Wang, D.; Ni, Y.; Liu, B.; Ding, H.; John, A.; Wain, L. V.; Johnson, S. R.; Maher, T. M.; Molyneaux, P. L.; Renzoni, E.; Saini, G.; Wells, A. U.; Morris-Rosendahl, D.; Jenkins, R. G.; Stewart, I.

Background Pulmonary fibrosis (PF) is a shared characteristic of chronic interstitial lung diseases of mixed aetiology. Previous studies on PF highlight a pathogenic role for common and rare genetic variants. This study aimed to identify rare pathogenic variants that are enriched in distinct biological pathways and dysregulated gene expression. Methods Rare variants were identified using whole genome sequencing (WGS) from two independent PF cohorts, the PROFILE study and the Genomics England 100K (GE100KGP) cohort, with the gnomAD database as a reference. Four pathogenic variant categories were defined: loss of function variants, missense variants, protein altering variants, and protein truncating variants. Gene burden testing was performed for rare variants defined as having a minor allele frequency <0.1%. Overrepresentation analysis of gene ontology terms and gene concept network analysis were used to interpret functional pathways. Integration of publicly available transcriptomic datasets was performed using weighted gene co-expression network analysis of idiopathic pulmonary fibrosis (IPF) lung tissue compared with healthy controls. Results Burden testing was performed on 507 patients from the PROFILE study and 451 PF patients from GE100KGP cohort, compared with 76,156 control participants from the gnomAD database. Ninety genes containing significantly more pathogenic rare variants in cases than in controls were observed in both cohorts. Fifty-six genes included missense variants and 87 genes included protein altering variants. For missense variants, HMCN1, encoding hemicentin-1, and RGPD1, encoding a protein with a RanBD1 domain, were highly associated with PF in both PROFILE (p=5.70E-22 and p=4.48E-51, respectively) and GE100KGP cohorts (p=2.27E-24 and p=1.59E-36, respectively). 56 of 90 genes with significant burden were observed within modules correlated with disease in transcriptomic analysis, including HMCN1 and RGPD1. Enriched functional categories from genetic and transcriptomic analyses included pathways involving extracellular matrix constituents, cell adhesion properties and microtubule organisation. Conclusions Rare pathogenic variant burden testing and weighted gene co-expression network analysis of transcriptomic data provided complementary evidence for pathways regulating cytoskeletal dynamics in PF pathogenesis. Functional validation of candidates could provide novel targets for intervention strategies.