Information thermodynamics and Maxwell's demon in cellular ion pumps

Bipartite stochastic thermodynamics is a powerful framework for analyzing a composite system's internal energy and information flows. While it has been used to study different biomolecular machines such as ATP synthase, ion-transporting proteins remain unexplored from this perspective, despite high-level functional similarity as stochastic molecular free-energy transducers. Here we use this framework to analyze the sodium-potassium pump and find considerable information flow, comparable to other molecular machines, with the ATP-consuming subsystem acting as a Maxwell demon (seemingly cheating the second law). Intriguingly, the information flow reverses direction during physiologically characteristic variation of ion concentrations and transmembrane voltage, suggesting distinct transduction modes before and after neuronal depolarization from an action potential. These findings establish ion pumps as model systems for information-thermodynamic analysis of biological free-energy transduction.