Regulatory Effect of SlyA on rcsB Expression in Salmonella enterica Serovar Typhimurium

The antagonistic behavior of RcsB and SlyA on virulence gene expression led us to hypothesize that there is interplay between both regulators in a regulatory network and these could be considered coordinators of this process. Here, we report that the SlyA virulence factor influences motility behavior by controlling rcsB transcription from the P_rcsB promoter. We also demonstrate that SlyA negatively affects the expression of the rcsB gene by direct binding to P_rcsDB and P_rcsB promoters. We suggest that different levels of RcsB act as a switch between the virulent and attenuated forms of Salmonella, where high concentrations of the regulator tend to tilt the balance toward the attenuated form and low concentrations or its absence tilt it toward the virulent form.

ABSTRACT

The Salmonella enterica serovar Typhimurium RcsCDB system regulates the synthesis of colanic acid and the flagellum as well as the expression of virulence genes. We previously demonstrated that the rcsC11 mutant, which constitutively activates the RcsB regulator, attenuates Salmonella virulence in an animal model. This attenuated phenotype was also produced by deletion of the slyA gene. In this work, we investigated if this antagonistic behavior is produced by modulating the expression of both regulator-encoding genes. We demonstrated that SlyA overproduction negatively regulates rcsB transcription. A bioinformatics analysis enabled us to identify putative SlyA binding sites on both promoters, P_rcsDB and P_rcsB, which control rcsB transcriptional levels. We also determined that SlyA is able to recognize and bind to these predicted sites to modulate the activity of both rcsB promoters. According to these results, SlyA represses rcsB transcription by direct binding to specific sites located on the rcsB promoters, thus accounting for the attenuated/virulence antagonistic behaviors. Moreover, we showed that the opposite effect between both regulators also physiologically affects the Salmonella motility phenotype. In this sense, we observed that under SlyA overproduction, P_rcsB is repressed, and consequently, bacterial motility is increased. On the basis of these results, we suggest that during infection, the different RcsB levels produced act as a switch between the virulent and attenuated forms of Salmonella. Thereby, we propose that higher concentrations of RcsB tilt the balance toward the attenuated form, while absence or low concentrations resulting from SlyA overproduction tilt the balance toward the virulent form.

IMPORTANCE The antagonistic behavior of RcsB and SlyA on virulence gene expression led us to hypothesize that there is interplay between both regulators in a regulatory network and these could be considered coordinators of this process. Here, we report that the SlyA virulence factor influences motility behavior by controlling rcsB transcription from the P_rcsB promoter. We also demonstrate that SlyA negatively affects the expression of the rcsB gene by direct binding to P_rcsDB and P_rcsB promoters. We suggest that different levels of RcsB act as a switch between the virulent and attenuated forms of Salmonella, where high concentrations of the regulator tend to tilt the balance toward the attenuated form and low concentrations or its absence tilt it toward the virulent form.

INTRODUCTION

Salmonella enterica serovar Typhimurium virulence is directly related to the presence of pathogenicity islands (SPIs) containing virulence genes. Salmonella virulence is controlled by a complex interplay between different transcriptional regulators that modulate the expression of these genes (1). Among the major regulators involved in this process are RcsB, PhoP, PmrA, SlyA, and OmpR (1–6). Several reports demonstrated that the above-mentioned interactions coordinately control the expression of the genes involved in pathogen adhesion, invasion, replication, or survival within macrophages, in a specific time and space (7).

The SlyA transcriptional factor belongs to the MarR (multiple antibiotic resistance regulator) regulator family, whose members are distributed in both Archaea and Bacteria. The MarR family members regulate a large number of cellular processes, including resistance to antibiotics, organic solvents, disinfectants, and oxidative stress-generating agents (8, 9). These kinds of regulators are linked to the induction of virulence in pathogenic bacteria in humans and plants (8). The slyA gene was isolated from S. Typhimurium and characterized to confer a hemolytic phenotype to Escherichia coli K-12 (10, 11). Further studies revealed that the slyA gene of S. Typhimurium is required to control the expression of the SPI-2 virulence genes (12). Furthermore, it was reported that this transcriptional factor is also involved in the regulation of genes required for adhesion and survival within macrophages as well for resistance to H₂O₂, magainin 2, and polymyxin B (6, 10, 13–16). Notably, Libby et al. (10) reported that the slyA mutant shows a virulence attenuation phenotype in mouse infection assays.

The RcsB protein is the transcriptional regulator that belongs to the RcsCDB phosphorelay system (17). This is an unusual two-component system, because it consists of three proteins: RcsC and RcsD, which acts as sensor proteins, and RcsB, which is the response regulator (18, 19). Other components of the system are the upstream regulators RcsF and IgaA (17). RcsF is a lipoprotein anchored to the outer membrane that can transduce stress signals to RcsC but is not fully required for RcsCDB system activation (20). We demonstrated that the Salmonella rcsB gene is controlled by two promoters, P_rcsDB and P_rcsB, whose activities depend on the bacterial growth phase, causing the rcsB regulator to be expressed with rcsD in a bicistronic transcript and/or as a monocistronic mRNA (21). In addition, we showed that high levels of the active RcsB form control the rcsDB transcript, since RcsB is able to bind to the P_rcsDB promoter, thus inhibiting its activity, but not to the P_rcsB promoter (22).

Interestingly, it was reported that the activation of the RcsB regulator affects both invasion and replication inside eukaryotic cells, repressing SPI-2 genes (23–25). Therefore, the presence of the RcsB factor causes Salmonella attenuation, while deletion of the rcsB gene increases the virulence phenotype (23, 26). In contrast, an opposite effect was reported for SlyA, whose overproduction or activation increases virulence via positive control of SPI-2 genes (12), as the virulence of a slyA gene mutant was attenuated in a mouse infection model (10). On the basis of the antagonistic behaviors of RcsB and SlyA on virulence, we decided to investigate if there is interplay between these transcriptional factors in a regulatory network that could account for the attenuated/virulent phenotype. Since both proteins are transcriptional factors, we hypothesized that such an effect would be based on a cross modulation of their encoding genes.

Here, we report that the SlyA transcriptional factor exerts negative and direct control on rcsB transcription. Our results revealed a putative SlyA binding sequence upstream of the P_rcsDB and P_rcsB promoters. We demonstrated that this regulator binds to DNA fragments containing these promoters. We also found that the repression of rcsB transcription, produced by SlyA overproduction, affected Salmonella motility behavior. Our results enable us to suggest that during the Salmonella infection steps, bacteria attain different RcsB protein levels, controlled by SlyA, which acts as a switch between the virulent and attenuated bacterial states. The virulent state is also favored during the stationary phase by increased SlyA concentrations, explaining the antagonist effects of the RcsB and SlyA regulators on Salmonella pathogenesis.

RESULTS

SlyA downregulates rcsB expression.

The hypothesis that RcsB and SlyA affect each other's activity led us to investigate the role of SlyA on rcsB expression. To test this assumption, we measured the β-galactosidase activity produced by the chromosomal transcriptional rcsB::lacZY fusion in wild-type and slyA genetic backgrounds. As shown in Fig. 1, β-galactosidase levels in the slyA mutant were 1.6-fold higher than those observed in the wild-type strain, suggesting that the SlyA regulator is able to repress transcription of the rcsB gene. To confirm that the effect displayed in the slyA mutant was not due to a polar effect produced by the gene deletion, we complemented in trans the slyA mutation. We observed that the rcsB expression was repressed by slyA overexpression, since the β-galactosidase levels were 3.2-fold lower in the wild-type strain in the presence of the pslyA plasmid and isopropyl-1-thio-β-d-galactopyranoside (IPTG). Moreover, in the slyA mutant harboring pslyA, β-galactosidase levels were also repressed by IPTG addition, while maximal repression levels were observed in the wild-type strain overexpressing the slyA and rcsB plasmidic genes (Fig. 1). We suggest that the negative effect of SlyA was independent of the RcsB autoregulatory mechanism exerted on its own transcription, since the levels of rcsB expression in the wild type and in the slyA mutant, both harboring prcsB*, were similar under arabinose induction (Fig. 1B). These results demonstrate that SlyA represses rcsB transcription, suggesting that SlyA acts on one or both of the rcsB promoters.

FIG 1.

rcsB gene transcription is modulated by the SlyA regulator. (A) β-Galactosidase activity of the rcsB::lacZY chromosomal transcriptional fusion was investigated in the wild-type (EG14932) and slyA (MDs1138) strains harboring the pslyA plasmid and grown at 37°C in LB medium in the absence (−) or presence (+) of 0.5 mM IPTG. (B) rcsB::lacZY expression was measured in the wild-type (EG14932) strain harboring pslyA, prcsB*, or both plasmids as well as in the slyA (MDs1138) mutant containing the prcsB* plasmid to analyze the effect of RcsB autoregulation. These strains were grown under the same conditions as in panel A, under induction (+) or not under induction (−) of slyA and rcsB plasmid genes with IPTG and arabinose, respectively. In these assays, the empty vectors were used as controls with IPTG and arabinose treatments, respectively. The data correspond to the averages from three independent experiments, performed in duplicate. The error bars correspond to standard deviations. *, P = 0.05 by Tukey’s test.

Bioinformatics analysis of the rcsB promoters.

To determine whether the rcsB promoters were affected by SlyA, we decided to analyze both rcsB promoter sequences. For this purpose, we performed an alignment between the P_rcsDB or P_rcsB promoter regions and the previously described SlyA consensus box (15), using the multiple expectation maximization for motif elicitation (MEME) tools (27). First, we studied the 248-bp region upstream of the rcsDB operon coding region, corresponding to the P_rcsDB promoter, and we observed the presence of a putative SlyA box located at bp −200 of the rcsD start codon and bp −127 of the conserved −35 box (Fig. 2A). It is important to mention that rcsD is the first gene in the rcsDB operon controlled by the P_rcsDB promoter. The results obtained showed that this putative binding site displays 66.67% identity with the SlyA consensus box, keeping intact the first half of the palindrome (Fig. 2A and B, lower portions). On the other hand, when we aligned the SlyA consensus box with the P_rcsB promoter region, we found another sequence that might serve as a SlyA binding site. This new site was located within the rcsD encoding sequence, at 8 bp upstream of the rcsB translational start codon (Fig. 2B). The location of this putative SlyA binding box suggests that it might function as a negative regulation site for rcsB transcription (Fig. 2B). Furthermore, this putative SlyA binding site showed a 75% DNA sequence identity with the palindromic consensus sequence reported by Stapleton et al. (15) (Fig. 2B). These results indicate that both transcript rcsDB and rcsB could be controlled by the SlyA virulence transcriptional factor.

FIG 2.

Sequence analysis of the rcsB promoter regions. DNA sequence (upper portions) and alignment (lower portions) of the SlyA consensus box with the rcsDB (A) and rcsB (B) promoter regions (248 bp and 163 bp, respectively). The start codon of the rcsDB operon or rcsB gene and −35 and −10 boxes are indicated with bold letters. The SlyA binding sequences found in this work are boxed. The numbering refers to the start codons of the rcsDB operon (A) and rcsB gene (B). In the sequence alignments, the conserved nucleotide positions of putative SlyA binding sites are boxed and the −35 and −10 regions are highlighted in bold.

SlyA overproduction affects the activity of P_rcsDB and P_rcsB promoters.

To examine the role of SlyA on the activity of each of the rcsB promoters, we decided to study rcsB expression under SlyA overproduction, using single promoter mutants (Fig. 3A). These assays were carried out using the rcsB::lacZY chromosomal transcriptional fusion, in which the rcsB gene was eliminated to avoid any effect of this regulator on its own expression (Fig. 3A). When the P_rcsDB mutant was analyzed, where rcsB is expressed by P_rcsB activity, we observed that the slyA overexpression produces a strong decrease in β-galactosidase levels with respect to those obtained without IPTG or with the empty vector (Fig. 3B). Interestingly, when rcsB was transcribed from the P_rcsDB promoter in the P_rcsB mutant, the β-galactosidase levels were decreased by slyA overexpression compared to those in the control without IPTG or with the empty vector (Fig. 3B). Moreover, the level of rcsB repression in each mutant as a result of the increased amount of SlyA was lower than that in the wild-type background grown under the same condition (Fig. 3B). These data demonstrate that SlyA modulates the activity of both promoters, P_rcsDB and P_rcsB, downregulating not only rcsB but also rcsD transcription.

FIG 3.

Effect of slyA overexpression on P_rcsDB or P_rcsB promoter activity. (A) Graphic representation of the mutant backgrounds in which rcsB transcriptional expression was determined. The empty spaces between brackets correspond to each promoter deletion. (B) β-Galactosidase activity of the rcsB::lacZY chromosomal transcriptional fusion was investigated in the wild-type (EG14932) strain and P_rcsDB (MDs1026) and P_rcsB (MDs1027) promoter mutants carrying the pslyA plasmid and grown to stationary phase at 37°C in LB medium in the absence (−) or presence (+) of 0.5 mM IPTG. The empty vector was used as control, with IPTG treatment. The data correspond to the averages from three independent experiments, performed in duplicate. The error bars correspond to standard deviations. *, P = 0.05 by Tukey’s test.

SlyA protein binds to rcsB promoters in vitro.

Based on the results described above, we investigated whether SlyA could bind to the P_rcsDB and P_rcsB promoters and modulate rcsB and rcsD expression. To this end, we performed electrophoretic mobility shift assays (EMSAs) using PCR products containing the P_rcsDB or P_rcsB promoter region, both of which harbor the putative SlyA binding sites previously identified (467 bp and 126 bp, respectively). In addition, we used 122-bp-deleted and 102-bp-deleted PCR products of the P_rcsDB and P_rcsB promoter regions, respectively, lacking the putative SlyA regulatory box as negative controls.

As shown in Fig. 4, the SlyA protein specifically bound to both P_rcsDB and P_rcsB promoters. We observed a slower migration band when each PCR product was incubated with the SlyA-His₆ protein. These bands were more intense at higher SlyA-His₆ concentrations, while the staining intensity of the band for the PCR product alone decreased. These observations suggested that the lower migration bands corresponded to the DNA/SlyA-His₆ complex, since these bands were not observed when the PCR products lacking the SlyA binding box were analyzed (Fig. 4). Remarkably, we noticed that high concentrations of SlyA-His₆ were required to observe the P_rcsB promoter shift compared to the optimal concentration to shift the P_rcsDB promoter (Fig. 4). Taken together, these results demonstrate that SlyA binds to the sequences in the P_rcsDB and P_rcsB promoter regions identified by the bioinformatics analysis.

FIG 4.

SlyA protein binds to the P_rcsDB and P_rcsB promoters. The electrophoretic mobility shift assay (EMSA) was performed using 2 pmol of P_rcsDB (A), P_rcsB (B) or control P_rcsDB PCR products and different SlyA-His₆ protein concentrations (100, 200, or 300 nM, from left to right); the control was performed without protein (−). The DNA fragments were separated on 6% polyacrylamide gels and visualized with ethidium bromide to detect the SlyA/DNA complex.

Physiological role of rcsB repression mediated by SlyA.

We hypothesized that SlyA repression of rcsB transcription affects the motility behavior of Salmonella, since both regulators also participate in the regulation of flagellar gene expression (6, 28–30). To investigate this assumption, we studied the swimming phenotype of the S. Typhimurium wild-type strain and rcsB and rcsC11 mutants, as well as the wild-type, rcsB, P_rcsDB, and P_rcsB strains harboring the pslyA plasmid. The rcsB and rcsC11 mutants were used as motility phenotype controls for the effect of RcsCDB system activation. As shown in Fig. 5, the rcsB mutant displayed a migration rate 1.43-fold greater than that of the wild-type strain, while the rcsC11 mutant did not swim, which is in agreement with previous results (28, 31). In this assay, the wild-type strain overexpressing slyA displayed a 1.25-fold lower migration rate than that without slyA induction (Fig. 5, with and without IPTG, respectively). We also detected that the motility of the P_rcsB mutant harboring pslyA decreased 1.34-fold in the presence of IPTG compared to that of the same mutant growing in the absence of IPTG (Fig. 5). Interestingly, an opposite motility phenotype was observed for the P_rcsDB mutant when the slyA gene was overexpressed, displaying a 1.3-fold increase in the migration rate compared to that of the same strain without IPTG addition (Fig. 5). Moreover, the increased migration levels were similar to those observed in the rcsB mutant overexpressing slyA (Fig. 5). These data indicate that the increased motility displayed by the P_rcsDB mutant might be due to the SlyA repression effect on P_rcsB, a weak activity promoter. Taken together, these results demonstrated that SlyA activation significantly affects the motility behavior only at low RcsB levels.

FIG 5.

Salmonella RcsB-dependent motility phenotype. (A) Motility rates were determined in the wild-type S. Typhimurium 14028s strain and rcsB (EG12711) and rcsC11 (EG14873) mutants (gray bars), as well as in wild-type, rcsB (EG12711), P_rcsDB (MDs1017), and P_rcsB (MDs1018) strains harboring the pslyA plasmid. Colonies from LB agar medium containing (+) or not containing (−) 0.5 mM IPTG were picked with a toothpick, placed in motility medium, and incubated at 37°C. The migration diameter was measured after 4 h of incubation. The data correspond to the average values from three independent experiments. The error bars correspond to standard deviations. *, P = 0.05 by Tukey’s test. (B) A picture of a representative swimming plate is shown. (C) dps gene transcription is modulated by the SlyA regulator in an RcsB-dependent pathway. β-Galactosidase activity of the dps::lacZY chromosomal transcriptional fusion was investigated in the wild-type (MDs1568) and rcsB (MDs1569) strains harboring the pslyA plasmid and grown at 37°C in LB medium in the absence (−) or presence (+) of 0.5 mM IPTG. Here, the strains containing the pUHE2-21 vector were used as the controls. The data correspond to the averages from three independent experiments, performed in duplicate. The error bars correspond to standard deviations.

To confirm the SlyA repression effect on rcsB, we used another RcsB-dependent gene, dps (32). When the levels of dps::lacZY transcription were determined, we observed that slyA overexpression decreased dps expression in the wild type but not in the rcsB strain (Fig. 5C).

Here, we have demonstrated that the repression of rcsB by the SlyA regulator results in the modulation of RcsB-dependent genes.

DISCUSSION

It is known that the Salmonella infection process is controlled by a network of transcriptional regulators acting to modulate many virulence genes. Numerous interplays are described: (i) PhoPQ and PmrAB systems interact to remodel the lipid A in order to resist the bactericidal action of cationic peptides and to replicate within macrophages (33, 34); (ii) PhoPQ and RcsCDB interact to coordinately control the expression of pag and ugd genes required for lipid A modification and colanic acid synthesis and several virulence genes (35, 36); (iii) PhoPQ and SpiR-SsrB interact to control the genes located in SPI-2, which are necessary for the maturation of Salmonella-containing vacuoles (SCV) within the host cell (2); (iv) SpiR-SsrB and OmpR/EnvZ interact to control genes encoding the SPI-2-type III secretion system, required for bacterial replication inside macrophages and for establishing systemic infection (5, 37, 38); (v) SirA/BarA interacts with CsrAB to induce hilA expression, whose product acts as a master regulator of the SPI-1 genes required to establish eukaryotic cell invasion (39). However, no RcsB/SlyA interaction has yet been reported.

Libby et al. (10) demonstrated that SlyA induces the SPI-2 virulence genes, as slyA deletion attenuated Salmonella virulence. However, an opposite effect was reported for RcsB, whose activation represses both invasion and intracellular replication genes, attenuating Salmonella virulence (23–25). Therefore, rcsB gene deletion from the bacterial chromosome induces a greater infective capacity. On the basis of this observation, we hypothesized that the opposite effects of SlyA and RcsB on virulence genes may be due to a cross regulation mechanism exerted by one of these factors on the gene that codes for the opposite factor, and vice versa.

In this work, we report for the first time that an overlapping regulatory effect on Salmonella motility behavior exists between SlyA and RcsB regulators. Previously, we demonstrated that activation of the P_rcsDB promoter starts in early-exponential-growth phase and is maintained along with growth, while the P_rcsB promoter becomes active in stationary phase (21). On the other hand, it was reported that physiological slyA gene expression was induced during the stationary-growth phase (13). Considering these data, we assumed that SlyA affects rcsB expression. We found that P_rcsDB and P_rcsB promoter sequences harbor SlyA regulatory boxes as imperfect palindrome sequences, with 66.67% and 75% identity, respectively, with the consensus. Accordingly, other authors also reported an imperfect but functional palindrome for SlyA binding at the ugtL and slyA promoter regions (15, 40), indicating that those motifs identified in the P_rcsDB and P_rcsB sequences might function as SlyA regulatory sites. We confirmed these results by EMSAs, where the SlyA was able to shift the mobility of P_rcsDB and P_rcsB PCR products. Meanwhile, deletion of SlyA binding nucleotides in the P_rcsDB and P_rcsB PCR products prevented SlyA from binding to these products. These results indicate that the predicted sequences are effectively the site of SlyA regulation on the rcsB promoters. It is important to note that the formation of the P_rcsB/SlyA-His₆ complex required higher protein levels than required for the P_rcsDB/SlyA complex. This result is consistent with a higher affinity of SlyA for the P_rcsDB promoter site. This possibility is supported by the different levels of identity of each SlyA binding box that were found when they were aligned with the SlyA consensus site described by Stapleton et al. (15).

Based on the results described above, we studied which of the rcsB promoters was modulated by SlyA. We demonstrated that overexpression of slyA resulted in decreased rcsB expression levels from both promoter mutants, suggesting that SlyA represses rcsB transcription mainly during stationary phase. Moreover, when the slyA gene was removed from the Salmonella chromosome, the rcsB expression level increased 1.6-fold. Similarly, rcsD expression is consequently repressed by large amounts of SlyA only when the P_rcsDB activity is affected as expected. In agreement with Dolan et al. (41), who reported that the SlyA/DNA complex produces a conformational change to DNA topology and its bending, we propose that this effect can prevent the recruitment of RNA polymerase to rcsB promoters and prevent their transcription.

To evaluate the physiological importance of SlyA on the RcsCDB system components, we analyzed motility, since SlyA and RcsB are involved in the control of this process. Our results led us to conclude that SlyA increases bacterial motility by repressing rcsB transcription from the P_rcsB promoter. Interestingly, Mouslim and Hughes (30) observed that SlyA had no effect on the motility phenotype. In fact, they reported that this regulator represses the activity of the P5_flhDC flagellar promoter, which in turn controls flhDC operon expression during the stationary phase. These results may differ from ours due to the fact that Mouslim and Hughes (30) only investigated the effect on motility with SlyA absence. In previous studies, we demonstrated that rcsB is constitutively expressed at very low levels by P_rcsB activity. However, these basal levels are not sufficient to maintain certain motility in the wild-type strain. Meanwhile, the complete deletion of the rcsB gene produces an exacerbated motility phenotype (22). Taking into account that (i) the activity of P1_flhDC ensures the synthesis of the flagellum to positively control motility, (ii) P1_flhDC is negatively affected by RcsB, even at basal levels, and (iii) high levels of SlyA repress the expression of rcsB from its two promoters, hence no repression of P1_flhDC takes place, we here propose that SlyA activation indirectly modulates Salmonella motility in a RcsB-dependent pathway. A similar effect was observed using another RcsB-dependent gene, dps; slyA overexpression decreased the levels of dps in the wild type but not in the rcsB strain. Therefore, our results complement the findings of Mouslim and Hughes (30) demonstrating that there is a negative interaction of SlyA and RcsB in the control of numerous virulence factors depending on the environment, exemplified by the regulation of flagella and motility behavior.

The results obtained in this work represent an important advance in the knowledge of the role of the RcsCDB system in Salmonella pathogenesis and of the role that SlyA plays in the control of the RcsCDB system components in this bacterium. Thus, we suggest that different levels of RcsB act as a switch between the virulent and attenuated forms of Salmonella, represented in the model in Fig. 6: high concentrations of the regulator tilt the balance toward the attenuated form (Fig. 6A), while low concentrations or their absence tilt it toward the virulent form (Fig. 6B). Based on our results, SlyA represents one additional mechanism by which Salmonella controls this balance, through the regulation of rcsB transcription.

FIG 6.

Model of the regulatory overlapping effect of RcsB and SlyA in Salmonella virulence control. Different levels of RcsB act as a switch between the virulent and attenuated states of the bacteria: high concentrations of the regulator tip the balance toward the attenuated state (A), while the absence or low concentrations of the regulator tip the balance toward the virulent state (B). SlyA represents one additional mechanism by which Salmonella controls this balance through the RcsCDB system, since SlyA represses rcsB expression, tilting the balance toward the virulent state. ↑, induction; ↓, repression.

MATERIALS AND METHODS

Bacterial strains, molecular techniques, and growth conditions.

The bacterial strains used in this work are listed in Table 1. Bacteria were routinely grown in Luria-Bertani (LB) broth or on agar plates. The double mutant strain construction was conducted by P22 phage-mediated transduction (42). Standard methods for DNA recombination and bacterial growth were used, as previously described (43). β-Galactosidase assays were performed according to the protocol developed by Miller (44). Kanamycin (50 µg/ml), ampicillin (50 µg/ml), and chloramphenicol (25 µg/ml) were added to the medium for antibiotic selection.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Descriptiona	Reference or source
Strains (S. enterica serovar  Typhimurium)
14028s	Wild type	47
EG14932	rcsB::lacZY	21
MDs1026	rcsB::lacZY PrcsDB::Cm	22
MDs1027	rcsB::lacZY PrcsB::Cm	22
EG14078	slyA::Cm	40
MDs1138	rcsB::lacZY slyA::Cm	This work
EG12711	rcsB::Cm	22
EG14873	rcsC11	23
MDs1017	PrcsDB::Cm	22
MDs1018	PrcsB::Cm	22
MDs1568	dps::lacZY	32
MDs1569	dps::lacZY rcsB::Cm	32
Plasmids
pUHE2-2lacIq	reppMB1 Apr lacIq	48
prcsB	pUHE2-21 lacIq containing rcsB gene	21
pBAD33	pBR322 ori, arabinose induction, Cm	49
prcsB*	pBAD33 containing rcsB gene	This work
pslyA	pUHE2-21 lacIq containing slyA gene	40
pACYCDuet-1	P15A ori lacI T7lac Cm	Novagen
pslyA-His6	pACYCDuet-1 containing slyA gene	This work

^aGene designations are summarized by Sanderson et al. (50).

Plasmid construction.

Plasmid pslyA-His6 was constructed by amplifying the slyA gene from wild-type strain 14028s chromosomal DNA with primers 8081 (CGGATCCTTGGAATCGCCACTAGGTTC) and 8082 (CCCAAGCTTAATCGTGAGAGTGCAATT). Then, the PCR product was digested with BamHI and HindIII and subsequently cloned into the corresponding sites of pACYC-Duet1 vector (Novagen), which encodes the six-histidine tag sequence. The correct slyA-His₆ DNA sequence was confirmed by sequencing.

SlyA purification.

E. coli BL21(DE3) harboring plasmid pslyA-His6 was grown at 37°C to an optical density at 600 nm (OD₆₀₀) of 0.2, and then 0.5 mM isopropyl-1-thio-β-d-galactopyranoside (IPTG) was added and the culture was incubated for another 3 h at 30°C. Cells were harvested by centrifugation and resuspended in 5 ml of purification buffer (300 mM NaCl, 30 mM NaH₂PO₄, pH 8.3) and then were disrupted with a French press. The cell lysate was mixed with a HIS-Select nickel affinity gel (Sigma-Aldrich) according to the manufacturer instructions and incubated overnight at 4°C. Then, the resin containing proteins was washed 3 times with lysis buffer including 15 and 30 mM imidazole, followed by centrifugation at 3,000 × g. Finally, the SlyA-H₆ protein was eluted from the nickel-nitrilotriacetic acid (Ni-NTA) resin using 100 mM imidazole, which was stored in glycerol (50% [vol/vol]) at −70°C.

Electrophoretic mobility shift assays.

EMSAs were carried out as previously described (45). Briefly, two fragments containing the putative SlyA binding sites of P_rcsDB and P_rcsB promoters (467 bp and 126 bp, respectively) were generated by PCR using primers Fwd P_rcsDB (CCGCTCGAGCGGGTTAAATTGATGAAATTC)/Rev P_rcsDB (CGCGGATCCTTATGTTTACGACCTGTAAAAG) and Fwd P_rcsB (CCGCTCGAG GGTACCCGGCAAGCAGTTATGTG)/Rev P_rcsB (CGCGGATCCGTATTGGGCTACCTTGCTACAG), respectively, with wild-type 14028s chromosomal DNA as the template. In addition, two smaller fragments of the P_rcsDB and P_rcsB promoters lacking the SlyA binding site were used as controls (122 bp and 102 bp, respectively), generated by PCR using the primer sets Fwd P_rcsDB control (CTCGAGCCCGTCCCGCCGACGGAGCGCG)/Rev P_rcsDB and Fwd P_rcsB/Rev P_rcsB control (TTGACGTAGGCGTCAATGTCGC), respectively. These DNA fragments were incubated with 100, 200, or 300 nM of SlyA-H₆ protein in binding buffer (25 mM Tris-HCl [pH 8], 50 mM NaCl, 5 mM MgCl₂, 5 mM dithiothreitol [DTT], and 10% glycerol) at room temperature for 20 min. The samples were run on a 6% polyacrylamide nondenaturing gel using 0.38× Tris-borate-EDTA buffer at 8 mA for 3 h. Gels were stained with ethidium bromide and visualized under UV irradiation.

Motility assays.

Swimming plates were performed according to the protocol previously described (46). Briefly, single colonies from overnight streak plates were placed into swimming agar using toothpicks (10 g/liter tryptone, 5 g/liter NaCl, and 0.35% agar) and incubated at 37°C for 4 h. An average of at least eight independent colonies from each strain were assayed.