Good afternoon everyone. I’m going to be talking about the accepted theory of coat colour genetics as it applies to our Afghan Hounds, at least as far as I understand it.

However, if you are expecting to leave this session with a complete definitive explanation of colour genetics and the ability to accurately predict the colours in your next litter, I have some bad news for you, as what I want to be stressing throughout this talk is that the theory is just that, a theory, and is still very speculative and in some aspects is almost certainly wrong.

In talking to Afghan owners over the years I have found that there is plenty of dogmatic opinion on colour inheritance but not much scientific analysis, and at the end of this session I will be making some suggestions for improving this situation.

The aspects of Afghan Hound colour I want to talk about are-

1. The breed and colour. A look at the breed standard, its interpretation, the original imports, attitudes to various colours over the years and the difficulty in deciding what colour an afghan actually is.
2. Colour prediction. A look at the uncertainties and difficulties in predicting the likely colours in a litter and what information you need to know to do this with reasonable accuracy.
3. Basic genetics. A brief explanation of basic genetic principles and terminology.
4. What is colour. What actually causes the colours we see in our dogs and what the genes do to produce those colours.
5. The genes. This is the real nitty gritty of the subject. A discussion of the various genes affecting coat colour and their effects both singly and in combination.
6. Conclusions, Suggestions for improving our knowledge of Afghan Hound colour genetics.
7. Pedigree example. Calculation of colour genes in an actual litter using two generations.
8. References. Sources of information on colour genetics.

The Standard: The English standard says simply "All colours are acceptable". and you can’t get much clearer than that, can you? Though we all know that in the show ring at least, some colours seem to be rather more acceptable than others! The American standard goes a bit further and declares white markings particularly on the face to be undesirable, though you have to wonder why they added this as we have it on fairly good authority that white face markings are highly regarded in their native country. A white chest blaze is probably more common than a lack of it, and a small number of white hairs on the forehead, the toes and the tip of the tail are also fairly common at birth though they often disappear as the puppy grows. In theory there is no reason why say a piebald or even Dalmatian spotted Afghan shouldn’t be shown, though I hope to show you later why this doesn’t happen.

Interpretations: The Connie Miller/Ed Gilbert book describes these white areas as poor pigmentation, though I prefer to regard them as just unpigmented, without making value judgements as these markings were present in many of the early Afghans and are not associated with any health problems. The same source also goes on to criticise eyelids that are not totally black, and brown noses. This is not realistic as d dilution blues cannot have any black pigment and will have blue/grey eyelids and noses. Likewise, dogs with the b brown dilution also cannot have black noses.

Imports: Looking at the colours of the Ghazni/Bell-Murray imports, or at least reading about them as colour photos don’t exist, we find the colour descriptions black, white, fawn, brindle, red, cream, grey, dark fawn, silver fawn and gold. Black masks are also mentioned. Pre-first world war colour descriptions included silver fawn, golden fawn and blue mouse!! Whatever colour that was! Black and tan isn’t mentioned, but soon appeared in English litters, Sirdar being a carrier along with some of the bitches. It would be interesting to know for certain just what colours were meant by these descriptions, particularly grey and the various fawns.

Puppy colours: On the face of it, you would think that as colour is (or should be) unimportant, there would be little interest in the subject, but almost the first question a breeder is asked on whelping a litter is "What colours did you get"! This sometimes isn’t an easy question to answer as Afghan pups are frequently born almost black regardless of their true colour, and some colours change considerably as the dog develops, often resulting in an adult dog’s colour being quite different to the colour on its registration documents. This is something to bear in mind when studying pedigrees to try and understand colour inheritance.

Colour fashions: Because there has been little or no selection for particular colours in Afghans, you can never be sure what colours are going to appear in a litter. Having said that, there have been fashions in Afghan colours from time to time. Dogs without black masks were unpopular at one time, as were brindles. Blacks have been regarded unfavourably as there was a feeling they were hard to succeed with in the ring, and you have probably heard of the so-called ‘Red Brigade’ in the UK.

This sort of thing is mostly fairly harmless, though it must be borne in mind that if dominant colours are strongly selected against they can be lost from the breed for good.

Colour myths: There was a theory put forward about 60 years ago by a very influential dog person in the UK that brindle was evidence of outcrossing between different breeds. This was before much was known about colour genetics and before people were aware that from a genetic standpoint all breeds are virtually indistinguishable and that they certainly all share the same genes affecting coat colour. It is true to say the brindle allele has been bred out of some breeds, and can only be reintroduced by crossing with a brindle from another breed but the theory is obviously false as regards Afghans as there were brindles in the original imports from Afghanistan eg Pushum.

You often hear breeders stating that the sire (or dam) is of more importance than the other half of the mating but hopefully I will be able to prove to you that this is not the case. Certainly the offspring are more likely to favour one or the other parent if one is more inbred than the other as there is less genetic diversity in that parent, but that is not the same thing.

You also often hear people say their dog cannot produce a particular colour because that colour has not been seen in the background. By this they usually mean that they haven’t seen it in the last 3 or 4 generations. However, recessive alleles can remain hidden for any number of generations until meeting up with another copy. It is theoretically possible that colours never seen in Afghans so far could suddenly appear if dogs with very different pedigrees were mated, especially if a native animal was used. This is especially true if more than one recessive allele is needed for a particular colour. The Weimaraner colour is a case in point as I will explain later. The occasional but very rare appearance of smooth coated Afghans could also be due to this sort of thing.

Uncertainties: Hopefully, by the time I have finished, you will be able to get a rough idea what colours you will get from a particular mating, but there will always be room for surprises because of the lack of complete understanding of coat colour genetics.

Chance and puppy colour: In addition, there could be a whole range of colours possible from any given mating and which ones actually appear, and in what proportions is a matter of chance. For instance, the theory often suggests that you would get either say 25% or 50% of pups of a particular colour, but you might actually get none! The theoretical predictions would work out quite close over a large sample, say 100 or more pups, but one litter is a very small sample from a large number of possibilities. What the theory usually can predict is what colours you cannot get from a particular mating.

Hidden recessives: It is important to realise that you can never totally predict the colours in a litter just by looking at the colours of the parents. This is because the hidden recessive genes they are carrying can have a dramatic effect on the colours you will get and most of the time you won’t know what these are. I will explain hidden recessives in a minute.

You need to find out as much as possible about these hidden recessives and the only way to do this is to look at the colours of the parent’s littermates; any other litters they might have produced; their parents and their littermates and so on for at least 3 generations. Even then it is sometimes not possible to resolve some of the uncertainties.

Before getting on to the genes themselves there are some basic principles and terminology which need to be understood before you can make sense of the actual colour genetics.

Chromosomes: I’m not sure how many of you are knowledgeable on genetic principles, so I will start with the basics. Bear with me or doze off for a bit if you know this already!

The total genetic blueprint of an animal is held in a group of structures inside the cell nucleus called chromosomes. With two exceptions, there are always two matched sets of chromosomes in each cell, one from each parent, dogs having 78 chromosomes in total.. The exceptions are the germ cells, sperm and egg, which have only one set of chromosomes, 39 in the case of the dog, so that when sperm and egg unite, there is again a full complement of two sets.

Genes: Genes are sections of DNA along these chromosomes, each gene having its own specific position on the chromosome known as its locus. Each cell therefore has two copies and only two, of every gene.

Alleles: There are usually two, sometimes more, varieties of each gene, known as alleles, but only one allele can occupy that gene’s locus.

Dominance: As every animal has two alleles of each gene, they may be both the same, or they may be different. If they are different the effect of one allele will usually over-ride that of the other, in which case it is said to be dominant, the other allele being recessive to it. In the case of multiple alleles there will usually be a clear cut order of dominance and you will only see the effect of the dominant allele in the dog. However, this dominance is sometimes incomplete and an animal with one dominant and one recessive allele will show an intermediate effect between the fully dominant and fully recessive forms. The dominant allele is usually the "original" one controlling a particular process and the other alleles are mutated versions giving an altered effect.

Modifiers: As a further complication, the effect of a particular allele may vary due to other genetic factors known as modifiers. These can be either plus or minus modifiers, increasing or decreasing the effect of the allele, so much so that the effect of minus modifiers on one allele can overlap the effect of plus modifiers on the next allele recessive to it. These modifiers could be such things as timing effects on processes of embryonic development or unrelated chemical processes which happen to influence hair structure or pigment production. I must admit, these modifiers sometimes seem to me to be a sort of universal fudge factor used to make the theory work out right!

The important things to remember from all this are that: a dog can only have 2 alleles from each locus; the 2 alleles cannot both come from the same parent, and it will never be possible to explain every last nuance of colour by simple genetics because of the unpredictable nature of modifiers, among other things.

Now, to get on to the actual business of colour:

What causes colour: The apparent colour of a dog (or any other animal) is actually caused by the way light reflects off melanin granules in the hair.

What the genes code for: The genes are responsible for the chemical processes which affect the presence or absence of melanin, the type of melanin, the size and shape of the melanin granules and their distribution in both in the inner core and outer layers of the hairs. The number of variables here begins to explain why there is so much variation in Afghan coat colour.

Melanin: There are genetically really only two colours in the dog, due to the presence of two types of melanin, eumelanin which appears black or brown and phyomelanin which appears red or yellow. White is the result of an absence of melanin-the hair is actually transparent.

Location of hair: In addition to this, different areas of the dog tend to be lighter or darker than others. Upper areas, face, head and saddle usually being darker than side coat and the undersides and inside of the legs being lighter still- more so in reds than blacks.

Type of hair: On long haired breeds like the Afghan, the long hair is also usually lighter than the short hair.Hair texture and surface properties, both of which vary considerably in Afghans, will also affect the apparent colour of the coat.

Body chemistry and diet: It is also likely that hormones and environmental factors such as diet can affect colour. We have a bitch, a B.M. gold, who would occasionally produce a darker band 2 or 3 cm wide in her side coat which would slowly work it’s way down as the hair grew. Obviously some kind of internal or external environmental effect, though we are not sure what.

This is a complete list of the gene series involved in coat colour in dogs, according to Little. Most of the names more or less describe their effects.

There is some dispute about the number of alleles in several of these series, particularly A and E, and even the existence of one or two series is disputed (P and R).

All dogs have these same gene series, but some breeds only have one allele of several of them. For instance Afghans have only one allele of M, R and T, the ones preventing their visual effects, so these genes can be ignored when discussing Afghan colours and all Afghans are genetically the same for these loci.

These gene series can be divided into three groups.

1. Genes responsible for the basic colour of the animal and where on the animal these colours appear. These are series A, S and T
2. Genes responsible for colour distribution or pattern. These are series E and M
3. Dilution genes. Series B, C, D, G and P

We will have a look at each of these series in turn.

The S series controls the amount of white on the dog. There are four alleles and there is considerable overlap of these and possibly some effect due to modifying factors.

Afghans almost certainly have only the dominant S allele, though a white chest blaze is common, as are small areas of white on the feet. The famous "thumbprint of Noah" comes under this heading also, and there is occasionally a very small white tip to the tail, particularly in blues for some reason. These small white areas are probably due to minus modifiers rather than the sⁱ allele, and one would have to see substantial white areas, say all the way up the front legs and possibly even around the neck before concluding that sⁱ was present.

You occasionally see larger areas of white on the lower legs, though this is not publicised as in spite of the standard, white markings are considered undesirable by many breeders. It is also not very aesthetically pleasing as the two legs usually have different amounts of white. This probably still comes under the dominant S allele with minus modifiers.

The A locus determines the relative amounts of red and black in the coat.

The dominant allele A allows dark pigment to form throughout the coat resulting in an all black (or brown or blue) animal provided this is allowed by the E series genes present.

a^y restricts the formation of dark pigment and results in a red (or gold or cream) animal.

a^g gives wild colour or wolf grey, as in say the Keeshond or the Siberian. It produces hairs with alternating bands of light and dark pigment along their length. This allele is thought not to occur in Afghans. Though there is a theory (by non-Afghan people) that it is responsible for domino. A study of matings involving dominos could settle this and I will show how later.

a^s black saddle and tail, tan everywhere else. There is some dispute as to whether this allele really exists. This coat pattern, which is occasionally seen in Afghans, could be just a version of a^t with plus modifiers.

a^t gives bi-colour, Black & Tan when two copies are present. I personally believe the domino pattern is a version of a^t with modifiers.

The dominance pattern of the A series alleles is not complete, so a dominant/recessive pair will show visible effects part way between the fully dominant and fully recessive forms. ie A-a^y may show a reddish tinge in the side coat particularly if e is also present. The so-called sunburn seen in some blacks may be due to this effect. a^y-a^t animals will show much heavier sabling than a^y-a^y and may even have areas of black. We have two bitches of this genetic type who have black areas on the back of their necks and black tails. a^y type golds also turn lighter with age.

Some breeds seem to have a recessive black rather than the dominant black seen in most breeds, for instance GSDs. I have seen no evidence that Afghans have this allele.

There is a theory that dominant A is not in the A series as in most mammals black is recessive but this may be quite hard to prove. There is also a theory that brindle belongs in this series which I will explain later.

All Afghans are t-t

Ticking, small clumps of dark hair in light areas of the coat, doesn’t occur in Afghans so we can ignore it. Though this is a bit surprising as it does occur in Salukis and also I believe, the Kirghiz Taigan, probably also a close relative of the Afghan.

This is the gene which causes the spots on the Dalmatian, and the reason we don’t see spotted Afghans is that they do not have the dominant T allele.

** In discussions after the Congress Garry Sinck told me he had owned an Afghan with a white foreleg which did show ticking, so it looks as though the above statement is not correct!

This series is where most of the controversies appear.

According to Little:

E^m is a sort of super dominant causing a black mask whenever it occurs. The extent of the mask varies a great deal, possibly due to modifiers. This allele also allows black to be expressed if specified by A locus genes.

E allows black to form wherever specified by alleles at the A locus.

e^br causes black stripes to appear in any red areas the animal may have. This effect also varies a great deal, from almost all black with just a few small areas of red to the other extreme where the black stripes are hard to spot. Again supposedly due to the magic modifiers!

As various dilution genes affect black and red differently, a wide range of colour combinations are possible in brindles.

e when present in double dose prevents black appearing anywhere in the coat. In an otherwise black and tan animal, there will be two different shades of tan in the normal B&T pattern.

As you can see, E^m and e^br seem to have different sorts of effects compared to E and e. It has been suggested by a number of people in other breeds, based on their personal experiences, that E and e are the only alleles at this locus and the other two are both located in other series. It would be interesting to see if comparable data could be found for Afghans. I will come back to this later.

All Afghans are m-m

You will all have probably seen the effect of this gene in the Australian cattle dog and the Great Dane.

The dominant merle gene M causes irregular patches of diluted dark pigment within which are blotches of undiluted dark coat. In double dose, it is associated with serious health problems.

Fortunately, Afghans have only the m allele.

According to Little, b does not exist in Afghans, though it certainly exists in Salukis. But I am reliably informed that chocolate Afghans have been seen, and several of the early English Afghans were described as Fawn (or silver fawn or even blue mouse) in pedigrees and the name fawn is usually used for the colour of Weimaraners which are b-b plus blue dilution d-d. Does anyone here think they have seen an Afghan this colour? As in Weimaraners, these dogs would have liver noses and yellow eyes which are also seen in Afghans, but the eye colour may be due to other factors than this gene.

This series is also not fully agreed on- there is some dispute about just how many of these alleles actually exist. Again, to give the Little explanation first:

C causes intense pigmentation, both red and black.

c^ch dilutes red to gold or cream but has no little or no visible effect on black. It could be that the Afghan colour known as black dilution has something to do with this allele.

c^d extreme dilution giving an all white animal with black nose and eye rims. This is sometimes seen in Afghans but is fairly rare. Khanabad White Warrior could have been one of these. Several of the early imports look and sound from descriptions as though they could have had this genetic make-up. There is some dispute about whether this allele really exists as the same effect could be caused by c^ch-c^ch with minus modifiers or by the action of extreme white spotting , though I think microscopic examination of the hair would reveal the difference between an extreme dilute cream and a true white.

c^a albino. This is extremely rare in dogs.

This locus is reasonably straightforward. In the case of a d-d animal, all black pigment, including nose, eye rims and pads is diluted to blue. The shade of blue is considerably affected by other genes though. This gene also affects red pigment, tending to dilute it to a pale silver though possibly only when c^ch is also present. This is often seen quite clearly in blue brindles.

When present with brown bb it causes the Weimaraner colour.

This gene causes the dilution seen in Kerry Blues. Dogs with G are born almost black and slowly dilute to blue/grey as they age.

The experts say G doesn’t exist in Afghans, but the colour changes seen in blues suggest that it does, and I believe it is probably responsible for black dilution. It is also one of those genes with incomplete dominance, so a G-g dog would show some lightening, though not as much as GG. Combinations of G and d could give rise to a huge variety of shades of blue, and in the case of brindles there would be even more possibilities due to the action of cch on the red.

There is apparently evidence in some breeds for the existence of a gene for intense pigmentation but I think it would be very difficult to separate its effects from the various dilution genes at other loci. It is probably best to ignore this with the current state of knowledge of Afghan colours.

The major area of contention regarding the Little theory is the E series. There is a strong suspicion that brindle is not in this series, and only last week I heard of an Afghan mating in which a B&T x cream produced a litter containing brindles. If brindle is in the E series this is impossible. With Afghans at least, you always have to be careful that brindle stripes may be present but very small in extent and/or very dilute and hard to see, but I understand that is unlikely in this case. A study of the dogs behind the sire and dam should settle this.

A large scale analysis has been done on greyhound colours by the keeper of the Australian & NZ greyhound studbooks, which contain data from about 30,0000 litters! This study seems to strongly suggest that brindle is in fact a member of the A series. This could make sense as this series controls the relative amounts of red and black. What the greyhound folks have done is analyse a large enough number of matings to obtain a statistically significant result. This is just what we need to do in Afghans. The key thing the greyhound study has shown is that blacks mated to fawns can produce blacks, fawns or brindles, but not fawn and brindle in the same litter. It would be interesting to see if this holds up for Afghan matings also. Unfortunately the fact that we see black and brindle afghans suggests that brindle is NOT in the A series.

The standard theory for the e series can be tested by looking at matings between B/M brindle and self masked gold sables. This should result in only masked or brindle dogs, but not the colour of either parent. Likewise, B/M brindle to B/M brindle must produce either brindle or masked offspring.

To try to establish the genetic makeup of domino it would be most interesting to see the result of domino x domino and domino x B & T matings. If domino is caused by a specific allele then the usual mendelian rules will apply and study of a large number of matings involving domino should make it possible to work out what allele is responsible and where in its series it belongs. If domino is black & tan with modifiers then there will probably be no particular inheritance pattern observed.

I would like to conclude by making an appeal to breeders to make information on colours they have produced available for study so we can try to resolve some of the unknowns. It is only by studying a large number of litters, together with accurate information on the colours obtained that the true nature of Afghan colour genetics can be determined.

As an example of how to work out the colour genotype of a dog and bitch and of their offspring, I have here the colours of a real mating. To start off with, I have listed the alleles which must be present to give the colours we see.

We can then fill in many of the initial blanks by making comparisons:

First, as the sire is d-d, all the pups must have at least one d allele

As one of the pups is a c^ch-c^ch cream, both parents must have this allele

As several pups have black masks and the sire doesn’t, then the dam must have Em

This mating would tend to rule out the possibility of C-c^ch being responsible for black dilution as in that case the dam would also be this colour

As the domino has e^br from the sire, she must have E from the dam, as she doesn’t have a mask and does have a mostly dark coat. Her dam must then be E^m-E . Likewise, the dam must have one copy of d to produce a blue domino pup. If domino is actually a modified B&T then both parents must be carrying a^t and all a^y pups must also.

This mating can also be used to demonstrate Mendelian principles. What we have is a black mated to a red. (a brindle is basically a red with striped overlay). We would expect 50% black and we get 6 out of 10 - not bad. If the brindle has only one e^br allele we would expect 25% brindle; as we can’t have 2½ puppies, 2 out of 10 is quite close!

This slide shows the next generation with the BM cream bitch from the previous litter mated to a BM black & brindle dog. We can go through the same process here. As there are several very dilute creams, the sire must be carrying c^ch

Virtually all writing on the subject of dog colour genetics is based on Clarence Little’s book, first published in the 1950’s, analyzing the results of a research programme started in the early 1900s. As far as I know, there is no more up to date research data than this. His conclusions were rather tentative as the research was still in progress, but no further results seem to have been published. More recently a number of breed specific books and articles have appeared which argue for changes to Little’s theories and I will be discussing these as we look at the genes in detail.