Loss of activity of the heat shock protein 90 (HSP90) chaperone leads to simultaneous loss of signaling proteins from many oncogenic pathways, ultimately leading to apoptosis in cancer cells. Inhibitors of HSP90 have been proposed for clinical use in a variety of solid tumors, in particular triple-negative breast cancer (TNBC), a cancer subtype that has proven difficult to treat with other targeted molecular drugs. However, HSP90 inhibitors have struggled to demonstrate efficacy in clinical trials, in part due to non-genetic resistance where signaling in survival pathways is reactivated and cells overcome loss of HSP90 activity. A major question remains as to whether this non-genetic resistance is a cell state induced by exposure to drug or one selected from pre-existing cell states. I hypothesize that HSP90 inhibition causes signaling network changes in a subset of cancer cells that allows them to maintain pro-survival signaling activity. Comprehensively measuring the signaling network in cells responding heterogeneously to drug exposure requires a systems- based approach with single-cell resolution. Mass cytometry enables us to measure up to 45 markers in individual cells in a high-throughput manner. To elucidate the mechanism for resistance to HSP90 inhibition in TNBC, I propose to quantify signaling activity in breast cancer cell lines treated with HSP90 inhibitors with mass cytometry. However, one major challenge to studying the process of drug response is that cells are destroyed during mass cytometry analysis so the same cells cannot be followed over time. To overcome this challenge, we have developed a computational method to trace the trajectories of cells responding to drug treatment in silico from high-dimensional single-cell data. My goal is to apply this computational method to dissect the mechanisms underlying resistance to HSP90 inhibitors, identifying specific signaling features unique to TNBC cells that go on to survive HSP90 inhibition. These candidate features will then be validated using pharmacological or genetic interventions to demonstrate their role in regulating resistance to HSP90 inhibition. The results of this analysis will identify strategies for effective use of HSP90 inhibitors by finding inhibitors that synergize with HSP90 inhibition to maximize death in breast cancer cells. The source of signaling activity that distinguishes HSP90 inhibitor-resistant cells will then be investigated by determining which genes are exclusively expressed in surviving cells. Then to further address whether drug resistance is induced or selected in these populations, I will interrogate the role of timing of activation of these genes in regulating survival. Moreover, I will test how varying the timing of interventions against candidate proteins or genes affects the size of the fraction of cells that are HSP90 inhibitor-resistant. We believe this work may give insight into more successful combinatorial therapeutic strategies for HSP90 inhibitors by identifying new synergistic interventions and effective drug regimens.
Though HSP90 inhibitors kill most triple-negative breast cancer cells, some cells survive treatment, ultimately making this drug ineffective in treating patients. My studies aim to develop systems-based approaches to identify which features in these resistant cells allow them to survive HSP90 inhibition. I will further characterize how and when to intervene to minimize resistance to HSP90 inhibition and maximize killing of cancer cells, thus identifying effective treatment strategies involving HSP90 inhibitors for triple-negative breast cancer.