Genetic diversity, evolution and selection in the major histocompatibility complex DRB and DQB loci in the family Equidae

Twelve toll-like receptor (TLR) genes in the family Equidae - comparative genomics, selection and evolution

Toll-like receptors (TLRs) represent an important part of the innate immune system. While human and murine TLRs have been intensively studied, little is known about TLRs in non-model species. The order Perissodactyla comprises a variety of free-living and domesticated species exposed to different pathogens in different habitats and is therefore suitable for analyzing the diversity and evolution of immunity-related genes. We analyzed TLR genes in the order Perissodactyla with a focus on the family Equidae. Twelve TLRs were identified by bioinformatic analyses of online genomic resources; their sequences were confirmed in equids by genomic DNA re-sequencing of a panel of nine species. The expression of TLR11 and TLR12 was confirmed in the domestic horse by cDNA sequencing. Phylogenetic reconstruction of the TLR gene family in Perissodactyla identified six sub-families. TLR4 clustered together with TLR5; the TLR1-6-10 subfamily showed a high degree of sequence identity. The average estimated evolutionary divergence of all twelve TLRs studied was 0.3% among the Equidae; the most divergent CDS were those of Equus caballus and Equus hemionus kulan (1.34%) in the TLR3, and Equus africanus somaliensis and Equus quagga antiquorum (2.1%) in the TLR1 protein. In each TLR gene, there were haplotypes shared between equid species, most extensively in TLR3 and TLR9 CDS, and TLR6 amino acid sequence. All twelve TLR genes were under strong negative overall selection. Signatures of diversifying selection in specific codon sites were detected in all TLRs except TLR8. Differences in the selection patterns between virus-sensing and non-viral TLRs were observed.

Rare and common variant discovery by whole-genome sequencing of 101 Thoroughbred racehorses

The Thoroughbred breed was formed by crossing Oriental horse breeds and British native horses and is currently used in horseracing worldwide. In this study, we constructed a single-nucleotide variant (SNV) database using data from 101 Thoroughbred racehorses. Whole genome sequencing (WGS) revealed 11,570,312 and 602,756 SNVs in autosomal (1–31) and X chromosomes, respectively, yielding a total of 12,173,068 SNVs. About 6.9% of identified SNVs were rare variants observed only in one allele in 101 horses. The number of SNVs detected in individual horses ranged from 4.8 to 5.3 million. Individual horses had a maximum of 25,554 rare variants; several of these were functional variants, such as non-synonymous substitutions, start-gained, start-lost, stop-gained, and stop-lost variants. Therefore, these rare variants may affect differences in traits and phenotypes among individuals. When observing the distribution of rare variants among horses, one breeding stallion had a smaller number of rare variants compared to other horses, suggesting that the frequency of rare variants in the Japanese Thoroughbred population increases through breeding. In addition, our variant database may provide useful basic information for industrial applications, such as the detection of genetically modified racehorses in gene-doping control and pedigree-registration of racehorses using SNVs as markers.