Advertisement

Bacillus subtilis Hydrolyzes CheY-P at the Location of Its Action, the Flagellar Switch*

  • Hendrik Szurmant
    Affiliations
    Department of Biochemistry, Colleges of Medicine and Liberal Arts and Sciences, University of Illinois, Urbana, Illinois 61801
    Search for articles by this author
  • Michael W. Bunn
    Affiliations
    Department of Biochemistry, Colleges of Medicine and Liberal Arts and Sciences, University of Illinois, Urbana, Illinois 61801
    Search for articles by this author
  • Vincent J. Cannistraro
    Affiliations
    Department of Biochemistry, Colleges of Medicine and Liberal Arts and Sciences, University of Illinois, Urbana, Illinois 61801
    Search for articles by this author
  • George W. Ordal
    Correspondence
    To whom correspondence should be addressed. Tel.: 217-333-9098 or 217-333-0268; Fax: 217-333-8868;
    Affiliations
    Department of Biochemistry, Colleges of Medicine and Liberal Arts and Sciences, University of Illinois, Urbana, Illinois 61801
    Search for articles by this author
  • Author Footnotes
    * This work is supported by the National Institutes of Health Grant RO1 GM54365. 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.
Open AccessPublished:August 14, 2003DOI:https://doi.org/10.1074/jbc.M306180200
      In this report we show that in Bacillus subtilis the flagellar switch, which controls direction of flagellar rotation based on levels of the chemotaxis primary response regulator, CheY-P, also causes hydrolysis of CheY-P to form CheY and Pi. This task is performed in Escherichia coli by CheZ, which interestingly enough is primarily located at the receptors, not at the switch. In particular we have identified the phosphatase as FliY, which resembles E. coli switch protein FliN only in its C-terminal part, while an additional N-terminal domain is homologous to another switch protein FliM and to CheC, a protein found in the archaea and many bacteria but not in E. coli. Previous E. coli studies have localized the CheY-P binding site of the switch to FliM residues 6–15. These residues are almost identical to the residues 6–15 in both B. subtilis FliM and FliY. We were able to show that both of these proteins are capable of binding CheY-P in vitro. Deletion of this binding region in B. subtilis mutant fliM caused the same phenotype as a cheY mutant (clockwise flagellar rotation), whereas deletion of it in fliY caused the opposite. We showed that FliY increases the rate of CheY-P hydrolysis in vitro. Consequently, we imagine that the duration of enhanced CheY-P levels caused by activation of the CheA kinase upon attractant binding to receptors, is brief due both to adaptational processes and to phosphatase activity of FliY.
      Peritrichously flagellated bacteria such as Escherichia coli and Bacillus subtilis move toward favorable environments by altering the rotation of their flagella between counterclockwise (CCW)
      The abbreviations used are: CCW
      counterclockwise
      GST
      glutathione S-transferase
      IPTG
      isopropyl-1-thio-β-d-galactopyranoside.
      1The abbreviations used are: CCW
      counterclockwise
      GST
      glutathione S-transferase
      IPTG
      isopropyl-1-thio-β-d-galactopyranoside.
      and clockwise (
      • Larsen S.H.
      • Reader R.W.
      • Kort E.N.
      • Tso W.W.
      • Adler J.
      ). CCW rotation propels the bacteria forward, while clockwise rotation causes the bacteria to change their swimming direction (
      • Berg H.C.
      • Brown D.A.
      ). Rotational direction is controlled by a two component signal transduction system, consisting of the histidine kinase CheA and the main response regulator CheY, as well as many other proteins (for review see Refs.
      • Macnab R.M.
      ,
      • Stock J.B.
      • Surette M.G.
      ,
      • Aizawa S.I.
      • Zhulin I.B.
      • Marquez-Magana L.
      • Ordal G.W.
      ). Some of these proteins are nearly universal among motile bacteria, and some are not. For instance the CheY-P phosphatase CheZ can only be found in some proteobacteria like E. coli (
      • Aizawa S.I.
      • Zhulin I.B.
      • Marquez-Magana L.
      • Ordal G.W.
      ). Conversely, B. subtilis expresses three proteins, CheC, CheD, and CheV, that are not found in E. coli (
      • Bischoff D.S.
      • Ordal G.W.
      ). However, they are widespread, at least in modified forms, throughout the bacteria. CheC and CheD are also found in the archaea (
      • Kirby J.R.
      • Kristich C.J.
      • Saulmon M.M.
      • Zimmer M.A.
      • Garrity L.F.
      • Zhulin I.B.
      • Ordal G.W.
      ). Recently, roles for CheD and CheV have been identified (
      • Karatan E.
      • Saulmon M.M.
      • Bunn M.W.
      • Ordal G.W.
      ,
      • Kristich C.J.
      • Ordal G.W.
      ), while the function of CheC has remained elusive. All chemotaxis proteins control directly or indirectly the levels of phosphorylation of CheY. Phosphorylated CheY (CheY-P) interacts with the flagellar switch consisting of FliM, FliN, and FliG in E. coli. FliM has been shown to directly interact with CheY-P and induce change of direction of flagellar rotation (
      • Welch M.
      • Oosawa K.
      • Aizawa S.
      • Eisenbach M.
      ). B. subtilis has the orthologs FliM and FliG, but instead of FliN has FliY, which is homologous to FliN only in its C terminus (
      • Bischoff D.S.
      • Ordal G.W.
      ). Its N-terminal domain is homologous to CheC (27% of residues identical, 48% similar) (Fig. 1A) (
      • Kirby J.R.
      • Kristich C.J.
      • Saulmon M.M.
      • Zimmer M.A.
      • Garrity L.F.
      • Zhulin I.B.
      • Ordal G.W.
      ). The fact that a Salmonella typhimurium fliN mutant could be partially complemented by B. subtilis fliY indicates that these two proteins share a common function as a component of the flagellar switch (
      • Bischoff D.S.
      • Ordal G.W.
      ).
      Figure thumbnail gr1
      Fig. 1Comparison of E. coli and B. subtilis switch proteins.A, diagram of switch protein homology. Black indicates FliM homology; white indicates CheC homology; light gray indicates FliN homology; and dark gray indicates FliG homology. B, alignment of N-terminal amino acids of E. coli FliM with B. subtilis FliM and FliY. The 29 most N-terminal amino acids are shown. The CheY-P binding region in E. coli FliM is underlined.
      Little information regarding the role of the N-terminal domain has been reported until now. Information on switch proteins requires detailed studies as simple deletion mutants are not flagellated and so reveal little information (
      • Yamaguchi S.
      • Aizawa S.
      • Kihara M.
      • Isomura M.
      • Jones C.J.
      • Macnab R.M.
      ). Interestingly, the Bacillus cereus genome encodes for a protein termed FliY, and its N terminus is homologous to that of B. subtilis FliY; however, it is missing the domain homologous to FliN (
      • Celandroni F.
      • Ghelardi E.
      • Pastore M.
      • Lupetti A.
      • Kolsto A.B.
      • Senesi S.
      ). A null mutant shows swarming defects but is flagellated and motile (and so this may not be a true flagellar switch protein (
      • Kirby J.R.
      • Kristich C.J.
      • Saulmon M.M.
      • Zimmer M.A.
      • Garrity L.F.
      • Zhulin I.B.
      • Ordal G.W.
      ,
      • Senesi S.
      • Celandroni F.
      • Salvetti S.
      • Beecher D.J.
      • Wong A.C.
      • Ghelardi E.
      ).
      The current study investigated the function of the N-terminal domain of FliY. Interestingly FliY shares homology with B. subtilis and E. coli FliM in its residues 6–15 (Fig. 1B), a significant finding because these residues have been described as the CheY-P binding area for E. coli FliM (
      • Bren A.
      • Eisenbach M.
      ). We show that FliM and FliY, the two flagella switch proteins in B. subtilis containing this putative binding region, are capable of binding phosphorylated CheY in vitro. Mutants deleted for their CheY-P binding region showed opposite phenotypes in vivo as seen in the tethered cell assay, confirming FliM's function in inducing CCW rotation of flagella. In vitro FliY substantially reduced the half-life of CheY-P. We conclude that FliY serves at least two purposes: 1) to dephosphorylate CheY-P and 2) to be a structural component of the flagellar switch.

      EXPERIMENTAL PROCEDURES

      Chemicals, Columns, Enzymes, and Growth Media—All chemicals were of reagent grade. All protein purification columns as well as PreScission protease and [γ-32P]ATP were purchased from Amersham Biosciences. Media were obtained from Difco. Luria-Bertani medium is 1% tryptone, 0.5% yeast extract, and 1% NaCl.
      Plasmid and Strain Construction—The strains and plasmids used are listed in Table I. All B. subtilis strains constructed were derived from the Che+ strain OI1085.
      Table IStrains and plasmids used in this study
      StrainRelevant phenotype or commentReference or source
      B. subtilis
       OI1085Che+, trpF7 hisH2 metC133(
      • Ullah A.H.
      • Ordal G.W.
      )
       OI2852fliY1::cat(
      • Bischoff D.S.
      • Ordal G.W.
      )
       OI2597fliM1::cat(
      • Zuberi A.R.
      • Bischoff D.S.
      • Ordal G.W.
      )
       OI3941ΔfliY2This work
       OI3954ΔfliM2This work
       OI3957Δ(fliY cheY)201This work
       OI4104ΔfliM2 amyE5720::fliM3This work
       OI4105ΔfliM2 amyE5720::fliM4This work
       OI3941ΔfliY2 amyE5720::fliY3This work
       OI4106ΔfliY2 amyE5720::fliY4This work
       OI4107Δ(fliY cheY)201 amyE5720::fliY4This work
      E. coli
       BL21Protease-deficient expression hostAmersham Biosciences
       TG1Cloning hostAmersham Biosciences
       GJ1158Host for pT7-derived plasmids; induction by NaCl(
      • Bhandari P.
      • Gowrishankar J.
      )
      PlasmidsCommentReference or source
      pGEX-6P-2GST-tag expression plasmidAmersham Biosciences
      pT7-5Expression plasmid(
      • Tabor S.
      • Richardson C.C.
      )
      pT7-6Expression plasmid(
      • Tabor S.
      • Richardson C.C.
      )
      pT7-7Expression plasmid(
      • Tabor S.
      • Richardson C.C.
      )
      pAZ249pT7-5::fliM3 [wild type](
      • Zuberi A.R.
      • Bischoff D.S.
      • Ordal G.W.
      )
      pHS114pT7-5::fliM4 [fliMΔ6-15]This work
      pDB32pT7-6::fliY3 [wild type](
      • Bischoff D.S.
      • Ordal G.W.
      )
      pHS115pBluescript SK::fliY4 [fliYΔ6-15]This work
      pDB6pT7-7::`fliM-fliQ'(
      • Bischoff D.S.
      • Ordal G.W.
      )
      pBluescript SK-Cloning plasmidStratagene
      pDR67amyE integration plasmid with IPTG-inducible pSpac promoter(
      • Ireton K.
      • Rudner D.Z.
      • Siranosian K.J.
      • Grossman A.D.
      )
      pEB112B. subtilis-E. coli shuttle vector(
      • Leonhardt H.
      • Alonso J.C.
      )
      pHS101pGEX-6P-2::cheA1 [wild type]This work
      pHS102pGEX-6P-2::cheY1 [wild type]This work
      pHS103pBluescript SK-::`flgE-fliY'This work
      pHS104pBluescript SK-::`flgEfliM-fliY'This work
      pHS105pEB112::`flgEfliM-fliY'This work
      pHS106pT7-7::`fliMfliY-fliQ'This work
      pHS107pT7-7:: `fliMfliYΔcheY-fliQ'This work
      pHS108pEB112::`fliMfliY-fliQ'This work
      pHS109pEB112::`fliMfliYΔcheY-fliQ'This work
      pHS110pDR67::fliY3This work
      pHS111pDR67::fliY4This work
      pHS112pDR67::fliM3This work
      pHS113pDR67::fliM4Δ6-15This work
      pHS116pGEX6P-2-cheY2 [cheYD54A]This work
      pDB21pEB112-cheY1(
      • Bischoff D.S.
      • Ordal G.W.
      )
      pDB56pEB112-cheY2(
      • Bischoff D.S.
      • Bourret R.B.
      • Kirsch M.L.
      • Ordal G.W.
      )
      ΔfliM, ΔfliY, and ΔfliYΔcheY strains were constructed by gene conversion. Plasmids containing at least 500 flanking base pairs to fliM, fliY, or fliYcheY were subjected to long PCR excluding most of the appropriate genes in-frame. PCR products were ligated, and the corresponding inserts, deleted for the appropriate gene(s), were subcloned into pEB112. The ΔfliM construct was transformed into strain OI2597 selecting for KanR and screening for CmS, followed by segregation of plasmid indicated by loss of KanR, to result in strain OI3954. The ΔfliY and ΔfliYΔcheY constructs were transformed into OI2852, and transformants were treated the same way, resulting in strains OI3941 and OI3957, respectively.
      fliMΔ6–15 and fliYΔ6–15 were constructed by long PCR of the plasmids pAZ249 (resulting in pHS114, overexpression plasmid for fliMΔ6–15) and pDB6, respectively. Wild type and mutant genes were amplified with primers introducing 5′-SphI and 3′-BglII sites and cloned into pDR67. These constructs were linearized with AatII and transformed into OI3954 for fliM and fliMΔ6–15, into OI3941 for fliY and fliYΔ6–15, and into OI3957 for fliYΔ6–15 selecting for CmRamyE. This procedure gave strains OI4104, OI4105, OI3942, OI4106, and OI4107, respectively. fliYΔ6–15 was subcloned into HindIII and EcoRI sites of pBluescriptSK, resulting in pHS115, the expression plasmid for the corresponding protein.
      Glutathione S-transferase (GST) fusion constructs of cheA, cheY, and cheYD54Awere made by amplifying the three genes, introducing a 5′-BamHI excluding the start codon and 3′-EcoRI sites. The PCR construct was inserted into the respective sites of pGEX-6P-2. For cheA any linker amino acid codons not essential for the PreScission protease cleavage site were removed by long PCR, leaving only glycine and proline codons at the 5′ end following the cleavage site.
      Protein Overexpression and Purification—Proteins were overexpressed in E. coli strain BL21 for IPTG-inducible plasmids and in GJ1158 for T7 plasmids. Strains were grown in LB until reaching A600 = 0.6, and protein expression was induced with 1 mm IPTG for BL21 strains and 0.3 m NaCl for GJ1158 strains. Induction times and temperatures were 4 h at 37 °C for GST-CheA and for wild type and mutant FliY, 16 h at 25 °C for wild type and mutant FliM, and 48 h at 16 °C for wild type and mutant GST-CheY.
      CheY and CheA were purified as GST-tagged proteins over a 2 ml glutathione-Sepharose column by gravity flow, essentially as specified by the supplier (Amersham Biosciences). The purified protein was cleaved by PreScission protease, again as specified by the supplier (Amersham Biosciences). Following dialysis, protease and GST tag were removed by applying the reaction over another glutathione column, producing essentially pure CheA or CheY. This was dialyzed into TKMD buffer (50 mm Tris-HCl, 5 mm MgCl2, 50 mm KCl, 0.2 mm dithiothreitol, and 10% glycerol, pH 8.0) and brought to the appropriate concentration. Cells expressing fliY were lysed and subjected to Polymine P treatment as described for T7 RNA polymerase except that FliY was precipitated at an ammonium sulfate saturation of 55% and redissolved in buffer 1 (50 mm Tris-HCl, pH 8, 1 mm EDTA, 1 mm DTT, 5% glycerol, 20 mm NaCl, 20 mg/liter phenylmethylsulfonyl fluoride) (
      • Davanloo P.
      • Rosenberg A.H.
      • Dunn J.J.
      • Studier F.W.
      ). The protein solution was dialyzed against three changes of this buffer. The extract was applied to a 20 ml DEAE column previously equilibrated in buffer 1, and a linear gradient from 20 to 500 mm NaCl was applied. Peak-fractions eluting around 175 mm NaCl contained FliY and were concentrated to 5 ml. Solid ammonium sulfate was added to a final concentration of 1 m. This sample was applied to an octyl-Sepharose column (20 ml) that had been equilibrated in buffer 2 (50 mm Tris-HCl, pH 7.7, 1 m (NH4)2SO4, 1 mm EDTA). The column was washed with 100 ml of that buffer. A 200 ml descending linear gradient from 1 to 0 m (NH4)2SO4 and a simultaneous ascending gradient from 0 to 80% ethylene glycol in buffer 2 were then applied at 0.2 ml/min. FliY eluted toward the very end of this gradient. Peak fractions were dialyzed against three changes of TKMD buffer and concentrated to desired volume. Following this procedure, FliY is essentially pure as observed by Coomassie Blue staining of a SDS-PAGE gel.
      CheY Phosphorylation—Acetyl-32P was synthesized as described (
      • McCleary W.R.
      • Stock J.B.
      ). 200 pmol of CheY, GST-CheY, or GST-CheYD54A were incubated in the presence of 20 mm acetyl-32P for 30 s to 20 min in a total volume of 15 μl buffer 3 (TKMD, pH 7.5, without glycerol). Reactions were stopped by adding the same volume of 2× SDS buffer (100 mm Tris/HCl, pH 6.8, 20% glycerol, 10% mercaptoethanol, 4% SDS) containing 100 mm EDTA. 10 μl of these samples were subjected to SDS-PAGE and phosphorimaging
      GST Pull-down Assay—All steps were performed at 4 °C. Unless noted otherwise, centrifugation speed was 500 × g. 50 μl of glutathione beads were placed into a 1.5-ml spin column and centrifuged for 1 min. The beads were washed three times with buffer 3. 200 μl of primary protein (GST, GST-CheY, or GST-CheYD54A) at a concentration of 0.25 mg/ml was added to the resin and incubated for 45 min, inverting constantly. The column was washed three times, and E. coli extract expressing B. subtilis FliM, FliMΔ6–15, FliY, or FliYΔ6–15 at a concentration around 0.25 mg/ml (as estimated by SDS-PAGE and Coomassie Blue staining) was added either in the presence or in the absence of 20 mm acetyl phosphate. The samples were incubated for another 45 min and washed three times with buffer 3, containing 20 mm acetyl phosphate for the appropriate samples. 50 μl of 2× SDS buffer was added to the resin, and the spin column was heated to 90 °C for 10 min. The resin was spun down at 1000 × g for 1 min, and the supernatant was diluted to 100 μl. The samples were subject to SDS-PAGE, and the binding of primary and secondary protein was confirmed by Coomassie Blue staining and Western blot analysis.
      Western Blot Analysis for FliY and FliM—Western blots were performed essentially as described for McpB (
      • Kirby J.R.
      • Niewold T.B.
      • Maloy S.
      • Ordal G.W.
      ). The dilution for anti-FliM antibody was 1:20,000, and the dilution for anti-FliY antibody was 1:50,000.
      Tethered Cell Assay—The tethered cell assay was essentially performed as described (
      • Kirby J.R.
      • Saulmon M.M.
      • Kristich C.J.
      • Ordal G.W.
      ); however, FliY and FliM expression was induced with 0.1 mm IPTG 3 h after starting the culture. Each experiment depicts the average bias of a population of 20 cells.
      Dephosphorylation Assay—25 μm CheA was phosphorylated by incubation with [γ-32P]ATP (33 μm, 12.5 μCi/μl) for 30 min in a total volume of 120 μl of TKMD. The remaining [γ-32P]ATP was removed by desalting using a Bio-Rad Micro-Bio-Spin column. The resulting CheA-32P was added to premixed CheY-FliY solutions, resulting in a final concentration of 20 μm CheY, 10 μm CheA-P, varying concentrations of FliY as indicated, and 5 mm cold ATP in TKMD. Aliquots were taken at the indicated times by pipetting 10 μl of each reaction into 10 μl of 2× SDS buffer containing 100 mm EDTA, and the proteins were separated on a 12% SDS-PAGE. The gel was exposed to a phosphorimaging screen and developed using a Storm 860 PhosphorImager from Amersham Biosciences. All experiments were done at least in duplicate.

      RESULTS

      CheY and GST-CheY Get Phosphorylated by Acetyl Phosphate in Vitro—Acetyl phosphate has been described as a phosphodonor for E. coli CheY (
      • McCleary W.R.
      • Stock J.B.
      ). To show that it also phosphorylates B. subtilis CheY we incubated CheY, GST-CheY, and GST-CheYD54A, a putative phosphorylation site mutant with acetyl-32P. Reactions were stopped after 20 min and were subjected to SDS-PAGE and phosphorimaging (Fig. 2). Only CheY and GST-CheY but not the phosphorylation site mutant showed phosphorylation, indicating that acetyl phosphate is indeed a phosphodonor for B. subtilis CheY and that residue Asp-54 is the site of phosphorylation.
      Figure thumbnail gr2
      Fig. 2CheY and GST-CheY phosphorylation by acetyl phosphate. Shown is a phosphorimage of CheY (lane 1), GST-CheY (lane 2) and GST-CheYD54A (lane 3), which were incubated with acetyl-32P for 20 min.
      B. subtilis FliY and FliM Bind CheY in the Presence of Acetyl Phosphate in Vitro—In order to test whether FliY and FliM are capable of binding CheY, we used a GST-pull down assay. GST, GST-CheY, and GST-CheYD54A were bound to glutathione beads and incubated with protein extracts from FliY or FliM overexpression strains. The experiment was performed in the presence or absence of acetyl phosphate. Eluted protein was subjected to SDS-PAGE and analyzed by Coomassie Blue staining (data not shown) and immunoblotting (Fig. 3). Coomassie Blue staining verified that all of the lanes contained comparable amounts of GST or wild type and mutant GST-CheY. Immunoblotting showed that FliY (Fig. 3A) and FliM (Fig. 3B) binding could be detected only in the presence of acetyl phosphate and only for GST-CheY but not GST-CheYD54A, indicating that both proteins bind CheY in its phosphorylated state.
      Figure thumbnail gr3
      Fig. 3GST pull-down assay of FliM and FliY with GST-CheY in the presence and absence of acetyl phosphate as indicated.A and B, primary protein was purified GST (lanes 1 and 2), GST-CheY (lanes 3–5), or GST-CheYD54A (lanes 6 and 7). Secondary protein was extract of E. coli GJ1158-expressing FliY (A) or FliM (B). Negative control was GJ1158 extract without expression plasmid as secondary protein (lane 5). Lane 8 contained purified FliY or FliM extract for reference. C and D, primary protein was purified GST (lanes 1 and 2) or GST-CheY (lanes 3–5). Secondary protein was FliYΔ6–15 (C) or FliMΔ6–15 (D). Negative control was GJ1158 extract without expression plasmid as secondary protein (lane 5). Lane 6 contained purified FliYΔ6–15 or FliMΔ6–15 for reference. Protein was visualized immunologically.
      The N-terminal Amino Acid Residues 6–15 in Both FliY and FliM Are Important for CheY-P Interaction—To verify that N-terminal residues 6–15 in FliY and FliM are indeed important for CheY-P binding, as is the case for E. coli FliM (
      • Bren A.
      • Eisenbach M.
      ), we overexpressed fliYΔ6–15 and fliMΔ6–15 in E. coli. Both of these mutant proteins were subjected to the same binding assay as described above. Neither FliYΔ6–15 nor FliMΔ6–15 were capable of binding CheY-P to detectable levels either in the absence or presence of acetyl phosphate (Fig. 3, C and D, respectively), implying that these residues are important for CheY-P binding as predicted.
      N-terminal Deletions in FliM and FliY Promote Opposite Phenotypes in the Tethered Cell Assay in Vivo—In order to address the purpose of the interaction of FliY and FliM with CheY-P in vivo, we constructed two strains with in-frame deletions of most of fliM and fliY, respectively. Wild type fliM and fliY or the binding site mutants fliMΔ6–15 and fliYΔ6–15 were added back into the amyE locus of the appropriate strain under an IPTG-inducible promoter. Expression levels of wild type and mutant fliM and fliY were similar to levels in wild type (OI1085) in the presence of 0.1 mm IPTG as observed by immunoblotting (not shown). In the presence of 0.1 mm IPTG in swarm plates, the fliM+ and fliY+ alleles complemented the respective null mutants to OI1085 swarm diameters but the fliMΔ6–15 and fliYΔ6–15 alleles did not (not shown). Tethered cell assays were performed to explore for the effect of these mutant alleles and to identify the CCW rotational bias of the flagella. As expected, the fliMΔ6–15 mutant showed a very clockwise bias and no response to the addition or removal of the attractant asparagine (Fig. 4A), indicating that the role of CheY-P interaction with FliM is to promote an increase of smooth swimming behavior. Surprisingly, the fliYΔ6–15 mutant produced the opposite phenotype, a very high CCW bias without notable response to the addition and removal of asparagine (Fig. 4B). In order to address the possibility that the mutation had created a lock-in phenotype only capable of rotating CCW, we deleted cheY in that strain. If a lock-in phenotype had been created, CCW behavior should have been independent of the presence of CheY. This strain however produced a very low CCW bias (Fig. 4C), indicating that CheY is necessary to promote the high bias in the fliY mutant strain. To test whether CheY or CheY-P is responsible for the high CCW bias of the fliY mutant strain, we added the phosphorylation point mutant cheYD54A back into the cheY fliYΔ6–15 strain. The CCW rotational bias of this strain remained low (Fig. 4D) while cheY+ was able to restore the high bias of this strain (not shown). Therefore it is likely that CheY-P, not unphosphorylated CheY is responsible for the high CCW bias in the fliYΔ6–15 mutant.
      Figure thumbnail gr4
      Fig. 4Tethered cell assay of switch protein mutants.A, fliMΔ6–15; B, fliYΔ6–15; C, cheY fliYΔ6–15; D, cheYD54AfliYΔ6–15. Each graph represents the average CCW rotational bias of a population of 20 cells. Downward and upward arrows indicate the addition and removal of 0.5 mm of the attractant asparagine, respectively.
      FliY Promotes Enhanced Hydrolysis of CheY-P in Vitro—As mentioned, the most likely explanation for the high bias in the fliYΔ6–15 mutant is high CheY-P levels in this strain. This could have been because of the lack of CheY-P hydrolysis activity of FliY if FliY had such an activity. In order to address this possibility, we incubated CheA with [γ-32P]ATP, removed the ATP, and exposed the resulting CheA-32P to CheY and various amounts of FliY. The more FliY was present, the less stable CheY-P became (Fig. 5). A similar assay performed using FliM instead of FliY did not show any effect on CheY-P stability (not shown), indicating that only FliY is capable of performing this function.
      Figure thumbnail gr5
      Fig. 5CheY-P hydrolysis assay. Shown are time points tracking dephosphorylation of CheY-32P. Lanes 1 and 10 contained 10 mm CheA-P before the addition of 20 mm CheY. 15-, 60-, 120-, and 240-s time points were taken in the presence of 0 μm (lanes 2–5), 0.2 μm (lanes 6–9), 1 μm (lanes 11–14), and 5 μm FliY (lanes 15–18).
      FliY Homologues Contain a Repetitive Unit in Their Sequence with Several Highly Conserved Residues—FliY homologues can be found in many Gram-positive organisms as well as in Thermotoga maritima and in some spirochetes. In addition, the chemotaxis protein CheC, homologous to the N terminus of FliY is widely spread throughout the bacteria and archaea. Alignment of N-terminal FliY sequences reveals two regions of high similarity, which appear to be repetitive units of each other (Fig. 6). These units are both 31 amino acid residues long and show a few conserved positions. Position 1 is either a serine or an aspartate; position 5 is a glutamate; position 9 is an asparagine; position 13 is a glycine; position 17 is a threonine; and position 31 is a proline (Fig. 6). Interestingly, most of these residues are also conserved throughout different CheCs (not shown). The significance of this observation needs yet to be investigated; however, we believe that this might indicate that some of these residues are involved in the function of FliY to increase the rate of CheY-P hydrolysis. It also might indicate that CheC and FliY share a common function.
      Figure thumbnail gr6
      Fig. 6Alignment of two repetitive units in selected FliY homologues. Conserved residues are in boldface and shaded gray. The two repetitive units are framed residues 35–74 and residues 132–171 following B. subtilis numbering. B. s., B. subtilis; B. h., Bacillus halodurans; B. c., B. cereus; O. i., Oceanobacillus iheyensis; T. t., Thermoanaerobacter tengcongensis; C. a., Clostridium acetobutylicum; T. m., T. maritima; L. i., Listeria innocua; and L. m., Listeria monocytogenis.

      DISCUSSION

      Using both in vivo and in vitro assays, we were able to identify a role for the N terminus of the B. subtilis flagellar switch protein FliY. We established that both FliY and FliM are capable of binding CheY-P in vitro and that this interaction is greatly reduced on deleting amino acid residues 6–15 in either FliY or FliM. In vivo, strains with these respective mutations show opposite phenotypes as observed by the tethered cell assay. The fliYΔ6–15 phenotype could be explained by showing that FliY enhances the rate of CheY-P hydrolysis.
      In E. coli CheY-P is destabilized by the chemotaxis protein CheZ (
      • Hess J.F.
      • Oosawa K.
      • Kaplan N.
      • Simon M.I.
      ). The importance of this activity is documented by the fact that the cheZ mutant is not capable of performing chemotaxis (
      • Parkinson J.S.
      • Houts S.E.
      ). It was puzzling that no CheZ homologue could be found in B. subtilis and, as is known now, the majority of chemotactic bacteria. While FliY and CheZ share little to no sequence similarity and different secondary structures (shown for CheZ (
      • Zhao R.
      • Collins E.J.
      • Bourret R.B.
      • Silversmith R.E.
      ) and predicted for FliY by various different algorithms), they seem to have a similar function. This activity of FliY remained elusive for a long time due to the fact that it is a feature of a flagellar switch protein. No enzymatic activity has ever before been associated with the flagellar switch, and since deleting switch proteins from the genome results in unflagellated bacteria, no information regarding chemotaxis functions could have been forthcoming from the null mutant that was characterized (
      • Yamaguchi S.
      • Aizawa S.
      • Kihara M.
      • Isomura M.
      • Jones C.J.
      • Macnab R.M.
      ).
      It remains to be seen whether FliY and CheZ hydrolyze CheY-P in a similar manner. Based on the CheY-P-CheZ crystal structure, it has been suggested that CheZ functions by positioning a water molecule into the active site of CheY for nucleophilic attack. This function is carried out by the conserved Gln-147 residue (
      • Zhao R.
      • Collins E.J.
      • Bourret R.B.
      • Silversmith R.E.
      ). While no conserved glutamine residues can be found in the N terminus of FliY other than residue 8, which is part of the CheY-P binding area, several conserved residues, namely Asp-39, Glu-43, Asn-46, Ser-136, Glu-140, or Asn-143, are potential candidates for a similar role. Interestingly, all of the aforementioned residues are conserved not only among FliY proteins but also among CheC proteins and the latter three are also conserved in CheX proteins (not shown). At this point, we can only speculate whether these proteins share a function with FliY. However, it is not unprecedented that multiple phosphatases exist to dephosphorylate the same response regulator. An example can be found in B. subtilis itself. Multiple protein-aspartate phosphatases have been shown to dephosphorylate Spo0F-P (
      • Perego M.
      ,
      • Perego M.
      • Hanstein C.
      • Welsh K.M.
      • Djavakhishvili T.
      • Glaser P.
      • Hoch J.A.
      ).
      Regardless of whether CheC shares FliY's function, it seems reasonable to assume that FliY evolved earlier than CheZ. Both its N-terminal homologue CheC as well as its C-terminal homologue FliN can be found throughout almost all phyla of chemotaxis-performing bacteria, and CheC can be found even in some archaeal species, whereas CheZ is limited to the class of β- and γ-proteobacteria (
      • Aizawa S.I.
      • Zhulin I.B.
      • Marquez-Magana L.
      • Ordal G.W.
      ).
      An evolutionary link might be found in T. maritima MSB8. Its genome revealed a FliY-like protein (coordinates 706,930–707,957) (
      • Nelson K.E.
      • Eisen J.A.
      • Fraser C.M.
      ), which through an (authentic) frameshift appears to be cut into two proteins, a FliY N-terminal part and a FliN. It appears likely that the FliN part is actually expressed as both S. typhimurium fliN and B. subtilis fliY mutants are not flagellated (
      • Bischoff D.S.
      • Ordal G.W.
      ,
      • Yamaguchi S.
      • Aizawa S.
      • Kihara M.
      • Isomura M.
      • Jones C.J.
      • Macnab R.M.
      ), yet T. maritima MSB8 is motile (
      • Gluch M.F.
      • Typke D.
      • Baumeister W.
      ) and its genome does not seem to encode for an alternative FliN. This might be an example of how FliN originated from FliY by loss of its N-terminal domain.

      Acknowledgments

      We thank Dr. Shin-Ichi Aizawa for the gift of anti-FliM-antibody. We thank George Glekas for help with the tethered cell assay, Wei Yuan for help with some of the PCRs, Travis Muff for help with the GST pull-down assay, and all of the Ordal laboratory members for suggestions on the paper.

      References

        • Larsen S.H.
        • Reader R.W.
        • Kort E.N.
        • Tso W.W.
        • Adler J.
        Nature. 1974; 249: 74-77
        • Berg H.C.
        • Brown D.A.
        Antibiot. Chemother. 1974; 19: 55-78
        • Macnab R.M.
        Neidhardt F.C. Curtiss III, R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society of Microbiology, Washington, D. C.1996: 123-145
        • Stock J.B.
        • Surette M.G.
        Neidhardt F.C. Curtiss III, R. Ingraham J.L. Lin E.C.C. Low K.B. Magasanik B. Reznikoff W.S. Riley M. Schaechter M. Umbarger H.E. Escherichia coli and Salmonella: Cellular and Molecular Biology. American Society of Microbiology, Washington, D. C.1996: 1103-1129
        • Aizawa S.I.
        • Zhulin I.B.
        • Marquez-Magana L.
        • Ordal G.W.
        Sonenshein A.L. Losick R. Hoch J.A. Bacillus subtilis and its Closest Relatives: from Genes to Cells. American Society of Microbiology, Washington, D. C.2002: 437-452
        • Bischoff D.S.
        • Ordal G.W.
        Mol. Microbiol. 1992; 6: 23-28
        • Kirby J.R.
        • Kristich C.J.
        • Saulmon M.M.
        • Zimmer M.A.
        • Garrity L.F.
        • Zhulin I.B.
        • Ordal G.W.
        Mol. Microbiol. 2001; 42: 573-585
        • Karatan E.
        • Saulmon M.M.
        • Bunn M.W.
        • Ordal G.W.
        J. Biol. Chem. 2001; 276: 43618-43626
        • Kristich C.J.
        • Ordal G.W.
        J. Biol. Chem. 2002; 277: 25356-25362
        • Welch M.
        • Oosawa K.
        • Aizawa S.
        • Eisenbach M.
        Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 8787-8791
        • Bischoff D.S.
        • Ordal G.W.
        Mol. Microbiol. 1992; 6: 2715-2723
        • Yamaguchi S.
        • Aizawa S.
        • Kihara M.
        • Isomura M.
        • Jones C.J.
        • Macnab R.M.
        J. Bacteriol. 1986; 168: 1172-1179
        • Celandroni F.
        • Ghelardi E.
        • Pastore M.
        • Lupetti A.
        • Kolsto A.B.
        • Senesi S.
        FEMS Microbiol. Lett. 2000; 190: 247-253
        • Senesi S.
        • Celandroni F.
        • Salvetti S.
        • Beecher D.J.
        • Wong A.C.
        • Ghelardi E.
        Microbiology. 2002; 148: 1785-1794
        • Bren A.
        • Eisenbach M.
        J. Mol. Biol. 1998; 278: 507-514
        • Davanloo P.
        • Rosenberg A.H.
        • Dunn J.J.
        • Studier F.W.
        Proc. Natl. Acad. Sci. U. S. A. 1984; 81: 2035-2039
        • McCleary W.R.
        • Stock J.B.
        J. Biol. Chem. 1994; 269: 31567-31572
        • Kirby J.R.
        • Niewold T.B.
        • Maloy S.
        • Ordal G.W.
        Mol. Microbiol. 2000; 35: 44-57
        • Kirby J.R.
        • Saulmon M.M.
        • Kristich C.J.
        • Ordal G.W.
        J. Biol. Chem. 1999; 274: 11092-11100
        • Hess J.F.
        • Oosawa K.
        • Kaplan N.
        • Simon M.I.
        Cell. 1988; 53: 79-87
        • Parkinson J.S.
        • Houts S.E.
        J. Bacteriol. 1982; 151: 106-113
        • Zhao R.
        • Collins E.J.
        • Bourret R.B.
        • Silversmith R.E.
        Nat. Struct. Biol. 2002; 9: 570-575
        • Perego M.
        Mol. Microbiol. 2001; 42: 133-143
        • Perego M.
        • Hanstein C.
        • Welsh K.M.
        • Djavakhishvili T.
        • Glaser P.
        • Hoch J.A.
        Cell. 1994; 79: 1047-1055
        • Nelson K.E.
        • Eisen J.A.
        • Fraser C.M.
        Methods Enzymol. 2001; 330: 169-180
        • Gluch M.F.
        • Typke D.
        • Baumeister W.
        J. Bacteriol. 1995; 177: 5473-5479
        • Ullah A.H.
        • Ordal G.W.
        J. Bacteriol. 1981; 145: 958-965
        • Zuberi A.R.
        • Bischoff D.S.
        • Ordal G.W.
        J. Bacteriol. 1991; 173: 710-719
        • Bhandari P.
        • Gowrishankar J.
        J. Bacteriol. 1997; 179: 4403-4406
        • Tabor S.
        • Richardson C.C.
        Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 1074-1078
        • Ireton K.
        • Rudner D.Z.
        • Siranosian K.J.
        • Grossman A.D.
        Genes Dev. 1993; 7: 283-294
        • Leonhardt H.
        • Alonso J.C.
        J. Gen. Microbiol. 1988; 134: 605-609
        • Bischoff D.S.
        • Ordal G.W.
        J. Biol. Chem. 1991; 266: 12301-12305
        • Bischoff D.S.
        • Bourret R.B.
        • Kirsch M.L.
        • Ordal G.W.
        Biochemistry. 1993; 32: 9256-9261