K Locus, the DEFB103 Gene

  • Dominant Black (KB)
  • Brindle (kbr)
  • ky/ky
  • This webpage is part of a series on Dog Coat Color Genetics and was last updated on July 27, 2020 by Sheila Schmutz, Ph.D., when it was moved to a different server.

    The K Locus

    Little postulated that an allele AS causes solid or self-colored dogs. Such dogs could be the red Chow or some red Basenji. Later authors, such as Willis (1989) and Sponenberg (2001) have for some reason begun to call this AS allele "dominant black". Kerns et al. (2003, 2005) have shown that this "dominant black" is not caused by an agouti allele using DNA studies. "Dominant black", by their definition is a genotype that is epistatic to fawn, sable, etc. agouti phenotypes, but occurs at another locus, K (for blacK). Several dog breeders have suspected this for some time. The K locus is located on an entirely different chromosome than the agouti locus. Dr. Sophile Candille has studied the alleles in this series, and has been joined by Dr. Christopher Kaelin in this effort. The Stanford group decided to refer to the three alleles as KB, kbr, and ky as of April, 2007.

    Dominant Black

    KB is a critical allele in the formation of black pigment in at least 50 breeds such as the Great Dane, Greyhound, Whippet, etc. In many breeds, KB is fixed, meaning all dogs have only this allele. In some hunting breeds, such as the Labrador Retriever, all dogs are homozygous for KB. In such breeds, the E locus alone then determines whether the dog is black or red. In other hunting breeds such as the German Shorthaired Pointer, the German Wirehaired Pointer, etc. the KB/KB genotype is fixed and the B locus determines whether they are black or brown. Although Little originally thought A caused solid black, we suggest the KB allele is actually the allele causing this type of black. In this case, "solid" does not mean that no white markings occur. A few people call this "self black".

    In order to find this "new" gene K, it was necessary to recruit a large number of dogs. Great Danes, Boxers and Afghans were especially useful but one Akita and one Greyhound were also extremely important in identifying the mutation.

    The gene is a beta-defensin. This family of genes has been known to play an important role in immunity in humans, mice and other animals previously. However, this was the first time that a defensin gene has been shown to affect pigmentation though.

    The discovery of this DEFB103 gene, allows DNA testing to determine which dogs are homozygous for black, and which are heterozygous, in breeds that have black and fawn or black-and-tan varieties, such as Pug, Miniature Schnauzer, Briard, Japanese Chin, Italian Greyhound, etc..

    Knowing if your black, brown or blue dog is homozygous black may be helpful to breeders for other reasons too. Although all Weimaraners are meant to be KB/KB, occasionally a litter contains a couple pups that have pale tan leg markings. This immediately means that both solid colored parents must be KB/ky, but what of the other pups in the litter? They could be tested to determine which are KB/KB (Noa Safra, Pers. Comm.). This situation has arisen in some other breeds too, such as Wirehaired Pointing Griffons.

    A Great Dane breeder who is looking for the right stud dog for her Harlequin bitch and prefers a mate that is homozygous black would find this test helpful. Conversely an Afghan breeder who may be hoping the beautiful black stud dog being considered is not homozygous black, since a variety of coat color in the litter improves pup choice and sales.

    All of the following dog breeds are black, brown or grey because of KB and could be tested for homozygosity. Both the brindle mutation and the ky allele are recessive to KB. Distinguishing these two recessive alleles with a "simple" DNA test is not yet possible since the brindle mutation is a complex mutation.


    Brindle is a pattern of stripes which can be red and black, fawn and black, isabella and grey, etc. It occurs in Great Danes, Greyhounds, Boxers, Akitas, Staffordshire Bull Terriers, Afghan Hounds, Boston Terriers, etc. Although the cause of this pattern is not yet completely understood at the DNA level, much progress was made about 2008. However, the brindle mutation is very complex and diagnostic labs that offer other coat color tests are not able to test for the presence of the brindle allele. Some studies (examples below) have reported finding a KB allele in brindle dogs, but unfortunately this is a misinterpretation of the DNA results. Brindle dogs could be heterozygous and carry one brindle allele and one fawn allele, but again this is not possible to prove with DNA tests that are commercially available in 2020.

    • Ciampolini R, Cecchi F, Spaterna A, Bramante A, Bardet SM, Oulouden A. 2012. Characterization of different 5'-untranslated exons of the ASIP gene in black-and-tan Doberman Pinscher and brindle Boxer dogs. Animal Genetics 44:114-117.
    • Oguro-Okano M., Honda M., Yamazaki K. & Okano K. (2011) Mutations in the melanocortin 1 receptor, b-defensin 103 and agouti signaling protein genes, and their association with coat color phenotypes in Akita-inu dogs. Journal of Veterinary MedicalScience 73, 8538.

    Little attributed the brindle pattern to the E locus, but a DNA study of fawn and brindle Great Danes done in our lab does not support this. Likewise a study by Julie Kerns and colleagues showed that brindle was not caused by the agouti signal peptide or A locus.

    The Brindle allele is part of the K series and as such is on a different chromosome than is MC1R (the E locus). The presence of brindle is dominant to fawn so only one copy of kbr is required. However the dog must also have either an E or Em allele, as discussed below. Likewise since kbr is recessive to KB, no KB allele can be present in the dog. Therefore all brindle dogs are either kbr/kbr or kbr/ky.

    The E locus has been shown to encode the melanocortin 1 receptor (MC1R) gene. To have a brindle phenotype, a dog must have one Em, in which case it will be a brindle with a melanistic mask; or one E allele, in which case it will be brindle with no mask. Dogs that are e/e at this locus have a dysfunctional melanocortin 1 receptor and are unable to make black hairs anywhere on their body. (see red/black/brown page for more information about these alleles). This means that a dog with an e/e genotype could carry brindle and it would not be expressed. Such a dog could have brindle pups with the right mate, and has, based on our DNA studies.

    If the kbr allele is present in a fawn dog with a black mask, it would be expressed. Brindle can be carried and not expressed by a fawn or red dog with no mask that is e/e. However, most fawn dogs with no mask that show shading from the dorsal to ventral side of their body are E/E, ay/ay and since E allows the expression of brindle, such dogs could not be hidden carriers of brindle either.

    Occasionally one hears of two fawn or reddish dogs having brindle pups. This is possible if one is of e/e genotype (the brindle carrier), and the other has at least one E or Em allele. Afghan Hounds are one breed where this situation can arise. French Bulldogs are another breed where this situation arises. In the photo of the family below, LeRoy, on the left, is the fawn sire (e/e) and Ember is the fawn masked dam ( Em/e), of the 3 pups shown in a row below them. LeRoy is a hidden carrier of brindle which then shows in the phenotype of the 2 pups that inherited Ember's Em allele.

    Whereas it seems that all Great Danes have only E and Em alleles at MC1R and therefore in this breed there are no hidden brindle carriers.

    Although the brindle stripes on Banjo, the Great Dane above, are clear and distinct, brindle in some long haired breeds such as this Havanese doesn't look striped. Taz's puppy picture showed he had clear stripes but at 9 1/2, in full coat, one would be hard-pressed to see a stripe.

    In rare cases the brindle stripes are confined to the areas that would typically be tan in a black-and-tan dog. This Staffordshire Bull Terrier, Taisto, is such a dog. Taisto was mated to a fawn/red female who was ay/ay and therefore they had some pups that were "full body" brindle not just brindle on the tan points, at being recessive to ay.

    Dogs can also be brindle with white spots, as illustrated by Allie, a Whippet.

    The Brindle phenotype can be modified by the dilution genes so that the melanistic stripes could be grey, as occurs frequently in Greyhounds and Whippets. This type of grey is caused by the D locus or melanophilin gene. It is also possible that certain shades of brown are not distinguishable as stripes on dark fawn dogs.

    In some breeds the dogs are called red brindle or black brindle. The proportion of blackish stripes to reddish stripes varies among individuals, as illustrated by Whippets Ripley and Colors in the first row below, and Inkie, Joey, and Torch in the group photo. We are not yet certain if this is random or under genetic control.

    Algthough the photo above doesn't show much brindle on Torch, this is him in a face shot!

    The brindle Corgis Goldhawk, PQ, and Lookin also illustrate variation in the amount of black striping.

    Occasionally a brindle dog darkens with age as shown by Duncan, the Cairn Terrier.

    ky/ky Phenotypes

    Dogs which have two recessive alleles ky/ky can express a variety of phenotypes. All black-and-tan dogs or dogs with tan points are ky/ky, as illustrated by the Gordon Setter on the left. All fawn or sable dogs are ky/ky, whether they have a melanistic mask or not, as shown by the sable Tervuren on the right. Red dogs that have an e/e genotype however, could be any genotype at the K locus because e/e is epistatic to the K genotypes.

    The Stanford group chose to name this allele ky since it allows yellow or phaeomelanin pigment to show. Where it shows depends on the alleles at the agouti locus.

    back to Dog Coat Color Main Page