Plasmid 54 (2005) 241–248

www.elsevier.com/locate/yplas

Construction of plasmid-based expression vectorsfor Bacillus subtilis exhibiting full structural stability

Hoang Duc Nguyen a,b, Quynh Anh Nguyen b, Rita C. Ferreira c, Luis C.S. Ferreira c, Linh Thuoc Tran b, Wolfgang Schumann a,¤

a Institute of Genetics, University of Bayreuth, D-95440 Bayreuth, Germanyb Vietnam National University-Ho Chi Minh City, College of Natural Sciences, Faculty of Biology, Ho Chi Minh City, Viet Nam

c Departamento de Microbiologia, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo, SP, Brazil

Received 8 February 2005, revised 4 May 2005Available online 11 July 2005

Communicated by Margarita Salas

Abstract

A series of plasmid-based expression vectors have been constructed allowing stable intracellular expression of recom-binant proteins in Bacillus subtilis strains. These expression vectors are based on the recently described Escherichia coli–B. subtilis shuttle vector pMTLBS72 which replicates as theta circles. Besides the weak constitutive promoter PlepA, weinserted three diVerent controllable promoters: PgsiB which can be induced by heat and acid shock, and by ethanol, PxylAand Pspac which respond to the addition of xylose and IPTG, respectively. The versatility of these expression vectors wasdemonstrated by fusing their promoters to a reporter gene and by overexpression of the HtpG protein with three ofthem. All recombinant vectors exhibited full structural stability. 2005 Elsevier Inc. All rights reserved.

Keywords: Promoters lepA, gsiB, xylA, and spac; htpG

1. Introduction

High-level production of recombinant proteins isa prerequisite for their subsequent puriWcation. In

* Corresponding author. Fax: +49 921 552710.E-mail address: wschumann@uni-bayreuth.de (W. Schumann).

0147-619X/$ - see front matter 2005 Elsevier Inc. All rights reservedoi:10.1016/j.plasmid.2005.05.001

most cases for the production of heterologous pro-teins, Escherichia coli cells are used as a protein fac-tory (see recent review article by Schumann andFerreira, 2004). Attempts to develop Bacillus subtilisas a second protein factory where the recombinantproteins are secreted into the medium have often notbeen successful because of two major reasons: (i)structural instability of the recombinant plasmids

d.

242 H.D. Nguyen et al. / Plasmid 54 (2005) 241–248

and (ii) instability of the recombinant proteinssecreted into the medium due to degradation (Bronet al., 1991b; Ehrlich et al., 1991; Wong, 1995). Mostconvenient vector plasmids have been derived fromnatural plasmids detected in Staphylococcus aureussuch as pUB110 (Gryczan and Dubnau, 1978),pC194 (Ehrlich, 1977), and pE194 (Gryczan andDubnau, 1978). While these vector plasmids repli-cate stably in B. subtilis, addition of recombinantDNA can confer mainly structural and sometimessegregational instability, too. The molecular basisfor the structural instability is related to their repli-cation mode. These plasmids replicate as rolling cir-cles producing single-stranded DNA as anintermediate, and short direct repeats within thissingle-stranded DNA may lead to the deletion ofone of the two repeats and the intervening DNA(Bron et al., 1991a). This observation led to thedevelopment of vectors carrying an expression cas-sette sandwiched between the two halves of a non-essential gene such as amyE (Shimotsu and Henner,1986), thrC (Guérout-Fleury, 1996), lacA (Härtl,2001), and pyrD, gltA, and sacA (Middleton andHofmeister, 2004). These vectors allow stable inte-gration of any DNA sequence into the chromosomeat one of the genes mentioned. The second majorproblem is related to the production of extracellularproteases which recognise and degrade most heter-ologous proteins secreted into the medium. Theproblem has been largely solved by constructing B.subtilis strains carrying six (Wu et al., 1991) or eveneight (Wu and Wong, 2002) protease null mutations.An alternative to circumvent the problem of recom-binant protein instability would be to identify host

species largely devoid of extracellular proteases.Certain strains of Bacillus brevis allow stable secre-tion of recombinant proteins (Takagi et al.,1989a,b). Here, we report on the construction andanalysis of plasmid-based expression vectors whichare devoid of structural instability under the condi-tions tested. Cloning vectors without structuralinstability have already been described based on thenatural plasmids pAM�1 and pTB19 (Jannière,1990; Simon and Chopin, 1988), and pTA1060(Bron et al., 1987).

2. Materials and methods

2.1. Vector construction

We started from the E. coli–B. subtilis shuttlevector pMTLBS72, carrying the replication regionsfrom the newly discovered plasmid pBS72 and frompBR322 (Titok et al., 2003). First, we inserted theeYcient trpA transcriptional terminator sequence atthe AatII restriction site to ensure eYcient termina-tion of transcription immediately downstream of therecombinant genes. For this aim, we used the twocomplementary oligonucleotides (ON) describedpreviously (Kaltwasser et al., 2002), resulting inpHCMC01. This plasmid served as a backbone intowhich four diVerent expression cassettes wereinserted. First, we cloned the weak lepA promoterinto pHCMC01 using ON1 and ON2 (see Table 1),and chromosomal DNA as template resulting inpHCMC02 (Fig. 1A). This promoter acts constitu-tively and drives expression of the bicistronic lepA

Table 1Oligonucleotides used in this study

a GC clamps are shown in boldface; restriction sites are underlined.

Primer Sequence (5�–3�)a Description

ON1 GGCCATGGATCC TCACTCCTACTATTAAACGCAAAATAC 5� lepA promoterON2 GGCCATGGATCC TCACTCCTACTATTAAACGCAAAATAC 3� lepA promoterON3 GGCCATGAGCTC CTA TCG AGA CAC GTT TGG CTG 5� gsiB promoterON4 GGCCATGGATCC TTC CTC CTT TAA TTG GTG TTG GT 3� gsiB promoterON5 GGCCATGAGCTCCTAACTTATAGGGGTAACACTTAA 5� xylose cassetteON6 GGCCATAGATCTGGATCCCATTTCCCCCTTTGATTT 3� xylose cassetteON7 GGCCATGGTACCAGGCCTTACACAGCCCAGTCCAG 5� spac promoterON8 GGCCATGTCGACGGATCCTCACCTCCTTAAGCTTAATTGTTATCC 3� spac promoterON9 GGCCATGGTACCAAGCTAATTCCGGTGGAAACGAG 5� lacI geneON10 GGCCATGAGCTCAGGCCTTAACTCACATTAATTGCG 3� lacI gene

H.D. Nguyen et al. / Plasmid 54 (2005) 241–248 243

operon of B. subtilis (Homuth et al., 1996). Second,the gsiB promoter was chosen. This promoter isrecognised by the alternative sigma factor �B andhas been reported to be expressed at a very low levelunder physiological conditions and being inducedby diVerent stresses such as heat and acid stress, and

ethanol (Maul et al., 1995; Völker et al., 1994), andthe strong ribosome-binding site of gsiB has beenreported to be crucial for increased stability of thegsiB transcript (Jürgen et al., 1998). Using ON3 andON4 (Table 1), the promoter region of gsiB, includ-ing the DNA sequence coding for the strong ribo-

Fig. 1. Genetic and restriction map of the pHCMC expression vectors. (A) pHCMC02, (B) pHCMC03, (C) pHCMC04, and(D) pHCMC05. All vectors replicate both in E. coli and in B. subtilis, and contain the eYcient transcriptional terminator trpAimmediately downstream from the promoters as indicated by a bar. The cloning sites immediately downstream of the promoters areindicated.

244 H.D. Nguyen et al. / Plasmid 54 (2005) 241–248

some-binding site, was ampliWed and inserted intopHCMC01 yielding pHCMC03 (Fig. 1B). Third, thecomplete xylose-inducible cassette was ampliWedusing ON5 and ON6, and plasmid pAX01 (Härtl,2001) as template resulting in pHCMC04 (Fig. 1C).Fourth, the last expression vector contains theIPTG-inducible Pspac promoter, and was assembledby Wrst inserting Pspac using ON7 and ON8 followedby the lacI gene (ON9 and ON10) resulting inpHCMC05 (Fig. 1D). All inserts were veriWed byDNA sequencing. The cloning sites are indicatedimmediately downstream from the promoters.

2.2. Measurement of reporter gene expression

To assess the strength and inducibility of thefour promoters, these were transcriptionallyfused to either the reporter gene lacZ or/andbgaB, both coding for �-galactosidase. We had tochoose two diVerent reporter genes since the �-galactosidase activity encoded by lacZ disappearswithin a few minutes after heat and ethanol chal-lenge, while that coded for by bgaB is heat-stable(Mogk et al., 1996). On the other hand, after chal-lenge of the cells to acid stress, the �-galactosi-dase activity encoded by bgaB turned out to beunstable. To comply with these observations, thebgaB reporter gene was fused to all four promot-ers and lacZ to PgsiB in addition. All transcrip-tional fusions were introduced into B. subtilisstrain 1012 (Saito et al., 1979) and those contain-ing the PgsiB promoter in addition into strain IS58(Smith et al., 1980) as well. All recombinantstrains were grown in LB medium to the mid-log-arithmic growth phase, and aliquots were takenbefore (t D 0) and at diVerent time points afterinduction with the exception of pHCMC02 whichcarries the constitutive lepA promoter. �-Galacto-sidase activity was measured as described (Mogket al., 1996). All assays were repeated at leastthree times, and the replicates were yielded com-parable results. Mean values are given togetherwith the standard deviations.

2.3. Immunoblot analysis

Samples were prepared for sodium dodecyl sul-phate–polyacrylamide gel electrophoresis (SDS–

PAGE) and immunoblot analysis as described pre-viously (Homuth et al., 1996); 5 �g of total cellularprotein was applied per lane. Polyclonal serumagainst HtpG at a dilution of 1:10,000, a donkeyanti-rabbit immunoglobulin G–horseradish perox-idase conjugate (Amersham), and a chemilumines-cence reaction (ECL System, Amersham) wereused for detection.

2.4. Determination of structural and segregational stability

Structural stability was measured by growth of1012 cells harbouring the four diVerent pHCMCplasmids with the bgaB reporter gene in LBmedium in the presence of chloramphenicol(10 �g/ml) for about 100 generations involvingseveral subcultures. This procedure involvedgrowth of the four cultures into stationary phase,dilution to about 1000 cfu/ml which were grownagain into stationary phase, etc. Finally, cells wereplated on LB Xgal medium in the presence of theappropriate inducer and the number of blue colo-nies were counted. Segregational instability wasdetermined for the two plasmids, pHCMC04-bgaB and pHCMC05-bgaB, in a similar way, butcells were grown in the absence of chlorampheni-col. Aliquots were withdrawn at the number ofgenerations elapsed as shown in Fig. 3, plated onLB plates, and replica plated on LB plates con-taining chloramphenicol. This allowed the calcula-tion of antibiotic-resistant colonies remainingafter the indicated number of generations.

3. Results and discussion

3.1. Analysis of transcriptional fusions between the four promoters reporter genes

To analyse the versatility of the four diVerentexpression systems, the lacZ or/and bgaB reportergene was fused to the four promoters used in thiswork, and �-galactosidase activities were mea-sured. It should be emphasised that the enzymaticactivities obtained with lacZ and bgaB cannotdirectly be compared due to the diVerent enzy-matic characteristics. With PlepA, between 1.5 and

H.D. Nguyen et al. / Plasmid 54 (2005) 241–248 245

4.1 units (Table 2) of �-galactosidase activity werefound, conWrming the weak nature of this pro-moter. Therefore, this promoter can be used toconstitutively express genes at a low level, theproducts of which do not cause any harm to thecells. With PgsiB, between 12 and 20 units were mea-sured in the absence of any stress factor with strain1012, which increased about 13-fold after challengewith acid shock (Table 2). In contrast, ethanol andheat shock resulted only in about 3-fold induction.When the �-galactosidase activities were measuredin strain IS58, the background activity was reducedto 3–7 units, and increased about 7-fold after acidshock and 8- and 12-fold after heat and ethanolshock, respectively. This strain was chosen becauseit has been reported to express �B-dependent genesat a low basal level in the absence of stress factorsknown to induce this regulon (M. Hecker, personalcommunication). This observation could be con-Wrmed here, but the induction factor was lower ascompared to strain 1012. These results demon-strate that PgsiB may be a suitable promoter underconditions of acid shock. When two diVerent anti-gens were fused to PgsiB and the B. subtilis cells car-rying the recombinant plasmids fed orally to mice,the animals developed an immune response indi-cating that the promoter is induced in the acidicenvironment of the stomach and/or the largely glu-cose-free small intestine leading to the productionof the antigens (manuscript in preparation).

Table 2

Plasmid pHCMC04 carries an xylose-induciblepromoter and does not produce any �-galactosidaseactivity in the absence of the inducer (Table 2). Afteraddition of 0.5% xylose, up to 9 units were mea-sured. The B. subtilis 168 genome codes for genesinvolved in catabolism of xylose (xylAB), and thesetwo genes are under the transcriptional control ofthe XylR repressor (Kraus et al., 1994), but thisstrain cannot actively transport xylose into the cell(Schmiedel and Hillen, 1996). We asked whether thexylose which must be taken up by the cell to inducethe xylose expression cassette derived from B. mega-terium (Rygus et al., 1991) can be degraded aVectingproduction of the recombinant proteins. To accom-plish this goal, we deleted the xylAB genes (weampliWed about 300 bp up- and downstream of thexylAB genes, fused them to a tetracycline resistancemarker by recombinant PCR, and transformed thelinear amplicon into strain 1012), and compared the�-galactosidase activities produced in the wild-typestrain and its isogenic knockout induced with 0.1%xylose for 5 h. Since only a minor eVect was mea-sured, degradation of xylose by the xylAB operonunder these experimental conditions does not playany signiWcant role (data not shown). It has beenpublished that glucose acts as an anti-inducer bybinding to XylR (Dahl et al., 1995), we askedwhether increased expression could be observed inthe glucose-free NAPS medium (Kim et al., 1996).As shown in Table 2, the measured �-galactosidase

�-Galactosidase activities

Cells were grown in LB medium at 37 °C to the mid-logarithmic growth phase and then induced as indicated. The data given in paren-theses have been obtained by growth in NAPS medium (Kim et al., 1996).

Plasmid Promoter Inducer Reporter gene

Recipient strain

Time (min) after induction

0 5 10 20 30 60

pHCMC02 PlepA — bgaB 1012 2.4 § 0.9 2.6 § 0.5 4.1 § 1.8 3.6 § 1.3 1. 6 § 0.2 1.5 § 0.4pHCMC03 PgsiB Acid shock lacZ 1012 19.8 § 3.5 19.2 § 3.7 39.3 § 5.4 74.4 § 5. 5 188 § 63 255 § 48pHCMC03 PgsiB Acid shock lacZ IS58 7.7 § 0.3 11.1 § 2.0 24.3 § 2.6 33.6 § 4. 2 32.5 § 3.0 56.5 § 5.2pHCMC03 PgsiB Ethanol bgaB 1012 12.0 § 5.3 15.7 § 3.4 19.1 § 5.9 32.2 § 8. 2 20.9 § 1.7 32.8 § 1.3pHCMC03 PgsiB Ethanol bgaB IS58 3.1 § 0.5 20.8 § 5.1 19.2 § 3.7 30.6 § 7. 5 32.8 § 9.2 36.7 § 5.7pHCMC03 PgsiB Heat shock bgaB 1012 12 § 5.3 13 § 1.4 21.9 § 3.4 23.7 § 10. 2 20.4 § 3.0 33.7 § 1.1pHCMC03 PgsiB Heat shock bgaB IS58 3.1 § 0.5 18.6 § 9.0 22.8 § 6.8 20.3 § 3. 2 19.1 § 2.1 23.4 § 3.9pHCMC04 PxylA 0.5% xylose BgaB 1012 0.0 § 0.0

(0.17 § 0.01)4.7 § 1.7

(1.01 § 0.31)7.9 § 1.4

(2.54 § 0.35)9.6 § 1.6

(4.91 § 0.79)9.4 § 2.2

(6.49 § 0.82)8.9 § 0.8

(9.48 § 1.23)pHCMC05 Pspac 0.1 mM

IPTGbgaB 1012 0.0 § 0.0 2.5 § 0.4 2.6 § 0.1 3.5 § 0.7 4. 2 § 0.5 3.9 § 0.9

pHCMC05 Pspac 0.5 mM IPTG

bgaB 1012 0.0 § 0.6 2.9 § 0.9 4.5 § 1.7 5.9 § 1.7 6. 9 § 1.7 6.3 § 1.5

246 H.D. Nguyen et al. / Plasmid 54 (2005) 241–248

activities with plasmid pHCMC04 (given in paren-theses) do not diVer signiWcantly from those wheregrowth occurred in LB medium. We concludethat our LB medium contains only traces of glucoseif any. Plasmid pHCMC05 carries the IPTG-induc-ible promoter Pspac. As already observed withpHCMC04, no �-galactosidase activity was mea-sured in the absence of IPTG, and the activity

Fig. 3. Segregational stability of two recombinant plasmids.B. subtilis 1012 carrying pHCMC04-bgaB and pHCMC05-bgaB was analysed as described under Materials and methods.The number of chloramphenicol-resistant colonies is given.

reached 4.2 units after addition of 0.1 mM and 6.9units after addition of 0.5 mM IPTG (Table 2). Insummary, the vectors pHCMC03 to pHCMC05 canbe induced either by a stress factor or by addition ofan inducer.

3.2. Immunoblot analysis

While measurement of �-galactosidase activitiesjust yields an indication of the promoter strength,direct visualization of a recombinant protein byWestern blot provides a more direct measure ofprotein production levels. To achieve this goal, theheat shock gene htpG (Schulz et al., 1997) wasinserted into all three of them, while the empty vec-tor served as a control. B. subtilis strain 1012 carry-ing pHCMC03-htpG or pHCMC01 was grown insupplemented LB medium (see legends to Fig. 2)and challenged with the three stresses describedabove. Due to the presence of the wild-type copy ofhtpG in the chromosome, there is a backgroundlevel of HtpG protein (Fig. 2A, lanes 1, 3, 5, and 7)which increased after addition of the second copyon the plasmid (lane 2). After challenge with acidstress, the amount of HtpG protein was enhancedwhile that in the control strain remained constant

Fig. 2. Detection of the HtpG protein by immunoblot analysis. B. subtilis strain 1012 carrying one of the plasmids pHCMC01,pHCMC03-htpG, pHCMC04-htpG, or pHCMC05-htpG was grown either in LB medium (pHCMC01, pHCMC04-htpG, andpHCMC05-htpG), or LB medium supplemented with 0.5% glucose and 1.5 mM KH2PO4 (Mueller et al., 1992) at 37 °C to the mid-exponential growth phase. Then, the cultures were divided into subcultures where one was further grown without induction, whilethe others were either challenged with diVerent stress conditions (A), or induced with 1% xylose or 1 mM IPTG (B). Aliquots werewithdrawn immediately before treatment (t D 0) and 30 (t D 30), and 60 min (t D 60) after treatment. All samples were preparedfor sodium dodecyl sulphate–polyacrylamide gel electrophoresis and for immunoblot analysis as described previously (Homuthet al., 1996). Equal amounts of protein were applied per lane (3.5 �g in (A) and 4.7 �g in (B)), and the blots were probed with �HtpG. (A) pHCMC01 (lane 1) and pHCMC03-htpG (lane 2), untreated; pHCMC01 (lane 3) and pHCMC03-htpG (lane 4), acidshock; pHCMC01 (lane 5) and pHCMC03-htpG (lane 6), 4% ethanol; and pHCMC01 (lane 7) and pHCMC03-htpG (lane 8),48 °C. (B) pHCMC04-htpG uninduced, t D 0 (lane 1), t D 30 (lane 2), and t D 60 (lane 3); pHCMC05-htpG induced with 1% xylose,t D 30 (lane 4) and t D 60 (lane 5); pHCMC05-htpG uninduced, t D 0 (lane 6), t D 30 (lane 7), and t D 60 (lane 8); and pHCMC05-htpG induced with 1 mM IPTG, t D 30 (lane 9), and t D 60 (lane 10).

H.D. Nguyen et al. / Plasmid 54 (2005) 241–248 247

(compare lanes 3 and 4). A signiWcant strongerincrease was observed after ethanol and heat treat-ment (lanes 6 and 8). When the strain carryinghtpG fused to the xylose-inducible promoter(pHCMC04-htpG) was analysed after addition of1% xylose, a signiWcant amount of HtpG proteinaccumulated within 60 min (Fig. 2B, lane 5). Incontrast, lower amounts of HtpG accumulatedafter induction of the strain carrying pHCMC05-htpG with IPTG (lanes 7–10). In summary, all threeexpression vectors allow increased expression ofhtpG after challenge with the appropriate inducer,where the strain carrying PgsiB yielded the highestamount of recombinant protein after ethanol andheat stress, followed by the strain carrying thexylose cassette. To our surprise, the IPTG-induc-ible cassette produced only modest amounts ofrecombinant protein.

3.3. Structural and segregational stability of the pHCMC plasmids

To check for structural stability of the pHCMCplasmids, all four vectors containing thebgaB reporter gene were grown in LB mediumcontaining chloramphenicol for about 100 genera-tions (involving several subcultures) and thenplated on LB plates containing Xgal. It turned outthat with all four recombinant plasmids, all colo-nies exhibited a blue phenotype demonstratingstructural stability of these four recombinant plas-mids (data not shown). We also checked for segre-gational stability using pHCMC04-bgaB andpHCMC05-bgaB. Again, B. subtilis 1012 carryingthese plasmids was grown in LB medium in theabsence of chloramphenicol for up to 100 genera-tions, aliquots were taken after diVerent genera-tions, plated on LB plates, and checked for thepresence of the plasmid by replica plating onLB plates containing chloramphenicol. WhilepHCMC04-bgaB turned out to be fully stable overat least 100 generations (Fig. 3), which is in fullagreement with published data for the originalplasmid pMTLBS72 (Titok et al., 2003), the secondplasmid, pHCMC05-bgaB, suVered from segrega-tional instability (Fig. 3). After 60 generations,about 60% of the cells had lost the plasmid.The reason for this diVerence is not known, but

this result indicates that cells carrying certainrecombinant pHCMC plasmids should be grownin the presence of chloramphenicol to avoid loss ofthe plasmids.

For the Wrst time, four diVerent plasmid-basedexpression systems have been constructed andshown to exhibit structural stability for at least 100generations of growth. Experiments are in progressto improve this expression system to obtain alower background level in the absence of inducer,and a higher induction factor by increasing thepromoter and ribosome-binding strength, and thestability of the transcript. Furthermore, the expres-sion systems will be converted into expression-secretion systems. All four vector plasmids can beordered from the Bacillus Genetic Stock Center(http://bacillus.biosci.ohio-state.edu). The DNAsequences of the plasmids and their plasmid mapscan be found at http://btbgn1.uni-bayreuth.de/lsgenetik1/frames.htm.

Acknowledgments

We thank the DRL (VNM 01/013), the DAAD(415-probral/ale-02/21612), and the MOST (LifeScience-643204) for Wnancial support.

References

Bron, S., Bosma, P., Van Belkum, M., Luxen, E., 1987. Stabilityfunction in the Bacillus subtilis plasmid pTA1060. Plasmid14, 185–194.

Bron, S., Holsappel, S., Venema, G., Peeters, B.P., 1991a. Plas-mid deletion formation between short direct repeats inBacillus subtilis is stimulated by single-stranded rolling-cir-cle replication intermediates. Mol. Gen. Genet. 226, 88–96.

Bron, S., Meijer, W., Holsappel, S., Haima, P., 1991b. Plasmidinstability and molecular cloning in Bacillus subtilis. Res.Microbiol. 142, 875–883.

Dahl, M., Schmiedel, D., Hillen, W., 1995. Glucose and glucose-6-phosphate interaction with Xyl repressor proteins fromBacillus spp. may contribute to regulation of xylose utiliza-tion. J. Bacteriol. 177, 5467–5472.

Ehrlich, S.D., 1977. Replication and expression from Staphylo-coccus aureus in Bacillus subtilis. Proc. Natl. Acad. Sci. USA74, 1680–1682.

Ehrlich, S.D., Bruand, C., Sozhamannan, S., Dabert, P., Gros, M.F.,Jannière, L., Gruss, A., 1991. Plasmid replication and structuralstability in Bacillus subtilis. Res. Microbiol. 142, 869–873.

248 H.D. Nguyen et al. / Plasmid 54 (2005) 241–248

Gryczan, T.J., Dubnau, D., 1978. Construction and propertiesof chimeric plasmids in Bacillus subtilis. Proc. Natl. Acad.Sci. USA 75, 1428–1432.

Guérout-Fleury, A.M., Frandsen, N., Stragier, P., 1996. Plas-mids for ectopic integration in Bacillus subtilis. Gene 180,57–61.

Härtl, B., Wehrl, W., Wiegert, T., Homuth, G., Schumann, W.,2001. Development of a new integration site within theBacillus subtilis chromosome and construction of compati-ble expression cassettes. J. Bacteriol. 183, 2696–2699.

Homuth, G., Heinemann, M., Zuber, U., Schumann, W., 1996.The genes lepA and hemN form a bicistronic operon inBacillus subtilis. Microbiology 142, 1641–1649.

Jannière, L., Bruand, C., Ehrlich, S.D., 1990. Structurally stableBacillus subtilis cloning vectors. Gene 87, 53–61.

Jürgen, B., Schweder, T., Hecker, M., 1998. The stability ofmRNA from the gsiB gene of Bacillus subtilis is dependenton the presence of a strong ribosome binding site. Mol. Gen.Genet. 258, 538–545.

Kaltwasser, M., Wiegert, T., Schumann, W., 2002. Constructionand application of epitope- and GFT-tagging integrationvectors for Bacillus subtilis. Appl. Environ. Microbiol. 68,2624–2628.

Kim, L., Mogk, A., Schumann, W., 1996. A xylose-inducibleBacillus subtilis integration vector and its application. Gene181, 71–76.

Kraus, A., Hueck, C., Gärtner, D., Hillen, W., 1994. Cataboliterepression of the Bacillus subtilis xyl operon involves a ciselement functional in the context of an unrelated sequence,and glucose exerts additional xylR-dependent repression. J.Bacteriol. 176, 1738–1745.

Maul, B., Völker, U., Riethdorf, S., Engelmann, S., Hecker, M.,1995. �B-dependent regulation of gsiB in response to multi-ple stimuli in Bacillus subtilis. Mol. Gen. Genet. 248, 114–120.

Middleton, R., Hofmeister, A., 2004. New shuttle vectors forectopic insertion of genes into Bacillus subtilis. Plasmid 51,238–245.

Mogk, A., Hayward, R., Schumann, W., 1996. Integrative vec-tors for constructing single-copy transcriptional fusionsbetween Bacillus subtilis promoters and various reportergenes encoding heat-stable enzymes. Gene 182, 33–36.

Mueller, J.P., Bukusoglu, G., Sonenshein, A.L., 1992. Transcrip-tional regulation of Bacillus subtilis glucose starvation-inducible genes: control of gsiA by the ComP–ComA signaltransduction system. J. Bacteriol. 174, 4361–4373.

Rygus, T., Scheler, A., Allmannsberger, R., Hillen, W., 1991.Molecular cloning, structure, promoters and regulatory ele-ments for transcription of the Bacillus megaterium encodedregulon for xylose utilization. Arch. Microbiol. 155, 535–542.

Saito, H., Shibata, T., Ando, T., 1979. Mapping of genes deter-mining nonpermissiveness and host-speciWc restriction tobacteriophages in Bacillus subtilis Marburg. Mol. Gen.Genet. 170, 117–122.

Schmiedel, D., Hillen, W., 1996. A Bacillus subtilis 168 mutantwith increased xylose uptake can utilize xylose as sole car-bon source. FEMS Microbiol. Lett. 135, 175–178.

Schulz, A., Schwab, S., Versteeg, S., Schumann, W., 1997. ThehtpG gene of Bacillus subtilis belongs to class III heat shockgenes and is under negative control. J. Bacteriol. 10, 3103–3109.

Schumann, W., Ferreira, L.C.S., 2004. Production of recombi-nant proteins in Escherichia coli. Gen. Mol. Biol. 27, 442–453.

Shimotsu, H., Henner, D.J., 1986. Construction of a single-copyintegration vector and its use in analysis of regulation of thetrp operon of Bacillus subtilis. Gene 43, 85–94.

Simon, D., Chopin, A., 1988. Construction of a vector plasmidfamily and its use for molecular cloning in Streptococcus lac-tis. Biochimie 70, 559–566.

Smith, I., Paress, P., Cabane, K., Dubnau, E., 1980. Genetics andphysiology of the rel system of Bacillus subtilis. Mol. Gen.Genet. 178, 271–279.

Takagi, H., Kagiyama, S., Kadowaki, K., Tsukagoshi, N.,Udaka, S., 1989a. Genetic transformation of Bacillus breviswith plasmid DNA by electroporation. Agric. Biol. Chem.53, 3099–3100.

Takagi, H., Miyauchi, A., Kadowaki, K., Udaka, S., 1989b.Potential use of Bacillus brevis HPD31 for the production offoreign proteins. Agric. Biol. Chem. 53, 2279–2280.

Titok, M.A., Chapuis, J., Selezneva, Y.V., Lagodich, A.V.,Prokulevich, V.A., Ehrlich, S.D., Jannière, L., 2003. Bacillussubtilis soil isolates: plasmid replicon analysis and con-struction of a new theta-replicating vector. Plasmid 49, 53–62.

Völker, U., Engelmann, S., Maul, B., Riethdorf, S., Völker, A.,Schmid, R., Mach, H., Hecker, M., 1994. Analysis of theinduction of general stress proteins of Bacillus subtilis.Microbiology 140, 741–752.

Wong, S.-L., 1995. Advances in the use of Bacillus subtilis forthe expression and secretion of heterologous proteins. Curr.Opin. Biotechnol. 6, 517–522.

Wu, S.C., Wong, S.-L., 2002. Engineering of a Bacillus subtilisstrain with adjustable levels of intracellular biotin for secre-tory production of functional streptavidin. Appl. Environ.Microbiol. 68, 1102–1108.

Wu, X.-C., Lee, W., Tran, L., Wong, S.-L., 1991. Engineering aBacillus subtilis expression-secretion system with a straindeWcient in six extracellular proteases. J. Bacteriol. 173,4952–4958.