Metformin sensitizes anticancer effect of dasatinib in head and neck squamous cell carcinoma cells through AMPK-dependent ER stress

Abstract

Head and neck squamous cell carcinoma (HNSCC) is an important endemic disease in Taiwan with aggressive course and dismal outcome. Dasatinib is a Bcr-bl and Src kinase inhibitor that has potential against HNSCC. We recently disclosed that EGFR degradation is critical for dasatinib-induced apoptosis. Here, we further demonstrate that AMPK-dependent ER stress is responsible for this event. Dasatinib induced ER stress which mediated EGFR degradation in a c-cbl-dependent manner. AMPK activation induced by dasatinib might be due to ATP decrease through the up-regulation of pyruvate dehydrogenase kinase 4 (PDK4). Furthermore, activation of AMPK by metformin sensitized dasatinib-induced in vitro and in vivo anti-cancer effect. The correlation of AMPK activation and EGFR expression was seen in HNSCC cells and human tumor specimens. Our results disclose that AMPK-dependent ER stress plays a crucial role in the anti-cancer effect of dasatinib in HNSCC and further activation of AMPK by metformin might enhance dasatinib efficacy.

Figure 1. Dasatinib induced ER stress and EGFR degradation in HNSCC cells

(A) The effect of dasatinib on ER stress and EGFR expression. The expression of EGFR, p-eIF2α, eIF2α, and CHOP was evaluated in HNSCC cells treated with dasatinib (1uM) for indicated time. (B) The effect of 4-phenyl butyric acid (PBA) on dasatinib-induced EGFR degradation and apoptosis in Ca9-22 (left) or HSC3 (right) cells. Upper, SubG1 analysis of HNSCC cells treated with indicated drugs for 48hrs. Column, mean (n=3); bar, standard deviation; *, p<0.05 by paired Student's t-test. Lower, the expression of EGFR after treated with indicated drugs for 24hrs. (C) The effect of PERK knockdown on dasatinib-induced EGFR degradation. Cells were transfected with control or EGFR siRNA for 72 hours and then treated with dasatinib (1uM) for 24 hours. (D) The effect of ER stress on EGFR expression. Cells were treated with brefeldin-A (5ug/ ml) or tunicamycin (2ug/ml) for indicated time. The expression of proteins was evaluated by Western blotting. Representative of three independent experiments was shown.

Figure 2. C-cbl-lysosome pathway mediated ER stress-induced EGFR degeradation

(A) The effect of ER stress on c-cbl activation. Cells were treated with brefeldin-A (5ug/ml) or tunicamycin (2ug/ml) for indicated time. The expression of p-c-cbl, and c-cbl was evaluated. (B) The effect of c-cbl knockdown on brefeldin-A-induced EGFR down-regulation. Cells were treated with control or c-cbl siRNA and then treated with brefeldin-A for 24 hours. The expression of EGFR and c-cbl was evaluated. (C) The effect of ER stress on the association of c-cbl and EGFR. Cells were transfected with c-cbl-HA plasmid and treated with brefeldin-A for 2 hours. Cell lysates were immunoprecipitated with anti-HA antibodies. The immunoprecipitates were blotted with EGFR and c-cbl. (D) The effect of lysosome inhibitor NH₄Cl on brefeldin-A or tunicamycin-induced EGFR down-regulation. Cells were treated with indicated drugs for 24 hours, and the expression of EGFR was evaluated. (E,F) The localization of EGFR and c-cbl under ER stress. HSC3 cells were co-transfected with EGFR-RFP and c-cbl-GFP. The localization of EGFR (red) and c-cbl (green) of live cells treated with brefeldin-A (E) or dasatinib (F) was recorded by time-lapse confocal microscopy (upper panel), and the co-localized signal was pseudo-colored in yellow (lane 4). The ratio of co-localization was calculated by Zen software (Carl Zeiss) (lower panel)

Figure 3. AMPK activation mediated dasatinib--induced ER stress and EGFR degradation

(A) The effect of dasatinib on AMPK activation. Cells were treated with dasatinib (1uM) for indicated intervals. The expression of p-AMPK and AMPK was evaluated. (B) The effect of AMPK knockdown on dasatinib-induced EGFR degradation and ER stress. Cells were treated with control or AMPK siRNA and then with dasatinib for 24 hours. (C) The effect of AMPK activation on dasatinib-induced EGFR degradation. Cells were treated with dasatinib with or without AICAR (10uM) for 24 hours. The expression of EGFR p-eIF2α, and AMPK was evaluated. (D) The correlation between p-AMPK and EGFR expression. Left, the expression of EGFR, p-AMPK, and AMPK in HNSCC cells. Right, the correlation of p-AMPK and EGFR expression in resected human specimens. Pearson's correlation coeffcient=0.659; *, p<0.01

Figure 4. Dasatinib induced cellular ATP decrease and PDK4 up--regulation

(A,B) The effect of 6-hr or 18-hr dasatinib (1uM) on cellular ATP (A) and glucose (B) levels. *, p<0.05. (C) The expression of PDK4 and p-Erk in HNSCC cells treated with dasatinib (1uM) for indicated intervals.

Figure 5. Metformin enhanced anti--cancer effect of dasatinib

(A) The effect of dasatinib in combination with metformin for 48 hours at indicated doses on cellular growth inhibition. Cells were treated with dasatinib and metformin at a molar ratio of 1:10000. Growth inhibition was evaluated by MTT assay (upper panel). Dot, mean (n=3); bar, standard deviation. The combination index (CI) was calculated by median dose analysis (lower panel). CI smaller than one indicated synergism between dasatinib and metformin. (B) Metformin enhanced dasatinib-induced apoptosis. Cells were treated with dasatinib (1uM) in combination with metformin (10mM) for 48 hrs. Apoptosis was evaluated by subG1 analysis of fow cytometry. Column, mean (n=3); bar, standard deviation; *, p<0.05.

Figure 6. Figure 6: In vivo effect of metformin in combination with dasatinib in sensitive HSC3 tumor (A) and resistant SAS tumors (B)

Upper panel, growth curve of xenograft tumors treated with indicated drugs. Dot, mean (n=8); bars, standard error; *, p < 0.05 by unpaired Student's t-test. Lower panel, the expression of EGFR, p-AMPK and p-eIF2α of HSC3 (A) or SAS (B) tumors. The representative three tumors from each treatment group at the end of experiments were analyzed by Western blotting. (C) Schematic illustration of dasatinib-induced AMPK activation, ER stress, and EGFR degradation.