Convergent toxins

January 17, 2010 • 5:01 am

by Matthew Cobb

Convergent adaptations form one of the most striking classes of proof for evolution by natural selection. Radically different species, with common ancestors deep in the past, show near-identical adaptations to similar environments. Convergence can even be seen in species that are separated by vast depths of time, such as icthyosaurs (extinct marine reptiles) and dolphins, which show strikingly similarities in their bodily form, a consequence of adaptation to the environment (water) and their predatory role, based on a common tetrapod anatomy and musculature.

However, most examples of convergent adaptation remain at the level of form or function, rather than the genes involved. It’s possible that the same genes are involved in shaping a dolphin and an icthyosaur, but it’s unlikely we’ll ever know. An exception is the recent discovery of convergent mutations in species of sand lizards with white skins, all of which affect the melacortonin-1 receptor. But in a way, that isn’t too surprising – the melacortonin-1 receptor controls skin colour in these animals, and only a restricted number of molecular changes would give rise to an advantageous white form.

Even more striking is the announcement, shortly to appear in the pages of Current Biology, of identical skin toxins produced in two lineages of frogs, but on the basis of two different genes that diverged during the Cambrian, between 488 and 557 MY ago!

Both lineages of frogs – the Australian Litoria species and the Pipidae (which includes the model species Xenopus laevis) – secrete caerulein, a powerful toxin that induces vomiting, diarrhea and pancreatitis, amongst other things. Strikingly, the two sets of amphibians have very different ecologies and are separated by massive distances – the Australian/Papuan Litoria frogs tend to be terrestrial, whereas Xenopus and its African relatives are strictly aquatic. This means that their toxins are used to ward off very different predators, although the assumption is that in all cases the predators are vertebrates, which are all vulnerable to these toxins.

Researchers in Belgium and Australia, led by Kim Roelants, studied the genome of these frogs, and also looked at the proteins they produced, and discovered that although the caeruleins produced by the two sets of species are identical, they use very different genes to get there.

In the case of Xenopus and its relatives, the caerulein gene evolved through duplication of the cholecystokinin gene (cck); in the case of Litoria, the gene involved was gastrin. Both these genes are present in all vertebrates, but diverged during the Cambrian, and subsequently evolved into a series of genes with different functions in different lineages.

In both cases, gene duplication – when a gene gets mistakenly copied through a genetic accident – provided the raw material on which natural selection could work. With two versions of a gene, one is able to maintain the important function that originally led to its presence, will either evolve randomly (and eventually become a pseudogene, no longer functional) or will accidentally produce a product that is in some other way advantageous to the organism that carries it.

In the case of the cck gene, there were several such duplication events, including two within the lineage that led to Xenopus – there are now three cck genes in Xenopus.

Phylogeny and comparative genomics of the cck and gastrin genes in vertebrates. Black lines at bottom indicate periods during which gene duplication (D) took place. Orange lines = evolution of skin expression of the genes. You can ignore the numbers on the tree.

The authors explain this striking case of molecular convergence by the fact that the products of both cck and gastrin genes retained a common structure; this meant they could both function in the frog’s physiology and produce a compound that would repel a wide range of vertebrate predator (by affecting the same physiological processes).

In other words, natural selection was able to use gene duplication to maintain one function, and select another, focusing on the same tiny character, but in two very different genes.

[First posted at the z-letter]

16 thoughts on “Convergent toxins

  1. “You can ignore the numbers on the tree.”

    What, lay people don’t get to enjoy the bootstrap debate?

    1. OK, you asked for it! The original caption reads:
      “Branch labels represent bootstrap percentages obtained by maximum parsimony (left) and maximum likelihood (right); values < 50% are indicated by a slash (/)"

      If that isn't clear, then shake your neurons with these Wikipedia pages (I cannot vouch for the accuracy of them, but I presume they are more or less OK.) Good luck!
      Bootstrapping:
      http://en.wikipedia.org/wiki/Bootstrapping_(statistics)
      Maximum parsimony:
      http://en.wikipedia.org/wiki/Maximum_parsimony
      Maximum likelihood:
      http://en.wikipedia.org/wiki/Maximum_likelihood

  2. Remarkable!

    the products of both cck and gastrin genes retained a common structure; this meant they could both function in the frog’s physiology and produce a compound that would repel a wide range of vertebrate predator (by affecting the same physiological processes).

    So not as accidentally produced product as in other cases. Actually, as many (most, I guess) genes and certainly these ones are related by descent, I would push the contingency away from this part of the process to the one where the organism benefits from a new trait.

    [Not fully though, as this layman understands it. Just because there are a number of pathways doesn’t mean any will be taken.]

    And have I said how remarkable all of this is?

  3. But see, this just proves design, because how could anything that unlikely occur on entirely different genes?

    Or, it’s just “microevolution.”

    My point being, they have just decided that evolution could never happen, beyond an undetectable “permissible” extent, so they’re just going to deny the practical possibility of any evolution beyond that.

    The mere evidence of what evolution predicts is totally beside the point. ID exists in order to deny the possibility for what non-teleological evolution predicts will be observed (and is) to actually be able to occur (unless it’s deemed to be “microevolution”) except by magic. IOW, it exists in order to deny the meaning of the evidence as it would be ordinarily construed by science, the judiciary, and by most reasonable laypersons.

    So these data only make sense via non-teleological evolution? It doesn’t if the first premise of the ID that you accept is that non-teleological evolution cannot happen and that only intelligence can create life. Then it’s just more evidence of magic making life.

    The only conclusion ID ever allows is that God had to do it.

    Glen Davidson
    http://tinyurl.com/mxaa3p

  4. Very cool! Has someone nailed down a definition of “convergent evolution” that incorporates all these molecular differences?

    Without giving it much though, I wouldn’t quite call the lizard example “convergent evolution” in the usual sense of African euphorbes vs. American cacti – or all the marsupial vs. mammal examples. These recent examples feel different – one is just tweaking the same pathway to arise at the same end result in one case (not convergent evolution?), whereas the other seems to work… the same product resulting from two very different starting points.

    It seems like these could almost lie on some sort of continuum of definitions of “convergent evolution” tailored to how similar the end results are to the two organisms most
    recent common ancestor?

    1. Affecting existing/shared gene products via two different SNPs;

    2. Affecting existing/shared gene products via changes in 2 different steps in a pathway [lizard Mc1 receptor];

    3. Producing the same gene product via non-shared (or completely different) pathways [frog toxins];

    4. Maybe something like repeated instances of such molecular changes all affecting the same physical characteristic?

    Thoughts??

  5. But see, the odds of the same thing happening in different genes means that it couldn’t have been due simply to mutations. Or it’s microevolution.

    The whole point of ID is to deny that non-teleological evolution is possible and that intelligence is required. It is the premise upon which ID is predicated, hence no other conclusion is allowed.

    Which means that evidence they’ll accept for “microevolution” simply won’t be accepted for “macroevolution.” That’s the secondary goal of ID, to deny any and all evidence that actually points to known unintelligent processes having produced life, to say that it is impossible that anything but intelligence could possibly produce the evidence of a complete lack of foresight during evolution. Foresight was still necessary, despite any straightforward evidence for it.

    So people not in denial about the evidence can be convinced–but most already are. ID’s public goal is to prevent the use of normal scientific inference in the case of evolution, by saying that the predictions of “Darwinism” would actually require god to effect their appearance.

    Glen Davidson
    http://tinyurl.com/mxaa3p

  6. But surely that proves God, because he designed them to fit the place he put them. They just happen to need the same design.

    It’s like eyes. You have eyes. Spiders have eyes. You are not an insect, so could not have evolved from one. Therefore God must have given you both eyes.

  7. that’s pretty cool.

    That list of symptoms is interesting because gastrin and CCK function as hormones in the mammalian (vertebrate?) digestive system (gastrin from the stomach to upregulate the stomach, and CCK from the duodenum to crank up the pancreas and to squeeze the gall bladder*); they also play poorly undrstood roles as neurotransmitters or neuromodulators in the brain.

    *”squeezes the gallbladder” or “mover of the bag of bile” are pretty good translations of “cholecystokinin.”

    1. Oh, and also, CCK downregulates secretion and motility of the stomach, making the hormones partial antagonists.

  8. At one time, convergent evolution required similar niches. I guess that restriction is gone now.
    Anything can be called “convergent”.

    1. Here are the two very different niches:

      “Strikingly, the two sets of amphibians have very different ecologies and are separated by massive distances – the Australian/Papuan Litoria frogs tend to be terrestrial, whereas Xenopus and its African relatives are strictly aquatic.”

      “Convergent evolution”:
      http://www.sciencedaily.com/articles/c/convergent_evolution.htm
      “In evolutionary biology, convergent evolution is the process whereby organisms not closely related (not monophyletic), independently evolve similar traits as a result of having to adapt to similar environments or ecological niches.”.

Leave a Reply