Responses to the interspecies quorum-sensing signal autoinducer-2 (AI-2) regulate the patterns of gene expression that promote biofilm development. Escherichia coli also senses AI-2 as a chemoattractant, a response that requires the periplasmic AI-2-binding protein LsrB and the chemoreceptor Tsr. Here, we confirm, as previously observed, that under static conditions highly motile E. coli cells self-aggregate and form surface-adherent structures more readily than cells lacking LsrB and Tsr, or than ΔluxS cells unable to produce AI-2. This difference is observed both at 37 and 30 °C. Cells deleted for the genes encoding the lsrACDBFG operon repressor (ΔlsrR), or the AI-2 kinase (ΔlsrK), or an AI-2 uptake channel protein (ΔlsrC), or an AI-2 metabolism enzyme (ΔlsrG) are also defective in biofilm formation. The Δtsr and ΔlsrB cells are totally defective in AI-2 chemotaxis, whereas the other mutants show normal or near-normal chemotaxis to external gradients of AI-2. These data demonstrate that chemotaxis to external AI-2 is necessary but not sufficient to induce the full range of density-dependent behaviours that are required for optimal biofilm formation. We also demonstrate that, compared to other binding-protein-dependent chemotaxis systems in E. coli, low levels (on the order of ~250 molecules of periplasmic LsrB per wild-type cell and as low as ~50 molecules per cell in some mutants) are adequate for a strong chemotaxis response to external gradients of AI-2.
Herbicides are frequently released into both rural and urban environments. Commercial herbicide formulations induce adaptive changes in the way bacteria respond to antibiotics. Salmonella enterica sv. Typhimurium and Escherichia coli were exposed to common co-formulants of formulations, and S. enterica sv. Typhimurium was exposed to active ingredients dicamba, 2,4-D and glyphosate to determine what ingredients of the commercial formulations caused this effect. Co-formulants Tween80 and carboxymethyl cellulose induced changes in response, but the pattern of the responses differed from the active ingredients, and effect sizes were smaller. A commercial wetting agent did not affect antibiotic responses. Active ingredients induced changes in antibiotic responses similar to those caused by complete formulations. This occurred at or below recommended application concentrations. Targeted deletion of efflux pump genes largely neutralized the adaptive response in the cases of increased survival in antibiotics, indicating that the biochemistry of induced resistance was the same for formulations and specific ingredients. We found that glyphosate, dicamba, and 2,4-D, as well as co-formulants in commercial herbicides, induced a change in susceptibility of the potentially pathogenic bacteria E. coli and S. enterica to multiple antibiotics. This was measured using the efficiency of plating (EOP), the relative survival of the bacteria when exposed to herbicide and antibiotic, or just antibiotic, compared to survival on permissive media. This work will help to inform the use of non-medicinal chemical agents that induce changes in antibiotic responses.