Diabetes and pancreatic cancer: role of advanced glycation endproducts (AGEs) in tumor progression via RAGE activation

Background & Aims: Diabetes mellitus (DM) has been identified as a risk factor for a variety of malignancies, including pancreatic cancer (PaC). Many factors may promote PaC in DM, including hyperglycemia itself, dyslipidemia, chronic inflammation and insulin resistance/hyperinsulinemia, but the mechanistic link is unknown. Previous studies have demonstrated a tumor-promoting role for RAGE, the receptor for advanced glycation endproducts (AGEs), which levels are increased in diabetes. Here we test the hypothesis of a promoting role of AGEs in PaC through RAGE activation. Methods: We investigated the effects of the AGE Nε-carboxymethyllysine (CML) in human pancreatic ductal adenocarcinoma (PDA) cell lines and in a mouse model of Kras-driven PaC interbred with a bioluminescent model of proliferation. PaC–prone KCM (LSL-KrasG12D/+;Pdx-1-Cre;MITO) mice, which spontaneously develop autochthonous lethal PaC, were generated by interbreeding mutated KCM transgenic mice with MITO-Luc reporter mice. In this model, the promoter of the luciferase reporter is under the control of the nuclear factor (NF)-Y, which is specifically expressed in proliferating cells. Therefore, this animal model gave us the chance to monitor the tumor growth in vivo by bioluminescence imaging. Results were confirmed by histology examination. In vitro results: In the range of 10-100 μg/mL, CML promoted cell growth and RAGE expression (~1,15 to 1,4 times, p<0.01) in a concentration- and time-dependent manner in two PDA cell lines. At variance, normal human pancreatic duct epithelial cells showed a positive response to CML only at the lowest concentration (1 μg/mL) tested. Interestingly, the different proliferative response to CML observed in different cell lines was positively associated with basal expression levels of RAGE: i.e., the higher was the basal expression of RAGE of the cell line, the larger was the range of CML concentrations capable of providing an additional proliferative stimulus at 24 h. Mechanistically, the activation of the AGE/RAGE axis in PDA cells increased NK-κB p65 protein level and activation (~1,5 and 3 times, respectively, p<0,01), STAT3 phosphorilation (~3 times, p<0,01), and the transcriptional activity of its targets genes PIM1 and NFATc1 (~3 and 1,5 times, respectively, p<0.01), which have been involved in PaC promotion. These effects were completely counteracted by RAP (RAGE antagonist peptide), except for PIM1 expression and STAT3 phosphorilation, which were only partially, but significantly (more than halved, p<0,01), reduced. In vivo results: Five-week old KCM mice were daily treated intraperitoneally (i.p.) with 30 μg of CML-modified (CML mice) or native mouse serum albumin (MSA, vehicle; Ctr mice). At 11 weeks of age (6 weeks of treatment), more than half of CML mice (six of eleven) showed positive abdominal imaging and about 1/4 (three of eleven) had a palpable abdominal mass. At the histological analysis, invasive PaC was confirmed in eight of eleven (72.7%) CML mice, compared with only one of eleven (9.1%, Fisher's exact test p