Abstract
A prior genome-wide association (GWA; N = 200) including Thoroughbreds and stock horses implicated chromosome 6 (NC_009149.3) in owner-reported equine anhidrosis. A missense variant in KCNE4 (NC_009149.3:g.11813731A>G) was proposed as a risk allele, although its association with anhidrosis was not reported. Variant annotation and protein modelling in the original study suggested the G allele conferred risk. We reported no association of the G allele with anhidrosis in 50 horses phenotyped by an intradermal terbutaline sweat test (ITST); all horses produced sweat regardless of genotype. It later appeared the A allele was instead suggested to confer disease risk. To reassess this, we genotyped 20 ITST-tested Thoroughbreds including 9 with partial or complete anhidrosis. The KCNE4 A allele was not associated with phenotype when analyzed as a binary trait (p = 0.16) or when classifying affected horses as having either partial or complete anhidrosis (p = 0.21). The locus was also uninformative in four clinical cases (AA = 1, AG = 1, GG = 2). Reasoning that a true risk allele should be in linkage disequilibrium (LD) with the associated GWA SNV (AX-103822151; rs68656009), we evaluated whole-genome sequence (N = 2) from the initial publication. Both case and control were homozygous AA at the putative risk locus; the case was heterozygous at the GWA SNV. In public data (N = 897), LD between loci was low (r(2) = 0.19). In 369 Thoroughbreds, LD was 0.58. Restricting to Thoroughbreds in the original GWA (N = 85), we found no association of chromosome 6 with anhidrosis. Collectively, these data do not support a role of the KCNE4 variant or the GWA SNV in equine anhidrosis.