Cannabidiol suppresses proliferation and induces cell death, autophagy and senescence in human cholangiocarcinoma cells via the PI3K/AKT/mTOR pathway

2.1. CCA cell culture and cannabidiol (CBD) preparation

The CCA cell lines (KKU-055, KKU-100, and KKU-213B) were originally established from patients with opisthorchiasis-associated CCA by Sripa et al.¹⁶ These CCA cell lines were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 4.5 g/L glucose (Gibco, Grand Island, NY, USA), 10 % fetal bovine serum (FBS; Corning, Woodland, CA,USA), 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco) at 37 °C with 5 % CO₂ in a humidified incubator.

The CBD (#02A64217) used in this study was obtained from the Department of Medical Sciences Reference Standards, Bureau of Drug and Narcotic, Department of Medical Sciences, Ministry of Public Health, Nonthaburi, Bangkok, Thailand. The CBD stock solution was prepared by dissolving it in dimethyl sulfoxide (DMSO; PanReac AppliChem ITW Reagents, Darmstadt, Germany) and diluted with complete medium to make 2–20 μM final concentration.

2.2. Cell proliferation detection by sulforhodamine B (SRB) assay

The SRB assay was slightly modified as previously described.¹⁷ Cells (2 × 10³ cells/well) were seeded into flat-bottom 96-well plates and treated with various concentrations of CBD for 24, 48 and 72 h. Subsequently, cells were fixed with 40 % cold trichloroacetic acid for 1 h and stained with 0.4 % (w/v) SRB solution in 1 % acetic acid for 1 h. SRB dye was dissolved in 10 mM Tris buffer pH 10.5. Absorbance was measured at 492 nm using an ELISA reader (Tecan group Ltd., Männedorf, Switzerland).

2.3. Clonogenicity assay

CCA cells (1 × 10³ cells/well) were grown in 6-well plates and treated with various concentrations of CBD. The culture medium was changed every 2 days, and cells were grown for approximately 2 weeks. Subsequently, cells were stained, and absorbance was measured as previously described.¹⁷

2.4. Cell cycle analysis

CCA cells (2 × 10⁵ cells/well) were seeded into 6-well plates and treated with 10 μM CBD for 48 h. Subsequently, cells were fixed with 70 % ethanol and stained with FxCycle™ PI/RNase staining solution as described.¹⁷ The stained cells were then sorted and analyzed using a BD FACSCanto II flow cytometer (BD Biosciences, San Jose, CA, USA) and FlowJo™ software version 10.8.1 (Becton, Dickinson & Company, OR, USA).

2.5. Lipid droplet/vesicle formation

Lipid droplet/vesicle formation was assessed as previously mentioned.¹⁸ CCA cells (2 × 10⁵ cells/well) were seeded into 6-well plates overnight. Then, cells were treated with 10 μM CBD for 48 h. A Nikon Eclipse TS 100 inverted light microscope (Nikon, Tokyo, Japan) was used to visualize cell morphology.

2.6. Optical diffraction tomography (ODT) and fluorescence imaging

To localize the lipid droplet accumulation and vesicle formation, ODT and fluorescence imaging were performed as previously described.¹⁹ CCA cells were exposed to 10 μM CBD for 48 h and cultured in a TomoDish at a density of 2 × 10⁴ cells per dish. Live cells were imaged using a Mach-Zehnder interferometric microscope HT-2H (Tomocube, Inc., Daejeon, Republic of Korea). Fluorescence imaging was used for additional 3D visualization, reconstructed via deconvolution using TomoStudio software (Tomocube, Inc.).

2.7. Transmission electron microscopy (TEM)

CCA cells (2 × 10⁵ cells/well) were seeded in a 6-well plates and treated with 10 μM CBD for 48 h. Cells were processed according to the methods outlined in a previous article.²⁰ TEM photographs were obtained using a JEOL JEM 1011 microscope, equipped with a Morada camera (JEOL Ltd., Tokyo, Japan).

2.8. Mitochondrial ROS assay

CCA cells (2.5 × 10⁴ cells/well) were seeded onto 8-well chamber slides. After overnight incubation, cells were treated with 10 μM CBD for 48 h. Then, cells were washed with 1X PBS, exposed to 5 μM MitoSOX Red for 30 min at RT, followed by Hoechst 33,342 (#H3570, Invitrogen).²¹ Imaging was conducted using Zeiss LSM880 confocal scanning microscope, capturing ten fields of view/group (20 × magnification). Fluorescence intensity was analyzed using ImageJ software as previously described.²²

2.9. Mitochondrial membrane potential (MMP; ΔΨm) assay

The MMP assay was performed as previously described.²³ CCA cells were seeded in 96-well plates (1 × 10⁴ cells/well) and treated with 5 and 10 μM CBD for 48 h. A positive control was exposed to 50 μM carbonyl cyanide m-chlorophenyl hydrazone (CCCP) for 10 min. JC-1 dye (200 μM) was added to each well and incubated for 20 min. Confocal images were captured using a Zeiss LSM880 microscope with ten fields per experimental group at 20 × magnification. Fluorescence intensity was quantified using ImageJ software, and mean values were calculated with GraphPad Prism 9.0 (GraphPad Software, La Jolla, CA, USA).

2.10. Reactive oxygen species (ROS) detection

CCA cells (2 × 10⁵ cells/well) were seeded into 6-well plates. After overnight incubation, cells were treated with 10 μM CBD for 12–24 h and 1 mM hydrogen peroxide (H₂O₂) as a positive control for 3 h. Following treatment, cells were incubated with 25 μM 2′,7′-dichlorofluorescein diacetate (DCFH-DA) for 30 min in the dark,²⁴ and washed twice with 1X PBS. Subsequently, cells were transferred to a FACS tube, washed twice with 1X PBS, and analyzed for ROS production using a flow cytometer (BD FACSCanto™ II) with data analyzed using BDFACSDiva™ software.

2.11. Senescence-associated β-galactosidase (SA- β-gal) staining

The effect of CBD on cellular senescence in CCA cells was investigated by SA-β-gal staining as previously described.²⁵ Cells were seeded into 6-well plates and treated with CBD for 48 h, and then stained with senescence β-galactosidase staining kit (#9860, Cell Signaling Technology, Danvers, MA, USA) following the manufacturer’s guidelines. Stained cells were imaged using a light microscope with ten fields per well captured at 20× magnification. Scoring of stained cells was done with the experiment replicated twice independently, each with triplicate technical replicates. Percentage of SA-β-gal-positive cells relative to total cells was computed for quantification.

2.12. Immunofluorescence assay (IFA)

CCA cells (2.5 × 10⁴ cells/well) were systematically seeded onto sterile cover slides within 24-well plates and treated with 10 μM CBD for 48 h. Additionally, a subset of cells was pre-treated with chloroquine (CQ) at a concentration of 10 μM, administered 8 h prior to CBD exposure.^26,27 After fixing with 4 % paraformaldehyde and permeabilizing with 0.2 % Triton X-100 (Amresco, Solon, OH, USA), cells were blocked with PBS-5% BSA with 0.1 % Triton X-100 and incubated overnight with primary antibodies against LC3BI/II (1:250, #A7198, ABclonal, Wuhan, China) and p62 (1:250, #A7758, ABclonal). Subsequently, cells underwent incubation with secondary antibodies (1:10, #5230-0298, KPL Affinity Purified Antibody Fluorescein-Labeled Goat Anti-Rabbit IgG (H + L), SeraCare Life Sciences Inc., MA, USA). Concurrently, CellMask Deep Red, a specialized plasma membrane stain (#C10046, Invitrogen), was employed to highlight the cellular plasma membranes. Additionally, Hoechst 33,342 was applied for nuclear counterstaining purposes. Ten fields from images of each experimental group were acquired at a magnification of 100 × using a fluorescence imaging microscope (ECLIPSE Ni–U, Nikon Instruments Inc, Japan). Subsequently, fluorescence-intensity images were analyzed using ImageJ software²² (National Institutes of Health, Bethesda, MD, USA).

2.13. Western blotting

Protein extraction was performed using 1X RIPA reagent (#9806S, Cell Signaling Technology) and protein concentrations were determined using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific Inc., Rockford, IL, USA). SDS-PAGE technique was used to separate proteins (20–25 μg).²⁷ PVDF membranes were incubated with primary antibodies (dilution 1:1000) for: p62 (#A7758), LC3BI/II (#A7198), p21 (#A19094), and phospho-PI3K (#AP0854), all purchased from Abclonal; Bcl-2 (#ab7973) and β-actin (#ab3280), both purchased from Abcam; phospho-mTOR (Ser2448) (#5536), phospho-AKT (Ser473) (#2983), PI3K (#42925), AKT (#92725) and mTOR (#2983), all purchased from Cell Signaling Technology, and p53 (#10442-1-AP, from Proteintech, IL, USA). Then, membranes were incubated with secondary antibodies HRP-conjugated goat anti-rabbit IgG (1:2,000, #111035003, Jackson Immuno Research Inc., West Grove, PA, USA).

2.14. Xenograft model

KKU-100 cells (1 × 10⁶ cells) were injected subcutaneously into female knockout BALB/c Rag-2.Jak3 double-knockout mice.²⁸ After 21 days, mice were treated with either DMSO or 50 mg/kg·Bw of CBD (n = 5) daily for 4 days. Tumor sizes were measured at days 0, 3 and 5 following this, and the tumor volume was calculated using the formula: tumor volume = (length × width²)/2.

2.15. Immunohistochemistry study

Tumor tissues from mice were immunohistochemically analyzed by antigen retrieval with citrate buffer and Tris-EDTA buffer, blocking with 5 % FBS, and incubation with anti-proliferating cell nuclear antigen (PCNA) antibody (1:750, #Ab2426, Abcam), p21 antibody (1:50, #A19094, Abclonal), and GBL1 antibody (1:40, #A1863, Abclonal) overnight. The signal was detected with diaminobenzidine substrate and counterstained with Mayer’s hematoxylin.²⁷ Brown staining indicated positive cells. Ten fields (20 × magnification) were analyzed using ImageJ software.

2.16. Statistical analysis

The experiments were conducted in biological triplicate with results presented as mean ± SD. Statistical comparisons involved one-way ANOVA followed by Tukey’s test for multiple group comparisons and Student’s t-test for in vivo investigations with two cohorts.²⁹ A significance threshold of p < 0.05 was used. All analyses were performed using GraphPad Prism 9.0 for Mac.

3.1. CBD treatment suppresses CCA cell proliferation, induces cell cycle arrest and inhibits colony formation

The SRB assay showed that CCA cell proliferation in all three cell lines decreased following CBD treatment in a dose- and time-dependentmanners (Fig. 1A). The half-maximal inhibitory concentration (IC₅₀) values at 48 h of treatment were 11.34 ± 1.46, 10.62 ± 1.23, and 10.59 ± 1.29 μM for KKU-055, KKU-100 and KKU-213B, respectively. After 48 h of treatment, there were no differences in the IC₅₀ values among the three distinct CCA cell lines. Cell line KKU-055 was not used in the remaining investigations reported below. While, IC₅₀ for MMNK 1 was 12.73 ± 1.09, indicating that CBD has low toxic effect (Supplementary figure 1).

To further investigate the inhibitory effects of CBD on CCA cell proliferation, KKU-100 and KKU-213B cells were treated with 10 μM CBD for 48 h. Flow cytometric analysis revealed a significant increase in the proportion of KKU-100 and KKU-213B cells in the G1 phase (KKU-100; p < 0.001, KKU-213B; p < 0.001) after treatment with 10 μM CBD compared to the 0.3 % DMSO treatment (Fig. 1B and C).

To investigate the inhibitory effect of CBD on the proliferative capacity of a single CCA cell, a clonogenic assay was performed (Fig. 1D). CBD treatment inhibited colony formation in both CCA cell lines (KKU-100 and KKU-213B) in a dose-dependent manner. Treatment with CBD at 10 and 20 μM significantly inhibited colony formation of KKU-100 cells (p < 0.001 and p < 0.001, respectively) and KKU-213 (p < 0.01 and < 0.001, respectively) compared to vehicle control (0.3 % DMSO treatment) (Fig. 1E). Notably, 20 μM CBD exhibited greater efficacy than 10 μM CBD in inhibiting colony formation in both cell lines (Fig. 1D and E).

3.2. CBD induces lipid droplet and vesicle formation in CCA cells

Following a 48-h treatment with 10 μM CBD, KKU-100 CCA cells enlarged and contained various vesicles in the cytoplasm when observed using an inverted light microscope and 3D holotomography (Fig. 2A and B). This was confirmed by the observations obtained from TEM, which showed extensive cytoplasmic vacuolation, nucleolus enlargement, and the presence of autophagosomes (Fig. 2C). In addition, both CCA cell lines exhibited the presence of single- and double-membrane vacuoles.

3.3. CBD increases mitochondrial ROS, changes mitochondrial membrane potential (MMP; ΔΨm) and increases ROS production in CCA cells

Fig. 3 demonstrates the effect of CBD on mitochondrial ROS, changes MMP and increases ROS production in CCA cells. Induction of ROS generation in mitochondria was evaluated using MitoSOX Red as a fluorescent probe. Elevated orange-red fluorescence indicated increased mitochondrial ROS levels. Treatment with 10 μM CBD significantly boosted fluorescence intensity in KKU-100 and KKU-213B cell lines (p < 0.0001 in both cases) compared to vehicle control (Fig. 3A–C). Additionally, cells treated with vehicle control (0.3 % DMSO) exhibited a higher proportion of red fluorescence,indicating healthy mitochondria, whereas, cells treated with 50 μM CCCP or CBD displayed a higher proportion of green fluorescence,indicating mitochondrial damage, relative to red fluorescence (Fig. 3D). In CBD treatment, the red/green fluorescence ratio of KKU-100 cells significantly decreased to 1.12 ± 0.07 (p < 0.001) and 0.83 ± 0.08 (p < 0.001) with CBD at concentrations of 5 and 10 μM, respectively, when compared to vehicle control (1.53 ± 0.14). Similarly, the red/green ratio of KKU-213B cells was significantly lower at 1.01 ± 0.12 (p < 0.001) and 0.83 ± 0.06 (p < 0.001) by CBD treatment at concentrations of 5 and 10 μM, respectively, when compared to vehicle control (1.24 ± 0.10) (Fig. 3E).

Flow cytometry revealed that KKU-100 CCA cells displayed a slight increase in fluorescence intensity when treated with 5 μM CBD and a significant rise (p < 0.01) with 10 μM CBD after 12 h. This did not occur in KKU-213B cells. As expected, both CCA cell lines gave the highest fluorescence intensity with 1 mM H₂O₂, the positive ROS control, after 3 h (Fig. 3F).

3.4. CBD induces autophagy in CCA cells

We further examined the impact of co-administration of CBD and CQ, an autophagy inhibitor, on autophagy pathways through IFA (Fig. 4) and Western blot analysis. IFA revealed increased LC3 fluorescence following CBD and CQ co-administration compared to 0.3 % DMSO control (p < 0.0001) (Fig. 4A and 4C-D). CBD treatment also altered p62 fluorescence, with significant effects in KKU-213B (p = 0.025) but not KKU-100 (p = 0.9023) (Fig. 4B and 4E-F). In contrast, CBD treatment led to a significant reduction in fluorescence intensity of p62 compared to CQ alone or the combination of CQ with CBD.

We further confirmed this by Western blot analysis (Fig. 5). CQ alone or with CBD significantly increased LC3B–I and II expression compared to 0.3 % DMSO or CBD treatment alone. CBD at 10 μM notably elevated LC3B-II expression in both cell lines compared to 0.3 % DMSO alone (KKU-100: p = 0.011, KKU-213B: p = 0.049). CBD treatment exhibited the most pronounced LC3B-II/I ratio (p < 0.0001) (Fig. 5A–D). A significantly decreased expression of p62 levels was observed in CCA cell lines with CBD treatment compared to 0.3 % DMSO or CQ alone (Fig. 5A and E). Additionally, CBD treatment reduced Bcl-2 expression levels in a dose-dependent manner. Significant reductions in Bcl-2 expression were observed at various CBD concentrations (KKU-100: 10 μM, p = 0.036; KKU-213B: 5 μM, p = 0.027; 10 μM, p = 0.007) relative to vehicle control (Fig. 5F).

3.5. CBD suppresses the expression of the PI3K/Akt/mTOR pathway

To investigate the impact of CBD on the PI3K/Akt/mTOR pathway, Western blot analysis was conducted (Fig. 6). The basal expression of PI3K, Akt and mTOR in KKU-100 and KKU-213B cells did not differ between the control and CBD treatment groups. However, in KKU-100 cells treated with 5 and 10 μM CBD, there was a reduction in the expression of phosphorylated PI3K (p-PI3K), phosphorylated Atk (p-Atk) and phosphorylated mTOR (p-mTOR) at 24 h compared to the control group. Interestingly, relative to controls, KKU-100 cells treated with 10 μM CBD exhibited a significant decrease in the expression of p-PI3K (p = 0.039), p-AKT (p = 0.048), and p-mTOR (p = 0.037) at 48 h. Likewise, KKU-213B cells treated with CBD 10 μM showed a slight decrease in the expression of p-PI3K, p-AKT, and p-mTOR at 24 h, but a significant decrease in these proteins was observed at 48 h (p = 0.019, 0.006 and 0.030, respectively; Fig. 6A and D).

3.6. CBD promotes cellular senescence in CCA cells

To evaluate the effect of CBD on induction of cellular senescence in CCA cell lines, SA-β-gal staining was performed. SA-β-gal appeared deep blue as shown in Fig. 7A. Cellular senescence significantly increased in KKU-100 and KKU-213B cells treated with 5 μM and 10 μM (p < 0.0001) compared to the vehicle control (Fig. 7B and C). To confirm cellular senescence, a Western blot analysis was performed to examine the cellular senescence markers p21 and p53 (Fig. 7D). CBD treatment (10 μM) of KKU-100 and KKU-213B for 24 h elicited a significant upregulation in the expression of p21 (KKU-100: p < 0.0001) when contrasted with treatment involving 0.3 % DMSO (Fig. 7E). In addition, p53 expression was found to be significantly elevated in the groups treated with 10 μM CBD (KKU-100: p = 0.042, KKU-213B: p = 0.035; Fig. 7D and F), confirming the activation of cell senescence by CBD in CCA cells.

3.7. CBD suppresses CCA cell proliferation and promotes cellular senescence in vivo

The tumor volume of xenograft mice treated with 50 mg/kg·Bw was significantly lower upon evaluation on day 5 after treatment (p = 0.0357) when compared to vehicle control treatment (Fig. 8A and B). Additionally, an immunohistochemistry study was performed to confirm the effect of CBD on the suppression of CCA proliferation and induction of cellular senescence at the molecular level (Fig. 8C). The expression of PCNA, β-gal and p21 were assessed and its area and intensity were measured using ImageJ software (Fig. 8D). The expression of PCNA in the 50 mg/kg·Bw group was significantly reduced (p < 0.0001) relative to the vehicle control group. Immunoreactivity indicated a significant increase in β-gal levels in mice treated with 50 mg/kg·Bw of CBD (p < 0.0001). Consistently, a noteworthy elevation in p21 expression was observed in tumor tissues by CBD treatment at the same dosage (p < 0.0001).