How fancy cats evolved: The science of our most adorable pets

The Manx, the Munchkin -- without fancy cats, we'd have no memes. Tip your hat to the genius of evolution

Published May 25, 2015 7:00PM (EDT)

    (<a href=''>chanon sawangmek</a> via <a href=''>Shutterstock</a>)
(chanon sawangmek via Shutterstock)

Excerpted from "Domesticated: Evolution in a Man-Made World"

A number of fancy cat breeds began with a mutation—often confined to one individual in a litter—that had an obvious effect on the phenotype. The Scottish Fold, for example, was founded by a barn cat from Perthshire, Scotland, with peculiarly forward-bending ears. Someone decided it would be a good thing to perpetuate this mutation. The Manx, from the Isle of Man, has a skeletal mutation that causes the tailless condition, among its other effects. In this it somewhat resembles the Japanese Bobtail, a natural breed with a quite different mutation. Munchkin cats have a mutation that causes limb shortening analogous to that of the dachshund.

Polydactyl cats have extra toes and constitute a recognized breed in the United States, called the American Polydactyl. They seem to have originated in southwest England, from where they made the Atlantic crossing by ship to New England, where they are especially abundant. One important reason for their early success was the widespread belief among sailors that they brought good luck—another example of the role of human caprice in the domestication process. The record for polydactyly is 27 toes, set by a Canadian cat. Here’s hoping that the record isn’t broken.

There is another mutation, called radial hypoplasia (RH), or “hamburger feet,” which results in a different form of polydactyly, of a spiraling nature. A creative breeder in Texas sought to build on this deformity in constructing a “Twisty cat” breed, in which the spiraling extends to the bones of the forelimb. Twisty cats also have extremely short forelimbs and relatively long hind limbs, which cause them to sit like a squirrel—hence an alternative name, “squitten.” Twisty cats are banned in Europe on humanitarian grounds, but not in the United States; the same is true of the Munchkin. It is time that the United States caught up with the United Kingdom in this regard. The deliberate breeding of skeletally deformed breeds is unconscionable.

Some of the oddest-looking breeds result from a mutation that causes hairlessness. Actually, these cats aren’t completely hairless; they just look that way. The first such breed originated in 1966 from a single naked kitten, appropriately named Prune. It is a mystery to me why anyone would want to perpetuate this condition; I suspect it is simple neophilia.

Given the climate there, it is particularly perverse that the Sphinx is a Canadian breed. But then, two other notable hairless breeds, the Donskoy and Levkoy, were created in Russia and Ukraine, respectively. One hopes they are indoor cats. Other cat breeds were founded by less drastic mutations of the coat, including the Cornish Rex (downy hair), Devon Rex (short guard hair), Iowa Rex (dreadlocks), and American wirehair (dense wiry coat).

The other method for generating new cat breeds is hybridization with existing breeds. The Siamese is most commonly used as one part of the cross. For example, the Havana Brown was the result of a cross between Siamese and American shorthair, and the Himalayan represents a cross of Siamese and Persian. Second-, third-, and fourth-order hybridizations begun with Siamese hybrids and other breeds include the Ragamuffin, Ocicat, and California Spangled. Some notable hybrids that lack a Siamese component include the Australian Mist (part Abyssinian), the Nebelung (part Russian Blue), and the Burmilla (part Burmese). The Levkoy is noteworthy not only for its uncomeliness but for the fact that it was created from a cross of two mutant breeds (the ear-challenged Scottish Fold and the hair-challenged Donskoy). The mutant ante can be ever upped.

Some truly creative breeders decided to go outside of the domestic cat box in finding partners for hybridization. The Chausie is a cross between an Abyssinian and a jungle cat (Felis chaus). Since the jungle cat is in the same genus (Felis) as the wildcat and the domestic cat, it is not surprising that this match worked. But other crosses outside of the genus Felis are more ambitious. The Bengal is a cross between a domestic cat and a leopard cat (Prionailurus bengalensis). At least the leopard cat is about the same size as a domestic cat; not so two other out-of-genus crosses: the Caracat is a cross between an Abyssinian and a caracal (Caracal caracal); and the Savannah is a cross between a domestic cat and a serval (Caracal serval). Both caracals and servals are considerably larger than wildcats.


When you start a breed with a single mutant, you have a founder population of two: the mutant and the individual with which it mates. To maintain the mutation at high levels, you must mate close relatives—say, siblings, or mothers and fathers with sons or daughters. Either way, the result is intense inbreeding and the accumulation of deleterious recessive mutations—a phenomenon known as “inbreeding depression.” Indeed, inbreeding in some cat breeds begun in this way is as severe as in dog breeds, as reflected in breed-characteristic pathologies.

The opposite occurs when breeds from different species are crossed, as in the Savannah and Caracat. Here the problem is a lack of harmony of various sorts among the genomes—a condition known as “outbreeding depression.” Servals and domestic cats, for example, don’t have the same number of chromosomes, which creates fundamental problems in partitioning them during the creation of sperm and eggs. More subtly, certain suites of genes that work particularly well with each other are normally inherited more or less as a unit. These “coadapted gene complexes” are broken up with excessive outbreeding.

The optimal condition lies somewhere between these poles, when the porridge is neither too cold nor too hot. The “just right” porridge is called hybrid vigor. This is what you get in mongrel dogs and barn cats. (Sylvester and Smoke are American shorthairs whose mother was the latter.) You would also expect to get hybrid vigor from crossing two distinct cat breeds, such as were used to create the Himalayan (Siamese ~ Persian). And initially, you do. The problem is that only a relative few offspring of these crosses, which have the desirable characteristics, are used as breeders for the next generation. The intense artificial selection for these characteristics soon results in inbreeding depression again.

The so-called natural breeds were in the “just right” hybrid vigor mode until cat fanciers began to control their breeding in the twentieth century. The effects of these efforts are especially evident in the Siamese, long the most popular of the natural breeds. The Siamese in Europe and North America today are strikingly different from those found in Thailand, as I can attest from personal experience. The Thai Siamese is a larger animal and longer of leg. Though the Thai Siamese has the typical “oriental” lithe body, it is more muscular, and not nearly as thin as that of the western Siamese. In addition, its skull is larger and notably more rounded in shape. These differences reflect the effects of artificial selection in the West.

The first Siamese to arrive in the West—appropriately named Siam—was an 1878 gift to President Rutherford B. Hayes. Six years later the first breeding pair was imported to Britain, followed by several more imports of a small number of these cats. Most Siamese in Britain today may be the descendants of only 11 imported Siamese. This small founder population, with its inherent sampling error relative to the genes of the Thai Siamese, was then prone, by virtue of its small size and isolation, to further random divergence through genetic drift.

The novel Siamese were an immediate hit at cat shows, so they were newly subjected to artificial selection, by means of which they further diverged from the original type. This divergent evolution accelerated in the last half of the twentieth century because judges came to prefer longer, thinner cats with proportionally small heads of a triangular shape, topped by large ears, set wide to emphasize this triangularity—to which end the snout was also thinned and the eyes became more almond shaped. Within a few decades, traditional Siamese had disappeared from cat shows. Some breeders organized to preserve the “traditional” style of Siamese, which is now recognized by TICA (The International Cat Association) as a new breed, called Thai. Such are the inversions of the topsy-turvy world of cat breeders.

The effects of inbreeding have been dire. Siamese have cancer rates rivaling those of Bernese mountain dogs and other cancer-prone dog breeds. They are especially prone to breast cancer. Accordingly, the life span of the Siamese is considerably shorter—with a median length of 10–12 years in one study—than that of the average house cat (15–20 years). Other “natural breeds,” such as the Abyssinian, also have shortened life expectancies as a result of inbreeding. Those that live longest are prone to blindness by means of progressive retinal atrophy and other defects of premature aging.

Aside from the Siamese, the Persian and the Himalayan have been the breeds most modified by sustained artificial selection. In addition to their gorgeous long hair, these two breeds are notable for their brachycephalic (squashed) faces, first developed in the Persian and inherited in the Himalayan when it was created through Siamese ~ Persian crosses.

Since creation of the Himalayan, the brachycephaly has been further exaggerated in both breeds, with predictable results. Though neither breed rises to the level of bulldog grotesquerie and its concomitant ailments, they do suffer from breathing problems, chronic sinus infections, and, more generally, abbreviated lives.

In stark contrast, the American shorthair, of which Smoke and  Sylvester are exemplars (OK, just Smoke), is a natural breed that has remained a natural breed. Which means that American shorthairs have long bred with whomever they deemed desirable—and the females often find it desirable to mate with more than one male. They evolved, from a large founding population, by means of natural selection into the perfect domestic cat—robust, athletic, and low maintenance. If properly socialized, they make ideal house cats. As an added bonus, American shorthairs are among the best mousers, right up there with the legendary Egyptian Mau.

There is an attempt under way to create an even better mouser, which would be the first cat breed created for function rather than appearance. The breed is called American Keuda, which is an acronym for “Kitten Evaluation Under Direct Assessment.” The breed is being created from American shorthair barn cats. The only criterion for the breeding program is exceptional mousing ability. Inbreeding, which inevitably reduces this ability, is therefore kept to a minimum, as evidenced by the huge variability in coat colors. Interestingly, some Keudas have come to look very much like the Egyptian Mau, a cat breed that perhaps most resembles the ancestral Felis silvestris lybica, from which all domestic cats are descended.


Cat genomics is not nearly as far advanced as dog genomics; it is still in the kitten stage. The first complete cat genome sequence came from an Abyssinian named Cinnamon. Subsequently, 10 other breeds have been partially sequenced. There are clear geographic factors in the genetic similarities of cat breeds. The Southeast Asian breeds, for example, form a distinct cluster; the European and North American breeds form a less distinct cluster; and the Central Asian, West Asian, and North African breeds tend to clump as well. Exceptions, such as the Ragdoll, American Curl, Ocicat, Sphinx, Devon Rex, Cornish Rex, and Bengal, are generally Western breeds recently created through hybridization or major mutations.

Many of the major mutations affecting body type and coat coloration of domestic cats were identified in the pregenomic age by conventional linkage analysis. Here I will consider a few interesting recent discoveries concerning coat characteristics.

Recall that a mutation (in a gene called Fgf5) was responsible for long hair length in many dog breeds. A mutation in the same gene also appears to cause long hair in cats. Actually, four separate mutations in this gene can cause long hair in cats, each different from the mutation that causes long hair in dogs. This phenomenon—same gene, different mutation, similar phenotype—is actually quite common. It occurs when different mutations, causing different amino acid substitutions in the coded protein, disrupt biological activity in similar ways. Since each variant of a gene is called an allele, we can more concisely say that, in this case, different alleles result in the same phenotype. But it is more often the case that different mutations in the same gene have different developmental effects; that is, different alleles result in different phenotypes. Consider the tyrosine gene (TYR), which plays an important role in coat pigmentation. One mutation in this gene is largely responsible for the distinctive coloration of the Siamese: dark extremities, light body. This color pattern is due to the fact that the mutant allele is temperature-sensitive.

During development, the extremities are cooler than the rest of the cat and the TYR gene is more active; in the more central areas, where the body is warmer, the TYR gene is less active, given this mutation. A different mutation in this gene results in an allele that is less temperature-sensitive. The result is the Burmese color pattern, in which the nonextremities are more pigmented than in the Siamese. Different mutations, and hence alleles of a related gene, called TYRP1, cause chocolate coloration or albinism.

Like all other genes, Fgf5, TYR, and TYRP1 are all coding regions of DNA, in that they code for proteins. But as we saw in the previous chapter, much of the evolutionary action is in noncoding sequences that regulate the activity of genes. One such noncoding mutation is responsible for the polydactyl condition. The gene that it regulates is one of the most storied in all of developmental biology: sonic hedgehog (shh).

Sonic hedgehog is a master developmental regulatory gene that produces a protein molecule of a sort called a “morphogen,” which forms a concentration gradient by diffusion. The effects of this morphogen on the cells of the developing embryo depend on its concentration. In this way, sonic hedgehog plays an important role in the development of organs, brain, and limbs. Its activity is regulated by noncoding elements near the gene called “cis-regulatory elements.” The limb-specific cis-regulatory element is called ZRS. A mutation in ZRS that causes too much sonic hedgehog activity is responsible for the polydactyl condition.

The noncoding polydactyly mutation is an example of a genetic mechanism that also underlies several human developmental abnormalities. And this is but one instance in which breeder-induced cat miseries have served to advance human medicine. For many of the ailments of purebred cats, as for purebred dogs, are also found in humans—a legacy of our shared mammalian evolutionary history. Indeed, these medical applications provided much of the original rationale for both the canine and feline genome projects.

Over 250 hereditary diseases of domestic cats are homologous to human diseases. The goal is to identify the genetic substrates for these diseases in cats, and then look for the homologous genetic substrates in humans. Cat models are especially promising with respect to progressive retinal degeneration, cardiomyopathy, and inherited motor neuron disease. The cat may also prove a useful model for amyotrophic lateral sclerosis (ALS). Cats are already important models for several viral diseases, including HIV-AIDS, which is prevalent in free-ranging cats, as is feline leukemia and the feline equivalent of SARS. Research is under way on cats to determine the DNA variants that make some cats more susceptible to these infections.

Until the wildcat genome is sequenced, feline genomics cannot provide much information about the genetic alterations that facilitated domestication. We can predict, however, that these genetic alterations were more concerned with behavior than with anatomy and physiology. For it is in their behavior that domestic cats most differ from their wild ancestors.

Excerpted from "Domesticated: Evolution in a Man-Made World" by Richard C. Francis. Published by W.W. Norton and Co. Copyright © 2015 by Richard C. Francis. Reprinted with permission of the publisher. All rights reserved.

By Richard C. Francis

Richard C. Francis is a science journalist with a PhD in neurobiology from Stony Brook University. He is the author of the acclaimed books Epigenetics and Why Won’t Men Ask for Directions? Francis currently resides in northern California.

MORE FROM Richard C. Francis

Related Topics ------------------------------------------

Animals Books Cats Evolution Noble Beasts