Anti-tumor potential of combinatory GSK3 inhibition in human 3D models of pancreatic neuroendocrine tumors and patient-derived GEP-NET primary cultures.

Abstract

Activation of signaling pathways that regulate survival, proliferation, motility, inflammation, metabolism, and stemness fuel tumor growth, metastasis, and recurrence. Therapies targeting signaling pathway components, including candidates such as GSK3 and TNFα, drastically affect cellular viability in preclinical cancer models but have limited success in the clinic. However, in recent years, spheroids and organoids have been demonstrated to more accurately reflect tumor characteristics and to be better predictors of therapeutic response than monolayer cultures. Here, we used 3D models from the pancreatic neuroendocrine tumor (pNET) model BON1 to evaluate the effect of GSK3 inhibition along with TNFα or insulin and extended our results in primary gastroenteropancreatic (GEP-)NET culture. The multidimensional configuration of BON1 spheroids imparted aggressive characteristics and a lack of anti-proliferative effects upon single treatments. However, GSK3 inhibition alone resulted in dispersion of spheroids, indicating that GSK3 is necessary for cell-cell adhesions and participates in spheroid architecture. Interestingly, GSK3 inhibition in combination with TNFα or insulin led to drastically reduced cell proliferation. In fresh patient-derived 2D primary cultures from (GEP-)NETs, we demonstrate that insulin has tumor-promoting effects, while GSK3 inhibition and metformin display significant anti-tumor activity mediated through common effects on GSK3/insulin signaling. Both agents show strong efficacy in a patient-derived insulinoma without affecting the corresponding normal pancreatic tissue. We conclude that treatment efficacy depends on three-dimensional architecture and that combinatorial treatments which target cellular dispersion in addition to cellular viability might have beneficial clinical applications, but metastatic potential of remaining single cells needs further characterization before clinical implementation.