Genome-wide association study in Spanish identifies ADAM metallopeptidase with thrombospondin type 1 motif, 9 (ADAMTS9), as a novel asthma susceptibility gene

Abstract

Numerous candidate gene and linkage studies have identified loci associated with asthma susceptibility. Recently, the analysis of single nucleotide polymorphisms (SNPs) from the entire genome by means of genome-wide association studies (GWASs) has provided new insights about asthma pathogenesis. However, these GWASs have revealed a handful of firm asthma susceptibility genes, explaining a small proportion of the disease heritability.1Meyers D.A. Bleecker E.R. Holloway J.W. Holgate S.T. Asthma genetics and personalised medicine.Lancet Respir Med. 2014; 2: 405-415Abstract Full Text Full Text PDF PubMed Scopus (78) Google Scholar Genetic risk can vary among different populations. In previous GWASs Spanish subjects were poorly represented. We hypothesized that a GWAS of asthma in Spanish subjects examining the comprehensive catalogue of genetic variation from the 1000 Genomes Project could reveal new susceptibility genes. Here we performed the first GWAS of asthma in a Spanish population using a 3-stage design analyzing a total of 1180 cases and 1915 control subjects. A workflow with the study design is shown in Fig 1, and Methods are described in this article's Online Repository at www.jacionline.org. Demographic and clinical features of the samples are shown in Table E1 in this article's Online Repository at www.jacionline.org. Association analysis of 6,467,565 genetic variants with asthma in the discovery stage from the Genetics of Asthma (GOA) study in the Spanish population showed no major inflation effects caused by population stratification (genomic control value λGC = 1.038, see Fig E1 in this article's Online Repository at www.jacionline.org). Although none of the variants were associated at the genome-wide level, we identified 337 variants spanning 68 independent loci with suggestive significance at a P value of 5.0 × 10−5 or less (see Fig E1). For replication, we focused on 19 loci with suggestive significance both for genotyped and imputed variants and that were not identified in previous studies (ie, IL33 and IL1RL1-IL18R1).2Moffatt M.F. Gut I.G. Demenais F. Strachan D.P. Bouzigon E. Heath S. et al.A large-scale, consortium-based genomewide association study of asthma.N Engl J Med. 2010; 363: 1211-1221Crossref PubMed Scopus (1489) Google Scholar, 3Torgerson D.G. Ampleford E.J. Chiu G.Y. Gauderman W.J. Gignoux C.R. Graves P.E. et al.Meta-analysis of genome-wide association studies of asthma in ethnically diverse North American populations.Nat Genet. 2011; 43: 887-892Crossref PubMed Scopus (618) Google Scholar From those loci, with independence of the significance level of the imputed variants, we prioritized 1 genotyped variant for subsequent genotyping in replication sample 1 (see Table E2 in this article's Online Repository at www.jacionline.org). Only 2 SNPs were nominally associated with asthma with the same direction of effects as in the discovery stage (rs17199249 and rs9866261, see Table E3 in this article's Online Repository at www.jacionline.org). The SNP rs17199249 is located near bone morphogenetic protein receptor type 2 gene (BMPR2; see Fig E2 in this article's Online Repository at www.jacionline.org), and its minor allele (G allele) was a risk factor for asthma susceptibility in both the discovery (odds ratio [OR], 1.09; P = 3.03 × 10−5) and replication sample 1 (OR, 1.30; P = .039) studies. The rs9866261 is an intronic variant from ADAM metallopeptidase with thrombospondin type 1 motif, 9 gene (ADAMTS9; see Fig E2) with a minor allele (G allele) that is associated with risk for asthma both in the discovery (OR, 1.56; P = 2.71 × 10−5) and replication sample 1 (OR, 1.27; P = .026) studies. We followed up those 2 SNPs for replication in a third sample. At this stage, the association of rs9866261 from ADAMTS9 was replicated (OR, 1.93; P = .007; Table I). No heterogeneity was detected across the 3 stages (P = .165, Cochran Q test) and a fixed-effects meta-analysis attained near genome-wide significance (OR, 1.45; P = 1.31 × 10−7; Table I).Table IAssociation results of the 2 SNPs followed up in all 3 stages of the studyStagers17199249 (BMPR2)rs9866261 (ADAMTS9)MAF casesMAF control subjectsOR (95% CI)∗ORs refer to the minor allele of each polymorphism (G allele).P valueMAF casesMAF control subjectsOR (95% CI)∗ORs refer to the minor allele of each polymorphism (G allele).P valueDiscovery phase (352 cases/537 control subjects)0.1120.1031.09 (1.05-1.14)3.03 × 10−50.1640.1121.56 (1.27-1.92)2.71 × 10−5Replication 1 (482 cases/1209 control subjects)0.1100.0861.30 (1.01-1.67).0390.1580.1251.27 (1.03-1.92).026Replication 2 (346 cases/169 control subjects)0.0840.1070.78 (0.50-1.20).2570.1260.0701.93 (1.20-3.09).007Meta-analysis†Results from a fixed-effect meta-analysis are shown because no heterogeneity was found among studies based on the Cochran Q test (P value >.12). (1180 cases/1915 control subjects)——1.09 (1.03-1.16)3.25 × 10−3——1.45 (1.26-1.66)1.31 × 10−7BMPR2, Bone morphogenetic protein receptor type 2; MAF, minor allele frequency.∗ ORs refer to the minor allele of each polymorphism (G allele).† Results from a fixed-effect meta-analysis are shown because no heterogeneity was found among studies based on the Cochran Q test (P value >.12). Open table in a new tab BMPR2, Bone morphogenetic protein receptor type 2; MAF, minor allele frequency. Finally, we identified in the bibliography 55 SNPs from 35 loci that were associated with asthma at a P value of less than 5.0 × 10−8 in previous studies (see Table E4 in this article's Online Repository at www.jacionline.org). We found evidence of replication for 14 of those SNPs (1.99 × 10−6 ≤ P value ≤ .038, see Table E5 in this article's Online Repository at www.jacionline.org), 12 of which had consistent effects and were located in 6 firm susceptibility loci (IL1RL1-IL18R1, RAD50-IL13, IL33, LRRC32, ORMDL3-GSDMB, and IKZF3). ADAMTS9 is highly expressed in the lungs. This disintegrin and metalloproteinase is involved in the extracellular matrix turnover of versican, among other proteoglycans, a process that has a key role in embryogenesis, tissue morphogenesis, and maintenance. Increased production of large extracellular matrix versican has been implicated in asthma pathogenesis,4Nihlberg K. Andersson-Sjoland A. Tufvesson E. Erjefalt J.S. Bjermer L. Westergren-Thorsson G. Altered matrix production in the distal airways of individuals with asthma.Thorax. 2010; 65: 670-676Crossref PubMed Scopus (59) Google Scholar disease severity,4Nihlberg K. Andersson-Sjoland A. Tufvesson E. Erjefalt J.S. Bjermer L. Westergren-Thorsson G. Altered matrix production in the distal airways of individuals with asthma.Thorax. 2010; 65: 670-676Crossref PubMed Scopus (59) Google Scholar, 5Weitoft M. Andersson C. Andersson-Sjoland A. Tufvesson E. Bjermer L. Erjefalt J. et al.Controlled and uncontrolled asthma display distinct alveolar tissue matrix compositions.Respir Res. 2014; 15: 67Crossref PubMed Scopus (47) Google Scholar and airway remodeling contributing to the persistent airway obstruction and decrease in lung function observed in asthmatic patients.6Chiappara G. Gagliardo R. Siena A. Bonsignore M.R. Bousquet J. Bonsignore G. et al.Airway remodelling in the pathogenesis of asthma.Curr Opin Allergy Clin Immunol. 2001; 1: 85-93Crossref PubMed Scopus (126) Google Scholar The SNP rs9866261 or SNPs in high linkage disequilibrium (LD) with it are not expression quantitative trait loci, but this SNP is in high LD (r2 = 0.80) with rs9311896, which is located in a regulatory region of the gene in fetal lung tissue, as indicated by DNase I–hypersensitive experiments and by the presence of enhancer histone marks according to ENCODE data. Despite this evidence, ADAMTS9 has never been associated with asthma or any related traits by using GWASs or candidate-gene association studies. This could be due to the fact that this SNP was described for the first time by the 1000 Genomes Project. Therefore it is not present in HapMap, the reference data used for imputation in most of previous GWASs of asthma, including the 2 largest meta-analyses.2Moffatt M.F. Gut I.G. Demenais F. Strachan D.P. Bouzigon E. Heath S. et al.A large-scale, consortium-based genomewide association study of asthma.N Engl J Med. 2010; 363: 1211-1221Crossref PubMed Scopus (1489) Google Scholar, 3Torgerson D.G. Ampleford E.J. Chiu G.Y. Gauderman W.J. Gignoux C.R. Graves P.E. et al.Meta-analysis of genome-wide association studies of asthma in ethnically diverse North American populations.Nat Genet. 2011; 43: 887-892Crossref PubMed Scopus (618) Google Scholar However, 2 SNPs in high LD with rs9866261 (r2 ≥ 0.8) were tested in the GABRIEL consortium2Moffatt M.F. Gut I.G. Demenais F. Strachan D.P. Bouzigon E. Heath S. et al.A large-scale, consortium-based genomewide association study of asthma.N Engl J Med. 2010; 363: 1211-1221Crossref PubMed Scopus (1489) Google Scholar but were not associated with asthma in that meta-analysis (P = .343 for rs4688470 and P = .400 for rs9311896). This result could be due to the different ancestral composition of the case-control sample analyzed in our study, which included greater African ancestry compared with other European populations (see Fig E3 in this article's Online Repository at www.jacionline.org). ADAMTS9 has been previously associated with waist-hip ratio, type 2 diabetes, age-related macular degeneration, and anthropometric traits based on GWAS findings and in secondary analyses with obesity, cholesterol, and fasting insulin.7Heid I.M. Jackson A.U. Randall J.C. Winkler T.W. Qi L. Steinthorsdottir V. et al.Meta-analysis identifies 13 new loci associated with waist-hip ratio and reveals sexual dimorphism in the genetic basis of fat distribution.Nat Genet. 2010; 42: 949-960Crossref PubMed Scopus (718) Google Scholar, 8Fritsche L.G. Chen W. Schu M. Yaspan B.L. Yu Y. Thorleifsson G. et al.Seven new loci associated with age-related macular degeneration.Nat Genet. 2013; 45: 433-439Crossref PubMed Scopus (615) Google Scholar, 9Liu C.T. Monda K.L. Taylor K.C. Lange L. Demerath E.W. Palmas W. et al.Genome-wide association of body fat distribution in African ancestry populations suggests new loci.PLoS Genet. 2013; 9: e1003681Crossref PubMed Scopus (97) Google Scholar This result is not surprising given that there are other firm susceptibility genes, such as SMAD3, that have been implicated both in patients with asthma2Moffatt M.F. Gut I.G. Demenais F. Strachan D.P. Bouzigon E. Heath S. et al.A large-scale, consortium-based genomewide association study of asthma.N Engl J Med. 2010; 363: 1211-1221Crossref PubMed Scopus (1489) Google Scholar, 10Ferreira M.A. Matheson M.C. Tang C.S. Granell R. Ang W. Hui J. et al.Genome-wide association analysis identifies 11 risk variants associated with the asthma with hay fever phenotype.J Allergy Clin Immunol. 2014; 133: 1564-1571Abstract Full Text Full Text PDF PubMed Scopus (148) Google Scholar and those with type 2 diabetes.11Dong C. Tang L. Liu Z. Bu S. Liu Q. Wang Q. et al.Landscape of the relationship between type 2 diabetes and coronary heart disease through an integrated gene network analysis.Gene. 2014; 539: 30-36Crossref PubMed Scopus (11) Google Scholar Similarly, the association of ADAMTS9 with obesity and cholesterol levels supports the genetic and epidemiologic data linking both traits with asthma.12Visness C.M. London S.J. Daniels J.L. Kaufman J.S. Yeatts K.B. Siega-Riz A.M. et al.Association of childhood obesity with atopic and nonatopic asthma: results from the National Health and Nutrition Examination Survey 1999-2006.J Asthma. 2010; 47: 822-829Crossref PubMed Scopus (144) Google Scholar, 13Al-Shawwa B. Al-Huniti N. Titus G. Abu-Hasan M. Hypercholesterolemia is a potential risk factor for asthma.J Asthma. 2006; 43: 231-233Crossref PubMed Scopus (70) Google Scholar We cannot rule out that the association with asthma can be confounded by obesity because no body mass index data are available among control subjects in the GOA studies. However, among cases from all 3 stages, the SNP rs9866261 was not associated with body mass index (P = .354). This study has some limitations regarding the selection of control subjects. In the discovery stage no information regarding history of asthma or other respiratory diseases was available, and the control subjects differed in age distribution compared with the cases. This could lead to misclassification of asthma cases as control subjects, reducing the statistical power to detect an association. In the replication studies sex representation, age distribution, or both also differed by case-control status. However, we confirmed the association of 12 SNPs from 6 genes identified in previous GWASs, reinforcing the validity of the findings from our study. The most significant associations corresponded to the IL1RL1-IL18R1 and IL33 loci, where several variants would be significant, even after applying a conservative Bonferroni adjustment for multiple testing (0.05/55, P < 9.0 × 10−4). The lack of replication of some of the loci previously associated with asthma could be due to the different asthma subphenotype analyzed, the fact that some of previous associations are population-specific (ie, PYHIN1 in African Americans),3Torgerson D.G. Ampleford E.J. Chiu G.Y. Gauderman W.J. Gignoux C.R. Graves P.E. et al.Meta-analysis of genome-wide association studies of asthma in ethnically diverse North American populations.Nat Genet. 2011; 43: 887-892Crossref PubMed Scopus (618) Google Scholar the existence of different LD patterns among populations, or our limited statistical power. In conclusion, we identified a suggestive genome-wide significant association of ADAMTS9 with asthma in Spanish subjects. Replication in independent studies will be needed to establish the generalizability of this finding. Download .doc (.19 MB) Help with doc files Online Repository TextFig E2View Large Image Figure ViewerDownload Hi-res image Download (PPT)Fig E3View Large Image Figure ViewerDownload Hi-res image Download (PPT)