|
|
J Bacteriol. 2008 March; 190(5): 1751–1761. Published online 2007 December 21. doi: 10.1128/JB.01516-07. | PMCID: PMC2258692 |
Copyright © 2008, American Society for Microbiology Transcriptome Analysis of Sorbic Acid-Stressed Bacillus subtilis Reveals a Nutrient Limitation Response and Indicates Plasma Membrane Remodeling ![[down-pointing small open triangle]](/corehtml/pmc/pmcents/x25BF.gif) † Alex Ter Beek,* Bart J. F. Keijser,‡ Andre Boorsma,§ Anna Zakrzewska,¶ Rick Orij, Gertien J. Smits, and Stanley Brul Laboratory for Molecular Biology and Microbial Food Safety, Swammerdam Institute for Life Sciences, University of Amsterdam, Amsterdam, The Netherlands Received September 19, 2007; Accepted December 13, 2007. The weak organic acid sorbic acid is a commonly used food preservative, as it inhibits the growth of bacteria, yeasts, and molds. We have used genome-wide transcriptional profiling of Bacillus subtilis cells during mild sorbic acid stress to reveal the growth-inhibitory activity of this preservative and to identify potential resistance mechanisms. Our analysis demonstrated that sorbic acid-stressed cells induce responses normally seen upon nutrient limitation. This is indicated by the strong derepression of the CcpA, CodY, and Fur regulon and the induction of tricarboxylic acid cycle genes, SigL- and SigH-mediated genes, and the stringent response. Intriguingly, these conditions did not lead to the activation of sporulation, competence, or the general stress response. The fatty acid biosynthesis (fab) genes and BkdR-regulated genes are upregulated, which may indicate plasma membrane remodeling. This was further supported by the reduced sensitivity toward the fab inhibitor cerulenin upon sorbic acid stress. We are the first to present a comprehensive analysis of the transcriptional response of B. subtilis to sorbic acid stress. The food industry commonly utilizes sorbic acid and other weak organic acids as preservatives. Sorbic acid ( trans- trans-2,4-hexadienoic acid) is a six-carbon unsaturated fatty acid with a pK a of 4.76 and was first isolated from unripe berries of Rowan ( Sorbus aucuparia). The acid, or its anionic salt, is used in a variety of food products and has a broad range of antimicrobial activities against spoilage bacteria, yeasts, and molds ( 5, 17, 57). However, the exact mechanism by which sorbate inhibits microbial growth is not entirely understood. No single mechanism appears to explain its toxicity to various spoilage organisms. Depending on the pK a of the acid and the pH of the environment, in solution, sorbate exists in equilibrium between the dissociated state (S −) and the undissociated state (HS). The neutral HS is lipid permeable and able to diffuse into the cell, reaching an equilibrium when the inside and outside concentrations of HS are equal. Inside, a new equilibrium is formed between S − and HS, releasing protons into the cytosol. This may acidify the cytosol, causing an inhibition of many metabolic functions ( 11, 59). Furthermore, the lipophilic tail of the sorbate molecule has been shown to disrupt the membrane and interfere with membrane proteins ( 65). This, together with the entry of protons, could result in a loss of the proton motive force, disrupting oxidative phosphorylation and affecting the transport of nutrients ( 4, 27, 58). Also, the accumulation of S − in the cell could cause a rise in osmolarity and affect cytosolic enzymes ( 3, 77). In order to counteract the effects of sorbic acid, microorganisms use various resistance mechanisms. Saccharomyces cerevisiae uses H +-ATPases to pump out the excess protons at the cost of ATP to maintain pH homeostasis ( 36, 45) and induces a dedicated ATP binding cassette (ABC) transporter, Pdr12, to prevent the accumulation of the anion S − ( 35). These processes, however, may reduce energy resources significantly ( 8, 35, 36). Studies with benzoic acid showed that adapted S. cerevisiae and Zygosaccharomyces bailii cells reduce their permeabilities to benzoate ( 33, 71). Changes in fatty acid composition in sorbate-stressed Zygosaccharomyces rouxii cells have been reported ( 29). Z. bailii is able to degrade benzoate and sorbate ( 51), and species of Penicillium can decarboxylate sorbate to 1,3-pentadiene ( 42). Compared to yeasts, very little is known about specific weak-acid resistance mechanisms in bacteria. Depending on the species, bacteria can induce several systems to counteract a drop in the internal pH when encountering low pH stress. Among others, these include proton pumps, several decarboxylases (lysine, glutamate, and arginine), the production of urease, arginine deiminase, chaperones (e.g., DnaK and GroELS), and sigma factor (SigB, SigM, and RpoS)-mediated responses ( 5, 6, 14, 31, 67). However, the importance of low-pH stress response systems in weak-acid resistance development remains unclear. The gram-positive bacterium Bacillus subtilis is one of the organisms that causes food spoilage, and its growth is inhibited by sorbic acid ( 25). This rod-shaped bacterium commonly lives in the upper layers of soil and is therefore found on crops and in food products. Thus, we investigated the time-resolved genome-wide response of B. subtilis sublethally stressed with potassium sorbate (KS) using DNA microarray technology. We used the complementary methods of hierarchical clustering and T-profiler, adapted for B. subtilis, to analyze the data. Our results indicate that sorbic acid induces responses normally seen upon nutrient limitation. However, mild sorbic acid stress does not lead to the induction of the general stress response (GSR), sporulation, or competence. B. subtilis likely remodels its plasma membrane, possibly to reduce the entry of sorbic acid into the cell. Bacterial strains and growth conditions. All B. subtilis strains used in this study are derivatives of the laboratory wild-type (WT) strain PB2 ( trp2C). PB2, PB153 ( trpC2 sigBΔ 2:: cat), and PB198 ( amyE::P ctc- lacZ) were kindly provided by C. W. Price. Mutant strains ATB002 ( ureC), ATB008 ( sigL), and ATB003 ( padC) were obtained by transformation of strain PB2 with chromosomal DNA of strains SF168U ( trpC2 ureC:: spc) ( 15), QB5505 ( trpC2 sigL:: aphA3) ( 19), and BS783 ( trpC2 padC:: cat) ( 23), respectively. To obtain mutant strains ATB001 ( fabHB), ATB004 ( ycsF), ATB005 ( yhcA), ATB006 ( yhcB), and ATB007 ( yxkJ), WT strain PB2 was transformed with chromosomal DNA of strains YHFBd ( trpC2 fabHB::pMUTIN), YCSFd ( trpC2 ycsF::pMUTIN), YHCAd ( trpC2 yhcA::pMUTIN), YHCBd ( trpC2 yhcB::pMUTIN), and YXKJd ( trpC2 yxkJ::pMUTIN), respectively, which were all received from the Japanese Consortium for Functional Analysis of the B. subtilis Genome ( http://bacillus.genome.ad.jp/). Transformants were selected on Luria-Bertani (LB) agar plates containing appropriate antibiotics after overnight incubation at 37°C. Depending on the strain, the antibiotics used were chloramphenicol (6 μg/ml), erythromycin (0.5 μg/ml), spectinomycin (100 μg/ml), or kanamycin (10 μg/ml). Isolation of chromosomal DNA was performed according to methods described previously by Ward and Zahler ( 70), and transformations were carried out as described previously by Kunst and Rapoport ( 44). B. subtilis strains were cultivated in a defined minimal medium as described previously by Neidhardt et al. ( 56), as modified by Hu et al. ( 37). The medium was buffered with 80 mM 3-( N-morpholino)propanesulfonic acid (MOPS), and the pH was set to 5.9, 6.4, 7.4, or 7.8 with KOH. As carbon and nitrogen sources, 5 mM glucose, 10 mM glutamate, and 10 mM NH 4Cl were used. A fivefold (25 mM) increase in glucose or a 10-fold (100 μM) or a 25-fold (250 μM) increase in iron was used where indicated. All strains were grown exponentially, transferred into a SpectroMax Plus microtiter plate reader (Molecular Devices Corp.) at an optical density at 600 nm (OD 600) of 0.08 (which corresponds to an OD 600 of 0.2 in a 1-cm-path-length spectrophotometer), and stressed with various concentrations of KS ranging from 1.25 to 125 mM or 5 μg/ml cerulenin where indicated. Cells were further cultivated in the microtiter plate reader under rigorous shaking at 37°C for 180 min. All conditions were tested in the microtiter plate reader at least in duplicate, and biologically independent experiments were performed at least twice. Assay of β-galactosidase activity. PB198 ( amyE P ctc- lacZ) was grown exponentially in shake flasks in defined medium at pH 6.4 to an OD 600 of 0.2 and stressed with 3, 7, and 20 mM KS or 0.3 M NaCl. To determine the β-galactosidase activity, 1-ml samples were collected every 15 min for 1 h, frozen in liquid nitrogen, and stored at −20°C until further processing. The β-galactosidase assay was performed as described previously ( 41). Cells were permeabilized using 0.002% sodium dodecyl sulfate and 4% chloroform (final concentrations). LacZ activities were calculated as Miller units ( 48). Preparation of total RNA for transcriptome analysis and real-time reverse transcriptase (RT) PCR. An exponentially growing culture of B. subtilis WT strain PB2 was split into two cultures and inoculated in well-controlled batch fermentors (500-ml working volume) to an OD 600 of 0.05. The cultures were grown at 37°C in defined medium at pH 6.4 with an aeration rate of 0.5 liters/min and vigorous stirring (200 rpm). At an OD 600 of 0.2, one culture was stressed with 3 mM KS. Samples of 20 ml were withdrawn from both the treated and control cultures at 0, 10, 20, 30, 40, and 50 min after the addition of KS. Glucose levels and oxygen consumptions were obtained as described elsewhere previously ( 1). The cells were collected using a vacuum-filtering setup, immediately quenched in liquid nitrogen, and stored at −80°C prior to RNA extraction. The whole procedure took no longer than 50 s. Two biologically independent experiments were performed. Total RNA was isolated as described previously ( 40). Synthesis of labeled cDNA, hybridization, and scanning of the DNA microarrays. Superscript II RT (Invitrogen) was used to synthesize labeled cDNA from total RNA samples by the direct incorporation of Cy3- or Cy5-labeled dUTP into cDNA. The reaction mixture in first-strand buffer contained 12 μg of total RNA; 0.5 μg of random hexamers (GE Healthcare); 400 units of Superscript II RT; 10 mM dithiothreitol; 0.5 mM dATP, dCTP, and dGTP; 0.2 mM dTTP (GE Healthcare); and 0.07 mM Cy3- or Cy5-dUTP (GE Healthcare). Control and sorbate-treated samples were incorporated with Cy3- and Cy5-labeled dUTP, respectively. After cDNA synthesis, the RNA was hydrolyzed using 1.5 μl of 1 M NaOH for 10 min at 70°C. The pH was neutralized with 1.5 μl of 1 M HCl, and the labeled cDNA was purified by using QIAquick PCR purification spin columns (Qiagen). The efficiency of labeling was monitored spectrometrically on a Nanodrop apparatus (Isogen Life Science). The B. subtilis DNA microarrays were constructed as described previously by Keijser et al. ( 40). Each constructed array contained spots in duplicate with 4,100 gene-specific 65-mer oligonucleotides representing 4,100 of the 4,106 protein-coding genes in B. subtilis (as reported for the B. subtilis genome at http://genolist.pasteur.fr/SubtiList/). Hybridization and scanning was performed as described previously ( 40). Microarray data extraction and processing. Quantification of the hybridization signals from both Cy3 and Cy5 channels and background subtractions were carried out with ArrayVision 6.1 software (Imaging Research Inc.). First, the pixels with density values that exceeded four median absolute deviations above the median were removed, and the average of all pixels remaining in the spot was computed for each channel (the artifact-removed [ARM] density values). Second, the local background was calculated for each spot and subtracted from its ARM density value (resulting in the subtracted ARM density value). Spots with a signal-to-noise ratio (subtracted ARM divided by the standard deviation of the local background) smaller than 2.0 in both channels were excluded from further analysis. The remaining data were normalized in J-Express Pro 2.6 software (MolMine AS) using a global locally weighted scatterspot-smoothing algorithm ( 75). To avoid extreme intensity ratios, low-intensity fluorescence data were floored at a value corresponding to a signal-to-noise ratio of 2.0. The data were averaged and log 2 transformed, and missing values were replaced by the average of the closest values. Genes with more than two missing values in the time series were omitted. Since the variation in differential expression measurements depends on the fluorescent signal intensity (smaller variation at higher fluorescence intensity levels and larger variation at lower fluorescence intensity levels), we applied an intensity-dependent method to identify differentially expressed genes ( 74). A sliding window of 50 genes was selected to calculate a Z score from the local mean and standard deviation using the data in the R-I plot [log 10(Cy5 × Cy3) versus log 10(Cy5/Cy3)]. Genes more than 1.96 standard deviations away from the local average (| Z| > 1.96) were considered to be differentially expressed. This corresponds to a confidence level of 95%. Genes that showed significant expression at 0 min were excluded from further analysis, unless the gene showed the opposite significant expression in at least one of the other time points. After the processing of the microarray data, 3,909 genes remained for each time point, 459 of which were found to be significantly expressed. The degree of enrichment or depletion for a specific gene group in the given significantly up- or downregulated genes was quantitatively assessed using a hypergeometric distribution analysis ( 55). Gene groups were considered to be enriched or depleted when the calculated P value was below 0.01. Microarray data analysis. Hierarchical clustering ( 24) of the significantly regulated genes was used to identify groups of genes with similar transcription profiles. In J-Express Pro 2.6 (MolMine AS), all 459 genes showing significant expression during KS treatment were hierarchically clustered using the average linkage (weighted-pair group method using average linkages) clustering method and a Euclidian distance metric. To assess the contribution of the expression of genes from specific gene classes to the total gene expression of all 3,909 genes, we used T-profiler ( 7). T-profiler was adapted for the use of B. subtilis transcriptome data by implementing predefined gene groups from the following sources: the Database of Transcriptional Regulation in Bacillus subtilis (DBTBS) (release May 2006) ( 47), the Kyoto Encyclopedia of Genes and Genomes database (KEGG) (release May 2006) ( 39), the SubtiList database ( 54), and the stringently controlled genes ( 26). Gene groups regulated positively or negatively by a specific transcription factor were named accordingly. For transcription factors acting both as an activator and as a repressor, separate gene groups were made. The composition of the individual gene groups can be found in the above-mentioned sources. Relative quantification of gene expression using real-time RT-PCR. For real-time RT-PCR, RNA isolated for the DNA microarray was used to make cDNA. RT reactions were performed using Superscript II RT (Invitrogen) according to the manufacturer's instructions. Equal amounts of total RNA (5 μg) and 150 ng of random hexamers (GE Healthcare) were used in RT reactions. The amplification and detection of PCR product were performed with the 7300 real-time PCR system (Applied Biosystems). Primer Express 3.0 software (Applied Biosystems) was used to design specific primers (purchased from Isogen Life Science) for real-time PCR (see Table S1 in the supplemental material). Reactions were carried out in a 20-μl mixture consisting of 3 to 9 μM specific primers, 2 μl of 200-fold-diluted cDNA template, and SYBR green PCR master mix (Applied Biosystems). The cycling conditions were as follows: 1 cycle at 50°C for 2 min, 1 cycle at 95°C for 10 min, and 40 cycles at 95°C for 15 s and at 60°C for 1 min. Melting curves were used to monitor the specificity of the reaction. RNA of all time points and independent experiments used in the microarray analysis were analyzed with real-time PCR in duplicate. Because the amplification of the target and reference genes was tested and found to be approximately equal (not shown), the ΔΔ CT method could be used to calculate relative gene expressions ( 46). The expression levels of the investigated genes were determined relative to the untreated reference group. The ratios (2 −ΔΔCT) were calculated and log 2 transformed. The accA gene was used as the internal control, since the expression of this gene was constant under both control and stress conditions in both microarray and real-time PCR experiments. Microarray data accession number. Growth inhibition by sorbic acid can be attributed mainly to the undissociated form of the acid. Weak acids in solution are in equilibrium between their undissociated and dissociated form. To investigate whether undissociated (HS) or dissociated (S −) sorbic acid is responsible for growth inhibition of B. subtilis, the pH dependence of sorbate action was tested on exponentially growing cells in a defined minimal medium. By using a defined and buffered medium, the pH remained stable, and the unwanted presence of weak organic acids in undefined rich media such as LB broth was avoided. B. subtilis WT strain PB2 was grown exponentially in the presence of KS (0 to 40 mM) at pH 5.9, 6.4, 7.4, and 7.8 (Fig. ). Clearly, the reduction in the growth rate is KS concentration dependent. Note that the drop in the OD 600 observed at the end of the exponential phase coincides with the depletion of glucose (confirmed by high-performance liquid chromatography) (our unpublished data), which may be due to a diffusion limitation of oxygen in the microtiter plate experiments. Cells grown into stationary phase in the presence of excess glucose did not show this phenomenon (our unpublished data). Similar qualitative trends of the growth curves were observed for all pH values tested. However, at a higher pH, higher concentrations of KS were needed to similarly affect growth (our unpublished data). We plotted the percentages of growth inhibition against the concentrations of S − and HS and observed that the curves for HS had a clear overlap, in contrast to those for S − (Fig. ). This demonstrates that HS is largely responsible for the growth inhibition and that its effect is pH independent. However, a closer look at the lower concentration range of HS (Fig. , inset) reveals a pH dependence, suggesting that S − also contributes to the growth inhibition although to a much lesser extent. These findings are in general agreement with previously reported observations ( 17, 25). It is noteworthy that sorbic acid stress lowers the maximally obtained OD600 (Fig. ), which suggests a lower yield. In order to investigate this in more detail, we grew cells in well-aerated batch fermentors, stressed them with 3 mM potassium sorbate at pH 6.4, and measured the oxygen consumption and glucose levels. This treatment resulted in a 29% reduction in the growth rate. The values for the calculated yield on oxygen decreased 20%, and those for the calculated yield on glucose decreased 16% during 50 min of sorbic acid treatment. This suggests that sorbate lowers the energetic efficiency of respiration, which may be a consequence of a decreased proton gradient and/or cellular reprogramming. Compared to the untreated control, the glucose flux per cell calculated over this time interval decreased 10%, which indicates that the uptake of glucose and/or its further metabolism is hampered or adjusted by the cell. Time-resolved transcriptome analysis of sorbic acid-treated B. subtilis. To obtain a better understanding of the response of B. subtilis enduring sorbic acid stress, DNA microarray analysis was performed. We studied the changes in gene expression in cells exponentially grown in batch fermentors. Samples were taken at 10, 20, 30, 40, and 50 min after exposure to 3 mM KS at pH 6.4 (29% growth inhibition) and compared to an untreated control. Using a fluorescent signal intensity-dependent method (see Materials and Methods), we identified a total of 459 genes (11.2% of the genome) that were differentially expressed in at least one time point in comparison with the untreated samples (see Table S2 in the supplemental material). We used real-time RT-PCR to validate the results of the microarray and selected seven transcripts representative of the various temporal expression patterns observed (see Table S3 in the supplemental material). Microarray interpretation using hierarchical clustering and T-profiler. To analyze the data obtained from the microarray experiments, we used two complementary methods: hierarchical clustering ( 24), to identify groups of genes (clusters) with similar transcription profiles, and T-profiler ( 7), to determine significantly regulated gene groups. All 459 genes that showed significant changes were hierarchically clustered and divided into four main clusters using an arbitrary distance cutoff, indicated by the dashed blue line in the hierarchical tree (Fig. ). Subdivisions of these clusters are indicated by colored bars and letters next to each cluster. The complete hierarchical clustering, together with all gene names, their descriptions, and functional categories, is available in the supplemental material (see Fig. S1 in the supplemental material). T-profiler ( http://www.science.uva.nl/~boorsma/t-profiler-bacillusnew/), developed originally for the analysis of S. cerevisiae ( 78) and Candida albicans transcriptome data, was adapted for the analysis of genome-wide expression data for B. subtilis (see Materials and Methods). T-profiler optimally uses all data, in contrast to hierarchical clustering, where a cutoff for significance is applied. Importantly, T-profiler transforms transcriptional data of single genes into the behavior of gene groups, reflecting biological processes in cells. All gene groups with significant T values in any time point are presented in the supplemental material (see Tables S4 to S8 in the supplemental material). Global adaptive responses to sorbic acid stress in B. subtilis. B. subtilis has different global adaptive responses that can be induced when it encounters stress or starvation. The GSR, regulated by the sigma factor SigB, is induced by many different types of stress (e.g., glucose starvation, heat, low external pH, salt, and ethanol) and provides the cell with nonspecific, multiple, and preventive stress resistance ( 31). Rather unexpectedly, we found no evidence for the induction of the GSR in our microarray analysis. Groupwise analysis using T-profiler yielded no significant values for the SigB-regulated gene group (see Fig. S2A in the supplemental material), and SigB-regulated genes were not overrepresented among the 256 upregulated genes (2 out of 95 SigB-regulated genes) ( P = 0.080). In addition, the reporter strain PB198 ( amyE::P ctc- lacZ), used to monitor the induction state of the GSR, showed no induction (see Fig. S2B in the supplemental material). Only very strong inhibition of growth (71%) caused by 20 mM KS resulted in increased LacZ activity. A sigB mutant strain showed susceptibility to KS similar to that of the WT strain for all concentrations tested in liquid medium (Table 1). Additionally, long-term stress survival, as tested by spotting 10-fold dilution series of exponentially growing WT and sigB mutant cells on plates containing KS, revealed no difference, even at high KS concentrations (our unpublished data). We conclude that the GSR is not the key response of cells encountering sorbic acid stress. | TABLE 1.Susceptibilities of selected mutant strains to sorbic acid stressa |
When nutrients become limiting at the end of exponential growth and cells enter the stationary phase, the genes of the SigH regulon and the genes repressed by the transition-state regulator AbrB and the early-stationary-phase regulator CodY are activated to adapt to limiting conditions ( 63, 66). T-profiler showed a clear induction of transition-state and early-stationary-phase genes (Fig. ). Indeed, 6 out of 29 ( P = 0.0021) genes of the SigH regulon were significantly induced (Fig. , clusters 2A, 2E, and 4). Also, the repression of CodY- and AbrB-regulated genes was relieved immediately after treatment with sorbic acid (Fig. ). This observation is further supported by the significant induction of 18 out of 42 ( P < 0.0001) genes negatively regulated by CodY and the induction of 15 out of 59 ( P < 0.0001) genes repressed by AbrB as well as the downregulation of 6 out of 20 ( P < 0.0001) genes positively regulated by AbrB. These data indicated that the activity of AbrB decreased transiently in the first 20 min after sorbate stress. To corroborate the derepression of genes negatively regulated by CodY, we analyzed and confirmed the induction of the gene ybgE by real-time RT-PCR (see Table S3 in the supplemental material). This gene is known to be regulated only by CodY ( 52). The induction of transition-state and early-stationary-phase genes suggests that sorbic acid-stressed cells respond as if they experience nutrient limitation. In amino acid-, glucose-, or oxygen-limited cells, the (RelA-mediated) stringent response helps to prevent the waste of scarce nutrients ( 9, 26). Hierarchical clustering revealed that only 8 out of 55 ( P = 0.017) RelA-dependent positive stringent control genes were significantly induced and that 4 out of 86 ( P = 0.79) RelA-dependent negative stringent control genes were significantly repressed. However, upon considering all genes, as is done in T-profiler, significant T values for both (RelA-dependent) positive and negative stringent control gene groups were found (Fig. ). The significant downregulation of gene groups involved in translation and protein synthesis (Fig. ), as well as purine and pyrimidine synthesis (see Tables S4, S6, and S7 in the supplemental material), further reflects the induction of the stringent response and the observed reduction in growth. Nutrient limitation can trigger the onset of sporulation and the development of competence ( 22, 64). However, we found no concerted induction of genes controlled by the key competence regulator ComK (Fig. ), of other known regulators involved in competence, or of the functional category transformation/competence (our unpublished data). Sporulation was also not induced by sorbic acid. We found a clear downregulation of the KipR-regulated genes (Fig. , cluster 3E, and Fig. ), which may indicate the release of the “brake” on the sporulation-regulatory cascade. KipR is the negative regulator of the ycsFGI-kipIAR-ycsK operon ( 69). Inactivation of ycsF, the first gene of this operon, did not lead to an altered susceptibility to sorbic acid (Table 1). KipI is an inhibitor of KinA, the primary kinase in the phosphorelay necessary for the phosphorylation of the key transcription factor Spo0A that regulates the initiation of sporulation ( 64). However, no induction of genes positively regulated by the sporulation master regulator Spo0A was observed (Fig. ). There was a brief downregulation of genes repressed by Spo0A (8 out of 17) ( P < 0.0001), but this can be fully explained by the downregulation of the arginine biosynthesis genes controlled by AbrB and the arginine metabolism regulator AhrC (Fig. , cluster 3D; see Table S4 in the supplemental material). Furthermore, we did not observe an induction of other sporulation gene groups regulated by SpoIIID or SpoVT or belonging to the SigE, SigF, SigG, and SigK regulon (our unpublished data). Compared to the untreated control, no increased spore counts were detected in sorbic acid-stressed cultures (our unpublished data). In conclusion, we see no signs of the development of competence or induction of sporulation. Responses to counteract intracellular acidification. One of the most prominent effects of weak acids on the microbial cell may be cytosolic acidification. We discovered a brief but significant induction of the genes coding for the class I heat shock proteins GroES and GroEL (Fig. , clusters 1A and 2A). Chaperones are known to play a role in the maintenance of protein folding at a lowered pH, and the induction of groEL upon acid stress in lactobacilli, Streptococcus mutans, Clostridium perfringens, and Listeria monocytogenes was described previously ( 14). We investigated cellular functions that may counteract the putative pH drop and maintain pH homeostasis. We found no significant induction of ATPases and components of the respiratory chain. However, the capacity of the cell's ATPases and respiratory chain may well be sufficient to maintain pH homeostasis without regulation at the transcription level. We did observe the altered expression of three out of six genes ( P < 0.0001) belonging to the Gene Ontology group involved in the regulation of pH (GO:0006885). The following genes were significantly downregulated: nhaC (Na +/H + antiporter), yjbQ (similar to the Na +/H + antiporter), and yuiF (similar to hypothetical proteins). A highly upregulated gene coding for a proton symporter that transports both citrate and malate was yxkJ ( 43). The greater-than-12-fold induction of this gene (Fig. , cluster 4) might indicate its involvement in sorbic acid tolerance. However, a strain mutated for yxkJ was not more sensitive to sorbic acid than the WT (Table 1). Long-term stress survival on plates containing KS did not reveal sensitivity either (our unpublished data). Interestingly, we found all three genes of the ureABC operon, which code for the structural components of urease ( 15), in cluster 4 with the most upregulated genes (Fig. ). The regulation of this operon is complex, as it is regulated by PucR, TnrA, CodY, and GlnR, and expression is dependent on SigA and SigH ( 10, 72). In addition, the induction of ureA is also RelA dependent ( 26). Thus, the induction of the ureABC operon may be explained by the derepression of CodY-regulated genes and the induction of the stringent response (Fig. ). The upregulation of this operon might be involved in counteracting the possible acidification of the cytosol through the production of basic ammonia, as in some bacterial pathogens, ureases are known to facilitate survival in acidic environments ( 13, 49). However, a urease mutant was not KS sensitive in liquid cultures (Table 1) or solid cultures (our unpublished data). Although sorbic acid is not a phenolic acid, we surprisingly found strong upregulation of the phenolic acid decarboxylase padC ( 12) in cluster 4 (Fig. ). The genes with unknown function yveF and yveG, which lie upstream of padC in the genome, were also strongly induced and present in this cluster, which may indicate the coregulation of these three genes. The inactivation of padC did not lead to KS sensitivity. Surprisingly, the mutant strain showed a clear resistant phenotype at low KS concentrations both in liquid cultures (Table 1) and on solid cultures (our unpublished data). Responses to cope with a decreased proton gradient. The proposed influx of protons mediated by the diffusion of the weak acid over the cell membrane may lead to a decreased proton motive force. Also, the expulsion of protons to maintain pH homeostasis by ATPases and/or the respiratory chain could lead to a higher demand on energy resources, which should have a negative effect on the yield. The latter was indeed measured as reported above. Interestingly, we observed major changes in genes involved in carbon metabolism upon sorbic acid stress (Fig. ). Besides the induction of carbohydrate metabolism as well as fructose and mannose metabolism gene groups, we found a strong upregulation of genes negatively regulated by the carbon catabolite control protein CcpA after 20 min of sorbic acid exposure. Carbon catabolite derepression normally occurs when the preferred carbon source (glucose, fructose, and mannose) is depleted ( 21). However, glucose was still present in the medium and consumed by the cells during sorbic acid treatment (confirmed by high-performance liquid chromatography) (our unpublished data), although the glucose flux per cell decreased by 10%. Noticeably, a fivefold (25 mM) increase in the glucose concentration in medium did not alter the growth-inhibitory effect of sorbic acid (Table 2). | TABLE 2.Growth medium supplemented with either additional glucose or iron does not increase tolerance to sorbic acid stressa |
The expression of genes dependent on SigL (involved in alternative carbon and nitrogen metabolism) also produced significant T values after 20 min of sorbic acid exposure (Fig. ). The induction of the SigL-dependent levD gene was confirmed by real-time RT-PCR (see Table S3 in the supplemental material). However, the inactivation of the sigL gene did not change the susceptibility to sorbic acid compared to the WT (Table 1). Genes of the tricarboxylic acid (TCA) cycle were also induced gradually during sorbic acid stress (Fig. ). The TCA cycle can be fed by additional acetyl coenzyme A produced from the breakdown of the catabolic products acetate and acetoin. Indeed, we found a significant induction of the acetyl coenzyme A synthetase acsA, which is responsible for the degradation of acetate ( 30), and acuABC, which is involved in the activation of AcsA ( 28). Also, the genes acoA and acoR, involved in the breakdown of acetoin ( 2, 38), were significantly upregulated. In addition to these carbon metabolism-related gene functions, we observed a short downregulation followed by a strong upregulation of the gene group regulated by the central iron-regulatory protein Fur (ferric uptake repressor) (Fig. ). This transient profile was also clearly seen in clusters 1D and 2C of the hierarchical clustering, containing many Fur-regulated genes (Fig. ). In total, we found 15 out of 36 ( P < 0.0001) Fur-regulated genes significantly expressed in our analysis with similar transient expression profiles. The Fur regulon is normally derepressed under iron-limiting or anaerobic conditions ( 53, 76). We tested whether increased iron concentrations in the medium would result in an increase in the stress tolerance against sorbic acid. A 10-fold (100 μM) or 25-fold (250 μM) increase of iron in the medium could not decrease the growth inhibition by sorbic acid (Table 2). These combined observations suggest that although nutrients in the cell's environment are not limiting, the cells respond as if nutrients are scarce. Sorbic acid influences the biogenesis of the cell envelope. Since HS can dissolve in the cell membrane and is thought to affect membrane integrity, we expected to see a membrane-remodeling response or an adaptation of the cell envelope to limit HS entry. Functions associated with the cell surface or transport are controlled by extracytoplasmic function sigma factors, like SigM, SigW, and SigX ( 32). Indeed, the gene groups regulated by extracytoplasmic function sigma factors SigW and SigX showed altered expression upon sorbic acid treatment (Fig. ). We also detected a similar repression pattern for genes regulated by YvrH (Fig. ). YvrH is involved in the maintenance of the cell surface state through transcriptional regulation and controls genes containing a promoter site for SigX and sigX itself ( 62). The repression of SigW- and SigX-regulated genes upon sorbic acid stress suggests their expression under control conditions. This finding was corroborated by the absolute fluorescence levels. Although SigW and SigX are activated upon cell wall- and membrane-perturbing stresses, we did not have any indication that our control cells suffered stress. This was supported by a μ max of 0.93, no induction of the GSR, and no formation of heat-resistant spores in the control culture (our unpublished results). T-profiler analysis showed a strong induction of genes involved in the metabolism of lipids directly after sorbic acid exposure (Fig. ). Most fatty acid biosynthesis ( fab) genes, negatively regulated by FapR (YlpC) ( 61), were significantly upregulated. The fab initiation genes fabHB and fabD and elongation genes fabF, fabG, and fabI (Fig. , clusters 2A, 2E, and 4) showed significant induction, with similar expression profiles. The strong upregulation (19-fold) of fabHB was confirmed using RT-PCR (Table 2). In addition, we observed the significant induction of genes regulated by BkdR in both T-profiler analysis (Fig. ) and hierarchical clustering (Fig. , clusters 2D and E). BkdR is the activator of the bkd operon, consisting of seven genes involved in the synthesis of precursor molecules for branched-chain fatty acids ( 18). The induction of both the fab genes and the bkd operon in sorbic acid-treated cells will likely increase the number of long-chain and branched-chain fatty acids in the membrane ( 20, 61). We investigated the role of fatty acid synthesis in the resistance to sorbic acid stress by growing a β-ketoacyl-acyl carrier protein (ACP) synthase III mutant in the presence of sorbic acid. Strain ATB001 ( fabHB) did not reveal a sensitive phenotype (Table 1). The antibiotic cerulenin inhibits the β-ketoacyl-ACP synthases (FabHA, FabHB, and FabF) of the fatty acid chain elongation step ( 16). In addition, cerulenin induces the FapR regulon ( 60). Remarkably, the simultaneous addition of sorbic acid and cerulenin to exponentially growing WT cells significantly decreased the inhibitory effect of the antibiotic compared to cultures treated with cerulenin alone (Fig. ). This result indicates that the addition of KS may lead to a membrane adaptation that alters the cell's sensitivity to cerulenin. Possible extrusion of the anion. S. cerevisiae uses the pump Pdr12 to extrude the sorbate anion to prevent accumulation ( 35). Interestingly, we found the gene yhcA to be highly upregulated in the presence of KS (Fig. , cluster 4). YhcA is a multidrug resistance transporter homologue of the major facilitator superfamily. To investigate the role of this gene in sorbate stress tolerance, we tested mutant strain ATB005 ( yhcA) for sorbic acid sensitivity. Surprisingly, the yhcA mutant strain showed clear sorbate resistance, especially at low KS concentrations in liquid medium (Table 1) and on solid cultures (our unpublished data). yhcB is located downstream of yhcA in the genome (similar to the trp repressor binding protein). This gene also showed strong upregulation (Fig. , cluster 4) and was tested for sorbate stress tolerance. Mutant strain ATB006 ( yhcB) showed no altered sorbic acid susceptibility (Table 1). These data illustrate that the resistance of the yhcA mutant strain is caused by the inactivation of yhcA itself and not by affected downstream genes. Our study shows that mainly the undissociated form of sorbic acid (HS) is responsible for the inhibition of growth of B. subtilis (Fig. ). These data support the model that suggests that the neutral form of the acid enters the cell, where it dissociates, possibly acidifying the cytosol and likely contributing to its inhibitory effect ( 5, 17). Maintaining the internal pH is of crucial importance for the proper functioning of the cell. We observed several responses that may counteract the possible intracellular acidification caused by the dissociation of HS (e.g., the induction of the urease operon and the phenolic acid decarboxylase padC). However, thus far, we have no indication that these responses are functional in HS tolerance. Besides possible acidification of the cytosol, the entry of protons will lower the proton gradient. Consequently, this will reduce the cell's efficiency to produce ATP, which is supported by the observed lower yields on glucose and oxygen upon sorbic acid stress. The transcriptome analysis clearly revealed that cells stressed with sorbic acid respond as if they encounter nutrient limitation. We observed a clear induction of transition-state and early-stationary-phase genes regulated by AbrB, SigH, and CodY as well as an induction of the stringent response mediated by RelA (Fig. ). The derepression of the carbon catabolite control genes, regulated by CcpA, major changes in carbon metabolism, and the induction of the Fur regulon further suggest that even though glucose is still abundant in the medium, the cells experience nutrient limitation(s) (Fig. ). This may be caused by effects of the acid on glucose uptake systems but may also be the result of an altered intracellular environment and effects on enzymes involved in the subsequent glucose-phosphate catabolism. This possibility is supported by the decreased glucose flux seen in sorbic acid-treated cells. Future studies should aim at (short-term) measurements of intracellular metabolites and oxygen consumption in the absence and presence of sorbic acid. The addition of excess glucose or iron to the medium did not change sorbate susceptibility (Table 2), which seems to point out that limitations in the medium are not the cause of the nutrient limitation responses of the cell. Sorbic acid did not induce sporulation, competence (Fig. ), or the GSR (see Fig. S2 in the supplemental material). We conclude that the limiting conditions caused by 3 mM sorbic acid are not severe enough to switch on sporulation, competence, or the GSR or that other factors inhibit these adaptation pathways. Zhang and Haldenwang ( 79) recently showed that the induction of the GSR by nutritional stress is preceded by a drop in ATP levels. The absence of a GSR induction upon (mild) sorbic acid stress might indicate that ATP levels are not severely affected. The observed active state of the SigW and SigX regulon in control B. subtilis cells grown in defined (MOPS-based) medium is somehow surprising. Note that we opted for the use of an established Bacillus medium (see, e.g., references 37 and 73) and noted no signs of stress (see Results). The observed downregulation of SigW- and SigX-regulated genes upon sorbic acid stress may lead to a blockage of cell envelope remodeling and, consequently, altered cell envelope composition. Immediately after sorbic acid exposure, a clear activation of genes involved in the metabolism of lipids was observed (Fig. ). The induction of both the fab genes and the bkd operon will likely increase the number of long-chain and branched-chain fatty acids in the membrane ( 20, 61). An elongation of fatty acids in the membrane might lower the diffusion rate of sorbic acid into the cell. The simultaneous addition of more branched-chain fatty acids into the membrane through the action of the bkd operon will likely preserve fluidity. The inactivation of fabHB did not result in a sensitive phenotype for sorbic acid (Table 1). Noticeably, since fabHA codes for a β-ketoacyl-ACP synthase III as well, this gene is likely to take over the function of fabHB in the mutant strain tested. Recently, Thomaides et al. ( 68) illustrated that the fabHA fabHB double mutant is nonviable. The combined treatment of cultures with both KS and cerulenin, the inhibitor of the β-ketoacyl-ACP synthases, significantly lowered the inhibitory effect of cerulenin alone (Fig. ), supporting the inferred remodeling of the membrane in response to sorbic acid. Alteration of the plasma membrane composition upon sorbic acid stress in Z. rouxii was described previously ( 29). In addition, S. cerevisiae and Z. bailii cells adapted to benzoic acid have a reduced permeability to benzoate ( 33, 71). We suggest that in B. subtilis, the observed induction of the fab genes and the bkd operon is a direct response to sorbic acid that will likely alter the membrane composition such that the diffusion rate of the organic molecule over the membrane is lowered. Accumulation of the anion may also cause harm to the cell ( 17, 34). The strong upregulation of yhcA, encoding a major facilitator superfamily multidrug resistance transporter homologue, suggested a potential anion extrusion mechanism. The inactivation of yhcA, however, revealed a clear resistant phenotype upon stress with low concentrations of KS in defined minimal liquid medium (Table 1) and on solid cultures (our unpublished data) under our test conditions. In theory, the extrusion of S − by YhcA may create a futile cycle and deplete cellular energy pools, since the anion can reassociate in the extracellular environment with a proton and diffuse back into the cell, thereby reducing the proton gradient further. If such futile cycling indeed occurs, it may significantly contribute to the observed resistance of the yhcA mutant strain when tested in defined minimal medium. Alternative explanations, in which the YhcA protein would be the site at which sorbic acid enters the cells inadvertently and its deletion would thus lower the sorbic acid sensitivity of the cells, are also possible. Such events would be analogous to the recently reported acetic acid stress resistance mechanisms in yeast involving the Fps1 aquaglyceroporin protein ( 50). Interestingly, we have preliminary data showing that in rich medium, the yhcA mutant strain is more sensitive to KS on solid cultures (our unpublished data). Future studies should address the function of YhcA and its relationship to the differences observed when minimal or rich medium is used. In conclusion, sorbic acid induces responses normally seen upon nutrient limitation. Therefore, we suggest that the entry of the protonated acid and the subsequent lowering of the proton gradient increase the demand for energy. The cells do not (and likely cannot) increase the uptake rate of glucose and consequently experience nutrient limitation. The upregulation of the TCA cycle and the utilization of acetate and acetoin may provide sufficient energy to maintain growth but at a lower rate. Finally, the plasma membrane is remodeled, likely in an attempt to reduce the entry of sorbic acid into the cell. Whether the observed responses in B. subtilis upon sorbic acid stress are representative of weak acids in general remains to be elucidated. We thank Chester W. Price, Susan H. Fisher, Yasutaro Fujita, Kunio Yamane, Yoshito Sadaie, Haike Antelmann, and Georges Rapoport for sending strains. We acknowledge Jurgo Verkooijen and Tessa Dillerop-van der Hoeven of the Microarray Department of Amsterdam for the hybridization and scanning of the microarray slides, Muus de Haan for assistance with real-time RT-PCR, and Reuben Smith and Rik van Arnhem for technical assistance. Furthermore, we thank Klaas J. Hellingwerf for critically reading the manuscript and Joost Teixeira de Mattos as well as the anonymous reviewers for thoughtful suggestions. 1. Alexeeva, S., K. J. Hellingwerf, and M. J. Teixeira de Mattos. 2003. Requirement of ArcA for redox regulation in Escherichia coli under microaerobic but not anaerobic or aerobic conditions. J. Bacteriol. 185204-209. [PubMed] 2. Ali, N. O., J. Bignon, G. Rapoport, and M. Debarbouille. 2001. Regulation of the acetoin catabolic pathway is controlled by sigma L in Bacillus subtilis. J. Bacteriol. 1832497-2504. [PubMed] 3. Azukas, J. J., R. N. Costilow, and H. L. Sadoff. 1961. Inhibition of alcoholic fermentation by sorbic acid. J. Bacteriol. 81189-194. [PubMed] 4. Bauer, B. E., D. Rossington, M. Mollapour, Y. Mamnun, K. Kuchler, and P. W. Piper. 2003. Weak organic acid stress inhibits aromatic amino acid uptake by yeast, causing a strong influence of amino acid auxotrophies on the phenotypes of membrane transporter mutants. Eur. J. Biochem. 2703189-3195. [PubMed] 5. Beales, N. 2004. Adaptation of microorganisms to cold temperatures, weak acid preservatives, low pH, and osmotic stress: a review. Compr. Rev. Food Sci. Food Saf. 31-20. 6. Bearson, S., B. Bearson, and J. W. Foster. 1997. Acid stress responses in enterobacteria. FEMS Microbiol. Lett. 147173-180. [PubMed] 7. Boorsma, A., B. C. Foat, D. Vis, F. Klis, and H. J. Bussemaker. 2005. T-profiler: scoring the activity of predefined groups of genes using gene expression data. Nucleic Acids Res. 33W592-W595. [PubMed] 8. Bracey, D., C. D. Holyoak, and P. J. Coote. 1998. Comparison of the inhibitory effect of sorbic acid and amphotericin B on Saccharomyces cerevisiae: is growth inhibition dependent on reduced intracellular pH? J. Appl. Microbiol. 851056-1066. [PubMed] 9. Braeken, K., M. Moris, R. Daniels, J. Vanderleyden, and J. Michiels. 2006. New horizons for (p)ppGpp in bacterial and plant physiology. Trends Microbiol. 1445-54. [PubMed] 10. Brandenburg, J. L., L. V. Wray, Jr., L. Beier, H. Jarmer, H. H. Saxild, and S. H. Fisher. 2002. Roles of PucR, GlnR, and TnrA in regulating expression of the Bacillus subtilis ure P3 promoter. J. Bacteriol. 1846060-6064. [PubMed] 11. Brown, M. H., and I. R. Booth. 1991. Acidulants and low pH, p. 22-43. In N. J. Russell and G. W. Gould (ed.), Food preservatives. Blackie, Glasgow, United Kingdom. 12. Cavin, J. F., V. Dartois, and C. Divies. 1998. Gene cloning, transcriptional analysis, purification, and characterization of phenolic acid decarboxylase from Bacillus subtilis. Appl. Environ. Microbiol. 641466-1471. [PubMed] 13. Collins, C. M., and S. E. D'Orazio. 1993. Bacterial ureases: structure, regulation of expression and role in pathogenesis. Mol. Microbiol. 9907-913. [PubMed] 14. Cotter, P. D., and C. Hill. 2003. Surviving the acid test: responses of gram-positive bacteria to low pH. Microbiol. Mol. Biol. Rev. 67429-453. [PubMed] 15. Cruz-Ramos, H., P. Glaser, L. V. Wray, Jr., and S. H. Fisher. 1997. The Bacillus subtilis ureABC operon. J. Bacteriol. 1793371-3373. [PubMed] 16. D'Agnolo, G., I. S. Rosenfeld, J. Awaya, S. Omura, and P. R. Vagelos. 1973. Inhibition of fatty acid synthesis by the antibiotic cerulenin. Specific inactivation of beta-ketoacyl-acyl carrier protein synthetase. Biochim. Biophys. Acta 326155-156. [PubMed] 17. Davidson, P. M. 2001. Chemical preservatives and natural antimicrobial compounds, p. 593-627. In M. P. Doyle, L. R. Beuchat, and T. J. Montville (ed.), Food microbiology: fundamentals and frontiers, 2nd ed. ASM Press, Washington, DC. 18. Debarbouille, M., R. Gardan, M. Arnaud, and G. Rapoport. 1999. Role of bkdR, a transcriptional activator of the sigL-dependent isoleucine and valine degradation pathway in Bacillus subtilis. J. Bacteriol. 1812059-2066. [PubMed] 19. Debarbouille, M., I. Martin-Verstraete, F. Kunst, and G. Rapoport. 1991. The Bacillus subtilis sigL gene encodes an equivalent of sigma 54 from gram-negative bacteria. Proc. Natl. Acad. Sci. USA 889092-9096. [PubMed] 20. de Mendoza, D., G. E. Schujman, and P. S. Aguilar. 2002. Biosynthesis and function of membrane lipids, p. 43-55. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and its closest relatives: from genes to cells. ASM Press, Washington, DC. 21. Deutscher, J., A. Galinier, and I. Martin-Verstraete. 2002. Carbohydrate uptake and metabolism, p. 129-150. In A. L. Sonenshein, J. A. Hoch, and R. Losick (ed.), Bacillus subtilis and its closest relatives: from genes to cells. ASM Press, Washington, DC. 22. Dubnau, D. 1991. Genetic competence in Bacillus subtilis. Microbiol. Rev. 55395-424. [PubMed] 23. Duy, N. V., U. Mäder, N. P. Tran, J. F. Cavin, L. T. Tam, D. Albrecht, M. Hecker, and H. Antelmann. 2007. The proteome and transcriptome analysis of Bacillus subtilis in response to salicylic acid. Proteomics 7698-710. [PubMed] 24. Eisen, M. B., P. T. Spellman, P. O. Brown, and D. Botstein. 1998. Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 9514863-14868. [PubMed] 25. Eklund, T. 1983. The antimicrobial effect of dissociated and undissociated sorbic acid at different pH levels. J. Appl. Bacteriol. 54383-389. [PubMed] 26. Eymann, C., G. Homuth, C. Scharf, and M. Hecker. 2002. Bacillus subtilis functional genomics: global characterization of the stringent response by proteome and transcriptome analysis. J. Bacteriol. 1842500-2520. [PubMed] 27. Freese, E., C. W. Sheu, and E. Galliers. 1973. Function of lipophilic acids as antimicrobial food additives. Nature 241321-325. [PubMed] 28. Gardner, J. G., F. J. Grundy, T. M. Henkin, and J. C. Escalante-Semerena. 2006. Control of acetyl-coenzyme A synthetase (AcsA) activity by acetylation/deacetylation without NAD+ involvement in Bacillus subtilis. J. Bacteriol. 1885460-5468. [PubMed] 29. Golden, D. A., L. R. Beuchat, and H. L. Hitchcock. 1994. Changes in fatty acid composition of various lipid components of Zygosaccharomyces rouxii as influenced by solutes, potassium sorbate and incubation temperature. Int. J. Food Microbiol. 21293-303. [PubMed] 30. Grundy, F. J., D. A. Waters, T. Y. Takova, and T. M. Henkin. 1993. Identification of genes involved in utilization of acetate and acetoin in Bacillus subtilis. Mol. Microbiol. 10259-271. [PubMed] 31. Hecker, M., and U. Volker. 2001. General stress response of Bacillus subtilis and other bacteria. Adv. Microb. Physiol. 4435-91. [PubMed] 32. Helmann, J. D. 2002. The extracytoplasmic function (ECF) sigma factors. Adv. Microb. Physiol. 4647-110. [PubMed] 33. Henriques, M., C. Quintas, and M. C. Loureiro-Dias. 1997. Extrusion of benzoic acid in Saccharomyces cerevisiae by an energy-dependent mechanism. Microbiology 1431877-1883. [PubMed] 34. Hirshfield, I. N., S. Terzulli, and C. O'Byrne. 2003. Weak organic acids: a panoply of effects on bacteria. Sci. Prog. 86245-269. [PubMed] 35. Holyoak, C. D., D. Bracey, P. W. Piper, K. Kuchler, and P. J. Coote. 1999. The Saccharomyces cerevisiae weak-acid-inducible ABC transporter Pdr12 transports fluorescein and preservative anions from the cytosol by an energy-dependent mechanism. J. Bacteriol. 1814644-4652. [PubMed] 36. Holyoak, C. D., M. Stratford, Z. McMullin, M. B. Cole, K. Crimmins, A. J. Brown, and P. J. Coote. 1996. Activity of the plasma membrane H+-ATPase and optimal glycolytic flux are required for rapid adaptation and growth of Saccharomyces cerevisiae in the presence of the weak-acid preservative sorbic acid. Appl. Environ. Microbiol. 623158-3164. [PubMed] 37. Hu, P., T. Leighton, G. Ishkhanova, and S. Kustu. 1999. Sensing of nitrogen limitation by Bacillus subtilis: comparison to enteric bacteria. J. Bacteriol. 1815042-5050. [PubMed] 38. Huang, M., F. B. Oppermann-Sanio, and A. Steinbuchel. 1999. Biochemical and molecular characterization of the Bacillus subtilis acetoin catabolic pathway. J. Bacteriol. 1813837-3841. [PubMed] 39. Kanehisa, M., S. Goto, M. Hattori, K. F. Aoki-Kinoshita, M. Itoh, S. Kawashima, T. Katayama, M. Araki, and M. Hirakawa. 2006. From genomics to chemical genomics: new developments in KEGG. Nucleic Acids Res. 34D354-D357. [PubMed] 40. Keijser, B. J., A. Ter Beek, H. Rauwerda, F. Schuren, R. Montijn, H. van der Spek, and S. Brul. 2007. Analysis of temporal gene expression during Bacillus subtilis spore germination and outgrowth. J. Bacteriol. 1893624-3634. [PubMed] 41. Kenney, T. J., and C. P. Moran, Jr.1987. Organization and regulation of an operon that encodes a sporulation-essential sigma factor in Bacillus subtilis. J. Bacteriol. 1693329-3339. [PubMed] 42. Kinderlerer, J. L., and P. V. Hatton. 1990. Fungal metabolites of sorbic acid. Food Addit. Contam. 7657-669. [PubMed] 43. Krom, B. P., R. Aardema, and J. S. Lolkema. 2001. Bacillus subtilis YxkJ is a secondary transporter of the 2-hydroxycarboxylate transporter family that transports l-malate and citrate. J. Bacteriol. 1835862-5869. [PubMed] 44. Kunst, F., and G. Rapoport. 1995. Salt stress is an environmental signal affecting degradative enzyme synthesis in Bacillus subtilis. J. Bacteriol. 1772403-2407. [PubMed] 45. Lambert, R. J., and M. Stratford. 1999. Weak-acid preservatives: modelling microbial inhibition and response. J. Appl. Microbiol. 86157-164. [PubMed] 46. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(−delta delta C(T)) method. Methods 25402-408. [PubMed] 47. Makita, Y., M. Nakao, N. Ogasawara, and K. Nakai. 2004. DBTBS: database of transcriptional regulation in Bacillus subtilis and its contribution to comparative genomics. Nucleic Acids Res. 32D75-D77. [PubMed] 48. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. 49. Mobley, H. L., M. D. Island, and R. P. Hausinger. 1995. Molecular biology of microbial ureases. Microbiol. Rev. 59451-480. [PubMed] 50. Mollapour, M., and P. W. Piper. 2007. Hog1 mitogen-activated protein kinase phosphorylation targets the yeast fps1 aquaglyceroporin for endocytosis, thereby rendering cells resistant to acetic acid. Mol. Cell. Biol. 276446-6456. [PubMed] 51. Mollapour, M., and P. W. Piper. 2001. The ZbYME2 gene from the food spoilage yeast Zygosaccharomyces bailii confers not only YME2 functions in Saccharomyces cerevisiae, but also the capacity for catabolism of sorbate and benzoate, two major weak organic acid preservatives. Mol. Microbiol. 42919-930. [PubMed] 52. Molle, V., Y. Nakaura, R. P. Shivers, H. Yamaguchi, R. Losick, Y. Fujita, and A. L. Sonenshein. 2003. Additional targets of the Bacillus subtilis global regulator CodY identified by chromatin immunoprecipitation and genome-wide transcript analysis. J. Bacteriol. 1851911-1922. [PubMed] 53. Moore, C. M., and J. D. Helmann. 2005. Metal ion homeostasis in Bacillus subtilis. Curr. Opin. Microbiol. 8188-195. [PubMed] 54. Moszer, I., L. M. Jones, S. Moreira, C. Fabry, and A. Danchin. 2002. SubtiList: the reference database for the Bacillus subtilis genome. Nucleic Acids Res. 3062-65. [PubMed] 55. Motulsky, H. 1995. Intuitive biostatistics. Oxford University Press, New York, NY. 56. Neidhardt, F. C., P. L. Bloch, and D. F. Smith. 1974. Culture medium for enterobacteria. J. Bacteriol. 119736-747. [PubMed] 57. Piper, P., C. O. Calderon, K. Hatzixanthis, and M. Mollapour. 2001. Weak acid adaptation: the stress response that confers yeasts with resistance to organic acid food preservatives. Microbiology 1472635-2642. [PubMed] 58. Ronning, I. E., and H. A. Frank. 1987. Growth inhibition of putrefactive anaerobe 3679 caused by stringent-type response induced by protonophoric activity of sorbic acid. Appl. Environ. Microbiol. 531020-1027. [PubMed] 59. Salmond, C. V., R. G. Kroll, and I. R. Booth. 1984. The effect of food preservatives on pH homeostasis in Escherichia coli. J. Gen. Microbiol. 1302845-2850. [PubMed] 60. Schujman, G. E., K. H. Choi, S. Altabe, C. O. Rock, and D. de Mendoza. 2001. Response of Bacillus subtilis to cerulenin and acquisition of resistance. J. Bacteriol. 1833032-3040. [PubMed] 61. Schujman, G. E., L. Paoletti, A. D. Grossman, and D. de Mendoza. 2003. FapR, a bacterial transcription factor involved in global regulation of membrane lipid biosynthesis. Dev. Cell 4663-672. [PubMed] 62. Serizawa, M., K. Kodama, H. Yamamoto, K. Kobayashi, N. Ogasawara, and J. Sekiguchi. 2005. Functional analysis of the YvrGHb two-component system of Bacillus subtilis: identification of the regulated genes by DNA microarray and northern blot analyses. Biosci. Biotechnol. Biochem. 692155-2169. [PubMed] 63. Sonenshein, A. L. 2005. CodY, a global regulator of stationary phase and virulence in gram-positive bacteria. Curr. Opin. Microbiol. 8203-207. [PubMed] 64. Sonenshein, A. L. 2000. Control of sporulation initiation in Bacillus subtilis. Curr. Opin. Microbiol. 3561-566. [PubMed] 65. Stratford, M., and P. A. Anslow. 1998. Evidence that sorbic acid does not inhibit yeast as a classic ‘weak acid preservative. ’ Lett. Appl. Microbiol. 27203-206. [PubMed] 66. Strauch, M. A., and J. A. Hoch. 1993. Transition-state regulators: sentinels of Bacillus subtilis post-exponential gene expression. Mol. Microbiol. 7337-342. [PubMed] 67. Thackray, P. D., and A. Moir. 2003. SigM, an extracytoplasmic function sigma factor of Bacillus subtilis, is activated in response to cell wall antibiotics, ethanol, heat, acid, and superoxide stress. J. Bacteriol. 1853491-3498. [PubMed] 68. Thomaides, H. B., E. J. Davison, L. Burston, H. Johnson, D. R. Brown, A. C. Hunt, J. Errington, and L. Czaplewski. 2007. Essential bacterial functions encoded by gene pairs. J. Bacteriol. 189591-602. [PubMed] 69. Wang, L., R. Grau, M. Perego, and J. A. Hoch. 1997. A novel histidine kinase inhibitor regulating development in Bacillus subtilis. Genes Dev. 112569-2579. [PubMed] 70. Ward, J. B., Jr., and S. A. Zahler. 1973. Genetic studies of leucine biosynthesis in Bacillus subtilis. J. Bacteriol. 116719-726. [PubMed] 71. Warth, A. D. 1989. Transport of benzoic and propanoic acids by Zygosaccharomyces bailii. J. Gen. Microbiol. 1351383-1390. 72. Wray, L. V., Jr., A. E. Ferson, and S. H. Fisher. 1997. Expression of the Bacillus subtilis ureABC operon is controlled by multiple regulatory factors including CodY, GlnR, TnrA, and Spo0H. J. Bacteriol. 1795494-5501. [PubMed] 73. Wray, L. V., Jr., and S. H. Fisher. 2007. Functional analysis of the carboxy-terminal region of Bacillus subtilis TnrA, a MerR family protein. J. Bacteriol. 18920-27. [PubMed] 74. Yang, I. V., E. Chen, J. P. Hasseman, W. Liang, B. C. Frank, S. Wang, V. Sharov, A. I. Saeed, J. White, J. Li, N. H. Lee, T. J. Yeatman, and J. Quackenbush. 2002. Within the fold: assessing differential expression measures and reproducibility in microarray assays. Genome Biol. 3research0062. [PubMed] 75. Yang, Y. H., S. Dudoit, P. Luu, D. M. Lin, V. Peng, J. Ngai, and T. P. Speed. 2002. Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res. 30e15. [PubMed] 76. Ye, R. W., W. Tao, L. Bedzyk, T. Young, M. Chen, and L. Li. 2000. Global gene expression profiles of Bacillus subtilis grown under anaerobic conditions. J. Bacteriol. 1824458-4465. [PubMed] 77. York, G. K., and R. H. Vaughn. 1964. Mechanisms in the inhibition of microorganisms by sorbic acid. J. Bacteriol. 88411-417. [PubMed] 78. Zakrzewska, A., A. Boorsma, S. Brul, K. J. Hellingwerf, and F. M. Klis. 2005. Transcriptional response of Saccharomyces cerevisiae to the plasma membrane-perturbing compound chitosan. Eukaryot. Cell 4703-715. [PubMed] 79. Zhang, S., and W. G. Haldenwang. 2005. Contributions of ATP, GTP, and redox state to nutritional stress activation of the Bacillus subtilis σB transcription factor. J. Bacteriol. 1877554-7560. [PubMed]
|
This article has been 
