Main

Chronic lung diseases, such as asthma and chronic obstructive pulmonary disease (COPD), are common and often occur together with chronic gastrointestinal tract (GIT) diseases, such as inflammatory bowel disease (IBD) or irritable bowel syndrome (IBS)1,2. Up to 50% of adults with IBD and 33% of patients with IBS have pulmonary involvement, such as inflammation or impaired lung function, although many patients have no history of acute or chronic respiratory disease3,4. Furthermore, patients with COPD are 2–3 times more likely to be diagnosed with IBD4. Individuals with asthma have functional and structural alterations in their intestinal mucosa, and patients with COPD typically have increased intestinal permeability2,5. Although the mature GIT and respiratory tract have different environments and functions, they have the same embryonic origin and, consequently, have structural similarities. Thus, it is not surprising that the two sites might interact in health and disease (Fig. 1); however, the underlying mechanisms are not well understood.

Figure 1: Principles of gut–lung crosstalk in health and disease.
figure 1

A healthy intestinal microbiota maintains homeostatic local immune responses through the exposure of structural ligands (for example, lipopolysaccharide (LPS) and/or peptidoglycan) and secreted metabolites (for example, short-chain fatty acids (SCFAs)). Invading microorganisms and absorbed metabolites influence circulating lymphocytes and contribute to the regulation of systemic immune responses. When the gut microbiota is disturbed, for example, during infection or antibiotic exposure, the normal microbiota-derived signals are altered, which leads to a changed immune response. In early life, when the immune system is still developing, this disturbance can substantially alter the way in which the immune system perceives its surroundings in later life, which can lead to the development of chronic inflammatory disorders in the gut and lung. In adulthood, dysbiosis of the gut microbiota, for example, through exposure to cigarette smoke, can cause systemic inflammation and an outgrowth of opportunistic pathogens, which can lead to chronic inflammation at distal sites. Although the specific taxa, ligands, metabolites and/or host responses may differ in specific disease situations, these broad principles outline the role of the microbiota in gut–lung crosstalk.

PowerPoint slide

An emerging area of intense interest is the influence of the host-associated microbiota on local and systemic immunity. This is exemplified in germ-free mice, which lack an appropriately developed immune system and show mucosal alterations, both of which can be restored through colonization with gut microbiota6,7. The microbiota changes over time from birth, to adulthood and into old age, and in response to environmental factors, such as diet, and drug and environmental exposures8.

In this ever-expanding field, researchers are now investigating how the local microbiota influences immunity at distal sites, in particular how the gut microbiota influences other organs, such as the brain, liver or lungs. This has led to the coining of terms such as the 'gut–brain axis' and 'gut–lung axis'. For example, antibiotic-induced alterations in the gut microbiota in early life increase the risk of developing allergic airway disease9,10,11,12; such findings add to our understanding of the links between exposure to microorganisms and allergy and autoimmunity (Box 1). The mechanisms by which the gut microbiota affects the immune responses in the lungs, and vice versa, are being uncovered, but many questions remain. In this Opinion article, we summarize the emerging role of the microbiota in the gut–lung axis, highlighting gaps in our knowledge and the potential for therapeutic intervention.

Microbiota of the healthy gut and lungs

The GIT remains, by far, the best-studied host-associated microbial ecosystem, partly owing to its abundance of microorganisms and partly because the microbiota can be profiled through faeces, which is easily obtainable. Both the abundance and diversity of the commensal microbiota generally increase along the GIT, and there are site-specific variations in the mucosa and the lumen13,14. These differences are governed by the prevailing environment, including pH, the concentration of bile acids, digesta retention time, mucin properties and host defence factors15. Despite these variations, the GIT is dominated by members of only four bacterial phyla, Firmicutes, Bacteroidetes, Proteobacteria and Actinobacteria; with lesser and sporadic representation of other phyla, including Fusobacteria, Verrucomicrobia and Spirochaetes. This 'core' gut microbial community comprises up to 14 bacterial genera and 150 bacterial 'species', many of which have not yet been cultured16,17,18.

We are beginning to understand the lung microbiota through programmes such as the Lung HIV Microbiome Project, which is a multi-centre study that examines both individuals who are infected with HIV and uninfected individuals who have varying histories of lung and/or respiratory disease19. The lungs have a large surface area with high environmental exposure and are equipped with effective antimicrobial defences. Healthy lungs were long considered to be sterile; however, the advent of culture-independent approaches for microbial community profiling has resulted in the detection of microbial DNA in the lungs of healthy individuals19,20. These microorganisms probably reached the lungs from the oral cavity through microaspiration, as the taxonomic profiles of the two sites were similar19,20. Compared with surrounding sites, the lungs had a decreased abundance of Prevotella-affiliated taxa and an enrichment of Proteobacteria, specifically Enterobacteriaceae, Ralstonia spp., and Haemophilus spp.19, which may have resulted from host immunity and environment, such as redox state and oxygen availability. The lung microbiota might not be resident in healthy individuals, but rather transiently recolonized through microaspiration and breathing. The lungs have a comparatively low microbial biomass and remarkably similar composition to adjacent sites, even though the lungs are continuously exposed to entering microorganisms and their environmental conditions differ vastly from other body sites. These observations support the hypothesis that entry and selective elimination of a transient microbiota is the major determinant of microbial composition in the lungs, rather than resident and viable microorganisms. This does not negate the importance of host–microorganism interactions in the lungs, as evidenced by correlations between the composition of microbial communities and pulmonary inflammation and disease21. Rather, it highlights the delicate balance between microbial exposure and elimination; the possibility of dysbiosis at oral sites preceding and/or causing dysbiosis in the lungs and contributing to disease pathogenesis19; and the importance in distinguishing whether microbial DNA that is detected by culture-independent techniques is truly representative of viable bacteria in the lungs22. Technical challenges, such as low microbial biomass and bronchoscope contamination, constant seeding from oral and GIT sites, and mucociliary and immune clearance, have hindered the identification of a viable and resident, or a transiently recolonizing, microbiota in the lungs, as well as further research into host–microorganism interactions. Novel methods of sampling tissue with minimal contamination23, longitudinal studies to identify temporal changes in the microbiota, and the increasing use of metagenomic analyses to facilitate the cultivation of fastidious bacterial species24, will provide a clearer picture of the role of the respiratory microbiota and enable the better design of interventional studies to develop a more complete understanding of host–microorganism interactions in the lungs.

Interactions between the gut and lungs

Interactions of microorganisms between the sites. The epithelial surfaces of the GIT and respiratory tract are exposed to a wide variety of microorganisms; ingested microorganisms can access both sites and the microbiota from the GIT can enter the lungs through aspiration. Both the gut and respiratory mucosa provide a physical barrier against microbial penetration, and colonization with the normal microbiota generates resistance to pathogens; for example, through the production of bacteriocins15. Furthermore, a rapidly expanding collection of commensal gut bacteria, including segmented filamentous bacteria (SFB), Bifidobacterium spp. and members of the colonic Bacteroides genus, induce the production of antimicrobial peptides, secretory immunoglobulin A (sIgA) and pro-inflammatory cytokines. Non-pathogenic Salmonella strains downregulate inflammatory responses in GIT epithelial cells by inhibiting the ubiquitylation of nuclear factor-κB (NF-κB) inhibitor-α (IκBα)25, whereas some Clostridium spp. promote anti-inflammatory regulatory T cell (Treg cell) responses26. In the respiratory tract, Streptococcus pneumoniae and Haemophilus influenzae synergistically activate host p38 mitogen-activated protein kinase (MAPK) in a Toll-like receptor (TLR)-independent manner to amplify pro-inflammatory responses27. Conversely, non-pathogenic S. pneumoniae and other bacteria and their components can suppress allergic airway disease by inducing Treg cells28,29,30,31. In recipients of lung transplants, the microbiota of the respiratory tract alters immunity in the lungs. Firmicutes-dominated and Proteobacteria-dominated dysbiosis were associated with the expression of inflammatory genes in pulmonary leukocytes, whereas Bacteroidetes-dominated dysbiosis was linked to a gene expression profile that is characteristic of tissue remodelling32. In both cell culture32 and animal models33, the inflammatory response that is induced by pathogenic species is larger than the response induced by commensal microorganisms, which indicates that a diverse lung microbiota protects against pathology by 'diluting' the more pro-inflammatory stimuli of pathogens. Although the transfer of microorganisms from faecal suspensions has been used to determine the role of the gut microbiota, such techniques have not yet been used to transfer the respiratory microbiota between animals, which limits our understanding of their roles.

Several studies have shown the effects of GIT colonization with orally administered bacteria on lung function. Oral gavage of faecal suspensions in mice that were first treated with antibiotics provided short-term improvements in some, but not all, measures in an S. pneumoniae infection model34.Although the nature of this 'gut–lung axis' has been challenged owing to the potential confounding effects of faecal administration by oral gavage and antibiotic use35, the concept warrants systematic and controlled evaluation. In infants, the composition of the gut microbiota and caesarean section have been linked to atopic manifestations, and colonization by Clostridium difficile at one month of age was associated with wheeze and eczema throughout early life, and with asthma after 6–7 years36. Positive associations between the presence of 'beneficial' bacteria, such as Bifidobacterium longum, in the gut and a lower incidence of asthma have also been identified37, although larger and longer studies are required to evaluate these associations.

Considerable evidence suggests that host epithelia and other structural and immune cells assimilate information directly from microorganisms and from the concomitant local cytokine response to adjust inflammatory responses, and that this shapes immune responses at distal sites, such as the lungs38,39 (Fig. 2). There is less evidence of the direct transfer of microorganisms between sites, although the translocation of bacteria from the GIT to the lungs has been observed in sepsis and acute respiratory distress syndrome, in which barrier integrity is compromised40. In addition, some environmental factors, such as dietary fibre, can produce similar changes in the microbiota of the GIT and the lungs39. Whether these result from diet-driven changes in microbial metabolites, changes in innate immune responses or a combination of both remains to be determined.

Figure 2: Structural and functional similarities and differences between the gut and lung.
figure 2

The gut and respiratory tract epithelia have substantial differences in functional purpose and exist in different environments; however, they retain some anatomical similarities. Both are derived from the endoderm and consist of columnar epithelial cells with projections of microvilli (gut) or cilia (respiratory tract) that function as a physical barrier and as sentinels for the immune system in conjunction with associated lymphoid tissue. Both secrete mucus through goblet cells as well as secretory immunoglobulin A (sIgA; although less in the lung). The alveoli that are found in terminal airways in the lungs differ substantially, consisting of squamous epithelial cells that either secrete surfactant (type 2 alveolar cells) or function in gas exchange (type 1 alveolar cells). However, the similarities end here: the intestinal lumen is an oxygen-poor environment and functions to digest food and absorb nutrients. The movement of matter is unidirectional (mouth to anus), with the exception of reflux or vomiting. Furthermore, the pH, enzyme presence and structure vary along the gastrointestinal tract (GIT). By contrast, the respiratory tract and alveoli are oxygen-rich and movement is bidirectional (inhalation and exhalation). The temperature in the gut is relatively uniform at 37 °C, whereas the temperature in the respiratory tract differs depending on the proximity to the pharynx. Thus, it is unsurprising that the microbial life in each environment is distinct. Changes in diet and exposure to therapeutics and environmental particulates can directly affect the composition of the microbiota. Both the gut and lungs are able to influence each other's immune responses. Dendritic cells in the intestines and respiratory tract, and macrophages in the lungs, sample antigens in the lumen. Lymphocytes in the associated lymphoid tissues circulate through the lymphatic system to affect systemic immunity. Bacteria from the gut can travel to the lungs through aspiration of vomit or oesophageal reflux. During times of dysbiosis, disturbed epithelial integrity may enable bacteria and their components and metabolites to enter the circulatory system, which can cause systemic inflammation.

PowerPoint slide

Microbial species-specific effects on host immunity. The crucial role of the microbiota in lung homeostasis and immunity is demonstrated by the poor outcomes of germ-free mice that were exposed to acute infections41 and their susceptibility to allergic airway disease42. Current research is assessing the effects of selected members of the commensal gut microbiota on systemic immunity, including in the lungs, as well as the use of probiotics and prebiotics to prevent and treat acute and chronic pulmonary disease (Fig. 3). For example, SFB in the gut, when present naturally or introduced by probiotic dosing or co-housing of mice, stimulated pulmonary T helper 17 (TH17) responses and protection from S. pneumoniae infection and mortality43. Intriguingly, a respiratory microbiota enriched with oral-related taxa, such as Prevotella spp., Rothia spp. and Veillonella spp., was associated with TH17-mediated immunity in the lungs of healthy human hosts21. Whether these links are correlative or causative remain unclear. Exposure of mice to dog-associated house dust altered the caecal microbiota, and, in particular, increased the abundance of Lactobacillus johnsonii and other Firmicutes-related lineages, such as species in the Peptococcaceae and Lachnospiraceae families44. Mice that were either exposed to dog dust or inoculated with L. johnsonii showed decreases in respiratory tract TH2 cytokine responses, and L. johnsonii treatment protected against exposure to respiratory syncytial virus and allergens such as ovalbumin. Other examples of microbial influences on host immunity include the ability of various Bacteroides spp. to expand Treg cell populations or bias the TH1/TH2 phenotype in either direction in a strain-specific manner, or the suppression of host inflammatory responses by the common bacterial metabolites short-chain fatty acids (SCFAs), which act through free fatty acid (FFA) receptors and/or epigenetic regulation of immune cells45.

Figure 3: Immune system programming by microbiota.
figure 3

Secreted and structural components of microorganisms can influence the host immune response both locally and at distal sites. Microbial metabolites, such as short chain fatty acids (SCFAs), bind to free fatty acid receptors or promote epigenetic changes in host leukocytes, which induce anti-inflammatory responses and decrease inflammation. Virulence factors from pathogenic bacteria, such as Helicobacter pylori or Bacteroides fragilis, can downregulate host immune responses, whereas structural components from commensal bacteria can influence inflammatory responses through the activation of pattern recognition receptors. GGT, γ-glutamyl transpeptidase; IL-10, interleukin-10; LPS, lipopolysaccharide; LTA, lipoteichoic acid; NAP, neutrophil-activating protein; PSA, polysaccharide A; Treg cell, regulatory T cell; UreB, urease subunit-β; VacA, vacuolating cytotoxin A.

PowerPoint slide

In related human studies, seropositivity to the gut-specific pathogen Helicobacter pylori, in particular, cytotoxin-associated gene A positive (cagA+) strains, has long been linked with decreased incidence of asthma and allergy46,47,48. Conversely, two recent meta-analyses suggest that infection with H. pylori is positively associated with increased incidence of COPD and other chronic bronchial diseases49. Although these differences might be partly attributable to genetic, environmental and lifestyle factors, these findings raise the possibility that systemic immune responses that are triggered by H. pylori might have different roles in the aetiology of different lung disorders. Strain variations, in addition to the expression of cagA, might also affect Treg cell responses50.

Clearly, the incredible diversity and abundance of species in the gut microbiota results in many immunomodulatory signals, which have considerable combined effects on host health. Although much has been uncovered about the activity of specific bacterial species, current research has only just begun to assess the structure–function relationships of the gut and lung microbiota with host immunity.

Components and metabolites of the gut microbiota that influence the lung. Early studies showed that germ-free mice have reduced responsiveness to lipopolysaccharide (LPS)-induced pathology and that this oral tolerance to microbial components was due to interleukin-10 (IL-10)-mediated hypo-responsiveness; however, subsequent exposure to LPS was no longer tolerated and the immune response became similar to that seen in conventional mice51,52. Furthermore, a robust response to LPS by colonic macrophages could be restored by commensal microorganisms53.

Bacterial components can also have anti-inflammatory effects, which attenuate GIT pathology. Polysaccharide A (PSA) from Bacteroides fragilis induces the production of IL-10 by T cells and protects against intestinal inflammatory disease caused either chemically or by infection with Helicobacter hepaticus54. Sphingolipids, which are naturally occurring cell membrane components of many anaerobic gut genera including Bacteroides, decrease the number of invariant natural killer T cells in the colon — cells that have been implicated in the development of colitis55. The best-studied metabolites, SCFAs, are by-products of the microbial fermentation of dietary fibre, have anti-inflammatory properties, are a source of energy for colonocytes, and regulate fatty acid and lipid synthesis in the host56.

Much less is known about the influence of microbial components and metabolites at other sites, including the lungs. Decreases in Faecalibacterium spp., Lachnospira spp., Veillonella spp. and Rothia spp. in the gut, and the urine levels of some microbial bile acid metabolites correlate with the development of atopic wheeze in children, although whether they are a cause or a consequence of wheeze is not known12. Oral administration of SCFAs has been shown experimentally to alleviate allergic airway disease39,57. Microbial components and metabolites have been implicated in other disorders, such as tryptophan in brain health, PSA in disorders of the central nervous system and trimethylamine N-oxide in atherosclerosis, which further highlights their importance in extra-intestinal environments55. In studies of other diseases, Bacteroides spp. were associated with early-onset autoimmune diseases, which may be a consequence of potent activation of immunity by LPS produced by these bacteria58.

Gut microbiota and lung diseases

Asthma. An increased risk of asthma has been connected to the disruption of the gut microbiota in early life (Box 1), and several studies have sought to characterize the precise microbial constituents that are associated with the development of the disease in infants.

The overall composition of the gut microbial community is not altered in infants at risk of the development of asthma, but subtle transient changes in select taxa can be detected in the first few months of life12,59. Increased risk of asthma has been associated with an increase in the abundance of B. fragilis and total anaerobes in early life60, as well as reduced microbial diversity59 and decreases in Escherichia coli61 and the relative abundances of Faecalibacterium spp., Lachnospira spp., Rothia spp. and Veillonella spp.12, although these findings were not consistent across all studies. In addition, although models of allergic airway disease support the existence of a critical developmental window early in life42,62, only one study has provided direct evidence that restoring the altered gut microbiota through probiotic treatment can decrease susceptibility to asthma12.

Similarly, in adults, the overall composition of the faecal microbiota in individuals with allergic asthma does not differ from healthy individuals63,64. There are taxa-specific differences, such as the enrichment of Bifidobacterium adolescentis, which negatively correlated with the time since asthma diagnosis63. Interestingly, heat-inactivated Bifidobacterium spp. that were isolated from infants with allergic asthma induced larger pro-inflammatory responses than those isolated from healthy individuals65.

There are several proposed mechanisms through which the gut microbiota can attenuate the risk of developing asthma. Infants who were at risk of developing asthma had decreased levels of LPS in their faeces12, whereas PSA from B. fragilis protected against the development of allergic airways disease in mice by inducing IL-10 responses in T cells66. H. pylori alleviated murine allergic airway disease in several ways, namely through the direct activation of Treg cells by neutrophil-activating protein67, or indirectly through urease subunit-β, which promotes tolerogenic reprogramming of dendritic cells68. In addition, γ-glutamyl transpeptidase and vacuolating cytotoxin from H. pylori altered dendritic cell function, but did not require Treg cells to alleviate symptoms69. Commensal bacteria can also influence the development of asthma through the production and secretion of metabolites, specifically SCFAs. The risk of asthma in infants was associated with a decrease in the concentration of acetate in faeces12 and inversely correlated with serum acetate concentrations in their mothers when they were pregnant57. A high-fibre diet, which increased levels of SCFAs in serum and faeces, protected mice against the development of asthma symptoms, a phenomenon that could be replicated through the direct administration of acetate or propionate before disease onset to promote tolerogenic immune responses in dendritic cells and Treg cells39,57. The benefits of a high-fibre diet were associated with a decreased ratio of Firmicutes/Bacteroidetes and an enrichment of Bacteroidaceae in both the faeces and lungs, which highlights the necessity of investigating microbial communities at several body sites for a complete understanding of the influence of microorganisms on host health. These studies did not directly explore the relationship between the composition of the microbial communities at the two sites, or the relative importance of the gut or lung microbiota in protection against disease. Such studies would be valuable in determining which body site to target with therapeutic interventions. An important but understudied area is the role that interactions between microorganisms have in the development of asthma. For example, the loss of intestinal bacteria and the subsequent outgrowth of commensal fungi triggered prostaglandin E2-induced changes in alveolar macrophages and increased allergic airway inflammation70. Furthermore, gut helminth infection protected mice against allergic airway disease, which was associated with an increase in the abundance of Lachnospiraceae and other Clostridiales members, the production of SCFAs and subsequent robust Treg cell responses in the lungs71. Although the Treg cell-promoting capability of Clostridium spp. has previously been demonstrated in the colon26,72, it is increasingly being explored for the treatment of diseases at other body sites, including asthma73,74.

COPD. Respiratory microbiota research in COPD has assessed changes in the disease state and with smoke exposure, which is a major risk factor for the development of this disease. Interestingly, although the lung microbiota is similar in healthy smokers and non-smokers, the oral microbiota differs substantially between the two groups19. As enrichment of the lung microbiota with taxa from the oral cavity is associated with increased inflammation in smokers75, it is plausible that changes in the oral microbiota and a failure to effectively clear aspirated microorganisms contribute to disease development, and may help explain why only a subset of smokers develop COPD. Moreover, there are substantial differences between the lung microbiota of patients with COPD compared with 'healthy' smokers76,77, which led to the proposal that the respiratory microbiota may be useful in the early diagnosis of COPD. By contrast, no study to date has investigated changes in the gut microbiota of patients with COPD. Nevertheless, in 'healthy' smokers, the faecal microbiota is characterized by an increase in the abundance of Bacteroides–Prevotella78, and a decreased Firmicutes/Bacteroidetes ratio79 compared with non-smokers. These changes in the composition of the faecal microbiota have been associated with intestinal inflammation and IBD80,81. Smokers also have a decreased abundance of Bifidobacterium spp.79,82, and hence may lose the anti-inflammatory effects that are often associated with this genus.

The causes of smoking-associated changes in the composition of the gut microbiota are probably a combination of environmental, host and microbial changes, such as intestinal and immune disruption, impaired clearance of pathogens83,84, acidification of gastric contents85 and ingestion of bacteria that are present in cigarettes86. Furthermore, cigarette smoke can directly affect the virulence of bacteria87 and fungi88, as well as alter the growth and exopolysaccharide structure of known gut bacteria, such as Bifidobacterium animalis89, which may contribute to dysbiosis. Even after the cessation of smoking, many of the changes that cause dysbiosis persist for prolonged periods of time, and thus any therapeutic intervention to restore the gut microbiota may potentially require repeated administration to prevent relapse.

In the absence of longitudinal or interventional studies, it is difficult to ascertain whether changes in the gut or respiratory microbiota are a cause or a consequence of COPD. Most likely, both are true and operate simultaneously or at different stages of disease. Exposure to environmental stimuli and the onset of disease cause dysbiosis, which, in turn, probably contributes to disease progression. Moreover, defined probiotic use may benefit patients with COPD, particularly if used as an early preventive intervention. Oral administration of Lactobacillus casei improved the previously defective function of peripheral natural killer cells in adult male smokers90, whereas Bifidobacterium breve and Lactobacillus rhamnosus reduced lung pathology in a mouse model of COPD91 and reduced inflammatory responses in macrophages that were exposed to cigarette smoke extract in vitro92. Similarly, a diet that increased the production of SCFAs protected against elastase-induced inflammation and emphysema93. Although a causal relationship between SCFAs and protection in this study was not confirmed, both cigarette smoke94 and environmental particulate matter95 decreased SCFA concentrations in rodents, and cigarette smoke condensate decreased the production of SCFAs in vitro89. Furthermore, increased intestinal translocation of bacteria and their products occurred after exposure to particulate matter or the development of COPD2,95,96. Bacterial toxins, such as enterotoxin97 or LPS98, can contribute to the pathogenesis of COPD and microbiota-associated intestinal inflammation may become systemic and also contribute. The potential of SCFAs to improve intestinal barrier function may account for their benefits in animal models of COPD99,100, although this is yet to be explored in clinical studies.

Respiratory infections. The gut microbiota is broadly protective against respiratory infection, as its depletion or absence in mice leads to impaired immune responses and worsens outcomes following bacterial or viral respiratory infection34,41,101,102,103. Administration of SFB improved resistance to Staphylococcus aureus pneumonia43 and Bifidobacterium spp. protected against both bacterial104 and viral pulmonary infection in mice103,105. Lactobacillus spp. and Bifidobacterium spp.-based probiotics also improved the incidence and outcomes of respiratory infections in humans106,107,108,109.

Several aspects of experimental design influence the results of infection studies, including the route of administration of bacterial ligands102,110, the facility from which research animals are sourced43, the type of antibiotic used for the depletion of the microbiota62,102 and the infecting pathogen. For example, herpes simplex virus type 2 or Legionella pneumophila do not seem to be influenced by antibiotic-mediated depletion of the microbiota102.

Nevertheless, several important mechanisms by which the gut microbiota promotes the clearance of pathogens have been identified. Innate immune responses to bacteria in the lungs are greatly enhanced by exposure to nucleotide-binding oligomerization domain (NOD)-like receptor and TLR agonists in the GIT, which include peptidoglycan, LPS, lipoteichoic acid and CpG DNA41,101,110. Similarly, stimulation of TLRs by cell wall components and flagellin of gut bacteria is necessary for effective adaptive immune responses to influenza102,111, whereas the anti-inflammatory effects of oral SCFA administration are linked to reduced pulmonary pathology following both bacterial104,112 and viral113 infection in mice. However, the microbiota can also drive gut pathology in pulmonary infection. Influenza virus infection in mice increased the number of lung-derived CC-chemokine receptor 9 positive (CCR9+)CD4+ T cells, which preferentially migrate to the GIT under the guidance of C-C motif ligand 25 (CCL25) expressed on intestinal epithelial cells114. This resulted in the outgrowth of E. coli and the induction of aberrant TH17 responses and intestinal damage.

Conclusions and perspectives

Many studies have identified the presence of a lung microbiota in health and disease. However, we believe that the healthy lung microbiota may be transient and best described as a progression of taxa that are influenced by adjacent body sites and the external environment, rather than an actively reproducing core resident community. This is not down-playing the importance of a transient microbiota in healthy lungs, which could still have important roles in inflammatory responses whether viable or not. By contrast, the microbiota is more likely to be persistent and resident in the respiratory tract and lungs of individuals with respiratory disease, although whether it is a cause or a consequence of disease remains to be elucidated. Furthermore, the lung microbiota could affect, or be affected by, microorganisms or immune responses at distal sites.

The crosstalk between microorganisms and the host is complex and our current understanding of these interactions is in its infancy. It is unlikely that any one of these interactions is solely responsible for the functions of the microbiota, and alterations in any part of these relationships may be enough to affect health and disease. It is unclear whether changes in the microbiota at one site affect many distal sites equally, or whether these systemic effects might be specific to certain tissues. To date, no such broad study investigating these systemic widespread effects has been carried out.

Thus far, gut–lung microbiota studies have had two major limitations: the first is discerning causative over correlative effects and the second is timing. Most studies have been associative. Furthermore, culture-independent identification of microorganisms has not yet replaced the need to isolate and culture suspected opportunistic pathogens or probiotics to study their effects, and many members of the microbiota cannot be easily cultured. Thus, it is typically unclear whether changes that are observed in the microbiota are the cause or effect of disease. In the case of timing, most experimental data have described the effect of the gut microbiota on the development of lung disease and not on established lung disease. Longitudinal studies in humans and animals that associate changes in the microbiota with the severity of established chronic lung disease are required. Research into manipulations of the microbiota during lung disease is necessary to improve our understanding and inform the development of novel therapies (Box 2).

Increasingly, microbiota research is moving towards defining functional guilds. As taxonomic variation between sites and individuals is so large, and the microbiota consists of thousands of species, it is highly likely that there is redundancy between species in terms of their interactions with other microorganisms and in the metabolites that they produce. Thus, next-generation 'omics' approaches are required to identify functional guilds to aid in defining how the microbiota of the gut and the microbiota of the lungs interact with each other and influence health and disease.

In summary, the lung microbiota in a healthy individual may be transient and constantly re-seeded from the environment and cleared by the immune system, but may still influence health and disease. In respiratory diseases the lung microbiota probably becomes persistent and may be both a cause and a consequence of the disease, forming a pathogenic feedback loop. It is clear that bacterial components and metabolites in the gut and the lungs have the capacity to modulate systemic and local immunity, with specific taxa able to influence the pathogenesis of respiratory diseases, such as asthma, COPD and respiratory infections. Such relationships have been identified in other respiratory diseases, such as cystic fibrosis115, which, as a genetic disease, is a special case. Respiratory challenges with environmental factors such as pollution, cigarette smoke, antibiotics and diet, influence disease risk and probably drive pathogenesis through their ability to modulate the composition of the microbiota, although the mechanisms of these effects remain unknown. Further longitudinal studies and improved interventional experiments will help to elucidate the role of the microbiota and gut–lung crosstalk in respiratory disease, and will potentially lead to the identification of new and effective avenues for treatment.