Common to all mammalian species, Mitochondrial DNA (mtDNA) is inherited exclusively from the mother. Mutations of this mtDNA occur, often in adaptation for environmental conditions. Sets of mutations shared by a large number of individuals may have an effect on athletic performance but are commonly used to define maternally-linked populations known as clades or haplogroups, and sub-populations known as haplotypes . This has made mtDNA a very useful tool for studying the evolution of, and classifying, various species including the horse.
Using mtDNA mutations as markers for familial segregation, Vila et al. 2001 first defined six haplogroups, A through F, in the equine mitochondrial genome. Following Vila's nomenclature, Jansen et al. 2002 expanded on these findings and published further D-Loop sequence motifs/markers for those six haplogroups plus a seventh, and motifs /markers for sub-haplogroups in clades A,B,C, and D.
More recently and in a significantly larger study using the complete equine mitochondrial genome for 83 different breeds of horses, Achilli et al. 2012 closely defined mutational motifs/markers of haplogroups and haplotypes in the equine breed, creating their own nomenclature, A through R, as well as a mtDNA reference sequence (Genbank:JN398377) for comparative study. Thus we now have a reference population to base the mtDNA of all Thoroughbred racehorses against using the full ~16,000 base of the mitochondrial genome.
Previous studies of mitochondrial DNA (mtDNA) in the Thoroughbred racehorse, mostly limited to the smaller d-loop section, not the complete genome, have identified mismatches between the pedigree records of the General Stud Book, the American Stud Book and maternally inherited mtDNA. Studies by Hill, et al 2002, Harrison and Turrion-Gomez in 2006, and Bower, et al (Bower et al. 2011, Bower et al. 2012a, Bower et al. 2012b) has shown that the mismatches are quite extensive.
So here is the current state of play in terms of the Haplogroups established by Achilli and Jensen, and the realities of thoroughbred classification as per Bruce Lowe Numbers.
You can see from the above that there are a lot of mismatches between the Lowe numbers and actual mtDNA haplogroups, including a couple of studies that have found that various members of sub-branches (such as 23b) appear in two different mtDNA haplogroups indicating a relatively modern error in stud book records. For those that continue to use Lowe numbers as a method of selection, there are more accurate and precise methods of classification to use.
There are two other unpublished papers, one by Allan Davie and his group out of Southern Cross University in Australia (co-authored with a Chinese group) and one by us that have further information than what is commonly known about Thoroughbred mtDNA. The paper by Davie, et al looked at the full sequences of mtDNA to see if there was a particular haplotype/group that outperformed relative to its presence in the population. Our paper qualified the #1 family and actually found that it was significantly more accurate than previously thought, the errors that Bower, et al found that split the #1 family between halpogroup L and N weren't as common as proposed (the 1-U branch is the only one with real errors) and the proposed link between the #1 family and the A1, A4 and 25 families isn't warranted.
A commonly held belief is that certain Lowe numbers are more profitable in terms of performance than others. The most likely reality is that if one mitochondrial haplotype had any true benefit over another, after 300 years we would have selected for it by now so it was homogeneous through the breed.
So what do the numbers say?
We created a list of the last 200 G1 winners around the world from part 1 countries and then we took all the yearlings offered at the Keeneland September, Tatts October, Inglis Easter and Magic Millions yearling sales, sorted their dams alphabetically and took the first 800 of them (a random but commercial population). Here is the result.
From the numbers above you can see that there are only a couple of haplotypes that might be considered to outperform their population representation. The best of these is the D haplogroup which is the 5b, 5h and 11b Lowe numbers - horses like Wise Dan, Fiorente, Obviously and Hana's Goal. Even then, we are talking about small numbers as far as their representation in the population is concerned.
One thing that is apparent when we consider other species (including humans) is that while the mitochondrial DNA itself may not have any value in determining athletic performance the relationship between the mitochondrial haplotype and the nuclear genes that are responsible for the process of mitochondrial biogenesis are. Thus, fixing for a particular mtDNA haplotype may prove a profitable process for the yearling buyer.