Flagellum evolution in Nature Reviews Microbiology

From the table, I count 15 proteins with no known homologies, 2 of which are indispensable. What's with that?

Great job, Nick. I wonder whether the "30 unique proteins" canard is going to find its way in the new edition of "Of Pandas and People", that Dembski & C are supposed to be putting out shortly. That would be sweet.

I've just had a quick scan of the paper. Looks like being a good read. In the meantime (and I thought this was as good a place as any to put it) the following paper in today's Nature is well worth a glance:

Keightley, P.D. and Otto, S.P. (2006) Interference among deleterious mutations favours sex and recombination in finite populations. Nature, 443, 89-92.

Sex and recombination are widespread, but explaining these phenomena has been one of the most difficult problems in evolutionary biology. Recombination is advantageous when different individuals in a population carry different advantageous alleles. By bringing together advantageous alleles onto the same chromosome, recombination speeds up the process of adaptation and opposes the fixation of harmful mutations by means of Muller's ratchet. Nevertheless, adaptive substitutions favour sex and recombination only if the rate of adaptive mutation is high, and Muller's ratchet operates only in small or asexual populations. Here, by tracking the fate of modifier alleles that alter the frequency of sex and recombination, we show that background selection against deleterious mutant alleles provides a stochastic advantage to sex and recombination that increases with population size. The advantage arises because, with low levels of recombination, selection at other loci severely reduces the effective population size and genetic variance in fitness at a focal locus (the Hill---Robertson effect), making a population less able to respond to selection and to rid itself of deleterious mutations. Sex and recombination reveal the hidden genetic variance in fitness by combining chromosomes of intermediate fitness to create chromosomes that are relatively free of (or are loaded with) deleterious mutations. This increase in genetic variance within finite populations improves the response to selection and generates a substantial advantage to sex and recombination that is fairly insensitive to the form of epistatic interactions between deleterious alleles. The mechanism supported by our results offers a robust and broadly applicable explanation for the evolutionary advantage of recombination and can explain the spread of costly sex.

One thought (and it may be a nitpick) about the abstract. I'm not entirely convinced by this sentence:

"Here we explore the arguments in favour of viewing bacterial flagella as evolved, rather than designed, entities"

It seems to me that this is a little generous towards the ID crowd, implicitly crediting them with having an actual working scientific hypothesis for the origin of the bacterial flagella.

Am I just being pedantic?

Am I just being pedantic?

— SteveF

Oh, darn it. I just looked it up and it means both. Now I look like an idiot.

From one pedant to another, belabor is prefectly acceptable in this context:

"to explain, worry about, or work at (something) repeatedly or more than is necessary: He kept belaboring the point long after we had agreed."

Nick,

Your table lists two proteins (FlgD and FliE) that are indispensible but have no known homologies, not one as you say below the table. This doesn't affect your main points, of course.

By irreducible complexity I mean a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning." --Michael Behe, Darwin's Black Box.

The flagellar motor is composed of five proteins: MotA, MotB, FliG, FliN, and FliM. We'll discuss this more below, but right now it is worth pointing out that the type III systems have no homologs for MotA, MotB, or FliM.

I agree with Corkscrew, if you count "None" (FliO) as equal to "None Yet Known" then you've got 15, 2 (FlgB, FliE) that are indispensable.

What stands out more to my eye is the obvious correlation of "Absent" and "None". Can anyone elucidate that?

Congratulations Nick, you worked long and hard on that paper.

Could it be that real scientists are doing the work of the ID crowd for them? Of course not, because we are actually, bit by bit, clawing away at the "we don't know at the moment = we can never know (and don't want to know)" hypothesis.
I wonder, though, to what extent ID is useful (god I hated writing that) in that it provides fruitful new areas of investigation? Might this kind of thing lead to real breakthroughs of the cure for X variety? What do you think Nick? Presumably as a scientist you aren't wasting your time proving them wrong. And where do you go from here?
Great work. But you're right, they won't listen.

Not being an expert in these matters, I don't know much you relaxed the defaults to arrive at some of the homologies. I'm sure a rebuttal of this data will observe that only using the defaults flips the ratios around completely.

Does the paper explain the extent and significance of the variation from the defaults and why using the defaults was inappropriate? If the same criteria were applied to the other proteins in the table what would be the results? Especially the homologies identified by paragraph marks - from the brief explanation it sounds like these homologies are identified with extremely subjective criteria.

With all the pounding on ID advocates to define their terms and stick to a consistent definition, I can't say I'm happy about a table that conflates four different meaning of the word homology in order to make its point. I'm very open to hearing from Nick or any other expert on using BLAST that this conflation is acceptable and standard procedure when counting homologies. As a non-expert however, that doesn't jump out at me as true.

Thanks for the comments -- making various edits. Apologies for counting errors, that's what I get for posting in the middle of the night.

Great job. I've cited your argument on flagellum evolution often. I'll be sure to update my links to the peer article.

Scott Minnich, the leading flagellum expert in the ID camp, was severely wrong about the most basic data relevant to the origins of the flagellum, the flagship system of the ID movement.

Hmm, I thought the issue discussed in the Keightley and Otto paper had been resolved a while back. Long-term effects like deterioration aren't enough to overcome the twofold reproductive advantage of asexual species - the sexual species would get overrun before the problems became apparent. However, parasitism and disease, adapting as they do to target individual organisms, are able to do substantially more damage to identical children than to children produced sexually.

That's the version I heard anyway, obviously I bow to the superior experience of anyone with actual scientific qualifications.

Regarding BLASTing on non-default settings: not really. Of the four categories, the first three are completely standard techniques and the homologies are well-accepted. Of the 28 homologies I counted just now (including a dual homology), only five are based on "structural/other" criteria, and even so several of those are well-accepted.

Homology by technique:

BLASTP defaults: 7
PSI-BLAST defaults: 7
PSI-BLAST slightly relaxed: 9
Structural/other: 5

The PT server appears to be overloaded at the moment, so I am having trouble logging in to do edits. For the record, the correct counts are 15 and 2 for "unique" and "unique + inessential."

PS: Yes, I am aware that several of the above ID articles contain various Emergency Backup Arguments that say something like "Even if all of the flagellum proteins had homologs, the flagellum still couldn't evolve because of X."

September 7, 2006 Flagellum Evolution Nick Matzke at Panda's Thumb, what evidence is there that the type III secretion system appeared in nature before the flagellum? If the flagellum coopted the ttss then the ttss must predate the flagellum. The ttss mediates elaborate interactions with plant and animal hosts of the bacteria. The flagellum on the other hand is for locomotion, not parasitic or pathogenic relationships with more complex cells. The flagellum is useful absent more complex organisms in the environment while the ttss is not. It seems to me quite likely that the flagellum appeared in nature before the ttss. Probably billions of years before as the following supports:

More than factual details of your study, what is important here is that you have shown what good science does - it takes an unknown and chips away at it looking for answers. Since IDers can't prove that something is irreducibly complex in principle, their claims rely on the fact that we don't know everything there is to know. ID depends on science not going forward.

Good job taking away one of their icons. Of course, they'll just find another dozen things we don't know everything about and claim they're irreducibly complex...

One thought (and it may be a nitpick) about the abstract. I'm not entirely convinced by this sentence: "Here we explore the arguments in favour of viewing bacterial flagella as evolved, rather than designed, entities" It seems to me that this is a little generous towards the ID crowd, implicitly crediting them with having an actual working scientific hypothesis for the origin of the bacterial flagella. Am I just being pedantic?

I wonder, though, to what extent ID is useful (god I hated writing that) in that it provides fruitful new areas of investigation? Might this kind of thing lead to real breakthroughs of the cure for X variety?

— AnthonyK

Didn't take them long, according to the trackback below. Apparently the X for today is "Maybe the proteins weren't around before the flagellum, huh, did you think of that?" September 7, 2006 Flagellum Evolution Nick Matzke at Panda's Thumb, what evidence is there that the type III secretion system appeared in nature before the flagellum? If the flagellum coopted the ttss then the ttss must predate the flagellum. The ttss mediates elaborate interactions with plant and animal hosts of the bacteria. The flagellum on the other hand is for locomotion, not parasitic or pathogenic relationships with more complex cells. The flagellum is useful absent more complex organisms in the environment while the ttss is not. It seems to me quite likely that the flagellum appeared in nature before the ttss. Probably billions of years before as the following supports:

At least nine core flagellar proteins share common ancestry with core components of NF T3SSs. However, the scientific community is divided on the nature of their common ancestor. Some have argued that the NF T3SS was derived from the flagellum, based on the eukaryote-specific function of known NF T3SS, and the apparently limited distribution of known NF T3SSs relative to the broad phylogenetic distribution of flagella [34]. We, and others, have argued that the two systems are sister groups, as indicated by gene phylogenies and arguments based on parsimony, diverging through 'descent with modification' from a common, but simpler, ancestral secretion system [35, 36]. Regardless of the conclusion of this debate, the existence of NF T3SSs is 'proof of concept' that a flagellar subsystem can function for purposes other than motility. [...] 34. Nguyen, L. , Paulsen, I. T. , Tchieu, J. , Hueck, C. J. & Saier, M. H. Jr. Phylogenetic analyses of the constituents of Type III protein secretion systems. J. Mol. Microbiol. Biotechnol. 2, 125---144 (2000). 35. Pallen, M. J. , Beatson, S. A. & Bailey, C. M. Bioinformatics, genomics and evolution of non-flagellar type-III secretion systems: a Darwinian perspective. FEMS Microbiol. Rev. 29, 201---229 (2005). 36. Gophna, U. , Ron, E. Z. & Graur, D. Bacterial type III secretion systems are ancient and evolved by multiple horizontal-transfer events. Gene 312, 151---163 (2003).

The SpringerBot is, in his own words, an "autodidact." If he wants to know what's in the paper, he'll tell YOU what's in the paper.

this is another example of how I.D. is actually anti-science.
While they are certainly imaginative in their questions (a very good thing!!) They are imaginative for the purpose of trying to find a question so difficult and so complex that they feel it can't be answered. To them, unanswered questions are the ultimate goal, since it creates a safe haven for their "supernatural" answers. Finding a good difficult question is the last step in their "research" and it closes the door for them. (ie "how could the flagellum have evolved step by step?)

Science on the other hand has another reason to be imaginative in finding difficult and hard questions. Science is more about finding the correct question than the correct answer, because in science questions aren't used to shut doors, but to open new doors. In this regard, it's good to have the I.Dists around because sometimes they can offer up some imaginative questions that might not have been asked otherwise. Someone already mentioned this, but it looks to me like a lot of flagellum research is being done these days and I think at least a small part of that was influenced by the ID throwing down the challenge. (of course that's just a guess)

They [IDers] are imaginative [only] for the purpose of trying to find a question so difficult and so complex that they feel it can't be answered. To them, unanswered questions are the ultimate goal, since it creates a safe haven for their "supernatural" answers. Finding a good difficult question is the last step in their "research" and it closes the door for them.

Great post Nick, and congrats on the publication.

I'm sure the IDists will complain that only one side of the story is being given "airtime" in peer-reviewed journals, and that somehow that isn't "fair". However, it's important to note that journals are not required to publish both sides of an argument when one side is based on faulty logic and science. Both sides must maintain a level of scientific rigor to be taken seriously by a journal. It's obvious that IDists argue by omitting important information or outright lying. They shouldn't whine that journals are unimpressed by it.

Another thing that's important to remember is that the bacterial flagellum was chosen as their biochemical mascot specifically because it's so friggin complex. It's not as though the flagellum is the cornerstone of all life on earth. I doubt it's even required for the bacteria that have it. If they really wanted to look for evidence of design, they would focus on those systems that were actually critical to all organisms (DNA replication, transcription, translation, to name a few). It shouldn't be surprising that the flagellum would give the most difficulty in explaining its origins, that's why they chose it.

If IC were really a roadblock for evolution, then it would apply equally to the most simple IC systems as well as the most complex. It's funny that reading Nick's post, I can see where other IC systems, like the complement system, have already bypassed the major hurdles that the flagellum research is still in the process of passing. For example, the IDists argue that the TTSS evolved from the flagellum and not vice versa. However, in the case of the complement system, we can see the steady accumulation of parts from the most basic complement system in sea urchin to the most complex in mammals. With regards to ICness, there's no qualitative difference between the complement system and the bacterial flagellum, only quantitative (i.e. number of parts). That the IDists focus so much attention on one system and ignore the others speaks volumes as to the vacuity of their ideas.

Great stuff-- I don't have access to Nature but am looking forward to your forthcoming posts here on the subject.

Something related to this subject which I find very interesting but don't know all that much about is the existence of variations in the flagellum. I'm told there are variations in structure, and your post here discusses variations in even composition (absent proteins), between the flagella of different organisms. I don't know the extent of these variations, but it seems to me the mere existence of variations belies the ID model of the flagellum as an ancient, divinely engineered gift likely to have been handed down by God himself; the flagellum apparently continues to mutate, just like any other normal part of a living thing.

It also makes me wonder what else we can conclude from those variations. You discuss here the question of the ancestry of the NF T3SS system (a funny question in itself; why would Intelligent Design lead us to conclude the flagellum came first and the NF T3SS came after? Why couldn't the designer have bestowed some bacteria with a secretion system, and then it "degraded" into an outbound motor? Is it just because the one is more "useful"?) ...but I'm somewhat curious whether anything is known (or can be detected) about the ancestry in whatever variations in the finished flagellum itself exist from organism to organism.

Do there exist any examples of an organism whose bacterial flagellum is "unusual" (mutated) in some way, and this variation is not just the result of random genetic drift but was actually clearly selected for for some reason?

Is there somewhere I could find out more about such things?

Another question, are the flagella found in bacteria the same as the flagella found in complex eukaryotes (for example, sperm cells)?

Thanks.

It also makes me wonder what else we can conclude from those variations. You discuss here the question of the ancestry of the NF T3SS system (a funny question in itself; why would Intelligent Design lead us to conclude the flagellum came first and the NF T3SS came after? Why couldn't the designer have bestowed some bacteria with a secretion system, and then it "degraded" into an outbound motor? Is it just because the one is more "useful"?) ...but I'm somewhat curious whether anything is known (or can be detected) about the ancestry in whatever variations in the finished flagellum itself exist from organism to organism.

Why couldn't the designer have bestowed some bacteria with a secretion system, and then it "degraded" into an outbound motor?

It seems we can use the flagellum as a scourge.

"At most, one can show that ID is unparsimonious and unnecessary"

I think one can show more in the particular cases, IC for example.

As TLTB says, one can't show IC in principle, due to scaffolding et cetera. In fact one can't in general show which structure or algorithm is the simplest to solve a specific task. If the global simplest solution is an illdefined concept, so is the local simplest (IC) solution.

So IC is illdefined. This forces IDers to specially plead to look for specific examples, and use an argument from ignorance to motivate a doubt.

In particular cases, probably all of them, it seems one can show that ID is unparsimonious, unnecessary, illdefined, nonuniversal, and arguing from ignorance.

The question about which came first is raised in the 2000 paper by Nguyen et al. that is quoted at the current PD thread. These authors base their conclusion on "phylogenetic clustering".

They think so, do they? Well, clearly that just means they're elitist, groupthinking academics, out-of-touch with the American public and conspiring to suppress the theory of Intelligent Secretions.

I wonder, though, to what extent ID is useful (god I hated writing that) in that it provides fruitful new areas of investigation?

Serious scientists don't even KNOW what the ID claims are. Any more than they know what the flat-earther claims are. Serious scientists simply don't pay any attention to the nutters.

Anyway, though my halfhearted joke (about which really comes first in ID, the flagellum or the secretion system) was not really intended to garner response, I am legitimately curious about what the record has to say about the phylogeny of whatever variations on the standard bacterial flagellum exist.

Nothing like a little flagellation to start the working day.

Onya Nick.

And at Uncommon Descent, Dave Scott challenges a forthcoming article by Nick Matzke et. al. on a supposed evolutionary pathway for the bacterial flagellum. Readers might also want to read Stephen Meyer and Scott Minnich's related paper on the "co-option" argument, "Genetic Analysis of Coordinate Flagellar and Type III Regulatory Circuits in Pathogenic Bacteria," as presented to the Second International Conference on Design & Nature, Rhodes Greece.

Is there somewhere I could find out more about such things?

Another question, are the flagella found in bacteria the same as the flagella found in complex eukaryotes (for example, sperm cells)?

"I doubt it's even required for the bacteria that have it."

In "Unlocking Life's Mysteries" they show bacteria lounging around on the bottom of pond since they lack a flagellum.

I pointed out that this was pretty sneaky false propaganda to the ID'ist showing the video since anyone who has ever seen movies of bacteria know that they are pretty much neutrally buoyant. Brownian motion is enough to get them around pretty good.

They [bacterial vs eukaryote flagellae] are very different. They both stick out of the cell and wiggle in some fashion, that is about the only similarity. Kind of like bird wings and insect wings. See Table 3 of the Big Flagellum Essay for a summary of the differences.

Nick Matzke at Panda's Thumb, what evidence is there that the type III secretion system appeared in nature before the flagellum?

Bacterial type III secretion systems are ancient and evolved by multiple horizontal-transfer events, U. Gophna et al. / Gene 312 (2003) 151---163 Abstract: Type III secretion systems (TTSS) are unique bacterial mechanisms that mediate elaborate interactions with their hosts. The fact that several of the TTSS proteins are closely related to flagellar export proteins has led to the suggestion that TTSS had evolved from flagella. Here we reconstruct the evolutionary history of four conserved type III secretion proteins and their phylogenetic relationships with flagellar paralogs. Our analysis indicates that the TTSS and the flagellar export mechanism share a common ancestor, but have evolved independently from one another. The suggestion that TTSS genes have evolved from genes encoding flagellar proteins is effectively refuted. A comparison of the species tree, as deduced from 16S rDNA sequences, to the protein phylogenetic trees has led to the identification of several major lateral transfer events involving clusters of TTSS genes. It is hypothesized that horizontal gene transfer has occurred much earlier and more frequently than previously inferred for TTSS genes and is, consequently, a major force shaping the evolution of species that harbor type III secretion systems.

Matzke castigates Luskin for saying that 2/3 of the proteins have no definite homolog. But in his 2003 paper, Matzke claims that 26 out of 40 proteins lack definite homologs. That sounds like about 2/3 to me. From Matzke's evolution in Brownian Space- 17 have no potential homolog - (Flif, FlhA, FlhB, FliN, and 12 axial proteins) 2 have "little current supporting evidence" of homology - (FlgA and FliG) 8 have suggestive homology (inconclusive) - (FliHJOPQR, FlgI, FlgH) What's more, his latest paper claims 15 unique (no homolog) proteins. But many of those are proteins that he previously claimed to have known homologs for. (FlgJMN, FlhE, FliELOSTZ) Has he lost the data, or changed his mind? Matzke also uses several different levels of criteria for determining homology - from a BLASTP, to a "slightly relaxed" PSI-BLAST, to "structural/other". "Other" presumably means that Matzke can postulate a homolog without any BLAST or structural data. What gives? Spearing Luskin for coming to the same conclusion he did 3 years ago? Changing status of homology? Including postulated homology as evidence for actual homology? I'm no molecular biologist, but something seems fishy here. Cant wait to hear from some actual molecular biologists on this.

Matzke castigates Luskin for saying that 2/3 of the proteins have no definite homolog. But in his 2003 paper, Matzke claims that 26 out of 40 proteins lack definite homologs. That sounds like about 2/3 to me.

From Matzke's evolution in Brownian Space- 17 have no potential homolog - (Flif, FlhA, FlhB, FliN, and 12 axial proteins)

2 have "little current supporting evidence" of homology - (FlgA and FliG)

8 have suggestive homology (inconclusive) - (FliHJOPQR, FlgI, FlgH)

What's more, his latest paper claims 15 unique (no homolog) proteins. But many of those are proteins that he previously claimed to have known homologs for. (FlgJMN, FlhE, FliELOSTZ) Has he lost the data, or changed his mind?

Matzke also uses several different levels of criteria for determining homology - from a BLASTP, to a "slightly relaxed" PSI-BLAST, to "structural/other". "Other" presumably means that Matzke can postulate a homolog without any BLAST or structural data.

What gives? Spearing Luskin for coming to the same conclusion he did 3 years ago? Changing status of homology? Including postulated homology as evidence for actual homology? I'm no molecular biologist, but something seems fishy here. Cant wait to hear from some actual molecular biologists on this.

One more dumb question: Is the non-homologousness of the bacterial, eukariotic, and archaeal flagellae sufficient to conclude that the three kinds of flagellae would have each developed after the prokaryote/achaea [archaea]/eukariote [eukaryote] split?

Very interesting, thanks.

Bravo! One can see the flagellum myth being nibbled away, or sometimes chomped back, as with this installment, and need only see how long it takes before the ID set simply drop it from their mystery list and jomp to some other antievolutionary icon.

I just learned about this today---could somebody please remind me where did I say in a currently published essay that 2/3 of flagellar parts lack homologs and what does Nick say in response to Mike Gene's critique of his homology arguments? Comment by Casey Luskin --- September 7, 2006 @ 7:42 pm http://www.uncommondescent.com/index.php/archives/1563#comment-59157

Huh. The quote is from his "lugnuts" article as I referenced in the OP.

When we look at the type III system its genes are commonly clustered and found on large virulence plasmids. When they are in the chromosome, their GC content is typically lower than the GC content of the surrounding genome. In other words, there is good reason to invoke horizontal transfer to explain type III distribution. In contrast, flagellar genes are usually split into three or more operons, they are not found on plasmids, and their GC content is the same as the surrounding genome. There is no evidence that the flagellum has been spread about by horizontal transfer.

Does your paper assert this claim made by Mike Gene?

I posted this in reply to the Uncommon Descent thread. However, my posts there have generally show up days/weeks later or not at all. Therefore, I am cross-posting it here. Thank you for your indulgence.

Scott wrote... "C'mon folks, why are we even still discussing the rediculous [sic] notion that the flagellum could have come about via Darwinian mechanisms?"

The reason this subject is of interest because it is something scientifically tangible. It provides an opportunity to explore the predictive nature of both ID and evolution. More importantly, debating this topic helps inform and define what is meant by the term "Intelligent Design" by various proponents.

At its core, the flagellum embodies the concept that if it looks designed it probably is. It also presents the challenge/opportunity to evolution proponents to hypothesize how this came about naturally. Why should anyone be surprised when evolution proponents start making predictions about the origin of the flagellum and testing those predictions? In my opinion debating this is both very topical and constructive (as opposed to arguing about who is and isn't a true Christian).

I read with interest Joseph's comments (even checked out his blog). This is an example of the positive aspects in discussing this. I can understand the position that the existence of the flagellum isn't "anti-evolution" but is designed the same way that the earth/moon system is designed. They both came about via natural processes but, if I understand correctly, Joseph and other ID proponents are arguing that it is unlikely, if not impossible, that either came about from "sheer-dumb-luck".

Unless ID wants to remain a simple variant of a "God in the gaps" argument, its proponents need to encourage discovery and discussion. We should all be eager for "actual molecular biologist[s]" to study and discuss evidence of the natural process that lead to the flagellum. This, of course, might be problematic for those whose real agenda is the promotion of the supernatural.

UPDATE - My post showed up at Uncommon Descent. It looks like things are improving in that regard.

I also wanted to add my appreciation of Nick Matzke's work.
I honestly hope it will lead to positive and constructive discourse on this interesting subject.

I posted some replies to comments at Telic Thoughts.

I have posted a short update to my Big Flagellum Essay from 2003.

I have also added some of the more recent articles to the Flagellum Background page.