Commensal bacteria in the upper respiratory tract regulate susceptibility to infection

Abstract

The human body is host to several distinct microbial communities. Disruption of these communities increases susceptibility to a wide range of diseases, including respiratory tract infections. While commensal bacteria in the gut contribute to this effect, recent studies point to a role for commensals occupying the upper respiratory tract through direct pathogen killing and by modifying nasal and lung immune homeostasis. Clinical trials exploring ‘probiotic’ respiratory tract commensals are an exciting development in this area. Upper respiratory tract microbiome sequencing has revealed that destabilization of this community precedes infection, indicating that microbiome profiling of individuals has predictive value. Further investigation of respiratory tract commensal–host interactions will be critical to translate bacterial-mediated protection toward new therapeutic approaches for respiratory tract disease.

Introduction

The upper respiratory tract (URT) contains a well-documented bacterial community, or microbiome, residing in the nasal cavity and nasopharynx. Opportunistic bacterial pathogens, referred to here as opportunistic pathogens, are transient members of this community that cause illness upon invasion of other host tissues. Respiratory tract infections caused by opportunistic pathogens constitute a major burden of disease. Among these, pneumonia is the number one cause of death worldwide in children under five years old and the leading infectious cause of death in the elderly [1–3]. Otitis media, or ear infection, is the most frequent diagnosis for antibiotic prescription in young children [4]. The four predominant opportunistic pathogens of the URT are Streptococcus pneumoniae, Haemophilus influenzae, Moraxella catarrhalis and Staphylococcus aureus. While these bacteria often occupy the URT asymptomatically, colonization is a prerequisite for invasive disease [5–7]. Cooperative and competitive interactions between these bacterial pathogens influence susceptibility to infection [8–11]. In addition, co-infection with viral pathogens including influenza A, influenza B, or respiratory syncytial virus (RSV) pre-dispose bacterial invasion and are associated with more severe disease [12–14]. However, it is less clear how the remaining members of the URT microbial community modify pathogen acquisition.

Many bacteria inhabiting the gut and URT are never or only rarely associated with disease, referred to as commensals. URT commensals predominantly reside in the nasal cavity and nasopharynx, but also reach the lung through aspiration [15–17]. The URT microbiome is composed of several ‘core’ genera present in the majority of healthy individuals, the most abundant of which include Staphylococcus, Streptococcus, Corynebacterium, Prevotella, Veillonella, Propionibacterium, and Fusobacterium in adults with Moraxella also prominent in young children [17,18,19••,20]. It is clear that the microbiome as a whole contributes to resistance against diverse (fungal, viral and bacterial) lung infections, which are frequently more severe in antibiotic treated and germ-free animals [21–26]. However, the depletion (or absence) of bacteria from both the gut and URT obscures the contributions of microbiota from each site to immune homeostasis. Antibiotic treatment also differentially enriches genera in the URT as well as the gut [23], which may influence susceptibility to infection depending on which bacteria remain. In people, the increased abundance of Haemophilus, Streptococcus, and Staphylococcus following antibiotic therapy is likely in part due the rise in antibiotic resistance among opportunistic pathogens including H. influenzae, S. pneumoniae and S. aureus [20,27,28]. This review focuses on new developments in our understanding of how commensal bacteria regulate susceptibility to respiratory tract infection, with an emphasis on the role of the URT microbiota.

Shared mechanisms for the regulation of lung immunity by commensal bacteria

URT commensal bacteria occupy the same niche as opportunistic pathogens, making it difficult to distinguish between direct competition and indirect modulation of the immune response. In contrast, bacteria colonizing the gut must engage circulating immune factors to influence lung immunity. Some of the signaling pathways activated by gut commensals are also induced by bacteria in the URT. This is particularly evident for innate immune receptors, which recognize conserved bacterial ligands. For example, Toll-like receptor (TLR)4 recognition of bacterial lipopolysaccharide (LPS) improves protection against influenza A virus following either intranasal or rectal exposure in mice [23]. More recently, this was also shown for the innate immune Nod-like receptor (Nod)2[29••].Reconstitution of antibiotic treated mice with a compilation of potent Nod2-stimulators from either the URT (intranasal reconstitution) or gut (oral reconstitution) rescues protection against S. pneumoniae and Klebsiella pneumoniae infections [29••]. Nod2 activation primes alveolar macrophages in the lung to produce reactive oxygen species (ROS), contributing to pathogen clearance [29••]. Similarly, the respiratory tract commensal Staphylococcus sciuri promotes the adjuvanticity of cholera toxin in a Nod2-dependent manner [30]. In this case, stimulation of lung CD11c+ cells, the majority of which are alveolar macrophages, by intranasal cholera toxin requires activation of Nod2 by commensal bacteria including S. sciuri [30]. Alveolar macrophage ROS is also induced by oral treatment with the gut commensal Bifidobacterium longum, boosting protection against lethal K. pneumoniae lung infection in mice [31]. These studies demonstrate that recognition of either gut or URT commensals by innate immune receptors activates alveolar macrophages in the lung, improving resistance to lung infection.

Commensal bacteria also influence lung immune homeostasis by regulating production of the cytokines IL-17A and IL-22 from several immune cell types. This has been shown for gut segmented filamentous bacteria (SFB), which improve resistance against both S. aureus and the fungus Aspergillus fumigatus in mice by activating lung IL-17A⁺ T helper (Th)17 cells and IL-22 production [25,32]. SFB are also protective against lethal S. pneumoniae infection in Rag1^−/− mice, which lack mature T and B cells, indicating a role for other cell types [33]. Aside from Th17 cells, gut commensal bacteria regulate lung mucosal-associated invariant T (MAIT) cell production of IL-17A and innate lymphoid cell type 3 (ILC3) production of IL-22, increasing resistance in murine models of Mycobacterium tuberculosis and S. pneumoniae infections respectively [24,34]. The bacterial ligands from SFB and other gut commensals that are critical for regulating these responses, and whether these ligands are shared by URT commensals, have not been established. However, the observation that Nod2 signaling induced by URT bacteria correlates with IL-17A-dependent GM-CSF production in the lung supports a role for bacteria from both the lung and gut in modulating lung IL-17A [29••]. Together, these studies indicate that commensal URT and gut bacteria activate shared innate immune pathways that regulate resistance to lung infection.

It less clear whether URT commensal bacteria modulate Th1/Th2 immunity in the lung similar to bacteria from the gut. For example, short-chain fatty acids (SCFAs) produced by gut commensals are protective against Th2 associated allergic lung inflammation, without influencing accumulation of regulatory T cells (Tregs), in mice [35,36]. Gut bacteria including Lactobacillus rhamnosus also improve Th1 responses in the lungs of mice infected with influenza A virus, S. pneumoniae, and M. tuberculosis [37–39]. While the URT harbors anaerobic SCFA-producers including Prevotella, the concentrations of SCFAs found in the URT are extremely low [40], suggesting a limited role. Commensal bacteria found in the skin and nose can differentially induce Th1 cytokines in the lung [41], but the impact of these responses on pathogen challenge remains largely unexplored. Instead, URT commensal bacteria have been shown to improve resistance to infection by mechanisms unique to the environment of the respiratory tract, as discussed below.

Distinct mechanisms for the regulation of pathogen acquisition and invasion by URT commensal bacteria

URT commensal bacteria regulate the mucosal barrier, the initial site of pathogen exposure. For example, the URT commensal Staphylococcus epidermidis induces nasal epithelial production of interferon (IFN)-λ, which increases resistance to influenza A virus infection in mice [42••]. IFN-λ is observed in the nasal secretions of people colonized with S. epidermidis [42••], though it is unclear whether these levels are sufficient for protection in humans. Resistance to influenza A virus infection is also associated with modulation of type I IFN signaling in the lung epithelium by gut bacteria [43]. However, there is no evidence to date that gut commensals regulate the nasal mucosa similar to URT bacteria. Another way that S. epidermidis regulates the nasal epithelium is by the stimulation of antimicrobial peptide (AMP) production [44••]. This has direct consequences for pathogen resistance, as S. epidermidis blocks S. aureus and M. catarrhalis acquisition in an AMP-dependent manner in mice [44••]. In addition to stimulating mucosal immune factors, S. epidermidis produces bacteriocins that restrict the growth of S. aureus and M. catarrhalis [45,46] and protect against S. aureus colonization in mice [47]. Similarly, commensals closely related to H. influenzae produce bacteriocins and modify nasal pro-inflammatory cytokine production [48,49•], indicating that other URT commensals contribute to pathogen resistance through similar mechanisms. Collectively, these findings illustrate that URT commensal bacteria reduce susceptibility to pathogen acquisition by regulating the nasal mucosa.

URT commensals also influence the adaptive immune response, which has the potential for long-term consequences through the generation of immune memory. This has been shown most clearly for commensal Streptococcus species. Streptococcus mitis induces cross-reactive protection against S. pneumoniae in mice characterized by systemic antibody production and IL-17⁺ Th17 cells in the lung [50••]. Memory T helper cells, including Th17 cells, with cross-reactivity against S. mitis and S. pneumoniae have also been identified in humans [51]. Several commensal Streptococcus species express capsule, the predominant Streptococcus antigen, with genetic and antigenic similarities to that found in S. pneumoniae [52•,53,54•]. However, capsule-specific memory was not prevalent in a pool of cross-reactive memory T cells identified in humans [51], indicating importance for other antigens. Commensal Streptococcus species may thus contribute to baseline resistance against S. pneumoniae by promoting cross-reactive immunity. Taken together, URT commensal bacteria improve protection against respiratory tract pathogen colonization and infection through both direct competition and indirect immune modulation (Figure 1).

Figure 1.

URT commensal bacteria protect against respiratory tract infection. Colonization of the nasal epithelium with opportunistic pathogens is restricted by URT commensal bacteria through direct competition, for example by production of bacteriocins, lugdunin, and free fatty acids that restrict pathogen growth, and indirectly through modulation of the mucosal barrier, resulting in secretion of factors including antimicrobial peptides (AMPs) and IFN-l, which contribute to pathogen clearance (1). URT commensal bacteria also reduce pathogen invasion by promoting innate and adaptive lung immune responses including alveolar macrophage production of reactive oxygen species (ROS), IL-22, Th17 cells, and cross-reactive antibodies (2).

Treatment with URT commensals protects against infection in humans

The development of probiotics is centered on the concept that commensal bacteria with beneficial properties can be used to improve resistance to disease. While originally developed for intestinal diseases, probiotic gut bacteria also increase resistance to influenza A virus infection in mice [55–57], indicating the potential utility of probiotics for respiratory tract infections. More recently, clinical trials have explored the probiotic potential of URT commensal bacteria. In one study, serial treatment with nasal sprays containing commensal Streptococcus salivarius and Streptococcus oralis bacteria reduced acute otitis media recurrence and severity in children [58•,59]. Similarly, an oral spray containing S. salivarius and S. oralis reduced episodes of pharyngotonsillitis infection and need for antibiotic treatment in children infected with group A beta-hemolytic Streptococcus (GABHS) [60•]. Successful therapy with commensal Streptococcus bacteria relies on repeated inoculation over a period of several weeks to months, and lower doses are not effective [61], indicating the need for further development of this approach. The protective effect of these commensals may depend on direct pathogen competition in addition to the generation of cross-reactive immunity discussed above. Streptococcus commensals can disrupt pre-formed biofilms from several pathogens [62] and produce bacteriocins that kill S. pneumoniae [63], though these observations have not been confirmed in vivo. S. salivarius also restricts S. pneumoniae binding to a human epithelial cell line independent of bacteriocin production [64], further demonstrating that multiple mechanisms contribute to the inhibitory effect of these commensals on pathogen infection.

Other genera in the respiratory tract have similarly antagonistic commensal-pathogen interactions in humans. The nasal commensal Staphylococcus lugdunensis produces lugdunin, an antibiotic that is bactericidal for S. aureus and S. pneumoniae [65•]. While the therapeutic potential of lugdunin has not been evaluated in clinical trials, carriage of S. lugdunensis correlates with reduced S. aureus in people [65•]. In another example of competition between closely related bacteria, nasal inoculation of the commensal Neisseria lactamica reduces carriage of Neisseria meningitidis in young adults [66]. Similar to evidence for cross-reactive Streptococcus immunity, cross-reactive opsonophagocytic antibodies generated in people colonized with N. lactamica may contribute to this effect [67]. These studies identify URT commensal bacteria with probiotic potential for protection against infection with closely related opportunistic pathogens in humans.

URT microbiome composition predicts disease risk

URT microbiome sequencing has revealed that this community undergoes substantial changes during early development and in the context of disease. The identification of respiratory tract commensals that are predominant in healthy, but not infected, people has emerged as a common theme (Figure 2). For example, Corynebacterium propinquum, Dolosigranulum pigrum and Moraxella bacteria in the nasopharynx negatively correlate with lower respiratory tract infection (lower RTI) in young children [68•].

Figure 2.

The composition of the URT microbiome of the nasal cavity and nasopharynx influences susceptibility to respiratory tract infection. Scales represent how the balance of colonization with different types of bacteria reflects an individual’s risk of respiratory tract infection. In this example, the scale on the left depicts a higher burden of potentially beneficial commensal bacteria, which is associated with health, while scale on the right depicts a higher burden of opportunistic pathogens, associated with increased risk of respiratory tract infection. URT commensal bacteria that negatively correlate with disease include Corynebacterium, Dolosigranulum, S. epidermidis and S. lugdunensis. URT opportunistic pathogens that increase infection risk include M. catarrhalis, S. aureus, H. influenzae and S. pneumoniae.

In this analysis the species identity for Moraxella was not defined, making the role of the opportunistic pathogen M. catarrhalis unclear. Corynebacterium and Dolosigranulum are also largely absent in the nasopharynx of children with otitis media and other upper RTIs compared with healthy individuals [69•,70–72]. While it is tempting to speculate that these respiratory tract commensals are protective against pathogen acquisition and/or invasion, causation has not been established for most of these relationships. However, accumulating evidence suggests that Corynebacterium species directly compete with URT pathogens. Corynebacterium accolens produces free fatty acids that inhibit S. pneumoniae growth [73•], and cell-free medium from Corynebacterium striatum reduces S. aureus adhesion to epithelial cells [74•]. These findings are supported by the observation that live, but not inactivated, Corynebacterium pseudodiptheriticum reduces RSV and S. pneumoniae lung burdens in an infant rat co-infection model [75••]. In summary, microbiome sequencing has identified URT commensal bacteria with the potential to contribute to pathogen resistance.

In the absence of mechanistic studies in people, longitudinal analysis of URT microbiome profiles in the same individuals over time supports the concept that the composition of this community influences susceptibility to infection. One such study paired microbiome analysis and RTI incidence in infants throughout their first year of life [19••]. Infants with increased RTIs had shorter periods of colonization with Corynebacterium and Dolosigranulum species and more rapid dominance of Moraxella [19••]. This is consistent with Corynebacterium and Dolosigranulum as predominantly associated with the absence of disease, while the impact of Moraxella may be species or time dependent. Moraxella nonliquefaciens is also increased in children with acute sinusitis [76], indicating that species beyond M. catarrhalis are associated with disease. Importantly, changes in the URT microbiome precede the first RTIs in children [19••,20,77]. The predictive power of URT microbiome profiling is consistent with a direct relationship between the composition of this community and susceptibility to pathogen infection. In contrast to the consensus for RTIs considered as a whole, a separate cohort of respiratory tract commensal bacteria negatively correlate with tuberculosis (TB) [78•], indicating that the relationships between the URT microbiome and disease risk are pathogen specific.

Opportunistic pathogens destabilize the URT microbiome

While commensal URT bacteria contribute protective benefits, opportunistic pathogens have the opposite effect. For example, asymptomatic colonization with Streptococcus earlier in life correlates with a younger age of first lower RTI [20]. Pathogen colonization disturbs the microbial community of the URT even in the absence of disease presentation. This has been shown in children, as colonization with Haemophilus and Streptococcus is associated with reduced microbiome stability compared with Corynebacterium, Dolosigranulum and Moraxella [77]. The concept of microbiome disruption by opportunistic pathogens has been directly tested in a human challenge study, where establishment of S. pneumoniae carriage in healthy adults increased URT microbiome diversity [79]. These findings suggest a nuanced relationship between opportunistic pathogens and disease, where asymptomatic changes in the URT microbiome influence subsequent pathogen invasion. Whether these changes modulate invasion of colonizing pathogens or newly acquired opportunistic bacteria remains an important area for further study.

Conclusions and future directions

Despite substantial interest in the influence of the microbiome on human disease, the importance of commensal bacteria from sites beyond the gut remains poorly characterized. Given the dynamic relationships in the URT between commensal and pathogenic species within the same genera, more in-depth microbiome analysis may reveal species-specific relationships that have been previously overlooked. Models of respiratory tract infection that incorporate the manipulation of URT commensal species and pathogen co-infection will improve our ability to predict and manipulate these relationships. Collectively, the work highlighted here demonstrates that URT commensals influence susceptibility to infection through direct and indirect mechanisms, some of which are unique to the environment of the nasal mucosa. Further investigation of how URT commensal bacteria alter pathogen acquisition and immune homeostasis is critical for the development of new therapeutic approaches, the most promising of which include probiotics from and for the respiratory tract.