Thoroughbred horses are bred for competitive racing and undergo intense training regimes. The maintenance of physical soundness and desirable behavioural characteristics are critical to the longevity of a racing career. Horses intended for Flat racing generally enter training as yearlings and undergo introductory training prior to exercise conditioning for racing. This period requires rapid adjustment to a novel environment. As a prey animal, a horse's 'fight-or-flight' response is highly adapted, in which a well-understood component of this response, the hypothalamic-pituitary-axis, is activated in response to a stress stimulus, releasing cortisol. In the Thoroughbred, a significant difference in salivary cortisol concentrations between pre- and post-first time ridden (i.e., first backing) by a jockey have previously been identified. Here, to test the hypothesis that salivary cortisol concentrations may be used to objectively detect individual variations in the acute physiological stress response we investigate individual variation in cortisol response to training milestones. Saliva samples were collected from a cohort of n = 96 yearling Flat racehorses, at the same training yard, across three timepoints at rest: before entering the training yard (n = 66), within three days of entry to the training yard (n = 67) and following 2-3 weeks in the training yard (n = 50). Salivary cortisol concentration was measured using an ELISA. There was no significant difference in cortisol concentration (ANOVA, P > 0.05) across the samples collected at timepoints at rest. Samples were also collected before and 30 minutes after exposure to three novel training events: first time long-reined (n = 6), first time backed by a jockey (n = 34), and first time ridden on the gallops (n = 10). Mean salivary cortisol concentration after all three novel training events was significantly higher than prior to the training event (Paired t-test, P <0.005). The ranges of post-event salivary cortisol concentration across all timepoints suggest individual variation in the measured stress response, reflecting individual differences in stress response to the early training environment. This measure may be used as an objective assessment of the stress response of Thoroughbred racehorses during training.

Behavioural plasticity enables horses entering an exercise training programme to adapt with reduced stress. We characterised SNPs associated with behaviour in yearling Thoroughbred horses using genomics analyses for two phenotypes: (1) handler-assessed coping with early training events [coping] (n=96); and (2) variation in salivary cortisol concentration at the first backing event [cortisol] (n=34). Using RNA-seq derived gene expression data for amygdala and hippocampus tissues from n=2 Thoroughbred stallions, we refined the SNPs to those with functional relevance to behaviour by cross-referencing to the 500 most highly expressed genes in each tissue. The SNPs of high significance (q < 0.01) were in proximity to genes (coping - GABARAP, NDM, OAZ1, RPS15A, SPARCL1, VAMP2; cortisol - CEBPA, COA3, DUSP1, HNRNPH1, RACK1) with biological functions in social behaviour, autism spectrum disorder, suicide, stress-induced anxiety and depression, Alzheimer's disease, neurodevelopmental disorders, neuroinflammatory disease, fear-induced behaviours and alcohol and cocaine addiction. The strongest association (q=0.0002) was with NDN, a gene previously associated with temperament in cattle. This approach highlights functionally relevant genes in the behavioural adaptation of Thoroughbred horses that will contribute to the development of genetic markers to improve racehorse welfare.

The proportion of the genome containing runs of homozygosity (ROH) affects production traits in livestock populations. In European and Australasian Thoroughbreds inbreeding, quantified using ROH (FROH ), is associated with the probability of ever racing. Here, we measured FROH using 333 K SNP genotypes from 768 Thoroughbred horses born in North America to evaluate the effect of inbreeding on racing traits in that region. Among North American horses, FROH was not associated (p=0.518) with the probability of ever racing but was significantly associated with the number of race starts (p=0.002). Among raced horses, those with a 10% higher FROH than the mean inbreeding coefficient were predicted to have 3.5 fewer race starts compared to horses with a mean inbreeding coefficient. Considering the trend of increasing inbreeding and a decline in the average number of race starts per runner in North America, mitigating inbreeding in the population could positively influence racing durability.

BackgroundThousands of years of natural and artificial selection since the domestication of the horse has shaped the distinctive genomes of Chinese Mongolian horse populations. Consequently, genomic signatures of selection can provide insights into the human-mediated selection history of specific traits and evolutionary adaptation to diverse environments. Here, we used genome-wide SNPs from five distinct Chinese Mongolian horse populations to identify genomic regions under selection for the population-specific traits, gait, black coat colour, and hoof quality. Other global breeds were used to identify regional-specific signatures of selection.ResultsWe first identified the most significant selection peak for the Wushen horse in the region on ECA23 harbouring DMRT3, the major gene for gait. We detected selection signatures encompassing several genes in the Baicha Iron Hoof horse that represent good biological candidates for hoof health, including the CSPG4, PEAK1, EXPH5, WWP2 and HAS3 genes. In addition, an analysis of regional subgroups (Asian compared to European) identified a single locus on ECA3 containing the ZFPM1 gene that is a marker of selection for the major domestication event leading to the DOM2 horse clade.ConclusionsGenomic variation at these loci in the Baicha Iron Hoof may be leveraged in other horse populations to identify animals with superior hoof health or those at risk of hoof-related pathologies. The overlap between the selection signature in Asian horses with the DOM2 selection peak raises questions about the nature of horse domestication events, which may have involved a prehistoric clade other than DOM2 that has not yet been identified.

Selection for system-wide morphological, physiological, and metabolic adaptations has led to extreme athletic phenotypes among geographically diverse horse breeds. Here, we identify genes contributing to exercise adaptation in racehorses by applying genomics approaches for racing performance, an end-point athletic phenotype. Using an integrative genomics strategy to first combine population genomics results with skeletal muscle exercise and training transcriptomic data, followed by whole-genome resequencing of Asian horses, we identify protein-coding variants in genes of interest in galloping racehorse breeds (Arabian, Mongolian and Thoroughbred). A core set of genes, G6PC2, HDAC9, KTN1, MYLK2, NTM, SLC16A1 and SYNDIG1, with central roles in muscle, metabolism, and neurobiology, are key drivers of the racing phenotype. Although racing potential is a multifactorial trait, the genomic architecture shaping the common athletic phenotype in horse populations bred for racing provides evidence for the influence of protein-coding variants in fundamental exercise-relevant genes. Variation in these genes may therefore be exploited for genetic improvement of horse populations towards specific types of racing.

Abstract Background: Thousands of years of natural and artificial selection since the domestication of the horse has shaped the distinctive genomes of Chinese Mongolian horse populations. Consequently, genomic signatures of selection can provide insights into the human-mediated selection history of specific traits and evolutionary adaptation to diverse environments. Here, we used genome-wide SNPs from five distinct Chinese Mongolian horse populations to identify genes within genomic regions under selection for traits specific to Wushen (gait), Abaga Black (black coat colour) and Baicha Iron Hoof (hoof quality) horse, as well as other global breeds to identify regional-specific signatures of selection.Results: We first validated the use of the composite selection signals approach by identifying the most significant selection peak for the Wushen horse in the region on ECA23 harbouring DMRT3, the major gene for gait. We detected selection signatures encompassing several genes in the Baicha Iron Hoof horse that represent good biological candidates for hoof health, including the CSPG4, PEAK1, EXPH5, WWP2 and HAS3 genes. In addition, an analysis of regional subgroups (Asian compared to European) identified a single locus on ECA3 containing the ZFPM1 gene that is a marker of selection for the major domestication event leading to the DOM2 horse clade.Conclusions: Genomic variation at these loci in the Baicha Iron Hoof may be leveraged in other horse populations to identify animals with superior hoof health or those at risk of hoof-related pathologies. The overlap between the selection signature in Asian horses with the DOM2 selection peak raises questions about the nature of horse domestication events, which may have involved a prehistoric clade other than DOM2 that has not yet been identified.

Abstract BackgroundThoroughbred racehorse performance is largely influenced by a major quantitative trait locus at the myostatin (MSTN) gene which determines aptitude for certain race distances due to a promoter region insertion mutation influencing functional phenotypes in skeletal muscle. To develop an in vitro system for functional experiments we established three novel equine skeletal muscle cell lines reflecting the variation in phenotype associated with MSTN genotype (CC/II, CT/IN and TT/NN for SNP g.66493737C&gt;T/SINE insertion 227 bp polymorphism). Primary equine skeletal muscle myoblasts, isolated from Thoroughbred horse gluteus medius, were conditionally immortalised and evaluated to determine whether cell phenotype and metabolic function were comparable to functional characteristics previously reported for ex vivo skeletal muscle isolated from Thoroughbred horses with each genotype.ResultsPrimary myoblasts conditionally immortalized with the temperature sensitive SV40TtsA58 lentivirus vector successfully proliferated and could revert to their primary cell phenotype and differentiate into multinucleated myotubes. Skeletal muscle fibre type, MSTN gene expression, mitochondrial abundance, and mitochondrial function of the three MSTN genotype cell lines, were consistent with equivalent characterisation of ex vivo skeletal muscle samples with these genotypes. Furthermore, addition of coenzyme Q10 (CoQ10) to the cell lines improved mitochondrial function, an observation consistent with ex vivo skeletal muscle samples with these genotypes following supplementation with CoQ10 in the diet. ConclusionsThe observation that the phenotypic characteristics and metabolic function of the cells lines are equivalent to ex vivo skeletal muscle indicates that this in vitro system will enable efficient and cost-effective analyses of equine skeletal muscle for a range of different applications including understanding metabolic function, testing of nutritional supplements, drug test development and gene doping test development. In the multi-billion-euro international Thoroughbred horse industry research advances in the biological function of skeletal muscle are likely to have considerable impact. Furthermore, this novel genotype-specific system may be adapted and applied to human biomedicine to improve understanding of the effects of myostatin in human physiology and medicine.

Small effective population sizes and active inbreeding can lead to inbreeding depression due to deleterious recessive mutations exposed in the homozygous state. The Thoroughbred racehorse has low levels of population genetic diversity, but the effects of genomic inbreeding in the population are unknown. Here, we quantified inbreeding based on runs of homozygosity (ROH) using 297 K SNP genotypes from 6128 horses born in Europe and Australia, of which 13.2% were unraced. We show that a 10% increase in inbreeding (FROH) is associated with a 7% lower probability of ever racing. Moreover, a ROH-based genome-wide association study identified a haplotype on ECA14 which, in its homozygous state, is linked to a 32.1% lower predicted probability of ever racing, independent of FROH. The haplotype overlaps a candidate gene, EFNA5, that is highly expressed in cartilage tissue, which when damaged is one of the most common causes of catastrophic musculoskeletal injury in racehorses. Genomics-informed breeding aiming to reduce inbreeding depression and avoid damaging haplotype carrier matings will improve population health and racehorse welfare.

Abstract BackgroundThoroughbred racehorse performance is largely influenced by a major quantitative trait locus at the myostatin (MSTN) gene which determines aptitude for certain race distances due to a promoter region insertion mutation influencing functional phenotypes in skeletal muscle. To develop an in vitro system for functional experiments we established three novel equine skeletal muscle cell lines reflecting the variation in phenotype associated with MSTN genotype (CC/II, CT/IN and TT/NN for SNP g.66493737C&gt;T/SINE insertion 227 bp polymorphism). Primary equine skeletal muscle myoblasts, isolated from Thoroughbred horse gluteus medius, were conditionally immortalised and evaluated to determine whether cell phenotype and metabolic function were comparable to functional characteristics previously reported for ex vivo skeletal muscle isolated from Thoroughbred horses with each genotype.ResultsPrimary myoblasts conditionally immortalized with the temperature sensitive SV40TtsA58 lentivirus vector successfully proliferated and could revert to their primary cell phenotype and differentiate into multinucleated myotubes. Skeletal muscle fibre type, MSTN gene expression, mitochondrial abundance, and mitochondrial function of the three MSTN genotype cell lines, were consistent with equivalent characterisation of ex vivo skeletal muscle samples with these genotypes. Furthermore, addition of coenzyme Q10 (CoQ10) to the cell lines improved mitochondrial function, an observation consistent with ex vivo skeletal muscle samples with these genotypes following supplementation with CoQ10 in the diet. ConclusionsThe observation that the phenotypic characteristics and metabolic function of the cells lines are equivalent to ex vivo skeletal muscle indicates that this in vitro system will enable efficient and cost-effective analyses of equine skeletal muscle for a range of different applications including understanding metabolic function, testing of nutritional supplements, drug test development and gene doping test development. In the multi-billion-euro international Thoroughbred horse industry research advances in the biological function of skeletal muscle are likely to have considerable impact. Furthermore, this novel genotype-specific system may be adapted and applied to human biomedicine to improve understanding of the effects of myostatin in human physiology and medicine.

Although inspiratory muscle training (IMT) is reported to improve inspiratory muscle strength in humans little has been reported for horses. We tested the hypothesis that IMT would maintain and/or improve inspiratory muscle strength variables measured in Thoroughbreds during detraining. Thoroughbreds from one training yard were placed into a control (Con, n = 3 males n = 7 females; median age 2.2±0.4 years) or treatment group (Tr, n = 5 males, n = 5 females; median age 2.1±0.3 years) as they entered a detraining period at the end of the racing/training season. The Tr group underwent eight weeks of IMT twice a day, five days per week using custom-made training masks with resistance valves and an incremental threshold of breath-loading protocol. An inspiratory muscle strength test to fatigue using an incremental threshold of breath-loading was performed in duplicate before (T0) and after four (T1) and eight weeks (T2) of IMT/no IMT using a custom-made testing mask and a commercial testing device. Inspiratory measurements included the total number of breaths achieved during the test, average load, peak power, peak volume, peak flow, energy and the mean peak inspiratory muscle strength index (IMSi). Data were analysed using a linear mixed effects model, P≤0.05 significant. There were no differences for inspiratory measurements between groups at T0. Compared to T0, the total number of breaths achieved (P = 0.02), load (P = 0.003) and IMSi (P = 0.01) at T2 had decreased for the Con group while the total number of breaths achieved (P<0.001), load (P = 0.03), volume (P = 0.004), flow (P = 0.006), energy (P = 0.01) and IMSi (P = 0.002) had increased for the Tr group. At T2 the total number of breaths achieved (P<0.0001), load (P<0.0001), volume (P = 0.02), energy (P = 0.03) and IMSi (P<0.0001) were greater for the Tr than Con group. In conclusion, our results support that IMT can maintain and/or increase aspects of inspiratory muscle strength for horses in a detraining programme.