Asthma Microbiome Studies and the Potential for New Therapeutic Strategies

Abstract

Recent applications of culture-independent tools for microbiome profiling have revealed significant relationships between asthma and microbiota associated with the environment, gut or airways. Studies of the airway microbiome in particular represent a new frontier in pulmonary research. Although these studies are relatively new, current evidence suggests the possibility of new therapeutic strategies for the treatment or prevention of asthma. In this article, recent literature on microbiota and asthma are critically reviewed, with a particular focus on studies of the airway microbiome. Perspectives are presented on how growing knowledge of relationships between the microbiome and asthma is likely to translate into improved understanding of asthma pathogenesis, its heterogeneity, and opportunities for novel treatment approaches.

Introduction

The potential relationships between microbial exposures, infections, and asthma have long been of and continue to stimulate much scientific interest. Areas of investigation have spanned from the role of early life microbial exposures and gut microbial colonization in the development of allergy or asthma [1–6], to studies of airway infections in established asthma [7–12]. Historically, approaches for characterizing microbial exposures or infection patterns in samples of interest have included serologic studies, microbial cultures or microorganism-specific molecular tests.

In recent years, more detailed analysis of total microbial community composition has been made possible by the development of culture-independent techniques with greater sensitivity in detecting microbial species, in particular bacteria. Conventional culture methods have low sensitivity in bacterial detection for several reasons, including organisms that are fastidious to grow or are as yet non-culturable. However, a commonly applied culture-independent approach to detect bacteria takes advantage of features of broadly conserved bacterial genes to detect species in a sample using polymerase chain reaction amplification, without need for prior knowledge of the specific bacterial composition present. Coupled with high-throughput profiling platforms such as microarrays and next-generation sequencing, such culture-independent techniques have enabled much deeper characterization of microbiota associated with specific environments, including mammalian hosts. Such studies have generated unprecedented insight into the potential complexity and diversity of microbial communities that exist in the environment and within humans [13,14].

These developments have led to a new or re-direction of efforts to understand relationships between asthma and microbial exposures or infections. These include studies of microbiota in samples derived from different living environments [5,15]. Among human niches, the gut has been the most extensively studied, and relationships between gut microbiota and allergy or asthma have also been examined [2,6,16,17]. More recently, investigations of the airway microbiome in pulmonary health or disease have emerged (11–12, 18–22), advancing our knowledge about the diversity of airway microbiota associated with obstructive lung diseases. Ultimately the goal in human microbiome studies is to achieve an integrative understanding of how the “microbiome” – defined as the totality of microbes, their genetic elements and interactions in a given niche – contributes to the development or persistence of relevant diseases. Despite the nascency of microbiome-focused research in asthma, recently gathered insights are generating new hypotheses about the role of microbiota in the development, persistence or heterogeneity of asthma.

The overall purpose of this article is to discuss ways in which growing knowledge about relationships between specific microbiomes and asthma may lead to the development of new therapeutic strategies. A brief overview of technologies and related considerations in microbiome studies will first be discussed, followed by a summary of recent asthma-focused microbiome studies. Select studies on the environmental and gut microbiota in asthma are highlighted, as relevant reviews in these areas are available [15, 23–26]. Recent airway microbiome studies of asthma will be discussed in greater detail. Finally, perspectives will be offered on how current and anticipated future knowledge about microbiota associated with asthma could translate into improved and/or novel treatment strategies for asthma.

Technologies for microbiome studies

The vast majority of microbiome studies to date have focused on studies of bacteria. This is due in large part to well-developed techniques based on analysis of the 16S ribosomal RNA (rRNA) gene, a highly conserved locus of the bacterial genome. Properties of this gene that facilitate bacterial community analysis are the presence of broadly conserved sequences, which flank a number of hypervariable regions. Universal primers targeting conserved sequences can be used to amplify the 16S rRNA gene, and polymorphisms within the hypervariable regions allow for phylogenetic identification of species using existing large 16S rRNA gene sequence databases. A number of 16S rRNA-based molecular tools exist, with higher resolution bacterial community profiling achieved primarily with 16S rRNA-based phylogenetic microarray or next-generation sequencing approaches [27–33]. The common theme is that these tools enable interrogation of a sample without a priori expectation of the bacteria present, such that fastidious or difficult to culture species also can be detected. The diversity of bacterial species known to exist is great and reflected in current bacterial taxonomies (i.e. phylogenetic classification of organisms). As a starting reference point, taxonomic classification for some of the major bacterial groups comprising the human microbiome are shown in Table 1.

Table 1.

Examples of taxonomic classification for some of the major bacterial groups identified as members of the human microbiome, including the respiratory tract. Given the diversity of human microbiota, the table is not intended to be comprehensive. Listed bacterial phyla, classes, orders, families and genera include organisms that have been identified in lung microbiome studies.

Phylum	ACTINOBACTERIA	BACTEROIDETES	FIRMICUTES	PROTEOBACTERIA
Classes	Actinobacteria	Bacteroidia Cytophagia Flavobacteria Sphingobacteria	Bacilli Clostridia Erysipelotrichia	Alphaproteobacteria Betaproteobacteria Deltaproteobacteria Epsilonproteobacteria Gammaproteobacteria
Orders	Acidimicrobiales Actinomycetales Bifidobacteriales Rubrobacterales	Bacteroidales Flavobacteriales Sphingobacteriales	Bacillales Clostridiales Lactobacillales	Burkholderiales Campylobacterales Enterobacteriales Moraxellaceae Neisseriales Pasteurellales Pseudomonadales Sphingomonadales
Families	Actinomycetaceae Bifidobacteriaceae Corynebacteriaceae Mycobacteriaceae Nocardiaceae Streptomycetaceae	Bacteroidaceae Flavobacteriaceae Porphyromonadaceae Prevotellaceae Sphingobacteriaceae	Bacillaceae Clostridiaceae Enterococcaceae Lachnospiraceae Lactobacillaceae Ruminococcaeae Streptoccocaceae Veillonellaceae	Bartonellaceae Burkholderiaceae Enterobacteriaceae Helicobacteraceae Neisseriaceae Pasteurellaceae Pseudomonadaceae Sphingomonadaceae
Genera	Actinomyces Bifidobacterium Corynebacterium Mycobacterium Nocardia Streptomyces	Bacteroides Capnocytophaga Porphyromonas Prevotella Sphingobacterium	Clostridium Faecalibacterium Lachnospira Ruminococcus Streptococcus Veillonella	Acinetobacter Escherichia Enterobacter Haemophilus Moraxella Pseudomonas Sphingomonas

As with any biologic technique, available tools to study the microbiome each have respective advantages and disadvantages. Important considerations in interpreting results of microbiome studies include the depth of community characterization achieved. For example, in sequencing-based studies, the number of quality-filtered reads analyzed on a per sample basis can be important depending on the complexity of a sample type. Lower numbers of analyzed reads generally provide less community resolution, which can be tested statistically. For further insight various reviews have been published that discuss these considerations and also describe other approaches for bacterial microbiome studies [31–33].

Finally, it is important to note that sample collection, processing and preparation methods also can affect the community composition characterized. Biases can be introduced at multiple steps including environmental or reagent sources of contamination, the variable efficiency of DNA extraction methods as some organisms are more difficult to lyse [34], and choice of 16S rRNA primers for PCR reactions. In particular, 16S rRNA sequencing protocols often utilize primers that target specific hypervariable region(s) of the gene. It is recognized that certain hypervariable regions do not capture well particular bacterial groups or species. Thus certain organisms may not be identified even if present in a sample [35–36]. The relative importance of these and other technical considerations will depend on the research questions, including the specific microbiome and types of samples to be studied, as well as potential bacterial groups of interest.

Environmental microbiota and asthma

A number of environmental factors have been associated with either risk for or protection against asthma or allergic sensitization (Figure 1). Microbial exposures can be invoked as plausible mechanisms for many of these associations. For example, subjects living in environments that afford greater exposure to animals (dogs, cats, farm animals) generally have lower risks for allergy or asthma [1, 3–5, 15, 37]. Raw milk consumption also has been associated with decreased atopy or asthma in childhood [38], while antibiotic use during pregnancy or perinatally is linked to increased risk for asthma-related outcomes [39]. However, meta-analyses of pooled studies have also suggested less strong or even absent associations of animal exposure and antibiotic use with asthma [40–42].

Figure 1.

Areas in which microbiota are likely to have an influential role in asthma. Microbial exposures associated with early life environmental factors contribute to risks for developing allergy, early onset asthma, or both. Among adults with asthma, the airway microbiome may be most influential in phenotypes associated with predominantly non-Th2 patterns of airway inflammation.

Studies in which indicators of microbial load or diversity have been measured, support the argument that environmental microbiota are likely an important mediator, if not a causal factor, in asthma development. Levels of endotoxin (produced by gram-negative bacteria) in mattress dust are inversely related to atopic asthma and sensitization in children [4]. Culture-independent surveys of bacteria and fungi in mattress and settled bedroom dust also suggest that microbial diversity is negatively associated with asthma risk [5]. Collectively these findings lead to speculation about how microbial exposures from one’s environment may modify risk for asthma in early life [15]. As one possibility, early evidence from the PASTURE study suggests microbial exposures may influence DNA methylation patterns over the first few years of life [43]. Thus stimuli from environmental microbiota could potentially shape expression of host genes and/or immune responses without necessarily effecting changes in host microbiota composition.

Gut microbiome and asthma

Gut microbiota provide antigenic stimulation to the immune system, educating its development in early life [23, 24]. Thus the composition of the gut microbiota can play an important role in shaping immune phenotype [25–26, 44]. Earlier studies of infant stool specimens have found that gut colonization patterns within the first three months of life differed between infants who did or did not develop allergic sensitization at 12 months of age [2]. Specifically more Clostridia and fewer Bifidobacteria species were identified from atopic children compared to non-atopic children. Different species within a specific bacterial family also may have different immune-stimulatory effects, as has been reported for Bifidobacteria as well as Lactobacilli [16, 45]. For example among breast-fed infants, Bifidobacterium bifidum was the main Bifidobacterium species found in fecal specimens from non-allergic infants, while B. adolescentis and B. longum were more prevalent in those who developed allergy [16].

Recent studies have implicated other bacterial species or bacterial diversity in the gut with the development of asthma [6, 17]. In a prospective study of 117 children classified by the Asthma Predictive Index (API), the prevalence of Bacteroides fragilis and other anaerobic bacteria cultured from fecal samples taken at three weeks of age was higher in API-positive vs. API-negative subjects [17]. In a birth cohort study of 411 children at high-risk for asthma, stool samples collected at one and twelve months after birth were analyzed by 16S rRNA-based denaturing gradient gel electrophoresis (DGGE) and also by conventional cultures [6]. Reduced bacterial diversity, as estimated from DGGE band analysis, was inversely associated with allergic sensitization in the first six years of life, though not with the development of asthma. Collectively, evidence from studies of gut and environmental microbiota indicate that decreased exposure to a diversity of microbes, including specific microbial consortia, have negative implications for immune health that affect risks for allergy and asthma.

Insights from Animal Studies of the Microbiome

Studies described up to this point have been human investigations. Recent studies in animal models are noteworthy for their demonstration that gut microbiota are important drivers of immune responses that can modulate allergic inflammation and features of asthma in the airways [46–48]. For example, germ-free compared to specific pathogen-free mice exhibit increased accumulation of mucosal invariant natural killer T cells (iNKT) in the colonic lamina propria as well as in the lung [46]. These germ-free mice, in an ovalbumin-model of allergic asthma, also demonstrated increased airway resistance and eosinophils in lung lavage and tissue. Feeding of conventional microbiota to neonatal, but not adult, germ-free mice prevented iNKT cell accumulation in the lungs and mitigated any evidence of allergic inflammation and asthma. In another study, administration of vancomycin to neonatal mice led to reduced bacterial diversity and similarly enhanced susceptibility to allergic asthma in a murine model [48]. Finally, several mouse studies have demonstrated that oral administration of bacterial species with immunomodulatory properties can modulate features of allergic asthma in the lungs. These include oral supplementation with specific Bifidobacterium and Lactobacillus species, which have been shown to reduce Th2-cytokine production, eosinophilic inflammation and/or promote T-regulatory or Th17 immune responses [49–52]. Findings from these and other studies suggest that interventions delivered via the gut during a critical window of time may alter susceptibility to or attenuate allergic asthma.

Respiratory microbiota and asthma

Previous studies of respiratory bacterial infections and risk for asthma

Acute respiratory infections are known triggers of asthma exacerbation, but evidence of chronic airway colonization or infection by specific organisms also have been linked to asthma. Earlier studies, often based on serologic tests, suggested relationships between Mycoplasma pneumoniae and Chlamydophila pneumoniae infections and the onset of asthma [53–55]. Subsequent studies have applied PCR- based approaches to more directly evaluate the presence of these bacteria in respiratory specimens [7, 9]. In a study of 95 subjects with asthma and 58 healthy controls, C. pneumoniae was more frequently detected by PCR and/or immunoglobulin measurements in induced sputum from asthmatics, particularly among those with poorly controlled or non-atopic asthma [9]. Species-targeted PCR tests of endobronchial biopsies or bronchoalveolar lavage fluid found evidence of M. or C. pneumoniae in 56% of asthmatics, compared to only one healthy control [7]. However, despite many studies that have focused on M. pneumoniae or C. pneumoniae, their role in asthma remains inconclusive [56–58].

Recent studies have examined relationships between other airway-associated microbiota and asthma. In a study of 321 neonates, culture-based studies of hypopharyngeal samples collected at 1 month of age found that evidence of early upper airway colonization by Streptococcus pneumoniae, Haemophilus influenzae, and/or Moraxella catarrhalis, was significantly associated with persistent wheeze during the first 5 years of life, and with asthma at age 5 [10]. Blood eosinophils and total IgE at 4 years of age also were significantly higher in those with early colonization. These same species have been associated with acute wheezing episodes in children (overall odds ratio 2.9), independent of a similar association found with viral infections [59].

Recent studies of respiratory microbiota and asthma

Studies published within the past three years have begun to apply 16S rRNA-based sequencing and microarray techniques to profile more deeply airway microbiota and investigate their potential relationships to asthma. Results from these studies collectively indicate that the composition of bacterial microbiota detected from the lower respiratory tract differ in those with asthma compared to healthy subjects [11, 12, 60]. Whether a resident community of microbiota exists in the lower airways of healthy individuals, however, is unclear. A low level of bacterial burden is detectable, the composition of which tends to be similar to bacterial groups typically associated with the oropharyngeal niche [20]. Even with rigorous efforts to minimize oral contamination of lower airway specimens during their collection, it is challenging to discern whether detection of species traditionally viewed as oral flora represent carry-over contamination, transient microbial colonization from aspiration, or low-level permanent colonization. However, in the setting of obstructive airway diseases like asthma, the predominant bacterial communities associated with asthma have primarily represented bacterial groups that are not dominant members of the oral microbiome [12, 60].

One of the first studies to investigate the composition of bacterial airway microbiota among asthmatic patients utilized a traditional 16S rRNA clone ial sequencing approach to analyze respiratory samples from 24 adults (11 with asthma, 5 with COPD, 8 healthy controls) and 20 children (13 with difficult to control asthma, 7 healthy controls) [11]. Although this approach has much lower resolution in profiling bacterial communities compared to high-throughput microarray [61] or next-generation sequencing tools, members of the Proteobacteria phylum, in particular Haemophilus spp., were more commonly identified from bronchial brushings or lavage fluid from individuals with airway disease (asthma or COPD). In contrast, members of the Bacteroidetes phylum, such as Prevotella spp., were more frequently found in specimens from healthy subjects. All adult asthma and 60% of the COPD patients, however, were on inhaled corticosteroid therapies. Thus the impact of corticosteroid use on the findings is unclear. Potential differences in microbiota composition by airway geography also were explored in this study by comparing samples from the nasopharynx (NP), oropharynx and left upper lung lobe (LUL). In general NP bacterial communities were distinct from that found in the oropharynx or LUL. LUL communities also differed between patients with airway disease (asthma or COPD) versus healthy controls. Another limitation of this study was the grouping of COPD together with asthma patients in reported differences from healthy controls in airway bacterial community composition.

In a larger study of 75 adults including subjects with mild-moderate asthma and healthy controls, between-group differences in lower airway bacterial community composition also were observed on analysis of bronchial epithelial brushings [12]. A significant correlation between the severity of airway hyperresponsiveness and bacterial diversity also was found. This extended to the identification of approximately 100 bacterial taxa that were strongly correlated with this pathophysiologic feature of asthma (a taxon being defined here as a group of bacterial species with > 97% homology in their 16S rRNA gene sequences). These included members of the Proteobacteria phylum, including families containing organisms with pathogenic potential such as Pseudomonadaceae, Enterobacteriaceae, Burkholderiaceae, and Neisseriaceae. Other highly correlated taxa included species with known functional repertoires that could be hypothesized to contribute to asthma-related disease mechanisms [12]. All asthmatics examined in this study had mild to moderate asthma, but were also taking a standardized moderate dose of inhaled fluticasone therapy. Thus, like the earlier discussed study [11], the potential influence of inhaled corticosteroids on the findings is unknown. However, the depth of bacterial community resolution achieved in this study was much greater, using a 16S rRNA-based microarray platform with the capacity to detect both very low and high abundance bacterial communities in a sample with equal efficiency [61]. This study also found that asthmatic subjects who improved after six weeks of clarithromycin therapy had higher airway bacterial diversity at baseline, compared to those that did not respond to therapy. In summary this study was one of the first to describe relationships between particular clinical features of asthma and the bacterial airway microbiome, suggesting that members of the bronchial microbiota may exert important functional effects.

Results of a recent small study indicate that differences in the composition of bacterial airway microbiota that have been associated with asthma are likely related to the disease of asthma itself [60], and not entirely attributable to inhaled corticosteroid use. 16S rRNA pyrosequencing analysis of induced sputum collected from 10 non-asthmatic and 10 mild asthma patients, none of whom were regularly taking inhaled corticosteroid therapy, found Proteobacteria present in significantly higher proportions in asthmatic vs. non-asthmatic patients. Specific bacterial families found in higher relative abundance among asthmatics included Enterobacteriaceae and Neisseriaceae. Although only 60% of the quality-filtered sequence reads could be confidently assigned to a bacterial family using a well-known 16S rRNA database, calculated bacterial diversity also was higher in the asthmatic group. Despite limitations of a small sample size and potential biases in bacterial community profiling due to the sample processing procedures used, the findings overall are consistent with that of previous studies [11, 12] and provide further evidence that airway microbiome features associated with asthma in studies to date are likely due to the disease itself.

Finally, it clearly would be of interest to investigate other types of airway microbiota and their potential role in asthma, such as fungi. While there have been some efforts to characterize fungal microbiota in the context of asthma [5, 62], it is important to note that best molecular approaches to profile fungal microbiota in a global manner are still active areas of research. Moreover, there is a need for better fungal sequence reference databases.

Avenues for new therapeutic approaches

Despite the relatively recent development of lung microbiome research, particularly with respect to asthma, it is not premature to begin to consider how growing knowledge in this area could transform our understanding of asthma and translate into novel or improved treatment approaches. Asthma is heterogeneous, however, and a variety of asthma phenotypes and classification schemes have been proposed [63]. Since this heterogeneity likely reflects different underlying disease mechanisms, contributions of the microbiome may be different in the etiopathogenesis of asthma compared to in established asthma and specific phenotypes, as suggested in Figure 1. Thus, determining where in the spectrum of asthma heterogeneity microbiome contributions are most important would shape the development of microbiome-based approaches to asthma treatment.

Current evidence points towards three areas in which the microbiome may play an important role in asthma: 1) the development of asthma in childhood that tends to be linked to allergic sensitization, 2) asthma in adults that is not associated with Th2 inflammation, such as neutrophil-predominant patterns, and 3) treatment-resistant asthma such as to corticosteroid therapy. Possible microbiome-driven therapeutic strategies for each of these scenarios are proposed and discussed further below.

Early-life microbiome composition and asthma development

Evidence from environmental and gut microbiome studies strongly suggests that microbial factors shape allergic sensitization and asthma development in susceptible individuals. Hence early life interventions that promote a ‘healthy’ human microbiome constitution may have benefits that include the possibility of asthma prevention. Specific environmental interventions to increase exposure to microbial diversity may be impractical due to individual circumstances. Alternative strategies might include oral supplementation with prebiotic nutrients, specific live microbial species (probiotics), or microbiota-derived products, to promote a gut microbiome composition that induces tolerogenic immune responses [25, 44]. Though these ideas are not new, hurdles remain from scientific and regulatory standpoints to investigate and develop probiotic-based therapies [64]. As these issues are being addressed in the probiotics field, approaches that succeed in achieving the same goal but do not require delivery of viable organisms may be more tenable.

Neutrophil variant forms of asthma

A large proportion of adult asthma patients do not exhibit hallmarks of Th2-driven airway inflammation. Thus it is particularly plausible that the airway microbiome may be important in the pathogenesis of certain non-Th2-driven phenotypes of asthma. For example, although neutrophilia is commonly seen in conjunction with inhaled corticosteroid use [63], sustained colonization of the airways by specific microbial species could also independently promote neutrophilic inflammation. Evidence in support of this is the consistent finding across airway microbiome studies of increased prevalence of Proteobacteria in asthmatic patients. This large phylum encompasses many known pathogenic species responsible for both acute respiratory and gastrointestinal illnesses, including H. influenzae, Pseudomonas, Neisseria, and Burkholderia species, as well as the Enterobacteriaceae family. Enterobacteriaceae comprise a large family of species typically associated with the GI tract (Klebsiella pneumoniae being one exception) but that also have been identified in microbiome studies of COPD as well as asthma [12, 21]. Neutrophilic inflammation promoted by Proteobacterial species in the context of asthma has been shown. For example, H. influenzae infection in an ovalbumin-murine model of allergic asthma caused induction of Th17 immune responses that promoted neutrophilic but suppressed eosinophilic inflammation [65]. Moreover, evidence of chronic H. influenzae infection was found one month after inoculation, and the airway inflammation seen in H. influenzae-infected allergic mice was resistant to steroid treatment in contrast to the steroid-responsive decrease in inflammation seen in uninfected allergic mice [66]. Human dendritic cells also produce greater pro-inflammatory cytokines (IL-23, IL-12p70) in response to H. influenzae and M. catarrhalis, compared to exposure to commensal species such as Veillonella [67].

These studies indicate that airway infections by specific microbiota may induce, augment, or potentially even modify the predominant type of airway inflammation seen in asthma. Thus, a possible therapeutic strategy could be use of antimicrobials or vaccine therapies that target specific microbiota in asthmatics demonstrating evidence of such colonization. Conversely, microbiome-targeted treatment approaches could be construed from the perspective of how to promote a more functionally balanced airway microbiome, such that pathogenic microbiota are not able to exert dominant effects. This strategy has been studied and promulgated more in the gut microbiome literature [24–25, 68]. Studies of experimental respiratory infections also show that immune responses to pathogenic infections depend on the background microbiota present [69, 70]. Thus an understanding of microbiota that are negatively associated with features of asthma could be useful towards developing approaches to promote these microbiota, or specific functions they express that counteract detrimental inflammatory processes.

Treatment-Resistant Asthma

Treatment-resistant asthma is a particular challenge clinically often manifest in those with severe asthma. Novel therapeutic approaches are much needed. A number of factors may contribute to difficult to control asthma, including the possibility that prescribed therapies do not address underlying mechanisms of disease in patients. For example, inhaled corticosteroids are less effective in asthmatics without evidence of Th2-driven airway inflammation [71]. Another possible theory for treatment-resistant asthma is modulation of intended therapeutic effects by microbiota, whether at the host response level or of the therapy itself. Foundation for this proposed theory is based on hypothesized functions of microbiota we have noted in previous [12] and ongoing studies. Additional supportive evidence is extrapolated from the known capabilities of gut microbiota to transform therapeutic agents [72], and also from the field of bioremediation. Thus identification of microbiota that modulate responses to therapies may potentially lead to development of approaches that block these effects.

Conclusions

Studies of the lung microbiome using modern molecular tools are re-defining our perception and understanding of respiratory microbiota associated with asthma and other airway diseases. It is becoming increasingly clear that compared to healthy individuals, asthma is associated with a different lower airway microbiota composition. The asthmatic airway microbiome includes species with pathogenic potential as well as species with potential immunomodulatory or metabolic properties relevant to certain asthma-related disease mechanisms. As the field advances, it will be important to determine where microbiome contributions to asthma heterogeneity are significant. This would enable focused development of microbiome-driven therapies, such as manipulation of microbiota composition and/or their functional derivatives, as novel approaches to asthma treatment.