Straight and rigid flagellar hook made by insertion of the FlgG specific sequence into FlgE.

The bacterial flagellar hook connects the helical flagellar filament to the rotary motor at its base. Bending flexibility of the hook allows the helical filaments to form a bundle behind the cell body to produce thrust for bacterial motility. The hook protein FlgE shows considerable sequence and structural similarities to the distal rod protein FlgG; however, the hook is supercoiled and flexible as a universal joint whereas the rod is straight and rigid as a drive shaft. A short FlgG specific sequence (GSS) has been postulated to confer the rigidity on the FlgG rod, and insertion of GSS at the position between Phe-42 and Ala-43 of FlgE actually made the hook straight. However, it remains unclear whether inserted GSS confers the rigidity as well. Here, we provide evidence that insertion of GSS makes the hook much more rigid. The GSS insertion inhibited flagellar bundle formation behind the cell body, thereby reducing motility. This indicates that the GSS insertion markedly reduced the bending flexibility of the hook. Therefore, we propose that the inserted GSS makes axial packing interactions of FlgE subunits much tighter in the hook to suppress axial compression and extension of the protofilaments required for bending flexibility.

Identical folds used for distinct mechanical functions of the bacterial flagellar rod and hook.

The bacterial flagellum is a motile organelle driven by a rotary motor, and its axial portions function as a drive shaft (rod), a universal joint (hook) and a helical propeller (filament). The rod and hook are directly connected to each other, with their subunit proteins FlgG and FlgE having 39% sequence identity, but show distinct mechanical properties; the rod is straight and rigid as a drive shaft whereas the hook is flexible in bending as a universal joint. Here we report the structure of the rod and comparison with that of the hook. While these two structures have the same helical symmetry and repeat distance and nearly identical folds of corresponding domains, the domain orientations differ by ∼7°, resulting in tight and loose axial subunit packing in the rod and hook, respectively, conferring the rigidity on the rod and flexibility on the hook. This provides a good example of versatile use of a protein structure in biological organisms.

Assembly and stoichiometry of FliF and FlhA in Salmonella flagellar basal body.

The bacterial flagellar export apparatus is required for the construction of the bacterial flagella beyond the cytoplasmic membrane. The membrane-embedded part of the export apparatus, which consists of FlhA, FlhB, FliO, FliP, FliQ and FliR, is located in the central pore of the MS ring formed by 26 copies of FliF. The C-terminal cytoplasmic domain of FlhA is located in the centre of the cavity within the C ring made of FliG, FliM and FliN. FlhA interacts with FliF, but its assembly mechanism remains unclear. Here, we fused yellow fluorescent protein (YFP) and cyan fluorescent protein (CFP) to the C-termini of FliF and FlhA and investigated their subcellular localization by fluorescence microscopy. The punctate pattern of FliF-YFP localization required FliG but neither FliM, FliN, FlhA, FlhB, FliO, FliP, FliQ nor FliR. In contrast, FlhA-CFP localization required FliF, FliG, FliO, FliP, FliQ and FliR. The number of FlhA-YFP molecules associated with the MS ring was estimated to be about nine. We suggest that FlhA assembles into the export gate along with other membrane components during the MS ring complex formation in a co-ordinated manner.

Distinct roles of highly conserved charged residues at the MotA-FliG interface in bacterial flagellar motor rotation.

Electrostatic interactions between the stator protein MotA and the rotor protein FliG are important for bacterial flagellar motor rotation. Arg90 and Glu98 of MotA are required not only for torque generation but also for stator assembly around the rotor, but their actual roles remain unknown. Here we analyzed the roles of functionally important charged residues at the MotA-FliG interface in motor performance. About 75% of the motA(R90E) cells and 45% of the motA(E98K) cells showed no fluorescent spots of green fluorescent protein (GFP)-MotB, indicating reduced efficiency of stator assembly around the rotor. The FliG(D289K) and FliG(R281V) mutations, which restore the motility of the motA(R90E) and motA(E98K) mutants, respectively, showed reduced numbers and intensity of GFP-MotB spots as well. The FliG(D289K) mutation significantly recovered the localization of GFP-MotB to the motor in the motA(R90E) mutant, whereas the FliG(R281V) mutation did not recover the GFP-MotB localization in the motA(E98K) mutant. These results suggest that the MotA-Arg90-FliG-Asp289 interaction is critical for the proper positioning of the stators around the rotor, whereas the MotA-Glu98-FliG-Arg281 interaction is more important for torque generation.