Sites Required for GltC-Dependent Regulation of Bacillus subtilis ...

JOURNAL OF BACTERIOLOGY, Oct. 1995, p. 5686–5695 0021-9193/95/$04.0010 Copyright q 1995, American Society for Microbiology

Vol. 177, No. 19

Sites Required for GltC-Dependent Regulation of Bacillus subtilis Glutamate Synthase Expression BORIS R. BELITSKY,† PAUL J. JANSSEN,‡

AND

ABRAHAM L. SONENSHEIN*

Department of Molecular Biology and Microbiology, Tufts University School of Medicine, Boston, Massachusetts 02111 Received 9 June 1995/Accepted 1 August 1995

The Bacillus subtilis gltAB genes, coding for the two subunits of glutamate synthase, are transcribed divergently from the gltC gene, encoding a LysR-type transcriptional activator of gltAB. The predicted gltA and gltC transcription start sites are separated by 51 to 52 bp. A 15-bp, consensus binding site (Box I) for LysR-type proteins was found centered at position 264 with respect to the gltA transcription start. This site was shown by mutational analysis to be required both for GltC-mediated activation of gltA and for autorepression of gltC. Box II, which is similar to Box I, is centered 22 bp downstream of Box I and overlaps the 235 region of the gltA promoter. Box II was found to be essential for activation of gltA but not for gltC autoregulation. Introduction of approximately one additional helical turn of DNA between Box I and Box II enhanced gltA expression 7- to 40-fold under nonactivating conditions and about 2-fold under activating conditions. Expression of gltA was dramatically decreased when the distance between Box I and Box II was altered by a nonintegral number of helical turns of DNA. gltC autorepression was abolished by most of the inserts between Box I and Box II but was augmented by adding one helical turn. Glutamate synthase is an important enzyme of nitrogen metabolism in bacterial cells. In Bacillus subtilis it is the enzyme responsible for biosynthesis of glutamate, an essential component of the major pathway for ammonia assimilation and a direct nitrogen donor for the biosynthesis of amino acids and other nitrogen-containing compounds (34). The reaction catalyzed by glutamate synthase, transfer of an amide group from glutamine to 2-ketoglutarate, places the enzyme at the intersection of intracellular carbon and nitrogen metabolism. In some gram-negative bacteria, expression of glutamate synthase genes is regulated positively by the leucine-responsive protein, LRP (10), or negatively by the Nac protein (3); it may also be affected by the cyclic AMP (cAMP) receptor proteincAMP complex (26). In none of these cases is the complete pattern of glutamate synthase expression understood, however. In B. subtilis, transcription of the gltA and gltB genes, encoding the large and small subunits of glutamate synthase, respectively, is subject to nitrogen-source-dependent positive regulation by the product of the gltC gene (5). GltC protein, like the Nac protein of Klebsiella aerogenes (35), belongs to the LysR family of bacterial regulators of transcription. As is the case for many members of this family (with the notable exception of Nac), GltC is encoded just upstream of and in the opposite orientation to its target genes, gltA and gltB. In addition to stimulating transcription of gltAB, GltC represses transcription of its own gene (5). Such negative autoregulation is a feature shared by many proteins of the LysR family (31). The present study was undertaken to characterize the overlapping gltA and gltC promoter regions and to identify the DNA sequences involved in regulation of these genes.

MATERIALS AND METHODS Bacterial strains and culture media. Bacterial strains used in this study are listed in Table 1. B. subtilis strains were grown at 378C in TSS minimal medium with 0.5% glucose and 0.2% nitrogen source or in DS nutrient broth medium, supplemented with the appropriate antibiotics (12). B. subtilis strains carrying gltC-gltA double transcriptional fusions at the amyE locus were isolated after strain SMY had been transformed with plasmid pLG200 or its variants, with selection for resistance to chloramphenicol and screening for loss of a-amylase production, which indicated a double-crossover recombination event. The Amy phenotype was assayed with colonies grown overnight on TBAB–0.2% starch plates (12). L medium with appropriate antibiotics (23) was used for growth of Escherichia coli strains. DNA manipulations and transformation. Methods for plasmid isolation, use of restriction and DNA modification enzymes, and DNA ligation were as described by Sambrook et al. (29). B. subtilis chromosomal DNA was isolated by a published procedure (12). Preparation of electroporation-competent E. coli cells and plasmid transformation with a Bio-Rad GenePulser apparatus (Bio-Rad Laboratories) were as described by Dower et al. (8). Transformation of B. subtilis by chromosomal or plasmid DNA was performed according to the method of Dubnau and Davidoff-Abelson (9). Hybridizations to membrane-bound DNA fragments (Southern blots) were with Nytran membranes (Schleicher & Schuell), by using the protocol obtained from the supplier. 32P-labelled DNA probes were generated by using the Ambion DECAprime DNA labelling kit (Ambion, Inc.). Plasmid construction. The plasmids used in this work are described in Table 2. E. coli JM107 and DH5aF9 were used for routine plasmid transformation and isolation; strain BU1255 (dam) was used to isolate plasmid DNA for digestion with BclI. pPJ1 was constructed by cloning the 1.15-kb HindIII-EcoRI fragment of pDEB811 (5), containing a full-length gltC gene, the gltCA promoter region, and the 59 end of gltA, in pMa5-8 (37), cleaved with HindIII and EcoRI. The same 1.15-kb fragment cleaved from pPJ1 with SmaI and XbaI is present in pIPC10 (Fig. 1) as an insert (cloned in two steps) into the XbaI site of pINT320. Plasmid pINT320 was constructed in several steps and consists of the 1.9-kb BglI (filled in)-XbaI fragment of pMa5-8 (carries plasmid and phage f1 origins of replication) and the 1.6-kb SmaI-NheI fragment carrying the repU-neo region of pBEST501 (17). The internal 36-bp EcoRI fragment of pBEST501 was deleted prior to pINT320 construction. pIPC100 is a variant of pIPC10 with a deletion of the 0.2-kb ScaI-XbaI (filled in) fragment, and as a result, pIPC100 contains a 39-truncated gltC gene. The pLG102 vector for transcriptional analysis of divergent promoters was constructed in two steps. The entire gusA (uidA) gene of E. coli (18), encoding b-glucuronidase, was first cloned in pCB192 (33), cleaved with XbaI (filled in) and HindIII, as a 2.0-kb EcoRI (filled in)-HindIII fragment derived from pRAJ294 (the XbaI and EcoRI sites were reconstituted). The SmaI-SacI fragment of the resulting plasmid, containing the gusA gene, the pCB192 polylinker, and the 59 part of the E. coli lacZ gene, was cloned in ptrpBG1 (36) opened with EcoRI (filled in) and SacI, yielding pLG102 (Fig. 2). The lacZ gene in pLG102 is preceded by the ribosomal binding site of the E. coli lpp gene and the gusA gene contains its natural ribosome binding site, but we have not verified the

* Corresponding author. Mailing address: Department of Molecular Biology and Microbiology, Tufts University School of Medicine, 136 Harrison Ave., Boston, MA 02111. Phone: (617) 636-6761. Fax: (617) 636-0337. Electronic mail address: [email protected]. † Permanent address: Institute of Genetics and Selection of Industrial Microorganisms, Moscow 113545, Russia. ‡ Present address: Laboratorium voor Microbiologie, Universiteit Gent, B-9000 Gent, Belgium. 5686

VOL. 177, 1995

REGULATION OF GLUTAMATE SYNTHASE SYNTHESIS

5687

TABLE 1. Bacterial strains used Strain

E. coli JM107

DH5aF9

BU1255 CJ236 B. subtilis SMY SMY-S LG200 LG200S

Source or reference

Genotype

endA1 gyrA96 thi hsdR17 (rk2 mk1) supE44 relA1 l2 D(lacproAB) e142/F9 traD36 proAB lacIq lacZDM15 endA1 gyrA thi-1 hsdR17 (rk2 mk1) supE44 relA1 l2 D(lacZYA-argF)U169 deoR (f80dlacZDM15) recA1/F9 dam-3 dcm-6 gal lac ara thr leu/F1 dut-1 ung-1 thi-1 relA1 spoT1/F2pCJ105 (cat) Wild type DgltC::spc DamyE::[cat F(gltC9-gusA) F(gltA9-lacZ)] DgltC::spc DamyE::[cat F(gltC9gusA) F(gltA9-lacZ)]

44

Laboratory stock

Laboratory stock 21

P. Schaeffer SMY 3 pIPC10S SMY 3 pLG200 SMY-S 3 DNA of LG200

actual translational start points for b-galactosidase and b-glucuronidase in B. subtilis. To create gltC-gltA bidirectional fusions, the 0.29-kb BclI-BamHI fragment (Fig. 3) of pIPC10 (the 281-bp BclI-NsiI fragment of the gltC-gltA intergenic region with 13 additional nucleotides containing a BamHI site originating from subcloning events at its NsiI end) was subcloned in both orientations in pLG102, cleaved with BglII. Plasmid pLG200 contains gltA-lacZ and gltC-gusA fusions. Mutant derivatives of this region were cloned in pLG102 in the same manner. pLG700 and pLG702 were isolated after cloning of the 0.18-kb BglII-BamHI and 0.17-kb BclI-BamHI fragments of pIPC10 and pIPC100 derivatives, containing, respectively, the gltAp13 and gltAp41 mutations (see below) in pLG102, digested with BglII. To create pLG706, the 0.24-kb FokI (filled in)-PstI fragment of pIPC10 was first cloned in pBS(2) that had been cleaved with BamHI (filled in) and PstI. The reconstructed BamHI site at the BamHI-FokI junction was then used to clone the regulatory region as a 0.22-kb BamHI-(FokI)-BamHI fragment in the BglII site of pLG102. All three plasmids (pLG700, -702, and -706) contain gltA-lacZ and gltC-gusA fusions. Construction of a gltC mutant. A deletion-insertion mutation within the gltC gene was created by replacing the 0.66-kb BclI-ScaI fragment of pIPC10 (Fig. 1)

FIG. 1. Structure of pIPC10. Construction of pIPC10 was as described in Materials and Methods.

with a 1.2-kb BamHI-EcoRV spc cassette, excised from pJL73 (22). The deletion covers 75% of the gltC coding region and destroys the putative helix-turn-helix motif of gltC. The orientation of the spc gene in this construction coincides with the orientation of gltC and is opposite to that of the gltA gene. The resulting plasmid was introduced into B. subtilis SMY, and Spcr Neos transformants, arising from double-crossover homologous recombination events, were selected. The replacement of the chromosomal gltC gene by the DgltC::spc allele was confirmed by Southern blot hybridization. The phenotype of the SMY-S (DgltC::spc) mutant was identical to that of the previously described strain SX150 (gltC150::Tn917) (5). RNA isolation and primer extension. Cells of B. subtilis were grown in TSSglucose-ammonia medium until late exponential phase. Pelleted cells from a 5-ml culture were resuspended in 0.1 ml of 20% sucrose–150 mM NaCl–1 mM EDTA and treated with lysozyme (0.4 mg/ml) for 20 min at 378C. One milliliter of Trizol reagent (Gibco BRL, Life Technologies) was added and RNA was extracted according to the manufacturer’s instructions. Primer extension exper-

TABLE 2. Plasmids used Name

pMa5-8 pBEST501 ptrpBG1 pCB192 pRAJ294 pJL73 pJL74 pBS(2) pDEB811 pPJ1 pIPC10 pIPC100 pIPC10S pIPC12 pIPC12H pLG102 pLG200 pLG700 pLG702 pLG706 a

Marker(s)

bla bla neo bla cat lacZ bla 9lacZ 9galK bla gusA bla spc bla spc bla bla cat gltC1 gltA9 bla gltC1 gltA9 neo gltC1 gltA9 neo gltC9 gltA9 neo DgltC::spc gltA9 neo gltC1 gltAp129 neo gltC1 gltAp129 bla cat 9lacZ 9gusA bla cat F(gltC9-gusA) bla cat F(gltC9-gusA) bla cat F(gltC9-gusA) bla cat F(gltC9-gusA)

F(gltA9-lacZ)a F(gltA9-lacZ)a F(gltA9-lacZ)a F(gltA9-lacZ)a

The structure of the insert is presented in Fig. 5.

Source or reference

37 17 36 33 K. Wilson 22 22 Stratagene, Inc. 5 This work This work This work This work This work This work This work This work This work This work This work

FIG. 2. Structure of pLG102. Construction of pLG102 was as described in Materials and Methods. The lower part of the figure shows the sequence between the initiation codon of gusA and the EcoRI site within lacZ. Likely initiation codons of gusA and lacZ reporter genes are in boldface type. Only one strand (corresponding to the nontemplate strand for rightward transcription) is shown.

5688

BELITSKY ET AL.

J. BACTERIOL.

FIG. 3. Sequence of the gltCA intergenic region and mutations in this region. Likely initiation codons and the transcription initiation sites of the gltC and gltA genes and their 210 and 235 regions are in boldface type. Directions of transcription and translation are indicated by arrows. The Box I sequence is underlined, the Box II sequence is doubly underlined, and the Box III sequence is underlined with dots. The 0.28-kb BclI-NsiI fragment was used to construct pLG200 and its derivatives as part of the 0.29-kb BclI-BamHI fragment in which the last T of the NsiI site was replaced by the 13-bp BamHI site-containing fragment which originated from pDEB811 (5) and is shown below the sequence. Mutations are indicated by vertical lines. Numbers in parentheses refer to the allele numbers of the mutations.

iments were performed as described previously (19) by using avian myeloblastosis virus reverse transcriptase (Life Sciences, Inc., St. Petersburg, Fla.) and the following two oligonucleotides as primers: oBB13 (59-CAGGACGGTAGAG ACC), corresponding to positions 102 to 87 within the coding region of gltA (Fig. 3), and oDB3 (59-GCCCTCCTCTATGGTCTAGC), corresponding to positions 218 to 199, downstream of the NsiI site (Fig. 3 and data not shown). These and other oligonucleotides used in this work were synthesized by M. Berne, Tufts University Protein and Nucleic Acid Analysis Facility. Isolation of mutants with alterations in the gltCA regulatory region. The 2-bp substitution mutation, gltAp11, was constructed by the gapped duplex DNA method (37) by using oligonucleotide 59-CATCCATCTAGATCATTTTGAG (letters in boldface type indicate mutagenic nucleotides) as the primer and pPJ1 as the template; the gltAp11 mutation creates an XbaI site. Plasmid pIPC100 was used as the template for oligonucleotide-mediated mutagenesis by the method of Kunkel et al. (21). Single-stranded uracil-containing DNA of pIPC100 was isolated from E. coli CJ236 by using helper phage R408 (28). After hybridization with a 59-phosphorylated mutagenic primer, the template was incubated with native T7 DNA polymerase (form II; New England Biolabs, Inc.), lacking strand displacement activity (45), and T4 DNA ligase. The mixtures were used for transformation of E. coli strains. Mutagenic oligonucleotides were as follows: 59-GATCAAAAGATATCTCAAAATG for the gltAp12 1-bp insertion, creating an EcoRV site (Fig. 3); 59-CAAAAGAATCTCAAAATG AGATAGATGG (during synthesis of this 28-mer the substrate stocks for the internal 19 nucleotides, in boldface type, were spiked with an equimolar mixture of the three other deoxynucleoside triphosphates at 1.8% each) for gltAp15, gltAp16, and gltAp17 (the gltAp18 mutation was identified inadvertently during site-directed mutagenesis with the same oligonucleotide, probably as a result of a T7 DNA polymerase error); 59-CAAAATGAGXTAGATGG (where X represents an equimolar mixture of dCTP, dGTP, and dTTP) for gltAp22; 59-CAATA TATAATTTXYZTCAAAAGAATC (where X 5 55% A and 15% each C, G, and T; Y 5 55% G and 15% each A, C, and T; and Z 5 55% A and 15% each C, G, and T) for gltAp41-gltAp49 (the gltAp41 mutation introduced a BclI site); 59-TTGTAGGTTGTCAAAACGA for gltAp19; and 59-CAAAACGATTTAAA CAATA for gltAp20. All mutations (Fig. 3) were confirmed by DNA sequencing of the entire gltCA regulatory region. The 0.29-kb BclI-BamHI fragment of a mutant form of pIPC100, containing the gltAp12 mutation, was used to replace an analogous fragment of pIPC10 to

create pIPC12. After digestion of pIPC12 with EcoRV and religation in the presence of the same restriction enzyme, we isolated pIPC12 derivatives which contained mutations (gltAp13 or gltAp14) within the EcoRV site (Fig. 3). The gltAp13 deletion created a BglII site. To delete the Box I sequence (see Results), the 1.15-kb HindIII-BamHI fragment of the pPJ1 derivative that contains the gltAp11 mutation was first subcloned into pBS(2), opened with HindIII and BamHI. The gltAp28 (DBox I) deletion was obtained by replacing the 0.20-kb XbaI (filled in)-BamHI fragment of the resulting plasmid with the 0.19-kb EcoRV-BamHI fragment of pIPC12. To eliminate the unique HindIII site of pIPC12, it was filled in by using E. coli DNA polymerase I large (Klenow) fragment and religated, giving pIPC12H. The EcoRV site of pIPC12H was used as a target for insertion of 8-bp SalI (59GGTCGACC, gltAp85), HindIII (59-CAAGCTTG, gltAp86) or SacI (59CGAGCTCG, gltAp83) linkers and a 10-bp EcoRI (59-CGGAATTCCG, gltAp84) linker (New England Biolabs, Inc.). Tandem insertions of SalI (gltAp81) or HindIII (gltAp82) linkers were created by using linker excess. The 5- and 13-bp insertions (gltAp90 and gltAp89) were constructed after digestion of pIPC12H derivatives containing SacI or HindIII linkers with SacI or HindIII, treatment with Klenow fragment, and religation. In a similar procedure, the spc cassettes of pJL73 or pJL74 (22) were excised with SacI and either XhoI, SalI, or EcoRV and inserted as blunt-ended fragments (treated with Klenow fragment) into the unique EcoRV site of pIPC12H. The spc cassette was then removed from the resulting plasmids by double digestion with SpeI or SacII and either HindIII or EcoRV. Religation of the cleaved and blunt-ended (Klenow fragment- or mung bean nuclease-treated) plasmids resulted in the construction of a set of plasmids with inserts of different lengths [the remains of the pSKII(2) polylinker] at the EcoRV site of pIPC12H (gltAp52-77). DNA sequencing. The mutations and insertions were sequenced by the dideoxy chain termination method of Sanger et al. (30) by using vector- or glt-specific oligonucleotides. Plasmid double-stranded DNA to be sequenced was purified by a modification of the mini-alkaline lysis protocol (29) or by using the Wizard miniprep purification kit (Promega Corp.). A Sequenase reagent kit (United States Biochemical Corp.) was used according to the protocol of the manufacturer. DNA sequences were analyzed by using the package of programs provided by the University of Wisconsin Genetics Computer Group (7). Enzyme assays. Frozen pellets of cells harvested at various times during growth in TSS medium were thawed and made permeable by vigorous mixing in

VOL. 177, 1995


5689

Z buffer (23) containing 1% (vol/vol) toluene. b-Galactosidase activity was assayed at 288C and expressed as units per absorbance unit of the culture at 600 nm as described elsewhere (23); 1 U hydrolyzes 1 mmol of o-nitrophenyl-b-D-galactopyranoside per min. The b-glucuronidase assay (42) was modified to be similar to the b-galactosidase assay. p-Nitrophenyl-b-D-glucuronide (Sigma Chemical Company) at 6 mM was used as a substrate, the incubation temperature was 378C, and the accumulation of product was monitored by measuring A405. b-Glucuronidase specific activity was determined in a manner analogous to the calculation of Miller units for b-galactosidase. Nucleotide sequence accession number. The corrected nucleotide sequence of the gltC gene has been assigned GenBank accession number M28509.

RESULTS Identification of the gltA transcription start. The sequence of the 281-bp BclI-NsiI fragment of the B. subtilis chromosome containing the gltCA regulatory region is presented in Fig. 3. This sequence reflects the correction of several errors in the previously reported sequence of this region (5): C at position 297 [C(297)] now replaces G; two nucleotides, C between G(296) and T(295) and G between A(110) and C(111), are deleted in the present sequence; and G(11) is added to the sequence. The corrected sequence alters the first several codons of the gltC open reading frame (the first 9 amino acids of the previously reported GltC sequence [5] are replaced now by Met, Asp, and Val) and predicts a GltC polypeptide of 300 amino acids (calculated molecular mass, 34.0 kDa). As previously noted, GltC protein is homologous to almost all of the presently identified members of the LysR family of bacterial regulatory proteins (16, 32), but only two family members are more than 30% identical to GltC: orf278 (25) (35.0%) and Ipa24d (14) (31.4%) (data not shown). Both are from B. subtilis, but their functions are unknown. The putative helix-turnhelix region characteristic of LysR-type proteins is found at positions 18 to 37 of GltC. The 59 ends of gltA and gltC mRNAs were predicted previously by low-resolution S1 nuclease mapping (4) to be at position A(11) for gltA and positions A(253) and/or A(252) for gltC (Fig. 3). Using the primer extension method and two different oligonucleotide primers, we confirmed the apparent transcription start site of the gltA gene (Fig. 4). We could not detect any RNA product of the poorly expressed gltC gene (4) by this approach, but the location of the gltC promoter was confirmed by analyzing mutations in its predicted 235 and 210 regions (see below). The distance between the two transcriptional starts is 51 to 52 bp, and the 210 and 235 regions of the two promoters overlap extensively. The initiation codon of the gltA gene was tentatively identified as the ATG codon starting at position 57 (unpublished results), leaving 146 bp between the coding regions of gltA and gltC (Fig. 3). Use of a double promoter probe vector. To analyze the effects of mutations in the gltCA intergenic region on regulation of the gltA and gltC genes, we constructed a double promoter probe vector, pLG102. pLG102 has two promoterless reporter genes, lacZ and gusA, from E. coli, oriented in opposite directions and separated by a polylinker region (Fig. 2). The usefulness of the E. coli gusA gene as a reporter gene in B. subtilis has also been shown by Karow et al. (20). pLG102 replicates autonomously in E. coli but not in B. subtilis. After transformation of B. subtilis cells, the plasmid can be maintained by integration at the chromosomal amyE locus. No activity of b-galactosidase or b-glucuronidase above endogenous levels was detected in B. subtilis cells transformed with pLG102. In TSS minimal medium, these levels were 0.15 to 0.50 U for b-galactosidase (depending on the composition of the medium) and #0.1 U for b-glucuronidase.

FIG. 4. Determination of gltA transcription start site. Primer oDB3 was extended with avian myeloblastosis virus reverse transcriptase by using 20 mg of RNA from B. subtilis SMY grown in glucose-ammonia minimal medium or from Saccharomyces cerevisiae (Sigma Chemical Company) as the template. The sequence of the template strand of a pBB559 derivative containing the 5.2-kb ClaI-PstI fragment from the gltCA region (2, 4) obtained by using oDB3 as a primer and Sequenase DNA polymerase is shown to the left. The apparent transcription start site of gltA is shown in outlined type. The direction of gltA transcription is shown by the arrow. Primer oBB13 gave the same apparent mRNA 59 end (data not shown).

The entire gltCA regulatory region, as presented in Fig. 3 (positions 2160 to 1119 relative to the gltA transcription start point), was inserted in both orientations as a BclI-BamHI fragment at the BglII site of pLG102 (Fig. 2). The resulting plasmid, pLG200, was introduced into B. subtilis cells by transformation, and the region containing the cat gene and the two transcriptional fusions was integrated in single copy at the amyE locus of the chromosome by a double-crossover recombination event. The use of pLG200 and its derivatives allowed us to monitor expression of gltC and gltA simultaneously in the same strains. Expression of a gltA-lacZ fusion in B. subtilis cells, growing exponentially in minimal glucose medium with different nitrogen sources, is presented in Table 3. Expression was highly dependent on an intact gltC gene and was regulated by the nitrogen source in the same manner as that described for glutamate synthase activity and gltA mRNA synthesis (4, 5). The activity of the fusion was very low when proline or glutamate served as the only nitrogen source in the medium and increased .100-fold when ammonium ions were the only nitrogen source. (Proline presumably is transported more rapidly than glutamate and can be readily converted to the latter intracellularly.) A gltA-gusA fusion had similar (slightly higher) activity and was similarly regulated (data not shown). The expression level of a gltC-gusA fusion was low in wildtype cells, but it was elevated more than 10-fold in a gltC

5690

BELITSKY ET AL.

J. BACTERIOL. TABLE 3. Activities of gltA and gltC fusions in glucose minimal medium

Strain

b-Galactosidase activitya (gltA-lacZ fusion)

Genotype NH4Cl 1

LG200 glt 75.8 6 10.2 (8) LG200S DgltC::spc NAc

b

b-Glucuronidase activitya (gltC-gusA fusion)

NH4Cl 1 Pro

Glutamate

Pro

NH4Cl

NH4Cl 1 Pro

Pro

49.6 6 3.9 (2) 0.9 6 0.1 (2)

2.4 6 0.3 (5) 0.5 6 0.1 (2)

,0.6 (3) #0.3 (2)

0.7 6 0.1 (13) NAc

0.5 6 0.1 (2) 7.1 6 0.6 (4)

0.8 6 0.3 (2) 8.8 6 0.4 (2)

a Cells were grown in TSS-glucose minimal medium with different nitrogen sources (12). Enzyme activity was assayed and expressed in units as described in Materials and Methods. b Numbers in parentheses indicate the number of experiments for each strain and growth condition. c NA, not applicable (gltC null mutants are unable to grow unless provided with a source of glutamate).

mutant (Table 3), reflecting relief from negative autoregulation (5). Deletion analysis of the gltCA regulatory region. Three truncated versions of the gltCA regulatory region, representing deletions from position 2160 to positions 294, 257, and 246, respectively (Fig. 5), were also cloned in pLG102. The resulting fusions were integrated at the B. subtilis amyE locus. Deletion upstream of position 293 had no effect on the expression and regulation of either gltA-lacZ or gltC-gusA fusions (Fig. 5 and data not shown). Deletion to position 256 or 245 reduced gltA-lacZ expression 50-fold (Fig. 5), implicating the sequence(s) between positions 293 and 255 in positive regulation of the gltA promoter. The last two deletions remove nucleotides downstream from positions 15 and 27 with respect to the predicted gltC transcription start point (Fig. 3). Deletion downstream from position 15 (from pLG700) fully derepressed the gltC-gusA fusion (Fig. 5) and made gltC expression insensitive to GltC, i.e., the fusion had the same level of expression in wild-type and gltC

mutant cells (data not shown). We conclude that the same sequence (present in pLG706 but missing in pLG700) is necessary for positive regulation of gltA transcription and negative regulation of gltC. Deletion downstream from position 27 (from pLG702) completely abolished gltC-gusA expression in B. subtilis, even in a gltC mutant (Fig. 5 and data not shown), though the deletion leaves the 210 and 235 regions of the gltC promoter intact. This inactivation may be due to the lack of suitably located purines near the potential transcription start within a newly created gltC-gusA junction. Mutations in Box I. A 15-bp dyad symmetry element, designated Box I, can be found upstream of the gltA promoter between positions 271 and 257 (Fig. 3). The two 6-bp arms of this sequence form a perfect dyad and are interrupted by 3 bp. We introduced one double mutation, replacement of T and C at positions 270 and 269 by G and A [TC(270, 69)GA] (gltAp11), and five single mutations, T(270)C (gltAp22), DT(264) (gltAp15), T(262)A (gltAp17), G(261)T (gltAp16), and T(257)A (gltAp14) into Box I (Fig. 3) and integrated

FIG. 5. Effects of deletions in the gltCA regulatory region on gltA-lacZ and gltC-gusA expression. The BglII site in the upper diagram represents the point of insertion of gltCA regulatory fragments into the double reporter gene region of pLG102 (Fig. 1). The boundaries of the gltCA fragments are given with respect to the gltA transcription initiation site (Fig. 3). The shaded boxes between the gltC and gltA coding regions represent Boxes I and II. The restriction sites in parentheses were constructed during this work. The activities (in units, as described in Materials and Methods) of gltA-lacZ and gltC-gusA fusions in cells grown in glucose-ammonia minimal medium are indicated.

VOL. 177, 1995


5691

FIG. 6. Effects of mutations in the gltCA regulatory region on gltA-lacZ (A) and gltC-gusA (B) expression and comparison of effects of Box I mutations on gltA-lacZ and gltC-gusA fusions (C). Strains LG200 and LG200-S (B) or their derivatives containing mutant fusions were grown in glucose-ammonia minimal medium or in glucose-ammonia-proline medium, respectively. The dark stippled bars in panel B refer to LG200 derivatives (gltC1), and the light stippled bars refer to LG200-S derivatives (gltC::spc); for the A(217)C mutation, the activity of b-glucuronidase in the gltC mutant was 59.4 U.

divergent gltA and gltC fusions carrying these mutations at the amyE locus. These mutations, which span the entire sequence of Box I, decreased expression of the gltA-lacZ fusions by 1.5to 40-fold (Fig. 6A). The residual activity was, in each case, still regulated by the nitrogen source and dependent on GltC (data not shown). The same mutations relieved repression of the gltC promoter, increasing b-glucuronidase activity 2.1- to 9.2-fold; complete derepression of all mutant gltC-gusA fusions was seen to occur in a gltC mutant (Fig. 6B). The defect in gltA activation for each mutation roughly correlated with the defect in gltC autorepression, that is, with the level of gltC-gusA derepression (Fig. 6C). Replacement of nearly the entire 15-bp Box I sequence with only two nucleotides, GA (gltAp28 5 DBox I; Fig. 3), completely inactivated gltA expression and significantly derepressed the gltC gene. The activity of the mutant gltC-gusA fusion was not further increased in a gltC mutant (Fig. 6A and B). We conclude that Box I is required both for activation of the gltA promoter and for repression of the gltC promoter. These requirements for Box I are consistent with deletion analysis of the gltC-gltA regulatory region (Fig. 5). Mutations in Box II. Two additional 15-bp sequences with similarity to Box I can be found upstream of the gltA transcription start: Box II lies between positions 249 and 235 and is centered 22 bp downstream of Box I; Box III lies between positions 229 and 215 and is centered 42 bp downstream of Box I (Fig. 3). Five mutations, T(248)A (gltAp43), T(248)C (gltAp49), CT(247, 46)GC (gltAp47), T(246)A (gltAp41), and T(246)G (gltAp48), were introduced into the left arm of Box II (Fig. 3). The mutations reduced the activity of a gltA-lacZ fusion by 1.2to 50-fold (Fig. 6A). The residual gltA expression of Box II mutant strains was dependent on the nitrogen source and GltC (data not shown). The mutations in Box II did not relieve repression of the gltC promoter, however; in fact, two of them reduced gltC-gusA expression 2.5-fold, but all mutant fusions, with one exception, were derepressed in a gltC mutant to about the same extent as the wild-type fusion was. The T(246)G mutation completely inactivated the gltC promoter: the fusion was not expressed even in a gltC mutant (Fig. 6B). The deleterious effects of some Box II mutations on gltC-gusA activity

presumably reflect their direct interference with gltC promoter function because of the proximity of these mutations to the putative gltC 210 region and transcription start site (Fig. 3). Thus, Box II, like Box I, is necessary for full activation of gltA transcription but, in contrast to Box I, is not required for autorepression of the gltC gene. Mutations in Box III and the 235 region of the gltC promoter. A T(228)A mutation (gltAp20) was introduced into the Box III sequence (Fig. 3). Mutations at the analogous positions in Box I and Box II reduced gltA expression 25- to 50-fold (Fig. 3 and 6A). The T(228)A mutation, by contrast, increased activity of the gltA-lacZ fusion by 80%, probably by improving the gltA promoter (Fig. 6A). Little effect on gltC transcription was detected for this mutation (Fig. 6B). Two other mutations in Box III, A(215)C (gltAp18) and A(217)C (gltAp19), had no effect on activity of the gltA-lacZ fusion (Fig. 6A). Both of these mutations alter the 235 region of the gltC promoter. The A(215)C mutation, which is an alteration away from the consensus sequence for sA-dependent promoters (24), caused complete inactivation of gltC expression (Fig. 6B and data not shown). A(217)C, on the other hand, converted the gltC 235 region from TTTTCA to TTGTCA, a change towards consensus (24). As a result, the activity of the modified gltC promoter increased 20-fold and was still sensitive to autorepression (Fig. 6B and data not shown). We conclude that Box III is not involved in regulation of either gltA or gltC but is an integral part of the overlapping gltA and gltC promoters. Effects of spacing mutations on gltA expression. To test whether the distance between Boxes I and II is important for regulation of gltA, we introduced insertion and deletion mutations into the gltCA regulatory region (Fig. 7). Deletion of a single T nucleotide at position 257 (gltAp13) almost completely abolished gltA-lacZ expression. An even more pronounced effect was seen for a 1-bp A insertion (gltAp12) between positions 257 and 256, that is, immediately downstream of Box I, as well as for 3- to 5-bp insertions at the same position. On the other hand, 9- to 11-bp insertions (gltAp83-86), which separate Box I from Box II by about one turn of the DNA helix, either were without effect or increased gltA expression about twofold under activating conditions (i.e., with ammonia as the sole nitrogen source). The level of gltA-lacZ expression under non-

5692

BELITSKY ET AL.

J. BACTERIOL.

FIG. 7. Effect of alteration of distance between Box I and Box II on gltA-lacZ and gltC-gusA expression. Strain LG200 and its derivatives containing mutant fusions were grown in glucose-ammonia minimal medium. Enzyme activities are expressed in units as described in Materials and Methods. The sequence of the gltCA regulatory region is marked as in Fig. 6.

activating conditions (i.e., with proline as the sole nitrogen source) was elevated 7- to 40-fold (Fig. 7 and Table 4). An insertion, with a length close to 2 helical turns of DNA (gltAp62), allowed some expression of the gltA-lacZ fusion to occur (Fig. 7 and Table 4). Increasing the insertion length to 25 to 82 bp led to loss of gltA expression (Fig. 7). To show that the effects of spacing mutations on gltA-lacZ expression (at the amyE locus) correctly reflect regulation of the gltA gene, we introduced some of these mutations at the gltCAB locus. The presence upstream of the gltA gene of in-

sertion mutations which abolish gltA-lacZ expression (e.g., gltAp71, gltAp76, and gltAp77) conferred a glutamate requirement for growth. Box I or Box II point mutations at the gltCAB locus allowed growth without glutamate, though at reduced rates for some mutations (data not shown). Effects of spacing mutations on gltC expression. The Box I sequence was shown above to be the only dyad element of the gltCA regulatory region necessary for gltC autorepression. Insertion of 1 bp between positions 257 and 256 (equivalent to positions 16 and 15 relative to the predicted gltC transcription

TABLE 4. Effect of insertions between Box I and Box II on gltA-lacZ and gltC-gusA expression gltA-lacZ (b-galactosidase) activityb

gltC-gusA (b-glucuronidase) activityb

gltAp allele (size of insert [bp]a within a fusion)

NH4Cl

NH4Cl 1 Pro

Pro

NH4Cl/Pro ratio

NH4Cl 1 Pro

Pro

NH4Cl

NH4Cl 1 Pro

NH4Cl 1 Pro

gltAp1 gltAp83 gltAp85 gltAp86 gltAp84 gltAp62

75.8c 146 159 182 68.5 5.5

49.6 78.8 96.6 102 40.7 3.7

0.6 21.4 20.0 21.5 4.0 0.4

$120 6.8 8.0 8.5 17.2 $13

0.8 0.8 1.0 0.6 0.8 0.6

0.3 NDd NDd NDd NDd NDd

0.7 0.1 0.2 0.1 2.8 4.8

0.5 0.1 0.1 0.1 1.8 4.5

7.1 4.6 7.5 6.8 6.5 5.5

a

(9) (9) (9) (11) (21)

gltC1

gltC1

gltC::spc

gltC::spc

The sequences of mutant alleles are shown in Fig. 7. Strain LG200 or its derivatives with mutant fusions were grown in TSS-glucose minimal medium with different nitrogen sources (12). Enzyme activities were assayed and expressed in units as described in Materials and Methods. c All the numbers are averages of at least two experiments, and the mean errors did not exceed 15%. d ND, not determined. b

VOL. 177, 1995


start site) had no effect on gltC repression (Fig. 7). Insertions of 3 to 5 bp or much longer sequences (ranging from 11 to 82 bp) at the same position between Box I and Box II partially or completely relieved autorepression of the gltC-gusA fusion (Fig. 7). Insertions of 9 bp, which increased the activity of the gltA promoter, diminished the expression of the gltC promoter at least 3.5-fold (Fig. 7). This reduction was due to more effective repression rather than to a general defect in gltC expression, since these gltC fusions were fully derepressed in a gltC mutant (Table 4). DISCUSSION Regulatory sites within the gltCA intergenic region. The GltC protein is a member of the LysR family (16, 31) of DNA-binding regulatory proteins. These proteins usually activate transcription of their target genes, which are frequently linked to and transcribed divergently from the regulatory genes. Box I (Fig. 3) has all of the features of the consensus sequence attributed to binding sites for LysR-type proteins (15, 31): (i) the central nucleotide of Box I is located at position 264 with respect to the gltA transcription start; (ii) the Box I sequence contains the highly conserved motif T-N11-A; (iii) Box I has two 6-bp arms of perfect dyad symmetry, interrupted by 3 nucleotides; and (iv) the central part of Box I is AT rich. Deletion and mutational analysis of Box I clearly demonstrates that this sequence is an essential component of GltCmediated activation of gltA (Fig. 4 and 6). The integrity of Box I is also required for autorepression of the gltC gene. Therefore, this sequence behaves as predicted for a binding site of a protein which simultaneously activates gltA and represses gltC. We assume that this protein is GltC. In fact, mutations in gltC that compensate specifically for certain defects in Box I and Box II provide strong genetic evidence for such interaction (2), even though direct biochemical proof of specific DNA binding by GltC is lacking. By mutational analysis similar to that described in this paper, the involvement of Box I-like sequences in activation of target genes has been demonstrated for certain other LysR-type proteins, such as NahR (32), MetR (6, 39, 43), NodD (11, 15), and CatR (27). Several LysR-type proteins have been shown to bind to DNA regions that overlap the LysR consensus site (31), and this binding is decreased by the same mutations which impair activation of target genes (1, 15, 27, 32). We have shown here that a second dyad symmetry region (Box II [Fig. 3]), which is separated from Box I by two DNA helical turns and partially overlaps the 235 region of the promoter of the target gene, is required for activation of gltA transcription (Fig. 6A). The sequence of Box II is similar to that of Box I (10 matches of 15) and also conforms to the consensus LysR site. The existence of a second DNA site necessary for gene activation mediated by LysR-type proteins and located between the Box I-like site and the promoter of a target gene was described as a general feature of this mode of gene regulation (31). Binding to these sites or extended binding sites stretching from Box I-like sequences towards the 235 regions of target genes was demonstrated by footprinting experiments for several LysR-type proteins (31). The functional role of a Box II-like sequence, called the activation site, has been demonstrated genetically for genes regulated by TrpI (13), CatR (27), and MetR (43). The extent of similarity between the two binding sites varies for different LysR-type proteins. The relative spatial orientation of Box I and Box II is crucial for activation of the gltA gene. Any insertion or deletion that

5693

changes the spacing between the two boxes to a nonintegral number of helical turns strongly reduces nitrogen-dependent activation of gltA. On the other hand, insertions of 9 to 11 bp, which make the distance between the centers of Box I and Box II equal to 31 to 33 bp or about 3 helical turns, have no effect on or even elevate expression of the gltA gene. These results suggest that Box I and Box II do not constitute a single site but rather are distinct binding sites for regulatory proteins and that, to achieve activation of gltA, proteins bound at the two sites must interact with each other, a task that is best accomplished when the proteins lie on the same side of the DNA helix. By an analogous approach, the same conclusion was reached for the two binding sites of the LysR-type NodD protein upstream of the Rhizobium meliloti nodH gene (11); similar data were presented for the NodD1 protein of Bradyrhizobium meliloti (41). On the basis of the sequence similarities of Box I and Box II, it is likely that the same protein, presumably GltC (2), binds to both sites. GltC protein as an autorepressor. The location of Box I just downstream of the predicted transcription start site of the gltC gene offers an obvious explanation for GltC-mediated autorepression; binding of GltC presumably interferes with transcription initiation. In fact, Box I is necessary for gltC repression (Fig. 6B) and the levels of repression under different conditions most likely reflect the strength of GltC binding to Box I. The involvement of a Box I-like sequence in autorepression of a gene encoding a LysR-type protein also has been shown directly for metR (39). Since expression of gltC is not significantly regulated by the nitrogen source in the growth medium, we suggest that GltC binding to Box I is independent of the nitrogen-associated signal. Effector-independent binding to Box I-like sequences, called repression sites, has been proposed for many LysR-type proteins (31). When the relative rotational orientation of Box I and the RNA polymerase binding site is altered and the distance between them is increased by 3 to 5 bp, autorepression is significantly relieved, even though Box I is still within the DNA sequence assumed to be covered by RNA polymerase. The ability of GltC to repress gltC is restored or even increased if Box I is moved 9 to 11 bp downstream from the gltC promoter, which still allows direct interaction between GltC protein, bound to Box I, and RNA polymerase, seeking to initiate transcription from the gltC promoter. Since repression of gltC was largely abolished when Box I was moved more than 11 bp further downstream from the gltC promoter, GltC protein is probably unable to function as a roadblock for chain-elongating RNA polymerase. We infer that to repress its own transcription, GltC must have a specific rotational orientation relative to RNA polymerase and be positioned in close proximity to the gltC promoter. Lack of gltC repression as a result of the postulated interaction of GltC with Box II, overlapping the gltC 210 region, apparently is due to the absence of any steric hindrance for RNA polymerase function. Some LysR-type proteins in their nonactivating conformation can repress their target genes, because the level of expression of these genes under nonactivating conditions is higher in the absence of the regulatory proteins (31, 40). No repressing properties with respect to gltA expression can be observed for GltC, because the expression level of the gltA gene is lower in gltC mutants than it is in gltC1 cells under all growth conditions tested (Table 3 and data not shown). Model of GltC interaction with the gltCA regulatory region. On the basis of these results, we offer the following model for regulation of gltA and gltC. The GltC protein binds to Box I under all growth conditions, causing transcription of gltC to be repressed about 10-fold. When the intracellular environment

5694

BELITSKY ET AL.

signals glutamate limitation, GltC also binds to Box II, and this binding is facilitated by interactions with GltC molecules bound to Box I. Only GltC molecules bound to Box II are likely to interact with RNA polymerase attempting to initiate gltA transcription and thereby to stimulate transcription of gltA. Since both boxes have dyad symmetry, we imagine that each binds a dimer of GltC and that positive regulation of gltA depends on dimer-dimer interaction. If this model is correct, nitrogen source-dependent regulation could reflect the effects of metabolites on dimer-dimer interaction. In fusion constructs integrated at the amyE locus, the addition of 1 helical turn between Box I and Box II decreases gltC-gusA expression, increases expression of gltA-lacZ, and makes gltA expression less responsive to regulation by the nitrogen source. (A similar effect of such insertions was recently described for expression of NodD- and OxyR-dependent gene fusions [11, 38].) All of these effects can be explained by stronger GltC protein-protein interaction. When the same insertion mutations are placed at the gltCAB locus, the potential increase in gltA expression could be mitigated by stronger repression of the gltC gene. It seems, therefore, that the wildtype arrangement of Box I and Box II is suboptimal in terms of GltC interaction and gltA expression but might be viewed as a compromise that has evolved to allow enough GltC synthesis to permit adequate transcription of gltA in a manner that discriminates well between activating and nonactivating conditions. ACKNOWLEDGMENTS We are grateful to S. Brown, P. Serror, and A. Wright for careful reading of the manuscript; to S. Brown, L. Grinius, S. Jin, S. Melville, J. Mueller, P. Serror, F. Slack, and S. Wellington for many helpful discussions; and to K. Wilson and J. LeDeaux for gifts of plasmids. This work was supported by Public Health Service grant GM36718.

J. BACTERIOL.

15.

16. 17. 18. 19. 20.

21. 22. 23. 24.

25. 26.

27.

REFERENCES 1. Bartowsky, E., and S. Normark. 1993. Interaction of wild-type and mutant AmpR of Citrobacter freundii with target DNA. Mol. Microbiol. 10:555–565. 2. Belitsky, B. R., and A. L. Sonenshein. 1995. Mutations in GltC that increase Bacillus subtilis gltA expression. J. Bacteriol. 177:5696–5700. 3. Bender, R. A., P. M. Snyder, R. Bueno, M. Quinto, and B. Magasanik. 1983. Nitrogen regulation system of Klebsiella aerogenes: the nac gene. J. Bacteriol. 156:444–446. 4. Bohannon, D. E., M. S. Rosenkrantz, and A. L. Sonenshein. 1985. Regulation of Bacillus subtilis glutamate synthase genes by the nitrogen source. J. Bacteriol. 163:957–964. 5. Bohannon, D. E., and A. L. Sonenshein. 1989. Positive regulation of glutamate biosynthesis in Bacillus subtilis. J. Bacteriol. 171:4718–4727. 6. Byerly, K. A., M. L. Urbanowski, and G. V. Stauffer. 1991. The MetR binding site in the Salmonella typhimurium metH gene: DNA sequence constraints on activation. J. Bacteriol. 173:3547–3553. 7. Devereux, J., P. Haeberli, and O. Smithies. 1984. A comprehensive set of sequence analysis programs for the VAX. Nucleic Acids Res. 12:387–395. 8. Dower, W. J., J. F. Miller, and C. W. Ragsdale. 1988. High efficiency transformation of E. coli by high voltage electroporation. Nucleic Acids Res. 16:6127–6145. 9. Dubnau, D., and R. Davidoff-Abelson. 1971. Fate of transforming DNA following uptake by competent Bacillus subtilis. I. Formation and properties of the donor-recipient complex. J. Mol. Biol. 56:209–221. 10. Ernsting, B. R., J. W. Denninger, R. M. Blumenthal, and R. G. Matthews. 1993. Regulation of the gltBDF operon of Escherichia coli: how is a leucineinsensitive operon regulated by the leucine-responsive regulatory protein? J. Bacteriol. 175:7160–7169. 11. Fisher, R. F., and S. R. Long. 1993. Interactions of NodD at the nod box: NodD binds to two distinct sites on the same face of the helix and induces a bend in the DNA. J. Mol. Biol. 233:336–348. 12. Fouet, A., and A. L. Sonenshein. 1990. A target for carbon source-dependent negative regulation of the citB promoter of Bacillus subtilis. J. Bacteriol. 172:835–844. 13. Gao, J., and G. N. Gussin. 1991. Mutations in TrpI binding site II that differentially affect activation of the trpBA promoter of Pseudomonas aeruginosa. EMBO J. 13:4137–4144. 14. Glaser, P., F. Kunst, M. Arnaud, M.-P. Coudart, W. Gonzales, M.-F. Hullo, M. Ionescu, B. Lubochinsky, L. Marcelino, I. Moszer, E. Presecan, M.

28. 29.

30. 31. 32.

33. 34.

35. 36.

37.

38.

39.

40.

Santana, E. Schneider, J. Schwezer, A. Vertes, G. Rapoport, and A. Danchin. 1993. Bacillus subtilis genome project: cloning and sequencing of the 97 kb region from 3258 to 3338. Mol. Microbiol. 10:371–384. Goethals, K., M. Van Montagu, and M. Holsters. 1992. Conserved motifs in a divergent nod box of Azorhizobium caulinodans ORS571 reveal a common structure in promoters regulated by LysR-type proteins. Proc. Natl. Acad. Sci. USA 89:1646–1650. Henikoff, S., G. W. Haughn, J. M. Calvo, and J. C. Wallace. 1988. A large family of bacterial activator proteins. Proc. Natl. Acad. Sci. USA 85:6602– 6606. Itaya, M., K. Kondo, and T. Tanaka. 1989. A neomycin resistance cassette selectable in a single copy state in the Bacillus subtilis chromosome. Nucleic Acids Res. 17:4410. Jefferson, R. A., S. M. Burgess, and D. Hirsh. 1986. b-Glucuronidase from Escherichia coli as a gene fusion marker. Proc. Natl. Acad. Sci. USA 83: 8447–8451. Jin, S., and A. L. Sonenshein. 1994. Transcriptional regulation of Bacillus subtilis citrate synthase genes. J. Bacteriol. 176:4680–4690. Karow, M. L., P. Glaser, and P. J. Piggot. 1995. Identification of a gene, spoIIR, that links the activation of sE to the transcriptional activity of sF during sporulation in Bacillus subtilis. Proc. Natl. Acad. Sci. USA 92:2012– 2016. Kunkel, T. A., J. D. Roberts, and R. A. Zakour. 1987. Rapid and efficient site-specific mutagenesis without phenotypic selection. Methods Enzymol. 154:367–382. LeDeaux, J. R., and A. D. Grossman. 1995. Isolation and characterization of kinC, a gene that encodes a sensor kinase homologous to the sporulation sensor kinases KinA and KinB in Bacillus subtilis. J. Bacteriol. 177:166–175. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Moran, C. P., Jr. 1993. RNA polymerase and transcription factors, p. 653– 667. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and other gram-positive bacteria: biochemistry, physiology, and molecular genetics. American Society for Microbiology, Washington, D.C. Ogasawara, N., S. Nakai, and H. Yoshikawa. 1994. Systematic sequencing of the 180 kilobase region of the Bacillus subtilis chromosome containing the replication origin. DNA Res. 1:1–14. Oliver, G., G. Gosset, R. Sanchez-Pescador, E. Lozoya, L. M. Ku, N. Flores, B. Becerril, F. Valle, and F. Bolivar. 1987. Determination of the nucleotide sequence for the glutamate synthase structural genes of Escherichia coli K-12. Gene 60:1–11. Parsek, M. R., R. W. Ye, P. Pun, and A. M. Chakrabarty. 1994. Critical nucleotides in the interaction of a LysR-type regulator with its target promoter region. catBC promoter activation by CatR. J. Biol. Chem. 269:11279– 11284. Russel, M., S. Kidd, and M. R. Kelly. 1986. An improved filamentous helper phage for generating single-stranded plasmid DNA. Gene 45:333–338. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74:5463–5467. Schell, M. A. 1993. Molecular biology of the LysR family of transcriptional regulators. Annu. Rev. Microbiol. 47:597–626. Schell, M. A., and E. F. Poser. 1989. Demonstration, characterization, and mutational analysis of NahR protein binding to nah and sal promoters. J. Bacteriol. 171:837–846. Schneider, K., and C. F. Beck. 1986. Promoter-probe vectors for the analysis of divergently arranged promoters. Gene 42:37–48. Schreier, H. J. 1993. Biosynthesis of glutamine and glutamate and assimilation of ammonia, p. 281–298. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and other gram-positive bacteria: biochemistry, physiology, and molecular genetics. American Society for Microbiology, Washington, D.C. Schwacha, A., and R. A. Bender. 1993. The nac (nitrogen assimilation control) gene from Klebsiella aerogenes. J. Bacteriol. 175:2107–2115. Shimotsu, H., and D. J. Henner. 1986. Construction of a single-copy integration vector and its use in analysis of regulation of the trp operon of Bacillus subtilis. Gene 43:85–94. Stanssens, P., C. Opsomer, Y. M. McKeown, W. Kramer, M. Zabeau, and H.-J. Fritz. 1989. Efficient oligonucleotide-directed construction of mutations in expression vectors by the gapped duplex DNA method using alternating selectable markers. Nucleic Acids Res. 17:4441–4454. Toledano, M., I. Kullik, F. Trinh, P. T. Baird, T. D. Schneider, and G. Storz. 1994. Redox-dependent shift of OxyR-DNA contacts along an extended DNA-binding site: a mechanism for differential promoter selection. Cell 78:897–909. Urbanowski, M. L., and G. V. Stauffer. 1989. Genetic and biochemical analysis of the MetR activator-binding site in the metE metR control region of Salmonella typhimurium. J. Bacteriol. 171:5620–5629. Wang, L., J. D. Helmann, and S. C. Winans. 1992. The A. tumefaciens transcriptional activator OccR causes a bend at a target promoter, which is

VOL. 177, 1995


partially relaxed by a plant tumor metabolite. Cell 69:659–667. 41. Wang, S.-P., and G. Stacey. 1991. Studies of the Bradyrhizobium japonicum nodD1 promoter: a repeated structure for the nod box. J. Bacteriol. 173: 3356–3365. 42. Wilson, K. J., S. G. Hughes, and R. A. Jefferson. 1992. The Escherichia coli gus operon: induction and expression of the gus operon in E. coli and the occurrence and use of GUS in other bacteria, p. 19. In S. R. Gallagher (ed.), GUS protocols: using the GUS gene as a reporter of gene expression. Academic Press, Inc., New York.

5695

43. Wu, W.-F., M. L. Urbanowski, and G. V. Stauffer. 1995. Characterization of a second MetR-binding site in the metE metR regulatory region of Salmonella typhimurium. J. Bacteriol. 177:1834–1839. 44. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103–119. 45. Zhou, C., L. Abaigar, and A. Y. Jong. 1990. A protocol for using T7 DNA polymerase in oligonucleotide site-directed mutagenesis. BioTechniques 8:503.