Background
Brain tumors are among the most disabling and resource-consuming diseases in neurology [
1‐
4]. A common clinical manifestation of brain tumors is brain tumor-related epilepsy; in younger patients, with a long life expectancy, epilepsy is often the major problem, whereas in patients with high grade gliomas, seizures are part of a more complex scenario involving cognitive deterioration and motor/sensory deficits [
5‐
7].
The management of brain tumor-related epilepsy must keep in consideration issues pertaining to both epileptology and neuro-oncology. Side effects of anti-epileptic drugs (AED) have in fact a relevant role; indeed, enzyme-inducing anticonvulsant drugs (EIAED) may interfere with the metabolism of chemotherapeutic agents [
8,
9]. Moreover, antiepileptic drugs per se may display an activity on brain cancer cells, as suggested by in vitro experiments [
10].
In 2009 Jaeckle had analysed prospectively the EIAED use in correlation with the outcome in patients with newly diagnosed glioblastoma (GBM). They found that overall survival (OS) and progression-free survival (PFS) were significantly longer in patients receiving EIAED compared to patients receiving non-EIAED [
11].
On the other hand, in 2011 Weller et al. examined the impact of the interaction between AED use and chemo-radiotherapy on survival in patients with newly diagnosed GBM. The OS of patients who were receiving AED at baseline versus not-receiving any AED was similar. However, patients receiving valproic acid (VPA) alone appeared to derive more survival benefit from chemo-radiotherapy (HR 0.39, 95% CI: 0.24–0.63) than patients receiving EIAED (HR 0.69, 95% CI: 0.53–0.90) or patients not receiving any AED (HR 0.67, 95% CI: 0.49–0.93) [
12].
In this context, particular attention has been devoted to two non-EIAED: VPA, known for a long time as a histone deacetylase inhibitor [
13] and levetiracetam (LEV).
In 2013, Guthrie investigated the role of VPA as an antitumor agent in the management of patients with GBM. The results showed that patients treated with AED had a significantly longer survival than those who were not. Moreover, patients receiving VPA had a significantly longer survival than both those who did not receive AED and those who received other AED [
14].
Kerkhof studied the effect of VPA on survival of patients with newly diagnosed GBM. The group using VPA and temozolomide (TMZ) during at least 3 months had a significantly longer median survival compared with what observed in the group not using VPA or using another AED (69 vs 61 weeks) [
15].
As far as LEV is concerned, in 2010 Bobustuc hypothesized that AED may modulate O-6 methylguanine-DNA methyltransferase (MGMT), a DNA repair protein that has an important role in tumor cell resistance to alkylating agents, and LEV was reported as the most potent MGMT inhibitor among several AED [
10].
Recently, Kim analysed the benefit of LEV compared with other AED as a chemosensitizer to TMZ for patients with GBM [
16]. The median PFS and OS for patients who received LEV in combination with TMZ were significantly longer than those for patients who did not receive LEV (6.7 vs 9.4 months and 16.7 vs 25.7 months respectively).
The possible actions of antiepileptic drugs on brain cancer cells include activity on cell proliferation, apoptosis and migration [
17]. A number of intracellular pathways are involved in these activities, among which microRNA (miRNA) are gaining increasing attention. Many data suggest that miRNA are key components of a wide range of biological processes [
18‐
20] and in a recently published study, in which miRNAs expression profiling was analyzed in a cohort of patients affected by low grade-gliomas, miR-196b has been identified as a predictive marker of seizures occurrence [
21].
Among drugs recently introduced in the management of epilepsy, both lacosamide (LCM) and brivaracetam (BRV) are devoid of enzyme-inducing activity on the cytochrome system, being good candidates for introduction in the management of brain tumor epilepsy. Lacosamide has an inhibitory activity on histone deacetylase [
22], making it worthwhile investigating its in vitro effects on brain cancer, while brivaracetam, a parent compound of levetiracetam, might share with it the same biological effects and both have been recently described to have high brain permeability [
23,
24].
In the present study, we investigated in vitro on a human glioma cell line, the effects of brivaracetam and lacosamide on biological parameters involved in tumor growth, resistance to therapies and invasiveness.
Methods
Cell cultures
The human glioma cell lines U87MG, SW1483 and T98G and human fibroblasts were purchased from ATCC (LGC Standards S.r.l.). Cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco) supplemented with 10% foetal bovine serum (FBS, Gibco) and 1% penicillin/streptomycin (Gibco).
Human umbilical vein endothelial cells (HUVECs) primary cultures were isolated from healthy donors [
25] and used in experiments up to the tenth passage. Purity of cell cultures was higher than 95% as assessed by flow cytometry after CD31 staining (polyclonal phycoerythrin-conjugated goat anti-human CD31, BD Bioscience). HUVECs were cultured in complete Endothelial Basal Medium (Lonza) supplemented with growth factors and antibiotics (EGM Single Quots, Lonza).
Drugs
BRV and LCM were kindly provided by UCB Pharma. Drugs were dissolved in distilled water at a concentration of 100 mM for BRV and 50 mM for LCM and successively diluted in complete medium to the necessary experimental concentrations (BRV: 100-200-400-600-800-1000-1200-1800-2000 μM; LCM: 100-200-400-600-800-1000-1200-1600-2000-2400 μM).
Cytotoxicity
Cytotoxicity of BRV and LCM was studied by cell proliferation assay following the manufacturer’s protocol (MTS assay, Molecular Probes).
Inhibitory concentrations (IC) were calculated from the regression line that associates percentage of inhibition and drug concentrations as follows: 100-(100 x average n.cells x C/n. cell basal level). Where C = drug concentration [range 0–2500 μM].
Apoptosis
Evaluation of apoptotic cells was performed in treated cells (IC20 BRV or IC20 LCM at 24–48–72 h) and in untreated cells (control) using Annexin V binding assay (Immunostep) for flow citometry (FacsVantage SE, Becton Dickinson) following the manufacturer’s protocol.
Chemoresistance
Treated and untreated cells were fixed and permeabilized using Cytofix/CytopermTM Fixation/Permeabilization Kit (BD Biosciences) for the evaluation of the expression of chemoresistance-associated molecules. Cells were then stained with the following antibodies: mouse-antihuman MRP1 (Monosan), MRP3 (Abcam), GSTπ (LSbio), P-gP (Chemicon) and rat–anti-human MRP5 (Kamiya) primary antibodies for 1 h at 4 °C and subsequently 30 min at 4 °C with a secondary conjugated antibody for flow cytometry: goat-anti-mouse fluorescein isothiocyanate-conjugated antibody (BD Biosciences) for MRP1, MRP3, GSTπ and P-gP and goat anti rat-biotin streptavidin phycoerythrin conjugated antibody for MRP5. Cells were analyzed by flow-cytometry.
Cell cycle
Cells (8 · 10^5) were plated in triplicate and cultured for 24 h. LCM 300 μM (IC20) and BRV 400 μM (IC20) were then added for 72 h. At each time-point cells were harvested and fixed in ethanol 80% at 4 °C for 30 min, washed and stained with PI 50 μg/ml in PBS overnight at 4 °C. DNA content was evaluated using flow-cytometry.
MiRNA expression
Total RNA was extracted using the TRIZOL (Gibco). The concentration and purity of total RNA were assessed using a Nanodrop TM 1000 (Nanodrop Technologies). Total RNA (100 ng) was labelled and hybridized to Human miRNA Microarray V.19 (Agilent). Scanning and image analysis were performed using the Agilent DNA Microarray Scanner (P/N G2565BA) equipped with extended dynamic range (XDR) software according to the Agilent miRNA Microarray System with miRNA Complete Labeling and Hyb Kit protocol manual. Feature Extraction Software (Version 10.5) was used for data extraction from raw microarray image files using the miRNA_105_Dec08 FE protocol.
Formaldehyde cross-linking and chromatin immunoprecipitations were performed as previously described. The chromatin solution was immunoprecipitated with anti-H4K8ac (Cell Signaling), anti-H3K9me3 (Cell Signaling).
Array analysis was performed using Matlab (The MathWorks Inc.). Signals were extracted using Agilent Feature Extraction, quantile normalized and log2-trasformed. Paired and unpaired T-test were applied to evaluate significantly deregulated miRNAs. For signature selection we considered as significant p values less than 0.01. A False Discovery Rate procedure for multiple comparisons was also included in the analysis. Hierarchical Clustering and Principal Component Analysis were used to evaluate the efficacy of the selected signature.
Target prediction was assessed by using several prediction software included in the web server tool MirWalk2.0 (
http://zmf.umm.uni-heidelberg.de/apps/zmf/mirwalk2/). Prediction was considered reliable if confirmed by at least three different software. Predicted targets were used for pathway analysis.
qRT-PCR analysis
10 ng of RNA was reverse-transcribed using the TaqMan microRNA Reverse Transcription Kit (Applied Biosystem) and Real time-PCR of miR expression was carried out using ABI Prism 7000 Sequence Detection System (Applied Biosystems). The PCR Reactions were initiated with a 10 min incubation at 95 °C followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. RTq-PCR quantification of miRNA expression was performed using TaqMan MicroRNA® Assays (Applied Biosystems) according to the manufacturer’s protocol. RNU48 was used as endogenous control to normalize microRNA expression. All reactions were performed in duplicate.
Transfection
For mature miR-195-5p or miR-107 expression, we used Pre-miRNA Precursor-Negative Control (Ambion) and Pre-miRNA195-5p (Ambion) or Pre-miRNA107 at final concentration of 5nM. For miR-195-5p and miR-107 depletion we used miRCURY LNA microRNA inhibitor control (Exiqon) and hsa-miR-195-5p miRCURY LNA (Exiqon) or hsa-miR-107 miRCURY LNA (Exiqon) at final concentration of 10nM. U87MG cells were transfected using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer’s instructions. For miRNAs depletion experiments, after 48 h of transfection cells were treated with IC20 BRV or IC20 LCM for 48 h.
Immunoblotting analysis
Cells were lysed in buffer consisting of 50 mM Tris-HCl pH 8, with 1% NP-40 (Igepal AC-630) 150 mM NaCl, 5 mM EDTA and fresh protease inhibitors. Protein concentrations were determined by colorimetric assay (Bio-Rad). Western blotting was performed using the following primary antibodies: mouse monoclonal anti-Tubulin (Santa Cruz Biotechnology), mouse monoclonal anti-Gapdh (Santa Cruz Biotechnology), rabbit polyclonal anti-p21 (Santa Cruz Biotechnology), rabbit polyclonal anti-Cyclin A (Santa Cruz Biotechnology), mouse monoclonal anti-Cyclin E (Santa Cruz Biotechnology), rabbit monoclonal anti-EGFR (Cell Signaling Tecnology, C74B9), rabbit polyclonal anti-N-Cadherin (Abcam). Secondary antibodies used were goat anti-mouse and goat anti-rabbit conjugated to horseradish peroxidase (Santa Cruz Biotechnology).
Cell proliferation assay
U87MG cells (6 × 104) were transfected in triplicated as indicated. Cells were collected and counted at 0–24–48–72 h after transfection.
Migration assay
Migration was measured using a 24-well plate with a non-coated 8-mm pore size filter in the insert chamber (BD Falcon). Cells were transfected with Pre-miRNA Precursor-Negative Control or the Pre-miRNA107, or the Pre-miRNA195-5p (Ambion), or treated with BRV or LCM at IC20. After 48 h from transfection or treatments, cells were resuspended in DMEM medium without FBS and seeded into the insert chamber. Cells were allowed to migrate for 12 h into the bottom chamber containing 0.7 ml DMEM medium containing 10% FBS in a humidified incubator at 37 °C in 5% CO2. Migrated cells that had attached to the outside of the filter were visualized by staining with DAPI and counted.
Statistical analysis
Statistical analyses were performed by Pearson correlation coefficient for cytotoxicity assay and by Student-t test for apoptosis, molecular analysis and cell cycle. Unless differently specified, level of significance was set at p < 0.05.
Discussion
Epilepsy is a frequent complication in patients with brain tumors, therefore AED are widely used for seizure control in addition to surgery, chemotherapy and radiotherapy. For this reason a possible intrinsic antineoplastic effect of these drugs would be useful. In addition to the anticonvulsant mechanism, some AED have been previously described to exert a cytotoxic effect that, added to the effects of conventional chemo-radiotherapy, possibly impacts also on the survival of these patients [
10,
16,
17]. Our results show that BRV and LCM in vitro exert a dose-dependent cytotoxic effect on various glioma cell lines and this effect was concomitant with the modulation of a number of miRNAs.
Although we could not completely clarify the mechanism of action, our data suggest that, even at low doses (IC20), the two drugs exert a role in blocking cell cycle progression of glioma cells possibly trough up-regulation of miR-195-5p. Indeed, in our experimental conditions, these two drugs seem to impinge on G1/S phase transition of cell cycle, as demonstrated by reduction in cyclin A protein levels and increase in p21 and cyclin E protein levels after treatments.
The tumor-suppressor activity of miR-195-5p in glioma cells and other tumor models has been previously characterized [
29‐
31]. Zhang et al. reported that overexpression of miR-195-5p was able to induce the arrest of cell cycle progression in G1/S transition in U87 [
31] and similar results have been described by Hui et al. in other cellular models of human glioma [
29]. An increase in p21 expression has been also previously reported to be associated to cell cycle arrest in lung, colorectal, thyroid and kidney cell lines [
32‐
34]. In agreement with these data, our results confirm the role of miR-195-5p in the suppression of cell proliferation in human glioma cells.
Other AED have been previously reported to interfere with cell cycle in cancer cells [
35‐
37]. Bobustuc and colleagues reported that levetiracetam inhibited in vitro human glioma cell proliferation through p53-mediated MGMT inhibition, thus increasing glioma cell sensitivity to temozolomide [
10]. Moreover, in a recent clinical study, Kim and colleagues showed that patients with glioblastoma receiving temozolomide-based chemotherapy and levetiracetam for seizure control, experienced a significant survival benefits [
16]. Our in vitro data suggest that some characteristics of the parent drug levetiracetam might be shared by BRV, however further study is needed to verify the effect of BRV on MGMT expression.
Moreover, our results show that BRV and LEV may modulate other miRNAs. Over-expression of miR-107 has been also demonstrated to inhibit glioma cell growth increasing apoptosis [
28] and to inhibit cell migration/invasion [
26,
27]. In our experimental conditions after treatment with LCM or BRV, the percentage of apoptotic cells was barely detectable therefore too low to be responsible of the observed cytotoxicity. On the other hand, our data confirm that overexpression of miR-107 reduces the ability of glioma cells to migrate in an in vitro assay.
Exposure to BRV and LCM seems to have no impact on chemoresistance. Indeed ATP-dependent drug efflux transporters or multidrug resistance proteins have been localized in different tissues including the blood brain barrier and they affect absorption, distribution and excretion of different drugs [
38‐
41]. Although preliminary, our results on endothelial and glioma cells suggest that the administration of these two drugs would not significantly modify glioma cells chemoresistance or the availability of chemotherapic drugs at the blood/tumor interface.
The meaning of the decreased expression of N-cadherin and EGFR following miR-107 overexpression deserves further investigation.
In U87 cells the decrease in N-cadherin protein levels correlates with a reduced migratory ability of the cells upon ectopic expression of miR-107 as upon treatment with BRV or LCM. The role of N-cadherin in the migration of mesenchymal cells is not completely defined. Recently, Guo et al. observed a reduction in the expression of mesenchymal related protein, such as N-cadherin, that determined a mesenchymal-to-epithelial transition of U87 and a consequent inhibition of cell migration [
42]. This is in line with our observations.
Amplified expression of EGFR and of its mutated variant v3, has been extensively studied as a possible target for anti-tumor therapy, although clinical trials focused on this treatment approach have so far yielded unsatisfactory results [
43]. A positive correlation between EGFR expression and migration ability has been previously reported in glioma cell lines and in neuronal stem cells therefore our results would further support the role of EGFR in cell motility [
44,
45].
Conclusions
Even if at concentration higher than those recommended for epilepsy control, our study provides evidence for possible effects of two novel anti-epileptic drugs, both devoid of enzyme-inducing activity, on glioblastoma cell proliferation/migration, at least partly mediated by enhanced expression of 2 miRNAs.
Further studies are needed to better delineate the multiple biological effects of these drugs in glioma and their possible impact on clinical management and outcome in brain tumor-related epilepsy.
Acknowledgments
The authors wish to thank Dr. Chiara Calatozzolo and Dr. Fiona Zucchetti for their precious help in cytotoxicity tests and in discussing the results.