Association of exacerbation phenotype with the sputum microbiome in chronic obstructive pulmonary disease patients during the clinically stable state

Abstract

Abstract Background Chronic obstructive pulmonary disease (COPD) is a progressive, life-threatening lung disease with increasing prevalence and incidence worldwide. Increasing evidence suggests that lung microbiomes might play a physiological role in acute exacerbations of COPD. The objective of this study was to characterize the association of the microbiota and exacerbation risk or airflow limitation in stable COPD patients. Methods The sputum microbiota from 78 COPD outpatients during periods of clinical stability was investigated using 16S rRNA V3-V4 amplicon sequencing. The microbiome profiles were compared between patients with different risks of exacerbation, i.e., the low risk exacerbator (LRE) or high risk exacerbator (HRE) groups, and with different airflow limitation severity, i.e., mild to moderate (FEV1 ≥ 50; PFT I) or severe to very severe (FEV1 < 50; PFT II). Results The bacterial diversity (Chao1 and observed OTUs) was significantly decreased in the HRE group compared to that in the LRE group. The top 3 dominant phyla in sputum were Firmicutes, Actinobacteria, and Proteobacteria, which were similar in the HRE and LRE groups. At the genus level, compared to that in the LRE group (41.24%), the proportion of Streptococcus was slightly decreased in the HRE group (28.68%) (p = 0.007). However, the bacterial diversity and the proportion of dominant bacteria at the phylum and genus levels were similar between the PFT I and PFT II groups. Furthermore, the relative abundances of Gemella morbillorum, Prevotella histicola, and Streptococcus gordonii were decreased in the HRE group compared to those in the LRE group according to linear discriminant analysis effect size (LEfSe). Microbiome network analysis suggested altered bacterial cooperative regulation in different exacerbation phenotypes. The proportions of Proteobacteria and Neisseria were negatively correlated with the FEV1/FVC value. According to functional prediction of sputum bacterial communities through Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt) analysis, genes involved in lipopolysaccharide biosynthesis and energy metabolism were enriched in the HRE group. Conclusion The present study revealed that the sputum microbiome changed in COPD patients with different risks of exacerbation. Additionally, the bacterial cooperative networks were altered in the HRE patients and may contribute to disease exacerbation. Our results provide evidence that sputum microbiome community dysbiosis is associated with different COPD phenotypes, and we hope that by understanding the lung microbiome, a potentially modifiable clinical factor, further targets for improved COPD therapies during the clinically stable state may be elucidated.