Integrative genomic analyses and Fine-Mapping of the 5q31.1 Cytokine Locus Links Food Allergy Risk to Asthma and Atopic Dermatitis Through Shared Causal Variants.
文献摘要
Background Food allergy (FA) is a common immune-mediated disorder, yet its genetic architecture and underlying mechanisms remain incompletely understood. Objectives We aimed to characterize the genetic architecture of FA, refine association signals through multi-ancestry fine-mapping, and investigate shared genetic and regulatory mechanisms linking FA with asthma and atopic dermatitis (AD). Methods We performed a multi-ancestry GWAS meta-analysis of FA including more than 16,000 cases and over 1 million controls of European, African, and East Asian ancestry. We used sequential conditional analysis and Bayesian fine-mapping to identify independent association signals and prioritize candidate causal variants. We further evaluated shared genetic architecture with asthma and AD through conditional colocalization analysis and integrated FA-associated variants with cis-eQTL and single-nucleus ATAC/RNA-seq data from blood and skin to nominate candidate target genes and disease-relevant cell types. Results This trans-ancestry design enabled fine-mapping of the 5q31.1 locus (IL4, IL5, and IL13), revealing multiple independent disease-associated signals, including a secondary genome-wide significant variant (rs6873732; p ≤ 5 × 10 -8 ) and a tertiary signal surpassing Bonferroni significance (rs4608913; p = 9.1 × 10 -6 ). Bayesian fine-mapping identified credible causal variants that also associate with asthma and atopic dermatitis (AD). Genetic colocalization demonstrated that the primary FA signal at 5q31.1 aligns with a secondary AD signal (rs11242122), while the secondary and tertiary FA signals at 5q31.1 colocalize with two independent asthma signals (rs12652920 and rs77450473). Outside 5q31.1, we confirmed a genome-wide significant association at 1q21.3 and identified candidate loci at 11q13.5 and 18q21.33-q22.1, each showing strong evidence of shared causal variants (PP > 0.8) with asthma and AD. To connect genetic risk to cellular mechanisms, we integrated single-nucleus ATAC/RNA-seq from blood and skin. At 5q31.1, three variants within the primary credible set overlapped accessible chromatin in basophils and T cells, while seven tertiary-signal variants overlapped open chromatin in basophils, monocytes, keratinocytes, and CD8 + T cells. Integration with cis-eQTL data demonstrated that primary and tertiary FA signals were associated with increased SLC22A5/IL13 expression and upregulation of IRF1, respectively. Conclusion Together, these findings refine the genetic architecture of FA, reveal shared and disease-specific mechanisms with related allergic diseases, highlight the value of integrating ancestry-diverse GWAS with multi-omics data to prioritize regulatory mechanisms in complex immune-mediated conditions.