Quantifying metabolic and clonal dynamics of evolving cancer cells

Scientists and clinicians have known for decades that cancer evolves within each patient as new clones emerge and the tumor progresses and spreads. Nonetheless, the molecular adaptations of clones evolving in dynamic environments remain difficult to define. Further, clones evolve for years prior to diagnosis and treatment, yet we know relatively little about cancer evolution before therapy. To directly quantify subclonal dynamics, we combined several novel methodologies to rigorously and comprehensively quantify human cancer cell dynamics by 1) evolving cancer cell populations that are comparable in size to clinically detectable human tumors (billions of cells) in highly controlled experimental conditions over extended periods, 2) analyzing cellular phenotypes and environmental parameters in real-time, 3) imposing precisely defined selective pressures such as treatment or metabolite depletion, 4) and enabling longitudinal sampling and characterization of cancer cells and their respective microenvironments. In combination with bioreactor culturing, we used genetic barcoding to quantify clonal dynamics occurring within each population and track the evolutionary fate of subclonal lineages. The size and duration of these cultures allowed us to consistently generate many samples from the same population over time, which is critical for modeling subclonal evolution occurring in patients and to identify adaptive mechanisms that are convergent or heterogeneous across clones. Thus far, we have successfully barcoded 14 human cell lines across solid and hematologic tumor types - including pancreatic, colorectal, leukemia, and others - and cultured these lines to evolve and expand unique clones from each population. We hypothesized that the survival of cancer cells under environmental constraints, including treatment-induced stress or nutrient-depleted microenvironments, is driven by intrinsic adaptive mechanisms that enable treatment resistance and continued evolution. Our results showed that cancer cells established highly dense and proliferative populations, with clones distinctly evolving across evolutionary pressures, including glucose limiting conditions and chemotherapy, while exhibiting ongoing reprogramming of gene and protein expression. Despite maintaining the media flow, oxygen level, and pH of each culture, cancer cells nonetheless depleted select metabolites, thereby fostering an imbalanced nutrient environment that is analogous to human tumors. In addition, we isolated clones from various timepoints across experiments to validate barcodes, quantify phenotypes, and enable experiments with defined subclone proportions. The impact of this work is to directly quantify the evolution of subclones under defined selective pressures, which is significant because the adaptive mechanisms we discover are expected to reveal novel therapeutic targets for enhancing treatments.