Introduction

Glioblastoma (GBM) is the most aggressive and prevalent form of brain cancer originating from astrocytes. It is categorized as a World Health Organization grade IV tumor (Louis et al. 2021; Bleeker et al. 2012; Tan et al. 2020). GBM has a formidable invasion capacity, causes structural and functional damage to the brain (So et al. 2021), and has a poor prognosis (Tran and Rosenthal 2010). Even after surgical removal, temozolomide chemotherapy, and radiation treatment, the average survival time remains at 15 months (Grochans et al. 2022). In addition, even if surgical removal is performed, recurrence is inevitable due to pre-existing infiltration in other areas (Kang et al. 2010). Numerous factors contribute to GBM invasion, including cellular interactions with the matrix, intercellular junctions, and molecular signaling pathways (Stetler-Stevenson et al. 1993; Martin et al. 2013). Among these factors, studies have highlighted the significant roles of the extracellular matrix (ECM), calcium (Ca2+), and glutamate signaling as the primary mechanisms regulating GBM invasion (So et al. 2021).

GBM upregulates the expression of various calcium signaling molecules, including G protein-coupled receptors (GPCR) and receptor tyrosine kinase (RTK) (So et al. 2021). Additionally, calcium moves from the extracellular space to the cytosol of GBM via store-operated channels (SOCs), Ca2+-permeable transient receptor potential (TRP) channels, voltage-gated calcium channels (VGCCs), and P2X7 receptors, leading to glioma invasion and migration (So et al. 2021). Calcium also enters into the GBM through the glutamate receptor α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) and N-methyl-d-aspartate (NMDA) receptors (Radin and Tsirka 2020; Corsi et al. 2019). These two receptors are activated in the GBM by neuronal and glial paracrine glutamate (Lyons et al. 2007). Glutamate, the most important excitatory neurotransmitter involved in neuronal activity, promotes GBM migration, invasion, and proliferation (So et al. 2021). GBM releases glutamate instead of taking up it, thereby inducing the excitotoxic death of surrounding neurons and creating space for invasion (Ye and Sontheimer 1999; Chung et al. 2005; Sontheimer 2008; Noch and Khalili 2009). Additionally, GBM uses glutamate in an autocrine manner through the Xc-system, which is highly expressed in GBM (Lyons et al. 2007; Groot and Sontheimer 2011). Indeed, the release of glutamate from the GBM can trigger the activation of Ca2+-permeable receptors such as AMPA or NMDA receptors. This activation leads to Ca2+ oscillations within the GBM cells or in neighboring GBM cells, subsequently facilitating the enhanced mobility of these cells. (So et al. 2021; Lyons et al. 2007).

System Xc− (or xCT), encoded by the SLC7A11 gene, serves as a cystine/glutamate transporter (Sato et al. 1999; Wang et al. 2003). The extracellular release of glutamate through the system Xc− transporter leads to its binding to glutamate receptors, subsequently eliciting a Ca2+ response within the cell. System Xc− is a protein composed of two chains, a specific light chain (xCT, SLC7A11) and a heavy chain, 4F2 (SLC3A2), connected by a disulfide bridge (Lewerenz et al. 2012). Sulfasalazine, a Food and Drug Administration-approved inhibitor of system Xc−, is currently under active investigation for its potential application in glioma treatment (Bridges et al. 2012). A previous study showed that glioma invasion was reduced in the presence of sulfasalazine (Lyons et al. 2007). Furthermore, sulfasalazine serves as an inhibitor of sepiapterin reductase and displays antibacterial and anti-inflammatory effects (Costigan et al. 2012; Goldman and Peppercorn 1975). This multifaceted nature has led to its application in the treatment of inflammatory bowel disease (Nielsen 1982; Das 1989) including ulcerative colitis and Chron’s disease (Sutherland et al. 1993; Klotz et al. 1980). Although the precise mechanism of action remains unclear, the reported side effects include depression, infertility, and immune thrombocytopenia (Cantarini et al. 2007). Therefore, in this study, we used a gene silencing method with shRNA rather than a pharmacological treatment.

Using the U-87 MG glioblastoma cell line, our initial experiments verified the reduction in glutamate release in cells treated with sulfasalazine. Subsequently, no significant variations were observed in the calcium response and proliferation compared with the control group. After generating and applying system Xc−-specific shRNAs, we observed a decrease in GBM cell invasion. Despite the observed reduction in invasion, no apparent differences were detected in terms of proliferation or calcium response. In summary, the targeted inhibition of system Xc− in GBM effectively diminishes glutamate release, subsequently contributing to a reduction in the invasion.

Materials and methods

Cell culture

U-87 MG cells were purchased from ATCC (Manassas, VA, USA, #HTB-14). The cells were maintained in Dulbecco’s Modified Eagle Medium (Corning, NY, USA, #10-013-CVRC) supplemented with 10% FBS (Gibco, NY, USA, #16000-044), and 1% penicillin–streptomycin solution (Cytiva, Seoul, Korea, #SV30010). All cells were grown at 37 °C in a humidified atmosphere containing 5% CO2.

Glutamate assay

Extracellular glutamate concentrations were measured using a glutamate assay kit (Abcam, Cambridge, England, #ab83389). U-87 MG cells (1.0 × 106) were seeded on culture plates in under specified conditions, and the assay was conducted after 24 h of incubation. The cells were treated with 250 µM of sulfasalazine (Tocris, Bristol, England, #4935) 6 h after seeding and incubated for another 24 h. The media containing glutamate was harvested 24 h later. The harvested medium was centrifuged at 13,000 rpm for 1 min to remove cell debris, and only the supernatant was used. The concentration was calculated according to the manufacturer’s instructions. The glutamate standard was prepared using 0.1 M glutamate and the assay buffer included in the kit. The standards were duplicated at concentrations of 0, 2, 4, 6, 8, and 10 nM. Both the standard and the sample were added to the well at 50 µL each, and the sample was diluted 1/10 and 1/20 in the assay buffer. 100 µL of reaction mix was added to every well, and the mix consisted of 90 µL of assay buffer, 8 µL of developer, and 2 µL of enzyme. After adding the mix, the sample was protected from light and reacted at 37 °C for 30 min. Absorbance was obtained by measuring OD450 nm. The concentration was calculated using the formula: \(Glutamate Concentration=\left(\frac{Sa}{Sv}\right)*D\). (Sa = Amount of sample (nmol) from the standard curve, Sv = Volume of sample (μL) added into the well, D = Sample dilution factor.)

Calcium imaging

For calcium imaging, all cells were cultured at a uniform density on poly-d-lysine-coated (Sigma-Aldrich, MO, USA, #P6403) 12-mm Ø cover glasses (#0111520; Paul Marienfeld, Lauda-Königshofen, Germany). Groups subjected to 250 µM sulfasalazine treatment (Tocris, Bristol, England, #4935) were incubated overnight. The external solution, designed for optimal imaging conditions, contained 150 mM NaCl (Sigma-Aldrich, MO, USA, #S9888), 10 mM HEPES (Carl Roth, Karlsruhe, Germany, #230-907-9), 3.2 mM KCl (Sigma-Aldrich, MO, USA, #P9541), 2 mM CaCl2 (Sigma-Aldrich, MO, USA, #C3881), 2 mM MgCl2 (Sigma-Aldrich, MO, USA, #M2670), 22 mM sucrose (Carl Roth, Karlsruhe, Germany, #200–334-9), and 5.6 mM glucose (Sigma-Aldrich, MO, USA, #G8270), and pH was adjusted to 7.4. All cells underwent a 40-min incubation in an external solution containing 5 μM Fura-2 AM (Invitrogen, MA, USA, #F1201) plus 1 μM pluronic acid (Sigma-Aldrich, MO, USA, #P2443). Imaging Workbench version 9.0.14.0 was used for the acquisition of intensity images and ratio calculations. The ligand TFLLR (synthesized by PEPTRON, Daejeon, Korea) was utilized in the study.

Proliferation assay

Cell proliferation was measured using a Quanti-Max WST-8 Cell Viability Assay Kit (BIOMAX, Guri, Korea, #QM1000). A seeding density of 1 × 105 U-87 MG cells was maintained on culture plates for each experimental condition. Following a 24 h incubation period, a proliferation assay was performed according to the manufacturer’s instructions. OD 450 values were determined for the naïve or scrambled shRNA groups by randomly pairing naïve and sulfasalazine groups or scrambled shRNA and system Xc-shRNA groups into one set.

shRNA cloning

The Xc-shRNA oligo system was synthesized by Cosmo Genetech (Daejeon, Korea). The cloning process following the shRNA sequences was used for cloning: (Shin et al. 2017). pSicoR-mCh-empty (#21907; Addgene), XhoI (#R0146S; BioLabs), HpaI (#R0105S; BioLabs), rCutSmart (BioLabs, MA, USA, #B6004S) and LaboPass Gel, and a PCR Clean-up kit (Cosmo Genetech, Daejeon, Korea, #CMA0112) were used for vector digestion and elution. NEB buffer (BioLabs, MA, USA, #B7002S), T4 polynucleotide kinase (Enzynomics, Daejeon, Korea, #M005S), T4 polynucleotide kinase buffer (Enzynomics, Daejeon, Korea, #B305), ATP (Sigma, MO, USA, #A26209), T4 DNA ligase (Enzynomics, Daejeon, Korea, #M001S), and T4 DNA ligase buffer (Enzynomics, Daejeon, Korea, #B304) were used for shRNA oligo annealing, phosphorylation, and construct ligation. The ligation product was transformed using DH5α competent cells (Enzynomics, Daejeon, Korea, #CP011). LB Broth (LPS solution, Daejeon, Korea, #LB-05), Bacto Agar (Becton Dickinson, NJ, USA, #214010), and ampicillin (Cayman Chemical, MI, USA, #14417) were used for E. coli growth, and the LaboPass plasmid mini kit (Cosmo Genetech, Daejeon, Korea, #CMP0112) was used to extract the plasmids from E. coli. The shRNA sequence was confirmed using Cosmo Genetech (Daejeon, Korea).

Candidate 1 sense: 5′-GCTGAATTGGGAACAACTATA-3′

Candidate 2 sense: 5′-GCAGTTGCTGGGCTGATTTAT-3′

Lentivirus production

The lentivirus was generated following the Addgene Lentiviral Guide. Transfection involved the introduction of 0.75 μg/mL psPAX2 (Addgene, MA, USA, #12260), 0.25 μg/mL pMD2.G (Addgene, MA, USA, #12259), and 1 μg/mL pSicoR-shRNA-mCh vector into HEK 293T cells using lipofectamine 3000 (Thermo Fisher Scientific, MA, USA, # L3000001). Opti-MEM (Gibco, NY, USA, #31985070) was used to ensure transfection efficiency, and HEK 293T cells were grown in an antibiotic-free medium. The packaging mix was injected into the HEK 293T cells at 80% density. Viruses were collected at 24-h intervals for 3 days. To enhance infection efficiency, polybrene (Merck Millipore, MA, USA, #TR-1003-G) was added at 1 μL/mL. The virus was aliquoted into 1 mL portions in cryotubes and stored at − 80 °C. Lentivirus cells were then introduced into U-87 MG at 20% density, and the medium was replaced 24 h later to promote cell growth and improve expression efficiency.

qRT-PCR

Total RNA was extracted 72 h after shRNA lentivirus infection using the RNeasy Mini Kit (QIAGEN, Hilden, Germany; #74106). RNA at 1,000 ng/μL per condition was synthesized into cDNA using SuperScript III First-Strand (Invitrogen, MA, USA, #18,080–051) and MiniAmp Thermal Cycler (Thermo Fisher Scientific, MA, USA, #A37834). qRT-PCR was performed using PowerSYBR Green PCR Master Mix (Applied Biosystems, MA, USA, #4367659) and a CFX96 Optics Module (Bio-Rad, CA, USA). Gene expression was normalized using the ΔΔCt method compared with the scrambled shRNA group. The primers used in the qRT reaction were as follows.

Human system Xc− forward: 5′-CAAGGTGCCACTGTTCATCC-3′

Human system Xc− reverse: 5′-CCACCTGGGTTTCTTGTCC-3′

Human GAPDH forward: 5′-GTCTCCTCTGACTTCAACAGCG-3′

Human GAPDH reverse: 5′-ACCACCCTGTTGCTGTAGCCAA-3′

Invasion assay

For each experimental condition, 1.0 × 105 U-87 MG cells were seeded on transwell plates (SPL, Pocheon, Korea, #35224) embedded with VitroGel hydrogel matrix (TheWell Bioscience, NJ, USA, #VHM01). The transwell chambers were filled with serum medium (2%), while 20% serum medium was placed outside the transwell chambers. Subsequently, the cells were incubated for 48 h. Treatment with 250 µM sulfasalazine (Tocris Bioscience, Bristol, England, #4935) was administered for 48 h. DPBS (WELGENE, Gyeongsan, Korea, #LB001-01), 3.7% formaldehyde (Sigma-Aldrich, MO, USA, #252549), methanol (Samchun Chemical Co., Ltd., Seoul, Korea, #M0588), and 4ʹ,6-diamidino-2-phenylindole, dihydrochloride (DAPI) (Thermo Fisher Scientific, MA, USA, #62248) were used. The invaded areas were measured using ImageJ software (Ver. 1.53 k). Three images were randomly captured per well, and the average of the three regions was compared with the average of naïve cell wells. The data were expressed as the percentage of invasion.

Statistical data analysis

Photograph area analysis was conducted using Image J (Ver. 1.53 k). All bar graphs, trace figures, and statistical analyses were generated using GraphPad Prism (Ver. 9.4.1). For comparisons involving two subjects, a two-tailed unpaired Student’s t-test was performed. For comparisons involving more than two subjects, a one-way analysis of variance was followed by a Dunnett’s multiple comparisons test. All data were expressed as the mean with the standard error of the mean. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, and ns were not considered significant.

Results

Sulfasalazine inhibiting glutamate release from U-87 MG

Previous studies have demonstrated the glioma-reducing properties of sulfasalazine (an inhibitor of system Xc−) (Lyons et al. 2007). We also compared the amount of glutamate released from U-87 MG cells treated with sulfasalazine. The glutamate assay revealed a notable reduction in glutamate release in the sulfasalazine-treated group (Fig. 1A). This finding confirmed the inhibitory effect of sulfasalazine on system Xc−, which decreases glutamate release from U-87 MG.

Fig. 1
figure 1

Sulfasalazine reduces glutamate release independent of calcium response. A The bar graph shows the concentration of glutamate in each condition (n = 6, respectively). The individual dot indicates the data from a single well. B The images for calcium imaging from U-87 MG loaded with Fura 2-AM. Scale bar = 20 μm. The scale bar on the right represents the normalized 340/380 nm ratio. C Traces of calcium imaging from control cells (left, n = 73) and cells treated with 250 μM sulfasalazine (right, n = 74). Scale bars = 0.01 ratio, 50 s time. The black trace depicts the calcium signal for individual cells, while the red trace represents the average value of the cells. The blue horizontal bar denotes the application of 30 μM TFLLR at the indicated time. D The bar graph illustrates the normalized 340/380 nm peak. Each dot represents the data from a single cell. All data are expressed as the mean with standard error of the mean. All data were analyzed by two-tailed unpaired Student’s t-test. ****P < 0.0001

Full size image

Calcium is a signaling molecule that influences the mobility and invasiveness of GBM (So et al. 2021). Therefore, calcium imaging was conducted to determine whether the GPCR pathway-related calcium response is affected by sulfasalazine (Fig. 1B). In this study, we used TFLLR to induce an activity-dependent calcium response, employing Fura 2-AM as a calcium indicator, as described previously (Roe et al. 1990). TFLLR functions as a selective peptide agonist that targets protease-activated receptor 1 (PAR1) (Hollenberg et al. 1997; Woo et al. 2012). PAR1, a member of the G protein-coupled receptor (GPCR) family, exhibits coupling with various G proteins such as Gαi, Gαq/11, and Gα12/13. Consequently, the presence of TFLLR triggers an intracellular calcium response due to this coupling. No significant difference was found in the intracellular calcium response between the sulfasalazine-treated and control groups (Fig. 1C and D). This finding suggests that glutamate release, which is reduced by sulfasalazine, does not affect GPCR-mediated intracellular calcium response in U-87 MG cells.

Decreased glutamate release in U-87 MG by silencing system Xc−

Although sulfasalazine is a commonly used treatment for glioma, it is nonspecific to system Xc− and has various side effects (Bridges et al. 2012; Cantarini et al. 2007). To address this, we used a short hairpin RNA (shRNA) to specifically silence system Xc. Figure 2A shows a schematic diagram of shRNA production and lentivirus infection. To validate the knockdown efficiency, qRT-PCR was performed in triplicate (Fig. 2B). Among the candidates, Candidate 1 exhibited superior knockdown efficiency, showing statistical significance Consequently, Candidate 1 was selected as the Xc-shRNA system for subsequent experiments.

Fig. 2
figure 2

Knockdown of SLC7A11 in U-87 MG by shRNA. A Schematic image of the procedure of shRNA production and lentiviral infection. B Bar graph depicting the relative expression in each condition (n = 3, respectively). Each dot represents the data from a single tube, randomly paired and calculated. Data are expressed as the mean with standard error of the mean. All data were analyzed using one-way analysis of variance followed by Dunnett’s multiple comparisons test. *P < 0.05

Full size image

Next, a glutamate assay was conducted to assess the knockdown effect achieved by the application of shRNA. The results revealed a significant reduction in glutamate release within the system Xc-shRNA group compared with that in the scrambled shRNA group (Fig. 3A). Additionally, calcium imaging using TFLLR was performed to determine whether the intracellular calcium response was altered by the introduction of Xc-shRNA (Fig. 3B). No difference was found in the calcium response between the scrambled shRNA and Xc-shRNA groups (Fig. 3C and D). These collective findings indicate that glutamate release is reduced in U-87 MG cells by the application system Xc-shRNA and is not related to the GPCR-mediated intracellular calcium response.

Fig. 3
figure 3

System Xc− knockdown by shRNA reduces glutamate release independent of calcium response. A The bar graph shows the concentration of glutamate in each condition (n = 3, respectively). Each dot represents the data from a single well. B Representative images of DIC, RFP, and Fura 2-AM from the left. Scale bar = 25 μm. C Calcium imaging traces from the from cells expressing scrambled shRNA (left, n = 153) and system Xc- shRNA (right, n = 51). Scale bars = 0.02 ratio, 50 s time. The black trace represents the calcium signal for individual cells, while the red trace denotes the average value of the cells. The blue horizontal bar indicates the application of 30 μM TFLLR at the specified time. D The bar graph shows the normalized 340/380 nm peak. Each dot represents the data from a single cell. All data are expressed as the mean with standard error of the mean. All data were analyzed by two-tailed unpaired Student’s t-test. **P = 0.0034

Full size image

Regulation of U-87 MG invasion influence by glutamate released by system Xc−

To determine the impact of reduced glutamate release resulting from SLC7A11 gene silencing on the invasion of GBM cells, an invasion assay was performed. Compared with the scrambled shRNA group, the Xc-shRNA-expressing group showed a significant reduction in invasion (Fig. 4A and B). The invasiveness index, calculated as the percentage of shRNA-expressing cells (mCherry+) relative to the total number of invading cells (DAPI+), demonstrated a notable decrease from 77.9% in the scrambled shRNA group to 9.7% in the Xc-shRNA group.

Fig. 4
figure 4

Inhibition of system Xc− via shRNA reduces the invasion of U-87 MG. A The representative images for the invasion assay were obtained by staining U-87 MG with DAPI in each expression condition. Scale bar = 500 µm. B The percentage of shRNA expression/DAPI stained cells was presented in each condition (n = 6, respectively). C The bar graph shows the proliferation O.D. values (n = 6, respectively). Each dot represents the data from a single well. All data are expressed as the mean with standard error of the mean. All data were analyzed using two-tailed unpaired Student's t-test. ****P < 0.0001

Full size image

An additional proliferation test was conducted; however, no significant differences were observed between the experimental groups. (Fig. 4C). These findings indicate that the glutamate released by system Xc− does not significantly influence the proliferation of GBM but specifically affects its invasive capacity. Moreover, it demonstrated a similar effect to sulfasalazine treatment on invasion, proliferation, and calcium response, implying its potential as an alternative to sulfasalazine for the treatment of GBM.

Discussion

Figure 5 presents a schematic summary of the findings of this study. Both the pharmacological approach using sulfasalazine and the gene-silencing method using shRNA demonstrated efficacy in inhibiting glutamate release from U-87 MG cells. In alignment with the results of previous studies using sulfasalazine (Lyons et al. 2007), the inhibition of system Xc− by shRNA reduced the invasion of GBM cells. Furthermore, the activation of Gαq/11 using TFLLR showed no difference in calcium response among all experimental groups. This observation suggests that the autocrine signaling of glutamate through system Xc− in GBM is more closely associated with the AMPA or NMDA receptor than with the GPCR pathway.

Fig. 5
figure 5

Schematic diagram of the study process

Full size image

Glutamate plays a key role in the invasion process of glioblastoma (So et al. 2021; Lyons et al. 2007; Groot and Sontheimer 2011). Recent studies have highlighted the significance of ion movement mediated by AMPA or NMDA receptors in GBM, suggesting these receptors as potential therapeutic targets (Corsi et al. 2019; Nandakumar et al. 2019). Notably, the inhibition of AMPA receptors in GBM induces apoptosis (Ishiuchi et al. 2002). However, the role of NMDA receptors remains controversial. Some studies suggested their absence in glioma cells (Lyons et al. 2007; Stepulak et al. 2009), while others indicated that inhibiting NMDA receptors through antagonists reduces invasion (Nandakumar et al. 2019; Müller-Längle et al. 2019). Therefore, targeting AMPA or NMDA receptors has traditionally been considered for GBM treatment; however, the complexity of the signaling pathways involved poses challenges to the development of effective treatment strategies (Corsi et al. 2019; Ishiuchi et al. 2007). Beyond these receptors, the kynurenic acid receptor has also been identified as a regulator of GBM mobility (Walczak et al. 2014). Furthermore, the metabotropic glutamate receptor (mGluR) family in GBM warrants attention, given the limited research conducted on this aspect.

System Xc− facilitates both glutamate export and cystine import. The physiological role of cysteine in cells is important. A study conducted a few years ago revealed that cystine uptake via system Xc− triggered cell death under glucose-deprived conditions (Goji et al. 2017). Additionally, another study suggested that excessive cysteine absorption may lead to a phenomenon known as disulfidoptosis in GBM (Yan et al. 2023). Consequently, system Xc− is intricately linked to both the glutamate and cystine metabolic pathways, making it challenging to draw definitive conclusions. Further research is required to fully elucidate their physiological functions.

This study used sulfasalazine to pharmacologically inhibit system Xc−. Sulfasalazine is commonly utilized for the treatment of rheumatoid arthritis, ulcerative colitis, and Crohn’s disease. It was approved for medical use in the United States in 1950 (Case 2001; Stuart et al. 2009; Pearl et al. 1986). It is mainly used to treat idiopathic urticaria, psoriasis, cirrhosis, and amyloidosis (McGirt et al. 2006; Gupta et al. 1990; Oakley et al. 2005; Brumshtein et al. 2015). Recently, it has been used in the treatment of pancreatic cancer, small cell lung cancer, and prostate cancer therapy (Lo et al. 2010; Guan et al. 2009; Doxsee et al. 2007), with ongoing investigations into its therapeutic viability for GBM (Bridges et al. 2012; Sontheimer and Bridges 2012). However, approximately 20–25% of patients using the medication have reported side effects (Stuart et al. 2009; Das et al. 1973). Symptoms such as nausea, vomiting, malaise, headache, and hemolysis are prevalent. The incidence of relatively mild side effects mainly depends on the dose used. Less frequent but more severe side effects that can affect almost any organ system have also been reported. These conditions include rashes, fever, joint pain, blood abnormalities, hepatitis, pneumonia, and male infertility (Peppercorn 1984). Hence, our findings suggest that employing gene-silencing methods might offer more promising avenues than conventional pharmacological drugs for targeting specific pathways or mechanisms of disease intervention. The potential expression of GBM-specific shRNAs holds promise, suggesting that achieving an ideal treatment for GBM might be attainable, provided that such targeted approaches can be effectively harnessed.

Finally, the implications of these studies should be extended to the treatment of human GBM. Therefore, validating the inhibition of invasion in an in vivo model or in patient-derived GBM samples is imperative. Based on our hypothesis, we anticipated reduced invasion and prolonged survival in the Xc knockdown group in GBM xenografts performed on nude mice. In addition, parallel in vitro experiments should corroborate decreased glutamate release and invasion of GBM samples obtained from patients. Although research on system Xc− currently presents various possibilities and lacks clarity, further investigations to unravel its physiological functions could potentially open up promising avenues for GBM treatment.