Vol. 170, No. 6

Transcriptional Regulation of a Promoter in the men Gene Cluster ofBacillus subtilis

PAUL MILLER,t JOHN MUELLER, KEVIN HILL, AND HARRY TABER*

Department of Microbiology, Immunology, and Molecular Genetics, Albany Medical College, Albany, New York 12208

Received 21 September 1987/Accepted 22 March 1988

The control of men gene expression during growth and sporulation of Bacillus subtilis was examined at thetranscriptional level. Two different approaches were used. (i) Steady-state levels of men-specific mRNA were

measured directly. (ii) A men'-lacZ gene fusion was constructed. In both cases, it was observed that men

promoter activity was maximal at the onset of sporulation and declined soon thereafter. These kinetics were

similar to the pattern of menaquinone accumulation previously observed. Expression from the men promoterwas independent of the presence of the products of the spoOA and spoOH genes and was enhanced by additionof glucose and glutamine to the culture medium. DNA sequence analysis of the promoter region revealed a

potential recognition site for the principal vegetative form of RNA polymerase but not for any of the knownminor polymerase forms. The functionality in vivo of the promoter sequence was confirmed by high-resolutionSl nuclease mapping of the transcript start site. An additional sequence element was identified that is sharedby the sdhA, ciG, and ctaA promoters and may indicate a common regulatory mechanism in the expression ofthese genes.

During the process of sporulation, Bacillus subtilis cellsundergo several dramatic changes involving metabolism andgene expression. The latter changes are characterized by a

modification of the transcriptional machinery, with concom-itant activation of a large class of sporulation-specific (spo)genes, apparently in a controlled temporal fashion (20). Thisprogram of gene expression is largely dependent on theexpression of a subset of early sporulation genes encoded bythe spoO loci.

Metabolic alterations that coincide with the onset ofsporulation include derepression of tricarboxylic acid (TCA)cycle enzymes, as well as changes in the composition of theelectron transport chain (3, 7, 10, 14). Presumably, thesechanges reflect a need to maintain adequate energy suppliesin the absence of a rapidly metabolizable carbon source andconstitute a condition necessary for sporulation. It is rea-

sonable to suppose that common control elements regulatethese switches in the patterns of metabolism and geneexpression. However, little is known about the mechanismswhich control either TCA or electron transport gene expres-sion. In probably the best studied of these functions, it hasbeen shown that modulations in aconitase activity, theproduct of the citB gene, are largely the result of transcrip-tional control (5, 27).

Previously, we described the cloning of the genes thatencode the enzymes involved in the biosynthesis of mena-

quinone (MK; vitamin K2), an essential component of the B.subtilis electron transport chain (25). Cellular levels of thissmall lipophilic molecule increase steadily as cells approachsporulation, reaching maximal levels shortly after To (theonset of sporulation) and falling off thereafter (8). Therequirement of this compound for sporulation was demon-strated by using a conditional mutant that can be renderedpartially MK deficient. Such a situation severely compro-

* Corresponding author.t Present address: Laboratory of Molecular Genetics, National

Institute of Child Health and Human Development, Bethesda, MD20892.

mises sporulation while leaving vegetative growth and res-

piration unaffected (7).To examine the involvement of transcriptional mecha-

nisms in controlling MK accumulation, we used cloned mensequences as hybridization probes to study the levels of menmRNA in vegetative and sporulating cells. In addition, a

men'-lacZ transcriptional gene fusion was constructed tofacilitate these analyses. Quantitation of men transcripts andf-galactosidase assays indicated that, at the onset of sporu-lation, transcription from the men locus increases in a

manner analogous to accumulation of MK. Studies with themen'-lacZ fusion in a spoOA or spoOH background indicatedthat these early sporulation loci do not control men geneexpression. Finally, DNA sequence analysis of the men

promoter region revealed significant homology only with thenormal vegetative promoter elements recognized by the UA(u43) form of RNA polymerase.

MATERIALS AND METHODS

Bacterial strains and plasmids. The bacterial strains usedare shown in Table 1. Strain RB938 was constructed bytransforming strain RB929 to Met' under congression con-ditions and screening for Spo- recombinants. StrainsRB1013 and RB1014 were constructed by transformation ofstrain RB929 with DNA from strain ZB480, which containsTn917 linked to spoOH. Selection was for macrolide-lincos-amide-streptogramin resistance (MLSr), encoded by the ermgene of Tn917. Lac' recombinants were identified byscreening with methylumbelliferyl-p-D-galactoside. PlasmidpAI46 has been described previously (25). This plasmidcontains a 650-base-pair (bp) HindIll fragment in which islocated a promoter required for the expression of men genes

(Fig. 1). A men'-lacZ transcriptional gene fusion was createdby cloning this fragment into the integrable lacZ fusionvector pDEB1 (31). The structure of this plasmid, pAI103, isshown in Fig. 2. Strains containing chromosomally inte-grated single copies of pAI103 were generated by transfor-mation of appropriate recipients and selection for chloram-phenicol resistance.

2742

JOURNAL OF BACTERIOLOGY, June 1988, p. 2742-27480021-9193/88/062742-07$02.00/0Copyright © 1988, American Society for Microbiology

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

EXPRESSION OF B. SUBTILIS men GENES 2743

TABLE 1. Bacterial strains

Strain Genotype Source

RB1 trpC2 This laboratoryRB896 c1(men'-lacZ) cat trpC2 RB1 x pAI103RB47 trpC2 lys-3 metBJO This laboratoryRB929 '1(men'-lacZ)cat trpC2 lys-3 metB1O RB47 x pAI1031S9 trpC2 pheAl spoOA12 BGSCaRB938 t(men'-lacZ)cat trpC2 lys-3 spoOA12 RB929 x DNA of 1S9ZB480 chr::Tn917fQHU146 trpC2 pheAl spoOHAHindIII P. ZuberRB1013 F(men'-lacZ)chr::Tn9J7flHU146cat trpC2 lys-3 metBlO RB929 x DNA of ZB480RB1014 1(men'-lacZ)chr::Tn9J7fZHU146cat trpC2 Iys-3 metBJO spoOHAHindIII RB929 x DNA of ZB480

a BGSC, Bacillus Genetic Stock Center, Ohio State University, Columbus.

Culture media. The standard medium for cultivation of B.subtilis strains was LB (LB is 1% tryptone [Difco Labora-tories, Detroit, Mich.], 0.5% yeast extract [Difco], and 0.5%NaCI). Chloramphenicol, when needed, was added to 5,ug/ml. Sporulation was achieved by growing cells in thedouble-strength nutrient sporulation medium (2x NSM) ofSchaeffer et al. (30).DNA manipulations and transformation. The methods used

for restriction endonuclease digestions were those suggestedby the supplier. DNA ligation reactions were performed asdescribed by Struhl (33). Chromosomal DNA was isolatedfrom B. subtilis as described by Saunders et al. (29). Small-scale plasmid isolation from Escherichia coli was performedas described by Birnboim and Doly (2); for large-scalepreparations, a further purification by equilibrium centrifu-gation in CsCl gradients was performed. E. coli strains weretransformed with plasmid DNA as described by Hanahan(13), and B. subtilis transformations were as described byPiggot (Spores Newsl. 8:246-250).

Assay of ,l-galactosidase in men'-lacZ fusion-containingstrains. Cells were harvested and frozen overnight at -70°C.Thawed cells were suspended in Z buffer, 0.1% toluene wasadded, and the emulsion was vortexed before being placedon ice for 10 min (35). ,-Galactosidase activity was mea-sured by the o-nitrophenylgalactoside procedure and ex-pressed in Miller units (24). Endogenous enzyme activitieswere 2 to 3 Miller units and did not show variation during theperiod of sampling.RNA purification and analysis. For isolation of RNA,

cultures of strain RB1 were grown in 2x NSM and 80-mlsamples were removed at the desired times. The sampleswere chilled rapidly by swirling for 10 to 30 s in liquidnitrogen, the cells were pelleted by centrifugation, and thepellets were frozen in liquid nitrogen. Extraction of RNAwas performed by a modification of a protocol of A. L.Sonenshein and C. Mathiopoulos (unpublished data). Thecell pellets were thawed and converted to protoplasts at 40C.

I

H, P H650 bp

Y PIH E SI1I

FIG. 1. Physical map of the B. subtilis men region, showing the

direction (arrow) of transcription (32). The 650-bp HindIII fragmentcontaining the origin of transcription and putative promoter isshown in expanded form. Abbreviations: H, HindIII; P, PstI; E,EcoRI; S, Sacd; K, KpnI; B, BamHI; C, ClaI.

After pelleting, the protoplasts were lysed in 8.0 ml ofLETSbuffer (100 mM LiCl, 10 mM EDTA, 10 mM Tris hydrochlo-ride [pH 7.4], 1% sodium dodecyl sulfate), and 10 ml of 4 Mguanidinium isothiocyanate was added and mixed. Afterextraction with phenol-chloroform, LiCl was added to 0.2M, and the nucleic acid was precipitated with 2.5 volumes of95% ethanol at -20°C. The precipitate was collected bycentrifugation and digested with RNase-free DNase (RQ1DNase; Promega, Madison, Wis.). To map the 5' terminus ofthe men transcript, a DNA probe (indicated in the text) waslabeled at its 5' end by the sequential action of calf intestinalalkaline phosphatase (Boehringer Mannheim Biochemicals,Indianapolis, Ind.) and T4 polynucleotide kinase (BethesdaResearch Laboratories, Gaithersburg, Md.), in conjunctionwith [_y-32PIATP (>3,000 Ci/mmol; New England NuclearCorp., Boston, Mass.), as described by Arrand (1). Theend-labeled probe was purified from the unincorporatedlabel by three successive ethanol extractions. S1 mappingwas performed with the double-stranded probe by themethod of Favaloro et al. (9). S1 nuclease-resistant productswere analyzed by electrophoresis in a 6% polyacrylamide-8M urea sequencing gel. In general, 50 pug of B. subtilis RNAand 50,000 cpm of probe were used for each analysis. For

H _p rpB/ w~~~~~~rpA

pAI103(16.5kb)

CATFIG. 2. Structure of the men'-lacZ fusion plasmid pAI103. The

heavy bar denotes the HindlIl fragment from pAI46. The pDEB1vector sequences are indicated by the solid line. The construction ofpAI103 is described in Materials and Methods. Abbreviations: Amp,ampicillin resistance determinant; CAT, chloramphenicol acetyl-transferase gene; ori, origin of plasmid replication for E. coli. H,HindlIl sites; P, PstI site; kb, kilobases.

E H KE BI I 11 1----

VOL. 170, 1988

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

2744 MILLER ET AL.

HindIII +1

MGMet

Gm ACA AM GIG CW wr AMl T CA AS AS MT C C GCM TI AAAVal Tlr mIr Val Gln Arg fhr Fhe Arg Lys Lys Val Lau His Ala Iau His Iys

T,aqIGOC M GAA G1CAMC CAT OCr GI'C TTA AIMAMC TAT T3 AAla Lys Glu Val Asn His Ala Val Ieu Ile Ser Tyr Ser

74

144

198

238 s

FIG. 3. Nucleotide sequence of 238 bp of DNA surrounding themen promoter. The -35 and -10 recognition sites for the vegetativeB. subtilis RNA polymerase are underlined. Also indicated are thetranscriptional start site (+1), a putative ribosome-binding site(double underlining), and the N-terminal portion of the proximalopen reading frame. The HindIll site at the beginning of thesequence corresponds to the leftmost HindlIl site in Fig. 1. Over-lining indicates a sequence shared with the citG, sdh, and ctaApromoters (see the text).

time course experiments, equal amounts ofRNA were addedto each hybridization reaction, as normalized by A260 andvisual examination of fluorescence intensity of 16S and 23SrRNA in ethidium bromide-stained agarose gels. The DNAsequencing ladder used to identify the precise mRNA startsite was obtained by the dideoxy method (28) with a men-specific synthetic oligonucleotide whose 5' end corre-sponded exactly to the 5' end of the ClaI site labeled in theseexperiments. That is, the sequencing ladder and Si probehad identical 5' ends; see Kudo et al. (19). The probe wasdetermined to be in excess under these conditions (data notshown).DNA sequencing. The nucleotide sequence of the promoter

region of the men cluster was determined on both strands bythe dideoxy chain termination method (28). Two procedureswere used. Procedure 1 was the double-stranded sequencingmethod of Chen and Seeburg (4) performed with a kitsupplied by International Biotechnologies, Inc. PlasmidpAI46 (25) was used as a template for sequencing. Procedure2 involved the use of the M13 single-stranded phage vectorsM13mpl8 and mp19 (22). In this case, the upstream HindIII-ClaI fragment of pAI46 (Fig. 1) was cloned into HindIII-AccI-digested M13 vectors, and single-stranded DNA waspurified from these recombinant phages before sequencing.

RESULTS

Nucleotide sequence of the men promoter region. Previ-ously, we described the cloning of the men gene cluster on aseries of overlapping plasmids (25). On the basis of plasmidintegration and transcriptional orientation studies, it wasconcluded that sequences necessary for initiation of mengene transcription are located within a 650-bp Hindlll frag-ment, with the direction of transcription as shown in Fig. 1.Since low-resolution S1 mapping experiments (data notshown) suggested that the initiation site for mRNA synthesismapped within 75 bp of the upstream HindIII site shown inFig. 1, the DNA sequence was obtained starting from thisHindIII site. The first 238 bp of this sequence are shown inFig. 3. A likely recognition site for the vegetative crA form ofRNA polymerase was observed upstream from the mRNAstart site identified at position 47 by S1 mapping (cf. Fig. 4).These recognition sequences exhibit four-of-six homology inthe -35 region and a perfect six-of-six match in the -10domain. Further downstream, sequences associated with theinitiation of protein synthesis (26) were observed. A strongribosome binding site, GGAGG, with a AG value of -14.4kcal, was found at position 129, followed by an ATG codon

3AA\A

A'.

H CL

H 308 cL _.. ..-

262... w 1

A C G T 1 2 3

Probe

Protected fragment

FIG. 4. Identification of the 5' end of the men transcript bynuclease Si mapping. End-labeled probes were prepared frompAI46 and made radioactive as described in Materials and Methods.The probe was the HindIII-ClaI fragment, labeled at the 5' end ofthe Clal site as indicated below the gel. Fifty thousand cpm of probewas hybridized to 50 ,ug ofRNA isolated from B. subtilis cells at Toin sporulation medium. S1 nuclease-resistant products were ana-lyzed by electrophoresis through a 6% polyacrylamide-8 M ureasequencing gel. Lanes: 1, probe hybridized with B. subtilis To RNAand treated with Si nuclease; 2, probe hybridized with S. cerevisiaetRNA and treated with Si nuclease; 3, probe alone (untreated); A,C, G, and T, sequencing reactions run as single-base size markerswith a men-specific synthetic oligonucleotide. The 5' end of thesequencing primer corresponds to the 5' end of the labeled Clal sitein the probe. Thus, the sequencing ladder corresponds exactly to theSi probe in orientation and 5' end. The sequence of the noncodingstrand in the region surrounding the protected 5' end of the menmRNA is shown. The arrow indicates the base at which mentranscription begins. The band migrating more slowly than theprotected fragment in lanes 1 and 2 corresponds to the full-lengthprobe not completely digested with Si nuclease.

at position 142. Additionally, Magnusson and co-workersobserved two regions of homology between the B. subtiliscitG (fumarase) and sdh (succinate dehydrogenase) promot-ers (21). One of these sequences is the undecamer TTCT-TATGAAA from the sdh promoter, which is conserved with9-of-11 homology in the citG promoter (23). This samesequence was found beginning at position 15 in the sequenceshown in Fig. 3, disturbed only by insertion of an AT doubletafter the first C of the sdh motif. This sequence overlaps theproposed -35 region of the men promoter.

Identification of the 5' end of the men transcript. Todetermine precisely the 5' end(s) of men-specific transcriptswithin the sequence shown in Fig. 3, a high-resolution Sinuclease protection experiment was performed. Strain RB1was grown in nutrient sporulation medium, and at time To(the measured end of exponential growth) the cells wereharvested and RNA was purified. This RNA was thenhybridized to a HindIII-Clal probe end labeled at the Clalsite as shown in Fig. 4. After treatment of the hybridization

J. BACTERIOL.

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

EXPRESSION OF B. SUBTILIS men GENES 2745

products with Si nuclease, the samples were analyzed on a6% polyacrylamide-8 M urea sequencing gel. A protectedfragment was identified that begins at sequences surroundinga G residue located 47 nucleotides downstream from theHindIlI site (shown at the left end of the cloned men regionin Fig. 4; cf. Fig. 1). Thus, it appears that the men transcriptinitiates within this region, confirming previous genetic andphysical data (25). Nuclease protection experiments withprobes containing additional sequences downstream fromthe ClaI site indicate that this transcript extends past thePstI and HindIll sites shown in Fig. 1 and toward the menstructural genes (data not shown).

Quantitation of men transcript levels during growth andsporulation. Cellular levels of menaquinone increase as cellscommence the sporulation process, peaking shortly after To.These levels then diminish steadily as the culture continuesthrough sporulation (8). To determine whether this alterationin quinone concentration is the result of a transcriptionalcontrol process, men-specific mRNA levels were assessed atvarious stages of growth and development by using the Simapping technique of the previous experiment. Briefly, RB1cells were grown in 2x NSM, and at various times, sampleswere removed and the RNA was extracted. This RNA wasthen hybridized with the HindIII-ClaI probe used in themRNA mapping experiment described above (Fig. 4). Nu-clease Si-resistant hybrids were analyzed on a sequencinggel as described before. The results show that the level ofmen-specific transcripts increases as the cells progress fromexponential phase to To (Fig. 5, lanes 1 and 2). After To, thetranscript levels decrease as the cells progress throughsporulation.

Analysis of men transcriptional patterns with a men'-lacZfusion. To observe the patterns of men transcription morereadily, the promoter region from the men cluster was fusedto the structural gene for E. coli ,-galactosidase. To generatethis construct, the 650-bp HindIII fragment containing themen promoter region (Fig. 1) was cloned into the HindIII siteof the integrable lacZ fusion vector pDEB1 (31), generatingthe plasmid pAI103 (Fig. 2). This plasmid does not replicatein B. subtilis but contains a chloramphenicol acetyltrans-ferase gene that is active in gram-positive bacteria. Chlor-amphenicol-resistant transformants of B. subtilis were ob-tained by integration of pAI103 into the men region of thechromosome. As expected, all such integrants had Men'Spo+ phenotypes, indicating that the integration event didnot disrupt the men cluster. Southern blot analysis was usedto verify that a single copy of pAI103 had been integrated(data not shown). Initially, wild-type strains were used asrecipients in integration experiments and the levels of 1B-galactosidase produced during growth and sporulation ofthese integrants were then measured. The levels of P-galactosidase activity, and thus of men promoter activity, inthese strains increase modestly in the To- to-T1 period andthen steadily decrease after this time (Fig. 6). We found thatdifferent men' spo+ strains gave slightly different men'-lacZkinetics during exponential growth in sporulation medium.For example, strain RB896 (RBlfpAI103) had less 1-galac-tosidase activity at T-1 (data not shown) than did strainRB929 or RB1013 (Fig. 6), although the kinetics in thepost-TO period were very similar for all men' spo+ strains.The pattern of 1-galactosidase expression by men'-lacZfusion-containing strains followed the pattern of men mRNAaccumulation shown in Fig. 5 and indicates that the men'-lacZ fusion vector is a suitable means for analyzing menpromoter activity.

Effect of spoO mutations on men transcription. As men-

A C G T 1 2 3 4 5 6 7 8

FIG. 5. Quantitative Si mapping of men transcripts duringgrowth and sporulation of B. subtilis RB1. The probe was theHindIII-ClaI fragment shown in Fig. 3. Fifty thousand cpm of thisprobe was hybridized with 50 ,ug of each RNA sample, followed bydigestion with nuclease Si (400 U/ml) and analysis of protectedhybrids on a 6% sequencing gel. Lanes: 1, RNA from vegetativecells; 2, RNA from cells grown to To; 3, RNA from cells grown to T1;4, RNA from cells grown to T2; 5, RNA from cells grown to T3; 6,RNA from cells grown to T4; 7, S. cerevisiae RNA; 8, probe alone(untreated); A, C, G, and T, sequencing reactions used as molecularweight markers as described in the legend to Fig. 4. The bandmigrating more slowly than the protected fragment corresponds tothe full-length probe not completely digested with Si nuclease (lanes1 to 7). The numbers on the right indicate fragment size, in basepairs.

tioned previously, the spoO genes represent a group of locithat are necessary for the switch from vegetative to sporu-lation-specific gene expression (20). To determine whetherthe products of these early sporulation genes are necessaryfor the changes in men expression described above, spoOAand spoOH strains containing pAI103 were constructed. The4(men'-lacZ) spoO strains were cultured in 2x NSM, and13-galactosidase activity was measured during growth andsporulation. The result is shown in Fig. 6, with men'-lacZexpression plotted as a function of hours before and after To.The data (spoOA, Fig. 6A; spoOH, Fig. 6B) show that thepattern of men'-lacZ expression in these spoO strains wassimilar to that of their isogenic spo+ counterparts, indicatingthat the spoOA and spoOH gene products do not influenceexpression from this men promoter.Response of men expression to altered nutritional condi-

tions. The experiments with the spoO mutants just describedsuggest that the men promoter may not respond to develop-mental signals per se but instead to physiological conditionsestablished during the growth phases of the culture. Toexamine this possibility, strain RB896 was grown in LB, amedium in which sporulation is largely repressed. Analysisof 1-galactosidase levels from this culture (Fig. 7) showedthat the overall pattern of 1-galactosidase activity was verysimilar to that of sporulating cultures, i.e., a maximum in theTo 5-to-T1 0 period, followed by a decline. In addition, a 24-hculture of the fusion-containing strain grown in LB producedfewer than 1% spores, whereas the same strain in sporula-tion medium resulted in greater than 90% spores, as deter-mined by phase-contrast microscopy. Thus, the increase inmen promoter activity appears to reflect the growth phase ofthe culture rather than the developmental situation. Sinceelectron transport function is related to the TCA cycle incellular metabolism, it was of interest to determine whetherthe activity of the men promoter could be modulated bynutritional variations in the way that the TCA cycle gene

VOL. 170, 1988

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

2746 MILLER ET AL.

2,-z

-

CZ

3

204z

-

-2.0 0.0 2'.0 4'.0TIME (hours)

FIG. 6. Activity profiles of P-galactosidase in B. subtilis spo+and spo integrants of pAI103, containing the men'-IacZ fusion.Cultures were grown in 2 x NSM, and samples were removed at theindicated times. The time scale is given relative to To, the measuredend of exponential-phase growth. Enzyme assays were performedas described in Materials and Methods. (A) Symbols: O,, strainRB929 (spo+); 0, strain RB938 (spoOAl2). (B) Symbols: El, strainRB1013 (spo+); 0, strain RB1014 (spoOHAHindIII).

citB (encoding aconitase) has been shown to respond. ThecitB gene is repressed in a synergistic manner by glucose andglutamine (27). To study this relationship', strain RB896 wasgrown in LB broth supplemented with 5% glucose and 0.2%glutamine, samples of cells were removed at various timesduring the growth and early stationary phases and analyzedfor ,B-galactosidase activity. The result is shown in Fig. 7.Although the maximal values for ,B-galactosidase activity atT05 were similar in supplemented and unsupplemented LB,the effect of glucose-glutamine supplementation was to pre-vent the decline in activity associated with growth in unsup-plemented medium. On further analysis of this phenomenon,it was clear that glucose alone had a larger effect than didglutamine alone. We also observed that 0.5% glucose gav'e aresponse quantitatively similar to that produced by 5.0%glucose (data not shown). From these data we conclude thatthe men promoter is affected by 'metabolic signals' in amanner distinct from that of at least one TCA cycle gene,citB.Changes in ,B-galactosidase activity during growth phase

transitions or in different media could have resulted fromaltered stability of the enzyme. To test this possibility, strainRB896 was grown in LB or LB containing 5% glucose and0.2% glutamine and treated with 100 ,ug of chloramphenicolper ml at To, T1, and T2. After addition of chloramphenicol,

20_z

-J-j

~ 0

_-2.0 0.0 2.0 4.0TIME (hours)

FIG. 7. Effects of glucose and glutamine on men expression.Activity profiles of P-galactosidase in strain RB896 (spo+ 4imen'-lacZ) grown in LB (0), LB-0.2% glutamine (O), LB-5% glucose(A), or LB-0.2% glutamine-glucose (*). The time scale and enzymeassays were as indicated in the legend to Fig. 6.

P-galactosidase activity was measured at 3-min intervals for15 min. Rates of enzyme decay measured in this way werevery similar for cultures grown in either medium (data notshown).

DISCUSSION

In the accompanying paper (25), we described the cloningof most of the B. subtilis sequences that encode the mengene cluster. Using integrable plasmids containing portionsof these men sequences, we identified by genetic and phys-ical criteria the approximate 5' boundary of the men tran-scriptional unit. If the positioning of this boundary is correct,an additional gene, previously undescribed for this locus,may lie between the men promoter and the first structuralgene, menCD, mapped by plasmid rescue of mutations. Thisinference was made because we observed that an interval ofapproximately 1.6 kilobase pairs ofDNA occurs between the5' end of the transcription unit and the furthest upstreammenCD allele (unpublished data). Here we present directevidence for the presence of an initiation site for men mRNAsynthesis. This transcriptional initiation site lies very closeto the upstream HindIII site previously defined as beingoutside of the men transcriptional unit (25; Fig. 1). Webelieve that this mRNA molecule represents a primary mentranscript for the following reasons. (i) As shown previously,integrant plasmids that contain sequences extending up-stream as far as the PstI site (Fig. 1) generate transformantswhich exhibit partial men auxotrophy as compared with thefully Men' transformants generated by integrant plasmidswhose upstream boundary is the 5' Hindlll site. (ii) ThemRNA molecule identified in the Si mapping experimentspresented here extends downstream from the PstI site (Fig.4; unpublished data).The kinetics of men-specific mRNA synthesis in cells

grown in 2x NSM (Fig. 5) generally resemble the pattern ofMK accumulation observed in cells cultured under similarconditions (8). This observation suggests that the control ofMK accumulation under these conditions is largely tran-scriptional. The conclusion is supported by results obtainedwith a men'-IacZ transcriptional fusion, integrated in a singlecopy at the men locus. Changes in men transcription were

J. BACTERIOL.

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

EXPRESSION OF B. SUBTILIS men GENES 2747

not dependent on recognized sporulation-specific controlmechanisms, since mutations in two early developmentalgenes, spoOA and spoOH, had no effect on the pattern of menexpression. Moreover, analysis of men promoter activity inthe fusion-containing strain when cultivated in LB showed a

similar pattern of expression, even though this latter mediumsupports sporulation very poorly. Thus, it appears that thechanges observed in the level of men transcription, andpresumably the level of MK biosynthesis, reflect primarilythe growth phase and not the developmental status of theculture.Rosenkrantz et al. observed that transcription of the citB

gene is repressed in a synergistic manner by glucose andglutamine (27). We did not observe repression of men butinstead found that glucose-glutamine allowed expression tobe maintained at about 75% of the maximal value up to atleast T4 of the postexponential period, at a time whenexpression in unsupplemented medium had declined to lessthan 10%. Although we report here data obtained in nonspo-

rulation medium (LB), similar results were obtained withsporulation medium (data not shown). We also found thatdecoyinine, which activates citB expression in the presence

of glucose and glutamine (5), did not increase men'-lacZexpression under similar conditions (H. Taber and C. Gus-tafson, unpublished data). The response of men expressionis somewhat reminiscent of the ctc gene (18; see also theDiscussion in reference 5). Expression of both was increasedby glucose and glutamine; also, like ctc, men expression isstimulated in TCA cycle mutants (K. Hill, unpublisheddata). However, the similarity breaks down when the kinet-ics. of expression in sporulation medium, and responses tothe presence of spoOA and spoOH mutations (stimulatory toctc; no effect on men) are considered. In addition, we haveno evidence that minor forms of RNA polymerase are

involved in men transcription, as they are for ctc.There may be additional effects on men expression that

reflect the complexity of the TCA cycle. The citB gene

product, aconitase, is a component of the first half of theTCA cycle, the reactions of which result in the conversion ofglycolytic products to o-ketoglutarate. This latter compoundis an important intermediate in the biosynthesis of theglutamate family of compounds. Thus, addition of glutamineto culture medium can feed back to repress the first half ofthe TCA cycle (15, 27). Increased levels of a-ketoglutarate,resulting from glutamine supplementation, may lead to acti-vation of the second half of the TCA cycle. This latteroxidative process requires an active electron transport chainto reoxidize the reduced coenzymes (NADH and flavinadenine dinucleotide) utilized in the conversion of a-keto-glutarate to oxaloacetate. Regulation of the men genes may,

as a consequence, have more in common with genes thatcontrol the dicarboxylic acid part of the TCA cycle.DNA sequence analysis of the men promoter region

revealed structures associated with the binding site for theprincipal vegetative form of RNA polymerase (Eo.A) but notwith that of any other polymerase form (6, 26). This suggeststhat men expression is not subject to sporulation-specificcontrol mechanisms. It is interesting that the spacing be-tween the -10 and -35 sequences is only 16 bp, in contrastto the 17-bp spacing normally found in B. subtilis. Thisreduced spacing also exists for the argC promoter (32) andmay explain the weak activity of both of these promoters inspite of good homology to consensus sequences in the -10region. The low activity of the men promoter may also bedue to lack of an invariant G in the -35 region. Of thegram-positive promoters analyzed by Graves and Rabino-

witz (11), 72% have a G at this position; furthermore,replacement of this G in E. coli promoters is associated withpromoter down mutations (16). In Fig. 3, signals for -theinitiation of translation are located downstream from thepromoter region. These signals include a strong ribosome-binding site, as well as an appropriately spaced ATG codonfor the start of protein synthesis (12). The assignments arestrengthened by our recent construction of a men'-IacZprotein fusion; activation of P-galactosidase was observed inE. coli and B. subtilis when the first 32 codons of thepromoter-proximal gene were fused to codon 8 of lacZ,which implies that this ATG (+95) is functionally active invivo (Hill et al., in preparation).

In their analysis of the sdhA promoter region, Magnussonet al. (21) observed two short sequences which also havebeen found in a highly conserved form in the citG promoter(25). One of these sequences, TTCTTATGAAA, is alsofound as a nearly perfect element in the men promoter. Thisundecamer, which starts at position 15 of the sequenceshown in Fig. 3, is identical to the sdhA copy, except forinsertion -of an AT doublet after the first C residue in the menversion. Interestingly, this sequence element overlaps theproposed -35 sequence of the men promoter. Recently, wehave discovered a similar sequence in the promoter region ofctaA, a controlling gene for cytochrome aa3 oxidase in B.subtilis (J. Mueller and H. Taber, manuscript in prepara-tion). In the ctaA example, the sequence occurs with 8-of-11homology in the AT-rich region upstream from the -35sequence. This sequence, which is common to the sdhA,citG, ctaA, and men promoters, may represent a binding sitefor a regulatory factor. Indeed, if one examines the last ninenucleotides of the men sequence element, TTTCTGAAA,symmetry is observed around a central T residue. Suchsequences have been associated with the binding of dimericregulatory molecules, for example, the GCN4 protein ofSaccharomyces cerevisiae (17 and references therein). Site-directed mutagenesis of this region should reveal any signif-icance of this element.

Sporulation probably requires a high level of metabolicproficiency, including the ability to synthesize MK at a levelhigher than that needed for growth. The metabolic status ofthe cell is an important part of the signaling mechanism thatactivates the developmental sequence. The effects of nutri-tional supplementation on men expression suggest a uniqueresponsiveness to metabolic conditions by comparison withthe citB gene. Sequence analysis of the men promoter,however, suggests possible common regulatory elementswith-the sdhA, citG, and ctaA genes, to which it is relatedmetabolically. We may expect to find that the men genecluster possesses control circuits characteristic of not onlycatabolic but also biosynthetic genes, which are now becom-ing the subject of close scrutiny in B. subtilis (34).

ACKNOWLEDGMENTS

We thank P. Zuber for strain ZB480, A. L. Sonenshein forplasmid pDEB1, A. Hinnebusch and R. Wek for help with the S1mapping experiments, C. Gustafson for P-galactosidase assays, andM. Cavanaugh for preparation of the manuscript.

This work was supported by Public Health Service grantGM34931 from the National Institute of General Medical Sciences.

LITERATURE CITED1. Arrand, J. E. 1985. Preparation of nucleic acid probes, p. 34. In

B. D. Hames and S. J. Higgins (ed.), Nucleic acid hybridization,a practical approach. IRL Press, Oxford.

2. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction

VOL. 170, 1988

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

2748 MILLER ET AL.

procedure for screening recombinant plasmid DNA. NucleicAcids Res. 7:1513-1523.

3. Chaix, P., and J. F. Petit. 1957. Influence du faux de croissancesur la constitution du spectra hematinique de B. subtilis. Bio-chim. Biophys. Acta 25:481-486.

4. Chen, E. Y., and P. H. Seeburg. 1985. Supercoil sequencing: afast and simple method for sequencing plasmid DNA. DNA4:165-170.

5. Dingman, D. W., M. S. Rosenkrantz, and A. L. Sonenshein.1987. Relationship between aconitase gene expression and spo-rulation in Bacillus subtilis. J. Bacteriol. 169:3068-3075.

6. Doi, R. H., and L.-F. Wang. 1986. Multiple procaryotic ribonu-cleic acid polymerase sigma factors. Microbiol. Rev. 50:227-243.

7. Farrand, S. K., and H. W. Taber. 1973. Physiological effects ofmenaquinone deficiency in Bacillus subtilis. J. Bacteriol.115:1035-1044.

8. Farrand, S. K., and H. W. Taber. 1974. Changes in menaqui-none concentration during growth and early sporulation inBacillus subtilis. J. Bacteriol. 117:324-326.

9. Favaloro, J., R. Freisman, and R. Kamen. 1980. Transcriptionmaps of polyoma virus-specific RNA: analysis by two-dimen-sional nuclease Si gel mapping. Methods Enzymol. 65:718-750.

10. Fortnagel, P. 1970. The regulation of aconitase and isocitratedehydrogenase in sporulation mutants of Bacillus subtilis. Bio-chim. Biophys. Acta 222:290-298.

11. Graves, M. C., and J. C. Rabinowitz. 1986. In vivo and in vitrotranscription of the Clostridium pasteurianum ferridoxin gene.J. Biol. Chem. 26:11409-11415.

12. Hager, P. W., and J. C. Rabinowitz. 1985. Translational speci-ficity in Bacillus subtilis, p. 1-32. In D. Dubnau (ed.), Themolecular biology of the bacilli, vol. 2. Academic Press, Inc.,New York.

13. Hanahan, D. 1985. Techniques for transformation of E. coli, p.109-135. In D. M. Glover (ed.), DNA cloning-a practicalapproach, vol. 1. IRL Press, Oxford.

14. Hanson, R. S., J. Blicharska, and J. Szulmajster. 1964. Obser-vations on the regulation of the synthesis of the tricarboxylicacid cycle enzymes of Bacillus subtilis Marburg. Biochem.Biophys. Res. Commun. 17:690-695.

15. Hanson, R. S., and D. P. Cox. 1967. Effect of different nutri-tional conditions on the synthesis of tricarboxylic acid cycleenzymes. J. Bacteriol. 93:1777-1787.

16. Hawley, D. K., and W. R. McClure. 1983. Compilation andanalysis of E. coli promoter DNA sequences. Nucleic Acid.Res. 11:2237-2255.

17. Hill, D. E., I. A. Hope, J. P. Mackee, and K. Struhl. 1986.Saturation mutagenesis of the yeast his3 regulatory site: require-ments for transcriptional induction and for binding by GCN4activator protein. Science 234:451-457.

18. Igo, M. M., and R. Losick. 1986. Regulation of a promoter thatis utilized by minor forms of RNA polymerase holoenzyme in

Bacillus subtilis. J. Mol. Biol. 191:615-624.19. Kudo, T., J. Yoshitake, C. Kato, R. Usami, and K. Horikoshi.

1985. Cloning of a developmentally regulated element fromalkalophilic Bacillus subtilis DNA. J. Bacteriol. 161:158-163.

20. Losick, R., P. Youngman, and P. J. Piggot. 1986. Genetics ofendospore formation in Bacillus subtilis. Annu. Rev. Genet.20:625-669.

21. Magnusson, K., M. K. Philips, J. R. Guest, and L. Rutberg.1986. Nucleotide sequence of the gene for cytochrome b558 ofthe Bacillus subtilis succinate dehydrogenase complex. J. Bac-teriol. 166:1067-1071.

22. Messing, J. 1983. New M13 vectors for cloning. MethodsEnzymol. 101:20-78.

23. Miles, J. S., and J. R. Guest. 1985. Complete nucleotidesequence of the fumarase gene citG of Bacillus subtilis 168.Nucleic Acids Res. 13:131-139.

24. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

25. Miller, P., A. Rabinowitz, and H. Taber. 1988. Molecularcloning and preliminary genetic analysis of the men gene clusterof Bacillus subtilis. J. Bacteriol. 170:2735-2741.

26. Moran, C. P., N. Lang, S. F. J. LeGrice, G. Lee, M. Stephens,A. L. Sonenshein, J. Pero, and R. Losick. 1982. Nucleotidesequences that signal the initiation of transcription and transla-tion in Bacillus subtilis. Mol. Gen. Genet. 186:339-346.

27. Rosenkrantz, M. S., D. W. Dingman, and A. L. Sonenshein.1985. Bacillus subtilis citB gene is regulated synergistically byglucose and glutamine. J. Bacteriol. 164:155-164.

28. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.USA 74:5463-5467.

29. Saunders, C. W., B. J. Schmidt, M. S. Mirot, L. D. Thompson,and M. S. Guyer. 1984. Use of chromosomal integration in theestablishment and expression of blaZ, a Staphylococcus aureusP-lactamase gene, in Bacillus subtilis. J. Bacteriol. 157:718-726.

30. Schaeffer, P., J. Millet, and J.-P. Aubert. 1965. Catabolicrepression of bacterial sporulation. Proc. Natl. Acad. Sci. USA54:704-711.

31. Schreier, H. J., and A. L. Sonenshein. 1986. Altered regulationof the glnA gene in glutamine synthetase mutants of Bacillussubtilis. J. Bacteriol. 167:35-43.

32. Smith, M. C. M., A. Mountain, and S. Baumberg. 1986. Se-quence analysis of the Bacillus subtilis argC promoter region.Gene 49:53-60.

33. Struhl, K. 1985. A rapid method for creating recombinant DNAmolecules. Biotechniques 3:452-453.

34. Zalkin, H., and D. J. Ebbole. 1988. Organization and regulationof genes encoding biosynthetic enzymes in Bacillus subtilis. J.Biol. Chem. 263:1595-1598.

35. Zuber, P., and R. Losick. 1983. Use of a lacZ fusion to studydevelopmental regulation by the spoO genes of Bacillus subtilis.Cell 35:275-283.

J. BACTERIOL.

on June 3, 2020 by guesthttp://jb.asm

.org/D

ownloaded from