Triclosan is a broad-spectrum antimicrobial agent that has been incorporated into many household and medical products. Bacteria with high levels of triclosan resistance were isolated from compost, water, and soil samples. Two of these bacteria, Pseudomonas putida TriRY and Alcaligenes xylosoxidans subsp. denitrificans TR1, were able to use triclosan as a sole carbon source and clear particulate triclosan from agar. A decrease in triclosan concentration was measured by HPLC within 6 h of inoculation with strain TriRY and 24 h with strain TR1. Bioassays demonstrated that triclosan was inactivated in liquid cultures and/or embedded in plastic by the growth of strain TriRY and strain TR1, permitting the growth of triclosan-sensitive bacteria.
Triclosan is a broad-spectrum antimicrobial agent that is incorporated into household and medical products as diverse as surgical drapes, antibacterial soaps, kitchenware, and paints. Widespread use of this compound has led to the detection of triclosan and its derivatives in the environment [1–3]. Triclosan targets fatty acid biosynthesis through specific interference with the NADH binding site [4,5] of the enoyl-ACP reductase FabI in Escherichia coli and InhA in Mycobacterium smegmatis. Alteration of this target enzyme provides low-level resistance but such bacteria remain susceptible to commercial triclosan concentrations. Bacteria that possess FabK, an enoyl-ACP reductase not affected by triclosan, instead of the susceptible FabI, are intrinsically resistant to higher levels of triclosan . Overexpression of a multidrug efflux pump locus also leads to low-level triclosan resistance in laboratory and clinical strains of E. coli. The high-level intrinsic resistance of Pseudomonas aeruginosa to triclosan is through active efflux of the compound . Triclosan is modified by industrial strains of the fungi Trametes versicolor and Pycnoporus cinnabarinus through conjugation of sugar moieties, leading to reduced toxicity . A consortium of bacteria has been reported to metabolize triclosan through cometabolism . In order to identify bacteria from environmental sources able to tolerate high levels of triclosan, samples from water, soil, and compost were collected. From these samples many different bacteria with moderate to high levels of triclosan resistance were isolated. The mechanism of triclosan resistance in two of these organisms was investigated through biochemical analysis and bioassays. The ability of one of the organisms to colonize triclosan-embedded plastic and form biofilms was also studied.
2Materials and methods
2.1Culture growth, identification, and characterization
Compost samples were collected from the Allegheny College Environmental Science Garden. Bacteria with high levels of triclosan resistance were isolated from compost samples by dilution plating on 1% triclosan test agar (1% TTA (tryptic soy agar; Difco, Detroit, MI, USA) supplemented with 10 g l−1 triclosan (KIC Chemicals, Armonk, NY, USA)). The pH of the agar directly surrounding the bacterial growth was monitored daily using pH paper. Each organism was characterized by its growth characteristics, standard staining and biochemical reactions, as well as through fatty acid analysis (Microbial ID, Newark, DE, USA). The ability of these bacteria to use triclosan as a nutrient was determined by growing cultures in M9 mineral salts medium (Sigma, St. Louis, MO, USA) with 2 g l−1 triclosan as the sole carbon source.
Detoxification was determined by growing two triclosan-resistant strains, Pseudomonas putida TriRY and Alcaligenes xylosoxidans TR1, at 24°C and 225 rpm in tryptic soy broth (TSB; Difco) containing 0.4 mg l−1 triclosan along with two controls, an uninoculated flask, and one inoculated with P. aeruginosa. After 48 h, cultures and controls were filter sterilized through Millipore (Bedford, MA, USA) type GS membranes. A portion of the filtrate from each culture was inoculated with either E. coli K12 or Staphylococcus aureus, bacteria susceptible to low levels of triclosan . Growth of these bacteria at 37°C and 225 rpm was monitored by measuring OD590nm for 12 h.
Growth of S. aureus on plastic previously colonized by strain TR1 was also examined. Pieces (1 cm2) of clear antibacterial plastic containing triclosan (WalMart, Bentonville, AR, USA) were surface-sterilized in 10% sodium hypochlorite and incubated in a Petri dish containing 10 ml TSB inoculated with 0.025 ml of an overnight culture of A. xylosoxidans TR1 at 24°C. After 10 days the plastic was washed in sterile distilled water and observed using phase contrast microscopy. A portion of the sample was surface-sterilized for 2 min as above then rinsed four times in sterile distilled water and allowed to air-dry under sterile conditions. The cleaned plastic samples were then placed in Petri dishes with 10 ml of TSB inoculated with an overnight culture of S. aureus. Controls of uninoculated plastic and samples not treated with strain TR1 were included. S. aureus biofilm formation was observed after 10 days of growth at 24°C by phase contrast microscopy following a sterile water wash. A catalase test was used as a biochemical marker for S. aureus (as opposed to A. xylosoxidans) colonization of the plastic.
2.3Chromatographic analysis of triclosan
Conical tubes containing 5 ml of 0.2% TTB (triclosan test broth; tryptic soy broth (Difco) supplemented with 2 g l−1 triclosan) were inoculated with 3×106 cells of an overnight culture of A. xylosoxidans TR1, P. putida TriRY, or E. coli K12, and incubated with shaking at 225 rpm and 24°C. Periodically, tubes were removed and centrifuged at 6000 rpm for 5 min. Ethanol (3 ml) was added to the pellets; the samples were vortexed for 30 s, and centrifuged for an additional 5 min. The supernatant from the second centrifugation was then added to the initial supernatant and high pressure liquid chromatographic (HPLC) analysis performed. The Waters Breeze system with an Xterra RP18 5 μm particle diameter (4.6×150 mm) column (Waters, Milford, MA, USA) with a guard column was used. Samples were eluted isocratically with a mobile phase of 60 acetonitrile:40 water:0.5 acetic acid with a flow rate of 1 ml min−1 (injection size 20 μl). The column was maintained at 30°C and triclosan was detected at an absorbance of 280 nm. To ensure that the peak detected at 11 min was triclosan, a sample was spiked with 2 g l−1 triclosan solution in 95% ethanol. Triclosan concentrations were determined by comparison to a standard curve of triclosan dissolved in 95% ethanol.
3Results and discussion
Two triclosan-resistant organisms were identified as P. putida TriRY and A. xylosoxidans subsp. denitrificans TR1 through standard biochemical tests and fatty acid analysis. When grown on 1% TTA, which is supersaturated with triclosan and appears cloudy due to suspended crystals, a zone of clearing which expanded over time was produced around the colonies (Fig. 1). There were two possible explanations for the zone of clearing. The first was that triclosan was being solubilized by the resistant organisms through production of a solvent or raising the pH of the medium . The second was that triclosan was being removed from or converted into a more soluble form within the agar. Both organisms were able to utilize triclosan as a sole carbon source in M9 minimal medium (data not shown) and no significant pH changes were recorded in the medium, supporting removal of triclosan as the probable clearing mechanism.
To determine if triclosan was being solubilized or degraded in the 1% TTA plates, P. putida TriRY and A. xylosoxidans TR1 were grown in 0.2% TTB. A laboratory strain of P. aeruginosa which is resistant to triclosan through active efflux  was used as a control. Lack of E. coli and S. aureus growth in the control tubes indicated that triclosan was not bound to the filter or bacterial cells. Both E. coli and S. aureus were able to grow in media that had initially been inoculated with P. putida TriRY and A. xylosoxidans TR1 (data not shown). This indicates that no active triclosan remained after triclosan-resistant bacteria were allowed to grow on media containing the antibacterial agent. A. xylosoxidans TR1 was able to colonize a triclosan-impregnated plastic (Fig. 2A) while triclosan-sensitive S. aureus did not form a biofilm after 10 days (Fig. 2B). S. aureus was able to colonize the plastic pretreated with A. xylosoxidans TR1 following removal of the TR1 biofilm (Fig. 2C). S. aureus colonization was confirmed through a positive reaction for catalase. This indicates that triclosan was removed from both liquid and solid substrates by A. xylosoxidans subsp. denitrificans TR1.
HPLC analysis demonstrated that triclosan decreased in the liquid growth medium within 6 h by P. putida TriRY and 12 h by A. xylosoxidans TR1. The concentration of triclosan decreased about 10-fold within the experimental period (Fig. 3). No decrease in triclosan was seen in uninoculated samples or those inoculated with E. coli (data not shown). Conjugated products of triclosan, as previously identified in triclosan detoxification by T. versicolor and P. cinnabarinus, were not observed in P. putida TR1 or A. xylosoxidans TR1 culture extracts at 280 nm, since there was a single peak with the identical retention time of triclosan. Combined with the growth of P. putida TriRY and A. xylosoxidans TR1 in medium containing triclosan as a sole carbon source and the ability of these isolates to detoxify triclosan-containing media and plastics, degradation appears to be the mechanism of triclosan detoxification by these bacteria.
The ability of P. putida and A. xylosoxidans to degrade halogenated phenolic compounds is well documented. A. xylosoxidans subsp. denitrificans dechloronates 2,4-dichlorobenzoate and uses the resulting product as a sole source of carbon and energy . The production of dioxygenase by this organism enables it to degrade 1,3-dichlorobenzene . P. putida can degrade trichlorobenzenes  and chlorocatechols through the ortho or meta pathways . These studies indicate that strains related to TR1 and TriRY produce enzymes capable of triclosan degradation. The data presented in this paper support the hypothesis that triclosan resistance in A. xylosoxidans TR1 and P. putida TriRY is through degradation of the antibacterial agent.
Both P. putida and A. xylosoxidans are responsible for a steady rise in nosocomial infections since 1975, although neither is currently considered a major human pathogen [19–21]. A. xylosoxidans is resistant to the antiseptics chlorhexidine, benzethonium chloride, and aklyldiaminoethylglycine [21,22], and both organisms are resistant to benzalkonium chloride as well as multiple antibiotics [22,23]. Use of these compounds or triclosan in hospital disinfection may thus not be effective to combat nosocomial transfer of A. xylosoxidans or P. putida.
The fact that both A. xylosoxidans TR1 and P. putida TriRY form biofilms on antibacterial plastics has serious implications. Biofilms may be more persistent in environments where triclosan is used as a disinfectant and thus available to serve as foci of inoculation and infection. Also, if products that are exposed to the common soil bacterium A. xylosoxidans subsp. denitrificans do not retain their antibacterial properties, potentially harmful organisms such as S. aureus may grow on those surfaces. To our knowledge this is the first report of the removal of active triclosan from both liquid media and solid materials by pure bacterial cultures.
This work was supported by National Institute of Health Grant AI047854-01. We thank Margaret Nelson for her review of the manuscript.