abstract Tremendous progress has been made in recent years in the elucidation of the canine coat color pathways. Along with the clarity that these discoveries have provided for understanding the color of most dogs have come some exceptions. Among these are the “sable” English Cocker Spaniel, and the “grizzle” Saluki. In both cases, animals share a phenotype which appears to be a modified Agouti phenotype (at/at) with reduced eumelanin expression throughout the black or brown regions, but particularly on the snout, leading to a reverse mask. Two separate mutations have been identified in the same transmembrane domain of MC1R which may be responsible for the phenotype. This would certainly seem to be the case for the English Cockers. The smaller amount of data for the Salukis and some other breeds is consistent, as well. Both of these mutations are hypomorphic, and intermediate in the dominance hierarchy between E and e.

The “sable” English Cocker Spaniel Phenotype

Background Testing for AY Samples from a sable English Cocker dam and her litter of black puppies were presented to Vetgen with a request to test for the Agouti A y mutation which is responsible for “sable” coloration in many breeds including American Cocker Spaniels. All of the dogs tested negative for this mutation, at which point a photograph of the dam was requested. Sequence Agouti gene The color and pattern indicated an Agouti phenotype of some sort, so the coding region of the Agouti gene was sequenced to look for novel mutations in the dam. No mutations were identified, so the assumption at this point was that the phenotype was derived from an Agouti promoter mutation similar to that responsible for the a t phenotype. Test K To support the assumption that an agouti allele was responsible for the phenotype, the dam was tested for the dominant black (K B ) mutation in CBD103 with expectation of a k y k y genotype, which would allow expression of agouti phenotypes. Suprisingly, the dam’s genotype was K B K B which based on existing knowledge of canine coat color should have rendered her black.

HOW DO THE A, B, D, E, and K LOCI AFFECT EACH OTHER IN DETERMINING COAT COLOR? The cartoon below illustrates the interactions of the genes at these five loci in a hierarchy in terms of their role in coat color. If a circle is filled with color, it means the color of the dog has been determined at that point. If a circle is still white, it means information about an additional gene is required. The first locus to look at is the E locus. The gene at this locus is responsible for black masks when present as well as most shades of yellow and red. Any dog that is “ee” will be some shade of yellow to red, and everything happening at the A, B, D, and K loci will be hidden until the next generation. If the dog has any E or E m alleles, then it will not be yellow and we must look next at the K locus. There are three versions, or alleles, of the K locus: K B, k br, and k y. If a dog has even a single copy of K B (K B K B, K B k, K B k br ) it will be solid colored in the pigmented areas, and we go directly to the B locus to determine color. Everything happening at the A locus in these dogs is hidden until the next generation. If a dog is k br k br, or k br k y, it will nearly always be brindled and we look next at the A locus to see the background color and pattern of that brindling. If a dog is k y k y, it will not be brindled, and we go next to the A locus to see which alleles are expressed. The A locus has at least four alleles. There are direct tests available for A Y (fawn, or sable) and “a” (recessive black), and in many breeds these two tests can be used to yield information about a t (tan points). There is no direct test for the Wild type (think wolf pattern) allele which is designated A w. Any dog which has at least one copy of A Y (and no K B ) will be fawn or sable, either with or without brindling. Any dog that is aa (with no K B ) will be black. Any dog that is a t a t or a t a (with no K B ) will have tan points, either with or without brindling. The next stop is the B locus. Any dog which is bb will be have brown fur in those areas that would otherwise be black. This holds true for both solid colored and agouti-patterned animals.

EE, Ee, EE m, E m E m, E m e ee E Locus BB, Bb bb B Locus kykykyky k br k br, k br k y K B K B, K B k y, K B k br K Locus A y A y, A y a t, A y a a t a t, a t a aa A Locus aa A y A y, A y a t, A y a a t a t, a t a

Sequence analysis of Cbd103 (K) and Mc1R (E) coding regions Both exons of the Cbd103 gene were sequenced to see if any novel K locus mutations may be present. None were identified. In sequencing the MC1R gene a mutation was identified that substitutes an Asparagine for Aspartate at residue 84 of the protein. This mutation was found in all “sable” English Cocker Spaniels.

PHENOTYPES AND GENOTYPES

“Grizzle” Salukis and “Domino” Afghan Hounds A second mutation in the same transmembrane domain has been identified in dogs with a very similar phenotype to that seen in the “sable” English Cockers. The allele appears to behave in the same manner i.e. recessive to E but dominant to e.

Transmembrane domain 2 mutations in Humans There have been over 60 natural mutations described in the human Mc1r gene with about 20% of these clustered in transmembrane domain 2. Many of these have functional significance. One such mutation, D84E, alters the same residue as the English Cocker mutation and is known to be responsible for red hair.

summary The mutation responsible for the “sable” phenotype in English Cocker Spaniels appears to be a hypomorphic MC1R allele (e h ) which falls into the heirarchy as follows: E M > E > e h > e The full designation is e h(D84N) which differentiates the allele from other probable e h alleles which will likely include the residue 78 mutation in Salukis and Afghan Hounds.

acknowledgements Special thanks go to breeders Sally Durham, who presented us with our first samples, and especially Sten and Eva Bergquist who provided valuable insight in the form of both numerous samples and historical breeding data.