Osteosarcoma Genetic Screening Project Achieves a Milestone
By John Dillberger, DVM
Last month we reached a couple of important milestones on the Osteosarcoma Genetic Screening Project. That's when I received the 200th Deerhound blood sample and the 20th sample from a dog with osteosarcoma. The second milestone is especially important because it gives us enough samples from affected dogs to begin the search for a genetic marker.
How significant are these milestones? What do they mean for you? Simply this: IF there is a single gene that increases a Deerhound's risk of osteosarcoma, then we could have a genetic screening test for osteosarcoma in our breed before the end of 2004. Five years of work collecting and banking DNA samples have given us the raw material we need to find out if there's a genetic basis for osteosarcoma in our breed and to produce a simple test for it.
That's really exciting news! Imagine being able to tell from a cheek swab or blood sample if a dog's genetic makeup gives him an increased risk of developing this cancer. Imagine knowing when you decide to breed a dog whether or not he's at risk for a cancer that, if it does occur, won't show up until years later. Imagine being able to test his offspring within days of birth to see whether or not they inherited the genetic factor. Imagine being able to identify the gene itself, learn how it works, and develop a way to prevent osteosarcoma from ever occurring or treat it effectively when it does.
I cannot overemphasize the magnitude of what we've accomplished. Taken together, the DNA bank and pedigree and health information that we've assembled is a unique scientific resource with great potential for unlocking some of the secrets of osteosarcoma-and not just osteosarcoma, but other canine health problems, too. Researchers at North Carolina State University are so excited by the possibilities that they want to make our project the poster child for their new Genetics Research Institute and have recruited a graduate student in veterinary genetics solely to pursue this project. And with this kind of interest comes not only the expertise of top researchers, but money. Our work over these years is why we now can attract such attention and resources.
Given that other breeds have thousands more owners and dogs than ours, why did we achieve what they haven't? In plain words, we had a fortunate conjunction of circumstances. First, I've spent most of my professional life in cancer research as a pathology resident, graduate student, and pharmaceutical industry scientist. That's not only kept me abreast of emerging knowledge about the genetic basis of cancer, but has also kept me in contact with the research community. Second, the past decades have seen the development of new techniques in molecular biology and genetics that make it possible to create genetic maps and genetic tests. Third, our breed has a single dominant form of cancer that occurs (unfortunately) with a fairly high incidence-a fact we all know by experience but that was confirmed and quantified in the Deerhound Health Survey. Fourth, we have a high percentage of owners and breeders (That's you, reader!) who have been willing to put their money (and time) where their mouth is-to get samples, provide information, and pay to have samples banked. And fifth, we had the foresight to conceive of this project and the patience and stamina to embark on it with no guaranteed results and a timetable measured in years.
Well, success still isn't guaranteed, but the timetable has shrunk to months instead of years. And don't think that our contribution is over-far from it. But from this point on, we're not alone. We've got some of the best canine geneticists, oncologists, and molecular biologists as partners.
In case you're new to Deerhounds, I'll briefly recap the history and goal of the Osteosarcoma Genetic Screening Project. The project is based on the suspicion that a single gene greatly increases the risk of osteosarcoma in Deerhounds, so that dogs that inherit the gene are likely to get osteosarcoma and dogs that don't aren't. On that assumption, the SDCA began in 1998 to collect and 'bank' (store) DNA samples from Scottish Deerhounds that could be used to develop a genetic screening test for osteosarcoma risk.
Deerhound owners and breeders have been submitting blood samples, health information, and pedigree information to me. I enter the information into a database and deliver the blood samples to the laboratory of veterinary oncologist Marlene Hauck at the North Carolina State University Veterinary School, who has been our partner in this project since it began. In Dr. Hauck's laboratory, DNA is harvested from white blood cells in the sample, separated into two vials, and stored in separate freezers, just for insurance.
We've now got samples from over 200 hounds stored at the North Carolina State Veterinary School. Among those are samples from 20 dogs that developed osteosarcoma and about 30 dogs over 11 years old that didn't. Because many of the samples are from young hounds, some of whom will develop osteosarcoma as they age, we can expect to have more samples from affected dogs as time goes on. But the important point is that we've now got enough of the right sort of samples to begin looking for a genetic marker.
To translate our DNA samples into a genetic screening test will take expertise and money. Fortunately, we're at the right place at the right time. While we've been collecting DNA samples, NC State University set about establishing a Genetic Research Institute to investigate the genetic basis of animal diseases. Dr. Hauck is excited about making our project one of the first that the institute tackles. To that end, she's recruited graduate student Dr. Jeffrey Phillips, a geneticist who will devote much of his time this year solely to the osteosarcoma project. Moreover, she's brought in Matthew Breen, newly recruited to the NC State University and a co-investigator in an ongoing study of the genetics of osteosarcoma in several other breeds.
Dr. Hauck also is actively seeking funds and collaborations to support Dr. Phillips's work on our project. Potential funding sources include internal university money for cross-departmental activities, grants available from the American College of Veterinary Internal Medicine, AKC Canine Health Foundation grant money, and funds from our own SDCA. Among the more important collaborators will be Dr. Matthew Breen, whom I mentioned already, and Dr. Elaine Ostrander, who has been instrumental in creating the canine genetic map and is actively investigating the genetic basis of osteosarcoma in other breeds.
How to Help
If you have a dog that develops osteosarcoma, then please contact me for instructions on providing a blood sample for DNA. Each sample we get from an affected dog increases our chances of finding a genetic marker.
If you've provided samples already, then expect to be asked soon for a health update on the dogs from which the samples came. Also, don't be surprised if you're asked to provide an additional blood sample from a dog, if he's alive. As samples analysis begins, we may find that we'd like more DNA from a particular dog than we have in the bank.
If a dog in the DNA bank develops osteosarcoma, then you may be asked to provide a sample of the tumor itself. Those tissue samples could be valuable when it comes to searching for a particular gene involved in osteosarcoma. A tissue sample would be taken at tumor removal surgery (if that's done) or when the dog is euthanized. Your veterinarian will receive instructions on how to handle and ship the sample, and costs will be reimbursed through grant money. Genetic Maps, Markers, and Screening Tests
For those interested in knowing more about how genetic screening tests are developed, I'll give a brief explanation. If you want to know more or if my explanation confuses more than it helps, then don't hesitate to call or write. I'll be happy to discuss further.
By the late 80s, researchers were beginning to use molecular genetic techniques to create genetic maps for humans and various animal species, including dogs. These maps were based on short unique stretches of DNA called genetic markers. (There are other names for these unique stretches of DNA, but I'll call them markers for this discussion.) Researchers could determine on which chromosome each marker resided, in what order the markers on a given chromosome occurred, and approximately how far apart they were. You can think of each chromosome as a very long string of Christmas tree lights, with each bulb a potential marker that lights up when it is discovered. At first relatively few markers were known and they often were far apart, so that large stretches of the DNA were dark. Over the years, more and more markers have identified so that now the entire long length of DNA is lit up with them.
Don't confuse markers with genes. Genes are long stretches of DNA-well, much longer than markers, at any rate. A marker can lie inside a gene or between genes. A single gene can contain many markers.
One way to visualize the relationship of genetic maps, genetic markers, chromosomes, and genes is to imagine the genetic map as a map of a very long Greyhound bus route (I couldn't resist the sighthound analogy.) that runs from coast to coast. Each state that the route traverses is a chromosome, each county it traverses is a gene, and each town is a genetic marker. Thus, each state (chromosome) contains many genes (counties), and each gene contains many markers (towns). For this analogy, we'll consider that genes lie side by side without any space between them, although that isn't actually true.
Your bus route map shows the location and name of each town, but doesn't show the county lines. You can't tell where one county ends and the next begins. In other words, you know the location and name of each "marker" on your map and the distance between them but have no idea about the location or size of any "gene."
If your map shows only towns with populations over 100,000, then it will have few markers and little detail and will be like early genetic maps. Some genes (counties) won't have any marker in them at all. However, if your map shows towns with populations over 100 people, then it will have many thousands of markers and great detail. Most genes will have at least one marker in them and many genes will have dozens. That's where the dog genetic map stands now-it's quite detailed with several thousand markers.
Now imagine that you're a medical researcher who suspects that people who live in a particular county are more likely to develop osteosarcoma. You'd like to test your suspicion and, if it's correct, then you'd also like to locate the county so you can study what it is about living there that puts people at risk. Unfortunately you can't ask people the name of their home county because they don't know-but they do know the names of the nearest towns with bus stations. So you interview 20 people who have osteosarcoma and 20 people who don't and collect bus-station town names. If your original suspicion was correct, then you'll find a few town names (markers, you'll remember) that show up again and again in the answers from people who have osteosarcoma but rarely if ever in the answers from people who don't. You might even find one town name that osteosarcoma-sufferers always name but that's never mentioned by people free of the disease.
Now you can go to your bus-route (or genetic) map and locate the general area in a specific state (the general region of a particular chromosome) where this dangerous county (gene) lies. With time and a little luck, you may be able to identify the county itself and pin down its boundaries. If so, then you can begin to investigate why people from that county are so susceptible to osteosarcoma. You'll be able to generate ideas to reduce that risk or eliminate it altogether, and you might learn things that would help you treat osteosarcoma more effectively.
But you don't have to do any additional research to start identifying people who are at risk of developing osteosarcoma-you can do that immediately by looking at the towns named by osteosarcoma sufferers and picking the name that shows up most often. Then to determine if somebody is at increased risk of developing osteosarcoma, you need only ask them if they live in or near that town. The closer they live to it, the more likely it is that they also live inside the dangerous county. What you've got is a simple, one-question, screening test for osteosarcoma risk.
What I've just described is the principle behind most of the genetic screening tests developed to date for people and dogs: map markers in DNA from affected and unaffected individuals, identify a marker that's closely associated with the disease, and then use that marker as the basis for a genetic screening test. In a few cases, the responsible gene itself has been identified and the genetic test actually looks for the gene instead of a marker, but if the marker lies inside the gene then a marker-based test is as accurate as a gene-based one.
Believe it or not, we've finished the difficult part of our search for an osteosarcoma genetic marker. There is no shortcut to collecting an adequate number of DNA samples to start the search, and we've accomplished that. Now it's a matter of applying research techniques that are well developed and readily available.
I haven't addressed in this article how we'd use a genetic screening test for osteosarcoma if we're lucky enough to have one by the end of 2004. That's a subject we should think about as individuals and debate when we get together, especially at this year's National Specialty where Dr. Gerald Bell will speak very much to this very subject. Among other things, we'll likely want to undertake some sort of pedigree analysis to shed light on the mode of inheritance of the osteosarcoma marker/gene in our breed. But that's a subject for another column...
Copyright 2003 by Dr. John Dillberger, P.O. Box 910, Creedmoor, NC 27522-0910. Reprinted with permission. All rights reserved.
For more information on osteosarcoma in Scottish Deerhounds, or in other breeds of dog, please click on the links on the right.