Novel Y Chromosome Retrocopies in Canids Revealed through a Genome-Wide Association Study for Sex

The lack of an annotated reference sequence for the canine Y chromosome has limited evolutionary studies, as well as our understanding of the role of Y-linked sequences in phenotypes with a sex bias. In genome-wide association studies (GWASs), we observed spurious associations with autosomal SNPs when sex was unbalanced in case-control cohorts and hypothesized that a subset of SNPs mapped to autosomes are in fact sex-linked. Using the Illumina 230K CanineHD array in a GWAS for sex, we identified SNPs that amplify in both sexes but possess significant allele frequency differences between males and females. We found 48 SNPs mapping to 14 regions of eight autosomes and the X chromosome that are Y-linked, appearing heterozygous in males and monomorphic in females. Within these 14 regions are eight genes: three autosomal and five X-linked. We investigated the autosomal genes (MITF, PPP2CB, and WNK1) and determined that the SNPs are diverged nucleotides in retrocopies that have transposed to the Y chromosome. MITFY and WNK1Y are expressed and appeared recently in the Canidae lineage, whereas PPP2CBY represents a much older insertion with no evidence of expression in the dog. This work reveals novel canid Y chromosome sequences and provides evidence for gene transposition to the Y from autosomes and the X.