Europe PMC
Nothing Special   »   [go: up one dir, main page]

Europe PMC requires Javascript to function effectively.

Either your web browser doesn't support Javascript or it is currently turned off. In the latter case, please turn on Javascript support in your web browser and reload this page.

This website requires cookies, and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our privacy notice and cookie policy.

Identification of candidate genes responsible for innate fear behavior in the chicken.

Author information

Affiliations

  • 1. Laboratory of Animal Genetics and Breeding, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan.
      Authors
    • Ochiai T 1
    • Sakaguchi M 1
    • Ishikawa A 1
    (3 authors)
  • 2. Laboratory of Animal Behavior and Physiology, Graduate School of Integrated Sciences for Life, Hiroshima University, Higashi-Hiroshima, Hiroshima 739-8528, Japan.
      Authors
    • Kawakami SI 2
    (1 author)

ORCIDs linked to this article

G3 (Bethesda, Md.), 01 Feb 2023, 13(2):jkac316
https://doi.org/10.1093/g3journal/jkac316 PMID: 36454218 PMCID: PMC9911055

This article is in the Europe PMC Open access subset. Refer to the copyright information in the article for licensing details.
Free full text in Europe PMC

Abstract 

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.

Free full text 

Logo of gthreeLink to Publisher's site
G3 (Bethesda). 2023 Feb; 13(2): jkac316.
Published online 2022 Dec 1. https://doi.org/10.1093/g3journal/jkac316
PMCID: PMC9911055
PMID: 36454218

Identification of candidate genes responsible for innate fear behavior in the chicken

Takayuki Ochiai, Marina Sakaguchi, Shin-Ichi Kawakami, and Akira Ishikawacorresponding author
Author information Article notes Copyright and License information Disclaimer

Associated Data

Supplementary Materials
Data Availability Statement

Abstract

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.

Keywords: QTL, candidate gene, chicken, innate fear behavior

Introduction

Large numbers of quantitative trait loci (QTLs) affecting traits of agricultural, medical, and biological importance have been mapped to chromosomal regions across most of the genomes of animals including livestock, poultry, model animals, and humans. However, it is difficult to identify causal genes or causal genetic variants for common QTLs with relatively small phenotypic effects, as reviewed previously (Mackay 2001; Keane et al. 2011; Albert and Kruglyak 2015; Ishikawa 2017; Visscher et al. 2017). QTLs can usually be identified either by conventional genome-wide QTL analysis based on linkage mapping in a three-generation crossbred population or by a genome-wide association study based on linkage disequilibrium in a single outbred population. Due to linkage disequilibrium, the genomic interval of each identified QTL is generally not small enough to positionally pinpoint a single candidate gene for the QTL. In recent years, various QTL mapping methods have been developed utilizing next-generation sequencing (Durbin et al. 2010; Daetwyler et al. 2014; Akbari et al. 2021), which can map QTLs more precisely than the conventional methods described above. However, most single nucleotide polymorphisms (SNPs) identified by conventional methods and next-generation sequencing are located in noncoding regions that usually contain elements that regulate gene expression, such as promoters, enhancers, and cis-regulatory elements. Identification of causal genes for QTLs therefore remains a major challenge. Recently, it has become increasingly clear that a single analysis is not sufficient for prioritizing candidate genes for the QTL and that different analyses need to be integrated to obtain multiple lines of evidence that consistently support one or a few candidate genes (Mackay 2001; Ishikawa 2017).

The above-described situation applies to the behavioral traits of chickens (Gallus gallus). Many behavioral QTLs have been reported (Buitenhuis et al. 2004; Schütz et al. 2004; Johnsson et al. 2016; Fogelholm et al. 2019; Ishikawa et al. 2020), and most of the QTLs are deposited in the Chicken QTL Database (Hu et al. 2019). Approximately 15 candidate genes for anxiety and tonic immobility behaviors were identified in QTL analyses, followed by advanced intercross lines between wild and domestic chickens (Johnsson et al. 2016; Fogelholm et al. 2019). However, no causal genes for behavioral QTLs have been reported in chickens.

We previously identified QTLs affecting innate open-field behavior on chicken chromosomes 2, 4, and 7 in an F2 intercross population between the G line of the White Leghorn breed (WL-G) and the native Japanese Nagoya breed (NAG; Ishikawa et al. 2020). Among the QTLs, the chromosome 4 QTL was localized in the 95% confidence interval between 14 and 35 Mb and affected seven open-field traits, suggesting that it is a major open-field locus. Interestingly, the allele derived from NAG increased open-field activity, even though NAG was more sensitive than WL-G to open-field fear (Sakaguchi and Ishikawa 2020).

In the present study, to identify candidate genes for the chromosome 4 QTL described above, we conducted an integrated approach of transcriptome analysis, haplotype frequency analysis, and two association analyses using phenotypically extreme individuals selected from the same F2 population as that used in the initial QTL analysis. The integrated approach eliminated the need to develop additional new crossbreeding populations for fine mapping of the initial QTL interval, which is time-consuming and labor-intensive, and successfully identified two strong candidate genes for the QTL.

Materials and methods

Animals

The NAG, WL-G, F1, and F2 chicks used in this study were previously produced (Ishikawa et al. 2020; Sakaguchi and Ishikawa 2020). After conducting the open-field test at 1 day of age, the body weight of the chicks was recorded, and the diencephalon was collected from each chick euthanized by decapitation and stored at −80°C. Chicks were given water from hatching to 1 day of age, but not food (Ishikawa et al. 2020; Sakaguchi and Ishikawa 2020). All chicks used in this study were handled in accordance with the guidelines of the Animal Research Committee of Nagoya University. All animal experiments were approved by the Animal Research Committee at the Graduate School of Bioagricultural Sciences, Nagoya University (authorization number AGR2019016) and conducted in accordance with the committee's guidelines. This study was also performed in compliance with the ARRIVE guidelines.

Open-field test

The open-field test was previously performed on F2 chicks at 1 day after hatching (Ishikawa et al. 2020). Briefly, each chick was placed in the lower left corner of a novel arena (54 cm × 79 cm × 30 cm) that had a periphery zone within 15 cm from the edge of the arena and a center zone (39 cm × 64 cm) inside the periphery zone. The behavior of each chick was videotaped for 10 minutes, and the videotaped recordings were analyzed using SMART v3.0 software (Panlab Harvard Apparatus, California, USA) to obtain data on 14 behavioral traits. The 14 traits obtained were as follows: number of entries in the center zone, latency of the first entrance to the center zone, total time in the center zone, distance in the periphery zone, distance in the center zone, total distance, resting time, slow time, fast time, mean speed, mean speed without resting, maximum speed, parallel index, and number of excrements.

Among the 14 traits obtained, only seven traits that were affected by the QTL (Ishikawa et al. 2020) were used in this study. Of the seven traits, four traits (total distance, resting time, mean speed, and mean speed without resting) were significantly affected and the other three traits (distance in the periphery zone, slow time, and fast time) were suggestively affected. Whether raw data for the seven traits are statistically affected by environmental factors such as sex and body weight has been tested previously (Ishikawa et al. 2020). Since the previous test found that only the hatching date significantly affected all seven traits, and other environmental factors such as sex and body weight had no effect at P < 0.05, the raw data were adjusted for hatching date. Adjusted data for seven traits from 241 F2 individuals (125 males and 116 females) were used in this study.

Principal component analysis

The adjusted seven open-field trait data were subjected to principal component analysis using a correlation matrix of JMP Pro software version 15.2.1 (SAS Institute Japan Ltd., Tokyo, Japan). The first principal component scores, the second principal component scores, and loading factors for the seven traits were calculated by JMP Pro software.

QTL analysis

Using the first principal component scores obtained above and 881 SNP markers developed previously (Ishikawa et al. 2020), a single-QTL genome scan was performed on 241 F2 individuals using the Haley–Knott regression method with the function scanone of R/qtl software version 3.6.3 (Broman and Sen 2009). A logarithm of odd (LOD) score was calculated at a 1-cM step. Genome-wide 1, 5, and 10% significance threshold levels were obtained by 10,000 permutation tests of R/qtl. The 95% confidence interval of a QTL detected was estimated by a 1.8-LOD drop method. The percentage of phenotypic variance explained by the QTL and the additive and dominance effects of the QTL were calculated by the function fitqtl of R/qtl. The mode of inheritance of the QTL was estimated by the degree of dominance as previously described (Kenney-Hunt et al. 2006). A two-QTL genome scan was performed using the function scan two of R/qtl. Genome-wide 10% significance threshold levels for full and additive two-QTL models were obtained by 500 permutation tests of R/qtl.

RNA-seq analysis

Total RNA was extracted from diencephalons of the top and bottom three F2 individuals of each sex using Trizol reagent (Life Technologies, Japan) and the NucleoSpin RNA kit (Takara Bio, Otsu, Japan) according to the manufacturer's instructions. The diencephalon contains the hypothalamus, which contributes to the hypothalamic–pituitary–adrenal axis that controls stress and fear responses (Matteri et al. 2000). The concentration of the total RNA obtained was measured by the Quant-iT RNA Board-Range BR Assay Kit (Thermo Fisher Scientific, Tokyo, Japan) using a Qubit Fluorometer (Thermo Fisher Scientific, Tokyo, Japan). The three RNA samples were pooled by extreme groups per sex, and the two extreme pooled RNA samples of each sex were used for RNA-seq analysis. RNA-seq analysis and subsequent sequence data analysis were outsourced to Eurofins Genomics (Tokyo, Japan). Briefly, RNA-seq analysis was performed with the next-generation sequencer Illumina Hiseq4000 using pair-end sequencing of 101-bp lengths. Adapter sequences and low-quality reads were removed using the trimmomatic software version 0.36. The cleaned reads were aligned to the chicken RefSeq GRCg6a (https://www.ncbi.nlm.nih.gov/data-hub/genome/GCF_000002315.5/) using BWA software version 0.7.17 (Li and Durbin 2009). The read counts of each transcript were compared between top and bottom groups in each sex after normalization by the Trimmed Mean of M-values (TMM) method using edgeR software version 3.16.5 (Robinson et al. 2010). The genes with log2FC > 0.26 (>1.2-fold) and <−0.26 (<0.83-fold) in either sex and with the same expression direction in both sexes were considered to be up- and downregulated in the bottom group, respectively.

Quantitative real-time PCR analysis

Total RNA was extracted from diencephalons of NAG (n = 10), WL-G (n = 10), F1 (n = 10) chickens, and top F2 (n = 20) and bottom F2 (n = 19) chickens ranked by the first principal component score. cDNA was synthesized from 1.0 µg of total RNA using the PrimerScript RT reagent Kit with gDNA Eraser (Takara Bio, Otsu, Japan) according to the manufacturer's instructions. Quantitative real-time PCR analysis was conducted by an Applied Biosystems StepOnePlus Real-Time PCR system (Thermo Fisher Scientific, Tokyo, Japan) with TB Green Premix Ex Taq II (Tli RNaseH Plus) (Takara Bio, Otsu, Japan). The thermal conditions of real-time PCR were initial denaturation at 95°C for 30 seconds, 40 cycles of denaturation at 95°C for 5 seconds, annealing and extension at 60°C for 30 seconds, and additional extension for 60 seconds. To find appropriate internal control genes among the four genes Pol II, TBP, ACTB, and GAPDH, the amplification efficiency of the target and control genes was simultaneously calculated using a quantitative relative standard curve with four concentrations of serial dilutions (10, 2, 0.4, and 0.08 ng/µl). Pairs of target and control genes that showed similar amplification efficiency were analyzed using the 2−ΔΔCt method. Primer sequences are listed in Supplementary Table 1. All samples were analyzed in triplicate.

Before comparing the obtained gene expression levels among NAG, WL-G and F1 groups, the effects of group, sex, and their interaction on expression levels were tested using two-way analysis of variance (ANOVA) in JMP Pro software. Expression levels were adjusted only for sex, which was significant at P < 0.05 (Supplementary Table 2), using a linear model in JMP Pro software. Using adjusted expression levels for sex, the NAG, WL-G, and F1 groups were compared by one-way ANOVA followed by Tukey's honestly significant difference post hoc test in JMP Pro software. Similarly, the statistical significance of the effects of group, sex, and their interaction on expression levels were tested before comparison of expression levels between the top and bottom F2 groups. After adjusting the expression data for effects significant at P < 0.05 (Supplementary Table 3), adjusted expression levels between the two groups were compared by Student's t-test in JMP Pro software.

Haplotype frequency analysis

Diplotypes of the top and bottom F2 individuals were determined on the basis of genotypes of 16 SNP marker loci (see Table 1) within the 95% confidence interval of the above-mapped QTL. The marker loci were previously developed and simultaneously genotyped by RAD-seq analysis (Ishikawa et al. 2020). Haplotype frequencies were determined on the basis of diplotypes and compared between the two groups of top and bottom F2 individuals by Pearson's chi-square test. F2 individuals with recombinant haplotypes were excluded from the test.

Table 1.

A list of 35 genes that tend to be differentially expressed in the 95% confidence interval of the chromosome 4 QTL detected by RNA-seq analysis.

GenePosition (bp)Log2FCa
MaleFemale
ZBTB33 16,473,770–16,484,173−0.14−0.40
NDUFA1 16,493,404–16,494,1710.300.27
NKRF 16,532,474–16,542,4290.530.07
SLC25A43 16,562,212–16,583,458−0.52−0.23
PASD1 17,394,750–17,501,799−0.09−0.27
GPR50 17,602,861–17,634,184−2.03−0.76
MAMLD1 17,785,536–17,877,708−0.10−0.34
TMEM185A 18,045,407–18,057,0710.180.38
LOC112532367 18,108,763–18,120,4330.050.32
TLR2A 20,006,893–20,028,9560.570.40
SFRP2 20,063,343–20,067,714−0.30−0.75
LRAT 20,396,173–20,399,2240.810.33
NPY2R 20,537,583–20,547,5750.270.19
FAM198B 21,766,896–21,787,710−1.24−0.08
NAF1 23,312,484–23,342,3980.180.48
NPY5R 23,386,250–23,395,197−0.30−0.54
APELA 23,781,390–23,786,4790.600.36
TLL1 24,058,915–24,221,0460.030.34
PALLD 25,014,616–25,202,319−0.45−0.21
AADAT 25,220,146–25,235,2040.150.42
MFAP3L 25,261,123–25,283,188−0.28−0.02
NAA15 29,258,294–29,298,615−0.08−0.37
LOC422442 29,323,717–29,327,017−0.30−0.39
LOC112532278 29,766,125–29,771,905−0.42−0.11
LOC101749214 31,375,703–31,405,9980.130.28
SLC10A7 31,386,080–31,566,794−0.06−0.45
LOC107051782 32,317,753–32,365,112−0.40−0.73
MAB21L2 33,007,846–33,009,5730.360.40
SLC20A2 34,163,904–34,231,930−0.27−0.15
PLK4 34,340,512–34,358,347−0.31−0.16
MFSD8 34,356,519–34,369,982−0.09−0.32
JADE1 34,537,748–34,598,3980.330.13
BTC 34,616,757–34,622,3700.420.87
MMRN1 35,448,636–35,506,195−0.05−0.99
CCSER1 35,498,952–36,242,028−0.03−0.50

The physical map positions are based on the chicken RefSeq GRCg6a. SNP marker loci located on the 16 genes in bold were used for haplotype frequency analysis of top and bottom F2 individuals.

Fold change.

Correlation analysis

To eliminate the influence of diplotype, a conditional correlation analysis between gene expression levels and the first principal component scores of top and bottom 39 F2 individuals combined was performed using a linear model conditional on diplotype in JMP Pro software.

Sequence analysis

In two pooled F2 RNAs used for RNA-seq analysis, a synonymous SNP was found in the coding region of NPY5R, a candidate gene for the QTL, as described below. The SNP genotypes of NAG and WL-G breeds were determined by direct sequence analysis of PCR products amplified from genomic DNA and cDNA. Genomic DNA was extracted from the blood of one NAG male and four WL-G females used as direct grandparents of the F2 QTL mapping population (Ishikawa et al. 2020) and 10 F1 individuals (n = 5 of each sex) used for RT-qPCR analysis, using a DNeasy Blood and Tissue kit (Qiagen, Tokyo, Japan) according to the manufacturer's protocol. The 205-bp SNP region of genomic DNA and cDNA was amplified on an Applied Biosystems Veriti Thermal Cycler (Thermo Fisher Scientific, Tokyo, Japan) using a pair of primers (Supplementary Table 1) and Quick Taq HS DyeMix (Toyobo, Osaka). PCR conditions were 94°C for 2 minutes, followed by 40 cycles of 94°C for 30 seconds, 55°C for 30 seconds, and 68°C for 1 minute, and 68°C for 20 seconds. The PCR products were purified with a Wizard SV Gel and PCR Clean-Up System (Promega, Tokyo, Japan). Sequence analysis of the PCR products purified was outsourced to Eurofins Genomics (Tokyo, Japan). Both strands of the PCR products were determined on an Applied Biosystems 3730xl DNA Analyzer (Thermo Fisher Scientific, Tokyo, Japan) using a BigDy Terminator v3.1 Cycle Sequencing kit (Thermo Fisher Scientific, Tokyo, Japan).

Allelic imbalance analysis

Using the genomic DNA and cDNA sequence data of 10 F1 individuals obtained above, the relative peak heights of sequence waveforms were manually measured according to the method described by Ge et al. (2005) to examine the imbalance between the two SNP alleles of NPY5R expression. For both genomic DNA and cDNA, the ratio of relative peak heights of mutant to wild-type SNP alleles was calculated. Mean height ratios among the four groups by DNA and sex combination were compared with Kruskal–Wallis test followed by Steel–Dwass post hoc test in JMP Pro software because the group variances were not equal at P < 0.0001 (Welch test). In the absence of imbalance between the two alleles, the peak height ratio was expected to be 1. The mean peak height ratio of each of the genomic DNA and cDNA were compared with the expected value of 1 by Wilcoxon signed-rank test in JMP Pro software.

Results

Using 241 F2 individuals between WL-G and NAG, we integrated five analyses and performed them sequentially to identify strong candidate genes for the open-field QTL on chicken chromosome 4. The workflow diagram of the integrated approach is shown in Fig. 1.

An external file that holds a picture, illustration, etc. Object name is jkac316f1.jpg

Overview of an integrated approach for identification of candidate genes for the open-field QTL on chicken chromosome 4 using a segregating F2 population between WL-G and NAG chicken breeds.

Re-analysis of the QTL map location

To confirm the exact map location of the open-field QTL on chromosome 4, we performed principal component analysis, followed by QTL analysis, using only seven open-field traits affected by the QTL. Principal component analysis revealed that the first and second axes of principal components explained 90.2 and 6.2% of the total open-field variance, respectively (Supplementary Fig. 1). For the first principal component, resting time had a negative factor loading and the other six traits had positive factor loadings, indicating that F2 chicks with higher positive first principal component scores were more active in a novel open-field arena than those with lower negative first principal component scores. As shown in their movement trajectories shown in Fig. 2a, the top-ranking F2 individuals in the distribution of the first principal component scores were very active in the open-field arena and the bottom-ranking F2 individuals were inactive for both sexes. Only the first principal component scores that explained most of the total variance were used as a quantitative trait for genome-wide QTL analysis described below.

An external file that holds a picture, illustration, etc. Object name is jkac316f2.jpg

Results of QTL analysis for the first principal component scores for seven open-field traits in an F2 population between WL-G and NAG breeds: a) histograms showing the distribution of the first principal component scores by sex. The orange and blue bars indicate top and bottom individuals, respectively, that were used for gene expression analysis. Insets show movement trajectories of typical top and bottom individuals in the open-field arena. b) LOD curve plot on chromosome 4. The blue, red, and black dashed horizontal lines indicate genome-wide 1, 5, and 10% significance threshold levels, respectively, obtained by 10,000 permutation tests. The double-headed arrow shows the 95% confidence interval of the LOD peak. c) Effect plots (mean ± SEM) for the C04_016 and C04_023 SNP markers nearest the LOD peaks that exceeded 10 and 5% levels, respectively. The letters A and B represent alleles derived from WL-G and NAG breeds, respectively.

Using the first principal component scores and 881 SNP markers, simple interval mapping was performed on the 241 F2 individuals by the function scanone of R/qtl software. The LOD scores for genome-wide 1, 5, and 10% significance threshold levels were estimated to be 4.6, 3.9, and 3.6, respectively. A QTL with a peak LOD score of 4.2 exceeded the genome-wide 5% threshold level, and it was detected at 93 cM on chromosome 4 (Fig. 2b), at which the nearest SNP marker C04_023 (29.5 Mb) was located. The 95% confidence interval of the QTL was estimated to be 14–35 Mb. This QTL explained 7.7% of the phenotypic variance. The additive and dominance effects (mean ± SEM) in standard deviation units were 1.00 ± 0.23 and 0.02 ± 0.31, respectively. The degree of dominance (ratio of the dominance effect to the additive effect) was calculated to be 0.02, clearly showing an additive mode of inheritance of the QTL. The NAG-derived allele at the C04_023 marker locus increased the first principal component score (Fig. 2c).

An additional suggestive LOD peak (LOD = 3.6) exceeding the genome-wide 10% level was located at 83 cM near the C04_016 marker locus (24.7 Mb) on chromosome 4 (Fig. 2b). The NAG-derived allele at the C04_016 locus increased the first principal component score (Fig. 2c). To confirm the presence of the suggestive QTL, a two-QTL genome scan was performed by the function scan two of R/qtl. The genome-wide 10% significance threshold levels for the full and additive models of the two-QTL scan were estimated to be 8.7 and 4.3, respectively. No statistical evidence showing the presence of the additional suggestive QTL (LOD = 2.3 and 0.7) was obtained at the 10% threshold levels.

RNA-seq analysis

Based on the chicken RefSeq GRCg6a, 333 genes were present in the 95% confidence interval of the open-field QTL on chromosome 4. To find genes that tended to be differentially expressed among the 333 genes, RNA-seq analysis was performed using pooled diencephalic RNAs extracted from the top and bottom three F2 individuals ranked by the first principal component score in each sex. In total, 70,569,862 reads were obtained and approximately 78% of the reads were mapped to chicken RefSeq GCRg6a. RNA-seq analysis revealed that 35 genes tended to be differentially expressed in the 95% confidence interval (Table 1). Among the 35 genes, 15 tended to be upregulated and 20 tended to be downregulated in the bottom group. The LOC112532278 gene, which tended to be differentially expressed, was excluded from the next prioritization analysis because it completely overlapped on the physical map with the non-differentially expressed gene RNF150, making it difficult to design primers that accurately quantify the expression of LOC112532278 and RNF150 separately.

Quantitative real-time PCR analysis

To confirm the expression of the 34 genes (excluding LOC112532278), quantitative real-time PCR analysis was performed in NAG and WL-G and their F1 chicks. Two-way ANOVA revealed that 13 genes had significant sex effects on the expression levels, but no genes had significant breed-by-sex interaction effects at nominal P < 0.05 (Supplementary Table 2). After adjusting the expression levels of 13 genes for sex, it was found that 16 genes showed significant differential expression levels in NAG, WL-G, and F1 chicks (Table 2). One-way ANOVA revealed that the expression of three genes (TLL1, AADAT, and LOC101749214) was significantly upregulated in WL-G compared with that in NAG at P < 0.05 (Tukey's honestly significant difference test). Conversely, the expression of 10 genes (NDUFA1, SLC25A43, PASD1, TMEM185A, TLR2A, FAM198B, NPY5R, MFSD8, MMRN1, and CCSER1) was significantly downregulated in WL-G. For these 13 up- and downregulated genes, F1 expression was between the two parental breeds. However, NAA15, PLK4, and BTC expression levels were not significantly different between the two parental breeds. The F1 expression for PLK4 and BTC was significantly different from the expression in either parental breed. Uniquely, the NAA15 expression in F1 was significantly lower than that in both parental breeds. No significant differences in expression were found for the remaining 18 genes (Table 2).

Table 2.

Diencephalic expression levels of 34 genes quantified by quantitative real-time PCR analysis using NAG, WL-G, and their F1 chicks.

GeneNAGF1WL-G P value
ZBTB33 1 ± 0.050.99 ± 0.051.11 ± 0.050.24
NDUFA1 1 ± 0.07ª0.47 ± 0.07b0.11 ± 0.07c1.8 × 10−09
NKRF 1 ± 0.060.92 ± 0.080.78 ± 0.070.091
SLC25A43 1 ± 0.05a0.73 ± 0.05b0.56 ± 0.05b4.5 × 10−06
PASD1 1 ± 0.06ª0.93 ± 0.06ab0.79 ± 0.06b0.042
GPR50 1 ± 0.050.98 ± 0.051.01 ± 0.050.37
MAMLD1 1 ± 0.060.97 ± 0.061.08 ± 0.060.38
LOC112532367 1 ± 0.140.95 ± 0.140.76 ± 0.140.44
TMEM185A 1 ± 0.07ª0.73 ± 0.07b0.67 ± 0.07b0.0042
TLR2A 1 ± 0.06ª0.84 ± 0.06ab0.77 ± 0.06b0.024
SFRP2 1 ± 0.120.88 ± 0.120.73 ± 0.120.29
LRAT 1 ± 0.041.01 ± 0.050.93 ± 0.060.49
NPY2R 1 ± 0.060.96 ± 0.060.81 ± 0.060.077
FAM198B 1 ± 0.05ª0.63 ± 0.05b0.49 ± 0.05b1.3 × 10−07
NAF1 1 ± 0.061.00 ± 0.061.03 ± 0.060.93
NPY5R 1 ± 0.04ª0.79 ± 0.04b0.79 ± 0.04b3.9 × 10−04
APELA 1 ± 0.051.04 ± 0.050.95 ± 0.050.45
TLL1 1 ± 0.04a1.02 ± 0.04ab1.16 ± 0.04b0.021
PALLD 1 ± 0.050.89 ± 0.050.87 ± 0.050.19
AADAT 1 ± 0.06a1.22 ± 0.06b1.26 ± 0.06b0.015
MFAP3L 1 ± 0.070.91 ± 0.070.96 ± 0.070.62
NAA15 1 ± 0.04ª0.78 ± 0.04b0.92 ± 0.04a6.8 × 10−04
LOC422442 1 ± 0.080.94 ± 0.080.77 ± 0.080.11
LOC101749214 1 ± 0.04a1.11 ± 0.40ab2.43 ± 0.40b0.033
SLC10A7 1 ± 0.050.84 ± 0.050.92 ± 0.050.068
LOC107051782 1 ± 0.080.98 ± 0.080.80 ± 0.080.13
MAB21L2 1 ± 0.081.13 ± 0.080.87 ± 0.080.073
SLC20A2 1 ± 0.081.17 ± 0.080.96 ± 0.080.15
PLK4 1 ± 0.06ª0.76 ± 0.06ab0.79 ± 0.06b0.016
MFSD8 1 ± 0.07ª0.77 ± 0.07a0.50 ± 0.07b2.1 × 10−04
JADE1 1 ± 0.080.99 ± 0.080.78 ± 0.080.10
BTC 1 ± 0.10a1.49 ± 0.10b1.18 ± 0.10ab0.0037
MMRN1 1 ± 0.16a2.09 ± 0.16b3.00 ± 0.16c8.1 × 10−09
CCSER1 1 ± 0.03ª0.85 ± 0.03b0.84 ± 0.03b0.0014

Data (n = 10/breed) are presented as mean ± SEM. The data were adjusted for sex, when the effect of sex on the expression data was significant at nominal P < 0.05 (Supplementary Table 2). The P values were obtained by one-way ANOVA.

Means with different letters are significantly different between the two groups at P < 0.05 by Tukey's honestly significant difference test.

Haplotype frequency analysis

If causal gene(s) for the open-field QTL were involved in the 95% confidence interval, the frequencies of haplotypes derived from NAG and WL-G for the confidence interval should differ significantly between the top 20 and bottom 19 F2 chicks ranked by the first principal component score. Haplotype frequency analysis was performed for the 16 loci confirmed by real-time PCR analysis using the two extreme F2 groups after the removal of individuals with haplotype recombination. At 11 loci (FAM198B, NPY5R, TLL1, AADAT, NAA15, LOC101749214, PLK4, MFSD8, BTC, MMRN1, and CCSER1), the haplotype frequencies calculated from three diplotypes, NAG/NAG, NAG/WL, and WL/WL, were significantly different between the two extreme groups (Table 3). As expected, the frequency of the NAG haplotype was significantly higher in the top group than in the bottom group. At the other five loci, no significant differences in haplotype frequency were observed between the two groups (Table 3).

Table 3.

Haplotype frequencies calculated from three diplotypes at 16 loci between the top and bottom F2 groups.

GeneHaplotypeTopBottomChi-square value P value
NDUFA1 NAG0.62 (21)0.40 (12)3.020.082
WL-G0.38 (13)0.60 (18)
SLC25A43 NAG0.62 (21)0.40 (12)3.020.082
WL-G0.38 (13)0.60 (18)
PASD1 NAG0.62 (21)0.40 (12)3.020.082
WL-G0.38 (13)0.60 (18)
TLR2A NAG0.65 (22)0.42 (15)3.730.054
WL-G0.35 (12)0.58 (21)
TMEM185A NAG0.65 (22)0.42 (15)3.730.054
WL-G0.35 (12)0.58 (21)
FAM198B NAG0.72 (26)0.42 (15)6.850.0088
WL-G0.28 (10)0.58 (21)
NPY5R NAG0.72 (26)0.42 (15)6.850.0088
WL-G0.28 (10)0.58 (21)
TLL1 NAG0.72 (26)0.42 (15)6.850.0088
WL-G0.28 (10)0.58 (21)
AADAT NAG0.72 (26)0.44 (15)5.690.017
WL-G0.28 (10)0.56 (19)
NAA15 NAG0.76 (26)0.44 (14)7.390.0066
WL-G0.24 (8)0.56 (18)
LOC101749214 NAG0.74 (25)0.44 (16)6.100.014
WL-G0.26 (9)0.56 (20)
PLK4 NAG0.74 (25)0.44 (16)6.100.014
WL-G0.26 (9)0.56 (20)
MFSD8 NAG0.74 (25)0.44 (16)6.100.014
WL-G0.26 (9)0.56 (20)
BTC NAG0.74 (25)0.44 (16)6.100.014
WL-G0.26 (9)0.56 (20)
MMRN1 NAG0.68 (27)0.44 (16)4.100.043
WL-G0.33 (13)0.56 (20)
CCSER1 NAG0.68 (27)0.44 (16)4.100.043
WL-G0.33 (13)0.56 (20)

The number of haplotypes is shown in parentheses. Individuals with haplotype recombination were excluded from the analysis. The P values were obtained by Pearson's chi-square test (df = 1).

Association analyses

For the 11 genes that passed the above haplotype analysis, two association analyses were conducted using the top 20 and bottom 19 F2 individuals. One analysis was carried out to compare expression levels in the diencephalon between the two extreme F2 groups, while the other analysis was carried out to examine the conditional correlation of diplotypes between the expression levels and first principal component scores of the 39 extreme F2 individuals combined. In the former expression analysis, only LOC101749214 out of the 11 genes was significantly upregulated in the bottom group compared with that in the top group at P < 0.05 (Student's t-test; Table 4; Fig. 3c). NPY5R expression was marginally significant, with a trend toward lower expression in the bottom group than in the top group. For the other nine genes, no significant differences in expression were found between the two extreme groups (Table 4). In the latter conditional correlation analysis of diplotypes, among the 11 genes, only NPY5R showed a significant positive correlation between the expression levels and the first principal component scores (Fig. 3d). LOC101749214 expression showed a marginal negative correlation with the first principal component scores. For the other nine genes, no significant correlations were detected between gene expression and first principal component scores (Supplementary Fig. 2).

An external file that holds a picture, illustration, etc. Object name is jkac316f3.jpg

Results of four analyses for NPY5R and LOC101749214 genes: a) quantitative real-time PCR analysis in NAG, WL-G, and their F1 hybrids. Data are presented as ± SEM. P values were obtained by one-way ANOVA. The means with different letters (a and b) were significantly different between the two groups at P < 0.05 (Tukey's honestly significant difference test). b) Analysis of haplotype frequency between the top and bottom F2 groups. P values were obtained by the chi-square test. c) Quantitative real-time PCR analysis in the top and bottom F2 groups. P values were obtained by Student's t-test. d) A conditional correlation analysis of diplotypes between gene expression levels and the first principal component scores for top and bottom F2 individuals combined.

Table 4.

Diencephalic expression levels of 11 genes in the top and bottom F2 groups.

GeneTopBottom P value
FAM198B 1 ± 0.06 (18)0.88 ± 0.06 (18)0.18
NPY5R 1 ± 0.05 (18)0.87 ± 0.05 (18)0.060
TLL1 1 ± 0.06 (18)0.96 ± 0.06 (18)0.69
AADAT 1 ± 0.08 (18)1.10 ± 0.08 (17)0.39
NAA15 1 ± 0.09 (17)0.95 ± 0.09 (16)0.71
LOC101749214 1 ± 0.11 (17)1.37 ± 0.11 (18)0.028
PLK4 1 ± 0.10 (17)0.94 ± 0.09 (18)0.67
MFSD8 1 ± 0.07 (17)1.00 ± 0.07 (18)0.98
BTC 1 ± 0.09 (17)1.06 ± 0.09 (18)0.65
MMRN1 1 ± 0.11 (20)1.18 ± 0.12 (18)0.28
CCSER1 1 ± 0.04 (20)0.99 ± 0.04 (18)0.81

Data are presented as mean ± SEM. The number of individuals is shown in parentheses. The data were adjusted for group, sex, and/or sex-by-group interaction when their effects on expression were significant at nominal P < 0.05 (Supplementary Table 3). The P values were obtained by Student's t-test.

By the integrated analyses, the first 333 genes in the 95% confidence interval of the open-field QTL were successfully narrowed down to two genes of NPY5R and LOC101749214. The results of a series of analyses for the two genes are summarized in Fig. 3 and are overviewed in Fig. 1.

SNP analysis of candidate gene coding regions

Examination of the RNA-seq data in pooled F2 samples revealed no SNPs in the coding region of LOC101749214. However, there was only one synonymous SNP (T < C) at g: 23,392,923 bp on the NPY5R coding region. The top samples had the same codon GTT as the chicken RefSeq GRCg6a, encoding Valine. The bottom samples had the codon GTC encoding the same amino acid. Sequence analysis of genomic DNA from one NAG male and four WL-G females used as direct grandparents for the F2 population revealed that NAG was clearly homozygous for the reference allele T and WL-G for the mutant allele C (Fig. 4, a and b). This SNP was consistent with the previously reported SNP rs316511723.

An external file that holds a picture, illustration, etc. Object name is jkac316f4.jpg

Sequence results for the SNP rs316511723 region of the NPY5R gene: a) genomic DNA of four WL-G females (W#) with a clear homozygous genotype (C/C). b) Genomic DNA of a NAG Male chicken with a clear homozygous genotype (T/T). c) Genomic DNA of an F1 chicken obtained from a cross between the W3 WL-G female and the NAG male in (b). The F1 is a clear heterozygote (C/G). d) cDNA of the same F1 chicken. For all F1 chickens obtained from four WL-G females and the NAG male in (b) (Supplementary Fig. 3). Arrows show the nucleotide of the SNP.

Allelic imbalance analysis

Sequence analysis of the rs316511723 region of NPY5R was performed using genomic DNA and cDNA from 10 F1 individuals used for the real-time PCR analysis described earlier (Fig. 3a). Genomic DNA sequence analysis showed that all F1 individuals were clearly heterozygous for both C and T alleles, and these two alleles had nearly the same peak height (Fig. 4c and Supplementary Fig. 3a). On the other hand, cDNA sequence analysis showed that all 10 F1 individuals were heterozygous for both alleles, but the peak height of the T allele appeared to be slightly higher than that of the C allele (Fig. 4d and Supplementary Fig. 3b). Kruskal–Wallis test showed that the mean ratios of relative peak heights of the C allele to the T allele among the four groups by DNA and sex combination were significant at P < 0.001 (Table 5). In both sexes, the mean peak height ratio of cDNA was significantly lower than that of genomic DNA at P < 0.05 (Steel–Dwass test). The height ratio of both genomic DNA and cDNA did not differ significantly between the sexes (Table 5). However, there was no significant difference from the expected height ratio of 1, assuming an allelic ratio of 1:1, in any of the four groups (Wilcoxon signed-rank test).

Table 5.

Relative peak height ratios of the C and T alleles of the NPY5R SNP.

SexGenomic DNAcDNA
Male1.1 ± 0.02a0.8 ± 0.01b
Female1.1 ± 0.02a0.7 ± 0.04b

Data (n = 5/DNA/sex) are presented as mean ± SEM. The mean ratios among the four groups by DNA and sex combination were significant at P < 0.001 (Kruskal–Wallis test).

Means with different letters are significantly different between two of the four groups at P < 0.05 by Steel–Dwass test.

Discussion

In the present study, the location of the open-field QTL on chicken chromosome 4 was re-analyzed using the first principal component scores that summarized the phenotypic variances of the seven open-field traits affected by the QTL, confirming the existence of a significant QTL that was previously mapped for individual open-field traits (Ishikawa et al. 2020). The map location and the 95% confidence interval length of the QTL were the same in the present study and the previous study. Furthermore, the NAG-derived QTL allele was confirmed to unexpectedly reduce open-field fear, implying that this QTL is not a locus for the timid temperament seen in the NAG breed.

The two-QTL genome scan did not provide statistical evidence of an additional suggestive QTL that has the same sign of the phenotypic effect as the significant QTL described earlier. However, the possibility of the existence of an additional QTL cannot be ruled out as the two-QTL scan is known to be effective in separating two linked QTLs with opposing phenotypic effects (Broman and Sen 2009). The possibility might be supported by the detection of two candidate genes: NPY5R and LOC101749214.

Using individual RNA samples for RNA-seq analysis may yield more robust results than the results from pooled RNA samples used in this study. However, pooling RNA samples would reduce differences between individual RNA samples and would facilitate obtaining gene expression levels between the two extreme groups. To minimize as much as possible the loss of power to identify differences in gene expression, we used lower fold changes (>1.2-fold and <0.83-fold) as thresholds. In addition, we performed real-time PCR experiments using different populations (NAG, WL, and F1). In fact, RNA-seq studies using pooled hypothalamic RNA samples have been performed to discover differences in gene expression (e.g. Sharma et al. 2018; Chen et al. 2020; Adlanmerini et al. 2021; Mohr et al. 2021).

In the present study, one synonymous SNP (T < C) was found in the coding region of NPY5R. Allelic imbalance analysis did not provide clear evidence of differential expression of NPY5R between the T and C alleles. Even if allelic imbalance did exist, no biological effect would be observed because the SNP is synonymous with no amino acid change.

Although the biological function of LOC101749214 has not been reported, this gene is believed to overlap in the antisense direction with the ion transporter SLC10A7 at its 3′ end. SLC10A7 expression was marginally different among NAG, WL-G, and F1 chickens. On the other hand, NPY5R effects on fear and anxiety behaviors have been reported in rats. For example, NPY5R in the amygdala had an anxiolytic effect on social interaction behavior (Sajdyk et al. 2002). The infusion of the specific Y5 agonist [cPP]hPP into the paraventricular nucleus of the hypothalamus mediated a reduction in anxiety-related behavior in a light/dark box test (Martinetz et al. 2019). Rats that received an intracerebroventricular injection of the specific Y5 agonist [cPP(1–7), NPY(19–23), Ala(31), Aib(32), and Gln(34)]hPP exhibited anxiolytic-like behavior in an open-field test (Sørensen et al. 2004). These results of previous studies suggest that NPY signaling via NPY5R acts on regulating anxiety, and the results also indicate the difficulty in identifying the main brain regions involved in the control of anxiety-related behaviors. This difficulty may be true for chickens, since NPY5R has been reported to be widely expressed in various regions of the brain (He et al. 2016).

Transcriptional analysis in the hypothalamus of newly hatched chicks during fasting and delayed feeding suggested that increased NPY5R mRNA may be consistent with increased appetite in fasted chicks (Higgins et al. 2010). In chicks at 2 days of age, an intracerebroventricular injection of procine [Leu31, Pro34]-NPY, which has a high affinity to chicken NPY2R and NPY5R, significantly increased food intake in neonatal chicks (Ando et al. 2001), suggesting that NPY2R and NPY5R may regulate feeding behavior in chickens. In the present study, NPY5R expression in F2 chicks was positively correlated with the first principal component scores, and although F2 chicks were not fed from hatching until 1 day of age when the open-field test is conducted, there was no significant difference in body weights at 1 day of age, suggesting that F2 chicks with higher NPY5R expression may be more active foragers in the open-field arena.

In strategies to identify candidate genes, consideration is often given to preferential expression in tissues related to a trait of interest. However, this is not always true. In cows, DGAT1, a causal gene for milk yield and composition QTLs, is not preferentially expressed in tissues associated with milk production (Ron and Weller 2007). Therefore, NPY5R remains a valid candidate gene.

In general, two traditional approaches have been used to search for candidate genes for a QTL of interest (Ishikawa 2017). One approach is based on positional cloning, which aims to narrow the initial large QTL interval down to a very small interval with only one gene by the development of crossbred animals such as congenic strains or advanced intercross lines. In a previous study in mice using this approach, Usp46 was identified as a causal gene for immobile behavior obtained by tail suspension and forced swimming tests (Tomida et al. 2009). However, positional cloning is not applicable in all cases, as it is usually difficult to narrow down to a very small interval due to low recombination in the QTL interval. The other approach is the positional and functional gene approach in which a search is made for candidate genes based on information on phenotype-related genes for which functions and expression patterns are already known. Examples of this approach have been reported for bovine DGAT1 (Grisart et al. 2002), sheep GDF8 (Johnson et al. 2005), and pig IGF2 (Van Laere et al. 2003). Since this approach relies on known gene function, if the causal gene function is unknown, the true causal gene may be missed. The success stories for the two traditional approaches appear to consist of large-effect causal genes that explain 25–50% of the phenotypic variance. Compared with the two traditional approaches, our approach in the present study has the following advantages. The approach does not require narrowing down the QTL interval by further fine mapping using newly developed populations. The approach is not biased toward known gene function, in the sense that it was conducted without any prior consideration of gene function in the chick diencephalon; that is, it is a no-hypothesis approach. Furthermore, as demonstrated in the present study, the approach may be applicable to QTLs with small effects that explain less than 10% of the total phenotypic variance.

We believe that the most important aspect of our success was the use of the F2 segregating population for a series of analyses, which allowed us to exclude unlikely candidate genes that were included in the 95% confidence interval of the QTL at each analysis step (Fig. 1). First, the QTL location was re-analyzed using only the traits affected by the QTL, enabling us to accurately rank and select F2 individuals at the top and bottom extremes of the phenotypic distribution. Second, following RNA-seq analysis using two pooled RNAs of extreme individuals, quantitative real-time PCR 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. Third, the haplotype frequency analysis in the top and bottom F2 extremes allowed genes not associated with the phenotype to be filtered out, resulting in a further reduction of the number of candidate genes from 16 to 11. Finally, following an association analysis of gene expression between the two extreme groups, a conditional correlation analysis of diplotypes between gene expression and phenotype of the extreme individuals successfully revealed two strong candidate genes, NPY5R and LOC101749214. Recently, methylation QTL and expression QTL analyses were performed using the same advanced intercross population between White Leghorn and red junglefowl to map the effects of chicken domestication (Höglund et al. 2020). Thus, it was suggested that the same chicken population could be used to efficiently prioritize candidate genes. Currently, we plan to generate chickens with knockouts of the NPY5R and LOC101749214 genes, respectively, using a CRISPR/Cas9 system to verify the biological functions of the two genes.

Conclusion

By our integrated approach using an F2 segregating population between WL-G and NAG breed chickens, we successfully identified NPY5R and LOC101749214 genes as strong candidate genes for the chromosome 4 QTL affecting innate open-field behavior, without further fine mapping the QTL interval in detail. Our integrated approach may be useful in the search for candidate genes for QTLs with small effects that explain less than 10% of the phenotypic variance in other breeds of chickens and other animal species including model animals.

Supplementary Material

jkac316_Supplementary_Data

Acknowledgments

We are grateful to the members of our Graduate School of Nagoya University for their help in animal care.

Contributor Information

Takayuki Ochiai, Laboratory of Animal Genetics and Breeding, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan.

Marina Sakaguchi, Laboratory of Animal Genetics and Breeding, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan.

Shin-Ichi Kawakami, Laboratory of Animal Behavior and Physiology, Graduate School of Integrated Sciences for Life, Hiroshima University, Higashi-Hiroshima, Hiroshima 739-8528, Japan.

Akira Ishikawa, Laboratory of Animal Genetics and Breeding, Graduate School of Bioagricultural Sciences, Nagoya University, Chikusa-ku, Nagoya 464-8601, Japan.

Data availability

The RNA-seq data have been deposited in the DDBJ Sequence Read Archive under the accession number DRA010275 (https://ddbj.nig.ac.jp/resource/sra-submission/DRA010275; Direct FTP site, https://ddbj.nig.ac.jp/public/ddbj_database/dra/fastq/DRA010/DRA010275). The sequence data for the SNP region of the NPY5R gene in WL-G and NAG breeds have been deposited in DDBJ under the accession numbers LC735722 (http://getentry.ddbj.nig.ac.jp/getentry/na/LC735722/?Format=flatfile&filetype=html&trace=true&show_suppressed=false&limit=10) and LC735723 (http://getentry.ddbj.nig.ac.jp/getentry/na/LC735723/?format=flatfile&filetype=html&trace=true&show_suppressed=false&limit=10), respectively. The phenotypic, genotypic, and other data are available at https://doi.org/10.18999/2003797.

Supplemental material available at G3 online.

Funding

This study was supported in part by JSPS KAKENHI Grant Number 22H02497 to AI.

Author contributions

AI conceived, designed, and supervised the study. MS, SK, and AI sampled the brains. MS and AI prepared the other materials. TO conducted the experiments. TO and AI analyzed the results and wrote the draft manuscript. AI wrote the final version of the manuscript. All authors read and approved the manuscript.

Literature cited

  • Adlanmerini M, Nguyen HCB, Krusen BM, Teng CW, Geisler CE, Peed LC, Carpenter BJ, Hayes MR, Lazar MA. Hypothalamic REV-ERB nuclear receptors control diurnal food intake and leptin sensitivity in diet-induced obese mice. J Clin Invest. 2021;131(1):e140424. 10.1172/JCI140424. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Akbari P, Gilani A, Sosina O, Kosmicki JA, Khrimian L, Fang Y-Y, Persaud T, Garcia V, Sun D, Li A, et al. . Sequencing of 640,000 exomes identifies GPR75 variants associated with protection from obesity. Science. 2021;373(6550):eabf8683. 10.1126/science.abf8683. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Albert FW, Kruglyak L. The role of regulatory variation in complex traits and disease. Nat Rev Genet. 2015;16(4):197–212. 10.1038/nrg3891. [Abstract] [CrossRef] [Google Scholar]
  • Ando R, Kawakami SI, Bungo T, Ohgushi A, Takagi T, Denbow DM, Furuse M. Feeding responses to several neuropeptide Y receptor agonists in the neonatal chick. Eur J Pharmacol. 2001;427(1):53–59. 10.1016/s0014-2999(01)01201-8. [Abstract] [CrossRef] [Google Scholar]
  • Broman KW, Sen S. A Guide to QTL Mapping with R/qtl. New York (NY): Springer; 2009. [Google Scholar]
  • Buitenhuis AJ, Rodenburg TB, Siwek M, Cornelissen SJB, Nieuwland MGB, Crooijmans RPMA, Groenen MAM, Koene P, Bovenhuis H, van der Poel JJ. Identification of QTLs involved in open-field behavior in young and adult laying hens. Behav Genet. 2004;34(3):325–333. 10.1023/B:BEGE.0000017876.82142.73. [Abstract] [CrossRef] [Google Scholar]
  • Chen X, Wyler SC, Li L, Arnold AG, Wan R, Jia L, Landy MA, Lai HC, Xu P, Liu C. Comparative transcriptomic analyses of developing melanocortin neurons reveal new regulators for the anorexigenic neuron identity. J Neurosci. 2020;40(16):3165–3177. 10.1523/JNEUROSCI.0155-20.2020. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Daetwyler HD, Capitan A, Pausch H, Stothard P, van Binsbergen R, Brøndum RF, Liao X, Djari A, Rodriguez SC, Grohs C, et al. . Whole-genome sequencing of 234 bulls facilitates mapping of monogenic and complex traits in cattle. Nat Genet. 2014;46(8):858–865. 10.1038/ng.3034. [Abstract] [CrossRef] [Google Scholar]
  • Durbin RM, Altshuler D, Durbin RM, Abecasis GR, Bentley DR, Chakravarti A, Clark AG, Collins FS, De La Vega FM, Donnelly P, et al. . A map of human genome variation from population-scale sequencing. Nature. 2010;467(7319):1061–1073. 10.1038/nature09534. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Fogelholm J, Inkabi S, Höglund A, Abbey-Lee R, Johnsson M, Jensen P, Henriksen R, Wright D. Genetical genomics of tonic immobility in the chicken. Genes (Basel). 2019;10(5):341. 10.3390/genes10050341. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Ge B, Gurd S, Gaudin T, Dore C, Lepage P, Harmsen E, Hudson TJ, Pastinen T. Survey of allelic expression using EST mining. Genome Res. 2005;15(11):1584–1591. 10.1101/gr.4023805. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Grisart B, Coppieters W, Farnir F, Karim L, Ford C, Berzi P, Cambisano N, Mni M, Reid S, Simon P, et al. . Positional candidate cloning of a QTL in dairy cattle: identification of a missense mutation in the bovine DGAT1 gene with major effect on milk yield and composition. Genome Res. 2002;12(2):222–231. 10.1101/gr.224202. [Abstract] [CrossRef] [Google Scholar]
  • He C, Zhang J, Gao S, Meng F, Bu G, Li J, Wang Y. Molecular characterization of three NPY receptors (Y2, Y5 and Y7) in chickens: gene structure, tissue expression, promoter identification, and functional analysis. Gen Comp Endocrinol. 2016;236:24–34. 10.1016/j.ygcen.2016.04.019. [Abstract] [CrossRef] [Google Scholar]
  • Higgins SE, Ellestad LE, Trakooljul N, McCarthy F, Saliba J, Cogburn LA, Porter TE. Transcriptional and pathway analysis in the hypothalamus of newly hatched chicks during fasting and delayed feeding. BMC Genomics. 2010;11(1):162. 10.1186/1471-2164-11-162. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Höglund A, Henriksen R, Fogelholm J, Churcher AM, Guerrero-Bosagna CM, Martinez-Barrio A, Johnsson M, Jensen P, Wright D. The methylation landscape and its role in domestication and gene regulation in the chicken. Nat Ecol Evol. 2020;4(12):1713–1724. 10.1038/s41559-020-01310-1. [Abstract] [CrossRef] [Google Scholar]
  • Hu ZL, Park CA, Reecy JM. Building a livestock genetic and genomic information knowledgebase through integrative developments of animal QTLdb and CorrDB. Nucleic Acids Res. 2019;47(D1):D701–D710. 10.1093/nar/gky1084. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Ishikawa A. A strategy for identifying quantitative trait genes using gene expression analysis and causal analysis. Genes (Basel). 2017;8(12):347. 10.3390/genes8120347. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Ishikawa A, Sakaguchi M, Nagano AJ, Suzuki S. Genetic architecture of innate fear behavior in chickens. Behav Genet. 2020;50(6):411–422. 10.1007/s10519-020-10012-0. [Abstract] [CrossRef] [Google Scholar]
  • Johnson PL, McEwan JC, Dodds KG, Purchas RW, Blair HT. A directed search in the region of GDF8 for quantitative trait loci affecting carcass traits in Texel sheep. J Anim Sci. 2005;83(9):1988–2000. 10.2527/2005.8391988x. [Abstract] [CrossRef] [Google Scholar]
  • Johnsson M, Williams MJ, Jensen P, Wright D. Genetical genomics of behavior: a novel chicken genomic model for anxiety behavior. Genetics. 2016;202(1):327–340. 10.1534/genetics.115.179010. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Keane TM, Goodstadt L, Danecek P, White MA, Wong K, Yalcin B, Heger A, Agam A, Slater G, Goodson M, et al. . Mouse genomic variation and its effect on phenotypes and gene regulation. Nature. 2011;477(7364):289–294. 10.1038/nature10413. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Kenney-Hunt JP, Vaughn TT, Pletscher LS, Peripato A, Routman E, Cothran K, Durand D, Norgard E, Perel C, Cheverud JM. Quantitative trait loci for body size components in mice. Mamm Genome. 2006;17(6):526–537. 10.1007/s00335-005-0160-6. [Abstract] [CrossRef] [Google Scholar]
  • Li H, Durbin R. Fast and accurate short read alignment with Burrows–Wheeler Transform. Bioinformatics. 2009;25(14):1754–1760. 10.1093/bioinformatics/btp324. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Mackay T. The genetic architecture of quantitative traits. Annu Rev Genet. 2001;35(1):303–339. 10.1146/annurev.genet.35.102401.090633. [Abstract] [CrossRef] [Google Scholar]
  • Martinetz S, Meinung C-P, Jurek B, von Schack D, van den Burg EH, Slattery DA, Neumann ID. De novo protein synthesis mediated by the eukaryotic elongation factor 2 is required for the anxiolytic effect of oxytocin. Biol Psychiatry. 2019;85(10):802–811. 10.1016/j.biopsych.2019.01.010. [Abstract] [CrossRef] [Google Scholar]
  • Matteri R, Carroll J, Dyer C. Neuroendocrine responses to stress. In: Moberg GP, Mench JA, editors. The Biology of Animal Stress. Boston (MA): CABI Publishing; 2000. p. 43–76. [Google Scholar]
  • Mohr MA, Wong AM, Sukumar G, Dalgard CL, Hong W, Wu TJ, Wu YE, Micevych PE. RNA-sequencing of AVPV and ARH reveals vastly different temporal and transcriptomic responses to estradiol in the female rat hypothalamus. PLoS ONE. 2021;16(8):e0256148. 10.1371/journal.pone.0256148. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Robinson MD, McCarthy DJ, Smyth GK. Edger: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics. 2010;26(1):139–410. 10.1093/bioinformatics/btp616. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Ron M, Weller JI. From QTL to QTN identification in livestock–winning by points rather than knock-out: a review. Anim Genet. 2007;38(5):429–439. 10.1111/j.1365-2052.2007.01640.x. [Abstract] [CrossRef] [Google Scholar]
  • Sajdyk TJ, Schober DA, Gehlert DR. Neuropeptide Y receptor subtypes in the basolateral nucleus of the amygdala modulate anxiogenic responses in rats. Neuropharmacology. 2002;43(7):1165–1172. 10.1016/S0028-3908(02)00234-4. [Abstract] [CrossRef] [Google Scholar]
  • Sakaguchi M, Ishikawa A. Tonic immobility and open field responses in Nagoya, White Leghorn, and White Plymouth Rock chicks. J Poult Sci. 2020;57(3):183–191. 10.2141/jpsa.0190085. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
  • Schütz KE, Kerje S, Jacobsson L, Forkman B, Carlborg Ö, Andersson L, Jensen P. Major growth QTLs in fowl are related to fearful behavior: possible genetic links between fear responses and production traits in a Red Junglefowl × White Leghorn intercross. Behav Genet. 2004;34(1):121–130. 10.1023/B:BEGE.0000009481.98336.fc. [Abstract] [CrossRef] [Google Scholar]
  • Sharma A, Singh D, Das S, Kumar V. Hypothalamic and liver transcriptome from two crucial life-history stages in a migratory songbird. Exp Physiol. 2018;103(4):559–569. 10.1113/EP086831. [Abstract] [CrossRef] [Google Scholar]
  • Sørensen G, Lindberg C, Wörtwein G, Bolwig TG, Woldbye DP. Differential roles for neuropeptide Y Y1 and Y5 receptors in anxiety and sedation. J Neurosci Res. 2004;77(5):723–729. 10.1002/jnr.20200. [Abstract] [CrossRef] [Google Scholar]
  • Tomida S, Mamiya T, Sakamaki H, Miura M, Aosaki T, Masuda M, Niwa M, Kameyama T, Kobayashi J, Iwaki Y, et al. . Usp46 is a quantitative trait gene regulating mouse immobile behavior in the tail suspension and forced swimming tests. Nat Genet. 2009;41(6):688–695. 10.1038/ng.344. [Abstract] [CrossRef] [Google Scholar]
  • Van Laere AS, Nguyen M, Braunschweig M, Nezer C, Collette C, Moreau L, Archibald AL, Haley CS, Buys N, Tally M, et al. . A regulatory mutation in IGF2 causes a major QTL effect on muscle growth in the pig. Nature. 2003;425(6960):832–836. 10.1038/nature02064. [Abstract] [CrossRef] [Google Scholar]
  • Visscher PM, Wray NR, Zhang Q, Sklar P, McCarthy MI, Brown MA, Yang J. 10 Years of GWAS discovery: biology, function, and translation. Am J Hum Genet. 2017;101(1):5–22. 10.1016/j.ajhg.2017.06.005. [Europe PMC free article] [Abstract] [CrossRef] [Google Scholar]
Articles from G3: Genes | Genomes | Genetics are provided here courtesy of Oxford University Press

Full text links 

Read article at publisher's site: https://doi.org/10.1093/g3journal/jkac316

Read article for free, from open access legal sources, via Unpaywall: https://academic.oup.com/g3journal/article-pdf/13/2/jkac316/49148551/jkac316.pdfUnpaywall

Citations & impact 

Impact metrics

3
Citations
Jump to Citations

Citations of article over time

Alternative metrics

Altmetric item for https://www.altmetric.com/details/139603212
Altmetric
Discover the attention surrounding your research
https://www.altmetric.com/details/139603212

Article citations

Data 

Data behind the article

This data has been text mined from the article, or deposited into data resources.

BioStudies: supplemental material and supporting data

Similar Articles 

To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.

Funding 

Funders who supported this work.

JSPS (1)