Although genetically engineered mouse models that recapitulate multiple aspects of human histology have advanced our understanding of PDAC biology, little progress has been made in the treatment of PDAC. This striking lack in translation requires the development of novel experimental systems that reflect both heterogeneity and plasticity and recapitulate treatment efficacy ex vivo and reveal the molecular basis of individual responses.

A highly promising novel model system to fulfill these needs of translational pancreatic cancer research are PDAC patient-derived organoids (PDOs) (Dutta et al., 2017). Cells isolated from surgical specimens or endoscopic biopsies are cultured in a three-dimensional matrix where they give rise to cystic multicellular structures with a hollow lumen, thereby preserving crucial aspects of tumor morphology. These PDOs are generated in a culture system where the selection pressure is low, allowing multiple subclones of the tumor to be represented within the culture. For comparative studies and as controls, normal tissue can also be cultured as organoids in this assay. The Reichert laboratory has recently derived more than 52 PDO lines, thereby providing an unprecedented representation of inter-tumor heterogeneity (also see S01). Thus, in contrast to traditional human and mouse model systems, this new, dynamic and constantly expanding experimental system is a powerful tool to address the current bottleneck in translational PDAC research.

Epithelial-to-mesenchymal transition (EMT) and the reverse process, mesenchymal-to-epithelial transition (MET), are important aspects of epithelial plasticity and play a major role in conferring intra- and intertumoral heterogeneity. The dynamic regulation of EMT and MET programs poses one of the greatest challenges in effective targeting of PDAC: epithelial plasticity is critical for cancer cells to acquire necessary traits to go through the metastatic cascade as well as to convey resistance to therapy.

For this proposal, we will combine the power of the PDO experimental system to generate a plasticity score, dissect the molecular basis of plasticity and then target the cellular programs regulating it. For this purpose, we will utilize a “plasticity gene signature” containing 76 transcripts that was derived by Dr. Reichert using a system biology approach comparing distinct biological processes involving epithelial plasticity, namely pancreatic embryonic development, inflammation, carcinogenesis (Reichert et al., 2013b) (Reichert et al., 2013a). As a functional in vitro readout, the Scheel lab has developed a unique high-throughput-compatible assay to quantify epithelial plasticity that was adopted from primary human mammary epithelial cells (Linnemann et al., 2015; Schmidt et al., 2015). In a collaborative effort, we will utilize PDOs to perform genetic and drug screens directed against plasticity drivers. Druggable plasticity drivers have already been identified. Candidates will be genetically and pharmacologically validated in PDO-derived xenografts (PDOXs). In addition, newly identified drugs will be tested in combination with established chemotherapeutics to evaluate therapy-sensitizing effects of plasticity-inhibition.

Targeting the epithelial plasticity program provides a promising strategy to arrest the metastatic cascade and to overcome chemotherapeutic resistance.