Transcriptome Analysis of Genes Controlled by luxS /Autoinducer-2 in Salmonella enterica Serovar Typhimurium

Abstract

The enteric pathogen Salmonella enterica serovar Typhimurium uses autoinducer-2 (AI-2) as a signaling molecule. AI-2 requires the luxS gene for its synthesis. The regulation of global gene expression in Salmonella Typhimurium by luxS/AI-2 is currently not known; therefore, the focus of this study was to elucidate the global gene expression patterns in Salmonella Typhimurium as regulated by luxS/AI-2. The genes controlled by luxS/AI-2 were identified using microarrays with RNA samples from wild-type (WT) Salmonella Typhimurium and its isogenic ΔluxS mutant, in two growth conditions (presence and absence of glucose) at mid-log and early stationary phases. The results indicate that luxS/AI-2 has very different effects in Salmonella Typhimurium depending on the stage of cell growth and the levels of glucose. Genes with p ≤ 0.05 were considered to be significantly expressed differentially between WT and ΔluxS mutant. In the mid-log phase of growth, AI-2 activity was higher (1500-fold) in the presence of glucose than in its absence (450-fold). There was differential gene expression of 13 genes between the WT and its isogenic ΔluxS mutant in the presence of glucose and 547 genes in its absence. In early stationary phase, AI-2 activity was higher (650-fold) in the presence of glucose than in its absence (1.5-fold). In the presence of glucose, 16 genes were differentially expressed, and in its absence, 60 genes were differentially expressed. Our microarray study indicates that both luxS and AI-2 could play a vital role in several cellular processes including metabolism, biofilm formation, transcription, translation, transport, and binding proteins, signal transduction, and regulatory functions in addition to previously identified functions. Phenotypic analysis of ΔluxS mutant confirmed the microarray results and revealed that luxS did not influence growth but played a role in the biofilm formation and motility.

Introduction

Q uorum sensing, a form of cell-to-cell communication in bacteria, involves signaling molecules called autoinducers (AIs) through which bacteria respond to their environment based on their cell density (Pillai and Jesudhasan, 2006). Among the various quorum sensing systems (Whiteley et al., 1999; Winzer and Williams, 2001; Sperandio et al., 2003; Xavier and Bassler, 2003; Miller et al., 2004) used by bacteria, the AI-2 system is used by both Gram-positive and Gram-negative bacteria and requires the luxS gene for the synthesis of the signaling molecule. AI-2, a derivative of 4,5-dihydroxy-2,3-pentanedione, has been found to exist in two interconvertible structures (Xavier and Bassler, 2005)—a furanosyl borate diester (2S,4S)-2-methyl-2,3,4-tetrahydroxytetrahydrofuran (S-THMF) in Vibrio harveyi (Chen et al., 2002; Miller et al., 2004) and a nonboronated form (2R,4S)-2-methyl-2,3,4-tetrahydroxytetrahydrofuran (R-THMF) in Salmonella Typhimurium (Miller et al., 2004), a foodborne pathogen. Salmonella Typhimurium is known to detect and produce the cell signaling molecule AI-2, which is produced from S-adenosylmethionine by a three-enzymatic step reaction, of which the last step is catalyzed by luxS. AI-2 is known to regulate several key functions such as motility, biofilm formation, virulence, antibiotic production, and cell division in several pathogenic bacteria (Sperandio et al., 1999, 2001; Fong et al., 2001, 2003; Lyon et al., 2001; Miller et al., 2002; McNab et al., 2003).

LuxS has been identified in more than 55 species of bacteria including Gram-positive and Gram-negative bacteria (Surette and Bassler, 1999; Xavier and Bassler, 2003). Studies have shown that in Salmonella Typhimurium, luxS is responsible for the regulation of the lsr operon that helps in the internalization of AI-2. This has been hypothesized to be a mechanism to control AI-2 levels in the vicinity of a cell or to prevent AI-2 signaling by other bacterial species in its environment (Taga et al., 2003). Recently it has been reported that in Salmonella Typhimurium, luxS is also responsible for virulence gene expression (Choi et al., 2007) and polarization of the flagellar phase variation (Karavolos et al., 2008). We have previously shown using microarray studies that AI-2 influenced virulence in Salmonella Typhimurium (Widmer et al., 2007). We have also shown using proteomic analysis of Salmonella Typhimurium that AI-2 is responsible for a variety of cellular processes (Soni et al., 2008).

In this study, global gene regulation by luxS/AI-2 in Salmonella Typhimurium was identified by comparing gene expression of Salmonella Typhimurium strain 87-26254 wild type (WT) and its isogenic ΔluxS mutant in two growth conditions at both mid-log and early stationary phase using microarrays. Genes regulated by AI-2 were identified by comparing gene expression in a ΔluxS mutant of Salmonella Typhimurium and ΔluxS mutant supplemented with cell-free supernatant of WT Salmonella Typhimurium. Phenotypic studies were conducted to examine the effects of luxS mutation on cell growth, motility, and biofilm formation.

Materials and Methods

Bacterial strains and growth conditions

The bacterial strain used in this study was Salmonella Typhimurium (accession no. 87-26254), a poultry isolate obtained from the National Veterinary Service Laboratory, Ames, IA. A ΔluxS mutant of this strain, designated PJ002 (luxS::cat), was subsequently generated in our laboratory (Widmer et al., 2007). Escherichia coli no. 5, an environmental isolate, was used as a positive control for the AI-2 bioassays because of its ability to produce elevated levels of the AI-2 (Qin et al., 2004). Strains of Salmonella Typhimurium (87-26254 and PJ002) and E. coli no. 5 were grown at 37°C in Luria–Bertani (LB) broth and LB broth plus 0.5% glucose. V. harveyi strain BB170 (luxN::Tn5 sensor 1⁻, sensor 2⁺), a reporter used to determine AI-2 activity, was grown at 30°C with aeration in AI bioassay medium (Bassler et al., 1993).

Preparation of cell-free supernatant

Overnight cultures of Salmonella Typhimurium (87-26254 and PJ002) grown in LB broth at 37°C were inoculated in 100 mL LB broth and in LB broth with 0.5% glucose (1:100) and grown at 37°C with shaking (250 rpm). Cell-free supernatants (CFS) were collected at different time points by removing the cells from the growth medium by centrifugation at 10,000 g for 2 min. The supernatants were then passed through 0.22 μm syringe filters (Corning^®, Corning, NY) and stored at −20°C. CFS of E. coli no. 5 grown in fresh LB broth with 0.5% glucose was prepared from 3.5-h culture using the same procedure that was used for Salmonella Typhimurium.

Bioluminescence assay

Salmonella Typhimurium CFS obtained from both WT and mutant grown in LB broth and LB broth with 0.5% glucose were tested for the presence of AI-2 activity using the V. harveyi reporter strain BB170, which responds only to AI-2 (Surette and Bassler, 1998). Luminescence assays were performed as outlined elsewhere (Surette and Bassler, 1998). Luminescence was measured by quantifying light production of V. harveyi with a Wallac Victor 1420 multilabel counter (Perkin Elmer, Boston, MA). The ratios of the luminescence of the test samples (reporter strain with Salmonella Typhimurium CFS) to the control (reporter strain without CFS) were reported as relative AI-2 activity (fold induction).

Growth curves

Overnight cultures of WT and mutant grown at 37°C were inoculated (1:100) in fresh LB medium and fresh LB medium supplemented with 0.5% glucose. Two hundred microliters of the bacterial suspensions was added to each well of 96-well microtiter plate (Corning) and maintained for 8 h at 37°C in a Tecan plate reader (Tecan U.S. Research, Triangle Park, NC). Optical density at 600 nm of each sample was measured every 10 min. Before each measurement, the plate was shaken at low speed for 7 min to prevent settling of the culture. Sixteen wells were used for each treatment, and the experiment was repeated twice. Growth curves were generated in Excel (Microsoft, Redmond, WA) by plotting the average OD against time.

Swimming and swarming assay

The swimming and swarming motility assays were performed in LB broth (Difco) containing 0.3% and 0.5% agar, respectively, both in the presence and absence of 0.5% glucose. The overnight cultures of WT and mutant strains grown in LB broth were inoculated with sterile toothpicks in the center of swimming and swarming assay plates and incubated for 7 h at 37°C. The diameters of the motility halos were measured. At least five replicate plates were used for each condition, and statistical significance was calculated using Student's t-test.

Biofilm assay

The biofilm assay was performed in sterile round-bottom 96-well polystyrene plates (Nalge Nunc International, Rochester, NY) according to O'Toole and Kolter (1998) with some modifications. Overnight cultures of WT and mutant strains were diluted 1:100 in LB medium and LB medium supplemented with 0.5% glucose, and 100 μL of the aliquots was added to eight wells. The plates were incubated at 37°C for 36 h without shaking. After incubation, the suspension cultures were removed, and the plates were washed with distilled water. The biofilms were stained with 125 μL of 0.1% crystal violet (Fisher, Hanover Park, IL) per well for 15 min. The excess dye was removed by washing with distilled water. Dye associated with attached biofilms was dissolved in 200 μL of 95% ethanol, and 125 μL was transferred to an optically clear flat-bottom 96-well plate (Corning), and OD₅₉₀ was measured.

Total RNA isolation

Overnight cultures of Salmonella Typhimurium WT and mutant strains were inoculated in LB broth and LB broth with 0.5% glucose at 37°C. Cells were grown in triplicate and harvested at mid-log and early stationary phase (3 and 7 h, respectively) for RNA extraction. Total RNA was isolated from the cultures using an RNeasy mini kit (Qiagen, Valencia, CA) according to the manufacturer's protocol. RNAprotect bacteria reagent (Qiagen) was added to the cultures to stabilize the RNA before isolation. The RNase-free DNase set (Qiagen) was used for on-column DNase digestion to remove residual genomic DNA. The quantity and quality of RNA were examined using a ND-1000 spectrophotometer (NanoDrop^® Technologies, Wilmington, DE) and bioanalyser 2100 (Agilent Technologies, Palo Alto, CA), respectively.

cDNA synthesis and labeling

cDNA synthesis and purification were performed using “microbial RNA aminoallyl labeling” (ftp://ftp.jcvi.org/pub/data/PFGRC/pdf_files/protocols/M007.pdf), a standard operating procedure of The J. Craig Venter Institute (JCVI, formerly The Institute for Genomic Research). Ten micrograms of total RNA was used to synthesize cDNA using a random primer for reverse transcription (Invitrogen, Carlsbad, CA). Purified cDNAs from the WT and mutant were each labeled with Cy-3 mono-Reactive Dye and Cy-5 mono-Reactive Dye (GE Health Care, Piscataway, NJ) and were processed using a dye-swapping design. A total of 12 microarray slides were used for each time point. The 12 slides were divided as follows: four slides (technical replicates) were used for each of the three biological replicates. Each set of four slides had a different DNA sample. Among the four slides that used the same DNA, two slides had the same dye, and the dye was swapped for the other two. The labeling mixtures were purified using a QIAquick PCR purification kit (Qiagen).

Microarray hybridization, washing, and scanning

Salmonella Typhimurium LT2 genome microarrays (Versions 4) developed by JCVI and provided by the Pathogen Functional Genome Resource Center of the National Institutes of Health were used in this study. These arrays are 70 mer-oligo arrays consisting of 4504 open reading frames each in five replicate spots. Equal amounts of labeled cDNA from the treatment and control were used to hybridize the arrays. The labeled cDNA was hybridized following “hybridization of labeled cDNA probes” (ftp://ftp.jcvi.org/pub/data/PFGRC/pdf_files/protocols/M008.pdf), a standard operating procedure used at JCVI. The hybridization was carried out overnight at 42°C in a water bath using Corning hybridization chamber. After hybridization, the slides were washed and scanned using a GenePix 4100A scanner (Molecular Devices Corporation, Sunnyvale, CA) at 532 nm (Cy3 channel) and 635 nm (Cy5 channel), and the images were stored for further analysis.

Microarray data analysis

The data from three individual experiments (four slides per experiment including dye swap) were initially filtered for spot quality (signal uniformity, signal to background ratio, threshold intensity) using GenePix Pro 5.0 (Axon). Visually flagged spots as well as spots with a median signal value less than the sum of the local background median plus three standard deviations were omitted from analysis (Hegde et al., 2000). Array data were normalized, and their statistical significance was evaluated using Acuity 4.0 (Molecular Devices). To identify genes differentially expressed between different treatment groups, a Student's t-test was performed and the false discovery rate was calculated using the Benjamini–Hochberg method (Benjamini and Hochberg, 1995) in Acuity. Genes with p ≤ 0.05 were considered to be significantly expressed differentially between WT and ΔluxS mutant. The microarray data analysis procedures used in this study were fully minimum information about a microarray experiment compliant. The complete dataset and raw data are available at NCBI Gene Expression Omnibus database (accession no. GSE7558). Although a twofold cutoff is generally used for the analysis of microarray data, a less than twofold change in relative transcript levels can be biologically significant (Hughes et al., 2000; Ichikawa et al., 2000). Hence, genes that are significantly up- or downregulated with a lesser cutoff induction ratio have been included in Tables 2 –5.

Supportive microarray studies

Genes identified in the microarray studies could be regulated either by luxS or AI-2. To identify the genes influenced by AI-2, another microarray study was conducted using the ΔluxS mutant (PJ002) with CFS from WT (wCFS) and mutant CFS (mCFS). The AI-2 activity of both wCFS and mCFS was measured using the AI-2 bioassay. Overnight culture of PJ002 was inoculated (1:100) in fresh LB broth with 10% of wCFS (treatment) and LB broth with 10% of mCFS (control) and grown at 37°C for RNA extraction. Cells were grown in triplicate and harvested after 3 h of growth. Total RNA was isolated from the cultures using an RNeasy mini kit (Qiagen) according to the manufacturer's protocol. Synthesis of cDNA, labeling, hybridizations, scanning of slides, and data analysis were performed using the same procedures as described above. Genes with p ≤ 0.05 were considered to be significantly expressed differentially between wCFS and mCFS.

Quantitative real-time reverse transcriptase–polymerase chain reaction

Quantitative real-time polymerase chain reaction (PCR) was used to validate microarray results. Relative transcript levels of selected genes were measured using SYBR^® Green PCR kit (Applied Biosystems) according to the manufacturer's recommendations. Aliquots from the same RNA samples (three biological replicates) used for microarray experiments were used for cDNA synthesis and subsequent quantitative real-time reverse transcriptase PCR (QRT-PCR). For each of the genes tested, primers were designed using Primer Express software (Applied Biosystems) to amplify products from 75 to 150 bp. 16S rRNA gene fragment was used as an internal control to normalize expression values. The reactions were performed on an ABI Prism 7900HT Sequence Detection System (Applied Biosystems) with the following cycle parameters: one cycle of 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min. The difference (fold change) in the initial concentration of each transcript (normalized to 16S rRNA) with respect to the WT was calculated according to the comparative C_T method according to the 2^−ΔΔC _T calculation (Livak and Schmittgen, 2001). A melting curve was completed following each reaction to confirm that no template-independent amplification was present.

Results

Bioluminescence assay

The bioluminescence bioassay used to check the AI-2 activity showed higher activity in WT in the presence of glucose (1500-fold) (Fig. 1A, B) in the mid-log phase (3 h) than in its absence (450-fold). In the early stationary phase (7 h), AI-2 activity was still high in the presence of glucose (650-fold) but was comparatively very low in the absence of glucose (1.5-fold). The Salmonella Typhimurium ΔluxS mutant PJ002 did not produce luminescence in both the presence and absence of glucose (Fig. 1A, B), confirming that the PJ002 strain does not produce AI-2 molecules in the absence of the luxS gene.

FIG. 1.

Growth and autoinducer-2 (AI-2) activity of Salmonella Typhimurium wild type (WT) and its isogenic ΔluxS mutant (PJ002). (A) Luria–Bertani (LB) medium supplemented with glucose 0.5%; (B) LB medium without glucose. The assays were carried out in triplicate. Bars indicate AI-2 activity, and lines indicate growth. Each bar represents mean ± standard error of the mean of all the measurements. Arrows point to the time points at which samples were collected for RNA extraction.

Effect of luxS gene deletion on growth, motility, and biofilm formation in Salmonella Typhimurium

Presence or absence of glucose did not affect the growth rates of both the ΔluxS mutant and its isogenic parent. Swimming and swarming motility was significantly (p < 0.05) reduced in the ΔluxS mutant strain compared with the WT (Fig. 2). Moreover, the mutant strain also formed less biofilm than the WT significantly (p < 0.0001) both in LB medium and LB medium supplemented with 0.5% glucose (Fig. 3). Significant inhibition in motility (p < 0.05) and biofilm formation (p < 0.0001) was observed in the WT in the presence of glucose (Figs. 2 and 3) due to catabolite repression. Microarray analysis revealed that in the absence of glucose, 27 of the 35 motility-related genes (Table 2) and 9 of the 13 biofilm formation–related genes (Table 3) were downregulated in the mid-log phase in the absence of both luxS and AI-2. Eleven motility-related genes and four biofilm formation–related genes had the same direction of regulation (downregulation) in both the microarray studies and the confirmatory microarray studies, indicating that these genes could be regulated by AI-2 (Tables 2 and 3).

FIG. 2.

Effect of luxS deletion on Salmonella Typhimurium motility. Swimming and swarming zones were measured 7 h after inoculating 0.3% and 0.6% LB agar plates, respectively, in both absence and presence of glucose (0.5%). Zone diameters (mean ± standard deviations; n = 5) are listed for each strain. The zone diameters for each strain in swimming motility (A) WT (30.5 ± 6.5), (B) ΔluxS (15.0 ± 1.6), (C) WT (13.2 ± 2.3), and (D) ΔluxS (10.7 ± 1.6); swarming motility (A) WT (22.7 ± 3.5), (B) ΔluxS (16.2 ± 1.5), (C) WT (0), and (D) ΔluxS (0) are shown. Mutant was deficient in motility both in the presence and absence of glucose.

FIG. 3.

Effects of luxS mutation and glucose on biofilm (mean OD ± SE; p < 0.0001, n = 3) formation in Salmonella Typhimurium. Biofilm assays were performed in 96-well plates, and eight wells were assigned for each strain and condition. Biofilms of WT and ΔluxS mutant were determined using crystal violet binding assay after 36 h of growth at 37°C in the presence and absence of glucose. Each bar represents mean ± standard error of the mean of all the measurements. Presence of glucose inhibited biofilm formation in WT. Mutant was deficient in biofilm formation both in the presence and absence of glucose.

Microarray identification of luxS-regulated genes

Differential expression of genes in LB medium without glucose was about 12% of the entire transcriptome (547 genes) at mid-log phase. This was substantially higher than differential regulation of genes (13 genes) in LB medium with glucose (Table 1). At early stationary phase, the gene modulation pattern was similar to that at mid-log phase with a higher number of genes differentially expressed in LB medium without glucose (60 genes) (data not shown) than in LB medium with glucose (16 genes) (Table 1). Of the 547 genes that were differentially regulated at mid-log phase in the absence of glucose, 217 were repressed and 330 were promoted (selected genes shown in Tables 2 –6). After statistical analysis of the microarray data, the differentially expressed genes in the WT compared with the ΔluxS mutant were broadly categorized into functional groups based on the clusters of orthologous groups (Fig. 4). Genes in 21 functional categories were regulated by luxS/AI-2: unclassified genes (25 up and 8 down), enzymes of unknown function (17 up and 13 down), general function predicted (29 up and 23 down), secondary metabolite biosynthesis, transport, catabolism (2 up and 0 down), inorganic ion transport and metabolism (20 up and 15 down), lipid transport and metabolism (3 up and 9 down), coenzyme transport and metabolism (10 up and 8 down), nucleotide transport and metabolism (6 up and 11 down), amino acid transport and metabolism (28 up and 15 down), carbohydrate transport and metabolism (31 up and 7 down), energy production and conversion (35 up and 4 down), posttranslational modification, protein turnover, chaperones (13 up and 5 down), intracellular trafficking and secretion (3 up and 6 down), cell motility (9 up and 1 down), cell wall/membrane biogenesis (8 up and 12 down), signal transduction mechanisms (25 up and 2 down), defense mechanisms (3 up and 3 down), cell cycle control, mitosis and meiosis (0 up and 2 down), replication, recombination, and repair (3 up and 17 down), transcription (38 up and 14 down), and translation (6 up and 37 down) (Fig. 4).

FIG. 4.

Categories of genes regulated by luxS/AI-2. luxS/AI-2-regulated genes are grouped into functional categories based on the Salmonella Typhimurium genome annotation (http://cmr.jcvi.org/cgi-bin/CMR/shared/MakeFrontPages.cgi?page=geneattribute). The histograms represent the number of genes in each category promoted or repressed by luxS mutation in the mid-log phase in LB without glucose.

Table 1.

List of Genes Differentially Regulated Between Wild Type (W) and Δ luxS Mutant (M) by luxS/Autoinducer-2 at Mid-Log and Early Stationary Phase in Luria–Bertani Broth with 0.5% Glucose in Salmonella Typhimurium

Genes	Putative identification of JCVI database	Fold change mid-log W/M	Fold change early stationary W/M
luxS	AI-2 production protein LuxS	5.1	5.4
fljB	Flagellin	2.7	2.5
intN	Integrase	2.2	–
fhuF	Ferric hydrozamate transport	2.0	–
pgsA	Phosphatidylglycerophosphate synthetase (CDP-1,2-diacyl-sn-glycero-3-phosphate phosphatidyl transferase)	2.0	–
pduO	Propanediol utilization: B12 related	−2.0	–
STM3772	Putative phosphotransferase system enzyme IIA	2.5	–
STM1669	Homology to invasin C of Yersinia intimin	−2.0	–
STM2617	Gifsy-1 prophage: similar to antiterminator Q	−2.1	–
STM1255	Putative ABC transporter periplasmic-binding protein	−2.2	–
dmsA	Anaerobic dimethyl sulfoxide reductase, subunit A	−2.2	–
glgB2	Maltooligosyl trehalose trehalohydrolase	−2.4	–
hscC	Putative heatshock protein, homolog of hsp70 in	−3.2	–
STM0293	Putative cytoplasmic protein	–	2.4
yfiD	Putative formate acetyltransferase	–	2.2
ynfK	Putative dethiobiotin synthase	–	2.1
focA	Putative FNT family, formate transporter (formate channel)	–	2.0
yccD	Putative cytoplasmic protein	–	2.0
STM3142	Putative ferrichrome-binding periplasmic protein	–	2.0
malM	Periplasmic protein of mal regulon	–	−2.0
STM3191	Putative arylsulfate sulfotransferase	–	−2.1
glxR	Tartronic semialdehyde reductase	–	−2.2
ynfM	Putative MFS familty transport protein	–	−2.4
STM3600	Putative sugar kinases, ribokinase family	–	−3.3
STM2358	Putative cytoplasmic protein	–	−2.2
STM3770	Putative phosphotransferase system enzyme IIC	–	−2.2
STM3257	Putative tagatose 6-phosphate kinase 1	–	2.7

Table 2.

Salmonella Typhimurium Genes Involved in Motility That Are Regulated by luxS/Autoinducer-2 at Mid-Log Phase in the Absence of Glucose

Genes	Putative identification of JCVI database	Fold change (W/M)	Fold change (wCFS/mCFS)
Chemotaxis proteins
Two-component system proteins
cheA	Chemotaxis protein CheA	−1.3	1.0
cheB	Chemotaxis-specific methylesterase	1.3	−1.0
STM2314	Putative chemotaxis signal transduction protein	3.5	1.1
MCPs
tcp	Methyl-accepting transmembrane citrate/phenol chemoreceptor	2.0	−1.0
STM3216	Putative methyl-accepting chemotaxis protein	1.8	1.1
trg	Methyl-accepting chemotaxis protein III	2.1	1.1
aer	Aerotaxis sensor receptor	2.5	−1.1
Flagellar assembly proteins
Type III secretion
fliH	Flagellar assembly protein H	−2.1	−1.0
fliO	Flagellar biosynthesis protein FliO	−1.8	−1.0
Motor/switch
motA	Flagellar motor protein MotA	2.0	−1.0
motB	Flagellar motor protein MotB	2.1	−1.0
C-ring
fliG	Flagellar motor switch protein G	−2.3	−1.0
fliN	Flagellar motor switch protein FliN	1.2	−1.0
M, S, P, and L rings
flgA	Flagellar basal body P-ring biosynthesis protein FlgA	1.5	−1.0
Rod, hook, and filament
flgB	Flagellar basal body rod protein FlgB	−2.4	−1.1
flgE	Flagellar hook protein FlgE	1.3	1.1
flgJ	Peptidoglycan hydrolase	−2.1	−1.0
flgK	Flagellar hook-associated protein FlgK	2.2	1.1
flgL	Flagellar hook-associated protein FlgL	1.4	1.0
fliD	Flagellar capping protein	2.0	1.2
fliK	Flagellar hook-length control protein	1.2	1.0
fljB	Flagellin	−1.3	−1.0
Regulation
flhC*	Regulator of flagellar biosynthesis, acts on class	1.4	−1.0
flhD*	Regulator of flagellar biosynthesis, acts on class	1.3	−1.0
flgM	Anti-sigma28 factor FlgM	2.7	−1.1
Chaperones
fliJ	Flagellar biosynthesis chaperone	1.5	1.1
Pilus system
Fimbrial proteins
fimA	Fimbrin	2.0	−1.0
fimD	Outer membrane usher protein precursor	−2.1	1.0
Putative motility
lpfA	Long polar fimbrial protein A precursor	1.4	−1.1
stdC	Putative fimbrial chaparone	1.4	−1.0
cheM	Methyl-accepting chemotaxis protein II	1.6	1.0
stiA	Putative fimbrial subunit	1.2	−1.0
lpfE	Long polar fimbrial minor protein	1.2	−2.2
yifL	Putative outer membrane lipoprotein	1.2	−1.0
stbA	Putative fimbrial major subunit	1.2	1.1

Differentially regulated genes between wild type (w) and ΔluxS mutant (M) and between wild-type cell-free supernatants and ΔluxS mutant cell-free supernatants are listed. Genes that have the same direction of regulation in both W/M and wCFS/mCFS are regulated by AI-2 and are shown in bold.

wCFS, wild-type cell-free supernatants; mCFS, mutant CFS.

Table 3.

Salmonella Typhimurium Genes Involved in Biofilm Formation That Are Regulated by luxS/Autoinducer-2 at Mid-Log Phase in the Absence of Glucose

Genes	Putative identification of JCVI database	Fold change (W/M)	Fold change (wCFS/mCFS)
csgA	Curlin major subunit, coiled surface structures cryptic	1.1	−1.1
csgC	Putative curli production protein	1.2	1.0
bcsA	Glycosyltransferase, probably involved in cell wall biogenesis	−1.8	1.0
bcsC	Endo-1,4-D-glucanase	−1.8	−1.0
yhjN	Putative cellulose synthase	−1.6	−1.0
fimA	Major type 1 subunit fimbrin (pilin)	2.0	−1.0
fimD	Outer membrane usher protein	−1.3	1.1
fimZ	Fimbrial protein Z, putative transcriptional regulator (LuxR/UhpA)	1.5	−1.1
fimW	Putative fimbrial protein	1.4	1.0
fimY	Putative regulatory protein	1.3	1.1
lpfA	Long polar fimbria	1.4	−1.1
lpfE	Long polar fimbrial minor protein	1.2	−2.2
yojN	Putative sensor/kinase in regulatory system	1.6	1.1

Differentially regulated genes between wild type (W) and ΔluxS mutant (M) and between wild-type cell-free supernatants and ΔluxS mutant cell-free supernatants are listed. Genes that have the same direction of regulation in both W/M and wCFS/mCFS are regulated by AI-2 and are shown in bold.

Table 4.

Salmonella Typhimurium Genes Involved in Virulence Regulated by luxS/Autoinducer-2 at Mid-Log Phase in the Absence of Glucose

Genes	Putative identification of JCVI database	Fold change W/M	Fold change wCFS/mCFS
hha	Hemolysin expression modulating protein (involved in environmental regulation)	1.9	1.0
pagP	PhoPQ-activated gene	−1.4	1.3
mviM	Putative virulence factor	2.2	1.3
cobB	Putative nicotinate-nucleotide dimethylbenzimidazolephosphoribosltransferase, homolog of virulence factor	1.6	1.2
pagC	PhoP regulated: reduced macrophage survival	1.5	−1.8
hnr	Response regulator in protein turnover: mouse virulence	−1.4	−1.3
sirC	Regulation of invasion genes	−2.3	−1.2
virK	Virulence gene homologous sequence to virK in	−2.5	1.4
avrA	Putative inner membrane protein	1.4	−3.1
STM2868	Putative cytoplasmic protein	−1.8	−2.1
orgA	Putative flagellar biosynthesis/type III secretory pathway protein	−1.7	−2.7
prgI	Cell invasion protein cytoplasmic	−1.4	−9.6
hilD	Regulatory helix-turn-helix proteins, araC family	1.2	−3.3
hilA	Invasion genes transcription activator	−2.0	−6.1
sptP	Protein tyrosine phosphate	−1.3	−5.9
sicP	Chaparone, related to virulence	−1.4	−5.0
STM2880	Putative cytoplasmic protein	−1.4	−2.9
sipB	Cell invasion protein	1.3	−2.8
spaR	Surface presentation of antigen secretory proteins	−1.5	−9.2
invC	Surface presentation of antigen secretory proteins	−1.5	−11.2
invB	Surface presentation of antigen secretory proteins	−1.4	−7.5
invA	Invasion protein	−1.6	−6.8
invE	Invasion protein	−1.4	−10.5
ygdP	Putative invasion protein NTP pyrophosphohydrolase	1.4	1.0
fis	Site-specific DNA inversion stimulation factor	−4.6	−1.0
hfq	Host factor I for bacteriophage Q beta	1.4	1.2

Table 5.

Salmonella Typhimurium Genes in lsr Operon That Are Regulated by luxS/Autoinducer-2 at Mid-Log Phase and Early Stationary Phases in Both the Absence and Presence of 0.5% Glucose

		Fold change W/M
		LB broth		LB broth + glucose
Genes	Putative identification of JCVI database	Mid-log	Early stationary	Mid-log	Early stationary	Fold change wCFS/mCFS
lsrA (ego)	Putative ABC-type sugar, aldose transport system, ATPase	1.7	2.0	1.4	−1.9	−1.1^a
lsrB (yneA)	Putative ABC superfamily (peri_perm), sugar transport protein	2.2	2.7	–	1.1	−1.1^a
lsrC (ydeY)	Putative ABC superfamily (membrane), sugar transport protein	2	2.3	1.1^a	–	1.0^a
lsrD (ydeZ)	Putative ABC superfamily (membrane), sugar transport protein	1.9	3.4	1.3	1.1	−1.1^a
lsrE (STM4080)	Putative ribulose-5-phosphate 3-epimerase	1.3	2.0	–	–	1.1
lsrF (yneB)	Putative fructose-1,6-bisphosphate aldolase	1.5	2.1	−1.3^a	−1.2	1.1
lsrG (yneZ)	Putative inner membrane protein	1.7	1.6	1.1	–	−1.1
lsrK (ydeV)	Putative sugar kinase	1.4	1.7	1.3	1.2	−1.0
lsrR (ydeW)	Putative transcriptional repressor	1.3^a	1.3	1.2	–	−2.7

Differentially regulated genes between wild type (W) and ΔluxS mutant (M) and between wild-type cell-free supernatants and ΔluxS mutant cell-free supernatants are listed.

p > 0.05.

LB, Luria–Bertani.

Table 6.

Salmonella Typhimurium Genes in rbs Operon That Are Regulated by luxS/Autoinducer-2 at Mid-Log Phase and Early Stationary Phases in Both the Absence and Presence of 0.5% Glucose

		Fold change W/M
		LB broth		LB broth + glucose
Genes	Putative identification of JCVI database	Mid-log	Early stationary	Mid-log	Early stationary	Fold change wCFS/mCFS
rbsA	ABC superfamily (atp_bind), D-ribose high-affinity transport protein	1.1	1.1^a	–	1.1^a	−1.1^a
rbsB	ABC superfamily (peri_perm), D-ribose transport protein	2.0	–	–	–	1.1
rbsC	ABC superfamily (membrane), D-ribose high-affinity transport protein	−1.2^a	1.2^a	–	1.1^a	1.0
rbsD	D-ribose high-affinity transport system membrane-associated protein	1.3^a	–	–	1.1^a	1.2
rbsK	Ribokinase	2.2	–	–	1.1^a	1.2
rbsR	Transcriptional repressor for rbs operon (GalR/LacI family)	2.5	–	–	–	−1.1^a

Differentially regulated genes between wild type (W) and ΔluxS mutant (M) and between wild-type cell-free supernatants and ΔluxS mutant cell-free supernatants are listed.

p > 0.05.

Validation of microarray results using QRT-PCR

To verify the microarray results, QRT-PCR was performed on a selected number of genes identified as luxS/AI-2-regulated, and the results are shown in Fig. 5. The magnitude of fold change was different in most of the genes when compared with that obtained using the microarray analysis. However, all of the genes used in this study had the same pattern of regulation (either up or down) as that of the microarray results. Multivariate analysis of the QRT-PCR results using JMP 6.0 (SAS Institute) showed a pairwise correlation coefficient of 0.933, to the microarray data (p = 0.0021).

FIG. 5.

Validation of microarray data of selected luxS/AI-2-regulated genes using real-time reverse transcriptase–polymerase chain reaction (PCR). The differences in expression of seven genes that are regulated by luxS/AI-2 in the mid-log phase in the absence of glucose were log₂ transformed and plotted against each other (microarray against reverse transcriptase–polymerase chain reaction).

Discussion

LuxS is involved in the production of AI-2, the cell–cell signaling molecule used by Salmonella Typhimurium. luxS/AI-2 is known to be involved in various cellular processes in Salmonella Typhimurium, but its influence on global gene expression has not been studied. This is the first study to report the impact of luxS mutation on the global gene regulation of Salmonella Typhimurium and to phenotypically characterize it based on growth, motility, and biofilm formation.

In this study, no difference in growth rates was observed between WT and ΔluxS mutant strains of Salmonella Typhimurium in LB with or without glucose supplementation (Fig. 1A, B). Similar observations were reported in E. coli K-12 (Wang et al., 2005). However, Sperandio et al. (2001) reported that the growth rate of a luxS mutant of enterohemorrhagic E. coli O157:H7 (EHEC) was faster than its WT. In Streptococcus pyogenes, media-dependent growth defect was observed in luxS mutants (Lyon et al., 2001). Results from our experiment suggest that luxS/AI-2 does not play a role in growth in Salmonella Typhimurium under the conditions tested.

WT Salmonella Typhimurium produced high levels of AI-2 in both log (1500-fold) and stationary (650-fold) phases in the presence of glucose (Fig. 1A). Earlier studies have reported that maximal AI-2 is produced between mid- and late log phases in the presence of glucose and that AI-2 activity is reduced with the onset of the stationary phase because of the depletion of glucose from the medium (Surette and Bassler, 1998, 1999). Contrary to what may be expected that higher levels of AI-2 result in a greater number of regulated genes, we only observed a small number of differentially expressed genes in Salmonella Typhimurium in the presence of glucose (Table 1). Of the total 4622 genes, only 13 were differentially expressed with 6 induced and 7 repressed. The minimal differential gene expression could be attributed to the abundance of glucose or catabolite repression (Wang et al., 2005; Xavier and Bassler, 2005), suggesting that Salmonella Typhimurium uses cell signaling molecule for survival in glucose-limiting conditions.

In the absence of glucose, there was a 450-fold induction of AI-2 activity in the mid-log phase (Fig. 1B), but a large number of genes were differentially expressed. Other studies have also reported that AI-2 production is reduced in the absence of glucose (Surette and Bassler, 1998; Taga et al., 2001; Wang et al., 2005; Xavier and Bassler, 2005). Also, microarray analysis of luxS-dependent gene regulation in E. coli K-12 showed that more genes were differentially expressed in the absence of glucose (Wang et al., 2005). We observed that 12% of the entire transcriptome was regulated with differential expression of genes in every set of clusters of orthologous groups at mid-log phase (Fig. 4). Comparison of gene expression patterns at mid-log and early stationary phases showed that 31 genes had a biphasic response, which means that they were differentially expressed (either expressed or repressed) in both mid-log and stationary phases (data not shown).

Earlier studies have suggested that luxS plays an important metabolic function in the recycling of S-adenosyl homocysteine (Winzer et al., 2002). In this study, we observed that among the genes that were regulated by luxS, genes encoding metabolism formed the largest group (Fig. 4). We also found that luxS not only plays a major role in metabolism but also regulates genes involved in other cellular processes in Salmonella Typhimurium (Fig. 4).

Flagellar biosynthesis in E. coli and Salmonella Typhimurium is regulated by more than 50 genes (Macnab, 1996), among which flhDC is the master regulator. In Salmonella Typhimurium, proteins MotA, MotB, FliG, FliM, and FliN have been identified as direct components of the rotary motor present in the cytoplasmic membrane and known to drive the bacterial flagellum (Macnab, 1996). In this study, among the genes that were differentially expressed (WT/ΔluxS), we observed downregulation of several key genes involved in motility including master regulator flhDC and components of the rotary motor motA, motB, fliN. Karavolos et al. (2008) reported that luxS affects flagellar phase variation in Salmonella Typhimurium. Studies in E. coli K-12 (Wang et al., 2005) found no difference in motility between WT and ΔluxS mutant. In this study, phenotypic studies showed that ΔluxS mutant was less motile both in the presence and absence of glucose. Supportive microarray studies showed that among the motility genes that were downregulated, some were regulated by AI-2 (Table 2). Our observations indicate that both luxS and AI-2 influence motility in Salmonella Typhimurium.

Biofilms are considered as a developmental process as they are formed in steps that require intercellular signaling. Microarray studies showed that among the genes involved in biofilm formation that were differentially expressed (WT/ΔluxS), most were downregulated in the absence of luxS/AI-2 (Table 3). Prouty et al. (2002) reported that quorum sensing was important for full biofilm formation. They also reported that flagellar genes were important for biofilm formation and that fimbriae genes played a negative role in biofilm formation (Prouty et al., 2002). In this study, microarrays comparing the differences in gene expression between ΔluxS mutant supplemented with WT CFS and ΔluxS mutant (wCFS/mCFS) showed similar direction of regulation in certain genes as in WT/ΔluxS mutant. Because AI-2 is present in both microarray studies, genes with a similar direction of regulation could be considered as influenced by AI-2. Phenotypic studies conducted to observe biofilm formation showed that, irrespective of the growth condition, biofilm formation was significantly less in the ΔluxS mutant (Fig. 3). Our microarray and phenotypic studies indicate that both luxS and AI-2 are involved in the biofilm process in Salmonella Typhimurium.

Cell–cell signaling and microbial pathogenesis are known to be closely linked (de Kievit and Iglewski, 2000), but their relationship in Salmonella Typhimurium is still unknown. Factor for inversion stimulation (fis) has been shown to be involved in DNA transcription, DNA replication at oriC, transposition, invasion of HEp-2 cells, and transcriptional activation and repression of several genes (Kelly et al., 2004; Hirsch and Elliott, 2005a, 2005b). In pathogenic strains of E. coli and Salmonella Typhimurium, fis has been reported to modulate virulence gene expression (Kelly et al., 2004). We observed that in the absence of glucose, fis showed a decreased WT/ΔluxS expression (−4.6-fold) at mid-log phase during which AI-2 is present (Table 4). In E. coli, it has been reported that hha, which encodes the regulator of the hemolysin operon and known to mediate environmental regulation of virulence factors in Pseudomonas aeruginosa (Whiteley et al., 1999), was comparatively induced 11.1-fold in the presence of AI-2 (DeLisa et al., 2001). Similarly, in this study we found that hha increased 1.9-fold in the presence of AI-2 at mid-log phase (Table 4). fis is a positive regulator of hilA, an activator of invasion genes, while hha is a negative regulator of hilA. HilA was found to be downregulated (twofold) at mid-log phase which could be due to the repression of fis and expression of hha (Table 4). This suggests that luxS/AI-2 regulates certain key virulence genes in Salmonella Typhimurium. None of the above mentioned genes (fis, hha, hilA) exhibited differential WT/ΔluxS expression at early stationary phase (data not shown). It must be noted that in the WT, AI-2 is present at mid-log phase (650-fold) and degraded or uptaken by cells at early stationary phase (1.5-fold) and hence AI-2 could be considered as the reason for the differential regulation of these genes at the mid-log phase. Further, microarray studies using ΔluxS mutant in the presence of CFS of WT Salmonella Typhimurium showed that key virulence genes were downregulated in the presence of AI-2. Higgins et al. (2007) have reported that the major AI-2 in Vibrio cholerae repressed the expression of virulence genes which is similar to the finding in this study that AI-2, the AI in Salmonella Typhimurium, represses the expression of its virulence genes.

In the WT, all the genes in the lsr operon were expressed at both mid-log and early stationary phases in the absence of glucose (Table 5). Significant induction of rbsB (periplasmic-binding protein), rbsK (cytoplasmic kinase), and rbsR (repressor) genes in the rbs operon were observed in the WT at mid-log phase (Table 6). Taga et al. (2001) and Xavier and Bassler (2005) found uptake of AI-2 even after the mutation of the lsrB gene, which is the receptor for AI-2 and therefore predicted the possibility of an alternative transport system. Recently, James et al. (2006) suggested the role of rbsB as a transporter of AI-2 in Actinobacillus actinomycetemocomitans. Interestingly, we observed that in contrast to the lsr operon that was induced at both time points, rbsB, rbsK, and rbsR were induced at mid-log phase when AI-2 was present and had no differential expression at early stationary phase when AI-2 had been degraded or uptaken by the cells (Fig. 1B). RbsB is known to be a LuxP and LsrB homolog (Taga et al., 2001; James et al., 2006). It is a part of the rbs operon that encodes six genes (rbsD, rbsA, rbsC, rbsB, rbsK, and rbsR) that form the ribose transport system (Iida et al., 1984; Bell et al., 1986). AI-2 synthesized by luxS is a ribose derivative, suggesting that the alternate transport system predicted (Taga et al., 2001; Xavier and Bassler, 2005) could be the rbs transport system in Salmonella Typhimurium.

The microarray results were validated using QRT-PCR on a few selected genes. Though the magnitude of the fold change obtained using QRT-PCR was different from the microarray analysis, the direction of regulation was the same with a high correlation coefficient (0.93) (Fig. 5). The high degree of concordance between the QRT-PCR and microarray results, and the use of multiple biological and technical replicates adds sufficient power to our analysis and confidence in these results.

Conclusions

The objective of this study was to examine the effect of luxS/AI-2 on global gene expression of Salmonella Typhimurium. Our results indicate that luxS and AI-2 are involved in the expression of genes that are related to motility, biofilm formation, virulence, translation, transcription, and other key cellular functions. The results show that AI-2 could play a vital role in the repression of key virulence-related genes in Salmonella Typhimurium.

Footnotes

Acknowledgments

This work was supported by Hatch grant H-8708 for the Texas AgriLife Research. We thank Bonnie L. Bassler for generously providing the V. harveyi BB170 reporter strains used in this study. DNA microarrays were obtained through National Institute of Allergy and Infectious Diseases (NIAID) Pathogen Function Genomics Resource Center, managed and funded by the Division of Microbiology and Infectious Disease, NIAID, National Institutes of Health, Department of Health and Human Services (DHHS), and operated by JCVI (formerly The Institute for Genomic Research).

Disclosure Statement

No competing financial interests exist.

References

Bassler

, Wright

, Showalter

, Silverman

. Intercellular signalling in Vibrio harveyi: sequence and function of genes regulating expression of luminescence. Mol Microbiol, 1993; 9:773–786.

Bell

, Buckel

, Groarke

et al. The nucleotide sequences of the rbsD, rbsA, and rbsC genes of Escherichia coli K12. J Biol Chem, 1986; 261:7652–7658.

Benjamini

, Hochberg

. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Royal Stat Soc B Meth, 1995; 57:289–300.

Chen

, Schauder

, Potier

et al. Structural identification of a bacterial quorum-sensing signal containing boron. Nature, 2002; 415:545–549.

Choi

, Shin

, Ryu

. Implication of quorum sensing in Salmonella enterica serovar typhimurium virulence: the luxS gene is necessary for expression of genes in pathogenicity island 1. Infect Immun, 2007; 75:4885–4890.

de Kievit

, Iglewski

. Bacterial quorum sensing in pathogenic relationships. Infect Immun, 2000; 68:4839–4849.

DeLisa

, Wu

, Wang

, Valdes

, Bentley

. DNA microarray-based identification of genes controlled by autoinducer 2-stimulated quorum sensing in Escherichia coli. J Bacteriol, 2001; 183:5239–5247.

Fong

, Chung

, Lamont

, Demuth

. Intra- and interspecies regulation of gene expression by Actinobacillus actinomycetemcomitans LuxS. Infect Immun, 2001; 69:7625–7634.

Fong

, Gao

, Demuth

. luxS and arcB control aerobic growth of Actinobacillus actinomycetemcomitans under iron limitation. Infect Immun, 2003; 71:298–308.

10.

Hegde

, Qi

, Abernathy

et al. A concise guide to cDNA microarray analysis. Biotechniques, 2000; 29:548–550552–544, 556 passim.

11.

Higgins

, Pomianek

, Kraml

et al. The major Vibrio cholerae autoinducer and its role in virulence factor production. Nature, 2007; 450:883–886.

12.

Hirsch

, Elliott

. Fis regulates transcriptional induction of RpoS in Salmonella enterica. J Bacteriol, 2005a. 187:1568–1580.

13.

Hirsch

, Elliott

. Stationary-phase regulation of RpoS translation in Escherichia coli. J Bacteriol, 2005b. 187:7204–7213.

14.

Hughes

, Roberts

, Dai

et al. Widespread aneuploidy revealed by DNA microarray expression profiling. Nat Genet, 2000; 25:333–337.

15.

Ichikawa

, Norris

, Bangera

et al. Interaction of Pseudomonas aeruginosa with epithelial cells: identification of differentially regulated genes by expression microarray analysis of human cDNAs. Proc Natl Acad Sci USA, 2000; 97:9659–9664.

16.

Iida

, Harayama

, Iino

, Hazelbauer

. Molecular cloning and characterization of genes required for ribose transport and utilization in Escherichia coli K-12. J Bacteriol, 1984; 158:674–682.

17.

James

, Shao

, Lamont

, Demuth

. The Actinobacillus actinomycetemcomitans ribose binding protein RbsB interacts with cognate and heterologous autoinducer 2 signals. Infect Immun, 2006; 74:4021–4029.

18.

Karavolos

, Bulmer

, Winzer

et al. LuxS affects flagellar phase variation independently of quorum sensing in Salmonella enterica serovar typhimurium. J Bacteriol, 2008; 190:769–771.

19.

Kelly

, Goldberg

, Carroll

et al. A global role for Fis in the transcriptional control of metabolism and type III secretion in Salmonella enterica serovar Typhimurium. Microbiology, 2004; 150:2037–2053.

20.

Livak

, Schmittgen

. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods, 2001; 25:402–408.

21.

Lyon

, Madden

, Levin

, Stein

, Caparon

. Mutation of luxS affects growth and virulence factor expression in Streptococcus pyogenes. Mol Microbiol, 2001; 42:145–157.

22.

Macnab

. Flagella and motility. Escherichia coli and Salmonella: Cellular and Molecular Biology. Neidhardt

, Curtis

III , Ingraham

, Lin

ECC

et al. Washington, DC: American Society for Microbiology Press, 1996; 123–145.

23.

McNab

, Ford

, El-Sabaeny

et al. LuxS-based signaling in Streptococcus gordonii: autoinducer 2 controls carbohydrate metabolism and biofilm formation with Porphyromonas gingivalis. J Bacteriol, 2003; 185:274–284.

24.

Miller

, Skorupski

, Lenz

, Taylor

, Bassler

. Parallel quorum sensing systems converge to regulate virulence in Vibrio cholerae. Cell, 2002; 110:303–314.

25.

Miller

, Xavier

, Campagna

et al. Salmonella Typhimurium recognizes a chemically distinct form of the bacterial quorum-sensing signal AI-2. Mol Cell, 2004; 15:677–687.

26.

O'Toole

, Kolter

. Flagellar and twitching motility are necessary for Pseudomonas aeruginosa biofilm development. Mol Microbiol, 1998; 30:295–304.

27.

Pillai

, Jesudhasan

. Quorum sensing: how bacteria communicate. Food Technol, 2006; 60:42–50.

28.

Prouty

, Schwesinger

, Gunn

. Biofilm formation and interaction with the surfaces of gallstones by Salmonella spp. Infect Immun, 2002; 70:2640–2649.

29.

Qin

, Luo

, Farrand

. Domains formed within the N-terminal region of the quorum-sensing activator TraR are required for transcriptional activation and direct interaction with RpoA from agrobacterium. J Biol Chem, 2004; 279:40844–40851.

30.

Soni

, Jesudhasan

, Cepeda

et al. Autoinducer AI-2 is involved in regulating a variety of cellular processes in Salmonella Typhimurium. Foodborne Pathog Dis, 2008; 5:147–153.

31.

Sperandio

, Mellies

, Nguyen

, Shin

, Kaper

. Quorum sensing controls expression of the type III secretion gene transcription and protein secretion in enterohemorrhagic and enteropathogenic Escherichia coli. Proc Natl Acad Sci USA, 1999; 96:15196–15201.

32.

Sperandio

, Torres

, Giron

, Kaper

. Quorum sensing is a global regulatory mechanism in enterohemorrhagic Escherichia coli O157:H7. J Bacteriol, 2001; 183:5187–5197.

33.

Sperandio

, Torres

, Jarvis

, Nataro

, Kaper

. Bacteria-host communication: the language of hormones. Proc Natl Acad Sci USA, 2003; 100:8951–8956.

34.

Surette

, Bassler

. Quorum sensing in Escherichia coli and Salmonella Typhimurium. Proc Natl Acad Sci USA, 1998; 95:7046–7050.

35.

Surette

, Bassler

. Regulation of autoinducer production in Salmonella Typhimurium. Mol Microbiol, 1999; 31:585–595.

36.

Taga

, Miller

, Bassler

. Lsr-mediated transport and processing of AI-2 in Salmonella Typhimurium. Mol Microbiol, 2003; 50:1411–1427.

37.

Taga

, Semmelhack

, Bassler

. The LuxS-dependent autoinducer AI-2 controls the expression of an ABC transporter that functions in AI-2 uptake in Salmonella Typhimurium. Mol Microbiol, 2001; 42:777–793.

38.

Wang

, Li

, March

, Valdes

, Bentley

. luxS-dependent gene regulation in Escherichia coli K-12 revealed by genomic expression profiling. J Bacteriol, 2005; 187:8350–8360.

39.

Whiteley

, Lee

, Greenberg

. Identification of genes controlled by quorum sensing in Pseudomonas aeruginosa. Proc Natl Acad Sci USA, 1999; 96:13904–13909.

40.

Widmer

, Jesudhasan

, Dowd

, Pillai

. Differential expression of virulence-related genes in a Salmonella enterica serotype Typhimurium luxS mutant in response to autoinducer AI-2 and poultry meat-derived AI-2 inhibitor. Foodborne Pathog Dis, 2007; 4:5–15.

41.

Winzer

, Hardie

, Burgess

et al. LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy-5-methyl-3(2H)-furanone. Microbiology, 2002; 148:909–922.

42.

Winzer

, Williams

. Quorum sensing and the regulation of virulence gene expression in pathogenic bacteria. Int J Med Microbiol, 2001; 291:131–143.

43.

Xavier

, Bassler

. LuxS quorum sensing: more than just a numbers game. Curr Opin Microbiol, 2003; 6:191–197.

44.

Xavier

, Bassler

. Regulation of uptake and processing of the quorum-sensing autoinducer AI-2 in Escherichia coli. J Bacteriol, 2005; 187:238–248.