Mythunderstanding DNA Tests
First published in Double Helix Network News, Fall 1996

by C.A. Sharp

Genetics used to be an arcane little science that took up very little of our time once we got through high-school biology. Even those non-scientists who dealt with it from time to time, like dog breeders, didn’t concern themselves with learning much beyond simple Mendelian traits, like how to get red merles or how not to get hemophilia. And you never had to pay for it.

In recent years the explosion of genetic knowledge and technology has brought genetic science front and center, providing a variety of products and services that touch the lives--and pocketbooks--of the general public. Dog breeders find themselves faced with decisions on whether or not to have their dogs DNA tested for various things Those letters--DNA--are uttered regularly on news programs and are likely to turn up almost anywhere in your daily newspaper, with the possible exception of the real estate section.

The problem with all the mass media coverage and subsequent word-of-mouth between people who suddenly find themselves (or their dogs) personally connected with the subject, is that misinformation and misunderstandings abound. This article is an attempt to dispel some of the misunderstandings which have surfaced among Australian Shepherd people.

If I send samples on my dogs for DNA testing to a club-sponsored program, someone connected with the club might leak information about my dogs' genes.

Polly Politician, who currently sits on the board, is out to get you because your dogs always beat hers in competition. Unbeknownst to the staff or anyone else, Polly sneaks into the office late one night and pulls up your dog's DNA profile. What dirt will she find there? Can she tell that Blue's great-grandma had cataracts?

No. The Aussie DNA testing programs currently in operation are for paternity testing only. The test searches out a set of pre-selected "markers," short segments of DNA. The entire set of markers makes up only a very tiny portion of the dog's total DNA--less than one hundredth of a percent. It provides no information at all about other DNA, including that cataract gene that might or might not have been passed down three generations to Blue.

Polly's efforts will be a complete waste of time--unless you have been sloppy or deceitful in your breeding practices. But if that's the case, the registry will already be in contact with you. The DNA paternity testing programs currently open to Aussies--whether through ASCA, AKC or UKC--cannot tell anybody anything except whether your dog's parents are or are not as represented.

Even as specific tests for various hereditary disease are developed, there is little chance anyone could access the results to do you dirt. Disease testing will be done by laboratories. You will either have your vet draw a blood sample or mail in swabs or other tissue sources and the lab will run the test and give you or your vet the results.

The most wonderful thing about all this DNA testing that the clubs are doing is how it will someday end genetic disease!

Unfortunately, it won't. The current testing programs are for paternity only. The markers used aren't even genes. (Not every part of the DNA is.) No record is being made of anything but the markers, so they can't be used in the search for disease genes.

The only possible windfall for genetic disease research is the fact that ASCA continues to store the blood samples submitted to its testing program. If it is maintained, that "blood bank" has the potential for providing data to researchers. But if ASCA should stop using blood, or decide not to keep the stored samples, the potential resource will be lost. If ASCA starts allowing samples other than blood, as is very likely to happen, and most Aussie owners opt to use something else, then their dogs will not become part of that "blood bank," because hair and cheek swabs don't provide as much tissue as a blood draw and little or nothing will be left after the paternity test is run.

I know Joe Trialer's dogs are actually cross-bred with one of the hairless breeds. Everybody knows hairless dogs don't have pre-molars and neither do Joe's. I want to make him DNA test his dogs to prove what he's done.

Whatever caused Joe's toothless wonders, the only thing DNA testing will prove is whether the parents of his dogs are or are not as listed on the papers. As long as every dog he turns in for paternity testing is out of the parents it's registration lists, they will pass.

While it is theoretically possible that every single breed might have a unique DNA sequence somewhere in its genome, it will be a very long time before genetic researchers have the time or the money to pursue that kind of trivia. At present it is difficult to distinguish pure wolf from wolf-hybrid from dog. Distinguishing one breed from another--especially when one of those breeds is as recent in development as the Aussie--is a much more difficult task.

All the time we hear about "false positive" and "false negative" lab tests. Some of this DNA stuff has to be wrong sometimes, too.

False readings are very unlikely. Paternity tests use multiple markers to minimize the possibility of error or mutation skewing the result. The possibility that the test result would be incorrect is too minuscule to cause any serious concern.

Human error, however, is another matter. As with any test, if the sample is mishandled the results can be incorrect. Blood samples, due to the amount collected and method of collection, are least prone to contamination or mislabelling. Those who do their own sampling, via cheek swabs or hair pulls, need to follow the directions provided and take care that each sample is packaged and labelled before the next is collected.

There are people who fake OFA and CERF tests all the time. They'll fake the DNA tests, too.

Suzy Slimeball knew how to deal with all this bothersome testing. Her black tri stud dog, Gonzo's Got It, was blind as a bat and had lousy hips. Not to worry, her next-door-neighbor, Denise Ditzy, had a black tri dog with lousy conformation and temperament but excellent eyes and hips. Suzy borrowed Killer whenever she needed to have Got It tested. She also used Killer once on a bitch someone had shipped to her for breeding because Got It wasn't in the mood.

When the litter was born the bitch owner got suspicious. Not to worry, said Suzy, Killer can stand in on the DNA test, too. All went well for a while, but about a year later another bitch owner bred to Got It. He was in the mood this time. But two days later, so was the bitch owner's male, who convinced her six-year-old he really needed to get out of his crate. The bitch owner had the whole litter tested to figure out whose pups were whose. Both males and her bitch had already been tested. Funny thing, three of the seven pups were sired by her own dog, but the other four didn't match up with Got It. The bitch owner demanded an investigation. Funny thing, Got It’s DNA turned out to be different than what Suzy had originally submitted.

Not to worry, Suzy Slimeball gave up breeding Aussies (those registry people were so nasty) and switched to a rare breed that wasn't yet associated with any major registry.

There are a bunch of labs that do DNA paternity testing. I have the right to use whichever one I want.

Yes, you do. BUT...the problem is that you can't compare the results you get at your chosen lab to the results you get from any other lab. Dogs have so much DNA to choose from that literally dozens of sets could be chosen by different labs with no overlap.

At present, there is no standardized marker set for canine paternity testing. Each lab uses different markers, so comparing results from one to the other is comparing apples to oranges to rutabagas. If your dog was tested at one lab and he's bred to a bitch tested at another, the pups would have to be tested at both labs to verify each parent separately. If the owner of a pup wanted to use a different lab than either the bitch or stud owner, the results would be meaningless because there would be noting to compare the puppies’ results with.

Sometime in the next few years one of two things is going to happen to standardize all canine paternity testing. There is an international committee of scientists who have been given he task of selecting standard sets of markers for different species of domestic animal. They've already completed the one for cattle. From all indications, they plan to choose "public domain" markers--meaning nobody owns them and can therefore charge licensing fees for their use--so any lab anywhere in the world could do a test compatible with one done anywhere else. Prices would be determined by market forces--meaning if you didn't like the price one place you could go to somewhere cheaper.

That's the first possibility. The second is that the commercial lab that is currently striving to "corner the market" in canine paternity testing by signing up as many major registries as possible may actually succeed in obtaining a virtual monopoly before the standard set of markers is selected. This company's markers are not public domain, so they could either keep all the business for themselves or charge other labs for the use of the set. They get to set their price and we get to pay it.

All that stored blood in ASCA's data base ought to be telling us more than who the sire and dam are.

Under the right circumstances (enough samples, willingness of owners to cooperate) the collection of blood samples ASCA is maintaining could at some point provide useful information about a number of genetic issues. However the actual data base is not the blood, it's a computer record of the paternity markers found in each sample. The club receives hard copies of the test results (the markers each dog has) but the computer data base is maintained by the lab. It is used not only for record-keeping, but for running comparisons between a tested dog and it's alleged parents. In theory, once enough data is gathered, one could even compare a dog of unknown parentage to all the animals in the base to identify it's actual parents, if they had been tested.

Donald Deceiver is telling everyone that his dog, WTCH Grabem Slabem is free of genetic defects. He says Slabem's DNA test came up with nothing but perfect, normal markers. He showed me the test results, but it's just a bunch of letters and numbers.

Maybe Slabem's test did have nothing but "perfect" markers, but no test currently available can tell anyone everything about a dog's genetic makeup. It is unlikely that such a test will be available for many years. Slabem's markers may tell you that his parents are as represented or that he doesn't have one of the very few diseases for which DNA tests are available (none of them specifically designed for Aussies) but there is no way to completely clear any individual dog of carrying unwanted genes.

Understanding what markers are and why they are used is also important. Markers are not genes, but sequences of DNA that occur at specific places (loci) and, ideally, have several different forms (alleles.) Each locus will be given a designation (A, B. C, etc.), as will each allele (1, 2, 3, etc.). The designations are not detailed descriptions of the markers, those would looks something like this: TCAGGGACCTCAGCAGCAG....only very much longer. Now you can understand why they use a shorthand designation.

In the case of paternity testing, the sire's and dam's markers are compared to the pup's. If the pup has markers that could not have come from one (or both) of its parents, then its breeder was either very sloppy or dishonest. The use of multiple markers eliminates the need to worry about mutations causing a mismatch. One marker might have changed, but two is unlikely, and mutations in more than that at the same time is so improbable that it can be dismissed as a practical impossibility.

With disease testing the technique is a little different. Genes can be pesky little things to pin down, though with work one can figure out approximately where they are on the DNA. Once this is established, markers with multiple alleles are selected that are on either side of the area of the gene locus. Because of their proximity (and the closer they are, the better), a particular set of alleles will nearly always be associated with the disease gene. Those markers may differ from breed to breed, or within major sub-populations of a single breed (which is why we need Aussie-specific tests). DNA does something called "recombine." It's a shuffling of the genetic material before it gets separated out for distribution in sperm and eggs. This happens with good-sized chunks of DNA rather than tiny pieces, so the markers selected for any given gene will almost always go with it. The selection of multiple markers helps determine whether recombination might have separated the gene from its markers. If there are changes in the more distant markers, but those on either side of the gene locus are unchanged, the gene is there. But if one of the adjacent markers has changed, there is a possibility that the gene has also moved.

Slabem’s markers may be "perfect," but Donald's promotion of his dog is anything but.

Drawing blood for DNA tests is a nuisance and costs too much. Cheek swabs or hair samples are a whole lot easier, cheaper and just as good.

Easier and cheaper, yes. Just as good? Not necessarily. Only a small amount of tissue is needed for the test, which is why a little pulled hair or a cheek swab will work. However, if something goes wrong at the lab (not a common occurrence but it can happen--people drop things, flip the wrong switch, the power goes out, etc.), there may not be enough material to do a second test. With a blood draw, there is plenty.

Someday there will be a set of standard markers for paternity testing. When that happens, if they are different from what a dog was originally tested with the dog will have to be re-tested in order to remain a useful part of the data pool--if it's still alive. But, if there is stored blood on file, all the dogs, living, dead or just plain missing, whose blood has been kept can be re-evaluated with the new markers.

###

DNA testing for paternity and disease will become commonplace as years go by. But now, while it’s still so new, we dog-breeding consumers of this new product need to read the "labels" carefully so we understand what it is--and isn't, and what it can--and can't--do.

Copyright 1996 C. A. Sharp. All rights reserved. C.A. Sharp is editor of the "Double Helix Network News", the quarterly newsletter for those interested in genetics and hereditary disease in the Australian Shepherd.

Return Home