A Caulobacter crescentus Extracytoplasmic Function Sigma Factor Mediating the Response to Oxidative Stress in Stationary Phase
http://www.100md.com
《细菌学杂志》
Departamento de Bioquímica, Instituto de Química, Universidade de So Paulo, So Paulo, Brazil
ABSTRACT
Alternative sigma factors of the extracytoplasmic function (ECF) subfamily are important regulators of stress responses in bacteria and have been implicated in the control of homeostasis of the extracytoplasmic compartment of the cell. This work describes the characterization of sigF, encoding 1 of the 13 members of this subfamily identified in Caulobacter crescentus. A sigF-null strain was obtained and shown to be severely impaired in resistance to oxidative stress, caused by hydrogen peroxide treatment, exclusively during the stationary phase. Although sigF mRNA levels decrease in stationary-phase cells, the amount of F protein is greatly increased at this stage, indicating a posttranscriptional control. Data obtained indicate that the FtsH protease is either directly or indirectly involved in the control of F levels, as cells lacking this enzyme present larger amounts of the sigma factor. Increased stability of F protein in stationary-phase cells of the parental strain and in exponential-phase cells of the ftsH-null strain is also demonstrated. Transcriptome analysis of the sigF-null strain led to the identification of eight genes regulated by F during the stationary phase, including sodA and msrA, which are known to be involved in oxidative stress response.
INTRODUCTION
Bacteria are widespread organisms capable of colonizing highly distinct environments and of surviving various types of stresses. This capacity is in part conferred by the presence of alternative sigma factors, which permit the rapid adaptation of bacteria to environmental changes by switching their genetic programs in response to internal or external signals. Alternative sigma factors of the extracytoplasmic function (ECF) subfamily were initially described as a divergent group of the 70 family involved in responses to changes in the extracytoplasmic compartment of the cell (23). Later, members of this class were also shown to regulate cytoplasmic functions (34). ECF sigma factors have been intensively characterized in different bacterial species and have been implicated in responses to a variety of stresses, such as heat shock, oxidative and osmotic stresses, and stationary-phase survival (14).
Characterization of ECF sigma factors from multiple bacteria revealed several common features. Alignment of –35 and –10 sequences of promoters recognized by different ECF sigma factors shows a high degree of conservation at the –35 region, which very often contains an "AAC" motif (14, 29). The –10 determinant is more variable, probably due to the reduced similarity in region 2.4 among distinct ECF sigma factors (23, 29).
ECF sigma factors are frequently cotranscribed with their cognate regulators, which in most cases are inner membrane anti-sigma factors, which bind to the sigma factors and inhibit their association with the RNA polymerase core enzyme. The operon may contain other genes coding for regulators of the anti-sigma factor activity, as described in Escherichia coli and Pseudomonas aeruginosa (7, 28, 43). The cascade that leads to induction of ECF sigma factor activity has been extensively studied for E of E. coli. In this model, the presence of extracytoplasmic induction signals triggers a proteolytic cascade in the periplasm that leads to degradation of the anti-sigma factor RseA, releasing E to bind the RNA polymerase (1, 17). This mode of regulation coordinates changes in the extracytoplasmic compartment of the cell with alterations in gene expression. However, a soluble anti-sigma factor has also been described in Streptomyces coelicolor (RsrA), and it regulates R, an ECF sigma factor involved in the response to disulfide stress in the cytoplasm (19, 34). In this case, the anti-sigma factor RsrA is inactivated by intramolecular disulfide bond formation during disulfide stress (19).
Caulobacter crescentus is a gram-negative, -class proteobacterium characterized by its dimorphic cell cycle (26, 40). As a free-living organism present in nutritionally poor aquatic environments, C. crescentus might have developed many regulatory mechanisms to cope with nutrient deprivation and environmental stresses and may be found in nature in a state resembling the stationary phase of growth. C. crescentus, like all -proteobacteria studied so far, lacks an rpoS homologue, responsible for stationary-phase and general stress responses in E. coli and other bacteria. However, a stationary-phase-induced resistance to heat shock, oxidative stress, and alkaline/acid stress has been described in C. crescentus (51). Analysis of the complete genome sequence of this bacterium has lead to the identification of 16 open reading frames encoding putative sigma factors, 13 of them belonging to the ECF subclass (31). The other three genes have already been characterized: rpoD (70), rpoN (54), and rpoH (32) (3, 24, 39, 52). The presence of a large number of ECF sigma factors in C. crescentus suggests that these genes may exert important roles in the regulation of environmental stress responses in this bacterium.
The present work describes the isolation and characterization of a gene encoding an ECF sigma factor of C. crescentus, called sigF, and proposes a role for this gene in the response to oxidative stress during the stationary phase in C. crescentus. Regulation of sigF gene expression during the exponential and stationary phases of growth is also described. Data presented here demonstrate that F is involved in the regulation of at least eight genes, two of them (msrA and sodA) clearly involved in resistance to oxidative stress in bacteria.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. The strains and plasmids used in this work are described in Table 1. C. crescentus strains were grown at 30°C in PYE complex medium (36). Plasmids were propagated in E. coli strain DH5 and mobilized into C. crescentus by bacterial conjugation using E. coli strain S17-1. E. coli strains were grown at 37°C in LB broth (41).
Cloning of sigF gene and construction of sigF mutant. The sigF gene from C. crescentus was isolated from a partial genomic library constructed by digestion of NA1000 DNA with NcoI and SacI restriction enzymes and ligation of restriction fragments 4 to 6 kb in size into the vector pUCBM21. After transformation of E. coli, the positive clone containing the sigF gene was isolated through colony blot screening of the resulting library. The probe used was obtained by PCR using primers E5i and E3i (Table 2), which was subcloned into the pUCBM21 cloning vector, checked by automated DNA sequencing, and excised by EcoRI/BamHI digestion for subsequent labeling with [-32P]dCTP. The plasmid containing the sigF gene and flanking regions (insert size of about 5.2 kb) was called pRB80. A 1.3-kb ApaI internal fragment was ligated into the vector pUCBM21 digested with the same enzyme, generating pRB82.
The sigF in-frame deletion mutant was obtained by PCR amplification of the 3' and 5' regions of the gene flanking the portion to be deleted and ligation of the resulting fragments into pUCBM21. The 5' region (736 bp) and 3' region (278 bp) were amplified using plasmid pRB82 as a template. Oligonucleotide sigF5' (Table 2) and M13 universal primer were used for amplification of the 5' region. The 3' region was amplified with M13 reverse primer and the oligonucleotide sigF3' (Table 2). The 5' and 3' fragments obtained were digested with EcoRI/XhoI and HindIII/XhoI, respectively, and ligated into pUCBM21 digested with EcoRI/HindIII, generating pCM20. This construct represents a deletion of 345 bp of the sigF coding region flanked by 39 bp and 123 bp of upstream and downstream coding fragments, respectively. The insert was excised by EcoRI/HindIII digestion and ligated into pNPTS138. The resulting construct, denominated pCM21, was transferred to C. crescentus by conjugation, and replacement of the wild-type sigF with the deleted copy was achieved by two homologous recombinations events, resulting in the sigF mutant SG16.
A conditional mutant was obtained by insertion of the sigF coding region under the control of the PxylX promoter in the xylX locus of strain SG16. The sigF coding region was obtained by PCR amplification using pRB82 as template and pXNE5' and pXNE3' as primers (Table 2). The resulting fragment was digested with EcoRI (restriction site added in the PCR primers) and cloned into pNPT228XNE, generating a translational fusion to the sixth codon of the xylX gene (16). This construct, denominated pCM22, was introduced into the xylX locus of strain SG16 by conjugation and homologous recombination, selecting kanamycin-resistant colonies. One colony was chosen for subsequent experiments. All constructs containing PCR-generated fragments were confirmed by automated DNA sequencing before subsequent cloning steps.
Stress sensitivity tests and long-term viability assays. Overnight C. crescentus cultures were diluted to an optical density at 600 nm (OD600) of 0.1 in PYE and grown at 30°C until exponential phase (OD600 of 0.6) or for 12 h and 24 h. For each condition tested, an aliquot of untreated control cells was taken before exposure to stress conditions. Sensitivity to oxidative stress in exponential, 12-h, and 24-h cultures was determined by addition of hydrogen peroxide (2.5 mM or 15 mM final concentration). For heat shock sensitivity tests, cultures were transferred to a water bath at 48°C and incubated with gentle shaking. Sensitivity to alkali and acid was determined by growing cells in PYE medium at pH 9.5 and 4.0, respectively, as described by Wortinger et al. (51). For all conditions tested, aliquots of cells were taken at indicated time points and serial dilutions were plated on PYE-agar plates. Viability of the parental strain NA1000 was determined as a control in all experiments.
Promoter activity assays. For construction of a transcriptional fusion of the sigF promoter region to the lacZ gene, the region upstream of sigF coding region was amplified by PCR using pRB80 as a template and primers PESigE2 and E5 (Table 2). The resulting 686-bp fragment was cloned into pUCBM21 and ligated into pRKlac290 after EcoRI/HindIII digestion, generating pRB70. C. crescentus strain NA1000 harboring pRB70 was grown overnight in PYE medium, diluted to an OD600 of 0.1, and samples were taken at different time points for determination of -galactosidase activity, as described previously (27). Experiments were performed in duplicates and repeated on three different occasions.
RNA extraction. Total RNA for microarray experiments, primer extension, and reverse transcription-PCR (RT-PCR) analysis was extracted by the Trizol method, as described by the manufacturer (Invitrogen). Aliquots of cells (five aliquots of 2 ml from each treatment) from cultures of different C. crescentus strains in exponential phase (OD600 of 0.6), or after 12 h (OD600 of 1.3) and 24 h (OD600 of 1.3) of growth at 30°C, were collected by centrifugation in a microcentrifuge for 1 min. The cell pellets were resuspended in 1 ml of Trizol Reagent (Invitrogen), and after the extraction procedure, the integrity of the RNA was checked by agarose gel electrophoresis.
Primer extension experiments. The transcriptional start site of the sigF gene was determined by primer extension assay. An 18-mer oligonucleotide (named PesigE2, Table 2) complementary to nucleotides 33 to 50 of the sigF gene (see Fig. 2) was 5' end labeled with [-32P]ATP and T4 polynucleotide kinase and hybridized to 50 μg of total RNA isolated from NA1000 and SG16 cells grown at 30°C or after a 15-min exposure to 48°C at the exponential phase of growth. Annealing was carried out at 55°C for 16 h in 25 μl of 100 mM piperazine-N, N'-bis (2-ethanesulphonic acid) buffer, pH 7.0, containing 1 M NaCl and 5 mM EDTA. The nucleic acids were ethanol precipitated, and reverse transcription was carried out for 1 h with 200 U of Superscript II reverse transcriptase (Invitrogen) at 42°C in the presence of 1 mM dithiothreitol, 10 U RNase inhibitor, 1 mM each deoxynucleoside triphosphate, and enzyme buffer. RNA was digested by incubation for 30 min at 37°C in the presence of 30 μg RNase A, followed by phenol-chloroform extraction. The extension products were detected by electrophoresis on denaturing sequencing gels followed by autoradiography. The size of the extension products was estimated by using a DNA sequencing ladder, obtained with M13mp18 DNA and universal primer –40.
Immunoblot experiments. Total protein extracts of NA1000 and mutant strains were prepared by centrifugation of 1-ml aliquots of cells, taken at different intervals during growth or after treatment with various stresses, and resuspension in Laemmli sample buffer followed by boiling for 5 min. Equal amounts of total protein were separated by denaturing polyacrylamide gel electrophoresis with sodium dodecyl sulfate, and the resolved proteins were transferred to nitrocellulose filters. Membranes were incubated overnight at 4°C with a 1:1,000 dilution of anti-F in 10 mM Tris-HCl, pH 8.0, containing 150 mM NaCl, 0.02% Tween 20, and 0.03% Triton X-100. The blots were developed using ECL enhanced chemiluminescence kit (GE Healthcare). Immunoblots were quantified by densitometry scanning of the X-ray films using ImageQuant (Molecular Dynamics).
The anti-F antibody was obtained by three immunizations of rabbits, with approximately 500 μg of recombinant F protein in each injection. Recombinant F protein was overproduced in E. coli as an N-terminally His6-tagged fusion protein. The sigF gene was amplified by PCR using primers XNE 5' and XNE 3' (Table 2), and the resulting fragment was ligated into pGEM T-Easy (Promega), generating construct pCM23. The insert was excised from pCM23 by EcoRI digestion and cloned into pPROEXHTa (Invitrogen). The resulting construct, denominated pCM24, was introduced by transformation into E. coli DH5, and expression was induced during the exponential phase by addition of 1 mM isopropyl--D-thiogalactopyranoside. After incubation for 3 h at 37°C, the cells were harvested and lysed by sonication in 10 mM Tris-HCl buffer (pH 7.5) containing 100 mM NaCl, 10 mM imidazole, and 10 mg/ml phenylmethanesulfonyl fluoride. The cell lysate was centrifuged (10,000 x g for 5 min), and the His6-F fusion protein was detected in the insoluble pellet, as inclusion bodies, as described for the ECF subfamily members W from Bacillus subtilis and CarQ from Myxococcus xanthus (5, 15). His6-F protein was purified from inclusion bodies using 0.3% Sarkosyl, as previously described (25). Immunoblot experiments using the anti-F antiserum showed that F protein migrates as a 25-kDa band, which is missing from total protein extracts obtained from sigF mutant strain, SG16. As described for other sigma factors, this apparent molecular mass is greater than the 19.8 kDa expected by the F primary sequence (15). E. coli rabbit anti-FtsH serum (kindly provided by T. Ogura, Kumamoto University, Japan) was used at a 1:2,000 dilution and Caulobacter crescentus rabbit anti-Rho serum (kindly provided by M. V. Marques) was used at a 1:1,000 dilution.
Semiquantitative RT-PCR. RNA samples used in semiquantitative RT-PCR experiments were treated with 1 U/μg of amplification-grade DNase I (Invitrogen), according to the manufacturer's instructions, and the absence of residual DNA contamination was confirmed by PCR. After DNase treatment, the RNA samples were quantified in a spectrophotometer and analyzed by agarose gel electrophoresis.
For quantitative amplification of each transcript, RT-PCRs were performed using the Access-Quick RT-PCR system (Promega) with a reduced number of cycles, avoiding reaching the plateau of the PCRs. The number of cycles to be used for each amplicon was determined by taking samples at a distinct number of cycles during the PCR amplification step and analyzing the products obtained by agarose gel electrophoresis. Reactions comparing expression of each gene in NA1000 and SG16 cells were performed with a fixed number of cycles and different amounts of total RNA (100 and 200 ng). The PCR products were resolved through agarose gel electrophoresis, and the DNA fragments were transferred to nylon membranes, as previously described (41). Membranes were hybridized to 32P-labeled fragments obtained by PCR amplification with the same primers used for RT-PCR experiments and exposed to Kodak X-Omat films. The hybridization bands were quantified by densitometry scanning of the films using ImageQuant (Molecular Dynamics).
Microarray preparation and data analysis. Design of oligonucleotides, hybridization, and image acquisition procedures are described online at www.stanford.edu/group/caulobacter/oligoArrays. Four distinct biological RNA samples from each strain were prepared and labeled with either Cy3 or Cy5 in a dye-swap design. The fluorescence intensity of Cy3 and Cy5 for each spot was determined by the equation: mean (foreground) – median (background). Spots containing flags or displaying a mean foreground value lower than the 90% quantile of the background were excluded. Normalization was performed by the LOWESS method as described previously (20). Normalized results of all replicates for each gene were used for the estimation of a normal distribution using a robust approach (median and median absolute deviation) and the probability of a change of 1.5-fold or more (log2 SG16/NA1000, >0.58, or log2 SG16/NA1000, <–0.58) was estimated for genes presenting at least three replicates. Genes showing a probability equal to or greater than 85% were classified as differentially expressed between the two strains. The log ratios for the selected genes are presented in Table S1 in the supplemental material.
RESULTS
Cloning and disruption of the sigF gene from C. crescentus. To better understand the control of gene expression in C. crescentus, we started the characterization of one of the putative ECF sigma factors identified in this bacterium (31). The putative 540-bp open reading frame CC3253 (TIGR nomenclature), which was denominated sigF, was isolated from a partial genomic library, as described in Materials and Methods. A global similarity search of the NCBI database using BLAST demonstrated a high degree of similarity of F to ECF sigma factors not yet characterized, especially from other members of the subdivision of proteobacteria. A global alignment of the predicted F amino acid sequence with the best hit obtained from BLAST showed a 47.2% identity to a Burkholderia fungorum putative ECF sigma factor named Bcep02004893 (NCBI Microbial Genomes Annotation Project). The closest match between an already characterized ECF sigma factor and the predicted F protein is Mycobacterium tuberculosis SigE, which is 28.65% identical to the C. crescentus F sequence (53). This gene was also found in other mycobacteria and was shown to be involved in survival of extreme heat shock, oxidative and acid stress, and detergent treatment (53). Similarity searches of the predicted F protein using the Conserved Domain Database from BLAST have demonstrated the presence of conserved regions 2 and 4 from 70 family members, encompassing amino acids 33 to 93 and 123 to 172, respectively (not shown).
A sigF-null mutant was obtained by constructing an in-frame internal deletion of 345 bp in the sigF coding region, deleting amino acids 25 through 139, corresponding to the entire region 2 and the beginning of region 4 (see Materials and Methods). This mutant strain, denominated SG16, did not show any difference in cell morphology, motility, or colony formation when compared to the parental strain NA1000, indicating that sigF is not essential for survival during normal growth. In addition, a previous analysis using a mutant with sigF interrupted by a spectinomycin resistance cassette has indicated that sigF is not necessary for survival of distinct treatments during the exponential growth phase, including osmotic treatment, acid and alkaline stresses, oxidative stress, and cold shock (data not shown). This insertion mutant was not used for further studies, as sigF probably forms an operon with a downstream gene encoding a conserved hypothetical protein (CC3252), and any phenotype observed in the mutant strain could be the result of changes in the expression of the latter gene. Thus, all the results further described in this work have been obtained with the in-frame deletion mutant SG16.
sigF mutant is sensitive to oxidative stress in the stationary phase of growth. Since no phenotype was observed in the exponential growth phase, we decided to investigate whether sigF could be important for the adaptation to the stationary phase, previously described in C. crescentus (51). The growth curves of NA1000 and SG16 were found to be identical in either complex medium (PYE; Fig. 1A) or minimal medium (M2; not shown), indicating that the absence of F does not affect the C. crescentus growth rate. Analysis of viability and morphology of SG16 cells during long periods of incubation in rich medium (15 days) also showed the same profile as that observed for NA1000 parental cells, including the acquisition of an elongated helical morphology after the ninth day (data not shown) (51).
Sensitivity of SG16 cells to distinct stress conditions during the stationary phase was then analyzed using 24-h cultures. SG16 stationary cells did not present increased sensitivity to heat shock (45°C) or acid stress (pH 4.0) compared to the parental strain NA1000 (data not shown). We also investigated the susceptibility of sigF mutant cells to exposure to high concentrations of hydrogen peroxide. As previously described (51), overnight cultures of NA1000 are quite resistant to 15 mM hydrogen peroxide, with approximately 45% of the cells remaining viable after 120 min of exposure (Fig. 1B). In contrast, SG16 mutant cells did not show this increased resistance to oxidative stress and viability decreased very rapidly after 60 min of exposure to 15 mM hydrogen peroxide, reaching about 0.2% of viable cells after 120 min (Fig. 1B). Sensitivity to hydrogen peroxide was also assessed in 12-h cultures, when cells are at the onset of the stationary phase (Fig. 1A). At this stage of growth, NA1000 cells have not acquired the resistance to this treatment observed in 24-h cultures, and after 15 min of exposure to 15 mM hydrogen peroxide, only about 5% of the parental cells were still viable (not shown). However, SG16 mutant cells were already more sensitive than parental cells to hydrogen peroxide at this stage, with only 0.1% of the mutant cells forming colonies after the same treatment (not shown). No increased sensitivity of mutant cells was observed with hydrogen peroxide treatment in exponential-phase cultures (not shown), which indicates that F becomes essential for survival of oxidative stress as C. crescentus cells enter the stationary growth phase.
We also constructed a conditional mutant, SG16C, carrying a single functional copy of sigF under the control of a xylose-inducible promoter in the xylX locus of the chromosome of SG16 strain. There is no residual expression of the sigF gene when SG16C cells are grown in the absence of xylose, as observed by RT-PCR analysis and immunoblot experiments using an anti-F polyclonal antiserum (data not shown). This conditional mutant strain was used in complementation tests, in order to confirm that the susceptibility of the SG16 strain to hydrogen peroxide during the stationary phase is solely due to the absence of sigF. Overnight cultures of SG16C grown in the absence of xylose were pretreated for 150 min with 0.1% xylose plus 0.1% glucose (induction of sigF expression) or 0.2% glucose (no expression of sigF) before exposure to 15 mM hydrogen peroxide. As depicted in Fig. 1C, addition of xylose before exposure to hydrogen peroxide resulted in increased resistance of mutant cells to oxidative stress, reaching levels close to those observed in parental cells. In contrast, cultures without xylose were still highly sensitive to hydrogen peroxide.
sigF transcription is induced by extreme heat shock during the exponential phase and is autoregulated. As ECF sigma factors have been implicated in the response to extreme heat shock (8), we investigated whether induction of sigF transcription occurs after exposure to extreme temperatures. To address this question, we performed RT-PCR analysis using RNA extracted from cells exposed to 48°C during exponential growth, and our results showed an increase of more than 2.5-fold in sigF mRNA levels after exposure to 48°C for 15 min (Fig. 2A). However, no difference in survival was observed between the parental strain and the sigF mutant strain after exposure to heat shock at 48°C during exponential growth (not shown). These results indicate that, although sigF may be involved in this stress response, its role is not essential. One possibility is that absence of F is compensated for by some of the other ECF sigma factors present in C. crescentus. Indeed, results from our laboratory have shown that other ECF sigma factors are induced by heat stress in this bacterium (data not shown).
Taking advantage of the high level of sigF expression after treatment of the cells at 48°C during exponential growth, we performed primer extension assays to determine the sigF transcription start site. A 52-nucleotide extension product was clearly observed with RNA isolated from NA1000 cells heat shocked at 48°C (Fig. 2B). However, no extension product was detected using RNA from untreated cultures, which is consistent with the RT-PCR data. The size of the extension product mapped the transcription start site to a guanine located 2 nucleotides upstream of the putative translation initiation site proposed by the TIGR annotation. Analysis of the sequence upstream of the putative transcription start site revealed the presence of a candidate –35 promoter region similar to –35 elements recognized by ECF sigma factors in other bacteria (Fig. 2B), which includes the highly conserved "AAC" motif (14). Similarity in the –10 region is difficult to assess because no consensus for this region has been defined for ECF sigma factor promoters.
The fact that most ECF sigma factors described to date are positively autoregulated led us to investigate if this was also true for sigF. To address this question, we performed primer extension experiments using RNA isolated from the sigF mutant, SG16. This was possible because the oligonucleotide used in the assays recognizes a region that was not affected in the sigF mutant. No induction of sigF expression was observed after exposure of sigF mutant cells to 48°C, as the 52-nucleotide band corresponding to the extension product was not detected using RNA isolated from SG16 cells (Fig. 2B). These results suggest that sigF positively autoregulates its own expression during heat shock.
To further confirm that F is capable of positively autoregulating its own expression, we analyzed the effect of increased F levels upon the activity of its own promoter. For this, we cloned the sigF gene in a medium-copy-number vector (pJS14) under the control of a constitutive promoter (Table 1). F levels are more than twofold higher in strain NA1000 harboring this construct (named pCM30), as observed in immunoblots (data not shown). This increased sigF expression did not result in differences in growth rate or viability during normal growth (data not shown). A transcriptional fusion of the sigF promoter to the lacZ gene was then transferred to the strain containing the pCM30 construct, and -galactosidase assays were performed. A 2.5-fold increase in sigF promoter activity was observed in the strain harboring the sigF extra copies, in comparison to NA1000 containing the pJS14 vector alone. This result confirms data obtained from primer extension analysis, discarding the possibility that the extension product was not detected in the mutant strain due to a lower stability of the truncated sigF transcript.
Analysis of sigF gene expression in response to the stationary phase and oxidative stress. Since sigF is proposed to have a role in the response to oxidative stress only during the stationary phase of growth, we analyzed transcription of sigF as a function of the growth phase by RT-PCR experiments. The results demonstrated a significant decrease in sigF mRNA levels upon entry into the stationary phase (12 h of growth at 30°C), which is even more pronounced after 24 h of growth (Fig. 3A). Then, F protein levels as a function of the growth rate were also examined by immunoblots. The amount of F protein in total protein extracts is low during the exponential growth phase (4 h) and increases gradually as a function of cell density, reaching maximal levels in stationary-phase cells (12 and 24 h of growth at 30°C) (Fig. 3B). As shown in Fig. 3B, F protein levels are nearly eightfold higher in stationary-phase cells, when compared to early-exponential-phase cells (4 h). As a control, membrane blots were also incubated with anti-Rho polyclonal antibody, since Rho protein levels do not change significantly as a function of Caulobacter cell density (M. V. Marques, personal communication) (Fig. 3B, lower panel). As sigF mRNA levels decrease during the stationary phase (Fig. 3A), we can conclude that sigF expression is posttranscriptionally controlled in C. crescentus.
sigF gene expression in response to hydrogen peroxide treatment during the exponential and stationary phases was also analyzed. We could not observe induction of sigF transcription in RT-PCR experiments after exposure of NA1000 cells to a wide range (0.06 mM to 15 mM) of H2O2 concentrations, in either the exponential or stationary phase (not shown). Accordingly, there was no increase in F protein levels, as determined in immunoblot experiments, when cells were exposed to different concentrations of hydrogen peroxide during the exponential or stationary phases of growth (not shown).
Effect of endogenous proteases on F protein levels. As our results pointed to a posttranscriptional regulation of sigF expression, we assessed F protein levels in some C. crescentus protease mutant strains characterized previously. Initially, we analyzed the levels of F protein in ClpP and ClpX conditional mutant strains, UJ270 and UJ271, respectively (16). The ClpP/ClpX complex is known to regulate S protein levels in a growth-phase-dependent manner in E. coli (44). The strains used contain a single functional copy of clpP or clpX under the control of a xylose-inducible promoter. As previously described (16), total depletion of ClpX or ClpP is observed after 6 h of growth in the absence of the xylose; therefore, F levels were assessed in cultures grown in the presence or absence of xylose for 8 h (late exponential growth phase) (Fig. 4A). F protein levels remained the same when either ClpP or ClpX was depleted (Fig. 4A), indicating that F levels are not regulated by these proteases.
We also tested the effect of a null mutation in the gene encoding the FtsH protease on F levels during different growth phases. FtsH is an inner membrane ATP-dependent protease which was shown to be required for cell differentiation, growth under salt and heat stresses, and for stationary-phase survival in C. crescentus (11). As depicted in Fig. 4B, F levels are 2.8-fold higher in exponential-phase cultures of the ftsH-null mutant strain UJ945, when compared to the levels observed in the parental strain NA1000, indicating that FtsH is either directly or indirectly involved in the control of F protein levels. Although F levels are still higher in UJ945 after 24 h of growth than those in NA1000, the difference is less pronounced at this time point (Fig. 4B), which suggests that FtsH activity has a more prominent effect in the reduction of F levels in exponential-phase cells. This is consistent with the gradual decrease in FtsH protein levels as cells enter the stationary phase, as determined in immunoblot assays. After 12 and 24 h of growth, a reduction of 20 and 50% in FtsH levels is observed, respectively, compared to the amount observed in exponential-phase cells (Fig. 4C). Thus, reduced FtsH-mediated proteolysis of F during the stationary phase may be an important factor leading to accumulation of F during this growth stage.
Analysis of F protein stability. To further confirm the results described above, we analyzed F stability in exponential- and stationary-phase cultures of NA1000 and the ftsH-null strain. For this purpose, protein synthesis was inhibited by addition of chloramphenicol and F levels were monitored at distinct time points after antibiotic treatment. Experiments were performed using 25 μg of chloramphenicol per ml, as this concentration was reported to immediately block protein synthesis in C. crescentus (33).
As observed in Fig. 5A, F protein levels gradually decrease in exponential-phase cultures of strain NA1000, reaching 30% of the amount observed at time zero, 1 h after the addition of chloramphenicol. In contrast, this protein is highly stabilized in stationary-phase cells (24-h cultures) and no significant change in its levels was observed during the same time interval (Fig. 5B). Similar results were obtained when exponential-phase cells were compared to 12-h cultures (data not shown). In addition, Rho protein levels were also assessed as a control in all experiments, and no difference in its stability was observed between exponential- and stationary-phase cells.
In agreement with the observation that F protein levels are higher in ftsH mutant cells, the sigma factor is also stabilized in exponential-phase cultures of the protease mutant strain, as depicted in Fig. 5C. In addition, no significant decrease in F protein levels was observed in 24-h cultures of strain UJ945 at the times assessed, in a profile similar to that observed for strain NA1000 at this growth stage (Fig. 5D).
These results demonstrate that stabilization of F protein occurs in stationary-phase cultures and corroborate the proposed role for FtsH in the negative regulation of F levels in exponential-phase cultures.
Transcriptome analysis of F-regulated genes. To identify F-regulated genes, we compared the global expression profiles of sigF deletion mutant SG16 with those of the parental strain NA1000 in microarray experiments. Since our data indicated a role for F during the stationary phase, differences in global gene expression were assessed during this phase, using cultures grown for 12 h and 24 h at 30°C, representing the early and late stationary phases, respectively. Any gene showing a reduction in expression of 1.5-fold or more in the SG16 strain in comparison to NA1000 was defined as being regulated by F, either directly or indirectly. A total of eight genes were identified through this analysis (Table 3).
Genes of interest were selected for validation of microarray data by semiquantitative RT-PCR. RNA samples from exponentially growing cultures were also used to assess gene expression by RT-PCR, in order to verify if the gene of interest shows higher expression levels upon entry into the stationary phase and if it is also F regulated during exponential growth.
Microarray data demonstrated that gene msrA, encoding the enzyme peptide methionine sulfoxide reductase (MsrA), displays a 1.9-fold reduction in expression in 24-h cultures of the mutant strain SG16 (Table 3). MsrA is an antioxidant enzyme, which reduces oxidized methionine sulfoxide [Met(O)] residues in proteins back to methionine. These residues occur as a consequence of aerobic metabolism, due to formation of reactive oxygen and nitrogen species (ROS and RNI, respectively) (50). RT-PCR analysis confirmed the reduced expression of msrA in the mutant strain during the stationary phase and showed that the gene is also dependent on F during exponential growth (Fig. 6A). Although the difference in msrA expression in 12-h cultures was not considered statistically significant in our microarray experiments (Table 3), results from RT-PCR analysis did show markedly reduced levels of msrA mRNA at this time point. Altogether, these results indicate that F regulates msrA expression during the exponential and stationary phases. In addition, RT-PCR experiments showed that transcription of msrA is not induced upon entry into stationary phase in C. crescentus (Fig. 6A).
Another gene analyzed by RT-PCR was sodA, which encodes the enzyme (Mn2+)-superoxide dismutase. This enzyme catalyzes the dismutation of O2·– to molecular oxygen (O2) and H2O2, which in turn, is detoxified by catalases and peroxidases. Consistent with microarray data, expression of sodA was shown in RT-PCR experiments to be highly reduced in 12-h cultures of strain SG16 compared to the parental strain. After 24 h of growth, sodA mRNA levels were shown to be quite similar in both mutant and parental cells, in agreement with microarray results (Fig. 6B). The RT-PCR analysis also showed that sodA expression is highly reduced in the parental strain NA1000 during the stationary phase (12 h and 24 h) in comparison to expression during exponential growth (4 h). As SodA activity has been mainly implicated in responses to superoxide radical, we investigated the sensitivity of SG16 cells to paraquat and xanthine oxidase, which are cytoplasmic and extracytoplasmic generators of superoxide, respectively. However, we could not detect increased sensitivity of mutant cells to these treatments using 12-h cultures (data not shown).
Semiquantitative RT-PCR assays were also performed to confirm differential expression of CC3255, which encodes a conserved hypothetical protein, between the two strains. This is the first gene of a probable operon located upstream of sigF and in the opposite orientation, which also contains CC3257 (whose microarray data also showed reduced expression in strain SG16). As shown in Fig. 6C, we could confirm the lower expression of CC3255 in 12-h and 24-h cultures of the SG16 strain. In addition, expression of CC3255 is also dependent on F during exponential growth, as its mRNA levels are significantly diminished in the sigF mutant in 4-h cultures. RT-PCR data also showed that CC3255 expression is highly induced upon entry into the stationary phase (12 h) in the wild-type background, decreasing again after 24 h of growth.
Gene CC3572, encoding a putative carbonic anhydrase, was also down-regulated in 12-h and 24-h cultures of strain SG16 in our microarray data. This enzyme catalyzes the interconversion of CO2 and HCO3– (bicarbonate anion), and evidence from studies of eukaryotic carbonic anhydrases indicate a role for these enzymes in protection against oxidative stress (13, 37, 57). However, we could not confirm reduced expression of this gene in strain SG16 in semiquantitative RT-PCR experiments (data not shown).
DISCUSSION
This work reports the first characterization of an ECF sigma factor from C. crescentus, named F, which is essential for the oxidative stress response during the stationary phase of growth. The response to these growth conditions may be specially important for C. crescentus physiology, as this bacterium is submitted to high oxygen concentrations in its natural habitat due to association with cyanobacteria (2, 22) and it is mostly found in diluted, nutritionally poor environments (36). Although some aspects of these environmental responses have been described previously (42, 46, 51), the only regulator identified to date is skgA, involved in katG induction in the stationary phase (38).
Despite its role in oxidative stress, sigF expression is not induced by H2O2 treatment. In contrast, F levels are induced upon entry into the stationary phase of growth, indicating that growth rate is the main factor that regulates F expression. A similar situation was observed for Salmonella enterica serovar Typhimurium rpoE expression (47). Stationary-phase cells are submitted to distinct stresses, and transition to this growth phase causes induction of many genes whose products are involved in protection against such adverse conditions. However, a growing body of evidence indicates that oxidative injury is the major challenge faced by stationary-phase cells and that most genes induced by stasis are involved in defense against oxidative stress (32). Thus, stationary-phase induction of sigF expression can be explained by its role in regulating responses to cope with oxidative stress. This assumption is substantiated by our transcriptome analysis, which identified at least two genes involved in resistance to oxidative stress as F dependent: msrA and sodA.
SodA activity has been shown to be involved in resistance to H2O2, as accumulation of O2·– leads to higher levels of free iron in the cell, which can react with H2O2 through the Fenton reaction, producing the highly reactive compound OH– (48). Our results show that the sodA gene is F dependent only in 12-h cultures, indicating that a regulator distinct from F is responsible for sodA levels during the exponential phase. In addition, sodA transcript levels are restored in the sigF mutant after 24 h, which may be due to the action of another sigma factor, compensating for the absence of F at this stage.
The msrA gene, which encodes another important antioxidant enzyme, has been shown to be necessary for survival of H2O2 exposure in a series of microorganisms, including E. coli (10, 30). Interestingly, C. crescentus msrA was observed to be F dependent at all stages of growth analyzed. No increased sensitivity to H2O2 was observed in strain SG16 during the exponential phase, which may indicate that correct expression of msrA (CC1039) may be more necessary for protection against oxidative stress during stationary phase. C. crescentus possesses an msrA paralog, named CC0994, which may be sufficient for protection against oxidative damage during the exponential phase.
Our transcriptome analysis led to the identification of a total of eight F-regulated genes. This result may suggest that F, like other ECF sigma factors described, regulates a small, defined set of genes under specific conditions. In addition, the absence of sigF does not lead to the complete absence of expression of the target genes, indicating that they are probably recognized by at least one additional sigma factor. These other sigma factors may have distinct or overlapping promoter specificity with F and could compensate for the absence of F at some target genes. The overlap between ECF regulons has been described for B. subtilis X and W (15) and was also suggested for S. coelicolor R (35). This may account for the minor changes in expression of target genes in a sigF-null background, complicating the identification of F-dependent genes using microarrays.
We performed a search for putative promoter sequences upstream of the F-regulated genes, using the program FINDPATTERNS (Genetics Computer Groups). This search was performed based on similarity to the –35 region of the sigF promoter determined in this study, since the sigF gene was shown to be positively autoregulated. This region is more conserved among ECF-dependent promoters than the –10 element (14). For some genes, we could identify putative –35 sequences and we also identified similarity at the –10 region (Table 4). However, for a more accurate identification of F promoters, determination of the transcription start sites for each gene is required.
Our results demonstrate that sigF possesses a complex pattern of regulation under distinct growth conditions. Analysis of sigF transcription in response to heat shock during the exponential phase and overexpression of sigF indicate that F positively autoregulates its own transcription under these conditions. This result was expected, since positive autoregulation of ECF gene expression has been demonstrated for most members of this subfamily. However, a distinct control of sigF transcription seems to occur during the stationary phase, since the increase observed in F protein levels does not lead to larger amounts of sigF mRNA. On the contrary, there is a significant reduction in sigF transcript levels during the stationary phase. These data suggest that the sigF promoter possesses a different selectivity during this stage, being less efficiently transcribed by EF or being regulated by another sigma factor, which may compete for the same promoter element and/or for the RNA polymerase core. Indeed, results from our laboratory indicate that at least two other members of the ECF subfamily are induced during the stationary phase in C. crescentus (data not shown). Distinct selectivity at the sigF promoter may be accounted for by the global changes that occur during transition between the exponential and stationary phases, including alterations in the DNA topology, the presence of nucleoid-associated proteins, and activators or repressors.
The sigF gene is probably cotranscribed with CC3252, coding for a conserved hypothetical protein, which may act as an anti-sigma factor. This is suggested by the presence of transmembrane domains in the protein encoded by CC3252 and by the fact that its orthologues in other bacteria are always found in operons containing an ECF sigma factor with a high degree of similarity to sigF as the first gene. Accordingly, analysis of CC3252 expression by RT-PCR also showed reduced transcription during the stationary phase, in a pattern similar to sigF (data not shown). In this sense, reduced transcription of the operon in the stationary phase may be an advantageous mechanism, preventing the consequent inactivation of F due to increased expression of its putative anti-sigma factor. This may be especially important during the stationary phase as a rapid shutoff of the response is not needed. We are presently investigating the involvement of the putative anti-sigma factor in the control of F activity by constructing a null strain in CC3252 as well as an overexpressing strain.
In addition, the posttranscriptional induction of sigF gene expression during the stationary phase may be a more advantageous mechanism of regulation in a growth stage when metabolic activities, including transcription and translation, are greatly diminished. Similarly, induction of E. coli rpoS gene expression during glucose starvation occurs most significantly by increased protein stability, in spite of decreased transcription and translation (55).
We have also demonstrated that FtsH protease is directly or indirectly involved in the regulation of F levels. In a similar manner, levels of B. subtilis W, a member of the ECF subfamily, are increased in an ftsH knockout strain (54). In C. crescentus, FtsH was previously shown to control the cellular levels of the heat shock sigma factor 32 in the presence of DnaK (6, 11). Our results also indicate that F is less susceptible to FstH action during the stationary phase, thus being more stable. By analogy to the 32 and S regulation models, F may associate with some accessory factor that promotes its degradation during the exponential phase (4, 56). In addition, we showed that FtsH is present at lower levels during the stationary phase, which may indicate lower activity of this protease at this stage. Furthermore, the increase in the amount of F observed in the ftsH mutant led to higher F activity, since msrA transcript levels determined in RT-PCR experiments increased from 1.5- to 2-fold in the protease mutant strain during the exponential phase, compared to the amount present in the parental strain (data not shown). Thus, degradation of F by FtsH should be an important regulatory component of its activity. Nevertheless, other proteases may be involved in the regulation of F levels, as F still accumulates in the ftsH mutant during the stationary phase, although to a lesser extent (see Fig. 4B). In a similar manner, a set of ATP-dependent proteases, including HslVU, act synergistically with FtsH in the negative regulation of 32 levels in E. coli (18).
Since the sigF-null mutant does not display a deficiency in stationary-phase survival, as observed for the stationary-phase sigma factor rpoS in E. coli (21), other sigma factors may also be involved in stationary-phase responses in C. crescentus. Thus, characterization of other ECF subfamily members and their regulons and regulators is of great importance for the understanding of C. crescentus physiology and responses to stress.
ACKNOWLEDGMENTS
We are indebted to Lucy Shapiro and Harley McAdams (Stanford University) for making possible the realization of the microarray experiments and to Lucy Shapiro for critical reading of the manuscript. We also thank Maliwan Meewan for performing the microarray assays and Alison Kay Hottes and Ann Reisenauer for helpful advice during the design and analysis of these experiments. We thank Urs Jenal for generously providing ftsH, clpX, and clpP mutant strains; Teru Ogura for anti-FtsH antiserum; and Marílis V. Marques for anti-Rho antiserum. We also thank Tie Koide and Ricardo Z. N. Vêncio for help with microarray data analysis.
REFERENCES
Alba, B. M., J. A. Leeds, C. Onufryk, C. Z. Lu, and C. A. Gross. 2002. DegS and YaeL participate sequentially in the cleavage of RseA to activate the sigma(E)-dependent extracytoplasmic stress response. Genes Dev. 16:2156-2168.
Allen, H. L. 1971. Primary productivity, chemoorganotrophy, and nutritional interactions of epiphytic algae and bacteria on macrophytes in the littoral of a lake. Ecol. Monogr. 41:97-127.
Anderson, D. K., N. Ohta, J. Wu, and A. Newton. 1995. Regulation of the Caulobacter crescentus rpoN gene and function of the purified sigma 54 in flagellar gene transcription. Mol. Gen. Genet. 246:697-706.
Blaszczak, A., C. Georgopoulos, and K. Liberek. 1999. On the mechanism of FtsH-dependent degradation of the sigma 32 transcriptional regulator of Escherichia coli and the role of the Dnak chaperone machine. Mol. Microbiol. 31:157-166.
Browning, D. F., D. E. Whitworth, and D. A. Hodgson. 2003. Light-induced carotenogenesis in Myxococcus xanthus: functional characterization of the ECF sigma factor CarQ and antisigma factor CarR. Mol. Microbiol. 48:237-251.
da Silva, A. C. A., R. C. G. Simo, M. F. Susin, R. L. Baldini, M. Avedissian, and S. L. Gomes. 2003. Downregulation of the heat shock response is independent of DnaK and 32 levels in Caulobacter crescentus. Mol. Microbiol. 49:541-553.
De las Penas, A., L. Connolly, and C. A. Gross. 1997. The sigmaE-mediated response to extracytoplasmic stress in Escherichia coli is induced by RseA and RseB, two negative regulators of sigma E. Mol. Microbiol. 24:373-385.
Erickson, J. W., and C. A. Gross. 1989. Identification of the sigma E subunit of Escherichia coli RNA polymerase: a second alternate sigma factor involved in high-temperature gene expression. Genes Dev. 3:1462-1471.
Evinger, M., and N. Agabian. 1977. Envelope-associated nucleoid from Caulobacter crescentus stalked and swarmer cells. J. Bacteriol. 132:294-301.
Ezraty, B., L. Aussel, and F. Barras. 2005. Methionine sulfoxide reductases in prokaryotes. Biochim. Biophys. Acta 1703:221-229.
Fischer, B., G. Rummel, P. Aldridge, and U. Jenal. 2002. The FtsH protease is involved in development, stress response and heat shock control in Caulobacter crescentus. Mol. Microbiol. 44:461-478.
Gober, J. W., and L. Shapiro. 1992. A developmentally regulated Caulobacter flagellar promoter is activated by 3' enhancer and IHF binding elements. Mol. Biol. Cell 3:913-916.
Gotz, R., A. Gnann, and F. K. Zimmerman. 1999. Deletion of the carbonic anhydrase-like gene NCE103 of the yeast Saccharomyces cerevisiae causes an oxygen-sensitive growth defect. Yeast 15:855-864.
Helmann, J. 2002. The extracytoplasmic function (ECF) sigma factors. Adv. Microb. Physiol. 46:47-110.
Huang, X., K. L. Fredrick, and J. D. Helmann. 1998. Promoter recognition by Bacillus subtilis W: autoregulation and partial overlap with the X regulon. J. Bacteriol. 180:3765-3770.
Jenal, U., and T. Fuchs. 1998. An essential protease involved in bacterial cell-cycle control. EMBO J. 17:5658-5669.
Kanehara, K., K. Ito, and Y. Akiyama. 2002. YaeL (EcfE) activates the sigma(E) pathway of stress response through a site-2 cleavage of anti-sigma(E), RseA. Genes Dev. 16:2147-2155.
Kanemori, M., K. Nishihara, H. Yanagi, and T. Yura. 1997. Synergistic roles of HslVU and other ATP-dependent proteases in controlling in vivo turnover of 32 and abnormal proteins in Escherichia coli. J. Bacteriol. 179:7219-7225.
Kang, J. G., M. S. Paget, Y. J. Seok, M. Y. Hahn, J. B. Bae, J. S. Hahn, C. Kleanthous, M. J. Buttner, and J. H. Roe. 1999. RsrA, an anti-sigma factor regulated by redox change. EMBO J. 18:4292-4298.
Koide, T., P. A. Zaini, L. M. Moreira, R. Z. N. Vecio, A. Y. Matsukuma et al. 2004. DNA microarray-based genome comparison of a pathogenic and a nonpathogenic strain of Xylella fastidiosa delineates genes important for bacterial virulence. J. Bacteriol. 186:5442-5449.
Lange, R., and R. Hengge-Aronis. 1991. Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol. Microbiol. 5:49-59.
Lapteva, N. A. 1987. Ecological features of distribution of bacteria of the genus Caulobacter in freshwater bodies. Mikrobiologiya 56:677-684. (In Russian; English translation, 56:537-543.)
Lonetto, M., K. Brown, K. Rudd, and M. Buttner. 1994. Analysis of the Streptomyces coelicolor sigE gene reveals the existence of a subfamily of eubacterial RNA polymerase factors involved in the regulation of extracytoplasmic functions. Proc. Natl. Acad. Sci. USA 91:7573-7577.
Malakooti, J., and B. Ely. 1995. Principal sigma subunit of the Caulobacter crescentus RNA polymerase. J. Bacteriol. 177:6854-6860.
Marshak, D. R., J. T. Kadonaga, R. Burgess, M. W. Knuth, J. R. Brennan, and S. Lin. 1996. Strategies for protein purification and characterization. A laboratory course manual, p. 214-216. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
McGrath, P. T., P. Viollier, and H. H. McAdams. 2004. Setting the pace: mechanisms tying Caulobacter cell-cycle progression to macroscopic cellular events. Curr. Opin. Microbiol. 7:192-197.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Missiakas, D., M. P. Mayer, M. Lemaire, C. Georgopoulos, and S. Raina. 1997. Modulation of the Escherichia coli E (RpoE) heat-shock transcription-factor activity by the RseA, RseB and RseC proteins. Mol. Microbiol. 24:355-371.
Missiakas, D., and S. Raina. 1998. The extracytoplasmic function sigma factors: role and regulation. Mol. Microbiol. 28:1059-1066.
Moskovitz, J., M. A. Rahman, J. Strassman, S. O. Yancey, S. R. Kushner, N. Brot, and H. Weissbach. 1995. Escherichia coli peptide methionine sulfoxide reductase gene: regulation of expression and role in protecting against oxidative damage. J. Bacteriol. 177:502-507.
Nierman, W. C., T. V. Feldblyum, M. T. Laub, I. T. Paulsen, K. E. Nelson, J. A. Eisen, J. F. Heidelberg, M. R. Alley, N. Ohta, J. R. Maddock, I. Potocka, W. C. Nelson, A. Newton, C. Stephens, N. D. Phadke, B. Ely, R. T. DeBoy, R. J. Dodson, A. S. Durkin, M. L. Gwinn, D. H. Haft, J. F. Kolonay, J. Smit, M. B. Craven, H. Khouri, J. Shetty, K. Berry, T. Utterback, K. Tran, A. Wolf, J. Vamathevan, M. Ermolaeva, O. White, S. L. Salzberg, J. C. Venter, L. Shapiro, and C. M. Fraser. 2001. Complete genome sequence of Caulobacter crescentus. Proc. Natl. Acad. Sci. USA 98:4136-4141.
Nystrom, T. 2004. Stationary-phase physiology. Annu. Rev. Microbiol. 58:161-181.
Osley, M. A., and A. Newton. 1978. Regulation of cell cycle events in asymmetrically dividing cells: functions required for DNA initiation and chain elongation in Caulobacter crescentus. J. Bacteriol. 135:10-17.
Paget, M. S., J. G. Kang, J. H. Roe, and M. J. Buttner. 1998. sigma R, a RNA polymerase sigma factor that modulates expression of the thioredoxin system in response to oxidative stress in Streptomyces coelicolor A3(2). EMBO J. 17:5776-5782.
Paget, M. S., V. Molle, G. Cohen, Y. Aharonowitz, and M. J. Buttner. 2001. Defining the disulphide stress response in Streptomyces coelicolor A3(2): identification of the sigmaR regulon. Mol. Microbiol. 42:1007-1020.
Poindexter, J. S. 1964. Biological properties and classification of the Caulobacter group. Bacteriol. Rev. 28:231-295.
Raisanen, S. R., P. Lehenkari, M. Tasanen, P. Rahkila, P. L. Harkonen, and H. K. Vaananen. 1999. Carbonic anhydrase III protects cells from hydrogen peroxide-induced apoptosis. FASEB J. 13:513-522.
Rava, P. S., L. Somma, and H. M. Steinman. 1999. Identification of a regulator that controls stationary-phase expression of catalase-peroxidase in Caulobacter crescentus. J. Bacteriol. 181:6152-6159.
Reisenauer, A., C. D. Mohr, and L. Shapiro. 1996. Regulation of a heat shock 32 homolog in Caulobacter crescentus. J. Bacteriol. 178:1919-1927.
Ryan, K. R., and L. Shapiro. 2003. Temporal and spatial regulation in prokaryotic cell cycle progression and development. Annu. Rev. Biochem. 72:367-394.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Schnell, S., and H. M. Steinman. 1995. Function and stationary-phase induction of periplasmic copper-zinc superoxide dismutase and catalase/peroxidase in Caulobacter crescentus. J. Bacteriol. 177:5924-5929.
Schurr, M. J., H. Yu, J. M. Martinez-Salazar, J. C. Boucher, and V. Deretic. 1996. Control of AlgU, a member of the E-like family of stress sigma factors, by the negative regulators MucA and MucB and Pseudomonas aeruginosa conversion to mucoidy in cystic fibrosis. J. Bacteriol. 178:4997-5004.
Schweder, T., K.-H. Lee, O. Lomovskaya, and A. Matin. 1996. Regulation of Escherichia coli starvation sigma factor (s) by ClpXP protease. J. Bacteriol. 178:470-476.
Simon, R., U. Prieffer, and A. Puhler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. Biotechnology 1:784-790.
Steinman, H. M., F. Fareed, and L. Weinstein. 1997. Catalase-peroxidase of Caulobacter crescentus: function and role in stationary-phase survival. J. Bacteriol. 179:6831-6836.
Testerman, T. L., A. Vazquez-Torres, Y. Xu, J. Jones-Carson, S. J. Libby, and F. C. Fang. 2002. The alternative sigma factor E controls antioxidant defences required for Salmonella virulence and stationary-phase survival. Mol. Microbiol. 43:771-782.
Touati, D. 2000. Iron and oxidative stress in bacteria. Arch. Biochem. Biophys. 373:1-6.
Tsai, J. W., and M. R. K. Alley. 2000. Proteolysis of the McpA chemoreceptor does not require the Caulobacter major chemotaxis operon. J. Bacteriol. 182:504-507.
Weissbach, H., F. Etienne, T. Hoshi, S. H. Heinemann, W. T. Lowther, B. Matthews, G. St. John, C. Nathan, and N. Brot. 2002. Peptide methionine sulfoxide reductase: structure, mechanism of action, and biological function. Arch. Biochem. Biophys. 397:172-178.
Wortinger, M. A., E. M. Quardokus, and Y. V. Brun. 1998. Morphological adaptation and inhibition of cell division during stationary phase in Caulobacter crescentus. Mol. Microbiol. 29:963-973.
Wu, J., and A. Newton. 1996. Isolation, identification and transcriptional specificity of the heat shock sigma factor 32 from Caulobacter crescentus. J. Bacteriol. 178:2094-2101.
Wu, Q.-L., D. Kong, K. Lam, and R. N. Husson. 1997. A mycobacterial extracytoplasmic function sigma factor involved in survival following stress. J. Bacteriol. 179:2922-2929.
Zellmeier, S., U. Zuber, W. Schumann, and T. Wiegert. 2003. The absence of FtsH metalloprotease activity causes overexpression of the W-controlled pbpE gene, resulting in filamentous growth of Bacillus subtilis. J. Bacteriol. 185:973-982.
Zgurskaya, H. I., M. Keyhan, and M. Matin. 1997. The S levels in starving Escherichia coli cells increases solely as a result of its increased stability, despite decreased synthesis. Mol. Microbiol. 24:643-651.
Zhou, Y., S. Gottesman, J. R. Hoskins, M. R. Maurizi, and S. Wickner. 2001. The RssB response regulator directly targets sigma(S) for degradation by ClpXP. Genes Dev. 15:627-637.
Zimmerman, U. J., P. Wang, X. Zhang, S. Bogdanovich, and R. Forster. 2004. Anti-oxidative response of carbonic anhydrase III in skeletal muscle. IUBMB Life 56:343-347.(Cristina E. Alvarez-Marti)
ABSTRACT
Alternative sigma factors of the extracytoplasmic function (ECF) subfamily are important regulators of stress responses in bacteria and have been implicated in the control of homeostasis of the extracytoplasmic compartment of the cell. This work describes the characterization of sigF, encoding 1 of the 13 members of this subfamily identified in Caulobacter crescentus. A sigF-null strain was obtained and shown to be severely impaired in resistance to oxidative stress, caused by hydrogen peroxide treatment, exclusively during the stationary phase. Although sigF mRNA levels decrease in stationary-phase cells, the amount of F protein is greatly increased at this stage, indicating a posttranscriptional control. Data obtained indicate that the FtsH protease is either directly or indirectly involved in the control of F levels, as cells lacking this enzyme present larger amounts of the sigma factor. Increased stability of F protein in stationary-phase cells of the parental strain and in exponential-phase cells of the ftsH-null strain is also demonstrated. Transcriptome analysis of the sigF-null strain led to the identification of eight genes regulated by F during the stationary phase, including sodA and msrA, which are known to be involved in oxidative stress response.
INTRODUCTION
Bacteria are widespread organisms capable of colonizing highly distinct environments and of surviving various types of stresses. This capacity is in part conferred by the presence of alternative sigma factors, which permit the rapid adaptation of bacteria to environmental changes by switching their genetic programs in response to internal or external signals. Alternative sigma factors of the extracytoplasmic function (ECF) subfamily were initially described as a divergent group of the 70 family involved in responses to changes in the extracytoplasmic compartment of the cell (23). Later, members of this class were also shown to regulate cytoplasmic functions (34). ECF sigma factors have been intensively characterized in different bacterial species and have been implicated in responses to a variety of stresses, such as heat shock, oxidative and osmotic stresses, and stationary-phase survival (14).
Characterization of ECF sigma factors from multiple bacteria revealed several common features. Alignment of –35 and –10 sequences of promoters recognized by different ECF sigma factors shows a high degree of conservation at the –35 region, which very often contains an "AAC" motif (14, 29). The –10 determinant is more variable, probably due to the reduced similarity in region 2.4 among distinct ECF sigma factors (23, 29).
ECF sigma factors are frequently cotranscribed with their cognate regulators, which in most cases are inner membrane anti-sigma factors, which bind to the sigma factors and inhibit their association with the RNA polymerase core enzyme. The operon may contain other genes coding for regulators of the anti-sigma factor activity, as described in Escherichia coli and Pseudomonas aeruginosa (7, 28, 43). The cascade that leads to induction of ECF sigma factor activity has been extensively studied for E of E. coli. In this model, the presence of extracytoplasmic induction signals triggers a proteolytic cascade in the periplasm that leads to degradation of the anti-sigma factor RseA, releasing E to bind the RNA polymerase (1, 17). This mode of regulation coordinates changes in the extracytoplasmic compartment of the cell with alterations in gene expression. However, a soluble anti-sigma factor has also been described in Streptomyces coelicolor (RsrA), and it regulates R, an ECF sigma factor involved in the response to disulfide stress in the cytoplasm (19, 34). In this case, the anti-sigma factor RsrA is inactivated by intramolecular disulfide bond formation during disulfide stress (19).
Caulobacter crescentus is a gram-negative, -class proteobacterium characterized by its dimorphic cell cycle (26, 40). As a free-living organism present in nutritionally poor aquatic environments, C. crescentus might have developed many regulatory mechanisms to cope with nutrient deprivation and environmental stresses and may be found in nature in a state resembling the stationary phase of growth. C. crescentus, like all -proteobacteria studied so far, lacks an rpoS homologue, responsible for stationary-phase and general stress responses in E. coli and other bacteria. However, a stationary-phase-induced resistance to heat shock, oxidative stress, and alkaline/acid stress has been described in C. crescentus (51). Analysis of the complete genome sequence of this bacterium has lead to the identification of 16 open reading frames encoding putative sigma factors, 13 of them belonging to the ECF subclass (31). The other three genes have already been characterized: rpoD (70), rpoN (54), and rpoH (32) (3, 24, 39, 52). The presence of a large number of ECF sigma factors in C. crescentus suggests that these genes may exert important roles in the regulation of environmental stress responses in this bacterium.
The present work describes the isolation and characterization of a gene encoding an ECF sigma factor of C. crescentus, called sigF, and proposes a role for this gene in the response to oxidative stress during the stationary phase in C. crescentus. Regulation of sigF gene expression during the exponential and stationary phases of growth is also described. Data presented here demonstrate that F is involved in the regulation of at least eight genes, two of them (msrA and sodA) clearly involved in resistance to oxidative stress in bacteria.
MATERIALS AND METHODS
Bacterial strains, plasmids, and growth conditions. The strains and plasmids used in this work are described in Table 1. C. crescentus strains were grown at 30°C in PYE complex medium (36). Plasmids were propagated in E. coli strain DH5 and mobilized into C. crescentus by bacterial conjugation using E. coli strain S17-1. E. coli strains were grown at 37°C in LB broth (41).
Cloning of sigF gene and construction of sigF mutant. The sigF gene from C. crescentus was isolated from a partial genomic library constructed by digestion of NA1000 DNA with NcoI and SacI restriction enzymes and ligation of restriction fragments 4 to 6 kb in size into the vector pUCBM21. After transformation of E. coli, the positive clone containing the sigF gene was isolated through colony blot screening of the resulting library. The probe used was obtained by PCR using primers E5i and E3i (Table 2), which was subcloned into the pUCBM21 cloning vector, checked by automated DNA sequencing, and excised by EcoRI/BamHI digestion for subsequent labeling with [-32P]dCTP. The plasmid containing the sigF gene and flanking regions (insert size of about 5.2 kb) was called pRB80. A 1.3-kb ApaI internal fragment was ligated into the vector pUCBM21 digested with the same enzyme, generating pRB82.
The sigF in-frame deletion mutant was obtained by PCR amplification of the 3' and 5' regions of the gene flanking the portion to be deleted and ligation of the resulting fragments into pUCBM21. The 5' region (736 bp) and 3' region (278 bp) were amplified using plasmid pRB82 as a template. Oligonucleotide sigF5' (Table 2) and M13 universal primer were used for amplification of the 5' region. The 3' region was amplified with M13 reverse primer and the oligonucleotide sigF3' (Table 2). The 5' and 3' fragments obtained were digested with EcoRI/XhoI and HindIII/XhoI, respectively, and ligated into pUCBM21 digested with EcoRI/HindIII, generating pCM20. This construct represents a deletion of 345 bp of the sigF coding region flanked by 39 bp and 123 bp of upstream and downstream coding fragments, respectively. The insert was excised by EcoRI/HindIII digestion and ligated into pNPTS138. The resulting construct, denominated pCM21, was transferred to C. crescentus by conjugation, and replacement of the wild-type sigF with the deleted copy was achieved by two homologous recombinations events, resulting in the sigF mutant SG16.
A conditional mutant was obtained by insertion of the sigF coding region under the control of the PxylX promoter in the xylX locus of strain SG16. The sigF coding region was obtained by PCR amplification using pRB82 as template and pXNE5' and pXNE3' as primers (Table 2). The resulting fragment was digested with EcoRI (restriction site added in the PCR primers) and cloned into pNPT228XNE, generating a translational fusion to the sixth codon of the xylX gene (16). This construct, denominated pCM22, was introduced into the xylX locus of strain SG16 by conjugation and homologous recombination, selecting kanamycin-resistant colonies. One colony was chosen for subsequent experiments. All constructs containing PCR-generated fragments were confirmed by automated DNA sequencing before subsequent cloning steps.
Stress sensitivity tests and long-term viability assays. Overnight C. crescentus cultures were diluted to an optical density at 600 nm (OD600) of 0.1 in PYE and grown at 30°C until exponential phase (OD600 of 0.6) or for 12 h and 24 h. For each condition tested, an aliquot of untreated control cells was taken before exposure to stress conditions. Sensitivity to oxidative stress in exponential, 12-h, and 24-h cultures was determined by addition of hydrogen peroxide (2.5 mM or 15 mM final concentration). For heat shock sensitivity tests, cultures were transferred to a water bath at 48°C and incubated with gentle shaking. Sensitivity to alkali and acid was determined by growing cells in PYE medium at pH 9.5 and 4.0, respectively, as described by Wortinger et al. (51). For all conditions tested, aliquots of cells were taken at indicated time points and serial dilutions were plated on PYE-agar plates. Viability of the parental strain NA1000 was determined as a control in all experiments.
Promoter activity assays. For construction of a transcriptional fusion of the sigF promoter region to the lacZ gene, the region upstream of sigF coding region was amplified by PCR using pRB80 as a template and primers PESigE2 and E5 (Table 2). The resulting 686-bp fragment was cloned into pUCBM21 and ligated into pRKlac290 after EcoRI/HindIII digestion, generating pRB70. C. crescentus strain NA1000 harboring pRB70 was grown overnight in PYE medium, diluted to an OD600 of 0.1, and samples were taken at different time points for determination of -galactosidase activity, as described previously (27). Experiments were performed in duplicates and repeated on three different occasions.
RNA extraction. Total RNA for microarray experiments, primer extension, and reverse transcription-PCR (RT-PCR) analysis was extracted by the Trizol method, as described by the manufacturer (Invitrogen). Aliquots of cells (five aliquots of 2 ml from each treatment) from cultures of different C. crescentus strains in exponential phase (OD600 of 0.6), or after 12 h (OD600 of 1.3) and 24 h (OD600 of 1.3) of growth at 30°C, were collected by centrifugation in a microcentrifuge for 1 min. The cell pellets were resuspended in 1 ml of Trizol Reagent (Invitrogen), and after the extraction procedure, the integrity of the RNA was checked by agarose gel electrophoresis.
Primer extension experiments. The transcriptional start site of the sigF gene was determined by primer extension assay. An 18-mer oligonucleotide (named PesigE2, Table 2) complementary to nucleotides 33 to 50 of the sigF gene (see Fig. 2) was 5' end labeled with [-32P]ATP and T4 polynucleotide kinase and hybridized to 50 μg of total RNA isolated from NA1000 and SG16 cells grown at 30°C or after a 15-min exposure to 48°C at the exponential phase of growth. Annealing was carried out at 55°C for 16 h in 25 μl of 100 mM piperazine-N, N'-bis (2-ethanesulphonic acid) buffer, pH 7.0, containing 1 M NaCl and 5 mM EDTA. The nucleic acids were ethanol precipitated, and reverse transcription was carried out for 1 h with 200 U of Superscript II reverse transcriptase (Invitrogen) at 42°C in the presence of 1 mM dithiothreitol, 10 U RNase inhibitor, 1 mM each deoxynucleoside triphosphate, and enzyme buffer. RNA was digested by incubation for 30 min at 37°C in the presence of 30 μg RNase A, followed by phenol-chloroform extraction. The extension products were detected by electrophoresis on denaturing sequencing gels followed by autoradiography. The size of the extension products was estimated by using a DNA sequencing ladder, obtained with M13mp18 DNA and universal primer –40.
Immunoblot experiments. Total protein extracts of NA1000 and mutant strains were prepared by centrifugation of 1-ml aliquots of cells, taken at different intervals during growth or after treatment with various stresses, and resuspension in Laemmli sample buffer followed by boiling for 5 min. Equal amounts of total protein were separated by denaturing polyacrylamide gel electrophoresis with sodium dodecyl sulfate, and the resolved proteins were transferred to nitrocellulose filters. Membranes were incubated overnight at 4°C with a 1:1,000 dilution of anti-F in 10 mM Tris-HCl, pH 8.0, containing 150 mM NaCl, 0.02% Tween 20, and 0.03% Triton X-100. The blots were developed using ECL enhanced chemiluminescence kit (GE Healthcare). Immunoblots were quantified by densitometry scanning of the X-ray films using ImageQuant (Molecular Dynamics).
The anti-F antibody was obtained by three immunizations of rabbits, with approximately 500 μg of recombinant F protein in each injection. Recombinant F protein was overproduced in E. coli as an N-terminally His6-tagged fusion protein. The sigF gene was amplified by PCR using primers XNE 5' and XNE 3' (Table 2), and the resulting fragment was ligated into pGEM T-Easy (Promega), generating construct pCM23. The insert was excised from pCM23 by EcoRI digestion and cloned into pPROEXHTa (Invitrogen). The resulting construct, denominated pCM24, was introduced by transformation into E. coli DH5, and expression was induced during the exponential phase by addition of 1 mM isopropyl--D-thiogalactopyranoside. After incubation for 3 h at 37°C, the cells were harvested and lysed by sonication in 10 mM Tris-HCl buffer (pH 7.5) containing 100 mM NaCl, 10 mM imidazole, and 10 mg/ml phenylmethanesulfonyl fluoride. The cell lysate was centrifuged (10,000 x g for 5 min), and the His6-F fusion protein was detected in the insoluble pellet, as inclusion bodies, as described for the ECF subfamily members W from Bacillus subtilis and CarQ from Myxococcus xanthus (5, 15). His6-F protein was purified from inclusion bodies using 0.3% Sarkosyl, as previously described (25). Immunoblot experiments using the anti-F antiserum showed that F protein migrates as a 25-kDa band, which is missing from total protein extracts obtained from sigF mutant strain, SG16. As described for other sigma factors, this apparent molecular mass is greater than the 19.8 kDa expected by the F primary sequence (15). E. coli rabbit anti-FtsH serum (kindly provided by T. Ogura, Kumamoto University, Japan) was used at a 1:2,000 dilution and Caulobacter crescentus rabbit anti-Rho serum (kindly provided by M. V. Marques) was used at a 1:1,000 dilution.
Semiquantitative RT-PCR. RNA samples used in semiquantitative RT-PCR experiments were treated with 1 U/μg of amplification-grade DNase I (Invitrogen), according to the manufacturer's instructions, and the absence of residual DNA contamination was confirmed by PCR. After DNase treatment, the RNA samples were quantified in a spectrophotometer and analyzed by agarose gel electrophoresis.
For quantitative amplification of each transcript, RT-PCRs were performed using the Access-Quick RT-PCR system (Promega) with a reduced number of cycles, avoiding reaching the plateau of the PCRs. The number of cycles to be used for each amplicon was determined by taking samples at a distinct number of cycles during the PCR amplification step and analyzing the products obtained by agarose gel electrophoresis. Reactions comparing expression of each gene in NA1000 and SG16 cells were performed with a fixed number of cycles and different amounts of total RNA (100 and 200 ng). The PCR products were resolved through agarose gel electrophoresis, and the DNA fragments were transferred to nylon membranes, as previously described (41). Membranes were hybridized to 32P-labeled fragments obtained by PCR amplification with the same primers used for RT-PCR experiments and exposed to Kodak X-Omat films. The hybridization bands were quantified by densitometry scanning of the films using ImageQuant (Molecular Dynamics).
Microarray preparation and data analysis. Design of oligonucleotides, hybridization, and image acquisition procedures are described online at www.stanford.edu/group/caulobacter/oligoArrays. Four distinct biological RNA samples from each strain were prepared and labeled with either Cy3 or Cy5 in a dye-swap design. The fluorescence intensity of Cy3 and Cy5 for each spot was determined by the equation: mean (foreground) – median (background). Spots containing flags or displaying a mean foreground value lower than the 90% quantile of the background were excluded. Normalization was performed by the LOWESS method as described previously (20). Normalized results of all replicates for each gene were used for the estimation of a normal distribution using a robust approach (median and median absolute deviation) and the probability of a change of 1.5-fold or more (log2 SG16/NA1000, >0.58, or log2 SG16/NA1000, <–0.58) was estimated for genes presenting at least three replicates. Genes showing a probability equal to or greater than 85% were classified as differentially expressed between the two strains. The log ratios for the selected genes are presented in Table S1 in the supplemental material.
RESULTS
Cloning and disruption of the sigF gene from C. crescentus. To better understand the control of gene expression in C. crescentus, we started the characterization of one of the putative ECF sigma factors identified in this bacterium (31). The putative 540-bp open reading frame CC3253 (TIGR nomenclature), which was denominated sigF, was isolated from a partial genomic library, as described in Materials and Methods. A global similarity search of the NCBI database using BLAST demonstrated a high degree of similarity of F to ECF sigma factors not yet characterized, especially from other members of the subdivision of proteobacteria. A global alignment of the predicted F amino acid sequence with the best hit obtained from BLAST showed a 47.2% identity to a Burkholderia fungorum putative ECF sigma factor named Bcep02004893 (NCBI Microbial Genomes Annotation Project). The closest match between an already characterized ECF sigma factor and the predicted F protein is Mycobacterium tuberculosis SigE, which is 28.65% identical to the C. crescentus F sequence (53). This gene was also found in other mycobacteria and was shown to be involved in survival of extreme heat shock, oxidative and acid stress, and detergent treatment (53). Similarity searches of the predicted F protein using the Conserved Domain Database from BLAST have demonstrated the presence of conserved regions 2 and 4 from 70 family members, encompassing amino acids 33 to 93 and 123 to 172, respectively (not shown).
A sigF-null mutant was obtained by constructing an in-frame internal deletion of 345 bp in the sigF coding region, deleting amino acids 25 through 139, corresponding to the entire region 2 and the beginning of region 4 (see Materials and Methods). This mutant strain, denominated SG16, did not show any difference in cell morphology, motility, or colony formation when compared to the parental strain NA1000, indicating that sigF is not essential for survival during normal growth. In addition, a previous analysis using a mutant with sigF interrupted by a spectinomycin resistance cassette has indicated that sigF is not necessary for survival of distinct treatments during the exponential growth phase, including osmotic treatment, acid and alkaline stresses, oxidative stress, and cold shock (data not shown). This insertion mutant was not used for further studies, as sigF probably forms an operon with a downstream gene encoding a conserved hypothetical protein (CC3252), and any phenotype observed in the mutant strain could be the result of changes in the expression of the latter gene. Thus, all the results further described in this work have been obtained with the in-frame deletion mutant SG16.
sigF mutant is sensitive to oxidative stress in the stationary phase of growth. Since no phenotype was observed in the exponential growth phase, we decided to investigate whether sigF could be important for the adaptation to the stationary phase, previously described in C. crescentus (51). The growth curves of NA1000 and SG16 were found to be identical in either complex medium (PYE; Fig. 1A) or minimal medium (M2; not shown), indicating that the absence of F does not affect the C. crescentus growth rate. Analysis of viability and morphology of SG16 cells during long periods of incubation in rich medium (15 days) also showed the same profile as that observed for NA1000 parental cells, including the acquisition of an elongated helical morphology after the ninth day (data not shown) (51).
Sensitivity of SG16 cells to distinct stress conditions during the stationary phase was then analyzed using 24-h cultures. SG16 stationary cells did not present increased sensitivity to heat shock (45°C) or acid stress (pH 4.0) compared to the parental strain NA1000 (data not shown). We also investigated the susceptibility of sigF mutant cells to exposure to high concentrations of hydrogen peroxide. As previously described (51), overnight cultures of NA1000 are quite resistant to 15 mM hydrogen peroxide, with approximately 45% of the cells remaining viable after 120 min of exposure (Fig. 1B). In contrast, SG16 mutant cells did not show this increased resistance to oxidative stress and viability decreased very rapidly after 60 min of exposure to 15 mM hydrogen peroxide, reaching about 0.2% of viable cells after 120 min (Fig. 1B). Sensitivity to hydrogen peroxide was also assessed in 12-h cultures, when cells are at the onset of the stationary phase (Fig. 1A). At this stage of growth, NA1000 cells have not acquired the resistance to this treatment observed in 24-h cultures, and after 15 min of exposure to 15 mM hydrogen peroxide, only about 5% of the parental cells were still viable (not shown). However, SG16 mutant cells were already more sensitive than parental cells to hydrogen peroxide at this stage, with only 0.1% of the mutant cells forming colonies after the same treatment (not shown). No increased sensitivity of mutant cells was observed with hydrogen peroxide treatment in exponential-phase cultures (not shown), which indicates that F becomes essential for survival of oxidative stress as C. crescentus cells enter the stationary growth phase.
We also constructed a conditional mutant, SG16C, carrying a single functional copy of sigF under the control of a xylose-inducible promoter in the xylX locus of the chromosome of SG16 strain. There is no residual expression of the sigF gene when SG16C cells are grown in the absence of xylose, as observed by RT-PCR analysis and immunoblot experiments using an anti-F polyclonal antiserum (data not shown). This conditional mutant strain was used in complementation tests, in order to confirm that the susceptibility of the SG16 strain to hydrogen peroxide during the stationary phase is solely due to the absence of sigF. Overnight cultures of SG16C grown in the absence of xylose were pretreated for 150 min with 0.1% xylose plus 0.1% glucose (induction of sigF expression) or 0.2% glucose (no expression of sigF) before exposure to 15 mM hydrogen peroxide. As depicted in Fig. 1C, addition of xylose before exposure to hydrogen peroxide resulted in increased resistance of mutant cells to oxidative stress, reaching levels close to those observed in parental cells. In contrast, cultures without xylose were still highly sensitive to hydrogen peroxide.
sigF transcription is induced by extreme heat shock during the exponential phase and is autoregulated. As ECF sigma factors have been implicated in the response to extreme heat shock (8), we investigated whether induction of sigF transcription occurs after exposure to extreme temperatures. To address this question, we performed RT-PCR analysis using RNA extracted from cells exposed to 48°C during exponential growth, and our results showed an increase of more than 2.5-fold in sigF mRNA levels after exposure to 48°C for 15 min (Fig. 2A). However, no difference in survival was observed between the parental strain and the sigF mutant strain after exposure to heat shock at 48°C during exponential growth (not shown). These results indicate that, although sigF may be involved in this stress response, its role is not essential. One possibility is that absence of F is compensated for by some of the other ECF sigma factors present in C. crescentus. Indeed, results from our laboratory have shown that other ECF sigma factors are induced by heat stress in this bacterium (data not shown).
Taking advantage of the high level of sigF expression after treatment of the cells at 48°C during exponential growth, we performed primer extension assays to determine the sigF transcription start site. A 52-nucleotide extension product was clearly observed with RNA isolated from NA1000 cells heat shocked at 48°C (Fig. 2B). However, no extension product was detected using RNA from untreated cultures, which is consistent with the RT-PCR data. The size of the extension product mapped the transcription start site to a guanine located 2 nucleotides upstream of the putative translation initiation site proposed by the TIGR annotation. Analysis of the sequence upstream of the putative transcription start site revealed the presence of a candidate –35 promoter region similar to –35 elements recognized by ECF sigma factors in other bacteria (Fig. 2B), which includes the highly conserved "AAC" motif (14). Similarity in the –10 region is difficult to assess because no consensus for this region has been defined for ECF sigma factor promoters.
The fact that most ECF sigma factors described to date are positively autoregulated led us to investigate if this was also true for sigF. To address this question, we performed primer extension experiments using RNA isolated from the sigF mutant, SG16. This was possible because the oligonucleotide used in the assays recognizes a region that was not affected in the sigF mutant. No induction of sigF expression was observed after exposure of sigF mutant cells to 48°C, as the 52-nucleotide band corresponding to the extension product was not detected using RNA isolated from SG16 cells (Fig. 2B). These results suggest that sigF positively autoregulates its own expression during heat shock.
To further confirm that F is capable of positively autoregulating its own expression, we analyzed the effect of increased F levels upon the activity of its own promoter. For this, we cloned the sigF gene in a medium-copy-number vector (pJS14) under the control of a constitutive promoter (Table 1). F levels are more than twofold higher in strain NA1000 harboring this construct (named pCM30), as observed in immunoblots (data not shown). This increased sigF expression did not result in differences in growth rate or viability during normal growth (data not shown). A transcriptional fusion of the sigF promoter to the lacZ gene was then transferred to the strain containing the pCM30 construct, and -galactosidase assays were performed. A 2.5-fold increase in sigF promoter activity was observed in the strain harboring the sigF extra copies, in comparison to NA1000 containing the pJS14 vector alone. This result confirms data obtained from primer extension analysis, discarding the possibility that the extension product was not detected in the mutant strain due to a lower stability of the truncated sigF transcript.
Analysis of sigF gene expression in response to the stationary phase and oxidative stress. Since sigF is proposed to have a role in the response to oxidative stress only during the stationary phase of growth, we analyzed transcription of sigF as a function of the growth phase by RT-PCR experiments. The results demonstrated a significant decrease in sigF mRNA levels upon entry into the stationary phase (12 h of growth at 30°C), which is even more pronounced after 24 h of growth (Fig. 3A). Then, F protein levels as a function of the growth rate were also examined by immunoblots. The amount of F protein in total protein extracts is low during the exponential growth phase (4 h) and increases gradually as a function of cell density, reaching maximal levels in stationary-phase cells (12 and 24 h of growth at 30°C) (Fig. 3B). As shown in Fig. 3B, F protein levels are nearly eightfold higher in stationary-phase cells, when compared to early-exponential-phase cells (4 h). As a control, membrane blots were also incubated with anti-Rho polyclonal antibody, since Rho protein levels do not change significantly as a function of Caulobacter cell density (M. V. Marques, personal communication) (Fig. 3B, lower panel). As sigF mRNA levels decrease during the stationary phase (Fig. 3A), we can conclude that sigF expression is posttranscriptionally controlled in C. crescentus.
sigF gene expression in response to hydrogen peroxide treatment during the exponential and stationary phases was also analyzed. We could not observe induction of sigF transcription in RT-PCR experiments after exposure of NA1000 cells to a wide range (0.06 mM to 15 mM) of H2O2 concentrations, in either the exponential or stationary phase (not shown). Accordingly, there was no increase in F protein levels, as determined in immunoblot experiments, when cells were exposed to different concentrations of hydrogen peroxide during the exponential or stationary phases of growth (not shown).
Effect of endogenous proteases on F protein levels. As our results pointed to a posttranscriptional regulation of sigF expression, we assessed F protein levels in some C. crescentus protease mutant strains characterized previously. Initially, we analyzed the levels of F protein in ClpP and ClpX conditional mutant strains, UJ270 and UJ271, respectively (16). The ClpP/ClpX complex is known to regulate S protein levels in a growth-phase-dependent manner in E. coli (44). The strains used contain a single functional copy of clpP or clpX under the control of a xylose-inducible promoter. As previously described (16), total depletion of ClpX or ClpP is observed after 6 h of growth in the absence of the xylose; therefore, F levels were assessed in cultures grown in the presence or absence of xylose for 8 h (late exponential growth phase) (Fig. 4A). F protein levels remained the same when either ClpP or ClpX was depleted (Fig. 4A), indicating that F levels are not regulated by these proteases.
We also tested the effect of a null mutation in the gene encoding the FtsH protease on F levels during different growth phases. FtsH is an inner membrane ATP-dependent protease which was shown to be required for cell differentiation, growth under salt and heat stresses, and for stationary-phase survival in C. crescentus (11). As depicted in Fig. 4B, F levels are 2.8-fold higher in exponential-phase cultures of the ftsH-null mutant strain UJ945, when compared to the levels observed in the parental strain NA1000, indicating that FtsH is either directly or indirectly involved in the control of F protein levels. Although F levels are still higher in UJ945 after 24 h of growth than those in NA1000, the difference is less pronounced at this time point (Fig. 4B), which suggests that FtsH activity has a more prominent effect in the reduction of F levels in exponential-phase cells. This is consistent with the gradual decrease in FtsH protein levels as cells enter the stationary phase, as determined in immunoblot assays. After 12 and 24 h of growth, a reduction of 20 and 50% in FtsH levels is observed, respectively, compared to the amount observed in exponential-phase cells (Fig. 4C). Thus, reduced FtsH-mediated proteolysis of F during the stationary phase may be an important factor leading to accumulation of F during this growth stage.
Analysis of F protein stability. To further confirm the results described above, we analyzed F stability in exponential- and stationary-phase cultures of NA1000 and the ftsH-null strain. For this purpose, protein synthesis was inhibited by addition of chloramphenicol and F levels were monitored at distinct time points after antibiotic treatment. Experiments were performed using 25 μg of chloramphenicol per ml, as this concentration was reported to immediately block protein synthesis in C. crescentus (33).
As observed in Fig. 5A, F protein levels gradually decrease in exponential-phase cultures of strain NA1000, reaching 30% of the amount observed at time zero, 1 h after the addition of chloramphenicol. In contrast, this protein is highly stabilized in stationary-phase cells (24-h cultures) and no significant change in its levels was observed during the same time interval (Fig. 5B). Similar results were obtained when exponential-phase cells were compared to 12-h cultures (data not shown). In addition, Rho protein levels were also assessed as a control in all experiments, and no difference in its stability was observed between exponential- and stationary-phase cells.
In agreement with the observation that F protein levels are higher in ftsH mutant cells, the sigma factor is also stabilized in exponential-phase cultures of the protease mutant strain, as depicted in Fig. 5C. In addition, no significant decrease in F protein levels was observed in 24-h cultures of strain UJ945 at the times assessed, in a profile similar to that observed for strain NA1000 at this growth stage (Fig. 5D).
These results demonstrate that stabilization of F protein occurs in stationary-phase cultures and corroborate the proposed role for FtsH in the negative regulation of F levels in exponential-phase cultures.
Transcriptome analysis of F-regulated genes. To identify F-regulated genes, we compared the global expression profiles of sigF deletion mutant SG16 with those of the parental strain NA1000 in microarray experiments. Since our data indicated a role for F during the stationary phase, differences in global gene expression were assessed during this phase, using cultures grown for 12 h and 24 h at 30°C, representing the early and late stationary phases, respectively. Any gene showing a reduction in expression of 1.5-fold or more in the SG16 strain in comparison to NA1000 was defined as being regulated by F, either directly or indirectly. A total of eight genes were identified through this analysis (Table 3).
Genes of interest were selected for validation of microarray data by semiquantitative RT-PCR. RNA samples from exponentially growing cultures were also used to assess gene expression by RT-PCR, in order to verify if the gene of interest shows higher expression levels upon entry into the stationary phase and if it is also F regulated during exponential growth.
Microarray data demonstrated that gene msrA, encoding the enzyme peptide methionine sulfoxide reductase (MsrA), displays a 1.9-fold reduction in expression in 24-h cultures of the mutant strain SG16 (Table 3). MsrA is an antioxidant enzyme, which reduces oxidized methionine sulfoxide [Met(O)] residues in proteins back to methionine. These residues occur as a consequence of aerobic metabolism, due to formation of reactive oxygen and nitrogen species (ROS and RNI, respectively) (50). RT-PCR analysis confirmed the reduced expression of msrA in the mutant strain during the stationary phase and showed that the gene is also dependent on F during exponential growth (Fig. 6A). Although the difference in msrA expression in 12-h cultures was not considered statistically significant in our microarray experiments (Table 3), results from RT-PCR analysis did show markedly reduced levels of msrA mRNA at this time point. Altogether, these results indicate that F regulates msrA expression during the exponential and stationary phases. In addition, RT-PCR experiments showed that transcription of msrA is not induced upon entry into stationary phase in C. crescentus (Fig. 6A).
Another gene analyzed by RT-PCR was sodA, which encodes the enzyme (Mn2+)-superoxide dismutase. This enzyme catalyzes the dismutation of O2·– to molecular oxygen (O2) and H2O2, which in turn, is detoxified by catalases and peroxidases. Consistent with microarray data, expression of sodA was shown in RT-PCR experiments to be highly reduced in 12-h cultures of strain SG16 compared to the parental strain. After 24 h of growth, sodA mRNA levels were shown to be quite similar in both mutant and parental cells, in agreement with microarray results (Fig. 6B). The RT-PCR analysis also showed that sodA expression is highly reduced in the parental strain NA1000 during the stationary phase (12 h and 24 h) in comparison to expression during exponential growth (4 h). As SodA activity has been mainly implicated in responses to superoxide radical, we investigated the sensitivity of SG16 cells to paraquat and xanthine oxidase, which are cytoplasmic and extracytoplasmic generators of superoxide, respectively. However, we could not detect increased sensitivity of mutant cells to these treatments using 12-h cultures (data not shown).
Semiquantitative RT-PCR assays were also performed to confirm differential expression of CC3255, which encodes a conserved hypothetical protein, between the two strains. This is the first gene of a probable operon located upstream of sigF and in the opposite orientation, which also contains CC3257 (whose microarray data also showed reduced expression in strain SG16). As shown in Fig. 6C, we could confirm the lower expression of CC3255 in 12-h and 24-h cultures of the SG16 strain. In addition, expression of CC3255 is also dependent on F during exponential growth, as its mRNA levels are significantly diminished in the sigF mutant in 4-h cultures. RT-PCR data also showed that CC3255 expression is highly induced upon entry into the stationary phase (12 h) in the wild-type background, decreasing again after 24 h of growth.
Gene CC3572, encoding a putative carbonic anhydrase, was also down-regulated in 12-h and 24-h cultures of strain SG16 in our microarray data. This enzyme catalyzes the interconversion of CO2 and HCO3– (bicarbonate anion), and evidence from studies of eukaryotic carbonic anhydrases indicate a role for these enzymes in protection against oxidative stress (13, 37, 57). However, we could not confirm reduced expression of this gene in strain SG16 in semiquantitative RT-PCR experiments (data not shown).
DISCUSSION
This work reports the first characterization of an ECF sigma factor from C. crescentus, named F, which is essential for the oxidative stress response during the stationary phase of growth. The response to these growth conditions may be specially important for C. crescentus physiology, as this bacterium is submitted to high oxygen concentrations in its natural habitat due to association with cyanobacteria (2, 22) and it is mostly found in diluted, nutritionally poor environments (36). Although some aspects of these environmental responses have been described previously (42, 46, 51), the only regulator identified to date is skgA, involved in katG induction in the stationary phase (38).
Despite its role in oxidative stress, sigF expression is not induced by H2O2 treatment. In contrast, F levels are induced upon entry into the stationary phase of growth, indicating that growth rate is the main factor that regulates F expression. A similar situation was observed for Salmonella enterica serovar Typhimurium rpoE expression (47). Stationary-phase cells are submitted to distinct stresses, and transition to this growth phase causes induction of many genes whose products are involved in protection against such adverse conditions. However, a growing body of evidence indicates that oxidative injury is the major challenge faced by stationary-phase cells and that most genes induced by stasis are involved in defense against oxidative stress (32). Thus, stationary-phase induction of sigF expression can be explained by its role in regulating responses to cope with oxidative stress. This assumption is substantiated by our transcriptome analysis, which identified at least two genes involved in resistance to oxidative stress as F dependent: msrA and sodA.
SodA activity has been shown to be involved in resistance to H2O2, as accumulation of O2·– leads to higher levels of free iron in the cell, which can react with H2O2 through the Fenton reaction, producing the highly reactive compound OH– (48). Our results show that the sodA gene is F dependent only in 12-h cultures, indicating that a regulator distinct from F is responsible for sodA levels during the exponential phase. In addition, sodA transcript levels are restored in the sigF mutant after 24 h, which may be due to the action of another sigma factor, compensating for the absence of F at this stage.
The msrA gene, which encodes another important antioxidant enzyme, has been shown to be necessary for survival of H2O2 exposure in a series of microorganisms, including E. coli (10, 30). Interestingly, C. crescentus msrA was observed to be F dependent at all stages of growth analyzed. No increased sensitivity to H2O2 was observed in strain SG16 during the exponential phase, which may indicate that correct expression of msrA (CC1039) may be more necessary for protection against oxidative stress during stationary phase. C. crescentus possesses an msrA paralog, named CC0994, which may be sufficient for protection against oxidative damage during the exponential phase.
Our transcriptome analysis led to the identification of a total of eight F-regulated genes. This result may suggest that F, like other ECF sigma factors described, regulates a small, defined set of genes under specific conditions. In addition, the absence of sigF does not lead to the complete absence of expression of the target genes, indicating that they are probably recognized by at least one additional sigma factor. These other sigma factors may have distinct or overlapping promoter specificity with F and could compensate for the absence of F at some target genes. The overlap between ECF regulons has been described for B. subtilis X and W (15) and was also suggested for S. coelicolor R (35). This may account for the minor changes in expression of target genes in a sigF-null background, complicating the identification of F-dependent genes using microarrays.
We performed a search for putative promoter sequences upstream of the F-regulated genes, using the program FINDPATTERNS (Genetics Computer Groups). This search was performed based on similarity to the –35 region of the sigF promoter determined in this study, since the sigF gene was shown to be positively autoregulated. This region is more conserved among ECF-dependent promoters than the –10 element (14). For some genes, we could identify putative –35 sequences and we also identified similarity at the –10 region (Table 4). However, for a more accurate identification of F promoters, determination of the transcription start sites for each gene is required.
Our results demonstrate that sigF possesses a complex pattern of regulation under distinct growth conditions. Analysis of sigF transcription in response to heat shock during the exponential phase and overexpression of sigF indicate that F positively autoregulates its own transcription under these conditions. This result was expected, since positive autoregulation of ECF gene expression has been demonstrated for most members of this subfamily. However, a distinct control of sigF transcription seems to occur during the stationary phase, since the increase observed in F protein levels does not lead to larger amounts of sigF mRNA. On the contrary, there is a significant reduction in sigF transcript levels during the stationary phase. These data suggest that the sigF promoter possesses a different selectivity during this stage, being less efficiently transcribed by EF or being regulated by another sigma factor, which may compete for the same promoter element and/or for the RNA polymerase core. Indeed, results from our laboratory indicate that at least two other members of the ECF subfamily are induced during the stationary phase in C. crescentus (data not shown). Distinct selectivity at the sigF promoter may be accounted for by the global changes that occur during transition between the exponential and stationary phases, including alterations in the DNA topology, the presence of nucleoid-associated proteins, and activators or repressors.
The sigF gene is probably cotranscribed with CC3252, coding for a conserved hypothetical protein, which may act as an anti-sigma factor. This is suggested by the presence of transmembrane domains in the protein encoded by CC3252 and by the fact that its orthologues in other bacteria are always found in operons containing an ECF sigma factor with a high degree of similarity to sigF as the first gene. Accordingly, analysis of CC3252 expression by RT-PCR also showed reduced transcription during the stationary phase, in a pattern similar to sigF (data not shown). In this sense, reduced transcription of the operon in the stationary phase may be an advantageous mechanism, preventing the consequent inactivation of F due to increased expression of its putative anti-sigma factor. This may be especially important during the stationary phase as a rapid shutoff of the response is not needed. We are presently investigating the involvement of the putative anti-sigma factor in the control of F activity by constructing a null strain in CC3252 as well as an overexpressing strain.
In addition, the posttranscriptional induction of sigF gene expression during the stationary phase may be a more advantageous mechanism of regulation in a growth stage when metabolic activities, including transcription and translation, are greatly diminished. Similarly, induction of E. coli rpoS gene expression during glucose starvation occurs most significantly by increased protein stability, in spite of decreased transcription and translation (55).
We have also demonstrated that FtsH protease is directly or indirectly involved in the regulation of F levels. In a similar manner, levels of B. subtilis W, a member of the ECF subfamily, are increased in an ftsH knockout strain (54). In C. crescentus, FtsH was previously shown to control the cellular levels of the heat shock sigma factor 32 in the presence of DnaK (6, 11). Our results also indicate that F is less susceptible to FstH action during the stationary phase, thus being more stable. By analogy to the 32 and S regulation models, F may associate with some accessory factor that promotes its degradation during the exponential phase (4, 56). In addition, we showed that FtsH is present at lower levels during the stationary phase, which may indicate lower activity of this protease at this stage. Furthermore, the increase in the amount of F observed in the ftsH mutant led to higher F activity, since msrA transcript levels determined in RT-PCR experiments increased from 1.5- to 2-fold in the protease mutant strain during the exponential phase, compared to the amount present in the parental strain (data not shown). Thus, degradation of F by FtsH should be an important regulatory component of its activity. Nevertheless, other proteases may be involved in the regulation of F levels, as F still accumulates in the ftsH mutant during the stationary phase, although to a lesser extent (see Fig. 4B). In a similar manner, a set of ATP-dependent proteases, including HslVU, act synergistically with FtsH in the negative regulation of 32 levels in E. coli (18).
Since the sigF-null mutant does not display a deficiency in stationary-phase survival, as observed for the stationary-phase sigma factor rpoS in E. coli (21), other sigma factors may also be involved in stationary-phase responses in C. crescentus. Thus, characterization of other ECF subfamily members and their regulons and regulators is of great importance for the understanding of C. crescentus physiology and responses to stress.
ACKNOWLEDGMENTS
We are indebted to Lucy Shapiro and Harley McAdams (Stanford University) for making possible the realization of the microarray experiments and to Lucy Shapiro for critical reading of the manuscript. We also thank Maliwan Meewan for performing the microarray assays and Alison Kay Hottes and Ann Reisenauer for helpful advice during the design and analysis of these experiments. We thank Urs Jenal for generously providing ftsH, clpX, and clpP mutant strains; Teru Ogura for anti-FtsH antiserum; and Marílis V. Marques for anti-Rho antiserum. We also thank Tie Koide and Ricardo Z. N. Vêncio for help with microarray data analysis.
REFERENCES
Alba, B. M., J. A. Leeds, C. Onufryk, C. Z. Lu, and C. A. Gross. 2002. DegS and YaeL participate sequentially in the cleavage of RseA to activate the sigma(E)-dependent extracytoplasmic stress response. Genes Dev. 16:2156-2168.
Allen, H. L. 1971. Primary productivity, chemoorganotrophy, and nutritional interactions of epiphytic algae and bacteria on macrophytes in the littoral of a lake. Ecol. Monogr. 41:97-127.
Anderson, D. K., N. Ohta, J. Wu, and A. Newton. 1995. Regulation of the Caulobacter crescentus rpoN gene and function of the purified sigma 54 in flagellar gene transcription. Mol. Gen. Genet. 246:697-706.
Blaszczak, A., C. Georgopoulos, and K. Liberek. 1999. On the mechanism of FtsH-dependent degradation of the sigma 32 transcriptional regulator of Escherichia coli and the role of the Dnak chaperone machine. Mol. Microbiol. 31:157-166.
Browning, D. F., D. E. Whitworth, and D. A. Hodgson. 2003. Light-induced carotenogenesis in Myxococcus xanthus: functional characterization of the ECF sigma factor CarQ and antisigma factor CarR. Mol. Microbiol. 48:237-251.
da Silva, A. C. A., R. C. G. Simo, M. F. Susin, R. L. Baldini, M. Avedissian, and S. L. Gomes. 2003. Downregulation of the heat shock response is independent of DnaK and 32 levels in Caulobacter crescentus. Mol. Microbiol. 49:541-553.
De las Penas, A., L. Connolly, and C. A. Gross. 1997. The sigmaE-mediated response to extracytoplasmic stress in Escherichia coli is induced by RseA and RseB, two negative regulators of sigma E. Mol. Microbiol. 24:373-385.
Erickson, J. W., and C. A. Gross. 1989. Identification of the sigma E subunit of Escherichia coli RNA polymerase: a second alternate sigma factor involved in high-temperature gene expression. Genes Dev. 3:1462-1471.
Evinger, M., and N. Agabian. 1977. Envelope-associated nucleoid from Caulobacter crescentus stalked and swarmer cells. J. Bacteriol. 132:294-301.
Ezraty, B., L. Aussel, and F. Barras. 2005. Methionine sulfoxide reductases in prokaryotes. Biochim. Biophys. Acta 1703:221-229.
Fischer, B., G. Rummel, P. Aldridge, and U. Jenal. 2002. The FtsH protease is involved in development, stress response and heat shock control in Caulobacter crescentus. Mol. Microbiol. 44:461-478.
Gober, J. W., and L. Shapiro. 1992. A developmentally regulated Caulobacter flagellar promoter is activated by 3' enhancer and IHF binding elements. Mol. Biol. Cell 3:913-916.
Gotz, R., A. Gnann, and F. K. Zimmerman. 1999. Deletion of the carbonic anhydrase-like gene NCE103 of the yeast Saccharomyces cerevisiae causes an oxygen-sensitive growth defect. Yeast 15:855-864.
Helmann, J. 2002. The extracytoplasmic function (ECF) sigma factors. Adv. Microb. Physiol. 46:47-110.
Huang, X., K. L. Fredrick, and J. D. Helmann. 1998. Promoter recognition by Bacillus subtilis W: autoregulation and partial overlap with the X regulon. J. Bacteriol. 180:3765-3770.
Jenal, U., and T. Fuchs. 1998. An essential protease involved in bacterial cell-cycle control. EMBO J. 17:5658-5669.
Kanehara, K., K. Ito, and Y. Akiyama. 2002. YaeL (EcfE) activates the sigma(E) pathway of stress response through a site-2 cleavage of anti-sigma(E), RseA. Genes Dev. 16:2147-2155.
Kanemori, M., K. Nishihara, H. Yanagi, and T. Yura. 1997. Synergistic roles of HslVU and other ATP-dependent proteases in controlling in vivo turnover of 32 and abnormal proteins in Escherichia coli. J. Bacteriol. 179:7219-7225.
Kang, J. G., M. S. Paget, Y. J. Seok, M. Y. Hahn, J. B. Bae, J. S. Hahn, C. Kleanthous, M. J. Buttner, and J. H. Roe. 1999. RsrA, an anti-sigma factor regulated by redox change. EMBO J. 18:4292-4298.
Koide, T., P. A. Zaini, L. M. Moreira, R. Z. N. Vecio, A. Y. Matsukuma et al. 2004. DNA microarray-based genome comparison of a pathogenic and a nonpathogenic strain of Xylella fastidiosa delineates genes important for bacterial virulence. J. Bacteriol. 186:5442-5449.
Lange, R., and R. Hengge-Aronis. 1991. Identification of a central regulator of stationary-phase gene expression in Escherichia coli. Mol. Microbiol. 5:49-59.
Lapteva, N. A. 1987. Ecological features of distribution of bacteria of the genus Caulobacter in freshwater bodies. Mikrobiologiya 56:677-684. (In Russian; English translation, 56:537-543.)
Lonetto, M., K. Brown, K. Rudd, and M. Buttner. 1994. Analysis of the Streptomyces coelicolor sigE gene reveals the existence of a subfamily of eubacterial RNA polymerase factors involved in the regulation of extracytoplasmic functions. Proc. Natl. Acad. Sci. USA 91:7573-7577.
Malakooti, J., and B. Ely. 1995. Principal sigma subunit of the Caulobacter crescentus RNA polymerase. J. Bacteriol. 177:6854-6860.
Marshak, D. R., J. T. Kadonaga, R. Burgess, M. W. Knuth, J. R. Brennan, and S. Lin. 1996. Strategies for protein purification and characterization. A laboratory course manual, p. 214-216. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y.
McGrath, P. T., P. Viollier, and H. H. McAdams. 2004. Setting the pace: mechanisms tying Caulobacter cell-cycle progression to macroscopic cellular events. Curr. Opin. Microbiol. 7:192-197.
Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Missiakas, D., M. P. Mayer, M. Lemaire, C. Georgopoulos, and S. Raina. 1997. Modulation of the Escherichia coli E (RpoE) heat-shock transcription-factor activity by the RseA, RseB and RseC proteins. Mol. Microbiol. 24:355-371.
Missiakas, D., and S. Raina. 1998. The extracytoplasmic function sigma factors: role and regulation. Mol. Microbiol. 28:1059-1066.
Moskovitz, J., M. A. Rahman, J. Strassman, S. O. Yancey, S. R. Kushner, N. Brot, and H. Weissbach. 1995. Escherichia coli peptide methionine sulfoxide reductase gene: regulation of expression and role in protecting against oxidative damage. J. Bacteriol. 177:502-507.
Nierman, W. C., T. V. Feldblyum, M. T. Laub, I. T. Paulsen, K. E. Nelson, J. A. Eisen, J. F. Heidelberg, M. R. Alley, N. Ohta, J. R. Maddock, I. Potocka, W. C. Nelson, A. Newton, C. Stephens, N. D. Phadke, B. Ely, R. T. DeBoy, R. J. Dodson, A. S. Durkin, M. L. Gwinn, D. H. Haft, J. F. Kolonay, J. Smit, M. B. Craven, H. Khouri, J. Shetty, K. Berry, T. Utterback, K. Tran, A. Wolf, J. Vamathevan, M. Ermolaeva, O. White, S. L. Salzberg, J. C. Venter, L. Shapiro, and C. M. Fraser. 2001. Complete genome sequence of Caulobacter crescentus. Proc. Natl. Acad. Sci. USA 98:4136-4141.
Nystrom, T. 2004. Stationary-phase physiology. Annu. Rev. Microbiol. 58:161-181.
Osley, M. A., and A. Newton. 1978. Regulation of cell cycle events in asymmetrically dividing cells: functions required for DNA initiation and chain elongation in Caulobacter crescentus. J. Bacteriol. 135:10-17.
Paget, M. S., J. G. Kang, J. H. Roe, and M. J. Buttner. 1998. sigma R, a RNA polymerase sigma factor that modulates expression of the thioredoxin system in response to oxidative stress in Streptomyces coelicolor A3(2). EMBO J. 17:5776-5782.
Paget, M. S., V. Molle, G. Cohen, Y. Aharonowitz, and M. J. Buttner. 2001. Defining the disulphide stress response in Streptomyces coelicolor A3(2): identification of the sigmaR regulon. Mol. Microbiol. 42:1007-1020.
Poindexter, J. S. 1964. Biological properties and classification of the Caulobacter group. Bacteriol. Rev. 28:231-295.
Raisanen, S. R., P. Lehenkari, M. Tasanen, P. Rahkila, P. L. Harkonen, and H. K. Vaananen. 1999. Carbonic anhydrase III protects cells from hydrogen peroxide-induced apoptosis. FASEB J. 13:513-522.
Rava, P. S., L. Somma, and H. M. Steinman. 1999. Identification of a regulator that controls stationary-phase expression of catalase-peroxidase in Caulobacter crescentus. J. Bacteriol. 181:6152-6159.
Reisenauer, A., C. D. Mohr, and L. Shapiro. 1996. Regulation of a heat shock 32 homolog in Caulobacter crescentus. J. Bacteriol. 178:1919-1927.
Ryan, K. R., and L. Shapiro. 2003. Temporal and spatial regulation in prokaryotic cell cycle progression and development. Annu. Rev. Biochem. 72:367-394.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Schnell, S., and H. M. Steinman. 1995. Function and stationary-phase induction of periplasmic copper-zinc superoxide dismutase and catalase/peroxidase in Caulobacter crescentus. J. Bacteriol. 177:5924-5929.
Schurr, M. J., H. Yu, J. M. Martinez-Salazar, J. C. Boucher, and V. Deretic. 1996. Control of AlgU, a member of the E-like family of stress sigma factors, by the negative regulators MucA and MucB and Pseudomonas aeruginosa conversion to mucoidy in cystic fibrosis. J. Bacteriol. 178:4997-5004.
Schweder, T., K.-H. Lee, O. Lomovskaya, and A. Matin. 1996. Regulation of Escherichia coli starvation sigma factor (s) by ClpXP protease. J. Bacteriol. 178:470-476.
Simon, R., U. Prieffer, and A. Puhler. 1983. A broad host range mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. Biotechnology 1:784-790.
Steinman, H. M., F. Fareed, and L. Weinstein. 1997. Catalase-peroxidase of Caulobacter crescentus: function and role in stationary-phase survival. J. Bacteriol. 179:6831-6836.
Testerman, T. L., A. Vazquez-Torres, Y. Xu, J. Jones-Carson, S. J. Libby, and F. C. Fang. 2002. The alternative sigma factor E controls antioxidant defences required for Salmonella virulence and stationary-phase survival. Mol. Microbiol. 43:771-782.
Touati, D. 2000. Iron and oxidative stress in bacteria. Arch. Biochem. Biophys. 373:1-6.
Tsai, J. W., and M. R. K. Alley. 2000. Proteolysis of the McpA chemoreceptor does not require the Caulobacter major chemotaxis operon. J. Bacteriol. 182:504-507.
Weissbach, H., F. Etienne, T. Hoshi, S. H. Heinemann, W. T. Lowther, B. Matthews, G. St. John, C. Nathan, and N. Brot. 2002. Peptide methionine sulfoxide reductase: structure, mechanism of action, and biological function. Arch. Biochem. Biophys. 397:172-178.
Wortinger, M. A., E. M. Quardokus, and Y. V. Brun. 1998. Morphological adaptation and inhibition of cell division during stationary phase in Caulobacter crescentus. Mol. Microbiol. 29:963-973.
Wu, J., and A. Newton. 1996. Isolation, identification and transcriptional specificity of the heat shock sigma factor 32 from Caulobacter crescentus. J. Bacteriol. 178:2094-2101.
Wu, Q.-L., D. Kong, K. Lam, and R. N. Husson. 1997. A mycobacterial extracytoplasmic function sigma factor involved in survival following stress. J. Bacteriol. 179:2922-2929.
Zellmeier, S., U. Zuber, W. Schumann, and T. Wiegert. 2003. The absence of FtsH metalloprotease activity causes overexpression of the W-controlled pbpE gene, resulting in filamentous growth of Bacillus subtilis. J. Bacteriol. 185:973-982.
Zgurskaya, H. I., M. Keyhan, and M. Matin. 1997. The S levels in starving Escherichia coli cells increases solely as a result of its increased stability, despite decreased synthesis. Mol. Microbiol. 24:643-651.
Zhou, Y., S. Gottesman, J. R. Hoskins, M. R. Maurizi, and S. Wickner. 2001. The RssB response regulator directly targets sigma(S) for degradation by ClpXP. Genes Dev. 15:627-637.
Zimmerman, U. J., P. Wang, X. Zhang, S. Bogdanovich, and R. Forster. 2004. Anti-oxidative response of carbonic anhydrase III in skeletal muscle. IUBMB Life 56:343-347.(Cristina E. Alvarez-Marti)