Transfer learning associates CAFs with EMT and inflammation in tumor cells in human tumors and organoid co-culture in pancreatic ductal adenocarcinoma [Bulk RNA-seq]

Pancreatic ductal adenocarcinoma (PDAC) is an aggressive malignancy characterized by an immunosuppressive tumor microenvironment enriched with cancer associated fibroblasts (CAFs). This study provides a convergence approach to identify tumor cell and CAF interactions through the integration of single-cell data from human tumors with human organoid co-culture. Analysis of a comprehensive atlas of PDAC single-cell RNA sequencing (scRNA-seq) data associates CAF density with increased inflammation and epithelial-mesenchymal transition (EMT) in epithelial cells. Transfer learning to transcriptional data from patient-derived organoid and CAF co-cultures provides in silico validation of CAF induction of inflammatory and EMT epithelial cell states. Further experimental validation in co-cultures demonstrates integrin beta 1 (ITGB1) and vascular endothelial factor A (VEGF-A) interactions with Neuropilin-1 (NRP1) mediating CAF and epithelial crosstalk. This study introduces transfer learning from human single-cell data to organoid co-culture for experimental validation of discoveries of cell-cell crosstalk, and identifies fibroblast-mediated regulation of EMT and inflammation. Patient matched organoids and cancer associated fibroblasts were generated from PDAC surgical speciments gathered at JHU. Samples were cultured in either monoculture or coculture conditions using a matrigel substrate. For bulk-RNA sequencing, samples were FACS sorted based on EPCAM (for organoids) and FAP (for CAFs) expression prior to RNA isolation. **Submitter declares that the raw data will be deposited in dbGaP due to patient privacy concerns***