Evolution of the Bacterial Flagella

Introduction

I recently ran across some material that bears directly on this question. In this article I'll outline the construction of eubacterial flagella, it's relationship to other systems, and end with a speculative scenario for the evolution of eubacterial flagella.

First off, there is no "the" bacterial flagella. Eubacteria and archebacteria have flagella that look almost identical, except that they are composed of completely different, non-homologous proteins (and they are both different from the eukaryote flagellum)(3,12). I'll ignore the archebacterial system as it is based on class IV pilins and clearly not IC (3,12).

Given that the eubacterial and archebacterial flagellar systems are unrelated, the eubacterial flagella must have arisen after the eubacterial/archebacterial split.

Within the eubacteria there are at least two, possibly three, flagella systems(1,2). In the eubacterial system, the best studied flagella are those of E. coli and S. typhimurium, and I'll briefly outline the structure of these as the "canonical" eubacterial flagella (but see below).

The eubacterial flagellum is a biochemical assemblage that deserves the appelation "molecular machine", and does resemble something a human would design (2,7,11). There is the helical filament (=propellor), the hook (=universal joint), the rod (=drive shaft), the S-P ring (=bushing around the rod, only in gram negative bacteria), and the SMC ring complex, the "motor" which includes the stator and the rotor. The entire assembly is hollow, including the actual filament. The significance of this will become apparent later. The rotor, hook and fillament are made of (non-identical) helical proteins that self assemble to give hollow, cylindrical structures (in the case of the filament, the cylinder is helical so that it acts as a "screw propellor" when it rotates.

Many eubacteria can switch the direction of rotation of the propellor (and hence the direction they travel in) and the "switch" mechanism appears to be part of the motor complex.

While around 50 genes are involved in the construction and regulation of the canonical eubacterial only around 18-20 form the actual motor-switch/shaft/propellor complex. I'll now describe the substructures in somewhat boring detail, but please bear with me.

Structure of eubacterial flagella

I could find less information on the hook (universal joint). This is formed by (at least) FlgE, there is very limited sequence similarity between the FlgE's of Salmonella sp. and Heliobacter sp. (33% identity), and it is not clear that they are derived from the same ancestral protein. FlgE mutants do not form flagella.

I'll ignore the L and P rings (FlgH, FlgI), as these can be absent or present without any effects on flagella function, so can't be part of the IC core :-)

The rod (driveshaft) is comprosed of a complex of the proteins FlgG, FlgB, FlgC, FlgF and FliJ, P, Q, R, I don't have any information about muations of these.

The M ring is formed from FliF, again, I don't have any info on mutations of this.

The C ring motor complex is formed from FliG, FliM, FliN and in some species, MotA and B, in others PomA, PomB, MotX and MotY.

Mot A and B forms a proton pump which provides the power of the motor, MotB also serves to anchor the motor to the cell. Deletion of Mot A or B paralyses the cell, however, it may be possible to overcome this paralysis by over expression of Mot A (in MotB mutants), FliG and FliM (2,7,11).

In Rhodobacter sphaeroides, the genes equivalent to MotA and MotB have only 19% sequence identity, so it is not clear if these proteins are divergent modifcations of the ancestral MotA MotB or convergent evolution of an unrelated motor(11,13). The Rhodbacter motor doesn't switch as does the other motors, but turns on and off, and re-orientation is via brownian motion (13). This suggests that the switch complex is not part of the IC core :-)

In Vibro species, the motor is a sodium, rather than proton pump, composed of pomA, pomB, MotX and MotY, the MotX,Y are unrelated to MotA or B, but PomA seems to be related to MotA from R. spheroides, and R. spheroides MotA can partialy restore swiming in PomA paralysed mutants (1,9).

FliG is involved in torque generation, transducing the proton gradient into rotational motion in a poorly understood way. It may also have a role in switching (2,7,11). FliM is involved in switching, but probably not in torque generation.

FliN is probably not involved directly in either torgue generation or switching, and may be a stabilizing protein. In Bacillus sp it is replaced by the protein FliY, which intriguingly resemebles a fusion between FliM and N.

Confused yet? This site has some nice diagrams and discussion of the eubacterial flagellum, but is a little out of date (especially on the roles of FliN and FliG). The flagellum certainly is complex, even irreducibly complex in this canonical form, but can it evolve? Well, there is still a lot unknown about eubacterial flagella, but we can get some ideas, and a plausible pathway from the literature.

Homologies with other systems

Importantly, the type III secreteory system forms a "rivet" structure identical to the rod and SMC ring complex of the flagellum (8,10,11). Furthermore, the switching/torque generation system FliG,FliN/Y, has homologs in virtually every type III secretory system examined so far(8,10,11)! Proteins exported by this system are shunted through the hollow SMC ring and through the rod to the outside of the cell(8,10). In flagellum assembly, flagellins and hook proteins are shunted to the outside of the cell via the rod and ring complex. The proteins attach to the outer rim of the rod and self assemble into a tubular structure that will become the hook + fillament, and flagellar proteins pass through this tube as it grows(11).

There is no aparent homolog of the motor (MotAB) in type III secretory systems.

Intrigingly, several of the type III secretory systems have tubular structures attached to the rod. The Hrp secretion system forms basal ring/rod system with a pilus that strongly resembles the flagellar system. It is not clear if the Hrp pilus has any relation to the flagellar filament(10). However, E coli has a filamentous structure attached to one of its type III secretory systems which has significant similarity to the flagellar filament(4).

Proposed evolutionary pathways

The proto-flagellar filament arose next as part of the protein secretion structure (cf the Pseudomonas pilus, the Salmonella filamentous apendages and the E coli filamentous structures), finally (as suggested by the presence of at least two, if not three, indpendent motors) an ion pump which was doing something else [see note] became associated with this structure and motility (presumably weak) occured fortuitosly. Even today MotAB can freely dissociate and re-associate with the flagellar structure. This early, limited motility was later refined into the more compentent system we see today. Alternatively, the ion pumps became linked to the proto-flagella to provide extra "power" to pump proteins out of the complex, and flagella motion occured via fortuitous mutations in the linkers Fli G,N,M later on. Regulation and switching can be added on later as there are modern eubacteria that lack these and function well in their environments.

Significantly, flagellar and other non-flagellar proteins are exported into culture medium via the flagellum, and MotA deletion slows transport through the flagellum, suggesting that the flagella still functions as a protein secretion system(5,6). Furthermore, the genes for "rivet" rod and SMC ring complex form a single transcription unit, are the orientation and order of these genes are very similar between the type III secretory systems and the flagellum(8).

This re-enforces the idea that eubacterial flagella and type III transport arose from a common ancestral system, and that motility arose as the co-option of an exitsing system that was doing something else. It would not be the first time that a secretory system was co-opted into motility, as the cyanobacteria have co-opted a carbohydrate export system to produce gliding motion.

The above discussion dealt with the accquisition of the main components, the molecular details of the actual changes are less certain, given how little we know about the details of the channeling of the protons and how FliG generates rotation. A speculative idea for the evolutionary linkage of the protoFliG to motor function is as follows. FliG, the torque generator, is a Y shaped molecule, and the rotor sits on the arms of the Y. The protoFLiG was originally a support system for the transport machinary. Adding the MotAB system to the protoflagellar complex adds protons to drive the transported substrates faster up the hollow rotor. Then a mutation alters protoFliG so that proton acceptance by FliG causes one of the arms of the Y to rotate (as seen in other molecules), then this would rotate the rotor, and hence the filament, and motion commences. As I said this is speculative, and a more detailed analysis of FliG and the FliG homologs, plus other components of the system, is needed to get a clearer picture.

Conclusions

This is a very tentative sketch, but it does seem that a fully detailed evolutionary explanation for eubacterial flagella is not so distant. While the details of the motor/rotor/filament system assembly seem reasonably clear, the details of the evolution of the FliG,M,N torque generating sytem are lacking, as we know little about how these systems generate torque.

It would be very interesting to see if addition of MotAB to a type III secretory system produces torque without any further modifcation. Can you replace FliG with HrpQ (or other homologs) in the flagella and still get torque? Can you substitute motors between Virio and E. coli? This is a rich field for research.

The eubacterial flagellar system is also interesting as it shows how misleading "design" thinking is. In this case what is defined as IC depends on our point of view. When viewed as a motile stucture, the flagella is IC. Remove the motor, it stops functioning, remove the hook (universal joint) it stops functioning, remove the fliament it ... well, it still works sort of :-). Viewing the flagellum as a motor, and an IC motor at that, provides no insights into the origin or functioning of this structure.

But view it as a secretory structure, it is NOT IC, remove the filament and it still works, remove the hook and it still works, remove the motor and it still works, not as well as with the motor, but it still works.

Thus, if the flagella is a secretory system that has been co-opted for a motile function (while still retaining some of it's secretory function), then the ICness of the system is in the mind of the beholder, and a clear path for it's evolution is opened up.

Evolution of clotting and immunoglobulin systems:
See this site for an overview.
Clotting: Doolittle and references therein, and George Actons post of the month.
Immune system: A review of the evolution of the immune system and references therein.

Previous web based works on the bacterial flagellar system:
Of course, I had missed some peoples writings (don't you just hate that). A good begining that covers some issues for flagella and cillia that I ignore is found here. And Howard Hershey covers a lot of the same ground as I do in a series of posts to Talk.Origins begining with this post.

References:

2. Berry RM, and Armitage JP. (1999). The bacterial flagella motor. Adv Microb Physiol, 41, 291-337.

3. Faugy DM and Farrel K, (1999 Feb) A twisted tale: the origin and evolution of motility and chemotaxis in prokaryotes. Microbiology, 145, 279-280.

4. Delahay RM, Knutton S, Shaw RK, Hartland EL, Pallen MJ, Frankel G. (1999 Dec 10) The coiled-coil domain of EspA is essential for the assembly of the type III secretion translocon on the surface of enteropathogenic Escherichia coli. J Biol Chem, 274(50):35969-74.

5. Komoriya K, Shibano N, Higano T, Azuma N, Yamaguchi S, Aizawa SI. (1999 Nov) Flagellar proteins and type III-exported virulence factors are the predominant proteins secreted into the culture media of Salmonella typhimurium. Mol Microbiol, 34(4):767-79.

6. Young GM, Schmiel DH, Miller VL. (1999 May 25) A new pathway for the secretion of virulence factors by bacteria: the flagellar export apparatus functions as a protein-secretion system. Proc Natl Acad Sci U S A, 96(11):6456-61.

7. DeRosier DJ. (1998 Apr 3). The turn of the screw: the bacterial flagellar motor. Cell, 93, 17-20.

8. Hueck CJ (1998 Jun) Type III Protein Secretion Systems in Bacterial Pathogens of Animals and Plants. Microbiol Mol Biol Rev, 62, 379-433.

9. Asai Y, Kojima S, Kato H, Nishioka N, Kawagishi I, and Homma M. (1997Aug). Putative channel components for the fast-rotating sodium-driven flagellar motor of a marine bacterium. J Bacteriol, 179, 5104-10.

10. He SY. (1997 Dec) Hrp-controlled interkingdom protein transport: learning from flagellar assembly. Trends Micro, 5, 489-495.

11. Harshey RM and Toguchi A (1996 Jun) Spining Tails: homologies amongst bacterial flagellar systems. Trends Micro, 4, 226-231.

12. Jarrel KF, Bayley DP and Kostyukova AS (1996 Sep) The Archael Flagellum: a unique motility structure. J Bacteriol, 178, 5057-5064.

13. Shah DS, and Sockett RE. (1995 Sep). Analysis of the motA flagellar motor gene from Rhodobacter sphaeroides, a bacterium with a unidirectional, stop-start flagellum. Mol Microbiol, 17, 961-9.

14. Behe, MJ. Darwin's Black Box: The Biochemical Challenge to Evolution (New York: The Free Press, 1996)