University of Groningen Nonribosomal peptide synthesis in ... · growing peptide chain (Mootz and Marahiel, 1997). In addition, an increasing number of hybrid synthetases have been

University of Groningen

Nonribosomal peptide synthesis in Bacillus subtilisDuitman, Erwin Hans

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2003

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Duitman, E. H. (2003). Nonribosomal peptide synthesis in Bacillus subtilis. Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 12-09-2020

Transcriptional regulation mycosubtilin operon

55

Chapter four

Transcriptional regulation of the lipopeptide synthetasesin Bacillus subtilis ATCC6633

Abstract

The Gram positive bacterium B. subtilis ATCC6633 produces two differentlipopeptides, surfactin and mycosubtilin. In the present study the transcriptionalregulation of the surfactin- and mycosubtilin synthetase operons in B. subtilisATCC6633 was compared to the transcriptional regulation of the surfactin synthetaseoperon in B. subtilis 168. This study revealed that expression of both operons in B.subtilis ATCC6633 is several hundred fold lower than that of the surfactin synthetaseoperon in B. subtilis 168. The promoter activity of the mycosubtilin synthetaseoperon in the B. subtilis 168 derivative, B. subtilis 8G5, did not differ from that in B.subtilis ATCC6633. In addition, and in contrast to the surfactin synthetase operon,which showed highest expression in minimal medium, the mycosubtilin synthetaseoperon showed highest expression in TY-medium. The transcriptional regulation ofthe surfactin synthetase operon in B. subtilis ATCC6633 is comparable to that in B.subtilis 168 and expression is mainly dependent on ComA. CodY does not effect theexpression of the surfactin synthetase operon in B. subtilis ATCC6633. In contrast,expression of the mycosubtilin synthetase operon in B. subtilis ATCC6633 isindependent of ComA but still seems to be regulated via quorum sensing as PhrCstrongly stimulated expression. Spo0H also influenced expression of themycosubtilin synthetase operon and addition of PhrC to the culture mediumbypassed the effects of a spo0H deletion. Finally, AbrB represses expression of themycosubtilin synthetase operon as deletion of abrB resulted in increasedexpression.

Introduction

Many bacteria and fungi produce small,modified peptides that are synthesizednonribosomally by large multienzymecomplexes, peptide synthetases. Owing toimportant medical properties of several ofthese peptides, and the promisingengineering prospects of the correspondingpeptide synthetases, there is a growinginterest in these multienzyme complexes. Allpeptide synthetases exhibit a modularstructure in which each module, consistingof a number of domains, performs all the

necessary reactions to modify andincorporate one specific amino acid into thegrowing peptide chain (Mootz and Marahiel,1997). In addition, an increasing number ofhybrid synthetases have been identified thatconsist of peptide synthetase modulesjoined to fatty acid or polyketide synthetasemodules (Duitman et al., 1999; Du andShen, 2001).

Most nonribosomally synthesizedpeptides produced by the Gram-positivebacterium B. subtilis are cyclic peptides thatcontain a fatty acid modification such assurfactin, fengycin and the members of the

Chapter 4

56

Figure 4.1. Transcriptional regulation of the surfactin synthetase operon in B. subtilis 168or derivative strains.

iturin family (Arima et al., 1968; Maget-Danaand Peypoux, 1994; Vanittanakom et al.,1986). B. subtilis ATCC6633 produces twolipopeptides, surfactin and mycosubtilin(Duitman et al., 1999). Surfactin consists ofa heptapeptide with the sequence Glu - Leu- D-Leu - Val - Asp - D-Leu - Leu closed to alactone ring by a ß-hydroxy fatty acid andexhibits strong antiviral and hemolyticactivities together with a limited antibacterialactivity (Arima et al., 1968; Bernheimer andAvigad, 1970; Maget-Dana and Peypoux,1994; Tsukagoshi et al., 1970). Mycosubtilinconsists of a heptapeptide with thesequence Asn - D-Tyr - D-Asn - Gln - Pro -Glu - D-Ser - Thr closed to an amide ring bya ß-amino fatty acid and exhibits strongantifungal and hemolytic activities (Bessonet al., 1979; Besson et al., 1989; Maget-

Dana and Peypoux, 1994; Peypoux et al.,1986; Walton and Woodruff, 1949).

In contrast to the structure andfunction of peptide synthetases, to date littleresearch has been performed to studytranscriptional regulation of the operonsencoding peptide synthetases. Only thetranscriptional regulation of the surfactinsynthetase operon in Bacillus subtilis 168and derivatives of this strain has beenstudied extensively because of its role indevelopment of genetic competence(Cosmina et al., 1993; Hamoen et al., 1995).Expression of the surfactin synthetaseoperon, srfA, is growth-phase and mediumdependent, increasing sharply at thetransition from exponential to stationarygrowth and is highest when cells arecultured in minimal medium (van Sinderenet al., 1990). Expression of the surfactin

Transcriptional regulation mycosubtilin operon

57

synthetase operon is mainly governed by atwo-component regulatorysystem, ComA and ComP, in whichphosphorylation of ComA stimulates thebinding of this response regulator to thepromoter of srfA and induces the expressionof this operon (Fig. 4.1; Nakano and Zuber,1991; Roggiani and Dubnau, 1993). Thephosphorylation state of ComA is dependenton two processes, phosphorylation by thehistidine kinase ComP, anddephosphorylation by the responseregulator aspartyl-phosphate phosphataseRapC. ComP senses the accumulation ofthe pheromone ComX in the medium and ata critical ComX concentrationphosphorylates itself (Magnuson et al.,1994; Solomon et al., 1996). Subsequently,autophosphorylated ComP, phosphorylatesComA (Hahn and Dubnau, 1991; Weinrauchet al., 1990). RapC performs the oppositereaction and dephosphorylates ComA and,like all members of the Rap family, theactivity of this phosphatase is dependent onthe intracellular concentration of anaccompanying pentapeptide (Grossman,1995; Lazazzera et al., 1997; Solomon etal., 1996). For RapC this is PhrC, previouslyknown as the competence-stimulating factor(CSF) due to its role in development ofgenetic competence ((Grossman, 1995;Lazazzera et al., 1997; Solomon et al.,1996). Expression of phrC is dependent ofSpo0H (σH), a minor sigma factor involved inthe transcription of several genes which areexpressed at the beginning of thesporulation process (Solomon et al., 1996).After processing and secretion into themedium, PhrC is taken up again by theoligopeptide permease Spo0K encoded bythe opp operon ((Grossman, 1995;Lazazzera et al., 1997; Rudner et al., 1991).In addition, expression of srfA is repressedwhen the growth medium containsincreasing concentrations of casamino acidsdue to the activity of the nutritional repressor

CodY, which directly binds to the srfApromoter (Serror and Sonenshein, 1996).The transition state regulator AbrB alsoinfluences expression of srfA, as a deletionof abrB causes a slight decrease inexpression of srfA and constitutiveexpression of spo0H (Weir et al., 1991).Finally, it has been observed that themutation DegUhy, causing hyper-phosphorylation of this response regulator,results in decreased expression of srfA. Themechanism behind this phenomenon is notclear (Hahn and Dubnau, 1991).

Although little information isavailable, other peptide synthetase operonsseem to be governed by comparableregulation cascades. For example,expression of the operon encoding proteinsinvolved in the biosynthesis of bacilysin ofB. subtilis and the lichenysin A synthetaseoperon of B. licheniformis are alsodependent on ComA (Yakimov andGolyshin, 1997; Yazgan et al., 2001). Inaddition, it is known that surfactin and iturinA are coproduced in a number of B. subtilisstrains and that interactions between bothlipopeptides can cause synergistic effectson the biological properties of iturin A(Maget-Dana et al., 1992; Sandrin et al.,1990; Thimon et al., 1992). Together theseresults suggest comparable expressionpatterns and transcriptional regulation ofpeptide synthetase operons in Bacillusspecies.

In the present work we investigatedthe expression and regulation of themycosubtilin- and surfactin synthetaseoperons in B. subtilis ATCC6633 andcompared this with that of both operons inB. subtilis 168-8G5. This was done bymutating known regulators of the surfactinsynthetase operon, and mutating a numberof regulatory proteins involved in generalpost-exponential regulatory pathways of B.subtilis. Finally, by using Tn10 mutagenesisa number of genes were identified that

Chapter 4

58

affected expression of the mycosubtilinsynthetase (mycS) operon.

Materials and methods

General methods and materials. Bacterialstrains and plasmids used in this study arelisted in table 4.1. All molecular cloning andPCR procedures were carried out usingstandard techniques (Sambrook etal.,1989). Restriction endonucleases wereobtained from Roche diagnostics GmbH(Mannheim, Germany) or New EnglandBiolabs (Beverly, USA). All other enzymesand chemicals used in this study wereobtained from Roche diagnostics GmbH(Mannheim, Germany), Sigma (St. Louis,USA) and Merck KGaA (Darmstadt,Germany). Oligonucleotides used for PCRor sequence determination were obtainedfrom Gibco BRL (Paisley, UK) or AmershamPharmacia Biotech (Little Chalfont, UK) andare listed in Table 4.2. Syntheticoligopeptides (Phr's) used in this study wereobtained from Genecor International (PaloAlto, USA). TY-medium for growth of E. coliand B. subtilis, and sporulation- and minimalmedium for growth of B. subtilis wereprepared as described previously (Biswal etal., 1967; Schaeffer et al. 1965; Spizizen,1958). Plasmid DNA isolation used forsequence determination was performed withmini-prep columns from Roche diagnosticsGmbh (Darmstadt, Germany) according tothe manufacturer’s protocol. Plasmid DNAused for transformation of competent B.subtilis, and B. subtilis chromosomal DNAwere purified according to protocols ofBirnboim et al. and Venema et al.,respectively (Birnboim and Doly 1979;Venema et al., 1965). Total RNA isolationsfor RT-PCR were performed using the "Highpure RNA Isolation Kit" from Rochediagnostics. RT-PCR to determine thetranscriptional start of the mycosubtilin

synthetase operon was performed usingsuperscript reverse transcriptase fromRoche diagnostics and primers DF1, DF2and FF2.Transformations. Escherichia coli wastransformed according to the method ofMandel and Higa (Mandel and Higa, 1970).Transformation of competent B. subtilis wasperformed as described by Spizizen(Spizizen, 1958). Poorly competent B.subtilis strains, such as B.subtilis ATCC6633, were transformed usingthe method of Duitman et al., or usingprotoplast transformation as described byChang and Cohen (Chang and Cohen,1979; Duitman et al., submitted forpublication).Reporter gene fusions. To compare theexpression and regulation of the surfactinand mycosubtilin synthetase operon of B.subtilis strain ATCC6633 with that of the twooperons in B. subtilis strains 168, reportergene fusions were made. For this theplasmid pLGW300 was used, whichcontains the ribosomal binding site of the B.subtilis spoVG gene fused to a promoterlesslacZ gene (van Sinderen et al., 1990).

For the construction of themycosubtilin lacZ fusion in B. subtilisATCC6633 an internal part of the mycAgene, obtained by PCR using the primersMA1 and MA2, was digested with BamHIand EcoRI and cloned into pLGW300linearized with the same restrictionenzymes. Using E. coli as an intermediatehost, the obtained plasmids, containing theinternal part of mycA were transformed to B.subtilis ATCC6633 and transformants wereselected on minimal agar plates containing5 µg/ml kanamycin and 4 µg/ml 5-bromo-4-chloro-3-indolyl-ß- D-galactopyranoside (X-gal). Transformants were tested for theabsence of mycosubtilin production, andproper integration was verified by PCRusing primers LZ1, MA1.

Transcriptional regulation mycosubtilin operon

59

Table 4.1: Strains and plasmids.Relevant genotype/characteristic Source or reference

Strains:B. subtilis (168-)7G5 derivative of B. subtilis 168, surfactin+ van Sinderen et al., 1993B. subtilis (168-)8G5 derivative of B. subtilis 168, surfactin- Bron and Venema, 1972B. subtilis AG665 Cmr, ∆spo0H Jaacks et al., 1989B. subtilis ATCC6633 mycosubtilin+, surfactin+ Garrido et al., 1982B. subtilis BD1777 Cmr, ∆comA Guillen et al., 1989B. subtilis BV12E12 Km r, mycA-lacZ This workB. subtilis BV12E15 (8G5) Cmr, Km r, srfAD-lacZ This workB. subtilis BV12E16 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆comA This workB. subtilis BV12E18 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆abrB This workB. subtilis BV12E20 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆sinR This workB. subtilis BV12E22 (ATCC6633) Spr, Km r, mycA-lacZ, cssS::Sp This workB. subtilis BV12E24 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆degU This workB. subtilis BV12E25 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆srfAA This workB. subtilis BV12E27 (8G5) Km r, mycProm-lacZ, surfactin- This workB. subtilis BV12E28 (7G5) Km r, mycProm-lacZ, surfactin+ This workB. subtilis BV12E29 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆spo0K This workB. subtilis BV12E31 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆spo0H This workB. subtilis BV12E32 (ATCC6633) Cmr, Km r, srfAD-lacZ, ∆spo0K This workB. subtilis BV12E33 (ATCC6633) Cmr, Km r, srfAD-lacZ, ∆spo0H This workB. subtilis BV12E35 (8G5) Km r, mycProm-lacZ, ∆comA, surfactin- This workB. subtilis BV12E39 (8G5) Km r, mycProm-lacZ, ∆abrB, surfactin- This workB. subtilis BV12I11 (ATCC6633) Cmr, Km r, srfAD-lacZ, ∆comA This workB. subtilis BV12I37 (ATCC6633) Cmr, Km r, mycA-lacZ, ∆codY This workB. subtilis BV15D29 Spr, Km r, cssS::Sp Hyyrylainen et al., 2001B. subtilis IS432 Cmr, ∆sinR Gaur et al., 1991B. subtilis JH12586 Cmr, ∆abrB Perego et al., 1988B. subtilis KI566 Cmr, ∆Spo0K Rudner et al., 1991E. coli MC1061 F+, araD139, ∆ (ara-leu)7696, ∆ (lac)X74, galU,

galK, hsdR2, mcrA, mcrB, rspL.Wertman et al., 1986

E. coli XL-Blue endA1 gyrA96 thi hsdR17(rK- mK-) supE44relA1 lac/F’ proAB lacIq lacZ∆M15 Tn10

Stratagene

S. cerevisiae G910 Veenhuis et al., 1986Plasmids:pGSP12 Em r, contains comK under the control of its

own promotervan Sinderen et al., 1995

pHV1248 Cm r, Em r, contains Tn10 mini-transposon Petit et al., 1990pLGW300 Km r, contains a promoterless spo0V-lacZ

fusion.van Sinderen et al., 1990

pUC18 Ap r Yanisch-Perron et al., 1985pUC19C Ap r, Cm r Lab collection, unpublished

Also a transcriptional lacZ fusion wasmade downstream of the mycC gene in B.subtilis ATCC6633, which still producesmycosubtilin, to determine the possibleeffects of mycosubtilin production onexpression of the mycosubtilin synthetaseoperon. In this case the 3'-part of the mycCgene, obtained by PCR using primers MC1and ML1, and digested with BamHI, wasligated into pLGW300, linearized withBamHI and SmaI, and transformed to E.

coli. The correct orientation of the insert wasverified with PCR using primers LZ1 andMC1, and plasmid DNA was isolated fortransformation of B. subtilis ATCC6633. B.subtilis ATCC6633 transformants wereselected on minimal agar plates containing5 µg/ml kanamycin and 4 µg/ml X-gal. Thetransformants obtained were tested for theproduction of mycosubtilin and correctintegration was verified by PCR usingprimers LZ1 and MC2.

Chapter 4

60

Table 4.2: Oligonucleotidesgene or region: sequence:abrB 5'- GGA AAC CCT CAC TGC GAA AGA AC -3'abrB 5'- GCT GTT ATT TCG GTA GTT TCC AAG AC -3'cmr 5'- GAC AAT TGG AAG AGA AAA GAG -3'cmr 5'- GAA TGT TTT AGA TAC ACC ATC AAA AAT TG -3'cmr * 5'- CTC TTT TCT CTT CCA ATT GTC -3'cmr * 5'- TAC ATC ATT CTG TTT GTG ATG -3'comA 5'- ACC TGG CCT CGC CGC GGT TTC -3'comA 5'- CAT TCC ATG AGC ATG GGG CTT TCT G -3'codY 5'- AGC TGG CTT TAG GAG AGC TCG AAG -3'codY 5'- AGA ACG ATT CAC CTT TTG ACT GAA G -3'codY 5'- ATC GAG TCT AGA TCA TTA GGA ATG -3'codY 5'- TGC CGC AGC TTG CAG CAT GGA G -3'cssS 5'- GCT CTA GAA TTG CCG TCT CCT CGT ATCG -3'cssS 5'- CGC GGA TCC AGC AGA CCT TGT CAG AGA A -3'dacC-FenF 5'- CCT CCA CCT AGA GCT ATT CAA C -3'dacC-FenF 5'- GAT GCG TTT GTA TGT CTC TGA GGG TG -3'degU 5'- CTA AAA ACA ACC TGG AGA GGG ATC -3'degU 5'- CAT TCG GCT TGC TGG GCA TGA AAG -3'fenF 5'- TCC GTC TGC AGG TCT TGT AGC GGT TGG ATA TAA AC -3'fenF 5'- CATGCTGTTGATCCGAGTTC -3'lacZ 5'- CCA GCT GGC GAA AGG GGG -3'mycA 5'- AAT TTG ATG ATA TCT ATT CTC -3'mycA 5'- CCG GAA TTC GAC CAC TTT CTG TCT CTG G -3'mycC 5'- TTC TGG ATC CAG CAC CTT ATG TTC GAT CAG -3'ppsA 5'- ATA TCC TCG AGA AGG ATA AAC ACA ACC ATC TTC AC -3'sinR 5'- GAA GCT ACA GAG TGG AAC GGC TTG -3'sinR 5'- GGT TGA ATT AAT GGT GGA AGC CAA AG -3'spo0H 5'- GAA ATC GGC CCG GGA GCT TC -3'spo0H 5'- GAG CTG TAT GTG AAT TGC AAG TAC -3'spo0K 5'- GAA TGT TCT GCA TGG CCT AGG CTC -3'spo0K 5'- CTG AGG ATT TAG CCG TAA GGA GCTG -3'spo0K 5'- TTG TGA CGA GGA CTC CTG CTA AAG -3'spo0K 5'- GAG CCT AGG CCA TGC AGA ACA TTC -3'srfAD 5'- GGC GGA ATG ATC ACC TTC AG -3'srfAD 5'- CAG CCG CCA TGA CGA TTC CC -3'yoxA 5'- TAT TTG GGA ACG CCG GCC ATC AAA G -3'

Underlined bases indicate a restriction enzyme recognition site and an asterisk (*) indicates a Cy5 labeledprimer.

To study the expression andtranscriptional regulation of the mycosubtilinsynthetase operon in B. subtilis 168-8G5 weinserted a transcriptional fusion of themycosubtilin promoter region including the5'-part of fenF and a lacZ gene into thegenome of the B. subtilis 168-8G5. As thisstrain does not contain the mycosubtilinsynthetase operon this fusion was insertedat the position identical to that occupied bythe mycosubtilin synthetase operon in B.subtilis strain ATCC6633. As the homology,between B. subtilis ATCC6633 and B.subtilis 168-8G5 is to low in this region, thefollowing approach was used to effect

efficient recombination. First the dacCregion of B. subtilis 168-8G5, obtained byPCR using primers DC1 and YA1, and theplasmid pLGW300 were both restricted withPstI and XhoI, and after ligation the ligationproducts were transformed to E. coli. Thecorrect orientation of the insert was verifiedwith PCR using primers LZ1 and YA1, andcorrect plasmids were isolated. These werelinearized with EcoRI and PstI and ligated inthe presence of 15 % polyethyleneglycol8000 to the mycosubtilin promoter region,obtained by PCR using primers DC2 andFF1, and restricted with EcoRI and PstI.Because plasmids containing the

Transcriptional regulation mycosubtilin operon

61

mycosubtilin promoter were not stable in E.coli, the obtained ligation products weredirectly transformed to competent B. subtilisstrain 168-8G5 cells as described byDuitman et al. (Duitman et al., submitted forpublication). Transformants were selectedon minimal agar plates containing 5 µg/mlkanamycin and 4 µg/ml X-gal, and correctintegration of the lacZ-reporter gene fusionwith the mycosubtilin promoter was verifiedwith PCR using primers LZ1 and YA2.

To examine whether the expressionand transcriptional regulation of the surfactinsynthetase operon in B. subtilis ATCC6633might be comparable to that in B. subtilis168 transcriptional fusions were made in B.subtilis ATCC6633 and B. subtilis 168-8G5.For this the 3'- end of the srfAD gene,obtained by PCR using primers SD1 andSD2, was restricted with BclI and the correctfragment (340 basepairs) was isolated fromgel and ligated into pLGW300 linearizedwith BamHI. The ligation products weretransformed to E. coli and correct orientationof the inserts was determined with PCRusing primers SD1 and LZ1. From thetransformants harbouring the correctplasmids, plasmid DNA was isolated totransform B. subtilis ATCC6633, B. subtilis168-8G5 and B. subtilis 168-7G5.Transformants were selected on minimalagar plates containing 5 µg/ml kanamycinand 4 µg/ml X-gal, and correct integration ofthe lacZ-reporter gene fused after the srfADgene was verified with PCR using primersSD2 and LZ1.ß-galactosidase assays. ß-galactosidaseactivity measurements were used asmeasures for the expression levels of thelipopeptide synthetase operons in strainsharbouring the lacZ fusions. For this B.subtilis was cultured in TY-, sporulation- orminimal medium at 37ºC under continuousshaking at 300 rpm. Overnight cultures werediluted 100-fold with the same medium andsamples were taken at hourly intervals for

optical density reading at 600 nm and ß-galactosidase activity determinations at 420nm. Calculation of ß-galactosidase unitswere carried out essentially as described byMiller (Miller, 1979), and expressed as unitsper OD600 (U/ OD600).Mutational analysis. To determine thepossible involvement of known regulatoryproteins in the transcriptional regulation ofthe surfactin and mycosubtilin synthetaseoperons in B. subtilis ATCC6633, the genesencoding these regulators were mutated inthe strains harbouring lacZ reporter genefusions. This was done by transformationwith chromosomal DNA of B. subtilis strainsalready harbouring the desired mutationmarked with an antibiotic resistance gene,and deletion were verified with PCR usingprimers AB1, AB2, CA1, CA2, CM1, CM2,CS1, CS2, DC1, SH1, SH2, SK1, SK2, SK3,SK4, SR1 and SR2. If this approach was notsuccessful or no suitable strain wasavailable, as was the case for codY anddegU, mutations were introduced as follows.From the plasmid pUC18 the lacZ gene wasalmost completely removed by digestionwith NdeI and EcoRI and the remaininglinearized plasmid was made blunt withKlenow enzyme. The linearized plasmid wasligated to a DNA fragment, obtained withPCR using primers CY1, CY2, DU1 andDU2, consisting of the gene to be deletedwhich was flanked by 1000 basepairs toeffect replacement recombination. Aftertransformation to E. coli, plasmids carryingthe insert were digested with restrictionenzymes to remove the gene to be deleted(degU) and made blunt using klenowenzyme or T4 DNA polymerase. If this wasnot possible (codY), the plasmid was usedas template for PCR, using primers CY3and CY5 located at the 5'- and 3'-ends ofcodY, which produced a linearized plasmidthat harbours the flanking sequences butlacks codY. The linear plasmid DNA wasthen ligated to a chloramphenicol resistance

Chapter 4

62

marker including the necessary signals forexpression, which was obtained frompUC19C by restriction with HindII, andtransformed to E. coli. Transformants wereselected on TY-agar plates with 5 µg/mlchloramphenicol and the orientation of thechloramphenicol resistance marker wasdetermined with PCR using primers CY1,CY2, DU1, DU2 and CM1. Plasmids inwhich the chloramphenicol marker had thesame orientation as the deleted gene wereisolated and transformed to B. subtilisstrains harbouring the lacZ reporter genefusions. Correct integration and deletion ofthe desired gene, which could only resultfrom double cross-over events, was verifiedwith PCR using primers CY1, CY2, DU1,DU2, CM1 and CM2.Tn10 mutagenesis. To identify unknownregulators of the mycosubtilin synthetaseoperon, B. subtilis ATCC6633 wassubjected to random mutagenesis using theminitransposon Tn10 active in B. subtilis asdescribed by Petit et al. (Petit et al., 1990).As B. subtilis ATCC6633 turned out to lyseat 51ºC, Tn10 mutagenesis was performedin B. subtilis 168-8G5 containing themycosubtilin promoter-lacZ fusion.Screening for regulation mutants thatexhibited a lower- or higher expression ofthe mycosubtilin synthetase operon wasdone by means of blue/white screening onminimal-agar plates with 5 µg/mlchloramphenicol and X-gal. From mutantswith altered levels of reporter geneexpression, chromosomal DNA was isolatedfor further analyses. The chromosomal DNAwas transformed to B. subtilis 168-8G5containing the mycosubtilin promoter lacZfusion, and transformants were selected onminimal agar plates containing 5 µg/mlchloramphenicol and 4 µg/ml X-gal. Thiswas done to ensure linkage between theobserved phenotype and the insertion of theTn10 mini-transposon. If the phenotype wasconfirmed the chromosomal DNA was

restricted with BamHI or EcoRI and ligatedinto pUC18 linearized with BamHI or EcoRI.The ligation products were transformed to E.coli XL1-blue and transformants wereselected on TY-agar plates containing 5µg/ml chloramphenicol. Transformants werecultured overnight in TY medium with 5µg/ml chloramphenicol, and the plasmidswere isolated for sequence determinationusing the ALFexpress II automatedsequence system from AmershamPharmacia Biotech. Reactions wereperformed using the Thermo sequenaseCy5 Terminator Kit on the DNA 625Labstation both from Amersham PharmaciaBiotech (Little Chalfont, UK). Theoligonucleotides used for sequencing wereCM3 and CM4, located in thechloramphenicol gene of the mini-transposon Tn10.Bioassays for lipopeptide production.The production of the lipopeptides surfactinand mycosubtilin were measured usingbioassays essentially performed asdescribed by Besson et al. and Mulligan etal., respectively (Besson et al., 1979:Mulligan et al., 1984). For mycosubtilinproduction Saccharomyces cerevisiae G910and for surfactin production sheep bloodwere used as indicators (Veenhuis et al.,1987).

Results

Expression of the surfactin synthetaseoperon in B. subtilis ATCC6633 is low. InB. subtilis 168, expression of the surfactinsynthetase operon, srfA, is highest inminimal medium and induced at the end ofthe exponential growth phase (van Sinderenet al., 1990). To examine whether theexpression of srfA in B. subtilis ATCC6633shows a similar medium- and growth phasedependency, srfAD-lacZ reporter genefusions were made in B. subtilis ATCC6633

Transcriptional regulation mycosubtilin operon

63

Figure 4.2. Activity of the promoter of the surfactin synthetase operon of B. subtilis ATCC6633 (A) and B.subtilis 8G5 (B) in different media. Samples were harvested at hourly intervals for cell density determination(OD600) and specific ß-galactosidase activity determination (A420). The time scale refers to hours before andafter the transition from the exponential to the stationary growth phase (defined as T0). Expression of srfAC-lacZ was measured in TY-medium (□), sporulation medium (△) and minimal medium (○).

and B. subtilis 168-8G5, and the activity ofthe promoter of srfA was measured in TY-,minimal- and sporulation medium at hourlyintervals during 12 hours after inoculation(Fig. 4.2). The results show that theexpression of srfA in B. subtilis ATCC6633,in the three used media, is about 200-foldlower than that in B. subtilis 8G5 (Fig. 4.2;van Sinderen et al., 1990). Althoughdramatically low in B. subtilis ATCC6633,the growth phase- and medium dependencefor expression of srfA is comparable in bothstrains. Also, in B. subtilis ATCC6633expression of srfA is induced at the end ofthe exponential growth phase and reaches amaximum level 2 to 4 hours after thetransition from exponential to stationarygrowth. As in B. subtilis 168-8G5,expression is the highest in minimal mediumwhere it reaches a maximum level of about1.5 U/OD600. Expression in richer mediasuch as TY- and sporulation medium is

lower and reaches maximum levels of about0.8 U/OD600.Expression of the surfactin synthetaseoperon in B. subtilis ATCC6633 is ComAdependent. Although the medium- andgrowth phase dependence for theexpression of srfA in B. subtilis ATCC6633and B. subtilis 168-8G5 appear to besimilar, the expression levels of the surfactinsynthetase operon show large differences.To study whether regulation of srfA in thetwo strains share similar components,known regulators of this operon in B subtilis168 were deleted from B. subtilisATCC6633 containing the srfAD-lacZreporter gene fusion, and the effects onexpression in minimal medium weremeasured. First, the gene encoding themain activator of srfA, comA, was deletedand expression of srfA in this mutant wasexamined (Fig. 4.3A). Clearly deletion ofcomA completely abolished surfactin

Chapter 4

64

Figure 4.3. Effect of various mutations on expression of the surfactin synthetase operon of B. subtilisATCC6633 measured in minimal medium. Samples were harvested at hourly intervals for cell densitydetermination (OD600) and specific ß-galactosidase activity determination (A420). The time scale refers to hoursbefore and after the transition from the exponential to the stationary growth phase (defined as T0). A) (□) wildtype, (△)△comA and (○)△codY. B) (□) wild type, (△)△spo0H and (○)△spo0K.

expression in minimal medium. Figure 4.3Aalso shows that deletion of the geneencoding the nutricial repressor CodYresulted in only a marginally increasedexpression of srfA in minimal medium underthe used conditions (Fig. 4.3A). In minimalmedium enriched with 2- or 4-fold increasedconcentrations of casamino acids, this effectwas somewhat stronger, but still marginal(Data not shown). We also deleted spo0Hand spo0K that effect expression of srfA inB. subtilis 168 in an indirect way. Inaccordance with these results, deletion ofspo0H indeed strongly decreased theexpression of srfA (Fig. 4.3B). Howevercompared to the effect of a deletion ofspo0H, partial deletion of spo0K had lesseffect on expression of srfA. Together theseresults indicate that the transcriptionalregulation of srfA in B. subtilis ATCC6633 ispartially similar with that of this operon in B.subtilis 168. In both strains transcription ofsrfA is dependent on ComA, but the roles of

CodY and Spo0K in B. subtilis ATCC6633seem less important than in B. subtilis 168Expression of the mycosubtilinsynthetase operon differs from thesurfactin synthetase operon. To studywhether expression of the mycosubtilinsynthetase operon, mycS, shows similarmedium- and growth phase dependence assrfA, a mycA-lacZ reporter gene fusion wasmade in B. subtilis ATCC6633. In addition,as the level of expression of srfA in B.subtilis 8G5 is much higher compared tothat in B. subtilis ATCC6633 we also studiedwhether increased expression of mycS inB. subtilis 168-8G5 might be observed.To this purpose, the mycS promoter regionwas fused to a lacZ reporter gene andintegrated into the B. subtilis 168-8G5chromosome at a position identical to thatoccupied by mycS in the B. subtilisATCC6633 chromosome.

To identify the mycS promoter regionthe transcriptional start of mycS had to

Transcriptional regulation mycosubtilin operon

65

be determined. However, probably due tothe very low activity of the mycS promoter,primer extension analysis was notsuccessful. Subsequently, RT-PCR wasperformed which identified a 30 basepairregion, 180 basepairs upstream of the startcodon of the fenF gene, harbouring thetranscriptional start (Fig. 4.4). A perfect -10sequence for a σA-dependent promoter ispresent in this region, which coincides withthe approximate position of thetranscriptional start, but no clear -35sequence for such a promoter could bedetected. However, to ensure the integration

of the complete promoter region in B.subtilis 168-8G5, the complete intergenicregion between dacC and fenF was used forthe lacZ reporter gene fusion (Duitman etal., 1999).

Using these lacZ reporter genefusions the expression of mycS in B. subtilisATCC6633 and B. subtilis 168-8G5 wasmeasured in TY-, minimal- and sporulationmedium at hourly intervals during 12 hoursafter inoculation (Fig. 4.5). In B. subtilisATCC6633 expression of mycS is highest inTY-medium where it reaches a maximumlevel of about 1.8 U/OD600 (Fig. 4.5A).

=========> <=========dacC AGTCAATAATAGCGTCAGACCTGTTTTCCAAGCGAAAACAGGT

TTTTTTATGTCCTTCGTTACGTCAAGCATTCTTTATCATTCCATATATAC

GAATAATATACATCTAACATATAAAACATTAAACGTTTTTAGAAAATTAA

AACAACTGAAAAATTAGTTTAGAAGACCTTGCAAAAAATGGTACTTTCAT

CACTTCATGTTCAAAATTACGACAAAAAACGAAACGAAAACTAAAGAAAT ================> <=============AATACTATTTTCAAAAAAGTACAATGAAAAATTTACATTTTTTATTGTAC===TTTAATAAATATACGTTAATATAGTGCATATATGGATTATATAGCCATAT

AATTTATTTTTCTGTCATTATTTCACTTTCTTAACCTCTTATTTTAGAAC

TGAGGGAATTGTTGAGCCGAACATCTTATTCTTTACCTTGCCCAAAAGAA

AGTACAGTTTATCGACCTGTTGGTTGTTCCAGTGTTTTTTGCAGGTGATT

CCAGCATTCATAGTTGCAGAAGTATTATCTAAGTCATTGTTAAACCTGTA

TCGTCTTAGCCAGCCTTATTTTGGCGTGGTAATTATAGGCCAATCTCAAG

ACTAGCAGGACAAGCTTCCGCAAACAATAGCTTCTAAAGTAGAAACTCCA

TATGTCGCGTTGCTCAATCAAGTAAAGTTTTTTGGCGAGTTGCAACGTCA

TCTGTGTCCGTCCCTGGCAGTTTTATCTGCCGCGGGCTTCTCACCCTTGC

TCTTGTTTTGTTTTCCTCCACCTAGAGCTATTCAACTATAATCAAACGCA

ATCAAATCTTGAACACCCTCAGAGACATACAAACGCATCAATTAAAAAAA

GACGTTTAATCGTTAGGCTTCCATTATTTGAGCTGCAATTATGACAATGA

TCCCATATGCATGTTTTTGTGATGATGATGCTATGACGGTACAAAAGTAT

ATGACATGTATCCGTTCGAAAAGATTGGAGGGAGCTAATGAAT fenF

Figure 4.4. Mycosubtilin promoterregion. The underlined basesindicate the region of thetranscriptional start determined withRT-PCR, the gray areas indicatethe end of the dacC gene and thebeginning of the fenF gene, andarrows (===> <===) indicateputative transcriptional terminatorsor inverted repeats with unknownfunctions. The bold and underlinedbases indicate the putative -10recognition sequence for a σA

dependent promoter.

Chapter 4

66

Figure 4.5. Promoter activity of the mycosubtilin synthetase operon in B. subtilis ATCC6633 (A) and B. subtilis8G5 (B) in different media. Samples were harvested at hourly intervals for cell density determination (OD600)and specific ß-galactosidase activity determination (A420). The time scale refers to hours before and after thetransition from the exponential to the stationary growth phase (defined as T0). Expression of mycA-lacZ wasmeasured in TY-medium (□), sporulation medium (△) and minimal medium (○).

In minimal medium almost no expression ofmycS was observed, and in sporulationmedium expression of mycS reaches amaximum level of about 1.0 U/OD600. Incontrast to the expression of srfA, which isexpressed at a very high level in B. subtilis168-8G5 (Fig. 4.2B), the expression level ofmycS in B. subtilis 168-8G5 is comparableto that in B. subtilis ATCC6633 (Fig. 4.5B).Also in B. subtilis 168-8G5, the highestexpression of mycS is in TY-medium, whereexpression reached a maximum level ofabout 1.8 U/OD600, and almost noexpression was detected in minimalmedium. Similar to B. subtilis ATCC6633 inB. subtilis 168-8G5 mycS expression insporulation medium was intermediate andreached a maximum level of about 1.0U/OD600. The growth phase dependence ofexpression of mycS is comparable to that ofsrfA in both B. subtilis ATCC6633 and B.subtilis 8G5. Expression in both strainsstarts at the end of the exponential growth

phase and reaches a maximum level 2 to 4hours after the transition from exponential tostationary growth.Expression of the mycosubtilinsynthetase operon in B. subtilisATCC6633 and B. subtilis 168-8G5 is notComA dependent. The differences inmedium requirements for optimalexpression between srfA and mycS as wellas the differences in expression levelbetween both operons in the two testedstrains suggest differences in the regulationof srfA and mycS. To study thesedifferences in regulation and to identifyunknown regulators of mycS, genesencoding regulatory proteins were mutatedin B. subtilis ATCC6633 containing themycA-lacZ reporter gene fusion. To thispurpose the genes encoding responseregulators involved in post exponential geneexpression, comA, cssS and degU weredeleted but none of these deletions had anyeffect on expression of mycS in TY-medium

Transcriptional regulation mycosubtilin operon

67

Figure 4.6. Effect of various mutations on the expression of the mycosubtilin synthetase operon of B. subtilisATCC6633 measured in TY-medium. Samples were harvested at hourly intervals for cell densitydetermination (OD600) and specific ß-galactosidase activity determination (A420). The time scale refers to hoursbefore and after the transition from the exponential to the stationary growth phase (defined as T0). A) (□) wildtype and (△)△abrB. B) (□) wild type, (△)△spo0H, and (○)△spo0K.

(data not shown). Also the genes encodingthe transition state regulator abrB, thenutritional repressor codY and thepleiotropic regulator of late growthprocesses sinR were deleted and the effectson expression of mycS in TY-medium weredetermined. Only deletion of the abrB generesulted in a 4- to 5-fold increase inexpression of mycS in B. subtilisATCC6633, suggesting an important role ofthis protein in transcriptional regulation ofmycS (Fig. 4.6A). As the expression ofmycS in the abrB deletion strain remainedgrowth phase dependent, AbrB alone doesnot seem to be pivotal to the growth phasedependency of the expression of mycS.Similar results were obtained in B. subtilis168-8G5 containing the mycS promoterregion-lacZ reporter gene fusion, and also inthis strain deletion of comA did not effect themycS promoter activity and deletion of abrBresulted in increased promoter activity of themycS promoter (data not shown).

As of all tested known regulatoryproteins, involved in regulation of medium-or growth phase dependent expression, onlyAbrB is involved in regulation of expressionof mycS, we tried to identify other regulatoryproteins of this operon using Tn10 randommutagenesis (Petit et al., 1990). However,this method contains a selection step at51°C and at this temperature B. subtilisATCC6633 cultures turned out to lysecompletely. For this reason and becausetranscriptional regulation of mycS seemedto be similar in B. subtilis ATCC6633 and B.subtilis 168-8G5, we used B. subtilis 168-8G5 containing the mycS promoter region-lacZ reporter gene fusion. Using this methodwe identified the following genes encodingproteins that decreased the expression ofmycS: leuB, spo0H, ybfG, yhjR and ykrB.The mechanism underlying the effects ofLeuB, YbfG, YhjR and YkrB on expressionof mycS is unclear and possibly indirect asLeuB is involved in leucine biosynthesis,

Chapter 4

68

Figure 4.7. Effect of addition of various Phr's on the expression of the mycosubtilin synthetase operon of B.subtilis ATCC6633 (A) and B. subtilis ATCC6633 containing a spo0H deletion (B) measured in TY-medium.Samples were harvested at hourly intervals for cell density determination (OD600) and specific ß-galactosidaseactivity determination (A420). The time scale refers to hours before and after the transition from the exponentialto the stationary growth phase (defined as T0). A) (□) wild type, (△) wild type + 10 µM PhrC. B) (□) wild type,(△)△spo0H + 0 µM PhrC, (○)△spo0H + 1 µM PhrC, (◇)△spo0H + 5 µM PhrC.

YbfG and YhjR are putative membraneproteins and YkrG has homology withformylmethionine deformylases (Kunst etal., 1997).

The gene spo0H encoding the sigmafactor H (σH) is involved in the expression ofsrfA in B. subtilis 168. This sigma factor isessential for the expression of PhrC, whichstimulates srfA as it inhibits thephosphatase of ComA, RapC (Solomon etal., 1996). To examine the influence of σH,we deleted spo0H in B. subtilis ATCC6633containing the lacZ reporter gene fusion inmycA. In addition, spo0K was deleted andthe effect on expression of mycS in TY-medium of both mutants was measured(Fig. 4.6B). Indeed, the expression of mycSdecreased by a factor 2 when spo0H wasdeleted, but partial deletion of spo0K did notinfluence expression of mycS. These resultsare comparable to the results obtained withsrfA in B. subtilis ATCC6633.

As has previously been shown forsrfA, the effects of σH on expression ofmycS in B. subtilis ATCC6633 might be dueto the absence of PhrC, as the phrCpromoter is a σH dependent promoter. If thiswould also apply for mycS, addition ofsynthetic PhrC should result in increasedexpression of this operon and bypass theeffects of a spo0H mutation. To examinethis possibility, 10 µM synthetic PhrC wasadded to cultures of B. subtilis ATCC6633,containing a mycA-lacZ reporter gene fusionand the effects on expression of mycS inTY-medium was measured. Also 10 µM ofall other chromosomally encoded Phr's(PhrA, E, F, G, I, K) were added todetermine whether they influencedexpression of mycS in TY-medium.Although some of the other synthetic Phr'sdid cause a slight decrease of expression ofmycS (data not shown), only 10 µM yieldeda twofold increased expression of mycS

Transcriptional regulation mycosubtilin operon

69

(Fig. 4.7A). Addition of increasingconcentrations of PhrC to cultures of B.subtilis ATCC6633 containing a mycA-lacZreporter gene fusion and a deletion ofspo0H, indeed bypassed the effects of theabsence of σH in TY-medium (Fig. 4.7B).Already a concentration of only 1 µM ofPhrC almost completely neutralized theeffect of the spo0H deletion on expressionof mycS. These results clearly demonstratethat the effects of a spo0H deletion on theexpression of mycS was due to therequirement of σH, encoded by spo0H, forthe expression of PhrC. In addition, thisresult demonstrates that the mycS promoteris not σH dependent. If it were, addition ofPhrC would not have bypassed the effectsof the absence of σH on expression of mycS.

Discussion

In the work described here we studiedwhether expression and regulation of srfA

and myc41S in B. subtilis ATCC6633 arecomparable to the expression andregulation of srfA in B. subtilis 168.Regulation of srfA in B. subtilis 168 hasbeen extensively studied because srfA alsoharbours the comS gene, a small proteininvolved in the development of geneticcompetence (Cosmina et al., 1993; Hamoenet al., 1995).

The present study revealed that thelevel of expression of srfA in B. subtilisATCC6633 is about 200-fold less than thatin B. subtilis 168-8G5. This low level ofexpression of srfA, and thus of comS, mightalso explain the near absence of geneticcompetence in B. subtilis ATCC6633, themore so because it has been shown thatintroduction of a comS-bearing plasmid,which results in higher levels of ComS,increased competence development (Liu etal., 1996). In contrast to the level ofexpression, the medium and growth phasedependence of expression were similar inthe two strains. Similar to B. subtilis 168,

also in B. subtilis ATCC6633 expression ofsrfA was highest in minimal medium, startedat the end of the exponential growth phase,and reached a maximum level 2 to 4 hoursafter the transition from the exponential tothe stationary growth phase. Apart from thelevel of expression, the transcriptionalregulation of srfA showed little differences inboth strains.

Earlier studies in B. subtilis 168 haveshown that expression of srfA is mainlydependent on ComA (Nakano and Zuber,1991; Roggiani and Dubnau, 1993).Apparently, this also applies to B. subtilisATCC6633 as a deletion of comA preventedsrfA expression. Deletion of spo0H in B.subtilis ATCC6633 also resulted indecreased expression of srfA, probably dueto the σH dependent expression of PhrC.However, partial deletion of spo0K,encoding the oligopeptide permeasenecessary for the uptake of PhrC, had little

influence on expression of srfA in B. subtilisATCC6633. Presumably, B. subtilisATCC6633 employs another oligopeptidepermease capable of taking up PhrC (Koideand Hoch, 1994). The role of CodY inregulation of srfA in B. subtilis ATCC6633remains uncertain, as a deletion of codYhad little effect on expression under theconditions used.

The expression of mycS in B. subtilisATCC6633 is also low compared toexpression of srfA in B. subtilis 168. Inaddition, also the medium requirements foroptimal expression of the mycS weredifferent. In contrast to srfA, which ismaximally expressed in minimal medium,mycS showed highest expression in TY-medium, in which growth phase dependentexpression was measured. In contrast to theexpression level of srfA, the expressionlevel of mycS in B. subtilis 168-8G5remained similar to that of mycS in B.

Chapter 4

70

subtilis ATCC6633. These results stronglysuggest that the transcriptional regulation ofmycS differs quite considerably from that ofsrfA. This was corroborated by theobservation that expression of mycS isindependent of ComA.

To identify possible other regulatorsof mycS, deletions of the pleiotropicregulators abrB, codY, cssS, degU and sinRwere made. Of these regulators, onlydeletion of AbrB effected expression ofmycS, and resulted in a 3- to 4-fold higherexpression level of this operon. However,expression in the abrB deletion straincontinued to be growth phase dependentsuggesting the involvement of otherregulatory proteins in the temporaryexpression of mycS.

In an effort to identify such unknownregulators Tn10 mutagenesis was adapted.Only one protein was identified with aplausible role in the transcriptionalregulation of mycS of B. subtilis ATCC6633.This was σH encoded by spo0H. The lowerexpression of mycS in a spo0H deletionstrain, which has also been observed forsrfA (Grossman, 1995; Hahn and Dubnau,1991), was apparently caused by theabsence of PhrC, as the addition ofsynthetic PhrC to the culture mediumresulted in increased levels of expressionand bypassed the effects of the spo0Hdeletion. Addition of most otherchromosomal located Phr's (PhrA, E, F, G, I,K) to the culture medium had little if anyeffect on the expression of mycS.

The addition of synthesized Phr's tothe culture medium might be an interestingoption for quick determination whetherexpression of certain genes or operons areunder cell density control in those B. subtilisstrains, which are refractory to genetictransformation. Although we recentlydeveloped two methods that facilitatetransformation of such B. subtilis strains,mutational analyses in these strains can still

be a problem due to lack of homology to B.subtilis 168. By adding synthesized Phr's wecould show that expression of mycS in B.subtilis ATCC6633 is dependent on theconcentration of PhrC in the medium, whichis measured via quorum sensing, and thusdependent on cell density. Finally, a partialdeletion of spo0K did not affect expressionof mycS as strongly as a spo0H deletion inthis strain, which is comparable to theresults obtained with srfA. This lack of effectin comparison to deletion of spo0H suggeststhe presence of another oligopeptidepermease, capable of taking up PhrC, in B.subtilis ATCC6633.

Based on this work, and work done inthe past on the transcriptional regulation ofsrfA in B. subtilis 168 (Grossman, 1995;Hahn and Dubnau, 1991; Lazazzera et al.,1997; Nakano and Zuber, 1991; Roggianiand Dubnau, 1993; Solomon et al., 1996;Weinrauch et al., 1990), we propose amodel for the transcriptional regulation ofsrfA and mycS in B. subtilis ATCC6633 (Fig.4.8). In this model we assume a putativerole for an unknown regulator, which inphosphorylated form binds to themycosubtilin promoter and inducesexpression. We also assume that thisresponse regulator is dephosphorylated byRapC, which activity is inhibited by PhrC(Lazazzera et al., 1997, Solomon et al.,1996). The most likely candidate for thesecond oligopeptide permease, required forthe uptake of PhrC, is the oligopeptidepermease encoded by the app operon,which in B. subtilis 168 is inactive due to thepresence of frame-shift mutations (Koideand Hoch, 1994).

The role of CodY in thetranscriptional regulation of both operonsremains unclear. Based on studies in B.subtilis 168, CodY was expected to represstranscription of srfA (Serror andSonenshein, 1996). However, deletion ofcodY did not result in increased

Transcriptional regulation mycosubtilin operon

71

Figure 4.8. Model for the regulation of expression of the operons encoding the surfactin and mycosubtilinsynthetase in B. subtilis ATCC6633. This model is based on the work described here as well as earlier workperformed in B. subtilis 168 derivatives. The proteins depicted in the gray rectangle are putative and furtherresearch is necessary to determine whether the App operon, RapC and an unknown regulator (Xxx) areinvolved in the transcriptional regulation of the mycosubtilin synthetase operon.

expression of srfA in B. subtilis ATCC6633,suggesting that CodY does not play animportant role in transcriptional regulation ofat least srfA in this strain.

The regulatory role of the otherproteins identified with Tn10 mutagenesison the expression of mycS, LeuB, YbfG andYhjR, remains obscure. Further research isnecessary to determine how these proteins,which are not potential regulatory proteinsbased on homology, do influence theexpression of mycS.

A number of the regulators of mycSidentified in this study are also involved inthe transcriptional regulation of other

peptide synthetase operons in variousBacillus species. It has been shown thatAbrB probably represses expression of thetyrocidine operon of Bacillus brevis and thatAbrB binds to the putative tyrocidinpromoter (Marahiel et al., 1987; Robertsonet al., 1989). In addition, these studiesrevealed that this repression is relieved bythe activity of Spo0A. Also the expression ofanother peptide synthetase operon of B.brevis, the gramicidin synthetase operon, isdependent on Spo0H, although in this casethe gramicidin synthetase operon promoteris a σH-dependent promoter (Marahiel et al.,1993). Finally, expression of the operons

Chapter 4

72

encoding proteins involved in thebiosynthesis of bacilysin of B. subtilis andthe lichenysin A synthetase operon of B.licheniformis are dependent on ComA(Marahiel et al., 1993; Yazgan et al., 2001).These results, together with the resultsobtained in this study, show that thetranscriptional regulation of peptidesynthetase operons is closely linked to thetranscriptional regulation of other starvationinduced processes like the development ofgenetic competence and sporulation.

However, further research remains tobe done to fully understand the complexregulatory pathways of peptide synthetaseoperons and to identify other signals thatmay stimulate production of nonribosomalsynthesized peptides. Often operonsencoding proteins involved in thebiosynthesis of ribosomally synthesizedantibiotic peptides, such as nisin, areautoinduced by the presence of theantibiotic peptide in the medium (Kuipers etal., 1995). However, we neither observeddifference in the expression of srfA nor ofmycS in the presence or absence ofsurfactin and mycosubtilin in the culturemedium (data not shown). Alternatively,lipopeptides could function as siderophores,which would agree with the observation thatthe expression of srfA and mycS isincreased upon addition of higher Fe3+

concentrations. Indeed, addition ofincreasing Fe3+ concentrations resulted inincreased expression of srfA and mycS,although the higher cell densities observedcould explain this induction (data notshown).

Acknowledgements

This work was supported by the EuropeanUnion (EU Grant PL950176). We thank Dr.Eugenio Ferrari (Genecor International, PaloAlto, USA) for the generous gift Phr

pentapeptides. Parts of this work weresupported by the European Union (EUGrant PL950176).

References

Arima, K., Kakinuma, A. and Tamura, G.1968. Surfactin, a crystalline peptidelipidsurfactant produced by Bacillus subtilis:isolation, characterization and its inhibitionof fibrin clot formation. Biophys. Res.Commun. 31: 488-494.]Bernheimer, A. W. and Avigad, L. S.1970. Nature and properties of a cytolyticagent produced by Bacillus subtilis. J. Gen.Microbiol. 61: 361-369.Besson, F., Peypoux, F., Michel, G. andDelcambe. L. 1979. Antifungal activity uponSaccharomyces cerevisiae of iturin A,mycosubtilin, bacillomycin L and of theirderivatives; inhibition of this antifungalactivity by lipid antagonists. J Antibiot. 32:828-833.Besson, F., Quentin, M. J. and Michel, G.1989. Action of mycosubtilin on erythrocytesand artificial membranes. Microbios 59:137-143.Birnboim, H. C. and Doly, J. 1979. A rapidalkaline extraction procedure for screeningrecombinant plasmid DNA. Nucleic AcidsRes. 7: 1513-1523.Biswal, N., Kleinschmidt, A. K., Spatz, H.C. and Trautner, T. A. 1967. Physicalproperties of the DNA of bacteriophageSP50. Mol. Gen. Genet. 100: 39-55.Bonmatin, J. M., Genest, M., Labbe, H.and M. Ptak, M. 1994. Solution three-dimensional structure of surfactin: a cycliclipopeptide studied by 1H-NMR, distancegeometry, and molecular dynamics.Biopolymers 34: 975-986.Bron, S. and Venema, G. 1972. Ultravioletinactivation and excision repair in Bacillussubtilis. I. construction and characterizationof a transformable eightfold auxotrophic

Transcriptional regulation mycosubtilin operon

73

strain and two ultraviolet-sensitivederivatives. Mutat. Res. 15: 1-10.Chang, S. and Cohen, S. N. 1979. Highfrequency transformation of Bacillus subtilisprotoplasts by plasmid DNA. Mol. Gen.Genet. 168: 111-115.Cosmina, P., Rodriguez, F., de Ferra, F.,Grandi, G., Perego, M., Venema, G. andvan Sinderen, D. 1993. Sequence andanalysis of the genetic locus responsible forsurfactin synthesis in Bacillus subtilis. Mol.Microbiol. 8: 821-831.Duitman, E. H., Hamoen, L. W., Rembold,M., Venema, G., Seitz, H., Saenger, W.,Bernhard, F., Reinhardt, R., Schmidt, M.,Ullrich, C., Stein, T., Leenders, F. andVater, J. 1999. The mycosubtilin synthetaseof Bacillus subtilis ATCC6633: amultifunctional hybrid between a peptidesynthetase, an amino transferase, and afatty acid synthase. Proc. Natl. Acad. Sci.USA 96: 13294-13299.Duitman, E. H., Venema, G., Boven, L. G.and Hamoen, L. W. 2002. Molecular toolsto facilitate transformation of Bacillussubtilis. (submitted for publication).Du, L. and Shen, B. 2001. Biosynthesis ofhybrid peptide-polyketide natural products.Curr. Opin. Drug. Discov. Devel. 4: 215-28.Garrido, N., Becerra, J., Marticorena, C.,Oehrens, E., Silva, M. and Horak, E. 1982.Antibiotic properties of ectomycorrhizae andsaprophytic fungi growing on Pinus radiataD. Don I. Mycopathologia 77: 93-98.Gaur, N.K., Oppenheim, J. and Smith, I.1991. The Bacillus subtilis sin gene, aregulator of alternate developmentalprocesses, codes for a DNA-binding protein.J. Bacteriol. 173: 678-86Grossman, D. 1995. Genetic networkscontrolling the initiation of sporulation andthe development of genetic competence inBacillus subtilis. Annu. Rev. Genetics 29:477-508.Guillen, N., Weinrauch, Y. and Dubnau, D.A. 1989. Cloning and characterization of the

regulatory Bacillus subtilis competencegenes comA and comB. J. Bacteriol. 171:5354-5361.Hahn, J. and Dubnau, D. 1991. Growthstage signal transduction and therequirements for srfA induction indevelopment of competence.J. Bacteriol. 173: 7275-7282.Hamoen L. W., Eshuis, H., Jongbloed, J.,Venema, G. and van Sinderen, D. 1995. Asmall gene, designated comS, locatedwithin the coding region of the fourth aminoacid-activation domain of srfA, is requiredfor competence development in Bacillussubtilis. Mol. Microbiol. 15: 55-63.Hyyrylainen, H. L., Bolhuis, A.,Muukkonen, L., Koski, P., Vitikainen, M.,Sarvas, M., Pragai, Z., Bron. S., van Dijl,J. M. and Kontinen, V. P. 2001. A noveltwo-component regulatory system inBacillus subtilis for the survival of severesecretion stress. Mol. Microbiol. 41: 1159-1172.Jaacks, K. J., Healy, J., Losick, R. andGrossman, A. D. 1989. Identification andcharacterization of genes controlled by thesporulation-regulatory gene spo0H inBacillus subtilis. J. Bacteriol. 171: 4121-4129.Koide, A. and Hoch, J. A. 1994.Identification of a second oligopeptidetransport system in Bacillus subtilis anddetermination of its role in sporulation. Mol.Microbiol. 13: 417-426.Kunst, F., Ogasawara, N., Moszer, I. et al.1997. The complete genome sequence ofthe Gram-positive bacterium Bacillussubtilis. Nature 390: 249-256.Lazazzera, B. A., Solomon, J. M. andGrossman, A. D. 1997. An exportedpeptide functions intracellularly to contributeto cell density signaling in B. subtilis. Cell89: 917-925.Liu, L. I., Nakano, M., Hyoung Lee, O. H.and Zuber, P. 1996. Plasmid-amplifiedcomS enhances competence and

Chapter 4

74

suppresses sinR in Bacillus subtilis. J.Bacteriol. 128: 5144-5152.Maget-Dana, R., Thimon, L., Peypoux, F.and Ptak, M. 1992. Surfactin/iturin Ainteractions may explain the synergisticeffect of surfactin on the biologicalproperties of iturin A. Biochimie 74: 1074-1051. Maget-Dana, R. and Peypoux, F. 1994.Iturins, a special class of pore-forminglipopeptides: biological and physicochemicalproperties. Toxicology 87: 51-174.Magnuson, R., Solomon, J. andGrossman, A. D. 1994. Biochemical andgenetic characterization of a competencepheromone from B. subtilis. Cell 77: 207-216.Mandel, M. and Higa, A. 1970. Calsium-dependent bacteriophage DNA infection. J.Mol. Biol. 53: 159-162.Marahiel, M. A., Nakano, M. M. and Zuber,P. 1993. Regulation of peptide antibioticproduction in Bacillus. Mol. Microbiol. 7:631-636.Marahiel, M. A., Zuber, P, Czekay, G. andLosick, R. 1987. Identification of thepromoter for a peptide antibioticbiosynthesis gene from Bacillus brevis andits regulation in Bacillus subtilis. J. Bacteriol.169: 2215-2222.Miller, J. H. 1979. . Experiments inmolecular genetics. Cold Spring HarborLaboratory Press, Cold Spring Harbor, N. Y.Mootz H. D. and Marahiel, M. A. 1997.Biosynthetic systems for nonribosomalpeptide antibiotic assembly. Curr. Opin.Chem. Biol. 1: 543-551.Mulligan, C. N., Cooper, D. G. andNeufield, R. J. 1984. Selection of microbesproducing biosurfactants in media withouthydrocarbons. J. Ferment. Technol. 62:311-314.Nakano, M. M. and Zuber, P. 1991. Theprimary role of ComA in stablishment of thecompetent state in Bacillus subtilis is to

activate expression of srfA. J. Bacteriol.173: 7269-7274.Perego, M., Spiegelman, G. B. and Hoch,J. A. 1988. Structure of the gene for thetransition state regulator, abrB: regulatorsynthesis is controlled by the spo0Asporulation gene in Bacillus subtilis. Mol.Microbiol. 2: 689-699.Petit, M. A., Bruand, C., Jannière andEhrlich, S. D. 1990. Tn10-derivedtransposons active in Bacillus subtilis. J.Bacteriol. 172: 6736-6740.Peypoux, F., M. T. Pommier, D. Marion,M. Ptak, B. C. Das, and Michel, G. 1986.Revised structure of mycosubtilin, apeptidolipid antibiotic from Bacillus subtilis.J. Antibiotics 39: 636-641.Kuipers, O. P., Beerthuyzen, M. M., deRuyter, P. G., Luesink, E. J. and de Vos,W. M. 1995. Autoregulation of nisinbiosynthesis in Lactococcus lactis by signaltransduction. J. Biol. Chem. 270: 27299-27304.Robertson, J. B., Gocht, M., Marahiel, M.A. and P. Zuber. 1989. AbrB, a regulator ofgene expression in Bacillus, interacts withthe transcription initiation regions of asporulation gene and an antibioticbiosynthesis gene. Proc. Natl. Acad. Sci.USA 86:8457-8461.Roggiani, M. and Dubnau, D. 1993. ComA,a phosphorylated response regulator proteinof Bacillus subtilis, binds to the promoterregion of srfA. J. Bacteriol. 175: 3182-3187.Rudner, D. Z., Le Deaux, J. R., Ireton, K.and Grossman, A. D. 1991. The spo0Klocus of Bacillus subtilis is homologous tothe oligopeptide permease locus and isrequired for sporulation and competence. J.Bacteriol. 173: 1388-1398.Sambrook, J. E., Fritsch, F. and Maniatis,T. 1989. Molecular cloning: a laboratorymanual, 2nd ed. Cold Spring HarborLaboratory Press, Cold Spring Harbor, N. Y.Sandrin. C., Peypoux, F. and Michel, G.1990. Coproduction of surfactin and iturin A,

Transcriptional regulation mycosubtilin operon

75

lipopeptides with surfactant and antifungalproperties, by Bacillus subtilis. Biotechnol.Appl. Biochem. 12: 370-375.Schaeffer, p., Millet, I. and Aubert, J.1965. Catabolite repression of bacterialsporulation. Proc. Natl. Acad. Sci. USA 54:704-711.Serror, P. and Sonenshein, A. L. 1996.CodY is required for nutritional repression ofBacillus subtilis genetic competence. J.Bacteriol. 178: 5910-5915.Solomon, J. M., Lazazzera, B. A. andGrossman, A.D. 1996. Purification andcharacterization of an extracellular peptidefactor that affects two differentdevelopmental pathways in Bacillus subtilis.Genes Dev. 10: 2014-2024.Spizizen, J. 1958. Transformation ofbiochemically deficient strains of Bacillussubtilis by deoxyribonucleate. Proc. Natl.Acad. Sci. USA 44: 1072-1078.Thimon, L., Peypoux, F., Maget-Dana, R.and Michel, G. 1992. Interactions ofbioactive lipopeptides, iturin A and surfactinfrom Bacillus subtilis. Biotechno. Appl.Biochem. 16: 144-151.Tsukagoshi, N., Tamura, G. and Arima, K.1970. A novel protoplast-bursting factor(surfactin) obtained from Bacillus subtilisIAM 1213. II. The interaction of surfactinwith bacterial membranes and lipids.Biochim. Biophys. Acta 196: 211-214.Vanittanakom, N., Loeffler, W., Koch, U.and Jung, G. 1986. Fengycin--a novelantifungal lipopeptide antibiotic produced byBacillus subtilis F-29-3. J. Antibiot. 39: 888-901.van Sinderen, D., Galli, G., Cosmina, P.,de Ferra, F., Withoff, S., Venema, G. andGrandi, G. 1993. Characterization of thesrfA locus of Bacillus subtilis: only thevaline-activating domain of SrfA is involvedin the establishment of genetic competence.Mol. Microbiol. 8: 833-841van Sinderen, D., Withoff, S., Boels, H.and Venema, G. 1990. Isolation and

characterization of comL, a transcription unitinvolved in competence development ofBacillus subtilis. Mol. Gen. Genet. 224: 396-404.van Sinderen, D. and Venema, G. 1994.ComK acts as an autoregulatory controlswitch in the signal transduction route tocompetence in Bacillus subtilis. J. Bacteriol.176: 5762-70.Venema, G., Pritchard, R. H. andVenema-Schröder, T. 1965. Fate oftransforming deoxyribonucleic acid inBacillus subtilis. J. Bacteriol. 89: 1250-1255.Veenhuis, M., Mateblowski, M., Kunau, W.H. and Harder, W. 1987. Proliferation ofmicrobodies in Saccharomyces cerevisiae.Yeast 3: 77-84.Walton, R. P. and Woodruff, H. B. 1949. Acrystalline antifungal agent, mycosubtilin,isolated from subtilin broth. J. Clin. Invest.28: 924-926.Weinrauch, Y., Penchev, R., Dubnau, E.,Smith, I. and Dubnau, D. 1990. Bacillussubtilis regulatory gene product for geneticcompetence and sporulation resemblessensor protein members of the bacterialtwo-component signal-transductionsystems. Genes Dev. 4: 860-872.Weir, J., Predich, M., Dubnau, E., Nair, G.and Smith, I. 1991. Regulation of spo0H, agene coding for the Bacillus subtilis sigma Hfactor. J. Bacteriol. 173: 521-529.Wertman, K. F., Wyman, A. R. andBotstein, D. 1986. Host/vector interactionswhich affect the viability of recombinantphage lambda clones. Gene 49: 253-262.Yakimov, M. M. and Golyshin, P. N. 1997.ComA-dependent transcriptional activationof lichenysin A synthetase promoter inBacillus subtilis cells. Biotechnol. Prog. 13:757-761.Yanisch-Perron, C., Vieira, J. andMessing, J. 1985. Improved M13 phagecloning vectors and host strains: nucleotidesequences of the M13mp18 and pUC19vectors. Gene 33: 103-119.

Chapter 4

76

Yazgan, A., Ozcengiz, G. and Marahiel, M.A. 2001. Tn10 insertional mutations ofBacillus subtilis that block the biosynthesisof bacilysin. Biochim. Biophys. Acta 1518:87-94.