Inducible cell-to-cell signaling for tunable dynamics in microbial communities

Abstract

The last decade has seen bacteria at the forefront of biotechnological innovation, with applications including biomolecular computing, living therapeutics, microbiome engineering and microbial factories. These emerging applications are all united by the need to precisely control complex microbial dynamics in spatially extended environments, requiring tools that can bridge the gap between intracellular and population-level coordination. To address this need, we engineer an inducible quorum sensing system which enables precise tunability of bacterial dynamics both at the population and community level. As a proof-of-principle, we demonstrate the advantages of this system when genetically equipped for cargo delivery. In addition, we exploit the absence of cross-talk with respect to the majority of well-characterized quorum sensing systems to demonstrate inducibility of multi-strain communities. More broadly, this work highlights the unexplored potential of remotely inducible quorum sensing systems which, coupled to any gene of interest, may facilitate the translation of circuit designs into applications.

Fig. 1. Design and characterization of the p-coumaric acid mediated iQS strain.

a Diagram of the iQS genetic circuit. The chemical structures of the molecules involved in the synthesis of the QS molecule are shown at the top. b Diagram to illustrate predicted dynamics associated with the inducible quorum sensing system as a function of population density and external inducer concentration. c Fluorescence microscopy images showing a composite of phase-contrast and GFP fluorescence in microfluidic traps. The data from the raw fluorescence values reflect the iQS dynamics predicted in part b. d Data from microplate reader experiment obtained by culturing the iQS strain in different p-coumaric acid concentrations. All data points represent mean ± standard deviation of three independent replicates. e Microplate reader experiment data obtained by culturing the wild-type E. Coli  strain in a range p-coumaric acid concentrations. All data points represent mean (solid line) ± standard deviation of three independent replicates (shaded areas). Source data are provided as a Source Data file.

Fig. 2. Characterization of the inducible syncronized lysis circuit (iSLC).

a Genetic diagram of the iSLC strain. b Simulations of the mathematical model showing three different dynamics at low (blue), medium (orange) and high (purple) inducer values. Medium values are predicted to result in sustained oscillations. Left: steady state maximum and minimum cell population values are plotted for a range of inducer concentrations. Right: simulated time traces for three representative p-coumaric acid values predict three emergent population dynamics. c Comparison between the iSLC and SLC dynamics based upon the model simulations in part b. d Diagram of the microfluidic device used to generate the inducer gradient. e Heatmaps representing the fluorescence time traces of all 14 traps present per column of the device. GFP signal is used as a proxy for population density. Four different inducer conditions are shown: low, medium, high, extra high, respectively. f Top: representative time series images from the fluorescence channel with three different inducer concentrations: low, medium, high, respectively. Bottom: fluorescence time traces plotted together with computer simulations of the mathematical model. For the simulation (dashed lines) time units are arbitrary, therefore the correspondence is strictly qualitative. Source data are provided as a Source Data file.

Fig. 3. Characterization of the iSLC kill switch properties in microfluidics and liquid culture.

a Illustrated iSLC strain kill switch mechanism. b Example time traces (n = 104) extracted from the transmitted light channel (gray). Solid black line represent the mean. At time zero the cells were induced with 500 nM p-coumaric acid. c Movie stills (4x) of the microfluidic chip before (top) and 6 h after (bottom) induction with 500 nM p-coumaric acid. Left side shows magnified images (x30) of a single representative trap. d Cell viability measured by CFU count following addition of the killing signal (p-coumaric acid) in liquid culture. Individual data points are represented by circles, the bars represent mean ± standard deviation of the three independent replicates. Source data are provided as a Source Data file.

Fig. 4. iQS enabled modulation of orthogonal multi-strain dynamics.

a Schematics representing co-culture of the two strains used with orthogonal inducible (iSLC) and non-inducible (SLC) quorum sensing, respectively. b Heatmaps representing the fluorescence time traces of all 14 traps present per column of the device. Top rows show the GFP values and bottom rows the CFP values. Fluorescence signals are used as a proxy for population density. c Shown at the top are Movie stills of the co-culture for three inducer concentrations at multiple time points. The corresponding fluorescence time traces for GFP and CFP are plotted at the bottom. Source data are provided as a Source Data file.