JAC Advance Access originally published online on November 13, 2006
Journal of Antimicrobial Chemotherapy 2007 59(2):254-263; doi:10.1093/jac/dkl469
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Antibiotic treatment in vitro of phenotypically tolerant bacterial populations
Department of Medical Biochemistry and Microbiology, Uppsala University Box 582, S-751 23 Uppsala, Sweden
*Correspondence address. Health Protection Scotland, Section for HAI & IC, 1 Cadogan Square, Cadogan Street, Glasgow G2 7HF, UK. Tel: +44-141-2822927; Fax: +44-141-8470399; E-mail: camilla.wiuff{at}hps.scot.nhs.uk
Received 2 August 2006; returned 4 October 2006; revised 19 October 2006; accepted 21 October 2006
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
Objectives: Most pharmacodynamic models used for design of treatment regimens are based on time–kill data obtained with normal cells in the susceptible state without taking into account the killing kinetics of the antibiotic-tolerant cells in the population. We compared the microbiological efficacy of six antibiotics against tolerant cells and by mathematical modelling explored the potential clinical implications of tolerance.
Methods: Tolerant cells were obtained by filtration of bacterial cultures of Escherichia coli MG1655 after antibiotic exposure. Killing kinetics of the tolerant cells was compared with that of exponentially growing naive cells. To examine the nutrient dependency of the reversion from the tolerant state to the susceptible state, tolerant cells were re-suspended in Luria–Bertani and PBS and re-exposed to antibiotics. A mathematical model was used to explore the clinical implications of antibiotic tolerance.
Results: Streptomycin was the most efficient drug against tolerant cells. Ciprofloxacin and ampicillin had intermediate activity against tolerant cells while rifampicin, tetracycline and erythromycin had poor activity against tolerant cells. No correlation could be established between the microbiological efficacies against susceptible and tolerant cells. Reversion from tolerance to susceptibility was dependent on the presence of nutrients and growth. Computer simulations demonstrated that the efficacy of antibiotics against tolerant cell populations has a large influence on treatment outcome.
Conclusions: The in vitro killing kinetics of tolerant cells is antibiotic-dependent and different from that of cells in the susceptible state. This difference in efficacy could have an influence on treatment outcome and tolerance should therefore be studied further in vivo.
Keywords: antibiotic tolerance , persistence , treatment regimens , killing kinetics , pharmacodynamic models
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
When susceptible bacteria are exposed to bactericidal concentrations of an antibiotic most of the population is killed within 1–2 h. Subsequently the killing rate declines slowly to values approaching zero and typically some bacteria survive even in the presence of active antibiotic. These surviving cells, present in all bacterial populations, are phenotypic variants of the wild-type that are tolerant to the antibiotics, supposedly without any underlying genetic change.1 This phenomenon was observed for the first time by Bigger who failed to sterilize Staphylococcus pyogenes cultures with penicillin.2 Incomplete in vitro eradication of bacterial populations, possibly due to tolerance, has subsequently been observed for virtually all antibiotics in clinical use.3–6 Recently the mechanism of phenotypic tolerance (or persistence) has been studied in vitro and by mathematical modelling.7–11 However, the clinical relevance of phenotypic tolerance has not been investigated.
In a previous paper, we suggested a mathematical model of phenotypic tolerance based on experimental data obtained in vitro with strain Escherichia coli O18:K1:H7 and five different classes of antibiotics.11 The model assumes that bacterial cells exist in two physiological states, an antibiotic-susceptible state, S, and an antibiotic-tolerant state, T. At any cell division there is a probability, f, of generating a tolerant cell, which is independent of the state of the parental cell. Since the tolerant cells are more refractory to the bactericidal action of the antibiotic they are enriched during drug exposure. Numerical solutions to this mathematical model showed that the presence of small tolerant sub-populations could lead to treatment failure under some circumstances.
Current pharmacodynamic models do not take into account the phenomenon of phenotypic tolerance even though it could potentially be a significant factor in the clearance of infections. Furthermore, the killing of susceptible cells has been described in detail but these experimental and theoretical studies have largely ignored the presence of surviving tolerant cells.12 In order to design more realistic antibiotic treatment regimens that take account of the presence of tolerant cells, we have studied the killing of tolerant cells by exposing them to different classes of antibiotics to evaluate and compare the efficacy of those antibiotics against tolerant cells. Furthermore, we compared single drug treatment with the successive treatment with two different classes of antibiotics.
Recently Balaban et al.7 studied single cells in a microfluidic device and observed that tolerant cells have reduced growth rates relative to the majority of the bacterial population before and during exposure to antibiotics. Wild-type cells enter the tolerant state continuously during antibiotic exposure and revert to the susceptible state upon removal of the antibiotic and transfer into the fresh medium.7,8,11 Based on these observations it is assumed that reversion to the susceptible state occurs when the tolerant cells switch to fast (normal) growth. To further explore the reversion to the susceptible state and its possible effect on treatment outcome, we studied the reversion to susceptibility in nutritionally rich and poor media. In conclusion, this study aims to characterize the killing and persistence of tolerant cells and their reversion to the susceptible state in further detail in order to develop more effective antibiotic treatment regimens.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
Bacterial strain and media
A strain of E. coli MG1655 ATCC number 47076 (designated DA5438) was used for all experiments. In most experiments bacteria were grown in Luria–Bertani broth (LB) at 37°C with aeration at 200 rpm. In reversion experiments bacteria were incubated in PBS as well. Total cell densities (cfu/mL) were estimated by diluting 0.1 mL samples in 0.85% saline and plating on LB agar. The lower level of detection was 10 cfu/mL.
Antibiotics and susceptibility testing
Ampicillin sodium salt, streptomycin sulphate salt and tetracycline hydrochloride min 95% were from Sigma-Aldrich and ciprofloxacin, erythromycin and rifampicin were from Fluka, BioChemika. Etest was carried out according to the instructions of the manufacturer under aerobic conditions at 37°C and 16 h of incubation (AB Biodisk, Solna, Sweden).
Tolerance experiments
Overnight cultures were diluted 1:100 into 60 mL of pre-warmed LB medium and incubated for 1 h at 37°C reaching a density of 
2 x 107 cfu/mL. At this time antibiotics at bactericidal concentrations were added to the cultures, which were incubated for 3 h. Cells surviving 3 h of antibiotic exposure were defined as being in the tolerant state. Surviving cells were caught on filters by pressing volumes of 10 mL of culture through a 0.45 µm pore size membrane filter (HT-450 Tuffryn®, Pall Life Sciences) attached to a syringe. The membrane filters were washed in 0.85% saline before they were transferred to a flask containing fresh medium and antibiotics. The bacteria on the membranes were re-suspended in fresh medium by vigorous vortex mixing of the flasks. The surviving cells were then re-exposed either to increasing concentrations of the same antibiotic or to increasing concentrations of another antibiotic. cfu/mL values were determined during the re-exposure at 1 h intervals for 3 h. As controls exponentially growing cultures (adjusted to the densities of the surviving cells after 3 h initial exposure) were exposed to a low concentration of the antibiotic being tested against the tolerant cells.
Reversion experiments
- Growth during reversion. Overnight cultures were diluted 1:100 into pre-warmed LB medium and incubated for 1 h at 37°C. These cultures were exposed to bactericidal concentrations of ampicillin, ciprofloxacin or streptomycin. After 3 h of exposure surviving cells (in the tolerant state) were caught on filters, washed in saline and transferred into either PBS or LB medium and incubated for 3 h at 37°C. Growth of tolerant cells in these two media was followed by estimating the cfu/mL from samples obtained at 0, 0.5, 1, 2 and 3 h of incubation.
- Drug susceptibility during reversion. Tolerant cells were obtained as under (i). After transfer of the tolerant cells to either PBS or LB, antibiotics at bactericidal concentrations were added. Reversion to the susceptible state was evaluated by measuring the reduction in cfu/mL of the cultures after 1 h of exposure at 37°C.
Time–kill curves
Overnight cultures of MG1655 were diluted 1:10 000 in fresh pre-warmed LB medium and incubated for 2 h at 37°C without antibiotics. The cultures were then divided into 10 mL aliquots and transferred to smaller flasks. Antibiotics were added at concentrations corresponding to 5, 10 and 20 times the MIC. These cultures were incubated at 37°C, shaking at 200 rpm, and sampled to estimate cfu/mL at 0, 20, 40, 60, 90 and 120 min. Control cultures with no antibiotics were set up in parallel and sampled at the same times.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
Tolerance experiments
In order to study the killing kinetics of tolerant cells, we re-exposed cells that survived a 3 h exposure to a single antibiotic (Figure 1). Cells surviving 3 h exposure to an antibiotic are in the following referred to as cells tolerant to that antibiotic. Controls were exponentially growing cells with no prior history of antibiotic exposure. The controls were exposed to the lowest concentration of antibiotics used to treat the tolerant cells.
|
Cells tolerant to ciprofloxacin were killed at a rather slow rate when re-exposed to a wide range of ciprofloxacin concentrations. Increasing the concentration of ciprofloxacin up to 1 mg/L only had a modest effect on the killing rate and the total decline in the number of tolerant cells (Table 1). Cells tolerant to ampicillin were also refractory to a secondary exposure to ampicillin (Figure 1b). Increasing the concentration of ampicillin up to 256 mg/L in the second exposure resulted in a gradual increase in the killing rate and in the total decline of cells. Cells surviving an initial exposure to streptomycin were less refractory to a secondary exposure to streptomycin than tolerant cells enriched in the presence of ampicillin or ciprofloxacin (Figure 1c). Increasing the concentration of streptomycin gradually increased the killing rate and the total decline in tolerant cells substantially (Table 1).
|
The effect of switching drug to another class of antibiotic in the re-exposure was examined by exposing ampicillin- and ciprofloxacin-tolerant cells to five other antibiotics (Figure 2 and Table 2). As controls, naive cells were exposed to the two lowest concentrations of antibiotics used to treat tolerant cells (except for ciprofloxacin controls which only included one concentration). Cells tolerant to ciprofloxacin were killed by streptomycin in a concentration-dependent manner, reaching a maximum effect at 96 mg/L whereas ampicillin only reduced the density of ciprofloxacin-tolerant cells with maximum 1.4 log (Figure 2a and b, and Table 2). The ciprofloxacin-tolerant cells were refractory to secondary exposures at most concentrations of rifampicin, tetracycline and erythromycin (Figure 2c–e and Table 2).
|
|
Cells tolerant to ampicillin were killed by streptomycin at increasing killing rates with increasing drug concentrations (Figure 2g and Table 2) while the tolerant cells were refractory to increasing concentrations of ampicillin, rifampicin and tetracycline. Increasing the concentration of these antibiotics did not increase the killing rate against tolerant cells considerably. The effect of erythromycin against both tolerant and naive cells was bacteriostatic rather than bactericidal which complicated the analysis of these experiments.
Reversion experiments
To examine the reversion from the tolerant state to the susceptible state and evaluate its dependence on nutrients, we transferred ciprofloxacin- and ampicillin-tolerant cells to PBS, a medium lacking nutrients, and LB, a nutritionally rich medium, and studied their change in susceptibility with time. Cells surviving exposure to ciprofloxacin and ampicillin were collected on filters after 3 h of antibiotic exposure and transferred to PBS or LB. When the tolerant cells were transferred to LB, the growth of the tolerant cells resumed immediately and reached exponential growth after 1 h (Figure 3a). When the tolerant cells were transferred to PBS, only transient slow (or no) growth was observed. The ampicillin-tolerant cells re-suspended in LB were immediately killed by low concentrations of ampicillin and ciprofloxacin whereas the cells transferred to PBS were refractory to both drugs (Figure 3b). The same was observed for streptomycin-tolerant cells (not shown).
|
To investigate the duration of the tolerant state in rich and poor medium and the change in susceptibility, tolerant cells were transferred to fresh medium (either PBS or LB) and samples were taken out from these re-suspended cultures (at 1 h intervals) and re-exposed to antibiotics. Figure 4 shows the ratios of the density of surviving cells (after 1 h re-exposure) relative to the initial density of cells (a ratio of 1 indicates that all cells survived). Cells initially exposed to ciprofloxacin remained refractory to a secondary exposure of ciprofloxacin when re-suspended in PBS for at least up to 3 h (Figure 4a). Ciprofloxacin-tolerant cells re-suspended in LB became successively more susceptible to a secondary exposure of ciprofloxacin. Cells initially exposed to ampicillin gradually reverted to increasing susceptibility with time when re-suspended in both LB and PBS (Figure 4b). However, a relatively higher number of ampicillin-tolerant cells persisted in PBS than in LB in the 2 and 3 h re-exposures. Streptomycin-tolerant cells remained somewhat refractory to the secondary antibiotic exposures in PBS, while the susceptibility immediately increased upon re-suspension in LB.
|
MIC determinations
MIC of the MG1655 strain used for these studies was determined using Etests according to standard protocols. The strain was very susceptible to ciprofloxacin, ampicillin, streptomycin and tetracycline but less susceptible to rifampicin and erythromycin (Table 3). Streptomycin, which was the most efficient drug against tolerant cells, had an MIC that was higher than the MIC of ciprofloxacin and tetracycline. Thus, MIC does not appear to be a good predictor of the microbiological efficacy against tolerant cells.
|
Time–kill curves
In addition to determining the MICs, time–kill kinetics were determined with exponentially growing E. coli cultures to test whether there exists any correlation between the microbiological activity against exponentially growing cells and tolerant cells (Figure 5 and Table 3). To enable a direct comparison of each antibiotic, the cultures were exposed to 5, 10 and 20 times the MIC. At the highest concentration tested (20x MIC) streptomycin had the (numerically) highest killing rate of all the antibiotics but not the greatest total (log) decline in cfu/mL. Ciprofloxacin, which had an intermediate microbiological efficacy against tolerant cells, had a killing rate and a total decline in cfu/mL at 20x MIC that were larger than the values obtained for streptomycin. Ampicillin, which also had an intermediate efficacy against tolerant cells, had a modest killing rate at the highest concentration while the decline in cfu/mL was of the same magnitude as streptomycin. Both rifampicin and erythromycin, which had poor activity against tolerant cells, produced intermediate killing rates and intermediate declines in cfu/mL against exponentially growing cells. Tetracycline had poor activity against both tolerant and exponentially growing cells. Thus, neither the killing rate nor the total decline in bacterial density for exponentially growing cultures correlates well with the parameters obtained for the tolerant cells.
|
Computer simulations of fluoroquinolone treatment illustrate the importance of the killing rates against tolerant cells
The in vitro exposures of tolerant cells indicated that survival of tolerant cells in the presence of relatively high antibiotic concentrations potentially could cause incomplete clearance of infections or, in the worst case, treatment failure. To examine the effect of killing rate against tolerant cells, we ran computer simulations employing a heuristic model with arbitrary units of the stepwise evolution of fluoroquinolone resistance. The model is not intended to directly predict treatment outcome in a clinical situation but rather to illustrate the potential impact of phenotypic tolerance on resistance development. Since the model has no immune system component it is mimicking a worst-case scenario. The model is an extension of a previously published model11 and details of the model are given in the Appendix [available as Supplementary data at JAC Online (http://jac.oxfordjournals.org/)].
We simulated a 5 day treatment regimen with dosing every 12 h with a fixed dose. Between the doses the antibiotic concentration was assumed to decline exponentially. First-step mutants with reduced drug susceptibility and second-step drug-resistant mutants were generated randomly using a Monte Carlo procedure. The graphs in Figure 6 depict the change in the total bacterial density (including the appearance of resistant mutants) during a treatment period at three different killing rates against tolerant cells. Cells in the normal state and the tolerant state are depicted as separate lines in the graphs.
|
At a killing rate of 1.0 against tolerant cells the total bacterial populations increased in density during the treatment period and the infection was not cleared (Figure 6a). At an intermediate killing rate of 1.5 the total bacterial populations were eradicated after 72 h of treatment (Figure 6b). At an elevated tolerant cell-killing rate of 2 the total bacterial populations were eradicated after only 24 h of treatment (Figure 6c). In repetitive runs of these simulations drug resistance evolved in 9 of 50 runs at a tolerant cell-killing rate of 1.0, while no resistance evolved in 50 runs at tolerant cell-killing rates of 1.5 or 2.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
Antibiotic resistance caused by stable genetic changes has been studied extensively since the discovery of antibiotics whereas resistance caused by phenotypic changes in bacteria has received little attention.13,14 Current pharmacodynamic models used for the design of antibiotic treatment regimens do not take into account the presence of tolerant cells, partly due to the lack of data on treatment of tolerant cells. Recently a renewed interest in phenotypically based antibiotic resistance and its possible implications on antibiotic treatment has emerged.7–11,15–18 Phenotypic tolerance is likely to play a significant role in most types of bacterial infections. In particular, bacteria growing in biofilms in the respiratory and urinary tracts and near prostheses and implants are known to produce high frequencies of tolerant cells which are refractory to antibiotics.17–20 Computer simulations of antibiotic treatment of planktonic bacterial populations indicate that killing of both the susceptible and the tolerant members of a bacterial population is essential to clear the infection, and that a low killing rate against tolerant cells (Figure 6) or a high frequency of tolerant cells potentially can cause treatment failure.11 The simulation results also suggest that tolerant cells can act as a reservoir of cells from which genetically altered resistant mutants can be generated.
In this study, we compared the in vitro microbiological efficacy of six antibiotics against tolerant cells enriched in the presence of ciprofloxacin, ampicillin or streptomycin. A large difference in the ability to kill tolerant cells was observed when comparing the six antibiotics. Streptomycin produced remarkably higher killing rates than any of the other drugs examined in both the single drug treatments and the treatment with two successive drugs. Tolerant cells, obtained after exposure to ciprofloxacin, ampicillin or streptomycin, were killed at increasing rates by streptomycin reaching maximum killing rates at 96 mg/L. Ciprofloxacin and ampicillin killed tolerant cells at intermediate rates in both the single drug treatments and the treatments with two successive drugs. Rifampicin, erythromycin and tetracycline produced the lowest killing rates against tolerant cells at most concentrations tested. Thus, the in vitro and simulation data presented here indicate that, due to tolerance, eradication of an infection might be impaired in particular in patients with reduced immune function and at infection sites with poor blood flow. The modest efficacy of ciprofloxacin against tolerant cells was somewhat unexpected since fluoroquinolones are known to have activity against non-growing cells. This observation speaks against the argument that the arrested growth of tolerant cells is the direct and only reason for the refractoriness against antibiotics. The reason for streptomycin's high efficacy against tolerant cells is presently unknown.
A comparison of the microbiological efficacy against tolerant and exponentially growing cells (Tables 1–3), revealed no clear correlation of efficacy. Thus, a drug with high efficacy against growing cells does not necessarily have a high efficacy against the tolerant members of a bacterial population. Thus, in the case of infections with high frequencies of tolerant cells it might be necessary to treat with a combination of two or more antibiotics to eradicate all members of the infecting bacterial population. Traditionally pharmacodynamic parameters used for the design of treatment regimens are determined on growing cells with no prior history of antibiotic exposure. These findings suggest that killing kinetics of tolerant cells should be included in future in vivo studies.
Since the killing of tolerant cells requires very high antibiotic concentrations, reversion to the susceptible state in the absence of a drug is of major importance. It has been shown that tolerant cells revert to the susceptible state whenever the antibiotic is removed and the cells are suspended in a fresh nutritionally rich medium,7,8,11 suggesting reversion is related to a shift in growth rate from slow/no growth to normal growth. Our data show that reversion to the susceptible state is dependent on the presence of nutrients in the medium and that the relative frequencies of tolerant cells go down with time in nutritionally rich medium (LB medium) when cell division occurs. In poor medium (PBS) where essentially no net cell division occurs the reversion to susceptibility was less rapid than in rich medium. However, there was some variation in the duration of tolerant state between the different antibiotics. In general cells tolerant to ciprofloxacin seemed to stay longer in the non-susceptible state compared with cells tolerant to ampicillin and streptomycin. The extended duration of the tolerant state after the drug has been removed from the environment is in vivo likely to lead to slow reversion rates since nutrients often are scarce and bacterial generation times much longer than in the laboratory medium. Thus, infections located in tissues with poor nutrient supply, e.g. soft tissue (abscesses), bone and joint infections, are likely to maintain high frequencies of tolerant cells.
In conclusion, the results obtained in this study suggest that the killing kinetics of tolerant cells could be of importance for the clinical treatment outcome. The difficulties with eradicating tolerant bacteria might be particularly problematic in immunosuppressed patients. Further in vivo studies are necessary to demonstrate the clinical significance of this phenomenon.
|
Transparency declarations |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
None to declare.
|
Supplementary data |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
The Appendix is available as Supplementary data at JAC Online (http://jac.oxfordjournals.org/).
|
Acknowledgements |
|---|
We would like to thank Professor Bruce R. Levin, Department of Biology, Emory University, Atlanta, USA, for inspiring us to do this study and for his advice and contribution to the mathematical modelling. The study was supported by the Swedish Research Council and Uppsala University, Sweden.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionTransparency declarationsSupplementary dataReferences |
|---|
1 Shah D, Zhang Z, Khodursky A, et al. (2006) Persisters: a distinct physiological state of E. coli. BMC Microbiol 6:53.[CrossRef][Medline]
2 Bigger JW. (1944) Treatment of staphylococcal infections with penicillin by intermittent sterilisation. Lancet ii:497–500.
3 Craig WA and Ebert SC. (1990) Killing and regrowth of bacteria in vitro: a review. Scand J Infect Dis Suppl 74:63–70.[Medline]
4
den Hollander JG, Knudsen JD, Mouton JW, et al. (1998) Comparison of pharmacodynamics of azithromycin and erythromycin in vitro and in vivo. Antimicrob Agents Chemother 42:377–82.
5
Fung-Tomc JC, Gradelski E, Valera L, et al. (2000) Comparative killing rates of fluoroquinolones and cell wall-active agents. Antimicrob Agents Chemother 44:1377–80.
6
Odenholt I, Cars T, Lowdin E. (2000) Pharmacodynamic studies of trovafloxacin and grepafloxacin in vitro against Gram-positive and Gram-negative bacteria. J Antimicrob Chemother 46:35–43.
7
Balaban NQ, Merrin J, Chait R, et al. (2004) Bacterial persistence as a phenotypic switch. Science 305:1622–5.
8 Keren I, Kaldalu N, Spoering A, et al. (2004) Persister cells and tolerance to antimicrobials. FEMS Microbiol Lett 230:13–8.[CrossRef][Web of Science][Medline]
9
Kussell E, Kishony R, Balaban NQ, et al. (2005) Bacterial persistence: a model of survival in changing environments. Genetics 169:1807–14.
10
Miller C, Thomsen LE, Gaggero C, et al. (2004) SOS response induction by ß-lactams and bacterial defense against antibiotic lethality. Science 305:1629–31.
11
Wiuff C, Zappala RM, Regoes RR, et al. (2005) Phenotypic tolerance: antibiotic enrichment of noninherited resistance in bacterial populations. Antimicrob Agents Chemother 49:1483–94.
12 Drusano GL. (2004) Antimicrobial pharmacodynamics: critical interactions of ‘bug and drug’. Nat Rev Microbiol 2:289–300.[CrossRef][Web of Science][Medline]
13
Greenwood D. (1985) Phenotypic resistance to antimicrobial agents. J Antimicrob Chemother 15:653–5.
14
Levin BR. (2004) Microbiology. Noninherited resistance to antibiotics. Science 305:1578–9.
15
Keren I, Shah D, Spoering A, et al. (2004) Specialized persister cells and the mechanism of multidrug tolerance in Escherichia coli. J Bacteriol 186:8172–80.
16
Kaldalu N, Mei R, Lewis K. (2004) Killing by ampicillin and ofloxacin induces overlapping changes in Escherichia coli transcription profile. Antimicrob Agents Chemother 48:890–6.
17 Sufya N, Allison DG, Gilbert P. (2003) Clonal variation in maximum specific growth rate and susceptibility towards antimicrobials. J Appl Microbiol 95:1261–7.[CrossRef][Medline]
18 Cogan NG. (2006) Effects of persister formation on bacterial response to dosing. J Theor Biol 238:694–703.[CrossRef][Web of Science][Medline]
19
Spoering AL and Lewis K. (2001) Biofilms and planktonic cells of Pseudomonas aeruginosa have similar resistance to killing by antimicrobials. J Bacteriol 183:6746–51.
20
Lewis K. (2001) Riddle of biofilm resistance. Antimicrob Agents Chemother 45:999–1007.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. I. Udekwu, N. Parrish, P. Ankomah, F. Baquero, and B. R. Levin Functional relationship between bacterial cell density and the efficacy of antibiotics J. Antimicrob. Chemother., April 1, 2009; 63(4): 745 - 757. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||