|
Author
|
Topic: ID research problem: bacterial flagellum and type III secretory system
|
William A. Dembski
Member
Member # 7
|
posted 06. May 2002 22:55
When I recently spoke at the University of Toronto, Larry Moran, a molecular biologist on faculty there, claimed that the type III secretory system (a type of pump) was an evolutionary precursor to the bacterial flagellum. Larry, as many evolutionists, thinks that co-optation is the way you get irreducibly complex systems like the bacterial flagellum. Basically, systems targeted for other uses get co-opted into systems with novel uses. In this way a pump supposedly gets co-opted for use in a flagellum.
Now, it seems that there is an experiment that one can do to test the co-optation hypothesis. There are about 40 or so genes needed to form a bacterial flagellum. Of these 10 are homologous to the 10 genes needed for the type III secretory system (the exact numbers are not important, but that's what memory serves me).
It would therefore be interesting to take the 10 homologous genes for the type III secretory system and substitute them for the corresponding genes for the bacterial flagellum to see whether a working flagellum results. Alternatively, one could take the homologous genes for the bacterial flagellum and substitute them for the corresponding genes for the type III secretory system to see whether a working secretory system results.
CONJECTURE ON THE BASIS OF ID: In neither case will there be a functioning system. These systems are too tightly specified and won't tolerate the discrepancies between the homologues.
CONJECTURE ON THE BASIS OF DARWINIAN THEORY: It's likely that functioning systems will result since homologous proteins should be robust under substitution.
Such an experiment would not decisively confirm one theory over the other. Still, it would provide epistemic support. Also, the fact that such an experiment arises out of taking seriously the possibility of design suggests that intelligent design itself needs to be taken seriously as a research project. Mike Gene has argued as much also.
Finally, as a postscript, it's worth noting that the evolutionary biology community is not clear which came from which, the flagellum from the secretory system or vice versa. According to Nguyen L, Paulsen IT, Tchieu J, Hueck CJ, Saier MH Jr. (Phylogenetic analyses of the constituents of Type III protein secretion systems, J Mol Microbiol Biotechnol 2000 Apr;2(2):125-44), the flagellum came first based on phylogeny comparisons of sequences.
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 08. May 2002 22:06
Hi Bill,
Interesting idea. But I forsee the objection that the 'homologues' in the secretory system may have co-evolved away to the point that they would no longer be expected to function in the flagellum (i.e. the selection pressure in the secretory system would be different from the flagellum). Not saying I necessarily agree, but this seems to be an objection waiting to happen.
But, on a related vein, it is clear that these two systems are related in *some* way. That is undeniable. However, simply pointing out sequence similarity between a few component parts does *not* qualify as proof that one system likely evolved from another. All it says is that the two systems share some similarities. That is all.
However, we have to account for the similarity somehow, so from a naturalistic perspective, it makes sense that the *assumption* would be made that the two are related by descent. I, however, propose another way of looking at this.
I propose that both of these systems are indeed real IC systems that share a few common parts. However, I further propose that they are, nontheless, truly separate systems that are also IC with respect to *each other*, meaning that the transition from one of these systems to another is similar to the transition from any other protein system. In other words, the same barriers that keep other collections of genes from happening to assemble into a flagellum are the same barriers that keep the secretory system from turning into the flagellum, or vice versa.
So, I propose that what we may have here is an example of common design. Designers can (and often do) borrow parts from one system to implement in another. So finding component parts in one IC system borrowed from another IC system may indicate to us that the same designer (or designers who communicated) was at work.
I suspect that as we uncover more examples of IC, we're going to see more borrowing of parts between separate systems. It makes sense from a design perspective. If you've made a good part, there is no reason not to reuse it.
Thanks,
Jay
IP: Logged
|
|
Columbo
Member
Member # 113
|
posted 09. May 2002 01:47
Greetings Bill,
After reading through your post, and Jay's reply, it occurred to me that another approach, perhaps slightly easier to accomplish, would be to subtract one set of homologous genes from either system (flagellum or pump) and see if the remainder functions. Doesn't it make sense that if some set of genes were co-opted into some other working set, that the added one could in principle be substracted again?
Columbo
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 09. May 2002 09:57
Mr. Dembski,
Maybe I’m jumping ahead a bit, but if Darwinian evolution is discredited as an explanation of the development or evolution of bacterial flagellum, such a finding would not eliminate all naturalistic explanations. At least in my opinion, it seems possible that bacterial flagellum could be ‘designed’ or developed by what by might be called cell level or genomic intelligence.
Consideration of this possibility would begin by breaking the design problem into the materials problem( the 10 proteins required) and the assembly problem (putting the ten proteins together in the proper order). My comments here are limited to a consideration of the assembly problem.
The assembly problem can defined or modeled as a ‘change of state’ problem. Each step in the assembly process represents a ‘change in state’ or ‘action which changes the assembly conditions’. I like to visualize change of state problems in terms of multiple decision point mazes. At each step in the assembly process, a option or action must be selected from a set of possible options.
For the discussion here, we don’t need to know what the exact assembly instructions are or how many options are available at each change in state. It need only be noted that these assembly instructions exist and can be observed and analyzed since ‘assembly of flagellum’ is a commonly occurring and observable process.
Having defined the assembly problem in this manner, there are two questions to be addressed- 1)How is the assembly problem solved? And 2)Is cell intelligence or processing capacity adequate to solve the problem?
There are, as far as I am aware, two basic methods of solving the assembly problem. The first is ‘to know the solution’, or to have the solution stored in some format. If the assembly problem was solved by some type of external intelligence, then one might expect the solution or code for the solution was inserted directly into the bacteria.
The second basic method is a variation of the trial and error method. In simple terms, many different combinations of possible assembly instructions are tried, and when a successful combination is found, the successful instructions are ‘saved’ for repeated use in the future.
Before considering whether cell intelligence is adequate to solve the assembly problem for flagellum, it is useful to note that biological systems routinely solve and store solutions to this type of change of state problem. Writing a book, writing a song, learning to properly hit a baseball or golf ball, can all be described as complex assembly problems.
It is also useful to note that there are all sorts of tricks and gimmicks which can reduce ‘designing a complex assembly process’ to a set of relatively simple design problems. Again it is worth noting that the ‘design a flagellum assembly’ problem, while complex, is not unimaginably complex. Furthermore, there are known processes in nature which routinely solve equally complex design problems.
If the ‘design a flagellum assembly’ problem is viewed as a change of state problem, the second question is "Do cells have the intelligence or processing power to solve the problem?" If intelligence is defined as the ability to perform trial and error problem solving, then the question "do cells have the ability or intelligence to solve the design a complex assembly problem" can be reduced to a series of specific technical questions such as "Do cells have the ability to store assembly instructions?’, "Do cells have the ability to test or try different assembly processes?", "Can cell distinguish between successful and unsuccessful assembly processes?".
If cells have the ability or intelligence to solve the ‘design a flagellum assembly process’ then the processes required to perform this design should be identifiable and verifiable. If the cells do not have the ability or intelligence to solve this problem, then, at least in my opinion, it should be possible to demonstrate that the required processes do not exist. It would appear that the existence or non-existence of this specific type of cell intelligence can be demonstrated by verifiable scientific analysis.
One last ‘philosophy of science’ comment. If it can be demonstrated that cells do not have the ability to design assembly instructions, then it would appear that those instructions must have been ‘introduced’ into bacteria. Personally, I can’t see how, from a scientific or logical perspective, how there is any difference between "assembly instructions were introduced by an intelligent designer" or "assembly instructions were introduced by some extremely rare ‘random’ event". I would take this observation one step further, and suggest that while ‘one time historical events’ undoubtedly occur and have significant impacts, there is no material difference between attributing these events to ‘acts of god’ or ‘acts of random chance’. Scientific analysis and scientific theories, while recognizing the existence of one occurrence facts, are primarily concerned with processes with multiple, observable, verifiable, and testable occurrences.
To briefly summarize my comments, I am suggesting that the "design of bacterial flagellum" is more than just a test of Darwinian theory. Cell or genomic intelligence may provide a naturalistic explanation of ‘design generated by intelligence’, and the bacterial flagellum would seem to offer an excellent opportunity to test for the existence or non-existence of such an intelligence.
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 10. May 2002 09:04
I'd be interested in seeing the results of the experiment suggesting by Bill and expect that eventually someone will do this. However, I have to agree with Jay that if the parts could not be swapped out, this could be explained by divergence through some kind of neutral co-evolutionary drift. After all, it may also hold true that a part from a flagellum cannot substitute for its flagellar homolog in a distantly related species. Thus, one might want to identify a type III part of the flagellum in E coli and see how well it can replace its homolog in other flagella and type III secretory systems. While this may not work for or against a darwinian view, it might inform us about the degree of CSI contained with these parts.
I would predict, however, that the type III versions of FliF, FliG, and FliN cannot substitute for their flagellar cousins (these proteins make up the M-ring and C-ring, structures which must be built before the secretory component is added). This is because it seems rather clear to me that the hypothesis of the type III system evolving from the flagellum is on much stronger ground than the notion that the type III systems represents a "precursor" of the flagellum. Since evolving from some flagellum, the type III transport system appears to have lost its ability to engage in rotary transport. The flagellar motor is composed of five proteins: MotA, MotB, FliG, FliN, and FliM. It is worth pointing out that the type III systems have no homologs for MotA, MotB, or FliM. The Mot proteins are essential components of the motor, as they are membrane proteins that fulfill two functions: they transport ions to provide the energy for rotation and serve as the stator against which the rotor (FliG, FliN, FliM) moves. What's more, the type III rotor components have significantly changed. The type III homolog of FliN shares sequence similarity only with in its C-terminal 80 amino acids. And the sequence similarities between the FliG homologs are almost non-existent. Furthermore, there have been significant changes in FliF. FliF forms the M ring (the "mounting plate"), which is associated with and above the C-ring composed of FliG, M, and N. FliF in flagella is composed of 500+ amino acids, but in the type III homolog, both the C- and N-terminal domains thought to be involved in forming the MS ring are missing. All that is left in common between them is a central region of about 90 amino acids.
Jay raises the hypothesis of common design to explain the similarities. While this hypothesis may hold true for certain systems, my hunch is that it doesn't hold true here. While attempts to account for the origin of the flagellum through the type III system are vague, completely ad hoc, and quite insufficient, the evolution of the type III system from the flagellum is rather straightforward, involving loss of parts and stream-lining around an already existing function. Type III systems are embedded in all flagella. And there are recent data that show both that flagella are involved in virulence and secrete virulence factors. Thus, while the type III systems Moran has in mind are a far distance from the flagellum, flagella already are, for all practical purposes, type III secretory systems.
IP: Logged
|
|
Jay
Member
Member # 268
|
posted 10. May 2002 12:21
Hi Mike,
You obviously have a better grasp of the kind of components that each of these systems contain, so I'd like to pick your brain just a bit more!
You say: "Type III systems are embedded in all flagella. And there are recent data that show both that flagella are involved in virulence and secrete virulence factors. Thus, while the type III systems Moran has in mind are a far distance from the flagellum, flagella already are, for all practical purposes, type III secretory systems."
This, however, would seem to imply that the flagellum is not as IC as we think, as an apparently simpler core is still functional. Is this a correct reading?
I was under the impression that the type III system did contain a simpler core of proteins than the flagellum, but that it also contained new proteins and new sub-domains within shared proteins that, in essence, made it another separate IC system.
I envisioned that if we simply lost parts from the flagellum we would still have to add on several more interacting parts in order to get the new function, making the transition from flagellum to type III difficult for the same reasons that it is difficult to get a flagellum in the first place (but just maybe a bit easier since some proteins are already there).
Currently, most of the papers dealing with phylogeny of systems like this, especially deep phylogeny, take any stretches of similarity between any parts of the two systems to mean homology of the two systems. But again I suspect, from a design standpoint, that we're going to find systems that do share some parts, or pieces of parts, but are nontheless IC with respect to each other, and/or their natural histories will better suggest that they are truly separate systems, and likely always have been. Likewise, I think we'll find experimental data to support their isolation from each other. This would, again, make perfect sense from a design standpoint. The intelligent mixing and matching of parts within different, separate systems seems like a very likely outcome from common design.
So basically, I'm not yet convinced that these two systems are best explained by one giving rise to the other as I still wonder whether IC barriers between the two exist. On the other hand, I can't say that I know enough about these two to really hedge any bets on this, as I've mostly been looking at other systems in drawing this inference. But you never know.
Thanks,
Jay
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 10. May 2002 23:59
Hi Jay,
When we consider these machines, I focus only on the parts that make up the machine and not the various proteins used to express the genes. For example, while many scientific reviews cite the bacterial flagellum as being made up of 40-50 genes, such a part count involves all the various transcription factors, proteins that regulate the transcription factors, and soluble chaperones. And sometimes, chemotaxis proteins get thrown into the mix. But if we focus on the structural components alone, the typical flagellum is made up of about 21-23 proteins (depending on whether it is found in gm- or gm+ bacteria). And when we turn to the structural components of the type III secretory systems (TTSS), they are typically composed of about 14 proteins. Of those 14, it is fairly well accepted that 10 are homologous the exporting machinery, the flagellar M ring, and the flagellar C ring.
Of the remaining four, two are probably homolgous to the P ring and L ring. This, of course, is further support for the TTSS-from-flagella hypothesis, as these rings are only found in gm- negative bacteria, just as the TTSS is restricted to gm- bacteria.
The remaining two (which form the needle) might very well be homologs to the hook protein and its chaperone. There are some very interesting similarities here. The length of both the needle and the hook are tightly regulated. In flagella, the gene fliK is very important in this regard. FliK mutants show a "polyhook" phenotype, with greatly elongated hooks of undefined lengths (needless to say, these are nonmotile). Overproduction of hook protein also gives the same phenotype. If we turn to the TTSS, there is a protein fairly similar to fliK known as InvJ. Mutants of InvJ show a "polyneedle" phenotype, where the needle is grown very long in an uncontrolled fashion.
Basically, what you have with the TTSS is a stripped down flagellum. Much of evolution is about reducing complexity (remember the lesson of microsporidia). Remove the motor. Remove the filament and its cap. And probably remove the drive shaft. You form a basal body from the M ring, C ring, P ring, and L ring, and retain the hook and grow it extra long. Remember that fully functional flagella can already function as TTSS involved in virulence. For example, some studies have shown that paralyzed flagella are needed for virulence in at least two different species of bacteria. Thus, the evolution of the type III machinery would probably not need to involve cooption or changing functions.
IP: Logged
|
|
warren_bergerson
Member
Member # 262
|
posted 11. May 2002 15:01
I am getting a little confused about the goal of this experiment. The first part of the experiment is obviously to define the assembly process involved in producing flagellum. As Gene’s comments clearly show, a great deal is known about the assembly process, and the assembly process is very complex.
Is the purpose of the experiment to "figure out how this very complex assembly process or design developed or evolved?"’ or is the purpose of the experiment to "Test or compare two competing theories?".
If the purpose is to test two theories, what are the theories? Exactly what Darwinian process or mechanism is supposed to be at work? How many billions of years, on average, would it have taken to evolve the observe process? Is ID predicting an ‘external intelligence’ or an internal intelligence? If an external intelligence, was it a one step or multi-step process?
My earlier comments we directed at the suggestion that flagellum or the assembly process which produces flagellum, developed or evolved by what I call the adaptive paradigm. That paradigm at least provides a step by step, testable model of an evolutionary process? I have a hard time seeing what model or theory is being tested with the proposed experiment.
IP: Logged
|
|
Doubting Thomas
Member
Member # 1214
|
posted 27. March 2004 15:14
William A. Dembski wrote:
quote:
When I recently spoke at the University of Toronto, Larry Moran, a molecular biologist on faculty there, claimed that the type III secretory system (a type of pump) was an evolutionary precursor to the bacterial flagellum. Larry, as many evolutionists, thinks that co-optation is the way you get irreducibly complex systems like the bacterial flagellum. Basically, systems targeted for other uses get co-opted into systems with novel uses. In this way a pump supposedly gets co-opted for use in a flagellum.
Dr. Dembski:
Please see the paper whose abstract I posted today in Literature Reviews. It presents evidence that the TTSS and the bacterial flagellum evolved from a common ancestor.
quote:
Finally, as a postscript, it's worth noting that the evolutionary biology community is not clear which came from which, the flagellum from the secretory system or vice versa. According to Nguyen L, Paulsen IT, Tchieu J, Hueck CJ, Saier MH Jr. (Phylogenetic analyses of the constituents of Type III protein secretion systems, J Mol Microbiol Biotechnol 2000 Apr;2(2):125-44), the flagellum came first based on phylogeny comparisons of sequences.
The response to the paper you cite by Nguyen et. al. can be found in the more recent paper I posted on page 160. On page 155: quote:
Thus the suggestion that the TTSS evolved from flagella {...;Nguyen et al, 2000) by what can only be called 'reductive evolution' receives no topological support from the phylogenetc trees [presented in our paper].
[ 27. March 2004, 15:58: Message edited by: Doubting Thomas ]
IP: Logged
|
|
zenheadache
Member
Member # 1233
|
posted 21. April 2004 20:58
Is there any solid evidence of cooptation occurring anywhere in the development of the flagellar motor, or is it all just conjecture for this particular machine?
Dr. Dembski wrote: Larry, as many evolutionists, thinks that co-optation is the way you get irreducibly complex systems like the bacterial flagellum. Basically, systems targeted for other uses get co-opted into systems with novel uses. In this way a pump supposedly gets co-opted for use in a flagellum.
I am curious about this notion of cooptation. Might these "systems targeted for other uses" also be systems that have "novel" uses before being coopted? That is, could it be the case that IC systems are coopted to become other IC systems? And if that were so, how would it help us to explain the gradual development of an IC system by means of another IC system?
It seems possible to imagine lots of things about how the flagellar motor evolved when there is an unclear or incomplete picture of its history.
Finally, can someone tell me why is it NOT possible for the TTSS to be a molecular "factory" producing the entire flagellar motor?
IP: Logged
|
|
Mike Gene
Member
Member # 149
|
posted 16. June 2004 07:14
From: Milton H. Saier, Jr. Evolution of bacterial type III protein
secretion systems. TRENDS in Microbiology Vol.12 No.3 March 2004
quote:
In considering the logistics of the conclusions made by Gophna et al. [10], it should be noted that Fla systems have been able to diverge in structure so as to span either one membrane or two in the envelopes of Gram-positive and Gram-negative bacteria, respectively. So why haven’t TTSSs? The most plausible explanation considers that TTSSs arose late from pre-existing Gram-negative bacterial Fla systems. Independent evolution of bacterial and eukaryotic flagella does not imply simultaneous evolution; because bacteria were present on Earth at least a billion years before eukaryotes [24], survival-promoting organelles such as flagella would be expected to have arisen in the former organisms long before those in the latter organisms. Although Chlamydia are distantly related to Proteobacteria, either they or the Proteobacteria could have acquired their TTSSs by horizontal transfer; there is no reason to suppose that they were acquired by vertical transmission and that the evolution of TTSSs therefore predated the split between Proteobacteria and Chlamydia. In fact, there is a tantalizing possibility that TTSSs might have evolved in ancestors of the chlamydias because some TTSS components in the chlamydias are placed at the bottom of the phylogenetic trees of flagellar export proteins and TTSS proteins [25]. In addition, because chlamydias are obligate parasites, recent gene transfer events might have been less probable in these organisms than in proteobacterial pathogens that harbor TTSSs. Regardless, none of these considerations suggests that TTSSs predated Fla systems. If ancient TTSSs functioned to secrete proteins for purely bacterial functions other than motility, why have such functions apparently been lost? Branch lengths on phylogenetic trees merely reflect degrees of sequence divergence, and it is well established that different protein families have undergone sequence divergence at very different rates [26,27]. In the case of TTSSs, rapid evolutionary divergence relative to Fla systems would be expected, first, because of the documented lateral transfer of these systems between organisms with very different properties, and second, because of the different types of substrate proteins required for establishing symbiotic or pathogenic relationships with the different eukaryotic host organisms. Based on these considerations, it is premature to accept the suggestion of Gophna et al. [10] that TTSSs were the precursors of Fla systems.
You can actually strengthen Saier’s argument significantly:
1. Flagella are not only found in Proteobacteria and Gram-positive Firmicites, .but also in spirochetes and the deepest branching eubacteria – Aquifex and Thermatoga. TTSS are lacking in these.
2. TTSS are associated with symbiotic relationships, often including parasitism. It has been well-documented that such relationships often entail degenerate evolution (think of the tape worm or microsporidia). Thus, the TTSS as a degenerate flagellum fits the context of the overall data quite well.
3. The similarities between the TTSS and flagella extend beyond the structural homologs and include mechanisms of regulation and assembly of the filament/needle. At this point, the “common ancestor” hypothesis becomes superfluous, as the common ancestor is looking more and more like a full-blown TTSS, leading us to ask which came first, the TTSS or the flagellum. Seen in the context, the later explanation is much better supported.
4. If you look at Gophna’s FliI tree closely, the TTSS homologs do indeed seem to nest within FliI sequence. Given the FliI is the ATPase, we might expect this one to best reflect a relationship.
IP: Logged
|
|
The Deuce
Member
Member # 992
|
posted 17. June 2004 14:50
Jay and Mike's comments have got me thinking of another possible experiment: strip a flagellum down to its TTSS homologs. See how much extra work it takes to get the resulting structure to work as a fully functioning TTSS. Try to figure out which circumstances, if any, the resulting structure would be useful in, whether the necesssary changes are likely to happen on their own, whether intermediate forms would be viable, etc.
IP: Logged
|
|
|
Printer-friendly view of this topic