σI from Bacillus subtilis: Impact on Gene Expression and Characterization of σI-Dependent Transcription That Requires New Types of Promoters with Extended -35 and -10 Elements

Abstract

The σ^I sigma factor from Bacillus subtilis is a σ factor associated with RNA polymerase (RNAP) that was previously implicated in adaptation of the cell to elevated temperature. Here, we provide a comprehensive characterization of this transcriptional regulator. By transcriptome sequencing (RNA-seq) of wild-type (wt) and σ^I-null strains at 37°C and 52°C, we identified ∼130 genes affected by the absence of σ^I Further analysis revealed that the majority of these genes were affected indirectly by σ^I The σ^I regulon, i.e., the genes directly regulated by σ^I, consists of 16 genes, of which eight (the dhb and yku operons) are involved in iron metabolism. The involvement of σ^I in iron metabolism was confirmed phenotypically. Next, we set up an in vitro transcription system and defined and experimentally validated the promoter sequence logo that, in addition to -35 and -10 regions, also contains extended -35 and -10 motifs. Thus, σ^I-dependent promoters are relatively information rich in comparison with most other promoters. In summary, this study supplies information about the least-explored σ factor from the industrially important model organism B. subtilisIMPORTANCE In bacteria, σ factors are essential for transcription initiation. Knowledge about their regulons (i.e., genes transcribed from promoters dependent on these σ factors) is the key for understanding how bacteria cope with the changing environment and could be instrumental for biotechnologically motivated rewiring of gene expression. Here, we characterize the σ^I regulon from the industrially important model Gram-positive bacterium Bacillus subtilis We reveal that σ^I affects expression of ∼130 genes, of which 16 are directly regulated by σ^I, including genes encoding proteins involved in iron homeostasis. Detailed analysis of promoter elements then identifies unique sequences important for σ^I-dependent transcription. This study thus provides a comprehensive view on this underexplored component of the B. subtilis transcription machinery.

Keywords: RNA-seq; RNAP; iron metabolism; promoter; sigma factor.

FIG 1

Spot assays of B. subtilis wt, ΔsigI-rsgI, and ΔrsgI strains on agar plates at 37°C and 52°C. (a) Serial dilutions of mid-logarithmic phase cultures (OD₆₀₀, ∼0.45) of wt, ΔsigI-rsgI, and ΔrsgI strains were spotted on LB agar plates and incubated at 37°C and 52°C for 40 h. The experiment was repeated three times with the same result. (b) Colony morphology of wt and ΔsigI-rsgI strains.

FIG 2

Cell shapes of the B. subtilis wt, ΔsigI-rsgI, and ΔrsgI cells imaged by scanning electron microscopy (SEM). (a) wt cells grown in LB at 37°C. (b) wt cells grown in LB at 52°C. (c) ΔsigI-rsgI cells grown in LB at 37°C. (d) ΔsigI-rsgI cells grown in LB at 52°C. (e) ΔrsgI cells grown in LB at 37°C. (f) ΔregI cells grown in LB at 52°C. The experiment was performed twice with identical results. Bars, 5 μm.

FIG 3

Genes in B. subtilis affected by σ^I. B. subtilis ΔsigI-rsgI and wt strains were grown at 37°C and 52°C in LB broth to an OD₆₀₀ of ∼0.45. RNA was extracted and libraries were prepared for transcriptome sequencing (RNA-seq). (a) Genes positively regulated by σ^I. These genes were downregulated in the ΔsigI-rsgI strain compared to in the wt strain. (b) Genes negatively regulated by σ^I. These genes were upregulated in ΔsigI-rsgI compared to wt.

FIG 4

The σ^I factor is involved in iron metabolism in B. subtilis. Growth curves of the ΔsigI-rsgI (triangles), ΔrsgI (squares), and wt (circles) strains grown in defined MOPS medium at 37°C in the presence (filled shapes) or absence (empty shapes) of FeCl₃. The experiment was repeated three times. The error bars show ± standard deviation (SD).

FIG 5

Multiple-round in vitro transcription assays with promoter regions of σ^I-regulated genes and RNAPσ^I. (a) Alignment of σ^I-dependent promoters. The −10 and −35 elements and +1 position for PsigI (16), PmreBH, and PbcrC (21) are in red. (b) Transcription was performed with the RNAPσ^I holoenzyme and the RNAP core. Transcription with the RNAP core was used to assess potential contamination of the RNAP core with σ factors. Promoter PsigI was used as a control and its transcription was set as 1. Pveg (σ^A dependent) was used as a negative control for RNAPσ^I. Primary data (radioactively labeled transcripts resolved on polyacrylamide [PAA] gels) are shown below the graph. The error bars show averages from three independent experiments ± SD. (c) The σ^I consensus logo was created from the 8 promoter sequences shown in panel a. Conserved promoter elements are indicated above the logo.

FIG 6

Multiple-round in vitro transcription assays with mutated PsigI promoter region. (a) Fragments of the PsigI promoter region used for evaluating the importance of the identified sequence elements. Fragment 1 contains the native PsigI promoter region. Fragments 2 to 9 contain mutations in −10, −35, extended −35 (Ext −35), extended −10 (Ext −10) elements, and their combinations. Mutations are highlighted in red. Fragments 10 and 11 contain control neutral (N) mutations highlighted in blue boxes. The −10 and −35 regions and +1 position are in bold and underlined. Extended elements are in bold. (b) In vitro transcription with specific double substitutions in PsigI −10 and −35 σ^I binding sites, extended −10, and extended −35 elements. Transcription was performed with the RNAP core (lane B, blank assay) and RNAPσ^I holoenzyme on PCR products as a template. The blank assay was used to demonstrate that the RNAP core was devoid of contaminating σ factors. Primary data (radioactively labeled transcripts resolved on PAA gel) are shown below the quantification graph; the vertical black line indicates the border between two PAA gels used for illustration. The dotted lines indicate lanes from the same gel electronically positioned next to each other. Error bars show averages from three independent experiments ± SD.

FIG 7

Phylogenetic tree of B. subtilis σ^I homologs. The list of the homologs is in Fig. S8 in the supplemental material. The phylogenetic tree was inferred with RAxML, and the best-scoring maximum likelihood tree is shown. Numbers denote bootstrap values (percent), as reported by RAxML. The scale bar represents expected number of substitutions per site. Cbo, Clostridium botulinum; Csa, Clostridium saccharobutylicum; Chy, Carboxydothermus hydrogenoformans; Rth, Rumuniclostridium thermocellum; Cce, Clostridium cellulolyticum; Dha, Desulfitobacterium hafniense; Ban, Bacillus anthracis; Bth, Bacillus thuringiensis; Bce, Bacillus cereus; Gth, Geobacillus thermodenitrificans; Bsu, Bacillus subtilis; Bam, Bacillus amyloliquefaciens; Spn, Streptococcus pneumoniae; Bli, Bacillus licheniformis; Bme, Bacillus megaterium; Cab, Chlamydia abortus; Sth, Symbiobacterium thermophilum; Hmo, Heliobacterium modesticaldum.