Most of the time I get fairly bored by newly released scientific papers. Many are just re-hashes of old work with a different angle while occasionally I see one that I am stunned that it got past the peer review process (like this paper here where they failed to validate their hypothesis in an independent sample set, but amazingly still got the paper published). This week however, a new scientific paper was published that has really piqued my interest.
The paper, which you can read here, comes out of the Animal Health Trust in England and is the first paper that has identified changes in genetic code (SNPs) associated with increase fracture risk in thoroughbreds.
We have previously discussed how scientists have taken a look at the soundness of thoroughbreds on a statistical/heritable basis. This study specifically looks at the genetic basis of soundness. The study had 269 fracture cases which were horses that sustained catastrophic distal limb fractures while racing on UK racecourses, necessitating euthanasia. The 253 control horses were over 4 years of age, were racing during the same time period as the cases, and had no history of fracture at the time the study was carried out.
43,417 SNPs were employed to perform a genome-wide association study and significant genetic variation associated with fracture risk was found on chromosomes 9, 18, 22 and 31. Three SNPs on chromosome 18 and one SNP on chromosome 1 reached genome-wide significance (p < 0.05). Examination of the linkage disequilibrium (LD) among SNPs showed that the most significant SNPs on chromosome 18 fall into an LD block containing 10 SNPs in total.
The 10 SNP haplotype GGAGGCTAAA on chromosome 18 appeared in 73.8% of the cases (unsound) and 85.6% of the control (sound) horses and has a protective effect, with logistic regression showing that controls are 1.95 times at less risk of fracture than cases. Alternatively TGGAATTAAG, a risk haplotype, is at low frequency in the cases (just 2.9%) and, at least in their data set, absent from the controls. While it is rare in the population, cases with this haplotype are at 3.39 times higher risk of fracture than controls.
Haplotype block 1 (located between 62.01 – 62.15 Mb on chromosome 18) contains the two most significant SNPs from the genome-wide association study in BIEC2-416680 and BIEC2-416681. The third SNP that reached statistical significance was BIEC2-417495 which is located in haplotype block 4 (67.18 – 67.20 Mb on ECA 18). Interstingly this latter SNP is in linkage disequilibrium (r2 = 0.8) with the myostatin (MSTN) gene, which is associated with optimal racing distance and performance in thoroughbreds. There is however only moderate LD (r2 < 0.3) between BIEC2-417495 and the two SNPs in haplotype block 1 which are significantly associated with catastrophic fracture risk.
Obviously you'd like to see other peer reviewed papers corroborate their findings especially in the North American breeding population where soundness is a real issue (is anyone at the Grayson-Jockey Club Research Foundation reading this???). While there was low LD between the high risk SNPs and the myostatin gene, it would be interesting to see if variations within myostatin meant higher risk of fracture for those with the TGGAATTAAG haplotype. It is quite possible that an antagonistic effect is at work here that limits the improvement of the breed (i.e while we select for speed, we select also for higher fracture risk). This study also does show that while fracture risk in the Thoroughbred horse is a complex condition that includes training regimes, track surfaces and other variables, it has an underlying genetic basis to it.