The Use of Machine Learning Methodologies to Analyse Antibiotic and Biocide Susceptibility in Staphylococcus aureus

Background

The rise of antibiotic resistance in pathogenic bacteria is a significant problem for the treatment of infectious diseases. Resistance is usually selected by the antibiotic itself; however, biocides might also co-select for resistance to antibiotics. Although resistance to biocides is poorly defined, different in vitro studies have shown that mutants presenting low susceptibility to biocides also have reduced susceptibility to antibiotics. However, studies with natural bacterial isolates are more limited and there are no clear conclusions as to whether the use of biocides results in the development of multidrug resistant bacteria.

Methods

The main goal is to perform an unbiased blind-based evaluation of the relationship between antibiotic and biocide reduced susceptibility in natural isolates of Staphylococcus aureus. One of the largest data sets ever studied comprising 1632 human clinical isolates of S. aureus originated worldwide was analysed. The phenotypic characterization of 13 antibiotics and 4 biocides was performed for all the strains. Complex links between reduced susceptibility to biocides and antibiotics are difficult to elucidate using the standard statistical approaches in phenotypic data. Therefore, machine learning techniques were applied to explore the data.

Results

In this pioneer study, we demonstrated that reduced susceptibility to two common biocides, chlorhexidine and benzalkonium chloride, which belong to different structural families, is associated to multidrug resistance. We have consistently found that a minimum inhibitory concentration greater than 2 mg/L for both biocides is related to antibiotic non-susceptibility in S. aureus.

Conclusions

Two important results emerged from our work, one methodological and one other with relevance in the field of antibiotic resistance. We could not conclude on whether the use of antibiotics selects for biocide resistance or vice versa. However, the observation of association between multiple resistance and two biocides commonly used may be of concern for the treatment of infectious diseases in the future.     

Biofilms and Biocides: Are there Implications for Antibiotic Resistance?

Considerable controversy surrounds the use of biocides in an ever increasing range of consumer products and the possibility that their indiscriminate use might reduce biocide effectiveness and alter susceptibilities towards antibiotics. These concerns have been based largely on the isolation of resistant mutants from in vitro monoculture experiments. To date, however the emergence of biocide-resistant strains in-vivo has not been reported and a number of environmental survey studies have failed to associate biocide use with antibiotic resistance. This article gives an overview of the issues as they currently stand and reviews data generated in our laboratory over the last five years where we have used laboratory microcosms of the environment and oral cavity to better understand the possible effects of real-life biocide exposure of these high risk ecosystems. In general, whilst biocide susceptibility changes can be demonstrated in pure culture, especially for E. coli towards triclosan, it has not been possible to reproduce these effects during chronic, sublethal dosing of complex communities. We conclude from this review that whilst the incorporation of antibacterial agents into a widening sphere of personal products may not overtly impact on the patterns of microbial susceptibility observed in the environment, the precautionary principle suggests that the use of biocides should be limited to applications where clear hygienic benefits can be demonstrated.     

Susceptibility of antibiotic-resistant cocci to biocides

Biocide resistance has hitherto been a poorly studied subject, possibly due to the belief that such resistance was rare and clinically insignificant. Various recent findings, however, have underlined the importance of biocide resistance as a clinically relevant phenomenon. Outbreaks of biocide-resistant organisms in hospitals have been described and the genetic mechanism for resistance to quaternary ammonium compounds (QACs) in Staphylococcus aureus has now been elucidated. Mycobacteria resistant to commonly used endoscope disinfectants are now commonly reported and have caused numerous adverse clinical events. Cross-resistance between triclosan and antituberculous drugs has been demonstrated in other strains of mycobacteria. This is related to a common mechanism of action. The work presented here describes studies into the biocide resistance of antibiotic-resistant cocci and attempts to create biocide-resistant strains in vitro. Strains of staphylococci (including methicillin-resistant Staph. aureus (MRSA)) and enterococci (including vancomycin-resistant enterococci (VRE)) had their susceptibility to biocides assayed using broth macro dilution methods and resistant strains were selected by serial subculture on biocide-containing media. Mutants were created with relative ease; for instance, triclosan minimal bactericidal concentrations (MBCs) increased from 0.002 to 3.12 mg l(-1). Some strains of MRSA which have intermediate resistance to glycopeptides were demonstrated to have decreased susceptibility to some biocides. Biocide resistance amongst enterococci was demonstrated although there was no clear correlation between biocide and antibiotic resistance. The exact mechanisms of resistance in these strains are still being studied but it is clear that biocide resistance is an important clinical phenomenon.     

nalD encodes a second repressor of the mexAB-oprM multidrug efflux operon of Pseudomonas aeruginosa

The Pseudomonas aeruginosa nalD gene encodes a TetR family repressor with homology to the SmeT and TtgR repressors of the smeDEF and ttgABC multidrug efflux systems of Stenotrophomonas maltophilia and Pseudomonas putida, respectively. A sequence upstream of mexAB-oprM and overlapping a second promoter for this efflux system was very similar to the SmeT and TtgR operator sequences, and NalD binding to this region was, in fact, demonstrated. Moreover, increased expression from this promoter was seen in a nalD mutant, consistent with NalD directly controlling mexAB-oprM expression from a second promoter.     

Mechanisms of bacterial biocide and antibiotic resistance

Resistance to antibiotics is increasingly commonplace amongst important human pathogens. Although the mechanism(s) of resistance vary from agent to agent they typically involve one or more of: alteration of the drug target in the bacterial cell, enzymatic modification or destruction of the drug itself, or limitation of drug accumulation as a result of drug exclusion or active drug efflux. While most of these are agent specific, providing resistance to a single antimicrobial or class of antimicrobial, there are currently numerous examples of efflux systems that accommodate and, thus, provide resistance to a broad range of structurally unrelated antimicrobials--so-called multidrug efflux systems. Resistance to biocides is less common and likely reflects the multiplicity of targets within the cell as well as the general lack of known detoxifying enzymes. Resistance typically results from cellular changes that impact on biocide accumulation, including cell envelope changes that limit uptake, or expression of efflux mechanisms. Still, target site mutations leading to biocide resistance, though rare, are known. Intriguingly, many multidrug efflux systems also accommodate biocides (e.g. triclosan) such that strains expressing these are both antibiotic- and biocide-resistant. Indeed, concern has been expressed regarding the potential for agents such as triclosan to select for strains resistant to multiple clinically-relevant antibiotics. Some of the better characterized examples of such multidrug efflux systems can be found in the opportunistic pathogen Pseudomonas aeruginosa where they play an important role in the noted intrinsic and acquired resistance of this organism to antibiotics and triclosan. These tripartite pumps include an integral inner membrane drug-proton antiporter, an outer membrane- and periplasm-spanning channel-forming protein and a periplasmic link protein that joins these two. Expression of efflux genes is governed minimally by the product of a linked regulatory gene that is in most cases the target for mutation in multidrug resistant strains hyperexpressing these efflux systems. Issues for consideration include the natural function of these efflux systems and the therapeutic potential of targeting these systems in combating acquired multidrug resistance.     

Comparative analysis of antibiotic and antimicrobial biocide susceptibility data in clinical isolates of methicillin-sensitive Staphylococcus aureus, methicillin-resistant Staphylococcus aureus and Pseudomonas aeruginosa between 1989 and 2000

AIMS: To analyse population minimum inhibitory concentrations (MICs) data from clinical strains of Staphylococcus aureus and Pseudomonas aeruginosa for changes over a 10-year period and to look for correlations between the antimicrobials tested. METHODS AND RESULTS: Data from the MIC study of 256 clinical isolates of Staph. aureus [169 methicillin-sensitive Staph. aureus (MSSA), 87 methicillin-resistant Staph. aureus (MRSA)] and 111 clinical isolates of Ps. aeruginosa against eight antimicrobial biocides and several clinically relevant antibiotics was analysed using anova, Spearman-Rho correlation and principal component analysis. Comparisons suggest that alterations in the mean susceptibility of Staph. aureus to antimicrobial biocides have occurred between 1989 and 2000, but that these changes were mirrored in MSSA and MRSA suggests that methicillin resistance has little to do with these changes. Between 1989 and 2000 a sub-population of MRSA has acquired a higher resistance to biocides, but this has not altered the antibiotic susceptibility of that group. In both Staph. aureus and Ps. aeruginosa several correlations (both positive and negative) between antibiotics and antimicrobial biocides were found. CONCLUSIONS: From the analyses of these clinical isolates it is very difficult to support a hypothesis that increased biocide resistance is a cause of increased antibiotic resistance either in Staph. aureus or in Ps. aeruginosa. SIGNIFICANCE AND IMPACT OF THE STUDY: The observation of negative correlations between antibiotics and biocides may be a useful reason for the continued use of biocides promoting hygiene in the hospital environment.     

Comparative proteomic analysis of Salmonella tolerance to the biocide active agent triclosan

Concern has been expressed about the overuse of biocides in farm animal production and food industries. Biocide application can create selective pressures that lead to increased tolerance to one or more of these compounds and are concomitant with the emergence of cross-resistance to antibiotics. A triclosan sensitive Salmonella enterica serovar Typhimurium and the isogenic triclosan tolerant mutant were studied at the proteomic level in order to elucidate cellular mechanisms that facilitate biocide tolerance. 2-D differential fluorescent gel electrophoresis (DIGE) compared protein profiles of parent and mutant Salmonella, in the presence and absence of triclosan. Differentially expressed proteins were identified by mass spectrometry and divided into two groups: Group A describes proteins differentially expressed between susceptible and triclosan tolerant Salmonella and includes the known triclosan target FabI which contained a mutation at the triclosan target binding site. Group B identified proteins differentially expressed in response to triclosan exposure and defines a general cell defence network. Only four proteins were common to both groups highlighting the diverse range of pathways employed by Salmonella to counteract biocides. These data suggest that sub-lethal concentrations of triclosan induce discernible changes in the proteome of exposed Salmonella and provide insights into mechanisms of response and tolerance.     

Comparative susceptibility of resident and transient hand bacteria to para-chloro-meta-xylenol and triclosan

AIMS: To determine the susceptibility of planktonic and biofilm-grown strains of resident and transient skin bacteria to the liquid hand soap biocides para-chloro-meta-xylenol (PCMX) and triclosan. METHODS AND RESULTS: Freshly isolated hand bacteria were identified by partial 16S rRNA gene sequencing. Two resident and three transient strains, as well as four exogenous potential transient strains, were selected for biocide susceptibility testing. The minimum inhibitory concentrations (MIC) and minimum bactericidal concentrations (MBC) of planktonic cells were determined. Resident and transient strains showed a range of susceptibilities to both biocides (PCMX, MIC 12.5-200 mg x l(-1), MBC 100-400 mg x l(-1); triclosan, MIC 0.6- > 40 mg x l(-1), MBC 1.3- > 40 mg x l(-1)). Strains were attached to polystyrene plates for 65 h in 96-well microtitre plates and challenged with biocide to determine the biofilm inhibitory concentration and biofilm eradicating concentration. For all strains tested, biofilms were two- to eightfold less susceptible than planktonic cells to PCMX. CONCLUSIONS: Very few transients were detected on the hand. Transients were not more sensitive than residents to the biocides and susceptibility to PCMX and triclosan was strain dependent. Biofilm-grown strains were less susceptible to PCMX than planktonic cells. SIGNIFICANCE AND IMPACT OF THE STUDY: The study provides increased knowledge about the susceptibility of skin bacteria to biocides present in typical liquid antibacterial hand soaps and suggests that the concentration of biocide employed in such products is in excess of that required to kill the low numbers of transient bacteria typically found on skin.     

Efficacy of biocides used in the modern food industry to control salmonella enterica, and links between biocide tolerance and resistance to clinically relevant antimicrobial compounds

Biocides play an essential role in limiting the spread of infectious disease. The food industry is dependent on these agents, and their increasing use is a matter for concern. Specifically, the emergence of bacteria demonstrating increased tolerance to biocides, coupled with the potential for the development of a phenotype of cross-resistance to clinically important antimicrobial compounds, needs to be assessed. In this study, we investigated the tolerance of a collection of susceptible and multidrug-resistant (MDR) Salmonella enterica strains to a panel of seven commercially available food-grade biocide formulations. We explored their abilities to adapt to these formulations and their active biocidal agents, i.e., triclosan, chlorhexidine, hydrogen peroxide, and benzalkonium chloride, after sequential rounds of in vitro selection. Finally, cross-tolerance of different categories of biocidal formulations, their active agents, and the potential for coselection of resistance to clinically important antibiotics were investigated. Six of seven food-grade biocide formulations were bactericidal at their recommended working concentrations. All showed a reduced activity against both surface-dried and biofilm cultures. A stable phenotype of tolerance to biocide formulations could not be selected. Upon exposure of Salmonella strains to an active biocidal compound, a high-level of tolerance was selected for a number of Salmonella serotypes. No cross-tolerance to the different biocidal agents or food-grade biocide formulations was observed. Most tolerant isolates displayed changes in their patterns of susceptibility to antimicrobial compounds. Food industry biocides are effective against planktonic Salmonella. When exposed to sublethal concentrations of individual active biocidal agents, tolerant isolates may emerge. This emergence was associated with changes in antimicrobial susceptibilities.     

Exposure of Salmonella enterica serovar Typhimurium to high level biocide challenge can select multidrug resistant mutants in a single step

Background

Biocides are crucial to the prevention of infection by bacteria, particularly with the global emergence of multiply antibiotic resistant strains of many species. Concern has been raised regarding the potential for biocide exposure to select for antibiotic resistance due to common mechanisms of resistance, notably efflux.

Methodology/Principal Findings

Salmonella enterica serovar Typhimurium was challenged with 4 biocides of differing modes of action at both low and recommended-use concentration. Flow cytometry was used to investigate the physiological state of the cells after biocide challenge. After 5 hours exposure to biocide, live cells were sorted by FACS and recovered. Cells recovered after an exposure to low concentrations of biocide had antibiotic resistance profiles similar to wild-type cells. Live cells were recovered after exposure to two of the biocides at in-use concentration for 5 hours. These cells were multi-drug resistant and accumulation assays demonstrated an efflux phenotype of these mutants. Gene expression analysis showed that the AcrEF multidrug efflux pump was de-repressed in mutants isolated from high-levels of biocide.

Conclusions/Significance

These data show that a single exposure to the working concentration of certain biocides can select for mutant Salmonella with efflux mediated multidrug resistance and that flow cytometry is a sensitive tool for identifying biocide tolerant mutants. The propensity for biocides to select for MDR mutants varies and this should be a consideration when designing new biocidal formulations.     

Evaluation of Epidemiological Cut-Off Values Indicates that Biocide Resistant Subpopulations Are Uncommon in Natural Isolates of Clinically-Relevant Microorganisms

To date there are no clear criteria to determine whether a microbe is susceptible to biocides or not. As a starting point for distinguishing between wild-type and resistant organisms, we set out to determine the minimal inhibitory concentration (MIC) and minimal bactericidal concentration (MBC) distributions for four common biocides; triclosan, benzalkonium chloride, chlorhexidine and sodium hypochlorite for 3319 clinical isolates, with a particular focus on Staphylococcus aureus (N = 1635) and Salmonella spp. (N = 901) but also including Escherichia coli (N = 368), Candida albicans (N = 200), Klebsiella pneumoniae (N = 60), Enterobacter spp. (N = 54), Enterococcus faecium (N = 53), and Enterococcus faecalis (N = 56). From these data epidemiological cut-off values (ECOFFs) are proposed. As would be expected, MBCs were higher than MICs for all biocides. In most cases both values followed a normal distribution. Bimodal distributions, indicating the existence of biocide resistant subpopulations were observed for Enterobacter chlorhexidine susceptibility (both MICs and MBCs) and the susceptibility to triclosan of Enterobacter (MBC), E. coli (MBC and MIC) and S. aureus (MBC and MIC). There is a concern on the potential selection of antibiotic resistance by biocides. Our results indicate however that resistance to biocides and, hence any potential association with antibiotic resistance, is uncommon in natural populations of clinically relevant microorganisms.     

Cellular impermeability and uptake of biocides and antibiotics in gram-negative bacteria

The principal targets for antibacterial agents reside at the cytoplasm and cytoplasmic membrane, damage to other structures often arising from initial events at these loci. The gram-negative bacteria offer a complex barrier system to biocides and antibiotics, regulating, and sometimes preventing, their passage to target regions. Routes of entry differ between hydrophobic and hydrophilic agents, often with a structure dependency; specialized uptake mechanisms are exploited and portage transport can occur for pro-drug antibacterials. Uptake isotherms offer insight into the sorption process and can sometimes shed light on biocide mechanisms of action. The multi-component barrier system of gram-negative bacteria offers opportunities for phenotypic resistance development where partitioning or exclusion minimizes the delivery of an antibacterial agent to the target site. Active efflux processes are recognized as increasingly relevant mechanisms for resistance, potentially offering routes to biocide:antibiotic cross-resistance. These mechanisms may be targeted directly in an attempt to compromise their role in microbial survival.     

Active efflux,a common mechanism for biocide and antibiotic resistance

Energy-driven drug efflux systems are increasingly recognized as mechanisms of antibiotic resistance. Chromosomally located or acquired by bacteria, they can either be activated by environmental signals or by a mutation in a regulatory gene. Two major categories exist: those systems energized by proton motive force and those dependent on ATP. The pumps may have limited or broad substrates, the so-called multiple drug resistance pumps, which themselves form a number of related families. The multiple antibiotic resistance (mar) locus and mar regulon in Escherichia coli and other members of the Enterobacteriaceae is a paradigm for a generalized response locus leading to increased expression of efflux pumps. One such pump, the AcrAB pump extrudes biocides such as triclosan, chlorhexidine and quaternary ammonium compounds as well as multiple antibiotics. In Pseudomonas aeruginosa, a number of multidrug efflux pumps export a broad range of substrates. Since bacteria expressing these pumps thwart the efficacy of both kinds of therapeutic agents which control infectious diseases--biocides which prevent transmission of infectious disease agents and antibiotics which treat and cure infectious diseases--they are of particular concern. The prudent use of antibiotics and biocides will guard against the selection and propagation of drug-resistant mutants and preserve the efficacy of these valuable anti-infective agents.     

Bacterial target sites for biocide action

Although biocides have been used for a century, the number of products containing biocides has recently increased dramatically with public awareness of hygiene issues. The antimicrobial efficacy of biocides is now well documented; however, there is still a lack of understanding of their antimicrobial mechanisms of action. There is a wide range of biocides showing different levels of antimicrobial activity. It is generally accepted that, in contrast to chemotherapeutic agents, biocides have multiple target sites within the microbial cell and the overall damage to these target sites results in the bactericidal effect. Information about the antimicrobial efficacy of a biocide (i.e. the eta-value) might give some useful indications about the overall mode of action of a biocide. Bacteriostatic effects, usually achieved by a lower concentration of a biocide, might correspond to a reversible activity on the cytoplasmic membrane and/or the impairment of enzymatic activity. The bacteriostatic mechanism(s) of action of a biocide is less documented and a primary (unique?) target site within the cell might be involved. Understanding the mechanism(s) of action of a biocide has become an important issue with the emergence of bacterial resistance to biocides and the suggestion that biocide and antibiotic resistance in bacteria might be linked. There is still a lack of understanding of the mode of action of biocides, especially when used at low concentrations (i.e. minimal inhibitory concentration (MIC) or sublethal). Although this information might not be required for highly reactive biocides (e.g. alkylating and oxidizing agents) and biocides used at high concentrations, the use of biocides as preservatives or in products at sublethal concentrations, in which a bacteriostatic rather than a bactericidal activity is achieved, is driving the need to better understand microbial target sites. Understanding the mechanisms of action of biocides serves several purposes: (i) it will help to design antimicrobial formulations with an improved antimicrobial efficacy and (ii) it will ensure the prevention of the emergence of microbial resistance.     

Possible implications of biocide accumulation in the environment on the prevalence of bacterial antibiotic resistance

The lethality of biocides depends upon their interaction with a number of distinct biochemical targets. This often reflects reactive chemistry for any given agent, such as thiol oxidation. Susceptibility may vary markedly between different target organisms, and changes within the more sensitive targets can alter the inhibitory effect. The multiplicity of potential targets, however, usually dictates against the development of overt resistance to concentrations used for hygienic applications. Similarly, although changes in cellular permeability toward such agents, mediated either by envelope modification or the induction of efflux-pumps may reduce susceptibility, they rarely influence the outcome of treatments at use-concentration. It has recently been proposed that chronic exposure of the environment to biocides used in a variety of commercial products might expose some microbial communities to subeffective concentrations causing emergence of resistant clones. Such resistance might relate to mutational changes in the most susceptible target or to regulatory mutants that cause the constitutive expression of certain efflux pumps. Although selection of organisms with such modifications is unlikely to influence the effectiveness of the biocides, changes in their susceptibility to third-party antibiotics can be postulated. This is particularly the case where a cellular target is shared between a biocide and an antibiotic, or where induction of efflux is sufficient to confer antibiotic resistance in the clinic. Although such linkage has been demonstrated in the laboratory in pure culture, it has not been documented in environments commonly exposed to biocides. In nature, the effects of chronic stressing with biocides are complicated by competition between microbial community members that may result in clonal expansion of naturally insusceptible clones. Received 11 March 2002/ Accepted in revised form 16 August 2002     

The MexJK efflux pump of Pseudomonas aeruginosa requires OprM for antibiotic efflux but not for efflux of triclosan

Using the biocide triclosan as a selective agent, several triclosan-resistant mutants of a susceptible Pseudomonas aeruginosa strain were isolated. Cloning and characterization of a DNA fragment conferring triclosan resistance from one of these mutants revealed a hitherto uncharacterized efflux system of the resistance nodulation cell division (RND) family, which was named MexJK and which is encoded by the mexJK operon. Expression of this operon is negatively regulated by the product of mexL, a gene located upstream of and transcribed divergently from mexJK. The triclosan-resistant mutant contained a single nucleotide change in mexL, which caused an amino acid change in the putative helix-turn-helix domain of MexL. The MexL protein belongs to the TetR family of repressor proteins. The MexJK system effluxed tetracycline and erythromycin but only in the presence of the outer membrane protein channel OprM; OprJ and OprN did not function with MexJK. Triclosan efflux required neither of the outer membrane protein channels tested but necessitated the MexJ membrane fusion protein and the MexK inner membrane RND transporter. The results presented in this study suggest that MexJK may function as a two-component RND pump for triclosan efflux but must associate with OprM to form a tripartite antibiotic efflux system. Furthermore, the results confirm that triclosan is an excellent tool for the study of RND multidrug efflux systems and that this popular biocide therefore readily selects mutants which are cross-resistant with antibiotics.     

Significance of biocide usage and antimicrobial resistance in domiciliary environments

Recent events have raised awareness of the need for effective hygiene in the home. Not least is the requirement to reduce antibiotic resistance by reducing the need for antibiotic prescribing. Current evidence suggests that improved hygiene in the domestic setting could have a significant impact. Recently, it has been suggested that widespread biocide usage, particularly in consumer products, may be a contributory factor in antibiotic resistance. In developing home hygiene policies, however, it is important that biocide use as an integral part of good hygiene practice is not discouraged in situations where there is real benefit. Although laboratory data indicate possible links, it is necessary to assess whether and to what extent biocide exposure could contribute to antibiotic resistance in clinical practice. The extent to which reduced susceptibility to biocides resulting from biocide exposure could compromise their 'in-use' effectiveness must also be considered. Equally, it is important that changes in susceptibility induced by biocide exposure are assessed relative to those induced by antibiotic exposure or the phenotypic changes induced by 'normal' environmental 'stresses'. It is proposed that to be effective, home hygiene policy should be based on the concept of risk assessment and risk prevention. Using this approach, critical risk situations are identified and appropriate hygiene procedures applied to reduce risks. This may involve either soap and water cleaning, or cleaning combined with a disinfection process. A 'targeted' hygiene approach not only provides the most effective means of preventing infectious disease, it also offers a means of addressing concerns about 'too much hygiene' and 'too many antibacterials' amongst a public who have lost confidence regarding appropriate hygiene for their home environment.     

Toxicity of triclosan, penconazole and metalaxyl on Caulobacter crescentus and a freshwater microbial community as assessed by flow cytometry

Biocides are widely used for domestic hygiene, agricultural and industrial applications. Their widespread use has resulted in their introduction into the environment and raised concerns about potential deleterious effects on aquatic ecosystems. In this study, the toxicity of the biocides triclosan, penconazole and metalaxyl were evaluated with the freshwater bacterium Caulobacter crescentus and with a freshwater microbial community using a combination of single- and double-stain flow cytometric assays. Growth of C. crescentus and the freshwater community were repressed by triclosan but not by penconazole or metalaxyl at concentrations up to 250 μM. The repressive effect of triclosan was dependent on culture conditions. Caulobacter crescentus was more sensitive to triclosan when grown with high glucose at high cell density than when grown directly in sterilized lake water at low cell density. This suggests that the use of conventional growth conditions may overestimate biocide toxicity. Additional experiments showed that the freshwater community was more sensitive to triclosan than C. crescentus , with 10 nM of triclosan being sufficient to repress growth and change the phylogenetic composition of the community. These results demonstrate that isolate-based assays may underestimate biocide toxicity and highlight the importance of assessing toxicity directly on natural microbial communities. Because 10 nM of triclosan is within the range of concentrations observed in freshwater systems, these results also raise concerns about the risk of introducing triclosan into the environment.