A Peer-reviewed,
open access bimonthly journal
pISSN 2092-7355   eISSN 2092-7363
    Allergy Asthma Immunol Res. 2020 Sep;12(5):806-820. English.
    Published online May 11, 2020.
    https://doi.org/10.4168/aair.2020.12.5.806
    Copyright © 2020 The Korean Academy of Asthma, Allergy and Clinical Immunology • The Korean Academy of Pediatric Allergy and Respiratory Disease
    Original Article

    Phenotypic and Molecular Characterization of Risk Loci Associated With Asthma and Lung Function

    Mehmet Karaca,1 Nazente Atceken,2 Şefayet Karaca,3 Ersoy Civelek,4 Bülent E. Şekerel,5 and Renato Polimanti6
      • 1Department of Biology, Faculty of Science and Arts, Aksaray University, Aksaray, Turkey.
      • 2Graduate School of Natural and Applied Sciences, Aksaray University, Aksaray, Turkey.
      • 3Department of Nutrition and Dietetics, Faculty of Health Science, Aksaray University, Aksaray, Turkey.
      • 4Pediatric Allergy and Immunology Clinic, Ankara Child Health and Diseases Hematology Oncology Research Hospital, Ankara, Turkey.
      • 5Pediatric Allergy and Asthma Unit, Faculty of Medicine, Hacettepe University, Ankara, Turkey.
      • 6Department of Psychiatry, Yale University School of Medicine and VA CT Healthcare Center, West Haven, CT, United States.
    • Correspondence to Mehmet Karaca, PhD. Department of Biology, Faculty of Science and Arts, Aksaray University, Bahçesaray District, 68100 Aksaray, Turkey. Tel: +90-382-2882227; Fax: +90-382-2882299; Email: mkaraca@aksaray.edu.tr
    Received January 08, 2020; Revised February 25, 2020; Accepted February 27, 2020.

    This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (https://creativecommons.org/licenses/by-nc/4.0/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

    Abstract

    Purpose

    Respiratory diseases have a highly multifactorial etiology where different mechanisms contribute to the individual's susceptibility. We conducted a deep characterization of loci associated with asthma and lung function by previous genome-wide association studies (GWAS).

    Methods

    Sixteen variants were selected from previous GWAS of childhood/adult asthma and pulmonary function tests. We conducted a phenome-wide association study of these loci in 4,083 traits assessed in the UK Biobank (n = 361,194 participants). Data from the Genotype-Tissue Expression (GTEx) project were used to conduct a transcriptomic analysis with respect to tissues relevant for asthma pathogenesis. A pediatric cohort assessed with the International Study of Asthma and Allergies in Children (ISAAC) Phase II tools was used to further explore the association of these variants with 116 traits related to asthma comorbidities.

    Results

    Our phenome-wide association studies (PheWAS) identified 206 phenotypic associations with respect to the 16 variants identified. In addition to the replication of the phenotypes tested in the discovery GWAS, we observed novel associations related to blood levels of immune cells (eosinophils, neutrophils, monocytes, and lymphocytes) for the asthma-related variants. Conversely, the lung-function variants were associated with phenotypes related to body fat mass. In the ISAAC-assessed cohort, we observed that risk alleles associated with increased fat mass can exacerbate allergic reactions in individuals affected by allergic respiratory diseases. The GTEx-based analysis showed that the variants tested affect the transcriptomic regulation of multiple surrounding genes across several tissues.

    Conclusions

    This study generated novel data regarding the genetics of respiratory diseases and their comorbidities, providing a deep characterization of loci associated with asthma and lung function.

    Keywords
    Allergy; asthma; genetics; lung function; transcriptome; loci

    INTRODUCTION

    Respiratory diseases are a major cause of mortality and morbidity worldwide, and children are particularly vulnerable to them.1 Understanding the molecular processes responsible for respiratory diseases can consistently improve our ability to prevent and treat the disorders, reducing their harmful consequences. Genetic studies have demonstrated that pediatric respiratory diseases and lung function are heritable and several risk variants have been identified by genome-wide association studies (GWAS).2 Understanding the mechanisms underlying the genetic associations identified remains a challenge. Genetic variants can affect the regulation of multiple genes and be associated with multiple phenotypes via pleiotropic mechanisms including phenotypic causal relationships or shared biological pathways.3 To date, numerous methods and datasets are available to deepen the biological meaning of genetic associations identified in large GWAS.4 Information regarding tissue-specific transcriptomic regulation can be used to disentangle which are the genes and the tissues involved in the disease pathogenesis.5 Phenome-wide association studies (PheWAS) can permit us to detect the full phenotypic spectrum associated with a specific genetic variant and to extricate the possible causal relationships linking the genetic variation to the phenotype of interest.6, 7

    In the present study, we conducted phenomic and transcriptomic analyses with respect to 16 variants (Table 1) identified by GWAS of childhood/adult asthma and forced vital capacity.8, 9, 10 Four of the variants investigated were already identified by GWAS of autoimmune diseases including rheumatoid arthritis,11 ulcerative colitis,12 eczema,13 juvenile idiopathic arthritis,14 and systemic sclerosis.15 To fully understand the phenotypic spectrum associated with the investigated variants, we conducted a PheWAS including 4,083 phenotypic traits assessed in 361,194 participants from UK Biobank.16 We leveraged data from the Genotype-Tissue Expression (GTEx) project, which includes expression quantitative trait locus (eQTL) information with respect to 49 tissue sites,17 to identify changes induced by these variants in tissue-specific transcriptomic regulation. Finally, we tested associations of the investigated variants with phenotypic traits related to pediatric respiratory diseases and their comorbidities derived from a cohort of 919 Turkish children using International Study of Asthma and Allergies in Children (ISAAC) Phase II tools.18 Through these multiple resources, we provided a deep phenotypic and molecular characterization of risk loci, unraveling the mechanisms by which these genetic variants are associated with asthma and lung function.

    Table 1
    Variants associated with asthma and lung function

    MATERIALS AND METHODS

    UK biobank

    The UK Biobank cohort is an open-access resource available to investigate a wide range of severe and life-threatening illnesses and also normal-range traits.16 This project has recruited more than 500,000 people assessed through detailed web-based questionnaires on their diet, cognitive function, work history, health status, and other relevant phenotypes. In our analysis, we considered GWAS datasets related to 361,194 unrelated participants of European descent. These data were generated testing 4,083 phenotypes using regression models available in Hail (available at https://github.com/hail-is/hail) including the first 20 ancestry principal components, sex, age, age,2 sex×age, and sex×age2 as covariates. Details regarding QC criteria, GWAS methods, and the GWAS data are available at https://github.com/Nealelab/UK_Biobank_GWAS/tree/master/imputed-v2-gwas. Based on these data, we conducted a phenome-wide investigation, applying a Bonferroni multiple testing correction accounting for the number of variants (n = 16) and phenotypes tested (n = 4,083).

    To verify that the associations observed were not false-positive results, we conducted phenome-wide scans with respect to randomly selected variants matched to the variants investigated on the basis of 4 genomic characteristics: i) minor allele frequency, ii) linkage disequilibrium (LD) proxies, iii) distance to the nearest gene, and iv) gene density. This control set of matched variants was selected using the SNPsnap tool19 according to the following parameters: 1,000 Genomes European reference panel (which is the closest reference panel among those available in SNPsnap); LD distance cutoff of R2 = 0.05; ± 5% point deviation; ± 50% of gene density relative deviation; ± 50% of relative deviation of the distance to nearest gene; and ± 50% of relative deviation of LD proxies. For each variant investigated (Table 1), we extracted up to 100 matched single-nucleotide polymorphism (SNP), excluding human leucocyte antigen region due to its complex LD structure.

    GTEx

    GTEx version 8 (V8) data were used to analyze the effect of alleles investigated on the expression of the surrounding genes (± 1 Mb of the gene transcription starting site) across human tissues.17 The association between genetic variants and gene expression was conducted using a linear regression analysis considering the covariates (i.e., top-3 ancestry principal components, genotyping array platform, Probabilistic Estimation of Expression Residuals factors, and sex) applied to gene-cis-SNP pairs using FastQTL20 and assuming an additive model. A detailed description of the GTEx methods (i.e., preprocessing, expression quantification, and association analysis) used is available at https://gtexportal.org/home/documentationPage#staticTextAnalysisMethods. In our analysis, we initially considered single-tissue eQTL associations surviving genome-wide multiple testing correction (false discovery rate; FDR, q ≤ 0.05). To systematically evaluate the effect across multiple tissues, we considered multi-tissue data calculated using Meta-Tissue.21 This meta-analytic approach permits us to calculate a posterior probability (m value) that an effect exists in each of the tissues tested. M values > 0.9 were considered to indicate whether the tissue was predicted to show the eQTL association.

    ISAAC cohort

    The subjects studied (n = 919) were selected from a cross-sectional study conducted using the ISAAC Phase II tools on schoolchildren from 5 city centers (Van, Manisa, Ankara, Antalya, and Trabzon) located in 5 regions of Turkey (East, Aegean, Central Anatolian, Mediterranean, and Black Sea regions, respectively) between September 15, 2005 and May 30, 2006.18 The procedures used in this study conform to the tenets of the Declaration of Helsinki and received approval from the Ethical Review Board of Ankara Child Health and Diseases Research Hospital (2017-041). Written parental and student consent was obtained separately for each participant. The study was approved by the appropriate ethics committees: the Ankara Child Health and Diseases Research Hospital, the Turkish Ministry of Health, and the central and provincial directors of the Ministry of Education and city governors. Further details regarding the sampling and phenotype assessments, and comorbidity and phenotypic correlation in this cohort are available in previous publications.7, 22, 23 Genotyping was performed using the Kompetitive Allele-Specific PCR (KASP) technique.24 DNA 5–10 ng were used per well, and polymerase chain reaction (PCR) reactions were carried out in the presence of no-template controls (NTCs) at a 10-µL final volume (GeneAmp PCR System 9700, Applied Biosystems). The KASP master mix, assay mix, and cycling conditions were based on manufacturers' protocols (available at http://www.kbioscience.co.uk). KlusterCaller™ software was used to evaluate genotyping data. In the analysis of the ISAAC cohort, Plink 1.07 tool25 was used to implement logistic and linear regression analyses for the associations between genetic variants (additive model) and phenotypic traits (binary and quantitative, respectively). Quantitative traits were normalized using appropriate Box-Cox power transformations before being entered into the analysis. We adjusted the association analysis according to the following covariates: age, sex, and sampling center. Due to the small sample size, since the ISAAC cohort was used for final external validation of the variants that were surviving Bonferroni multiple testing correction in both UK Biobank and GTEx analyses, we considered a nominal threshold (P < 0.05) as a significance criterion.

    Data availability

    Data supporting the findings of this study are available within this article and its additional files. UK Biobank GWAS summary association data are available at https://github.com/Nealelab/UK_Biobank_GWAS/tree/master/imputed-v2-gwas. GTEx data are available at https://gtexportal.org/home/.

    RESULTS

    In the UK Biobank PheWAS, we identified 206 genetic associations surviving Bonferroni multiple testing correction (P < 7.65 × 10−7; Supplementary Table S1). While asthma risk alleles showed the strongest associations with traits included “medical information” category (Fig. 1A), variants associated with lung function showed the strongest association with phenotypes included in the physical characteristics (Fig. 1B). Additionally, asthma risk alleles were associated with a larger effect size than those observed for lung-function associated loci. The blood levels of immune system cells showed the strongest significance with the asthma risk alleles: eosinophil percentage (rs928413, P = 7.61 × 10−264), neutrophil count (rs3894194, P = 4.26 × 10−197), leukocyte count (rs3894194, P = 8.53 × 10−171), monocyte percentage (rs3894194, P = 7.36 × 10−75), and lymphocyte percentage (rs3894194, P = 1.47 × 10−57). The variants associated with blood immune cell levels were identified by the asthma GWAS,8, 9 but not by the lung function GWAS.10 Among the lung function-associated variants, rs4237643 showed the largest number of phenotypic associations (n = 33). This variant was identified by the lung function GWAS and, in the UK Biobank cohort, we observed a replication for these phenotypes (forced vital capacity [FVC], P = 1.28 × 10−8; forced expiratory volume in 1 second [FEV1], P = 2.83 × 10−7). However, rs4237643 showed much stronger associations with phenotypes related to fat mass distribution such as arm fat mass (P = 7.16 × 10−23), whole-body fat mass (P = 4.17 × 10−22), body mass index (P = 2.54 × 10−21), and leg fat mass (P = 1.02 × 10−20). A similar phenotypic spectrum (i.e., associations with body mass phenotypes, but not with respiratory diseases or blood immune cell levels) was also observed with respect to the other variants identified by lung-function GWAS.

    Fig. 1
    Manhattan plot of the phenome-wide scan conducted in the UK Biobank with respect to asthma risk alleles (A) and lung function-associated variants (B). As indicated in the legend, the phenotypic categories are color coded.
    FVC, forced vital capacity; FEV1, forced expiratory volume in 1 second; ICD, International Statistical Classification of Diseases and Related Health Problems.

    Beyond these associations, we also observed additional findings that may be related to a broader effect of these variants across human health and disease domains (Table 1). Multiple variants were associated with basal metabolic rate (rs1430193, P = 2.39 × 10−8; rs4237643, P = 1.24 × 10−10; rs6923462, P = 3.56 × 10−7), number of self-reported non-cancer illnesses (rs1558641, P = 3.09 × 10−8; rs6871536, P = 9.2 × 10−9; rs928413, P = 3.92 × 10−7), and dietary habits (cheese intake: rs4237643, P = 2 × 10−7; salt added to food: rs3894194, P = 1.56 × 10−7). Additional associations were observed for rs1430193 with multiple hernia phenotypes (hernia P = 1.14 × 10−15; inguinal hernia P = 1.67 × 10−21; femoral hernia P = 3.79 × 10−7), for rs744910 with corneal resistance factor (P = 1.48 × 10−7), for rs2305480 with age at menopause (P = 5.74 × 10−7), and for rs2305480 with mouth ulcers (P = 4.01 × 10−7) (Table 2).

    Table 2
    Broader phenotypic associations with asthma and lung-function risk variants

    Our control analysis conducted with respect to randomly selected variants matched to the asthma and lung-function associated variants for multiple genomic features showed that the empirical probability to observe a phenome-wide association surviving multiple testing correction by chance is 0.2%.

    Considering single-tissue eQTL, we observed that 14 out of 16 investigated variants are a genome-wide significant eQTL at least with respect to a gene in a tissue. A total of 385 eQTLs reached genome-wide multiple testing correction (FDR, q ≤ 0.05; Supplementary Table S2). Rs1342326 and rs6923462 were the only 2 variants that did not show significant effects on gene expression. Among the significant eQTLs, we observed that the asthma risk alleles tend to be eQTLs for multiple genes across several tissues, while variants associated with lung function are eQTLs for a reduced number of genes and tissues (Fig. 2A). Across the tissues and the genes tested with respect to each variant, we observed a wide range of normalized effect sizes (NES; Fig. 2B). Considering the 2 tissues that expected to be more relevant for asthma pathogenesis and inter-individual variability of lung function, 7 variants showed genome-wide significance with multiple genes in whole blood and lung tissue. The strongest eQTL association was observed for rs2305480 with respect to GSDMB gene expression in whole blood (G allele, NES = −0.3, P = 7.1 × 10−54; Fig. 3A). Although it showed a weaker association, rs2305480-GSDMB eQTL was also genome-wide significant in lung tissue (NES = −0.18, P = 5.7 × 10−14). Considering the multi-tissue analysis, rs2305480-GSDMB eQTL (multi-tissue P = 8.32 × 10−152; Supplementary Fig. S1) showed a posterior probability greater than 90% in additional tissues including spleen (m = 1), Epstein-Barr virus-transformed lymphocytes (m = 1), small intestine (m = 1), colon transverse (m = 1), and stomach (m = 0.99). Beyond GSDMB, rs2305480 affects the transcriptomic regulation of additional genes, including ORMDL3 (whole blood: NES = −0.38, P = 4.4 × 10−51; lung: NES = −0.22, P = 1.2 × 10−16; multi-tissue eQTL P = 2.35 × 10−154), GSDMA (lung: NES = 0.44, P = 6.3 × 10−16; multi-tissue eQTL P = 8.44 × 10−217), IKZF3 (whole blood: NES = 0.07, P = 1.2 × 10−7; multi-tissue eQTL P = 3.83 × 10−13), and PGAP3 (lung: NES = −0.12, P = 3.2 × 10−7; multi-tissue eQTL P = 9.2 × 10−113). A similar eQTL pattern was also observed for rs3894194, which is partially in LD and closely located to rs2305480. On chromosome 11, rs4237643 regulated the transcriptomic of HSD17B12 in both whole blood (T allele, NES = 0.46, P = 2.4 × 10−49; Fig. 3B) and lung (NES = 0.52, P = 3.6 × 10−46). The multi-tissue eQTL analysis showed that rs4237643-HSD17B12 association has a posterior probability greater than 90% in 48 tissues available in GTEx V8 (multi-tissue eQTL P < 10−300; Supplementary Fig. S2). Rs4237643 also showed regulatory effects on 2 long non-coding RNAs, RP11-613D13.10 (lung: NES = 0.6, P = 3.7 × 10−42; whole blood: NES = 0.35, P = 4.8 × 10−20; multi-tissue eQTL P < 10−300) and RP11-613D13.5 (lung: NES = 0.25, P = 1.3 × 10−11; multi-tissue eQTL P = 5.84 × 10−155). Rs1558641 affects the transcriptomic regulation of MFSD9 in whole blood (G allele; NES = 0.35, P = 1.3 × 10−8; Fig. 3C) and IL18R1 in lung (NES = −0.18, P = 5.9 × 10−5). Although it reached significance in the multi-tissue eQTL analysis (P = 7.87 × 10−7), whole blood is the only tissue with a posterior probability greater than 90% in the rs1558641-MFSD9 eQTL (m = 1). Conversely, rs1558641-IL18R1 association (multi-tissue eQTL P = 2.35 × 10−44) showed a posterior probability greater than 90% in lung (m = 1) together with 22 additional tissues available in GTEx (Supplementary Fig. S3). A similar eQTL pattern was observed for rs3771166, which is closely located and in partial LD with rs1558641 (distance: 220 kb, LD r2 = 0.13 in the European population of the 1,000 Genomes Project reference panel).26 Rs744910 showed genome-wide significant effect on AAGAB transcriptomic regulation in whole blood (G allele; NES = 0.1, P = 4.9 × 10−5; Fig. 3D). The multi-tissue rs744910-AAGAB eQTL analysis showed that only whole blood and esophagus mucosa have a posterior probability greater than 90% (multi-tissue eQTL P = 7.92 × 10−6; whole blood m = 0.98; esophagus mucosa m = 1; Supplementary Fig. S4). Similarly, rs2284033 affects ELFN2 transcriptome profile in lung (G allele; NES = 0.15, P = 4.6 × 10−5; Fig. 3E) and the multi-tissue eQTL analysis showed that this association is specific for lung tissue (lung m = 0.90; Supplementary Fig. S5).

    Fig. 2
    eQTL analysis of asthma risk alleles (triangles) and lung function-associated variants (circles). (A) The relationship between the number of tissues and genes regulated for each locus investigated. (B) Relationship between normalized effect size and statistical significance for each locus investigated.
    eQTL, expression quantitative trait locus; FVC, forced vital capacity; NES, normalized effect sizes.

    Fig. 3
    Most significant eQTLs with respect to whole blood and lung tissue.
    eQTL, expression quantitative trait locus.

    Considering the 11 variants that showed significant eQTL effects and significant associations in the phenome-wide scan, we tested their association with 116 phenotypic traits (Supplementary Table S3) derived from a pediatric cohort (n = 919) assessed using ISAAC Phase II tools. We observed 60 genetic associations with respect to 10 of the variants investigated (P < 0.05; Table 3). Comparing asthma risk alleles with respect to lung-function associated variants, we observed that risk alleles are associated with a larger number of ISAAC-derived phenotypes thanks (6.5 ± 1.41 vs. 2.67 ± 2.3, respectively). The strongest findings were observed with respect to rs3894194 (an asthma risk allele), which was associated with eight different phenotypes and showed an association with increased risk of being diagnosed with asthma bronchitis at least twice (P = 8.68 × 10−4). Beyond other asthma definitions and comorbidities, rs3894194 was associated with reactions to the skin prick test for cat and horse atopy (P = 0.031 and P = 0.015, respectively). With respect to the only lung-function associated locus, rs4237643 was associated with four ISAAC-derived phenotypes, including being diagnosed with allergic bronchitis at least twice (P = 0.017), mean expiratory flow (FEV) after 480s of hypertonic saline inhalation (P = 0.023), absolute neutrophil count (P = 0.035), and FEV after bronchodilator drug inhalation subsequently to bronchial challenge test (BCT; P = 0.045).

    Table 3
    Associations of asthma and lung-function risk variants with supporting evidence from phenome-wide and eQTL analyses in the pediatric cohort assessed using ISAAC Phase II tools

    DISCUSSION

    Asthma and comorbid respiratory diseases have a highly multifactorial etiology where different pathogenic mechanisms contribute to the individual's susceptibility.27 In the present study, we explored the phenotypic and molecular spectrum of risk loci associated with asthma and lung function. There is a well-known relationship between respiratory diseases and lung function28 also including altered lung function in asthmatic patients.29 However, there is limited knowledge regarding molecular mechanisms linking these phenotypic traits. We leveraged multiple resources to provide a deep characterization of the loci investigated, and verified similarities and differences among them. We leveraged 3 different datasets: i) UK Biobank, a large-scale cohort representative of the characteristics of the general population; ii) GTEx, a dataset providing broad information regarding the role of genetic variation in regulating gene expression in multiple tissues; and iii) a cohort specifically recruited and assessed to investigate asthma and allergies in children.

    Our PheWAS conducted in the UK Biobank identified more than 200 phenotypic associations with respect to asthma and lung function-associated loci. In both cases, we replicated the original GWAS phenotypes (asthma and lung function, respectively), but observed much stronger associations with respect to additional traits. Asthma risk loci were strongly associated with blood levels of immune cells and particularly with eosinophil level. This is in line with the known role of immune cells in inflammatory processes at the basis of asthma and allergy pathogenesis.30 Additionally, eosinophils in asthmatic inflammation have been a recognized risk factor for many years and eosinophilic airway inflammation seems to be associated with risk of severe asthma exacerbation and loss of asthma control with inhaled corticosteroid withdrawal.31 Our data indicate that risk loci associated with asthma phenotypes are involved in the regulation of immune systems. Based on these findings, we hypothesize that causal mechanisms responsible for the genetic associations reported in the asthma discovery GWAS8, 9 are related to this regulatory effect: individuals predisposed to have altered immune function have an increased risk for developing asthma. In line with previous studies that observed the effect of eosinophils and other immune cells in both adult and childhood asthma,32 we observed that the blood levels of immune cells were correlated with risk loci associated with both adult and childhood asthma, suggesting that immune alterations are a pathogenic pathway shared between the different disease-onset phenotypes. Investigating the effect of asthma risk alleles on transcriptomic regulation of the surrounding genes, we were able to gain more information about molecular mechanisms responsible for the association between asthma risk and immune alterations. Our eQTL analyses showed that asthma risk alleles tend to affect multiple genes across several tissues. Among the variants investigated, the asthma-increasing risk allele rs2305480*G is associated with the down-regulation of GSDMB, ORMDL3, and PGAP3 and the up-regulation of GSDMA and IKZF3 across multiple tissues. Accordingly, we can hypothesize that the effect of rs2305480 on asthma risk and the blood levels of immune cells is due to mechanisms related to the function of multiple genes, including the GSDMB and GSDMA regulation of apoptosis in epithelial cells,33, 34 ORMDL3 regulation of the innate immune system,35 and IKZF3 regulation of B lymphocyte proliferation and differentiation.36, 37 Among the other variants tested, the asthma-increasing risk rs1558641*G allele was associated with the up-regulation of MFSD9 and down-regulation of IL18R1. The gene MFSD9 is involved in transmembrane transporter activities38 and rs1558641 appears to affect its regulation specifically in whole blood. Conversely, rs1558641 regulates IL18R1 gene expression across multiple tissues. This gene encodes for the cytokine receptor that binds interleukin 18 (IL18), and it is essential for IL18-mediated signal transduction.39 Accordingly, we hypothesize that rs1558641 association with asthma and the blood levels of immune cells is due to the regulatory function of the gene IL18R1.

    Unlike asthma risk loci, variants associated with lung function did not show any association with blood immune cells. Conversely, we observed a consistent association with several traits related to body fat mass together with replication of the lung-function traits. Obesity and pathologic accumulation of fat mass are risk factors for both allergic and non-allergic respiratory diseases.40, 41 Our data support the hypothesis that the causative mechanisms responsible for the genetic associations reported in the lung-function discovery GWAS10 are related to the effect of body fat mass: individuals with risk variants associated with increased fat mass have an increased predisposition of reduced lung function. Among the lung function-associated variants, rs4237643*T allele is associated with the up-regulation of HSD17B12 gene which encodes 17-beta-hydroxysteroid dehydrogenase, an enzyme involved in the long-chain fatty acid elongation cycle and the metabolism of steroid hormones.42 HSD17B12 locus has been associated with body mass index,43 type 2 diabetes,44 and cardiovascular diseases.13 Accordingly, we hypothesize the effect of lung function-associated loci investigated is due to the mediated pleiotropy,45 where variants are associated with altered metabolic processes that lead to increased body fat and obesity, which consequently affects lung function. This hypothesis is supported by studies showing the accumulation of adipose tissue causes pulmonary complications directly (mechanical effect of fat accumulation in the chest and abdominal regions) and indirectly (altered immune function due to the numerous cytokines produced by adipocytes).46, 47, 48

    In the pediatric cohort assessed using the ISAAC Phase II tools, we observed that asthma risk alleles are associated with a broader range of asthma- and allergy-related phenotypes than lung function-associated variants. However, we observed that 2 lung function- and fat mass-associated variants (i.e., rs2863171 and rs4237643) are associated with atopy, asthmatic and allergic bronchitis, and the blood levels of immune cells. Although these results will be needed to be confirmed in a larger cohort, these findings may support the effect of increased fat mass in the positive feedback loop between local inflammation in adipose tissue and altered immune response.49

    In conclusion, this study generated novel data regarding the genetics of respiratory diseases and their comorbidities, providing a deep characterization of loci associated with asthma and lung function based on multiple sources. Our findings support that asthma and lung function have distinct genetic architecture, but addictive mechanisms may be present, linking these phenotypes.

    SUPPLEMENTARY MATERIALS

    Supplementary Table S1

    Significant associations surviving Bonferroni multiple testing correction (P < 7.65 × 10−7) in the UK Biobank PheWAS

    Click here to view.(57K, xls)

    Supplementary Table S2

    eQTLs surviving genome-wide multiple testing correction

    Click here to view.(66K, xls)

    Supplementary Table S3

    List of the phenotypes investigated in the pediatric cohort assessed using the ISAAC Phase II tools

    Click here to view.(34K, xls)

    Supplementary Fig. S1

    Multi-tissue eQTL results of rs2305480 with respect to GSDMB.

    Click here to view.(348K, ppt)

    Supplementary Fig. S2

    Multi-tissue eQTL results of rs4237643 with respect to HSD17B12.

    Click here to view.(325K, ppt)

    Supplementary Fig. S3

    Multi-tissue eQTL results of rs1558641 with respect to IL18R1.

    Click here to view.(323K, ppt)

    Supplementary Fig. S4

    Multi-tissue eQTL results of rs744910 with respect to AAGAB.

    Click here to view.(329K, ppt)

    Supplementary Fig. S5

    Multi-tissue eQTL results of rs2284033 with respect to ELFN2.

    Click here to view.(317K, ppt)

    Notes

    Disclosure:There are no financial or other issues that might lead to conflict of interest.

    ACKNOWLEDGMENTS

    This study was supported by grants (2017-024 and 2018-057) from the Scientific Research Unit of Aksaray University. The study sponsor had not any involvement regarding the study design, data evaluation, writing of the manuscript and, its submission for publication. The authors are indebted to the field workers, those who collaborated in the participating centers, and all the parents and school staff who assisted in the survey. The authors thank the participants and investigators of the UK Biobank and the Neale lab for generating the genome-wide data used in the present study.

    References

      1. Zar HJ, Ferkol TW. The global burden of respiratory disease-impact on child health. Pediatr Pulmonol 2014;49:430–434.
      1. Ober C. Asthma genetics in the post-GWAS era. Ann Am Thorac Soc 2016;13 Suppl 1:S85–S90.
      1. Karaca S, Erge S, Cesuroglu T, Polimanti R. Nutritional habits, lifestyle, and genetic predisposition in cardiovascular and metabolic traits in Turkish population. Nutrition 2016;32:693–701.
      1. Pasaniuc B, Price AL. Dissecting the genetics of complex traits using summary association statistics. Nat Rev Genet 2017;18:117–127.
      1. Greene CS, Krishnan A, Wong AK, Ricciotti E, Zelaya RA, Himmelstein DS, et al. Understanding multicellular function and disease with human tissue-specific networks. Nat Genet 2015;47:569–576.
      1. Polimanti R, Kranzler HR, Gelernter J. Phenome-wide association study for alcohol and nicotine risk alleles in 26394 women. Neuropsychopharmacology 2016;41:2688–2696.
      1. Karaca S, Civelek E, Karaca M, Sahiner UM, Ozgul RK, Kocabas CN, et al. Allergy-specific phenome-wide association study for immunogenes in Turkish children. Sci Rep 2016;6:33152.
      1. Bønnelykke K, Sleiman P, Nielsen K, Kreiner-Møller E, Mercader JM, Belgrave D, et al. A genome-wide association study identifies CDHR3 as a susceptibility locus for early childhood asthma with severe exacerbations. Nat Genet 2014;46:51–55.
      1. Moffatt MF, Gut IG, Demenais F, Strachan DP, Bouzigon E, Heath S, et al. A large-scale, consortium-based genomewide association study of asthma. N Engl J Med 2010;363:1211–1221.
      1. Loth DW, Soler Artigas M, Gharib SA, Wain LV, Franceschini N, Koch B, et al. Genome-wide association analysis identifies six new loci associated with forced vital capacity. Nat Genet 2014;46:669–677.
      1. Eyre S, Bowes J, Diogo D, Lee A, Barton A, Martin P, et al. High-density genetic mapping identifies new susceptibility loci for rheumatoid arthritis. Nat Genet 2012;44:1336–1340.
      1. McGovern DP, Gardet A, Törkvist L, Goyette P, Essers J, Taylor KD, et al. Genome-wide association identifies multiple ulcerative colitis susceptibility loci. Nat Genet 2010;42:332–337.
      1. Kichaev G, Bhatia G, Loh PR, Gazal S, Burch K, Freund MK, et al. Leveraging polygenic functional enrichment to improve GWAS power. Am J Hum Genet 2019;104:65–75.
      1. Hinks A, Cobb J, Marion MC, Prahalad S, Sudman M, Bowes J, et al. Dense genotyping of immune-related disease regions identifies 14 new susceptibility loci for juvenile idiopathic arthritis. Nat Genet 2013;45:664–669.
      1. Terao C, Kawaguchi T, Dieude P, Varga J, Kuwana M, Hudson M, et al. Transethnic meta-analysis identifies GSDMA and PRDM1 as susceptibility genes to systemic sclerosis. Ann Rheum Dis 2017;76:1150–1158.
      1. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK biobank: an open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med 2015;12:e1001779
      1. GTEx Consortium; Laboratory, Data Analysis &Coordinating Center (LDACC)—Analysis Working Group; Statistical Methods groups—Analysis Working Group; Enhancing GTEx (eGTEx) groups; NIH Common Fund; NIH/NCI. Genetic effects on gene expression across human tissues. Nature 2017;550:204–213.
      1. Civelek E, Cakir B, Boz AB, Yuksel H, Orhan F, Uner A, et al. Extent and burden of allergic diseases in elementary schoolchildren: a national multicenter study. J Investig Allergol Clin Immunol 2010;20:280–288.
      1. Pers TH, Timshel P, Hirschhorn JN. SNPsnap: a web-based tool for identification and annotation of matched SNPs. Bioinformatics 2015;31:418–420.
      1. Ongen H, Buil A, Brown AA, Dermitzakis ET, Delaneau O. Fast and efficient QTL mapper for thousands of molecular phenotypes. Bioinformatics 2016;32:1479–1485.
      1. Sul JH, Han B, Ye C, Choi T, Eskin E. Effectively identifying eQTLs from multiple tissues by combining mixed model and meta-analytic approaches. PLoS Genet 2013;9:e1003491
      1. Civelek E, Cakir B, Orhan F, Yuksel H, Boz AB, Uner A, et al. Risk factors for current wheezing and its phenotypes among elementary school children. Pediatr Pulmonol 2011;46:166–174.
      1. Civelek E, Yavuz ST, Boz AB, Orhan F, Yuksel H, Uner A, et al. Epidemiology and burden of rhinitis and rhinoconjunctivitis in 9- to 11-year-old children. Am J Rhinol Allergy 2010;24:364–370.
      1. He C, Holme J, Anthony J. SNP genotyping: the KASP assay. Methods Mol Biol 2014;1145:75–86.
      1. Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MA, Bender D, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 2007;81:559–575.
      1. 1000 Genomes Project Consortium. Auton A, Brooks LD, Durbin RM, Garrison EP, Kang HM, et al. A global reference for human genetic variation. Nature 2015;526:68–74.
      1. Bønnelykke K, Ober C. Leveraging gene-environment interactions and endotypes for asthma gene discovery. J Allergy Clin Immunol 2016;137:667–679.
      1. Lange P, Celli B, Agustí A, Boje Jensen G, Divo M, Faner R, et al. Lung-function trajectories leading to chronic obstructive pulmonary disease. N Engl J Med 2015;373:111–122.
      1. McGeachie MJ, Yates KP, Zhou X, Guo F, Sternberg AL, Van Natta ML, et al. Patterns of growth and decline in lung function in persistent childhood asthma. N Engl J Med 2016;374:1842–1852.
      1. Kubo M. Innate and adaptive type 2 immunity in lung allergic inflammation. Immunol Rev 2017;278:162–172.
      1. Price DB, Rigazio A, Campbell JD, Bleecker ER, Corrigan CJ, Thomas M, et al. Blood eosinophil count and prospective annual asthma disease burden: a UK cohort study. Lancet Respir Med 2015;3:849–858.
      1. Holtzman MJ. Asthma as a chronic disease of the innate and adaptive immune systems responding to viruses and allergens. J Clin Invest 2012;122:2741–2748.
      1. Qiu S, Liu J, Xing F. ‘Hints’ in the killer protein gasdermin D: unveiling the secrets of gasdermins driving cell death. Cell Death Differ 2017;24:588–596.
      1. Saeki N, Usui T, Aoyagi K, Kim DH, Sato M, Mabuchi T, et al. Distinctive expression and function of four GSDM family genes (GSDMA-D) in normal and malignant upper gastrointestinal epithelium. Genes Chromosomes Cancer 2009;48:261–271.
      1. Boeck A, Landgraf-Rauf K, Vogelsang V, Siemens D, Prazeres da Costa O, Klucker E, et al. Ca2+ and innate immune pathways are activated and differentially expressed in childhood asthma phenotypes. Pediatr Allergy Immunol 2018;29:823–833.
      1. Somasundaram R, Prasad MA, Ungerbäck J, Sigvardsson M. Transcription factor networks in B-cell differentiation link development to acute lymphoid leukemia. Blood 2015;126:144–152.
      1. Tsai DY, Hung KH, Chang CW, Lin KI. Regulatory mechanisms of B cell responses and the implication in B cell-related diseases. J Biomed Sci 2019;26:64.
      1. Perland E, Hellsten SV, Schweizer N, Arapi V, Rezayee F, Bushra M, et al. Structural prediction of two novel human atypical SLC transporters, MFSD4A and MFSD9, and their neuroanatomical distribution in mice. PLoS One 2017;12:e0186325
      1. Yu Q, Chang HC, Ahyi AN, Kaplan MH. Transcription factor-dependent chromatin remodeling of Il18r1 during Th1 and Th2 differentiation. J Immunol 2008;181:3346–3352.
      1. Pinto LM, Alghamdi M, Benedetti A, Zaihra T, Landry T, Bourbeau J. Derivation and validation of clinical phenotypes for COPD: a systematic review. Respir Res 2015;16:50.
      1. Gonzalez-Garcia M, Caballero A, Jaramillo C, Maldonado D, Torres-Duque CA. Prevalence, risk factors and underdiagnosis of asthma and wheezing in adults 40 years and older: a population-based study. J Asthma 2015;52:823–830.
      1. Kemiläinen H, Adam M, Mäki-Jouppila J, Damdimopoulou P, Damdimopoulos AE, Kere J, et al. The hydroxysteroid (17β) dehydrogenase family gene HSD17B12 is involved in the prostaglandin synthesis pathway, the ovarian function, and regulation of fertility. Endocrinology 2016;157:3719–3730.
      1. Turcot V, Lu Y, Highland HM, Schurmann C, Justice AE, Fine RS, et al. Protein-altering variants associated with body mass index implicate pathways that control energy intake and expenditure in obesity. Nat Genet 2018;50:26–41.
      1. Mahajan A, Taliun D, Thurner M, Robertson NR, Torres JM, Rayner NW, et al. Fine-mapping type 2 diabetes loci to single-variant resolution using high-density imputation and islet-specific epigenome maps. Nat Genet 2018;50:1505–1513.
      1. Hackinger S, Zeggini E. Statistical methods to detect pleiotropy in human complex traits. Open Biol 2017;7:170125
      1. Makki K, Froguel P, Wolowczuk I. Adipose tissue in obesity-related inflammation and insulin resistance: cells, cytokines, and chemokines. ISRN Inflamm 2013;2013:139239
      1. Dixon AE, Peters U. The effect of obesity on lung function. Expert Rev Respir Med 2018;12:755–767.
      1. Mafort TT, Rufino R, Costa CH, Lopes AJ. Obesity: systemic and pulmonary complications, biochemical abnormalities, and impairment of lung function. Multidiscip Respir Med 2016;11:28.
      1. de Heredia FP, Gómez-Martínez S, Marcos A. Obesity, inflammation and the immune system. Proc Nutr Soc 2012;71:332–338.
    Cited by
    Crossref 1
    Google Scholar 
    PubMed Central 1
    Scopus 1
    Web of Science 1
    MeSH Terms
    Adipose Tissue
    Alleles
    Asthma*
    Child
    Cohort Studies
    Genetics
    Genome-Wide Association Study
    Humans
    Hypersensitivity
    Lung*
    Monocytes
    Neutrophils
    Phenotype
    Respect
    Respiratory Function Tests
    Transcriptome
    Metrics
    Share
    Links to
    Figures
    slide
    slide
    slide

    1 / 3

    Tables
    slide
    slide
    slide

    1 / 3

    ORCID IDs
    Funding Information
    PERMALINK