Interactions of the bacteriome, virome, and immune system in the nose

Abstract

Emerging evidence suggests that the nasal microbiome may influence host susceptibility to initial development and severity of respiratory viral infections. While not as extensively studied as the microbiota of the alimentary tract, it is now clearly established that the microbial composition of this niche is influenced by medical, social and pharmacological influences, predisposing some sub-populations to respiratory infections. The resulting specific microbial profiles may explain variance in susceptibility to viral infection. This review summaries the evolution and constituents of the commensal nasal microbiome; the bacterial-virus, bacterial-host and interbacterial interactions which potentiate disease; and considers the effects of interventions such as vaccination and probiotics.

Introduction

In 2015, there were an estimated 17.2 billion cases of upper respiratory infections alone (GBD 2015 Disease and Injury Incidence and Prevalence Collaborators 2016). Despite their high incidence, the pathophysiology of respiratory viral infections is yet to be fully understood. It includes complex interactions between viruses, bacteria, and the immune system.

Knowing whether and how the nasal microbiome (NM) influences the severity and progression of viral respiratory diseases is essential in developing new therapies. This knowledge may in turn forecast virus-infected patients’ susceptibility to secondary bacterial infections which are associated with high mortality. This narrative literature review aims to summarize the evolution of the nasal microbiome, as well as characterize bacterial-bacterial, viral-bacterial, and bacterial-immune interfaces.

Much of the literature focusses exclusively on the nasopharynx, the anatomical area comprising the back of the nose and articulating with the oropharynx. Where not otherwise stated, the ‘nasal microbiome’ here refers to the microbiota of the mucosa-lined nasal cavity and nasopharynx (and exclude the keratinized anterior nares and/or oropharynx), as the most accurate sampling techniques, namely the nasopharyngeal swab and nasal aspirate, by default sample both these sites (Flynn et al. 2021).

Nasal microbiota development

The microbiome of the nasal cavity begins to differentiate from other neonatal niches as early as the first week of life. Colonization by Staphylococcus aureus, Corynebacterium, and Dolosigranumlum peak shortly at birth, and these genera are accompanied by Moraxella catarrhalis/nonliquefaciens,  Streptococcus pneumoniae, and Haemophilus influenzae (Bosch et al. 2016, Salter et al. 2017). High abundance of Corynebacterium, Dolosigranulum, and Moraxella species are associated with increased microbiome stability (Biesbroek et al. 2014). Profiles dominated by these taxa were associated with lower patient-reported frequency of respiratory tract infections (RTIs) within the first two years of life. However, increased prevalence of both Haemophilus, Streptococcus, and Moraxella are associated with decreased microbiome stability and higher risk of RTIs even when adjusting for potential demographic environmental and therapeutic factors, thus suggesting a pathobiont role of Moraxella (Biesbroek et al. 2014, Teo et al. 2015). In adolescence, an unsurprising shift to lipophilic genera Propionibacterium, Corynebacterium, and Turicella are noted in the skin-like anterior nares, but trends further back in the mucosa-lined nasal cavity are yet to undergo longitudinal study in this age group (Oh et al. 2012).

The majority of studies of the nasal microbiome follow pediatric populations. A study of 100 adult nasopharynges found differentiation into four distinct seemingly stable profiles, three dominated by Moraxella, Fusobacterium, or Streptococcus, respectively and one more heterogeneous still (De Boeck et al. 2017). A comparable analysis of the oropharyngeal microbiome in elderly pneumonia patients and healthy controls grouped 8 similar ‘clusters’ of similar microbiomes along the spectrum from health to pneumonia, and it is likely the nasal microbiome follows similar trends (de Steenhuijsen Piters et al. 2015).

External factors affecting nasal microbiome development

A variety of other factors influence the NM across childhood into adulthood (Fig. 1) (Flynn and Dooley 2021). Infants originating from vaginal delivery (as opposed to caesarian section) and being breastfed were associated with favorable Corynebacterium/Dolosigranulum dominant profiles, and are considered beneficial in protecting from development of respiratory disease (Dominguez-Bello et al. 2010, Li et al. 2014, Bosch et al. 2016, Shilts et al. 2016). In the gut microbiome, caesarian-associated microbial changes were associated with significantly lower circulating levels of chemokines CXCL10 and CXCL11 to 24 months of age; a degree of immune modulation may be at play in the respiratory system also (Jakobsson et al. 2013). Indeed the risk of asthma seems to be related inversely to initial age of streptococcal colonization, and greater abundance of Moraxella catarrhalis and Streptococcus pneumoniae exacerbate the severity of asthma attacks (Kloepfer et al. 2014, Teo et al. 2018).

Figure 1.

Factors affecting composition of the Nasal Microbiome during childhood.

Children with siblings are significantly more likely to have a Moraxella-dominant than seemingly protective Corynebacterium/Dolosigranulum profiles, suggesting that these overall trends in childhood occur as part of groups rather than merely within individuals (Hasegawa et al. 2016). A positive association has been seen with smoking and relative abundance of Corynebacterium and Staphylococcus (De Boeck et al. 2017).

Two widely used vaccines, the bacterial pneumococcal conjugate vaccine (PCV) and the viral live attenuated influenza vaccine (LAIV), have been shown to exert pressures on specific genera, such as increased prevalence of Staphylococcus aureus and Haemophilus influenzae (Spijkerman et al. 2012, Tarabichi et al. 2015). PCVs (7-, 10-, and 13-valent) eliminate the most prevalent pneumococcal serotypes. However, children vaccinated with the PCV vaccine demonstrate replacement with a greater abundance of non-vaccinated pneumococcal serotypes (Gladstone et al. 2017, Gursoy et al. 2019, Sime et al. 2019). Notably, within the presence of live influenza vaccine, the optimal levels of low mucosal inflammation and minimal levels of dysbiosis have been seen with ‘low-dense’ pneumococcal carriers as opposed to non-carriers, implying some degree of upregulated immune homeostasis provided by this pathobiont (de Steenhuijsen Piters et al. 2019).

The microbiome of infants in the Netherlands appear to fluctuate over the seasons independently of antibiotic use and viral coinfection. Absolute abundance of Fusobacteria and Fusobacteria increase notably in autumn and winter compared to relative increase in Bacteroidetes and Firmicutes in the spring (Bogaert et al. 2011). In a similar cohort in Switzerland showed an annual variation in Shannon diversity, troughing and peaking in June and September, respectively (Mika et al. 2015).

Protective Effects of Nasopharyngeal Microbiota against Viral Respiratory Infections

Several hypothesized mechanisms exist by which bacteria individually or corporately convey protection against respiratory viral infections:

Immune Modulation

Immune modulation in infection with influenzavirus is well studied. Less severe responses inadequately inhibit viral proliferation, but stronger responses may cause immune-driven macroscopic lung injury. These responses are elicited by toll-like receptors (TLR) particularly TLR4, 7, 8, and 9, Similar nod-like receptors trigger a platform known as the ‘inflammasome,’ a cascade of pro inflammatory mediators, which are complemented by retinoic acid inducible gene-I receptors to alert the immune system of Influenza A infection. It is suspected that influenza A alters the antibacterial activity of unconventional T-cells such as natural killer cells paving the way for opportunistic pathogens Staphylococcus aureus and streptococcus pneumoniae (Paget and Trottein 2019). The abundance of Staphylococcus aureus seem to be inversely related to other key genera such as Dolosigranulum and Simonsiella within Influenzavirus patients, suggesting that they are causal or at least diagnostically indicative of health (Manohar et al. 2020). Proteolytic activation of hemagglutinin via proteases may enhance the pathogenicity of Influenzavirus. This activation is achieved by trypsin-like activity in bronchoalveolar lavage fluids and may be direct (e.g. Staphylococcus aureus and Aerococcus viridans) or indirect via the combined application of protease and virus (e.g. Pseudomonas aeruginosa) (Scheiblauer et al. 1992). The presence of some bacteria and absence of others may affect the ability of certain viruses to evade host immune responses. In vitro, Haemophilus influenzae stimulates increased ICAM-1 and TLR3 expression on airway epithelial cells, facilitating Rhinovirus entry (Sajjan et al. 2006). The presence of Moraxella catarrhalis was also found to downregulate TLR3 expression in vitro, possibly compromising hosts’ antiviral protection (Heinrich et al. 2016).

Bacterial–bacterial and Bacterial–virus synergism and competition

There are various protective mechanisms by which specific bacteria can suppress potentially pathogenic co-inhabitants during respiratory viral infections. One method is direct competition—a significant inverse relationship has been observed between Dolosigranulum and streptococci in children with invasive pneumococcal disease (Camelo-Castillo et al. 2019). Indeed, Dolosigranulum has demonstrated remarkable genomic stability and lack of virulence over the last 20 years, offering it as a potential future probiotic (Flores Ramos et al. 2021). Moraxella appears to work synergistically with Rhinovirus to cause milder infections (Tozzi et al. 2021). Whilst increased Moraxella diversity at 6 months is associated with decreased alpha-diversity and thus decreased stability, it is unclear whether such in stability is caused by moraxella itself or whether it merely overgrows in the favorable ecology of such dysbiosis, especially given its previously noted role as a key constituent of the balanced microbiome (Chapman et al. 2020). Similarly, children prone to frequent admissions with viral upper respiratory tract and anatomically related otitis media infections demonstrated overgrowth of a combination of Moraxella catarhallis, Streptococcus pneumoniae, and non-typable Haemophilus influenzae as much as any one species, further supporting a more generalized dysbiotic model than a single culprit pathogen (Xu et al. 2017). Furthermore, the absence of Haemophilus influenzae, Moraxella catarrhalis and Staphylococcus aureus and presence of pneumococci in bronchial epithelial cells may increase susceptibility to Human metapneumovirus (Verkaik et al. 2011). Respiratory syncitial virus has been outlined as frequently associated with nasal microbiome dysbiosis. The nasal microbiome during RSV infection in children has been compared to the Moraxella/Haemophilus/Streptococcus-dominant profile noted in childhood asthma, however healthy adults inoculated with the virus did not demonstrate disruption of the microbial community nor inflammatory cytokines (Morens et al. 2008, Mika et al. 2015). Findings like these undermine classical models of pathogen infection where inoculation with a pathogen causes disease. On the other hand, influenza A positive patients have demonstrated increased abundance of Staphylococcus aureus, Phyllobacterium spp., Moraxella spp., Corynebacterium spp and a decrease in abundance of Pseudomonas spp compared with healthy controls (Wen et al. 2018). This, however, was within a severe ICU treated cohort, so such dysbiosis may be more closely a reflection of disease severity than triggered by influenza carriage alone.

Infection-causing microbiota

The emergence of whole genome sequencing has allowed the respiratory microbiome to be viewed in an entirely new light, calling for a renewed understanding of traditional disease models of culprit pathogens. When comparing the entire microbiomes of children admitted to intensive care with respiratory disease versus healthy controls, individual bacteria and viruses poorly distinguished health from disease (Man et al. 2019). Even in the recent case of severe SARS-CoV-2 respiratory disease, the effects of this pathogen appeared to be mediated by dysbiosis of the lung and even gut microbiome resulting in advanced respiratory disease, an ecological disruption which appeared to vary with comorbidity, baseline microbial profiles intubation and oral hygiene (Battaglini et al. 2021).

Disturbances in the upper respiratory tract at a whole-microbiome level can result in the overgrowth of specific bacterial strains, increasing susceptibility to potential pathogens and thus an opportunity for viral invasion (Murphy et al. 2009). Indeed, in one model summarizing this pathobiome, competitive interactions occurred between Staphylococcus aureus, Streptococcus pneumoniae, and Haemophilus influenze, with the latter two demonstrating partial synergism, and viruses seemed to work hand-in-hand with one or two individual pathobionts alone (Bosch et al. 2013). This is a notable deviance from the Kochian model of infective disease, whereby a single pathogen is associated with disease and is therefore to be targeted.

The SARS-CoV-2 pandemic has renewed interest in the observable microbial trends upon infection with this virus. It has, however, become apparent that ‘infection with’ (i.e. presence of detectable virus) is not associated with dysbiosis in and of itself in a non-hospitalised population (Pelucchi et al. 2012, Qin et al. 2020). Within those hospitalized with SARS-CoV-2, nasopharyngeal swabs from SARS-CoV-2 positive patients demonstrated significantly lower alpha diversity than SARS-CoV-2 negative patients, reflecting a similar analysis of bronchioalveolar lavage specimens, where the overall microbiomes reflected those with community acquired pneumonia (Sajjan et al. 2006, Rosas-Salazar et al. 2016). Along a scale of self-reported respiratory disease severity within a SARS-CoV-2 population, ordinal logistic regression found a significant increase in microbial richness with increasing severity, but not with alpha diversity (Shilts et al. 2022). Another study of 18 ‘symptomatic’ patients found no significant alteration in microbial diversity but a significant decrease in abundance of Fusobacterium Periodonticum. These inconsistent results are accompanied by confusion over the role of subsequently abundant genera in these dysbiotic patients and whether these represent coinfection or mere opportunistic overgrowth. Similarly, lack of clear outcome measures (is dysbiosis a determinant or indicator of disease) has frustrated meaningful interpretations of these results. In future clear profiling of dysbiosis against objective stratified measures of respiratory disease will be needed to infer meaningful associations.

Nasopharyngeal Microbiota and Secondary Bacterial Infections

While the precise mechanism by which viruses predispose the host to bacterial disease remains unclear, a combination of in vitro and human analyses suggest virus-induced epithelial damage, immunosuppression and inflammation may facilitate suppression of bacterial clearance and promotion of bacterial adhesion and penetration (Fig. 2) (Hament et al. 2004, Avadhanula et al. 2006, Qin et al. 2020).

Figure 2.

Viral mechanisms that promote the pathogenesis of secondary bacterial infections

Secondary bacterial infections are considered a major cause of fatalities and morbidity in viral pandemics of the 20th and 21st century (Chertow and Memoli 2013). For example, 75% of confirmed Influenzavirus fatalities during the Spanish Flu Pandemic showed bacteriological and histological evidence of bacterial pneumonia, particularly due to Staphylococcus aureus or Streptococcus pneumoniae (Morens et al. 2008). A small analysis of SARS-CoV-2 patients without severe symptoms found no significant alteration in the nasal microbiome, suggesting this pandemic pathogen too requires cooperation from a dysbiotic NM to effect disease (De Maio et al. 2020).

Future Directions

Many of the above factors discussed can cause NM dysbiosis, potentiating upper respiratory tract viral infections. To prevent such infections, vaccinations and probiotics may be used to prevent colonization or overgrowth of the NM by pathogens (Bassetti et al. 2017). Antibiotics are valuable in treating severe for upper respiratory tract infections; however, they also have harmful effects which need to be mitigated in clinical practice. For suspected streptococcal throat infection, estimated responsiveness to antibiotics has been stratified into no antibiotic/delated antibiotic/immediate antibiotic groups for clinicians and adopted into mainstream clinical practice. Similar tools do not exist for infections of the sinuses or lower airways however (Pelucchi et al. 2012). Sinus disease also requires considerable further study as representing a large disease burden and a frequent reason for the prescription of antibiotics (Dekker et al. 2015, Feng et al. 2021). Chronic sinusitis as a long-term condition has been posed as a target for probiotic therapy. Enterococcus faecalis, and Lactobacillus rhamnosus have decreased frequency of exacerbations of rhinitis and SNOT-20 symptom scores respectively in the form of oral supplementation, while the SNOT-22 test has been improved by topical Lactococcus lactis (Mukerji et al. 2009, Habermann et al. 2011, Endam et al. 2020). This emphasis on gut flora continues in novel probiotic therapies to protect against viral respiratory infection, with protection being procured through upregulation of a series of immune mediators such as enhanced defensin expression, increased interferon-γ and interleukin-2, activation of plasmacytoid dendritic cell and the TLR-dependent inflammatory response (Sundararaman et al. 2020). It is yet unclear whether probiotics can consistently prevent otitis media (Scott et al. 2018). It is to be hoped that we will develop new strategies and tools to allow us to nurture our natural microbiome and move beyond the human-microbial arms race of the 20^th century.

Originality

The results of this review were presented in poster format at the Student and Foundation Doctors in Otolaryngology Conference (SFO) 2021 by ENT UK.

A more extensive previous version of this review was produced as part of a student selected component (SSC) in the MBChB programme at Edinburgh University. Key recent studies in the field of the upper respiratory microbiome and their references were appraised during the module and provided the basis for this review. A full description of this evidence-based research module has been published at the following link:

Matthew F Flynn, Developing critical appraisal and evidence synthesis skills in future microbiologists,  FEMS Microbiology Letters, Volume 368, Issue 18, September 2021, fnab114,  https://doi.org/10.1093/femsle/fnab114.

Conflict of interest statement

None declared.