Exploring genomes of domestic animals using Next Generation Sequencing

Domestic animals represent a diverse group of populations comprising millions of genetic variants in their genomes. In particular, due to their extensive diversity among breeds and high similarity within breeds, they have been shown to be a unique model for investigations of phenotypic variations. Whole genome sequencing of highly selected horse breeds and non-breed horses kept under free range conditions revealed an average number of 3.4 million SNVs and 0.9 million indels, which could not be found in public databases. Private variants exclusively identified in non-breed horses indicated a natural selection of these horses for metabolic processes and morphogenesis, whereas breed horses harbored private variants in performance-associated genes affecting muscle contraction, neurological processes and ion/cation transport. Comparative analysis with bead chip genotyping data confirmed a high detection accuracy of next generation sequencing data of more than 95%. Further investigations of horse breeds and non-breeds for ROH regions indicating potential signatures of selection revealed an average number of 3492 ROHs in windows of a minimum of 50 consecutive homozygous SNPs and 292 ROHs in windows of a minimum of 500 consecutive homozygous SNPs. Only three ROHs were found to be common in whole genome sequencing data of all investigated 10 individuals, harboring the KITLG gene. This gave evidence for a positive natural and artificial selection of breeds as well as non-breeds for reproduction traits. Filtering of whole genome data from 11 horses for high-impact variants in genes involved in male reproductive processes resulted in 17 variants with no homozygous mutant genotype in investigated fertile stallions. After validation in further 337 fertile Hanoverian stallions, 9 high-impact variants remained and were considered as potentially deleterious factors for stallion fertility. Furthermore, a splice-site disruption variant in NOTCH1 was shown to be associated with the de-regressed breeding values of the paternal component of the pregnancy rate per estrus (EBV-PAT). Thus, this variant was proposed to be a significant locus for Hanoverian stallion reproduction, highly affected by targeted selection for stallion fertility. Genomic changes due to artificial selection for conformational parameters were further investigated Shetland ponies. Potential signatures of selection for extremely small body size were identified in both miniature and standard-sized ponies. NGS data revealed four synergistically interacting variants in ROH regions, which were proposed to limit the height at the withers in Shetland ponies to 87 cm (34.25 inches). We found evidence that these variants were Shetland pony-specific variants, providing a miniature size under specific genotypic combinations. Potential signatures of selection were also identified in Lundehund dogs, investigated for the Lundehund syndrome. A missense mutation was found in whole genome sequencing data of affected Lundehund dogs in a region 2 Mb proximal to a ROH. The affected gene LEPREL1 was estimated to be a precursor of inflammatory effects promoting the Lundehund syndrome in this breed. Similar observations were made in Shar-Pei dogs. Analyses for potential signatures of selection resulted in a large ROH region exclusively identified in whole genome sequencing data of two SPAID-affected Shar-Pei. A missense variant located in this region in MTBP with a predicted damaging effect was shown to be highly associated with SPAID, whereas a significant association with a previously identified CNV near HAS2 could not be confirmed. Affected dogs revealed a characteristic Shar-Pei fever as well as other signs of inflammation including arthritis and dermatitis. Further scans of NGS data for potential candidate mutations, successfully detected disease traits and desired phenotypes in horses and cattle. We identified a missense mutation in ACAN in a Miniature Shetland pony with a dwarf phenotype suggesting this variant as a novel mutation for a so far not observed dwarfism phenotype associated with malformations in Shetland ponies. Further studies on coat characteristics in horses revealed two variants associated with curly hair in KRT25 and SP6. Mutant KRT25 allele was found to be masking SP6 effect, resulting in curly hair accompanied with hypotrichosis. This gave evidence for an epistatic effect of KRT25 on SP6 for the development of curly hair and hypotrichosis in horses. In cattle, whole genome sequencing data of a curly-coated German Angus compared to 51 cattle from seven different breeds resulted in a causative variant in KRT27, which has also been observed in curly-coated German Fleckvieh, suggesting a transmission of this variant from German Fleckvieh to German Angus. The results of this work give an insight into the wide spectrum of genome analyses using NGS data and elucidate various genetic effects on the development of domestic animals identified as genomic footprints of selection.