Siramesine induced cell death of glioblastoma through inactivating the STAT3-MGMT signaling pathway

Abstract

Background

Glioblastoma (GBM), the most prevalent invasive primary brain tumor in adults, is characterized by high mortality, frequent recurrence, and short survival time. The effectiveness of temozolomide (TMZ) in treating GBM is compromised by chemoresistance. Elevated levels of O⁶-methylguanine DNA methyltransferase (MGMT) and the activation of signal transduction and activator of transcription 3 (STAT3) have been associated with GBM’s resistance to TMZ chemotherapy. Although siramesine (Sira) has demonstrated antitumor activity in various cancers, its potential therapeutic effect in gliomas remains uncertain.

Methods

Cell counting kit-8 (CCK-8), healing assay, and clone formation assay were utilized to assess the cell viability, proliferation, and migration of glioma cells. RNAseq, molecular docking analysis, pull-down assay, qPCR, and pharmacologic treatment were employed to explore the cell signaling pathway. Tumor growth and STAT3 signaling pathway proteins in U87-MG cell-derived xenografts from nude mice were examined using Western blot.

Results

Our study demonstrated that Sira reduces cell viability, inhibits proliferation and migration, and induces autophagy in GBM cells. Sira promotes GBM cell death by binding to STAT3 and inhibiting phosphorylation of STAT3(Y705). Combination therapy of TMZ and Sira synergistically induced cell death and inhibited GBM cell proliferation and migration, potentially linked to decreased p-STAT3(Y705) and MGMT levels. Furthermore, using STAT3 signaling activators and inhibitors, we confirmed that reduced MGMT levels were mediated by STAT3 inactivation. Cell-derived xenografts from nude mice revealed that Sira did not affect glioma growth. However, it did inhibit the JAK2-STAT3-MGMT signaling pathway and decrease glioma stem cell properties.

Conclusion

Our findings suggest that Sira inhibits the STAT3-MGMT signaling pathway, slowing tumor growth and increasing sensitivity to TMZ in GBM. These results lay a solid foundation for developing new GBM therapies, with Sira holding promise as a candidate for combined TMZ chemotherapy in GBM.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-025-06693-y.

Introduction

Glioblastoma (GBM) stands as the most prevalent invasive primary brain tumor in adults, with a median survival time of only 12.1 months [1, 2]. At present, GBM is predominantly addressed through surgical resection, radiotherapy, and temozolomide (TMZ) chemotherapy. Nonetheless, owing to its rapid development, pronounced heterogeneity, chemoresistance, and high relapse rate, treatment options for GBM remain limited [3]. As a result, there is considerable interest in researching new treatment targets and biomarkers for precise treatment.

Signal transducer and activator of transcription 3 (STAT3) is recognized for its crucial role in the cellular response to external signals, such as growth factors and cytokines, and is considered a critical pathogenic factor in GBM [4, 5]. Activation of STAT3 promotes tumor cell proliferation, growth, and evasion of cell death in gliomas. It is closely linked to the malignancy grade and prognosis of GBM [6, 7]. However, the potential use of STAT3 in GBM treatment requires further exploration. Persistent phosphorylation of STAT3 has been identified in various cancers [8]. Generally, STAT3 is recruited from the cytoplasm to the cell membrane and interacts with various cytokine receptors through the Src Homology 2 (SH2) domain. This process results in phosphorylating its Tyr705 residue, leading to dimerization and translocation to the nucleus, where it binds to the palindromic DNA consensus sequence [8]. In most cases, hyperactivation of STAT3 promotes tumor progression by regulating biological processes such as epithelial-mesenchymal transition, proliferation, metastasis, cell cycle progression, stem cell formation, and resistance to therapy, all of which are associated with a poor clinical prognosis [9–12]. Thus, inhibiting STAT3 phosphorylation could be an effective anti-cancer strategy. Consequently, developing an inhibitor that blocks STAT3 phosphorylation, dimerization, nuclear translocation, or STAT3 DNA binding activity is warranted.

The ability of cancer cells to evade the caspase-dependent apoptotic pathway has prompted the investigation of innovative anticancer approaches targeting lysosomes [13]. Siramesine (Sira, Lu-28-179), a lysosomotropic agent, has been discovered to directly destabilize lysosomal membranes, leading to lysosomal dysfunction. These biological effects result in increased membrane permeability, the release of cathepsin into the cytoplasm, and, ultimately, cathepsin-mediated cell death [14, 15]. Initially intended for the treatment of anxiety and depression, Sira is well-tolerated and non-toxic in humans [16]. Moreover, Sira has exhibited high toxicity to tumor cells both in laboratory tests and animal studies [15, 17]. Its minimal side effects and potential involvement in inducing caspase-independent cell apoptosis have positioned it as a promising anticancer medication [18, 19]. In a study by Stine S. Jensen, Sira was observed to eliminate standard glioma cell lines effectively, induce tumor cell death, and reduce the formation of secondary spheroids in patient-derived spheroid cultures [20]. Similarly, a study by Villalpando-Rodriguez demonstrated that the combination of Sira and lapatinib synergistically induced cell death in the glioma cell line U87-MG [21]. Sira prompted cell death in immortalized and tumorigenic cells through lysosomal cathepsin leakage and oxidative stress [15]. However, further research is required to elucidate the mechanism of Sira in gliomas fully.

Given the role of lysosomal dysfunction and STAT3 signaling in TMZ resistance [7, 14, 15, 22], we hypothesize that Sira could enhance the efficacy of TMZ by disrupting lysosomal integrity and inhibiting STAT3 signaling. Specifically, we propose that Sira-induced lysosomal disruption could impair the cellular machinery responsible for DNA repair, including MGMT, thereby sensitizing GBM cells to TMZ-induced DNA damage. Furthermore, the inhibition of STAT3 signaling by Sira could downregulate O⁶-methylguanine DNA methyltransferase (MGMT) expression and other STAT3-regulated genes involved in chemoresistance, leading to a synergistic effect when combined with TMZ.

In this study, we aim to investigate the potential of Sira to enhance the efficacy of TMZ in GBM. we initially determined the in vitro concentration of Sira by calculating the mean inhibitory concentration (IC₅₀) values for three GBM cell lines and one astrocyte cell line HA1800. Subsequently, we assessed the impact of Sira on GBM cell migration and proliferation. We then delved into the potential molecular mechanisms through RNAseq, molecular docking analysis, pull-down assay, and surface plasmon resonance (SPR) assay. Additionally, we examined the efficacy of the drug combination of Sira and the first-line chemotherapeutic agent TMZ in GBM cells using combination index (CI) analysis, wound healing assay, and clone formation assay. Finally, the effects of the drug combinations were validated in a U87-MG cell-derived xenograft model (CDX) to confirm the potential preclinical use of this innovative drug combination.

Materials and methods

Cell lines and reagents

The GBM cell lines U87-MG and U251-MG (10th and 18th generation) used in this study were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). The T98G cells (8th and 15th generation) were sourced from iCell Bioscience Inc. (Shanghai, China). Non-neoplastic controls included the HA1800 human glial cell line (Jining Shiye Co., Ltd., Shanghai, China), HT22 murine hippocampal neuronal cells (Shanghai Zhongqiaoxinzhou Biotech Co., Ltd., Shanghai, China), and SH-SY5Y human neuroblastoma cells (QuiCell Co., Ltd., Shanghai, China). All cell lines were authenticated via short tandem repeat (STR) profiling. The cells were cultured in high glucose DMEM (Cat. No. 04‑052‑1ACS; Biological Industries) with 10% FBS (Cat. No. 04‑001‑1ACS; Biological Industries) and 1% penicillin-streptomycin (Cat. No. P1400; Beijing Solarbio Science & Technology Co., Ltd.) at 37˚C in a humidified incubator with 5% CO₂.

The reagents used were Sira (Shanghai Selleck Chemicals Co., Ltd.; Cat. No. S6756), TMZ (Shanghai Selleck Chemicals Co., Ltd.; Cat. No. S1237), Colivelin (MedChemExpress LLC Co., Ltd; Shanghai, China, Cat. No. HY-P1061), AG490 (MedChemExpress LLC Co., Cat. No. HY-12000), C188-9 (MedChemExpress LLC Co., Ltd; Shanghai, China, Cat. No. HY-112288), Wortmannin (WOR, Selleck Chemicals Co., Ltd.; Shanghai, China; Cat. No. S2758; 10 µM), Chloroquine (CQ, Selleck Chemicals Co., Ltd.; Cat. No. S6999; 20 µM), Z-VAD-FMK (MedChemExpress LLC Co., Ltd.; Cat. No. HY-16658B; 20 µM). The inhibitors were applied for 1 h before adding Sira. The GBM cell lines were exposed to the inhibitors and Sira for 48 h before assessing cell proliferation and viability. The selected concentrations of Sira and TMZ were justified based on prior literature [14, 20], preliminary dose-response experiments, and clinical relevance. The authors ensured that the concentrations used were pharmacologically relevant, covered both sub-lethal and therapeutic doses, and accounted for cell line-specific sensitivities and resistance profiles.

Immunofluorescence and confocal microscopy

U87-MG and T98G glioma cells were cultured on glass coverslips in six-well plates. The cells were treated with Sira (5 µM for U87-MG, 10 µM for T98G) or DMSO (solvent control) for 48 h at 37 °C. Subsequently, the cells were washed with ice-cold phosphate-buffered saline (PBS), fixed with freshly prepared 4% paraformaldehyde (PFA) in PBS at room temperature (RT) for 15 min, and permeabilized with 0.3% Triton X-100 in PBS for 20 min. Afterward, the samples were blocked with 5% BSA for 1 h at RT and then incubated with phosphorylated STAT3 at tyrosine 705 (p-STAT3 [Y705]) antibody (Abcam Co., Ltd.; Shanghai, China, Cat. No. EPR23968-52, 1:2000) or STAT3 antibody (Huaan Biotechnology Co, Ltd.; Hangzhou, China, CatNo. ET1607-38, 1:2000) at 4 °C overnight, followed by incubation with the corresponding fluorescein-labeled secondary antibody for 2 h at RT. The nuclei were stained with DAPI for 15 min at RT. Finally, the coverslips were mounted in a 50% glycerol-PBS anti-fade mounting medium. The images were captured using an OLYMPUS confocal microscope (magnification, x100 for T98G, x200 for U87-MG) and analyzed using Adobe Photoshop CS6 and Image-Pro Plus (v6.0; Media Cybernetics, Inc.).

Cell counting kit-8 (CCK-8) assay

Cell viability was assessed using the CCK-8 assay kit (Meilun Biotechnology Co., Ltd., Dalian, China; Cat. No. MA0218) following the manufacturer’s protocol. The experimental procedure was conducted as follows: U87-MG and U251-MG cells were seeded at a density of 5.0 × 10³ cells/well, while T98G cells were seeded at 3.0 × 10³ cells/well in 96-well plates. Four time points (0 h, 12 h, 24 h, and 48 h) were established for analysis. Cell suspensions were prepared according to the specified cell densities, with triplicate wells for each condition. After inoculation, the plates were incubated for 24 h in a cell culture incubator to ensure cell attachment. Following the attachment period, the culture medium was aspirated and replaced with fresh medium containing various drug concentrations (200 µL/well). The time of drug addition was designated as 0 h. Based on preliminary experiments, the optimal incubation time for CCK-8 reagent was determined. Subsequently, 10 µL of CCK-8 reagent was added to each well, followed by a 2-hour incubation in the dark. The absorbance was then measured at 450 nm using a Multiskan Mk3 microplate reader (Thermo Fisher Scientific, Inc.). The cell viability was calculated using the following formula:

Cell viability (%) = [(As - Ab) / (Ac - Ab)] × 100%

where As represents the absorbance of the experimental well, Ab denotes the absorbance of the blank well, and Ac indicates the absorbance of the control well. All data were recorded and analyzed for subsequent interpretation.

Colony formation assay

GBM cells were seeded in six-well plates with 200 cells per well and incubated for 24 h with or without the specified drug treatment. Afterward, the medium was replaced with a fresh growth medium. After two weeks, the colonies were fixed in 100% methanol and stained with 0.1% crystal violet in 20% methanol for 15 min. The number of colonies was quantified using Image J software (1.48v, National Institutes of Health, USA).

Wound-healing assay

To assess the impact of Sira on the migration of GBM cells, an in vitro wound healing experiment was conducted. GBM cells were cultivated in 24-well plates to establish a monolayer, and wounds were created via incisions made with sterile 200-µl plastic micropipette tips. The cells were subsequently exposed to various medications for 12 h. The migration area was then photographed at the exact location and analyzed by comparing the final slit width with the control group. For each condition, triplicate wells were imaged at three standardized positions, and migration rates were averaged across technical replicates within each biological experiment.

Quantitative real-time reverse transcription polymerase chain reaction analysis (qRT-PCR)

In order to assess the impact of Sira on mRNA expression of STAT3-related genes, a qRT-PCR analysis was carried out. GBM cells were plated in a six-well plate (2.5 × 10⁴ cells/well for U87-MG and U251-MG, 2.0 × 10⁴ cells/well for T98G). After 48 h of incubation, the media was replaced, and the GBM cells were cultured for an additional 48 h under the corresponding drug treatment. Subsequently, total RNA was extracted following the manufacturer’s protocols (Invitrogen, Life Technologies, Grand Island, NY, USA). CDNA was synthesized from the isolated RNAs (2 µg) using Superscript reverse transcriptase (Cowin Biotech Co., Ltd.; Jiangsu, China, Cat. No. CW2569M) and then amplified with Platinum Taq polymerase using the Superscript One-Step RT-PCR Kit (Yeasen Biotechnology Co., Ltd.; Shanghai, China, Cat. No. 11202ES08). The relative expression of SHP1, SHP2, PIAS3, and SOCS3 was analyzed by qRT-PCR, with the results expressed as relative expression normalized to GAPDH. The following primers were utilized: SHP1 (forward), 5′-GCA GAT GGC GTG GCA GGA G-3′; SHP1(reverse), 5′-CCG CAG TTG GTC ACA GAG TAG G-3′; SHP2 (forward),5′-TGA CAA CGA AAG AAG TGG AGA GA GG-3′; SHP2 (reverse), 5′-CCT AAC ACG CAT GAC GCC ATA TTC-3′; PIAS3 (forward), 5′-GCT GGG CGA ATT AAA GCA CA-3′; PIAS3 (reverse), 5′-TTC CGT CCA CTC TTG TTC CG-3′; SOCS3 (forward), 5′- ATG GTC ACC CAC AGC AAG TT-3′; SOCS3 (reverse), 5′-GTC ACT GCG CTC CAG TAG AA-3′; GAPDH (forward),5′-TGT GTC CGT CGT GGA TCT GA-3′; GAPDH (reverse), 5′-TTG CTG TTG AAG TCG CAG GAG-3′.

Molecular docking assay

Molecular docking studies were conducted using Autodock Vina 4.2.6 to assess the binding between Sira and the STAT3 protein. The UniProt ID for the STAT3 protein and its structure were retrieved from the UniProt database (https://www.UniProt.org/) and the PDB database (https://www.rcsb.org/), respectively. The protein structure of STAT3 was then downloaded and preprocessed to remove water and organic components using PyMOL software. The 2D structure of Sira was obtained from PubChem (https://pubchem.ncbi.nlm.nih.gov/). The PubChem CID of Sira was 9,829,526, and its molecular weight was 454.6 g/mol. Subsequently, the molecular docking was performed using AutoDock Vina, and the AutoDockTools 1.5.7 package [23] was utilized to generate the docking input files. The top-ranked pose based on the Vina docking score was selected, and visual analysis was carried out using PyMOL 1.7.6 software (http://www.pymol.org/).

SPR assay

Ligand immobilization

SPR analysis was performed on a Biacore T200 instrument (29147020, Cytiva) using a CM7 sensor chip (Cytiva, Cat. No. 29147020). The chip was oriented with the labeled surface facing upward and inserted into the instrument. Channel 4 was activated with a 10 µL/min injection of 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC, GE Healthcare) /N-hydroxysuccinimide (NHS, GE Healthcare) for 800 s. Ligand immobilization proceeded in two phases: (1) 20 µg/mL STAT3 protein (UA, Cat. No. UA080435) in sodium acetate (pH 4.5) at 10 µL/min for 300 s, followed by (2) 40 µg/mL STAT3 protein (UA, Cat. No. UA080435) in sodium acetate (pH 4.0) at 10 µL/min for 900 s. Residual reactive groups were deactivated with ethanolamine (10 µL/min, 800 s). Channel 1 served as a reference and underwent identical activation/blocking without ligand coupling.

STAT3-Sira interaction assay

Solvent optimization: The running buffer was replaced with 1×PBS-P+ (Cytiva, Cat. No. 28995084) containing 5% DMSO (v/v), and the system was primed to ensure equilibration.

Binding kinetics: Serially diluted Sira (prepared in 96-well plates, Maclin, Cat. No. 147817-50-3) was injected over channels 1 (reference) and 4 (active) at 30 µL/min, with 60 s association and 60 s dissociation phases. Triplicate injections were performed for each concentration.

Data analysis

Reference-subtracted sensorgrams were analyzed using BIAcore T200 Evaluation Software. Equilibrium dissociation constants (K_D) were derived by global fitting to a 1:1 Langmuir binding model, incorporating corrections for mass transport and baseline drift.

Datasets

We obtained TCGA datasets, which included GBM phenotypes, GBMLGG gene expression RNA sequencing (TCGAseq), and GBMLGG phenotypes, from the official TCGA website (https://portal.gdc.cancer.gov/). The RNA sequencing data of diffuse gliomas (CGGAseq) was also downloaded from the CGGA website (http://www.cgga.org.cn/). We also downloaded the GSE16011 dataset, comprised of the Series Matrix File (GSE16011mic) and GPL8542, from the Gene Expression Omnibus (GEO) (https://www.ncbi.nlm.nih.gov/geo/). In our analysis, we utilized data from 614 cases in TCGAseq, 321 cases in CGGAseq, and 284 cases in GSE16011mic to study STAT3 mRNA expression in gliomas. Furthermore, we examined the data of 399 primary glioma radiotherapy cases in TCGAseq, 126 primary glioma chemoradiation cases in CGGAseq, and 263 glioma cases in GSE16011mic to analyze the overall survival (OS) of glioma patients.

Western blot (WB)

The GBM cell’s total protein was harvested and lysed in a cold RIPA buffer with protease inhibitors after drug treatment. The protein concentration was quantified using the Pierce BCA Assay Kit (Solarbio Science & Technology Co., Ltd.; Beijing, China, Cat. No. PC0020). Equal amounts of protein were loaded onto SDS-PAGE gels using the Bio-Rad Vertical Electrophoresis System and subsequently transferred to polyvinylidene fluoride (PVDF) membranes (Merck Millipore Co., Ltd.; Ireland, Cat. No. IPVH85R). After blocking with 5% milk in Tris-buffered saline (0.1% Tween) for 1 h, the PVDF membranes were incubated with the primary antibodies overnight at 4 °C. The primary antibodies used were as follows: anti-STAT3 (Huaan Biotechnology Co., Ltd.; Cat. No. ET1607-38, 1:2000), anti-p-STAT3(Y705) (Abcam Co., Ltd.; Cat. No. EPR23968-52, 1:2000), anti-JAK2 (Huaan Biotechnology Co., Ltd.; Cat. No. ET1607-35, 1:3000), anti-p-JAK2(Y1007 + Y1008) (Huaan Biotechnology Co., Ltd.; Cat. No. ET1607-34, 1:3000), anti-MGMT (Huaan Biotechnology Co., Ltd.; Cat. No. ET1701-55, 1:2000), anti-STAT1 (Selleck Co., Ltd.; Cat. No. F0263, 1:1000), anti-rH2AX (Cell Signaling Technology Co., Ltd.; Cat. No. 9718T, 1:1000), anti-AKT (Cell Signaling Technology Co., Ltd.; Cat. No. 9272 S, 1:1000), and anti-p-AKT (Cell Signaling Technology Co., Ltd.; Cat. No. 4060 S, 1:2000). Following washing with Tris-Buffered Saline with Tween 20 (TBST), the PVDF membranes were then incubated with the corresponding biotinylated horseradish peroxidase (HRP) secondary antibodies for 1 h at RT. The protein bands were visualized using an enhanced chemiluminescent substrate (Dalian Meilun Biology Technology Co., Ltd.; Cat. No. MA0186). The relative gray values of the protein bands were subsequently calculated using Image J software.

Pull-down assay

Commercial Sepharose 4B beads (Cat. No. S8711) were purchased from Beijing Solarbio Science & Technology Co, Ltd (Beijing, China). Sira-Sepharose 4B beads were subsequently prepared according to the manufacturer’s instructions. Cell lysates (500 µg) were treated with Sira-Sepharose 4B beads or with Sepharose 4B beads only in 1×lysis buffer (5 M Tris-HCl pH 7.5, 5 mM EDTA, 150 mM NaCl, 1 mM dithiothreitol, 10% NP-40, 2 mg/ml bovine serum albumin, 20× protease inhibitor [1 tablet each]) and rotated overnight at 4 °C. The beads were washed with a wash buffer (1 M Tris–HCl pH 7.5, 0.5 M EDTA, 1 M NaCl, 1 M dithiothreitol, 10% NP-40, and 0.1 M PMSF). The STAT3 proteins bound to the beads were subsequently analyzed by WB.

Combination index (CI) calculation

The interaction between Sira and TMZ was evaluated using the Chou-Talalay method. In this method, the combination index (CI) was determined as follows: CI = dsx/Dsx + dtx/Dtx, where dsx and dtx are the doses of Sira in combination with TMZ, and Dsx and Dtx are the doses required for Sira and TMZ to achieve the same effect individually. CI values less than 1, equal to 1, and greater than 1 indicate synergistic, additive, and antagonistic effects, respectively.

In vivo xenograft assay

A total of 21 male nude mice (4–5 weeks old, BALB/c) were purchased from Sibeifu Biotechnology Co., Ltd. (Beijing, China). The mice were then randomly divided into four groups, with six mice in the diluent control group and five mice in each of the drug treatment groups. Subcutaneous implantation of 5 × 10⁶ U87-MG cells was performed into the axilla of the right forelimb of the nude mice. One week post-implantation, the mice were subjected to intraperitoneal injections of saline (administered daily), Sira (1 mg/kg, dissolved in PEG300, injected every other day), TMZ (20 mg/kg, dissolved in saline, injected daily), and a drug combination of Sira (1 mg/kg, dissolved in PEG300) and TMZ (20 mg/kg in saline) (TMZ was injected daily, and the combination of the two drugs was injected every other day). The mice’s body weight and tumor size were monitored every two days, and tumor growth was calculated using the formula: long diameter × (short diameter^2) / 2. Upon completion of the experiment, all mice were euthanized via intraperitoneal administration of 1.25% tribromoethanol (0.2 ml/10 g, dissolved in saline, Cat. No. M2910, Nanjing Aibei Biotechnology Co., Ltd.) followed by cervical dislocation, and the tumor tissue was extracted. Tumor tissue lysate or ice-cold sections were prepared from the tumors for WB or hematoxylin and eosin (H&E) analysis. The present study was approved by the Life Science Ethics Committee of Zhengzhou University (Zhengzhou, China). It should be noted that all animal experimentation, including the euthanasia of animals, adhered to the Guidance on the operation of the Animals (Scientific Procedures) Act 1986 and associated guidelines, EU Directive 2010/63 for the protection of animals used for scientific purposes or the NIH (National Research Council) Guide for the Care and Use of Laboratory Animals.

Statistical analysis

The data was presented as mean ± standard deviation. This study used GraphPad Prism (version 6.02) for statistical analysis. All experiments were conducted with a minimum of three replicates. The difference between the two groups was determined using Student’s t-test. Variance (one-way or two-way ANOVA) was carried out analyzed for appropriate multivariate analyses. Post hoc tests were used to assess the statistical significance of the means, with a threshold of P < 0.05 considered statistically significant.

Results

Sira inhibits viability, migration, and proliferation of GBM cells

The chemical structure of Sira hydrochloride is shown in Fig. 1A. To evaluate the effect of Sira on GBM cells, we used cell lines U87-MG, U251-MG, and T98G to investigate cell viability. Cells were treated with increasing concentrations of Sira (0-100 µM) for 48 h, and cell viability was measured using the CCK-8 assay to determine the IC₅₀ of Sira. The concentration-response curves showed that the IC₅₀ values of Sira in U87-MG, U251-MG, and T98G cells were 8.875 µM, 9.654 µM, and 7.236 µM, respectively (Fig. 1B-D). We also investigated TMZ and found that the IC₅₀ values of TMZ in the three glioma cell lines were 887.9 µM, 1174 µM, and 2519 µM, respectively (Fig. 1E-G). The results indicated that lower Sira concentrations achieve equivalent cytotoxicity compared to TMZ. The viability of GBM cells decreased in a dose and time-dependent manner after treatment with Sira, as shown in Fig. 1H and I. Additionally, the morphological features of cell death, such as cytoplasmic contraction, cell debris, and floating cells, were observed in the groups treated with Sira (supplementary Fig. S1A). Furthermore, we observed a lower sensitivity of Sira on normal glial cells, with cell viability decreasing to approximately 71.7% after administration of 100 µM-Sira (Fig. 1J). Expanded toxicological profiling demonstrated IC₅₀ values of 12.55 µM in HT22 hippocampal neurons, 22.79 µM in SH-SY5Y neuroblastoma cells, and 19.05 µM in HEK293T epithelial cells (Supplementary Fig. S2)—representing a 2.6- to 3.2-fold therapeutic window compared to its IC₅₀ in GBM cell lines (7.24–9.65 µM). The impact of Sira on GBM cell migration was investigated using a wound-healing assay, which showed a significant decrease in cell migration in cells treated with Sira (Fig. 1K-N). The clonogenic ability of GBM cells also decreased significantly after treatment with Sira (Fig. 1O-R). Additionally, we found that the apoptosis inhibitor Z-VAD-FMK could not reverse the decrease in cell viability caused by Sira (supplementary Fig. S1B-D), indicating that Sira did not induce cell death by activating apoptotic cascades. Overall, these results demonstrate that Sira effectively inhibits the viability, migration, and proliferation of GBM cells.

Fig. 1.

Siramesine (Sira) significantly reduced cell viability, migration, and colony formation of glioblastoma (GBM) cells. (A) The molecular structure of Sira HCl. (B-D) The IC₅₀ value of Sira and (E-G) temozolomide (TMZ) in GBM cells after 48 h treatment. (H) GBM cell viability was assessed after treatment with varying concentrations of Sira for 3 h. (I) The viability of U87-MG cells was evaluated at 24, 48, and 72 h post-treatment with varying concentrations of Sira. (J) The viability of HA1800 was treated with varying concentrations of Sira for 48 h. (K-N) GBM cell migration was assessed using a wound-healing assay. (O-R) Colony formation assay was performed on GBM cells treated with different drug concentrations for 10 days. Data represents the mean ± standard deviation of at least three independent experiments *P < 0.05, **P < 0.01, ***P < 0.001

Sira induces protective autophagy in GBM cells

Sira has been shown to stimulate autophagy while blocking the degradation of substrates [14, 18, 24]. It was uncertain whether Sira would have the same effect in GBM cells. Analysis using WB revealed that treatment with higher concentrations of Sira (50 µM) for 3 h increased the synthesis and processing of LC3B-II and Beclin1 in U87-MG cells (Fig. 2A-C) and the expression of LC3B-II, Beclin1, and ATG5 in U251-MG cells (Fig. 2F-I). LC3B-II (microtubule-associated protein 1 light chain 3 beta-II) is a lipidated form of LC3B that is specifically associated with autophagosome membranes and serves as a marker for autophagosome formation [25]. Beclin1, a core component of the class III phosphatidylinositol 3-kinase (PI3K) complex, plays a critical role in the initiation of autophagy by promoting vesicle nucleation [26]. ATG5 (autophagy-related protein 5) is essential for autophagosome elongation and maturation, as it forms a complex with ATG12 and ATG16L1 to facilitate the conjugation of LC3B to phosphatidylethanolamine, a step required for autophagosome membrane expansion [27]. The upregulation of these proteins by Sira suggests its role in promoting autophagic activity. However, the autophagy substrate P62 levels decreased significantly in both GBM cells (Fig. 2E and J). P62, also known as SQSTM1, is a selective autophagy substrate that is degraded during autophagic flux. A significant reduction in P62 levels typically indicates enhanced autophagic degradation, as P62 is incorporated into autophagosomes and subsequently degraded in autolysosomes [28, 29]. Therefore, the observed decrease in P62 supports the conclusion that Sira activates autophagy and promotes autophagic degradation in GBM cells. CQ is a late autophagy inhibitor that blocks autophagic flux by inhibiting autophagosome-lysosome fusion and inducing the accumulation of LC3B-II [14]. Simultaneous treatment of GBM cells with Sira and CQ significantly increased the accumulation of LC3B-II. The increase in LC3B-II levels in the presence of CQ suggests enhanced autophagic flux. CQ inhibits lysosomal degradation, leading to the accumulation of LC3B-II when autophagic flux is active. The higher LC3B-II levels in Sira-treated cells compared to controls indicate that Sira increases autophagosome formation rather than merely blocking autophagosome-lysosome fusion. This interpretation is further supported by the concomitant decrease in P62 levels, which reflects increased autophagic degradation [30], suggesting that Sira treatment increases autophagic flux (Fig. 2K-M). Similar results were also obtained in GBM cells treated with lower concentrations of Sira (5 µM for U87-MG cells and 10 µM for T98G cells) (supplementary Fig. S3). We also investigated whether Sira induces protective autophagy or autophagic cell death. The CCK-8 assay showed that both the early autophagy inhibitor WOR and the late autophagy inhibitor CQ enhanced the cell death induced by Sira treatment (Fig. 2N-P). The fact that inhibition of autophagy enhances Sira-induced cell death strongly suggests that the autophagy triggered by Sira is protective rather than destructive. If autophagy were contributing to cell death (i.e., autophagic cell death), inhibiting it would likely reduce cell death. However, the opposite was observed: inhibiting autophagy increased cell death, indicating that autophagy is helping the cells survive the cytotoxic effects of Sira. These findings align with the well-established paradigm of autophagy as a stress-adaptive mechanism and are consistent with recent studies demonstrating that cytoprotective autophagy mitigates lysosomal-targeted therapies in cancer. For instance, Chen et al. [31] showed that autophagy sustains cancer cell survival by clearing lysosomal damage-induced organellar debris, a mechanism that may transiently buffer acute stress. Our conclusions further corroborate foundational work by Ostenfeld et al. [14], who identified Sira as a lysosomotropic agent inducing cytoprotective autophagosome accumulation. Together, these results position Sira-induced autophagy as a transient resilience mechanism in GBM, highlighting the therapeutic potential of combining autophagy inhibitors with Sira to overcome adaptive resistance.

Fig. 2.

Siramesine (Sira) induced cytoprotective autophagy in glioblastoma (GBM) cells. (A-E) Autophagic proteins in U87-MG cells were analyzed by western blot (WB) with the indicated antibodies. (F-J) Autophagic proteins in U251-MG cells were analyzed by WB with the indicated antibodies. (K-M) Autophagic proteins LC3B and P62 in U87-MG cells were analyzed by WB after treatment with Sira and chloroquine (CQ). (N-P) Effect of autophagy inhibitors wortmannin (WOR) and CQ on Sira-induced decrease in viability of GBM cells. The data presented is the mean ± standard deviation of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001, ns = not significant

Sira interacts with STAT3 in GBM cells

In our investigation, we delved into the molecular mechanism of Sira in GBM cells. First, we treated U87-MG cells with DMSO or Sira (10 µM) and conducted mRNA sequencing. The analysis revealed that the differentially expressed genes were enriched in cytokine receptor binding and cytokine-cytokine receptor interaction, as evidenced by both Gene Ontology (GO) functional analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis (Fig. 3A, B). Furthermore, KEGG analysis highlighted the Janus kinase (JAK)-STAT signaling pathway as downstream of cytokine and cytokine receptor interaction (Fig. 3C). Subsequently, we validated the bioinformatic data with qRT-PCR, which confirmed the upregulation of several cytokines (IL-1β, IL-6, TNF-α, IL-8, IL-10, TNSF14, CXCL2, and CXCL5) following Sira treatment in U87-MG cells (Fig. 3D). Given the known pivotal role of STAT3 in tumorigenesis [30, 32], our bioinformatic analysis also demonstrated higher STAT3 mRNA expression in GBM tissues (WHO grade IV gliomas, G4) compared to grade III gliomas (WHO grade III gliomas, G3) and grade II gliomas (WHO grade II gliomas, G2), as well as non-tumorous tissue. Furthermore, we found a negative correlation between STAT3 mRNA expression and OS in glioma patients (supplementary Fig. S4). We then explored the potential interaction between Sira and STAT3 through a molecular docking assay, which indicated binding energy of -8.2 kcal/mol for Sira with STAT3 (Fig. 3E). This was supported by the results of a pull-down assay, where STAT3 was one of the proteins pulled down by the Sira-Sepharose 4B probe (Fig. 3F, G). To quantitatively validate this interaction, SPR assays were conducted, revealing a dissociation constant (K_D) of 8.712*10^− 4 M for Sira-STAT3 binding (Fig. 3H). The measurable K_D underscores direct and physiologically relevant interaction, aligning with prior molecular docking and pull-down results. Collectively, these data implicate STAT3 as a critical molecular target through which Sira modulates GBM cell activity.

Fig. 3.

Siramesine (Sira) combined with STAT3 in glioblastoma (GBM) cells

Sira suppresses the constitutive activation of STAT3 in GBM cells

The activation of STAT3 relies on tyrosine phosphorylation at position 705(Y705), which leads to dimerization, nuclear translocation, DNA binding, and gene transcription [33]. In our study, we examined whether Sira inhibits the phosphorylation of STAT3 at the 705 position. We discovered that Sira effectively inhibited the phosphorylation of STAT3 at the Tyr705 site in U87-MG and T98G cells in a time- and dose-dependent manner (Fig. 4A-D). There was no significant change in constitutive STAT3, but in U87-MG cells, STAT3 significantly decreased after 48 h of treatment with Sira (10 µM) (Fig. 4A-D). The JAK2/STAT3 pathway plays a significant role in developing malignant tumors by affecting the activation state of downstream effector molecules [34]. Sira significantly suppressed constitutive JAK2 expression at 12 h, 24 h, and 48 h after treatment (Fig. 4A-D). However, the level of phosphorylation of JAK2 (p-JAK2)/JAK2 did not change significantly (Fig. 4A-D). PIAS3, SOCS3, SHP1, and SHP2 are negative regulators of STAT3 signaling that can effectively slow down tumor progression [35, 36]. We investigated these regulators’ expression to further understand the molecular mechanism for inhibiting STAT3 signaling in GBM cells treated with Sira. The qRT-PCR results demonstrated that SHP1, SHP2, and PIAS3 were upregulated by Sira (10 µM) in U87-MG and U251-MG. In T98G cells, SHP2 was upregulated, while SHP1 and PIAS3 did not show noticeable changes. The SOCS3 gene was only upregulated in U87-MG cells (Fig. 4E). Translocation of p-STAT3(Y705) to the nucleus is crucial for the activation of STAT3 signaling [37]. Therefore, we further investigated the subcellular localization and relative intensity of p-STAT3(Y705) by immunofluorescence staining. The results revealed that both whole-cell and nuclear expression of p-STAT3(Y705) decreased significantly after treatment with Sira (Fig. 4F-K). To determine if the decrease in cell viability induced by Sira is related to the STAT3 signaling pathway, we used colivelin, a STAT3 activator, and AG490, a JAK2/STAT3 signaling inhibitor, in a combination treatment with/without Sira for 24 h in GBM cells. We found that colivelin (40 µM) reversed the decrease in cell viability induced by Sira (Fig. 4L). Conversely, AG490 (20 µM and 40 µM) further worsened the Sira-induced decrease in GBM cell viability (Fig. 4M). Notably, Sira exhibited superior potency to the STAT3 inhibitor C118-9, with IC₅₀ values of 8.875 µM (U87-MG), 9.654 µM (U251-MG), and 7.236 µM (T98G)—3- to 6-fold lower than C118-9’s IC₅₀ (23.64 µM, 47.06 µM, and 55.64 µM in T98G, respectively; supplementary Fig. S5). This enhanced efficacy implies multi-target mechanisms, potentially involving lysosomal membrane permeabilization and caspase-independent death pathways, beyond STAT3 inhibition. Supporting this hypothesis, Sira upregulated STAT1 and γ-H2AX (DNA damage marker) in T98G/U87-MG cells, while modulating AKT signaling in U87-MG/U251-MG lines (supplementary Fig. S6). In summary, our findings establish that Sira directly targets STAT3 signaling through: inhibition of Tyr705 phosphorylation, impairment of nuclear translocation and upregulation of negative regulators (SHP1/2, PIAS3, SOCS3). These mechanisms collectively contribute to its potent anti-GBM activity, while additional pathways involving STAT1, DNA damage response (γ-H2AX), and AKT modulation likely enhance its therapeutic efficacy through multi-targeted actions.

Fig. 4.

Siramesine (Sira) inhibited the STAT3 signaling pathway in glioblastoma (GBM) cells. (A, B) Proteins of the JAK2-STAT3 signaling pathway in GBM cells were evaluated with the indicated concentrations of Sira treatment for 0 h, 12 h, 24 h, and 48 h. (C, D) JAK2-STAT3 signaling pathway proteins in GBM cells were evaluated with the indicated concentrations of Sira treatment for 48 h. (E) QRT-PCR evaluation of PIAS3, SOCS3, SHP1, and SHP2 levels in GBM cells after 48 h of Sira treatment. (F, I) Expression and subcellular localization of p-STAT3(Y705) in GBM cells were examined at the indicated concentration of Sira treatment for 48 h. (G, J) Statistical analysis of total cellular expression of p-STAT3(Y705) in GBM cells. (H, K) Statistical analysis of nuclear expression of p-STAT3(Y705) in GBM cells. (L, M) The indicated concentrations of the STAT3 activator colivelin or the JAK2-STAT3 pathway inhibitor AG490 were used to assess the effect of the STAT3 pathway on GBM cell viability. Data presented as mean ± standard deviation of at least three independent experiments. Scale bars: 20 µm in Fig. 4F, and 100 µm in Fig. 4I. *P < 0.05, ***P < 0.001

Sira and TMZ synergistically inhibit the viability of GBM cells

The effectiveness of Sira on GBM cells with minimal harm to normal astrocytes led us to study the impact of Sira when combined with TMZ, the primary chemotherapy drug for GBM treatment. Initially, GBM cells were treated with increasing concentrations of TMZ for 48 h, both with and without a lower dose of Sira (5 µM). We determined the optimal drug concentrations in a pilot experiment. For U87-MG cells, we selected TMZ concentrations of 50 µM, 100 µM, 250 µM, and 500 µM. For U251-MG, we used TMZ concentrations of 500 µM, 750 µM, 1000 µM, and 1250 µM. For T98G, the TMZ concentrations were 50 µM, 100 µM, 250 µM, and 500 µM, respectively. Our results revealed that simultaneous treatment with Sira and TMZ significantly reduced cell viability at each concentration of TMZ, with the combined treatment showing significantly lower viability compared to TMZ alone. Additionally, the inhibitory effect of the drug combination was also higher than that of Sira alone when TMZ was used at 200 µM and 500 µM in U87-MG cells, and at 1250 µM in U251-MG, as well as at each concentration in T98G cells (Fig. 5A-C). We also assessed the cytotoxicity of the drug combination. Compared with the glioma cells, there was a relatively lower toxic effect on HA1800 cells (Fig. 5D). The IC₅₀ value of TMZ in combination with Sira was 322.9 µM, 264 µM, and 510 µM in U87-MG, U251-MG, and T98G cells, respectively (Fig. 5E-G). Subsequently, we analyzed the Combination Index (CI) of Sira and TMZ in GBM cells. The CI values less than 1 indicated a synergistic effect, a CI equal to 1 indicated an additive effect, and a CI greater than 1 indicated antagonism between the two drugs [38, 39]. Our results showed that the combination of TMZ and Sira resulted in a CI of 0.92, 0.73, and 0.88 for U87-MG, U251-MG, and T98G cells, respectively, all of which were less than 1 (Fig. 5H-J), indicating a synergistic effect between Sira and TMZ. GBM cells treated simultaneously with TMZ and Sira exhibited typical morphological features of cell death, such as cytoplasmic contraction, cell debris, and floating cells, when observed under a microscope at 48 h (supplementary Fig. S7). In conclusion, these findings suggest that Sira and TMZ have a synergistic effect in inhibiting the viability of GBM cells.

Fig. 5.

Combining Siramesine (Sira) with temozolomide (TMZ) increased the chemosensitivity of glioblastoma (GBM) cells. (A-C) The viability of GBM cells was evaluated with the combination of Sira (5 µM) and the indicated concentrations of TMZ treatment for 48 h. (D) The viability of HA1800 was evaluated with the combination of Sira (5 µM) and the indicated concentrations of TMZ treatment for 48 h. (E-G) The IC₅₀ value of TMZ after combined use of Sira in GBM cells. (H-J) The Chou-Talalay index (CI) was used to evaluate the effect of TMZ in combination with Sira in GBM cells. CI < 1, = 1, or > 1 indicated synergistic, additive, or antagonistic effects. Data presented as mean ± standard deviation of at least three independent experiments. **P < 0.01, ***P < 0.001

TMZ and Sira synergistically inhibit GBM cell migration and proliferation

We investigated the antitumor effect of the combination therapy of TMZ and Sira by performing in vitro wound healing assays to determine the ability of the cells to migrate. A combination of Sira (5 µM) and TMZ (500 µM) reduced the migration rates of U87-MG and U251-MG, both of which were higher than that of TMZ alone (Fig. 6A-D). However, the inhibitory effect was insignificant in T98G cells (Fig. 6E, F). We then performed a colony formation assay to validate the antitumor effect further. The results showed that the number of colonies decreased significantly in the combination group compared to the single-agent group (Fig. 6G-L).

Fig. 6.

The combination of Siramesine (Sira) with temozolomide (TMZ) inhibited the migration and colony formation of glioblastoma (GBM) cells. (A-F) The migration of GBM cells was detected using a wound-healing assay. The edges of the scratches were marked with white lines. (G-L) Colony formation assay was performed on GBM cells treated with the indicated drug concentrations for 10 days. Data are presented as mean ± standard deviation of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

TMZ and Sira synergistically inhibit the activation of STAT3(Y705) and reduce MGMT levels in GBM cells

Additionally, we examined the molecular mechanism responsible for the increased cell death observed in the group receiving the drug combination. The findings revealed that simultaneous administration of TMZ (500 µM) and Sira (5 µM) led to a significant decrease in the expression of p-STAT3(Y705) in U87 MG cells and similarly reduced the expression in T98G cells compared to when each drug was used alone (Fig. 7A, E, F, J). We also observed a significant decrease in the total amount of STAT3 in T98G cells in the drug combination group (Fig. 7F, I). P-JAK2 was weakly expressed in U87-MG cells under all treatment conditions (Fig. 7A, C). In T98G cells, p-JAK2 was reduced by simultaneous treatment with TMZ and Sira, but not significantly (Fig. 7F, H). These results indicated that the combined therapy of Sira and TMZ inhibited the activation of STAT3 in GBM cells. MGMT is associated with TMZ resistance in GBM [40–42]. Patients with high MGMT expression are not effectively treated and have a poor prognosis [41]. Therefore, we analyzed MGMT expression to assess its significance for the chemosensitivity of GBM cells to TMZ after combination drug therapy. As shown in Fig. 7K, T98G cells expressed a higher level of MGMT protein, while it was barely detected in U87-MG cells, which may explain the differential chemosensitivity of GBM cells to TMZ: T98G was less sensitive to TMZ than U87-MG (Fig. 1E, G). In T98G cells, both Sira and TMZ decreased MGMT expression. At the same time, the combination therapy was even more effective in reducing MGMT levels, possibly explaining why TMZ in combination with Sira reversed the chemoresistance of T98G to TMZ (Fig. 7K, M). It has been reported that the expression of MGMT positively correlates with that of p-STAT3 in GBM [43, 44]. Therefore, we aimed to examine the correlation between STAT3 activation status and MGMT levels. T98G cells were treated with Sira (5 µM) in combination with different concentrations of colivelin, a selective STAT3 signaling activator [44], and AG490, a selective inhibitor of STAT3 phosphorylation [45]. WB analysis revealed that p-STAT3(Y705) levels in T98G cells significantly increased with 0.5 µM, 1 µM, and 5 µM colivelin and decreased with 20 µM and 40 µM AG490 in a dose-dependent manner. Meanwhile, MGMT levels increased with colivelin and decreased with AG490 in a dose-dependent manner (Fig. 7N, O). Correlation analysis showed that p-STAT3(Y705) levels were positively correlated with those of MGMT (Fig. 7P). In summary, these results suggest that Sira, in combination with TMZ, improves chemosensitivity by inhibiting STAT3 activation and MGMT expression in chemoresistant GBM cells.

Fig. 7.

The combination of Siramesine (Sira) with temozolomide (TMZ) inhibited the STAT3-O⁶-methylguanine-DNA methyltransferase (MGMT) signaling pathway in glioblastoma cells. (A-J) Proteins of the JAK2-STAT3 signaling pathway in GBM cells were examined upon treatment with Sira (5 µM) in combination with TMZ (500 µM) for 48 h. (K-M) MGMT expression in GBM cells was examined upon 48-hour treatment with Sira (5 µM) in combination with TMZ (500 µM). (N, O) The indicated concentrations of colivelin or AG490 were used to evaluate the regulation of the STAT3 signaling pathway on the expression of MGMT in GBM cells. (P) Correlation analysis between the level of p-STAT3(Y705)/STAT3 and the expression of MGMT. Data presented as mean ± standard deviation of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

Sira inhibits the JAK2-STAT3-MGMT signaling pathway in a CDX tumor model

Since Sira inhibited the proliferation of GBM cells in vitro and increased their chemosensitivity, we investigated whether these results could be translated in vivo. We established a human glioma xenograft model in nude mice with U87-MG cells. Once the tumors reached a size of 500 mm³, the mice were randomly assigned to one of the following four treatment groups and treated for two weeks: Diluent control (vehicle); TMZ (25 mg/kg daily i.p.); Sira (1 mg/kg every other day i.p.); or TMZ + Sira. The experimental protocol is shown in Fig. 8A. The mice’s body weight and tumor volume were recorded every other day. The final body weight of the Diluent control and the Sira groups increased slightly compared to the initial stage. However, it was insignificant (Fig. 8B). We speculate that the discrepancy may be related to tumor growth in these two groups. No significant difference in tumor volume was found between the Diluent control group and the group treated with Sira alone. In contrast, the tumors treated with TMZ or a combination of TMZ and Sira were significantly smaller than the control tumors (Fig. 8C). We then examined the isolated tumors. The tumor size in the group treated with Sira alone was similar to that in the control group treated with Diluent. At the same time, it decreased significantly in the group treated with TMZ alone and in the group treated with the drug combination (Fig. 8D, E). Consistent with this, tumor weight increased in the Sira group but not significantly more than in the Diluent control group (Fig. 8F). The tumor weight decreased significantly in the TMZ group or the drug combination group (Fig. 8D-F). These results suggest that Sira at a lower dose (1 mg/kg) did not affect glioma growth in vivo. On the contrary, TMZ has a significant inhibitory effect on tumor growth in vivo (Fig. 8D-F), suggesting that U87-MG glioma was sensitive to TMZ treatment in vivo. Furthermore, we investigated the JAK2-STAT3-MGMT signaling pathway and glioma stem cell (GSC) properties in CDX gliomas using WB. Surprisingly, Sira alone could effectively downregulate the expression of JAK2/p-JAK2, STAT3/p-STAT3(Y705), MGMT, and the GSC markers SOX2 and CD133 (Fig. 8G-O). Similarly, the expression of JAK2, STAT3, MGMT, SOX2, and CD133 decreased significantly after TMZ treatment (Fig. 8G-O). In contrast, only JAK2, MGMT, and SOX2 decreased significantly in the drug combination group (Fig. 8G-O). Finally, we prepared frozen sections of liver, lung, spleen, and kidney of the four treatment groups to investigate the drugs’ therapeutic efficacy and possible toxic side effects. H&E examination showed no noticeable histopathologic changes in the above organs in each group (supplementary Fig. S8). Overall, although Sira at a lower dose (1 mg/kg every other day i.p.) did not affect glioma growth, it was able to suppress the JAK2-STAT3-MGMT signaling pathway and the properties of GSC.

Fig. 8.

Siramesine (Sira) inhibited the JAK2-STAT3-MGMT signaling pathway and stem cell properties of glioma in vivo. (A) The schematic of the experiment using U87-MG cells as xenograft. U87-MG cells were injected subcutaneously into the right axilla and treated with systemic treatments after tumor formation. Twenty-one nude mice were randomly divided into four groups, including Diluent control (n = 6), Siramesine (Sira) alone (n = 5), temozolomide (TMZ) alone (n = 5), and Sira + TMZ (n = 5). Drug treatment was administered over two cycles. (B) Body weight data of the mice during drug treatment. (C) Recordings of tumor volume of mice during drug treatment. (D) Images of xenografts. The subcutaneous tumors were circled with red lines. (E) The isolated tumors of four groups were taken two days after the last treatment. (F) Weight of isolated tumors. (G-L) Protein expressions of the JAK2-STAT3-MGMT pathway in the isolated tumors of the four groups. (M-O) Expression of glioma stem cell (GSC) proteins SOX2 and CD133 in the isolated tumors of four groups. Data presented as mean ± standard deviation of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001

Discussion

TMZ is a cornerstone in GBM treatment, but resistance often leads to treatment failure, underscoring the need for strategies to enhance TMZ sensitivity [46, 47]. Drug repurposing, particularly with FDA-approved drugs like Sira (Lu 28–179), offers a promising avenue due to their established safety profiles [48–50]. Sira, originally used for central nervous system disorders, has shown anticancer potential, particularly in crossing the blood-brain barrier (BBB) [50, 51]. This study explored Sira’s efficacy and mechanisms, both alone and in combination with TMZ, in GBM treatment. Key findings include: (1) Sira inhibits GBM cell viability, migration, and clone formation by downregulating STAT3 signaling; (2) Sira synergistically enhances TMZ’s antitumor effects in vitro; (3) The combination suppresses MGMT expression via p-STAT3 downregulation; and (4) Sira inhibits the JAK2-STAT3-MGMT pathway and reduces GSC properties in vivo.

Sira exerts antitumor activity through multimodal STAT3 suppression

Our findings establish STAT3 as a primary molecular target through which Sira exerts its antitumor effects in GBM. While constitutive STAT3 activation drives tumor progression and therapeutic resistance in multiple malignancies [52, 53], Sira employs a dual mechanism to disrupt this oncogenic signaling: (1) direct binding to STAT3 (confirmed by molecular docking, pull-down and SPR assays; Fig. 3E-H) leading to dose- and time-dependent inhibition of Tyr705 phosphorylation (Fig. 4A-D), and (2) upregulation of endogenous suppressors PIAS3, SHP1, and SHP2 (Fig. 4E) known to deactivate STAT3 independently of upstream regulators [35, 36, 54–56]. Notably, while Sira reduced JAK2 expression levels, it did not alter JAK2 phosphorylation status, suggesting its STAT3 inhibition operates through mechanisms extending beyond canonical JAK2 kinase activity. This multimodal STAT3 targeting underlies Sira’s superior efficacy compared to the classical STAT3 inhibitor C188-9 across glioma models (Supplementary Fig. S5), potentially circumventing compensatory resistance mechanisms common to single-pathway inhibitors.

The therapeutic impact is amplified through lysosomal membrane permeabilization and caspase-independent cell death pathways [13–15]. This creates a synergistic “double-hit” effect: direct STAT3 inhibition disrupts survival signaling while lysosomal leakage induces oxidative stress (via iron release) and impairs DNA repair capacity (evidenced by γ-H2AX elevation; Supplementary Fig. S6). Such organellar stress may deplete nucleotides or redox regulators [57], exacerbating STAT3 pathway vulnerabilities. Crucially, preserved cytotoxicity in apoptosis-resistant models highlights the clinical relevance of these non-canonical death mechanisms in treating therapy-evading GBM.

Synergistic effects of Sira and TMZ

The combination of Sira and TMZ demonstrates synergistic activity by concurrently targeting STAT3-mediated resistance pathways and DNA repair mechanisms. While TMZ monotherapy reduced MGMT expression (a STAT3-regulated resistance factor [43, 58, 59]) more effectively than Sira alone in T98G models (Figs. 7K and 8G), the combination achieved enhanced suppression. Notably, Sira potentiated TMZ efficacy even in MGMT-negative U87-MG cells, suggesting additional resistance-overcoming mechanisms such as: Lysosomal permeabilization impairing DNA repair through ATP/pH dysregulation; Oxidative stress amplification via iron-mediated Fenton reactions; STAT1 activation and PI3K-AKT pathway suppression (Supplementary Fig. S6), given the established STAT3-AKT cross-talk [60].

RNAseq analysis revealed significant PI3K-AKT pathway modulation by Sira (Fig. 3B), with validation experiments confirming p-AKT reduction (Supplementary Fig. S6). This multimodal pharmacology—simultaneously destabilizing survival signaling (STAT3/AKT), inducing oxidative damage, and compromising stress adaptation systems—positions Sira as a next-generation agent capable of preventing adaptive resistance. Future studies should incorporate biomarkers of lysosomal integrity [61], oxidative stress [62], and drug efflux [63] to fully characterize this therapeutic synergy.

Sira targets GSCs and the JAK2-STAT3-MGMT pathway

Although Sira exhibited limited tumor growth inhibition in vivo, it significantly reduced the expression of key stemness regulators, including JAK2, STAT3, and GSC markers, indicating its potential to compromise glioma stemness [20, 64, 65]. This finding is supported by previous studies demonstrating Sira’s efficacy in targeting CSCs across multiple cancer types [31, 64, 66]. However, our data suggest that higher therapeutic doses or combination strategies may be necessary to achieve durable tumor eradication. Notably, Sira’s dual role in autophagy modulation—both as an inducer and as a sensitizer to autophagy inhibitors—reveals a unique therapeutic window for GBM treatment [67–70].

Future directions

The anti-GBM effects of Sira could likely be further enhanced through modifications in dosage, formulation, or combination with other targeted therapies. Future studies should explore higher doses of Sira to determine the optimal therapeutic window. Dose-escalation studies in preclinical models (e.g., CDX and patient-derived xenograft [PDX]) could help identify the minimum effective dose and maximum tolerated dose [71]. Novel drug delivery systems, such as nanoparticles, liposomes, or biodegradable polymers, could improve Sira’s pharmacokinetics and biodistribution. Conjugating Sira with BBB-penetrating peptides or using focused ultrasound to temporarily disrupt the BBB could improve its delivery to brain tumors [72]. Combining Sira with other therapeutic agents could target multiple pathways simultaneously, overcoming resistance and enhancing efficacy. Potential combinations include: PARP inhibitors [73], immune checkpoint inhibitors [74], epigenetic modulators [75], combining Sira with agents that specifically target GSCs, such as Notch, Wnt, or Hedgehog pathway inhibitors, could further enhance its efficacy. Conduct pharmacokinetics (PK) /pharmacodynamics (PD) studies to measure Sira’s distribution, metabolism, and tumor tissue concentrations in preclinical models.

Overall, while the current study provides evidence of Sira’s ability to modulate the STAT3-MGMT pathway and target GSCs in vivo, we acknowledge the limitations of the low dose and CDX model. We will address these issues in future work by investigating higher doses, improving formulations, quantifying drug delivery, utilizing PDX models, and combining with other targeted therapies to strengthen the clinical relevance of our findings. These strategies would address the limitations of Sira’s bioavailability, tumor penetration, and resistance mechanisms, potentially leading to more effective and durable responses in GBM patients. Future studies should focus on these approaches to fully exploit Sira’s therapeutic potential.

While JAK2 is a classical upstream kinase for STAT3, our findings imply that Sira’s primary mechanism involves direct STAT3 binding and enhancement of endogenous STAT3 suppressors, rather than relying solely on JAK2 pathway modulation. The reduction in JAK2 expression could reflect feedback regulation or off-target effects, but this requires further investigation. Kinase activity assays and genetic manipulation (e.g., JAK2 knockdown) will elucidate Sira’s indirect effects on JAK2 signaling.

Systematic toxicological evaluation must extend beyond immortalized lines to primary human neurons, iPSC-derived cultures, and organotypic brain slices, distinguishing tumor-selective efficacy from lineage-specific toxicity. Expansion to diverse somatic cell types (endothelial cells, fibroblasts) will further delineate off-target liabilities. Current in vitro data reveal a 2–3-fold therapeutic index (GBM IC₅₀: 7.24–9.65 µM vs. neuronal/renal IC₅₀: 12.55–22.79 µM), supporting preferential cytotoxicity but necessitating in vivo validation to mitigate neurotoxic/nephrotoxic risks.

Conclusion

Our findings establish Sira as a multimodal therapeutic agent in GBM, exerting antitumor effects through direct STAT3 pathway inhibition and lysosomal destabilization. Sira suppresses GBM cell viability, migration, and clonogenicity by targeting STAT3 phosphorylation (Tyr705), upregulating endogenous suppressors (PIAS3, SHP1/2), and impairing STAT3 nuclear translocation. Critically, Sira synergizes with TMZ to overcome chemoresistance via dual suppression of STAT3 activation and MGMT expression, particularly in TMZ-refractory models. While in vivo efficacy was dose-limited, Sira’s ability to attenuate STAT3-MGMT signaling and reduce GSC markers (SOX2, CD133) underscores its potential to disrupt tumor recurrence drivers. Future studies should prioritize pharmacological optimization—including dose escalation, BBB-penetrating formulations, and rational combinations with DNA repair inhibitors or immunotherapies—to translate Sira’s mechanistic promise into clinically actionable regimens for GBM.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

CCK-8

Cell counting kit-8

CDX

Cell-derived xenograft

CGGA

Chinese Glioma Genome Atlas

CI

Combination index

CQ

Chloroquine

DMSO

Dimethyl sulfoxide

GBM

Glioblastoma

GO

Gene Ontology

GSCs

Glioma stem cells

HE

Hematoxylin and eosin

HRP

Horseradish Peroxidase

IC₅₀

Median inhibitory concentration values

JAK

Janus kinase

KEGG

Kyoto Encyclopedia of Genes and Genomes

LGG

Lower Grade Glioma

MGMT

O⁶-methylguanine DNA methyltransferase

MMR

Mismatch repair

OS

Overall survival

PB

Phosphate buffer

PBS

Phosphate buffered saline

PDX

Patient-derived xenograft

PEG300

Polyethylene glycol 300

PMSF

Phenylmethanesulfonyl fluoride

p-STAT3(Y705)

Phosphorylated STAT3 at tyrosine 705

PVDF

Polyvinylidene fluoride

QPCR

Quantitative Real-Time Reverse Transcription Polymerase Chain Reaction

RT

Room temperature

RT-PCR

Reverse Transcription-Polymerase Chain Reaction

SH2

Src homology 2

STAT3

Signal transducer and activator of transcription 3

TCGA

The Cancer Genome Atlas

TMZ

Temozolomide

WOR

Wortmannin

Author contributions

Anhui Yao: Funding acquisition, Liyun Jia and Anhui Yao: Conceptualization, Xiaohang Cui: Methodology and Resources, Hengzeng Li and Yahui Wu: Software, Jinquan Lv, Chi Zhang, and Yue Chen: Validation, Liyun Jia: Writing- Original draft preparation. Liyun Jia: Visualization, Investigation and Supervision, Liyun Jia and Anhui Yao: Writing- Reviewing and Editing.

Funding

The present study was supported by the opening project of State Key Laboratory of Explosion Science and Technology (Beijing Institute of Technology) (grant no. KFJJ23-09 M), the China Postdoctoral Science Foundation (grant no. 2019M653946), the research project of 988th Hospital of Joint Logistic Support Force of PLA (grant no. YNZX2024005), and the joint construction project of Henan Province (grant no. LHGJ20230703). The funding sources had no involvement in study design, in the collection, analysis, and interpretation of data, in the report’s writing, and in the decision to submit the article for publication.

Data availability

All data generated or analyzed during this study are included in this published article and its supplementary information files.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.