Introduction

Chronic obstructive pulmonary disease (COPD) is a leading cause of morbidity and mortality worldwide and is characterized by persistent airflow obstruction and respiratory symptoms.¹ Despite the heterogeneous pathogenesis of COPD, diagnostic and therapeutic approaches remain rather circumscribed,² driving the need for further comprehension of the underlying mechanisms of COPD etiology.

Over the past decade, the term “gut-lung axis” has gained significant attention as a way to elucidate the complex interplay between the host, microbiome, and respiratory diseases. Recently, there has been increasing interest in the potential role of the gut-lung axis in the pathogenesis of COPD. Growing evidence suggests that bidirectional communication between the 2 organs plays a crucial role in COPD progression and exacerbation.^3,4 In fact, dysbiosis, or an imbalance in the microbiome, contributes to COPD development,^5,6 and disruption in intestinal barrier integrity and function was found in COPD patients but not in non-COPD control individuals.^7,8 An expanding corpus of research indicates that the constitution of both the gut and respiratory microbiome is distinct in individuals with COPD compared to healthy individuals. This underscores the plausible notion that alterations in the composition of the microbiome may lead to distinct host immune reactions and clinical presentations via intricate interplay with the host.

The rapid advancement in technologies has played a crucial role in comprehending the microorganisms residing in our body, their functionalities, and their effects on human health and disease.⁹ Sequencing of the 16S rRNA gene is a widely used method to classify microbial taxa due to its ability to provide a fast and cost-effective analysis of microbial composition.¹⁰ The commonly used indices to evaluate overall patterns of variation in the microbiome are α-diversity and β-diversity, which represent the diversity of the microbial community within individual samples and the dissimilarity in microbial composition within pairs of samples, respectively.^10,11 Investigating microorganisms that exhibit differential abundance between the groups of interest is another commonly employed analytical approach. Exploring the correlation between modified taxa and disease phenotypes can aid in identifying microbial indicators between groups and directing treatment.¹¹ Furthermore, the swift advancement of novel technologies based on exact sequence variants, integrating metagenomics and metabolomics, enables precise identification of particular species and analysis of microbial metabolism and functions.¹¹

The objective of this review is to provide a concise overview of the role both gut and respiratory microbiome play in the development of COPD. This will be accomplished by compiling current literature on the microbiome profile in stable and exacerbated cases of COPD, as well as exploring the biological mechanisms through a discussion of relevant experiments conducted on murine models.

Dysbiosis of the Gut Microbiome in COPD

Gut Microbiome in Health

The microbial inhabitants in the gastrointestinal tract constitute the most abundant microbial group within the human body, so the gastrointestinal system represents the most extensively investigated microbiome in scientific inquiry. The structural composition of the gastrointestinal microbiome encompasses the microbiota inhabiting the oral cavity, esophagus, stomach, small intestine, and colon. In both rodents and humans, the regions of the cecum and proximal colon harbor the greatest microbial biomass, while the small intestine contributes to a lesser but still significant extent.^12,13 Distinct microbial compositional profiles are discernible along the length of the gastrointestinal tract. Major phyla found in the oral cavity and esophagus include Firmicutes, Bacteroidetes, Proteobacteria, Actinobacteria, and Fusobacteria. Because of the high acidic environment in the stomach, only a limited number of microorganisms can survive in the stomach and duodenum. Some of the common genera include Bacillales incertae sedis, Streptococcaceae, and Enterobacteriaceae. The intestine is dominated by Proteobacteria and Lactobacillales. Last, the featured phyla in the terminal ileum and colon are Bacteroidetes, Firmicutes, Verrucomicrobia, Proteobacteria, and Actinobacteria.¹³

Relationship Between the Gut Microbiome and Stable COPD

Studies have revealed that the composition of the gut microbiome in COPD patients varies significantly from that of healthy individuals. When assessed by β-diversity, significant differences in the composition of the gut microbiome were found in stable COPD individuals compared to healthy individuals, although the α-diversity between the 2 groups showed no significant difference.^6,14 When compared within COPD patients, both the bacterial richness and diversity failed to gauge the severity of stable COPD.^6,15 However, a decline in lung function within the same patient may be related to altered α-diversity. A 1-year follow-up study¹⁶ of COPD patients uncovered a compelling correlation: an increase in microbial diversity was concomitant with a marked reduction in forced expiratory volume in 1 second (FEV₁) exceeding 40ml. (It has been observed that the annual reduction in FEV₁ for nonsmokers is approximately 30ml. In contrast, for individuals who partake in smoking, this annual diminution fluctuates between 40 and 60ml.)^17,18

Regarding the differential abundance of specific taxa, the most mentioned alteration in the gut microbiome of COPD is a reduction of the phylum Bacteroidetes^6,15,16 and its genus Bacteroides,¹⁴ which is related to decreased FEV₁ and forced vital capacity (FVC).^15,16 (Table 1). Other genera observed to be diminished in COPD include Roseburia and Lachnospira.¹⁴ The phylum Firmicutes proliferated in COPD patients and was related to decreased lung function.^6,16 Dominating bacterial genera observed in fecal samples of COPD patients but not in healthy individuals are Prevotella,⁶ Streptococcus, Rothia, Romboutsia, and Intestinibacter.¹⁴ Several members from the genera Prevotella and Streptococcus were found to be negatively associated with FEV₁.^14,16 When compared within COPD patients, the genera Fusobacterium and Aerococcus had greater abundance in stage 3-4 COPD when compared to stage 1-2 COPD.¹⁵

In addition to microbial profiling, a significant difference was also found in the metabolome of gut microbiota between healthy individuals and stable COPD patients.¹⁴ It is interesting to note that the severity of COPD was found to be linked with the amount of short-chain fatty acids (SCFAs), which are gut microbial metabolites that demonstrate protective effects on respiratory diseases.^6,19 SCFAs are byproducts of the fermentation of dietary fibers that are released into the lumen and peripheral circulation.²⁰ In a mouse model of emphysema, SCFAs demonstrated significant prophylactic potential against the advance and intensity of emphysema, concurrently attenuating inflammatory responses.^21,22 A comprehensive study involving a cohort of 35,339 Swedish women revealed that long-term consumption of dietary fibers was associated with a 30% lower risk of COPD.²³ SCFAs have also exhibited therapeutic potential in asthma by mitigating airway inflammation and reducing the production of immunoglobulin E, interleukin (IL)-4, and IL-8, along with immune cell counts.²⁴ It is conceivable that individuals with more severe conditions lost the protective effects of specific metabolites generated by commensal intestinal microbiota.

Relationship Between the Gut Microbiome and COPD Exacerbations

The dysbiosis of the gut microbiome in acute exacerbation of COPD (AECOPD) patients is rarely investigated. In terms of microbial diversity and richness, contradictions exist. A study reported that decreased richness and diversity of gut microbiota were found in AECOPD patients but not in healthy individuals and stable COPD patients,²⁵ while another study found no alterations in microbial diversity with varied abundances of bacteria.²⁶ Phylogenetically, increased abundances of Bacteroidetes and Proteobacteria, as well as decreased abundances of Firmicutes and Actinobacteria, were observed in AECOPD patients but not in stable COPD and healthy individuals. Moreover, weighted gene co-expression networks revealed a significant correlation between members of Firmicutes and Actinobacteria and lower lung function as well as higher levels of inflammatory markers, suggesting that these phyla may contribute to the progression of COPD by exacerbating inflammation.²⁵ (Table 2)

Relationship Between the Gut Microbiome and Host Immunity in COPD

Scientific evidence underscores a correlation between alterations in the abundance of specific gut microbiota in COPD patients and systemic inflammation markers, alongside inflammatory cell counts. (Table 1 and Table 2) Therefore, it appears plausible that gut microbiota may contribute to COPD pathogenesis through the modulation of systemic inflammation. (Figure 1) Members from the phyla Firmicutes and Actinobacteria were reported to have a highly significant relationship with blood inflammatory indices, including IL-6, IL-8, tumor necrosis factor-alpha (TNF-α), procalcitonin, and C-reactive protein, in COPD, suggesting that these bacterial taxa may be involved in the development and progression of COPD by exacerbating systemic inflammation.²⁵ Another study revealed that a higher abundance of a member from Bacteroides was linked to better lung function and lower blood eosinophil counts. This is notable because high blood eosinophil counts are known¹⁵ to be associated with an increased risk of COPD exacerbations, mortality, and reduced FEV₁. The above changes are likely to arise from the epigenetic regulation of immune cells by SCFAs, which can bind to free fatty acid receptors (FFARs) on immune cells such as neutrophils, monocytes, and macrophages, causing an inhibition of the expression of proinflammatory factors, including IL-6, IL-1β, and TNF-α.²⁷ In addition to attaching to FFARs, SCFAs exert anti-inflammatory effects by inhibiting histone-deacetylase complexes and suppressing nuclear factor-kappa B (NF-κB) signaling.²⁸ (Figure 2) Notably, an elevated abundance of Bacteroidetes and a decreased abundance of Firmicutes were associated with higher levels of SCFAs,²⁹ which is consistent with the aforementioned negative correlation between SCFA levels and COPD severity.⁶ In other words, the reduction in Bacteroidetes and Firmicutes results in decreased levels of blood SCFAs, which deprives COPD patients of the protective effect of SCFAs against systemic inflammation. This ultimately leads to an increase in lung inflammation and may be a key mechanism by which alterations in the gut microbiome contribute to COPD. Another proposition that needs further inquiry is whether there exists specific gut microbiota that do not confer protection against COPD through the secretion of SCFAs.

Relationship Between Antibiotic Therapy and the Gut Microbiome in COPD

Antibiotics are frequently used in the management of COPD, especially for addressing pneumonia during acute exacerbations. Nevertheless, their direct influence on microorganisms poses a significant challenge in analyzing gut microbial sequencing. The effect of antibiotic use on gut microecology in COPD patients remains uncertain, primarily because most studies have excluded individuals who recently used antibiotics.^14-16,25 This lack of information also extends to severe AECOPDs, despite the pressing need for further research and medical support for this specific population. Li et al found that antibiotics led to a significant decrease in gut bacterial abundance in mice, resulting in dysbiosis and reduced levels of SCFAs, which played a role in COPD progression.⁶ However, in clinical practice, the complexities of different pathogenic bacteria and bacterial loads in the lungs of AECOPD patients make it challenging to simply characterize the pros and cons of antibiotics. Further investigation is needed to better understand the impact of antibiotics on gut microecology in COPD. Longitudinal study designs could prove beneficial in providing more comprehensive insights into the long-term effects.

Dysbiosis of the Respiratory Microbiome in COPD

Respiratory Microbiome in Health

Lungs were previously thought to be completely sterile until researchers detected bacterial DNA in healthy lung samples.³⁰ In contrast to the gut, the respiratory tract is a low-biomass environment with microbial density diminishing in a gradient from the upper to the lower respiratory tract.³¹ The most prevalent phyla of bacteria found in the respiratory tract are Firmicutes and Bacteroidetes, while Prevotella, Streptococcus, and Veillonella are the most common genera.³² The pulmonary microorganism community may have originated from the upper respiratory tract through microaspiration and been further shaped by host defense mechanisms; thus, a healthy respiratory microbiome is more likely to be a result of intermittent aspiration of taxa rather than reproduction of the core resident community.^4,20,32 Several sampling methods are available to profile the respiratory microbiome, including nasal brushing, induced or expectorate sputum, bronchoalveolar lavage, epithelial brushing, and bronchial biopsies, while sputum has been a preferred sampling method in terms of accessibility and difficulties with bronchoscopy in COPD patients.³¹ The application of sputum sampling also eases the implementation of longitudinal studies, for example, comparing samples before and after exacerbations.^33-50 However, contamination from microorganism-rich oral cavities must be considered, especially since the propensity for contamination increases in environments with a low biomass, such as the lower respiratory tract.^20,51 Therefore, sputum profiling cannot entirely encapsulate the composition of the pulmonary microbiome. Previous studies have suggested that oral antiseptic rinsing before collecting sputum samples may reduce contamination in sputum cultures.^52,53 However, the role of oral antiseptic rinsing in sputum microecological sequencing remains unexplored. Additional research is required to standardize protocols for spontaneous and induced sputum sampling.

Relationship Between the Respiratory Microbiome and Stable COPD

Studies have shown significant deviation of the respiratory microbiome in relatively severe COPD patients, whereas relatively milder patients had a respiratory microbiome more comparable to that of healthy individuals. When compared within stable COPD, the bacterial biodiversity can reflect the severity of COPD, as sputum samples from severe patients were found to have a decreased α-diversity compared to those from milder patients.^54-56 When compared with healthy individuals, several studies highlighted a decrease in α-diversity in stable COPD compared to healthy individuals,^41,57-60 while others reported no significant changes or even a higher α-diversity.^61-63 A meta-analysis focusing on the alteration of α-diversity in COPD found that there was no significant but a slight trend toward decreased microbial diversity measured by the Chao1 index, which is an index that measures microbial richness but not evenness.⁶⁴ The conflicting findings in α-diversity may be attributed to variations in the severity of the enrolled patients, as those with more severe symptoms are likely to have an altered respiratory microbiome with a declined microbial diversity. In terms of β-diversity, principal coordinate analysis was able to separate lung microbiota samples from healthy controls and COPD patients.^61,65,66 Taken together, changes in microbial richness and evenness indicate diminished commensal microbiota and proliferated potentially pathogenic microbiota, resulting in a deviated and restricted pattern in the respiratory microbiome of severe COPD patients.

In regard to the distribution of particular taxa, the most frequently mentioned alterations in the respiratory microbiome of COPD include an increase in the abundance of Proteobacteria (including Haemophilus,^41,60,62 Moraxella,^41,59,62 and Pseudomonas^58,60,67-69) and a decrease in the abundance of Prevotella^58-60,70 (Table 1). An elevated abundance of Proteobacteria was observed in COPD patients compared to healthy controls,⁶² and dominance of the phylum was correlated with more severe COPD.^56,71 This is likely related to the fact that COPD patients have a respiratory microbiome colonized by the pathogenic bacteria Haemophilus influenzae and Moraxella catarrhalis,^36,72 as proliferation of representative pathogens (including H. influenzae, M. catarrhalis, Streptococcus pneumoniae, and Pseudomonas aeruginosa) was detected in patients with more severe symptoms.⁴² However, it is worth mentioning that more studies have demonstrated an increase in multiple species from the same genera, resulting in an overall increase in the whole genera (Haemophilus^41,60,62 and Morexella^41,59,62), which indicates a potential synergistic pathogenic effect among congeners. Correspondingly, Haemophilus⁴⁴ and Pseudomonas⁶⁷ were found to be related to more severe airflow limitation. The intricate interactions among taxa have resulted in greater heterogeneity of the respiratory microbiome in COPD, which can be classified into subgroups characterized by distinct microbial dominance. For example, a study found that the abundance of H. influenzae was positively associated with S. pneumoniae but negatively associated with P. aeruginosa.⁴⁸

A common pattern of intraspecies interaction is that the proliferation of potential bacteria is accompanied by a reduction of commensal bacteria. For instance, Moraxella³⁷ and Pseudomonas⁶³ were found to be negatively related to the α-diversity in sputum samples. The commensal bacteria reported to be reduced in COPD are Prevotella^58-60,68,70 and Veillonella,^68,72 of which Prevotella was found to be associated with better lung function and alleviated symptoms.^59,73,74 Other genera that were reduced in COPD include Treponema,^60,67 Actinomyces,^54,58and Fusobacteria.^60,70 Despite being considered a commensal genus, the relationship between Streptococcus and COPD is inconsistent based on various studies. Several studies have observed an increased abundance of Streptococcus in COPD,^{41,59,69,70,73} which has been linked to decreased lung function^73,74 and could be associated with infection by the potentially pathogenic S. pneumoniae.^36,72 In contrast, other studies reported a decline in the abundance of Streptococcus in COPD.^62,68 A plausible explanation for this discrepancy might be attributed to the divergent roles performed by distinct species within the Streptococcus genus. This is fundamental because 16S rRNA analysis is limited to differentiation at the genus level and is, therefore, incapable of distinguishing between individual species and their respective functionalities. Another possible reason is that the specific microbiota itself can develop contradictory functionalities; for example, S. pneumoniae is a highly adapted commensal but has the ability to cause severe disease according to specific bacterial and host factors.⁷⁵ This complex and dynamic relationship between the host and its microbiota is a key focus of current research to better understand and manage various diseases.

Relationship Between the Respiratory Microbiome and COPD Exacerbations

The common feature of the respiratory microbiome during COPD exacerbations is decreased α-diversity and Proteobacteria dominance. When compared with stable COPD, decreased α-diversity was found in COPD exacerbations using 16S rRNA sequencing^26,33,66,76 and metagenomic sequencing.⁷⁷ A meta-analysis found that there was a slight trend toward decreased α-diversity in COPD exacerbations compared to stable COPD.⁶⁴ In accordance with β-diversity, contradictory results also exist.^{26,34,37,49,66,67} A decrease in α-diversity was correlated with higher frequency in exacerbations^76,77 and higher mortality,⁷⁸ whereas deviation of microbial composition was linked to reduced FEV₁ and FVC during exacerbations.³⁷ Notably, remodeling of the respiratory microbiome is not commonly seen in all exacerbation cases but only in a subgroup^33,34,37,76; for example, a study found that significant deviation of pulmonary microbial composition was only present in 41% of all patients with exacerbations.³⁷ In fact, great heterogeneity in the respiratory microbiome usually exists within COPD patients during exacerbations,^34,35 indicating the existence of various triggers of AECOPD.

Consistent with findings in stable COPD, Proteobacteria dominance (especially Haemophilus and Moraxella) and diminishment of commensal bacteria (especially Veillonella) were also present in AECOPD (Table 2). Several studies reported a slight trend of increased Proteobacteria and decreased Actinobacteria and Firmicutes,^26,33-35,77 of which Proteobacteria dominance was related to increased mortality.⁵⁶ The Proteobacteria dominance is possibly due to an elevation in both representative and nonrepresentative pathogenic taxa from the genera Moraxella (including M. catarrhalis)^{26,33,39,46,47,50,76,77,79} and Haemophilus (including H. influenzae),^{26,33,35,39,46,47,50,76,79} suggesting that synergic effects of bacteria from the same genus have contributed to AECOPDs. The risk of exacerbations was elevated with the presence of H. influenzae and M. catarrhalis.^45,46,76 During exacerbations, the colonization of the Haemophilus genus is often found concurrently with a reduction of other commensal bacteria^33,76 but an elevation in human rhinovirus (HRV).^39,46 In contrast, the Moraxella genus tends to trigger exacerbations in an independent manner without viral effects.³⁹ In agreement with these results, a diminishment of commensal bacteria was reported in exacerbations, especially the genus Veillonella.^37,77 A study found that the decreased abundance of Veillonella and increased abundance of Staphylococcus were independent predictors of mortality in AECOPDs, elevating the risk by 13.5- and 7.3-fold, respectively.⁷⁸ As previously mentioned, the respiratory microbiome of patients with AECOPDs exhibits great heterogeneity, with different studies reporting diverse altered taxa (including Acinetobacter, Actinomyces, Ehrlichia, Fusobacteria, Perlucidibaca, Prevotella, Pseudomonas, and Sphingomonas).^26,34,76-79 This implies that the respiratory microbiome in AECOPD may undergo varying degrees of change and remodeling. Overall, increases in the abundance of Moraxella, Haemophilus, and HRV, as well as decreases in α-diversity and the abundance of Veillonella, are associated with COPD exacerbation frequency, severity, and mortality.

Relationship Between the Respiratory Microbiome and Host Immunity in COPD

Several studies have shown that dysbiosis in the respiratory microbiome is linked to pulmonary inflammation pathways. Overall, the diversity and community organization of the respiratory microbiome were strongly correlated with sputum chemokine (C-X-C motif) ligand (CXCL)8/IL-8.³³ Host gene expression analysis found that decreased microbial diversity was related to emphysematous destruction, remodeling of the bronchiolar and alveolar tissue, and infiltration of the tissue by CD4<⁺ T cells.⁸⁰ The genera related to pathogenic states, Haemophilus and Moraxella, have been shown to correlate with exacerbated lung inflammation in COPD. Specifically, Haemophilus is more commonly associated with stable COPD, whereas Moraxella is more commonly associated with AECOPD. Both Haemophilus and Moraxella were found to be related to the excessive production of several chemokines in sputum samples, including IL-1β, IL-8, IL-10, and TNF-α.^{33,42,44,50,59,72,81} The bacterial load of H. influenzae was found to be an independent predictor of sputum IL-1β and TNF-α in stable COPD, while the bacterial load of M. catarrhalis was correlated with elevated sputum IL-1β and TNF-α concentrations in COPD exacerbations.^41,72 In addition to proinflammatory mediators, the abundance of Moraxella and Haemophilus was also found to correlate with neutrophil counts in sputum^33,36,38 or blood³⁸ in both cross-sectional and longitudinal studies.

The type of inflammation is linked to particular microbial communities. Neutrophilic inflammation is associated with Proteobacteria dominance, particularly Haemophilus and Moraxella, which is correlated with host interferon and proinflammatory signaling pathways, Eosinophilic inflammation, on the other hand, may be related to a diverse microbial profile dominated by Firmicutes or Bacteroidetes.^41,56 Additionally, the dominance of Firmicutes, especially Streptococcus, has a positive relationship with peripheral eosinophil counts.⁵⁶ Another study found that peripheral eosinophil levels ≥2% were correlated with a higher diversity in the respiratory microbiome,⁶⁷ which can partially explain the negative correlation between Proteobacteria and eosinophil counts in blood,⁵⁶ as the colonization of Haemophilus has been linked to decreased microbial diversity in the respiratory microbiome.^33,76 Consistently, a study classified 3 clusters of COPD and asthma patients: cluster 1 was characterized by Proteobacteriadominance, increased proinflammatory mediators (IL-1β, IL-6, TNF-α, vascular endothelial growth factor, etc.), and increased neutrophil counts in both blood and sputum. Cluster 2 was characterized by Bacteroidetesdominance, increased type 2 mediators (IL-5, IL-13, chemokine C-C motif ligand [CCL]13, CCL17, and CCL26), and increased eosinophil counts. The features of cluster 3 were Actinobacteria andFirmicutes dominance and increased type 1 mediators (CXCL10, CXCL11, and interferon-gamma).⁸²

The clinical decision-making process is directly influenced by the type of inflammation present. For instance, genomic markers of type 2 inflammation are associated with elevated eosinophils (1 of the main effector cells of type 2 inflammation) and a favorable response to inhaled corticosteroids (ICS). As a result, eosinophil counts were the first blood biomarker incorporated into the Global initiative for chronic Obstructive Lung Disease (GOLD)⁸³ guidelines. This addition was based on evidence that eosinophil counts aid in determining the role of ICS treatment in patients prone to symptom exacerbation.¹ Therefore, the integration of microecological data will contribute to our understanding of the molecular mechanisms and characterization of various COPD subtypes. Alterations in chemokines and inflammatory cells found in sputum samples have the potential to act as biomarkers, enabling the identification of subsets of COPD with distinctive features in pathogen growth and inflammatory pathways.

It is plausible that activation of neutrophilic inflammatory pathways by Proteobacteria contributes to pathogenic phenotypes in COPD. Toll-like receptors (TLRs) have the ability to recognize and modify the inflammatory response of the host by detecting conserved microbial-associated molecular patterns.⁸⁴ The surface structures of gram-negative bacterial colonies, such as lipopolysaccharide (LPS) and polysaccharide A (PSA), can develop either beneficial or harmful effects. In terms of beneficial effects, it has been shown that PSA generated by colonies of gram-negative bacteria holds the capacity to rectify immune deficiencies in gnotobiotic mice and inhibit the progression of experimental colitis, although the underlying molecular processes that enable these effects require further research.⁸⁵ On the other hand, proinflammatory LPS interacts with the host myeloid-differentiation-2/Toll-like receptor 4 receptor complex, hence triggering a cascade of events that result in the activation of the transcription factor NF-κB, which then leads to the expression of genes responsible for producing proinflammatory cytokines and chemokines.⁸⁶ Similar proinflammatory effects can also be induced in a TLR3- or TLR4-dependent manner, which are associated with COPD aggravation⁸⁷ (Figure 2). Proinflammatory LPS challenge has been linked to bacterial exacerbations of COPD characterized by elevated pulmonary and systemic inflammation.⁸⁸ In an in vivo study in mice, H. influenzae caused severe COPD-like inflammation in a TLR2-independent manner, which was characterized by airway neutrophilia, neutrophilic cytokine/chemokine profile, and lung immunopathology.⁸⁹ Furthermore, enhanced expression of TLR4 and NOD-like receptors in the bronchial epithelium of severe COPD patients has been observed to correlate with exacerbated bronchial inflammation and an increased P. aeruginosa bacterial load. This relationship may potentially be involved in the fundamental pathogenesis of COPD.⁹⁰ These results are consistent with the aforementioned findings of increased abundance in gram-negative bacteria such as Haemophilus, Moraxella, and Pseudomonas in the respiratory microbiome of COPD patients. The link between lung pathology and microbiome alterations has also been observed in clinical studies. For instance, deviation of the respiratory microbiome was found to be related to severe subtypes determined by structural alterations (either airway or emphysema type changes) in CT scans.⁷⁰ Similarly, colonization of H. influenzae was found to be associated with bronchiectasis in COPD, and the degree of emphysema was correlated with the IL-8 concentration in sputum samples.⁴³

The diminishment of commensal bacteria suggests a loss of protective effects provided by these microorganisms, which may contribute to COPD. Contrary to the effect of Proteobacteria, several commensal bacteria were found to have anti-inflammatory effects. The most mentioned commensal genus that is altered in COPD, Prevotella, was found to be associated with mild neutrophilic airway inflammation in an in vivo study.⁸⁹ Another study discovered that protection against S. pneumoniae generated by Prevotella was achieved by identification of Prevotella lipoproteins through TLR2, which triggered TNF-α production and elevated neutrophils.⁹¹ Prevotella is thought to cause only a low to moderate level of inflammation, which could be the reason why it is tolerated by the respiratory immune system and develops protective effects. However, further research is needed to fully understand the intricate relationship between Prevotella and the immune system in the respiratory tract.^89,91 S. pneumoniae has a similar ability to trigger TLR2-related responses,⁷⁵ yet it is worth mentioning that the effect of TLR priming varies depending on the infection. For example, TLR4 priming with LPS derived from Escherichia coli enhances protection against lung infection caused by Klebsiella pneumoniae and influenza A virus. However, this is not the case with S. pneumoniae. To the best of our knowledge, proliferation of the genus was linked to reduced lung function.^73,74 The intrinsic mechanism of various effects induced by these bacterial colonies requires further investigation.

Relationship Between Clinical Therapy and the Respiratory Microbiome in COPD

Antibiotic treatment in cases of AECOPDs has demonstrated a beneficial impact, leading to a reduction in the abundance of Proteobacteria and facilitating the restoration of microbial diversity. A study by Wang et al highlighted that steroid treatment alone reduced respiratory microbial diversity and increased the Proteobacteria: Firmicutes ratio, whereas antibiotic treatment had the opposite effect in AECOPD patients.³³ Similarly, Huang et al observed a decrease in Proteobacteria within AECOPD patients treated with antibiotics alone.³⁵ However, the beneficial impacts are not necessarily sustained, especially when patients require prolonged antibiotic use for conditions like recurrent acute exacerbations. In patients with cystic fibrosis, antibiotic use has been associated with a decline in bacterial diversity in the respiratory microbiome over a decade, independent of age and lung function.⁹² In addition, different types of antibiotics, pathogenic bacterial strains, and bacterial loads can influence sequencing results, and the long-term impact of antibiotics on respiratory microecology of COPD patients with recurrent exacerbations remain unclear. Further accumulation of long-term clinical data is necessary.

Gut-Lung Axis in COPD

Evidence of Gut-Lung Crosstalk in COPD

Emerging evidence has indicated the possible interaction between the gut microbiome and respiratory microbiome. The respiratory microbiome is thought to be partly shaped through microaspiration from the oropharynx,^20,93 indicating that it can be affected by both the upper respiratory tract and gastrointestinal tract. While there is currently no evidence supporting direct transfer between the gut and respiratory microbiomes, the correlation between the gut and respiratory microbiome may be attributed to the potential transfer of metabolites between these sites. An analysis of cystic fibrosis in infancy found a strong correlation between the changes in bacterial composition over time in both the gut and respiratory tract and that microbial colonization of the respiratory tract can be anticipated by prior colonization of the gut.⁹⁴ In regard to COPD, the genus Streptococcus proliferated in both the lung and gut, indicating a transfer of microbiota through the oropharynx.¹⁴ In contrast, the genus Prevotella presented contradictory abundance and correlation with lung function depending on its location. The genus proliferated in the guts of COPD patients and healthy smokers^15,95 but was reduced in the lungs.^58-60,70 The varying microenvironments and metabolite transfer across the body may account for these differences. Examining the intrinsic mechanisms that allow the same microbiota to carry out various abundances and functions depending on the local microenvironment may aid in better understanding the gut-lung axis.

Recent studies are starting to reveal a potential link between the environments of the lungs and gut COPD. This suggests that these 2 organ systems might interact or influence each other’s health and disease states. In an in vivo study, mice that received transplantation of feces from COPD patients developed inflammation in their lungs.⁶ Another study revealed that the transplantation of normal fecal microbiota could alleviate COPD pathogenesis by reducing lung and intestinal inflammation, as well as restoring abnormal amino acid metabolism in sera.⁵ On the other hand, patients with COPD had considerably higher occurrences of inflammatory bowel disease, and the risk rose along with the severity of COPD.^96,97 Consistently, reduced integrity and function of the intestinal barrier were observed in COPD patients,^7,8 suggesting that chronic lung inflammation has a systemic effect on the gut ecosystem.

The Intrinsic Mechanisms of Gut-Lung Crosstalk in COPD

Colonization of potentially pathogenic microorganisms in the lungs plays an important role in the advancement and deterioration of COPD. As mentioned before, gram-negative bacterial colonies can trigger proinflammatory TLR signaling through surface structures, including LPS, which leads to elevated pulmonary and systemic inflammation.^86,88,89 Proinflammatory LPS may have further effects on intestinal homeostasis, as decreased fecal secretory immunoglobulin A levels and enlargement of Peyer’s patches were observed in cigarette smoke- and LPS-induced murine models.⁹⁸ This suggests that exacerbated systemic inflammation induced by the lung microbiota may have disrupted the balance of the intestinal environment, including its microecology.

The proper balance of microorganisms in the gut has a significant impact on regulating innate and acquired immunity. Harmful factors such as cigarette smoke, pollution particles, and an unbalanced diet cause dysbiosis of both the gastrointestinal and respiratory microbiome, depriving patients of the protective benefits of a healthy microecology.⁹⁹ Loss of these modifications may further aggravate systemic and lung inflammation, creating a vicious cycle. The modifications of host immunity by gut microbiota are achieved primarily through the secretion of SCFAs and the recognition of microbe-associated molecular patterns. (Figure 1) Evidence has shown that COPD patients experience a loss of protective effects from SCFAs produced by gut microbiota.^27,28 Since SCFAs are present in human lungs, the absence of the substrates needed for SCFA synthesis by fermenting bacteria indicates that gut microbiota-produced SCFAs can transfer into the lungs through systemic circulation.²⁸ The bacterial phylum that is less abundant in the gut microbiome of COPD patients, Bacteroidetes, was associated with higher levels of SCFAs,²⁹ which is consistent with lower levels of SCFAs in COPD.^6,100 SCFAs have been found to develop anti-inflammatory properties by modifying the Th2 response, modulating pathways of pulmonary ILC2s, and inhibiting M2 macrophage polarization in airway hyperreactivity and airway inflammation models.²⁸ In regard to recognition of microbial surface structures, the capsular polysaccharide component PSA produced by members from the genus Bacteroides in the gut has developed anti-inflammatory properties and the ability to correct immune defects.⁸⁵ Thus, the decline in Bacteroides in stable COPD may represent a loss of PSA-mediated protective effects. An increased ratio of Firmicutes/Bacteroidetes, which was observed in stable COPD, has previously been associated with elevated lung IL-17 and IL-22 responses.¹⁰¹ Further research is necessary to determine whether the above anti-inflammatory responses are compromised in COPD patients.

Conclusions and Perspectives

In conclusion, specific patterns in the gut and respiratory microbiome were discovered in both stable COPD and COPD exacerbations. The commensal microorganisms associated with protective effects were diminished, whereas potentially pathogenic microorganisms proliferated, leading to activated systemic and pulmonary inflammation. As a consequence, deteriorated pulmonary function, enhanced severity, increased onset of exacerbations, and elevated mortality were observed. Further investigations on the relationship between the gut microbiome and clinical manifestations of COPD, the intrinsic mechanisms of protective or harmful effects induced by certain genera, and the transformation from current knowledge on gut and respiratory microbiome dysbiosis to therapeutic strategies are warranted.

Acknowledgements

Author contributions: ZXC wrote the main manuscript text and prepared Figures 1 and 2. JZ reviewed the manuscript. Both authors reviewed and approved the final version of the manuscript for publication.

Declaration of Interests

The authors report no conflicts of interest in this work.