PROJECT SUMMARY The epithelial-mesenchymal transition (EMT) is a developmental process that is aberrantly reactivated in pancreatic ductal adenocarcinoma (PDAC) to promote disease progression and chemoresistance. PDAC tumors and cell lines typically contain a heterogeneous mixture of transformed cells displaying epithelial or mesenchymal characteristics, complicating efforts to understand the regulatory mechanisms that govern this important phenotypic switching. The observation that EMT can be initiated by a variety of different growth factors, low oxygen tension, and matrix-mediated signaling strongly suggests that multiple signaling pathways cooperate to drive robust EMT and raises the possibility that EMT heterogeneity is explained by the ability of only some cells to activate robustly the pathways that cooperate to drive EMT. Another potential, but not necessarily mutually exclusive, explanation for phenotypic heterogeneity is that some PDAC cells are primed to undergo EMT due to transcriptional differences that enable utilization of specific transcription factors or signaling pathways. Based on our preliminary data, we hypothesize that certain PDAC cells are transcriptionally primed to undergo EMT and that EMT heterogeneity further depends upon cell-to-cell variations in kinase-regulated signaling processes within cell populations. The objective of the work proposed here is to test these hypotheses through the development of quantitative systems biology methods to study the basis of EMT heterogeneity regulation via transcriptional and kinase-mediated signaling processes. In Aim 1, an iterative immunofluorescence imaging pipeline will be developed to gather multiplexed signaling data on populations of PDAC cells treated with different EMT agonists. Based on preliminary studies, we propose to measure markers for seven distinct signaling pathway nodes and two EMT markers to create a dataset with nine features measured for thousands of cells for each experimental condition. We will then apply a mutual information data science approach for the quantitative identification of the signaling pathways that cooperate to drive robust EMT. Model predictions will be tested using small molecule inhibitors and siRNA-mediated knockdowns. In Aim 2, we will use genetic barcoding for the transcriptomic profiling of EMT-resistant or -compliant lineages within cell populations. Single-cell RNA sequencing data from cells before and after EMT induction will be analyzed to identify transcriptional states that preferentially enable PDAC cells to undergo the mesenchymal transition. The relevance of candidate transcripts for explaining EMT priming will be tested through knockdown experiments. The methods developed in this work will be broadly applicable to the study of EMT in other cancer settings and to the study of alternative types of phenotypic switching. Moreover, the specific results will have implications for the design of novel combination therapies for PDAC ba...