Yesterday a really cool paper came out in the journal Nature that demonstrates why evolutionary theory is so useful and fruitful in biology. A team of researchers has recreated an ancestral bacterial protein to determine that the ancestral bacteria grew in hot water around 3.5 billion years ago.
How can you reconstruct the protein sequence of organisms long gone? Using evolutionary theory, you can take the genomes of existing related organisms, and work backwards to the sequence of their common ancestor. This has been done to reconstruct ancestral steroid receptors, archosaur visual pigments and ancient coral fluorescent proteins. Of course, there are several technical issues involved in reconstructing the phylogeny and the ancestral sequence. You have to have the right phylogeny to construct the gene of the ancestral organism, and there are a number of issues that revolve around the rate of nucleotide replacement and biases in the frequencies of amino acids over time. However, these can be addressed in a number of ways, including using powerful statistical techniques and testing alternative phylogenies.
Gaucher, Govindarajan and Ganesh examined the thermostability of Elongation factor Tu (EF Tu) proteins reconstructed by phylogenetic analysis. EF Tu is critical in protein synthesis, and there are lots of EF Tu sequences available to build reasonably reliable phylogenies. Importantly, the stability of Ef Tu is known to correlate with environmental temperature in extant organisms. So the stability of reconstructed EF Tu should give us a readout of the environmental temperature at the time the ancestral organism lived.
The authors were able to reconstruct the EF Tu proteins at several nodes in the phylogenetic tree, corresponding to different times between 500 million years ago and 3.5 billion years ago. The reconstructed proteins show a progressive increase in thermostability, suggesting that bacteria living around 3.5 billion years ago were living in very hot water. This analysis can’t determine whether the organisms were living it hot springs, or if the entire oceans were hot, but the thermostability of the proteins correlate strikingly with oceanic temperatures inferred from oxygen and silicon isotope ratios.
Figure from Gaucher et al, Nature (2008) 451, 708
This analysis, deeply rooted in evolutionary concepts, has allowed us to infer something important about the environments of ancient creatures. Also, by generating functional proteins, we can better understand the mechanisms of the enzymes themselves. Contrast this with Intelligent Design, which has yet to make a useful prediction of any kind.
Also note that functional proteins were produced. This is also true of the reconstructed ancestral steroid receptors, archosaur visual pigments and ancient coral fluorescent proteins. The Intelligent Design movement in general has wedded itself to the idea that functional proteins are “extremely isolated” and that you require an intelligent designer to get from one isolated island of function to another. Thus, according to ID, you cannot infer the ancestral structure of a protein from existing organisms. The successful reconstruction of multiple, function proteins by evolutionary phylogenetics represents a major blow to this ID concept.
References
- Eric A. Gaucher, Sridhar Govindarajan & Omjoy K. Ganesh (2008) Palaeotemperature trend for Precambrian life inferred from resurrected proteins Nature 451: 704-707
- Thornton JW, Need E, Crews D (2003) Resurrecting the ancestral steroid receptor: ancient origin of estrogen signaling. Science 301: 1714–1717.
- Chang BS, Jonsson K, Kazmi MA, Donoghue MJ, Sakmar TP (2002) Recreating a functional ancestral archosaur visual pigment. Mol Biol Evol 19: 1483–1489.
- Chang BS, Ugalde JA, Matz MV (2005) Applications of ancestral protein reconstruction in understanding protein function: GFP-like proteins. Methods Enzymol 395: 652–670.
Michael Behe would say that this is an uninteresting result. Completely trivial. Not surprising–no, not at all. In fact it’s downright boring. Getting sleepy now, … zzzzz.
There are older references as well concerning ancestral reconstruction. All of them are exciting to me. It’s like the ultimate form of paleontology and you don’t have to get sunburned doing it!
Very nice result. Thanks, Ian, for bringing this to our attention.
LOL!
Sure it does. It predicts your result. But not just that; it also predicts that the two protein lineages arose independently, that the proteins diverged in-vivo by “saltation,” and that an ancestral protein had all the features of descendants and that mere loss of “information” caused the divergence. ID also predicts any time frame you want for the divergence, abiogenesis, even origin of the Universe. Even if what you want is “5 minutes ago.”
Oh, you said useful prediction. Never mind.
But it’s still a protein! (Well, someone had to say it).
Seriously, the ID crowd will probably just say that you can never be sure if you are right or not unless you find the “missing links”. Of course, some will eventually be found and sequenced, so then they will have to retract their statements and admit that science was right and that they were wrong all along. What? Oh yea, I forgot, they’ll never read the papers. Never mind.
They have a trivial tactic for dealing with no-longer-missing links, because the more links you have, the more “gaps” there are between them.
That sounds like they’re working a jigsaw puzzle that has lots and lots of pieces, and no picture on the box to show what it’s gonna look like when done.
Henry
I proposed doing the same thing decades ago. Not with bacteria. With dinosaurs. In theory with enough phylogenetic sequence information we should be able to reconstruct the genome of…dinosaurs.
From there it is a simple genetic engineering problem to recreate large extinct reptiles.
I’m tempted to say that Spielberg and Crichton stole my idea for Jurassic Park, but they came up with it independently and much later.
In practice one doesn’t need the whole genome, just the controlling morphogenetic/developmental biology loci. It will most likely be harder than it sounds at first glance but Jurassic Park is coming within the grasp of modern science.
That’s a seriously hot ocean! Thanks for the informative post.
But you have to already have faith in evolution and naturalism to get such a result. You know, sort of like having to have faith in the second law of motion before you can calculate the thrust of a rocket.… Oh, uh, never mind that.
The most important thing is that you can’t ever show that a designer didn’t have something to do with the course of the changing proteins, so ID will always be science. I’m ready for my awards for pointing this out, and the textbook orders for my book. They’ll save money, too, since it only takes a couple of pages to point out that a designer could have done things without being detectable, and to figure out that God did it from that.
And you guys thought science was hard, and a lot of work. All of that useless sequencing and hypothesizing, and it’s still altogether too plain that you can’t rule out a designer. Don’t you feel foolish now!
Glen D
http://tinyurl.com/3yyvfg
Technically with sequence data it is quantitized so it is possible to get around the multiplying gaps sophistry.
A not-so-trivial analog to that - and a Catch-22 for any scientific theory - is that the more evidence there is, the more there is to cherry pick to make the theory look weak to those who don’t have the time or interest to check that the evidence isn’t cherry-picked.
That’s why I think that Pope John Paul II’s phrase “convergence, neither sought nor fabricated” is the one of the most powerful defenses of evolution against misrepresentation. Anti-evolution activists know that the only way to make evolution look weak is to seek and fabricate “evidence.” And now with ID, to cover up the fact that potential alternatives have diverged into a hopeless mess of mutual contradictions.
ooo,ooo,ooh
oh yeah, bones only paleontology is long gone
But, where are the Neutral Gaps ?
I think that the study may be being misinterpreted. Really what is being argued is that the earliest-diverging lineages of *extant* life lived in hot environments. Taking a step from there to saying that the oceans were hot is another step. One might argue an alternative view, e.g. that what this study shows is that the ancestors of modern life developed some key features while living in high-temperature environments, and subsequently spread out and beat out all of the other microbial life forms then existing.
(I have also seen it argued that the basal-thermophile observation is an artefact due to DNA & amino acid sequences changing more rapidly due to e.g. adaptation from moderate to extreme environments. I assume the authors thought about that & dismissed it for reasonable reasons but I haven’t looked into it.)
Also – I have seen claims that global glaciation events are known in the precambrian geological record, how does that match up with the geological evidence for higher temperatures?
The Hadean eon lasted until 3.8 billion years ago:
http://en.wikipedia.org/wiki/Hadean
and is currently thought to be life-free. The Paleoarchean era is currently thought to contain the earliest fossils but there is no evidence of glacial activity:
http://en.wikipedia.org/wiki/Paleoarchean
The earliest definite evidence for massive glaciation (of which I am aware) is much later, during the Neoprotozoic era:
http://en.wikipedia.org/wiki/Neoproterozoic
by which time life forms had been evolving for about 3 billion years.
Oh, oh!
Don’t you get to use the “blogging reviewed literature” thingy? I think you should, if only to piss-off Casey Luskin. Piss-on, piss-off, kind of like the Disko Kid.
Ahem, I hate to break up the singing to the choir, but regarding:
I am still waiting for an answer. How can you reconstruct the protein sequence? This is a legitimate question. More specifically, how can you know that you’ve reconstructed the protein sequence?
If I simply assert that super studly researchers at the von Trap institute for linguistics have reconstructed the pre-historic word for dwelling by using linguistic theory, and extrapolating backward in time words for dwelling including: woning, σπίτι, Wohnung, домашн, domestico, moradia, κατοικία, and жилище, and came up with wπмmestοιщеung, many would consider this little more than pseudo-scientific technobabble.
The simplest way to do the reconstruction William is using an established phylogeny and then using a parsimony analysis. The more sophisticated approach would involve Maximum likelihood methods. Either approach faces substantial uncertainty at poorly conserved residues, but these days it is feasible to analyze a number of these mutations to check for what presumably in most cases (though not all because of the potential for epistasis) are neutral or near neutral variants. The statistical methods and molecular techniques have improved but the idea is atleast as old as an Emile Zuckerkandl and Linus Pauling paper on hemoglobin from the 60’s
“More specifically, how can you know that you’ve reconstructed the protein sequence?”
We can never “know”, we can only make reasonable inferences based on current data.
“If I simply assert that super studly researchers at the von Trap institute for linguistics have reconstructed the pre-historic word for dwelling by using linguistic theory, and extrapolating backward in time words for dwelling including: woning, σπίτι, Wohnung, домашн, domestico, moradia, κατοικία, and жилище, and came up with wπмmestοιщеung, many would consider this little more than pseudo-scientific technobabble.”
Imagine that you have a robust linguistic theory of which letters tend to change into other letters and at what likelihood. At every stage of the reconstruction, imagine that you have ways to check if the reconstructed word would have been likely to be a real, usable word in a real language, or a non-viable chimera (as your ancestral word seems to be).
That’s why they said (and I reported, emphasis added)
There was a lot of glaciation around 600 Million years ago (I’ve even stood on some of these glacial deposits in the Flinders Ranges), but around 3.5 Gya there is no current evidence for glaciation. And most of the period covered by the reconstruction is outside the glaciation times.
William,
In fact linguists have done precisely that, just google proto-indo-european language. By gathering all languages of the indo european language tree, and also written records of extinct indo-european languages, focusing on words that would exist since the start, such as ‘mother’, ‘father’, ‘goat’, ‘dwelling’, etc. Linguists has reconstructed the best approximation of the progenator of indo-european languages. Along the way the most probable phylogeny is also reconstructed. Since writting came after PIE (proto-indo-european language) any written version of PIE is merely a transliteration.
As I said above you use phylogenetic methods to work backwards from extant sequences. The details get technical, and comments can’t display diagrams, so I can’t use diagrams to explain, but this paper and this paper should give you a handle on it (not the testing they do to check the methodology).
Reconstructing languages using comparative methodologies is a well established field, reconstructing proteins is actually methodologically simpler.
Why, you test it of course. Some of the tests are mentioned right here:
- protein functionality
- correlation of stability (which is truly beautiful, though I gather from the discussion not a rigorous test).
And I assume in other types of phylogenetic reconstructions later and independent reconstructions would end up close to earlier ones, thus constituting a test. I dunno if one can make such a test here, though.
You got to love science, who takes its results through as many tests as it can come up with. As opposed to, say, creationism btw.
Exactly. And the fact that ID is science, and not “creationism,” is plainly demonstrated by the numerous tests, since 1996, that Behe has conducted to support his “front-loaded ancestral cell” hypothesis (the only clear anti-evolution position any major IDer has taken, BTW, much to the chagrin of YECs). From those dozens of peer-reviewed publications on the subject, the one that clinched it for me was the human pseudogene for chlorophyll.
What? He didn’t do that? Oh. Never mind.
Nick (Matzke) said: “Also – I have seen claims that global glaciation events are known in the precambrian geological record, how does that match up with the geological evidence for higher temperatures?”
Global mean temperatures are may not be very relevant. Shortly after an initial crust solidified, there were almost certainly places that were very hot for a very long time. The mid-atlantic ridge is one such contemporary example.
Orr’s article, which I have read before, is very efficient. I think I’m in a position to enjoy it even more the second time around. For example, Orr plays a bit fast and loose with the dating of the first cell, but can draw the correct conclusion anyway.
So, sensu Dawkins, my question to Behe would become if Behe is expressing a non-functional pseudomeme for interlocking complexity (“irreducible complexity”) that His Designer (HD) provided the Primal Soul, why didn’t HD provide creationists with functional memes instead?
Thanks to you and others. Sounds like an interesting parlor trick all and all, but short of a time machine, it doesn’t sound exactly falsifiable.
William Wallace
Probably not strictly falsifiable, but if the resultant protein was non-functionable it would be strong grounds for considering it to be not ancestral. Similarly if different research groups, starting from different organisms, came up with very different ‘ancestral’ proteins.
BTW, what does ID/Creationism have to say on the topic?
The same procedure of phylogenetic sequence analysis can be used to recreate the genomes of extinct animals.
You will know it works when you see a Tyranosaurus chasing a herd of duck billed dinosaurs.
[There are independent checks on this sort of analysis. Many or most changes at random are more likely to result in a nonfunctional or lowered function protein, so that the protein is functional is strong evidence. It is also orders of magnitude harder to make a protein more thermostable than to make one temperature sensitive. That the protein is more thermostable is intriguing. One can also phylogenetically predict the sequence of extant creatures and then check by routine sequencing.]
They “recreated”.….?!?
I don’t think you should use the word “creation.”
It should have said they “randomly facilitated a random repeat of something that happened…”
Of course we all know it could have never randomly happened. There has to be some creation… just like everything else.
Why not? The Big Bang created a universe. Evolution created a biosphere.
Creationists create lies. Although to be fair, they aren’t very good at it and end up recycling the same old fallacies for centuries and millenia.
William Wallace,
Skimming your new blog I can see a little better why the DI is giving up the pretense that their ideas are motivated by anything but a religious right ideology. Nevertheless I respect your religion and political POVs; I probably agree with you more than not on the latter, and I also am a theist if not a member of any organized religion.
My questions for you, however, have nothing to do with any of that, or any of the ID sound bites that you like, or any problems with “Darwinism” you may have. Rather I’m only interested in whether you agree with Michael Behe that:
1. Life on Earth has a 3-4 billion year history.
2. Modern humans are related to other animals, plants, etc.
Note that neither answer requires that you accept a Darwinian mechanism. And even if you are genuinely unsure, you seem confident enough in your other ideas that you could easily offer a “best guess.” If you have answered those questions elsewhere, a link will suffice.
Thanks in advance.
Update