SEARCH

SEARCH BY CITATION

Keywords:

  • antibiotic tolerance;
  • persisters;
  • ofloxacin;
  • stationary phase;
  • biofilm

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Persister cells are phenotypic variants that are extremely tolerant to high concentrations of antibiotics. They constitute a fraction of stationary phase cultures and biofilm populations of numerous bacterial species, such as the opportunistic pathogen Pseudomonas aeruginosa. Even though persisters are believed to be an important cause of incomplete elimination of infectious populations by antibiotics, their nature remains obscure. Most studies on persistence have focused on the model organism Escherichia coli and only a limited number of persistence genes have been identified to date. We performed the first large-scale screening of a P. aeruginosa PA14 mutant library to identify novel genes involved in persistence. A total of 5000 mutants were screened in a high-throughput manner and nine new persistence mutants were identified. Four mutants (with insertions in dinG, spuC, PA14_17880 and PA14_66140) exhibited a low persister phenotype and five mutants (in algR, pilH, ycgM, pheA and PA14_13680) displayed high persistence. These genes may serve as new candidate drug targets in the combat against P. aeruginosa infections.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Pseudomonas aeruginosa is an opportunistic pathogen that causes severe infections in immunocompromised patients. Especially in individuals with cystic fibrosis, these infections often switch to a chronic nature, seriously increasing the case fatality rate (Lyczak et al., 2002). Apart from its intrinsic tolerance against a broad spectrum of antibiotics, emerging resistance and biofilm formation of this bacterium are becoming increasingly problematic in health care units (Costerton et al., 1999; Hota et al., 2009). The presence of specialized persister cells in biofilms makes it virtually impossible to completely eradicate the bacterial population. These persister cells represent a small fraction of phenotypic variants within the population that remain viable even after prolonged treatment with high doses of antibiotics. Because they revert back to a normal state after reinoculation in antibiotic-free medium and the overall antibiotic susceptibility of their offspring remains the same as that of the original population, these cells are not considered to be mutants. Instead, they exhibit a phenotypic tolerance to the antibiotic used (Lewis, 2008).

Although persister cells were discovered as early as 1944 by Bigger (1944) when analyzing the penicillin mode of action, they were ignored for a long time. The first mutants identified to be altered in their persistence were the high persistence (hip) mutants in Escherichia coli, discovered in 1983 (Moyed & Bertrand, 1983). From this point on, interest in persisters grew, and gradually, additional genes were discovered to be involved in persistence (for reviews on persistence, see Levin & Rozen, 2006; Lewis, 2008). Most research has focussed on E. coli persistence, but the phenomenon has been observed in numerous bacterial species, such as P. aeruginosa (Spoering & Lewis, 2001; Keren et al., 2004a; Harrison et al., 2005), Staphylococcus aureus (Woolfrey et al., 1986), Streptococcus pneumoniae (Novak et al., 1999) and even in the yeast Candida albicans (Jabra-Rizk et al., 2004; LaFleur et al., 2006). Although efforts were undertaken for large-scale screenings in the search for persistence genes, they have so far been performed only in E. coli and have yielded relatively few results (Hu & Coates, 2005; Spoering et al., 2006; Hansen et al., 2008).

Persister cells typically represent a small fraction of slowly or nondividing cells that are less vulnerable to antibiotics than the bulk of the population. However, the mechanism underlying this phenomenon is still largely unknown. Over the years, several hypotheses have been postulated. One possible explanation lies in the dormant nature of persisters. Most antibiotics kill bacteria by corrupting fundamental cellular processes such as translation, transcription or cell wall synthesis (Lewis, 2008). By downregulating their metabolism, the persister cells might ‘escape’ this corruption, rendering them insusceptible to antimicrobial agents targeting functions vital to cell growth. It has been established that overproduction of toxins known to block cell metabolism causes the persister fraction to increase in the population (Vazquez-Laslop et al., 2006). Stochasticity in the expression of these toxins, resulting from fluctuations in transcription or translation, could then cause only a few cells to become persistent, leaving the bulk of the population in a fast-growing state.

Even though P. aeruginosa persistence is of great clinical relevance, the organism has been the subject of few molecular studies on this matter. Currently, the only persistence genes that have been discovered in P. aeruginosa are the global regulators spoT, relA and dksA (Viducic et al., 2006) and rpoS (Murakami et al., 2005). Here, we describe the first large-scale screening of a P. aeruginosa mutant library to identify genes that contribute to the persistence phenomenon in this pathogen. A screening procedure was designed to select, in a high-throughput manner, mutants exhibiting a significant difference in persister fraction after prolonged treatment with the fluoroquinolone antibiotic ofloxacin. Several mutants with either an increased or a decreased number of surviving persister cells were identified and subjected to sequence analysis.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Bacterial strains, media, growth conditions and plasmids

Pseudomonas aeruginosa strain PA14 was used for the library construction. For comparison, specific PAO1 mutants were ordered from the Manoil library (Jacobs et al., 2003). Both P. aeruginosa PA14 and PAO1 were cultured in trypticase soy broth (TSB) or solidified medium (1.5% agar) at 37 °C. The following antibiotics were used: gentamicin (45 μg mL−1), ofloxacin (5 μg mL−1) and tetracycline (200 μg mL−1). Escherichia coli strains were grown in Luria–Bertani (LB) broth or solidified medium (1.5% agar) at 37 °C. The following antibiotics were used: gentamicin (30 μg mL−1) and kanamycin (50 μg mL−1).

Construction of a PA14 mutant library

A mutant library of P. aeruginosa PA14 was constructed by random insertional mutagenesis using a pTnMod-OGm plasposon (Dennis & Zylstra, 1998). The plasposon was transferred to PA14 by conjugation as described by D'Hooghe et al. (1995) with the following adjustments: biparental mating was performed by mixing equal volumes of an overnight culture of the E. coli donor strain carrying the pTnMod-OGm and an exponential phase culture of the acceptor strain PA14. The mating mixture was centrifuged, resuspended in 10 mM MgSO4, plated on LB plates without selection and incubated for overnight growth at 30 °C. Conjugants were recovered from the plate by suspending them in 1 mL TSB, and plated on TSB plates with gentamicin and kanamycin. After overnight growth at 37 °C, the colonies were picked up and suspended in separate wells of a microtiter plate in 100 μL TSB, with gentamicin and kanamycin. After overnight growth at 37 °C, 100 μL glycerol 50% was added, and the microtiter plates were sealed and kept at −80 °C. Five thousand mutants were constructed in total. The presence of the plasposon was confirmed for a random selection of 40 mutants.

Persistence assay

The mutants were inoculated in microtiter plates with 200 μL TSB without antibiotics, sealed with a permeable membrane and grown overnight at 37 °C and shaking at 200 r.p.m. After overnight growth, 90 μL of the culture was transferred to a new microtiter plate to which 10 μL of an ofloxacin stock solution was added to achieve a final concentration of 5 μg mL−1. As a control, 90 μL of the overnight culture was treated in a separate microtiter plate with sterile water. For the confirmation of the persistence phenotype after selection in the screening, the PA14 and PAO1 mutants were grown to stationary phase in test tubes with 5 mL TSB at 37 °C shaking at 200 r.p.m. The treatment was performed in test tubes with 1 mL culture and 10 μL of an ofloxacin stock solution to achieve a final concentration of 5 μg mL−1. The control treatment with water was also performed in a test tube with 1 mL culture and 10 μL water. All the ofloxacin and control treatments were performed at 37 °C, shaking at 200 r.p.m., during 5 h.

Selection of persistence mutants

After ofloxacin treatment, the cultures were diluted 100-fold and incubated in an automated OD plate reader (Bioscreen C, Oy Growth Curves Ab Ltd). The linear relationship between the number of cells incubated and the lag phase of the growth curve was confirmed in the context of this research (see Supporting Information, Fig. S2). Because the plate reader can measure 200 samples simultaneously, both the treated cultures and the controls could be measured in parallel.

Based on the lag phase of the growth curves generated by the plate reader, mutants with an altered number of persister cells were selected. This selection comprises the 5% most extreme values of lag phases when analysing all the growth curves of the complete mutant library. The normal distribution of the number of cells incubated in the plate reader was checked and confirmed for each screened microtiter plate of the library. The selected mutants were subjected twice to independent repeats of the experiment to confirm their phenotype.

Minimal inhibitory concentration (MIC50) determination

The MIC50 was defined as the minimal antibiotic concentration needed to inhibit growth by 50% and was determined with a standard macrodilution procedure. Ofloxacin concentrations ranged from 10 to 0.02 μg mL−1 in a twofold dilution series and the test was carried out in TSB at 37 °C, shaking at 200 r.p.m.

Identification of the plasposon insertion location

To determine the genomic region of the plasposon insertion, the general procedure of subcloning and subsequent sequencing was carried out as described by Moris et al. (2005). In some cases, subcloning was ineffective and genomic sequencing was performed instead.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Screening of a PA14 mutant library for mutants with altered persistence

A library of 5000 P. aeruginosa PA14 mutants was constructed by random insertion of the pTnMod-OGm plasposon (Dennis & Zylstra, 1998) and screened for persistence mutants. To select mutants affected in genes that play a role in persistence, the selection procedure was based on differential survival after prolonged treatment with ofloxacin compared with the wild type. The mutants from the library were grown overnight to stationary phase and challenged with ofloxacin at a concentration of 5 μg mL−1 (10-fold the MIC50) for 5 h. The fluoroquinolone ofloxacin was chosen because the persistence assay is performed on nongrowing stationary phase cells and the antibiotic is lethal to nondividing cells. After treatment, the cells were diluted in growth medium and further incubated in an automated OD plate reader to generate growth curves of the surviving cells. The selection for persistence mutants was based on a significant difference in lag phase compared with the wild-type cells. Mutants with a lower or a higher number of persister cells will display a longer or a shorter lag phase, respectively, when they are recultured after ofloxacin treatment compared with the wild type (see Fig. S1). This procedure enabled us to perform the screening in a high-throughput manner.

After an initial round of screening, 126 mutants were selected that were significantly altered in their persister fraction compared with the wild type. The phenotype of these mutants was confirmed by independently repeating the persistence assay using plate counts. MIC50 values were determined and no resistant mutants were found among the selected mutants. Finally, nine mutants consistently displaying an altered persister fraction were retained. The relative persister fraction was determined for each mutant in at least four independent experiments, the mean values and 25th and 75th percentiles of which are displayed in Fig. 1. Four mutants showed a decreased persister fraction, ranging from a 2.8- to a 16-fold reduction compared with the wild-type strain. Five mutants exhibited a higher number of persister cells after treatment, between 2.6- and 18.4-fold the number of persister cells compared with the wild type.

image

Figure 1.  Mean relative persister fraction of the nine selected Pseudomonas aeruginosa mutants. The persister fraction is defined as the number of surviving cells after treatment with ofloxacin, divided by the number of cells of the control condition. The relative persister fraction for each mutant is the persister fraction of the mutant divided by that of the wild type (a value of 1 in the y-axis). Each data point in the graph is calculated as the inverse logarithm of the mean of the logarithmic values of these relative persister fractions of separate experiments (each experiment was independently repeated at least four times). Each number in the x-axis represents the corresponding CMPG mutant strain.

Download figure to PowerPoint

Identification of the selected mutants

Sequence analysis revealed the plasposon insertion site in the selected mutants to be located in genes of diverse functional classes, such as enzymes and regulators involved in various cellular processes (Table 1).

Table 1.   Mutants with an altered persistence phenotype and their plasposon insertion location
Mutant IDPA14 locus*GenePredicted function of gene productRelative persister fraction
  • *
  • The persister fraction is defined as the number of surviving cells after treatment with ofloxacin, divided by the number of cells of the control condition. The relative persister fraction for each mutant is the persister fraction of the mutant divided by that of the wild type. Each value in the table is calculated as the inverse logarithm of the mean of the logarithmic values of these relative persister fractions of separate experiments (each experiment was independently repeated at least four times).

Decreased persistence
 CMPG13400PA14_50840dinGPutative DNA-helicase0.06
 CMPG13401PA14_66140 Conserved hypothetical protein0.13
 CMPG13402PA14_03900spuCPutrescin aminotransferase0.23
 CMPG13403PA14_17880 Acetyl-CoA acetyltransferase0.36
Increased persistence
 CMPG13404PA14_13680 Putative short-chain dehydrogenase2.60
 CMPG13405PA14_69470algRAlginate biosynthesis regulatory protein3.32
 CMPG13406PA14_04150ycgMPutative fumarylacylacetoacetate hydrolase family protein9.05
 CMPG13407PA14_05330pilHType IV pilus response regulator17.16
 CMPG13408PA14_23280pheAFused chorismate mutase–prephenate dehydratase18.39

CMPG13400 has a persister fraction 16 times lower than that of the wild type. The plasposon is inserted in a putative Rad3-related DNA-helicase, similar to E. coli dinG (Voloshin et al., 2003). In E. coli, DinG is involved in stress-induced DNA-repair and P. aeruginosa PAO1 dinG is one of the genes activated upon ciprofloxacin treatment in a LexA-dependent manner (Cirz et al., 2006). Debbia et al., (2001) suggested that persister cells achieve a dormant state as a result of repairing spontaneous errors of DNA synthesis, a mechanism that is known to block cell division. They showed that the persistence of an E. coli lexA knock-out mutant (defective in DNA repair mechanisms) against several antibiotics was significantly decreased. This corroborates the phenotype of CMPG13400 as plasposon insertion in dinG indeed lowers the persister fraction. Analysis of the expression profile of isolated persisters (Keren et al., 2004b) also revealed elevated expression of genes involved in the SOS response.

In CMPG13402, the persister fraction is lowered 4.4-fold compared with that of the wild type. The plasposon is inserted in spuC, encoding a putrescine aminotransferase. It is involved in putrescine and spermidine utilization and probably regulates the ABC transporter SpuDEFGH that is used for spermidine uptake. A spuC mutant is unable to grow in minimal growth medium with putrescine as the sole carbon and nitrogen source, illustrating its essential role in the uptake of putrescine (Lu et al., 2002). Both putrescine and spermidine are natural polyamines that play a regulatory role in many cellular processes, such as protecting the cell from external toxic conditions. Exogenous addition of these polyamines decreases the antibiotic susceptibility against quinolones in P. aeruginosa PAO1 (Kwon & Lu, 2006), corroborating the low persistence phenotype of our mutant CMPG13402.

CMPG13403, displaying a persister fraction 2.8 times lower than that of the wild type, has an insertion in a gene encoding a PaaJ-like acetyl-CoA acetyltransferase involved in fatty acid and phospholipid metabolism. It is currently not understood how this acyltransferase contributes to persistence.

In CMPG13405, the plasposon is inserted in the response regulator gene algR, resulting in a higher persister fraction than the wild type. AlgR is the response regulator of a two-component regulatory system. Its function has been extensively studied in P. aeruginosa and it seems to play a role in a variety of processes, such as twitching motility (and therefore biofilm formation), rhamnolipid production, virulence and conversion to a mucoid phenotype by overproduction of alginate (Gooderham & Hancock, 2009). This mucoid phenotype is an important clinical feature, because it is correlated with a shift to infections of a chronic nature (Qiu et al., 2008). Because of the broad regulatory spectrum of AlgR, it is difficult to speculate about which mechanism(s) might be involved in persistence.

In CMPG13408 (persister fraction 18.4 times higher than that of the wild type), the plasposon is located in pheA, a fused chorismate mutase–prephenate dehydratase involved in phenylalanine biosynthesis and metabolism (Fischer et al., 1991).

A second two-component response regulator, pilH, was selected from the mutant library. This mutant CMPG13407 also develops more persisters than the wild type. Like AlgR, PilH is a global regulator in P. aeruginosa and is involved in twitching motility by controlling the synthesis of type IV pili (Barken et al., 2008). Previous screenings for persistence mutants in E. coli also revealed the involvement of global regulators such as DksA, SsrS-YgfA, DnaKJ, HupAB and IhfAB (Hansen et al., 2008).

The screening also resulted in the identification of several mutants in hitherto uncharacterized genes. The cellular processes that are affected in these mutants are unknown. In CMPG13401 (exhibiting a low persistence phenotype), the plasposon is inserted in a gene encoding a hypothetical protein that is highly conserved among Pseudomonas species, but not in other bacteria. The function of this gene is not known, but it is located in a large operon with genes that encode proteins involved in cell envelope biogenesis (Walsh et al., 2000). CMPG13404 has 2.6 times more persister cells than the wild type and carries an insertion in a putative short-chain dehydrogenase. Finally, in CMPG13406, the plasposon is located in ycgM, coding for a putative fumarylacetoacetate hydrolase. This insertion causes the persister fraction to be nine times higher than that of the wild type.

Characterization of the persistence phenotype in corresponding PAO1 mutants

To confirm the observed phenotypes described above, we determined the persistence fraction of several corresponding PAO1 mutants (Jacobs et al., 2003). The PAO1 mutants were subjected to the same persistence assay as the PA14 mutant strains (see Materials and methods). The respective mean relative persister fractions of these PAO1 mutants are summarized in Table 2. The results support the proposed involvement of these genes in P. aeruginosa persistence, suggesting a conserved role among Pseudomonas spp. The persistence phenotype of the PA14 mutants CMPG13402, CMPG13405, CMPG13406 and CMPG13408 could not be confirmed for PAO1 mutants 52194 (PA0299), 51130 (PA5261), 1963 (PA0318) and 1037 (PA3166), respectively. This may result from structural differences of the transposons, differences in transposon insertion site or strain-specific effects.

Table 2.   Mean relative persister fraction of PAO1 mutant strains corresponding to PA14 persistence mutants identified in this study
PAO1 mutant ID*PAO1 geneMean relative persister fractionCorresponding PA14 mutant
  • *

    The PAO1 mutant strains were obtained from the PAO1 mutant library distributed by the Manoil Group from the University of Washington Genome Center (Jacobs et al., 2003).

  • The persister fraction is defined as the number of surviving cells after treatment with ofloxacin, divided by the number of cells of the control condition. The relative persister fraction for each mutant is the persister fraction of the mutant divided by that of the wild type. Each value in the table is calculated as the inverse logarithm of the mean of the logarithmic values of these relative persister fractions of separate experiments (each experiment was independently repeated at least four times). The numbers in parentheses indicate the values for the corresponding PA14 mutant.

19640PA10450.02 (0.06)CMPG13400
3892PA50020.02 (0.13)CMPG13401
42693PA35890.46 (0.36)CMPG13403
49849PA040910.41 (17.16)CMPG13407

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Despite the presence of persister cells in biofilm populations formed by some of the most health-threatening pathogens, such as P. aeruginosa and S. aureus, little is known about their nature. Several efforts have been undertaken to isolate and characterize persister cells, but because of their low abundance in bacterial populations, and the fact that they switch back to a normal state after reinoculation, analysis of these phenotypic variants has proved very challenging. The use of standard screening methods to identify persistence genes, such as screening of knock-out libraries (Hu & Coates, 2005; Hansen et al., 2008), has indeed not been straightforward (for a review, see Lewis, 2008). Alternative routes that have been taken are, for example, the study of overexpression libraries (Spoering et al., 2006), or the use of advanced microscopic techniques to study the persisters at a single-cell level (Balaban et al., 2004).

Notwithstanding the limited success of previous screenings, we performed a high-throughput screening of a newly constructed mutant library of P. aeruginosa PA14 in order to broaden the knowledge on persistence in this pathogen. Out of 5000 candidates, nine new persistence mutants were identified; four with a lower and five with a higher persister fraction compared with that of the wild type. These mutants were affected in genes belonging to diverse functional classes, including global regulators (CMPG13405 and CMPG13407), and enzymes involved in different cellular processes such as amino acid synthesis and metabolism (CMPG13408), DNA-repair (CMPG13400), nutrient uptake (CMPG13402) or phospholipid metabolism (CMPG13403). Further analysis of these newly discovered persistence genes may, in the future, lead to new candidate targets for improved drug development.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

V.N.D.G. and C.I.K. are recipients of a fellowship from the Institute for the Promotion of Innovation through Science and Technology in Flanders (IWT-Vlaanderen). N.V. received a fellowship from the Fund for Scientific Research – Flanders (FWO). The work was supported by a grant from the FWO (G.0299.07). The authors would like to thank Dr Michael Jacobs from the University of Washington Genome Center for kindly providing the PAO1 mutant strains (Jacobs et al., 2003). We also thank Prof. Maarten Jansen from the Department of Numerical Analysis and Applied Mathematics at the K.U. Leuven for his assistance in the statistical analysis of the experimental data.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results and discussion
  6. Conclusions
  7. Acknowledgements
  8. References
  9. Supporting Information

Fig. S1. Illustration of the mutant selection procedure. Growth curves of a randomly selected persistence mutant (triangles) and of PA14 wild type (circles) incubated in an automated optical density (OD) plate reader, both after control treatment (open symbols) and after treatment with ofloxacin (closed symbols). The cultures were grown to stationary phase in Trypticase Soy Broth (TSB) in microtiterplates at 37°C. After overnight growth, 90 μl of the culture was treated with 10 μl ofloxacin in a final concentration of 5 μg ml-1. A control treatment was performed in parallel with 90 μl culture and 10 μl sterile water. Both treatments were carried out at 37°C during 5 hours, after which the cultures were diluted 100 times in TSB and incubated in an automated OD plate reader for two days. 1T (calculated as 1T = 1t wild type (dashed line) - 1t mutant (straight line)) was used as an estimation for the difference in lag phase between ofloxacin-treated cultures of mutant and wild type and was used as a parameter for statistical selection of mutants.

Fig. S2. Linear relationship between Colony Forming Units and lag phase of growth curves PA14 wild type was grown overnight at 37°C in TSB, dilutions were made in TSB and were plated on TSB agar to perform colony counts after overnight growth at 37°C. Incubation in an automated OD plate reader was carried out in parallel. The time needed to achieve an OD of 0.6 (= t(OD = 0.6)) was selected as an indication of the lag phase. In the graph, the CFU (colony forming units) per ml are displayed in relation to the corresponding t(OD = 0.6). This relationship is a linear function, as the regression displayed in the graph indicates. This enabled us to use t(OD = 0.6) as an estimation of the number of cells incubated in the automated OD plate reader after ofloxacin and control treatment in the mutant library selection protocol.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

FilenameFormatSizeDescription
FML_1657_sm_suppinfoFigs1and2.pdf278KSupporting info item
FML_1657_sm_suppinfolegends.doc24KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.