Frequency-dependent interactions determine outcome of competition between two breast cancer cell lines

Tumors are highly dynamic ecosystems in which diverse cancer cell subpopulations compete for space and resources. These complex, often non-linear interactions govern continuous spatial and temporal changes in the size and phenotypic properties of these subpopulations. Because intra-tumoral blood flow is often chaotic, competition for resources may be a critical selection factor in progression and prognosis. Here, we quantify resource competition using 3D spheroid cultures with MDA-MB-231 and MCF-7 breast cancer cells. We hypothesized that MCF-7 cells, which primarily rely on efficient aerobic glucose metabolism, would then dominate the population under normoxic conditions; and MDA-MB-231 cells, which exhibit high levels of glycolytic metabolism, would dominate under hypoxic conditions. In spheroids with single populations, MCF-7 cells exhibited equal or superior intrinsic growth rates (density-independent measure of success) and carrying capacities (density-dependent measure of success) when compared to MDA-MB-231 cells in both normoxic and hypoxic conditions. However, when the populations were mixed, MDA-MB-231 cells outcompeted MCF-7 cells under all conditions with competition dynamics determined by frequency-dependent interactions. When grown together, MDA-MB-231 suffered little from MCF-7 while MCF-7 cells experienced a disproportionately large reduction in growth rates in the presence of MDA-MB-231 cells. We frame these results in a game-theoretic (frequency-dependent) model of cancer cell interactions and conclude that competition assays can demonstrate critical density-independent, density-dependent and frequency-dependent interactions that likely contribute to in vivo outcomes.