Fig 2: Specificity of pancreatic ductal adenocarcinoma–driven β-cell depletion.(A, B, C, D) Formalin-fixed paraffin-embedded sections from control, pancreatitis (PPT), HFD-fed (HFD), or KC mice (n = 6) were immunostained with antibodies to insulin, glucagon, somatostatin, or PP and revealed by immunofluorescence. Representative pictures of normal or remnant islets are shown (left). INS+, GCG+, SST+, or PP+ cells in all islets from six different sections were counted, and results are presented as mean of the total number of hormone-positive cells per islet (right). Bar, 200 μM. Data are expressed as dot plot with a line at the median and whiskers showing SD. Statistical significance was estimated by two-way ANOVA. *P < 0.05; ***P < 0.001; ns, nonsignificant.

Fig 3: High-fat diet (HFD) induces intra-islet glucagon-like peptide-1 (GLP-1) system activation. (A) Representative images of immunofluorescence staining for insulin (green) and prohormone convertase 1/3 (PC1/3) (red) in the pancreatic islets of mice that were fed a low-fat diet (LFD) (n=6) or a HFD (n=7) for 8 weeks. The arrows indicate the insulin-negative and PC1/3-positive cells. (B) PC1/3 mRNA expression in the pancreata of mice from each group was detected by qPCR. (C) Plasma GLP-1 levels in the mice were detected by insulin enzyme-linked immunosorbent assay (ELISA). LFD compared with HFD; **p<0.01; ***p<0.001.

Fig 4: Prolonged exposure to palmitate induces the activation of the glucagon-like peptide-1 (GLP-1) system in isolated mouse islets. After the islet cells were incubated with 0.5 mmol/l palmitate for the indicated periods of time, the GLP-1 levels in the (A) cell medium and (B) lysates were determined by insulin enzyme-linked immunosorbent assay (ELISA), and (C) prohormone convertase 1/3 (PC1/3) mRNA expression in cell lysates was determined by qPCR and (D) the PC1/3 protein level was determined by immunoblot anlaysis. n=3 separate islet preparations; **p<0.01; ***p<0.001.

Fig 5: Inactivation of TGF-β signaling suppresses pancreatic ductal adenocarcinoma (PDAC)–driven β-cell depletion.(A) Formalin-fixed paraffin-embedded (FFPE) sections from control, KC, or KSC mice were stained with H&E or subjected to immunohistochemistry analysis using antibodies to CK19 or JunB. Representative pictures of normal or PDAC tissues are shown. Bar, 200 μM. (B) FFPE sections from control, KC, or KSC mice (n = 6) were immunostained with anti-insulin antibody and revealed by immunofluorescence (IF). Then, INS+ cells in all islets from six different sections were counted, and results are presented as percentage of INS+ cells relative to the total cell number in islets. Bar, 200 μM. (C) FFPE sections from control, KC, or KSC mice (n = 6) were immunostained with anti-glucagon antibody and revealed by IF. Representative pictures of normal or remnant islets are shown (left). Percentage of remnant islets is shown. GCG+ cells in all islets from six different sections were counted, and results are presented as percentage of GCG+ relative to the total cell number in islets (middle) or as mean of the total number of GCG+ cells per islet cells (right). Bar, 200 μM. (D) FFPE sections from control, KC, or KTC mice were stained with H&E or immunostained with antibodies to CK19 or JunB and revealed by immunohistochemistry. Representative pictures of normal or PDAC tissues are shown. (E, F, G) FFPE sections from control, KC, or KTC mice (n = 6) were immunostained with antibodies to insulin or glucagon and revealed by IF. (E, F, G) Then, INS+ or GCG+ cells in all islets from six different sections were counted, and results are presented as percentage of INS+ or GCG+ cells relative to the total cell numbers in islets (E, F) or as mean of the total number of GCG+ cells per islet cells (G). For (B, C [middle], E, G), data are expressed as mean ± SEM. For (C [right], F), data are expressed as dot plot with a line at the median and whiskers showing SD. For all graphs, statistical significance was estimated by one-way ANOVA. ***P < 0.001; ns, nonsignificant.