Integrated genome-wide association, coexpression network, and expression single nucleotide polymorphism analysis identifies novel pathway in allergic rhinitis

Abstract

Background: Allergic rhinitis is a common disease whose genetic basis is incompletely explained. We report an integrated genomic analysis of allergic rhinitis.

Methods: We performed genome wide association studies (GWAS) of allergic rhinitis in 5633 ethnically diverse North American subjects. Next, we profiled gene expression in disease-relevant tissue (peripheral blood CD4+ lymphocytes) collected from subjects who had been genotyped. We then integrated the GWAS and gene expression data using expression single nucleotide (eSNP), coexpression network, and pathway approaches to identify the biologic relevance of our GWAS.

Results: GWAS revealed ethnicity-specific findings, with 4 genome-wide significant loci among Latinos and 1 genome-wide significant locus in the GWAS meta-analysis across ethnic groups. To identify biologic context for these results, we constructed a coexpression network to define modules of genes with similar patterns of CD4+ gene expression (coexpression modules) that could serve as constructs of broader gene expression. 6 of the 22 GWAS loci with P-value ≤ 1x10-6 tagged one particular coexpression module (4.0-fold enrichment, P-value 0.0029), and this module also had the greatest enrichment (3.4-fold enrichment, P-value 2.6 × 10-24) for allergic rhinitis-associated eSNPs (genetic variants associated with both gene expression and allergic rhinitis). The integrated GWAS, coexpression network, and eSNP results therefore supported this coexpression module as an allergic rhinitis module. Pathway analysis revealed that the module was enriched for mitochondrial pathways (8.6-fold enrichment, P-value 4.5 × 10-72).

Conclusions: Our results highlight mitochondrial pathways as a target for further investigation of allergic rhinitis mechanism and treatment. Our integrated approach can be applied to provide biologic context for GWAS of other diseases.

Figure 1

Study flow for the integrated genome-wide association, coexpression network, and expression single nucleotide polymorphism analysis of allergic rhinitis. CHS = Children’s Health Study, CAMP = Childhood Asthma Management Program, CAG = Chicago Asthma Genetics Study, CSGA = Collaborative Studies on the Genetics of Asthma, SARP = Severe Asthma Research Program, GALA1 = Genetics of Asthma in Latinos, MCCAS = Mexico City Childhood Asthma Study, GRAAD = Genomic Research on Asthma in the African Diaspora and Barbados, SAPPHIRE = Study of Asthma Phenotypes and Pharmacogenomic Interactions by Race-Ethnicity. Detailed descriptions of the individual studies have been previously described [25].

Figure 2

Manhattan plot of the genome-wide association and meta-analysis results for allergic rhinitis showing ethnicity-specific findings.

Figure 3

Results of the genome-wide association studies of allergic rhinitis, meta-analysis, and GWAS tagging of the coexpression network.

Figure 4

CD4+ lymphocyte coexpression network with detail of the brown coexpression module. A. Each circle represents a gene. Weighted gene coexpression analysis identified groups of genes with similar patterns of gene expression and interconnectivity (coexpression modules). The 41 coexpression modules identified are labeled by color. Pathways associated with the largest coexpression modules are denoted in the legend. B. Interconnectivity of the brown coexpression module is shown in detail. Tagged by 6 allergic rhinitis GWAS loci, this coexpression module was highly enriched for allergic rhinitis-associated eSNPs (3.4-fold enrichment, FDR-adjusted P value = 2.6 × 10⁻²⁴) and also highly enriched for pathways related to mitochondrial function (8.6-fold enrichment, FDR-adjusted P value = 4.5 × 10⁻⁷²). Genes containing allergic rhinitis-associated eSNPs are marked in brown, with those containing eSNPs with lowest P-value for association between genotype and gene expression marked with greatest brown saturation. Genes in pathways related to mitochondrial function are marked by diamonds with blue outline. Higher correlation between gene expression is shown with thicker and darker edges.

Figure 5

eSNP enrichment and pathway analysis of coexpression modules tagged by GWAS loci.