Logic of optimal collective migration in heterogeneous tissues

Collective cell migration is a critical process in embryogenesis and cancer invasion. Recent work has shown that uniform tissues can undergo sharp rheological transitions, with collective motion emerging above a critical cell motility. In vivo, however, migration typically involves multiple populations with distinct motile and adhesive properties, and how this heterogeneity shapes collective dynamics remains unclear. Here, using two different vertex model implementations, we show that migration of heterogeneous clusters through tissues is maximized at intermediate adhesion strength: too little and the cluster fragments, too much and cluster cell cohesion suppresses the rearrangements needed for forward motion. We test our model against recent and new data on zebrafish mesendoderm invasion, where graded Nodal signalling regulates both motility and adhesion differences. By mapping measured Nodal levels to mechanical parameters, the model not only reproduces migration outcomes across homogeneous and heterogeneous clusters, but also discriminates between alternative adhesion rules. Strikingly, the inferred parameters place the system near the predicted optimum, where adhesion is strong enough to maintain cohesion yet graded enough to allow selective coupling among heterogeneous neighbors. These results identify an optimal balance between cohesion and interfacial remodeling as a general principle coordinating collective invasion in heterogeneous tissues. Significance statementCells often migrate collectively during embryonic development and cancer invasion, but tissues are rarely uniform and different cells differ both in their adhesion and activity. Using models of tissue mechanics, we show that collective invasion is maximized at an intermediate level of adhesion within the migrating cluster cells: too little and the cluster falls apart, too much and it cannot advance. We test this principle against experiments in zebrafish gastrulation, where a signaling gradient simultaneously controls both cell motility and adhesion. The model reproduces migration outcomes across a range of experiments and identifies the adhesion rule cells use to selectively stick to neighbors. These results reveal a simple mechanical logic for how heterogeneous cell collectives coordinate invasion.