Article Figures & Data
Figures
- Figure 1.
A, continued smoking reduces overall survival in patients with resected pancreatic cancer. Patients with pancreatic adenocarcinoma were followed postoperatively after pancreatic resection and a multivariate Cox proportional hazards model was used to evaluate the effects of continued tobacco abuse on overall survival. Continued smoking correlated with reduced survival in pancreatic cancer (HR 1.93; P = 0.040). P < 0.05 was considered statistically significant. B, TAS induces gemcitabine chemoresistance and Id1 expression in pancreatic cancer cells. TAS cells were cultured from human pancreatic tumors using the outgrowth method. Pancreatic cancer cells were cocultured with or without TAS in the presence or absence of gemcitabine at concentrations corresponding to LD50 values for each pancreatic cancer cell line. C, TAS coculture induces Id1 expression in pancreatic cancer cells. β-Actin was used as a loading control. D–F, TAS cells secrete HGF while pancreatic cancer cells express c-Met and Id1. D, qRT- pancreatic cancer R (left) and ELISA (right) were performed for HGF expression and secretion in TAS and pancreatic cancer cells. Data, mean ± SD of three independent experiments (*, P < 0.01 TAS vs. pancreatic cancer cell lines). E, protein lysates from TAS and pancreatic cancer cell lines indicated were harvested and probed for Id1 (right) and c-Met (left). β-Actin was used as a loading control. F, immunofluorescent stains viewed at 20× magnification were performed on cells in culture of the indicated cell type. c-Met is depicted here in red, while phalloidin staining appears in green. Phalloidin stains are displayed to indicate the characteristic stress fiber organization of TAS cells. Goat anti-mouse AF647-stained controls are displayed in the far right column. Scale bar, 10 μm.
- Figure 2.
Nicotine augments expression of HGF and c-Met in TAS and pancreatic cancer cells, respectively. A, TAS were stimulated with nicotine and RNA, supernatants or protein lysates were harvested for ELISA (left) and qRT-PCR (right) for analysis of HGF expression. B, pancreatic cancer cells were stimulated with nicotine and Western blot analysis of c-Met induction was performed. β-Actin was used as a loading control. C, c-Met activity was evaluated in pancreatic cancer cells alone and direct TAS/pancreatic cancer cell coculture at 48 hours with and without nicotine stimulation (1 μmol/L) and crizotinib treatment (200 nmol/L). Phospho-Met staining (top, green) was superimposed upon phase-contrast images with DAPI stains shown separately (bottom, blue). D, qRT-PCR was also utilized to evaluate the expression of nicotinic AchR subunits within the TAS in the absence of nicotine exposure. Data, mean ± SD of three independent experiments.
- Figure 3.
TAS-mediated induction of Id1 expression in pancreatic cancer cells is c-Met dependent. A (left to right), pancreatic cancer cell lines were transfected with c-Met or control siRNA and either exposed to TAS-conditioned media or media control. Protein lysates were harvested, and the expression of c-Met was evaluated by Western blot. B, HGF induces Id1 expression in pancreatic cancer cells. C (left to right), pancreatic cancer cell lines were transfected with c-Met or control siRNA and were exposed to HGF, and the expression of Id1 was evaluated by Western blot. D (left and right), crizotinib inhibits HGF induction of c-Met phosphorylation in pancreatic cancer cells. E (left and right), pancreatic cancer cell lines were exposed to c-Met inhibitor crizotinib, TAS-conditioned media, or combination. Protein lysates were harvested and probed for Id1 expression by Western blot. β-Actin was used as a loading control. Data, immunoblots of three independent experiments.
- Figure 4.
Nicotine augments tumor growth and metastasis in a patient-derived pancreatic adenocarcinoma xenograft model. A, primary human pancreatic cancer specimens and normal pancreas controls were subjected to IHC analysis of α5 and α7 nicotinic AchR subunit staining. Images are viewed at 10× magnification. B, animals with palpable patient-derived tumors were administered physiologic doses of nicotine or control. Tumor dimensions were measured with calipers 3 times per week, and volumes were calculated over 45 days. Data, mean ± SEM (*, P < 0.05 vs. control). C–E, xenograft lysates from nicotine-treated and control animals were subjected to qRT-PCR for HGF expression (C), ELISA (D), and Western blot analysis (E) for c-Met activation. F, two representative lungs from nicotine-treated and control animals were subjected to H&E analysis, viewed at 20× magnification.
- Figure 5.
Activated c-Met in pancreatic cancer specimens correlates with reduced postoperative survival. A, intratumoral c-Met activation was measured using a phospho-Met ELISA on tumor lysates, and representative samples were verified using IHC (insets). Kaplan–Meier survival curves were generated through stratification of expression or not of phospho-Met within the primary tumor of 26 patients with pancreatic adenocarcinoma for whom clinical follow-up was obtained. Significance was determined using log-rank analysis. B, a summary model is depicted by which nicotine stimulates HGF secretion in TAS while directly inducing c-Met expression in pancreatic cancer cells, thereby augmenting c-Met activation in a paracrine manner.
Tables
- Table 1.
Survival analysis
HR (95% CI) P Univariate analysis (n = 90) Operation performed (PD) DP (n = 14) 0.70 (0.44–1.73) 0.701 T stage (T1–2) T3–4 (n = 82) 2.08 (0.76–5.71) 0.157 N stage (N0) N1 (n = 63) 1.38 (0.82–2.32) 0.233 Positive lymph node ratio 2.99 (0.91–9.81) 0.071 Poor tumor differentiation (n = 46) 2.05 (1.26–3.36) 0.004a Smoking status (never smokers) Former smokers (n = 38) 1.12 (0.64–1.94) 0.694 Current smokers (n = 19) 1.78 (0.96–3.32) 0.070 Multivariate analysis (n = 90) T stage (T1–2) T3–4 (n = 82) 1.61 (0.56–4.61) 0.376 Positive lymph node ratio 2.99 (0.87–10.2) 0.081 Poor differentiation (n = 46) 2.03 (1.21–3.40) 0.007a Smoking status (never smokers) Former smokers (n = 38) 1.41 (0.80–2.50) 0.239 Current smokers (n = 19) 1.93 (1.03–3.63) 0.040a NOTE: Continued smoking is associated with reduced survival in patients with resected pancreatic cancer. Patients with a history of resected pancreatic adenocarcinoma and data regarding dose of tobacco exposure were followed postoperatively. A Cox proportional hazards model was used to analyze the effect of prognostic clinicopathologic variables as well as smoking status on overall survival. Positive lymph node ratio was defined as the number of lymph nodes positive for malignancy divided by the total number of lymph nodes examined. T stage was dichotomized due to a small sample size of T1–2 patients. Significance was considered for P < 0.05a.
Abbreviation: CI, confidence interval.
Supplementary Data
- Supplemental Figure S1 - Supplemental Figure S1. Induction of c-Met by nicotine occurs at the post-transcriptional level.
- Supplemental Figure S2 - Supplemental Figure S2. Nicotine-mediated induction of c-Met is Src-dependent.
- Supplemental Figure S3 - Nicotine treatment does not lead to c-Met activation in PC cells.
Volume 22, Issue 7
