Antioxidant properties of antiepileptic drugs levetiracetam and clonazepam in mice brain after in vitro-induced oxidative stress

Antioxidant properties of antiepileptic drugs levetiracetam and clonazepam in mice brain after in vitro-induced oxidative stress Aline de Albuquerque Oliveira, Maria Isabel Linhares, Adriano José Maia Chaves Filho, Emiliano Ricardo Vasconcelos Rios, Camila Nayane de Carvalho Lima, Edith Teles Venancio, Alana Gomes de Souza, Klistenes Alves de Lima, Francisca Cléa Florenço de Sousa, Danielle Macedo Gaspar and Marta Maria de França Fonteles*


INTRODUCTION
The role of free radical-mediated reactions in human neuropathology continues to attract interest. The production of free radicals is associated with injury to cell structures and the pathogenesis of many neurological disorders, such as Alzheimer's disease, Parkinson's disease, amyotrophic lateral sclerosis and epilepsy (Li et al., 2013;Kong and Lin, 2010). Over-accumulation of reactive oxygen species (ROS) caused by imbalance between the generation and elimination of these entities frequently results in severe harmful effects to cells, a *Corresponding author.   3366-8333. Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution License 4.0 International License phenomenon known as oxidative stress (OS) (Gutteridge and Halliwel, 2010;Kong and Li, 2010). Furthermore, in sufficiently high concentrations, nitrogenous compounds, mainly nitric oxide (NO), can also damage cellular structures in a process known as nitrosative stress (NS). This occurs through nitrosation of amines and thiols, which also results in an excess of oxidative species and subsequent molecular damage (Aguiar et al., 2012).
In this context, OS and NS are possible mechanisms implicated in the pathogenesis of epilepsy (Chang and Yu, 2010). Studies have already verified that status epilepticus changes redox potential and decreases the level of ATP, which can lead to a collapse in brain energy production and supply (Aguiar et al., 2012;Wasterlain et al., 1993). Moreover, it was demonstrated that an increase in mitochondrial OS and NS occurs after persistent seizures and promotes subsequent dysfunction of this organelle and cell death (Freitas et al., 2010;Ueda et al., 1997). Therefore, since mitochondrion is the main source of ATP for neurons and has an important role in the homeostasis of intracellular calcium, injury to it may strongly affect neuronal excitability and synaptic transmission (Chang and Yu, 2010). Further support for the role of free radicals in seizures comes from the successful use of exogenously administered antioxidants, as vitamin C and vitamin E, to protect the brain against seizure-induced damage Murashima 1998;Santos et al., 2009;Ueda et al., 1997).
The second-generation antiepileptic drug levetiracetam (LEV) is a new molecule that is clearly differentiated from conventional antiepileptic drugs by its pharmacologic properties and mechanisms of action (Gibbs et al., 2006;Lyseng-Williamson 2011;Ueda et al., 2009). Although the exact molecular mechanism of action of LEV remains uncharacterized, it is known that this drug binds to synaptic vesicle protein 2A. By binding to this vesicle, this drug appears to act as a modulator of synaptic vesicle exocytosis, leading to direct inhibition of pre-synaptic neurotransmitter release (Rigo et al., 2002). Moreover, LEV also has been previously demonstrated to protect against oxidative stress-induced neurotoxicity in several models of seizures Ueda et al., 2009;Zona et al., 2001).
Benzodiazepines, such as clonazepam (CNZ), are established and important agents for the treatment of seizures and epilepsy (Talarek and Fidecka, 2003). Studies have suggested the antioxidant properties of some benzodiazepines. CNZ has been shown to have a protective effect against free radical mediated brain damage in models of diabetic and depressive-like behavior in mice, which involve an oxidative stress process (Haeser et al., 2007;Wayhs et al., 2013 a ; Wayhs et al., 2013 b ). Moreover, nitric oxide synthase (NOS) inhibitors can potentiate the anticonvulsant action of benzodiazepines, which suggests the participation of NO system, and possibly of the NS, in the neuroprotective effect of these drugs (Borowicz et al., 2000;Talarek and Oliveira et al. 279 Fidecka, 2003). However, although the literature suggests that the anticonvulsant effects of antiepileptic drugs (AEDs) could be related to their antioxidant properties, the basis of these properties is still unclear. Therefore, this work aimed to investigate the antioxidant activity of the novel anticonvulsant levetiracetam and benzodiazepine anticonvulsant clonazepam. The study also aimed to investigate how these AEDs could counteract oxidative stress in brain tissues. The investigation of the antioxidant activity of these AEDs and their possible neuroprotective effect may provide stronger evidences for the hypothesis of the involvement of oxidative damage in the pathophysiology of epilepsy.

Animals
Experimentally naive, male Swiss mice Mus musculus, from the Animal House of the Federal University of Ceará, weighing 25 to 30 g, were used. The animals were maintained at a controlled temperature (24±2°C) with a 12-h dark/light cycle and food and water ad libitum. Mice were caged in groups of 8 in a 41 × 34 × 16 cm cage.
The animals were used according to the NIH Guide for the Care and Use of Laboratory Animals. The experiments were performed after approval of the protocol by the Ethics Committee on Animal Research of the Federal University of Ceará (with protocol number 59/07), in accordance with the Ethical Principles in Animal Research adopted by the Brazilian College of Animal Experimentation (COBEA). Moreover, the experiments were performed to minimize the number of rats and their suffering, following the ethical doctrine of the three "R"s: Reduction, Refinement and Replacement.

Assessment of antioxidant capacity in vitro
Antioxidant activity of the compounds was evaluated by measuring the production of thiobarbituric acid (TBARS), as an indicator of lipid peroxidation, reduced glutathione (GSH) levels and catalase activity. In order to assess the effects of these drugs on nitric oxide production, nitrite-nitrate levels were determined. Mice were decapitated and the brains were removed rapidly under standard conditions at 4°C. The whole brain, apart from the cerebellum, was homogenized in 50 mM potassium phosphate buffer (pH 7.4) and the concentration adjusted to 1 g wet weight of brain per 60 ml. Then, 250 l of the brain homogenates was maintained in the absence or presence of LEV (50, 100 or 200 g/ml), CNZ (50, 100 or 200 g/ml) or vitamin C (VIT C) (200 g/ml) for 24 h at 10°C. Following this, oxidative stress was induced by incubation of the brain homogenates for 1 h at 37°C with samples without oxidative stress acting as controls (Auddy et al., 2003;Mattei et al., 1998;Stocks et al., 1974).

Measurement of TBARS
The homogenates were incubated in a water bath for 1 h at 37°C. After incubation, assessment of the levels of thiobarbituric acid concentration was performed according to the method described by Houng et al. (1998), with absorbance measured at a wavelength of 532 nm and expressed as micromol of malondialdehyde (MDA)/mg of protein. The protein concentration was determined by using Lowry assay (Lowry et al., 1951).

Nitrite-nitrate determination
For the assessment of nitrite-nitrate, derived from nitric oxide (NO), 100 l of Griess reagent (1% sulfanilamide dissolved in 5% H3PO4, 0.1% N-(1-naphthyl)-ethylenediamine dihydrochloride and distilled water, 1:1:1:1) was added to 100 l of brain homogenate or to 100 l of NaNO2 at concentrations ranging from 0.75 to 100 M (standard curve) or for the blanks to 100 l of the buffer used in the homogenates. The absorbance was measured with a reader plate at 560 nm (Green et al., 1981).

Measurement of catalase activity
This was measured by using hydrogen peroxide to generate H2O and O2. The substrate mixture contained 0.3 ml of hydrogen peroxide in 50 ml of 0.05 M phosphate buffer, pH 7.0. The sample aliquot (20 l) diluted with phosphate buffer was added to 980 l of the substrate mixture. Absorbances were read after 1 and 6 min at 230 nm. A standard curve was established using purified catalase (Sigma MO, USA) and the results were expressed as M/min/mg protein (Chance and Maehly, 1955). Protein measurement was performed with the Lowry assay (Lowry et al., 1951).

Determination of glutathione (GSH) levels
GSH levels were evaluated to estimate endogenous defenses against oxidative stress. The method was based on the reaction of Ellman's reagent (DTNB) with free thiol groups. The measurement of this parameter was performed according to the test steps of Sedlak and Lindsay (1998). GSH concentration was determined by the absorbance at 412 nm and was expressed as g of GSH/g of protein. The protein concentration was determined with the Lowry assay (Lowry et al., 1951).

Statistical analysis
Data were expressed as mean ± standard error of mean (SEM). Analysis of variance (ANOVA) was used for the statistical analysis followed by Student-Newman-Keuls test to identify differences between experimental groups. Values of p < 0.05 were considered significant. The results of VIT C group was compared with internal groups for each antiepileptic drug, LEV or CNZ, studied in this work.
No alteration in GSH levels was observed in the samples after the in vitro induced oxidative stress. An increase in GSH level was shown at the lowest concentration of CNZ (50 g/ml) compared to the control group (p < 0.05). An increase in GSH was also demonstrated with the two lowest tested concentrations of LEV (50 and 100 g/ml) as compared to the control (p < 0.0001) and to the stress control (p<0.001). An increase in GSH levels also occurred in samples with VIT C as compared to the control (p < 0.001) and to the stress control (p<0.001). Results of GSH levels: LEV (Control: 34.23+ 3.14; Stress control: 38.   Nitrite-nitrate level in mice brain homogenates submitted or not (control) to oxidative stress in the absence (stress control) or presence of different concentrations of levetiracetam (50, 100 or 200 g/ml) or VIT C (200 g/ml), after 1 h of incubation of brain homogenate at 37ºC. Each bar represents the mean ± SEM, n=8, ***p<0.0001 vs. control group; § § § p<0.0001 vs. stress control group (one-way ANOVA followed by post-hoc Student-Newman-Keuls test).  Catalase activity in mice brain homogenates submitted or not (control) to oxidative stress in the absence (stress control) or presence of different concentrations of levetiracetam (50, 100 or 200 g/ml) or VIT C (200 g/ml), after 1 h of incubation of brain homogenate at 37ºC. Each bar represents the mean ± SEM, n=8, ***p<0.0001 vs. control group; § § § p<0.0001 vs. stress control group (one-way ANOVA followed by post-hoc Student-Newman-Keuls test).

DISCUSSION
In this study, to further investigate the antioxidant properties of levetiracetam and clonazepam, these drugs were tested in an established and specific in vitro model of oxidative stress, proposed by Stocks et al. (1974) and modified by Mattei et al. (1998) and Huong et al. (1998). This model is based on the fact that brain homogenates autoxidize spontaneously and reproducibly when heated to 37°C for 1 h, and this preparation may be used to assay the antioxidant effect of several agents in neuronal substrate (Huong et al., 1998;Mattei et al., 1998;Stocks et al., 1974). Moreover, to verify the potential antioxidant capacity of these AEDs in this model, the results obtained were compared with these drugs with the powerful watersoluble antioxidant vitamin C (VIT C). Previous studies have demonstrated not only the well-recognized antioxidant effects of VIT C, but also its anticonvulsant action, suggesting that this drug behaves as a potential neuronal protective agent, capable of attenuating the behavioral and pro-oxidative changes involved in development of seizures (Santos et al., 2009;. Our results showed, as expected, that samples submitted to this in vitro stress model showed an increase in MDA and nitrite-nitrate concentrations and higher levels of catalase activity when compared with control samples not submitted to heating. These findings clearly indicate that an oxidative process occurred in brain homogenates submitted to this model. Furthermore, the elevation in free radical formation can be accompanied by a rapid compensatory increase in the activity of free radical scavenging enzymes (Ferreira and Matsubara, 1997;Bellissimo et al., 2001). The high level of catalase activity observed in our results suggests the development of an adaptive mechanism against oxidative stress and further supports that this process occurs in this model.
In a similar fashion to VIT C, treatment with LEV and CNZ, at all doses tested, was able to prevent the elevation of MDA levels induced by in vitro stress model, with values of MDA similar to those observed in the control group. Regarding nitrite-nitrate concentrations, Figure 8. GSH increased especially with treatment with the lower concentration tested of CNZ and with VIT C. GSH level in mice brain homogenates submitted or not (control) to oxidative stress in the absence of different concentrations of clonazepam (50, 100 or 200 g/ml) or VIT C (200 g/ml), after 1 h of incubation of brain homogenate at 37ºC. Each bar represents the mean ± SEM, n=8, *p<0.05,***p<0.0001 vs. control group; § § § p<0.0001 vs. stress control group (oneway ANOVA followed by post-hoc Student-Newman-Keuls test).
both AEDs demonstrated the capacity to reduce the changes induced by heating, mainly at the lowest doses. VIT C was also able to reduce considerably the nitritenitrate concentrations in brain homogenates. Catalase activity was restored to control levels with pre-incubation with LEV and CNZ, and the same effect was observed in samples incubated with VIT C. Finally, the GSH amount was increased by the incubation with both AEDs, especially in lowest doses, and also by incubation with VIT C, when compared with groups submitted to stress procedure.
The degree of lipid peroxidation in tissues is commonly measured by determining the amount of mean product of this process, MDA, and this analysis may relate directly to the level of ROS-mediated injury to a tissue (Kunz et al., 2008;Schihiri, 2014). Free radical-mediated lipid peroxidation proceeds by a chain mechanism, that is, once it was initiated, free radicals can oxidize both lipid molecules in biological membranes and low density lipoproteins. This phenomenon has been shown to induce disturbance to membrane organization and functional loss/modification of proteins and DNA (Frantseva et al., 2000;Schihiri, 2014).
The generation of oxidative damage appears to result from excitotoxicity mechanisms already recognized for participating in several pathological conditions that lead to the development of seizures (Dubinsky et al., 1995;Duchen, 2000). In this context, overactivation of the glutamate receptors, especially the NMDA receptors, would lead to an increase in intracellular calcium, resulting to mitochondrial dysfunction and generation of ROS and pro-apoptotic factors, fundamental for the damage and cell death (Patel, 2004).
In this context, it has been found that NO is produced in response to N-methyl-D-aspartate (NMDA) receptor stimulation and may be involved in the modulation of neuronal damage (Ferrer et al., 2000;Mülsch et al., 1994). Moreover, the literature has suggested that seizures induce alterations in NO metabolism, increasing the production of its reactive metabolites (nitrite and nitrate). These metabolites, in turn, can interact with NMDA glutamatergic receptors and potentiate its excitotoxicity action on the CNS (Santos et al., 2009). Therefore, elevated nitrite-nitrate levels may be implicated in lipid peroxidation and others oxidative damages on brain tissues submitted to injury situations, as seizures.
Interestingly, according to our findings, LEV and CNZ, similarly to VIT C, was able to reduce the oxidative stress in brain homogenates by decreases of both TBARS and nitrite-nitrate concentrations. Together, these results reinforces the participation of nitrergic pathway, mainly through NO, in oxidative damage in brain tissues and the ability of these antiepileptic drugs, at least in part, through their antioxidant properties to protect the neurons of oxidative damage.
Furthermore, there is an endogenous ROS scavenging enzyme system, involving the cooperative action of superoxide dismutase (SOD), catalase and glutathione peroxidase (GSH-Px) enzymes. The oxidative stress involved, for example, in status epilepticus and seizures can alter the activity of these enzymes, such catalase, promoting an adaptive cellular response to increased free radicals milieu (Ferreira and Matsubara, 1997;Freitas et al., 2004;Bellissimo et al., 2001). In this context, preincubation with LEV and CNZ, preserved the catalase activity in control levels. VIT C was also able to preserve the catalase activity at levels similar to controls samples. These results indicated a possible reduction in toxic reactive substrates in the samples incubated with both AEDs and suggest that these drugs, similarly to a powerful exogenous antioxidant, could help the brain cells to counteract the stress-induced reactive oxygen species overproduction and the oxidative damage.
In the brain, GSH is thought to play a central role in defense against reactive oxygen species. This substance can either directly detoxify reactive oxygen species as can act as a substrate for several peroxidases (Dean et al., 2011). Under conditions of overproduction of free radicals or a deficiency of antioxidant systems, GSH is consumed and its levels fall (Wang and Cynader, 2000). In the present work, no alterations in GSH concentrations were observed in the group of samples submitted to oxidative stress compared to control. However, an increase in this biomarker was induced by pre-incubation with the two lowest tested concentrations of LEV and by the lowest tested concentration of CNZ. An important augment in GSH amount also occurred in samples treated with VIT C. This effect may be probably attributed the capacity of these compounds to avoid the depletion of endogenous antioxidants reserves by the reactive species. Additionally, some of these drugs may potentiate the biosynthesis of brain antioxidants, such as GSH, increasing their amounts. In this context, according to Al-Shorbagy et al. (2013), LEV would be able to up-regulate the action of cystine/glutamate exchanger and to increase the cysteine concentrations in glial cells, the limiting substrate to production of GSH (Al-Shorbagy et al., 2013;Arakawa and Ito, 2007;Kau et al., 2008).
Our findings demonstrate no concentration dependent effects of both LEV and CNZ on the parameters investigated. In fact, in general, the lowest concentrations tested present the highest antioxidant capacity and this property became distinctly smaller with the higher concentration of LEV and CNZ. These results suggest that high concentrations of these drugs could inhibit part of the endogenous antioxidant defense. In accordance with this, previous reports pointed to pro-oxidative effects related to long-term therapy or to administration of high doses of AEDs. In this context, increase in TBARS and decrease in the levels of thiols were detected in blood samples from patients under long-term AED therapy (Alshafei et al., 2013;Devi et al., 2008;Tan et al., 2009). Therefore, this study reinforces that the prooxidant/antioxidant balance in brain is modulated by antiepileptic drugs. In addition, the ability of these compounds to reduce brain damage caused by seizures and their biochemical changes (that is., markers of oxidative stress) further supports the involvement of free radicals in seizure generation and highlights the possible role for antioxidants as adjuncts to antiepileptic drugs for better seizure control (Devi et al., 2008).

Conclusion
In the present work, it was demonstrated, by an in vitro model, the antioxidant ability of two antiepileptic drugs, with different mechanism of action, and that this property is mediated, at least in part, by reduction of lipid peroxidation and nitrite-nitrate contents, preservation of catalase activity at control parameters and increase of GSH levels. These results allow us to infer the participation of this effect in the neuroprotective mechanism of these drugs, and reinforce, even indirectly, the hypothesis of oxidative damage in the pathophysiology of epilepsy.
Additionally, our findings also suggest a potential therapeutic use of antioxidant compounds as alternative or complementary tools to the conventional treatment for this disease. However, more studies need to be performed to elucidate the relation between antioxidant activity and protection against seizure development and epilepsy.