Pulmonary EV miRNA profiles identify disease and distinct inflammatory endotypes in COPD

INTRODUCTION: Chronic obstructive pulmonary disease (COPD) is a heterogeneous condition without effective disease modifying therapies. Identification of novel inflammatory endotype markers such as extracellular vesicles (EVs), which are important intercellular messengers carrying microRNA (miRNA), may enable earlier diagnosis and disease stratification for a targeted treatment approach. Our aim was to identify differentially expressed EV miRNA in the lungs of COPD patients compared with healthy ex-smokers and determine whether they can help define inflammatory COPD endotypes. METHODS: EV miRNA were isolated and sequenced from ex-smoking COPD patients and healthy ex-smoker bronchoalveolar lavage fluid. Results were validated with RT-qPCR and compared to differential inflammatory cell counts. RESULTS: Expression analysis identified five upregulated miRNA in COPD (miR-223-3p, miR-2110, miR-182-5p, miR-200b-5p and miR-625-3p) and three downregulated miRNA (miR-138-5p, miR-338-3p and miR-204-5p), all with a log2 fold change of >1/−1, FDR < 0.05. These miRNAs correlated with disease defining characteristics such as FEF 25–75% (a small airways disease measure) and DLCO% (a surrogate measure of emphysema). Receiver operator curve analysis demonstrated miR-2110, miR-223-3p, and miR-182-5p showed excellent combinatory predictive ability (AUC 0.91, p < 0.0001) in differentiating between health and mild COPD. Furthermore, miR-223-3p and miR-338-3p correlated with airway eosinophilia and were able to distinguish “pure eosinophilic” COPD from other airway inflammatory subtypes (AUC 0.94 and 0.85, respectively). DISCUSSION: This is the first study to identify differentially expressed miRNA in COPD bronchoalveolar lavage fluid EVs. These findings suggest specific lung derived EV miRNA are a strong predictor of disease presence even in mild COPD. Furthermore, specific miRNA correlated with inflammatory cell numbers in COPD, and may have a role in defining inflammatory endotypes for future treatment stratification.