Frustration and Percolation in the Collective Dynamics of Biological Neuron Networks

Robust collective responses to external temporal signals are essential for the normal function of multicellular systems, yet the mechanisms by which cells coordinate their dynamics in the presence of intrinsic heterogeneity and noisy signaling remain poorly understood. Here, we investigate how neuronal cell populations encode and transmit information carried by periodic ATP stimuli through collective calcium dynamics. Using microfluidic devices and micropatterning to precisely control stimulus timing and cell–cell connectivity, we combine quantitative experiments with causal network inference and mathematical modeling. Granger causality analysis reveals that interacting cells self-organize into spatially decentralized and temporally stationary networks that mediate information transfer via gap junctions. The structure and connectivity of these causal networks depend sensitively on the temporal profile of the external stimulus: short driving periods, or long periods with small duty fractions, lead to weakened connectivity and fragmented network topology. At the same time, we find a nontrivial role of physical coupling in shaping collective dynamics: while isolated cells synchronize more strongly at long driving periods due to prolonged entrainment, highly connected networks can instead desynchronize, despite increased gap-junction expression. A theoretical framework based on coupled excitable units, operating near a coupling-induced bifurcation, reproduces both the emergence of optimal information-transfer networks and the counterintuitive desynchronization at strong coupling. The model predicts maximal effective connectivity at an intermediate communication strength and explains how co-culturing with gap-junction-deficient cells can restore synchronization, consistent with experimental observations. Together, these results demonstrate that collective cellular responses to temporal stimuli are jointly regulated by external driving and internal cell–cell communication, revealing a trade-off between synchronization and information transfer in multicellular networks.