Article Text
PDF
Abstract
Background Diet has a crucial role in the gut microbiota, and dysbiosis in the gut and lungs has been suggested to be associated with chronic obstructive pulmonary disease. We compared the diet, microbiome and metabolome between asymptomatic smokers and those with emphysema.
Methods We enrolled 10 asymptomatic smokers with preserved lung function and 16 smokers with emphysema with severe airflow limitation. Dietary intake information was gathered by a self-reported questionnaire. Sputum and faecal samples were collected for microbial and metabolomics analysis. A murine model of emphysema was used to determine the effect of metabolite supplementation.
Results Despite having a similar smoking history with emphysema patients, asymptomatic smokers had higher values of body mass index, fibre intake and faecal acetate level. Linear discriminant analysis identified 17 microbial taxonomic members that were relatively enriched in the faeces of asymptomatic smokers. Analysis of similarity results showed dissimilarity between the two groups (r=0.287, p=0.003). Higher acetate level was positively associated with forced expiratory volume in one second in the emphysema group (r=0.628, p=0.012). Asymptomatic smokers had a greater number of species associated with acetate and propionate (r>0.6) than did those with emphysema (30 vs 19). In an emphysema mouse model, supplementation of acetate and propionate reduced alveolar destruction and the production of proinflammatory cytokines, and propionate decreased the CD3+CD4+IL-17+ T-cell population in the lung and spleen.
Conclusion Smokers with emphysema showed differences in diet, microbiome and short-chain fatty acids compared with asymptomatic smokers. Acetate and propionate showed therapeutic effects in a smoking-induced murine model of emphysema.
- emphysema
- COPD pathology
Data availability statement
The data that support the findings of this study are available from the corresponding author (SWL), upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
While smoking is a major risk factor for chronic obstructive pulmonary disease (COPD), a significant proportion of smokers do not develop COPD. Dysbiosis of the gut microbiota can serve as the trigger for several immune disorders, which may be treated through microbial modulation.
WHAT THIS STUDY ADDS
The correlations among fibre intake, acetate, propionate and short-chain fatty acid-producing bacteria were more remarkable in asymptomatic smokers than in smokers with emphysema. Supplementation of acetate and propionate in a mouse model of emphysema attenuated the inflammatory response and the development of emphysema.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
The gut microbiome and metabolome appear to be associated with the pathogenesis of COPD. Further large-scale studies are warranted to determine the therapeutic potential of acetate and propionate in COPD.
Introduction
Chronic obstructive pulmonary disease (COPD) is a chronic progressive disease that significantly contributes to worldwide morbidity and mortality.1 Cigarette smoking is the major risk factor for the development and progression of COPD. However, a considerable proportion of smokers do not develop clinically significant airflow limitations,2 3 and a noticeable proportion of patients with COPD are never-smokers. COPD also shows variations in its progression,4 5 with patients with more symptoms, low body mass index (BMI) or frequent exacerbation showing more dismal outcomes.6–8 Smoking cessation can slow the progression of COPD,9 but not stop it completely. Genetic predisposition,10 exercise,11 environment12 and lung development13 can each partially explain such heterogeneity and variability in the development and progression of COPD.
The interactions between the host and microbes are critical for maintaining immune homeostasis. Imbalances in the gut microbiota, or dysbiosis, can serve as the trigger for several immune disorders.14 Accordingly, modulation of the microbiota has already been applied in the treatment of diseases, which are mostly related to infections15 and gastrointestinal conditions.16 Recent studies showed that gut microbes and the lung have an intimate interaction, and that microbial dysbiosis in the gut and the lung are related to chronic respiratory diseases.17–19 The therapeutic potential of microbial modulation in respiratory diseases is not yet evident, but accumulating evidence supports its possible benefits in conditions such as asthma.20 For COPD, epidemiological studies have supported the role of gut health,21 and Jang et al reported the therapeutic potential of gut microbiota modulation in emphysema.22
Despite the recent advances in the management of COPD, such interventions have been shown to be insufficient to reduce the rapid decline in lung function combined with uncontrolled inflammation, which is mainly due to the limited understanding of the pathogenesis of COPD.1 Inflammatory responses to noxious stimuli such as smoking, one of the major contributors to COPD, can differ according to the patients. In this study, we hypothesised that there is a significant interindividual difference in the diet (prebiotics), gut microbiota (probiotics) and its metabolites (postbiotics) between asymptomatic smokers and patients with severe COPD (emphysema). To test this hypothesis, we compared the composition of the diet, gut microbiome and metabolome in asymptomatic smokers and those with severe emphysema. Based on the information gathered from clinical samples, we also conducted an animal experiment to test the possibility of the therapeutic application of gut metabolites in the treatment of COPD.
Methods
Study populations
Men older than 18 years with a smoking history of ≥30 pack-years and available results of pulmonary function test at a stable state within 3 months were eligible to participate in this study. Among them, we enrolled those with severe emphysema and asymptomatic controls between September and November 2019. Severe emphysema was defined as the presence of severe airflow limitation of postbronchodilator forced expiratory volume in one second (FEV1)/forced vital capacity (FVC)<0.5 and FEV1<50% of predicted value, and definite evidence of emphysema in more than two lobes on CT scan of the chest at the time of the diagnosis. The control group was defined as asymptomatic individuals with a relatively preserved lung function (ie, postbronchodilator FEV1/FVC≥0.6 and FEV1≥70% of predicted value). We excluded individuals with a history of antibiotic exposure within 2 months and underlying diseases such as malignancy, congestive heart failure or diabetes mellitus on insulin therapy.
DNA extraction and bacterial metagenomic analysis
DNA was extracted from the sputum and faecal samples using a commercial DNA isolation kit (MO BIO, Carlsbad, California, USA) according to the manufacturer’s instructions. Further analysis methods were described in the online supplemental file.
Supplemental material
Quantitative metabolome measurement
Faeces samples (10–20 mg) were freeze-dried for 24 hours using a benchtop manifold freeze drier and stored at −80°C until analysis. The samples were prepared using a previously described procedure23 and analysed using an LC-MS/MS (liquid chromatography with tandem mass spectrometry) system equipped with a 1290 HPLC (high-performance liquid chromatography, Agilent, Waldbronn, Germany) and QTRAP5500 mass spectrometry (AB Sciex, Toronto, Canada). The detailed procedures are described in the online supplemental file.
Murine model of emphysema
C57BL/6 mice were purchased from Orient Bio (Seongnam, Korea). Experimental scheme was designed to examine the effect of sodium acetate (NaA) or sodium propionate (NaP) on smoking-exposed emphysema, and detailed animal procedure and analysis were described in the online supplemental file.
Statistical analysis
The normality of the variables was tested by the Shapiro-Wilk normality test. Normally distributed continuous variables are presented as mean±SD and their differences between two groups were analysed by Student’s t-test. Other continuous variables, which were non-parametrically distributed are expressed as median values with IQR, were analysed with the use of the Mann-Whitney test for comparisons. Clinical, dietary and metabolic factors associated with emphysema were analysed by univariate logistic regression, then statistically significant factors (p<0.1) were adopted for multivariate analysis. A partial correlation was used to account for the confounding effect of BMI, as it was the main covariate. Basic analyses were performed by SPSS V.24 (IBM, Armonk, New York, USA) and visualised using GraphPad Prism V.7.00 (GraphPad Software, La Jolla, California, USA). Complex analyses for specific clads of the microbiome and their associations with short-chain fatty acids (SCFAs) were performed with R software, V.3.6.2 (R Foundation for Statistical Computing, Vienna, Austria). We also performed multiple linear regression after adjustment for BMI. The results from analysis of similarity (ANOSIM) based on Bray-Curtis matrix between two groups were visualised principal coordinate analysis (PCoA) graph. The results of the animal experiments were analysed using one-way analysis of variance (ANOVA) and are expressed as mean±SE of the mean. Dunnett’s test was used for multiple comparisons. Differences with p<0.05 were regarded as statistically significant.
Results
Participants
A total of 26 participants (10 asymptomatic controls (online supplemental table S1) and 16 emphysema patients) were included in the study, and their baseline characteristics are described in table 1. Age, height and smoking history in pack-years were not significantly different between the two groups; however, the BMI was significantly lower in the emphysema group (26.78 vs 19.48, p<0.001). Despite having a similar smoking history to the control group, the emphysema group had a significantly lower FEV1 per predicted value (86.80%±7.18% vs 26.38%±9.44%, p<0.001). Of the emphysema group, 10 (62.5%) had experienced severe exacerbation in the previous year and 5 (31.3%) required long-term oxygen therapy.
Baseline characteristics of the study population
Diet intake
Table 2 shows the nutritional profile of the study population. Most nutritional intakes were not significantly different between the two groups. Fibre intake was lower in the emphysema group, although without statistical significance (36.96±13.51 g vs 27.79±10.83 g, p=0.069); moreover, the proportion of participants whose fibre intake was less than the recommended daily amount was lower in the asymptomatic controls than in the emphysema patients (1/10 (10%) vs 6/16 (37.5%), p=0.13).
Dietary intake of asymptomatic smokers and emphysema patients
Microbiome analysis
A total of 52 sputum and faecal samples (26 each) were sequenced; as a result, 443 species and 502 species were identified in the sputum samples (online supplemental figure S1A) and the faecal samples (online supplemental figure S2A), respectively. In the sputum samples, the most dominant phyla were Firmicutes, and Bacteroidetes and Actinobacteria were also common in the asymptomatic smoker group; in contrast, the phylum Firmicutes and Proteobacteria were dominant in the emphysema group (online supplemental figure S1B). The two groups did not show significant differences in the diversity indices such as the Gini-Simpson Index, the Shannon Index, Chao1 and Amplicon Sequence Variants (online supplemental figures S1C and online supplemental table S2). In the faecal samples, the phylum Firmicutes and Bacteroidetes were major phyla in both groups and there was no statistical difference in the diversity indices as well (online supplemental figure S2 and online supplemental table S3).
The linear discriminant analysis (LDA) effect size method identified a total of 20 bacterial groups in the sputum and 35 groups in the faeces that were differentially abundant (ie, LDA Score of >2.0) between the two groups are shown in figure 1A (sputum) and figure 1B (faeces). Notably, at the bacterial family level, the sputum samples of emphysema patients showed significant enrichment of Moraxellaceae and Eubacteriaceae that belong to phylum Proteobacteria and Firmicutes, respectively; in contrast, the sputum samples of the controls showed significant enrichment of species Lancefieldella parvula and Bifidobacterium longum that belong to the phylum Actinobacteria. The faecal samples of asymptomatic controls showed distict 17 bacterial groups including Prevotella copri, Dialister succinatiphilus and Catenibacterium mitsuokai. These results are also shown as cladograms (online supplemental figure S3). PCoA results showed significant dissimilarity between the two groups in both sputum (online supplemental figure S4A) and faecal samples (figure 1C). The results of microbiome analysis according to smoking status showed that faecal microbial differences were more strongly associated with disease status than with smoking status (ie, current vs ex-smoker) (online supplemental figures S5 and S6).
Results of microbial analysis. (A,B) Linear discriminant analysis (LDA) effect size method for microbiome analysis. Histogram of the LDA scores of the microbial taxa in the sputum (A) and faecal samples (B) that are differentially abundant between asymptomatic controls and emphysema patients. LDA scores represent bacterial groups overabundant in each group. (C,D) Results of clustering at the species level in the faecal samples via principal coordinate analysis using analysis of similarity based on the Bray-Curtis matrix. The test statistic R and p value are also presented. Unadjusted result (C) and adjusted result (D) for body mass index. Each dot represents the bacterial microbiome of individual participants, and the ellipses indicate the 95% CIs. Arrows describe taxa with a correlation coefficient greater than 0.1 (r2>0.1). PC, principal coordinate.
Faecal metabolome analysis
Previous studies described the beneficial role of fibre in generating SCFAs after fermentation by the gut microbiota.24 25 Therefore, we analysed the faecal metabolites in 26 patients by focusing on the levels of SCFAs. The proportion of the major SCFAs—acetate, propionate and butyrate—was 64.9% in asymptomatic controls and 40.4% in the emphysema group (figure 2A). Specifically, the concentration of acetate was significantly higher in asymptomatic smokers than in emphysema patients (1.78±0.19 vs 1.17±0.19 peak area/dried faeces (mg), p=0.035, figure 2B). The histogram by fibre intake quartiles suggests the presence of a positive correlation between dietary fibre intake and the concentrations of acetate and propionate (figure 3A). The metabolomic profiles of amino acids and phospholipids did not show significant differences between the two groups except for PE aa C32:0 and PE aa C32:1 (online supplemental figure S7).
Short-chain fatty acids (SCFAs) in faecal samples of asymptomatic controls and emphysema patients. (A) Relative proportions of SCFAs in the two groups. (B) Histogram of the concentrations of SCFAs. Area ratio, analytic peak area/internal stand peak area. *P<0.05.
Correlation analysis between fibre intake and short-chain fatty acids (SCFAs). (A) Concentrations of SCFAs according to the fibre intake quartiles. (B) Scatter plots with partial correlation coefficients(r) adjusting for body mass index and the line of best fit (95% CI) between fibre intake and representative SCFAs.
Associations between metabolome, microbiome and clinical parameters
Participants with emphysema showed significant differences in clinical parameters, dietary intake and faecal metabolites compared with asymptomatic smokers. Among the variable parameters, possible risk factors associated with emphysema were identified in univariate analysis (online supplemental figure S8A). BMI, fibre intake and faecal acetate level were notably associated with emphysema in the univariate model (p<0.1) and were adopted for the multivariate analysis (online supplemental figure S8B). BMI was significantly associated with a reduced risk of emphysema (OR, 0.52; 95% CI, 0.31 to 0.89; p<0.05). Additional ANOSIM and PCoA results adjusted for BMI revealed a modest but statistically significant dissimilarity between asymptomatic controls and those with emphysema in faecal samples (r=0.287, p=0.003) (figure 1D).
Partial correlation analysis showed a positive association between fibre intake and faecal levels of acetate, which was evident only in the control group with a partial correlation coefficient of 0.764 (p=0.017). Propionate showed a modest correlation in the control group, although without statistical significance (r=0.524, p=0.148). Butyrate did not show a significant correlation with fibre intake in both groups (figure 3B). The correlation analysis of SCFA with lung function showed acetate level was positively associated with FEV1 (r=0.592, p=0.002). Interestingly, this positive association was greater in the emphysema group (r=0.628, p=0.012) (figure 4A). The acetate level showed a correlation with FVC as well (r=0.486, p=0.014) (figure 4B). The propionate level had a marginal correlation with FVC in the emphysema group (r=0.488, p=0.065, figure 4D).
Correlation analysis of lung function, acetate and propionate. (A,B) Associations between acetate and FEV1 and FVC. (C,D) Associations between propionate and FEV1 and FVC. Area ratio, analytic peak area/internal stand peak area; FEV1, forced expiratory volume in one second; FVC, forced vital capacity. Partial correlation coefficient (r) was derived after adjusting for body mass index.
In faecal samples from asymptomatic controls, fibre intake was positively associated (r>0.6) with 16 microbial species such as Prevotella stercorea, Duncaniella freteri and C. mitsuokai; in contrast, emphysema patients did not show significant associations between fibre intake and the abundance of specific microbial species (figure 5). Under the same criteria (r>0.6), acetate and propionate levels were positively associated with 30 species in asymptomatic smokers, which included SCFA-producing families such as Lachnospiraceae (Anaerobutyricum hallii, Blautia obeum, Blautia wexlerae, Dorea longicatena, Faecalicatena orotica, Ruminococcus gnavus, Merdimonas faecis), Oscillospiraceae (Fournierella massiliensis, Ruminococcus callidus) and Lactobacillaceae (Leuconostoc gelidum, Lactobacillus nagelii) (figure 5A). Meanwhile, in faecal samples from emphysema patients, acetate and propionate levels were associated with a reduced number of species (n=19) including the families Lactobacillaceae (Lactiplantibacillus plantarum, Lactobacillus delbrueckii, Weissella koreensis) and Lachnospiraceae (Blautia coccoides, Blautia faecicola, Blautia hydrogenotrophica, Blautia luti, Blautia schinkii) (figure 5B). A negative association of microbial species was found mainly with butyrate in asymptomatic smokers (online supplemental figure S9), and six species showed a positive association with butyrate in emphysema.
Correlation analysis of fibre intake, metabolites and microbiota in the faeces. (A,B) Relationship between the gut microbiome, fibre intake and short-chain fatty acids (SCFAs) in asymptomatic controls (A) and in emphysema (B). The degree of correlation was determined by Pearson’s correlation r with p<0.05. Species with r >0.6 are shown and the numbers in each box denote the r values. Red colour denotes positive associations between the species and fibre intake or SCFAs, and blue colour denotes negative associations.
We adopted PICRUSt2 (phylogenetic investigation of communities by reconstruction of unobserved states) metagenomic predictions to analyse the functional differences between asymptomatic smokers and emphysema patients. Significance was defined at a false discovery rate<0.05. In faecal samples, 369 pathways were found, with 16 pathways being more abundant in the emphysema group compared with asymptomatic smokers (online supplemental figure S10A). The largest difference in abundance was observed in pathways belonging to the precursor metabolites and energy superclass. In sputum samples, 479 pathways were discovered, and 17 pathways showed statistically significant differences in abundance, with only 2 pathways being more abundant in asymptomatic smokers (online supplemental figure S10B). At least 17 pathways are assumed to be related to SCFA production, of which 6 pathways are associated with acetate formation, 7 pathways with propionate production and 5 pathways with butyrate production (online supplemental figure S11).
In addition, microbial networks were analysed based on SParse and Compositionally Robust Inference,26 which showed that microbial interactions were more active in asymptomatic smokers than in emphysema patients (online supplemental figures S12 and S13 and online supplemental table S4).
Supplemental material
Supplemental material
SCFA administration ameliorates alveolar destruction and inflammatory response in a smoking-induced emphysema model
Based on our findings that acetate and propionate were more predominant in the faecal samples from asymptomatic smokers than those from emphysema patients, we examined whether acetate and propionate modulate the degree of alveolar destruction induced by cigarette smoking in a murine model of emphysema. Mice exposed to cigarette smoke were administered either NaA or NaP (figure 6A).27 28 Although the administration of NaA or NaP did not result in significant differences in body weight in mice exposed to cigarette smoke (figure 6B), administration of NaA and NaP significantly decreased the degree of airspace enlargement in mice exposed to cigarette smoke (figure 6C). Importantly, the lungs of mice supplemented with NaA and NaP showed decreased levels of interferon (IFN)-γ (figure 6D), which is associated with the severity of emphysema by driving the production of proinflammatory cytokines and chemokines.29 Furthermore, the number of CD3+CD4+IFN-γ+, CD3+CD4+IL-17+ and CD3+CD4+CD25+Foxp3+ T cells were significantly lower in the lungs of the NaP group compared with the smoking group (figure 6E). In addition, supplementation of propionate decreased the percentage of CD3+CD4+IL-17+ T cells in the spleen (figure 6F). Then, we examined inflammatory cell infiltration, the levels of proinflammatory cytokine and physiological parameters. NaA and NaP administration decreased the number of lung-infiltrating neutrophils and lymphocytes (online supplemental figure S14A,B), and the levels of IL-6, IL-1β, TNF-α, iNOS and BAL fluid protein (online supplemental figure S14C-J). Physiologic parameters showed a similar trend, although without statistical significance (online supplemental figure S14K). In addition, we also found a consistent effect through intrarectal administration of NaA and NaP (online supplemental figure S15). We further examined the levels of interleukin (IL)-6 and IL-1β in bone marrow-derived macrophages and BEAS-2B cells; consistent with the in vivo results, pretreatment of NaA and NaP nullified the poly (I:C)-induced increase in IL-6 and IL-1β levels (online supplemental figure S16). These results suggest the protective effect of NaA and NaP on the development of smoking-induced emphysema.
Short-chain fatty acid administration ameliorates alveolar destruction and inflammatory response in a murine model of emphysema. (A) Experimental scheme. Mice (n=4–6 per group) were exposed to cigarette smoke for 4 weeks and poly (I:C) was intranasally administered. Sodium acetate (NaA) or sodium propionate (NaP) was administered in drinking water and intraperitoneally. (B) Body weight changes. (C) H&E analysis of lung lobes. Representative sections from five independent experiments are shown. Scale bar=250 µm. Bar graph on the right shows the mean interalveolar septal wall distance of lung tissues. (D) Relative mRNA (left) and protein (right) levels of interferon (IFN)-γ in the lungs of experimental groups. Isolation of lung cells (E) and splenocytes (F). The number or percentage of CD3+CD4+IFN-γ+ T cells, CD3+CD4+IL-17+ T cells or CD3+CD4+FoxP3+ T cells in each group are shown. Values are mean±SEM. *P<0.05 vs SM, **p<0.01 vs SM. ***p<0.001 vs SM, ****p<0.0001 vs SM. CTL, control; NaA=SM+NaA; NaP=SM+NaP; SM, smoking.
Discussion
While the role of the gut microbiome in the systemic inflammatory response has been extensively investigated, its function in respiratory diseases and serious conditions such as COPD has not been well addressed. This study analysed the clinical, gut microbiome and metabolome data of smokers according to the presence of severe emphysema, and found that despite similar smoking history and age, the two groups showed significant differences in the lung function phenotypes. Interestingly, the correlations between fibre intake, acetate, propionate and SCFA-producing bacteria were more pronounced in asymptomatic controls as well. Based on these clinical data, we tested the potential therapeutic effects of acetate and propionate administration in a murine model of smoking-induced emphysema and observed that administration of acetate and propionate protected the mice against alveolar destruction and inflammatory response in the lung.
Our study identified that the respiratory and faecal microbiota ecologies were different in patients with COPD. Sputum samples from patients with COPD showed an increased abundance of Proteobacteria, Moraxellaceae and Eubacteriaceae. These results are consistent with those of other reports that sputum analysis identified a decreased diversity and increased abundance of Moraxella, Streptococcus, Veillonella, Eubacterium and Prevotella in COPD30 and that Proteobacteria including Moraxella and Haemophilus became dominant when exacerbated COPD.31 There were no species that could be used to specifically characterise either asymptomatic smokers or those with COPD by being commonly present in both the sputum and faecal samples from one group and not in those of the other group. LDA showed that P. copri, which is abundant in non-Western populations consuming a plant-rich diet, was overrepresented in the faecal samples of healthy controls.32
Gut metabolites in COPD were also rarely addressed in the literature. Conversely, a high fibre diet is known as one of the protective factors against the development of COPD.33 As SCFAs are major microbial end products from dietary fibres, we hypothesised that there might be significant associations between the metabolome, microbiome and clinical parameters. The positive association of SCFA with fibre intake was only noted in asymptomatic controls and that of acetate reached statistical significance despite the small number of participants. These results are of note considering that SCFAs carry a significant role in the regulation of the host immune system and oxidative stress.34 We also discovered that the association between fibre intake and both the metabolome and microbiome was more pronounced in healthy controls. Interestingly, there was a significant correlation between acetate levels and FEV1, particularly in the emphysema group. This finding raises the intriguing possibility that as airflow limitation becomes more severe in emphysema, there may be a greater acetate depletion. Acetate and propionate are produced from the fermentation of dietary fibre by gut bacteria such as Lactobacillus spp, Blautia spp, Bacteroides spp, Bifidobacterium spp, Prevotella spp and Ruminococcus spp.35 The number of species that were closely correlated with SCFAs (r>0.6) was also fewer in patients with emphysema. In this context, the lack of association between fibre intake and the abundance of SCFAs and SCFA-producing bacteria in patients with emphysema may provide a clue to the pathogenesis of COPD.
The modulation of gut microbiota is relatively easier than that of respiratory microbiota, and the use of prebiotics and probiotics has been investigated using in vivo and in vitro models of COPD.22 36 Based on the results from human samples, we applied postbiotics, acetate and propionate, which have not been tried in the treatment of COPD. Considering the immunoregulatory function of acetate and propionate, such as inhibition of neutrophil infiltration and inflammatory cytokine production,37–39 we expected that a possible mechanism by which SCFA reduces lung destruction in the mouse model of emphysema is by preventing inflammatory cytokine production and the immune cell recruitment. Indeed, supplementation of acetate and propionate reduced infiltrated number of immune cells and the production of IL-6 and IL-1β in the lungs of emphysema mice (online supplemental figure S8). Therefore, appropriate production of SCFA may be essential to prevent disease development against external inflammatory stimuli such as smoking. Further studies using germ-free or antibiotic-treated mice would be useful to better understand the effects of SCFA on the development of emphysema.
Our study had several limitations. First, the respiratory and gut microbiomes are highly dynamic, and their results can vary depending on various confounders and covariates. This variability poses a challenge when characterising the microbiome using a single cross-sectional study design. Therefore, future longitudinal studies will be help provide further evidence and substantiate the findings of this study. The use of inhalers, especially ICS, might be another confounding factor in this study.40 Despite this limitation, our results showed consistent results with previous findings in patients with COPD. Second, we did not identify specific microbiota with therapeutic potentials (ie, probiotics). We identified a total of 46 species that were related to fibre intake, acetate or propionate, suggesting that multiple species were associated with the response to prebiotics and the production of postbiotics. Therefore, intervention using one or two potentially beneficial microbial species may not be enough to modify the disease course. Third, a relatively small number of participants were involved in this study. Performing sample size calculations for each hypothesis in multidimensional data would increase the reliability and power of the study. Furthermore, demonstrating the beneficial effects of a high-fibre diet or SCFAs is an additional challenge, demanding careful and meticulous design in the context of clinical trials. Nevertheless, we were still able to identify meaningful results on the relationship between clinical data, microbiota and metabolome in participants with distinct phenotypes regarding emphysema. Fourth, further studies are necessary to investigate the transmission of T cells and SCFAs from the gut or spleen to the lungs to elucidate the underlying mechanisms of emphysema improvement with SCFA administration.
Taken together, our results suggest that identifying the differences in dietary fibre intake, microbiome and SCFAs may help understand the pathogenesis and individual variability of COPD. We expect that further longitudinal and larger-scale studies will provide concrete evidence of fibre intake, acetate and propionate as novel treatments for COPD.
Data availability statement
The data that support the findings of this study are available from the corresponding author (SWL), upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants. The study was approved by the institutional review board of Asan Medical Center (2018-0980). The protocols of this study were registered in clinicaltrials.gov (NCT03755505). Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank the patients who participated and were willing to provide their samples for this study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
SHL and JK are joint first authors.
SHL and JK contributed equally.
Contributors SWL developed the concept and designed the experiment. SWL and YJ enrolled participants and collected clinical samples. JK, SHL and SWL developed the image reconstruction and analysis methods used in the study. JK, NHK, O-HK, C-HS and S-HH performed animal experiments. SJK and HJY performed metabolome analysis. SY, SEL and SYJ performed microbiome analysis. SHL, JK and SWL wrote the first draft of the manuscript and all authors contributed to redrafting the manuscript. The guarantor is SWL.
Funding This study was supported by the National Research Foundation of Korea (NRF) grant funded by the Korean government (MSIT) (No. 2020R1A2C1008431, 2023R1A2C2006688, RS-2023-00222687, SWL), Basic Science Research Program through the NRF of Korea funded by the Ministry of Education (No. 2020R1I1A1A01069464, JK) and the Bio & Medical Technology Development Program of NRF funded by the Korean government (MSIT) (No. 2022M3A9G8017220).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.