What Six-Toed Cats Tell Us About Genetic Development
Neil Shubin on What We Can Learn from Mutations
Sailors once believed that six-toed cats could bring good luck on ships. These so-called mitten cats were thought to make better mousers because their broad paws could balance them while at sea. Stanley Dexter, a sea captain, had a litter of these cats and gave one to his pal Ernest Hemingway, who was living in Key West at the time. This kitten, Snow White, gave rise to a lineage of six-toed cats that thrives to this day at the Hemingway estate. Besides being a highlight for tourists, these cats have played a role in a new conception of the workings of the genome.
People, too, occasionally have extra fingers and toes. About one in every thousand people is born with an extra digit in the hand or foot. In an extreme case, in 2010, a boy in India was born with thirty-four digits. Extra fingers can appear on the thumb side or the pinky side, or in split and forked fingers. Additional digits on the thumb side, known as preaxial polydactyly, are particularly important biologically.
In the 1960s scientists working on chicken eggs were probing how wings and legs are made in the embryo during development. Limbs emerge from the embryo’s body as tiny buds, looking like small tubes. Over a few days—the number varies by species—the bud grows, bones begin to form, and the growing end becomes shaped like a broad paddle. Digits, wrists, and ankle bones form inside this expanded surface.
Scientists discovered that by removing or moving the cells inside the paddle area, they could tweak the number of digits that form. If they excised a small strip of tissue from the terminal end, development of the limb stopped. If they cut out this strip during early development, the embryo formed a limb with few digits or none at all. If they extracted the strip at slightly later stages, the embryo might lack only a single digit. The stage of development at which the experiment is done matters: early removal has more dramatic effects on the embryo than later removal.
John Saunders and Mary Gassling from the University of Wisconsin, for reasons lost to time, extracted a tiny slice of tissue from the base of the growing paddle of a limb bud. This patch is nondescript—nothing about it looks unusual. It sits on the side of the paddle where the pinky will ultimately form. The researchers took this sliver of tissue, less than a millimeter long, and grafted it onto the opposite side of the limb bud, at the base of the paddle where the first digit would form. After sealing up the embryo in the egg, they let it complete development.
The embryo that emerged was a complete surprise. It looked like any normal chick, with a beak, feathers, and wings. But its wings, unlike normal wings with a pattern of three elongated fingers, had as many as six fingers. Something inside that little patch of cells contained instructions to make fingers.
Other labs soon got into the act. In the 1970s a group from England put tiny strips of tinfoil between the patch of tissue and the rest of the limb bud. The wings that emerged had fewer digits than normal. The foil served as a barrier between the patch and other cells. The implication is that some compound emanates from that patch of cells, diffuses across the developing limb, and stimulates digits to form. When a foil barrier stops that diffusion, fewer digits develop, and when the barrier is placed at a different point in the limb, more digits form. But what is the compound that is released?You can waste a lot of time looking for something nearby when you should be looking really far afield.
In the early 1990s three laboratories, working independently, used new techniques to isolate the protein and the gene that makes it. The gene makes a protein during limb development that diffuses across the paddle of the limb bud. As it does so, the researchers found, it tells groups of cells which digits to form. High levels of the protein make a pinky, or fifth digit. Low levels make a first digit, or thumb. Intermediate levels form the digits in between. One of these groups of researchers named the gene Sonic hedgehog, a nod both to a gene known as hedgehog at work in other species and to a video game popular at the time.
But what tells the gene to make fewer or more digits? Are there switches at work for the Sonic hedgehog gene that influence the evolution of digits? Answering this question would be a key for understanding how genes build bodies and how they evolve.
As with most important moments in life and science, this story begins with an accident.
In the late 1990s a team of geneticists in London were inserting snippets of DNA into the genomes of mice to study brain formation. These fragments are part of a little molecular machine researchers make to attach to DNA and to serve as a marker for its activity. Every now and then something goes wrong with this kind of experiment. The fragment can land anywhere in the genome. If it lands in a biologically important part of the genome, a mutant can form. That’s what happened with this team’s experiment: some of the injected mice developed normal brains but had deformed fingers and toes. In fact, one of the mice had extra digits and very broad paws not unlike Hemingway’s mitten cats. The team was able to generate an entire family line of these mutants and, by scientific convention, give them a name. They called them Sasquatch, after the big-footed creature of the paranormal world.
Since their mutants were now useless for the study of brains, the team wondered if any limb biologists might be interested in them. They set up a poster at a scientific meeting announcing their results. Posters at conferences are sometimes thought to contain the B-list of scientific results, as the best ones get presented as talks. But posters also have a social element; people mill around and science gets discussed. It’s been my experience that more collaborations begin over posters than after talks.
The poster showed a type of polydactyly that was known to arise from a mutation in Sonic hedgehog: the extra fingers were on the pinky side. These mutations happen because Sonic hedgehog is turned on in the wrong side of the limb. So the obvious next step was to look at the activity of Sonic in the mutants, experiments that the team did to present in their poster. After they accidentally made the mutant, they looked at the tiny developing limbs under the microscope. The activity of Sonic in the mutants was abnormally expanded, just as you would have expected in this kind of polydactyly. These observations led to the hypothesis that the mutant Sasquatch had been produced by the snippet inserting into, or very near, the Sonic hedgehog gene.
The team didn’t attract a limb biologist to their poster, but Robert Hill, a distinguished geneticist at Edinburgh, randomly walked by and saw the photos of the Sasquatch mutant. From that, a new research program began.
Hill’s lab had gained renown for understanding the workings of the genome in eye development. Through that work, his team, including the young scientist Laura Lettice, had developed a toolkit to probe the genome to find fragments of DNA. Since they knew the DNA sequence of the snippet, they had to chug through the whole genome looking for where it ended up residing. Lettice was just starting her career and still quite green, but she had the patience and the skill set necessary to pull it off.
The team used a simple trick to identify the general location of the mutation on the strand of DNA. They attached a dye to a small molecule that was complementary to the piece of DNA that made the mutant. The idea was that this sequence would home in on the mutation, attach to it, and voilà, the dye would light up at that location. Since the mutation was affecting the activity of Sonic hedgehog, it was likely to be found in one of two places: in the gene itself or in the control region immediately adjacent to it, like the control regions Jacob and Monod had discovered in bacteria.
The reaction did not affect the gene of Sonic hedgehog. That area was not lit up by the dye. Whatever was affecting Sonic hedgehog in the limb, and causing polydactyly, wasn’t a mutation of the gene or, correspondingly, a change in its protein. The team concluded, as Jacob and Monod had, that one of the adjacent control regions was affected. But when they looked, they saw that this area was completely normal. So if neither the gene nor the adjacent switch was affected, what was the cause of the mutation?
As anybody who has ever tried to recover a model rocket on a windy day knows, you can waste a lot of time looking for something nearby when you should be looking really far afield. Hill, Lettice, and the team started trudging through the entire genome until they saw the signal. The snippet inserted was almost a million bases away from the Sonic hedgehog gene. That’s an enormous amount of genetic real estate between the site of the mutation and the site of the Sonic gene. Thinking they must be wrong, they repeated the process and reanalyzed the results. But try as they might, the result stood. A small region one million bases from the gene somehow controlled the activity of Sonic hedgehog. It was like finding the switch for a light in a living room in Philadelphia on a wall in a garage in suburban Boston.
Maybe changes to this remote site were the source of the extra digits? The team tracked down every six-fingered person or cat they could find—polydactylous patients in Holland, a child in Japan, even Hemingway’s cats—and examined their DNA. And in every single one, they found a slight mutation in that region one million bases away from the Sonic hedgehog gene. Somehow, a little mutation at the far end of the genome causes a change in the activity of Sonic, turning it on broadly across the limb, leading to additional fingers and toes.
While sequencing the pattern of As, Ts, Cs, and Gs in this special region, they found this stretch of DNA to be very dis- tinctive. It is about fifteen hundred bases long, and its sequence is comparable among different creatures. People have the region in the exact same place as mice do, about one million bases away from the gene. So do frogs, lizards, and birds. It is present in everything with appendages, even in fish. Salmon have it, as do sharks. Every creature that has the Sonic hedgehog gene active in the development of its appendages, whether limbs or fins, has this control region almost one million bases away. Nature was telling scientists something important with this odd genomic arrangement.
At first glance, it is a wonder that polydactylous cats and people even survive to birth. Sonic hedgehog does not merely control limbs during embryonic development; it is a master gene controlling the development of the heart, spinal cord, brain, and genitals as well. Sonic is like a general tool that development pulls from its toolkit to make diverse organs and tissues. Accordingly, a mutation in the Sonic hedgehog gene should affect every structure where it is active; mutants would have deformed spinal cords, hearts, limbs, faces, and genitalia, among other organs. But what kind of animal would likely arise from a mutation in the Sonic hedgehog gene? Since so many aberrant tissues would likely be produced by a mutation in Sonic hedgehog, the answer would certainly be a dead one.A change in a protein can affect the body everywhere that protein is found.
But the way Sonic hedgehog is controlled during development ensures this outcome doesn’t happen. Why? Mutations in the limb-control region only affect limbs. That’s why polydactylous people with this kind of Sonic hedgehog mutation have normal hearts, spinal cords, and other structures: the switch that controls the activity of the gene is specific only to a particular tissue, leaving the rest unaffected.
Imagine a house with many rooms, each with its own thermostat. A change to the furnace will affect the temperature in every single room, but changing a single thermostat will affect only the room it controls. The same relationship is true for genes and their control regions. Just as a change in the furnace will affect the entire house, an alteration in a gene, and the protein that is produced, can affect the entire body. A global change would be catastrophic, producing dead ends in evolution. But since the genetic control regions are specific to tissues, like a thermostat in a room, a change in one organ won’t affect any others. Mutants can be viable, and evolution can work.
Two kinds of genomic changes can play a role in evolutionary transformations. In the first, changes in genes can cause new proteins to form. A mutation in the sequence of As, Ts, Gs, and Cs in DNA could bring about a change in the amino acid chain that makes protein. If the DNA mutation causes a different amino acid to form along that string, then a new protein can be produced. This clearly happens in many of the major proteins of the body, such as the hemoglobin genes that Zuckerkandl and Pauling studied. The key point is that a change in a protein can affect the body everywhere that protein is found.