0022-3565/05/3122-866–874$20.00 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 312, No. 2 Copyright © 2005 by The American Society for Pharmacology and Experimental Therapeutics 68825/1193701 JPET 312:866–874, 2005 Printed in U.S.A.

First Demonstration of a Functional Role for Central Nervous System Betaine/␥-Aminobutyric Acid Transporter (mGAT2) Based on Synergistic Anticonvulsant Action among Inhibitors of mGAT1 and mGAT2

H. Steve White, William P. Watson, Suzanne L. Hansen, Scott Slough, Jens Perregaard, Alan Sarup, Tina Bolvig, Gitte Petersen, Orla M. Larsson, Rasmus P. Clausen, Bente Frølund, Erik Falch, Povl Krogsgaard-Larsen, and Arne Schousboe

Anticonvulsant Drug Development Program, Department of Pharmacology and Toxicology (H.S.W.), University of Utah, Salt Downloaded from Lake City, Utah; Departments of Neuropharmacology (W.P.W., S.L.H., S.S.) and Medicinal Chemistry (J.P.) H. Lundbeck A/S, Valby, Denmark; and Departments of Pharmacology (A.Sa., T.B., G.P., O.M.L., A.Sc.) and Medicinal Chemistry (R.P.C., B.F., E.F., P.K.-L.), The Danish University of Pharmaceutical Sciences, Copenhagen, Denmark Received April 7, 2004; accepted June 9, 2004 jpet.aspetjournals.org

ABSTRACT In a recent study, EF1502 [N-[4,4-bis(3-methyl-2-thienyl)-3- Frings audiogenic seizure-susceptible mouse and the pentyle- butenyl]-3-hydroxy-4-(methylamino)-4,5,6,7-tetrahydrobenzo netetrazol seizure threshold test. In contrast, combination of the [d]isoxazol-3-ol], which is an N-substituted analog of the GAT1- two mGAT1-selective inhibitors, and LU-32-176B, re- selective GABA uptake inhibitor exo-THPO (4-amino-4,5,6,7- sulted in only an additive anticonvulsant effect. Importantly, the tetrahydrobenzo[d]isoxazol-3-ol), was found to inhibit GABA combination of EF1502 and tiagabine did not result in a greater

transport mediated by both GAT1 and GAT2 in human embryonic than additive effect in the rotarod behavioral impairment test. In at ASPET Journals on May 3, 2019 kidney (HEK) cells expressing the mouse GABA transporters subsequent in vitro studies conducted in HEK-293 cells express- GAT1 to 4 (mGAT1–4). In the present study, EF1502 was found to ing the cloned mouse GAT transporters mGAT1 and mGAT2, possess a broad-spectrum anticonvulsant profile in animal mod- EF1502 was found to noncompetitively inhibit both mGAT1 and ␮ els of generalized and partial epilepsy. When EF1502 was tested the betaine/GABA transporter mGAT2 (Ki of 4 and 5 M, respec- in combination with the clinically effective GAT1-selective inhibitor tively). Furthermore, in a GABA release study conducted in neo- tiagabine [(R)-N-[4,4-bis(3-methyl-2-thienyl)-3-butenyl]nipecotic cortical neurons, EF1502 did not act as a substrate for the GABA acid] or LU-32-176B [N-[4,4-bis(4-fluorophenyl)-butyl]-3-hydroxy- carrier. Collectively, these findings support a functional role for 4-amino-4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol], another GAT1- mGAT2 in the control of neuronal excitability and suggest a pos- selective N-substituted analog of exo-THPO, a synergistic rather sible utility for mGAT2-selective inhibitors in the treatment of than additive anticonvulsant interaction was observed in the epilepsy.

Reduction of GABA-mediated inhibitory neurotransmis- tion or inhibition of high-affinity transport have been dem- sion is associated with seizure activity and drugs that elevate onstrated to possess anticonvulsant activity (see Dalby, 2003; synaptic GABA levels either by inhibition of GABA degrada- Sarup et al., 2003). For example, the GABA-transaminase inhibitor and the GABA-transport inhibitor tiaga- bine [(R)-N-[4,4-bis(3-methyl-2-thienyl)-3-butenyl]nipecotic This work was supported by Danish State Medical Research Council Grant 20-00-1011, the Lundbeck Foundation, and National Institute of Neurological acid] are clinically effective antiepileptic drugs (for review Disorders and Stroke Grant NO1-NS-9-2313. and references, see Ben-Menachem, 2002; Kalviainen, 2002). Article, publication date, and citation information can be found at Since the advent of cloning of several GABA transporters http://jpet.aspetjournals.org. doi:10.1124/jpet.104.068825. from different species including mouse, rat, and human, in-

ABBREVIATIONS: GAT, GABA transporter; SKF 89976A, (R)-N-(4,4-diphenyl-3-butenyl) ; SKF 100330A, N-(4,4-diphenyl-3-butenyl)gu- vacine; CI-966, N-2-bis[4-(trifluoromethylphenyl) methoxy]ethyl ; exo-THPO, 4-amino-4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol; EF1502, N-[4,4- bis(3-methyl-2-thienyl)-3-butenyl]-3-hydroxy-4-(methylamino)-4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol; LU-32-176B, N-[4,4-bis(4-fluorophenyl)-butyl]- 3-hydroxy-4-amino-4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol; HEK, human embryonic kidney; PTZ, pentylenetetrazol; ANOVA, analysis of variance; EF1500, N-[4,4-bis (3-methyl-2-thienyl)-3-butenyl]-3-hydroxy-4-amino-4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol; NNC 05-2045, 1-(3-9H-carbazol-9-yl)- 1-propyl)-4-(4-methoxyphenyl)-4-piperidinol; NNC 05-2090, 1-(3-(9H-carbazol-9-yl)-1-propyl)-4-(2-methoxyphenyl)-4-piperidinol. 866 Function of Central Nervous System Betaine/GABA Transporter 867 terest has been focused on the development of inhibitors of afforded protection against maximal electroshock, audio- the GABA transporter (GAT) 1. In support of this principle, a genic, and kindled seizures (Dalby et al., 1997) Unfortu- number of GAT1-selective inhibitors have been developed nately, these inhibitors, which structurally are not GABA and demonstrated to possess anticonvulsant activity in ani- mimetics (Thomsen et al., 1997), were shown to have affinity mal seizure models (Swinyard et al., 1991; Pavia et al., 1992; for ␣-1 adrenergic and D-2 dopamine receptors (Dalby et al., Suzdak and Jansen, 1995; White et al., 2002; Dalby, 2003). 1997). As such, it is difficult to make any definitive conclu- For example, the first peripherally active GABA uptake in- sion regarding the mechanism underlying their anticonvul- hibitor was SKF 89976A [(R)-N-(4,4-diphenyl-3-butenyl) sant action. nipecotic acid], an N-substituted diphenyl butenyl analog of The present study was undertaken in an effort to obtain nipecotic acid (Swinyard et al., 1991). The corresponding detailed information about newly developed N-substituted an- analogs of guvacine, SKF 100330A [N-(4,4-diphenyl-3-bute- alogs of exo-THPO (4-amino-4,5,6,7-tetrahydrobenzo[d]isox- nyl)guvacine] and CI-966 [N-2-bis[4-(trifluoromethylphenyl) azol-3-ol; see Fig. 1) that have been shown to inhibit to varying methoxy]ethyl guvacine], supported the principle that mod- degrees GABA transporters other than GAT1 (Clausen et al., ulation of GABA transport would result in an anticonvulsant 2005). Of particular interest is the analog EF1502 [N-[4,4-bis(3- action (Swinyard et al., 1991; Pavia et al., 1992). Unfortu- methyl-2-thienyl)-3-butenyl]-3-hydroxy-4-(methylamino)- nately, subsequent clinical trials with CI-966 revealed the 4,5,6,7-tetrahydrobenzo[d]isoxazol-3-ol], which was found to potential for serious psychotic adverse effects in certain pa- equipotently inhibit both mGAT1 and mGAT2 and was com- tients (Radulovic et al., 1993). pletely devoid of activity at a multitude of neurotransmitter Tiagabine is a second generation analog of nipecotic acid systems including adrenergic, GABAergic, and dopaminergic Downloaded from that contains the dithienyl ring system in place of the diphe- receptors (radioligand binding study of 74 different receptors nyl rings and was developed clinically without any obvious and ion channels; data not shown). The role of mGAT2 within psychotropic adverse effects. To date, tiagabine is the only the brain is poorly understood. In addition to transporting clinically available GABA transport inhibitor used for the GABA, this transporter is thought to regulate the uptake of treatment of epilepsy. However, since its introduction into the osmo-regulator betaine (for review and references, see the world-wide market, a number of individual case reports Borden et al., 1995). However, it is not clear whether modu- jpet.aspetjournals.org have suggested that tiagabine, if not titrated correctly or lation of mGAT2 contributes to control of neuronal excitabil- employed in patients with pre-existing psychiatric disorders, ity. As such, the present investigation employed a number of may possess a similar adverse-effect profile (Trimble et al., animal seizure and epilepsy models to characterize the anti- 2000) as that predicted for CI-966 (Radulovic et al., 1993). convulsant profile of this particular exo-THPO analog and to Although acute psychosis has not been substantiated in more elucidate a functional role for the betaine/GABA transporter recent studies (Sackellares et al., 2002), tiagabine use has mGAT2. The findings obtained from isobolographic analysis been associated with an increased incidence of depression of results obtained from combination studies conducted with at ASPET Journals on May 3, 2019 (for review, see Kalviainen, 2002). These patient reports EF1502 and tiagabine or LU-32-176B [N-[4,4-bis(4-fluoro- demonstrate the continued need for less toxic third genera- phenyl)-butyl]-3-hydroxy-4-amino-4,5,6,7-tetrahydrobenzo- tion GABA transport inhibitors. [d]isoxazol-3-ol] suggests a functional role for mGAT2 in the In this regard, it would seem advantageous if such trans- control of seizure activity. These findings provide the basis port inhibitors would exhibit an inhibition profile different for this report. from that of tiagabine, i.e., an affinity for transporters other than GAT1, such as GAT2 to 4. To this end, Dalby et al. (1997) have demonstrated that such inhibitors may be inter- Materials and Methods esting as antiepileptic drugs. For example, it was reported Materials. Plastic tissue culture dishes were purchased from that inhibitors of GAT3 and GAT4 (Thomsen et al., 1997) Taconic, Europe (Ry, Denmark), and fetal calf serum was purchased

Fig. 1. Chemical structures of nipecotic acid, exo-THPO, N-methyl-exo-THPO, and their li- pophilic derivatives tiagabine, LU-32-176B, EF1500, and EF1502. 868 White et al.

Ͼ from SeraLab (Crawley, UK). Poly-D-lysine (mol wt. 300.000), Ki values were calculated from the determined Vmax values of the ϭ trypsin, soybean trypsin inhibitor, DNase, cytosine arabinoside and control and test situation using the following equation: Ki I/[Vmax/ 1 Ϫ amino acids were obtained from Sigma-Aldrich (St. Louis, MO), V max 1], where I is the inhibitor concentration, Vmax is the control 1 insulin from NOVO-Nordisk A/S (Bagsvaerd, Denmark), and peni- value, and V max is the value determined in the presence of inhibitor. cillin from Leo Pharmaceuticals (Ballerup, Denmark). [3H]GABA GABA Release. Cerebral cortical neurons in culture were pre- (79.0 Ci/mmol) was from DuPont-NEN (Frankfurt, Germany). Exo- loaded with [3H]GABA (1 ␮M, 0.1 ␮Ci) for 30 min in the presence of THPO and its N-substituted analogs were synthesized as described 10 ␮M vigabatrin to irreversibly inactivate GABA-transaminase, previously (Clausen et al., 2005). Tiagabine was generously provided thereby blocking GABA (Drejer et al., 1987; Gram et al., by Sanofi Synthelabo (Brøndby, Denmark). All other chemicals were 1988). Individual cultures (35-mm Petri dishes) were subsequently of the purest grade available from regular commercial sources. The placed in a superfusion apparatus (Drejer et al., 1987) at 37°C accession numbers for mouse GAT1 and GAT2 are M92378 and equipped with peristaltic pumps, and the cells were superfused at a M97632, respectively. flow rate of 2 ml/min. Fractions from the outlet were collected every Cultured Neurons. Cerebral cortical neurons were isolated and 30 s, and at the end of the experiments, radioactivity was determined cultured from 15-day-old mouse embryos that were obtained from in all fractions. During the superfusion, either 200 ␮M nonradioac- Bomholtgaard (Ry, Denmark). After dissociation of the tissue by tive GABA or 200 ␮M EF1502 was added to the superfusion medium trypsinization and trituration in a DNase solution containing soy- for 2 min. Results were expressed as counts per minute per fraction bean trypsin inhibitor as described by Hertz et al. (1989), the cells collected. It should be noted that since the baseline of the GABA were plated onto NUNC 35-mm Petri dishes. After 48 h in culture, 20 release during the entire superfusion period was constant no major ␮M (final concentration) cytosine arabinoside was added to the cul- loss of intracellular [3H]GABA occurred during this period. ture medium to prevent astrocytic proliferation (Larsson et al., Animals. Male albino CF no. 1 mice (18–25 g; Charles River Downloaded from 1985). Cells were cultured for 7 to 8 days, at which time the neurons Laboratories, Inc., Wilmington, MA) and male and female Frings had become functionally differentiated with properties resembling audiogenic seizure-susceptible mice (18–25 g; Anticonvulsant Drug mature GABAergic neurons such as high activity of glutamate de- Development Program, University of Utah, Salt Lake City, UT) were carboxylase and pronounced vesicular GABA release (Hertz and used as experimental animals for studies conducted at the Univer- Schousboe, 1987). sity of Utah. For these studies, all animals were allowed free access Subcloning and Expression of GABA Transporters. The to both food (S/L Custom Lab Diet-7) and water except when they

cDNAs encoding the two murine GABA transporters GAT1 and were removed from their cages for the experimental procedure. Mice jpet.aspetjournals.org GAT2 (mGAT1 and mGAT2, respectively; Liu et al., 1992, 1993; were housed in a temperature and humidity controlled facility and Lopez-Corcuera et al., 1992) were subcloned into the mammalian were maintained on 12-h light/dark cycle beginning at 6:00 AM. expression vector, pCis, as described previously (Bolvig et al., 1999). Furthermore, all mice were housed, fed, and handled in a manner GAT-1pBSK was digested with XbaI, blunt-ended with Klenow en- consistent with the recommendations in the Health, Education, and zyme, and subsequently digested with NheI. The 2.2-kbp fragment Welfare publication (National Institutes of Health) 86-23, Guide for containing GAT-1 was then ligated into the 5Ј-ClaI (blunt-ended the Care and Use of Laboratory Animals. Animals were euthanized with Klenow ) and 3Ј-XbaI sites of the pCis vector. The in accordance with Public Health Service policies on the humane

1.9-kbp XbaI (5Ј)-NheI (3Ј) GAT-2 fragment from GAT-2pBSK was care of laboratory animals. at ASPET Journals on May 3, 2019 ligated into the XbaI site of the pCis vector. GATpCis cDNAs were For those studies conducted at H. Lundbeck A/S, male Wistar rats transformed into XL1-Blues bacteria (Stratagene, La Jolla, CA), and (280–320 g at start of experiment) from M&B (Laven, Denmark) plasmids were prepared by QIAGEN (Valencia, CA) column purifi- were used for amygdala kindling. For the i.v. pentylenetetrazol cation. Human embryonic kidney (HEK)-293 cells were maintained (PTZ) and i.p. pilocarpine tests, male NMRI mice (18–22 g; M&B) in complete growth medium (minimum Eagle’s medium with Earle’s were used. All animals were allowed to adapt to the laboratory salts, supplemented with 5% fetal calf serum, 1% anti-Pichia pasto- environment for at least 1 week prior to testing. Animals were ris lysyl oxidase, and 1% Glutamax I, pH 7.3 under 5% CO2). GAT- housed on a 12-h light/dark cycle, beginning at 6:00 AM. All exper- pCis transfections were carried out as described previously (White et iments were conducted between 8:00 AM and 2:00 PM. Experimental al., 2002). work was performed under a license from the Danish Ministry of [3H]GABA Uptake. The uptake of [3H]GABA in the recombinant Justice and in accordance with EC Directive 86/609/EEC and the cell systems was investigated essentially as described previously Danish law regulating experiments on animals. (Bolvig et al., 1999; White et al., 2002). The incubations, performed The test substances were either suspended or dissolved in 0.5% at 37°C in phosphate-buffered saline, were initiated by exchanging methylcellulose by trituration in a mortar and pestle and sonicating the culture medium with incubation medium leaving the cells at- for 10 min. All substances were administered i.p. or s.c. in a volume tached to the bottom of the wells during the entire procedure and of 0.01 ml/g body weight in mice and 0.05 ml/g body weight in rats. terminated (after 3 min) by rapid wash with nonradioactive incuba- Anticonvulsant Testing. The anticonvulsant profile of EF1502 tion medium. In the kinetic experiments, the GABA concentration was established by several different anticonvulsant tests that in- was varied over the range 1 to 1000 ␮M. The concentrations of cluded audiogenic, 6-Hz psychomotor, amygdala kindled, PTZ, and uptake inhibitors present during incubations in the kinetic assays pilocarpine seizures. are stated in Table 2. After incubation, the cells were dissolved in 0.4 Audiogenic Seizures. The ability of LU-32-176B, EF1502, and M KOH, and radioactivity (Schousboe et al., 1977) and protein con- tiagabine to prevent sound-induced seizures in the Frings AGS- centrations (Lowry et al., 1951) were determined. Protein contents susceptible mouse model was assessed following i.p. administration. were related to bovine serum albumin used as the standard. The For each of the analogs examined, time-response curves were kinetic parameters, Vmax and Km, of the carrier-mediated, high- constructed using a dose that produced a submaximal anticonvul- affinity GABA uptake in HEK-293 cells expressing mGAT1 and sant effect in a group of four mice per time point tested (5, 15, 30, 60, mGAT2 were calculated (Larsson et al., 1986a,b) by means of a and 120 min). Dose-response studies were then conducted at the computer program for unconstrained minimization or via the com- subsequently determined time of peak effect of each test substance. puter program Grafit v3.0 (Erithacus Software, Horley, Surrey, UK), For these studies, individual mice were placed into a Plexiglas cyl- ϭ ϩ ϩ fitting to the following equation: [V (VmaxS)/(Km S)] KS. The inder (diameter, 15 cm; height, 18 cm) fitted with an audio trans- nonsaturable influx component (KS) varied somewhat between ducer (Model AS-ZC; FET Research and Development, Salt Lake batches but was in the range of 10Ϫ3 to 10Ϫ2 ml/min/mg. There was City, UT) and exposed to a sound stimulus of 110 decibels (11 KHz) no systematic variation of this component depending on the presence delivered for 20 s. Sound-induced seizures are characterized by wild or absence of inhibitors. In the case of noncompetitive inhibition, the running followed by loss of righting reflex with fore limb and hind Function of Central Nervous System Betaine/GABA Transporter 869

limb tonic extension. Mice not displaying hind limb tonic extension volume of 0.2 ml/0.1 kg prior to stereotaxic implantation of tripolar were considered protected. electrodes (Plastics One, Roanoke, VA) into the right basolateral The anticonvulsant activity for each of the test compounds was amygdala (AP, Ϫ2.8 mm; L, Ϫ4.8; V, Ϫ8.7, measured from bregma; then quantitated in groups of four to eight mice per dose level. Paxinos and Watson, 1997). After surgery, the animals received the Varying doses of each substance were administered to groups of mice analgetic Rimadyl (5 mg/kg) and were subsequently allowed to re- until at least two points were established between the limits of 0 and cover for 2 weeks before the kindling protocol was initiated. 100% protection. Animals were stimulated daily with monophasic square wave In addition to assessing the percentage of protection following pulses (2 s, 50 Hz) using an Ellegaard Systems (Faaborg, Denmark) exposure to audiogenic stimulation, the effect of each analog on the stimulator. The intensity of the stimulus for kindling was deter- audiogenic seizure severity was determined by assigning a score to mined by increasing the stimulus in 25-␮A steps until an afterdis- the seizure phenotype as follows: 0, no response; 1, wild running for charge was observed (electroencephalogram recorded using Spike 2 Ͻ10 s; 2, wild running for Ͼ10 s; 3, all-limb clonus; 4, fore limb software); this current afterdischarge threshold was then used for extension; and 5, fore limb and hind limb extension. kindling. Stimulations were then delivered daily until five consecu- The dose of drug required to produce the desired endpoint in 50% tive grade 5 seizures were observed; at this point, the animals were

of animals tested (i.e., the ED50) and the 95% confidence interval considered to be fully kindled. were then calculated by Probit analysis of the dose-response data Assessment of Drug Effects on Afterdischarge Threshold. using a computer program based on the method described by Finney The threshold for induction of an afterdischarge was determined in (1971). kindled rats using an ascending stairstep procedure; using 10 ␮Aas EF1502 and LU-32-176B were also tested in combination with the initial stimulus, steps of 10 ␮A were applied at 1-min intervals tiagabine in the Frings AGS mouse. For these studies, varying doses until an afterdischarge was observed. Afterdischarges were defined Downloaded from (administered as a fraction of their ED50 value) of one of the drugs as being electroencephalogram spikes at least twice the amplitude of were administered i.p. The effect of a given dose level on the efficacy the prestimulus recording, with a frequency greater than 1/s, for a

of the second drug was then established by redetermining the ED50 minimum duration of 3 s. for the second drug. The time of drug administration was optimized The effect of EF1502 on afterdischarge threshold was determined so that the audiogenic seizure test was conducted at the time to peak 24 h later, 30 min after administration of EF1502 (0.3–2.5 mg/kg effect for each experimental drug. The results were plotted as func- s.c.). The afterdischarge threshold was determined using 10-␮A ␮ tion of their ED50, and 95% confidence intervals and isobolos were steps, with an initial stimulus 20 A less than that previously jpet.aspetjournals.org constructed in an effort to determine the nature of the observed measured for each rat. In addition to observing the afterdischarge interaction between EF1502 and tiagabine or LU32-176B (i.e., addi- threshold, the duration of the afterdischarge, the severity of the tive, antagonistic, or synergistic). seizure according to the Racine scale (Racine, 1972), and the number Six-Hertz Psychomotor Seizure Test. The ability of EF1502 to of rats displaying fully generalized seizures (Racine scale Ն 3) were

prevent seizures induced by 6-Hz corneal stimulation (3-s duration) also recorded. The ED50 and 95% confidence interval for protection was assessed at the time to peak effect. Current was applied by an against the fully kindled secondarily generalized seizure (Racine apparatus similar to that originally described (Woodbury and Dav- scale Ն 3 was then calculated by Probit analysis; Finney, 1971).

enport, 1952). EF1502 was evaluated at a current sufficient to pro- Pilocarpine Seizure Test. Groups of eight mice were injected at ASPET Journals on May 3, 2019 duce a seizure in 97% of the control animals tested (i.e., 22 mA). Prior with pilocarpine (250 mg/kg i.p.) 30 min after s.c. administration of to placement of the corneal electrodes, a drop of anesthetic/electro- EF1502 or vehicle; the dose of pilocarpine was chosen from a previ- lyte solution (0.5% tetracaine hydrochloride in 0.9% saline) was ously constructed dose-response curve to induce convulsions in 97% applied to the eyes of each animal. of control animals (data not shown). The percentage of animals per Six-Hertz seizures are characterized by a minimal clonic phase group that exhibited clonic convulsions within 30 min was recorded, that is followed by stereotyped, automatistic behaviors described and results were compared by Fisher’s exact probability test. originally as being similar to the aura of human patients with partial Behavioral Impairment. For those studies conducted at the seizures (for historical review and references, see Barton et al., University of Utah, minimal motor impairment was identified in 2001). Animals not displaying this behavior were considered pro- mice by the rotarod procedure as described previously (Dunham and tected. Miya, 1957). Inability of the mouse to maintain its equilibrium for Subcutaneous PTZ-Induced Seizures and i.v. PTZ Seizure one min in each of three trials on this rotating rod (6 rpm) was used Threshold. EF1502 was tested for its ability to prevent a minimal as an indication of such impairment. The effect of EF1502 on rotarod clonic seizure induced by the chemoconvulsant PTZ administered s.c. impairment was then quantitated in groups of four to eight mice per at a dose of 85 mg/kg. Absence of a 3-s clonic episode was used as the dose level. Varying doses of each substance were administered to endpoint for protection in these three tests. PTZ was dissolved in groups of mice until at least two points were established between the 0.9% saline and injected in a volume of 0.01 ml/g body weight in mice. limits of 0 and 100% impairment. The dose of drug required to Animals were observed for at least 30 min for the presence or produce the desired endpoint in 50% of animals tested (i.e., the

absence of a seizure. TD50), and the 95% confidence interval was then calculated by a Infusion of the chemoconvulsants PTZ (5 mg/ml, 0.5 ml/min) computer program based on the method described by Finney (1971). through the lateral tail vein induces clonic seizure with a loss of In combination studies conducted at H. Lundbeck A/S designed to righting reflex in mice. The threshold for producing these seizures, assess whether EF1502 enhanced the behavioral toxicity of tiagabine, a calculated as milligrams per kilogram, was determined in groups of modified rotarod procedure was employed. For this test, groups of eight 7 to 12 mice 30 min after pretreatment (i.p.) with EF1502 (four–five untrained mice were tested for six runs (30 s) on a 75-mm-diameter doses) and the corresponding vehicle. The threshold doses for the rotarod rotating at 17 rpm (Rotamex 4/8; Columbus Instruments, OH) drug treatment groups were normalized to the vehicle group and 30 min after i.p. administration of drug or vehicle. For each mouse, the expressed as a percent increase from control. In experiments where total time spent on the rotarod was recorded, and mean Ϯ S.E.M. was EF1502 was coadministered with tiagabine, they were administered calculated for each group. Significant differences between drug- and at the same time by i.p. injection on opposite sides of the abdomen. vehicle-treated groups were determined by one-way ANOVA. All data are reported as mean Ϯ S.E.M., and statistical analysis was performed by one-way ANOVA. Results Amygdala Kindling. Rats were anesthetized with Hypnorm (5 mg/ml midazolam), Dormicum (10 mg/ml fluanisone and 0.315 Inhibition of GABA Uptake. The inhibition profiles of

mg/ml fentanyl), and H2O in a ratio of 1:1:2 administered s.c. in a tiagabine and the exo-THPO analogs (see Fig. 1) on GABA 870 White et al. uptake in neurons and astrocytes as well as HEK cells trans- clonic seizures induced by s.c. PTZ and limbic seizures in- fected with the cloned mouse GABA transporters mGAT1 to duced by 6-Hz corneal stimulation. In the s.c. PTZ test,

4 obtained in a recently conducted study (Clausen et al., EF1502 prevented clonic seizures (ED50 18.6 mg/kg). In ad- 2005) are summarized in Table 1. It can be seen that EF1502 dition, EF1502, at a dose as low as 2.5 mg/kg i.p., produced a (the dithienyl-butenyl analog of N-methyl-exo-THPO), in con- 17% increase in the i.v. PTZ threshold. In this particular test, trast to tiagabine (the dithienyl-butenyl analog of nipecotic EF1502 was equipotent to tiagabine (results not shown). acid), exhibited pronounced inhibitory activity toward EF1502 was also found to prevent limbic seizures in two mGAT2, whereas mGAT3 and mGAT4 were only weakly different rodent models of partial epilepsy (i.e., the 6-Hz psy- inhibited. It should be noted that EF1500 as well as LU-32- chomotor seizure test and the amygdala kindled rat model). 176B, both of which are lipophilic analogs of exo-THPO and Compared with results obtained in the Frings audiogenic sei- not of N-methyl-exo-THPO, did not display any affinity for zure model, EF1502 was only slightly less potent in the 6-Hz mGAT2–4 but only inhibited GABA transport mediated by psychomotor seizure test than in the Frings audiogenic seizure mGAT1. All of the compounds potently inhibited GABA up- model (ED50 4.4 versus 10.4 mg/kg i.p., respectively). take in neurons and astrocytes. Of the drugs tested, tiagab- In the amygdala kindled rat model of partial epilepsy, ine was approximately 10-fold more potent than the exo- EF1502 produced a dose-dependent reduction in both the THPO analogs. No neuronal or glial selectively was observed electrographic afterdischarge duration and behavioral sei- for any of the compounds tested. The kinetics of the inhibi- zure score without affecting the afterdischarge threshold (Ta- tion was investigated for EF1502, an analog of N-methyl-exo- ble 4). At the highest dose tested (2.5 mg/kg), EF1502 signif- THPO, with regard to GABA uptake in HEK cells expressing icantly decreased the afterdischarge duration by 50% and Downloaded from mGAT1 or mGAT2. As shown in Table 2, EF1502 was found reduced the seizure score from 5 Ϯ 0to2.6Ϯ 0.6. to act as a noncompetitive inhibitor of both mGAT1 and Combination Studies. In an effort to assess whether the mGAT2. mGAT2 inhibitory properties of EF1502 contributed to its Based on this information, cultured neocortical neurons anticonvulsant activity, an isobologram study was conducted in preloaded with [3H]GABA were used to investigate whether the Frings audiogenic seizure-susceptible mouse. As shown in

GABA release could be stimulated by EF1502. The cells were Fig. 3A, when the two mGAT1-selective inhibitors (LU-32-176B jpet.aspetjournals.org superfused, and the addition of a high concentration of non- and tiagabine) were combined, an additive anticonvulsant effect radioactive GABA (i.e., 200 ␮M) to the superfusion medium was observed (i.e., experimental results fell along the predicted led to a pronounced stimulation of release of radioactive line of additivity). In contrast, when the mixed mGAT1 and GABA caused by homoexchange (Fig. 2). When 200 ␮M mGAT2 inhibitor EF1502 was combined with tiagabine or LU- EF1502 was used in place of GABA, no increase in the release 32-176B, a synergistic interaction was observed (Fig. 3, B and of [3H]GABA was seen (Fig. 2). These results demonstrate a C; i.e., experimental results fell to the left of the line of additiv- 3 lack of heteroexchange between [ H]GABA and EF1502. ity). In a subsequent study, the combination of minimally active at ASPET Journals on May 3, 2019 Anticonvulsant Profile. EF1502 was evaluated in a bat- doses of EF1502 and tiagabine were found to produce a highly tery of anticonvulsant models to assess its ability to block both significant synergistic anticonvulsant action in the i.v. PTZ generalized and partial seizures in mice and rats (Table 3). threshold test (Fig. 4). Collectively, these results support the EF1502 was effective in preventing sound-induced seizures in hypothesis that mGAT2 plays a role in regulating brain excit- the Frings audiogenic seizure-susceptible mouse model of reflex ability and that an inhibitor of mGAT2 possesses anticonvul- epilepsy. EF1502 was found to block both tonic extensor and sant activity. clonic seizure components of the audiogenic seizure. Although Behavioral Impairment. The effect of EF1502 on motor less potent (ED50 4.4 versus 0.55 mg/kg i.p. for EF1502 and performance was evaluated in two different variants of the tiagabine, respectively), EF1502 was equally effective to tiaga- rotarod impairment model. In the first study, the ability of a bine in this reflex seizure model. As shown in Table 3, the mouse to maintain balance on a 2.5-cm-diameter knurled rod anticonvulsant activity of EF1502 was associated with the R- rotating at 6 rpm was evaluated. The dose producing impair- isomer; i.e., the S-isomer was found to be inactive at doses up to ment in 50% of the mice tested (i.e., TD50) was found to be 57

20 mg/kg. In a separate model, EF1502 was found to afford mg/kg (Table 3). In contrast, the TD50 for tiagabine in this protection against pilocarpine-induced seizures with 100% of test was 1.3 mg/kg i.p. In the second model, the total time the animals being protected at 10 mg/kg. In the same test, spent on a 7.5-cm-diameter rod rotating at 17 rpm was as- tiagabine was slightly more potent offering complete protection sessed over six separate sessions, each lasting 30 s. As shown against clonic seizures at 2.5 mg/kg. in Fig. 5, doses of EF1502 as high as 40 mg/kg did not EF1502 displayed dose-dependent protection against negatively impair performance in this test. In contrast, tiaga-

TABLE 1

IC50 values for inhibition of GABA uptake in neurons, astrocytes, and cloned mouse GABA transporters expressed in HEK cells Values are from Clausen et al., (2005).

IC50 Values Compound Neurons Astrocytes mGAT1 mGAT2 mGAT3 mGAT4

␮M EF1502 2 2 4 22 Ͼ300 Ͼ300 EF1500 4 2 3 130 Ͼ100 Ͼ100 LU-32-176B 2 1 4 Ͼ100 300 Ͼ300 Tiagabine 0.36 0.18 0.8 300 Ͼ300 800 Function of Central Nervous System Betaine/GABA Transporter 871

TABLE 2 Kinetic characterization of GABA transport in HEK cells expressing mGAT1 or mGAT2 by EF1502 HEK cells transiently expressing mGAT1 and mGAT2 were prepared as detailed under Materials and Methods, and the kinetic analysis of GABA transport was performed in the absence or presence of 5 ␮M EF1502 using a series of GABA concentrations as described under Materials and Methods.

Km Vmax Ki Type Control EF1502 Control EF1502

␮M nmol ϫ minϪ1 ϫ mgϪ1 ␮M mGAT1 12 Ϯ 417Ϯ 5 1.65 Ϯ 0.19 0.73 Ϯ 0.11* 4 Noncompetitive mGAT2 55 Ϯ 652Ϯ 9 0.41 Ϯ 0.04 0.21 Ϯ 0.02* 5 Noncompetitive * Statistically significant differences (p Ͻ 0.01) from control (Students t test).

of the lipophilic side chain induces a molecular restraint that interferes with the interaction between the GABA mimetic part of the molecule and the GABA binding site of the carrier protein. Since such an alternative binding site would not have any affinity for GABA, this kinetic analysis would in- dicate a mechanism of action of EF1502 that is noncompeti- tive. Such a model would be compatible with the finding that tiagabine, having the same lipophilic side chain (Fig. 1), Downloaded from Fig. 2. Cerebral cortical neurons were cultured for 7 days and preloaded binds to a site overlapping but not identical to the GABA with [3H]GABA as described under Materials and Methods. The culture dishes were subsequently placed in the superfusion system (see Materials recognition site (Braestrup et al., 1990). and Methods), fractions were collected, and the radioactivity was deter- Given the unique properties of EF1502 as an mGAT1 and mined. As indicated by the shaded bars, the cells were exposed to 200 ␮M mGAT2 GABA uptake inhibitor, it was of interest to assess ␮ GABA (fractions 7–10) or 200 M EF1502 (fractions 23–26) to assess a whether it possessed any unique attributes as an anticonvul- possible exchange with the intracellular pool of radioactively labeled GABA. Results are expressed as the average of four independent super- sant. Results obtained from a series of studies in well estab- jpet.aspetjournals.org fusion experiments Ϯ S.E.M. values. lished anticonvulsant models demonstrated that EF1502 possessed a broad-spectrum anticonvulsant profile consistent bine produced marked impairment at doses of 5 mg/kg and with that of tiagabine. Furthermore, the results from these higher (Fig. 5). Interestingly, when EF1502 was tested in studies suggest that EF1502, like tiagabine, exerts its anti- combination with tiagabine, the degree of impairment was convulsant action by elevating seizure threshold. not greater than that observed with tiagabine alone (Fig. 5). Further examination found EF1502 to be active in two

models of human partial epilepsy; i.e., 6-Hz psychomotor at ASPET Journals on May 3, 2019 seizure test and the amygdala kindled rat. The former sei- Discussion zure test possesses a unique pharmacological profile in that In contrast to the parent compounds THPO and N-methyl- the characteristic seizure induced by low frequency stimula- THPO, of which the latter has been shown to be a glial- tion is refractory to phenytoin, carbamazepine, lamotrigine, selective GABA uptake inhibitor (Falch et al., 1999; White et and topiramate but not to valproic acid and the novel anti- al., 2002), the lipophilic analogs of these compounds de- epileptic drug levetiracetam (Barton et al., 2001). Another scribed here did not exhibit cell-type selectivity (Clausen et model where EF1502 showed marked anticonvulsant activity al., 2005). However, EF1502, the lipophilic analog of N-meth- was the pilocarpine model of acute generalized seizures. Car- yl-exo-THPO, was found to display significant affinity for bamazepine, lamotrigine, and phenytoin are ineffective in mGAT2 as well as mGAT1. This was not the case for EF1500, this model, even at doses causing marked ataxic effects its non-N-methylated counterpart (Clausen et al., 2005). (Watson, 2001). To characterize EF1502’s potential for the That this N-methyl group (see Fig. 1) is an important deter- treatment of partial epilepsy, it was evaluated in the tradi- minant of GAT2 inhibitory activity is underlined by the find- tional amygdala kindled rat model of partial epilepsy. In this ing that exchanging the lipophilic moiety of the molecule (i.e., model, it was demonstrated to decrease electrographic sei- EF1500 and LU-32-176B) did not by itself affect the affinity zure duration and behavioral seizure severity. These results of the compound toward mGAT1 and mGAT2. suggest that EF1502 would be effective against primary par- The finding that EF1502 was unable to stimulate the re- tial seizures and partial seizures secondarily generalized. lease of preloaded GABA in cultured neocortical neurons Finally, it should be noted that the anticonvulsant activity of strongly suggests that it does not act as a substrate for the EF1502 was restricted to its R-stereoisomer. This finding is carrier. This is in line with the finding that it acted as a consistent with previous results, wherein inhibition of GABA noncompetitive inhibitor of GABA uptake in HEK cells ex- uptake was also found to be stereoselective (Clausen et al., pressing mGAT1 or mGAT2. It also agrees with the previous 2005). Interestingly, EF1502 seems to possess a tolerability observation that analogous lipophilic analogs of nipecotic profile that is superior to that of tiagabine. For example, in acid, which in itself is a substrate for the carrier (Larsson et the rotarod performance tests conducted, doses substantially al., 1983), do not act as substrates (Larsson et al., 1988; exceeding anticonvulsant effective doses did not seem to pro- Braestrup et al., 1990). It is of interest to note that N-methyl- duce any notable impairment. In contrast, at a dose that was THPO was previously found to be a competitive inhibitor of only slightly higher than its anticonvulsant dose in the i.v. mGAT1 (White et al., 2002), whereas in the present study, its PTZ threshold test, tiagabine produced a significant dose- lipophilic analog, EF1502, was found to be a noncompetitive dependent impairment in the rotarod test (Fig. 5). Thus, inhibitor of mGAT1. This strongly suggests that the addition based on these initial findings, it would seem that EF1502 872 White et al.

TABLE 3 Anticonvulsant profile of EF1502 following systemic administration

For each of the anticonvulsant studies conducted, the median effective (ED50) or median toxic (TD50) was determined at the time to peak effect according to the methods described in the text. The pilocarpine- and amygdala-kindled rat studies were conducted at H. Lundbeck A/S. All other anticonvulsant studies were conducted at the University of Utah.

Strain/Species Test Time of Test, Route ED50 or TD50 95% Confidence Interval PI mg/kg CF#1 mouse Rotarod 30Ј, i.p. 57.2 37.9–91.4 scPTZ 30Ј, i.p. 18.6 12.4–27.1 3.1 6Hz 30Ј, i.p. 10.4 6.8–15.5 5.5 Frings mouse Rotarod 30Ј, i.p. 18.2 14.7–22.9 AGS 30Ј, i.p. 4.4 (R,S) 3.2–5.5 4.1 AGS 30Ј, i.p. 3.1 (R) 2.0–3.9 AGS 30Ј, i.p. Ͼ20 (S) NMRI mouse Pilocarpine 30Ј, s.c. 100% protection at 10 mg/kg Wistar rat Amygdala kindling 30Ј, s.c. 0.8 0.07–4.9 Ͼ4

PI, protective index (TD50/ED50).

TABLE 4 Effect of EF1502 on amygdala-kindled rat The effect of EF1502 on afterdischarge threshold (ADT), afterdischarge duration (ADD), and behavioral seizure score (BSS) was determined 30 min after subcutaneous Downloaded from administration of varying doses of EF1502 according to the procedures described under Materials and Methods. Parametric one-way ANOVA was used to assess the effect of treatment for ADT and ADD, whereas one-way ANOVA on ranks was used for the seizure score.

Dose ADT ADD BSS (Racine Scale 0–5)

␮As Vehicle 78.3 Ϯ 25.2 81 Ϯ 9.5 5 Ϯ 0 0.30 mg/kg 55.7 Ϯ 11.3 101 Ϯ 15.6 4.6 Ϯ 0.4

0.63 mg/kg 66.9 Ϯ 10.1 54.8 Ϯ 18.5 3.1 Ϯ 0.5 jpet.aspetjournals.org 2.5 mg/kg 112 Ϯ 20.1 40.5 Ϯ 15.6 2.6 Ϯ 0.6* One-way ANOVA F(3,28) ϭ 2.2; p Ͼ 0.1 F(3,28) ϭ 3.02; p Ͻ 0.05 H ϭ 10.9 (3 df); p ϭ 0.01 * p Ͻ 0.05 compared with vehicle using Dunn’s method for multiple comparisons versus vehicle. offers some beneficial improvement over the GAT1-selective anticonvulsant activity. It is worth noting that the synergy inhibitor tiagabine. between EF1502 and tiagabine or LU-32-176B could be re-

At the present time, the role of mGAT2 within the brain is lated to pharmacokinetic interaction between these com- at ASPET Journals on May 3, 2019 not understood. This transporter is thought to regulate the pounds. Such an interaction could theoretically lead to an uptake of the osmo-regulator betaine (for references, see increase in the intracerebral concentration and an apparent Borden et al., 1995). In an effort to determine whether inhi- supra-additive effect. However, the fact that synergism was bition of mGAT2 might contribute the anticonvulsant action observed after combination of EF1502 with either of the two of EF1502, three separate combination studies with tiagab- mGAT1 inhibitors, but was not seen with the combination of ine were undertaken. The first of these, conducted in the tiagabine and LU-32-176B, the structures of which are to- Frings audiogenic seizure-susceptible mouse, clearly demon- tally different (Fig. 1), makes this highly unlikely. In addi- strated that the combination of the GAT1 and GAT2 inhibitor tion, rotarod impairment was not increased when EF1502 EF1502 with the GAT1-selective inhibitors tiagabine or LU- was combined with tiagabine. Such an effect would have been 32-176B results in a synergistic anticonvulsant action (Fig. 3, predicted by a purely pharmacokinetic interaction. B and C). This conclusion is supported by the negative find- Based on Northern blot studies in human brain and in situ ing wherein the combination of two GAT1-selective inhibitors hybridization studies in mouse brain, the (i.e., tiagabine and LU-32-176) in the same model resulted in (i.e., mGAT2) seems to be primarily localized to the extra- only an additive effect. That this synergism is indeed related synaptic component (Borden et al., 1995). However, ultra- to the difference in the affinity for mGAT1 and mGAT2 and structural immunocytochemical studies have yet to add more not to differences in chemical structures of the GABA-mi- details to these findings. On the other hand, when microin- metic moiety is supported by the fact that the GAT1 inhibi- jected into cultured hippocampal neurons, mGAT2 was re- tors tiagabine and LU-32-176B, which are structurally dif- ported to colocalize with MAP2 but not with synapsin (Ahn et ferent from each other, both with regard to their lipophilic al., 1996). Moreover, after transfection of mGAT2 into Madin side chain and their GABA mimetic moiety (Fig. 1), did not Darby canine kidney cells, mGAT2 is sorted to the basolat- exhibit synergism. This is further substantiated by the find- eral surface of the cells (Ahn et al., 1996). These findings ing that synergism could be demonstrated with EF1502 in support a somatodendritic localization of mGAT2, and it combination with either of the structurally different mGAT1 seems reasonable to hypothesize that mGAT2 may play a role inhibitors tiagabine and LU-32-176B. Additional support for in regulating extrasynaptic GABA levels. Thus, inhibiting this conclusion is provided by the combination study con- mGAT2 would be expected to increase extrasynaptic GABA ducted in the i.v. PTZ seizure threshold test. In this study, that could then activate nonsynaptic GABAA and GABAB minimally effective doses of EF1502 and tiagabine, when receptors. administered together, resulted in a greater than additive Interestingly, EF1502 was found to possess an anticonvul- effect. Collectively, these results suggest that inhibition of sant profile similar to that of the nonselective GAT2–4 in- mGAT2 by EF1502 contributes significantly to its overall hibitors NNC 05-2045 and NNC 05-2090 (Dalby et al., 1997). Function of Central Nervous System Betaine/GABA Transporter 873

Fig. 4. Synergistic anticonvulsant effect of EF1502 and tiagabine in the i.v. PTZ seizure threshold test. The effect of EF1502 (1.38 mg/kg) and tiagabine (0.63 and 1.38 mg/kg) on i.v. PTZ seizure threshold was deter- Downloaded from mined 30 min after i.p. administration in three separate groups of mice (n ϭ 7–12 mice/group). The combined effect of EF1502 and tiagabine on i.v. PTZ seizure threshold was then assessed. At the doses tested, EF1502 and tiagabine did not increase i.v. PTZ threshold when administered alone; however, the combination of EF1502 and tiagabine resulted in a significant elevation of seizure threshold above that which would have been predicted by a purely additive interaction. Results were normalized to the vehicle-treated group and are expressed as the mean Ϯ S.E.M. jpet.aspetjournals.org percentage increase in seizure threshold. Statistical significance was determined by one-way ANOVA. at ASPET Journals on May 3, 2019

Fig. 3. Additive anticonvulsant effect of LU-32-176B when combined with tiagabine (A) and synergistic anticonvulsant effect of EF1502 when com- bined with tiagabine (B) or LU-32-176B (C) in the Frings audiogenic seizure-susceptible mouse. For these studies, varying doses (adminis- Fig. 5. Lack of negative interaction between EF1502 and tiagabine in the tered as a fraction of their predetermined ED value) of LU-32-176B, 50 rotarod performance test. Thirty minutes after vehicle or drug adminis- tiagabine, or EF1502 were administered i.p. The effect of any given dose tration, groups of untrained mice (n ϭ 8 mice/group) were tested in six level on the anticonvulsant efficacy of the secondary drug was established separate trials, each lasting for 30 s (180 s total). For this study, individ- by redetermining the ED . The time of each drug administration was 50 ual mice were placed on a 75-mm-diameter rotarod rotating at 17 rpm. optimized so all anticonvulsant testing was conducted at the time to peak The cumulative time (seconds) Ϯ S.E.M. that the mouse spent on the rod effect for each drug. The results are plotted as a function of their ED 50 during the six individual trials was then averaged and plotted as a and 95% confidence intervals. The predicted line of additivity and 95% function of the dose of EF1502 (filled squares) or tiagabine (filled trian- confidence interval connect the individual ED values for each drug 50 gles) administered. For the combination study, separate groups of mice tested independently. received 15 mg/kg EF1502 and a single dose of tiagabine (1.3, 2.5, 10, and 20 mg/kg; inverted triangle). Both drugs were administered s.c. Thirty minutes after drug administration, their performance on the rotarod was In this study, the anticonvulsant activity of these compounds assessed, and the results were plotted as the mean time (seconds) Ϯ was attributed to inhibition of mGAT3 and mGAT4 despite S.E.M. spent on the rotating rod. the fact that these two inhibitors were somewhat more potent inhibitors of mGAT2 (Thomsen et al., 1997) Hence, given that EF1502 lacks affinity for mGAT3 and mGAT4, an alternative study conducted. Lack of a synergistic effect in this study conclusion would be that inhibition of mGAT2 did indeed might suggest that a selective mGAT2 inhibitor would offer contribute to the anticonvulsant efficacy of these two inhib- some distinct advantage over a nonselective mGAT1/mGAT2 itors of GABA transport. Clearly, the development of a purely inhibitor when it comes to tolerability. However, this hypoth- selective, peripherally active, mGAT2 inhibitor could resolve esis will only be substantiated or refuted at the time that a this issue. peripherally active mGAT2-selective inhibitor becomes avail- It is of particular interest that synergy was obtained in the able for testing. Overall, the results of the present investiga- anticonvulsant studies but not in the rotarod impairment tion suggest that EF1502 possesses a favorable anticonvul- 874 White et al. sant and tolerability profile and support the continued tured cerebral cortex interneurons and in cerebral cortex in vivo. Int J Dev Neurosci 3:177–185. development of EF1502 and its analogs. Larsson OM, Falch E, Krogsgaard-Larsen P, and Schousboe A (1988) Kinetic char- acterization of inhibition of gamma-aminobutyric acid uptake into cultured neu- rons and astrocytes by 4,4-diphenyl-3-butenyl derivatives of nipecotic acid and Acknowledgments guvacine. J Neurochem 50:818–823. We thank Lone Rosenquist, Kirsten Thuesen, and Laura Webb for Larsson OM, Griffiths R, Allen IC and Schousboe A (1986a) Mutual inhibition kinetic analysis of gamma-aminobutyric acid, taurine and beta-alanine high-affinity expert technical assistance, Irene Kamerath for editorial assistance, transport into neurons and astrocytes: evidence for similarity between the taurine and Nathan Nelson (Tel Aviv University, Israel) for providing the and beta-alanine carriers in both cell types. J Neurochem 47:426–432. mGAT1 and mGAT2 clones used in this study. Larsson OM, Hertz L, and Schousboe A (1986b) Uptake of GABA and nipecotic acid in astrocytes and neurons in primary cultures: changes in the sodium coupling ratio during differentiation. J Neurosci Res 16:699–708. References Liu QR, Lopez-Corcuera B, Mandiyan S, Nelson H, and Nelson N (1993) Molecular Ahn J, Mundigl O, Muth TR, Rudnick G, and Caplan MJ (1996) Polarized expression characterization of four pharmacologically distinct gamma-aminobutyric acid of GABA transporters in Madin-Darby canine kidney cells and cultured hippocam- transporters in mouse brain. J Biol Chem 268:2106–2112. pal neurons. J Biol Chem 271:6917–6924. Liu QR, Mandiyan S, Nelson H, and Nelson N (1992) A family of genes encoding Barton ME, Klein BD, Wolf HH, and White HS (2001) Pharmacological character- neurotransmitter transporters. Proc Natl Acad Sci USA 89:6639–6643. ization of the 6 Hz psychomotor seizure model of partial epilepsy. Epilepsy Res Lopez-Corcuera B, Liu QR, Mandiyan S, Nelson H, and Nelson N (1992) Expression 47:217–227. of a mouse brain cDNA encoding novel gamma-aminobutyric acid transporter. Ben-Menachem E (2002) Vigabatrin, in Antiepileptic Drugs, 5th ed (Levy RH, Matt- J Biol Chem 267:17491–17493. son RH, Meldrum BS, and Perucca E eds) pp 855–863, Lippicott Williams & Lowry OH, Rosebrough NJ, Farr AL, and Randall RJ (1951) Protein measurement Wilkins, Philadelphia. with the Folin phenol reagent. J Biol Chem 193:265–275. Bolvig T, Larsson OM, Pickering DS, Nelson N, Falch E, Krogsgaard-Larsen P, and Pavia MR, Lobbestael SJ, Nugiel D, Mayhugh DR, Gregor VE, Taylor CP, Schwarz Schousboe A (1999) Action of bicyclic isoxazole GABA analogues on GABA trans- RD, Brahce L, and Vartanian MG (1992) Structure-activity studies on benzhydrol- porters and its relation to anticonvulsant activity. Eur J Pharmacol 375:367–374. containing nipecotic acid and guvacine derivatives as potent, orally-active inhibi-

Borden LA, Smith KE, Gustafson EL, Branchek TA, and Weinshank RL (1995) tors of GABA uptake. J Med Chem 35:4238–4248. Downloaded from Cloning and expression of a betaine/GABA transporter from human brain. J Neu- Paxinos G and Watson C (1997) The Rat Brain in Stereotaxic Coordinates. Academic rochem 64:977–984. Press, San Diego. Braestrup C, Nielsen EB, Sonnewald U, Knutsen LJ, Andersen KE, Jansen JA, Racine RJ (1972) Modification of seizure activity by electrical stimulation: II. Motor Frederiksen K, Andersen PH, Mortensen A, and Suzdak PD (1990) (R)-N-[4,4- seizure. Electroencephalogr Clin Neurophysiol 32:281–294. bis(3-methyl-2-thienyl)but-3-en-1-yl]nipecotic acid binds with high affinity to the Radulovic L, Woolf T, Bjorge S, Taylor C, Reily M, Bockbrader H, and Chang T (1993) brain gamma-aminobutyric acid uptake carrier. J Neurochem 54:639–647. Identification of a pyridinium metabolite in human urine following a single oral Clausen RP, Moltzen EK, Perregaard J, Lenz SM, Sanchez C, Falch E, Frølund B, dose of 1-[2-[bis[4-(trifluoromethyl)phenyl]methoxy]ethyl]- 1,2,5,6-tetrahydro-3- Bolvig T, Sarup A, Larsson OM, et al. (2005) Selective inhibitors of GABA uptake: pyridinecarboxylic acid monohydrochloride, a gamma-aminobutyric acid uptake

synthesis and molecular pharmacology of 4-N-methylamino-4,5,6,7-tetrahydro- inhibitor. Chem Res Toxicol 6:341–344. jpet.aspetjournals.org benzo[d]isoxazol-3-ol analogues. Bioorg Med Chem, in press. Sackellares JC, Krauss G, Sommerville KW, and Deaton R (2002) Occurrence of Dalby NO (2003) Inhibition of gamma-aminobutyric acid uptake: anatomy, physiol- psychosis in patients with epilepsy randomized to tiagabine or placebo treatment. ogy and effects against epileptic seizures. Eur J Pharmacol 479:127–137. Epilepsia 43:394–398. Dalby NO, Thomsen C, Fink-Jensen A, Lundbeck J, Sokilde B, Man CM, Sorensen Sarup A, Larsson OM, and Schousboe A (2003) GABA transporters and GABA- PO, and Meldrum B (1997) Anticonvulsant properties of two GABA uptake inhib- transaminase as drug targets. Curr Drug Targets CNS Neurol Disord 2:269–277. itors NNC 05-2045 and NNC 05-2090, not acting preferentially on GAT-1. Epilepsy Schousboe A, Svenneby G, and Hertz L (1977) Uptake and metabolism of glutamate Res 28:51–61. in astrocytes cultured from dissociated mouse brain hemispheres. J Neurochem Drejer J, Honore T, and Schousboe A (1987) Excitatory amino acid-induced release 29:999–1005. of 3H-GABA from cultured mouse cerebral cortex interneurons. J Neurosci Suzdak PD and Jansen JA (1995) A review of the preclinical pharmacology of tiagabine: a potent and selective anticonvulsant GABA uptake inhibitor. Epilepsia

7:2910–2916. at ASPET Journals on May 3, 2019 Dunham MS and Miya TA (1957) A note on a simple apparatus for detecting 36:612–626. neurological deficit in rats and mice. J Am Pharm Assoc Sci Ed 46:208–209. Swinyard EA, White HS, Wolf HH, and Bondinell WE (1991) Anticonvulsant profiles Falch E, Perregaard J, Frølund B, Søkilde B, Buur A, Hansen LM, Frydenvang K, of the potent and orally active GABA uptake inhibitors SK&F 89976-A and SK&F Brehm L, Bolvig T, Larsson OM, et al. (1999) Selective inhibitors of glial GABA 100330-A and four prototype antiepileptic drugs in mice and rats. Epilepsia 32: uptake: synthesis, absolute stereochemistry and pharmacology of the enantiomers 569–577. of 3-hydroxy-4-amino-4,5,6,7-tetrahydro-1,2-benzisoxazole (exo-THPO) and ana- Thomsen C, Sorensen PO, and Egebjerg J (1997) 1-(3-(9H-Carbazol-9-yl)-1-propyl)- logues. J Med Chem 42:5402–5414. 4-(2-methoxyphenyl)-4-piperidinol, a novel subtype selective inhibitor of the Finney DJ (1971) Probit Analysis, Cambridge University Press, London. mouse type II GABA-transporter. Br J Pharmacol 120:983–985. Gram L, Larsson OM, Johnsen AH, and Schousboe A (1988) Effects of , Trimble MR, Rusch N, Betts T, and Crawford PM (2000) Psychiatric symptoms after vigabatrin and on release of endogenous and exogenous therapy with new antiepileptic drugs: psychopathological and seizure related GABA from cultured neurons. Epilepsy Res 2:87–95. variables. Seizure 9:249–254. Hertz E, Yu ACH, Hertz L, Juurlink BHJ, and Schousboe A (1989) Preparation of Watson WP (2001) Pilocarpine-induced convulsions in mice are effectively prevented primary cultures of mouse cortical neurons, in A Dissection and Tissue Culture by GABA-modulating drugs. Soc Neurosci Abstr 27:551.555. Manual for the Nervous System (Shahar A, de Vellis J, Vernadakis A, and Haber White HS, Sarup A, Bolvig T, Kristensen AS, Petersen G, Nelson N, Pickering DS, B eds) pp 183–186, Alan R. Liss, New York. Larsson OM, Frølund B, Krogsgaard-Larsen P, et al. (2002) Correlation between Hertz L and Schousboe A (1987) Primary cultures of GABA-ergic and glutamatergic anticonvulsant activity and inhibitory action on glial gamma-aminobutyric acid neurons as model systems to study neurotransmitter functions: I. Differentiated uptake of the highly selective mouse gamma-aminobutyric acid transporter 1 cells, in Model Systems of Development and Aging of the Nervous System (Ver- inhibitor 3-hydroxy-4-amino-4,5,6,7-tetrahydro-1,2-benzisoxazole and its N- nadakis A, Privat A, Lauder JM, Timiras PS, and Giacobini E eds) pp 19–31, M. alkylated analogs. J Pharmacol Exp Ther 302:636–644. Nijhoff Publishing Company, Boston. Woodbury LA and Davenport VD (1952) Design and use of a new electroshock seizure Kalviainen R (2002) Tiagabine: clinical efficacy and use in epilepsy, in Antiepileptic apparatus and analysis of factors altering seizure threshold and pattern. Arch Int Drugs, 5th ed (Levy RH, Mattson RH, Meldrum BS, and Perucca E eds) pp Pharmacodyn Ther 92:97–104. 699–704, Lippincott Williams & Wilkins, Philadelphia. Larsson OM, Drejer J, Hertz L, and Schousboe A (1983) Ion dependency of uptake Address correspondence to: Dr. H. Steve White, University of Utah, Anti- and release of GABA and (RS)-nipecotic acid studied in cultured mouse brain convulsant Drug Development Program, Department of Pharmacology and cortex neurons. J Neurosci Res 9:291–302. Toxicology, 20 S. 2030 E., Room 408, Salt Lake City, UT 84112. E-mail: Larsson OM, Drejer J, Kvamme E, Svenneby G, Hertz L, and Schousboe A (1985) [email protected] Ontogenetic development of glutamate and GABA metabolizing in cul-