Glioblastoma is the most common and aggressive form of primary brain tumor in adults, characterized by rapid progression, heterogeneity, and resistance to conventional therapies. Despite advances in treatment modalities, including maximal surgical resection, radiation therapy, and alkylating chemotherapy with temozolomide, the prognosis for glioblastoma patients remains dismal, with a median survival of only 12–15 months and a 5-year survival rate of less than 5% (Jaoude et al., 2019). Thus, novel treatment options are needed. One of the possible options for improving therapeutic outcomes in glioblastoma is to create synergistic pro-oxidant therapy, which could be effective in such concentrations that are not harmful to normal cells. Such an approach allows selective targeting of glioblastoma cells without destructive effects on normal astrocytes and neurons.

Oxidative and nitrosative stress, as well as inflammation, contribute to the mechanisms of gliomagenesis, tumor progression and drug resistance (Ostrowski and Pucko, 2022). The rapid growth of glioblastoma cells creates prolonged hypoxia which results in oxidative stress, hindering oxidative phosphorylation and increasing reactive oxygen species (ROS) production in the mitochondria. This leads to the reprogramming and development of glioblastoma stem cells (GSC). These cells in particular show signs of high oxidative stress, partly because of their high metabolic demand and hypoxic conditions (Schroeder et al., 2020). In response to variations in oxygen and ROS levels, nuclear respiratory factors and hypoxia-inducible factors are activated, all of which start tumor metabolic reprogramming (Krawczynski et al., 2020). It is also the Achilles’ heel of glioblastoma cells, because the pro-oxidative therapies, such as radio- and chemotherapy, further overload the mitochondria and intracellular environment with too many ROS, ultimately leading to apoptosis. In fact, the majority of anticancer treatments and targeted therapies, function by either directly or indirectly producing ROS (Farhan, 2024). Therefore, the pro-oxidant effects of new drug regimens are of high value. On the other hand, the safety of such new therapies to normal astrocytes and neurons is crucial, as those cells are prone to oxidative stress-induced damage, which can lead to neurocognitive side effects.

In this regard, plant-derived agents present an interesting option, as they possess antioxidant properties, but can also exhibit pro-oxidant behavior, depending on the context (Farhan, 2024). In contrast to synthetic chemotherapeutics, such as temozolomide, plant-derived compounds possess unique anticancer properties and very often high selectivity for tumor cells, without affecting normal cells (Farhan, 2024; Majchrzak-Celińska and Studzińska-Sroka, 2024). In this context, cannabidiol (CBD) have been found to exert interesting anticancer effects, which are accompanied by its neuroprotective and anti-inflammatory properties (Rybarczyk et al., 2023a). CBD is a multi-mechanism agent that modulates many signaling pathways in cancer cells and is considered a non-opioid analgesic. Recently, CBD is being evaluated in clinical trials as an anti-glioblastoma agent (Rybarczyk et al., 2023b; Feng et al., 2024). On-going reports provide evidence of very promising results which rises hopes that the implementation of CBD as a new anti-glioblastoma agent can be expected in the nearest future.

Moreover, it is important to consider neuroinflammation as a key factor contributing to gliomagenesis, when thinking of new anti-glioblastoma therapies. Neuroinflammation is promoted by hypoxic tumor microenvironment, increased oxidative stress, and an immune suppressive milieu (Alghamri et al., 2021). Glioblastoma cells express and secrete immune suppressive chemokines and cytokines, including interleukin (IL)-6, IL-10, transforming growth factor (TGF)-β, and galectin-1, which act on infiltrating immune cells to hijack them by inducing a protumor cellular phenotype (Roesler et al., 2021). Tumor cells also overexpress cyclooxygenase-2 (COX-2) and its major product within the brain, prostaglandin E2 (PGE2) (Qiu et al., 2017; Dean and Hooks, 2023). The latter acts on its E-prostanoid (EP) receptors to regulate cell proliferation, migration, apoptosis, and angiogenesis, contributing to tumor progression (Qiu et al., 2017). In our previous study we showed that COXIBs, in particular celecoxib and 2,5-dimethylcelecoxib (2,5-DMC), counteract the hyperactivated Wnt pathway and COX2/PGE2/EP4 pathway, providing therapeutic benefit both when used alone or in a combination with temozolomide (Majchrzak-Celińska et al., 2021). Moreover, our most recent study showed that combining tinostamustine with celecoxib more effectively induces apoptosis and halts cellular migration, as compared to tinostamustine alone (Pawlak and Majchrzak-Celińska, 2025). Together with mounting evidence for the benefits of COXIBs in glioblastoma treatment, there is also a growing need for further exploration of possible other drug combinations involving COXIBs, which might accelerate glioblastoma research.

Thus, this study aimed to examine whether CBD + celecoxib and CBD + 2,5-DMC have a higher anti-glioblastoma efficacy as compared to individual compounds. In this regard we evaluated their cytotoxicity against U-138 MG glioblastoma cell line and human astrocyte HA cell line. We also explored the involvement of ROS production and the antioxidant Nrf2 response with canonical and non-canonical pathways. Nrf2 nuclear translocation and binding, followed by the analysis of Nrf2-pathway downstream proteins were applied to verify the cellular response to applied treatments. Finally, the impact of the analyzed compounds and their combinations on cellular motility, autophagy and apoptosis was evaluated.

This study proposes a new, dual therapy, consisting of CBD with either celecoxib or 2,5-DMC, as an effective anti-glioblastoma treatment inducing oxidative stress-related cytotoxicity, apoptosis and diminishing cellular motility.

Combination therapy, a treatment modality that combines two or more therapeutic agents, is nowadays a cornerstone of cancer therapy (Mokhtari et al., 2017). It is a way to increase the anticancer effect due to synergistic or additive effects exerted by compounds. Recently, the synergistic inhibitory effect of celecoxib and 2,5-DMC plus paclitaxel or cisplatin on human cervix HeLa and SiHa cells was reported (Robledo-Cadena et al., 2024). Previously, we also found that the combination of CBD with ibuprofen or diclofenac brings therapeutic benefit concerning vulvar squamous cell carcinoma (Krajka-Kuźniak et al., 2022). Additionally, regarding the glioblastoma research, our group recently developed poly (lactic-co-glycolic acid)-based nanoparticles loaded with etoricoxib and CBD, which exhibited synergistic effect against two glioblastoma cell lines (Kuźmińska et al., 2023). In this study, we provide evidence that CBD and celecoxib, both at low molar concentrations (up to ∼25 µM), exert synergistic cytotoxic effect against glioblastoma cells. It is important to note, that both compounds cross the blood-brain barrier, so they can potentially exert their synergistic action within the tumor mass (Novakova et al., 2014; Zhu et al., 2018; Calapai et al., 2020). Similar syngery profile was found for the second tested drug combinations, i.e., CBD + 2,5-DMC, in the low molar concentrations (up to ∼25 µM).

In our study we found that the analyzed compounds exert their cytotoxic actions through the induction of oxidative stress. Although initially appearing counterintuitive, due to the association of oxidative stress with malignant transformation, it remains clear, that the induction of oxidative stress in glioblastoma cells brings therapeutic benefit (Ostrowski and Pucko, 2022; Cesca et al., 2024). Further increasing the ROS levels beyond the level that is acceptable by glioblastoma cells, leads to cancer cell death, as indicated in our apoptosis flow cytometry and Western blot caspase-3 results (Cesca et al., 2024). Indeed, all currently used cancer therapies, including radiotherapy and chemotherapy, exert their anticancer effects through generation of ROS that are derived from the ionization or excitation of the water constituent of the cells, resulting in the formation of aqueous free radicals and ROS, including superoxide radicals and H₂O₂ (Olivier et al., 2021). However, the neurocognitive side effects of those currently used therapies on normal cells are pronounced (Lawrie et al., 2019; Jiang et al., 2023). Indeed, the exposure of healthy brain tissue to radiation or chemotherapy adversely affects brain plasticity and repair processes causing neurocognitive impairment (Lawrie et al., 2019). Therefore, the aspect of potential neurotoxicity should be considered when thinking of new therapies for glioblastoma.

Our study provides evidence that the pro-oxidative effect produced by CBD, celecoxib and 2,5-DMC or their combination exerted on glioblastoma cells, is accompanied by the minimal effects on normal astrocytes. Within the synergistic, and potentially available at the tumor site, concentrations (5 and 10 µM), the single compounds did not create significant cytotoxicity in HA cell line. When used in combination and at a concentration of 10 μM, they, however, reduced the percentage of living cells to ∼90%, which can still be regarded as relatively safe. In a recent study, where the neuron-astrocyte sandwich coculture were used to investigate the neurotoxicity of CBD, it was found that 15 and 30 µM CBD caused viability decrease and morphological damage in the neuron-alone culture, whereas these toxic effects were significantly attenuated by the support of astrocytes in the neuron-astrocyte coculture (Kim et al., 2023). However, in a study by Jurič et al., where the impact of submicromolar (0.1, 0.5, 1, and 5 µM) concentrations of CBD on perinatal rat cortical neurons and astrocytes were evaluated, it was found that during this critical period of time, CBD might induce toxic effects (Jurič et al., 2022). In this regard, the authors reported that in astrocytes, 0.5 µM CBD caused an immediate short-term dysregulation of ΔΨm, followed by ATP depletion with transient activation of caspase-8 and progressive activation of caspase-9 and caspase-3/7, leading to early apoptosis and subsequent necroptosis (Jurič et al., 2022).

The central signaling pathway linking oxidative stress and neuroinflammation is the nuclear factor-2 erythroid related factor-2 (Nrf2) pathway. Nrf2 is a transcription factor that regulates cellular redox status through endogenous antioxidant systems with simultaneous anti-inflammatory activity (Staurengo-Ferrari et al., 2019). Our study provides evidence, that glioblastoma cells when treated with CBD, celecoxib and 2,5-DMC activate the Nrf2 pathway-dependent antioxidant response, however the response is only around 25% stronger as compared to control cells. In this regard, the most pronounced effects were observed as a response to combinatory treatment, with both CBD + celecoxib and CBD + 2,5-DMC, especially when 10 µM concentration was used. In this case we observed upregulation of mRNA level of Nrf2, nuclear translocation of Nrf2 and increased binding to the Nrf2 consensus sequence. Similar results were also observed as a response to 10 µM CBD treatment, however, the in this case we did not find lower levels of cytosolic Nrf2 protein level, which usually accompanies higher levels of Nrf2 in the nucleus. Nevertheless, the ELISA test confirmed higher binding affinity of Nrf2 to DNA in cells treated with CBD, as compared to DMSO-treated cells. Moreover, the nuclear translocation was also detected following the treatment with 10 µM celecoxib and 10 µM 2,5-DMC, however without significant increase in the binding affinity of Nrf2.

Glial tumor cells exhibit high levels of ROS, but also high level of antioxidant response signaling pathways, including Nrf2 pathway. Constitutive Nrf2 activation is a hallmark of cancer cells and promotes their migration through accumulation of the BACH1 transcriptional regulator (Occhiuto et al., 2023). Whether or not it is better to activate or to inhibit Nrf2 pathway during glioblastoma treatment remains elusive. Vast majority of studies provides evidence that inhibiting Nrf2 pathway is a better option, as it can sensitize glioblastoma cells to chemotherapy and radiation treatments (Rocha et al., 2020; Awuah et al., 2022). However, recent advances in the understanding of Nrf2 go beyond the classical understanding. Currently, studies reveal the potential clinical application of Nrf2 activators in glioblastoma treatment (Kim and Jeon, 2022). In this regard, Nrf2 activators were shown to inhibit proliferation and induce differentiation and apoptosis in various cancer cells, including glioblastoma (Tsai et al., 2021; Kim and Jeon, 2022). In this context, RTA 404, a new trifluoroethylamide derivative of 2-cyano-3,12-dioxoolean-1,9-dien-28-oic acid was found to inhibit proliferation, cell migration, cell cycle progression, and induce apoptosis in GBM840 and U87 MG cells in vitro, possibly through its inhibition of N-cadherin and E-cadherin expression via its inhibition of the AKT pathway (Tsai et al., 2021). Nrf2 activators, such as dimethyl fumarate, are tested in clinical trials, in the context of glioblastoma (Talebi et al., 2023). Moreover, recently, de Sousa et al. found that high levels of Nrf2 sensitize temozolomide-resistant glioblastoma cells to ferroptosis via ABCC1/MRP1 upregulation (de Souza et al., 2022). The complex nature of the Nrf2 pathway still awaits elucidation (Rybarczyk et al., 2023a).

Here we provide evidence, that oxidative stress generated by CBD + celecoxib, and CBD + 2,5-DMC, is accompanied by mild Nrf2 pathway activation. Similar results regarding celecoxib, however, in the context of cardiovascular disease, were published by Al-Rashed. In this study, the treatment of human endothelial cells with celecoxib led to COX-2 independent signaling via phosphorylation of AMPK, resulting in the nuclear translocation of Nrf2 and AMPK-CREB-Nrf2-dependent signaling activation (Al-Rashed et al., 2018). As far as CBD is concerned, our results align with a study of Singer et al. who found that CBD induces Nrf2 activation, which in turn induces antioxidant response genes expression (Singer et al., 2015). According to our best knowledge, the impact of 2,5-DMC on Nrf2 signaling in glioblastoma cells has not been evaluated by other research groups before.

Furthermore, we also show that in response to Nrf2 pathway activation following CBD, celecoxib and 2,5-DMC treatment, SOD1 is being upregulated, potentially accelerating the conversion of the superoxide anion to H₂O₂ and oxygen (Wang et al., 2018). However, we also show, that CAT, encoding H₂O₂ neutralizing enzyme, is not being upregulated, which might explain the cytotoxicity of these compounds, despite the Nrf2 mild activation.

The results of our study also reveal an increase in the protein level of p62, in response to CBD, celecoxib and dual therapy of glioblastoma cells with these two agents. Additionally, we observed similar upregulation of p62 following the treatment with CBD + 2,5-DMC. p62 is the key protein of the non-canonical Nrf2 pathway, coordinating the ubiquitin-proteasome system and autophagy (Staurengo-Ferrari et al., 2019; Kumar et al., 2022). Overall, p62’s domains provide a scaffold that directs substrates to autophagosomes and facilitates the autophagic process (Kumar et al., 2022). Therefore, when autophagy is induced, a corresponding decrease in p62 levels is observed. Conversely, p62 accumulation indicate autophagy inhibition (Puissant et al., 2012). Taking into consideration that autophagy can be regarded as one of glioblastoma therapy resistance mechanisms, the observed increase in the protein level of p62 as a response to the treatment with CBD, celecoxib and 2,5-DMC is a promising result (Khan et al., 2021).

Recent studies indicate that CBD influences both autophagy and apoptosis, which aligns with our finding (Fu et al., 2023). Our results are, however, in contradiction with the data published by Giannotti et al., who found that low-dose CBD makes U-87 MG cells more vulnerable to cytotoxic effects, reducing cell viability and mitochondrial dynamics while increasing autophagic flux and redox systems (Giannotti et al., 2025). The discrepancy between our and Giannotti’s results might arise from different glioblastoma cell line used as a model. In our study, instead of autophagy, we observed apoptosis induction, as presented by caspase-3 Western blot results. In this context, all analyzed compounds and combinations significantly upregulated caspase-3, indicating an ongoing apoptotic degradation of cells.

In addition to p62, we also analyzed LC3 protein, a widely recognized marker of autophagy. We observed accumulation of LC3 upon treatment with CBD, celecoxib, 2,5-DMC and its combinations, suggesting that autophagic processes are being triggered in U-138 MG cells. Interestingly, the increase in LC3 was accompanied by elevated p62 levels, which may indicate a blockade of autophagic flux rather than its full activation. Such a scenario has been reported in glioblastoma cells previously and is thought to contribute to apoptosis induction rather than to survival (Puissant et al., 2012; Khan et al., 2021). This dual regulation of LC3 and p62 points to a complex balance between autophagy and apoptosis following treatment with the investigated compounds.

We also evaluated the component of Akt/mTOR pathway, a critical negative regulator of autophagy. Treatment with CBD and its combinations reduced levels of mTOR protein expression, consistent with the accumulation of LC3 and suggesting suppression of the Akt/mTOR axis. Given that mTOR inhibition promotes autophagy initiation, our findings indicate that the compounds may trigger autophagic signaling, but incomplete flux - as evidenced by p62 accumulation - may redirect cells toward apoptosis instead of survival. This interplay between mTOR suppression, p62 accumulation, and LC3 induction highlights a possible mechanism by which CBD, celecoxib, and 2,5-DMC tilt the balance from cytoprotective autophagy toward apoptotic cell death in glioblastoma.

Eventually we would like to point that the combination treatment with CBD + celecoxib, but also CBD + 2,5-DMC influence cancer cell migration. In this regard, all 10 µM solutions of CBD, celecoxib and 2,5-DMC and their combinations significantly halted cell migratory potential. This might result from the anti-inflammatory properties of celecoxib, or the Nrf2 pathway modulation by all three compounds, since this pathway is also involved in migration and invasion of glioma cells (Pan et al., 2013).

In conclusion, the combination of CBD with celecoxib or 2,5-DMC represents a promising therapeutic strategy for glioblastoma. While CBD alone induces cytotoxicity, ROS production, and apoptosis, our synergy analysis demonstrates that combining CBD with celecoxib or 2,5-DMC allows effective killing of GBM cells at lower CBD concentrations. This dose reduction could potentially reduce neurotoxicity risks. Thus, by leveraging the synergistic effects of these compounds, it may be possible to overcome the challenges of therapy resistance, while providing lowest-possible side effects for glioblastoma patients. Future studies should focus on validating these findings in animal models and exploring their clinical applicability.