Silvestrol (MCE, cat: HY-13251, NJ, USA) was purchased from MCE, and dissolved in dimethyl sulfoxide (Beyotime, ST038, Shanghai, China). 4% paraformaldehyde (Biosharp, BL539A, GuangDong, China), LY3214996 (MCE, HY-101494, NJ, USA). Antibodies: β-actin, cleaved caspase3, extracellular regulated kinase 1/2 (ERK1/2), p-ERK1/2, mitogen-activated protein kinase kinase (MEK), p-MEK, Goat Anti-Rabbit IgG (H + L) HRP (Affinity, LOT: AF7018, AF7022, AF1015, AF0155, AF8035, AF6385, S0001, WuHan, China), neuregulin 1 (NRG-1) (66492-1-Ig, protein tech, Wuhan, China).

U87 MG and U251 glioblastoma cell lines were obtained from Wuhan Prosser Life Technology Co., Ltd. China. The glioblastoma cell line was cultured in Dulbecco’s modification of Eagle’s medium (BasalMedia, L110KJ, Shanghai, China) medium containing 10% fetal bovine serum (Procell, 164210-500, Wuhan, China) and 1% Penicillin-Streptomycin Solution (Procell, PB180120, Wuhan, China) at 37°C in a 5% CO₂ incubator.

The half maximal inhibitory concentration (IC50) of silvestrol on U87 MG and U251 cells for 48 h was determined using the cell counting kit-8 (CCK-8) (DOJINDO, CK04, Kumamoto, Japan). U87 MG and U251 cells were seeded in 96-well plates at 2000 cells per well. After culturing the cells overnight, they were treated with different concentrations of silvestrol for 48 h. 10 μL of CCK-8 solution was then added to each well and incubated for 3 h at 37°C. Subsequently, their optical density values at 450 nm were measured using a microplate reader (BioTek, SYNERGY/H1, Vermont, Germany).

After the IC₅₀ was determined, 0, 2.5, and 10 nM silvestrol were treated with cells for 24, 48, and 72 h, respectively, to assess silvestrol’s impact on cell proliferation, and the rest of the operations were carried out as above.

87 MG and U251 cells were plated in 96-well plates at a density of 5 × 10⁴ cells per well. After treatment with drugs for 24 h, 100 μL of EDU (Beyotime, C0088S, Shanghai, China) working solution (20 μM) was added and cells were cultured for an additional 2 h. Following this, cells were fixed with 4% paraformaldehyde (Beyotime, P0099, Shanghai, China) for 15 min at room temperature. Afterward, they were washed with 200 μL of immunostaining strong permeabilizer (Beyotime, P0097, Shanghai, China) and incubated at room temperature for ten minutes. Subsequently, 50 μL of click reaction solution was added, and cells were incubated in the dark at room temperature for 30 min. Finally, cells were counterstained with Hoechst 33342 for 10 min, and blocked, and fluorescence images were captured using a fluorescence microscope (Leica, DMi8, Hessen, Germany).

In 6-well plates, cells were seeded at 1000/well, and after overnight adherence, different concentrations of drugs were added. After 7–14 days, the cell culture medium was removed, washed with PBS, and then washed with 4% paraformaldehyde. The cells were fixed and stained with Giemsa staining solution (Beyotime, C0131, Shanghai, China) for 30 min. Lastly, the cells were washed with PBS three times and photographed.

The apoptosis rate of U87 MG cells was detected using an Annexin V-FITC/PI apoptosis kit (Multi Sciences, C0131, Hangzhou, China). U87 MG cells were harvested following a 24-h drug treatment, washed with pre-cooled PBS at 4°C, and resuspended in a buffer. Subsequently, the Annexin V-FITC/PI mixture was added, gently mixed and incubated for 15 min in the dark. Cell analysis was performed using a flow cytometer (Beckman, CytoFLEX LX, CA, USA).

In 6-well plates, U87 MG and U251 underwent drugs for 24 h and were collected. Cells were washed with PBS, fixed with 75% ethanol, and fixed overnight at 4°C. Before staining, washed with PBS, then added 500 μL PI/RNase Staining Buffer (BD, 550825, NJ, USA) and incubated for 15 min in the dark. Detected using a flow cytometer (Beckman, CytoFLEX LX, CA, USA).

Following a 24-h drug treatment of U87 MG and U251 cells, the cells were collected for Western blot (WB) experiments, with experimental procedures referenced from previous reports [29]. The reagents and materials used in this study are as follows: RIPA lysis buffer (Beyotime, P0013B, Shanghai, China), ultrasensitive multi-function imager (Cytiva, Amersham ImageQuant 800, MA, USA).

In 6-well plates, U251 was cultured until the cell density reached approximately 70%–80%. Slowly create straight scratches in the cell monolayer with a 200 μL pipette tip. After washing away floating cells with PBS, the cells were treated with various drugs in a serum-free medium and then incubated. The microscope was used to capture scratch images of each group at the same location at both 0 and 24 h.

U87 cell grouping: silvestrol (10 nM) and silvestrol (10 nM) + LY3214996 (ERK inhibitor). Total RNA was extracted using TRIzol reagent (Invitrogen, 15596018CN, CA, USA). Concentrations were quantified using Illumina Novaseq 6000 sequencing platform and cDNA was built libraries by starting material Poly (A)+ RNA, which finally was purified with AMPure XP beads. IlluminaHiSeq platform (BioMarker Technologies, Beijing, China) sequenced the library. Reference gene source: homo_sapiens (http://asia.ensembl.org/Homo_sapiens/Info/Index); Reference genome version: grch38. We normalized fragments per kilobase of transcript per million fragments mapped (FPKM) = cDNA fragments/[mapped fragments (millions) × transcript length (kb)].

Genes upregulated (|FC-1| ≥ 0.5 and p adjust < 0.05) or downregulated (|FC-1| ≥ 0.5 and p adjust < 0.05) by ERK inhibitors (LY3214996) were designated as the RNA-UP and RNA-DOWN gene sets, respectively. Additionally, differentially expressed genes in glioblastoma compared to normal tissues were identified as the GEPIA-UP and GEPIA-DOWN gene sets using the gene expression profiling interactive analysis (GEPIA) database.

The intersection of transcriptomic data and the GEPIA database was obtained to identify uniformly upregulated and downregulated genes for subsequent gene ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment. Furthermore, the association between hub genes and clinical survival rates was analyzed using the tumor immune estimation resource (TIMER) database.

Cell pellets were resuspended in TRNzol Universal Reagent (TianGen, DP424, Beijing, China) and incubated on ice for 30 min. One-fifth of the volume (200 μL) of chloroform was added, followed by gentle inversion for 15 times. The mixture was placed on ice for 10 min to allow phase separation. After centrifugation (4°C, 10000 rpm, 15 min) (Thermo, pico21/21R, CA, USA), the upper aqueous phase was carefully collected and mixed with an equal volume of isopropanol. The mixture was placed on ice for 10 min and then centrifuged (4°C, 10000 rpm, 10 min) to obtain RNA precipitate. The RNA pellet was resuspended in 700 μL of 75% ethanol, followed by centrifugation (4°C, 7500 rpm, 5 min) to wash the RNA. After air-drying on ice, the RNA sample was dissolved in 20 μL of RNase-free water. The PrimeScript™ RT reagent Kit with gDNA Eraser (TaKaRa, RR047A, Kusatsu, Japan) was utilized for the reverse transcription of RNA into cDNA. Real-time fluorescence quantitative PCR was conducted using the GoTaq® qPCR Master Mix (Promega, A6001, WI, USA). The sequence of primers is as follows: β-actin(human)-F: CATGTACGTTGCTATCCAGGC; β-actin(human)-R: CTCCTTAATGTCACGCACGAT; NRG1(human)-F: CGGTGTCCATGCCTTCCAT; NRG1(human)-R: GTGTCACGAGAAGTAGAGGTCT; MANBA(human)-F: TGAGCTGCGTTTCCAGTCAG; MANBA(human)-R: ACATGGCATTCACCCTTCTGC.

After reaching an appropriate cell density in a 96-well plate, U87 cells were fixed with 4% paraformaldehyde for 15 min. Following fixation, cells were washed with PBS and treated with 0.3% Triton X-100 for 10 min. Rinse the wells three times with PBS, then add 50 μL of immunostaining blocking solution (Beyotime, P0102, Shanghai, China), and incubate at room temperature for 60 min. The staining solution containing NRG-1 antibody was added, ensuring thorough wetting of the cells. Incubation was conducted overnight at 4°C, followed by three PBS washes. Subsequently, secondary antibodies (BBI, D110061, and D110090, Shanghai, China) were added and incubated for 1 h, followed by another round of PBS washes. DAPI (Beyotime, C1002, Shanghai, China) was diluted 1:100 in PBS and added (10 μL) for nuclear counterstaining. After 15 min of incubation in the dark, specimens were washed five times with PBST (5 min each) to remove excess DAPI. Finally, the localization and expression of NRG-1 antibody in the cells were observed by a fluorescence microscope (Leica, DMi8, Hessen, Germany).

Comparisons were performed using Prism (Graphpad Software Inc, GraphPad 8.0.1, CA, USA). All data were expressed as the mean ± SD and the one-way ANOVA and Tukey’s multiple comparison test were used to determine the statistical significance of the differences among groups. p < 0.05 was considered to indicate a statistically significant difference.

Initially, half maximal inhibitory concentration (IC50) of silvestrol was determined in two glioblastoma cell lines over 48 h, revealing IC50 values of 6.379 nM for U251 and 9.154 nM for U87 MG. Compared to the control group, the suppressive effect of silvestrol on brain glioblastoma cells was observed to increase significantly with higher drug concentrations, indicating a clear dose-response relationship (p < 0.05) (Figs. 1A and 1B). Based on these findings, the following experiments were conducted using 10 nM as the high-dose group, 4 nM as the mid-dose group, and 2.5 nM as the low-dose group. Cell proliferation assays demonstrated that silvestrol’s inhibitory effect on both glioblastoma cell types intensified over time and with increasing concentration (Figs. 1C and 1D). Additionally, colony formation assays provided further proof of silvestrol’s suppressive effects, showing a significant decrease in the cloning efficiency of glioblastoma cells at both high and low doses of silvestrol compared to the control group (Figs. 1E and 1F). Silvestrol was also found to hinder the migration of U251 cells, with higher concentrations of silvestrol correlating with a greater percentage of wound closure in a dose-dependent manner (Fig. 1G).

EDU proliferation assays revealed that silvestrol decreased the proportion of EDU-positive cells in both U251 and U87 MG cell lines (Figs. 2A and 2B). Flow cytometry analyses showed that silvestrol triggered cell cycle arrest in the G2 phase for U251 cells (Fig. 2C) and in the G1 phase for U87 MG cells (Fig. 2D). This arrest obstructed the seamless transition of cells into subsequent phases, effectively impeding DNA synthesis in both types of glioblastoma cells and disrupting normal cell division. Furthermore, levels of the apoptosis-indicative protein, cleaved caspase3, were significantly increased (Figs. 2E and 2F), with this elevation demonstrating a dose-dependent relationship (p < 0.05). Additional flow cytometry analyses of apoptosis indicated that silvestrol enhanced both early and late apoptosis in U87 MG cells (Fig. 2G). Remarkably, the group treated with a high dose of silvestrol showed a substantially higher total rate of apoptosis compared to both the control and low-dose groups (p < 0.05).

We detected a significant and rapid elevation in extracellular regulated kinase (ERK) phosphorylation levels in both U251 and U87 MG glioblastoma cells within 2 h of treatment with silvestrol at a concentration of 10 nM (p < 0.05). This phosphorylation surge returned to baseline levels after 2 h. Moreover, the phosphorylation of MEK in both U251 and U87 MG cells exhibited a notable increase (p < 0.05) (Figs. 3A and 3B). Additionally, upon administering silvestrol (4 nM) for 1 h, the phosphorylation levels were diminished when an ERK inhibitor, LY3214996, at a concentration of 80 nM was applied (Figs. 3C and 3D). These observations confirm that silvestrol treatment leads to the activation of the Ras/Raf/MEK/ERK signaling pathway and that LY3214996 is effective in inhibiting this activation of the ERK pathway.

The application of an ERK inhibitor introduced several significant changes. Firstly, it reduced the antiproliferative effects of silvestrol, as evidenced in Figs. 4A and 4B. It also lessened the cell cycle arrest caused by silvestrol, as illustrated in Figs. 4C and 4D, and improved the colony formation capability of U87 MG cells, as indicated in Fig. 4E. Moreover, the presence of the ERK inhibitor decreased the reduction in U251 cell migration induced by silvestrol, leading to a notably lower percentage of wound healing distance compared to the silvestrol medium-dose group (Fig. 4F). However, it is critical to note that the increase in levels of the apoptosis-related protein, cleaved caspase3, remained largely unaffected by the treatment with silvestrol (4 nM) in conjunction with the ERK inhibitor, as shown in Figs. 4G and 4H. The ERK inhibitor did not significantly influence the early and late apoptosis in U87 MG cells promoted by silvestrol, as depicted in Fig. 4I. Despite this, the overall rate of apoptosis experienced a significant increase (p < 0.05).

The overlap between RNA-UP and GEPIA-UP gene sets comprised 37 genes, whereas the intersection of RNA-DOWN and GEPIA-DOWN gene sets included 29 genes (Fig. 5A). Both the upregulated and downregulated hub genes implicated in catalytic activities, binding-related molecular functions, and a range of biological processes (Fig. 5B). In particular, genes that were upregulated were linked to immune diseases, cardiovascular diseases, and genetic information processing. In contrast, downregulated genes were associated with neurodegenerative diseases and resistance to anti-tumor drugs (Fig. 5C). Both hub-up genes and hub-down genes were also verified by UALCAN online tools and GSCA database (Fig. 5D and Table 1).

Moreover, TIMER database was used to analyze genes related to cumulative survival rates, and the results revealed that patients with high expression of the chromosome 1 open reading frame 226 (C1ORF226) gene experienced significantly higher survival rates (Log-rank p = 0.032). Conversely, patients with low expression levels of mannosidase beta A (MANBA) (Log-rank p = 0.015), IQ motif and Sec7 domain 2 (IQSEC2) (Log-rank p = 0.018), NRG1 (Log-rank p = 0.002) also showed significantly improved survival rates (Fig. 5E). Notably, C1ORF226 and MANBA were part of the upregulated hub gene set, while IQSEC2 and NRG1 were included in the downregulated hub gene set (Fig. 5D).

The expression levels of MANBA mRNA were significantly reduced by silvestrol at a concentration of 4 nM, and this reduction was reversed upon the introduction of LY3214996, as depicted in Fig. 6A. In contrast, silvestrol markedly increased the mRNA levels of NRG-1, and once again, the addition of LY3214996 reversed this effect, as shown in Fig. 6B. Interestingly, Western blot analysis revealed that silvestrol not only significantly decreased the protein levels of NRG-1 but also further reduced these levels upon the addition of LY3214996 (Fig. 6C). Immunofluorescence staining of cells displayed consistent patterns, with similar alterations in the levels of NRG-1 (Fig. 6D).