Sodium pyruvate (cat. no. P5280), lactate (cat. no. 1614308), D-glucose (cat. no. NIST917C), GlcNAc (cat. no. A3286), RPI-1 (cat. no. R8907), azaserine (cat. no. A4142), tunicamycin (cat. no. T7765) and OSMI-1 (cat. no. SML1621) were purchased from MilliporeSigma. Glioma and normal brain tissue samples were collected from the Department of Neurosurgery, First Affiliated Hospital of Zhengzhou University between August 2019 and September 2021. Tumors were classified histopathologically according to the 2016 World Health Organization classification (23). The study was approved by the ethics committee of the First Affiliated Hospital of Zhengzhou University (approval no. 2020-KY-155).

U251 and U87 human glioma cell lines were purchased from the American Type Culture Collection, the cell lines were authenticated by short tandem repeat (STR) analysis (HKGENE, Inc.). All cell lines were normally cultured in complete Dulbecco's modified eagle medium (DMEM; Thermo Fisher Scientific, Inc.) supplemented with 10 mM glucose, 10% fetal bovine serum (HyClone; Cytiva), 100 U/ml penicillin-streptomycin (HyClone; Cytiva) and 2 mM glutamine in a humidified atmosphere of 5% CO₂ at 37°C. Cells in the mid-log phase of growth were used for the experiments.

U251 and U87 human glioma cells were transfected at 80% confluence with SCAP siRNA (cat. no. sc-36462; Santa Cruz Biotechnology, Inc.), hypoxia-inducible factor 1 (HIF-1) siRNA (cat. no. sc-35561; Santa Cruz Biotechnology, Inc.), SREBP-1 siRNA (cat. no. sc-36557; Santa Cruz Biotechnology, Inc.) and control siRNA (cat. no. sc-37007; Santa Cruz Biotechnology, Inc.) using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Briefly, prior to treatment, Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.)/siRNA complexes were prepared in Opti-MEM medium, at the ratio of 1 μl Lipofectamine 2000 per 20 pmole siRNA. Cells then were then treated at 37°C for 48 h before being replaced with complete medium for the desired duration. The transfection efficiency of the cells was verified by western blotting.

Cell proliferation assays was performed as previously described (24). Briefly, U251 (1×10⁴/cells/200 μl), U87 (1×10⁴/cells/200 μl) glioma cells were seeded onto 96-well microplate and cultured at 37°C for 24 h and then treated with target compounds at given concentrations at 37°C for indicated periods. The cytotoxicity to glioma cells was determined with an MTT assay. Viability was expressed as a ratio to the absorbance value at 490 nm of the control cells, and the OD value was measured using a microplate reader (BioTek Instruments, Inc.).

A synergy analysis was performed as previously described (24). Briefly, the Chou-Talalay method and CalcuSyn software (version 1.0) (25) were used to determine the dose effect of combination therapy. For this synergy analysis, RPI-1 was combined with Fatostatin at a constant ratio for glioma cells at a dosage determined by the IC₅₀ of each drug. Interaction was quantified based on a combination index (CI) to assess synergism (CI <1), additive effect (CI=1), and antagonism (CI >1).

The collected U251, U87 glioma cells and tissue lysates were prepared using RIPA buffer (Beyotime Institute of Biotechnology) and total protein concentration was quantified using a bicinchoninic acid (BCA) assay kit. For the SCAP glycosylation analysis, the protein samples were treated with PNGase F according to the manufacturer's instructions (MilliporeSigma). Equal protein amounts (30–50 μg) were electrophoresed on 10% sodium dodecyl-sulfate polyacrylamide gel electrophoresis gels and the separated proteins were transferred to polyvinylidene difluoride membranes. After blocking with 5% skim milk at room temperature for 2 h, the membranes were probed with primary antibodies against acetyl CoA carboxylase (ACC; 1:1,000; cat. no. 3662, Cell Signaling Technology, Inc.), HIF-1 (1:1,000; cat. no. 36169, Cell Signaling Technology, Inc.), SCAP (1:1,000; cat. no. 13102, Cell Signaling Technology, Inc.), SREBP-1 (1:1,000; cat. no. sc-365513, Santa Cruz Biotechnology, Inc.), fatty acid synthase (FASN) (1:1,000; cat. no. 3180, Cell Signaling Technology, Inc.), RET (1:1,000; cat. no. 14556, Cell Signaling Technology, Inc.), phosphorylated (p)-RET (1:1,000; cat. no. SAB4504530, MilliporeSigma), ERK (1:1,000; cat. no. 5013, Cell Signaling Technology, Inc.), p-ERK (1:1,000; cat. no. 4370, Cell Signaling Technology, Inc.), stearoylCoA desaturase-1 (SCD1) (1:1,000; cat. no. 2794, Cell Signaling Technology, Inc.), β-tubulin (1:1,000; cat. no. 2128, Cell Signaling Technology, Inc.) and lamin B (1:1,000; cat. no. 13435, Cell Signaling Technology, Inc.), at 4°C for 12 h. Subsequently, the membranes were incubated with HRP-conjugated goat anti-rabbit IgG (1:5,000; cat. no. ZB-2301; OriGene Technologies, Inc.) or HRP-conjugated goat anti-mouse IgG (1:5,000; cat. no. ZB-2305, ZSGB-BIO; OriGene Technologies, Inc.) secondary antibodies at room temperature for 2 h. Immunoreactivity was visualized using the Odyssey Infrared Imaging System (LI-COR Biosciences).

RT-qPCR assays was performed as previously described (24). Briefly, U251 (1×10⁶/cells/5 ml) and U87 (1×10⁶/cells/5 ml) glioma cells were plated in 60 mm dishes and allowed to grow to 60–70% confluence and then treated with target compounds at given concentrations at 37°C for 24 h. Total RNA from the U251 and U87 glioma cells was extracted using a TRIzol^® reagent (Invitrogen, Thermo Fisher Scientific, Inc.). Following which, the RNA was reverse-transcribed to cDNA using Trans-Script First-Strand cDNA Synthesis SuperMix (TransGen Biotech Co. AT301); both procedures were performed according to the manufacturer's instructions. RT-qPCR reactions were performed using a SYBR green PCR master mix (TransGen Biotech, China) on a MxPro-Mx3005P real-time PCR system (Agilent Technologies, USA) and β-tubulin was used as the control. The qPCR conditions were as follows: Initial denaturation at 95°C for 30 sec; followed by 40 cycles of denaturation at 95°C for 15 sec, annealing at 60°C for 30 sec and extension at 72°C for 30 sec. The relative expression of target genes was calculated using the 2^-ΔΔCq method (26). The following primers were used: β-tubulin, F: 5′-GTG GTA CGG AAG GAG GCA GAG A-3′, R: 5′-AAC GGA GGC AGG TGG TGA CA-3′; GDNF, F: 5′-TCA CTG ACT TGG GTC TGG G-3′, R: 5′-TCA AAG GCG ATG GGT CTG C-3′; SREBP-1, F: 5′-CCA TGG ATT GCA CTT TCG AA-3′, R: 5′-CCA GCA TAG GGT GGG TCA AA-3′; SCD1, F: 5′-CAC TTG GGA GCC CTG TAT GG-3′, R: 5′-TGA GCT CCT GCT GTT ATG CC-3′; FASN, F: 5′-TGA GCA CAG ACG AGA GCA CCT T-3′, R: 5′-CGA TGT TGT AGA TGG CGG CTG AG-3′; ACC, F: 5′-TTC ACT CCA CCT TGT CAG CGG A-3′, R: 5′-GTC AGA GAA GCA GCC CAT CAC T-3′.

Immunofluorescence analysis was performed as previously described (24). Briefly, the treated cells were immunostained with an antibody to SREBP-1 (1:100; cat. no. sc-365513, Santa Cruz Biotechnology, Inc.) at 4°C overnight and subsequently incubated with fluorochrome-conjugated secondary antibody (1:100; cat. no. ZF-0311; OriGene Technologies, Inc.) for 0.5 h at room temperature in darkness. The nuclei were counterstained with DAPI at room temperature for 20 min. The SREBP-1 expression was monitored by confocal microscopy. Quantitative evaluation of SREBP-1 nuclear intensity was performed with ImageJ (v 1.8, National Institutes of Health).

U251 (1×10⁵/cells/500 μl) and U87 (1×10⁵/cells/500 μl) glioma cells were plated in 48-well microplate and cultured at 37°C for 24 h and then treated with target compounds at given concentrations at 37°C for 24, 48 or 72 h. Subsequently, 50 μM 2-NBDG (cat. no. 72987; MilliporeSigma) was added to the cells at 37°C for 1 h and the U251 and U87 glioma cells were washed in Hank's balanced salt solution buffer for three times. The fluorescent intensity was then measured using laser confocal microscopy at excitation and emission wavelengths of 467 and 542 nm, respectively.

The preprocessed level 3 RNA-seq data and corresponding clinical information of cancer patients were collected from The Cancer Genome Atlas (TCGA) database (http://cancergenome.nih.gov/) and the normal samples RNA data were acquired from the Genotype-Tissue Expression (GTEx) databases (https://www.gtexportal.org/).

The experiments were independently performed in triplicate and the results were presented as mean values ± standard deviation. Unpaired student's t-test was used to analyze the differences between two groups and one-way analysis of variance (ANOVA) followed by Tukey's test was used for the comparison among multiple groups. Patients were divided into high and low groups according to the 50% cut off point of GDNF and SREBP-1 expression and Kaplan-Meier survival analysis was used to analyzed significance between groups. All statistical analyses and experimental graphs were performed by GraphPad Prism version 8.0 software (GraphPad Software, Inc.). P<0.05 was considered to indicate a statistically significant difference.

The relative expression level of GDNF mRNA in normal brain and in low- and high-grade glioma tissues was determined by RT-qPCR. The results indicated that GDNF mRNA expression was upregulated in glioma compared to normal tissue (Fig. 1A). In addition, GDNF mRNA levels increased with pathological grade of glioma tissue (Fig. 1B). The GDNF mRNA levels between normal brain tissue and glioma tissue were then compared using RNA sequencing (RNA-seq) data from the GTEx database and The Cancer Genome Atlas (TCGA) database (http://cancergenome.nih.gov/). The results also showed that GDNF expression were upregulated in glioma compared to normal human tissue (Fig. 1C) and high GDNF gene expression was associated with poor prognosis in glioma (Fig. 1D).

In the presence of GDNF, SREBP-1 was activated (nSREBP-1) in a dose- and time-dependent manner. In addition, there was an increase in the protein levels of FASN, SCD1 and ACC, which are downstream targets genes of SREBP-1 (Fig. 1E and F). The immunofluorescence analysis showed that the nuclear fluorescence intensity of the SREBP-1 signal was significantly higher in U251 glioma cells treated with GDNF than in control cells (Fig. 1G). The RT-qPCR results showed that GDNF stimulation enhanced the SREBP-1 expression and activated the expression of SREBP-1 regulated genes involved in lipid metabolism (Fig. 1H). However, SREBP-1 expression in gliomas and its relationship with tumor malignancy remains to be elucidated.

The mRNA expression of SREBP-1 was more enriched in glioma than in normal human brain tissues, according to the RNA-seq data from the GTEx database and The Cancer Genome Atlas (TCGA) database (Fig. 1I). Therefore, it was decided to explore the prognostic value of SREBP-1 in gliomas based on the TCGA datasets. The results showed that glioma patients with higher SREBP-1 expression presented worse overall survival than those with lower SREBP-1 expression (Fig. 1J). Furthermore, the results of the present study showed that GDNF activates SREBP-1 through the RET/ERK signaling pathway (Fig. 1K). Therefore, GDNF pharmacologically blocked the activity of RET/ERK signaling with RPI-1 (GDNF/RET inhibitor), which significantly reduced the SREBP-1 activity (nSREBP-1; Fig. 1K). The MTT assay demonstrated that GDNF significantly promoted glioma cell proliferation in a dose- and time-dependent manner and that the inhibition of RET/ERK signaling could significantly reverse this biological effect (Fig. 1L and M).

GDNF was overexpressed in glioma and was associated with poor clinical outcome and highly expressed GDNF promoted the expression of SREBP-1, which is a transcription factor with a central role in lipid metabolism. Accordingly, patients with high expression of SREBP-1 presented poor prognosis. Although the present study revealed that SREBP-1 was activated by the GDNF/RET/ERK signaling pathway, the mechanisms underlying the oncogenic signaling to the SREBP-1 function remains to be elucidated.

Tumorigenesis is associated with increased glucose consumption and lipogenesis. Studies have suggested that elevated GDNF/RET signaling is associated with enhanced glucose uptake or lipogenesis in tumorigenesis (21,22). The present study demonstrated that GDNF promoted lipid metabolism by upregulating SREBP-1. GDNF also promoted glucose absorption (Fig. 2A and B) in a dose- and time-dependent manner and it pharmacologically blocked the activity of RET/ERK signaling with RPI-1, thereby significantly preventing glioma cells to absorb glucose (Fig. 2C). To investigate whether glucose was involved in SREBP-1 activation, U87 and U251 glioma cells were plated with GDNF with and without glucose and SREBP processing was analyzed by RT-qPCR, western blot and immunofluorescence microscopy. As shown in Fig. 2 D and E, GDNF stimulation presented no effect on the activation of SREBP-1 in a glucose-free medium, despite the strong activation of the RET/ERK signaling pathway. By contrast, GDNF stimulation promoted SREBP-1 activity in the presence of glucose and it pharmacologically blocked the activity of RET/ERK signaling with RPI-1, which completely abolished the GDNF-mediated activation of SREBP-1 expression. The fluorescence imaging indicated that GDNF stimulation was unable to promote the nuclear translocation of SREBP-1 in the absence of glucose and the addition of glucose restored the GDNF-mediated SREBP-1 nuclear translocation (Fig. 2F). Although GDNF did not elevate SREBP-1 activity without glucose, RT-qPCR analysis showed that it still promoted SREBP-1 mRNA expression and was inhibited by the RET inhibitor RPI-1. However, there was no change in the downstream target gene expression of SREBP-1 (Fig. 2G). Combined with the results described above, this suggested that GDNF/RET/ERK promoted SREBP-1 mRNA, protein expression and glucose absorption and that glucose is important for the activation of SREBP-1.

To investigate the glucose function in SREBP-1 activation, glucose and its intermediate metabolites, namely N-acetylglucosamine (GlcNAc; HBP), lactate or pyruvate (glycolysis pathway), were added in a glucose-free medium to U251 and U87 glioma cells, respectively. The results showed that GlcNAc was as effective as glucose in enhancing SREBP-1 activity, whereas lactate and pyruvate presented no effect (Fig. 2H). GFPT is the rate-limiting enzyme of HBP and glioma cells were treated with azaserine (GFPT inhibitor). As expected, azaserine inhibited the glucose-mediated SREBP-1 activity, but did not inhibit the SREBP-1 activity mediated by GlcNAc (Fig. 2I). The addition of GlcNAc, which has been widely used to increase HBP production, restored SREBP-1 protein activity in U87 and U251 glioma cells, which was previously reduced by the azaserine treatment (Fig. 2J). However, HBP inhibition presented no effect on the expression of SREBP-1 mRNA (Fig. 2K). In addition, in both glioma cell lines tested, the toxicity of azaserine was at least partially reversed by GlcNAc supplementation (Fig. 2L). These results demonstrated that GDNF/RET can promote glucose absorption and subsequently activate SREBP-1 by accelerating HBP synthesis.

Proteases cleaving SREBPs are activated by SCAP (10). In the present study, the knockdown of SCAP using siRNA reduced the GDNF- and glucose-mediated activation of SREBP-1 (Fig. 3A). Nohturfft et al (27) and Cheng et al (28) show that the N-glycosylation status of SCAP affects its protein function. However the NetNGlyc server prediction of glycosylation sites using artificial neural networks (http://www.cbs.dtu.dk/services/NetNGlyc/) showed that SCAP presented both N- and O-glycosylation sites (Fig. 3B). UDP-N-acetyl glucosamine (UDP-GlcNAc), which is the end product of HBP, is the substrate for O- and N-glycosylations (29). Therefore, whether GDNF regulated SCAP levels by regulating its N- or O-glycosylation was investigated. Tunicamycin (inhibitor of N-glycosylation) and OSMI-1 (inhibitor of O-glycosylation) were added to U251 and U87 glioma cells, respectively, in a GNDF/glucose medium. As shown in Fig. 3C, tunicamycin inhibited the protein activity of SREBP-1, whereas OSMI-1 did not. Immunofluorescence analysis showed that in U251 glioma cells treated with tunicamycin, the nuclear fluorescence intensity of the SREBP-1 signal was significantly lower than in control cells and there was no clear change in the OSMI-1 group (Fig. 3D). To further confirm the function of SCAP N-glycosylation for the regulation of SREBP-1 activity, U251 glioma cells were cultured with GDNF with and without glucose. The N-glycosylation was investigated using PNGase F and the differences between glycosylated and deglycosylated proteins were evaluated by western blot tests. As shown in Fig. 3E, the treatment combining GDNF and glucose induced more total SCAP proteins and its glycosylated forms, which was associated with elevated SREBP-1 activation. As expected, GDNF- and glucose-mediated SCAP N-glycosylation and SREBP-1 activity were simultaneously inhibited by the RET inhibitor (RPI-1), GFPT inhibitor (azaserine) and N-glycosylation inhibitor (tunicamycin) (Fig. 3F-H). These results demonstrated that GDNF elevated the SREBP-1 activity through mechanisms involving the upregulation of SCAP N-glycosylation.

Although the results indicated that GDNF/RET/ERK signaling promotes glucose absorption and SREBP-1activation, the mechanisms through which GDNF/RET promotes glucose absorption are still unknown. The western blot results showed that the hypoxia-inducible factor 1 (HIF-1) protein levels increased significantly when U251 and U87 glioma cells were treated with GDNF in glucose medium and the GDNF-induced changes in HIF-1 expression were associated with SREBP-1 activation (Fig. 4A). HIF-1 is crucial for the reprogramming of cancer metabolism as it activates the transcription of genes that encode glucose transporters and glycolytic enzymes (30). In the present study, the knock- down of HIF-1 using siRNA reduced the GDNF-mediated glucose absorption (Fig. 4B), which was associated with the terminated SREBP-1 activation (Fig. 4C). Although the present study showed again that GDNF induced SREBP-1 activation depended on upregulated RET/ERK/HIF-1 signaling pathway, but knockdown of HIF-1 using siRNA had no effect on GDNF induced RET/ERK expression (Fig. 4C). Immunofluorescence analysis also showed that the nuclear fluorescence intensity of the SREBP-1 signal was significantly lower in U251 glioma cells treated with siHIF-1 than in the cells of the GDNF group (Fig. 4D). RT-qPCR analysis showed that knockdown of HIF-1 using siRNA reduced the SREBP-1 downstream target gene expression, but presented no apparent effect on SREBP-1 mRNA expression (Fig. 4E). In addition, GlcNAc supplementation restored the SREBP-1 protein activity (Fig. 4F) and cell toxicity (Fig. 4G) in U251 and U87 glioma cells, which were previously reduced by the siHIF-1 treatment. Therefore, it was hypothesized that the GDNF/RET/ERK signaling pathway regulated the expression of SREBP-1 and HIF-1, whereas the SREBP-1 activity is regulated by HIF-mediated glucose absorption (Fig. 5A).

SREBP-1 functions as a transcription factor that activates specific genes involved in cholesterol and fatty acid metabolism. SREBP-1 and its downstream target gene can be effectively regulated by GDNF and the knockdown of SREBP-1 can reduce GDNF-mediated SREBP-1 and its downstream target gene expression (Fig. 5B) and it can completely reverse the GDNF-induced cell proliferation (Fig. 5C). Fatostatin, a chemical inhibitor of the SREBP pathway (31), shows high antitumor activity for a number of cancers (32,33), but its effects on glioma cells are largely unknown. The present study showed that fatostatin reversed the GDNF induced SREBP-1 activity (Fig. 5D). The nuclear fluorescence intensity of the SREBP-1 signal was significantly lower in U251 glioma cells treated with GDNF and fatostatin than in GDNF-treated cells (Fig. 5E). In addition, in both glioma cell lines, GDNF-induced cell activity was completely reversed by the supplementation with fatostatin, which inhibited the growth of glioma cells in a dose- and time-dependent manner (Fig. 5F).

It is often ineffective to treat cancer only using traditional methods involving the inhibition of a single oncogene pathway or enzyme (7). Therefore, the combination of drugs and chemotherapeutic agents is becoming a popular therapeutic option. Accordingly, because GDNF/RET regulates SREBP-1 activity, the present study investigated if the GDNF/RET inhibitor could enhance the cytotoxicity of the SREBP-1 inhibitor. A proliferation assay was performed in which glioma cells were treated with RPI-1 and fatostatin at a constant ratio according to their respective IC₅₀. The combination of RPI-1 and fatostatin provided a antiproliferative effect stronger than that of single agents and showed synergistic effect when they were used in combination [combination index (CI)<1.0; Fig. 5G]). These results suggest that the inhibition of SREBP-1 combined with GDNF/RET signaling pathway might be a new therapeutic method for glioma.