Dog coat genetics
This article may be too technical for most readers to understand. Please help improve it to make it understandable to non-experts, without removing the technical details. (June 2017) (Learn how and when to remove this template message)
Modern breeds of dog exhibit a diverse range of coat colorings, patterns, lengths and textures. In recent years, the understanding of the genetic basis for coat coloring and patterning and coat length and texturing has increased significantly. Genes occur in pairs, and for each gene a dog will either be:Homozygous, that is, Inherited the same gene from each parent, and therefore likely to express that trait; or Heterozygous, that is, Inherited a different variant from each parent, and would require understanding of the individual gene to know how the dog will look.
- 1 Genes associated with coat color
- 1.1 A (agouti) locus
- 1.2 B (brown) locus
- 1.3 D (dilute) locus
- 1.4 E (extension) locus
- 1.5 H (harlequin) locus
- 1.6 K (dominant black) locus
- 1.7 M (merle) locus
- 1.8 S (spotting) locus
- 1.9 Theoretical genes for color and pattern
- 1.10 Interactions of color and pattern genes
- 2 Genes associated with hair length, growth and texture
- 3 Other related genes
- 4 Genetic testing and phenotype prediction
- 5 See also
- 6 Notes
- 7 References
- 8 External links
Genes associated with coat color
Coat colour in dogs consists of specific patterns of:
- Eumelanin — black, chocolate brown, grey or taupe pigment;
- Phaeomelanin — tan pigment, including all shades from red to gold to cream; and/or
- No melanin — white, the lack of pigment.
There are currently eight known genes within the canine genome that are associated with coat color. Each of these genes occurs in at least two variants, or alleles, which accounts for the variation in coat color among animals. Each of these genes exists at a fixed location, or locus, of the animal's genome. The loci associated with canine coat color are:
A (agouti) locus
The alleles at the A locus are related to the production of agouti signalling protein (ASIP) and determine whether an animal expresses an agouti appearance, and if so what type, by controlling the distribution of pigment in individual hairs. There are four known alleles that occur at the A locus:
- Ay = Fawn or sable (tan with black whiskers and varying amounts of black-tipped and/or all-black hairs dispersed throughout) - fawn typically referring to dogs with clearer tan and sable to those with more black shading
- aw = Wild-type agouti (each hair with 3-6 bands alternating black and tan) - also called wolf sable
- at = Tan point (black with tan patches on the face and underside) - including saddle tan (tan with a black saddle or blanket) 
- a = Recessive black (black, inhibition of phaeomelanin)
- ayt = Recombinant fawn (black with tan patches on the face and underside) has been identified in numerous Tibetan Spaniels and a Tibetan Mastiff. Its hierarchical position is not yet understood.
Most texts suggest that the dominance hierarchy for the A locus alleles appears to be as follows: Ay > aw > at > a; however, research suggests the existence of pairwise dominance/recessiveness relationships in different families and not the existence of a single hierarchy in one family.
- Ay is incompletely dominant to at, so that heterozygous individuals have more black sabling, especially as puppies and Ayat can resemble the awaw phenotype. Other genes also affect how much black is in the coat.
- aw is the only allele present in many Nordic spitzes, and is not present in most other breeds.
- at includes tan point and saddle tan, both of which look tan point at birth. Modifier genes in saddle tan puppies cause a gradual reduction of the black area until the saddle tan pattern is achieved.
- a is only present in a handful of breeds. Most black dogs are black due to a K locus allele.
B (brown) locus
The alleles at the B locus are related to the production of tyrosinase related protein 1 (TYRP1) and determine the degree to which an animal expresses tyrosinase, an enzyme related to the production of melanin, in its coat and skin (including the nose and paw pads). There are four known alleles that occur at the B locus:
- B = Black melanin
- b = Brown melanin - such as chocolate or liver (includes several alleles - bs, bd and bc)
B is dominant to b.
- An animal that has at least one copy of the B allele will have a black nose, paw pads and eye rims and (usually) dark brown eyes.
- An animal with any matched or unmatched pair of the b alleles will have brown, rather than black, hair, a liver nose, paw pads and eye rims, and hazel eyes. Phaeomelanin is slightly reduced from redder to oranger tan. Only one of the alleles is present in the English Setter (bs), Doberman Pinscher (bd) and Italian Greyhound (bc), but in most breeds with any brown allele 2 or all 3 are present. It is unknown whether the different brown alleles cause specific shades or hues of brown.
D (dilute) locus
The alleles at the D locus (the melanophilin gene or MLPH) are related to the dilution of eumelanin and phaeomelanin and determine the intensity of pigmentation. There are two known alleles:
- D = Not diluted (black or brown eumelanin, reddish or orangish tan phaeomelanin)
- d = Diluted (black diluted to bluish grey ranging from light blue-grey to dark steel, chocolate brown diluted to taupe or "Isabella", yellowish tan phaeomelanin)
D is completely dominant to d.
- An animal that is homozygous for d will have a diluted coat colour, slight to moderate dilution of the paw pads and eye rims towards bluish grey if B/- or taupe if b/b, and slight to moderate reduction of eye colour from brown towards amber in a B/- animal, or from hazel towards light amber in a b/b animal. Homozygosity of d is sometimes accompanied by hair loss and recurrent skin inflammation, a condition referred to as either color dilution alopecia (CDA) or black hair follicular dysplasia (BHFD) depending upon the breed of dog.
E (extension) locus
The alleles at the E locus (the melanocortin receptor one gene or MC1R) determine whether an animal expresses a melanistic mask, as well as determining whether an animal can produce eumelanin in its coat. There are three known, plus two more theorized, alleles that occur at the E locus:
- Em = Mask (a eumelanin mask is added to the face)
- EG = Grizzle (if atat and not KB/-, tan underparts with a dark overlay covering the top and sides of the body, head and tail, and the outside of the limbs) - also called domino
- E = Normal extension (pattern expressed as per alleles present at A and K loci)
- eh = Cocker sable (if KB/- and may require atat, tan with a dark overlay covering the top and sides of the body, head and tail, and the outside of the limbs)
- e = Recessive or clear fawn (tan, inhibition of eumelanin)
The dominance hierarchy for the E locus alleles appears to be as follows: Em > EG > E > eh > e.
- E allows normal expression of eumelanin and/or phaeomelanin according to the alleles present at the A and K loci.
- Em allows similar pattern expression to E except any tan (phaeomelanin) areas on the mask area are replaced with eumelanin (black/etc.) The mask can vary from the muzzle, to the face and ears, to a larger area with shading on the front and sides as in the Belgian Tervuren. The mask Em is unaffected by the greying gene G and will remain dark in a G/- animal while the rest of the dog pales, such as in Kerry Blue Terriers. Some puppies are born with a mask which fades away within a few weeks of birth: these puppies do not have the Em allele and their temporary mask is due to sabling.
- An animal that is homozygous for e will express a red to yellow coat regardless of most alleles at other loci. Eumelanin is inhibited, so there can be no black hairs anywhere, even the whiskers. Pigment on the nose leather can be lost at the middle (Dudley nose). In combination with a/a (phaeomelanin inhibitor), an e/e dog will be white to off-white; in combination with U/U or U/u, an e/e dog will be off-white or cream.
- The Grizzle allele has been studied only in Salukis and Afghan Hounds, the latter in which it is referred to as "Domino", but also occurs in the Borzoi. Its placement in the dominance hierarchy has not been solidified. Black with fawn-tan points (at/at E/-) is instead dark-sable with extended clear-tan points (at/at EG/-). Brindle affects fawn and sable areas, resulting in black with bridled-tan points (at/at E/- Kbr/-) or brindle with clear-tan points (at/at EG/- Kbr/-). Expression of EG is dependent upon the animal being homozygous for at and not possessing Em or KB. EG is theorized to have no effect on the phenotype of non-at/- nor KB dogs and to be allelic to Em and e.
- The eh sable extension allele has been studied only in English Cocker Spaniels and produces sable in the presence of dominant black KB and tan point at/at. Its expression is dependent upon the animal not possessing Em nor E nor being homozygous for e. eh is theorized to be on the E locus and to have no effect on ky/ky dogs. All cocker spaniels are homozygous for at, so it is unknown how the gene may function in the presence of other A-series alleles.
H (harlequin) locus
DNA studies have not yet isolated the gene at the H locus, but the traits associated with it have been mapped to chromosome 9. The H locus is a modifier locus (of the M locus) and the alleles at the H locus will determine if an animal expresses a harlequin vs merle pattern. There are two alleles that occur at the H locus:
- H = Harlequin (if M/-, patches of full colour and white)
- h = Non-harlequin (if M/-, normal expression of merle)
H/h heterozygotes are harlequin and h/h homozygotes are non-harlequin. Breeding data suggests that homozygous H/H is embryonic lethal and that therefore all harlequins are H/h.
- The Harlequin allele is specific to Great Danes. Harlequin dogs (H/h M/m) have the same pattern of patches as merle (h/h M/m) dogs, but the patches are white and harlequin affects eumelanin and phaeomelanin equally. H has no effect on non-merle m/m dogs.
K (dominant black) locus
- KB = Dominant black (black)
- kbr = Brindle (black stripes added to tan areas)
- ky = Phaeomelanin permitted (pattern expressed as per alleles present at A and E loci)
The dominance hierarchy for the K locus alleles appears to be as follows: KB > kbr > ky.
- KB causes a solid eumelanin coat (black, brown, grey or taupe) except when combined with e/e (tan or white), Eh/- (Cocker sable) or Em/- G/- and appropriate coat type (light eumelanin with dark eumelanin mask)
- kbr causes the addition of eumelanin stripes to all tan areas of a dog except when combined with e/e (no effect) or EG/- atat non-KB/- (eumelanin and sabled areas become striped, tan areas remain tan)
- ky is wild-type allowing full expression of other genes.
M (merle) locus
The alleles at the M locus (the silver locus protein homolog gene or SILV, aka premelanosome protein gene or PMEL) determine whether an animal expresses a merle pattern to its coat. There are two alleles that occur at the M locus:
- M = Merle (patches of full colour and reduced colour)
- m = Non-merle (normal expression)
M and m show a relationship of both co-dominance and no dominance.
- On heterozygous M/m merles, black is reduced to silver on ~50% of the animal in semi-random patches with rough edges like torn paper. The fraction of the dog covered by merle patches is random such that some animals may be predominantly black and others predominantly silver. The merle gene is “faulty” with many merle animals having one odd patch of a third shade of grey, brown or tan.
- On homozygous M/M “double merles”, black is replaced with ~25% black, ~50% silver and ~25% white, again with random variation, such that some animals have more black or more white.
- Eumelanin (black/etc.) is significantly reduced by M/m, but phaeomelanin is barely affected such that there will be little to no evidence of the merle gene on any tan areas or on an e/e dog. However, the white patches caused by M/M affect both pigments equally, such that a fawn double merle would be, on average, ~75% tan and ~25% white.
- The merle gene also affects the skin, eye colour, eyesight and development of the eye and inner ear. Merle M/m puppies develop their skin pigmentation (nose, paws, belly) with speckled-edged progression, equally evident in e/e merles except when extensive white markings cause pink skin to remain in these areas. Blue and part-blue eyes are common.
- Both heterozygosity and homozygosity of the merle gene (i.e., M/m and M/M) are linked to a range of auditory and ophthalmologic abnormalities. Most M/m merles have normal-sized eyes and acceptably functional eyesight and hearing; most M/M double merles suffer from microphthalmia and/or partial to complete deafness.
S (spotting) locus
The alleles at the S locus (the microphthalmia-associated transcription factor gene or MITF) determine the degree and distribution of white spotting on an animal's coat. There is disagreement as to the number of alleles that occur at the S locus, with researchers postulating either two or four alleles. The four alleles postulated are:
- S = Solid color (small areas of white may appear on chest, toes or tail tip)
- si = Irish-spotting (white on muzzle, forehead, feet, legs, chest, neck and tail)
- sp = Piebald (varies from coloured with Irish spotting plus at least one white marking on the top or sides of the body or hips, to mostly white which generally retains patches of colour around the eyes, ears and tail base)
- sw = Extreme piebald spotting (extremely large areas of white, almost completely white)
S is incomplete dominant (towards co-dominant) to sp. DNA studies have not yet confirmed the existence of all four alleles, with some research suggesting the existence of at least two alleles (S and sp) and other research suggesting the possible existence of a third allele (si).
- S/sp heterozygotes usually have some white at birth on the chest and toes, which may be covered by ticking as the puppy grows. Animals of this genotype also commonly display pseudo-Irish spotting; in fact most Irish-spotted dogs are so due to heterozygosity for solid and piebald.
- A few breeds (e.g., Boston Terrier) are fixed for Irish spotting and therefore theorized to possess a different allele on the S locus (si) or an allele on a completely separate gene.
- It has been suggested that what appears to be the result of an sw allele is in fact the result of plus and minus modifiers acting on one of the other alleles. It is thought that the spotting that occurs in Dalmatians is the result of the interaction of three loci (the S locus, the T locus and F locus) giving them a unique spotting pattern not found in any other breed.
- White spotting also affects skin, causing pink patches.
- White spotting can cause blue eyes, microphthalmia, blindness and deafness; however, because pigmentation is generally retained around the eye/ear area, this is rare except among sw/sw dogs (or extreme versions of sp/sp if sw does not exist).
Theoretical genes for color and pattern
There are at least six additional theoretical loci thought to be associated with coat color in dogs. DNA studies are yet to confirm the existence of these genes or alleles but their existence is theorised based on breeding data:
C (colored) locus
The alleles at the theoretical C locus are thought to determine the degree to which an animal expresses phaeomelanin, a red-brown protein related to the production of melanin, in its coat and skin. Five alleles are theorised to occur at the C locus:
- C = Full color (animal expresses phaeomelanin)
- cch = Chinchilla (partial inhibition of phaeomelanin resulting in decreased red pigment)
- ce = Extreme dilution (inhibition of phaeomelanin resulting in extremely reduced red pigment)
- cb, cp = Blue-eyed albino/Platinum (almost total inhibition of phaeomelanin resulting in near albino appearance)
- ca = Albino (complete inhibition of phaeomelanin production, resulting in complete inhibition of melanin production)
The C locus in dogs is not well understood and the theorised alleles are based on those present in other species. True albinism has not been conclusively shown to exist in dogs. It is thought that an animal that is heterozygous for the C allele with any of the c alleles will express a result somewhere between the two alleles.
White in Doberman Pinschers and albino-like animals of Asian/Tibetan companion breeds have a phenotype resembling a C locus dilution, but all tested animals have been C/C. The gene responsible is recessive and not at the C locus.
F (flecking) locus
The alleles at the theoretical F locus are thought to determine whether an animal displays small, isolated regions of white in otherwise pigmented regions (not apparent on white animals). Two alleles are theorised to occur at the F locus:
- F = Flecked
- f = Not flecked
It is thought that F is dominant to f.
G (progressive greying) locus
The alleles at the theoretical G locus are thought to determine if progressive greying of the animal's coat will occur. Two alleles are theorised to occur at the G locus:
- G = Progressive greying (melanin lost from hairs over time)
- g = No progressive greying
It is thought that G is dominant to g.
- The greying gene affects both eumelanin, and to a lesser extent phaeomelanin. In the presence of Em/- the eumelanin mask will be unaffected and remain dark. Grey dogs are born fully coloured and develop the greying effect over several months. New hairs are grown fully coloured but their colour fades over time towards white. Greying is most evident in continuous-growing coats (long + wire + curly) as individual hairs remain on the dog long enough for the colour to be lost. In short-haired dogs, hairs are shed out and re-grown before the colour has a chance to change.
- Premature greying, in which the face/etc. greys at a young age is not caused by G and has not been proven to be genetic.
I (intensity) locus
The alleles at the theoretical I locus are thought to affect phaeomelanin expression. Two alleles are theorised to occur at the I locus:
- I = Intense red, not diluted
- i = Not intense red
It is thought that I and i interact with semi-dominance, so that there are three distinct phenotypes. I/i heterozygotes are paler than I/I animals but darker than i/i animals.
- i results in light-coloured phaeomelanin such as gold, yellow, buff and apricot. This gene is the most common cause of lighter tans, and unlike d/d, it allows the skin and eyes to remain dark.
T (ticking) locus
The alleles at the theoretical T locus are thought to determine whether an animal displays small, isolated regions of pigment in otherwise s-spotted white regions. Two alleles are theorised to occur at the T locus:
- T = Ticked
- t = Not ticked
It is thought that T is dominant to t. Ticking may be caused by several genes rather than just one. Patterns of medium-sized individual spots, smaller individual spots, and tiny spots that completely cover all white areas leaving a roan-like or merle-like appearance (reserving the term large spots for the variation exclusive to the Dalmatian) can each occur separately or in any combination.
- The effect of the ticking gene(s) is to add back little coloured spots to areas made white by piebald spotting (-/s) or the limited white markings of S/S animals. It does not affect white areas that were caused by a/a e/e or M/M or M/m H/h. The colour of the tick marks will be as expected or one shade darker. Tick marks are semi-random, so that they vary from one dog to the next and can overlap, but are generally present on the lower legs and heavily present on the nose.
U (urajiro) locus
The alleles at the theoretical U locus are thought to limit phaeomelanin production on the cheeks and underside. Two alleles are theorised to occur at the U locus:
- U = Urajiro
- u = Not urajiro
It is thought that U is dominant to u but incomplete with homozygosity required for complete dilution to off-white and heterozygotes displaying a darker cream. The urajiro pattern is expressed in the tan (phaeomelanin) areas of any dog who is not e/e. In e/e dogs, the urajiro gene causes dilution of the entire dog to off-white or cream.
Interactions of color and pattern genes
Alleles at the Agouti (A), Extension (E) and Black (K) loci determine colour pattern (eumelanin vs phaeomelanin):
|Fawn or Sable
aw/aw, aw/at or aw/a
at/at or at/a
E/E or E/e
eh/eh or eh/e
Kbr/Kbr or Kbr/ky
|black & brindled tan
E/E or E/e
|brindle||brindle||black & brindled tan||black|
EG/EG, EG/E or EG/e
|?||?||brindle & tan||?|
|fawn or sable
|black & tan
E/E or E/e
|fawn or sable||wolf sable||black & tan||black|
EG/EG, EG/E or EG/e
|* Note that the black mask on a black dog is only phenotypically evident in presence of G/- and an appropriate coat type.|
† eh and EG are only included in the table where their interactions are known.
Alleles present at the Brown (B) and Dilution (D) loci determine melanin colour:
(D/D or D/d)
B/B or B/b
|Taupe or "Isabella" eumelanin|
|* Note that phaeomelanin is frequently diluted by intensity factor of theoretical I locus.|
Alleles present at the Intensity (I), Urajiro (U), Greying (G) and Albino (C-like) loci determine melanin shade.
Alleles present at the Merle (M) and Harlequin (H) loci cause patchy reduction of melanin to half (merle), zero (harlequin) or both (double merle).
Alleles present at the Spotting (S), Ticking (T) and Flecking (F) loci determine white markings.
Genes associated with hair length, growth and texture
Every hair in the dog coat grows from a hair follicle, which has a three phase cycle, as in most other mammals. These phases are:
- anagen, growth of normal hair;
- catagen, growth slows, and hair shaft thins; and
- telogen, hair growth stops, the follicle rests, and the old hair falls off—is shed. At the end of the telogen phase, the follicle begins the cycle again.
Most dogs have a double coat, each hair follicle containing 1-2 primary hairs and several secondary hairs. The primary hairs are longer, thicker and stiffer, and called guard hairs or outer coat. Each follicle also holds a variety of silky- to wiry-textured secondary hairs (undercoat) all of which are wavy, and smaller and softer than the primary hair. The ratio of primary to secondary hairs varies at least six-fold, and varies between dogs according to coat type, and on the same dog in accordance with seasonal and other hormonal influences. Puppies are born with a single coat, with more hair follicles per unit area, but each hair follicle contains only a single hair of fine, silky texture. Development of the adult coat begins around 3 months of age, and is completed around 12 months.
Research indicates that the majority of variation in coat growth pattern, length and curl can be attributed to mutations in four genes, the R-spondin-2 gene or RSPO2, the fibroblast growth factor-5 gene or FGF5, the keratin-71 gene or KRT71 and the melanocortin 5 receptor gene (MC5R). The wild-type coat in dogs is short, double and straight.
L (length) locus
- L = Short coat
- l = Long coat
L is dominant to l. A long coat is demonstrated when a dog has pair of recessive l alleles at this locus. The dominance of L > l is incomplete, and L/l dogs have a small but noticeable increase in length and finer texture than closely related L/L individuals. However, between breeds there is significant overlap between the shortest L/L and the longest L/l phenotypes. In certain breeds (German Shepherd, Alaskan Malamute, Cardigan Welsh Corgi), the coat is often of medium length and many dogs of these breeds are also heterozygous at the L locus (L/l).
W (wired) locus
The alleles at the W locus (the R-spondin-2 gene or RSPO2) determine the coarseness and the presence of "facial furnishings" (e.g. beard, moustache, eyebrows). There are two known alleles that occur at the W locus:
- W = Wire (hair is coarse and facial furnishings present)
- w = Non-wire (hair is not coarse and facial furnishings are not present)
W is dominant to w, but the dominance of W > w is incomplete. W/W dogs have coarse hair, prominent furnishings and greatly-reduced shedding. W/w dogs have the harsh wire texture, but decreased furnishings, and overall coat length and shedding similar to non-wire animals.
Animals that are homozygous for long coat (i.e., l/l) and possess at least one copy of W will have long, soft coats with furnishings, rather than wirey coats.
R (curl) locus
- R = Straight
- r = Curly
The relationship of R to r is one of no dominance. Heterozygotes (R/r) have wavy hair that is easily distinguishable from either homozygote. Wavy hair is considered desirable in several breeds, but because it is heterozygous, these breeds do not breed true for coat type.
Corded coats, like those of the Puli and Komondor are thought to be the result of continuously growing curly coats (long + wire + curly) with double coats, though the genetic code of corded dogs has not yet been studied. Corded coats will form naturally, but can be messy and uneven if not "groomed to cord" while the puppy's coat is lengthening.
Interaction of length and texture genes
|COAT TYPE GENE
L/L or L/l
(e.g., Akita, Greyhound)
(e.g., Chesapeake Bay Retriever)
|Short curly |
(Curly Coated Retriever? (unproven))
(e.g., Pomeranian, Cocker Spaniel)
(e.g., Boykin Spaniel)
|Long curly |
(e.g., Irish Water Spaniel)
W/W or W/w
(e.g., Shih Tzu, Bearded Collie)
(e.g., Bichon Frise, Portuguese Water Dog, SCWT)
|Long curly with furnishings or |
Corded (e.g., Poodle, Puli, Komondor)
L/L or L/l
(e.g., Border Terrier, Scottish Terrier)
(e.g., Wire Fox Terrier)
(e.g., Wirehaired Pointing Griffon)
Breed exceptions to coat type
Breeds in which coat type Is not explained by FgF5, RSPO2 and KRT71 genes:
Genotypes of dogs of these 3 breeds are usually L/L or L/l, which does not match with their long-haired phenotype. The Yorkshire and Silky Terriers share common ancestry and likely share an unidentified gene responsible for their long hair. The Afghan Hound has a unique patterned coat that is long with short patches on the chest, face, back and tail. The Irish Water Spaniel may share the same pattern gene, although unlike the Afghan Hound, the IWS is otherwise genetically a long-haired (fixed for l/l) breed.
The alleles on the melanocortin 5 receptor gene (MC5R) determine whether an animal will have neotenous retention of a puppy-like coat type. The locus has not been assigned a common name or letter, but has been called the shedding gene or single coat gene. There are two known alleles that occur at this locus:
- The mutant allele (reduced shedding, single coat or minimal undercoat, reduced length)
- The wildtype allele (normal shedding, double coat, normal length)
The mutant allele is incomplete dominant. With respect to coat texture, shedding, follicle density and number of secondary hairs per follicle, heterozygotes closely resemble animals homozygous for the mutant allele, with minor differences. With respect to coat length and the prominence of fringing and furnishings, the relationship between the two alleles is more complex and dependent on the alleles present at the L and W loci:
- In short-haired dogs (L/- w/w), this gene causes the smooth coat type that is common in hounds and pointers. Coat length is significantly reduced in animals homozygous for the smooth-coat allele, and of intermediate length in heterozygotes. Heterozygosity for long coat (L/l) dulls the effect on coat length. Typically, the undercoat is completely absent. Very few breeds (e.g., Chihuahua) have both smooth and non-smooth coat types.
- In long-haired dogs (l/l w/w), this gene causes fringed coats (e.g., Saluki, long-haired pointers). Coat length is reduced on the body, but lengthened on the feathering. Fringed coats may have an unbristled undercoat. An overall long single coat (e.g., Papillon, Japanese Chin) requires additional lengthening modifier genes.
- In wire-haired dogs (L/- W/-), this gene causes short-wire coats (e.g., Border Terrier, Wire-haired Dachshund, German Wirehaired Pointer) only when homozygous, and has no effect on length when heterozygous. Short-wire coats may have a bristled undercoat.
- In shaggy-haired dogs (l/l W/-), this gene causes a soft single coat (e.g., Coton de Tulear, Soft Coated Wheaten Terrier, Dachshunds of mixed longhair/wirehair parentage) which varies by breed from cottony to silky. The minimal undercoat of fringed and short-wire coats originates from a different subset of secondary hairs, and is lost when a dog has the alleles for both long and wire hair.
- In dogs with long curly coats with furnishings (l/l W/- R/R), this gene causes a single long curly coat with furnishings that will not cord (e.g., most Poodles), as proper formation of cords requires a double coat.
Remaining influences of length (e.g., setter vs Cocker Spaniel), texture (e.g., setter vs mountain dog vs spitz or Bearded Collie vs Old English Sheepdog) and abundance of undercoat (e.g., Labrador Retriever vs Keeshond) are likely polygenic.
Some breeds of dog do not grow hair on parts of their bodies and may be referred to as "hairless". Examples of "hairless" dogs are the Xoloitzcuintli (Mexican Hairless Dog), the Peruvian Inca Orchid (Peruvian Hairless Dog) and the Chinese Crested. Research suggests that hairlessness is caused by a dominant allele of the forkhead box transcription factor (FOXI3) gene, which is homozygous lethal. There are coated homozygous dogs in all hairless breeds, because this type of inheritance prevents the coat type from breeding true. The hairlessness gene permits hair growth on the head, legs and tail. Hair is sparse on the body, but present and typically enhanced by shaving, at least in the Chinese Crested, whose coat type is shaggy (long + wire). Teeth are affected as well, and hairless dogs have incomplete dentition.
The American Hairless Terrier is unrelated to the other hairless breeds and displays a different hairlessness gene. Unlike the other hairless breeds, the AHT is born fully coated, and loses its hair within a few months. The AHT gene, serum/glucocorticoid regulated kinase family member 3 gene (SGK3), is recessive and does not result in missing teeth. Because the breed is new and rare, outcrossing to the parent breed (the Rat Terrier) is permitted to increase genetic diversity. These crosses are fully coated and heterozygous for AHT-hairlessness.
Some breeds (e.g., Rhodesian Ridgeback, Thai Ridgeback) have an area of hair along the spine between the withers and hips that leans in the opposite direction (cranially) to the surrounding coat. The ridge is caused by a duplication of several genes (FGF3, FGF4, FGF 19, ORAOV1 and sometimes SNP), and ridge is dominant to non-ridged.
Genetic testing and phenotype prediction
The genes responsible for the determination of coat colour also affect other melanin-dependent development, including skin colour, eye colour, eyesight, eye formation and hearing. In most cases, eye colour is directly related to coat colour, but blue eyes in the Siberian Husky and related breeds, and copper eyes in some herding dogs are not known to be related to coat colour.
The development of coat colour, skin colour, iris colour, pigmentation in back of eye and melanin-containing cellular elements of the auditory system occur independently, as does development of each element on the left vs right side of the animal. This means that in semi-random genes (M merle, s spotting and T ticking), the expression of each element is independent. For example, skin spots on a piebald-spotted dog will not match up with the spots in the dog's coat; and a merle dog with one blue eye can just as likely have better eyesight in its blue eye than in its brown eye.
Loci for coat colour, type and length
All known genes are on separate chromosomes, and therefore no gene linkage has yet been described among coat genes. However, they do share chromosomes with other major conformational genes, and in at least one case, breeding records have shown an indication of genes passed on together.
|ASIP||24||Ay, aw, at, a||Agouti||Sable, wolf-sable, tan point, recessive black; as disproven|
|TYRP1||11||B, bs, bd, bc||Brown||Black, 3 x chocolate / liver|
|SLC45A2||4||C||Colour||No alleles known; possible tyrosinase gene||STC2, GHR(1)|
& GHR(2) size
|MLPH||25||D, d||Dilution||Black/chocolate, blue/isabella|
|MC1R||5||Em, Eg, E, eh, e||Extension||Black mask, grizzle, normal extension, cocker-sable, recessive fawn|
|PSMB7||9||H, h||Harlequin||Harlequin, non-harlequin|
|DEFB103||16||KB, Kbr, ky||blacK||Dominant black, brindle, agouti-enabler|
|FgF5||32||L, l||Longcoat||Short coat, long coat|
|PMEL||10||M, m||Merle||Double merle, merle, non-merle||HMGA2 size|
|KRT71||27||R, r||cuRlycoat||Straight coat, curly coat|
|MITF||20||S, si, sp||Spotting||Solid, Irish spotting, piebald spotting; sw not proven to exist|
|RSPO2||13||W, w||Wirecoat||Wire coat, non-wire coat|
|MC5R||1||n/a||Shedding||Single coat/minimal shedding, double coat/regular shedding||C189G bobtail|
|--||3||-||-||No coat genes yet identified here.||IGF1R size|
|--||7||-||-||No coat genes yet identified here.||SMAD2 size|
|--||15||-||-||No coat genes yet identified here.||IGF1 size|
There are size genes on all 39 chromosomes, 17 classified as "major" genes. 7 of those are identified as being of key importance and each results in ~2x difference in body weight. IGF1 (Insulin-like growth factor 1), SMAD2 (Mothers against decapentaplegic homolog 2), STC2 (Stanniocalcin-2) and GHR(1) (Growth hormone receptor one) are dose-dependent with compact dwarfs vs leaner large dogs and heterozygotes of intermediate size and shape. IGF1R (Insulin-like growth factor 1 receptor) and HMGA2 (High-mobility group AT-hook 2) are incomplete dominant with delicate dwarfs vs compact large dogs and heterozygotes closer to the homozygous dwarfed phenotypes. GHR(2) (Growth hormone receptor two) is completely dominant, homozygous and heterozygous dwarfs equally small, larger dogs with a broader flatter skull and larger muzzle. It is believed that the PMEL/SILV merle gene is linked to the HMGA2 size gene, meaning that alleles are most often inherited together, accounting for size differences in merle vs non-merle litter mates, such as in the Chihuahua (merles usually larger) and Shetland Sheepdog (merles frequently smaller).
- Labrador Retriever coat colour genetics
- Cat coat genetics
- Equine coat color genetics
- Farm-Fox Experiment
- Researchers have not yet assigned a letter to this locus and "R" has been selected based on the use of the term "Rex" for curled hair in domestic cats.
- Schmutz, S. M.; Berryere, T. G. (December 2007). "Genes affecting coat color and pattern in domestic dogs: a review". Animal Genetics. 38 (6): 539–549. doi:10.1111/j.1365-2052.2007.01664.x. PMID 18052939.
- Edouard Cadieu; Mark W. Neff; Pascale Quignon; Kari Walsh; Kevin Chase; Heidi G. Parker; Bridgett M. VonHoldt; Alison Rhue; Adam Boyko; Alexandra Byers; Aaron Wong; Dana S. Mosher; Abdel G. Elkahloun; Tyrone C. Spady; Catherine André; K. Gordon Lark; Michelle Cargill; Carlos D. Bustamante; Robert K. Wayne; Elaine A. Ostrander (October 2009). "Coat Variation in the Domestic Dog Is Governed by Variants in Three Genes". Science. 326 (5949): 150–153. doi:10.1126/science.1177808. PMC . PMID 19713490.
- Dreger, Dayna L.; Schmutz, Sheila M. (2011). "A SINE Insertion Causes the Black-and-Tan and Saddle Tan Phenotypes in Domestic Dogs". Journal of Heredity. 102: S11–S18. doi:10.1093/jhered/esr042. PMID 21846741.
- Dreger DL, Parker H, Ostrander E, Schmutz SM. The involvement of RALY in a complex gene interaction producing the saddle tan phenotype in dogs. A presentation at Advances in Canine and Feline Genomics and Inherited Diseases 2012 Conference, Visby, Sweden. June 1, 2012.
- Julie A. Kerns; J. Newton; Tom G. Berryere; Edward M. Rubin; Jan-Fang Cheng; Sheila M. Schmutz; Gregory S. Barsh (October 2004). "Characterization of the dog Agouti gene and a nonagouti mutation in German Shepherd Dogs". Mammalian Genome. 15 (10): 798–808. doi:10.1007/s00335-004-2377-1. ISSN 0938-8990. PMID 15520882.
- Ute Philipp; Henning Hamann; Lars Mecklenburg; Seiji Nishino; Emmanuel Mignot; Anne-Rose Günzel-Apel; Sheila M Schmutz; Tosso Leeb (June 2005). "Polymorphisms within the canine MLPH gene are associated with dilute coat color in dogs". BMC Genetics. 6 (34): 34. doi:10.1186/1471-2156-6-34. ISSN 1471-2156. PMC . PMID 15960853.
- Dayna L. Dreger; Sheila M. Schmutz (Jun 2010). "A New Mutation in MC1R Explains a Coat Color Phenotype in 2 Old Breeds: Saluki and Afghan Hound". Journal of Heredity. 101 (5): 644–649. doi:10.1093/jhered/esq061. PMID 20525767.
- Leigh Anne Clark; Alison N. Starr; Kate L. Tsai; Keith E. Murphy (July 2008). "Genome-wide linkage scan localizes the harlequin locus in the Great Dane to chromosome 9". Gene. 418 (1–2): 49–52. doi:10.1016/j.gene.2008.04.006. PMID 18513894.
- Sophie I. Candille; Christopher B. Kaelin; Bruce M. Cattanach; Bin Yu; Darren A. Thompson; Matthew A. Nix; Julie A. Kerns; Sheila M. Schmutz; Glenn L. Millhauser; Gregory S. Barsh (November 2007). "A β-Defensin Mutation Causes Black Coat Color in Domestic Dogs". Science. 318 (5855): 1418–1423. doi:10.1126/science.1147880. PMC . PMID 17947548.
- Leigh Anne Clark; Jacquelyn M. Wahl; Christine A. Rees; Keith E. Murphy (January 2006). "Retrotransposon insertion in SILV is responsible for merle patterning of the domestic dog". PNAS. 103 (5): 1376–1381. doi:10.1073/pnas.0506940103. ISSN 0273-1134. PMC . PMID 16407134.
- "George Strain on Merle" (PDF). Merle Poms. Retrieved 27 October 2011.
- Sheila M. Schmutz; Tom G. Berryere; Dayna L. Dreger (June 2009). "MITF and White Spotting in Dogs: A Population Study". Journal of Heredity. 100 (Supplement 1): 566–574. doi:10.1093/jhered/esp029.
- Winge, Ojvind (1950). Inheritance in Dogs: With Special Reference to the Hunting Breeds. Catherine Roberts (translator). Ithaca, N.Y.: Comstock Publishing. p. 194.
- Little, Clarence Cook (1957). The Inheritance of Coat Color in Dogs. New York: Comstock Publishing. p. 194. ISBN 0-87605-621-4.
- Karlsson E. K.; Baranowska I.; Wade C. M.; Salmon Hillbertz N. H.; Zody M. C.; Anderson N.; Biagi T. M.; Patterson N.; Pielberg G. R.; Kulbokas E. J. III; Comstock K. E.; Keller E. T.; Mesirov J. P.; von Euler H.; Kämpe O.; Hedhammar A.; Lander E. S.; Andersson G.; Andersson L.; Lindblad-Toh K. (November 2007). "Efficient mapping of mendelian traits in dogs through genome-wide association". Nature Genetics. 39 (11): 1304–1306. doi:10.1038/ng.2007.10. PMID 17906626.
- Edward J. Cargill1, Thomas R. Famula, Robert D. Schnabel, George M. Strain & Keith E. Murphy (July 2005). "The color of a Dalmatian's spots: Linkage evidence to support the TYRP1 gene". BMC Veterinary Research. 1 (1): 1. doi:10.1186/1746-6148-1-1. ISBN 1-74661-481-1. PMC . PMID 16045797.
- Why white dogs are white Uppsala University 2014
- Baranowska Körberg, Izabella; Sundström, Elisabeth; Meadows, Jennifer R. S.; Rosengren Pielberg, Gerli; Gustafson, Ulla; Hedhammar, Åke; Karlsson, Elinor K.; Seddon, Jennifer; Söderberg, Arne; Vilà, Carles; Zhang, Xiaolan; Åkesson, Mikael; Lindblad-Toh, Kerstin; Andersson, Göran; Andersson, Leif (2014). "A Simple Repeat Polymorphism in the MITF-M Promoter is a Key Regulator of White Spotting in Dogs". PLoS ONE. 9 (8): e104363. doi:10.1371/journal.pone.0104363. PMC . PMID 25116146.
- Sheila M. Schmutz (December 27, 2008). "Coat Color Alleles in Dogs". Retrieved September 12, 2010.
- Pamela A. Davol (January 29, 2001). "B/b, E/e, and Beyond: A Detailed Examination of Coat Color Genetics in the Labrador Retriever". Retrieved September 11, 2010.
- Evans, Howard E.; de Lahunta, Alexander (August 7, 2013). Miller's Anatomy of the Dog (Fourth ed.). Saunders. pp. 71–73. ISBN 978-1437708127.
- Miller's Anatomy of the Dog
- D. J. E. Housley; P. J. Venta (August 2006). "The long and the short of it: evidence that FGF5 is a major determinant of canine 'hair'-itability". Animal Genetics. 37 (4): 309–315. doi:10.1111/j.1365-2052.2006.01448.x. PMID 16879338.
- Hayward, J. J. et al. Complex disease and phenotype mapping in the domestic dog. Nat. Commun. 7:10460 doi: 10.1038/ncomms10460 (2016)
- Hayward JJ et al. 2016 Complex disease and phenotype mapping in the domestic dog. Nat. Commun. 7, 10460
- Droegemueller, C; Karlsson, EK; Hytšnen, MK; Perloski, M; Dolf, G; Sainio, K; Lohi, H; Lindblad-Toh, K; Leeb, T (2008). "A mutation in hairless dogs implicates FOXI3 in ectodermal development". Science. 321: 1462.
- Salmon Hillbertz N H C, Isaksson M, Karlsson E K, Hellmen E, Rosengren Pielberg G, Savolainen P, Wad C M, von Euler E, Gustafson U, Hedhammar A, Nilsson M, Lindblad-Toh K, Andersson L, Andersson G (2007): "Duplication of FGF3, FGF4, FGF 19 and ORAOV1 causes hair ridge and predisposition to dermoid sinus in, Ridgeback dogs" in Nature Genetics, Vol. 39, Number 11, November 2007.
- "Vet Gen". Veterinary Genetic Services. 2010. Retrieved September 12, 2010.
- "Breeders Assistant". Premier Pedigree Software. 2009. Retrieved September 12, 2010.
- Hayward, J. J. et al. Complex disease and phenotype mapping in the domestic dog. Nat. Commun. 7:10460 doi: 10.1038/ncomms10460 (2016). supplementary figures and tables
- Hytönen et al (2008), "Ancestral T-Box mutation is present in many, but not all, short-tailed dog breeds", Journal of Heredity, Advance Access published online on October 14, 2008, doi:10.1093/jhered/esn085 
- Rimbault M, Beale HC, Schoenebeck JJ, Hoopes BC, Allen JJ, Kilroy-Glynn P, et al. Derived variants at six genes explain nearly half of size reduction in dog breeds. Genome Res. 2013;23:1985–1995. doi: 10.1101/gr.157339.113
- Schmutz, Sheila M. (March 4, 2010). "Dog Coat Color Genetics". University of Saskatchewan. Retrieved September 12, 2010.