The genetic architecture of innate fear behavior in chickens is poorly understood. Here, we performed quantitative trait loci (QTL) analysis of innate responses to tonic immobility (TI) and open field (OF) fears in 242 newly hatched chicks of an F 2 population between the native Japanese Nagoya breed and the White Leghorn breed using 881 single nucleotide polymorphism markers obtained by restriction site-associated DNA sequencing. At genome-wide 5% significance levels, four QTL for TI traits were revealed on chromosomes 1-3 and 24. Two of these loci had sex-specific effects on the traits. For OF traits, three QTL were revealed on chromosomes 2, 4 and 7. The TI and OF QTL identified showed no overlaps in genomic regions and different modes of inheritance. The three TI QTL and one OF QTL exerted antagonistic effects on the traits. The results demonstrated that context-dependent QTL underlie the variations in innate TI and OF behaviors.

Innate fear responses to tonic immobility (TI) and open field (OF) were evaluated in newly hatched chicks of three breeds with distinct breed origin and genetic relationships. The breeds studied were Nagoya (NAG), a native Japanese breed; White Leghorn (WL), a representative of layers; and White Plymouth Rock (WPR), a parental breed of common broilers. The TI test revealed that WL was the most sensitive to extensive fear evoked by the TI test among the three breeds, followed in order by WPR, and NAG. In contrast, the OF test revealed that NAG was the most sensitive to mild fear evoked by the OF test, followed in order by WPR, and WL. The different fear responses between NAG and WL were supported by minimal phenotypic correlations between TI and OF traits in each breed. These results demonstrated that NAG and WL breeds exhibit extreme and opposite responses to TI and OF fears.

Identifying the genes responsible for quantitative traits remains a major challenge. We previously found a major QTL on chromosome 4 affecting several innate fear behavioral traits obtained by an open field test in an F2 population between White Leghorn and Nagoya breeds of chickens (Gallus gallus). Here, an integrated approach of transcriptome, haplotype frequency and association analyses was used to identify candidate genes for the QTL in phenotypically extreme individuals selected from the same segregating F2 population as that used in the initial QTL analysis. QTL mapping for the first principal component, which summarizes the variances of all affected behavioral traits in the F2 population, revealed the behavioral QTL located at 14-35 Mb on chromosome 4 with 333 genes. After RNA-seq analysis using two pooled RNAs from extreme F2 individuals, real-time qPCR analysis in the two parental breeds and their F1 individuals greatly reduced the number of candidate genes in the QTL interval from 333 to 16 genes. Haplotype frequency analysis in the two extreme F2 groups further reduced the number of candidate genes from 16 to 11. After comparing gene expression in the two extreme groups, a conditional correlation analysis of diplotypes between gene expression and phenotype of extreme individuals revealed that NPY5R and LOC101749214 genes were strong candidate genes for innate fear behavior. This study illustrates how the integrated approach can identify candidate genes more rapidly than fine mapping of the initial QTL interval and provides new information for studying the genetic basis of innate fear behavior in chickens.