Charged residues of the rotor protein FliG essential for torque generation in the flagellar motor of Escherichia coli¹

Abstract

The FliG protein of Escherichia coli is essential for assembly and function of the flagellar motor. Certain mutations in FliG give a non-motile, or Mot⁻, phenotype, in which flagella are assembled but do not rotate. Mutations with this property are clustered in a C-terminal segment of FliG that is stable when expressed alone, and thus probably constitutes an independently folded domain. Previously, we suggested that this domain forms the rotor portion of the active site for torque generation in the motor. In this work, we have used a mutational approach to identify the amino acid residues in the C-terminal domain of FliG that are most important for motor function. Site-directed mutagenesis was used to replace each of the conserved residues in this domain with alanine, and the effects on motor function were measured. Because charged residues have often been suggested to have important roles in torque generation, conserved charged residues were changed individually and in all pairwise combinations. The results show that three charged residues of FliG, Arg279, Asp286 and Asp287, are directly involved in torque generation. Mutations in these residues cause motility defects that suggest that they function jointly, in an active site whose most important property is a specific arrangement of charges. Two other charged residues, Lys262 and Arg295, may also be involved in torque generation, but are less critical than Arg279, Asp286 or Asp287. Uncharged residues of the FliG motility domain do not appear to have direct roles in torque generation, although some are needed for the stability of the protein or for normal clockwise/counter-clockwise switching. The Mot⁻ mutations of fliG isolated previously by random mutagenesis do not alter the putative active-site residues, but render the proteins abnormally susceptible to proteolysis, suggesting significantly altered conformations or reduced stabilities.

Introduction

Motile cells of Escherichia coli are propelled by flagella, thin helical filaments each driven at the base by a reversible rotary motor anchored in the cell membrane (reviewed by Macnab 1992, Macnab 1996, Schuster and Khan 1994, Blair 1995, Berg 1995). The energy for rotation comes from the transmembrane gradient of protons Larsen et al 1974, Manson et al 1977, or in certain species sodium ions (Hirota & Imae, 1983). The flagellar motor is thus a device that converts chemical energy stored in a transmembrane ion gradient into the mechanical energy of rotation. The molecular mechanism of this energy conversion is not understood.

About 50 proteins are required for the assembly and operation of the flagella of Escherichia coli or Salmonella typhimurium (Macnab, 1992). Most of these are needed for assembly of the flagellum or for the control of switching between clockwise (CW) and counter-clockwise (CCW) directions of rotation. Genetic analyses suggest that only a few flagellar proteins are closely involved in torque generation Enomoto 1966, Hilmen and Simon 1976, Silverman and Simon 1976, Yamaguchi et al 1986a, Yamaguchi et al 1986b. These are MotA, MotB, FliG, FliM and FliN. Certain mutations in each of these proteins can give a Mot⁻ (non-motile) phenotype, in which flagella are assembled and appear normal in intact cells but do not rotate.

MotA and MotB are integral membrane proteins that together form a proton channel Blair and Berg 1990, Blair and Berg 1991, Wilson and Macnab 1988, Stolz and Berg 1991, Garza et al 1995. MotB is believed also to anchor the channel to the peptidoglycan layer Chun and Parkinson 1988, Blair et al 1991, DeMot and Vanderleyden 1994, Garza et al 1996a, Garza et al 1996b; this would imply that MotA and MotB are components of the stator, or non-rotating part, of the motor. Each flagellar motor contains several MotA/MotB complexes, which can function independently Block and Berg 1984, Blair and Berg 1988, and which are probably arranged in a circle around the basal-body MS-ring (Khan et al., 1988).

FliG, FliM and FliN are required for flagellar assembly and torque generation, and for the control of switching between CW and CCW directions of rotation Yamaguchi et al 1986a, Sockett et al 1992, Irikura et al 1993. Intergenic suppression analyses (Yamaguchi et al., 1986b) and ultrastructural studies Khan et al 1992, Francis et al 1992, Francis et al 1994, Zhao et al 1995, Zhao et al 1996 suggest that they function in a complex at the base of the flagellum, often called the “switch complex.” A genetic fusion of FliM to FliN has been characterized in detail, and has properties that suggest that these proteins are associated with each other in the flagellum (Kihara et al., 1996). Recently, in vitro binding studies (Tang et al., 1996) and studies using the two-hybrid system in yeast (Marykwas et al., 1996) have provided direct evidence of a complex containing FliG, FliM and FliN. Ultrastructural studies Francis et al 1992, Francis et al 1994, Khan et al 1992, Zhao et al 1995, and a study of binding interactions in vitro (Oosawa et al., 1994), showed that FliG is associated with the basal-body MS-ring, and that FliM and FliN are components in a large ring-shaped or bell-shaped structure extending from the MS-ring into the cytoplasm. Collectively, the evidence strongly supports the view that FliG, FliM and FliN are all components of the rotor. Physiological studies are also consistent with a rotor location for FliG, FliM and FliN; underexpression or overexpression of these proteins has effects that suggest that they function together, in a complex distinct from MotA and MotB (Tang and Blair 1995, Tang et al 1995; our unpublished results).

We recently reported evidence that suggests that FliG is more closely involved than FliM or FliN in torque generation (Lloyd et al., 1996). Specifically, we found that the Mot⁻ mutants of fliM or fliN could be made motile to various degrees, either by overexpressing the mutant proteins or by overexpressing another of the switch-complex proteins. The fliM and fliN Mot⁻ mutations thus appear to affect the binding of one or more of the switch-complex proteins to their sites in the flagellum, rather than torque generation per se. In contrast, several of the fliG Mot⁻ mutations abolish motor rotation, while allowing good flagellation, at all levels of expression of the mutant protein or other switch-complex proteins. These mutations, which disrupt torque generation specifically, are clustered in a C-terminal domain of FliG. On the basis of these and other observations, we suggested that the C-terminal domain of FliG forms the rotor portion of the site of torque generation (Lloyd et al., 1996). Consistent with this proposal is the observation that certain defects in the stator components MotA and MotB can be suppressed by mutations in FliG Garza et al 1996a, Garza et al 1996b, and a small deletion in FliG can be suppressed by a mutation in MotB Yamaguchi et al 1986a, Irikura et al 1993.

Here, we have used a mutational approach to identify the residues in the C-terminal domain of FliG that are most important for torque generation. Evolutionarily conserved residues were identified by comparison of the available FliG sequences, and these residues were then replaced by alanine (and in some cases by other residues as well) using site-directed mutagenesis. Because electrostatic interactions among charged groups have been invoked in a number of proposals for the mechanism of torque generation in the flagellar motor (e.g. see Glagolev and Skulachev 1978, Murata et al 1989, Blair 1990, Berry 1993), charged residues were given closest attention in this mutational analysis, being changed singly and in all pairwise combinations. The results show that three charged residues, Arg279, Asp286 and Asp287, are directly involved in torque generation. We suggest that these residues form an active site for torque generation on the rotor of the flagellar motor, and that the arrangement of charged groups within this active site is its most important feature.