In situ structures of the spirochete flagellar motor by cryo-ET reveal two distinct modes of interactions between the rotor ring and stator units. Together with new cryo-EM structures of the isolated stator units, this work provides insights into the mechanisms of torque generation and directional switch.
Motility is essential for bacterial survival and host infection. In many bacteria, the flagellum is the organelle responsible for rapid movement toward more desirable environments, but bacterial flagella are distinct from eukaryotic flagella, which are composed of microtubules and dynein and function as linear motors. The bacterial flagellum is a huge protein nanomachine, composed of more than 30 different proteins, each present in a few to tens of thousands of copies. There are three structural parts: the basal body, working as a rotary motor; the filament, a helical screw propeller; and the hook, a universal joint connecting the filament to the motor1,2. The flagellar motor consists of a rotor ring and multiple stator units surrounding the rotor, and it converts the electrochemical potential difference across the cell membrane into mechanical work with almost 100% efficiency3. The maximum rotation speed measured for bacterial flagella was 1,700 revolutions per second4, which is much faster than that of a Formula One racing car engine.
The structure and function of the bacterial flagellum have been intensely studied, with Salmonella enterica serovar Typhimurium and Escherichia coli being the main model systems1,2. Their flagellar basal bodies consist of the C ring, MS ring, LP ring and rod (Fig. 1a). The MS–C rings act as the rotor of the flagellar motor, whereas the rod and LP ring function as the drive shaft and bushing, respectively, to support high-speed rotation. The rod is a rigid, straight tube connected directly to the MS ring and transmits motor torque to the flagellar filament via the hook. The LP ring surrounds the distal part of the rod via smooth interfaces and a small gap, consistent with its bushing function5. The MS ring, formed by FliF, is the transmembrane core of the rotor. The C ring (formed by FliG, FliM and FliN) is attached to the cytoplasmic face of the MS ring through FliF–FliG interactions, and it is essential for torque generation and regulation of motor switching. Motor torque is generated by electrostatic interactions between the stator unit and the C ring. The stator unit is a transmembrane complex formed by MotAB family proteins; MotA has four transmembrane α-helices (TMHs), while MotB has a single TMH. MotAB proteins also work as a cation channel and are classified into three groups based on the coupling ion and sequence similarity: H+-coupled MotAB, Na+-coupled PomAB and Na+-coupled MotPS6. The stator complex also has a periplasmic domain, formed by the C-terminal region of MotB, which contains a peptidoglycan (PG) binding motif (MotBPGB), and a cytoplasmic domain, formed by the central region of MotA (MotACD). Motor torque is thought to be generated by the cyclic association and dissociation of MotACD with motor protein FliG, which is located at the top of the C ring, facing the membrane and the M ring (Fig. 1a).
The structure of the isolated flagellar basal body has been studied by cryo-EM and single-particle image analysis7, but the resolution has been limited to around a few nm due to the structural instability of the C ring and difficulties in purifying the basal body at sufficiently high concentration for cryo-EM data collection. In addition, the stator complexes cannot be isolated with the basal body by detergent solubilization of the membrane fraction due to the weak binding between the stator and C ring. Structural analysis of the entire flagellar basal body has been carried out by electron cryotomography (cryo-ET), but the thickness of the Salmonella and E. coli cell bodies (1 μm) and the relatively small numbers of flagella per cell (4–7) have hampered high-resolution cryo-ET and subtomogram averaging analyses in those species8. Now, Chang et al. have applied cryo-ET to the flagellar motor of the Lyme disease spirochete, Borrelia burgdorferi, which has multiple flagellar motors at the cell poles and is only ~200-nm thick, allowing efficient collection of high-resolution tomograms. They could thus visualize the in situ flagellar motor structure, including the stator–rotor interaction, in unprecedented detail9.
The flagellar motor is bidirectional, occasionally switching the direction of rotation between counter-clockwise (CCW) and clockwise (CW). Chang et al. analyzed the B. burgdorferi motor in both CCW and CW states and detected structural changes involved in the directional switching of motor rotation. Although the resolution was limited to around 2 nm, what they visualized is quite intriguing. The cylindrical wall-like structure of the C ring shows large conformational changes, both within its wall plane and in the tilt angle of the wall relative to the motor axis, closing the upper, membrane-facing edge to a diameter of 55 nm in the CCW state and opening it to 62 nm in the CW state (Fig. 1b). The stator complex appears to be fixed in a stable position around the rotor in both states, with multiple MotACD subunits forming a somewhat conical tube. In the CCW state, the C ring structure allows the FliG subunits on its top edge to interact with part of the MotACD tube that is closer to the motor axis; in the CW state, the FliG subunits interact with the opposite part of the MotACD tube, located further away from the motor axis.
On the basis of these two modes of rotor–stator interaction, Chang et al. propose a torque generation mechanism involving the CW rotation of the MotA tube around the MotB dimer at the center of the MotAB complex. The TMHs from these two MotB subunits presumably form two proton channels surrounded by the TMHs of MotA. Since MotB is anchored to the PG layer via MotBPGB and possibly to the LP ring as well, the tube formed by MotA subunits (TMHs and MotACD) could rotate around MotB by a conformational change of the entire MotAB complex, coupled with proton translocation through the channel. The CW rotation of MotACD could generate the torque for C ring rotation in either CCW or CW directions through different interaction modes with FliG at the top of the C ring, that is, either with the inner side of the tube (closer to the motor axis) or with the opposite outer side (Fig. 1b). This is just like a two-cogwheel gear system, composed of small and large cogwheels, with cogs at the cylinder edges that can switch their relative positions; the small cogwheel (MotA) always generates torque in one direction, but it can mesh with the large cogwheel (FliG in the C ring) via either its internal or external edges, hence driving rotation of the large cogwheel in either direction.
Interestingly, two other groups have recently determined structures of MotAB stator complexes isolated from different bacterial species (Clostridium sporogenes, Bacillus subtilis and Campylobacter jejuni). The structures were determined to near-atomic resolution by cryo-EM and revealed that the complex is formed by a pentamer of MotA and a dimer of MotB, with some indications of conformational changes suggesting that the MotA pentamer rotates around the two TMHs of MotB10,11. These two groups also propose a two-cogwheel gear model as a torque generation mechanism for the flagellar motor. The overall structure of these stator complexes is similar to that revealed by the cryo-ET analysis of the B. burgdorferi motor, making these proposals quite plausible.
Is the two-cogwheel gear mechanism actually the one that drives flagellar motor rotation? For this to be the case, proton translocation through the stator channel and motor rotation must be tightly coupled, but such coupling remains to be quantitatively examined, for example, by state-of-the-art single-molecule analysis of the motor dynamics, including the rotation of the MotA pentamer around the MotB dimer. Previous cryo-EM images of the flagellar basal body from a Salmonella mutant strain with a deletion fusion of FliF and FliG showed that the C ring inner lobe that connects the C ring wall and the M ring is missing altogether, with the top of the C ring so closely associated with the M ring and membrane12 that the gap between them appears to be too small to accommodate MotACD for CW rotation; however, the motor remains functional in this strain, with its rotation biased to the CW direction13. In any case, near-atomic resolution structural analyses of the C ring–stator complex in both CCW and CW states are also required to fully understand the MotA–FliG interactions involved in torque generation. We are stepping into a place of wonder through a door that has just opened before us.
Minamino, T. & Namba, K. J. Mol. Microbiol. Biotechnol. 7, 5–17 (2004).
Nakamura, S. & Minamino, T. Biomolecules 9, 279 (2019).
Berg, H. C. Annu. Rev. Biochem. 72, 19–54 (2003).
Magariyama, Y. et al. Nature 371, 752 (1994).
Fujii, T. et al. Nat. Commun. 8, 14276 (2017).
Minamino, T., Terahara, N., Kojima, S. & Namba, K. Mol. Microbiol. 109, 723–734 (2018).
Thomas, D., Francis, N. R., Xu, C. & DeRosier, D. J. J. Bacteriol. 188, 7039–7048 (2006).
Kawamoto, A. et al. Sci. Rep. 3, 3369 (2013).
Chang, Y. et al. Nat. Struct. Mol. Biol. https://doi.org/10.1038/s41594-020-0497-2 (2020)
Deme, J. C. et al. Nat. Microbiol. https://doi.org/10.1038/s41564-020-0788-8 (2020).
Santiveri, M. et al. Cell https://doi.org/10.1016/j.cell.2020.08.016 (2020).
Thomas, D. R., Morgan, D. G. & DeRosier, D. J. Proc. Natl Acad. Sci. USA 96, 10134–10139 (1999).
Sakai, T. et al. Mbio 10, e00079–19 (2019).
The author declares no competing interests.
About this article
Cite this article
Namba, K. A proposed gear mechanism for torque generation in the flagellar motor. Nat Struct Mol Biol 27, 1004–1006 (2020). https://doi.org/10.1038/s41594-020-00514-0