Bortezomib inhibits growth and sensitizes glioma to temozolomide (TMZ) via down-regulating the FOXM1–Survivin axis
High-grade glioma (HGG) is a fatal human cancer. Bortezomib, a proteasome inhibitor, has been approved for the treatment of multiple myeloma but its use in glioma awaits further investigation. This study aimed to explore the chemotherapeutic effect and the underlying mechanism of bortezomib on gliomas.
U251 and U87 cell viability and proliferation were detected by 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2-H-tetrazolium bromide (MTT) assay, tumor cell spheroid growth, and colony formation assay. Cell apoptosis and cell cycle were detected by flow cytometry. Temozolomide (TMZ)-insensitive cell lines were induced by long-term TMZ treatment, and cells with stem cell characteristics were enriched with stem cell culture medium. The mRNA levels of interested genes were measured via reverse transcription-quantitative polymerase chain reaction, and protein levels were determined via Western blotting/immunofluorescent staining in cell lines and immunohistochemical staining in paraffin-embedded sections. Via inoculating U87 cells subcutaneously, glioma xenograft models in nude mice were established for drug experiments. Patient survival data were analyzed using the Kaplan–Meier method.
Bortezomib inhibited the viability and proliferation of U251 and U87 cells in a dose- and time-dependent manner by inducing apoptosis and cell cycle arrest. Bortezomib also significantly inhibited the spheroid growth, colony formation, and stem-like cell proliferation of U251 and U87 cells. When administrated in combination, bortezomib showed synergistic effect with TMZ in vitro and sensitized glioma to TMZ treatment both in vitro and in vivo. Bortezomib reduced both the mRNA and protein levels of Forkhead Box M1 (FOXM1) and its target gene Survivin. The FOXM1–Survivin axis was markedly up-regulated in established TMZ-insensitive glioma cell lines and HGG patients. Expression levels of FOXM1 and Survivin were positively correlated with each other and both related to poor prognosis in glioma patients.
Bortezomib was found to inhibit glioma growth and improved TMZ chemotherapy efficacy, probably via down-regulating the FOXM1–Survivin axis. Bortezomib might be a promising agent for treating malignant glioma, alone or in combination with TMZ.
KeywordsGlioma Bortezomib FOXM1 Survivin Temozolomide (TMZ) Chemotherapy
forkhead box M1
glioma stem cells
short interfering RNA
reverse transcription-quantitative polymerase chain reaction
Food and Drug Administration
American Type Culture Collection
Dulbecco’s modified Eagle’s medium
World Health Organization
The Cancer Genome Atlas
50% inhibitory concentration
phosphate buffer saline
analysis of variance
negative regulator of FOXM1
High-grade glioma (HGG) is one of the leading causes of cancer mortality in adults and impose a great challenge on its treatment [1, 2]. Owing to the introduction of the alkylating agent temozolomide (TMZ) and the adoption of radiotherapy with concomitant and adjuvant TMZ treatment, the median survival of patients with glioblastoma multiforme (GBM) has been prolonged from 12.1 to 14.6 months . However, the overall clinical effect of this regimen is still disappointing, mostly due to the inherent or induced resistance to TMZ therapy . Thus, more comprehensive understanding of the progression and resistance mechanisms and novel therapeutic targets are urgently needed for the clinical management of this fatal tumor .
The ubiquitin–proteasome system plays an important role in the regulation of cell growth and survival, and the 26S proteasome is an essential component for degrading 80–90% of dysfunctional proteins and preventing their intracellular accumulation [6, 7]. Tumor cells are characterized by uncontrolled proliferation and rapidly accumulation of abnormal proteins, and timely degradation of these substrates is essential for cancer cell growth and survival. In accordance, aberrant activation of the proteasome has been widely observed in various types of cancers and implicated in the development and progression of carcinogenesis . However, this highly dependence upon proteasome activity makes tumorigenic cells more sensitive to proteasomal inhibition than normal cells, contributing to specific targeting of tumor cells by proteasome inhibitors (PIs) [9, 10]. Bortezomib (PS-341/Velcade) was the first US Food and Drug Administration (FDA) approved PI used in the treatment of newly diagnosed multiple myeloma, relapsed/refractory multiple myeloma, and mantle cell lymphoma . The FDA approval of bortezomib for the treatment of multiple myeloma provided a “proof of concept” that targeting the ubiquitin–proteasome pathway was a viable route for cancer treatment. Accumulating studies have shown that bortezomib was an active antitumor agent in a variety of solid malignancy models, both in vitro and in vivo [12, 13].
Aberrant high proteasomal activity was also found in glioma cells, especially in the group resembling glioma stem cells (GSCs), suggesting that bortezomib could be a potential chemotherapeutic agent for malignant gliomas . Although few previous studies have indicated the killing effect of bortezomib on glioma cells [14, 15], more comprehensive investigations on the chemotherapeutic role of bortezomib in glioma treatment as well as the related molecular mechanism are still in urgent need. As such, in this study, we initially treated U251 and U87 cell lines with different concentrations of bortezomib and determined the consequent alterations in cell viability, proliferation, apoptosis, cell cycle distribution, colony formation, and stem cell characteristics. We then explored the effects of bortezomib on the efficacy of TMZ chemotherapy by testing bortezomib and TMZ combined treatment in both cell lines and glioma xenograft models. In our study, the “FOXM1–Survivin” axis was supposed to be an important target of bortezomib. So, we investigated the regulatory role of bortezomib on the FOXM1–Survivin axis with in vitro cell lines and in vivo xenograft models. In addition, we also measured the expression level of the FOXM1–Survivin axis in clinical glioma samples and analyzed its relation with patient prognosis. We expect that our findings would help, to a certain extent, to clarify the chemotherapeutic effect and underlying mechanism of bortezomib treatment on gliomas.
Materials and methods
Cell lines and cell culture
Human glioblastoma-derived U87, U251, LN229, A172, SF295, and astrocytoma-derived SF268 cell lines were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA, passages 5-15), verified for purity using the ATCC cell line authentication service, and routinely tested for mycoplasma. All cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Gibco) and 1% penicillin/streptomycin (Invitrogen, Carlsbad, CA, USA). All cells were cultured under 37 °C in a humidified atmosphere with 5% CO2.
Clinical glioma samples
Baseline clinical information of the investigated glioma patients
Number of cases
Exponentially growing cell lines were digested and seeded into 96-well plates with 4 × 103 cells/well. After treatment with different concentrations of bortezomib (Selleck Chemicals, Houston, TX, USA), TMZ (Selleck Chemicals), or their combination, cell viability was detected. After adding 20 µL MTT reagent (5 mg/mL) in each well and another 4 h normal culture, the medium was carefully removed, and 100 µL formazan solution was added in each well. The optical density (OD) was measured at 570 nm using an Ultra Multi-functional Microplate Reader (Tecan, Durham, NC, USA). Cell proliferation inhibition rates and survival rates were used to represent the inhibiting effect of different treatments on cell viability, and they were calculated using the following formulae: cell proliferation inhibition rate = 100% × [mean OD value of control group − mean OD value of treatment group]/mean OD value of control group; cell survival rate = 100% × [mean OD value of treatment group/mean OD value of control group]. The 50% inhibitory concentration (IC50) of drug used was calculated with the method of “log(inhibitor) vs. normalized response-Variable slope” using GraphPad Prism 7.0 (GraphPad Software, San Diego, CA, USA). Quantitative analysis of dose–effect relationships and calculation of combination index were performed by CompuSyn (ComboSyn, Inc., Paramus, NJ, USA).
Colony formation assay
Glioma cells were seeded into 6-well culture plate with 200 cells/well and cultured for 10 days. Colonies were washed with cold phosphate buffer saline (PBS) and fixed with 4% paraformaldehyde. Images were taken on a digital microscope (OLYMPUS, Ishikawa, Japan). Those colonies composed of more than 15 cells were counted manually. The number of colonies was represented by the average number from five random fields.
Tumor cell spheroid assay, enrichment of cells with GSC characteristics, and induction of TMZ-insensitive cell lines
Exponentially growing cells were digested and added into a U-bottom 96-well plate at a concentration of 1 × 103 cells/well in 100 μL medium. After centrifuging at 1000×g for 5–10 min, the cells were cultured for another 24 h. The top half medium was carefully replaced with fresh medium containing drug at day 1, and with normal medium at days 4 and 8. Images of spheroids were taken every 2 days. The surface (superficial) area of spheroids on planar images was used to represent the size of real spheroids and was measured using the Image-pro Plus 6.0 (Media Cybernetics, Rockville, MD, USA). The medium for stem cell culture was composed of 20 ng/mL epidermal growth factor, 20 ng/mL basic fibroblast growth factor, 1% N-2 supplement (500×), 1% Glutamax, 0.2% heparin, and 1% penicillin/streptomycin in DMEM/F12ham. After culturing for 24 h with normal medium with or without bortezomib, the cells were digested and seeded into 6-well plates with 2 × 103 cells/well in 1 mL stem cell culture medium. 500 μL fresh stem cell culture medium was added every 3 days. Images were taken every 2 days. To induce TMZ-insensitive U251 and U87 cell lines, U251 and U87 cells were cultured in 10-cm dishes under a 10-day insensitivity-inducing process with normal medium at days 1, 2, 6, and 7, and with medium containing 200 or 500 μmol/L TMZ at days 3, 4, 5, 8, 9, and 10. The process was conducted for at least 3 cycles. Digestion and splitting were conducted when tumors cells reached 100% confluence in one dish.
Flow cytometry detecting cell apoptosis and cell cycle
Cell apoptosis and cell cycle were detected with the Annexin V-FITC Apoptosis Detection Kit (C1062S, Beyotime Biotechnology, Shanghai, China) and the Cell Cycle and Apoptosis Analysis Kit (C1052, Beyotime Biotechnology), performed according to the manufacturer’s instructions . Cell apoptosis and cell cycle were measured and analyzed by a flow cytometry machine (FACS Calibur™, BD Biosciences, San Jose, CA, USA).
The culture medium of 85% confluent 293T cells was replaced with Opti-MEM 2 h before plasmid transfection. Using Lipofectamine 2000, we initially transfected 293T cells with a Lenti-easy packaging mix and FOXM1 overexpression (OE) plasmid (GV270-FOXM1) or control empty vector (EV) (GeneChem, Shanghai, China). After transfection for 8 h, the cell medium was replaced with normal medium and cultured for another 48 h. Lentivirus particles were collected from cell medium by centrifuging (1500 rpm for 5 min) and filtering (0.22 μm) processes.
Stable overexpression and transient knockdown of FOXM1
For overexpressing FOXM1, 50% confluent U251 and U87 cells in 6-well plates were cultured in 1 mL medium with lentivirus particles and 5 μg/mL polybrene (GeneChem). 12 h later, the cell medium was replaced with 2 mL fresh normal medium and cultured for another 48 h. Medium with 2 μg/mL puromycin was used for selecting stably transfected cells, and this puromycin-containing medium was refreshed every 3 days for at least a 9-day selection process. For FOXM1 transient knockdown, 50% confluent U251, U87, and LN229 cells were transfected with FOXM1-short interfering RNA (siRNA) oligonucleotide (CUCUUCUCCCUCAGAU AUAdTdT) or control siRNA (RiboBio, Guangzhou, Guangdong, China). After transfection for 12 h, fresh normal medium was added, and cells were cultured for another 48 h before performing the following experiments.
Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)
Genes’ specific primers used for RT-qPCR
F 5′- GACCCCTTCATTGACCTCAAC-3′
Total protein was extracted with RIPA buffer, and 35–50 μg samples were loaded after measuring their concentration using a Pierce BCA kit (Thermo Fisher Scientific, Waltham, MA, USA) and separated by 7.5% or 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Polyvinylidene difluoride membranes were blocked with 5% fat-free milk for 1 h at room temperature and incubated in primary antibodies (Additional file 1: Table S1) at 4 °C overnight. After washing with tris buffered saline with Tween 200, the membranes were incubated in horseradish peroxidase (HRP)-conjugated anti-mouse or anti-rabbit secondary antibodies (Additional file 1: Table S1) for 1 h at room temperature. Protein bands were detected using the enhanced chemiluminescence system (Thermo Fisher Scientific).
Cellular immunofluorescent staining
Cells cultured in 24-well dishes were washed with cold PBS, fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100 (Sigma-Aldrich, St. Louis., MO, USA) for 10 min at room temperature. After blocking with 10% goat or donkey serum and washing, the cells were incubated with primary rabbit monoclonal antibodies to human FOXM1 and Survivin (Additional file 1: Table S1) at 4 °C overnight. The cells were incubated with Alexa Fluor488-conjugated secondary antibody (Additional file 1: Table S1) for 1 h at room temperature in the dark. Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) for 1 min. Fluorescence was visualized on a fluorescence microscope (FV-1000, OLYMPUS).
Immunohistochemistry (IHC) analysis and scoring
Tissue samples were formalin-fixed and paraffin-embedded. Slides were deparaffinized and rehydrated via successive immersion in the following solutions: 100% xylene I (10 min), 100% xylene II (10 min), 100% ethanol I (5 min), 100% ethanol II (5 min), 95% ethanol (5 min), 85% ethanol (5 min), 80% ethanol (5 min), 75% ethanol (5 min), 80% ethanol (5 min), double distilled water I (10 min), and double-distilled water II (10 min). The slides were then boiled in 0.01 mol/L citrate buffer at 99 °C for 20 min, and endogenous peroxidase activity was blocked with 0.3% H2O2 in methanol for 30 min. Goat serum (Beyotime Biotechnology) was used to block the antibody at room temperature for 10 min. After overnight incubation at 4 °C in primary antibodies (Additional file 1: Table S1), the slides were exposed to HRP-labeled secondary antibodies (Additional file 1: Table S1) for 1 h at room temperature and developed with 3,3′-diaminobenzidine system. Staining intensity was accessed by a designated member of our group using the Image-pro Plus 6.0 (Media Cybernetics) and was represented by the mean density, using the formula mean density = integrated optical density/area of interest. Here, the IHC intensity of each slide was determined by the average “mean density” of at least 3 images from it, and the protein expression was assessed as “weak” (IHC intensity was 0 to 0.15), “moderate” (IHC intensity was 0.16 to 0.25), “strong” (IHC intensity was 0.26 to 0.35), and “very strong” (IHC intensity was above 0.35).
Subcutaneous glioma xenograft model
All experiments involving mice were performed under the ethical criteria of the Third Military Medical University Animal Care and Use Committee, and guidelines for the Care and Use of Laboratory Animals (NIH publications Nos. 80-23, revised 1996) were seriously conducted during the whole process. To establish xenograft model of glioma in mice, 5 × 106 human U87 cells suspended in 80 μL PBS were inoculated subcutaneously into the right hindlimb interior root of BALB/c nude mice (4-week old, female, purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China). About 5–8 days later, the mice bearing tumor around 50 mm3 were selected and randomized into a control group, bortezomib group, TMZ group or bortezomib + TMZ group. Mice in the control group received equivalent drug vehicle (PBS and dimethyl sulfoxide [DMSO]), mice in bortezomib group received 0.25 mg/kg bortezomib every 3 days (intraperitoneal injection [i.p.]), mice in TMZ group received 5 mg/kg TMZ on a 5 days on/2 days off regimen (4 cycles in total, i.p.), and mice in bortezomib + TMZ group received 0.25 mg/kg bortezomib every 3 days and also 5 mg/kg TMZ on a 5 days on/2 days off regimen. Tumor volume was measured every 3 days with a caliper and calculated using the formula tumor volume (mm3) = (length × width2)/2. About 28 days after the first treatment, all mice were euthanized, and the tumor bumps were carefully removed, weighed, and processed for IHC staining.
The SPSS software, version 13.0 (SPSS, Inc., Chicago, IL, USA) and GraphPad Prism 7.0 (GraphPad Software) were used for statistical analysis. Data are presented as the mean ± standard deviation. The nonparametric unpaired t test was utilized to calculate the P value of the difference between 2 independent datasets. One-way analysis of variance (ANOVA) was used to analyze the significance among three or more independent datasets, and the Fisher’s Least Significant Difference method was used for multiple comparisons when the probability for ANOVA was statistically significant. Methods of nonparametric statistics such as the Mann–Whitney and Kruskal–Wallis tests were used when variances did not pass the Levene test for normality or homogeneity. For experiments with over 2 groups and repeated measurements at different time points, their data were processed by the repeated measure two-way ANOVA with Bonferroni post-test. Correlations of protein or mRNA expression between FOXM1 and Survivin were performed using the Pearson R test. The Kaplan–Meier method was used to estimate survival rates. P < 0.05 was considered statistically significant.
Bortezomib inhibited cell viability and induced apoptosis and cell cycle arrest in glioma cells
Cell apoptosis after bortezomib treatment
Cell apoptosis rate (%)
5.78 ± 1.34
2.56 ± 0.44
3.22 ± 0.26
10 nmol/L Bor
14.01 ± 1.33
5.68 ± 1.42
8.33 ± 1.51
20 nmol/L Bor
23.07 ± 3.21
13.49 ± 1.97
9.58 ± 2.50
11.62 ± 2.12
6.02 ± 1.08
5.58 ± 1.86
10 nmol/L Bor
28.03 ± 31.34
21.08 ± 1.52
6.95 ± 0.21
20 nmol/L Bor
41.17 ± 2.55
31.48 ± 2.97
10.23 ± 0.93
Cell cycle alteration after bortezomib treatment
Percentage of cells in different cell cycle phases (%)
43.87 ± 2.77
38.12 ± 3.50
18.01 ± 2.23
54.44 ± 6.32
33.42 ± 3.22
12.15 ± 1.22
70.83 ± 7.54
22.42 ± 2.52
6.74 ± 2.10
45.97 ± 2.99
37.06 ± 2.52
16.97 ± 1.88
53.1 ± 64.3
32.41 ± 2.21
14.43 ± 1.90
76.35 ± 6.54
14.89 ± 1.50
8.76 ± 2.22
Bortezomib inhibited spheroid growth, colony formation, and stemness of glioma cells
Bortezomib down-regulated FOXM1–Survivin axis in glioma cells
To confirm whether Survivin was an important downstream effector of FOXM1, we firstly measured the mRNA levels of FOXM1 and Survivin in six glioma cell lines: U251, U87, LN229, A172, SF295, and SF268. A close correlation between FOXM1 mRNA level and Survivin mRNA level was indicated in them (Pearson r = 0.92, P = 0.01) (Fig. 3c). Further, Survivin mRNA and protein levels were significantly up-regulated by FOXM1 overexpression and markedly reduced by FOXM1 knockdown in U251 and U87 cells (P < 0.05) (Fig. 3d). In accordance with FOXM1 alteration, the mRNA and protein levels, as well as immunofluorescent staining of Survivin were significantly reduced after bortezomib treatment (P < 0.05 or P < 0.01) (Fig. 3e). In summary, bortezomib down-regulated the FOXM1–Survivin axis in U251 and U87 cells, and this might be an important molecular mechanism of its chemotherapeutic effects, alone or in combination with other agents.
Bortezomib sensitized glioma cells to TMZ
Combination index of the temozolomide (TMZ) in combination with different concentrations of bortezomib (Bor)
Cell apoptosis after treatment with bortezomib (Bor) and temozolomide (TMZ)
Cell apoptosis rate (%)
20 nmol/L Bor
11.23 ± 2.22
4.75 ± 0.92
6.48 ± 0.51
200 μmol/L TMZ
11.06 ± 2.63
7.26 ± 1.52
3.80 ± 0.76
Bor + TMZ
40.96 ± 3.88
11.68 ± 0.97
29.28 ± 2.56
20 nmol/L Bor
8.84 ± 1.99
4.06 ± 0.54
4.78 ± 0.88
200 μmol/L TMZ
14.52 ± 1.36
6.68 ± 0.25
7.84 ± 1.92
Bor + TMZ
51.52 ± 3.59
20.33 ± 2.58
31.22 ± 2.18
Cell cycle alteration after treatment with bortezomib (Bor) and temozolomide (TMZ)
Percentage of cells in different cell cycle phases (%)
20 nmol/L Bor
52.5 ± 3.99
35.98 ± 2.96
11.52 ± 1.66
200 μmol/L TMZ
54.45 ± 4.88
32.77 ± 3.25
12.78 ± 1.42
Bor + TMZ
78.31 ± 5.54
13.83 ± 1.58
7.86 ± 1.26
20 nmol/L Bor
45.67 ± 2.99
30.73 ± 2.53
23.63 ± 1.88
200 μmol/L TMZ
66.44 ± 4.33
26.19 ± 2.20
7.37 ± 1.94
Bor + TMZ
81.40 ± 6.51
12.22 ± 1.54
6.42 ± 2.25
FOXM1–Survivin axis was up-regulated in TMZ-insensitive glioma cell lines
Bortezomib inhibited glioma growth and enhanced TMZ efficacy in vivo
The FOXM1–Survivin axis was up-regulated in glioma and related to poor prognosis
Bortezomib, a novel boronic acid dipeptide that inhibits the 26S proteasome activity, has already shown some chemotherapeutic effects against gliomas in vitro and in vivo [18, 19, 20]. Consistent with the findings by Yin et al. , our study demonstrated a significant decrease in cell survival and an increase in cell apoptosis in GBM cell lines treated with bortezomib at a concentration as low as 10 nmol/L (Fig. 1a, c). We also found that a low concentration of bortezomib significantly inhibited spheroid growth, colony formation, and stem-like cell proliferation of U251 and U87 cells (Fig. 2a–c). GSCs are defined as a highly tumorigenic cell subset responsible for tumor progression and drug resistance and is expected to be a critical therapeutic target in various malignancies [22, 23]. Interestingly, recent studies have shown that GSCs were more sensitive to PIs than non-stem differentiated controls or neural stem/precursor cells [24, 25]. Thus, the chemotherapeutic effect of bortezomib on HGG cell lines might result from the specific deletion of GSCs, or at least the GSC sub-population was preferentially inhibited/killed by bortezomib.
In regard to the molecular mechanism underlying the chemotherapeutic effect of bortezomib on glioma cells, generally, it is assumed that the pharmacological inhibition of the proteasome led to toxic accumulation of misfolded and abnormal proteins in cells . For instance, several studies have reported reduced activation of nuclear factor-kappa B after inhibiting IκB degradation by PIs . P53 has been also widely studied in exploring the molecular mechanism of PI-induced tumor apoptosis . Many studies have demonstrated that proteasome inhibition led to growth suppression and apoptosis of tumor cells via inducing overwhelming endoplasmic reticulum stress . Some other possible mechanisms have also been proposed, but the exact mechanism by which PIs exert anti-tumor activity is still poorly understood, especially in glioma. In the present study, we found that bortezomib treatment led to a significant reduction in both mRNA and protein expression of FOXM1 (Fig. 3a), which is very important for tumor malignant behaviors . In support of our findings, similar results were also reported by Andrei L. Gartel and colleagues. They initially found two novel PIs (Siomycin A and thiostrepton), then a serial of conventional PIs that could down-regulate not only the transcriptional activity but also the protein and mRNA levels of FOXM1 [31, 32, 33]. The FOXM1 auto-regulation loop, whereby FOXM1 binds to its own promoter and induces its own transcription , seems to be a feasible mechanism for PI-induced alteration in both protein and mRNA levels. In our present study, another oncogene, the anti-apoptotic factor Survivin, was found to be regulated by FOXM1 in glioma cells (Fig. 3d, e). FOXM1 has been found to directly regulate the transcription of Survivin and the X-linked inhibitor of apoptosis protein in breast cancer cells, providing good support to our findings .
The present study demonstrated that down-regulating the FOXM1–Survivin axis could be an important molecular mechanism for using bortezomib in treating gliomas. However, more details involving bortezomib and FOXM1 interaction await clarification. Bortezomib and other PIs may change the activity of FOXM1 directly or indirectly. In regard to direct action, bortezomib or its metabolic derivatives could attach to the substructure of FOXM1 protein and attenuate or abrogate its function, or intervene in the DNA-binding activity of the FOXM1 domain . As to indirect action, even more possibilities exist. As the activity of FOXM1 protein is strictly regulated by post-translation modification, certain kinds of modification may make the protein lost its initial function and be a “disturber” of normal protein activity. The principle biological function of PIs is repressing the function of the proteasome system . Bortezomib might cause accumulation of this “disturbing” FOXM1 and subsequently attenuated DNA-binding and transcriptional activity of normal FOXM1 . In agreement to this “inhibiting” hypothesis, Andrei L. Gartel  proposed the “negative regulator of FOXM1 (NRFM)” theory: the expression of a putative NRFM is accumulated after treatment with PIs then, this NRFM binds to FOXM1 and inhibits its transcriptional activity on the promoter of target genes, including FOXM1. Although several tumor suppressors, such as heat shock protein 70 and sterile alpha motif-pointed domain-containing E26 transformation-specific transcription factor, were found to suppress FOXM1 auto-regulation by inhibiting the activity of FOXM1 binding to its own gene promoter [39, 40], “NRFM” remains to be verified. Contrary to the above discussed “accumulation” effects by inhibiting proteasome activity, bortezomib may also accelerate the “reduction” of proteins by inducing autophagy/lysosomal degradation, which was reported in various tumor cells, including glioma cell lines [41, 42, 43, 44]. Interestingly, the degradation of FOXM1 partially depends on lysosome under normal conditions . Thus, bortezomib might reduce FOXM1 by enhancing lysosomal degradation of its protein. Another less possible mechanism is the down-regulation of “positive regulator of FOXM1”, such as transcription factors for FOXM1. Again, not much has been revealed about this hypothesis.
As summarized in a review by Hanahan et al. , FOXM1 is implicated in most hallmarks of cancer. In regard to glioma, FOXM1, as an oncogenic transcription factor, also plays important roles in glioma progression and maintenance of GSC characteristics [47, 48, 49]. Studies have also reported that FOXM1 promoted glioma resistance to TMZ chemotherapy via regulating DNA damage repair gene Rad51 and replication factor C5 [50, 51]. As one of the classic anti-apoptotic proteins, Survivin plays important role in glioma progression, recurrence, and chemoradiotherapy resistance [52, 53, 54]. In our TMZ-insensitive U251 and U87 glioma cell lines, the expression levels of FOXM1 and Survivin were markedly up-regulated (Fig. 5a, b). This might have resulted from the enrichment of clones with high levels of FOXM1 and Survivin, while clones with lower FOXM1 and Survivin expression were deleted by continuous TMZ screening. Thus, a highly activated FOXM1–Survivin axis could greatly promote glioma cell proliferation and malignant transformation, and enhance resistance and survival of glioma cells under stress.
However, the high dependence on the FOXM1–Survivin axis may make glioma cells very susceptible to agents specifically targeted to this oncogenic axis, such as bortezomib. As such, bortezomib can significantly enhance the sensitivity of glioma cells to TMZ treatment both in vitro (Fig. 4a–d) and in vivo (Fig. 6a–c). Interestingly, via analyzing clinical samples from our department and public datasets from TCGA , we found that FOXM1 and Survivin were widely overexpressed in glioma patients (Fig. 7a, b, d). On one hand, extensive up-regulation of the FOXM1–Survivin axis in HGG patients might be the basis of possible clinical use of bortezomib in the future; on the other hand, the activation level of the FOXM1–Survivin axis in certain patients could be a good predictor for their sensitivity to bortezomib, alone or in combination with TMZ. However, several problems have to be tackled before the use of bortezomib and other PIs in glioma patients, such as the drug ability to penetrate the blood–brain barrier, unspecific targeting, and dose limited toxicity of drug [55, 56]. Further investigation is needed to further enhance the clinical use of bortezomib and to develop next-generation PIs with more satisfying clinical efficacy with advanced sensitivity, specificity, and safety [57, 58].
In summary, our findings showed that PI bortezomib had a sensitive chemotherapeutic effect on glioma cells. Low concentrations of bortezomib significantly inhibited the proliferation, spheroid growth, colony formation, and stem cell characteristics of glioma cells by inducing apoptosis and cell cycle arrest. Bortezomib also demonstrated a synergistic effect with TMZ and sensitized glioma to TMZ treatment in vitro and in vivo. Mechanistically, bortezomib down-regulated the FOXM1–Survivin axis, which was also found to be up-regulated in glioma patients and was related to poor prognosis. Our findings provide an important research basis for instigating further investigation on bortezomib or other PIs in glioma therapy.
We highly appreciate the help from Mrs. Rong Xin and Miss Lu Jiang (Central Laboratory of Xinqiao Hospital, Third Military Medical University) with paraffin-embedded sections and IHC staining.
SQL, ZZ, and JXC have contributed to the conception and design of this research; JHT, LY, QFX, QRL, and LRZ have conducted experiments involving in vitro cells and in vivo mice models and clinical samples; JHT, GHH, ZXZ, YX, and LD participated in data interpretation, statistical analysis, and literature research; All authors have participated in the writing and revision of this manuscript. All authors agreed to be accountable for all aspects of this research work. All authors read and approved the final manuscript.
This work was primarily supported by Grants from the National Natural Science Foundation of China (NSFC-81972360).
Ethics approval and consent to participate
This study was reviewed and approved by the Ethics Committee of the Third Military Medical University. Informed consents were obtained from all patients or his/her guardians. All experiments involving mice were performed under the rules of Third Military Medical University Animal Care and Use Committee, and guidelines for the Care and Use of Laboratory Animals (NIH publications Nos. 80-23, revised 1996) were seriously conducted during the whole process.
Consent for publication
Consents for publication were obtained from all patients or his/her guardians.
The authors declare that they have no competing interests.
- 3.Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJB, Janzer RC, et al. Effects of radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet Oncol. 2009;10:459–66.PubMedCrossRefGoogle Scholar
- 14.Bota DA, Alexandru D, Keir ST, Bigner D, Vredenburgh J, Friedman HS. Proteasome inhibition with bortezomib induces cell death in GBM stem-like cells and temozolomide-resistant glioma cell lines, but stimulates GBM stem-like cells’ VEGF production and angiogenesis. J Neurosurg. 2013;119:1415–23.PubMedPubMedCentralCrossRefGoogle Scholar
- 16.The Cancer Genome Atlas (TCGA). The TCGA Research Network, National Cancer Institute at the National Institutes of Health. 2019. https://www.cancer.gov/tcga. Accessed 21 May 2019.
- 20.Kong XT, Nguyen NT, Choi YJ, Zhang G, Nguyen HN, Filka E, et al. Phase 2 study of bortezomib combined with temozolomide and regional radiation therapy for upfront treatment of patients with newly diagnosed glioblastoma multiforme: safety and efficacy assessment. Int J Radiat Oncol Biol Phys. 2018;100(5):1195–203.PubMedPubMedCentralCrossRefGoogle Scholar
- 27.Qureshi AA, Zuvanich EG, Khan DA, Mushtaq S, Silswal N, Qureshi N. Proteasome inhibitors modulate anticancer and anti-proliferative properties via NF-kB signaling, and ubiquitin—proteasome pathways in cancer cell lines of different organs. Lipids Health Dis. 2018;17:62.PubMedPubMedCentralCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.