Cationic antimicrobial peptides elicit a complex stress response in Bacillus subtilis that involves ECF-type sigma factors and two-component signal transduction systems

Milla Pietiäinen, Marika Gardemeister, Maria Mecklin, Soile Leskelä, Matti Sarvas and Vesa P. Kontinen

Vaccine Development Laboratory, National Public Health Institute, Mannerheimintie 166, FIN-00300 Helsinki, Finland

Correspondence
Vesa P. Kontinen
Vesa.Kontinen{at}ktl.fi


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Stress responses of Bacillus subtilis to membrane-active cationic antimicrobial peptides were studied. Global analysis of gene expression by DNA macroarray showed that peptides at a subinhibitory concentration activated numerous genes. A prominent pattern was the activation of two extracytoplasmic function sigma factor regulons, SigW and SigM. Two natural antimicrobial peptides, LL-37 and PG-1, were weak activators of SigW regulon genes, whereas their synthetic analogue poly-L-lysine was clearly a stronger activator of SigW. It was demonstrated for the first time that LL-37 is a strong and specific activator of the YxdJK two-component systems, one of the three highly homologous two-component systems sensing antimicrobial compounds. YxdJK regulates the expression of the YxdLM ABC transporter. The LiaRS (YvqCE) TCS was also strongly activated by LL-37, but its activation is not LL-37 specific, as was demonstrated by its activation with PG-1 and Triton X-100. Other strongly LL-37-induced genes included yrhH and yhcGHI. Taken together, the responses to cationic antimicrobial peptides revealed highly complex regulatory patterns and induction of several signal transduction pathways. The results suggest significant overlap between different stress regulons and interdependence of signal transduction pathways mediating stress responses.


Abbreviations: ECF, extracytoplasmic function; PLL, poly-L-lysine; TCS, two-component system


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Our purpose in this study was to characterize stress responses of Bacillus subtilis to cationic antimicrobial peptides. Antimicrobial peptides represent an ancient form of weapon in host defence mechanisms and their ubiquitous existence in cells and organisms of all types suggests important roles for them in innate immunity and defence against microbial invasion (reviewed by, for example, Yeaman & Yount, 2003). Natural antimicrobial peptides are typically amphipathic and positively charged, and they contain well-defined {alpha}-helical or {beta}-sheet structures (see also below). Mammalian antimicrobial peptides include defensins, protegrins and cathelicidins. Antimicrobial peptides typically attach to membrane surfaces of invading pathogens and disturb membrane integrity (Zasloff, 2002). In the case of nisin and epidermidin, however, a specific target has been identified. They interact with the membrane-bound peptidoglycan precursors and disturb cell wall biosynthesis (Breukink et al., 1999; Brotz et al., 1998). Modification of the net negative charge of the bacterial cell surface by adding positively charged residues to teichoic and lipoteichoic acids helps bacteria to avoid being killed by antimicrobial peptides. It has been shown that inactivation of the dlt operon of Staphylococcus aureus and consequent lack of D-alanine substitution in teichoic and lipoteichoic acids results in increased negative charge of the cell surface and increased sensitivity to defensins, protegrins and other antimicrobial peptides (Peschel et al., 1999). The absence of D-alanine substitution in the anionic polymers of Listeria monocytogenes also increases sensitivity to antimicrobial peptides and decreases virulence (Abachin et al., 2002).

Extracytoplasmic function (ECF) sigma factors are regulatory components by which bacteria control gene expression in response to environmental stress. There are seven different ECF-type sigma factors in the Gram-positive model bacterium B. subtilis (Helmann, 2002; Kunst et al., 1997). Several stress conditions activate the SigW regulon. These include alkaline shock (Wiegert et al., 2001), inhibition of the cell wall synthesis by antibiotics such as vancomycin and disturbance of the integrity of the cell membrane by detergents (Cao et al., 2002b). The SigW sigma factor is associated with the membrane-bound SigW anti-sigma factor when the Bacillus cell is not exposed to environmental stress (Schobel et al., 2004). Under stress conditions, the anti-sigma factor is proteolytically degraded, resulting in the release of SigW from the membrane and binding to gene promoters of the regulon (Schobel et al., 2004). A similar pattern is anticipated for other ECF-type sigma factors. SigM is required for combating stress due to antibiotic effects on the cell wall, ethanol, heat, acid and superoxide (Thackray & Moir, 2003). It is also essential for survival in environments containing high concentrations of salt (Horsburgh & Moir, 1999), suggesting that it is required for maintaining the integrity of the cell envelope. Alternative sigma factors are not the only regulatory systems that are involved in stress tolerance: two-component systems (TCSs) also have a role in controlling gene expression in environmental changes. TCSs are signalling devices composed of a membrane-bound sensor kinase and a response regulator. B. subtilis two-component regulation has recently been reviewed (Ogura & Tanaka, 2002).

We studied stress responses to two naturally occurring antimicrobial peptides, LL-37 and PG-1, and their synthetic analogue poly-L-lysine (PLL), using a DNA macroarray and real-time RT-PCR. Human LL-37 is 37 amino acid residues long and belongs to the cathelicidin family of antimicrobial peptides (Johansson et al., 1998; Turner et al., 1998). It is an amphipathic and {alpha}-helical peptide that probably disrupts the lipid bilayer by a toroidal pore mechanism (Henzler Wildman et al., 2003; Johansson et al., 1998). The porcine protegrin PG-1 is composed of 18 amino acid residues, including four cysteines, and forms a two-stranded antiparallel {beta}-sheet linked by a {beta}-turn (Aumelas et al., 1996; Fahrner et al., 1996). The four cysteines of PG-1 form two disulphide bonds, which are important for the {beta}-sheet conformation and antimicrobial activity (Harwig et al., 1996). The synthetic peptide PLL differs from the natural peptides in that it is not amphipathic. Its mode of action on membranes is unclear.

It was found that the antimicrobial peptides induced ECF-type sigma factor regulons in a complex manner. Several genes that are regulated by two-component signal transduction systems were also induced. Most interestingly, LL-37 strongly upregulated, via the YxdJK TCS, the yxdLM genes encoding an ABC-type transporter of unknown function. The yvcRS and bceAB (ytsCD) genes, which encode ABC transporters highly homologous with YxdLM, were also moderately upregulated. Interestingly PG-1, PLL and Triton X-100 did not induce the expression of any of these ABC-transporter genes.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial growth conditions.
B. subtilis cells were grown in Luria–Bertani (LB) medium or in BFA minimal medium at 37 °C with vigorous shaking. The BFA medium is a modified Spizizen's minimal salts medium (Anagnostopoulos & Spizizen, 1961) containing glutamine instead of ammonium sulphate as the nitrogen source. When needed, kanamycin and erythromycin were added at concentrations of 10 µg ml–1 and 1 µg ml–1, respectively. B. subtilis strain 168 and its derivatives were used in all experiments. The sensitivity of strains to four antimicrobial peptides, HNP-1 (American Peptide Company), PG-1 (Bachem), LL-37 (Ale Närvänen, University of Kuopio, Finland) and PLL (Sigma-Aldrich) were tested with the Bioscreen C Microbiology Reader (Labsystems). The culture volume was 150 µl and the number of bacteria in the inoculum was approximately 106 ml–1. These tests were also repeated in shake-flask cultures at approximately 108 bacteria (ml inoculum)–1.

Mutant constructions.
The sigW : : neo and sigM : : pMUTIN4 mutations were introduced into strains by transformation with chromosomal DNA of the B. subtilis strains HB4247 (kindly supplied by J. D. Helmann) and MJH003 (Horsburgh & Moir, 1999), respectively. The sigW sigM double mutant was constructed by transforming the strain IH8342 containing the sigW : : neo mutation with the chromosome of MJH003. The yvcQ gene was inactivated with pMUTIN4, as described by Vagner et al. (1998). The yxdJ and yxdK null mutations were obtained from Naotake Ogasawara (Nara Institute of Science and Technology, Nara, Japan).

RNA isolation, labelling with 33P and DNA macroarray analysis.
For RNA isolations, strains were grown in the BFA minimal medium containing 100 mM NaCl in shake-flask cultures. Cell densities were measured with a Klett colorimeter. Antimicrobial peptides were added at 60 Klett units and samples for RNA isolation were harvested after 20 min from 4 ml cell culture by centrifugation. Control samples without peptides were treated in a similar manner. Cells were resuspended in 400 µl ice-cold culture medium and transferred to screw-capped Eppendorf tubes containing 1·5 g glass beads, 50 µl 10 % SDS, 50 µl 3 M sodium acetate and 500 µl phenol/chloroform/isoamylalcohol (25 : 24 : 1 by vol.). The tubes were frozen in liquid nitrogen, followed by vigorous shaking for 6 min with a face-grinding machine and centrifugation at 10 000 r.p.m. for 5 min. The water phase was mixed (Vortex) with 1 vol. chloroform and centrifuged at 14 000 r.p.m. for 2 min. Next, the water phase was mixed with 2 vols Roche lysis/binding buffer, and the RNA extraction was continued with the Roche High Pure RNA Isolation Kit according to the manufacturer's instructions.

DNA macroarray analysis was carried out using Panorama B. subtilis gene array filters and specific cDNA labelling primers (Sigma Genosys). The Panorama B. subtilis gene array contains duplicate spots of PCR products representing currently known B. subtilis genes. Prior to cDNA synthesis, the quality of RNA was confirmed using Northern blotting. For cDNA synthesis, 10 µg RNA was used, and the synthesis was performed as described by Wiegert et al. (2001). The SuperScript II reverse transcriptase was purchased from Gibco-BRL. cDNA was purified with MicroSpin G-25 columns (Amersham Pharmacia Biotech) and the labelling efficiency was determined with a liquid scintillation counter. Prehybridization, hybridization and washing of the filters were performed according to the manufacturer's instructions. The DNA array filters were exposed overnight on phosphor screens and the screens were scanned with a Fluorescent Image Analyser FLA-2000 (Fujifilm). Hybridization signal intensities were quantified with the ArrayVision software (Imaging Research), as described by Wiegert et al. (2001). Data were filtered to avoid false positives by excluding genes with a signal-to-noise ratio <3 (Array Vision software) and normalized by dividing the intensity of each spot by the mean intensity of all the spots. Each experiment was carried out twice with RNA isolated from two independent cultures. Genes were regarded as induced when the induction ratio was >2 in both experiments.

Quantitative real-time RT-PCR.
For real-time RT-PCR, RNA was isolated similarly as for the DNA array. RT reactions were carried out with the Omniscript Reverse Transcriptase Kit (Qiagen) according to the manufacturer's instructions with the exception of an additional DNase I (Roche) treatment. An equal amount of RNA (2 µg) was used in each RT reaction. Primers used in RT reactions were random hexamers (0·15 µg ml–1) provided by Roche. The absence of chromosomal DNA in the RNA preparations was verified with a control sample that was not treated with RT, but was otherwise treated in a similar manner to the RT-treated samples. Real-time PCR reactions were carried out with specific primer pairs using the SYBR green PCR master mix (Applied Biosystems). Primers were designed with the Primer Express software (Applied Biosystems) and purchased from Sigma Genosys or TAGC Copenhagen. Sequences of the PCR primers for the genes studied are shown in Table 1. The amplification and detection of PCR products were performed with the ABI PRISM 5700 sequence detection system (Applied Biosystems). The cycling conditions were: 1 cycle at 50 °C for 2 min, 1 cycle at 95 °C for 10 min, 40 cycles at 95 °C for 15 s and at 60 °C for 1 min. The threshold cycle (Ct) is the first cycle at which the fluorescence becomes detectable above the background and is inversely proportional to the logarithm of the initial number of template molecules. Ct values of known quantities of B. subtilis chromosomal DNA were plotted for each primer pair to obtain standard curves. The standard curves allowed us to convert the Ct values of each amplified gene in the cDNA preparations to relative numbers of cDNA molecules. These cDNA values were normalized with the value of gyrA, which was constant in different growth conditions and phases (data not shown).


View this table:
[in this window]
[in a new window]
 
Table 1. Primers used in real-time PCR analyses

 

   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Optimization of experimental conditions for the transcriptome analysis
Four different types of peptides were chosen to study the response to antimicrobial peptides. Human cathelicidin LL-37 and porcine protegrin PG-1 are cathelin-associated {alpha}-helical and {beta}-sheet structures, respectively (Turner et al., 1998), and human defensin HNP-1 belongs to {alpha}-defensins with a triple-stranded {beta}-sheet structure (Lehrer & Ganz, 2002). PLL was chosen as a synthetic analogue of cationic peptides (Vaara & Vaara, 1983). These antimicrobial peptides were first used at various concentrations in a Bioscreen assay to determine the range of concentrations needed to inhibit the growth of B. subtilis. Then, the effects of peptide concentrations that inhibited growth but did not kill bacteria were determined in shake-flask cultures (Fig. 1). Bacteria were grown in BFA minimal medium containing an additional 100 mM NaCl to enhance the microbicidal effect of the peptides (Turner et al., 1998). No effect on growth was detected with HNP-1 for any of the tested concentrations (up to 9 µg ml–1) and it was omitted from further analyses. For RNA isolations, bacteria were cultured in shake flasks and the antimicrobial peptides were added at 60 Klett units at concentrations of 1·5 µM (LL-37), 50 nM (PG-1) or 1 mM (PLL).



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 1. The cationic peptides LL-37, PG-1 and PLL inhibit the growth of B. subtilis. Bacteria were grown in BFA minimal medium in shake flasks, and culture densities were measured with a Klett colorimeter. The cationic peptides were added at the cell density of 60 Klett units (arrow).

 
Cationic antimicrobial peptides induce a subset of SigW- and SigM-regulated genes
We studied the effects of the antimicrobial peptides on gene expression in B. subtilis using DNA macroarrays containing all the ORFs of the B. subtilis genome. Bacteria were exposed to peptide stress for 20 min, after which total RNA was isolated and the DNA macroarray analysis was carried out. In parallel, control cultures without a peptide addition were similarly treated. Two independent array experiments from separate cultures with each peptide treatment were performed. The array data were analysed with the ArrayVision and Microsoft Excel programs. Genes with a twofold induction ratio or higher and signal-to-noise ratios >3 in both independent experiments were considered to be induced.

Altogether, the LL-37 treatment induced 96 genes (Table 2, only the first gene of an operon is listed), including several genes that are regulated by the SigW and SigM ECF sigma factors (Huang et al., 1999; Cao et al., 2002a; Thackray & Moir, 2003; Asai et al., 2003). Of the 30 verified promoters of the SigW regulon (Cao et al., 2002a), only 10 were expressed at elevated levels (greater than twofold) in LL-37-treated cells, with the fold-induction ratios ranging from 2·4 to 14·7, suggesting that LL-37 is a weak SigW inducer. The promoters that are directly regulated by SigM are less well known, but it was observed that 15 candidate promoters of the SigM regulon (Asai et al., 2003), including sigM itself, were upregulated by LL-37 (Table 2). Many of the induced genes are involved in extracytoplasmic functions such as synthesis of the cell wall. These genes included pbpE, encoding a penicillin-binding protein (4·5-fold induction), wapA, encoding a cell wall-associated protein (2·5-fold induction), murG, involved in cell wall formation (6·1-fold-induction) and maf, required for septum formation (2·5-fold induction). The upregulated genes also included bcrC (ywoA), which is dependent on several sigma factors and is required for bacitracin resistance (Cao & Helmann, 2002) (4·1-fold induction), and the gene encoding the penicillin-binding protein ponA (2·3-fold induction). The most strongly induced gene was liaI (yvqI) (Mascher et al., 2004), which was induced 58-fold (Table 2); other genes of the liaIHGFSR (yvqIHGFEC) gene cluster were also upregulated, but to a lesser extent (not shown). liaI (yvqI) is not known to be dependent on any ECF sigma factor. Furthermore, yrhH (14·7-fold), encoding a putative methyltransferase, yxdL (22·7-fold) and yhcG (14·7-fold), encoding putative ATP-binding components of ABC transporters, and yoeB (9·1-fold), encoding a putative exported protein of unknown function, were strongly upregulated.


View this table:
[in this window]
[in a new window]
 
Table 2. Genes induced by LL-37, PG-1 and PLL

 
The second natural peptide, PG-1, induced 58 genes. In a similar manner to LL-37, PG-1 also activated a subset of the genes of the ECF regulons (Table 2). Twelve genes of the SigW regulon, for which there are 30 verified promoters, and five genes predicted to belong to the SigM regulon (Asai et al., 2003) were upregulated. The gene induction pattern resembled that of LL-37, but the induction ratios were lower.

Genes that were induced by both peptides were araE, bcrC (ywoA), dltB, liaI, murG, pbpE, pspA, spoOM, wprA, yceC, yeaA, yjbC, yoeB, yqeZ and yuaG (Table 2). Interestingly, the genes of the yxdLM operon, yrhH and yhcG, which were highly induced by LL-37, were not induced by PG-1 (see also below).

PLL was also an ECF inducer and the gene induction pattern resembled that of the other antimicrobial peptides, but characteristic differences were also observed (Table 2). Among 86 upregulated genes, there were 23 and 8 genes that belonged to the SigW and SigM regulons, respectively. Thus, compared to the response to LL-37 (above), it seems that PLL is clearly a stronger activator of the SigW regulon. There were also several genes that were induced at high levels by PLL, but not at all by LL-37 or PG-1, including csfB (6·6-fold), yaaN (5·1-fold), yfhL (5·6-fold), yocA (4·5-fold) and yrzI (12·4-fold), encoding proteins of unknown function. In addition, genes involved in purine, pyrimidine and ribosomal protein synthesis were strongly induced, a phenomenon not seen with the natural peptides. Furthermore, some genes that were induced at high levels by either or both of the natural peptides were not induced by PLL, notably liaI (yvqI) and other genes of the lia (yvq) cluster, yxdL, which was strongly induced by LL-37, and yhcG and yoeB.

Decreased expression of several genes was also observed. However, the experimental setup, short time of exposure to the peptides, and very different degradation rates of mRNAs hampered the interpretation of these results and they were not analysed in this study.

Cross-talk in signal transduction pathways mediating stress responses induced by a cationic peptide
We also carried out the DNA array analysis with sigM and sigW knockout mutants using LL-37 for the induction. The gene induction patterns of the sigma mutants and wild-type strain were compared in scatter plots (Fig. 2). The LL-37 treatment elicited a significantly lower number of induced genes in both sigma mutants than in the wild-type. The genes induced in the sigM and sigW mutants are listed in Table 3. This reduced stress response was not due to decreased stress in the sigma mutants, since the effective concentration of LL-37 causing the growth inhibition was the same in all three strains (see below). In the sigM mutant, as expected, the genes of the SigM regulon were not induced. Surprisingly, the array data also showed either that the SigW-regulated genes were not induced or that their induction levels were clearly lower than in the wild-type strain. Analogous results were obtained with the sigW mutant. Interestingly, numerous genes which are not known to belong to these two sigma factor regulons were also induced in a SigM- or SigW-dependent manner. These results suggest significant cross-talk between the SigM and SigW regulons, and some other regulon(s) responding to LL-37. Furthermore, a significant observation was the clearly higher level of expression of several genes in the sigW mutant compared to that of the wild-type (Fig. 2b). This set included (Table 3) genes encoding endo-1,4-{beta}-glucanase (bglC; 4·6-fold), a protein synthesizing {alpha}-1,4-glucan using ADP-glucose (glgA; 5·1-fold) and 6-phospho-{alpha}-glucosidase (glvA/malA; 3·9-fold).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 2. Comparison of the induction ratios of the wild-type strain and the sigM (a) and sigW (b) mutants treated with LL-37. The genes induced in the wild-type and/or a sigma mutant by LL-37 are highlighted (closed triangle, open circle or square). The genes belonging to the SigM or SigW regulons are marked with open circles and squares, respectively. The induction ratios of other genes are marked with grey circles and include some high induction ratios due to the use of non-filtered data in this graphical comparison; the filtering (see Methods) of data eliminated these high induction ratios. WT, wild-type.

 

View this table:
[in this window]
[in a new window]
 
Table 3. Induction of gene expression in cells of the sigW and sigM mutants treated with LL-37

 
In order to study the overlap between the stress regulons further, we determined the expression levels of some of the induced genes by real-time RT-PCR. The liaI and yxdL genes were chosen due to their strong induction with LL-37. The yuaG gene (in the yuaFGI operon) was induced most strongly by PG-1. It encodes a putative flotillin known to belong to the SigW regulon (Huang et al., 1999; Cao et al., 2002a). Other chosen genes included racX (in an operon with pbpE), a SigW-dependent gene (Cao et al., 2002a), and yjbC and bcrC (ywoA), which belong to the SigM regulon (Cao & Helmann, 2002; Ohki et al., 2003a). The yjbC gene is also regulated by SigW and SigX (Cao et al., 2002a; Ohki et al., 2003a; Thackray & Moir, 2003). Furthermore, the expression of radC (in an operon with maf), which may be regulated by SigM (Asai et al., 2003), and of wprA and ypuA, which may be expressed independently of the ECF sigma factors (Table 2), was determined. The RT-PCR analysis was carried out with the wild-type strain and the sigM and sigW mutants. A sigM sigW double mutant was also used to analyse the expression of some genes. Bacteria were treated in a similar manner to that employed in the array experiments with LL-37 or PG-1. Since the array analysis revealed that the induction of radC and ypuA with PG-1 was low, their induction with this peptide was not determined. Samples were collected for analysis at two different time points, 10 and 20 min after the addition of the antimicrobial peptides. All mRNA measurements were performed two to four times.

In the wild-type strain treated with LL-37, liaI and yxdL were the most highly induced genes (Table 4), as was also the case in the array analysis. Higher induction ratios were seen in the 10 min samples than in the 20 min samples, indicating that the induction was fast and transient. As shown in Fig. 3, expression was dependent on the dosage of LL-37, as demonstrated with the liaI gene (Fig. 3a), and decreased (also that of liaH) from the maximal level (10 min time point) back to the uninduced level in about 2 h (Fig. 3b). Consistent with the induction of the liaIHGFSR gene cluster, the LiaH protein appeared in the proteome of cytoplasmic proteins, as demonstrated by two-dimensional gel electrophoresis and spot identification by matrix-associated laser desorption ionization–time of flight mass spectrometry (MALDI-TOF) (data not shown). The other genes were induced to a lesser extent, and in only half of them was the induction transient (Table 4). The normalized mRNA levels (not shown) indicated that liaI, yxdL and yuaG were expressed at a very low level in non-treated cells (basal expression level). The basal expression level of wprA, yjbC, ypuA and radC was fairly high, and bcrC (ywoA) and racX were expressed at intermediate levels. The genes that were expressed at a low level in non-treated cells exhibited the strongest induction in peptide-treated cells. Normalized mRNA levels of the genes varied approximately twofold from one experiment to another and between non-treated wild-type and ECF sigma mutant cells (not shown). Considerable experimental variation was especially observed in the induction ratios of the liaI and yxdL genes (Table 4). The RT-PCR displayed clearly higher induction ratios than the DNA array, but there was a good overall consistency of the induction pattern in these two types of assay.


View this table:
[in this window]
[in a new window]
 
Table 4. The dependency of a set of LL-37-induced genes on SigM and SigW

 


View larger version (12K):
[in this window]
[in a new window]
 
Fig. 3. Dependence of liaI expression on the concentration of LL-37 and kinetics of LL-37-induced expression of liaI and liaH. Bacteria were grown in BFA minimal medium. (a) The induction of liaI in cells stressed with various concentrations of LL-37. (b) LL-37 (1·5 µM) was added at the cell density of 60 Klett units and the expression of liaI (closed triangle) and liaH (open square) was determined by RT-PCR at various time points of the LL-37 treatment. The induction ratio is fold induction in LL-37-treated cells compared to that in non-treated cells.

 
The liaI gene, which was induced 1530-fold in the wild-type (10 min time point) with LL-37, was also upregulated in the sigM and sigW mutants, but in both of them the induction was clearly lower, 198- and 287-fold, respectively (Table 4). In DNA arrays, the difference in the induction ratios was more dramatic, 58-fold in the wild-type and twofold in the mutants. The other highly induced gene, yxdL, was also upregulated significantly less in the sigM (248-fold) and sigW (347-fold) mutants compared to the wild-type strain (704-fold), consistent with the array results above. Only a minor additive effect on liaI and yxdL expression was observed in the double mutant (Table 4). The yuaG (yuaFGI) gene was induced 18-fold (10 min time point), whereas no induction was seen in the sigW mutant, consistent with its previously verified SigW dependency (Cao et al., 2002a) (see also below). The induction of yuaG was also reduced in the sigM mutant (sevenfold at the 10 min time point), suggesting moderate SigM dependency. The yjbC and bcrC (ywoA) genes were induced in the wild-type strain 6–11-fold, whereas in the sigM sigW double mutant almost no induction was seen (Table 4), indicating that these genes are induced in an ECF-dependent manner. However, the inactivation of single ECF sigma factors reduced the induction ratios only moderately or not at all. LL-37 induced yjbC expression three- to fivefold in the sigM mutant and six- to ninefold in the sigW mutant. The induction ratios of bcrC were 3–6 (sigM) and 6–9 (sigW). racX induction was affected by both sigma mutations (see also the effects of PG-1 and Triton X-100 on racX expression below). These results suggest that both SigM and SigW regulate either directly or indirectly the yjbC, bcrC (ywoA) and racX genes. LL-37 upregulated the expression of radC, wprA and ypuA four- to 12-fold in the wild-type. Reduced induction ratios of radC and ypuA in the sigM mutant suggest that their induction by LL-37 is at least partially mediated by SigM (Table 4). The wprA gene exhibited induction patterns that rather suggest the independence of the sigma factors (see also the effects of PG-1 and Triton X-100 on wprA expression below).

The induction ratios of all genes studied were significantly lower in PG-1-treated cells than in LL-37-treated cells (Table 5), suggesting that the stress caused by the PG-1 treatment was less severe; for example, the induction ratio of liaI with LL-37 was several-fold higher than with PG-1. Despite the lower induction level in PG-1-treated cells, the sigM and sigW mutations reduced the induction of liaI as in LL-37-treated cells (Table 5). The concentration of PG-1 used induced the yjbC and bcrC (ywoA) genes very weakly in all three strains. PG-1 treatment induced yuaG expression in a similar manner to LL-37 treatment, and the strong SigW dependency and moderate SigM dependency of yuaG were also observed with PG-1 (Table 5). In the wild-type, racX was induced four- to ninefold, whereas no induction was observed in the sigW mutant, suggesting SigW-dependent regulation of racX. In the sigM mutant, racX was induced two- to sixfold by PG-1. The inactivation of the sigma factors did not impair the induction of wprA, consistent with an induction mechanism that is independent of SigW and SigM.


View this table:
[in this window]
[in a new window]
 
Table 5. Induction of a set of genes by PG-1 and their dependence on SigM and SigW

 
Triton X-100 induces sigma regulons, but in a pattern different from that of antimicrobial peptides
In order to find out whether the above genes are also induced by a membrane-disrupting agent with no presumed specificity, we treated cells with 0·005 % Triton X-100 and determined expression levels of the genes in the wild-type strain and the ECF sigma factor mutants by real-time RT-PCR. The effects of cationic peptides on membranes may be somewhat different from those of detergents (Henzler Wildman et al., 2003), although the opposite view has also been put forward (Oren et al., 1999).

A similar induction of liaI expression (790-fold at the 10 min time point) and slightly reduced induction ratios in the sigM and sigW mutants, as with LL-37, were observed (Table 6). The strong dependency of yuaG on SigW was also demonstrated with the detergent, but the moderate SigM dependency was not observed. The induction of three other genes (radC, ypuA and bcrC) was also partially dependent on both sigma factors. radC was expressed in the wild-type strain at two- to fivefold, ypuA at two- to threefold and bcrC (ywoA) at threefold higher levels than in the sigma mutants (Table 6). In the wild-type strain and sigM mutant, a clearly stronger induction of racX was observed when cells were treated with Triton X-100 (19–61-fold) than when they were treated with the antimicrobial peptides (four- to ninefold). In a similar manner to that observed with PG-1, no induction was seen in the sigW mutant, consistent with the SigW dependency of racX. The sigma mutations did not impair the induction of yjbC in Triton X-100-treated cells, in contrast to LL-37-treated cells. Furthermore, the wprA gene was not induced by Triton-X-100.


View this table:
[in this window]
[in a new window]
 
Table 6. Induction ratios in cells treated with Triton X-100

 
Inactivation of SigM and SigW does not increase the sensitivity of B. subtilis to antimicrobial peptides
The sensitivity of sigW and sigM mutants and the sigM sigW double mutant to LL-37, PG-1 and PLL was tested. Growth experiments were performed in shake-flask cultures in the same conditions as in the array experiments. The growth was also studied with several peptide concentrations with Bioscreen. No significant difference in the sensitivity of the sigW, sigM or sigM sigW mutants to the antimicrobial peptides was detected and the growth arrest caused by the antimicrobial peptides was similar to that of the wild-type. However, it was obvious that the sigW and sigM sigW mutants grew slower than the wild-type (in the absence of peptide) in the BFA minimal medium containing 100 mM NaCl, indicating the importance of the SigW regulon in these conditions (data not shown). The growth of the sigM mutant resembled that of the wild-type.

yxdL gene expression is specifically induced by LL-37
A highly interesting observation was that the yxdL gene, which was strongly induced by LL-37 (704-fold), was not induced by PG-1 (Table 5). The lack of induction was seen in both the DNA array and RT-PCR analyses, and it was true for both wild-type and sigM and sigW mutant cells. The yxdL gene was not induced by Triton X-100 either (Table 6). These results suggest that there is a strong specificity in the induction mechanism of yxdL.

LL-37 causes upregulation of three paralogous ABC-transporter genes via TCS-mediated signalling dedicated to the regulation of transporter expression
The yxdL gene encodes the putative ATP-binding component of an ABC transporter of unknown function. The downstream gene yxdM, which most probably forms an operon with yxdL, encodes the permease component of the ABC transporter. Immediately upstream from yxdLM there are the yxdJ and yxdK genes, which encode the components of a TCS of unknown function (see the organization of the gene cluster in Fig. 4). The yxeA gene, which encodes a conserved protein of unknown function, is located downstream from the yxdM gene and most probably belongs to the same operon as yxdLM (not shown in Fig. 4). The DNA array data also revealed that LL-37 strongly induced the yxeA gene (36-fold) in a similar manner to yxdLM, consistent with the operon organization. A sequence similarity search revealed that yxdL and yxdM are highly homologous with the corresponding genes of two other ABC transporters of B. subtilis, BceAB (formerly YtsCD) and YvcRS (Fig. 4; see also Joseph et al., 2002; Mascher et al., 2003; Ohki et al., 2003b). Interestingly, genes encoding TCSs are also located in the immediate upstream regions of the bceAB and yvcRS genes in a pattern similar to that of the yxdLM region. It has been shown that the BceRS TCS is able to sense extracellular bacitracin and induce the expression of the BceAB ABC transporter, which confers resistance to bacitracin (Mascher et al., 2003; Ohki et al., 2003b). The YxdLM ABC transporter as well as the YxdJK TCS exhibit homology with the corresponding proteins of the Yvc and Bce systems. The homology is highest between the ATP-binding components of the ABC transporters (about 50 % identity), and, in the following order, is less between the response regulators of TCS, the sensor kinases of TCS and the permease components of the ABC transporters (Fig. 4). The Yxd proteins exhibit slightly higher similarity to the Yvc proteins than to the Bce proteins.



View larger version (12K):
[in this window]
[in a new window]
 
Fig. 4. Organization of three pairs of genes encoding homologous ABC transporters and in their immediate upstream region three similar pairs of genes encoding homologous TCSs. The identity values of the amino acid sequences of the deduced protein products are indicated. a, Response regulator of TCS; b, sensor histidine kinase of TCS; c, ATP-binding component of ABC transporter; d, permease component of ABC transporter.

 
The DNA array analysis showed that the yvcR and yvcS genes were upregulated 3·5- and sixfold in LL-37-treated cells, respectively, suggesting that this antimicrobial peptide may affect not only yxdL expression but also its paralogues. The regulation of the three paralogous ABC transporter genes was studied by determining their expression in the wild-type strain and mutants of the upstream histidine kinase genes (yxdK, yvcQ and bceS) treated with LL-37 using real-time RT-PCR. In the wild-type strain, the yxdM gene was induced strongly (about 300-fold at the 10 min time point) in a similar manner to the yxdL gene (Table 7), further indicating that these genes form an operon (together with yxeA). Since no induction of the upstream yxdJ gene was observed, the TCS genes must belong to a different transcriptional unit than yxdLM, in a manner similar to that by which the bceRS genes are transcribed separately from bceAB (Joseph et al., 2002; Ohki et al., 2003b). In the yxdK mutant, hardly any induction of yxdL and yxdM was seen, indicating that the YxdJK TCS regulates the expression of the yxdLM operon (Table 7). The induction of yxdL was also slightly reduced (about 25 %) in both the yvcQ and the bceS mutant, suggesting cross-regulation between the systems; in other words, the paralogous TCSs also regulate the promoters of the non-cognate paralogous ABC-transporter genes. The yvcR gene was induced by LL-37, but only about fivefold, consistent with the DNA array result for yvcRS induction. The inactivation of the upstream sensor kinase gene yvcQ only partially abolished the induction of yvcR (2·2-fold induction at the 10 min time point). Similar partially abolished induction was also seen in the yxdK mutant (3·4-fold), consistent with cross-regulation (Table 7). In the bceS mutant, the induction of yvcR was comparable to that of the wild-type. LL-37 did not induce yvcP expression, indicating that yvcPQ and yvcRS are different transcriptional units. The DNA array results suggested that the bceAB operon is not induced by LL-37. The RT-PCR analysis, however, revealed that bceA is also induced about sixfold in LL-37-treated cells (Table 7). The bceS mutation partially abolished the induction (3·3-fold), but the yvcQ and yxdK mutations did not affect it.


View this table:
[in this window]
[in a new window]
 
Table 7. Expression regulation and cross-regulation of the yxdLM operon and its homologues yvcRS and bceAB via the YxdK, YvcQ and BceS TCSs

 
In order to find out whether or not the YxdLM or YvcRS ABC transporters have a role in peptide resistance, the sensitivity of yxdL and yxdL yvcR mutants to LL-37 was tested. Neither the single mutant nor the double mutant showed increased sensitivity to LL-37.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In this study, we carried out genome-wide transcription analyses of the stress responses of B. subtilis to the cationic antimicrobial peptides LL-37 and PG-1, which are natural peptides, and PLL, a synthetic peptide. The responses to each peptide were highly complex, including activation of several signal transduction pathways.

The antimicrobial peptides induced expression of 96 (LL-37), 58 (PG-1) and 86 (PLL) genes in B. subtilis. In this complex response, some patterns were recognized. A prominent feature was a high proportion of induced genes belonging to the SigW and SigM ECF-type sigma factor regulons. However, only subsets of these sigma factor regulon genes were induced by the antimicrobial peptides. The non-amphipathic PLL was the most effective peptide in activating the SigW regulon, as evidenced by the induction of 23 out of the 30 verified promoters of the regulon [see Cao et al. (2002a) for the SigW regulon]. The amphipathic peptides LL-37 and PG-1 upregulated only about one-third of the 30 SigW-regulated genes. In a similar manner, only subsets of putative SigM-regulated promoters were induced in peptide-treated cells. These results suggest that SigW and SigM are involved in the stress responses to the antimicrobial peptides. However, the high numbers of induced genes that are expressed independently of SigW and SigM suggest that probably several other signal transduction pathways and regulators also mediate the stress responses.

It has been shown that the ECF sigma factor regulons are partly overlapping (Cao et al., 2002b; Huang et al., 1998). Nevertheless, the effect of sigW or sigM mutation on the number of induced genes in cells treated with LL-37 was striking. Not only the genes belonging to the mutated sigma regulon but also those under the control of other ECF sigma factors and those expressed independently of these sigma factors were poorly induced after peptide treatment. This phenomenon may be partly due to the increased basal expression level of several genes in the sigW and sigM mutants. Consequently, the additional stress of cationic peptides may not have caused further induction. Yet this alone does not explain why so few genes were induced in the sigma factor mutants, since the basal expression level was elevated in the sigma mutants only in the case of about 30 % of the genes induced in the wild-type. The RT-PCR analysis revealed decreased induction ratios for bcrC, liaI, radC, racX, ypuA and yxdL in both the sigma mutants. However, it is apparent that not all the affected genes are directly regulated by the sigma factors. It has been shown that liaI is regulated by the LiaRS (YvqCE) TCS (Mascher et al., 2003; H.-L. Hyyryläinen and others, unpublished results). The effects of ECF sigma factor mutations on liaI expression in cells treated with alkaline (Wiegert et al., 2001) or cationic antimicrobial peptides (this study) are most probably indirect. It is possible that the inactivation of one ECF sigma factor results in disturbance of the sensory function of other membrane-associated stress sensors. The similarity of the lethal doses of the cationic peptides in the sigma factor mutants and the wild-type, however, suggests that the degree of stress in these strains was the same. These results suggest that the stress response to cationic antimicrobial peptides is very complex. The functional overlap of several sigma factors and other types of regulators may also explain why the inactivation of SigW, SigM or both of them did not make cells sensitive to antimicrobial peptides.

PLL is expected to interact with the negatively charged cell wall and head groups of the membrane phospholipids (carpet or detergent-like mechanism; Yeaman & Yount, 2003), but may not penetrate deeper into the membrane interior. LL-37 and PG-1 penetrate into the membrane and disturb its integrity by forming pores (Henzler Wildman et al., 2003; Oren et al., 1999; Yang et al., 2000). We hypothesize that SigW-regulated promoters are activated by antimicrobial peptides by the interaction of the latter with the cell membrane surface and the cell wall, rather than by deeper effects inside the membrane. This conclusion is consistent with the strong induction of the SigW regulon by cell wall antibiotics (Cao et al., 2002b).

The araE, bcrC (ywoA), dltB, pbpE, pspA (ydjF), yceC, spo0M, yeaA, yjbC, yqeZ and yuaG genes were induced by all three peptide treatments. Each of these genes belongs to at least one ECF sigma factor regulon (Table 2). Some of these genes are involved in interactions with antimicrobial compounds interfering with the cell wall or membrane: pbpE encodes a penicillin-binding protein, yceC is similar to the tellurium resistance proteins, and bcrC (ywoA) encodes a bacitracin permease (Cao & Helmann, 2002; Podlesek et al., 1995). The dlt operon including dltB is involved in the D-alanine esterification of lipoteichoic and wall teichoic acids (Perego et al., 1995), which increases bacterial resistance to cationic antimicrobial peptides (Peschel et al., 1999; Cao & Helmann, 2004).

The DNA array and real-time RT-PCR analyses revealed that not only the ECF sigma factors but also TCSs have a major role in sensing antimicrobial peptides. Most importantly, LL-37 induced the genes of three ABC-type transporters; yxdLM was induced strongly (about 700-fold) and its close homologues yvcRS and bceAB were induced moderately (about sixfold). All these ABC-transporter genes are regulated by TCSs. The TCSs are encoded by genes in the immediate upstream regions of the ABC transporter genes, as evidenced by the lack/decrease of the induction in TCS mutants (Mascher et al., 2003; Ohki et al., 2003b; this study) or demonstrated by primer extension and DNase protection experiments (Pascale et al., 2004). In this study, the results suggested some low-level cross-talk between the three TCSs.

It has been shown that the expression of bceAB is induced more than 200-fold by bacitracin via BceRS TCS-mediated signalling (Mascher et al., 2003) and that the BceAB ABC transporter has a role in bacitracin resistance (Ohki et al., 2003b). In contrast, the YxdLM and YvcRS ABC transporters are of unknown function. These TCSs may be involved in sensing conditions inside the cell membrane (Mascher et al., 2003).

This study demonstrates for the first time an activator of the YxdJK TCS: LL-37. Our results also indicate that Triton X-100 and PLL do not activate YxdJK, suggesting distinctly different modes of action for LL-37 and detergents/detergent-like molecules, in contrast to what has been claimed (Oren et al., 1999). LL-37 is most probably a pore-forming peptide (Henzler Wildman et al., 2003), and the penetration of this amphipathic molecule into the membrane is probably crucial for the activation of the YxdJK TCS. The very short extracytoplasmic loop of the YxdK sensor is consistent with the conclusion that YxdJK senses signals inside the membrane, possibly by direct interaction with LL-37, and not signals on the membrane surface or cell wall, or membrane disturbance as such.

PG-1 did not activate YxdJK either, being a pore-forming peptide (Yang et al., 2000), suggesting that the pore formation may not be required for the activation of YxdJK. It may be essential to YxdJK activation that a peptide interacts directly and appropriately with the YxdK sensor in the membrane. In contrast, the LiaRS (YvqCE) TCS, which regulates liaIHG expression (Mascher et al., 2004), was strongly activated by both peptides (LL-37 and PG-1) as well as by Triton X-100. Several other stress treatments, such as alkaline shock (Wiegert et al., 2001), vancomycin (Cao et al., 2002b) and secretion stress (H.-L. Hyyryläinen and others, unpublished results) also activate LiaRS (YvqCE). This further confirms that YxdJK senses a narrow range of signals or peptides, while LiaRS broadly senses various stress conditions. Interestingly, however, PLL did not activate LiaRS. These peptide ligands of known structure with differences in their specificity give an excellent future opportunity to study the structure–function relationships of these TCSs.

In addition to yxdLM and liaIH (yvqIH), some other genes of unknown function were also strongly induced by cationic antimicrobial peptides (DNA array). These included yrhH, which was induced by both LL-37 and PLL. The yrhH gene encodes a putative methyltransferase. yhcG was strongly upregulated by LL-37 (14·7-fold) and enhanced expression levels of several other genes of the yhc operon were also observed. yhcG encodes an ABC-transporter ATP-binding protein. The yhcH gene, which was induced 4·6-fold, also encodes a putative (second) ABC-transporter ATP-binding protein, and the yhcI gene, which was induced 5·1-fold, encodes a putative ABC-transporter permease homologous with bacitracin permeases. The putative roles of these ABC transporters in the removal of LL-37 from cells should be studied in the future.


   ACKNOWLEDGEMENTS
 
We thank the groups of Wolfgang Schumann and Anne Moir for the sigW : : neo (constructed in the laboratory of J. D. Helmann) and sigM : : pMUTIN4 mutants, respectively. This work was supported by grants from the European Union (QLK3-CT-1999-01455) and the Academy of Finland (72592/2000).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Abachin, E., Poyart, C., Pellegrini, E., Milohanic, E., Fiedler, F., Berche, P. & Trieu Cuot, P. (2002). Formation of D-alanyl-lipoteichoic acid is required for adhesion and virulence of Listeria monocytogenes. Mol Microbiol 43, 1–14.[CrossRef][Medline]

Anagnostopoulos, C. & Spizizen, J. (1961). Requirements for transformation of Bacillus subtilis. J Bacteriol 81, 741–746.

Asai, K., Yamaguchi, H., Kang, C. M., Yoshida, K., Fujita, Y. & Sadaie, Y. (2003). DNA microarray analysis of Bacillus subtilis sigma factors of extracytoplasmic function family. FEMS Microbiol Lett 220, 155–160.[CrossRef][Medline]

Aumelas, A., Mangoni, M., Roumestand, C., Chiche, L., Despaux, E., Grassy, G., Calas, B. & Chavanieu, A. (1996). Synthesis and solution structure of the antimicrobial peptide protegrin-1. Eur J Biochem 237, 575–583.[Abstract]

Breukink, E., Wiedemann, I., van Kraaij, C., Kuipers, O. P., Sahl, H. & de Kruijff, B. (1999). Use of the cell wall precursor lipid II by a pore-forming peptide antibiotic. Science 286, 2361–2364.[Abstract/Free Full Text]

Brotz, H., Josten, M., Wiedemann, I., Schneider, U., Gotz, F., Bierbaum, G. & Sahl, H. G. (1998). Role of lipid-bound peptidoglycan precursors in the formation of pores by nisin, epidermin and other lantibiotics. Mol Microbiol 30, 317–327.[CrossRef][Medline]

Cao, M. & Helmann, J. D. (2002). Regulation of the Bacillus subtilis bcrC bacitracin resistance gene by two extracytoplasmic function sigma factors. J Bacteriol 184, 6123–6129.[Abstract/Free Full Text]

Cao, M. & Helmann, J. D. (2004). The Bacillus subtilis extracytoplasmic-function {sigma}X factor regulates modification of the cell envelope and resistance to cationic antimicrobial peptides. J Bacteriol 186, 1136–1146.[Abstract/Free Full Text]

Cao, M., Kobel, P. A., Morshedi, M. M., Wu, M. F., Paddon, C. & Helmann, J. D. (2002a). Defining the Bacillus subtilis sigma (W) regulon: a comparative analysis of promoter consensus search, run-off transcription/macroarray analysis (ROMA), and transcriptional profiling approaches. J Mol Biol 316, 443–457.[CrossRef][Medline]

Cao, M., Wang, T., Ye, R. & Helmann, J. D. (2002b). Antibiotics that inhibit cell wall biosynthesis induce expression of the Bacillus subtilis sigma (W) and sigma (M) regulons. Mol Microbiol 45, 1267–1276.[CrossRef][Medline]

Fahrner, R. L., Dieckmann, T., Harwig, S. S., Lehrer, R. I., Eisenberg, D. & Feigon, J. (1996). Solution structure of protegrin-1, a broad-spectrum antimicrobial peptide from porcine leukocytes. Chem Biol 3, 543–550.[CrossRef][Medline]

Harwig, S. S., Waring, A., Yang, H. J., Cho, Y., Tan, L. & Lehrer, R. I. (1996). Intramolecular disulfide bonds enhance the antimicrobial and lytic activities of protegrins at physiological sodium chloride concentrations. Eur J Biochem 240, 352–357.[Abstract]

Helmann, J. D. (2002). The extracytoplasmic function (ECF) sigma factors. Adv Microb Physiol 46, 47–110.[CrossRef][Medline]

Henzler Wildman, K. A., Lee, D. K. & Ramamoorthy, A. (2003). Mechanism of lipid bilayer disruption by the human antimicrobial peptide, LL-37. Biochemistry 42, 6545–6558.[CrossRef][Medline]

Horsburgh, M. J. & Moir, A. (1999). Sigma M, an ECF RNA polymerase sigma factor of Bacillus subtilis 168, is essential for growth and survival in high concentrations of salt. Mol Microbiol 32, 41–50.[CrossRef][Medline]

Huang, X. & Helmann, J. D. (1998). Identification of target promoters for the Bacillus subtilis sigma X factor using a consensus-directed search. J Mol Biol 279, 165–173.[CrossRef][Medline]

Huang, X., Fredrick, K. L. & Helmann, J. D. (1998). Promoter recognition by Bacillus subtilis sigma W: autoregulation and partial overlap with the sigma X regulon. J Bacteriol 180, 3765–3770.[Abstract/Free Full Text]

Huang, X., Gaballa, A., Cao, M. & Helmann, J. D. (1999). Identification of target promoters for the Bacillus subtilis extracytoplasmic function sigma factor, sigma W. Mol Microbiol 31, 361–371.[CrossRef][Medline]

Johansson, J., Gudmundsson, G. H., Rottenberg, M. E., Berndt, K. D. & Agerberth, B. (1998). Conformation-dependent antibacterial activity of the naturally occurring human peptide LL-37. J Biol Chem 273, 3718–3724.[Abstract/Free Full Text]

Joseph, P., Fichant, G., Quentin, Y. & Denizot, F. (2002). Regulatory relationship of two-component and ABC transporter systems and clustering of their genes in the Bacillus/Clostridium group, suggest a functional link between them. J Mol Microbiol Biotechnol 4, 503–513.[CrossRef][Medline]

Kunst, F., Ogasawara, N., Moszer, I. & 148 other authors (1997). The complete genome sequence of the gram-positive bacterium Bacillus subtilis. Nature 390, 249–256.[CrossRef][Medline]

Lehrer, R. I. & Ganz, T. (2002). Defensins of vertebrate animals. Curr Opin Immunol 14, 96–102.[CrossRef][Medline]

Mascher, T., Margulis, N. G., Wang, T., Ye, R. W. & Helmann, J. D. (2003). Cell wall stress responses in Bacillus subtilis: the regulatory network of the bacitracin stimulon. Mol Microbiol 50, 1591–1604.[CrossRef][Medline]

Mascher, T., Zimmer, S. L., Smith, T. A. & Helmann, J. D. (2004). Antibiotic-inducible promoter regulated by the cell envelope stress-sensing two-component system LiaRS of Bacillus subtilis. Antimicrob Agents Chemother 48, 2888–2896.[Abstract/Free Full Text]

Ogura, M. & Tanaka, T. (2002). Recent progress in Bacillus subtilis two-component regulation. Front Biosci 7, d1815–d1824.[Medline]

Ohki, R., Tateno, K., Okada, Y., Okajima, H., Asai, K., Sadaie, Y., Murata, M. & Aiso, T. (2003a). A bacitracin-resistant Bacillus subtilis gene encodes a homologue of the membrane-spanning subunit of the Bacillus licheniformis ABC transporter. J Bacteriol 185, 51–59.[Abstract/Free Full Text]

Ohki, R., Giyanto, Tateno, K., Masuyama, W., Moriya, S., Kobayashi, K. & Ogasawara, N. (2003b). The BceRS two-component regulatory system induces expression of the bacitracin transporter, BceAB, in Bacillus subtilis. Mol Microbiol 49, 1135–1144.[CrossRef][Medline]

Oren, Z., Lerman, J. C., Gudmundsson, G. H., Agerberth, B. & Shai, Y. (1999). Structure and organization of the human antimicrobial peptide LL-37 in phospholipid membranes: relevance to the molecular basis for its non-cell-selective activity. Biochem J 341, 501–513.[CrossRef][Medline]

Pascale, J., Guiseppi, A., Sorokin, A. & Denizot, F. (2004). Characterization of the Bacillus subtilis YxdJ response regulator as the inducer of expression for the cognate ABC transporter YxdLM. Microbiology 150, 2609–2617.[CrossRef][Medline]

Perego, M., Glaser, P., Minutello, A., Strauch, M. A., Leopold, K. & Fischer, W. (1995). Incorporation of D-alanine into lipoteichoic acid and wall teichoic acid in Bacillus subtilis. Identification of genes and regulation. J Biol Chem 270, 15598–15606.[Abstract/Free Full Text]

Peschel, A., Otto, M., Jack, R. W., Kalbacher, H., Jung, G. & Götz, F. (1999). Inactivation of the dlt operon in Staphylococcus aureus confers sensitivity to defensins, protegrins and other antimicrobial peptides. J Biol Chem 274, 8405–8410.[Abstract/Free Full Text]

Podlesek, Z., Comino, A., Herzog-Velikonja, B., Zgur-Bertok, D., Komel, R. & Grabnar, M. (1995). Bacillus licheniformis bacitracin-resistance ABC transporter: relationship to mammalian multidrug resistance. Mol Microbiol 16, 969–976.[Medline]

Schobel, S., Zellmeier, S., Schumann, W. & Wiegert, T. (2004). The Bacillus subtilis sigmaW anti-sigma factor RsiW is degraded by intramembrane proteolysis through YluC. Mol Microbiol 52, 1091–1105.[CrossRef][Medline]

Thackray, P. D. & Moir, A. (2003). SigM, an extracytoplasmic function sigma factor of Bacillus subtilis, is activated in response to cell wall antibiotics, ethanol, heat, acid, and superoxide stress. J Bacteriol 185, 3491–3498.[Abstract/Free Full Text]

Turner, J., Cho, Y., Dinh, N. N., Waring, A. J. & Lehrer, R. I. (1998). Activities of LL-37, a cathelin-associated antimicrobial peptide of human neutrophils. Antimicrob Agents Chemother 42, 2206–2214.[Abstract/Free Full Text]

Vaara, M. & Vaara, T. (1983). Polycations as outer membrane-disorganizing agents. Antimicrob Agents Chemother 24, 114–122.[Medline]

Vagner, V., Dervyn, E. & Ehrlich, S. D. (1998). A vector for systematic gene inactivation in Bacillus subtilis. Microbiology 144, 3097–3104.[Medline]

Wiegert, T., Homuth, G., Versteeg, S. & Schumann, W. (2001). Alkaline shock induces the Bacillus subtilis sigma (W) regulon. Mol Microbiol 41, 59–71.[CrossRef][Medline]

Yang, L., Weiss, T. M., Lehrer, R. I. & Huang, H. W. (2000). Crystallization of antimicrobial pores in membranes: magainin and protegrin. Biophys J 79, 2002–2009.[Abstract/Free Full Text]

Yeaman, M. R. & Yount, N. Y. (2003). Mechanisms of antimicrobial peptide action and resistance. Pharmacol Rev 55, 27–55.[Abstract/Free Full Text]

Zasloff, M. (2002). Antimicrobial peptides of multicellular organisms. Nature 415, 389–395.[CrossRef][Medline]

Received 10 November 2004; revised 1 February 2005; accepted 3 February 2005.



This Article
Abstract
Full Text (PDF)
Alert me when this article is cited
Alert me if a correction is posted
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Google Scholar
Articles by Pietiäinen, M.
Articles by Kontinen, V. P.
Articles citing this Article
PubMed
PubMed Citation
Articles by Pietiäinen, M.
Articles by Kontinen, V. P.
Agricola
Articles by Pietiäinen, M.
Articles by Kontinen, V. P.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
INT J SYST EVOL MICROBIOL MICROBIOLOGY J GEN VIROL
J MED MICROBIOL ALL SGM JOURNALS
Copyright © 2005 Society for General Microbiology.