Caffeine Sensitizes U87-MG Human Glioblastoma Cells to Temozolomide through Mitotic Catastrophe by Impeding G2 Arrest

Abstract

Temozolomide (TMZ) is the first-line chemotherapeutic agent in the treatment of glioblastoma multiforme (GBM). Despite its cytotoxic effect, TMZ also induces cell cycle arrest that may lead to the development of chemoresistance and eventual tumor recurrence. Caffeine, a widely consumed neurostimulant, shows anticancer activities and is reported to work synergistically with cisplatin and camptothecin. The present study aimed to investigate the effects and the mechanisms of action of caffeine used in combination with TMZ in U87-MG GBM cells. As anticipated, TMZ caused DNA damage mediated by the ATM/p53/p21 signaling pathway and induced significant G2 delay. Concurrent treatment with caffeine repressed proliferation and lowered clonogenic capacity on MTT and colony formation assays, respectively. Mechanistic study showed that coadministration of caffeine and TMZ suppressed the phosphorylation of ATM and p53 and downregulated p21 expression, thus releasing DNA-damaged cells from G2 arrest into premature mitosis. Cell cycle analysis demonstrated that the proportion of cells arrested in G2 phase decreased when caffeine was administered together with TMZ; at the same time, the amount of cells with micronucleation and multipolar spindle poles increased, indicative of enhanced mitotic cell death. Pretreatment of cells with caffeine further enhanced mitotic catastrophe development in combined treatment and sensitized cells to apoptosis when followed by TMZ alone. In conclusion, our study demonstrated that caffeine enhanced the efficacy of TMZ through mitotic cell death by impeding ATM/p53/p21-mediated G2 arrest.

Full text

Research Article
Caffeine Sensitizes U87-MG Human
Glioblastoma Cells to Temozolomide through Mitotic
Catastrophe by Impeding G2 Arrest
Ning Li ,PingdeZhang ,KarrieMeiYeeKiang,
Yin Stephen Cheng, and Gilberto Ka Kit Leung
Department of Surgery, Li Ka Shing Faculty of Medicine, e University of Hong Kong, Queen Mary Hospital, Hong Kong
Correspondence should be addressed to Gilberto Ka Kit Leung; gilberto@hku.hk
Received 7 March 2018; Accepted 3 June 2018; Published 28 June 2018
Academic Editor: Monica Fedele
Copyright ©  Ning Li et al. is is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Temozolomide (TMZ) is the rst-line chemotherapeutic agent in the treatment of glioblastoma multiforme (GBM). Despite its
cytotoxic eect, TMZ also induces cell cycle arrest that may lead to the development of chemoresistance and eventual tumor
recurrence. Caeine, a widely consumed neurostimulant, shows anticancer activities and is reported to work synergistically with
cisplatin and camptothecin. e present study aimed to investigate the eects and the mechanisms of action of caeine used in
combination with TMZ in U-MG GBM cells. As anticipated, TMZ caused DNA damage mediated by the ATM/p/p signaling
pathway and induced signicant G delay. Concurrent treatment with caeine repressed proliferation and lowered clonogenic
capacity on MTT and colony formation assays, respectively. Mechanistic study showed that coadministration of caeine and TMZ
suppressed the phosphorylation of ATM and p and downregulated p expression, thus releasing DNA-damaged cells from
G arrest into premature mitosis. Cell cycle analysis demonstrated that the proportion of cells arrested in G phase decreased
when caeine was administered together with TMZ; at the same time, the amount of cells with micronucleation and multipolar
spindle poles increased, indicative of enhanced mitotic cell death. Pretreatment of cells with caeine further enhanced mitotic
catastrophe development in combined treatmentand sensitized cells to apoptosis when followed by TMZ alone. In conclusion, our
study demonstrated that caeine enhanced the ecacy of TMZ through mitotic cell death by impeding ATM/p/p-mediated
G arrest.
1. Introduction
Glioblastoma multiforme (GBM) is the most common and
aggressive form of primary brain tumor, with an annual
incidence of two to three per million adults. It accounts
for more than half of all primary intracranial tumors [].
Surgical removal followed by radiation and chemotherapy
is the standard treatment [, ]. However, the recurrence
rate remains high, with a median survival of just over one
year, rendering GBM one of the most challenging tumors to
manage [].
Temozolomide (TMZ), an oral alkylating agent, is the
rst-line chemotherapeutics for GBM patients. Its mecha-
nism of action majorly lies in the alkylation of N- or O-
guanine residues or N- adenine residues within DNA which
leads to mismatches during subsequent DNA replication and
consequentially cycle arrest, autophagy, senescence, and cell
death []. Cell cycle arrest upon DNA damage is thought
to be a double-edged sword, however. On the one hand,
transient cell cycle delay allows DNA repair and is considered
an essential self-protective process in maintaining cellular
homeostasis and preventing tumorigenesis in normal tissues.
On the other hand, it oers time for tumor cells to erase
alkylated residues, correct mismatched base pairs, and even-
tually promote tumor cell survival, and thereby contributing
to chemoresistance and disease recurrence []. In this regard,
agents that overcome TMZ-induced cell cycle arrest may
potentiate its ecacy in GBM treatment.
Caeine is a widely consumed neurostimulant found in
many food products including coee, tea, and so drinks.
Hindawi
BioMed Research International
Volume 2018, Article ID 5364973, 10 pages
https://doi.org/10.1155/2018/5364973

BioMed Research International
F : Demonstration of experimental strategies. CTRL: control; CAF: caeine alone; TMZ: temozolomide alone; CT: temezolomide with
caeine pretreatment; TC: combined treatment of temozolomide and caeine; and CTC: combined therapy with caeine pretreatment.
Caeineexertsmultiplebiologicaleectssuchasraising
blood pressure [], aecting gastrointestinal motility [],
and increasing basal metabolic rate []. Because it readily
penetrates the blood-brain barrier, caeine can inuence
psychological performance, enhance long-term memory, and
decrease the risk of neurodegenerative diseases such as
Parkinson’s disease []. Recently, an inverse association
between caeine intake and the risk of brain tumors was
reported by two epidemiological studies from the United
States and Europe [, ]. Laboratory evidence further
showed that caeine alone had a suppressive eect on the
proliferation of human U-MG glioma cells in vitro and
tumor growth in vivo []. Moreover, caeine reduced the
migration of GBM cell lines by impairing focal adhesion com-
plex formation []. When used in conjunction with cisplatin
or camptothecin, enhanced cytotoxicity against human U
glioma cells was observed []. is synergistic cytotoxicity
was attributed to the attenuation of chemotherapy-induced
G cell cycle delay by caeine. Against this background, we
investigated the eects of TMZ with or without concomitant
caeine on U-MG human glioma cells in vitro.Wealso
studied the priming eect of caeine with a view to mimic
the real-life situation of high daily caeine intake prior to
the administration of TMZ. e hypotheses were (i) caf-
feine would enhance TMZ cytotoxicity by attenuating TMZ-
induced G arrest; (ii) pretreatment with caeine would
further enhance this eect.
2. Materials and Methods
2.1. Cell Culture and Drug Treatment. Human GBM cell
line U-MG (American Type Tissue Collection, Manassas,
VA, USA) was cultured in minimum essential medium-
alpha (MEM-𝛼) supplemented with % heat-inactivated
foetal bovine serum (FBS),  IU/ml penicillin, and 
𝜇g/ml streptomycin (all from Gibco, Life Technologies, Inc.,
Carlsbad, CA, USA) in a humidied incubator with %
air atmosphere and % carbon dioxide at ∘C. Caeine
powder was obtained from Sigma, and stock solution was
prepared in culture medium at the concentration of 
mM and kept in ∘C. TMZ was obtained from Schering-
Plough (Kenilworth, NJ, USA) and dissolved in dimethyl
sulfoxide (DMSO, Sigma-Aldrich, Saint Louis, MO, USA)
at the concentration  mM, further diluted in culture
medium to  mM, and stored in -∘C. For treatment,
caeine and TMZ were further diluted in medium to the nal
concentrations as stated below.
A four-day treatment was divided into two phases: one-
day pretreatment and a subsequent three-day treatment of
either TMZ alone, caeine alone, or both (Figure ). Overall,
six arms of cells were set up accordingly. e dosage of
caeine was  mM as determined by MTT assay, and the
dosage of TMZ we opted was its IC50 value dened in U-
MG cells in our previous work [].
2.2. MTT Assay. is was used to dene the experimental
dosage of caeine in this study as well as the ecacy of
dierent treatment strategies. Briey, cells were plated in -
well microplates at a density of  cells/well overnight for
attachment. Caeine with dierent concentrations ranging
from . to  mM was administered on the second day,
and cells were incubated for  h. On the fourth day,  𝜇l
MTT stocking solution (mg/ml, Sigma-Aldrich) was added
on the top of the culture medium, and cells were lysed by
adding  𝜇l DMSO. Absorbance was acquired at  nm
using spectrophotometer (MultiSkan FC, ermo Scientic).
To examine the combined eects of caeine and TMZ
on viability, cells were seeded at a  x 4/well in -
well plates for microscopic imaging or  x 3/well in -
well culture plates for MTT assay, respectively. Aer full
attachment, cells were treated with six dierent strategies
as stated. For microscopic observation, cells in -well plates
were washed with x phosphate buered saline (PBS) and
xed by % paraformaldehyde for image taking. MTT assay
was conducted to test the cytotoxicity of dierent treatment
as previously described. ree experiments were performed
separately in quadruplicates.
2.3. Colony Formation Assay. For longer-term observation,
a clonogenic assay was performed. Briey,  cells/well
were seeded in -well plates and incubated overnight to
allow full attachment. On the following four days, dierent
treatment strategies were applied as stated above except that
the concentration of TMZ was reduced from  𝜇Mto
𝜇M. All cells were released free on the h day and allowed

BioMed Research International 
to be maintained in fresh medium for another  days. en
cells were xed with % ethanol and stained with crystal
violet solution (g/L, Sigma-Aldrich). Colonies comprising
more than y cells in each well were counted and analyzed
under microscopy.
2.4. Cell Cycle Analysis. Cells were harvested and washed
with PBS and xed in % ethanol at ∘Covernight.On
the following day, cells were washed twice with PBS and
resuspended in staining solution containing  𝜇g/ml propid-
ium iodide and  𝜇g/ml RNase (both from ermo Fisher
Scientic Inc., Waltham, MA) in PBS. Aer maintaining the
reaction in the dark at ∘C for  min, cells were washed and
resuspended in PBS before analysis on BD FACSCalibur ow
cytometry; results were analyzed using FlowJo soware.
2.5. Immunoblotting. Treated cells were washed with ice-cold
PBS twice and collected for homogenization with lysis buer
(Cell Signaling, Beverly, MA) containing proteinase inhibitor
cocktail (Roche Diagnostics, Indianapolis, IN, USA) to obtain
total cellular protein. Protein samples (- 𝜇gperlane)
were electrophoresed on a % or % sodium dodecyl sulfate
polyacrylamide gel and transferred to polyvinylidene uoride
membranes. Aer blocking in % non-fat milk for  h, mem-
branes were incubated with primary antibodies including
p-ATM (:), p (:), p-p (:), p𝑊𝑎𝑓 1 / 𝐶𝑖𝑝1
(p, :), caspase- (:), cleaved caspase- (:),
and GAPDH (:) (all from Cell Signaling) overnight at
∘C. On the following day, anti-rabbit (for p-ATM, p-p,
p, caspase-, cleaved caspase- and GAPDH, :, Santa
Cruz, CA, USA) or anti-mouse (for p, :, Sigma-
Aldrich) peroxidase-conjugated secondary antibodies. Pro-
tein bands were detected with chemiluminescent reagents
(GE Healthcare, Buckinghamshire, UK) and then exposed to
X-ray lm.
2.6. Immunouorescent Staining. Cells were seeded on 
mm coverslips and incubated with corresponding treatments.
On the harvest day, cells were xed with -∘Cmethanol
for  min at room temperature. Cells were then incubated
with % normal goat serum (Dako Corp., Carpinteria,
CA) for  h, followed by incubation with 𝛼-tubulin pri-
mary antibody (:, Santa Cruz) overnight at ∘C. A
uorescein-conjugated secondary antibody (ermo Fisher
Scientic) was used for visualization of the signal. Aer  h
of incubation, DAPI counterstaining was performed, and cell
counting and the image obtaining were conducted under the
uorescent microscope. e entire procedure was performed
under light-proof conditions.
2.7. Cell Counting and Statistical Analysis. For cell counting,
twenty random elds for each sample were chosen. All
positive cells with multiple spinal poles or micronuclei were
counted and summed up as the nal number for statistical
analysis. e number of cells was quantied by using Image J
(version.;NIH,Bethesda,MA,USA)inablindedmanner.
All data in the text were expressed as mean ±SD, and
statistical analyses were performed using Prism  and SPSS
F : MTT results for caeine toxicity. Dosages b elow mM did
not aect the viability of U-MG cells. erefore,  mM was chosen
as the target dosage in the further experiments. ∗p<. compared
to control.
... Multiple t-test or one-way ANOVAs with Tukey's mul-
tiple comparisons test were performed to evaluate dierences
among groups. A pvalue <. was considered statistically
signicant.
3. Results
3.1. Caeine Enhances TMZ’s Chemoecacy in Both Short-
and Long-Term Observations. As shown in Figure , cell
viability was not signicantly aected by caeine alone below
adosageofmMduringafour-daycourse;mMwas
therefore selected for use in subsequent experiments. ree-
day incubation with  𝜇M TMZ reduced cell proliferation
to %, with a further % reduction aer cotreatment with
caeine (TMZ versus TC, p<., Figure (b)). Interestingly,
pretreatment with caeine for one day in advance further
enhanced the ecacy of cotreatment (TC versus CTC, p
<.). ough a slightly reduced cell survival was also
observed in CT cells (i.e., caeine followed by TMZ alone),
no statistical dierence was reached when compared with
that of TMZ alone (CT versus TMZ, p>.). Microscopic
observationsconcurredwithMTTassaysandrevealedthe
same trend (Figure (a)).
On colony formation assay,  𝜇MTMZaloneresultedin
an obvious decrease in the number of colonies in comparison
with control, and preexposure to caeine did not bring
additional benets (TMZ versus CT, p>.). Concomitant
treatment with TMZ and caeine (i.e., TC group) reduced
the number of colonies to % of those in CTRL and %
of those in the TMZ alone group, suggesting an enhanced
antiproliferative eect with the use of caeine. In line with the
MTT results, cell growth was further inhibited if cells were
given caeine one day before combined therapy (Figures (c)
and (d)). Altogether, these data suggested that caeine could
enhance TMZ cytotoxicity in both short and long terms.

BioMed Research International
(a) (b)
(c) (d)
F : Results of MTT (b) and colony formation assay (c,d). Both assays showed that combination of caeine and TMZ exerted enhanced
antiproliferative eect in comparison with TMZ alone. Furthermore, pretreatment of caeine produced additional benets only when it was
followed by combined treatment but not TMZ alone. Images from bright eld demonstrated the same trend (a). ∗,,and&representforp<
. when compared to control, TMZ, and TC, respectively.
3.2. Caeine Abrogates TMZ-Induced G2/M Arrest through
Inhibiting ATM-p53-p21 Pathway. TMZ causes DNA damage,
triggers repair responses, and induces signicant cell cycle
arrest in glioma cell lines; attenuating this cell cycle delay
may facilitate chemocytotoxicity []. Caeine is known as
a potent cell cycle modulator by regulating ATM-mediated
signaling pathway []. Hence, we next asked whether or not
this may explain the enhanced chemoecacy as described
above. Flow cytometry results showed that TMZ alone pro-
duced a dramatic cycle delay with more than half of the cells
beingtrappedinGphase(Figure).epresenceofcaeine
together with TMZ partially abrogated the G arrest, as the
number of arrested cells decreased by approximately %
when compared to that observed aer TMZ-alone treatment.
Meanwhile, the proportion of cells with multinucleation
(MN) increased with concomitant caeine (TMZ versus TC,
p<.). MN or polyploid peak was usually considered
to be a marker of improper cell divisions, characterized
by an increased amount of cells containing multiple sets
of chromosomes []. e obvious increase of MN peak
indicated that cotreatment with TMZ and caeine induced
the polyploid formation in U-MG cells. e results were
also in line with the aforementioned ndings that the largest
amount of cells with MN was seen when caeine was given
both before and during TMZ treatment. Accompanying these
results, we also witnessed a sub-G peak in four TMZ-treated
groups, indicating that TMZ caused apoptotic cell death.
Combined treatment again showed a larger amount of cells
located in sub-G area; however, caeine pretreatment did not
bring additional increase of sub-G population, be it followed
by TMZ alone or combined treatment (further discussed
below).
Immunoblotting results showed marked ATM phospho-
rylation in response to TMZ challenge, which concurrently
led to an activation of its downstream p and p. Combined
treatment with TMZ and caeine suppressed the activation
of ATM/p/p pathway; the phosphorylation of ATM and
p was restrained, and the expression of p was downreg-
ulated (TMZ versus TC). Moreover, this eect was further
augmented when caeine was given  h ahead of TMZ plus
caeine (TC versus CTC) (Figure ).
3.3. Caeine Induced Cell Death through Mitotic Catastrophe,
Which May Be Independent from Apoptosis. When cells with

BioMed Research International 
F : Representative cell cycle analysis in dierent treatment strategies. Combination of caeine and TMZ impeded TMZ-induced G
delay in U-MG cells and caused an increase in the sub-G and MN phase (TMZ versus TC/CTC, p<.). An additional increase in MN
proportion but not sub-G was seen in pretreatment plus combination strategy. ∗,,and&representforp<. when compared to control,
TMZ, and TC, respectively.
damaged DNA bypass the cell cycle checkpoint and enter the
mitotic phase, they may manifest mitotic catastrophe (MC),
which is dened as a mechanism of mitosis-linked cell death
due to inappropriate entry into mitosis []. In the present
study, two characteristic features of MC, micronucleation and
mitotic spindle disruptions, were determined. We found that
the majority of nuclei were oval-shaped in CTRL cells, while
cells with multiple nucleic fragments were observed in TMZ-
treated populations (Figure (a), DAPI staining). It was more
frequently seen in combined treatment (TC) and could be
further enhanced when pretreatment was introduced (CTC,
Figure (b)). Aberrant chromosome segregation represented
as multiple spindle poles is also a typical feature of MC.
In our studies, over % of cells showed multiple spindle
poles which indicated mis-segregation of chromosomes in
TMZ group; this compared with only around % found in
CTRL and CAF (Figures (a) and (b), DAPI and 𝛼-tubulin
counter staining). en, the number of cells going through
multipolar spindle poles doubled when TMZ was com-
bined with caeine. Consistent with our cell cycle analysis,
pretreatment with caeine only produced additional eect
when followed by combined therapy but not TMZ alone, as
there were % more cells suering from chromosome mis-
segregation in CTC in comparison with TC. ese results
suggested that combined therapy brought more signicant
disturbances to chromosome segregation and mitosis that
would be indicative of an essential involvement of MC.
We then further investigated the level of cleaved caspase-
, an apoptosis marker. TMZ induced caspase- cleavage,
which was augmented when cells were preexposed to caeine
(Figures (c) and (d)). Concurrent treatment then showed
a more powerful induction in caspase-dependent apoptosis,

BioMed Research International
F : Immunoblotting analysis showed that TMZ-induced activation of ATM/p/p pathway was repressed by concurrent caeine
(TMZ versus TC), and it was further inhibited by preexposure of caeine  h ahead of combined therapy (TC versus CTC). GAPDH was
used as an internal loading control. ∗,,and&representforp<. when compared to control, TMZ, and TC, respectively.
while pretreatment did not bring more benets in this case.
ese ndings were inconsistent with the MC alterations,
indicating that caeine could also work as an apoptotic syn-
ergist with TMZ independently from MC. Indeed, apoptosis
by sub-G (Figure ) did not show a clear association between
apoptosis and MC either, and that no dierence was observed
between TC and CTC.
4. Discussion
Recent epidemiological studies demonstrated the benets
of caeinated drink intake in decreasing the risk of brain
oncogenesis [, ]; caeine alone was also reported to
suppress the proliferation and migration of GBM cells both
in vitro and in vivo [, ]. Caeine’s synergistic eects with
radiotherapy and chemotherapy were also demonstrated in
adenocarcinoma cells [], hepatocellular carcinoma cells
[], and cervical carcinoma cells []. In GBM, Janss et
al. found that caeine was a potent sensitizer for cisplatin
and camptothecin, as it enhanced the cytotoxicity of both
drugsinUcells[].Inlinewiththesendings,our
results demonstrated that caeine at a noncytotoxic concen-
tration promoted the ecacy of TMZ in U-MG cells. is
chemosensitizing benet produced by concurrent caeine
treatment was likely related to its regulation of cell cycle
progressions. It has been known that cell cycle delay upon
checkpoint activation following DNA damaging treatments
such as TMZ can facilitate DNA repairs, promote cancer cells
survival, and lead to chemoresistance []. ese dormant
cells may later exit cell cycle temporarily while remaining
metabolically active and are also thought to be more resistant
to chemotherapy when they reenter the cell cycle and begin
to divide again aer a period of time []. In the case of
GBM, TMZ treatment produces marked cell cycle arrest
which could partially explain the high rate of TMZ tolerance
and disease relapse in clinical treatment []. Our previous
works using TMZ-resistant U-MG and D-MG lines also
revealed dramatic cell cycle arrest aer long-term TMZ
exposure, indicating that a close association indeed exists
between cell cycle delays and the onset of TMZ resistance
[]. Conversely, the attenuation of this cell cycle arrest may
potentially promote the ecacy of current chemotherapies
andreducetheincidenceofchemoresistantrelapse.
ATM is one of the essential DNA damage response (DDR)
kinases, and its activation can restore genomic integrity
in response to DNA instability involved in a variety of
cellular processes []. Activation of ATM leads to subse-
quent phosphorylation of downstream substrates, such as p
and p, and exerts its eects on DNA repair, cell death,
and, most importantly, cell cycle arrest []. erefore, its
activation in GBM following TMZ treatment is considered
to be responsible for the cell cycle delay due to DNA

BioMed Research International 
(a) (b)
(c) (d)
F : Representative immunouorescent images illustrating characteristic micronucleation and multiple spindle poles during MC
occurrence (a). Proportion of cells showing MC features increased aer TMZ treatment, which could be enhanced by concurrent exploitation
of caeine. Pretreatment of caeine only exerted additional eect when it was followed by combined treatment but not TMZ alone. ∗,,and
&representforp<. when compared to control, TMZ, and TC, respectively (b). While combination of caeine augmented TMZ-induced
caspase- cleavage, early exposure of caeine only produced additional benets to TMZ but not combined treatment. ∗,,and&represent
for p<. when compared to control, TMZ, and CT, respectively (c, d).
double strand breaks (DSBs). Pharmacological inhibition
of ATM reverses this G arrest and renders GBM cells
more susceptible to TMZ []. In the present study, TMZ
dramatically increased the phosphorylation of ATM and its
important downstream factors p and p as anticipated,
leading to a substantial G stage delay. Caeine is known as
an inhibitor of ATM kinase activity [], and cotreatment of
caeine suppressed the activation of ATM signaling pathway
and ameliorated TMZ-induced G arrest, which contributed
to the augmentation of TMZ chemoecacy. ese ndings
are in agreement with previous studies in which caeine
augmented radiotherapeutic benets by inhibiting ATM and
ATR activities in lung adenocarcinoma cells and leukemia
cells [].
We also observed an inhibition of phosphorylation of p ,
a well-known tumor suppressor, through the suppression of
ATM activation in combined TMZ plus caeine treatment.
Extensive studies had shown correlations between p func-
tional loss and tumorigenesis of liver cancer, lung cancer,
colon cancer, and GBM []. Activation of p plays a crucial
role in TMZ therapies in GBM treatment as it is believed to
mediate apoptotic cell death following ATM-mediated DDR
[]. In this regard, one might expect that downregulating
p would work against DNA damaging agents in the
treatment of GBM. However, current evidence suggests that
p participation in anticancer therapy exhibits a two-armed
eect; in addition to apoptosis, p mediates checkpoint
activation and induces cell cycle delay which is considered
to favor DNA repair and attenuate chemoecacy []. In the
present study, although the phosphorylation of p was sig-
nicantly suppressed, chemoecacy was augmented as both
mitotic cell death and apoptosis were enhanced following
treatment with both caeine and TMZ (TMZ versus TC).
is result agreed with several works which demonstrated
that GBM cells with inactive p were more sensitive to TMZ
treatment than those with wild-type p [, ].
We surmised, and our ndings suggested, that GBM
cells that reenter cell cycle following our treatments would
carryimpairedDNAandexperienceaberrantandincomplete
mitosis. e latter maylead to cell division failure and cellular
break down. is form of catastrophic cell division, or MC,
is characterized by the distinguishing features of giant cells

BioMed Research International
with micronuclei formation and multiple spindle poles, both
of which reect an abnormal segregation of chromosomes
[]. G arrest prevents MC by halting premature entry,
and MC will manifest when G arrest is abrogated. In
this work, we observed an inverse association between the
extent of G arrest and frequency of MC. e number of
cells showing large cell bodies with micronucleation and
multipolar spindles was larger in combined treatment groups
when compared with TMZ alone. is trend was in line
with the alterations in cell cycle analysis where G peak was
lowerandMNpeakwashigherinTCandCTCgroups
in comparison with TMZ alone. Our results supported the
notion that caeine produced its chemosensitizing eects
by promoting mitotic cell death through abrogating TMZ-
induced G delay.
So far, there is no general consensus on the relationship
between MC and apoptosis []. Caspases were reported
to be essential for the termination of MC, indicating that
premature mitotic failure might act as an intermediate
process leading to ultimate apoptotic cell death [, ].
On the other hand, in p-decient U- OS bone osteosar-
coma cells, MC could be induced by adeno-associated virus
in the absence of caspase activation and apoptosis [].
In another MDR-induction model in HeLa cells, apop-
tosis was signicantly suppressed, while cellular fractions
resulted from MC increased aer ionizing radiation []. e
obvious phenotypic dierences between apoptosis and MC
also supported the notion that these two cellular processes
may be independent. Herein, although combined treatment
enhanced both apoptosis and MC, early caeine exposure
only brought additional benets to apoptotic cell death when
followed by TMZ alone, which was in contrary with its
eect on MC development. Supportively, a previous study
using SH-SYY human neuroblastoma cell line showed that
pretreatment with caeine increased the sensitivity of cells
to doxorubicin-induced apoptosis secondary to an increased
production of mitochondrial-free radical []. Together with
our data, we propose that caeine pretreatment before
the administration of chemotherapeutics could sensitize
glioma cells to apoptosis that is likely independent from
MC.
Last but not least, it was interesting to observe that
caeine-pretreated cells showed an enhanced sensitivity to
combined treatment through MC augmentation but not to
TMZ alone. It has been reported that long and chronic
caeine exposure would sensitize organisms to some sort
of drugs such as methylphenidate [], but the mechanism
is far from clear. In the study by Susan et al., although
the caeine-induced sensitization of neuroblastoma cells to
doxorubicin was attributed to increased mitochondrial-free
radical production, a decrease of total reactive oxygen species
(ROS) production was also noticed []. Evasion of ROS
production by antioxidants was reported to facilitate MC
development [], indicative of a potential involvement of
ROS regulation in a setting similar to our study. Note should
be taken that potentiation by caeine to chemotherapy is
cell line- and agent-specic []; therefore, further studies
inotherGBMcelllinesaswellasinanimalmodelsare
needed.
F : Proposed mechanisms for caeine in sensitizing TMZ’s
ecacy. TMZcauses DNA damage, activates ATM/p/ppathway,
and induces signicant G arrest (full line). Caeine inhibits the
phosphorylation of ATM and thus suppresses the activation of DDR
pathway. G delay is abrogated, cells with damaged DNA enter
mitotic phase prematurely, and augmented MC occurs (dotted line).
Our results did not show a dependent association between MC and
apoptosis; hence further studies are warranted.
5. Conclusions
e present study demonstrated that caeine enhanced
TMZ’s chemoecacy through impeding G delay by inhibit-
ing ATM/p/p pathway and the promotion of mitotic
catastrophe (Figure ). Our results highlighted the following.
() Caution should be paid to the fact that cancer cells possess
dierent intrinsic responses to chemotherapies, and GBM is
known to be resistant to apoptosis. TMZ induces signicant
cell cycle arrest at an early stage of treatment rather than
cell death, which could account for the frequent clinical
relapses seen aer TMZ treatment. () Attenuating G delay
aer TMZ treatment enhances MC, which may then lead to
cellular breakdown in ways that may be either dependent or
independent from apoptosis. is particular property should
be further explored in the treatment of GBM. () Caeine
pretreatment augments the sensitizing ecacy of combined
treatment, indicating the potential benet of a high intake
of caeinated products before and during TMZ treatment. In
vivo conrmation of our ndings as well as epidemiological
studies is needed.
Data Availability
e data used to support the ndings of this study are
available from the corresponding author upon request.
Conflicts of Interest
e authors declare no conicts of interest regarding the
publication of this paper.

BioMed Research International 
References
[] K. Urba´nska, J. Sokołowska, M. Szmidt, and P. Sysa, “Review
Glioblastoma multiforme – an overview,” Wsp´
ołczesna Onkolo-
gia,vol.,pp.–,.
[] P.Zhang,S.Sun,N.Lietal.,“Rutinincreasesthecytotoxicity
of temozolomide in glioblastoma via autophagy inhibition,”
Journal of Neuro-Oncology,vol.,no.,pp.–,.
[] R.M.Young,A.Jamshidi,G.Davis,andJ.H.Sherman,“Current
trends in the surgical management and treatment of adult
glioblastoma,” Annals of Translational Medicine,vol.,p.,
.
[] A. M. Rulseh, J. Keller, J. Klener et al., “Long-term survival of
patients suering from glioblastoma multiforme treated with
tumor-treating elds,” World Journal of Surgical Oncology,vol.
, article , .
[] J.Zhang,M.F.G.Stevens,andT.D.Bradshaw,“Temozolomide:
Mechanisms of ac tion, repair and resistanc e,” Current Molecular
Pharmacology,vol.,no.,pp.–,.
[] W.P.Roos,A.D.omas,andB.Kaina,“DNAdamageandthe
balance between survival and death in cancer biology,” Nature
Reviews Cancer,vol.,no.,pp.–,.
[]J.W.Daniels,P.A.Mol´e, J. D. Sharath, and C. L. Stebbins,
“Eects of caeine on blood pressure, heart rate, and forearm
blood ow during dynamic leg exercise,” Journal of Applied
Physiology,vol.,no.,pp.–,.
[] P.J.Boekema,M.Samsom, G.P.vanBergeHenegouwen,and
A. J. Smout, “Coee and gastrointestinal function: Facts and
ction: A revie w,”Scandinavian Journal of Gastroenterolog y,vol.
, no. , pp. –, .
[] P. Koot and P. Deurenberg, “Comparison of changes in energy
expenditure and body temperatures aer caeine consump-
tion,” Annals of Nutrition and Metabolism,vol.,no.,pp.–
, .
[] A.Ascherio,S.M.Zhang,M.A.Hern´an et al., “Prospective
study of caeine consumption and risk of Parkinson’s disease in
men and women,” Annals of Neurology,vol.,no.,pp.–,
.
[] C. N. Holick, S. G. Smith, E. Giovannucci, and D. S. Michaud,
“Coee, tea, caeine intake, and risk of adult glioma in three
prospec tive cohort studies,” Cancer Epidemiology, Biomarkers &
Prevention,vol.,no.,pp.–,.
[] D. S. Michaud, V. Gallo, B. Schlehofer et al., “Coee and tea
intake and risk of brain tumors in the European Prospective
Investigation into Cancer and Nutrition (EPIC) cohort study,”
American Journal of Clinical Nutrition,vol.,no.,pp.–
, .
[]B.M.Ku,Y.K.Lee,J.Y.Jeongetal.,“Caeineinhibitscell
proliferation and regulates PKA/GSK𝛽pathways in UMG
human glioma cells,” Molecules and Cells,vol.,no.,pp.–
, .
[] Y. Chen, W.-C. Chou, Y.-M. Ding, and Y.-C. Wu, “Caeine
inhibits migration in glioma cells through the ROCK-FAK
pathway,” Cellular Physiology and Biochemistry,vol.,no.,
pp. –, .
[] A. J. Janss, C. Levow, E. J. Bernhard et al., “Caeine and
staurosporine enhance the cytotoxicity of cisplatin and camp-
tothecin in human brain tumor cell lines,” Experimental Cell
Research,vol.,no.,pp.–,.
[] S. Y. Lee, “Temozolomide resistance in glioblastoma multi-
forme,” Genes and Diseases,vol.,no.,pp.–,.
[] A. Blasina, B. D. Price, G. A. Turenne, and C. H. McGowan,
“Caeine inhibits the checkpoint kinase ATM,” Current Biology,
vol. , no. , pp. –, .
[] I.S.Smirnova,N.D.Aksenov,E.V.Kashuba,P.Payakurel,V.
V. Grabovetsky, A. D. Zaberezhny et al., “Hepatitis C virus core
protein transforms murine broblasts by promoting genomic
instability,” Cellular Oncology,vol.,no.,pp.–,.
[] H. Vakifahmetoglu, M. Olsson, and B. Zhivotovsky, “Death
through a tragedy: mitotic catastrophe,” Cell Death & Dieren-
tiation,vol.,no.,pp.–,.
[] A. R. Kadia and G. B. Shah, “Cisplatin resistance reversal by
disulram and caeine,” Journal of Pharmacology and Pharma-
cotherapeutics,vol.,no.,pp.–,.
[] Y.Kawano,M.Nagata,T.Kohnoetal.,“Caeineincreasesthe
antitumor eect of cisplatin in humanhepatocellular carcinoma
cells,” Biological & Pharmaceutical Bulletin,vol.,no.,pp.
–, .
[] D. Rybaczek and M. Kowalewicz-Kulbat, “Premature chro-
mosome condensation induced by caeine, -aminopurine,
staurosporine and sodium metavanadate in S-phase arrested
HeLa cells is associated with a decrease in Chk phosphory-
lation, formation of phospho-HAX and minor cytoskeletal
rearrangements,” Histochemistry and Cell Biology,vol.,no.
, pp. –, .
[] E.S.Newlands,M.F.G.Stevens,S.R.Wedge,R.T.Wheelhouse,
and C. Brock, “Temozolomide: a review of its discovery, chem-
ical properties, pre-clinical development and clinical trials,”
Cancer Treatment Reviews,vol.,no.,pp.–,.
[] S. Sun, T. S. Wong, X. Q. Zhang et al., “Protein alterations
associated with te mozolomide resistance in subc lones ofhuman
glioblastoma cell lines,” Journal of Neuro-Oncology,vol.,no.
, pp. –, .
[] A. Mar´echal and L. Zou, “DNA damage sensing by the ATM and
ATR k inase s,” Cold Spring Harbor Perspectives in Biology,vol.,
no. , .
[] E. Crescenzi, G. Palumbo, J. De Boer, and H. J. M. Brady,
“Ataxia telangiectasia mutated and pCIP modulate cell sur-
vival of drug-induced senescent tumor cells: Implications for
chemotherapy,”Clinical Cancer Research,vol.,no.,pp.–
, .
[] A. Nadkarni, M. Shrivastav, A. C. Mladek et al., “ATM inhibitor
KU- increases the TMZ responsiveness of only inherently
TMZ sensitive GBM cells,” Journal of Neuro-Oncology, vol. ,
no. , pp. –, .
[] J.N.Sarkaria,E.C.Busby,R.S.Tibbettsetal.,“Inhibitionof
ATM and ATR ki nas e act iviti es by the r adios ensit izing agent ,
caeine,” Cancer Research,vol.,no.,pp.–,.
[] P. A. J. Muller and K. H. Vousden, “P mutations in cancer,”
Nature Cell Biology,vol.,no.,pp.–,.
[] H. Hasegawa, Y. Yamada, H. Iha et al., “Activation of p
by Nutlin-a, an antagonist of MDM, induces apoptosis and
cellular senescence in adult T-cell leukemia cells,” Leukemia,
vol. , no. , pp. –, .
[] S. Martin, H. Janouskova, and M. Dontenwill, “Integrins and
p pathways in glioblastoma resistance to temozolomide,”
Frontiers in Oncology,vol.,.
[] M.D.Blough,D.C.Beauchamp,M.R.Westgate,J.J.Kelly,
and J. G. Cairncross, “Eect of aberrant p function on
temozolomide sensitivity of glioma cell lines and brain tumor
initiating cells from glioblastoma,” Journal of Neuro-Oncology,
vol. , no. , pp. –, .

 BioMed Research International
[] E. B. Dinca, K. V. Lu, J. N. Sarkaria et al., “p small-molecule
inhibitor enhances temozolomide cytotoxic activity against
intracranial glioblastoma xenogras,” Cancer Research,vol.,
no. , pp. –, .
[] M. M. Mc Gee, “Targeting the Mitotic Catastrophe Signaling
Pathway in Cancer,” Mediators of Inammation,vol.,
Article ID , .
[] S. Mansilla, W. Priebe, and J. Portugal, “Mitotic catastro-
phe results in cell death by caspase-dependent and caspase-
independent mechanisms,” Cell Cycle,vol.,no.,pp.–,
.
[] A. Skwarska, E. Augustin, and J. Konopa, “Sequential induction
of mitotic catastrophe followed by apoptosisin human leukemia
MOLT cells by imidazoacridinone C-,” Apoptosis,vol.,
no.,pp.–,.
[] H. Tauchi and S. Sawada, “Analysis of mitotic cell death
caused by radiation in mouse leukaemia ly cells: Apoptosis
is the ultimate form of cell death following mitotic failure,”
International Journal of Radiation Biology,vol.,no.,pp.
–, .
[] M. Fragkos and P. Beard, “Mitotic catastrophe occurs in the
absence of apoptosis in p-null cells with a defective G
checkpoint,” PLoS ONE,vol.,no.,ArticleIDe,.
[] A. C. Ruth and I. B. Roninson, “Eects of the multidrug
transporter P-glycoprotein on cellular responses to ionizing
radiation,” Cancer Research, vol. , no. , pp. –, .
[] S. Hall, S. Anoopkumar-Dukie, G. D. Grant et al., “Modulation
of chemotherapy-induced cytotoxicity in SH-SYY neurob-
lastoma cells by caeine and chlorogenic acid,” Toxico l o g y
Mechanisms and Methods,vol.,no.,pp.–,.
[] C. R. Boeck, V. B. Marques, S. S. Valvassori et al., “Early
long-term exposure with caeine induces cross-sensitization to
methylphenidate with involvement of DARPP- in adulthood
of rats,” Neurochemistry International,vol.,no.,pp.–,
.
[] I.V.Sorokina,T.V.Denisenko,G.Imreh,V.Gogvadze,and
B. Zhivotovsky, “Reactive oxygen species regulate a balance
between mitotic catastrophe and apoptosis,” e International
Journal of Biochemistry & Cell Biology,vol.,pp.–,.