Correspondence: Sima Yaron, Technion-Israel Institute of Technology, Faculty of Biotechnology and Food Engineering, Haifa 32000, Israel. Tel.: +972 4 829 2940; fax: +972 4 829 3399; e-mail: firstname.lastname@example.org
Triclosan is a biocide whose wide use has raised a debate about the potential benefits vs. hazards of the incorporation of antimicrobials in consumer products. The purpose of the present study was to determine whether exposure of biofilms of Salmonella enterica serovar Typhimurium to triclosan influences the tolerance of the bacteria towards antibiotics such as ciprofloxacin and vice versa. A synergistic antibiofilm activity was observed when the biofilms were treated with triclosan before or together with ciprofloxacin, and an additive activity was observed with planktonic cells. For example 500 μg mL−1 triclosan and 500 μg mL−1 ciprofloxacin reduced the number of viable cells in the biofilm by 1.6 and 0.5 log, respectively. However, the sequential treatment of 500 μg mL−1 triclosan followed by ciprofloxacin resulted in 4.8 log reduction. Combination indexes (CI) for biofilms treated with triclosan followed by ciprofloxacin were 0.7, 0.32 and 0.25 for reduction of 90%, 99% and 99.9%, respectively, indicating a synergism. For planktonic cells, CIs were 1±0.1, indicating an additive effect. Therefore, it was suggested that triclosan weakens the ability of biofilm-associated cells to survive exposure to ciprofloxacin in the biofilm, probably by improving the permeability or the activity of ciprofloxacin.
Triclosan is a bactericide frequently added to antiseptic products as diverse as toothpastes, deodorants, soaps and lotions (Russell, 2004). It is also used to reduce microbial loads from different surfaces such as food-processing plants (http://www.made-in-china.com), chopping boards or plastic lunchboxes (Russell, 2004). The frequent use of triclosan has resulted in recorded elevated concentrations of triclosan in groundwater (Kolpin et al., 2002; Singer et al., 2002) and in human plasma and milk (Allmyr et al., 2006, 2008). The reason for the wide usage of triclosan is its low toxicity (Reiss et al., 2009) and its activity against a broad range of bacteria. However, there is a debate about the benefit of triclosan-containing products in reducing bacteria compared with similar product without triclosan (Nole et al., 2000; Larson et al., 2003). Triclosan affects the cell membrane structure and function. Low concentrations interfere with bacterial nutrient uptake, whereas high concentrations induce leakage of intracellular components (Regos et al., 1979). Part of the antimicrobial activity is gained by integration of triclosan in the membrane and disruption of its packed structure (Villalain et al., 2001; Guillen et al., 2004). When triclosan enters the cell it also specifically inhibits the enzyme enoyl–acyl carrier protein reductase, which catalyzes the last step in bacterial type II fatty acid biosynthesis (McMurry et al., 1998a; Levy et al., 1999; Heath et al., 2000).
Bacteria use multiple mechanisms to overcome the antimicrobial activity of triclosan, including mutations in the enoyl reductase, alterations of the cell envelope, active efflux and expression of triclosan-degradative enzymes (Meade et al., 2001; Schweizer, 2001; Yazdankhah et al., 2006). In Escherichia coli and Salmonella enterica, overexpression of the AcrAB-TolC efflux pumps decreased the susceptibility to triclosan (McMurry et al., 1998b; Webber et al., 2008b). These efflux pumps are membrane proteins that transport a variety of toxic compounds out of the cell, and contribute to the development of a low-level resistance to various antibiotics, including tetracycline, chloramphenicol, cephalosporins, penicillins, nalidixic acid and fluoroquinolones (George & Levy, 1983; Cohen et al., 1989).
In a previous study, we observed that S. enterica serovar Typhimurium in a biofilm is resistant to triclosan. There was only 1 log reduction in biofilms treated with 1000 μg mL−1 triclosan, a concentration that is in the range of the triclosan concentrations in commercial preparations (600–20 000 μg mL−1) (Tabak et al., 2007). Resistance of Salmonella in the biofilm was attributed to low diffusion through the extracellular matrix and to an adaptive response, which was obtained by changes in the expression of genes such as acrAB and marA (encoding the AcrAB efflux pump and its activator MarA, respectively). Because the induced systems might provide further tolerance to triclosan and to other antimicrobials, we hypothesized that the exposure of biofilms to triclosan could influence the susceptibility to some antibiotics such as ciprofloxacin. The aim of the present study was to determine whether exposure of S. Typhimurium biofilm and planktonic cells to one antimicrobial agent influences the tolerance to the second.
Materials and methods
Bacterial strains and preparation of cells in different phases of growth
Salmonella enterica serovar Typhimurium ATCC 14028 and its mutants MAE52 and MAE190 were described previously (Romling et al., 2000; Gerstel & Romling, 2001; Zogaj et al., 2001). For biofilm formation, overnight cultures of MAE52 were diluted (1 : 30) in fresh Luria–Bertani (LB) broth without NaCl and incubated in 24-well microplates (1.5 mL) for 24 h at 37 °C with gentle shaking (130 r.p.m.). Under these conditions, MAE52 forms a biofilm at the air–liquid surface (pellicle). This biofilm is stable and can be easily removed from the interface (Tabak et al., 2007). As was shown before, each biofilm contains approximately 108 CFU (Scher et al., 2005, 2007). For various analyses, biofilms were gently removed from the surface of the broth with sterile tweezers and washed with 10 mL saline. Enumeration of the cells in the biofilms before and after each treatment was conducted by disruption of the biofilm with glass beads, followed by conventional plate counting as described previously (Scher et al., 2005).
MAE52 planktonic cells at the stationary phase were prepared by collecting the broth under the biofilm after 24 h of incubation at 37 °C. The stationary cultures of MAE190, which do not form biofilms due to deletion of the genes encoding for the production of cellulose (bcsA) and curli (agfBA), were prepared by incubation at 37 °C under the same conditions as MAE52. Bacteria at the logarithmic phase of growth were collected after 4 h of incubation. All planktonic cells were collected by centrifugation (4500 g, 15 min) and resuspended in saline to a final concentration of 108 CFU mL−1.
Effect of triclosan and ciprofloxacin on viability of biofilm and planktonic cells
To determine the effect of triclosan or ciprofloxacin on the viability of biofilm-associated cells, each prewashed biofilm was placed in 2 mL triclosan solution (0–1000 μg mL−1) at 25 °C or 37 °C, or in 2 mL ciprofloxacin solution (0–1000 μg mL−1) at 37 °C for 0, 30 and 60 min. At each time point, a biofilm was taken, washed twice with saline, disrupted with glass beads as described (Scher et al., 2005) and plated onto LB agar for counting.
To study the effect of triclosan before ciprofloxacin treatment, biofilms were placed in solutions containing 0–500 μg mL−1 triclosan for 30 min at 37 °C. Then, the biofilms were washed in saline three times and treated with ciprofloxacin (0–500 μg mL−1) for 1 h at 37 °C. To study the effect of ciprofloxacin before triclosan treatment, biofilms were placed in solutions containing 0–500 μg mL−1 ciprofloxacin for 1 h at 37 °C. The biofilms were then washed in saline and treated with triclosan (0–500 μg mL−1) for 30 min. To study the effect of ciprofloxacin together with triclosan, biofilms were placed in solutions containing 0–500 μg mL−1 ciprofloxacin and 0–500 μg mL−1 triclosan for 1 h at 37 °C. The treated biofilms were washed, disrupted with glass beads and diluted for plate counting. Sequential and combined treatments of triclosan and ciprofloxacin were also conducted with the planktonic cells at the logarithmic and stationary phases as described above. The treated planktonic cells were collected by centrifugation, washed three times and resuspended in saline, diluted 10-fold and plated.
To assess possible synergistic activity, the combination indexes (CI) were calculated according to the equation: CI=(D)T/(Dx)T+(D)C/(Dx)C, where (D)T and (D)C are doses of triclosan and ciprofloxacin in combination, and (Dx)T and (Dx)C are doses of triclosan and ciprofloxacin that produce x% effect when used alone. CI<1, =1, and >1 indicate synergism, additive effect and antagonism, respectively. The simplest definition for additive effect indicates that the effect of the two compounds together is greater than the effect of each alone, and synergism occurs when the combined effect exceeds that predicted by the sum of the individual actions of the compounds (Chou, 2006). For each set of experiments, we calculated the lowest concentrations that resulted in 90%, 99% and 99.9% reduction. Ciprofloxacin alone at a concentration as high as 1000 μg mL−1 resulted in approximately 80% reduction in the biofilm; thus, (Dx)C was displayed in all calculations as 1000 μg mL−1, an assumption that does not affect the conclusions about synergy when the calculated CI<1, because the real (Dx)C would reduce CI even more.
Each experiment was conducted at least three times in duplicate. Data were analyzed with microsoft excel version 7, and statistically processed using the one-way anova method, followed by the Tukey–Kramer test. P-values <0.05 were regarded as significant.
Susceptibility of S. Typhimurium in biofilm to sequential treatment with triclosan followed by ciprofloxacin
The minimal inhibitory concentration (MIC) of ciprofloxacin for planktonic S. Typhimurium (wt) used in this study was 0.125 μg mL−1, but a reduction of <0.7 log of viable cell counts was obtained in the biofilm after 1 h of incubation at 37 °C with the highest concentration used (1000 μg mL−1). The MIC of triclosan for planktonic cells was 0.5 μg mL−1. In biofilms, there was only a 1 log reduction even with 1000 μg mL−1 triclosan at 25 °C as was described (Tabak et al., 2007), but exposure to 500 and 1000 μg mL−1 triclosan at 37 °C resulted in 1.6 and 3 log reduction, respectively. Because treatments with triclosan at 37 °C were found to be more effective in killing the bacteria, we continued the experiments at 37 °C. When biofilms were treated with triclosan followed by ciprofloxacin at concentrations of 50–500 μg mL−1 triclosan and 250–500 μg mL−1 ciprofloxacin, a 4–5 log reduction was observed. This reduction in viability was much larger than the sum of both compounds treated alone, indicating a possible synergy (Fig. 1). For example, treatment with 500 μg mL−1 triclosan or ciprofloxacin reduced the CFU by 1.6 and 0.5 log, respectively. However, the sequential treatment with the same concentration of triclosan followed by ciprofloxacin resulted in a 4.8 log reduction of the viable cell count. Even lower concentrations (50 μg mL−1 triclosan followed by 250 μg mL−1 ciprofloxacin) resulted in a 3 log reduction, which is significantly higher than the effect of triclosan alone and ciprofloxacin alone, each compound at 500 μg mL−1 (P<0.05). The calculated CI indexes were CI=0.7 for 90% reduction, CI=0.32 for 99% and CI=0.25 for 99.9%, indicating a synergism. This synergy was also observed when the biofilms were treated with triclosan at 25 °C followed by a treatment with ciprofloxacin at 37 °C (data not shown).
Susceptibility of S. Typhimurium in biofilms to combined treatment of triclosan and ciprofloxacin
To determine whether triclosan and ciprofloxacin have synergistic activity when they are used together, the experiments described above were repeated with both agents at the same time. A very low effect on the viability of biofilm-associated cells (<1.5 log reduction) was observed when the cells were treated with triclosan and ciprofloxacin simultaneously, at ciprofloxacin concentrations of 0–16 μg mL−1 (Fig. 2). However, a significant antibiofilm activity was observed at higher concentrations of ciprofloxacin. For example, while 500 μg mL−1 triclosan reduced 1.6 log and 250 μg mL−1 ciprofloxacin reduced 0.25 log, the combined treatment resulted in a 5.0 log reduction (P<0.05). The calculated CI indexes were CI=1.0 for 90% reduction, CI=0.32 for 99% and CI=0.45 for 99.9%, indicating a synergism at higher concentrations and additive effect at the lower concentrations.
Susceptibility of S. Typhimurium to sequential treatment with ciprofloxacin followed by triclosan
A sequential treatment with ciprofloxacin followed by triclosan was also more effective against biofilms than the sum of each treatment alone at high concentrations of ciprofloxacin (>125 μg mL−1) (Fig. 3). Yet the decrease was smaller compared with treatments with triclosan before ciprofloxacin or with both compounds together, and the effect was less significant statistically. For example, 500 μg mL−1 triclosan and 250–500 μg mL−1 ciprofloxacin reduced 4.8 log when the biofilms were first treated with triclosan, 5.0 log in the mixed treatment and 3.8 log when they were first treated with ciprofloxacin (P<0.05). The calculated CI indexes were CI=1.0 for 90% reduction, CI=0.35 for 99% and CI=0.6 for 99.9%, indicating a synergism at the higher concentrations.
Susceptibility of planktonic cells to treatment with triclosan and ciprofloxacin
When planktonic cells were treated with triclosan and ciprofloxacin, an additive effect was observed (CIs were 1±0.1). In planktonic cells, the order of treatments did not have any effect on cell viability, and very similar results were obtained with all strains (wt, MAE52 and MAE190). A summary of the comparison between planktonic cells at the stationary phase of growth and biofilm-associated cells is shown in Fig. 4. Interestingly, although each compound had low activity against biofilm-associated cells as compared with planktonic cells, the combined treatment reached a similar reduction in biofilm and planktonic cells.
In 2005, the Non-Prescription Drug Advisory Committee of the US Food and Drug Administration (FDA) discussed the potential benefits and risks associated with antiseptic products marked as ‘antimicrobial’. Much of the debate regarding consumer antiseptic products was focused on the use of products that contain triclosan, and the conclusions of the FDA meeting resulted in a call for further research regarding the benefits and risks of consumer antiseptic products used in the community setting (Aiello et al., 2007). In a recent study, we showed that the ‘in use’ concentrations of triclosan are probably not efficient in killing Salmonella embedded in biofilms (Tabak et al., 2007). In the present research, we went further and tried to determine whether triclosan influences the activity of ciprofloxacin in the biofilm. However, we did not study the evolutionary process of emergence and selection of resistant mutants, as had been shown and discussed in the past (McMurry et al., 1998b; Mazzariol et al., 2000; Chuanchuen et al., 2001; Tkachenko et al., 2007), but focused on mechanisms of adaptation in the biofilms. To the best of our knowledge, the effect of a sequential or combined exposure of Salmonella in biofilms to biocides and antibiotics had not been investigated.
We investigated the dual effect of triclosan and ciprofloxacin (used together or sequentially) against S. Typhimurium biofilms. Initially, we determined the susceptibility of biofilm-associated Salmonella to ciprofloxacin, and confirmed that biofilm-associated Salmonella are less susceptible to ciprofloxacin as compared with planktonic cells. Previous articles also showed that different biofilm-associated cells resist exposure to ciprofloxacin (Desai et al., 1998; Anderl et al., 2000; Aiassa et al., 2006; Rodriguez-Martinez et al., 2007). The resistance to ciprofloxacin was not attributed to the low diffusion of ciprofloxacin through the biofilm matrix, because ciprofloxacin readily penetrated biofilms formed by different bacteria such as Klebsiella pneumoniae, E. coli or Pseudomonas aeruginosa (Anderl et al., 2000; Rodriguez-Martinez et al., 2007). Hence, it was suggested that ciprofloxacin failed to kill bacteria in the biofilm because the bacteria do not grow or grow slowly (Anderl et al., 2003).
In line with our previous studies, demonstrating that, in a biofilm, triclosan induces the transcription of the acrAB and marA (Tabak et al., 2007), we expected that triclosan would have increased the tolerance of biofilm-associated cells to ciprofloxacin, because overexpression of MarA or AcrAB leads to a low level of resistance to antibiotics such as quinolones (Hachler et al., 1991; Miller et al., 1994; Okusu et al., 1996; Alekshun & Levy, 1997; Randall et al., 2004). In contrast to our expectations, the results pointed towards synergistic activity between triclosan and ciprofloxacin. Exposure of biofilms to triclosan before ciprofloxacin resulted in a significant reduction in viable counts. This reduction was dependent on the concentrations of both triclosan and ciprofloxacin. A similar, but less notable, phenomenon was observed when the biofilms were treated with triclosan and ciprofloxacin together, because in this case an additive effect was observed at the lower concentrations and synergism at the higher concentrations. The lowest effect was observed when biofilms were treated with ciprofloxacin before triclosan. Furthermore, the biofilm did not give the embedded cells any advantage over planktonic cells in surviving the combined treatment, because both the biofilm and the planktonic cells were equally susceptible to this treatment. Similarly, Aaron et al. (2002) showed that while biofilm-grown P. aeruginosa were tolerant to the treatment with ciprofloxacin or to the combined treatment with tobramycin and piperacillin–tazobactam, the same isolates were sensitive to a combined treatment of tobramycin, piperacillin–tazobactam and ciprofloxacin. Two possible mechanisms may explain the synergistic activity in the biofilm: (1) bacteria use the same cellular mechanisms to adapt to ciprofloxacin and triclosan, such as the AcrAB efflux pumps (McMurry et al., 1998b; Webber et al., 2008b), and the double attack is beyond the capability of the function of these defense systems; (2) triclosan weakens the bacteria in a way that increases the bacterial susceptibility to ciprofloxacin, probably by increasing the permeability of the antibiotic. Based on evidence from the literature, and based on our results about the importance of the order of treatments, the second option is more likely. It is known that triclosan embeds into the membranes, weakens the van der Waals interactions between adjacent phospholipid chains and disrupts the packing of the phospholipids molecules (Villalain et al., 2001; Guillen et al., 2004). These subtle membrane structural perturbations may increase the permeability of ciprofloxacin. Moreover, alteration in the expression of genes that have a role in the cell membrane structure and function were observed in Staphylococcus aureus mutant strains resistant to triclosan and ciprofloxacin (Tkachenko et al., 2007).
The synergy in the biofilm also indicates that the low activity of ciprofloxacin in the biofilm can not only be explained by the slow growth of the bacteria as was suggested in the past (Anderl et al., 2003), because exposure to triclosan does not increase the growth rate of bacteria. It is possible that the low penetration of ciprofloxacin into the cells has a dominant contribution to this resistance. Thus, further investigation of the spatial location of triclosan and ciprofloxacin in biofilms formed not only on the air–liquid interface, but also on solids, their ability to invade the cells, and their effect on the membrane structure is needed.
It was shown that triclosan resistance is multifactorial and that there are at least three distinct mechanisms of triclosan resistance in Salmonella (Webber et al., 2008a, b). Our findings show that mechanisms of resistance in biofilms and particularly mechanisms of cross-resistance cannot be predicted from data about mechanisms of resistance in planktonic cells, and that observations about resistance due to the selection of specific mutants is not always correlated with an adaptive response in the biofilm. Thus, we conclude that knowledge about the consequences of the usage of biocides alone or combined with other antimicrobials is still incomplete, particularly in microbial biofilms.
This work was supported by the Technion Research and Development and Lando/Ben-David Funds.