Stochasticity in Mammalian Drug Resistance

Drug resistance is a global health crisis which kills 700,000 people each year, as it undermines the treatment of many infections and cancers. Despite recent advances, we still lack a complete understanding of such drug resistance processes, including the role of stochastic or “noisy” gene expression in mammalian cells. To study how stochastic gene expression affects drug resistance of mammalian cells, we combine mathematical modeling with synthetic biology. We develop a phenomenological model to explore how cellular survival depends on the interplay between the steepness of the drug’s concentration-effect curve (a fitness function) and a drug resistance gene’s expression noise. Next, we incorporate drug resistance mutations into a detailed model to predict adaptation time. Predictions from such models are tested experimentally with puromycin-treated Chinese Hamster Ovary (CHO) cells carrying synthetic negative and positive feedback gene circuits that control a puromycin antibiotic resistance gene. Overall, we show that high gene expression noise facilitates survival and evolution of resistance in high levels of drug, while the reverse is true in low levels of drug. However, we also find that these conclusions depend strongly on the steepness of the fitness function.