Editor: Ewa Sadowy
Increased tolerance of Staphylococcus aureus to vancomycin in viscous media
Article first published online: 29 AUG 2007
FEMS Immunology & Medical Microbiology
Volume 51, Issue 2, pages 277–288, November 2007
How to Cite
Kostenko, V., Ceri, H. and Martinuzzi, R. J. (2007), Increased tolerance of Staphylococcus aureus to vancomycin in viscous media. FEMS Immunology & Medical Microbiology, 51: 277–288. doi: 10.1111/j.1574-695X.2007.00300.x
- Issue published online: 29 AUG 2007
- Article first published online: 29 AUG 2007
- Received 14 February 2007; revised 25 May 2007; accepted 19 June 2007.First published online October 2007.
- Staphylococcus aureus;
- antimicrobial susceptibility;
- medium viscosity;
The increased viscosity observed in biofilms, adherent communities of bacterial cells embedded in a polymeric matrix, was hypothesized to induce increased tolerance of bacteria to antibiotics. To test this concept, planktonic Staphylococcus aureus cells were grown and exposed to vancomycin in media brought to specific viscosities in order to mimic the biofilm extracellular polymeric matrix. A viscous environment was observed to decrease the vancomycin susceptibility of planktonic S. aureus to levels seen for biofilms. Both planktonic S. aureus at a viscosity of 100 mPa s and staphylococcal biofilms were able to survive at >500 times the levels of the antibiotic effective against planktonic populations in standard medium. Time-dependent and dose-dependent viability curves revealed that more than one mechanism was involved in high S. aureus tolerance to vancomycin in viscous media. Increased viscosity affects antibiotic susceptibility by reducing diffusion and the mass transfer rate; this mechanism alone, however, cannot explain the increased tolerance demonstrated by S. aureus in viscous media, suggesting that viscosity may also alter the phenotype of the planktonic bacteria to one more resistant to antimicrobials, as seen in biofilms. However, these latter changes are not yet understood and will require further study.
Biofilms, structured adherent bacterial communities embedded in extracellular polymeric substances (EPS) formed on biotic and abiotic surfaces, play a major role in the pathogenesis in many Staphylococcus aureus infections (Fitzpatrick et al., 2005). Biofilms allow bacteria to persist at the site of infection by protecting them from the host immune system and providing reduced susceptibility to antimicrobial treatment (Ceri et al., 1999; Donlan & Costerton, 2002; Hall-Stoodley et al., 2004). Although the mechanisms that increase biofilm tolerance to antimicrobial agents are not fully elucidated, it has been postulated that several phenomena contribute to biofilm defense, including delayed penetration of antimicrobials through the biofilm matrix, slow growth of the bacterial cells within biofilms, and heterogeneity of the bacterial population within the biofilm including, phenotypic variability in the population and persister cell formation (Mah & O'Toole, 2001; Harrison et al., 2005b). The EPS, which typically comprise two-thirds of the biofilm mass and are composed of mainly water (up to 95–99%), but include bacterial polysaccharides, extracellular protein and DNA as well as excreted host cellular products such as mucopolysaccharides, fibrin and collagen (Sutherland, 2001; Whitchurch et al., 2002; Pawlowski et al., 2005), contribute to the observed increased tolerance of biofilms to antibiotics, due to alteration in mass transfer and changes in bacterial physiology (Resch et al., 2005). The EPS promote the attachment of bacterial cells to abiotic and biotic surfaces (Frank & Belfort, 2003; Sadovskaya et al., 2005), and mediate intercellular bacterial aggregation and communication (Burdman et al., 2000), and, hence, biofilm formation (Cramton et al., 1999). Furthermore, the retarded diffusion processes inside biofilms, which are strongly influenced by the structure and properties of EPS, and the resulting increased viscosity inside the biofilm (Mayer et al., 1999) lead to delayed antimicrobial access to the embedded bacterial cells (Donlan, 2000; Gilbert et al., 2002) or help to concentrate enzymes that inactivate antibiotics within the biofilm. For example, delayed penetration was observed for aminoglycosides, glycopeptides, and, in some studies, for β-lactam antibiotics (Donlan & Costerton, 2002). However, several other researchers have demonstrated no barrier to diffusion through a biofilm for many antibiotics (Mah & O'Toole, 2001; Walters et al., 2003; Rani et al., 2005).
This apparent contradiction in published results might be explained by structural, functional and hydrodynamic heterogeneity within biofilms. The biofilm consists of cell clusters embedded in EPS and separated by internal channels (Stoodley et al., 1994). Owing to mass transfer through channels, the true, local flux of antibiotics in the bacterial clusters and EPS may be overestimated. Several researchers demonstrated a reduction in diffusion coefficients in cell clusters relative to those in channels. This reduction depended on solute molecule size, shape and charge (de Beer & Stoodley, 1995; de Beer et al., 1997; Bryers & Drummond, 1998;Marcotte et al., 2004).
Increased environmental viscosity and subsequent diffusion limitation may also promote differences in the replication and metabolic activity of bacterial cells within the biofilm, due to nutrient and oxygen gradients as well as local cell density variation (Walters et al., 2003; Fux et al., 2005). This heterogeneity may lead to altered antibiotic susceptibility, as it is believed that bacteria with low metabolic activity are less susceptible to antibiotics than their counterparts with normal metabolic activity (Drenkard, 2003). Moreover, certain antibiotics such as vancomycin exhibit cell density-dependent effects and are much less effective against bacterial populations with high density (Cerca et al., 2005).
As the biofilm phenotype is probably sensitive to the environmental viscosity, it is hypothesized that a viscous environment can induce tolerance to antibiotics within planktonic bacterial populations to levels found in biofilms. Independently, Gilbert et al. (1998) and Wirtanen et al. (1998) showed that Pseudomonas aeruginosa and S. epidermidis grown in 30% poloxamer gel exhibited enhanced tolerance to biocides, whereas Chan (1998) observed increased tolerance of Pseudomonas and Candida to antimicrobials in viscous media supplemented with poly(vinylpirrolidone) (PVP). Strathmann et al. (2000) also demonstrated increased tolerance to antibiotics for bacterial populations grown in agarose beads. However, given the conditions used in these experiments, the effects of viscosity cannot be discriminated from those due to the polymer. Thus, the present experiments were conducted with media supplemented with polymers displaying different properties, in order to distinguish the effects of viscosity on the efficacy of vancomycin against planktonic S. aureus from the effects of the polymer. The susceptibility of methicillin-sensitive and methicillin-resistant S. aureus to the antibiotic in artificial viscous media was compared to the susceptibility level of their counterparts within biofilms or planktonic populations under standardized conditions.
Materials and methods
Microorganisms and media
Methicillin-sensitive S. aureus ATCC 29 213 and methicillin-resistant S. aureus UOC18 (ST5-MRSA-II, kindly provided by the Biofilm Research Group, University of Calgary, Canada) were stored cryogenically at −80°C on beads in cryovials (Microbank Prolab Diagnostics, USA). During the experiments, bacteria were cultivated in standard cation-adjusted Mueller–Hinton broth (CAMHB, Difco) and CAMHB supplemented with PVP (PVP40T, PVP360T), sodium alginate (A2158) or methylcellulose (MC7140) (all from Sigma-Aldrich, USA) to adjust media viscosity to 3, 5, 10 and 100 mPa s.
Antimicrobial susceptibility assays for planktonic bacteria
The minimum inhibitory concentration (MIC) and the minimum bactericidal concentration (MBC) of vancomycin (Sigma-Aldrich, USA), as well as viability dynamics of planktonic populations, were determined by serial twofold dilution of antibiotic (resulting in a log2 concentration gradient from 1024 to 0 μg mL−1) in standard and supplemented CAMHB according to the procedures recommended by the American Clinical and Laboratory Standards Institute (CLSI, 2006). Positive Displacement Microdispensers (VWR International, USA) were used for accurate dilution and delivery procedures performed in viscous media. Briefly, an inoculum of 1 × 106 CFU mL−1 for MIC and MBC assays and 1 × 107 CFU mL−1 for viability dynamics, determined from McFarland standards, were prepared from overnight cultures grown in standard or supplemented CAMHBs and were placed into wells of 96-well microtiter plates containing the corresponding media supplemented with vancomycin at desired concentrations. The challenged plates were incubated at 37°C for 24 h. OD620 nm in the wells was measured with an AD 340C Absorbance Detector (Beckman Coulter Inc., USA) to assess bacterial growth. The lowest concentration of vancomycin that completely inhibited the growth of the planktonic cultures was considered to be the MIC. The MBC corresponded to the lowest concentration of the antibiotic that killed 99.9% of cells in the initial planktonic population as determined by viable cell counts performed from the wells following MIC determination. The staphylococcal viability dynamics were determined by viable cell counts vs. the antibiotic concentrations after 24 h of exposure.
Additionally, the minimum eradication concentration (MEC) was the lowest concentration of vancomycin that eradicated the planktonic population. Briefly, the isolates (106 CFU mL−1) were exposed to vancomycin at concentrations that exceeded MBCs for 24 h. Then, 1 mL of the bacterial suspension was centrifuged at 7000 g for 15 min and washed twice in phosphate-buffered saline (PBS). The pellet was suspended in an equal amount of fresh Trypticase soy broth (TSB) and cultivated at 37°C for 24 h to determine regrowth of the population from surviving cells. Bacterial growth was assessed by measuring the OD620 nm with the AD 340C Absorbance Detector (Beckman Coulter Inc., USA).
Biofilms were grown in MBEC-P&G devices (Innovotech, Canada) according to the manufacturer's instructions and as described previously (Ceri et al., 1999). Briefly, single colonies from a fresh overnight streak plate were suspended in TSB to concentrations of 107 CFU mL−1 as described above. The sterile peg lid of the MBEC devices was submerged in the suspension and incubated in humidified conditions at 37°C with a rotation rate of 135 r.p.m. for 24 h. Biofilms formed on the pegs of the MBEC-P&G device were rinsed with PBS to remove loosely adherent planktonic bacteria. To verify biofilm formation, three pegs were sonicated in PBS with an Aquasonic sonicator (VWR International, model 170HT) for 5 min as described by Ceri et al. (1999). The disrupted biofilms were diluted serially and plated to verify viable cell numbers.
Scanning electron microscopy
To verify biofilm formation, pegs with biofilms were rinsed in PBS, fixed in 5% glutaraldehyde for 24 h at 4°C, washed twice with distilled water, and dehydrated in serial ethanol dilutions ranging from 30% to 100%. Dehydrated samples were dried in a critical point dryer and sputter coated with gold. Biofilms on pegs were viewed with an FEI ESME XL30 scanning electron microscope (Microscopy and Imaging Facility, University of Calgary).
Antimicrobial susceptibility assays for biofilms
Two-fold geometric dilutions of vancomycin (resulting in a concentration range from 1024 to 0 μg mL−1) were prepared in 96-well microtiter plates containing standard CAMHB. The peg lids with rinsed biofilms were submerged in challenge plates containing antibiotic. These plates were incubated in humidified conditions at 37°C for 24 h. Following the exposure to vancomycin, the biofilms were transferred from the challenge plates and rinsed in PBS to dilute out the antibiotic. Biofilm pegs were placed in 96-well microtiter plates containing 200 μL of TSB and sonicated for 5 min as above. A 20 μL sample of biofilm suspension from each well was diluted and plated to determine the viability (as above), the minimum biofilm inhibitory concentration, defined as the lowest concentration of vancomycin that inhibited biofilm growth, and the minimum biofilm bactericidal concentration (MBBC), defined as the lowest concentration of the antibiotic that killed 99.9% of bacterial cells within the biofilm. To determine the minimum biofilm eradication concentration (MBEC), which is the lowest concentration of vancomycin required to eradicate 100% of the biofilm population, recovery plates from disrupted biofilms were incubated at 37°C for 24 h to detect bacterial growth as described by Ceri et al. (1999).
Time-dependent efficacy of vancomycin against S. aureus
Planktonic bacterial cultures of 1 × 107 CFU mL−1 cultivated in either standard CAMHB or CAMHB supplemented with polymers and biofilms of 1–5 × 106 CFU mL−1 were exposed to vancomycin at concentrations of either 32, 128 or 1024 μg mL−1. Challenge plates were incubated at 37°C for 1, 2, 4, 8, 24, 36, 48 and 72 h, and this was followed by 10-fold dilution and plating of the bacterial cultures to estimate viable cell numbers. From these data, the killing time (tn) was determined as the time required for vancomycin to kill 99.9% of bacterial populations.
Media viscosity measurement
Stock solutions of PVP, sodium alginate and methylcellulose were prepared in CAMHB in concentrations ranging from 1% to 20%. The viscosity of the challenge media was measured with a Brookfield Syncro-Lectric viscometer (Brookfield Engineering Laboratories Inc., USA) according to the manufacturer's instructions. The measurements were performed at 37°C.
All experiments described above were performed nine times. Students t-test or a one-way anova was used to analyze the data.
Susceptibility of the planktonic S. aureus to vancomycin in standard and viscous media
The MIC, the MBC and the MEC of vancomycin were determined to estimate the susceptibility of methicillin-sensitive S. aureus ATCC 29 213 and methicillin-resistant S. aureus UOC18 grown in standard or viscous media. The isolates were grown planktonically in standard CAMHB and CAMHB supplemented with PVP, sodium alginate or methylcellulose to adjust the media viscosity to 3–100 mPa s. There was no significant difference between susceptibility patterns for both strains tested under all experimental conditions. The ranges of MIC, MBC and MEC values observed for nine independent replicates of every strain, along with average values, are summarized in Table 1.
|Experimental conditions||Viscosity (mPa s)||MIC (μg mL−1)||MBC (μg mL−1)||MEC (μg mL−1)||Fold tolerance*|
|CAMHB||NA||1–2 (1.50)||1–2 (1.67)||8–16 (10.00)||>100|
|CAMHB+PVP||3||1–2 (1.67)||4–8 (4.44)||8–16 (12.44)||>80|
|5||2–4 (3.11)||4–8 (6.67)||16–32 (18.00)†||>60|
|10||4–8 (4.89)||32–64 (46.2)†||128–256 (213.3)†||>5|
|100||8–16 (11.56)‡||128–256 (170.7)†,‡||512–1024 (682.7)†,‡||>1.5|
|CAMHB+sodium alginate||3||2||4–8 (5.33)||8–16 (10.67)||>90|
|5||2–4 (3.33)||8||16–32 (26.67)†||>50|
|10||4–8 (4.89)||128†||512–1024 (853.3)†||>1.2|
|100||8–16 (10.67)‡||128–512 (384.0)†,‡||512 to>1024 (950.9)†,‡||>1.1|
|CAMHB+methylcellulose||3||2–4 (2.67)||4||8–16 (10.67)||>90|
|5||2–4 (3.33)||4–8 (6.67)||32–64 (42.67)†||>25|
|10||4–8 (5.00)||64–128 (85.3)†||512–1024 (682.7)†||>1.5|
|100||8–16 (10.00)‡||128–256 (213.3)†,‡||512 to>1024 (877.7)†,‡||>1.2|
|Biofilm||NA||8–16 (12.4)||256||> 1024||NA|
In all instances, the assayed strains were susceptible to vancomycin in standard CAMHB at concentrations of 1–2 μg mL−1. In polymer-containing CAMHB at a viscosity of 3 mPa s, the susceptibility was similar to that seen in the standard CAMHB. Further increases in viscosity to 5–10 mPa s increased the vancomycin MICs to 3–5 μg mL−1, still within the breakpoint for S. aureus susceptibility to vancomycin of around 4 μg mL−1. At a viscosity of 100 mPa s, the MIC values ranged from 8 to 16 μg mL−1 and the S. aureus populations showed a vancomycin-intermediate phenotype.
Whereas MIC and MBC values for vancomycin against S. aureus were equivalent for planktonic cultures grown in standard CAMHB, increasing viscosity led to a significant change in MIC and MBC values (Table 1). The MBCs of the planktonic S. aureus populations in the polymer-containing CAMHB with low viscosity (3–5 mPa s) required vancomycin concentrations that were double those of the MIC values. At higher viscosity (10–100 mPa s), MBC values were 10–50 times higher than MICs. Despite the fact that vancomycin MBC values increased proportionally to viscosity, polymers demonstrated different effects at the same viscosity. In particular, the bactericidal activity of vancomycin against planktonic S. aureus was most sensitive to the effect of sodium alginate. In high-viscosity CAMHB (100 mPa s) with sodium alginate, the MBCs reached 400 μg mL−1 on average, whereas, in the presence of PVP or methylcellulose, MBCs were significantly, lower reaching 170 and 200 μg mL−1, respectively.
The MEC values required to eradicate the planktonic staphylococcal populations were 5–10 times higher than MICs in both standard and low-viscosity polymer-containing CAMHB. However, the level of tolerance increased as the viscosity was increased further (Table 1). For example, the eradication of the planktonic staphylococcal population by vancomycin in polymer-free and low-viscosity polymer-containing CAMHB (3–5 mPa s) required 10–40 μg mL−1 of the antibiotic, whereas bacteria survived at 512 μg mL−1 and higher at a viscosity of 100 mPa s. Whereas a viscosity of 100 mPa s enhanced the staphylococcal tolerance to vancomycin to biofilm-equivalent levels, the survival potential of the majority of planktonic populations in the presence of sodium alginate or methylcellulose significantly exceeded that for S. aureus incubated in PVP-containing CAMHB. For example, average populations survived at 950 and 870 μg mL−1, respectively, in sodium alginate and methylcellulose, whereas, in the presence of PVP, the staphylococcal population was eradicated by vancomycin at 680 μg mL−1.
Biphasic viability curves for planktonic S. aureus populations were observed upon exposure to increasing vancomycin levels in the viscous CAMHB, whereas the bacterial population in standard CAMHB gradually decreased as antibiotic concentration increased (Fig. 1). The planktonic population in the polymer-free CAMHB was completely eradicated by vancomycin at 16 μg mL−1, whereas, at the same antibiotic level, only an 2 log reduction in initial bacterial populations was observed in CAMHB with a viscosity of 10 mPa s. The killing curves then reached a plateau at vancomycin concentrations of 16–128 μg mL−1, and this was followed by an abrupt reduction in viable cell numbers up to complete eradication at 1024 μg mL−1. At a viscosity of 100 mPa s, the plateau was observed at an antibiotic level of 32–256 μg mL−1, although complete eradication was not achieved at 1024 μg mL−1 (maximum tested concentration of antibiotic).
Susceptibility of the S. aureus biofilms to vancomycin with respect to susceptibility patterns for planktonic populations
Staphylococcus aureus UOC18 and S. aureus ATCC 29 213 produced well-developed biofilms of 1–5 × 106 CFU mL−1 (verified by scanning electron microscope observation; Fig. 2) after 24 h of incubation. The minimum biofilm inhibition concentrations (MBICs), the MBBCs and the MBECs were determined to estimate the susceptibility of the S. aureus biofilm in comparison with analogous indices found for planktonic populations in standard and viscous CAMHB. Table 1 shows the ranges for MBIC, MBBC and MBEC values and the mean of nine independent replicates for each strain, along with the ratios for MBEC/MEC used to estimate the ‘fold tolerances’ for the vancomycin levels required to eradicate biofilm relative to the concentrations that completely killed the planktonic population (Harrison et al., 2005a).
The growth of bacterial cells within biofilms was inhibited by vancomycin at 8–16 μg mL−1, indicating that S. aureus biofilm exhibited a vancomycin-intermediate phenotype. These values were similar to MICs found for the planktonic population at viscosity of 100 mPa s, and contrast with the susceptibility patterns for planktonic bacteria in standard and low-viscosity CAMHB, which were significantly lower than the MBICs.
The MBBC for S. aureus was 250 μg mL−1 vancomycin, which was 130 times higher than MBCs for planktonic cultures in standard CAMHB, and 25–45 times higher than effective doses for planktonic bacteria in low-viscosity media. In contrast, the MBBC values were two to four times higher than the MBCs for planktonic bacteria at a viscosity of 10 mPa s and equal to the MBCs observed at 100 mPa s.
The vancomycin MBEC values for biofilms exceeded 1024 μg mL−1 (the maximum tested concentration), which corresponds to an 1000-fold difference in susceptibility levels when compared to planktonic populations cultivated in CAMHB according to the standard comparison parameters (MBECs vs. MICs). However, the ratios of the MBECs to the MECs were 80–100 in standard CAMHB and polymer-containing CAMHB with viscosity of 3 mPa s, and 25–60 in media with viscosity of 5 mPa s. In contrast to low-viscosity media, at high viscosity levels, the MECs of vancomycin were similar for biofilm and planktonic S. aureus populations. In fact, at viscosities of 10–100 mPa s, fold tolerance dropped to 1.1–5.0.
Like planktonic populations grown in viscous media, biofilms exhibited bilinear viability curves when exposed to vancomycin (Fig. 1). Specifically, after a gradual decrease in biofilm populations observed at 16–256 μg mL−1, the number of viable cells was constant up to 1024 μg mL−1. In contrast to planktonic populations with a <2 log surviving subpopulation, biofilms produced a 3–4 log subpopulation of survivors.
Time-dependent efficacy of vancomycin against S. aureus in standard and viscous media
It has been reported that reduced antimicrobial susceptibility in biofilms is mediated by delayed antimicrobial penetration and the presence of phenotypic variants or persister cells, among other factors (Keren et al., 2004; Harrison et al., 2005b; Chambless et al., 2006). An analysis of the time-dependent killing effects of vancomycin may clarify whether the influence of viscosity on the diffusion rate can account for the loss of susceptibility among S. aureus in viscous environments. To this end, planktonic staphylococci cultivated in either standard CAMHB or CAMHB supplemented with PVP, sodium alginate or methylcellulose and staphylococcal biofilm were exposed to vancomycin at 32 μg mL−1 (resistance threshold), 128 μg mL−1 (100 × MIC) or 1024 μg mL−1 (1000 × MIC) for 1–72 h.
In standard CAMHB, bacterial populations decreased sharply as antibiotic concentration and exposure time increased. For vancomycin concentrations of 32 μg mL−1, viable staphylococcal cells were not detected in standard CAMHB after 24 h of treatment (Fig. 3), whereas the survival period dropped to 8 h at 128 μg mL−1 and to <2 h at 1024 μg mL−1. In contrast, no killing of either strain of S. aureus was seen with vancomycin at 32 μg mL−1 for 2 h at a viscosity of 10 mPa s and for 4 h at 100 mPa s in the presence of PVP or sodium alginate. After this initial delay, further exposure of staphylococci to antibiotic caused a gradual decrease in the initial populations, such that the staphylococcal populations were reduced by 2–3 logs after a 24 h exposure. A reduction in the killing activity of vancomycin was detected after a 4–8 h exposure in the presence of sodium alginate, whereas the vancomycin killing curve was linear to 24 h in PVP-containing medium. The high concentration of bacterial cells after 24 h of treatment, however, did not allow further extrapolation of killing curve behavior. The viable S. aureus population in methylcellulose-containing medium did not show an initial delay, and diminished by 2–3 logs after a 2 h exposure to 32 μg mL−1 vancomycin (Fig. 3). Following a quick reduction in the initial population, the remaining population easily survived and maintained the same viable cell number regardless of treatment length.
At higher concentrations of vancomycin (128 or 1024 μg mL−1), no delay was observed, and the killing curves were similar in all the polymer-containing media (Fig. 3). Gradual reductions in the initial staphylococcal populations occurred during 2–4 h of exposure to vancomycin at 128 or 1024 μg mL−1 in viscous media, regardless of the polymer used. Furthermore, the eradication rate for the remaining subpopulation significantly decreased.
A delay period of 4–8 h in antibiotic activity against staphylococcal biofilms was observed for both 32 and 128 μg mL−1, whereas vancomycin at 1024 μg mL−1 caused a gradual reduction in the initial biofilm population. This behavior is similar to that observed for the planktonic populations at 10 and 100 mPa s.
To quantify the time-dependent efficacy of vancomycin in viscous media, the killing time (tn), defined as the time required for killing of 99.9% of the initial staphylococcal population by vancomycin, was determined (Fig. 4) (note that tn is calculated after the initial delay period). Whereas the majority of the bacterial populations in standard CAMHB were eradicated with 32 μg mL−1 after 3 h of treatment, the bacterial populations survived for 26–42 h at a viscosity of 10 mPa s and for 42–60 h at 100 mPa s. The antibiotic at 128 μg mL−1 killed the bacterial cells in the control medium within 2 h, whereas 99.9% killing in polymer-containing CAMHB required 12–24 h at a viscosity of 10 mPa s and 26–40 h at 100 mPa s. The killing time for vancomycin at 1024 μg mL−1 increased from 30 min in the control medium to 3–4 h at a viscosity of 100 mPa s. The time required to kill 99.9% in initial biofilm populations exceeded 72 h for vancomycin at 32 and 128 μg mL−1, and reached 18 h for vancomycin at 1024 μg mL−1.
Ratios of the killing time, tn, to the maximum killing time, tmax, observed in experiments were investigated in order to distinguish the effect of purely viscous (molecular) diffusion in polymer solutions on vancomycin efficacy. It is expected that if the reduced vancomycin efficacy is a direct result of slower molecular diffusion due to an increase in medium viscosity, then plotting tn/tmax vs. viscosity would lead to the collapse of all data for different polymers and antibiotic concentrations on a single curve (Bird et al., 1960). In fact, no significant differences in tn/tmax values were observed for all tested polymers at 32, 128 and 1024 μg mL−1 of the antibiotic (Fig. 4). The tn/tmax ratios increased proportionally from 0.2 in standard medium to 1.0 in polymer-supplemented media with a viscosity of 100 mPa s.
The trends observed for the maximum killing time, tmax, determined in the polymer media appear to be consistent with results for biofilms. Figure 4c shows tn at each medium viscosity, normalized by the killing time observed for S. aureus biofilm (tb=18 h for the present experiments), as a function of viscosity. The solid lines represent a best fit of the data obtained for experiments in polymeric media, and the broken line is the extrapolation to viscosity values associated with biofilms. For S. aureus biofilms, an effective viscosity of 3500±2900 Pas has been reported by Rupp et al. (2005) and Shaw et al. (2004).
In this study, we showed that planktonic methicillin-resistant S. aureus (MRSA) and methicillin-susceptible S. aureus (MSSA) isolates, sensitive to vancomycin in standard media, displayed a vancomycin-intermediate phenotype (MIC values were 8–16 μg mL−1) at a viscosity of 10–100 mPa s. It was also observed that the vancomycin MICs, MBCs and MECs were similar in both high-viscosity media and within biofilms, showing tolerance up to 1000-fold the MICs. These observations support the view that the increased viscosity within the biofilm microenvironment is an important contributor to the high tolerance of S. aureus to vancomycin.
In addition to the influence of viscosity, polymer-specific effects on susceptibility were observed. In particular, the highest tolerance rate was detected for planktonic S. aureus exposed to vancomycin in the presence of sodium alginate, where the bactericidal concentrations reached 500 μg mL−1 and eradication concentrations exceeded 1000 μg mL−1. In contrast, the bactericidal and eradication levels of the antibiotic were 250 and 1000 μg mL−1 in the presence of PVP at the same viscosity.
A widely used explanation for the influence of EPS on biofilm tolerance to high antimicrobial levels is the delayed access of the treatment agents to target cells, due to diffusion limitation (Gilbert et al., 2002) resulting from increased viscosity. In addition, increasing the viscosity results in higher localized concentrations of bacterial cells and their products (exopolymers and extracellular enzymes), which might moderate the access of the antibiotic to the more deeply placed cells, due to consumption and/or inactivation of the treatment agent (Gilbert et al., 2002). Moreover, biofilm EPS are believed to interact with certain antibiotics, and hence prevent or delay their contact with target bacterial cells (Stewart, 1996). For example, Dunne et al. (1993), using concentration measurements, and Jefferson et al. (2005), using fluorescently tagged vancomycin, observed that staphylococcal biofilms markedly decreased the diffusion rate of vancomycin when compared to planktonic cell suspensions.
In this study, to mimic the effect of the biofilm polymeric matrix, planktonic MSSA and MRSA were exposed to vancomycin in artificial viscous media supplemented with amorphous, monodispersed (PVP and sodium alginate) and semicrystalline (methylcellulose) polymers. The possible mechanisms involved in the loss of vancomycin efficacy against S. aureus in viscous environments were assessed by time-dependent and dose-dependent viability dynamics. The multiphase viability curves observed in this study indicate that more than one mechanism might be involved in high S. aureus tolerance to vancomycin in viscous media.
In contrast to the linear killing curve observed for S. aureus exposed to vancomycin in standard CAMHB, the killing effects of the antibiotic in CAMHB supplied with either PVP or sodium alginate showed 2–4 h latent periods. This ‘inactive treatment’ period fits well with a slow penetration model developed by Chambless et al. (2006) to describe a mechanism of EPS influence on biofilm susceptibility. This model showed that delayed antimicrobial penetration provided good protection to the biofilm up to 15 h, after which bacterial viability decreased sharply. A similar model was suggested by Cogan et al. (2005), where neutralizer–antibiotic reactions consumed the treatment agent through binding of the antibiotic to the polymers. In this model, the binding capacity of polymers must therefore decrease with increased antibiotic concentration. This behavior is also consistent with observations made in this study. After a latent period, the bacterial populations exposed to 32 μg mL−1 vancomycin in media supplemented with PVP or sodium alginate gradually decreased; the higher concentrations, however, readily killed bacteria. Similar behavior was observed in biofilms, where cell viability decreased after 4–8 h of exposure to vancomycin at 32 and 128 μg mL−1, whereas antibiotic at 1024 μg mL−1 killed bacteria without any delay. Therefore, the binding activity of biofilm EPS was higher than that seen with one of the artificial media, where biofilms allowed the temporal inactivation of vancomycin at 128 μg mL−1. This observation might be explained by the fact that biofilm viscosity was determined to be five orders of magnitude greater than the viscosity values used in this study (Rupp et al., 2005), and hence implicates viscosity as an important factor in increased tolerance of biofilm to antibiotics.
Vancomycin is a large hydrophilic molecule that consists of a glycosylated heptapeptide with five aromatic amino residues, which can interact with bacterial cells and polymers by means of hydrogen bonds, as is the case with sodium alginate, a polyanion with a large number of carboxylic groups, which allowed an ‘inactive treatment’ period of 4 h, due to inactivation of a certain portion of vancomycin, although increased antibiotic concentration and longevity of treatment decreased the binding activity of alginate chains. In contrast, active hydroxyl groups in methylcellulose, which demonstrated a different behavior, are partly substituted by methyl groups, which do not interact with the antibiotic. The ‘inactive treatment’ period observed for PVP is not likely to be related to antibiotic–polymer bonds, as PVP is considered to be inert with respect to vancomycin. It can be argued that the flexible polymer chains of PVP (and sodium alginate) are able to block vancomycin, whereas methylcellulose, which forms large, amorphous zones fixed in a crystalline network, cannot physically prevent the penetration of the antibiotic into target cells. Perry et al. (2006) also pointed out that the diffusion characteristics of polymeric systems depended on the nature (flexible vs. rigid colloidal) of the polymer used.
Independent of the mechanism involved in polymer–antibiotic interactions, increasing viscosity postponed the effects of vancomycin on bacterial cells. Comparative analysis of staphylococcal viability dynamics demonstrated that 99.9% killing of S. aureus at high viscosity of either artificial media or biofilms required a 10–40 times longer period of treatment than in standard medium.
Analyzing the influence of diffusion explicitly is complicated, as other parameters, including intrinsic viscosity, polymer chain configuration in solution, concentrations of polymer, permeability, and intermolecular and intramolecular forces, must be considered concurrently. To simplify this analysis, the diffusion time was linked to the killing time (tn) required to kill 99.9% of the bacterial population, as the bactericidal efficacy of the antibiotic is considered to be proportional to its ability to reach target cells. The killing time increased proportionally with viscosity. However, the delay in the bactericidal activity of vancomycin was affected by the polymers used and the concentrations of the antibiotic. To clarify the effect of increased viscosity, ratios of the killing time (tn) to the maximum killing time observed in experiments (tmax) were considered.
In purely viscous diffusive processes, the killing time would be proportional to the viscosity, such that by scaling tmax (maximum killing time observed in experiments), all of the data for the different polymers would fall on a single tn/tmax curve. Indeed, the scaled killing time (tn/tmax) of vancomycin bactericidal activity was observed to increase proportionally to viscosity in the same manner for different experimental conditions. Moreover, the killing time required for the biofilm, tb, is reasonably predicted by extrapolating the killing time, tn, results for the planktonic cultures to viscosity values observed in biofilm. The consistency of these indices for the tested polymers and biofilm implies that viscosity is an independent factor contributing to an altered pattern of susceptibility for staphylococci in viscous environments. Our data are consistent with previous observations, where restricted diffusion of latamoxef in biofilm-like agar-entrapped Escherichia coli correlated with enhanced tolerance of immobilized cells to antibiotic (Jouenne et al., 1994).
Although retarded diffusion of vancomycin to target cells can explain the temporary reduction of vancomycin efficacy, it cannot explain the high tolerance of the subpopulations that were recognized from dose-dependent or time-dependent killing curves as populations that were not affected by either dosage or longevity of treatment. The bilinear killing curves, in which rapid killing of the majority of the bacterial populations is followed by slow killing of the remaining populations, are consistent with the concept of persister cells, a dormant population that is considered to be able to survive dramatically high concentrations of an antibiotic and recover biofilm populations after antibiotic removal (Keren et al., 2004; Harrison et al., 2005a; Lewis, 2005; Roberts & Stewart, 2005). In this study, 0.1–1% of staphylococcal biofilm populations survived at 1024 μg mL−1 vancomycin. The same portion of a planktonic staphylococcal population cultivated in viscous CAMHB was able to survive at 100–500 μg mL−1, and a smaller portion of bacterial cells remained viable even at 1024 μg mL−1. Thus, planktonic S. aureus within viscous media produced small populations of cells that survived despite continued exposure to and increased concentration of the antibiotic, similar to a biofilm population, resulting in enhanced tolerance of the population to exposure to vancomycin. Although the mechanism of the formation and tolerance of this possible surviving population is unknown, it was observed that the size of the surviving population increased with increased viscosity, as persister cell numbers do within a biofilm (Harrison et al., 2005b).
One hypothesis to explain the decreased susceptibility of biofilms is that the increased viscosity of the polymeric matrix causes substrate limitation within biofilms that creates regions of inactive and less susceptible cells (Walters et al., 2003). This assumption might be an attractive explanation for the formation of dormant, less susceptible cells. According to this hypothesis, however, bacteria would lose their tolerance to the antimicrobial agent as substrate penetrates the biofilm and becomes available (Chambless et al., 2006). The present experiments provide at least a partial test of this concept. In the experiments with artificial media, the bacterial population mostly consisted of suspended aggregates of 50 μm or less (V. Kostenko, unpublished data), which should be easily penetrated by the substratum. The substrate limitations and the subsequent metabolic inactivation might therefore be negligible, as supported by the fact that S. aureus grew at the same rate in standard and high-viscosity media (Fig. 3). Gilbert et al. (1998) also demonstrated a similar growth rate for P. aeruginosa grown either in 30% poloxamer hydrogel or in liquid batch cultures. Nevertheless, bacteria within viscous media aggregated and formed microcolonies with high localized cell density (V. Kostenko, unpublished data), which may switch on density-dependent physiologic processes such as those observed in biofilms.
Thus, mimicking of the biofilm microenvironment by means of exposure of the planktonic bacterial populations to artificial polymer-supplemented media allowed modeling of the role of EPS in the loss of susceptibility of staphylococcal biofilms to vancomycin. It was established that a viscous environment decreased the susceptibility of planktonic S. aureus to the level seen for biofilms. Multiple mechanisms might be involved in this process. Increased viscosity promotes retarded diffusion of vancomycin through the polymer solutions. Furthermore, electrostatic interaction of the antibiotic with polymers and mechanical blocking of the antibiotic by the polymeric net decreased the possibility of the antibiotic reaching the target cells. Delayed penetration, however, is only able to temporarily reduce vancomycin efficacy, especially under increasing antibiotic load. The subpopulations with high tolerance to the antibiotic are more likely to be responsible for staphylococcal survival potential. The latter is consistent with the observation that P. aeruginosa grown in 30% poloxamer gel expressed outer membrane proteins that were observed for the biofilm rather than the planktonic phenotype (Gilbert et al., 1998).
According to this study, increased viscosity, along with other mechanisms, might stimulate altered gene expression within the planktonic population, and hence induce a biofilm-like phenotype among planktonic bacteria. Despite these assumptions, mechanisms of the interaction of polymers and increased viscosity with antibiotics and bacterial cells are not fully understood and require additional studies.
We acknowledge the support of the National Sciences and Engineering Research Council of Canada (R.J.M., H.C.), the Alberta Strategic Research Investment Program (H.C.) and Nucryst Pharmaceutical Inc. through the Industrial Research Chair program (R.J.M.).
- 1960) Transport Phenomena, Wiley International Edition. pp. 495–516. John Wiley and Sons Inc., New York. , & (
- 1998) Local macromolecule diffusion coefficients in structurally non-uniform bacterial biofilms using fluorescence recovery after photobleaching (FRAP). Biotechnol Bioeng 60: 462–473. & (
- 2000) Extracellular polysaccharide composition of Azospirillum brasilense and its relation with cell aggregation. FEMS Microbiol Lett 189: 259–264. , , , & (
- 2005) Comparative assessment of antibiotic susceptibility of coagulase-negative staphylococci in biofilm versus planktonic culture as assessed by bacterial enumeration or rapid XTT colorimetry. J Antimicrob Chemother 56: 331–336. , , , , , & (
- 1999) The Calgary biofilm device: new technology for rapid determination of antibiotic susceptibilities of bacterial biofilms. J Clin Microbiol 37: 1771–1776. , , , , & (
- 2006) A three-dimensional computer model of four hypothetical mechanisms protecting biofilms from antibiotics. Appl Environ Microbiol 73: 2005–2013. , & (
- 1998) The role of viscosity in Pseudomonas and Candida phenotype expression. Ph.D. thesis, University of Calgary, Calgary. (
- Clinical and Laboratory Standards Institute (2006) Methods for dilution antimicrobial susceptibility tests for bacteria that grow aerobically; Approved standard–7th edition. NCCLS document M7-A7. Clinical and Laboratory Standards Institute, Villanova, PA.
- 2005) Modeling physiological resistance in bacterial biofilms. Bull Math Biology 67: 831–853. , & (
- 1999) The intercellular adhesion (ica) locus is present in Staphylococcus aureus and is required for biofilm formation. Infect Immun 67: 5427–5433. , , , & (
- 1995) Relation between the structure of an aerobic biofilm and mass transport phenomena. Water Sci Tech 32: 11–18. & (
- 1997) Measurement of local diffusion coefficients in biofilms by microinjection and confocal microscopy. Biotechnol Bioeng 53: 151–158. , & (
- 2000) Role of biofilms in antimicrobial resistance. ASAIO J 46: S47–S52. (
- 2002) Biofilms: survival mechanisms of clinically relevant microorganisms. Clinical Microbiol Rev 15: 167–193. & (
- 2003) Antimicrobial resistance of Pseudomonas aeruginosa biofilms. Microbes Infect 5: 1213–1219. (
- 1993) Diffusion of rifampin and vancomycin through a Staphylococcus epidermidis biofilms. Antimicrob Agents Chemother 37: 2522–2526. , & (
- 2005) The genetics of staphylococcal biofilm formation–will a greater understanding of pathogenesis lead to better management of device-related infection? Clin Microbiol Infect 11: 967–973. , & (
- 2003) Polysaccharides and sticky membrane surfaces: critical ionic effect. J Membr Sci 212: 205–212. & (
- 2005) Survival strategies of infectious biofilms. Trends Microbiol 13: 34–40. , , & (
- 1998) The use of poloxamer hydrogels for the assessment of biofilm susceptibility towards biocide treatments. J Appl Microbiol 85: 985–990. , , , , & (
- 2002) Biofilms in vitro and in vivo: do singular mechanisms imply cross-resistance? Symp Ser Soc Appl Microbiol (31): 98S–110S. , & (
- 2004) Bacterial biofilms: from the natural environment to infectious diseases. Nat Rev Microbiol 2: 95–108. , & (
- 2005a) Persister cells mediate tolerance to metal oxyanions in Escherichia coli. Microbiology 151: 3181–3195. , , , , & (
- 2005b) Biofilms: a new understanding of these microbial communities is driving a revolution that may transform the science of microbiology. American Scientist 93: 508–515. , , & (
- 2005) Use of confocal microscopy to analyze the rate of vancomycin penetration though Staphylococcus aureus biofilms. Antimicrob Agents Chemother 49: 2467–2473. , & (
- 1994) Agar-entrapped bacteria as an in vitro model of biofilms and their susceptibility to antibiotics. FEMS Microbiol Lett 119: 237–242. , & (
- 2004) Persister cells and tolerance to antimicrobials. FEMS Microbiol Lett 230: 13–18. , , , & (
- 2005) Persister cells and the riddle of biofilm survival. Biochemistry (Moscow) 70: 267–274. (
- 2001) Mechanisms of biofilm resistance to antimicrobial agents. Trends Microbiol 9: 34–39. & (
- 2004) Solute size effects on the diffusion in biofilms of Streptococcus mutans. Biofouling 20: 189–201. , , , & (
- 1999) The role of intermolecular interactions: studies on model systems for bacterial biofilms. Int J Biolog Macromol 26: 3–16. , , , , , & (
- 2005) Bacterial biofilm formation on a human cochlear implant. Otol Neurotol 26: 972–975. , & (
- 2006) Fluorescence recovery after photobleaching as a probe of diffusion in starch systems. Biomacromolecules 7: 521–530. , & (
- 2005) Rapid diffusion of fluorescent tracers into Staphylococcus epidermidis biofilms visualized by time lapse microscopy. Antimicrob Agents Chemother 49: 728–732. , & (
- 2005) Differential gene expression profiling of Staphylococcus aureus cultivated under biofilm and planktonic conditions. Appl Environm Microbiol 71: 2663–2676. , , & (
- 2005) Modeling protection from antimicrobial agents in biofilms through the formation of persister cells. Microbiology 151: 75–80. & (
- 2005) Viscoelasticity of Staphylococcus aureus biofilms in response to fluid shear allows resistance to detachment and facilitates rolling migration. Appl Environ Microbiol 71: 2175–2178. , & (
- 2005) Extracellular carbohydrate-containing polymers of a model biofilm-producing strain, Staphylococcus epidermidis RP62A. Infect Immun 73: 3007–3017. , , , & (
- 2004) Commonality of elastic relaxation times in biofilms. Phys Rev Lett 93: 1–4. , , , & (
- 1996) Theoretical aspects of antibiotic diffusion into microbial biofilms. Antimicrob Agents Chemother 40: 2517–2522. (
- 1994) Liquid flow in biofilm systems. Appl Environ Microbiol 60: 2711–2716. , & (
- 2000) Artificial biofilm model–a useful tool for biofilm research. Appl Microbiol Biotechnol 54: 231–237. , & (
- 2001) Biofilm exopolysaccharides: a strong and sticky framework. Microbiology 147: 3–9. (
- 2003) Contributions of antibiotic penetration, oxygen limitation, and low metabolic activity to tolerance of Pseudomonas aeruginosa biofilms to ciprofloxacin and tobramycin. Antimicrob Agents Chemother 47: 317–323. , , , & (
- 2002) Extracellular DNA required for bacterial biofilm formation. Science 295: 1487. , , & (
- 1998) Performance evaluation of disinfectant formulations using poloxamer-hydrogel biofilm-constructs. J Appl Microbiol 85: 965–971. , , , & (