Open Access

Biocides are used in large amounts in industrial, medical, and domestic settings. Benzalkonium chloride (BAC) is a commonly used biocide, for which previous research revealed that Escherichia coli can rapidly adapt to tolerate BAC-disinfection, with consequences for antibiotic susceptibility. However, the consequences of BAC tolerance for selection dynamics and resistance evolution to antibiotics remain unknown. Here, we investigated the effect of BAC tolerance in E. coli on its response upon challenge with different antibiotics. Competition assays showed that subinhibitory concentrations of ciprofloxacin—but not ampicillin, colistin and gentamicin—select for the BAC-tolerant strain over the BAC-sensitive ancestor at a minimal selective concentration of 0.0013–0.0022 µg/mL. In contrast, the BAC-sensitive ancestor was more likely to evolve resistance to ciprofloxacin, colistin and gentamicin than the BAC-tolerant strain when adapted to higher concentrations of antibiotics in a serial transfer laboratory evolution experiment. The observed difference in the evolvability of resistance to ciprofloxacin was partly explained by an epistatic interaction between the mutations conferring BAC tolerance and a knockout mutation in ompF encoding for the outer membrane porin F. Taken together, these findings suggest that BAC tolerance can be stabilized in environments containing low concentrations of ciprofloxacin, while it also constrains evolutionary pathways towards antibiotic resistance.

Wastewater treatment plants (WWTP) are hotspots for the development and spread of antimicrobial resistance. Antimicrobial resistant bacteria (ARB) can persist in the environment for long periods of time, despite metabolic fitness costs that often arise with resistance. Recent research efforts are striving to uncover the role of bacterial metabolism for the ecology and evolution of antibiotic resistance. The aim of this study is to understand the ecological mechanisms, which allow ARB to persist in the environment. More specifically, we aim to identify metabolic niches that can select for and against resistant bacteria. 62 E. coli strains isolated from different WWTPs with different levels of resistance to 14 antibiotics and 3 disinfectants were assembled, sequenced, and phenotypically characterized. Next, genome scale metabolic models (GEMs) were constructed, and the growth of the strains was simulated in the presence of 298 different carbon sources. Furthermore, the growth rates of the isolates were measured in the presence of 3 carbon sources to verify the model predictions. Competition experiments with synthetic microbial communities consisting of a selection of 10 WWTP isolates, 5 of which were antibiotic resistant and 5 sensitive, were carried out in minimal medium with different carbon sources. Population dynamics modelling was used to simulate the competition of isolates under different conditions. The isolates have a wide range of susceptibility to the antibiotics, while disinfectants result in a narrower range of susceptibility. GEMs identified 40 carbon sources that can be utilized for growth only by a portion of all the isolates. The prediction accuracy of the GEMs was 93% in the case of D-malate. A range of WWTP isolates were identified which use D-malate as carbon source and are susceptible to specific antibiotics. In contrast, antibiotic-resistant WWTP isolates were identified that did use sucrose as carbon source but not D-malate. Competition experiments demonstrated that changing the carbon source of the medium from sucrose to D-malate resulted in selection against the resistant isolates. Modelling the competition between isolates under different conditions suggests that adding a carbon source to a bacterial community under specific conditions could exclude resistant bacteria from a microbial community. Our data suggest that changing the available carbon source could shift the selection advantage between resistant and susceptible bacterial strains. If this strategy is confirmed experimentally in complex microbial communities, it could be applied to reduce the number of ARB in environments such as wastewater.

Biocides are used as disinfectants and preservatives; one important active substance in biocides is benzalkonium chloride (BAC). BAC-tolerant bacterial strains can survive short treatments with high concentrations of BAC. BAC tolerance and resistance have been linked to antibiotic resistance. Here, the selection dynamics between a BAC-tolerant Escherichia coli strain and a sensitive wild type were investigated under four conditions: in the absence of antibiotics and in the presence of three different sub-inhibitory concentrations of the antibiotic ciprofloxacin in liquid cultures. The wild type was selected over the BAC-tolerant strain in the absence of antibiotics, while the BAC-tolerant strain was selected over the wild type at all ciprofloxacin concentrations investigated, with a minimum selection concentration (MSC) of 1/10th of the minimum inhibitory concentration (MIC) of the wild type. Furthermore, the evolvability of resistance of the two strains to inhibitory concentrations of ciprofloxacin was assessed by performing a serial dilution evolution experiment with gradually increasing ciprofloxacin concentrations. The wild type had a higher probability to develop resistance to ciprofloxacin than the tolerant strain. By the end of the evolution experiment both strains evolved to grow at the highest ciprofloxacin concentration investigated, which was 2048 ×MIC of the wild type. The importance of these results is highlighted by the fact that concentrations of ciprofloxacin well above the calculated MSC can be found in environmental samples such as hospital wastewaters and livestock slurry. In turn, BAC is used as a disinfectant in the same settings. Thus, the selection of BAC-tolerant strains at sub-inhibitory concentrations of ciprofloxacin can contribute to the stabilization and spread of BAC-tolerance in natural populations. The prevalence of such strains can impair the effects of BAC disinfections.

The growth of pathogenic bacteria in the host is a prerequisite for infectious diseases. Antibiotic drugs are used to impair bacterial growth and thereby treat infections. In turn, growth of bacteria is underpinned by their primary metabolism. Thus,it has long been recognized that the activity of antibiotics is determined by the metabolic state of cells. However, only recently researchers have begun to systematically interrogate the links between metabolism and resistance. In their recent study, Lubrano and colleagues (Lubranoet al, 2025) apply an elegant CRISPR-based approach to the model bacterium Escherichia coli to systematically screen the effect of 15,120 mutations in genes that encode for 346 proteins which are required for growth of E. coli (also referred to as ‘essential proteins’). The authors identified a multitude of mutations that reduce the susceptibility against two antibiotics related to two very distinct chemical classes; the β-lactam antibiotic carbenicillin and the aminoglycoside gentamicin. Strikingly, the majority of the identified mutations are directly linked to primary metabolism. The work highlights the importance of metabolism in order to understand antibiotic resistance mechanisms and the ecology and evolution of antibiotic resistance. In addition, the work provides leads to design metabolism-based intervention strategies to mitigate antibiotic resistance.

Background and Aim: Wastewater treatment plants are considered as hotspots for the development and spread of antimicrobial resistance. Antimicrobial resistant bacteria (ARB) can persist in the environment for long periods of time, despite metabolic fitness costs that can arise with resistance. We are interested in understanding the mechanisms, which allow ARB to persist in the environment. More specifically, we want to identify metabolic niches that can select for resistant bacteria. Procedure/Method: 62 E. coli strains isolated from different WWTPs in Norway were used. The isolates have different levels of resistance to 14 antibiotics. The susceptibility of the isolates to 3 disinfectants was determined. In addition, genome scale metabolic models (GEMs) were constructed and the growth of the strains was simulated in the presence of 198 different carbon sources. Furthermore, the growth rates of the isolates were measured in the presence of 3 carbon sources. Findings/Results: Many of the isolates have high resistance to multiple antibiotics but only few of the isolates have higher minimum inhibitory concentrations to the disinfectants, compared to an E. coli laboratory strain. With the GEMs, we identified 40 carbon sources that can be utilized for growth only by a portion of all the isolates. The prediction accuracy of the GEMs was 93% in the case of D-Malate. A group of 10 isolates was identified of which 5 isolates are resistant to ciprofloxacin, gentamicin and tetracycline and can grow on Sucrose but not on D-Malate and the other 5 isolates are susceptible to the same 3 antibiotics but can grow on D-Malate and not on Sucrose. Implications/Applications: Our data suggest that changing the available carbon source could shift the selection advantage between resistant and susceptible bacterial strains. If this strategy is confirmed experimentally, it could be applied to reduce the number of ARB in environments like wastewater.