The Combination of Δ9-Tetrahydrocannabinol and Cannabidiol Suppresses Mitochondrial Respiration of Human Glioblastoma Cells via Downregulation of Specific Respiratory Chain Proteins

2.2. Cell Culture

The human glioblastoma cell line U251MG (#09063001; RRID:CVCL_0021) was purchased from Sigma-Aldrich Corporation. The human glioblastoma cell line U138MG cell line (#ACC 291; RRID: CVCL_0020) was bought from DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany). Both cell lines were cultured in Dulbecco’s modified eagle medium (DMEM) with 4.5 g/L glucose and UltraGlutamine I (Lonza Cologne GmbH, Cologne, Germany). DMEM was supplemented with 10% heat-inactivated FBS, 100 U/mL penicillin and 100 μg/mL streptomycin. The cells were cultured in a humidified incubator at 37 °C and 5% CO₂.

All cell treatments with specific substances were performed in serum-free DMEM supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin, applied after washing the cells with DPBS. The test substances were dissolved in absolute ethanol (THC, CBD), DMSO (AM-251, AM-630, bafilomycin A₁, bortezomib, capsazepine and MG-132) or 1 M NaOH (SnPPIX). The final concentration of solvents in the incubation media of cells treated with the test substance and vehicle varied between experiments, but in no case exceeded 0.033% (v/v) for ethanol, 0.1% (v/v) for DMSO and 0.4 mM for NaOH. All incubation media of an experiment also contained the same amount of solvent.

2.3. Cellular Viability Assay

For viability analyses, two different assays were employed. In all experiments, U251MG and U138MG cells were seeded in 96-well plates at a density of 5000 cells per well in serum-containing DMEM and cultured for 24 h before washing with DPBS and treatment with the test substances in serum-free and phenol red free Gibco^TM DMEM with 4.5 g/L glucose and L-glutamine (#21063; Thermo Fisher Scientific Inc.) supplemented with 100 U/mL penicillin and 100 μg/mL streptomycin for 24 h. The use of serum-free media minimises loss of substance due to interaction with serum proteins and discrepancies in cell number over time due to different proliferation rates of cell lines.

The metabolic activity of the cells was determined using WST-1 reagent (Roche Diagnostics, Mannheim, Germany). The water-soluble tetrazolium WST-1 was added to a final dilution of 1:10 and cells were further incubated for 10 min before the absorbance was measured at 450 nm (wavelength correction at 690 nm) using a microplate reader.

For cell number analysis, cells were fixed with ice-cold absolute ethanol for 30 min before incubation with crystal violet staining solution (0.1% [w/v] crystal violet in 10% ethanol) for 30 min. After thoroughly washing off the excess dye, the stained cells were dissolved with 10% acetic acid and the dye intensity was measured at 570 nm with a microplate reader.

2.4. Cellular Apoptosis Assay

The analysis of caspase-3 and -7 activity was carried out using Caspase-Glo^® 3/7 Assay (#G8093; Promega, Walldorf, Germany). For this purpose, U251MG and U138MG cells were seeded in 96-well plates at a density of 5000 cells per well in serum-containing DMEM for 24 h. Cells were then washed with DPBS and incubated with the appropriate test substances for a further 24 h in serum-free and phenol red free DMEM. Caspase-Glo^® 3/7 Reagent was added in equal volume proportions according to the manufacturer’s instructions and incubated in the dark for 1 h. Luminescence was measured with a microplate reader.

2.5. Seahorse XFe Analysis

The oxygen consumption rate (OCR) and extracellular acidification rate (ECAR) were determined using a Seahorse XFe24 Analyser (Agilent Technologies, Inc., Waldbronn, Germany) according to the manufacturer’s instructions. U251MG and U138MG cells were seeded on Seahorse 24XFe plates (Agilent Technologies, Inc., Waldbronn, Germany) at a density of 25,000 cells per well and incubated in serum-containing DMEM for 24 h. After washing with DPBS, the cells were treated with the test substances in serum-free DMEM for 24 h. To start the Seahorse XFe analysis, media were changed on the cells to unbuffered XF base media, pH 7.4 (#103575-100, Agilent Technologies, Inc.) supplemented with 10 mM glucose, 2 mM glutamine and 1 mM pyruvate and incubated for 1 h. Compounds from the Seahorse XF Cell Mito Stress Test Kit (#133015-100, Agilent Technologies, Inc.) were added to the wells at final concentrations according to the manufacturer’s instructions (port A: 1.5 μM oligomycin (ATP synthase inhibitor), port B: 1.0 μM FCCP (carbonyl cyanide-4 (trifluoromethoxy) phenylhydrazone; uncoupling agent that collapses the proton gradient and disrupts the mitochondrial membrane potential), port C: 0.5 μM rotenone (complex I inhibitor) and 0.5 μM antimycin A (complex III inhibitor)). OCR and ECAR per well were normalised to the amount of protein determined after the experiment. To this end, 10 µL of lysis buffer (50 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1% [v/v] Triton^® X-100, 10% [v/v] glycerol, 10 µg/mL aprotinin, 1 µg/mL leupeptin, 1 mM orthovanadate and 1 mM PMSF) was added to each well and the lysates were collected. Protein concentration was determined using the Pierce™ BCA (bicinchoninic acid) Protein Assay Kit (Thermo Fisher Scientific Inc.). Calculation of basal respiration, ATP production-based respiration, spare capacity and proton leak was performed using the Seahorse XF Cell Mito Stress Test Report Generator (Agilent Technologies, Inc.).

2.6. Lactate Release Analysis

Lactate was determined in the supernatant of the cultured cells using the L-Lactate Assay Kit (#ab65330, Abcam, Cambridge, UK) according to the manufacturer’s instructions. Cell culture supernatants were obtained from U251MG cells seeded at a density of 25,000 cells per well in 48-well plates in 10% FBS containing DMEM for 24 h, which were washed with DPBS and subsequently treated with the test substances in serum-free DMEM for 24 h. Lactate concentrations were measured at 570 nm (reference wavelength 450 nm) in a microplate reader.

2.7. Total Cellular Protein Isolation

U251MG or U138MG cells were seeded in 6-well plates at a density of 200,000 cells per well and cultured in DMEM containing 10% heat-inactivated FBS for 24 h before treatment with substances in serum-free DMEM. After the respective incubation time, the medium supernatant was separated from the cells, and the cells were washed with DPBS, detached by trypsin-EDTA and collected again in the medium supernatant. Cells were harvested in a pellet after centrifugation for 5 min (200× g, 4 °C). The cell pellet was washed with DPBS and collected again after centrifugation (250× g, 4 °C) for 5 min. Lysis buffer (50 mM HEPES, 150 mM NaCl, 1 mM EDTA, 1% [v/v] Triton^® X-100, 10% [v/v] glycerol, 10 µg/mL aprotinin, 1 µg/mL leupeptin, 1 mM orthovanadate and 1 mM PMSF) was added to the cell pellet, followed by 30 min incubation on ice and subsequent centrifugation for 5 min (20,817× g, 4 °C). The resulting supernatant, containing total cellular protein, was collected and stored for further protein analysis. Protein concentration was measured using the Pierce™ BCA Protein Assay Kit (Thermo Fisher Scientific Inc.).

2.8. Mitochondrial Protein Isolation

U251MG or U138MG cells were seeded, treated and harvested as described in the whole cellular protein isolation experiments. The resulting cell pellet was washed in 0.9% NaCl and centrifuged for 5 min (500× g, 4 °C). To isolate mitochondria and to obtain the cytosolic protein fraction, the Qproteome Mitochondria Isolation Kit (Qiagen GmbH, Hilden, Germany) was used according to the manufacturer’s instructions for standard preparation. For extraction of mitochondrial protein, the resulting mitochondrial fraction was treated with lysis buffer, centrifuged and collected in a similar manner to the protocol listed for isolation of total cellular protein.

2.9. Western Blot Analysis

Equal amounts of protein were separated on a 15% SDS-polyacrylamide gel, transferred to a nitrocellulose membrane and incubated for 1 h in 5% (w/v) NFM in Tris-buffered saline containing 0.1% (v/v) Tween^® 20 (TBS-T buffer). After washing with TBS-T buffer, membranes were incubated overnight at 4 °C with primary antibodies in 1% (w/v) NFM. The following antibodies were purchased from Thermo Fisher Scientific Inc.: NDUFA9 (#459100, RRID:AB_2532223), OxPhos Complex IV subunit IV (#A21348; COX4, RRID:AB_2535839), ATP Synthase β subunit (#A21351; ATP5B, RRID:AB_221512) and OxPhos Human WB Antibody Cocktail (#45-8199, RRID:AB_2533836). Other antibodies were LC3A/B (#4108, RRID: AB_2137703) and Cytochrome c (#11940, RRID:AB_2637071), both from Cell Signaling Technology (Frankfurt/Main, Germany), SDHA (#ab14715, RRID:AB_301433) and VDAC (#ab14734, RRID:AB_443084) from Abcam (Berlin, Germany), HO-1 (#ADI-SPA-894F, RRID:AB_10631417) from Enzo Life Sciences and GAPDH (#G9545, RRID:AB_796208), β-actin (#A5441, RRID:AB_476744) and LONP1 (#HPA002192, RRID:AB_1079695) from Sigma-Aldrich Corporation. After washing with TBS-T buffer, membranes were incubated with secondary horseradish peroxidase-coupled antibodies (anti-rabbit antibody, #7074, RRID:AB_2099233; anti-mouse antibody, #7076, RRID:AB_330924; Cell Signaling Technology) in 1% (w/v) NFM in TBS-T buffer for 1 h at room temperature. To visualise antibody binding, a substrate solute for chemiluminescence detection (100 mM Tris hydrochloride, pH 8.5; 1.25 mM luminol; 200 µM p-coumaric acid; 0.09% [v/v] H₂O₂) was added and signal detection was performed using the ChemiDoc XRS gel documentation system manufactured by Bio-Rad Laboratories GmbH (Munich, Germany). Quantification of signal intensity was carried out using the 1-D analysis software Quantity One (Bio-Rad Laboratories GmbH). The signal of a specific protein band was normalised to the signal of the loading control. Accordingly, the protein levels of VDAC for mitochondrial proteins, GAPDH for cytosolic proteins or β-actin for total cellular protein membranes were also analysed. Protein levels were then calculated relative to the vehicle control. The Precision Plus Protein^TM Dual Colour Standard from Bio-Rad Laboratories GmbH was used to identify the molecular weight of the bands. The Western blots shown in the Figures are included in File S1 in uncropped form.

2.10. Electron Microscopy

For electron microscopy, U251MG or U138MG cells were seeded at a density of 200,000 cells per well in a 6-well plate and cultured in DMEM containing 10% heat-inactivated FBS for 24 h before treatment with substances in serum-free DMEM. Cells were then incubated with THC/CBD or vehicle for 24 h and subsequently treated in the same way as described for the protein extraction experiments. The cell pellets were fixed in a solution containing 2% glutaraldehyde and 1% paraformaldehyde in 0.1 M phosphate buffer pH 7.3, and were stored at 4 °C until further processing for embedding. Next, after two washes in 0.1 M phosphate buffer, the cells were mixed with prewarmed 2% low melting agarose (Sigma-Aldrich Corporation) in 0.05 M HEPES buffer at 40 °C, collected by centrifugation and enclosed as a pellet in the agarose after hardening. The specimen blocks were then post-fixed in a solution of 1% osmium tetroxide (Carl Roth GmbH + Co. KG, Karlsruhe, Germany) for 2 h and, after washes in distilled water, were dehydrated in an ascending series of acetone. Resin infiltration was started with an overnight incubation in a 1:1 mixture of acetone and Epon resin (48% Epon 812, 30% methylnadic anhydride, 20.7% 2-dodecenylsuccinic acid anhydride and 1.3% 2,4,6-Tris(dimethylaminomethyl)phenol; all components from Serva, Heidelberg, Germany) followed by pure Epon resin for 4 h. Specimen blocks were transferred to silicone rubber moulds with fresh resin and were cured in an oven at 60 °C for 2 days. Further processing included trimming of the resin blocks (Leica EM Trim2, Leica Microsystems, Wetzlar, Germany) and subsequent sectioning on an ultramicrotome (Ultracut S, Reichert, Wien, Austria) using a diamond knife (Diatome, Nidau, Switzerland). Semithin sections with a thickness of 0.5 µm were stained with toluidine blue to visualise and select the areas for ultrastructural inspection. Ultrathin sections of approximately 50–70 nm thickness were transferred to copper grids and stained with uranyl acetate and lead citrate prior to examination with a Zeiss EM902 transmission electron microscope operated at 80 kV (Carl Zeiss, Oberkochen, Germany). Digital images were acquired with a side-mounted 1x2k FT-CCD Camera (Proscan, Scheuring, Germany) using iTEM camera control and imaging software (Olympus Soft Imaging Solutions, Münster, Germany).

For quantification of mitochondrial density, all mitochondria were manually outlined in the images of the greatest magnification and the mean intensity measured with the open source software FIJI (https://imagej.net/software/fiji). In order to present the data as density, the measured intensities were subtracted from the maximal value of 65,535 which represents white in 16-bit images. The data were plotted as mean over all data points ± SEM.

2.11. Quantitative Reverse Transcriptase Polymerase Chain Reaction (RT-PCR)

U251MG or U138MG cells were seeded at 200,000 cells per well in a 6-well plate, cultured for 24 h in DMEM containing 10% heat-inactivated FBS, treated with compounds for another 24 h in serum-free DMEM, and then harvested as in the protein isolation experiments. Total RNA was isolated from the obtained cell pellet using the RNeasy Mini Kit from Qiagen GmbH (Hilden, Germany). For quantification of TFAM, the Applied Biosystems^® TaqMan^® Gene Expression Assay (Assay ID: Hs00273372_s1) and the Applied Biosystems^® TaqMan^® RNA-to-CT™ 1-Step Kit from Thermo Fisher Scientific Inc. were used according to the manufacturer’s instructions. Peptidylprolyl isomerase A (PPIA; Assay ID: Hs99999904_m1) was used as the housekeeping gene to normalise TFAM mRNA levels prior to comparison with the respective vehicle controls.

3.2. Combined Treatment with THC and CBD Suppresses Mitochondrial Respiration in Human Glioblastoma Cells, but Triggers Cell Line-Specific Differential Effects on Glycolysis

According to the WST assay, the combination of THC and CBD led to decreases in the viability of the treated cells at the level of metabolic activity. In order to assess changes to the cellular energy metabolism upon treatment with THC/CBD, we next determined the oxygen consumption rate (OCR) as an indicator of mitochondrial respiration and the extracellular acidification rate (ECAR) as an indicator of glycolysis. Measurements carried out with the Seahorse analyser revealed an inhibition of OCR by THC and CBD (2.5 µM each) in both cell lines, as well as a much more pronounced reduction when the corresponding combination was applied (Figure 2A,B). For the combination of 1 µM THC and CBD each, a significant OCR reduction could also be detected in U138MG cells (Figure 2B), whereby this cell line showed a comparatively higher basal respiration rate in vehicle-treated cells compared to U251MG cells. Interestingly, a divergent pattern was found for glycolysis. Here, an increase of ECAR was observed in U251MG cells incubated with the combination of 2.5 µM each of THC and CBD (Figure 2C), while a significant decrease in ECAR was registered by the administration of THC and CBD alone (2.5 µM each) as well as by the combinations of 1 and 2.5 µM of the two cannabinoids (Figure 2D). Accordingly, plotting OCR versus ECAR on an energy map showed bioenergetic profiles for the combination of THC and CBD (2.5 µM each) with a strong reduction of the oxidative phosphorylation compared to the single substances (Figure 2E,F) and a simultaneously higher (U251MG, Figure 2E) or lower (U138MG, Figure 2F) glycolytic activity. The increased ECAR in U251MG cells may reflect a compensatory attempt by the cells to regulate energy metabolism in response to reduced OCR, which, however, had no positive effects on cell survival (Figure 1). To confirm these results, lactate secretion from U251MG cells was also measured, demonstrating a significant increase upon treatment with a combination of 2.5 µM each of THC and CBD (Figure S1).

To gain a more detailed insight into the mitochondrial function, a mitochondrial stress test was performed, which in addition to basal respiration also includes the parameters ATP-linked respiration, maximal and reserve capacities, and non-mitochondrial respiration. This showed a statistically significant synergistic inhibition of the basal respiration by the combination of THC and CBD (2.5 µM each) in both U251MG (Figure 2G,I) and U138MG cells (Figure 2H,J). In U251MG cells, which exhibit a high proportion of ATP-linked respiration as well as strong FCCP-induced maximal respiration under vehicle treatment, both ATP-linked respiration and especially reserve capacity were strongly reduced by the combination of THC and CBD. In contrast, U138MG cells displayed a high non-mitochondrial respiration and, under vehicle treatment, only a small ATP-linked respiration as well as a slight induction of the respiration by FCCP. In these cells, THC/CBD also induced changes in the ATP-linked respiration and reserve capacity; however, these did not reach statistical significance. Overall, although both cell lines show differences in their energy metabolism, the combination of THC and CBD causes a similar OCR reduction in either cell line, which is mainly due to the loss of mitochondrial respiration.

3.3. Combined Treatment with THC and CBD Confers Structural Mitochondrial Changes and Promotes the Release of Cytochrome c

Reduced basal and ATP-linked respiration as well as spare and maximal capacities in THC/CBD-treated cells may be due to mitochondrial damage. To investigate this, electron microscopic examinations were carried out. The cytoplasm of cannabinoid-treated cells exhibited high vacuolisation (Figure 3A,B images on the right). Many vesicles were recognisable as phagosomes (Figure 3C,D), corresponding to the increase of autophagy upon treatment with THC/CBD shown before (Figure 1G,H). The proximity of phagosomes to mitochondria was prominent. The mitochondria in the THC/CBD-treated cells displayed marked enlargement as well as rounding, and the mitochondrial cristae showed marked structural damage (Figure 3C,D). Quantification revealed a significant loss of electron-dense material within the mitochondria of THC/CBD-treated cells as compared to the vehicle control (Figure 3E,F). Consistent with the microscopically detected mitochondrial damages, there was also a release of cytochrome c into the cytosol (Figure 3G,H). This release indicates a loss of mitochondrial membrane integrity and represents a known apoptosis triggering mechanism, consistent with the activation of effector caspases shown earlier (Figure 1E,F).

3.4. Combined Treatment with THC and CBD Mediates Downregulation of Specific ETC Proteins

Mitochondrial damage is often associated with a defective ETC. To identify the cause of mitochondrial respiratory impairment and mitochondrial integrity, protein concentrations of mitochondrial proteins, especially ETC complexes, were examined. Accordingly, Western blot analyses were performed after treatment of cells with different concentrations of THC and CBD as well as the combination of both cannabinoids (Figure 4A,B). A highly significant downregulation of NADH:ubiquinone oxidoreductase subunit A9 (NDUFA9, subunit of ETC complex I) was obvious in both cell lines in the combination group (2.5 µM THC and CBD each) as well as after single treatment with 5 µM THC and CBD each (Figure 4A,B). Remarkably, under the same conditions, the protein concentrations of ATP synthase subunit β (ATP5B, subunit of complex V), succinate dehydrogenase complex, subunit A (SDHA, subunit of complex II) and voltage-dependent anion-selective channel (VDAC, channel of the outer mitochondrial membrane) remained unaffected (Figure 4A,B). Therefore, a general change in mitochondrial mass can be ruled out and a specific effect on NDUFA9 can be concluded.

To support these findings, Western blot analysis was performed on mitochondria isolated from cells after treatment with a combination of 2.5 µM THC and CBD (Figure 4C–F). Again, NDUFA9, but not ATP5B, SDHA and VDAC showed a strong decrease in protein levels (Figure 4C,D). On the other hand, a decrease in protein of about 90% was also observed for cytochrome c oxidase subunit 4 (COX4, subunit of complex IV) (Figure 4C,D), which was of a similar magnitude to the decrease observed for NDUFA9.

Corresponding regulation of subunits of complexes I and IV was also confirmed using an OxPhos antibody set. Thus, at the mitochondrial level, a significant reduction of the NADH:ubiquinone oxidoreductase subunit B8 (NDUFB8, subunit of complex I) as well as of the cytochrome c oxidase subunit 2 (COX2, subunit of complex IV) could be registered in the presence of the combination of THC and CBD (Figure 4E,F); this indicates a disruption of the respective entire protein complex. In addition, a moderate inhibition of succinate dehydrogenase complex subunit B (SDHB, subunit of complex II) and ubiquinol-cytochrome c reductase core protein 2 (UQCRC2, subunit of complex III) was observed, whereby the SDHB decrease in U251MG cells was even comparable to the decrease of COX2. On the other hand, ATP synthase subunit α (ATP5A, subunit of complex V) exhibited very little reduction under THC/CBD treatment of U251MG cells, which was not noted in U138MG cells (Figure 4E,F).

3.5. Downregulation of NDUFA9 by Combined Treatment with THC and CBD Is Independent of Cannabinoid Receptor Activation, Autophagy and Proteasomal Degradation

Since the greatest decrease in treatment with THC and CBD was in complex I, represented by NDUFA9, different pathways that might be involved in this effect were investigated next.

First, the identification of the upstream trigger for the loss of complex I was addressed. To this end, a possible involvement of the cannabinoid receptors CB₁ and CB₂ and transient receptor potential vanilloid 1 (TRPV1) in the effect of the combination of THC and CBD was investigated using antagonists against CB₁ (AM-251), CB₂ (AM-630) and TRPV1 (capsazepine). The receptor antagonists were used at concentrations of 1 µM, which were previously shown to inhibit the activation of CB₁, CB₂ and TRPV1 [25,26,27,28]. With the exception of AM-251, which on its own led to a reduction in NDUFA9 that was significant in U138MG cells, AM-630 and capsazepine exhibited no marked effect on basal NDUFA9 expression (Figure 5A,B). More importantly, none of the antagonists showed an inhibitory effect on the downregulation of NDUFA9 induced by the combination of THC and CBD, so that a receptor-independent effect of the combination can be assumed.

Next, the role of autophagy in the reduction of complex I after THC/CBD exposure was investigated, as autophagy often correlates with mitochondrial damage [29,30]. However, simultaneous treatment of cells with THC/CBD and the late-phase autophagy inhibitor bafilomycin A₁ had no effect on NDUFA9 levels (Figure 5C,D), suggesting that autophagy occurs downstream of complex I loss.

Finally, a possible role of the ubiquitin-proteasome system was addressed, which may exert both protective and detrimental effects on mitochondrial biogenesis, including complex I [31]. Corresponding inhibitor experiments were carried out with the proteasome inhibitor bortezomib, an N-protected dipeptide whose boron atom binds with high affinity and specificity to the 26S proteasome [32]. As shown in Figure 5E,F, however, the synergistic inhibitory effect of the combination of THC and CBD could neither be abolished nor partially blocked in this way. In the same context, MG-132, another potent and cell-permeable proteasome inhibitor, also showed no effect on cannabinoid-induced NDUFA9 decline in U251MG cells (Figure S2). Since MG-132 also inhibits LONP1 [33], a mitochondrial protease that has been reported to be activated in response to stress [34], this result further suggests that LONP1 is not involved in NDUFA9 turnover. This view is also supported by the absence of cannabinoid-induced upregulation of LONP1 (Figure S3A,B).

Overall, it seems unlikely that induced protein degradation is the cause of the reduced NDUFA9 levels after treatment with THC/CBD.

The corresponding author’s preclinical experimental work on the effects of cannabinoids and chemotherapeutic agents is currently funded by the German Research Foundation (DFG, HI 813/9-1), the joint research project ONKOTHER-H (supported by the European Social Fund, reference: ESF/14-BM-A55-0002/18, and the Ministry of Education, Science and Culture of Mecklenburg-Vorpommern, Germany) and by the joint research projects RESPONSE FV18 (reference: 03ZZ0928A) and RESPONSE TV3 (reference: 03ZZ0933A) of the Federal Ministry of Education and Research (BMBF), Germany. Furthermore, we thank Ute Schulz for expert technical support with cell embedding and sectioning for electron microscopy.