Neuroscience Letters 492 (2011) 124128

Contents lists available at ScienceDirect

Neuroscience Letters
journal homepage: www.elsevier.com/locate/neulet

Gabapentin, A GABA analogue, enhances cognitive performance in mice

Ipek Komsuoglu Celikyurt , Oguz Mutlu, Guner Ulak, Furuzan Yildiz Akar, Faruk Erden
Kocaeli University, Department of Pharmacology, Faculty of Medicine, Psychopharmacology Laboratory, 41380, Kocaeli, Umuttepe, Turkey

a r t i c l e

i n f o

a b s t r a c t
Gabapentin is one of the new antiepileptic drugs (AEDs) launched recently. The advantage of new AEDs includes newer mechanism of action, broad spectrum of antiseizure effects, lesser drug interactions and fewer side effects. Gabapentin (GBP) a GABA analogue, is efcacious in several neurological and psychiatric conditions and it is conventionally used in the treatment of partial epilepsies. In this study, we aimed to evaluate the effects of GBP on learning and memory processes of naive mice in Morris water maze (MWM), passive avoidance (PA) and modied elevated plus maze (mEPM) tests. GBP (5 and 10 mg/kg, i.p.) was administered on the probe trial of MWM and on the acquisation session of PA and mEPM tests. In the MWM test, GBP (10 mg/kg) signicantly increased the time spent in target quadrant and GBP (5 and 10 mg/kg) signicantly decreased the distance to platform compared to control group. In the mEPM test, GBP (5 and 10 mg/kg) signicantly decreased the transfer latency compared to control group on the second day and in the PA test, GBP (5 and 10 mg/kg) signicantly prolonged retention latency compared to control group. Our results indicate that GBP has improving effects on spatial and emotional cognitive performance of naive mice in MWM, PA and mEPM tasks. 2011 Elsevier Ireland Ltd. All rights reserved.

Article history: Received 10 December 2010 Received in revised form 16 January 2011 Accepted 27 January 2011 Keywords: Morris water maze Passive avoidance Modied elevated plus maze Gabapentin Mice Memory

Antiepileptic drugs (AEDs) are a major treatment consideration for patients with epilepsy. The main concern in choosing an appropriate antiepileptic is efcient control of seizures. The most prevalent of the cognitive side effects observed in an antiepileptic drug therapy in clinical use are sedation, sommolence, distractibility, insomnia and diziness. Gabapentin, (1-(aminonomethy) cyclohexaneatic acid) (GBP), a GABA analogue was originally developed for the treatment of epilepsy and is widely used to relieve neuropathic pain [18]. It is also efcacious in migraine, tremor, social phobia and bipolar disorder [26]. GBP prevents seizures in a wide variety of animal models including generalized tonic-clonic and partial seizures [23] and currently in clinical use in patients with partial seizures resistant to other AEDs [11]. GBP prevents pain responses in several animal models of hyperalgesia and prevents neuronal death in vitro and in vivo by models of neurodegenerative disease like amyotrophic lateral sclerosis. Although GBP may have several different pharmacological actions, modulation of GABA and glutamate synthesis may also appear to be important [23]. Cognitive effects of GBP and other AEDs have been compared in a number of clinical studies. GBP as add-on treatment has been shown to have favorable effects on cognition in clinical studies in humans. In a doubleblind, add-on, crossover study of 27 patients

Corresponding author. Tel.: +90 262 303 71 02; fax: +90 262 303 80 03. E-mail address: ikomsu@hotmail.com (I.K. Celikyurt). 0304-3940/$ see front matter 2011 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.neulet.2011.01.072

with refractory partial seizures, it is reported that GBP has no negative side effects on cognition except for an increase in drowsiness at 2400 mg/day but not at lower doses [9]. Cognitive side effects associated with old antiepileptics are moderate and can frequently be clinically signicant. In a double-blind, randomized, crossover study comparing GBP to carbamazepine (CBZ) in healthy adults with 5 weeks treatment arms found statistically better performance for GBP, but none for CBZ [13]. In contrast, a 12-week, randomized, double-blind, parallel-group study of CBZ and GBP in healthy volunteers using quantitated EEG and a cognitive battery found no signicant differences between CBZ and GBP [19]. In a randomized double blind triple crossover study in patients with complex partial epilepsy, there is no cognitive differences found between carbamazepine and phenytoin, and phenobarbital produces greater cognitive differences than the two other drugs [12]. In a double-blind, randomized, placebo-controlled parallel study in healthy volunteers, the effects of topiramate and GBP on cognitive abilities were evaluated and GBP was signicantly better than topiramate on 50% of the 24 variables [20]. Cognitive effects of rst and second generation AEDs and comparative studies for the effects of them were also performed in animal models. It is reported that valproic acid reduces spatial working memory and cell proliferation in the hippocampus [27] and valproate administered after traumatic brain injury provides neuroprotection and improves cognitive function in rats [7]. In a study, the effect of carbamazepine and lamotrigine on cognitive function and oxidative stress in brain during chemical epilepto-

I.K. Celikyurt et al. / Neuroscience Letters 492 (2011) 124128

125

genesis in rats were investigated and the results showed that lamotrigine did not produce any change in cognitive function, while carbamazepine produced cognitive dysfunction [2]. There are few systematic data on the effects of GBP on specic cognitive domains. Altough preliminary results are promising with GBP for its effect on memory storage, the effects of GBP still remains to be explored on spatial and emotional learning and memory functions in mice. The aim of this study was to investigate the effects of GBP on spatial and avoidance cognitive performance in Morris water maze (MWM), passive avoidance (PA) and modied elevated plus maze (mEPM) tasks in naive mice. GBP was purchased from Sigma Chemical Company (Sigma, St. Louis, MO) and dissolved in saline. GBP was freshly prepared and administered by intraperitoneal (i.p.) route in a volume of 1 ml per 10 g body weight of mice. Control groups received the same volume of vehicle. Doses that do not affect locomotor activity of animals were selected and previous literature taken into consideration [1]. GBP was administered intraperitoneally (i.p.) 30 min before the tests. Male BALB/c mice (Istanbul University Research Center, DETAM, Turkey) weighing 3545 g were housed vesix per home cages (L 30 W 20 h 12.5 cm) in an animal colony facility for 2 weeks before the start of the experiment. The animals were maintained in constant room temperature (22 2 C) under a 12-h light/dark cycle (light onset at 07:00 h). Tap water and food pellets were provided ad libitum. All animals were naive to tests. Each mice was tested individually and only once. Experiments were conducted between 10:00 and 14:00 h. All procedures were in compliance with the European Community Council Directive of 24 November 1986 and ethical approval was granted by the Ethics Committee of Kocaeli University. (Number: KOU/HADYEK-6/1-13, Kocaeli, Turkey). MWM was a circular pool (90 cm diameter and 30 cm height) that was lled to a depth of 14 cm with water (22 C) and rendered opaque by the addition of small black balls. The pool was located in a semi-soundproof, dimmed test room illuminated with a table lamp (80 lux), a number of extra-maze visual cues, including a whiteblack-colored poster on the wall, a camera and the experimenter. Maze was divided into four quadrants, and three equally spaced points were used as starting positions around the edge of the pool. The order of release positions varied systematically throughout the experiment. An escape platform (6 cm diameter and 12 cm height) was located in one quadrant 1 cm above the water surface during the familiarization session and 1 cm below the water surface during other sessions. Videotracking conducted with a videocamera focused on the full diameter of the pool. Navigation parameters were analyzed using the ETHOVISION 3.1 video analysis 3 system (Noldus). Mice were trained during ve daily sessions (S1, S2, S3, S4, S5). One familiarization and four acquisition sessions were performed. During familiarization session and acquisition phase, each mice was given three trials. Delay between trials was 60 s, and a 1day interval was used between sessions. For each daily trial, mice was taken from the home cage and placed into the maze at one of three randomly determined locations with its head facing the center of maze. After the mouse had found and climbed on to the platform, trial was stopped and escape latency was recorded. If the mouse had not climbed onto the platform in 60 s, the trial ended automatically and the experimenter guided the mouse to the platform and an escape latency of 60 s was recorded [6]. Twenty-four hours after the last acquisition session, a probe trial was used to assess the mouse spatial retention of the location of the hidden platform. During this trial, platform was removed from the maze and the mouse was allowed to search the pool for 60 s. The percent of time spent in each quadrant was recorded. Distance to platform and time spent in escape platforms quadrant analyses were used as measures to determine spatial memory,

whereas the swim speed parameter was used to assess locomotor function. Experiments were conducted in a dimly lit, semi-soundproof room, illuminated with table lamp (80 lux). Transfer latency was used as an index of learning and memory [17]. Maze was made of wood and consisted of two open (29 cm long 5 cm wide) and closed arms (29 cm 5 cm with 15 cm high walls) forming a square cross with a 5 cm square center piece. In order to avoid falls the open arms was surrounded by a short (1 cm) plexiglass edge. The maze was elevated 40 cm above the oor. The open arms and central platform were painted white and enclosed arms were painted black. The animals were randomly assigned to different experimental and control groups. In the acquisition session (on day 1), each mice was gently placed at the distal end of an open arm of the apparatus facing away from central platform. The time it took for the mice to move from the open arm to either of the enclosed arms (transfer latency) was recorded. Training (repeated exposure of animal to open arms) shortens this parameter, possibly as a consequence of learning acquisition and retention. If the mice did not enter the enclosed arm within 90 s, it was excluded from further experimentation. The criterion of an animals entry into the enclosed arm was crossing with all four legs of an imaginary line separating the enclosed arm from the central space. After entering the enclosed arm, the mice was allowed to move freely in the maze regardless of open and enclosed arms for 10 s. Then, the mice was returned to its home cage. Retention session followed 24 h after the acquisition session (on day 2). The mice were put into the open arm and the transfer latency was recorded again. Animals were trained in a one-trial, step-through, PA apparatus for evaluating memory based on contextual fear conditioning and instrumental learning [15]. Decrease in retention latency indicates an impairment in memory in the PA task. The apparatus consisted of a box with an illuminated part (L 7 12.5 h 14 cm) and a dark part (L 24 12.5 h 14 cm), both equipped with a grid oor composed of steel bars (0.3 cm diameter) spaced 0.9 cm apart. The inhibitory avoidance task consisted of two trials. On the rst day of training, mice were placed individually into the light compartment and allowed to explore the boxes. The intercompartment door was opened after a 60-s acclimation period. In the acquisition trial, each mouse was placed in the illuminated compartment, which was lighted by a bright bulb (2000 lux). The animals received drugs prior to acquisition training. If the mouse stepped into the dark compartment (2/3 of the tail in the dark compartment), the door was closed by the experimenter and an inescapable foot shock (0.25 mA/1 s) was delivered through the grid oor of the dark compartment. A cutoff time of 5 min was selected. The time taken to enter the dark compartment (training latency) was recorded. Immediately after the shock, the mouse was returned back to the home cage. The retention trial started 24 h after the end of the acquisition trial. Each mouse was placed in the illuminated compartment as in the training trial. The door was opened after a 30-s acclimation period. The step-through latency in the retention trial (with a maximum 300 s cutoff time) was used as the index of retention of the learned experience. Shock was not applied at the retention trial. One-way analysis of variance (ANOVA) post-hoc Tukey test was used to analyse PA and MWM tests data. Wilcoxon t-test (rst day vs. second) and KruskalWallis ANOVA followed by Dunns test (vs. control group on second day) were preferred for the analyses of mEPM test. Data are expressed as the mean values SEM. p < 0.05 accepted as statistically signicant. GBP (5 mg/kg) had no effect on the time spent in target quadrant. GBP (10 mg/kg) increased the time spent in target quadrant compared to control (n = 910, p < 0.01 vs. control group ANOVA posthoc Tukey, F(2, 28) = 7.26, p = 0.0031) (Fig. 1a).

126

I.K. Celikyurt et al. / Neuroscience Letters 492 (2011) 124128

Fig. 2. Effects of GBP (5, 10 mg/kg, i.p.) administration on the transfer latencies (of rst day and second day) to the enclosed arm of mEPM test in mice. GBP or saline was administered 30 min before the rst days trial.

Fig. 1. (a) Effects of GBP (5, 10 mg/kg, i.p.) administration on the time spent in target quadrant in MWM test in naive mice. (b) Effects of GBP (5, 10 mg/kg, i.p.) administration on distance to platform in MWM test in naive mice. (c) Effects of GBP (5, 10 mg/kg, i.p.) administration on swim speed in MWM test in naive mice.

Fig. 3. (a) Effects of GBP (5, 10 mg/kg, i.p.) on the rst day latency of mice in passive avoidance task. (b) Effects of GBP (5, 10 mg/kg, i.p.) on the retention latency of mice during retention test of passive avoidance task.

GBP (5, 10 mg/kg) administration decreased the distance to platform compared to control (n = 910, p < 0.001 (vs. control) ANOVA posthoc Tukey, F(2, 28) = 49.33, p < 0.0001) (Fig. 1b). GBP (5, 10 mg/kg) did not affect the swim speed in MWM test compared to control (n = 910, ANOVA posthoc Tukey, F(2, 28) = 1.8, p = 0.18) (Fig. 1c). Mean TL of GBP (5, 10 mg/kg, i.p.) or vehicle (on rst day and second day), given 30 min before acquisition session in the mEPM in mice are presented in Fig. 2. GBP had no effect on the TL on the rst day (TL1) compared to that of the vehicle treated group (n = 810, KW = 5, 02, p = 0.08, KruskalWallis followed by Dunns test). 2 days TL signicantly decreased compared to 1 day TL within control and GBP groups

indicating that the mice remembered the presence of the enclosed arms in mEPM n = 810, p < 0.001 (rst day vs. second day, Wilcoxon t-test). On the 2nd day, GBP signicantly decreased the TL2. p < 0.01, p < 0.001 (vs. control on the second day), KruskalWallis ANOVA followed by Dunns test, KW = 17, 79, p = 0.0001). Fig. 3a shows the effect of GBP (5, 10 mg/kg, i.p.) on the rst day latency (STL) of mice in PA task. GBP did not affected rst day latency on training session of PA task (n = 910, F(2, 27) = 0.81, p = 0.45, ANOVA followed by post-hoc Tukey test) Fig. 3b. shows the effect of GBP (5, 10 mg/kg, i.p.) on the retention latency during retention test of PA task. GBP signicantly increased the retention latency (sec) on 2nd day compared to control and GBP (10 mg/kg) signicantly increased the retention latency (sec)

I.K. Celikyurt et al. / Neuroscience Letters 492 (2011) 124128

127

on 2nd day compared to GBP (5 mg/kg) p < 0.01, p < 0.001 vs. control, p < 0.01 (vs. 10 mg/kg), ANOVA posthoc Tukey (n = 910, F(2, 27) = 29, 716, p < 0.0001). We reported in this study that the acute and intraperitoneal administration of GBP (5, 10 mg/kg) positively affects spatial cognition in mEPM and MWM tests and facilitates emotional learning-memory function in PA test in naive mice; without affecting locomotor activity. The MWM test was described twenty years ago to investigate spatial learning and memory in laboratory animals and has become one of the most frequently used laboratory tools in behavioral neuroscience. In this experiment, animals have to nd behavioral strategies for swimming away from the pool wall and recognizing the platform as the only rescue from water. Spatial learning in general and MWM performance in particular appear to depend upon the coordinated action of different brain regions and neurotransmitter systems constituting a functionally integrated neural network. The MWM test has often been used in the validation of rodent models for neurocognitive disorders and the evaluation of possible treatments of them [5]. In this study, GBP (10 mg/kg) increased the time spent in target quadrant and GBP (5, 10 mg/kg) decreased the distance to platform without affecting spontaneous locomotor activity of animals which corresponds to a positive effect on spatial cognition. The elevated plus-maze test has a spatial component since the animals must remember the conguration of the open and enclosed arms and they should escape from the unsafe open arm more rapidly on the second trial. Shortened transfer latency on the second day is used as a parameter for retention or consolidation of memory while treatment with drugs prior to the rst day affects task acquisition [21,25]. In this study, second day transfer latencies signicantly decreased compared to rst day transfer latencies within control and GBP (5, 10 mg/kg) groups indicating that the mice remembered the presence of the enclosed arms in mEPM. PA behaviour was studied in a one trial learning, step-through type task utilizing the natural preference of mice for a dark environment. The PA is based on instrumental learning, and is a hippocampal and amygdala dependent test which shows long term emotional memory [14]. In this test, the animals learn to avoid an inescapable electrical shock, and longer retention latencies indicate a better learned experience. In this study, GBP (5, 10 mg/kg) dose dependently enhances learning and memory performance in PA test in mice. One trial PA procedure represents a model of learning where the effects of drugs can be studied by applying them at different stages of learning and memory process [24]. In a study, impairment on retention performance of an inhibitory avoidance task in mice induced by repeated administration of GBP, affected memory retrieval but not memory consolidation and it is reported that this impairment may be attributable to a reduction on central cholinergic activity [3]. Regarding different types of memory improvement by GBP [1,3,4], there may be some differences between the data processing pathways of memories and retention performance obtained in PA task. There are some reported data concerning the effects of GBP on cognition in animal models. Intraperitoneal administration of GBP facilitates retention performance of a one trial step-through inhibitory avoidance response in mice [1]. In a study with CF-1 mice, using an inhibitory avoidance task, GBP enhanced retention performance. In that study, these effects were prevented by atropine (neither methylatropine nor mecamylamine, or hexamethonium prevented the effects of posttraining GBP) and were not inuenced by neostigmine; suggested that central muscarinic cholinergic mechanism has a possible involvement in memory improving actions of GBP [4]. It is already known that, AEDs affect cognition by suppressing neuronal excitability or enhanc-

ing inhibitory neurotransmission [16]. GBP is a cyclized analogue of GABA but it does not interact with GABA receptors, nor does it inhibit GABA uptake or prevent the degradation of GABA. However, in vivo, GBP increases the GABA accumulation in the rat brain [10]. The exact mechanism of GBP action still remains to be investigated, but its therapeutic action on neuropathic pain is thought to involve voltage-gated N-type calcium ion channels. It is thought to bind to the 2 subunit of the voltage-dependent calcium channel in the central nervous system. The receptor prole is known as the direct blockade of subunits of voltage gated calcium channels and leads to reduction of release of multiple neurotransmitters [8]. GBP treatment alters the metabolism or concentration of glutamate, glycine or GABA in brain tissues. Also interact with the alpha2 delta auxiliary subunit of voltage gated calcium channels. GBP may modulate certain types of calcium currents [22] and apart from all these knowledge studies should be performed in order to clarify by which pharmacological mechanism does GBP use among these while affecting different types of cognitive performance in naive and epileptic animals. Cognitive functions in individuals with epilepsy may be inuenced by several factors among which underlying pathology, seizure type, and the detrimental effects of AEDs are most important. Therefore, in this regard, the importance of our study is that we try to evaluate the effects GBP on spatial and emotional learning and memory in non-epileptic (naive) mice which is independent of the disease or epilepsy itself. In conclusion GBP enhances cognitive performance in MWM, mEPM and PA tasks in naive mice. Further studies should be performed to investigate the effects of chronic GBP treatment and compare GBP with at least one conventional antiepileptic drug in the same tasks both in naive animals and in epilepsy-models. References
[1] G.B. Acosta, M.M. Boccia, C.M. Baratti, Gabapentin, an antiepileptic drug, improves memory storage in mice, Neurosci. Lett. 279 (2000) 173176. [2] T. Arora, A.K. Mehta, K.K. Sharma, P.K. Mediratta, B.D. Banerjee, G.R. Garg, A.K. Sharma, Effect of carbamazepine and lamotrigine on cognitive function and oxidative stress in brain during chemical epileptogenesis in rats, Basic Clin. Pharmacol. Toxicol. 106 (2010) 372377. [3] M.G. Blake, M.M. Boccia, G.B. Acosta, C. Hcht, C.M. Baratti, Opposite effects of a single versus repeated doses of gabapentin on retention performance of an inhibitory avoidance response in mice, Neurobiol. Learn. Mem. 87 (2007) 192200. [4] M.M. Boccia, G.B. Acosta, C.M. Baratti, Memory improving actions of gabapentin in mice: possible involvement of central muscarinic cholinergic mechanism, Neurosci. Lett. 311 (2001) 153156. [5] R. Brandeis, Y. Brandys, S. Yehuda, The use of the Morris Water Maze in the study of memory and learning, Int. J. Neurosci. 48 (1989) 2969. [6] M. Cachard-Chastel, S. Devers, S. Sicsic, M. Langlois, F. Lezoualch, A.M. Gardier, C. Belzung, Prucalopride and donepezil act synergistically to reverse scopolamine-induced memory decit in C57Bl/6j mice, Behav. Brain Res. 187 (2008) 455461. [7] P.K. Dash, S.A. Orsi, M. Zhang, R.J. Grill, S. Pati, J. Zhao, A.N. Moore, Valproate administered after traumatic brain injury provides neuroprotection and improves cognitive function in rats, PLoS One 5 (2010) e11383. [8] A. Davies, J. Hendrich, A.T. Van Minh, J. Wratten, L. Douglas, A.C. Dolphin, Functional biology of the alpha (2) delta subunits of voltage-gated calcium channels, Trends Pharmacol. Sci. 28 (2007) 220228. [9] J.P. Leach, J. Girvan, A. Paul, M.J. Brodie, Gabapentin and cognition: a double blind, dose ranging, placebo controlled study in refractory epilepsy, J. Neurol. Neurosurg. Psychiatry 62 (1997) 372376. [10] W. Lscher, D. Hocack, C.P. Taylor, Gabapentin increases aminooxyacetic acidinduced GABA accumulation in several regions of rat brain, Neurosci. Lett. 128 (1991) 150154. [11] M.J. McLean, B.E. Gidal, Gabapentin dosing in the treatment of epilepsy, Clin. Ther. 25 (2003) 13821406. [12] K.J. Meador, D.W. Loring, K. Huh, B.B. Gallagher, D.W. King, Comparative cognitive effects of anticonvulsants, Neurology 40 (1990) 391394. [13] K.J. Meador, D.W. Loring, P.G. Ray, A.M. Murro, D.W. King, M.E. Nichols, E.M. Deer, W.T. Goff, Differential cognitive effects of carbamazepine and gabapentin, Epilepsia 40 (1999) 12791285. [14] S. Monlen, A. Urquiza, M. Carmen Arenas, C. Vinader-Caerols, A. Para, Chronic administration of uoxetine impairs inhibitory avoidance in male but not female mice, Behav. Brain Res. 136 (2002) 483488.

128

I.K. Celikyurt et al. / Neuroscience Letters 492 (2011) 124128 [22] A. Stefani, F. Spadoni, G. Bernardi, Gabapentin inhibits calcium currents in isolated rat brain neurons, Neuropharmacology 37 (1998) 8391. [23] C.P. Taylor, Mechanisms of action of gabapentin, Rev. Neurol. (Paris) 153 (1997) 3945. [24] L. Trneckov, S. Hynie, P. Sda, Z. Hlink, I. Krejc, V. Klenerov, Effects of stress and of amphetamine on passive avoidance conditioning in rats, Gen. Physiol. Biophys. 24 (2005) 129142. [25] F. Yildiz Akar, G. Ulak, P. Tanyeri, F. Erden, T. Utkan, N. Gacar, 7-Nitroindazole, a neuronal nitric oxide synthase inhibitor, impairs passive-avoidance and elevated plus-maze memory performance in rats, Pharmacol. Biochem. Behav. 87 (2007) 434443. [26] L.T. Young, J.C. Robb, G.M. Hasey, G.M. MacQueen, I. Patelis Siotis, M. Marriott, R.T. Joffe, Gabapentin as an adjunctive treatment in bipolar disorder, J. Affect. Disord. 55 (1999) 7377. [27] J. Umka, S. Mustafa, M. ElBeltagy, A. Thorpe, L. Latif, G. Bennett, P.M. Wigmore, Valproic acid reduces spatial working memory and cell proliferation in the hippocampus, Neuroscience 166 (2010) 1522.

[15] S.O. Ogren, C. Johansson, O. Magnusson, Forebrain serotonergic involvement in avoidance learning, Neurosci. Lett. 58 (1985) 305309. [16] P. Ortinski, K.J. Meador, Cognitive side effects of antiepileptic drugs, Epilepsy Behav. 311 (2004) 153156. [17] D.S. Reddy, S.K. Kulkarni, Possible role of nitric oxide in the nootropic and antiamnesic effects of neurosteroids on aging- and dizocilpine-induced learning impairment, Brain Res. 799 (1998) 215229. [18] J.M. Rosenberg, C. Harrell, H. Ristic, R.A. Werner, A.M. de Rosayro, The effect of gabapentin on neuropathic pain, Clin. J. Pain 13 (1997) 251255. [19] M.C. Salinsky, L.M. Binder, B.S. Oken, D. Storzbach, C.R. Aron, C.B. Dodrill, Effects of gabapentin and carbamazepine on the EEG and cognition in healthy volunteers, Epilepsia 43 (2002) 482490. [20] M.C. Salinsky, D. Storzbach, D.C. Spencer, B.S. Oken, T. Landry, C.B. Dodrill, Effects of topiramate and gabapentin on cognitive abilities in healthy volunteers, Neurology 64 (2005) 792798. [21] A.C. Sharma, S.K. Kulkarni, Evaluation of learning and memory mechanisms employing elevated plus-maze in rats and mice, Prog. Neuropsychopharmacol. Biol. Psychiatry 16 (1992) 117125.