Skip to main content

Prediction and validation of novel SigB regulon members in Bacillus subtilis and regulon structure comparison to Bacillales members



Sigma factor B (SigB) is the central regulator of the general stress response in Bacillus subtilis and regulates a group of genes in response to various stressors, known as the SigB regulon members. Genes that are directly regulated by SigB contain a promotor binding motif (PBM) with a previously identified consensus sequence.


In this study, refined SigB PBMs were derived and different spacer compositions and lengths (N12-N17) were taken into account. These were used to identify putative SigB-regulated genes in the B. subtilis genome, revealing 255 genes: 99 had been described in the literature and 156 genes were newly identified, increasing the number of SigB putative regulon members (with and without a SigB PBM) to > 500 in B. subtilis. The 255 genes were assigned to five categories (I-V) based on their similarity to the original SigB consensus sequences. The functionalities of selected representatives per category were assessed using promoter-reporter fusions in wt and ΔsigB mutants upon exposure to heat, ethanol, and salt stress. The activity of the PrsbV (I) positive control was induced upon exposure to all three stressors. PytoQ (II) showed SigB-dependent activity only upon exposure to ethanol, whereas PpucI (II) with a N17 spacer and PylaL (III) with a N16 spacer showed mild induction regardless of heat/ethanol/salt stress. PywzA (III) and PyaaI (IV) displayed ethanol-specific SigB-dependent activities despite a lower-level conserved − 10 binding motif. PgtaB (V) was SigB-induced under ethanol and salt stress while lacking a conserved − 10 binding region. The activities of PygaO and PykaA (III) did not show evident changes under the conditions tested despite having a SigB PBM that highly resembled the consensus. The identified extended SigB regulon candidates in B. subtilis are mainly involved in coping with stress but are also engaged in other cellular processes. Orthologs of SigB regulon candidates with SigB PBMs were identified in other Bacillales genomes, but not all showed a SigB PBM. Additionally, genes involved in the integration of stress signals to activate SigB were predicted in these genomes, indicating that SigB signaling and regulon genes are species-specific.


The entire SigB regulatory network is sophisticated and not yet fully understood even for the well-characterized organism B. subtilis 168. Knowledge and information gained in this study can be used in further SigB studies to uncover a complete picture of the role of SigB in B. subtilis and other species.

Peer Review reports


The general stress response (GSR) in bacteria constitutes a vital trait for cells to adapt to and survive conditions such as temperature change, nutrient depletion, or exposure to reactive oxygen species in natural niches.

The GSR in Bacillus subtilis and many Bacillales members is under the transcriptional control of the alternative sigma factor B (σB or SigB) [1]. Different environmental or nutritional signals, such as heat, ethanol, salt, and glucose starvation can induce the SigB-mediated GSR, resulting in the expression of SigB-dependent genes and the production of proteins to protect cells from injuries [2,3,4]. The production of these proteins provides general protection to the cells and confers resistance to multiple stresses, allowing for rapid adaption to changing environments, thereby enhancing the survival of vegetative cells in extreme habitats [5].

The genes/proteins with SigB-dependent expression are defined as the SigB regulon members. Those that are directly regulated by SigB contain a promoter binding motif (PBM) consensus sequence GTTTAA-N15 (± 2 bp) – GGGTAT [6,7,8]. Those that are indirectly controlled by SigB do not have a SigB PBM and can either be controlled via SigB-dependent genes/proteins or regulated by other transcriptional regulators [8]. To date, 224 genes have been listed as members of the SigB regulon on Subtiwiki for B. subtilis [2, 3, 5, 9,10,11,12,13]. A recent study by Vohradsky et al. [8] further expanded the number to 411, with around 30% of the SigB regulon genes reported to lack a SigB PBM, and around 60–95% predicted to contain only a partial SigB PBM (i.e., only the − 35 or the − 10 promoter region, upstream of the AUG start codon).

Over time, more and more SigB regulon genes have been identified. This indicates that SigB regulatory networks are sophisticated and heavily interlinked with various other cellular mechanisms that are regulated by other transcriptional regulators. For example, SigB is indirectly involved in cellular responses such as sporulation and biofilm formation in B. subtilis [14, 15]. SigB-dependent genes may be expressed selectively to serve particular functions under certain conditions, e.g., 36% of the 411 experimentally confirmed and predicted genes are expressed during spore germination and outgrowth in B. subtilis [8], and roughly half are needed to cope with physical stresses such as exposure to ethanol, butanol, salt, or high/low temperatures [5]. In B. subtilis, distinct sets of SigB-dependent genes are expressed under different environmental conditions in nature. This may hold true as well for other Bacillus species like Bacillus cereus and Bacillus licheniformis, or other Bacillales members that contain SigB. In members of the B. cereus sensu lato group – which contain a common set of genes - it was found that the presence of varying SigB promoters gave rise to a unique SigB regulon structure per species, influencing pathogenesis via different strategies [16]. It is anticipated that Bacillales members that inhabit different niches may have SigB regulon structures that are different from B. subtilis and that these may have evolved to be mediated by different SigB activation routes.

Currently, four SigB activation routes are known for Bacillales members, namely, I) the stressosome RsbRST (Rsb = Regulator of SigB); II) the bipartite RsbQP, III) the two-component RsbKY system, and IV) direct activation (here refer to as Rsb-independent), as reviewed in Pané-Farré et al. [17] and Rodriguez Ayala et al. [18]. The stressosome signaling complex is formed by RsbR and its paralogs (RsbRB, RsbRC, RsbRD, and YtvA), RsbS serine phosphatase, and RsbT serine kinase [17, 19, 20]. The stressosome becomes phosphorylated upon exposure to environmental stressors, and RsbT is then released to dephosphorylate the RsbU phosphatase, leading to further dephosphorylation of the RsbV anti-sigma factor antagonist [21]. The dephosphorylated RsbV uncouples the binding between the anti-sigma factor RsbW and SigB, promoting the transcriptional activation of SigB and subsequently the expression of SigB regulon genes [19, 22, 23]. Activation of SigB via RsbQP involves the signal transfer from the α-hydrolase activator RsbQ to the RsbP phosphatase under nutritional stress (e.g., decrease in ATP, glucose starvation). RsbP then dephosphorylates RsbV, resulting in the same sequential SigB activation as for activation via the stressosome [24, 25]. The RsbKY two-component system includes the histidine sensor kinase RsbK and its cognate response regulator, RsbY [26]. By default, the methyltransferase (RsbM) methylates RsbK and negatively regulates SigB. Upon exposure to environmental/nutritional stressors, RsbK autophosphorylates and activates the RsbY phosphatase [27, 28], and subsequent SigB activation takes place in the same way as for the stressosome and the RsbQP module [26, 28, 29]. However, not all of the above-mentioned SigB activation systems are present in all Bacillales. Lastly, in low-temperature adaptation or nitrosative stress adaptation, SigB is activated independently from its regulators, RsbU, RsbP, and RsbV [30, 31].

To better understand the SigB regulon structures and functions in B. subtilis, this study employed B. subtilis 168 as a model, and used a newly derived SigB PBM to perform genome mining for novel SigB direct regulon members. The functionality of several predicted PBMs was verified using translational fusions to a reporter in a wild type (wt) and ΔsigB background. The SigB regulons in 18 different B. subtilis strains and 106 Bacillales genomes were also assessed and a Bacillales SigB PBM consensus sequence was obtained. Lastly, the absence and presence of the four SigB activation routes in the 106 Bacillales genomes were predicted.

Materials and methods

Sigma B (SigB) promoter binding motif (PBM) reconstruction

To identify novel SigB direct regulon members in B. subtilis 168, the SigB PBM was reconstructed as described by Wels et al. [32, 33] with slight modifications. In short, the respective operons of SigB regulon genes known to date in B. subtilis 168 were grouped according to their predicted operons (Supplementary Table S1). In total, 226 genes belonging to the SigB regulon were obtained for B. subtilis 168 on Subtiwiki [13], resulting in a compiled list with a total of 224 genes after removing duplicates. The operon structures of these genes were subsequently assessed and genes were allocated to the same operon when: 1) adjacent genes were on the same coding strand, 2) the intergenic region between adjacent genes was < 50 bp, and 3) no terminator was found between adjacent genes using Transterm (a tool to predict Rho-independent terminator) [34]. Regions 300 bp upstream of each operon and the full intergenic region (if < 300 bp) were then inspected to identify the SigB PBM, and used to derive a standard SigB PBM in MEME Suite [35, 36]. MEME was run using standard settings, with the following exceptions; −mod zoops (zero or one occurrence per sequence), −minw 10 (minimum width of 10), −maxw 50 (maximum width of 50), −dna (DNA molecule). The derived SigB PBM was used to repeat a search on the genome of B. subtilis 168 to obtain a new list of positive hit genes with a putative SigB PBM using the MAST search option in the MEME suite. The respective operons for these positive hit genes were predicted as aforementioned, and then a refined SigB PBM was built with MEME Suite. This refined SigB PBM with increased plasticity was used to search the genome of the 168 strain repeatedly, each with a different promoter spacer length, from N12 to N17. Promoter space length was increased/decreased by deleting or copying the least informative position in the position-specific scoring matrix (PSSM). The p-value indicating the confidence of predicted PBMs was set at 10− 5 (illustrated in Fig. 1). Each promoter hit sequence was manually curated (Supplementary Table S2). Different spacers were screened in this study as Vohradky et al. [8] indicated that genes controlled by SigB must contain both − 35 and − 10 binding motifs in the SigB PBM, and have a spacer length of 15 to 17 bp. Additionally, the similarity of the SigB PBM to the SigB consensus (indicated by the p-value) also took the nucleotide composition of spacers into account in this study. This is because the spacer compositions may influence the promoter binding strength and gene expression, and may promote co-recognition by different transcriptional regulators [37,38,39].

Fig. 1
figure 1

Flowchart of the reconstruction of SigB promoter binding motif for Bacillus subtilis. All SigB PBMs were predicted using the MEME Suite version 5.0.5 [35, 36], and MAST was used to screen the genome for a potential SigB PBM, as described by Wels et al. [32, 33]. The letter N indicates the number of base pairs present in the spacer region. STD stands for the standard motif, which was built based on all the listed SigB regulon genes on Subtiwiki up to October 2020 [13]. The p-value cut-off threshold, indicating the confidence level of the predicted motif, was set as 105. All final sequences predicted for B. subtilis 168 were manually curated and listed in Supplementary Table S2

Bacterial culturing conditions, media, chemicals, and DNA manipulations

All strains of B. subtilis and Escherichia coli used in this study were cultured in Lysogeny broth (LB) medium (Tritium Microbiologie, Eindhoven, The Netherlands), and propagated on LB agar plates unless stated otherwise. All incubations were performed at 37 °C, and all liquid cultures were incubated using shaking at 220 rpm. For standard DNA cloning, plasmids were prepared and isolated from TOP10 E. coli cells (Thermo Fischer Scientific, Bleiswijk, The Netherlands). Chemically competent E. coli cells were transformed via heat shock [40]. B. subtilis cells were transformed via natural competence in 1X MC competence medium (containing 200 μl of 10X MC plus 6.7 μl of 1 M MgSO4, 10 μl of 1% tryptophan, and 1.8 ml of sterile water). A stock solution of 10X MC was prepared with 14.036 g K2HPO4.3H2O, 5.239 g KH2PO4, 20 g glucose, 10 ml of 300 mM Na3C6H5O7, 1 ml of C6H8FeNO7, 1 g of casein hydrolysate, and 2 g of KC5H8NO4 to a total volume of 100 ml H2O. The amylase activity of B. subtilis transformants was tested on 1% starch plates and stained with iodine. For the β-galactosidase assay, B. subtilis was cultured in a C-minimal medium supplemented with 1 g/L glucose and 8 g/L potassium glutamate (CE) as described by Commichau et al. [41].

Oligonucleotides and KOD hot-start DNA polymerase were purchased from Merck (Zwijndrecht, The Netherlands). B. subtilis chromosomal DNA was isolated using the GenElute Bacterial Genomic DNA Kit (Merck). Plasmids were isolated using the GeneJET Plasmid Miniprep Kit (Thermo Fischer Scientific). PCR products were purified using the PCR Purification Kit (Qiagen, Hilden, Germany). All FastDigest restriction enzymes and T4 DNA ligase were purchased (Thermo Fischer Scientific) and used according to the manufacturer’s instructions. Bacterial culturing media were purchased (Tritium Microbiologie). DNA sequencing was performed by BaseClear B.V. (Leiden, The Netherlands).

Plasmids and reporter strains construction

The plasmids and strains constructed and used in this study are listed in Supplementary Table S3. The B. subtilis ΔsigB mutant was constructed using the long flanking homology recombination method as described in Kunst and Rapoport [42] with the selective marker for chloramphenicol resistance amplified from pNZ5319 [43]. The cre-recombinase plasmid pDR244 (purchased from the Bacillus Genetic Stock Centre, Columbus, USA) was used to excise the chloramphenicol cassette at the sigB locus [44] to obtain a ΔsigB clean deletion mutant (BY47).

Plasmids pCY22, pCY23, pCY26, pCY27, pCY31, pCY32, and pCY33 were constructed, containing translational reporter fusions of PyaaI, PywzA, PpucI, PgtaB, PylaL, PykaA, and PygaO, respectively (~ 150 bp upstream fragment), to lacZ. These SigB PBMs were selected as representative sequences as predicted in section “Sigma B (SigB) promoter binding motif (PBM) reconstruction” and were grouped into five categories irrespective of the spacer lengths: Category I (PrsbV): exact match at − 35 and − 10 regions; Category II (PytoQ, PpucI): exact match either at − 35 or − 10 region; Category III (PylaL, PygaO, PykaA): conserved GTTT at − 35 and NGG at − 10 region; Category IV (PyaaI): less conserved motif with high p-values; and Category V (PgtaB): with a duplicated − 35 or − 10 region. RsbV is a well-known SigB-induced gene and PrsbV was selected as a positive control because the PBM has a perfect match to the reported SigB consensus. PytoQ and PpucI were selected because both have conserved sequences at the − 35 and − 10 regions, but PpucI has a longer spacer of N17, and PytoQ has the same spacer length as the consensus, which is N14. These two PBMs were selected to check whether the difference in spacer length affects the promoter binding activity. For category three: PylaL, PygaO, and PykaA were selected because of the different spacer lengths of N16, N14, and N12, respectively. PywzA was selected because the gene ywzA has previously been reported by Petersohn et al. [3] to contain a SigB PBM, but the predicted PBM has not been tested so far. PyaaI (category IV) was selected because of a deviated third nucleotide position in the − 35 region from guanine to cytosine and has a higher p-value than the predicted PBMs in category III. PgtaB was selected because of the duplicated GTTTAA regions, and the predicted promoter sequence differed from the previous reports [2, 3].

The predicted SigB PBM (~ 150 bp in length) upstream of each mentioned gene, including the native ribosomal binding site, was amplified by Polymerase Chain Reaction (PCR) using the corresponding oligonucleotides as listed in Supplementary Table S4. Each amplified fragment was ligated into either pDG1728 (purchased from BGSC) or pAC7 amyE integration plasmid (courtesy of Dr. J. Stülke, University of Goettingen, Germany) with the restriction enzymes EcoRI and BamHI [45]. Additionally, the plasmid pBP638 (courtesy of Dr. F. Commichau, Cottbus-Senftenberg University, Germany) was used to study the activity of PytoQ and the plasmid pNW2205 (PrsbV -lacZ) (courtesy of Dr. N. Stanley Wall, University of Dundee, UK) was used as a positive control in the β-galactosidase-assay.

The constructed plasmids were transferred into chemically competent TOP10 E. coli cells using heat shock, and transformants were selected on LB agar supplemented with 100 μg/ml carbenicillin. Plasmids were isolated from the positive E. coli colonies, sequenced, and subsequently introduced into B. subtilis 168 wt and ΔsigB mutant (BY47) cells via natural competence. B. subtilis transformants were selected on Brain Heart Infusion (BHI) agar supplemented with antibiotics (either 250 μg/ml spectinomycin or 10 μg/ml kanamycin). Activities of the promoters in response to stresses known to trigger SigB activation were investigated.

β-galactosidase reporter assay

The activity of a promoter in vivo was determined using the β-galactosidase assay. B. subtilis strains carrying promoter-lacZ reporter fusions were grown in C-Glc medium [41], supplemented with 250 μg/ml spectinomycin or 10 μg/ml kanamycin. Overnight cultures in C-Glc were used to inoculate fresh C-Glc medium to an optical density of 0.05 at 600 nm (OD600), and allowed to grow to OD600 ~ 0.35. Cells were then divided into different portions and subjected to either ethanol stress (4%), NaCl (6%), or heat (upshift from 37 °C to 48 °C) for 10 min. The time point of 10 min was selected because the β-galactosidase activities did not change at different time points, i.e., 5 min, 10 min, 15 min, and 20 min after exposure to stressors using the PrsbV-lacZ positive control and the PgtaB -lacZ (data not shown). Moreover, the time point of 10 min was selected to prevent the influence of increased promoter activities when cells in the culture approach or enter the stationary phase. Even if growth is limited, e.g., in 4% ethanol, cells may grow slowly, so a later time point at 30 min or 60 min would not be ideal. The experimental setting has a limitation to validate promoters that require a longer time to respond. Cell pellets were collected before and after stress treatment and stored at − 20 °C. Quantitative studies of lacZ expression in B. subtilis were performed as described previously by Stannek et al. [46]. Briefly, cells were thawed and lysed with 400 μl Z-buffer (0.48 mM Na2HPO4. 2H2O; 0.32 mM NaH2PO4; 0.08 mM KCl; 8 μM MgSO4; 0.4 mM ß-mercaptoethanol; 200 μg lysozyme and 200 μg DNAse I) for 1 hour (h) at 37 °C. Lyzed cultures were centrifuged at 17,000 g to remove cell debris, and 100 μl of cell-free crude extract per sample was transferred into a new Eppendorf tube. 700 μl of Z-buffer without ß-mercaptoethanol was mixed with the 100 μl of crude extracts and incubated for 5 min at 28 °C. 800 μl of Z-buffer without ß-mercaptoethanol was used as a reference. Subsequently, 200 μl of 4 mg/ml of ortho-nitrophenyl-β-galactoside (ONPG) was added to all the crude extracts and the reference and allowed to react at 28 °C. When the cell extract turned visibly yellow, 500 μl of 62.5 mM Na2CO3 was added to stop the reaction. The absorption of samples at λ = 420 nm (absorption of O-nitrophenol) was measured. The protein concentration was determined via the Bradford assay [47, 48] at λ = 595 nm using the commercial Roti®-Quant Bradford solution (Carl Roth, Karlsruhe, Germany) according to the manufacturer’s protocol. The absorbance at A595nm and A420nm was corrected with the blank without cells. The specific ß- galactosidase activity (indicating the SigB activity) in Miller Units (MU)/mg protein was calculated using the formula:

$$\frac{\textrm{Units}}{\textrm{mg}\ \textrm{protein}}=1000\ \textrm{x}\frac{\ \textrm{A}420\ \textrm{nm}-\left(1.75\ \textrm{x}\ \textrm{A}550\textrm{nm}\right)}{\Delta \textrm{T}\ \textrm{x}\ \textrm{V}\ \textrm{x}\ \textrm{A}595}$$

One unit of β – galactosidase = the amount of enzyme produced to hydrolyze the chromogenic substrates ONPG to one nmol of O-nitrophenol (absorbs light at λ = 420 nm) per minute at 28 °C. V = 0.1 ml. Absorbance at 550 nm represents the scatter from cell debris, and multiplication with 1.75 gives the approximation of the scatter observed at 420 nm.

Bacillales core genome phylogenetic tree reconstruction, species-specific SigB PBM, and regulon structure prediction

To better understand the SigB regulon structure and function in B. subtilis and Bacillales members, a core genome phylogenetic tree of 18 B. subtilis strains (other than strain 168) and 106 Bacillales genomes was reconstructed (Supplementary Table S5). Species-specific SigB PBMs and SigB regulons were predicted and a Bacillales SigB consensus was derived as described below.

B. subtilis wild isolate strains and Bacillales members selection

The Bacillales members were selected when SigB operon genes had been described [17] and when they had been characterized for other properties (not necessarily related to SigB), such as producing high heat-resistance spores [49] or acid-tolerance [50]. The genomes of strains analyzed in this study included strains of B. subtilis of different origins (18 plus reference strain 168), Bacillus amyloliquefaciens (4), Bacillus vallismortis (4), Bacillus licheniformis (11), Bacillus velenzensis (1), Bacillus cereus (33), Bacillus coagulans (5), Bacillus thermoamylovorans (5), Bacillus pumilus (6), Anoxybacillus (6), Bacillus sporothermodurans (1), Geobacillus (14), Parageobacillus (7), Caldibacillus (2), Paenibacillus (2), Listeria (3) and Staphylococcus (3) which were extracted from the NCBI bacterial genome database, available at (

Core phylogenetic tree reconstruction and absence/presence of B. subtilis SigB regulon members in other Bacillales members

Orthologous groups were constructed using OrthAgogue (PMID24115168) [51]. Orthologous protein sequences with exactly one copy in all 125 Bacillales genomes were aligned with Muscle (Edgar, R.C. Nucleic Acids Res 32(5), 1792–97) and a core genome phylogenetic tree was constructed from the concatenated alignments via PhyML [52, 53]. Subsequently, all genes belonging to the SigB regulon in B. subtilis 168 [13] were selected as genes of interest (GOI) (Supplementary Table S1). The absence or presence of these GOI in the other 124 genome sequences was predicted via genome mining (Supplementary Table S5). Locus tags or gene symbols were used to identify the corresponding orthologous group of each GOI. Then, all GOI were clustered using the 1-(Spearman rank correlation of the gene copy number in each genome) as a distance measure, and neighbor-joining as the clustering algorithm (Phylip package, ( Lastly, the absence or presence of gene functions in a genome was predicted based on the absence or presence of one or more orthologs in that genome. The phylogenetic tree (“MLST trees”) heat map was visualized using iTOL (PMID27095192) [54].

Species-specific SigB PBMs and SigB regulon structures prediction for other Bacillales genomes

The predicted orthologs of the B. subtilis 168 SigB regulon members in 18 other B. subtilis strains and 106 Bacillales genomes (identified as described in section “Core phylogenetic tree reconstruction and absence/presence of B. subtilis SigB regulon members in other Bacillales members”) were established and used to derive a species-specific SigB PBM via procedures reported in section “Sigma B (SigB) promoter binding motif (PBM) reconstruction” (for an example illustrated for constructing the B. cereus SigB PBM see Fig. 2, step ii and iii). Subsequently, the species-specific SigB PBM was employed to screen for genes with potential SigB PBMs in the respective species. Operons for the new positive hit genes were predicted, and the promoter regions of these genes were used to build the species-specific SigB PBM 2 (step v, Fig. 2). Then, the genome of each species was screened repeatedly with the improved species-specific SigB PBMs with different promoter spacer lengths (N12 to N17) as described in section “Sigma B (SigB) promoter binding motif (PBM) reconstruction”, respectively. The screening thus resulted in a list of genes with putative SigB PBMs per species, forming the predicted SigB regulon for each inspected Bacillales genome. The overall presence/absence of the predicted species-specific SigB regulon members was cross-checked between all 125 genomes that were analyzed (Supplementary Table S6). The genome tree heat map for 125 Bacillales members (including B. subtilis 168) was generated using GENESIS 1.7.7 [55] (Supplementary Fig. S1).

Fig. 2
figure 2

Flowchart of the reconstruction of species-specific SigB promoter binding motif for other species belonging to the order of Bacillales. All SigB PBMs were predicted using the MEME Suite version 5.0.5 [35, 36], and MAST was used to screen the genome for a potential SigB PBM, as described by Wels et al. [32, 33]. The letter N indicates the number of base pairs present in the spacer region. STD stands for the standard motif, which was built based on all the listed SigB regulon genes on Subtiwiki up to October 2020 [13]. The Bacillales SigB promoter binding motif consensus was acquired from the predicted species-specific SigB PBM from the 125 Bacillales genomes, including the B. subtilis 168 strain

Bacillales Sigma B (SigB) sensing modules prediction

To evaluate the ability of various Bacillales to employ different SigB signaling modules, the occurrence of the three known SigB sensing modules (RsbRST, RsbQP, and RsbKY) was evaluated in 125 Bacillales genomes as described in section “Core phylogenetic tree reconstruction and absence/presence of B. subtilis SigB regulon members in other Bacillales members”. Briefly, genes encoding the proteins involved in these three known SigB signal transduction pathways for Bacillales [2,3,4,5, 9, 10, 12, 16, 17, 26, 56, 57], and the SigB regulators rsbX and rsbM were set as the GOI. The absence/presence of these GOI in 125 genomes was predicted via genome mining to check for the absence or presence of an orthologous protein (“Core phylogenetic tree reconstruction and absence/presence of B. subtilis SigB regulon members in other Bacillales members” section (Supplementary Table S7) and a heat map was generated as mentioned in section “Core phylogenetic tree reconstruction and absence/presence of B. subtilis SigB regulon members in other Bacillales members” and visualized with iTOL.

Results and discussion

Two-step SigB promoter binding motif derivation in B. subtilis 168

Based on the analysis of the B. subtilis genome 168 using the refined SigB PBM (see section “Sigma B (SigB) promoter binding motif (PBM) reconstruction”), 255 genes (some belonging to the same operon) with a putative SigB PBM were predicted, indicating that they may be directly regulated by SigB (Supplementary Table S2). Of these, 99 overlap with SigB regulon members described in the literature: 74 were listed as SigB regulon members on Subtiwiki [13] (Fig. 3), and an additional 25 were recently reported by Vohradsky et al. [8]. Thus, a total of 156 of the 255 genes with a predicted SigB PBM are novel putative SigB regulated genes that were not identified in earlier studies [2,3,4,5, 8,9,10,11,12]. To avoid possible false positives in this prediction, each of the predicted sequences was manually checked (Supplementary Table S2).

Fig. 3
figure 3

Novel SigB regulon genes detected for Bacillus subtilis. 255 SigB regulon genes with SigB PBMs were predicted, 99 overlapped with regulon genes already reported in literature, and 156 were newly identified in this study

The 255 genes that were predicted to have a SigB PBM in this study were grouped into five categories (Supplementary Table S2). Category I (green) contains 6 out of the 255 predicted genes with SigB PBMs that have an exact matching sequence to the previously reported SigB consensus motif (GTTTAA- N15 (± 2 bp) -GGGTAT) [6,7,8], demonstrating that a very small number of the predicted genes have Category I SigB PBM. These are the well-known SigB-regulated genes, i.e., sigB itself, the anti-SigB antagonist rsbV, the serine protein kinase rsbW, the phosphoserine phosphatase rsbX, the general stress gene ctc, and the acetyltransferase yjbC. These genes with the category I SigB PBM and spacer length of N14 (except yjbC with N13) have been shown to have extensive differential regulation under various conditions that lead to the induction of SigB [2,3,4, 9].

Category II (orange) contains 27 out of 255 genes with a − 35 binding motif that is identical to the consensus motif GTTTAA and only 1 to 2 base pairs variations in the − 10 binding motif, or a − 10 binding motif that is identical to the consensus motif GGGTAT and only 1 to 2 base pairs variations in the − 35 binding motif. This category of SigB PBMs deviates the least from the SigB consensus and is likely easily recognized by SigB. Many general stress genes such as the glucose starvation gene gsiB have the category II SigB PBMs (Supplementary Table S2). Eleven newly identified genes in this category were not listed in Subtiwiki. Of these, yjlB was also recently discovered as SigB dependent by Vohradsky et al. [8], who identified SigB PBMs with − 35 and − 10 binding sites with spacers of 13–15 bp (which they referred to as so-called Class I promoters). The genes pgcA, phoH, and yqhY that were reported by Vohradsky et al. [8] were not identified in our study, likely due to the use of different settings for the promoter searches. Moreover, the length of the spacer was up to 50 bp in the study of Vohradsky et al. [8].

Most of the genes (137 out of 255) detected in this study contain Category III (yellow) SigB PBMs, which have the conserved GTTT bases at the − 35 binding site and GG at the − 10 binding site in the promoter region. Many of these SigB regulon genes (~ 50%) were identified in earlier studies [2,3,4,5, 9], listed as SigB regulon genes in Subtiwiki, or reported by Vohradsky et al. [8]. These findings imply that SigB can recognize the binding motif well as long as the bases at the − 35 and − 10 binding sites are conserved, and the length of the spacer is between N12 and N17.

Category IV (grey) contains 76 genes with low level conserved SigB PBMs which have p-values close to the cut-off threshold (“Sigma B (SigB) promoter binding motif (PBM) reconstruction” section), likely indicating the presence of binding sequences for SigB and other transcriptional regulators. Several genes that are known to be coregulated by SigB and at least one other regulator were found in this group. These include, for instance, secG (preprotein translocase subunit), yaaI (general stress protein), and plsC (acylglycerolphosphate acyltransferase, biosynthesis of phospholipids). The plsC gene is known to be regulated by both SigB and FapR (a fatty acid synthetic gene repressor) to maintain membrane homeostasis [58]. Moreover, part of the SigB regulon overlaps with other regulons; it has been reported that SigB regulon members can be controlled by SigB and other regulators of which 36 have been reported (SigB and/or different regulators at different cellular states) [8, 13]. Therefore, the existence of genes that have less conserved SigB PBM or that do not have a SigB PBM but show SigB dependency suggested that cells may fine-tune response to integrate multiple signals in a variety of conditions.

Category V (blue) shows 9 out of 255 genes containing SigB PBM that deviated the most from the SigB consensus motif but were shortlisted likely due to the presence of either a duplicated − 35 or − 10 binding site or an occurrence of a single perfect − 35 or − 10 binding site (Supplementary Table S2). A previously described SigB regulon member, gtaB [3, 9], encoding the UTP-glucose-1-phosphate uridylyltransferase (a general stress protein), was found in this category. The duplicated GTTT region in the SigB PBM of gtaB might be the reason why the SigB PBM of gtaB had a low p-value (indicating higher similarity to the SigB consensus), suggesting that gtaB may be regulated by multiple regulators, via a versatile promoter sequence. To confirm this hypothesis, the predicted gtaB SigB PBM sequence was submitted to the database of transcriptional regulation in B. subtilis (DBTBS), and binding sequences of two other regulators SigA and DegU were found (Table 1). In the study of Vohradsky et al. [8], this gene was described to have a Class I promoter (i.e contained both − 35 and − 10 binding sites) and the SigB PBM was indicated at 96 bp upstream of ATG. However, a manual check on the genome of the strain 168 did not find a − 35 binding site.

Table 1 Nine selected promoters with predicted SigB promoter binding motifs.

Experimental validation of predicted SigB promoter binding motifs

To further validate these in silico results, promoters of genes belonging to the five different assigned categories were selected and their SigB-dependent activities were studied using a promoter-reporter approach in a wild type (wt) and ΔsigB background.

Eight predicted SigB PBMs (PytoQ, PpucI, PylaL, PygaO, PykaA, PywzA, PyaaI, PgtaB) and the well-known PrsbV were selected as representatives from different promoter categories (indicated in bold in Supplementary Table S2). The (putative) functions of these genes are presented in Table 1. SigB-dependent activities of translational promoter-lacZ fusions were determined in wt and ΔsigB mutants by measuring β- galactosidase activity upon temperature upshift from 37 °C to 48 °C, exposure to 4% ethanol, and exposure to 6% NaCl (in three independent experiments) (Fig. 4).

Fig. 4
figure 4

β-galactosidase activities of Category I-V predicted SigB promoters that were translationally fused to lacZ upon exposure to heat, ethanol, and salt stress. β-galactosidase activities of Category I-V predicted SigB promoters that were translationally fused to lacZ were investigated upon exposure to heat (37 °C > 48 °C), 4% (v/v) ethanol, and 6% (v/v) NaCl in wt cells and in the ΔsigB mutant of B. subtilis. Each data point represents a biologically independent replicate, and the bar indicates the average value for the three independent experiments. β-Galactosidase (LacZ) activities are presented in Miller units per milligram protein. The color-coded box at the right of each graph indicates the different categories to which the promoters belonged, based on the confidence level of the predicted PBM. Category I (green): the predicted PBM has the exact match at both − 35 and − 10 regions. Category II (orange): the predicted PBM has either an exact match at the − 35 region, with 1–2 bp variations at − 10, or vice versa. Category III (yellow): the predicted PBM has conserved GTTT bases at the − 35 region and the GG bases at the − 10 region. Category IV (grey): low-level homology compared with the conserved motif with borderline p-values; and Category V (Blue): with a duplicated − 35 or − 10 region. A PrsbV-lacZ activities; B PytoQ -lacZ activities; C PpucI -lacZ activities; D PylaL-lacZ activities; E PygaO-lacZ activities; F PykaA-lacZ activities. G PywzA-lacZ activities; H PyaaI -lacZ activities; I PgtaB -lacZ activities

Category I: PrsbV as the positive control

The rsbV gene is a well-known SigB regulon gene. Thus, PrsbV, which encompasses the sequence (GTTTAA-N14-GGGTAT) that exactly matches the SigB PBM consensus was used as the positive control and the Category I representative in the experimental validation. Exposure of wt cells containing PrsbV-lacZ to heat, ethanol, and salt resulted in 8-fold, 15-fold, and 6-fold induction of β-galactosidase activity compared to unstressed cells, respectively (Fig. 4A). In the ΔsigB mutant, no PrsbV -dependent β-galactosidase activity was observed in either stressed or unstressed conditions, showing that the observed promoter activity could be attributed to SigB induction, as expected for this positive control.

Category II: PytoQ shows SigB-dependent induction under ethanol stress, PpucI with longer spacer showed very mild SigB-dependent activity

Both ytoQ and pucI are newly identified putative SigB regulon genes in this study. PytoQ and PpucI, with predicted SigB PBM of (GTTTAA-N14-GGGTGA) and (GTTTAA-N17-GGGAAA), respectively, were used as the Category II representatives. These sequences have exact matches with the consensus − 35 binding site and both contained three conserved GGG bases at the − 10 binding site but have spacer lengths of 14 and 17 nucleotides, respectively.

Cells containing the PytoQ-lacZ (GTTTAA-N14-GGGTGA) showed SigB-dependent lacZ induction only upon exposure to 4% (v/v) ethanol, but not upon temperature upshift from 37 °C to 48 °C nor upon NaCl shock of 6% (v/v) (Fig. 4B). This gene is likely also regulated by at least one other regulator because the baseline activity of PytoQ at T0 before stress exposure was already ~ 100 MU/mg protein. The average PytoQ -dependent β-galactosidase activity after ethanol treatment was around 230 MU/mg protein for the wt carrying PytoQ-lacZ, which was about 130 MU/mg protein more than the activity at T0. The increase was lower (around 40 MU) in the ΔsigB mutant carrying PytoQ-lacZ. This difference between the wt and the ΔsigB mutant may imply that the product of the ytoQ gene is specific in response to stress caused by ethanol. Although the function of the ytoQ gene has not been fully elucidated, it was shown to be important under vitamin B6 starvation in B. subtilis [59]. Using the transcription regulator database DBTBS [6], we found that the predicted SigB PBM for ytoQ also contained two alternative PBMs: the first was for xre, which is the repressor of a phage-like bacteriocin, and the second was for codY, the repressor involved in the response to branched-chain amino acid limitation (see Table 1). These findings indicate that ytoQ may be regulated by SigB, CodY, and Xre under different conditions.

A mild, yet notable SigB-dependent response was observed for pucI (encoding allantoin permease) upon exposure to ethanol and NaCl stresses (Fig. 4C). After cells were exposed to 6% NaCl, β- galactosidase activities of wt PpucI-lacZ cultures increased around 4-fold compared to the control at T0. However, in the ΔsigB PpucI-lacZ cultures, this increase was notably higher (~ 6-fold), suggesting that SigB may partially involve in the negative regulation of pucI. Upon exposure to ethanol stress, β- galactosidase activities of wt PpucI-lacZ cultures did not change compared to the activity at T0, but the ΔsigB PpucI-lacZ mutant showed ~ 2- fold higher β-galactosidase activities than the wt PpucI-lacZ (see Fig. 4C). This result indicated that the mild increase in β-galactosidase activities as observed for the ΔsigB PpucI-lacZ mutant did not result from the exposure to ethanol but the deletion of the sigB gene.

The promoter of pucI was induced by NaCl in the wt and further induced in the ΔsigB mutant, implying that pucI may be co-regulated by other regulators/sigma factors as found in Table 1. pucI may also have roles in other stress conditions, triggered by other undiscovered stressors, or stressors other than heat, ethanol, and salt. It is noteworthy that the predicted SigB PBM of PpucI has an extended spacer (17 nucleotides) between the − 35 and − 10 binding motifs compared with the consensus spacer. This longer spacer region may affect the promoter strength by influencing the binding of the RNA polymerase, thereby affecting transcription. In transcription initiation, the bacterial RNA polymerase first locates the promoter, and its largest subunit (β-zipper) will interact with the spacer between the − 35 and − 10 elements to form a holoenzyme complex [60, 61].

Category III: PylaL showed mild SigB-dependent activity, PygaO and PykaA did not show changes and PywzA showed ethanol-specific SigB-dependent induction

For category III, PylaL (GTTTAT-N16-GGGAAT), PygaO (GTTTAT-N14-GGGAAT), PykaA (GTTTTT-N12-GGCTAT), and PywzA (GTTTAT-N14-AGGAAT) were selected based on the conserved GTTT at the − 35 binding site and the GG at the − 10 binding sites and different spacer lengths. ylaL, ygaO and ykaA are newly identified in this study to be potential SigB regulon candidates, in which ywzA is a known SigB regulon gene, used as a control of Category III.

Mild β-galactosidase activities were observed for PylaL but were unrelated to the stressors (Fig. 4D). In the wt PylaL-lacZ culture at T0 (before the exposure to heat, ethanol, or salt), no β-galactosidase activity was measured, whereas the ΔsigB PylaL-lacZ culture showed higher levels. The same results were observed in wt and the ΔsigB mutant after exposure to all three stressors, indicating that SigB may negatively affect ylaL. Based on the NCBI BlastP results, YlaL is 99.9% similar to the peptidyl-prolyl cis-trans isomerase and located next to the spore germination gene ylaJ. The predicted SigB PBM sequence for ylaL was also found to have a positive hit to the SigG binding motif with the consensus TGCATAT-N16-GATACTTA (DBTBS) (see Table 1), implying that YlaL may be coregulated by SigB and SigG, and may have a role in sporulation. The role of isomerase in sporulation was described in B. subtilis subs. spizizenii [49], and the involvement of SigB in sporulation control was also reported before. SigB is known to induce the expression of spo0E, which is a suppressor of spo0A and spoIIE genes which are required for sporulation initiation [14, 62]. Similarly, as indicated above for PpucI, the PylaL activities were relatively low, which may be attributed to the long spacer length of 16 nucleotides. Further experiments are thus needed to elucidate the potential regulation of PucI and YlaL by SigB.

PygaO and PykaA did not show evident changes in transcriptional activation in the three tested conditions despite having a relatively conserved SigB PBM (Fig. 4E and F). Baseline β-galactosidase activities for both promoters were seen before stress imposition at T0, and the PygaO and PykaA activities in ΔsigB mutant did not differ from the wt upon exposure to heat, ethanol, or salt stress. Results suggested that both ygaO and ykaA are likely controlled by other regulators than SigB (Table 1). Nonetheless, as only the three most commonly used SigB stressors were tested in this study, PygaO and PywzA may respond to other environmental or nutritional stressors.

In addition, the promoter activities of ywzA (a known SigB regulon gene predicted with a Category III SigB PBM) were verified via experiments. The β-galactosidase activity in wt PywzA-lacZ cultures increased 2-fold after heat treatment compared to cultures without treatment, but the induction after ethanol treatment stood out (20-fold higher) (Fig. 4G). No β-galactosidase activities were seen in the ΔsigB PywzA-lacZ cultures upon heat and ethanol stress, indicating that SigB was responsible for the increased expression in the wt cultures. Exposure to osmotic stress did not lead to changes in β-galactosidase activities in wt and ΔsigB PywzA-lacZ cultures. As the deletion of sigB abolished the activity of PywzA completely, ywzA is likely not co-regulated by other regulators under the conditions tested. However, alternative binding sites for codY (repressor in response to branched-chain amino acid limitation) and araR (repressor of the L-arabinose metabolic operon) were found for the predicted SigB PBM of ywzA (Table 1).

Of the newly predicted SigB regulon genes described so far (ytoQ, pucI, ylaL, ygaO, and ykaA), none of these five were identified as SigB regulated in previous studies, either via transposon mutagenesis [63,64,65], gel-based proteomics [30, 66,67,68,69], consensus promoter search [9], transcriptional profiling [2, 3], the combination of microarray and machine learning algorithm in defining the SigB regulon structure [5] or SigB modeling [8]. Our data show that three out of the five have SigB-dependent promoter activity, indicating that these genes may have been overlooked in earlier studies. The predicted PygaO and PykaA SigB PBM with high confidence did not show a clear SigB-dependent activation, suggesting that they might be induced by other stressors than heat, ethanol, or NaCl, or that the control by SigB is affected by the consensus and the spacer of the promoter.

Category IV and V: PyaaI and PgtaB showed SigB-dependent activities despite deviating considerably from the consensus

Lastly, PyaaI (GTTTTT-N14-GGCTAC), and PgtaB (GTTTTA-N14-CTTGTTTAA) were included as a representative from category IV and category V, respectively. Both yaaI and gtaB are known SigB regulon genes but were selected for verification in this study because the predicted SigB PBMs deviate considerably from the original SigB consensus (“Plasmids and reporter strains construction” section).

The PyaaI -LacZ activity in wt was induced the most upon exposure to ethanol stress, with a 35-fold increase, and heat stress resulted in a 4-fold increase compared with untreated wt PyaaI-lacZ cultures, wherease no difference of wt PyaaI -LacZ activity was observed under salt stress (Fig. 4H). This result suggests that the general stress gene yaaI plays an important role in protecting cells from damage caused by ethanol. The deletion of the sigB gene diminished the activity of PyaaI under all three conditions that were tested, suggesting that the expression of yaaI may be solely-dependent on SigB. However, an alternative binding site for codY was also found for the predicted SigB PBM of yaaI (Table 1), but the interaction of SigB, CodY and YaaI is yet to be explored.

The β-galactosidase activities of PgtaB were also investigated in wt and ΔsigB mutant (Fig. 4I). Petersohn et al. [3] reported that the putative SigB PBM for gtaB is located inside the gene coding region, but this study identified that the SigB PBM for gtaB is located upstream of the AUG start codon, containing the sequence GTTTTA-N14-GCTTGTTTAA. This SigB PBM met the selection criteria (“Sigma B (SigB) promoter binding motif (PBM) reconstruction” section) only because of the duplicated GTTT sequence at both − 35 and − 10 binding sites and it was chosen as a target to verify if SigB could bind to this predicted SigB PBM.

The wt PgtaB-lacZ culture displayed SigB-dependent induced β-galactosidase activity upon exposure to ethanol and salt stress despite the poor binding motif (Fig. 4I). At T0 before stress exposure, the baseline PgtaB activity in wt cells was already around 290 MU/mg protein, indicating that other transcriptional regulators may co-regulate this gene. A promoter sequence search using the DBTBS database revealed binding sites for two other alternative regulators, namely, SigA and DegU (Table 1). Despite the high baseline activity, an increase in PgtaB-dependent β-galactosidase activity was seen after ethanol (584 MU) and salt shock (333 MU), but no significant increase was observed upon heat treatment. Notably, the PgtaB-lacZ activity in the ΔsigB mutant at T0 was also ~ 3-fold lower than in the wt, indicating that SigB might play a role in regulating gtaB even under the control (presumably unstressed) condition. No induction was seen in the ΔsigB PgtaB-lacZ mutant in response to ethanol, salt, and heat stress. These results are in line with available transcriptomics data for B. subtilis 168 wt and ΔsigB mutant, showing more profound expression of gtaB upon ethanol or salt shock than upon heat shock [3]. This example demonstrated that SigB recognizes the predicted binding sequence, at least weakly, despite the large deviation from the SigB consensus sequence (GGGTAT) at the − 10 binding site.

The identification of putative SigB regulon members (section “Two-step SigB promoter binding motif derivation in B. subtilis 168”), of which nine predicted SigB PBMs of Category I-V were validated in section “Experimental validation of predicted SigB promoter binding motifs”, suggests that the SigB regulon in B. subtilis 168 may be even more extensive than currently thought. The number of theoretical SigB regulon genes was recently estimated to be 411 [8], and taken together with the predicted genes in this study, the total number may exceed 500 (Fig. 3). This large number of SigB regulon genes aligned with the notion that many SigB-regulated genes are also co-regulated by other transcriptional regulators, interlinking SigB regulation with other cellular processes [8]. Category III representatives that did not show SigB dependence in this study may respond to other so far unknown stressors or their promoter activities may be affected by the spacer length and/or compositions of the promoter, which requires further confirmation.

Functional distribution of known and predicted SigB regulon genes

The functions of the 156 predicted SigB regulon candidates in this study and all genes listed in Subtiwiki are presented in a functional distribution map (see Fig. 5). The list of genes with known functions (data extracted from Subtiwiki) is presented in Supplementary Table S2b. The sunburst map illustrates genes with and without SigB PBM (with shaded regions indicating genes with a predicted SigB PBM in this study; Fig. 4), and indicates that 30% of the genes encode for proteins involved in lifestyles (e.g., coping with stress, sporulation), 21% in information processing (e.g., protein synthesis and modification, transcription or translational regulation), 17% in metabolism regulation (e.g., biosynthesis of amino acid, lipids, utilization of carbon sources), 8% in cellular processing (e.g., transporter, exporter, homeostasis), 6% in phage-related function, and 18% constitute proteins with unknown functions.

Fig. 5
figure 5

Functional distribution map for the predicted and existing SigB regulon members. The sunburst map shows five known functional groups and a group with unknown functions, each labeled with different colors, of the predicted and existing SigB regulon members. The dotted regions in the map refer to genes with either a known SigB PBM or a predicted SigB PBM in each functional category. Blue represents genes involved in lifestyles (e.g., coping with stress, sporulation) (~ 30%); Red represents genes for information processing (e.g., protein synthesis and modification, transcription, or translational regulation) (~ 21%); Purple indicates genes for metabolism regulation (e.g., biosynthesis of amino acid, lipids, utilization of carbon sources) (~ 17%); Green indicates genes for cellular processing (e.g., transporter, exporter, homeostasis) (~ 8%); Turquoise includes genes for phage-related function (~ 6%); and Grey indicates genes with unknown functions (~ 18%). Underlying background data are shown in Supplementary Table S6

Many of the SigB regulon members that are involved in lifestyle management have generic functions in general stress protection. Some genes are likely regulated by SigB directly as SigB PBMs were predicted; some encode proteins with a role in resistance to toxins or antibiotics and others are linked to sporulation. Among the ~ 21% of members involved in information processing, many are well-known SigB-dependent genes (such as ctc, rsbV, rsbW, and sigB itself) that are involved in the regulation of gene expression, and many play a role in protein synthesis, modification, and degradation as well as DNA repair and recombination. Interestingly, a range of genes that have a putative SigB PBM are involved in cellular processes relating to nutrient transport, such as ABC transporters, ions transporters, and amino acids transporters. Other metabolic genes that are responsible for the biosynthesis of lipid synthesis, acquisition of amino acids, and utilization of different carbon sources also have predicted SigB PBMs (Fig. 5). The remaining genes were either prophages and mobile genetic elements, or genes encoding membrane proteins with undefined functions. Further functional investigation of these genes may help to better understand their involvement in stress response regulation in B. subtilis.

This study identified an additional 156 SigB regulon candidates with a putative SigB PBM in B. subtilis based on a computational approach using a more plastic SigB PBM. Multiple factors may have limited the detection of such genes in previous studies: 1) The use of a more restricted SigB PBM in earlier studies [6, 7]; 2) Utilization of the same stressors in global SigB- mediated GSR studies. Ethanol, heat, and salt were used mostly because of their potent SigB triggering response. Although other investigators studied the induction of SigB by acid, cold, antibiotics, reduced ATP, GTP, low oxygen, glucose limitation, blue light, red light, carbonyl cyanide m-chlorophenylhydrazone (CCCP), butanol, pH, low pressure, high-level iron, and oxidative stress, global transcriptomic or proteomic analyses were not performed in all studies [3, 5, 10, 14, 30, 31, 70,71,72,73,74,75,76,77,78,79,80]. Therefore, SigB-regulated genes that are specific for other stressors may have not been detected. Moreover, stimuli or stressors that play a role in unexplored ecological niches may trigger SigB as well; 3) Analysis restricted to the SigB induction response. The standard setting in many studies of the global transcriptomics or proteomics SigB stress response focused on the induction pattern of SigB, and thus often checked for the “loss-of-gene-function” in the ΔsigB mutant. Such approaches may overlook genes that may be negatively regulated by SigB; 4) SigB is likely active without “triggers” in the control condition and co-regulates other cellular mechanisms than the SigB GSR. Several known or predicted SigB candidate genes with SigB PBMs are involved in miscellaneous functions in B. subtilis, for instance, biofilm formation, sporulation, utilization of sugars, biosynthesis of amino acids, and homeostasis. Reder et al. [14] and Rothstein et al. [62] reported the negative regulation of SigB in sporulation initiation. Bartolini et al., [15] demonstrated the role of SigB in regulating biofilm growth rate via the interaction with the SinR transcriptional regulator. SigB was also shown to indirectly affect the expression of surfactin, a cyclic lipopeptide (biosurfactant) [81].

Other than the functions reported in B. subtilis, SigB can influence motility, virulence, and invasiveness in other Bacillales members, e.g., L. monocytogenes and S. aureus [78, 82, 83], indicating that the structure of SigB regulons between species may have diverged due to differences in physiological responses upon exposure to a broad range of stressors. Thus, to obtain a global outlook of the SigB general stress regulon in other species in the Bacillales order, putative SigB regulons of 18 other B. subtilis strains and 106 Bacillales genomes were predicted as described in section “Bacillales core genome phylogenetic tree reconstruction, species-specific SigB PBM, and regulon structure prediction”.

SigB regulon prediction for B. subtilis wild isolates and Bacillales genomes

Genomes of 18 B. subtilis wild isolates and 106 other Bacillales genomes including different Bacillus species, Listeria spp., and Staphylococcus spp. (Supplementary Table S5) were mined for the presence of SigB regulon members that had been identified in B. subtilis 168 (Supplementary Table S1). Based on the conserved protein sequences, the reconstructed phylogenetic tree heat map in Fig. 6 showed that nearly all genes that belong to the SigB regulon in B. subtilis 168 had orthologs in 18 other wild B. subtilis isolates, except for a small cluster of germination genes (yfkR, yfkS, yfkT) and a group of genes with unknown function (ykzN, ypuB, yydC) (details in Supplementary Table S5). Much more prominent differences were seen between SigB regulons of B. subtilis and other Bacillus species and Bacillales genomes, such as B. licheniformis, B. cereus, and other further related species like Geobacillus, Listeria, and Staphylococcus. Around 25% of the B. subtilis 168 SigB regulon genes were absent in B. licheniformis, around 50% were missing in B. cereus, and three quarters were lacking in Geobacillus, Listeria, and Staphylococcus (Fig. 6, Supplementary Table S5).

Fig. 6
figure 6

Genome tree heat map of 18 other B. subtilis strains and 106 Bacilalles genomes and the prediction of the presence of orthologs of B. subtilis 168 SigB regulon genes The phylogenetic genome tree heat map for 19 Bacillus subtilis genomes (including B. subtilis 168) and 106 Bacillales members was generated using iTOL (PMID27095192) [54]. The Y-axis in the genome tree shows the phylogenetic relationships of all genomes based on the core conserved protein sequences in each genome. The x-axis shows the clustering of B. subtilis 168 SigB regulon genes listed in Subtiwiki up to October 2020 (Supplementary Table S1). A green square indicates the presence of a target gene, the intensity of the green color indicates the gene copy number, and white indicates the absence of a gene. Underlying background data are shown in Supplementary Table S5

The prediction results showed that SigB regulates different sets of genes in different species, e.g., in B. cereus, L. monocytogenes, or S. aureus. Therefore, the SigB PBMs are likely species-specific and deviate from the SigB consensus of B. subtilis (GTTTAA-N15 (± 2 bp) -GGGTAT), or have the same PBM but is/are not present in front of the same genes. This assumption was further investigated by reconstructing species-specific SigB PBMs per inspected species (described in section “Species-specific SigB PBMs and SigB regulon structures prediction for other Bacillales genomes”) which is illustrated in Fig. 2. In a step-wise approach, 1) orthologous genes of the B. subtilis 168 SigB regulon members in other analyzed species were first predicted, 2) the predicted orthologous genes were grouped into operons, 3) promoters of these operons were used to reconstruct a species-specific SigB PBM, 4) the respective genome of a species was screened for the presence of the constructed species-specific PBM, resulting in a new list of genes with putative SigB PBMs, 5) operons for these genes with putative SigB PBMs were again predicted, and species-specific SigB PBM 2 was constructed, and lastly, 6) the respective genome of a species was repeatedly screened for the presence of the species-specific SigB PBM 2 with different spacers N12 – N17 (Fig. 2). The same procedures were performed for other species included in this study, and each species-specific SigB PBM was used to derive a Bacillales consensus (KTTTW- N12-N17- GGGWAW). The Bacillales consensus is less conserved at the first guanine nucleotide in the − 35 region when compared with the B. subtilis SigB consensus and contains more thymine than adenine nucleotides.

The predicted species-specific SigB regulon members with/without SigB PBM are presented in a heat map (Supplementary Fig. S1), showing the 1) absence/presence of predicted genes that are orthologous to the B. subtilis 168 SigB regulon members in 124 Bacillales genomes, and 2) the putative species-specific SigB regulon genes with or without a predicted SigB PBM. The full list of these genes is shown in Supplementary Table S6. Four major observations can be made based on the heat map: 1) many other Bacillales genomes contain genes that are orthologous to the SigB regulon members of B. subtilis 168, but they do not necessarily have a SigB PBM; 2) groups of predicted regulon genes with/without SigB PBM are species-specific; 3) a group of genes that are orthologous to the SigB regulon members of B. subtilis 168 or the predicted species-specific regulon genes do not have a SigB PBM, and lastly 4) a group of genes with/without SigB PBM are specific for Bacillus species but is absent in other Bacillales genomes. These results suggested two different possibilities for different species: divergent SigB PBMs or conserved PBMs may control different genes.

Overall, the results obtained confirm that SigB plays a role in adaptive stress response in many species, but that the actual cellular responses and genes involved are different for different species. Species-specific SigB regulons may correspond with distinct physiological responses of species when dealing with a broad range of stressors in their environments. Orthologs of SigB-regulated genes with a SigB PBM as found in B. subtilis were mainly found in other Bacillus species but did not necessarily contain an upstream SigB PBM, and the majority of the B. subtilis SigB regulon genes were absent in Listeria spp. and Staphylococcus spp.

These results are in line with the studies of Scott et al. [16], who reported on the divergence of the SigB GSR regulons within the B. cereus sensu lato group (containing species that are not included in this study: B. anthracis, B. mycoides, B. pseudomycoides, B. thuringiensis, B. weihenstephanensis and B. cytotoxicus). Four lineages of the SigB regulon were described in their study, and each lineage has arisen from the selection of a set of genes from the common gene pool, with the “reassignment” of a SigB promoter to these genes to support pathogenesis for different sensu lato members [16]. The extra members in addition to the SigB core regulon (consisting of ~ 20 members) was suggested to serve a distinct function in different habitats and support the phenotype of a specific member, such as enhancing pathogenic potential or increasing competence against other microorganisms in the soil [16]. Moreover, the SigB PBM predicted for B. cereus in this study (shown in Fig. 2) is highly similar to the one described in the study of Scott et al. [16] despite using different species of B. cereus group members (in this study only B. cereus genomes were used, Supplementary Table S6).

The Occurrence of SigB sensing modules in other B. subtilis strains and Bacillales genomes

Our analysis showed differences between the predicted SigB regulons for various Bacillales genomes. We furthermore examined the presence of the three well-known SigB signaling modules in Bacillales, i.e. RsbRST, RsbQP, and RsbKY (as described in section “Bacillales Sigma B (SigB) sensing modules prediction”).

The absence/presence of genes involved in sensing stressors, SigB transduction, activation, and regulation in 19 B. subtilis genomes and 106 Bacillales members is presented in Fig. 7. The complete datasets relating to the presence of the sensing modules are presented in Supplementary Table S7. The majority of the inspected genomes carried the sigB gene, however, it was absent in several species like B. thermoamylovorans, Parageobacillus thermoglucosidasius, and Anoxybacillus. These species likely evolved to utilize other stress sensing systems and were therefore excluded from further analyses.

Fig. 7
figure 7

Genome tree heat map of sensing modules of the SigB general stress for 19 Bacillus subtilis genomes and 106 Bacillales members. The heat map of the core genome tree of sensing modules of the SigB general stress for 19 Bacillus subtilis genomes and 106 Bacillales members was generated using iTOL (PMID27095192) [54]. The tree on the left shows the phylogenetic relationships of all genomes based on the core conserved protein sequences in each genome. The tree on the top shows the clustering of genes involved in SigB signal sensing. The red squares indicates the presence of a target gene, and the intensity of the red color indicates the gene copy number. White indicates the absence of a gene. Underlying background data are shown in Supplementary Table S7. The insert shows the summary of the general distribution of the three sensing modules for each species belonging to the Bacillales order. Stressosome refers to the rsbRST stressosome genes and stressosome downstream elements to rsbV, rsbW and sigB. The energy branch refers to the rsbQP genes and the two-component to the rsbKY genes. Inspected strains that contain only a single gene of a signaling module, e.g., the presence of an orphan rsbK gene without its cognate response regulator gene rsbY, or the presence of a rsbP phosphatase gene without its partner rsbQ, are referred to as having a partial two-component, or a partial energy system, respectively. Other SigB sensors and regulators included the blue light sensor (YtvA), the regulator of SigB methyltransferase BC1007 (renamed to RsbM by Chen et al., [27]), and the feedback regulator RsbX phosphoserine phosphatase. Underlying background data are shown in Supplementary Table S7

All species that were found to contain the sigB gene also carried the rsbV and rsbW genes. These three genes are highly conserved in Bacillales, including the Geobacillus and Paenibacillus species that have a high GC content. In general, the complete stressosome system (rsbRST and rsbU) and its feedback regulator (rsbX) were found in B. subtilis, B. vallismortis, B. licheniformis, B. amyloliquefaciens, B. coagulans, B. pumilus, B. valezensis, B. sporothermodurans, L. monocytogenes, and L. innocua (Fig. 7). The RsbQP module was found in B. subtilis, B. vallismortis, and Paenibacillus spp. and the two-component system (rsbKY) and its regulator rsbM were identified in B. cereus, Paenibacillus spp. and B. coagulans (Fig. 7).

The stressosome RsbRST SigB activation pathway

The RsbRST stressosome system was detected in many species but only the RsbRA ortholog is conserved (Fig. 7). Many species that contain the stressosome genes lacked one or more genes that encodes either the RsbRB, RsbRC, RsbRD or the YtvA ortholog. B. licheniformis, B. pumilus, and B. sporothermodurans did not contain the genes that encode the RsbRB and RsbRC orthologs, whereas B. coagulans did not have the genes encoding the RsbRC and RsbRD orthologs. Only B. subtilis, B. vallismortis, B. amyloliquafaciens, B. pumilus, L. monocytogenes, and L. innocua contained the gene encoding the fifth RsbR ortholog, YtvA, which is involved in the sensing of blue light. Even in B. subtilis, not every strain carried the genes encoding the same RsbR orthologs. Some of the B. subtilis food isolate strains (B4068 and B4073) lost the gene encoding the RsbRC ortholog and B. subtilis B4122 had two gene copies encoding the RsbRD ortholog (Fig. 7, Supplementary Table S7).

The occurrence of different types and numbers of RsbR orthologs results in the formation of heterogeneous stressosome complexes [84], thereby affecting specific stress sensing via the turrets (referring to the protein structure of the RsbR and its paralogs) with different ligands [17, 20]. Moreover, from an evolutionary point of view, different species may be exposed to certain stress conditions in particular ecological niches, putting selective pressures on retaining certain RsbR orthologs that are needed for sensing of a specific stress. Different RsbR orthologs can have distinct functions in mediating stress; RsbRC has for instance been shown to be responsible for a slow progressive stress response upon ethanol stress, whereas RsbRA mediates a fast transient response [85]. However, RsbRC was absent in most genomes that carry genes which encode stressosome members (Fig. 7), which implies that its role may not be essential, or that its function is redundant in the presence of RsbRA. In B. subtilis, it was shown that RsbRA and RsbRB orthologs are responsible for light sensing, and the role of RsbR can be complemented by RsbRC or RsbRD [86].

Additionally, the gene rsbX was found only in strains that contained the stressosome genes, confirming the reported function of the RsbX in forming a negative feedback loop by dephosphorylating the RsbR and RsbS, thereby resetting the activated stressosome to its original state [87]. The absence of rsbX in other strains that do not contain the stressosome genes was not a surprise as this feedback loop was probably not required.

The bipartite RsbQP SigB activation pathway

The entire RsbQP module was detected in B. subtilis, B. vallismortis, and Paenibacillus spp. (Fig. 7), indicating that the nutritional stress sensing branch is not restricted to just B. subtilis. The rsbQ gene was missing in the majority of the genomes, but the rsbP gene, encoding a PP2C phosphatase, was distributed more broadly. As rsbP and rsbY belong to the same OG1773 group, a “rsbP” ortholog was also found in the genomes of B. cereus, Geobacillus spp., Paenibacillus spp., and Parageobacillus spp. (see Fig. 7).

The finding was in line with the publication of Nadezhdin et al. [25] who suggested that the predicted RsbQP could be functional in other species, or react with other stressors, but not in the same way as described for B. subtilis [75, 77]. The observation that RsbQ is generally absent while RsbP is generally present in Bacillales (Supplementary Table S7) suggests RsbP may have other so far unidentified functions, and its interaction partner may not be limited to RsbQ. This speculation can be supported by reports on alternative functions of RsbP in earlier studies, which showed that RsbP is also involved in sensing red light [77] and that this protein interacted with the stressosome to mediate resilience toward oxidative and nitrosative stress in B. subtilis [88].

The two-component RsbKY SigB activation pathway

The two-component system encoded by rsbKY and rsbM was found in B. cereus, Paenibacillus spp., and B. coagulans. This sensing module was well-known to be specific to B. cereus and its group members [57], while a full RsbKYM system in B. coagulans has not been described previously. In addition, the rsbK (bc1008) gene in B. cereus was found to have an ortholog in B. subtilis and other group members like B. licheniformis, B. amyloliquefaciens, B. pumilus, and B. vallismortis (Fig. 7, Supplementary Table S7), but no cognate response regulator was detected adjacent to the predicted rsbK gene in these species. The functionality of an RsbK ortholog in B. subtilis and its potential role in the SigB activation pathway remains to be confirmed.

The suggestion of an alternative SigB activation pathway in B. subtilis is not new, as extreme heat and chill conditions have been reported to induce SigB either directly, or independently from RsbV [30, 74], and nitrosative stress has been reported to trigger SigB in the absence of RsbT or RsbP [31]. Moreover, the RsbW (the anti-sigma factor) exhibited high cross-phosphorylation activity by other kinases [89], and may thereby cause unexpected SigB activation in B. subtilis.


This study generated a SigB PBM that took spacer composition into account and has higher plasticity than the previously reported consensus sequence [6,7,8]. This was used to identify potential novel candidates that belong to the SigB regulon of B. subtilis. Of the 255 genes with predicted SigB PBMs as identified in this study, 99 genes have previously been reported in the literature, indicating the identification of 156 new putative SigB regulon members. The functionality of nine predicted SigB PBMs (including a positive control PrsbV) was further validated via experiments and results that were obtained showed that 1) some promoters containing the predicted SigB PBMs are stressor-specific; 2) spacer length likely influences the promoter activity with a spacer length of 14 bp appearing to be optimal; 3) less conserved SigB PBMs are still recognized by SigB (e.g., PgtaB) but may be co-regulated by other transcriptional regulators.

Furthermore, this study demonstrated the diversity of SigB regulons in different species of the Bacillales order. The various SigB regulons are likely linked with distinct strategies that are required for survival of different species in diverse ecological niches. While all Bacillales members can be present in soils, some inhabit salt lakes, hot springs, guts of invertebrates, etc. Thus, the strategies used to cope with stressors found in the soil environments likely overlap between these species, but different stress management strategies may be required in other niches. A Bacillales SigB consensus was predicted, with the sequence of KTT at the − 35 and the GG at the − 10 binding site, respectively. The SigB stress sensing modules were also species-specific and may even vary between different strains of the same species, likely due to the evolution of Bacillales members in specific habitats, demanding different needs to sense unique stressors.

Overall, the entire SigB regulatory network is sophisticated and not yet fully understood even for the well-characterized organism B. subtilis 168. Knowledge and information gained in this study can be used in further SigB GSR studies to uncover a complete picture of the role of SigB in B. subtilis and other species.

Availability of data and materials

All data generated or analysed during this study are included in this published article (and its supplementary information files). The whole genome sequences of the 125 genomes used are published on the NCBI Assembly Database at, with the respective genome accession numbers as shown in the list below.


Genome accession no.






























Genome accession no.


























































































Genome accession no.


























































































Genome accession no.











































General stress response


Gene of interest


Promoter binding motif






  1. Moran CP, Lang N, Banner CD, Haldenwang WG, Losick R. Promoter for a developmentally regulated gene in Bacillus subtilis. Cell. 1981;25:783–91.

    Article  CAS  Google Scholar 

  2. Petersohn A, Antelmann H, Gerth U, Hecker M. Identification and transcriptional analysis of new members of the sigmaB regulon in Bacillus subtilis. Microbiology (Reading, Engl). 1999a;145(Pt 4):869–80.

    Article  CAS  Google Scholar 

  3. Petersohn A, Brigulla M, Haas S, Hoheisel JD, Völker U, Hecker M. Global analysis of the general stress response of Bacillus subtilis. J Bacteriol. 2001;183:5617–31.

    Article  CAS  Google Scholar 

  4. Price CW, Fawcett P, Cérémonie H, Su N, Murphy CK, Youngman P. Genome-wide analysis of the general stress response in Bacillus subtilis. Mol Microbiol. 2001a;41:757–74.

    Article  CAS  Google Scholar 

  5. Nannapaneni P, Hertwig F, Depke M, Hecker M, Mäder U, Völker U, et al. Defining the structure of the general stress regulon of Bacillus subtilis using targeted microarray analysis and random forest classification. Microbiology (Reading, Engl). 2012;158:696–707.

    Article  CAS  Google Scholar 

  6. Sierro N, Makita Y, de Hoon M, Nakai K. DBTBS: a database of transcriptional regulation in Bacillus subtilis containing upstream intergenic conservation information. Nucleic Acids Res. 2008;36:D93–6.

    Article  CAS  Google Scholar 

  7. Coelho RV, de Avila E, Silva S, Echeverrigaray S, Delamare APL. Bacillus subtilis promoter sequences data set for promoter prediction in Gram-positive bacteria. Data Brief. 2018;19:264–70.

    Article  Google Scholar 

  8. Vohradsky J, Schwarz M, Ramaniuk O, Ruiz-Larrabeiti O, Vaňková Hausnerová V, Šanderová H, et al. Kinetic modeling and meta-analysis of the Bacillus subtilis SigB Regulon during Spore Germination and Outgrowth. Microorganisms. 2021;9:E112.

    Article  CAS  Google Scholar 

  9. Petersohn A, Bernhardt J, Gerth U, Höper D, Koburger T, Völker U, et al. Identification of sigma(B)-dependent genes in Bacillus subtilis using a promoter consensus-directed search and oligonucleotide hybridization. J Bacteriol. 1999b;181:5718–24.

    Article  CAS  Google Scholar 

  10. Waters SM, Robles-Martínez JA, Nicholson WL. Exposure of Bacillus subtilis to low pressure (5 kilopascals) induces several global regulons, including those involved in the SigB-mediated general stress response. Appl Environ Microbiol. 2014;80:4788–94.

    Article  CAS  Google Scholar 

  11. Arrieta-Ortiz ML, Hafemeister C, Bate AR, Chu T, Greenfield A, Shuster B, et al. An experimentally supported model of the Bacillus subtilis global transcriptional regulatory network. Mol Syst Biol. 2015;11:839.

    Article  Google Scholar 

  12. Schumann W. Regulation of bacterial heat shock stimulons. Cell Stress Chaperones. 2016;21:959–68.

    Article  CAS  Google Scholar 

  13. Zhu B, Stülke J. SubtiWiki in 2018: from genes and proteins to functional network annotation of the model organism Bacillus subtilis. Nucleic Acids Res. 2018;46:D743–8.

    Article  CAS  Google Scholar 

  14. Reder A, Gerth U, Hecker M. Integration of σB activity into the decision-making process of sporulation initiation in Bacillus subtilis. J Bacteriol. 2012;194:1065–74.

  15. Bartolini M, Cogliati S, Vileta D, Bauman C, Rateni L, Leñini C, et al. Regulation of biofilm aging and dispersal in Bacillus subtilis by the alternative sigma factor SigB. J Bacteriol. 2018.

  16. Scott E, Dyer DW. Divergence of the SigB regulon and pathogenesis of the Bacillus cereus sensu lato group. BMC Genomics. 2012;13:564.

    Article  CAS  Google Scholar 

  17. Pané-Farré J, Quin MB, Lewis RJ, Marles-Wright J. Structure and function of the stressosome signalling hub. Subcell Biochem. 2017;83:1–41.

    Article  CAS  Google Scholar 

  18. Rodriguez Ayala F, Bartolini M, Grau R. The Stress-responsive alternative Sigma factor SigB of Bacillus subtilis and its relatives: An old friend with new functions. Front Microbiol. 2020:11.

  19. Kim T-J, Gaidenko TA, Price CW. A multicomponent protein complex mediates environmental stress signaling in Bacillus subtilis. J Mol Biol. 2004;341:135–50.

    Article  CAS  Google Scholar 

  20. Marles-Wright J, Lewis RJ. The Bacillus subtilis stressosome. Commun Integr Biol. 2008;1:182–4.

    Article  Google Scholar 

  21. Yang X, Kang CM, Brody MS, Price CW. Opposing pairs of serine protein kinases and phosphatases transmit signals of environmental stress to activate a bacterial transcription factor. Genes Dev. 1996;10:2265–75.

    Article  CAS  Google Scholar 

  22. Benson AK, Haldenwang WG. Bacillus subtilis sigma B is regulated by a binding protein (RsbW) that blocks its association with core RNA polymerase. Proc Natl Acad Sci U S A. 1993;90:2330–4.

    Article  CAS  Google Scholar 

  23. Alper S, Dufour A, Garsin DA, Duncan L, Losick R. Role of adenosine nucleotides in the regulation of a stress-response transcription factor in Bacillus subtilis. J Mol Biol. 1996;260:165–77.

    Article  CAS  Google Scholar 

  24. Kaneko T, Tanaka N, Kumasaka T. Crystal structures of RsbQ, a stress-response regulator in Bacillus subtilis. Protein Sci. 2005;14:558–65.

    Article  CAS  Google Scholar 

  25. Nadezhdin EV, Brody MS, Price CW. An α/β Hydrolase and Associated Per-ARNT-Sim domain comprise a bipartite sensing module coupled with diverse output domains. PLoS One. 2011:6.

  26. de Been M, Tempelaars MH, van Schaik W, Moezelaar R, Siezen RJ, Abee T. A novel hybrid kinase is essential for regulating the sigma(B)-mediated stress response of Bacillus cereus. Environ Microbiol. 2010;12:730–45.

    Article  CAS  Google Scholar 

  27. Chen L-C, Chen J-C, Shu J-C, Chen C-Y, Chen S-C, Chen S-H, et al. Interplay of RsbM and RsbK controls the σB activity of Bacillus cereus. Environ Microbiol. 2012;14:2788–99.

    Article  CAS  Google Scholar 

  28. Chen J-C, Liu J-H, Hsu D-W, Shu J-C, Chen C-Y, Chen C-C. Methylatable signaling helix coordinated inhibitory receiver domain in sensor kinase modulates environmental stress response in Bacillus Cereus. PLoS One. 2015;10:e0137952.

    Article  CAS  Google Scholar 

  29. Chen J-C, Chang C-F, Hsu D-W, Shu J-C, Chen H-Y, Chen C-Y, et al. Temporal regulation of σB by partner-switching mechanism at a distinct growth stage in Bacillus cereus. Int J Med Microbiol. 2017;307:521–32.

    Article  CAS  Google Scholar 

  30. Brigulla M, Hoffmann T, Krisp A, Völker A, Bremer E, Völker U. Chill induction of the SigB-dependent general stress response in Bacillus subtilis and its contribution to low-temperature adaptation. J Bacteriol. 2003;185:4305–14.

    Article  CAS  Google Scholar 

  31. Tran V, Geraci K, Midili G, Satterwhite W, Wright R, Bonilla CY. Resilience to oxidative and nitrosative stress is mediated by the stressosome, RsbP and SigB in Bacillus subtilis. J Basic Microbiol. 2019;59:834–45.

    Article  CAS  Google Scholar 

  32. Wels M, Francke C, Kerkhoven R, Kleerebezem M, Siezen RJ. Predicting cis-acting elements of Lactobacillus plantarum by comparative genomics with different taxonomic subgroups. Nucleic Acids Res. 2006;34:1947–58.

    Article  CAS  Google Scholar 

  33. Wels M, Overmars L, Francke C, Kleerebezem M, Siezen RJ. Reconstruction of the regulatory network of Lactobacillus plantarum WCFS1 on basis of correlated gene expression and conserved regulatory motifs. Microb Biotechnol. 2011;4:333–44.

    Article  CAS  Google Scholar 

  34. Jacobs GH, Stockwell PA, Tate WP, Brown CM. Transterm—extended search facilities and improved integration with other databases. Nucleic Acids Res. 2006;34:D37–40.

    Article  CAS  Google Scholar 

  35. Bailey TL, Boden M, Buske FA, Frith M, Grant CE, Clementi L, et al. MEME Suite: tools for motif discovery and searching. Nucleic Acids Res. 2009;37:W202–8.

    Article  CAS  Google Scholar 

  36. Bailey TL, Johnson J, Grant CE, Noble WS. The MEME Suite. Nucleic Acids Res. 2015;43:W39–49.

    Article  CAS  Google Scholar 

  37. Liebeton K, Lengefeld J, Eck J. The nucleotide composition of the spacer sequence influences the expression yield of heterologously expressed genes in Bacillus subtilis. J Biotechnol. 2014;191:214–20.

    Article  CAS  Google Scholar 

  38. Gaballa A, Guariglia-Oropeza V, Dürr F, Butcher BG, Chen AY, Chandrangsu P, et al. Modulation of extracytoplasmic function (ECF) sigma factor promoter selectivity by spacer region sequence. Nucleic Acids Res. 2018;46:134–45.

    Article  CAS  Google Scholar 

  39. Han L, Cui W, Suo F, Miao S, Hao W, Chen Q, et al. Development of a novel strategy for robust synthetic bacterial promoters based on a stepwise evolution targeting the spacer region of the core promoter in Bacillus subtilis. Microb Cell Factories. 2019;18:96.

    Article  CAS  Google Scholar 

  40. Froger A, Hall JE. Transformation of plasmid DNA into E. coli using the heat shock method. J Vis Exp. 2007.

  41. Commichau FM, Gunka K, Landmann JJ, Stülke J. Glutamate metabolism in Bacillus subtilis: gene expression and enzyme activities evolved to avoid futile cycles and to allow rapid responses to perturbations of the system. J Bacteriol. 2008;190:3557–64.

    Article  CAS  Google Scholar 

  42. Kunst F, Rapoport G. Salt stress is an environmental signal affecting degradative enzyme synthesis in Bacillus subtilis. J Bacteriol. 1995;177:2403–7.

    Article  CAS  Google Scholar 

  43. Lambert JM, Bongers RS, Kleerebezem M. Cre-lox-based system for multiple gene deletions and selectable-marker removal in Lactobacillus plantarum. Appl Environ Microbiol. 2007;73:1126–35.

    Article  CAS  Google Scholar 

  44. Koo B-M, Kritikos G, Farelli JD, Todor H, Tong K, Kimsey H, et al. Construction and analysis of two genome-scale deletion libraries for Bacillus subtilis. Cell Systems. 2017;4:291–305.e7.

    Article  CAS  Google Scholar 

  45. Weinrauch Y, Msadek T, Kunst F, Dubnau D. Sequence and properties of comQ, a new competence regulatory gene of Bacillus subtilis. J Bacteriol. 1991;173:5685–93.

  46. Stannek L, Thiele MJ, Ischebeck T, Gunka K, Hammer E, Völker U, et al. Evidence for synergistic control of glutamate biosynthesis by glutamate dehydrogenases and glutamate in Bacillus subtilis. Environ Microbiol. 2015;17:3379–90.

    Article  CAS  Google Scholar 

  47. Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72:248–54.

    Article  CAS  Google Scholar 

  48. Kruger NJ. The Bradford Method For Protein Quantitation. In: Walker JM, editor. The protein protocols handbook. Totowa: Humana Press; 2009. p. 17–24.

    Chapter  Google Scholar 

  49. Berendsen EM, Boekhorst J, Kuipers OP, Wells-Bennik MHJ. A mobile genetic element profoundly increases heat resistance of bacterial spores. ISME J. 2016;10:2633–42.

    Article  CAS  Google Scholar 

  50. Patel MA, Ou MS, Harbrucker R, Aldrich HC, Buszko ML, Ingram LO, et al. Isolation and characterization of acid-tolerant, thermophilic bacteria for effective fermentation of biomass-derived sugars to lactic acid. Appl Environ Microbiol. 2006;72:3228–35.

    Article  CAS  Google Scholar 

  51. Ekseth OK, Kuiper M, Mironov V. orthAgogue: an agile tool for the rapid prediction of orthology relations. Bioinformatics. 2014;30:734–6.

    Article  CAS  Google Scholar 

  52. Guindon S, Gascuel O. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol. 2003;52:696–704.

    Article  Google Scholar 

  53. Guindon S, Dufayard J-F, Lefort V, Anisimova M, Hordijk W, Gascuel O. New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0. Syst Biol. 2010;59:307–21.

    Article  CAS  Google Scholar 

  54. Letunic I, Bork P. Interactive tree of life (iTOL) v3: an online tool for the display and annotation of phylogenetic and other trees. Nucleic Acids Res. 2016;44:W242–5.

    Article  CAS  Google Scholar 

  55. Sturn A, Mlecnik B, Pieler R, Rainer J, Truskaller T, Trajanoski Z. Client–Server environment for high-performance gene expression data analysis. Bioinformatics. 2003;19:772–3.

    Article  CAS  Google Scholar 

  56. van Schaik W, van der Voort M, Molenaar D, Moezelaar R, de Vos WM, Abee T. Identification of the σB Regulon of Bacillus cereus and conservation of σB-regulated genes in low-gc-content gram-positive bacteria. J Bacteriol. 2007;189:4384–90.

    Article  CAS  Google Scholar 

  57. de Been M, Francke C, Siezen RJ, Abee T. Novel sigmaB regulation modules of Gram-positive bacteria involve the use of complex hybrid histidine kinases. Microbiology (Reading, Engl). 2011;157:3–12.

    Article  CAS  Google Scholar 

  58. Albanesi D, de Mendoza D. FapR: from control of membrane lipid homeostasis to a biotechnological tool. Front Mol Biosci. 2016;3:64.

    Article  CAS  Google Scholar 

  59. Rosenberg J, Yeak KC, Commichau FM. A two-step evolutionary process establishes a non-native vitamin B6 pathway in Bacillus subtilis. Environ Microbiol. 2018;20:156–68.

    Article  CAS  Google Scholar 

  60. Yuzenkova Y, Tadigotla VR, Severinov K, Zenkin N. A new basal promoter element recognized by RNA polymerase core enzyme. EMBO J. 2011;30:3766–75.

    Article  CAS  Google Scholar 

  61. Lee J, Borukhov S. Bacterial RNA Polymerase-DNA Interaction—The driving force of gene expression and the target for drug action. Front Mol Biosci. 2016:3.

  62. Rothstein DM, Lazinski D, Osburne MS, Sonenshein AL. A Mutation in the Bacillus subtilis rsbU gene that limits rna synthesis during sporulation. J Bacteriol. 2017:199.

  63. Boylan SA, Redfield AR, Brody MS, Price CW. Stress-induced activation of the Sigma B transcription factor of Bacillus subtilis. J. Bacteriol. 1993;175:7931–37.

  64. Boylan SA, Rutherford A, Thomas SM, Price CW. Activation ofB acillus subtilis transcription factor sigma B by a regulatory pathway responsive to stationary-phase signals. J Bacteriol. 1992;174:3695–706.

  65. Boylan SA, Thomas MD, Price CW. Genetic method to identify regulons controlled by nonessential elements: isolation of a gene dependent on alternate transcription factor Sigma B of Bacillus subtilis. J Bacteriol. 1991;173:7856–66.

  66. Antelmann H, Scharf C, Hecker M. Phosphate starvation-inducible proteins of Bacillus subtilis: proteomics and transcriptional analysis. J Bacteriol. 2000;182:4478–90.

  67. Hahne H, Mäder U, Otto A, Bonn F, Steil L, Bremer E, et al. A comprehensive proteomics and transcriptomics analysis of Bacillus subtilis salt stress adaptation. Journal of Bacteriology. 2010;192:870–82.

  68. Höper D, Bernhardt J, Hecker M. Salt stress adaptation of Bacillus subtilis: a physiological proteomics approach. Proteomics. 2006;6:1550–62.

  69. Wolff S, Otto A, Albrecht D, Zeng JS, Büttner K, Glückmann M, et al. Gel-free and gel-based proteomics in Bacillus subtilis: a comparative study. Mol. Cell Proteomics. 2006;5:1183–92.

  70. Völker U, Maul B, Hecker M. Expression of the SigB-dependent general stress regulon confers multiple stress resistance in Bacillus subtilis. J Bacteriol. 1999;181:3942–8.

    Article  Google Scholar 

  71. Bernhardt J, Weibezahn J, Scharf C, Hecker M. Bacillus subtilis during feast and famine: visualization of the overall regulation of protein synthesis during glucose starvation by proteome analysis. Genome Res. 2003;13:224–37.

    Article  CAS  Google Scholar 

  72. Helmann JD, Wu MFW, Gaballa A, Kobel PA, Morshedi MM, Fawcett P, et al. The global transcriptional response of Bacillus subtilis to peroxide stress is coordinated by three transcription factors. J Bacteriol. 2003;185:243–53.

    Article  CAS  Google Scholar 

  73. Mascher T, Margulis NG, Wang T, Ye RW, Helmann JD. Cell wall stress responses in Bacillus subtilis: the regulatory network of the bacitracin stimulon. Mol Microbiol. 2003;50:1591–604.

    Article  CAS  Google Scholar 

  74. Holtmann G, Brigulla M, Steil L, Schütz A, Barnekow K, Völker U, et al. RsbV-Independent induction of the sigb-dependent general stress regulon of Bacillus subtilis during growth at high temperature. J Bacteriol. 2004;186:6150–8.

    Article  CAS  Google Scholar 

  75. Zhang S, Haldenwang WG. Contributions of ATP, GTP, and redox state to nutritional stress activation of the Bacillus subtilis σB transcription factor. J Bacteriol. 2005;187:7554–60.

    Article  CAS  Google Scholar 

  76. Gaidenko TA, Kim T-J, Weigel AL, Brody MS, Price CW. The blue-light receptor YtvA acts in the environmental stress signaling pathway of Bacillus subtilis. J Bacteriol. 2006;188:6387–95.

    Article  CAS  Google Scholar 

  77. Avila-Pérez M, van der Steen JB, Kort R, Hellingwerf KJ. Red light activates the sigmaB-mediated general stress response of Bacillus subtilis via the energy branch of the upstream signaling cascade. J Bacteriol. 2010;192:755–62.

    Article  CAS  Google Scholar 

  78. Ondrusch N, Kreft J. Blue and red light modulates SigB-dependent gene transcription, swimming motility and invasiveness in Listeria monocytogenes. PLoS One. 2011;6:e16151.

    Article  CAS  Google Scholar 

  79. Jurk M, Schramm P, Schmieder P. The blue-light receptor YtvA from Bacillus subtilis is permanently incorporated into the stressosome independent of the illumination state. Biochem Biophys Res Commun. 2013;432:499–503.

    Article  CAS  Google Scholar 

  80. Yu W-B, Ye B-C. Transcriptional profiling analysis of Bacillus subtilis in response to high levels of Fe(3+). Curr Microbiol. 2016;72:653–62.

    Article  CAS  Google Scholar 

  81. Bartolini M, Cogliati S, Vileta D, Bauman C, Ramirez W, Grau R. Stress-responsive alternative sigma factor SigB plays a positive role in the antifungal proficiency of Bacillus subtilis. Appl Environ Microbiol. 2019:85.

  82. Kim H, Marquis H, Boor KJ. SigmaB contributes to Listeria monocytogenes invasion by controlling expression of inlA and inlB. Microbiology (Reading, Engl). 2005;151:3215–22.

    Article  CAS  Google Scholar 

  83. Mitchell G, Fugère A, Pépin Gaudreau K, Brouillette E, Frost EH, Cantin AM, et al. SigB is a dominant regulator of virulence in Staphylococcus aureus small-colony variants. PLoS One. 2013:8.

  84. Delumeau O, Chen C-C, Murray JW, Yudkin MD, Lewis RJ. High-molecular-weight complexes of RsbR and paralogues in the environmental signaling pathway of Bacillus subtilis. J Bacteriol. 2006;188:7885–92.

    Article  CAS  Google Scholar 

  85. Cabeen MT, Russell JR, Paulsson J, Losick R. Use of a microfluidic platform to uncover basic features of energy and environmental stress responses in individual cells of Bacillus subtilis. PLoS Genet. 2017;13:e1006901.

    Article  CAS  Google Scholar 

  86. van der Steen JB, Ávila-Pérez M, Knippert D, Vreugdenhil A, van Alphen P, Hellingwerf KJ. Differentiation of function among the RsbR Paralogs in the general stress response of Bacillus subtilis with regard to light perception. J Bacteriol. 2012;194:1708–16.

    Article  CAS  Google Scholar 

  87. Chen C-C, Yudkin MD, Delumeau O. Phosphorylation and RsbX-dependent dephosphorylation of RsbR in the RsbR-RsbS complex of Bacillus subtilis. J Bacteriol. 2004;186:6830–6.

    Article  CAS  Google Scholar 

  88. Guldimann C, Boor KJ, Wiedmann M, Guariglia-Oropeza V. Resilience in the face of uncertainty: Sigma factor b fine-tunes gene expression to support homeostasis in gram-positive bacteria. Appl Environ Microbiol. 2016;82:4456–69.

    Article  CAS  Google Scholar 

  89. Shi L, Pigeonneau N, Ravikumar V, Dobrinic P, Macek B, Franjevic D, et al. Cross-phosphorylation of bacterial serine/threonine and tyrosine protein kinases on key regulatory residues. Front Microbiol. 2014;5:495.

    Article  Google Scholar 

Download references


We thank Fabian Commichau, Jörg Stülke, and Nicola Stanley-Wall for providing the plasmid pBP638, pAC7, and pNW2205, respectively, and thank Manisha Pandey for constructing plasmid pNW2205.


KCY received funding from the European Union’s Horizon 2020 research and innovation program under the Marie Skłodowska-Curie grant agreement No 721456.

Author information

Authors and Affiliations



JB and MW helped to perform the in silico analysis. KCY analyzed the in silico data, constructed all plasmids and strains used in this study, conducted all experiments, and wrote the manuscript. KCY, TA, and MWB designed experiments together, TA and MWB reviewed the manuscript. The author(s) read and approved the final manuscript.

Corresponding author

Correspondence to Marjon H J Wells-Bennik.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that there are no conflicts of interest.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Supplementary Table S1.

List of SigB regulon genes described to date October 2020 for Bacillus subtilis 168. Supplementary Table S2. The list of newly predicted SigB regulon genes with a SigB PBM in this study. Supplementary Table S2b. Functional distribution of SigB regulon candidate. Supplementary Table S3. Strains and plasmids constructed in this study. Supplementary Table S4. Oligonucleotides used in this study. Supplementary Table S5. Presence and absence of B. subtilis 168 SigB regulon genes described to date October 2020 in other Bacillales. Supplementary Table S6. Predicted putative species-specific SigB regulon genes in 19 Bacillus subtilis genomes and 96 Bacillales members. Supplementary Table S7. Presence and absence of genes involved in SigB sensing modules in Bacillales described to date October 2020.

Additional file 2: Supplementary Fig. S1.

Heat map of SigB regulon members with SigB promoter binding motifs in 125 Bacillales.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Yeak, K.Y.C., Boekhorst, J., Wels, M. et al. Prediction and validation of novel SigB regulon members in Bacillus subtilis and regulon structure comparison to Bacillales members. BMC Microbiol 23, 17 (2023).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: