Ovarian cancer cells exhibit diverse migration strategies on stiff collagenous substrata.

In homoeostasis, the shape and sessility of untransformed epithelial cells are intricately linked together. Variations of this relationship in migrating cancer cells as they encounter different microenvironments are as yet ill understood. Here, we explore the interdependency of such traits in two morphologically distinct invasive ovarian cancer cell lines (OVCAR-3 and SK-OV-3) under mechanically variant contexts. We first established a metric toolkit that assessed traits associated with cell motion and shape, and rigorously measured their dynamical variation across trajectories of migration using a Shannon entropic distribution. Two stiffness conditions on polymerized collagen I with Young's moduli of 0.5 kPa (soft) and 20 kPa (stiff) were chosen. Both the epithelioid OVCAR-3 and mesenchymal SK-OV-3 cells on soft substrata exhibited slow and undirected migration. On stiff substrata, SK-OV-3 showed faster persistent directed motion. Surprisingly, OVCAR-3 cells on stiffer substrata moved even faster than SK-OV-3 cells but showed a distinct angular motion. The polarity of SK-OV-3 cells on stiff substrata was well correlated with their movement, whereas, for OVCAR-3, we observed an unusual "slip" behavior, wherein the axes of cell shape and movement were poorly correlated. Whereas SK-OV-3 and OVCAR-3 showed greater mean deformation on stiffer substrata, the latter was anticorrelated with variation in angular motion or the mean deviation between shape and motility axis for SK-OV-3 but poorly correlated for OVCAR-3. Moreover, on softer substrata OVCAR-3 and SK-OV-3 were relatively rigid but showed greater shape variation (with OVCAR-3 showing a higher fold change) on stiffer substrata. Our findings suggest that greater deformability on stiffer milieu allow epithelioid cells to overcome constraints on the congruence in axis of shape and motion seen for mesenchymal cells and display distinct motile behaviors across this phenotypic spectrum.