Document resource
Rationale Idiopathic Pulmonary Fibrosis (IPF) is linked to over 35 genetic risk loci, but their contribution to disease mechanisms remains unclear. DNA methylation (DNAm) is a cell-type specific molecular mechanism that may mediate genetic risk by altering cell type specific gene expression (GE). We previously demonstrated widespread DNAm-associated GE alterations in IPF versus non-IPF fibroblasts derived from airways (AF) and parenchyma (PF). Here, we aimed to test the hypothesis that cell type specific GE is associated with IPF-associated SNPs, with potential modulation by DNAm.Methods DNA and RNA were isolated from AF (8 non-IPF, 8 IPF) and PF (14 non-IPF, 8 IPF) at passage 4 (AllPrep DNA/RNA Mini Kit, Qiagen). DNA was profiled using the Illumina HumanMethylation EPIC array. RNA expression was assessed using Affymetrix Human Gene 2.1ST Array. We analysed the dataset with a new focus on integrating genetic risk. Firstly, IPF associated genes linked to DNAm were overlapped with SNP associated genes. A direct link between DNAm and GE for overlapping genes was assessed via DNMT (5-aza-2′-deoxycytidine, 10 − 10-10− 6 M) and TET (Ten-Eleven Translocation-in-C35, 5×10− 7, 1×10-6, 5×10-6 and 1×10-5 M) inhibition in vitro, in PF from 3 IPF and 3 non-IPF donors. GE was assessed by qPCR. Secondly, we screened ±500 kb regions around each of the 35 IPF-associated SNPs for associated transcriptome changes, by linear modelling.Results Four genes; DSP, STMN3, FAM13A, and TERC, were identified as SNP-associated genes that overlapped with DNAm-associated differentially expressed genes. TET inhibition did not show any impact on GE. DNMT inhibition showed significant reduction in expression of FAM13A in non-IPF cells. Our analysis of ±500 kb regions around IPF-associated SNPs revealed 64 genes differentially expressed between AF and PF in IPF, highlighting cell-type specific regulatory effects of genetic risk loci.Conclusion Our integrative approach highlights an overlap between genetic risk loci and DNAm-associated GE changes in IPF. We observe fibroblast subtype-specific differences in GE near IPF-associated SNPs, though whether these effects are mediated by DNAm is to be further studied. These findings support the potential importance of epigenetic mechanisms in IPF pathogenesis and highlight the need for cell-type specific epigenetic analyses.