The gut microbiota established a close symbiotic relationship with our body that plays a critical role in health maintenance, promoting metabolism of dietary components, vitamin synthesis, preventing pathogen colonization, and regulating immune response (Jandhyala et al., 2015). The gut microbiota is a complex community of microorganisms inhabiting the gastrointestinal tract that in a health individual comprises 150 distinct bacterial species, mainly Bacteroidetes, Firmicutes, Proteobacteria, Actinobacteria, and Verrucomicrobia (Qin et al., 2010). Besides, fungi have been identified as an integral part of commensal flora, and the diversity of the gut microbiota in healthy subjects is limited to few genera, with a high prevalence of Saccharomyces cerevisiae, Malassezia restricta, and Candida albicans (Nash et al., 2017). Fungi are 100 to 1,000 times larger than bacteria, but less frequent than bacteria sequences. Fungi may compensate the bacterial functions of immune modulation and protect mucosal tissue damage at the physiological state and during gut microbiota disturbances (Jiang et al., 2017). In addition, bacteria can affect the function of fungi (Tso et al., 2018).

While the lung and gut originate from the same embryonic organ in the primitive foregut, the microbiota of lung and gut differ significantly in taxonomic composition, diversity, and function (Girosi et al., 2006). The lung microbiota is composed of transient microorganisms mainly derived from upper respiratory tract (URT) including nostrils, nasal passages, paranasal sinuses, nasopharynx, and oropharynx. While Bacteroidetes and Firmicutes are the most abundant bacterial phyla in both microbiotas, the lung and gut microbiota are very different at the species level. In the lungs, the genera Streptococcus spp., Veillonella spp., and Prevotella were abundant, whereas Bacteroides, Faecalibacterium, and Bifidobacterium are more prevalent in the gut (Sender et al., 2016). In a disease or dysbiotic state, other organisms such as viruses or fungi are present in the lung (Limon et al., 2017).

Lung microbiota plays key roles in the development and progression of COPD. The disease is associated with dysbiosis of the lung microbiota with the overgrowth of pathogenic bacteria, resulting in accelerated decline in lung function. Studies on the lung microbiota of COPD showed reduced lung microbial community diversity compared with those of healthy individuals, with an expansion of Hemophilus spp. Afipia, Brevundimonas, Curvibacter, Moraxella, Neisseria, and Undibacterium spp. of the Proteobacteria phylum, Corynebacterium spp. of the Actinobacteria phylum, Capnocytophaga spp. of the Bacteroidetes phylum, and Leptolyngbya spp. of the Cyanobacteria phylum (Erb-Downward et al., 2011; Garcia-Nunez et al., 2014; Huang et al., 2014; Millares et al., 2014; Wang et al., 2016). Furthermore, the severity of COPD is associated with less diverse lung microbiota and an expansion of pathogenic microbes (Garcia-Nunez et al., 2014; Wang et al., 2016; Mayhew et al., 2018; Wang Z. et al., 2018). The culture of bronchial secretion from the lower airway of stable COPD patients showed bacteria such as H. influenza, S. pneumoniae, and Moraxella catarrhalis. In contrast, the exacerbated COPD patients present an unstable lung microbiota with an increased relative abundance of Moraxella spp. of the Proteobacteria, but a decreased abundance of the Firmicutes (Rosell et al., 2005; Wang et al., 2016; Sun et al., 2020). Notably, increased Proteobacteria: Firmicutes ratio suggested that the cause of exacerbation is due to bacterial dysbiosis, whereas eosinophilic exacerbation exhibits a decrease in Proteobacteria: Firmicutes ratio (Wang et al., 2016). Disturbance in core lung microbiota during acute exacerbation of COPD (AECOPD) leads to the outgrowth of pathogen including Acinetobacter spp. and Klebsiella spp., indicating that commensal lung microbiota protects the lung from colonization of pathogenic microbes (Sun et al., 2020). Reduced microbial diversity in COPD showed negative correlation with CXCL8, indicating that reduced species diversity might be associated with clearance by neutrophil extracellular traps through invading host epithelial cells. Furthermore, mannose binding lectin (MBL) was shown to interferes with phagocytosis of H. influenza. Study by Dicker et al. (2018) showed that COPD patients with MBL deficiency have a more diverse lung microbiota and a lower risk of exacerbation. Additionally, viral infections such as rhinoviruses and influenza virus are detected in around 10%–15% of sputum sample from stable COPD patients in comparison with 30%–60% in severe COPD patients (Mallia et al., 2011; Utokaparch et al., 2014; Tan et al., 2016). Moreover, the mycobiome in the lung microbiota showed an increased relative abundance of Candida spp., Aspergillus, Candid, Phialosimplex, Penicillium, Cladosporium, and Eutypella in the lung of COPD patients (Charlson et al., 2012; Su et al., 2015). Thus, bacterial, viral, and fungal infections may lead to exacerbated airway and systemic inflammation (Sethi and Murphy, 2008).

The research on gut-lung axis in respiratory diseases such as asthma, COPD, cystic fibrosis, and respiratory infections suggested that these diseases might be prevented, or at least ameliorated by regulating gut microbial ecosystem through manipulation of gut microbiota (Zhang D. et al., 2020). The intestinal and respiratory microbiota may affect each other by changing mucosal immunity. On one hand, mice lacking an intestinal microbiota through antibiotic treatment are more susceptible to pneumonia and the alveolar macrophages showed an altered transcriptome which results in decreased phagocytic activity and bacterial clearance (Chen et al., 2011; Schuijt et al., 2016). On the other hand, bacterial and viral respiratory infections in mice lungs lead to gut dysbiosis (Sze et al., 2014; Samuelson et al., 2016). The murine model of sepsis suggested that dysbiosis in gut microbiota influenced composition of the respiratory microbiota through changes in circulating inflammatory cytokines and translocation of intestinal microbiota to the airway (Gauguet et al., 2015).

Clinical data have been available for COPD. The first study that showed increased gastrointestinal (GI) permeability in AECOPD (Sprooten et al., 2018) suggested the involvement of the gut microbiota in exacerbation of COPD. Mucosal epithelial integrity is critical for preventing the entry of harmful particles (e.g., bacteria and their product) into the systemic circulation. Disturbed intestinal integrity may contribute to increased permeability of the intestinal mucosal barrier, resulting in bacterial translocation and systemic inflammation (Turner, 2009). The changes in gut microbiota for COPD have also been observed. For instance, by using untargeted fecal metagenomics and metabolomics, a recent study reported that composition of intestinal microbiota and metabolome in COPD patient was significantly different from those of healthy controls, with an increased abundance of several Streptococcus parasanguinis_B in COPD (Bowerman et al., 2020). In an earlier study of AECOPD, it identified increased abundance of both Streptococcus parasanguinis_B and Streptococcus salivarius in fecal microbiome, while only Streptococcus parasanguinis_B was observed increased in sputum (Chan et al., 2015). Furthermore, in a study comparing both gut and lung microbiota in samples collected from 15 AECOPD patients over a 14-day period of antibiotic and steroid treatment, it only observed difference in abundance of certain phyla but not in diversity of core flora between these two sample types (Sun et al., 2020). These studies together suggested a potential cross-talk in the gut-lung axis that may play a critical role in the pathophysiology of COPD, but the mechanisms how the axis regulates microbiota remain unknown.

Cigarette smoking is the most significant risk factor for COPD (Rennard and Vestbo, 2008). Although 15%–20% smokers develop COPD in the United States, approximately 80% COPD patients being past or current smokers (Warren et al., 2014). Smoking is also associated with high mortality rate of COPD (Warren et al., 2014). Although the bacterial communities in bronchoalveolar lavage fluid (BALF) from smokers with COPD were similar to those from healthy smokers and non-smokers, the lung tissues of smokers with COPD showed the heterogeneity and diversity in the lung microbiota at different regions of abnormal lung (Erb-Downward et al., 2011), suggesting that long-term smoking may contribute to the development and exacerbation of COPD. In addition, tobacco smoke has been previously associated with infections of pathogenic microorganisms in the lungs, as well as persistent colonization of opportunistic pathogens in lower respiratory tract, which all in turn lead to exacerbation and progression of COPD (Bagaitkar et al., 2008; Garmendia et al., 2012). Pathophysiologically, this may be established by modification of ciliary movements, damage of tight junctions that increases epithelial permeability, and elevation of neutrophils, macrophages, and T cells along the airways of COPD patients (Garmendia et al., 2012). Notably, exposure to smoke may also modify relative abundance of microbiota of the lungs, as indicated by animal studies (Zhang et al., 2018). Cigarette smoking has been shown to reduce the diversity of microbes in the gut (Faith et al., 2013). However, such smoking-related diversity of gut microbiota has not been compared between COPD patients and healthy smokers.

Malnutrition is an important but overlooked problem in COPD (Mete et al., 2018). Around 20% COPD patients have weight loss and malnutrition in protein and calorie, which have negative impact on mortality (Hillas et al., 2015). A 5-year follow-up study showed that individuals that consume less fiber, vitamin, and folic acid are prone to develop airflow limitation and COPD (Jung et al., 2021). The gut microbiome produces a large number of metabolites, e.g., short-chain fatty acids (SCFAs), through fermentation of fiber. SCFAs are potent anti-inflammatory molecules which reduced chemotaxis and adherence in immune cells while increasing the release of anti-inflammatory cytokines and inducing apoptosis (Ratajczak et al., 2019). SCFAs are absorbed into the systemic circulation and can influence lung health. Trompette et al. (2014) showed that dietary fiber leads to altered composition of gut and lung microbiota, and mice with prolonged high-fiber feeding are protected against the allergic inflammation in the lung. A lack of fermentable fibers can lead to malnourishment of the microbiota, resulting in gut dysbiosis and affect lung physiology (Makki et al., 2018; Vaughan et al., 2019).

Other risk factors such as chronic asthma, low birth weight, childhood respiratory tract infections, pulmonary tuberculosis, and occupational exposures to dusts have also been reported to contribute to the pathogenesis of COPD (Barnes, 2016). The metagenomic analysis of fecal samples from patient with asthma demonstrated a decreased diversity in the gut microbiota compared to a healthy cohort (Wang Q. et al., 2018). In addition, the antibiotics and steroid treatments for COPD seem to have differential effects on lung and gut microbiota (Sun et al., 2020). After treatments with antibiotics and steroid, the abundance and diversity of sputum bacterial flora decreased initially and returned to normal on day 14, whereas the abundance of gut bacteria increased throughout the 14 days post-treatment without affecting diversity of gut microbiota. Of note, these COPD patients have a history of frequent antibiotic use; therefore, the prolonged antibiotic pressure has caused a long-term disturbance in the microbiota with a shift toward bacterial community resistant to antibiotics.

The nucleotide-binding oligomerization domain (NOD)-like receptor (NLR) family pyrin domain-containing protein (NLRP3) inflammasome is generally activated by a range of stimuli and acts as a critical regulator in the innate immunity (Kelley et al., 2019). The NLRP3 inflammasome promotes the recruitment of inflammatory cells and regulates immune response in the gastrointestinal tract and the lung (Donovan et al., 2020). Most bacterial infections in the gut could be cleared via gut immunity response. However, for those that escape from the clearance, some has been reported to transfer to circulation and travel to the lung, which results in activation of inflammation. The gut and lung microbiota can influence the function of NLRP3 inflammasome (Donovan et al., 2020). In addition, inflammasomes may also be activated by interactions between bacteria and pattern recognition receptors (PRRs) with alterations of inflammasome responses presented in different severities of COPD (Donovan et al., 2020). Activation of NLRP3 inflammasome mediates caspase-1 activation and the secretion of proinflammatory cytokines such as IL-1b/IL-18. This process was shown to link to the development of airway inflammation in COPD (Wang H. et al., 2018). In peripheral blood mononuclear cells (PBMCs) and bronchial tissues, the mRNA levels of NLRP3, Caspase-1, apoptosis-associated spec-like protein containing CARD (ASC), IL-18, and IL-1b were elevated in AECOPD comparing with smokers; however, the mRNA levels of these gene were decreased in COPD patients in stable stage (Wang H. et al., 2018), suggesting that NLRP3 inflammasome is associated with AECOPD.

Some RCTs have shown that targeting inflammasome-related effectors (i.e., IL-1a and IL-1b) failed to benefit COPD patients at moderate to severe stages. For instance, a 45-week monthly treatment on COPD patients via intravenously infusing canakinumab, a human anti-IL-1b monoclonal antibody, did not show statistical difference in forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC) between treatment and placebo groups (Rogliani et al., 2015). Furthermore, MEDI8968, a human IgG2 monoclonal antibody against IL-1 receptor 1, did not show any beneficial effects on lung function or quality of life of AECOPD patients within 52 weeks of treatment (Calverley et al., 2017). Similar results were obtained for MEDI2338, a human IgG1 monoclonal antibody that binds IL-18, and AZD9056, a P2X7 antagonist that binds human P2X7 receptor. However, some of these RCTs failed to assess the impact of these drugs on disease pathogenesis due to the lack of adequate endpoints; thus, further larger-scaled well-designed studies are required to draw a conclusion for these inflammasome-related effectors.

Immunoproteasome is a specialized type of proteasome required for intracellular protein degradation in immune cells. It also acts as a key regulator for immune cell differentiation, inflammatory activation, and autoimmunity. A recent study demonstrated that proteasome expression was upregulated in PBMCs isolated from severe COPD patients as compared to that from matching controls (Kammerl et al., 2021). The distinct activation of proteasome was observed in PBMCs of young male smokers and severe COPD individuals. The same study also showed that inhibiting immunoproteasome reduced proinflammatory cytokine expression in COPD-derived immune cells (Kammerl et al., 2021), suggesting that targeting immunoproteasome may be a potential treatment for severe COPD.

Epigenetic modifications have also been identified to participate in altering composition of gut microbiota and in the development of COPD (Zhang L. et al., 2020). However, the exact role of these epigenetic modifications in regulating gut and lung microbiota of COPD remains largely unclear. Firstly, DNA methylation is widely used in the study of epigenetic dysregulation in smokers with or without COPD. DNA methyltransferases (DNMTs) are enzymes catalyzing the addition of a methyl (CH₃) group to the DNA strands to regulate gene expression. DNMTs are highly sensitive to the availability of nutrients, and they are affected by metabolites produced by gut microbial species (Ansari et al., 2020). DNA methylation is associated with gene expression profile in COPD lung tissues (Lee et al., 2017), however, whether such an epigenetic modification alters COPD-related gut and lung microbiota remains to be further elucidated. Secondly, COPD patients have a reduced level of histone deacetylase 2 (HDAC2) compared with non-smokers (Barnes, 2009), which may results in expansion of inflammation. Other HDACs such as HDAC5, HDAC8, and SIRT1 are all decreased in COPD (Ito et al., 2005; Rajendrasozhan et al., 2008). It is known that gut microbiome can modulate the activity of HDACs by producing SCFAs (Yuille et al., 2018). Therefore, targeting HDAC complex may be a potential therapy for COPD (Zwinderman et al., 2019). Thirdly, increased evidence have shown that gut microbiome plays a key role in regulating microRNAs (miRNAs; Malmuthuge and Guan, 2021), whereas miRNAs are capable of suppressing or preventing COPD development (Zhang L. et al., 2020). Based on profiling analysis, approximately 140 miRNAs are dysregulated in COPD compared to healthy controls, whereas around 70 miRNAs are differentially expressed in COPD compared to smokers (Leidinger et al., 2011; Osei et al., 2015). Clinical data have shown that these miRNA targets were enriched in genes associated with cytokine production (Faiz et al., 2019; Hobbs and Tantisira, 2019). However, the profiling of miRNA has not been compared between lung and gut in COPD. Future studies are required to understand the immune-modulatory role of miRNA in the gut and lung axis.

Epigenetic modifications can be regulated by metabolites of gut microbiota including SCFAs, folates, biotin, and trimethylamine-N-oxide (TMAO). Increased circulating TMAO level was associated with increased mortality of COPD patients, which was independent of disease severity (Ottiger et al., 2018). However, the role of epigenetic modifications on the mechanisms of regulating TMAO levels is still unclear. There are several therapeutic strategies developed to lower TMAO levels to treat atherosclerosis, but with potential side effects (Yang et al., 2019). Whether these therapeutic strategies targeting TMAO levels can be used for COPD should be addressed in future studies. Despite TMAO, the study by Bowerman et al. (2020) also identified a list of metabolites from lipid, amino acids, or xenobiotic classes that are associated with COPD, suggesting these metabolites can be used as biomarkers for COPD diagnosis in the future studies.

Diet is a readily modifiable factor for COPD patients. A direct association between Western diet and increased risk for COPD has been shown in both genders (Varraso et al., 2015; Mekary, 2016; Young and Hopkins, 2018), indicating that diet is strongly associated with the risk of COPD. For consumptions of specific type of food, several studies have shown that total and cereal dietary fiber are inversely associated with COPD incidence, whereas results of fruits and vegetables are inconsistent (Varraso et al., 2010; Kaluza et al., 2018; Szmidt et al., 2020). Therefore, nutrition-based treatment, particularly high dietary fiber intake, may be a potential non-pharmaceutical approach for managing or preventing COPD.

A recent study demonstrated that dietary supplementation of probiotics Lactobacillus rhamnosus and Bifidobacteriumbre breve in COPD mice prevented the airway inflammation and lung damage (Aimbire et al., 2019). The probiotic treatment in these mice attenuated cellularity in BALF and restored the balance between cytokines and chemokines, resulting in a reduction of alveolar enlargement and collagen deposition. Hence, the probiotic treatment yielded positive outcomes in COPD mice, suggesting its potential therapeutic value for the disease, which needs to be testified in large-scaled RCTs for clinical use.

Fecal microbial transplantation in humans has shown systemic effects (Li et al., 2014). Dysbiosis in the gut can be modified through fecal transplantation in the cigarette-smoking-based model, indicating that gut microbiota may play a causal role on COPD pathogenesis (Lai et al., 2021). In particular, Parabacteriodies goldsteinii (P. goldsteinii) was shown to reduce intestinal inflammation, enhance mitochondrial functions, ribosomal activities in colon, restore amino acids metabolism in sera, and inhibit lung infection, at least partially through inhibiting toll-like receptor 4 (TLR4) signaling pathway via the lipopolysaccharide derived from P. goldsteinii (Lai et al., 2021). Thus, this bacterial strain or its derived lipopolysaccharide could be used as a new therapeutical agent for COPD. At present, no fecal transplantation studies have been conducted in COPD patients, hence, whether this approach is effective for COPD awaits further clinical investigations.