DNA microarray analysis of Bacillus subtilis sigma factors of extracytoplasmic function family


  • Kei Asai,

    1. Department of Biochemistry and Molecular Biology, Faculty of Science, Saitama University, Saitama, 338-8570 Saitama, Japan
    Search for more papers by this author
  • Hirotake Yamaguchi,

    1. Department of Biotechnology, Faculty of Engineering, Science, Fukuyama University, Fukuyama, Hiroshima 729-0292, Japan
    Search for more papers by this author
  • Choong-Min Kang,

    1. Department of Biotechnology, Faculty of Engineering, Science, Fukuyama University, Fukuyama, Hiroshima 729-0292, Japan
    Search for more papers by this author
    • 1

      Faculty of Life Science and Engineering, Youngdong University, Youngdong, South Korea.

  • Ken-ichi Yoshida,

    1. Department of Biotechnology, Faculty of Engineering, Science, Fukuyama University, Fukuyama, Hiroshima 729-0292, Japan
    Search for more papers by this author
  • Yasutaro Fujita,

    1. Department of Biotechnology, Faculty of Engineering, Science, Fukuyama University, Fukuyama, Hiroshima 729-0292, Japan
    Search for more papers by this author
  • Yoshito Sadaie

    Corresponding author
    1. Department of Biochemistry and Molecular Biology, Faculty of Science, Saitama University, Saitama, 338-8570 Saitama, Japan
      *Corresponding author. Tel.: +81 (48) 858 3399; Fax: +81 (48) 858 3384, E-mail address: ysadaie@molbiol.saitama-u.ac.jp
    Search for more papers by this author

*Corresponding author. Tel.: +81 (48) 858 3399; Fax: +81 (48) 858 3384, E-mail address: ysadaie@molbiol.saitama-u.ac.jp


Target gene candidates of the seven extracytoplasmic function (ECF) sigma factors of Bacillus subtilis have been surveyed using DNA microarray analysis of mRNA extracted from cells grown in Luria–Bertani broth, in which an ECF sigma factor gene was placed under the control of the spac promoter on multicopy plasmid pDG148 and overexpressed. The number of target candidates for each of the sigma factors varied greatly, and a total of 278 genes were selected. Interestingly, the above target gene candidates shared only one gene out of 94 target genes of the general stress sigma B that have been reported in the literature thus far. Furthermore, lacZ-fusion experiments based on the results of DNA microarray analysis indicated that each ECF sigma factor directs transcription of its own operon, with the exception of sigZ. The DNA microarray data collected in this study are available at the KEGG Expression Database web site (http://www.genome.ad.jp/kegg/expression/).


The complete sequencing of the entire Bacillus subtilis genome has revealed the presence of seven genes encoding sigma factors of extracytoplasmic function (ECF) family[1], which include SigM, SigV, SigW, SigX, SigY, SigZ, and YlaC [2–9]. They possess a common domain characteristic to a family of sigmas, that are generally considered to direct the transcription of genes responding to external signals, such as the denaturation of proteins or oxygen, salts, heat, and cold stresses [10,11]. In fact, some of the ECF family sigmas of B. subtilis are required in the presence of alkaline, salt, and heat stresses [2,12,13]. Furthermore, although some of the target genes of SigW and SigX are known [4,12,14–17] and SigM, SigW, SigX direct transcription of their own operons [2,4,16], the target genes of the other sigmas are still unknown.

DNA array technology has been a powerful tool for the transcriptome analysis of several microorganisms whose whole genome sequences have been determined. This method was also applied to the analysis of the global regulons of B. subtilis transcriptional regulators [18–28]. The DNA microarray analysis adopted in the studies to find target genes of all two-component regulatory systems of B. subtilis[29] involved an overproduction system of each of the regulatory proteins where the corresponding gene was placed downstream of the IPTG inducible spac promoter on a multicopy plasmid, pDG148 [30,31]. The genes, whose expression was affected by overproduction of a response regulator, were considered to be the target gene candidates.

In this study, we used the above DNA microarray analysis involving the overexpression system to survey the candidates for the target genes of seven ECF sigma factors of B. subtilis, because they are typical positive DNA-binding regulators for most of which stress signals are still unknown. We took strategies to mimic stress response by enforced expression of ECF sigma genes.

2Materials and methods

2.1Bacterial strains, plasmids, DNA manipulation, and culture media

All bacterial strains and plasmids used in this study are described in Table 1. DNA manipulation, PCR, media, reagents and enzymes have been described elsewhere[32].

Table 1.  Bacterial strains and plasmids used
Strain [reference]GenotypeSequences of primer pairs used for construction [forward, reverse (5′ to 3′)]
B. subtilis
E. coli
C600[32]thi-1 thr-1 leuB6 lacY1 tonA21 supE44 

2.2Construction of the plasmids carrying each of the ECF sigma factor genes

Derivatives of plasmid pDG148 carrying ECF sigma genes were constructed as follows. The DNA region containing the structure gene of each ECF sigma and its Shine–Dalgarno sequence was amplified by PCR using chromosomal DNA from strain 168 and the primer pairs listed in Table 1. Amplified DNA fragments were digested with Sal I and Sph I, and cloned into plasmid pDG148[31]. The DNA sample was used to transform Escherichia coli C600 to ampicillin (50 μg ml−1)-resistance to obtain derivatives of plasmid pDG148 (Table 1). After the nucleotide sequences in the cloned DNA regions were confirmed by sequence determination, wild-type B. subtilis strain 168 was transformed to kanamycin (5 μg ml−1)-resistance with plasmids carrying each of the ECF sigma genes.

2.3Construction of the B. subtilis strain carrying the lacZ gene fused to the promoter of each of the ECF sigma operons

A portion of the promoter region of the ECF sigma operon was PCR synthesized with a primer pair described in Table 1. The PCR product was digested with Eco RI and Bam HI, and cloned into plasmid pDL2[33]. Plasmid DNAs were used to transform B. subtilis strain 168 to chloramphenicol (5 μg ml−1)-resistance to obtain strains carrying lacZ-fusion with the promoter of each of the ECF sigma factor operons at the amyE locus (Table 1).

2.4Disruption of the sigW and sigX genes

Disruption of the sigW and sigX genes with the integrative plasmid pMUTIN was carried out as described elsewhere[34]. The primer pairs used for PCR synthesis of sigW and sigX are described in Table 1. Activity of β-galactosidase (β-Gal) was measured as described elsewhere[35].

2.5Cell growth and RNA isolation

The cells of B. subtilis strain carrying a derivative of plasmid pDG148 harboring each ECF sigma factor gene, were grown in LB medium at 37°C to early exponential phase of growth (A600= 0.1). IPTG was added at a concentration of 1 mM, and the cells were harvested after 2 h incubation. Total RNA was isolated from the cells by procedures previously described elsewhere[36].

2.6DNA microarray techniques

Fluorescently labeled cDNA probe used for hybridization to DNA microarray were prepared by a two-step procedure, and the hybridization and DNA microarray analysis were performed as previously described[36]. The microarrays used contained 39 calf thymus DNA spots as well as 4004 protein genes in the B. subtilis 168 trpC2 genome[1]. The mean Cy3 and Cy5 fluorescence intensities for each spot were calculated with the background taken as the average of the intensities of negative controls of 39 calf thymus DNA spots. After subtracting the background from all the intensities of the B. subtilis gene spots, and their normalization using total Cy3 and Cy5 intensities, we calculated the expression ratios. Protein genes which were upregulated more than three-fold were collected. To obtain reliable ratios, we ignored the spots of the intensities used as numerators for this calculation, which were less than the background, and replaced the intensities used as denominators with the standard deviation of the average intensity of the negative controls if they were less than it.

3Results and discussion

3.1Detection of protein genes which showed elevated transcription by overexpression of ECF sigmas

To find protein genes whose transcription might be directed by seven ECF sigma factors, cells carrying a plasmid pDG148 derivative with each sigma factor gene were grown for 2 h in the presence or absence of IPTG, and their RNAs were subjected to DNA microarray analysis. In this analysis, we picked out only those genes which had been upregulated more than three-fold upon the addition of IPTG.

As listed in Table 2, overexpression of sigM, sigV, sigW, sigX, sigY, sigZ, and ylaC upregulated 50, 98, 183, 10, 17, 3, and 6 genes respectively. These numbers include the ECF sigma genes themselves that were overexpressed. Seven SigM dependent genes induced by vancomycin[26] were found among 50 upregulated genes upon sigM overexpression. The 183 genes upregulated upon sigW overexpression included 46 of 61 target genes of σW described earlier [12,15], while overexpression of σX induced only ycgR among proposed target genes of σX[16]. Overexpression of sigZ upregulated only gsiB (a target of sigB)[22] and yrpG. SigZ-dependent transcription of yrpG was confirmed by β-Gal assay using lacZ-fusions (data not shown).

Table 2.  The target gene candidates of the ECF sigma factors of B. subtilisa
  1. aRNA preparation and DNA microarray experiments were repeated until the best-quality two-hybridization profiles were obtained. The expression ratio for each gene was taken as the average of two ratios obtained from the above hybridization profiles. The target gene candidates exhibited greater than three-fold are listed in the order of location on the chromosome. Each number indicates expression ratio.

  2. bThe underlined genes are the target genes of SigW proposed[15], whereas those with asterisks are target gene candidates of SigW revealed from DNA array analysis of the responding genes to alkali shock[12]. The genes with # are target genes of SigW induced by vancomycin[26].

SigM rpoC3.1yrrN4.0rplB6.4hemE3.5mreC4.5yngC3.6
mreB3.8ctaA3.2  ydcC12.9dinG4.4dltC6.0ylaD10.8
radC5.6murD3.1SigWb ydcF3.9panB3.3dltD12.0  
SigV yrhK10.2fusA3.3treA4.4yrrN4.9SigX   

It is remarkable that 278 protein genes upregulated by overexpression of seven ECF sigmas of B. subtilis described in this study have only one common gene, gsiB, among 94 target gene candidates of the general stress sigma, SigB of B. subtilis[22].

Overproduction of sigma protein might lead to a kind of stress, however, only ywoA and ycgR were commonly upregulated by only SigM, SigV, SigW and SigX overexpression. In addition, only one target gene of the general stress sigma B was upregulated by sigZ overexpression. Therefore it might not be possible that overexpression of ECF sigma gene causes serious stress.

3.2Positive autoregulation of the ECF sigma factor operons

Transcription of the presumed sigY-yxlC-yxlD-yxlE-yxlF-yxlG operon was upregulated by overexpression of the sigY gene (Table 2), and similar tendencies were also observed in the presumed sigM-yhdL-yhdK, sigV-yrhM, sigW-ybbM, sigX-rsiX and ylaA-ylaB-ylaC-ylaD operons. These DNA microarray data suggest that the ECF sigma operons might be transcribed by their own ECF sigma factor except for sigZ, which appears to constitute monocistronic operon.

In order to confirm the above DNA microarray results, we constructed B. subtilis strains carrying a lacZ-fusion with the promoter region of each of the presumed ECF sigma factor operons at the amyE locus (Table 1), and determined whether transcription of each of the ECF sigma operons was enhanced upon overexpression of the corresponding ECF sigma gene by measuring β-Gal. As for ylaC, the promoter region of ylaA was used, because ylaC is the middle gene of the presumed ylaA-ylaB-ylaC-ylaD operon.

As shown in Table 3, the addition of IPTG clearly induced β-Gal synthesis in the cells bearing plasmid pDG148 derivatives with sigM, sigV, sigY and ylaC, and carrying lacZ-fusions with their promoters, indicating that their transcription was positively regulated upon overexpression of these ECF sigma factors. However, β-Gal induction upon IPTG addition in the cells carrying the lacZ-fusions with those of the sigW and sigX operons was not so clear due to high background (-IPTG) (Table 3). To reduce background we constructed strains which carried disrupted sigW or sigX gene with integrative plasmid pMUTIN. The resultant strain carried lacZ gene downstream of sigW or sigX promoter[34]. Experiments with overexpression of sigW or sigX in the strains carrying disrupted sigW or sigX on their chromosomes clearly showed the induction of transcription from sigW or sigX promoters by SigW or SigX (Table 4). In the case of sigW, introduction of the plasmid carrying sigW induced expression of sigW operon even in the absence of overexpression of sigW (-IPTG). The magnitude of induction was low in both cases probably because of induction of anti-sigma genes, ybbM and rsiX located downstream of disrupted sigW and sigX genes respectively.

Table 3.  Monitor of transcription of the ECF sigma factor operons of B. subtilis using lacZ-fusionsa
  1. aThe cells of B. subtilis strains bearing a derivative of plasmid pDG148 harboring each ECF sigma factor gene and carrying a lacZ-fusion with the promoter of its own operon were grown at 37°C to early exponential phase of growth (A600= 0.1). After they were further incubated with and without 1 mM IPTG for 90 min, the cells were harvested, and the β-Gal activities were measured as described previously[35].

ExperimentOperonβ-Gal activity (Miller unit)
  +Ratio ±
Table 4.  Monitor of transcription of the sigW and sigX operons of B. subtilis using lacZ-fusionsa
  1. aThe cells of B. subtilis strain carrying disrupted sigW or sigX genes with pMUTIN and bearing a pDG148 derivative harboring sigW or sigX genes were grown at 37°C to early exponential phase of growth. Then they received increasing amounts of IPTG for 90 min. The β-Gal activity was measured as described elsewhere[35].

Strainβ-Gal activity (Miller unit)
 IPTG (μM)
BSU38 (sigW::pMUTIN)
− pDG148-sigW46.849.449.746.546.048.047.8
+ pDG148-sigW118.0132.9151.7157.9160.1149.2137.6
− pDG148-sigX61.968.
+ pDG148-sigX50.648.357.181.892.196.396.2

These results coincide with the previous findings that the sigW and sigX operons as well as the sigM operon possess the promoters recognized by their own sigma factors [2,4,16]. In contrast, IPTG addition did not induce β-Gal synthesis in the case of sigZ. These results together with the DNA microarray data indicated that the ECF sigma factor operons of B. subtilis are transcribed by their own sigma factors, except for sigZ. This is consistent with a general feature of ECF sigma factors. ECF sigma gene, in general, constitutes an operon with anti-sigma factor gene and directs transcription from its own promoter[11].

Our current strategy was designed to detect only such genes which responded solely to sigma factors overproduced. Some known target genes of SigW or SigX which were not detected in this study might require regulatory factors other than overexpressed sigmas. Moreover, it is apparent that the upregulated genes listed in Table 2 not only include direct target genes of seven ECF sigma factors, but also indirect target genes, the expression of which is a consequence of alterating the expression of the direct target genes.


We are grateful to I. Kiuchi, K. Matsumoto, T. Matsumoto, M. Matsunaga, K. Nakanishi, J. Takahashi, and S. Yamaki for technical assistance. This work was supported by a Grant-in-Aid for Scientific Research on Priority Area from the Ministry of Education, Science, Sports and Culture of Japan.