FliZ Acts as a Repressor of the ydiV Gene, Which Encodes an Anti-FlhD₄C₂ Factor of the Flagellar Regulon in Salmonella enterica Serovar Typhimurium

Abstract

YdiV acts as an anti-FlhD₄C₂ factor, which negatively regulates the class 2 flagellar operons in poor medium in Salmonella enterica serovar Typhimurium. On the other hand, one of the class 2 flagellar genes, fliZ, encodes a positive regulator of the class 2 operons. In this study, we found that the FliZ-dependent activation of class 2 operon expression was more profound in poor medium than in rich medium and not observed in the ydiV mutant background. Transcription of the ydiV gene was shown to increase in the fliZ mutant. Purified FliZ protein was shown in vitro to bind to the promoter region of the nlpC gene, which is located just upstream of the ydiV gene, and to repress its transcription. These results indicate that FliZ is a repressor of the nlpC-ydiV operon and activates the class 2 operons by repressing ydiV expression. Therefore, the fliZ and ydiV genes form a regulatory loop.

INTRODUCTION

Salmonella enterica serovar Typhimurium cells swim by means of rotating flagella through their environment. The individual flagellum consists of three structural parts, a basal body, a hook, and a filament. More than 50 genes are specifically required for flagellar formation and function (1, 2, 4, 26). These genes are organized into at least 15 operons, and their transcriptional expression forms a highly organized three-tiered cascade called a flagellar regulon (22, 27). The flhDC operon is the sole one belonging to class 1 and encodes FlhD and FlhC, which assemble into an FlhD₄C₂ heterohexamer acting as an activator of class 2 operons (31, 42). Class 2 contains operons encoding component proteins for the hook-basal body structure and the flagellum-specific type III export apparatus as well as a gene encoding the flagellum-specific sigma factor σ²⁸ (FliA), essential for class 3 expression (33). Class 3 operons encode proteins involved in filament assembly and flagellar function. σ²⁸ activity is negatively controlled by an anti-σ²⁸ factor, FlgM (15, 24, 34), which is excreted from the cell through the flagellum-specific type III export apparatus upon completion of hook-basal body assembly (11, 19).

Class 2 operons are transcribed by σ⁷⁰ RNA polymerase in the presence of the FlhD₄C₂ complex, which binds to the DNA region upstream of the class 2 promoter (13, 30, 31). Two genes within the flagellar regulon, fliT and fliZ, have been shown to be involved in fine control of the class 2 operons (23). The fliT and fliZ genes belong to the fliDST and fliAZ operons, respectively, both of which are transcribed from both class 2 and class 3 promoters (12, 21, 46). FliT acts as an anti-FlhD₄C₂ factor, which binds to the FlhD₄C₂ complex through interaction with the FlhC subunit and inhibits its binding to the class 2 promoter, resulting in decreased expression of the class 2 operons (45). On the other hand, FliZ was shown to be a positive regulator of class 2 operons (23) and to participate in a positive-feedback loop that induces a kinetic switch in class 2 operon expression (38). Although class 2 operon expression is inhibited in vivo by the fliZ mutation (23), transcription from the class 2 promoter occurs efficiently in vitro in the absence of FliZ (13, 31). This suggests that FliZ affects indirectly the activation process of the class 2 operons. On the basis of the observation that the fliZ mutation decreased the amount of FlhC without having a significant effect on the transcription and translation of the flhC gene, Saini et al. (37) proposed that FliZ is a posttranslational activator of FlhD₄C₂. However, the molecular mechanism underlying FliZ control of the class 2 operons has remained obscure.

The fliZ mutation has also been shown to impact several cellular processes unrelated to flagellar biogenesis in Salmonella. For example, the fliZ mutation decreases the transcriptional expression of Salmonella pathogenicity island 1 (SPI1) genes (14) through affecting the activity of their positive regulator, HilD (5, 17). On the other hand, the fliZ mutation increases type 1 fimbrial gene expression through affecting posttranscriptionally the activity of the positive regulator FimZ (6, 39). Furthermore, in Escherichia coli, FliZ was shown to affect curli expression by interfering with RpoS, the stationary-phase-specific σ factor (35). However, the precise molecular mechanisms underlying these FliZ controls are also still unclear.

Recently, we found that a nonflagellar gene, ydiV, encodes another anti-FlhD₄C₂ factor (41). YdiV binds to FlhD₄C₂ by interacting with the FlhD subunit, resulting in inhibition of class 2 operon expression. The intracellular level of YdiV is higher in poor medium than in rich medium, which allows YdiV to act as a mediator of the nutritional level of the environment in the expression of the flagellar regulon. Nutritional control of YdiV expression was shown to be executed mainly at the translational level (41). On the other hand, Wozniak et al. (43) reported that ydiV transcription is under the negative regulation of the flhDC genes, suggesting that some gene(s) within the flagellar regulon may act as a negative regulator of the ydiV gene.

In this study, we showed that FliZ-dependent activation of class 2 operon expression was more profound in poor medium than in rich medium and not observed in the ydiV mutant background. This suggested the possibility that FliZ control of class 2 operon expression is exerted via the regulatory pathway of ydiV expression. Gel mobility shift assay and in vivo and in vitro transcription analyses revealed that FliZ binds to the promoter region of the nlpC-ydiV operon and acts as a repressor of its transcription. Therefore, FliZ-dependent activation of class 2 operon expression is achieved by repression of the ydiV gene by FliZ.

MATERIALS AND METHODS

Bacterial strains, plasmids, chemicals, and culture media.

The Salmonella strains and plasmids used in this study are listed in Tables 1 and 2, respectively. All the Salmonella strains used were derivatives of a standard laboratory strain, LT2. Their construction procedures are described below. P22-mediated transduction was performed as described previously (25, 28). Unless otherwise specified, all the chemicals and DNA enzymes used were purchased from Nacalai Tesque (Kyoto, Japan) and Toyobo (Osaka, Japan), respectively. Oligonucleotide primers used were purchased from Life Technologies (Tokyo, Japan) and are summarized in Table 3. The rich and poor media used were Luria broth (LB) and M9 minimal medium (32) supplemented with 0.2% glycerol and 0.2% Casamino Acids (MGC), respectively (41). Hard-agar and motility agar plates were prepared by adding 1.2% and 0.25% agar (Shoei, Tokyo, Japan), respectively, to LB or MGC. Ampicillin, kanamycin, and tetracycline were used at final concentrations of 50, 50, and 20 μg/ml, respectively. Where required, arabinose or isopropyl-β-d-thiogalactopyranoside (IPTG) was added to the culture media at final concentrations of 0.2% or 0.1 mM, respectively.

Table 1.

Salmonella enterica strains used in this study

Strain	Relevant characteristic(s)	Source or reference
KK1004	LT2 Δ(hin-fljBA) ΔFels-2	25, 28
KK1004hDC	KK1004 flhD2140::Tn10	28
KK1004iA	KK1004 ΔfliA::FRT (in-frame)	This study
KK1004iZ	KK1004 ΔfliZ::FRT (in-frame)	This study
KK1004iAZ	KK1004 ΔfliAZ::FRT (in-frame)	This study
KK1004V	KK1004 ΔydiV::FRT (in-frame)	41
KK1004ViA	KK1004V ΔfliA::FRT (in-frame)	This study
KK1004ViZ	KK1004V ΔfliZ::FRT (in-frame)	This study
KK1004ViAZ	KK1004V ΔfliAZ::FRT (in-frame)	This study
KK1004 flhC-3F	KK1004 flhC-3×FLAG	41
KK1004 ydiV-lac	KK1004 ΔydiV::FRT::pKG137	41

Table 2.

Plasmids used in this study

Plasmid	Description	Source or reference
pKD13	FRT-kan-FRT cassette template, amp	7
pKD46	Red expression plasmid, amp	7
pCP20	Flp expression plasmid, amp cat	7
pQE80L	Expression vector, His6 tag, amp	Qiagen
pQE80L-flhCD	pQE80L flhC flhD (His6-FlhC FlhD)	This study
pQE80L-fliZ	pQE80L fliZ (His6-FliZ)	This study
pQE80L-ydiV	pQE80L ydiV (His6-YdiV)	41
pQE80L-ΔPT5-PnlpC	pQE80L ΔPT5 PnlpC-T0	This study
pQE80K	pQE80L Δamp::kan	Laboratory stock
pQE80K-fliZ	pQE80K fliZ (His6-FliZ)	This study
pFZY1	Promoter probe vector, amp	18
pFZY1-PfliA-lacZ	pFZY1 PfliA-lacZ	This study
pFZY1-PnlpC-lacZ	pFZY1 PnlpC-lacZ	This study
pFZY1-ΔPnlpC-lacZ	pFZY1-PnlpC-lacZ ΔPnlpC	This study

Table 3.

Oligonucleotide primers used in this study

Use and primer name	Nucleotide sequence (5′ to 3′)a
Construction of chromosomal deletion mutants
fliAH1P1	TAGAAACGGATAATCATGCCGATAACTCATTTAAC
	GCAGGGCTGTTTATCGTGTAGGCTGGAGCTGCTTC
fliAH2P4	GGTAGTCTATACGTTGTGCGGCACTTTTCGGGTGC
	GATCATGCGCGACCTATTCCGGGGATCCGTCGACC
fliZH1P1	TCGCACCCGAAAAGTGCCGCACAACGTATAGACT
	ACCAGGAGTTCTCATGGTGTAGGCTGGAGCTGCTTC
fliZH2P4	AGGTTTGCCACGTTTCACCAACACGACTCTGCTA
	CATCTTATGCTTTTTTATTCCGGGGATCCGTCGACC
Construction of plasmids
IAEIB	GGGGGATCCTCTGACGATACTCCGC
KpnIfliA0	GGGGGTACCGCTACAGGTTACATAA
IZP1	GGGGATCCACGGTGCAGCAACCTAAAAG
SalIfliZR	GGGGTCGACTTAATATATATCAGAA
HCf1B	GGGGATCCAGTGAAAAAAGCATTGTTCAGG
HCr1S	GGGGTCGACTTAAACAGCCTGTTCGATCTG
HDf1S	GGGGTCGACACATCACGGGGTGCGGCTA
HDr1P	GGGCTGCAGTCATGCCCTTTTCTTACGCGC
PnlpCf1E	GGGGAATTCATTTATGTCAGCCAGGAATTG
PnlpCf2E	GGGGAATTCCTACGGTTTGCGCTTTCGACG
NLPCF	ACCAGAATTCAAACAGAGGATTGTTGC
PnlpCr1B	GGGGGATCCAAAAACGCATGCCGCAACAAT
PnlpCr2B	GGGGGATCCCTGCAACTGATCGTTCAGACC

^aNucleotides corresponding to the created restriction enzyme cleavage sites are shown in italics.

Construction of chromosomal deletion mutants.

The λ Red recombination system (7) was used for construction of chromosomal in-frame deletion mutants. Double deletion mutants were constructed by P22-mediated transduction.

An in-frame deletion mutant of the fliA gene (KK1004iA) was constructed as follows. The kanamycin resistance Flp recombination target (FRT) cassette was PCR amplified from pKD13 with primers fliAH1P1 and fliAH2P4, which possess sequences homologous to the 5′- and 3′-end-flanking regions of the fliA gene, respectively. The amplified product was introduced into KK1004 harboring pKD46 by electroporation, and kanamycin-resistant colonies were selected on LB agar plates containing kanamycin and arabinose. After the correct replacement of the fliA gene with the kanamycin resistance gene cassette was confirmed, this mutation was moved to fresh KK1004 via P22-mediated transduction. The kan gene was removed through site-specific recombination between the flanking FRT sequences by transient exposure of the mutant cells to pCP20.

Similarly, an in-frame deletion of the fliZ gene (resulting in strain KK1004iZ) was carried out using primers fliZH1P1 and fliZH2P4, which possess sequences homologous to the 5′- and 3′-end-flanking regions of the fliZ gene, respectively. Through the same procedure, an in-frame deletion of the fliAZ operon (resulting in strain KK1004iAZ) was carried out using primers fliAH1P1 and fliZH2P4.

Plasmid construction.

The fliZ gene was PCR amplified with primers IZP1 and SalIfliZR using genomic DNA from KK1004 as a template. The amplified product was digested with BamHI and SalI and inserted into the corresponding site of pQE80L or pQE80K to obtain pQE80L-fliZ or pQE80K-fliZ, respectively. They encode an N-terminally hexahistidine-tagged FliZ protein (His₆-FliZ). Introduction of these plasmids restored full motility to the cells of KK1004iZ (ΔfliZ) (data not shown), indicating that the His₆-FliZ protein is functional.

The flhC and flhD genes were separately PCR amplified with primers HCf1B and HCr1S and HDf1S and HDr1P, respectively, using genomic DNA from KK1004 as a template. The amplified flhC DNA was digested with BamHI and SalI, while the amplified flhD DNA was digested with SalI and PstI. These two DNA fragments were inserted together into the BamHI/PstI site of pQE80L to obtain pQE80L-flhCD. This plasmid encodes an N-terminally hexahistidine-tagged FlhC protein (His₆-FlhC) and a nontagged FlhD protein. Introduction of this plasmid restored full motility to the cells of KK1004hDC (flhD::Tn10) (data not shown), indicating that both the His₆-FlhC and FlhD proteins encoded by this plasmid are functional.

The promoter region of the fliAZ operon was PCR amplified with primers KpnIfliA0 and IAEIB using genomic DNA from KK1004 as a template. The amplified product was digested with KpnI and BamHI and inserted into the corresponding site of pFZY1 to obtain pFZY1-P_fliA-lacZ. In this plasmid, the lacZ gene is transcribed from the class 2 and class 3 promoters of the fliAZ operon (P_fliA).

The promoter region of the nlpC gene was PCR amplified with primers PnlpCf1E and PnlpCr1B using the genomic DNA from KK1004 as a template. The amplified product was digested with EcoRI and BamHI and inserted into the corresponding site of pQE80L. In order to delete the T5 promoter on the vector sequence, the resulting plasmid was digested with XhoI and EcoRI, blunt ended with the Klenow fragment, and then self-ligated. In the resulting plasmid, pQE80L-ΔP_T5-P_nlpC, transcription initiates at the nlpC promoter (P_nlpC) and terminates at the T₀ terminator on the vector sequence, yielding an RNA product of 184 or 183 bases.

Similarly, the same region was PCR amplified with primers PnlpCf2E and PnlpCr2B or NLPCF and PnlpCr2B, digested with EcoRI and BamHI, and inserted into the corresponding site of pFZY1 to obtain pFZY1-P_nlpC-lacZ or pFZY1-ΔP_nlpC-lacZ, respectively. In pFZY1-P_nlpC-lacZ, the lacZ gene is transcriptionally fused to P_nlpC. pFZY1-ΔP_nlpC-lacZ has the same structure as pFZY1-P_nlpC-lacZ except that it lacks the P_nlpC region.

β-Galactosidase enzyme assay.

β-Galactosidase activity was assayed as described previously (20) using cells grown aerobically to exponential phase at 37°C in LB or MGC containing appropriate antibiotics. Each sample was assayed in triplicate.

Protein analysis.

Cells were grown to exponential phase at 37°C in LB or MGC, and the cultures were directly subjected to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE). SDS-PAGE and Western blotting of proteins were performed according to the method described previously (24). FliC (flagellin) was detected with an anti-FliC polyclonal antibody and an anti-rabbit horseradish peroxidase (HRP)-conjugated antibody (Santa Cruz Biotechnology, CA) using the ECL Plus Western blotting detection system (GE Healthcare, NJ). FLAG-tagged proteins were detected with an anti-FLAG M2 monoclonal antibody (Sigma, MO) and an anti-mouse HRP-conjugated antibody (Santa Cruz Biotechnology).

Affinity purification of His₆-FliZ.

His₆-FliZ protein synthesized in E. coli strain BL21 ΔslyD (laboratory stock) harboring pQE80L-fliZ was affinity purified according to the method described previously (44). The purified protein exhibited a single band at a position corresponding to approximately 21 kDa in an SDS-PAGE gel (data not shown).

In vitro transcription.

In vitro RNA synthesis was performed using E. coli RNA polymerase holoenzyme (Epicenter, WI) according to the method described previously (41, 44). The reaction mixture contained 40 mM Tris-HCl, pH 7.5, 150 mM KCl, 1 mM dithiothreitol (DTT), 10 mM MgCl₂, 50 U of RNase inhibitor, 100 μM (each) ATP, GTP, and CTP, 50 μM UTP with 4 × 10⁵ Bq [α-³²P]UTP (Institute of Isotopes, Budapest, Hungary), 1 μg of template DNA, 1 U of RNA polymerase, and various concentrations (0 to 1,250 nM) of His₆-FliZ. Mixtures without substrates were prepared in a 45-μl volume and incubated at 37°C for 10 min. The reaction was initiated by addition of 5 μl prewarmed substrate mixture. After incubation at 37°C for 10 min, the reaction was terminated by adding 50 μl of a stop solution (0.6 M sodium acetate [pH 5.5], 20 mM EDTA, 200 μg tRNA/ml). The transcripts were extracted with phenol and then precipitated with ethanol. The precipitated materials were electrophoretically separated in an 8% polyacrylamide gel containing 6 M urea. The labeled transcripts were visualized by autoradiography.

Gel mobility shift assay.

A 287-bp DNA fragment containing P_nlpC was PCR amplified from the KK1004 genomic DNA using primers PnlpCf2E and PnlpCr2B. The amplified product was gel purified and then labeled at the 5′ end with [γ-³²P]ATP (Institute of Isotopes) by T4 polynucleotide kinase and used as a probe. The DNA-binding activity of His₆-FliZ was assayed according to a method described previously (41). The binding reaction mixture (20 μl) contained a 0.2 nM concentration of the labeled probe DNA, 20 mM Tris-HCl (pH 8.0), 100 mM NaCl, 0.1 mM EDTA, 1 mM dithiothreitol, 1 mM MgCl₂, 0.1 mg/ml bovine serum albumin, 5% glycerol, and various concentrations (0 to 250 nM) of His₆-FliZ. The reaction mixture was incubated at 37°C for 30 min and subjected to electrophoresis on a native 5% polyacrylamide gel in 0.5× TBE buffer (25 mM Tris, 25 mM boric acid, 1 mM EDTA). Labeled DNAs were detected by autoradiography.

RESULTS

FliZ is required for efficient flagellar production in poor medium.

We showed previously that a disruption of the fliZ gene reduced class 2 operon expression more than 5-fold (23), whereas Saini et al. (37) reported that the fliZ mutation had only a moderate effect on class 2 operon expression. We noted that the culture media used were different between these two studies; that is, Kutsukake et al. used MGC, whereas Saini et al. used LB. This suggested the possibility that the effect of the fliZ mutation on flagellar production is influenced by the nutritional condition. In order to test this possibility, we examined first more carefully the effect of the fliZ mutation on class 2 transcription with the cells grown in LB and MGC. For this purpose, we used the P_fliA-lacZ transcriptional fusion gene on the single-copy plasmid pFZY1-P_fliA-lacZ. The P_fliA DNA contains both class 2 and class 3 promoters of the fliAZ operon. Therefore, in order to avoid transcription from the class 3 promoter, all the host strains that we used for this study possessed an in-frame deletion of the fliA gene. As shown in Table 4, the effect of the fliZ mutation on β-galactosidase activity was more pronounced in MGC (more than 5-fold) than in LB (less than 2-fold).

Table 4.

Effects of growth media and ydiV mutation on FliZ control of class 2 transcription^a

Medium	Genotypeb			Plasmid	β-Galactosidase activity (Miller units)
	ydiV	fliA	fliZ
LB	+	−	+	None	236 ± 8
	+	−	−	None	123 ± 21
	−	−	+	None	226 ± 60
	−	−	−	None	220 ± 26
MGC	+	−	+	None	31 ± 6
	+	−	−	None	6 ± 1
	−	−	+	None	248 ± 33
	−	−	−	None	251 ± 48
	+	−	−	pQE80K	76 ± 23
	+	−	−	pQE80K-fliZ	386 ± 60
	−	−	−	pQE80K	502 ± 59
	−	−	−	pQE80K-fliZ	519 ± 56

^aβ-Galactosidase activity was assayed with strains harboring pFZY1-P_fliA-lacZ.

^bStrains used were ydiV⁺ fliA fliZ⁺ strain KK1004iA, ydiV⁺ fliA fliZ strain KK1004iAZ, ydiV fliA fliZ⁺ strain KK1004ViA, and ydiV fliA fliZ mutant KK1004ViAZ. −, the gene is deleted.

Next, we examined motility and flagellin (FliC) production in LB and MGC. As shown in Fig. 1, the effect of the fliZ mutation on motility and flagellin level was also more pronounced in MGC than in LB. These results indicate that Salmonella cells require FliZ for efficient flagellar synthesis more strictly in poor medium than in rich medium.

Fig. 1.

Effect of the fliZ mutation on motility (A) and the steady-state level of flagellin (B). (A) Single colonies were stabbed onto an LB motility agar plate (left panel) or an MGC motility agar plate (right panel) and incubated for 4 h at 30°C. Strains used include KK1004 (wild type [WT]), KK1004iZ (ΔfliZ), KK1004V (ΔydiV), and KK1004ViZ (ΔydiV ΔfliZ). (B) When the bacterial growth reached logarithmic phase, the cells were sampled and separated by SDS-PAGE. Flagellin was detected by Western blotting with an anti-FliC polyclonal antibody. Strains used were the same as those used for panel A.

FliZ acts in a regulatory pathway upstream of YdiV.

Saini et al. (37) reported that the fliZ deletion caused an approximately 50% decrease in the FlhC-3×FLAG protein level in rich medium. Therefore, there is a possibility that inhibition of class 2 operon expression by the fliZ mutation in poor medium might be due to the decreased level of the FlhC protein. However, we could not observe any significant effect of the fliZ mutation on the FlhC-3×FLAG protein level in MGC (data not shown). Therefore, FliZ-dependent activation of class 2 operon expression is unlikely to be mediated by modulation of the intracellular FlhD₄C₂ level, at least in the strains used here.

Since YdiV was shown to be a mediator of the nutritional cue to flagellar synthesis (41), the above-mentioned results raised the possibility that YdiV is involved in the fliZ-mediated regulation of the flagellar regulon. In order to test this, the effect of the fliZ mutation on flagellar gene expression was examined in the absence of YdiV. As shown in Fig. 1 and Table 4, the negative effects of the fliZ mutation on motility, flagellin production, and class 2 gene transcription in MGC were all relieved in the ydiV mutant background. These results suggest the hypothesis that FliZ acts upstream of YdiV on the common regulatory pathway.

In order to further confirm this, the effect of FliZ overexpression from pQE80K-fliZ on class 2 operon transcription was examined with the strains harboring pFZY1-P_fliA-lacZ (Table 4). Introduction of the empty vector pQE80K enhanced P_fliA-lacZ expression. This is consistent with a previous report showing that the lacI gene on the plasmid upregulates at least some flagellar genes (8). In the ydiV⁺ background, FliZ overexpression enhanced P_fliA-lacZ expression 5-fold. In contrast, FliZ overexpression had no significant effect on P_fliA-lacZ expression in the absence of YdiV. These results conform to the hypothesis mentioned above. Taken together, we conclude that the FliZ protein is a direct or indirect antagonist of the anti-FlhD₄C₂ factor YdiV in transcriptional control of the class 2 flagellar operons. Interestingly, overexpression of FliZ in the ydiV⁺ strain did not attain a level of expression of P_fliA-lacZ equivalent to its expression level in the ydiV mutant. This is discussed below.

FliZ negatively regulates transcription of the ydiV gene.

As described in the introduction, Wozniak et al. (43) reported that ydiV transcription is under the negative regulation of the flhDC genes. Since their experiment was carried out in rich medium, we examined the effect of the flhD mutation on ydiV transcription in poor medium. As shown in Table 5, when the cells of the strain carrying the ydiV-lacZ transcriptional fusion gene on the chromosome were grown in MGC, the flhD::Tn10 mutation increased β-galactosidase activity 2.4-fold.

Table 5.

Effects of the flhD or fliZ mutation on the ydiV-lac expression^a

Medium	Plasmid	β-Galactosidase activity (Miller units) with mutationb
		None	flhD::Tn10	fliZ
LB	None	173 ± 61	404 ± 15	376 ± 34
MGC	None	156 ± 13	381 ± 10	290 ± 37
MGC	pQE80L	212 ± 19	336 ± 9	341 ± 9
MGC	pQE80L-flhCD	198 ± 17	191 ± 44	323 ± 20
MGC	pQE80L-fliZ	159 ± 10	188 ± 13	180 ± 22
MGC	pQE80L-ydiV	380 ± 6	385 ± 20	379 ± 35

^aAll strains used carried the ydiV-lacZ fusion gene on the chromosome, which was transduced from KK1004 ydiV-lac.

^bStrains used were wild-type KK1004, flhD::Tn10 strain KK1004hDC, and fliZ strain KK1004iZ.

Since the fliZ gene is under the positive control of FlhD₄C₂ (12, 23), FliZ is one of the most plausible candidates of the regulator involved in FlhD₄C₂ control of the ydiV gene. In order to test this, the effect of the fliZ mutation on ydiV-lacZ expression was examined (Table 5). As expected, the fliZ mutation increased β-galactosidase activity about 2-fold. Ectopic expression of FliZ from pQE80L-fliZ in the flhD::Tn10 mutant resulted in decreased β-galactosidase activity, whereas that of FlhC and FlhD from pQE80L-flhCD in the fliZ mutant had no significant effect on enzyme activity (Table 5). These results indicate that FlhD₄C₂ affects ydiV expression through its positive control of fliZ expression. It should be noted that the chromosomal ydiV-lacZ fusion gene was not completely turned off by FliZ expression from pQE80L-fliZ (Table 5). This is consistent with the result shown in Table 4. These results suggest that both FliZ-sensitive and FliZ-insensitive transcriptions are responsible for ydiV expression. This is discussed further below.

Because YdiV inhibits fliZ expression by acting as an anti-FlhD₄C₂ factor (41), the above-described result suggests that the ydiV gene should be autogenously activated. As expected, YdiV overexpression from pQE80L-ydiV enhanced ydiV-lacZ expression, and the flhD or fliZ mutation did not enhance further its expression (Table 5).

FliZ represses ydiV transcription from the nlpC promoter.

The ydiV gene is located just downstream of the nlpC gene (Fig. 2). Jonas et al. (16) reported results suggesting that two promoters located within the nlpC-ydiV intergenic region may be responsible for ydiV transcription. Recently, however, we showed that these promoters are inactive in our strains and that the ydiV gene is transcribed mainly from a promoter, P_nlpC, located upstream of the nlpC gene (41). Therefore, in this study, we examined the structure and function of P_nlpC further. A bioinformatics study using the software Neural Network Promoter Prediction (http://www.fruitfly.org/seq_tools/promoter.html) (36) suggested one σ⁷⁰-type promoter sequence, TTGAAA-N₁₈-TAAATT, whose predicted transcriptional start site is nucleotide C or A at bp position 30 or 29 upstream of the start codon of the nlpC gene (Fig. 2A). Unfortunately, our attempts to identify the transcriptional start site of the nlpC gene by primer extension analysis of mRNA or by the RNA ligation-mediated reverse transcription-PCR method were unsuccessful for unknown reasons. So, we analyzed the activity of P_nlpC using a single-copy promoter-probe vector, pFZY1.

Fig. 2.

Sequence of the nlpC promoter region (A) and the structure of the nlpC-ydiV operon (B). (A) The coding sequences of the btuD and nlpC genes are written in uppercase, whereas the sequence of their intergenic region is written in lowercase letters. The putative −35 and −10 sequences of P_nlpC are boxed. The predicted transcriptional start point of the nlpC promoter is labeled +1. The convergent arrows indicate the inverted repetitious sequences, which may act as a FliZ-binding site. Nucleotide sequences corresponding to the primers used for PCR amplification of the nlpC promoter region are underlined. (B) The btuD, nlpC, and ydiV genes are indicated by horizontal arrows. The chromosomal regions cloned on the recombinant reporter plasmids are shown by horizontal lines. The putative promoter sequence shown in panel A is present in the P_nlpC fragment but absent from the ΔP_nlpC fragment.

DNA fragments containing the upstream region of the nlpC gene with and without this promoter-like sequence were inserted into pFZY1 to obtain pFZY1-P_nlpC-lacZ and pFZY1-ΔP_nlpC-lacZ, respectively (Fig. 2B). In the wild-type cells grown in MGC, β-galactosidase activity was more than 20-fold higher from pFZY1-P_nlpC-lacZ (310 ± 37 Miller units) than from pFZY1-ΔP_nlpC-lacZ (15 ± 5 Miller units), suggesting that this promoter-like sequence is functional in Salmonella.

Next, we analyzed the effect of the fliZ mutation on β-galactosidase activity from pFZY1-P_nlpC-lacZ. In order to avoid any possible effect of YdiV on transcription from P_nlpC through the ydiV-fliZ regulatory loop, β-galactosidase activity was assayed in the ydiV mutant background. As a result, the enzyme activity was higher in the ydiV fliZ deletion strain (272 ± 51 Miller units) than in the ydiV fliZ⁺ strain (38 ± 9 Miller units), indicating that this promoter is responsive to negative regulation by FliZ. Note that, in the presence of FliZ, the enzyme activity was much lower in the ydiV mutant than in the wild-type strain. This indicates that YdiV activates its own expression by inhibiting FliZ expression. Furthermore, in the ydiV fliZ⁺ background, the enzyme activity from the chromosomal ydiV-lacZ fusion gene (Table 5) was higher than that from pFZY1-P_nlpC-lacZ, which suggests an additional promoter for ydiV expression on the chromosome. This promoter may be responsible for the FliZ-insensitive transcription of the ydiV gene described above.

FliZ is a repressor for the nlpC promoter.

P_nlpC activity was analyzed further by an in vitro transcription experiment using plasmid pQE80L-ΔP_T5-P_nlpC as a template. If the above prediction about the transcriptional start site is correct, a 184- or 183-nucleotide transcript is expected to be produced by transcription from P_nlpC in this plasmid. In the presence of σ⁷⁰ RNA polymerase, an RNA product of about 190 nucleotides was produced in addition to a 107-nucleotide RNA I derived from the vector sequence (Fig. 3, lane 2). This result indicates the validity of the above-mentioned prediction for P_nlpC.

Fig. 3.

Effects of FliZ on nlpC transcription in vitro. A plasmid carrying the T5 or nlpC promoter was transcribed in vitro with E. coli RNA polymerase holoenzyme with or without His₆-FliZ in the presence of [α-³²P]UTP. His₆-FliZ was added to the reaction mixture at the final concentrations indicated above the lanes. Synthesized RNAs were separated on an 8% polyacrylamide gel containing 6 M urea and detected by autoradiography. Positions of the T5, nlpC, and RNA I transcripts are indicated by arrows. nt, nucleotides.

In order to analyze the function of the FliZ protein in vitro, we purified the His₆-FliZ protein from E. coli cells harboring pQE80L-fliZ and examined its effect on in vitro transcription from P_nlpC. An empty vector, pQE80L, was also included in the reaction mixture as a control template, which produces a 230-nucleotide RNA from the T5 promoter. Addition of the purified His₆-FliZ protein into the reaction mixture reduced the amount of the nlpC transcript in a concentration-dependent manner, whereas that of the RNA product from the T5 promoter was not affected (Fig. 3, lanes 3 to 6). This result indicates that FliZ specifically represses transcription from P_nlpC. Interestingly, at the highest concentration of His₆-FliZ used in this study, the amount of RNA I was also somewhat reduced (Fig. 3, lane 6). This is discussed below.

FliZ is known to contain a SAM (sterile alpha motif)-like phage integrase domain (37), suggesting the possibility that FliZ has a DNA-binding ability. In order to test this, gel mobility shift assay of the ³²P-labeled DNA fragment containing P_nlpC was performed in the presence of the purified His₆-FliZ protein. A shifted band was observed in the presence of His₆-FliZ, and its amount increased in a concentration-dependent manner (Fig. 4, lanes 2 to 5). Band shift was not observed with the ³²P-labeled DNA fragment lacking P_nlpC (data not shown), indicating that FliZ binds specifically to the DNA region containing P_nlpC. Binding specificity was further examined by a competition experiment using unlabeled DNAs as competitors. An unlabeled DNA fragment containing P_nlpC behaved as an effective competitor (Fig. 4, lane 7), whereas an unlabeled DNA fragment without P_nlpC did not (Fig. 4, lane 8), indicating the high specificity of FliZ binding to the DNA region containing P_nlpC.

Fig. 4.

Gel mobility shift assay of the nlpC promoter DNA by FliZ. A 287-bp P_nlpC DNA fragment containing the nlpC promoter was 5′-end labeled with ³²P and used as a probe. Each mixture contained 0.2 nM labeled DNA. After incubation with His₆-FliZ at the final concentrations indicated above the lanes, the DNA-protein mixtures were separated on a native 5% polyacrylamide gel, and labeled DNA was detected by autoradiography. In the competition experiments (lanes 7 and 8), 2 nM concentrations of the unlabeled P_nlpC and ΔP_nlpC fragments shown in Fig. 2B were used. Positions of free and bound DNAs are indicated by arrows.

DISCUSSION

We showed that FliZ is required for efficient expression of the class 2 flagellar operons in poor medium in Salmonella (Table 4). Although Saini et al. (37) reported data suggesting that FliZ functions at the posttranslational level of FlhD₄C₂ by regulating its stability, we could not observe any significant effect of the fliZ mutation on the FlhC protein level (data not shown). Instead, we showed that FliZ binds to the promoter region of the nlpC-ydiV operon and represses its transcription (Fig. 3 and 4), which results in the decreased expression of the ydiV gene, encoding an anti-FlhD₄C₂ factor.

Our current model depicting a regulatory pathway involving FliZ and YdiV is summarized in Fig. 5A. According to this model, the fliZ and ydiV genes form a regulatory loop and are differentially expressed in response to nutritional cues. Since ydiV expression is enhanced in poor medium at the translational step (41), an increased level of the YdiV protein would activate its own transcription by inhibiting the FlhD₄C₂-dependent expression of the fliZ gene. Under this condition, the flagellar operons stay at a low expression level. On the other hand, in a rich nutritional environment where the amount of the YdiV protein is low, the FlhD₄C₂ complex is free from inhibition by YdiV, resulting in high expression of the FliZ protein, which in turn represses ydiV expression. This leads the flagellar operons to the actively expressed state. Thus, the FliZ-YdiV regulatory loop has a stable bipartite expression profile in switching between highly and weakly motile lifestyles in responding to nutrient availability (Fig. 5B).

Fig. 5.

Models illustrating the FliZ-dependent activation of the flagellar class 2 operons via control of the expression of the ydiV gene (A) and the interplay between YdiV and FliZ in motility control in response to nutritional conditions (B). Details are provided in the text.

In this study, we analyzed ydiV expression by focusing on the P_nlpC promoter, which is susceptible to FliZ repression. However, as described above, we suggest that an additional promoter insensitive to FliZ repression may also be involved in ydiV transcription. We anticipate that transcriptional readthrough from the btuCED operon upstream of the nlpC gene may be responsible for this transcription, because there is no apparent ρ-independent terminator-like sequence in the intergenic region between the btuD and nlpC genes (Fig. 2A).

FliZ is known to regulate the SPI1 genes positively (5, 14, 17) and the type 1 fimbrial genes negatively (6, 39). FliZ control of these genes was observed in rich medium, in which YdiV expression must be low. Therefore, it is reasonable to predict that this FliZ control may not be affected by the ydiV mutation. This prediction is consistent with the previous observation that the expression of one of the SPI1 genes, sipC, was not affected by the ydiV mutation (10). Therefore, we believe that FliZ control of these genes is not mediated by YdiV. FliZ may regulate these genes through a mechanism different from that of FliZ control of the flagellar genes.

FliZ from Xenorhabdus nematophila was shown to specifically bind to the promoter region of the flhDC operon and to function as an activator (29). Although X. nematophila FliZ shows 56% identity with Salmonella FliZ, they differ significantly from each other in their C-terminal sequences (data not shown). Since FliZ is not involved in the regulation of the flhDC operon in Salmonella (23), there is a possibility that the C-terminal portion of FliZ determines its DNA-binding specificity. This possibility is supported by our preliminary experiment showing that a C-terminally truncated FliZ protein does not retain DNA-binding ability (our unpublished result).

The C-terminal portion of FliZ contains a SAM-like phage integrase domain (37), which is also included in Cre, Flp, and XerD recombinases (9). These recombinases are known to recognize palindromic sequences (3, 40). Interestingly, a 14-bp palindromic sequence (CGTTTCTTGAAAAG, where underlining indicates complementary nucleotides) was found also at the nlpC promoter region (Fig. 2A). This sequence overlaps the putative −35 element of the nlpC promoter and thus may be a plausible target site of FliZ binding. A similar palindromic sequence (GAGTTCTTGAAGTG) exists around the −35 element of the RNA I promoter, whereas such a sequence is not observed at the T5 promoter region of pQE80L. This structural difference conforms to the results of Fig. 3 showing that the nlpC and RNA I promoters were susceptible to FliZ repression but that the T5 promoter was not. Identification of a recognition sequence of FliZ will help us to understand a molecular mechanism underlying FliZ control of the ydiV gene and reveal additional targets of FliZ on the bacterial genome.