Genetic tests, health tests, and how each might be used to make both healthcare and breeding decisions.
by John E. Dillberger, DVM
Reprinted from the November/December 2011 Claymore.
We are in the midst of a genetic test explosion in dogs. Almost every month, a research group announces that they have found a gene responsible for a particular disease in dogs, or that they are offering a genetic test for the disease, or both.
In the excitement generated by all of this progress, it is easy to confuse a genetic test with a health test. In some cases, a genetic test is a health test. Other times, it isn’t at all clear what, if anything, the genetic test result means for a dog’s health. And even when a genetic test is truly a health test, the results have to be interpreted with caution, just as one would interpret the results of any other sort of health test.
Most of us think of a genetic test result as definitive because we are taught that a dog’s genes are its destiny. For some disease genes, that’s true—if a dog inherits a particular form of the gene, then it will develop the disease. But many disease genes don’t work that way.
A dog’s genes are often described as a blueprint for what the dog will be, including its health. But a better analogy is to think of a dog’s genes as a set of possibilities for what the dog might become. Some genes confer a very strong possibility of disease—virtually 100%—but others have a much weaker influence. In truth, what a dog becomes—its conformation, personality, and health—is not determined solely by its genes. Rather, it is how the genes interact with each other and with the dog’s environment, over time, that determine the sort of life and health the dog enjoys.
In this month’s column, I will first define what I mean by genetic test and health test and then explain how each sort of test might be used to make healthcare decisions and breeding decisions.
Not all authors use the terms health test and genetic test in the same way. Here are my definitions:
Genetic test = a test that gives information about a dog’s DNA; specifically, whether or not the DNA contains a short sequence of base pairs. The sequence might be within, or close to, a particular version of a single gene. In either case, a positive result means that the dog’s DNA contains that version of the gene.
Health test = a test that gives information about either a dog’s phenotype or its genotype, where that information has implications for the dog’s current or future health.
DNA is a long string of base pairs—or actually, many strings, since each chromosome is its own long string of DNA. A genetic test scans your dog’s entire DNA to see if it contains a specific sequence (a relatively short series of base pairs) and, if so, whether the dog has one or two copies of that sequence. Ideally, the DNA sequence that is the basis of the test is found only in, or near, a specific version of a specific gene. In that case, the genetic test tells you if the dog has zero, one, or two copies of that version of the gene.
When a genetic test detects a version of a gene that is linked to a health problem, then that version of the gene is often referred to as a “disease gene.” This term can cause misunderstanding. It is easy to jump to the conclusion that if a dog has a disease gene, then it will develop the disease, but that is not always so. In reality, the implications of having the disease gene will depend upon the nature of the linkage between the disease gene and the disease. Bear with me for a few paragraphs, because this is where things can get murky.
Here are just some of the ways in which a disease gene may be related to a disease:
• Having even one copy of the disease gene means the dog will develop the disease.
• Having one copy of the disease gene has no effect on the dog’s health, but having two copies means the dog will develop the disease.
• Having one copy of the disease gene means the dog may develop a mild form of the disease, but having two copies of the disease gene means the dog will develop a more severe form of the disease.
• Having one or two copies of the disease gene does not mean the dog will develop the disease, but it means the dog is more likely to develop the disease. This often is expressed by saying the disease gene increases a dog’s risk of developing the disease.
• Having one or two copies of the disease gene means the dog is at risk of developing the disease, but only under certain circumstances—for instance, only if the dog also has a particular form of another gene, or only if the dog is exposed to a certain environmental factor at a certain time in its life.
As you can see, knowing the result of a genetic test doesn’t tell you anything useful unless you also know the relationship between the gene and the disease!
Unfortunately, a genetic test often becomes available long before it is clear how the disease gene is related to the disease for which the test is used. We have an example in our own breed: the Factor VII test. This test identifies an abnormal version of the gene that carries instructions for making Factor VII protein, which is involved in blood clotting. Many Deerhounds have one or two copies of the abnormal Factor VII gene; however, there is no evidence that having one or two copies of the gene puts a Deerhound at greater risk of bleeding. Deerhounds with two copies of the abnormal Factor VII gene have grown to adulthood without bleeding excessively when their adult teeth erupted, when they got cuts, or even when they had surgery. Possibly having one or two copies of the abnormal Factor VII gene does increase a Deerhound’s risk of bleeding, but only under certain circumstances that are not yet understood.
If you think about it for a minute, it makes sense that genetic tests often precede an understanding of how the gene and a disease are related. Genetic research on diseases goes basically like this. DNA is collected from many dogs in a breed, some with a disease and others that are not affected. The DNA is screened for a sequence that is present in sick dogs but not in unaffected dogs. If a sequence is found, then the gene in which the sequence resides is identified as a potential disease gene. The sequence then becomes the basis for a genetic test.
But of course, it is only when the potential disease gene is identified and the genetic test is available that the researcher can begin to explore the true relationship between the disease gene and the disease. In other words, the genetic test is itself a crucial tool for asking questions about the relationship between the disease and the gene. With the test, one can begin to ask basic questions, such as whether all dogs with the disease gene develop the disease, and whether all dogs that develop the disease have the disease gene.
Most of us are familiar with health tests. In Deerhounds, for example, we use a urine test to screen for the disease cystinuria, a blood test for bile acids to screen for a liver disease called portosystemic shunt, and an echocardiogram to screen for cardiomyopathy. These are all phenotypic tests. They tell us something about a dog’s traits, not about its genes.
A genetic test also can be a health test, if the relationship between the disease gene and the disease is understood. For example, a genetic test exists for copper storage disease in Bedlington Terriers. Bedlingtons with one copy of the disease gene remain healthy, but Bedlingtons with two copies get sick. Or at least, that was the original story. As more and more Bedlingtons were tested with the new genetic test, a few dogs were discovered that tested negative but still developed copper storage disease. This led to the discovery of a second disease-causing version of the same gene. Thus, there are now two genetic tests for copper storage disease in Bedlingtons, each detecting a different disease-causing version of the same gene.
Nor is this as complicated as things can get. In humans, the gene responsible for Lou Gehrig’s disease was found some years ago. To date, more than 130 disease-causing variations of this gene have been identified! Recently, a version of this same gene was identified as playing a role in degenerative myelopathy (DM) in several dog breeds, and a genetic test is available. The situation in humans suggests that there may be other disease-causing forms of this gene in dogs. If so, then it could take many years to figure out how each disease-causing version of the gene affects a dog’s chances of developing DM. We could end up with a large panel of genetic tests for DM.
We understand that phenotypic health tests have some margin for error and have to be interpreted. For example, just because a dog has a negative urine test for cystine does not mean that the dog will not develop cystinuria. But we tend to think that genotypic health tests are different. They aren’t! The results have to be interpreted in light of what we know about the relationship between the gene being tested and the disease to which it is related. And as I have explained, that knowledge often comes long after the genetic test is available.
Using Health Tests for Healthcare Decisions
Knowing that a dog might have a particular disease or might develop the disease later in life allows us to do many things. For instance, we can make decisions about how to manage and/or treat the disease. We also get an idea of what lies ahead for the dog and can plan accordingly. We might even be able to slow the onset of the disease or head it off entirely. And, if the disease might be heritable, we can share the information with others who have related dogs that might also be at risk.
What one does will be influenced by how confident one is about the relationship between the test result and the disease. Sometimes that relationship will be crystal clear. For example, a puppy with a very high serum bile acid concentration, particularly in two successive blood tests, is almost certain to have a portosystemic shunt. On the other hand, the relationship between test and disease might be ambiguous. For example, a dog with two abnormal Factor VII genes may be at some slight increased risk of bleeding under certain undefined circumstances.
Using Health Tests for Breeding Decisions
Here is where the uncertainty gets squared. Not only does one have to consider how a test result is related to a disease, but one also needs to consider how heritable the disease is.
To illustrate, I will invent a fictitious disease called chlorophyllitis, in which the hemoglobin in a dog’s red blood cells is gradually replaced by chlorophyll, turning the blood from red to green. This begins around 3 years of age, and is ultimately fatal. There is a blood test that can detect as little as 1% chlorophyll in a dog’s blood, so affected dogs can be identified. A research group has found a specific version of a hemoglobin gene that is present in dogs with chlorophyllitis but absent in dogs that don’t have the disease, and has dubbed it the green gene. The group offers a genetic test for the green gene. Using the test, the group has screened a large number of Deerhounds and found that 50% of Deerhounds have the green gene. They also have done pedigree analysis of several large groups of related Deerhounds and found that the disease is inherited as a “recessive” trait and is not sex-linked. In other words, dogs with two copies of the green gene develop chlorophyllitis but dogs with only one copy of the green gene live normal lives.
How is a breeder to use this information? S/he might decide not to breed any dog that has the green gene. But what if many of the Deerhounds with the green gene have highly desirable traits? In that case, there are several possibilities:
• A breeder might decide to mate a dog that has one copy of the green gene to a dog that does not have the gene and then test the pups, hoping that some of the desirable traits show up in pups that did not inherit the green gene. In this way, s/he could gradually reduce the incidence of the green gene in the breeding population, without running any risk of producing an affected pup.
• A breeder might decide to mate a dog with two copies of the green gene to a dog that does not have the gene, recognizing that all of the pups would have one copy of the green gene. S/he then could take a pup that has some desirable characteristics and breed it to a dog that does not have the gene, as in the first scenario. Once again, this would gradually reduce the incidence of the green gene in the breeding population, without running any risk of producing an affected pup.
• A breeder might decide to mate two dogs that each have one copy of the green gene, hoping that some of the desirable traits show up in pups that do not inherit the green gene. There will be a 1 in 4 chance of producing an affected pup, but the pup could be identified shortly after birth and culled.
This sounds simple enough. Hypothetical situations always do. Real-life situations are a lot more complicated.
In real life, breeding decisions are not based on a single trait, but on hundreds of traits. And for most of those traits, the mode of inheritance is unknown, the number of genes affecting the trait is unknown, how those genes and environmental factors interact to affect the trait is unknown, and there are no genetic tests to tell a breeder which versions of those genes a dog has.
And that is just for traits that are visible, like conformation and temperament! One knows what sort of coat or feet a Deerhound has. But when the trait is a disease, it often is invisible until the dog gets sick. That is the situation for some very important diseases in our breed, like gastric torsion and osteosarcoma. Phenotypic health tests do exist for a few Deerhound diseases, but even then, the relationship of the test result to the dog’s disease status may be far from clear, as is the case with the urine test for cystinuria.
In real life, a genetic test gives a breeder more information about one of the hundreds of traits to consider when planning each mating. And even that information can be ambiguous, depending upon how much is understood about the relationship between the disease gene and the disease.
With each passing year, more genetic tests for disease-related genes become available to augment traditional phenotypic health tests. These genetic tests are wonderful, sometimes frustrating, and potentially dangerous.
When it comes to healthcare decisions for an individual dog, genetic tests have a big advantage over phenotypic tests; namely, they can identify a future health problem before it begins, when there is time to plan, prepare for, and maybe even prevent the disease. But that is true only so far as we understand the relationship between the disease gene and the disease. Often such understanding comes only years after the genetic test becomes available.
When it comes to decisions about a planned mating or a breeding program, genetic tests offer breeders a new tool to help produce healthier dogs. But the benefit of any genetic test is tempered not only by the extent to which we understand the relationship between the disease gene and the disease, but also by the size of the breeding population, how widespread the disease gene is, and, most importantly, by the need to make each breeding decision on the basis of hundreds of traits, including other diseases, for which genetic status is unknown.