Genetic variation in the chicken genome: insights in selection

The chicken currently provides more than a quarter of the meat and nearly all eggs produced worldwide. For future improvements in production traits and animal welfare as well as to address future consumer demands it is necessary to understand the etiology and biology underlying production traits and diseases. The primary aim of the research described in this thesis was to investigate the utility of several molecular approaches to identify causative variants underlying a variety of traits in the chicken.
The general introduction in chapter 1 provides an overview of the domestication history of the chicken - with a particular focus on commercial chicken breeds - and describes the importance to identify causative variants underlying production traits and diseases. Furthermore, several different molecular techniques and methods are introduced that are being used to detect causative variants underlying monogenic and polygenic traits.
Linkage maps are essential for linkage analysis, important to study recombination rates and recombination hotspots within the genome and can assist in the sequence assembly of genomes. In chapter 2 we describe the construction of a new high-resolution linkage map of the chicken genome based on two chicken populations with a total of 1619 individuals. The two populations used are a purebred broiler line and a broiler x broiler cross. This high resolution allowed accurate identification of recombination hotspots in the chicken genome, including sex specific recombination. Furthermore, to improve the current reference genome (WASHUC2), 613 unmapped markers were included in the genome-wide assay that included a total of 17,790 SNPs. The resulting linkage map comprises 13,340 SNPs, of which 360 had not been assigned to a known chromosome on chicken genome build WASHUC2. The resulting linkage map is composed of 31 linkage groups, with a total length of 3,054 cM for the sex-average map of the combined population. Regional differences in recombination hotspots between the two mapping populations were observed for several chromosomes near the telomere of the p arm. The sex-specific analysis revealed that these regional differences were mainly caused by female-specific recombination hotspots in the broiler × broiler cross.
In chapter 3 we describe the molecular characterization of the locus causing the late feathering phenotype; a monogenic trait in chicken that results in a delayed emergence of flight feathers at hatch. The late feathering phenotype is beneficial to breeders as it can be used for sex typing at hatch. The locus has, therefore, been extensively used in diverse commercial chicken breeds. However, a retrovirus closely linked to the late feathering allele causes a negative pleiotropic effect on egg production and causes viral infections. Within this chapter we describe the identification of a 180 kb tandem duplication in the late feathering allele using a quantitative PCR approach. The tandem duplication results in the partial duplication of two genes; the prolactin receptor and the gene encoding sperm flagellar protein 2. Sequence analysis revealed that the duplication is identical in broiler, white egg-layer, and brown egg-layer lines. This information was also used to design a molecular test to detect this duplication, particularly in heterozygous individuals.
The recent advances in massive parallel sequencing technologies have enabled rapid and cost-effective detection of all genetic variants within genomes. The detection of all genetic variants within a genome has further increased our ability to identify causative variants underlying quantitative trait loci (QTL). In chapter 4, we combined a genome-wide association study with whole-genome resequencing to identify causative variants underlying the pulmonary hypertension syndrome (PHS), a polygenic trait in chicken. PHS is a metabolic disease that has been linked to intense selection on growth rate and feed conversion ratio of modern broilers (meat-type chicken). PHS has become one of the most frequent causes of mortality within the broiler industry and leads to substantial economic losses and reduced animal welfare. In total, 18 QTL regions were identified in the genome-wide association study. In order to detect causative variants underlying these QTL regions, we sequenced the genomes of twelve individuals. To maximize the detection of causative variants we selected the individuals based on extreme phenotypes for PHS. Within 8 QTL regions we identified a total of 10 genes that contain at least one variant that is predicted to affect protein function. Moreover, 7.62 million SNPs were detected within the twelve animals compared to the reference genome. These markers can be used in the development of future genome-wide assays.
Genomic regions that have undergone selection should contain loci that influence important phenotypic traits and will, therefore, include causative variant(s) that could aid in further future improvement of production traits and disease resistance. In chapter 5, we applied hitch-hiking mapping to make a broad assessment of the effects of selection histories in domesticated chicken. Towards this end, we sampled commercial chickens representing all major breeding goals from multiple breeding companies. In addition, we sampled non-commercial chicken diversity by sampling almost all recognized traditional Dutch breeds and a representative sample of breeds from China. The broad sample of 67 commercial and non-commercial breeds were assessed for signatures of selection in the genome using information of 57,636 SNPs that were genotyped on pooled DNA samples. Our approach demonstrates the strength of including many different populations with similar, and breed groups with different selection histories to reduce stochastic effects based on single populations. The detection of regions of putative selection resulted in the identification of several candidate genes that could aid in further improvement of production traits and disease resistance.
Finally, the general discussion in chapter 6 describes the main findings of this thesis. In this chapter recommendations are given for the best strategies to detect causative variants underlying monogenic or polygenic traits. All strategies can benefit substantially from the recent developments in massive parallel sequencing, although the high costs of this method currently prevent large scale studies. In order to perform powerful and cost-effective studies, several strategies are discussed that combine massive parallel sequencing with other existing methods and techniques. Furthermore, the limitations of the different strategies are addressed, as well as the improvements needed in the near future to identify causative variants underlying a variety of traits in, but not limited to, the chicken.