To further complicate things, the agouti gene has more than one promoter which seems to signal where on the body, or even on individual hairs, each pigment is laid down. Roughly, one seems to control ventral or belly color and another dorsal or back color. The simplest way to "see" this is on a black and tan dog......the back is black from eumelanin pigment being made and the belly is tan or red from phaeomelanin pigment being made. In some dogs banded hairs are produced over parts of the body, usually the back. With certain genotypes, the coat color changes from birth to adulthood, usually being born darker and then lightening.

In the summer of 2021, a manuscript appeared in "Nature Ecology & Evolution" based on a in-depth study of the promoter region of the dog ASIP gene. It was conducted by a large group of established researchers, from several universities.

This study showed that there are two types of promoters in canids. The Ventral Promoters (VP1 and VP2) affect the amount of phaeomelanin on the undersides or ventral surface and the Hair Cycle Promoters (HCP1-5) control the amount of phaeomelanin in the the a-banding or coloration of the hairs on the dorsal surface or loosely, the back. So each dog receives one of the 2 forms of the Ventral Promoter and one of the 5 forms of the Hair Cycle Promoter, if they are homozygous.

The combinations are numerous and somewhat complex. I have attempted to draw the Table below to try to capture the promoter information relative to the phenotype names they used to classify the dogs they studied. Note that not all possible combinations were found. Three different combinations were found in the dogs with black back.

This table is over-simplified because it only mentions the promoter combinations on a single chromosome, not both of them. I have attempted to use shades of gray to illustrate the dominance hierarchy, as I understand it, with the palest gray box being the lightest color but the top dominant.

Although the 2021 promoter study introduced new nomenclature for the traditional agouti phenotypes, I will try to correlate them to the names that dog breeders are more familiear with, from the time of C.C. Little's book, in the sections below.

The study suggests that some of the previously published alleles, based on mutations in ASIP that were published from research from our lab (discussed below), are not the causative mutations for the traditional phenotypes, but rather were in "linkage disequilibrium" because a few dogs were discovered that didn't have the expected phenotype from those alleles they carry. The authors say these discrpancies were "rare". Whether some commercial labs will try to use the information from this promoter study instead, is too difficult to predict. The companies that offer automated tests of many loci will not be able to do so because these promoter differences don't adapt well to those types of systems.

Note that "linkage disequilibrium" means that alleles, variants, or mutations go together much more often than would be expected by chance alone.

The promoter study did not include 14 dogs with the E^m allele in their verification however. The manuscript states that the authors couldn't unambiguously classify whether some masked dogs should be classified as Dominant Yellow or Shaded Yellow. This may imply that the E^m allele introduces black on more parts of the body than only the muzzle, as our earlier studies suggested.

• Bannasch DL, Kaelin CB, Letko A, Loechel R, Hug P, Jagannathan V, Henkel J, Roosje P, Hytonen MK, Lohi H, Arumilli M, DoGA consortium (Scandinavia), Minor KM, Mickelson JR, Droegemueller C, Barsh, GS, Leeb T. Dog colour patterns explained by modular promoters of ancient canid origin. Nature Ecology & Evolution. https://doi.org/10.1038/s41559-021-01524-x (August 2021, 18 pages). [Note that this manuscript is Open Access, meaning it can be read and even downloaded for free]

Agouti Alleles in the Traditional Sense

In about 2005 we mapped the agouti gene to dog chromosome 24. This gene undoubtedly has several alleles, but how many is still an open question. Some have been identified using DNA studies and tests for agouti phenotypes in some breeds. Although several books attempt to state the dominance hierarchy of the agouti alleles, until recently no breed was proven to have all the alleles. Most books suggest that it is a^y > a^w > a^t > a. Breeding data and DNA data from our collaborative study with Dr. Greg Barsh's group at Stanford supports this. The Eurasier (shown below) is one breed that has all four alleles (Dreger and Schmutz 2011).

• Dreger, D. L. Kaelin, C. B., Barsh, G.S., Schmutz, S.M. Proposed evolution of the agouti locus alleles in domestic dogs. 5th International Canine and Feline Genome and Hereditary Disease Conference, Baltimore, MD Sep 22-25, 2010

The table above, from the same poster presentation, helps emphasize that although the a^w allele is the wild type, it is no longer that common among dog breeds. We tried to study all common breeds where we believed it would be present in this study and identified only 9 such breeds.

The 2021 promoter study (Bannasch et al.) approaches the evolution of the agouti gene phenotypes based on the promoter region, rather than as we did using the coding exons. Their conclusion is that the HCP1 allele was introduced from a now extinct canid into wolves more than 2 million years ago. It persists in some very pale wolves, usually found in the far north.

Wild Type Black Banded Hairs (a^w)

Agouti signal peptide is also the gene that causes a wolf or coyote to have yellowish hair with black tips and base, or what is often called banded. The competition, in that case, is going on as the hair is growing, which results in a hair that changes color along its length. In breeds with very long hair, there may be even more alternating bands of phaeomelanin and eumelanin. This gene is likely also causing the change in hair color in the Malamute, Siberian Husky (shown above), the Elkhound, and some German Shepherd Dogs. The allele which causes this banding of hairs is sometimes called the wild type allele (aw), or wolf sable. In German Shepherds, like Fello at the right, this pattern is called sable. Although this allele was once fixed in dogs, if we assume that they all descended from the wolf, it is not a very common allele in dog breeds today. Bannasch et al. (2021) have subdivided this phenotype into two which they call "Banded yellow (SY)" and "Agouti (AG)". The suggest the main difference among most of these dogs is that the lighter SY dogs have HCP1 and the darker AG dogs have HCP2 (based on Fig.2).

Fawn or Clear Sable (a^y)

Another allele, ay, has been postulated by Little to explain fawn Whippets, Great Danes, and Pugs which are yellow in pigmentation over all of their body or if masked, most of the body. Latte, the Great Dane, illustrates fawn without mask and Pansy, the Pug, shows it with mask. In some breeds such as Basenjis, Dachshunds, and French Bulldogs there may be dogs that are "clear red" with the genotype e/e and others that are "fawn" ayay. Some breeds have black tips on their hairs into adulthood, like Pugs. Other breeds have large black tips on the hairs in puppyhood that grow out by the time the dog matures. Fawn (ay) is considered dominant to black-and-tan (at) and the wild type allele (aw).

This phenotype is called Dominant Yellow (DY) in the Bannasch et al. (2021) paper. All such dogs have VP1 and HCP1 promoter forms. They also consider this the dominant phenotype.

They studied over 350 dogs. They found 1 dog that was reported as "Dominant Yellow" even though an a^y allele was not present. This supports that the exon 4 mutation test that has been used by most commercial labs, continues to be quite an accurate predictor of the fawn genotype.

This same ay allele occurs in dogs with the coat color called sable in Tervuren, Malinois, Lakenois, Collie, and Shetland Sheepdog. In all these breeds some solid black hairs occur and a few hairs with alternating bands of phaeomelanin and eumelanin can be present. The sable, and probably all red coat colors of the Cardigan Welsh Corgi are also caused by this allele. Note that sable in German Shepherd Dog is a term used to describe the presence of predominantly banded hairs and is not caused by this allele, but instead by the aw allele as described above. Distinguishing the difference between Lady, the Shetland Sheepdog and Fello, the German Shepherd Dog above, is not simple from a photo or even from a distance in person. Note that the promoter study of 2021 would call some, most or all of these dogs "Shaded Yellow (SY)". They report they have VP2 instead of VP1.

Although many dog breeders have assumed that the differences in darkness of fawn or sable dogs is determined by whether they are homozygous or heterozygous for ay, it is likely that some to all of this variation is caused by another gene. The cover from the 2005 issue containing our article describing the ay allele shows 4 dogs that are all ay/ay, yet vary tremendously in shading: Peyton, a female Shetland Sheepdog; Rain, a male Belgian Tervuren; Bull, a male Mastiff; Ginny, a female Akita. It is suspected that dogs that are heterozygous for the promoters may have subtle differences in shade, compared to the homozygous dogs.

Brindle dogs which are full-body brindles, as opposed to dogs with only brindle undersides, have at least one a^y allele.

back to Dog Coat Color Main Page