More evidence for evolution: dead genes in sea snakes

Well, we don’t really need any more evidence for evolution, do we?  But it keeps pouring in, the latest in a paper from the Journal of Evolutionary Biology (link is to the abstract, and if you have journal access you’ll find the paper in the online “early edition”).

In WEIT I talk about olfactory receptor (“OR”) genes as evidence for evolution: these are genes that encode receptor proteins involved in smell, with each OR gene encoding a different protein.  These genes provide evidence for evolution because, as I describe on pp. 69-71 of my book, in terrestrial mammals most OR genes are intact and functional, but in their aquatic mammalian relatives, like dolphins, whales, and sea lions, many of the OR genes are inactivated by mutations (80% of them in dolphins!). (These inactive genes are called pseudogenes.)  Nevertheless, the DNA sequences of those dead genes clearly show their affinity to the active genes of terrestrial relatives.  OR genes are probably useless underwater  because they detect airborne and not waterborne odors.  Moreover, marine mammals that spend some time on land, like sea lions, have fewer OR pseudogenes than do more-aquatic species like whales and dolphins, presumably because sea lions need to smell things when they’re on land.

The new paper, by T. Kishida and T. Hikida from Kyoto University, offers a remarkable parallel to the mammal results, but from snakes.  To make a longish story short, the authors sequenced OR genes in two groups of sea snakes, those that give birth in the water, and hence never leave it, and those that are “oviparous”, and must leave the sea to lay eggs on land.  They also looked at OR genes from the snakes’ fully terrestrial relatives. (We know from other data that sea snakes evolved from land-dwelling ancestors.)

The results: just like in mammals.  In viviparous (live-bearing) sea snakes, about 30% of their OR genes had become nonfunctional pseudogenes, compared to only about 7% in their terrestrial relatives (elapids). That in itself is evidence for evolution, for what else could explain the presence of nonfunctional genes that are similar in DNA sequence — but broken by mutations — to active genes in relatives?  And the inactivation of OR genes in aquatic snakes makes real evolutionary sense; genes that are unnecessary — indeed, that are a metabolic burden since unneeded — get zapped by mutation.  The other cute result is that oviparous snakes that occasionally come ashore showed a lower proportion of OR pseudogenes genes (about 12%) than their viviparous cousins,  precisely as expected if they, like sea lions, must retain some sniffing ability on land.

I love pseudogenes, for they constitute some of the most irrefutable evidence for evolution–and not just microevolution, either. The transition from terrestrial to aquatic mammals, or terrestrial to aquatic snakes, surely represents macroevolution.  It’s hard to explain patterns like those of the OR genes by any form of creationism, unless you think the creator designed species to fool us into thinking they evolved!

Fig. 1.  An oviparous sea snake, the highly venomous banded sea krait, Laticauda colubrina.

h/t: Matthew Cobb


Kishida, T., and T. Hikida. 2010. Degeneration patterns of the olfactory receptor genes in sea snakes. J. Evol. Biol. early view

17 thoughts on “More evidence for evolution: dead genes in sea snakes

  1. But there’s little junk DNA in the genome, just like ID predicts. Cause it predicts, you know, whatever amount it turns out to be.

    And the dead genes are deterioration, plus there’s no reason why God can’t design useless DNA. What do you expect from God, good design?

    So see, genes can deteriorate, but when we find genes co-opted for other functions (you know, like non-teleological evolutionary theory predicts, which ID can’t predict) in the clotting cascade which couldn’t be hit as exact targets, we know that design was necessary since evolution can’t hit exact targets (never mind that evolutionary theory has no exact targets–the rubes don’t know that).

    So, uh, Jesus.

    Glen Davidson

  2. This past quarter I gave the Vitamin C pseudogene example to my beginning Cell Bio students (mostly nursing majors at the college where I teach). I actually heard jaws hit the floor in amazement that we and other primates have the gene that would make GLO if it were functional…

    The degenerate-DNA-because-of-the-fall-of-man argument is about the best the creationists can do, but it sure doesn’t explain why it happened to snakes. Oh, wait maybe that was part of the Lord’s punishment of the serpent for tempting Adam and Eve — deactivation of genes. Maybe???

    Doesn’t explain it in marine mammals, however…

  3. Of course it is evidence for evolution, but here’s what the creationist reply is going to be:
    1. All snakes are the same thing to a creationist.
    2. Therefore, there were only 2 snakes, each with fully-functional genes, on Noah’s ark.
    3. The pseudogenes have evolved since the flood, along with every species of snake.

    1. Re: #2, that argument only works until you get them to commit to a definition of a “kind.” Most of them will have to go with the Bible which says “reproduces according to its kind” and will give a somewhat surprisingly BSC-ish definition that reproduction isolates the different “kinds.” But then if, for example, cobras can’t breed with rattlesnakes, they have to be two different “kinds” and the hapless Biblical Creationist has to explain (1) how rattlesnakes (genus Crotalus) had to be on the Ark and how they got from Ararat to their current distribution — limited to N. and S. America OR (2) why suddenly new “kinds” that can’t interbreed can arise by evolution given that all the “kinds” had to be created by God.

      1. Hmm. Edit the last line to: “(2) why new “kinds” that can’t interbreed can suddenly arise by evolution when, before, all the “kinds” had to be created by God.

      2. I hate to explicate creationist bullshit, but I think you’re misunderstanding the “kinds” argument.

        The two snakes of each “kind” that were supposedly on the ark were likely none of the modern snakes that you see. They were archetypal snakes, and all modern snakes evolved from those two. The fact that cobras can’t breed with vipers doesn’t matter, because kind-creationists accept evolution (and attribute to it impossibly high speeds) and speciation (although they’re inconsistent about that and often say that speciation also is impossible) but say that evolution is limited to ‘within the kind’.

        So all of the “kinds” were created, but nature allows for variation and evolution of reproductive isolation within the kinds. Something like Cuvier’s debranchements, and equally outdated.

      3. Well, of course creationist arguments run the gambit, so I don’t disagree with you. But the ones who commit to the “kinds”-are-reproductively-isolated definition ask for trouble when they also commit to the no *new* “kinds” can arise by evolution. Not all of them say this, of course and I agree that many try to say that cobras and rattlers represent “evolution within a kind.” All I’m saying is that gets them in trouble because then they are stuck unable to explain gradual transitions that take you from, say, sarcopterygians to tetrapods, theropods to aviales, artiodactyls to cetaceans, etc. ad nauseum.

  4. “OR genes are probably useless underwater because they detect airborne and not waterborne odors.”

    That is probably only true of aquatic mammals. Most vertebrates are teleosts, and they certainly have ORs that detect water soluble compounds like amino acids. It annoys me very much that olfaction is characterized as “volatile” and gustation as “soluble.” Smell and taste are both just chemosensation…they are psychophysically different because they use different organs and neural processing in the animals we are most familiar with. It is a phylocentric distinction that tells us nothing about the nature of the stimulus. I am not sure but would bet a lot that many olfactory-driven mammals smell aerosolized soluble compounds. Interestingly, C. elegans has distinct sensory neurons for volatile or soluble chemicals, however I see no reason to call one of these “smell” and the other “taste,” thought that’s what the field does.

      1. Miko is right at one level – my students get a big “X” if they suggest that olfaction is the detection of airborne stimuli. Fish do indeed have separate senses of smell and taste.

        In fact, although your (and a snake’s) olfactory neurons are the only bits of their brain that are on the outside of their body (topologically speaking), the actual detection of the chemical stimuli takes place in a fluid – mucus if you’re a vertebrate, or the lumen if you’re an insect. To get through this watery barrier, hydrophobic molecules like most odorants, need to be chaperoned by “odorant binding proteins”, which then deliver the stimulus molecule to the receptor. Exactly how all this works is still unknown.

        HOWEVER, there *is* a difference between gustation and olfaction. Olfactory neurons project straight to the brain (indeed, they are part of the brain), while gustatory neurons generally have a synapse before the signal hits the clever bits.

        Secondly – and this is the topic of Jerry’s post – you can reliably distinguish between olfactory receptor molecules (ORs) and gustatory receptor molecules (GRs) by their protein structure, and therefore by the genes that code for them. And as you’d expect, there’s a great deal of difference between the ORs of aquatic organisms and those of terrestrial animals. Amphibians, you’ll not be surprised to learn, have some of both kinds. There are not the same changes in GRs.

        The prediction would be that while the ORs in sea snakes show high levels of degeneration, the same would not be true of GRs, which will not have been affected by the change of environment.

      2. Thanks for the clarification, Matthew, but I think there is still a phylocentric error in the chemosensory fields. Vertebrates have a clear smell/taste distinction in terms of sensory transduction and anatomical substrate, but NOT as you point out between volatile/soluble, which I hear from a lot of biologists. The error gets worse as you get to the diversity of chemosensation mechanisms in invertebrates. Most vertebrates are teleosts, but most animals are nematodes. In C. elegans, the literature refers to sensing volatiles as smell and soluble compounds as taste, though there is of course no homology to these senses in vertebrates (other than the use of GPCRs). This, I think, is just wrong. There are several classes of chemosensory neurons in worms, the “smell” and “taste” ones use similar sensory transduction. There are different sensory neurons specifically dedicate to volatile or soluble chemicals, but neither is “smell” or “taste,” because these modalities are properly defined by their vertebrate-specific anatomical substrates and transduction mechanisms, not the chemical nature of the stimuli. Too pedantic? Maybe. I don’t know why I worry about these things.

  5. Fascinating.

    I always learn something new from WEIT and appreciate all the authors who talk the time to explain and explore these topics. Much appreciated.

  6. One of the things I hear bleated by creationists is “show me an example of a genetic mutation that is beneficial to the organism!” Surely pseudogenes provide ample examples? If I remember correctly either they inactivate genes that could be harmful or ones that are unnecessary (e.g. OR genes in a marine animal). Thoughts?


Comments are closed.