Identification of quorum sensing-related regulons in Vibrio vulnificus by two-dimensional gel electrophoresis and differentially displayed reverse transcriptase PCR

Authors

  • Na-Ri Shin,

    1. Department of Infectious Diseases, BK21 for Veterinary Science and KRF Zoonotic Priority Research Institute, College of Veterinary Medicine, Seoul National University, Korea
    Search for more papers by this author
  • Deog Yong Lee,

    1. Department of Infectious Diseases, BK21 for Veterinary Science and KRF Zoonotic Priority Research Institute, College of Veterinary Medicine, Seoul National University, Korea
    Search for more papers by this author
  • Han Sang Yoo

    1. Department of Infectious Diseases, BK21 for Veterinary Science and KRF Zoonotic Priority Research Institute, College of Veterinary Medicine, Seoul National University, Korea
    Search for more papers by this author

  • Present address: Na-Ri Shin, Center for Infectious Diseases, Division of Biodefense Research, Korea National Institute for Health, Seoul, 122-701 Korea.

  • Editor: Kai Man Kam

Correspondence: Han Sang Yoo, Department of Infectious Diseases, College of Veterinary Medicine, Seoul National University, Seoul 151-742, Korea. Tel.: +82 2 880 1263; fax: +82 2 874 2738; e-mail: yoohs@snu.ac.kr

Abstract

Vibrio vulnificus is thought to employ a quorum-sensing system to control the expression of a global gene. In this study, proteomes and transcriptomes of a lacZ null mutant, VvSRΔZ, and a luxS–smcR double mutant, VvSRΔZSR, were compared with the parent strain, VvAR, by means of two-dimensional gel electrophoresis (2D-PAGE) and differentially displayed reverse transcriptase PCR (DDRT-PCR). 2D-PAGE analysis showed that 36 protein spots were differentially expressed, 14 of which have been identified by peptide-mass fingerprinting. The expression of eight cellular proteins was repressed by luxS and smcR mutation: Zn-dependent protease, 6-phosophofructokinase, periplasmic ABC-type Fe3+ transport system, deoxyribose-phosphate aldolase, phosphomannomutase, orotidine-5′-phosphate decarboxylase, uridylate kinase, and an unidentified protein. These proteins are involved in virulence, adaptation to environmental stress, biosynthesis of LPS, and cell multiplication. Phage shock protein A, a chemotaxis signal transduction protein, and an uncharacterized low-complexity protein were activated in the cellular components of the luxS-smcR mutant. However, only three proteins, of unknown function, were identified in the extracellular components of the mutants. Analysis of transcriptomes with DDRT-PCR showed that two genes, phosphoribosylformylglycinamidine synthase and ATP-dependent protease HslVU protease were regulated at the transcriptional level by luxS and smcR gene mutation. The results from this study show conclusively that luxS/smcR quorum sensing endows a global change in gene expression to V. vulnificus.

Introduction

Quorum sensing (QS) is an intercellular signalling mechanism by which bacteria monitor their own population density or that of other populations by recognizing local concentrations of chemical molecules, referred to as autoinducers, produced by the same bacterial species (Shin et al., 2005). The signal molecules regulate the expression of target genes at a minimal threshold stimulatory concentration. QS circuits have been identified in over 30 species of Gram-negative bacteria. In many Gram-negative bacteria, the bacterial expression of some virulence factors is regulated by the QS system. Gram-negative QS bacteria typically possess proteins homologous to the LuxI and LuxR proteins of Vibrio fischeri (Linkous & Oliver, 1999). Many studies have been undertaken to establish an association between QS and phenotypes in a number of bacterial species, and have revealed that QS controls processes including bioluminescence, biofilm formation, antibiotic synthesis, and the production of several virulence factors (Engebrecht et al., 1983; Bainton et al., 1992; Beck von Bodman & Farrand, 1995; Stintzi et al., 1998; Davies et al., 1998).

Vibrio vulnificus is a Gram-negative estuarine bacterium known to be a significant human pathogen. A V. vulnificus infection, acquired via direct contact or through the gastrointestinal route, is characterized by food-borne septicemia and skin infections, with ulcer and edema in many clinical cases (Kumamoto & Vukich, 1998; Shin et al., 2004). The infection is characterized by a high fatality rate, particularly in immunocompromised individuals and in those with underlying conditions such as hemochromatosis, liver cirrhosis, or alcoholism (Nealson & Hastings, 1979; Brent et al., 1991; Bullen et al., 1991; Chan et al., 1994). Most of the fatal cases are caused by septic shock (Linkous & Oliver, 1999; Strom & Paranjpye, 2000). Also, it was demonstrated as in vitro and in vivo models in our previous studies (Shin et al., 2004, 2005).

Vibrio vulnificus produces a number of factors implicated in virulence and pathogenesis, such as capsular polysaccharide, lipopolysaccharide (LPS), elastase, cytolysin, metalloprotease, siderophores, and phospholipase (Wright et al., 1981; Hanne & Finkelstein, 1982; Gray & Kreger, 1985; Shao & Hor, 2001; Jeong et al., 2003), and uses several genes for surviving or decreasing environmental stress (Wright et al., 1981; Jeong et al., 2003; Park et al., 2004; Rhee et al., 2004).

A few works have recently identified the role of QS in V. vulnificus pathogenesis. smcR appears to play an important role in starvation adaptation and in the regulation of many stationary phase-regulated genes, including some virulence factors (McDougald et al., 2001; Shao & Hor, 2001). The luxS mutant also showed a decrease of in vivo virulence and regulation of protease and haemolysin production (Kim et al., 2003). In addition, luxS and smcR in V. vulnificus could play a role in bacterial survival in the host by enhancing proliferation and adjustment to environmental changes (Shin et al., 2004, 2005). Analyses of bacterial proteomes and/or regulons have been established as standard tools for the study of diverse cellular functions and regulations. They have also frequently been used for studies of the disease process and of bacterium−host interactions. Recently, proteomes related to Vibrio spp. (Kan et al., 2004; Xu et al., 2005; Lee at al., 2006) or quorum-sensing (Arevalo-Ferro et al., 2003; Riedel et al., 2003) have been analysed, and several proteins possibly linked to pathogenesis have been characterized. These findings suggest that QS in V. vulnificus constitutes a global regulatory system. However, there have been no studies on the luxS/smcR QS regulons (QSR) of V. vulnificus.

On the basis of this knowledge, differentially displayed reverse transcriptase PCR (DDRT-PCR) and proteomic analysis were employed to identify luxS- and smcR-controlled mRNA and proteins in V. vulnificus, and several QS regulons were demonstrated in this study.

Materials and methods

Preparation of differentially expressed proteins

Bacterial strain and culture conditions

The strains used in this study were VvAR, VvSRΔZ, and VvSRΔZSR (Shin et al., 2005). All strains were cultured in Luria–Bertani media (Difco, Detroit, MI) supplemented with 2.5% NaCl (LBS) with vigorous shaking at 30°C. Cells in the mid to late log phase were collected by spectrophotometrical monitoring at 600 nm.

Sample preparation for cellular proteins

For the preparation of cellular proteins, bacterial cells from 20-mL cultures were harvested by centrifugation (5000 g for 30 min), and washed twice with phosphate-buffered saline (PBS, pH 7.4). After resuspension of the cells in 100 mL of lysis buffer [7 M urea, 2 M thiourea, 4% Cholamidopropyl dimethylammoniopropane sulfate (CHAPS), 70 mM dithiothreitol (DTT)], cells were disrupted by sonication at 4°C for 15 s. The supernatants were collected for analysis by centrifugation at 100 000 g at 4°C for 1 h. The protein concentrations were measured using a BCA protein assay kit (Pierce, Rockford, IL).

Sample preparation for extracellular proteins

For the preparation of extracellular proteins, the supernatant of 200 mL of bacterial culture was separated by centrifugation at 5000 g for 30 min. After filtering the supernatant with a 0.45-μm filter, the proteins in the supernatant were precipitated with saturated ammonium sulfate at 4°C overnight and harvested by centrifugation at 12 000 g for 10 min at 4°C, and the precipitants were resuspended in ddH2O. After removal of the remaining salts by dialysis, proteins in the solutions were thoroughly freeze-dried. The protein pellets were solubilized in 100 μL of lysis buffer after freeze drying.

Analysis and identification of differentially expressed proteins

Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE)

2D-PAGE was carried out in a vertical electrophoresis (Amersham Pharmacia Biotech), IPGPhor, for the first-dimensional isoelectric focusing using immobiline DryStrip (24 cm, pH 4–7, Amersham Pharmacia Biotech, Piscataway, NJ), and with a Ettan DALT 2, 2-DE system (Amersham Pharmacia Biotech) for the second-dimensional SDS-PAGE (11% gel). A total amount of 800 or 100 μL of cellular or extracellular protein, respectively, was loaded onto the IPG (immobilized pH 4–7) strips by in-gel rehydration (Amersham Pharmacia Biotech) using IPGPhor (18 cm, Amersham Pharmacia Biotech). Isoelectric focusing was carried out successively at 0−300 V for 3 h. After isoelectric focusing, strips were incubated with equilibration buffer 1 (50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, BPB, 2% DTT) and buffer 2 (50 mM Tris-HCl, pH 8.8, 6 M urea, 30% glycerol, 2% SDS, BPB, 2% DTT, 2.5% iodoacetamide) for 15 min each. The equilibrated strip was placed on polyacrylamide gradient slab gel (7–17% gradient). Separation was performed at 10 mA gel−1 in running buffer (25 mM Tris, pH 8.8, 198 mM glycine, and 0.1% SDS) until P-bromophenacyl bromide (BPB) reached the bottom of the gel. After electrophoresis, spots of cellular proteins were visualized by staining with a Coomassie brilliant blue R protocol that is compatible with the matrix-assisted laser desorption/ionization-time of flight mass spectrometer (MALDI-TOF MS), and extracellular protein spots were visualized by silver staining for high sensitivity. The stained gels were scanned using a densitometer 800 (Bio-Rad, Hercules, CA). The digitalized image was analysed with pdquest software (version 6.2, Bio-Rad).

MALDI-TOF MS and database searches

Proteins were identified by peptide mass fingerprinting methods (Jensen et al., 1999; Pandey et al., 2000). Briefly, the selected protein spots were cut from the gel by spot cutters (Bio-Rad). After excised gel spots had been washed with 100 μL of ddH2O for 5 min, 100 μL of 100% acetonitrile, 0.1 M ammonium bicarbonate, and 100% acetonitrile were added, and removed in turn after shaking for 5 min. The gel pieces were dried in a speed vacuum concentrator for 5 min and were rehydrated with 20 μL of 50 mM ammonium bicarbonate for 45 min on ice. After removal of the solution, 30 μL of 50 mM ammonium bicarbonate containing 0.2 μL of modified trypsin (Promega, Madison, WI) was added. The digestion was performed overnight at 37°C. After the removal of residual trypsin, the peptides were desalted using a C18 nanoscale (porus C18) column. The peptides were eluted with 0.8 μL of matrix solution including 70% acetonitrile (Merck), 0.1% TFA (Merck), and 20 mg mL−1α-cyano-4-hydroxycinamic acid (Sigma). The eluted peptides were spotted onto a stainless-steel target plate. Masses of peptides were determined using a MALDI-TOF MS (Micromass, Manchester, UK). Calibration was performed using internal mass of trypsin auto digestion product (2211.105 mz−1). Peptide masses were matched with the theoretical peptides of all proteins in the NCBI database using mascot software and profound software.

Identification of differentially expressed mRNA

Bacterial strains and culture

Three strains, VvAR, VvSRΔZ, and VvSRΔZSR, were grown in LBS with vigorous shaking at 37°C. Overnight cultures of the strains were inoculated in 20 mL of fresh LBS to a concentration of 2 × 107 CFU mL−1. The fresh cultures of V. vulnificus were grown under the same conditions, and the bacteria were collected by centrifugation for total RNA extraction at the mid-logarimic stage by spectrophotometrical monitoring at 600 nm.

Total RNA isolation and cDNA synthesis

Total RNA was extracted from bacterial cells with a RNeasy mini kit (Qiagen Co.). Three microliters of total RNA was mixed with 10 μM ramdom hexamer (Invitrogen), and incubated at 80°C for 3 min. After being chilled on ice for 2 min, the tube was added to 20 μL of total volume with the following reagents: 2 μL of 10 × buffer (Invitrogen), 1 μL of 10 mM dNTP mix (Invitrogen), 0.5 μL of RNase inhibitor (40 U μL−1) (Invitrogen), and 1 μL M-MLV reverse transcriptase (200 U μL−1) (Invitrogen). The mixture was incubated at 42°C for 90 min, heated at 94°C for 2 min, and chilled on ice for 2 min. Remaining RNA was removed by incubation with 1 μL of RNase H (Invitrogen) at 37°C for 20 min. The synthesized cDNA was stored at −20°C until use.

DDRT-PCR

DDRT-PCR was carried out according to the manufacturer's protocol with a GeneFishing™ DEG kit (Seegene Co., Korea). Briefly, 2 μL of first-strand cDNA was mixed with the following reagents in a PCR tube for the PCR reaction: 5 μL of 10 × buffer without MgCl2, 2 μL of 5 μM arbitrary ACP (one of arbitrary ACPs), 1 μL of 5 μM arbitrary ACP (another of arbitrary ACPs), and 4 μL of 2.5 mM dNTP. The total volume was adjusted to 49.5 μL using ddH2O. After the PCR mixture had been placed in a preheated thermal cycler (Perkin Elmer Co., Foster, CA), cDNA was amplified using a hot-start PCR. After initial denaturation at 94°C for 3 min, two cycles of denaturation at 94°C for 40 s, annealing at 50°C for 3 min and polymerization at 72°C for 1 min were performed, and 40 cycles of amplification reaction at 94°C for 40 s, 65°C for 40 s and 72°C for 40 s followed. Final polymerization was carried out at 72°C for 5 min. The amplified PCR products were separated in 2.5% agarose gel, and visualized under a UV trans-illuminator after staining with ethidium bromide. PCR products expressed differentially among the strains were extracted from the gel using a gel extraction kit (Qiagen Co.). Nucleotide sequences of the products were blasted with database in GenBank.

Confirmation of differentially expressed mRNA

The specific primers for identified genes were synthesized based on the accession numbers gil27359902 and gil27360771 (Table 1). The PCR mixture contained 1 μL of 10 μM primers, 5 μL of 10 × PCR buffer, 3 μL of 25 mM MgCl2, 4 μL of 10 mM dNTPs, and 0.5 μL of Taq polymerase with 2 μL cDNA in a 50-μL reaction volume. PCR amplification of cDNA was performed at 94°C for 5 min, followed by 20 cycles as follows: 94°C for 30 s, 55°C for 30 s and 72°C for 50 s. A final extension was performed at 94°C for 7 min. Expression levels of the genes were analysed by comparison of intensity of PCR products using a Gel Doc XR (Bio-rad, Co.) after agarose gel electrophoresis.

Table 1.   The specific primers for genes identified by DDRT-PCR
BandPrimerNucleotide sequences (5′–3′)PCR
product
(bp)
AForwardATGGTTTTACGTACATCTTCAAC500
ReverseACCTTCTTGATCACCATTGGTG 
BForwardATGAGTCAGCTTTACTGCC233
ReverseTGGCGTCGAATGCAACTTG 

Results

Growth of V. vulnificus and prefractionation of cellular compartments

Proteomes of VvAR and its mutants were separated into cellular and extracellular proteomes for analysis using 2D-PAGE. Cellular proteins were prepared by lysis of the harvested cells employing sonication and separation of the crude extract from the cell debris by centrifugation. Extracellular proteins were prepared by a combination of centrifugation, and freeze drying followed by precipitation with ammonium sulfate and dialysis.

Comparison of protein profiles

In order to characterize the quorum sensing of V. vulnificus, the protein expression pattern of the wild type was compared with those of its luxS- and smcR-deficient mutants by 2D-PAGE. A pH gradient from 4 to 7 was used to analyse protein samples from three independent growths. A number of spots were detected in colloidal Coomassie-stained gel after separation of cellular proteins, while extracellular proteins were visualized by silver staining for higher sensitivity. Figures 1 and 2 show the cellular and extracellular protein patterns, respectively, of VvAR, VvSRΔZ, and VvSRZSR. The 2D gel of cellular proteins comprises protein patterns with a high density of spots in the neutral range (Table 2 and Fig. 1). In contrast, the spots appeared to be mostly located in the acidic range in the profile of extracellular proteins (Fig. 2). Matching and comparing the respective 2D-PAGE maps with the image allowed the identification of the respective differentially expressed protein. In cellular protein profiles, 22 protein spots were found showing different expression patterns in the genotypes, and 11 of these 22 spots were significantly regulated by luxS and smcR mutation (Fig. 3 and Table 2). The intensities of cellular protein spots were compared with each strain (Tables 2 and 3). The expression level of each spot was different depending on the mutation of VvAR. The rest spots were up- or down-regulated by lacZ deletion, and VvSRΔZ alone was expressed differentially among the three strains. The optical densities of proteins not regulated by luxS/smcR quorum sensing are shown in Table 3. In the extracellular protein profiles, 14 candidates were found. Of these candidates, the intensities of five spots increased whereas those of four spots decreased by luxS and smcR mutation (Fig. 4).

Figure 1.

 Electrophoretic analysis of cellular proteins of Vibrio vulnificus mutants using 2D-PAGE.

Figure 2.

 Electrophoretic analysis of extracellular proteins of Vibrio vulnificus mutants using 2D-PAGE.

Table 2.   ODs of Vibrio vulnificus cellular protein spots regulated by luxS/smcR quorum sensing
ExpressionSpot no.Strain
VvARVvSRΔZVvSRΔZSR
Decrease20.440.250.06
30.640.510.28
40.390.410.14
50.630.430.18
70.570.410.27
120.950.950.24
150.410.370.12
200.290.140.08
Increase100.200.290.40
180.120.110.26
190.130.160.39
Figure 3.

 Comparative analysis of 2D gel images for cellular proteins among the strains.

Table 3.   Optical densities of Vibrio vulnificus cellular protein spots expressed differentially among the genotypes but nonregulated by luxS/smcR quorum sensing
Most expressed strainSpot no.Strain
VvARVvSRΔZVvSRΔZSR
VvAR60.760.490.46
80.270.170.13
90.80.510.48
110.430.310.25
130.530.150.26
160.610.30.24
VvSRΔZ & VvSRΔZSR170.210.360.33
VvSRΔZ10.240.670.22
VvSRΔZSR140.660.480.96
Expression in only VvARAO
BO
Figure 4.

 Comparative analysis of 2D gel images for extracellular proteins among the strains.

Identification of quorum sensing-regulated proteins

Because the sequence of the V. vulnificus genome is complete, application of MALDI-TOF is suitable for the identification of the luxS- and smcR-regulated proteins. The proteins selected from the comparison of intensity were identified by peptide mass fingerprinting. The peptide masses were then searched in the NCBI database. Significance was endowed with protein scores greater than 74 (P<0.05) belonging to organism V. vulnificus. Tables 4 and 5 show the identified cellular and extracellular proteins, respectively, and their NCBI numbers.

Table 4.   Identification of differentially expressed cellular proteins by MALDI-TOF
Spot no.NCBI no.Protein identified
2gi∣27366846Predicted Zn-dependent protease
3gi∣273646586-phosphofructokinase
4gi∣27365023ABC-type Fe3+ transport system,
Periplasmic component
5gi∣27366253Orotidine-5′-phosphate decarboxylase
7gi∣27365209Uridylate kinase
12gi∣27365053Phosphomannomutase
15gi∣27365085Deoxyribose-phosphate aldolase
20 Not matched
10gi∣27366158Phage shock protein A
18gi∣27365304Chemotaxis signal transduction protein
19 Uncharacterized low-complexity protein
(from Q-TOF ms)
Table 5.   Identification of differentially expressed extracellular proteins by MALDI-TOF
NCBI no.Protein identified
gi∣27366101Conserved hypothetical protein
gi∣27367131Outer membrane receptor protein
gi∣4927284Secretory peroxidase

Identification of genes transcriptionally regulated by quorum sensing

About 40 arbitrary primers were combined for DDRT-PCR, which of display results were similar among the strains in most of combination. Amplification with arbitrary primers numbers 19 and 20 revealed about 2000 and 332 bp of differentially expressed PCR products in VvSRΔZSR. Two differentially displayed amplicons were detected, identified as phosphoribosylformyl-glycinamidine (FGAM) synthase and ATP-dependent protease HslUV, by sequencing and blast (Table 6), and confirmed by PCR amplification with specific primers of the genes. The gene encoding FGAM synthase showed higher expression at the transcriptional level in VvSRΔZSR than those in VvAR and VvSRΔZ, whereas mRNA expression of the HslVU gene was lower in VvSRΔZSR than that in the parent strain (Fig. 5).

Table 6.   Identification of mRNA expressed differentially by luxS/smcR quorum-sensing mutation
Band
(bp)
NCBI
no.
Name of proteinHomology
(%)
  1. The mRNA was sequenced and blasted in GenBank.

2000gi∣27359902Phosphoribosylformyl-glycinamidine96
(FGAM) synthase, synthase domain 
332gi∣27360771ATP-dependent protease HslVU,94
ATPase subunit 
Figure 5.

 Confirmation of two genes identified by DDRT-PCR in 2.0% agarose gel. These amplicons were amplified with specific primers. Lane M, 100-bp DNA ladder; lanes 1, 2 and 3, PCR products from VvAR, VvSRΔZ and VvSRΔZSR cDNA with specific primers for the FGAM synthase gene; lanes 4, 5 and 6, PCR products from VvAR, VvSRΔZ and VvSRΔZSR cDNA with specific primers for the FGAM synthase gene. Upper and lower panels are a agarose gel and intensity pattern of the PCR products, respectively.

Discussion

Quorum sensing has been defined as a global regulatory system in various bacterial species. In V. vulnificus, luxS and smcR are thought to participate in the quorum-sensing system. The behaviour of luxS- or smcR-deficient V. vulnificus strains have been well characterized by recent studies (McDougald et al., 2001; Shao & Hor, 2001; Kim et al, 2003; Shin et al, 2005). However, little is known regarding the precise proteins involved in the events regulated by quorum sensing in V. vulnificus.

Based on the knowledge gained by recent studies, differently regulated regulons in the luxS- or smcR-mutants of V. vulnificus were investigated by 2D-PAGE and DDRT-PCR. From the analysis, 14 proteins were identified as quorum sensing-regulated proteins. Employment of 2D-PAGE revealed that quorum-sensing genes of V. vulnificus could relate to the expression of both extracellular and cellular proteins. The results from this study were similar to previous findings showing a number of apparently unrelated functions, including the production of virulence factors and interaction with its host luxS- or smcR- regulated genes (McDougald et al., 2001; Shao & Hor, 2001; Kim et al, 2003; Shin et al, 2005).

Of the identified proteins, V. vulnificus Zn-dependent protease (VVP) is known as an important virulence determinant for skin lesions as a result of the induction of permeability-enhancing and hemorrhagic reactions in vivo (Miyoshi & Shinoda, 1997). Down-regulation of the V. vulnificus protease by luxS or smcR agreed with previous reports on the decrease in protease activity (McDougald et al., 2001; Shao & Hor, 2001; Kim et al, 2003).

Some of the proteins needed to adapt to starvation, 6-phosphofructokinase (PFK), periplasmic ABC-type Fe3+ transport system, and deoxyribose-phosphate aldolase, were also repressed by the mutation of luxS and smcR. Essential roles of PFK in glucose metabolism have been shown for Vibrio alginolyticus (Blatch et al., 1990) and Bacillus sphaericus (Alice et al., 2002). The ABC-type Fe3+ transport system is an essential factor in pathogenicity under oxidative stress and iron limitation by iron acquisition via siderophore in many pathogenic bacteria (Brown et al., 2002; Qian et al., 2002; Braun, 2003; Nachin et al., 2003). Deoxyribose-phosphate aldolase is also known as a key enzyme in the deoxynucleoside catabolism, allowing the utilization of deoxyribose 5-P through glycolysis and leading to the utilization of the pentose moiety as carbon and energy source (Sgarrella et al., 1997). This information suggests that quorum sensing in V. vulnificus could be associated with the up-regulation of enzymes needed to obtain energy sources such as carbon and iron more efficiently and completely from a limited environment. Furthermore, our previous study supported the role of quorum sensing in bacterial survival in the host by enhancing adjustment in vivo (Shin et al., 2005).

Lipopolysaccharide (LPS) is known to be a representative component in septic shock as it induces a massive production of proinflammatory mediators such as cytokines and NO. Phosphomannomutase (PMM), which was identified in this study, was required for the biosynthesis of bacterial exopolysaccharides, namely alginate and LPS (Regni et al., 2002). Furthermore, the requirement of the enzyme for virulence, synthesis of O-antigen polysaccharide and resistance to serum has been shown using a pmm-deficient mutant of V. furnissii (Kim et al., 2003). The decrease in the expression of the enzyme by mutation agrees with our previous studies that show a decrease in the production of proinflammatory cytokines and NO and in virulence in mice (Shin et al., 2004, 2005).

Decreases in the expression of orotidine-5′-phsophate decarboxylase, an essential enzyme for uracil biosynthesis (Houk et al., 2001), and uridylate kinase, which may be related to cell division (Landais et al., 1999), in the luxS/smcR mutant also indicate an association of quorum sensing with cell proliferation.

Three cellular proteins that were induced as a result of luxS and smcR mutation, namely phage shock protein (Psp) A, a chemotaxis signal transduction protein, and an uncharacterized low-complexity protein, were identified in this study. Phage shock protein (Psp) was strongly induced in Escherichia coli by exposure to adverse conditions such as environmental stress (Weiner & Model, 1994). Most motile bacteria are capable of directing their movement in response to chemical gradients, a behaviour known as chemotaxis. The movement is usually related to nutritional stress. The proteins involved in this system were previously shown to be activated 3.4- to 6.3-fold in the luxO mutant in V. cholerae (Zhu et al., 2002), which is consistent with our results. It implies that external stress could be compensated by luxS and smcR mutation via activation of the related genes.

Extracellular proteins were also regulated by quorum sensing. Three proteins showing a difference in expression were identified by peptide-mass fingerprinting from 14 candidates found by 2D-PAGE, even though it was difficult to identify the spots owing to the low density of the activated or inhibited proteins visualized. Unfortunately, the functions of the proteins have not yet been revealed.

The proteins identified from this study were not, however, closely matched with those from other studies (Kan et al., 2004; Xu et al., 2005; Lee et al., 2006). This phenomenon might be a result of the expression of these proteins in different environmental conditions.

Two proteins, phosphoribosylformyl-glycinamidine (FGAM) synthase and ATP-dependent protease (HslVU), were also identified using DDRT-PCR as regulons controlled by quorum sensing at the transcriptional level. FGAM synthase and HslVU protease are the enzymes involved in the de novo synthesis of purines, and the degradation of an inhibitor of cell division, SulA, respectively (Gu et al., 1992; Seong et al., 1999, 2000). These results also support the idea that quorum sensing could regulate the multiplication of V. vulnificus.

In summary, a central role for LuxS and SmcR in the global regulation of gene expression in V. vulnificus has been demonstrated in this study. luxS and smcR might be capable of controlling the positive expression of genes related to virulence, adaptation in stressful environments, LPS biosynthesis, and cell division. In addition, other genes are correlatively activated to compensate for them when these quorum-sensing genes are damaged.

Acknowledgements

This study was supported by KOSF, Brain Korea 21, KRF (KRF-2006-005-J02901) and the Research Institute for Veterinary Science, Seoul National University, Korea.

Ancillary