Vol. 175, No. 22JOURNAL OF BACTERIOLOGY, Nov. 1993, P. 7130-71370021-9193/93/227130-08$02.00/0Copyright ©) 1993, American Society for Microbiology

Characterization of the Gene Encoding an IntracellularProteinase Inhibitor of Bacillus subtilis and Its Role in

Regulation of the Major Intracellular ProteinaseYASUHIRO SHIGA, HIDEO YAMAGATA, AND SHIGEZO UDAKA*

Department ofApplied Biological Sciences, Faculty ofAgriculture, Nagoya University, Chikusa, Nagoya 464, Japan

Received 3 June 1993/Accepted 9 September 1993

The gene (ipi) for an intracellular proteinase inhibitor (BsuPI) from Bacillus subtilis was cloned and foundto encode a polypeptide consisting of 119 amino acids with no cysteine residues. The deduced amino acidsequence contained the N-terminal amino acid sequence of the inhibitor, which was chemically determinedpreviously, and showed no significant homology to any other proteinase inhibitors. Analysis of the transcrip-tion initiation site and mRNA showed that the ipi gene formed an operon with an upstream open reading framewith an unknown function. The transcriptional control of ipi gene expression was demonstrated by Northern(RNA) blot analysis, and the time course of transcriptional enhancement roughly corresponded to the resultsobserved at the protein level. Strains in which the ipi gene was disrupted or in which BsuPI was overexpressedconstitutively sporulated normally. Analysis of the time course of production of the intracellular proteinaseand proteinase inhibitor in these strains suggested that BsuPI directly regulated the major intracellularproteinase (ISP-1) activity in vivo.

Bacillus subtilis cells produce extracellularly at least fourserine proteinases (Apr, Epr, Bpf, and Vpr) and three metal-loproteinases (NprA, NprB, and Mpr), all of which have beenshown to be nonessential for normal growth and sporulation(15, 44, 50). As for intracellular proteinases, four serineproteinases, ISP-1, ISP-2 (generally considered to be a trypsin-like proteinase), ISP-3, and a membrane-bound proteinase (29,41-43, 46), and a spore-germinating proteinase (Gpr) (48) areknown. Among them, ISP-1 is the major intracellular protein-ase, accounting for 80% of the intracellular azocollagen orazocasein hydrolytic activity, and is the most well characterized(5). ISP-1 was thought to function in the degradation andturnover of bulk proteins during sporulation (5). However, thegene encoding ISP-1 has already been cloned (17) and amutant deficient in it was shown to sporulate normally (1).Expression of the ISP-1 gene was shown to be regulated at thetranscriptional level indirectly by SpoOA and various regula-tory proteins (16, 34). Transcription was enhanced at To, theearliest stage of sporulation, but ISP-1 activity was not de-tected in cells until about 4 h later (34). Although theN-terminal 17-amino-acid sequence of the primary transla-tional product was not found in purified ISP-1, it was reportedthat this cleavage was an artifact and not related to theactivation of ISP-1 (39). These results suggested that ISP-1activity was masked posttranslationally and revealed whenneeded.On the other hand, ISP-1 was found to be inhibited by an

intracellular proteinaceous inhibitor of B. subtilis (26). Thisinhibitor (designated BsuPI, in accordance with our proposal[40]) was purified by two different groups (27, 31) and analyzedat the protein level (30). Although there are some differencesbetween the results obtained by the groups, BsuPI was re-ported to be a very thermostable protein with a molecular massof about 16 kDa and the decrease in its cellular amountseemed to cause an increase in ISP-1 activity (27, 42). Nishinoand Murao reported that BsuPI strongly inhibited ISP-1

* Corresponding author.

activity and was inactivated by a membrane-bound proteinaseunder acidic conditions when tested in vitro (30). These resultssuggested that BsuPI inhibited ISP-1 activity through complexformation at an early stage of sporulation and that later BsuPIwas inactivated by a membrane-bound proteinase, resulting inhigh ISP-1 activity. Therefore, for characterization of the exactfunction of BsuPI with special reference to sporulation, cloningof the BsuPI-encoding gene was deemed to be essential.

In addition, it was also reported that BsuPI contained nocysteine residues. Since the structural relationship between thedisulfide bridge and the reactive site in serine proteinaseinhibitors from eukaryotic organisms (20, 33) and Streptomycesspecies (28) is already well known, it would be interesting todetermine how BsuPI inhibits ISP-1. In this report, cloning ofthe BsuPI-encoding gene and characterization of its structureand function are described.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and transformation.Escherichia coli XL1-blue (Stratagene) and BL21(DE3) (No-vagen) were grown at 37°C in 2YT and M9 (37) or M9ZB (47),respectively, and then transformed by the method of Hanahan(13). B. subtilis DB104 (15) was grown at 37°C in L broth (37)or 2 x SG (21) and then transformed by the method of Changand Cohen (7) or Sadaie and Kada (35). When required,ampicillin, kanamycin, and erythromycin were added at con-centrations of 50, 5, and 10 p.g/ml, respectively. For molecularcloning, pBR322 (37) and its derivative pBR-AN3 (49) wereused, and for sequencing, M13mpl8 and M13mpl9 (37) wereused. For gene expression in E. coli, pET-1 Id (47) was used. Inthe gene disruption experiment, plasmid pUCI 18 (37) wasused as the vector. spoOA mutants UOT-1611 (spoOAXHBtrpC2 lys-I nprE18 nprR2 aprEA3), whose spoOA mutation isparticularly stable, and UOT-1277 (spoOA12 trpC2 lys-1 nprE18nprR2 aprEA3) were kindly supplied by Y. Kobayashi and F.Kawamura, respectively.

Preparation of DNA. Chromosomal DNA of B. subtilis wasprepared by the method of Saito and Miura (36). Large- and

7130

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PROTEINASE INHIBITOR GENE FROM BACILLUS SUBTILIS 7131

A

Hd Ec Ec Nc Ps Ha Nc Hd

I I I i I I I(Hf) Ns

ORF 2 BsuPIORF I

B

II

promoter

Restriction map and deletion mapjregion BsuPI-encoding gene. (A) Restri

BsuPI-encoding gene, ipi. The gray, white, an

ORF1, ORF2, the BsuPI-encoding gene, re

map indicates the transcrifAbbreviations for restriction enzyi

Ec, EcoRI; Hc, HincII; Hd, Hindlll, Hf, Hinfl; NPstI; Sp, SphI (only one of the several Hinfl

fragments used for promoter mappwith lacZ. The names of the

given middle. 3-Galactosidase activity vcolony L-broth agar containing X-4

indicates the position of the presum

small-scale plasmid preparations were perfiby Birnboim (2).

Cloning of the BsuPI-encoding gene andeoxyoligonucleotide 5'-ATGGA(AG)A/(AG)GT(ACGT)GT-3', which correspondamino acids of the N-terminal sequence of

Nishino et al. (31), was synthesized at th(Research of Nagoya University and used a

labelling with a MEGALABEL kit (Takar;B. subtilis chromosomal DNA was dig

restriction enzymes and then separated b)electrophoresis. Southern blot analysis (formed with 6.6 x 106 cpm of the pihybridization buffer, with incubation at 42blotted nylon membrane was washed three

(1x SSC is 0.15 M NaCl plus 0.015

containing 0.1% sodium dodecyl sulfate (room temperature. The 1.9-kb PstIfragme]positive signal, was recovered from the agGene Clean II (Bio 101, Inc.), inserted irplasmid pBR-AN3, and used to transform iampicillin resistance. For selection of positihybridization (37) was performed with thecloning of the upstream region of the BsuP0.5-kb PstI-HindIII fragment (Fig. 1) w;

Multiprime DNA labeling system (Amersham) and used as aprobe. DNA sequencing was performed with an M13 Sequenc-

I kb ing Kit, Version 2.0 (Takara Shuzo Co.). We named theBsuPI-encoding gene ipi (a gene coding for intracellular

Hc Sp Ps proteinase inhibitor).Expression of the BsuPI-encoding gene in E. coli and B.

subtilis. An expression vector for BsuPI was constructed asfollows. The 1.4-kb HindIII-PstI fragment (Fig. 1) of theBsuPI-encoding gene was shortened with exonuclease BAL 31from the Hindlll site to the PstI site and treated with the

B-Galactosidase Klenow fragment of DNA polymerase I in the presence of allPlasmid activity four deoxynucleoside triphosphates. An XbaI linker was li-

gated to the fragment and then digested with both XbaI andpPL-HdB - Pstl. The digested fragments were inserted into pUC119 andpPL-ENc - sequenced. Among them, a 1.1-kb fragment truncated up topPL-PNs + nucleotide position 1,109 (see Fig. 2), which contained the

putative ribosomal binding site and the translational initiationpPL-PHf + codon for BsuPI, was obtained. This DNA fragment, DH5, waspPL-BHf + used for expression of BsuPI. For expression in E. coli, the

DH5 fragment was inserted between the XbaI and HindlIlpPL-HdHf - sites of pET-lld, and the resultant plasmid, pET-DH5, waspPL-dBNc - introduced into E. coli BL21(DE3).pPL-dNcHd - For expression in B. subtilis, the DH5 fragment was ligated

with the 0.4-kb Sau3AI-AluI fragment of pUB110, whichcontained the promoter of the kanamycin resistance gene, andthen introduced into the PvuII site of pUB110. B. subtilis

ping of the promoter DB104 was transformed with this plasmid (pUB-DH5).ction map around the Purification of BsuPI expressed in E. coli. E. coliId black bars indicate BL21(DE3) harboring pET-DH5 was cultured in 1 liter of-spectively. The arrow M9ZB medium for 3.5 h, and then transcription from the T7ptional start point and promoter was induced by addition of isopropyl-3-D-thiogalac-'m NcoIts: NspVBams topyranoside (IPTG) (47). Incubation was continued for arc, Ncoi; Ns, NspV; Ps, further 4.5 h. Cells were harvested by centrifugation (6,000 xing. These fragments g, 10 min, 40C) and dissolved in 25 ml of distilled water. Theresulting plasmids are cell suspension was then passed through a French pressure cell,

vas detected as blue- and cell debris was removed by centrifugation (25,000 X g, 10Gal. The closed bar at min, 4°C). The supernatant was heated at 70°C for 10 min, andied promoter region. then the denatured proteins were removed by centrifugation.

Then 1 N HCl was added to adjust the pH of the supernatantto 4.0. After incubation at 37°C for 30 min, the denaturedproteins were removed as described above. The pH of the

ormed as described solution was adjusted to 7.0 with 1 N NaOH, and then solidammonium sulfate was added to 80% saturation. The precip-

id sequencing. The itate obtained on centrifugation was dissolved in a minimumk(CT)CA(AG)GA volume of buffer A (20 mM Tris-HCl buffer, pH 7.5) and thenIs to the first seven dialyzed against the same buffer overnight at 4°C. After.BsuPI reported by removal of insoluble materials by centrifugation, the dialysatee Center for Gene was applied to a DEAE-cellulose column (2.6 by 3.5 cm)5S a probe after 32p preequilibrated with buffer A. Elution was performed with ana Shuzo Co.). NaCl gradient (0 to 0.5 M) in buffer A, and BsuPI was eluted,ested with various in the 0.22 M fraction. The BsuPI-containing fractions were0.7% agarose gel then combined and dialyzed against buffer B (20 mM triethyl-

37) was then per- amine, pH 7.3). After removal of the insoluble materials, therobe in 10 ml of dialysate was applied to a Mono-Q HR5/5 column (Pharmacia)2°C overnight. The preequilibrated with buffer B. Elution was performed with antimes with 6 x SSC NaCl gradient (0 to 0.35 M) in buffer B. BsuPI was eluted inMI sodium citrate) the 0.2 M fraction and detected as a single band on SDS-15%SDS) for 5 min at polyacrylamide gel electrophoresis (PAGE) (19), which wasnt, which showed a followed by staining with Coomassie brilliant blue.)arose gel by using Determination of the N-terminal amino acid sequence ofito the PstI site of recombinant BsuPI. The N-terminal amino acid sequence ofE. coliXL1-blue to purified recombinant BsuPI was determined with a gas phaseive colonies, colony protein sequencer (ABI 470-120A). The sample for sequencesame probe. For analysis was prepared by the method of Matsudaira (22).'I-encoding gene, a Analysis of BsuPI mRNA. Total RNA of B. subtilis cells was

as labeled with a prepared by the hot-phenol method (45) with the modifica-

- m

VOL. 175, 1993

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7132 SHIGA ET AL.

tions described below. B. subtilis was cultured in 500 ml of Lbroth in a 2-liter flask at 37°C, and at various phases of growth,50 ml of the culture was removed, mixed with 0.5 ml of ice-cold1 M NaN3-0.1 M MgCl2, and then rapidly chilled on ice-water.Cells were harvested by centrifugation (6,000 x g, 10 min, 4°C)and then suspended in 5 ml of ice-cold TES (30 mM Tris-HCIbuffer [pH 8.0], 5 mM EDTA [pH 8.0], 50 mM NaCl)containing 25% sucrose and 3 mg of lysozyme per ml. Incuba-tion on ice-water was continued for 15 min. After centrifuga-tion (3,500 x g, 20 min, 4°C), the spheroplast pellet was

resuspended in 3 ml of 0.02 M sodium acetate (pH 5.5)-0.5%SDS-1 mM EDTA, and then hot phenol extraction was

performed three times. After ethanol precipitation, the pelletwas dissolved in 100 [l of diethylpyrocarbonate-treated dis-tilled water and then kept at - 80°C until use.

Northern (RNA) blot analysis was performed as describedpreviously (37). A 30-[Lg sample of total RNA prepared from

cells at various phases of growth was subjected to electrophore-sis on 1% agarose containing formaldehyde and then blottedonto a Hybond-N+ membrane (Amersham). Hybridizationwas performed at 65°C overnight in 10 ml of 0.5% SDS-5 x

SSC-5 x Denhardt's solution containing 50 pLg of sonicatedsalmon sperm DNA per ml with the multiprime-labeled 1.3-kb

BamHI-SphI fragment as a probe (2.0 x 106 cpm). The

membrane was washed twice with 0.2x SSC-0.1% SDS at

65°C for 20 min each time.Primer extension analysis was performed as described by

Debarbouille and Raibaud (10). A deoxyoligonucleotide with

the sequence 5'-TCATGTGGCGAGTGTGAATG-3', which

is complementary to the sequence from nucleotides 821 to 840

(see Fig. 2), was synthesized at the Center for Gene Research

of Nagoya University and then used as a primer. A 100-[Lgsample of total RNA extracted from T1.5 (1.5 h after To) cells

was hybridized with 50 ng of primer at 30°C overnight and used

for the assay. The length of the primer-extended product was

determined by comparison with sequencing ladders generatedby using the same primer and the single-stranded DNA

template containing the BsuPI-encoding gene ranging from

nucleotides 339 to 2260 (1.9-kb PstI fragment).Construction of an in-frame lacZ fusion vector as a pro-

moter probe. The promoterless lacZ gene was obtained from

pMC1403 (6), digested with both BamHI and Dral, and then

inserted into BamHI-SmaI-digested pUC118 (pUC-lacZ).From the BsuPI-encoding gene and its flanking region, a 0.9-kb

PstI-Hinfl fragment for an in-frame ipi-lacZ fusion and a 1.0-kb

HindIII-BamHI fragment for orfl-lacZ were prepared, treated

with T4 polymerase in the presence of all four deoxynucleosidetriphosphates, and then inserted into the 5'-upstream regionof the lacZ gene in pUC-lacZ. In-frame fusions were selectedon the basis of the formation of blue colonies on 5-bromo-

4-chloro-3-indolyl-,3-D-galactopyranoside (X-Gal)-containingplates and confirmed by nucleotide sequencing. The fusion

genes thus constructed were isolated and inserted into appro-priate sites of pUB110, to which the multicloning site region ofpNU210 (49) was incorporated, in the direction avoidingreadthrough from any promoters of the plasmid (24) and thenintroduced into B. subtilis. r-Galactosidase activity, due to thepromoter activity of the inserted fragment, in B. subtilis was

judged on the basis of the formation of blue colonies on X-Galplates, and the activity in liquid culture was measured by themethod of Miller (25).

Disruption of the BsuPI-encoding gene by homologousrecombination. pUB110 was partially digested with HaeIII,and the 1.0-kb DNA fragment containing the whole kanamycinresistance gene was recovered from the agarose gel afterelectrophoresis. The 1.0-kb HindIII-SphI fragment of the

BsuPI-encoding gene (Fig. 1) was inserted between the same

restriction sites of pUCI 18, and then the kanamycin resistancegene was introduced into the NspV site of the BsuPI-encodinggene of the plasmid. A 1.2-kb ClaI-SacI fragment containingthe erythromycin resistance gene obtained from pHW1 (14)was introduced into the SacI-AccI site of this plasmid(pYEK13). B. subtilis was then transformed with pYEK13 bythe method of Sadaie and Kada (35), and kanamycin-resistantand erythromycin-sensitive colonies were selected. Chromo-somal DNAs of these clones, which were expected to berecombinants on both side of the disrupted BsuPI-encodinggene on pYEK13, were extracted, and insertion of the kana-mycin resistance-encoding gene into the chromosomal BsuPI-encoding gene was confirmed by Southern blot analysis.Among the clones with disrupted BsuPI gene, number 42(designated 42pi) was chosen for further analysis.

Sporulation of BsuPI-defective and -overproducing strains.B. subtilis DB104, 42pi, and DB104(pUB-DH5) were inocu-lated into 5 ml of 2 x SG medium at an initial A660 of 0.08 andthen incubated for 24 h at 37°C with shaking. Heat-resistantspores were counted by plating cells on tryptose blood agarbase supplemented with 0.5% glucose after heating the culturefor 10 min at 80°C. Strains 42pi and DB104(pUB-DH5), on theother hand, were replicated onto L agar containing kanamycinto confirm maintenance of the kanamycin resistance gene.

Determination of intracellular proteinase activity and pro-teinase inhibitor activity. B. subtilis DB104, 42pi and DBI04(pUB-DH5) were inoculated into 500 ml of 2 x SG mediumand then cultured at 37°C. At various phases of growth, 5 ml ofthe culture was removed and cells were harvested by centrifu-gation (3,500 x g, 10 min, 4°C). Cell extracts were obtained as

described by Ruppen et al. (34).Intracellular proteinase activity was determined as described

by Chavira et al. (8). One proteinase unit was defined as theamount of proteinase that hydrolyzed 1 pg of azocollagen(Azocoll; Sigma) per min.

Intracellular proteinase inhibitor activity was determined as

described previously (40). Samples were prepared after heattreatment at 95°C for 10 min and removal of denaturedproteins by centrifugation. Proteinase activity was then mea-

sured as described above. One inhibitor unit was defined as theamount of inhibitor that caused a 50% reduction of azocolla-genase activity (2 U) of subtilisin Carlsberg (Sigma) under theassay conditions used. Subtilisin was used as the proteinase forthe assay because it was quantitatively inhibited by BsuPI (31).

Nucleotide sequence accession number. The nucleotide se-

quence data reported here will appear in the DDBJ, EMBL,and GenBank nucleotide sequence data bases under accessionnumber D1631 1.

RESULTS

Cloning and sequencing of the BsuPI-encoding gene, ipi. Anoligonucleotide probe corresponding to the N-terminal aminoacid sequence of BsuPI was used to determine that the genefor BsuPI is located on a 1.9-kb PstI fragment as described inMaterials and Methods. The 1.9-kb PstI fragments were iso-lated, inserted into the PstI site of pBR-AN3, and used totransform E. coli. Colony hybridization with the above-de-scribed probe identified four positive colonies containing a

1.9-kb PstI fragment among approximately 1,000 ampicillin-resistant colonies. Sequence analysis revealed that this frag-ment contained three open reading frames (ORFs), and thelongest one was cloned incompletely. Therefore, the 1.3-kbHindlIl fragment containing the 5'-upstream region of thisORF was also cloned and sequenced. These ORFs were 840,

J. BACTERIOL.

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PROTEINASE INHIBITOR GENE FROM BACILLUS SUBTILIS 7133

228, and 357 nucleotides long, respectively. The second ORF(ORF2) was contained within the first ORF (ORF1), in adifferent reading frame. The deduced amino acid sequence ofthe product of the third ORF matched that of BsuPI, as

reported by Nishino et al. (31), indicating that ORF3 encodesBsuPI (Fig. 1 and 2).The deduced amino acid sequence of BsuPI contained no

cysteine residues, as reported previously (31). Although it wasreported that BsuPI contained no histidine residues (31), thededuced amino acid sequence of BsuPI contained one histi-dine residue. Putative ribosomal binding sites complementaryto the 3' sequence of the 16S rRNA of B. subtilis (3) werefound just upstream of the putative translation initiationcodons (ATG for ORFI and the BsuPI-encoding gene andTTG for ORF2) of these ORFs, although that of ORF2 wasless matched. An 11-bp-long palindromic sequence (AG =- 108 kJ/mol), followed by a T stretch, was found downstreamof the BsuPI-encoding gene and may function as a p-indepen-dent transcriptional terminator. However, no palindromic se-

quence was found in the region between ORFI (or ORF2) andthe ipi gene. Southern blot analysis revealed only one BsuPI-encoding gene in the B. subtilis genome (data not shown).

Production of recombinant BsuPI in E. coli. The BsuPI-encoding gene was cloned downstream of an IPTG-inducibleT7 promoter in E. coli expression vector pET-Ild to createplasmid pET-DH5 as described in Materials and Methods.After IPTG induction, the cell lysate of E. coli BL21(DE3)harboring pET-DH5 contained significant inhibitory activitytoward subtilisin Carlsberg, while the lysate of uninduced cellsexhibited no activity. The recombinant protein was soluble andhad a molecular mass of 16 kDa, corresponding to the mass ofBsuPI estimated by SDS-PAGE. This protein, purified as a

single band on SDS-PAGE as described in Materials andMethods, retained inhibitory activity against subtilisin. TheN-terminal amino acid sequence of the purified inhibitor,MENQEVVLXI (X = not identified), was determined byEdman degradation and matched that of native BsuPI exceptfor the first methionine, which was absent in the native form(31).Northern blot analysis of BsuPI mRNA. On Northern blot

analysis of RNA prepared from B. subtilis at various phases ofgrowth (Fig. 3), an RNA of about 0.9 kb was specificallydetected with the BsuPI gene probe (1.3-kb BamHI-SphIfragment in Fig. 1) but not with the 5'-upstream probe (1.0-kbHindIII-BamHI fragment) or the 3'-downstream probe (0.4-kbSphI-PstI fragment) (data not shown). As can be seen in Fig. 3,an about 1.6-kb RNA was also detected, but this signal wasdetected with all of the probes described above (data notshown), suggesting that the band resulted from nonspecifichybridization with the probes. These results also suggest thatthe whole transcription unit of the BsuPI-encoding gene iscontained in the 1.3-kb BamHI-SphI fragment (see below) andthe 0.9-kb RNA is the mRNA for BsuPI synthesis. Expressionof the BsuPI-encoding gene was enhanced at about To andcontinued to be so to at least T1 5 (Fig. 3). The amount of 16SrRNA was also monitored as a control and found to beunchanged throughout T_ I to T 55 (data not shown). Theresults obtained with the BsuPI mRNA are in general agree-ment with the results obtained at the protein level (42) up to T3and suggest that BsuPI expression is controlled mainly at thetranscriptional level.

Determination of the promoter region and transcriptioninitiation site. To determine the transcription initiation site forthe BsuPI-encoding gene, a series of ipi-lacZ or orfl-lacZin-frame fusion plasmids were constructed (Fig. 1) and intro-duced into B. subtilis. Among them, B. subtilis harboring

pPL-PNs, pPL-PHf, and pPL-BHf formed blue colonies onX-Gal plates but B. subtilis harboring pPL-HdHf did not, evenafter being cultured for up to 7 days. Furthermore, promoteractivity was completely lost on deletion of the 130-bp BamHI-NcoI fragment or the 170-bp NcoI-HindIII fragment. Theseresults indicate that the 0.3-kb BamHI-HindIII fragment con-tains the promoter region and the sequence around the NcoIsite of this fragment is important for transcription of theBsuPI-encoding gene. pPL-HdB did not form a blue colony,suggesting that ORFI is not transcribed in B. subtilis.Upstream from the BsuPI-encoding gene, several sequences

similar to the SpoOA-binding sequence (4) were found (Fig. 2).To examine the role of SpoOA in the expression of theBsuPI-encoding gene, pPL-PHf was introduced into spoOAmutant strains (UOT-1611 and UOT-1277), but these strainsshowed as high a level of P3-galactosidase activity as thewild-type strain, indicating that the promoter was functional inthe absence of SpoOA. This showed that SpoOA, which isknown to control expression of the ISP-i-encoding geneindirectly (16, 34), is not necessary for expression of theBsuPI-encoding gene and, moreover, u factors involved down-stream of the SpoOA cascade may not be relevant to BsuPIexpression (4, 12, 16).

Primer extension analysis revealed that transcription of theBsuPI-encoding gene starts at nucleotides 623 and 624, whichare about 40 bp upstream of the NcoI site described above(Fig. 4). In the region 5' upstream from this site, a - 10sequence for CD (GACGGTAT/consensus GCCGATAT) wasfound but a -35 sequence was not. No other promoter-likesequence for u factors of B. subtilis was found (18). Todetermine whether ORFI is transcribed, a deoxyoligonucle-otide with the sequence 5'-CGCCGATGATGACGATGAGTA-3', complementary to the sequence from nucleotides 274to 294 (Fig. 2), was used as a primer for reverse transcriptase.However, no primer-extended product was observed. Thisresult suggests that ORFi is not transcribed under the culturalcondition employed.

Effects of constitutive overexpression and disruption of theBsuPI-encoding gene on sporulation. Strains containing adisruption of the BsuPI-encoding gene were constructed byinsertion of a kanamycin resistance gene. Southern blot anal-ysis was performed on chromosomal DNAs extracted from 12kanamycin-resistant and erythromycin-sensitive clones, ob-tained as described in Materials and Methods, by using a0.4-kb HindIII-NspV fragment labelled with 32P as a probe.Among them, 10 had the 1.0-kb insert of the kanamycinresistance gene within the BsuPI-encoding gene, as expected(data not shown).To determine the relationship of growth phase-specific

expression of the BsuPI-encoding gene to sporulation of B.subtilis, heat-resistant spores of the wild-type strain, DB104;one of the insertion mutants, 42pi; and a strain overexpressingthe BsuPI-encoding gene, DB104(pUB-DH5), were counted asdescribed in Materials and Methods. While the numbers ofviable cells of these three strains per milliliter were 9.9 x 108,2.1 X 108, and 1.1 x 10', respectively, the numbers of sporesper milliliter were 5.9 x 108, 8.0 x 107, and 3.9 x 108,respectively. As for B. subtilis 42pi and DB104(pUB-DH5),more than 98% of the colonies derived from spores retainedthe kanamycin resistance gene. Thus, the three strains cansporulate efficiently, indicating that BsuPI is not necessary forsporulation of B. subtilis.

Interaction of BsuPI with intracellular proteinase in vivo.The time course of the intracellular proteinase and proteinaseinhibitor activities was analyzed by using cell extracts of B.subtilis DBI04, 42pi, and DBI04(pUB-DH5). As shown in Fig.

VOL. 175, 1993

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7134 SHIGA ET AL. J. BACTERIOL.

F,coRlGAATCTTTATCGCCGAGAGAAACCCTGAGCGGAATAAACGGAATGCCTTTTTCTTCAAAATAAGAACGAGGCAAATCAGTGGCGCTGTCAGCGATGAGATGAACTGTCATATTGCTTCC

CTTCTTTCCGTTATCGCTCGTTATCATGAGAGTTAAGCCAACTATAT'CATATTAATTTGCCAAAATCGTGTTCAAATATTTGTTTTAAGGGAAAACATAATAAAGAAGAAAGAGCGGAGG

NcoI Pat!

M V L D Q T KKKL L I V I I G A L L N A A G L N L F L I P A D V Y A S G F TORFI

CAGGTGTTGCCCAGC TTTTATCAAGCGTCGTTGATCAATATGCCCCCTTTTACATATCGACGGGAACGCTCTTGTTCCTCTTAAACATTCCGGTCGGTATTTTAGGATGGCTGAAGGTCGG V A Q L L S S V V D Q Y A P F Y I S T G T L L F L L N I P V G I L G W L K V G

BamHIGCAAATCGTTTACAGTGTACAGCATTTTAAGSICSVACTCACTACTTCTGTTTATGGGGATCCTGCCGGAAACAAGCCTGTCAEATGACATTTTGCTGAACGCGGTGTTCGGCGGCGTCA

K S F T V Y S I L S V A L T T L F M G I L P E T S L S H D I L L N A V F G G V I

-10 44 Ncol

60 1 TTTiCClCTALCGGATCGL TAACTLAAL11ATAA GACT1TL.Al 1CTTJiA.1iKLiLiluI W swiLii.LIA 1LUPALiTt'i'LihiirnIaRAI.iTi3i..JI.IT'..JXJ'JL1S A D G I G L T L K Y G A S T G G L D I V A A V L A K W K D K P V G T Y F F I L

8 D721t TGAACGGGATTATCATC:TTGAC(:GCAtIGATTATTGCAAGGlTGGGAGAAAGCATTATATACCCTGGTTACACTATATGTGACAACGAGGGTGATcrGACGCCATTCACACTCGCCACATGA-

N G I I I L T A G L L Q G W E K A L Y T L V T L Y V T T R V I D A I H T R H M KM G E S I I Y P G Y T I C D N E G D R R H S H S P H E0RF2

Hindlil841 AGCTTACGGCAATGATTGTGACGAAAAAAGCGGACGAA~ATCAAG;GAAGCCATTTA4CGGAAAAATGGTG;CGCrGGCATCACAACTCTTCCGGCAAAAGGArcATTTACGAACGAACAGAAAG

L T A M I V T K K A D E I K E A I Y G K M V R G I T T L P A K G A F T N E Q K EA Y G N D C D E K S G R N Q G S H L R K N G A R H H N S S G K R S I Y E R T E R

AAATGATGATTATCGTCATCACGAGGTATGAACTGTACGATTTAGAAAAGATCGTCAAAGAAGTTGATCCAAAAGCATTTACGAACATCGTTICAAACGACAGGGATTTTTGGTTTCTTTAM M I I V I T R Y E L Y D L E K I V K E V D P K A F T N I V Q T T G I F G F F TN D D Y R H H E V *

BSD.......

D

CAAAAGACTGACGAAAGCAGGT,GGAAGTICGT'GAAACGGCTGCTTGTGATGCTCCTCCCCGTGCTGCTTTTGATAGGCTGCGGGAAAGATGAGCAGACAGAACCCGATAAGGAGGTATCAGK D *

(Ilinfi)

M E N Q E V V L S I D A I Q E P E Q I K F N H S L K N Q S E R A I E F Q. F S T8supI

NspV1321CA3GG2CTTTA TCGTGTATATCWAGCAC GATACCGTTATT G TGTTTACGATTTTCCTGACGTTGTCTGACAT

G Q K F E L V V Y D S E H K E R Y R Y S K E K M F T Q A F Q N L T L E S G E T Y

D F S D V W K E V P E P G T Y E V K V T F K G R A E N L K Q V Q A V Q Q F E V K

HGnCIIAATAAGAGGCTATGGCGAGTCGCCATAGCCTTTTTTTATTGCAGTTTGTCCAGCTGATTTTGAAATTGATGCAACTGTTGACGTTCCGCATCTGTCGAATTGGCATACGCAGAAGATACA*,̂~ ~ ~ ~ ~~~~~~~~~~~~~~~id

GCATTTTTGGCTCTCAGAATCGCTTGCTCCTGATCGCCTTCCTCAAAGCCGTTTGTCACCATGACCGCCTGTTGAACAAATGATTTTGCCTGCTGAAACAGTTCATTTCTCATTTAATAT

SphlTCATACTGCCTTTCAGCCGAGTCTCGTGCCGCGGCATGCTGGTGAGCTTCTTCTAAAGTAGACCGATATGGAAATCTTGCGGCGTGTAGCCGATGGAGTCTTTACCCTGCTGAATAAACC

PaliCCGCAGCCTGCAAATTCAAGCATTTCCAAATCGTTATCCTCATCTCCAAAAGCAATGATCCGCTCCCTCGGCACGCCGTAATAATCGCTGATTTTCTGCAG

FIG. 2. Nucleotide sequence of the gene encoding BsuPI and the deduced amino acid sequences of three ORFs. Only the antisense strand is

shown. The vertical arrows above the DNA sequence indicate the positions of the 5' termini of the BsuPI-encoding gene transcripts determined

by primer extension analysis (see Fig. 4). The possible UD-type -10 region is doubly underlined. Putative ribosome-binding sites (S D) are

underlined. Putative initiation codons and termination codons are shown in bold letters. An 11-nucleotide palindromic sequence is indicated byhorizontal arrows below the DNA sequence. The region considerably homologous to the SpoOA-binding site is indicated by dots above the

nucleotide sequence.

121

241

361

481

961

1081

1201

1441

1561

1681

1801

1921

2041

2161

TTCTGGATTTGCTTCGTTTACCCATTTTGAGACTGCCCCCTTCCATACTTTGACGAGATGAGTCGTCTTCAATAGTATGGTGGGGTAAAGTAAGGTTTATGAGTGGGAAAAATTAATACG

CACCTTCTTTAAAGTGAGAAATATTCTTT,CAAAAACCTTGCCACGCCGTCCTCTTCATTTGT'GGCCGTAGTTCGGTTGGCAATICTGCTT,GACCGCGTCAATTCCATTTCCCATTGCAACG

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PROTEINASE INHIBITOR GENE FROM BACILLUS SUBTILIS 7135

T T0T1T2.5T4T5.5 T G C A 1 2

23S-~

_L^ "_ -0- BsuPI mnRNA

FIG. 3. Northern blot analysis of BsuPI-encoding gene transcripts.Total RNAs, 30 ju.g of each, of B. subtilis prepared from the cells atvarious phases of growth (T_ l, To, Ti, T2.5, T4, and Ts55) were subjectedto electrophoresis on a 1% agarose gel containing formaldehyde. Theinternal markers, 16S and 23S rRNAs, were visualized by ethidiumbromide staining, and their positions are indicated.

5, no proteinase inhibitor activity was detected in the BsuPI-defective strain (42pi). In this strain, proteinase activity wasdetected earlier than in the parental strain and the timing of itsappearance was in agreement with that of ISP-1-encoding geneexpression (34). In contrast, proteinase activity was not de-tected in the BsuPI-constitutive strain. These data demonstratethat BsuPI controls the major intracellular proteinase, ISP-1,in vivo and that no other proteinase inhibitor for ISP-1 exists inB. subtilis.

DISCUSSION

By using the cloned BsuPI-encoding gene, we were able topursue the role of BsuPI during growth and sporulation of B.subtilis. The transcriptional control of BsuPI gene expressionwas demonstrated by Northern blot analysis, and the timecourse of transcriptional activation roughly corresponded tothe results obtained at the protein level (Fig. 3 and 5).Furthermore, recombinant BsuPI with N-terminal methionineexhibited inhibitory activity toward subtilisin Carlsberg. Theseresults demonstrated that BsuPI was active without posttrans-lational modification, such as removal of the N-terminal me-thionine and proteolytic processing, as was observed in thecase of the proteinase inhibitor of B. brevis (40). Therefore,growth phase-specific transcriptional activation of the BsuPI-encoding gene may have some significance at the early stage ofsporulation, regardless of its necessity for sporulation. North-ern blot analysis also revealed that the BsuPI mRNA has along 5'-upstream control region and that it might be tran-scribed as an operon with ORF2 (Fig. 2).The transcriptional initiation site of the BsuPI-encoding

gene was detected by primer extension analysis. Although thefusion assay was performed under multicopy conditions, tran-scriptional activation with the same timing was also observed

C

GGC

TTA*A*CA 7sTT

A I w4FIG. 4. Primer extension analysis of BsuPI-encoding gene tran-

scripts. Lanes: 1 and 2, products of the primer extension reaction withRAV-2 and avian myeloblastosis virus reverse transcriptase, respec-tively; T, G, C, and A, sequence ladders for comparison. The positionsof the 3' ends of the primer extension products are indicated byasterisks.

on time course analysis of 3-galactosidase production in B.subtilis harboring pPL-PHf (data not shown). These findings,together with the results obtained with the fusion gene shownin Fig. 1, suggest several possibilities concerning the control ofBsuPI expression. For example, the product of ORF2 may beessential for BsuPI-encoding gene expression, which is similarto the finding that genes for a proteinase and its activatorprotein were in the same operon as reported in B. stearother-mophilus (32). The positively charged amino acid residuesabundant in the ORF2 product might be significant for bindingto DNA.Although only the - 10 consensus sequence recognized by

(J is found in the upstream sequence of the transcriptionalinitiation site and CjD is presumed to function in vegetative cells(18), a mechanism involving cooperation of aD and somefactor, like that of (A and SpoOA in sporulating cells, might benecessary for expression of the BsuPI-encoding gene (38).Further study is needed to identify which u factor recognizesthe promoter of the BsuPI-encoding gene.Time course analysis of production of the intracellular

proteinase and proteinase inhibitor in the BsuPI-deficient and-constitutive strains showed that BsuPI strongly inhibited themajor proteinase (ISP-1) in vivo at least until T7, so that a lagbetween transcription of the ISP-I-encoding gene and theappearance of ISP-1 activity was seen (Fig. 5). However, ISP-1activity increased at T4, even during BsuPI-encoding genetranscription (Fig. 3), and an abrupt increase of ISP-1 activityat T4 was observed in the wild-type strain, as well as in theBsuPI-deficient strain (Fig. 5). These results cannot be ex-plained simply by BsuPI inactivation by the membrane-boundproteinase, as was reported previously (30, 42, 43), since theISP-1 increase was observed even in the BsuPI-deficient strain.The reason for these changes in ISP-1 activity is not clear, butthey might have some relevance to the finding that ISP-1 isinhibited less by BsuPI from cells at later stationary phases(39).The BsuPI-deficient strain and the BsuPI-overexpressing

strain sporulated efficiently. This indicates that BsuPI is not

VOL. 175, 1993

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7136 SHIGA ET AL.

A 10

::L.

u

c:

-

ci

8-

6-

4-

2-

0-

B

0

..

0.._0E

*r

c

._Z0

C;

T_5 T_4 T_3 T_2 T_ To T1 '2 13 14

stage of sporulation

'1'X r Xf '1 'I' 'r T15 14 13 '-2 10 1 12 13 4

stage of sporulatioii

FIG. 5. Intracellular proteinase and BsuPI production by BsuPImutants during sporulation. Samples were prepared at various stagesof growth, and proteinase and proteinase inhibitor activities were

measured as described in Materials and Methods. (A) Intracellularproteinase production. (B) BsuPI production. The vertical scale iscontracted in the inset for monitoring of BsuPI production by B.subtilis DB104(pUB-DH5). Symbols: 0, B. subtilis DB1O4(pUBl 1O);A, B. subtilis 42pi; O, B. subtilis DB104(pUB-DH5).

important for normal sporulation. Since an ISP-1-deficientstrain also sporulates normally (1), the function of BsuPI was

assumed to involve inhibition of ISP-1 activity, which might beharmful only at an early stage of sporulation. It may be notedthat with a longer culture period (3 days), the BsuPI-deficientstrain tended to lyse more easily than the normal strain inmedium containing bonito extract instead of beef extract.

BsuPI was reported to contain no cysteine residues, as

judged from amino acid composition analysis. Although therelationship between the disulfide bridge and the reactive siteof proteinase inhibitors has been well characterized for eukary-otic inhibitors (20, 33), no such relationship is known forbacterial inhibitors other than those from Streptomyces spp.

(28). Recently, proteinase inhibitor genes were cloned fromgram-negative bacteria, E. coli (9, 23), and Pseudomonasaeruginosa (11). Their deduced amino acid sequences con-tained cysteine residues. We cloned an extracellular inhibitor(BbrPI)-encoding gene from B. brevis which also contained nocysteine residues (40). Primary structure comparison seemedto be the first step for identification of the reactive site andcharacterization of the inhibitory mechanism for these cys-teine-less inhibitors from bacilli, but BsuPI shows no homologyto other inhibitors or even to BbrPI. This suggests that theseinhibitors evolved through different pathways.

ACKNOWLEDGMENTS

We are grateful to Yasuo Kobayashi for helpful comments anddiscussions. We also thank Roy H. Doi for critical reading of themanuscript.

REFERENCES1. Band, L., D. J. Henner, and M. Ruppen. 1987. Construction and

properties of an intracellular serine protease mutant of Bacillussubtilis. J. Bacteriol. 169:444-446.

2. Birnboim, H. C. 1983. A rapid alkaline extraction method for theisolation of plasmid DNA. Methods Enzymol. 100:243-255.

3. Bott, K. F., and M. A. Hollis. 1982. Nucleotide sequence ofintergenic spacer DNA of three 16S-23S RNA gene sets from B.subtilis, p. 3-10. In A. T. Ganesan, S. Chang, and J. A. Hoch (ed.),Molecular cloning and gene regulation in bacilli. Academic Press,Inc., New York.

4. Burbulys, D., K. A. Trach, and J. A. Hoch. 1991. Initiation ofsporulation in B. subtilis is controlled by a multicomponentphosphorelay. Cell 64:545-552.

5. Burnett, T. J., G. W. Shankweiler, and J. H. Hageman. 1986.Activation of intracellular proteinase in Bacillus subtilis cellsduring sporulation. J. Bacteriol. 165:139-145.

6. Casadaban, M. J., J. Chou, and S. N. Cohen. 1980. In vitro genefusions that join an enzymatically active 3-galactosidase segmentto amino-terminal fragments of exogenous proteins: Escherichiacoli plasmid vectors for the detection and cloning of translationalinitiation signals. J. Bacteriol. 143:971-980.

7. Chang, S., and S. N. Cohen. 1979. High frequency transformationof Bacillus subtilis protoplasts by plasmid DNA. Mol. Gen. Genet.168:111-115.

8. Chavira, R., Jr., T. J. Burnett, and J. H. Hageman. 1984. Assayingproteinases with azocoll. Anal. Biochem. 136:446-450.

9. Chung, C. H., H. E. Ives, S. Almeda, and A. L. Goldberg. 1983.Purification from Escherichia coli of a periplasmic protein that is apotent inhibitor of pancreatic proteases. J. Biol. Chem. 258:11032-11038.

10. Debarbouille, M., and 0. Raibaud. 1983. Expression of theEscherichia coli malPQ operon remains unaffected after drasticalteration of its promoter. J. Bacteriol. 153:1221-1227.

11. Duong, F., A. Lazdunski, B. Cami, and M. Murgier. 1992. Se-quence of a cluster of genes controlling synthesis and secretion ofalkaline protease in Pseudomonas aeruginosa: relationships toother secretory pathways. Gene 121:47-54.

12. Grossman, A. D. 1991. Integration of developmental signals andthe initiation of sporulation in B. subtilis. Cell 65:5-8.

13. Hanahan, D. 1983. Studies on transformation of Escherichia coliwith plasmids. J. Mol. Biol. 116:557-580.

14. Horinouchi, S., and B. Weisblum. 1982. Nucleotide sequence andfunctional map of pE194, a plasmid that specifies inducibleresistance to macrolide, lincosamide, and streptogramin type Bantibiotics. J. Bacteriol. 150:804-814.

15. Kawamura, F., and R. H. Doi. 1984. Construction of a Bacillussubtilis double mutant deficient in extracellular alkaline andneutral proteases. J. Bacteriol. 160:442-444.

16. Klier, A., T. Msadek, and G. Rapoport. 1992. Positive regulation inthe gram-positive bacterium: Bacillus subtilis. Annu. Rev. Micro-biol. 46:429-459.

17. Koide, Y., A. Nakamura, T. Uozumi, and T. Beppu. 1986. Cloningand sequencing of the major intracellular serine protease gene of

J. BACTERIOL.

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

PROTEINASE INHIBITOR GENE FROM BACILLUS SUBTILIS 7137

Bacillus subtilis. J. Bacteriol. 167:110-116.18. Kuroda, A., and J. Sekiguchi. 1993. High-level transcription of the

major Bacillus subtilis autolysin operon depends on expression ofthe sigma D gene and is affected by a sin (flaD) mutation. J.Bacteriol. 175:795-801.

19. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

20. Laskowski, M., Jr., and I. Kato. 1980. Protein inhibitors ofproteinases. Annu. Rev. Biochem. 49:593-626.

21. Leighton, T. J., and R. H. Doi. 1971. The stability of messengerribonucleic acid during sporulation in Bacillus subtilis. J. Biol.Chem. 246:3189-3195.

22. Matsudaira, P. 1987. Sequence from picomole quantities ofproteins electroblotted onto polyvinylidene difluoride membranes.J. Biol. Chem. 262:10035-10038.

23. McGrath, M. E., W. M. Hines, J. A. Sakanari, R. J. Fletterick, andC. S. Craik. 1991. The sequence and reactive site of Ecotin. J. Biol.Chem. 266:6620-6625.

24. McKenzie, T., T. Hoshino, T. Tanaka, and N. Sueoka. 1986. Thenucleotide sequence of pUBIIO: some salient features in relationto replication and its regulation. Plasmid 15:93-103.

25. Miller, J. H. 1972. Experiments in molecular genetics, p. 352-355.Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.

26. Millet, J. 1977. Characterization of a protein inhibitor of intracel-lular protease from Bacillus subtilis. FEBS Lett. 74:59-61.

27. Millet, J. 1979. Characterization of an inhibitor of the intracellularprotease from Bacillus subtilis. Biochimie 61:385-391.

28. Murao, S., T. Ikenaka, B. Tonomura, and K. Hiromi. 1985.Introduction, p. 1-14. In K. Hiromi, K. Akasaka, Y. Mitsui, B.Tonomura, and S. Murao (ed.), Protein protease inhibitor-thecase of Streptomyces subtilisin inhibitor (SSI). Elsevier SciencePublishers B.V., Amsterdam.

29. Murao, S., Y. Shimizu, M. Kameda, and T. Nishino. 1982.Inhibition of sporulation of genus Bacillus by various microbialprotease inhibitors. Agric. Biol. Chem. 46:3075-3077.

30. Nishino, T., and S. Murao. 1986. Interaction of proteinaceousprotease inhibitor of Bacillus subtilis with intracellular proteasesfrom the same strain. Agric. Biol. Chem. 50:3065-3073.

31. Nishino, T., Y. Shimizu, K. Fukuhara, and S. Murao. 1986.Isolation and characterization of a proteinaceous protease inhib-itor from Bacillus subtilis. Agric. Biol. Chem. 50:3059-3064.

32. Nishiya, Y., and T. Imanaka. 1990. Cloning and nucleotide se-quences of the Bacillus stearothermophilus neutral protease geneand its transcriptional activator gene. J. Bacteriol. 172:4861-4869.

33. Read, R. J., and M. N. G. James. 1986. Introduction to the proteininhibitors: X-ray crystallography, p. 301-336. In A. J. Barrett andG. Salvesen (ed.), Proteinase inhibitors. Elsevier Science Publish-ers B.V., Amsterdam.

34. Ruppen, M. E., G. L. Van Alstine, and L. Band. 1988. Control ofintracellular serine protease expression in Bacillus subtilis. J.Bacteriol. 170:136-140.

35. Sadaie, Y., and T. Kada. 1983. Formation of competent Bacillussubtilis cells. J. Bacteriol. 153:813-821.

36. Saito, H., and K. Miura. 1963. Preparation of transforming DNAby phenol treatment. Biochim. Biophys. Acta 72:619-629.

37. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, 2nd ed. Cold Spring Harbor Labo-ratory, Cold Spring Harbor, N.Y.

38. Satola, S. W., J. M. Baldus, and C. P. Moran, Jr. 1992. Binding ofSpoOA stimulates spolIG promoter activity in Bacillus subtilis. J.Bacteriol. 174:1448-1453.

39. Sheehan, S. M., and R. L. Switzer. 1990. Intracellular serineprotease 1 of Bacillus subtilis is formed in vivo as an unprocessed,active protease in stationary cells. J. Bacteriol. 172:473-476.

40. Shiga, Y., K. Hasegawa, A. Tsuboi, H. Yamagata, and S. Udaka.1992. Characterization of an extracellular protease inhibitor ofBacillus brevis HPD31 and nucleotide sequence of the correspond-ing gene. Appl. Environ. Microbiol. 58:525-531.

41. Shimizu, Y., T. Nishino, and S. Murao. 1983. Purification andcharacterization of a membrane bound serine protease of Bacillussubtilis IFO 3027. Agric. Biol. Chem. 47:1775-1782.

42. Shimizu, Y., T. Nishino, and S. Murao. 1984. Inhibition ofsporulation of Bacillus subtilis by MAPI, a serine protease inhibi-tor, and interaction of MAPI with membrane bound protease.Agric. Biol. Chem. 48:365-372.

43. Shimizu, Y., T. Nishino, and S. Murao. 1984. Control of proteaseactivity during sporulation of Bacillus subtilis. Agric. Biol. Chem.48:3109-3114.

44. Sloma, A., G. A. Rufo, Jr., K. A. Theriault, M. Dwyer, S. W. Wilson,and J. Pero. 1991. Cloning and characterization of the gene for anadditional extracellular serine protease of Bacillus subtilis. J.Bacteriol. 173:6889-6895.

45. Snead, R., L. Day, T. Chandra, M. Mace, Jr., D. W. Bullock, andS. L. C. Woo. 1981. Mosaic structure and mRNA precursor ofuteroglobin, a hormone-regulated mammalian gene. J. Biol.Chem. 256:11911-11916.

46. Srivastava, 0. P., and A. I. Aronson. 1981. Isolation and charac-terization of a unique protease from sporulating cells of Bacillussubtilis. Arch. Microbiol. 129:227-232.

47. Studier, F. W., A. H. Rosenberg, J. J. Dunn, and J. W. Dubendorff.1990. Use of T7 RNA polymerase to direct expression of clonedgenes. Methods Enzymol. 185:60-89.

48. Sussman, M. D., and P. Setlow. 1991. Cloning, nucleotide se-quence, and regulation of the Bacillus subtilis gpr gene, whichcodes for the protease that initiates degradation of small, acid-soluble proteins during spore germination. J. Bacteriol. 173:291-300.

49. Udaka, S., and H. Yamagata. 1993. High-level secretion of heter-ologous proteins by Bacillus brevis. Methods Enzymol. 217:23-33.

50. Wu, X.-C., W. Lee, L. Tran, and S.-L. Wong. 1991. Engineering aBacillus subtilis expression-secretion system with a strain deficientin six extracellular proteases. J. Bacteriol. 173:4952-4958.

VOL. 175, 1993

on Novem

ber 17, 2020 by guesthttp://jb.asm

.org/D

ownloaded from