|Home | About | Journals | Submit | Contact Us | Français|
Lactobacillus plantarum ctsR was characterized. ctsR was found to be cotranscribed with clpC and induced in response to various abiotic stresses. ctsR deletion conferred a heat-sensitive phenotype with peculiar cell morphological features. The transcriptional pattern of putative CtsR regulon genes was examined in the ΔctsR mutant. Direct CtsR-dependent regulation was demonstrated by DNA-binding assays using recombinant CtsR and the promoters of the ctsR-clpC operon and hsp1.
In order to respond to stressful conditions, microorganisms have evolved multiple adaptive mechanisms. Indeed, external signals can elicit dramatic changes in the expression pattern of a variety of stress-related genes encoding proteins thought to improve adaptation to the changing environment. Heat shock induces the synthesis of a specific set of proteins, known as heat shock proteins (HSPs). Some of these, such as GroEL, DnaK, small HSP, and several Clp ATPases, are molecular chaperones that facilitate the proper folding of cellular proteins (32). Others, such as the Clp ATP-dependent protease, degrade incorrectly folded proteins (14).
In Bacillus subtilis, the model organism of Gram-positive bacteria, the heat shock response involves at least six different classes of heat-inducible genes that are distinguished by their regulatory mechanisms (25). Class III genes are transcriptionally controlled by the class III stress gene repressor, CtsR, which binds to a specific heptanucleotide direct repeat (5′ RGTCADN NAN RGTCADN 3′) referred to as the CtsR box (6).
Among Gram-positive bacteria, CtsR mainly regulates the expression of genes encoding the Clp ATPases and the ClpP protease (25). However, in staphylococci, the CtsR regulon also comprises the groESL and dnaK operons (3) and, in Oenococcus oeni, even a small-HSP gene (15). Recently, we have reported that the Lactobacillus plantarum ftsH gene, encoding a membrane-bound metalloprotease, is a novel member of the CtsR stress response regulon, representing an exception with respect to other Gram-positive species (11).
Lactobacillus plantarum, a facultative heterofermentative lactic acid bacterium (LAB), is extremely widespread in the environment. It is a natural inhabitant of the human gastrointestinal tract and widely used as a starter and probiotic in many food processes for which stress conditions such as heat, cold, and acidity are common. The survival of L. plantarum under harsh and diverse conditions indicates that it has developed several tolerance and resistance mechanisms (20, 28). Although some stress genes have been characterized in this species (9, 27, 29, 30), little is known about their transcriptional regulation.
Here we characterize ctsR in relation to stress response in L. plantarum.
ctsR of L. plantarum WCFS1 (20) was studied by DNA sequence analysis, 5′ rapid amplification of cDNA ends (RACE), and reverse transcription-PCR (RT-PCR). The genomic organization of ctsR is shown in Fig. S1 in the supplemental material. The ctsR open reading frame encodes a 155-amino-acid (aa) protein with a predicted molecular mass of 17.6 kDa. The protein shares 45% aa sequence identity with CtsR from B. subtilis and displays even higher similarity with respect to CtsRs from other Lactobacillus species. The highest levels of identity are found in the HTH (helix-turn-helix) motif region (residues 25 to 48; 56 to 92% identity), which is responsible for specific DNA sequence recognition and binding; in the amino-terminal domain (residues 6 to 15; 80 to 90% identity); and in the central glycine-rich region (residues 59 to 70; 92 to 100% identity), playing a putative heat-sensing function (7).
The transcriptional start site was determined by 5′-RACE analysis of total RNA extracted from exponentially growing bacterial cells (optical density at 600 nm [OD600] = 0.6) following incubation at 40°C for 10 min. The initiation site was mapped to position −31 relative to the translational start codon (see Fig. S1 in the supplemental material). Oligonucleotides used for 5′-RACE analysis and other experiments in this study are listed in Table S1 in the supplemental material. Putative −35 and −10 hexameric boxes were identified at proper distances. The promoter contains a 17-bp sequence (AGTCAAT ATT AGTCAAA) closely resembling the suggested consensus heptanucleotide direct-repeat CtsR binding site RGTCADN NAN RGTCADN, so far defined in several Gram-positive bacteria (6). This potential CtsR operator overlaps the transcriptional start site, consistent with a role for CtsR as a repressor and in agreement with previously characterized σA-dependent promoters under CtsR-mediated negative control (6, 25). RT-PCR analysis, using appropriate primers spanning ctsR and the downstream clpC coding region, demonstrated that ctsR and clpC are cotranscribed (data not shown). A canonical Rho-independent terminator was found downstream of clpC, indicating a bicistronic operon organization. In contrast to that of B. subtilis but similar to those of streptococci (26) and LAB, such as Lactococcus lactis (35) and O. oeni (15), L. plantarum's ctsR-clpC operon lacks genes encoding potential CtsR modulators (McsA and McsB [modulators of CtsR repression]) (19), thus suggesting that regulation of CtsR might require alternative pathways.
ctsR mRNA levels were analyzed by quantitative RT-PCR (qRT-PCR) following the exposure of exponential-phase cultures (OD600 nm = 0.6) to a range of abiotic stresses, mimicking conditions commonly encountered by this bacterial species (for details, see Table S2 in the supplemental material). Total RNAs were extracted from control and stressed cultures, reverse transcribed, and subjected to qRT-PCR with appropriate primers (see Table S1 in the supplemental material) on an Applied Biosystems 7300 real-time PCR instrument, using SYBR green I detection and the ldhD and gyrA genes as internal controls (10, 11, 12). Strong induction of ctsR expression was observed in response to heat shock (35-fold increase) and osmotic stress (14-fold increase); more marked induction resulted from combined heat and osmotic stress (120-fold increase). A weaker increase (2-fold) in mRNA level was also observed in ethanol-stressed cultures. This finding substantiates the results of previous studies demonstrating that the ctsR-clpC operon is induced by stresses other than heat shock (1, 23).
To understand the relevance of CtsR in L. plantarum, the phenotype of a previously generated ctsR deletion mutant strain was investigated (11). Growth rates of wild-type and mutant strains were analyzed by OD600 monitoring and direct plate counting under the following different culture conditions: at 28°C (control temperature); at 40°C (heat stress); and at 28°C in MRS broth containing 2% ethanol (vol/vol), 0.05% (wt/vol) porcine bile, or 0.8 M NaCl or in HCl-acidified MRS broth (pH 5). In comparison to the wild type, the ΔctsR strain displayed a similar growth rate when cultivated under optimal temperature and medium conditions. However, a growth impairment became evident in cultures of the mutant strain growing either under heat stress or in the presence of ethanol (Fig. 1A and B), suggesting the involvement of ctsR in adaptation to these kinds of stress. By spotting serial dilutions of exponentially growing cultures on MRS plates, we noted that incubation at 40°C strongly inhibited the growth of the mutant strain without affecting the wild type (Fig. (Fig.22).
To learn more about the role of CtsR under heat shock conditions, the ability of the mutant to tolerate a short intense heat stress and to induce thermotolerance was examined. When exponentially growing cultures were challenged for 15 min at 50°C, the number of surviving cells, as determined by CFU counting, was not significantly different between the wild type and the ΔctsR strain (data not shown). Similarly, no significant differences in recovery and growth capacity rates were observed when cells were pre-exposed to 40°C for 30 min and subsequently shifted to 50°C for 15 min, suggesting that ctsR inactivation does not affect the ability of L. plantarum to induce thermotolerance. It is well known that pre-exposure at a sublethal temperature enables L. plantarum to better survive for several minutes at the lethal temperature of 50°C (32). Our data indicate that CtsR does not play an essential role in this adaptation process; however, they highlight a crucial role of CtsR for coping with prolonged heat stress. This is a novel, unexpected finding, since CtsR inactivation generally causes derepression of heat stress genes, thus rendering cells more prepared to face stressful conditions. Indeed, in most Gram-positive bacteria analyzed so far, ctsR mutations increased heat resistance and/or general stress tolerance (4, 16, 17, 18, 24, 36).
The cell morphology of wild-type L. plantarum and ΔctsR mutant strains was analyzed before and after heat stress exposure by scanning electron microscopy (SEM) (Fig. (Fig.3),3), using a Philips XL30 ESEM scanning electron microscope as reported before (5). Under optimal growth temperature conditions, cells exhibited the characteristic rod-shaped, smooth-surface morphology of L. plantarum, and no relevant difference between wild-type and ΔctsR strains could be observed. In contrast, after heat shock, compared to those of the wild type, the mutant cell envelopes looked stiffer and presented somewhat fissured surfaces. These intriguing features suggest that the cell wall might be critically damaged in the ΔctsR strain, pointing to a novel function of CtsR for cell wall integrity control in L. plantarum. Indeed, in B. subtilis, key enzymes of cell wall biosynthesis have been proven to be specific substrates of Clp proteolytic complexes (21), which, in turn, are transcriptionally regulated by CtsR.
To identify members of the L. plantarum CtsR regulon, a detailed DNA motif analysis of the L. plantarum WCFS1 genome (20) was carried out using the CtsR operator consensus sequence (6) and Artemis version 9 software (Sanger Institute; http://www.sanger.ac.uk/Software/Artemis). Potential CtsR binding sites were found in the promoter regions of 10 L. plantarum genes (see Fig. S2 in the supplemental material), including ftsH, for which involvement of CtsR in transcriptional regulation was previously reported (11). Putative CtsR boxes were detected upstream from ctsR-clpC, clpB, clpE, clpP, and hsp1, encoding one of the three small heat shock proteins (HSP20 family) of L. plantarum (20, 29, 34). Likely CtsR operators were also found in the promoters of four genes annotated as lp_0836 (nrpR1), lp_1995, lp_2090, and lp_2942 (20). Except for nrpR1, these latter genes do not seem to encode heat stress-related functions, as confirmed by the lack of appreciable transcriptional induction following heat shock (data not shown). Conversely, hsp1 and clp genes were substantially induced by thermal upshift (data not shown), confirming their involvement in heat stress response mechanisms.
The transcript levels of all the putative CtsR-controlled genes in both the ΔctsR and wild-type strain were investigated by qRT-PCR (Fig. (Fig.4).4). Total RNA was purified from exponentially growing cultures (cultivated at 28°C), retrotranscribed, and analyzed using appropriate oligonucleotides (see Table S1 in the supplemental material) and the real-time PCR procedure described above. The transcription of five of the genes investigated, namely, hsp1, clpC (examined as part of the ctsR-clpC operon), clpB, clpE, and clpP, was significantly derepressed in the mutant relative to their transcription in the wild-type control, with transcript levels increasing 12-, 7-, 2-, 2.6-, and 2.5-fold, respectively. Conversely, the expression of the other four genes was considerably repressed (approximately 10-fold lower). These results strongly suggest a negative regulation exerted by CtsR on its own operon, as well as on hsp1 and clp genes, in agreement with previously characterized CtsR regulons (2, 4, 8, 15, 24, 35). On the other hand, since the transcription of nrpR1, lp_1995, lp_2090, and lp_2942 appears affected by CtsR inactivation, we cannot rule out that the expression of these genes might be controlled by CtsR activity either directly or indirectly.
An in vitro approach was used to demonstrate the direct interaction of CtsR with its putative target sites in the hsp1 and ctsR-clpC promoters. CtsR of L. plantarum was cloned, overexpressed, and purified as described elsewhere (6, 11) and used in gel shift assays with biotin-labeled, PCR-generated DNA fragments corresponding to the promoter regions of the ctsR-clpC operon and hsp1 (Fig. 5A and B, respectively). Recombinant CtsR was able to bind specifically to the promoters, with progressive displacement of the free DNA fragment as the protein concentration increased. The binding was specific for both promoter regions, as shown by the absence of any protein-DNA complex when CtsR was incubated with promoter fragments lacking the CtsR target sequence.
These results confirm the quantitative transcriptional analysis of hsp1 and clpC in the ΔctsR strain, indicating that CtsR controls the expression of hsp1 and that of its own operon by interacting directly with their promoters. A small heat shock gene named Lo18 was also reported to be controlled by CtsR in O. oeni (15); therefore, L. plantarum hsp1 would represent an additional example of ctsR-mediated regulation of small HSPs.
Regulatory mechanisms of stress response genes have been investigated in several Gram-positive bacteria and studied in detail in the genetic model B. subtilis. Conversely, very little is known about the transcriptional regulation of stress genes in L. plantarum, a versatile LAB which is likely to have evolved multiple and complex adaptive mechanisms. Comparative genomics revealed that CtsR and its target sequences are highly conserved among Gram-positive bacteria with low G+C content (6, 26). However, the CtsR-dependent stress response may vary significantly in terms of the composition of the regulon, combination of regulatory patterns, and control of CtsR activity (2, 3, 15, 35).
The results of this work indicate, for the first time, that L. plantarum CtsR is necessary for survival under stress conditions and, most importantly, plays a critical role for growth at high temperature. Since the heat sensitivity of the ΔctsR strain became detectable only after long incubation periods and not after short heat shocks, CtsR seems to be involved in coping with a prolonged heat stress regimen. Interestingly, our phenotypic analyses suggest that CtsR does not contribute to the induction of thermotolerance. The growth impairment and the intriguing morphological changes observed in the ΔctsR cells under heat stress might derive from a transcriptional dysregulation consequential to the ctsR inactivation and point to a novel protective function of CtsR for cell wall homeostasis.
Similarly to CtsRs of other Gram-positive bacteria (6, 15, 22, 23, 35), L. plantarum CtsR autoregulates its own transcription together with that of the cotranscribed clpC. However, the lack of McsA and McsB homologues in the ctsR-clpC operon, as well as in the rest of the L. plantarum genome, suggests that, in this species, the regulation of CtsR activity might differ significantly from that in B. subtilis.
Sequence analysis of the L. plantarum WCFS1 genome allowed us to identify potential CtsR regulon members. The results of transcriptional analysis indicated that clp genes are under CtsR control, thus confirming the regulatory pattern observed for clp in most Gram-positive bacteria (6, 13, 15, 35) but in contrast to that in several closely related Lactobacillus species in which clp gene transcription has been suggested to be CtsR independent (31, 33). Besides clp, CtsR controls the expression of a shs gene in L. plantarum: thus far, only one other small HSP gene, Lo18 of O. oeni, had been included in a ctsR regulon (15). The occurrence of likely CtsR operators upstream from four apparently heat stress-unrelated genes of L. plantarum raises the possibility that CtsR might control other activities. However, further studies are needed and will be undertaken to shed light on this aspect.
Overall, our results confirm the high conservation of the CtsR regulatory system among Gram-positive bacteria but hint at possible differences in the regulation of CtsR activity and suggest that control by CtsR might extend to other major cellular functions.
P. Hols is a research associate at the FNRS. Work in the group of T. Msadek was supported by research funds from the European Commission (grant BACELL Health LSHG-CT-2004-503468), the Centre National de la Recherche Scientifique (CNRS URA 2172), and the Institut Pasteur (GPH 9). V. Capozzi was partially supported by a SfAM Research Development Fund grant award from the English Society of Applied Microbiology (SfAM).
Published ahead of print on 20 November 2009.
†Supplemental material for this article may be found at http://jb.asm.org/.