Identification and characterization of a stress-responsive promoter in the macromolecular synthesis operon of Bacillus subtilis

Induction of σ^A in the B. subtilis Ts sigA mutants harbouring pCX2Fx plasmids is caused by the loss of σ^A function at high temperature and is mainly from the plasmid-borne sigA

We have reported that there is an induction of σ^A in B. subtilis DB1005 (pCX2F5) but not in B. subtilis DB2 (pCX2F) under heat stress (Chang et al., 1997). As the chromosomal and the plasmid-borne sigA genes in both B. subtilis strains were controlled by the same transcriptional and translational signals, respectively, we suspected that the difference in σ^A induction between these two B. subtilis strains was attributed to their difference in σ^A activity at high temperature. To confirm this idea, the levels of σ^A and GroEL induction in various B. subtilis strains harbouring the corresponding pCX2Fx plasmids (Fig. 1A and Table 1) were examined by Western blot analyses. The level of GroEL induction was used as an indicator of σ^A activity, as it is transcriptionally controlled by σ^A (Chang et al., 1994; Yuan and Wong, 1995). As shown in Fig. 2, the levels of σ^A induction were high in the Ts sigA mutants such as B. subtilis DB1005 (pCX2F5), DB1001 (pCX2F1) and DB1009 (pCX2F9). However, no clear induction of σ^A was observed for B. subtilis DB1008 (pCX2F8), DB1002 (pCX2F2) and DB2 (pCX2F), which were relatively resistant to elevated temperature (data not shown). Moreover, no induction of GroEL was observed for the Ts sigA mutants (Fig. 2). These results suggest that the induction of σ^A is triggered by low σ^A activity. Presumably, a common regulatory network is adopted by B. subtilis, with which the effect of the defective σ^A on cells can be overcome at high temperature.

Where was the Ts sigA gene induced? Was it mainly from the chromosomal or the plasmid copy of sigA gene? To answer this question, the pDA4 plasmid (Fig. 1B and Table 1) that bears a truncated sigA gene controlled by the same transcriptional and translational signals as those on the pCX2Fx plasmids was introduced into B. subtilis DB2 and DB1005. The levels of intact and truncated σ^A thus expressed in the cells of B. subtilis DB2, DB1005, DB2 (pDA4) and DB1005 (pDA4) were then analysed before and after heat shock. Our data showed that the levels of intact σ^A expressed from the chromosomes of B. subtilis DB2, DB1005 and DB2 (pDA4) were fairly constant during the transition in temperature (Fig. 3). However, the level of truncated σ^A was induced exclusively in B. subtilis DB1005 (pDA4), indicating that the induction of σ^A at elevated temperature was mainly from the ‘plasmid-borne’sigA allele.

σ^B and its cognate promoter are required for the induction of plasmid-borne sigA at elevated temperature

What are the elements responsible for the induction of plasmid-borne sigA? As the sigA genes on the pCX2Fx plasmids were controlled by the same P1P2 promoter-containing Sau3A DNA fragment (Figs 1A and 4A) (Wang and Doi, 1987) and the Ts σ^As lost their activities at elevated temperature, we suspected that the induction of σ^A from pCX2Fx must be dependent on elements other than σ^A and its cognate P1P2 promoters, similar to that reported for the B. subtilis csh203::Tn917lac mutant in which the minor σ factor, σ^H, becomes essential for σ^A expression during vegetative growth (Hick and Grossman, 1995). To prove this idea, we examined the DNA sequence of the P1P2 promoter-containing Sau3A DNA fragment. A nucleotide sequence homologous to the consensus sequence of the B. subtilisσ^B-type promoters (Moran et al., 1982; Kalman et al., 1990; Boylan et al., 1991; Mueller et al., 1992; Varon et al., 1993; Antelmann et al., 1995; Engelmann et al., 1995; Maul et al., 1995) was found just upstream of the −35 hexamer of the P2 promoter (Fig. 4A); it matched exactly at least at the five positions most important for the activity of the well-characterized σ^B-dependent ctc promoter (Fig. 4B) (Tatti and Moran, 1984; Ray et al., 1985). This putative promoter was designated P7. We thought that the induction of σ^A was ascribed to transcription from this promoter by RNA polymerase holoenzyme containing σ^B. To confirm this idea, we constructed an integration vector, pCTT5, in which the internal HindIII fragment (90 bp long) of sigB was replaced with a tetracycline (tet ) resistance cassette. We then linearized the plasmid and integrated it into the chromosomes of B. subtilis DB2 and DB1005 to obtain the sigB-disrupted mutants (see Experimental procedures). The sigB mutants with the genetic backgrounds of B. subtilis DB2 and DB1005 were designated as B. subtilis DB2004 and DB5004 respectively. The effects of sigB disruption on σ^A induction were examined in B. subtilis DB2 (pCX2F), DB1005 (pCX2F5), DB2004 (pCX2F) and DB5004 (pCX2F5) under heat stress (Fig. 5). Again, heat induction of σ^A was observed in B. subtilis DB1005 (pCX2F5). However, the induction was almost completely abolished by sigB disruption, as seen in B. subtilis DB5004 (pCX2F5). The lack of σ^A induction in B. subtilis DB5004 (pCX2F5) indicates that the induction requires σ^B, with which P7 can probably be transcribed.

Transcription of P7 by σ^B-RNA polymerase holoenzyme in vivo

To confirm that P7 was transcribed by σ^B-RNA polymerase holoenzyme under heat stress, we took advantage of the primer extension assay with which RNA transcripts generated from different initiation sites can be visualized in just a single experiment. In this assay, the total RNA samples isolated from the tested B. subtilis strains before and after heat shock were primed with the synthetic primer pE7B (see Experimental procedures). As shown in Fig. 6, the + 1 sites of RNA transcripts generated from the P2 promoters of all the tested B. subtilis strains lined up at the same position as that reported previously (Wang and Doi, 1987) at both 37°C and 49°C. However, two extra RNA transcripts with the + 1 sites located 39 and 40 nucleotides, respectively, upstream of the P2 initiation site were induced significantly in B. subtilis DB1005 (pCX2F5) at 49°C (lanes M and 5). The two + 1 sites corresponded to an A and a C residue preceded by the P7 promoter DNA sequence, 5′-GGGTTTTT-13 bp-GGGAAT-3′ (Fig. 4). Such induction was also detected in B. subtilis DB1005 (pCX2F5) at 37°C (lane 2) and B. subtilis DB2 (pCX2F) at 49°C (lane 4) after a prolonged exposure of the sequencing gel (data not shown), whereas no such induction was detectable in B. subtilis DB5004 (pCX2F5) (lanes 3 and 6) at both 37°C and 49°C even after a prolonged exposure. These results were consistent with the exclusive σ^A induction in B. subtilis DB1005 (pCX2F5) at elevated temperature, as shown in Fig. 5. They also demonstrated that there was a σ^B-type promoter upstream of P2 in the B. subtilis MMS operon. Unexpectedly, P2 induction was also observed in both B. subtilis DB1005 (pCX2F5) and DB5004 (pCX2F5) at 49°C (lanes 5 and 6), and it was more significant with the latter strain. The mechanism responsible for the induction is interesting and will be discussed later.

P7 can be transcribed in vitro by reconstituted RNA polymerase holoenzyme containing σ^B

To confirm the σ specificity of P7, in vitro transcription was performed with reconstituted σ^B-RNA polymerase holoenzyme. The presence of strong T1T2 terminators (Jurgen et al., 1981) upstream and downstream of P1P2 and P7 on the template plasmid pCT20 (Fig. 1C and Table 1) enabled us to detect the RNA transcripts generated from both P2 (≈ 348 nucleotides) and P7 (≈ 388 nucleotides) precisely. The data are shown in Fig. 7. As expected, the P2 transcript was observed when pCT20 was transcribed with reconstituted RNA polymerase holoenzyme containing wild-type σ^A at both 37°C and 49°C (lanes 1 and 6). Likewise, the P7 transcript was synthesized in the presence of σ^B (lanes 3 and 8). Together with the data from σ^A induction and primer extension assays (Figs 5 and 6), it was clear that P7 was absolutely a σ^B-type promoter in the B. subtilis MMS operon. Moreover, a higher amount of P2 transcript was obtained when σ^B was mixed with the wild-type σ^A (lanes 2 and 7) than with the Ts σ^A (lanes 4 and 9), suggesting that the Ts σ^A factor was less efficient than the wild-type counterpart in competition with σ^B for core binding. To our surprise, a very strong RNA band (278 nucleotides in length) was observed in all assays containing σ^B (lanes 2, 3, 4, 7, 8 and 9). It seemed that a second σ^B-type promoter was present in the MMS operon. As the RNA transcript produced from this putative promoter (designated as P8) was shorter than that generated from P2, we assumed that it would be located downstream of P2. As shown in Fig. 4, the putative P8 promoter has a very good −10 consensus sequence despite the possession of a non-conserved −35 sequence and a longer spacer region (15 bp). As no transcript was generated from this promoter in vivo (Fig. 8, lanes 1–8), we thought that it might not be of any physiological significance. However, the synthesis of P8 transcript in vitro without further addition of σ^B (Fig. 7, lanes 1, 5, 6, 10 and 11) suggests that our core enzyme is contaminated with σ^B during purification with the cationic exchanger Bio-Rex 70. A similar phenomenon was also reported for the purification of core RNA polymerase using phosphocellulose (Haldenwang and Losick, 1979).

Induction of P7 after the imposition of heat, ethanol and salt stresses

Is the induction of P7 restricted to the Ts sigA mutants under heat stress? Are there other stresses that could also induce the expression of P7 in B. subtilis? To answer these questions, total RNA was extracted from the cultures of B. subtilis DB2 and DB2004 after the imposition of heat, ethanol or salt stress. RNA transcript generated from P7 was then examined using primer extension assays (Fig. 8). Our data showed that P2 was constitutively expressed in B. subtilis DB2 and DB2004 (lanes 1–8); however, no P1 transcript was observable, probably because of its fairly low expression in B. subtilis (Qi et al., 1991). P7 was not expressed until the imposition of heat (lane 2), ethanol (lane 3) and salt (lane 4) stresses, with the least induction observed for heat shock. Minor induction of P7 was also observed in B. subtilis under oxidative stress (data not shown). The induction of P7 was completely abolished when sigB was disrupted (lanes 6–8). Similar results were obtained with B. subtilis DB2 and DB2004 bearing the pCX2F plasmid (data not shown). Thus, it is certain that P7 is transcribed by σ^B-RNA polymerase holoenzyme and is responsive to stresses.

Bacterial strains, plasmids and culture conditions

The sources and properties of bacterial strains and plasmids used in the study are listed in Table 1. The method for the construction of B. subtilis DB1008 and DB1009 was the same as that reported previously (Chang and Doi, 1993). The overlapping primers for the construction of the two mutant sigA genes were as follows: B8 5′-TCGGCCGCGGCGCGTGTAATCGCC-3′; C8 5′-CACGCGCCGCGGCCGATCAGGCGAGAAC-3′; B9 5′-GGCGCGTGTGGCCGCCTGTCTGATCCAC-3′; C9 5′-GGCGGCCACACGCGCCATTGCCG-3′. Co-transformation and direct DNA sequencing were adopted to confirm the two sigA mutants. B. subtilis was grown in glucose minimal medium (GMM) at 37°C to a cell density of 5 × 10⁷ ml⁻¹ (A₅₅₀ = 0.4) before the imposition of stress according to the following schemes. Heat shock: the culture was transferred from 37°C to 49°C. Ethanol stress: ethanol was added at a final concentration of 4% (v/v). Salt stress: sodium chloride was added at a final concentration of 4% (w/v). Oxidative stress: hydrogen peroxide was added at a final concentration of 0.005% (v/v). The bacterial culture was sampled before and 10 min after the imposition of stress.

Western blot analyses of σ^A and GroEL

The method for Western blot analyses of σ^A and GroEL was similar to that reported previously (Chang and Doi, 1990). In brief, B. subtilis strains were grown in GMM at 37°C to an A₅₅₀ of 0.4 (referred to as time zero) and then shifted to 49°C. One millilitre of culture was harvested at each designated time point. After resuspending the cell pellet with 50 μl of 0.5 × SET buffer [(20% sucrose, 50 mM Tris-HCl, pH 7.6, 50 mM EDTA, 2 mM phenylmethylsulphonyl fluoride (PMSF)] and digestion of the cell wall with lysozyme, 50 μl of 2 × SAB (125 mM Tris-HCl, pH 6.8, 4% SDS, 20% glycerol, 0.002% bromophenol blue) was added to disrupt the cells. Five microlitres of the cell lysate was then used for Western blot analysis. Antibodies against σ^A and GroEL were prepared as mentioned previously (Chang and Doi, 1990; Chang et al., 1994).

Construction of B. subtilis strains with a disrupted sigB gene

To create the sigB-disrupted mutants, we first constructed the pCT30 plasmid by eliminating the HindIII site on pUC19. Then, a 1.9 kb DNA fragment containing rsbV, rsbW, sigB and rsbX was amplified from the chromosome of B. subtilis DB2 by polymerase chain reaction (PCR) using the two primers, PBO5 (5′-TGACTGCTGCAGAAGCTCATTGAGGAAC-3′) and PBO3 (5′-GTCATACTGCAGAGGATTCGTTAGCAAG-3′) (Kalman et al., 1990). After digestion with PstI, the DNA fragment was ligated with pCT30 to generate pCT32. In parallel, a 1.7 kb tetracycline resistance cassette was PCR amplified from the pHY300PLK plasmid (Ishiwa and Shibahara, 1985; 1986) using the primers PTC5 (5′-ACGTACAAGCTTGGGAACGGAAAAATTATTTTATTAA-3′) and PTC3 (3′-TCGGTGAAGCTTGAACTCTCTCCCAAAGTTGATC-3′). After digestion with HindIII, the DNA fragment was used to replace the 90 bp HindIII fragment located within the sigB gene on pCT32. The resulting plasmid, pCTT5, was linearized and integrated into the chromosomes of B. subtilis DB2 and DB1005 respectively. The sigB-disrupted mutants generated by a double cross-over recombination event were selected on tetracycline plates, verified by PCR, immunoprecipitation and σ^B activity.

Purification of total RNA

The procedure for the purification of total RNA was the same as that mentioned in the protocol of the High Pure RNA isolation kit from Boehringer Mannheim. The amount of RNA was quantified by determining its OD at 260 nm.

Primer extension assay

The primer used for the study was PE7B (5′-GAGTCGTATTAATTTCGCGGG-3′), which is complementary to the sequence of the T7 phage gene 10 promoter on pCX2F (Chang et al., 1997). Primer extension assay was performed using the following procedure: 5 μg of total RNA was mixed with 9.3 μl of DEPC-treated H₂O and 1 μl (1 μg μl⁻¹) of the primer PE7B. After denaturation of the sample at 65°C for 10 min, the mixture was cooled down slowly at 30°C for 1 h to allow the annealing of primer and RNA template. Then, 8.7 μl of reaction cocktail comprising 2 μl of dG labelling mixture (USB sequencing kit), 0.5 μl of RNasin (20 units; Promega), 0.2 μl of avian myeloblastosis virus reverse transcriptase (40 units; Bethesda Research Laboratories), 2 μl of [α-³⁵S]-dATP (>1000 Ci mmol⁻¹; Amersham) and 4 μl of RT buffer (250 mM Tris-HCl, pH 8.3, 375 mM KCl) was added to the annealed sample. The reaction mixture was incubated further at 42°C for 10 min before adding 1 ml of 10 mM dNTP to allow the extension of DNA chain. The reaction was stopped 2 h later by adding 20 μl of stop solution (95% formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol FF). Five microlitres of the final mixture was denatured at 95°C for 3 min and electrophoresed on a 6% polyacrylamide gel containing 8 M urea.

Overproduction of σ^B

To overproduce σ^B, we amplified the sigB gene from the B. subtilis DB2 chromosome by PCR using the two linker primers, PSIGB5 (5′-ATCGAGGGATCCAAGGAGATATACATATGACACAACCATCAAAAACTACG-3′) and PSIGB3 (5′-GAGATCCTGCAGTCATTACATTAACTCCATCGAG-3′) (Binnie et al., 1986). The forward primer, PSIGB5, contained the ribosome binding site for the T7 phage gene 10 protein as well as a BamHI site in the 5′ end, while the reverse primer, PSIGB3, contained a PstI site. After digesting the sigB gene with BamHI and PstI, the DNA fragment was cloned into the compatible sites of the overexpression vector pT7-5 (Tabor, 1990) to form the pOB2 plasmid. Overexpression of σ^B in E. coli BL21(DE3) was performed as reported previously (Chang and Doi, 1990).

Purification of σ^B and preparation of anti-σ^B antibody

The cells collected from 1 l of overexpression culture were suspended in 40 ml of buffer L [10 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM dithiothreitol (DTT), 1 mM PMSF, 200 mM NaCl, 10% glycerol, 130 μg ml⁻¹ lysozyme] followed by incubation on ice for 30 min and disruption with a sonicator. The cell debris containing the overexpressed σ^B protein in the form of an inclusion body was collected by centrifugation at 8000 r.p.m. (4°C for 30 min) and then resuspended with 3 ml of buffer L. To purify σ^B for antibody preparation, a equal volume of 2 × SAB buffer was added to the protein sample before separation on SDS–polyacrylamide gels. The separated σ^B protein was cut out from the gel and collected with an eluter (Isco) using TAE buffer (40 mM Tris-acetate, pH 8.0, 1 mM EDTA). The σ^B protein thus purified had a purity over 95%. After being concentrated, the protein was mixed with adjuvant and injected into rabbit to raise antibody. For in vitro transcription assays, the purified σ^B protein was precipitated with four volumes of cold acetone (−20°C) by centrifugation at 10 000 r.p.m. for 20 min, and then washed twice with 1 ml of 80% acetone containing 0.15 M NaCl. This would allow the removal of SDS from the protein sample as much as possible. The washed protein pellet was dried, dissolved in 20 μl of 6 M guanidine-HCl and diluted stepwise with an equal volume of dilution buffer (50 mM Tris-HCl, pH 7.9, 0.1 mM EDTA, 1 mM DTT, 0.15 M NaCl, 5% glycerol) until the concentration of guanidine-HCl reached 0.06 M. The diluted protein sample was further incubated overnight at 4°C, concentrated with a Centricon-10 tube (Amicon) and finally dialysed against storage buffer (20 mM Tris-HCl, pH 7.9, 10 mM MgCl₂, 1 mM EDTA, 0.1 mM DTT, 1 mM PMSF, 0.2 M KCl, 50% glycerol).

In vitro transcription

Core enzyme and σ^A were prepared as described previously (Fukuda and Doi, 1977; Halling et al., 1977; Davison et al., 1979; Chang and Doi, 1990). Procedures for the preparation of reconstituted RNA polymerase and in vitro transcription assay were modified from a previously published version (Chang and Doi, 1990). The RNA polymerase holoenzyme was obtained by mixing 15 μl of core enzyme (1 μg) with 10 μl of purified σ (1.5 μg) on ice for 10 min. Afterwards, 25 μl of the supercoiled pCT20 plasmid (1.5 μg) was added to the enzyme solution, and the mixture was further incubated on ice for 10 min. After this, the mixture was transferred to the designated temperature (37°C or 49°C), and 70 μl of reaction cocktail prewarmed at the designated temperature was added to start the transcription reaction. The final concentration of each component in the reaction mixture was 40 mM Tris-HCl, pH 7.9, 10 mM MgCl₂, 150 mM KCl, 0.4 mM DTT, 5% glycerol, 0.2 mM each of UTP, CTP, GTP, ATP and 6 μCi [α-³²P]-ATP. After incubation of the reaction mixture at the designated temperature for 1 h, the sample was immediately placed on ice. Four micrograms of E. coli tRNA was then added to the reaction mixture before extraction with phenol–chloroform. One hundred microlitres of RNA sample in the upper aqueous phase was pipetted into a tube, and the RNA transcripts were precipitated with 10 μl of 3 M sodium acetate and 250 μl of 100% ethanol at −70°C. The pelleted RNA was then dissolved in stop buffer (95% formamide, 20 mM EDTA, 0.05% bromophenol blue, 0.05% xylene cyanol FF), heated at 95°C for 3 min and run on a denaturing polyacrylamide gel. Autoradiography was performed after the electrophoresis.