Bacillus subtilis contains seven extracytoplasmic-function factors that activate partially overlapping regulons . We here identify four additional members of the ^X regulon, pbpX [penicillin-bindingprotein], ywnJ, the dlt operon [D-alanylation of teichoic acids], and the pss ybfM psd operon [phosphatidylethanolamine biosynthesis].Modification of teichoic acids by esterification with D-alanineand incorporation of phosphatidylethanolamine into the cellmembrane have a common consequence: in both cases positivelycharged amino groups are introduced into the cell envelope.The resulting reduction in the net negative charge of the cellenvelope has been previously implicated as a resistance mechanismspecific for cationic antimicrobial peptides . Consistent withthis notion, we find that both sigX and dltA mutants are moresensitive to nisin than wild-type cells . We conclude that activationof the ^X regulon serves to alter cell surface properties toprovide protection against antimicrobial peptides.

 
Bacillus subtilis encodes seven extracytoplasmic-function [ECF] factors . Most studies to date have focused on three: ^X, ^W, and ^M [reviewed in reference 19] . sigX and its downstream genersiX [encoding the anti-^X factor] were originally observed tobe homologous [but not orthologous] to Escherichia coli fecIand fecR, which are involved in expression of ferric citratetransport genes [37] . Although expression of sigX in E . colican complement a fecI mutant [4], the B . subtilis sigX mutant is not affected in any known ferri-siderophore uptake systems[20].

To understand the function of ^X, we identified several ^X-regulated genes using a consensus promoter search method [22] . In theseinitial studies, we characterized six genes that are preceded by promoters recognized by ^X: sigX, abh [an AbrB homolog], csbB[a membrane-bound glucosyl transferase] [2], divIC [a membrane-boundcell-division initiation protein], lytR [a negative regulatorof autolysin] [30], and rapD [a response regulator aspartatephosphatase] [43] . These results suggested that ^X modulatesaspects of cell envelope metabolism . Interestingly, most ^X-controlled genes are also transcribed by other forms of holoenzyme . For example, csbB has an additional ^B-dependent promoter, lytR andrapD both have additional ^A-dependent promoters, and sigX itselfis preferentially transcribed from an upstream ^A-dependent sitein addition to the ^X-dependent autoregulatory promoter [20,22] . Moreover, in some cases [e.g., abh and divIC] the promoter activated by the E^X holoenzyme can also be recognized by theE^W holoenzyme at least in vitro [21] . Similarly, the recently defined bcrC gene [a bacitracin resistance gene] is transcribed from a promoter that is recognized by either ^X or ^M [7, 38].The unknown function of many ^X-controlled genes makes it difficultto predict a phenotype for the sigX mutant . This challenge isexacerbated by the fact that many of the genes are expressedfrom multiple promoters or by multiple holoenzyme forms activatingthe same promoter . The latter observation also makes DNA microarrayapproaches difficult, since many genes that can be activatedby ^X are also expressed by ^X-independent pathways.

In this study we have used both promoter consensus search andin vitro runoff transcription-macroarray analysis [ROMA] [8] to identify four additional ^X-dependent operons: dltABCDE, pssAybfM psd, pbpX, and ywnJ . Both the dlt and the pssA operonsencode enzymes that modulate cell surface charge [D-alanylation of teichoic acids and biosynthesis of phosphatidylethanolamine [PE], respectively], and PbpX is a low-molecular-weight penicillin-binding protein of unknown function . These results lead us to propose that one function of ^X is to regulate cell surface modificationas a defense against cationic antimicrobial peptides.

 
Bacterial strains, plasmids, oligonucleotides, and growth conditions. All bacterial strains, plasmids, and oligonucleotides used inthis study are listed in Table 1 . B . subtilis and E . coli strainswere grown at 37°C with vigorous shaking in Luria broth [LB] medium [50] unless otherwise indicated . For E . coli, 100µg of ampicillin/ml was used to select for Amp^r, and 200µg of spectinomycin/ml was used to select for Spc^r . ForB . subtilis, antibiotics used for selection were as follows:100 µg of spectinomycin/ml for Spc^r, 10 µg of kanamycin/mlfor Kan^r, 8 µg of neomycin/ml for Neo^r, and 1 µgof erythromycin/ml and 25 µg of lincomycin/ml for macrolide-lincomycin-streptograminB resistance [MLS^r].

 
Construction of mutants. CU1065 chromosomal DNA was amplified with primers #427 and #428.The PCR fragment was digested with SacI and PstI and ligatedinto pGEM-cat-3Zf[+] [59] to generate plasmid pMC57 . pMC57 wasdigested with HincII and SnaBI and ligated with a Spc^r cassette[PCR amplified from pKF59 [5] using T₃ and T₇ primers] to generatepMC58 . B . subtilis CU1065 was transformed with pMC58 [linearizedwith ScaI] with selection for Spc^r to generate HB0048 [dltA::spc].Thus, a 630-bp internal fragment of dltA was replaced with aSpc^r cassette.

Primers #371 and #372 were used to amplify an internal fragmentof dltA [490 bp] . The PCR fragment was digested with EcoRI andBamHI and cloned into pMUTIN4 [56], generating plasmid pMC59. This plasmid was inserted into CU1065 by Campbell integration and selection for MLS^r to generate strain HB0038 [dltA::pMUTIN].

CU1065 was transformed with chromosomal DNA from SDB01 [psd::neo] [33] and SDB02 [pssA::spc] [33] to generate the psd::neo [HB4519]and pssA::spc [HB4520] mutants, respectively . The dltA pssA[HB0094] and dltA psd [HB0095] double mutants were generatedby using chromosomal DNA from SDB02 and SDB01 to transform HB0038[dltA::pMUTIN] and HB0048[dltA::spc], with selection for [MLS^r plus Spc^r] and [Spc^r plus Neo^r], respectively.

Construction of promoter-cat-lacZ fusions. The putative promoter regions were amplified and cloned intopJPM122 [51] . The sequence of the promoter region in each plasmidwas verified by DNA sequencing [Cornell DNA sequencing facility].The promoter fusions were introduced into the SPßprophage by double-crossover recombination, in which each pJPM122derivative was linearized by digestion with ScaI and used totransform B . subtilis ZB307A [60] with selection for Neo^r . SPß lysates were prepared by heat induction and used to transduce various recipient strains, and ß-galactosidase activitywas measured on each sample in early stationary phase [when ^X activity is at a maximum] [21] as described by Miller [34].

Purification of RNAP and factors. Preparation of B . subtilis core RNA polymerase [RNAP] and ^A, ^D, ^X, ^W and proteins was previously described [10, 20, 21,25, 31].

ROMA. The ROMA experiment was performed as described previously [8].A typical transcription reaction [50 µl] contains 1.3pmol of core RNAP, 16 pmol of ^X, 15 pmol of , and 1 µgof digested genomic DNA mixed in transcription buffer [20 mM Tris-HCl [pH 8.0], 10 mM MgCl₂, 50 mM KCl, 0.5 mM dithiothreitol,0.1-mg/ml bovine serum albumin, 5% [vol/vol] glycerol, and RNasinfrom Promega [10 U/reaction]] and NTP mixture [40 nmol of ATP,GTP, CTP, and 8 nmol of [-³³P]UTP [3,000 Ci/mmol] from NEN].The Panorama B . subtilis gene arrays [catalog no . PRBS0002]were purchased from Sigma-Genosys Biotechnologies, Inc.

In vitro runoff transcription assays for candidate genes. A typical runoff reaction mixture [20 µl] contains 0.36pmol of core RNAP, 4.5 pmol of ^X, 4.2 pmol of , 0.04 pmol ofPCR-amplified template DNA [normally the same fragments usedfor generating promoter fusions] and NTP mixture [10 nmol of ATP, GTP, CTP, 1 nmol of UTP, and 0.6 pmol of [-³²P]UTP [3,000Ci/mmol]].Reactions were incubated and processed as described for theROMA experiments [8].

Primer extension assays. RNA was either purified from in vitro runoff transcription reactionmixtures or extracted from late-log-phase B . subtilis cellsusing phenol-chloroform extraction as described previously [22].A PCR fragment containing the promoter region studied was sequencedusing the same primer to index the transcription start site.

Nisin MIC assays. Nisin was obtained from Sigma Chemical Co . and dissolved in20 mM HCl . Overnight cultures were diluted 1:100 into freshLB medium in the presence of nisin at the indicated concentration.After incubation for 6 h with shaking, the optical density at600 nm [OD₆₀₀] was measured.

Autolysis test. CU1065 [wild type], sigX::spc, dltA::spc, and dltA::pMUTIN strainswere grown in LB or minimal medium [9] . Cells were harvested at exponential growth phase [OD₆₀₀, 0.7], washed twice withcold Tris-HCl buffer [pH 7.1], and resuspended in 50 mM Tris-HClbuffer [pH 7.1] containing 0.05% Triton X-100 . Incubation wasat 37°C, and autolysis was monitored by measuring the decreaseof OD₆₀₀ at 30-min intervals.

Northern blot analysis. Primers #537 and #538 were used to amplify an internal fragment[570 bp] of psd from CU1065 chromosomal DNA . After HindIII digestion, the fragment was labeled with [-³²P]dATP by the 3' fill-in methodusing a Klenow fragment of DNA polymerase [Exo^-; New EnglandBioLabs] . The probe was hybridized with membranes containingtotal RNA from wild-type, sigX, and rsiX strains [same RNA sampleused for primer extension; see above] . The NorthernMax formaldehyde-basedsystem [Ambion, Inc.] was used to perform the Northern analysis.Ten micrograms of total RNA was denatured and loaded on 1% formaldehydeagarose gel . Hybridization was performed at 42°C overnight.The second day, the blot was washed twice with low-stringencybuffer [2x SSC [1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate]]at room temperature followed by two washes with high-stringencybuffer [0.1x SSC] at 42°C . The blot was wrapped in plasticwrap and exposed to a Phosphor screen [Molecular Dynamics].

 
Defining the ^X regulon using promoter consensus search. Previously, saturation mutagenesis of the sigX autoregulatorypromoter was used to identify those bases important for ^X-dependent promoter recognition . The resulting consensus was used to search the partially sequenced B . subtilis genome [63% complete at the time the analysis was done [22]] to identify candidate promoters.While this approach identified several ^X target genes, otherswere subsequently found to be primarily controlled by ^W or ^M, which recognize promoters with closely related sequences [22, 35].

The availability of the complete genome sequence [28], togetherwith a better understanding of the rules for promoter recognitionby ^X and its paralog ^W [21, 48], encouraged us to repeat the consensus search procedure . Although we explored the use of several different search patterns to identify likely candidates,one of the most successful searches used the degenerate consensus tgtaACtttt n_12-13 CG[A,T]C to screen the SubtiList database [36] for those sites within 250 bp of an annotated start codon.This search pattern is based on the observation that many identified ^X-dependent promoters contain a T-rich region in the downstreamportion of the -35 element and the AC base pairs are highlyconserved . We allowed up to three mismatches in this extended-35 element [in the positions in lowercase] and none in the-10 element . By including those promoters with -10 elementsof either CGTC or CGAC, we expected to identify some sites alreadydefined as largely dependent on ^W . The resulting list of candidatepromoters includes one site with no mismatches [the sigX autoregulatorysite], three with one mismatch [preceding lytR, ywnJ, and dltA], and four with two mismatches [divIC, ydjA, abh, and yrhH; underlinedsites are known to be at least partially ^X dependent in vivoor in vitro] [21, 22] . Among the 15 sites with three mismatchesin the -35 element, we focused our attention on those precedingpssA and pbpX, since these genes are of known function [Table2].

 
Defining the ^X regulon using ROMA. As a complementary mechanism to identify candidate ^X targetgenes, we performed in vitro ROMA analysis [8] . The ROMA approachgenerates ³³P-labeled runoff transcripts using ^X-containing holoenzyme to transcribe total genomic DNA that has been restricted with either EcoRI or HindIII . The resulting runoff transcriptsare then used to probe a DNA macroarray [Sigma/GenoSys] containing4,107 B . subtilis open reading frames . Candidates [genes whosesignal became stronger in the presence of ^X] were chosen forfurther analysis if they had a particularly strong signal orif they were associated with a plausible promoter site as identifiedby the consensus search approach described above . In additionto signals corresponding to promoters known to be recognized by ^X [e.g., sigX, csbB, lytR, divIC, and bcrC] [Fig . 1], theROMA experiment revealed strong signals for the pssA and ywnJgenes and a weaker signal for dltA . Note that in many cases,these same genes appeared in reactions using ^W holoenzyme insteadof ^X [Fig . 1], consistent with the known overlap between the sets of promoters recognized by these two factors [22].

 
Confirmation of promoters for pbpX and ywnJ. We used a reporter fusion to demonstrate that pbpX is dependent on sigX in vivo, with a further reduction in expression in the sigX sigW double mutant [Fig . 2A] . DNA microarray analyses revealthat the expression of pbpX decreased 2.7-fold in the sigX mutant[data not shown] but not in the sigW mutant [8] . The reporterfusion for the putative ywnJ promoter had very low activity,so in vitro transcription was used to demonstrate that thissite could be recognized by both the ^X and ^W holoenzymes [Fig.2B] . Transcription initiates at the expected site, as measuredby primer extension mapping of the resulting in vitro transcripts[Fig . 2C].

 
The dltABCDE operon is largely dependent on ^X. The B . subtilis dltABCDE operon is responsible for D-alanine esterification of both lipoteichoic acids [LTA] and wall teichoic acids [WTA] [44] . Transcription of the dltABCDE operon was originallyproposed to be largely [70%] ^D dependent, with the residualactivity perhaps due to two putative ^A-dependent promoters [Fig.3A, P₁ and P₄] [44].

 
To examine the regulation of the dlt operon, we integrated two lacZ transcriptional fusions ectopically at the SPßlocus . The P_dltA1-cat-lacZ fusion consists of a 730-bp fragment[from -630 bp to +100 bp relative to the start codon] and includesall four putative promoters [P₁ to P₄] . The P_dltA2-cat-lacZfusion consists of a shorter fragment [from -390 bp to +100bp] and includes P₂ through P₄ [Fig . 3B] . Results with bothpromoter fusions indicate that expression is reduced by about 85% in the sigX mutant and slightly increased in the rsiX mutant[defective in the anti- factor that targets ^X [20]], confirmingthe existence of a ^X-dependent promoter in this region [Fig.3C] . Expression was unaffected in the sigW and rsiW mutant strains,despite the presence of a CGTC motif is the predicted -10 region[see Discussion] . Under our growth conditions, ^D does not seemto play a role in dlt transcription, since the activities fromboth fusions neither decreased in a sigD mutant nor increasedin a flgM [anti-^D] mutant . Moreover, the first putative ^A-dependent promoter [P₁] apparently did not contribute to the dlt transcription,since expression from P_dltA1 and P_dltA2 was similar . The residual activity [11 Miller units] in the sigX mutant might be due torecognition of the ^X-dependent promoter by another ECF factoror might be due to another promoter [maybe P₄].

We extended these in vivo results using in vitro runoff transcription and primer extension assays . When the long PCR fragment [P_dltA1] was incubated with RNAP core enzyme in the presence of ^A, ^X, ^W or ^D, appropriately sized transcripts were generated by boththe ^X and ^D holoenzymes [Fig . 4A] . Although ^W could weakly recognizethe ^X-dependent site in vitro [Fig . 4A], it did not appear toplay a major role in vivo [Fig . 3C] . While E^D could initiatetranscription from the ^D-dependent promoter in vitro, in vivotranscripts were detected only for the ^X-dependent promoter[Fig . 4B] . Primer extension reactions indicate that transcriptionof dltA initiates primarily from a G residue 11 bases downstreamof the -10 CGTC motif and secondarily from an A residue 2 basesupstream . Both signals became stronger in the rsiX mutant andwere greatly reduced in the sigX mutant . No other strong startsites were visible in this region . We conclude that dlt expressionis largely dependent on ^X in vivo.

 
The pssA ybfM psd operon is partially controlled by ^X. Our identification of a candidate ^X-dependent promoter upstreamof the pssA gene suggests a possible role for ^X in regulatingthe phospholipid content of the membrane . The pssA gene is partof a predicted operon including ybfM and psd . Together, thePssA and Psd proteins catalyze the synthesis of PE . Okada etal . [39] proposed two putative ^A-dependent promoters [P₁ andP₂] upstream of pssA, and our results suggest a third ^X-dependent promoter [P₃] [TGTAAC-N_16-CGTCaa] [Fig. 5A].

 
To test the contribution of each promoter to pssA expression, we constructed two lacZ fusions: one contains the complete promoter region [P₁, P₂, and P₃], the other contains only P₃ [Fig . 5B]. ß-Galactosidase assays demonstrate that about one-halfof the expression derives from the ^X-dependent promoter [P₃],with the other half from the region containing P₁ and P₂ [Fig.5C].

Recognition of P₃ by ^X holoenzyme was confirmed in vitro byrunoff transcription assays [Fig . 6A] . A faint, larger bandwas observed in reactions containing E^A, probably resultingfrom one of the ^A-dependent promoters . We used primer extensionassays to localize the transcription start site for the ^X holoenzymeto an A residue 10 bp downstream from CGTC [Fig. 6B] . A weaktranscript was detected in the wild type [CU1065] but not inthe sigX mutant strain . The amount of transcript increased inthe rsiX mutant, as expected for a ^X-dependent promoter.

 
To test whether pssA and psd are in one operon, we conducted Northern blot analysis using a ³²P-labeled internal fragment of the psd gene as a probe . A large transcript [1.9 kb] wasdetected, consistent with an mRNA extending from pssA throughpsd . A smaller transcript [850 bp] likely corresponds to thepsd gene and may have been produced by RNA processing, sinceit varies in intensity with the full-length transcript [Fig.6C] . The density of both bands decreased about 50% in the sigXmutant and increased in the rsiX mutant, consistent with theprevious conclusion that ^X contributes 50% of the expressionof P_pssA.

sigX mutants are altered in autolysis and sensitivity to cationic antimicrobial peptides. Since ^X regulates both D-alanylation of teichoic acids and PE biosynthesis, we tested the effects of a sigX mutation on two phenotypes previously shown to be affected by cell surface charge: autolysis and resistance to cationic antimicrobial peptides.We first compared the Triton X-100-induced autolysis rates [58] of the wild type and the sigX and dltA::spc mutants . Autolysinsare a group of positively charged cell wall hydrolytic enzymesthat bind more avidly to the cell wall of dlt mutant strains[58] . As expected, both the sigX and dltA mutants have a twofoldincrease in the rate of autolysis compared to the wild type[Fig . 7A] . Similar results were observed with cells grown inLB or minimal medium.

 
Cationic antimicrobial peptides [CAMPs] are a broadly distributed family of peptides that kill bacteria . Many are thought to actby accumulating within the cytoplasmic membrane to a critical concentration that allows the assembly of structures that permeabilize the cell [16-18] . To test whether ^X plays a role in resistanceto CAMPs, we measured the MICs of nisin for the wild type andthe sigX, dltA, pssA, and psd mutants: a positively charged[+3] peptide produced by Lactococcus lactis [24] . As expected,the sigX and dltA mutants were more sensitive to nisin than the wild type [Fig . 7B] . The psd mutant had only slightly increasedsensitivity, while the pssA mutant was unaffected . A psd dltAdouble mutant behaved much like the dltA single mutant . In additionto nisin, the sigX and dltA mutants were more sensitive to severalother tested CAMPs [S . Farmer and R . Hancock, personal communication]but not to gramicidin, a neutral peptide . In contrast, the mutantswere unaltered in their sensitivity to vancomycin, tunicamycin,or lysozyme [data not shown], although dlt mutants have been previously reported to display an increased susceptibility to methicillin [57].

 
Using a promoter consensus search and ROMA approaches, we have defined four additional ^X-dependent operons . Together with theresults of our previous analyses [22], we conclude that mostmembers of the ^X regulon control processes related to the compositionor metabolism of the cell envelope . For example, LytR is a negativeregulator of autolysin activity [30], CsbB is a membrane-bound glucosyl transferase likely involved in cell wall biosynthesis[2], PbpX is a penicillin-binding protein, DltA, DltB, DltCand DltD are responsible for D alanylation of the WTA and LTA[44], and PssA and Psd are enzymes for PE biosynthesis [33]. We note that most of these operons are expressed from complex promoter regions: lytR is controlled by both ^A and ^X, csbB isalso regulated by ^B, pbpX is partially regulated by ^W, and ^X and ^A each contribute to pssA ybfM psd expression . Perhaps dueto this overlapping regulation, the sigX mutant strain doesnot display dramatic growth defects under most tested conditions, although some increased sensitivity to oxidative stress andheat stress has been noted [20] . Here, we have extended the phenotypes of the sigX mutant to include increased rates of autolysis and increased sensitivity to cationic antimicrobial peptides.

Promoter recognition by ^X. As noted previously, ^X recognizes -10 elements with sequenceCGaC, ^W recognizes CGTa, and both can recognize CGTC [lowercasereflects a noncritical base for recognition] [21, 22, 48] . InTable 2 we compile the 11 promoters that are the best candidatesfor regulation by ^X in vivo . Note that some of these sites canalso be recognized by either ^W or ^M . For example, both bcrC[7, 38] and pbpX [Fig . 2A] seem to be under dual control in vivo, and ^W recognizes several other sites in vitro [Table 2]. In addition, a number of other promoters previously studied[22] can be recognized by ^X in vitro, but an in vivo role for ^X has not been documented, and it seems likely that they maybe primarily dependent on ^W or ^M for in vivo expression [22].Indeed, in B . subtilis W23, expression of the divIC gene ispartially ^M dependent [35].

Inspection of Table 2 allows a refinement of our previous modelsfor promoter discrimination among ECF factors . Specifically,we note that most of the newly characterized promoters identifiedin this study contain a CGTC -10 motif, previously shown tobe also recognized by ^W holoenzyme . However, all four promoters[abh, divIC, pbpX, and ywnJ] that are also recognized by ^W sharea common extended -10 region of CGTCta . In contrast, the otherthree [rapD, dltA, and pssA] that are recognized only by ^X havea -10 region of "CGTCaa [Table 2] . This is consistent with theobservation that the highly specific autoregulatory sites forsigX and sigW contain -10 elements of "CGACaa" and "CGTAta,"respectively . Furthermore, in a previous promoter mutagenesisstudy we found that changing the sigX promoter [-10] regionCGACaa to CGTCaa resulted in a site that retained high selectivityfor ^X . In contrast, when the sigW [-10] region CGTAta was changedto CGTCta, both ^X and ^W could recognize this promoter [48].We therefore conclude that [i] the preferred -10 consensus sequencesfor ^X [CGaCaa] and ^W [CGTata] differ in two positions [italics]rather than one position and [ii] there is considerable overlapbetween these two regulons.

Biological role of ^X and the ^X regulon. Distinctive aspects of the gram-positive bacterial cell envelopeinclude the presence of a thick cell wall containing peptidoglycan,WTA, and LTA [14] . In B . subtilis 168 strains, the negativelycharged teichoic acids contain an alternating glycerol phosphatecopolymer, whereas in B . subtilis W23 strains ribitol replacesglycerol . Recent results indicate that ribitol-based teichoicacid synthesis in W23 strains is regulated by both ^X and ^M [29,35].

In general, the WTA and LTA polymers are highly modified by esterification on the sugar residues with sugar, amino sugar,or amino acid substituents . For B . subtilis, LTA chains contain between 24 and 33 glycerol phosphate monomers and carry, onaverage, 0.35 to 0.55 D-alanine constituents and 0.2 to 0.4 glycosyl substituents per monomer [14] . The D-Ala residues onLTA are subject to rapid turnover, both by spontaneous hydrolysisand by transesterification to WTA [14, 26] . D-Alanylation ofWTA and LTA is catalyzed by the products of the dlt operon. The dltA and dltC genes encode the D-alanine-D-alanyl carrierprotein ligase [Dcl] and the D-alanyl carrier protein [Dcp],respectively . DltB and DltD may function in transport and theactual esterification reaction [45].

The modification with D-Ala introduces free amino groups [NH₃⁺]into the cell envelope and thereby reduces the net negativecharge of the surface [44] . Genetic studies with several microorganismsindicate that dlt mutants are pleiotropic, with phenotypes includingaltered patterns of autolysis, increased sensitivity to CAMPs[1, 45-47], altered colonization properties [11], altered carbohydratemetabolism [52], enhanced UV sensitivity, and loss of acid tolerance[3] . In addition, D-alanylation affects protein folding and secretion [23, 54] . Our results suggest that conditions leadingto activation of the ^X regulon will lead to enhanced expressionof the dlt operon and thereby result in a decrease in the netnegative charge of the cell wall . The factors that activateexpression of the ^X regulon are not well defined, but they arelikely to act through the RsiX anti-, shown previously to inhibit ^X activity [20] . It has also been shown that transposon insertionsin the yitG multidrug efflux system, the manA gene encodingmannose-6-phosphate isomerase, the srfAB surfactin biosynthesisgene, the ytxJ general stress protein, the ywpH single-strandedDNA-binding protein, and the yogA alcohol dehydrogenase locusalso lead to enhanced expression from a ^X-dependent promoter[55], although the significance of these observations is notyet clear.

The bacterial cell membrane also contains a net negative charge due to the abundance of anionic phospholipids . However, PE,a neutral [zwitterionic] lipid, makes up as much as 50% of theB . subtilis membrane [33] . The biosynthesis of PE in B . subtilis is carried out by two membrane-localized enzymes: CDP-diacylglycerol-dependentphosphatidylserine [PS] synthase [PssA] and phosphatidylserinedecarboxylase [Psd] . The genes [pssA and psd] encoding the enzymesare separated by another gene, ybfM, on the chromosome . Allthree genes are cotranscribed [Fig . 6C] . The psd mutant containsno PE and accumulates PS, while the pssA mutant contains noPE or PS . The absence of PE in B . subtilis cells does not haveany adverse effects on cell growth, probably due to compensationfrom increased glucosyldiacylglycerol content in the membrane[33] . Interestingly, the E . coli psd gene has recently beenshown to have a ^E-dependent promoter [49], suggesting that thissystem may also be controlled, at least in part, by an ECF factor in this organism.

In B . subtilis, ^X may serve to regulate the net charge in thecell envelope by affecting the expression of both the dlt andpssA operons [Fig . 8], and this, in turn, may affect sensitivity to CAMPs . CAMPs share several common features, including broad-spectrum antimicrobial activity and cationic charge at physiologicalpH [16-18] . CAMPs act by an initial electrostatic binding tothe anionic moieties on the microbial membrane, followed bymembrane disruption [16-18] . In eukaryotes, CAMPs [includingdefensins] are the major form of defense against bacterial infectionand are induced by bacteria or lipopolysaccharides [12, 17].

 
Bacteria can acquire resistance to CAMPs by modification oftheir surface properties, although in general the underlyingregulatory mechanisms have not been described . For example,a nisin-resistant Listeria monocytogenes strain contains elevatedlevels of zwitterionic PE and a reduction in phosphatidylglyceroland cardiolipin [13] . Similarly, a nisin-resistant strain of the rumen bacterium Streptococcus bovis has decreased negative surface charge [32] . A recent study found that Staphylococcusaureus achieves resistance to defensins and CAMPs by modifyinganionic membrane lipids with L-lysine [27] . In gram-negativebacteria, resistance often involves modification of lipopolysaccharides.For example, CAMP resistance in Salmonella enterica involvesaddition of palmitate or 4-aminoarabinose to lipid A, a processregulated by the PmrA-PmrB two-component regulatory system [53].

In this study, we demonstrate that sensitivity of B . subtilis to CAMPs is affected by an ECF factor that contributes to theexpression of two operons that modulate surface charge [Fig.8] . Other ^X regulon proteins [e.g., LytR, PbpX, and CsbB] mayalso participate in this adaptive response . In other bacteria,related cell wall homeostasis functions may be controlled bytwo-component regulatory systems instead of, or in additionto, ECF factors . For example, the Streptomyces coelicolor CseC-CseB two-component system activates expression of ^E, in responseto unknown signals, which then functions to modify cell wallstructure [40, 41] . In Streptococcus agalactiae, up-regulationof the dlt operon when D-alanine incorporation into LTA is deficient is controlled by the DltS-DltR two-component system [46] . Ourstudies provide evidence linking ECF factors to the biosynthesisand modification of the cell envelope and suggest that theseregulatory proteins may participate in an inducible defenseresponse providing resistance to CAMPs.

 
We would like to thank previous lab members who purified proteins used in this study: Y . L . Juang [RNAP and ^A], Y . F . Chen [^D], X . J . Huang [^X and ^W], and F . J . Lopez de Saro [] . Thanks alsogo to Y . Chai for construction of the P_pssA-cat-lacZ fusions,J . Qiu for construction of the P_pbpX-cat-lacZ fusion, S . Farmerand R . Hancock for tests of CAMP sensitivity, and K . Matsumotofor providing the original pssA and psd mutant strains.

This work was supported by NIH grant GM-47446 [to J.D.H.].

 
* Corresponding author . Mailing address: Department of Microbiology, Wing Hall, Cornell University, Ithaca, NY 14853-8101 . Phone: [607] 255-6570 . Fax: [607] 255-3904 . E-mail: jdh9@cornell.edu .

1. Abachin, E., C . Poyart, E . Pellegrini, E . Milohanic, F . Fiedler, P . Berche, and P . Trieu-Cuot. 2002 . Formation of D-alanyl-lipoteichoic acid is required for adhesion and virulence of Listeria monocytogenes. Mol Microbiol . 43:1-14.
2. Akbar, S., and C . W . Price. 1996 . Isolation and characterization of csbB, a gene controlled by Bacillus subtilis general stress transcription factor ^B . Gene 177:123-128.
3. Boyd, D . A., D . G . Cvitkovitch, A . S . Bleiweis, M . Y . Kiriukhin, D . V . Debabov, F . C . Neuhaus, and I . R . Hamilton. 2000 . Defects in D-alanyl-lipoteichoic acid synthesis in Streptococcus mutans results in acid sensitivity . J . Bacteriol . 182:6055-6065 .
4. Brutsche, S., and V . Braun. 1997 . SigX of Bacillus subtilis replaces the ECF sigma factor FecI of Escherichia coli and is inhibited by RsiX . Mol . Gen . Genet . 256:416-425.
5. Bsat, N., A . Herbig, L . Casillas-Martinez, P . Setlow, and J . D . Helmann. 1998 . Bacillus subtilis contains multiple Fur homologues: identification of the iron uptake [Fur] and peroxide regulon [PerR] repressors . Mol . Microbiol . 29:189-198.
6. Cao, M., B . A . Bernat, Z . Wang, R . N . Armstrong, and J . D . Helmann. 2001 . FosB, a cysteine-dependent fosfomycin resistance protein under the control of ^W, an extracytoplasmic-function sigma factor in Bacillus subtilis. J . Bacteriol . 183:2380-2383 .
7. Cao, M., and J . D . Helmann. 2002 . Regulation of the Bacillus subtilis bcrC bacitracin resistance gene by two extracytoplasmic function sigma factors . J . Bacteriol . 184:6123-6129 .
8. Cao, M., P . A . Kobel, M . M . Morshedi, M . F . Wu, C . Paddon, and J . D . Helmann. 2002 . 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.
9. Chen, L., L . P . James, and J . D . Helmann. 1993 . Metalloregulation in Bacillus subtilis: isolation and characterization of two genes differentially regulated by metal ions . J . Bacteriol . 175:5428-5437.
10. Chen, Y.-F., and J . D . Helmann. 1995 . The Bacillus subtilis flagellar regulatory protein ^D: overproduction, domain analysis, and DNA-binding . J . Mol . Biol . 249:743-753.
11. Clemans, D . L., P . E . Kolenbrander, D . V . Debabov, Q . Zhang, R . D . Lunsford, H . Sakone, C . J . Whittaker, M . P . Heaton, and F . C . Neuhaus. 1999 . Insertional inactivation of genes responsible for the D-alanylation of lipoteichoic acid in Streptococcus gordonii DL1 [Challis] affects intrageneric coaggregations . Infect . Immun . 67:2464-2474 .
12. Cole, A . M., and T . Ganz. 2000 . Human antimicrobial peptides: analysis and application . BioTechniques 29:822-826, 828, 830-831.
13. Crandall, A . D., and T . J . Montville. 1998 . Nisin resistance in Listeria monocytogenes ATCC 700302 is a complex phenotype . Appl . Environ . Microbiol . 64:231-237 .
14. Fischer, W. 1988 . Physiology of lipoteichoic acids in bacteria . Adv . Microb . Physiol . 29:233-302.
15. Fredrick, K., and J . D . Helmann. 1996 . FlgM is a primary regulator of D activity, and its absence restores motility to a sinR mutant . J . Bacteriol . 178:7010-7013.
16. Hancock, R . E., and D . S . Chapple. 1999 . Peptide antibiotics . Antimicrob . Agents Chemother . 43:1317-1323.
17. Hancock, R . E., and G . Diamond. 2000 . The role of cationic antimicrobial peptides in innate host defences . Trends Microbiol . 8:402-410.
18. Hancock, R . E., and M . G . Scott. 2000 . The role of antimicrobial peptides in animal defenses . Proc . Natl . Acad . Sci . USA 97:8856-8861 .
19. Helmann, J . D. 2002 . The extracytoplasmic function [ECF] sigma factors . Adv . Microb . Physiol . 46:47-110.
20. Huang, X., A . Decatur, A . Sorokin, and J . D . Helmann. 1997 . The Bacillus subtilis ^X protein is an extracytoplasmic function sigma factor contributing to survival at high temperature . J . Bacteriol . 179:2915-2921.
21. Huang, X., K . L . Fredrick, and J . D . Helmann. 1998 . Promoter recognition by Bacillus subtilis ^W: autoregulation and partial overlap with the ^X regulon . J . Bacteriol . 180:3765-3770 .
22. Huang, X., and J . D . Helmann. 1998 . Identification of target promoters for the Bacillus subtilis ^X factor using a consensus-directed search . J . Mol . Biol . 279:165-173.
23. Hyyrylainen, H . L., M . Vitikainen, J . Thwaite, H . Wu, M . Sarvas, C . R . Harwood, V . P . Kontinen, and K . Stephenson. 2000 . D-Alanine substitution of teichoic acids as a modulator of protein folding and stability at the cytoplasmic membrane/cell wall interface of Bacillus subtilis. J . Biol . Chem . 275:26696-26703 .
24. Jack, R . W., J . R . Tagg, and B . Ray. 1995 . Bacteriocins of Gram-positive bacteria . Microbiol . Rev . 59:171-200.
25. Juang, Y . L., and J . D . Helmann. 1994 . The delta subunit of Bacillus subtilis RNA polymerase . An allosteric effector of the initiation and core-recycling phases of transcription . J . Mol . Biol . 239:1-14.
26. Koch, H . U., R . Doker, and W . Fischer. 1985 . Maintenance of D-alanine ester substitution of lipoteichoic acid by reesterification in Staphylococcus aureus. J . Bacteriol . 164:1211-1217.
27. Kristian, S . A., M . Durr, J . A . Van Strijp, B . Neumeister, and A . Peschel. 2003 . MprF-mediated lysinylation of phospholipids in Staphylococcus aureus leads to protection against oxygen-independent neutrophil killing . Infect . Immun . 71:546-549 .
28. Kunst, F., N . Ogasawara, I . Moszer, A . M . Albertini, G . Alloni, V . Azevedo, M . G . Bertero, P . Bessieres, A . Bolotin, S . Borchert, R . Borriss, L . Boursier, A . Brans, M . Braun, S . C . Brignell, S . Bron, S . Brouillet, C . V . Bruschi, B . Caldwell, V . Capuano, N . M . Carter, S . K . Choi, J . J . Codani, I . F . Connerton, A . Danchin, et al. 1997 . The complete genome sequence of the Gram-positive bacterium Bacillus subtilis. Nature 390:249-256.
29. Lazarevic, V., F . X . Abellan, S . B . Moller, D . Karamata, and C . Mauel. 2002 . Comparison of ribitol and glycerol teichoic acid genes in Bacillus subtilis W23 and 168: identical function, similar divergent organization, but different regulation . Microbiology 148:815-824 .
30. Lazarevic, V., P . Margot, B . Soldo, and D . Karamata. 1992 . Sequencing and analysis of the Bacillus subtilis lytRABC divergon: a regulatory unit encompassing the structural genes of the N-acetylmuramoyl-L-alanine amidase and its modifier . J . Gen . Microbiol . 138:1949-1961.
31. Lopez de Saro, F . J., A . Y . Woody, and J . D . Helmann. 1995 . Structural analysis of the Bacillus subtilis delta factor: a protein polyanion which displaces RNA from RNA polymerase . J . Mol . Biol . 252:189-202.
32. Mantovani, H . C., and J . B . Russell. 2001 . Nisin resistance of Streptococcus bovis. Appl . Environ . Microbiol . 67:808-813 .
33. Matsumoto, K., M . Okada, Y . Horikoshi, H . Matsuzaki, T . Kishi, M . Itaya, and I . Shibuya. 1998 . Cloning, sequencing, and disruption of the Bacillus subtilis psd gene coding for phosphatidylserine decarboxylase . J . Bacteriol . 180:100-106 .
34. Miller, J . H. 1972 . Experiments in molecular genetics, p . 352-355 . Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
35. Minnig, K., J . L . Barblan, S . Kehl, S . B . Moller, and C . Mauel. 2003 . In Bacillus subtilis W23, the duet ^XM, two sigma factors of the extracytoplasmic function subfamily, are required for septum and wall synthesis under batch culture conditions . Mol . Microbiol . 49:1435-1447.
36. Moszer, I., P . Glaser, and A . Danchin. 1995 . SubtiList: a relational database for the Bacillus subtilis genome . Microbiology 141:261-268.
37. Ochs, M., S . Veitinger, I . Kim, D . Welz, A . Angerer, and V . Braun. 1995 . Regulation of citrate-dependent iron transport of Escherichia coli: fecR is required for transcription activation by FecI . Mol . Microbiol . 15:119-132.
38. Ohki, R., K . Tateno, Y . Okada, H . Okajima, K . Asai, Y . Sadaie, M . Murata, and T . Aiso. 2003 . 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 .
39. Okada, M., H . Matsuzaki, I . Shibuya, and K . Matsumoto. 1994 . Cloning, sequencing, and expression in Escherichia coli of the Bacillus subtilis gene for phosphatidylserine synthase . J . Bacteriol . 176:7456-7461.
40. Paget, M . S., E . Leibovitz, and M . J . Buttner. 1999 . A putative two-component signal transduction system regulates sigmaE, a sigma factor required for normal cell wall integrity in Streptomyces coelicolor A3[2] . Mol . Microbiol . 33:97-107.
41. Paget, M . S . B., L . Chamberlin, A . Atrih, S . J . Foster, and M . J . Buttner. 1999 . Evidence that the extracytoplasmic function sigma factor E is required for normal cell wall structure in Streptomyces coelicolor A3[2] . J . Bacteriol . 181:204-211 .
42. Perego, M. 1993 . Integrational vectors for genetic manipulation in Bacillus subtilis, p . 615-624 . In A . L . Sonenshein, J . A . Hoch, and R . Losick [ed.], Bacillus subtilis and other gram-positive bacteria: biochemistry, physiology, and molecular genetics . American Society for Microbiology, Washington, D.C.
43. Perego, M., P . Glaser, and J . A . Hoch. 1996 . Aspartyl-phosphate phosphatases deactivate the response regulator components of the sporulation signal transduction system in Bacillus subtilis. Mol . Microbiol . 19:1151-1157.
44. Perego, M., P . Glaser, A . Minutello, M . A . Strauch, K . Leopold, and W . Fischer. 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 .
45. Peschel, A., M . Otto, R . W . Jack, H . Kalbacher, G . Jung, and F . Gotz. 1999 . Inactivation of the dlt operon in Staphylococcus aureus confers sensitivity to defensins, protegrins, and other antimicrobial peptides . J . Biol . Chem . 274:8405-8410 .
46. Poyart, C., M . C . Lamy, C . Boumaila, F . Fiedler, and P . Trieu-Cuot. 2001 . Regulation of D-alanyl-lipoteichoic acid biosynthesis in Streptococcus agalactiae involves a novel two-component regulatory system . J . Bacteriol . 183:6324-6334 .
47. Poyart, C., E . Pellegrini, M . Marceau, M . Baptista, F . Jaubert, M . C . Lamy, and P . Trieu-Cuot. 2003 . Attenuated virulence of Streptococcus agalactiae deficient in D-alanyl-lipoteichoic acid is due to an increased susceptibility to defensins and phagocytic cells . Mol . Microbiol. 49:1615-1625.
48. Qiu, J., and J . D . Helmann. 2001 . The -10 region is a key promoter specificity determinant for the Bacillus subtilis extracytoplasmic-function sigma factors ^X and ^W . J . Bacteriol . 183:1921-1927 .
49. Rezuchova, B., H . Miticka, D . Homerova, M . Roberts, and J . Kormanec. 2003 . New members of the Escherichia coli sigmaE regulon identified by a two-plasmid system . FEMS Microbiol . Lett . 225:1-7.
50. Sambrook, J., E . F . Fritsch, and T . Maniatis. 1990 . Molecular cloning: a laboratory manual, 2nd ed . Cold Spring Harbor Press, Cold Spring Harbor, N.Y.
51. Slack, F . J., J . P . Mueller, and A . L . Sonenshein. 1993 . Mutations that relieve nutritional repression of the Bacillus subtilis dipeptide permease operon . J . Bacteriol . 175:4605-4614.
52. Spatafora, G . A., M . Sheets, R . June, D . Luyimbazi, K . Howard, R . Hulbert, D . Barnard, M . el Janne, and M . C . Hudson. 1999 . Regulated expression of the Streptococcus mutans dlt genes correlates with intracellular polysaccharide accumulation . J . Bacteriol . 181:2363-2372 .
53. Tamayo, R., S . S . Ryan, A . J . McCoy, and J . S . Gunn. 2002 . Identification and genetic characterization of PmrA-regulated genes and genes involved in polymyxin B resistance in Salmonella enterica serovar Typhimurium . Infect . Immun . 70:6770-6778 .
54. Thwaite, J . E., L