In Silico Identification of DNMT Inhibitors for the Treatment of Glioblastoma

Abstract

Background/Objectives: Gliomas are the most common tumours of the central nervous system (CNS), classified into grades I to IV based on their malignancy. Genetic and epigenetic alterations play a crucial role in glioma progression. DNA methyltransferases (DNMTs) are vital enzymes responsible for DNA methylation, with DNMT1 and DNMT3 catalysing the addition of a methyl group to the 5-carbon of cytosine in CpG dinucleotides. Targeting DNMTs with DNA methyltransferase inhibitors (DNMTi) has become a promising therapeutic approach in tumour treatment. In this study, in silico screening tools were employed to evaluate potential inhibitors of DNMT1, DNMT3A, and DNMT3B for the treatment of glioblastoma multiforme (GBM). Methods: The Gene2Drug platform was used to screen compounds and rank them based on their capacity to dysregulate DNMT genes. PRISM viability assays were performed on 68 cell lines, and DepMap data were analyzed to assess the antitumor activities of these compounds and their target genes. Candidate drug similarity was evaluated using DSEA, and compounds with p < 1 × 10⁻³ were considered statistically significant. Gene-compound interactions for DNMT1, DNMT3A, and DNMT3B were confirmed using Expression Public 24Q2, while Prism Repositioning Public data were analyzed via DepMap. Results: Glioblastoma cell lines showed sensitivity to compounds including droperidol, demeclocycline, benzthiazide, ozagrel, pizotifen, tracazolate, norcyclobenzaprine, monocrotaline, dydrogesterone, 6-benzylaminopurine, and nifedipine. SwissTargetPrediction was utilised to identify alternative molecular targets for selected compounds, revealing high-probability matches for droperidol, pizotifen, tracazolate, monocrotaline, dydrogesterone, and nifedipine. Conclusions: Integrating computational approaches with biological insights and conducting tissue-specific and experimental validations may significantly enhance the development of DNMT-targeted therapies for gliomas.

1. Introduction

Although gliomas are the most common tumors of the central nervous system (CNS), and are classified into grades I to IV according to their malignancy level [1,2], their hospital admission rates remain lower compared to other cancers such as malignant neoplasm of lymphoid, hematopoietic, and related tissue [3]. Glioblastoma (GBM) is a highly aggressive, infiltrative, mutative, therapeutically non-responsive, complex, and lethal disease in adults, making it challenging to treat [4]. Genetic and epigenetic alterations play a crucial role in the progression of gliomas. DNA methylation is an epigenetic modification that plays a significant role in gliomas by transcriptional silencing of tumor suppressor genes (TSGs) [5]. Therefore, DNA methylation changes are substantial in treatment planning for glioblastoma multiforme. Several tumor suppressor genes, including MGMT, PTEN, p14ARF, RASSF1A, p16, p73, and Rb, are frequently hypermethylated in gliomas [1].

DNA methyltransferases (DNMTs), including DNMT1, DNMT3A, and DNMT3B are key enzymes responsible for DNA methylation, where they catalyze the addition of a methyl group to the 5-carbon cytosine in a CpG dinucleotide. DNMT1 catalyzes methylation of newly synthesized strands by interacting with hemi-methylated DNA during DNA replication, leading to maintenance of the DNA methylation state. DNMT3A and DNMT3B provide de novo DNA methylation during embryogenesis [6]. Overexpression of DNMTs contributes to cancer initiation and progression by reducing the expression of TSGs in various cancers [5]. Targeting DNMTs with DNA methyltransferase inhibitors (DNMTi) has emerged as a promising strategy for tumor treatment [4,7]. Several DNMT inhibitors, including arsenic trioxide, clofarabine, decitabine, and azacitidine, have been approved to treat haematological malignancies [8].

The Food and Drug Administration (FDA) and the European Medicines Agency (EMA) have approved the nucleoside analogues decitabine and azacitidine for the treatment of myelodysplastic syndrome (MDS), acute myeloid leukaemia (AML), and chronic myelomonocytic leukaemia (CMML). Other cytidine analogues, such as SGI-110 and 5-fluoro-2′-deoxycytidine, are currently undergoing clinical trials for various malignancies [9]. These agents covalently trap DNMTs on DNA to block their function and prevent the methylation of CpG sites, resulting in DNA damage, dose-limiting toxicities, and apoptosis [10]. Although they are approved for use in the treatment of haematological tumours and myelodysplastic syndrome (MDS), they have several side effects, such as fever, thrombocytopenia, and febrile neutropenia [11].

Recent research has shifted focus toward developing non-nucleoside DNMT inhibitors. Several non-nucleoside DNMT inhibitors, including the natural compounds curcumin and epigallocatechin (EGCG), the oligonucleotide MG98, and the local anaesthetic procaine, are undergoing clinical trials for cancer treatment [9]. While most DNMT inhibitors remain in the preclinical phase, some have progressed to clinical trials for anti-tumor research.

The current standard treatment for glioblastoma includes maximal surgical resection followed by radiotherapy and concomitant chemotherapy [12]. Temozolomide, an oral alkylating agent, is commonly used in combination with radiotherapy to treat GBM. However, its use is associated with several adverse effects, including haematological toxicity [13]. Moreover, resistance to temozolomide and radiotherapy remains a major clinical challenge. A key mechanism of resistance involves the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT), which removes alkyl groups from the O6 position of guanine. By reversing this DNA damage, MGMT prevents the formation of mismatched base pairs and thereby reduces the cytotoxic effects of the drug [12,14]. These limitations highlight the urgent need to explore alternative therapeutic approaches for GBM. Drug repurposing, which is identifying new uses for existing medications, offers a promising strategy to overcome treatment resistance and improve outcomes in GBM patients.

Several nucleoside and non-nucleoside DNMTi have been reported to exhibit toxicity or unstable bioavailability, which hampers their use as radiosensitizers in clinics and has led to the discovery of novel biostable DNMTi. Psammaplin A, a bromotyrosine-derived disulfide dimer, is a natural compound that has been reported to exhibit radiosensitising effects as a non-nucleoside DNMTi in human lung cancer and glioblastoma cells in vitro. However, its short half-life in vivo precluded its clinical use [15]. Non-nucleoside inhibitor phthalimido-alkanomide derivative (M17) was reported to significantly induce radiosensitization in all glioblastoma cells without affecting normal astrocytes. However, several obstacles have been suggested that would prevent clinical trials of MA17 in GBM patients [16]. Novel psammalin A derivatives, such as MA14, MA16, and MA22, were reported to have significant radiosensitising effects as non-nucleoside DNMTi for DNMT1, with markedly improved biostability in human lung cancer and glioblastoma cells [17]. Consequently, the discovery of novel DNMTi with non-toxicity, good biostability, and radiosensitising effect is significantly necessary for GBM treatment.

To date, there exist nearly seventy pipelines dedicated to the development of DNA methyltransferase inhibitors (DNMTi) globally. Among these, azacitidine, decitabine, clofarabine, and arsenic trioxide have received regulatory approval for the treatment of haematological malignancies. Several DNMTi have advanced to the clinical stage for anti-tumour research; however, the majority remain in the pre-clinical phase.

Drug repurposing is utilised for non-cancer pharmaceuticals within the realm of cancer therapy, providing numerous advantages in treatment approaches. Integrating computational methods with biological insights and conducting tissue-specific and experimental validations can significantly advance DNMT-targeted therapies for gliomas. The present study employed computational tools to investigate the potential anti-cancer effects of DNMT1, DNMT3A and DNMT3B inhibitors in GBM.

2. Materials and Methods

In silico screening tools were employed to evaluate the potential utility of DNMT1, DNMT3A, and DNMT3B inhibitors in glioma therapy. Gene2drug was used to facilitate this screening by identifying candidate compounds targeting DNMT1 based on their biological and molecular functionalities. Specifically, Gene2drug ranked 1309 molecules for their potential to modulate DNMT1, 1308 molecules for DNMT3A, and 1308 molecules for DNMT3B.

Following this, PRISM viability assays were performed using 68 cell lines in the DepMap website tool (https://depmap.org/portal/ (accessed on 25 March 2025)). The antitumor activities of the candidate drugs were analyzed alongside their target gene dependencies using DepMap data [18,19,20]. To classify glioma cells according to their sensitivity to these compounds, the Expression Public 24Q4 dataset was compared with PRISM Repurposing Public 24Q2 [18,19,20].

Sensitivity scores and gene dependency metrics (Chronos and CERES scores) were obtained from DepMap for the candidate compounds. Correlations between gene expression levels and drug response were evaluated using Spearman’s correlation and ANOVA, as appropriate. Compounds demonstrating selective cytotoxicity against glioma cell lines were prioritized for further investigation [18,19,20].

2.1. Analysis of Gene-Drug Interactions

The Gene2drug database was employed to identify candidate compounds targeting the DNMT1, DNMT3A, and DNMT3B genes based on their biological and molecular functions [21]. Compounds were ranked according to their p-values generated by Gene2drug, reflecting their predicted ability to modulate the target genes. Sensitivity data for these candidate compounds were obtained from DepMap. To ensure disease-specific relevance, only glioma-derived cell lines were included in downstream analyses.

2.2. In Silico Analysis of Compounds

SwissADME was employed for in silico analysis of compounds identified as potential DNMT1, DNMT3A, and DNMT3B inhibitors with high sensitivity in glioma cell lines. This complementary online platform enables the evaluation of pharmacokinetic properties, drug-likeness (predicting oral bioavailability), and medicinal chemistry friendliness, including PAINS alerts, for small molecules [22,23,24,25,26]. In addition, target prediction analyses were performed using the SwissTargetPrediction web tool [23,24,26].

3. Result

The lists of potential compounds for DNMT1, DNMT3A and DNMT3B were sourced from Gene2drug.

3.1. The Responsiveness of Glioblastoma Cells to Potential Drugs

Sensitivity data for glioblastoma cells to potential compounds targeting DNMT1, DNMT3A, and DNMT3B were obtained from PRISM. The relationship between these compounds and their respective target genes was evaluated using Expression Public 24Q4, in comparison with PRISM Repurposing Public 24Q2, via DepMap data. A comprehensive analysis was performed on 1309 compounds targeting DNMT1, 1308 compounds targeting DNMT3A, and 1308 compounds targeting DNMT3B using this integrated approach.

Glioma cells exhibited significant sensitivity to Tracazolate, Vorinostat:Navitoclax (4:1), and Norcyclobenzaprine for DNMT1 dysregulation (

Table 1. The list of compounds targeting DNMT1.

Compound Name	Pearson	Spearman	Slope	Intercept	p-Value (Linregress)
droperidol	−0.470	−0.535	−2.75 × 10−1	9.11 × 10−1	8.81 × 10−3
demeclocycline	−0.436	−0.401	−2.53 × 10−1	6.16 × 10−1	5.54 × 10−3
benzthiazide	−0.465	−0.365	−2.33 × 10−1	2.01 × 10−1	9.54 × 10−3
ozagrel	0.529	0.515	2.28 × 10−1	−6.23 × 10−1	3.77 × 10−3
pizotifen	0.590	0.605	2.71 × 10−1	−9.04 × 10−1	7.53 × 10−4

Compound Name: Name of the compound predicted to modulate DNMT1 activity. Pearson: Pearson correlation coefficient between compound sensitivity and DNMT1 expression. Negative values indicate an inverse relationship; positive values indicate a direct relationship. Spearman: Spearman rank correlation coefficient, indicating monotonic relationships. Slope: Slope of the linear regression between DNMT1 expression and sensitivity; negative slope indicates decreased sensitivity with higher expression. Intercept: Y-intercept of the regression line. p-value (linregress): Statistical significance of the linear regression; values < 0.05 indicate significance.

), and to Droperidol, Demeclocycline, Benzthiazide, Ozagrel, and Pizotifen for DNMT3A dysregulation (

Table 2. The list of compounds targeting DNMT3A.

Compound Name	Pearson	Spearman
tracazolate	−0.497	−0.522
vorinostat:navitoclax (4:1)	0.486	0.490
norcyclobenzaprine	−0.494	−0.447

Compound Name: Name of the compound predicted to modulate DNMT1 activity. Pearson: Pearson correlation coefficient between compound sensitivity and DNMT1 expression. Negative values indicate an inverse relationship; positive values indicate a direct relationship. Spearman: Spearman rank correlation coefficient, indicating monotonic relationships. Slope: Slope of the linear regression between DNMT1 expression and sensitivity; negative slope indicates decreased sensitivity with higher expression. Intercept: Y-intercept of the regression line. p-value (linregress): Statistical significance of the linear regression; values < 0.05 indicate significance.

). For DNMT3B dysregulation, slight sensitivity was observed to Tracazolate, Monocrotaline, Dydrogesterone, 6-Benzylaminopurine, and Nifedipine (

Table 3. The list of compounds targeting DNMT3B.

Compound Name	Pearson	Spearman	Slope	Intercept	p-Value (Linregress)
tracazolate	−0.507	−0.485	−3.63 × 10−1	4.98 × 10−1	4.98 × 10−3
monocrotaline	0.454	0.310	2.65 × 10−1	−7.41 × 10−1	3.72 × 10−3
dydrogesterone	0.525	0.614	3.22 × 10−1	−3.98 × 10−1	2.88 × 10−3
6-benzylaminopurine	−0.571	−0.567	−2.72 × 10−1	4.77 × 10−1	1.21 × 10−3
nifedipine	−0.497	−0.573	−4.40 × 10−1	7.14 × 10−1	1.29 × 10−3

Compound Name: Name of the compound predicted to modulate DNMT1 activity. Pearson: Pearson correlation coefficient between compound sensitivity and DNMT1 expression. Negative values indicate an inverse relationship; positive values indicate a direct relationship. Spearman: Spearman rank correlation coefficient, indicating monotonic relationships. Slope: Slope of the linear regression between DNMT1 expression and sensitivity; negative slope indicates decreased sensitivity with higher expression. Intercept: Y-intercept of the regression line., p-value (linregress): Statistical significance of the linear regression; values < 0.05 indicate significance.

).

Table 1 showed that Droperidol shows a negative correlation (Pearson = −0.470, Spearman = −0.535, p = 0.00881), suggesting that higher DNMT1 expression is associated with decreased sensitivity to this compound. Demecycline and Benzthiazide also show negative correlations, indicating similar inverse relationships with DNMT1 expression. Ozagrel and Pizotifen exhibit positive correlations (Pearson = 0.529 and 0.590, respectively), indicating that glioblastoma cells with higher DNMT1 expression are more sensitive to these compounds, with Pizotifen demonstrating high statistical significance (p = 0.000753). Overall, this analysis identifies compounds with both positive and negative associations between DNMT1 expression and cellular sensitivity, providing a basis for prioritizing candidates for further in vitro validation.

For DNMT3A-targeting compounds (Table 2), glioblastoma cells showed variable sensitivity. Tracazolate and Norcyclobenzaprine exhibited negative correlations with DNMT3A expression (Pearson = −0.497 and −0.494; Spearman = −0.522 and −0.447), indicating reduced sensitivity at higher expression levels. Conversely, Vorinostat:Navitoclax (4:1) showed a positive correlation (Pearson = 0.486, Spearman = 0.490), suggesting enhanced sensitivity in cells with higher DNMT3A expression. These findings highlight compounds with differential effects depending on DNMT expression levels, providing a prioritized list of candidates for subsequent experimental validation in glioblastoma models.

For DNMT3B-targeting compounds (Table 3), glioblastoma cells showed variable sensitivity. Tracazolate, 6-Benzylaminopurine, and Nifedipine exhibited negative correlations with DNMT3B expression (Pearson = −0.507, −0.571, −0.497; Spearman = −0.485, −0.567, −0.573), indicating decreased sensitivity at higher expression. In contrast, Monocrotaline and Dydrogesterone showed positive correlations (Pearson = 0.454 and 0.525; Spearman = 0.310 and 0.614), suggesting increased sensitivity in cells with elevated DNMT3B expression. All correlations were statistically significant (p < 0.01).

3.2. In Silico Draggability Analysis of Compounds

Bioavailability radars were used to illustrate the pharmacological profiles of the selected compounds. The pink-highlighted area in the graphs represents the optimal range for key physicochemical properties—including lipophilicity, molecular size, polarity, solubility, saturation, and flexibility—which are critical for oral bioavailability and potential anti-tumor activity beyond the gastrointestinal tract [22,23,24,25,26].

In addition, SwissTargetPrediction was employed to identify potential alternative molecular targets for the selected compounds. High-probability matches were identified for Droperidol, Pizotifen, Tracazolate, Monocrotaline, Dydrogesterone, and Nifedipine (Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6). In contrast, Demeclocycline, Benzthiazide, Ozagrel, Norcyclobenzaprine, and 6-Benzylaminopurine were excluded because they did not meet the SwissADME criteria [23]. Furthermore, drug combinations such as Vorinostat:Navitoclax (4:1) could not be evaluated in SwissADME due to the nature of the combination [22,23,24,25,26].

4. Discussion

Glioblastoma multiforme (GBM), classified as a grade IV astrocytoma, is one of the most aggressive and malignant primary brain tumors [1]. GBM exhibits significant genomic, transcriptomic and epigenomic heterogeneity, contributing to therapy resistance and cancer recurrence [27,28]. Studies have shown that isocitrate dehydrogenase 1 (IDH1) deficiency can lead to the glioblastoma CpG island methylator phenotype (G-CIMP) by altering the methylome. As a result, the interaction between genetic and epigenomic alterations is significant in glioblastoma but is still unclear [29]. Understanding the molecular mechanisms underlying GBM development is crucial for identifying early diagnostic markers and developing effective therapeutic strategies [13,30]. Additionally, DNA methylation has been found to influence the treatment response of GBM patients, further highlighting its clinical significance [31].

DNA methylation is a crucial epigenetic modification for embryonic development and the maintenance of specific cell functions. There are several DNMTs, including DNMT1, DNMT3A/DNMT3B, that are responsible for DNA methylation, thereby regulating gene expression. Hypermethylation, which involves the methylation of CpG dinucleotides in gene promoter regions, suppresses gene expression. On the other hand, demethylation of the genome, known as global DNA hypomethylation, activates transposon elements and chromosomal instability. Aberrant DNA methylation—global hypomethylation, and especially hypomethylation-induced activation of proto-oncogenes and hypermethylation-induced inactivation of tumour suppressor genes—play a significant role in the occurrence, invasion, metastasis, and treatment resistance of different cancers, including glioblastoma [11].

It has been suggested that satellite two pericentromeric DNA sequences are hypomethylated due to decreased DNMT3A expression, while DNMT1 expression is increased in human glioblastomas [32]. Studies have determined that EZH2 and DNMT1/3B expression is elevated, with DNMTs playing a regulatory role in EZH2 expression in GBMs. High expression levels of EZH2, DNMT1, and DNMT3B were associated with poor prognosis in GBM patients [33,34]. The overexpression of DNMT1 and DNMT3B was associated with alterations in the methylation status of apoptosis-associated genes, including survivin, TMS1, and caspase-8, contributing to apoptotic resistance and a poor prognosis in GB.

The suppression of DNMT1 and DNMT3B was reported to decrease H3K9 and H3K27 methylation levels in the promoters of the DUSP5, SDC2, and TMTC1 genes, thereby increasing their transcriptional activity without altering promoter methylation in GBM [34,35]. Additionally, DNMT1 and EZH2 have been implicated in glioblastoma progression through the epigenetic silencing of miR-200b/a/42 [36]. Furthermore, regulator of chromosome condensation 2 (RCC2) has been reported to enhance proliferation and radioresistance in glioblastoma through p-STAT3-mediated activation of DNMT1 transcription [37].

TRDMT1, 5-methylcytosine RNA methyltransferase with limited DNA methylation capacity, was reported to influence genes involved in cell cycle, apoptosis, and DNA damage response. TRDMT1 knockout (KO) reduced total DNA 5-mC level, as well as and DNMT1 levels and activity, overcoming chemotherapy resistance in glioblastoma cells [28]. The reprogramming factors Oct4 and Sox2 directly induced DNA methyltransferase (DNMT) promoter transactivation, leading to the downregulation of miRNAs such as miRNA-148a, which promotes stem-like and tumor-propagating properties and treatment resistance in GBM [38].

The interaction between NUP37 and DNMT1 has been reported to promote glioma cell proliferation, invasion, and migration. Notably, NUP37 depletion led to a decrease in DNMT1 protein levels in glioma cells [39]. The increased expression of DNMT1, DNMT3A, and DNMT3B, along with global DNA hypomethylation and PPARγ hypermethylation, has been linked to the development of GBM. A positive correlation between global DNA hypomethylation and reduced PPARγ protein expression has been observed in GBM patients, although no significant relationship was found between DNMTs and PPARγ expression [30].

Furthermore, miR-148-3p expression was negatively correlated with DNMT1 expression but positively correlated with RUNX3 expression in GBM. The miR-148-3p was reported to regulate the DNMT1-RUNX3 axis by directly inhibiting DNMT1 expression, thereby suppressing proliferation, migration, invasion, and epithelial–mesenchymal transition (EMT) in GBM [40]. Overexpression of DNMT1 has also been reported to inhibit the expression of p21 tumor suppressor genes in gliomas. The miR-29b was suggested to indirectly inhibit DNMT1 expression by targeting Sp1 in gliomas [41]. Additionally, SH2 domain-containing adaptor protein F (SHF) has been shown to inhibit STAT3 dimerization, thereby blocking the STAT3-DNMT1 interaction and impairing GBM tumorigenesis [42]. A recent study demonstrated that a curcumin-chitosan nanocomplex decreased the expression of DNMT1, 3A, and 3 B in glioblastoma cells, suggesting a novel therapeutic approach [43]. It revealed that miR-29s significantly inhibited proliferation, migration, and invasion, and promoted apoptosis by blocking the expression of DNMT3A and DNMT3B in the U87MG glioblastoma cell line [44].

Here, we utilised in silico tools to identify potential compounds for targeting DNMT1, DNMT3A, and DNMT3B in glioblastoma multiforme (GBM). Glioblastoma cell lines demonstrated sensitivity to various compounds, including droperidol, demeclocycline, benzthiazide, ozagrel, pizotifen, tracazolate, norcyclobenzaprine, monocrotaline, dydrogesterone, 6-benzylaminopurine, and nifedipine. The SwissTargetPrediction tool was employed to identify potential alternative molecular targets for these selected compounds, revealing high-probability matches for droperidol, pizotifen, tracazolate, monocrotaline, dydrogesterone, and nifedipine.

Droperidol is a butyrophenone-derived dopamine D2 receptor antagonist used to improve chemotherapy-associated nausea, vomiting, and postoperative pain in oncological settings [45]. It was reported that droperidol can lead to hypotension by inhibiting alpha-adrenergic receptor [46]. Upregulation of EVPL and downregulation of ENTPD3 have been demonstrated as molecular properties associated with an abnormal immune state in both type 2 diabetes mellitus (T2DM) and colorectal cancer (CRC). Droperidol was determined to be inhibitor of EVPL and agonist of ENTPD3. Therefore, it was suggested that droperidol may be a promising drug for T2DM and CRC [47].

Ozagrel is a thromboxane A2 (TXA2) synthase inhibitor. Ozagrel is an antiplatelet agent and also an effective agent for the treatment of bronchial asthma, occlusive vascular disease and diabetic nephropathy [48]. Ozagrel was demonstrated to improve endothelial dysfunction, oxidative stress, and neuroinflammation in a rat model of streptozotocin diabetes induced dementia [49]. It was suggested that ozagrel can prevent the 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK)-mediated stemness of lung cancer stem cells (LCSCs) by blocking thromboxane synthase (TXS). Ozagrel was reported to induce apoptosis by increasing caspase-3 activity and reducing surviving expression in T24 and transitional cell carcinoma TCC-SUP bladder cancer cells [50]. It was suggested that ozagrel may be a potential agent for targeting the human cytochrome P450 enzyme CYP4Z1 in breast cancer, as indicated in an in silico model [51]. Here, we demonstrated the possible anticancer effects of Ozagrel in GBM cell lines through computational analysis.

Pizotifen is a serotonin or 5-hydroxytryptamine (HT)_2A receptor antagonist used for the treatment of vascular headaches such as migraine. Pizotifen was demonstrated to inhibit viability, migration, and invasion, while inducing apoptosis in gastric cancer. In addition, it was suggested that pizotifen may inhibit migration and invasion by inhibiting the Wnt/B-catenin-EMT signalling pathway [52]. Another study indicated that pizotifen may prevent proliferation and migration by inhibiting Wnt signaling pathway in colon cancer HCT116 cells [53]. It has been declared that pizotifen could show a neuroprotective effect [54]. Pizotifen was reported to decrease the transcriptional activity of β-catenin by inhibiting HTR2C, thereby suppressing the epithelial-to-mesenchymal transition (EMT)- mediated metastasis of human cancer cells in a zebrafish xenotransplantation model. Here, we demonstrated the potential anticancer effect of pizotifen in GBM cell lines using in silico analysis. Nifedipine (NIFE) is a dihydropyridine L-type calcium channel blocker, the most effective and safe drug based on the World Health Organization (WHO) list of essential medicines. NIFE is used for all types of hypertension. NIFE have been indicated to prevent NFAT2 nuclear translocation, thereby suppressing colorectal cancer progression and immune resistance [55,56]. It was suggested that nifedipine can induce apoptosis by increasing the anti-tumour effects of cisplatin on cisplatin-sensitive human glioblastoma U87-MG cells and cisplatin-resistant human glioblastoma U87-MB-CR cells lacking a Ca2+-dependent endonuclease [57]. Cisplatin/nifedipine treatment was reported to increase lysosome-associated membrane protein (LAMP)-1 expression, leading to apoptosis in human glioblastoma cell lines, U-373 MG and LN-Z308 [58]. It was reported that nifedipine significantly inhibits the invasion of U87-MG cells, but has a slightly inhibitory effect on the invasion of U251-MG cells. It was suggested that nifedipine can inhibit invasion by blocking high voltage-activated classes of Ca2+ channels implicated in cellular motility pathways [59]. Dydrogesterone is a stereoisomer of progesterone (inverse progesterone) that was determined to be a safe and effective therapeutic agent for the treatment of endometriosis. It was suggested that dydrogesterone could treat endometriosis by inhibiting the proliferation of ectopic endometrium, inducing apoptosis-mediated atrophy of ectopic endometrium, and reducing inflammation [60]. Dydrogesterone is also used in the treatment of menstrual disorders as well as postmenopausal hormone replacement. Dydrogesterone has been documented to infrequently induce adverse effects such as weight gain and oedema, attributable to its low antagonistic activity on glucocorticoid and mineralocorticoid receptors in comparison to progesterone [61]. It has also been reported that dydrogesterone can promote a healthy pregnancy by modulating pro-inflammatory cytokines. Consequently, it is recommended that dydrogesterone may serve as an effective and safe therapeutic approach in cases of recurrent spontaneous miscarriages [62].

Monocrotaline (MCT) is a plant-derived compound traditionally used in research to induce pulmonary hypertension in animal models [63]. The ETAR/miR-27b-3p/FBXW7/KLF5/GLI1 pathway was indicated to contribute to MCT-induced pulmonary arterial hypertension (PAH) development in rats [64]. MCT-induced PAH was reported to be suppressed through the inhibition of miR-27b in rats [65]. Although the low toxicity of MTC, it was demonstrated that intermediate metabolites derived from metabolised MTC in the liver can lead to liver damage by binding to proteins containing thiol (-SH), hydroxyl (-OH), and amino (-NH) groups, as well as RNA and DNA [66]. Monocrotaline-mediated autophagy was reported to induce apoptosis by inhibiting the PI3K/AKT/mTOR pathway in rat hepatocytes [67]. An in silico docking study showed that monocrotaline can have antineoplastic activity for hepatocellular carcinoma by impacting p53, HGF, and TREM1. In vitro experimental validations of the in silico study suggested that monocrotaline is a promising agent with high efficacy and low side effects for hepatocellular carcinoma [68]. Interestingly, it demonstrated a moderate positive correlation with DNMT3B inhibition in glioma cell lines. While monocrotaline’s toxicity limits its clinical use, these results raise the possibility of structural derivatives being developed as selective DNMT3B-targeting agents in GBM [63].

Tracazolate is a non-benzodiazepine anxiolytic compound that modulates GABAergic activity [69]. Tracazolate has anticonvulsant, anxiolytic, sedative, and muscle relaxant effects and was used only in research. The tracazolate was suggested to be an effective and safe agent for the treatment of postpartum depression (PPD) in humans [70]. It demonstrated consistent inverse correlations with both DNMT1 and DNMT3B, suggesting broad inhibitory potential across DNMT isoforms. This makes tracazolate a promising candidate for further study in epigenetic drug repurposing for glioma treatment [69].

The adverse effects of DNA methyltransferase (DNMT) inhibitors can be attributed to two main factors. First, on-target effects arise from the inhibition of DNMT enzymes, which may disrupt normal epigenetic regulation and gene expression patterns, potentially leading to unintended biological consequences [71,72]. Second, off-target or compound-related effects are associated with the intrinsic chemical properties of nucleoside analogs, including their incorporation into DNA or RNA, which can induce cytotoxicity, DNA damage, and mutagenesis [73,74]. Recognizing this distinction allows for a more nuanced and balanced discussion of the safety profile of DNMT inhibitors. Moreover, our preliminary data will guide subsequent in vitro studies aimed at elucidating the underlying molecular mechanisms.

It is also crucial to note that the results shown here are predictions. The actual sensitivity of cells to these compounds can only be determined through experimental validation in cell culture systems. Other factors, such as cellular context, metabolic activity, and the presence of compensatory pathways may also influence this sensitivity. Therefore, while the current data provide valuable insights and hypotheses, they should be seen as predictive rather than definitive, highlighting the need for follow-up in vitro studies to confirm and refine these observations.

Nevertheless, this study’s limitations are attributed to its exclusive reliance on in silico data and the restricted number of cell lines analysed. Verifying the in silico predictions may necessitate animal models, cell culture studies, and, when applicable, patient-derived samples. Integrating these in silico insights with prevailing biological knowledge and conducting additional analyses that consider tissue-specific variables, in collaboration with experimental biologists, can produce a more comprehensive understanding. We evaluated 1308 compounds across 68 cell lines. Glioblastoma cell lines showed sensitivity to compounds including droperidol, demeclocycline, benzthiazide, ozagrel, pizotifen, tracazolate, norcyclobenzaprine, monocrotaline, dydrogesterone, 6-benzylaminopurine, and nifedipine. SwissTargetPrediction was utilized to identify alternative molecular targets for selected compounds, revealing high-probability matches for Droperidol, pizotifen, tracazolate, monocrotaline, dydrogesterone, and nifedipine.

5. Conclusions

In this research, we employed in silico tools and demonstrated the potential application of Droperidol, pizotifen, tracazolate, monocrotaline, dydrogesterone, and nifedipine in glioblastoma cell lines. Furthermore, target prediction analysis identified cancer-related molecular targets exhibiting very high probability scores, thereby suggesting the polypharmacological profiles of these agents. These bioinformatic predictions will necessitate validation in subsequent studies to augment our comprehension of the therapeutic significance of these compounds as potential candidates for repurposing in cancer treatment therapies.