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

p53-dependent reduction in viability and growth of CRC cells in vitro and in vivo

The viability of HCT116 p53wt, HCT116 p53^−/−, SW480 p53mut, LS174T p53wt cells after treatment with CBD (5–20 µM), as determined using the CCK-8 assay, was reduced in all cell lines in a dose-dependent manner after a 24 h (Fig. ​(Fig.2a)2a) and 48 h (Fig. ​(Fig.2b)2b) incubation. However, the sensitivity of cells to CBD was cell type dependent: HCT116 p53wt (IC50 = 14.67 µM; 95% CI 13.42–15.79), HCT116 p53^−/− (IC50 = 24.26 µM; 95% CI 21.33–34.88), LS174T p53wt (IC50 = 7.918 µM; 95% CI 7.039–8.867) and SW480 p53mut (IC50 = 16.58 µM; 95% CI 14.73–19.17) cells. The data revealed that p53 deficient (HCT116 p53^−/−) and p53 mutant (SW480 p53mut) cells are less sensitive to a CBD treatment than p53 wild-type cells when compared to each other (HCT116 p53wt vs. HCT116 p53^−/−, LS174T p53wt vs. SW480) (Fig. ​(Fig.2a).2a). However, a significant difference between HCT116 p53wt and SW480 p53mut was observed only with a CBD concentration of 20 µM and an incubation period of 24 h (p < 0.0001) (Fig. 2a, b). Unlike p53 deletion (HCT116 p53^−/−), the two-point mutations (R273H/P309S) in p53 (SW480) triggers a diverse sensitivity to CBD which is probably due to a partially retained p53 function [17]. Based on the IC50 values, a concentration of 15 µM was used as the maximum concentration for in vitro experiments with HCT116 p53wt and p53−/− cells, while LS174T and SW480 cells were treated with 20 µM CBD to enable a better pairwise comparison of intercellular responses to CBD.

We further investigated the antitumoral effect of CBD in a xenograft tumor mouse model. For this, HCT116 p53wt and HCT116 p53^−/− cells were injected subcutaneously (s.c.) into immunodeficient SCID mice following a whole-body irradiation with 3 Gy. To avoid methanol induced toxicity in mice, DMSO instead of methanol was used as the vehicle for CBD in the mouse experiment. We could demonstrate that irrespective of the vehicle (methanol or DMSO), CBD exerts comparable effects on cell viability of HCT116 p53wt cells, as demonstrated in a CCK8 assay (Supplementary Fig. 1a). Tumor-bearing mice were treated five times a week with CBD with the non-toxic dose of 20 mg/kg for 5 weeks as illustrated schematically in Fig. ​Fig.2c.2c. A significant reduction in the volume of HCT116 p53wt tumors was observed in mice after CBD treatment compared to the DMSO vehicle control from day 37 onwards (Fig. ​(Fig.2d).2d). In contrast, the volume of HCT116 p53^−/− tumors was not significantly reduced by the CBD treatment at any time point (Fig. ​(Fig.2e),2e), suggesting that a long-term CBD treatment only decreases the growth of p53wt, but not p53^−/− tumors. Preliminary studies demonstrated that the growth of HCT116 p53wt and HCT116 p53^−/− cell-derived tumors was comparable when 2.0 × 10⁶ p53wt and 2.8 × 10⁶ p53^−/− cells were injected s.c., respectively (Supplementary Fig. 1b). The treatment with CBD or the vehicle DMSO did not elicit any negative side effects in mice since the body weight of the mice was comparable in all treatment groups over a period of 40 days (Supplementary Fig. 1c). Representative examples of tumors treated either with vehicle or CBD are illustrated in Fig. ​Fig.2f.2f. Immunohistochemical analysis revealed a cell vacuolization in H&E sections of HCT116 p53wt tumors following CBD treatment, which was barely seen in HCT116 p53^−/− tumors (Fig. ​(Fig.2g).2g). Moreover, CBD induced a weak elevation in caspase-3 expression in HCT116 p53wt tumors which is indicative for apoptosis, whereas almost no apoptosis was apparent in HCT116 p53^−/− tumors (Fig. ​(Fig.2g).2g). With respect to DNA strand breaks, cleaved PARP1 was only accumulated in HCT116 p53wt cells upon a treatment with CBD at a concentration of 15 µM. This finding corresponds to an increase in the proportion of TUNEL positive cells (Fig. 2h, i). Similar to HCT116 p53^−/− cells, there was no significant difference in cleaved PARP1 expression after CBD treatment in SW480 p53mut cells (Fig. ​(Fig.2h2h and Supplementary Fig. 2a). However, in contrast to HCT116 p53wt cells, in LS174T p53wt cells neither cleaved PARP1 nor cleaved-caspase-3 were found to be upregulated (Supplementary Fig. 2a, b).

CBD induces a G0/G1 cell cycle arrest and a p53-dependent over production of ROS

CDK2 kinase plays a crucial role in regulating the G1-S transition. Mitotic cell arrest occurs when CDK2 kinase activity is inhibited by the CDK-specific inhibitor p21 [18]. We found that CBD in a concentration range of 7.5 and 15 µM induces a significant G0/G1 phase arrest which was more pronounced in p53wt (Fig. ​(Fig.3a)3a) cells than p53^−/− or p53mut cells, concomitant with a p53-dependent CDK2 downregulation and a p53-independent p21 upregulation (Fig. 3b–d). A similar trend was found in another LS174T p53wt cells (Fig. ​(Fig.3e).3e). As expected, CBD had no significant impact on the expression of cell cycle proteins in SW480 p53mut cells (Fig. ​(Fig.3f).3f). p53 is a transcription factor that acts as an upstream regulator for reactive oxygen species (ROS) in response to stress by activating or repressing several ROS-regulating genes, such as glutathione peroxidase (GPX) and p53-induced genes (PIGs) [19] which act in a pro- or an antioxidative environment. Our results indicate that the significant increase in ROS production induced by CBD is proportional to the p53 functional activity, since a nuclear translocation of p53 (Fig. 3g, h) together with a ROS accumulation following CBD treatment only occurs in p53wt cells (Fig. ​(Fig.3h),3h), but not in p53 deficient cells (Fig. ​(Fig.3h3h).

Heat shock protein 70 (Hsp70) attenuates CBD-induced apoptosis by inhibiting the p53-dependent caspase-8/9/3 pathway

Hsp70, a member of the HSP70 family encoded by HSPA1, is very strongly upregulated by heat stress and a large variety of other stress stimuli including toxic chemicals, particularly heavy metals such as arsenite, cadmium, copper, mercury, etc. Heat shock proteins were originally discovered by Ferruccio Ritossa in the 1960s when a lab worker accidentally boosted the incubation temperature of Drosophila (fruit flies). When examining the chromosomes, Ritossa found a “puffing pattern” that indicated the elevated gene transcription of an unknown protein [6, 7]. This was later described as the “Heat Shock Response” and the corresponding proteins were termed as “Heat Shock Proteins” (HSPs). It is well accepted that many tumor cells overexpress Hsp70 and that this overexpression correlates with resistance to apoptosis-inducing agents, whereas a downregulation of Hsp70 results in an increased sensitivity toward these agents [20]. Hsp70 can interfere with apoptosis pathways by blocking the aggregation of the functional apoptosome [21]. PES-CI, an Hsp70 inhibitor, binds to the substrate domain of Hsp70 which is important for the binding of client proteins, and thereby represses the enzymatic activity of the anaphase-promoting complex/cyclosome (APC/C) [22]. Previous study showed that PES interacts with the major stress-inducible Hsp70, but not with the constitutively expressed Hsc70 [23]. To better understand the potential role of Hsp70 on the antitumoral effects of CBD, we therefore interrogated the potential interaction partners of Hsp70 using the PES-CI inhibitor. As expected, PES-CI reduces the viability of all CRC cells in a dose-dependent manner (Fig. ​(Fig.4a),4a), with the sensitivity of HCT116 p53wt (IC₅₀ = 10.33 µM; 95% CI 4.937–13.18), HCT116 p53^−/− (IC₅₀ = 14.96 µM; 95% CI 12.59–17.91), LS174T (IC₅₀ = 8.451 µM; (95% CI 5.201–9.902) and SW480 (IC₅₀ = 15.78 µM; 95% CI 13.93–17.86) to PES-IC depending on the p53wt status. A co-treatment of CRC cells with the PES-CI and CBD results in distinct morphological changes which was more pronounced in p53wt cells. In general, the cells became round, detached from the surface of the culture flask and underwent apoptosis (Supplementary Fig. 3) which was documented by a disruption of the mitochondrial membrane potential (Fig. ​(Fig.4b),4b), concomitant with a substantial increase in the proportion of Annexin V/PI double positive, apoptotic cells (Fig. 4c–g). In contrast, the population of Annexin-V single positive cells was only significantly enhanced in p53wt cells, but not in p53^−/− or p53mut CRC cells (Fig. 4c–g). We further determined cleaved caspase-8/9/3 by flow cytometry to confirm apoptosis induction in wild-type p53 CRC cells. We demonstrated that CBD treatment alone considerably increased the levels of reactive oxygen species (ROS), but had only a moderate effect on the percentage of cells positive for Annexin-V and cleaved caspase-8/9/3 (Fig. 5a–d). However, when CBD was used in combination with the Hsp70 inhibitor PES-CI, the expression of cleaved caspase-9/3 was significantly elevated in cells harboring wild-type p53 (HCT116 p53wt, LS174T p53wt), but not in p53 deficient or p53 mutant cells (HCT116 p53^−/− and SW480 p53mut) (Fig. 5d, e). Caspase-8 displays a pivotal role in the p53-independent apoptotic signaling pathway following CBD treatment, and Hsp70 inhibition significantly accelerated caspase-8 dependent apoptosis (Fig. ​(Fig.5b).5b). As discussed previously [24, 25], the antioxidant N-acetyl cysteine (NAC) attenuates the antitumoral effect of CBD, as demonstrated by a reduction in the proportion of Annexin-V positive cells and the levels of cleaved caspase-8 (p = 0.0060), cleaved caspase-9 (p = 0.0545), cleaved caspase-3 (p = 0.0215) in HCT116 p53wt cells (Fig. ​(Fig.55).

Protective macroautophagy induced by CBD is related to ROS accumulation and wild-type p53 acts as a potential upstream regulator

Associated with the heat shock (stress protein) system, macroautophagy is a regulatory mechanism which maintains cellular protein homeostasis by sequestering and transporting large protein aggregates and damaged or senescent organelles to lysosomes for degradation.

Microtubule-associated protein 1A/1B-light chain 3 (LC3) is a ubiquitously distributed soluble protein which is incorporated into the expanded phagosome upon binding to phosphatidylethanolamine [26]. In this process, phagosome-bound LC3BII acts as a pillar for proteins bound to ubiquitinated substrates [27]. During the macroautophagy flux, autophagosomes fuse with lysosomes, which subsequently hydrolyze and degrade the content inside a cell [26]. The ability of PES-Cl to inhibit macroautophagy has been documented in several different autophagy assays [22, 28]. As shown in Fig. ​Fig.6a,6a, LC3BII appeared to be elevated in HCT116 p53wt cells, but not in HCT116 p53^−/− cells. p62 behaves as a linker in the macroautophagy process, and its aberrant aggregation indicates an impairment of the autophagy degradation pathway. To distinguish whether the accumulation of these proteins is caused by an increased macroautophagy or by a dysregulated auto-lysosomal degradation, we examined the levels of LC3BI and LC3BII in HCT116 p53wt cells 24 h after exposure to CBD in the presence and absence of BafA1 (50 nM), an autophagy inhibitor which halts the autophagic flux by inhibiting late-stage fusion between autophagosomes and lysosomes [29]. A co-incubation with BafA1 significantly enhances the CBD-induced increase of these proteins in HCT116 p53wt cells (Fig. ​(Fig.6b),6b), which means that CBD accelerates the macroautophagy process. Similar results were shown in Fig. ​Fig.6c.6c. The number of LC3B vesicles in p53wt cells were significantly increased after CBD treatment, while they did not significantly alter in p53 deficient cells. In line with a previous report [28], a 24 h co-incubation with CBD (15 µM) and PES-CI prevented the autophagy flux in both HCT116p53wt and HCT116 p53^-/-cells, whereas p62 overexpression was primarily dependent on the Hsp70 inhibitor in PES-CI in HCT116 p53^−/− cells. Additionally, NAC slightly suppressed the CBD-induced macroautophagy, which was induced by ROS in HCT116 p53wt cells, as it attenuated the p62 and LC3BII expression as well as LC3B vesicles without BafA1 (p = 0.3294 and p = 0.3821, respectively; Fig. 6a, c) or with BafA1 (p = 0.2118 and p = 0.1422, respectively; Fig. ​Fig.6b).6b). Autophagy was once known to be triggered by the lack of p53 or mutant p53 mostly during the G1 phase and to a lesser extent in the S phase to avoid that cells enter the G2/M phase [30]. However, our study found that the p53 inhibitor pifithrin-α (20 µM) decelerates the cytosolic expression of p62 (p = 0.026) and LC3BII (p = 0.2155) which induced by CBD after a 24 h cotreatment with BafA1 (Fig. ​(Fig.6d).6d). Based on the results of the viability assay, this kind of autophagy displays a cytoprotective effect (Fig. ​(Fig.6e).6e). In summary, our data indicates that CBD stimulates protective macroautophagy partially though ROS accumulation and it is likely that wild-type p53 displays a role in this process as an upstream regulating factor.

Screening of potential targets of CBD and enrichment assay

Well-reported pharmacological targets of cannabidiol were identified using the Drugbank [58], Swiss Target Prediction [59] and SuperPred [60] databases. The Genecard [61] database was used to screen for pathogenic targets in CRC. Potential targets for CBD in CRC were analyzed using Venn diagrams and the interactome network, functional processes and molecular pathways visualized using Metascape [62].

Cells and cell culture

The HCT116 (p53wt) was purchased from ATCC and HCT116 p53 double knockout (HCT116 p53^−/−) human colon adenocarcinoma cell lines was kindly provided by Prof. Bert Vogelstein (Johns Hopkins University, Baltimore, USA) and maintained in McCoy’s 5A medium (Sigma-Aldrich). The SW480 human adenocarcinoma colorectal cancer cell line (ATCC#CCL-228™; ATCC, USA) which carries two-point mutations (R273H/P309S) was cultured in Dulbecco’s modified Eagle’s medium (DMEM, Sigma-Aldrich) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin/streptomycin (Sigma-Aldrich). LS174T (p53wt, ATCC#CL-188™; ATCC, USA) [63] human adenocarcinoma colorectal cancer cells were cultured in DMEM supplemented with 10% (v/v) FBS and 1% (v/v) penicillin/streptomycin (Sigma-Aldrich), 2 mM L-glutamine (Sigma-Aldrich) and 1 mM sodium pyruvate (Sigma-Aldrich) at 37 °C in a 5% CO₂ atmosphere. After seeding at the desired density, cells were incubated overnight prior to the experiments. Cells were routinely checked for mycoplasma contamination.

Cell viability assay

The effect of CBD on the viability of cells was determined using a CCK-8 (Cell Counting Kit-8,) assay (Sigma-Aldrich). For this, cells were incubated with culture medium containing 5, 10, 15, 20 µM of CBD for 24–48 h, after which the absorbance at 450 nm was measured using a PerkinElmer 2030 multilabel reader (PerkinElmer LAS GmbH, Germany). The absorbance of the cells incubated in medium alone was used as a control (survival rate: 100%).

Cell cycle analysis

The cell cycle distribution of CBD-treated cells was determined by flow cytometry. Fluorescence detection of propidium iodide (PI)-DNA complexes was determined by flow cytometry (FACSCalibur™ flow cytometer BD Biosciences, Heidelberg, Germany). The distribution of cells in different stages of the cell cycle was analyzed using ModFit LT™ software (Scripps Research, La Jolla, CA, USA). A minimum of 30,000 cells was analyzed.

Apoptosis assay

An Annexin V-FITC/PI double staining assay was performed according to the manufacturer’s instructions (TACS® Annexin V Kits, R&D Systems). Briefly, after 24 h of CBD treatment, cells were collected and incubated with TACS Annexin V-FITC in binding buffer containing propidium iodide (Incubation Reagent) for 15 min at room temperature. Fluorescence intensity was measured using a FACSCalibur™ flow cytometer (BD Biosciences), and the apoptotic rates of CBD-treated cells were analyzed using BD FACSDiva™ software (version 6.1.3; BD Biosciences). A minimum of 10,000 cells were analyzed.

Mitochondrial membrane potential assay

Mitochondrial membrane potential was determined using the JC-I Mitochondrial Membrane Potential Assay Kit (ab113850, Abcam), according to the manufacturer’s instructions. Fluorescence intensity of cells was measured using a FACSCalibur™ flow cytometer (BD Biosciences). A minimum of 30,000 cells were analyzed.

Total reactive oxygen species (ROS) measurement

Total intracellular ROS levels were determined using DCFDA/H2DCFDA Cellular ROS Assay Kit (ab133851; Abcam). The DCFDA assay protocol is based on the diffusion of DCFDA/H2DCFDA/DCFH-DA/DCFH into the cell. It is then deacetylated by cellular esterases to a non-fluorescent compound, which is later oxidized by ROS into highly fluorescent 2’,7’-dichlorofluorescein (DCF). Cellular fluorescence (excitation/emission, ~485 nm/~535 nm) was quantified by flow cytometry. A minimum of 10,000 cells were analyzed.

Caspase activity assay

Measurements of activated caspase-8/9/3 were performed according to the manufacturer’s instructions using FITC-Caspase-3 antibody (345815; R&D Systems) or FITC-Caspase 8/PE-Caspase-9 antibodies (ab65615/ab65618; Abcam). Fluorescence intensity was determined by flow cytometry. A minimum of 10,000 cells were analyzed.

DNA double strand break labeling by flow cytometry (TUNEL assay)

The TUNEL assay was performed using BrdUTP analysis following the instructions of the assay kit (TUNEL Assay Kit, Thermo Fisher). Fluorescence intensity was determined by flow cytometry (FACSCalibur™ flow cytometer BD Biosciences, Heidelberg, Germany). A minimum of 10,000 cells were analyzed per sample.

Immunofluorescence

Cells were fixed with 4% (w/v) formaldehyde in phosphate buffered saline (PBS), and then permeabilized by incubation with 0.15% (v/v) Triton X-100 in PBS, after which cells were incubated overnight at 4 °C with monoclonal antibodies to p53 (murine IgG2a clone DO-1, sc-126, Santa Cruz), LC3B—autophagosome marker (rabbit polyclonal, ab48394, Abcam) or Nuclear Factor-Like 2 (Nrf2, murine IgG1 clone A-10, sc-365949, Santa Cruz). Primary antibody binding was detected using IgG (H + L) goat anti-mouse Alexa Fluor™ 488 Superclonal™ recombinant polyclonal secondary antibody (A28175, Invitrogen). Nuclei of labeled cells were counterstained with 40, 6-diamidino-2-phenyIindole (DAPI; 1 µg/ml for 1 min). Fluorescence images were taken using a Leica TCS SP8 confocal microscope. Fiji software (https://imagej.net/software/fiji/, accessed on 22 April 2021) [64] was used for quantification.

Immunoblot analysis

Cell lysates were prepared, separated by SDS-PAGE and blotting performed as described previously [65]. The protein content was determined using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific). The following antibodies directed against the indicated antigens were used in immunoblotting experiments: p53 (murine IgG2a monoclonal clone DO-1, Santa Cruz), LC3B—autophagosome marker (rabbit polyclonal, ab48394, Abcam), SQSTM1/p62 (recombinant rabbit monoclonal, ab211324, Abcam), p21 Waf1/Cip1 (rabbit monoclonal, 2947, Cell Signaling Technology), CDK2 (recombinant rabbit monoclonal, ab32147, Abcam), PARP1 (rabbit polyclonal, 9542, Cell Signaling Technology), Hsp70 (murine IgG1 monoclonal clone cmHsp70.1, multimmune GmbH), Nrf2 inhibitor keap1 (keap1, rabbit monoclonal, 8047, Cell Signaling Technology), β-Actin (murine IgG2a monoclonal antibody clone AC-74, A2228, Sigma-Aldrich). Primary antibody binding was detected using horseradish peroxidase (HRP)-conjugated rabbit anti-mouse immunoglobulins (P0260, Dako-Agilent) and HRP-conjugated swine anti-rabbit immunoglobulins (P0217, Dako-Agilent) secondary antibodies and a Pierce™ ECL Western Kit (Thermo Fisher Scientific). Blots were digitally imaged using a ChemiDoc™ Touch Imaging System (Bio-Rad). Fiji software (https://imagej.net/software/fiji/, accessed on 22 April 2021) [64] was used for quantifying Western blot signals.

In vivo tumor xenograft model

Female SCID mice purchased from the Pasteur Institute, Iran was kept under standard laboratory conditions and all experiments were performed according to the requirements of a project license (EE/1401.2.24.105658/SCU.AC.IR) issued by the Faculty of Veterinary Medicine Animal Ethics Committee of the Shahid Chamran University of Ahvaz, Iran. Specifically, animals had ad libitum access to food and water during maintenance under standard conditions (22 °C, 50% relative humidity, and 12 h light/dark cycles). Mice were adapted to the standard housing conditions for 1 week before the start of the experiments. All animal procedures were performed in compliance with the revised Animals Directive 2010/63/EU of the European Union.

Mice were randomly divided into two groups. HCT116 p53wt (2 × 10⁶) and the slower growing HCT116 p53^−/− cells (2.8 × 10⁶) were subcutaneously injected into 8-week-old female SCID mice in 100 µl McCoy’s 5 A medium 24 h after a 3 Gy whole-body irradiation. Tumor size, as measured using a caliper and body weight were measured twice per week. CBD (PhytoLab GmbH&Co.KG) was dissolved in a solution (5% DMSO, 5% Tween 80 (P4780, Sigma), 90% stroke-physiological saline solution) injected intraperitoneally (20 mg/kg, i.p.) 5 times a week for 5 weeks in total, from day 5 onwards. Mice were sacrificed on day 40 by isoflurane. Four to six mice were included in each group.

Immunohistochemistry

Tissue was fixed in formalin overnight and embedded in paraffin. Blocks were sectioned in 2 μm slices and stained with hematoxylin (Mayer’s hematoxylin) and eosin (eosin y-solution 0.5% (v/v) aqueous) to visualize tissue structure according to standard protocols. Caspase-3 staining using the antibody rabbit anti-caspase 3, cleaved (9661, Cell Signaling Technology) was performed to determine the extent of apoptosis. Biotin-conjugated secondary antibodies were incubated for 1 h at room temperature. Nuclear counterstaining was done using hematoxylin.

The authors want to thank Dr. Ritu Mishra and Hasanzadeh Kafshgari, Morteza for their technical support in the performance of the experiments. We thank Prof Alan Graham Pockley for proof-reading of the English.