The Reality of the B-Locus in an Actual Breed Population
LET'S PAUSE NOW for a moment while we relate our chosen example a little more closely to reality. The "B" and "b" alleles are in fact real genes. These genes influence the ability of the dog's body chemistry to synthesise a compound called tyrosine-melanin. A dog that cannot synthesise tyrosine-melanin will not have black pigment anywhere on its body. A homozygous liver-nosed "bb" individual may have dark-pigmented hair in its coat, but that pigment will be yellowish-red to reddish-brown in colour, not black. We called the B-locus a nose-colour gene only because that is where its effect is most obvious! The dominant "B" gene is the common, normal allele in the general population of Canis familiaris as a species. The recessive "b" is usually an uncommon rarity.
However, we frequently see the homozygous recessive "bb" in the show ring among some sleddog breeds. The most obvious example is the CKC/AKC Siberian Husky. "bb" dogs are called "coppers" or "reds" in that breed; these colours became popular in the show ring during the 1960s and subsequent decades. Originally they were quite scarce, an uncommon novelty. Today they are common. Unquestionably the relative allele frequency of "B" and "b" has changed over the years, or rather has been altered by the actions of breeders. There is no question of this change being anything other than a matter of fashion or fad. The breed standard of the Siberian Husky allows all colours and white; it states no preference for any particular colour. Therefore this change of allele frequency in the Siberian Husky has not been motivated by any desire to "improve the breed," despite the insistence of most breeders that such is the goal of their breeding programmes. In fact, we must consider that this particular change represents the encouragement of a metabolic defect! We said that the normal canine body chemisty is represented by the "BB" or "Bb" genotypes and in both cases the individual is capable of synthesising the tyrosine-melanin compound that allows black pigment in the hair coat and the skin. Since the "copper" Siberian with its recessive "bb" genotype is incapable of synthesising tyrosine-melanin, we could legitimately say that it is the victim of a metabolic disorder -- a "genetic defect." Of course, nobody calls this trait a genetic defect, simply because the copper coat which it expresses is a common and much-admired feature in the show ring!
There is a curious feature of the "bb" genotype that is neither well-known nor much discussed among Siberian Husky fanciers. Let's recall that the breed mythos of the Siberian Husky depicts the breed as a fast and enduring sleddog with a past history of many triumphant victories in dogsled racing. Let's also recall that the "b" allele is generally distributed among northern breeds; copper-coloured dogs are regularly seen among Alaskan Malamutes and Canadian Eskimo Dogs as well as Siberian Huskies. Siberian Huskies have regularly contributed to unregistered "Alaskan husky" racing sleddog bloodlines, so we can assume that the "b" allele is also present in that genome. Nevertheless, this author recently observed the premier speed racing event of the entire North American continent, the Open North American Championship races in Fairbanks, Alaska, where the elite of "world-class Alaskan husky" racing sleddogs gather to decide the fastest team over a grueling race in the March sunshine of twenty miles the first day, twenty miles the second day, and thirty miles the last day, using teams which typically vary from fourteen to twenty-two dogs running at speeds of an average 17.5 to over 20 miles per hour. One surprising fact became obvious at that 1998 running of the ONAC: out of more than twenty teams involving well over three hundred top-flight racing sleddogs, there was not one single "bb" dog to be seen!
If the "bb" genotype has become common and generally accepted within the Siberian Husky show dog population, yet is virtually absent among the elite of high-speed dogsled racing, the conclusion must be drawn that the homozygous recessive of the B-locus gene has metabolic effects that go beyond restriction of black pigment in the coat and skin, effects that decrease overall metabolic efficiency of the dog with respect to its capacity to demonstrate sustained racing speed over a twenty- to thirty-mile race course.
(I do not claim that "bb" genotype sleddogs are all totally incapable of racing. Here and there an individual can be found that is apparently able to participate in a competitive racing team. Nevertheless these individuals are extremely uncommon, compared to the frequency of their appearance in the Siberian Husky show dog population. It is difficult to escape the conclusion that in some way the homozygous "bb" genotype represents a metabolic liability to the dog, whether directly through the action of the homozygote genotype, or through some other gene that has been influenced by the selection process that gave rise to the "bb" homozygote.)
The Problem of Gene Linkage
THIS LAST POSSIBILITY must be examined closely, because it relates intimately to the inadvertent damage that I believe is often done by show dog breeders' selection for homozygous recessive traits of superficial appearance. Our example of the B-locus may or may not be an instance of direct genetic damage through such selection, of consciously selecting for a trait that turns out to be harmful to the physiological well-being of the dog. More significant, I think, is the possibility that, in the course of selecting for homozygous recessive traits of appearance, other beneficial genes are lost or other harmful genes are fixed.
How could these things occur? Quite simply, I think. We must not forget that the units that we call "genes" are not simple, single discrete entities. The usual popular understanding of practical genetics has led everyone to think of them in that way. Nevertheless, nobody has ever seen a "gene" as such. Scientists, when pressed to define what is meant by a gene, will usually confess that this is a convenient conceptualisation that doesn't conform too closely with the biochemical reality of cell structures. There is no structure within a cell that can be identified as a single gene! There is only genetic material, consisting of DNA strands that are found in several locations, but most importantly in the chromosomes of the cell nucleus. Chromosomes are not even visible most of the time. They become apparent during the process of cell division. It is certain that the bulk of what we refer to as genes occur on the chromosomes. However, a canine cell is said to possess 78 chromosomes. The number of individual "genes" in the canine genome (the sum total of the dog's genetic material) is thought to be on the order of one hundred thousand or more. It is important to realise that this number is only an estimate. The exact number of genes in the dog species is not known. Perhaps it cannot be known with any certainty, because it is difficult to determine the exact nature of a gene. One gene may control more than one trait, or conversely, a single trait may be influenced by a constellation of genes. The exact and particular nature of a gene defies definition.
Given the foregoing situation, it is absolute lunacy for a breeder to assume that he can selectively deal with individual genes in isolation. When gametogenesis takes place (the process by which the dog's reproductive organs generate spermatozoa or ova), the chromosomes within the nuclear material separate and divide. An individual dog possesses 39 pairs of chromosomes, each pair representing one chromosome from its sire and one from its dam. That dog, as a parent, can pass on only one member of each pair to each of its progeny. It can pass on the chromosome from its own sire, or the one from its dam, but not both. (In some cases, which we'll examine more closely below, the chromosome it passes on can be partly from one parent and partly from the other, but this is the exception, not the rule. In any normal case, any given DNA sequence passed on by a parent will come either from its sire or its dam but never from both.)
Typically, then, the smallest unit of selection that can be dealt with in any given mating is not an individual gene, but the chromosome upon which that gene happens to be found! This is of the utmost importance, yet one almost never hears this fact mentioned. No matter, anytime a breeder selects a parent animal with reference to a particular trait, he is not selecting just for that trait -- he is selecting for every single trait located on the same chromosome as the gene controlling the target trait. (Really, along with that he must accept that he is also selecting for all the other traits exemplified in that particular parent, plus those that are genetically present but not visible or expressed.) If he does not like some other trait on another chromosome of that parent (and remember that he has no direct perception of the location of the individual "genes" or of which chromosome they may be found on), he can always hope to select from among the progeny of the mating an individual that expresses the contribution of the other parent more strongly.
It is possible, through selection, to separate chromosomes from one parent or the other. It is much less possible to separate out individual genes located on the same chromosome. All the genes found on the same chromosome are effectively "linked." To separate them is feasible only if a process called crossing over happens to have occurred opportunely. Sometimes, in the course of separation and division of chromosomes within the cell nucleus, the two chromosomes of a pair become entangled in such a manner that, when they separate, the two resultant bodies are spliced versions with part of their length from one chromosome of the pair and part from the other. The probability that a given gene will thus become unlinked from the other genes on its chromosome of origin depends entirely upon its position on the chromosome. If it is located near either end of the chromosome, there is little probability of the gene becoming unlinked from the others through crossing over; if it is located near the midpoint, the probability is highest. There is no way, short of meticulous and indirect laboratory methods, for anyone to determine the probability that one gene will become unlinked from another gene on the same chromosome, nor even to determine readily that this has actually taken place.
I must apologise to the reader if the foregoing discussion has left her confused. The actual machinery of genes and chromosomes is still incompletely understood; it is still quite difficult to relate the reality of DNA sequences to the expression of visible or measurable traits in individual animals. These uncertainties are quite frustrating to the dog breeder who seeks a practical understanding of genetics as it relates to breeding and selection. Yet I think it is important to understand what we do not know, and what cannot readily be done or known, especially when we look at the selection process. If we imagine individual genes as being the size of garden peas, we should like to have some means of picking them out and moving them around that would be the size and accuracy of a pair of tweezers; whereas in reality, the selection tool at our disposal would be more like the bucket of a front-end loader! If we think to use such a tool to select a single gene and leave the others undisturbed, we are self-deceived. There is simply no way that the breeder can deal with single traits in isolation.
This is particularly true of homozygous recessive traits. The only way to achieve homozygosity where it is not already the case is through inbreeding and selection. Inbreeding is certain to influence many traits other than the target trait. Think about it: if, in order to achieve expression of one recessive gene, one must necessarily eliminate the other allele, is it not virtually inevitable that in the process of eliminating one dominant allele, other genes, particularly those located on the same chromosome, are bound to be eliminated as well? Other genes, as well as the desired recessive, fixed? Genes the exact nature and effects of which the breeder may remain totally ignorant? Think, too, about this: selection for a recessive trait quite frequently involves selecting just one individual for future breeding out of a litter; every time two parents produce one progeny that turns out to be the sole representative of those parents in future breeding, the gross amount of genetic diversity in that line is cut in half, because one individual can have only half the available genetic material from each parent!
Genetic Drift and the Closed Stud Book Breed Registry System
DOES THE READER now begin to comprehend the probable extent of genetic depauperisation inherent in the modern system of dog breeding and purebred registration? For as long as we continue to use the current system of registration, inbreeding and artificial selection, it is inevitable that the overall loss of genetic diversity will be sustained and significant. If the loss of genetic material cannot be stopped, then there is only other effective remedy possible -- and that is periodically to replace the lost material!
Yet consider further: the major umbrella registry organisations, CKC and AKC, operate under a completely closed stud book system. They make no provision whatever for periodic addition of new genetic material to the breeds they register. In every case, candidates for CKC and AKC registration are expected to descend from the original foundation stock of the breed in question -- one hundred percent. Oh, yes, the major kennel clubs accept imports from each other's stud books and from those of (some!) FCI countries -- but always on the understanding that the imports descend exclusively from the original breed foundation stock in all pedigree lines. Breeders of purebred AKC and CKC dogs apparently consider that "breed purity" is paramount, more important than genetic health, species soundness, or good temperament.
Most equine breed associations and other domestic livestock registries have "grading-up" provisions, whereby a purebred may be mated with a non-purebred and the progeny be registered (for example) as "half-Arab", a half-bred mated with a purebred to produce a "three-quarter-Arab", and the three-quarter-bred mated with a purebred to produce -- purebred Arab horse progeny! The Animal Pedigree Act 1988 of Canada provides that in order to be considered legally purebred, an animal must be at least seven-eighths descended from the original foundation stock of its breed, meaning that the grading-up process is enshrined in Canadian agricultural law, guaranteeing purebred status to progeny of three-quarter-bred and purebred parents. Yet The Canadian Kennel Club persists in requiring one hundred percent descent from original foundation stock, thereby ensuring that no breed in its stud book can ever be genetically refreshed. No provision whatever is made for new genetic input to existing breeds. The Club justifies its position by referring to the "integrity of its stud book" and the "reliability of its records" (despite the fact that it has only a much-abused "honour system" to ensure that parentage is correctly reported).
The closed stud book "breed purity" system has one undeniable major result: it is that, as long as this system remains in effect, the operation of genetic drift (and of other factors that accelerate the effects of genetic drift) can never be undone or counteracted. The net result is that each generation of AKC and CKC purebred dogs born has less genetic diversity. An inexorable tendency toward genetic impoverishment is build into the national all-bred canine registries. Presumably the ideal is for every purebred dog breed to become homozygous recessive at every possible locus, with a net genetic diversity of zero; for each individual of a breed to be a genetic laser-copy of whatever reigning show dog stud is deemed to be the living representative of the breed standard; for every dog breed to be like the cheetah, genetically depauperate, with a single homozygous genetic blueprint completely shorn of diversity, utterly vulnerable to the least change in its environment, having no possible response to new diseases, pollutants or other threats.
We are already halfway there! In many breeds it is common for the inbreeding coefficient in many litters to be on the order of fifty percent. That means the probability that any given gene received from the sire of a litter is identical by descent with the corresponding gene received from the dam is fifty-fifty -- a tossup! Every other gene is homozygous! That means half the original genetic diversity has already been eliminated from AKC and CKC dog breeds since they first began to be registered. To eliminate another half is not likely to take another hundred years, though. The effects of this breeding/registry system are cumulative; the net loss of genetic diversity probably rises on a skew curve. For many breeds it may already be too late. Random genetic drift, accelerated by sustained inbreeding and a high level of artificial selection for superficial traits, has already fixed in a homozygous state many harmful genes that in a natural, wild canid population would rarely if ever occur in such a state. Evidence that this is true is seen in the alarming upsurge in genetic diseases that have become a major concern in today's purebred dog community. Dog breeders, in their quest for exquisite breed type and absolute breed purity, have stripped their animals of the genetic diversity with which nature originally equipped the dog for his health and hardiness and as insurance against new environmental threats. How ironic it is that now, at the turn of the millennium, when all humanity are declared to be brothers regardless of race, when on every side all living creatures are most threatened by new virus diseases, widespread environmental pollutants, increased solar radiation, decreased oxygen supply, etc., those who profess love for their dogs insist upon depriving them of the genetic diversity they so badly require, and all in the name of a racist ideal which, if it were suggested in a human context, would be universally condemned.
