|Home | About | Journals | Submit | Contact Us | Français|
Salmonella flgG point mutations produce filamentous rod structures whose lengths are determined by FliK. FliK length variants produce rods with lengths proportional to the corresponding FliK molecular size, suggesting that FliK controls the length of not only the hook but also the rod by the same molecular mechanism.
The bacterial flagellum is a motility organelle (2, 10). The flagellum is composed of three main structures: (i) the basal body, an engine embedded in the membranes and cell wall which includes a rotor and stator embedded in the cytoplasmic membrane; (ii) a long, external filament that acts as a propeller; and (iii) a hook acting as a universal joint that connects them. The flagellar base also incorporates a type III secretion system to export flagellar proteins during flagellar assembly (1, 11, 12). The exported flagellar proteins polymerize onto the growing tip of the extracellular substructures of the flagellum. The flagellar type III secretion system proteins form the three principal axial structures of the flagellum: the rod, hook, and filament. The rod acts as a drive shaft to transmit torque generated at the cytoplasmic membrane to the filament. The rod is further divided into the proximal rod, composed of FlgB, FlgC, and FlgF, and the distal rod, composed of FlgG (7, 13).
The four flagellar axial structures all grow to defined lengths, with different length control mechanisms for each. The length of the filament is believed to be self-limited, being controlled by hindered diffusion. This model is supported by the fact that filament growth slows down with length, reaching a zero-growth rate at a length of 15 μm (6). A molecular ruler, FliK, controls hook length by catalyzing a substrate specificity switch from hook type to filament type substrates at a hook length of ca. 55 nm (3). The loss of FliK results in a polyhook phenotype, with many hooks extending up to a micrometer in length. The self-limited polymerization of the distal-rod protein FlgG is thought to determine final rod length (4). Mutations within the flgG gene result in extended rod structures, termed filamentous rods. The loss of FliK in a filamentous-rod mutant strain results in a polyrod phenotype, with rods up to a micrometer in length.
In this paper, we provide the first reported measurements of the flagellar rod structures in wild-type strains, as well as the lengths of rods produced in various fliK backgrounds (Table (Table1),1), and show that FliK directly controls filamentous-rod length in flgG mutant strains as well as the rod-hook length in flgG+ strains.
This study focuses on accurate measurements of the distal rod to elucidate rod length control. First, we measured the distance between the distal edge of the MS ring and the distal edge of the L ring as the length of the whole rod, which was 21.5 nm (the average for 105 samples) (Fig. (Fig.1A).1A). Second, we measured the lengths of the proximal rods of polyrods (see below) and found a consistent length of 10 nm (the average for 51 samples) (Fig. (Fig.1B).1B). Then we subtracted the proximal-rod length from the whole-rod length to denote the length of the distal rod only (11.5 nm).
flgG point mutations give rise to filamentous rods (4). The resulting mutant rods gave a variety of flagellar structures, depending on the mutation sites (Fig. (Fig.1C).1C). Most (about 60%) of the flagellar structures had a basal structure: a filamentous rod attached to an MS ring (class A). The remaining structures (about 40%) had a basal structure attached to filaments; 39% had filamentous rods attached with no apparent hook structure in between (class B), and 1% had filamentous rods attached to hook-filament structures (class C).
In the cases of class A and class B structures, the length of the filamentous outer rod was regulated at an average ± standard deviation (SD) of 48.8 ± 12.5 nm (n = 154) (4), in contrast to the wild-type hook length of 55.0 ± 6.0 nm (5). It would be reasonable to assume that FliK may control the rod length by the same mechanism used for hook length. In the case of the rare (1%) class C structures, however, the rod length and the hook length concomitantly changed. Interestingly, the total length of the outer-rod-hook structure was constant at 74.0 ± 5.0 nm (n = 106) (4).
The filamentous-rod and hook component proteins, FlgG and FlgE, are homologous in amino acid sequence (4). However, the transition from rod to hook polymerization requires that the hook cap (FlgD) replace the rod cap (FlgJ). Apparently, this occurs only rarely in filamentous-rod mutants to produce the class C structures and not in the cases of class A and class B structures. The high frequency of class A structures suggests that the replacement of the rod cap by the filament cap (with no intermediate hook cap) is not as efficient as the replacement of the rod cap by a hook cap, as occurs in wild-type flagellar assembly.
When a complete deletion of the fliK gene was introduced into a flgG(G65V) filamentous-rod mutant, cells produced rods with no apparent length control, termed polyrods. The length distribution of the polyrods was 143.1 ± 89.0 nm (n = 180) (4), a much wider range than that observed for the filamentous rods (see above) yet narrower than the range reported for polyhooks (5, 9). Occasionally, flgG ΔfliK double mutants gave rise to polyrods with polyhooks, indicating that the mutated rod subunits could still initiate hook polymerization. Interestingly, even in this case, the outer-rod length exhibited apparent control at 69.1 ± 21.8 nm (n = 102) (4), which is shorter than the average length observed for polyrods and for normal rod-hook structures.
We have previously constructed chimeras of FliK and YscP of various lengths and shown that the lengthening of FliK results in an increase in the hook length (14). We used the same set of FliK-YscP chimeras to test the effect of FliK on the rod length. Intact flagella were isolated, and rod lengths were measured as described previously (5). The rod lengths measured were increased in concert with an increase in FliK length, as observed previously for hook lengths (14), but with a wider distribution (Table (Table22 and Fig. Fig.2).2). It is important that the hooks also have distal rods attached, which would add 11.5 nm to their lengths. Thus, FliK variants exhibit rod-hook structures longer than filamentous rod structures before catalyzing the secretion specificity switch. Regardless, the data show that there is a correlation between the FliK length and the rod length, confirming that FliK controls the rod length in a manner similar to the control of hook length.
Why does the distal-rod length terminate at 21.5 nm in the wild type when it is clear from the filamentous-rod mutants that polymerization can continue well beyond that length? We have shown that proper rod termination is required for normal PL ring assembly, which is required for the hook to polymerize outside the cell (4). The inability to terminate rod length at 21.5 nm results in periplasmic flagella. It has been assumed that the PL ring is a bushing in the peptidoglycan-outer membrane (15). Our data suggest that the PL ring is required to form a pore in the outer membrane and perhaps facilitate the transition from rod polymerization to hook polymerization (4). We presume that termination of a rod at 21.5 nm is required to place the distal end of the rod perpendicular to the outer membrane to facilitate PL ring pore formation within the outer membrane. Since the loss of normal rod length termination is seen only in flgG mutants, it appears that the normal signal for FlgG to terminate polymerization is an intrinsic property of FlgG itself. The length of the wild-type distal rod is 11.5 nm, which corresponds to the thickness of four turns of the FlgG basic helix. This geometry would place one subunit of FlgG on top of another, but no more.
This work was supported by a CREST project grant from the Japan Science and Technology Agency (to S.-I.A.) and by PHS grant GM56141 from the NIH (to K.T.H.).
Published ahead of print on 7 August 2009.