Modulation of GPR55 and PINK1-Mediated Cytotoxicity in Glioblastoma Multiforme Models by Cannabis sativa and Piper nigrum Metabolites: A Potential Role for PD-L1

1. Introduction

Gliomas are the most common brain tumors, representing approximately 80% of malignant tumors in the central nervous system (CNS) [1,2]. The annual incidence of primary malignant brain tumors is approximately 7 per 100,000 individuals, and it increases with age, peaking in individuals aged 75-84 years [3]. In this context, glioblastoma (GB) is the most common malignant brain tumor in the adult population, being more frequent in men than in women [4]. GB has a progression-free survival (PFS) of around 7 months, with an average overall survival (OS) of less than 15 months [3,5,6].

The molecular diagnosis of GB requires the presence of wild-type isocitrate dehydrogenase (IDHwt) enzyme, along with microvascular proliferation and/or necrosis. Additionally, at least one of the following three criteria must be met: amplification of chromosome 7 and monosomy of chromosome 10, leading to amplification of the epidermal growth factor receptor (EGFR), and mutations in the telomerase reverse transcriptase (TERT) promoter [7,8].

Etiological factors associated with GB include genetic factors such as Li-Fraumeni syndrome, Turcot syndrome, neurofibromatosis type 1, among others, as well as mutations in genes like TP53, PTEN, and EGFR [9]. Ionizing radiation is also a recognized risk factor for GB development [3]. Exposure to low-frequency electromagnetic fields and the use of mobile phones have been hypothesized to correlate with an increased risk of gliomas, including GB; however, the available data is inconclusive, and no definitive association between these factors and the occurrence of GB has been demonstrated to date [3].

Regarding treatment, the Stupp protocol remains the standard of care, which includes maximal surgical resection, radiotherapy (RT), and chemotherapy, typically with temozolomide (TMZ) [10,11]. Despite these treatments, the prognosis remains poor, and the 5-year survival rate continues to be extremely low. Currently, therapeutic approaches such as tumor treating fields (TTF) [12], immunotherapy, and agents like bevacizumab [13] are being investigated, especially in the case of recurrences [14]. However, no significant improvements in survival have been achieved since the Stupp protocol became the standard of care in 2005 [8].

In this context, GB, as a terminal pathology, presents numerous challenges not only in terms of overall survival but also in maintaining patients’ quality of life [15]. Due to its anatomical location, high recurrence rate, and the adverse effects associated with standard therapies, GB imposes a significant burden of morbidity and mortality [15]. Consequently, beyond identifying potential molecular vulnerabilities to refine diagnosis and prognosis, there is a pressing need for novel therapeutic strategies. These strategies should adopt a multimodal or combinatorial approach to effectively address the tumor’s high recurrence and intrinsic heterogeneity. In this regard, proteins such as PTEN-induced kinase 1 (PINK1), as well as components of the endocannabinoid system (ECS), including the G protein-coupled receptor 55 (GPR55), may offer promising avenue at least in part due to the wide range of cellular processes they regulate [16,17,18,19,20,21,22,23,24,25,26].

PINK1 encodes a serine/threonine kinase originally linked to the pathogenesis of early-onset Parkinson’s disease [27]. It has been shown to play a protective role against neuronal cell death under both physiological and stress conditions [27]. Dysfunction in PINK1 has been associated not only with neurodegenerative diseases but also with cardiovascular, pulmonary, inflammatory disorders, diabetes, and cancer [20,27,28,29].

One of the most widely studied functions of PINK1 is its role in initiating mitophagy. It is often overexpressed to maintain mitochondrial homeostasis, particularly during chemoresistance and targeted therapy in various cancer types, acting as a tumor-protective factor [30]. Furthermore, growing evidence indicates that PINK1’s activity is not limited to mitophagy. Depending on the subcellular and submitochondrial context, it may also be involved in regulating cell survival [20], metabolism [31], immune responses [21], among other processes [32].

GPR55, a non-classical cannabinoid receptor coupled to G proteins, is expressed in various tissues and organs, including the central nervous system (striatum, hypothalamus, microglia), spleen, bone marrow, platelets, neutrophils, and lymphocytes, suggesting a role in glucose regulation, energy balance, cancer progression, inflammation, and immune responses [23,24]. It has also been identified in several cancer types, such as glioma, prostate, breast, and pancreatic cancer [22,23,33,34,35,36].

Lysophosphatidylinositol (LPI), considered its endogenous ligand, is part of a family of bioactive lipids and has been found at elevated levels in cancer patients, correlating with poor prognosis [22,37,38]. These findings suggest that targeting GPR55 could hold therapeutic potential for cancer, pain, and bone-related diseases [38,39]. As a component of the ECS, a ubiquitous lipid signaling network that modulates key physiological activities across the nervous, endocrine, immune, circulatory, gastrointestinal, and reproductive systems [40], its dysregulation has been associated with inflammation, neurodegeneration, and cancer [36,42,43,44].

In parallel, Cannabis sativa and Piper nigrum derivatives have a long-standing traditional use for various diseases, including cancer [45,46]. Notably, cannabinoid-based drugs such as Marinol^® (since the 1980s) and Cesamet^® (a synthetic THC analog) have been approved for nausea and vomiting induced by chemotherapy in some countries [47]. A substantial body of in vitro and in vivo studies supports their potential anticancer effects, including antiproliferative, cytotoxic, and adjuvant properties [48,49,50,51].

Taking all this into account, previous work by our group [42] has identified deregulation of PINK1 and GPR55 in malignant glioma, suggesting their potential as therapeutic targets for GB control and progression. In the present study, we performed functional assays included MTT, WST, and LDH for cell viability, apoptosis assays, mitochondrial membrane potential assessments, RT-qPCR, immunofluorescence (IF) and western blot to evaluate whether Cannabis sativa and Piper nigrum compounds directly affect components of the PINK1 and GPR55 signaling pathways.

Although the cytotoxic effects of Cannabis in various cancer models are widely documented, and its application in gliomas and GB has reached clinical stages (NCT05753007, NCT05629702, NCT03529448), its true therapeutic value and safety profile remain subjects of ongoing debate and emerging research, such as the study presented herein.

2. Materials and Methods

2.1. Compounds

Cannabigerol (CBG, CAS No. 25654-31-3, purity 99%), cannabidiol (CBD, CAS No. 13956-29-1, purity 99%), cannabichromene (CBC, CAS No. 20675-51-8, purity >50%), and non-psychoactive cannabinoid fractions were obtained from Lasanta SAS, a Colombian medicinal cannabis company based in Bogotá, Cundinamarca. These compounds were derived from Cannabis plants legally registered with the Colombian Agricultural Institute (Instituto Colombiano Agropecuario, ICA, Bogotá, Cundinamarca; Resolution No. 00010394 of 2019). An ethanolic extract and essential oil of Piper nigrum were also prepared by GIFFUN, using commercially sourced fruits cataloged in the Medellín Herbarium (Voucher No. AIP 9171 JAUM-101230). Additional compounds used for comparative purposes included temozolomide (TMZ, CAS No. 85622-93-1) and osimertinib (CAS No. 1421373-65-0), both of which were available through the inventory of the Universidad Nacional de Colombia.

Ethidium bromide and acridine orange (EB/AO Thermo Fisher Scientific, Cat. No. 15585011 and Cat. No. A1301) were available at the Universidad Nacional de Colombia.

2.2. Sample Preparation

Stock solutions were prepared by dissolving 5 mg of each pure compound in 1 mL of filtered dimethyl sulfoxide (DMSO; 0.22 μm pore size), resulting in a final concentration of 5 mg/mL. From each stock solution, a 5 μL aliquot was diluted into 1 mL of culture medium (RPMI or DMEM) supplemented with 2% fetal bovine serum (FBS), yielding a working concentration of 25 μg/mL. Serial dilutions were then performed to generate dose–response curves for each compound. For the cannabinoids and Piper nigrum fraction, the effective concentration was adjusted based on the known purity of the fraction to ensure consistency across experimental conditions. This allowed for accurate standardization of dosing across samples with varying levels of purity.

2.3. Cell Lines and Primary Culture

The cell lines utilized in this study included U87MG (provided by the Universidad Nacional de Colombia), and CCF-STTG1 (donated by Universidad de los Andes). All cell lines were maintained in Dulbecco’s Modified Eagle Medium (DMEM) or RPMI medium, supplemented with 10% fetal bovine serum (FBS), 1% L-glutamine, and 1% penicillin/streptomycin. Cells were cultured in T-75 or T-25 flasks and incubated at 37 °C in a humidified atmosphere containing 5% CO₂. For primary glioblastoma cultures, one tumor sample was collected from patients following ethical approval granted by the institutional ethics committee (Resolution 20252000-006214-1). The process was coordinated by the principal investigator in collaboration with clinical staff at Hospital El Tunal / Subred Integrada de Servicios de Salud Sur E.S.E. at Bogotá, Colombia. Primary cultures were established under sterile conditions in a laminar flow hood. Tumor tissues were washed with phosphate-buffered saline (PBS) containing 2% penicillin/streptomycin, mechanically cleaned to remove necrotic areas, and finely minced into ~1 mm fragments. The tissue fragments were enzymatically dissociated using trypsin at 37 °C for 15–20 minutes. The enzymatic reaction was stopped by adding twice the volume of supplemented DMEM, and the suspension was centrifuged at 1000 rpm for 3 minutes. The resulting cell pellet was resuspended in fresh medium, further dissociated by pipetting, and seeded into sterile culture dishes. All cultures were labeled with cell line identification, passage number, date, and operator initials, and maintained at 37 °C in a 5% CO₂ incubator.supplemented DMEM (containing 15% FBS and 1% penicillin/streptomycin), and the cell suspension was centrifuged at 1000 rpm for 3 minutes. The pellet was resuspended in fresh medium, mechanically dissociated by pipetting, and seeded in sterile culture dishes. All cultures were labeled with cell line ID, passage number, date, and operator initials, and incubated at 37 °C with 5% CO₂.

2.4. Cell Viability Tests

2.4.1. MTT Assay

Cells were seeded at a density of 5 × 10³ cells per well in 96-well plates and incubated overnight at 37 °C in a 5% CO₂ atmosphere to allow adherence. Cells were then exposed to treatments for 12 hours. Following treatment, 0.83 mg/mL of MTT reagent (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) was added to each well and incubated for 4 hours at 37 °C to facilitate the formation of formazan crystals. After incubation, the supernatant was carefully removed, and the resulting crystals were solubilized using dimethyl sulfoxide (DMSO). Absorbance was measured at 540 nm using a spectrophotometer.

2.4.2. WST Assay

Cell viability was assessed using the WST-1 reagent (Abcam, Cambridge, UK; Cat. ab65473). A volume of 10 µL of WST-1 was added to each well of a 96-well plate containing cells, followed by incubation at 37 °C for 1 hour. The plate was then placed on an orbital shaker for 15 minutes to ensure uniform mixing. Absorbance was measured at 440 nm using a microplate reader.

2.4.3. LDH Assay

Cytotoxicity was further assessed by measuring lactate dehydrogenase (LDH) release using an LDH ELISA kit (Abcam, Cat. ab183367). Subconfluent cells were mock-treated or treated with samples (e.g., Piper nigrum extract) for 24 hours. Post-incubation, culture supernatants were collected and centrifuged at 2,000 × g for 10 minutes at room temperature. A volume of 50 µL of the clarified supernatant was used per well in accordance with the manufacturer’s instructions. Absorbance was recorded at 490 nm, and the concentration of extracellular LDH was calculated based on a standard curve included in each experiment. The mock-treated control condition was normalized to 1.0, and all other experimental conditions were expressed relative to this control.

All assays were performed in triplicate. Cell viability was calculated as a percentage relative to untreated control wells. Absorbance data were processed using Magellan™ software (Tecan, v7.2), and dose–response curves were used to determine half-maximal inhibitory concentration (IC₅₀) values for each compound.

2.4.4. Ethidium Bromide and Acridine Orange (EB/AO) Staining Assay

For all cell lines, 96-well plates were centrifuged at 1,000 RPM (129 × g) for 5 minutes using a centrifuge equipped with inserts for 96-well plates. A dye mix containing 100 µg/mL acridine orange and 100 µg/mL ethidium bromide in PBS was prepared, and 8 µL of this solution was added to each well. Cells were immediately observed under a fluorescence microscope to assess viability and cell death. Each condition was tested in triplicate, and a minimum of 100 cells per sample were counted for analysis.

2.5. RT-qPCR for Gene Expression

Total RNA was isolated from treated cell cultures previously grown in 6-well plates for 24 h using the RNeasy Mini Kit (Qiagen), following the manufacturer’s instructions. RNA purity and concentration were assessed via OD260/280 using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific). For each sample, cDNA synthesis and amplification were performed using the Luna^® Universal One-Step RT-qPCR Kit (New England Biolabs). Reactions (20 µL total volume) contained 10 µL of 2X Reaction Mix, 1 µL of 20X WarmStart RT Enzyme Mix, 0.8 µL each of forward and reverse primers (10 µM), template RNA (<1 µg), and nuclease-free water to volume. Reactions were assembled on ice and run in triplicate.

Thermocycling was conducted on a StepOnePlus™ Real-Time PCR System (Applied Biosystems) with the following program: reverse transcription at 55 °C for 10 min; initial denaturation at 95 °C for 1 min; 40–45 amplification cycles consisting of 95 °C for 10 sec and 60 °C for 30 sec with plate read; and a final melt curve analysis from 60 °C to 95 °C. Relative gene expression was determined using the 2^−ΔΔCt method, normalizing to β-Actin. Primer specificity was verified by melt curve analysis.

2.6. Immunofluorescence

Cells were seeded onto sterile glass coverslips placed in 6-well plates. Fixation was performed with 100 μL of 4% (w/v) glutaraldehyde and 4% (w/v) sucrose in 1× PBS supplemented with 1 mM Ca²⁺ and Mg²⁺ for 15 min at room temperature (RT). After fixation, cells were washed twice with 1× PBS (5 min each) and permeabilized with 100 μL of 0.1% (v/v) Tween-20 in PBS for 20–30 min at RT, followed by two additional washes with PBS.

Non-specific binding was blocked with 100 μL of TPBS (0.1% Tween-20 in PBS) for 1–2 h at RT. Primary antibody (mouse) was applied at 50–100 μL per well and incubated overnight at 4 °C in a humidified chamber. The following day, coverslips were washed three times with TPBS (5 min each) and incubated with Alexa Fluor 488-conjugated anti-mouse secondary antibody (50–100 μL per well) for 1–2 h at RT, protected from light. After two washes with TPBS, cells were stained with phalloidin (50–100 μL per well, 20 min, RT) and subsequently washed with PBS.

Nuclear counterstaining was performed with DAPI (5 mM, 10 min, RT), followed by two final washes with PBS. Coverslips were then mounted on slides for fluorescence imaging.

2.7. Western Blot Analysis

Protein extracts were obtained from treated cell cultures seeded in 6-well plates and incubated for 24 h. Lysis was performed using RIPA buffer supplemented with protease and phosphatase inhibitors (Thermo Fisher Scientific). Protein concentration was measured using the BCA assay. Equal amounts (30 µg) of total protein were resolved on 10% SDS-PAGE gels and transferred onto PVDF membranes.

Membranes were blocked in 5% non-fat dry milk in TBS-T for 1 hour at room temperature, followed by overnight incubation at 4 °C with primary antibodies (PINK1 and GPR55) targeting the proteins of interest. After washing, membranes were incubated with HRP-conjugated secondary antibodies for 1 hour. Detection was performed using an ECL kit, and bands were visualized using chemiluminescence. Densitometric analysis was conducted using ImageJ software, and β-actin was used as the internal loading control.

2.8. Mitochondrial Potential

Mitochondrial mass and membrane potential (ΔΨm) were assessed using MitoTracker dyes (Thermo Fisher Scientific). Cells were cultured to appropriate confluence, washed with PBS, and incubated with a pre-warmed, serum-free medium containing either MitoTracker Green FM (50–200 nM) to evaluate mitochondrial mass, or MitoTracker Orange/Red CMXRos (50–200 nM) to assess mitochondrial membrane potential, at 37 °C for 30–60 minutes. Following incubation, cells were washed twice with PBS to remove excess dye and analyzed by fluorescence microscopy. Nuclear counterstaining was performed with Hoechst 33342 (5–10 µg/mL, 10 minutes at 37 °C). In these assays, reduced accumulation of MitoTracker Orange/Red indicated mitochondrial depolarization, while MitoTracker Green staining provided an internal control of total mitochondrial content.

3. Ethics Statement

The study was conducted in full compliance with the ethical principles of the Declaration of Helsinki. All in vitro and molecular studies using patient-derived glioblastoma cells were performed under a protocol approved by the Institutional Ethics Committee. Written informed consent was obtained from all patients prior to the collection of tumor tissue samples for research purposes.

4. Statistical Analysis

All data are presented as the mean ± standard error of the mean (SEM) from at least three independent experiments. Statistical significance was determined using one-way analysis of variance (ANOVA), followed by a post-hoc test to compare differences between treatment groups. A p-value of less than 0.05 was considered statistically significant. Specific significance levels are indicated in the figures as follows: *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001. All statistical analyses were performed using GraphPad Prism software.

5. Results

Cellular and Molecular Effects of Cannabinoids and Piper Nigrum in Glioblastoma

To determine the half-maximal inhibitory concentration (IC₅₀) of the tested compounds, we first performed cell viability assays using the MTT and WST methods. Treatments with cannabichromene (CBC), cannabigerol (CBG), and the ethanolic extract of Piper nigrum (PiperOH) resulted in a dose-dependent reduction in viability across glioblastoma cell lines and patient-derived primary cultures (Tables 1 and S1). Among the compounds tested, cannabinoids showed the strongest cytotoxicity, in several cases surpassing the positive controls temozolomide (TMZ) and osimertinib (O).

Table 1. Viability assays of GB cell lines and Primary culture of selected compounds from Cannabis sativa and Piper nigrum.

PiperOH (ethanolic extract of Piper nigrum), Primary GB culture derived from patients (GBP2) NC*: Non cytotoxic — : Non tested.

To further validate these findings, we conducted lactate dehydrogenase (LDH) release assays (Figure 1). The increased extracellular lactate levels were consistent with cell death and correlated with the decreased viability observed in MTT and WST assays. Notably, distinct response patterns were observed among the different glioblastoma models, with cannabinoid- and Piper nigrum-derived compounds showing variable but significant effects. In addition to single treatments, drug combinations were evaluated, including a 1:1 mixture of temozolomide and cannabigerol (TMZ+CBG), as well as a 1:1 mixture of PiperOH and cannabigerol (Piper+CBG). As positive controls, we included TMZ, the standard-of-care alkylating agent for glioblastoma, and osimertinib (O), a tyrosine kinase inhibitor that targets epidermal growth factor receptor (EGFR), frequently overexpressed in glioblastoma.

Figure 1. LDH assays in selected GB models under different treatments:, a) U87MG, b) CCF-STTG1, c) GBP2, Treatment used across the GB models: Cannabichromene (CBC), Cannabidiol (CBD), Cannabigerol (CBG), Piper nigrum ethanolic extracts (Piper), Temozolomide (TMZ), Osimertinib (O). List of treatment mixed applied: TMZ + Cannabigerol (TMZ+CBG), PiperOH+Cannabigerol (Piperr+CBG), ns: non significant.

Finally, to explore the mode of cell death, we performed apoptosis assays using ethidium bromide/acridine orange (EB/AO) staining. EB/AO demonstrated that treatments and TMZ modulated cell viability and death across the glioblastoma cell lines U87MG and CCF-STTG1. In all two models, a notable decrease in the viable cell population was observed following treatments with CBG, PiperOH (Piper) and TMZ, which corresponded with an increase in the apoptotic and necrotic cell fractions (Figure 2). While apoptosis was evident in several conditions, additional non-apoptotic death pathways were particularly evident following treatment with CBG and PiperOH, suggesting multiple mechanisms of cytotoxicity.

Figure 2. Ethidium bromide and acridine orange (EB/AO) staining assay. (a-c) Cell population distribution was assessed by EB/AO staining in glioblastoma cell lines. U87MG (a) and CCF-STTG1 (b) after 24 hours of treatment. The cell populations are distributed as percentages of viable cells (green bars), apoptotic cells (yellow bars), and necrotic cells (red bars). (c) Representative fluorescence microscopy images of CCF-STTG1, and U87MG cells subjected to the different treatments. The images show characteristic cell morphology for viable cells (green nuclei), apoptotic cells (condensed yellow/orange nuclei), and necrotic cells (red nuclei). The scale bar for all images is 30 µm.

Differential Gene and Protein Expression in Glioblastoma Models following Cannabinoid and PiperOH Treatments.

After assessing cell viability and cytotoxicity in the different glioblastoma models, we next explored the potential mechanisms by which the selected cannabinoids (CBG, CBC) and the ethanolic fraction of Piper nigrum (PiperOH-Piper) exert their effects on PINK1, GPR55, and related signaling pathways, the agents assessed in this study are listed in Table 2. To this end, we performed PCR analysis and Immunofluorescence assays of panel of genes, including PINK1, GPR55, cannabinoid receptors 1 and 2 (CNR1, CNR2), phosphoinositide-dependent protein kinase-1 (PDK1) and O-6-methylguanine-DNA methyltransferase (MGMT) (Table 2; Figures 3 and 4). Additionally in supplementary material we include analysis of PDL1 too. Across all tested models, some treatments induced dose-dependent alterations in the relative expression of these genes compared with control conditions.

Table 2. Genes and proteins analyzed, their main functions in cancer, and representative references.

Figure 3. PCR Assays in Glioblastoma and Their Modulation by Cannabinoids, PiperOH (Piper) and Temozolomide. The relative mRNA expression of key genes was evaluated by PCR across three distinct glioblastoma models: the cell lines U87MG and CCF-STTG1, and a patient-derived primary culture (GBP2). The data for each gene was normalized to β-actin and are presented as the mean ± SEM. Expression of PINK1, GPR55, CNR1, CNR2, PDK1 and MGMT in U87MG (a). U87MG cells: Cells were treated with CBG, PiperOH, and TMZ. CBG treatment significantly increased the expression of PDK1 (p < 0.0001), MGMT (p = 0.0095), CNR1 (p = 0.0212), CNR2 (p < 0.0001), and PINK1 (p < 0.0001). In contrast, CBG significantly decreased the expression of GPR55 (p = 0.0009). TMZ decreased GPR55 expression (p = 0.0010), but had no significant effect on the other genes. (b) GBP2 cells: Cells were treated with CBC, CBG, PiperOH, and TMZ. TMZ treatment significantly increased PDK1 expression (p = 0.0451) while significantly decreasing PINK1 expression (p = 0.0088). CBG (p = 0.0302) and PiperOH (p = 0.0310) also significantly decreased PINK1 expression. Abbreviations are: a.u. (arbitrary units) and ns (not significant). Statistical significance compared to the control group is denoted by asterisks: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.

MGMT serves as a crucial biomarker for predicting how glioblastoma will respond to treatment, specifically with temozolomide. CNR1 is widely expressed in the central nervous system, there are currently no approved cannabinoid-based therapies for brain cancer

We analyzed the relative mRNA expression of several cancer-related genes in U87MG and CCF-STTG1 cell lines, as well as in a patient-derived primary culture (GBP2). The results are summarized in Figure 3 for U87MG and GBP2, and in Supplementary material Figure S1 for CCF-STTG1.

Figure 3 illustrates the complex transcriptional responses of key cancer-related genes, including PINK1, GPR55, CNR1, CNR2, PDK1, and MGMT, to various cannabinoid treatments, PiperOH and TMZ in distinct glioblastoma models. The data, normalized to β-actin, are presented as mean relative mRNA expression (a.u.) ± SEM. In established glioblastoma cell lines (U87MG, CCF-STTG1) , CBG robustly modulates the expression of several pro-tumorigenic genes, including GPR55, and PDK1. Notably, CBG consistently increased the expression levels across both cell lines. TMZ, surprisingly, shows minimal impact on the expression of most of these genes in these cell line models.

In a patient-derived primary glioblastoma culture (GBP2), the response profile diverges significantly. Here, PDK1 expression is markedly induced by TMZ (p = 0.0451), while PINK1 expression is concurrently downregulated by TMZ (p = 0.0088), CBG (p = 0.0302), and PiperOH (Piper) (p = 0.0310)

Finally, we performed immunofluorescence (IF) assays in U87MG and GBP2 cells using selected treatments from this study to evaluate ten proteins: PINK1, GPR55, CNR1, CNR2, PDK1 and its phosphorylated form (p-PDK1), phosphorylated AKT (p-AKT), LC3B, cleaved caspase-3, and MGMT (Figures 4 and S2). CCF-STTG1 cells were excluded from these analyses due to their distinct tumor origin and molecular background compared to U87MG and the patient-derived GBM culture (GBP2), as they are derived from an astrocytoma rather than a glioblastoma, which may influence their biological behavior and treatment response. Treatment with CBG induced modulation of p-PDK1 and p-AKT levels compared with controls. In parallel, LC3B expression was elevated, consistent with enhanced autophagic activity, whereas cleaved caspase-3 was markedly upregulated following exposure to all compounds, indicating activation of apoptotic pathways. MGMT expression across treatments was dependent on the model.

Figure 4. Gene and protein expression modulation in U87MG and GBP2 a) Representative immunofluorescence (IF) images of PINK1, GPR55, and additional markers in U87MG cells under different treatments. (b) IF analysis of PINK1, GPR55, Caspase-3, and LC3B, together with Western blot (WB) of PINK1 and GPR55 with densitometric quantification, in U87MG cells treated with different treatments.(c) IF analysis of PINK1, GPR55, Caspase-3, and LC3B, and WB of PINK1 and GPR55 with quantification, performed in primary glioblastoma cultures (GBP2) under the same treatments. While LC3B expression was elevated, indicating autophagic activation. Cleaved caspase-3 was consistently upregulated across treatments. Figure 4b-c. Western blot analysis of PINK1 and GPR55 in U87MG and GBP2 cells following treatment with control, CBG, CBC, PiperOH, and TMZ. PINK1 expression increases after CBC treatment in the U87MG model but in GBP2 after CBG increases too, in both models, TMZ reduces his expression. By contrast, GPR55 regulation was found to be model-dependent; while its expression was downregulated by CBG, CBC, and PiperOH treatments in U87MG cells, it was conversely upregulated in GBP2 cells with the same treatments.

To complement these findings, we performed Western blot (WB) analysis of PINK1 and GPR55 in U87MG and GBP2 cells under four treatment conditions (Figure 4b,c) and a Mitochondrial potential assay of U87MG and GBP2 (Figure 5). Both CBG and CBC Modulate GPR55 expression in both models. But type of modulation (e.g., reduce or increase) varied depending on the model.

Figure 5. Effects of Cannabinoids, PiperOH and Temozolomide on Mitochondrial Potential and Mass in Glioblastoma Models. The effects of treatments on mitochondrial potential and morphology were evaluated in U87MG, and GBP2 cells. (a,c) Representative fluorescence microscopy images of U87MG cells following treatment. Cells were stained with Hoescht 33342 (blue) for nuclei, Phalloidin (green) for F-actin filaments, MitoTracker Green (green) for mitochondrial mass, and MitoTracker Orange (red) for mitochondrial membrane potential. Images illustrate changes in cellular morphology, cytoskeletal integrity, and mitochondrial health across different treatment conditions. (b,d) Quantitative analysis of mitochondrial potential, mass and their ratio. their.Data are presented as mean relative fluorescence intensity per cell (a.u.) ± SEM. The ratio of mitochondrial potential to mitochondrial mass (ΔΨm/Mt Mass) is also shown. Abbreviations: Cannabichromene (CBC), Piper (PiperOH), CBG (cannabigerol), TMZ (temozolomide), TMZ + CBG 1:1 (T+G). MitoTracker, a.u. (arbitrary units). Statistical significance is denoted by asterisks: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

a)

U87MG

Staining of glioblastoma cells with Hoechst 33342 (nuclei, blue), MitoTracker Green (mitochondrial mass), and MitoTracker Orange (mitochondrial membrane potential) 24h.

• c) GBP2

Staining of glioblastoma cells with Hoechst 33342 (nuclei, blue), MitoTracker Green (mitochondrial mass), and MitoTracker Orange (mitochondrial membrane potential) 24h.

d)

6. Discussion

An intriguing question in cancer biology concerns the pleiotropic effects of cannabinoids across diverse diseases, ranging from neurodegeneration [56] to cancer [47]. We hypothesize that these broad effects are driven, at least in part, by the molecular targets of cannabinoids, which regulate key processes in cellular homeostasis. Our study provides compelling evidence that cannabinoids such as CBG and CBC, as well as Piper nigrum extract, exert cytotoxic effects on GB cells by disrupting mitochondrial function. These effects are marked by an alteration of mitochondrial membrane potential (ΔΨm), LDH release levels, and morphological changes consistent with apoptosis [57]. These data, together with the information shown in Figure 5, position mitochondrial dysfunction as a potential mechanism underlying cannabinoid-induced GB cell death.

We first found that cannabinoids and PiperOH significantly affect the viability and cytotoxicity of the studied models (Table 1, Table 1S, and Figure 1). Although these effects are context-dependent, as is the case in many other cancer models, both compound groups demonstrate dose-dependent effects superior to those of the chemotherapy control (3-20 ng/μL of cannabinoids and PiperOH vs. doses greater than 50 ng/μL of TMZ).

Additionally, while apoptosis is the primary mechanism of cell death in most cases, our results (Figure 2) for the cannabinoids CBC and CBG indicate they also induce other forms of non-apoptotic cell death (e.g., autophagy, ferroptosis). This observation aligns with our hypothesis that modulating proteins like PINK1, which regulates both apoptosis and mitophagy, is a key mechanism of action. Notably, in the CCF-STTG1 model (Figure 2b), a mixed GB model (anaplastic astrocytoma grade IV), TMZ also induces a non-apoptotic form of cell death, which is not observed in the U87MG cell line (Figure 2a).

Building on previous in silico and in vitro studies from our group [42] that identified the potential for cannabinoids and Piper nigrum derivatives to interact with PINK1 and GPR55 as molecular targets, we conducted a series of experiments. We measured gene expression levels (RT-qPCR) in U87MG, CCF-STTG1, and GBP2 cell lines. For the U87MG and GBP2 models, we performed a comprehensive protein characterization of PINK1 and GPR55 using immunofluorescence (IF) and Western blot (WB). We also evaluated other proteins by IF, such as Caspase-3 (apoptosis) and LC3b (to assess autophagy/mitophagy levels), to complement our acridine orange (AO/EO) assay results.

Our findings reveal that gene and protein expression levels for PINK1 and GPR55 vary significantly depending on the model and the compound (Figures 3 and 4, and Supplementary Figures S1 and Figure S2). In U87MG cells, PINK1 mRNA levels were increased, but this change was not statistically significant at the protein level via IF or WB. Conversely, in GBP2 cells, CBG reduced PINK1 levels in both RT-qPCR and IF analyses, yet a statistically significant increase in protein levels was observed via WB.

For CBC, no significant differences were found via IF in U87MG cells, but an increase in PINK1 protein levels was observed via WB. The opposite trend was noted in GBP2 cells, where no changes were seen via RT-qPCR or WB, but a decrease was found via IF. For PiperOH, no significant differences were found in U87MG cells, but a decrease in PINK1 levels was observed via RT-qPCR and IF in GBP2 cells, though not in WB. Finally, it is noteworthy that the positive control, TMZ, yielded highly consistent results across both models showing a reduction in PINK1 levels, which not only aligns with existing literature [58] but also suggests a potential link between PINK1 and TMZ resistance in glioblastoma [59].

Regarding GPR55, we observed that both mRNA (RT-qPCR) and protein (Western Blot) levels decreased in U87MG cells following treatment with CBG, CBC, and PiperOH. Conversely, in GBP2 cells, these same treatments had an opposite effect: a reduction in gene expression and protein levels via immunofluorescence, but an increase in protein levels measured by WB This finding demonstrates that compounds derived from Cannabis spp and Piper nigrum modulate this protein in a cell subtype-dependent manner. While GPR55 is considered an oncogene, this differential modulation and its antitumor effects are consistent with observations from other studies, which suggest that both agonism [23] and antagonism [60] of this receptor can have therapeutic value.

Complementary to these findings, we evaluated other potential markers that may interact with or be members of the PINK1 and GPR55 signaling pathways (Figure 6) and that play important roles in the tumor context. We investigated the expression of Caspase 3 and LC3b to better understand the involvement of apoptosis and autophagy, respectively, as mechanisms of cell death. We also analyzed PDK1 and AKT, key proteins involved in the modulation of cellular metabolism, growth, and survival. Additionally, we evaluated CNR1 and CNR2, the primary receptors of the endocannabinoid system, which are known to have multiple functions in cellular homeostasis, proliferation, and death [36,52]. Finally, MGMT, a critical protein for DNA repair and a key player in glioblastoma therapy and prognosis. As hypomethylation of its promoter leads to increased expression and subsequent chemotherapy resistance, MGMT constitutes a crucial biomarker in glioblastoma management [61,62].

Figure 6. Schematic Representation of Putative Interactions Between Analyzed Markers and Their Role in Glioblastoma Pathophysiology.

Mitochondrial Homeostasis and Signaling Pathway Interactions

We first analyzed the levels of Caspase 3 and LC3b, finding that the tested compounds generally increased both markers of apoptosis and autophagy/mitophagy, respectively. Among them, CBG demonstrated the most significant activity across all models. We also highlight a notable finding from Figure 4b: while TMZ alone did not activate LC3b, its co-administration with CBG (in a 1:1 ratio, TMZ+CBG) generated a synergistic effect that significantly increased LC3b levels. This suggests a potential co-adjuvant role for CBG. This, along with the results in Figure 5, where we observed that all compounds (including PiperOH) impair mitochondrial membrane potential, strongly indicates that these agents disrupt key components of mitochondrial homeostasis. Notably, in the U87MG model, CBG treatment led to a decrease in mitochondrial potential while simultaneously causing an increase in mitochondrial mass. This suggests a potential mitochondrial damage effect where these organelles lose their functionality, and as a compensatory mechanism, the cell produces more mitochondria, albeit with deficient function. This phenomenon could be partially explained by the modulation of proteins like PINK1, which is crucial for mitochondrial homeostasis.

Regarding the relationship between GPR55 and PINK1 and other analyzed proteins, we found consistent, yet opposing, changes in AKT expression across U87MG and GBP2 following treatment with CBG, PiperOH, and TMZ (see Supplementary Figure S2). In U87MG cells, CBG and PiperOH reduced the expression of p-AKT, which aligns with our viability and cytotoxicity data (Table 1, 1S, and Figure 2). A decrease in p-AKT expression is generally associated with a reduction in tumor growth, while its hyperactivation supports it. Since AKT is a downstream effector of both the PI3K/AKT and GPR55/PINK1 pathways, these results are biologically coherent. In the case of TMZ, the observed increase in p-AKT may represent a compensatory mechanism by glioblastoma cells to mitigate the effects of the chemotherapeutic agent.

Another crucial pathway linked to AKT and PINK1/GPR55 is PDK1, a key protein involved in the Warburg effect, a hallmark of glioblastoma metabolism [52]. We observed in both U87MG and GBP2 cells that CBG decreased PDK1 levels in U87MG, but this was not the case in GBP2, as measured by both immunofluorescence and RT-qPCR. Conversely, CBC and the PiperOH fraction led to an increase in PDK1 levels in U87MG and a decrease in GBP2. The use of TMZ+CBG also had a differential effect in the cell models, once again highlighting the inherent heterogeneity of glioblastoma. These findings could correlate with the modulation of PINK1, which regulates mitochondrial metabolism, and the GPR55-PI3K/AKT axis, which also alters PDK1 levels. This inverse relationship could be explained by the fact that PDK1 promotes glycolysis, while PINK1 facilitates oxidative phosphorylation. Notably, TMZ exhibited a similar profile, inducing PDK1 in U87MG as a tumor survival mechanism against TMZ-induced damage, which is consistent with existing literature [63].

We also analyzed the levels of phosphorylated PDK1 (p-PDK1), as illustrated in Figure S2. Notably, both CBG and PiperOH reduced p-PDK1 levels in U87MG cells, while no significant differences were found with the other treatments. Conversely, in GBP2 cells, CBC, TMZ, and the combined T+G treatment lowered p-PDK1 levels. It is worth noting that while CBG and PiperOH increased overall PDK1 levels in GBP2, they decreased its phosphorylated form. Furthermore, CBC and TMZ+CBG consistently reduced both total and phosphorylated PDK1 levels in GBP2, while CBG had a similar effect on both forms in U87MG. These findings underscore the complex and model-dependent regulation of this key metabolic protein.

In our experiments, we also analyzed the canonical endocannabinoid receptors, CB1 (CNR1) and CB2 (CNR2). Although the endocannabinoid system is ubiquitous and can be considered part of the larger bioactive lipid system, it has few first-line clinical applications, largely due to its complex and pleiotropic nature. In the context of glioblastoma, these receptors are crucial for maintaining tumor survival and proliferation, given their relationship with GPR55. In U87MG cells, we found no significant differences in the expression of CNR1 or CNR2 across treatments. However, in GBP2 cells, we observed a general decrease in CNR1 levels across all treatments, while CNR2 levels were only significantly reduced by the TMZ+CBG combination. This suggests that the compound’s anti-tumor effects are not primarily mediated by these canonical receptors in the U87MG model but may involve the modulation of CNR1 in GBP2 cells, a point that requires further dedicated investigation.

Finally, we analyzed MGMT (O-6-methylguanine-DNA methyltransferase), a DNA repair enzyme that removes O-6-methylguanine adducts, thereby protecting against DNA damage induced by alkylating chemotherapeutic agents like TMZ. A higher degree of MGMT promoter methylation predicts greater chemosensitivity due to low MGMT repair activity. Clinically, patients with MGMT promoter methylation have a median overall survival of 21.7 months, compared to 12.7 months in the unmethylated population [61]. We found that protein levels of MGMT were decreased by all treatments in U87MG cells, but only by the TMZ+CBG combination in GBP2 cells. This reduction is significant due to MGMT’s critical role in the TMZ resistance. Conversely, while we found an increase in MGMT mRNA levels in U87MG treated with CBG, and in GBP2 treated with PiperOH and TMZ, further studies are needed to determine if this is a cellular response to injury or another compensatory mechanism.

Within the supplementary material of this work (Figure S1), we present the RT-qPCR results for the CCF-STTG1 glioblastoma cell line. In this model, the use of CBG altered (increased the levels of) nearly all evaluated markers, with the exception of CNR1 and MGMT, for which no significant changes were observed. Additionally, we analyzed the levels of PD-L1, a crucial marker of immune response and a prognostic factor for immunotherapy. CBG demonstrated varied effects on PD-L1, increasing its levels in U87MG and CCF-STTG1 cells while decreasing them in GBP2. This finding holds great significance. Traditionally, the immunomodulatory properties of cannabinoids have been primarily attributed to the activation of the CNR2 receptor and their antioxidant activity [64]. However, given the known molecular promiscuity and complex pharmacodynamic interactions of these compounds, our results suggest an alternative mechanism. The modulation of PD-L1 by CBG could be mediated through its interaction with PINK1, which is consistent with studies in other tumor models that have shown PD-L1 expression can be regulated by PINK1 via mechanisms such as ubiquitination or mitochondrial iron accumulation [65,66,67].

The data presented here, along with the well-established clinical role of the MGMT gene, highlights the critical challenges in GB therapy. In approximately 40–45% of GBs, MGMT promoter hypermethylation silences gene expression, conferring sensitivity to TMZ and extending overall survival (median ~24.5 months vs. ~14 months for unmethylated tumors) [61,62]. However, most GBs retain an active MGMT promoter, leading to efficient DNA repair and profound TMZ resistance. This critical challenge highlights a major clinical bottleneck, as the standard-of-care Stupp protocol for GB has remained largely unchanged for the past 25 years. Our prior work [42] demonstrates that PINK1 is highly expressed in ~70% of GB cases, while absent in normal astroglia. This suggests that modulating PINK1 could be a more plausible therapeutic strategy, provided a suitable modulator is found, as it would benefit a significantly larger patient population than the 30% that respond to TMZ based on MGMT status. By analogy, our data suggest that PINK1 may serve as both a prognostic and predictive biomarker, with pleiotropic functions (Figure 7) that vary according to tumor context. Its role in modulating the phosphoproteome and interacting with key oncogenic drivers in GB such as EGFR [68] underscores its potential as a novel therapeutic target, which warrants further investigation.

Figure 7.PINK1 in the context of cancer and its relationship with the lipidome. PINK1 has been studied primarily in the setting of neurodegenerative diseases such as Parkinson’s disease, where loss of function leads to impaired mitophagy, defective mitochondrial fitness, and alterations in its kinase and ubiquitin-mediated activities. In cancer, however, PINK1 exhibits context-dependent roles that may act as analogues or antitheses of its neuroprotective functions. In glioblastoma and other tumors, dysregulated PINK1 activity intersects with lipid metabolism and mitochondrial dynamics, linking altered mitophagy to tumor survival, metabolic reprogramming, and therapeutic resistance.

Our work extends beyond the classical endocannabinoid system (ECS) by highlighting the involvement of GPR55, a receptor highly expressed in GB. Importantly, GPR55 is also a key player in neuroinflammation and a promising therapeutic target for epilepsy [69], a significant and debilitating comorbidity in GB patients. Beyond mitochondrial regulation, the cannabinoid-sensitive receptor GPR55 represents another critical survival pathway in GB. Interestingly, the effect of CBG, CBC, and PiperOH on GPR55 levels was cell-line specific. These treatments led to a reduction in GPR55 expression in U87MG cells but an increase in GBP2 cells, underscoring a divergent regulatory response.

These findings align with the expanded concept of the “endocannabinoidome” (eCBome), a network of lipid mediators, enzymes, and receptors extending beyond the plasma membrane to intracellular organelles such as mitochondria [70]. This intracellular localization may help explain the varied and context-dependent effects of cannabinoids. We propose that cannabinoids coordinate mitochondrial quality control via PINK1 while simultaneously modulating GPR55-dependent signaling, thereby linking lipid-based communication with metabolic reprogramming.

In summary, our study provides new insights into the molecular actions of cannabinoids in GB, highlighting mitochondrial dysfunction, PINK1, and GPR55 as key mechanistic hubs (Figure 8). These findings reinforce the potential of natural compounds as adjuncts to current therapy, while also emphasizing the complexity of their pleiotropic effects. Future work should validate these results in in vivo models and further dissect the interplay between the eCBome, mitochondrial metabolism, and the tumor immune microenvironment. Such efforts may ultimately establish cannabinoids as mediators capable of reprogramming GB cell survival and reshaping the therapeutic landscape of this devastating disease.

Figure 8. Proposed model of cannabinoid-mediated effects in glioblastoma via PINK1 and GPR55. Owing to their physicochemical properties, cannabinoids can act at multiple cellular levels: (1) through membrane receptors such as GPR55 or tyrosine kinase receptors (TKRs, e.g., EGFR), and (2) by passive diffusion or endocytosis, enabling interactions with subcellular organelles, including mitochondria, and intracellular proteins such as PINK1. This diversity of entry routes and molecular targets allows cannabinoids to exert context-dependent effects. In glioblastoma models, these include the induction of apoptosis, autophagy, and ferroptosis, reflecting the broad spectrum of signaling pathways modulated by distinct cannabinoids under specific pathophysiological conditions.

7. Perspectives

A growing body of evidence highlights the potential of natural products as therapeutic agents against complex diseases, including cancer. Within this framework, cannabinoids emerge as promising candidates due to their diverse biological activities. However, research in this area has historically been constrained by the legal status of cannabis, which limits both experimental investigation and clinical translation, as well as by concerns regarding psychotropic effects. Our findings contribute to overcoming these barriers by focusing on non-psychotropic cannabinoids, such as CBC and CBG, which demonstrated anticancer activity in glioblastoma models. These results underscore the therapeutic relevance of non-psychoactive cannabinoids and support their continued exploration as potential adjuncts or alternatives in oncology.

An intriguing aspect of cannabinoids lies in their virtually universal applicability across multiple pathologies [71]. For example, CBD can exert pleiotropic effects, functioning as both a neuroprotective and cytotoxic agent depending on the cellular context [72]. Beyond receptor abundance within ECS, one plausible explanation involves mitochondrial modulation. Indeed, cannabinoids have been shown to alter mitochondrial membrane potential in neurodegenerative disease models, thereby enhancing mitophagy and protecting neurons from oxidative stress [43,44,45,46,47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70,71,72,73]. Conversely, in cancerous cells, similar mitochondrial modulation can drive cytotoxicity by forcing tumor cells into apoptosis [74].

Some cannabinoids thus act as metabolic and lipidomic modulators, enabling them to produce context-dependent outcomes. In glioblastoma, for instance, cannabinoids can inhibit glycolytic regulators such as PDK1, disrupt mitochondrial polarization, and alter the cellular lipidome. Since cannabinoids are themselves bioactive lipids, their ability to remodel the lipidome has profound implications: virtually any structure containing membranes plasma, nuclear, mitochondrial, or vesicular can become a potential target. This provides a mechanistic explanation for their pleiotropy but also highlights the need for further research to precisely define the context-dependent utility of cannabinoids, particularly in cancer where membrane remodeling and metabolic plasticity are tightly linked to tumor progression [75,76].

From an evolutionary perspective, cancer and glioblastoma in particular can be viewed as a breakdown of the endosymbiotic balance that defines eukaryotic identity. Aggressive tumors disrupt this integration by favoring dedifferentiated metabolic states characterized by suppressed mitochondrial respiration and enhanced aerobic glycolysis, a phenomenon known as the Warburg effect. This not only sustains rapid proliferation but also impairs immune and inflammatory regulation [77,78].

In this context, the ECS acts as a regulatory interface capable of influencing the crosstalk between nuclear and mitochondrial subsystems. Receptors such as CNR1 and GPR55, located at both plasma and intracellular membranes, can modulate mitochondrial dynamics, immune tolerance, and metabolic signaling. Importantly, PINK1, a mitochondrial kinase essential for mitophagy, emerges as a convergence node. Our data and recent literature suggest that CBG and CBC are able to modulate PINK1 and GPR55. This conceptual framework positions non-psychotropic cannabinoids not merely as cytotoxic agents but as metabolic-immune modulators capable of rebalancing disrupted cellular funtion in glioblastoma.

8. Conclusions

Our study provides a strong rationale for exploring compounds derived from Cannabis sativa and Piper nigrum as modulators of glioblastoma biology. We’ve shown that these compounds exert significant cytotoxic effects, mediated by modulating the key proteins PINK1 and GPR55. Importantly, our research focused exclusively on non-psychoactive/psychotomimetic cannabinoids such as CBG and CBC, avoiding the usual challenges associated with the use of THC. This approach overcomes a major obstacle often attributed to Cannabis sativa-based therapies. Our results position PINK1 and GPR55 as critical molecular axes, opening the door to a new therapeutic paradigm that bridges phytochemistry with precision oncology. The ability of these compounds to impact mitochondrial homeostasis and metabolic reprogramming suggests a promising path forward to overcome treatment resistance and ultimately offer new options for patients with this devastating disease.