Vol. 166, No. 1JOURNAL OF BACTERIOLOGY, Apr. 1986, p. 238-2430021-9193/86/040238-06$02.00/0Copyright © 1986, American Society for Microbiology

Characterization of the Developmentally Regulated Bacillus subtilisGlucose Dehydrogenase Gene

KEITH A. LAMPEL,1* BRENDA URATANI,'t G. RASUL CHAUDHRY,'t ROBERT F. RAMALEY,2 ANDSTUART RUDIKOFF3

Laboratory of Molecular Biology, National Institute of Neurological and Communicative Disorders and Stroke,' andLaboratory of Genetics, National Cancer Institute,3 Bethesda, Maryland 20892, and Department ofBiochemistry,

University of Nebraska Medical Center, Omaha, Nebraska 681052

Received 12 November 1985/Accepted 14 January 1986

The DNA sequence of the structural gene for glucose dehydrogenase (EC 1.1.1.47) of BaciUus subtilis wasdetermined and comprises 780 base pairs. The subunit molecular weight of glucose dehydrogenase as deducedfrom the nucleotide sequence is 28,196, which agrees well with the subunit molecular weight of 31,500 asdetermined from sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The sequence of the 49 aminoacids at the NH2 terminus of glucose dehydrogenase purified from sporulating B. subtilis cells matched theamino acid sequence derived from the DNA sequence. Glucose dehydrogenase was purified from an Escherichiacoli strain harboring pEF1, a plasmid that contains the B. subtilis gene encoding glucose dehydrogenase. Thisenzyme has the identical amino acid sequence at the NH2 terminus as the B. subtilis enzyme. A putativeribosome-binding site, 5'-AGGAGG-3', which is complementary to the 3' end of the 16S rRNA of B. subtilis,was found 6 base pairs preceding the translational start codon of the structural gene of glucose dehydrogenase.No known promoterlike DNA sequences that are recognized by B. subtilis RNA polymerases were presentimmediately preceding the translational start site of the glucose dehydrogenase structural gene. The glucosedehydrogenase gene was found to be under sporulation control at the trancriptional level. A transcript of 1.6kilobases hybridized to a DNA fragment within the structural gene of glucose dehydrogenase. This transcriptwas synthesized 3 h after the cessation of vegetative growth concomitant to the appearance of glucosedehydrogenase.

Bacillus subtilis is an aerobic, gram-positive bacteriumwhich, upon encountering deprivation of some essentialnutrient, undergoes morphological and biochemical changesleading to the development of an endospore (see reference 8for review). Mutations that blocked each of the five morpho-logically defined stages of spore development (spoO, spoII,spoIII, spoIV, and spoV) have been characterized (24).However, little is known about the control of the genesresponsible for this cellular differentiation.

Losick (14) and Doi (4) have shown that several sigmafactors are associated with B. subtilis core RNA polymer-ases, two of which are produced during vegetative growth(sigma 55 and sigma 37), and at least one of which is madeonly 2 h after sporulation has started (sigma 29) (15). Theyalso found that for several sporulation genes, DNA promotersequences that differ from those used by the major vegeta-tive sigma factor, sigma 55, are used. The sequential expres-sion of many sporulation genes makes it unlikely that sigmafactors of RNA polymerase solely dictate the transcriptionof all developmental loci. The study of the regulatory mech-anisms of developmental gene expression is facilitated by theinvestigation of known gene products under sporulationcontrol.Glucose dehydrogenase (,-D-glucose: NAD(P)+ 1-

oxidoreductase) is a developmentally regulated enzyme ofB.

* Corresponding author.t Present address: Crop Genetics International, Dorsey, MD

21076.t Present address: Department of Soil Science, University of

Florida, Gainesville, FL 32604.

subtilis. It catalyzes the oxidation of glucose to glucono-lactone with the concomitant reduction ofNAD or NADP toNADH or NADPH, respectively. The enzyme is synthe-sized only during sporulation and found only in the forespore(9). No enzyme activity or cross-reacting material to anti-serum against purified glucose dehydrogenase is detectedduring vegetative growth (35). The structural gene for theglucose dehydrogenase (gdh) locus is located on the Bacilluschromosome between mtlB and aroI (3), nearer to the mtlBgene at map position 9.3 (13).A 4.0-kilobase (kb) B. subtilis EcoRI DNA fragment

containing the structural gene and regulatory region ofglucose dehydrogenase was isolated from a lambda Charon4A phage library of J. Hoch (6) and was inserted into plasmidpBR322. This plasmid, designated pEF1, was transformedinto Escherichia coli cells allowing the expression of enzy-matically active glucose dehydrogenase during vegetativegrowth (35).

In this paper, we describe the nucleotide sequence of thestructural gene for glucose dehydrogenase and its locationwithin the 4.0-kb EcoRI fragment of pEF1. Hybridization ofa DNA fragment from within the glucose dehydrogenasegene with total RNAs from exponentially growing cells andat time points after the end of exponential growth showedthat one transcript is synthesized and only from sporulatingcells. The amino acid sequence of glucose dehydrogenasededuced from the DNA sequence agrees with the NH2-terminal amino acids of glucose dehydrogenase purified fromsporulating B. subtilis cells and from E. coli transformedwith the plasmid pEF1. The purification of the enzymeproduced in E. coli is described, and its properties were

238

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

B. SUBTILIS GLUCOSE DEHYDROGENASE GENE 239

TABLE 1. Purification of B. subtilis glucose dehydrogenase fromE. coli carrying plasmid pEF1

Procedure Total Ua Total Sp actb Yield(~i.mol/min) protein (mg) (%)

Cell extract 35,600 44,900 0.80 100DEAE-Sephacel 29,100 15,900 1.82 83Sephacryl S-300 26,500 6,700 3.95 74Hydroxyapatite 18,500 617 30.1 52Phenyl-Sepharose 17,100 33 520 48Chromatofocusing 15,600 27 578 44Second phenyl-Sepharose 14,000 26 539 39Concentrated enzyme 12,800 24 525 36

a A unit of activity is defined as the amount of enzyme needed to produce1 p,g of NADPH at 340 nm. Assays were done at room temperature (22 to25°C).

bSpecific activity is defined at micromoles per minute per milligram ofprotein.

compared with those of the glucose dehydrogenase purifiedfrom sporulating B. subtilis cells.

MATERIALS AND METHODS

Bacterial strains, plasmids, phages, and media. Glucosedehydrogenase was purified from sporulating cells of B.subtilis 61297 (26) and E. coli HB101 harboring the plasmidpEF1 (35), which carries the structural gene for glucosedehydrogenase. E. coli JM101 and JM103 were used topropagate phages M13mplO and M13mp8, respectively (20),and were grown in YT medium (21). Plasmid pEF60 is derivedfrom pEF1 by digesting pEF1 with the restriction endonu-clease AccI, isolating the largest fragment, and recircularizingit with T4 DNA ligase. B. subtilis 60015 (metC2 trpC2) wasgrown in nutrient sporulation medium described previously(34). E. coli strains containing plasmids were grown in LBmedium (21) supplemented with tetracycline (20 ,ug/ml) orampicillin (50 ,ug/ml).Methods. DNA sequences were determined by the

dideoxynucleotide chain termination method (28) with a15-mer primer (5'-AGTCACGACGTTGTA-3') from Be-thesda Research Laboratories, Inc., and [a-32P]dATP fromAmersham Corp. The sequencing reactions were separatedon 0.4-mm-thick 8% polyacrylamide gels. To determine theamino acid sequence of glucose dehydrogenase isolated fromB. subtilis and E. coli cells, automated Edman degradationswere performed on the proteins with a modified Beckman890C sequencer (10, 37) by using a 0.25 M Quadrol bufferprogram. With both enzyme preparations, sequencing wascarried out in the presence of Polybrene. Phenylthiohydan-toin amino acids were identified by high-pressure liquidchromatography as described previously (38).

Plasmids were isolated by the method of Birnboim andDoly (1). Standard procedures for cloning and ligation reac-tions were used (18). Restriction endonuclease digestionswere performed as described by the manufacturer. Com-puter analyses of the DNA sequences were performed byusing the NUCALN program of D. J. Lipman and W. J.Wilbur at the National Institutes of Health. Glucose dehy-drogenase activity was assayed as described previously (26).Protein concentration was determined by the method ofBradford (2). Total cellular RNA was separated on a 1%agarose-formaldehyde gel and electrophoresed in 10 mMsodium phosphate buffer, pH 7.0, with 6% formaldehyde.The DNA-RNA hybridizations were performed as describedby Silverman et al. (31). The DNA probes were nick trans-

lated with [a-32P]ATP by using a kit purchased from Be-thesda Research Laboratories, Inc.

Isolation of RNA. B. subtilis 60015 was grown in nutrientsporulation medium supplemented with potassium acetate (5mM final concentration) at 37°C with shaking. Cells (5 to 10ml) were harvested at designated times by centrifugation at7,000 x g for 5 min at 4°C. The cell pellets were suspendedin 2 ml of 50 mM Tris-hydrochloride buffer (pH 8.0)-10 mMvanadyl ribonucleoside complex (Bethesda Research Labo-ratories)-100 ,ug of lysozyme. The cell mixture was incu-bated at 37°C for 10 min, followed by the addition of 1 ml of2% sodium dodecyl sulfate solution. After the solution hadcleared, the RNA was extracted with phenol-chloroform(1:1), ethanol precipitated, dried, and stored at -70°C untilneeded.

Purification of glucose dehydrogenase. E. coli cells weregrown in 300 liters of LB medium containing 50 ,ug ofampicillin per ml at the Fermentation and Cell GrowthFacility of the National Institutes of Health. The culture wasaerated with 1 liter of air per min per liter of medium, andDow Corning silicone antifoam A was added. The cells werechilled at the end of exponential growth and harvested bycentrifugation and then stored at -70°C until used. Frozencells (500 g [wet weight]) were suspended in 1.5 liters of 5mM imidazole buffer (pH 6.5) containing 20% (wt/vol) glyc-erol, 0.1 ,ug of phenylmethylsulfonyl fluoride per ml, 10 mMEDTA, and 0.1 mM 2-mercaptoethanol. After the cells hadthawed, the mixture was adjusted to pH 6.5 with glacialacetic acid, and 0.5 g of lysozyme was added. The mixturewas placed in an ice bath and was sonicated at 5-minintervals for a total of 30 min. The cellular debris wasremoved by two centrifugations, one at 10,000 x g for 30min, then at 76,000 x g for 120 min. The supernatant wascarefully aspirated and constituted the cell-free extract indi-cated in Table 1.The procedures employed for the purification of plasmid-

encoded glucose dehydrogenase were the same as thosefrom sporulating cells of B. subtilis 61297 (26), except thatthe second columns of Sephacryl S-300 and hydroxyapatiteas well as the w-amino hexyl Sepharose were not required.The inclusion of 20% (wt/vol) glycerol in all buffers wasnecessary to retain the glucose dehydrogenase enzymaticactivity.

Materials. Restriction endonucleases and DNA-modifyingenzymes were purchased from Bethesda Research Labora-tories and International Biotechnologies, Inc. Thedideoxynucleotide sequencing reagents were obtained fromBethesda Research Laboratories. Chromatography materi-als were purchased from Pharmacia Fine Chemicals, exceptfor hydroxyapatite, which was obtained from LKB Instru-ments.

RESULTSLocation and characterization of the structural gene of

glucose dehydrogenase. A partial restriction map of the4.0-kb EcoRI B. subtilis fragment ofpEF1 is shown in Fig. 1.The restriction endonuclease PvuI cleaves the 4.0-kb DNAinto four fragments labeled A, B, C, and D. To determine thelocation of gdh, specific fragments were systematically de-leted from the 4.0-kb EcoRI fragment by restrictionendonucleases, and the resulting plasmids were assayed forthe ability to express glucose dehydrogenase when trans-formed into E. coli HB101. pEF60 was constructed bydigesting pEF1 with AccI and recircularizing the largestfragment, which still contained 3.6 kb of the original 4.0-kbB. subtilis DNA (Fig. 2). An E. coli strain containing pEF60

VOL. 166, 1986

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

240 LAMPEL ET AL.

was able to synthesize active glucose dehydrogenase (0.69,umol/min per mg of protein). Glucose dehydrogenase activ-ity was also detected after PvuI fragments B and C weredeleted from pEF60 (4.06 j±mollmin per mg of protein). Noactivity was seen after PvuI fragment D was deleted, nor wasany activity detected when the EcoRV-EcoRI fragmentwithin PvuI fragment D was removed. Glucose dehydroge-nase activity from an E. coli strain with pEFi was 1.06pLmollmin per mg of protein.

Fujita et al. (9) have demonstrated that glucose dehydro-genase is not detected until the cells are 2 to 3 h pastlogarithmic growth in nutrient sporulation medium. To de-termine whether the expression of the glucose dehydroge-nase gene was under translational or transcriptional control,RNA was isolated from cells at different stages of growthand hybridized against an HpaII fragment from within thestructural gene of glucose dehydrogenase (see Fig. 4, posi-tions 348 to 727). One RNA band was detected from t3 to t5,which is 3 to 5 h beyond exponential growth of the cell. Nohybridization was detected with RNA from vegetative cells(Fig. 3).

Purification and amino acid sequence at the NH2 terminus ofglucose dehydrogenase. Glucose dehydrogenase was purifiedto apparent homogeneity from sporulating B. subtilis cells(26) and from E. coli cells carrying the plasmid pEFI. Asummary of the purification scheme is presented in Table 1.The enzyme synthesized from the cloned B. subtilis geneappears to have properties identical to those of the enzymepurified from B. subtilis (Table 2). Both enzyme preparationsshowed cross-reactivity with antibody prepared against glu-cose dehydrogenase from B. subtilis (data not shown).The sequence of the first 49 amino acids at the NH2

terminus of glucose dehydrogenase, purified fromsporulating B. subtilis cells, was determined by automatedEdman degradations. The sequence of the first 25 aminoacids from the NH2 terminus of glucose dehydrogenase,purified from E. coli cells harboring the plasmid pEF1 whichcontains the B. subtilis gdh gene, was also determined.These amino acid sequences were identical and matchedperfectly with the amino acid sequence deduced from thenucleotide sequence of the glucose dehydrogenase structuralgene (Fig. 4).

RP An Iu

0 2 3 4

A P B P C P D

,'/r/

XbaIni aiTaqIPvuI

100bpFIG. 1. Restriction map of the 4.0-kb B. subtilis DNA in pEF1.

(a) Restriction endonucleases AccI (A), EcoRI (R), EcoRV (E), PvuI(P), and XbaI (X) recognition sites are shown. (b) PvuI digests the4.0-kb DNA into four fragments, A, B, C, and D. (c) The strategy todetermine the nucleotide sequence of the structural gene of glucosedehydrogenase is diagrammed. Arrows indicate the length anddirection of the DNA sequence.

b

c

A-R AccIA

A

M = B. subti/is DNAA = AccIR = EcoRI

A R A1.,1 kb

A A1 1.6 kb

A R A5.7 kb

J Ligase

RU(5.7kb)IAR A

FIG. 2. Contruction of pEF60. Plasmid pEF1 was digested withthe restriction endonuclease AccI. The largest fragment (5.2 kb)generated from this digest was recircularized with T4 DNA ligase.The open box represents the B. subtilis DNA that has the structuralgene of glucose dehydrogenase and its regulatory components. Tworestriction sites are shown; EcoRI (R) and AccI (A). The bla(P-lactamase) and tet (tetracyline) genes are from pBR322.

Nucleotide sequence of gdh. The 2.0-kb XbaI fragmentwithin pEF1 (Fig. 1) was isolated and either cloned directlyinto the XbaI site of M13mplO or digested with TaqI orHpaII and then cloned into the AccI site of M13mp8.Fragment D was subcloned into the PvuI-EcoRI site ofpBR322. The PstI-EcoRI fragment of this plasmid wasisolated and cloned into the PstI-EcoRI site of M13mp8. TheDNA sequence from the PstI site of pBR322 through thePvuI site of fragment D was determined. The sequencingstrategy is outlined in Fig. 1. Most of the gdh gene wassequenced in both directions, except the COOH terminus, inwhich two independent XbaI fragments cloned in the samedirection were sequenced and were identical. The completenucleotide sequence of the structural gene for glucose dehy-drogenase is shown in Fig. 4.The coding sequence of gdh was located by matching the

known NH2-terminal amino acid sequence of glucose dehy-drogenase with the translated nucleotide sequence. Theopen reading frame of 780 base pairs (bp) extends from theinitiation codon ATG at position 71 to the TAA stop codonat 850. Following the end of the gene is an inverted repeat(calculated AG value of -22.4 kcal/mol) which may functionas a terminator.

Preceding the initiation codon of the glucose dehydroge-nase gene is the sequence 5'-AGGAGG-3' (Fig. 4), whichhas strong complementarity to the Shine-Dalgarno sequence(30). No known Bacillus promoter-like sequence was locatedpreceding this ribosome-binding site.

DISCUSSIONGlucose dehydrogenase has been purified to apparent

homogeneity both from sporulating cells of B. subtilis (26)and from E. coli cells containing the pEFi plasmid carryingthe B. subtilis glucose dehydrogenase structural gene. TheB. subtilis enzyme has a native molecular weight of approx-imately 126,000 and is composed of four subunits of 31,500molecular weight. The native and subunit molecular weightsizes and enzymatic properties of glucose dehydrogenaseisolated from E. coli transformants were identical to thatpurified from sporulating cells of B. subtilis (Table 2). TheNH2-terminal amino acid sequence of both purified proteinsagreed with each other and with the amino acid sequencededuced from the nucleotide sequence (Fig. 4).

J. BACTERIOL.

PX IX .

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

B. SUBTILIS GLUCOSE DEHYDROGENASE GENE 241

V Tj T2 T3 T4T5

-1 .6Kb

FIG. 3. DNA-RNA hybridization blot. Cellular RNA was iso-lated from B. subtilis cells during vegetative growth and t, throught5, and 20 jig was separated on a 1% agarose gel, and transferred toa nitrocellulose filter. A 32P-labeled HpaII DNA fragment (106cpm/ml) from within the glucose dehydrogenase structural gene was

hybridized against the filter. RNAs that bound to the labeled DNAfragment were visualized by autoradiography.

We have determined the location of the structural gene forglucose dehydrogenase within PvuI fragment D ofpEF1 (seeFig. 1) by deleting specific fragments of the 4.0-kb EcoRI B.subtilis DNA and measuring the ability of transformed E.coli cells to synthesize glucose dehydrogenase. The nucleo-tide sequence between the PvuI and XbaI sites of PvuIfragment D was determined. The coding sequence of glucosedehydrogenase was identified by matching the nucleotidesequence with that of the known amino acid sequence at theNH2 terminus. The structural gene of glucose dehydroge-nase is 780 bp long and encodes for a 260 amino acid peptidewhich has a calculated molecular weight of 28,196. 'thiscompares well with the 31,500 molecular weight of purifiedglucose dehydrogenase estimated by sodium dodecyl sul-fate-polyacrylamide gel electrophoresis (26).The sequence at the 3' end of the 16S rRNA of B. subtilis,

3'-UCUUUCCUCCACUAG-5', is complementary to theribosome-binding (Shine-Dalgarno) site of gram-positivemRNAs (19). The nucleotide sequence 5'-AGGAGG-3' thatprecedes the initiation codon of gdh by 6 bp is complemen-tary to the 3' end of the B. subtilis 16S rRNA. The freeenergy (AG) of base pairing between the puttative ribosomebinding site of gdh and the 3' end of 16S rRNA is -16.6kcal/mol, according to the method of Tinoco et al. (33) ofestimating the free energy of double helical RNA structures.In E. coli, the consensus sequence of 5'-AGGA-3' has a AGvalue of -9.4 kcallmol, somewhat lower than most reportedfor B. subtilis ribosome-binding sites. It has been proposedby McLaughlin et al. (19) that in B. subtilis a strongerinteraction at the ribosome-binding site is needed for effi-cient translation of gram-positive mRNAs.There are no DNA sequences preceding the ribosome-

binding site ofgdh that are known to be recognized by any B.subtilis RNA polymerases. There is an open reading framepreceding the glucose dehydrogenase structural gene endingwith a termitiation codon (TAA) 19 bp before the translationstart site of gdh. This open reading frame is 855 bp long and,with the gdh gene, comprises a polycistronic transcriptionalunit which is transcribed only during sporulation (K. A.Lanmpel, B. Uratani, and R. H. Lipsky, manuscript inpreparation). The size of the combined open reading framesis nearly the same size as the transcript detected in theNorthern hybridization blot. Recently, several othersporulation specific genes were found to be a part of apolycistronic operon (5, 7, 23).

Fujita et al. (9) have shown that glucose dehydrogenase isnot detected in B. subtilis cells until nearly 3 h after the endof exponential growth. We isolated total RNA from cells atvarious stages of growth and hybridized them against a DNAfragment from within the structural gene to determinewhether the regulation of synthesis of the gene was at thelevel of transcription of translation. One band was seen onthe autoradiogram from t3 to t5 (Fig. 3) which is the sametime that glucose dehydrogenase activity was detected. Thedata strongly suggest that the regulation of glucose dehydro-genase expression is at the level of transcription, althoughthe actual control mechanism is not known. Several inves-tigators (16, 25) have suggested that the transcription ofsome genes during sporulation occurs by the association ofsporulation specific sigmta factors to RNA polymerase.The amino acid sequence of the B. subtilis glucose dehy-

drogenase is very similar to the amino acid sequence of B.megaterium glucose dehydrogenase recently reported byJany et al. (11) with the apparent retention of the same aminoacids in the active site and subunit-binding site. There is aconservation of 85% of the amino acid sequence when thetwo sequences were aligned. There is extensive amino acidsequence homology in the NH2-terminal portion of the twoenzymes (Fig. 5), as might be expected since this area of theenzymes contains the putative coenzyme binding site (11).

TABLE 2. Comparison of properties of glucose dehydrogenasepurified from E. coli with plasmid pEF1 and sporulating B. subtilis

cells

Glucose dehydrogenase from:Property

E. coli with pEFl Sporulating B. subtilisa

Native molecular weight 126,000 126,000

Subunit molecular weight 31,500 31,500

Isoelectric point pH 4.7 pH 4.7

KmbmD-glucose 12.7 ± 1.5 mM 12.5 ± 1.3 mM2-deoxyglucose 11.3 ± 1.4 mM 10.2 ± 1.2 mMD-glucosamine 83.2 ± 10 mM 85.0 ± 8.0 mM

NAD or glycerol Required Requiredprotection

Yield of enzyme 24.4 mg 2.0 mg(per 500 g [wet wt])

Spactcat 37C 1,313 920

a Data from reference 26.b Km determinations were made at 37°C.c Specific activity is defined as micromoles per minute per milligram of

protein.

VOL. 166, 1986

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

242 LAMPEL ET AL.

* * * * * * *CTGATTATTG CCGCCGCCGT ATTCTTAGGA ATCGCCAAAA CAAATTCATA ACAAATGG GAGGIATGTAT 70

rbs

* *ATG TAT CCG GAT TTA AAA GGAMET Tyr Pro Asp Leu Lys Gly

AAA GTC GTC GCT ATT ACA GGA GCT GCT TCA GGG CTC 127Lys Val Val Ala Ile Thr Gly Ala Ala Ser GIy Leu

* * * * * *

GGA AAG GCG ATG GCC ATT CGC TTC GGC AAG GAG CAG GCA AAA GTG GTT ATC AAC TATGly Lys Ala Met Ala Ile Arg Phe Gly Lys Glu Gin Ala Lys Val Val Ile Asn Tyr

184

* * *.* * *

TAT AGT AAT AAA CAA GAT CCG AAC GAG GTA AAA GAA GAG GTC ATC AAG GCG GGC GGT 241Tyr Ser Asn Lys Gln Asp Pro Asn Glu Val Lys Glu Glu Val Ile Lys Ala Gly Gly

* * * * *GAA GCT GTT GTC GTC CAA GGA GAT GTC ACG AAA GAG GAA GAT GTA AAA AAT ATC GTG 298Glu Ala Val Val Val Gin Gly Asp Val Thr Lys Glu Glu Asp Leu Lys Asn Ile Val

* * * * * *CAA ACG GCA ATT AAG GAG TTC GGC ACA CTC GAT ATT ATG ATT AAT AAT GCC GGT CTT 355Gln Thr Ala Ile Lys Glu Phe Gly Thr Leu Asp Ile Met Ile Asn Asn Ala Gly Leu

* * * *GAA AAT CCT GTG CCA TCT CAC GAA ATG CCG CTC AAGGlu Asn Pro Val Pro Ser His Glu Met Pro Leu Lys

GAT TGG GAT AAA GTC ATC GGCAsp Trp Asp Lys Val Ile Gly

* * * * *ACG AAC TTA ACG GGT GCC TTT TTA GGA AGC CGT GAA GCG ATT AAA TAT TTC GTA GAA 469Thr Asn Leu Thr Gly Ala Phe Leu Gly Ser Arg Glu Ala Ile Lys Tyr Phe Val Glu* * * * *

AAC GAT ATC AAG GGA AAT GTC ATT AAC ATG TCC AGT GTG (OAO GCG TTT CCT T CCG 526Asn Asp Ile Lys Gly Asn Val Ile Asn Met Ser Ser Val H Ala Phe Pro LrJ Pro

* * * * * *TTA TTT GTC CAC GCG GCA AGT AAA GGC GGG ATA AAG CTG ATG ACA GAA ACA TTA 583Leu Phe Val His IIyd Ala Ala Ser Lys Gly Gly Ile Lys Leu Met Thr Glu Thr Leu

* * *GCG TTG GAA TAC GCG CCG AAG GGC ATT CGCAla Leu Glu Tyr Ala Pro Lys Gly Ile Arg

* * *AAC ACG CCA ATC AAT GCT GAA AAA TTC GCTAsn Thr Pro Ile Asn Ala Glu Lys Phe Ala

GTC AAT AAT ATT GGG CCA GGT GCG ATCVal Asn Asn Ile Gly Pro Gly Ala Ile

GAC CCT / CAG AAA GCT GAT GTA GAAAsp Pro s GIn Lys Ala Asp Val Glu

* * * * * *

AGC ATG ATT CCA ATG GGA TAT ATC GGC GAA CCG GAG GAG ATC GCC GCA GTA GCA GCC 754Ser Met Ile Pro Met Gly Tyr Ile Gly Glu Pro Glu Glu Ile Ala Ala Val Ala Ala

* * * * * *

TGG CTT GCT TCG AAG GAA GCC AGC TAC GTC ACA GGC ATC ACG TTA TTC GCG GAC GGC 811Trp Leu Ala Ser Lys Glu Ala Ser Tyr Val Thr Gly Ile Thr Leu Phe Ala Asp Gly

GGT ATG ACA CAA TAT CCT TCA TTC CAG GCA GGC CGC GGT TAA ACA TAAGly Met Thr Gln r Pro Ser Phe Gln Ala Gly Arg Gly

859

* * * * * * *

A AAGOGAOCCA GAOATGAOAT OTGGATCGOT TTCTTTATTA GGCACGCTTT TTOTTTACAA GTGCTGCT 928

* * * * *

AC AGATAAAATA ATGGCAGCCG CAGACAGCAC CATCGCTGTA ATGTIA1GA 979

FIG. 4. Nucleotide sequence of the structural gene of glucose dehydrogenase and the corresponding amino acid sequence. The completenucleotide sequence of the gdh gene is shown. Below each codon triplet is the corresponding amino acid deduced from the nucleotidesequence. The ribosome binding site (rbs) preceding the translational start codon (ATG) of gdh is enclosed in a box. The arrows indicate theregion of dyad symmetry. Amino acids that are proposed as being part of the active site (0) and subunit binding area (O) of B. megateriumare indicated. This sequence was plotted through the DEC-10 computer at the National Institutes of Health using the programs Graph andPlot by Shapiro and Senapathy (M. Shapiro and P. Senapathy, Nucleic Acids Res., in press).

The amino acid sequences of the glucose dehydrogenasesfrom both Bacilli will be compared elsewhere (P. Fortnagel,K. A. Lampel, K.-D. Neitzke, and E. Freese, manuscript inpreparation).

Jornvall et al. (12) have also compared the primary aminoacid sequence ofB. megaterium glucose dehydrogenase withother alcohol-polyol dehydrogenases. They showed that theB. megaterium glucose dehydrogenase shared a 23% aminoacid sequence homology with the ribitol dehydrogenase fromKlebsiella aerogenes. We also compared the amino acidsequence of the B. subtilis glucose dehydrogenase and the K.aerogenes ribitol dehydrogenase (the NH2-terminal aminoacid sequence is shown in Fig. 5) and have determined thatthese two enzymes have 25% homology. The similarities in

amino acid sequence and folding patterns of severaldehydrogenases support the premise that dehydrogenaseshave a common origin (27).The role of glucose dehydrogenase, either in sporulation

or germination, has yet to be clearly defined (24, 26). Strauss(32) has shown that sporulation is not affected in a mutantwhich lacks glucose dehydrogenase activity. He found thatthe gdh- mutant, which germinates on alanine, is unable togerminate in media with a combination of glucose, fructose,asparagine, and KCl (GFAK), a known germinant for B.subtilis spores (36). These genetic data are consistent withbiochemical studies (26), indicating that glucose dehydroge-nase in the spore is needed for germination in glucose. Theconversion of glucose to gluconolactone, catalyzed by glu-

B. subtiIis P KGKVVAITGAASGLG A IRFlGKMEQAKVVI NKQDP§E KB. megaterium I EJKILE EIGIY]IGJ AIRFATEKIAKVV1R JK[1DEANSVLILEK. aerogenes MKHSVSS T S_ TLLGAG G KLNKL AE

FIG. 5. Comparison of the amino acid sequence at the NH2 terminus for glucose dehydrogenases from B. subtilis and B. megaterium andribitol dehydrogenase from K. aerogenes. The three amino acid sequences are aligned and regions of homology are enclosed in boxes. Thisfigure was plotted by using the program of Shapiro and Senapathy (Shapiro and Senapathy, in press).

412

640

697

J. BACTERIOL.

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from

B. SUBTILIS GLUCOSE DEHYDROGENASE GENE 243

cose dehydrogenase, generates NAD(P)H from NAD(P),which may serve as a source of reducing equivalence forgermination (29).

Several developmental genes have been cloned and se-quenced (see reference 17 for review). In most cases, thegene product has yet to be defined. Since the gene product ofgdh is known and easily assayed, the availability of thiscloned developmental gene may aid in future gtudies todetermine the regulatory mechanisms of B. subtilis sporula-tion-specific genes.

ACKNOWLEDGMENTSWe gratefully acknowledge Janet G. Pumphrey for assistance with

the amino acid sequencing and Marvin Shapiro for assistance with thecomputer-drawn graphics, and we are indebted to Enid Galliers forpreparing the illustrations.

LITERATURE CITED1. Birnboim, H. C., and J. lDoly. 1979. A rapid alkaline extraction

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

2. Bradford, M. M. 1976. A rapid and sensitive method for thequantitdtion of microgram quantities of protein utili7ing theprnnciple of protein dye binding. Anal. Biochem. 72:248-254.

3. Chaudhry, G. R., Y. S. Halpern, C. Saunders, N. Vasantha, B.Scphmidt, and E. Freese. 1984. Mapping of the glucose de4ydro-genase gene in )iacillus subtilis. J. Bacteriol. 160:607-611.

4. Doi, R. H. 1982. RNA polymerase of Bacillus subtilis, p. 72-110.In D. A. Dubnau (ed.), The molecular biology of the Bacilli.Academic Press, Inc., New York.

5. Ferrari, F. A., K. Trach, and J. A. Hoch. 1985. Sequenceanalysis of the spoOB locus reveals a polycistronic transcriptionunit. J. Bacteriol. 161:556-562.

6. Ferrari, J., D. J. Henner, and J. A. Hoch. 1981. Isolation ofBacillus subtilis genes from a Charon 4A library. J. Bacteriol.146:430432.

7. Fort, P., and J. Errington. 1985. Nucleotide sequence andcomplementation analysis of a polycistronic sporulatiqp bperon,spoVA, in Btcillus subtilis. 1. Gen. Microbiol. 131:1®1-1105.

8. Freese, E. 1981. Initiation of bacterial sporulation, p. 1-12 InH. S. Levinson, A. L. Sonenshein, and D. J. Tipper (ed.),Sporulation and germination, American Society for Microbiol-ogy, Washington, D.C.

9. Fujita, Y. R., R. Ramaleyj and E. Freese. 19i7. Location andproperties of glucose dehydrogenase in sporulating cells andspores of Bacillus subtilis. J. Bacteriol. 132:282-293.

10. HunkapilHer, M. W., and L. E. Hood. 1978. Direct microse-quence analysis of polypeptides using an improved sequenator,a nonprotein carrier (Polybrene), and high pressure liquid chro-matography. Biochemistry 17:2124-2133.

11. Jany, K.-D., W. Ulmer, M. Froschle, and G. Pfleiderer. 1984.Complete atnino acid sequence of glucose dehydrogenase fromBacillus megaterium. FEBS Lett. 165'6-10.

12. Jornvall, H.; H. von Bahr-Lindstromi K.-D. Jany, W. Ulmer,and M. Froschle. 1984. Extended superfamily of short alcohol-polyol-sugar dehydrogenases: structural similarities betweenglucose and ribitol dehydrogenases. FEBS Lett. 165:190-196.

13. Lepe&ant-Kejzlarova, J., J.-A. Lepesant, J. Walle, A. Billault,and R. Dedonder. 1975. Revision of the linkage map of Bacillussubtilis 168: indications for circularity of the chromosome. J.Bacteriol. 121:823-834.

14. Losick, R. 1981. Sigma factors, stage 0 genes, and sporulation,p. 48-56. In H. S. Levinson, A. L. Sonenshein, and D. J.Tipper (ed.), Sporulation and germination, American Societyfor Microbiology, Washington, D.C.

15. Losick, R. 1982. Sporulation genes and their regulation, p.179-202. In D. A. Dubnau (ed.), The molecular biology of theBacilli, Academic Press, Inc., New York.

16. Losick, R.,and J. Pero. 1981. Cascades of sigma factors. Cell25:582-584.

17. Losick, R., and P. Youngman. 1984. Endospore formation in

Bacillus, p. 63-88. In R. Losick and L. Shapiro (ed.), Microbialdevelopment. Cold Spring Harbor Laboratory, Cold SpringHarbor, N.Y.

18. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual, p. 383. Cold Spring HarborLaboratory, Cold Spring Harbor, N.Y.

19. McLaughlin, J. R., C. L. Murray, and J. C. Rabinowitz. 1981.Unique features in the ribosome binding site sequence of theGram-positive Staphylococcus aureus P-lactamase gene. J.Biol. Chem. 256:11283-11291.

20. Messing, J., and J. Viera. 1982. A new pair of M13 vectors forselecting either DNA strand of double digest restriction frag-ments. Gene 19:269-276.

21. Miller, J. 1972. Experiments in molecular genetics, p. 433. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

22. Piggot, P. J., and J. G. Coote. 1976. Genetic aspects of bacterialendospore formation. Bacteriol. Rev. 40:908-962.

23. Piggot, P. J., C. A. M. Curtis, and H. DeLencastre. 1984. Use ofintegrational plasmid vectors to demonstrate the polycistronicnature of a transcriptional unit (spoIIA) required for sporulationof Bacillus subtilis. J. Gen. Microbiol. 130:2123-2136.

24. Prasad, C., M. Diesterhaft, and E. Freese. 1972. Initiation ofspore germination in glycolytic mutants of Bacillus subtilis. J.Bacteriol. 110:321-328.

25. Price, C. W., and R. H. Doi. 1985. Genetic mapping of rpoDimplicates the major sigma factor of Bacillus subtilis RNApolymerase in sporulation initiation. Mol. Gen. Genet.201:88-95.

26. Ramaley, R. F., and N. Vasantha. 1983. Glycerol protection andpurification of Bacillus subtilis glucose dehydrogenase. J. Biol.Chem. 258.12558-12565.

27. Rossman, M. G., A. LiUas, C.-I. Branden, and L. J. Banaszak.1975. Evolutionary and structural relationships amongdehydrogenases, p. 61-102. In H. Boyer (ed.), The enzymes,vol. XI, 3rd ed. Academic Press, Inc., New York.

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. Setlow, B., and P. Setlow. 1977. Levels of oxidized and reducedpyridine nucleotides in dormant spores and during growth,sporulation, and spore germination of Bacillus megatarium. J.Bacteriol. 129:857-865.

30. Shine, J., and L. Dalgarno. 1974. The 3'-terminal sequence ofEscherichia coli 16S ribosomal RNA: complementarity to non-sense triplets and ribosome binding sites. Proc. Natl. Acad. Sci.USA 71:1342-1346.

31. Silverman, S. J., M. Rose, D. Botstein, and G. Fink. 1982.Regulation of HIS4-lacZ fusions in Saccharomyces cerevisiae.Mol. Cell. Biol. 2:1212-1219.

32. Strauss, N. 1983. Role of glucose dehydrogenase in germinationof Bacillus subtilis. FEMS Microbiol. Lett. 20:379-384.

33. Tinoco, I., P, N. Borer, B. Dengler, M. D. Levine, 0. C.Uhlenbeck, D. M. Crothers, and J. Gralla. 1973. Improvedestimation of secondary structure in ribonucleic acids. Nature(London) New Biol. 246:40-41.

34. Vasantha, N., and E. Freese. 1980. Enzyme changes duringBacillus subtilis sporulation caused by deprivation of guaninenucleotides. J. Bacteriol. 144:1i19-1125.

35. Vasantha, N., B. Uratani, R. Ramaley, and E. Freese. 1983.Isolation of a developmental gene of Bacillus subtilis and itsexpression in Escherichia coli. Proc. Natl. Acad. Sci. USA80:785-789.

36. Wax, R., and E. Freese. 1968. Initiation of the germination ofBacillus subtilis spores by a combination of compounds in placeof alarline. J. Bacteriol. 95:433438.

37. Wittman-Liebold, B., H. Graffunder, and H. Kohis. 1976. Adevice coupled to a modified sequenator for the automatedconversion of anilinothiazolinones into PTH amino acids. Anal.Biochem. 75:621-633.

38. Zimmerman, C. L., E. Appella, and J. J. Pisano. 1977. Rapidanalysis of amino acid phenylthiohydantoins by high-performance liquid chromatography. Anal. Biochem. 77:569-573.

VOL. 166, 1986

on February 28, 2019 by guest

http://jb.asm.org/

Dow

nloaded from