This loss and imbalance of ERAP haplotype diversity

This loss and imbalance of ERAP1/2 haplotype SM-164 cost among dog breeds mirrored the situation observed for DLA class I and II haplotypes (Pedersen et al., 2015a,b). A diminution in the repertoire of ERAP1 and ERAP2 haplotypes, especially when coupled with similar diminutions in DLA class I polymorphisms, would be counter to normal evolution. Maintaining the total ancestral diversity of ERAPs and MHC class I receptors is accomplished by a process known as balancing selection (Garrigan and Hedrick, 2003). Andrés et al. (2009) concluded “that although balancing selection may not have an obvious impact on a large proportion of human genes; it is a key force in affecting the evolution of a number of genes in humans.” Under normal circumstances, a new allele that favors survival would be expected to dominate over the less fit ancestral allele. Balancing selection prevents this by what is known as heterozygote advantage (Sellis et al., 2011). Natural selection in the MHC is largely pathogen driven; the objective being temporary expansion of certain receptor haplotypes rather than permanent selection of one haplotype over another. Although balancing selection has not been the most pervasive force in shaping the human genome, it has played an essential role in maintaining polymorphisms that are critical for species survival in whatever form evolution will dictate. There was evidence for balancing selection among ERAP1 and ERAP2 haplotypes when looking at all dogs as a whole. However, the positive selection forces created by breed development, which has been most intense in the last 150 years or so, has negated balancing selection forces and limited the number and specific form of ERAP1 and ERAP2 haplotypes inherited by descent in each breed of dogs. This non-balancing selection within the ERAPs and MHC class I receptors might help to explain differences in susceptibility to cancers, infectious disease, and autoimmune disorders between different pure breeds of dogs. Whether this is indeed the case remains to be determined by further studies, but the necessary tests are now available.
The original goals for the study did not include canid evolution. However, the dog ERAP1 and ERAP2 haplotypes provided a large and highly polymorphic region of the genome, which has been under strong linkage disequilibrium and subjected to balanced selection until recently. Three major ERAP1/2 clades were identified among all of the dogs tested, which mirrored the three major population clades derived from mitochondrial sequences (Vilà et al., 1997Savolainen et al., 2002). Other genetic evidence indicates that dogs descended from at least 51 female Gray wolf founders (Webb and Allard, 2010), and that all dog breeds descend from 67 paternal lineages (Bannasch et al., 2005). These figures are in range to the 45 most common ERAP1/2 haplotypes that occur in the phylogenetic tree of the dogs that were tested.
Although the contemporary Gray wolf is presumed to share ancestry with the modern dog, the paucity of wolf specific or wolf/dog shared haplotypes in the genetic tree of present day dogs requires comment. Dog specific haplotypes may have predominated because the SNP polymorphisms used for testing were of dog origin. However, many dog ERAP1 and ERAP2 SNPs and SNP haplotypes were both shared and unique among various Caninae. The paucity of wolf-specific ERAP1/2 haplotypes may also be a reflection of the estimated 16-fold reduction in effective population size of wolves beginning soon after their genetic divergence from dogs (Freedman et al., 2014). The larger presence of coyote (five) or coyote/dog (six) haplotypes compared to wolf (three) or wolf/dog (four) haplotypes in the dog genetic tree also suggests a genetic contribution of coyote-like canids to dog ancestry, evoking an ancestral species less like modern Gray wolves than assumed by sharing of mitochondrial DNA (Vilà et al., 1997). At least one recent study suggests that present-day wolves are genetically monophyletic from indigenous dogs in the same regions and that the true nearest wild ancestor of dogs may be extinct (Freedman et al., 2014).