Global Transcriptional Response of Bacillus subtilis to Heat Shock

Abstract

In response to heat stress, Bacillus subtilis activates the transcription of well over 100 different genes. Many of these genes are members of a general stress response regulon controlled by the secondary sigma factor, B , while others are under control of the HrcA or CtsR heat shock regulators. We have used DNA microarrays to monitor the global transcriptional response to heat shock. We find strong induction of known B -dependent genes with a characteristic rapid induction followed by a return to near prestimulus levels. The HrcA and CtsR regulons are also induced, but with somewhat slower kinetics. Analysis of DNA sequences proximal to newly identified heat-induced genes leads us to propose ϳ70 additional members of the B regulon. We have also identified numerous heat-induced genes that are not members of known heat shock regulons. Notably, we observe very strong induction of arginine biosynthesis and transport operons. Induction of several genes was confirmed by quantitative reverse transcriptase PCR. In addition, the transcriptional responses measured by microarray hybridization compare favorably with the numerous previous studies of heat shock in this organism. Since many different conditions elicit both specific and general stress responses, knowledge of the heat-induced general stress response reported here will be helpful for interpreting future microarray studies of other stress responses.DNA microarray technology provides a powerful tool for the analysis of global transcriptional responses elicited by various physical and chemical stresses. One challenge in this sort of analysis is to distinguish stress-specific responses from more general stress responses. For example, in Bacillus subtilis, many different stresses (including heat shock, osmotic stress, and energy stress) activate the large general stress response controlled by the B transcription factor (20,44,45). While others have used two-dimensional protein gels to classify cellular stress responses (55, 58), DNA-based methods have several advantages: they can be rapidly adapted to new organisms, they provide greater coverage of the genome, and data processing is comparatively easy to automate. Ultimately, it may be possible to integrate both technologies, at least for well-studied model organisms (19,41,56).We have initiated a series of studies to characterize the global transcriptional responses of B. subtilis, a model grampositive microorganism. Here, we document the heat-induced general stress response. Heat shock was chosen for this initial study since it is arguably the best-studied stress response in this organism and includes activation of the large general stress response under the control of B (20, 44). In addition, a subset of antibiotics that inhibit translation have been reported to induce heat shock genes in other organisms (57). It is anticipated that knowledge of transcriptional responses to antimicrobial compounds will be useful for both antibacterial discovery and characterization (47).Analysis of the transcriptional profile of B...

Defining the structure of the general stress regulon of Bacillus subtilis using targeted microarray analysis and random forest classification

“…In our analysis, these genes were strongly induced not only under low-temperature growth but also under all the conditions tested. This example also highlights the value of our study, because these genes were previously identified only by Petersohn et al (2001), and would have been missed if the focus had been on the 67 overlapping genes.Around 95 % of genes which were described as being SigBdependent in at least two of the three previous transcriptional profiling studies (Helmann et al, 2001; Petersohn et al, 2001; Price et al, 2001) were validated in the present study (Fig. 6).…”

“…The probes in the custom array targeted a total of 350 genes. The selection of genes was based on: (1) 67 genes that were described to be SigB-dependent in all the three previous transcriptional profiling studies (Helmann et al, 2001; Petersohn et al, 2001; Price et al, 2001), (2) 185 genes that were reported to be SigB-dependent in one or two of the three studies, (3) 52 genes reported in other studies to be probably SigB-dependent (Petersohn et al, 1999b(Petersohn et al, , 2001 Price et al, 2001), (4) 10 additional genes of interest based on unpublished data, and (5) 36 genes that were selected as controls as not exhibiting SigB-dependent expression but encoding a variety of other cellular stress-adaptation reactions (for details see Supplementary Table S1). …”

“…Various stresses were applied to the cultures incubated at 37 uC with shaking at 210 r.p.m. : ethanol stress [the culture was treated with a final concentration of 4 % (v/v) ethanol (Völker et al, 1995)]; salt stress [solid sodium chloride was added to a final concentration of 4 % (w/v) (Petersohn et al, 2001)]; butanol stress [the culture was treated with a final concentration of 1 % (v/v) butanol (Helmann et al, 2003)]; heat stress [the culture was shifted from 37 to 48 uC (Helmann et al, 2001)]; and oxidative stress [the culture was treated with hydrogen peroxide at a final sublethal concentration of 60 mM (Helmann et al, 2003)]. Samples were taken Fig.…”

“…Several research groups have identified regulon members by various approaches, including transposon mutagenesis (Boylan et al, 1991(Boylan et al, , 1992(Boylan et al, , 1993, gel-based proteome analysis or consensus promoter search-driven extensive oligonucleotide hybridization screening (Petersohn et al, 1999a; Völker et al, 1994). After the genome sequence of B. subtilis became available in 1997 (Kunst et al, 1997), a comprehensive search for SigB regulon members was performed in three genome-wide transcriptional profiling studies (Helmann et al, 2001; Petersohn et al, 2001; Price et al, 2001). These three DNA array studies were performed to identify SigB regulon members by testing subsets of well-described stress conditions.…”

“…They compared a wild-type strain and a sigB mutant strain cultivated in complex LuriaBertani (LB) medium and postulated 193 genes to be SigBdependent. Helmann et al (2001) conducted experiments exposing a wild-type strain to heat shock (shift from 37 to 48 u C) in minimal medium and used DNA microarray glass slides. That study identified 123 genes as members of the SigB regulon.…”