Bimodal stringence-mediated response of metal-detecting luminescent whole cell bioreporters: Experimental evidence and quantitative theory

Abstract

Whole cell luminescent bacterial sensors are used to monitor the bioavailability and toxicity of metals in aquatic media. Any rationale of the time response of such metal-detecting biosensors necessarily asks for the so-far missing measurement and modeling of the impacts of medium nutritional quality on lux-reporter expression and bioluminescence production. In this work, improvement of nutritional conditions in bioluminescence assays by increasing amino acid concentration is shown to generate a transition from metal concentration-dependent monomodal to bimodal bell-shaped signal. The long-term component of the latter features a stringence-mediated adaptation of the cells to deficiency in nutrients supply from the medium. The demonstration is based on the analysis of the ∼17 h response of bioluminescent Escherichia coli engineered with luxCDABE reporter genes placed under the transcriptional control of a cadmium (0−22 nM bulk solution concentration) inductible-PzntA promoter or of the ribosomal RNA rrnB P1 promoter repressed by (p)ppGpp alarmones synthesized by cells in nutritional stress. The time-dependence of normalized PzntA-luxCDABE biosensor signal is successfully reproduced by our recent theory elaborated for the kinetics of bioluminescence emission by metal-responsive biosensors. The formalism allows assessment of luciferase half-life and of the time variations of cells photoactivity with medium composition. The theoretically-determined long-term kinetics of cell photoactivation is remarkably supported by independent measurements on the rrnB P1-luxCDABE bioreporter. Results clarify the origin of the mismatch between maxima in photoactive biomass and in bioluminescence over time, and they offer practical options for addressing contaminant toxicity via cell photoactivity derived from deconvolution of normalized biosensor response.