Dissecting the lung transcriptome of pulmonary fibrosis-associated pulmonary hypertension

Abstract

Integrative multiomics can help elucidate the pathophysiology of pulmonary fibrosis (PF) associated pulmonary hypertension (PH) (PF-PH). Weighted gene co-expression network analysis (WGCNA) was performed on a transcriptomic dataset of explanted lung tissue from 116 patients with PF. Patients were stratified by pulmonary vascular resistance (PVR) and differential gene expression analysis was conducted. Gene modules were correlated with hemodynamics at the time of transplantation and tested for enrichment in the lung transcriptomics signature of an independent pulmonary arterial hypertension (PAH) cohort. We found 1,250 differentially expressed genes between high and low PVR groups. WGCNA identified that black and yellowgreen modules negatively correlated with PVR, while the tan and darkgrey modules positively correlated with PVR. Additionally, the tan module showed the strongest enrichment for an independent PAH gene signature, suggesting shared gene expression patterns between PAH and PF-PH. Pharmacotranscriptomic analysis using the Connectivity Map implicated the tan and darkgrey modules as potentially pathogenic in PF-PH, given their combined module signature demonstrated a high negative connectivity score for Treprostinil, a medication used in the treatment of PF-PH, and a high positive connectivity score for Bone morphogenetic protein loss of function. Pathway enrichment analysis revealed that inflammatory pathways and oxidative phosphorylation were downregulated, whereas epithelial mesenchymal transition was upregulated in modules associated with increased PVR. Our integrative systems biology approach to the lung transcriptome of PF with and without PH identified several PH-associated co-expression modules and gene targets with shared molecular features with PAH warranting further investigation to uncover potential new therapies for PF-PH.