and Chloride-Induced Protein in the Moderately Halophilic Bacterium ...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 2007, p. 371–379 0099-2240/07/$08.00⫹0 doi:10.1128/AEM.01625-06 Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Vol. 73, No. 2

Autoinducer-2-Producing Protein LuxS, a Novel Salt- and Chloride-Induced Protein in the Moderately Halophilic Bacterium Halobacillus halophilus䌤 Xaver Sewald,1† Stephan H. Saum,1,2 Peter Palm,3 Friedhelm Pfeiffer,3 Dieter Oesterhelt,3 and Volker Mu ¨ller1,2* Section Microbiology, Department Biology I, Ludwig Maximilians University, Munich, Germany1; Molecular Microbiology & Bioenergetics, Institute of Molecular Biosciences, Johann Wolfgang Goethe University Frankfurt, Frankfurt am Main, Germany2; and Max Planck Institute of Biochemistry, Department of Membrane Biochemistry, Martinsried, Germany3 Received 13 July 2006/Accepted 31 October 2006

The moderately halophilic bacterium Halobacillus halophilus carries a homologue of LuxS, a protein involved in the activated methyl cycle and the production of autoinducer-2, which mediates quorum sensing between certain species. luxS of H. halophilus is part of an operon that encodes an S-adenosylmethionine-dependent methyltransferase, a cysteine synthase, and a ␤-cystathionine lyase. Expression of luxS was growth phase dependent, with maximal expression in the mid-exponential growth phase. In addition, transcription of luxS was strictly salt dependent; maximal mRNA concentrations were observed with 2.0 M NaCl in the growth medium. Chloride ions stimulated luxS transcription by a factor of three. Western blot analyses demonstrated a growth phase- and salinity-dependent production of LuxS. Moreover, cellular LuxS levels were strictly chloride dependent. Maximal accumulation of LuxS was observed at 0.5 to 1.0 M Clⴚ and depended on the salinity.

spores, activation of transport of the compatible solute glycine betaine, motility, and flagellum production, were identified as Cl⫺-dependent processes (11, 21, 24). To identify components of the regulatory network, we compared the cellular protein contents of cells grown under different conditions. These analyses revealed six proteins that were upregulated by the chloride concentration. Their N termini were sequenced, and this sequence was used to identify similar proteins in the database (22). Because the cellular functions of the deduced proteins were rather dissimilar, we speculated about the presence of a chloride regulon by which the cells sense and respond to changes in the salinity of the medium (18). Based on analysis of a short stretch of 18 N-terminal amino acids, one protein upregulated by the chloride concentration was tentatively identified as LuxS. This was a rather interesting finding, since LuxS is a well-characterized protein in several bacteria, where it has dual functions: first, it is involved in the metabolism of S-adenosylmethionine (SAM) by the so-called “activated methyl cycle” (AMC), and second, one of the products of the LuxS-catalyzed reaction leads to the production of autoinducer-2 (AI-2), which is involved in quorum sensing in gram-negative and gram-positive bacteria (3, 10, 27, 30, 33, 34). In contrast to species-specific communication signals, AI-2 mediates interspecies communication (3, 27, 34). A function of LuxS in regulatory networks is interesting in the context of a chloride regulon involved in salt perception and response in H. halophilus. Although quorum sensing has not been described for moderate halophiles, it may be involved in salt perception, a process still not understood for moderate halophiles in general. These considerations have prompted us to analyze this protein and its encoding gene in more detail. Here we address the issue of the genomic organization of luxS and its regulation.

The moderately halophilic bacteria are a specialized group of organisms that are strictly salt dependent for growth. About 0.5 M of NaCl is required for optimal growth, and most interestingly, cells are able to grow over a rather wide range of external salinities (0.5 to 2.0 M), at similar rates, to comparable yields (31). The cellular responses to elevated environmental salinities are plentiful and involve structural changes in cell walls and changes in the cellular protein composition as well as the accumulation of solutes to prevent loss of cytoplasmic water (13, 19, 20). This behavior must involve signaling of the different salt concentrations and the presence of regulatory networks leading to differential gene expression and protein activation (16, 29, 32). Halobacillus halophilus is a moderately halophilic, gram-positive bacterium isolated from a salt marsh at the North Sea coast of Germany (4). It grows optimally at 0.5 to 2.0 M NaCl and requires Na⫹ for growth, a feature that it shares with other halophiles and nonhalophiles (1, 9, 17). An outstanding feature of H. halophilus is its chloride dependence for growth (23). H. halophilus is the first bacterium for which a chloride dependence has been demonstrated. H. halophilus accumulates compatible solutes, but not KCl, to counterbalance the external salt concentration (28), and therefore a function of Cl⫺ solely as an intracellular anionic osmolyte is excluded. Apart from growth, different physiological activities, such as germination of endo* Corresponding author. Mailing address: Molecular Microbiology & Bioenergetics, Institute of Molecular Biosciences, Johann Wolfgang Goethe University, Max-von-Laue-Str. 9, 60438 Frankfurt am Main, Germany. Phone: 49-69-79829507. Fax: 49-69-79829306. E-mail: [email protected]. † Present address: Bacteriology, Max von Pettenkofer Institute, Munich, Germany. 䌤 Published ahead of print on 3 November 2006. 371

372

APPL. ENVIRON. MICROBIOL.

SEWALD ET AL. TABLE 1. Oligonucleotides used in this study

Oligonucleotide

Sequence (5⬘–3⬘)

Use

LuxSI.1

ATGACACAGATGAACGTAGAA

LuxSI.2 luxS_RTI.1_rev ubiE_RTI.1_for ubiE_RTI.2_for cysK_RTI.2_rev cysK_RTI.2_for cm_RTI.1_rev LuxS(5⬘)E LuxS(3⬘)S PatpD1 PatpD2

GTACGTCTTTCAGTGTGTTCT TCATTACGTTTGTTAAGCAGTTC TCTCGATGAGCTGAAGGAAAT CTCCGGTTATGGAATTTGGT GTCGAACCAGAAGGCTCTGTC GTCGAACCAGAAGGCTCTGTC AGGATTCGTCGGCGTCTCAAC TTTTTTGAATTCGTGAAAGCTCCTTATATCCGC TTTTTTGTCGACAACTTCTGTAGCATTTAGTAC GGTTAGTGGAATTCGCCC TCTGAAAGCTGCAGCCATTA

The cellular function of LuxS in the context of the chloride regulon is also discussed. MATERIALS AND METHODS Organisms and cultivation. The bacterial strains used in this study were Halobacillus halophilus (DSMZ 2266) and Escherichia coli DH5␣. H. halophilus was maintained on 8 g nutrient broth (NB; Becton Dickinson, Heidelberg, Germany) per liter supplemented with 0.05 M magnesium sulfate and with sodium chloride, sodium nitrate, or sodium glutamate, as indicated. The pH was adjusted to pH 7.5. Cultures were grown at 30°C with agitation at 170 rpm. E. coli DH5␣ was grown on Luria broth (LB) containing 10 g peptone, 5 g yeast extract, and 10 g sodium chloride per liter and supplemented with ampicillin at a final concentration of 100 ␮g/ml. The cultures were incubated on a rotary shaker at 37°C at 140 rpm. DNA cloning. For generation of polyclonal antibodies, luxS was cloned into the EcoRI and SalI sites of pET21a(⫹) by using the oligonucleotides LuxS(5⬘)E (TTTTTTGAATTCGTGAAAGCTCCTTATATCCGC) and LuxS(3⬘)S (TTTT TTGTCGACAACTTCTGTAGCATTTAGTAC), resulting in the plasmid construct pLuxS1. Correct plasmid constructs were confirmed by DNA sequencing using an ABI Prism 310 genetic analyzer (Applied Biosystems, Darmstadt, Germany). PCR was performed in an MJ minicycler (Biometra, Go ¨ttingen, Germany) with Taq-PCR core kits (QIAGEN, Hilden, Germany). Plasmid DNA was isolated using a QIAprep Spin miniprep kit (QIAGEN, Hilden, Germany), chromosomal DNA was isolated as described previously (12), and PCR products were purified with a nucleotide removal kit (QIAGEN, Hilden, Germany) or Qiaex II gel extraction kit (QIAGEN, Hilden, Germany). Protein purification and generation of polyclonal antibodies. The plasmid construct pLuxS1 was used for purification of the fusion protein LuxS-His tag. The construct contains a region downstream of luxS encoding a His tag consisting of six histidine residues and a promoter region for T7 polymerase upstream of luxS. pLuxS1 was chemically transformed, as described previously (25), into E. coli BL21(DE3) containing a chromosomally located and IPTG (isopropyl-␤-Dthiogalactopyranoside)-inducible gene encoding a T7 polymerase. The fusion protein, LuxS-His6, was purified using a chelating Sepharose Fast Flow column (Amersham Biosciences, Freiburg, Germany) following the instructions of the supplier. The success of the purification was checked by sodium dodecyl sulfatepolyacrylamide gel electrophoresis (SDS-PAGE) according to a previously described method (26), and 975 ␮g of the purified fusion protein was used to raise polyclonal antibodies in rabbits (SEQLAB, Go ¨ttingen, Germany). Probe construction and labeling. Chromosomal DNA of H. halophilus was prepared as described previously (12). For the atpD probe, a genomic library was constructed by partial digestion of chromosomal DNA with MboI and cloning into the BamHI restriction site of pUC18. An atpD-containing clone was identified from this genomic library by colony hybridization (2), using the corresponding gene of Acetobacterium woodii as a probe (14). atpD was cloned into the PstI site of pUC18, using the oligonucleotides PatpD1 (GGTTAGTGGAATTCG CCC) and PatpD2 (TCTGAAAGCTGCAGCCATTA). For the luxS probe, a 327-bp fragment of luxS was amplified using the oligonucleotides LuxSI.1 (AT GACACAGATGAACGTAGAA) and LuxSI.2 (GTACGTCTTTCAGTGTGT TCT). The DNA fragments were radiolabeled with [␥-32P]dATP (Hartmann Analytic GmbH, Braunschweig, Germany), using a random primer DNA labeling system

RT-PCR (primer 1 in Fig. 1), probe construction for luxS Probe construction for luxS RT-PCR (luxS primer in Fig. 1) RT-PCR (primer 2 in Fig. 1) RT-PCR (primer 3 in Fig. 1) RT-PCR (orf4 primer in Fig. 1) RT-PCR (bridges orf4-orf5) RT-PCR (bridges orf4-orf5) Construction of polyclonal antibodies Construction of polyclonal antibodies Probe construction for atpD Probe construction for atpD

(Invitrogen, Karlsruhe, Germany). Following 32P labeling, probes were separated from unincorporated nucleotides by using a QIAquick nucleotide removal kit (QIAGEN, Hilden, Germany). The specificities of the probes were confirmed by Southern blot analyses as described previously (25). Northern blots. Cells were grown in NB supplemented with 0.05 M magnesium sulfate and different salts and concentrations, as indicated. A 40-ml portion of the culture was harvested by centrifugation and resuspended in 200 ␮l TE buffer (10 mM Tris, 1 mM EDTA [pH 8.0]). Six milligrams of lysozyme was added, and the cells were incubated at room temperature for 3 min. Following centrifugation, the sediment was resuspended in 1 ml of PeqGOLD RNApure (peqLAB, Erlangen, Germany) heated up to 65°C for total RNA isolation. After incubation for 10 min at room temperature, the samples were frozen with liquid nitrogen. After thawing of the samples, 100 ␮l of chloroform was added, and the following steps were performed according to the instructions of the manufacturer (peqLAB, Erlangen, Germany). RNAs were dissolved in water (RNase-free through diethyl pyrocarbonate treatment), and the total RNA concentration was determined by spectroscopy. RNA preparations contained 0.6 to 1 mg RNA/ml. Denaturing agarose gel electrophoresis of RNAs in the presence of formaldehyde, transfer to Hybond N nylon membranes (Amersham Biosciences, Freiburg, Germany), and Northern blot hybridization were performed essentially as described previously (25). Finally, the blots were visualized using Kodak storage phosphor screens and a Storm 860 laser scanner (GE Healthcare Europe GmbH, Munich, Germany). Densitometric analyses were performed with Image Quant software. RT-PCR. Residual DNA contamination of the total RNA samples was removed by DNase I treatment (Boehringer, Biberach, Germany). The success of this treatment was checked by PCR. RNAs were considered DNA-free when no product was produced under conditions where a clear product was gained with chromosomal DNA as the template. Two micrograms of this isolated total RNA from a mid-exponential-phase H. halophilus culture grown on NB (supplemented with 2 M NaCl) was used in a reverse transcriptase (RT) reaction. For the experiment, an Omniscript RT kit (QIAGEN, Hilden, Germany) and the indicated oligonucleotides were used. The RT reaction was performed according to the method in the instruction manual. Subsequently, a PCR with the indicated oligonucleotides was performed, using the cDNA from the RT reaction as the template. The oligonucleotides used in this experiment are listed in Table 1. SDS-PAGE and immunoblots. Cells were harvested by centrifugation (8,000 ⫻ g, 15 min), washed once with 1 ml of washing buffer (0.05 M magnesium sulfate and an appropriate salt concentration), and resuspended, after centrifugation (10,000 ⫻ g, 4 min), in denaturing buffer as described previously (26). The samples were boiled for 15 min. Proteins were separated by SDS-PAGE on 12.5% gels and transferred to nitrocellulose membranes (Whatman, Dassel, Germany), using a semidry blotting chamber (Bio-Rad, Munich, Germany). Membranes were blocked with 1% skim milk powder in PBST (140 mM NaCl, 10 mM KCl, 6.4 mM Na2HPO4, 2 mM KH2PO4, 0.05% Tween 20) for 1 h, washed four times in PBST for 20 min, and incubated with antisera (0.75 ␮l/ml PBST) for 2 h. The membranes were washed three times with PBST for 20 min and then incubated for 1 h with protein A-horseradish peroxidase conjugate, followed by three additional washing steps (10 min) with PBST. All steps were performed at room temperature. Luminescence was detected using a chemiluminescence blotting substrate from Roche Molecular Biochemicals (Mannheim, Germany), and signals were detected by autoradiography with Kodak X-OMAT-AR film.

LuxS PROTEIN OF H. HALOPHILUS

VOL. 73, 2007

373

FIG. 1. Genomic organization of luxS. Panel A displays the genomic organization of luxS, and panel B shows the results of PCR analyses to determine the transcriptional organization. Total RNA from H. halophilus cells grown in NB in the presence of 2 M NaCl was isolated and transcribed into cDNA, using the indicated primers. The resulting cDNAs were used for further PCR analyses with primers 1, 2, and 3 in combination with the luxS or orf4 primer. Primer binding sites are indicated with arrows. The resulting PCR products were separated in 0.8% agarose gels. The resulting fragments are shown in the right part of panel B. As a positive control, chromosomal DNA was used in the same PCRs (left part of panel B). As a negative control, the PCRs were performed with the same RNAs with which cDNAs were produced to ensure that the RNAs were free of DNA prior to transcription to cDNAs (data not shown).

DNA sequence determination. DNA sequences were retrieved from the genome sequence of H. halophilus, which will be described elsewhere. Nucleotide sequence accession numbers. DNA sequences were deposited in the GenBank database with the accession numbers EF088800 (orf1), EF088801 (orf2), EF088802 (luxS), EF088803 (orf4), EF088804 (orf5), and EF088805 (orf6).

RESULTS Genetic organization of luxS in H. halophilus. Previously, we reported the chloride-dependent regulation of several proteins in H. halophilus (22). One was tentatively identified as a homologue of LuxS based on BLAST analyses with the N-terminal sequence MQMNVEVFNLDHTKVKAP. We have now used the N-terminal sequence to identify the corresponding gene in the genome of H. halophilus. The sequence exactly matches the sequence derived from only one open reading frame. This open reading frame is 459 bp long and starts with an ATG codon. Upstream of the ATG is a well-placed and conserved Shine-Dalgarno sequence. The deduced protein sequence is very similar to those of the LuxS proteins of several other bacteria (see below), and therefore the gene was designated luxS. Inspection of the genomic sequence did not reveal another open reading frame that encodes a protein with similarity to LuxS. luxS is part of a cluster of four genes (Fig. 1). Upstream of luxS is an open reading frame of 642 bp (orf2) that

is separated from luxS by only 3 bp. Downstream of luxS are two open reading frames: orf4 is separated by only 5 bp from luxS, and orf5 is separated from orf4 by only 18 bp. The cluster is flanked at the 5⬘ end by orf1 and at the 3⬘ end by orf6, which are separated from the cluster by 160 and 115 bp, respectively. In front of orf2 is a putative ␴A-dependent promoter. This promoter lies 48 bp upstream of the translational start point (ATG) of orf2 and is 90% identical to the consensus sequence for the corresponding ␴ factor in Bacillus subtilis. No similarity to a ␴B-dependent promoter (essential in the stress response) was found. Additionally, a putative terminator sequence was identified downstream of orf5 by inspection of the DNA sequence. Here the sequence shows a dyadic symmetry that is able to form a stem-loop at the RNA level, followed by seven thymidine bases. These data suggest that orf2, luxS, orf4, and orf5 form a transcriptional unit. To determine the transcriptional organization of luxS, cells were grown in 2 M NaCl in NB to late exponential growth phase, and total RNA was isolated and reverse transcribed, using gene-specific primers that bind in luxS or orf4. PCR with primers that bind in orf4 or luxS and primers 1 to 3 (Fig. 1A; Table 1) gave amplification products of the expected sizes (Fig. 1B). There was no amplification product by using the orf4 primer and primer 2, but there was a product by using the orf4

374

SEWALD ET AL.

FIG. 2. orf4 and orf5 are part of one transcriptional unit. Total RNA from H. halophilus cells grown in NB in the presence of 2 M NaCl was isolated and transcribed into cDNA. The resulting cDNAs were used for further PCR analysis, using the primers cysK_RTI.2_for and cm_RTI.1_rev (B). The resulting fragment is indicated with an arrow. It bridges the gap between orf4 and orf5 and has a size of 770 bp. As a positive control, chromosomal DNA was used as a template (A).

primer and primer 3. The inability to amplify a product with the orf4 primer and primer 2 might result from inefficient binding of primer 2 and/or premature transcription termination from site 2. These data clearly show that orf2, luxS, and orf4 are part of one transcriptional unit. This was confirmed using specific, 32P-labeled primers for cDNA synthesis (data not shown). Furthermore, we were able to amplify a region that covers part of orf4 and orf5 (Fig. 2), but we were never able to obtain products that bridge orf1 and orf2 or orf5 and orf6. Therefore, it was concluded that orf2, luxS, orf4, and orf5 comprise an operon. Properties of the products of the luxS operon of H. halophilus. The deduced LuxS protein has 152 amino acids and a predicted molecular mass of 17.3 kDa and is very similar to the LuxS proteins from Oceanobacillus iheyensis (153 amino acids [aa]; 72% identity and 90% similarity), Staphylococcus aureus (156 aa; 72% identity and 86% similarity), and Chromoha-


FIG. 4. Growth of Halobacillus halophilus in the presence of different NaCl concentrations. H. halophilus cells were cultivated in NB in the presence of 0.5 (■), 1.0 (F), or 2.0 (}) M NaCl. Samples were taken at the early exponential, mid-exponential, or stationary growth phase (indicated by arrows) for RNA isolation and Northern blot analysis.

lobacter salexigens (155 aa; 70% identity and 86% similarity) (Fig. 3). Orf2 has a predicted molecular mass of 24.2 kDa and is similar to S-adenosylmethionine-dependent methyltransferases from Bacillus licheniformis (45% identity and 67% similarity), Bacillus sp. (46% identity and 66% similarity), and Geobacillus kaustophilus (48% identity and 65% similarity). Orf4 and Orf5 have deduced molecular masses of 32.9 kDa and 41.4 kDa, respectively. They are similar to cysteine synthases (from Bacillus cereus [68% identity and 81% similarity], Geobacillus kaustophilus [65% identity and 79% similarity], and Bacillus thuringiensis [67% identity and 81% similarity]) and cystathionine-␤-lyases (from Bacillus sp. [72% identity and 84% similarity], Bacillus licheniformis [68% identity and 82% similarity], and Geobacillus kaustophilus [68% identity and 82% similarity]), respectively. orf1 encodes a putative ATP-dependent RNA helicase with a deduced molecular mass of 44.4 kDa, and orf6 encodes a

FIG. 3. Alignment of LuxS sequence from H. halophilus with those of similar proteins deposited in databases. Identical amino acids are indicated by asterisks. The alignment was performed with ClustalW.

VOL. 73, 2007


375

FIG. 5. Transcription of luxS is growth phase and salinity dependent. H. halophilus cells were cultivated in the presence of 0.5, 1.0, or 2.0 M NaCl, and samples were taken in the early exponential, mid-exponential, or stationary growth phase in order to isolate total RNA (see Fig. 4). For Northern blot analysis, 10 ␮g of total RNA was separated in a 1.5% denaturing agarose gel and blotted on a nylon membrane. As a loading control, a specific probe for atpD, which encodes the ␤-subunit of the F1F0 ATP synthase of H. halophilus, was used.

putative DnaK suppressor protein; both are not in an obvious functional relationship with the luxS cluster. Expression of luxS is growth phase and NaCl dependent in H. halophilus. To determine the cellular levels of luxS transcripts, Northern blotting using a luxS-specific probe was performed. Southern analyses confirmed the specificity of the probe (data not shown). As a control, the level of atpD, which encodes the ␤ subunit of the F1F0 ATP synthase, was monitored. H. halophilus was grown in NB at a salinity of 0.5, 1.0, or 2.0 M NaCl, and samples were taken at early or mid-exponential growth phase and at stationary phase (Fig. 4). Regulation of luxS transcription was clearly growth phase dependent. luxS transcripts were close to the detection limit in early exponential (optical density at 600 nm [OD600] ⫽ 0.3) or early stationary (OD600 ⫽ 7) phase (Fig. 5A and C). However, in exponentially growing cells (OD600 ⫽ 2), luxS transcripts

were detectable. As shown in Fig. 5B, fragments of 0.7, 1.3, 3.2, and 4.0 kbp reacted with the probe. The 4.0-kb fragment corresponds to a fragment that could cover orf2, luxS, orf4, and orf5, and the smaller fragments may represent fragments that arose by (nonspecific or specific) degradation or by transcription initiation from a different promoter. The smallest fragment, of 700 bp, had the strongest intensity, and based on its size, it may represent luxS only. Most interestingly, expression of luxS was strongly salt dependent. Quantification revealed an 11-fold increase of the 700-bp fragment from 0.5 to 2.0 M salt. Cellular levels of LuxS are salt dependent. To confirm the salt-dependent expression of luxS, the cellular levels of LuxS in cells grown to mid-exponential growth phase were determined by Western blotting using a LuxS-specific polyclonal antiserum. It is evident from Fig. 6A that the cellular concentration of LuxS increased with increasing salinity. Maximal levels

376

SEWALD ET AL.


FIG. 6. Salt and chloride dependence of cellular LuxS levels. H. halophilus was grown at 0.5, 1.0, or 2.0 M NaCl (A) or at a constant salinity of 2 M while the Cl⫺ concentration was adjusted to 0.5 M, 1.0 M, or 2.0 M (supplied as NaCl) (B). To gain isosmotic conditions, aliquots of NO3⫺ were added. Cells were harvested in the mid-exponential growth phase (see Fig. 4), and 20 ␮g of protein was used for Western blot analysis with a LuxS-specific antibody.

were detected in cells grown at 2.0 M NaCl, which corresponds nicely to the results of Northern blot analyses. It should be mentioned that the growth rate is in the same range at NaCl concentrations between 0.5 and 2 M (23). Interestingly, three bands reacted with the antiserum. The smallest one was a bit smaller than expected (15-kDa apparent mass versus 17.4-kDa deduced mass). This could be due to inherent problems with determining molecular masses by SDS-PAGE or to posttranslational modification events. The signals resulting from bigger proteins could represent higher-order states of LuxS (dimers and tetramers). The results of the aforementioned experiments are in line with the hypothesis that the cellular concentration of LuxS is salt dependent. However, given the chloride dependence of gene expression and protein production in H. halophilus (22), the effects could also be interpreted to result from an increased chloride concentration. To address this possibility, the experiments were repeated with cells grown at a constant chloride concentration of 0.5 M (supplied as NaCl) but with an increasing salt concentration (by the addition of NaNO3). These experiments revealed the same magnitudes in the increases of luxS transcripts and cellular LuxS levels as those seen before (data not shown). Taken together, these results clearly demonstrate a NaCl dependence of cellular LuxS levels. Transcript levels of luxS are chloride dependent. To address a potential role of chloride in the regulation of luxS transcription, cells had to be grown in the absence and presence of chloride. It should be remembered that H. halophilus does not grow in the presence of Na2SO4 but can adapt to 1 M NaNO3 after prolonged incubation, although growth is much slower and does not reach the final optical densities observed in the presence of NaCl (23). Cells were grown with 1.0 M NaCl or 1 M NaNO3, and cellular levels of luxS transcripts were determined by Northern blot analyses. As shown in Fig. 7, luxS transcripts were increased threefold in the presence of chloride. When the chloride concentration was varied from 0.5 M to 2.0 M at a constant salt concentration of 2.0 M (by appro-

FIG. 7. Transcription of luxS is Cl⫺ dependent. H. halophilus was cultivated in complex medium in the presence of 1 M NaCl or 1 M NaNO3. Total RNA was isolated from cells that were grown to the mid-exponential growth phase. For Northern blot analysis, 10 ␮g of total RNA was loaded per lane and separated in a 1.5% agarose gel. After blotting of the RNA on a nylon membrane, it was hybridized with a probe specific for luxS. As a loading control, a probe specifically binding atpD, encoding the ␤-subunit of the F1F0 ATP synthase, was used.


VOL. 73, 2007

377

FIG. 8. AMC and biosynthesis of homocysteine in H. halophilus as deduced from the genome sequence. For explanations, see the text. 1, serine transacetylase; 2, cysteine synthase; 3, cystathionine ␥-synthase; 4, cystathionine ␤-lyase; 5, methyltransferase(s); 6, adenosylmethionine synthase; 7, methyltransferase; 8, adenosylhomocysteine nucleosidase; 9, S-ribosylhomocysteine lyase; 10, S-adenosylhomocysteine hydrolase.

priate addition of NaNO3) or from 0.5 to 1.0 M at a constant salt concentration of 1.0 M, no further increase in luxS transcripts was observed (data not shown). These experiments demonstrate an increase of luxS transcripts in the presence of chloride, but maximal levels were already observed at 0.5 M NaCl when the total salt concentration was set at 1.0 or 2.0 M salt. Cellular levels of LuxS are chloride dependent. Previously, we had shown that the cellular level of the structural component of the flagellum, flagellin (FliC), was much more affected by chloride than was the transcript level of fliC (24). Therefore, we determined the effect of chloride on the cellular levels of LuxS in H. halophilus grown in NB in the presence of different salts. In cells grown in the presence of 1.0 M NaNO3, LuxS was hardly detectable. However, in cells grown with 0.5 M NaCl plus 0.5 M NaNO3, LuxS was readily detectable. A further increase of the Cl⫺ concentration to 1 M did not lead to a concomitant increase in the LuxS level. These experiments were repeated at a salinity of 2 M, which was shown to be optimal for LuxS production. The salt concentration was kept constant by appropriate addition of NaNO3. Again, in the absence of Cl⫺, there was no LuxS detectable in the cells, but the level increased dramatically at 1.0 and 2.0 M Cl⫺ (Fig. 6B). These experiments clearly demonstrated a strict chloride dependence of the cellular level of LuxS. DISCUSSION The data presented here demonstrate for the first time the presence of LuxS, a protein involved in the AMC and in quorum sensing, in a moderate halophile. The cellular LuxS concentration was strongly dependent on the salinity of the medium. At 0.5 M NaCl, cellular LuxS levels were close to the detection limit. At 1.0 M NaCl, levels increased dramatically, but maximal cellular LuxS concentrations were observed at 2.0 M NaCl. The growth rates remained comparable under these

conditions, so growth rate-dependent regulation can be excluded. In addition, LuxS levels were strictly chloride dependent. The highest levels were already observed at 0.5 M Cl⫺ (with 1.0 M sodium salts) or 1.0 M Cl⫺ (with 2 M sodium salts). Furthermore, LuxS levels were dependent on the growth phase, with a maximal concentration in exponentially growing cells. Obviously, the LuxS level is regulated by different stimuli, including salinity, chloride, and growth phase. Regulation by various factors is indeed observed very often and reflects the requirement of cross talk between different regulatory networks. Expression of luxS in some other prokaryotes is regulated by the growth phase, and genome-wide expression profiling with E. coli revealed that luxS expression is part of a general stress response system (7). An interesting scenario could involve regulation of the AMC by the growth phase and regulation of the AI-2-associated functions by chloride. Growth phase dependence of LuxS production has been observed before in gram-negative and gram-positive organisms (8, 15), but salt- and chloride-dependent regulation of LuxS production is demonstrated here for the first time. The only other component of the chloride regulon of H. halophilus characterized to date is FliC. Motility of H. halophilus is chloride dependent, and it turned out that the production of the major component of the flagellum, flagellin (FliC), is strongly chloride dependent (24). There is a common motif in the regulation of FliC and LuxS: gene expression was only marginally stimulated by Cl⫺, but synthesis of the proteins was strictly dependent on Cl⫺. Therefore, different regulatory layers, i.e., transcription, posttranscription, translation, or posttranslation, might be simultaneously and/or differently affected by Cl⫺. Precedence for different layers of regulation is plentiful. One well-studied system is the synthesis of the E. coli alternative sigma factor ␴s, whose production is regulated on different levels (22). LuxS mutations abolish AI-2 production, and this leads, for example, to Escherichia coli K-12 or Campylobacter jejuni cells

378

SEWALD ET AL.

with decreased motility or Neisseria meningitides cells impaired in the development of full meningococcal virulence (7). It did not escape our notice that these or similar processes are also connected in H. halophilus. Chloride was shown to regulate motility and to alter the cell wall composition of H. halophilus, as determined by electron microscopy (24). Whether chloride affects LuxS, which then regulates FliC and proteins involved in cell wall maintenance, or whether all three are regulated in parallel remains to be established. Unfortunately, a genetic system for H. halophilus is not yet available, and therefore this question cannot be addressed by mutant studies. LuxS is part of the AMC, which serves in the cycling of building blocks for SAM biosynthesis (Fig. 8) (30). The cellular function of SAM is that of a methyl group donor for various purposes, giving rise to a methylated product and S-adenosylhomocysteine (SAH). SAH is converted to homocysteine (HCY) by one of two possible routes: in one pathway, SAH is hydrolyzed in a single-step reaction to HCY by SAH hydrolase, and in another pathway, SAH is converted by the enzyme S-adenosylhomocysteine nucleosidase (Pfs) to S-ribosylhomocysteine, which is converted by LuxS to 4,5-dihydroxypentane2,3-dione (DPD) and HCY. The latter is methylated to yield methionine, which is the precursor again of SAM. A pfs gene is also present on the chromosome of H. halophilus, and preliminary data suggest that its expression is also salt dependent. A SAH hydrolase gene could not be found in the genome, indicating that H. halophilus uses the two-step pathway for detoxification of SAH. In some bacteria, notably gram-positive bacteria, luxS is part of an operon encoding proteins, such as CysK, MetA, MetB, and MetE, that might be involved in restoration of the cellular HCY pool from cysteine. The luxS operon of H. halophilus encodes proteins potentially involved in the AMC cycle. Orf4 is similar to cysteine synthase/o-acetylserine lyase, an enzyme involved in the biosynthesis of cysteine from O-acetylserine and H2S. Orf2 is similar to SAM-dependent transferases and therefore might be involved in demethylation of SAM to SAH. Orf5 is similar to cystathionine-␤-lyase (EC 4.4.1.8). The enzyme catalyzes the reaction L-cystathionine ⫹ H2O 3 L-homocysteine ⫹ NH3 ⫹ pyruvate and could therefore be involved in the production of homocysteine from cystathionine. This is, at least to the best of our knowledge, the first case in which luxS is organized in an operon together with three genes with potential function in the AMC or in generating building blocks for the AMC. The by-product of the LuxS-catalyzed reaction, DPD, is the precursor of AI-2. It was beyond the scope of this paper to prove the production of AI-2 by H. halophilus, but this seems likely. Quorum-sensing signals confer multicellular behavior to planktonic microbes. This includes, for example, the production of pathogenicity factors, secretion of catabolic enzymes, biofilm formation, and motility (5, 6, 15, 34). The last two behaviors might also be regulated by LuxS in H. halophilus. Biofilm formation has not been observed in H. halophilus, but at least the cell surface structure changes with the chloride concentration. This is at least an indication of chloride-induced morphological changes. Morphologically, cells of H. halophilus build sarcina-like structures. These are motile, and motility requires the coordinated synthesis of flagella in individual cells in the sarcinas. Again, the production of flagella is chloride dependent.


In summary, LuxS is the second identified protein of H. halophilus that is produced only in the presence of chloride. LuxS, via its involvement in quorum sensing, might be a central player in the chloride regulon of H. halophilus. ACKNOWLEDGMENT This study was supported by a grant from the German-Israel Foundation for Scientific Research and Development (GIF). REFERENCES 1. Albers, S. V., J. L. C. M. van de Vossenberg, A. J. Driessen, and W. N. Konings. 2001. Bioenergetics and solute uptake under extreme conditions. Extremophiles 5:285–294. 2. Buluwela, L., A. Forster, T. Boehm, and T. H. Rabbitts. 1989. A rapid procedure for colony screening using nylon filters. Nucleic Acids Res. 17:452. 3. Chen, X., S. Schauder, N. Potier, A. Van Dorsselaer, I. Pelczer, B. L. Bassler, and F. M. Hughson. 2002. Structural identification of a bacterial quorumsensing signal containing boron. Nature 415:545–549. 4. Claus, D., F. Fahmy, H. J. Rolf, and N. Tosunoglu. 1983. Sporosarcina halophila sp. nov., an obligate, slightly halophilic bacterium from salt marsh soils. Syst. Appl. Microbiol. 4:496–506. 5. Daines, D. A., M. Bothwell, J. Furrer, W. Unrath, K. Nelson, J. Jarisch, N. Melrose, L. Greiner, M. Apicella, and A. L. Smith. 2005. Haemophilus influenzae luxS mutants form a biofilm and have increased virulence. Microb. Pathog. 39:87–96. 6. Daniels, R., J. Vanderleyden, and J. Michiels. 2004. Quorum sensing and swarming migration in bacteria. FEMS Microbiol. Rev. 28:261–289. 7. DeLisa, M. P., and W. E. Bentley. 2002. Bacterial autoinduction: looking outside the cell for new metabolic engineering targets. Microb. Cell Fact. 1:5. 8. DeLisa, M. P., J. J. Valdes, and W. E. Bentley. 2001. Mapping stress-induced changes in autoinducer AI-2 production in chemostat-cultivated Escherichia coli K-12. J. Bacteriol. 183:2918–2928. 9. Dimroth, P. 1994. Bacterial sodium ion-coupled energetics. Antonie Leeuwenhoek 65:381–395. 10. Doherty, N., M. T. Holden, S. N. Qazi, P. Williams, and K. Winzer. 2006. Functional analysis of luxS in Staphylococcus aureus reveals a role in metabolism but not quorum sensing. J. Bacteriol. 188:2885–2897. 11. Dohrmann, A. B., and V. Mu ¨ller. 1999. Chloride dependence of endospore germination in Halobacillus halophilus. Arch. Microbiol. 172:264–267. 12. Errington, J. 1984. Efficient Bacillus subtilis cloning system using bacteriophage vector ⌽ 105J9. J. Gen. Microbiol. 130:2615–2628. 13. Falb, M., F. Pfeiffer, P. Palm, K. Rodewald, V. Hickmann, J. Tittor, and D. Oesterhelt. 2005. Living with two extremes: conclusions from the genome sequence of Natronomonas pharaonis. Genome Res. 15:1336–1343. 14. Forster, A., R. Daniel, and V. Mu ¨ller. 1995. The Na⫹-translocating ATPase of Acetobacterium woodii is a F1F0-type enzyme as deduced from the primary structure of its ␤, ␥ and ε subunits. Biochim. Biophys. Acta 1229:393–397. 15. Lombardia, E., A. J. Rovetto, A. L. Arabolaza, and R. R. Grau. 2006. A LuxS-dependent cell-to-cell language regulates social behavior and development in Bacillus subtilis. J. Bacteriol. 188:4442–4452. 16. Mukhopadhyay, A., Z. He, E. J. Alm, A. P. Arkin, E. E. Baidoo, S. C. Borglin, W. Chen, T. C. Hazen, Q. He, H. Y. Holman, K. Huang, R. Huang, D. C. Joyner, N. Katz, M. Keller, P. Oeller, A. Redding, J. Sun, J. Wall, J. Wei, Z. Yang, H. C. Yen, J. Zhou, and J. D. Keasling. 2006. Salt stress in Desulfovibrio vulgaris Hildenborough: an integrated genomics approach. J. Bacteriol. 188:4068–4078. 17. Mu ¨ller, V., S. Aufurth, and S. Rahlfs. 2001. The Na⫹ cycle in Acetobacterium woodii: identification and characterization of a Na⫹ translocating F1F0ATPase with a mixed oligomer of 8 and 16 kDa proteolipids. Biochim. Biophys. Acta 1505:108–120. 18. Mu ¨ller, V., and S. H. Saum. 2005. The chloride regulon of Halobacillus halophilus: a novel regulatory network for salt perception and signal transduction in bacteria, p. 303–310. In N. Gunde-Cimerman, A. Oren, and A. Plemenitas (ed.), Adaption to life at high salt concentrations in Archaea, Bacteria, and Eukarya. Springer, Dordrecht, The Netherlands. 19. Oren, A., F. Larimer, P. Richardson, A. Lapidus, and L. N. Csonka. 2005. How to be moderately halophilic with broad salt tolerance: clues from the genome of Chromohalobacter salexigens. Extremophiles 9:275–279. 20. Roberts, M. F. 2005. Organic compatible solutes of halotolerant and halophilic microorganisms. Saline Systems 1:5. 21. Roessler, M., and V. Mu ¨ller. 2001. Chloride dependence of glycine betaine transport in Halobacillus halophilus. FEBS Lett. 489:125–128. 22. Roessler, M., and V. Mu ¨ller. 2002. Chloride, a new environmental signal molecule involved in gene regulation in a moderately halophilic bacterium, Halobacillus halophilus. J. Bacteriol. 184:6207–6215. 23. Roessler, M., and V. Mu ¨ller. 1998. Quantitative and physiological analyses of chloride dependence of growth of Halobacillus halophilus. Appl. Environ. Microbiol. 64:3813–3817. 24. Roessler, M., G. Wanner, and V. Mu ¨ller. 2000. Motility and flagellum syn-


VOL. 73, 2007

25.

26.

27.

28.

29.

thesis in Halobacillus halophilus are chloride dependent. J. Bacteriol. 182: 532–535. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Scha ¨gger, H., and G. von Jagow. 1987. Tricine-sodium dodecyl sulfatepolyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166:369–379. Schauder, S., K. Shokat, M. G. Surette, and B. L. Bassler. 2001. The LuxS family of bacterial autoinducers: biosynthesis of a novel quorum-sensing signal molecule. Mol. Microbiol. 41:463–476. Severin, J. 1993. Kompatible Solute und Wachstumskinetik bei halophilen aeroben heterotrophen Eubakterien. Ph.D. thesis. University of Bonn, Bonn, Germany. Steil, L., T. Hoffmann, I. Budde, U. Vo ¨lker, and E. Bremer. 2003. Genome-

30. 31. 32. 33.

34.

379

wide transcriptional profiling analysis of adaptation of Bacillus subtilis to high salinity. J. Bacteriol. 185:6358–6370. Vendeville, A., K. Winzer, K. Heurlier, C. M. Tang, and K. R. Hardie. 2005. Making ‘sense’ of metabolism: autoinducer-2, LuxS and pathogenic bacteria. Nat. Rev. Microbiol. 3:383–396. Ventosa, A., J. J. Nieto, and A. Oren. 1998. Biology of moderately halophilic aerobic bacteria. Microbiol. Mol. Biol. Rev. 62:504–544. Weber, A., and K. Jung. 2002. Profiling early osmostress-dependent gene expression in Escherichia coli using DNA macroarrays. J. Bacteriol. 184: 5502–5507. Winzer, K., K. R. Hardie, N. Burgess, N. Doherty, D. Kirke, M. T. Holden, R. Linforth, K. A. Cornell, A. J. Taylor, P. J. Hill, and P. Williams. 2002. LuxS: its role in central metabolism and the in vitro synthesis of 4-hydroxy5-methyl-3(2H)-furanone. Microbiology 148:909–922. Xavier, K. B., and B. L. Bassler. 2003. LuxS quorum sensing: more than just a numbers game. Curr. Opin. Microbiol. 6:191–197.