L. Delrosario1, C. G. Flesher1, N. A. Baker1, R. W. O’Rourke1, T. Frankel1 1University Of Michigan,Department Of Surgery,Ann Arbor, MI, USA
Introduction: Type 2 diabetes mellitus (DM) is a risk factor for pancreatic ductal adenocarcinoma (PDAC) but mechanisms underlying this association remain poorly understood. Multiple reports implicate adipocytes and preadipocytes as putative promoters of PDAC in humans and mice. To test whether preadipocytes or adipocytes regulate PDAC tumorigenic functions in the context of DM, we explored the effects of conditioned media prepared from preadipocytes from nondiabetic (NDM) and diabetic patients on PDAC cells.
Methods: Preadipocytes were isolated from omental adipose tissue obtained from obese NDM and DM patients undergoing bariatric surgery. Serum-free conditioned media was prepared from these preadipocytes cultured in vitro for 72 hours. Sphere assays using T3M4 and Capan1 PDAC lines were performed in the presence or absence of conditioned media, and sphere surface area was measured over 2-3 weeks. Modifications of the sphere assays, including boiling of conditioned media and addition of neutralizing antibodies, was performed. A proteome profiler array was used to identify differences in chemokine expression between DM and NDM preadipocytes.
Results: Conditioned media from human DM preadipocytes, compared to media from NDM preadipocytes, increased PDAC cell sphere surface area in both T3M4 and Capan1 cell lines. Conditioned media from mature differentiated human adipocytes did not regulate PDAC cell sphere area. Sphere surface area correlated directly with serum hemoglobin A1c (HbA1c) but not gender or body mass index of the human subjects from which preadipocytes were derived. Boiling of preadipocyte supernatants abrogated their pro-proliferative effects, suggesting a peptide or protein mediator, and ultracentrifugation of preadipocyte supernatants localized activity to the non-exosome fraction. Proteome profiler array analysis identified different proteins that were differentially expressed by DM and NDM preadipocytes, among which chitinase-3-like protein 1 (CHI3L1) was increased over 2-fold in DM relative to NDM preadipocyte supernatants. Neutralizing antibody to human CHI3L1 attenuated the increase in PDAC sphere surface area in response to DM preadipocyte conditioned media in a dose-dependent manner.
Conclusion: Preadipocyte-PDAC cell crosstalk regulates PDAC cell proliferation in a DM-specific and CHI3L1-dependent manner. These results suggest that the positive clinical correlation between DM and PDAC may be mediated by preadipocytes expressing higher levels of CHI3L1. Preadipocytes and CHI3L1 represent targets for further research directed towards developing novel therapies for PDAC based on manipulation of the tumor microenvironment.