Abstract
Pancreatic ductal adenocarcinoma (PDAC) is characterized by its broad, primary resistance to genotoxic
chemotherapy, targeted therapy, and immunotherapy. The singular driving oncogene of PDAC is KRAS, which
is mutated in 95% of cases. Unfortunately, inhibitors of the specific mutant KRAS alleles found in PDAC are
not yet clinically available, and all attempts to target downstream effector pathways have been stymied by
feedback loops and other resistance mechanisms.
Work from our collaborators found that inhibition of the MEK pathway, a key KRAS effector, results in
compensatory upregulation of autophagy, a survival pathway. Combination MEK and autophagy inhibition was
shown to have synergistic effects in reducing viability of PDAC cell lines and led to tumor responses in model
systems and individual case reports from human patients. However, only a subset of patients have responded
to MEK/autophagy inhibitors in early clinical trials. Among those who respond, most tumors progress within a
few months. Nevertheless, this is potentially the first example of an active regimen lacking genotoxic agents.
Published mechanisms for the synergy between MEK and autophagy inhibitors in PDAC do not readily
explain the rapid emergence of resistance, do not account for cellular heterogeneity within the malignant
epithelial compartment, and do not address the role of the tumor stroma in the response to therapy. The
overall goal of this proposal is take a holistic approach to understanding cellular responses to MEK/autophagy
inhibition by studying the effects of treatment on every cell type in the tumor. We will do this by performing
single cell RNA sequencing and analyzing the data using single cell regulatory network analysis – an
advanced computational framework designed to systematically measure the activity of every transcription
factor in each individual cell. This innovative new approach enables us both to identify specific cellular
subtypes impacted by treatment, and to rapidly identify the mechanistic drivers of observed responses. We will
deploy this approach on tumor samples from four different model systems: genetically engineered (KPC) mice,
human patient derived xenografts (PDXs), patient-derived tumor “explant” models (freshly cultured intact tissue
slices), and samples from an active clinical trial of MEK/autophagy inhibition in PDAC patients.
Extensive preliminary data in human cell lines, human PDAC explants, and KPC mouse tumors
demonstrate that MEK/autophagy inhibition primarily impacts specific subsets of three major cell types:
epithelial cells, fibroblasts, and macrophage/monocytes. Our three aims will study the effects and mechanisms
of MEK/autophagy inhibition in each cell type, with the goals of annotating phenotypes, learning whether
effects are cell autonomous, determining the transcription factors driving treatment response, understanding
impacts on the remaining cells, and identifying new therapeutic combinations to potentiate anticancer effects.