Decoding stress-induced adaptation of cancer cells in spatial confinement by integrating live-cell imaging and mathematical modeling

The adaptive strategies employed by cancer cells to combat stress when grown in 2D cultures have been extensively studied, including applying evolutionary game theory. However, relatively little is known about these strategies when the cells are grown in confinement that resemble the tumor microenvironment more closely. Here, we used cisplatin-sensitive and -tolerant lung cancer cells that were grown as monotypic or heterotypic spheroids in a microfluidics system and monitored in real time. Competition and cooperation between these cells were modeled by estimating the parameters through the best fit of observed measurements using a modified version of the Cellular Potts model. The results indicated that the two sensitive and tolerant populations exhibited different cooperative and competitive strategies under changing environmental conditions. Importantly, we found that in addition to initial cell type composition, the population fitness was governed by the spatial organization, intercellular interactions, and dynamic remodeling of the local microenvironment by these cells. Further, tolerant cells exhibited similarities to a persister trait and appeared to cooperate with sensitive cells, enabling them to evade chemotherapy.