Helicobacter pylori FlhA Binds the Sensor Kinase and Flagellar Gene Regulatory Protein FlgS with High Affinity

ABSTRACT

Flagellar biogenesis is a complex process that involves multiple checkpoints to coordinate transcription of flagellar genes with the assembly of the flagellum. In Helicobacter pylori, transcription of the genes needed in the middle stage of flagellar biogenesis is governed by RpoN and the two-component system consisting of the histidine kinase FlgS and response regulator FlgR. In response to an unknown signal, FlgS autophosphorylates and transfers the phosphate to FlgR, initiating transcription from RpoN-dependent promoters. In the present study, export apparatus protein FlhA was examined as a potential signal protein. Deletion of its N-terminal cytoplasmic sequence dramatically decreased expression of two RpoN-dependent genes, flaB and flgE. Optical biosensing demonstrated a high-affinity interaction between FlgS and a peptide consisting of residues 1 to 25 of FlhA (FlhA_NT). The K_D (equilibrium dissociation constant) was 21 nM and was characterized by fast-on (k_on = 2.9 × 10⁴ M⁻¹s⁻¹) and slow-off (k_off = 6.2 × 10⁻⁴ s⁻¹) kinetics. FlgS did not bind peptides consisting of smaller fragments of the FlhA_NT sequence. Analysis of binding to purified fragments of FlgS demonstrated that the C-terminal portion of the protein containing the kinase domain binds FlhA_NT. FlhA_NT binding did not stimulate FlgS autophosphorylation in vitro, suggesting that FlhA facilitates interactions between FlgS and other structures required to stimulate autophosphorylation.

IMPORTANCE The high-affinity binding of FlgS to FlhA characterized in this study points to an additional role for FlhA in flagellar assembly. Beyond its necessity for type III secretion, the N-terminal cytoplasmic sequence of FlhA is required for RpoN-dependent gene expression via interaction with the C-terminal kinase domain of FlgS.

INTRODUCTION

Helicobacter pylori is an epsilonproteobacterium that can cause significant pathologies in the stomach (1–3). Approximately 50% of the world population is infected with H. pylori, although only a small fraction of infected individuals have symptoms. H. pylori possesses 2 to 6 polar flagella that are used to burrow through the mucus layer lining the stomach epithelium. Colonization of the gastric mucosa is dependent on motility, as nonflagellated mutants are unable to colonize (4).

The flagellum itself is a complex structure comprised of the basal body, hook, and filament (5). The basal body is located within the cell envelope and contains the flagellar protein export apparatus, the various flagellar rings, the rod, and motor components. The export apparatus is responsible for the secretion of axial components of the flagellum and consists of six proteins located within the inner membrane (FliO, FliP, FliQ, FliR, FlhA, and FlhB) plus three cytoplasmic proteins that bring flagellar substrates to the integral membrane component of the export apparatus (FliH, FliI, and FliJ). The C ring (or switch complex) is located at the cytoplasmic side of the inner membrane. In H. pylori it consists of four proteins, FliG, FliM, FliN, and FliY. In addition to controlling the rotational direction of the flagellum, the C ring works with the soluble components of the export apparatus to bring flagellar substrates to the export apparatus for secretion (6, 7). The rod proteins are the first proteins exported and are followed by hook proteins. The hook serves as a universal joint between the rod and the filament, transmitting torque from the motor to the filament. The filament is assembled after the completion of the hook and involves the minor flagellin FlaB and the major flagellin FlaA.

Flagellar biogenesis is a complex process that involves the coordinated expression of over 50 structural and regulatory genes with assembly of the nascent flagellum. In H. pylori, temporal expression of flagellar genes is controlled by the three sigma factors found in the bacterium: RpoD (σ⁸⁰), RpoN (σ⁵⁴), and FliA (σ²⁸). Transcription of the early flagellar genes, which encode components of the basal body, is regulated by RpoD. Genes whose products are needed for flagellar biogenesis following assembly of the basal body include components of the hook and a minor flagellin, and transcription of these genes is regulated by RpoN. Transcription of the RpoN-dependent genes is regulated by the two-component system consisting of the sensor kinase FlgS and the response regulator FlgR (8, 9). FlgS undergoes autophosphorylation in response to an unknown signal and subsequently transfers the phosphate to FlgR. Based on characterization of other RpoN-dependent genes (10), phosphorylation of FlgR likely promotes multimerization of the protein, which allows it to interact productively with RpoN-RNA polymerase holoenzyme to stimulate transcription. Components of the export apparatus are required for transcription of the RpoN-dependent genes, as deletions of flhB, fliO, and flhA result in decreased expression of these genes in H. pylori (11–13). The transcription of genes whose products are required late in flagellar biogenesis, which include the major flagellin, is regulated by FliA. The activity of FliA is regulated by the anti-sigma factor FlgM which, when bound to FliA, represses its activity (14). In Salmonella, inhibition of FliA activity is relieved when FlgM is secreted through the nascent flagellum as a filament-type substrate (15). FlgM has not been shown to be exported in H. pylori, but it may bind FlhA to alleviate the repression of FliA (16).

In this study, we sought to identify the activating signal that is sensed by FlgS to initiate signal transduction, resulting in expression of the RpoN regulon. In a previous study, we showed that a truncated form of the export apparatus component FlhA, consisting of only the first 77 residues of the protein, is sufficient to support transcription of RpoN-dependent genes (13). The 25-amino-acid residues at the N terminus of FlhA (FlhA_NT) are predicted to be exposed on the cytoplasmic side of the membrane. They are the only significant components of the 77-amino-acid sequence likely to be exposed to cytoplasmic FlgS, since putative transmembrane domains span residues ∼26 to 48 and 50 to 69. We show here that an FlhA variant lacking FlhA_NT is unable to support transcription of RpoN-dependent genes, suggesting a role for FlhA_NT in regulating expression of the RpoN regulon.

Using optical biosensing, we show that a peptide corresponding to FlhA_NT binds FlgS with nanomolar affinity via the C-terminal half of FlgS. In vitro phosphorylation studies suggest that binding of the FlhA_NT peptide does not stimulate FlgS autophosphorylation. Our results suggest a mechanism by which transcription of RpoN-dependent genes is initiated via an early flagellum assembly checkpoint that involves the amino-terminal segment of FlhA.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Escherichia coli strains were grown at 37°C in Luria-Bertani (LB) broth or on LB agar. Kanamycin (30 μg/ml) or ampicillin (100 μg/ml) was added to the medium when appropriate. H. pylori strains were grown on tryptic soy agar (TSA) supplemented with 10% horse serum at 37°C under an atmosphere consisting of 2% O₂, 5% CO₂, and 93% N₂. Kanamycin (30 μg/ml) or chloramphenicol (30 μg/ml) was added to the medium for culturing H. pylori when appropriate.

Strain construction.

ΔflhA mutants were constructed in H. pylori strains B128 and ATCC 43504 as previously described (13). Briefly, the region corresponding to 90 nucleotides upstream of the start codon through codon 77 of flhA was replaced with a chloramphenicol transacetylase (cat) cassette. Genomic DNA from H. pylori 26695 was extracted using the Wizard genomic DNA purification kit (Promega) and used as the template for creating flhA alleles. The ΔflhA mutant was complemented with flhA alleles expressed from the native flhA promoter and carried on the shuttle vector pHel3 (17). Primers SphI flhA F2 and KpnI flhA R2 were used to amplify flhA and its native promoter. This region was cloned into pCR2.1 and subcloned into pHel3 via the SphI and KpnI sites. The resulting plasmid containing the flhA region was called pflhA. Another plasmid was generated to create an in-frame deletion of codons 2 to 24 (pflhA_ΔNT). Primers SphI PflhA F2 and flhA24 R were used to amplify the flhA promoter and the start codon of flhA. Primers flhA24 F and KpnI flhA R2 were used to amplify codons 25 to 56 nucleotides downstream of the flhA stop codon. Primer flhA24 F contains sequences reverse and complementary to flhA24 R. This region was used to perform overlapping PCR to create an in-frame deletion of codons 2 through 24. The resulting amplicon was cloned into pCR2.1 and subcloned into pHel3, and the resulting plasmid was named pflhA_ΔNT. Plasmids pflhA and pflhA_ΔNT were introduced into the ΔflhA mutant by natural transformation. All constructs were confirmed by PCR and sequencing of the resulting amplicon. Primers used for strain construction are listed in Table 1.

TABLE 1.

Primers and peptides used in this study

Primer or peptide	Sequence
Primers
SphI flhA F2	ATG TCA GCA TGC ATA ACG ACC TCC TAA TTG GTA A
KpnI flhA R2	CAT GGT ACC CAA TGT GTG AAA GGT GGT AAA C
flhA24 R	CAT AAT CTA AAA TCA ATG CCT
flhA24 F	AGG CAT TGA TTT TAG ATT ATG GAC TTA GCC CTT GTG GTC
FlgS F	GTC TCG CAT ATG AAA AAA TCC AAG CAC TTA AAA C
FlgS R	CCT ATC GGA TCC AAA TTA AGA AGC GTT AAG AAT
gyrA F	GCT AGG ATC GTG GGT GAT GT
gyrA R	TGG CTT CAG TGT AAC GCA TC
flgE F	TGC GAA CGT GAA TAC CAC TG
flgE R	GTC ATT CTG CCC TGC TAA CC
flaB F	ATC GCC GCT TTA ACT TCT CA
flaB R	CGC CAT CCC ACT AGA ATC AT
Peptides
FlhANT	Biotin-MANERSKLAFKKTFPVFKRFLQSKD
FlhANTN	Biotin-MANERSKLAFKKTF
FlhANTC	Biotin-PVFKRFLQSKD

Plasmids encoding fragments of flgS were synthesized by DNA 2.0 (Menlo Park, CA). The parent vector pJ414 (DNA 2.0) contains a T7 promoter and a terminator sequence flanking a multiple cloning site. DNA sequences encoding residues 1 to 169 and residues 170 to 230 along with an N-terminal His tag were cloned into the NdeI and BamHI site of this vector to create the plasmids pflgS_N and pflgS_C, respectively. For expression and purification, pflgS_N and pflgS_C were transformed into E. coli BL21(DE3) pLysS.

Motility assay.

Motility of H. pylori cells was assayed on soft agar plates consisting of Mueller-Hinton broth and 0.4% noble agar. After autoclaving, sterile heat-inactivated horse serum (10% final concentration), FeSO₄ (10 μM final concentration), and MES (20 mM final concentration) were added to the medium. A sterile toothpick was used to inoculate cells into the center of the agar. Three inocula were used per strain. Plates were incubated at 37°C under an atmosphere of 2% O₂, 5% CO₂, and 93% N₂. Diameters of the halos formed by cells migrating from the point of inoculation were measured after 1 week. Diameter averages were tested for significance with Student's t test.

qRT-PCR.

RNA was extracted as previously described (13). Briefly, H. pylori cells were grown on TSA supplemented with 10% horse serum for 18 h. Cells were harvested and resuspended in 100 μl of distilled water. RNA was extracted using the Aurum total RNA minikit (Bio-Rad), and contaminating DNA was removed using the Turbo DNA-free kit (Ambion). Single-stranded cDNA was synthesized from 200 ng RNA using the iScript cDNA synthesis kit (Bio-Rad). Relative transcript levels of flaB, flgE, and flaA were determined using quantitative reverse-transcription PCR (qRT-PCR) as described in reference 12. Primers used for this study are listed in Table 1. Significance was determined using the Student t test.

FlgS purification.

flgS was amplified from H. pylori ATCC 43504 using primers FlgS F and FlgS R with Pfu polymerase. The amplicon was cloned into the NdeI and the BamHI sites of pET19b. The resulting construct (pflgS) was verified by PCR and sequencing. For expression and purification of FlgS, pflgS was transformed into E. coli KRX cells. Cultures were grown at 37°C in Terrific broth to an optical density at 600 nm (OD₆₀₀) of 1. Rhamnose (0.1% final concentration) and isopropyl-β-d-thiogalactopyranoside (IPTG) (1 mM final concentration) were added to the culture to induce expression of His-FlgS. After overnight growth at room temperature, cells were harvested by centrifugation and resuspended into buffer A (50 mM phosphate buffer, pH 8.0, 500 mM NaCl, 10 mM imidazole, 10% glycerol, 0.05% Tween 20). All purification steps were carried out at 4°C. Cells were lysed by three passages through a French pressure cell at 10,000 kPa. Unlysed cells were removed by centrifugation for 15 min at 6,000 × g. His-FlgS was purified using His-Pur nickel-nitrilotriacetic acid (Ni-NTA) resin (Thermo Scientific). The His-Pur Ni-NTA resin was equilibrated in buffer A and then incubated with the lysate for 1 h on an end-over-end tube rotator. The resin containing any bound proteins was separated from unbound proteins by centrifugation and washed with buffer B (50 mM phosphate buffer, pH 8.0, 500 mM NaCl, 25 mM imidazole, 10% glycerol, 0.05% Tween 20). The resin mixture then was transferred to a column for further washing with buffer B. His-FlgS was eluted from the resin using buffer C (50 mM phosphate buffer, pH 8.0, 500 mM NaCl, 250 mM imidazole, 10% glycerol, 0.05% Tween 20). For phosphorylation studies, eluted proteins were dialyzed in 10 mM HEPES, 150 mM NaCl, 10% glycerol, 1 mM dithiothreitol, 0.05% Tween 20, 100 mM potassium thiocyanate, and 0.1 mM EDTA, pH 7.4. For biosensing experiments, FlgS was exchanged into binding buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 0.05% Tween 20, 10% glycerol, 1 mM dithiothreitol). Protein concentration was determined by a bicinchoninic acid protein assay (Pierce) or Bradford assay. FlgS_N and FlgS_C were purified as described for FlgS, with minor modifications.

Optical biosensing.

All biolayer interferometry (BLI) experiments were performed on a FortéBio (Menlo Park, CA) Octet QK biosensor using streptavidin (SA) sensors. Assays were performed in 96-well microplates at 25°C. All volumes were 200 μl. Ligand peptides used in this study were synthesized with an N-terminal biotin group (Biomatik, Cambridge, Canada). Their sequences are shown in Table 1. After loading ligands onto SA sensors, a baseline was established in binding buffer prior to monitoring association at various analyte concentrations. Dissociation subsequently was measured in buffer only. Nonspecific binding was measured by screening sensors not exposed to ligand versus analytes under conditions identical to those of the binding assays and with the same analyte samples. Raw data were analyzed with GraphPad Prism, and constants were determined by fits to appropriate models.

In vitro phosphorylation assay.

The phosphorylation state of FlgS was monitored using a phosphate affinity polyacrylamide (Phos-tag) gel electrophoresis system (Wako Chemistry USA, Inc.) as previously described (18). A buffer containing 50 mM Tris-HCl (pH 7.5), 50 mM KCl, and 20 μM MgCl₂ was used for all dilutions. Reaction mixtures contained 2 μg FlgS and 200 nM FlhA_NT and were initiated with 1 mM ATP in a total volume of 25 μl. Reaction mixtures were incubated at room temperature for 2 min, 5 min, and 10 min before stopping the reaction with the addition of 6× SDS-PAGE loading dye (168 mM Tris base, pH 6.8, 7% SDS, 0.3% bromophenol blue, 34% glycerol, 4% 2-mercaptoethanol). Phos-tag gels were made according to the supplier's protocols. Fifteen microliters of the reaction mixture was loaded onto the gel. Gels were stained with Coomassie brilliant blue to visualize both unphosphorylated and phosphorylated forms of FlgS.

RESULTS

The amino terminus of FlhA is required for motility in H. pylori ATCC 43504.

FlhA is required for transcription of the RpoN regulon (13). It is a large protein of 733 amino acid residues in length and is a member of the membrane-integrated portion of the flagellar protein export apparatus. Like its homolog in Salmonella, hydropathy analysis of FlhA predicts a small N-terminal cytoplasmic region, followed by eight membrane-spanning helices and a large cytoplasmic region (Fig. 1A). A strain carrying a complete deletion of flhA (the ΔflhA strain) does not support transcription of the RpoN regulon. However, a strain capable of expressing only the first 77 residues of FlhA (the flhA77 strain) containing the N-terminal cytoplasmic region and the first two transmembrane segments is able to support wild-type levels of RpoN-dependent gene transcription (13). Therefore, we hypothesized that the N-terminal cytoplasmic region of FlhA (FlhA_NT) plays a role in regulation of the RpoN regulon, perhaps via interaction with a regulatory protein, such as FlgS.

FIG 1.

(A) Transmembrane topography as inferred by hydropathy analysis. FlhA_TM, the transmembrane domain; FlhA_C, cytoplasmic domain. The sequence deleted (residues 2 to 24) for construct pflhA_ΔNT is indicated at the left. (B) Motility assays. Top row, the ΔflhA mutant from parent strain 43504 transformed with empty vector, pflhA, and pflhA_ΔNT. Bottom row, 43504 flhA deletion strain transformed with the same plasmids as the wild type. (C) Strain-specific motility differences. H. pylori strains 43504 and B128 were used as the parental strains. The parent strain for flhA deletion is indicated by color, with motility results quantified for each transformation. Error bars indicate standard errors from three separate trials.

To test this hypothesis, we monitored the ability of a plasmid bearing the full-length flhA allele (pflhA) and a plasmid bearing an flhA allele which lacked codons 2 through 24 (pflhA_ΔNT) to complement motility defects of the ΔflhA mutant. The plasmid-borne flhA alleles possessed the native flhA promoter. Plasmids were transformed into wild-type and ΔflhA mutant cells. Transformation with pflhA partially restored motility in the ΔflhA background (the resulting strain was called the ΔflhA/pflhA strain) (Fig. 1B). Introduction of pflhA in the wild-type background (resulting in the WT/pflhA strain) reduced motility, possibly due to elevated levels of FlhA from the plasmid-based expression method. We have obtained similar results in previous experiments where overexpression of an export apparatus protein in otherwise wild-type cells resulted in decreased motility in soft agar medium (12, 19). In contrast to pflhA, the introduction of pflhA_ΔNT in the ΔflhA mutant (the ΔflhA/pflhA_ΔNT strain) did not enhance the motility of the strain (Fig. 1B), suggesting that the N-terminal portion of FlhA is required for flagellar biogenesis. Introduction of pflhA_ΔNT in the wild-type background (the WT/pflhA_ΔNT strain) inhibited motility. We infer from these results that the truncated FlhA_ΔNT variant is stably expressed and inserted into the export apparatus, where it interferes with the function of full-length FlhA proteins in H. pylori ATCC 43504.

Complementation differences were observed between the two background strains, B128 and ATCC 43504. In both ΔflhA strains, transformation with pflhA restored motility, whereas pflhA_ΔNT did not. However, significant differences in inhibition of motility upon transformation of the wild type were observed, particularly by pflhA_ΔNT (10-mm halo in ATCC 43504, 22-mm halo in B128). Differences in membrane assembly consistent with earlier reports may account for the observation that truncated forms of FlhA were able to associate with the membrane in ATCC 43504 but not in B128 (13).

The N terminus of FlhA is required for RpoN-dependent flagellar gene expression.

To determine whether the FlhA_ΔNT allele is required for transcription of the RpoN regulon, quantitative reverse transcription-PCR (qRT-PCR) was used to monitor expression of two RpoN-dependent genes, flaB and flgE. Transcript levels of flaB and flgE in the ΔflhA mutant were consistent with our previous results (13), where there was a very low basal level of transcription of flaB and transcription of flgE was about 6-fold lower than wild-type levels (Fig. 2). Introduction of pflhA into the ΔflhA mutant partially restored transcript levels, while the introduction of pflhA_ΔNT did not restore transcript levels in the ΔflhA mutant. These findings suggest FlhA_NT is required to initiate signal transduction resulting in transcriptional activation of the RpoN regulon. Since FlhA_NT is predicted to be exposed on the cytoplasmic side of the membrane, we postulated that FlgS recognizes FlhA_NT as part of an assembly checkpoint to initiate signal transduction.

FIG 2.

qRT-PCR shows restoration of RpoN-dependent gene transcription upon transformation with pflhA. Transcription of flaB and flgE is dramatically reduced in the ΔflhA mutant compared to that in the wild type. Transcription of both is partially restored in the ΔflhA/pflhA strain but not the ΔflhA/pflhA_ΔNT strain. Some transcription of flgE is evident. Significance was determined by using the two-sample t test. Asterisks indicate a P value of <0.05 when the ΔflhA strain is compared to the wild type, and the circle indicates a P value of <0.05 when the ΔflhA/pflhA_ΔNT strain is compared to the ΔflhA/pflhA strain.

The N terminus of FlhA binds FlgS via its C-terminal kinase domain.

To test the hypothesis that FlgS recognizes FlhA_NT, a synthetic peptide consisting of the first 25 residues of FlhA was used as a ligand to measure interactions with FlgS using biolayer interferometry (BLI). BLI is an optical biosensing technique similar to the better-known surface plasmon resonance (SPR) (20). It allows for measurement of intermolecular interactions in real time (21–23). Ligand molecules are tethered to fiber-optic sensors and then exposed to various concentrations of an analyte. Binding is measured by the shift of the interference pattern of white light reflected from the end of the sensor. After an association phase, sensors are moved to a buffer-only solution and dissociation is monitored. By examining several different analyte concentrations and fitting raw data to global binding models, rate and affinity constants can be determined for ligand-analyte interactions.

As shown in Fig. 3, binding at five different FlgS concentrations ranging from 31 to 500 nM readily fit a one-state global model. K_D (equilibrium dissociation constant) was 21 nM, with subsidiary fast-on kinetics (k_on) of 2.9 × 10⁴ M⁻¹ s⁻¹ and slow-off kinetics (k_off) of 6.2 × 10⁻⁴ s⁻¹. Goodness-of-fit parameters indicate excellent fits, e.g., the standard error of the K_D is 440 pM. Controls of analyte screened against sensors without ligands resulted in negligible nonspecific binding. Microscale thermophoresis (24), a solution method not requiring the tethering of one of the binding partners to a surface, yielded a K_D of 62 nM, in reasonable agreement with the BLI results (see Fig. S1 in the supplemental material).

FIG 3.

BLI analysis of FlhA_NT binding to FlgS. Raw data are shown in black, with fits to a global one-state model as red lines. Analyte concentrations were 31, 62, 125, 250, and 500 nM.

To more specifically determine which region of FlgS is involved in binding, interactions of FlhA_NT were analyzed against fragments of FlgS. Residues 1 to 169 (termed FlgS_N) did not bind FlhA_NT. However, residues 170 to 377 (termed FlgS_C) bound with high affinity (Fig. 4). A one-state global fit of five FlgS_C concentrations gave a k_on of 2.1 × 10⁴ M⁻¹s⁻¹ and k_off of 7.1 × 10⁻⁴ s⁻¹ for an overall K_D of 33 nM. Fits to the raw data were excellent, but not as good as those for full-length FlgS, e.g., the standard error for K_D was 1 nM. However, fits to a parallel two-state model yielded K_Ds of 24 nM and 17 μM, with the majority of the amplitudes accounted for in the high-affinity K_D (see Fig. S2 in the supplemental material), suggesting that the secondary event is an artifact and that the affinity of FlgS_C is nearly identical to that of full-length FlgS (see Discussion).

FIG 4.

BLI analysis of FlhA_NT binding to FlgS_C (residues 170 to 377). Raw data are shown in black, with fits to a global one-state model as red lines. Analyte concentrations were 62, 125, 250, 500, and 1000 nM.

An initial attempt to discern critical residues for binding was performed with synthetic peptides representing residues 1 to 14 (FlhA_NTn) and 15 to 25 (FlhA_NTc). Neither evinced binding to FlgS, as a response was indistinguishable from background noise (Fig. 5), suggesting that the binding site is disrupted in the smaller peptides.

FIG 5.

FlhA peptides binding to 1.2 μM FlgS (A) and 1.0 μM FlgS_C (B). Binding to FlhA_NT ligand is shown in black, binding to sensor only (no ligand) is in yellow, and binding to FlhA_NT fragment peptides is shown in red for FlhA_NTN and green for FlhA_NTC.

FlhA_NT does not stimulate autophosphorylation of FlgS.

To test the hypothesis that interaction with FlhA_NT stimulates autophosphorylation of FlgS, 200 nM FlhA_NT was incubated with 1.78 μM FlgS. Phosphorylated proteins were separated from unphosphorylated proteins using the Phos-tag gel system where phosphorylated proteins migrate slower through the gel than unphosphorylated proteins. Under our assay conditions, we observed a low level of phosphorylation of FlgS in the presence of ATP (Fig. 6, lane 2). The addition of FlhA_NT did not increase the amount of phosphorylated protein, indicating that FlhA_NT does not act by itself to stimulate the autokinase activity of FlgS (Fig. 6, lane 3).

FIG 6.

Phosphorylation of FlgS monitored by an in vitro phosphorylation assay. Phosphorylated proteins were separated from nonphosphorylated proteins using the Phos-tag gel system. Additions of FlgS, FlhA_NT, and ATP are indicated above the image.

DISCUSSION

Like its homologs in Salmonella and other species, H. pylori flhA is required for motility, presumably to effect flagellar type III secretion (19, 25). The N-terminal soluble cytoplasmic sequence also is required in both Salmonella and H. pylori, as its deletion eliminated the ability of flhA to complement ΔflhA mutations in these bacteria (Fig. 1B) (19). In the study by McMurry et al., an FlhA variant lacking residues 18 to 22 was able to complement an flhA mutant in Salmonella, while another variant that lacked residues 2 to 22 was unable to complement the same flhA mutant (19). We observed similarities in which expression of H. pylori FlhA in an flhA mutant was able to complement motility, whereas the expression of a truncated form of FlhA lacking the N-terminal segment (FlhA_ΔNT) was unable to complement motility (Fig. 1B). While for technical reasons (e.g., lack of antiserum directed against H. pylori FlhA) fractionation and protease protection assays could not be performed, the observation that FlhA_ΔNT inhibits wild-type motility strongly suggests proper assembly of H. pylori FlhA_ΔNT into the export apparatus. FlhA_NT has been proposed to interact with FliI, as either overproduction of FliI or bypass mutations in FlhA_NT improved motility in a ΔfliH mutant background, suggesting that FlhA_NT has a role in substrate export as well (26).

H. pylori FlhA also is involved in regulating the transcription of flagellar genes, as a complete deletion of flhA abolished transcription of the RpoN-dependent genes examined (13). The same study also showed that the first 77 residues were sufficient to support RpoN-dependent expression at about 60% of the level for the wild type. In the present study, we focused on the sequence N terminal to the first transmembrane segment. qRT-PCR results indicated that the first 25 residues are necessary for transcription of the RpoN-dependent genes flaB and flgE (Fig. 2).

Transcription of the RpoN-dependent genes is regulated by the FlgS/FlgR two-component system in which FlgS must interact with a signal to initiate signal transduction, which culminates in transcriptional activation of the RpoN regulon. Our results suggest that the signal sensed by FlgS involves FlhA_NT, which binds FlgS with high affinity (Fig. 3). Experiments with fragmentary peptides suggest that almost all of the N-terminal sequence is necessary or that binding determinants are in the center of the sequence (Fig. 5). FlhA_NT binding to FlgS_C (Fig. 4) demonstrated that the FlhA binding site on FlgS lies within the C-terminal half of the protein that contains both the histidine kinase A domain (HisKA; residues ∼170 to 230) and the histidine kinase-like ATPase domain (HATPase_c; residues ∼280 to 377) of the protein. While a simple model fit well, the complexity observed for FlhA_NT-FlgS_C could be better fit by a parallel two-state model (see Fig. S2 in the supplemental material). The second state, relatively slow on and fast off, was of low affinity and accounted for no more than about 10% of the dissociation amplitude. While it could represent some biologically relevant event, such as a conformational change, maybe even one governed by the N-terminal domain, we consider it more likely a result of expressing the C-terminal fragment of FlgS alone, perhaps due to alterations of the binding site.

The high affinity of FlgS for FlhA_NT suggests the interactions between these proteins are biologically relevant. Consistent with this hypothesis, the results of the qRT-PCR assays (Fig. 3) suggest the N-terminal segment of FlhA is necessary for transcription of the RpoN-dependent genes flaB and flgE. Failure of the FlhA_NT peptide to stimulate FlgS autokinase activity under the assay conditions (Fig. 6) can be attributed to several reasons. For example, FlhA_NT may not be the signal that stimulates FlgS autokinase activity, or it may be unable to stimulate FlgS activity outside its native context within the export apparatus. Alternatively, FlhA_NT may be part of a larger complex that functions as the signal for stimulating FlgS autokinase activity. Such a complex could simply include other copies of FlhA_NT, as the export apparatus is predicted to contain nine FlhA subunits (27), each of which could bind a FlgS monomer, or FlhA_NT may need to work in conjunction with other proteins to stimulate FlgS autophosphorylation. Good candidates for such auxiliary proteins are the MS ring protein FliF and the C ring protein FliG, as Boll and Hendrixson showed that FlgS could be cross-linked to these proteins in vivo in the epsilonproteobacterium Campylobacter jejuni (28). In support of this supposition, FlhA_NT, the C-terminal domain of FliF (FliF_C), and FliG likely are closely associated with each other, as FlhA has been shown to interact with FliF (29) and FliF_C has been shown to interact with FliG (30, 31). We plan to investigate these possibilities in subsequent studies.