PT - JOURNAL ARTICLE AU - Maxime Lamontagne AU - Wim Timens AU - Ke Hao AU - Yohan Bossé AU - Michel Laviolette AU - Katrina Steiling AU - Joshua D Campbell AU - Christian Couture AU - Massimo Conti AU - Karen Sherwood AU - James C Hogg AU - Corry-Anke Brandsma AU - Maarten van den Berge AU - Andrew Sandford AU - Stephen Lam AU - Marc E Lenburg AU - Avrum Spira AU - Peter D Paré AU - David Nickle AU - Don D Sin AU - Dirkje S Postma TI - Genetic regulation of gene expression in the lung identifies <em>CST3</em> and <em>CD22</em> as potential causal genes for airflow obstruction AID - 10.1136/thoraxjnl-2014-205630 DP - 2014 Nov 01 TA - Thorax PG - 997--1004 VI - 69 IP - 11 4099 - http://thorax.bmj.com/content/69/11/997.short 4100 - http://thorax.bmj.com/content/69/11/997.full SO - Thorax2014 Nov 01; 69 AB - Background COPD is a complex chronic disease with poorly understood pathogenesis. Integrative genomic approaches have the potential to elucidate the biological networks underlying COPD and lung function. We recently combined genome-wide genotyping and gene expression in 1111 human lung specimens to map expression quantitative trait loci (eQTL). Objective To determine causal associations between COPD and lung function-associated single nucleotide polymorphisms (SNPs) and lung tissue gene expression changes in our lung eQTL dataset. Methods We evaluated causality between SNPs and gene expression for three COPD phenotypes: FEV1% predicted, FEV1/FVC and COPD as a categorical variable. Different models were assessed in the three cohorts independently and in a meta-analysis. SNPs associated with a COPD phenotype and gene expression were subjected to causal pathway modelling and manual curation. In silico analyses evaluated functional enrichment of biological pathways among newly identified causal genes. Biologically relevant causal genes were validated in two separate gene expression datasets of lung tissues and bronchial airway brushings. Results High reliability causal relations were found in SNP–mRNA–phenotype triplets for FEV1% predicted (n=169) and FEV1/FVC (n=80). Several genes of potential biological relevance for COPD were revealed. eQTL-SNPs upregulating cystatin C (CST3) and CD22 were associated with worse lung function. Signalling pathways enriched with causal genes included xenobiotic metabolism, apoptosis, protease–antiprotease and oxidant–antioxidant balance. Conclusions By using integrative genomics and analysing the relationships of COPD phenotypes with SNPs and gene expression in lung tissue, we identified CST3 and CD22 as potential causal genes for airflow obstruction. This study also augmented the understanding of previously described COPD pathways.