Old Hands and New Fins

i-5d9c314d52cff4ad7884fc0b25807f5a-paddlefish agape.jpgThe paddlefish is a surreal giant, with a spatula-shaped nose that some scientists believe it uses to sense the electric fields of its prey, which it sucks up like a whale. You might not think of it as an animal that has much to offer in our quest to understand ourselves. But in fact, underneath the glaring differences, paddlefish and humans share some surprising similarities. And those similarities are precious clues to how our distant ancestors evolved hands and feet.

Back 400 million years ago, our ancestors swam with fins. The descendants of those early fish split off into two main branches. One became the ray-finned fishes--a group that makes up the vast majority of fish alive today. The other, sometimes called the lobefins, produced many species, but only three lineages of lobefins survive today. One is the coelacanths, which survive in a few isolated spots in the Indian Ocean. One is the lungfishes, which struggle against extinction in rivers and ponds in Australia, Africa and South America. The third lineage is by far the most successful: the fish with feet, or also known as the tetrapods. Tetrapods include amphibians, reptiles, and mammals. We belong to their ranks.

Starting in the late 1800s, paleontologists began to piece together clues as to how the fins of our ancestors evolved into our hands and feet. They found fossil lobefins from around 370 million years ago--that had a few stout bones in their fins that corresponded to the long bones of tetrapod limbs. (For more on this history, check out my first book, At the Water's Edge.) But some of the most important clues are quite fresh. Last year, Neil Shubin of the University of Chicago and his colleagues discovered a tetrapod-like fish in the Arctic called Tiktaalik that had particularly tetrapod-like limbs. Bones in its fins corresponded not only to the long bones of our limbs, but also to some of the bones in our wrists and ankles. Yet those bones were surrounded by fin rays that today you'd expect to see on a fish. It's a vestige of a lost time in evolution, when fins were becoming feet.

Shubin and his colleagues can also find evidence for how limbs evolved much closer to home--in their laboratories. A tetrapod embryo, be it chicken, mole, or human, produces four tiny buds along its sides, each of which then grows into a limb. In the late 1980s scientists began to pinpoint the genes that control the limb bud's development. The results of these experiments (carried out on chicks) were cause for a little deja vu. Scientists had already discovered that many of these genes play important roles in the development of other parts of the body.

For example, the head-to-tail axis of the vertebrate body is divvied up into different regions by a set of genes called Hox genes. Different combinations of proteins made from Hox genes can specify a section of the brain, a particular sort of backbone, or the point at which a tail branches off from the body. Thus it appeared that Hox genes, which in vertebrates had originally set the coordinates for the body axis, was borrowed at some point in the past to set the coordinates of the tetrapod limb.

The logical thing to do next was to see how fish fins are built. And in the mid-1990s, French scientists did so. The fish they chose was the zebrafish. By then zebrafish had become wonderfully easy for scientists to work on. They have many advantages over other fish--for one thing, their embryos are transparent, making it easy to see the stains that mark different kinds of proteins. It turned out that zebrafish, like us, use Hox genes to build fins. And the pattern of Hox genes they use turned out to be practically identical to the pattern found in tetrapod limb buds, at least in the earliest stages.

But as each kind of embryo develops further, the similarity ends. Zebrafish fins stop producing proteins from Hox proteins, and the skeletal bones in their fins never get very big. Insead, they become embedded in other tissues, such as rays, which form most of their fin. In the tetrapod embryo, on the other hand, cells at the forward edge of the limb bud keep making proteins from their Hox genes. But the pattern in which they do so is different from the earlier one shared by fish and tetrapods. And it's in this region, with its new Hox pattern, where hands and feet eventually form.

So here, it seemed, was the pivot on which the evolution of the hand had turned. By redeploying Hox genes in a still newer way, a limb-like fin had sprouted a hand (or, to be more precise about it, an autopod). There was just one problem. When scientists pick out a model organism, they sometimes put more value on how easy it is to study than where it fits on the evolutionary tree. A Drosophila fly is no ordinary insect, for example, but the end result of a lot of very peculiar evolution. And zebrafish are no ordinary fish. They belong to a very young, very successful branch of ray-finned fish, called the teleosts. That's not to say that scientists can't learn a great deal about biology from Drosophila or zebrafish, nor is that to say that those lessons can't apply to humans. But it's a fact that can't be ignored.

i-f718d1de1c5c4c3fd1ad3a15f2f4cff3-tol paddlefish 300.jpgThis fact has prompted several teams of scientists to turn their attention from the zebrafish to the paddlefish. The paddlefish belongs to a much older branch of the ray-finned fish. Along with paddlefish, this lineage is represented today only by a few species, such as sturgeon. Scientists have now figured out how to rear paddlefish embryos, and how to prepare them in order to see which genes turn on in the cells of the developing fin.

In today's issue of Nature, Shubin and fellow Chicago biologists Marcus Davis and Randall Dahn report a surprising result from their own study on paddlefish. In some ways, it turns out, the fins of paddlefish are a lot more like our hands than the fins of zebrafish.

When the fin of a paddlefish starts to form, Hox genes become active in the bud in the same pattern shared by zebrafish and tetrapods. But when zebrafish fins stop, the paddlefish fin keeps going. Its Hox genes produce the same late-stage pattern that defines our own hands and feet. The paddlefish doesn't go on to form hands and feet, of course. Instead, it forms a set of skeletal bones that's more developed than the skeletal bones in the fins of zebrafish and other teleosts. Around that core of skeletal bones, the paddlefish then grows other tissues to form a familiar fish fin, which it can use for swimming. The Chicago team conclude that when it comes to Hox patterns, it's zebrafish that are the new twist on limb development, not us tetrapods.

The scientists summarize their findings in an evolutionary tree, which I've reprinted below. The common ancestor of ray-finned fish and tetrapods, an early bony fish, had fins in which Hox genes continued to be expressed late in development. But it still grew full-fledged fins made of rays and other fin tissues. This late Hox pattern is still found today in the limb buds of tetrapods and paddlefish. In the ancestors of zebrafish and other teleosts, however, this late pattern disappeared.

Something else disappeared in our own lineage. As Tiktaalik and other fossils document, the outer fringe of fin tissues gradually shrank until it disappeared altogether. What was left were the skeletal bones, which grew and branched, evolving wrists, ankles, fingers and toes. Tetrapods became able to walk, first underwater and then on land. Later, these autopods evolved into bird wings, into whale flippers, into horse hoofs, into human hands. But they are still built with the help of gene patterns that existed dozens of millions of years before the first true hands and feet existed, patterns that can be found in distant relatives like the paddlefish.

Teleosts and tetrapods took opposite paths, the new research indicates, and along the way each found its own sort of success.

i-c4c6bf2c2c5ef417551198137f68e445-paddlefish limb evolution.jpg


More like this

Tiktaalik: music to my ears. Tiktaalik is the lilting name of a newly discovered fossil fish with fingers. It lived 380 million years ago in the northern reaches of Canada, back when the northern reaches of Canada were tropical coastal wetlands not far from the equator. Tiktaalik's discoverers (Ted…
Paleontologists have uncovered yet another specimen in the lineage leading to modern tetrapods, creating more gaps that will need to be filled. It's a Sisyphean job, working as an evolutionist. This creature is called Tiktaalik roseae, and it was discovered in a project that was specifically…
Shubin had a tough act to follow, coming after Kingsley's great talk. I'm sure it will be good, though — last night I got a tour of his lab, saw the original Tiktaalik specimens and some new ones, and some of his work in progress (which I won't tell you about until it's published), so I'm confident…
In addition to AIG, Reasons to Believe and all the other DI folks falling all over themselves to say nonsensical things about Tiktaalik roseae, Casey Luskin has now jumped into the fray with this silly post at the DI's blog. His argument can be summed up thusly: even though this find fills in a gap…

The always-surprising freakshow that is biological reality is so awesome, so exciting and interesting, only a dullard could even wish for ID to be true. Great post, Carl. Between you and P-Zeddy, I feel in on some fantastic discoveries.

By Greg Peterson (not verified) on 23 May 2007 #permalink

A better-than-average Loom append ... and that's high praise!

Thanks again ...

By Scott Belyea (not verified) on 23 May 2007 #permalink

...only a dullard could even wish for ID to be true....

The example of evolution of fins into hands can support theory of Nomogenesis or professor's John Davison Prescribed Evolutonary Hypothesis as well. It would mean that the mentioned evolution is predominantly caused by different activity of HOX genes which are so to say present from the beginning. There is no need for random mutation and natural selection - and consequntly for neodarwinism - at all. It could mean that information is stored from the beginning and in the right time is only released as a new phenotype.

Thanks for the interesting post. This type of concise prose style post is why I have now read At The Waters Edge, Evolution - The Triumph Of An Idea, and look forward to reading more of your books.

Carl i know that this dosen t have nothing to do with this topic but i can´t wait for your next book.. When it will be released?

By Jose Miguel (not verified) on 23 May 2007 #permalink

Terrific post, the kind of thing that makes me a keen reader of your blog and a number of others in the Celestial Assembly of Sciencebloggers. The reference to Hox genes links up nicely with a PZ piece in the last day or so. Three years ago I had never heard of Hox genes, now I think of them whenever I look at my hands and arms.

By John Monfries (not verified) on 23 May 2007 #permalink

>It could mean that information is stored from the beginning and in the right time is only released as a new phenotype

Stored where? The proteins that make these things happen are coded for by DNA. You can't take a paddlefish genome and flip a couple of switches and produce a chicken, as the necessary proteins depend on DNA that simply is not found in the paddlefish genome. There is plenty of evidence of differences between species depending on mutations in the original DNA. Differences between the DNA sequences of humans, chimpazees and macaque monkeys show many of these differences.

What this research implies to me is that in terms of limb development, zebrafish are more highly evolved than humans.

Consider this cuttlefish [cephalopod] comment [viewable in the video] from PBS NOVA [WGBH Boston] 'Kings of Camouflage':

"But the strangest of them all is a creature called the flamboyant cuttlefish. For a start, it's walking, not swimming."


Jose [comment 5]: It's looking like May 2008 for the E. coli book

VMartin, I am NOT a biologist, and I depend on the talented science writing of such folks as Carl, PZ Myers, and Sean Carroll for views on how evolution works, so I could be mistaken (and any mistake I make in what I say next is mine, not theirs), but my understanding is, when it comes to genetically encoded traits, the rule is "use it or lose it." Carroll explains this explicitly in "The Making of the Fittest," and provides cogent examples from nature. Information that is not being used in the present cannot be stored for the future. It will naturally undergo cumulative mutations and wind up corrupted and useless (as opposed to mutations that are advantageous in the present and selected for by natural or sexual selection, and can then be passed down through inheritance). For genetic traits to be "front-loaded" as you suggest, and then somehow preserved, would be a miracle, not science. Once you've posited that kind of miracle, why not just go the whole distance and hypothesize that the universe was created 6,000 years ago in seven days? The frontloading/storing approach appears superficially more sophisticated, but in terms of what is known from biology, it amounts to something very similar.

By Greg Peterson (not verified) on 24 May 2007 #permalink

("...our ancestors swan with fins...")

Greg--Agreed. The phenotype of the hand was not mysteriously stored in the Hox genes of the common ancestor of ray-finned fishes and tetrapods. The Hox genes evolved, as well as the genes they interact with, to give rise to hands, paddlefish fins, etc. But there is still enough conserved similarity between them to recognize their shared ancestry.

The Science Daily synopsis of this research characterizes it as a surprise to biologists - http://www.sciencedaily.com/releases/2007/05/070523132701.htm
I don't understand why you think this discovery is some sort of confirmation of evolutionary theory. All it confirms is the mounting evidence that the vast majority of genes are found highly conserved in a wide array of species. And rather than evolution producing more complexity, even your graphic shows gene "losses" along the supposed evolutionary tree. But the thing I find most interesting is how you would still think that genes have much of anything to do with the body plans of species. Certainly Hox genes are involved in development, but the same Hox genes are found in many species as this research shows. The reason that we don't look like paddlefish is because of the Transposons, microRNA, and other non-coding DNA that you've characterized as "genomic parasites" in previous posts. See the following: http://www.sciencedaily.com/releases/2007/04/070423185538.htm

Wade [13]: I'm afraid you're mistaken.

If you compare the Hox genes in a human, a zebrayfish, and a lamprey, you do not see an identical set of genes. In some cases the genes have been dupicated. (Actually, in the zebrafish lineage they may have duplicated twice, if I remember correctly.) In other cases, certain Hox genes have disappeared. And they have also acquired mutations that influence how they are controlled by genes and control other genes. So it's wrong to think of them as nothing more than a pre-existing set of genes. It's a family of genes that has been evolving for millions of years.

As for "losses," you're misreading the tree. It's the specific pattern of expression found in the ancestor that have been lost. The genes are still in the genome, playing other functions in the developing embryo. At the same time, however, new mutations arose in each lineage. But those were not the focus of this particular study.

As for non-coding DNA, there's overwhelming evidence that transposons etc. make new copies in a virus-like fashion, having started out in many cases as invading viruses. Perhaps a few percent of these genomic parasites have evolved functions that are beneficial to the host.


by which measure are tetrapods more successful than ray-finned fishes? number of species? number of habitats conquered? raw number of individuals alive? It's not a critique, I am just curious whether there really is a measure of success on which everyone agrees on.

Also: are lungfishes so badly doing, if they're present in three continents? Is their risk of extinction due to man / other predators, or are they only present in very marginal habitats?


Carroll explains this explicitly in "The Making of the Fittest," and provides cogent examples from nature. Information that is not being used in the present cannot be stored for the future. It will naturally undergo cumulative mutations and wind up corrupted and useless (as opposed to mutations that are advantageous in the present and selected for by natural or sexual selection, and can then be passed down through inheritance).

I don't know who Carroll is. Anyway Jaroslav Flegr (who translated carl Zimmer's Parasite Rex into Czech) wrote in his latest book "Frozen evolution" 2006 that there are many genes that are not expressed alone or that have cumulative effect so to say. It means that there are let say 10 genes and expression of one of them depends on the 9 remaining - so called gene interactions. These gene interactions are very complicated - there are many alleles of the genes involved so the outcome is always somehow specific. Many of alleles are not expressed and yet present for many ganerations. Then suddenly due specific combination of alleles of interacted genes they become "visible". So probably it is not correct idea that
genes/alleles become corrupted if they are not used. Otherwise eugenic should be possible and scientific.

Flegr came to the same conclusion as Broom, Huxley and Davison - who is banned here btw - that evolution is frozen (darwinian newspeak for finished). Of course Flegr give his brand new - as far as I can judge - theory of the phenomena. It should be due "frozen plasticity" of species. It is also crux of his probably originalthought that domesticated animals are "evolutionary" youngest,
just developed. The other, older species are so "frozen" that they cannot be domesticated whatever effort you make using selection!

An interesting idea. Flegr is heretic and yet somehow evolutionary scientists who dont need God.

For genetic traits to be "front-loaded" as you suggest, and then somehow preserved, would be a miracle, not science.

It is the same miracle as the fact, that homo sapiens would have evolved due "random mutation". Natural selection play no role in evolution according Punnet, Schindewolf, Heikertinger etc. - it just removes extremities. Flegr idea correspond with Davison idea that sexual selection probably hindern further development and stabilize species.

The frontloading/storing approach appears superficially more sophisticated, but in terms of what is known from biology, it amounts to something very similar.

Of course my opinion is that evolution was directed rocess towards origin of mankind. Once mankind arouse evolution is finished. Its creative forces are not taking ffect any more. Its like when you study processes in a cell of dying man and want from it to establish process and forces in development of zygote. No way. And yet
information is still there, from the beginning (zygote) till the end - the death of an organism. The same for phylogeny.

(I quoted Flegr because he translated Rex and because Charles Uni Prague seem to be source of darwinian heretics. The most prominent from them is professor Zdenek
Neubauer who publicized world wide - also in Nature - and then became an antidarwinist. He called neodarwinists to be sorcerers of modern era who kill beauty of life.

Neubauer: A brief consideration on the meaning of the lysogenic conversion. Nature (1967) 213:1263

Oppenheim, Neubauer, Calef, Antirepressor: A new element in genetical regulation. Nature (1970) 226:31


Thank you for a high quality post. This is truly fascinating stuff. You say that tetrapod embryos develop four buds. It will be interesting to know if ray-finned fish embryos develop the same number of buds. This might then suggest that all fish have the same number of fins. If less, are some modified (or suppressed)? If more, what is going on? Would this clash with the requirements of fluid dynamics, like fins for stabilization and balance?

By Vijay Sundaram (not verified) on 30 May 2007 #permalink


Excellent post. I am a lowly lay-person, non scientist, but I eat this stuff up. After reading "Fish With Hands and Whales With Feet," several months ago, I have been hooked on your books, and your blog. Thanks to that first book I read, I even remembered what HOX genes were. It made me feel very smart when I dropped this information around a Biology teacher, friend of mine.

It will be interesting to know if ray-finned fish embryos develop the same number of buds. (#17)
Fish seem to have three sets of paired fins: pectoral, pelvic and anal. The homology between pectoral fins and arms is clear, but is it certain that the pelvic fins and legs are comparable? What happened to the anal fins? Or is the number of paired fins variable among fishes? In Shubin's group other recent Nature article (18 January 2007), about Sonic hedgehog expression in Chondrichthyes, a skeleton was shown for the midline dorsal fin, that looked much as if it had the general build of a pectoral fin skeleton. Has anyone any idea when fins earliest appeared in fishes, and with what type of skeleton?

VMartin, my creationist cliche-o-meter exploded when I read your post.
"Once mankind arouse evolution is finished."
Gaaaaaah!!! NO NO NO NO NO!!!! That's NOT how evolution works at all!!! There IS no pinnacle of evolution-that's just more regurgitated creationist claptrap about an "evolutionary ladder", which is really a thinly vieled version of the Great Chain of Being. http://en.wikipedia.org/wiki/Great_chain_of_being

By Laser Potato (not verified) on 23 Aug 2007 #permalink

Comforting for me to know there is ONE other person in my County (the guy from Wickliffe, KY -- I'm in Kevil, KY) actually believes in SCIENCE & isn't a fundamentalist right wing "CREATIONIST"!

This is the Bible Belt, & my son has married into a family which is fundamentalist Christian. They took my 4 yr. old granddaughter to that "CREATION" Museum in Ohio?

I asked my daughter in law, didn't she have Science in High School? She said, "Yes, but I didn't believe a word of it!" Which led to, "You're not a "Christian" if you believe in "Evolution", the World was created in 6 days, the Bible says so, case closed -- and I'll be going to "HELL" when I die.