The therapeutic potential of repurposed mebendazole, alone and in synergistic combination with ONC201, in the treatment of diffuse midline glioma

Abstract

H3K27-altered diffuse midline glioma (DMG) is a universally fatal disease with no available therapeutic strategies apart from palliative radiotherapy. Repurposing marketed non-cancer drugs in oncology is emerging as a fast-tracking approach to speed up the development of new treatment options, urgently needed for DMG. Repurposed anthelmintic mebendazole (MBZ) is in the spotlight against brain tumors, because it joins promising anticancer properties with high neuropenetrance, favorable pharmacokinetic and safety profile. Although MBZ is undergoing Phase I/II trials against brain tumors, including DMG, MBZ anticancer properties and the underlying mechanisms of actions have poorly been characterized in DMG preclinical models. We found that MBZ robustly reduced cell viability in six out of seven DMG cell lines with either K27M-mutated or wild-type H3. All IC₅₀ values (range 102 to 958 nM) fell in a clinically attainable range. The antiproliferative MBZ properties were mediated by an arrest of DMG cells in the G₂/M phase with a concomitant upregulation of the key cell cycle regulators p21 and p27, whereas p53 upregulation and activation were cell context-dependent. At the same growth-inhibitory concentrations, MBZ triggered apoptotic cell death, as evidenced by higher levels of the apoptotic markers caspase-3 and PARP cleavage. Consistently, Annexin V-Propidium iodide (PI) double staining showed MBZ dose-dependent increase in both stages of apoptosis. Of interest, the combination of MBZ with the first-in-class imipridone ONC201 sinergistically increased the antiproliferative effects in two DMG cell lines as assessed by combination scores with different algorithms, showing additive effects in two others cell lines. Mechanistically, the combination potentiated the proapoptotic activity of either MBZ or ONC201, while not changing the cytokinetic perturbations induced by the single drugs. Finally, one pair of ONC201-sensitive and ONC201-resistant DMG cell lines with acquired resistance showed same responsiveness to MBZ with similar values of IC₅₀ and E_max. In conclusion, MBZ demonstrates high growth-inhibitory/proapoptotic activity, chemosensitization property to ONC201 and the ability to overcome ONC201 resistance in DMG cell cultures, proposing as a new low-toxicity therapeutic for DMG, with a potential to be used in second-line treatment and/or in combination protocols.

Introduction

Diffuse midline gliomas (DMG) represent the leading cause of cancer-related mortality in children. According to the latest World Health Organization (WHO) classification of Central Nervous System (CNS) tumors, this group of neoplasms are categorized as DMG H3 K27-altered based on global hypomethylation of lysine 27 of histone H3 (H3K27) as a driving force for spurious gene expression and uncontrolled tumor growth [1]. This epigenetic alteration results from two mutually exclusive mechanisms, that is H3 K27M-mutation occurring in the majority of DMGs and Enhancer of Zeste Homologs Inhibitory Protein (EZHIP) over-expression with wild-type H3 present in almost all the remaining tumors [1-3].

Due to location within vital areas of the midline structures of the brain and infiltrative nature, surgical resection of DMG tumors is extremely challenging, and focal radiotherapy (RT) is the only life- prolonging treatment. Such approach, however, shows a marginal influence on the course of disease as median overall survival (OS) is 9-11 months, and only 2% of children survive 5 years after diagnosis [4,5]. Over 250 clinical trials spanning several decades and testing a vast array of therapeutic interventions, including cytotoxic chemotherapy, chemoradiation, molecularly targeted therapy and immunotherapy, have not improved patient outcomes [3]. Thus, new chemotherapeutic options are urgently needed to defeat these universally fatal tumors [6,7].

In the recent years, drug repurposing - i.e. the use of already Food and Drug Administration (FDA) -approved drugs for indications other than their originally intended use - is emerging as a fast-tracking, cost-effective approach to expedite the developmental timelines of new therapeutic strategies [8,9]. In this scenario, mebendazole (MBZ) has garnered interest as one of the most promising repurposed drugs against brain cancers [10-12]. It has shown more than 40 years of safe use in humans as an anthelmintic drug and serendipitously been discovered having anticancer properties against animal brain tumor models. In addition to its well-known destabilizing effects on microtubule polymerization of fast-dividing cells [13], MBZ inhibits a number of oncogenic signaling pathways and cancer-related cellular processes, such as angiogenesis and epithelial-to-mesenchymal transition, while triggering apoptosis and cell cycle arrest. In cultured glioblastoma (GBM) and medulloblastoma (MB) cell lines dosed with MBZ, the IC₅₀s for cell viability vary from 0.1 to 0.5 μM [10,14], concentrations that are clinically attainable [15]. In vivo administration of MBZ prolongs survival of mice bearing intracranial GBM [10] and MB [16], with concomitant reduction in vascularity in tumor, but not in normal tissues. Of interest, in vitro high-throughput screening testing large mechanistically annotated collections of approved and investigational compounds have identified MBZ among the top potency-selected agents active against cell lines derived from childhood solid tumors [17], including DMG [18].

Therefore, due to its broad mechanisms of actions as anticancer agent and proven synergy with chemotherapy and RT in the preclinical setting, as well as its high blood brain barrier penetrability and favorable pharmacokinetic and safety profile in the clinical setting, MBZ meets many of the ideal features for a repurposed drug for the treatment of CNS tumors [12,19,20]. Indeed, Phase I/II clinical trials are currently undergoing in both adults (NCT01729260) and children (NCT02644291, NCT01837862) to investigate the anticancer effect of MBZ, alone or in combination with standard chemotherapy drugs, for the treatment of brain tumors including DMG. Moreover, computational multiomics-based approaches used for biomarker-guided drug selection have identified MBZ as one of the medicines of choice for personalized treatments of patients with DMG to block specific signaling [21] or in backbone therapies [22].

However, thorough mechanistic investigations of the anticancer properties of MBZ in DMG cell cultures has not been performed. A deeper knowledge of the cellular and molecular processes underlying the antitumor effects of MBZ and of its pharmacological interactions with other agents are essential for the inclusion of MBZ in more efficacious treatment protocols against DMG [11]. Emerging evidence shows that DMGs display large intertumoral and intratumoral heterogeneity, highlighting the need for combination therapies to overcome clonal diversity and the occurrence of resistance.

With these premises, we characterized the effects of MBZ across a panel of DMG H3K27M-mutated and H3 wild-type cell lines, by utilizing cell survival analyses, cell cycle profiling and markers of cell death. In addition, we performed combination experiments to investigate MBZ’s role in chemosensitizing DMG cells to ONC201, a promising investigational compound that is currently in advanced phase clinical trials against H3K27-altered gliomas showing preliminary survival benefits [3]. Finally, the ability of MBZ to overcome ONC201 resistance is highlighted.

Materials and methods

Cell culture and reagents

Primary DMG cell culture models (SU-DIPG-VI(E), SU-DIPG-XIII-FL, SU-DIPG-XIII-P*, SU-DIPG-XLVIII, SU-DIPG-XXXVIII) were kindly provided by Dr. Michelle Monje, Standford University. SF8628 and SF7761 were purchased from Merck Millipore (Burlington, MA, USA). SF8628 cells were maintained in Dulbecco’s modified Eagle medium (DMEM) High glucose (Euroclone) supplemented with 10% fetal bovine serum (FBS) and GlutaMAX-I (Thermo Fisher, Invitrogen brand, Carlsbad, CA, USA) and grown in adherent culture conditions. All the other cell lines were grown in tumor stem serum-free media, made up of Neurobasal-A medium and DMEM/F-12 1:1, N-(2-Hydroxyethyl)piperazine-N’-(2-ethanesulfonic acid) (HEPES) buffer, sodium pyruvate, minimum essential medium (MEM) nonessential amino acids, antibiotic-antimycotic solution, GlutaMAX-I, B-27 supplement minus vitamin A (all purchased from ThermoFisher), supplemented with human fibroblast growth factor 2 (FGF2), human epidermal growth factor (EGF), human platelet-derived growth factor A (PDGFA), human platelet-derived growth factor B (PDGFB) (from Shenandoah Biotechology Inc., Warwick, PA, USA), and 0.2% Heparin (STEMCELL Technologies, Vancouver, BC, Canada).

The ONC201-resistant variant, referred to as SU-DIPG-XIII-R, was established by continuous exposure of the parental line to stepwise increasing ONC201 concentrations and maintained in a medium always containing 0.7 µM of the selecting agent.

MBZ (HY-17595, CAS No 31431-39-7, Polymorph C) and ONC201 (CAS No 1616632-77-9) were purchased from MedChemExpress (LLC, Monmouth Junction, NJ, USA), and dissolved in dimethyl sulfoxide (DMSO) to a stock concentration of 10 mM. Serial dilutions were made in cell culture media prior to cell treatments.

Cell viability assay

All suspension cell lines were plated in 96-well plates at a density of 5000 cells/well, whereas SF8628 were plated at a density of 1000 cells/well. After 24 hours, the cells were treated with vehicle DMSO or serial dilutions of each drug in triplicate or quadruplicate for 72 hours. Cell viability was determined by CellTiter-Glo (CTG, Promega, Madison, WI, USA) and bioluminescence imaging was measured using the TECAN imaging system (Tecan Spark, Tecan Austria GmbH Grödig, Austria). The half maximal inhibitory concentration (IC₅₀), the 95% confidence interval (95% CI) and maximal effect (E_max, defined as the cell proliferation inhibition percentage at the highest dose) were calculated using the sigmoidal dose-response function (variable slope, four parameters) of data with the GraphPad Prism version 8.00 for Windows (GraphPad Software Inc., San Diego, CA, USA).

Cell numbers of live and dead cells were assessed by an automated cell counter (NucleoCounter 100TM, ChemoMetec, Allerød, Hovedstaden, Denmark), that discriminates between live and dead cells through staining of nuclei with propidium iodide.

Drug combination and synergy assessment

For combination drug testing, cells were treated with each drug as a single-agent or in combination for 72 hours and cell viability was evaluated using the CellTiter-Glo assay. The pharmacological interaction between MBZ and ONC201 was calculated using the Compusyn software (ComboSyn, Inc.) and expressed as Combination Index (CI) based on the Chou - Talalay median-effect equation method, where a CI value < 1 is synergistic and a CI > 1 indicates antagonism [23].

To avoid biased synergy estimation, we also used the SynergyFinder 3.0 free web-application [24], which enables to quantify the combinatorial effects with multiple synergy reference models, each of them formulated under different assumptions of single-drug behavior, including Bliss excess, Loewe additivity, highest single-agent (HSA) and zero interaction potency (ZIP) [24,25]. The conventional metrics are: antagonism < -10; additivity, from -10 to 10; synergism > 10.

Flow cytometry

Cells were seeded at the appropriate concentrations to maintain them in a proliferative phase throughout the experiment. After 24 hours, cells were challenged with either vehicle or drug(s) at designated concentrations for 72 hours. At the end of the treatment period, the cells were collected, dissociated and then filtered using Cell Strainer (EASYstrainer 70 µm, Greiner Bio-One, Kremsmünster, Austria). Thereafter, cells were resuspended and fixed adding ice-cold 80% Ethanol dropwise while vortexing and stored at 4°C. At the moment of analysis, fixed cells were stained using a cell cycle kit according to the manufacturer’s instructions (Beckman coulter). After 15 minutes incubation at room temperature and protected from direct light exposure, fluorescence of treated and control samples was measured with flow cytometry (CytoFLEX S, Beckman Coulter, Milan, Italy).

Western blotting

SU-DIPG-XLVIII and SF8628 were seeded in the appropriate vessels and challenged 24 hours later with drug(s) at the designated concentrations for different time intervals. Vehicle-treated cells were used as controls. At the end of the treatment period, the cells were harvested, washed with ice-cold phosphate buffered saline (PBS) and disrupted in Pierce RIPA buffer with Halt Protease inhibitor cocktail (Thermo Fisher Scientific). Cell lysates (15 μg/lane) were resolved by precast SDS-PAGE gel (Bio-Rad Laboratories, Inc., Hercules, CA, USA) and transferred to a nitrocellulose membrane. The membranes were probed with the following primary antibodies overnight at 4°C: phospho-p53 (S15) (9284S, 1:1000), p21 Wafl/Clp1 (12D1) (2947S, 1:1000), cleaved Caspase 3 (Asp175) (5A1E) (9664, 1:1000), GAPDH (5174, 1:1000), all from Cell Signaling Technology (Danvers, MA, USA); p53 (DO-1) (sc-126, 1:1000), p27 Klp-1 (F-8) (sc-1641, 1:500) and cleaved PARP (Sc-56196, 1:1000) were from Santa Cruz Biotechnology, Dallas, TX, USA. All blots were incubated with the corresponding secondary horseradish peroxidase conjugated anti-mouse IgG (PI-2000, 1:10000) or anti-rabbit IgG (PI-1000, 1:10000) from Vector Laboratories (Newark, CA, USA). Immunoreactive bands were developed by enhanced chemiluminescence (ECL prime, GE Healthcare, Amersham, United Kingdom). Densitometry analysis of the bands was performed using the Image Lab software version 6.0.1 (Bio-Rad Laboratories Inc., Hercules, CA, USA).

Apoptosis

Cell apoptosis analysis was performed by treating appropriate numbers of SU-DIPG-XLVIII (1 × 10⁶) and SF8628 (0.25 × 10⁶) cells with different drug(s) concentrations for 72 hours. At the end of the incubation period, the samples were collected, dissociated in single cells and stained with the Dead Cell Apoptosis Staining kit (Thermo Fisher Scientific) containing Fluorescein isothiocyanate (FITC) - Annexin V conjugates and propidium iodide (PI), according to the manufacturer’s instructions. After incubation in the dark for 15 minutes at room temperature, the fluorescence of the treated samples was measured with flow cytometry (CytoFLEX S, Beckman Coulter, Milan, Italy).

Results

MBZ restrains the proliferation of DMG cells

We evaluated the antiproliferative effects of MBZ in 7 cell cultures of DMG representing neurosphere (SU-DIPG-XIII-FL, SU-DIPG-VI, SU-DIPG-XLVIII, SU-DIPG-XIII-P*, SF7761) and adherent models (SF8628 and SU-DIPG-XXXVIII). SU-DIPG-XIII-P*and SU-DIPG-XIII-FL (the latter indicated thereafter as SU-DIPG-XIII) have been derived from the primary pontine tumor and the secondary frontal lobe lesion from the same patient, respectively [26,27]. The DMG panel encompasses the complete genetic background of the tumor, including lines with H3.1K27M (SU-DIPG-XXXVIII), wild-type H3 (SU-DIPG-XLVIII), and the hallmark mutation H3.3K27M (all the remaining lines). Cytotoxicity was assessed after 72 hours of exposure to increasing MBZ concentrations (from 50 to 5000 nM) by the CTG assay. Dose-dependent response to MBZ was observed in 6 out of 7 DMG lines, with IC₅₀ values spanning from 102 nM to 958 nM. The median IC₅₀ was 290 nM, whereas the median value of E_max (%), indicating the maximum effect corresponding to the minimum measured viability [28], was 71% (range: 58 to 82%) (Figure 1A, 1B). SU-DIPG-VI cells were the less responsive to MBZ, yelding the lowest E_max (35%). Importantly, the IC₅₀ values that we found in vitro are attainable in a clinical range [15]. For example, in a recent Phase 1 dose-escalation study in patients with high-grade glioma, plasma concentrations of MBZ reached 1 μM with long-term safety and acceptable toxicity [29].

Figure 1.

MBZ reduces cell viability of DMG cell lines. (A) Dose-response curves and IC₅₀ graphic representation (dotted line) of MBZ effects on the indicated DMG cell lines. Cell viability was measured by CTG assay after 72 hours drug exposure. Results represent cell viability percentage respect to vehicle-treated control cells for three independent experiments performed in triplicate (mean ± SE). (B) IC₅₀ values with 95% confidence intervals and E_max (%). Cell viability and cell death (%) in the SU-DIPG-XLVIII (C) and SF8628 (D) cell lines at equitoxic doses of MBZ as determined by cell counting. Cells were treated with the corresponding IC₅₀ (1 µM and 0.25 µM, respectively) for 72 hours, when viable and non-viable cells were counted by NucleoCounter. Each bar represents the mean ± SD from three independent experiments (*P < 0.05; **P < 0.01; mean ± SD; two-tailed Student’s t test). n.r.: not reached.

To further characterize the pharmacological efficacies of MBZ, we determined cell viability and cell death by cell counting in SU-DIPG-XLVIII (Figure 1C) and SF8628 (Figure 1D), that exhibited different sensitivities to MBZ in terms of IC₅₀ (1 µM and 0.25 µM, respectively), but high and comparable E_max values (82 and 80%, respectively). Recently, growing evidence has shown that DMG with wild-type and K27M-mutated H3 have different epigenomic and transcriptional landscapes, and possible differences in therapeutic strategies [30], advocating focused preclinical characterization of their chemosensitivity profile. MBZ concentrations equal to the respective IC₅₀ values determined a reduction in viable cell number by about 50% in both lines as expected, that was accompanied by a two-fold increase in the number of non-viable cells. All the subsequent experiments were performed at these equitoxic concentrations of MBZ, unless otherwise specified.

MBZ induces perturbations of cell cycle in DMG cells by targeting cell checkpoint proteins p21 and p27

To determine the contribution of cytostasis vs apoptosis in the reduction in cell number induced by MBZ, we evaluated the perturbations of cell cycle in SU-DIPG-XLVIII and SF8628 cultures after exposure to the corresponding IC₅₀ for 24, 48 and 72 hours. Overall, the alterations of cytokinetic profile induced by MBZ were similar in the two lines, although of different entity: in SU-DIPG-XLVIII, an almost two-fold accumulation of cells in the S and G2/M phases was observed as soon as 24 hours of drug exposure (Figure 2A), that further increased at 48 hours and persisted up to 72 hours, with a concomitant marked reduction in the G0/G1 population at the corresponding time points. In the SF8628 line, changes in the cell cycle distribution were globally modest (Figure 2B), being an increase in the G2/M compartment at 24 hours up to 72 hours the most overt effect.

Figure 2.

MBZ alters the cell cycle progression and the expression of regulatory proteins p53, p27 and p21. (A, B) Time-course analysis of the effects of MBZ on cell cycle distribution in SU-DIPG-XLVIII (A) and SF8628 (B) cell lines. Cells were treated with MBZ at the respective IC₅₀ (1 µM and 0.25 µM, respectively) for the indicated periods of time. The percentages of the total cell population in the different phases of the cell cycle were assessed by flow cytometry. (C, D) Western blot analysis of total lysates from SU-DIPG-XLVIII (C) and SF8628 (D) treated with vehicle or equitoxic doses (IC₅₀) of MBZ for 24, 48 and 72 hours. Blots were probed with the indicated antibodies. GAPDH was used as a loading control.

Parallel Western blotting experiments carried out upon identical experimental conditions showed that MBZ upregulated the expression of the tumor suppressor proteins p53 and p21 in a time-dependent manner in SU-DIPG-XLVIII and this effect was already evident after 24 hour - exposure, whereas a frank increase in p27 protein level was observed at 48 hours to peak thereafter. Phosphorylation of p53 at Ser 15 was consistent with p53 modulation, suggesting that MBZ promotes p53 transcription factor activity, likely leading to the upregulation of the checkpoint proteins (i.e. p21) and cell cycle arrest (Figure 2C) [31]. In SF8628, carrying a mutation in the TP53 gene [32], p27 and mostly p21 showed a mild, time-dependent upregulation (Figure 2D), whereas there was no modulation of p53 nor of its activated form, in agreement with the little cell cycle changes observed. Thus, MBZ appears to block DMG cells at G2/M phase, regardless of their p53 status, although the effects on cell cycle are more evident in the presence of wild-type p53.

MBZ triggers apoptotic cell death in DMG cells

Next, we analyzed MBZ-induced caspase 3 activation and poly (ADP-ribose) polymerase (PARP) cleavage, two hallmarks of the apoptotic machinery. Both markers increased time-dependently in the two MBZ-treated lines as soon as 24 hours (Figure 3A, 3B), indicating an early onset of the apoptotic process. To further investigate the mechanisms of cell death set in motion by MBZ, we used Annexin V-PI double staining to distinguish early from late apoptotic and necrotic cells after drug challenging of SU-DIPG-XLVIII (Figure 3C) and SF8628 (Figure 3D) at their IC₂₅s (500 nM and 100 nM, respectively) and IC₅₀s for 72 hours. MBZ potently induced both stages of apoptosis dose-dependently in the two lines, which showed different sensitivity: at the highest equitoxic dose, apoptotic cells were approximately 10-fold and 4-fold more in MBZ-treated SU-DIPG-XLVIII and SF8628 cultures, respectively, compared to the corresponding control cells.

Figure 3.

MBZ triggers apoptotic cell death in DMG cells. (A, B) Time course of the effects exerted by MBZ on the cleavage of PARP and the activation of caspase 3 in SU-DIPG-XLVIII (A) and SF8628 (B). Cells were exposed to equitoxic concentrations (IC₅₀) of MBZ for 24, 48 and 72 hours. Cell lysates were subjected to immunoblot analysis with the indicated antibodies. (C, D) Flow cytometry analysis of MBZ-treated SU-DIPG-XLVIII (C) cells and SF8628 cells (D) stained with annexin V-FITC and PI. Cells were incubated with vehicle or equitoxic concentrations (IC₂₅ and IC₅₀) of MBZ for 72 hours.

Combinations of MBZ and ONC201 show synergistic activity against DMG

MBZ has been shown to act synergistically with drugs with different mechanisms of actions, including alkylating and molecularly-target agents [11,19]. So far, no study has thoroughly addressed the combinatorial effects of MBZ with ONC201, a promising investigational compound that is currently in advanced phase clinical trials against DMG and other cancers [33]. Initially designed as a bitopic antagonist of dopamine receptor D2/3, ONC201 has subsequently been found to activate the mitochondrial caseinolytic protease P (ClpP) [3,34] and to exert antineoplastic activity in different cancer models [35].

Because of the distinct mechanisms of action of MBZ and ONC201, we hypothesized that their combination could likely be synergistic. To examine the interaction between these two compounds, we firstly generated drug-response curves for ONC201 alone and in combination with a sub IC₅₀ MBZ across four representative patient-derived DMG cell cultures namely SU-DIPG-XLVIII, SF8628, SU-DIPG-XIII-P* and SU-DIPG-XIII. All combinations were tested for 72 hours before assessment of cell viability. Encouragingly, ONC201 demonstrated a shift to the left of the dose-response curve in the four lines when combined with a low-dose of MBZ (Figure 4). Together these data suggest increased efficacy of ONC201 when combined with MBZ.

Figure 4.

Low doses of MBZ potentiates the antiproliferative effects of ONC201. Dose-response curves of cell viability of DMG cell cultures after 72-hour exposure to ONC201 alone (green line) or with sub-IC₅₀ of MBZ (red line). Cell viability was measured by CTG assay and expressed as percentage respect to vehicle-treated control cells. MBZ concentrations were as follows: 500 nM for SU-DIPG-XLVIII, 100 nM for all the other lines.

Next, as a follow-up confirmatory screen, we performed a checkerboard assay, in which various concentrations of each drug were tested in duplicate both alone and in combinations for 72 hours. Each plate contained 6 × 5-dose matrix dilutions of MBZ and ONC201. Overall, cell viability in each culture was reduced when the two compounds were used in combination rather than as monotherapy (Figure 5A). Combination analyses with the CalcuSyn 2.0 software (Biosoft) demonstrated consistent synergy (CI < 1) between MBZ and ONC201 across doses of drug combination and cell lines (Figure 5B).

Figure 5.

Combinations of MBZ and ONC201 show synergistic activity against DMG cells. (A) Cell viability compared to DMSO control after 72-hour exposure of four patient-derived DMG cell cultures to increasing doses of MBZ (grey), ONC201 (green), or both (red). (B) Synergy CI scores (CompuSyn 2.0) across doses for MBZ/ONC201 combinations from the individual viability measurements in (A). Horizontal dotted line indicates a CI = 1, where points below the line indicate synergy and points above the line indicate antagonism.

We further validated the combination efficacy of MBZ with ONC201 by taking advantage from the Web application SynergyFinder, that allows to calculate and compare synergy scoring with four different reference models: HSA, Loewe, Bliss and ZIP [25,36]. MBZ/ONC201 combination was synergistic in SU-DIPG-XLVIII (Figure 6A) and SF8628 (Figure 6B) when assessed by all the algorithms (score > 10). Instead, in SU-DIPG-XIII-P* (Figure 6C) and SU-DIPG-XIII (Figure 6D) additivity was consistently found across all the models (score -10 to 10), with the exception of a Loewe score of 18.56 in SU-DIPG-XIII-P*, and a HSA score of 13.65 in SU-DIPG-XIII showing strong synergistic interaction. In the 2D synergy maps, algorithms identified the dose regions with the highest synergy score (highlighted by a white frame), that, in our case, corresponds to doses around the respective IC₅₀ value of each drug in the different lines. For this reason, these IC₅₀ values were selected for subsequent molecular studies.

Figure 6.

Comparison of the pharmacological interactions between MBZ and ONC201 calculated with different reference models confirms synergistic/additive effects of the combination in DMG cells. The combinatorial inhibitory effect of the MBZ and ONC201 combinations were analyzed by Synergy Finder from the individual viability measurements (shown in Figure 5A) in SU-DIPG-XLVIII (A) SF8628 (B), SU-DIPG-XIII-P* (C) and SU-DIPG-XIII (D). Each panel shows the 2D and 3D synergy maps as well as the table with synergy scores calculated with HSA, Bliss, Loewe and ZIP algorithms. The 2D and 3D synergy maps highlight synergistic and antagonistic dose regions in red and green colors, respectively. The white frame in each 2D plot represent the most synergistic area. The white dot in (A) indicates the drug combination used for the subsequent mechanistic studies.

Cotreatment with ONC201 potentiates the pro-apoptotic effects of MBZ in DMG cell lines

We next wondered whether the synergism between MBZ and ONC201 was mediated by an increase in cytostasis, in apoptosis or in both mechanisms. To this end, SU-DIPG-XLVIII cells were treated for 72 hours at doses of MBZ and ONC201 that resulted in the highest synergy score. FACS analysis revealed that the percentage of cells in the G2/M phase was even lower after co-treatment with MBZ-ONC201 than after treatment with the most effective single-drug agent, that is MBZ alone (Figure 7A). Concordantly, MBZ and MBZ+ONC201 augmented the expression of p53 and p27 proteins to a similar extent, being the MBZ-induced p27 and p21 upregulation even reversed by the combination (Figure 7B). Importantly, MBZ+ONC201 induced the cleavage of PARP and of caspase 3 much more than either agent alone (Figure 7C). Similar data were obtained through the Annexin V-PI double staining, that showed that cotreatment markedly increased the percentage of apoptotic cells compared to monotherapy, while minimally affecting the proportion of necrotic cells (Figure 7D).

Figure 7.

Cotreatment with ONC201 potentiates the pro-apoptotic effects of MBZ in DMG cell lines, while minimally affecting cell cycle distribution and checkpoint proteins. (A) Effects of both MBZ and ONC201, alone and in combination, on cell cycle distribution in SU-DIPG-XLVIII. Cells were treated with vehicle, MBZ (1 µM), ONC201 (1.5 µM) and both agents for 72 hours. The percentages of the total cell population in the different phases of the cell cycle were assessed by flow cytometry. (B, C) Western blot analysis of total lysates from SU-DIPG-XLVIII under the same conditions as in (A). Blots were probed with antibodies against regulatory proteins p53, p27 and p21 (B) and apoptosis markers (C). GAPDH was used as a loading control. (D) Flow cytometry analysis of SU-DIPG-XLVIII cells treated as in the above (A), stained with annexin V-FITC and PI.

MBZ overcomes ONC201 resistance in DMG cells

ONC201 has been proved to be beneficial in early phase trials (NCT03416530 and NCT03134131) in pediatric patients with H3 K27M-mutant DMG [37]. However around 20-30% of patients with nonrecurrent DMG show disease progression, suggesting that intrinsic or acquired resistance to ONC201 often occurs in clinical use.

To ascertain whether MBZ can be used in second-line therapy when upfront treatment with ONC201 fails, we evaluated the growth-inhibitory effects of ONC201 and MBZ in parental and ONC201-selected (SU-DIPG-XIII-R) cell lines. In this line, ONC201 resistance was associated with lower levels of the ONC201 target ClpP (Figure 8A, inset). ONC201 treatment for 72 hours demonstrated an IC₅₀ of 1.497 µM in SU-DIPG-XIII cells, in line with the literature data [38]. By contrast, IC₅₀ was not reached in SU-DIPG-XIII-R treated with the same doses of ONC201, and E_max value was 32% as opposed to 72% in parental cells (Figure 8A). Of interest, sensitivity to MBZ was unaltered in the two lines as evidenced by similar values of IC₅₀ and E_max, suggesting that MBZ overcomes acquired resistance to ONC201 (Figure 8B).

Figure 8.

MBZ overcomes ONC201-resistance in SU-DIPG-XIII-R. (A, B) Graphic representations and tables of IC₅₀, 95% CI, and E_max of the antiproliferative effects of ONC201 (A) and MBZ (B) on paired ONC201-sensitive and -resistant SU-DIPG-XIII cell lines. Cell viability was measured by CTG assay after 72-hour drug exposure and expressed as percentage respect to vehicle-treated control cells. n.r.: not reached. Inset (A), western blot analysis of the expression of ClpP and of housekeeping GAPDH proteins in SU-DIPG-XIII (WT) and SU-DIPG-XIII-R (R).

Discussion

Non-cancer medicines repurposed in oncology are emerging as an alternative approach to improve treatment protocols by reducing toxicity, enhancing efficacy and antagonizing drug resistance [39].

In search of novel and more effective therapeutic strategies for children with DMG, we investigated the multifaceted therapeutic potential of repurposed anthelmintic MBZ in DMG cell lines. We show that MBZ restrains cell proliferation in DMG cell lines by both reducing cell cycle progression and triggering apoptotic cell death. Previously, high-throughput drug screening of 2706 investigational and marketed compounds targeting 860 distinct cellular mechanisms uncovered that MBZ was one of the 371 most potent agents against 4 DMG cell lines (JHH-DIPG-1, SU-DIPG-XIII, SU-DIPG-XVII, SU-DIPG-XXV) [18]. However, no other cell-based studies beyond cytotoxicity assays were performed. We extended this panel of cultures by including six DMG cell lines so far uncharacterized for MBZ sensitivity, together with the previously reported SU-DIPG-XIII, used as a reference line. The IC₅₀ values that we found fell in the range of the published values, and most importantly, of the concentrations attainable with the clinic dosages, thus highlighting the efficacy of MBZ at clinically relevant concentrations [40] against a vast array of DMG lines with different genetic backgrounds, including an H3 wild-type DMG model.

The antiproliferative MBZ properties were mediated by an arrest of DMG cells in the G2/M phase with a concomitant upregulation of the key cell cycle regulators p21 and p27, which occurred regardless of the TP53-status of the cell lines, although they were more evident in those carrying wild-type TP53. In addition to the cytostatic effect, MBZ was able to reduce DMG cell survival by triggering apoptosis-related pathway. Our findings are in line with data from a wide variety of tumor models [12], in which mitotic arrest followed by caspase 3-mediated apoptotic cell death are among the main antitumor mechanisms induced by MBZ.

Cell cycle arrest at G2/M phase has been documented in acute lymphoblastic leukemia [11,41], GBM [41], triple-negative breast cancer [42] and lung cancer [43,44], whereas induction of the S-phase has inconsistently been reported [42]. In GBM cells, MBZ dose-dependently arrested the cell cycle at G2/M phase through the p53/p21/cyclin B1 pathway [41]. However, this study explored only a short treatment up to 24 hours. Importantly, our data, in addition to confirming published findings in other tumor cell lines, including brain tumor cell lines, document a never-before-reported, long lasting cytokinetic effect of MBZ, up to 72 hours.

Although both cytostatic and cytotoxic effects of MBZ have been placed in the context of functional p53 [19], p53-mutated cells engage apoptosis as well. For example, both wild-type and mutant p53 melanoma cell lines were shown to be sensitive to MBZ [45]. By the same token, MBZ treatment caused post-translational p53 stabilization and the downstream expression of p21 and MDM2 in lung cancer cell lines [40,43]. However, p53-null lung cancer cells exposed to MBZ also apoptose through intrinsic and extrinsic pathways, as indicated by activation of caspase-9 and caspase-8. Similarly, MBZ growth-arrested ovarian cancer cell cultures regardless of their p53-status [46]. In our models, MBZ treatment was able to induce cell-cycle arrest and apoptosis regardless of TP53 status and protein modulation, suggesting that the antiproliferative/proapoptotic response to MBZ does not uniquely rely on p53 in DMG. These findings are of importance for DMG, where a relevant portion of tumors express mutant TP53 [4,7,47].

Because of its pleiotropic nature, it is likely that MBZ activates programmed cell death by inhibiting crucial prosurvival signaling, such as PI3K/AKT and RAS/RAF/MEK/ERK, that play an important role in phosphorylating various apoptosis-regulating factors [11,48-50]. In melanoma models, MBZ decreases the phosphorylation of MEK and ERK as well as that of the ERK downstream targets related to stress response and translation, including ElK1 and RSKs [51]. Dose-dependent hypophosphorylation of ERK1/2 and AKT by MBZ has also been documented in acute myeloid leukemia cells (AML), which suggests that the growth-inhibitory activity of MBZ on AML cells may be mediated in part via inhibition of these signaling [52]. In a more recent publication, MBZ-induced ROS accumulation in non small cell lung cancer cells inhibits STAT3 signaling and downregulates BCL-xl antiapoptotic protein, with concomitant upregulation of apoptotic markers [53]. MBZ’ effects on the expression levels of antiapoptotic proteins, including XIAP and BCL2, have also been found in other models [54].

As for cell cycle progression, p21 has reportedly been documented as a key effector of the growth-inhibitory properties of MBZ in numerous cancer cell lines. However, other cell cycle progression proteins, including MYC [55,56], MYB [52], MAF [57], and cyclin D1 [49], to cite some, are downregulated after MBZ challenging. Importantly, it has been shown that H3K27M drives cancer by epigenetic transcriptome remodelling, that ultimately leads to the activation of a RAS/MYC axis [58]. Therefore, MBZ targeting of these pathways may be a potential therapeutic option for H3K27M-driven cancers.

Until now monotherapies have unequivocally been unsuccessful in patients with a hard-to-treat disease such as DMG [33], except for ONC201 that has shown emerging clinical efficacy in early phase trials [34,37], where a significant increase in median OS has been reported from historical 11.9 months to 21.7 months in ONC201-treated patients. However, children eventually succumb with some even failing upfront treatment, indicating that intrinsic and/or acquired resistance to ONC201 occurs in DMG cells [59]. Combining drugs with different mechanisms of action for synergistic interactions optimizes treatment efficacy and safety [60], and could be particularly beneficial for DMGs that are characterized by a high degree of intratumoral and intertumoral heterogeneity [61,62].

To build upon the preliminary promising efficacy of ONC201 and MBZ in the treatment of DMG, we explored the therapeutic potential of the combination of these two compounds. By using a dose-response matrix design, we found that combined treatments exerted a greater growth- inhibitory activity in four DMG lines compared with either single agent, resulting synergistic in two cultures and additive in the other two. Synergy between drugs vary upon several issues, such as cell-context and drug-class, introducing bias and reproducibility crisis in cancer drug combination discovery firstly in preclinical in vivo studies and secondly in the clinical setting [25,60,63]. Therefore, we validated the effects of the combination MBZ/ONC201 using four reference models developed on the basis of different assumptions, and, importantly, we found high consistency among all the interaction scores, further corroborating the combinatorial efficacy of these two compounds.

Mechanistically, the superior efficacy of MBZ/ONC201 cotreatment in DMG compared to single agents was mediated by enhanced apoptotic signaling rather than mitotic arrest, as evidenced by increased levels of activated caspase 3 and PARP cleavage, that exceeded by far the sum of single agent-induced upregulation of these cell death markers. This may suggest that the combination triggers two different apoptotic pathways, ONC201 relying on ClpP-dependent activation of the integrated stress response (ISR) and up-regulation of the TNF-related apoptosis-inducing ligand (TRAIL) pathway [64], while MBZ triggering activation of caspase-mediated intrinsic and extrinsic mitochondrial pathways [19]. Very importantly, we found the highest scores of sensitivity/additivity at concentrations of both MBZ and ONC201 [8] within the clinically achievable serum levels and regardless of the H3 K27 status of the cell lines. Although ONC201 is in clinical trials for DMG harboring the canonical H3K27M mutation, growing evidence suggests tumor-agnostic indication for ONC201 administration in patients, including wild-type H3 DMG.

To date, a few preclinical studies have found synergistic effects combining ONC201 with other agents, such as those targeting kinases [59], epigenetic modification enzymes [38], and proteasome [38]. Very recently, the combination of ONC201 and PTC596, an investigational antineoplastic small molecule that binds to the colchicine sites of tubulin similarly to MBZ, has demonstrated a strong synergistic growth-inhibitory effects on patient-derived DMG cultures, regardless of their H3K27M-status [65]. Together, these and our findings indicate that co-targeting tubulin and ClpP may be strategic for a patient population that continues to search for effective therapeutic strategies of combating difficult-to-treat disease.

MBZ’s ability to overcome ONC201-resistance suggests that MBZ may block or bypass alternative mechanisms carried out by cancer cells to escape drug treatment. Different molecular mechanisms may contribute to ONC201 resistance, including ClpP protein expression levels and/or mutation. Ectopic overexpression of ClpP sensitizes blood cancer cells to ONC201 [35], whereas loss of ClpP by CRISPR knockout confers resistance [59,66]. ONC201 resistance has also been linked to a D190A mutation occurring in ONC201- selected cell lines [66]. In our DMG model, prolonged exposure to ONC201 determined a substantial decrease in ClpP protein level, consistent with the relationship between ONC201 sensitivity and the expression of its target ClpP. A plausibile explanation for MBZ ability to circumvent the decrease sensitivity to ONC201 is the specific mechanism of action of each agent, targeting MBZ microtubules, while ONC201 being an activator of ClpP-mediated degradation of misfolded proteins. Our finding are in lines with works by others, who documented that ONC201-resistant cells retain similar responsiveness to other drugs, such as adriamycin and vincristine [35], the latter inhibiting tubulin polymerizations as MBZ does.

A limitation of our study is the lack of preclinical in vivo validation, that hampers the assessment of the therapeutic potential of MBZ in a more reliable environment, where the complex tumor architecture along with pharmacokinetics factors may impact on drug delivery to the tumor sites and the response to drug. This limitation is partly due to the very limited number of tumorigenic DMG cell lines and their very slow rate of growth in vivo, which makes preclinical models very cumbersome. We tried to compensate for this limitation by using a large panel of genetically different DMG cell lines, most of which grow in 3D cultures as floating neurospheres, the in vitro models that more reliably resemble in vivo tumors [67]. In addition, literature data from preclinical and clinical studies document high blood-brain-barrier permeability to MBZ and ONC201, suggesting that our results in vitro may likely translate to in vivo setting.

Active research is currently underway to further our understanding of MBZ anticancer mechanisms of action as well as of its clinical efficacy. Although MBZ demonstrates potent effects as a potential new low-toxicity therapeutic for DMG, further investigations in DMG in vivo models are warranted to confirm the preclinical anti-neoplastic activity of MBZ and its safety in combination with other drugs for future translation into a clinical setting.

Conclusion

To the best of our knowledge, this study is the first to thoroughly investigate the anticancer activity of repurposed MBZ in a panel of DMG lines, both as monotherapy and in combination with ONC201. Our study provides the cellular and molecular insights for a novel potential combination strategy that, joining together two low-toxicity, preclinically effective, anticancer compounds, exerts synergistic growth-inhibitory and pro-apoptotic activity in H3K27-mutated and H3-wild-type DMG cell lines. In addition to its chemosensitizing activity to ONC201, we highlight the preclinical potential of MBZ as a second line treatment opportunity in advanced DMG setting.

Disclosure of conflict of interest

None.