Putative Hook Protein in Thermotoga maritima

Thermotoga maritima are thermophilic bacteria that branch deep in the eubacterial tree [1] and also possess flagella. In fact, even though these bacteria are distantly related to such bacteria as B. subtilis and E. coli, they share all the basic structural genes for the flagellum (everything from fliF to flagellin). However, when surveying the flagellar genes, I discovered the hook gene (flgE) is annotated as an "authentic frameshift" [2]. How is one to interpret this? Is this really an authentic frameshift (rendering flgE a pseudogene) or has there been some error associated with the sequencing?

This protein, which mediates the torque generated by the motor to the filament, is found in all other bacteria with flagella, shows good sequence conservation, and genetic studies have shown it to be essential for forming a functioning flagellum [3,4]. Thus, from an IC perspective, we would predict that Thermotoga does indeed express the flgE hook protein and there is some type of sequencing error. But that would be a risky prediction given that TIGR already lists flgE as containing an "authentic frameshift."

So I snooped around a bit and eventually found an interesting web page. [5] It turns out the Joint Center for Structural Genomics is looking at the "pseudogenes" in Thermatoga. They have already successfully expressed and crystallized three other proteins previously described as "authentic frameshifts," including the motor switch protein, FliY. Currently, flgE has only been amplified. Thus, I will go out on a limb, with IC in hand, and predict that eventually flgE will likewise be found.

1. http://www.tigr.org/CMR2/BackGround/btm.html

2. http://www.tigr.org/tigr-scripts/CMR2/GenePage.spl?locus=TM0674

3. "We now report the construction of a specific flgE mutant of T. denticola ATCC 35405 following electroporation utilizing an erythromycin resistance cassette inserted into an flgE DNA fragment. The resulting mutant displays no visible motility and lacks periplasmic flagella as would be predicted from inactivation of the gene for the flagellar hook protein." - Li H, Ruby J, Charon N, Kuramitsu H. 1996 Gene inactivation in the oral spirochete Treponema denticola : construction of an flgE mutant. J Bacteriol 178(12):3664-7

4. "Mutants of H. pylori and H. mustelae defective in hook production generated by allele replacement were non-motile and devoid of flagellar filaments but produced both flagellin subunits, which were localized in the soluble fraction of the cell." - O'Toole PW, Kostrzynska M, Trust TJ. 1994. Non-motile mutants of Helicobacter pylori and Helicobacter mustelae defective in flagellar hook production. Mol Microbiol 14(4):691-703

5. http://bioinfo-core.jcsg.org/dasp/tm_pseudogenes.htm

[ 28 February 2002: Message edited by: Mike Gene ]

Thanks for the welcome.

The general logic is as follows. Say you have a biological system X involving 10 components (proteins and/or RNA). You hypothesize that this system is IC, that is, all 10 components are needed for function F in system X. Removal of various parts in a few highly studied species support this contention (as removal of one part is associated with the complete loss of F). But you further hypothesize that this set of 10 components is not species-specific, but instead, reflects the generic constraints imposed by "system X" in light of the function (a top-down engineering perspective). This would allow you to predict that wherever you looked, if you found variants of system X, and system X is functioning, you will still find parts 1-10.

In this example, I found an example of system X that apparently is missing part 7 (Thermotoga "missing" the hook protein). I say apparently, because the parts are identified indirectly through their encoded instructions (genes); the gene for one flagellar part is noted as having a frameshift mutation, which would essentially render the part nonexistent. From an IC perspective, this would be like encountering a claim that a functioning mousetrap had been discovered that lacked a hammer. One can either concede the hammer is therefore not part of the IC system or infer that the claim is mistaken. If you have a good feel for the design of the mousetrap and understand the role each part plays, you'd likely reject the claim.

I don't think a Darwinian approach can provide such guidance. To explain the origin of IC machines, this approach assumes proteins are essentially low information structures and/or are rather easily coopted into new protein complexes with novel functions. From a purely Darwinian approach, one could argue that the hook has been replaced by some other protein that previously was not associated with the flagellum. Or one might argue that even though the hook protein has lost all its "hook" information (due to the frameshift mutation), by chance, it stills retains some form of structure capable of substituting for the canonical hook function.

Put simply, IC gives us reason to expect the hook protein in Thermotoga to be expressed and used (in spite of the contrary data that exist today). The Darwinian perspective, as far as I can tell, provides no expectation either way, but would simply account for whatever is found.

For instance, let's say it turns out that the Thermotoga flgE gene is indeed a pseudogene.
Would that invalidate ID as a theory? Unlikely, since there are so many other irreducibly complex (IC)-looking biological systems.
In fact, it wouldn't even invalidate the IC status of flagella, since you could now claim that the flgE-deficient flagellum is the prototypical IC machine, and all the others just happen to have a fancy add-on (flgE) that is really unnecessary.
After all, this is what happened to the other IC paradigm system, the complement cascade, which was shown to miss one component in urochordates (Smith et al, Immunol. 156, 593, 1996;
Ji et al, Proc. Natl. Acad. Sci USA 94, 6340, 1997; Nonaka and Azumi, Dev Comp Immunol 23, 421, 1999).

You wrote:

quote:

---

I think what matters is not much the ability to make a prediction ... but rather what you would do once you test that prediction and get a negative result....

What kind of prediction could you make that would invalidate the entire concept of Irreducible Complexity?

---

I think that science would progress in the ID paradigm, just as it does now. As counterexamples would accumulate, either paradigmatic explanations would be made clear, or
ad hoc theories would be appended to the more fundamental theory. At some point, if the whole program was weak, it would become unstable, just as Kuhn proposed. If it proved robust, then the appended theories would just prove to be refinements on the original idea. Do you see a problem with this way of understanding the progression of science?

I see a contrast between: 1. a prediction that when shown wrong, completely undermines a scientific notion or 2. a prediction using a scientific notion, with the knowledge that if this scientific notion repeatedly fails to produce positive hypothesis, you throw it out

Most scientists are trying to model and understand the world using scientific concepts. I would venture to say that Mike Gene is on to something in his use of "Utility" in the title. If a scientific notion tends to help us make accurate predictions (has utility) then we keep using it, if not, we throw it out. I don't think that a scientific notion gets thrown out based on a "one strike and your out" policy.

OK. Now that I've given you my sense of the issue, I do have a bone to pick. In your post you made the following unnacceptable comment:

"anybody can do that, even young-earth creationists" - attacking a group of people

Important things first: you are right, and I was wrong, about the general use of predictions in science. Of course, one can make scientific predictions that may turn out to be wrong without having to throw away the underlying theory.
However, it is also true that predictions that do not challenge any theory (whether the one you support, or an opposite one) are of little use to the advancement of science, at least on a large scale (which I think is what the ID movement wants to do, at this point).

Since ID theory has staked so much on the concept of Irreducible Complexity, emphasizing its role in particular in a few well-chosen biological systems, I think it is reasonable to ask for a basic outline of how they would go about testing their theory in a stringent fashion, which I think Mike Gene's prediction in this case fails to do (especially in the light of the "failure" of the complement system to hold up to a similar prediction).

If I may try to put it in a positive way, I think ID might indeed turn out to be an important addition to our understanding of biology (maybe even a paradigm shift, who knows?), but to do so its proponents have to come up with ways ID theory can be put to some tough test. Since both Gene and Dembski have obviously thought hard about the matter, I would like to hear their opinion.

quote:

---

Originally posted by Columbo:
As counterexamples would accumulate, either paradigmatic explanations would be made clear, or
ad hoc theories would be appended to the more fundamental theory.

---

But my point is exactly: what would a counterexample look like? After all, we have plenty of biological systems which may appear IC at first sight, but whose evolution by gene duplication and divergence is quite obvious and likely (for instance, the myoglobin/hemoglobin oxygen transport system).

So, is the C2-deficient complement system in urochordates a failure of ID theory? After all, doesn't the urochordate complement system appear itself IC, still? Wouldn't a flgE-less flagellum also be IC? Do we need to have the ENTIRE picture of flagellar protein or complement evolution on the table, before we can say either system is not IC, or is a plausible evolutionary path sufficient?

I'll offer my perspective, limited though it is. Let's see where it get's us. I'm a mechanical engineer, not a biologist, so you'll forgive me if my response leans toward the philosophic.

You wrote:

quote:

---

But my point is exactly: what would a counterexample look like? After all, we have plenty of biological systems which may appear IC at first sight, but whose evolution by gene duplication and divergence is quite obvious and likely (for instance, the myoglobin/hemoglobin oxygen transport system).

---

[It seems to me that the ID project is not attempting to deny or discard the mechanism fundamental to Darwinian evolution. (I hope this is not patronizing...) What is fundamental to ID is the idea that certain degrees of complexity and specificity are categorically outside the domain of chance and law. {Additional comment: In less impressive cases, it is not clear whether an intelligent agent did the work. For example, I can make a pile of sand look like something nature did, but nature cannot make a machine like I can.} These cases, and these only constitute clear examples of intelligent agent intervention. I think that your expectations for a clear test of ID should require 1) a quantified description of the information in what appears to be an IC system. 2) a computation along the lines of Dembski's explanatory filter applied to the system in question. 3) Peer review and testing of the results, and publication. 4)Further investigation, to falsify the test. The problem with gene duplication and divergence, (as well as point mutation, and virtually all other known mutation processes) as I see it, is that they do not generate any novel CSI. They either copy, remove, mangle, or insignificantly modify, the morphologies they effect. A really novel contrivance is e.g. blood clotting where there wasn't any, or avian lungs structures to replace reptilian. (As described by Michael Denton.)]

... Wouldn't a flgE-less flagellum also be IC? Do we need to have the ENTIRE picture of flagellar protein or complement evolution on the table, before we can say either system is not IC, or is a plausible evolutionary path sufficient?

[A system is IC if the removal of any component renders it non-functional. If removing the flgE (protein?) does not do so, then the system was not IC. Keep looking for components to remove. If you get to a single molecule, then clearly there was no clothing on that Emperor!]

Are these legitimate questions at all, or am I just rambling?

I'd certainly say so. I hope mine are too. Good talking to you.

Columbo

[ 01 March 2002: Message edited by: Columbo ]

[ 01 March 2002: Message edited by: Columbo ]

The example I provided was raised as a specific counter to a specific claim, namely, that the concept of IC is useless. I have not found this to be true and provided just one example. My objectives are rather modest. My interest is whether a teleological perspective, an alternative approach to mainstream science, is capable of generating a sustained, coherent research program focused on the biotic world. IC is one tool and I am testing its utility as I explore the maze of data generated by the biological sciences.

You write:

quote:

---

However, it is also true that predictions that do not challenge any theory (whether the one you support, or an opposite one) are of little use to the advancement of science, at least on a large scale (which I think is what the ID movement wants to do, at this point).

---

My interest here is not in "the advancement of Science, at least on the large scale." In fact, I'm not quite sure how to make sense of this claim. My interest is in simply determining if a better understanding of the bacterial flagellum comes from hypothesizing a teleological origin. If these flagella were originally designed, what did the originally designed flagella look like? Why were they designed? Why were they designed as they were designed? Does their design suggest design elsewhere in the original cells?

You correctly note that if the flgE gene in Thermotoga is indeed a pseudogene, this will not indicate the bacterial flagellum is non-IC; it would merely mean that flgE was incorrectly scored as an IC component. It is common, even in science, for a hypothesis to be refined. This may not sit well with some, but my focus is on reconstructing the original flagellum. Now, if flgE in Thermotoga is a pseudogene, this would not, by itself, be sufficient to dislodge it as a member of the original flagellum (as it may be a more recent state), but it would raise a significant caveat, suggesting other examples (or hook-less flagella) might turn up. If this happened, one would certainly have to revisit any attempt to reconstruct the originally designed state.

In other words, Charlie, you should not read my posting as some type of challenge to non-teleological thinking. If you are trying to understand my intent, you should read it as an opportunity to peek inside another head and view how it interprets differently and independently of concerns about any challenge. If this different approach develops its own positive track record, spawns a coherent story, and succeeds in uncovering new facts about the biotic world, then it may be time to come back to that challenge.

In the end, I am simply left with some facts. I encountered the putative pseudogene in Thermotoga as a function of exploring the flagellum through an ID perspective (the TeleoLogic section of my web page has a few essays about this and many more to come). Yet my growing convictions about the flagellum in light of ID and IC provided an impetus to dig further and I did find something encouraging. Clearly, the type of thing the Joint Center for Structural Genomics is doing is something an "ID lab" would do. And my comment about the Darwinian perspective should neither be viewed as a challenge. I just didn't find it useful in trying to decide about TIGR's annotation of this gene. Would you not agree?

[ 01 March 2002: Message edited by: Mike Gene ]

I should clarify it was not "mere IC" at work here. As I obliquely explained with my analogy:

From an IC perspective, this would be like encountering a claim that a functioning mousetrap had been discovered that lacked a hammer. One can either concede the hammer is therefore not part of the IC system or infer that the claim is mistaken. If you have a good feel for the design of the mousetrap and understand the role each part plays, you'd likely reject the claim.

That said, you clearly are significantly more cautious than the leading ID proponents. While you may think that the ID theory goal is not “advancement of science on a grand scale”, the Discovery Institute seems to think just that, calling ID theory "A New Science for a New Century", and clearly stating:
"Design theory promises to revitalize many long-stagnant disciplines by recognizing mind, as well as matter, as a causal influence in the world. It also promises, by implication, to promote a more holistic view of reality and humanity, thus helping to reverse some of materialism's destructive cultural consequences."
Basically, in the biological sciences, most of the leading ID theorists very clearly aim at displacing, even replacing, Darwinism as the paradigm on which modern biology is built.

Now, from your post, you seem to be making the much more modest statement that design theory could be useful in making some informed predictions about IC molecular machines (it's not clear to me whether you believe, like Behe and others, ID also explains IC biochemical pathways). Basically, unless I am mistaken, you see design theory as a supplement to Darwinism, mostly on issues that Darwinism by its own nature is “neutral” about (of course Darwinism does not predict we are going to find every single intermediate in every single molecular evolutionary pathway). Hardly something that can reverse materialistic science, although still an empirically "useful" concept.

quote:

---

"That said, you clearly are significantly more cautious than the leading ID proponents. While you may think that the ID theory goal is not “advancement of science on a grand scale”, the Discovery Institute seems to think just that, calling ID theory "A New Science for a New Century", and clearly stating:
"Design theory promises to revitalize many long-stagnant disciplines by recognizing mind, as well as matter, as a causal influence in the world. It also promises, by implication, to promote a more holistic view of reality and humanity, thus helping to reverse some of materialism's destructive cultural consequences."
Basically, in the biological sciences, most of the leading ID theorists very clearly aim at displacing, even replacing, Darwinism as the paradigm on which modern biology is built.

---

However, what's most important, I think my post is relevant to what Mike was saying initially, that is that ID, unlike what some of its opponents have said, can be useful to make some informed predictions about IC systems. Which I am willing to agree with.

The point I want to get across, though, is that if ID will have to limit itself to be a complement to darwinism in the exploration of a subset of molecular machines, as Mike seemed to propose, then it will be quite a small achievement. Furthermore, by conceding that a failure in his prediction would not really affect ID theory significantly, Mike is leaving again the door open for the classic Darwinist objection that ID theory is not falsifiable.

Therefore, I am trying to start a discussion to see whether indeed ID theory can be used for more ambitious (and by definition, more "risky") testable predictions, with broad implications for our understanding of biology. After all, Darwin did just that when he identified the major obstacle to his theory in the fact that heredity, at the time, seemed to be blending rather than particulate. If this indeed turnd out to be the case, he said, his theory would have gone down the drain (well, maybe he said it differently). Anyway, that was a gutsy prediction, with broad biological implications, and going against the status of the knowledge of the time. This is the kind of prediction that would result in a quantum leap for ID theory, if ID proponents could come up with one.