Lung microbiome composition and bronchial epithelial gene expression in patients with COPD versus healthy individuals: a bacterial 16S rRNA gene sequencing and host transcriptomic analysis.

Background: Chronic obstructive pulmonary disease (COPD) is associated with airway inflammation and bacterial dysbiosis. The relationship between the airway microbiome and bronchial gene expression in COPD is poorly understood. We aimed to identify differences in the airway microbiome from bronchial brushings in patients with COPD and healthy individuals and to investigate whether any distinguishing bacteria are related to bronchial gene expression.

Methods: For this 16S rRNA gene sequencing and host transcriptomic analysis, individuals aged 45-75 years with mild-to-moderate COPD either receiving or not receiving inhaled corticosteroids and healthy individuals in the same age group were recruited as part of the Emphysema versus Airways Disease (EvA) consortium from nine centres in the UK, Germany, Italy, Poland, and Hungary. Individuals underwent clinical characterisation, spirometry, CT scans, and bronchoscopy. From bronchoscopic bronchial brush samples, we obtained the microbial profiles using 16S rRNA gene sequencing and gene expression using the RNA-Seq technique. We analysed bacterial genera relative abundance and the associations between genus abundance and clinical characteristics or between genus abundance and host lung transcriptional signals in patients with COPD versus healthy individuals, and in patients with COPD with versus without inhaled corticosteroids treatment.

Findings: Between February, 2009, and March, 2012, we obtained brush samples from 574 individuals. We used 546 of 574 samples for analysis, including 207 from healthy individuals and 339 from patients with COPD (192 with inhaled corticosteroids and 147 without). The bacterial genera that most strongly distinguished patients with COPD from healthy individuals were Prevotella (median relative abundance 33·5%, IQR 14·5-49·4, in patients with COPD vs 47·7%, 31·1-60·7, in healthy individuals; p<0·0001), Streptococcus (8·6%, 3·8-15·8, vs 5·3%, 3·0-10·1; p<0·0001), and Moraxella (0·05%, 0·02-0·14, vs 0·02%, 0-0·07; p<0·0001). Prevotella abundance was inversely related to COPD severity in terms of symptoms and positively related to lung function and exercise capacity. 446 samples had assessable RNA-seq data, 257 from patients with COPD (136 with inhaled corticosteroids and 121 without) and 189 from healthy individuals. No significant associations were observed between lung transcriptional signals from bronchial brushings and abundance of bacterial genera in patients with COPD without inhaled corticosteroids treatment and in healthy individuals. In patients with COPD treated with inhaled corticosteroids, Prevotella abundance was positively associated with expression of epithelial genes involved in tight junction promotion and Moraxella abundance was associated with expression of the IL-17 and TNF inflammatory pathways.

Interpretation: With increasing severity of COPD, the airway microbiome is associated with decreased abundance of Prevotella and increased abundance of Moraxella in concert with downregulation of genes promoting epithelial defence and upregulation of pro-inflammatory genes associated with inhaled corticosteroids use. Our work provides further insight in understanding the relationship between microbiome alteration and host inflammatory response, which might lead to novel therapeutic strategies for COPD.

Funding: EU Seventh Framework Programme, National Institute for Health Research.