Stochastic Modeling of Contact Inhibition of Locomotion

When cells collide with each other, they often exhibit "contact inhibition of locomotion" (CIL), where cells polarize and crawl away from the point of contact. Experimentally, CIL outcomes are probabilistic, which deterministic reaction-diffusion equations cannot explain. To address this, we use a stochastic model of polarization dynamics to study how intrinsic noise from the finite number of molecules affects a cell's ability to respond to contact. Our model describes diffusive motion of Rho GTPases as well as stochastic switching between the cell membrane and cytosol, with the switching rate being regulated by cell-cell contact. Our main question is how well cells detect and respond to contact and which properties of the cell's polarity and the cell-cell contact are responsible for higher sensitivity. We examine critical parameters that influence CIL, such as the location of cell-cell contact, the width of the junction, interaction duration, and the strength of regulation of Rho GTPase switching by cell-cell contact. We find that deterministic and stochastic models converge when molecule numbers are large. However, in our stochastic model, the cell's direction of polarity will wander with a diffusion coefficient that depends on the number of Rho GTPases, so even if there is no cell-cell contact, the cell will randomly repolarize. The effect of cell-cell contact can become undetectable relative to random repolarization if the contact size is small or the interaction is weak or transient.