|
Author
|
Topic: Back to the Topo II - where does it end?
|
Nel
Member
Member # 614
|
posted 06. May 2003 15:16
Nic,
IC systems can have mini-IC systems, as I showed with the motor complex and even the ExbBD-TonB complex. I'm not convinced that there is much homology here, nor that the ExbBD complex is even relevant as an evolutionary precusor to the bacterial motor. Nonetheless, the point of "unselectable" steps, which is why I brought ExbBD up, still stands even if we were to replace the MotAB-FliG system with the ExbBD-TonB system. They both seem to be equivalently irreducibly complex.
There doesn't seem to be any evidence that the Exb complex predates the flagellar motor. It is a bit simpler, but is found in very few types of bacteria. Furthermore, the similarity that Kojima et. al. cite was just between MotA and ExbB . FliG doesn't seem to have a homolog in the Exb complex.
With respect to electrostatic interaction, I don't think we can say that the channel is internal to ExbBD. TonB might function in opening the channel or causing some kind of structural change. ExbB has only 3 membrane segments as opposed to MotA which has four.
The electrostatic model of the bacterial flagellar motor seems to be a bit more then a hunch, in fact, that electrostatic model behind flagella rotation seems to be near certainty that it is correct:
Bren, Anat and Michael Eisenbach. "How Signals Are Heard during Bacterial Chemotaxis: Protein-Protein Interactions in Sensory Signal Propagation." Journal of Bacteriology. 182:6865-6873 (2000).
[/QUOTE]
I think even Blair, the author of the paper you cite for the ExbBD complex has done some work on this:
Zhou, J.D., Lloyd, S.A., and Blair. D.F. (1998) Electrostatic interactions between rotor and stator in the bacterial flagellar motor. Proc. Natl. Acad. Sci. USA 95, 6436-6441
quote:
Mutational studies of the rotor protein FliG and the stator protein MotA showed that both proteins contain charged residues essential for motor rotation. This suggests that functionally important electrostatic interactions might occur between the rotor and stator.
So it may not even matter that MotAB is homologous to ExbBD, if the electrostatic interactions are completely essential. Thus, I think Mike's objection should be taken seriously here with respect to the logistical problems with random mutations and ion channels.
I would also like to see the mutational studies that you allude to which show some flexibility in the motor functionality.
With respect to the F-ATP synthase, the illustration you link to above is quite outdated if not completely erroneus. The F-ATP synthase is not just the alpha and beta subunits, nor is it a single Fo subunit. And fliI is not homologous to the entire complex, as you previously stated. ATP synthase is composed of 2 components: F1 component (the ATPase) and the F0 component (the ion channel). However, the F1 component is broken up into 5 parts: alpha, beta, gamma, delta, and epsilon. And F0 is broken up into 3 parts: a, b, and c. Thus the ATP synthase is an 8-part IC system. All the parts needed for ATP synthesis to work: The catalytic chamber, which is alpha and beta, the rotary piston which is gamma and epsilon, the rotary motor which is a and c, and the stator, which is b and delta. b, delta, epsilon, and gamma also function as the "joint" holding F1 and F0 together. There is a lot more to the story than this but this should suffice for now.
quote:
The proton-transporting ATP synthase, F0F1, consists of a catalytic
sector, F1 or F1-adenosine triphosphatase (ATPase) (alpha 3,beta 3,
gamma 1,delta 1, epsilon 1), and a proton pathway, F0 (a1, b2, c12)(1, 2). "
Mechanical Rotation of the c Subunit Oligomer in ATP Synthase (F0F1):
Direct Observation Yoshihiro Sambongi et. al. Science 1999 November 26; 286: 1722-1724.Epsilon, when it is not attached to F0, actually inhibits ATP synthase activity. (Arch Biochem Biophys 210, 421-436)
Julie Thomas has several essays written from talk.origins which I have to put back up on the net, I keep putting it off for some odd reason. ![[Mad]](mad.gif)
These type of discussions of the actual things you are claiming homology to is important to get into the detail of, because most of the time things are still treated as black boxes where things seem easy to evolve via a stochastic process (i.e. the entire fliI is homologous to the entire F-ATP synthase is simply not true, things are a lot more complex than that.)
Although the point of FliI homology seems moot upon careful reading of the paper you cite:
Source: Blocker A, Komoriya K, Aizawa S. Proc Natl Acad Sci U S A 2003 Mar 18;100(6):3027-30. Type III secretion systems and bacterial flagella: Insights into their function from structural similarities.
I read the paper last night, and it seems you are quite wrong about this. There is no evidence for a Spa47 protein complex, a fliI that is homologous to the b subunit of the F-ATP synthase. What is shown in green in fig 3 in the paper is only an imaginary model of Spa47 based on the F1 structure and that model can be successfully docked to the real NC outline where they would imagine it would have to sit to pump secrete proteins into the channel. The evidence here is pretty weak.
The evidence that the rod structure of the bacterial flagellum would unlikely function with a single protein was not addressed. All three gene products are found in all bacterial flagella, we would expect, if the other two proteins were redundant, to find some plasticity in that area. I endorse subfunctionalization but thats irrelevant, I don't think it applies here, as the number of gene products and their sizes have held constant despite billions of years of mutation and selection. There might be, as Mike notes, a specification there.
As Mike notes, there is a problem with treating the gram-negative and gram-positive bacterial flagella as the same:
quote:
The problem faced with coopting a drive-shaft in a gram-negative bacterium is that a partial rod that does not span the periplasm and penetrate the cell wall and outer membrane provides no obvious utility. This seems to be thorny problem that will escape our analysis since the Ur-IC scoring assumes all eubacteria are related through a common ancestor and thus factors out the gram-negative-specific features.
Even without the ur-IC scoring that Mike does, the lack of L and P rings in gram-positive bacteria doesn't make larger the probability calculation for the flagellum, as it was done for E Coli.
With respect to adhesion organelles, I dont have time to tread though an entire thread to find an example of a flagellum acting as an adhesion organelle, but something tells me that you are just talking about a pilus, which I already discussed. My reason for thinking this is that since this was brought up in the immunity thread, PapD of the pilus has Ig-like domains. It is misleading to call pilus "bacterial flagellum".
[ 16. May 2003, 15:10: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 06. May 2003 15:38
Well that stinks, over half my post was cut in half for some reason, maybe it was too long. So I decided to respond to the rest in a second post.
Alternatives to common design are not at-random invokations. For example there is no good reason why the designer would "re-invent the wheel" and there is every good reason why he would reuse "design principles". In fact, re-use is currently the dominant paradigm in many areas of computer science. Michael Denton's paper points to evidence that structural homologies are no good. There are examples of common design principles at work with little to no sequence homology, Tubulin and FtsZ; kinesin and myosin, or even NaiI (to go back to the topic of this thread). So here the tables are turned, the specific similarities here are hard to reconcile with a theory of natural selection.
Now I have yet to read your paper on secretin but I have no idea what piece of the puzzle you are referring to as L and P rings are simply not needed if you don't have an outermembrane. And that leads me to this comment:
quote:
Even here we are a far cry from Dembski's strawman of getting flagellum parts from the mythical, mystical "protein supermarket".
First of all, we are still as far away from Dembski's calculation as far as I can throw you. For you are invoking the existence of a protein supermarket where all these proteins are doing other things. That is what Dembski means by the protein supermarket. So it kind of baffles me that you would call it "mystical". Secondly, what you are doing is pointing to mere similarity, imaginary similarity, or other multi-part machines, there is still that sea of non-functionality that Dembski's calculation points to, in fact, none of it has changed, not a single order of magnitude.
With regard to plasticity, archael flagella seem to be irrelevant as they are completely different from eubacterial flagellum, and even represent a seperate kingdom. What is interesting again, that it is precisely the fact, even if you think that some pores would do, some sticky out things, some transporters, that we don't see plasticity among eubacterial flagella. And this is not predicted by co-option stories. What we should see, especially taking into account the fact that all these things exist that you say may likely forge a bacterial flagellum, and yet we see all flagellum with the same parts ABCDEF, not AB and other parts, or CD and other parts. It's universal.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 06. May 2003 16:40
Yet another reply for the archive:
[A brief aside: Nelson, your sentences often run-on and seem rather breathless. This makes them difficult to interpret sometimes. Just a heads-up...]
Regarding the mechanism of flagellar rotation, Nelson really should have read Kojima and Blair's 2001 paper before citing older papers in favor of Nelson/Mike Gene's favored model:
quote:
With respect to electrostatic interaction, I don't think we can say that the channel is internal to ExbBD. TonB might function in opening the channel or causing some kind of structural change. ExbB has only 3 membrane segments as opposed to MotA which has four.
The electrostatic model of the bacterial flagellar motor seems to be a bit more then a hunch, in fact, that electrostatic model behind flagella rotation seems to be near certainty that it is correct:
Bren, Anat and Michael Eisenbach. "How Signals Are Heard during Bacterial Chemotaxis: Protein-Protein Interactions in Sensory Signal Propagation." Journal of Bacteriology. 182:6865-6873 (2000).
I think even Blair, the author of the paper you cite for the ExbBD complex has done some work on this:
Zhou, J.D., Lloyd, S.A., and Blair. D.F. (1998) Electrostatic interactions between rotor and stator in the bacterial flagellar motor. Proc. Natl. Acad. Sci. USA 95, 6436-6441
quote:
--------------------------------------------------
Mutational studies of the rotor protein FliG and the stator protein MotA showed that both proteins contain charged residues essential for motor rotation. This suggests that functionally important electrostatic interactions might occur between the rotor and stator.
--------------------------------------------------
So it may not even matter that MotAB is homologous to ExbBD, if the electrostatic interactions are completely essential. Thus, I think Mike's objection should be taken seriously here with respect to the logistical problems with random mutations and ion channels.
LOL! Virtually everything you argue ("near certainty" that the electrostatic model is correct, the channel is not internal to MotAB) is contradicted by the very Kojima and Blair article I've been citing.
Kojima and Blair write that based on the results of their 2001 paper, they think that the proton channel is internal and that the MotAB conformational-change model is probably correct:
quote:
In the years since the discovery of flagellar rotation, many hypotheses for the mechanism have been proposed (reviewed in ref 43). The models are diverse, but can be classified according to whether the proton pathway includes elements of both the rotor and stator or is confined to just the stator (Figure 1). Because the mutational studies found no critical titratable residues on the rotor, we currently favor models in which protons remain within the stator. In this case, proton flow must be coupled to rotation by some means other than direct proton-rotor contact. Our hypothesis is that protonation of Asp32 in MotB drives conformational changes in the stator, which work on the rotor to drive rotation.
[...]
Here, we test for conformational changes in the MotA/MotB complex by using limited proteolysis. Patterns of proteolysis of MotA were compared in wild-type MotA/MotB complexes and complexes with mutations in key residues of one or both of the proteins. The results support a mechanism in which the stator undergoes significant changes in conformation. [bold added]
Indeed, the separability of the motor as an independent subunit appears to be what led them to look this is what led them to look for independently-functioning homologs in simpler contexts:
quote:
The occurrence of significant conformational change in the stator has implications not only for the present-day mechanism but also for the evolution of the flagellar motor. A membrane complex that undergoes proton-driven conformational changes could perform useful work in contexts other than (and simpler than) the flagellar motor, and ancestral forms of the MotA/MotB complex might have arisen independently of any part of the rotor. We queried the sequence database using the sequence of the best-conserved part of MotA (the segment containing membrane segments 3 and 4) from Aquifex aeolicus, a species whose lineage is deeply branched from other bacteria. In addition to the expected MotA homologues, the search returned a protein sequence from the archaeal species Methanobacterium thermoautotrophicum (protein MTH1022) that shows significant sequence similarity not only to MotA but also to the protein ExbB (Figure 9).
As for mutational flexibility:
quote:
Residues of MotA that are important for function include two charged residues in the cytoplasmic domain, Arg90 and Glu98, that interact with the functionally important charged residues of FliG (39, 40). Like the charged residues of FliG, Arg90 and Glu98 of MotA function redundantly, and charge appears to be their key property. Two Pro residues of MotA located at the cytoplasmic ends of membrane segments, Pro173 and Pro222, are also important for rotation and might function to regulate conformational changes occurring during the torque-generating cycle (41). In MotB, an aspartic acid residue near the cytoplasmic end of the membrane segment, Asp32, is conserved and critical for motor rotation. A survey of conserved residues in MotA, MotB, FliG, FliM, and FliN found that no other titratable residue is critical for motor rotation (42). Asp32 is likely to have a direct role in the conduction of protons.
It appears that, basically, you need the Asp32 to transfer the proton through the channel internal to MotA/B, and then a significant variety of charged residues on the C-ring (FliG/M/N) and MotA can transfer the conformational change in MotA to "push on" the C-ring. This seems rather less daunting than Mike Gene's characterization which you quoted:
quote:
Of all the ways to mutate an ion channel, the number of ways that would result in its interacting with the base of some filament is surely in the distinct minority. And of all the ways to mutate an ion channel that gloms onto a filament, the number of ways to mutate it such that rotation does not occur is probably much higher than the number of ways to elicit some rotation...This [mutation] allows some ion channel to glom onto the base of a filament and open its channel and expose the ion flow to the proto-rotor in such a way that a set of electrostatic interactions just happen to form and elicit significant rotation. Suffice it to say that such an improbable mutation has never been observed in nature or the lab.
This is not the only place where Kojima and Blair tripped up Mike Gene, indeed, he was originally skeptical of the existence of nonflagellar MotAB homologs:
quote:
MotA/MotB, on the other hand, could plausibly exist as some ion channel prior to the existence of the flagella, but there is no evidence of this.
I'll address the other points when I have time. Like I said, why should we take your probability calculation seriously, when it is based on incorrect assumptions about the mechanism of rotation, the independence of the MotAB complex homologs, etc.? Why doesn't your calculation include the possibility of cooption of MotAB from a pre-existing, simpler system? Why does your calculation assume that proteins are drawn at random from a mythical "protein supermarket" when we have reasonable hypotheses for the prior functions of many of these proteins, and we have a large body of empirical evidence for the efficacy of natural processes to produce new genes by cooption and modification of old genes?
[edit to fix double-quote]
[ 06. May 2003, 17:04: Message edited by: yersinia ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 06. May 2003 17:50
Nic,
None of your quotes, not even the very existence of the ExbBD complex, contradicts the electrostatic model of the bacterial flagellum, or anything I said. None of your quotes necessarily implies that the channel is internal. It doesn't even say it. As I said, TonB might function in opening the channel or it might be part of the channel itself. It doesn't have to be internal. As far as mutational redundancy, all that the quote is showing is a couple of redundant residues, but there is no evidence that the motor complex would likely be "flexible" in terms of mutation for precisely what Mike Gene stated, especially taking into account that the model for the ExbBD complex is consistent with the model for the bacterial flagellum. The proton flow (protonation of Asp32 in MotB) is also proposed to cause a conformation change in the cytoplasmic loop of MotA. This loop, which contains the charged residues that interact with oppositely charged residues in FliG, is somehow making and breaking contacts with the rotor as a consequence of these conformational changes. These interactions, in turn, apply force to the rotor that manifests itself as motor torque. It seems to me that the models for the function of the TonB/ExbBD and Tol complexes are completely consistent with the electrostatic model of the bacterial motor, once again vindicating what Mike Gene said about them.
What exactly, in the calculation, is incorrect about MotAB, what does it have to do with a complex that is quite consistent with not only the ICness of the motor complex, but with it's model for rotation, to the fact that if you remove fliG you won't get a flagellum, much less torque? Even if we replace it with the motor complex of the flagellum, you would still have the same imporobability and unselectable steps. It doesn't even effect the electrostatic model of the motor. What is the relevance of these quotations? There is no evidence that the ExbBD system predates the flagellum. And if you assume that it has then you definitely pointing to a protein supermarket where the proteins are doing other things. Why do you call it mythical?
I see no reason to reject convergence for the ExbBD complex.
[ 06. May 2003, 17:54: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 06. May 2003 20:30
quote:
None of your quotes, not even the very existence of the ExbBD complex, contradicts the electrostatic model of the bacterial flagellum, or anything I said. None of your quotes necessarily implies that the channel is internal. It doesn't even say it. As I said, TonB might function in opening the channel or it might be part of the channel itself. It doesn't have to be internal. As far as mutational redundancy, all that the quote is showing is a couple of redundant residues, but there is no evidence that the motor complex would likely be "flexible" in terms of mutation for precisely what Mike Gene stated, especially taking into account that the model for the ExbBD complex is consistent with the model for the bacterial flagellum. The proton flow (protonation of Asp32 in MotB) is also proposed to cause a conformation change in the cytoplasmic loop of MotA. This loop, which contains the charged residues that interact with oppositely charged residues in FliG, is somehow making and breaking contacts with the rotor as a consequence of these conformational changes. These interactions, in turn, apply force to the rotor that manifests itself as motor torque. It seems to me that the models for the function of the TonB/ExbBD and Tol complexes are completely consistent with the electrostatic model of the bacterial motor, once again vindicating what Mike Gene said about them.
Nelson, don't you understand that the electrostatic model postulates that the proton passes between the rotor (C-ring) and stator (MotA/B), and that this moving positive charge (acting against, say, diagonal rows of positive charges; MG has a nice picture in his essay even) rotates the C-ring? Whereas the other basic model postulates that the proton passes just through the stator and conformational changes convey force to the rotor? And that Kojima and Blair favor the latter for several reasons which they cite?
This is clear as day from this quote which I previously posted:
quote:
In the years since the discovery of flagellar rotation, many hypotheses for the mechanism have been proposed (reviewed in ref 43). The models are diverse, but can be classified according to whether the proton pathway includes elements of both the rotor and stator or is confined to just the stator (Figure 1). Because the mutational studies found no critical titratable residues on the rotor, we currently favor models in which protons remain within the stator. In this case, proton flow must be coupled to rotation by some means other than direct proton-rotor contact. Our hypothesis is that protonation of Asp32 in MotB drives conformational changes in the stator, which work on the rotor to drive rotation.
And yet you are clinging to your Mike Gene quote, which depends upon the electrostatic model which you also cling to, because your argument against the cooption of a simple pre-existing system to provide the motor of the flagellum depends upon it.
I have already reviewed the statistically significant sequence similarity (already established as homology in peer-reviewed journal articles, what more do you want?) between Mot and Exb, archived here:
link to ARN thread
...and showed that ExbB is as widely or more widely distributed than MotA.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 07. May 2003 06:13
Nelson writes,
quote:
With respect to adhesion organelles, I dont have time to tread though an entire thread to find an example of a flagellum acting as an adhesion organelle, but something tells me that you are just talking about a pilus, which I already discussed. My reason for thinking this is that since this was brought up in the immunity thread, PapD of the pilus has Ig-like domains. It is misleading to call pilus "bacterial flagellum".
I'm not sure what you're talking about with a pilus, although a great many (most?) pili of various kinds do have adherence as their function or one of their functions.
My point is that it appears many flagella also have an adhesion function. The paper I was referring to from the immune system thread was the first paper I cited on page 1, sorry that wasn't clear:
Giron JA, Torres AG, Freer E, Kaper JB. The flagella of enteropathogenic Escherichia coli mediate adherence to epithelial cells. Mol Microbiol 2002 Apr;44(2):361-79
There are others, e.g.:
Gavin R, Rabaan AA, Merino S, Tomas JM, Gryllos I, Shaw JG. Lateral flagella of Aeromonas species are essential for epithelial cell adherence and biofilm formation. Mol Microbiol. 2002 Jan;43(2):383-97
One more:
de Oliveira-Garcia D, Dall'Agnol M, Rosales M, Azzuz ACGS, Martinez MB, Girón JA (2002).Characterization of Flagella Produced by Clinical Strains of Stenotrophomonas maltophilia. Emerging Infectious Diseases [serial online]
quote:
Although adherence to abiotic surfaces is a property of both environmental and clinical S. maltophilia isolates, little information has been available to elucidate the nature of the surface factors involved in this phenomenon. Flagella have been associated with biofilm formation in other bacteria (18,20–22), where they can perform three basic roles: a) act as an adhesin promoting intimate attachment to the surface; b) generate force to subjugate the repulsive forces between bacteria and surface; and c) promote spread of the bacteria throughout the surface (20).
I bring it up because the question is whether or not intermediate stages had selectable function. It appears that an adhesive pili is one example of such a selectable intermediate stage, even though it wouldn't need motors or motility function.
Based on the above quote, we might even put forward a hypothesis: a certain species of bacteria living in a pre-flagella world makes its living by adhering to particles (bits of silt or whatever). When it reproduces, brownian motion spreads it a small distance in a random direction before it grows its primitive Type III pili and adheres to another silt particle. Occasional disturbances (such as waves or tides or new silt layers) wipe out most of the population at regular intervals, creating a classic early-successional ecological system (think weeds in a continually disturbed field).
Amongst the typical huge number of detrimental mutations that go nowhere in this population, a very small proportion of the mutations cause a member of the population to have a mutant ExbB homolog that attaches to the base of the pili rather than whatever it was originally attached to. This serves only to give random, undirected motion. However, experiments even with bacteria with partially-disabled flagella or flagella stuck on "tumble" show that they have diffusion coefficients orders of magnitude higher than dead or otherwise immotile bacteria (which still have diffusion coefficients large relative to body size). So, even undirected motility will serve to increase diffusion-distance-before-sticking, allowing that genotype to spread further and faster than the rest of the population (mechanisms to increase random diffusion of offspring are ubiquitous in life BTW, think dandelion seeds and go from there). Thus, this mutation comes to dominate and we have our very primitive proto-flagellum. Only gradually does all the fancy stuff get added -- finer co-adaptation of rotor and motor, chemotaxis sensors, switching, hook/flagellin distinction, etc.
Hmm, don't think that Dembski took anything like this into account in his calculation.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 07. May 2003 15:51
Nic,
With respect to the electrostatic model, you didn't read my reply nor did you respond to it. However, before I get back to that, let me make my main point in this reply.
The improbability of mutations building this system does not "depend" on the electrostatic model. Any co-option event where a mutation has to build anything that must be precisely built is improbable regardless of the mechanism of how that works (unless it is a wildly simple mechanism). My foray into Blair's discussion and the electrostatic model above is not necessarily to rescue the electrostatic model, because as you can see, each one has many things in common. As Mike Gene says:
quote:
What all models share in common is the theme of specificity, whereby specific interactions between the stator and rotor are required to elicit rotation.
Notice the improbablity of what a mutation would do building this primitive motor:
quote:
The torque generation function associated with this protein is restricted its C-terminal domain. Random mutations were introduced into the gene coding for this protein and yielded a set of mutants with flagella that did not rotate. In fact, even if the mutant protein was overexpressed, motility was not restored. All but one of these mutants involved the loss of a hydrophobic amino acid and were found to make the proteins subject to degradation. [7] This likely means that these mutations altered the conformation of FliG such that its charged residues directly involved in torque generation were no longer properly positioned. The mutations involved residues at positions 234, 237, 249, 252, 257, and 306. Remember that a mutation in any one of these sites resulted in complete loss of motility. In addition, small deletions also had the same effect: deletions of residues 280-285 and 292-295. With this in mind, let us return to the EFM hypothesis.
Now, my last post in response to you showed that I don't have to rewrite Mike's concern because the model that Dr. Blair is proposing is consistent with the problems associated with it, in my opinion. The probability of your co-option event however, does not depend on the truth of any one of the three models because they all require a high degree of specificity.
In Blair's model, the proton doesn't bind to any site on the rotor, but binds to a site in the stator and affects its conformation, that is clear from the quote. However, that conformational change affects parts of the stator that contact the rotor. Now lets compare this to what the model looks like for the bacterial motor. The proton binding site for both is Asp32, this site is also proposed to cause a conformation change in the cytoplasmic loop of MotA in the electrostatic model. Either way, there must be access to the proton through the ion channel. For both, it cannot be internal because it must be exposed to solvent to allow proton access and a clear avenue. It is the loop, which contains the charged residues that interact with oppositely charged residues in FliG, that is in contact with the rotor as a consequence of these conformational changes. These interactions, in turn, apply force to the rotor that provide torque for the motor.
[ 07. May 2003, 16:07: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 07. May 2003 16:16
Nelson says,
quote:
Either way, the rotor must have access to the proton through the ion channel.
But Kojima and Blair 2001 says, in a chunk previously quoted by me,
quote:
Because the mutational studies found no critical titratable residues on the rotor, we currently favor models in which protons remain within the stator. In this case, proton flow must be coupled to rotation by some means other than direct proton-rotor contact. (bold added)
Reposting quotes to repeatedly correct the same mistake that you repeat over and over is becoming tiresome, Nelson.
nic
PS: Also note that "the mutational studies found no critical titratable residues on the rotor" implies that the amount of specificity required by the rotor is not massive. A "lock and key" fit ain't necessary, more just having some appropriate charges on the outside that the stator can act against.
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 07. May 2003 16:30
Nic,
I edited that portion before you posted that reply, it's not even relevant to my discussion of it or the problems related to it.
With respect to the mutational studies, as far as Asp32 is concerned, function requires Asp there. And as Mike has shown:
quote:
The torque generation function associated with this protein is restricted its C-terminal domain. Random mutations were introduced into the gene coding for this protein and yielded a set of mutants with flagella that did not rotate. In fact, even if the mutant protein was overexpressed, motility was not restored. All but one of these mutants involved the loss of a hydrophobic amino acid and were found to make the proteins subject to degradation. [7] This likely means that these mutations altered the conformation of FliG such that its charged residues directly involved in torque generation were no longer properly positioned. The mutations involved residues at positions 234, 237, 249, 252, 257, and 306. Remember that a mutation in any one of these sites resulted in complete loss of motility. In addition, small deletions also had the same effect: deletions of residues 280-285 and 292-295.
And it's not just having any change outside the stator, it requires conformational change that affects parts of the stator that contact the rotor.
[ 07. May 2003, 20:38: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 07. May 2003 16:54
Nic,
With respect to adhesion function, none of the papers show anything different from what I already discussed. In each paper, they talk about the export machine of bacterial flagellum, or pilus. In both cases, we still have the "sea of nonfunctionality" that Dembski's calculation refers to. This is true, even if there are alternative functions all the way to the flagellum, each alternative function obviously requires several parts to function, so you doing nothing more then what H. Allen Orr objected to:
quote:
Second, we might think that some of the parts of an irreducibly complex system evolved step by step for some other purpose and were then recruited wholesale to a new function. But this is also unlikely. You may as well hope that half your car's transmission will suddenly help out in the airbag department. Such things might happen very, very rarely, but they surely do not offer a general solution to irreducible complexity.
For example, you have the problem of the 6 part export machine. There is no evidence that any subset of this export machine carries out alternative function. And as I stated with the pili, With that you continue to introduce more unselectable steps, the irreducible complexity of the folding of P Pilus, not to mention the sophisticated mechanisms, donor strand exchange and donor strand complementation. The pilus itself is made up of 5 parts, PapK PapA,PapE,PapK, and PapG. Furthermore, the pilus doesn't seem to be able to secrete proteins, and the biggest difference between flagella and pili is that flagella are built from the top to the bottom, whereas pili are built from the bottom to the top.
Now you keep saying things like "I wonder why Dembski didn't take that into account". Now, can you show, in the peer reviewed literature, a paper that shows how natural selection and random mutation, taking all those imaginary homologs and multi-part machines into account with alternative functions, a detailed pathway leading up to the flagellum that Dembski could have worked with. What do we find in the peer reviewed literature?
We find the evidence that type III systems evolved from the flagellum, and there are good reasons for this. Even taking all these similarities into account, it is quite easy to see how the "sea of nonfunctinality" between each mutli-part machine (especially when proposing a sticky-out thing when in fact it would be useless in a "brownian storm", again the question of minimal function where the actual components don't even matter, it's what those components can do), most likely none of these alternative machines will confer a selective advantage until we arrive at the fully functional flagellum. Pure chance events come in full force when you attempt to fortitously have multi-part machines interacting with multi-part machines. It's nothing but a tornado in a junk yard.
[ 07. May 2003, 16:57: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 08. May 2003 02:20
Nelson writes,
quote:
Nic,
I edited that portion before you posted that reply, it's not even relevant to my discussion of it or the problems related to it.
Thanks, I will leave it up for context. Let's be clear, then, that the objections of Mike Gene which you cited and previously confidently defended ("near certainty"), based on the idea that the proton channel was between the rotor and stator and that therefore both parts would have to be instantly co-adapted in order to function, are now moot. E.g. here is Mike Gene quote:
quote:
The Fortuitous Interaction
The EFM hypothesis envisions some ill-defined fortuitous interaction between some unknown ion channel and the basal body of the non-motile filament. Then somehow, motility spontaneously emerges and selection takes over from here. Once again, pure chance must bring about the new function. But how likely is this?
First, I'm having a hard time envisioning this. The rotor must have access to the proton/ion through the ion channel. What type of fortuitous change is going to pry open this ion channel and then glom it onto the proto-rotor? It would seem that of all the ways to mutate an ion channel, such a change would represent only a very small minority of all possible changes.
In fact, this very same theme repeats itself with all necessary parts of this fortuitous interaction. Of all the ways to mutate an ion channel, the number of ways that would result in its interacting with the base of some filament is surely in the distinct minority. And of all the ways to mutate an ion channel that gloms onto a filament, the number of ways to mutate it such that rotation does not occur is probably much higher than the number of ways to elicit some rotation.
Thus, as with the first cooption event, we need another special mutation. This one allows some ion channel to glom onto the base of a filament and open its channel and expose the ion flow to the proto-rotor in such a way that a set of electrostatic interactions just happen to form and elicit significant rotation. Suffice it to say that such an improbable mutation has never been observed in nature or the lab.
But you have now apparently conceeded that Kojima and Blair (2001) makes the "ion channel [must] glom onto the base of a filament and open its channel and expose the ion flow to the proto-rotor" and objection an unrealistic one.
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 08. May 2003 05:55
quote:
Nic,
With respect to adhesion function, none of the papers show anything different from what I already discussed. In each paper, they talk about the export machine of bacterial flagellum, or pilus.
I don't understand what you are saying here at all. Here is the point: non-motile cell extensions have function even though they are non-motile. Some flagella still perform this function in addition to having their motility function. This function remains even if they have no motors.
quote:
In both cases, we still have the "sea of nonfunctionality" that Dembski's calculation refers to. This is true, even if there are alternative functions all the way to the flagellum
This must set another record for self-contradiction! If there are alternative functions "all the way to the flagellum" then the "sea of nonfunctionality" either doesn't exist or is crossed by a land bridge.
quote:
, each alternative function obviously requires several parts to function
So? They function independently from being flagella, that is the point. Dembski's calculation, like basically all antievolutionary calculations, assume that everything had to come together at once, in one step. This is how ridiculously low probabilities are reached in calculations.
quote:
so you doing nothing more then what H. Allen Orr objected to:
This is the one quote that you have saying that cooption is unlikely, and it's not even clear what is meant as he's never expanded on these words, discussed the literature on cooption in detail, etc. Orr himself refers to the change-of-function between swim bladders and lungs in that very review, which sounds to me rather like "half your car's transmission will suddenly help out in the airbag department." And even if Orr meant exactly what you think he meant, I can cite dozens of other scientists of equal authority who think that cooption is indeed common, because they have written numerous articles and amassed a lot of evidence in favor of it. For example:
The Origin of "Information" via natural causes
[url=]The Origin of "Information" via natural causes -- citations in the literature[/url]
Co-option/change of function -- Citations of this in the literature
Origins of morphological novelty
quote:
For example, you have the problem of the 6 part export machine. There is no evidence that any subset of this export machine carries out alternative function. And as I stated with the pili, With that you continue to introduce more unselectable steps, the irreducible complexity of the folding of P Pilus, not to mention the sophisticated mechanisms, donor strand exchange and donor strand complementation. The pilus itself is made up of 5 parts, PapK PapA,PapE,PapK, and PapG.
There are lots of kinds of pili based on lots of kinds of transport systems, you keep bringing up one specialized system (based on a Type I transporter) over and over like it was some kind of talisman. Type III virulence systems (called Type III pili in some articles, like this the Cornelius and Van Gijsegem paper mentioned earlier).
It's not very clear, but by repeated mention of the P pilus I assume that you are trying to argue that a pili extension, excluding the export system at the base, would *have* to be made of at least 5 parts in order to function. But there is no "logic of pili" that requires this -- while a rotary motor must, logically, have at least a stator, rotor, and filament, a simple pili could really be made of one repeated subunit. Figure 1B of Blocker et al. 2003 (Pubmed) depicts a generalized T3SS where the extracellular extension (pili) is made up of MxiH/PrgI (homologs in 2 different organisms) for the entire stalk and "CaoX?" at the tip, perhaps a cap and/or adhesion protein at the tip (flagellar cap proteins are known to have adhesive functions in some bacteria, if you were interested). The article says that it is not even clear that T3SS require a cap protein as none has yet been detected. The online supporing text says that only some T3SS bother to add an additional filament to the top of the needle.
quote:
Furthermore, the pilus doesn't seem to be able to secrete proteins, and the biggest difference between flagella and pili is that flagella are built from the top to the bottom, whereas pili are built from the bottom to the top.
But of course Type III pili secrete just fine out the top. And archaeal flagella are built from the bottom like some pili. So there is no problem for pili to secret or be built at the top and no problem for flagella to be built from the base. I recall pointing this out before in this very thread, but you bring up the argument again as if you didn't have to address the counterargument.
quote:
Now you keep saying things like "I wonder why Dembski didn't take that into account". Now, can you show, in the peer reviewed literature, a paper that shows how natural selection and random mutation, taking all those imaginary homologs and multi-part machines into account with alternative functions, a detailed pathway leading up to the flagellum that Dembski could have worked with. What do we find in the peer reviewed literature?
Dembski hasn't even dealt with the implications of the "imaginary" homologs published in Kojima and Blair, with direct commentary on evolutionary implications, in Biochemistry in 2001. But Dembski, and you, are arguing that his calculation "sweeps the field clear" of natural hypotheses and that therefore design is the logical conclusion. This is far more ambitious than simply pointing out that the literature does not yet contain the evolutionary pathway for every molecular system whose mere structure and function have yet to be fully understood.
Even where there is extensive peer-reviewed literature on the evolution of an IC system, e.g. the entire field of evolutionary immunology, neither you nor Dembski has shown any respect for it, despite your respective wild assertions about there being no literature on the evolution of IC systems.
And, no one ever published Dembski's straw-man "the flagellum assembled all at once" model, and yet you and Dembski are treating it as if it's the only conceivable route for natural processes to take.
quote:
We find the evidence that type III systems evolved from the flagellum, and there are good reasons for this. Even taking all these similarities into account, it is quite easy to see how the "sea of nonfunctinality" between each mutli-part machine (especially when proposing a sticky-out thing when in fact it would be useless in a "brownian storm", again the question of minimal function where the actual components don't even matter, it's what those components can do),
Well, I already told you about the diffusion coefficient difference between dead bacteria and bacteria stuck on "tumble" with completely undirected motility, but I guess you won't believe me unless I give you the numbers. From Berg, 1993, Random Walks in Biology, p. 93:
Diffusion coefficient of a bacterium with flagellum stuck on "run" (rotation is only due to Brownian motion): 2 x 10^-5 cm2/sec
Diffusion coefficient of a normal E.coli bacterium with flagellum doing "run" and "tumble" at normal proportions: 4 x 10^-6 cm2/sec (however, the normal bacterium can bias runs in the right direction to be longer, leading to directional drift, i.e. chemotaxis).
Diffusion coefficient of a bacterium with flagellum stuck on "tumble" is described as "smaller" and that of a dead or paralyzed bacterium is described as "much smaller". The number calculated for the dead/paralyzed bacterium is 2 x 10^-9 cm2/sec, so a "stuck on tumble" bacterium is above this but below the number for the non-chemotactic "stuck on run" bacterium.
Even undirected motility, even in a non-smooth-run fashion, appears to improved dispersal capability quite a bit.
quote:
[...] most likely none of these alternative machines will confer a selective advantage until we arrive at the fully functional flagellum. Pure chance events come in full force when you attempt to fortitously have multi-part machines interacting with multi-part machines. It's nothing but a tornado in a junk yard.
No, Dembski's calculation is like assembling a plane by putting a tornado in a junkyard. But no one advocates that model. What evolution proposes is (vaguely) more like a "tornado" coming through a junkyard and making various modifications to parts, sometimes linking two systems in some way. True, most of these go nowhere, but the occasional pairing that is successful can be selected and therefore will persist for another bunch of rounds of modifications and selection, which may eventually successfully pair one double system with another system of one or more parts, etc. The analogy still sucks but this gives some idea of how many important factors Dembski's random-assembly-all-at-once calculation leaves out.
[ 08. May 2003, 05:56: Message edited by: yersinia ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. May 2003 14:22
Nic writes:
quote:
Thanks, I will leave it up for context. Let's be clear, then, that the objections of Mike Gene which you cited and previously confidently defended ("near certainty"), based on the idea that the proton channel was between the rotor and stator and that therefore both parts would have to be instantly co-adapted in order to function, are now moot. E.g. here is Mike Gene quote:
This is simply false. It seems as though you are simply ignoring my post and repeating yourself. Obviously in the quote cited, Mike was assuming the electrostatic model. But as I already showed (and Nic doesn't respond to) the problem is still quite visible even with Blair's model. Here is the important thing to remember, is that I think the electrostatic model for the bacterial flagellar motor is almost near certainty that it is correct. Still the issue is this with any of the three models:
quote:
What all models share in common is the theme of specificity, whereby specific interactions between the stator and rotor are required to elicit rotation.
http://www.idthink.net
With all three models there is still a rotor-stator interface, and this is obvious from the mutational studies that Mike Gene cites.
Nic still hasn't responded to this.
Nic also writes:
quote:
But you have now apparently conceeded that Kojima and Blair (2001) makes the "ion channel [must] glom onto the base of a filament and open its channel and expose the ion flow to the proto-rotor" and objection an unrealistic one.
Unfortunately I conceded no such thing. The complaint is extremely realistic, given that the electrostatic model is still on the table, just because you disagree with it, doesn't mean it disappears. Also, it's realistic given that it's consistent with Dr. Blair's model, even if Asp is not on the rotor. An ion channel still has to glom onto the base of a filament and open its channel since the proton-binding site, which is now moved to the stator, has to be exposed to solvent to allow proton access and egress. A "lock and key" fit is still extremely essential between the stator and the rotor, since once the proton binds to the site on the stator, it must introduce conformational changes on those parts of the stator that contact the rotor. This is not a resilient contact either, it's pretty stringent. Once again here is Mike:
quote:
What all models share in common is the theme of specificity, whereby specific interactions between the stator and rotor are required to elicit rotation.
I also missed responding to this:
Nic writes:
quote:
I have already reviewed the statistically significant sequence similarity (already established as homology in peer-reviewed journal articles, what more do you want?) between Mot and Exb, archived here:
link to ARN thread
...and showed that ExbB is as widely or more widely distributed than MotA.
I'm sorry but this is simply false. How can ExbB be as widely or more widely distributed than MotA? Here is what is wrong with your excercise, true ExbB proteins are only involved in driving outer-membrane transport . So they should only be found in gram-negative organisms. Sure there are lots of those. But flagella are found in gram negatives and gram positives . If you "tree out" the genes (ask where authentic motA genes and authentic exbB genes are found, on a phylogenetic tree of all bacterial lineages) you'll find MotA genes on more branches.
Note, I'll reply to Nic's second post in a bit (later or tommorrow), have some things to attend to at the moment.
[ 08. May 2003, 14:55: Message edited by: Nelson_Alonso ]
IP: Logged
|
|
yersinia
Member
Member # 324
|
posted 08. May 2003 15:44
quote:
I'm sorry but this is simply false. How can ExbB be as widely or more widely distributed than MotA? Here is what is wrong with your excercise, true ExbB proteins are only involved in driving outer-membrane transport . So they should only be found in gram-negative organisms. Sure there are lots of those. But flagella are found in gram negatives and gram positives . If you "tree out" the genes (ask where authentic motA genes and authentic exbB genes are found, on a phylogenetic tree of all bacterial lineages) you'll find MotA genes on more branches.
Who says that gram-positives are the most basal lineage? This is far from clear. All we can go on is the current taxonomy. When the MotA-ExbB homolog family is plotted on this taxonomy, as is done in this handy "Species distribution for MotA_ExbB" tree generated by Pfam, and the various links to the hits are followed, we get MotA_ExbB homologs, identified as things other than MotA, in these fundamental lineages:
Bacteria
- Fusobacteria
- Cyanobacteria (which lack flagella entirely AFAIK)
- Deinococcus-Thermus
- Aquificae
- Chlamydiae
- Proteobacteria
- Chlorobi
...each of which has equal rank to the group Firmicutes, aka gram-positive bacteria.
And furthermore, there are several hits in Archaea, which lack bacterial flagella entirely, and only have one membrane, and are an outgroup to the whole of Bacteria in the taxonomy. Obviously those proteins are doing something -- there is no reason to assume that all ExbB homologs have to do something at the outer membrane.
In other words, when I said,
quote:
[my link to the ARN thread] showed that ExbB is as widely or more widely distributed than MotA
...I was simply conveying the current data to you, although to be precise I should have said "nonflagellar MotA_ExbB homologs" instead of just "ExbB".
Swing and a miss, once again...
[ 08. May 2003, 15:47: Message edited by: yersinia ]
IP: Logged
|
|
Nel
Member
Member # 614
|
posted 08. May 2003 18:29
Unfortunately, just enough time for a couple of comments right now but a few comments seems to be enough. Again fuller replies will have to come later.
Can you show me an ExbB protein present in gram-positive bacteria? Of course if you change what you said from ExbB being as widely distributed as MotA to nonflagellar imaginary homologs of ExbB/MotA then you're talking about two different things. It seems to be irrelevant to what I said about authentic MotA and ExbB proteins. If your entire argument is based on the function of ExbB/D, which is restricted to the gram-negatives, your argument seems to get quite ambiguous. Even if gram-positives weren't basal, MotA is still far more widely distributed than ExbB, since it is also found in gram-positives (although there is good evidence that gram-positives are basal).
Furthermore, I have more to say about that link, however, for starters, the entire tree is based on a hypothetical protein (much like your Spa47 blunder).
quote:
Hypothetical protein MTH671.
The most ironic thing about your post is that you say "swing and a miss once again", but I'm still waiting for the first miss.
IP: Logged
|
|
|
Printer-friendly view of this topic