LATE-BREAKING ABSTRACT: Network-based meta-analysis of lung, sputum and blood transcriptomics in COPD

Abstract

Background Chronic Obstructive Pulmonary Disease (COPD) is characterized by pulmonary and systemic manifestations. Multiple studies have identified airway or blood biomarkers associated with the severity of airflow limitation, but it is not clear whether these biomarkers directly represent the pathobiology in the lung. Objective To identify pathobiological mechanisms shared between different compartments in COPD patients. Methods Weighted Gene Co-expression Network Analysis (WGCNA) was used to meta-analyse four gene expression datasets from Caucasian former smokers with COPD: two lung tissue datasets (n=70 and n=124), and one each of sputum (n=121) and peripheral blood (n=121). Results Eighteen (86%) of the twenty-one co-expression modules identified were preserved between the two lung tissue datasets; seven (33%) of them were preserved also in sputum, and five also (24%) in blood. One module (labelled “yellow”) was preserved across all datasets and associated with FEV 1 % in lung and sputum but not in blood. This module was enriched in ontologies related to mitochondrial function, cell cycle, ubiquitination and RNA processing. Two other modules (brown and magenta) were preserved and associated with FEV 1 % across lung and sputum datasets (but not in blood). These modules included genes related to ion homeostasis, T cell response, cytokine signalling and RNA processing. Conclusions Our study identified several biological functions in mRNA co-expression modules associated with FEV 1 % in lung and sputum but not in blood. These results suggest that the systemic manifestations of COPD likely constitute an independent component of the disease. Funding Unristricted grant form GSK, U.S. NIH P01 HL105339 and FIS PI15/00799.