Edited by A.M. George
Role for rpoS gene of Pseudomonas aeruginosa in antibiotic tolerance
Article first published online: 9 JAN 2006
FEMS Microbiology Letters
Volume 242, Issue 1, pages 161–167, January 2005
How to Cite
Murakami, K., Ono, T., Viducic, D., Kayama, S., Mori, M., Hirota, K., Nemoto, K. and Miyake, Y. (2005), Role for rpoS gene of Pseudomonas aeruginosa in antibiotic tolerance. FEMS Microbiology Letters, 242: 161–167. doi: 10.1016/j.femsle.2004.11.005
- Issue published online: 9 JAN 2006
- Article first published online: 9 JAN 2006
- Received 11 August 2004, Revised 29 September 2004, Accepted 1 November 2004
- Pseudomonas aeruginosa;
- rpoS gene
The alternative sigma factor, RpoS has been described as a central regulator of many stationary phase-inducible genes and a master stress-response regulator under various stress conditions. We constructed an rpoS mutant in Pseudomonas aeruginosa and investigated the role of rpoS gene in antibiotic tolerance. The survival of the rpoS mutant cells in stationary phase was ∼70 times lower when compared with that of the parental strain at 37 °C for 2 h after the addition of biapenem. For imipenem, the survival was ∼40 times lower. Heat stress promoted an increase in the survival of the parental strain to biapenem, but the same was not found to be the case for the rpoS mutant. Our results indicate that rpoS gene is involved in tolerance to antibiotics in P. aeruginosa during the stationary phase and heat stress. However, under osmotic stress, tolerance to biapenem was not dependent on the rpoS gene.
It is well known that bacterial eradication frequently fails after antibacterial chemotherapy, despite the fact that the antibiotics used are active against the causative organisms in vitro . Thus, the antibiotic activity of antibiotics in vitro does not always reflect the same activity in vivo. In bacteria, significant physiological changes occur depending on various environmental conditions, such as heat shock, nutrient starvation, the presence of hydrogen peroxide, high osmolarity, and growth phase [2,3]. These environmental conditions can trigger the induction of stress-response genes.
It was reported previously that stationary-phase cells exhibit higher tolerance to antibiotics than do logarithmic phase ones [4,5], one of the reasons for this being their slow growth rate . It has been suggested that stationary-phase cells express different stress-response genes, which allow them to survive severe environmental conditions. The antibiotic tolerance of stationary-phase cells might be affected by the genes which are under control of RpoS.
The alternative sigma factor, RpoS positively regulates many genes in stationary phase [7,8] and is considered to be a master stress-response regulator. Genes involved in heat and osmotic induction are also known to be controlled by RpoS [9,10]. The rpoS gene of P. aeruginosa has been cloned and sequenced . Like that in E. coli, the RpoS level in P. aeruginosa increases upon entry of the cells into the stationary phase. RpoS affects the production of extracellular alginate, exotoxin A, and biofilm formation [12,13]. Furthermore, RpoS affects the expression of more than 40% of all quorum-controlled genes identified by transcriptome analysis .
Antibiotic tolerance is defined as the ability of bacteria to survive but not grow in the presence of antibiotics above their MIC [15–17]. We previously reported that adherent bacteria on solid surfaces were already tolerant to antibiotics , and we later isolated a biapenem-tolerant mutant of P. aeruginosa by transposon insertion . However, little is known about the relationship between stress response and antibiotic tolerance. The aim of this study was to investigate the effect of several types of stress conditions on antibiotic tolerance of Pseudomonas aeruginosa, an important opportunistic pathogen.
2Materials and methods
2.1Bacterial strains and growth conditions
The bacterial strains and plasmids used in this study are shown in Table 1. P. aeruginosa SM21 is a streptomycin-resistant strain derived from PAO1. LB broth was used for strain construction with the following supplements as required: ampicillin, 50 μg ml−1; gentamicin, 15 μg ml−1 (for Escherichia coli) or 200 μg ml−1 (for P. aeruginosa); streptomycin 1000 μg ml−1 (for P. aeruginosa); sucrose 5% (for P. aeruginosa).
|Strain or plasmid||Relevant characteristic||Source or reference|
|SM21||Smr derivative of PAO1||This study|
|KMS1||SM21 rpoS::Gmr||This study|
|S17-1||thi endA recA hsdR with RP4-2Tc::Mu-Km::Tn7 integrated in chromosome|||
|KMS3||S17-1 derivative harboring pKMS3||This study|
|pACΩGm||pACYC184 derivative carrying Ω fragment, Gmr|||
|pMOB3||Kmr Cmr 5% sucrose sensitive|||
|pGEM-T||Apr and TA cloning vector||Promega|
|pKMS1||rpoS in pGEM-T||This study|
|pKMS2||rpoS::Gmr in pGEM-T||This study|
|pKMS3||rpoS::Gmr and mob in pGEM-T||This study|
2.2Construction of the rpoS mutant
For DNA manipulations, such as those involving restriction digestion, ligation, transformation, and agarose gel electrophoresis, standard procedures  were followed.
In order to obtain an rpoS gene knockout mutant, we amplified the rpoS gene by PCR using primers rpoS-1s (5′-TAAGCCTGTCGATCCACTGCAATA-3′) and rpoS-1a (5′-TAAATTCACCGAGCGTTTTTGGCA-3′) from P. aeruginosa PAO1 DNA used as a template. The plasmid pKMS1 was constructed by insertion of the rpoS gene into the pGEM-T vector (Promega). Then the gentamicin-resistance gene from the Ω fragment of pACΩGm  was inserted into the BamHI site of pKMS1 to yield pKMS2. Plasmid pKMS3 was constructed by insertion of the NotI fragment of the MOB cassette of pMOB3  into the NotI-digested plasmid pKMS2. The resulting plasmid with the MOB cassette and insertion of Gmr cassette in the rpoS gene was used to transform S17-1, which served as a mobilizer strain . Conjugational transfer was achieved by the method of Simon et al. . After conjugational transfer of pKMS3 into SM21, we selected the gentamycin-resistant strain KMS1, which contains the rpoS::aac1 Gmr insertion in place of rpoS gene. The insertion of aac1 was confirmed by PCR using primers rpoS-2s (5′-CATCCAGGTCGGTCAAGCTATCCA-3′) and rpoS-2a (5′-GGAAGTCTGGCCGAACATCACCGA-3′). DNA sequencing of the PCR product was performed with the BigDye Terminator Cycle Sequencing Ready Reaction Kit and ABI PRISM 3100 (Applied Biosystems, Inc.).
The MICs (minimal inhibitory concentrations) and MBCs (minimal bactericidal concentrations) were determined by the microbroth dilution method, as previously described .
2.4Concentration-dependent killing study
Liquid cultures were incubated for 12 h at 37 °C with aeration until the stationary phase was reached. Cells were harvested by centrifugation and resuspended in fresh LB broth at the original concentration (∼109 cells ml−1). They were then incubated with a given antibiotic of various concentrations for 2 h. Viable cell number was determined before and after drug addition by subculturing the cells on LB agar plates for 24 h. Colony-forming units were expressed as the mean of duplicate plates. The killing curve experiments were performed at least three times, and the data in this paper are representative results obtained from one of these experiments.
2.5Time-dependent killing study
For the time-dependent killing study, the same culture conditions were applied, as specified above. Logarithmic-phase cultures were obtained by inoculating the cells into fresh medium and incubating them for 2.5 h at 37 °C with aeration. The cells were then sedimented and resuspended in fresh LB broth (∼108 cells ml−1). Both stationary-phase and logarithmic-phase cells were incubated with a given antibiotic for various periods of time (0–120 min). Viable cell number was determined before and after drug addition by subculturing the cells on LB agar plates for 24 h.
2.6Time-dependent killing study under heat or osmotic stress
Logarithmic-phase cultures were prepared by incubating cells at 30 °C with aeration. The cells were then sedimented and resuspended in fresh LB broth at the original concentration (∼108 cells ml−1) and incubated at 30 or 43 °C for 15 min. Following the addition of biapenem (32 μg ml−1), the cells were further incubated at either of the temperatures mentioned above for various periods of time (0–30 min).
Other logarithmic-phase cells incubated at 37 °C were sedimented and resuspended in fresh LB broth at the original concentration with or without 0.8 M NaCl, and incubated at 37 °C for 15 min, after which 32 μg ml−1 of biapenem was added. Viable cell number was determined by counting viable cells before and after drug addition.
The MICs and the MBCs of biapenem, imipenem, meropenem, panipenem, piperacillin, cefepime, ofloxacin, and ciprofloxacin were determined. The MIC and the MBC of the rpoS mutant and the parent were found to be almost the same (Table 2).
|Antibiotics||Parental strain||rpoS mutant|
|MIC (μg ml−1)||MBC (μg ml−1)||MIC (μg ml−1)||MBC (μg ml−1)|
3.2Killing rate by biapenem and imipenem
Fig. 1(a) shows the susceptibility of the rpoS mutant and the parental cells in stationary phase to different concentrations of biapenem after incubation for 2 h with the drug. At 0.5 μg ml−1 of biapenem, there was slight difference between the rpoS mutant and the parental strain; but at 4 μg ml−1, the survival of the rpoS mutant was 10 times lower than that of the parent. At 32, 64, and 128 μg ml−1, the survival of the rpoS mutant was over 100 times lower than that of the parent.
As a result of the dose-dependent killing study, the susceptibility of the rpoS mutant and the parent was determined by conducting time-dependent killing in the presence of 32 μg ml−1 of biapenem. Fig. 2 shows the time-dependent killing study of logarithmic-phase (A) and stationary-phase (B) cells. Logarithmic-phase cells were killed more rapidly than stationary phase ones; and the rpoS mutant cells were killed more rapidly than the parental cells, especially in the case of the logarithmic phase ones. In the stationary phase, the survival of the rpoS mutant was ∼70 times less than that of the parental strain at 2 h after the addition of biapenem. It was apparent that the effect of killing was due to antibiotics, not to natural dying (data not shown).
The killing curves of logarithmic-phase cells (A) and stationary-phase cells (B) with 64 μg ml−1 of imipenem are shown in Fig. 3. As was the case for biapenem, logarithmic-phase parental cells were killed more rapidly by imipenem than the stationary-phase parental cells. However, the stationary-phase cells of the rpoS mutant were killed as rapidly as the logarithmic-phase ones.
3.3Killing curves for ofloxacin
The results of the concentration-dependent killing study on the stationary-phase cells in the present of ofloxacin for 2 h are shown in Fig. 1(b). With 8 μg ml−1 of ofloxacin, the survival of the rpoS mutant was 10 times lower than that of the parental strain.
Fig. 4 shows the killing curves for logarithmic-phase cells (A) and stationary-phase cells (B) after incubation with 8 μg ml−1 of ofloxacin for various periods of times. Parental cells in either phase were killed rapidly by ofloxacin. The rpoS mutant cells in the logarithmic phase were killed more rapidly than the parent cell up to 20 min, although no apparent difference was observed later. The stationary-phase rpoS mutant cells were killed more rapidly than the parental-strain ones regardless of length of exposure time. Up to this point, these results suggest that the rpoS gene affects antibiotic tolerance with respect to the growth phase.
3.4Effect of heat and osmotic stress
In addition, we also evaluated the effect of heat and osmotic stress.
The survival rate of the parental strain in logarithmic phase was ∼5 times higher at 43 °C than that seen at 30 °C after 30 min in the presence of 32 μg ml−1 of biapenem. In contrast, there was no significant difference in the survival rate of the rpoS mutant between 30 and 43 °C (Fig. 5(a)).
When exposed to osmotic stress by the addition of 0.8 M NaCl, both the parental strain and the rpoS mutant demonstrated 10 times higher survival (Fig. 5(b)).
RpoS positively regulates many genes in stationary phase [7,8]. Moreover, even in logarithmic phase cells, RpoS is also involved in stress adaptation. In E. coli, there are about 100 genes under the control of RpoS . In P. aeruginosa, the results of a recent microarray analysis indicate that RpoS regulates 722 genes in stationary phase but not in logarithmic phase . The synthesis of RpoS is regulated at transcriptional, translational, and post translational levels by many regulatory factors, such as guanosine 3′,5′-bispyrophophate (ppGpp), H-NS, and ClpP .
The results presented here lead us to propose a role for the rpoS gene in the tolerance of stationary-phase cells to antibiotics. In the parental strain, the tolerance of stationary-phase cells to antibiotics was observed in the presence of either biapenem or imipenem (Figs. 2 and 3). However, the parental stationary-phase cells were not tolerant to ofloxacin, which had a strong bactericidal effect on nongrowing cells (Fig. 4). In logarithmic-phase cells, the rpoS gene effected the tolerance to biapenem and imipenem, but not that to ofloxacin. In stationary-phase cells, the survival of the rpoS mutant was lower than that of the parental strain in the presence of ofloxacin, as well as in that of biapenem or imipenem. Our findings thus suggest that the rpoS gene is involved in the tolerance of stationary-phase cells to these antibiotics (Figs. 2–4).
The potent antimicrobial activity of the carbapenems depends on efficient binding to the lethal penicillin-binding proteins (PBPs) targets. The carbapenems make the strongest binding to PBP 2 of P. aeruginosa. In addition, the affinity with PBP 1b of biapenem was higher than that of imipenem . It is probable that the tolerance in the parental strain was not attributable to the binding affinity with PBP(s), because there were little differences in the survival ratio of the rpoS mutant between biapenem and imipenem. The efficiency of rpoS mRNA translation was increased by ppGpp in E. coli. In E. coli, The hipA mutant, which exhibit a reduction in the rate of killing by ampicillin, increases the basal level of (p)ppGpp synthesis, allowing a significantly greater percentage of cells in a population to enter a persistent state . It is likely that the tolerance to carbapenems is related to rpoS.
The target of ofloxacin is DNA gyrase which is unique in catalyzing the negative supercoiling of DNA and is essential for efficient DNA replication, transcription, and recombination . It is likely that promoters of environmental stress sensitive genes, regulated by rpoS, were affected by DNA gyrase . It has been shown that the selectivity for stationary phase-specific promoters by rpoS increased concomitantly with the decrease in DNA superhelicity . It is probable that the tolerance to ofloxacin results in conformational alteration of DNA superhelicity in rpoS specific promoter region through ofloxacin binding.
It was demonstrated by transcription profile of E. coli treated with ampicillin or ofloxacin, that there was a significant overlap in genes whose transcription was affected. Twenty-two genes were induced and 139 genes were transcriptionally repressed by both antibiotics . It is likely that there is an overlap in the pathways of killing by β-lactams and quinolones. It is possible that there are common genes which effect the process in antibiotic activity and are regulated by ppGpp and rpoS in stationary phase.
It was earlier reported that an rpoS mutant of P. aeruginosa showed an increase in the killing rate by exposure to heat, low pH, high osmolarity, hydrogen peroxide or ethanol [12,32]. Cochran et al.  found that RpoS and AlgT played a transient role in protecting thin biofilms from hydrogen peroxide but not from monochloramine.
In E. coli, RpoS increased in response to a temperature shift from 30 to 42 °C . In our experiment, the rpoS mutant exhibited lower tolerance under heat stress (Fig. 4(a)). In E. coli, the osmotic induction of rpoS can be triggered by the addition of 0.3 M NaCl, and stimulation of rpoS translation and a change in the half-life of RpoS from 3 to 50 min both contribute to this osmotic induction . However, the results obtained in this study suggest that osmotic stress increased antibiotic tolerance but that the antibiotic tolerance was independent of the rpoS gene. The role of the rpoS gene in P. aeruginosa might thus be slightly different from that in E. coli. Twenty-four putative sigma factors have been detected in the genome of P. aeruginosa; however, in E. coli, only 7 sigma factors were found .
We previously isolated an biapenem-tolerant mutant of P. aeruginosa by Tn 1737 KH insertion. The survival of the mutant 3 h after the addition of biapenem to culture of the cells in stationary phase was about 1000 times higher than that of the parental strain. After biapenem addition, survival of adherent cells on a plastic surface was higher than that of planktonic cells .
Our results indicate that the rpoS gene is a participant, but not the only one, in the tolerance of stationary-phase cells to antibiotics and of logarithmic-phase cells to antibiotics triggered by heat stress. Under osmotic stress, tolerance to biapenem was not dependent on the rpoS gene, therefore suggesting that some other stress response gene(s) might be involved in this tolerance.
We gratefully thank Dr. Herbert P. Schweizer for providing plasmids pMOB3 and pACΩGm.
- Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning; A Laboratory Manual, second ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
- Levine, C., Hiasa, H., and Marians, K.L. (1998) DNA gyrase and topoisomerase. IV: biochemical activities, physiological roles during chromosome replication, and drug sensitivities, Biochem. Biophys. Acta 1400, 29–43