About X-frogs and failing at regeneration

Not the usual mad squee, but here’s a neat little system for studying regeneration that I quite liked today. I normally think about regeneration in terms of amputated limbs, mutilated hearts, decapitated flatworms. But you can induce a kind of “regeneration” in a less drastic and rather more tricksy way, at least in some animals. In newts and salamanders, you can create a small, superficial wound on the side of a limb, then manipulate a nearby nerve into it and add some skin from the other side of the limb.

The poor hurt limb then decides you’ve actually cut something off and tells the wound to grow a new limb. If you don’t add skin, regeneration begins but doesn’t progress very far; if you don’t add a nerve, nothing happens. IIRC you can also make extra heads in some worms in a similar way, but I digress. The figure below from Endo et al. (2004) illustrates just how well the procedure can work. The top row shows stages in the development of the extra limb, while D shows the stained skeletons of the original and new limbs. I’d say that’s a pretty good looking forearm and hand!



That this trick works is in itself a very interesting insight into the nature of regeneration, as it helps us figure out exactly what it is that triggers various steps of regeneration as opposed to a simple healing process (Endo et al., 2004).

Clawed frogs (Xenopus) have been staples of embryology for a long time, but they are also quite fascinating from a regeneration point of view. One, they can regrow their limbs while they are tadpoles, but mostly lose the ability as they mature. They also have a really weird thing going on with their tadpole tails, which they can regenerate early on, then can’t, then can again (Slack et al., 2004). Huh? O.o

Two, their adult limb regeneration ability is not totally absent: it’s somewhere between salamanders’ (oh, whatever, fine, I can do that!) and ours (uh… as long as I’m a baby and it’s just a fingertip?). In a frog, an amputated arm or leg doesn’t simply heal over, but the… thing that grows out of the stump is just a simple cartilaginous spike with no joints or muscles. It’s as if the system was trying very hard to remember how to form a limb but kind of got distracted.

We are obviously interested in creating superhumans with mad regeneration skillz, which also makes us interested in how and why animals lose this seemingly very useful ability*. (Bely (2010) wrote a lovely piece on this not at all simple question.) So: Xenopus yay!

Now, Mitogawa et al. (2014) have devised a skin wound + nerve deviation system to grow little extra limb buds in adult frogs. As you might expect, it doesn’t work nearly as well as it does in axolotls: you need three nerves rather than one, and it only induces a bud about half the time, but it works well enough for research purposes.

The bud (technically, a blastema when you’re talking about regeneration) looks like a good regeneration blastema: it’s got the seemingly undifferentiated cells inside, it’s got the thickened epidermis at the tip that teams up with the nerves to give developmental instructions to the rest of the thing, and it expresses a whole bunch of genes that are turned on in normal limb blastemas.

(Totally random aside: thanks to Chrome’s spell checker, I have discovered that “blastema” is an anagram for “lambaste”.)

One area where this blastema-by-trickery fails is making cartilage, which is one of the few proper limb things the defective spike regenerates in frogs do contain. There’s no simple way of coaxing a complete spike out of these blastemas. The researchers tried the skin graft thing from axolotls (which can already form cartilage without the skin graft), but they still only got a little blastema with no cartilage. To get a skeleton, however crappy,  you need to cut out muscles and crack the underlying bone, which kind of defeats the purpose of the whole exercise IMO. At that point, you might as well just chop off the arm.

Below: the best a frog can do. Development of blastema-like bumps and “spike limbs” on the upper arm from Mitogawa et al. (2014). Compared to the fully formed accessory limbs of axolotls, the things you can see in B-D here are not terribly impressive, but they may be just the “transitional form” we need!

The failure of skin grafts alone at inducing cartilage, however, does hint at the things that go wrong with regeneration in frogs. Mitogawa et al. speculate that newt and axolotl limbs produce factors that frogs can only get from damaged bone. Broken bones even in us form a cartilaginous callus as they begin to heal, and unlike the cartilage in the extra limbs of axolotls, the cartilage in frog spikes is directly attached to the underlying bone.

They also point out that if you add proteins called BMPs to amputated mouse arms, which are otherwise even shitter at regeneration than frog arms, a surprising amount of bone formation occurs. (“BMP” stands for bone morphogenetic protein, which is a big clue to their function.) So it looks like there may be a kind of regeneration gradient from mammals (need bone damage AND extra BMP), through frogs (need bone damage, take care of BMPs themselves) to salamanders (don’t need either).

I should point out that salamanders and frogs are equally closely related to us, so this isn’t a proper evolutionary gradient, but given all the ways in which we and amphibians are fundamentally similar, our loss of regenerative ability may well have evolved through a similar stage to where frogs are now. Neat!

(I just wish they stopped calling us “higher vertebrates”. That phrase annoys me right up the fucking wall, because, and I may have said this before, EVOLUTION IS NOT A GODDAMNED LADDER. The group they are referring to has a perfectly good name that doesn’t imply ladder thinking. Amniotes, people. Or mammals, if you mean mammals, but I think if they’d meant mammals they would have said mammals. End grump.)


*I mean “us” in a very general sense. I think regenerative medicine is the coolest thing in medicine since vaccines and antibiotics, but I personally don’t think that the evolution of regenerative ability needs medical considerations to make it interesting. Whatever. I’m not exactly a practically minded person 😛



Bely AE (2010) Evolutionary loss of animal regeneration: pattern and process. Integrative and Comparative Biology 50:515-527

Endo T et al. (2004) A stepwise model system for limb regeneration. Development 270:135-145

Mitogawa K et al. (2014) Ectopic blastema induction by nerve deviation and skin wounding: a new regeneration model in Xenopus laevis. Regeneration 2:11

Slack JMW et al. (2004) Cellular and molecular mechanisms of regeneration in Xenopus. Philosophical Transactions of the Royal Society B 359:745-751

Thumbs down, what?

Bird fingers confuse me, but the explanations confuse me more, it seems.

I didn’t mean to post today, but I’ve just read a new review/hypothesis paper about the identities of the stunted little things that pass for fingers in the wings of modern birds. The review part is fine, but I’m not sure I get the difference between the hypothesis Čapek et al. (2013) are proposing and the hypothesis they are trying to replace/improve.

To recap: the basic problem with bird fingers is that fossil, genetic and developmental evidence seem to say different things about them.

1. Fossils: birds pretty clearly come from dinosaurs, and the early dinosaurs we have fossils of have five fingers on their hands with the last two being reduced. Somewhat closer to birds, you get four fingers with #4 vestigial. And the most bird-like theropods have only three fingers, which look most like digits 1, 2 and 3 of your ordinary archosaur. (Although Limusaurus messes with this scheme a bit.)

2. Embryology: in developing limb buds, digits start out as little condensations of tissue, which develop into bits of cartilage and then finger bones. Wing buds develop a short-lived condensation in front of the first digit that actually forms, and another one behind the last “surviving” digit. Taking this at face value, then, the fingers are equivalent to digits 2, 3 and 4.

3. Genetics: In five-fingered limbs, each digit has a characteristic identity in terms of the genes expressed during its formation. The first finger of birds is most like an ordinary thumb, both when you focus on individual genes like members of the HoxD cluster and when you take the entire transcriptome. However, the other two digits have ambiguous transcriptomic identities. That is, bird wings have digit 1 and two weirdos.

Add to this the fact that in other cases of digit loss, number one is normally the first to go and number four stubbornly sticks around to the end, and you can see the headache birds have caused.

So those are the basic facts. The “old” hypothesis that causes the first part of my confusion is called the frame shift hypothesis, which suggests that the ancestors of birds did indeed lose digit 1, as in the digit that came from condensation 1 – but the next three digits adopted the identities of 1-2-3 rather than 2-3-4. (This idea, IMO, can easily leave room for mixed identities – just make it a partial frame shift.)

Čapek et al.’s new one, which they call the thumbs down hypothesis, is supposedly different from this. This is how the paper states the difference:

The FSH postulates an evolutionary event in which a dissociation occurs between the developmental formation of repeated elements (digits) and their subsequent individualization.


According to the TDH no change of identity of a homeotic nature occurs, but only the phenotypic realization of the developmental process is altered due to redirected growth induced by altered tissue topology. Digit identity stays the same. Also the TDH assumes that the patterning of the limb bud, by which the digit primordia are laid down, and their developmental realization, are different developmental modules in the first place.

(Before this, they spent quite a lot of words explaining how the loss of the original thumb could trigger developmental changes that make digit 2 more thumb-like.)

I…. struggle to see the difference. If you’ve (1) moved a structure to a different position, (2) subjected it to the influence of different genes, (3) and turned its morphology into that of another structure, how exactly is that not a change in identity?

Maybe you could say that “an evolutionary event” dissociating digit formation and identity is different from formation and identity being kind of independent from the start, but I checked Wagner and Gauthier’s (1999) original frame shift paper, and I think what they propose is closer to the second idea than the first:

Building on Tabin’s (43) insight, we suggest causal independence between the morphogenetic processes that create successive condensations in the limb bud and the ensuing developmental individualization of those repeated elements as they become the functional fingers in the mature hand, thus permitting an opportunity for some degree of independent evolutionary change.

Am I missing something? I feel a little bit stupid now.



Čapek D et al. (2013) Thumbs down: a molecular-morphogenetic approach to avian digit homology. Journal of Experimental Zoology B, published online 29/10/2013, doi: 10.1002/jez.b.22545

Wagner GP and Gauthier JA (1999) 1,2,3 = 2,3,4: A solution to the problem of the homology of the digits in the avian hand. PNAS 96:5111-5116

What might have been possible

(Of fins, genes, fossils and the nature of evidence)


Behold the lengthy going-on about limbs, developmental genetics and semi-philosophical stuff that I promised! I mentioned that this was long in the making. Ironically, that means I’m not sure I managed to make it coherent… Then again, my blog subtitle does warn you about certain “meanderings” 😉


I previously mentioned that limbs kind of brought me to evo-devo. I haven’t closely followed the subject since, but a recent paper (Schneider et al., 2011) brought it back to my attention. Aside from the nostalgia, the evolution of limbs is also a perfect excuse for me to ruminate on some of the issues I consider important in evo-devo – such as the meaning of evidence, the role of “model organisms” and the nature of homology and novelty. (Some of this I touched on in my treehopper post)

I love developmental genetics. Davis et al. (2007), which through the blurred glasses of hindsight I’ll call the paper that made me an evo-devo nerd, is a genetic study. Genes are really exciting for us evolutionists because they obey different rules from the traits they control. Especially for regulatory genes – those that affect the activity of other genes –, gene sequence doesn’t correspond to the appearance of the organism in any straightforward way. The same circuitry of regulatory genes can also control the development of quite different structures, because most of the actual work is done by their target genes. Therefore, genes can often preserve connections we can no longer see in higher-level traits. (My favourite combination of evidence is genes plus fossils, but bear with me a little…)

The gist of Davis et al. (2007) is as follows. Hands and feet (collectively known as the autopod) are unique to tetrapods, or vertebrates with legs. There’s a special pattern of Hox gene activity that controls autopod formation. This pattern was missing from the fish that had been examined at the time. However, those fish are quite different from the distant ancestors they shared with tetrapods. There are living fish whose fin skeletons include bits that might correspond to digits, and there are many fossil examples. These include, as it later turned out, not just the iconic fishapod Tiktaalik, but also its slightly less tetrapod-like relative Panderichthys (Boisvert et al., 2008). Hence the question: did common lab animals like zebrafish lose the bones and the genetic circuitry, and did the bones of the autopod evolve from particular bones of the ancestral fin, or did tetrapods invent something new?

The answer is almost certainly the former, Davis et al. (2007) tell us after they find the tetrapod kind of Hox gene expression in the fins of a comparatively “primitive” ray-finned fish (Ray-fins are one of the two main groups of bony fishes. The zebrafish is a ray-fin, as are other familiar fish like cod and tuna. The other group – lobe-fins – include lungfish, coelacanths and tetrapods themselves). Around the same time, other teams found similar patterns in lungfish (Johanson et al., 2007), which are probably the closest living relatives of tetrapods, and sharks (Freitas et al., 2007), which are only distantly related to any of the creatures mentioned above.

Schneider et al. (2011), which caused this post, found that some DNA elements that regulate the Hox genes in the autopod are shared by tetrapods, ray-fins and sharks (ergo, probably all living vertebrates with fins or limbs). Together with the evidence from fossil and modern skeletons, this suggests that the digits of tetrapods evolved from pre-existing fin bones by tweaking an ancient genetic program. Fins and limbs really are variations on a single ancient theme. (Illustration of “fishapod” fins and early tetrapod limbs below is by Dennis C Murphy, from Devonian Times)

It is at this conclusion that we come to the stuff Hox gene expression can’t tell us. Knowing that radial bones (or cartilages, as the case may be) in fins and digits in limbs are “really the same thing” in some way is one thing. But radials and digits are not that similar, and neither is a shark’s fin and a newt’s leg. Maybe you’re interested in how one became the other, how fins suited to balancing and manoeuvring in water became limbs suited to plodding along on land. The autopod-like Hox pattern doesn’t say, since it’s basically the same in appendages that look very different, a perfect example of what I said about regulatory genes a few paragraphs back. Clearly, Hox genes define a distinct part of the fin or limb, but they don’t give detailed instructions on how to flesh it out. The details depend on the genes under Hox control.

Unseen ancestors

Now, fins and limbs are relatively easy, because we have a really quite awesome fossil record of their history (and also, some cool computer models :D). But the same lessons we can learn from their example apply equally to countless other cases where the fossil record is silent. Shared expression patterns of “master” genes and genetic pathways are often used to infer things about ancestors that aren’t known from fossils at all (De Robertis, 2008 is a nice review of such pathways). How far can we take such inferences? What does the fact that arthropods, vertebrates and segmented worms all seem to use some of the same genetic pathways to generate their bodies from repeating units (e.g. Stollewerk et al., 2003; Pueyo et al., 2008; Rivera and Weisblat, 2009)? Was their common ancestor as obviously segmented as an earthworm, did it just have a few repeated body parts like a chiton, or maybe nothing more than the basic ladder-like nervous system* of bilaterian animals? Or perhaps even less?

[*Photo of the nervous system of a planarian flatworm stained with a fluorescent dye, by the Agata group.]

The fossil record of early animal evolution (or rather, the lack of it) argues that this common ancestor was relatively small and simple (Erwin and Davidson, 2002). We know that quite different structures can be underpinned by the same “master” genes. Given this, can we really say anything meaningful about such long-extinct creatures? Well, we certainly can. They probably had the genetic circuitry their descendants share today. But what does that say about their body plans?

The answer may not be too far from “fuck all”. That’s why I chose a quote from Tabin et al. (1999) for the title of this post. I couldn’t agree more when they write, “developmental genetics only tells us what characters might have been possible”. I love finding out where we and the other creatures with whom we share this planet came from. That’s why I’m in this business. But there is only so much that any given type of evidence can tell us. And this is why I think the fossil record is so important. Like Erwin and Davidson (2002) argue, it can help us distinguish between “might have beens” in sometimes surprising ways.

Same difference

All of this puts the whole concept of homology into a slightly unsettling new perspective. Homologues (often spelled “homologs” nowadays) are supposed to be traits (genes, organs, behaviours etc.) that are derived from the same ancestral trait. The original concept of homology was defined for whole organs/body parts. Now, what do we do with organs that are made by the same genetic networks? Some of them show obvious historical continuity with the organs of other organisms. A bird’s wing is clearly homologous to my arm, on probably every level imaginable. They are connected by similar position on the body, similar basic structure, similar development and developmental genetics, and a rich fossil record. But that absolutely need not be the case.

Some butterflies use the same genetic circuitry to put eyespots on their wings that insects in general use to subdivide their wings into different regions (Keys et al., 1999). It would be quite absurd to call wings and eyespots homologous because of that – but in a very real sense, the gene network underpinning both is “the same thing”. And there is everything in between. Eyes, for example, share common “master” developmental genes including Pax6/eyeless. They were probably built around homologous cell types (i.e. photoreceptors) in most animals that have them (e.g. Arendt, 2003). Nonetheless, the highly complex eye structures of, say, a squid, a dragonfly and a falcon almost certainly evolved independently. And then there are the strange, confused identities of bird fingers that I talked about on previous occasions. Thus, when we ask the question: “are these two structures homologous?” – there is often no simple yes/no answer. At the very least, you have to ask: at what level?

An ode to diversity

The whole autopod business also highlights the dangers of extrapolation. Scientists believed that the autopod-specific Hox code was invented by tetrapods because their staple experimental fish didn’t have it. But life is a huge, diverse bush. Every twig has its own unique quirks, and we can’t take any of them to represent everything on its branch in every respect. In fact, some of the most popular lab animals – fruit flies, nematode worms, the aforementioned zebrafish – are also among the quirkier denizens of the planet. This is why I find it really really important not to limit ourselves to a few well-worn “model organisms”, not to draw sweeping conclusions from them. Although our common ancestry means that fruit flies or nematodes will in many ways help us understand ourselves, there is no guarantee. Comparative biology thrives on diversity.

(Of course, I say that as an evolutionary biologist working on a non-model organism. I may be somewhat biased ;))



Arendt D (2003) Evolution of eyes and photoreceptor cell types. The International Journal of Developmental Biology 47:563-571

Boisvert CA et al. (2008) The pectoral fin of Panderichthys and the origin of digits. Nature 456:636-638

Davis MC et al. (2007) An autopodial-like pattern of Hox expression in the fins of a basal actinopterygian fish. Nature 447:473-476

De Robertis EM (2008) Evo-devo: Variations on ancestral themes. Cell 132:185-195

Erwin DH & Davidson EH (2002) The last common bilaterian ancestor. Development 129:3021-3032

Freitas R et al. (2007) Biphasic Hoxd gene expression in shark paired fins reveals ancient origin of the distal limb domain. PLoS ONE 2:e754

Johanson Z et al. (2007) Fish fingers: digit homologues in sarcopterygian fish fins. Journal of Experimental Zoology Part B 308:757-768

Keys DN et al. (1999) Recruitment of a hedgehog regulatory circuit in butterfly eyespot evolution. Science 283:532-534

Pueyo JI et al. (2008) Ancestral Notch-mediated segmentation revealed in the cockroach Periplaneta americana. PNAS 105:16614-16619

Rivera AS & Weisblat DA (2009) And Lophotrochozoa makes three: Notch/Hes signaling in annelid segmentation. Development Genes and Evolution 219:37-43

Schneider I et al. (2011) Appendage expression driven by the Hoxd Global Control Region is an ancestral gnathostome feature. PNAS 108:12782-12786

Stollewerk A et al. (2003) Involvement of Notch and Delta genes in spider segmentation. Nature 423:863-865

Tabin CJ et al. (1999) Out on a limb: Parallels in vertebrate and invertebrate limb patterning and the origin of appendages. Integrative and Comparative Biology 39:650-663

The bare bones of fins and limbs

Perhaps the central question in developmental biology is how cells that start out as identical end up making bodies with complex shapes and a multitude of different tissues. And perhaps the central question in evo-devo is how such bodies can change into other bodies during the course of evolution. A really cool paper by Zhu et al. (2010) probes a little bit at both, and shows how relatively simple rules can produce results that are surprisingly similar to what we observe in nature.

The authors modelled the development of limb (or fin) bones in vertebrates. They used a simple model made up of the following:

  1. a virtual limb bud (let me call them “simbuds” hereafter) growing continuously
  2. a signal spreading from the tip of the bud that tells “cells” to keep growing but wanes over time (mimicking the role of the apical ectodermal ridge in real limb buds)
  3. two equations describing the activity of (1) genes that make cells differentiate into bone (“activators”), and (2) genes that prevent cells from doing so (“inhibitors”)

The shape of the simbud could be set at the start, and so could the values of all the parameters in the activator and inhibitor equations.

This is much more simple than real limb develompent. It says nothing about cell movement, and it condenses the effect of genes other than the bone activators and inhibitors into two little parameters in the equations. Yet running it with pretty much any initial settings produces something vaguely limb-like, and some sets of parameters give you simbuds that look eerily like real limbs.

Development of a simbud mimicking a chicken wing, next to drawings of the real thing.

Or fins. Or mutant limbs. Or transitional fossils.

Fully developed simbuds resembling various fossil fins: Brachypterygius was a marine reptile from the Jurassic; the other four are more or less close relatives of tetrapods, among them the famous "fishapod" Tiktaalik.

The similarity is not perfect, of course – but the model is not perfect either. Overall, it’s still pretty amazing what a variety of very realistic limb skeletons you can get out of such a simple setup – and how much you can achieve just by varying small things like how wide the limb bud is to begin with or how strongly two gene networks interact. Evolving fins into limbs should be a piece of cake for a system like that!



Zhu J, Zhang Y-T, Alber MS, Newman SA (2010) Bare Bones Pattern Formation: A Core Regulatory Network in Varying Geometries Reproduces Major Features of Vertebrate Limb Development and Evolution. PLoS ONE 5:e10892