Drift–diffusion model of stress-induced aging in bacterial recovery from starvation and drug exposure

Abstract

After starvation or drug exposure, genetically identical bacteria resume growth after broadly distributed lag times, and the distribution lengthens with the duration of the stress. This aging has been attributed to a broad spectrum of relaxation times within every cell. We develop the alternative in which each cell relaxes with a single rate and the rates differ between cells. A heritable physiological state is displaced by the stress and recovers at a cell-specific rate; the population is described by the distribution of these rates, estimated without assuming its form. Fitted to eight data sets from three laboratories, the model separates a clock common to all cells, which stretches the bulk of the distribution, from a slowing of a minority of cells, supported by a calibrated test after abrupt serine hydroxamate. We discuss the questions that the present data leave open, among them whether aging eventually saturates, and propose experiments that would resolve them.