E. coli exploits growth arrest to quickly adapt its phenotype under carbon starvation

Due to nutrient scarcity, bacteria often face starvation. Despite the major clinical and ecological implications, little is known about how starved bacteria set their phenotype. Using microfluidics combined with fluorescence time-lapse microscopy, we measure growth and gene expression in individual E. coli cells as we switch them from exponential growth to a carbon-free medium. Quantifying the volumic production of proteins across time, we find that the immediate growth arrest induced by starvation triggers a consistent gene expression program in the cells. Some promoters sharply halt expression, others show an exponential decay of expression on a ten-hours time scale, and a third subset exhibits a burst in activity before decaying exponentially. These patterns are remarkably homogeneous, with only moderate variability in time dynamics across cells. Furthermore, we find that protein decay rate also decreases exponentially with time. By combining the observed time-dependent protein production and decay rates we show how protein concentrations deep into starvation are determined mainly by the expression dynamics in the first hours of growth arrest. Finally, by inhibiting gene expression at various points in starvation, we demonstrate that cell's ability to withstand stress deep into starvation depends strongly on this early gene expression program. Our work provides foundations for quantitative studies on cell maintenance and the emergence of specific phenotypes when nutrients become scarce.