Genetic Evidence for the Actin Homolog Gene mreBH and the Bacitracin Resistance Gene bcrC as Targets of the Alternative Sigma Factor SigI of Bacillus subtilis

Abstract

The Bacillus subtilis sigI gene, which is a member of the class VI heat shock genes of the B. subtilis heat shock stimulon, encodes an alternative sigma factor whose regulon is poorly defined. In this study, by using a binary vector system, we showed that B. subtilis SigI could drive expression of a transcriptional fusion between the sigI regulatory region from Bacillus licheniformis, Bacillus sp. strain NRRL B-14911, B. subtilis, or Bacillus thuringiensis and the xylE reporter gene in B. subtilis. The transcriptional initiation sites of these fusions in B. subtilis were mapped by primer extension analyses. A putative consensus promoter sequence probably recognized by the B. subtilis SigI was thus deduced. Using a consensus sequence-based search procedure, we found putative σ^I promoters preceding the actin homolog gene mreBH and the bacitracin resistance gene bcrC of B. subtilis. Overexpression of the B. subtilis sigI gene could specifically stimulate expression of both an mreBH promoter region-bgaB fusion and a bcrC promoter region-bgaB fusion. Expression of these two fusions at the amyE locus of the B. subtilis chromosome was heat inducible and SigI dependent as revealed by sigI gene disruption experiments. Primer extension analysis showed that the identified mreBH and bcrC transcriptional start sites were at appropriate distances from their σ^I promoter elements. This further supports the notion that SigI can directly regulate mreBH and bcrC expression. Taken together, these results strongly suggest that mreBH and bcrC are new members of the SigI regulon.

The responses of the soil bacterium Bacillus subtilis to various environmental stresses such as high temperature (16, 33), alkaline media (7, 13), high osmolarity (14), oxidative stress (8, 16, 27, 33), and antibiotic stress (3, 5, 6, 9, 23) involve a variety of sigma factors. Induction of expression of genes encoding these sigma factors or activation of these sigma factors can direct core RNA polymerase to coordinate differential gene expression that mediates various cellular responses to allow for survival and persistence in harsh and adverse environments.

The B. subtilis sigI gene encodes an alternative sigma factor of the σ⁷⁰ family (22). Transcription of sigI is heat inducible (50 or 52°C) (37) and depends on SigI (1). Disruption of the sigI gene led to a temperature-sensitive phenotype: the sigI mutant could not grow on LB plates at 55°C and did not form colonies on supplemented minimal medium plates at 54°C. The sigI gene was initially proposed to be a member of the class IV heat shock genes of B. subtilis (37) but was later reassigned into class VI of the B. subtilis heat shock stimulon (30). The B. subtilis sigI gene has orthologs in other Bacillus species, such as Bacillus licheniformis ATCC 14580 (26) (GenBank accession number YP_078683), Bacillus thuringiensis serovar Israelensis ATCC 35646 (ZP_00741267), and Bacillus sp. strain NRRL B-14911 (ZP_01172170). The sigI gene and its downstream rsgI (formerly ykrI) gene are cotranscribed and constitute an operon. The rsgI gene encodes a putative transmembrane protein that can interact with SigI and functions as an anti-σ^I factor (1). To date, little is known about the SigI regulon. In this study, we have derived a putative consensus sequence of the σ^I promoter and demonstrate that the actin homolog gene mreBH (10) and the bacitracin resistance gene bcrC (20) are targets of SigI.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. Escherichia coli and B. subtilis cells were grown in Luria-Bertani (LB) medium (28). Antibiotics were used at the following concentrations (μg/ml): ampicillin, 100 (for E. coli); chloramphenicol, 5; tetracycline, 10; and erythromycin, 1 (for B. subtilis).

TABLE 1.

Bacterial strains and plasmids

Strain or plasmid	Descriptiona	Reference or sourceb
E. coli DH5α	F− φ80dlacZΔM15 Δ(lacZYA-argF) recA1 gyrA endA1 relA1 supE44 hsdR17	Laboratory stock
B. subtilis strains
168	trpC2	Laboratory stock
BM1140	trpC2 amyE::PmreBH-bgaB	This study
BM1168	trpC2 amyE::P“bcrC”-bgaB	This study
BM1199	trpC2 amyE::PmreBH-bgaB sigI::pGS1651	This study
BM1201	trpC2 amyE::P“bcrC”-bgaB sigI::pGS1651	This study
B. licheniformis ATCC 14580		ATCC
B. thuringiensis serovar Israelensis ATCC 35646		ATCC
Bacillus sp. strain NRRL B-14911		ARSCC
Plasmids
pDL	Integrative vector for generation of transcriptional fusions to bgaB; Apr Cmr	34
pHY300PLK	Shuttle vector; Apr Tcr	Takara
pLC4	Promoter probe vector with xylE as the reporter gene; Apr Cmr	25
pRN5101	Vector used for gene disruption; Apr Emr	12
pGS1217	pLC4 carrying the promoter region of B. subtilis sigI	This study
pGS1395	pLC4 carrying the promoter region and the N-terminal region of B. thuringiensis sigI	This study
pGS1564	pLC4 carrying the promoter region and the N-terminal region of B. licheniformis sigI	This study
pGS1604	pLC4 carrying the promoter region and the N-terminal region of Bacillus sp. strain NRRL B-14911 sigI	This study
pGS1651	pRN5101 carrying an internal region of the B. subtilis sigI gene	This study
pGS1671	pHY300PLK carrying the B. subtilis sigI gene	This study
pGS1722	pDL carrying the promoter region and the N-terminal region of B. subtilis mreBH	This study
pGS1740	pDL carrying the σI promoter region of B. subtilis bcrC (not including the σX, σM, and σV promoters)	This study
pGS1761	pHY300PLK carrying the B. subtilis sigV gene	This study

^aAp^r, ampicillin resistant; Cm^r, chloramphenicol resistant; Em^r, erythromycin resistant; Tc^r, tetracycline resistant.

^bATCC, American Type Culture Collection; ARSCC, Agricultural Research Service Culture Collection.

Construction of plasmids.

To construct the sigI-disruptive plasmid pGS1651, a 0.3-kb DNA fragment containing the internal region of sigI and flanked by HindIII and BamHI sites was amplified by PCR and cloned between HindIII and BamHI sites of the thermosensitive replicative plasmid pRN5101 (12).

Various DNA fragments containing the regulatory region plus a small portion of N terminus (or not) of the sigI gene from B. subtilis, B. thuringiensis, B. licheniformis, or B. sp. strain NRRL B-14911 were amplified by PCR and cloned individually between the EcoRI and HindIII (or BamHI) sites of pLC4 (25) to generate plasmids pGS1217, pGS1395, pGS1564, and pGS1604, respectively. The lengths of PCR-amplified DNA fragments in plasmids pGS1217, pGS1395, pGS1564, and pGS1604 are 0.32, 0.71, 0.41, and 0.42 kb, respectively.

To construct a plasmid that overproduces SigI or SigV in B. subtilis, 0.8-kb and 0.56-kb DNA fragments carrying the Shine-Dalgarno sequence plus the sigI and sigV coding regions, respectively, and flanked by EcoRI and HindIII sites were amplified by PCR and cloned between EcoRI and HindIII sites of pHY300PLK (Takara Shuzo Co.) to generate plasmids pGS1671 and pGS1761, respectively. The promoter of the tetracycline resistance gene present in pHY300PLK can drive sigI and sigV expression in B. subtilis.

To construct an mreBH promoter region-bgaB fusion, a 0.4-kb DNA fragment was amplified by PCR (using primers 5′-GCGGAATTCGATGTACCTCAATATATCG-3′ and 5′-GCGGGATCCACTGCGACAACAGA-3′) and cloned between EcoRI and BamHI sites of plasmid pDL (34) to generate plasmid pGS1722. To construct a bcrC promoter region-bgaB fusion containing only the σ^I promoter (and not the σ^X, σ^M, and σ^V promoters) of bcrC, a 0.25-kb DNA fragment was amplified by PCR (using primers 5′-GCGGGATCCTGTTGATCAAGTGACA-3′ and 5′-GCAGGTACCAGATTGTCTGAAATAACGG-3′) and cloned between BamHI and KpnI sites of plasmid pDL to generate plasmid pGS1740.

Construction of strains with a bgaB transcriptional fusion integrated at the amyE locus.

pDL-derived integrative plasmids pGS1722 and pGS1740 were individually introduced into B. subtilis cells by protoplast transformation. Chloramphenicol-resistant candidate integrants were spotted onto LB plates containing 0.2% soluble starch and screened by staining with iodine solution (0.5% I₂, 1% KI) as previously described (21) for a lack of halos due to loss of amylase activity. This could confirm that there was correct double-crossover integration at the amyE locus and generated strains BM1140 and BM1168, respectively.

Disruption of the chromosomal sigI gene.

Disruption of the chromosomal sigI gene of B. subtilis by integration of the thermosensitive replicative plasmid pRN5101-derived pGS1651 through a single crossover event was performed as previously described (12). B. subtilis BM1199 and BM1201 were constructed by introducing plasmid pGS1651 into BM1140 and BM1168, respectively, by the method of protoplast transformation. Transformants were first grown at the permissive temperature of 30°C and then transferred to the nonpermissive temperature of 39°C. Finally, integrants were selected on LB agar plates at 39°C for resistance to erythromycin. The correctness of integrants was verified by PCR.

RNA extraction and primer extension analysis.

Total RNA was extracted from B. subtilis cells grown in LB medium to an absorbance at 600 nm of 0.5 by a previously described method (36). The transcriptional start site was determined by primer extension as previously described (17) using synthetic oligonucleotides 5′-ACTACTCATTAATGATAGCT-3′ for B. subtilis sigI, 5′-ATAGCTAATTTGATAGGGAT-3′ for B. licheniformis sigI, 5′-CCCACTTAATCTAGTAAATT-3′ for B. thuringiensis sigI, 5′-CACCTCTCCACCAGCAC-3′ for Bacillus sp. strain NRRL B-14911 sigI, 5′-GTTCAAATAATCACCTTTTAC-3′ for B. subtilis bcrC, and 5′-GCTGTTCCTAAGTCAATC-3′ for B. subtilis mreBH.

β-Galactosidase activity assay.

Activity of the thermostable β-galactosidase BgaB was measured as previously described (29) but with some modifications. A portion (at least 2 ml) of the bacterial culture was pelleted and resuspended in an equivalent volume of chilled BgaB buffer (25 mM potassium phosphate [pH 6.4], 50 mM KCl, and 1 mM MgSO₄). The absorbance of the resuspended cells was measured at 600 nm. For cell permeabilization, 50 μl of 0.1% sodium dodecyl sulfate and 100 μl of chloroform were added to 1 ml of resuspended cells. Permeabilization was achieved by vortexing. The reaction was then initiated by adding 0.2 ml of o-nitrophenyl-β-d-galactopyranoside at a concentration of 4 mg per ml. After incubation at 55°C for 30 min, 0.5 ml of 1 M Na₂CO₃ was added to stop the reaction. The reaction mixture was centrifuged to remove cell debris and chloroform. Absorbance of the supernatant was recorded at 420 nm and 550 nm. BgaB activity is given in Miller units (19).

Other methods.

Genomic DNAs from a variety of Bacillus species were isolated as previously described (24). Transformation of B. subtilis cells by the protoplast method was carried out as previously described (11). Assays of catechol 2,3-dioxygenase (XylE) activity were performed according to an established method (25). Protein concentrations were determined by the bicinchoninic acid protein assay method according to the instructions of the manufacturer (Pierce Biotechnology, Inc.) with bovine serum albumin as the standard.

RESULTS

Effects of overexpression of B. subtilis sigI on expression of transcriptional fusions of sigI regulatory regions from various Bacillus species to an xylE reporter gene in B. subtilis.

In an attempt to define a consensus sequence for the σ^I promoter and to identify target genes for the sigma factor SigI of B. subtilis, we adopted a different approach. As mentioned above, the B. subtilis sigI gene and its downstream rsgI gene, which encodes an anti-σ^I factor, are cotranscribed and constitute an operon. Expression of the sigI operon is subject to positive autoregulation by SigI (1). Since a similar genetic organization of the sigI operon exists in B. licheniformis ATCC 14580, B. thuringiensis serovar Israelensis ATCC 35646, and Bacillus sp. strain NRRL B-14911, it is possible that a positive autoregulatory mechanism for expression of the sigI operons is also conserved in these Bacillus species. Since overall amino acid sequence similarities and identities between the B. subtilis SigI and SigI from the other three Bacillus species range from 94 to 70% and 83 to 48%, respectively, it is possible that the σ^I promoter sequence that can be recognized by SigI from these Bacillus species is also conserved.

To examine whether the B. subtilis SigI could recognize and activate the σ^I promoters of sigI genes from various Bacillus species, we used a binary-vector system in B. subtilis. In this system, one vector (plasmid pGS1671) constitutively produced the B. subtilis SigI and the other vector carried the transcriptional fusion of the regulatory region of sigI from one of these Bacillus species to the xylE reporter gene. As shown in Table 2, the specific activity of XylE from each of the B. subtilis strains carrying the sigI-overexpressing plasmid was significantly higher than that observed in the corresponding B. subtilis strains carrying the control vector pHY300PLK. In a control experiment, a DNA fragment showing no promoter activity was fused to xylE. The result revealed that sigI overexpression had no effect on XylE activity in this B. subtilis strain (data not shown). Taken together, these results suggest that the B. subtilis SigI can specifically recognize and activate the σ^I promoters of sigI genes from these Bacillus species.

TABLE 2.

Effects of overexpression of B. subtilis sigI on expression of various promoter-xylE transcriptional fusions in a binary-vector system of B. subtilis

Plasmid 1a	Plasmid 2	XylE sp act (mU/mg)b
pGS1671 (B. subtilis sigI-expressing plasmid)	pGS1217 (carrying B. subtilis sigI promoter region-xylE fusion)	292.1 ± 23.7
pHY300PLK (control vector)	pGS1217	144.6 ± 15.1
pGS1671	pGS1564 (carrying B. licheniformis sigI promoter region-xylE fusion)	249.4 ± 24.6
pHY300PLK	pGS1564	93.5 ± 8.2
pGS1671	pGS1395 (carrying B. thuringiensis sigI promoter region-xylE fusion)	157.3 ± 16.9
pHY300PLK	pGS1395	26.8 ± 4.5
pGS1671	pGS1604 (carrying Bacillus sp. strain NRRL B-14911 sigI promoter region-xylE fusion)	488.1 ± 47.6
pHY300PLK	pGS1604	91.2 ± 9.4

^asigI was overexpressed from the pHY300PLK-based plasmid pGS1671.

^bThe specific activity of XylE was measured with crude extracts prepared from B. subtilis cells carrying a binary-vector system and grown at 37°C to an absorbance at 600 nm of 1.0. Each value is the mean ± standard error from three independent experiments.

Mapping of transcriptional initiation sites of σ^I promoters of sigI genes from various Bacillus species in B. subtilis.

We next used various synthetic oligonucleotides as primers to map transcriptional initiation sites of σ^I promoters of sigI genes from various Bacillus species by primer extension analyses. RNAs were isolated from B. subtilis cells harboring the above-mentioned binary vectors. As shown in Fig. 1, a major SigI-responsive extension product could be detected for each sigI gene from these Bacillus species. The transcriptional initiation site of the σ^I promoter of each sigI gene could thus be located by comparison with each corresponding DNA sequencing ladder (Fig. 1 and Table 3).

FIG. 1.

Primer extension analysis of transcriptional initiation sites of σ^I promoters of sigI from various Bacillus species in B. subtilis. (A) B. subtilis sigI. (B) B. licheniformis sigI. (C) B. thuringiensis sigI. (D) Bacillus sp. strain NRRL B-14911 sigI. Total RNA was isolated from B. subtilis cells carrying binary vectors. Lanes 1, extension product for B. subtilis cells carrying the control vector pHY300PLK plus a plasmid containing a sigI promoter region-xylE fusion; lanes 2, extension product for B. subtilis cells carrying the sigI-overexpressing plasmid pGS1671 plus a plasmid containing a sigI promoter region-xylE fusion. Dideoxy sequencing ladders obtained with the same primer used for each primer extension analysis are shown in the first four lanes (G, A, T, and C) of each panel. The sequences shown are complementary to those read from ladders. Each arrow indicates a transcriptional initiation site.

TABLE 3.

Alignment of SigI-dependent promoters

Gene (species)	Promoter regiona	5′ Untranslated region (nucleotides)
sigI (B. subtilis)	acgcataaaACCCCCttAAttctttagaaaggcaCGAAatcatgtataGaacg	116
sigI (B. licheniformis)	ggcgggaaaACCCCCttAAtcgtcaaacagatcaCGAAttgttctaggAgatc	115
sigI (B. thuringiensis)	tattttttgACCCCCatAAaactatgtattcctcCGAAtatgtatagtgAaga	376
sigI (Bacillus sp. strain NRRL B-14911)	caccaaaccACCCCCaaAAgctcctctttccgggCGAAgctttatgttaGcag	95
bcrC (B. subtilis)	gcacagaatCCCCCCagAAaccgcgattcctcttCGAAttctcttcaAgcgcc	92
mreBH (B. subtilis)	ccgacgcacACCCCCaaAAatcgcagtatttctgAGAAactttaaatgTagaa	39
Consensus	---------ACCCCC--AA---------------CGAA----------

^aThe −35 region, two consecutive A nucleotides downstream of the −35 region, and the −10 region (from left to right) are indicated in bold uppercase letters. The identified transcriptional initiation sites (+1) are denoted by underlined bold uppercase letters.

Alignment of putative σ^I promoter sequences of sigI genes from various Bacillus species.

We next made an alignment of the nucleotide sequences upstream of these identified transcriptional initiation sites in order to deduce a putative consensus sequence for the σ^I promoter. Since SigI is an alternative sigma factor of the σ⁷⁰ family, σ^I promoters are thought to contain −35 and −10 consensus elements (22). As shown in Table 3, the putative consensus sequence for the σ^I promoter is ACCCCC for the −35 region and CGAA for the −10 region. The spacing between the −35 and −10 regions is 19 nucleotides. It is interesting to note that there is a conserved sequence “AA” within the spacer and close to the −35 region.

Computer search for B. subtilis genes preceded by putative σ^I promoter elements.

Identification of target genes in the B. subtilis genome by using a consensus sequence-directed computer search has been reported for the B. subtilis sigma factor SigX (15). In this study, we used the consensus sequence of the putative σ^I promoter elements deduced from the analyses described above to search the SubtiList database (http://genolist.pasteur.fr/SubtiList/) for possible SigI target genes. The search pattern we used was ACCCCC-N₂-AA-N₁₅-CGAA (N represents any base), where one mismatch was allowed in either ACCCCC or CGAA. In an initial search, the search regions were arbitrarily restricted to being within 200 bp upstream of a predicted gene. This search identified bcrC (formerly ywoA; bacitracin-resistant protein), lytE (cell wall hydrolase), mreBH (cell shape-determining protein), ysfD (possible glycolate oxidase subunit), and ytpS (possible DNA translocase) as potential target genes.

Effects of overexpression of B. subtilis sigI on expression of candidate target genes.

To test whether sigI overexpression could stimulate expression of these candidate target genes, PCR-amplified DNA fragments containing putative σ^I promoter elements of these candidate genes were individually cloned into the promoter probe vector pDL (34), which was designed to create transcriptional fusions to bgaB (encoding a thermostable β-galactosidase). It was previously shown that the bcrC gene could be directly regulated by three extracytoplasmic function sigma factors: SigX, SigM, and SigV (4, 18, 35). Therefore, the DNA fragment used in this report did not include the σ^X, σ^M, and σ^V promoter regions so as to avoid possible interference. The generated plasmids carrying bgaB fusions were individually introduced into B. subtilis and integrated into the chromosome by a double crossover at the amyE locus. The pHY300PLK-based plasmid pGS1671 was then introduced into these strains to constitutively overexpress SigI. The results showed that sigI overexpression could stimulate expression of both the bcrC promoter region-bgaB fusion and the mreBH promoter region-bgaB fusion (Fig. 2A and B) but could not stimulate expression of a lytE, ysfD, or ytpS promoter region-bgaB fusioin at the amyE locus (data not shown). In a control experiment, we used B. subtilis strains carrying either the bcrC promoter region-bgaB fusion or the mreBH promoter region-bgaB fusion at the amyE locus plus the pHY300PLK-based plasmid pGS1761, which constitutively overexpressed the sigma factor SigV (32). It turned out that sigV overexpression could not increase BgaB activity in these strains (data not shown). Together, these results suggest that SigI can specifically stimulate bcrC and mreBH expression and that the lytE, ysfD, and ytpS genes are not SigI targets.

FIG. 2.

Effects of overexpression of B. subtilis sigI on expression of bcrC and mreBH promoter region-bgaB transcriptional fusions. (A) Expression of a bcrC promoter region-bgaB fusion at the amyE locus of BM1168 carrying the sigI-overexpressing plasmid pGS1671 (filled circles) or the control vector pHY300PLK (filled squares). (B) Expression of an mreBH promoter region-bgaB fusion at the amyE locus of BM1140 carrying the sigI-overexpressing plasmid pGS1671 (filled circles) or the control vector pHY300PLK (filled squares). Cells were grown in LB medium for various times as indicated. Growth was monitored by measuring optical density at 600 nm (OD₆₀₀) and is indicated by open symbols. The values shown are means from two independent experiments. Individual values did not differ by more than 15% from the means.

SigI-dependent heat induction of bcrC and mreBH.

It was previously shown that expression of the B. subtilis sigI gene was heat inducible (37). This prompted us to test whether expression of SigI target genes could also be induced by heat. We used the chromosomal bgaB transcriptional fusions at the amyE locus as described above to evaluate the heat inducibility of these target genes. As shown in Fig. 3A and B, both bcrC and mreBH expression could be induced by heat. To investigate whether heat induction of bcrC and mreBH was SigI dependent, we disrupted the chromosomal sigI gene in these strains. The result showed that sigI disruption almost totally abolished heat induction of the bcrC promoter region-bgaB fusion and the mreBH promoter region-bgaB fusion (Fig. 3A and 3B). This indicates that heat induction of bcrC and mreBH can be SigI dependent.

FIG. 3.

SigI-dependent heat induction of a bcrC promoter region-bgaB fusion and an mreBH promoter region-bgaB fusion. Cells were grown at 37°C to an absorbance at 600 nm of 0.3 and then divided into two parts. One part remained at 37°C, and the other was transferred to 51°C at time zero. The cells were grown for various times as indicated. (A) BgaB activity from the bcrC promoter region-bgaB fusion was measured in the wild-type strain BM1168 at 37°C (open circles) or 51°C (filled circles) or in the sigI-disrupted strain BM1201 at 37°C (open triangles) or 51°C (filled triangles). (B) BgaB activity from the mreBH promoter region-bgaB fusion was measured in the wild-type strain BM1140 at 37°C (open circles) or 51°C (filled circles) or in the sigI-disrupted strain BM1199 at 37°C (open triangles) or 51°C (filled triangles). Growth was monitored by measuring optical density at 600 nm (OD₆₀₀) and is indicated by dashed lines. The values shown are means from two independent experiments. Individual values did not differ by more than 15% from the means.

Mapping of transcriptional initiation sites of σ^I promoters of the bcrC and mreBH genes.

To further demonstrate that SigI could regulate bcrC and mreBH expression directly, we performed primer extension analysis to determine the transcriptional start sites of σ^I promoters of the bcrC and mreBH genes. RNAs were isolated from B. subtilis cells carrying a bcrC promoter region-bgaB fusion or mreBH promoter region-bgaB fusion at the amyE locus plus the sigI-overexpressing plasmid pGS1671 or the control vector pHY300PLK. Figure 4A and B show that a major SigI-responsive extension product could be detected for the bcrC and mreBH genes. The corresponding transcriptional start sites were at appropriate distances from the σ^I promoter elements of bcrC and mreBH (Table 3). This further supports the notion that SigI can directly regulate bcrC and mreBH expression.

FIG. 4.

Primer extension analysis of transcriptional initiation sites of σ^I promoters of the bcrC and mreBH genes. (A) Extension product for B. subtilis cells carrying a bcrC regulatory region-bgaB fusion at the amyE locus plus the control vector pHY300PLK (lane 1) or the sigI-overexpressing plasmid pGS1671 (lane 2). (B) Extension product for B. subtilis cells carrying an mreBH promoter region-bgaB fusion at the amyE locus plus the control vector pHY300PLK (lane 1) or the sigI-overexpressing plasmid pGS1671 (lane 2). Dideoxy sequencing ladders obtained with the same primer used for each primer extension analysis are shown in the first four lanes (G, A, T, and C) of each panel. The sequences shown are complementary to those read from ladders. Each arrow indicates a transcriptional initiation site.

DISCUSSION

When the first paper proposing that the B. subtilis ykoZ gene encodes a putative sigma factor, SigI, was published in 2001, the authors mentioned that DNA microarray experiments were in progress so as to define the SigI-dependent regulon (37). However, no further microarray data have been reported since then. To date, no SigI target gene except sigI itself has been identified (1). In this report, we have successfully used a different approach to demonstrate that the bcrC and mreBH genes are new members of SigI regulon.

The B. subtilis bcrC gene encodes an undecaprenyl pyrophosphate phosphatase that is important for cell wall biosynthesis (2) and for bacitracin resistance (20). Bacitracin is an antibiotic that binds to undecaprenyl pyrophosphate and prevents its dephosphorylation into undecaprenyl phosphate. Undecaprenyl phosphate is a lipid carrier that is required for cell wall biosynthesis (31). The B. subtilis mreBH gene encodes a bacterial homolog of actin that is important for cell morphogenesis. An mreBH mutant exhibits a cell wall-related defect (10). The critical role of B. subtilis SigI in regulation of bcrC and mreBH implies that one function of SigI is related to the maintenance of cell envelope integrity and homeostasis during heat stress. It was previously observed that a B. subtilis sigI mutant (BFA 251) exhibited a temperature-sensitive phenotype (37). The B. subtilis sigI mutant that we constructed exhibited a similar phenotype. However, this phenotype was not observed in either the bcrC or the mreBH mutant (data not shown), suggesting that another, not-yet-identified SigI-dependent gene(s) may contribute to heat resistance.

In addition to the initial computer search in which the search region was limited to 200 bp upstream of a predicted gene, we also carried out a more extensive search by not restricting the search region to 200 bp upstream of a predicted gene. We also did additional computer searches by allowing one mismatch in the −35 region, one mismatch in the −10 region, and one or two mismatches in the highly conserved AA sequence simultaneously. Since it is unusual that there is a string of five consecutive Cs in the consensus sequence of the −35 region of the putative σ^I promoter, we selected those genes containing at least four consecutive Cs in the −35 region as potential candidate genes. By using these criteria, we were able to find more potential target genes, including the coxA (spore cortex protein), dnaJ (heat shock protein), mbl (MreB-like protein), metA (homoserine O-succinyltransferase), rapK (response regulator aspartate phosphatase), sacX (negative regulatory protein of SacY), and ywjE (possible cardiolipin synthetase) genes. We tested each of them and found that sigI overexpression could not stimulate their expression (data not shown).

The B. subtilis bcrC gene was known to be directly regulated by three extracytoplasmic function sigma factors: SigX, SigM, and SigV (4, 18, 35). It is now clear that SigI is the fourth sigma factor that can directly regulate bcrC expression. Involvement of multiple sigma factors in controlling bcrC expression allows bcrC to respond to various environmental stresses through different pathways. For examples, bacitracin induction of bcrC expression is SigM dependent but not SigX dependent (4), and heat induction of bcrC can be mediated through SigI, as revealed in this report.

Several attempts to carry out in vitro runoff transcription experiments by using purified His-tagged SigI and His-tagged core RNA polymerase have been unsuccessful. Since overexpression of sigI in Escherichia coli produced only insoluble SigI protein in inclusion bodies, we tried several different methods to refold denatured SigI protein, including (i) dialysis of a denatured SigI-containing solution against buffers, (ii) gradual dilution of a denatured SigI-containing solution with buffers, and (iii) on-column refolding of denatured SigI by using a descending concentration gradient of urea. Unfortunately, the SigI proteins prepared by these methods could not work in the in vitro transcription experiments. Several possibilities may explain the failure of these attempts: (i) a lack of some general or specific chaperone(s) that might be required in vivo to help folding of SigI to a biologically active form, (ii) a lack of an unidentified protein that might be required in vivo to facilitate the association of SigI with core RNA polymerase, or (iii) a lack of an unknown protein that might be required in vivo for the σ^I holoenzyme to bind specifically to its promoter DNA. Nevertheless, this question remains to be clarified.