Genome-wide imputation study identifies novel HLA locus for pulmonary fibrosis and potential role for auto-immunity in fibrotic idiopathic interstitial pneumonia

Abstract

Background: Fibrotic idiopathic interstitial pneumonias (fIIP) are a group of fatal lung diseases with largely unknown etiology and without definitive treatment other than lung transplant to prolong life. There is strong evidence for the importance of both rare and common genetic risk alleles in familial and sporadic disease. We have previously used genome-wide single nucleotide polymorphism data to identify 10 risk loci for fIIP. Here we extend that work to imputed genome-wide genotypes and conduct new RNA sequencing studies of lung tissue to identify and characterize new fIIP risk loci.

Results: We performed genome-wide genotype imputation association analyses in 1616 non-Hispanic white (NHW) cases and 4683 NHW controls followed by validation and replication (878 cases, 2017 controls) genotyping and targeted gene expression in lung tissue. Following meta-analysis of the discovery and replication populations, we identified a novel fIIP locus in the HLA region of chromosome 6 (rs7887 P meta = 3.7 × 10(-09)). Imputation of classic HLA alleles identified two in high linkage disequilibrium that are associated with fIIP (DRB1*15:01 P = 1.3 × 10(-7) and DQB1*06:02 P = 6.1 × 10(-8)). Targeted RNA-sequencing of the HLA locus identified 21 genes differentially expressed between fibrotic and control lung tissue (Q < 0.001), many of which are involved in immune and inflammatory response regulation. In addition, the putative risk alleles, DRB1*15:01 and DQB1*06:02, are associated with expression of the DQB1 gene among fIIP cases (Q < 1 × 10(-16)).

Conclusions: We have identified a genome-wide significant association between the HLA region and fIIP. Two HLA alleles are associated with fIIP and affect expression of HLA genes in lung tissue, indicating that the potential genetic risk due to HLA alleles may involve gene regulation in addition to altered protein structure. These studies reveal the importance of the HLA region for risk of fIIP and a basis for the potential etiologic role of auto-immunity in fIIP.

Keywords: Gene expression; HLA association; Imputation; Pulmonary fibrosis; RNA-Seq.

Fig. 1

Overview of Study Design. A discovery imputation GWAS among 1616 cases and 4683 controls was followed by validation and replication genotyping in 878 cases and 2017 controls. One novel locus was identified on Chromosome 6p21. Classical HLA alleles were imputed using genotyped SNPs among the GWAS cases and controls and were tested for association with fIIP. Lung tissue gene expression was compared between a subset of GWAS cases and non-GWAS controls and across genotypes for the most significant 6p21 SNP, rs7887. Gene expression was also compared across genotypes of the most significant HLA alleles (DQB1*06:02, DRB1*15:01) within the cases with lung tissue expression data

Fig. 2

Locus-specific plot for HLA region corresponding to discovery imputation GWAS results. The –log₁₀ P values (y axis) of the SNPs are shown according to their chromosomal positions (x axis). The estimated recombination rates (cM/Mb) from the HapMap Project (NCBI Build 36) are shown as light blue lines, and the genomic locations of genes within the regions of interest in the NCBI Build 36 human assembly are shown as arrows. SNPs shown in red, orange, green, light blue and blue have r ² ≥ 0.8, r ² ≥ 0.6, r ² ≥ 0.4, r ² ≥ 0.2 and r ² < 0.2 with the most highly-associated SNP, respectively. SNPs with no r ² information with most-highly associated SNP shown in grey. Circles correspond to genotyped SNPs, squares correspond to imputed SNPs. P-values correspond to discovery cohort statistical evidence only; meta-analysis p-values can be found in Table 1 and Additional file 1: Tables S1-S3

Fig. 3

Targeted RNA-Seq Expression Differences between IPF and Control Lung. The size of the dots corresponds to q-value; larger dots have smaller q-values. The color of the dots corresponds to the direction of expression changes; genes in blue have lower expression in case lung tissue compared to control lung tissue, genes in red have higher expression