestingly, Molyneaux et al. (60) compared sputum in COPD patients at baseline and during exacerbations and noted that sputum acquired after viral exposure demonstrated a shift toward the Proteobacteria phylum, a potential explanation for the increased presence of Pseudomonas spp. noted in COPD exacerbations (52). Sequencing-based studies of tissue from COPD patients have demonstrated an increase in the Firmicutes community in severe disease (GOLD stage 4) attributable to an increase in the Lactobacillus genus (41). Phagocytosis of Lactobacillus spp. by human macrophages reduces the effects of cigarette smoke–related inflammation, potentially suggesting that these species are beneficial modifiers of smoking-related lung disease (61). Animal models of respiratory syncytial virus infection have demonstrated that the antiviral response within the lung mucosa can be modulated by the administration of Lactobacillus rhamnosus species prior to infection (62). Thus, changes in the microbiome may represent an adaptation to try to protect the lung from respiratory viral infection. However, we speculate that, as suggested above, once pathogenic infection does occur, these beneficial changes may be lost.

Studies of airway microbiota in asthma have established that composition is altered when compared with controls. Hilty et al. (20) identified asthma patients with more frequent Proteobacteria (particularly Haemophilus) in the bronchial tree compared with controls. The authors also noted a decrease in Bacteroidetes, especially Prevotella species, in asthmatic airways. Studies of asthma severity identified similar altered microbiomes with a predominance of Proteobacteria and found that the airway microbiome of asthmatic patients was more diverse that of nonasthmatic controls (63). Huang et al. (64) reported an association between bronchial hyperresponsiveness and community diversity and composition, secondary to an increased abundance of Proteobacteria. Bronchial hyperresponsiveness is accentuated during exacerbation and is an accurate predictor of future exacerbations (65). Alterations to the microbiome appear similar in both mild and severe disease and are specifically associated with features of the disease. No studies have analyzed the airway microbiome in asthma exacerbations. However, an estimated 80% of asthma exacerbations are associated with viral infection (66, 67). The host microbiome interaction may be crucial in the development of asthma. Ege et al. (68) have demonstrated that children with broad microbial exposures (i.e., traditional farming) were protected from asthma and atopy in childhood. Further studies reported an association with high-fiber diet and a reduced risk of asthma (69). The proposed mechanism was related, in part, to an altered immune response. Mice fed a low-fiber diet exhibited reduced levels of short-chain fatty acids with increased allergic inflammation, whereas mice with a high-fiber diet had elevated levels of short-chain fatty acids and were protected against allergic inflammation. Short-chain fatty acid propionate treatment of mice resulted in the generation of macrophages and dendritic cells with enhanced phagocytic properties but an attenuated capacity to initiate a Th2 response, a crucial component of allergic inflammation (69). Furthermore, studies have suggested that resident microbiota may promote Th17-dependent neutrophil inflammation in a murine model of OVA-induced asthma (70). Similarly, an experimental model of allergic airway inflammation is exacerbated by the administration of antibiotics during early life. This correlated with a reduction in microbial load and diversity (71).

IPF is a chronic fatal remodeling disease of the lung parenchyma of unknown etiology (72). The natural history of IPF is characterized by exacerbations that contribute greatly to disease-related morbidity and mortality. Recent work has highlighted a potential role for both viral and bacterial infection in the pathogenesis of IPF (73–77). Unlike with asthma or COPD, the current definition of acute exacerbations of IPF excludes active infectious pathogens (78). Disease progression in IPF is characterized by an alteration in the microbiome with a relative increased abundance of Streptococcus and Staphylococcus taxonomic groups (79). This has particular relevance given recent work identifying pneumolysin, a Pneumococcus-produced toxin, that mediates fibrotic progression in animal models via injury of the alveolar epithelium (73). Further study by Molyneaux et al. (76) described an increased bacterial burden in the bronchoalveolar lavage fluid of IPF patients compared with controls using culture-independent techniques. These communities were enriched with Haemophilus, Streptococcus, Neisseria, and Veillonella. The greater the bacterial burden in these patients, the greater the independent association with IPF disease progression. Recent trials of trimethoprim/sulfamethoxazole have demonstrated benefit with improved Medical Research Council dyspnea scores, quality of life, and even all-cause mortality (hazard ratio, 0.21; 95% confidence interval, 0.06–0.78; p = 0.02) (80). This further supports a role for bacterial burden in disease progression. Aspects of host defense and innate immunity also have putative roles in IPF disease progression (81–86). We have previously suggested that defective TLR3 signaling promotes IPF disease progression (81). The L412F polymorphism (rs3775291) of TLR3 results in a functional defect in primary lung fibroblasts from IPF patients. This defect leads to aberrant inflammation and blunted IFN responses to TLR3 activation by synthetic dsRNA (and likely pathogen-associated molecular patterns, although this was not directly examined). Genotyping studies of two independent cohorts of IPF patients confirmed an association between this polymorphism with increased mortality and functional decline. However, the interaction between the pulmonary microbiome, host defense, acute infection, and IPF disease progression remains unclear. Ultimately, the questions of whether chronic lung disease promotes microbiome alterations or microbiome changes modify chronic lung disease remain to be answered.

The manifestations of CF in the lung involve the progressive development of bronchiectasis and obstructive lung disease. Central to disease progression are exacerbations of CF bronchiectasis, which are responsible for significant mortality, morbidity, and accelerated disease progression (87). Exacerbations of disease are attributed to infection by specific pathogens that are cultured from sputum during exacerbations and clinical stability. These pathogens commonly include Staphylococcus aureus and Pseudomonas aeruginosa (88). The evidence to support the use of antibiotics directed against these pathogens is sparse. The bacterial density of sputum is not altered during CF exacerbations when antibiotics are administered (89). Clinical trials have not reported an association between the clinical response during antibiotic therapy and the in vitro susceptibility of the cultured bacteria to the administered antibiotics (90, 91). Indeed, culture-independent analysis has primed a revision of our long-held understanding of the bacterial pathogenesis of CF lung disease. Studies have consistently reported that CF exacerbations are not associated with increased bacterial density or attenuated diversity (92–94). However, evidence would support a loss of diversity with increasing age and disease severity, which was strongly associated with cumulative antibiotic exposure (95, 96). The emergence of new pathogens may have implications for our understanding of the microbiome and lung disease interaction in CF. Nontuberculosis Mycobacterium, in particular M. abscessus, is associated with increased mortality and morbidity in CF (97–99). Guidelines for the management of M. abscessus have been published, and long-term treatment with broad-spectrum antibiotics is required in these cases (100). The consequences for the lung microbiome for this treatment remain unknown. This is of particular relevance given the limited treatment benefit for M. abscessus infection in CF (100). The relationship between the microbiome, antibiotic exposure, exacerbation, and ultimate disease progression will thus require further careful study in CF.

Exactly how microbiota may regulate innate immunity in health and disease is an area of active investigation, and very little is known about how the lung microbiota may specifically regulate lung immunity or the development of bronchial-associated lymphoid tissue. There is growing appreciation for the fact that the gut commensal microbiota is an important regulator of the innate immune system (101, 102). The bacterial biomass of the intestine dwarfs the relative mass seen in the lung (103). In healthy adults the intestine microbiota consists predominantly of three phyla: Bacteroides, Prevotella, and Ruminococcus (104). There is evidence to support a crucial early period during life where intestinal microbiome development is important for the regulation of an appropriate immune response in the lung. CF and asthma are examples of chronic lung disease where disease course and susceptibility are influenced by shifts in the composition of the gut microbiota (105, 106). Furthermore, in the absence of normal gut biota, the host is more susceptible to pulmonary infections, including Listeria monocytogenes (107), Klebsiella pneumoniae (108), and viruses (109). This raises the interesting possibility that exacerbations of chronic lung disease may arise from impaired innate and adaptive immune function secondary to alterations in the host gut microbiota. As mentioned above, patients with progressive IPF show evidence of enhanced burden of Streptococcus and Staphylococcus species in the lung, and previous studies have shown that the ability of neutrophils from microbiota-depleted mice to kill S. pneumoniae and S. aureus are reduced (110). It is not known at present whether the accumulation of these species in the lung correlates with loss of gut microbial communities during IPF disease progression, but this is an interesting area for future study (Fig. 1). Support for such a concept comes from recent work showing that bacterial activation of Nod-like receptors in the gut leads to enhanced production of reactive oxygen species within alveolar macrophages, the sentinel innate immune cell within the lung (101), implying that conditions associated with loss of gut bacterial homeostasis (e.g., antibiotic use) could provide a window of opportunity for lung immunity to be impaired. In COPD, exacerbations can occur due to viral infections (60, 111), and resulting pathogenesis could be the result of dysbiosis leading to altered airway microbiota and disproportionate inflammation. Although loss of gut commensal signaling may impair lung innate immunity in this disorder, cigarette smoke directly and indirectly contributes to impaired innate immunity in the lung via alterations in ciliary function, mucus, innate immune cell phagocytosis, and via direct enhancement of bacterial virulence (e.g., enhanced biofilm formation) (reviewed in Ref. 112).These changes could impact the ability of respiratory pathogens to exacerbate COPD.

Given the propensity for viruses to precipitate lung disease exacerbations, it is interesting to note the potential impact of respiratory viral infection on the intestinal microbiota. Wang et al. (113) reported that influenza infection may lead to alterations in intestinal microbiota with a reduction in Lactobacillus and Lactococcus and an outgrowth of Enterobacteriaceae. As noted above, this may lead to a loss of beneficial microbiota for smoking-related disease. The authors demonstrated that these shifts in intestinal microbiota were not secondary to lytic influenza gut infection. This injury was mediated by Th17 cells, and IL-17 neutralization resulted in attenuated injury. Additionally, antibiotic-mediated depletion of intestinal microbiota led to attenuated intestinal injury. Interestingly, this study also highlighted the importance of an effector T cell that arose in the lung after infection and then migrated to the small intestine to provide IFN-γ to alter the gut microbiome. Ultimately the alterations in the gut microbiota stimulated epithelial-derived IL-15 to promote the Th17 response. It is possible that IL-17 responses arising in the gut may further impact lung disease (114). IL-17 is involved in the elimination of certain pathogens (115) and is implicated in the pathogenesis of several pulmonary pathologies including asthma, sarcoidosis, obliterative bronchiolitis, CF, and bone marrow transplant–related pneumonitis (116–120).

IL-17 may also play a central role in the dynamic change that occurs within the pulmonary microbiota of COPD. Yadava et al. (121) reported the impact of experimental change on the lung microbiota in an emphysema animal model. Specific pathogen-free and axenic mice were challenged with LPS/elastase for 4 wk. Microbiota diversity and abundance was decreased in the LPS/elastase model with an abundance of Pseudomonas, Lactobacillus, and a depletion of Prevotella. Loss of bacterial load was associated with attenuated IL-17 production. The intranasal application of microbiota-enriched fluid to axenic mice enhanced IL-17 production. The neutralization of IL-17 in mice harboring microbiota led to dampened inflammation and reduced disease burden. Several studies have implicated IL-17 in hepatic fibrosis, and certain experimental models of pulmonary fibrosis are IL-17A–dependent (120, 122, 123). Furthermore, studies examining the development of intestinal fibrosis have reported an association with alterations in the microbiota and Th17 responses (124). The intestine is a known source of Th17 cells through binding of segmented filamentous bacteria to intestinal epithelial cells (125). The case may be similar in the lung. Gauguet et al. (126) have demonstrated that intestinal segmented filamentous bacteria have the ability to promote pulmonary innate immunity through the induction of IL-17 to provide resistance to S. aureus pneumonia in animal models. This constitutes further evidence to support a gut/lung microbiome axis that may be pivotal in modulating the innate immune response of the lung. However, direct evidence of a gut/lung axis promoting exacerbation of chronic lung disease is limited to experimental data to date and requires further study.

Growing evidence suggests that alterations in the lung and/or gut microbiota characterize chronic lung diseases and may allow for exacerbations caused by endogenous microbiota alterations or susceptibility to new infection (Fig. 1). We speculate that impairment in lung innate immunity caused by microbial dysbiosis may promote susceptibility of the host to infections that can exacerbate chronic lung diseases. Furthermore, shifts in cytokine profiles mediated by changes in the microbiota may also promote epithelial injury and fibrotic outcomes. Overall, there appears to be a vital cross-talk between the gut and lung mucosa and the microbial communities within. The device through which this cross-talk may be achieved remains unknown. Possible instruments include translocation of gut microbiota (including potential pathogens) through blood or lymphatics, modulation of circulating or lung-resident effector immune and inflammatory cells, or alterations in systemic cytokine profiles. These results highlight the need for careful future human studies that will characterize not only the lung, but also the gut microbiota during periods of disease stability versus exacerbations. Additionally, murine models may allow us to interrogate the PRRs and cytokine signaling pathways that promote exacerbations.

The authors have no financial conflicts of interest.