A Statistical Physics Approach to Cellular Response Under Strong Perturbations

Statistical physics has been successful in accounting for phenomena involving large numbers of components by employing a probabilistic approach, thereby offering predictions for the collective properties of systems. Biological cells, comprising a vast array of interacting components such as proteins, RNA molecules, and metabolites, present a complex network. This network, intricately shaped by evolution, often resists a straightforward statistical physics characterization due to its highly specific interaction patterns. In this study, we demonstrate that under conditions of acute but non-lethal stress, the perturbed state of a cell can be modeled effectively through random network dynamics. Such strong perturbations seem to randomize the network dynamics, thus allowing the use of a statistical physics framework. Our experimental data, observing the recovery dynamics of bacteria following intense perturbations, aligns with the concept of physical aging in disordered systems. Further experimental investigations into gene expression in single cells and populations levels corroborate the predictions made by our model. In particular, it allows distinguishing between bacteria at growth arrest because of chaotic dynamics from those that reach growth arrest by biological regulation. This novel predictive description of cellular behavior under acute perturbations holds potential for novel approaches in combating bacterial infections and managing cancer relapse post-treatment.