ABSTRACT
The gene vvpE, encoding the virulence factor elastase, is a member of the quorum-sensing regulon in Vibrio vulnificus and displays enhanced expression at high cell density. We observed that this gene was repressed under iron-rich conditions and that the repression was due to a Fur (ferric uptake regulator)-dependent repression of smcR, a gene encoding a quorum-sensing master regulator with similarity to luxR in Vibrio harveyi. A gel mobility shift assay and a footprinting experiment demonstrated that the Fur-iron complex binds directly to two regions upstream of smcR (−82 to −36 and −2 to +27, with respect to the transcription start site) with differing affinities. However, binding of the Fur-iron complex is reversible enough to allow expression of smcR to be induced by quorum sensing at high cell density under iron-rich conditions. Under iron-limiting conditions, Fur fails to bind either region and the expression of smcR is regulated solely by quorum sensing. These results suggest that two biologically important environmental signals, iron and quorum sensing, converge to direct the expression of smcR, which then coordinates the expression of virulence factors.
INTRODUCTION
Pathogenic bacteria are equipped with complicated signal transduction systems to sense various environmental factors and swiftly adapt themselves to survive and propagate while in competition with host cells. Each signal transduction pathway for a given environmental factor has been studied intensively, but relationships among these pathways still remain to be defined.
Quorum sensing is a signal transduction process by which microorganisms sense the population of the same or related species and communicate with each other via diffusible signal molecules, generally called autoinducers (1). As with numerous other pathogenic bacteria, quorum sensing plays pivotal roles in regulating virulence factors in Vibrio vulnificus. V. vulnificus is a Gram-negative bacterium that causes fatal primary septicemia after the ingestion of contaminated seafood or during wound infection (2–5). Quorum sensing in V. vulnificus is not well understood but appears to be closely related to that of Vibrio harveyi and Vibrio cholerae (6). V. vulnificus harbors homologs of LuxPQ, a sensor for a borate diester autoinducer (AI-2), and carries a luxS gene, encoding the AI-2 synthase (7). A homoserine lactone autoinducer (AI-1) has been detected in one strain of V. vulnificus (8); however, in many other well-studied strains, such as YJ016, CMCP6, and MO6-24/O, whose genome sequences have been completely determined (9–11), the effort to identify an AI-1 compound or a gene responsible for AI-1 biosynthesis has been unsuccessful. An analysis of the genome sequences of these three strains has uncovered homologs of luxU and luxO, which encode proteins responsible for the transduction of signals via a phosphorelay from a sensor protein. These signals are funneled to the master regulator, SmcR, a homolog of LuxR in V. harveyi (7, 12), which subsequently induces the expression of vvpE, a gene encoding an elastase (13), and represses yegD, a gene encoding a chaperone (14). SmcR also represses the expression of hlyU (15), a gene encoding an activator that induces the expression of the virulence factors vvhAB and rtxA, encoding a hemolysin and an Rtx protein (a multifunctional autoprocessing toxin) (16), respectively.
Iron is an essential element for living organisms and is required for many biological metabolic pathways, including oxygen transport, photosynthesis, the trichloroacetic acid cycle, and respiration (17). However, the solubility of iron is extremely low at neutral pH; hence, biologically available iron is scarce and most organisms struggle to obtain iron. On the other hand, an excess of intracellular iron is deleterious to cells because it leads to the production of toxic free radicals. For these reasons, cells must meticulously control intracellular iron levels (18). Iron also plays an important role in the pathogenicity of bacteria. In pathogens such as Escherichia coli, V. cholerae, and Corynebacterium diphtheriae, iron levels dictate the expression of virulence-associated genes (19–21), and the production of those virulence factors reaches a maximal level when the concentration of iron is lower than that required for optimal growth (22). For the iron-associated regulation of genes, many bacteria employ ferric uptake regulator (Fur), a small protein that, in complex with iron, regulates multiple genes by binding to upstream sequences called Fur boxes (23–25). Genes in the Fur regulon have been identified as relevant not only for iron uptake/utilization, including siderophores and the ton system (17, 23), but also for pathogenicity, including a Shiga-like toxin and the Pseudomonas exotoxin A (19, 26). The Fur protein of V. vulnificus functions as a homodimer of approximately 16-kDa monomers and affects the expression of diverse genes, including those for iron utilization and superoxide dismutase (27).
Even though both quorum sensing and iron-dependent regulation have been studied intensively for Vibrio species, no relationships between these two global regulatory pathways have been explored. Recently, we showed that the Fur-iron complex and quorum sensing in V. vulnificus coordinately regulate siderophore production to achieve appropriate intracellular levels of iron (28). Under iron-limiting conditions, vvsAB, encoding vulnibactin, a siderophore of V. vulnificus, showed no significant expression at low cell density, but it was expressed fully at high cell density through SmcR-mediated induction. Under iron-rich conditions, vvsAB expression was repressed regardless of cell density. These results suggested that the two environmental signals, iron and cell density, are important for modulation of this virulence factor and that these two regulatory circuits are linked. In this study, we further investigated the effect of iron on the expression of virulence factors known to respond to quorum sensing in V. vulnificus. We found that signals from both iron levels and population density are funneled to the master regulator, SmcR, a homolog of V. harveyi-type LuxR, which then coordinately regulates the expression of virulence factors.
MATERIALS AND METHODS
Strains, plasmids, and culture conditions.The bacterial strains and plasmids used in this study are listed in Table 1. E. coli strains were cultured in Luria-Bertani (LB) broth supplemented with appropriate antibiotics at 37°C. V. vulnificus strains were cultured in LB broth or thiosulfate citrate bile salt sucrose (TCBS) agar at 30°C. When necessary, either ferrous sulfate (25 μM) as an iron source or 2,2′-dipyridyl (100 μM) as an iron chelator was added exogenously to the LB broth when the A600 value of the culture reached approximately 0.1.
Bacterial strains and plasmids used in this study
Assay of proteolytic activity.Strains of V. vulnificus were cultured overnight in LB medium and subcultured in fresh LB broth containing 100 μM 2,2′-dipyridyl or 25 μM FeSO4. Protease activity was measured quantitatively using the assay described previously (29). Specific activities were normalized to cell density.
Bioluminescence assays.Overnight cultures of V. vulnificus strains grown in LB were inoculated into fresh LB medium. To assess the effect of iron, either 25 μM FeSO4 or 100 μM 2,2′-dipyridyl was added, and samples were diluted 125-fold with LB broth. At various growth stages, 0.006% (vol/vol) n-decylaldehyde (in 50% ethanol) was added and luminescence was measured using a luminometer (Mithras LB 940; Berthold, Bad Wildbach, Germany) as previously described (28). Specific transcriptional level was expressed as the luminescence units normalized to cell density (relative luminescence units [RLU]).
Western blot hybridization of SmcR.For the SmcR expression analysis, overnight cultures of V. vulnificus MO6-24/O(pRK415), HLM101(pRK415), or HLM101(pRK-fur) grown in LB were subcultured into fresh LB medium and treated with either 100 μM 2,2′-dipyridyl or 25 μM FeSO4 when the A600 value of the culture reached approximately 0.1. Cells at stationary phase (A600 of ≈1.5) were washed and resuspended in phosphate-buffered saline (PBS) containing either 100 μM chelator or 25 μM FeSO4. Then, 60 μg of each lysate was subjected to SDS-PAGE and transferred to a Hybond P membrane (GE Healthcare Life Sciences, Piscataway, NJ). The membrane was incubated with polyclonal rat antisera against purified SmcR (28) (1:1,000 dilution in blocking solution) and subsequently with goat anti-rat immunoglobulin G-horseradish peroxidase (HRP) (1:2,000) (Santa Cruz Biotechnology, Santa Cruz, CA). SmcR expression was visualized using the Western blotting Luminol reagent (Santa Cruz Biotechnology, Santa Cruz, CA).
Cloning of fur and construction of vvpE-luxAB and smcR-luxAB transcriptional fusions.The 673-bp DNA fragment comprising the promoter region and the coding region of fur was amplified by PCR using the primers fur_comF and fur_comB (Table 2). The resulting product was cloned into pRK415 to construct pRK-fur. The 416-bp upstream region (−358 to +58 with respect to the translation start site) of vvpE and the 461-bp upstream region (−347 to +114 relative to the translation start site) of smcR were amplified by PCR using the primers vvpE_tcF and vvpE_tcB and primers smcR_tcF and smcR_tcB (Table 2), respectively, and cloned into the pGEM-T Easy vector (Promega, Madison, WI). The resulting plasmids were digested with KpnI and XbaI and cloned into pHK0011 (30) to generate pHvvpE and pHsmcR. To construct single crossovers of smcR-luxAB fusions into V. vulnificus, each lux fusion was amplified by PCR using the primers smcRscF and luxscB (Table 2) and cloned into the SphI- and SalI-digested pDM4 plasmid (31) using the In-fusion HD cloning kit (Clontech Laboratories, TaKaRa Bio Inc., Shiga, Japan) to generate pDM4-smcRlux. Each plasmid was conjugated into V. vulnificus MO6-24/O wild type, HLM101, and HS031.
Primers used in this study
Expression and purification of Fur.A DNA fragment encoding 149 amino acids of Fur was amplified by PCR using the primers furOEF and furOEB (Table 2) and subcloned into the pASK-IBA7 vector, resulting in expression of Fur fused to a Strep-tag at the N terminus. The resulting vector, named pASK-IBA7-Fur, was transformed into E. coli BL21(DE3), and expression of the Strep-tagged Fur was induced with 0.2 μg/ml anhydrotetracycline. After centrifugation, bacterial pellets were suspended in buffer W (100 mM Tris-Cl, 150 mM NaCl, and 1 mM EDTA), sonicated, and centrifuged at 7,000 rpm for 10 min. The resulting supernatant was purified using Strep-Tactin affinity resin (IBA BioTAGnology, Göttingen, Germany), and specifically bound protein was eluted with buffer E (100 mM Tris-Cl, 150 mM NaCl, 1 mM EDTA, and 2.5 mM desthiobiotin) according to the manufacturer's protocol. The eluted protein was separated on a 12% sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE) to assess purity. The purified Fur protein was dialyzed using Spectra/Por molecular porous membrane tubing (molecular weight cutoff [MWCO] of 10,000; Spectrum Laboratories Inc., Rancho Dominguez, CA) with buffer A (50 mM Tris-Cl [pH 8.0], 100 mM NaCl, 1 mM MgCl2, and 2 mM dithiothreitol) and concentrated using the Vivaspin 6 instrument (Vivagen, Seoul, South Korea). The protein concentration was determined by the Lowry method (32).
Site-directed mutagenesis of Fur boxes in the upstream region of smcR.The 398-bp DNA fragment of the smcR upstream region (−284 to +114 with respect to the translation start site) was amplified by PCR using the primers smcR_F and smcR_tcB (Table 2). The resulting product was ligated to the pGEM-T easy vector, resulting in pGEM-SmcR. Each of the regions (I and II) containing Fur boxes was mutagenized using the primer set smcR_smIF and smcR_smIB and the primer set smcR_smIIF and smcR_smIIB (Table 2), respectively, and both of the regions were mutagenized using both sets of primers and the QuickChangeII site-directed mutagenesis kit (Agilent Technologies, Santa Clara, CA). The resulting plasmids were named pGEM-Sm1, pGEM-Sm2, and pGEM-Sm1/2, respectively. Using each of these constructs as templates, the 461-bp upstream region (−347 to +114 with respect to the translation start site) was amplified by PCR using the primers smcR_tcF and smcR_tcB (Table 2). The resulting products were digested with KpnI and XbaI and cloned into pHK0011 to generate pHSm1, pHSm2, and pHSm1/2.
Gel shift assay.A 319-bp DNA fragment of the region upstream of smcR (−284 to +35 with respect to the translation start site) was amplified by PCR using the primers smcR_F and 32P-labeled smcR_B (Table 2). For gel shift assays, 10 ng of the labeled probe was incubated with increasing amounts of purified Fur protein (0 to 30 nM) in a 20-μl reaction mixture in binding buffer (33) containing 10 mM Tris-borate (pH 7.5), 100 μg/ml bovine serum albumin, 5% (vol/vol) glycerol, 40 mM KCl, 1 mM MgCl2, and 1 μg poly(dI-dC) and supplemented with either 100 μM MnCl2 or 100 μM EDTA for 30 min at 30°C. The binding reaction was terminated by the addition of 3 μl loading buffer, and samples were resolved on a 6% neutral polyacrylamide gel. To measure the affinity of Fur for smcR, sm1, sm2, and sm1/2, the 319-bp regions (−284 to +35 with respect to the translation start site) were amplified by PCR using the primers smcR_F and 32P-labeled smcR_B (Table 2) and pGEM-SmcR, pGEM-Sm1, pGEM-Sm2, and pGEM-Sm1/2 as templates. Each of the labeled probes (10 ng) was incubated with increasing amounts of purified Fur protein (0, 3, 4.5, 6, 12, 24, and 30 nM), and gel shift assays were performed as described above. Gels were exposed to a BAS-MP 2040s imaging plate (Fujifilm, Tokyo, Japan), and scanned by using a BAS-1500 instrument (Fujifilm, Tokyo, Japan). The intensities of the bands were measured using the Multi Gauge software program, version 3.0 (Fujifilm, Tokyo, Japan). The percentage of probe bound to Fur relative to unbound probe was calculated at each Fur concentration. The Kd value was obtained using the Prism program (Graphpad Software Inc., San Diego, CA).
DNase I footprinting analysis.An end-labeled, 319-bp DNA fragment of the smcR upstream region (−284 to +35 with respect to the translation start site) was amplified using the primers smcR_F and 32P-labeled smcR_B (Table 2). To determine the Fur binding site, 200 ng of the amplified smcR upstream region was incubated with increasing amounts of purified Fur (0 to 1 μM) in 20 μl of binding buffer (10 mM Tris-borate [pH 7.5], 100 μg/ml bovine serum albumin, 5% [vol/vol] glycerol, 40 mM KCl, 1 mM MgCl2, 1 μg poly[dI-dC]) with either 100 μM MnCl2 or 100 μM EDTA as a chelator for 1 h at 30°C. After 1 h, 0.4 unit of DNase I (Promega, Madison, WI) was added, and the reaction mixture was incubated at 30°C for 2 min. The reaction was terminated by the addition of 25 μl of stop buffer (20 mM EDTA), and 150 μl of deionized water was added. After the addition of 500 μl ethanol, samples were precipitated at −80°C for more than 2 h and then centrifuged. DNA pellets were washed with 70% ethanol and resuspended in 10 μl loading buffer (0.1 M NaOH-formamide [1:2], 0.1% xylene cyanol, 0.1% bromophenol blue). The samples and the sequencing ladder generated with 32P-labeled SmcR_B were denatured for 5 min at 95°C, chilled on ice, and separated on a 6% polyacrylamide sequencing gel. The sequencing ladders were prepared using an AccuPower DNA sequencing kit (Bioneer, Daejeon, South Korea).
Quantitative RT-PCR (qRT-PCR) analysis.RNA was isolated from V. vulnificus or V. cholerae using the RNeasy minikit (Qiagen, California, USA) and the RNase-free DNase set (Qiagen, California, USA). After quantification of purified RNA using a biophotometer (Eppendorf, Hamburg, Germany), cDNA was synthesized from 1 μg of RNA using the PrimeScript RT reagent kit (TaKaRa Bio Inc., Shiga, Japan), following the manufacturer's directions. cDNA (2 μl) was analyzed by reverse transcriptase PCR (RT-PCR) on a Light Cycler 480 II real-time PCR system (Roche Applied Science, Penzberg, Upper Bavaria, Germany) using LightCycler 480 DNA SYBR green I master (Roche Applied Science). RT-PCR was carried out in triplicate in a 96-well plate (Roche Applied Science) using the primers shown in Table 2. The gene encoding NAD-dependent glyceraldehyde-3-phosphatase of V. vulnificus was used as an endogenous loading control for the reactions. Quantification was carried out using the Light Cycler 480 II real-time PCR system software program.
RESULTS
Under iron-rich conditions, transcription of vvpE is repressed.Our previous study concluded that the Fur-iron complex regulates the expression of vvsAB, an operon encoding a siderophore important for virulence in V. vulnificus (28). From this observation, we predicted that iron may also affect the expression of other virulence factors in this pathogen; therefore, we examined the effect of iron on the expression of another known virulence factor, elastase, encoded by vvpE (30). In the absence of iron, elastase activity in a wild-type strain (MO6-24/O) increased when cells entered stationary phase (at A600 ∼2.2) (Fig. 1A), demonstrating that expression of this protein is controlled by quorum sensing, as previously reported (13). When we measured elastase activity under iron-limiting conditions in a fur deletion strain with or without complementation by wild-type fur on a plasmid, there was no significant difference from that of a wild-type strain. However, when cells were grown under iron-rich conditions, the timing of the increase in elastase activity differed between the wild type and a fur mutant. In the wild-type strain, activity increased later in stationary phase (A600 ∼2.7) than in the fur deletion mutant (A600 of ∼2.2). In the fur mutant strain complemented with wild-type fur on a multicopy plasmid, induction of activity was delayed even further (A600 of ∼3.0). To follow up on these results, we assessed the transcription of vvpE using a lux fusion (Fig. 1B). Consistent with what we observed for enzyme activity, the enhancement of transcription of vvpE in wild-type V. vulnificus in stationary phase (A600 of 2 to ∼2.7) was significantly delayed under iron-rich conditions compared to that under iron-limiting conditions. In the fur deletion mutant, transcription was induced much earlier regardless of iron availability. Together, these results suggest that the expression of vvpE, which is regulated by quorum sensing, is also affected by the presence of iron through repression by Fur.
Expression of vvpE is repressed by iron. (A) VvpE activity in wild-type MO6-24/O(pRK415) (circles), HLM101 (a fur deletion isotype) (triangles), or HLM101(pRK-fur) (squares) under iron-limiting or iron-rich conditions. Relative protease activity is expressed as protease activity normalized to cell density. Open symbols represent protease activities under iron-limiting conditions. Solid symbols represent activities under iron-rich conditions. (B) Luciferase activities of V. vulnificus wild type MO6-24/O (circles) and HLM101 (triangles) harboring vvpE-luxAB fusions under iron-limiting or iron-rich conditions. Relative luminescence units (RLU) represent the luminescence values normalized to cell density (optical density at 600 nm [OD600]). Open symbols represent luciferase activities under iron-limiting conditions. Solid symbols represent the activities under iron-rich conditions. Data are the average values from three independent experiments, and error bars denote standard deviations.
The Fur-iron complex represses the expression of SmcR.It is well known that the transcription of vvpE is positively regulated by the quorum-sensing master regulator, SmcR (13). We hypothesized that iron, together with Fur, affects the expression of vvpE indirectly by regulating the expression of smcR. To verify this hypothesis, we assessed the effect of iron on the expression of smcR using an smcR-luxAB transcriptional reporter fusion in cells grown for varied cell densities (Fig. 2). At early exponential phase, the transcription of smcR in a wild-type strain was barely detectable, irrespective of the iron concentration. At stationary phase (A600 of ∼2.0), smcR was fully expressed in the absence of iron but was expressed at only about half that level under iron-rich conditions. When cells reached late stationary phase, the repression was almost completely relieved (Fig. 2A). In contrast, in a fur deletion mutant, the transcription of smcR under iron-rich conditions is derepressed regardless of growth phase. Complementation of the fur deletion mutant with a fur gene on a plasmid restored the iron-dependent repression of smcR.
The Fur-iron complex represses transcription and translation of smcR. (A) Luciferase activities of V. vulnificus MO6-24/O(pRK415) (circles), HLM101(pRK415) (triangles), and HLM101(pRK-fur) (squares) harboring smcR-luxAB fusions under iron-limiting (open symbols) or iron-rich (closed symbols) conditions. Relative luminescence units (RLU) represent the luminescence values normalized to cell density (OD600). Data are the average values from three independent experiments, and error bars denote standard deviations. (B) Western hybridization of total protein extracts from V. vulnificus MO6-24/O(pRK415), HLM101(pRK415), and HLM101(pRK-fur) using polyclonal antisera against SmcR. Lanes 1 and 2, V. vulnificus MO6-24/O(pRK415); lanes 3 and 4, HLM101(pRK415); lanes 5 and 6, HLM101(pRK-fur). Lanes 1, 3, and 5 represent cells grown under iron-limiting conditions (100 μM 2,2′-dipyridyl), and lanes 2, 4, and 6 represent cells grown under iron-rich conditions (25 μM FeSO4). Total protein was extracted from cells at an OD600 of 2.0, and 60 μg of this protein was loaded into each lane. Relative intensities of bands measured by a densitometer are indicated.
Repression of smcR expression at low cell density is achieved by degradation of smcR mRNA through the action of small RNA molecules called Qrrs, which are activated by phosphorylated LuxO (6, 34). For this reason, measurements of smcR transcription using a reporter fusion may not accurately represent the levels of the active SmcR protein; therefore, we measured the levels of the SmcR protein directly to assess the effect of iron (Fig. 2B). Western hybridization using polyclonal rat antisera against SmcR showed that the levels of SmcR are lower in the presence of iron and that a mutation in fur reversed this effect. Complementation of the mutant with fur on a plasmid again led to decreased levels of SmcR in the presence of iron (Fig. 2B). These results suggest that Fur, in the presence of iron, represses the expression of the quorum-sensing master regulator SmcR, which consequently represses the expression of vvpE, and the effect was relieved at late stationary phase.
Fur binds directly to the promoter region of smcR under iron-rich conditions.It is well known that the Fur-iron complex acts as a negative regulator (25), and we assumed that this complex would repress transcription of smcR by binding to a cis-acting element of smcR. To test this possibility, gel shift assays and DNase I footprinting analyses were carried out. A gel shift assay was performed using a 32P-labeled 319-bp fragment of DNA from the region upstream of smcR to determine whether the Fur-iron complex binds directly to the promoter region (Fig. 3A). In the presence of 100 μM manganese (Mn), which is a typical substitute for ferrous iron in in vitro binding reactions (35–37), Fur binds to this sequence (Fig. 3A); however, in the presence of the chelator EDTA, Fur was unable to bind (Fig. 3B). To identify the exact location within this region where Fur binds, a DNase I footprinting analysis was carried out using purified Fur. In the presence of Mn, two distinct regions within this sequence were protected by Fur, while no protection was observed when EDTA was added. One region (region I) includes nucleotides −82 and −36 with respect to the transcriptional start site and therefore is located just upstream of the −35 promoter region (Fig. 3C). The second region (region II) includes nucleotides −2 and +27 and therefore overlaps with the transcriptional start site and spans the start of the transcript (Fig. 3C). The nucleotide sequences of these two protected regions are similar to the typical consensus sequence recognized by Fur and possibly form hairpin structures to accommodate homodimers of Fur (25) (Fig. 4).
Binding of Fur to the region upstream of smcR. Shown here are the results of gel shift assays using a radiolabeled 319-bp fragment from upstream of smcR and purified Fur with either 100 μM MnCl2 (A) or EDTA as a chelator (B) added to the medium. Lanes 1 to 5, Fur concentrations of 0 nM, 3 nM, 6 nM, 12 nM, and 30 nM, respectively; lanes 6 to 8, 30 nM Fur with unlabeled probe as a competitor at 1 ng, 10 ng, and 100 ng, respectively. (C) DNase I protection of the upstream region of smcR by Fur. Lanes 1 to 5, 0 nM, 131 nM, 265 nM, 533 nM, and 1 μM Fur, respectively, incubated with 200 ng of 3′-labeled DNA with either 100 μM MnCl2 or 100 μM EDTA; lanes G, A, T, and C represent the corresponding sequencing ladder. Regions I and II denote regions protected from DNase I digestion (+1 represents the transcriptional start point).
Comparison of regions I and II within the sequence upstream of smcR containing the consensus Fur box and the mutagenized bases. Nucleotides in bold were protected by the Fur-iron complex in the footprinting analysis, and dots denote putative hairpin-forming nucleotides. Putative −35 and −10 regions and the translational start site are indicated. Mutagenized sequences, Sm1 and Sm2, are also indicated.
This footprinting analysis suggested that two regions upstream of smcR, regions I and II, are protected by Fur and that Fur binds region I with a higher affinity than region II. To confirm that Fur binds in this location, we mutagenized bases in one or both of the two regions such that the hairpin structures were disrupted (Fig. 4) and measured the binding of purified Fur to these mutated sequences. As shown in Fig. 5A, the Kd of Fur for the wild-type sequence was ∼8.2 nM in the presence of iron. Mutations in either region led to higher Kd values for Fur, with∼12.6 nM for region I (Sm2) and ∼38.0 nM for region II (Sm1). Mutations in both of the regions combined (Sm1/2) completely abolished Fur binding. In the absence of iron, binding of Fur to either region was not detected (data not shown). These results suggest that regions I and II are both bound by the Fur-iron complex but that binding occurs with a much higher affinity at region I than at region II.
Regions I and II are important for repression of smcR by the Fur-iron complex. (A) Affinity of the Fur-iron complex to either wild-type or mutagenized Fur-binding regions in the upstream region of smcR. Binding of the Fur-iron complex to the wild-type sequence, Sm1, Sm2, and Sm1/2 was quantitatively measured as described in Materials and Methods. The y axis represents the percentage of Fur-iron binding relative to binding of the wild-type sequence, and the x axis represents the concentration of Fur. (B) Expression of the luxAB fusion driven by the wild-type or mutagenized region upstream of smcR. Cells were grown in the presence (25 μM FeSO4) or absence (100 μM 2,2′-dipyridyl) of iron, and the resulting luciferase activity was measured. RLU (relative luminescence units) indicates luciferase activity normalized to cell density (OD600). “X” indicates a mutated Fur binding region. Data are average values from three independent experiments, and error bars denote the standard deviations (*, P < 0.005; **, P < 0.05).
Transcriptional fusions were constructed between the reporter genes luxAB and the wild type smcR upstream region (pHSmcR), each of the mutated regions (pHSm1 and pHSm2), or the double mutant (pHSm1/2). Expression of the reporter genes in each of these constructs was quantitatively measured in the presence or absence of iron. When regions I and II were both wild type, the presence of iron resulted in half as much expression of the reporter than was observed in the absence of iron (Fig. 5B). Mutations in either of the two Fur binding regions lessened the repressive effect, and mutations in both of the Fur binding regions resulted in expression equivalent to that of the wild-type sequence in the absence of iron, suggesting that Fur was unable to bind in this case. There was more expression of the reporter gene when region II was mutated than when region I was mutated, suggesting that region II is more important for Fur repression of smcR.
The Fur-iron complex affects the expression of yegD and hlyU by repressing smcR.In the study described above, we showed that the expression of vvpE is decreased in the presence of iron due to the regulation of smcR expression. Our finding that the Fur-iron complex repressed the expression of smcR, the primary quorum-sensing regulator gene, led us to examine the effect of iron on the expression of other genes previously reported to be under the regulation of SmcR. Two of these genes, yegD, encoding a molecular chaperone, and hlyU, encoding an activator, are negatively regulated at the transcriptional level by SmcR (14, 16). We measured the expression of these genes in cells at stationary phase for the wild type (MO6-24/O), a fur deletion mutant (HLM101), and an smcR deletion mutant (HS031) in the presence or absence of iron using quantitative RT-PCR. Under iron-rich conditions, the expression of each of these genes was enhanced in wild-type cells (Fig. 6A and B). However, in a fur deletion mutant, expression levels did not increase upon the addition of iron, but in trans complementation of fur on a plasmid restored the response of these genes to iron. In an smcR deletion strain, the expression of these two genes was higher. Again, in trans complementation of smcR into this mutant reduced expression levels to those of the wild type. In summary, these results demonstrate that iron, in cooperation with Fur, led to the increased expression of yegD and hlyU by repressing smcR. Together, our results indicate that members of the quorum-sensing regulon are regulated by the Fur-iron complex and that this regulation is accomplished by the modulation of SmcR levels.
Transcription of yegD and hlyU is affected by the Fur-iron complex through regulation of smcR. Comparisons of transcriptional levels of yegD (A) or hlyU (B) in wild-type V. vulnificus MO6-24/O(pRK415), HLM101(pRK415), HLM101(pRK-Fur), HS031(pRK415), and HS031(pRK-SmcR) at stationary phase under iron-limiting (empty bars) or iron-rich (solid bars) conditions by quantitative RT-PCR using the primers shown in Table 2. Overnight cultures were washed and subcultured in LB medium supplemented with either 100 μM 2,2′-dipyridyl or 25 M ferrous sulfate and grown to log phase (A600 of 1.0). RNA levels were quantified using the comparative threshold cycle (ΔΔCT) method, and RNA fold change was normalized to the value for MO6-24/O harboring pRK415 in the absence of iron. The data are average values from three independent samples, and error bars denote the standard deviations.
DISCUSSION
Cell density and iron availability are important environmental variables affecting the physiology of cells and are particularly important for the pathogenicity of virulent bacteria. Each of these two environmental factors elicits its own cognate signal transduction pathway and presumably affects the pathway of the other. Iron may influence the quorum-sensing pathway by affecting the growth rate of cells. Iron-poor conditions limit the growth of cells and result in a low cell density, whereas excessive amounts of iron may limit growth by generating toxic radicals (17, 24). When a pathogen enters a host where available iron is scarce, iron influences the growth of the pathogen and likely affects quorum sensing. On the other hand, cell density may directly affect the availability of iron. Cells at high density compete for available iron sources, leading to iron-limiting conditions and the subsequent constraint of growth. Therefore, pathogenic bacteria need to sense those two environmental conditions both temporally and spatially and must orchestrate the appropriate signal pathways to control gene expression accordingly in order to optimize physiological conditions and promote survival in the host.
Among Pseudomonas species, in which a quorum-sensing regulatory pathway similar to that from V. fischeri is employed, there are examples for which the Fur-iron complex affects the quorum-sensing regulatory pathway. In Pseudomonas syringe, a mutation in Fur reduces the production of the autoinducer N-acyl homoserine lactone (N-AHL) and reduces the expression of several quorum-sensing-associated genes, and conversely, N-AHL influences expression of fur (38). In Pseudomonas aeruginosa, Fur affects the production of Pseudomonas quinolone signal (PQS), which controls the synthesis of secondary metabolites, extracellular enzymes, and virulence factors. In this bacterium, iron also affects the expression of the quorum-sensing regulator LasR (39). However, the molecular mechanism for a relationship between the iron-associated signaling pathway and the quorum-sensing pathway in these Pseudomonas species has not yet been investigated. In fact, for most bacteria in which a quorum-sensing regulatory pathway similar to that of V. harveyi is employed, interactions between quorum sensing and Fur-mediated iron regulation have yet to be elucidated. Previously, we showed that the expression of vvsA, a gene encoding vulnibactin, is modulated cooperatively by SmcR and the Fur-iron complex. This result led us to extend our study to the effect of iron on components associated with quorum sensing. In this work, we showed that in V. vulnificus, cell density and iron availability are monitored, and those two signals converge on smcR, a master regulator for the quorum-sensing pathway with similarity to V. harveyi luxR, to orchestrate the expression of virulence genes. The iron-dependent regulation of smcR is achieved through binding of the Fur-iron complex to the upstream regions of the regulator gene, as demonstrated by both gel shift and footprinting experiments.
The Fur-iron complex binds to two distinct regions (regions I and II) upstream of smcR (Fig. 4), each of which has sequence similarity to the consensus Fur binding site and potentially forms the hairpin structure that is a common feature of the Fur box (23–25). Both of these two regions are required for the full repression of smcR by Fur, and the presence of two distinct binding regions appears to be necessary for fine-tuning of smcR repression in response to iron concentrations. Region I had a significantly higher affinity for Fur than region II, suggesting that the Fur-iron complex binds first to region I and then, as the intracellular concentration of iron increases, binds to region II. Regardless of binding affinity, region II appears to be more important for the regulatory activity of Fur (Fig. 5), perhaps because this region includes the transcriptional initiation site that would be blocked when the Fur-iron complex is bound. Even though there are two Fur binding sites, the expression of smcR was not completely repressed upon the addition of iron but rather was about a half the level observed in the absence of iron (Fig. 5). Complete repression of smcR expression is achieved only through the action of quorum-sensing regulation responding to low cell density regardless of iron, suggesting that the Fur-iron complex does not tightly repress smcR. The expression of both smcR and vvpE is low under iron-rich conditions through early stationary phase, but repression is relieved by late stationary phase (Fig. 1 and 2A and Fig. 2). To explain this observation, we considered the possibility that intracellular iron is depleted by late stationary phase and that the repression of smcR may be relieved as well. However, supplementation with excess iron to a culture in late stationary phase did not lead to complete repression of either smcR or vvpE (data not shown). Another possible hypothesis, in which SmcR represses the expression of fur in a culture at high cell density, was rejected based on findings of our previous study, which clearly showed that SmcR does not affect the expression of fur (28). Likewise, we ruled out a role for RpoS, which may have participated because of its role in stationary-phase-dependent gene regulation (37), because the effect of Fur-iron on expression of smcR was the same in wild-type and rpoS mutant strains (data not shown). Taken together, these data suggest that the derepression of smcR in late stationary phase is not due to any genetic or physiological effects. It is likely that the repression of smcR by the Fur-iron complex is intrinsically “leaky.” Such weak repression may be advantageous for cells because it allows the quorum-sensing regulation to be operative even under iron-rich conditions. If repression of smcR by the Fur-iron complex were strong, then quorum-sensing regulation would not be possible in the presence of iron. By employing a weak inhibitor of smcR, cells are able to sense both population density and iron concentrations simultaneously and coordinately control the expression of virulence factors.
Previous studies using DNA microarrays to screen for genes in either V. vulnificus or V. cholerae that are expressed differentially in various iron concentrations (33, 40) did not identify smcR, hapR of V. cholerae, which is homologous to V. harveyi luxR and is the equivalent of smcR, or genes in the quorum-sensing regulon. However, in these studies, total gene expression was measured in cells during the exponential phase of growth. During this phase, the mRNAs of smcR and hapR are degraded through the action of the small RNA Qrrs (6, 34), and therefore any effect of the Fur-iron complex would not be observed. In contrast, our studies were carried out in cells at various growth stages. At high cell density, the mRNAs of smcR and hapR are not degraded, and the effect of the Fur-iron complex on each of these master regulators and on genes influenced by these master regulators can be observed.
Orchestrated gene regulation via SmcR in response to both cell density and iron availability would make it possible to vary the expression of the virulence factors VvpE, RtxA, and VvhA spatially as well as temporally (Fig. 7). VvpE was expressed at the highest level under iron-limiting conditions at high cell densities, whereas RtxA and VvhA were expressed at the highest level at low cell densities regardless of iron concentrations. This result leads to speculation about the possible roles these factors play during the infection process. VvpE, an elastase, may be expressed within a biofilm, where cell density reaches the highest level and iron is depleted due to the confined space. This enzyme may be employed to destroy human cells in order to acquire iron at the initial stage of infection following dispersal of pathogens from the biofilm. After elastase mediates iron scavenging, RtxA1 and VvhA may be employed for subsequent attacks on host cells and further propagation. Monitoring the expression pattern of these genes in an animal model, especially within biofilms, may be an effective way to verify this hypothesis.
Model for the orchestrated regulation of virulence genes by SmcR in response to iron and population density in V. vulnificus. At low cell densities (LCD), the smcR transcript is degraded through the action of both Hfq and small RNA molecules (Qrrs), which themselves are induced by phosphorylated LuxO (34). Therefore, under these conditions, irrespective of iron, vvpE is not expressed but hlyU and yegD are expressed (13–15). Under iron-rich conditions, vvsAB is directly repressed by the Fur-iron complex (28). At high cell densities (HCD) under iron-limiting conditions, Qrrs are not expressed and apo-Fur does not repress smcR; therefore, SmcR is fully expressed. As a consequence, vvpE is fully induced and yegD, hlyU, and vvsAB are repressed. Under iron-rich conditions, the Fur-iron complex represses smcR but binds weakly, and thus quorum-sensing regulation is still functional. As a result, vvpE is partially repressed, and yegD and hlyU are both partially enhanced. The activator encoded by hlyU then increases the expression of the virulence factors RtxA and VvhA (15). Meanwhile, under these same conditions, the expression of vvsAB is strongly repressed directly by the Fur-iron complex (28).
In summary, we demonstrated at a molecular level that intracellular signaling in response to two important environmental factors, population density and iron, converge to fine-tune the levels of the V. harveyi LuxR-type master regulator SmcR in V. vulnificus. This pattern of regulation makes it possible to orchestrate the expression of virulence factors efficiently and promptly when changes in the environment are detected.
ACKNOWLEDGMENT
This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (grant number 2012-0005731), Ministry of Education, Science & Technology, Republic of Korea.
FOOTNOTES
- Received 24 March 2013.
- Returned for modification 7 May 2013.
- Accepted 22 May 2013.
- Accepted manuscript posted online 28 May 2013.
- Copyright © 2013, American Society for Microbiology. All Rights Reserved.
