Mapping Tumor-Stroma-ECM Interactions in Spatially Advanced 3D Models of Pancreatic Cancer.

Bioengineering-based tumor models are increasingly important as tools for studying disease progression and therapy response for many cancers, including the deadly pancreatic ductal adenocarcinoma (PDAC) that exhibits a tumor/tissue microenvironment of high cellular/biochemical complexity. Therefore, it is crucial for models to capture that complexity and to enable investigation of the interplay between cancer cells and factors such as extracellular matrix (ECM) proteins or stroma cells. Using polyurethane (PU) scaffolds, we performed a systematic study on how different ECM protein scaffold coatings impact the long-term cell evolution in scaffolds containing only cancer or only stroma cells (activated stellate and endothelial cells). To investigate potential further changes in those biomarkers due to cancer-stroma interactions, we mapped their expression in dual/zonal scaffolds consisting of a cancer core and a stroma periphery, spatially mimicking the fibrotic/desmoplastic reaction in PDAC. In our single scaffolds, we observed that the protein coating affected the cancer cell spatial aggregation, matrix deposition, and biomarker upregulation in a cell-line-dependent manner. In single stroma scaffolds, different levels of fibrosis/desmoplasia in terms of ECM composition/quantity were generated depending on the ECM coating. When studying the evolution of cancer and stroma cells in our dual/zonal model, biomarkers linked to cell aggressiveness/invasiveness were further upregulated by both cancer and stroma cells as compared to single scaffold models. Collectively, our study advances the understanding of how different ECM proteins impact the long-term cell evolution in PU scaffolds. Our findings show that within our bioengineered models, we can stimulate the cells of the PDAC microenvironment to develop different levels of aggressiveness/invasiveness, as well as different levels of fibrosis. Furthermore, we highlight the importance of considering spatial complexity to map cell invasion. Our work contributes to the design of models with variable, yet biomimetic, tissue-like properties for studying the tumor microenvironment's role in cancer progression.