The Morphodynamics of 3D Migrating Cancer Cells

During cancer cell migration, cells on a 2D surface experience uniform mechanical cues whereas in a fibrous 3D environment there exists heterogeneity at cellular and sub-cellular levels, leading to dramatically different invasion strategies. Here, we physically characterize morphodynamics (the temporal fluctuations of cell shape) rather than real-space migration alone. By studying morphodynamics, we show that 3D cell migration is accompanied by spontaneous and rapid shape changes regulated by the extra-cellular matrix (ECM), in contrast to 2D migration. We employ machine learning to classify cell shape into five different morphological phenotypes corresponding to different migration modes. We systematically characterize the phenotype evolutions including occurrence probability, dwell time, transition flux, and 3D migrational characteristics. By tuning ECM density, pore-size, and fiber-alignment, we show local mechanical influence on migration mode switching. We find mechanosensing mediated by rho-signaling is important to the resultant morphodynamics. We demonstrate cell morphodynamics as an information-rich biomarker that is directly regulated by cell mechanosensing and contributes critically to the cell motility.