Abstract B003: Developing an all-human model of perineural invasion in pancreatic cancer

Abstract

Perineural invasion (PNI), the infiltration of nerves by tumor cells, is a hallmark of pancreatic ductal adenocarcinoma (PDAC); however, despite being a clinically relevant phenomenon, the mechanisms underlying PNI remain poorly understood. Most currently available in vitro and ex vivo models rely on murine-sourced nerve tissue; while these models are invaluable for investigating nerves as an understudied component of the tumor microenvironment, there remains a critical need for a relevant all-human model to advance studies of tumor-nerve crosstalk. To meet this gap in the field, we developed an all-human model of perineural invasion to identify components of the neural niche and tumor microenvironment that drive PNI in pancreatic cancer. We have established a collaboration with Gift of Life Michigan to utilize organ-donor derived dorsal root ganglion (DRG) and celiac ganglion (CG) to source nerves for our model. As an alternative, scalable, complementary source of nerves, we have leveraged a doxycycline inducible “i-Neuron” system that allows for the rapid differentiation of inducible pluripotent stem cells into neurons. Using labelled patient-derived tumor cells and cancer-associated fibroblasts (CAFs) for co-culture with neurons, we tracked tumor cell migration and innervation. We observed tumor cell innervation and migration towards DRGs in these co-cultures as well as differences in the amount of innervation exhibited across different patient-derived tumor cells. Furthermore, we noted the extensive migration of CAFs when cultured with neurons. Additional computational analysis of RNA sequencing data highlighted the impact of CAFs when co-cultured with tumor cells and neurons: compared to tumor cells or CAFs alone, neurons exhibited the greatest upregulation of inflammatory markers and expressed markers of neuronal injury. In summary, we have observed that donor-derived nerve ganglia and iPSC-derived neurons can be leveraged to generate all-human models of perineural invasion. Further studies are required to determine the role of CAFs in tumor cell invasion and proliferation as well as why innervation differs across tumors from multiple patients. Nevertheless, our advances towards building an all-human model for perineural invasion in pancreatic cancer appear promising.