Link tests for PRA

IN recent years animal breeders witnessed the arrival of DNA link tests. These tests were received with mixed reactions. Some breeders preferred to welcome the new technology, accepting any result obtained in good faith while expressing relief that heritable disease may soon be eliminated from the breed. Other breeders rejected link tests and proclaimed an age-old truth; that a good breeder will keep heritable disease at a minimum through careful mate selection.

Both reactions to link testing represent extremes of opinion and are responses that arise naturally when information is either lacking or difficult to digest. The fact that link tests do not test for a particular gene or disease is the most common cause of confusion among dog breeders and needs clarification.

Progressive retinal atrophy (PRA) is a late onset disease affecting many dog breeds and produces retinal damage culminating with total blindness. At some point in canine ancestry a gene that was critical for the development of normal retina (this gene hereafter referred to as the ancestral gene) mutated. The mutation probably occurred in a single animal although similar mutations may have happened throughout history. The mutation in itself was not life threatening and produced no ill effects either in its founder or immediate progeny. Some progeny did of course carry this gene and it may have been many generations before two mutant genes were paired in a mating between two carriers ... the outcome being a dog that expressed PRA and became blind.

During the original PRA mutation event, the fragment of DNA that codes for the ancestral gene became smaller, splitting into two DNA fragments. One that codes for the PRA gene and another smaller fragment (the link fragment) that probably serves no function at all. It is this small non-functional fragment that the PRA link test is looking for.

Researchers have found that the onset of PRA disease and the link fragment of DNA are statistically associated. That is, a positive test for the DNA fragment is found in dogs that later develop the diseased state of PRA. Hence the PRA link test does not test for the PRA gene itself, being positive only for a smaller fragment of associated DNA.

It would be easy to conclude that a reliable test for PRA is indeed available. Further examination may, however, revoke any such assumption. Although the onset of PRA disease is associated with the link fragment, the reverse may not be true. That is, the link fragment may not be associated with all types of PRA (it isn't). There may have been several spontaneous mutations, each producing separate populations isolated geographically that maintain a PRA type gene differing to the gene of interest in the PRA link test. Dogs with these genes would most likely test negative for PRA using available link test methods but may still carry or develop the disease.

The frequency at which the ancestral gene remains in the general population is unknown and without this knowledge we cannot determine with certainty whether the PRA link test is detecting the link fragment or is finding that similar DNA sequence located in the DNA of the ancestral gene. The link test itself requires that the ancestral gene has evolved with time and no longer contains a similar DNA sequence to that of the link fragment.

This we can be sure has happened in the majority of dog populations world wide. Other populations may have retained the ancestral gene and with the advent of international breeding, the ancestral gene is likely to be dispersed rather thinly in every country. The presence of the ancestral gene or any DNA fragment that resembles the link fragment will give rise to false positives. Mean-ing a dog may test positive for PRA but not have the gene at all and therefore not develop the disease nor pass the gene to progeny.

We have seen that a dog may test negative and carry the PRA gene. Conversely a dog may test positive yet the PRA gene is absent. Perhaps this reflects only a small margin of error and the vast majority of test results are in fact correct. That would be reassuring but is difficult to prove, as decades need to pass in order to match test results with disease development.

Some statistics are readily available. For instance, it has been recorded that as much as 75 percent of all dogs tested by the PRA link test are affected and of these, 25 percent will develop the disease and become blind. In contrast, the incidence of observed PRA varies between countries from around seven percent to about 12 percent. This discrepancy may easily be explained as kennels with a history of PRA development probably utilise the testing service more frequently than other kennels.

Perhaps the social grapevine is still the best way to ascertain the accuracy of PRA link tests. Stories of the dog that was given an all clear for PRA and later developed PRA have completed many rounds of the rumour circuit. DNA link testing does work most of the time and this should be carefully considered. The actual percentage of accurate results at present cannot be determined, but time will reveal these details soon enough. At that time we will all know if the money was well spent and that the tests are reliable.

The ambiguity of the PRA link test casts a dark shadow over other link tests developed to detect some of the many heritable diseases that afflict dogs. We must be careful not to generalise, thereby associating other tests with the inconclusive nature of the PRA link test. Each link test needs to be assessed according to its own merit. The implications for the breeder are clear. Make an informed decision whether to use these tests. Decide how much emphasis to place on the result. After all, a positive could actually be a real positive, not a false positive. Any breeder ignoring a positive result will undoubtedly entertain doubts about the animal from time to time and this may influence a kennel breeding strategy or bias the breeder against the animal. But to cull a dog based on a test result alone would not only be cruel, but a detriment to the breed.

Experience, skill, and knowledge of specific breed genetics are as important now as they have always been. - K Scott