Collective migration of cells in geometric spaces: Intrinsic correlation length racing against extrinsic confinement size

Many experiments showed that cell groups can exhibit various impressive motion modes (e.g., global rotation and local swirling) in different-sized geometrical spaces, with the underlying mechanisms still unclear. However, recent experiments showed that different types of cells perform distinct motion modes (i.e., global rotation and coherent oscillation) in the same constraint conditions, which brings further puzzles on the movements of collective cells. Here, by comprehensively considering the social interactions between adjacent cells and internal mechanical processes of the cell itself, we propose an active vertex model to bring physical insights into collective cell migration in geometrical constraints. Using this model, we can reproduce all reported motion modes of cells in circular constraints observed in existing experiments. Specifically, as the confinement size decreases, local swirling and global rotation modes successively appear in cell groups with finite correlation length, while eccentric rotation and coherent oscillation modes successively emerge in cell groups with scale-free correlations. We demonstrate that these distinct modes are subtly coded by the intrinsic correlation length of cell groups and the extrinsic confinement size. Unexpectedly, we discover that, in small-size confinements, cell groups with scale-free correlations can spontaneously transition from eccentric rotation into coherent oscillation modes, because the strong social interactions reconcile the movements of all cells. In addition, we propose a velocity indicator that can directly distinguish migration modes emerging in small-scale confinements. These findings are in broad agreement with many experiments, and shed light on the spatiotemporal dynamics of active matter.