Mutants with defective phosphatase activity show no phosphorylation-dependent oligomerization of CheZ. The phosphatase of bacterial chemotaxis.

CheZ is the phosphatase of CheY, the response regulator in bacterial chemotaxis. The mechanism by which the activity of CheZ is regulated is not known. We used cheZ mutants of Salmonella typhimurium, which had been isolated by Sockett et al. (Sockett, H., Yamaguchi, S., Kihara, M., Irikura, V. M., and Macnab, R. M.(1992) J. Bacteriol. 174, 793-806), for cloning the mutant cheZ genes, overexpressing and purifying their products. We then measured the phosphatase activity, binding to CheY and to phosphorylated CheY (CheYP), and CheYP-dependent oligomerization of the mutant CheZ proteins. While all the mutant proteins were defective in their phosphatase activity, they bound to CheY and CheYP as well as wild-type CheZ. However, unlike wild-type CheZ, all the four mutant proteins failed to oligomerize upon interaction with CheYP. On the basis of these and earlier results it is suggested that (i) oligomerization is required for the phosphatase activity of CheZ, (ii) the region defined by residues 141-145 plays an important role in mediating CheZ oligomerization and CheYP dephosphorylation but is not necessary for the binding to CheYP, (iii) the oligomerization and hence the phosphatase activity are regulated by the level of CheYP, and (iv) this regulation plays a role in the adaptation to chemotactic stimuli.

signal transduction (24) implicated the occurrence of such a regulation mechanism.
In addition to its phosphatase activity, CheZ was found to be involved in two other processes: binding to CheY (25,26) (mainly to its phosphorylated form (26)), and oligomerization upon binding to CheYϳP (27). With the goal of finding out whether the oligomerization is involved in CheZ regulation, this study examines, by biochemical analysis of mutant CheZ proteins, the inter-relationship between the different functions of CheZ.

EXPERIMENTAL PROCEDURES
Bacterial Strains-The strains used in this study are listed in Table I.
Cloning of Mutant cheZ Genes onto an Overexpressing Vector-For the overexpression of mutant CheZ proteins we used a collection of cheZ mutants generated and characterized by Sockett et al. (28). The cheZ mutant alleles were amplified from total DNA of the respective Salmonella typhimurium strain by polymerase chain reaction using the primers 5Ј-CCGAATTCATGATGCAACCATCTATCAAGCC-3Ј and 5Ј-CCG-GATCCTTAACAGCCAAGACTGTCCAGCA-3Ј, which contained added EcoRI and BamHI sites, respectively, at their 5Ј end. The amplified cheZ-containing fragments were digested with EcoRI and BamHI, and ligated with pBTac1 (Boehringer Mannheim) predigested with EcoRI and BamHI (Fig. 1). The resultant plasmids (Table II) overexpressed the wild-type and mutant CheZ proteins under the control of the tac promoter. The existence of the mutations in the cloned cheZ genes and the lack of additional mutations that might be caused by polymerase chain reaction were confirmed by DNA sequencing.
Overexpression and Purification of CheZ-Wild-type CheZ and the mutant proteins CheZ141FI and CheZ143DE were overexpressed in RP3098. CheZ110LP was overexpressed in RP1616, and CheZ145TM in BW3. The purification of CheZ was carried out as described in the preceding paper (27) except that, in the case of CheZ110LP and CheZ145TM, buffer A used for washing the Sepharose CL-6B column, contained 225 mM NaCl (instead of 275 mM), and the elution of CheZ was carried out by a gradient of 225-450 mM NaCl (instead of 275-450 mM). The overexpression and purification of CheY from S. typhimurium was described in the preceding paper (27).
Sensitivity of CheZ to Proteolysis-CheZ (83 M) in Tris-HCl (50 mM, pH 7.9) and MgCl 2 (5 mM) were incubated at room temperature (22°C) with trypsin (Sigma T-8642, 4 g/ml). Samples of 10 l were removed at the indicated time points, quenched by addition of 3 l of ϫ5 concentrated sample buffer and 10 min boiling, and analyzed by 15% SDSpolyacrylamide gel electrophoresis.
Phosphatase Activity of CheZ-The phosphatase activity of CheZ was assayed by monitoring the steady-state level of CheY phosphorylation in the presence of [ 32 P]acetyl phosphate (AcP) as described earlier (27).
Binding of CheZ to CheY-Binding of CheZ to CheY immobilized onto CNBr-activated Sepharose beads was measured as follows. Immobi-* This study was supported by Grant 93-00211 from the United States-Israel Binational Science Foundation (BSF), Jerusalem, Israel. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
lized CheY (50 M) or immobilized bovine serum albumin (as a control) were prepared as described (26). The reaction mixture (200 l), consisting of CheY-or BSA beads, Tris-HCl (50 mM, pH 7.9), MgCl 2 (7.7 mM), glycerol (5.3%), bovine serum albumin (2.7 mg/ml), [ 3 H]CheZ (0.2 M, 48,000 -60,000 cpm), and, where indicated, AcP (18 mM), was incubated for 15 min at room temperature (22°C). The beads were then pelleted by a brief centrifugation, and washed twice by 0.5 ml of Tris-HCl (50 mM, pH 7.9), MgCl 2 (5.0 mM), and AcP (18 mM, only when it had been included also in the assay mixture). The bound protein was extracted from the beads by agitating them in 400 l of SDS (10%) for 1 h, after which the beads were pelleted and the amount of [ 3 H]CheZ in the supernatant was determined by scintillation counting.

Cloning of Mutant cheZ Genes, and Overexpression and Purification of Their
Products-In order to determine whether any of the functions of CheZ (binding to CheY, phosphatase activity, and oligomerization) are correlated, we chose to analyze these functions in several mutant CheZ proteins. To this end we used S. typhimurium cheZ point mutants isolated and characterized by Sockett et al. (28) (Table I). Flagellar rotation in all these mutants is CW biased, indicating that these cheZ alleles code for CheZ proteins with impaired ability to antagonize the CW causing activity of CheY in vivo. To produce the proteins, we amplified the mutant genes by polymerase chain reaction directly from the chromosomal DNA of the mutant strains, and cloned the amplified genes into the expression vector pBTac ( Fig. 1 and Table II). The plasmids containing the cloned genes indeed overexpressed the mutant CheZ proteins, thus enabling us to purify the proteins to near homogeneity (Fig. 2).
The Phosphatase Activity of Mutant CheZ Proteins-The phosphatase activity of each of the mutant CheZ proteins was determined by measuring its effect on the level of CheY phosphorylation under steady-state conditions. All the mutant proteins had a significantly lower phosphatase activity than wildtype CheZ (Fig. 3). For example, in the presence of 2 M CheZ, the level of CheYϳP under steady-state conditions was 7% in the wild type versus 75, 76, 62, or 37% in CheZ110LP, CheZ141FI, CheZ143DE, or CheZ145TM, respectively (100% is the steady-state level of CheYϳP in the absence of CheZ). This means that CheZ110LP and CheZ141FI had over 10-fold lower phosphatase activity than the wild type, and CheZ 143DE and CheZ145TM had 9-and 5-fold lower activities, respectively.
Binding of Mutant CheZ Proteins to CheY-The apparent defects in the phosphatase activity of the mutant CheZ proteins could result from one of the following possibilities: (i) a reduced binding to CheYϳP; (ii) an increased binding to non-phosphorylated CheY, as a result of which the CheZ-CheY complex could not dissociate and CheZ could not become available to other CheYϳP molecules (CheZ was used in catalytic concentrations); or (iii) a reduced catalytic activity. To distinguish between these possibilities, we measured the binding of the mutant CheZ proteins to both phosphorylated and non-phosphorylated CheY. For this purpose the binding of [ 3 H]CheZ to CheY immobilized onto CNBr-activated Sepharose beads was determined in the presence or absence of the phosphodonor AcP (29). We used a large excess of CheY and AcP, thus making the loss of CheYϳP, resulting from the phosphatase activity, negligible. As shown in Fig. 4, all the mutant CheZ proteins were bound to the CheY beads to an extent similar to that of wildtype CheZ under both phosphorylating and non-phosphorylating conditions. These results, obtained with subsaturating concentrations of CheZ, indicate that the mutant CheZ proteins are not defective in their binding to either CheYϳP or CheY. Thus, by way of elimination, it seems that the third possibility, i.e. a reduced catalytic activity, is the cause of the apparent defects in phosphatase activity of the mutants.
Oligomerization of the Mutant CheZ Proteins-As before (27), we used cross-linking to determine the occurrence of CheZ oligomerization. Under non-phosphorylating conditions, i.e. in the absence of AcP, similar cross-linking products were obtained with the wild-type and mutant proteins. Thus, in the absence of CheY, the major band was that of the CheZ dimer (Fig. 5, lanes 2, 5, 8, 11, and 14; the molecular size of the   2. Gel electrophoresis of the purified mutant CheZ proteins. Each of the purified mutant CheZ proteins (5 g) were analyzed on a 15% SDS-polyacrylamide gel electrophoresis. Lanes 1-5 are wildtype CheZ, CheZ110LP, CheZ141FI, CheZ143DE, and CheZ145TM, respectively. Lane 6 contains the indicated molecular size markers (in kDa). monomer is 23.9 kDa (30)); in the presence of CheY but absence of AcP, higher molecular size products (up to ϳ70 kDa) were also observed (lanes 3, 6, 9, 12, and 15). However, under phosphorylating conditions there was a marked difference between the wild type and the mutants: all the mutant proteins (lanes 7, 10, 13, and 16), unlike the wild-type protein (lane 4), did not form the high molecular size oligomer with CheYϳP. (In the case of CheZ110LP, other forms of CheZ, higher than the dimer but lower than the CheZ-CheYϳP oligomer, were also observed, possibly because of its global structure defects (see below).) These results demonstrate that the mutant CheZ proteins are unable to oligomerize. When taken together with the conclusion reached earlier that all the mutants have a reduced catalytic activity, the results further suggest that the oligomerization is involved in the phosphatase activity of CheZ.
The Effect of the Mutations on the Global Structure of CheZ-To determine whether the strong effects of the point mutations on the oligomerization and the phosphatase activity of CheZ were the consequence of local effects or global perturbation of CheZ structure, we measured the susceptibility of the proteins to limited trypsin proteolysis. The rationale behind this approach was that perturbation in the structure of CheZ is expected to make the protein less compact and thereby to expose more sites to trypsin action, with a consequent faster proteolysis. As shown in Fig. 6, the mutant proteins CheZ141FI, CheZ143DE, and CheZ145TM exhibited similar proteolysis patterns as wild-type CheZ, indicative of no major structural changes in these three mutant proteins. These patterns are similar to those observed earlier in wild-type CheZ by Stock and Stock (30). In contrast, CheZ110LP was significantly less resistant to trypsin proteolysis and was almost completely degraded to small fragments already within 1.5 min of incubation with trypsin; the wild-type protein and the other mutant proteins remained almost intact during this time period. This result suggests that the 110LP mutation causes global perturbation in the structure of CheZ. DISCUSSION In this study we demonstrated that the phosphatase activity of CheZ is correlated with its ability to oligomerize upon interaction with CheYϳP. The study also confirmed the correlation between the phosphatase activity of CheZ and the CCW bias of flagellar rotation, and it provided an insight into the involvement of specific CheZ residues in the functions of the protein. These issues are discussed below.
Relation between CheZ-CheYϳP Binding, Direction of Flagellar Rotation, and Phosphatase Activity-We studied four mutant CheZ proteins that in vivo are unable to antagonize the CW causing activity of CheYϳP (28). We observed that all these mutant proteins were, on the one hand, severely impaired in their ability to dephosphorylate CheY (Fig. 3) but, on the other hand, apparently normal in their ability to bind to CheY (Fig. 4). This indicates that CheZ-CheYϳP binding is not sufficient for CheYϳP dephosphorylation. The observation that CW biased cheZ mutants are defective in the phosphatase activity of CheZ (this study) taken together with the observation that CCW biased cheZ mutants have phosphatase activity higher than wild-type CheZ (31) indicates that the direction of flagellar rotation is tightly dependent on the phosphatase activity of CheZ.
Involvement of Specific CheZ Residues in the Functions of the Protein-Three of the four mutant alleles of CheZ were clustered between residues 141 and 145. These mutant proteins were normal in binding CheYϳP (Fig. 4) but were severely impaired in their ability to oligomerize in its presence (Fig. 5). Furthermore, even conservative substitutions such as Phe 3 Ile (at position 141) or Asp 3 Glu (at position 143) were sufficient for CheZ inactivation. This suggests that the region defined by residues 141-145 plays an important role in mediating CheZ oligomerization and CheYϳP dephosphorylation, but is not necessary for the binding to CheYϳP. Indeed, this region is part of a conserved domain in CheZ of E. coli, S. typhimurium, and Pseudomonas aeruginosa (30,32,33). The substitution Leu 3 Pro (at position 110) in the fourth mutant was also a substitution of a conserved residue. This mutation had apparently a global effect on the structure of CheZ, as evident from the increased susceptibility of the protein to proteolysis (Fig. 6). Since proline residues are known to break ␣-helices, it is possible that Leu 110 is in an ␣-helix crucial for the structure of CheZ. In accordance with a global structural effect, the mutation in residue 110, unlike the mutations in residues 141-145, did not have an "all or none" effect on the oligomeric state (Fig. 5). This may be attributed to the indirect effect of this mutation on the domain responsible for the oligomerization.
Correlation between the Phosphatase Activity and the Oligomerization of CheZ-All the mutant CheZ proteins that we studied were defective in both the phosphatase activity and the oligomerization (Figs. 4 and 5), suggesting that oligomerization is required for the expression of the phosphatase activity. This suggests that the phosphatase activity of CheZ, like the oligomerization of CheZ (27), is dependent on the level of CheY phosphorylation. This is in line with an earlier observation (which was not understood at the time) that, when the level of CheY phosphorylation is very low (ϳ1%), it is essentially inde-pendent of the CheZ concentration (in the range 5-100 M) (26). This earlier observation that CheZ is not active at low levels of CheY phosphorylation, taken together with the observations of this study, suggests that the phosphatase activity is positively regulated by the level of CheY phosphorylation.
What Is the Physiological Role of CheZ Oligomerization?-A reasonable possibility is that any modulation of the level of CheY phosphorylation by chemotactic stimuli is counterbalanced by a delayed modulation of the phosphatase activity of CheZ. The delay may be caused by the oligomerization, which is presumably rate-limiting. Thus, according to this possibility, an increase in the level of CheY phosphorylation would increase the CW bias of flagellar rotation and, in parallel, would cause a relatively slow oligomerization of CheZ with a resultant increased phosphatase activity and a decreasing CW bias (Fig. 7). Conversely, a decrease in the level of CheY phosphorylation would decrease the CW bias and, in parallel, would cause a relatively slow dissociation of the CheZ oligomer with a resultant decreased phosphatase activity and increased CW bias. This proposed adaptation mechanism would ensure that the phosphorylation level is partially set back close to the prestimulus level. Such a mechanism may be part of the methylation-independent adaptation (34 -36) (see for review, Refs. 3 and 37). The notion of CheZ involvement in adaptation is well in line with earlier in vivo studies in which cheZ mutants were found to adapt to both attractants (28,35,38) and repellents ( Fig. 7 versus Fig. 10 in Ref. 39) significantly slower than wild-type bacteria. 2 Note Added in Proof-Recently Wang and Matsumura (Mol. Microbiol. (1996) in press) found, in line with our observations and suggestion that CheZ oligomerization increases its phosphatase activity, that the multimeric complex between CheZ and the short form of CheA has an increased phosphatase activity.