Cat breeder desire consistency of conformation in their breeding stock. Their breeding programmes are aimed at fixing certain traits and removing others, but every now and again an unwelcome trait surfaces. This might be a visible trait such as curly hair or an inherited disease such as polycystic kidney disease. Breeders are then faced with a number of questions:

- is it a mutation that has been in the breeding stock for several generations and has shown up because carriers were mated together? If so, is it necessary of feasible to identify all carriers (or probable carriers) and remove them from the breeding program?

- is it the cumulative effect of several different mutations?

- is it a new mutation that has occurred spontaneously in this individual or in one or other of its parents? If so, is the mutation desirable or undesirable?

- has an unwanted genetic trait been accidentally fixed in the breed?

HOW MUTATIONS HAPPEN

Mutations are changes in an organism's DNA to produce a new version of a gene. The mutant gene may function in a different way and the effects range from minor through to devastating. The same gene may play a part in several processes and an apparently minor change, such as a different coat colour, may have other less visible impacts.

Mutations occur naturally due to transcription errors when cells divide (a bit like typos if you copy a piece of text), the mispairing of homologous chromosomes (sometimes they match wrongly so that loops or repeats occur) or damage caused by natural UV radiation. Paramutations occur when genes are silenced – the gene is still there and is unchanged, but it is not activated.

Mutations can occur anywhere in the DNA and in any cell. When they occur in somatic cells (body cells except the eggs and sperm) they may lead to cancer, signs of ageing or to cells that don't function correctly. Mutations in germ cells (eggs and sperm) will be passed on to the next generation and affect the offspring.

Some mutations are corrected by the cell's DNA replication machinery (enzymes) which limit the rate of mutation. Other mutations cannot be corrected and result in change. Although this article mainly looks at problem mutations, mutation plays an important part in evolution and adaptation.

Mutation must be regarded as unavoidable. The rate of mutation can be increased by UV radiation and toxins that can damage cells at the DNA level, but even eliminating these factors won't eliminate mutation altogether.

Gene mutation rates in sexually reproducing animals are normally less than 1 in 100,000 i.e. a particular gene may be mutated in 1 out of every 100,000 egg or sperm. Many of those eggs or sperm will never result in an embryo. Most mammals have around 80,000 to 100,000 genes so it's likely that each offspring born will have a new mutation in one of their genes. Again, there's a catch – there's a lot of "junk DNA" in the chromosomes so the mutation may occur in a stretch of unused DNA. Since offspring inherit 2 copies of most genes, there's a good chance the other copy will be an undamaged version (but their descendents may be less lucky as we'll find out). Fewer than 1 in 3 mutations will turn out to be lethal.

Whether a mutant genes is expressed or not depends on whether it is dominant (only one copy is needed for the trait to show up) or recessive (2 copies are needed for the trait to show up) or on the X or Y chromosomes (will show up unequally depending on whether the offspring is male or female).

GENETIC DRIFT

In sexually reproducing animals, each offspring inherits one copy of the gene from each of its parents (the exceptions are the genes on the X and Y chromosomes). It is incredibly unlikely that the same mutation will occur simultaneously in both parents so the offspring only inherits 1 copy of the mutant gene. It is heterozygous for that genetic trait.

If the gene is a dominant one, the animal only needs 1 copy for the trait to be exhibited. If the mutation is deleterious (harmful) it will quickly be eliminated from the population. In the wild, if that trait is somehow disadvantageous, the animal is unlikely to survive to breeding age and pass the gene on.

If the mutation is advantageous, the animal will be more successful in its environment and leave more descendents, about half of whom will have copies of that gene. The prevalence of the gene will increase in the wild, perhaps eventually outnumbering or replacing the unmutated version 25 or more generations down the line.

In pets and livestock, the mutation, whether advantageous or disadvantageous in the wild, might be considered aesthetically or economically pleasing and selective breeding may create an entire strain of animals with the same trait. Many modern breeds would have problems surviving in the wild due to their changed conformation, but are highly popular as pets.

Many mutations are neutral – neither good nor bad, but merely different – for example the wide variety of colours in random-bred cat populations. They don't currently confer a survival advantage, but at the same time they aren't disadvantageous, however, should the environment change, those genes might prove advantageous or disadvantageous.

Recessive mutations are not detected in the first mutant offspring because their effects are masked by the dominant version of the gene inherited from the other parent. Consequently, natural or artificial selection doesn't yet occur. When enough individuals carry a copy of the recessive mutation, those carriers become more likely to meet up and breed, resulting in homozygous offspring that exhibit the mutant trait. When (or if) 2 mutants are bred together, their offspring will always display the mutant trait. As with dominant mutations, if the recessive trait is advantageous these individuals are more likely to survive and breed. Longhair is a recessive gene in cats, but has proven advantageous in cold climates hence populations of longhairs arose in the USA (Maine Coon), Siberia and Norway.

It is possible to selectively breed in favour of a recessive trait, but impossible to eliminate carriers of the recessive gene unless there are DNA tests able to detect carriers who have only a single copy of that gene. There are a number of genetic defects in cats where screening tests are available.

The random nature of genetic events results in drift. If you breed a black cat to a grey cat, you'd expect the offspring to be black as this is the dominant trait. Since some black cats carry the recessive grey gene (are heterozygous), the offspring may be a mix of black cats and grey cats. You'd expect half the kittens to be black and half to grey, but there isn't always a 50/50 split (the average of a hundred litters from the same parents might give a 50/50 split). You might get 75% black and only 25% grey. If all of those offspring go on to breed, the frequency of black cats drift up in the next generation. But because many of those offspring carry the hidden grey gene, the frequency of grey may go up in a later generation.

In a large populations, these fluctuations won't be huge. Even if no grey cats are born for a couple of generations, the grey gene will never go away completely – it's still lurking there as a hidden recessive and sooner or later a throwback will appear when 2 heterozygous black cats breed. Only rarely do recessive mutations reach significant levels in a population and when this happens, it is usually due to some factor favouring those individuals e.g. the recessive mutation gives the animals an advantage (longhair in cold climates).

In small populations, the fluctuations tend to be much greater and eventually one version of the gene will take over and the trait will become fixed in that population. This is more likely to happen if the population is isolated and there is no new blood introduced. A trait can become fixed within 25 generation (around 100 years for cats). The Manx tailless mutation has a number of damaging side effects but because the mutant population was on an island it became a fixed trait. On the mainland (in Cornwall) it died out.

SELECTIVE BREEDING, GENETIC LOAD AND THE FOUNDER EFFECT

The previous section assumed that all of the animals were equally likely to breed and have offspring. This is not the case in domestic animals and livestock. Some males will be selected to fix traits in the breed and their offspring will greatly outnumber the offspring of other males (some of which may never be allowed to breed). For better or worse, the chosen stud's mutations (both the ones you can see and the ones you don't know about) will be passed on to more individuals. If he has a new mutation, this will quickly become widespread in the population. As a whole, the population will lose genetic diversity.

Each individual in the population probably carries a "genetic load" of 3 or 4 significantly harmful recessive genes. If an individual inherited 2 copies of a "lethal" or "deferred lethal" gene, they would be stillborn or die young. This is not a problem where those genes are distributed evenly around a large population because the chances of 2 carriers meeting up is small.

If a group of those individuals became isolated, the situation changes. An individual might carry a gene that is rare in the larger population e.g. 1 in 100,000. If that individual is isolated with only 9 others, that gene now occurs in 1 in 10. It gets passed on to descendents and becomes widespread in the small population, increasing the chance of 2 carriers meeting up and breeding. At the same time, variants of many genes will be lost because the founding stock of 10 animals didn't carry them. This effect is also seen where religious groups isolate themselves genetically by not marrying outside of their own faith. Any harmful genes carried by the founders become widespread within the faith in later generations. This is the founder effect and it greatly influences the genetic health of animal breeds and human communities alike.

Sometimes a mutation or hidden recessive turns up in an established breed and is considered attractive for example longhair in the Abyssinian gave us Somali. The longhaired offspring and their parents (who carry the mutant gene), are used to found a new breed. If any of those breed founders carry impairing genes, those genes are bred into the new breed. The founder effect may happen several times in the history of a breed. The first derivative breed may harbour another attractive mutation resulting in yet another derivative breed.

Breeds that have only small numbers risk losing valuable genes as the breeders work to fix "type". Some genes are closely linked (next door neighbours on a chromosome) and by selecting for one trait, the breeder inadvertently fixes a harmful mutation with it. Several breeds have breed-specific genetic problems, such as enzyme deficiencies or kidney disease, as a result. If the trait is recessive, a breeder might only see it once and assume it is a random new mutation. Only when the trait pops up again and again within the breed do breeders realise potentially damaging genes are present.

In the past, lack of exchange of information allowed some harmful recessive genes to become common in breeds. In the days of the internet, it is easier to discuss these mutations with other breeders or via a hereditary disorders list (such as the one maintained by the Feline Advisory Bureau) to see if the breed has a tendency to exhibit this trait.

SELECTION – NATURAL, ARTIFICIAL AND IN WHICH GENE’S FAVOUR?

Selection becomes possible and effective when a breeder can identify phenotypes (outward appearance) within the population. The increasing use of DNA screening to look for genetic markers is now also making it possible to select for or against certain genotypes. Already breeders can screen for carriers of a number of hereditary diseases in cats. X-rays have been used for some years for screening certain dog breeds for potential hip dysplasia problems.

Where genetic tests are not yet available, it is harder to use selection in dealing with undesirable mutations. Some mutations are silent i.e. have no noticeable effect on proteins they code for. Some mutations occur in non-coding DNA and don't make any functional products (however we don't yet know if those areas of DNA have other uses). A few mutations are beneficial. There is a problem if the mutations impairs a normal function, but not seriously enough for this to be noticed. These are easily overlooked and can become widespread.

Polygenes are another problem. Many traits, including some genetic diseases, are controlled by multiple interacting genes rather than one single mutation. In selecting for a particular look, breeders may accidentally be accumulating more of the polygenes in each generation. The cats become homozygous (have 2 identical copies) for a greater number of those genes until a critical point is reached and the cumulative effect of those genes can be seen. It's like a game of Jenga – you can remove quite a lot of blocks without the stack falling down, but it will reach a point where removing just one more block (or becoming homozygous for just one more polygene that affects a particular protein) will bring it all crashing down.

There are probably many genes that have mutations that cause minor impairments and cannot be identified. Many of these will already be fixed in a breed (because the gene was present in at least one of the foundation cats) without breeders realising there is a slightly defective gene. If that carrier is a popular stud, or to a lesser extent a prolific breeding female, that gene will become more wide spread in the next generation.

Lipinski et al (The ascent of cat breeds: genetic evaluations of breeds and worldwide random-bred populations, Genomics (2007)) showed a number of cat breeds have little genetic diversity due to intense selective breeding. This can cause inbreeding depression where fertility and lifespan are adversely affected. The long term survival of the Singapura and Sokoke (each derived from a very small foundation stock), may depend on careful outcrossing to improve genetic diversity without affecting the look of the breed. See The Pros and Cons of Inbreeding for a detailed discussion on inbreeding depression.

CONTROLLING UNWANTED MUTATIONS

There are several techniques for controlling unwanted mutations once they are identified. While many of the mutations can't be eliminated from a breed, it may be possible to reduce the further spread of the genes.

Linechasing: This is the old fashioned method and still useful in tracing potential carriers or working out where a gene came from. It involves studying the pedigrees of cats that have the undesirable or damaging mutation and identifying potential carriers: e.g. siblings and parents and preventing these from being bred in future. The siblings of the parents and any other offspring those cats have had are also potential carriers. The problem is that this is using a sledgehammer to crack a nut. Not all of the related cats will carry the trait.

DNA Screening: Some conditions can now be screened for and genetic markers for the undesirable gene identified. Linechasing can identify related cats and genetic screening can refine this by only eliminating proven carriers of the gene. In some breeds a gene is so widespread that all individuals should be screened before they are bred. Gangliosidosis GM1, a recessive gene found in Korat and Siamese cats is lethal when 2 copies of the gene are inherited, but the enzyme deficiency can also be detected in cats with only 1 copy of the gene so it is possible to avoid breeding from carriers.

Sometimes the undesirable gene can be traced to a single influential cat whose stud services were much in demand several generations earlier. Removing his descendents from a breeding programme might wipe out entire breed lines, some of whom would not have inherited the gene. The sledgehammer approach simply isn't feasible. Widespread genetic testing (if a test is available) isn't cost effective but a combination of linechasing to identify potential carriers and testing identified cats is more practical. They should be tested prior to breeding them, which in practice means testing prior to selling them for breeding.

CONCLUSION

Mutation cannot be prevented. Dominant mutations may be eliminated because these show up in the phenotype. Recessive mutations cannot be eliminated because these are carried from generation to generation without showing up until two carriers are bred together. Even when a recessive mutation surfaces, it may not be recognised as a gene that is already widespread in the gene pool until several cats are affected.

Breeders should make use of existing web-based lists of breed-specific genetic abnormalities, not just as a reader but also contributing information where a genetic abnormality is recognised or suspected in their breed-lines. In the past there has been a tendency to sweep things under the carpet, but openness is more beneficial to the cat fancy and the continued health of the cats themselves.