Introduction
Glioblastoma (GBM), a WHO grade IV astrocytoma, is the most common [
1] and most aggressive type of primary brain tumour. Current treatment consists of tumour resection (where possible), followed by ionising radiotherapy combined with concomitant and adjuvant temozolomide (TMZ) chemotherapy [
2]. TMZ is a methylating agent that creates lesions in DNA, the most cytotoxic of which is O
6-methylguanine [
3]. This multimodal treatment regimen is rigorous, yet prognosis remains poor and although TMZ chemotherapy improves survival in a subset of patients, 75% die within 2 years and the vast majority of patients experience disease recurrence [
4].
O
6-methylguanine-DNA methyltransferase (MGMT) is a DNA repair protein, which removes O
6-methylguanine adducts from damaged DNA [
5]. This reaction is irreversible and once bound to the alkyl group MGMT is ubiquitinated and destroyed by the proteasome [
6]. The cytotoxic effect of TMZ chemotherapy is therefore influenced by the ability of tumour cells to re-synthesise MGMT and maintain steady state levels of the protein. Previous work has shown that
MGMT promoter methylation resulting in gene silencing and resultant low levels of MGMT protein, increases sensitivity to TMZ and is associated with improved patient survival. Unfortunately direct inhibition of MGMT using small molecule inhibitors such as lomeguatrib has not been successful as a clinical application because it also increases haematological toxicity [
2,
7,
8]. However, downregulation of MGMT remains an attractive therapeutic strategy for patients with tumours exhibiting unmethylated
MGMT promoters if it could be achieved in a tumour specific manner.
Common mutations in GBM cells include genetic changes that result in a loss of PTEN function and EGFR amplification [
9], both of which can generate hyperactive phosphoinositide 3-kinase (PI3K)/mTOR signalling. mTOR is a serine/threonine kinase that belongs to the PI3K-related kinase family and interacts with proteins to form two distinct complexes in mammals, mTORC1 and mTORC2 [
10]. mTORC1, when active, regulates protein translation through the phosphorylation of 4EBP1. Phosphorylation of 4E-BP1 prevents it binding to eIF4E, which enables eIF4E to participate in the formation of the eIF4F complex on the m
7 GTP cap structure of mRNA and mediate small ribosomal subunit binding and subsequent protein translation [
11]. The hyperactivity of this pathway therefore results in increased protein synthesis, promoting cell growth and proliferation.
Because of this there has been interest in the use of mTOR inhibitors in combination with radiation and TMZ in the treatment of GBM. The rationale for combining mTOR inhibitors with TMZ treatment is based on the reasoning that the lesions in DNA caused by TMZ will result in a depletion of cellular MGMT protein levels. When coupled with mTOR inhibition, not only would there be a decrease in MGMT levels, but the tumour cell would be compromised in its ability to synthesise new protein, thus sensitising the cells to further TMZ treatment. In addition to this, tumour cells should be specifically targeted with this course of treatment, due to the tumour cells oncogenic addiction to the PI3K/mTOR signalling pathway. This would avoid the current drawbacks faced during direct inhibition of MGMT, as it would avoid MGMT depletion in healthy cells, and therefore avoid undesirable cytotoxicity.
In this work, we have used Western blotting to examine the effect of inhibiting mTORC1/2 signalling on steady state MGMT protein levels in T98G GBM cells, a cell line which exhibits relatively high MGMT expression compared to other glioblastoma cell lines [
12,
13]. KU0063794 is a specific mTORC1 and mTORC2 inhibitor that does not display significant activity against similar kinases such as PI3K, ERK1/2, or p38MAPK [
14].
The findings described in this paper are of both biochemical and potential clinical interest. The research highlights how important it is to identify how DNA repair proteins are translated and maintained as proteins, which is an important consideration, especially when manipulating them for clinical benefit.
Materials and methods
Cell culture
T98G (ECACC) cells were cultured to confluency in minimum essential medium (MEM) supplemented with 5% non-essential amino acids (Invitrogen, UK), 10% foetal calf serum (Biosera, UK) and 5% GlutaMax (Invitrogen, UK). U87 (ECACC) cells were cultured in the same manner but with Hyclone foetal calf serum.
Concentrations of drug treatments
Cells were treated with the following concentrations of drugs: 10 μM KU0063794, 10 μM TMZ and 10 μM emetine.
Following treatment, cells were isolated in a cooled microfuge and washed briefly with 0.5 ml ice-cold PBS. Pellets were resuspended in 100 μl ice-cold lysis buffer (20 mM MOPS-KOH, pH 7.2, 20 mM sodium fluoride, 1 μM microcystin LR, 75 mM KCl, 2 mM MgCl2, 2 mM benzamidine, 2 mM Na3VO4, complete protease inhibitor mix (−EDTA (Roche, UK), 0.1% (v/v) SDS), with the addition of 0.5% (v/v) Igepal and 0.5% (w/v) deoxycholate (DOC) with vortexing. Cell debris was removed by centrifugation in a microfuge for 5 min at 4°C and the resultant supernatants recovered.
Western blotting
Samples containing equal amounts of protein (10 μg) were resolved by polyacrylamide gel electrophoresis (SDS-PAGE) and processed as described previously [
15,
16]. Briefly, membranes were blocked using TBS-Tween containing 3% (w/v) BSA for 1 hour and incubated with antisera diluted in the same overnight at 4°C. Following washing in TBS-Tween, membranes were incubated with horseradish peroxidase-conjugated secondary antibody and signals developed using ECL reagent. The antiserum used were raised against: eIF4A, phospho-eIF4E, eIF4G, eIF4E [
15,
16]; MGMT, phospho-eIF2α, (Abcam, UK); 4E-BP1, phospho-4E-BP S65, phospho-4E-BP1 T70, phospho-p70 S6K T389, phospho- ERK1/2, phospho-p38 MAPK, phospho-rpS6 S240/44, phospho-Akt T308, p21
cip1 (Cell Signaling, UK).
[35S]-methionine labelling of cellular protein
One hour prior to harvesting, T98G cells were pulse-labelled with [35S]-methionine (MP Biomedicals, UK; 10 μCi/ml) and cell extracts prepared as above and 5 μl of extract was spotted onto Whatman filter papers. Filters were transferred to 10% (v/v) trichloroacetic acid (TCA) containing 5 mM unlabelled methionine for 15 minutes then boiled in 5% (v/v) TCA. Once cooled, the filters were washed in ethanol, then acetone, dried and subjected to liquid scintillation counting. Protein synthesis, expressed as cpm/ μg protein, is shown as a % of the rate obtained in cells incubated in the absence of drugs.
MTS assay
Cells were cultured in a 96 well plate in 100 μl of complete medium. Following incubation with drugs, 20 μl of MTS/PMS solution (Promega, UK) was added, and left for 2 hours at 37°C. The colour reaction, reflecting metabolic activity was quantified by measuring absorbance at A490 using the control wells for comparison. All assays were carried out in triplicate.
Flow cytommetry
Following incubation of cells as described in the figure legends, cells were washed with warm PBS then trypsinised and removed from culture dishes. Cells were washed in PBS and fixed in 70% ethanol/30% PBS at 4°C overnight. Immediately before analysis, the cells were washed in PBS and re-suspended in 500 μl PBS containing 0.1 mg/ml RNase A (Sigma, UK) and 30 μg/ml propidium iodide (Invitrogen, UK) and analysed by a FACSCanto™ flow cytometer (BD Biosciences) using BD FACSDiva™ software. Data shown are representative of those obtained in at least three separate experiments.
Quantitative RT-PCR
RNA was extracted from cell extracts using an RNA easy mini-kit (Qiagen, UK) as per manufacturer’s instructions. RNA concentration was then quantified using a Nanodrop and 1 μg of RNA was used for cDNA synthesis using the Promega ImpromII kit. The SYBR real-time PCR system (Kapa Biosystems, UK) was used to quantify transcript abundance for genes of interest and18S mRNA was used as a control. Template equivalent to 5 ng of RNA in cDNA library per reaction was added to each 20 μl reaction with a final primer concentration of 200 nM per reaction. Crossing thresholds were determined using MxPro software (Agilent), and fold-difference in RNA quantity was calculated using the relative quantification method (2-ΔΔct).
Immunofluorescence
Cells were plated on glass coverslips 8 hours before treatment. Cells were then fixed with 4% (w/v) paraformaldehyde for 20 minutes at room temperature and permeabilised for 5 minutes in 0.1% Triton X-100/PBS. Cells were blocked for 20 minutes with 5% (w/v) gelatin/PBS in a humidified chamber. Cells were incubated with anti-MGMT (Abcam, UK) for 1 hour at room temperature also in a humidified chamber. Alexa Fluor 555-conjugated anti-rabbit secondary antibody was added for 1 hour at room temperature. Alexa Fluor 488 phalloidin was also used in the secondary antibody incubation. Following further extensive washing, nuclei were stained with DAPI (4’, 6-diamidino-2-phenylindole) for 5 minutes. After a further two washes, coverslips were mounted on microscope slides with Mowiol mounting solution (0.2 M Tris/HCl (pH 8.5), 33% (w/v) glycerol, 13% (w/v) Mowiol and 2.5% (w/v) DABCO (1, 4-diazobicyol [
2]-octane)) and sealed with clear nail polish. Images were collected on a Zeiss Axiovert LSM510 scanning confocal microscope using a × 100 objective. Single stain, bleed-though controls and antibody cross-reaction controls were prepared for each sample (results not shown).
m7GTP-Sepharose affinity chromatography
To isolate eIF4E and associated proteins [
15,
16] cells were lysed as described and aliquots containing 100 μg were incubated with 30 μl (50% v/v in 1 mg/ml cytochrome c) m
7GTP-Sepharose resin for 15 mins on ice, with mixing. The resin was washed twice in 200 μl buffer (20 mM MOPS/KOH pH 7.2, 75 mM KCl, 2 mM benzamidine, 7 mM 2-mercaptoethanol, 2 mM MgCl
2, 0.1 mM GTP, complete protease inhibitor mix (−EDTA), 25 mM NaF). To visualise eIF4E and associated proteins, the resin was washed twice with 200 μl buffer then eluted in SDS-PAGE sample buffer containing 10% w/v β-mercaptoethanol prior to Western blotting.
Polysome profiles
Cells were treated with 100 μg/ml of cycloheximide (Sigma, UK) for 3 mins before harvesting. The cells were washed twice in PBS containing 100 μg/ml cycloheximide and lysed in hypertonic lysis buffer (10 mM Tris/HCl pH7.5, 10 mM KCl, 15 mM MgCl2, 1%(v/v) Igepal, 0.5%(v/v) DOC, 40 mM β-glycerophosphate, complete protease inhibitor mix (−EDTA), 2 mM DTT, 100 μg/ml cycloheximide, 10 U/ml RNase inhibitor) for 15 minutes on ice and homogenised every 5 mins. The cell lysate was centrifuged at 14,000 rpm for 3 mins and the supernatant was separated on a 11.2 ml sucrose gradient (10-60% (w/v) in hypertonic lysis buffer) for 130 mins at 38, 000 rpm using a Beckman SW40 rotor and fractionated using an ISCO, monitoring A260 nm through a flow cell. For analysis of GAPDH and MGMT mRNA, the mRNA from pooled fractions was extracted and quantitative RT-PCR was used to determine mRNA distribution across each fraction.
Discussion
GBM is the most aggressive adult brain cancer and TMZ, an oral methylating chemotherapeutic agent, is part of the standard treatment, often accompanied by radiotherapy [
2‐
4]. TMZ exerts its antitumour effect by methylating guanine at a number of sites [
22]. The N
7 guanine and N
3 adenine adducts caused by TMZ are repaired by the base excision repair pathway, but the O
6-methyl-guanine (O
6MeG) lesion cannot be repaired in this fashion, making it the most potently cytotoxic adduct [
22]. O
6MeG pairs with thymine and leads to GC-AT transitions, beginning a cycle of futile DNA mismatch repair which causes DNA double-stranded breaks and, ultimately causes tumour cell death [
6,
22,
23].
MGMT is a DNA repair protein that repairs the O
6MeG lesion created by TMZ by a process of irreversible binding and subsequent degradation of the protein [
5,
23]. When O
6MeG is repaired by MGMT, the majority of cells become resistant to O
6MeG-triggered apoptosis [
4‐
6]. Hence the level of MGMT activity correlates with resistance of tumour cells to TMZ [
24]. As mutations in the mTOR signalling pathway in GBM frequently result in hyperactive PI3-K/mTOR signalling, promoting cell survival and protein synthesis [
9,
10,
20], we investigated whether mTOR kinase activity has a role in regulating levels of MGMT protein expression in human glioblastoma cells.
As predicted, we show that TMZ does not reduce MGMT protein levels in T98G cells, so the large amount of MGMT in the cells is effectively repairing any lesions created by TMZ. This suggests that, in agreement with previous studies [
25‐
27], on its own TMZ is not sufficiently effective in cellular environments containing MGMT overexpression.
In contrast to cells treated with TMZ alone, when protein loading was accounted for in three separate experiments, our data show that in cells treated with both KU0063794 and TMZ, there was an increase in MGMT protein expression. This occurred even though global protein synthesis rates were reduced by around 80% under these conditions. These data suggest that although dual mTOR inhibition was indeed impairing the formation of the translation initiation complex, reducing global protein synthesis rates and causing a disaggregation of polysomes from mRNA, protein levels were still increasing. To decipher if this unexpected maintenance of MGMT protein levels was reliant on an alternative translation mechanism, we assessed the translation status of MGMT mRNA. This demonstrated that during conditions of reduced global translation rates, and therefore reduced mRNA association with polysomes, MGMT mRNA remained associated with heavy polysomes, although GAPDH mRNA shifted to sub polysomal fractions. This is strong evidence that MGMT mRNA might be translated using a cap-independent mechanism, or preferentially recruited to ribosomes under conditions of low levels of eIF4F [
28], but in the absence of increased levels of eIF2α phosphorylation (Figure
3). This is further supported by the presence of a number of GC-rich short stem loops in the annotated 3’-UTR of the
MGMT mRNA, which may also have a role in this process. Additional extensive biochemical analysis would be needed to determine if this is the case. Alternative mechanisms controlling translation have been identified under hypoxic stress conditions, which rely on protein complexes involving the alternative eIF4E isoform eIF4E2, specifically recruiting target mRNAs to polysomes during periods of reduced cap dependant translation [
29]. These findings could compliment the preferential translation of MGMT under stress conditions, but further investigations would need to be carried out to confirm this.
In addition to the mechanisms regulating translational control of MGMT mRNA, the stability of the MGMT protein was also investigated, as this is of paramount importance when considering its manipulation for effective chemotherapeutic treatment. The inhibition of protein synthesis elongation with emetine in the presence of TMZ caused a dramatic decrease in MGMT protein level, reflecting the rapid degradation of the protein following DNA repair activity (Figure
5). However, the combination of KU0063794 and TMZ compared with TMZ alone resulted in a lengthened half-life of MGMT protein to around 40 hours. This is an important observation, as an increase in MGMT stability has the potential to greatly impede treatment of GBM with DNA methylating agents.
Previous work has used the mTORC1 inhibitor rapamycin to inhibit the mTOR/PI3K pathway in an attempt to inhibit the growth of GBM [
30‐
32]. It was therefore a logical progression to investigate specific mTOR signalling inhibition in GBM cells and their possible use in conjunction with TMZ. Even though research in this area is surprisingly limited, recent studies have indicated combination treatments involving compounds that negatively regulate the mTOR signalling pathway may impact MGMT regulation [
33], further compounding the need for research to increase the efficacy of TMZ. Indeed a number of clinical trials combining mTOR inhibitors with radiation and TMZ have been initiated and are recruiting patients (e.g. everolimus (RTOG 0913) and BKM120).
The data presented here reveal that in vitro, rather than inhibiting translation of MGMT, mTOR inhibition promotes steady state levels of the MGMT protein and counteracts the depleting effects of TMZ. We therefore conclude that further in vivo analysis of mTOR inhibition is necessary to determine if this effect is also found in patients exhibiting GBM. If so, mTOR inhibition would not be a useful adjunct to TMZ therapy in a clinical setting and could exacerbate tumour growth.
Competing interests
The author’s declare that they have no competing interests.
Authors’ contributions
Sarah Smalley, Simon Morley and Anthony Chalmers designed the research. Sarah Smalley carried out all experiments and drafted the paper. All authors read and approved the final manuscript.