Glioblastoma (GB) represents the most common and aggressive form of malignant primary brain tumour associated with high rates of morbidity and mortality. In the present study, we considered the potential use of idebenone (IDE), a Coenzyme Q10 analogue, as a novel chemotherapeutic agent for GB. On two GB cell lines, U373MG and U87MG, IDE decreased the viable cell number and enhanced the cytotoxic effects of two known anti-proliferative agents: temozolomide and oxaliplatin. IDE also affected the clonogenic and migratory capacity of both GB cell lines, at 25 and 50 µM, a concentration equivalent to that transiently reached in plasma after oral intake that is deemed safe for humans. p21 protein expression was decreased in both cell lines, indicating that IDE likely exerts its effects through cell cycle dysregulation, and this was confirmed in U373MG cells only by flow cytometric cell cycle analysis which showed S-phase arrest. Caspase-3 protein expression was also significantly decreased in U373MG cells indicating IDE-induced apoptosis that was confirmed by flow cytometric Annexin V/propidium iodide staining. No major decrease in caspase-3 expression was observed in U87MG cells nor apoptosis as observed by flow cytometry analysis. Overall, the present study demonstrates that IDE has potential as an anti-proliferative agent for GB by interfering with several features of glioma pathogenesis such as proliferation and migration, and hence might be a drug that could be repurposed for aiding cancer treatments. Furthermore, the synergistic combinations of IDE with other agents aimed at different pathways involved in this type of cancer are promising.
Glioblastoma (GB) represents the most common and aggressive form of malignant primary brain tumour and is associated with high rates of morbidity and mortality . Despite the substantial advances in neurosurgical techniques in combination with radio/chemotherapy, the median overall survival time of GB patients remains only ∼8–15 months, and it has not changed significantly over the past four decades . This is reflected in the limited success of recent phase III clinical trials making the treatment of GB one of the greatest challenges in neuro-oncology . In the attempt to improve treatment outcomes of GB patients and to increase their survival rate and quality of life, a diverse range of therapeutic strategies are being explored. These include immunotherapy, nanoparticles encapsulating anti-cancer agents, gene therapy along with the substantial need for exploring and developing new, effective and safe chemotherapeutic agents . An important prerequisite for the success of any drug for this disease is that of crossing the blood–brain barrier (BBB) even though this barrier is disrupted at the brain–tumour interface . One such compound that has been shown to cross the BBB following oral administration using 14C radiolabel in both rats and dogs is idebenone (IDE) [6,7]. IDE is currently exploited by the pharmaceutical industry to treat age-related cognitive disorders including Alzheimer's disease due to its powerful antioxidant properties [8,9], and it has recently been used with success for the treatment of several mitochondrial-related neuromuscular disorders, especially Leber's hereditary optic neuropathy and Friedrich's ataxia [10–14]. Chemically, IDE is structurally similar to the naturally occurring Coenzyme Q10 (Figure 1), in that both possess a benzoquinone moiety involved in electron transport, but their hydrophobic tails differ in length and composition. The shorter tail of IDE seems to be the ideal length for favouring partitioning into the mitochondrial membrane and for a better BBB permeation compared with Coenzyme Q10 . It, therefore, has a more favourable pharmacokinetic profile and, in some cases, is considered a better therapeutic agent than its natural analogue [16,17].
Recent research suggests that IDE may also have potential use as an anti-cancer agent. Tai et al.  studied the effect of IDE on human dopaminergic neuroblastoma SHSY-5Y cells, demonstrating that concentrations ≥25 µM were cytotoxic and that the mechanism of cell death was apoptotic in nature. Seo et al.  showed that in PC-3 prostate cancer cells and in CFPAC-1 pancreatic ductal adenocarcinoma cells, IDE reduced cell proliferation, inhibited cell migration and induced apoptosis by inhibiting anoctamin 1 (ANO1), a calcium-activated chloride channel which is significantly increased in various tumours. These are the only two studies that have specifically investigated the effects of IDE on human cancer cells to date. Both demonstrated that it was effective, highlighting the potential this compound has as an anti-proliferative agent if studied more extensively on other cancer cell lines. For these reasons, IDE could be an interesting candidate for investigation against GB.
Therefore, the aim of the study was to investigate the influence of IDE on growth, regulation and migration of two human GB cell lines, U87MG and U373MG, to determine whether IDE might be a potential new anti-cancer agent.
Materials and methods
Cell culture and reagents
Idebenone (Tocris, U.K.) and temozolomide (TMZ, Sigma, U.S.A.) were both prepared as a 100 mM stock solution in dimethyl sulfoxide (DMSO), whereas oxaliplatin (OX, Tocris, U.K.) was prepared as a 10 mM stock solution in sterile water. They were all aliquoted and stored at −20°C until use. The following antibodies were purchased from different sources: anti-p21/WAF1/Cip74 (#05-655, EMD Millipore, U.S.A.), anti-β-Actin (#Ab119716, Abcam, U.S.A.), anti-Casp3 (#HPA002643, Sigma, U.S.A.), anti-rabbit and anti-mouse IgG-horseradish peroxidase (HRP) (#sc-2004 and #sc-2005, respectively, Santa Cruz Biotechnology, U.S.A.). All other analytical grade chemicals were purchased from Sigma–Aldrich (U.S.A.).
Human GB cell lines, U373MG and U87MG, were procured from ECACC and are commonly used as models of GB harbouring a range of different genetic lesions . They were cultured in Eagle's minimum essential medium supplemented with 10% foetal bovine serum (FBS), 100 units/ml penicillin, 100 µg/ml streptomycin and 2 mM l-glutamine, under standard cell culture conditions (37°C, 5% CO2, humidified atmosphere), passaged every 4/5 days and used within 8–20 passages. Prior to each experiment, they were seeded in appropriate plates/dishes at a seeding density of 2.4 × 104 cells/cm2 and treated according to each assay protocol. Appropriate controls were included throughout including the use of maximum concentration of vehicle that the cells were exposed to which did not exceed 0.05% for DMSO. This concentration did not cause any observable harmful effects on the cells based on cell morphology and cell growth.
3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide and trypan blue exclusion assays and cell growth analysis
Cells were seeded onto 96-well plates and allowed to grow for 48 h after which medium was replaced with that containing increasing concentrations of the chosen drugs or their combinations (100 µl final volume). For combination studies, the drugs were added simultaneously. After different exposure times (24–72 h), the effects of the compounds on cell viability were determined using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay . Ten microlitres of MTT solution dissolved in phosphate-buffered saline (PBS, 5 mg/ml) were added to each well and cells were incubated for a further 3 h at 37°C. Medium was removed and replaced with 100 µl of DMSO to solubilise the crystals. The optical density of each well was determined at 570 nm on a microplate reader (Tecan Sunrise™), and viable cell count was assessed as a percentage relative to untreated cells. The IC50 values for each drug were calculated using the GraphPad Prism 7 XML Project (GraphPad Software, Inc., San Diego, CA, U.S.A.).
For cell growth analysis, cells were seeded in 35 mm cell culture dishes in 2 ml medium in duplicate and allowed to grow for 48 h, after which medium was replaced with that containing different concentrations of IDE and incubated for a further 48 h. Cells were then harvested by trypsinisation, resuspended in 1 ml PBS and counted on an automated cell counter (Beckman Z2 Coulter Particle Count and Size Analyzer). For determining the number of dead/dying cells, the trypan blue exclusion dye assay was used. The same procedure as described above for cell growth analysis was followed, except that cells were counted under light microscopy on a haematocytometer after staining with 0.1% trypan blue solution in PBS at a 10× dilution factor. The number of dead/dying cells (stained) was assessed as a percentage of the total cell number for each treatment.
Colony formation assay
The colony-forming assay was performed to evaluate the effects of IDE on the clonogenic capacity of U373MG and U87MG cells. Cells were seeded in 60 mm cell culture dishes in duplicate (200 and 400 cells, respectively/2.5 ml medium) and after 4 h (to allow cells to attach), 2.5 ml of appropriate concentrations of IDE were added to reach the desired final concentrations in 5 ml. Cells were then left in a sealed incubator for 2 weeks, after which they were carefully rinsed with PBS, fixed and stained with 2 ml 0.5% crystal violet solution in 50% methanol for 30 min, rinsed again carefully with tap water and left to dry in normal air at room temperature. Colonies containing more than 50 cells were counted as representative of clonogenic cells under a Carl Zeiss™ Stemi 2000-C stereo microscope. The surviving fraction, which is the number of colonies that arise after cell treatment expressed in terms of plating efficiency, was determined according to the formula reported in Franken et al. :
Cell migration assay
The effect of IDE on cell migration was assessed using the scratch/wound assay as previously developed by Valster et al. . The cells were grown to confluence in a six-well cell culture plate for 72 h, washed twice with medium without FBS, and a scratch was performed with the tip of a sterile 200 µl pipette tip to create a defined, uniform scratch in the centre of the well. The medium with suspended cells was removed and replaced with medium containing 0.5% FBS with or without different concentrations of IDE. Closure of the wounds by migrating cells was observed under a digital inverted microscope (EVOS XL, AMG), right after the scratch and at 24 and 48 h of incubation, and images were taken in the same field by marking the wells underneath. An average of six images/well/time point were taken, and the gap surface area of the wound was analysed using ImageJ software and expressed as the percentage of the area at time 0 h.
Western immunoblotting analysis
Cells were seeded onto 60 mm diameter dishes in duplicate, and after 48 h, the medium was replaced with different concentrations of IDE and incubated for a further 48 h. Cells were then harvested and the lysates in RIPA buffer were stored at −80°C. After thawing, brief vortexing and centrifugation at 16 000×g for 15 min at 4°C, the protein concentration was determined on the supernatant using the BCA protein assay according to the manufacturer's instructions (Pierce BCA Protein Assay Kit, Thermo Scientific, U.S.A.). Samples containing equal amounts of protein were separated by 12% SDS–PAGE, and the proteins were transferred overnight at 4°C on polyvinylidene difluoride membranes and probed for the proteins of interest as previously reported . The primary antibodies were used at the following dilutions: anti-p21 (1:750), anti-pro-caspase3 (1:500), anti-β-actin (1:5000), while the secondary ones were 1:7000 for anti-rabbit and 1:3000 for anti-mouse HRP-labelled antibodies. The protein bands were detected using an enhanced chemiluminescent substrate (SuperSignal West Dura, Thermo Scientific, U.S.A.) and captured on a Genoplex VWR Bio imager (VWR, U.S.A.). Protein bands were quantified using ImageJ software, and the data are reported as the percentage of intensity of the band of the protein of interest compared with the intensity of the β-actin band (control).
Flow cytometry analyses
Cells were seeded onto six-well plates and allowed to grow for 48 h after which medium was replaced with that containing different concentrations of IDE or camptothecin (1 µM) used as positive control, and incubated for 24 h. Cells were then harvested by trypsinisation and combined with floating cells collected from the medium, pelleted and washed with PBS. For Annexin V/propidium iodide (PI) staining, the Annexin V-FITC Kit (Miltenyi Biotec, Germany) was used and the assay was performed according to the manufacturer's protocol. Briefly, after washing cells with 1× Annexin-binding buffer, cells were resuspended in 100 µl 1× Annexin-binding buffer to which 10 µl of Annexin V-FITC was added and incubated for 15 min at room temperature. Cells were then washed with 1 ml of 1× binding buffer, and to the cell pellet resuspended in 500 µl of this same buffer, 5 µl of PI solution was added directly before measurement on a BD Fortessa flow cytometer (BD Biosciences, U.S.A.). Emission of Annexin V-FITC was detected at 530 nm and PI fluorescence was collected at 670/14 nm with excitation at 488 nm. For the detection of apoptotic events, the percentage of the population was evaluated on single cells, which are positive for AV-FITC or PI, using the FlowJo-V10 analysis software (FlowJo LLC, U.S.A.).
Cell cycle analysis was performed by measuring the changing amount of DNA associated with each phase of the cell cycle. Cellular DNA was labelled with DNA-binding fluorochrome and subsequent fluorescence was measured to determine the relative DNA content and cell cycle position. Briefly, after harvesting the cells as described above, the cell pellets were fixed with 1 ml of 70% ethanol overnight at −20°C. Cells were then pelleted at 2500 g/5 min and washed with 1 ml of phosphate–citrate buffer (0.2 M Na2HPO4, 0.1 M citric acid, pH 7.8). The cell pellet was resuspended in 100 µl of RNase A (100 µg/ml in PBS) and incubated for 15 min at 37°C before adding 400 µl of PI solution (50 µg/ml in PBS) directly to it. After 1 h incubation in the dark at room temperature, the DNA content was analysed as PI fluorescence emission at 610 nm using the 561 nm laser on the BD Fortessa analyser. Cell cycle phases were evaluated using the cell cycle module of the analysing software FlowJo-V10.
All experiments were repeated independently at least three times. IC50 values of compounds were analysed using GraphPad Prism 7 using non-linear sigmoidal curve fitting with the normalised response. In the case of OX, the Excel add-in ED50V10 was used as this gave better curve fitting for obtaining the IC50. Statistical differences were analysed using one-way ANOVA followed by Dunnett's post-hoc analysis or Student's t-test using the GraphPad Prism 7 XML Project (GraphPad Sofware, Inc., San Diego, CA, U.S.A.). Significant differences were defined as P < 0.05. Excess over bliss (EOB) analysis was performed to determine the drug combinations effect at each combination dose according to Liu et al. , where an EOB score >0 is considered synergism, =0 independent/additive and <0 antagonism.
IDE decreased viable cell number in GB cells and enhanced the anti-tumour effects of TMZ and OX
The effects of increasing concentrations of IDE on the viable cell number of U373MG and U87MG cells were determined using the MTT assay (Figure 2A,B). At concentrations of IDE ≥20 µM, a statistically significant decrease in viable cell number was observed for both cell lines at almost all exposure times compared with untreated controls. The half maximal inhibitory concentrations (IC50) at 48 h were 84.5 ± 5.2 µM for U373MG and 74.4 ± 2.7 µM for U87MG. At longer exposure times, IDE became increasingly toxic with <30% of viable cells remaining after 96 h exposure at 50 µM (the values of IC50 for U373MG at 72 and 96 h are 31.3 ± 1.9 and 41.1 ± 3.1 µM, respectively, whereas those for U87MG are 38.7 ± 1.7 and 26.6 ± 1.8 µM, respectively).
Effect of IDE, OX and TMZ on viable cell number of human glioblastoma cells.
To compare the effects of IDE with known anti-cancer agents, cells were also exposed to TMZ and OX for 48 h (Figure 2C,D). Exposure to OX leads to a statistically significant decrease in U87MG viable cell number compared with untreated cells at concentrations ≥250 µM, while, in U373MG cells, this was observed starting from 350 µM (Figure 2C). The values of IC50 for U87MG and U373MG are 342.4 ± 2.4 and 476.8 ± 12.4 µM, respectively, suggesting that U373MG cells are more resistant to OX. However, both these values are about five times higher compared with those obtained for IDE which are >70 µM for both cell lines. TMZ had little effect in both cell lines (Figure 2D). The value of IC50 was >500 µM for both cell lines which is seven times greater than that of IDE. Thus, under our experimental conditions, IDE exhibits greater toxicity than both known anti-cancer agents alone, and U373MG cells are more resistant to these drug treatments than U87MG cells. Others have also reported that U373MG cells are more resistant to drug treatment which is consistent with our data [26,27].
Since combining drugs is one of the major strategies used for improving clinical outcomes of GB , we explored whether IDE could modulate the effects of TMZ or OX. In the case of U87MG cells, the combination of IDE and OX leads to a greater decrease in cell viability than either IDE or OX alone. However, the results of EOB analysis reported in Figure 3 suggest that this combination does not have impressive synergism since all values are <0.1. In U373MG cells, the combination of 50 µM IDE and OX also leads to a greater decrease than OX alone, but not to IDE alone (Figure 3A,B). The co-presence of OX in this case seems to increase the number of viable cells compared with IDE alone. In fact, this combination appears to have an antagonistic effect according to EOB analysis. When IDE was combined with TMZ, a greater dose-dependent reduction (20–50% decrease) in cell viability was also observed than when either were used alone in both cells lines (Figure 3C,D), indicating a synergistic effect, albeit unimpressive, as found by EOB analysis.
Effect of IDE in combination with anti-cancer agents on viable cell number of human glioblastoma cells.
IDE inhibited growth of GB cells and affected their clonogenic and migratory capacity
To determine the effects of IDE on cell growth, three concentrations corresponding to a low, medium and relatively high dose (10, 25 and 50 µM) of IDE were chosen based on the results reported in Figure 2A,B (48 h). There was a decrease in total cell number for both cell lines with increasing concentrations of IDE (Figure 4). In the case of U373MG cells, IDE appears to have a cytostatic effect, especially at 50 µM since the cell number in its presence after 48 h incubation was almost identical as before IDE addition at time 0 (no growth). Instead for U87MG cells, IDE resulted more growth inhibitory since the cell number was reduced after IDE addition but always higher than the starting cell number. In the presence of IDE, there was also a modest increase in trypan blue-positive staining, indicating increased cell death in the presence of 25 and 50 µM IDE after 48 h (Table 1). This increase is consistent with the trend observed with the MTT assay.
Effect of IDE on the growth of human glioblastoma cells.
|Cell line||Idebenone (µM)||No. of dead/dying cells (% of total cell no.)|
|U373MG||0||3.67 ± 1.15|
|10||7.17 ± 1.04|
|25||9.83 ± 2.47|
|50||15.00 ± 5.12*|
|U87MG||0||7.33 ± 1.15|
|10||8.30 ± 2.89|
|25||13.60 ± 3.14*|
|50||17.83 ± 4.37*|
|Cell line||Idebenone (µM)||No. of dead/dying cells (% of total cell no.)|
|U373MG||0||3.67 ± 1.15|
|10||7.17 ± 1.04|
|25||9.83 ± 2.47|
|50||15.00 ± 5.12*|
|U87MG||0||7.33 ± 1.15|
|10||8.30 ± 2.89|
|25||13.60 ± 3.14*|
|50||17.83 ± 4.37*|
The effects of IDE on cell survival were also assessed. The treatment of GB cells with IDE reduced the surviving fraction in a dose-dependent manner. After the 2-week incubation period hardly, any colonies were observed in cells treated with 50 µM IDE (Figure 5A,B). Since the number of colonies is a reliable indicator of the survival potential of these cells, the results indicate that IDE at concentrations ≥25 µM drastically reduces the ability of GB cells to survive.
Effect of increasing concentrations of IDE on the clonogenic survival of human glioblastoma cells.
GBs are known to be highly invasive and infiltrative tumours which are hallmarks of this type of disease; therefore, the possible anti-migratory effect of IDE using the wound healing assay was also investigated [29,30]. In untreated cells, after 24 and 48 h cells migrated into the wound gap reducing its surface area (Figure 6A). In the presence of 10 and 25 µM IDE, however, cell migration diminished by 45 and 65%, respectively, for U373MG cells, and by 5 and 34% for U87MG cells, respectively, at 24 h (Figure 6B,C). At 48 h, IDE at both concentrations and in both cell lines significantly reduced cell migration compared with the untreated control at the same time point. The U87MG cells were more migratory than U373MG cells in accordance with the observation of others on these two cell lines [31,32].
Effect of IDE on migration of human glioblastoma cells.
IDE reduced the expression of caspase-3 and p21-inducing apoptosis and cell cycle arrest
The next question was does IDE induce the effects observed through cell cycle dysregulation and/or apoptosis induction? To examine this, the expression of p21 and caspase-3, respectively, was examined by Western immunoblotting (Figure 7). Under our experimental conditions, IDE does not seem to exert its effects via apoptosis, at least not in U87MG cells. In these cells, caspase-3 is more expressed than in U373MG cells, and there was no significant difference compared with the untreated control. This was also confirmed by flow cytometry analysis (Figure 8A,B) using dual staining with Annexin V/PI where no appreciable differences were observed between the control and IDE treated cells. Camptothecin used as a positive control did, however, induce apoptosis as can be observed by the significant increase in cells in late apoptosis compared with the control. In U373MG cells, a decline in caspase-3 protein expression was evident and significant at 50 µM IDE (Figure 7B). This finding was confirmed by flow cytometry (Figure 8C,D), which showed that IDE at 25 and 50 µM significantly increased the percentage of early apoptotic cells by almost 2-fold compared with the untreated control as well as the percentage of cells in late apoptosis at the highest concentration. From the results shown in Figure 7A,C, it also appears that IDE affects the cell cycle since, in both cell lines, there was a decreasing trend in the expression of p21 with an increasing concentration of IDE. This is especially the case in U373MG cells which is consistent with the flow cytometry data showing single staining with PI for DNA cell cycle content and distribution (Figure 9A,B). In this cell line, a significant dose-dependent decrease in cell population in the G1 phase (54 and 42% at 25 and 50 µM IDE, respectively, vs. 66% of control) and 1.5- to 2-fold increase in the S-phase (31 and 41% at 25 and 50 µM IDE, respectively, vs. 21% of control) were observed in the presence of IDE (Figure 9B). In U87MG cells, as observed for caspase-3, p21 was more expressed compared with U373MG cells. However, despite the significant decrease in p21 expression in these cells at 50 µM, the DNA distribution analysed by flow cytometry revealed no apparent changes in the presence of IDE in three independent experiments (results not shown).
Effect of IDE on expression of caspase-3 and p21 in human glioblastoma cells.
Annexin V-FITC/PI flow cytometric analysis of apoptosis in human glioblastoma cells.
Flow cytometric analysis of cell cycle parameters in human glioblastoma cells.
The main purpose of the present study was to investigate the potential anti-cancer effect of IDE on two human GB cell lines. GB is one of the most resistant tumours to conventional cytotoxic therapies; therefore, current studies concentrate on the development of novel agents for use either alone or in combination with standard chemotherapy and radiotherapy. In the present study, we demonstrate that IDE decreased cell viability in a time- and concentration-dependent manner and that it was cytotoxic at concentrations similar to those reported by others on both human and non-human cancer cells [18,19,33,34]. Furthermore, in a separate study, IDE had no effect on a normal cell line consisting of colonocytes (CCD841CON), whereas it proved to be cytotoxic in a colorectal cancer cell line (SW480) (results not shown). Interestingly, when IDE was co-administered with the two well-known anti-cancer agents, TMZ and OX, a greater decrease in cell viability was observed in both cell lines, especially with TMZ. This improved effect resulted marginally synergistic. Since IDE appears to enhance the cytotoxic effects of TMZ, this novel combination for GB therapy merits further investigation, especially as combinations of other drugs and natural compounds with TMZ are being explored continuously [27,31,35–40]. OX has been occasionally used for treating GB but limited due to its side effects . However, it was chosen in this study for comparison with IDE, since there have been indications recently for repurposing platinum-based chemotherapies for multi-modal treatment of GB. Hence, it could be more widely used for GB treatment in the future . Besides this aspect, IDE also proved to be more potent than both cytotoxics, with IC50 value at least five times lower than those of the known drugs. The reduced number of colonies in the gold standard colony-forming assay provided further evidence of the growth inhibitory and hence survival effects of IDE. To the best of our knowledge, this is the first report showing the ability of IDE to hamper with cell survival in the long term. In the context of preventing recurrence, this is important as the capacity for unlimited proliferation of all stem cells must be eradicated. In the present study, we also established that IDE inhibits cell migration as previously observed in prostate cancer cells . This anti-migratory/metastatic effect of IDE on GB cells could help to contain spreading of a GB tumour in vivo.
In the attempt to address the possible mechanisms underlying the effects displayed by IDE on glioma cells, we found that apoptosis is probably not a major pathway responsible for the above outcomes, at least not in U87MG cells, where no major decline in pro-caspase-3 was observed nor was there any indication from the flow cytometric data. Caspase-3 belongs to the executioner family of cysteine–aspartic acid proteases (caspases) and plays a dominant role in the hallmark caspase cascade characteristic of the apoptotic pathway . Upon activation, it is cleaved into its active 17 and 12 kDa fragments which lead to a concomitant decrease in intensity of the uncleaved band at 32 kDa during immunoblot analysis. This could explain the dose-dependent decrease in protein expression observed in U373MG cells in the presence of IDE. Indeed, in U373MG cells, IDE appears to induce modest apoptosis as also confirmed by Annexin/PI staining analysis using flow cytometry. These results are in accordance with two previous reports on the direct effects of IDE on cancer cells which both describe an apoptotic effect of IDE [18,19]. Seo et al.  attribute their observations to the fact that IDE blocks the ANO1 calcium–chloride channel, but it has no effect on cancer cells which do not express ANO1. However, in U87MG cells, we failed to observe evidence of IDE-induced apoptosis and this could be due to the p53 status of the two cell lines. p53 is a well-known tumour suppressor protein which, when active, induces many genes linked to diverse functions such as cell cycle regulation, DNA repair mechanisms and those related to apoptosis . U87MG cells have a wild-type p53 gene, but do not express the functional protein to any measurable extent because of Mdm2 overexpression which destabilises it , whereas U373MG cells have a mutant p53 gene . Lack of p53 activity in U87MG cells could thus prevent the induction of p53-dependent apoptosis whether IDE is present or not, explaining our results. In the case of U373MG, dysfunctional p53 activity due to the mutated gene would make these cells more sensitive to high concentrations of IDE which could then respond by apoptosis, as indicated by the decreased expression of pro-caspase-3 at 50 µM IDE and by the flow cytometric analysis. This divergent apoptotic response to IDE possibly due to the p53 status of the two cell lines is similar to that described by Datta et al.  on the same cells in the presence of cisplatin. The higher protein levels of caspase-3 and p21 expressed in U87MG cells, compared with U373MG cells, may also reflect this different status, similarly to the observations by Ravizza et al. for p21 .
It is more likely that under our experimental conditions, IDE exerts its anti-proliferative effects by interfering with cell cycle regulation, since, in both cell lines, a decline in protein expression of the cyclin-dependent kinase (CDK) inhibitor, p21 (known as p21WAF1/Cip1) was evident especially at high concentrations. This protein is uniquely positioned in the cell cycle to function as both a sensor and an effector of multiple anti-proliferative signals in response to a variety of cellular and environmental signals to promote tumour suppressor activities, both dependently and independently of the classical p53 tumour suppressor pathway. Usually, it is assumed that p21 down-regulation or repression increases cell cycle progression and proliferation due to disinhibition of cyclin/CDK complexes . However, this is not always the case as indicated by several reports in which p21 functions as a positive cell cycle regulator. Indeed, in U373MG cells, we observed using flow cytometry, a dose-related increase in cell population in the S-phase and a concomitant decrease in cells in the G1 phase of the cell cycle, suggesting that IDE is responsible for accumulation of cells in S-phase. A similar S-phase arrest concomitant with p21 down-regulation has also been observed by others in human cells under different treatment regimes [48–50]. During S-phase, replication can cease in response to DNA damage or stress to the replication process. However, while the former response induces arrest through different mechanisms involving ATM protein kinases and invoking p53 and p21 response, the response to replicative stress arrests all cells regardless of p53 status and is not accompanied by p21 induction . Since we do not observe IDE-induced p21 in GB cells, we expect that IDE is affecting them mainly through replicative stress. The correlation between reduced expression of p21 and impairment of cell proliferation as observed in our study has been shown in several cell models ranging from HaCaT keratinocytes , smooth muscle cells , endothelial cells , and colon and liver cancer cells [55,56] exposed to different stimuli although the reason for this has not always been clarified. The mechanism by which IDE down-regulates p21 in glioma cells remains to be elucidated. However, the evidence so far suggests that in U373MG cells, IDE-induced S-phase arrest is linked to p21 down-regulation and that this plays an important part in IDE-induced apoptosis. This is supported by the fact that in several systems, p21 down-regulation has been shown to trigger apoptosis [57–59]. In U87MG cells, despite observing p21 decrease in the presence of IDE at 50 µM, we could not link this to any changes in DNA cell cycle distribution, suggesting that our observations are cell line-specific. These differential responses between the two cell lines may depend on their p53 status as recently reviewed by Georgakilas et al.  who depicts p21 as an onco-suppressor or an onco-promotor depending on cell type, cellular localisation, p53 status, and the type and level of genotoxic stress. The fact that IDE down-regulates p21 expression in U373MG and U87MG cells, which are p53-deficient/mutant, implies that IDE could repress the oncogenic potential of these cells via p21 inhibition. Repression of p21 by IDE could also explain the anti-migratory effect observed in the present study, since p21 appears to be essential for cell migration as reported in bladder cancer cells induced by the inflammatory cytokine IL-20 . Further investigations are clearly required to understand mechanistically the effects of IDE observed in this research.
Overall, the present study demonstrates that IDE has potential as an anti-proliferative agent for GB by interfering with several features of glioma pathogenesis such as proliferation and migration. The human safety of IDE is well established, and a daily dose of 60 mg/kg/day has been shown to reach a transient concentration in plasma equivalent to 29.6 µM . This is a concentration close to those used in this study which were effective (25–50 µM). Recently, the repurposing of existing drugs has attracted considerable attention , because it is advantageous, in time and cost saving. Therefore, IDE, besides its current use in mitochondrial-related neuromuscular and neurodegenerative diseases, could be repurposed for aiding cancer treatments especially as it can cross the BBB. For example, its analogue Coenzyme Q10 has already been reported to be a promising candidate either alone or in combination for the prevention and treatment of breast cancer . Atovaquone, another CoQ10 analogue and an FDA-approved anti-malarial drug, is another example which is being considered for repurposing because of its anti-proliferative effect against MCF7 Cancer Stem-like Cells . The future treatment of malignant gliomas will likely involve synergistic combinations of agents aimed at different pathways in the molecular pathogenesis of this type of cancer. In this context, the results of the present study on IDE appear promising providing the preliminary experimental basis for exploring it further.
excess over bliss
foetal bovine serum
E.D. and H.M.W. designed the study. E.D. performed the experiments, interpreted the data and wrote the manuscript. H.M.W. interpreted the data and revised the manuscript. R.Y. helped with FACS experiments, analysis and interpretation. All authors have read and approved the final version.
E.D. was funded by an internal research grant from the Polytechnic University of the Marche provided by MIUR (Italian Ministry of University and Research). Funding for FACS experiments was provided by a grant from the University of Aberdeen.
The Authors declare that there are no competing interests associated with the manuscript.
Present address: Department of Life and Environmental Sciences, Polytechnic University of the Marche, Ancona, Italy.