- Email: [email protected]
Gabapentin increases aminooxyacetic acid-induced GABA accumulation in several regions of rat brain
Neuroscience Letters, 128 (1991) 150-154 © 1991 Elsevier Scientific Publishers Ireland Ltd. 0304-3940/91/$ 03.50 ADONIS 030439409100343W
150
NSL 07861
Gabapentin increases aminooxyacetic acid-induced GABA accumulation in several regions of rat brain W o l f g a n g L6scher ~, D a g m a r H 6 n a c k I a n d Charles P. T a y l o r z ~Department of Pharmacology, Toxicology and Pharmacy, School of Veterinary Medicine, Hannover ( F.R.G.) and 2parke Davis Pharmaceutical Research Division, Warner-Lambert Company, Ann Arbor, M I (U.S.A.) (Received 1 February 1991; Revised version received 20 March 1991; Accepted 10 April 1991)
Key words." ~,-Aminobutyric acid; Gabapentin; Substantia nigra; Epilepsy Gabapentin exerts anticonvulsant effects in different animal models of seizure states and in epileptic patients with different seizure types, but the mechanism of action of these effects is unknown. In the present study, the y-aminobutyric acid (GABA) accumulation induced by aminooxyacetic acid (AOAA) was used as a method to study the effects of gabapentin on regional turnover of GABA in the rat brain. Gabapentin was administered at a dose of 23 mg/kg i.p. (the ED95 against tonic electroconvulsions in rats) I, 2 and 8 h prior to injection of AOAA, 100 mg/kg, i.p. Gabapentin significantly increased the AOAA-induced GABA accumulation in most of the 12 brain regions examined, but the time course of the increases in GABA accumulation differed from region to region. Regions in which the time course of the increase in GABA accumulation was similar to the anticonvulsant time course of gabapentin included substantia nigra, amygdala and thalamus. The data suggest that an effect of gabapentin on GABA synthesis might be involved in its mechanism of anticonvulsant action.
Gabapentin (1-(aminomethyl)cyclohexaneacetic acid) is an amino acid which was designed so that, unlike yaminobutyric acid (GABA), it would penetrate the blood-brain barrier but retain as much as possible of the chemical and physical properties of GABA [1]. In preclinical investigations, gabapentin was shown to exert pronounced anticonvulsant effects at non-toxic doses in various seizure models, including traditional models, such as pentylenetetrazol-induced clonic convulsions in mice and maximal electroshock seizures (MES) in rats [1]. As could be predicted from its preclinical efficacy, add-on trials in epileptic patients indicated that gabapentin possesses antiepileptic properties against various types of seizures [1, 2]. Indeed, gabapentin appears to be one of those novel antiepileptics which might offer advantages to drugs currently used in antiepileptic therapy
[8]. Despite the numerous preclinical and clinical studies which have been undertaken with gabapentin, the pharmacological mechanism of action of this drug is still unidentified. In spite of its structural similarities to GABA, it has no affinity to GABAA or GABAB recepCorrespondence: W. L6scher, Department of Pharmacology, Toxicology and Pharmacy, School of Veterinary Medicine, Bfinteweg 17, D3000 Hannover 71, F.R.G.
tors, and effects on benzodiazepine receptors are found only at high (millimolar) concentrations [1]. Similarly, inhibition of the GABA-degrading enzyme GABA-aminotransferase (GABA-T) is seen only at high (millimolar) concentrations of gabapentin, and the drug does not interfere with the GABA uptake system [1]. After administration of gabapentin to mice, the GABA concentrations determined in nerve terminals (synaptosomes) isolated from the brains of the animals were not elevated [1]. The only (indirect) evidence which might indicate a GABAergic effect comes from experiments in which gabapentin was shown to exert similar inhibitory effects on the release of dopamine, noradrenaline and 5-hydroxytryptamine as GABA or the GABAB receptor agonist baclofen, although interaction experiments indicated that the presynaptic site of action of gabapentin might be different from that of GABA and baclofen [13, 14]. Whereas the available biochemical data indicate that gabapentin does not exert effects on GABA receptors or GABA degradation and uptake, a possible action on GABA synthesis has not been studied as yet. In the present paper, inhibition of GABA-T by aminooxyacetic acid (AOAA) was used as a method for estimation of regional GABA turnover in rats [7, 11]. With this method, the regional profile and temporal pattern of gabapentin's
151
effects on GABA accumulation were determined. The dosage of gabapentin chosen for respective experiments (23 mg/kg i.p.) was the MES ED95 in rats, determined 2 h after dosing. For all experiments, female rats of the Wistar strain, weighing 200-220g, were used. GABA turnover was estimated after i.p. injection of AOAA hemihydrochloride, 100 mg/kg (dissolved in saline and injected at a volume of 2 ml/kg). As recently shown [11], at this dosage AOAA rapidly and completely inhibits the GABA-degrading enzyme GABA-T in the 12 brain regions examined in this study without reducing the activity of the GABA synthesizing enzyme, glutamate decarboxylase (GAD). The increase in GABA concentrations induced with i.p. injection of AOAA, 100 mg/kg, is rapid in onset, allowing one to estimate GABA turnover rates from the initial rate of GABA accumulation, i.e. during the first 30-60 min after AOAA injection [11]. For the present study, 9 groups of 6 age-matched rats each were used. Except for the last group, two groups were used per experimental day. More than two groups could not be used per day, since two groups already meant that 144 individual brain regions had to be dis-
sected and processed through the biochemical assays (see below). The first two groups were used to reestablish the extent of GABA accumulation induced by AOAA. For this purpose, one group was injected with saline (2 ml/kg i.p.) and the other group with AOAA and both groups were decapitated after 1 h. For the gabapentin experiments, gabapentin, 23 mg/kg, was injected i.p. l, 2 and 8 h prior to AOAA (i.e. 2, 3 and 9 h prior to sacrifice), using groups of 6 rats per time. Gabapentin was freshly dissolved in saline and injected at a volume of 2 ml/kg. Together with each gabapentin-treated group, a second group of 6 rats was injected with saline (2 ml/kg i.p.) l, 2 and 8 h prior to AOAA, so that the GABA accumulation with and without gabapentin could be compared directly. All animals were killed by decapitation 1 h after AOAA injection. At the last experimental day, one control group without gabapentin or AOAA was used for redetermination of basal GABA levels; in these animals, in which saline (2 ml/kg) was injected i.p. 1 h prior to decapitation, GABA levels were not significantly different from those determined at the first experimental day. In order to avoid variation in GABA turnover by circadian rhythms, all groups of rats were killed at the same
GABA accumul.ation induced by AOAA [ nmo[/mg protein] 110-
[--] Control. • l h r afterAOAA
100, 90. 80. 70-
~r
605040-
0
Olfactory Frontal bulb corfex
Corpus Hippo- Amygdala Thalataus Hype- Tectum Subsf. sfriatum campus thalamus nigra
Pens
Mad. obl.
CerQl:~lla¢ cortex
Fig. 1. Effect o f A O A A , 100 mg/kg i.p., on regional G A B A levels in rat brain. Data are means _+ S.D. of 6 saline- and 6 AOAA-pretreated rats. G A B A levels were determined I h after saline or A O A A injection. Significance of differences is indicated by asterisks ( P < 0.001).
152
time in the morning (between 8.00 h and 9.00 h a.m.). After decapitation, the brains were rapidly removed and dissected on a cold plate at - 13°C into 12 brain regions within 4 min after decapitation. The dissection technique was in principle that described by L6scher et al. [9] with some modifications described recently [11]. Following dissection, the individual (bilaterally pooled) regions were rapidly homogenized in 2 ml of ice-cold 80 % ethanol. Previous experiments with and without pretreatment of rats with the GAD-inhibitor, 3-mercaptopropionic acid, have shown that no postmortem increase of GABA occurs when~ brain regions are dissected at - 1 8 ° C within 4 min after decapitation and are then rapidly homogenized in ethanol [9]. Further processing of the homogenates and determination of GABA by the enzymatic 'GABAse' method was carried out as described in detail elsewhere [9]. Significance of differences in GABA levels between the two groups of each experimental day was calculated by Student's t-test. As shown recently [11], at the dose administered (100 mg/kg i.p.), AOAA induced significant increases of GABA levels in all 12 brain regions examined, ranging from 70 to 141% (Fig. 1). When GABA turnover rates were calculated from the difference between the GABA concentration after treatment with AOAA and the control level, the following turnover rates (in nmol/h/mg protein) were obtained: olfactory bulb, 24.7; frontal cortex, 12.4; striatum, 14.2; hippocampus, 23.7; amygdala, 25.6; thalamus, 22.7; hypothalamus, 26.6; tectum, 31.1; substantia nigra, 48.7; pons, 9.4; medulla, 11.7; cerebellum, 14.3. These estimations of regional turnover rates correspond to values previously reported for the AOAA method and other available procedures for estimating GABA turnover, such as local injection of GABA-T inhibitors or i.v. infusion of ]3C-labelled glucose [11]. We have shown in previous studies that the differences in regional GABA turnover rates determined by the AOAA method are correlated significantly with the differences of GAD activities in the respective brain regions, suggesting that the AOAA-induced GABA accumulation can be used as an index of regional GABA synthesis rates [11]. Pretreatment with gabapentin, 23 mg/kg 1 h prior to AOAA, significantly increased the AOAA-induced GABA accumulation in olfactory bulb, frontal cortex, corpus striatum, amygdala, thalamus, and substantia nigra by 16-53 %, suggesting an increase in GABA turnover in these regions (Fig. 2). As indicated by previous experiments [1], the increase in AOAA-induced GABA accumulation found in gabapentin-pretreated rats was certainly not due to an effect of gabapentin on GABA degradation or basal GABA levels. When gabapentin was administered 2h prior to
AOAA, except in amygdala, thalamus and pons the GABA accumulation was not significantly higher than that determined with AOAA alone, indicating a rapid decrease of gabapentin's effect (Fig. 2). However, administration of gabapentin 8 h prior to AOAA led to significant increases in GABA accumulation in most brain regions, including several regions (hypothalamus, tectum, medulla oblongata, cerebellar cortex) in which no increase had been observed with shorter pretreatment times. Thus, these data indicated regional differences in
EffectofgLbapentinonAOAA-induced GABA accumulation
[*/,]
Gabapentin lhl--I 2h B~ 8 h D prior to AOAA
200~
150 III °__
.
10050"'l
0 Olfactory Frontal bulb
cortex
.
.
.
.
.
.
.
,I
Corpus stri aturn
I%]
200.
.
.
.
.
.
Hippo- Amygdata Thalamus campus *** ***
et
150 ~t
100 "o 50
......... r 1.... Hypo- Tectum Substantia Pons thalamus nig~
Medulla Cerebettar
oblongata
cortex
Fig. 2. Effect of gabapentin on regional G A B A accumulation induced by A O A A in rats. Gabapentin, 23 mg/kg i.p., was injected 1, 2 or 8 h prior to A O A A , 100 mg/kg. G A B A accumulation was determined 1 h after injection of A O A A as shown in Fig. 1. The effects of gabapentin on AOAA-induced G A B A accumulation are shown in percent compared to the individual control groups which were treated with A O A A alone together with each gabapentin-pretreated group. For these calculations, the G A B A levels determined after A O A A alone in the respective control groups were used as 100% (indicated by hyphenate lines). Data are means + confidence limits for 95% probability (calculated by Fieller's theorem) of 6 rats per time. Significant increases of G A B A accumulation in groups treated with gabapentin plus A O A A compared to the individual time-matched group treated with A O A A alone are marked by asterisks (*P < 0.05: **P < 0.02; ***P < 0.001 ).
153
the temporal pattern of gabapentin's effects on GABA turnover. In some regions, i.e., olfactory bulb, cortex, and striatum, the action of gabapentin appeared to be biphasic (Fig. 2). The late effects of gabapentin on GABA accumulation were unexpected since the half-life of the drug in plasma after i.v. or oral administration in rats is only about 2 h, and there is no evidence for accumulation of active metabolites [15]. In order to prove the possible functional significance of the differences in regional alterations in GABA accumulation with respect to the anticonvulsant effects of gabapentin, the effects of gabapentin on electrically induced seizures were examined at different times after i.p. injection. As already mentioned above, the dose (23 mg/kg i.p.) used for the biochemical determinations was the ED95 in the MES test in rats (determined 2 h after injection; EDs0 at this time was 3.5 mg/kg i.p.). Time course studies using the MES test in rats showed that gabapentin exerted its peak effect after 2 h, and anticonvulsant activity rapidly decreased thereafter. With respect to the time of peak anticonvulsant effect, i.e. 2 h after injection, it should be noted that the initial increases in GABA accumulation were also measured 2 h after administration of gabapentin (because of the 1-h GABA accumulation period with AOAA), while 3 h after gabapentin no increased GABA accumulation was found in most regions (Fig. 1). Nine hours after injection, i.e. at the time at which the late effects on GABA accumulation had been determined in several brain regions, gabapentin did not protect rats in the MES test. Indeed, the drug did not even increase the threshold for tonic electroconvulsions at this time (determined with an up-and-downmethod [8] in groups of 15 female rats). These data thus demonstrate that gabapentin exerts potent anticonvulsant effects in rats after 2 h, but this effect is completely lost after 9 h. The only brain regions for which a similar time course of gabapentin's neurochemical effects was found (i.e. significant initial increase in GABA accumulation, but no effect after 9 h) were the substantia nigra, the amygdala and the thalamus (Fig. 2). In recent GABA turnover experiments with valproate (VPA) in rats, a clinically established antiepileptic drug with a spectrum of anticonvulsant activity similar to that of gabapentin, increases in GABA synthesis rates were found in substantia nigra and corpus striatum, but not in other brain regions, indicating a regional selectivity in VPA's effects [7]. The increase in GABA turnover induced by VPA in some brain regions was related to findings that VPA is capable of increasing the GAD activity of brain tissue homogenates from treated mice or rats [4, 12]. Indeed, Phillips and Fowler [12] reported that treatment of rats with VPA increased GAD activity in some regions, including the midbrain, whereas no sig-
nificant increases were found in other regions, e.g. the cortex, suggesting that the effect of VPA on GAD activity may be specific for certain brain regions. The present data on gabapentin might indicate that this drug is also capable of increasing GABA synthesis, but in contrast to VPA, this effect seems not to be restricted to one or two brain regions. Interestingly, both drugs caused significant increases of AOAA-induced GABA accumulation in substantia nigra, a region which has repeatedly been related to the mechanism of action of anticonvulsant drugs, especially those which act through facilitation of GABAergic neurotransmission [3, 16]. With respect to the late effects of gabapentin on GABA accumulation in several brain regions, it should be noted that delayed effects on GABA metabolism have been described also for VPA [5]. These delayed effects of VPA on GABA metabolism could be one explanation for experimental and clinical findings showing a time delay of days or even weeks between onset of chronic treatment with VPA and attainment of the maximum anticonvulsant effect [6, 10]. Thus, further studies, including experiments with chronic administration, should be carried out to study the delayed effects of gabapentin in more detail. In conclusion, the present study suggests that gabapentin is capable of increasing GABA synthesis in several brain regions. However, the biochemical basis of these increases and their time course remain to be clarified. We thank Prof. W. Lehmacher (Department of Biostatistics, School of Veterinary Medicine, Hannover) for advice in the statistical evaluation of the present data. We appreciate the skillful technical assistance of Mrs. Gramer, Mrs. Pieper and Mrs. Bartling. The experiments were initiated partly in response to suggestions by R. Silverman (Northwestern University, U.S.A.) that 3substituted GABA compounds may increase the activity of GABA-synthesizing enzyme(s).
1 Bartoszyk, G.D., Meyerson, N., Reimann, W., Satzinger, G. and von Hodenberg, A., Gabapentin. In B.S. Meldrum and R.J. Porter (Eds.), New Anticonvulsant Drugs, John Libbey, London, 1986, pp. 147-163. 2 Chadwick, D. and the U.K. gabapentin study group, Gabapentin in partial epilepsy, Lancet, 335 (1990) 1114-1117. 3 Gale, K., Progression of generalization of seizure discharge: anatomical and neurochemical substrates, Epilepsia, (Suppl. 2) 29 (1988) S15-$34. 4 L6scher, W., Valproate induced changes in GABA metabolism at the subcellular level, Biochem. Pharmacol., 30 (1981) 1364-1366. 5 L6scher, W., Anticonvulsant and biochemical effects of inhibitors of GABA aminotransferase and valproic acid during subchronic treatment in mice, Biochem. Pharmacol., 31 (1982) 837-842.
154 6 L6scher, W., Valproic acid. In H.-H. Frey and D. Janz (Eds.), Anticonvulsant Drugs, Handbook of Experimental Pharmacology, Vol. 74, Springer, Berlin, 1985, pp. 507 536. 7 Ltscher, W., Valproate enhances GABA turnover in the substantia nigra, Brain Res., 501 (1989) 198-203. 8 Ltscher, W. and Schmidt, D., Which animal models should be used in the search for new antiepileptic drugs? A proposal based on experimental and clinical considerations, Epilepsy Res., 2 (1988) 145-181. 9 Ltscher, W., Vetter, M., Miiller, F., Bthme, G. and StoltenburgDidinger, G., Development of a synaptosomal model to determine drug-induced in vivo changes in GABA-levels of nerve endings in 11 brain regions of the rat, Neurochem. Int., 6 (1984) 441-451. l0 Ltscher, W., Fisher, J.E., Nau, H. and Htnack, D., Valproic acid in amygdala-kindled rats: alterations in anticonvulsant efficacy, adverse effects and drug and metabolite levels in various brain regions during chronic treatment, J. Pharmacol. Exp. Ther., 250 (1989) 1067-1078. 11 Ltscher, W., Htnack, D. and Gramer, M., The use of inhibitors of GABA-transaminase for the estimation of GABA turnover in
12
13
14
15
16
various brain regions of rats: a re-evaluation of aminooxyacetic acid, J. Neurochem., 53 (1989) 1737-1750. Phillips, N.I. and Fowler, L.J., The effects of sodium valproate on ~-aminobutyrate metabolism and behaviour in naive and ethanolamine-o-sulphate pretreated rats and mice, Biochem. Pharmacol., 31 (1982) 2257-2261. Reimann, W., Inhibition by GABA, baclofen, and gabapentin of dopamine release from rabbit caudate nucleus: are there common or different sites of action?, Eur. J. Pharmacol., 94 (1983) 341-344. Schlicker, E., Reimann, W. and Gtthert, M., Gabapentin decreases monoamine release without affecting acetylcholine decrease in the brain, Arzneim. Forsch. (Drug. Res.), 35 (1985) 1347-1349. Vollmer, K.-O., yon Hodenberg, A. and Ktlle, E.U., Pharmacokinetics and metabolism of gabapentin in rat, dog and man, Arzneim. Forsch. (Drug Res.), 36 (1986) 830-839. Waszczak, B.L., Lee, E.K. and Waiters, J.R., Effects of anticonvulsant drugs on substantia nigra pars reticulata, J. Pharmacol. Exp. Ther., 239 (1986) 606-611.
150
NSL 07861
Gabapentin increases aminooxyacetic acid-induced GABA accumulation in several regions of rat brain W o l f g a n g L6scher ~, D a g m a r H 6 n a c k I a n d Charles P. T a y l o r z ~Department of Pharmacology, Toxicology and Pharmacy, School of Veterinary Medicine, Hannover ( F.R.G.) and 2parke Davis Pharmaceutical Research Division, Warner-Lambert Company, Ann Arbor, M I (U.S.A.) (Received 1 February 1991; Revised version received 20 March 1991; Accepted 10 April 1991)
Key words." ~,-Aminobutyric acid; Gabapentin; Substantia nigra; Epilepsy Gabapentin exerts anticonvulsant effects in different animal models of seizure states and in epileptic patients with different seizure types, but the mechanism of action of these effects is unknown. In the present study, the y-aminobutyric acid (GABA) accumulation induced by aminooxyacetic acid (AOAA) was used as a method to study the effects of gabapentin on regional turnover of GABA in the rat brain. Gabapentin was administered at a dose of 23 mg/kg i.p. (the ED95 against tonic electroconvulsions in rats) I, 2 and 8 h prior to injection of AOAA, 100 mg/kg, i.p. Gabapentin significantly increased the AOAA-induced GABA accumulation in most of the 12 brain regions examined, but the time course of the increases in GABA accumulation differed from region to region. Regions in which the time course of the increase in GABA accumulation was similar to the anticonvulsant time course of gabapentin included substantia nigra, amygdala and thalamus. The data suggest that an effect of gabapentin on GABA synthesis might be involved in its mechanism of anticonvulsant action.
Gabapentin (1-(aminomethyl)cyclohexaneacetic acid) is an amino acid which was designed so that, unlike yaminobutyric acid (GABA), it would penetrate the blood-brain barrier but retain as much as possible of the chemical and physical properties of GABA [1]. In preclinical investigations, gabapentin was shown to exert pronounced anticonvulsant effects at non-toxic doses in various seizure models, including traditional models, such as pentylenetetrazol-induced clonic convulsions in mice and maximal electroshock seizures (MES) in rats [1]. As could be predicted from its preclinical efficacy, add-on trials in epileptic patients indicated that gabapentin possesses antiepileptic properties against various types of seizures [1, 2]. Indeed, gabapentin appears to be one of those novel antiepileptics which might offer advantages to drugs currently used in antiepileptic therapy
[8]. Despite the numerous preclinical and clinical studies which have been undertaken with gabapentin, the pharmacological mechanism of action of this drug is still unidentified. In spite of its structural similarities to GABA, it has no affinity to GABAA or GABAB recepCorrespondence: W. L6scher, Department of Pharmacology, Toxicology and Pharmacy, School of Veterinary Medicine, Bfinteweg 17, D3000 Hannover 71, F.R.G.
tors, and effects on benzodiazepine receptors are found only at high (millimolar) concentrations [1]. Similarly, inhibition of the GABA-degrading enzyme GABA-aminotransferase (GABA-T) is seen only at high (millimolar) concentrations of gabapentin, and the drug does not interfere with the GABA uptake system [1]. After administration of gabapentin to mice, the GABA concentrations determined in nerve terminals (synaptosomes) isolated from the brains of the animals were not elevated [1]. The only (indirect) evidence which might indicate a GABAergic effect comes from experiments in which gabapentin was shown to exert similar inhibitory effects on the release of dopamine, noradrenaline and 5-hydroxytryptamine as GABA or the GABAB receptor agonist baclofen, although interaction experiments indicated that the presynaptic site of action of gabapentin might be different from that of GABA and baclofen [13, 14]. Whereas the available biochemical data indicate that gabapentin does not exert effects on GABA receptors or GABA degradation and uptake, a possible action on GABA synthesis has not been studied as yet. In the present paper, inhibition of GABA-T by aminooxyacetic acid (AOAA) was used as a method for estimation of regional GABA turnover in rats [7, 11]. With this method, the regional profile and temporal pattern of gabapentin's
151
effects on GABA accumulation were determined. The dosage of gabapentin chosen for respective experiments (23 mg/kg i.p.) was the MES ED95 in rats, determined 2 h after dosing. For all experiments, female rats of the Wistar strain, weighing 200-220g, were used. GABA turnover was estimated after i.p. injection of AOAA hemihydrochloride, 100 mg/kg (dissolved in saline and injected at a volume of 2 ml/kg). As recently shown [11], at this dosage AOAA rapidly and completely inhibits the GABA-degrading enzyme GABA-T in the 12 brain regions examined in this study without reducing the activity of the GABA synthesizing enzyme, glutamate decarboxylase (GAD). The increase in GABA concentrations induced with i.p. injection of AOAA, 100 mg/kg, is rapid in onset, allowing one to estimate GABA turnover rates from the initial rate of GABA accumulation, i.e. during the first 30-60 min after AOAA injection [11]. For the present study, 9 groups of 6 age-matched rats each were used. Except for the last group, two groups were used per experimental day. More than two groups could not be used per day, since two groups already meant that 144 individual brain regions had to be dis-
sected and processed through the biochemical assays (see below). The first two groups were used to reestablish the extent of GABA accumulation induced by AOAA. For this purpose, one group was injected with saline (2 ml/kg i.p.) and the other group with AOAA and both groups were decapitated after 1 h. For the gabapentin experiments, gabapentin, 23 mg/kg, was injected i.p. l, 2 and 8 h prior to AOAA (i.e. 2, 3 and 9 h prior to sacrifice), using groups of 6 rats per time. Gabapentin was freshly dissolved in saline and injected at a volume of 2 ml/kg. Together with each gabapentin-treated group, a second group of 6 rats was injected with saline (2 ml/kg i.p.) l, 2 and 8 h prior to AOAA, so that the GABA accumulation with and without gabapentin could be compared directly. All animals were killed by decapitation 1 h after AOAA injection. At the last experimental day, one control group without gabapentin or AOAA was used for redetermination of basal GABA levels; in these animals, in which saline (2 ml/kg) was injected i.p. 1 h prior to decapitation, GABA levels were not significantly different from those determined at the first experimental day. In order to avoid variation in GABA turnover by circadian rhythms, all groups of rats were killed at the same
GABA accumul.ation induced by AOAA [ nmo[/mg protein] 110-
[--] Control. • l h r afterAOAA
100, 90. 80. 70-
~r
605040-
0
Olfactory Frontal bulb corfex
Corpus Hippo- Amygdala Thalataus Hype- Tectum Subsf. sfriatum campus thalamus nigra
Pens
Mad. obl.
CerQl:~lla¢ cortex
Fig. 1. Effect o f A O A A , 100 mg/kg i.p., on regional G A B A levels in rat brain. Data are means _+ S.D. of 6 saline- and 6 AOAA-pretreated rats. G A B A levels were determined I h after saline or A O A A injection. Significance of differences is indicated by asterisks ( P < 0.001).
152
time in the morning (between 8.00 h and 9.00 h a.m.). After decapitation, the brains were rapidly removed and dissected on a cold plate at - 13°C into 12 brain regions within 4 min after decapitation. The dissection technique was in principle that described by L6scher et al. [9] with some modifications described recently [11]. Following dissection, the individual (bilaterally pooled) regions were rapidly homogenized in 2 ml of ice-cold 80 % ethanol. Previous experiments with and without pretreatment of rats with the GAD-inhibitor, 3-mercaptopropionic acid, have shown that no postmortem increase of GABA occurs when~ brain regions are dissected at - 1 8 ° C within 4 min after decapitation and are then rapidly homogenized in ethanol [9]. Further processing of the homogenates and determination of GABA by the enzymatic 'GABAse' method was carried out as described in detail elsewhere [9]. Significance of differences in GABA levels between the two groups of each experimental day was calculated by Student's t-test. As shown recently [11], at the dose administered (100 mg/kg i.p.), AOAA induced significant increases of GABA levels in all 12 brain regions examined, ranging from 70 to 141% (Fig. 1). When GABA turnover rates were calculated from the difference between the GABA concentration after treatment with AOAA and the control level, the following turnover rates (in nmol/h/mg protein) were obtained: olfactory bulb, 24.7; frontal cortex, 12.4; striatum, 14.2; hippocampus, 23.7; amygdala, 25.6; thalamus, 22.7; hypothalamus, 26.6; tectum, 31.1; substantia nigra, 48.7; pons, 9.4; medulla, 11.7; cerebellum, 14.3. These estimations of regional turnover rates correspond to values previously reported for the AOAA method and other available procedures for estimating GABA turnover, such as local injection of GABA-T inhibitors or i.v. infusion of ]3C-labelled glucose [11]. We have shown in previous studies that the differences in regional GABA turnover rates determined by the AOAA method are correlated significantly with the differences of GAD activities in the respective brain regions, suggesting that the AOAA-induced GABA accumulation can be used as an index of regional GABA synthesis rates [11]. Pretreatment with gabapentin, 23 mg/kg 1 h prior to AOAA, significantly increased the AOAA-induced GABA accumulation in olfactory bulb, frontal cortex, corpus striatum, amygdala, thalamus, and substantia nigra by 16-53 %, suggesting an increase in GABA turnover in these regions (Fig. 2). As indicated by previous experiments [1], the increase in AOAA-induced GABA accumulation found in gabapentin-pretreated rats was certainly not due to an effect of gabapentin on GABA degradation or basal GABA levels. When gabapentin was administered 2h prior to
AOAA, except in amygdala, thalamus and pons the GABA accumulation was not significantly higher than that determined with AOAA alone, indicating a rapid decrease of gabapentin's effect (Fig. 2). However, administration of gabapentin 8 h prior to AOAA led to significant increases in GABA accumulation in most brain regions, including several regions (hypothalamus, tectum, medulla oblongata, cerebellar cortex) in which no increase had been observed with shorter pretreatment times. Thus, these data indicated regional differences in
EffectofgLbapentinonAOAA-induced GABA accumulation
[*/,]
Gabapentin lhl--I 2h B~ 8 h D prior to AOAA
200~
150 III °__
.
10050"'l
0 Olfactory Frontal bulb
cortex
.
.
.
.
.
.
.
,I
Corpus stri aturn
I%]
200.
.
.
.
.
.
Hippo- Amygdata Thalamus campus *** ***
et
150 ~t
100 "o 50
......... r 1.... Hypo- Tectum Substantia Pons thalamus nig~
Medulla Cerebettar
oblongata
cortex
Fig. 2. Effect of gabapentin on regional G A B A accumulation induced by A O A A in rats. Gabapentin, 23 mg/kg i.p., was injected 1, 2 or 8 h prior to A O A A , 100 mg/kg. G A B A accumulation was determined 1 h after injection of A O A A as shown in Fig. 1. The effects of gabapentin on AOAA-induced G A B A accumulation are shown in percent compared to the individual control groups which were treated with A O A A alone together with each gabapentin-pretreated group. For these calculations, the G A B A levels determined after A O A A alone in the respective control groups were used as 100% (indicated by hyphenate lines). Data are means + confidence limits for 95% probability (calculated by Fieller's theorem) of 6 rats per time. Significant increases of G A B A accumulation in groups treated with gabapentin plus A O A A compared to the individual time-matched group treated with A O A A alone are marked by asterisks (*P < 0.05: **P < 0.02; ***P < 0.001 ).
153
the temporal pattern of gabapentin's effects on GABA turnover. In some regions, i.e., olfactory bulb, cortex, and striatum, the action of gabapentin appeared to be biphasic (Fig. 2). The late effects of gabapentin on GABA accumulation were unexpected since the half-life of the drug in plasma after i.v. or oral administration in rats is only about 2 h, and there is no evidence for accumulation of active metabolites [15]. In order to prove the possible functional significance of the differences in regional alterations in GABA accumulation with respect to the anticonvulsant effects of gabapentin, the effects of gabapentin on electrically induced seizures were examined at different times after i.p. injection. As already mentioned above, the dose (23 mg/kg i.p.) used for the biochemical determinations was the ED95 in the MES test in rats (determined 2 h after injection; EDs0 at this time was 3.5 mg/kg i.p.). Time course studies using the MES test in rats showed that gabapentin exerted its peak effect after 2 h, and anticonvulsant activity rapidly decreased thereafter. With respect to the time of peak anticonvulsant effect, i.e. 2 h after injection, it should be noted that the initial increases in GABA accumulation were also measured 2 h after administration of gabapentin (because of the 1-h GABA accumulation period with AOAA), while 3 h after gabapentin no increased GABA accumulation was found in most regions (Fig. 1). Nine hours after injection, i.e. at the time at which the late effects on GABA accumulation had been determined in several brain regions, gabapentin did not protect rats in the MES test. Indeed, the drug did not even increase the threshold for tonic electroconvulsions at this time (determined with an up-and-downmethod [8] in groups of 15 female rats). These data thus demonstrate that gabapentin exerts potent anticonvulsant effects in rats after 2 h, but this effect is completely lost after 9 h. The only brain regions for which a similar time course of gabapentin's neurochemical effects was found (i.e. significant initial increase in GABA accumulation, but no effect after 9 h) were the substantia nigra, the amygdala and the thalamus (Fig. 2). In recent GABA turnover experiments with valproate (VPA) in rats, a clinically established antiepileptic drug with a spectrum of anticonvulsant activity similar to that of gabapentin, increases in GABA synthesis rates were found in substantia nigra and corpus striatum, but not in other brain regions, indicating a regional selectivity in VPA's effects [7]. The increase in GABA turnover induced by VPA in some brain regions was related to findings that VPA is capable of increasing the GAD activity of brain tissue homogenates from treated mice or rats [4, 12]. Indeed, Phillips and Fowler [12] reported that treatment of rats with VPA increased GAD activity in some regions, including the midbrain, whereas no sig-
nificant increases were found in other regions, e.g. the cortex, suggesting that the effect of VPA on GAD activity may be specific for certain brain regions. The present data on gabapentin might indicate that this drug is also capable of increasing GABA synthesis, but in contrast to VPA, this effect seems not to be restricted to one or two brain regions. Interestingly, both drugs caused significant increases of AOAA-induced GABA accumulation in substantia nigra, a region which has repeatedly been related to the mechanism of action of anticonvulsant drugs, especially those which act through facilitation of GABAergic neurotransmission [3, 16]. With respect to the late effects of gabapentin on GABA accumulation in several brain regions, it should be noted that delayed effects on GABA metabolism have been described also for VPA [5]. These delayed effects of VPA on GABA metabolism could be one explanation for experimental and clinical findings showing a time delay of days or even weeks between onset of chronic treatment with VPA and attainment of the maximum anticonvulsant effect [6, 10]. Thus, further studies, including experiments with chronic administration, should be carried out to study the delayed effects of gabapentin in more detail. In conclusion, the present study suggests that gabapentin is capable of increasing GABA synthesis in several brain regions. However, the biochemical basis of these increases and their time course remain to be clarified. We thank Prof. W. Lehmacher (Department of Biostatistics, School of Veterinary Medicine, Hannover) for advice in the statistical evaluation of the present data. We appreciate the skillful technical assistance of Mrs. Gramer, Mrs. Pieper and Mrs. Bartling. The experiments were initiated partly in response to suggestions by R. Silverman (Northwestern University, U.S.A.) that 3substituted GABA compounds may increase the activity of GABA-synthesizing enzyme(s).
1 Bartoszyk, G.D., Meyerson, N., Reimann, W., Satzinger, G. and von Hodenberg, A., Gabapentin. In B.S. Meldrum and R.J. Porter (Eds.), New Anticonvulsant Drugs, John Libbey, London, 1986, pp. 147-163. 2 Chadwick, D. and the U.K. gabapentin study group, Gabapentin in partial epilepsy, Lancet, 335 (1990) 1114-1117. 3 Gale, K., Progression of generalization of seizure discharge: anatomical and neurochemical substrates, Epilepsia, (Suppl. 2) 29 (1988) S15-$34. 4 L6scher, W., Valproate induced changes in GABA metabolism at the subcellular level, Biochem. Pharmacol., 30 (1981) 1364-1366. 5 L6scher, W., Anticonvulsant and biochemical effects of inhibitors of GABA aminotransferase and valproic acid during subchronic treatment in mice, Biochem. Pharmacol., 31 (1982) 837-842.
154 6 L6scher, W., Valproic acid. In H.-H. Frey and D. Janz (Eds.), Anticonvulsant Drugs, Handbook of Experimental Pharmacology, Vol. 74, Springer, Berlin, 1985, pp. 507 536. 7 Ltscher, W., Valproate enhances GABA turnover in the substantia nigra, Brain Res., 501 (1989) 198-203. 8 Ltscher, W. and Schmidt, D., Which animal models should be used in the search for new antiepileptic drugs? A proposal based on experimental and clinical considerations, Epilepsy Res., 2 (1988) 145-181. 9 Ltscher, W., Vetter, M., Miiller, F., Bthme, G. and StoltenburgDidinger, G., Development of a synaptosomal model to determine drug-induced in vivo changes in GABA-levels of nerve endings in 11 brain regions of the rat, Neurochem. Int., 6 (1984) 441-451. l0 Ltscher, W., Fisher, J.E., Nau, H. and Htnack, D., Valproic acid in amygdala-kindled rats: alterations in anticonvulsant efficacy, adverse effects and drug and metabolite levels in various brain regions during chronic treatment, J. Pharmacol. Exp. Ther., 250 (1989) 1067-1078. 11 Ltscher, W., Htnack, D. and Gramer, M., The use of inhibitors of GABA-transaminase for the estimation of GABA turnover in
12
13
14
15
16
various brain regions of rats: a re-evaluation of aminooxyacetic acid, J. Neurochem., 53 (1989) 1737-1750. Phillips, N.I. and Fowler, L.J., The effects of sodium valproate on ~-aminobutyrate metabolism and behaviour in naive and ethanolamine-o-sulphate pretreated rats and mice, Biochem. Pharmacol., 31 (1982) 2257-2261. Reimann, W., Inhibition by GABA, baclofen, and gabapentin of dopamine release from rabbit caudate nucleus: are there common or different sites of action?, Eur. J. Pharmacol., 94 (1983) 341-344. Schlicker, E., Reimann, W. and Gtthert, M., Gabapentin decreases monoamine release without affecting acetylcholine decrease in the brain, Arzneim. Forsch. (Drug. Res.), 35 (1985) 1347-1349. Vollmer, K.-O., yon Hodenberg, A. and Ktlle, E.U., Pharmacokinetics and metabolism of gabapentin in rat, dog and man, Arzneim. Forsch. (Drug Res.), 36 (1986) 830-839. Waszczak, B.L., Lee, E.K. and Waiters, J.R., Effects of anticonvulsant drugs on substantia nigra pars reticulata, J. Pharmacol. Exp. Ther., 239 (1986) 606-611.