Neuroprotective Role of Cannabinoid CB₁ and GPR55 Receptors in a Cell Model of Multiple Sclerosis

Abstract

Multiple sclerosis (MS) is a chronic neuroinflammatory disorder characterized by demyelination, oligodendrocyte (OLG) loss, and progressive neurological decline. While current disease-modifying therapies primarily target immune responses, they offer limited neuroprotective or remyelinating effects. Through its receptors, including cannabinoid receptor type 1 (CB₁R) and the G protein-coupled receptor GPR55, the endocannabinoid system (ECS) has emerged as a promising therapeutic target due to its roles in neuronal survival and glial function. To better study the functional role of these receptors and their potential as targets for remyelinating therapies, we sought a human cell model that lacks endogenous expression of CB₁R and GPR55. Quantitative PCR confirmed that human OLG-derived HOG cells do not express these receptors, providing a clean background for transfection-based functional studies. Using this system, we found that expressing either CB₁R or GPR55 was sufficient to confer resistance to cuprizone (CPZ)-induced cytotoxicity, while co-expression produced a comparable level of protection but uncovered a distinct pharmacological interaction consistent with functional receptor crosstalk. In situ proximity ligation assays confirmed the formation of CB₁R/GPR55 receptor complexes in co-transfected cells. The protective effects were evident even under serum-containing conditions, suggesting that serum components may provide basal receptor activation. In serum-free conditions, selective activation of CB₁R with arachidonyl-2'-chloroethylamide (ACEA) or GPR55 with CID1792197 enhanced cell survival, while the CB₁R antagonist SR141716 blocked both ACEA- and CID1792197-induced protection. These cross-antagonistic interactions support the presence of heteromer-dependent signaling. Altogether, our findings identify CB₁R/GPR55 heteromers as novel modulators of OLG resilience and highlight their potential as therapeutic targets for promoting neuroprotection and remyelination in demyelinating diseases such as MS.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12035-026-05780-5.

Introduction

Multiple sclerosis (MS) is a chronic, immune-mediated neurodegenerative disorder of the central nervous system (CNS) characterized by progressive demyelination, axonal injury, and irreversible neurological disability [1–5]. Globally, MS is one of the leading causes of non-traumatic incapacity in young adults, with a significantly higher incidence in women than men [6, 7]. Epidemiological studies point to a rising prevalence in developed nations, a trend linked to genetic predisposition compounded by environmental risk factors such as vitamin D deficiency, Epstein–Barr virus (EBV) infection, smoking, and gut microbiota dysbiosis [8–10].

The pathogenesis of MS involves autoimmune activation, particularly of CD4⁺ T lymphocytes and B cells, which target myelin and trigger oligodendrocyte (OLG) apoptosis, neuroinflammation, and gliosis [1–3, 11, 12]. This leads to the formation of demyelinated plaques and progressive neurological decline, manifesting initially in a relapsing–remitting pattern before often transitioning to a secondary progressive stage [13–15]. Although disease-modifying therapies (DMTs), mostly immunomodulatory, have reduced relapse rates and delayed disability, no curative treatment exists and benefits in remyelination or neuroregeneration are limited [9, 16–18]. This gap has prompted recent studies to explore alternative molecular targets beyond traditional immune modulation, shifting the focus toward neuroprotective mechanisms aimed at preserving or restoring CNS function [19–21]. Notably, Sativex^® (nabiximols) is among the most widely approved symptomatic treatments for MS-related spasticity. This oromucosal spray, derived from a standardized Cannabis sativa L. extract, contains Δ⁹-tetrahydrocannabinol (THC) and cannabidiol (CBD) in a 1:1 ratio [22–24]. The sustained clinical success of this formulation over more than a decade has attracted increasing attention to the endocannabinoid system (ECS), which plays a crucial regulatory role in the CNS by modulating synaptic plasticity, neuroinflammation, neuronal survival, and homeostatic responses to injury [25–28]. Consequently, there is increasing interest in the clinical potential of phytocannabinoids and of synthetic cannabinoids, particularly in targeting cannabinoid receptors, as promising candidates for neuroprotective interventions in MS [20, 29, 30].

Central to the therapeutic potential of the ECS are its primary receptors, cannabinoid receptor type 1 (CB₁) and type 2 (CB₂), which mediate diverse neurophysiological and immunomodulatory effects. CB₁R are among the most abundantly expressed G protein-coupled receptors (GPCR) in the CNS, particularly in regions such as the hippocampus, cortex, basal ganglia, and cerebellum [31–34]. They are predominantly localized at presynaptic terminals, where they modulate neurotransmitter release through G_i/o-coupled mechanisms by reducing calcium influx and enhancing potassium efflux [35, 36]. The cascade decreases intracellular cAMP and activates MAPK (ERK1/2, p38, JNK) and PI3K/Akt signaling. These pathways contribute to motor coordination, cognitive function, pain modulation, and regulation of oxidative stress [36–38]. CB₂R, primarily an immune-system receptor, is expressed by microglia and astroglia and is strongly upregulated during CNS injury or inflammation [39–43]. CB₂R also couples to G_i/o proteins, modulating immune responses, reducing neuroinflammation, and contributing to neuronal cell survival [29, 44, 45]. Beyond these classical receptors, GPR55 has emerged as a putative third cannabinoid receptor of growing interest due to its structural divergence and distinct signaling profile. Despite its endogenous ligand is believed to be lysophosphatidylinositol (LPI), GPR55 also responds to various phytocannabinoids and synthetic ligands originally developed for CB₁R, such as rimonabant and AM251 [46–48]. Unlike CB₁R and CB₂R, GPR55 primarily signals through G_α13 and RhoA, activating ERK1/2 and mobilizing intracellular calcium, processes involved in cell proliferation, migration, and cytoskeletal dynamics [36, 47, 49–51]. Although less studied, GPR55 is expressed in several CNS regions, including the striatum and the hippocampus, and has been associated with glial reactivity and neuroinflammatory responses [34, 36, 43, 52, 53].

Given this complex receptor landscape and their distinct signaling mechanisms, attention has turned to how the ECS contributes to the pathophysiology and/or treatment of MS [54, 55]. In this context, CB₁R activation has been shown to promote neuronal and OLG survival, modulate glial reactivity, and reduce oxidative stress, contributing to a neuroprotective milieu [56–59]. In contrast, GPR55 expression is often elevated under inflammatory and demyelinating conditions, suggesting a complex or dualistic role that may depend on context and ligand availability [60]. Nevertheless, the complexity of cannabinoid signaling extends beyond classical receptor-ligand interactions. In fact, an emerging paradigm in cannabinoid pharmacology is receptor heteromerization, which confers novel functional properties distinct from those of individual receptors [61–63]. CB₁R/GPR55 heteromers, for instance, have been identified in neuronal cultures and brain tissue from rodents and primates, particularly in the striatum, where they exhibit biased agonism, cross-antagonism, and unique calcium signaling dynamics [34, 64, 65]. This molecular interaction introduces a new layer of regulation in ECS signaling that may be particularly relevant in pathological states [43, 64]. Although OLGs in vivo are known to express CB₁R, and GPR55 expression has been reported in glial cells, evidence for their co-expression in OLG lineage cells is still lacking. Notably, a recent study by Menéndez-Pérez et al. (2024) provided the first evidence of CB₁R/GPR55 heteromer expression in post-mortem brain tissue of MS patients, specifically in the prefrontal cortex. Using a powerful technique, namely in situ proximity ligation assay (PLA), the authors demonstrated a significant upregulation of these heteromers in MS samples compared to controls [66].

There is a well-established murine model that recapitulates key MS pathologies [67–69]. Administration of the copper-chelating agent cuprizone (CPZ) induces reproducible demyelination in the CNS of mice, affecting both white and gray matter. This effect is mediated by the selective triggering of OLG apoptosis, astrocyte activation, and the generation of proinflammatory microglia [70–74]. Complementary to studies in animal models and to adapt to the methods that replace, reduce, and refine animal use, cell-based CPZ cytotoxicity models are increasingly used to explore the molecular mechanisms underlying MS, thanks to its simplicity, high reproducibility, and applicability in drug screening [75, 76]. The aim of the present study was to investigate the neuroprotective potential of CB₁ and/or GPR55 receptors using the CPZ-induced cytotoxicity model in OLGs of human origin. Specifically, we assessed the effects of selective synthetic cannabinoid ligands, acting as agonists or antagonists of CB₁R and GPR55, to elucidate receptor-dependent mechanisms of action. In addition, we examined possible pharmacological interactions between these receptors that could modulate cellular responses to CPZ-induced damage.

Materials and Methods

Drugs

N-(2-Chloroethyl)−5Z,8Z,11Z,14Z-eicosatetraenamide (ACEA; 1319 Tocris Bioscience, Bristol, UK), 5-(4-chlorophenyl)−1-(2,4-dichlorophenyl)−4-methyl-N-1-piperidinyl-1H-pyrazole-3-carboxamide (SR141716, rimonabant; 168,273–06-1, Cayman Chemical, Michigan, USA), and CID1792197 from PubChem [46], in-house synthesized (> 99% purity), were used in this study. ACEA was obtained as a pale-yellow oil and prepared as a 10 mM stock solution by diluting in ethanol. Stock solutions of SR141617 and CID1792197 were similarly made at 10 mM in dimethyl sulfoxide (DMSO). All aliquots were stored at − 20 °C until needed.

RNA Purification

Cells were collected from culture dishes using scrapers and TRIzol (15,596,018, Invitrogen, Paisley, Scotland, UK). Total RNA was then isolated with the RNeasy Mini Kit (74,104, Qiagen, Valencia, CA, USA), including an on-column DNase digestion (79,204, Qiagen, Valencia, CA, USA), according to the manufacturer’s protocol.

Quantitative Real-Time PCR

First-strand cDNA was synthesized from 1 μg of total RNA using random primers and the SuperScript III kit (11,752,050, Invitrogen, Paisley, Scotland, UK) in a final volume of 20 μL, following the manufacturer’s instructions. Quantitative real-time PCR (qRT-PCR) was performed by combining the cDNA with SYBR Green PCR Master Mix (a25778, Applied Biosystems, Carlsbad, CA, USA) and 0.3 μM of the corresponding forward and reverse primers (Table 1). Gene expression was quantified using a 7300 Real-Time PCR System (Applied Biosystems, Carlsbad, CA, USA). The cycling program included an initial denaturation at 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 1 min. A dissociation curve from 60 to 95 °C was included at the end of the run to confirm the specificity of the amplified products. Relative gene expression levels were determined using a standard curve, and all data were normalized to the housekeeping genes 18S and GAPDH.

Table 1.

Primers used for qRT-PCR in this study

Oligonucleotide		Sequence
CNR1-Fw	––––	CAGATACCACCTTCCGCACC
CNR1-Rev	––––	GTCTCCCGCAGTCATCTTCT
GPR55-Fw	––––	CAGTCCACATCCCCACCTTC
GPR55-Rev	––––	TGGAGGTGGCAGCATAATCG
GAPDH-Fw	––––	AAATCCCATCACCATCTTCC
GAPDH-Rev	––––	GACTCCACGACGTACTCAGC
18S-Fw	––––	ATGCTCTTAGCTGAGTGTCCCG
18S-Rev	––––	ATTCCTAGCTGCGGTATCCAGG

The annealing temperature was 60 °C for all primers. 18S and GAPDH RNAs were used as housekeeping genes

Fusion Proteins and Expression Vectors

Human CB₁R and GPR55 cDNAs were cloned into pcDNA3.1 and amplified without their stop codons using sense and antisense primers containing unique EcoRI or BamHI restriction sites. The resulting fragments were subcloned in-frame with EYFP into the BamHI and EcoRI sites of an EYFP-expressing vector (EYFP-N1; enhanced yellow variant of GFP; Clontech, Heidelberg, Germany), producing plasmids encoding CB₁R-YFP and GPR55-YFP, with the fluorescent tag fused to the C-terminal end of each receptor. Receptor expression was confirmed by confocal microscopy, and receptor functionality was assessed via ERK1/2 activation assays.

Cell Line Culture and Transfection

The HOG cell line, established from a human oligodendroglioma by Dr. A. T. Campagnoni (University of California, UCLA, Berkeley, CA, USA) [77] was generously provided by Dr. J. A. López-Guerrero (Universidad Autónoma de Madrid, Madrid, Spain) [78]. Cells were maintained in low-glucose DMEM supplemented with pyruvate, HEPES (22,320–022, Invitrogen, Paisley, Scotland, UK), 10% (v/v) heat-inactivated fetal bovine serum (FBS) (10,270–106, Invitrogen, Paisley, Scotland, UK), and penicillin/streptomycin at 100 units/mL (17-602E, Invitrogen, Paisley, Scotland, UK). Cultures were kept at 37 °C in a humidified incubator with 5% CO₂ and were subcultured twice weekly once they reached 80–90% confluency, with passages limited to 20 to preserve cell characteristics.

For transient expression, cells in 10 cm dishes were transfected with the appropriate fusion protein cDNAs (see figure legends) using Lipofectamine 3000 (3000–015, Invitrogen, Paisley, Scotland, UK) according to the manufacturer’s protocol. Parallel control cultures received Lipofectamine 3000 without DNA. Cell viability was assessed relative to untreated cells, and no significant cytotoxic effects were detected.

Cell Treatments

Cell treatments were performed 48 h after transfection. For MTT reduction assays, 3,000 cells were seeded in each well of 96-well plates, whereas 30,000 cells per well were plated on 8-well chambered cover glasses (Nunc^® Lab-Tek^® II, Sigma-Aldrich, St. Louis, MO, USA) for immunofluorescence and proximity ligation assays (PLA). Cytotoxicity was triggered by treating cells with CPZ at increasing concentrations (50–1000 μM; see figure legends). A 30 mM CPZ stock solution (C9012-25G, Sigma-Aldrich, St. Louis, MO, USA) was freshly prepared by dissolving CPZ powder in a 50% ethanol/medium mixture and incubating it at 60 °C with shaking (225 rpm) for 15–20 min until full dissolution. Final working concentrations were obtained by diluting the stock in complete medium to keep ethanol content at a minimum [75, 76].

For experiments involving CB₁R and GPR55 ligands, cells cultured in serum-free DMEM for 2 h, were either left untreated or treated with the CB₁R agonist ACEA (100 nM) and the GPR55 agonist CID1792197 (1 μM) for 30 min prior to the addition of freshly prepared CPZ (500 μM). In experiments using the CB₁R antagonist, cells were pre-incubated with the SR141716 (rimonabant; 250 nM) for 20 min before agonist treatment. When multiple compounds were applied, they were added simultaneously to the medium. Concentrations and incubation times were chosen based on previous studies on receptor pharmacology and our own prior experience [46, 64].

Control groups received either the same volume of fresh medium or the corresponding vehicle.

MTT Reduction Assay

Cell viability was assessed using the 3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide (MTT) reduction assay, which relies on mitochondrial NADH-dependent oxidoreductase activity as an indicator of mitochondrial function. After completion of the treatments, 10 µL of MTT solution (5 mg/mL in phosphate-buffered saline, PBS; 5655, Sigma-Aldrich, St. Louis, MO, USA) was added to each well. Following a 4-h incubation, 100 µL of lysis buffer (20% SDS, 50% dimethylformamide, pH 4) was added, and plates were incubated overnight at 37 °C. Absorbance was measured at 570 nm using a Multiskan EX Microplate Reader (ThermoFisher Scientific, Waltham, MA, USA). Background readings from wells containing only medium were subtracted, and cell viability was expressed as a percentage relative to untreated control/vehicle cells.

Immunofluorescence

HOG cells expressing CB₁R and GPR55 were washed three times with PBS and fixed in 4% paraformaldehyde for 15 min. Following fixation, cells were rinsed three times and permeabilized with 0.05% Triton X-100 at room temperature for 15 min. To block nonspecific binding, cells were incubated with 1% bovine serum albumin for 30 min at room temperature. For CB₁R detection, cells were incubated overnight at 4 °C in a humidified chamber with a rabbit anti-CB₁R antibody (1:500; PA1-745, Thermo Scientific, Rockford, USA). After three PBS washes, cells were incubated for 30 min at room temperature with a biotinylated horse universal secondary antibody (1:40; Vector, PK-8800, Vector Laboratories Inc., Newark, NJ, USA), followed by streptavidin Alexa Fluor^® 550 conjugate (1:500; S2138, Invitrogen, Paisley, Scotland, UK). GPR55 was visualized via its YFP fusion, allowing direct detection through its intrinsic fluorescence. Nuclei were counterstained with TOPRO^®−3 (1:500; T3605, Invitrogen, Paisley, UK), and samples were mounted with Neomount Fluo (NB-23–00158-2, NeoBioTech, Nanterre, France). Negative controls were included to verify the absence of nonspecific labeling or signal amplification.

Imaging was carried out with a Leica TCS-SP8X spectral confocal microscope (Leica Microsystems, Mannheim, Germany) fitted with a 63X apochromatic oil-immersion objective (N.A. 1.4) and laser lines of 405 nm and 561 nm. For every field, stacks of images were collected across five Z planes at 1 μm steps in two separate channels (each corresponding to one fluorophore).

In SituProximity Ligation Assay

For PLA experiments, HOG cells that expressed CB₁R, GPR55, or both receptors (HOG-CB₁R, HOG-GPR55, HOG-CB₁R/GPR55) were first fixed with 4% paraformaldehyde for 15 min, rinsed in PBS containing 20 mM glycine to neutralize aldehyde residues, and then permeabilized in the same buffer with 0.05% Triton X-100 for 5 min. The PLA probes were generated by coupling a rabbit anti-CB₁R antibody (PA1-745, Thermo Scientific, Rockford, USA) to a PLUS oligonucleotide (Duolink^® In Situ Probemaker PLUS DUO92009, Sigma-Aldrich, St. Louis, MO, USA) and a rabbit anti-GPR55 antibody (10,224, Cayman Chemicals, Michigan, USA), which recognizes the human 207–2019 sequence, to a MINUS oligonucleotide (Duolink^® In Situ Probemaker MINUS DUO92010, Sigma-Aldrich, St. Louis, MO, USA), following the instructions provided by the manufacturer. After blocking with the solution from the PLA kit for 1 h at 37 °C, cells were incubated overnight at 4 °C with the antibody-linked probes at a final concentration of 75 µg/mL. Non-transfected HOG cells were used to confirm specificity.

Receptor clusters were visualized with the Duolink II in situ PLA detection kit (Duolink^® In Situ Detection Reagents Red DUO92008, Sigma-Aldrich, St. Louis, MO, USA). The samples were incubated with the ligation solution for 1 h at 37 °C, rinsed, and afterwards exposed to the amplification solution in a humidified chamber for 100 min at 37 °C. Nuclei were labeled using TOPRO^®−3 (1:500; T3605, Invitrogen, Paisley, UK), and coverslips were mounted with Neomount Fluo (NB-23–00158-2, NeoBioTech, Nanterre, France). To rule out nonspecific staining and amplification of signal, negative controls were performed. Negative controls included HOG cells transfected with CB₁R alone or GPR55 alone, as well as empty-vector (lipofectamine-only) treated cells lacking receptor expression. These conditions were processed in parallel and showed no detectable PLA signal, confirming assay specificity.

Images were acquired on a Leica TCS-SP8X spectral confocal microscope (Leica Microsystems, Mannheim, Germany) equipped with a 63X apochromatic oil-immersion objective (N.A. 1.4) and 405 nm and 561 nm laser lines. For each microscopic field, stacks of images were collected over five Z planes with 1 μm intervals in two separate channels (one for each fluorophore).

Data Analysis

After obtaining the raw data, outliers were identified using GraphPad Prism’s ROUT method and excluded before analysis. The outlier detection procedure was applied uniformly across all conditions. The resulting outlier-free datasets were then evaluated for normality and homogeneity of variances prior to statistical analysis. Statistical comparisons were performed using one-way or two-way analysis of variance (ANOVA), as appropriate for the experimental design, followed by Bonferroni’s post hoc test for multiple group comparisons. In directly comparative experiments, planned comparisons versus the corresponding control condition were applied. Differences were considered statistically significant at p < 0.05. All statistical analyses were performed using GraphPad Prism version 8 (San Diego, CA, USA).

Results

Heterologous Expression of CB₁R and GPR55 HOG Cells

Prior to these experiments, we confirmed by qPCR assays that HOG cells do not express endogenous CB₁R or GPR55 (Supplementary Fig. S1). The absence of amplification signal for both receptors rules out any potential interference from endogenous expression, thereby providing a suitable system to evaluate receptor function and potential heteromer formation upon transfection. Accordingly, HOG cells were transiently transfected with the plasmids encoding for CB₁R-Rluc and GPR55-YFP fusion proteins (see details in Materials and Methods). Immunofluorescence was performed to verify successful co-expression; CB₁R was detected using an anti-CB₁R antibody (red fluorescence, Fig. 1A), while GPR55 was visualized directly through the intrinsic fluorescence of its YFP tag (green fluorescence, Fig. 1B). Both receptors were primarily expressed at the plasma membrane, as shown by fluorescence, and were colocalized in several regions within the cell (yellow signal, Fig. 1C), suggesting close spatial proximity and potential heteromer formation. Notably, receptor expression was observed throughout the entire cell, including the cytoplasm and cellular extensions, indicating widespread distribution.

Fig. 1.

Confocal laser fluorescence microscopy images showing the expression of CB₁R-Rluc (A) and GPR55-YFP (B) in co-transfected HOG cells. CB₁R was detected by immunofluorescence using an anti-CB₁R antibody (red fluorescence), while GPR55 fused to YFP was visualized based on its intrinsic fluorescent properties (green fluorescence). Colocalization of both receptors is shown in yellow (C). Cell nuclei were stained with TOPRO^®−3 (blue). Scale bar: 10 µm. Detail 100x

Based on colocalization data suggesting a direct interaction, we used PLA to probe for CB₁R/GPR55 heteromerization. A distinct punctate red fluorescence, indicative of receptor–receptor interactions, was observed exclusively in cells co-transfected with both receptors (Fig. 2D). In contrast, non-transfected cells and cells expressing either CB₁R or GPR55 alone showed no detectable PLA signal, displaying only the TOPRO^®−3 nuclear counterstain (Fig. 2A, B, C).

Fig. 2.

Confocal laser fluorescence microscopy images showing the presence or absence of CB₁R/GPR55 heteromers (red fluorescent puncta) detected by in situ proximity ligation assay in HOG cells (A), HOG cells transfected with GPR55 (B), HOG cells transfected with CB₁R (C) and HOG cells co-transfected with both receptors (D). Nuclei were stained with TOPRO^®−3 (blue). Scale bar: 10 µm. Detail 100x

Neuroprotective Effect of CB₁R and GPR55 Expression Against Cuprizone-Induced Cytotoxicity in HOG cells

We next examined the potential neuroprotective role of CB₁R and GPR55 in a cellular model of CPZ-induced cytotoxicity. HOG cells were transiently transfected with CB₁R, GPR55, or both receptors (CB₁R/GPR55) and subsequently exposed to increasing concentrations of CPZ (50–1000 µM) for 24 h. Viability was assessed by the MTT reduction assay that measures mitochondrial activity. It is important to note that independent experiments were conducted separately for each transfection condition and were not performed as direct side-by-side comparisons.

In non-transfected cells, CPZ produced a significant reduction in viability, with losses of approximately 20% at concentrations above 200 µM (Fig. 3A). CPZ-induced toxicity in these cells did not follow a linear concentration–response relationship across the concentration range tested. A marked decrease in MTT reduction was observed already at 200 µM CPZ, with no significant additional decrease at 500–1000 µM. By contrast, cells expressing CB₁R, GPR55, or both receptors exhibited marked resistance to CPZ-induced toxicity (Fig. 3B–D). We next compared these conditions side-by-side within the same experimental runs (Fig. 4) to enable an unbiased assessment of relative efficacy. In this matched analysis, expression of CB₁R or GPR55 alone conferred resistance to CPZ-induced toxicity, and co-expression did not further increase the magnitude of this effect. Notably, this cytoprotective response was observed in the absence of exogenous agonists, consistent with the possibility that basal receptor activity contributes to cell survival in HOG cells. However, under serum-starved conditions CPZ toxicity was restored, suggesting that one or more serum-derived factor(s) may provide endogenous activation or otherwise support receptor-dependent signaling at CB₁R and/or GPR55.

Fig. 3.

Cuprizone effect on cell viability in HOG-CB₁R (B), HOG-GPR55 (C), and double-transfected HOG-CB₁R/GPR55 (D) cells treated with increasing concentrations of the toxic (50–1000 µM) for 24 h. Cell viability was assessed by the MTT reduction assay. Non-transfected HOG cells under the same conditions are shown as controls (A). Every condition was assayed in at least 5 different experimental sessions using 3 wells per condition per group. Each condition was normalized to its corresponding control/vehicle (100%) thus generating at least 15 raw data per condition. Data represent the mean ± SEM of normalized percentages after excluding outliers (see Methods). Vehicle treatment did not significantly affect viability; therefore, cuprizone-treated conditions were compared directly with the corresponding control group. Statistical significance was determined by one-way ANOVA followed by Bonferroni’s post hoc test with planned comparisons versus control (*p < 0.05, **p < 0.01, ***p < 0.001 versus control)

Fig. 4.

Comparative effect of cuprizone on cell viability in non-transfected HOG (blue bars), HOG-CB₁R (green bars), HOG-GPR55 (red bars), and double-transfected HOG-CB₁R/GPR55 (yellow bars) cells treated with increasing concentrations of the toxic (50–1000 µM) for 24 h. Cell viability was assessed by the MTT reduction assay. Every condition was assayed in at least 5 different experimental sessions using 3 wells per condition per group. Each condition was normalized to its corresponding control/vehicle (100%) thus generating at least 15 raw data per condition. Data represent the mean ± SEM of normalized percentages after excluding outliers (see Methods). Vehicle treatment did not significantly affect viability; therefore, cuprizone-treated conditions were compared directly with the corresponding control group. Statistical significance was determined by two-way ANOVA with Bonferroni's post-hoc test (^#p < 0.05; ^{# #}p < 0.01; ^{# # #}p < 0.001 versus non-transfected control cells)

To further explore the role of receptor activation, we next evaluated the neuroprotective potential of pharmacological stimulation of CB₁R and GPR55. HOG cells, either non-transfected or expressing CB₁R, GPR55, or both receptors, were pretreated with the selective CB₁R agonist arachidonyl-2'-chloroethylamide (ACEA, 100 nM) or the GPR55 agonist CID1792197 (1 µM), administered individually or in combination with the CB₁R antagonist SR141716 (250 nM). These assays were performed under serum-starved conditions. It is important to note that no selective antagonist for GPR55 is commercially available.

In cells expressing either CB₁R or GPR55 alone, agonist treatment did not elicit a consistent or receptor-specific enhancement of viability (Fig. 5A–B). Although modest statistically significant changes were observed under some conditions, these effects were small, variable, and did not reveal a robust agonist-driven survival profile in single-receptor expressing cells. In contrast, in cells co-expressing CB₁R and GPR55, ACEA treatment induced a modest but significant decrease in mitochondrial activity, which was fully reversed by the CB₁R antagonist SR141716, thereby implicating CB₁R in this response (Fig. 5C). Notably, pharmacological modulation by the GPR55 agonist CID1792197 was altered in the presence of CB₁R antagonism across receptor-expressing conditions, revealing a complex pattern of cross-regulation rather than a uniform blockade. This behavior is consistent with the atypical pharmacological properties described for CB₁R/GPR55 receptor complexes and reflects current limitations in selectively dissecting GPR55 signaling in the absence of a highly specific antagonist [79–82].

Fig. 5.

MTT reduction assay in HOG-CB₁R (A), HOG-GPR55 (B), and HOG-CB₁R/GPR55 (C) cells treated with the CB₁R agonist, ACEA (100 nM), or the GPR55 agonist CID1792197 (1 μM), in the presence or absence of the CB₁R antagonist SR141716 (250 nM). Every condition was assayed in at least 3 different experimental sessions using 2 wells per condition per group. Each condition was normalized to its corresponding control (100%) thus generating at least 6 raw data per condition. Data represent the mean ± SEM of normalized percentages after excluding outliers (see Methods). Statistical analysis was performed by a one-way ANOVA followed by Bonferroni’s post hoc test with planned comparisons versus the respective control group (*p < 0.05, **p < 0.01, ***p < 0.001 versus control; ^&&p < 0.01, ^&&&p < 0.001 versus the respective agonist treatment)

We next evaluated whether pharmacological activation of CB₁R or GPR55 could counteract CPZ-induced cytotoxicity. As shown in Fig. 6, treatment with the CB₁R agonist ACEA or the GPR55 agonist, CID1792197, significantly attenuated CPZ-induced loss of viability in HOG cells co-expressing both receptors (Fig. 6D). In cells transfected with CB₁R alone, ACEA fully restored cell viability to control levels, an effect that was abolished by co-treatment with the selective CB₁R antagonist, SR141716 (Fig. 6B). This protective effect was not observed in non-transfected cells (Fig. 6A). By contrast, in cells expressing GPR55 alone, cannabinoid ligands did not enhance survival beyond the modest effects associated with receptor expression itself (Fig. 6C).

Fig. 6.

MTT assay in HOG (A), HOG-CB₁R (B), HOG-GPR55 (C), and HOG-CB₁R/GPR55 (D) cells pretreated with the CB₁R agonist ACEA (100 nM) or the GPR55 agonist CID1792197 (1 µM), in the presence or absence of the CB₁R antagonist SR141716 (250 nM), 30 min prior to exposure to cuprizone (500 µM, 24 h). Every condition was assayed in at least 3 different experimental sessions using 2 wells per condition per group. Each condition was normalized to its corresponding control/vehicle (100%) thus generating at least 6 raw data per condition. Data represent the mean ± SEM of normalized percentages after excluding outliers (see Methods). Statistical analysis was performed by a one-way ANOVA followed by Bonferroni’s post hoc test with planned comparisons versus the respective control group (*p < 0.05, **p < 0.01, ***p < 0.001 versus control; ^#p < 0.05, ^{# # #}p < 0.001 versus cuprizone; ^&&&p < 0.001 versus the respective agonist treatment)

Together, these results highlight the atypical pharmacological profile associated with CB₁R/GPR55 receptor co-expression and support the presence of functional receptor interactions that modulate pharmacological responses to CPZ-induced toxicity.

Discussion

This study provides novel evidence that GPR55 and the cannabinoid CB₁ receptor exert significant neuroprotective effects against CPZ-induced cytotoxicity in human oligodendroglial cells, with comparable levels of protection observed upon individual or combined receptor expression, alongside evidence for a distinct pharmacological interaction when both receptors are co-expressed. These findings significantly extend our previous work on CB₁R/GPR55 heteromerization [34, 43, 64, 65] and underscore the therapeutic potential of targeting this receptor complex in demyelinating pathologies like MS.

Before addressing the receptor-dependent mechanisms underlying the observed neuroprotection, some methodological aspects of the CPZ-based in vitro model should be considered. The lack of a clear concentration-dependent decrease in viability in CPZ-treated HOG cells deserves specific consideration. In this study, cell viability was assessed using the MTT reduction assay, which primarily reflects mitochondrial metabolic activity rather than absolute cell number or membrane integrity. CPZ is known to induce early mitochondrial dysfunction in oligodendroglial cells, leading to a rapid decrease in metabolic activity that may plateau over a wide concentration range without necessarily reflecting incremental cell death. Similar non-linear concentration–response profiles have been previously reported in CPZ-based in vitro models and are thought to reflect a threshold effect on mitochondrial function rather than progressive cytotoxicity [75]. Therefore, the steady state viability observed at higher CPZ concentrations likely reflects saturation of mitochondrial impairment rather than resistance to CPZ toxicity.

Within this context, expression of CB₁R, GPR55, or both receptors conferred substantial resistance to CPZ-induced toxicity. Although we did not detect a statistically greater effect of co-expression under these conditions, the data robustly support protection with receptor expression, and future work will address whether specific stimulus contexts reveal synergistic effects. In this sense, one plausible explanation for the absence of an additive effect of co-expression in the baseline CPZ assay is that, under serum-containing conditions, receptor-dependent protective signaling may already be partially engaged by endogenous factor(s) present in fetal bovine serum (FBS). Indeed, protection was observed in the absence of exogenous agonists but was lost under serum-starved conditions, arguing against a purely constitutive mechanism and instead supporting the presence of a serum-dependent permissive “tone” (e.g., bioactive lipids/endocannabinoid-related mediators) capable of sustaining CB₁R and/or GPR55 signaling. This interpretation is particularly relevant for GPR55, whose pharmacology has been widely described as atypical, context-dependent (including functional selectivity), and still incompletely defined, with ongoing debate regarding ligand classification and receptor “identity” [83–85]. Under such conditions, basal activation and/or pathway saturation could compress the dynamic range of a viability readout and obscure incremental gains expected from co-expression, even if receptor–receptor interactions meaningfully alter signaling quality without substantially changing the overall magnitude of the response. Importantly, because all groups were assayed under identical media composition and normalized within each experiment, this interpretation not compromise the validity of the matched comparisons but instead provides context for the absence of additivity.

Rather than increasing the magnitude of the response, co-expression revealed evidence of receptor–receptor interaction consistent with heteromer formation and atypical pharmacological modulation. This aligns with our prior reports of unique pharmacological properties in these heteromers, including cross-antagonism and biased signaling [64, 65]. The present pharmacological data further cement this interpretation: the response to a GPR55-specific agonist was significantly attenuated by a CB₁R antagonist. This cross-receptor modulation is a recognized hallmark of GPCR heteromerization [79–82, 86] and highlights an atypical signaling profile that could be exploited for next-generation therapeutics in neurodegenerative and demyelinating diseases.

Interpretation of these pharmacological interactions must take into account current limitations in the available pharmacological tools for GPR55. The absence of a truly selective and well-characterized GPR55 antagonist is a recognized constraint in the field and complicates the dissection of GPR55-specific signaling using classical pharmacological approaches. In the present study, we therefore employed a strategy consistent with current best practice, using selective agonists in combination with CB₁R antagonism to assess functional modulation and cross-regulatory behavior. While genetic approaches such as GPR55 knockdown or knockout would provide complementary mechanistic insight, these strategies represent a distinct experimental expansion and are more appropriately addressed in future studies. Importantly, the pharmacological effects observed here are sufficient to support functional interaction and atypical signaling behavior consistent with receptor-complex formation rather than isolated receptor activity.

The translational implications for MS are considerable. Current standard-of-care therapies primarily target the immune component of the disease but offer limited efficacy in promoting remyelination or protecting neural cells from damage [9, 18, 87, 88]. Our results propose a paradigm shift by targeting oligodendroglial resilience directly. Therefore, developing ligands that selectively target the CB₁R/GPR55 heteromer to modulate its specific signaling output may represent a more precise and effective strategy to enhance OLG survival in the hostile inflammatory milieu of MS lesions.

Despite these promising insights, several considerations must be acknowledged. Firstly, the use of the human HOG oligodendroglial cell line represents a deliberate reductionist approach. Although HOG cells are derived from an oligodendroglioma and do not fully recapitulate the complexity of primary OLGs, their lack of endogenous CB₁R and GPR55 expression provides a unique experimental advantage. This feature allows receptor-specific effects to be examined in isolation, without interference from endogenous cannabinoid signaling, thereby enabling a clear functional assessment of CB₁R, GPR55, and their co-expression. While this simplified system cannot model the full cellular and molecular environment of the CNS, it is well suited to dissect receptor-dependent mechanisms under controlled conditions. CB₁R expression has been documented in cells of the OLG lineage [56–59], where it contributes to survival and differentiation, and GPR55 expression has been reported in glial populations, with upregulation under inflammatory conditions [60]. However, whether OLGs co-express both receptors in vivo remains unresolved. To date, the only available evidence for CB₁R/GPR55 complexes in the human brain comes from our recent study of post-mortem MS tissue, in which these heteromers were detected in the prefrontal cortex [66]. Accordingly, our reductionist model serves as a unique tool to uncover mechanistic principles that may underlie OLG resilience in demyelinating conditions, while highlighting the need for further studies in primary cultures, animal models, and patient-derived samples. The apparent protective effect observed in cells expressing CB₁R and GPR55, even in the absence of exogenous agonists, is unlikely to reflect constitutive receptor activity, as our data under serum-free conditions do not support this interpretation. Accordingly, future studies employing targeted lipidomics in serum/plasma (and in commonly used culture sera) will be valuable to determine whether endocannabinoids or structurally related bioactive lipids are present at concentrations compatible with CB₁R and/or GPR55 engagement, potentially contributing to a permissive basal signaling tone. However, such mechanistic dissection is beyond the scope of the present study. Serum is known to contain bioactive lipids, including endocannabinoids and structurally related molecules, which may contribute to basal receptor tone and partially activate CB₁R and/or GPR55 in the absence of exogenous agonists. Importantly, all experimental conditions were assessed under identical media composition and normalized within each experiment, ensuring that relative comparisons between conditions remain valid and are not confounded by serum-derived effects. This leads to a novel hypothesis: that the loss of these receptors in vivo could itself be detrimental to OLG resilience [89–91]. However, it is important to note that this immortalized cell line cannot fully capture the complexity of primary OLGs or the intricate cellular crosstalk with astrocytes, microglia, and neurons that defines the in vivo CNS environment. Secondly, the acute CPZ-induced cytotoxicity model, while useful, replicates aspects of demyelination but not the chronic immune-mediated pathogenesis characteristic of MS [74, 92]. Thirdly, the field is constrained by a limited pharmacological toolkit for GPR55, particularly the absence of a highly selective and potent antagonist, which hinders the precise dissection of individual receptor contributions within the complex [46]. Finally, while we demonstrate protection against acute toxicity, future work must determine if this translates to enhanced remyelination and long-term functional recovery in more complex systems.

By demonstrating receptor-dependent protection and receptor–receptor interaction in a human oligodendroglial model, this work provides a rationale for further exploration of heteromer-specific cannabinoid signaling in demyelinating conditions. Future research must now validate these results in vivo, using advanced animal models and ultimately human clinical samples, to translate this promising mechanism into a viable strategy for halting progression and promoting repair in MS. In this regard, preclinical studies such as that of Reynoso-Moreno et al., who demonstrated that the selective endocannabinoid reuptake inhibitor WOBE437 reduces disease progression in a mouse model of MS, further underscore the therapeutic promise of cannabinoid system modulation in demyelinating disorders [93]. Future studies using primary OLG cultures, organotypic systems, or in vivo demyelination models (e.g. Experimental Autoimmune Encephalomyelitis models) will be required to validate the translational relevance of these findings.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Conceptualization was agreed by EMP and RF, who also participated in the design of the project and analyzed the results; EMP, JMCM performed the majority of the experiments; CMP, SVC, RRS, RP participated in a significant number of experiments; EMP, AN, RP and RF participated in the software, data analysis and preparing the final figures; EMP and RF wrote the first draft of the manuscript which was further edited by all co-authors, who agreed with the final submission. All authors have read and agreed to the published version of the manuscript.

Funding

Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This research was funded by Fondo de Investigaciones Sanitarias (FIS), which belongs to the Spanish National Instituto de Salud Carlos III though the project (PI15/00601) (Co-funded by European Regional Development Fund and European Social Fund “Investing in your future”).

Data Availability

The datasets generated and/or analyzed during the current study are not publicly available due to privacy/ethical restrictions but are available from the corresponding author upon reasonable request.

Declarations

Ethics Approval

Not applicable.

Competing interests

The authors declare no competing interests.