The human GBM cell line U-87 MG (U87; glioblastoma of unknown origin) was purchased from The Cell Bank of Type Culture Collection of the Chinese Academy of Sciences and was authenticated by Short Tandem Repeat profiling as described previously (23). U87 cells were cultured in DMEM (HyClone; Cytiva) supplemented with 10% FBS (Clark Bioscience), 100 U/ml penicillin and 100 µg/ml streptomycin (Biosharp Life Sciences) at 37°C in a cell culture incubator (Thermo Forma 4131; Thermo Fisher Scientific, Inc.) with 5% CO₂.

To determine the effect of α-Ketoglutarate (α-KG; Sigma-Aldrich; Merck KGaA) on cell migration, U87 cells were treated with 1 or 2.5 mM α-KG supplemented-medium during culture at 37°C for 24 h. To investigate the effect of the IDH1/α-KG axis on the PI3K/AKT/mTOR signaling pathway, rapamycin (MedChemExpress), an mTOR-specific inhibitor, was used to treat wild-type or IDH1-overexpressing U87 cell cultures at 37°C for 24 h to block the PI3K/AKT/mTOR signaling pathway. Mock control-treated cells [treated with DMSO (1:200)] were used as controls.

IDH1 overexpression plasmids (pcDNA3-cMyc-IDH1) were generated by inserting IDH1 coding gene fragments into pcDNA3 vectors (Invitrogen; Thermo Fisher Scientific, Inc.). The IDH1 coding gene fragment was amplified from the human cDNA library, which was produced by the reverse transcription of total RNA extracted from 293T cells, using the following primer pairs: Forward (containing the BamHI enzyme site and the cMyc tag sequence), 5′-CCGGATCCGCCACCATGGAGCAGAAGCTGATCTCAGAGGAGGACCTGATGTCCAAAAAAATCAGTGGCG-3′ and reverse (containing the EcoRI enzyme site), 5′-CGCGAATTCTTAAAGTTTGGCCTGAGCTAGT-3′. The amplified fragment was then ligated into the pcDNA3 vector at the BamHI and EcoRI sites. pcDNA3-cMyc-IDH1 plasmids were sequenced to confirm its validity using T7 and SP6 general sequencing primers at General Biosystems, Inc. The pcDNA3 empty vector was used as a control. IDH1 overexpression (OE; IDH1-OE) and control (Ctrl) cell lines were constructed by stably transfecting U87 cells with IDH1 overexpression plasmids (pcDNA3-cMyc-IDH1) and empty vectors (pCDNA3), respectively. IDH1 overexpression was verified by western blotting with antibodies against the cMyc tag and IDH1.

A total of 2 IDH1 shRNA plasmids (shIDH1-1 and shIDH1-2) were produced by inserting shRNA sequences designed against IDH1 into a pLL4.0 vector, which was produced by replacing the GFP expression cassette of the pLL3.7 vector (cat. no. 11795; Addgene, Inc.) with a Neomycin expression cassette. The shRNA sequences were placed downstream of the U6 promoter in the pLL4.0 vector. The shRNAs were designed using the siRNA at WHITEHEAD website (http://jura.wi.mit.edu/bioc/siRNAext) and were synthesized at General Biosystems, Inc. The sense and antisense strands were annealed into double-stranded DNA fragments, which were subsequently phosphorylated. The phosphorylated shRNA fragments were ligated into the pLL4.0 vector at the HpaI and XhoI sites. The shRNA plasmids were sequenced to confirm their validity. Scramble shRNA (shScramble) was used as a control. The sequences of the shRNA sense and antisense strands are listed in Table SI.

The plasmid transfections were performed using 1.5 µg plasmid/well and Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol, upon the cells reaching 70–90% confluence (8.0×10⁵ cells/well). Following transfection at 37°C for 24 h, the culture medium was changed and selection was performed beginning at 48 h post-transfection to generate stable cell lines; 400 µg/ml geneticin (Sigma-Aldrich; Merck KGaA) or 1 µg/ml puromycin (Sangon Biotech Co., Ltd.) was added to the culture medium of cells transfected with the pCDNA3- or pLL4.0-based plasmid.

U87 cell lysates were extracted using Cell Lysis Buffer (Beyotime Institute of Biotechnology) supplemented with protease inhibitors (cOmplete, Mini, EDTA-free Protease Inhibitor Cocktail; Roche Diagnostics), according to the manufacturer's protocol. Protein concentrations were measured using the BCA protein assay kit (Biosharp Life Sciences), according to the manufacturer's protocol, and were adjusted to the same concentration in each set of experiments. Protein samples (4 µg/lane) were subjected to electrophoresis on 10% SDS-PAGE gels and then transferred onto PVDF membranes. The membranes were blocked with 5% bovine serum albumin (Sigma-Aldrich; Merck KGaA) at room temperature for 1 h and incubated with primary antibodies overnight at 4°C. The membranes were then incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies at room temperature for 1.5 h and reacted with chemiluminescent substrates (Biosharp Life Sciences). The signals were captured by the Tanon 5200 Imaging system (Tanon 5200; Tanon Science and Technology Co., Ltd.) and the expression levels were analyzed using ImageJ 1.52a software (National Institutes of Health). The antibodies and their dilutions were as follows: cMyc (1:1,000; Thermo Fisher Scientific, Inc.; cat. no. A21280), IDH1 (1:1,000; Hangzhou HuaAn Biotechnology Co., Ltd.; cat. no. EM40705), GAPDH (1:2,000; Biosharp Life Sciences; cat. no. BL006B), phosphorylated (p)-AKT1 (Ser473; 1:1,000; Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. MA120325), AKT (1:1,000; Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 44609G), mTOR (1:1,000; Cell Signaling Technology, Inc.; cat. no. 2983), p-mTOR (Ser2448; 1:1,000; Cell Signaling Technology, Inc.; cat. no. 9205), goat anti-mouse HRP-conjugated immunoglobulin (Ig) G (1:2,000; Biosharp Life Sciences; cat. no. BL001A) and donkey anti-rabbit HRP-conjugated IgG (1:2,000; Invitrogen; Thermo Fisher Scientific, Inc.; cat. no. 31458).

Total RNA was extracted from cells using the Total RNA Isolation reagent (Biosharp Life Sciences). Reverse transcription was performed using the FastKing RT kit (Tiangen Biotech Co., Ltd.), according to the manufacturer's protocol, and qPCR was performed using the Powerup SYBR Master mix (Applied Biosystems; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol, on an RT-qPCR machine (Bio-Rad CFX96 Touch; Bio-Rad Laboratories, Inc.). The following thermocycling conditions were used for the qPCR: Pre-culture at 50°C for 120 sec; initial denaturation at 95°C for 120 sec; 40 cycles of annellation at 95°C for 15 sec and elongation at 60°C for 60 sec; and a default dissociation step of 95°C for 15 sec and then heated from 60–95°C in incremental steps of 0.2°C for 15 sec. GAPDH served as an internal control. Gene expression levels were quantified using the 2^−ΔΔCq method (24). The primers used for RT-qPCR (IDH1 and GAPDH) are listed in Table SI.

Transwell migration assays were performed as previously reported (25). Transwell inserts with 8.0-µm pore polycarbonate membranes (Corning Inc.) and 24-well plates were utilized. U87 cells were initially seeded at 2.5×10⁴ cells/well in a 100 µl serum-free DMEM, supplemented with 100 U/ml penicillin and 100 µg/ml streptomycin, in the upper chamber. DMEM supplemented with 10% FBS, 100 U/ml penicillin and 100 µg/ml streptomycin was plated into the lower chambers. Following incubation for 20 h at 37°C in a cell culture incubator with 5% CO₂, the migratory cells were fixed with 4% paraformaldehyde at room temperature for 10 min and then 100% methanol for 10 min at room temperature. Next, the cells were stained with 0.05% crystal violet for 30 min at room temperature. Images were captured using the bright field channel of an Olympus IX71 fluorescence microscope (magnification, ×100; Olympus Corporation).

U87 cells were seeded in 6-well plates at a seeding density of 6×10⁵ cells/well and cultured overnight at 37°C in a cell culture incubator with 5% CO₂. Wounds were generated by scratching cells with 0–200 µl pipette tips. Following scratching, cells were briefly washed three times with PBS and cultured in DMEM for 48 h. The cells were harvested at 0 and 48 h post-scratch. The cells were washed twice with cold PBS and fixed with chilled methanol for 10 min on ice. Cells were incubated with a staining solution that contained 0.05% crystal violet and 25% methanol for 20 min at room temperature and washed with distilled water. Images were captured with the bright field channel of an Olympus IX71 fluorescence microscope (magnification, ×40; Olympus Corporation). The width of the wounds was measured by Canvas X software (version 19; Canvas GFX) and the migratory distances were calculated using the following equation: Scratch width at 0 h-scratch width at 48 h. All migratory distances were normalized to the migratory distances of the control groups at 0 h.

α-KG levels in U87 cells were measured using the a-Ketoglutarate Colorimetric/Fluorometric Assay kit (BioVision, Inc.) according to the manufacturer's protocol. The values were determined by measuring absorbance at 570 nm using a TECAN Microplate Reader (Tecan Group, Ltd.).

TCGA data analysis was performed using the UALCAN website (ualcan.path.uab.edu/index.html) (26). GBM samples were selected for analysis; the analysis included 156 primary GBM samples and 5 normal samples. The transcript per million (TPM) of IDH1, AKT, PTEN, CDK2, Myc, MDM2, SNAIL2, N-cadherin, Vimentin, TWIST1, ZEB1 and RAC1 in the above samples were analyzed using the UALCAN website and plotted as boxplots.

Experiments were performed in triplicate. All data were normalized to the controls and presented as the mean ± standard deviation, unless otherwise stated. Student's t-tests were used for two-group comparisons and one-way ANOVA, followed by post-hoc Tukey tests, was used for multiple-group comparisons. P<0.05 was considered to indicate a statistically significant difference. Significant differences in the statistical analyses are labeled as ‘*’ in all the figures. All statistical analyses were performed using GraphPad Prism software (version 8; GraphPad Software, Inc.).

Previously, numerous studies have revealed that IDH1 site mutations, particularly the IDH1 R132H mutation, promoted tumorigenesis (13,18–21). However, in primary GBM, IDH1 mutations were not the main cancer-causing factor (15). The analysis of TCGA data demonstrated a significant increase in IDH1 expression in samples from patients with primary GBM compared with adjacent normal samples (Fig. 1A). IDH1 knockdown and scramble control cell lines were constructed by stably transfecting U87 cells with two IDH1 shRNA expression plasmids (shIDH1-1 or shIDH1-2) or shScramble expression plasmids, respectively. IDH1 knockdown was verified by western blotting (Fig. 1B and C) and RT-qPCR (Fig. S1A). Wound healing assays were performed on IDH1 knockdown and scramble shRNA control cells. At 48 h post-scratch, shIDH1-1 and shIDH1-2 groups demonstrated relatively delayed migration compared with the scramble shRNA group (Fig. 1D and E). Additionally, Transwell migration assays were performed with the same cell lines, the results of which demonstrated that IDH1 knockdown had fewer cells that successfully migrated through the membrane of the Transwell inserts (Fig. 1F and G). In summary, the results indicated that IDH1 downregulation repressed the migration of primary GBM cells.

As the downregulation of IDH1 repressed primary GBM cell migration, the effect of ectopic IDH1 expression on primary GBM cell migration was investigated. The results revealed that cMyc was strongly expressed in the IDH1-OE group; however, it was not detected in the Ctrl group (Fig. 2A). Furthermore, IDH1 expression was significantly increased in the IDH1-OE group compared with the Ctrl group (Fig. 2B). Similarly, IDH1 mRNA levels were increased in the IDH1-OE group, as indicated by RT-qPCR (Fig. S1B). IDH1 expression did not increase markedly, which may be due to IDH1 being endogenously abundant in U87 cells (15). A wound healing assay was then performed on IDH1-OE and control cell lines. Significantly faster migration was observed in the IDH1-OE group (Fig. 2C and D). The results of the Transwell migration assays also demonstrated that IDH1 overexpression promoted U87 cell migration (Fig. 2E and F). In conclusion, IDH1 overexpression facilitated the migration of primary GBM cells.

Since IDH1 catalyzes the conversion of isocitrate to α-KG and CO₂, whether α-KG levels in primary GBM cells were associated with IDH1 expression was examined. α-KG levels were significantly downregulated in the IDH1 knockdown groups (Fig. 2G) and significantly upregulated when IDH1 was overexpressed (Fig. 2H). Whether α-KG levels affected primary GBM cell migration was then investigated. U87 cells were treated with 1 and 2.5 mM α-KG in the medium during culture. The migration of the U87 cells was promoted by the addition of α-KG in a dose-dependent manner in the wound healing and Transwell assays (Fig. 2I-L). These results indicated that α-KG levels may mediate the changes in migration in primary GBM cells caused by IDH1 level alterations.

Primary GBM is a type of cancer with the highest mortality partly due to its relatively high migration rate (27,28). TCGA database analysis revealed that PI3K/AKT/mTOR pathway activity was significantly increased in primary GBM. AKT and phosphatase and tensin homolog, which are key components of the PI3K/AKT/mTOR pathway, were significantly upregulated and downregulated in primary GBM, respectively (Fig. S2A). The expression of the genes downstream of the PI3K/AKT/mTOR pathway and whose expression is regulated by the pathway was then examined. Cyclin-dependent kinase 2, Myc and mouse double minute 2 homolog were significantly upregulated in primary GBM, which are proteins that promote the cell cycle and repress apoptosis (Fig. S2B) (29). Snail family transcriptional repressor 2, N-cadherin and vimentin were also significantly upregulated, which facilitate the epithelial-mesenchymal transition (EMT) process (Fig. S2C). Additionally, twist-related protein 1, zinc finger E-box-binding homeobox 1 and Ras-related C3 botulin toxin substrate 1 were upregulated, which are proteins that enhance cell migration and metastasis (Fig. S2D) (29–31). The results indicated that the PI3K/AKT/mTOR pathway was hyperactivated in primary GBM, which is associated with the cell cycle, apoptosis, EMT, cell migration and metastasis processes (32–36).

As PI3K/AKT/mTOR pathway activity was increased in primary GBM, whether IDH1 and α-KG expression affected the PI3K/AKT/mTOR pathway was investigated. The results of western blotting revealed that IDH1 knockdown and overexpression resulted in the downregulation and upregulation of p-AKT (Ser473) and the p-mTOR (Ser2448), respectively, in primary GBM cells (Fig. 3A-D and G-J). By treating primary GBM cells with 1 and 2.5 mM α-KG, p-AKT (Ser473) and p-mTOR (Ser2448) were significantly upregulated in a dose-dependent manner (Fig. 3E, F and K, L). The results indicated that IDH1 and α-KG levels regulated PI3K/AKT/mTOR pathway activity in primary GBM cells.

As PI3K/AKT/mTOR pathway activity was regulated by changes in the IDH1 and α-KG levels, whether IDH1/α-KG regulated primary GBM cell migration by modulating the PI3K/AKT/mTOR pathway was examined. Wild-type U87 cells were treated with rapamycin, which is an mTOR-specific inhibitor, to block the PI3K/AKT/mTOR pathway. Mock control-treated cells were used as controls. Western blotting results reported that p-mTOR (Ser2448) was significantly inhibited by rapamycin treatment compared with controls (Fig. S3A and B). Additionally, rapamycin treated U87 cells exhibited significantly delayed cell migration compared with controls, indicating that blocking the PI3K/AKT/mTOR pathway led to the repression of cell migration (Fig. S3C and D). Furthermore, IDH1-overexpressing U87 cells and α-KG-treated U87 cells were treated with rapamycin. Western blotting results demonstrated that p-mTOR (Ser2448) was repressed by rapamycin in both treatment groups (Fig. 4A-D). The increased cell migration caused by IDH1 overexpression and α-KG supplementation were also reversed following rapamycin treatment (Fig. 4E-H). The results indicated that the IDH1/α-KG axis regulated primary GBM cell migration by modulating the PI3K/AKT/mTOR pathway (Fig. 5).