INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Vibrio vulnificus is a halophilic estuarine bacterium that causes fatal septicemia and necrotizing wound infections. Primary septicemia occurs following ingestion of raw seafood contaminated with V. vulnificus. V. vulnificus preferentially affects persons with underlying hepatic diseases, a heavy alcohol drinking habit, and other immunocompromised conditions. Primary septicemia shows a rapidly progressing and fulminant course, which results in a high mortality rate of over 50% despite aggressive antimicrobial and supportive shock therapies (38, 39, 48).

The optimal natural habitat of V. vulnificus is an estuary. The bacterium normally flourishes in estuarine seawater, shellfish, and plankton during warm months (8, 35, 50, 54). V. vulnificus is concentrated in oysters and probably in other shellfish as well (21, 49). V. vulnificus opportunistically causes primary septicemia when contaminated shellfish is eaten raw by susceptible patients. This opportunist should experience a very dramatic change in environmental parameters during the infection process. Successful infection by pathogenic bacteria, in general, is established by coordinate expression of various virulence factors in vivo. Expression of virulence factors is controlled by environmental cues. Pathogenic bacteria possess elegant regulatory systems that sense and react to fluctuations in environmental parameters such as temperature, osmolarity, pH, iron concentration, CO₂ concentration, etc. (12, 27).

Many pathogens employ novel signal transduction systems in regulating the virulence gene expression (12). Toxigenic V. cholerae has the toxRS system for that purpose (30). The genes toxR and toxS are clustered in an operon and encode transmembrane proteins ToxR and ToxS, respectively (10, 29, 32). ToxR regulates expression of multiple V. cholerae virulence factors such as the cholera toxin (ctx), toxin-coregulated pilus (tcp), and accessory colonization factor (acf) genes (9, 47). The activity of ToxR is further enhanced by ToxS, which interacts with the former protein in the periplasmic space and is thought to stabilize it (10, 36). The toxRS system seems to play universally important roles in the survival and host-microorganism interaction of Vibrio species. V. parahaemolyticus and V. fischeri also have homologs of the V. cholerae toxRS (toxRS_Vc) system (25, 43). In the present study, we identified the toxRS homolog in V. vulnificus and showed the functional homology of ToxRS_Vv with ToxRS_Vc. We also observed that ToxR_Vv regulates the production of hemolysin, the most potent exotoxin produced by the organism (14).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, media, plasmids, and reagents. The Escherichia coli and V. vulnificus strains and plasmids used in this study are listed in Table 1. V. vulnificus type strain ATCC 29307 was used for the toxRS_Vv cloning experiment. For mutant construction and functional studies of ToxRS_Vv, we used the highly virulent clinical isolate V. vulnificus MO6-24/O. E. coli and V. vulnificus strains were grown in LB medium (28) and 2.5% NaCl brain heart infusion medium (6), respectively. Antibiotics were used at the following concentrations: ampicillin, 100 µg/ml; chloramphenicol, 30 µg/ml; tetracycline, 20 µg/ml; kanamycin, 30 µg/ml.

Amplification of a putative toxRS_Vv DNA fragment by degenerate PCR. A DNA fragment used as an authentic probe for the screening of plasmid libraries was amplified by degenerate PCR. By comparing the toxRS genes of V. cholerae and V. parahaemolyticus, we designed a set of degenerate primers (sense primer SCH961 [5'-CGATTAGGNAGCAACGAAAGCCG-3'] and antisense primer SCH962 [5'-GTCACTNCCCCANTANAACC-3']) encompassing parts of toxR and toxS. Chromosomal DNA of V. vulnificus ATCC 29307, prepared as described elsewhere (24), was used as the template for PCR amplification. PCR was performed with a water-cooled thermal cycler (Single Block Easycycler; Ericomp Inc., San Diego, Calif.).

Cloning of toxRS_Vv from plasmid libraries. The chromosomal DNA of V. vulnificus ATCC 29307 was fully digested with various restriction enzymes, electrophoresed, and transferred to nitrocellulose membranes (Bio-Rad, Hercules, Calif.). Southern blot analysis was performed as described by Sambrook et al. (44). The insert DNA of pCMM9 and the two fragments resulting from its HindIII digestion were used as probes. The 222-bp PCR product of the vvh hemolysin gene was used as a control probe (24). The probes were labeled by [³²P]dCTP (DuPont NEN, Boston, Mass.) using the nick translation system or the random priming system of Promega (Madison, Wis.), depending upon the probe length.

Expression of ToxR and ToxS as fusion proteins. Expression of the toxR_Vv and toxS_Vv genes as proteins was tested by using the Glutathione S-Transferase (GST) Gene Fusion System (Amersham Pharmacia Biotech, Piscataway, N.J.). For this study, fragments encoding toxR_Vv and toxS_Vv were amplified by PCR and cloned into the pCRII TA cloning vector. For directional cloning into GST fusion vectors at the BamHI-EcoRI site, the primers were designed to contain BamHI or EcoRI restriction sites at their 5' regions. toxR_Vv was amplified by primers GSTR1 (5'-CGGGATCCTCATGAAGTAATACCGGCA-3') and GSTR2 (5'-CGGAATTCTTATTTACAGATAGAACC-3'). toxS_Vv was amplified by primers GSTS1 (5'-CGGGATCCATGAAACTTCGAATTGCT-3') and GSTS2 (5'-CGGAATTCTCAGTTAGAAAACAGTAC-3'). Boldface and underlined letters denote restriction enzyme recognition sites and the putative start or stop sites of the genes, respectively. toxR_Vv was cloned into pGEX-3X, and toxS_Vv was cloned into pGEX-2T. The resulting plasmids were transformed into E. coli DH5, and the transformants were induced by isopropyl--D-thiogalactopyranoside (IPTG; Sigma, St. Louis, Mo.) in accordance with the manufacturer's protocol. The expressed GST-ToxR and GST-ToxS proteins were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) of the bacterial lysate and subsequent Coomassie staining.

Production of anti-ToxR serum. For every batch culture, expression of the fusion protein was confirmed by SDS-PAGE and Coomassie staining. After induction, the culture was harvested by centrifugation. To prepare a bacterial lysate for affinity column chromatography, the pellet was resuspended in lysis buffer (50 mM Tris-Cl [pH 7.5], 150 mM NaCl, 0.5 mM EDTA [pH 8.0], 1 mM benzamidine HCl, 0.2 mM phenylmethylsulfonyl fluoride, 1% Triton X-100, 0.1% -mercaptoethanol) and sonicated (Vibra Cell Sonicator Model CV26; Sonics & Materials Inc., Danbury, Conn.) on an ice bed. After sonication, the GST-ToxR_Vv fusion protein in the lysate was purified by the affinity chromatographic method recommended by the manufacturer (Amersham Pharmacia Biotech).

Western blot analysis of ToxR_Vv expression. ToxR expression in the V. vulnificus and E. coli backgrounds was analyzed by Western blot analysis using the rabbit antiserum as the primary polyclonal antibody. E. coli VM2 harboring pCMM610 was used as the positive control. Protein bands separated by SDS-PAGE were transferred to nitrocellulose paper (Bio-Rad). The paper was first treated with blocking buffer (5% skim milk, 0.1% Tween 20 in PBS) for 1 h and then washed three times with washing buffer (1% skim milk, 0.5% bovine serum albumin, 0.1% Tween 20 in PBS) for 10 min each time. The antiserum diluted 1:500 was incubated with the washed paper strip for 1 h with gentle agitation. After washing, alkaline phosphatase-conjugated anti-rabbit immunoglobulin G antibody (Sigma) diluted 1:15,000 was reacted for 1 h. After thorough washing, the strip was dipped in a 5-bromo-4-chloro-3-indolylphosphate-nitroblue tetrazolium (Sigma) substrate solution. All reaction and washing steps were performed at room temperature.

Stimulation of ctx promoter by ToxR_Vv. The functional similarity of ToxR_Vv to its V. cholerae counterpart was tested in an E. coli background. P_BAD expression plasmids pCMM601 and pCMM610 were transformed into E. coli VM2 harboring a ctx::lacZ fusion in the chromosome (30). E. coli VM2 transformed with either pCMM610 (toxR_Vv) or pCMM601 (toxRS_Vv) was induced by 0.2% arabinose as described by Guzman et al. (15). -Galactosidase assays (28) of the culture aliquots collected at appropriate time intervals were carried out in triplicate. The same experiments were done with pCMM641 (toxRS_Vc) and pCMM651 (toxR_Vc) as positive controls.

Interaction of ToxRS_Vv with vvhA in an E. coli background. Regulation of vvhA gene expression by ToxR_Vv or ToxRS_Vv was investigated in an E. coli background. E. coli DH5 harboring pJL961, a derivative of pACYC184 containing a 3.4-kb vvhA fragment from pCVD702 (56), was transformed with pCMM601, pCMM610, pCMM641, or pCMM651. The cotransformants were grown in NZY broth (15) containing both chloramphenicol and ampicillin. Induction of the P_BAD promoter by arabinose was done as described above. At appropriate time intervals after induction, culture supernatants were collected for hemolysin assay. Hemolysin activity in the supernatant was assayed in triplicate by the tube method as previously described (45). Western blot analysis of the ToxR_Vv levels in the E. coli background was performed to address whether coexpression of ToxS_Vv increased ToxR_Vv stability and contributed to significant hemolysin production while the expression of ToxR_Vv alone could not induce hemolysin production.

Construction of a toxR mutant. To investigate the physiological role of toxR in hemolysin production by V. vulnificus, a toxR mutant of strain MO6-24/O was constructed by inserting a suicide vector into the chromosome using the strategy originally described by Miller and Mekalanos (31). A KpnI-HindIII toxR_Vv fragment from pCMM9 was ligated to suicide vector pNQ705 (33). The toxR_Vv fragment lacks about 100 and 300 bp at the 5' and 3' ends, respectively, of the toxR_Vv open reading frame. The ligated DNA was used to transform competent cells of E. coli SY327  pir. Chloramphenicol-resistant (Cm^r) transformants were selected, and plasmid DNA was analyzed by restriction mapping to identify a plasmid, designated pCMM700, carrying the correct insert. Plasmid pCMM700 was transformed into E. coli SM10  pir and subsequently transferred to V. vulnificus MO6-24/O by conjugation. Cm^r transconjugants, which contain the mobilized plasmid integrated into the genome by homologous recombination, were selected on thiosulfate-citrate-bile-sucrose (TCBS) agar plates containing chloramphenicol at 2 µg/ml. Insertional mutation was confirmed by Southern hybridization and PCR analysis of the chromosomal DNAs of the mutant and the wild type (26, 31). The truncated toxR_Vv insert was labeled with the nonradioactive digoxigenin labeling kit (Boehringer Mannheim) and used as the probe for Southern blot hybridization. The difference in hemolysin production between the mutant and parent wild-type strains was determined by culturing both strains in 2.5% NaCl heart infusion broth at 37°C and 220 rpm. Hemolytic activity in the supernatant was measured as described above.

OMP profile. Changes in the outer membrane protein (OMP) profile of the toxR_Vv mutant were tested. OMPs were purified by a method reported elsewhere (7). Briefly, bacterial cells were harvested by centrifugation at 3,000 × g for 15 min. The pellets were washed three times with 0.8 M sucrose in 50 mM Tris-HCl buffer (pH 8.0). Spheroplasts were made by suspending the pellets in 50 mM Tris-HCl buffer (pH 8.0) containing lysozyme at 10 µg/ml and 2.5 mM EDTA and incubated at 37°C for 30 min. Spheroplasts were pelleted by centrifugation at 8,000 × g for 15 min. The supernatants were collected and pelleted at 30,000 × g for 30 min. The resulting pellets were treated with 1.5% Sarkosyl at room temperature for 20 min. After this treatment, OMPs were pelleted at 30,000 × g for 30 min. The OMP preparations were analyzed by SDS-12% PAGE.

Nucleotide sequence accession number. The complete toxRS_Vv and partial htpG_Vv DNA sequences have been deposited in the GenBank database under accession number AF166120.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Identification and analysis of toxRS_Vv. By using degenerate primers, ca. 900-bp DNA fragments were amplified from the chromosomal DNAs of V. vulnificus ATCC 29307, V. cholerae ATCC 14033, and V. parahaemolyticus ATCC 17802. All three of the Vibrio species gave the same size of amplification product. The ca. 900-bp DNA fragment amplified from V. vulnificus ATCC 29307 was cloned into the pCRII vector and sequenced. Restriction and DNA sequence analysis showed that the insert was 854 nucleotides in length and had a HindIII site yielding 566- and 288-bp fragments after digestion with the restriction enzyme (Fig. 1). The DNA sequence of the insert showed 61.2, 46.9, and 47.8% homologies to the corresponding regions in the toxRS genes of V. parahaemolyticus, V. cholerae, and V. fischeri, respectively. This insert was used, with or without HindIII digestion, as the authentic probe in further studies to identify the chromosomal copy of toxRS_Vv.

View larger version (12K): [in this window] [in a new window]   FIG. 1.   Restriction map of the 2.8-kb HindIII-HindIII-BglII DNA fragment cloned from V. vulnificus ATCC 29307. The fragment was cloned as separate 1.2-kb HindIII and 1.6-kb HindIII-BglII fragments. The arrows indicate the locations and directions of transcription of two complete open reading frames (toxRVv and toxSVv) and one partial open reading frame (htpGVv). Restriction sites are abbreviated as follows: AI, Alw44I; BII, BglII; CI, ClaI; HIII, HindIII; PI, PvuI; RV, EcoRV. The hatched boxes indicate fragments of the PCR-amplified DNA product used as the authentic probes. The whole product or the two fragments resulting from HindIII digestion were labeled by [32P]dCTP and used in the genomic Southern blot hybridization analysis.

View larger version (63K): [in this window] [in a new window]   FIG. 2.   Nucleotide sequence of the locus containing the toxRSVv genes. The arrows indicate the directions of transcription of open reading frames. The partial open reading frame of htpGVv runs opposite to the toxRSVv operon. Putative Shine-Dalgarno sequences are underlined. The underlined, italicized sequences were designated the putative initiation or stop codons of the open reading frames.

View larger version (81K): [in this window] [in a new window]   FIG. 3.   Comparison of ToxRVv with ToxRVc and ToxRVp at the deduced amino acid sequence level. The amino acid sequences were aligned for maximum homology using the MacDNASIS software (Hitachi). Dashes indicate missing sequences. Identical sequences are indicated by a black background. Solid underlining indicates the region supposed to be the transcription activation domain that binds to target promoter sequences (24, 31). Broken underlining indicates the putative transmembrane domain of ToxRVv (24, 31).

View larger version (47K): [in this window] [in a new window]   FIG. 4.   Comparison of ToxSVv with ToxSVc and ToxSVp at the deduced amino acid sequence level. The amino acid sequences were aligned for maximum homology using the MacDNASIS software (Hitachi). Dashes indicate missing sequences. Identical sequences are indicated by a black background. Broken underlining indicates the putative transmembrane domain of ToxSVv (24, 28).

View larger version (48K): [in this window] [in a new window]   FIG. 5.   Comparison of HtpGVv with HtpGVc (31) and HtpGVf (42) at the deduced amino acid sequence level. The amino acid sequences were aligned for maximum homology using the MacDNASIS software (Hitachi). Dashes indicate missing sequences. Identical sequences are indicated by a black background.

Expression of ToxR and ToxS. We tested whether the putative toxR_Vv and toxS_Vv open reading frames could express proteins of the expected sizes. We expressed the genes as GST fusion proteins. When fusion plasmids pCMM709 (toxR_Vv) and pCMM820 (toxS_Vv) were induced with IPTG, ToxR_Vv and ToxS_Vv were successfully expressed. The molecular masses of ToxR_Vv and ToxS_Vv were estimated to be 32 and 19 kDa, respectively, on the gel (Fig. 6A). The molecular masses of ToxR and ToxS were estimated to be the same by DNA sequence analysis.

View larger version (82K): [in this window] [in a new window]   FIG. 6.   Observation of toxRSVv expression by SDS-PAGE (A) and Western blot analysis (B). (A) Expression of ToxRVv and ToxSVv as GST fusion proteins. The molecular masses of ToxR and ToxS were estimated to be 32 and 19 kDa, respectively. Molecular size markers are indicated on the left. In lanes A through H, each pair represents lysates of E. coli DH5 containing corresponding plasmids before and after IPTG induction as follows: A and B, pGEX-3X; C and D, pCMM709; E and F, pGEX-2T; G and H, pCMM820. Arrowheads indicate expressed fusion proteins. Asterisks indicate GST bands expressed after induction. (B) Western blot analysis of ToxRVv expression in V. vulnificus MO6-25/O. Lanes: A, E. coli DH5 harboring pCMM610 without arabinose induction; B, same as lane A except that the cells were induced by 0.2% arabinose; C, V. vulnificus MO6-24/O, the isogenic wild-type strain; D, V. vulnificus CMM981, the toxRVv null mutant. The molecular mass on the left is that of ToxRVv.

Stimulation of the ctx promoter by ToxR_Vv. The V. cholerae ctx promoter is known to be activated by ToxR, even in an E. coli background (30). In order to verify that the putative ToxR_Vv protein is functionally related to ToxR_Vc, we tested whether ToxR_Vv could cross-activate the ctx promoter incorporated in the chromosome of E. coli VM2 (30). This E. coli strain harbors a recombinant lambda phage containing the upstream regulatory region of the ctx operon and the nucleotides encoding the first 23 amino acids of the A subunit of cholera toxin fused to the -galactosidase coding region (lacZ). In the present study, we placed toxR_Vv (pCMM610), toxRS_Vv (pCMM601), toxR_Vc (pCMM651), and toxRS_Vc (pCMM641) under the control of the promoter P_BAD. Using these constructs, we could compare the ctx activation potency of ToxR_Vv to that of its native counterpart, ToxR_Vc. The transformants harboring ToxR expression vectors were evaluated both before and after arabinose induction. E. coli VM2, either with or without the vector pBAD24, provided negative controls. ToxR_Vv could activate the ctx promoter with an efficiency comparable to that of native ToxR_Vc (Fig. 7). Stimulation of the ctx promoter by either ToxRS_Vv or ToxR_Vv showed no significant difference in almost all of the experiments. However, the induction of ToxRS_Vc resulted in the greatest expression of the ctx promoter. The -galactosidase activity induced by ToxRS_Vc was usually 100 to 150 U higher than in other induction experiments at the same time point, while ToxRS_Vv showed no such enhancement.

View larger version (24K): [in this window] [in a new window]   FIG. 7.   Cross-activity of toxRVc and toxRVv in the stimulation of the ctx promoter in an E. coli background. Both genes were cloned into expression vector pBAD24 and transformed into E. coli VM2 with a ctx promoter-lacZ fusion in the chromosome. Cultures were induced by 0.2% arabinose at time zero. Each error bar indicates the standard error of the mean of triplicate experiments. Symbols: × and , E. coli VM2 without plasmids, before and after induction, respectively;  and , VM2 carrying pBAD24, before and after induction, respectively;  and , VM2 carrying pCMM610, before and after induction, respectively;  and , VM2 carrying pCMM651, before and after induction, respectively.

Stimulation of vvhA gene expression by ToxRS_Vv in E. coli. The effect of ToxRS_Vv on the expression of vvhA gene was examined in an E. coli background. The vvhA gene was subcloned into pACYC184, which is compatible with the pBAD24 vector, to construct pJL961. Plasmid pJL961 was cotransformed with an expression vector bearing toxRS_Vv, toxR_Vv, toxRS_Vc, or toxR_Vc under the control of the P_BAD promoter into E. coli DH5. Only ToxRS_Vv could significantly stimulate hemolysin production from the vvhA gene in the E. coli background (Fig. 8A). Repeated experiments showed that toxRS_Vv induction increased hemolysin production 5- to 10-fold. In contrast to the results of ctx promoter induction experiments, both ToxR_Vv and ToxS_Vv were required to stimulate hemolysin production.

View larger version (27K): [in this window] [in a new window]   FIG. 8.   Stimulation of the vvhA gene by ToxRSVv in an E. coli background. Compatible plasmid pJL961 carrying the 3.4-kb vvhA gene fragment was cotransformed with a plasmid carrying toxRSVv (pCMM601), toxRVv (pCMM610), toxRSVc (pCMM641), or toxRVc (pCMM651) under the control of the PBAD promoter. (A) Hemolytic activity in the culture supernatant was assayed at the indicated time intervals. Cultures were induced with 0.2% arabinose. Each error bar indicates the standard error of the mean of triplicate experiments. Symbols: ×, pJL961 only;  and , pCMM601 before and after induction, respectively;  and , pCMM610 before and after induction, respectively;  and  with solid line, pCMM641 before and after induction, respectively;  and  with broken line, pCMM651 before and after induction, respectively. (B) ToxRVv levels in the bacterial cells were measured by Western blot analysis. Top, pCMM610 after induction; bottom, pCMM601 after induction. The numbers above the photographs are the times after induction.

Effect of a toxR_Vv mutation on hemolysin production. A V. vulnificus toxR_Vv mutant strain was constructed by inserting a suicide plasmid into the chromosomal copy of toxR_Vv by homologous recombination. Null mutation was confirmed by Western blot analysis of the whole-cell lysate using rabbit anti-ToxR_Vv serum (Fig. 6B).

View larger version (40K): [in this window] [in a new window]   FIG. 9.   Effect of toxRVv mutation on V. vulnificus hemolysin production and its complementation by the wild-type (wt) gene. (A) Hemolysin production profiles of the toxRVv mutant (CMM981) and isogenic wild-type strain MO6-24/O. Peak hemolysin production by the mutant decreased to about half of the level of the isogenic wild-type strain. (B) Partial complementation of the hemolysin production defect of the mutant in trans by plasmid-encoded wild-type toxRSVv. Hemolysin production was significantly restored by complementation. Percent peak hemolysin production of the mutant before or after complementation was compared with that of the wild type. The symbols * and ** indicate statistically significant differences from the wild type by the Student t test at P < 0.005 and P < 0.001, respectively. The mean and the standard error of the mean were calculated from triplicate experiments.

Effect of a toxR_Vv mutation on the OMP profile. In V. cholerae, ToxR was reported to activate ompU expression and reciprocally inhibit ompT expression (9, 31, 47). One of the most prominent phenotypic changes after the toxR_Vc mutation was the alteration in OMP expression (31). Mutation of toxR_Vv also resulted in a change in the OMP profile, but the pattern of change was different. V. vulnificus MO6-24/O showed two major OMP bands of 36 and 38 kDa on SDS-PAGE. The toxR_Vv mutation resulted in reciprocal changes in the expression of the two major OMP bands. V. cholerae OmpT and OmpU show molecular masses of 40 and 38 kDa. Contrary to the case with V. cholerae, the high-molecular-mass OMP was positively regulated by toxR_Vv and the toxR mutation increased the expression of the low-molecular-mass OMP (Fig. 10A). It is noteworthy that stationary cells grown at 25°C, rather than 37°C, showed a more prominent difference in the OMP profile.

View larger version (26K): [in this window] [in a new window]   FIG. 10.   Effect of toxRVv mutation on OMP expression. (A) OMP profiles of toxRVv mutant CMM981 and isogenic wild-type strain MO6-24/O. V. vulnificus strains were grown in 2.5% NaCl heart infusion at 25°C to stationary phase. OMPs were prepared and analyzed by SDS-12% PAGE. The two major bands appeared to have molecular masses of 38 and 36 kDa. Lanes: A, CMM981; B, MO6-24/O. The values on the left are molecular masses in kilodaltons. (B) Amino-terminal amino acid sequences of the two major OMP bands and their similarities to proteins in the sequence databases. Boldface letters indicate identical sequences. Amino acid residue numbers are shown at the beginning and end of each sequence. ZP2, mouse zona pollucida sperm-binding protein 2 precursor.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In the present study, we cloned and sequenced the toxRS_Vv operon from V. vulnificus type strain ATCC 29307. We also proved that ToxR_Vv is expressed in V. vulnificus at the expected size and showed that ToxRS_Vv could upregulate hemolysin production in both the E. coli and V. vulnificus backgrounds.

Pathogenic Vibrio species such as V. cholerae, V. parahaemolyticus, and V. vulnificus cycle between two distinct environments: the human body and the estuarine environment. Maintaining life in such different environments requires many different gene products and a finely tuned regulatory system for the genes. The virulence regulation of V. cholerae is under the master control of a cascade regulatory system in which cytoplasmic membrane proteins ToxR and TcpP control expression of the AraC-type transcriptional activator ToxT (2, 9, 16, 17, 18, 19). Many virulence factors, such as cholera toxin and the toxin-coregulated pilus, are under the control of this ToxR-TcpP regulon. More than 20 V. cholerae genes proved to constitute the ToxR-TcpP regulon (47). Several environmental factors, such as temperature, medium pH, osmolarity, and aeration are known to serve as the regulatory signals of the ToxR regulon (31, 40, 51).

ToxR and TcpP of V. cholerae are cytoplasmic membrane proteins with a cytoplasmic amino-terminal DNA-binding domain and a periplasmic carboxy-terminal domain that interacts with the other transmembrane proteins ToxS and TcpH, respectively (10, 16, 31, 32). The regulatory network in the ToxR and TcpP regulon seems to be multilayered and hierarchical. Most of the genes in the regulon are controlled by a second AraC type of transcription activator, ToxT, which is located at a novel pathogenicity island and is itself positively regulated by ToxR and TcpP (10, 16, 19, 20, 34). ToxT then directly activates a number of virulence genes, such as those producing the toxin-coregulated pilus and accessory colonization factors (9, 47). Recently, the global regulator cyclic AMP (cAMP)-cAMP receptor protein (CRP) was also reported to influence the expression of the ToxR regulon under various environmental conditions (46, 47). Our previous report that V. vulnificus hemolysin production was influenced by temperature and/or salinity change and was inhibited by exogenous glucose (S. E. Lee, C. M. Kim, S. Y. Kim, S. J. Kim, P. Y. Ryu, H. C. Lee, J. S. Oh, S. S. Chung, and J. H. Rhee, Abstr. 98th Gen. Meet. Am. Soc. Microbiol. 1998, abstr. B-172, p. 84, 1998) suggests the possible existence of a multilayered regulatory network in V. vulnificus.

In 1993, Lin et al. discovered that V. parahaemolyticus has a homolog of the V. cholerae toxRS operon which mediates environmentally induced regulation of the TDH (25). Cherwonogrodzky et al. (3, 4, 5) reported that TDH production by V. parahaemolyticus could also be modulated by cultural parameters, as observed in V. cholerae. These observations encouraged the Nishibuchi group to speculate on the existence of the ToxR_Vc-like signal-transducing regulator in V. parahaemolyticus (25). In that report, they showed that ToxR_Vp promoted the expression of the tdh gene. Subsequently, Reich and Schoolnik cloned other toxRS genes from V. fischeri (43). The existence of a ToxT homolog system in V. parahaemolyticus and V. fischeri has yet to be reported. In the present study, we found that the ToxRS_Vv proteins were homologous to previously reported ToxRS proteins. The operon was found in all of the clinical and environmental V. vulnificus isolates studied by colony blot and Southern hybridization assays. Low-stringency colony blot hybridization analyses showed the presence of toxRS_Vv homologs in various Vibrio spp. (J. H. Rhee, S. H. Shin, and S. E. Lee, unpublished data). These findings suggest that the toxRS operon played important roles in the survival of numerous Vibrio species throughout evolution. Many virulence factors of Vibrio spp. might have come under the control of ToxR serendipitously. This hypothesis was also proposed by Lin et al. (25) on the basis of the finding that the tdh genes of V. parahaemolyticus appeared to be transferred among strains as transposon-like units (52) and were discovered only in subpopulations of the bacterium, while the toxRS_Vp genes were harbored by every V. parahaemolyticus strain tested.

As was the case for V. parahaemolyticus TDH production, production of V. vulnificus hemolysin was affected by changes in environmental parameters (Lee et al., 98th Gen. Meet. Am. Soc. Microbiol.). The experimental results gave us a strong clue to the existence of toxRS homologs in V. vulnificus. The most remarkable parameters that change in the course of the infection process should be temperature and salinity (osmolarity). The optimal temperature and salinity of estuaries that give the highest isolation rate for V. vulnificus have been reported to be about 25°C (35, 53) and over 23 (50), respectively. When V. vulnificus was cultured at a salinity of 2.5%, it produced the hemolysin poorly. More hemolysin is produced at 37°C than at 25°C. At a salinity of 0.9%, the hemolysin was produced more efficiently than at 2.5%. Shifting from 2.5% salinity and 25°C (reflecting the estuarine environment) to 0.9% salinity and 37°C (representing the human internal milieu) resulted in a 2.5-fold increase in hemolysin production (Lee et al., 98th Gen. Meet. Am. Soc. Microbiol.). These findings contradict reports that cholera toxin and TDH are produced more efficiently at 30°C than at 37°C and under high osmolarity (4, 13, 31). V. vulnificus responded to the temperature and/or salinity shift by increasing hemolysin production. This suggests that V. vulnificus senses environmental parameter changes during the infection process and can become more virulent in the human body. The results also suggested that V. vulnificus should have a common signal recognition-transduction system that senses changes in temperature and/or salinity and modulates hemolysin production. The ToxRS_Vv system was identified as the strongest candidate for the signal transduction system and proved to play an important role in the modulation of hemolysin production in the present study. However, the operational mode of the ToxRS_Vv system may be different from that of V. cholerae or V. parahaemolyticus because temperature and osmolarity affect downstream response contradictorily. This contradiction was also noted in the regulation of OMP expression by ToxRS. In V. vulnificus, the high-molecular-mass (38-kDa) OMP was positively regulated and the low-molecular-mass (36-kDa) OMP was negatively regulated by ToxRS_Vv. On the other hand, V. cholerae OmpU (low molecular mass [38 kDa]) was positively regulated and OmpT (40 kDa) was negatively regulated by ToxR_Vc. The 38-kDa OMP of V. vulnificus was found to be a homolog of OmpU_Vc by amino-terminal amino acid sequencing. The amino-terminal amino acid sequence of the 36-kDa OMP showed no homology to known OMPs in the molecular biological databases. The question of whether the OMP serves as a functional homolog of OmpT_Vc should be answered by further molecular biological studies.

The finding that hemolysin production is under the control of a transmembrane signal transducer may explain the raison d'être of the exotoxin. Wright and Morris proposed that hemolysin might not play a major role in the pathogenesis of V. vulnificus septicemia (55). They reported that the 50% lethal dose of the vvh-inactivated mutant for mice was no different from that of the wild-type strain. Whether the mouse septicemia model exactly reproduces the human disease needs further investigation. Previously, our group showed that hemolysin induced vasodilatation independently of nitric oxide synthase at a concentration more than 100-fold lower than that needed to produce cytotoxicity or hemolysis (22, 23). Onset of septic shock usually coincides with or precedes the earlier skin manifestations such as petechia, ecchymosis, or vesicles (38, 39). Hemolysin is highly cytotoxic and was proposed to play a significant role in producing typical skin manifestations (14). Pharmacological vasodilatation and hypotension seem to occur prior to hemolysin-induced overt damage of endothelial cells and surrounding tissues. We suppose that in vivo hemolysin production should be tightly regulated at the time and place needed. We propose that the ToxRS_Vv system is part of the regulatory system.

It is widely accepted that ToxS stabilizes ToxR in a conformation that is optimal for transcriptional activation, possibly as a heterodimer or as a ToxR homodimer (10, 11, 36). In the present study, we found that ToxRS_Vv functionally mimics ToxRS_Vc in stimulating the ctx promoter in an E. coli background. This finding is supported by the significant homologies of the proteins. ToxRS_Vc showed the highest promoting activity, as expected. ToxR_Vv alone was as potent as ToxR_Vc in promoting the ctx gene. The effect of ToxRS_Vv was not superior to that of ToxR_Vv. In contrast, in the E. coli background, only overexpression of toxRS_Vv, but not toxR_Vv alone, toxRS_Vc, or toxR_Vc alone, resulted in significant activation of the vvhA gene. These results suggest an important role of ToxS_Vv in the optimal activation of the vvhA promoter but not in that of the ctx promoter. For optimal activity on the vvhA promoter, ToxR_Vv may need the greater stabilization provided by ToxS_Vv binding. But this possibility was ruled out by Western blot analyses of ToxR_Vv levels in the E. coli background. The amount of ToxR_Vv was not affected by the presence or absence of ToxS_Vv when the former protein was overexpressed. Under physiological conditions, ToxR forms a heterodimer with ToxS and become stabilized. ToxR-ToxR homodimers were also formed when ToxR was overexpressed (36). The homodimers were supposed to be as resistant as heterodimers to destruction, since overexpressed ToxR_Vv alone showed a band intensity similar to that of overexpressed ToxRS_Vv in the present study. These results indicate that ToxS_Vv might play an active role in regulating vvhA expression other than mere stabilization of ToxR_Vv. Pfau and Taylor (42) showed that the DNA-binding and transcription activities could be separated and that chaperone-like activity of ToxS may be required for the formation of the transcription activation complex but not the ToxR-DNA complex at the ctx promoter. At the vvhA promoter region, proper formation of a ToxR-ToxS_Vv complex might be a prerequisite for both DNA binding and formation of the transcription activation complex. Another explanation for the fact that only ToxRS_Vv could activate the vvhA promoter in the E. coli background is that ToxR_Vv might not be a major transcription activator of the vvhA gene. Under physiological conditions, full stabilization of ToxR_Vv by ToxS_Vv would take effect in a small-scale activation or fine tuning of the vvhA promoter while some other major transcription factor(s) is responsible for regulating hemolysin production on a large scale(s). The hypothesis comes from the result that the toxR_Vv null mutation decreased hemolysin production to only half of that of the isogenic wild-type strain. Another factor, such as CRP, would play the major role as a transcription activator. We recently showed that a vvhA-luxCDABE fusion reporter requires cAMP and CRP in an E. coli background (1). In that report, we also showed that a putative chemical crp mutant of V. vulnificus did not produce hemolysin. ToxR_Vv would play an auxiliary role along with CRP as a fine tuner of vvh expression, while cAMP and CRP determine whether the switch of the hemolysin gene should be turned on or off. The presence of 0.5% glucose in the culture medium totally shuts off hemolysin production (1; Lee et al., 98th Gen. Meet. Am. Soc. Microbiol.). In the 5' untranslated region of vvhBA, a region showing homology to the consensus CRP binding sequence was noted (56). We are currently investigating other factors acting at the promoter region of vvhA.

ToxR_Vc binds to tandem TTTTGAT repeats present in the upstream region of the ctx promoter (32, 41). The ToxR_Vc binding site of the toxT upstream DNA has no TTTTGAT repeats but is rich in inverted-repeat elements (17, 18). In the vvhA upstream DNA sequence, neither TTTTGAT repeats nor the inverted repeats present around the toxT upstream ToxR_Vc-binding region were observed in the vvhA promoter region (56). To elucidate the mechanism by which the vvhA promoter is activated by ToxR_Vv, the ToxR_Vv-binding sites at the promoter region should be mapped and their interaction with the transcription activator needs to be investigated by using an in vitro transcription assay system.