Cross-feeding percolation phase transitions of inter-cellular metabolic networks

Abstract

Intercellular exchange networks are essential for the adaptive capabilities of populations of cells. While diffusional exchanges have traditionally been difficult to map, recent advances in nanotechnology enable precise probing of exchange fluxes with the medium at single-cell resolution. Here we introduce a tiling-based method to reconstruct the dynamic unfolding of exchange networks from flux data, subsequently applying it to an experimental mammalian co-culture system where lactate exchanges affect the acidification of the environment. We observe that the network, which initially exhibits a dense matrix of exchanges, progressively breaks up into small disconnected clusters of cells. To explain this behaviour, we develop a two-parameter Maximum-Entropy multicellular metabolic model that incorporates diffusion-driven exchanges through a set of global constraints that couple cellular behaviors. The model predicts a transition from a densely interconnected network to a sparse, motif-dominated state as glucose and oxygen consumption levels shift. We characterize such a crossover both numerically, revealing a power-law decay in the cluster-size distribution at the critical transition, and analytically, by computing the critical line through a mean-field approximation based on percolation theory. By comparing empirical data with theoretical predictions, we find that populations evolve towards the sparse phase by remaining near the crossover point between these two regimes. These findings offer new insights into the collective organization driving the adaptive dynamics of cell populations.