SUMMARY Acquired resistance to anti-cancer therapeutics has proven to be one of the largest hurdles in cancer cell biology because cancer cells have the remarkable ability to adapt to diverse conditions. For example, when essential metabolic processes are blocked, some cancer cells die, but subsets of cells can survive and acquire resistance. The organelle recycling process, autophagy, provides an excellent paradigm to study metabolic adaptations in cancer. Many cancer cells are addicted to autophagy to maintain homeostasis and regenerate nutrients, but previous work highlighted the ability of rare cells to rapidly adapt and acquire new dependencies on alternate metabolic pathways. Rapid and transient adaptations to stress that manifest in the metabolome, epigenome and transcriptome have been understudied. This proposal suggests that resistance mechanisms are more complex than just pre-existing genetic differences between heterogeneous tumor cells, but instead include rapid signaling events, broad stress and metabolic responses, epigenetic changes, and the acquisition of new genetic mutations. How and when each of these factors contribute to resistance remains unknown. Many studies analyze adapted populations after they have undergone selection. The approach taken here is different: these studies aim to observe the process of selection and adaptation in action. The proposed projects will develop a set of novel tools and model systems to track the dynamic interactions of rapid signaling, stress and metabolic responses, along with transcriptional changes, epigenetic changes, and genetic alterations – all with temporal precision. Despite decades of studies on therapeutic resistance, fundamental questions remain. For example, it is critical to determine whether: A) cancer cells undergo a change in state and adapt in response to a treatment, or B) a treatment simply selects for a pre-existing state that is heterogenous and already resistant. It is critical to differentiate the dynamics between these two models to determine whether a given resistance mechanism should be targeted as a combination therapy (a consequence of Model A), or instead used as a biomarker for patient selection (a consequence of Model B). Some patients respond remarkably well to autophagy inhibition and the field is desperate for both biomarkers associated with these patients to improve patient selection, and for ways to prevent therapy resistance. To this end, these studies will facilitate the development of better autophagy-targeting cancer therapeutics. Moreover, understanding the temporally dynamic contributions of different kinds of adaptations will generate new models of cancer cell drug resistance, beyond those that model autophagy modulation.