Figure 6

The degree of cell clustering allows us to estimate q . All results correspond to experiments where cells were pretreated with Mitomycin–C to prevent cell proliferation. Isolated cells were identified in several regions along the experimental transects, as shown to scale in (A). The scale bar corresponds to 1.5 mm. The inset in (A) illustrates a schematic representation of our definition of an isolated cell that occupies a circular region of at least radius 18 μ m that contains no other cells. The inset in (A) is not to scale. Experimental snapshots in (B–C) show Propidium Iodide images at t=0 and t=48 hours, respectively, for experiments initialised with 30,000 cells. The scale bar corresponds to 0.1 mm. Red crosses indicate cells which were identified as isolated cells by the image analysis software. Snapshots of the model simulations are shown in (D–F), at t=0 hours, at t=48 hours with weak cell–to–cell adhesion, q=0.1 and at t=48 hours with strong cell–to–cell adhesion, q=0.9. Simulations were performed using D=243μ m²hour⁻¹. Red circles correspond to isolated cells, while black circles indicate all other connected cells. Results in (G–H), show the time evolution of the average percentage of isolated cells for experiments initialised with 20,000 and 30,000 cells, respectively. The error bars correspond to one standard deviation from the mean (experimental – black, model – red, blue and green) and all simulation results were averaged over twenty realisations. Equivalent simulations of the mathematical model with no cell–to–cell adhesion (red lines) and strong cell–to–cell adhesion are superimposed (green lines). The best match solutions using q=0.3 and q=0.5, respectively, are shown in blue. The error surfaces in (I) and (J) show Error _C (D,q), given by Equation 5, for various values of D and q. Simulation results are averaged over twenty identically–prepared realisations. The red, green and blue crosses in (I) and (J) correspond to the solutions superimposed in (G) and (H), respectively.