Functional specialization and spontaneous oscillations in the chemotaxis system of Escherichia coli

Genetically identical bacterial populations exhibit substantial phenotypic variation. Using programmable microfluidics and single-cell FRET, we recorded chemotaxis responses to a composite stimulus from hundreds of Escherichia coli cells. By simultaneously inferring key sensory parameters in the standard nonlinear model of bacterial chemotaxis, we characterized the joint parameter distribution across the isogenic population. Surprisingly, ~5% of cells displayed spontaneous oscillations in chemosensory activity under constant stimulation. These oscillations likely reflect critical coupling within receptor clusters, as predicted by theory. Incorporating the inferred parameters into agent-based simulations revealed regions of parameter space that map onto specialized navigation tasks. Together, our results link population-level parameter distributions to task performance and uncover an oscillatory phenotype with putative critical origins.