Leslie A. Lyons
Department of Population Health & Reproduction, School of Veterinary Medicine, University of California-Davis, Davis, CA
Overview of the Issue
Genetic testing is becoming more prevalent in many companion animal species and is an increasingly important diagnostic tool for veterinarians. At least seven1-5 genetic tests for diseases and phenotypic traits have become available for the cat within the past 12 months alone. The low-resolution genetic sequencing of the cat genome should greatly facilitate the development of additional tests at an even faster pace. DNA-based tests for several inborn errors of metabolism have been available for cats for many years6-10, however, DNA-based tests for more common diseases that afflict large breed populations, such as hypertrophic cardiomyopathy and polycystic kidney disease, have been recently announced and currently are, or may soon be, commercially available from more than one testing facility3,11. Genetic tests for several recessive coat color variations can be used by breeders to develop more efficient breeding programs. A DNA marker panel for cat parentage and identification has been internationally standardized and recognized, allowing the verification of pedigrees and individuals worldwide. Different laboratory techniques can be used to assay the same particular genetic mutation, thus, veterinarians and breeders need to recognize the inherent differences of these assays and the associated error rates. In addition to the important value of genetic tests to the breeder and veterinarian, these potential error rates for genetic testing will be discussed. The newly available tests for the domestic cat will be used as examples for these discussions. Breeders and veterinarians will learn the pros and cons of genetic testing and be reminded of the cooperation required between investigators, veterinarians and breeders to develop a test for the commercial market.
The reduction of genetic variation is an inherent concern for any domesticated breed or isolated population. Loss of variation tends to promote inbreeding depression within a breed, which may be expressed as health issues and defects. Generally, all breeds experience population bottlenecks, founder effects, selection, reduced migration, inbreeding, and random genetic loss. During the development of animals that will "breed true" and produce individuals with desired characteristics and phenotypes, deleterious or undesired traits can also increase in frequency within breed populations along with the desired, breed defining traits. These accidental traits can "hitch-hike" along with good traits within a breed via random chance, by being in close physical proximity on a chromosome to a desired characteristic, or by producing a desired trait in the carrier state. Thus, although breeders may have accidentally caused the increase in frequency of deleterious traits, with the advent of DNA-testing, breeders have a highly accurate means to determine trait status, even prior to trait onset, and hence they have the responsibility and a more effective means to properly manage deleterious traits within the population.
Along with the breeder, veterinarians have the important role of supporting proper breed management decisions. Veterinarians need to determine when a genetic test is appropriate, interpret the results for each genetic test, understand the inherent limitations and errors of a test, and provide additional health information to support informed breeding decisions. Each aspect must be considered, along with others more pertinent to the breeder, for a sustainable and healthy breeding program.
Phenotypes and Phenocopies
The phenotype is the outward appearance of an individual. What an individual looks like is a combination of environment and genetically inherited affects. Both desired and undesired traits can be mimicked by a wide variety of combinations of nature and nurture, thus, the initial challenge to breeders and veterinarians is to confirm that a phenotype is due to the expected interactions of genes and the environment. A trait that is clearly inherited as a single gene recessive trait, such as brown coat color in the domestic cat, can also be mimicked by poor nutrition. This mimicry is known as a phenocopy.
Commonly, a health issue can be accidentally lumped into the same category and be misreported within breeds. Polycystic kidney disease (PKD) in cats is associated with cystic kidneys and frequently a cystic liver. The disease often causes renal failure at an early age. Breeders can make the error of considering renal failure a differential for PKD instead of confirming that the renal failure is due to severe kidney cysts. Renal cysts alone are not a clear diagnosis for PKD, as the normal population can have minor and asymptomatic kidney or liver cysts. Thus, breeders and veterinarians need to be reminded that a proper list of differentials needs to be considered in order to differentiate between a true phenotype and a false phenocopy.
Hypertrophic cardiomyopathy (HCM) is a disease that often has false reports by breeders. Not all heart murmurs are due to HCM, however this is a common leap as HCM is reported at a high frequency in some breeds. Proper differentials are needed to decipher dilated cardiomyopathy (DCM) and other cardiac issues from HCM, and not all abnormal echocardiograms are a result of HCM. Hence a little knowledge can often lead to false conclusions in the breeding world.
Hallmarks for Genetic Diseases
DNA-tests have been developed for simple, single gene traits in the cat. Simple or single gene traits have a predictable inheritance because their mode of inheritance is clearly, dominant, recessive, or co-dominant. These traits may or may not be sex-linked. Inherited diseases can be deciphered from sporadic or idiopathic phenocopies by several hallmarks common to genetic traits. Firstly, genetic traits will tend to have a higher prevalence within a breed. Thus, particular breeds should be recognized as high risk for some diseases. Secondly, the trait will have a highly consistent presentation. The presentations may be multiple, but fairly consistent. Thirdly, if appropriate, the trait will have bilateral presentation. PKD is defined by having multiple cysts in both kidneys. Thirdly, the disease onset will be early for the type of disease. Hence, renal failure in a 4 year old Persian cat with bilateral kidney cysts is likely PKD. Lymphomas are common in cats, however, mediastinal lymphoma in the Oriental Shorthair cat breed generally afflicts individuals by 2 years of age. This early onset disease with in a breed that always presents as a mediastinal tumor is highly indicative of a heritable condition.
Along with phenocopies, genetic diseases must be categorized clearly due to disease heterogeneity. In humans there are several forms of kidney disease, which have polycystic kidneys. At least three different genes can cause polycystic kidney disease and lead to renal failure. In humans, the different genes have different modes of inheritance. Approximately 85% of human PKD is caused by a dominant mutation in the gene that makes polycystin-1, PKD1. However, PHKD1 causes recessive, juvenile onset of PKD and PKD2 (polycystin-2) causes a milder form of the autosomal dominant disease. The clinical presentation of these three forms of PKD is so overlapping that DNA-tests are often required to decipher which form of the disease is present.
In addition to 85% of PKD being caused by the gene PKD1 in humans, the disease is also caused by different mutations within the gene. Hence, most newly identified human families tend to have a new, sporadic mutation. Thus, genetic testing for PKD in humans is confined to within particular families or very limited populations and a variety of PKD alleles occur within the human population.
Genetic Test Development
Phenocopies and disease heterogeneity are common and recognized complications to the development of a genetic test, however, explaining these aberrations to a breeder often causes more distress as the information can be quickly misrepresented and falsely spread in the breeder community. Genetic tests are generally developed within a particular breed or population, and then slowly expanded to other breeds and populations once scientific standards have been rigorously performed.
The newly developed PKD test for cats is an excellent example for defining standards for genetic test development. The Persian breed is one of the original breeds of the cat fancy and is recognized in a wealth of colors and varieties. Being an older breed, this may predict that Persians have been bottlenecked and inbred for a longer period of time than other cat breeds, suggesting a more inbred population. However, the large amount of color variations and varieties and the large worldwide popularity of the breed suggest a large foundation stock, less genetic loss by random chance, and increased migration, as compared to other breeds. Each of these aspects of the breed all lead to increased genetic diversity. Thus, although Persians are a breed, it is difficult to predict if a disease may be due to a single mutation that has descended and spread through the population, or if Persians are more similar to a Caucasian human population, which is large and outbred, and hence they may have more than one cause of a common phenotype.
Such is the concern with a disease like PKD, where it is known to be caused by several different genes and even different alleles within the same gene for a large outbred population, like humans. Several large pedigrees of Persian cats were used to identify the mutation for feline PKD. However, in the world of genetic testing, finding the mutation is only the first step. Although the pedigrees used to find the mutation were large, they represented cats from only a few Persian lines. Hence, the next task was to confirm that Persian PKD was caused by the same mutation in all Persians. All Persians not only implies cats in the United States, but cats from throughout the world. Thus, clinical and research veterinarians and testing facilities from throughout the world had to cooperate to confirm that the Persian PKD mutation was consistent throughout the world and what is considered and mutation that is "identical by descent".
Immediately upon the announcement of the PKD test, different cat breeds "came out of the woodwork" and expressed concerns about PKD. Several breeds represent Persian varieties, such as Exotic Shorthairs and Himalayans. These breeds can be clearly included in the genetic testing. Some other breeds are predictable candidates for testing, such as breeds that have recently used Persians, like Scottish Folds and Selkirk Rexes. But, when more esoteric breeds, such as Maine Coons, British Shorthairs, and Burmese, also came forward, the investigators now had to become familiar with not only breeding practices for a given association within their home country, but nearly every breeding practice for every association for every country. Only rigorous testing standards can help decipher the breed relationships and support the transfer of a genetic test to different breeds. One breed, the British Shorthair, has been examined in detail for PKD and has helped develop the standards used by our laboratory. This study required the interaction and cooperation of breeders, researchers and veterinarians from the United States, Australia and Europe.
Genetic Testing Standards
To enable the PKD test to be used in other breeds, our laboratory has set rigorous standards that consider the possibilities of phenocopies, disease heterogeneity, sporadic mutations, "de novo" mutations and breeding practices. In order to transfer a genetic test to a different breed, the following criteria have been established and we attempt to follow these standards before releasing tests for other breeds:
1. Standard clinical signs and diagnoses must first be confirmed. In the case of PKD, proof of cystic kidneys by ultrasound has been the gold standard. For PKD, the cysts need to be multiple in one kidney or present in both kidneys. Once these diagnoses were confirmed, samples were considered as "voucher" specimens for DNA testing.
2. The DNA test had to match the clinical diagnoses.
3. Once an individual was confirmed as positive by clinical and genetic diagnoses, related individuals were collected to confirm the mode of inheritance. This step is taken to help prove if a mutation is a new, sporadic event, or an inherited mutation. New sporadic events pose less risk to the population than an inherited mutation. Proving the mode of inheritance also supports that the disease is caused by the same mutation and potentially inherited from Persians.
4. A variety of cats from different lines need to be examined to support the test for the breed. For British Shorthairs, we were cautious as to whether this mutation was due to outcrossing with Persians, or a "de novo" mutation within the breed. Hence, some cats could have PKD from Persian crosses, while others have a new mutation. Some breeders perform unacceptable outcrosses, thus this step also helps to determine breed risk.
5. Families from new breeds were proven by parentage testing. Known and unknown accidents happen, thus, all research pedigrees are confirmed by parentage testing.
6. Some percentage of the breed must be examined. Hence, again, is this a localized problem, or an inherent problem within the breed. Different breeding practices for different associations and countries need to be considered. Hence, PKD in British Shorthairs from Australia may not imply that all British Shorthairs in other countries are at risk if Persians have not been a common outcross. This percentage is a slippery slope and should ideally represent the effective breeding population. This aspect of the testing may be the most difficult to establish.
7. The breed should have a clearly defined outcrossing program with cats that have a high risk for the disease, such as Persians, Exotic Shorthairs and Himalayans. Parentage testing should be used to confirm this outcrossing.
A genetic test is based on a mutation that causes a trait or markers that are highly associated with a trait. Associated markers inherently have a higher error rate than mutation tests and this information should be published by the facilities offering association tests. Many laboratory techniques can be efficiently used to assay for mutations and each technique have its own potential errors. Thus, even though a mutation test should be as close to 100% accurate as possible, the errors in the type of assay will lower the accuracy.
A complete review of different laboratory methods used for genetic testing is beyond the scope of this discussion, however, a few cases can be presented. Recently, the mutation that causes the "pointing" phenotype in Siamese and other breeds of cats has been published. This mutation disrupts a restriction enzyme site and hence the restriction enzyme, MYL4, can be used to assay for the mutation. Restriction fragment length polymorphisms are very routine and a low cost and low technology method for genetic testing. However, a Dominant white Siberian cat was tested for "points" and clearly should have been a "pointed" cat, however, this cat had yellow eyes. The dominant white would have masked the pointing phenotype, but all "pointed" cats should have blue eyes. Sequence analysis of this cat revealed that the cat had only one allele for "points" and the other allele wildtype, however, a different nucleotide in the restriction site was altered. This DNA variation did not cause any amino acid change, hence does not cause a phenotypic affect in the cat and is just a random DNA variant. Thus, the genetic sequencing of this cat proved that the Dominant white Siberian cat only carries points, which the genetic assay suggested the cat was pointed. In this particular case, coat and eye colors helped to identify a false positive test. In the case of other traits, these other clues may not be available, hence, the veterinarian needs to understand the possible error of a genetic test.
Direct sequencing is generally the "gold standard" for confirm DNA mutations, but is a laborious and expensive endeavor, thus most genetic tests are not routinely evaluated by sequencing. But even with sequencing and any other PCR-based assay, DNA variants in the primer regions may cause "allele drop-out" and hence cause an individual to appear homozygous when they are actually heterozygous, but the second allele has not been amplified. Various higher through-put assays are now preferred over RFLP testing, but in the end, the veterinarian needs to recognize that a mutation test will still have an error rate that is more associated with the testing method and the efficiency of the laboratory.
Basically, veterinarians should always ask about error rates and never accept 100% accuracy as an answer. Breeders and veterinarians should not be bashful about asking labs to reveal their methods, provide the supportive data and reveal their efficiency. With cat DNA testing in its infancy, strong efforts should be made by researchers, veterinarians and breeders to promote accurate testing by efficient testing facilities.
1. Eizirik E, Yuhki N, Johnson WE, et al. Molecular genetics and evolution of melanism in the cat family. Curr Biol 2003;13:448-53.
2. Goree M, Catalfamo JL, Aber S, et al. Characterization of the mutations causing hemophilia B in 2 domestic cats. J Vet Intern Med 2005;19:200-4.
3. Lyons LA, Biller DS, Erdman CA, et al. Feline polycystic kidney disease mutation identified in PKD1. J Am Soc Nephrol 2004;15:2548-55.
4. Lyons LA, Foe IT, Rah HC, et al. Chocolate coated cats: TYRP1 mutations for brown color in domestic cats. Mamm Genome 2005;16:356-66
5. Lyons LA, Imes DL, Rah HC, et al. Tyrosinase mutations associated with Siamese and Burmese patterns in the domestic cat (Felis catus). Anim Genet 2005;36:119-26.
6. Crawley AC, Muntz FH, Haskins ME, et al. Prevalence of mucopolysaccharidosis type VI mutations in Siamese cats. J Vet Intern Med 2003;17:495-8.
7. Fyfe JC, Kurzhals RL, Lassaline ME, et al. Molecular basis of feline beta-glucuronidase deficiency: an animal model of mucopolysaccharidosis VII. Genomics 1999;58:121-8.
8. Haskins M, Jezyk P, Giger U. Diagnostic tests for mucopolysaccharidosis. J Am Vet Med Assoc 2005;226:1047; author reply 1047
9. Hubler M, Haskins ME, Arnold S, et al. Mucolipidosis type II in a domestic shorthair cat. J Small Anim Pract 1996;37:435-41.
10. Martin DR, Krum BK, Varadarajan GS, et al. An inversion of 25 base pairs causes feline GM2 gangliosidosis variant. Exp Neurol 2004;187:30-7.
11. Meurs K, Sanchez X, David R, et al. Identification of a missence mutation in the cardiac myosin binding protein C gene in a family of Maine Coon Cats with hypertrophic cardiomyopathy. American College of Internal Veterinary Medicine 2005.