Tufts University School of Veterinary Medicine, Dermatology & Allergy Services, Walpole, MA, USA; Bizvet, Inc., Westborough, MA, USA
© 2005 Lowell Ackerman [No part of this material may be reproduced or distributed without express written consent of author].
When an animal is born, it receives half of its genetic blueprint from the sire and half from the dam. This combination of genes is what accounts for the animal being a truly unique individual, not a clone of the parents. Chromosomes contain about 50,000 to 100,000 different genes. A gene is that portion of DNA that codes for a specific sequence of amino acids, which in turn make proteins, enzymes, or polypeptides.
For any one genetic character, then, each parent contributes one version of the gene for that character, which we call an allele. The location of a gene on a chromosome is its locus.
In the Irish setter example, the PRA disease allele is called rcd1 for the specific form of rod-cone dysplasia seen in that breed. When an animal inherits the same version of an allele from both parents, we say it is homozygous for that trait. When the alleles differ in a gene pair, we say the animal is heterozygous for that trait. For an X-linked trait, affected males are hemizygous for the trait because they possess only one X chromosome, which they get from their mother. Whatever the combination, the pairing of actual genes is what constitutes the genotype.
Based on this pairing of genes, we have rules to determine the impact of genetic combinations on progeny. This is most applicable when a single gene pair determines how the trait will be expressed. This is known as monogenic inheritance, of which PRA in Irish setters is an example. Other traits, such as hip dysplasia, are caused by the product of multiple gene effects. This is known as polygenic inheritance. In a monogenic trait involving one pair of genes, four genotypic outcomes are possible, but only three phenotypic expressions, as alleles from mother and father combine in offspring. When two gene pairs are involved, 16 genotypic combinations are possible, and, assuming all genotypes are expressed equally, nine possible phenotypes (for example, coat color in the Labrador retriever). For every n gene pairs involved in a trait, there are 4n possible genotypic outcomes and 3n possible phenotypic expressions.
Because accurately predicting genotypes with polygenic traits is difficult, and because environment can have profound influences, the genetic involvement in a trait is typically expressed as heritability (h2), which is a mathematical representation of the variance in breeding values divided by phenotypic variance. The heritability of a trait can vary from 0 (no heritable component) to 1 (complete inheritance).
When two copies of a disease-causing gene (one from each parent) are required to cause a specific problem, we say that trait is recessive. Thus, PRA in Irish setters is recessive because, to manifest the disease, an animal must inherit a defective gene from both parents. If the parents of an Irish setter with PRA appear phenotypically normal, they both must be carriers (heterozygous for a recessive character) of the trait, because each contributes a disease-causing gene to their offspring. And, because the trait is recessive, both carrier parents appear normal.
When only one copy of a gene is necessary for a trait to be expressed, we say that trait is dominant. In our PRA example, the gene for normal retinal development is dominant. That's why carriers look outwardly normal even though they carry an abnormal allele as well as a normal one.
Genes don't cause diseases. They code for proteins that may have developmental or maintenance roles. A golden retriever with X-linked muscular dystrophy, similar to Duchenne's muscular dystrophy in humans, inherits a gene that codes for a defective or absent form of dystrophin protein, which, in turn, is an important component of muscle.
When we consider traits such as hip dysplasia, cardiomyopathy, seborrhea, allergies, epilepsy, diabetes mellitus, glaucoma, and colitis, no easy genotypic information is forthcoming. If we presume that the problem is likely to be caused by more than one gene, we'll label it polygenic, but most of the time, this is impossible to prove without genotype testing. If we see a condition such as epilepsy, which might affect more than one member of a breed line, we might describe it as familial, without knowing precisely how the trait is passed within that family. When a disorder appears to be more common in some breeds than others, we call this a breed predisposition. Once again, in most instances we don't know the exact mode of inheritance.
Purebred dog populations have been subjected to strong selection, which has resulted in extreme differences between breeds and decreased heterogeneity within breeds. This also explains why some lines of German shepherds have no incidence of hip dysplasia whatsoever and other lines are severely plagued by the condition. That's why quoting breed predispositions is so difficult. Problems run in family lines more than the breed in general.
Veterinary medicine has also contributed to propagation of genetic disease in purebred dog lines. Without effective treatments for conditions such as demodicosis, malocclusion, allergies, entropion, and others, many animals might not have an appearance that would make breeders rush to breed them. Medical intervention that alters phenotype positively is more likely to propagate genotype. By the same token, as veterinary medical science has reduced the frequency of infectious, parasitic, and nutritional diseases in the dog, problems of a genetic nature become more significant.
A project is under way to map all the genes in all the chromosomes in the domestic dog, creating a reference text of the canine genome. A preliminary canine map already has been constructed from more than 400 highly informative canine-specific markers, known as microsatellite markers. The next job is to determine the order and spacing of genes and traits of interest on the chromosomes of the canine genome.
Genetic diseases usually cause problems because the mutation creates a poorly functioning facsimile of the normal gene product. This often happens from innocent-appearing mishaps that result in an altered product. Considering that mutations happen by chance, that they can affect any bases in a DNA sequence for a peptide, and then they pass to future generations, it shouldn't be surprising that similar disorders in different breeds can result from very different gene mutations. That's why the DNA test for PRA in Irish setters won't work in miniature poodles. Although the final clinical result is similar, the underlying genetic disorder couldn't be more different. When you combine the incidence of mutations with the fact that 70% of all mutations are recessive, it isn't difficult to see how they can be propagated.
Linkage and LOD Scores
When two loci are very close on a chromosome, their recombination fraction is low and the loci are said to be linked. Linkage is often expressed in terms of a LOD score, which is the log of the odds supporting linkage between two markers or between a marker and a disease gene. A LOD score of three or greater is considered sufficient evidence that linkage exists. Copper toxicosis in Bedlington terriers is tested using a genetic linkage test. The test, however, does have limitations: A known affected dog in the pedigree must be available to test related dogs, the marker must have genetic variation, and occasionally the gene and marker can become separated. At this point in time, with current technology and identified markers, there is about an 85% possibility of a disease gene being close enough to a marker to detect linkage.
To be the most specific, a test might actually look at the gene sequence. This involves cloning the gene, which is an involved and expensive process, but it is an extremely accurate assessment. The PRA test in Irish setters is an assay that tests the actual gene. Accordingly, the specificity is near 100%.
That's got to be good news for breeders of Labrador retriever breeders, because labs are also prone to PRA but a clinical diagnosis can't be made until the dogs are 4 years of age, and not even by electroretinography until they are 18 months of age. Unfortunately, the specificity of DNA testing is also one of its limitations. The gene that causes PRA in Labrador retrievers (prcd) is different from the one in Irish setters (rcd1), so the same DNA test can't be used. Miniature schnauzers have yet another form of PRA (pd), and so do Siberian huskies (XLPRA), collies (rcd2), and others. In Norwegian elkhounds, two different forms of PRA (erd, rd) have been identified. Accordingly, although the DNA test for PRA in Irish setters is a breakthrough, it will take time before genetic disorders are fully identified on an individual breed basis, and tests subsequently developed. For the most part, the research is being driven by individual breed clubs that are providing funds for researchers to work on specific problems in their breeds.
1. Ackerman, L: The Genetic Connection: A Guide to Health Problems in Purebred Dogs. AAHA Press, 2000.