Investigating the genetic and genomic basis of osteochondrosis in Thoroughbred horses from Australia and New Zealand

Abstract

Osteochondrosis (OC) is a skeletal disorder that occurs in young, growing animals. It is defined as a focal disturbance of endochondral ossification, visible as small, discrete areas of abnormal bone and/or cartilage tissue occurring at typical sites on the articular surfaces of joints. In the Australian and New Zealand Thoroughbred horse populations, OC is one of the most common skeletal lesions reported to occur in young horses. The presence of OC lesions can negatively affect a horse's welfare, monetary value, and athletic performance. Approximately 1,300 diagnostic reports written to assist stud managers in their sales process were used as a data source to determine the prevalence of OC and other skeletal lesions in a population of Australasian Thoroughbred weanlings and yearlings. The prevalence and distribution of skeletal lesions and injuries described in these reports were consistent with those found in studies of related populations. Overall, 20.5% of yearlings were reported as having OC, and the most common sites for OC lesions were the lateral trochlear ridge of the distal femur, the medial femoral condyle, and the sagittal ridge of the third metacarpal bone. Based on these data, analyses were carried out to determine the extent to which non-genetic factors contribute to the prevalence of OC and other skeletal lesions in Australasian Thoroughbred yearlings; to estimate the heritability of OC and other skeletal lesions overall and at particular anatomical sites; and to estimate phenotypic and genetic correlations between pairs of skeletal lesion traits within and between joints. Correlations between Estimated Breeding Values (EBVs, the estimated sum of additive gene effects for each horse) were used as a proxy for conventional genetic correlation, which can not be determined on the underlying scale with currently available software for binary traits such as those in the current study. Non-genetic factors were found to contribute to the occurrence of some skeletal lesion traits in this population, but the effect of these factors was not consistent between traits. Breeding values were found to contribute significantly to the occurrence of OC, some OC component traits, and bone chip(s) or fragment(s) (FRAG) occurring proximal palmar/plantar to the first phalanx (PPP1) in the hind fetlocks. Heritability estimates for these traits ranged from 0.10 to 0.22. Not all OC traits had positive phenotypic or EBV correlations with one another. However, positive EBV correlations were found within a group of traits including stifle OC lesions and FRAG occurring PPP1 in the hind fetlocks. This group included multiple traits that were among the most prevalent in this population, that are known to have negative impacts on the financial value and/or race performance of affected horses. It appears to be a good potential target for genetic selection. Single Nucleotide Polymorphism (SNP)-based case-control genome-wide association studies (GWAS) were carried out for 11 OC traits, FRAG occurring PPP1 in the hind fetlocks, and for the chestnut coat colour as a positive control using the Illumina Equine SNP50 beadchip, in a group of 140 horses with skeletal lesion data from the current study. A check for population stratification identified one large cluster comprising the majority of the population and two small outlier clusters, each comprising the offspring of a single Australian-born Thoroughbred stallion. Despite only a small number of cases being available for these analyses, genome-wide significant quantitative trait loci were found on chromosome 30 for lysis at the sagittal ridge of the third metacarpal bones in the fore fetlocks (within the large cluster only), and on chromosome 3 for the positive control chestnut coat colour (within all three clusters of this population, and within the large cluster only). A pedigree analysis of the Australian Thoroughbred population was carried out using pedigree data provided by the Australian Stud Book. The impact of past changes in breeding practice, including changes in sire usage and the origins of imported breeding stock, was examined via trends in the rate of inbreeding and loss of genetic variability due to unequal use of founders, population bottlenecks and genetic drift over time. There has been a low rate of loss of genetic variability in the Australian Thoroughbred population since 1973. This rate of loss is now increasing and is likely to increase further in coming decades. The importation of breeding stock from traditional sources (New Zealand, Europe and North America) is no longer increasing genetic variability. The number of sires is decreasing and their co-ancestry is increasing. Selective breeding could be used to reduce the occurrence of many OC traits and FRAG occurring PPP1 in the hind fetlocks. The group of genetically associated traits that includes stifle OC and FRAG occurring PPP1 in the hind fetlocks appears to be a particularly good target for genetic selection, where minimisation of these lesions would result in financial benefit to breeders as well as improving the welfare of the horses. In recent decades, selective breeding techniques in first world agricultural animal species have generally implemented selection based on EBVs. More recently, techniques have been developed to include SNP data in the calculation of EBVs, resulting in Genomic Estimated Breeding Values (GEBVs) that have the potential to be substantially more accurate than EBVs. The data required in order to implement genetic selection in this population is easily available, but the creation and maintenance of such a program would require ongoing financial investment. This investment could come from either the industry as a whole, or individual breeders who are open to embracing genetic and/or genomic technologies that are new to the Thoroughbred horse industry. There is also the potential to extend any genetic selection program to include selection for particular athletic traits, or selection against other disorders with a genetic component.