Address correspondence and reprint requests to Dr. M.W. Decker at Neuroscience Research, Department R4N5, Bldg. AP-9A/3, Abbott Laboratories, 100 Abbott Park Road, Abbott Park, IL 60064-6125, U.S.A. E-mail: firstname.lastname@example.org
Current address of Dr. Bennani: Vertex Pharmaceuticals, Cambridge, MA 01239, U.S.A.
Current address of Dr. Jolly: Lilly Research Labs, Greenfield, IN 46104, U.S.A.
Summary: Purpose: The objective of this study was to characterize the antiseizure and safety profiles of ABT-769 [(R)-N-(2 amino-2-oxoethyl)spiro[2,5]octane-1-carboxamide].
Methods: ABT-769 was tested for protection against maximal electroshock and pentylenetetrazol-induced seizures in the mouse and for suppression of electrically kindled amygdala seizures and spontaneous absence-like seizures in the rat. The central nervous system safety profile was evaluated by using tests of motor coordination and inhibitory avoidance. The potential for liver toxicity was assessed in vitro by using a mitochondrial fatty acid β-oxidation assay. Teratogenic potential was assessed in the mouse.
Results: ABT-769 blocked maximal electroshock, subcutaneous pentylenetetrazol and intravenous pentylenetetrazol–induced seizures with median effective dose (ED50) values of 0.25, 0.38, and 0.11 mmol/kg, p.o., respectively. No tolerance was evident in the intravenous pentylenetetrazol test after twice-daily dosing of ABT-769 (0.3 mmol/kg, p.o.) for 4 days. ABT-769 blocked absence-like spike–wave discharge (ED50, 0.15 mmol/kg, p.o.) and shortened the cortical and amygdala afterdischarge duration of kindled seizures (1 and 3 mmol/kg, p.o.). The protective indices (ED50 rotorod impairment/ED50 seizure protection) were 4.8, 3.2, and 10.9 in the maximal electroshock, subcutaneous pentylenetetrazol and intravenous pentylenetetrazol seizure tests, respectively. ABT-769 did not affect inhibitory avoidance performance (0.1–1 mmol/kg, p.o.). ABT-769 did not affect mitochondrial fatty acid β-oxidation or induce neural tube defects.
Conclusions: ABT-769 is an efficacious antiseizure agent in animal models of convulsive and nonconvulsive epilepsy and has a favorable safety profile. ABT-769 has a broad-spectrum profile like that of valproic acid. Its profile is clearly different from those of carbamazepine, phenytoin, lamotrigine, topiramate, vigabatrin, and tiagabine.
The search for new antiepileptic drugs (AEDs) is driven by the need for drugs that are effective in pharmacoresistant epilepsies and by the need for drugs that have far better safety and tolerability profiles than those of the current generation of drugs (1). Approximately one third of epilepsy patients do not experience adequate seizure control with current medications, and significant safety and tolerability issues abound among the AEDs and often curtail their use. ABT-769 [(R)-N-(2 amino-2-oxoethyl)spiro[2.5]octane-1-carboxamide] (Fig. 1) was synthesized with the objective of identifying a broad-spectrum AED with a favorable safety and tolerability profile. ABT-769 is a valproic acid (VPA) analogue possessing a spirobicyclo [2.5]octane core with a pendant glycinamide substituent.
VPA has a unique and desirable efficacy profile among the AEDs. It is often described as a broad-spectrum drug because it has efficacy in the treatment of both partial and generalized epilepsy in adults and children (2). The broad-spectrum profile of VPA distinguishes it not only from other first-generation antiepileptic compounds, such as phenytoin (PHT) and carbamazepine (CBZ), but also from many of the newer, second-generation compounds. Not only is VPA an effective AED, but divalproex, a 1:1 molar compound of sodium valproate and VPA, is approved for the treatment of the manic episodes associated with bipolar disorder and the prophylaxis of migraine headache (3). VPA also has shown potential in the treatment of impulsivity, aggression, agitation, and personality disorders (4,5). It is well documented that VPA has many different mechanisms of action, and it has been suggested that multiple mechanisms of action account for its broad-spectrum antiseizure profile (6).
Hepatic failure and teratogenicity are the main toxic effects of VPA. Although rare in occurrence, they are serious adverse events. The structural features accounting for the teratogenicity and hepatotoxicity of VPA and related derivatives were considered in the design of ABT-769. Radatz and colleagues (7) have reported several in vivo structure–activity relation studies on the propensity of numerous VPA analogues to produce neural tube–related birth defects in the mouse (7). Structural attributes considered essential for teratogenicity include the presence of a free carboxylic acid group and a branched sp3α-carbon atom containing a proton. Compounds lacking any of these structural requirements typically do not cause neural tube malformations, although two recent studies documented that taurinamide and amantadine amide derivatives of VPA did induce significant neural tube defects (8,9). Meanwhile, the hepatotoxicity of VPA may be largely attributed to its 4-ene and 2,4-diene metabolites, which have terminal double bonds (10). These olefinic metabolites interfere with the process of fatty acid β-oxidation and can ultimately cause microvesicular steatosis (10,11).
The glycinamide moiety of ABT-769, in analogy with the nonteratogenic VPA derivative TV-1901 (12), was incorporated into ABT-769 to increase antiseizure potency and reduce the likelihood for induction of neural tube defects. Cyclic VPA analogues, including N-methyl–tetramethylcyclopropyl carboxamide, tend not to be teratogenic (13). It is possible that these structural elements present in ABT-769 may also preclude the formation of terminal alkene-containing metabolites, thereby circumventing the hepatotoxic liabilities associated with VPA.
ABT-769 was evaluated in the maximal electroshock (MES) and pentylenetetrazol (PTZ) seizure tests in the mouse to provide a general characterization of its antiseizure pharmacology. ABT-769 was then evaluated for its effects on electrically kindled amygdala seizures and nonconvulsive absence-like seizures in the rat. Duration of action and tolerance also were studied. The potential for adverse CNS effects of ABT-769 was examined in tests of motor coordination and inhibitory avoidance learning. Toxicity profiling focused on liver toxicity and teratogenicity.
ABT-769 [(R)-N-(2–amino-2-oxoethyl)spiro[2.5]octane-1-carboxamide], its enantiomer, A-425760, and the glycine acid metabolite (Fig. 1) were synthesized at Abbott Laboratories (North Chicago, IL, U.S.A.) (14). Sodium VPA, PTZ, and diazepam (DZP) were obtained from Sigma Chemical Co. (St. Louis, MO, U.S.A.).
Antiseizure profile and side-effect liability in the mouse
Adult, male, CD-1 mice (22–25 g) were obtained from Charles River Laboratories (Portage, MI, U.S.A.) for use in tests of electrically and chemically induced seizures, inhibitory avoidance, and motor impairment. Mice were housed under standard lighting conditions of 12 h on/12 h off, with lights on at 6 a.m. Food and water were provided ad libitum. Studies were performed under protocols approved by Abbott Laboratories Institutional Animal Care and Use Committee.
Compound administration for antiseizure studies
ABT-769, A-425760, VPA (sodium valproate), and DZP were suspended in 0.5% methylcellulose for oral administration. PTZ was dissolved in saline for subcutaneous and intravenous administration. Compounds were administered to the mouse orally in a volume of 10 ml/kg of vehicle.
Maximal electroshock–induced seizures
ABT-769 was tested for its ability to block MES-induced tonic hindlimb extensor seizures in the mouse. A supramaximal current (50 mA; 0.4-s duration; pulse width, 0.5 ms; 60 pulses/s, ECT Unit (Ugo Basile #7801) was applied via corneal electrodes that were coated with electrocardiogram electrolyte (Signa Crème; Parker Labs 1708, Columbus, OH, U.S.A.) to ensure good contact. Mice were observed after stimulation for the onset of tonic hindlimb extension and death. A mouse was considered to have had a tonic extensor seizure only if it experienced a prolonged extension (>90 degrees from plane of body) of the hind legs. Animals not displaying tonic hindlimb extension were considered protected. Twenty mice were used per dose group. ABT-769 (0.1, 0.2, 0.3, and 0.4 mmol/kg, p.o.), A-425760 (0.2, 0.3, 0.4, and 0.5 mmol/kg, p.o.), and VPA (0.5, 0.75, 1, 1.5, 1.8, 2, and 4 mmol/kg, p.o.) were administered orally 30 min before electroshock. The major metabolite of ABT-769, a glycine acid derivative, was tested at a dose of 1.6 mmol/kg, p.o. and i.p.
ABT-769 was tested for its ability to block clonic and tonic seizures induced by the subcutaneous injection of PTZ (85 mg/kg, s.c.). During the experiment, the mice were housed individually in clear polycarbonate cages for observation. PTZ was injected just below the nape of the neck. The animals were then observed for 15 min. The number of mice that exhibited seizures and the time to the first clonic convulsion and subsequent sustained tonic convulsion were recorded. Twenty mice were used per dose group. ABT-769 (0.1, 0.3, 0.5, and 0.9 mmol/kg, p.o.), A-425760 (0.2, 0.5, 1, and 1.5 mmol/kg, p.o.), and VPA (1, 1.3, 1.5, 2, 2.5, and 3 mmol/kg, p.o.) were administered orally 30 min before PTZ injection.
ABT-769 was tested for its ability to block clonic seizures induced by the intravenous administration of PTZ (32 mg/kg, i.v., bolus of 10 ml/kg through the tail vein). The number of mice having clonic seizures within 30 s after PTZ injection was recorded. Ten mice were used per dose group. ABT-769 (0.05, 0.1, 0.2, and 0.4 mmol/kg, p.o.), A-425760 (0.4, 0.8, and 1.2 mmol/kg, p.o.), and VPA (0.5, 1, 2, and 3 mmol/kg, p.o.) were administered orally 30 min before the intravenous PTZ challenge. The glycine acid metabolite of ABT-769 was tested at a dose of 1.6 mmol/kg, p.o. and i.p., administered 30 min before the PTZ challenge. To determine the duration of action in the mouse, protection against PTZ-induced seizures was assessed at 0.5, 1, 2, 3, 4, and 5 h after dosing with ABT-769 (0.6 mmol/kg, p.o.) and at 0.5, 1, 2, 3, and 4 h after dosing with VPA (4 mmol/kg, p.o.).
Antiseizure effects after repeated dosing also were evaluated in the mouse by determining intravenous PTZ seizure threshold. Mice were treated twice daily with ABT-769 (0.3 mmol/kg, p.o.), VPA (4 mmol/kg, p.o.), DZP (0.002 mmol/kg, p.o.), or vehicle for 4 days. On day 5, vehicle or a single dose of ABT-769 (0.3 mmol/kg, p.o.), VPA (2 mmol/kg, p.o.), or DZP (0.002 mmol/kg, p.o.) was administered, and mice were challenged with an intravenous PTZ injection. The dose of PTZ was titrated by using a modified up–down procedure within a group of animals to determine the dose that evoked clonic seizures in 50% of the animals (15). Fourteen mice were used per dose group. The first animal in each dose group was injected with one of five doses of PTZ (21, 24, 27, 32, or 37 mg/kg, i.v.) and observed for the induction of clonic seizures. If the first animal did not convulse, the second animal received the next larger dose of PTZ. Conversely, if the first animal convulsed, the second animal received the next smaller dose of PTZ. The dose of PTZ was thus increased or decreased until all 14 mice were tested. For each dose of PTZ, the number of mice with convulsions and the total number tested were recorded. The median effective dose (ED50) values were derived by using the Litchfield–Wilcoxon method and refer to the sensitivity to PTZ. Thus, higher ED50 values indicate increased protection from seizures.
Data analysis for MES and PTZ studies
The percentage of mice protected from seizures was determined, and ED50 values and 95% confidence intervals were calculated. ED50 values in the subcutaneous PTZ test were determined by using linear regression, and ED50 values in the MES and intravenous PTZ tests were determined by using PROBIT analysis.
Side-effect liability: acute motor impairment
Acute motor impairment was assessed in the mouse by using an Omnitech Omni Rotor (Accuscan Instruments, Columbus, OH, U.S.A.). The apparatus consists of four rubber-coated rods, 33 mm in diameter and 111 mm wide. The rods were isolated from each other in separate enclosures and mounted 365 mm above a wire-grid floor. Thirty minutes after dosing, four mice, one per enclosure, were placed onto the stationary rod, facing away from the experimenter. The rotation of the four rotors increased simultaneously from 0 to 40 rotations per minute over a 120-s test period. Mice fell from the rotating rod onto the wire grid and halted the corresponding timer. Each mouse was tested twice in succession, and a mean latency in seconds to fall from the rotating rod was noted. ABT-769 was administered at doses of 0.3, 1, and 3 mmol/kg, p.o., and VPA was administered at doses of 1, 2, 3, 4, 5, and 6 mmol/kg, p.o. Control mice received the vehicle only but were otherwise treated identically. In total, 12 mice were used in each dose group. Compounds were dosed orally in 0.2% hydroxypropylmethylcellulose in distilled water at a volume of 10 ml/kg. The percentage impairment was calculated according the formula:
Side-effect liability: determination of learning and memory impairment
Four automated, independently controlled, shuttlebox systems (ShuttleScan II; Accuscan Instruments with ShuttleFlex 2.10 software) were used to assess recall in an inhibitory avoidance task. A shuttlebox consists of two compartments separated by a guillotine door, independently controlled overhead lightbulbs in each compartment to create a lighted side and a darkened side, and a shock grid floor.
A mouse was placed into the right-hand compartment of the shuttlebox 30 min after dosing with ABT-769 (0.1, 0.3, and 1 mmol/kg, p.o.), VPA (0.3, 1, and 3 mmol/kg, p.o.), or DZP (0.001, 0.004, and 0.01 mmol/kg, p.o.). Control mice received the vehicle (10 ml/kg) only but were otherwise treated identically. In total, 10 mice were used in each dose group. After a 30-s habituation period, the overhead light in the occupied compartment was turned on, and the guillotine door separating the two compartments was opened. When the mouse traversed to the dark side, the door closed automatically, and a 0.5-mA current was applied to the stainless steel grid floor for 5 s. The latency to transfer from the lighted compartment to the dark compartment was recorded. The mouse was removed when shock delivery was complete and returned to the home cage. Twenty-four hours later, the mouse was once again placed into the right hand compartment of the same shuttlebox, and the procedure was repeated without shock. The latency to transfer from the lighted compartment to the darkened compartment was recorded once again. The grid floor was wiped with a fresh, dry towel between each trial, and a criterion time of 180 s applied. Latency data for the training period were analyzed by analysis of variance (ANOVA). Latency data for the recall period were analyzed by using Kruskal–Wallis and individual Mann–Whitney tests because of the nonparametric nature of the data. Compounds were dosed orally in 0.2% hydroxypropylmethylcellulose and 0.1% Tween 80 in distilled water at a volume of 10 ml/kg.
Spike–wave discharge in the EEG: absence-like seizures in the rat
Male WAG/Rij (Wistar derived) rats age 4–6 months (400–500 g) were obtained from Charles River (Germany). They were delivered to the laboratory ≥2 to 3 weeks before implantation. The animals were housed under standard lighting conditions of 12 h on/12 h off, with lights on at 6 a.m. Food and water were provided ad libitum. After surgery, the animals were individually housed in a postsurgical room under the same environmental conditions as before surgery. Food and water were provided ad libitum. This study was performed under a protocol approved by Abbott Laboratories Institutional Animal Care and Use Committee.
Rats were surgically implanted under pentobarbital (PTB) anesthesia (50 mg/kg, i.p.) with permanently indwelling stainless steel screw-type cortical electrodes for EEG recording. The method used was similar to that described by Peeters (16). The active recording electrode was placed in the skull over the dura at −2 mm AP, and 4 mm L from bregma. A ground electrode was placed in the skull over the cerebellum. Active and ground electrodes were connected to a miniature electrical plug that was cemented to the skull with dental acrylic. Postoperative care was provided. Rats were allowed to recover from surgery for ≥14 days before experiments were initiated.
Each rat was administered either three doses of ABT-769 (0.1, 0.3, and 1 mmol/kg, p.o.) or VPA (1, 3, and 6 mmol/kg, p.o.). Each rat also was administered the vehicle to establish baseline seizure activity. Each dose of ABT-769, VPA, or vehicle was administered on different days with ≥3 days separating each dosing and recording session. With this design, each rat served as its own control (vehicle treatment). Twenty minutes after dosing, rats were placed in cages in which the EEG was recorded for 2 h. The vehicle for ABT-769 and VPA was 0.2% hydroxypropylmethylcellulose and Tween-80 (2 drops/ml HPMC). Compounds were administered at an oral dosing volume of 5 ml/kg. The EEG signals were acquired with a cabling system that allowed unrestricted movement within the cage. EEG amplifiers (Astro-Med Grass-Telefactor, West Warwick, RI, U.S.A.) and a computer were used to condition and archive the microvolt EEG signals.
In repeated dosing experiments, ABT-769 (0.3 mmol/kg, p.o.) was administered once daily for 7 days. EEG was recorded daily as described earlier after each dose and then daily for 5 days after the last dose on day 7.
After the 2-h recording session, the EEG was analyzed off-line by using computer-based analysis software (Stellate Systems, Montreal, Canada). Absence-like 6-to 9-Hz spike–wave activity was detected by an algorithm that uses frequency, amplitude, and waveform to identify the seizure burst. The accuracy of this algorithm was previously verified by manual inspection of EEG records. The total amount of seizure time for each rat during the 2-h recording period was derived from this off-line analysis and was used as a measure to characterize the antiseizure effects of the drugs. Statistical significance was evaluated by using a repeated-measures ANOVA with Fisher's protected least significant difference post hoc test for comparisons between treatments.
Electrically kindled amygdala seizures in the rat
Kindled seizure experiments were performed by Porsolt and Partners Pharmacology (Paris, France). Adult, male Wistar rats (376–621 g) were obtained from Elevage Janvier (Le Genest-Saint-Isle, France). They were delivered to the laboratory ≥4 days before implantation and group housed (up to six per cage). The animals were housed under standard lighting conditions of 12 h on/12 h off, with lights on at 7 a.m. Food and water were provided ad libitum. After electrodes were implanted, the animals were individually housed in a postsurgical room under the same environmental conditions as before surgery. Studies were performed in accordance with a currently valid license for experiments on vertebrate animals issued by the French Ministry for Agriculture and Forestry to Mr. Sylvain Roux (Study Director, Porsolt and Partners Pharmacology, France).
Rats were anesthetized with sodium PTB (55 mg/kg, i.p., plus supplemental doses of 5–10 mg/kg as needed), and surgery was performed under aseptic conditions. Rats were implanted with two surface titanium miniature screw-type EEG electrodes, bilaterally, over the frontoparietal cortex and with two twisted platinum-iridium wire electrodes placed stereotaxically into the left amygdala nucleus (Paxinos and Watson interaural target coordinates: AP +6.5 mm, L +4.8 mm, V +1.5 mm). A third electrode was placed over the right parietooccipital cortex to serve as ground. Postoperative care was provided. Rats were allowed to recover from surgery for ≥10 days before the kindling procedure was initiated. At the completion of the experiments, animals were deeply anesthetized with PTB (>100 mg/kg, i.p.) and given constant-current pulse (1 mA, 10 s) to mark the location of deep electrodes. Electrode placement was verified by microscopic examination.
The kindling procedure was similar to that described by Meldrum and Durmuller (17). In brief, the rats were stimulated twice daily (2-s rectangular pulses; pulse width, 1 ms; amplitude, 500 μA; frequency, 50 Hz) until fully kindled. Rats showing at least three consecutive stage five seizures were regarded as fully kindled.
Fully kindled animals were administered vehicle, ABT-769, or VPA in their home cages. The rats were then individually transferred to the testing apparatus. Behavioral and EEG signals were recorded for 10 min, overlapped by a 2-s stimulation episode at 3 min from the start of recording. On day 1, a vehicle control recording/stimulation test was performed to establish baseline. On day 2, 24 h after the control test, compounds were administered, and the rats were recorded/stimulated 60 and 120 min after dosing. Electrical stimulations were separated by 1 h during testing in kindled rats to avoid potential for refractoriness. On day 3, the rats were recorded/stimulated once at 24 h after dosing on day 2. The animals received the three doses of ABT-769 or VPA in a random order. ABT-769 and VPA were dosed orally in 0.2% hyroxypropylmethylcellulose in distilled water and in distilled water, respectively. Compounds were administered at an oral dosing volume of 5 ml/kg.
Cortical and amygdala afterdischarge durations were measured in seconds from the start to the end of the electrical seizure. Behavioral seizure scores, ranging from 0 to 5, were defined as follows:
0, no visible response
1, facial automatism (licking, chewing, etc.)
2, head nodding, head jerks, eye blinking, facial clonus
3, 2 + unilateral forelimb clonus
4, 2 + rearing, bilateral forelimb clonus
5, 4 + animal falling over
Afterdischarge scores were statistically analyzed by the Wilcoxon signed rank test. Scores at each time point measured were compared individually with control values. Behavioral seizure scores were analyzed by paired Student's t test. As for the seizure scores, each time point measured was compared individually with the corresponding control values. No statistical comparisons were made between postadministration time points, between different doses, or between different parameters.
Reproductive toxicity in the mouse
The embryonic and fetal developmental effects of ABT-769 were evaluated in the female Crl:CD-1(ICR)BR mouse. The study was performed by the Argus Division of Charles River Laboratories (Horsham, PA, U.S.A.).
Virgin females, ≥60 days old and weighing ∼25–30 g, were bred overnight and examined the next morning for the presence of vaginal plugs. Females observed to have a copulatory plug were considered to be at day 0 of gestation. ABT-769 was administered orally at dosages of 2, 4, and 8 mmol/kg/day. Mice were dosed (15 dams per dose group) from gestation days 6 to 15. Satellite animals were dosed (12 mice per dose group) for plasma ABT-769 determinations. Blood samples were collected at gestation days 6 and 15 at 1.5, 4, 8, and 24 h after the first daily-dose administration. The main study animals were caesarean sectioned on gestation day 18 and examined for the number and distribution of corpora lutea, implantation sites, and uterine contents. A gross necropsy of the thoracic, abdominal, and pelvic viscera was performed. Fetuses were weighed and examined for gross external, visceral, and skeletal alterations and gender.
Because liver toxicity associated with VPA appears to be related to the inhibition of the β-oxidation of fatty acids and disruption of mitochondrial respiration (10,11), the effects of ABT-769 were compared with those of VPA on mitochondrial fatty acid β-oxidation by using the method of Otto et al. (18). Mitochondria were prepared from rat liver after killing with pentobarbital (Nembutal). Livers were removed and placed in 20 ml of buffer containing 220 mM mannitol, 70 mM sucrose, 5 mM MOPS, and 2 mM EGTA 0.01% delipidated BSA, pH 7 MSE/BSA). The liver was then minced and washed with the MSE/BSE buffer twice. The minced liver was homogenized with a Polytron homogenizer with the 12-mm generator for 10 s at a power setting of 19. The homogenate was centrifuged at 400 g for 10 min, and the resulting supernatant was centrifuged at 5,700 g for 10 min at 4°C. The resulting supernatant was discarded and the pellet resuspended in MSE/BSE. The suspension was rehomogenized with a Potter-Elvehjem homogenizer and centrifuged at 400 g, and the supernatant was decanted into a clean centrifuge tube. This supernatant was centrifuged at 5,700 g for 10 min, and the resultant pellet resuspended in 0.5 ml of MSE/BSE; protein concentration was determined by the method of Bradford (19).
Mitochondria were suspended in a media containing 94 mM sucrose, 60 mM KCl, 10 mM potassium phosphate, 5 mM MgCL2, 3 mM ATP, 0.10 mM coenzyme A, 1 mM DTT, 0.4 mM L-carnitine, and 1.5% BSA, pH 7.2. ABT-769 or VPA at concentrations of 100, 300, and 1,000 μM was dissolved in media and added to the suspension. The reaction was initiated by the addition of universally labeled [14C]palmitic acid, 7.5% BSA, supplemented with 5 mM cold palmitic acid to a final activity of 0.5 μCi/ml and 0.5 mM cold palmitate. The reaction was stopped by adding 1.2 ml 70% PCA, and mixing vigorously with a vortex mixer and placing the sample on ice. After two hexane extractions, 250 μl of acid precipitable material was counted on a Beckman scintillation counter with 10 ml scintillant. The amount of palmitate oxidized (nmoles per minute per milligram protein) was calculated and expressed as a percentage of the vehicle control.
Analysis of plasma and brain concentrations by HPLC-MS/MS
Blood and brain samples were collected from groups of mice and rats killed after testing. Plasma was separated by centrifugation. Plasma and brain samples were stored frozen (−20°C) until analysis. Brain tissue was homogenized with two parts water before analysis. A measured aliquot of plasma or homogenate (100–250 μl) was combined with an aliquot of internal standard and 2–5 volumes of acetonitrile. The samples were vortexed for ∼1 min, followed by centrifugation (15 min ×∼2,500 rpm; 4°C). The supernatant was transferred to a clean polypropylene tube and evaporated to dryness with a gentle stream of dry nitrogen over low heat (∼35°C). The samples were reconstituted by vortexing with 150 μl mobile phase. Spiked plasma and brain homogenate standards were analyzed simultaneously with the samples.
ABT-769 or A-425760, the respective glycine metabolites, and the internal standard were separated from coextracted contaminants on a 50 × 3-mm 5-μm Aquasil C18 column (Keystone Scientific, Inc.) eluted with an acetonitrile:1% acetic acid (∼56:44 by volume) at a flow rate of 0.38 ml/min. Analysis was performed on a Sciex API2000 Biomolecular Mass Analyzer with a turbo-ionspray interface. Analytes were ionized in the positive ion mode with a source temperature of ∼450°C. Detection was in the multiple reaction monitoring (MRM) mode at m/z 211.0 → 194.3 for ABT-769/A-425760, m/z 212.1 → 137.2 for the glycine acid metabolite, and m/z 239.4 → 222.5 for the internal standard.
Similar sample-extraction methods were used for the quantitation of VPA in both plasma and brain samples. VPA and the internal standard were separated from coextracted contaminants on a 50 × 3-mm 5-μm Clipeus C18 column (Higgins Analytical) eluted with an acetonitrile:1% acetic acid (∼60:40 by volume) at a flow rate of 0.4 ml/min. Analysis was performed on a Sciex API2000 Biomolecular Mass Analyzer with a turbo-ionspray interface. Analytes were ionized in the positive ion mode with a source temperature of ∼450°C. Detection was in the single-ion monitoring mode (SIM) at m/z 145.0 for VPA and m/z 155.1 for the internal standard.
The peak areas of the selected analytes and internal standard were determined using the Sciex Analyst or TurboQuan software. The plasma drug concentration of each sample was calculated by least squares linear regression analysis (weighted as 1/concentration) of the peak area ratio (parent/internal standard) of the spiked plasma or brain homogenate standards versus concentration. The methods, generally evaluated over the concentration range from 0 to 80 μg/ml(g), were characterized by mean accuracy values from 95–107% of theory for the analysis of triplicate standards at six to eight separate concentrations. The limit of quantitation was estimated to be ∼0.2 μg/ml(g) from a 0.25-ml sample.
Maximal electroshock and pentylenetetrazol seizures
ABT-769 was fully efficacious against MES-induced seizures and against subcutaneous PTZ and intravenous PTZ-induced seizures in the mouse at 30 min after oral administration (Table 1, Fig. 2). ABT-769 was more potent than VPA in these tests. The trend for ABT-769 to be more potent than its enantiomer, A-425760, most likely reflects a difference in pharmacokinetics because comparable plasma concentrations of 28 and 30 μg/ml of ABT-769 and A-425760, respectively, were measured at the ED84 values (0.3 and 1.4 mmol/kg, p.o., respectively) in the intravenous PTZ test.
Table 1. Effects of ABT-769, A-425760, and valproic acid in the maximal electroshock and pentylenetetrazol seizures tests in the mouse
ED50 (95% confidence interval) mmol/kg, p.o.
The effects of repeated administration of ABT-769, VPA, and DZP on antiseizure efficacy were evaluated in separate studies (Table 2). ABT-769 and VPA were equally effective in raising the seizure threshold in the intravenous PTZ test after either short-term or repeated treatment. In contrast, DZP was less effective after repeated treatment in both studies.
Table 2. Effects of ABT-769, valproic acid, and diazepam on intravenous pentylenetetrazol seizure threshold in the mouse after twice-daily administration for 4 days
Pretreatment: Twice-daily dosing for 4 days (mmol/kg, p.o.)
The efficacy of supramaximal doses of ABT-769 (0.6 mmol/kg, p.o.) and VPA (4 mmol/kg, p.o.) was compared as a function of time in the intravenous PTZ test. As shown in Fig. 3, comparable durations of action were achieved with the two compounds. It should be noted that the dose of VPA was more than sixfold greater than the dose of ABT-769.
The terminal amide of ABT-769 is hydrolyzed in vivo to form the glycine acid derivative of ABT-769. To determine whether the glycine acid derivative is an active metabolite of the parent, it was evaluated in the MES and intravenous PTZ tests. A maximum of 20–30% protection was observed after either oral or intraperitoneal administration of 1.6 mmol/kg of the glycine acid derivative in the intravenous PTZ seizure test. This dose produced concentrations of 31.1 μg/ml and 1.1 μg/g of the glycine derivative in plasma and brain, respectively. Even less efficacy (0–10%) was observed at this dose in the MES seizure test. After a dose of 0.3 mmol/kg, p.o., of ABT-769 that produced 60–70% efficacy in the mouse intravenous PTZ and MES seizure tests, concentrations of 5.3 μg/ml and 0.75 μg/g of the glycine derivative were measured in plasma and brain, respectively. Thus although it is possible that even higher brain concentrations of the glycine acid derivative of ABT-769 would be efficacious, the concentrations generated by metabolism of ABT-769 were not active.
Pharmacodynamic/Pharmacokinetic relation in the mouse
A dose of 0.3 mmol/kg, p.o., of ABT-769 that did not produce full efficacy at any time was used to study the relation between antiseizure efficacy and the plasma and brain concentrations of ABT-769. Antiseizure efficacy in the MES and intravenous PTZ seizure tests was determined as a function of time. Plasma and brain concentrations were then determined at the same time points in a separate set of mice. The data reveal that the peak antiseizure activity in both seizure tests occurred at 30 min after dosing, which was the time point used to generate the dose–response data already described, and that the duration of action was closely related to both plasma and brain concentrations (Fig. 4A). Another way of examining these data is to determine the efficacy and concentration at each time point and to plot efficacy as a function of concentration. These plots of efficacy as a function of concentration (a concentration–response graph) show that the concentration in both plasma (Fig. 4B) and brain (Fig. 4C) was highly correlated with efficacy in both seizure models (R2 values, all >0.95).
Spike–wave discharge: absence-like seizures
ABT-769 produced efficacy comparable to that produced by VPA against absence-like spike–wave discharge in the WAG/Rij rat, but it was more potent than VPA after oral administration (Fig. 5). The ED50 values estimated from regression analysis were 0.15 and 2.6 mmol/kg, p.o., for ABT-769 and VPA, respectively. Neither ABT-769 nor VPA affected slow-wave activity (1–4 Hz) associated with somnolence (data not shown).
ABT-769 (0.3 mmol/kg, p.o., once daily) blocked spike–wave discharges on each day of the 7-day repeated-dosing study. During the first session conducted after discontinuation of treatment (i.e., 24 h after the last treatment on day 7), slow-wave discharge activity had still not returned to baseline. Spike–wave discharge activity returned to baseline levels 3–5 days after discontinuation of ABT-769 (data not shown).
Electrically kindled amygdala seizures
The results obtained in the electrically kindled seizure test are shown in Table 3. A dose of 3 mmol/kg, p.o., of ABT-769 significantly reduced the behavioral seizure score and significantly shortened cortical and amygdala afterdischarge duration at 60 and 120 min after dosing. Cortical and amygdala afterdischarge durations also were significantly shortened by a dose of 1 mmol/kg of ABT-769 at 120 min after dosing. Doses of 3 and 6 mmol/kg, p.o., of VPA significantly reduced the behavioral seizure score at 60 and 120 min after dosing. Cortical and amygdala afterdischarge durations were significantly shortened at 60 and 120 min after a dose of 6 mmol/kg, p.o., of VPA and at 120 min after a dose of 3 mmol/kg, p.o., of VPA.
Table 3. Effects of ABT-769 and valproic acid on behavioral seizure score and on cortical and amygdala afterdischarge duration in the amygdala kindled rat
ABT-769 60 min before first stimulation
ap < 0.05; bp < 0.01; cp < 0.001.
Average behavioral seizure score (mean ±(SEM)
0.3 mmol/kg, p.o.
5.0 ± 0.0
4.3 ± 0.6
3.7 ± 0.8
5.0 ± 0.0
1.0 mmol/kg, p.o.
5.0 ± 0.0
4.5 ± 0.3
3.0 ± 0.6
5.0 ± 0.0
3.0 mmol/kg, p.o.
5.0 ± 0.0
2.3 ± 0.6a
1.4 ± 0.4a
5.0 ± 0.0
Average cortical afterdischarge duration in s (mean ± SEM)
0.3 mmol/kg, p.o.
102.3 ± 23.3
74.8 ± 19.5
52.0 ± 16.2
102.5 ± 8.9
1.0 mmol/kg, p.o.
110.9 ± 10.6
96.1 ± 15.5
19.3 ± 9.3b
109.5 ± 3.6
3.0 mmol/kg, p.o.
92.0 ± 9.9
32.8 ± 12.6b
23.3 ± 9.7c
101.8 ± 8.9
Average amygdala afterdischarge duration in s (mean ± .SEM)
0.3 mmol/kg, p.o.
86.8 ±. 18.4
1.0 mmol/kg, p.o.
3.0 mmol/kg, p.o.
87.5 ± 9.6
26.0 ± 8.7b
15.8 ± 6.2c
99.9 ± 8.5
Valproic acid 60 min before first stimulation
Average behavioral seizure score (mean ± SEM)
1.0 mmol/kg, p.o.
5.0 ± 0.0
4.6 ± 0.4
4.1 ± 0.6
5.0 ± 0.0
3.0 mmol/kg, p.o.
5.0 ± 0.0
1.7 ± 0.9a
2.1 ± 0.7a
5.0 ± 0.0
6.0 mmol/kg, po
5.0 ± 0.0
0.6 ± 0.3a
1.9 ± 0.8a
4.9 ± 0.1
Average cortical afterdischarge duration in s (mean ± SEM)
1.0 mmol/kg, p.o.
88.5 ± 8.3
86.7 ± 15.4
64.4 ± 12.0
97.7 ± 4.6
3.0 mmol/kg, p.o.
93.2 ± 4.5
51.3 ± 29.1
15.6 ± 4.5c
88.5 ± 12.4
6.0 mmol/kg, p.o.
78.4 ± 9.7
18.6 ± 5.5c
23.1 ± 7.4b
69.4 ± 12.7
Average amygdala afterdischarge duration in s (mean ± SEM)
1.0 mmol/kg, p.o.
85.8 ±. 6.6
76.1 ±. 13.5
3.0 mmol/kg, p.o.
6.0 mmol/kg, p.o.
76.6 ± 9.2
15.6 ± 5.4b
23.4 ± 8.3b
69.4 ± 12.0
Evaluation of side-effect liability
As shown in Fig. 6, ABT-769 and VPA produced a dose-related impairment of rotorod performance in the mouse at doses above those effective in the MES and intravenous PTZ seizure tests. The ED50 values for rotarod impairment are 1.2 and 4.4 mmol/kg, p.o., for ABT-769 and VPA, respectively. The protective indexes of ABT-769, which were calculated by dividing the ED50 for rotorod impairment by the ED50 for seizure protection, are 4.8, 3.2, and 10.9 in the MES, subcutaneous PTZ, and intravenous PTZ seizure tests, respectively. The protective indexes of ABT-769 were greater than those for VPA, which were 3.7, 2.6, and 5.7 in the MES, subcutaneous PTZ, and intravenous PTZ seizure tests, respectively.
The effects of ABT-769 and VPA on inhibitory avoidance are shown in Fig. 7. Doses of 0.1, 0.3, and 1 mmol/kg, p.o., of ABT-769 (Fig. 7A) administered before training did not significantly affect recall 24 h after footshock. The dose of 1 mmol/kg is 4, 2.6, and 9 times the ED50 for ABT-769 in the MES, subcutaneous PTZ, and intravenous PTZ seizure tests, respectively. Doses of 0.3 or 1.0 mmol/kg, p.o., of VPA (Fig. 7B) did not influence performance, but impaired performance was observed with a higher dose of 3 mmol/kg, p.o., that produces plasma concentrations substantially higher than those achieved clinically. The dose of 3 mmol/kg, p.o., is 2.5, 1.7, and 3.9 times the ED50 for VPA in the MES, subcutaneous PTZ, and intravenous PTZ seizure tests, respectively. A dose of 0.01 mmol/kg, p.o., of DZP given before training significantly attenuated recall 24 h after footshock; lower doses of 0.001 and 0.004 mmol/kg, p.o., did not impair performance (data not shown).
Reproductive toxicity in the mouse
Dosages of 4 and 8 mmol/kg/day of ABT-769 were maternally toxic, producing dose-dependent incidences of adverse clinical observations (ataxia, decreased motor activity, and ptosis at 4 and 8 mmol/kg/day; bradypnea, increased motor activity, impaired proprioceptive positioning, lost or impaired righting reflex, and coma at 8 mmol/kg/day.) Most or all of these signs appeared to be an exaggeration of the pharmacodynamic effects of ABT-769.
The observed maternal toxicity at the doses of 4 and 8 mmol/kg/day of ABT-769 demonstrated that the reproductive toxicity study was performed in a dose range that provided an adequate exposure of the fetus to ABT-769. Cmax and AUC values in the 2-, 4-, and 8-mmol/kg/day dosage groups were determined on gestation days 6 and 15. Cmax values ranged from about 5 to 10 times the effective plasma concentration in the mouse MES seizure test, and the AUC values ranged from 0.5 to 3.5 times the value projected to maintain efficacy in the mouse.
An increased incidence of treatment-related fetal alterations (reduction in fetal body weight, malformations and minor delays in ossification) were observed at the 8-mmol/kg/day dose. These alterations are very common in this mouse strain at maternally toxic doses of any test substance (20–22). Neural tube defects (exencephaly) were not observed at any dose.
The effects of ABT-769 and its enantiomer, A-425760, and VPA on the oxidation of palmitate in mitochondria isolated from rat liver are shown in Table 4. ABT-769 and its enantiomer did not inhibit β-oxidation at concentrations of 100, 300, and 1,000 μM. In contrast, VPA inhibited β-oxidation in the same concentration range. In this assay, percentages of control values <80% are interpreted as indicating meaningful interference with normal β-oxidation. In addition, the glycine acid and carboxylic acid metabolites of ABT-769 did not inhibit mitochondrial β-oxidation up to concentration of 1,000 μM (data not shown).
Table 4. Effects of ABT-769, its enantiomer (A-425760), and valproic acid on β-oxidation in mitochondria isolated from rat liver
Values shown are mean and standard deviations of the percentage control values.
VPA, valproic acid.
ABT-769 emerged from a program whose goal was to identify a broad-spectrum AED with a favorable safety and tolerability profile. VPA was chosen as a target because it is recognized as a major AED that is effective against many types of seizures and also has proven efficacy in migraine prophylaxis and bipolar illness. Thus in the studies reported here, VPA was used as the primary comparator, and toxicity profiling focused on the liabilities of VPA that could best be addressed preclinically. Taken together, the results show that ABT-769 has the profile in animals of a broad-spectrum AED and suggests that it lacks liabilities, such as liver toxicity and neural tube defects, associated with VPA. Moreover, ABT-769 does not have the narrow CNS therapeutic index in animals that is characteristic of many AEDs.
ABT-769 displayed a spectrum of antiseizure activity in rodent seizure models similar to that produced by VPA. Like VPA, it protected against MES- and PTZ-induced seizures, suggesting that it can prevent seizure spread and raise the seizure threshold in these tests. ABT-769 was clearly more potent than VPA after oral administration. The duration of action of ABT-769 in the mouse was comparable to that of VPA. ABT-769, like VPA, maintained efficacy after repeated dosing, which is consistent with the low tolerance liability of VPA in the clinic. The efficacy of ABT-769 in the mouse seizure models was correlated to the plasma and brain concentrations of the compound. The effective plasma concentration of ABT-769 in the mouse is 17 μg/ml in the MES seizure test. Full efficacy was observed at doses that did not impair motor coordination or inhibitory avoidance learning and did not increase slow-wave EEG activity. These findings suggest that seizure protection is not accompanied by generalized inhibition of brain activity and function, although motor impairment is observed at super-efficacious doses.
Rats of the WAG/Rij strain spontaneously express EEG spike-and-wave patterns that are similar to those of nonconvulsive human absence (petit mal) epilepsy (23). ABT-769 blocked absence-like activity in the WAG/Rij rat. It was clearly more potent than VPA in this test. No sign of tolerance to ABT-769 was found during a 7-day repeated dosing study. Spike–wave discharge activity was blocked until 3 to 5 days after the last dose of ABT-769, suggesting the possibility that it produced residual pharmacodynamic changes in the brain, a characteristic that has been reported for VPA (24,25).
ABT-769 reduced electrically kindled amygdala seizures in a dose-related manner. Electrically kindled seizures, a model of focal seizures in epilepsy, have been used to identify new AEDs with the potential for treating complex partial seizures (2,26). The effects on the average behavioral seizure score and on the cortical and amygdala afterdischarge duration were similar for ABT-769 and VPA. Both compounds shortened electrical seizure duration in the amygdala and the cortex, indicating that they raised the seizure threshold but did not prevent seizure spread from the amygdala to the cortex.
For comparative purposes, the profiles in animal seizure models of several first- and second-generation AEDs are shown in Table 5. ABT-769 has an anticonvulsant profile in key animal-seizures tests like that of VPA, phenobarbital, felbamate, and the benzodiazepines. This group of drugs has a broad spectrum of antiseizure activity in animals and is effective against focal and generalized seizures in human epilepsy. ABT-769 does not produce tolerance like the benzodiazepines and does not have the sedative-like profile of the barbiturates. The anticonvulsant profile of ABT-769 is clearly different than that of PHT and CBZ, which are ineffective in absence and myoclonic seizures. PHT, CBZ, tiagabine, and gabapentin exacerbate absence-like seizures in the WAG/Rij rat (16,23,27–29), whereas ABT-769 clearly blocks absence-like seizures in this animal model.
Table 5. Antiseizure profile of ABT-769 and other key antiepileptic drugs
Protection in animal seizures models
+, Effective; ±, inconsistent efficacy; NE, not effective; EX, exacerbates.
Data from Loscher (2,6), Peeters et al. (16), van Luijtelaar, Coenen (23), Coenen et al. (27), van Luijtelaar et al. (28), van Rijn et al. (29), and Constanti et al. (30).
Hepatotoxicity and teratogenicity, two serious, but very infrequent, adverse events associated with VPA pharmacotherapy, have been carefully examined in animals. Although a complete understanding of the liver toxicity of VPA is not yet available, it appears to be related to the inhibition of the β-oxidation of fatty acids and disruption of mitochondrial respiration (10,11). Moreover, metabolism of VPA to hepatotoxic metabolites, such as the 2,4-diene, 4-ene, and 3-keto-4-ene, is likely involved (10). In the studies reported here, ABT-769, its enantiomer, and main metabolites, in contrast to VPA, did not inhibit β-oxidation at concentrations ≤1,000 μM, suggesting that ABT-769 has a very low liability for drug-induced liver toxicity.
Mice were selected for the reproductive toxicology study on ABT-769 because VPA produces a high rate of neural tube defects in this species, although the particular neural tube defect produced in the mouse, exencephaly, differs from that most common in humans, spina bifida. VPA can produce spina bifida in the mouse, but the incidence of this particular defect is low (31). Therefore exencephaly was used as a surrogate measure of this particular liability. In contrast to VPA, ABT-769 did not cause neural tube defects in the mouse at maternally toxic doses that were selected to provide an adequate exposure of the fetus to ABT-769.
Taken together, the effects of ABT-769 in MES and PTZ seizure models in the mouse, in the WAG/Rij rat, and in the amygdala-kindled rat indicate that ABT-769 has the potential to be a broad-spectrum AED in humans. Efficacy in these antiseizure models predicts activity against generalized tonic–clonic seizures, secondarily generalized seizures, and absence seizures (2,26). The mechanism of action of ABT-769 has not been studied, but the broad-spectrum activity of ABT-769 suggests that multiple mechanisms of action may be involved (6). The efficacy observed with ABT-769 in preclinical seizure models also suggests that it has potential in the treatment of other CNS disorders. Several drugs developed initially for epilepsy have beneficial effects in a number of common neurologic and psychiatric disorders including bipolar disorder, migraine, neuropathic pain, and movement disorders (32). It has been reported that one third of patients currently taking AEDs do so for the treatment of diverse CNS disorders other than epilepsy (33). For example, clinical data support a broad utility of VPA in psychiatry beyond the bipolar mania for which it has labeling (4,5). Thus given the increasingly diverse range of clinical utility being recognized with AEDs, it is likely that ABT-769, because it is a broad-spectrum antiseizure compound based on VPA chemistry, will have beneficial effects for the treatment of a variety of neurologic and psychiatric disorders.