Healthy or unhealthy? Boy or girl? These are the first questions we ask as we welcome each new puppy into the world. But face it; at the back of our minds is the thought we don't wish to utter out loud for fear of it sounding too trivial: What color?
When it comes to color, most sighthounds have a lot of choices. And while that sets breeders free to ignore color in their mate selections, it also means we tend to have more gaps about color inheritance in our breeds. Fortunately, some sighthound breeds, most notably salukis and Afghan hounds, have caught the attention of the world's foremost color genetics researchers, Sheila Schmutz, Ph.D., of the University of Saskatchewan in Saskatoon, Canada, along with her former student, Dayna Dreger, Ph.D., now of the Ostrander lab at the National Human Genome Research Institute/National Institute of Health. Dr. Dreger presented a seminar on saluki color genetics at the 2013 Saluki Club of America National specialty; the information here is based in large part on that presentation as well as Dr. Schmutz's website (http://homepage.usask.ca/~schmutz/dogcolors.html).
Coat color is surely not the most important genetic aspect of our dogs, but it's one of the most accessible. And even breeders who claim not to care about color are often heard speculating about what colors a proposed breeding will bring. Let's face it: if you don't like black-fringed reds, you'd better know you have a good chance of getting one if you breed to one! Um...well, except...it's not always that easy!
Here's why: Several genes at different chromosomal locations (loci) interact to produce the wide variety of colors we see. Dreger describes this interaction as a sort of layering effect where you can envision layer after layer of gene action affecting the dog's ultimate color and pattern. We'll look at those basic layers as first, the ones controlling the distribution of black versus red pigment; second, the ones modifying the intensity of that pigment; and third, the ones preventing any pigment and so essentially covering parts of the dog in white.
But first, some basics. Mammals have two types of melanin that make up their hair color: the brownish black eumelanin and the reddish yellow phaeomelanin. Several genes act together to determine the extent to which eumelanin and phaeomelanin are distributed throughout the coat. When you see black, chocolate or gray hairs, you're seeing eumelanin at work. When you see cream, gold or red hairs, you're seeing phaeomelanin at work. When you see absolutely white hairs, you're often the lack or dilution of both eumelanin and phaeomelanin.
These genes are located at different chromosomal locations, or "loci." At some loci there may be only two alternate forms of the gene, or "alleles." In others, there may be several different forms. Usually one allele is dominant to the other, or in the case of several alleles, there is a hierarchy of dominance. In some other cases, alleles may be co-dominant, meaning that if there is one of each allele, the color will be influenced by both alleles.
Geneticists use upper case letters to denote dominant alleles, and lower case to denote recessive ones. Each dog carries two alleles for each locus, and they are listed with a slash between them, such as J/j . If the allele isn't known, a dash is used in place of a letter: A/-. When several alleles exist at a locus, a superscript (ay/at) may be used instead of upper case versus lower case.
Base Colors: Black versus Red
At the base of the layering are what Dreger refers to as base colors. These are found at three different loci: A (agouti), E (extension) and K (dominant black).
A: The A (agouti) locus is technically called the Agouti Signaling Protein (ASIP) gene. It has four known alleles, but only two (ay and at) are found in salukis:
ay: Results in a sable coat color. The best known examples are black-fringed fawns and black-fringed reds. These dogs may have a very little amount of sabling on them, restricted to the ear tips, for example, or they may have extensive sabling over their entire bodies. What used to be commonly called mahogany grizzles are actually heavily sabled dogs, not grizzles at all. Dogs with ay are generally born dark; red sables may be almost black, and fawns are often described as being the color of wet cardboard. The black hair then gradually diminishes until the adult color is reached. In some cases, the adult coat has no noticeable black pigment, and may appear to be indistinguishable from cream (e/e, discussed later). These dogs have black vibrissae throughout life. Because ay is dominant to at, sable salukis can be either ay/ay or ay/at. <<photo 1: sable>>>
at: Results in a tan-pointed dog. Black and tan salukis as well as chocolate and tan salukis result from at. Because at is recessive to ay, a dog must have two copies of at to be tan-pointed. <<photo 2: black & tan>>
The other two variants (alleles) at this locus are not found in salukis. They are aw, which results in wolf sable type coloring with banded hairs; and a, which results in recessive black but is seen mostly in herding breeds.
So basically the A locus determines whether your saluki will be sable or tan-pointed, with sable dominant over tan-pointed.
E: The E (extension) locus is technically called the Melanocortin 1 Receptor (MC1R) gene. It has four known alleles, and salukis have all of them. In order of dominance they are:
EM: Results in a dark (usually black) mask. This is not a common allele in salukis, but is very common in Afghan hounds. The mask can be present on either a sable (ay/-) or black and tan (at/at) background. <<photo 3: black masked>>
EG: Results in the pattern known as grizzle in salukis and domino in Afghan hounds. But here's where it gets interesting. It can only make a grizzle (or domino) dog if the dog is also at/at. In other words, EG results in grizzle by modifying a dog that is already tan-pointed. A sable (ay/-) dog can have the EG allele, but it will not be grizzle in color. <<photo 4: grizzle>>
E: Allows either eumelanin or phaeomelanin to be expressed, according to what alleles at other loci signal. In other words, if a dog is ay/- (sable) then E would result in a sable dog. If the dog is at/at, E would result in a tan-pointed dog.
e: Allows only phaeomelanin to be produced when the dog has two copies of the allele (e/e); in other words, only shades of cream to red---no black or chocolate. So even if a dog is ay/ay or at/at, it will be a shade of cream (from almost white to red) instead of sable or tan-pointed. These dogs always have whitish vibrissae. <<photo 5: cream or self red>>
K: The K (dominant black) locus in technically called the Beta-Defensin 103 (DEFB103) gene. It's a fairly recent discovery as earlier researchers believed dominant black was a dominant member of the A locus family of alleles. Salukis (and Afghans) can have any of the three possible alleles:
KB: Results in a solid black dog. Technically, it is solid eumelanin, as it could also be solid chocolate, or have white on it, depending on other modifiers, but in practice the dog will be solid black or chocolate or blue, with or without white. It will not be sable, black and tan or otherwise tan-pointed. No matter what other alleles are present elsewhere, with the exception of a e/e cream genotype, the KB allele essentially overrides them. <<photo 6: solid black >>
kbr: Results in alternating jagged stripes of eumelanin and phaeomelanin---also known as brindle. These stripes can only be seen against a light background, so a black and tan dog with kbr/- would only have brindle stripes evident in its tan areas. A grizzle would have brindling most evident on its legs and face, where a sable would show brindling over its entire body. <<photo 7: brindle>>
ky: Allows for the expression of the A locus alleles. Almost all AKC salukis are ky/ky. However, some recent imports tend to be brindle or solid black, with brindle proving to be controversial in the breed.
So these are the base colors. How can we use them to describe salukis we see? Let's look at a few salukis and see what alleles they might---or in some cases must---have.
Black-fringed red or black-fringed fawn: <photo 8>This dog must have at least one ay allele, so it's either ay/ay or ay/at. The E alleles are a bit more difficult, but because it has some black hair we know it can't be e/e. It doesn't have a black mask so it's not EM. But we don't know whether it has E versus EG alleles because the EG would only have been expressed if it had a tan-pointed pattern to act upon. At the K locus we know it must be ky/ky because it's neither self black nor brindle. So we have ay/- -/- ky/ky.
Grizzle: <photo 9> We know that a grizzle dog must have at least one allele for grizzle, EG, and that EM is dominant to EG, so cannot be present. We don't know if it's EG/EG, EG/E or EG/e, so we just say EG/-. And we know that the EG allele only makes grizzle if the dog would otherwise be tan-pointed, so we know the dog is at/at. And because it is neither solid black nor brindle, we know it is ky/ky. So we have at/at EG/- ky/ky.
Black and tan: <photo 10>>To be tan pointed this dog must be at/at. It is neither black masked nor grizzle, so it cannot be EM or EG. It has black hair, which means eumelanin is expressed, which in turn means it cannot be e/e. It must be either E/E or E/e. This dog is not solid black and not brindle, so it's ky/ky. So this dog is at/at E/- ky/ky.
Cream/clear red: <photo 11> Cream and red are tricky, because they can be either e/e or E/- and ay/-. Remember that the e/e genotype effectively negates the genes at several other loci, including A and K loci. So if the dog is e/e, it doesn't matter if the dog is also KB/- (dominant black), kbr/- (brindle), ay/- (sable) or at/at (tan pointed) because it doesn't allow any eumelanin (black pigment) to be produced, so the dog will be some shade of cream to red. These e/e dogs never have any black hairs on them, even in puppyhood.
Cream or red salukis can also result from having at least one EG, E, or EM allele, along with one ay allele as well as two ky alleles. Remember most salukis are ky/ky so unless you're dealing with lines with solid black or brindle you don't have to include that in your calculations. And very few salukis are EM (masked), although many Afghans are. A black-masked red or cream Afghan will be red or cream by virtue of being EM/- and ay/-, not e/e. But a self-colored one could still be either E/- or e/e. If the dog ever had black hairs on it, it is E/-. Otherwise it is impossible to tell a cream or red caused by e/e from one caused by E/- with ay/- simply by looking.
And some creams can even be EG/-, as long as they are not also at/at. Without the black and tan pattern to modify, the EG cannot result in grizzle. Canine color genetics expert Sheila Schmutz, Ph.D., says on her saluki color page (http://homepage.usask.ca/~schmutz/SalukiColor.html): "There must be another gene, yet undiscovered, that causes cream in Saluki. In Afghans, dogs with e/e or with an E or EM or EG allele can all be cream. This seems to also be true in Saluki."
The depth of coloration, from almost creamy white to deep red, is not associated with whichever of these mechanisms causes red/cream. The depth of pigmentation must instead lie with an as yet undiscovered gene. The same is true of the coloration of the tan points on a black and tan. Some are a deep rich red; others an almost silvery white. Nor do we know what causes the depth of coloration and contrast on grizzles. Some are almost black and white; others barely have any grizzling visible, the contrast is so subtle. The best we can do is to take the old fashioned approach and say "like begets like." You're more likely to get light creams from light creams, dark grizzles from dark grizzles and so on...but nothing's guaranteed.
Modifiers: Black versus Brown
The next layer of coloration involves dilutions and modifiers. Several loci are involved, but saluki have only one that varies.
B: The B (brown) locus is technically called the Tyrosinase Related Protein (TYRP1) gene. It determines whether the eumelanin on the dog is black or brown. <<photo 12: chocolate dog>>
B: Results in black eumelanin.
b: Results in brown (also called chocolate or liver) eumelanin. This allele is one of the few that salukis and Afghan hounds don't share, as chocolate Afghans are frowned upon. A chocolate saluki is b/b, and also has liver nose leather, lips and eye rims.
For any color that has black in it, the b/b combination changes that black to chocolate, so you can have a chocolate and tan instead of black and tan, or a chocolate grizzle instead of black grizzle, or even a chocolate brindle. When a cream is b/b, then the nose, lips and eye rims are still liver.
Other loci that act as modifiers in other breeds are the saddle tan, which changes the black and tan pattern to one with a black saddle; dilution D that changes black to gray; and progressive gray G that causes black hair to gray at a young age. In salukis and Afghans, these loci are not variable, and all salukis and Afghans have the dominant member of them. And yes, there are gray or blue salukis and Afghan hounds, but according to Dreger, they are not caused by alleles at any as yet identified loci. While the gene for the D locus has been identified as Melanophilin (MLPH), and the RALY gene has been shown to impact the saddle tan patterning, the genes for the I and G loci have not yet been identified.
Another gene, dilution I, changes phaeomelanin red pigment to cream. "This may be the locus that dilutes to cream in Salukis," says Dreger. "But since we don't know what the gene is, we can't verify which variants are present in Salukis. We just know there has to be 'something' that causes the change in darkness of phaeomelanin, and for now, we can only call it Locus I, since the gene is unknown."
Spotting: Colored versus White
People tend to describe parti-colors as white dogs with spots, when in fact they should be looked upon as colored dogs with blanked out areas. Look at virtually any parti-color saluki and you can see the coat pattern hidden behind the white. Dreger suggests envisioning the dog as having a white sheet thrown over it, with holes cut in the sheet through which you can see the real color. So you can have grizzle partis, black and tan partis, cream partis, sable partis---any color you can have without white, you can have with white. Parti-colors are fairly common in salukis, but extremely rare in Afghan hounds, as parti-color Afghans are not accepted by the AKC standard.
In Clarence Little's classic 1957 work, The Inheritance of Coat Color in Dogs, he described the S, or spotting, locus as consisting of at least four different variants:
S: Solid: Allows the full base color pattern to show with minimal white.
si: Irish spotted: White is confined to the tail tip, feet or legs, and maybe muzzle and collar.
sp: Piebald spotted: Definite spots and patches are seen.
sw: Mostly white: No or few patches of color, often confined to the head or tail base.
But according to Dreger and Schmutz, salukis (along with whippets and a few other non-sighthound breeds) don't fit this classic pattern. Instead, salukis seem to fit the following pattern of inheritance:
s: piebald or mostly white
si: Irish spotted
Note that the dominance hierarchy differs from Little's, such that si is the most recessive member, and S and s can be co-dominant, meaning that S/s looks different than either S/S or s/s. And to really make it really confusing, these S/s Salukis appear to be Irish spotted! The Microopthalmia-associated Transcription Factor (MITF) gene has been shown to affect whitespotting patterns in some breeds, however, the gene for this locus in Salukis has not been identified, so the entire story is still awaiting discovery. <<Photo 13: Irish marked>> <<photo 14: extreme white spotted>>
Many, but not all, of the alleles described here have been mapped, and DNA tests exist for some of them. You can find out if your dog carries b (chocolate) or EG (grizzle), for example. This might be an interesting exercise for a popular stud dog, but for most of us, color isn't the main reason guiding our breeding choices. It is, however, the subject of much speculation and often, surprise, at whelping time---and for some of us that bit of mystery is half the fun!
Sidebar: Test Your DNA I.Q.!
Can you describe the color these dogs should be?
1) ay/at e/e ky/ky b/b S/S
2) at/at EG/e ky/ky B/b s/s
3) ay/at EG/EG ky/ky B/B S/S
4) at/at E/e ky/ky b/b si/si
5) at/at EG/EG kbr/ky B/B S/S
6) ay/at EM/e KB/kbr B/b s/s
7) You breed the following pair: at/at e/e to ay/ay EG/EG
Will you get grizzle puppies?
8) You breed the following pair: ay/at e/e to ay/at EG/e
What colors could you get?
9) You have a cream saluki with a black nose. One of its parents was chocolate grizzle. The other was black and tan. What can you surmise about its genotype?
10) You breed two chocolate grizzle parti-colors together. What colors should you expect?
1) Solid cream with a liver nose. The e/e negates the genes at the A locus, The b/b codes for brown, but because of e/e the only place it is expressed is as brown is on the nose, lips and eye rims.
2) Grizzle parti-color. The combination of at/at with EG/- creates grizzle. The s/s determines the grizzle will only show in patches with the majority of the dog being white in color.
3) Either sable or red/cream. The ay means that the EG cannot act on the black and tan pattern to create grizzle. And although it seems like the dog should be sable, some such salukis appear similar to creams and reds.
4) Irish marked chocolate and tan.
5) Solid grizzle with brindle showing up in the light areas.
6) Spotted (solid) black dog.
7) Your puppies will all be ay/at EG/e, so no, you will not get grizzles. You will instead get sables or possibly clear reds or creams.
8) You could get the following combinations of alleles:
ay/ay e/e = cream/clear red
ay/at e/e =cream/clear red
at/at e/e = cream/clear red
ay/ay EG/e = sable or cream/clear red
ay/at EG/e = sable or cream/clear red
at/at EG/e = grizzle
9) We know the grizzle parent had to be at/at EG/- ky/ky b/b S/-. The black and tan parent had to be at/at E/- ky/ky B/- S/-. Your dog must be at/at -/- ky/ky B/b S/-. Both parents were at/at because they had to be in order to be either grizzle or black and tan. Your dog's black nose tells you she is not b/b, but you know she had to have inherited one b allele from her chocolate parent. The only way that your dog can be cream in color and have been produced by these two parents is if both parents also carried the e allele, which is possible based on the colors that they are expressing. Your dog is e/e cream. Any other combination of available alleles at the E locus, in addition to being at/at at the A locus, would have resulted in your dog being black and tan or grizzle, with or without a black mask. The final genotype of your dog is at/at e/e ky/ky B/b S/-.
10) Expect a litter of parti-colors, in either chocolate grizzle, chocolate and tan or cream with a liver nose. The parents must be at/at EG/- ky/ky b/b s/s, but since they could be EG/ EG or EG/E or EG/e you don't know exactly what the E locus will code for.
Acknowledgement: Thank you to Dr. Dayna Dreger for reviewing the text and for helpful comments.