INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In its natural environment, Bacillus subtilis spends most of its life in a starving or nongrowing state because of different growth-limiting and stress conditions. As a soil bacterium, B. subtilis is exposed to runoff into the sea, and it is largely found in coastal waters, estuarine sediments and other saline habitats. It dominates the marine flora to such an extent that it could be considered a primary inhabitant of the oceans (3, 27). To adapt to drastic variations of environmental conditions including increasing saline concentrations, B. subtilis has developed a highly sophisticated regulatory network that involves transcriptional modulation of large sets of genes controlling cellular differentiation (26) and the induction of a set of proteins called general stress proteins (GSPs) or stress-specific proteins (11).

At least three distinct mechanisms of salt stress induction have been identified in B. subtilis. Transcription of the so-called class II heat shock genes, encoding GSPs, is activated by the alternative sigma factor ^B. This regulation involves two separate pathways which respond either to environmental stresses, including salt stress, or energy limitation, through a dual multicomponent network (15, 32). Class II heat shock genes encode a ^B-independent group of GSPs including FtsH (5) and the ClpP, ClpC, and Lon proteases (16, 28, 33). Among these proteins involved in stress response, ClpC was clearly shown to be required for tolerance to salt stress (16). The DegS-DegU two-component system was suggested to be involved in sensing salt stress, as exemplified by the effect of high salt concentration on the production of degradative enzyme synthesis and the expression of comG, a late competence gene, known to be controlled by DegU (18). In contrast to the first two mechanisms, which provide nonspecific stress resistance, the DegS-DegU regulatory pair seems to be subjected to salt-specific induction. Interestingly, the DegS-DegU signal transduction system also plays a key role in the complex network that governs post-exponential-phase responses under growth-limiting conditions (21).

As a first step toward a better understanding of B. subtilis behavior in saline environments, we have sought promoters that are differentially expressed in low-salt and high-salt conditions. Here, we report the isolation of the B. subtilis wapA promoter, from which expression is relatively strong in low-salt medium and completely repressed by high salt concentrations. The wapA gene encodes a wall-associated protein which belongs to a large family of high-molecular-weight, surface-associated proteins involved in various cellular processes, including surface hydrophobicity, pathogenicity, wall metabolism, secretion, and cell adhesion (34). We further show that salt stress repression of wapA is mediated by the DegS-DegU two-component system and propose a tentative target site for DegU-mediated regulation based on mutation/deletion analysis of the wapA regulatory region together with comparison of promoters known to be controlled by DegU.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains and genetic techniques. All strains used in this study are listed in Table 1.

1

1

1

1

1

1

Growth media. LSM medium contains 1.7% Bacto Agar (Difco), 0.2% casein hydrolysate (Oxoid), 0.5% glucose, 100 mM potassium phosphate buffer (pH 7.0), 3 mM MgSO₄, 22 mg of ferric ammonium citrate per liter, and 50 mg of L-tryptophan per liter, supplemented with 80 mg of X-Gal (5-bromo-4-chloro-3-indolyl--D-galactopyranoside) per liter to detect -galactosidase activity. HSM medium is LSM medium containing 0.7 M disodium succinate added from a 30% disodium succinate solution (pH 7.0). MB liquid medium contains 100 ml of a 10 × MB solution (tryptone, 100 g per liter; yeast extract, 50 g per liter) per liter, 3 mM MgSO₄, and 50 mg of L-tryptophan per liter, supplemented with 100 mM disodium succinate (low-salt MB) or 0.7 M disodium succinate (high-salt MB). Bacterial growth was followed by measuring the optical density at 600 nm in liquid cultures.

Nucleic acid manipulations. The B. subtilis genomic library used in this study was constructed as follows. Chromosomal DNA from wild-type strain 168 was partially digested with Sau3A, and fragments between 0.2 and 1.8 kb in size were ligated at the dephosphorylated BamHI site of the B. subtilis integrative vector pAZ7 to create transcriptional fusions with the lacZ reporter gene. E. coli TG1 was transformed with the ligation mixture, and plasmid DNA was extracted from a pool of 4,000 clones. After transformation of B. subtilis 168, integration into the chromosome at the amyE locus and screening on low-salt and high-salt media, the insert of the unique plasmid of interest was recovered by PCR using oligonucleotides complementary to regions located on each side of the insert and containing BamHI restriction sites. The resulting fragment was then cloned and sequenced.

Plasmid constructions. The following lacZ transcriptional fusions were constructed by using the pJM783 vector for integration at the original locus of the fused promoter and the pJM115 vector for double-crossover integration at the amyE locus (23). Plasmids pJM783 and pJM115 are chloramphenicol-resistant and kanamycin-resistant derivatives of pDH32, which harbors the lacZ reporter gene translated from the spoVG ribosome binding site (25). Plasmids pWP252 and pWP253 were obtained by subcloning the wapA promoter region and 5' end of the coding sequence comprised between oligonucleotides OMD26 (5'-GAATGAATTCGCTGAGGGTACGGATATTG-3') and OMD28 (5'-GTAGGGATCCCTAGTACATCGGCTGGCAC-3') at the EcoRI and BamHI sites (underlined) of pJM115 and pJM783, respectively. Plasmid pWP259 is a variation of pWP252 that carries a shorter insert, ending near the wapA start codon, which was obtained by PCR with the oligonucleotides OMD26 and WP3BA (5'-GCAAGGATCCTTTTAAAGTTTCGCCTCTTTC-3'). Plasmid pWP259.15 is a pWP259 derivative harboring the A₃₈G mutation isolated from pWP266.15 after PCR random mutagenesis (see below). A wapA'-lacZ fusion devoid of sequences located downstream of the transcription start site (pWP263) was constructed in pJM115 by using oligonucleotides OMD26 and WP+1 (5'-AAAGGGATCCTTACTCAATAATCTTAACTAG-3'). Plasmid pWP265 is pWP263 carrying the A₃₈G mutation. Finally, plasmid pWP280 carries the wapA promoter region devoid of sequences located upstream from the putative DegU target site, obtained with oligonucleotides OMD261 (5'-TCGCGAATTCTTAAAATAAGATAAATTTTCTAGAAA-3') and WP3BA.

Site-directed and PCR random mutagenesis of the wapA 5' regulatory region. Site-directed mutagenesis of the putative DegU binding site in the promoter region of wapA was carried out by PCR. A 150-bp upstream fragment (A) ending at the mutation site (TT_23/22AA) was amplified by using oligonucleotides OMD26 and 5'-TTAAGAAGACTGTTATATCATTACAATATTTTTC-3' and digested with EcoRI and BbsI to liberate 5'-TTAT-3' asymmetric overhang carrying the mutations to be introduced. The downstream fragment (B) was amplified with the oligonucleotides 5'-TATTGAAGACATATAACAGTCTAGTTAACATTATTG-3' and WP3BA. Digestion with BamHI and BbsI liberated 5'-ATAA-3' cohesive ends. A three-way ligation was performed between the pJM115 integrational vector containing the lacZ reporter gene (23) cleaved with EcoRI and BamHI and fragments A and B described above. The recombinant clone was designated pWP267, and the insert was sequenced to verify that no additional mutation had been introduced in the course of PCR amplification and cloning processes.

-Galactosidase assay. B. subtilis strains harboring lacZ fusions were assayed for -galactosidase activity as previously described (20).

Mapping of mRNA start site by primer extension. The synthetic oligonucleotide OMD25 (5'-GCCAACACTAAAAATGCTGCAATGAACC-3') was 5'-end labeled with [-³²P]ATP (110 TBq mmol¹) by using T4 polynucleotide kinase. Forty micrograms of RNA and 1 pmol of labeled primer were annealed in a total volume of 18 µl of reverse transcriptase buffer (50 mM Tris-HCl, 8 mM MgCl₂, 30 mM KCl, 1 mM dithiothreitol [pH 8.5]). The mixture was incubated for 3 min at 60°C and then slowly cooled at room temperature. One microliter (25 U) of avian myeloblastosis virus (Boehringer) reverse transcriptase and 1 µl of a deoxynucleoside triphosphate solution (20 mM each) were added. After 30 min at 37°C, reactions were stopped by the addition of 5 µl of 95% formamide, 10 mM EDTA, 0.3% xylene cyanol, and 3.3% bromophenol blue. Two microliters of each reaction was loaded on a 6% polyacrylamide sequencing gel.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Transcription of wapA is repressed under salt stress conditions. A B. subtilis genomic library was constructed by using the pAZ7 integrative vector that contains the lacZ reporter gene and genomic fragments between 0.5 and 1.8 kb in size to obtain a promoter library (see Materials and Methods for details). A preliminary screening of 150 clones was performed on high-salt and low-salt media to detect promoters affected by salt stress. One of the clones (strain QB4871) appeared to carry a promoter strongly expressed at low salt concentration as observed on LSM medium (minimal medium containing 100 mM disodium succinate) while significantly repressed in HSM (0.7 M disodium succinate). After PCR rescue and sequencing, the corresponding insert was found to harbor the promoter region and 5' end of the coding sequence of wapA, encoding a wall-associated protein (8). Analysis of the wapA and adjacent chromosomal regions indicated that wapA is preceded by an open reading frame (ORF1) ending 165 bp upstream of the wapA start codon and that the ORF1-wapA intergenic region contained a putative ^A-dependent promoter. To determine whether the observed phenotype was conferred by a promoter located directly upstream of the wapA ORF, wapA'-lacZ fusions were made by using a PCR-amplified fragment encompassing the ORF1-wapA intergenic region. The resulting transcriptional fusions, carried on plasmids pWP252 and pWP253, were integrated in the chromosome of the wild-type strain 168 at the amyE and wapA loci to give strains QB4950 and QB4951, respectively. In both cases, wapA-driven expression of -galactosidase was significantly decreased under salt stress conditions, as observed on HSM compared to LSM. This was further confirmed by monitoring the -galactosidase activity of strain QB4950 grown in low-salt and high-salt MB liquid media. The results shown in Fig. 1 indicate that wapA is transcribed from a promoter located in the ORF1-wapA intergenic region and subjected to a drastic repression under high-salt conditions. The transcription start site, determined by primer extension analysis, was found to be located at nucleotide position 35 relative to the wapA start codon (Fig. 3, lane 1). Putative 35 and 10 boxes typical of ^A promoters were detected in the corresponding regions.

View larger version (19K): [in this window] [in a new window]   FIG. 1.   Effect of salt stress on -galactosidase expression driven from a wapA'-lacZ transcriptional fusion integrated at the amyE locus. B. subtilis strain QB4950 was grown in MB medium (1% tryptone, 0.5% yeast extract, 50 µg of tryptophan per ml, 1 mM MgSO4) containing 100 mM disodium succinate () or 0.7 M disodium succinate (). The corresponding growth curves are shown with open (high salt) and closed (low salt) triangles. OD, optical density at 600 nm.

wapA belongs to the DegS-DegU regulon affected by salt stress. To determine whether wapA regulation could be categorized as one of the three salt stress induction mechanisms described above, expression of the wapA'-lacZ fusion in low-salt and high-salt media was tested in sigB, clpC, and degU deletional mutants. Only inactivation of the degU response regulator had an effect on the expression of wapA. More precisely, either deleting degU or replacing degU with the degU146 allele, which produces a nonphosphorylatable form of DegU, caused a strong derepression of wapA in high-salt medium (Fig. 2). These results confirmed that the DegS-DegU two-component system is involved in adaptation to salt stress and also indicated that transcription of wapA is repressed by the phosphorylated form of DegU under high-salt conditions. Interestingly, none of these two degU mutations (degU or degU146) had a significant effect on the expression of wapA in LB or low-salt medium (data not shown), indicating that DegU-mediated control of wapA occurs mostly under conditions of salt stress. These observations were confirmed by primer extension analysis using mRNA isolated from cultures of the wild-type and degU mutant strains in low-salt and high-salt media. As shown in Fig. 3, deletion of degU had no effect on either the position of the transcription start site or the apparent amount of wapA transcript under low-salt conditions (lanes 1 and 2). In agreement with the -galactosidase results shown in Fig. 1 and 2, the apparent amount of wapA transcript in the wild-type strain was strongly decreased by high salt concentrations whereas wapA transcription in the degU null mutant was independent of the salt concentration.

View larger version (21K): [in this window] [in a new window]   FIG. 2.   Effects of degU mutations on the expression of wapA'-lacZ under salt stress conditions. B. subtilis QB4871, QB4883, and QB4955 were grown in high-salt MB medium. Symbols: , QB4871 (wild type); , QB4883 (degU::erm); , QB4955 (degU146).

View larger version (60K): [in this window] [in a new window]   FIG. 3.   Determination of the transcription start site of wapA by primer extension analysis using an oligonucleotide complementary to the DNA sequence from positions +59 to +33 relative to the wapA start codon. The products of a sequencing reaction generated with the same oligonucleotide were run in parallel. mRNA was isolated from low-salt (lanes 1 and 2) and high-salt (lanes 3 and 4) cell cultures of the wild-type strain 168 and the degU strain QB4487, respectively. Equivalent amounts of total RNA were used in all primer extension experiments.

A consensus sequence upstream of DegU-regulated genes? In an attempt to define a putative target sequence for DegU control, the promoter regions of genes known to be regulated by DegS and DegU were aligned, and the consensus sequence AGAAN₁₁TTCAG was inferred (Fig. 4). To validate our observations, site-directed mutagenesis of the wapA promoter region was undertaken to transform the TTCAG conserved motif, which lies between the 35 and 10 boxes (Fig. 5), into AACAG, thereby modifying two invariant nucleotides in the putative DegU target site. The corresponding wapA[TT_22/23AA]-lacZ fusion, carried on plasmid pWP267, was integrated at the amyE locus of strain 168 and tested for its ability to be repressed under salt stress conditions. The -galactosidase phenotype observed on HSM solid medium as well as -galactosidase assay in liquid high-salt MB medium revealed that expression of wapA[TT_22/23AA]-lacZ was significantly derepressed in high-salt medium compared to the wild-type fusion (Table 2), indicating that the TT_23/22AA mutations prevented full repression of wapA transcription under salt stress conditions. However, the degU32(Hy) mutation, which encodes a highly stable phosphorylated form of DegU (4), was able to restore full repression whereas inactivation of degU led to an additional derepression (Table 2), suggesting that the putative DegU target site had been only partially altered by the TT_23/22AA modification and/or that another site within the wapA regulatory region is involved in DegU-mediated regulation.

View larger version (59K): [in this window] [in a new window]   FIG. 4.   Comparison of nucleotide sequences located upstream of genes reported to be regulated by DegU. Numbers preceding the sequences indicate positions of the leftmost nucleotide relative to the transcriptional start site; those in parentheses correspond to positions relative to the translational start site. The yxjJI operon was sequenced by Glaser et al. (10) in the framework of the genome sequencing project, and its transcription was shown to be repressed by DegU (4a).

View larger version (9K): [in this window] [in a new window]   FIG. 5.   Localization of random mutations leading to high-salt-resistant expression of wapA. The mutations indicated above the nucleotide sequence conferred specific derepression under salt stress conditions, while mutations that exhibited a nonspecific increase of expression in both low- and high-salt media are indicated below the sequence. Asterisks mark nucleotides modified by site-directed mutagenesis. The black-boxed sequences correspond to the conserved motif found upstream of genes known to be regulated by DegU, also referred to as the upstream site or site I. The dots under nucleotides in the upstream region indicate a potential dyad symmetry. The gray boxes highlight the 6-bp direct repeats encompassing the second focus of mutations (site II).

View this table: [in this window] [in a new window]   TABLE 2.   Effect of TT(23/22)AA and degU mutations on the repression of wapA transcription by salt stressa

Isolation of random mutations that prevent repression of wapA by salt stress. To confirm the results obtained by site-directed mutagenesis and further define the target of wapA repression by salt stress, a PCR random mutagenesis was performed on the 5' regulatory region of wapA as described in Materials and Methods. A library of potentially mutated PCR fragments was obtained and subcloned in the replicative vector pHT304-18Z (1) in front of the lacZ reporter gene. The resulting library was transferred into B. subtilis 168 and screened for clones that were no longer sensitive to high-salt repression and had therefore retained a Lac⁺ phenotype on HSM solid medium. Among 12,000 clones examined, over 50 colonies displayed a light to dark blue color; by comparison, the control strain harboring the wild-type wapA regulatory region (plasmid pWP266) remained white on HSM. Plasmid DNA was isolated from the positive clones, designated pWP266.1 to pWP266.48S; the corresponding inserts were sequenced, and only those clones that carried a single mutation were retained. Further screening allowed us to distinguish two classes of mutants. The majority (26 of 30) were specifically derepressed on HSM, while their expression remained unchanged on low-salt or sporulation medium. On the other hand, 4 clones of 30 exhibited a nonspecific increase of expression on all media tested. Accordingly, these mutants appeared to carry a promoter-up mutation (TAAGATTATGAT) that improved the 10 box recognized by ^A. A compilation of these mutations is presented in Fig. 5. Interestingly, the high-salt-specific mutations appeared to fall into two separate clusters: one corresponding to the conserved sequence thought to be involved in DegU binding, and a second focus of high mutation frequency located downstream from the transcription start site within the 5' untranslated region (Fig. 5). Examination of the DNA sequence encompassing the second cluster of mutations allowed the detection of a direct repeat of two 6-bp motifs (TATTAC) separated by a three-nucleotide spacer. Shortening of this spacer by 1 bp (mutant 3S) also led to a derepressed phenotype. We selected a subset of representative mutants of each class in each region and determined the corresponding -galactosidase activities in low-salt and high-salt media. The results presented in Table 3 and Fig. 6 confirm the phenotypes observed on plates. It must be noted that pWP266 plasmids are pHT304 derivatives which are present at about four copies per equivalent chromosome (2).

View larger version (14K): [in this window] [in a new window]   FIG. 6.   Schematic representation of the wapA regulatory region together with the different constructs designed to alter or delete each one or both of the putative sites involved in repression by salt stress. Plasmid pWP259 contains the entire wild-type 5' regulatory region of wapA transcriptionally fused to lacZ. In pWP280, sequences located upstream of the two clusters of random mutations have been removed. Plasmid pWP263 is a pWP259 derivative that lacks the downstream site (site II). Plasmid pWP267 harbors the TT22/23AA site-directed mutation, pWP259.15 carries the A38G mutation isolated in site I by random mutagenesis, and pWP265 combines deletion of site II and mutation A38G in site I. Each plasmid was integrated at the chromosomal amyE locus of the wild-type, degU, and degU32(Hy) strains. The corresponding -galactosidase activities were determined in high-salt liquid cultures as described for Table 3 and are shown at the right. The data relative to the wapA[TT22/23AA] mutant obtained by site-directed mutagenesis are shown for easier comparison.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

As a widespread soil bacterium also found in coastal and estuarine sediments, as well as in marine and freshwater habitats (27), B. subtilis must be able to adapt to transient or permanent high salt concentrations in its environment. Extensive studies on general and salt stress-specific proteins (see reference 11 for a review) together with the recent advances in functional analysis of the newly sequenced B. subtilis genome indeed indicate that B. subtilis modulates the expression of an impressive number of genes in response to salt stress (2a). However, the molecular mechanisms that govern salt stress-specific regulation remain to be fully elucidated. Recently, expression of genes that are controlled by the DegS-DegU two-component system was shown to be affected under salt stress conditions, such as 1 M NaCl or 1 M KCl. The stimulatory or inhibitory effect proved to be DegS-DegU dependent, as well as salt specific since it could not be obtained during growth at high osmolarity such as 0.5 M lactose (18). A similar effect was also obtained when NaCl was replaced with disodium succinate, indicating that chloride anions were not involved in this salt-mediated regulation (16a).

In the course of a screening for novel promoters modulated by salt stress, the wapA gene has been isolated and shown to undergo negative regulation by the DegS-DegU regulatory pair in high-salt medium. Since inactivation of degU had no effect on wapA transcription at low salt concentrations (data not shown), we propose that the negative control exerted on wapA under salt stress conditions is mediated at least partially by DegU. We also observed that the DegU(Hy)32 protein, of which the phosphorylated form is highly stable, remains able to repress wapA transcription at low salt concentrations (data not shown). This finding supports the hypothesis that salt stress increases the amount of phosphorylated DegU in the cell, by either increasing its synthesis or stimulating DegS-mediated phosphorylation of DegU. As already pointed out, DegU is a pleiotropic regulator involved in various post-exponential-phase responses. It positively controls the synthesis of certain exoenzymes such as the sacB-encoded levansucrase, which produces branched polymers called levans (17). Accordingly, it has been observed that salt stress increases sacB transcription (18). In addition, DegU is a positive regulator of competence and DNA uptake. Thus, the DegS-DegU signal transduction system could be viewed as a means to modulate the synthesis, degradation, or entry into the cells of various macromolecules, thereby participating to osmoregulation and adaptation upon osmotic upshock. In the case of B. subtilis WapA, no precise function could be assigned to the protein based on the absence of any particular phenotype of a wapA null mutant (8). Nevertheless, wall-associated proteins are generally involved in pathogenicity (e.g., the major antigen A from Streptococcus mutans) (7), wall metabolism, secretion, adhesion, or other cell surface-associated properties (34). It could be hypothesized that WapA increases cell wall permeability by functioning as a sieve, or an ion pore or channel, and that turning off its synthesis may prevent excessive exchanges between the outside medium and the cytoplasmic compartment.

Based on a conserved region previously detected by Dubnau upstream from the comC, comG, and recA genes (6), alignment of several promoters known to be controlled by the DegS-DegU system allowed the identification of the conserved sequence AGAAN₁₁TTCAG. It must be pointed out that in the case of sacB, the two detected motifs (Fig. 4) are centered around a 20-bp region shown to be strictly required for DegU-mediated activation, by successive 5' deletions in the control region of sacB (4a, 12). A computer search of the B. subtilis genome has revealed sequences that perfectly match the consensus presented above in promoter regions not known to be regulated by DegU to date, such as rapA, encoding an aspartate phosphatase that functions to prevent sporulation (24), and paiB, a negative regulator of sporulation and degradative enzyme synthesis (13). Since both rapA and paiB null mutants are able to sporulate in glucose-enriched medium (13, 22) as is the degU32(Hy) mutant, one could hypothesize that phosphorylated DegU lies upstream of rapA and paiB in the regulatory cascade leading to catabolite repression of sporulation.

Site-directed mutagenesis of the putative DegU target, by changing the two invariant T residues in the TTCAG motif to obtain AACAG, led to a partial derepression of wapA transcription in high-salt medium compared to the transcriptional levels of the wild-type wapA'-lacZ in a degU null mutant (about 1900 Miller units). Deleting degU in the wild-type or wapA' [TT_23/22AA] wapA'-lacZ strain had a similar effect, supporting the hypothesis that the elevated levels observed in strain QB4963 (wapA[TT_23/22AA]) were due to a decreased efficiency of DegU repression. At least two reasons can be envisioned to explain the partial effect of the TT_23/22AA mutation: (i) it may only slightly decrease the binding capability of DegU to the modified target or (ii) DegU could bind to some additional site(s) located in the wapA control region to further downregulate its transcription. The observation that DegU(Hy)32, a mutant version of DegU that exhibits a hyperstable phosphorylated state, remains able to fully repress transcription of wapA [TT_23/22AA] is compatible with either of these explanations. Complete elimination or drastic modification of the putative target of DegU proved impossible since it overlaps the 35 box that specifies ^A binding.

Random PCR mutagenesis of the control region of wapA confirmed the role of the putative DegU binding site (site I) in repression of wapA transcription at high salt concentrations and allowed the detection of a second site of repression located downstream from the transcription start site (site II [Fig. 5]). The relatively low occurrence of mutations that were obtained in site I could be due to possible interference of such mutations with the binding of ^A to the 35 promoter box. Using constructs designed to alter site I and/or delete site II, we have shown that the negative control at the two sites is cumulative in the wild-type background as well as in the degU null mutant. This suggests the existence of a second mechanism of repression by salt stress occurring at site II and DegU independent, since deletion of site II causes an additional derepression in a degU background (compare -galactosidase activities conferred by constructs pWP259 and pWP263 or by pWP259.15 and pWP265 in the degU strain). Furthermore, derepression observed in the absence of the downstream operator site proved to be salt specific and could not be observed at low salt concentrations (data not shown). It therefore seems unlikely that the removal of site II, located downstream from the transcription start site, generally affects transcription efficiency by altering the 5' untranslated sequence of the wapA transcript in pWP263 and derived plasmids. The hypothesis of a second repressor is also reinforced by the observation that transcriptional levels of wapA'-lacZ in low-salt medium (Fig. 1) are twofold higher than those observed at high salt concentrations in the absence of active DegU (Fig. 2), suggesting the existence of residual negative control in the absence of DegU.

However, the degU32(Hy)-encoded protein remained able to fully repress transcription of wapA'-lacZ fusions that carried either an intact or a modified site I (Fig. 6). In addition to the wapA mutant promoters presented in Fig. 6, regulation by the degU32(Hy)-encoded protein was examined on each PCR mutant in the putative operator site I (pWP259.21S, -13S, and -32S [Fig. 5]). None of these mutations was able to eliminate or otherwise affect the ability of DegU(hy) to repress wapA transcription at high salt concentrations. Since DegU(Hy) only partially represses wapA at low salt concentrations (not shown), the effect of each mutation located in site I was examined in low-salt medium in the degU32(Hy) background. A slight resistance to DegU(Hy) repression could be observed with pWP259.21S and 32S compared to the wild-type promoter (twofold decrease), suggesting that at least those two mutations affect indeed the ability of DegU(Hy) to repress wapA. On the other hand, deletion of site II also appeared to slightly affect repression by the DegU(Hy) protein (Fig. 6). Although the significance of these last data might be questionable, the possibility that DegU itself is involved directly or indirectly in regulation at the downstream site cannot be ruled out. A computer search of the B. subtilis genome revealed sequences that match the TATTACN₃TATTAC motif upstream from aprE, srfA, and sacX, suggesting that the site I-site II combination may not be unique to the wapA regulatory region. Transposon mutagenesis will be undertaken in an attempt to identify an additional repressor and characterize the second mechanism of salt stress repression exerted on wapA in combination with DegU.