MEMORY IN CELLULAR RESPONSES TO FLUCTUATING STRESS

Bacteria use a range of mechanisms to survive in stressful conditions. However, these mechanisms are typically costly and bacteria need to optimize their use to attain the long-term advantage. In environments which change sufficiently slowly cells favor stochastic switching over sensor-mediated responses, while for rapid fluctuations cells can lock into a single phenotypic state and therefore avoid the costs of switching or sensing. However, between these two extremes lies a wide range of intermediate fluctuation timescales which were analytically inaccessible, yet critical to predict the physiological adaptations of bacteria. By developing a mathematical formalism that predicts the optimal bacterial responses in randomly changing environments, in this intermediate regime we identified a class of survival strategies, responses with memory, that minimize the impact of random fluctuations on bacterial growth and survival.

This work outlines the environmental conditions under which evolving physiological memory would be beneficial and the evolutionary changes through which this memory can be acquired. By analogy to classical physical systems, we represented the main result with a phase diagram of optimal response strategies. Classification of the transitions between different strategies suggests the ways of evolving the optimal response in a gene circuit and provides a powerful predictive tool of gene network optimization in fluctuating conditions.