Cannabidiol-induced crosstalk of apoptosis and macroautophagy in colorectal cancer cells involves p53 and Hsp70

Abstract

Although it has been established that cannabidiol (CBD), the major non-psychoactive constituent of cannabis, exerts antitumoral activities, the exact mechanism(s) via which tumor cells are killed by CBD are not well understood. This study provides new insights into the potential mechanisms of CBD-induced mutual antagonism of apoptosis and macroautophagy using wild type (HCT116 p53wt, LS174T p53wt), knockout (HCT116 p53^-/-) and mutant (SW480 p53mut) human colorectal cancer cells (CRC). CBD causes a more pronounced loss in the viability of p53wt cells than p53^-/- and p53mut cells, and a 5-week treatment with CBD reduced the volume of HCT116 p53wt xenografts in mice, but had no effect on the volume of HCT116 p53^-/- tumors. Mechanistically, we demonstrate that CBD only significantly elevates ROS production in cells harboring wild-type p53 (HCT116, LS174T) and that this is associated with an accumulation of PARP1. CBD-induced elevated ROS levels trigger G0/G1 cell cycle arrest, a reduction in CDK2, a p53-dependent caspase-8/9/3 activation and macroautophagy in p53wt cells. The ROS-induced macroautophagy which promotes the activation of keap1/Nrf2 pathway might be positively regulated by p53wt, since inhibition of p53 by pifithrin-α further attenuates autophagy after CBD treatment. Interestingly, an inhibition of heat shock protein 70 (Hsp70) expression significantly enhances caspase-3 mediated programmed cell death in p53wt cells, whereas autophagy-which is associated with a nuclear translocation of Nrf2-was blocked. Taken together, our results demonstrate an intricate interplay between apoptosis and macroautophagy in CBD-treated colorectal cancer cells, which is regulated by the complex interactions of p53wt and Hsp70.

Fig. 1. Screening of potential targets of CBD based on the target gene enrichment assay.

a The Venn diagram identifies overlapping target genes affected in CRC after CBD treatment. b Hierarchy of 124 overlapping target genes acquired from the Venn analysis based on the p values of the Pathway and Process Enrichment Analysis. c Protein–protein interaction network and identified MCODE components. d List of most affected genes identified by the TRRUST database.

Fig. 2. CBD reduces the viability of p53wt CRC cells in vitro and inhibits tumor growth in vivo.

a, b Cell viability was determined using CCK8 assay kit 24 h (left) or 48 h (right) after CBD treatment. Data are expressed as the percentage of cell viability compared to control. Statistical differences of HCT116 p53^−/− was evaluated by compared to HCT116 p53wt (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) and LS174T compared to SW480 (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). Two-way ANOVA was used for analysis. Results represent the mean values of three independent experiments (n = 3). c Schematic diagram of workflow for data in vivo. d–g HCT116 cells were subcutaneously implanted into SCID mice which received vehicle alone or CBD (20 mg/kg) intraperitoneally 5 times per week over 5 weeks. Tumor size (d, e) were measured twice per week. f Representative examples of tumors treated either with vehicle or CBD are illustrated. g The hematoxylin and eosin (H&E) stained sections for observation of morphological changes in cells and immunohistochemistry of mice tumor tissue for detection of cleaved-caspase 3 to identify apoptotic cells directly after CBD treatment (Scale bar: 100 μm). h Immunoblot analysis of cytosolic PARP1 and cleaved-PARP1 expression in HCT116 p53wt and HCT116 p53^−/− cells 24 h after CBD treatment. i TUNEL assay has been performed to detect apoptotic cells after CBD treatment using a colorimetric TUNEL system. t-Test was used. Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). All data are expressed as the mean ± SD of three independent experiments.

Fig. 3. CBD induces G0/G1 cell cycle arrest.

a Flow cytometry was used to determine the cell cycle distribution in CBD-treated CRC cells. The quantification of each cell cycle phase is shown in the adjacent bar chart. b The expression levels of CDK2 and p21 protein in CBD-treated CRC cells were detected by Western blotting. c–f The ratio of protein levels was normalized to the values of the control. g HCT116 p53wt and p53^−/− cells stained for the expression of p53 were analyzed by confocal microscopy (Scale bar: 25 μm). Quantification of mean fluorescence intensity per cell was presented as adjacent bar charts. p53(green). Nuclei, DAPI (blue). Statistical differences of each group were evaluated by compared to the control group (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001). t-Test was used. Results are shown as mean ± SEM (HCT116 p53wt n = 36–43, HCT116 p53−/− n = 53–72). h After a 24 h CBD treatment, intracellular reactive oxygen species (ROS) levels were determined using the DCFDA assay. One-way ANOVA analysis was used. Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). All data are expressed as the mean ± SD of three independent experiments.

Fig. 4. The Hsp70 inhibitor PES-CI potentiates the antitumor effect of CBD.

a Viability of CRC cells treated with PES-CI. b CBD-induced reduction in mitochondrial membrane potential (MMP) (%) in HCT116 cells. c Apoptosis determined by Annexin V/PI staining. d, e Percentage of apoptotic CRCs (Annexin V positive: UR quad% + LR quad%). f, g Percentage of early apoptotic Annexin V positively stained CRCs (LR quad%). Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). One-way ANOVA, two-way ANOVA analysis or t-test was used. All data are expressed as the mean ± SD of three independent experiments.

Fig. 5. Hsp70 inhibition enhances the p53-dependent cleaved caspase-8/9/3 pathway.

a Flow cytometry was used to monitor ROS generation in both p53wt and p53^−/− HCT116 cells by staining with DCF-DA after CBD (15 µM) treatment in combination with PES-CI (IC50 value accordingly) or ROS scavenger NAC (2.5 mM). b–d Cleaved caspase-8/9/3 levels as measured by flow cytometry. e The p53-dependent activation of caspase-3 in LS174T and SW480 cells after treatment with CBD and PES-CI. f NAC attenuated apoptosis induced by CBD in HCT116 p53wt cells. Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). One-way ANOVA, two-way ANOVA or t-test was used. All data are expressed as the mean ± SD of three independent experiments.

Fig. 6. Protective autophagy induced by elevated ROS levels after CBD treatment in HCT116 p53wt cells.

a Immunoblot of cytosolic Hsp70, p62 and LC3BII expression levels in HCT116 p53wt and HCT116 p53^−/− cells after CBD treatment (15 µΜ) and a co-treatment with the Hsp70 inhibitor PES-CI (IC₅₀ value correspondingly) and/or the ROS scavenger NAC (2.5 mM). Adjacent bar charts show the quantification of Hsp70, p62 and LC3BII expression upon a combined treatment with the different reagents in HCT116 p53wt and p53^−/− cells. b Representative immunoblot showing the expression of intracellular p62 and LC3B 24 h after co-incubation with BafA1 (50 nM), a quantification of the p62 and LC3BII expression level are shown in the adjacent bar chart. c Quantification of LC3B vesicles using confocal fluorescence microscopy. (Left) Exemplary confocal images of LC3B (green) expressing HCT116 p53wt and HCT116 p53^−/− cells. Nuclei, DAPI (blue). Scale bar: 100 µm. (Right) Quantification of the LC3B vesicles per cell of HCT116 cells. Statistical differences of each group were evaluated by compared to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). One-way ANOVA was used. Results are shown as mean ± SEM (HCT116 p53wt n = 20–85, HCT116 p53^−/− n = 9–80). d The amount of p62 and LC3B expression in CBD-treated HCT116 p53wt cells with or without p53 inhibitor pifithrin-α (20 uM) and BarfA1. The relative expression levels of p62 and LC3B are shown in the adjacent bar graph. e CBD-induced cytoprotective autophagy is associated with a reduced cell viability following co-treatment with BafA1, as determined with the CCK-8 viability assay in HCT116 p53wt cells. Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). One-way ANOVA was used. All data are expressed as the mean ± SD of at least three biological replicates.

Fig. 7. p53-associated oxidative stress activates the keap1-Nrf2 pathway.

a Immunoblot of keap1 and Actin. A quantification of the keap1: actin ratio is shown in the adjacent bar chart. b, c Inhibition of Nrf2 by ML385 (5 uM, 24 h) enhances the antitumoral effect of CBD via an increase in apoptosis in p53wt cells, whereas only a moderate effect is observed for p53 knockout cells. Statistical differences of each group were evaluated by comparing the control (*p ≤ 0.05, **p ≤ 0.01 and ***p ≤ 0.001) or to the other group (^#p ≤ 0.05, ^##p ≤ 0.01 and ^###p ≤ 0.001). One-way ANOVA was used. All data are representative of three independent experiments. d HCT116 p53wt and p53^−/− cells stained for Nrf2 after different treatments as determined by confocal microscopy (Scale bars: 25 μm).

Fig. 8. Potential mechanism of CBD in antitumor effect.

Both the protective autophagy route and the programmed cell death pathway are activated in response to mitochondrial malfunction and ROS overproduction, which are caused by p53 nuclear translocation after CBD treatment. Hsp70-mediated autophagy degrades keap1, consequently, unbound Nrf2 is released for nuclear translocation and blocks p53-regulated apoptosis.