Reversion of chemoresistance by endocannabinoid-induced ER stress and autophagy activation in ovarian cancer

Abstract

The difficulty of detection at an early stage and the ease of developing resistance to chemotherapy render ovarian cancer (OVC) difficult to cure. Although several novel cancer therapies have been developed recently, drug resistance remains a concern since chemotherapy remains as the most commonly used treatment for cancer patients. Therefore, there is an urgent need to reclaim potential combination treatments for OVC. So far, there have been several research targeting the endocannabinoid system (ECS) in cancer. Among the various cannabinoid-based drugs, endocannabinoids, which are lipid molecules generated in the body, have been reported to produce many anti-tumor effects; however, research investigating the anti-chemoresistance effect of endocannabinoids in OVC remains unclear. In this study, we aimed to combine endocannabinoids, anandamide (AEA), and 2-arachidonoylglycerol (2-AG) with chemotherapeutic drugs as a combination approach to treat OVC. Our results showed that OVC cells expressed both cannabinoid receptors (CBR), CB1 and CB2, suggesting the possibility of endocannabinoid system (ECS) as a target. We found that the anti-chemoresistance effect mediated by endocannabinoids was caused by upregulation of ceramide levels, leading to severe endoplasmic reticulum (ER) stress and increased autophagy in chemoresistant cancer cells. Therefore, chemoresistant cancer cell growth was inhibited, and cell apoptosis was induced under combined treatments. Based on our results, endocannabinoids overcomed chemoresistance of OVC cells in vitro. Our findings suggest that drugs targeting ECS may have the potential to be adjuvants for chemotherapy by increasing the efficacy of chemotherapeutic drugs and decreasing their side effects.

Keywords: ER stress; Ovarian cancer; autophagy; chemoresistance; endocannabinoid system.

Figure 1

Existence of CB1 and CB2 in OVC. A. Kaplan-Meier plots showed the high and low expression levels of CNR1 and CNR2 genes in OVC cohort study. Log rank p-value and hazard ratio (HR) are shown in the figures. B. Representative images of CB1 or CB2 immunoreactivity in ovarian tissues. Images of serum control were presented on the left panel. In the middle and the right panels, brown color intensity represented the expression of CB1 and CB2. Enlarged images were shown in the lower right corner. Dotted lines displayed the location of normal epithelium. Scale bars, 100 μm. C. Representative images taken by confocal microscopy showed that both ES2 and IGROV1 cell lines expressed CB1 and CB2 (green). Nucleus was stained by Hoechst dye (blue). D. Comparison of CB1 and CB2 Western blotting between cancer cell lines (ES2 and IGROV1) and normal cell lines (HOSE11-12, HOSE6-3, IOSE398). GAPDH was served as the internal control. E, F. CNR1 and CNR2 mRNA expression levels in OVC cells. The Ct value of CNR1 and CNR2 was compared to GAPDH mRNA level in each cell line, and the comparative Ct value of CNR1 and CNR2 in chemoresistant cells was normalized with wild-type cells. The Y axis showed the fold change of CNR1 and CNR2 mRNA expression levels between wild-type and chemoresistant cells. The experiments were repeated for at least three times. Bars represent mean ± SEM. WT: wild-type, non-chemoresistant cell line, CP: cisplatin-resistant cell line, TX: paclitaxel-resistant cell line.

Figure 2

Cell toxicity of endocannabinoids and chemotherapeutic drugs on OVC cell growth. A, B. AEA and 2-AG were treated for 48 h at 0-100 μM on ES2 and IGROV1 cell lines. C and D. Cisplatin (0-40 μM) and paclitaxel (0-0.2 μg/mL) were treated for 48 h on both ES2 and IGROV1 cell lines. E-H. Different concentration of cisplatin or paclitaxel were combined with endocannabinoids to treat both ES2 and IGROV1 cell lines. In the left and middle images, red solid line represents CP- or TX-resistant cell lines. Yellow/blue or green/purple solid lines represent the addition of IC₃₀ or IC₅₀ values of AEA or 2-AG respectively on cisplatin- or paclitaxel-resistant cell lines. Gray dotted line represents wild-type cell line. In the right panels, IC₅₀ values of cisplatin or paclitaxel were analyzed and compared between wild-type or chemoresistant cell lines. Black asterisk symbolized the IC₅₀ comparison between wild-type and resistant cell lines, while red asterisk represented the IC₅₀ comparison between the treatment of chemotherapeutic drugs alone and combined treatments with several conditions in resistant cell lines. ***P<0.001 by one-way ANOVA. All experiments were repeated for three independent experiments. Bars represent mean ± SEM.

Figure 3

Endocannabinoids up-regulate ceramide levels and induce ER stress in chemoresistant cells. A. Chemoresistant cancer cells were pre-treated with 2 μM myriocin for 1 h and were incubated with IC₃₀ of AEA or 2-AG for 24 h. Antibody against ceramide was used, and the nucleus was stained by Hoechst dye (blue). Representative images were shown, and all images were taken by confocal microscopy. B-E. The fold change of ceramide intensity was quantified. The results were analyzed from three independent experiments. F-I. Wild-type and cisplatin-resistant cancer cells were pre-treated with 2 μM myriocin for 1 h and were incubated IC₃₀ of AEA or 2-AG for 48 h. Antibodies against Grp78 and CHOP were used for immunoblotting analysis. β-actin was served as the internal control. Arrowhead indicates the position of CHOP proteins on the immunoblots. Bars represent mean ± SEM. *P<0.05, **P<0.01, ***P<0.001 by one-way ANOVA.

Figure 4

Activation of ER stress in chemoresistant cancer cells under the treatments of endocannabinoids and chemotherapeutic drugs. A-D. Cisplatin-resistant ES2 and IGROV1 cell lines were treated with 2 mM cisplatin and IC₃₀ of AEA and 2-AG for 12, 24 and 48 h. Antibodies against ER stress-related proteins, including Grp78, ATF6α, phospho-IRE1α (p-IRE1α), IRE1α, phospho-eIF2α (p-eIF2α) and eIF2α were used in the immunoblotting analysis. 10 μg/mL tunicamycin was served as positive control (+) of ER stress. β-actin was served as the internal control. FL: full-length form, CL: cleaved form.

Figure 5

Basal level autophagy in OVC cells and enhancement of autophagy under combination of endocannabinoids and chemotherapeutic drugs. (A, B) Representative LC3 immunoblots and quantification of relative protein level of LC3-II from 3 independent experiments were shown. The protein level of LC3-II was divided by control group in each cell lines. Cells were treated with 5 and 10 μM CQ for 24 h. Actin was served as internal control. Bars represent mean ± SD. (C, D) 10 μM cisplatin, 0.01 μg/mL (for ES2-TX) or 0.1 μg/mL (for IGROV1-TX) paclitaxel and IC₃₀ of AEA or 2-AG were used for treating cisplatin- or paclitaxel-resistant ES2 and IGROV1 cell lines. All drugs were incubated for 48 h. Autophagy was detected by LC3 antibody (green), and the nucleus was stained by Hoechst dye (blue). The images were taken by confocal microscopy. (E, F) Quantification of LC3 puncta per cell from (A and B) was analyzed. The experiments were repeated for three independent experiments. Bars represent mean ± SEM. (G-J) Western blot analysis of LC3 in ES2 and IGROV1 cell lines treated with AEA/2-AG and cisplatin/paclitaxel for 36 h. *P<0.05, **P<0.01, ***P<0.001 by one-way ANOVA.

Figure 6

Endocannabinoids combined with chemotherapeutic drugs induce apoptosis in chemoresistant cancer cells. A. Chemoresistant cancer cells were treated with 2 mM cisplatin, 0.01 μg/mL (for ES2-TX) or 0.1 μg/mL (for IGROV1-TX) paclitaxel and IC₃₀ of AEA or 2-AG for 48 h. Annexin V and propidium iodide were used to quantify apoptosis. Representative images of flow cytometry were shown. B-E. Quantification of the percentage of apoptotic cells was analyzed from three independent experiments. The bars represent mean ± SEM. F. Schematic diagram shows the mechanism about how endocannabinoids reverse chemoresistance in OVC. *P<0.05, **P<0.01, ***P<0.001 by one-way ANOVA.