LABORATORY INVESTIGATION

J Neurosurg 134:701–710, 2021

Riluzole enhances the antitumor effects of temozolomide

via suppression of MGMT expression in glioblastoma
Tetsuya Yamada, MD,1,2 Shohei Tsuji, BS,1 Shinsuke Nakamura, PhD,1 Yusuke Egashira, MD, PhD,2
Masamitsu Shimazawa, PhD,1 Noriyuki Nakayama, MD, PhD,2 Hirohito Yano, MD, PhD,2
Toru Iwama, MD, PhD,2 and Hideaki Hara, PhD1
Molecular Pharmacology, Department of Biofunctional Evaluation, Gifu Pharmaceutical University; and 2Department of
1

Neurosurgery, Gifu University Graduate School of Medicine, Gifu, Japan

OBJECTIVE  Glutamatergic signaling significantly promotes proliferation, migration, and invasion in glioblastoma (GBM).
Riluzole, a metabotropic glutamate receptor 1 inhibitor, reportedly suppresses GBM growth. However, the effects of
combining riluzole with the primary GBM chemotherapeutic agent, temozolomide (TMZ), are unknown. This study aimed
to investigate the efficacy of combinatorial therapy with TMZ/riluzole for GBM in vitro and in vivo.
METHODS  Three GBM cell lines, T98G (human; O6 -methylguanine DNA methyltransferase [MGMT] positive), U87MG
(human; MGMT negative), and GL261 (murine; MGMT positive), were treated with TMZ, riluzole, or a combination of
both. The authors performed cell viability assays, followed by isobologram analysis, to evaluate the effects of combi-
natorial treatment for each GBM cell line. They tested the effect of riluzole on MGMT, a DNA repair enzyme causing
chemoresistance to TMZ, through quantitative real-time reverse transcription polymerase chain reaction in T98G cells.
Furthermore, they evaluated the efficacy of combinatorial TMZ/riluzole treatment in an orthotopic mouse allograft model
of MGMT-positive GBM using C57BL/6 J mice and GL261 cells.
RESULTS  Riluzole displayed significant time- and dose-dependent growth-inhibitory effects on all GBM cell lines as-
sessed independently. Riluzole enhanced the antitumor effect of TMZ synergistically in MGMT-positive but not in MGMT-
negative GBM cell lines. Riluzole singularly suppressed MGMT expression, and it significantly suppressed TMZ-induced
MGMT upregulation (p < 0.01). Furthermore, combinatorial TMZ/riluzole treatment significantly suppressed tumor growth
in the intracranial MGMT-positive GBM model (p < 0.05).
CONCLUSIONS  Riluzole attenuates TMZ-induced MGMT upregulation and enhances the antitumor effect of TMZ in
MGMT-positive GBMs. Therefore, combinatorial TMZ/riluzole treatment is a potentially promising novel therapeutic regi-
men for MGMT-positive GBMs.
https://thejns.org/doi/abs/10.3171/2019.12.JNS192682
KEYWORDS  glioblastoma; MGMT; riluzole; temozolomide; combinatorial therapy; oncology

G
lioblastoma (GBM) is the most aggressive and ating agents including TMZ frequently occurs owing to
one of the most common tumors of the central the effect of O6-methylguanine DNA methyltransferase
nervous system. Despite current combinatorial (MGMT), a DNA repair enzyme upregulated in cancers.
treatment modalities, including resection, radiation ther- MGMT inhibits the effects of TMZ by eliminating the
apy, and chemotherapy, the prognosis for patients with TMZ-induced alkylation adducts from the O6 position in
GBM remains poor owing to the tumor’s heterogeneity, guanine.18 MGMT-positive GBM cells, which are GBM
high proliferation, and invasiveness.29 Temozolomide cells with unmethylated MGMT promoters, exhibit che-
(TMZ), an alkylating agent capable of crossing the blood- moresistance to TMZ through this repair mechanism.
brain barrier, is almost the only chemotherapeutic agent This highlights the need for additional effective chemo-
used to treat GBM and has been reportedly highly effec- therapeutics for GBM that are potentially applicable for
tive in some cases.12 However, chemoresistance to alkyl- increasing the efficacy of TMZ, perhaps by targeting

ABBREVIATIONS  CCK-8 = Cell Counting Kit-8; CI = combination index; DMEM = Dulbecco’s modified Eagle medium; DMSO = dimethyl sulfoxide; FBS = fetal bovine
serum; GBM = glioblastoma; IC20 = 20% inhibitory concentration; IFN-β = interferon-β; MGMT = O6-methylguanine DNA methyltransferase; RT-PCR = reverse transcription
polymerase chain reaction; TMZ = temozolomide.
SUBMITTED  October 1, 2019.  ACCEPTED  December 30, 2019.
INCLUDE WHEN CITING  Published online March 13, 2020; DOI: 10.3171/2019.12.JNS192682.

©AANS 2021, except where prohibited by US copyright law J Neurosurg  Volume 134 • March 2021 701

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

Yamada et al.

MGMT. Concurrent administration of the small nucleo- well) with culture medium supplemented with 10% FBS.
side inhibitor, O6-benzylguanine, and TMZ was previous- After 24 hours of incubation, the medium was changed
ly reported in a clinical trial to restore chemosensitivity in to DMEM supplemented with 10% FBS and the drugs in
patients with recurrent TMZ-resistant malignant glioma accordance with the following 4 conditional groups: no-
by mimicking the target of MGMT.23 Unfortunately, how- treatment control (dimethyl sulfoxide [DMSO]), TMZ (10,
ever, TMZ/O6-benzylguanine combination therapy was 20, 30, 40, 50, 60, 100, 150, 200, 300, 400, 500, or 600
also found to cause severe myelosuppression.23 Despite μM), riluzole (1, 2, 3, 4, 5, 10, 15, 20, 25, 30, or 35 μM),
several clinical trials assessing the impact of TMZ com- and the combinatorial treatment (TMZ and riluzole). Af-
bined with other drugs, none of the combinations tested ter 24, 48, or 72 hours of treatment, CCK-8 assays were
so far have prolonged the overall survival of GBM pa- performed to determine cell viabilities. The absorbance of
tients in comparison with TMZ administration alone.7,13 each well was measured at 450 nm with a reference wave-
Glutamate is one of the major excitatory neurotransmit- length of 660 nm using a microplate reader.
ters regulating neuronal and synaptic activity, and it plays
an important role in activating proliferation, migration, Isobologram Analysis
and invasion among GBM cells.10 Glutamate activates not To determine whether riluzole has a synergistic or ad-
only the ionotropic glutamate receptors (iGluRs), such as ditive effect with TMZ on each cell line, we conducted
the α-amino-3-hydroxy-5-methyl-4-isoxazole propionate isobologram analysis. We selected a 20% inhibitory con-
receptors (AMPARs), N-methyl-D-aspartate receptors centration (IC20) as the reference concentration, which is
(NMDARs), and kainic acid receptors (KARs), but also defined as the drug concentration reducing the cell viabil-
the G-protein-coupled metabotropic glutamate receptors ity of a treated group to 20% of the level of the no-treat-
(mGluRs), a family containing eight subtypes (mGluR1 ment control group.
to mGluR8).36 Riluzole is a selective mGluR1 antagonist
Based on the results of the cell viability assay, the IC20
of the benzothiazole class, approved by the U.S. FDA for
of TMZ and riluzole was determined for each cell line.
treating amyotrophic lateral sclerosis.28,36 Moreover, rilu-
We evaluated the growth-inhibitory effects upon combi-
zole reportedly inhibits glutamate release via inactivation
of voltage-dependent sodium channels and serves as a natorial treatment in comparison with the control group
noncompetitive antagonist of NMDARs.11 Recent studies by measuring cell viabilities after 72 hours of treatment.
have reported that riluzole inhibits GBM progression by Only the concentration combinations exhibiting 20%
suppressing the PI3K/AKT/mTOR pathway.36 Although growth-inhibitory effects were plotted on isobolograms.
riluzole monotherapy reportedly did not improve the sur- On isobologram graphs, the points plotted below, on, or
vival of rats transplanted with GBM, it enhanced their sen- above the theoretical line of additivity indicate synergism,
sitivity to radiotherapy.33 However, no studies have investi- additivity, or antagonism of the drug combination, respec-
gated the efficacy of combinatorial therapy with TMZ and tively.1 The combination index (CI) was determined using
riluzole for GBM, nor has there been any evidence regard- the following equation: CI = [C]T20/IC20T + [C]R20/IC20R.
ing the effectiveness of riluzole against MGMT-induced [C]T20 and [C]R20 represent the concentrations of TMZ
chemoresistance. and riluzole, respectively, that exhibit 20% inhibition on
This study aimed to evaluate the effect of riluzole on cell growth upon combinatorial treatment. CI < 1, CI = 1,
MGMT expression and assess the effectiveness of combi- and CI > 1 indicate synergism, additivity, and antagonism
natorial treatment with TMZ and riluzole against GBM in of the drug combination, respectively.1
vitro and in vivo.
Quantitative Real-Time Reverse Transcription Polymerase
Chain Reaction Analysis
Methods To evaluate the effect of riluzole on the expression of
Cell Culture MGMT mRNA expression, we performed quantitative
We used the human GBM cell lines T98G and U87MG real-time reverse transcription polymerase chain reaction
and one murine GBM cell line, GL261. T98G cells were (RT-PCR) analysis. T98G cells were seeded in a 12-well
purchased from the JCRB Cell Bank, and U87MG cells plate (2.0 × 104 cells/well) with DMEM supplemented with
were from the American Type Culture Collection. GL261 10% FBS. After 24 hours of incubation, the medium was
cells were kindly provided by Dr. Masanao Saio (Gunma changed, and GBM cells were cultured under the follow-
University, Japan). ing conditions for 48 hours: no-treatment control (DMSO),
As previously described,21 T98G, U87MG, and GL261 TMZ (300 μM), riluzole (1, 5, or 10 μM), or a combination
cells were cultured in Dulbecco’s modified Eagle medium of TMZ (300 μM) and riluzole (5 μM). After 48 hours
(DMEM) with low glucose (Nacalai Tesque), supplement- of treatment, RNA was isolated from T98G cells using
ed with 10% fetal bovine serum (FBS; Valeant), 100 U/mL Nucleo Spin RNA II (Takara). RNA concentrations were
penicillin, and 100 μg/ml streptomycin in 10-cm dishes determined using NanoVue Plus (GE Healthcare Japan).
(BD Biosciences) at 37°C in 5% CO2. Single-strand cDNAs were synthesized from the isolated
RNAs via reverse transcription with a PrimeScript RT Re-
Cell Viability Assay agent Kit (Perfect Real Time; Takara). Quantitative real-
Cell Counting Kit-8 (CCK-8) cell viability assays were time RT-PCR was performed using TB Green Premix Ex
conducted as previously described.21 T98G, U87MG, and Taq II (Tli RNaseH Plus; Takara) and a TP800 Thermal
GL261 cells were seeded in a 96-well plate (2.0 × 103 cells/ Cycler Dice Real Time System (Takara). All aforemen-

702 J Neurosurg  Volume 134 • March 2021

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

 Yamada et al.

tioned procedures were carried out in accordance with the ability assays were conducted on two MGMT-positive cell
manufacturer’s instructions. A human MGMT quantita- lines (T98G and GL261) and one MGMT-negative cell
tive PCR primer pair was purchased from Sino Biologi- line (U87MG).6,37 When administered alone, riluzole (5,
cal Inc. The PCR primer sequences for GAPDH (internal 10, or 20 μM) had significant time- and dose-dependent
control) were as follows: 5′-TGT GTC CGT CGT GGA inhibitory effects on the proliferation of all GBM cell lines
TCT GA-3′ (forward) and 5′-TTG CTG TTG AAG TCG tested herein (Fig. 1A–C). Cell viabilities were significant-
CAG GAG-3′ (reverse). The cycling conditions were in ac- ly lower in the treatment groups than in the no-treatment
cordance with the manufacturer’s protocol. The results are control group at the following riluzole concentrations and
expressed as relative gene expression levels normalized to time points in each cell line: T98G cells after 48 and 72
that of GAPDH. hours of treatment with 20 μM (p < 0.01); U87MG cells
after 48 hours of treatment with 20 μM (p < 0.01) and after
Establishment of the Orthotopic Mouse Allograft Model of 72 hours of treatment with 5, 10, and 20 μM (p < 0.01); and
MGMT-Positive GBM GL261 cells after 24, 48, and 72 hours treatment with 20
All animal procedures were conducted in accordance μM (p < 0.01) and 48 hours of treatment with 5 μM (p <
with the ARRIVE (Animal Research: Reporting of In 0.05).
Vivo Experiments) guidelines and were approved by the
animal experiment committee of Gifu Pharmaceutical Riluzole Sensitizes MGMT-Positive GBM Cells to TMZ
University, Japan. We treated two MGMT-positive (unmethylated MGMT
Eight-week-old male C57BL/6 J mice (Japan SLC, promoter) cell lines (T98G and GL261) and one MGMT-
Ltd.) were used herein. All mice were housed at 24°C un- negative (methylated MGMT promoter) cell line (U87MG)
der a 12-hour light-dark cycle. On day 0, the mice received with TMZ, riluzole, or a combination of the two for 72
orthotopic intracranial injections of 1.0 × 105 GL261 cells hours and evaluated cell proliferation via CCK-8 cell vi-
in 2 μL of PBS using a Hamilton microliter syringe placed ability assays (Fig. 2A–C). After 72 hours, TMZ exerted
in the left striatum at previously described coordinates21: significant dose-dependent inhibitory effects on the prolif-
1 mm anterior, 2 mm left lateral to the bregma, and 3 mm eration of all GBM cell lines tested herein. Cell viabilities
below the cranial surface. The mice were randomly di- were significantly decreased at 400 μM in T98G cells (p
vided into the no-treatment control group (n = 12) and < 0.01), 20 μM in U87MG cells (p < 0.01), and 50 μM in
treatment groups. The latter received TMZ on days 14–18 GL261 cells (p < 0.05), compared with the no-treatment
(50 mg/kg/day) (n = 12), or a combination of riluzole on control group. As described in Fig. 1, riluzole exerted sig-
days 13–18 (15 mg/kg/day) and TMZ on days 14–18 (50 nificant dose-dependent inhibitory effects on all GBM cell
mg/kg/day) (n = 12). All drugs were orally administered. lines tested herein after 72 hours of treatment. Cell viabili-
The dosages of TMZ and riluzole were based on previous ties were significantly decreased at 10 μM in T98G cells
reports.16,30,33 The mice were monitored daily during the (p < 0.05), 5 μM in U87MG cells (p < 0.05), and 15 μM
experiment, and those displaying severe neurological defi- in GL261 cells (p < 0.01), compared with the no-treatment
cits, severe pain, or anorexic symptoms were euthanized control. The effects of combinatorial therapy of TMZ and
with pentobarbital.
riluzole on GBM cell lines were evaluated at drug concen-
On day 20, 10 mice in each group were deeply anes-
thetized and transcardially perfused with 4% paraformal- trations largely ineffective when administered alone. The
dehyde (Wako). The brains were dissected out, fixed in results for each cell line are shown in Fig. 2 (A: T98G,
4% paraformaldehyde, and embedded in paraffin (Leica B: U87MG, C: GL261). In T98G human MGMT-positive
Biosystems). To determine the maximum cross-sectional cells, the combination of 300 μM TMZ and 5 μM riluzole
areas and volumes of the tumors, samples were cut into demonstrated a significant inhibitory effect on cell prolif-
5-μm-thick coronal sections, stained with H & E, and eration (p < 0.01; Fig. 2A). Furthermore, in GL261 murine
photographed using an All-in-One Fluorescence Micro- MGMT-positive cells, the combination of 40 μM TMZ
scope (BZ-X710; Keyence). and 5 μM riluzole demonstrated a significant inhibitory
effect (p < 0.01; Fig. 2C). In contrast, in U87MG human
Statistical Analysis MGMT-negative cells, the combination of 10 μM TMZ
and 3 μM riluzole did not significantly inhibit cell prolif-
Each in vitro experiment was performed in triplicate. eration (Fig. 2B).
Data are presented as mean ± SEM. Statistical analy- Thereafter, we determined whether treatment with 5
ses were performed using one-way ANOVA, followed μM riluzole enhanced the cytostatic effect of TMZ at oth-
by Dunnett’s test or Tukey’s test for post hoc analysis. er concentrations in T98G (Fig. 2D) and GL261 (Fig. 2E)
Thompson’s rejection tests were performed for individual cells. Both these cell lines were significantly sensitized to
values to eliminate outliers. p values < 0.05 were consid- TMZ in the presence of 5 μM riluzole at all TMZ concen-
ered statistically significant. All tests were conducted us- trations tested.
ing the statistical software JMP 13 (SAS Institute Inc.).
Riluzole Acts Synergistically With TMZ Against
Results MGMT-Positive GBM Cells
Riluzole Independently Inhibits GBM Cell Proliferation Further, we investigated whether riluzole treatment
We first evaluated the effects of riluzole on the growth synergistically enhances the cytostatic effect of TMZ on
of GBM cell lines in monolayer cultures. CCK-8 cell vi- each GBM cell line. Prior to this analysis, we preliminar-

J Neurosurg  Volume 134 • March 2021 703

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

Yamada et al.

FIG. 1. The effect of riluzole on GBM cell growth. CCK-8 cell viability assays were performed on two MGMT-positive GBM cell
lines (T98G, A, and GL261, C) and one MGMT-negative GBM cell line (U87MG, B) after 24, 48, and 72 hours (h) of treatment
with riluzole at 3 different concentrations (5, 10, and 20 μM). Each column and bar represents the mean ± SEM (n = 6). *p < 0.05
versus no treatment (NT); **p < 0.01 versus NT (one-way ANOVA followed by Dunnett’s test).

ily determined the IC20 of TMZ (Fig. 3A) and riluzole GL261 cells, the strongest synergism was observed with
(Fig. 3B) for each GBM cell line, using CCK-8 viability a combination of 34 μM TMZ and 6 μM riluzole, with a
assays. In T98G, U87MG, and GL261 cells, the IC20 for CI of 0.66. In U87MG cells, the lowest CI was 0.97, which
TMZ was 481 μM, 75 μM, and 170 μM, whereas the IC20 was observed with a combination of 45 μM TMZ and 7
for riluzole was 16 μM, 19 μM, and 13 μM, respectively. μM riluzole, whereas the other combinations displayed
We then conducted isobologram analysis to determine antagonism.
whether the effects of combinatorial treatment with TMZ
and riluzole were synergistic or additive (Fig. 3C–E). For
T98G and GL261 cells, the isobologram analysis demon- Riluzole Inhibits MGMT mRNA Expression in GBM Cells in
strated clear synergism at all drug concentrations tested a Dose-Dependent Manner
(Fig. 3C and E). In contrast, combinatorial treatment dis- To evaluate the effect of riluzole on MGMT mRNA
played additive or even antagonistic effects in U87MG levels, we performed quantitative real-time RT-PCR anal-
cells (Fig. 3D). The concentration combinations used to ysis. In T98G MGMT-positive GBM cells, riluzole down-
construct isobolograms and determine the CIs are shown regulated MGMT mRNA in a dose-dependent manner.
in Table 1. In T98G cells, the strongest synergism be- MGMT mRNA levels differed significantly between the
tween TMZ and riluzole was observed with the combina- nontreated control and 10-μM riluzole-treated cells after
tion of 96 μM TMZ and 8 μM riluzole, with a CI of 0.7. In 48 hours (p < 0.05; Fig. 4, left).

704 J Neurosurg  Volume 134 • March 2021

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

 Yamada et al.

FIG. 2. The effect of combining riluzole with TMZ on the growth of the GBM cell lines T98G, U87MG, and GL261. CCK-8 cell
viability assays were performed for two MGMT-positive GBM cell lines (T98G [A] and GL261 [C]) and one MGMT-negative GBM
cell line (U87MG [B]) after 72 hours of treatment with TMZ, riluzole, or a combination of the two at concentrations that were not
significantly effective during monotherapy. Each column and bar represents the mean ± SEM (n = 6). *p < 0.05 versus NT; **p <
0.01 versus NT. In two MGMT-positive GBM cell lines (T98G [D] and GL261 [E]), the efficacy of riluzole (5 μM) combined with 4
different concentrations of TMZ was evaluated using CCK-8 assays after 72 hours of treatment. Each column and bar represents
the mean ± SEM (n = 6). **p < 0.01 versus TMZ alone (Student t-test). n.s. = not significant (one-way ANOVA followed by Dun-
nett’s test).

Riluzole Inhibits TMZ-Induced MGMT Expression in GBM dose of maintenance therapy for GBM patients (150 mg/
Cells m2). The mean maximum cross-sectional area was 12.69 ±
Quantitative real-time RT-PCR analysis revealed signif- 2.47 mm2 in control mice, 8.23 ± 1.68 mm2 in mice treat-
icant upregulation of MGMT mRNA in T98G cells upon ed with 50 mg/kg of TMZ, and 5.30 ± 0.65 mm2 in mice
treatment with 300 μM TMZ for 48 hours (p < 0.001; Fig. treated with 50 mg/kg of TMZ combined with 15 mg/kg
4, right). We then investigated whether riluzole treatment of riluzole (Fig. 5B and C). The mean tumor volumes were
decreases TMZ-induced upregulation of MGMT. Combi- 30.02 ± 6.99 mm3, 15.55 ± 4.17 mm3, and 9.84 ± 1.91 mm3
natorial treatment with 300 μM TMZ and 5 μM riluzole for control, TMZ-treated, and combination-treated mice,
for 48 hours significantly downregulated MGMT mRNA respectively (Fig. 5D). TMZ monotherapy did not signifi-
in comparison with treatment with 300 μM TMZ alone (p cantly decrease the mean maximum cross-sectional area
< 0.01; Fig. 4, right). of the tumors or the mean tumor volumes in comparison
with the control group. However, combinatorial thera-
Riluzole Enhances the Inhibitory Effect of TMZ in an py with both TMZ and riluzole significantly suppressed
MGMT-Positive Mouse Model of Intracranial GBM not only the mean maximum cross-sectional area of the
We generated a mouse model of MGMT-positive GBM tumors, but also the mean tumor volumes in comparison
by injecting GL261 cells into the left striatum of adult mice. with the control group (p < 0.05). The mean maximum
The protocol for drug administration is illustrated in Fig. cross-sectional area of the tumors and the mean tumor
5A. In this study, two mice in each group were euthanized volumes were slightly but not significantly lower in mice
because of technical errors during drug administration or treated with combinatorial therapy than in those treated
symptoms of anorexia following drug administration. To with TMZ monotherapy.
evaluate the effects of the drug therapies, we measured the
maximum cross-sectional area and volume of the intra- Discussion
cranial tumors after treatment (Fig. 5B–D). The 50-mg/kg The present study investigated the effectiveness of
dose of TMZ was selected because it is equivalent to the combinatorial treatment with riluzole and conventional

J Neurosurg  Volume 134 • March 2021 705

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

Yamada et al.

FIG. 3. IC20 evaluation and isobologram analyses of the effects of different combinations of TMZ and riluzole on the GBM cell lines
T98G, U87MG, and GL261. A and B: The 20% inhibitory concentration (IC20) values for TMZ and riluzole were determined in the
T98G, U87MG, and GL261 cell lines through CCK-8 cell viability assays. Herein, the IC20 is defined as the drug concentration that
reduces cell viability to 20% of the no-treatment control cells after 72 hours. In the T98G, U87MG, and GL261 cell lines, the IC20
for TMZ was 481 μM, 75 μM, and 170 μM, and the IC20 for riluzole was 16 μM, 19 μM, and 13 μM, respectively. C–E: Isobologram
analyses of the effects of the combined TMZ and riluzole were conducted for T98G (C), U87MG (D), and GL261 (E) cell lines.
The “theoretical line of additivity,” which connects the IC20 values of the two drugs used in combination therapy, separates the x–y
plane into two parts. The points (dots) plotted below, on, or above the line indicate synergism, additivity, or antagonism of the drug
combination, respectively. In the T98G (C) and GL261 (E) cell lines, the isobolograms indicated clear synergism at all concentra-
tions tested. In the U87MG cell line (D), the isobologram indicated additivity or antagonism at all concentrations tested.

TMZ in terms of GBM cell growth in vitro and in vivo. during monotherapy significantly enhanced the effect of
The results indicate the following three novel aspects: 1) TMZ at various clinical concentrations. Furthermore,
riluzole synergistically enhanced the antitumor effect of isobolograms obtained through combinatorial treatment
TMZ in MGMT-positive GBM cells; 2) mechanistically, in MGMT-positive GBM cells displayed clear synergism
riluzole directly suppressed MGMT expression and sup- between TMZ and riluzole. These results clearly indicate
pressed TMZ-induced MGMT upregulation; and 3) syn- that riluzole synergistically enhances the antitumor effect
ergistic effects of riluzole were observed in a mouse al- of TMZ in MGMT-positive GBM cells.
lograft model of MGMT-positive GBM. In MGMT-positive GBM cells, treatment with TMZ
Glutamatergic systems potentially play important roles upregulates MGMT.17 Accordingly, MGMT is a potential-
not only in several neurological disorders, including epi- ly important therapeutic target for GBM therapy. Recent
lepsy,2 Alzheimer’s disease,32 Parkinson’s disease,4 and studies have reported that several agents suppress MGMT
amyotrophic lateral sclerosis8 but also in tumors such as in GBM cells via various pathways, such as the Wnt/β
melanoma and GBM.16,25 Recent studies have reported that catenin, Hedgehog/GLI1, and PI3K/AKT/NFκB signaling
inhibition of the glutamatergic pathway may be effective pathways.3,17,19 To our knowledge, previous studies have
for GBM treatment. Riluzole, a selective mGluR1 antago- reported that treatment with interferon-β (IFN-β), leveti-
nist, inhibits activation of the PI3K/AKT/mTOR pathway racetam, valproic acid, bortezomib, and disulfiram down-
and induces the death of GBM cells.36 However, to our regulate MGMT in GBM cells.5,20,22,24,31 In contrast, anoth-
knowledge, no studies have investigated the effective- er study reported that valproic acid upregulates MGMT.5
ness of combinatorial treatment of TMZ and riluzole. In Antagonists of the glutamatergic pathway, including rilu-
the present study, cell viability assays revealed significant zole, have not previously been reported to inhibit MGMT
synergistic effects between TMZ and riluzole in MGMT- expression in GBM cells. This study is the first to report
positive GBM cell lines but not in the MGMT-negative that riluzole monotherapy downregulates MGMT mRNA
GBM cell line. Indeed, in the MGMT-positive cell lines, in a dose-dependent manner in T98G MGMT-positive
treatment with riluzole at an ineffective concentration GBM cells. As shown in Fig. 4 right, TMZ monotherapy
706 J Neurosurg  Volume 134 • March 2021

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

 Yamada et al.

TABLE 1. Combination index analysis

Cell Lines TMZ (μM) Riluzole (μM) TMZ IC20 Equivalent Riluzole IC20 Equivalent Combination Index Judgment
481 0 1.0000 0.0000
289 4 0.6000 0.2500 0.85 Synergism
T98G: MGMT+ 192 7 0.4000 0.4375 0.84 Synergism
96 8 0.2000 0.5000 0.70 Synergism
0 16 0.0000 1.0000
75 0 1.0000 0.0000
60 6 0.8000 0.3157 1.12 Antagonism
45 7 0.6000 0.3684 0.97 Additivity
U87MG: MGMT−
30 14 0.4000 0.7368 1.14 Antagonism
15 18 0.2000 0.9474 1.15 Antagonism
0 19 0.0000 1.0000
170 0 1.0000 0.0000
136 0.5 0.8000 0.0385 0.84 Synergism
102 2 0.6000 0.1538 0.75 Synergism
GL261: MGMT+
68 4 0.4000 0.3077 0.71 Synergism
34 6 0.2000 0.4615 0.66 Synergism
0 13 0.0000 1.0000

significantly upregulated MGMT in T98G GBM cells; of GBM patients receiving combinatorial treatment with
MGMT was significantly downregulated when cells were TMZ. Thus far, several agents have reportedly exerted
treated with riluzol. These results suggest that combinato- combined effects with TMZ in experimental animal mod-
rial treatment with riluzole and TMZ potentially suppress- els.9,14,35 To our knowledge, IFN-α/β is the only agent that
es TMZ-induced MGMT upregulation in MGMT-positive enhances the effect of TMZ via suppression of MGMT
GBM. Moreover, the concentrations of riluzole used for in a mouse model of intracranial GBM.20,26 In the present
in vitro combinatorial treatments were within the range study, we generated a mouse model with a normal immune
of plasma concentrations of healthy volunteers in clinical system using the GL261 MGMT-positive mouse GBM cell
trials.15 line. Compared with control mice, TMZ monotherapy
A number of clinical trials have investigated the effica- did not significantly suppress tumor size in terms of both
cy of several combinatorial therapies with TMZ for GBM maximum cross-sectional area and volume. In contrast,
treatment, including bevacizumab and IFN-β.7,13 However, TMZ combined with riluzole significantly suppressed tu-
no agents have reportedly prolonged the overall survival mor growth in comparison with that seen in control mice.

FIG. 4. The effects of riluzole on MGMT mRNA levels in the MGMT-positive GBM cell line T98G. Left: Quantitative RT-PCR
(qRT-PCR) analysis was performed after 48 hours of treatment with riluzole at 2 different concentrations (5 and 10 μM). Each
column and bar represent the mean ± SEM (n = 4 or 5). *p < 0.05 versus NT (one-way ANOVA followed by Dunnett’s test). Right:
qRT-PCR analysis after 48 hours of treatment with 300 μM of TMZ displayed significant upregulation of MGMT mRNA expression.
However, treatment with 5 μM of riluzole significantly reduced the TMZ-induced MGMT upregulation. Each column and bar repre-
sents the mean ± SEM (n = 5). ***p < 0.001 versus NT; ##p < 0.01 versus TMZ alone (one-way ANOVA followed by Tukey’s test).

J Neurosurg  Volume 134 • March 2021 707

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

Yamada et al.

FIG. 5. The effects of combinatorial therapy with TMZ and riluzole on the growth of MGMT-positive GBM in vivo. A: Scheme show-
ing the time point of administration of TMZ and riluzole to mice. B: Photomicrographs showing the H & E–stained murine brain
coronal sections following treatment of each group. Original magnification ×4. Scale bar = 1 mm. The dashed lines indicate the
margins of the tumors. C and D: Bar graphs representing the mean maximum cross-sectional areas (C) and mean tumor volumes
(D) of each group. Each column and bar represents the mean ± SEM (n = 10); *p < 0.05 versus NT; n.s. = not significant (one-way
ANOVA followed by Dunnett’s test).

These results suggest that combinatorial treatment with pression.17,27,34 These previous studies indicate that riluzole
riluzole and TMZ enhances the antitumor effect of TMZ suppresses MGMT expression by inhibiting the mGluR1/
and may improve the prognosis in patients with MGMT- PI3K/AKT/NFκB signaling pathway. Further studies are
positive GBM via suppression of MGMT. However, clini- required to completely elucidate the effect of riluzole on
cal trials are required to confirm whether combinatorial NFκB signaling.
treatment with riluzole and TMZ is as effective in human
patients as in experimental animal models.
This study, to our knowledge, is the first to report that
Conclusions
riluzole enhances the effect of TMZ against MGMT-posi- This study shows that when administering TMZ mono-
tive GBM cells via suppression of MGMT expression (Fig. therapy, MGMT-positive GBM cells displayed MGMT
4). Previous studies have indicated that riluzole exerts its upregulation and chemoresistance to TMZ. Riluzole sup-
antitumor effects by inhibiting the mGluR1/PI3K/AKT/ pressed the TMZ-induced upregulation of MGMT and
mTOR signaling pathway.28,36 Other studies have reported enhanced the antitumor effect of TMZ in MGMT-positive
that a PI3K inhibitor reduces chemoresistance to TMZ in GBM in vivo. In conclusion, our results indicate that com-
GBM cells, that another PI3K inhibitor suppresses MGMT binatorial treatment of TMZ/riluzole is a potential ther-
expression via inhibition of PI3K/AKT/NFκB signaling, apy for the effective treatment of patients with MGMT-
and that inhibition of NFκB downregulates MGMT ex- positive GBM.
708 J Neurosurg  Volume 134 • March 2021

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

 Yamada et al.

References 20. Natsume A, Ishii D, Wakabayashi T, Tsuno T, Hatano H,

Mizuno M, et al: IFN-β down-regulates the expression of
  1. Azzalin A, Nato G, Parmigiani E, Garello F, Buffo A, Ma- DNA repair gene MGMT and sensitizes resistant glioma cells
grassi L: Inhibitors of GLUT/SLC2A enhance the action of to temozolomide. Cancer Res 65:7573–7579, 2005
BCNU and temozolomide against high-grade gliomas. Neo- 21. Ono Y, Chiba S, Yano H, Nakayama N, Saio M, Tsuruma
plasia 19:364–373, 2017 K, et al: Glycoprotein nonmetastatic melanoma protein B
  2. Barker-Haliski M, White HS: Glutamatergic mechanisms (GPNMB) promotes the progression of brain glioblastoma
associated with seizures and epilepsy. Cold Spring Harb via Na+/K+-ATPase. Biochem Biophys Res Commun 481:7–
Perspect Med 5:a022863, 2015 12, 2016
  3. Bi Y, Li H, Yi D, Bai Y, Zhong S, Liu Q, et al: β-catenin 22. Paranjpe A, Zhang R, Ali-Osman F, Bobustuc GC, Srivenu-
contributes to cordycepin-induced MGMT inhibition and gopal KS: Disulfiram is a direct and potent inhibitor of hu-
reduction of temozolomide resistance in glioma cells by in- man O6-methylguanine-DNA methyltransferase (MGMT) in
creasing intracellular reactive oxygen species. Cancer Lett
brain tumor cells and mouse brain and markedly increases
435:66–79, 2018
the alkylating DNA damage. Carcinogenesis 35:692–702,
  4. Blandini F, Porter RH, Greenamyre JT: Glutamate and Par-
2014
kinson’s disease. Mol Neurobiol 12:73–94, 1996
23. Quinn JA, Jiang SX, Reardon DA, Desjardins A, Vreden-
  5. Bobustuc GC, Baker CH, Limaye A, Jenkins WD, Pearl G,
burgh JJ, Rich JN, et al: Phase II trial of temozolomide plus
Avgeropoulos NG, et al: Levetiracetam enhances p53-medi-
O6-benzylguanine in adults with recurrent, temozolomide-re-
ated MGMT inhibition and sensitizes glioblastoma cells to
temozolomide. Neuro Oncol 12:917–927, 2010 sistant malignant glioma. J Clin Oncol 27:1262–1267, 2009
  6. Chahal M, Abdulkarim B, Xu Y, Guiot MC, Easaw JC, Sti- 24. Ryu CH, Yoon WS, Park KY, Kim SM, Lim JY, Woo JS, et
fani N, et al: O(6)-Methylguanine-DNA methyltransferase is al: Valproic acid downregulates the expression of MGMT and
a novel negative effector of invasion in glioblastoma multi- sensitizes temozolomide-resistant glioma cells. J Biomed
forme. Mol Cancer Ther 11:2440–2450, 2012 Biotechnol 2012:987495, 2012
  7. Chinot OL, Wick W, Mason W, Henriksson R, Saran F, 25. Rzeski W, Ikonomidou C, Turski L: Glutamate antagonists
Nishikawa R, et al: Bevacizumab plus radiotherapy-temo- limit tumor growth. Biochem Pharmacol 64:1195–1200,
zolomide for newly diagnosed glioblastoma. N Engl J Med 2002
370:709–722, 2014 26. Shen D, Guo CC, Wang J, Qiu ZK, Sai K, Yang QY, et al:
  8. Corona JC, Tovar-y-Romo LB, Tapia R: Glutamate excitotox- Interferon-α/β enhances temozolomide activity against
icity and therapeutic targets for amyotrophic lateral sclerosis. MGMT-positive glioma stem-like cells. Oncol Rep 34:2715–
Expert Opin Ther Targets 11:1415–1428, 2007 2721, 2015
  9. Da Ros M, Iorio AL, De Gregorio V, Fantappiè O, Laffi G, 27. Shi F, Guo H, Zhang R, Liu H, Wu L, Wu Q, et al: The PI3K
de Martino M, et al: Aldoxorubicin and Temozolomide com- inhibitor GDC-0941 enhances radiosensitization and reduces
bination in a xenograft mice model of human glioblastoma. chemoresistance to temozolomide in GBM cell lines. Neuro-
Oncotarget 9:34935–34944, 2018 science 346:298–308, 2017
10. de Groot J, Sontheimer H: Glutamate and the biology of glio- 28. Sperling S, Aung T, Martin S, Rohde V, Ninkovic M: Rilu-
mas. Glia 59:1181–1189, 2011 zole: a potential therapeutic intervention in human brain
11. Doble A: The pharmacology and mechanism of action of tumor stem-like cells. Oncotarget 8:96697–96709, 2017
riluzole. Neurology 47 (6 Suppl 4):S233–S241, 1996 29. Stupp R, Mason WP, van den Bent MJ, Weller M, Fisher
12. Friedman HS: Temozolomide in early stages of newly diag- B, Taphoorn MJ, et al: Radiotherapy plus concomitant and
nosed malignant glioma and neoplastic meningitis. Semin adjuvant temozolomide for glioblastoma. N Engl J Med
Oncol 27 (3 Suppl 6):35–40, 2000 352:987–996, 2005
13. Gilbert MR, Dignam JJ, Armstrong TS, Wefel JS, Blumen- 30. Viel T, Monfared P, Schelhaas S, Fricke IB, Kuhlmann MT,
thal DT, Vogelbaum MA, et al: A randomized trial of beva- Fraefel C, et al: Optimizing glioblastoma temozolomide che-
cizumab for newly diagnosed glioblastoma. N Engl J Med motherapy employing lentiviral-based anti-MGMT shRNA
370:699–708, 2014 technology. Mol Ther 21:570–579, 2013
14. Jensen KV, Cseh O, Aman A, Weiss S, Luchman HA: The 31. Vlachostergios PJ, Hatzidaki E, Stathakis NE, Koukoulis
JAK2/STAT3 inhibitor pacritinib effectively inhibits patient- GK, Papandreou CN: Bortezomib downregulates MGMT
derived GBM brain tumor initiating cells in vitro and when expression in T98G glioblastoma cells. Cell Mol Neurobiol
used in combination with temozolomide increases survival 33:313–318, 2013
in an orthotopic xenograft model. PLoS One 12:e0189670, 32. Wang R, Reddy PH: Role of glutamate and NMDA receptors
2017 in Alzheimer’s disease. J Alzheimers Dis 57:1041–1048,
15. Le Liboux A, Lefebvre P, Le Roux Y, Truffinet P, Aubeneau 2017
M, Kirkesseli S, et al: Single- and multiple-dose pharma- 33. Yohay K, Tyler B, Weaver KD, Pardo AC, Gincel D, Blakeley
cokinetics of riluzole in white subjects. J Clin Pharmacol J, et al: Efficacy of local polymer-based and systemic deliv-
37:820–827, 1997 ery of the anti-glutamatergic agents riluzole and memantine
16. Lee HJ, Wall BA, Wangari-Talbot J, Shin SS, Rosenberg S, in rat glioma models. J Neurosurg 120:854–863, 2014
Chan JL, et al: Glutamatergic pathway targeting in mela- 34. Yu Z, Chen Y, Wang S, Li P, Zhou G, Yuan Y: Inhibition of
noma: single-agent and combinatorial therapies. Clin Cancer NF-κB results in anti-glioma activity and reduces temozolo-
Res 17:7080–7092, 2011 mide-induced chemoresistance by down-regulating MGMT
17. Li M, Liang RF, Wang X, Mao Q, Liu YH: BKM120 sensi- gene expression. Cancer Lett 428:77–89, 2018
tizes C6 glioma cells to temozolomide via suppression of the 35. Yu Z, Zhao G, Xie G, Zhao L, Chen Y, Yu H, et al: Metfor-
PI3K/Akt/NF-κB/MGMT signaling pathway. Oncol Lett min and temozolomide act synergistically to inhibit growth
14:6597–6603, 2017 of glioma cells and glioma stem cells in vitro and in vivo.
18. Liu L, Gerson SL: Targeted modulation of MGMT: clinical Oncotarget 6:32930–32943, 2015
implications. Clin Cancer Res 12:328–331, 2006 36. Zhang C, Yuan XR, Li HY, Zhao ZJ, Liao YW, Wang XY, et
19. Melamed JR, Morgan JT, Ioele SA, Gleghorn JP, Sims- al: Anti-cancer effect of metabotropic glutamate receptor 1
Mourtada J, Day ES: Investigating the role of Hedgehog/ inhibition in human glioma U87 cells: involvement of PI3K/
GLI1 signaling in glioblastoma cell response to temozolo- Akt/mTOR pathway. Cell Physiol Biochem 35:419–432,
mide. Oncotarget 9:27000–27015, 2018 2015

J Neurosurg  Volume 134 • March 2021 709

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC

Yamada et al.

37. Zhu T, Shen Y, Tang Q, Chen L, Gao H, Zhu J: BCNU/PLGA Nakayama, Yano, Iwama. Reviewed submitted version of manu-
microspheres: a promising strategy for the treatment of glio- script: Hara, Tsuji, Nakamura, Egashira, Shimazawa, Nakayama,
mas in mice. Chin J Cancer Res 26:81–88, 2014 Yano, Iwama. Approved the final version of the manuscript on
behalf of all authors: Hara. Statistical analysis: Hara, Yamada,
Tsuji, Shimazawa. Administrative/technical/material support:
Disclosures Hara, Egashira. Study supervision: Hara, Nakamura, Shimazawa,
The authors report no conflict of interest concerning the materi- Iwama.
als or methods used in this study or the findings specified in this
paper. Correspondence
Hideaki Hara: Gifu Pharmaceutical University, Gifu, Japan.
Author Contributions hidehara@gifu-pu.ac.jp.
Conception and design: Hara, Yamada, Tsuji, Nakamura, Yano.
Acquisition of data: Yamada, Tsuji. Analysis and interpretation
of data: Yamada, Tsuji. Drafting the article: Yamada. Critically
revising the article: Hara, Nakamura, Egashira, Shimazawa,

710 J Neurosurg  Volume 134 • March 2021

Unauthenticated | Downloaded 04/06/21 06:57 PM UTC