Quantitative serological phenotypes reveal regulatory genetic variation underlying host immune responses to brucellosis in naturally infected sheep

Abstract Background Brucellosis is a globally important zoonosis with major impacts on the sheep industry and public health. However, current research efforts mainly rely on heterologous models and vaccine-induced immunity, which fail to accurately capture the host-specific immune mechanisms underlying natural Brucella infection in sheep. Moreover, most genetic studies of brucellosis susceptibility have modeled disease status as a binary trait. Methods Here, three serological assays were applied to 139 sheep from naturally infected populations to characterize brucellosis-specific humoral responses. A discovery-stage dual-phenotype genome-wide association study (GWAS) integrating quantitative traits and binary serostatus was performed, followed by expanded-population validation and epigenomic annotation. Results Serological phenotyping revealed high inter-assay concordance and a broad, continuous spectrum of antibody levels, supporting the use of quantitative immune traits for genetic analysis. Quantitative trait GWAS prioritized 35 single nucleotide polymorphisms (SNPs) consistently associated across antibody measurements, whereas only six SNPs overlapped with signals detected using the binary phenotype. Candidate genes located near these association signals included nuclear factor of activated T cells 2 ( NFATC2 ), diacylglycerol kinase iota ( DGKI ), and neuromedin U receptor 2 ( NMUR2 ). Functional enrichment analyses of genes associated with quantitative traits revealed assay-specific biological patterns, with genes involved in vesicle-mediated transport and lipid and fatty acid metabolic processes. In the expanded cohort, seven SNPs showed reproducible genotype-dependent associations with quantitative antibody responses and collectively explained a substantial proportion of the variance in a composite antibody phenotype derived from multiple assays. Based on 5-fold cross-validation, these SNPs jointly accounted for 22.48% of the phenotypic variance (CV-R² = 0.2248). However, due to some overlap between the discovery and validation samples, this estimate may be affected by the ‘winner’s curse’ and needs further confirmation in future independent populations. Epigenomic annotation suggested regulatory potential for these SNPs, supported by integrated chromatin accessibility and transcription factor binding features from multiple tissues, though direct functional validation is lacking. Conclusion These results suggest that quantitative serological phenotypes enhance the detection of candidate genetic variants associated with host immune responses to brucellosis under natural exposure conditions, providing a practical framework for genetic studies of complex infectious diseases in livestock. Functional validation is required to determine whether these variants play direct causal regulatory roles.