Previous Article | Next Article ![]()
Journal of Bacteriology, October 2007, p. 7007-7013, Vol. 189, No. 19
0021-9193/07/$08.00+0 doi:10.1128/JB.00896-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Department of Biochemistry, Colleges of Medicine and Liberal Arts and Sciences, University of Illinois, Urbana, Illinois 61801
Received 7 June 2007/ Accepted 24 July 2007
|
|
|---|
cheC adaptational defect but also generally lowered the counterclockwise flagellar rotational bias. The effect on rotational bias suggests that CheX reduced the overall levels of CheY-P in the cell and did not truly replicate the adaptational effects of CheC. Thus, CheX is not functionally redundant to CheC and, as outlined in the discussion, may be more analogous to CheZ. |
|
|---|
Bacterial motility is controlled by a modified two-component system, resulting in chemotaxis, the ability of an organism to direct its movement to more favorable environments (29). A bacterium is able to detect small changes in a gradient of a chemoeffector and to move to high levels of attractants or lower levels of repellants. The effectors are detected by transmembrane receptor proteins that regulate the activity of the cytoplasmic CheA histidine kinase aided by the coupling protein CheW (9). CheY is the main response regulator of the system and controls the rotation of the bacterium's flagella. In the chemotaxis system of the firmicute Bacillus subtilis, the binding of an attractant to the receptors activates CheA and generates higher levels of CheY-P (2, 10). CheY-P binds to the flagellar switch complex (34, 37) and induces the flagella to rotate counterclockwise, generating a smooth swimming event (22, 32). When CheA is deactivated, CheY-P levels fall and the flagella reverse, causing a tumble event. By this modulation between smooth swims and tumbles, the bacteria perform a biased random walk towards a better environment.
The chemotactic response is phasic, and a cell must adapt to an effector once the response is generated so that it is ready to respond to the next change in the gradient (4). The adaptation returns the rotational bias to a nearly even split between clockwise (CW) and counterclockwise (CCW) rotation. CheR, CheB, CheV, and CheC are proteins shown to be important in adaptation (14, 16, 17, 25, 26). The CheR methyltransferase and CheB methylesterase add and remove methyl groups from the receptors (known as methylated chemotaxis proteins), causing adaptation. The involvement of CheV and CheC has been studied, but the mechanisms of their involvement are not clear.
Aspartyl phosphate phosphatases terminate the signal of phosphorylated response regulator proteins in some two-component signal transduction systems (12, 24). CheZ is the phosphatase of the response regulator CheY-P in the extensively studied Escherichia coli chemotaxis system (39). However, the proteins CheC, FliY, and CheX have been shown to comprise a novel family of CheY-P phosphatases found in the Archaea, Firmicutes, and various other bacteria (32). This family of phosphatases share a consensus sequence (D/S-X3-E-X2-N-X21/22-P) (31) with four conserved residues thought to form the phosphatase active site (21). The functions of CheC and FliY have previously been studied in Bacillus subtilis chemotaxis (30, 31), while CheX has been examined in the spirochetes (11, 19). FliY is a component of the flagellar switch complex and is the major phosphatase in B. subtilis. CheC forms a heterodimer with CheD and is involved in adaptation. CheX forms a homodimer (21) with one active site per monomer (versus two for CheC and FliY) and is not found in the B. subtilis system.
Many bacteria, such as Bacillus halodurans, have chemotaxis systems closely resembling the B. subtilis system but that contain CheX in addition to CheC and FliY (21, 32, 33). The function of CheX in this three-phosphatase system has not been studied previously. Here we examined the in vitro CheX interaction with CheY-P and heterologously expressed CheX in B. subtilis to observe the effects of this third phosphatase on chemotaxis.
|
|
|---|
|
View this table: [in a new window] |
TABLE 1. Strains and plasmids used in this study
|
Mutations in B. halodurans cheX were generated via the QuikChange mutagenesis kit (Stratagene) in vector pTM74. We subcloned these mutants into pDR67, creating variants of pTM53; thus, all four cheX point mutants are under the control of the optimized RBS. We then transformed these plasmids into OI3135 by selecting for Cmr and amyE, creating strains OI4222, OI4223, OI4224, and OI4225.
We created pTM0 and pTM75 through PCR amplification of B. subtilis fliY and B. halodurans cheX, adding a 5' BamHI site and 3' EcoRI (fliY) or NotI (cheX) site. These DNA fragments were cloned into pGEX-6P-2 for the generation of glutathione S-transferase (GST)-tagged proteins in E. coli strain BL21.
Protein purification. All purified proteins are the B. subtilis type, except for CheX, which was cloned from B. halodurans. Proteins used in this study were purified as previously described (31). Briefly, GST fusions were expressed from BL21 lysates expressing proteins from pGEX-6P-2 variant plasmids. Five-milliliter GSTrap columns (GE Healthcare) were used with an AKTA Prime system (GE Healthcare) for purification as the manufacturer prescribed. All proteins were dialyzed into TKMD buffer (50 mM Tris, pH 8.0, 50 mM KCl, 5 mM MgCl2, 0.1 mM dithiothreitol, 10% [vol/vol] glycerol) for storage at –80°C.
GST pull down. The GST pull-down assay was performed essentially as described previously (20), except acetyl phosphate was selectively added to the incubations and wash buffers to generate GST-CheY-P.
Phosphate release assay. The phosphate release assay was performed essentially as described previously (6, 20, 31), except that 3 µM CheY was used. This assay measures inorganic phosphate (Pi) released from CheY at steady-state conditions. The dephosphorylation was observed over a range of phosphatase concentrations and was averaged over three to five experiments. Specific activities for each phosphatase were determined from the slope of the initial linear portion. Errors were calculated as standard errors of the means (SEM). The Pi was measured spectrophotometrically by using the EnzChek phosphate assay kit (Molecular Probes).
Swarm plate assay. The strains examined were inoculated onto tryptone swarm plates (1% [wt/vol] tryptone, 0.5% [wt/vol] NaCl, 0.27% [wt/vol] agar, 1 mM IPTG [isopropyl-ß-D-thiogalactopyranoside]) and incubated at 37°C for 4 h. The bacteria grow and metabolize the nutrients in the medium, creating a gradient detectable by the chemotaxis system. The bacteria thus form dense swarms radiating out from the point of inoculation. After the incubation, the diameters of the swarm are measured and compared to the diameter of the swarm formed by the wild type as a measure of chemotactic ability.
Capillary assay. The capillary assay was performed as described previously (1, 40) to measure chemotactic ability. Briefly, cells were grown to late exponential phase in minimal medium and tested for their ability to sense a gradient created by an attractant-filled capillary tube inserted into a pool of bacteria. Cells were grown in the presence of 1 mM IPTG to induce the expression of cheX and cheC under the control of the Pspac promoter. The assay was performed twice to insure reproducibility.
Tethered cell assay. The tethered cell assay was performed to determine flagellar rotational direction versus time essentially as described previously (5, 15). Bacterial cells were adhered by a flagellum to a microscope coverslip by an anti-flagellar antibody, and the rotational direction (CW or CCW) was tracked. After 2 min, 0.5 mM asparagine was added. Data averaged over a population resulted in a probability of CCW (smooth swim) rotation as an indication of CheY-P levels (CCW indicates high CheY-P, and CW indicates low CheY-P). Again, CheX expression was induced by the addition of 1 mM IPTG during growth.
|
|
|---|
![]() View larger version (57K): [in a new window] |
FIG. 1. GST-CheY pull down of FliY and CheX. FliY and CheX were retained after three washes only in the presence of both GST-CheY and acetyl phosphate. GST was used as a negative control.
|
![]() View larger version (13K): [in a new window] |
FIG. 2. Phosphate release assay comparing CheX, FliY, and CheC phosphatase activity. The rate of Pi release from 3 µM CheY was measured at various concentrations of CheX ( ), FliY ( ), and CheC ( ). Each value is the average of three to five trials; error bars represent SEM. The slope of the initial linear portion was used to calculate specific activities of 26 ± 2, 19 ± 1, and 1.7 ± 0.1 µM Pi min–1 µM–1 phosphatase for CheX, FliY, and CheC, respectively.
|
cheC (data not shown). Further, B. halodurans CheX was shown to effectively bind (Fig. 1) and dephosphorylate (Fig. 2) B. subtilis CheY-P.
The cheX gene was expressed under the control of the B. subtilis cheC RBS and with an optimized RBS (plasmids pTM50 and pTM53, respectively) in the
cheC strain OI3135. The B. subtilis chemotactic wild-type strain OI1085 harboring pTM53 was also observed. Western blots with an anti-FLAG antibody showed that expression levels of CheX-FLAG (OI4233) were approximately three times those of CheC-FLAG when both genes were under the control of the pDR67 Pspac promoter and the cheC RBS (data not shown). Expression levels of CheX from the optimized RBS (OI4219) were three times higher than those from the cheC RBS (OI4233).
The swarm plate assay was employed to test the effect of CheX on the B. subtilis chemotaxis system (31). With the cheC RBS, the cheX gene was able to partially complement the
cheC mutant up to 40% of the wild-type swarm size (Fig. 3A). This level was increased to 80% complementation when expressed with the optimized RBS (Fig. 3A). The same complementation effect was found for Clostridium acetobutylicum cheX (data not shown). Strain OI4218 expressing CheX in the wild type formed a swarm ring larger than that of the wild type (Fig. 3A).
![]() View larger version (29K): [in a new window] |
FIG. 3. Chemotaxis assays for strains expressing CheX. (A) Diameters of swarms on tryptone swarm plates expressed relative to that of the wild type (WT) (OI1085) strain. Bars represent the average of three swarms, and error bars represent SEM. (B) Capillary accumulations expressed relative to accumulation by the wild type. The experiment was performed in duplicate at the three proline concentrations indicated. The strains tested were OI1085 (wild type) (black bars), OI3165 ( cheC cheC+) (white bars), OI4218 (cheX+) (gray bars), OI4219 ( cheC cheX+) (vertically striped bars), and OI3135 ( cheC) (diagonally striped bars).
|
cheC strain. The resulting cheX-S87A, cheX-E91A, cheX-N94A, and cheX-P116A mutant strains were tested by swarm plate assay; none of these mutants had swarms much greater than that of the
cheC strain (Fig. 3A). By Western blotting, the four mutant cheX strains all showed CheX expression equal to that of the wild-type CheX (OI4219) (data not shown). Thus, all four conserved active site residues are crucial for CheX function.
To confirm the swarm assay results, the capillary assay was employed. Strain OI4219 was tested against the B. subtilis wild type,
cheC, and OI3165 (Fig. 3B). CheX was able to partially complement the
cheC strain, confirming the swarm plate data. However, when expressed in the wild type (OI4218), CheX inhibited chemotaxis.
CheX effect on flagellar rotation.
Since a chemotaxis phenotype was observed for the strains expressing CheX, the flagellar rotation was observed to see the effect of CheX on the chemotaxis system. The OI3135 (
cheC) strain had a near wild-type-level prestimulus bias and attractant addition peak, but the strain did not return to the prestimulus rotational bias after the addition of attractant, thus showing incomplete adaptation (25, 31). The OI4219 (
cheC cheX+) strain, however, was able to adapt after the addition of asparagine (Fig. 4). Therefore, CheX restores flagellar rotation to prestimulus levels after adaptation in the
cheC strain. However, the expression of CheX also lowered the pre- and poststimulus biases compared to that of the wild type (55% wild type versus 35% OI4219) (Fig. 4) as well as the attractant addition peak bias.
![]() View larger version (17K): [in a new window] |
FIG. 4. Effect of CheX expression on flagellar rotation. The strains tested were OI1085 (thick line), OI4219 (medium line), and OI3135 (thin line). The plus symbol on the graph represents the addition of 0.5 mM asparagine.
|
|
|
|---|
We have shown that the putative active site residues of CheC are necessary both for binding and dephosphorylating CheY-P (T. Muff and G. Ordal, unpublished data) and that these residues are necessary for CheX function in vivo (Fig. 3A). Since the four consensus residues (Ser, Glu, Asn, and Pro) are conserved among the members of the CheC-FliY-CheX family, they likely have the same function in each protein and form the phosphatase active sites of these proteins. Peculiarly, FliY cannot bind (30) or effectively dephosphorylate (32) CheY-P without its N-terminal CheY-P binding domain; though CheC and CheX, lacking this binding domain, are still able to bind and dephosphorylate CheY-P (T. Muff and G. Ordal, unpublished data) (Fig. 1). This result suggests that the active sites of CheC and CheX have higher binding affinities for CheY-P than do the active sites of FliY.
The ability of CheX to nearly complement the
cheC strain in both the swarm and capillary assays (Fig. 3A and B) suggests that these proteins not only share the in vitro CheY-P phosphatase ability but also may function similarly in the chemotactic cell. However, when CheX was expressed at levels found to nearly complement
cheC and the flagellar rotation was observed (Fig. 4), a different picture emerged. While CheX expression did allow the cells to adapt back to the prestimulus bias after attractant was added (removing the
cheC defect), the overall bias and excitation peak were reduced significantly. The CheX phosphatase action must reduce CheY-P levels in all phases of chemotactic signaling as opposed to CheC, which seems to be active only poststimulus. The restoration of adaptation by CheX in the
cheC strain then likely resulted from the overall reduction of CheY-P levels by CheX and caused the observed degree of complementation. Thus, CheC and CheX, though sharing CheY-P phosphatase activity, must not have completely redundant functions in chemotaxis.
Strain OI4218 (OI1085 cheX+) formed a swarm ring larger than that of the wild-type strain (Fig. 3A) but had inhibited chemotaxis in the capillary assay (Fig. 3B). Both of these results can likely be explained by the low CCW rotational bias resulting from CheX expression. A tumbly bias (predominantly CW flagellar rotation) reflects a reduced length of smooth swimming runs. Tumbly mutant strains have been shown to migrate better in the swarm plate medium than do other nonchemotactic mutants (38). The tumbly bias of strain OI4218 may allow it to migrate through the swarm medium faster than the wild type. However, tumbly free-swimming cells (as in the capillary assay) are at a disadvantage compared to the wild type because they have shorter runs.
CheC and FliY each have novel protein interactions (Fig. 5); CheC binds CheD (8, 26) and FliY associates with other flagellar proteins (3, 34). However, CheX has only been shown to dimerize (19, 21) and associate with CheY-P (Fig. 2) and CheA (27). The interaction with CheA is intriguing since CheZ, which also has limited protein interactions, was shown to be localized to the receptors by an alternatively translated version of CheA known as CheA-short (7). A CheX interaction with CheA may parallel the localization of CheZ by CheA-short. With CheX being a strong phosphatase and localized to the receptors (Fig. 5), it may be responsible for preventing the formation of a CheY-P gradient across the cell as was functionally proposed for CheZ (18, 36). Thus, CheX may be functionally analogous to CheZ. Though B. subtilis and B. halodurans have essentially identical chemotaxis components, B. subtilis evidently has evolved to a state where the requirement for CheX has been overtaken by CheC and FliY.
![]() View larger version (15K): [in a new window] |
FIG. 5. Diagram of the chemotaxis phosphatases. The binding of an attractant to the methylated chemotaxis proteins (MCPs) induces a conformational change across the cell membrane. This change causes the autophosphorylation of CheA (A) and the subsequent transfer of the phosphoryl group to CheY (Y). CheY binds to FliM in the flagellar switch and promotes CCW flagellar rotation. CheY is dephosphorylated by FliY at the flagella or by CheX (X) and CheC (C), perhaps at the receptor complex. D, CheD.
|
This work was supported by National Institutes of Health grant RO1GM54365 to GWO.
Published ahead of print on 3 August 2007. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»