Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Epilepsy is a collection of disease states with an incidence rate estimated at 20 to 70/100,000 per year and a point prevalence rate of 4 to 10/1000 in the general population (Shorvon, 1990). Despite the optimal use of available AEDs such as PB, PHT, CBZ and VPA through which most epileptic patients can have their seizures adequately controlled (Beghi and Perucca, 1995; Chadwick, 1994; Mattson, 1992; Pellock, 1994), many people with epilepsy fail to experience seizure control. It is generally estimated that 30% of epilepsy patients are refractory to available AED treatment (cf., Brodie and Dichter, 1996), and 15% should be considered for surgical treatment (Mattson, 1992).

Achievement of seizure control is restricted in part by the intrinsic central nervous system side effects of AEDs. This a considerable problem in the treatment of seizures in terms of compliance (Mattson, 1992). Current AEDs such as barbiturates, phenytoin and benzodiazepines also induce behavioral problems and cognitive disturbances in the treatment of children with epilepsy (Kälviäinen et al., 1996; Trimble and Cull, 1988). Furthermore, most treatment-resistant epilepsy patients (Mattson, 1992) are affected by temporal lobe and encephalopathy-related seizures (Gastaut et al, 1975). Thus, despite the development of novel AEDs such as felbamate, gabapentin, LTG and topiramate, all of which display some efficacy in these patients, new AEDs with more potent anticonvulsant activity and reduced side effects are still required.

PNU-151774E (fig. 1) is a structurally novel compound identified from an anticonvulsant program that was based on chemical modifications of -amino amides (Maj et al., 1993; Pevarello et al., 1998). It was selected from several derivatives for further investigation on the basis of its original chemical structure, promising activity in the MES test and favorable separation of anticonvulsant doses from those producing ataxia in the rotorod test.

View larger version (7K): [in this window] [in a new window]   Fig. 1.   Chemical structure of PNU-151774E, [(S)-(+)-2-(4-(3-fluorobenzyloxy) benzylamino) propanamide, methanesulfonate].

The aims of this study were to characterize the anticonvulsant effects and safety of PNU-151774E in several standard seizure models in rodents. The MES test as well as various chemically induced seizures (BIC, PIC, 3-MPA, PTZ, STRYC) were used to examine the efficacy of PNU-151774E in models of generalized seizures. In collaboration with the Anticonvulsant Screening Project (NIH-NINDS), the activity of PNU-151774E also was evaluated in the PTZi.v. infusion model of absence seizures. Behavioral toxicity and ataxia of this compound were evaluated with use of the rotorod test. The anticonvulsant activity of PNU-151774E, as well as its effect on motor function and ataxia, were compared with CBZ, PHT, VPA, DZP and LTG. Potential neurotoxic and cognitive side effects of PNU-151774E were evaluated on changes of spontaneous locomotor activity and passive avoidance responding.

The results of these studies show that PNU-151774E is an anticonvulsant in all electrically and chemically induced maximal seizure models. This activity of PNU-151774E was associated closely with brain levels of unchanged drug. PNU-151774E also showed low potential to induce locomotor or cognitive impairments. The high doses required to induce locomotor impairments compared with anticonvulsant doses in various seizure tests indicate that PNU-151774E has a wide therapeutic margin. PNU-151774E was not effective in blocking i.v. infused PTZ-induced seizures.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Experimental animals. Male Crl: CD-1(ICR)BR mice (19-21 g, Charles River, Como, Italy), male Crl: (WI)BR Wistar rats (190-210 g, Charles River, Como, Italy) and male Cr1: (CD)SDBR Sprague-Dawley rats (210 g, Charles River, Como, Italy) were used in studies carried out by Pharmacia & Upjohn. Animals were housed in groups of 6 to 10 and kept in a temperature (21°C ± 1°C) and relative humidity (60%) controlled room on a 12-h light/dark cycle (lights on between 0600 and 1800 h). Male albino CF No 1 mice (18-25 g, Charles River, Wilmington, MA) were used as experimental animals in studies carried out by the NIH-NINDS. All animals were maintained on a 12-h light/dark cycle and allowed free access to both food (Agway Prolab animal diet, NIH-NINDS and Mucedola rodent type 4RF21, Pharmacia & Upjohn) and water, unless otherwise indicated.

Maximal electroshock test. Wistar rats received an electroshock (160 mA for 0.2 s with a pulse train of 60 Hz having a pulse duration of 0.4 ms; ECT unit model 7801, Ugo Basile, Comerio, Italy) through intra-aural clip electrodes sufficient to produce a hindlimb tonic extensor response in at least 97% of control animals. Mice received a 28 mA shock of 0.7 s with a pulse train of 80 Hz having a pulse duration of 0.4 ms. Several doses of PNU-151774E and standard AEDs were administered to groups of 10 to 20 mice or rats per dose in a volume of 5 ml/kg 60 min p.o. or 30 min i.p. before induction of MES to calculate ED₅₀ values. The complete suppression of the hindlimb tonic extensor component of seizures was taken as evidence of anticonvulsant activity.

Chemically induced seizures. Mice were given doses of convulsant drugs that could induce tonic-extensor convulsions in at least 97% of controls. The doses used were: BIC, 0.6 mg/kg; PIC, 6 mg/kg; 3-MPA, 60 mg/kg; STRYC, 0.55 mg/kg; and PTZ, 85 mg/kg. Several doses of PNU-151774E and standard AEDs were administered orally to calculate ED₅₀ values in a volume of 5 ml/kg to groups of 10 to 20 mice 60 min before either i.v. administration of BIC and STRYC or s.c. injection of PIC or 3-MPA. PTZ also was administered s.c., and the test compounds were administered i.p. to mice 30 min before PTZ administration. Mice were replaced in single cages and observed for 30 min for the loss of righting reflexes lasting at least 3 s after PTZ treatment (cf., Loscher et al., 1991).

Evaluation of tolerance development potential. Groups of 10 to 20 mice (Pharmacia & Upjohn) and groups of 8 rats (NIH-NINDS) were used. Mice were treated orally for 4 consecutive days with the mouse MES ED₉₀ of PNU-151774E (20 mg/kg) calculated from data obtained at Pharmacia & Upjohn or vehicle. On the 5th day, the ED₅₀ was re-calculated after PNU-151774E treatment in both groups of mice treated either with PNU-151774E or vehicle. Repeated treatment with the anti-MES ED₉₀ of PB (25 mg/kg i.p.) was used as a positive control.

Behavioral ataxia and determination of TI. Groups of 10 to 20 mice trained to remain on a 3-cm diameter rod which rotated at 10 rpm (Ugo Basile, Varese, Italy) were used to evaluate the ability of PNU-151774E and standard AEDs to produce ataxia. These compounds were administered p.o. in a volume of 5 ml/kg 60 min before being placed on the rotorod. The number of animals falling during a 2-min test period were monitored. The toxic dose for each compound causing 50% of the animals to fall from the rotorod was calculated (TD₅₀). The TD₅₀ was related to the ED₅₀ of each anticonvulsant in the MES and chemically induced tonic extensor seizure tests to calculate the TIs (TD₅₀/ED₅₀) relative to each compound.

Spontaneous locomotor activity. Groups of 10 Wistar rats were used to examine the effects of PNU-151774E on spontaneous locomotor activity. The rats were housed singly 24 h before activity measures. The rats were treated orally in a volume of 10 ml/kg body weight with either PNU-151774E (10, 100, 400 and 700 mg/kg), LTG (4, 40, 160 and 200 mg/kg), the positive control haloperidol (1 mg/kg) or vehicle 60 min before the recording session. LTG was used as a reference standard in this as well as the passive avoidance test (see below) because this drug produces fewer psychomotor effects than other AEDs in clinical trials (Cohen et al., 1985). Testing was done between 1000 and 1500 h. Horizontal and vertical locomotor activity was monitored for 30 min with the Digiscan Animal Monitor (Omnitech Electronics, Columbus, OH) system consisting of four independent cages equipped with infrared movement-detecting sensors. Horizontal and vertical activity was defined as the total number of beam interruptions throughout a 30-min observation period divided in 5-min bins.

Passive avoidance. Groups of 10 Wistar rats were used to examine the effects of PNU-1551774E on passive avoidance responding. The rats were housed individually 24 h before training in a passive avoidance procedure. The rats were treated orally with PNU-151774E (10, 100, 400 mg/kg), LTG (4, 40, 160 mg/kg) or vehicle. The doses of PNU-151774E and LTG represent approximately 40 times the oral ED₅₀ of each in the MES test. MK-801 was used as a positive control and was administered i.p. (0.1 mg/kg) in a volume of 10 ml/kg body weight. Sixty (PNU-151774E) or 30 (MK-801) min after treatment, the rats were placed in the illuminated (20 W, 200 lux) portion of the passive avoidance apparatus consisting of a light and a dark chamber connected by a semielliptical doorway. Fifteen seconds after being placed in the light chamber, the door separating the two chambers was opened and the rats were allowed to enter and escape from the dark chamber where they received a mild scrambled foot-shock (0.3 mA, Coulbourn Instruments, Lehigh Valley, PA). The rats were replaced in their home cages at the end of this training session. A recall session was carried out 24 h later by the same procedure except that the rat was not foot shocked upon re-entering the dark chamber. The step-through latency required to re-enter the dark compartment was measured. A 300-s time limit was imposed for the training and recall time. The system and data collection were controlled by Spider Process Control software (Paul Frey, Ltd, Cambridge, England).

Plasma and brain levels. Thirty-six Sprague Dawley rats were fasted overnight before being given a single oral dose of PNU-15177E (10 mg/kg) in an aqueous solution (dose volume, 2 ml/kg). Food was returned to the animals 4 h after dosing, and tap water was available ad libitum throughout the study. Groups of three rats were sacrificed under ether anesthesia by bleeding from the abdominal aorta at 0.17, 0.33, 0.5, 1, 2, 4, 6, 8, 12, 16, 24 and 48 h after PNU-151774E. Blood was collected into heparinized tubes and immediately centrifuged at 1500 × g for 20 min at 4°C. Plasma was separated and frozen at 20°C until assayed. The brain was removed rapidly, washed in saline solution, dried on blotting paper, weighed and stored at 20°C pending analysis. The hypophysis and pineal gland were excised before weighing. PNU-151774E was determined in plasma and brain by a HPLC method with fluorimetric detection. PNU-151774E was extracted from 1 ml of plasma with diethyl ether (twice, 4 ml), then back-extracted with 0.25 ml of 25 mM H₃PO_4. After washing with 1 ml of n-hexane (twice), 200 µl of aqueous phase was submitted for HPLC analysis. The whole brain was homogenized in 3 ml of acetone and after centrifugation at 3000 × g for 8 min, the extraction step was repeated on the pellet. The acetone phase, or an aliquot, was evaporated to dryness at 37°C (N₂ stream). The residue was reconstituted with 0.6 ml of 0.15 M H₃PO₄ and washed three times with 1 ml of n-hexane. After neutralization, the aqueous phase was extracted with diethyl ether and then back-extracted with H₃PO₄ as described above for plasma. The HPLC system consisted of a SP8700 XR pump equipped with a Rheodyne sampling valve with a 200-µl loop (Thermo Separation Products, Fremont, CA), a Jasco model 821-FP fluorescence detector (Jasco, Tokyo, Japan) and a Chromject Integrator (Thermo Separation Products). Detection was performed at 224 nm (excitation) and 302 nm (emission). The limits of quantification were 10 ng/ml in plasma, and 20 ng/g in brain (wet tissue). The chromatographic separation was carried out with a 250 × 4 mm Erbasil S3 C18 column (3 µm) (Farmitalia Carlo Erba, Milan, Italy). The mobile phase was CH₃CN/50 mM KH₂PO₄, pH 2 (32:68, v/v); flow rate, 0.5 ml/min. PNU-151774E was eluted at about 16 min.

Drugs. In studies carried out by Pharmacia & Upjohn, PNU-151774E methanesulfonate (Department of Medicinal Chemistry, Pharmacia & Upjohn S.p.A., Milan Italy) was dissolved in distilled water. PHT (Fluka AG, Buchs SG, Switzerland), CBZ (Sigma Chemical Co., St. Louis, MO), DZP (Fabbrica Italiana Sintetici, Alte Montecchio, Italy), LTG base (Wellcome Research Laboratories, Beckenham, Kent, England), haloperidol (Sigma Chemical Co.) were suspended in methylcellulose. LTG isothionate (Department of Medicinal Chemistry, Pharmacia & Upjohn S.p.A.), VPA sodium (Sigma Chemical Co.), STRYC sulfate (Serva, New York), PIC (BDH, Milan, Italy), 3-MPA (Janssen, Geel, Belgium), PTZ (Sigma Chemical Co.), PB (Farmaceutica Milanese, Milan, Italy), (+) MK-801 (Research Biochemical, Inc., Natick, MA) were dissolved in distilled water. BIC (Fluka AG) was dissolved in 0.01 N HCl and brought to volume with distilled water. PNU-151774E was administered in either 0.9% saline or 0.5% methylcellulose in studies carried out by NIH-NINDS. PTZ was dissolved in 0.9% saline. Doses of PNU-151774E and reference standards are expressed in their salt-free form.

Statistical analysis. Probit analysis (Finney, 1964) was used to calculate the ED₅₀ of each compound (95% confidence interval) for the MES, rotorod and chemically induced tests. Mean values and the S.E.M. were calculated for the PTZi.v. test and P values were determined by the Student's t test. Locomotor activity data were analyzed by a split plot analysis of variance (ANOVA. Unistat, Version 3. Unistat Ltd, London, UK) for each compound with Dose being the independent factor and Time being repeated (Kirk, 1968). The significance of the differences between means was evaluated by Student-Newman-Keuls multiple comparisons. Passive avoidance data were analyzed according to a one-way ANOVA. The significance of the difference between means was evaluated by Dunnett's test for multiple comparison. The data are represented in terms of means ± S.E.M. Statistical levels of significance are indicated as * P < .05 and ** P < .01.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Maximal electroshock test. The data in table 1 describe the anticonvulsant activity of PNU-151774E and reference standards the MES test. PNU-151774E showed dose-dependent anticonvulsant activity in both mice and rats with ED₅₀ values of 4.1 and 6.9 mg/kg or 8.0 and 11.8 mg/kg after i.p. or p.o. administration in mice and rats, respectively. This anticonvulsant activity was more potent than or similar to that of the classical anticonvulsants, CBZ, PHT and VPA. PNU-151774E, however, was less potent than DZP or LTG.

Time course of MES anticonvulsant activity and plasma and brain concentrations of PNU-151774E. The peak effect of an oral dose of 10.0 mg/kg of PNU-151774E lay between 30 and 60 min after administration, and considerable activity was still observed 4 h after (fig. 2). Brain concentrations of PNU-151774E were approximately 10 times higher than plasma levels at all times. Maximal levels of PNU-151774E in both plasma and brain were already evident by 15 min after administration. Peak levels of 37 µM were reached in the brain between 30 and 60 min. These times correspond to the maximal anticonvulsant effect. The rate of decline of brain concentrations of PNU-151774E paralleled the decline of anticonvulsant activity.

View larger version (16K): [in this window] [in a new window]   Fig. 2.   Time course of the anticonvulsant effect of PNU-151774E in the MES test (); plasma (, µmol/l) and brain (, µmol/kg) levels of the unchanged drug after oral administration of 10 mg/kg in rats. Groups of at least 20 rats were used to assess the anticonvulsant activity at various time points. Groups of three rats were used to calculate the plasma and brain concentrations of PNU-151774E.

Chemically induced seizures. PNU-151774E blocked tonic extension seizures induced by all chemical convulsants tested (table 2). This anticonvulsant activity was similar to or more potent than PHT, VPA and LTG, but less potent than CBZ and DZP. LTG showed relative inactivity on PIC-, 3-MPA- and STRYC-induced convulsions at doses up to 40 mg/kg p.o. VPA was also active in these chemically induced seizures, but at very high doses. PHT was not effective in the STRYC and PIC test. DZP was the most active anticonvulsant tested. PNU-151774E did not significantly affect the threshold for minimal seizures induced by the intravenous infusion of PTZ when administered i.p. to mice at 8.4 and 22 mg/kg i.p. (table 3).

Development of tolerance to the anticonvulsant effects in the MES test. Administration of the oral MES ED₉₀ dose of PNU-151774E (20.0 mg/kg) did not alter the activity of this compound in treated mice after 4 days of repeated dosing. The acute ED₅₀ of PNU-151774E determined in a separate group of mice at the start of repeated dosing was 11.2 (8.9-14.2) mg/kg, whereas that determined in mice after having had 20.0 mg/kg administered daily for 4 days was 13.2 (9.9-17.7) mg/kg. This difference was not significant. In contrast, the ED₅₀ for PB in the MES in mice after repeated i.p. administration of 25 mg/kg increased from 10.5 (8.6-12.8) mg/kg to 15.2 (11.4-20.2) mg/kg. No tolerance was observed in rats after 4 days of oral administration of 15.4 mg/kg of PNU-151774E. Repeated administration of this dose protected 37.5% of rats subjected to the MES test.

Behavioral ataxia and determination of therapeutic indices. The data shown in table 4 describe the oral TD₅₀ of PNU-151774E and reference anticonvulsants on behavioral ataxia as determined by the rotorod test and the calculated TIs of PNU-151774E relative to the various anticonvulsant ED₅₀ values. PNU-151774E produced motor impairment in the rotorod at relatively high doses (TD₅₀ = 626 mg/kg). This low toxicity combined with high potency in the anticonvulsant tests resulted in very high TIs of 78.2 (MES), 23.3 (BIC), 10.3 (PIC), 29.1 (3-MPA) and 6.0 (STRYC) for PNU-151774E. These TIs were generally higher than those of the reference anticonvulsants.

Spontaneous locomotor activity. No significant changes in either horizontal and vertical locomotor activity of rats was observed after dosing with up to 400 mg/kg p.o. of PNU-151774E. However, significant reductions of 48% (horizontal locomotor activity, fig. 3A) and 56% (vertical locomotor activity, fig. 3B) were seen after administration of the highest dose tested (700 mg/kg p.o). Although LTG tended to reduce both horizontal (29%, fig. 3C) and vertical (34%, fig. 3D) locomotor activity at the highest dose tested (200 mg/kg p.o.), these reductions were not statistically significant. The neuroleptic haloperidol (1 mg/kg), which served as a positive control in this experiment, significantly reduced both horizontal (91.67%, fig. 3E) and vertical (95.98%, fig. 3F) activity.

View larger version (51K): [in this window] [in a new window]   Fig. 3.   The effect of PNU-151774E, LTG and haloperidol on total horizontal (A, C, E) and vertical (B, D, F) locomotor activity in rats. Horizontal and vertical spontaneous locomotor activity (mean number of photobeam interruptions ± S.E.M.) was tested for 30 min 1 h after oral administration of the compound or vehicle in groups of 10 rats per compound. The significance of the differences between means were evaluated by Student-Newman-Keuls where * = P < .05 and ** = P < .01.

Passive avoidance. As shown in fig. 4, PNU-151774E did not impair passive avoidance responding at any of the doses tested including the highest dose of 400 mg/kg (40 times the MES ED₅₀). LTG, on the other hand, impaired passive avoidance responding. This impairment was significant at the highest dose tested (160 mg/kg, which is equivalent to 40 times the MES ED₅₀ for LTG). The noncompetitive N-methyl-D-aspartate antagonist (+) MK-801, used as a positive control, significantly impaired passive avoidance responding at the standard dose of 0.1 mg/kg.

View larger version (31K): [in this window] [in a new window]   Fig. 4.   The effect of PNU-151774E, LTG and (+) MK-801 on passive avoidance responding. Groups of 10 rats were trained in the passive avoidance procedure 60 min after receiving PNU-151774E and LTG orally or 30 min after receiving (+) MK-801 intraperitoneally. The data represent the mean (± S.E.M.) step-through latency to re-enter a dark compartment associated with aversive footshock 24 h after training. Significance of the difference between means were evaluated by Dunnett's t after one-way ANOVA where * = P < .05.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

PNU-151774E is a structurally novel anticonvulsant agent with inhibitory properties at the batrachotoxin-sensitive site 2 of the Na⁺ channel (Pevarello et al., 1996). PNU-151774E also has glutamate release-inhibiting properties (Vaghi et al., 1997) and is a MAO_B inhibitor (Strolin-Benedetti et al., 1994). This compound originally was identified on the basis of activity in the MES procedure and its relatively low behavioral toxicity in the rotorod test (Maj et al., 1993; Pevarello et al., 1998). Its spectrum of anticonvulsant activity was characterized further in various seizure models, and compared with that of standard reference anticonvulsants in this study. Four important results emerged: (1) PNU-151774E is an effective anticonvulsant in rats and mice, with a favorable TI; (2) PNU-151774E was not active in the PTZi.v. infusion model of absence seizures, but also lacked proconvulsant effects in this test; (3) PNU-151774E has pharmacokinetic properties favorable for a compound aimed at central action on the nervous system; (4) repeated administration with PNU-151774E does not suggest the development of tolerance to this compound; and (5) PNU-151774E does not affect locomotion or cognitive function at therapeutic doses.

PNU-151774E is a very effective anticonvulsant in well-established procedures for inducing seizures with either electrical or chemical supramaximal stimuli in rodents. Mechanistically, this suggests a reduction of seizure spread throughout the brain (Pirreda et al., 1985) and would indicate anti-generalized tonic-clonic efficacy in the clinic.

Suppression of seizures induced by threshold or liminal stimuli is taken as an index of antigeneralized absence or Petit Mal activity, which is related to raising the local epileptogenic threshold (White et al., 1995a, b). PNU-151774E did not alter seizure threshold; a lack of effect which suggests that PNU-151774E would show no efficacy in blocking absence seizures. However, lack of effect in preclinical models of absence seizures is controversial and need not necessarily translate into lack of clinical efficacy. LTG, for example, is not effective in antagonizing PTZ-induced facial twitches (Miller et al., 1986) nor electroencephalogram changes in WAG/Rij rats (van Rijn et al., 1994), which are both models of absence seizures. Nevertheless, it is effective in the clinic (Ferrie et al., 1995). Thus, the potential of PNU-151774E as an anti-absence drug must await clinical testing.

Seizures are the result of an imbalance of a wide variety of biochemical systems in the brain. The ability of PNU-151774E to block seizures induced by glycine and GABAergic mechanisms, as well as against seizures induced by enhancement of glutamatergic excitation (Maj et al., 1996), indicates that a broad spectrum of action also may be expected clinically. This is especially relevant considering that PNU-151774E also shows activity in models of complex partial seizures. Complex partial or temporal lobe epilepsy is the most frequent and common type of epilepsy in humans, accounting for approximately 40% to 60% of the affected population (Gastaut et al., 1975; Shorvon, 1990), and is not controlled adequately by currently available antiepileptic drugs (Mattson, 1992). Complex partial seizures have been modeled both by kainic acid administration (Ben-Ari, 1985) or the electrical kindling of the amygdala (McNamara et al., 1980). PNU-151774E not only inhibited kainic acid-induced seizures, but also reduced the hippocampal neuronal cell loss induced by kainic acid at doses (1-30 mg/kg i.p.) used in models of generalized tonic-clonic seizures (Maj et al., 1996.). Similar antiseizure effects of PNU-151774E also have been observed after amygdala kindling. PNU-151774E reduced electroencephalogram seizure activity as well as the intensity of behavioral seizures (Maj et al., 1995) in this model. These results indicate that PNU-151774E also would have potential efficacy in the treatment of complex partial seizures.

Despite the limited spectrum of behavioral seizure manifestations in rodents, convulsive phenomena originating from either the forebrain or the hindbrain can be distinguished (Browning, 1987; Fromm, 1987; Piredda and Gale, 1985). The former type is modeled by bicuculline-induced fits, whereas the latter may be modeled by STRYC-induced seizures. PNU-151774E is effective in antagonizing STRYC seizures in rats. Because several medically intractable epileptic myocloni and seizures may originate in the brainstem, it will be important to investigate the effectiveness of PNU-151774E in these clinical seizures.

PNU-151774E has pharmacokinetic properties favorable for a compound aimed at central action on the nervous system. These include a rapid and effective absorption after oral administration, and a preferable affinity for the brain where drug levels were approximately 10-fold higher than in plasma (fig. 2). These levels were related strictly to the course of anticonvulsant activity. This high selective affinity of PNU-151774E for the brain is not shared by classical AEDs. Brain levels in the rat of CBZ (cf., Morselli, 1995) or PHT (cf., Treiman and Woodbury, 1995), for example, are 1.1 to 1.6 and 1 to 3 times the plasma concentration. Among the newly developed AEDs, LTG reached rat brain levels of 1.4 to 1.9 times the plasma concentration after oral treatment (Parsons et al., 1995). The greater separation of drug levels between plasma and brain shown by PNU-151774E suggests anticonvulsant activity associated with lower peripheral side effects.

Impaired locomotor activity and co-ordination are known problems with barbiturates and benzodiazepines, and are among the factors influencing poor compliance (Mattson, 1992). PNU-151774E did not affect spontaneous locomotor activity in rats except at its rotorod toxic dose (700 mg/kg p.o.), i.e., 70 times its MES ED₅₀ (fig. 3). Similar, although nonsignificant effects were observed after treatment with LTG. These results obtained with PNU-151774E support the observation that this compound has a wide therapeutic window and indicate that PNU-151774E is not likely to affect co-ordination in patients.

Impairment of mental function, particularly in the development age, is a frequently reported side effect of anticonvulsants (Kälviäinen et al., 1996; Trimble and Cull, 1988). PNU-151774E did not affect passive avoidance responding, even at doses 40 times its oral ED₅₀ in the MES test (fig. 4). The lack of effect of PNU-151774E contrasted with the impairment observed with LTG given at a similarly corresponding dose of 40 times its MES ED₅₀ (160 mg/kg). Although LTG generally is considered to be an AED with low cognitive-impairing properties (cf., Binnie, 1994; Gillham et al., 1996), case reports of cognitive impairments associated with LTG have been published (Bouman et al., 1997). However, this impairment was observed in a patient also treated with VPA and showing abnormally high serum LTG levels (13.6 mg/l with respect to a normal range up to 4 mg/l). The results obtained in the passive avoidance test indicate that both LTG and PNU-151774E have a wide margin of safety and suggest that toxic doses would be required before cognitive impairments are evident in the clinic.

Finally, development of tolerance and behavioral side effects are frequent limitations of standard AEDs. No significant decrease in MES anticonvulsant activity was noted after repeated treatment with ED₅₀ (rats) or ED₉₀ (mice) doses of PNU-151774E in mice or rats. This is in contrast to the development of tolerance to the anti-MES effects of PB (Schmidt et al., 1980). These data suggest a lack of tolerance potential with the dosing protocols used, although further studies with longer periods of administration clearly are needed to confirm this indication. Furthermore, the inability of PNU-151774E to decrease seizure threshold in the PTZi.v. infusion test suggests a lack of proconvulsant potential. This is an important consideration because several of the newly introduced AEDs are being reported to exacerbate or induce de novo absence and/or myoclonic attacks in patients (e.g., Baulac et al., 1997; Besag et al, 1995).

Overall these data indicate that the structurally novel anticonvulsant PNU-151774E prevents seizure spread in electrical and chemical seizure models. PNU-151774E has a potency similar to or greater than of most classical anticonvulsant drugs. The anticonvulsant activity of this compound parallels its brain levels. Na⁺ channel blocking activity present in PNU-151774E (Pevarello et al., 1996) is common to other AEDs (Macdonald and Kelly, 1993), and is likely to account for its anticonvulsant activity. However, PNU-151774E's glutamate release inhibitory property (Vaghi et al., 1997) as well as its ability to inhibit MAO_B activity (Strolin-Benedetti et al., 1994) may also contribute to its anticonvulsant effect (cf., Pranzatelli and Nadi, 1995; Loscher and Lehmann, 1996). Whether PNU-151774E's MAO_B inhibitory properties may provide additional benefits as a neuroprotectant should be explored further in subsequent studies of its mechanism of action. Among the newly developed AEDs, LTG is most similar to PNU-151774E in that it shares Na⁺ channel and glutamate release inhibitory effects (cf., Goa et al., 1993). LTG also has fewer effects on psychomotor functions than other AEDs in clinical trials (Binnie, 1994; Cohen et al., 1985; Gillham et al., 1996). The results obtained with PNU-151774E in these preclinical assessments of psychomotor impairments further indicate that PNU-151774E also would have a low potential for inducing behavioral side effects such as impairment in locomotor activity, coordination and cognition when tested in the clinic.