HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 25, 22251-22259, June 21, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/25/22251    most recent M110908200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Boukhvalova, M. S. Articles by Stewart, R. C. Search for Related Content PubMed PubMed Citation Articles by Boukhvalova, M. S. Articles by Stewart, R. C. Social Bookmarking                      What's this?

CheW Binding Interactions with CheA and Tar

IMPORTANCE FOR CHEMOTAXIS SIGNALING IN ESCHERICHIA COLI^*

Marina S. Boukhvalova, Frederick W. Dahlquist^§, and Richard C. Stewart^¶

From the  Department of Cell Biology and Molecular Genetics, University of Maryland, College Park, Maryland 20742 and the ^§ Institute of Molecular Biology and Department of Chemistry, University of Oregon, Eugene, Oregon 97403

Received for publication, November 14, 2001, and in revised form, February 7, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The initial signaling events underlying the chemotactic response of Escherichia coli to aspartic acid occur within a ternary complex that includes Tar (an aspartate receptor), CheA (a protein kinase), and CheW. Because CheW can bind to CheA and to Tar, it is thought to serve as an adapter protein in this complex. The functional importance of CheW binding interactions, however, has not been investigated. To better define the role of CheW and its binding interactions, we performed biochemical characterization of six mutant variants of CheW. We examined the ability of the purified mutant CheW proteins to bind to CheA and Tar, to promote formation of active ternary complexes, and to support chemotaxis in vivo. Our results indicate that mutations which eliminate CheW binding to Tar (V36M) or to CheA (G57D) result in a complete inability to form active ternary complexes in vitro and render the CheW protein incapable of mediating chemotaxis in vivo. The in vivo signaling pathway can, however, tolerate moderate changes in CheW-Tar and CheW-CheA affinities observed with several of the mutants (G133E, G41D, and 154ocr). One mutant (R62H) provided surprising results that may indicate a role for CheW in addition to binding CheA/receptors and promoting ternary complex formation.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The chemotaxis system of Escherichia coli allows this bacterium to control its movements in response to gradients of beneficial and noxious chemicals encountered as the cell swims through its environment (for reviews see Refs. 1-3). In this system, a well defined signal transduction pathway mediates communication between cell-surface receptor proteins and the flagellar motors, and this control allows the cell to migrate up gradients of chemoattractants or down gradients of chemorepellents. The transmembrane receptor proteins (also called MCPs)¹ sample the chemical environment and modulate, accordingly, the activity of CheA, an autophosphorylating protein histidine kinase (5-10). Phosphorylated CheA, in turn, directs activation of CheY by donating its phosphoryl group to it, and phospho-CheY then diffuses to the flagellar motors where it promotes clockwise flagellar rotation, an event that causes the cell to change its swimming direction (11-14). Binding of a chemoattractant molecule to the periplasmic domain of an MCP results in decreased CheA autokinase activity, reducing the intracellular pool of phospho-CheY and the probability of a change in swimming direction (7, 9, 15).

Phospho-CheA also directs activation of CheB, a methylesterase, in a similar manner; phospho-CheB removes methyl ester groups from the MCPs, a modification that contributes to sensory adaptation in the chemotaxis system (6, 16-18). Understanding how the MCPs regulate the autokinase activity of CheA is, therefore, central to understanding the mechanisms of signal transduction and sensory adaptation in this system. The magnitude of this regulation is impressive; the autokinase activity of MCP-coupled stimulated CheA is ~100-fold higher than that of uncoupled CheA and ~500-fold higher than that of MCP-coupled inhibited CheA (7, 9, 10, 19-22). How is this large dynamic range of regulation accomplished at a molecular level? Ames and Parkinson (10) have demonstrated that inhibition of CheA by MCPs could be mediated through a low affinity MCP·CheA complex. However, stimulatory MCP-CheA coupling requires CheW, a small cytoplasmic protein that promotes formation of ternary complexes involving MCPs, CheW, and CheA (5, 8, 9, 10, 21-26). MCP-mediated regulation of CheA activity (in response to binding of attractants/repellents to the MCPs) is thought to occur within this complex (7, 9, 24, 27) or perhaps by altering its composition (21-23).

Considerable progress has been made toward defining the composition and activities of the complex(es) formed by mixtures of CheW, CheA, and MCPs. Formation of a ternary complex was first demonstrated by the in vitro binding experiments of Gegner et al. (24), and subsequent electron microscopy studies supported the existence of such complexes in intact cells (25, 28). Initial in vitro experiments indicated a stoichiometry of 2:2:2 for the MCP·CheW·CheA complex (i.e. one receptor dimer and two CheW molecules bound per CheA dimer) and suggested that this complex was stable, forming and dissociating over a time course of many minutes (20, 24). However, more recent experiments indicate that several distinct ternary complexes are possible, including a 14:3:2 complex with very high CheA autokinase activity and a 2:2:2 complex in which the autokinase activity is very low (21, 22). Moreover, recent experiments suggest that the active ternary complex might be a dynamic assembly, forming and dissociating on a time scale comparable with that required for signal propagation by the chemotaxis signaling pathway (21, 22, 29). CheW might play a role in defining these dynamics, as suggested by the observation that addition of "extra CheW" to preformed ternary complexes causes rapid disassembly of the 14:3:2 complex (21, 22).

The results presented above give rise to numerous questions about the role of CheW in chemotaxis signal transduction. How does it promote ternary complex formation? How does it mediate regulation of CheA within the ternary complex(es)? However, the detailed biochemical role of CheW remains obscure, in large part because it is one of the least studied components of the chemotaxis system. To date, only one survey of cheW mutants has been published (30), and the only defined biochemical activities of CheW are its abilities to bind to MCPs and CheA. Analysis of genetic suppressors by Liu and Parkinson (30) established the in vivo importance of CheW-MCP interactions, and subsequent work by Gegner et al. (24) demonstrated the existence of a CheW·MCP complex of moderate affinity (K_d ~10 µM). In vivo evidence of CheW·CheA complexes came from immunoprecipitation studies by McNally and Matsumura (8), and subsequent work demonstrated binding of purified preparations of CheW to CheA with a K_d of 15 ± 2 µM and a binding stoichiometry of 1:1 (31).

The binding activities of CheW have led to the idea that CheW serves as an adapter protein; its role is to tether CheA to the MCPs and thereby enable appropriate modulation of CheA autokinase activity by the associated MCPs (24). In its simplest form, this "adapter model" predicts that cheW missense mutations that alter CheW activity in vivo should have a corresponding effect on the affinity of the protein for CheA and/or MCPs. Here we report our characterization of the altered binding properties of six mutant versions of CheW that were identified by virtue of their altered abilities to support chemotaxis in vivo under a variety of different conditions. Our results provide the first direct experimental support for the hypothesis that CheW binding interactions with MCPs and CheA are indeed important for its in vivo function. However, these results also suggest that the role of CheW may be more complex than simply tethering CheA to the MCPs; close-to-normal affinities for CheA and MCPs do not guarantee that CheW will function correctly in vivo.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials

Fluorescein-5-maleimide was purchased from Molecular Probes, Inc. (Eugene, OR), stored at 20 °C in the dark, and used under minimal lighting conditions. Polyethylene glycol 12,000 (PEG12,000) was purchased from Fluka and Sigma. All other chemicals were purchased from standard sources and were of reagent grade.

Strains and Plasmids

All bacterial strains used in this work are derivatives of E. coli K12. Strains RP437 (32) and RP3098 ((flhA-flhD) (33)) were kindly provided by J. S. Parkinson (University of Utah). Strain D392 (cheW (34, 35)) was used for swarm assays, whereas strain RP3098, lacking all chemotaxis genes, was used for purification of most of the proteins in this study. E. coli strain BL21 (DE3) (36) was used for production of (His)₆-LZ-Tar_c and CheY.

Plasmid pCW (31) containing the wild type cheW gene under control of the tac promoter was modified to generate plasmid pCnoW (lacking cheW gene but otherwise identical to pCW) and plasmid pCE (engineered to attach Met-(His)₆-Cys to the N-terminal end of CheW). To make pCnoW, a StyI-XbaI piece of pCW was excised, and the StyI and XbaI overhangs were ligated. This deletion removed codons 24-167 of cheW.

To create pCE, an oligonucleotide ((CATCAC)₃TGCGAATTC) was inserted at the NdeI site located at the cheW start codon in pCW. This oligonucleotide added coding sequences for a Met-(His)₆-Cys extension at the N-terminal end of CheW and introduced an EcoRI site upstream of the new, engineered cheW start codon. Wild type cheW and mutant cheW alleles were PCR-amplified from appropriate pCW vectors using primers that generated EcoRI and SalI sites upstream and downstream, respectively, of the cheW coding sequence. These PCR products were digested with EcoRI and SalI and ligated into corresponding sites of pCE.

The vector for overexpression of (His)₆-LZ-Tar_c (pDE:LZ-Tar_c) was created by ligating the GCN4-Tar_c coding sequence of pGCN4-pR (19) into pET14b. Plasmid pAR1:tar, encoding full-length Tar, was used to prepare membrane vesicles containing overproduced levels of E. coli Tar.

Plasmids pAR1:cheA and pT7:cheY (37) were used for overexpression of CheA and CheY, respectively. All plasmids created in the course of this work were verified by dideoxy sequencing (performed at the University of Maryland Center for Agricultural Biotechnology).

Mutagenesis and Screening

cheW mutants (except 154ocr) were obtained by random hydroxylamine mutagenesis of plasmid-encoded cheW following a procedure reported previously (5). Swarm assays (38) were used to screen transformants of cheW strain D392 for mutant cheW alleles.

Protein Purification and Preparation of Membranes

Wild type CheW and mutant CheW proteins were purified from the E. coli strain RP3098 transformed with plasmid pCE:cheW. CheW overexpression was induced by addition of 1 mM IPTG to cells in early exponential phase. 4 h after induction, cells were harvested by centrifugation. Cell pellets were resuspended in phosphate buffer (0.05 M sodium phosphate buffer (pH 8.0) with 1 mM imidazole, 0.3 M NaCl, and 2 µM 2-mercaptoethanol) and processed to produce protein extracts as described in the protocol of Hess et al. (11) for CheA preparation. Extracts were then loaded onto a nickel-nitrilotriacetic acid (Qiagen Inc., Santa Clarita, CA) chromatography column. After washes with 1 and 20 mM imidazole in phosphate buffer, protein was eluted from the column with 140 mM imidazole in phosphate buffer. Collected protein was dialyzed against TEGDP buffer, pH 7.5 (50 mM Tris, 0.5 mM EDTA, 10% glycerol, 0.5 mM dithiothreitol, and 0.1 mM phenylmethylsulfonyl fluoride), and loaded onto a DEAE column (Whatman). CheW was eluted from this column with a 0 to 0.3 M NaCl gradient in TEGDP buffer. Fractions containing CheW were identified by SDS-PAGE, concentrated to ~200 µM using a Centrifugal Filter Device (Millipore), and dialyzed against TEGDP buffer before storage at 80 °C.

(His)₆-LZ-Tar_c was purified from BL21(DE3) cells carrying plasmid pDE:LZ-Tar_c. Cells were grown at 30 °C in Luria Broth containing 100 µg ml¹ ampicillin until the culture reached early exponential phase and then induced with 125 µM IPTG. 3 h later cells were harvested by centrifugation. Cell pellets were resuspended in phosphate buffer, and protein extracts were produced as described above. Extracts were loaded onto a nickel-nitrilotriacetic acid column. After washing this column with 50 mM imidazole in phosphate buffer, we eluted (His)₆-LZ-Tar_c with 100 mM and 350 mM imidazole in phosphate buffer. Fractions containing (His)₆-LZ-Tar_c were then pooled and concentrated to ~200 µM using a Centrifugal Filter Device before dialysis against TEGDP buffer and storage at 80 °C.

Inner membrane vesicles containing Tar and control vesicles lacking receptors were prepared from E. coli strain RP3098 carrying pAR1:tar and pCnoW, respectively, according to the procedure of Osborne and Munson (39) as modified by Gegner et al. (24). Membrane vesicles were stored in TEND buffer, pH 7.5 (50 mM Tris, 0.5 mM EDTA, 0.1 M NaCl, and 0.5 mM dithiothreitol), at 20 °C.

CheA and CheY were purified from RP3098/pAR1:cheA and BL21/pT7:cheY, respectively, using previously published procedures (40).

The concentrations of purified proteins were estimated spectrophotometrically using calculated extinction coefficients (41) ((His)₆-Cys-CheW, ₂₈₀ = 6,085 M¹ cm¹; CheA, ₂₈₀ = 16,300 M¹ cm¹; CheY, ₂₈₀ = 8,250 M¹ cm¹; LZ-Tar_c, ₂₈₀ = 8,480 M¹ cm¹) and by staining of protein bands on SDS-PAGE (for membrane vesicles).

Swarm Assay

Fresh transformants of strain D392 carrying pCW (encoding wild type or mutant CheW) were stabbed into the center of tryptone swarm plates (38, 42) containing 0.3% Difco BactoAgar in tryptone broth (1% tryptone, 0.5% NaCl), 100 µg ml¹ ampicillin, and 0-50 µM IPTG. The diameter of the outermost ring of the resulting swarm colonies was measured after incubating the plates in a humid incubator at 30 °C.

CheW Labeling with Fluorescein

Wild type CheW and mutant CheW proteins (modified with a (His)₆-Cys tag), 200 µM, were spin-dialyzed against 50 mM Tris, pH 7.5, using Bio-Gel P-6 columns (Bio-Rad) and then treated with 2 mM fluorescein-5-maleimide for 30 min at 25 °C. Labeled protein (F*-CheW) was then spin-dialyzed against TEGD buffer (TEGDP buffer without phenylmethylsulfonyl fluoride) to remove the fluorescein that was not covalently attached to CheW. The labeled proteins were then stored at 20 °C.

Fluorescence Anisotropy Assay of CheW Binding to CheA

Fluorescence anisotropy experiments were carried out using a Jobin Yvon-Spex Fluoromax-2 Spectrofluorometer. Samples were maintained at 25 °C in a thermostated cuvette holder. The excitation and emission wavelengths were 492 and 517 nm, respectively, and the monochromators were set to give slit widths of 5 nm.

A solution of F*-CheW (0.8 µM) was prepared in TEND buffer containing 10% (w/v) PEG12,000 and then passed through a 0.45-µm filter. Aliquots of 110 µM CheA were added to a 1.25-ml sample of filtered F*-CheW solution in a cuvette. After each addition, fluorescence anisotropy was measured and subsequently plotted as a function of CheA concentration. The resulting binding isotherm was fitted using Equation 1 to determine the dissociation constant for CheW-CheA interaction (K_d^CheA),

(Eq. 1)

In Equation 1 (A  A_o)/(A_f  A_o) is the fraction of the total anisotropy change observed for a particular sample in a titration series; A_o is the initial anisotropy of F*-CheW before addition of CheA; A_f is a maximum anisotropy observed at saturation; and [CheA] is the concentration of free CheA in solution (equivalent to total [CheA] in these experiments because of the low concentrations of CheW and the relatively weak affinity of the CheW-CheA interaction). The best fit of Equation 1 to the CheW-CheA binding isotherm in the presence of PEG12,000 yielded a K_d^CheA value of 6.0 ± 0.2 µM. K_d^CheA values for CheW mutant proteins obtained by such analysis are given in Table I.

Assays to Monitor CheW binding to Tar in Membrane Vesicles

Direct Pull-down Assays-- Inner membrane vesicles containing Tar (final concentration 0-20 µM) or lacking receptors were incubated with F*-CheW (final concentration 2 µM) in TEND buffer containing 1 mg/ml BSA (to minimize nonspecific association of CheW with the membranes). These incubations were carried out in the dark for 10 min at 25 °C with gentle agitation. These membrane-protein mixtures were then subjected to centrifugation for 1 min at 13,500 rpm in a Hermle (Labnet) microcentrifuge to sediment the membranes and F*-CheW bound to the Tar in the membranes. An aliquot of the resulting supernatant was then diluted into TEND buffer, and the resulting fluorescence intensity was measured (Jobin Yvon-Spex Fluoromax-2 Spectrofluorometer with the excitation and emission wavelengths set as described above). The data generated by these "pull-downs" indicated a progressive decrease in the amount of F*-CheW present in the supernatant as the concentration of Tar was increased. To quantitatively analyze these results, the binding isotherms were fitted using Equation 2,

(Eq. 2)

In this equation (I  I_min)/(I_o  I_min) is the fraction of the total intensity change observed for a particular sample in a titration series; I_o represents the initial fluorescence intensity of F*-CheW solution before membrane addition; and I_min is the minimal fluorescence intensity remaining after complete depletion of supernatant (obtained by extrapolating binding curves to saturation). The best fit of Equation 2 yielded K_d^Tar = 10.8 ± 0.7 µM for wild type CheW (a value in good agreement with that reported by Gegner et al. (24)). This value of K_d^wtCheW,Tar was then used in calculations to determine K_d^Tar for different CheW mutant proteins from data obtained in a competition assay described below.

Pull-down Competition Assays-- Wild type F*-CheW (final concentration 2 µM) was premixed with 0-50 µM of a particular CheW mutant protein in TEND buffer containing 1 mg/ml BSA. Then inner membrane vesicles were added to the premix, generating a final Tar concentration of 11 µM. Reactions were incubated with agitation in the dark for 10 min at 25 °C. Vesicles were then sedimented, and the fluorescence emission intensity of the supernatant was measured as described above. The data generated by these experiments indicated a progressive increase in the concentration of F*-CheW present in the supernatant as increasing levels of competitor (mutant) CheW was added. To obtain an apparent K_d for the CheW-Tar interaction from this experiment, we plotted (I  I_o)/(I_f  I_o) versus [CheW^mutant] and fitted these plots using Equation 3,

(Eq. 3)

In this equation (I  I_o)/(I_f  I_o) is the fraction of the total intensity change observed for a particular sample in a titration series; I represents intensity remaining in the solution after sedimentation of receptor-containing vesicles; I_o is the initial intensity; I_f is a maximum fluorescence intensity corresponding to a complete release of F*-CheW from its association with receptor vesicles (observed at saturation); and [CheW] is total concentration of mutant CheW in solution. K_d^Tar values for the CheW mutants were then calculated using Equation 4,

(Eq. 4)

Applying this competition assay to wild type CheW generated a K_d^Tar value of 13 µM, in reasonable agreement with the value determined using the direct pull-down approach (11 µM, above) and in agreement with the value determined by Gegner et al. (24) using a similar approach.

Assays of CheA Activity in Ternary Complexes

Formation of kinase-activated ternary complexes was assayed by incubating 11 µM CheA, 0-70 µM CheW, and 50 µM LZ-Tar_c for 3 h at 30 °C in TEGD buffer containing 1 mg/ml BSA (21). Each protein mixture was then diluted 1:35 in TEND buffer containing 10% (w/v) PEG12,000, and CheA steady-state turnover was measured using a coupled ATPase system (43) as described in Ninfa et al. (9). Each reaction contained 2 mM ATP, 5 mM MgCl₂, and 50 µM CheY. Doubling the incubation time (from 3 to 6 h) and increasing the concentrations of the coupling assay reagents were found to have no effect on the outcome of these assays.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Isolation and in Vivo Characterization of cheW Mutants

Mutant strains of E. coli lacking functional CheW cannot accomplish chemotaxis; they exhibit an extreme smooth-swimming bias and cannot modulate their swimming pattern in response to chemoattractants or chemorepellents (32). This Che phenotype is readily observed as an inability to form expanding rings or "swarms" when colonies of these bacteria are grown in semisolid agar plates ("swarm plates") (42, 44). Chemotactic swarming ability can be restored to a cheW strain by providing a plasmid that encodes CheW and that directs expression of this protein at appropriate levels (23, 45). We used plasmid pCW (31) to enable IPTG-inducible expression of cheW over a wide range of levels in cheW host strain D392 (Fig. 1A). For example, D392/pCW cells grown in the absence of IPTG generated a CheW level ~30% of that present in wild-type cells; at 10 µM IPTG the level of CheW was ~2.5 times higher than in wild-type cells; 35 µM IPTG and 110 µM IPTG resulted in ~13-fold and a ~40-fold overexpression of CheW (relative to wild type), respectively. We examined the chemotactic abilities of cells under these induction conditions by monitoring their swarming abilities (Fig. 1B). In agreement with previous work (23, 45), we observed that full restoration of swarming ability required a CheW level close to that present in wild-type cells. If the level of CheW generated by the plasmid expression system exceeded the wild-type level by greater than 10-fold, then the chemotactic ability decreased, as reflected by a decrease in the rate of expansion of colonies in the swarm plates. This inhibitory effect of excess CheW is relatively small at moderate levels of overproduction (5-10-fold) but becomes quite noticeable as CheW levels reach 20-50 times the wild-type level.

View larger version (53K): [in this window] [in a new window]   Fig. 1.   CheW expression pattern and corresponding swarming ability of E. coli cells. A, Western blot analysis of CheW expression by wild type strain RP437 and by cheW mutant D392. Samples were prepared from cells grown to a mid-exponential phase in tryptone broth at 30 °C in the presence of indicated amounts of IPTG. The ~18-kDa band corresponds to CheW. The ~36-kDa band is an unknown protein cross-reacting with CheW antiserum. This band was used as a control of loading equivalency. B, a swarm plate assay of D392 cells expressing wild type CheW (1), "loss-of-function" mutant G57D (2), "weak titrator" mutant G41D (3), "strong titrator" mutant 154ocr (4), or no CheW (5). The cells were inoculated into 0.3% tryptone agar plates containing the indicated concentrations of IPTG. The plates were incubated at 30 °C for 10.5 h. The 110 µM IPTG plate was incubated then for an additional 12 h at 25 °C to more clearly display the differences between transformants 1 and 3.

To identify potentially informative mutant variants of CheW, we made use of swarm assays and the dominant-negative effect of CheW overproduction. We randomly mutagenized cheW (5), ligated the mutagenized gene into (unmutagenized) plasmid pCW, then introduced this pool of mutants into a cheW host strain, and screened these transformants using swarm assays at three different levels of CheW expression to identify three distinct sets of mutants. The mutant alleles causing aberrant chemotaxis behavior were then sequenced and subcloned (using an NdeI-XbaI fragment that carried only cheW coding sequences) into an unmutagenized pCW backbone and transformed into "fresh" D392 host cells (5); the resulting transformants were subjected to swarm assays to confirm that, for each mutant, the phenotype observed in the original screen was due to a particular cheW point mutation.

Screen to Isolate Loss-of-Function Mutants-- To identify CheW variants that were incapable of supporting chemotaxis, we screened D392/pCW transformants for a Che phenotype in swarm plates containing 10 µM IPTG, an inducer level chosen to produce optimal swarming for wild type CheW. Western blots of ~200 of these Che candidates indicated that, as anticipated, most of these mutations resulted in severely decreased levels of CheW and were therefore uninteresting for our purposes. However, three of our Che mutants encoded stable, full-length CheW proteins. These mutants were retained for detailed biochemical analysis as described below. The Che phenotype of one of these mutants is shown in Fig. 1B, and the amino acid sequence changes associated with the three Che mutations (deduced from DNA sequence) are shown in Table I.

We examined the abilities of these loss-of-function mutants to support swarming over a range of expression levels (Fig. 2). None of the loss-of-function mutants was able to restore a Che⁺ phenotype at any level of induction, but we did observe a range of severity in the phenotypes as follows. G57D had the least activity (swarm sizes indistinguishable from those observed with D392 carrying a control plasmid that lacked cheW); V36M and R62H generated swarm rates that were marginally better than this, but these were only 15-30% of the rate supported by wild-type CheW (Fig. 2A).

View larger version (23K): [in this window] [in a new window]   Fig. 2.   Swarm profiles of CheW mutants. Swarm plate assays were performed on the D392 (cheW) strain expressing wild type CheW (top gray line), no CheW (bottom gray line), and one of the following. A, loss-of-function mutants G57D (), V36M (), or R62H (); B, weak titrator mutants G133E () or G41D (); C, strong titrator mutant 154ocr (). Cells were inoculated near the center of 0.3% tryptone agar plates containing 0-50 µM IPTG. After 16 h of incubation at 30 °C, diameters of the swarms produced were measured. The solid lines connecting the data points are shown to aid differentiation among the data sets and have no theoretical significance.

Screen to Isolate Weak Titrator Mutants-- In a second screen of our cheW mutant pool, we set CheW expression levels at ~40 times the wild-type level by adding 110 µM IPTG to the swarm plates that we used to examine the chemotactic abilities of D392/pCW transformants. At this level of overproduction, wild-type CheW markedly inhibits swarming ability. We identified ~100 cheW mutants that exhibited much improved swarming ability relative to that observed for cells overexpressing wild-type cheW under these conditions. Western blots indicated that most of these mutations resulted in decreased levels of CheW, but two mutants had no noticeable effect on the size or levels of CheW; they were overexpressed to the same extent as wild-type CheW when grown in the presence of 110 µM IPTG. We concluded that the Che⁺ phenotype of these two mutants (Fig. 1B) was not due to diminished CheW protein levels but rather to a diminished ability of the mutant CheW proteins to participate in the interaction(s) responsible for the dominant-negative effects of CheW overproduction; the mutants were weakened in their ability to titrate some limiting component of the chemotaxis system. The amino acid changes caused by these mutations were deduced from DNA sequence analysis (Table I).

We examined the abilities of these weak titrator mutants to support swarming over a range of expression levels (Fig. 2B). At low-to-moderate levels of expression, the mutants were almost as effective as wild-type CheW in supporting chemotaxis, but (unlike wild-type CheW) they did not inhibit swarming even at elevated levels of expression induced by 50 µM IPTG.

Screen to Isolate Strong Titrator Mutants-- A third type of CheW variant was identified by screening for mutant alleles that supported close-to-normal swarming ability at low expression levels (5 µM IPTG) but greatly diminished swarming ability at moderately higher expression levels (35 µM IPTG). The 13-fold overexpression of wild-type CheW generated at 35 µM IPTG has only a small effect on the swarming ability of D392/pCW, but with several of our mutant alleles this level of overproduction reduced the swarming rate to less than 10% of that supported by wild-type CheW under comparable conditions (Fig. 1B). Western blots indicated that these mutations did not cause elevated levels of CheW. We hypothesized that the mutant phenotype resulted from an enhanced activity or increased interaction of the mutant CheW proteins with other components of the chemotaxis system. DNA sequencing results indicated that mutants identified in this screen carried frameshift or nonsense mutations affecting the C-terminal 13-20 codons. We chose one of these mutants, cheW154ocr, for detailed biochemical analysis. The nonsense mutation in this allele results in production of a CheW protein lacking the C-terminal 13 amino acids of the wild-type protein. The ability of cheW154ocr (later referred to as "154ocr") to support chemotactic swarming at low expression levels but not at moderate expression levels is shown in Fig. 2C.

Fluorescein-labeled CheW and Its Interactions with CheA and Tar

Previous work indicated three in vitro activities for purified CheW as follows: (i) binding to CheA; (ii) binding to chemotaxis receptor proteins, such as Tar; and (iii) formation of receptor·CheW·CheA complexes in which the autokinase activity of CheA is markedly higher than that observed for uncomplexed CheA. We sought to examine whether the mutations we had identified affected these in vitro activities. To facilitate binding assays, we generated a version of CheW that could be easily purified and easily labeled with a fluorescent tag. Thus, we engineered an N-terminal Met-(His)₆-Cys addition to CheW such that it could be purified by nickel-nitrilotriacetic acid chromatography. This protein was then treated with fluorescein-5-maleimide to generate fluorescein-labeled CheW (F*-CheW) in which the fluorescent label was covalently attached to the side chain thiol group of the added Cys (CheW lacks any naturally occurring cysteines). (His)₆-Cys-tagged versions of our mutant CheW proteins were also generated and labeled with fluorescein using this approach.

We used fluorescein-labeled CheW in fluorescence anisotropy experiments to assay binding of CheW to CheA. The anisotropy of the fluorescein emission signal of the labeled CheW increased significantly upon addition of CheA (Fig. 3A). This increase was sufficiently rapid that it was complete by the time of the first measurement following CheA addition and mixing (~30 s). A plot of the magnitude of the observed anisotropy change as a function of CheA concentration indicated a hyperbolic relationship. Addition of comparable amounts of BSA or CheA-(1-616), a mutant version of CheA lacking the CheW binding domain (46), gave rise to a diminutive change in anisotropy (less than ~8% of the change observed with full-length CheA). We concluded that the anisotropy change observed in mixtures of F*-CheW and CheA reflected a saturable binding interaction between CheA and the labeled CheW. Curve fitting of the titration results indicated a K_d of ~18 µM for the F*-CheW·CheA complex, and Scatchard analysis (47) indicated a binding stoichiometry of one molecule of CheA (one monomer equivalent) per CheW molecule. Both values agree well with previous estimates reported by Gegner and Dahlquist (31) based on Hummel-Dreyer chromatography experiments.

View larger version (25K): [in this window] [in a new window]   Fig. 3.   Fluorescence anisotropy studies of fluorescein-labeled (His)6-Cys-CheW. A, the effect of CheA on the fluorescence anisotropy (A = (I  I)/(I + 2I)). Increasing amounts of CheA were added to a 0.8 µM solution of fluorescein-labeled CheW in the presence () or absence () of 10% (w/v) PEG12,000. Fluorescence anisotropy was measured. The initial anisotropy (Ao) of F*-CheW was ~0.10 in the absence of PEG12,000 and ~0.11 in the presence of PEG12,000. Titrations of F*-CheW with CheA-(1-616) () and BSA () in the presence of PEG12,000 are also shown. The curves connecting the data points represent the best (least squares) fit of binding isotherms by Equation 1 as described under "Experimental Procedures." The inset shows Scatchard analysis of the CheW-CheA interaction in the presence of PEG12,000 (Y represents fractional saturation). B, Perrin plot of the anisotropy of F*-CheW. Anisotropy was measured for mixtures containing 0.6 µM fluorescein-labeled CheW alone () or in combination with 28 µM CheA (), or 28 µM BSA (). Sample viscosity was adjusted using sucrose over a range from 0.7 to 2.2 cPoise.

We also examined the effect of the polymer polyethylene glycol M_r 12,000 on the F*-CheW-CheA binding interaction reflected by the anisotropy change. Addition of PEG12,000 enhanced the binding affinity severalfold (Fig. 3A), decreasing the K_d of the F*-CheW·CheA complex to ~6 µM (compared with 18 µM in the absence of PEG). All subsequent CheA-CheW binding experiments were carried out in the presence of this "macromolecular crowding agent" (48-50).

To explore further the nature of the anisotropy change observed when CheA is added to F*-CheW, we examined the effect of solvent viscosity on the anisotropy of F*-CheW alone and in the presence of high concentrations of CheA. When the results of these experiments are presented in the form of Perrin plots (51, 52), the slopes of the lines reflect the effective size of the fluorescent macromolecule; larger macromolecules (such as a complex between F*-CheW and CheA) are expected to give rise to a smaller slope than are smaller macromolecules (such as uncomplexed F*-CheW). Our results agree with this qualitative prediction: the slope of the Perrin plot for the mixture of F*-CheW and CheA is one-third the slope observed for F*-CheW alone or F*-CheW in the presence of BSA (Fig. 3B). The extrapolated y axis intercepts of these plots indicate that the free rotation of the fluorescein probe (absent any contribution due to rotational diffusion of the CheW protein molecule) appears to be unaffected by formation of the complex between F*-CheW and CheA. Thus, the increased anisotropy observed in the presence of CheA reflects the decrease in rotational diffusion of the CheW protein as a result of its binding to CheA.

Fluorecein-labeled CheW was also a useful tool to monitor binding interactions between CheW and the chemotaxis receptor Tar. These experiments are similar to those utilized by Gegner et al. (24) to monitor binding of radioactively labeled CheW to Tsr. Inner membrane vesicles can be readily pelleted by centrifugation. We observed that when such vesicles, containing Tar, are mixed with F*-CheW, pelleting the vesicles pulls down much of the F*-CheW, and the resulting supernatant (post-centrifugation) is depleted of fluorescent material (Fig. 5A). The greater the vesicle/Tar concentration, the more F*-CheW is removed from the supernatant. The relationship between F*-CheW disappearance and Tar concentration is consistent with a saturable binding interaction occurring with a K_d of ~11 µM (a value in good agreement with that reported by Gegner et al. (24)). By contrast, vesicles lacking Tar (or any other chemotaxis receptor) were ineffective in such "pull-down" experiments. Scatchard analysis of CheW-Tar binding revealed a 1:1 stoichiometry of binding, consistent with previous findings (20, 24).

Effects of Mutations on CheW Binding to CheA and Tar

We used the assays described above to investigate the binding abilities of our set of mutant CheW proteins. These proteins, modified with an N-terminal Met-(His)₆-Cys addition, were purified, labeled with fluorescein-5-maleimide, and used in titration experiments. Fluorescence anisotropy experiments indicated that each of the mutant proteins exhibited decreased affinity for CheA, relative to that observed for wild type CheW (Table I). Binding curves of a representative set of mutants are shown in Fig. 4. Analysis of the binding isotherms observed for CheW mutant proteins yielded the K_d^CheA values shown in Table I. Our mutant set exhibited increases in K_d^CheA values ranging from moderate (e.g. a ~2.5-fold increase in K_d^CheA value for mutant R62H) to severe (e.g. mutant G57D exhibited no detectable binding to CheA).

View larger version (21K): [in this window] [in a new window]   Fig. 4.   Binding of the representative CheW mutant proteins to CheA. Titrations were performed by adding CheA to 0.8 µM solutions of fluorescein-labeled wild type CheW (), R62H (), V36M (), or G57D () in buffer containing 10% (w/v) PEG12,000. The results of titrations of F*-CheW with BSA in the presence of PEG12,000 are also shown (). Curve fits are as described for Fig. 3A.

To assess the effects of CheW mutations on the ability of the protein to interact with Tar, we monitored the ability of CheW mutants to compete with wild type F*-CheW for Tar-binding sites in membrane vesicles. The results of such experiments for a representative set of mutants are shown in Fig. 5B. Analysis of these binding competition experiments yielded the K_d^Tar values presented in Table I. These results indicate that five of the six CheW mutants had diminished affinity for Tar. This defect ranged from moderate (e.g. a ~2-fold increase in K_d^Tar value for the mutant R62H) to severe (no detectable binding observed for mutant V36M).

View larger version (23K): [in this window] [in a new window]   Fig. 5.   F*-CheW binding to Tar-containing membrane vesicles. A, direct pull-down assay. Increasing amounts of Tar-containing membrane vesicles were added to 2 µM solutions of fluorescein-labeled wild type CheW (), or 154ocr () in TEND buffer in the presence of 1 mg/ml BSA. As a control, vesicles lacking Tar were added to 2 µM F*-CheW (). After a 10-min incubation at 25 °C, membrane vesicles (and proteins bound to them) were sedimented by centrifugation, and the fluorescence intensity remaining in the supernatant was measured. The inset shows Scatchard analysis of CheW-Tar interaction (Y represents fractional saturation). B, competition assay to monitor displacement of fluorescein-labeled wild type CheW by mutant CheW proteins. 2 µM F*-CheWwt was added to 11 µM Tar (in membrane vesicles) in the presence of increasing amounts of the following: unlabeled wild type CheW (); CheW mutants G41D () or V36 M (); or BSA (). After a 10-min incubation at 25 °C, membrane vesicles and bound proteins were sedimented by centrifugation, and the fluorescence intensity remaining in the supernatant was measured. Curves represent the best fit of binding isotherms to Equations 2 (A) and 3 (B) as described under "Experimental Procedures."

The sixth mutant (154ocr) exhibited an intriguing set of results in Tar-binding experiments. This CheW mutant protein did not compete with the wild type CheW for receptor-containing vesicles. Instead of causing an increase in the fluorescence intensity in the supernatant, increasing amounts of 154ocr in the "competition" mixtures caused a further drop in the level of F*-CheW remaining in the supernatant (data not shown). This observation raised the possibility that CheW 154ocr might enhance the affinity of Tar for wild type CheW. To explore further the effects of this mutation, we monitored direct binding of fluorescein-labeled CheW 154ocr to Tar in membrane vesicles. The results of this experiment (Fig. 5A) indicate that the mutant 154ocr has an enhanced affinity for Tar, ~3-fold tighter than that observed for wild type CheW.

Effects of Mutations on CheW Participation in the Ternary Signaling Complex

The experiments described above examined the effects of mutations on pairwise interactions of CheW with CheA and CheW with Tar. However, to mediate chemotactic signal transduction, CheW operates in the context of a ternary signaling complex composed of CheW, CheA, and receptor (21, 22, 24, 26). We investigated the ability of our mutant CheW proteins to form and function in the ternary complex. This complex can be generated conveniently in vitro by incubating CheW with CheA and soluble MCP domain constructs (10, 20-22). Ternary complex formation in these mixtures can be monitored by examining the dramatic increase in CheA autokinase activity associated with complex formation (10, 19-22). We performed titration experiments in which we added increasing concentrations of CheW (mutant or wild-type) to solutions containing fixed levels of CheA and LZ-Tar_c, a construct that has a leucine zipper fused to the cytoplasmic (soluble) portion of Tar (19). Following an appropriate incubation interval (3 h) to allow complex formation, we assayed the resulting autokinase activities of these mixtures. As shown in Fig. 6, each of our mutant CheW proteins was less effective than wild-type CheW in these titrations. Two of the mutant proteins (G57D and V36M) exhibited no detectable ability to form an activated complex. The other mutant proteins were capable of complex formation (as indicated by CheA activation) but did so only at CheW concentrations significantly higher than that required for wild-type CheW.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   Effect of CheW mutant proteins on CheA kinase activity in the presence of Tar. Mixtures containing 11 µM CheA, 50 µM LZ-QEQE Tar (19), 1 mg/ml BSA, and 0-70 µM CheW were incubated at 30 °C for 3 h. Kinase activity of CheA was then measured by an enzyme-coupled ATPase assay (43). Results for wild type CheW (), 154ocr (), V36M (), and G57D () (A); R62H (), G41D (), G133E (), and wild type CheW (·) (B) are shown. The solid lines connecting the data points are shown to aid differentiation among the data sets and have no theoretical significance.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The initial events of the chemotaxis signaling cascade involve communication between the transmembrane MCPs, such as an aspartate receptor Tar, and the histidine autokinase CheA. CheW is necessary for this communication and has been proposed to serve as an adapter protein that physically couples receptors and CheA in the ternary MCP·CheW·CheA complexes (21, 24). The idea that CheW functions as an adapter is based on its ability to bind to CheA and MCPs in vitro. Although such a role is consistent with available information about CheW, it is important to note that, prior to the work presented here, the functional importance of CheW binding interactions with CheA and receptors had not been investigated, for example, by examining whether mutations that affect CheW function have corresponding effects on its ability to bind to CheA and/or receptors. We investigated the binding abilities of mutant CheW proteins that we identified by screening for aberrant chemotaxis behavior in vivo. We analyzed the ability of mutant proteins to promote ternary complex formation in vitro and to support chemotaxis in vivo under a variety of conditions. Despite the limited number of mutants analyzed here, several important patterns emerged.

CheW in Vivo Function Can Be Abolished by Mutations That Eliminate CheW Binding to CheA or Tar-- Among our loss-of-function (Che) mutants, we identified one (G57D) that lacked detectable binding to CheA and one (V36M) that could not bind to Tar. These two mutants were unable to promote formation of an activated ternary complex in vitro. Thus, the ability of CheW to bind CheA and Tar in pairwise interactions appears to be crucial for formation and/or activity of the ternary complex and for overall signaling in vivo. These results support the idea that both the CheW-CheA and the CheW-Tar binding interactions are necessary for ternary complex formation and that CheW serves as an adapter protein mediating communication between receptors and CheA.

Mutations That Diminish the Affinity of CheW for Tar and/or CheA Also Decrease Its Ability to Form an Activated Ternary Complex-- Four mutants (R62H, G41D, 154ocr, and G133E) exhibited moderate decreases in their affinity for CheA and Tar. These mutants were also defective in their ability to form an activated ternary complex in vitro, and the magnitude of this coupling defect paralleled the severity of binding defects as follows. Mutant G133E was least efficient in complex formation and had the highest K_d^CheA and K_d^Tar values in this subset of mutants; R62H was most efficient in terms of ternary complex formation and exhibited the lowest K_d^CheA and K_d^Tar values in this subset of mutants. This observation is consistent with the idea that the binding interactions monitored for the isolated pairwise binding experiments are related to (and perhaps identical to) the binding interactions that are involved in formation of the ternary complex.

The relationship between the activated ternary complex and the pairwise CheW·CheA and CheW·MCP complexes is not well established. Several distinct ternary complexes are possible (21): a very large, kinase-active complex with 14 molecules of MCP and 3 molecules of CheW per CheA dimer; a small kinase-inhibited complex having 2 MCP molecules and 2 CheW molecules per CheA dimer; and perhaps several complexes with sizes/stoichiometries intermediate between these two extremes. In mixtures of CheW, CheA, and MCPs, the resulting populations of active and inactive ternary complexes are sensitive to the relative concentrations of the proteins in a manner that is not, at present, easily understood. In particular, the presence of "too much" CheW can inhibit formation of the kinase-active complex and/or promote formation of the small, kinase-inhibited complex (7, 21, 22). This observation raised the possibility that CheW binding to CheA and MCPs in active complexes is qualitatively different from that in inactive ternary complexes. Our results argue against this possibility by supporting the hypothesis that binding contacts involved in pairwise CheW-CheA and CheW-Tar interactions are related to the binding contacts that mediate formation of the activated Tar·CheW·CheA complex.

Moderate Decreases in CheW Binding Affinities Are Tolerated by the in Vivo Signaling System-- The three "titrator" mutants isolated and analyzed here exhibited 3-12-fold reductions in their affinities for CheA and Tar. These affinity changes gave rise to corresponding decreases in the level of active ternary complexes in mixtures of CheA, CheW, and Tar. However, these mutations had little observable effect on the in vivo efficacy of the chemotaxis signaling pathway, as monitored by swarm assays, when CheW levels were set at wild-type levels. This observation suggests either that the chemotaxis signaling pathway can operate effectively with significantly diminished levels of the signaling complex or that in vivo some mechanism exists to compensate for the diminished affinities and to restore the level of the signaling complex with these weak-binding mutants. The adaptation machinery of the chemotaxis system might contribute to this tolerance by, for example, adjusting the methylation status of the receptor protein in a manner that modulates the activity of the signaling complex or the affinity of the receptor for CheW in a manner that compensates for the effects of these mutations (53, 54).

An Ability to Bind CheA and MCPs Does Not Guarantee Functionality of CheW-- Whereas diminished affinities of CheW for CheA and Tar do not preclude some mutant CheW variants from supporting chemotaxis, in one case (R62H) we observed complete loss of function in vivo despite only small effects on the affinity of CheW for CheA, the affinity of CheW for Tar, and the ability to form an active ternary complex. Taken together these results suggest that CheW has an activity/ability in addition to binding to CheA/receptor and that this activity/ability is affected in the R62H mutant. We are currently searching for additional examples of CheW mutants with a similar phenotype.

CheW-Tar Interaction Influences Inhibitory Effect of CheW Overexpression-- Although the chemotaxis system can operate efficiently over a wide range of CheW levels, there are limits to this tolerance, and the system becomes progressively less effective as CheW levels exceed the wild type level by over 10-fold (23, 45) (Fig. 1). The molecular mechanism underlying this inhibitory effect of excess CheW is unknown at present, but it seems likely that the "extra" CheW sequesters some key component of the signaling pathway, rendering that component inactive or insensitive to chemostimuli. Our results with the "weak titrator" and "strong titrator" mutants provide some insight into the nature of this inhibition and the target component sequestered by the excess CheW. In particular, we noted that weak titrator mutants exhibited ~4-fold decreases in their affinity for Tar (relative to that exhibited by wild type CheW), whereas the strong titrator mutant exhibited a ~3-fold increase in the affinity of this binding interaction. Both types of mutations affected CheA interactions in a similar manner, increasing K_d^CheA by 3-12-fold. These results suggest that the affinity of CheW for receptors is the key for determining the ability of excess CheW to inhibit chemotactic signaling, a finding that implies that the limiting component sequestered by CheW is the receptor population.

Role of C-terminal Segment of CheW-- The C-terminal region of CheW, encompassing residues 154-167, appears to exert a negative influence on CheW binding to Tar. Deletion of this region significantly increases the affinity of CheW for receptors and reduces its affinity for CheA. Moreover, when mixtures of CheW 154ocr and wild type (full-length) CheW are added to Tar, the truncated protein appears to facilitate tighter interaction between wild type (full-length) CheW molecules and receptor. This observation suggests that, at least with CheW 154ocr, cooperativity may contribute to CheW-MCP binding interactions, a possibility we are currently exploring in greater detail.

Location and Conservation of Mutation Sites-- When projected onto the NMR-derived structure of CheW isolated from Thermotoga maritima (55), each of our mutation sites lies on the solvent-exposed surface of CheW, consistent with the notion that these positions are accessible for binding interactions with CheA and MCPs. Moreover, each of our point mutations affected positions that are highly conserved among CheW proteins identified in different proteobacteria, and some of these positions (e.g. Gly-57, Arg-62, and Gly-133) are conserved among CheW proteins present in microbial species as diverse as archaea, firmicutes, and spirochaeta.² Although this conservation might reflect conservation of specific contact points at a CheW-CheA or CheW-MCP interface, it is not clear why such contact points could not accommodate, through extended evolution, mutual compensatory changes on the interacting partners. Why is it so important to have a specific amino acid at a contact point? Perhaps, such sites reflect selective pressure for CheW to maintain some activity or conformation that goes beyond simple tethering of CheA to the chemotaxis receptors.

	ACKNOWLEDGEMENTS

We thank Amy Roth, who contributed to identification of several of our cheW mutants, and Sandy Parkinson for numerous E. coli strains and for insightful criticism of this manuscript.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants GM52853 (to R. C. S.) and GM59544 (to F. W. D.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^¶ To whom correspondence should be addressed. Tel.: 301-405-5475; Fax: 301-314-9489; E-mail: rs224@umail.umd.edu.

Published, JBC Papers in Press, March 28, 2002, DOI 10.1074/jbc.M110908200

² M. S. Boukhvalova, unpublished observations.

	ABBREVIATIONS

The abbreviations used are: MCPs, methyl-accepting chemotaxis proteins (4); PEG12, 000, polyethylene glycol 12,000; IPTG, isopropyl -D-thiogalactoside; BSA, bovine serum albumin; LZ-Tar_c, a construct that has a leucine zipper fused to the cytoplasmic (soluble) portion of Tar (19); F*-CheW, fluorescein-labeled CheW.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES