Cell-Cell Variation in Chemotaxis Gain Implements a Simulated Tempering Strategy for Efficient Navigation in Complex Environments

Abstract

Bacterial chemotaxis is a major testing-ground for concepts in systems biology, including the role of variation between individuals. Bacteria are well-known to show cell-cell variation in their tumbling frequency and adaptation time, contributing to bet-hedging against fluctuations in the environment. Recently, large cell-cell variation was discovered also in the chemotaxis gain, also known as pathway sensitivity. Variation in gain presents a puzzle, because low gain impairs the sensing of gradients, and hence one would expect gain to be maximized in all cells. Here, we provide a functional explanation for gain variation by establishing a formal analogy between chemotaxis and statistical-physics and computer-science algorithms for sampling probability distributions. We show that variation in gain implements an approach called simulated tempering, which allows sampling of attractant distributions with many local maxima and minima. Periods of high gain allow bacteria to detect and climb gradients quickly, and periods of low gain allow them to detach from local maxima and scan for new peaks. Stochastic fluctuations in gain thus allows bacteria to navigate, colonize, and grow in environments that consist of many attractant patches. In contrast, chemotaxis strategies with a constant high gain get stuck around a local maximum. Phenotypic variability in pathway gain may more generally play an important functional role for organism navigation.