|Home | About | Journals | Submit | Contact Us | Français|
The pathogen Vibrio cholerae uses a large number of coordinated transcriptional regulatory events to transition from its environmental reservoir to the host and establish itself at its preferred colonization site at the host intestinal mucosa. The key regulator in this process is the AraC/XylS family transcription factor, ToxT, which plays critical roles in pathogenesis, including the regulation of two type IV pili, the anticolonization factor mannose-sensitive hemagglutinin and the toxin-coregulated pilus. Previously, it was thought ToxT required dimerization in order to effect transcriptional regulation at its cognate promoters. Here, we present evidence that ToxT directly represses transcription of the msh operon by binding to three promoters within this operon and that dimerization may not be required for transcriptional repression of target promoters by ToxT, suggesting that this regulator uses different mechanisms to modulate the transcriptional repertoire of V. cholerae.
Vibrio cholerae is a gram-negative bacterium normally found in aquatic reservoirs. It is also the etiologic agent of cholera, a severe dehydrating diarrhea that affects hundreds of thousands of people each year. Cholera is characterized by voluminous diarrhea caused by the action of the cholera toxin (CT), produced by vibrios colonizing the mucosa of the small intestine. The transitions between the stages of the infectious cycle of this pathogen, between colonization, toxin production, and dissemination back into the aquatic reservoir, is governed by a transcriptional cascade that regulates the production of virulence factors such as CT and the biogenesis of the colonization determinant the toxin-coregulated pilus (TCP). In addition to these provirulence factors, some elements, including the mannose-sensitive hemagglutinin (MSHA) pilus and quorum-sensing systems, which as “anticolonization factors” can have a negative influence on pathogenesis, are downregulated during infection (12, 16).
The keystone of the transcriptional regulatory system that governs the phenotypic shifts in this transition from aquatic organism to human pathogen is the AraC/XylS family transcription factor ToxT, which is the major regulator of pathogenesis activated in response to entry into the intestine (5, 11). At the end of this cascade are the biogenesis systems for two related but functionally antagonistic type IV pili, the TCP, and MSHA pili. The tight regulation of pili biogenesis is critical for V. cholerae infection, as TCP is necessary for colonization of the host intestinal mucosa (10, 26). The presence of MSHA pili, on the other hand, is actively detrimental to survival in the host, since it binds secretory immunoglobulin A antibodies that, synergistically with the mucin glycocalyx of the intestinal mucosa (17), excludes pathogens such as V. cholerae from the epithelium. We have previously demonstrated that ToxT plays an important part in V. cholerae fitness in infection by resolving the conflicting pressures of these two type IV pili functions by downregulating the production of MSHA, which if present act as potent anticolonization factors (12). In addition, ToxT activates elements of the TCP biogenesis pathway, including the prepilin leader peptidase TcpJ, which then posttranslationally degrades MshA prepilin, further suppressing MSHA function during infection (13). Although the mechanism of ToxT-mediated transcriptional activation has been extensively studied, very little is known about the requirements for anticolonization factor suppression, the latest documented role played by this versatile virulence regulator in V. cholerae.
In the present study, we demonstrate that ToxT directly regulates the expression of the MSHA biosynthesis gene cluster (msh genes) by binding to three different promoters within the locus. Interestingly, ToxT is able to repress the transcription of msh genes in the presence of the ToxT-dimerization inhibitor virstatin. Repression is maintained even by a truncated ToxT variant completely lacking the N-terminal dimerization/environmental sensing domain. Both virstatin and dimerization domain deletion abolish transcriptional activation at the TCP operon tcpA promoter, suggesting that ToxT uses different mechanisms in the transcriptional regulation of pro- and anticolonization factors.
Escherichia coli and V. cholerae strains were all propagated in LB with appropriate antibiotics at 37°C unless otherwise stated. To induce virulence gene expression in V. cholerae, cells were propagated in AKI media and growth conditions (15). All plasmids and oligonucleotides used in the present study are available upon request. The virstatin concentration used here was 100 μM.
pBAD-toxT plasmid (9) was transformed into the mutator XL1-Red E. coli by electroporation according to the manufacturer's instructions (Stratagene). Pools were then made of transformants, and plasmids were purified from each pool for transformation into a PtcpA-lacZ E. coli reporter strain. These were then plated on selective minimal agar containing triphenyl-tetrazolium chloride (TTC) to a final concentration of 0.0025% and lactose as a carbon source. Lac− colonies were then identified, and plasmids were extracted and transformed individually into E. coli containing a PmshH-lacZ plasmid. These cells were then plated again on TTC-lactose agar, and Lac− bacteria were identified. pBAD-toxT mutants that could not activate PtcpA but could still repress PmshH were then isolated and sequenced, and their activity was confirmed in PtcpA and PmshH reporter strains by β-galactosidase activity assays (20).
For Lux reporter assays, cells grown to late logarithmic phase and then withdrawn were normalized for growth using the optical density at 600 nm (OD600), and the luminescence was determined by using a Bio-Tek Synergy HT spectrophotometer. Lux expression is reported as light units/OD600.
E. coli cells containing a plasmid overexpressing maltose-binding protein fused to the N terminus of ToxT (MBP-ToxT) were grown at 37°C and induced with 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside) at 16°C. MBP-ToxT protein was purified through amylose columns according to the manufacturer's instructions (New England Biolabs). PCR products of ~400 bp containing mshH, mshI, mshB, and tcpA promoter regions were digested with EcoRI and end labeled using [α-32P]dATP and the Klenow fragment of DNA polymerase I. Binding reactions contained 0.1 ng of DNA and MBP-ToxT proteins in a buffer consisting of 10 mM Tris-HCl (pH 7.9), 1 mM EDTA, 1 mM dithiothreitol, 60 mM KCl, and 30 mg of calf thymus DNA/ml. After 20 min of incubation at 37°C, samples were size fractionated using 5% polyacrylamide gels in 0.5× TAE buffer (20 mM Tris-acetate, 1 mM EDTA; pH 8.5). Where indicated, the competitive nonradiolabeled PCR products (1 μg) were included in the reaction. The radioactivity of free DNA and ToxT-DNA complexes was visualized by using a Storm B840 PhosphorImager (Molecular Dynamics).
V. cholerae mannose-sensitive hemagglutination was assayed as described previously (6). Briefly, defibrinated sheep erythrocytes washed in phosphate-buffered saline (PBS) were resuspended in Krebs-Ringer Tris (KRT) buffer (7.5 g of NaCl, 0.383 g of KCl, 0.318 g of MgSO4·7H2O, and 0.305 g of CaCl2 in 1 liter of distilled H2O) to a final concentration of 10% (vol/vol). Erythrocyte suspensions were then mixed with mid-logarithmic-phase cultures of V. cholerae that had been serially diluted in KRT buffer. Samples checked for hemagglutination after 2 h.
Relevant V. cholerae strains were grown under inducing AKI and noninducing LB conditions at 37°C as previously described (13). Cultures were then normalized by using the OD600, and proteins were fractionated by size using sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Proteins were then transferred to a nitrocellulose membrane and immunoblotted using polyclonal rabbit αTcpA antibody or αMshA antibody.
ToxT is the central regulator of a well-studied transcriptional activation cascade whereby environmental cues are transmitted by the TcpPH and ToxRS membrane complexes to activate toxT transcription (3-5, 9, 10, 19, 21). ToxT is then produced and is able to activate transcription of the ctxAB genes responsible for CT production and the tcp genes responsible for TCP biogenesis, as well as the genes responsible for accessory colonization factors, while at the same time repressing the transcription of the anticolonization MSHA pilus (5, 12, 30). ToxT comprises two domains: a C-terminal DNA-binding domain and an N-terminal domain that has been implicated in dimerization and sensing on environmental stimuli (11, 23). Several studies have examined the mechanism of transcriptional activation by ToxT. ToxT is able to mediate transcriptional activation at some target promoters, such as that for the tcp operon, by binding individually at AT-rich “toxbox” motifs (29). Interactions between the N-terminal domains of ToxT are required for efficient activation of many virulence genes in the ToxT regulon; chimeric proteins consisting of ToxT DNA-binding domains fused to heterologous dimerization domains, as well as those with the ToxT dimerization domain fused to different DNA-binding domains, are able to bind the DNA corresponding to the appropriate DNA-binding domain but not the individual truncated forms (23). However, little is known about the requirements for ToxT-mediated repression of msh genes, the only set of genes ToxT is currently known to transcriptionally repress.
Previously, we have shown that ToxT represses a transcriptional msh-lacZ fusion in E. coli, suggesting that ToxT may directly regulate the expression of msh operon. To further study the relationship between ToxT and msh gene expression, we searched for sequences homologous to previously identified ToxT binding sites in the msh locus (Fig. (Fig.1A)1A) by using the PATSER program (27), taking into account known toxbox consensus sequences, as well as published results of mutagenesis of nucleotides within them (29). We identified potential ToxT binding sites in the mshI promoter and mshB promoters (Fig. (Fig.1A),1A), which were reported previously as regulating msh gene expression (18). The putative ToxT binding sites of mshI and mshB promoters are similar to the toxbox sequences in the aldA and acfD promoters, respectively (29). In addition, we also identified tandem ToxT binding sites located upstream of mshH similar to the binding sites upstream of tcpA (Fig. (Fig.1A).1A). Although mshH gene product may not be required for MSHA synthesis (18), it is predicted to form an operon with the downstream msh genes (22).
To test whether ToxT regulates these promoters containing putative ToxT binding sites, we constructed transcriptional Lux fusions to mshH, mshI, and mshB promoters and tested the effect of ToxT on Lux expression in a heterologous E. coli system. Figure Figure1A1A shows that ToxT was able to repress all of these promoters, suggesting that ToxT is capable of direct regulation at these sites. We then examined whether purified ToxT could bind to these putative regulatory regions in a gel retardation assay. Purified MBP-ToxT was able to shift DNA fragments containing each of the putative toxbox sequences identified by using PATSER (Fig. (Fig.1B),1B), as well as control DNA containing the toxbox found within the tcpA promoter. All of these mobility shifts could be abolished by the addition of specific unlabeled DNA, indicating that the binding of ToxT to these DNA sequences is specific. These data demonstrate that ToxT is able to bind several regulatory elements in order to negatively regulate msh gene transcription. We currently do not know the exact physiological significance of ToxT binding of three separate promoters within msh locus, but it is possible that repression at multiple regulatory regions may be exploited by V. cholerae to ensure that no or very little MSHA is produced in the early stage of infection.
To examine whether dimerization played a role in the transcriptional regulation of msh genes by ToxT, we examined the effects of a known inhibitor of ToxT dimerization, virstatin (14, 24), on the transcription of msh genes in E. coli (Fig. (Fig.1A).1A). Using Lux transcriptional fusion reporters (see Fig. Fig.1A),1A), we observed that PmshH-lux was repressed and that PtcpA-lux was activated when toxT was expressed via an arabinose-inducible PBAD promoter on a plasmid (pBAD-toxT) (9) (Fig. (Fig.2A).2A). Repression of PmshH-lux was maintained upon the addition of the ToxT dimerization inhibitor virstatin, but PtcpA-lux activity was completely abrogated in the presence of virstatin. The virstatin effects on expression of mshI and mshB were similar to that of mshH (data not shown).
ToxT represses the transcription of msh genes in V. cholerae after growth under toxT-inducing AKI conditions (15). To examine the effects of virstatin on ToxT-mediated regulation in V. cholerae, we immunoblotted for major pilin subunits TcpA and MshA using wild-type V. cholerae and a toxT mutant, both grown in AKI (Fig. (Fig.2B).2B). After virulence-inducing growth in AKI, wild-type V. cholerae activated production of TcpA and repressed production of MshA, while toxT mutants continued to express MshA even under toxT-inducing conditions. The addition of virstatin led to a loss of detectable TcpA production as reported previously (14, 24) but had no effect on ToxT repression of MshA production, suggesting that ToxT does not require dimerization in order to mediate transcriptional repression. Moreover, MSHA-dependent V. cholerae hemagglutination was not affected by growth with virstatin (Table (Table11).
To examine which regions of the ToxT protein were important for mediating transcriptional activation and repression, we generated a library of mutagenized toxT by passage of pBAD-toxT in E. coli XL1-Red mutator cells. We then screened for ToxT mutants that could no longer activate transcription but could still act as a repressor. Briefly, mutated pBAD-toxT plasmids were isolated from XL1-Red and used to transform an E. coli MC4100 containing a PtcpA-lacZ reporter plasmid. Transformants that caused a Lac− phenotype after growth on TTC-lactose indicator agar (25) (indicating that they contained mutated ToxT that lost the ability to activate the expression of tcpA) were then selected, and plasmids were isolated for transformation into PmshH-lacZ plasmid reporter-containing E. coli (12). Most of colonies were blue on the X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) plates, suggesting that mutations in toxT abolished ToxT activity completely. However, we were able to obtain five LacZ− mutants in which mutated ToxT could still function as an msh repressor. The effects of these mutations in toxT on transcriptional regulation were then confirmed by β-galactosidase reporter assays (20; data not shown). Sequence analysis revealed that all five ToxT mutants had an aspartate to glycine mutation at residue 141. This amino acid is within the first N-terminal amino acid region of ToxT that is proposed to be a dimerization domain.
When expressed from a PBAD promoter on a plasmid, ToxTD141G was able to still repress mshH, mshI, and mshB promoter activities in E. coli (Fig. (Fig.3B3B and data not shown) and to repress MshA protein synthesis (Fig. (Fig.3C)3C) and MSHA production (Table (Table1)1) in V. cholerae, but it was unable to activate tcpA expression in E. coli (Fig. (Fig.3B)3B) and complement production of TcpA in V. cholerae (Fig. (Fig.3C).3C). Further studies are required to test whether this D141G mutation affects the dimerization of ToxT.
ToxT seems to be able to bind independently to toxbox motifs upstream of most known ToxT-regulated promoters (28), either dimerizing after binding or interacting with other bound monomers to interact with RNA polymerase and activate transcription. The N-terminal region of ToxT seems to be critical for both dimerization and modulation of ToxT activity by environmental signals (2, 23, 24). D141 lies within this putative N-terminal dimerization domain.
To examine the role of the putative dimerization domain of ToxT in transcriptional regulation of msh genes, we constructed a variant of ToxT, ToxTΔN, that lacked the 2 to 144 N-terminal amino acid residues (Fig. (Fig.3A).3A). When expressed from a plasmid in E. coli, this variant of ToxT could not activate transcription of PtcpA-lux reporters but was still able to repress transcription of Pmsh-lux (including mshH, mshI, and mshB), further suggesting that negative regulation of the msh locus by ToxT is independent of dimerization (Fig. (Fig.3B3B and data not shown). In V. cholerae, ToxTΔN was unable to complement an in-frame toxT mutation for production of TCP (Fig. (Fig.3C,3C, top panel) but remained able to repress MshA expression (Fig. (Fig.3C,3C, bottom panel) and attenuate hemagglutination (Table (Table11).
These data suggest that ToxT is capable of direct negative regulation of the msh genes involved in the biogenesis of the anticolonization factor MSHA in both V. cholerae and a heterologous E. coli reporter system. Furthermore, ToxT is able to bind to the msh operon in up to three toxbox-like regulatory regions (Fig. (Fig.1).1). Transcriptional repression by ToxT is not affected by a known inhibitor of dimerization and transcriptional activation, virstatin. Mutations in the putative N-terminal dimerization domain are can abrogate activation at the ToxT-regulated tcpA promoter, but these ToxT alleles remain capable of repressing msh gene transcription. A complete N-terminal deletion mutant lacking the first 2 to 144 amino acids shows the same phenotype, further suggesting that ToxT-mediated transcriptional repression is independent of dimerization. Although monomeric ToxT-binding to cognate promoters has been reported previously in the case of activated promoters (28, 29), the data presented here offer the first indications of a mechanism for ToxT-mediated transcriptional repression. It is possible that ToxT may bind as independent monomers to a number of toxbox or toxbox-like motifs within the regulatory region of the msh operon, which may then occlude RNA polymerase and repress transcription. Additional in vitro and in vivo biochemical studies may yet yield insights into how and where ToxT can bind to the msh operon regulatory region. Studies of ToxT regulation at this locus thus offer fertile ground for future investigations into how this versatile regulator resolves the contradiction between the effects of colonization and anticolonization factor production during V. cholerae infection.
We thank Deborah Hung for providing virstatin and Bonnie Bassler for the pBBRlux plasmid.
This study was supported by NIH/NIAID R01 (AI072479) (to J.Z.) and an NSFC key project (no. 30830008) (to B.K.). R.V.K. acknowledges support from the Institute of Critical Technology and Applied Science at Virginia Tech.
Editor: A. Camilli
Published ahead of print on 21 January 2009.