Large-Scale Mitochondrial Genome Analysis of the Modern Horses: Perspectives on Their Domestication History

In spite of many global studies on the horse domestication, this topic remains controversial. To explore the domestication history of the horse, we analyzed complete mitochondrial genomes of 151 horses from 68 breeds, in which sequences from Mongolian (n = 2) and Jeju (n = 6) breeds were first determined in the present study. Here we used a Bayesian coalescent approach, and three different phylogenetic analysis methods, Bayesian inference, maximum likelihood, and neighbor-joining methods. We found that the domestic horse has multiple maternal origins, and that the Przewalski’s horse was not a direct ancestor of domestic horses. The horses were classified into one of the six major clades, and no apparent association was observed between breed or geographic location and heterogeneity of modern horses. The Mongolian breed was closely linked with four European (Andalusian, Oldenburg, Spotted Horse, and Trakhener) and one American (Bashkir Curly) breed, while the Jeju breed, which was not monophyletic, was closely related to four European (Icelandic Horse, Norwegian Fjord, Shetland, and Maremmano) and one Middle Eastern (Iranian) breed. Selection pressure analysis showed that the mitochondrial genome sequences were under purifying selection. Our Bayesian coalescent approach also showed that modern horses evolved at a rate of 1.487 × 10^–2 substitutions/site/My, and that their most recent common ancestor lived about 0.696 My ago. A Bayesian skyline plot revealed that the horse domestication events occurred approximately 2,000 - 3,500 years ago - during the Iron Age and that all modern horse lineages existed prior to domestication.