Multi-ancestry genome-wide association study reveals novel genetic signals for lung function decline

Abstract

Rationale: Accelerated decline in lung function contributes to the development of chronic respiratory disease. Despite evidence for a genetic component, few genetic associations with lung function decline have been identified.

Objectives: To evaluate genome-wide associations and putative downstream functionality of genetic variants with lung function decline in diverse general population cohorts.

Methods: We conducted genome-wide association study (GWAS) analyses of decline in the forced expiratory volume in the first second (FEV₁), forced vital capacity (FVC), and their ratio (FEV₁/FVC) in participants across six cohorts in the Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE) consortium and the UK Biobank. Genotypes were imputed to TOPMed (CHARGE cohorts) or Haplotype Reference Consortium (HRC) (UK Biobank) reference panels, and GWAS analyses used generalized estimating equation models with robust standard error. Models were stratified by cohort, ancestry, and sex, and adjusted for important lung function confounders and genotype principal components. Results were combined in cross-ancestry and ancestry-specific meta-analyses. Selected top variants were tested for replication in two independent COPD-enriched cohorts.

Measurements and main results: Our discovery analyses included 52,056 self-reported White (N=44,988), Black (N=5,788), Hispanic (N=550), and Chinese American (N=730) participants with a mean of 2.3 spirometry measurements and 8.6 years of follow-up. Functional mapping of GWAS meta-analysis results identified 361 distinct genome-wide significant (p<5E-08) variants in one or more of the FEV₁, FVC, and FEV₁/FVC decline phenotypes, which overlapped with previously reported genetic signals for several related pulmonary traits. Of these, 8 variants, or 20.5% of the variant set available for replication testing, were nominally associated (p<0.05) with at least one decline phenotype in COPD-enriched cohorts (White [N=4,778] and Black [N=1,118]). Using the GWAS results, gene-level analysis implicated 38 genes, including eight (XIRP2, GRIN2D, SATB1, MARCHF4, SIPA1L2, ANO5, H2BC10, and FAF2) with consistent associations across ancestries or decline phenotypes. Annotation class analysis revealed significant enrichment of several regulatory processes for corticosteroid biosynthesis and metabolism. Drug repurposing analysis identified 43 approved compounds targeting eight of the implicated 38 genes.

Conclusions: Our multi-ancestry GWAS meta-analyses identified numerous genetic loci associated with lung function decline. These findings contribute knowledge to the genetic architecture of lung function decline, provide evidence for a role of endogenous corticosteroids in the etiology of lung function decline, and identify drug targets that merit further study for potential repurposing to slow lung function decline and treat lung disease.

Keywords: Chronic Obstructive; Genomics; Lung Volume Measurements; Pulmonary Disease; Spirometry.

Figure 1.. Study Design.

Sex and ancestry-stratified GWAS analyses were performed separately in each cohort following a uniform analysis plan and summarized with inverse-variance weighted meta-analysis. Distinct variants with p-values passing genome-wide significance thresholds (p < 5E-08) were considered significant. Significant variants were tested for replication in two COPD-enriched cohorts following the discovery analysis plan. Replication was declared for variants with p-values passing the Bonferroni-adjusted significance threshold (0.05/361 distinct variants).

Figure 2.. Circular Manhattan Plot of Cross-Ancestry Results.

Genome-wide results for decline in FEV₁ (outer circle), FVC (middle circle), and FEV₁/FVC (inner circle) from the cross-ancestry analyses. Dotted red lines denotes the genome-wide significance threshold of p = 5E-08. Red circles represent variants passing genome-wide significance.

Figure 3.. Upset Plot Showing Overlap of Lung Function Decline-Associated Variants with Previously Reported Signals for Relevant Pulmonary Phenotypes.

Significant decline-associated variants were evaluated for prior associations with relevant pulmonary traits (COPD, death from respiratory disease, cross-sectional lung function, asthma/allergic disease, emphysema/chronic bronchitis, and other/non-specified respiratory disease). Dots connected by lines depict intersecting sets and vertical bars show the number of variants in each intersection. The first column shows the variants identified in this GWAS that were not previously reported for the pulmonary traits evaluated.

Figure 4.. Gene-Based Testing of Cross-Ancestry and Ancestry-Specific GWAS Results.

Genes implicated from gene analysis of our GWAS results. A) Z-score comparison across ancestries and decline phenotypes. Color corresponds to Z score value and circle size corresponds to p-value. Red boxes highlight genes with consistent and significant associations across ancestries or phenotypes. B) GTEx data for lung tissue. Note: KIAA1426 synonymous for JCAD; HIST1H2BI synonymous for H2BC10

Figure 5.. Enrichment of Processes Regulating Glucocorticoid and Mineralocorticoid Biosynthesis.

Annotation class analysis of GWAS results revealed significant enrichment of Gene Ontology (GO) biological processes for corticosteroid regulatory pathways. Fold enrichment calculated using the binomial (textured) and hypergeometric (untextured) methods are shown. Color represents false discovery rate adjusted p values.