UC San Diego

Despite recent advances in cancer treatment, resistance to therapy and metastatic progression remain critical drivers of mortality. One central mechanism underlying therapy resistance and metastasis is tumor heterogeneity. Within the tumor bulk, genetic and epigenetic diversity fuel variable responses to therapy and a spectrum of invasive potential. In particular, rare subpopulations of tumor cells that reactivate developmental signals are uniquely primed for therapy resistance and metastatic success. These cells, often referred to as cancer stem cells, are enriched for the ability to self-renew in the face of therapy, driving eventual relapse. Deepening our understanding of the molecular dependencies of these aggressive cells may provide new opportunities for therapeutic intervention. In collaboration with a fellow graduate student (Nikki Lytle) who led this project, we used transcriptional and epigenetic profiling paralleled by a genome-wide CRISPR analysis to map the molecular dependencies of pancreatic cancer stem cells. This integrated approach revealed an unexpected utilization of immuno-regulatory signals by pancreatic cancer cells, and identified the nuclear hormone receptor (RORγ) as a targetable dependency in pancreatic cancer stem cells. We expanded preclinical work to test RORγ inhibitors, providing new evidence that clinical grade RORγ inhibitors can block pancreatic cancer growth and deplete cancer stem cells in vivo. These studies also revealed a unique epigenetic landscape in cancer stem cells, suggesting upstream epigenetic regulation of stem cell fate. Thus, to follow this work, I used a curated functional screen for stem cell-enriched epigenetic factors, ultimately identifying the SWI/SNF subunit SMARCD3 as an epigenetic dependency in pancreatic cancer stem cells. Using diverse genetic mouse models, I showed that Smarcd3 dependency is bimodal, with a preferential impact in established tumors, improving survival and chemosensitivity in vivo. Finally, I leveraged genetically engineered mouse models to identify and test clinical inhibitors that target cancer stem cells. Using a genetic reporter for the stem cell signal Msi2, I helped conduct an image-based screen and found that clinical inhibitors of MEK signaling inhibited Msi2 and blocked CSC growth in vivo. Together, these studies generate a comprehensive molecular profile of the landscape and functional requirements of pancreatic cancer stem cells that may be used to identify new therapeutic targets in the future.