,

,1 ,1

*Department of Neuroscience, Faculty of Medicine, Norwegian University of Science and Technology, Norway

�Department of Pharmacology and Pharmacotherapy, Faculty of Pharmaceutical Sciences, University of Copenhagen, Copenhagen,

Denmark

Biosynthesis of GABA, the major inhibitory neurotransmitterin the mammalian brain, is dependent upon conversion of itsprecursor glutamate by glutamate decarboxylase (GAD, EC4.1.1.15). This enzyme exists in two isoforms, that is,GAD65 and GAD67, encoded for by distinct genes anddiffering according to regulation of activation and cellularlocalization (Erlander et al. 1991; Kaufman et al. 1991;Esclapez et al. 1994). Expression of GAD65 is predominantin the GABAergic synapse and has been suggested to beassociated with vesicles while GAD67 is evenly distributedthroughout the cytosol (Kaufman et al. 1991). Based on thedifferent regulation of activation and cellular localization, thetwo isoforms of GAD have been assigned distinct functions.However, the importance of this is still a matter ofuncertainty.

GABAergic neurotransmission is comprised of phasic andtonic forms of inhibition (Richerson 2004). The phasicinhibition is associated with the classical transient vesicularGABA release confined to the synapse while tonic inhibition

Received May 3, 2010; revised manuscript received September 21,2010; accepted September 27, 2010.Address correspondence and reprint requests to Helle S. Waagepe-

tersen, Department of Pharmacology and Pharmacotherapy, Faculty ofPharmaceutical Sciences, University of Copenhagen, 2 Universitetspar-ken, DK-2100 Copenhagen, Denmark. E-mail: hsw@farma.ku.dk1The major part of the work was performed in the laboratories of UrsulaSonnewald and Helle S. Waagepetersen.Abbreviations used: aCSF, artificial CSF; GAD, glutamate decarbox-

ylase; GC-MS, gas chromatography-mass spectrometry; SED, sponta-neous epileptiform discharge; TCA, tricarboxylic acid.

Abstract

GABA is synthesized from glutamate by glutamate decar-

boxylase (GAD), which exists in two isoforms, that is, GAD65

and GAD67. In line with GAD65 being located in the GABA-

ergic synapse, several studies have demonstrated that this

isoform is important during sustained synaptic transmission.

In contrast, the functional significance of GAD65 in the

maintenance of GABA destined for extrasynaptic tonic inhi-

bition is less well studied. Using GAD65)/) and wild type

GAD65+/+ mice, this was examined employing the cortical

wedge preparation, a model suitable for investigating extra-

synaptic GABAA receptor activity. An impaired tonic inhibition

in GAD65)/) mice was revealed demonstrating a significant

role of GAD65 in the synthesis of GABA acting extrasynap-

tically. The correlation between an altered tonic inhibition and

metabolic events as well as the functional and metabolic role

of GABA synthesized by GAD65 was further investigated

in vivo. Tonic inhibition and the demand for biosynthesis of

GABA were augmented by injection of kainate into GAD65)/)and GAD65+/+ mice. Moreover, [1-13C]glucose and

[1,2-13C]acetate were administered to study neuronal and

astrocytic metabolism concomitantly. Subsequently, cortical

and hippocampal extracts were analyzed by NMR spectros-

copy and mass spectrometry, respectively. Although seizure

activity was induced by kainate, neuronal hypometabolism

was observed in GAD65+/+ mice. In contrast, kainate evoked

hypermetabolism in GAD65)/) mice exhibiting deficiencies in

tonic inhibition. These findings underline the importance of

GAD65 for synthesis of GABA destined for extrasynaptic tonic

inhibition, regulating epileptiform activity.

Keywords: 13C isotopes, cortical wedge, glucose meta-

bolism, glutamate decarboxylase, hypometabolism, kainate.

J. Neurochem. (2010) 115, 1398–1408.

JOURNAL OF NEUROCHEMISTRY | 2010 | 115 | 1398–1408 doi: 10.1111/j.1471-4159.2010.07043.x

1398 Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

involves activation of extrasynaptic receptors (Rossi andHamann 1998; Farrant and Nusser 2005). Synthesis ofvesicular GABA mediating the transient phasic inhibition hasbeen suggested to be catalyzed at least in part by GAD65because of its cellular localization in the synaptic area andassociation with the vesicles (Erlander and Tobin 1991;Esclapez et al. 1994). GABA spillover from the synapseappears to mediate a prominent part of tonic inhibition (Rossiand Hamann 1998; Farrant and Nusser 2005) althoughrelease of cytosolic GABA presumably mediated via areversal of the high affinity GABA transporters may alsocontribute (Richerson 2004; Farrant and Nusser 2005; Wuet al. 2007). GAD67 has been proposed to govern synthesisof GABA responsible for tonic inhibition because of itshigher expression level and even distribution throughout theGABAergic neuron (Kaufman et al. 1991; Esclapez et al.1994; Bergado-Acosta et al. 2008).

Subsequent to release of neurotransmitter GABA and itsinteraction with receptors on the post-synaptic neuron,GABA is cleared from the synapse by uptake into the pre-synaptic neuron or into adjacent astrocytic processes(Schousboe et al. 2004). In both compartments GABA isdegraded to succinate by the combined actions of GABAtransaminase and succinate semialdehyde dehydrogenase(Schousboe and Waagepetersen 2006). Subsequent meta-bolism of succinate in the tricarboxylic acid (TCA) cycle isdependent upon a continuous supply of acetyl-CoA and isthereby energy generating. For maintenance of properneuronal GABA homeostasis, the extent of GABA uptakeinto astrocytes must be compensated for by a transfer ofglutamine from astrocytes to neurons because of the lack ofanaplerotic pathways in neurons (Waagepetersen et al.1999). Such interdependent substrate cycling betweenastrocytes and neurons, that is, the GABA/glutamate-glutamine cycle, is in line with a quantitative significanceof astrocytic glutamine as a precursor for GABA synthesis(Bradford et al. 1978; Paulsen et al. 1988; Battaglioli andMartin 1990, 1991; Sonnewald et al. 1993b; Waagepetersenet al. 2001; Rae et al. 2003). Moreover, the importance ofastrocytic glutamine for maintenance of neurotransmitterpools suggests that GABA synthesis from astrocyticglutamine is confined to the synaptic area (Behar andRothman 2001; Lebon et al. 2002).

The intricate balance between excitatory and inhibitorytransmission governed by glutamate and GABA, respec-tively, appears to be disturbed during seizures (Bradford1995; Schousboe and White 2009). In GAD65)/) mice, inwhich the gene encoding for GAD65 is deleted, the cerebralGABA content is reduced by 20–30% whereas no alterationswere observed in the cerebral glutamate content (Stork et al.2000; Walls et al. 2010). This may be associated with thepro-convulsive state of these mice in which seizures areeasily induced as well as the increased mortality related tothe development of spontaneous seizures (Asada et al. 1996;

Kash et al. 1997; Stork et al. 2000). In line with the synapticlocalization of GAD65 in the GABAergic neuron, thisisoform is essential specifically for the synthesis of GABAfrom glutamine via glutamate (Walls et al. 2010) and effectson synaptic activity in GAD65)/) mice have repeatedlybeen reported (Hensch et al. 1998; Tian et al. 1999; Choiet al. 2002). It appears that GAD65 is particularly importantwhen extensive synthesis of synaptic GABA is requiredduring peak demand (Hensch et al. 1998; Tian et al. 1999;Choi et al. 2002). In contrast to investigations of synaptic,phasic GABA transmission, there is a dearth of data on therole of GAD65 in maintenance of tonic inhibition. Recently,however, it was demonstrated using patch clamp thatGABAergic tonic inhibition is significantly reduced incortical slices from mice lacking GAD65 (Kubo et al.2009). Regulation of the extrasynaptic availability of GABA,that is, tonic inhibition, has been suggested to be importantfor seizure activity (Galvan et al. 2005; Keros and Hablitz2005; Madsen et al. 2009, 2010). Systemic or intracerebralinjections of kainate in rodents give rise to seizures andepileptiform discharges (Ben-Ari et al. 1981; Lothman andCollins 1981). Moreover, it has repeatedly been shown thatthe extrasynaptic tonic inhibition is significantly increasedfollowing exposure of hippocampal slices to kainate atconcentrations giving rise to seizures in vivo (Ben-Ari andCossart 2000; Semyanov 2004). Such increased tonicinhibition was suggested to be mediated by synaptic spilloverafter enhanced vesicular GABA release (Ben-Ari and Cossart2000) compatible with the excitatory action of kainate(Pinheiro and Mulle 2006).

In order to study the significance of GAD65 for synthesisof GABA essential for tonic inhibition, we employed thecortical wedge preparation, a model suitable for investigat-ing extrasynaptic GABA receptor activity (Ebert et al.2002; Storustovu and Ebert 2003). Spontaneous epilepti-form discharges (SEDs) were monitored in cortical wedgesfrom GAD65)/) and corresponding wild type GAD65+/+mice in the presence of the GABAA receptor agonistisoguvacine. In line with Kubo et al. (2009), we find thattonic inhibition is significantly reduced in GAD65)/) mice.The role of GAD65 for the biosynthesis of GABAmediating tonic inhibition and the metabolic consequencesof an elevated tonic inhibition was further investi-gated in vivo. A metabolic mapping was performed onGAD65)/) and GAD65+/+ mice treated with kainate toincrease tonic inhibition and the demand for GABA actingextrasynaptically. Both genotypes were injected with acombination of [1-13C]glucose and the astrocyte-specificsubstrate [1,2-13C]acetate (Waniewski and Martin 1998) forinvestigation of both neuronal and astrocytic metabolism.Information regarding the specific metabolic pathways wassubsequently provided from NMR spectroscopic and massspectrometric analysis of tissue extracts from cerebralcortex and hippocampus, respectively.

� 2010 The AuthorsJournal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408

GAD65 synthesizes GABA crucial for tonic inhibition | 1399

Materials and methods

Materials[1-13C]Glucose, [1,2-13C]acetate and D2O (99.9%) were bought

from Cambridge Isotope Laboratories (Woburn, MA, USA) and

ethylene glycol was from Merck (Darmstad, Germany). Kainate,

N,N-dimethylformamide and 1,2,3,6-tetrahydro-4-pyridine carbo-

xylic acid hydrochloride (isoguvacine) were purchased from Sigma-

Aldrich (St. Louis, MO, USA). Diazepam was bought from

Nycomed Denmark Aps (Roskilde, Denmark) and N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide as well as 1% tertbutyl-

dimethylchlorosilane were from Regis Technologies Inc. (Morton

Grove, IL, USA).

AnimalsThe GAD65)/) mice, generated as described by Kash et al. (1997),were bought from Jackson laboratories (Bar Harbor, ME, USA).

These mice have been backcrossed onto the C57BL/6 background

for 10 generations. Mice used in this study were generated from

heterozygous breeding and subsequently GAD65)/) and

GAD65+/+ mice were identified by genotyping. The breeding and

genotyping procedures were managed at Taconic (Ry, Denmark).

After delivery to the animal facilities at the Faculty of Pharmaceu-

tical Sciences, University of Copenhagen or at the Norwegian

University of Science and Technology the mice were kept under

standard conditions at ambient temperature of 20–22�C, air

humidity of 50–60%, and a 12-h light-dark cycle (lights on at

7 AM). Mice were acclimatized to the above conditions for at least

1 week before the experiments were performed. Within this

period as well as during the experiment, the animals had free

access to food and water. All protocols were approved by the

Norwegian National Animal Research Authority (07/58706) and

the animals were treated in compliance with the European

Convention (ETS 123 of 1986).

Cortical wedge preparationThe cortical wedge preparation from 15- to 16-week-old GAD65+/+

and GAD65)/) mice was obtained as previously described

(Harrison and Simmonds 1985; Ebert et al. 2002). Briefly, a wedge(400 lm thick) of brain tissue containing cerebral cortex and corpus

callosum was placed in a two-compartment bath, which was

electrically insulated with a grease gap. The left compartment

containing the cortex part was independently and continuously

superfused at 1 mL/min with O2/CO2 (95%/5%) saturated artificial

CSF (aCSF; 118 mM NaCl, 2.1 mM KCl, 1.2 mM KH2PO4,

11 mM D-glucose, 25 mM NaHCO3, 2.5 mM CaCl2). Similarly,

the right compartment containing the corpus callosum part was

continuously superfused, however, with aCSF devoid of Ca2+. Ag/

AgCl electrodes (Dri-RefTM, World Precision Instruments, Sarasota,

FL, USA) in each compartment were in contact with dishcloth tissue

providing electrical connection with the wedge. The difference in

electrical potential between the electrodes was monitored on a chart

recorder Yokogawa LR 4220E (Yokogawa Electric Corporation,

Tokyo, Japan), which was connected to a PC via a RS-232C

interface for digital sampling. The wedges were left for development

of SEDs for at least 2½ h. Characterizations of drug effects on the

SEDs were initiated when the SEDs ‡ 30 spikes per 12-min and

stable over a 40-min period. Stability was defined as no greater than

10% variation in the basal activity (frequency) of the SEDs between

two subsequent analyses. Diazepam was employed to reveal any

effect on synaptic GABA receptors. Moreover, different concentra-

tions of the GABAA receptor agonist isoguvacine were used to

investigate genotype differences in inhibitory capacity. Drugs were

dissolved in aCSF and the wedges were superfused with each drug

solution for 20 min. The sampled data were analyzed using the Mini

Analysis Program 6.0.3 (Synaptosoft Inc., Decatur, GA, USA). The

frequency of the SEDs was determined at equilibrium, that is, during

the last 12 min of each drug application.

Dose-finding for kainate in GAD65+/+ and GAD65)/) animalsKainate was used as a pharmacological intervention to intensify the

extracellular tonic inhibition. Accordingly, we wanted to employ a

dose that induces mild seizure activity but without giving rise to

generalized seizures within the time course of the experiment, that

is, 30 min (see below). In GAD65+/+ mice of a comparable strain,

intraperitoneal (i.p.) injection of 30 mg/kg kainate was previously

shown to induce stage 4–5 seizures (Zeng et al. 2007) on a modified

Racine scale (Racine 1972; Ben-Ari 1985). Therefore, the beha-

vioral effects caused by 25 mg/kg and 15 mg/kg kainate injected i.p.

were tested. Because of the higher susceptibility to seizures

previously demonstrated in GAD65)/) mice (Asada et al. 1996;Kash et al. 1997), an additional dose finding procedure for kainate

was necessary to investigate behavioral responses in this genotype.

Doses of 15 mg/kg, 7.5 mg/kg and 3.75 mg/kg kainate were tested

in GAD65)/) mice.

Metabolic studiesGAD65+/+ (n = 20) and GAD65)/) (n = 14) mice were used for

metabolic experiments at the age of 15–23 weeks and both genders

were used. The GAD65)/) and GAD65+/+ mice were given an i.p.

injection of 0.9% NaCl or kainate. GAD65)/) mice received a low

dose (3.75 mg/kg) of kainate whereas GAD65+/+ mice were given

either the same low dose or a high dose (15 mg/kg) of kainate. After

15 min the mice were injected (i.p.) with a combination of

[1-13C]glucose (543 mg/kg) and [1,2-13C]acetate (504 mg/kg) and

following another 15 min the brains of the animals were subjected

to microwave fixation, 4 kW, 1.7 s (Model GA5013, Gerling

Applied Engineering, Modesto, CA, USA) instantly inactivating

reactions and processes in the brain. Mice were observed for 30 min

post-injection of kainate or saline for incidence and severity of

seizures. Severity was graded according to a modified Racine scale

(Racine 1972; Ben-Ari 1985). After decapitation the cerebral

cortices and hippocampi were excised and kept at 75�C until

extraction employing 0.7% perchloric acid. Using a Vibra Cell

sonicator (Model VCX 750, Sonics & Materials, Newtown, CT,

USA) the samples were exposed to ultrasound for homogenization.

After centrifugation of the homogenous suspensions at 3000 g and

4�C for 5 min the precipitates were washed with distilled water and

centrifuged once again. The supernatants were pooled and pH was

adjusted to 6.5–7.5 before lyophilization.

1H and 13C NMRCortical and hippocampal extracts were both analyzed for content of

metabolites using 1H NMR. Only extracts of cortex tissue were

analyzed for 13C labeling by 13C NMR spectroscopy, as 13C

incorporation into amino acids in hippocampal tissue was not

Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

1400 | A. B. Walls et al.

quantifiable by NMR spectroscopy. Thus, in order to detect 13C

enrichment in extracts from hippocampus it was necessary to

employ gas chromatography-mass spectrometry (GC-MS) which is

more sensitive than 13C NMR. GC-MS analyses provide informa-

tion only about the number of labeled 13C atoms within a molecule,

that is, mono- or double labeling, but not about the position of the

isotope.

The lyophilized tissue extracts were reconstituted in 99% D2O

(deuterated water) containing 0.05% ethylene glycol as an internal

standard. Before transfer of the samples into 5 mm Shigemi NMR

microtubes (Shigemi Inc., Allison Park, PA, USA), pH was re-

adjusted to 6.5–7.5. The samples were analyzed using a BRUKER

DRX-600 spectrometer (BRUKER Analytik GmbH, Rheinstetten,

Germany) and the spectra were recorded at � 22� C. For the 1H

NMR spectra a pulse angle of 90� and a spectral width with 32 K

data points were employed. The acquisition time was 1.36 s and

relaxation delay was 10 s. Water suppression was achieved by

applying a low-power pre-saturation pulse at the water frequency.

The number of scans was 128 and 1024 for cortical and

hippocampal extracts, respectively.

Proton decoupled 13C NMR spectra of cortical extracts were

obtained using a 30� pulse angle and 30 kHz spectral width with

64 K data points employing an acquisition time of 1.08 s and a

relaxation delay of 0.5 s. The number of scans needed to obtain an

appropriate signal to noise ratio was typically 30 000. Relevant

peaks in the spectra were identified and integrated using

XWINNMR software (BRUKER BioSpin). The amounts of 13C

labeling and the total amounts of metabolites were quantified from

the integrals of the peak areas using ethylene glycol as an internal

standard. Factors for the nuclear Overhauser and relaxation effects

were applied to all 13C spectra. 13C Results for singlets were

corrected for naturally abundant 13C and amounts calculated from1H spectra were corrected for 13C containing metabolites.

Gas chromatography-mass spectrometryThe percentual distribution of 13C labeled mass isotopomers of

several metabolites in hippocampus was determined using GC-MS.

After 1H NMR spectroscopy, an aliquot of the sample was adjusted

to pH 2 and lyophilized. Organic acids and amino acids were

extracted into an organic phase of ethanol and benzene, dried under

air and reconstituted in N,N-dimethylformamide before they were

derivatized with N-methyl-N-(tert-butyldimethylsilyl)trifluoroaceta-

mide in the presence of 1% tertbutyldimethylchlorosilane (Mawhin-

ney et al. 1986). After derivatization compounds were analyzed in

an Agilent 6890GC gas chromatograph linked to an Agilent 5975 B

inert MSD with an electron ionization source (all from Agilent

Technologies Inc., Santa Clara, CA, USA). Results for labeled

metabolites were corrected for natural abundance of 13C.

Labeling patterns obtained from metabolism of [1,2-13C]acetate[1,2-13C]Acetate is selectively taken up by astrocytes (Waniewski

and Martin 1998) and its subsequent entrance into the TCA cycle

generates [4,5-13C]a-ketoglutarate which can be converted to

[4,5-13C]glutamate and [4,5-13C]glutamine. After relocation to the

neuronal compartments, [4,5-13C]glutamine is converted to

[4,5-13C]glutamate (Hassel et al. 1997; Shen et al. 1999; Qu et al.2000; Chowdhury et al. 2007) by the action of phosphate activated

glutaminase in the inner mitochondrial membrane (Laake et al.

1999; Kvamme et al. 2001). Subsequently, [4,5-13C]glutamate may

be decarboxylated to [1,2-13C]GABA by GAD in the cytosol of

GABAergic neurons (Walls et al. 2010).

Labeling patterns obtained from metabolism of [1-13C]glucoseGlucose is taken up into both the astrocytic and the neuronal

compartments; however, oxidative metabolism of glucose is

predominantly a neuronal event (Qu et al. 2000). Glycolytic

metabolism of [1-13C]glucose gives rise to [3-13C]pyruvate which

after conversion to [2-13C]acetyl-CoA enters the TCA cycle and

generates [4-13C]a-ketoglutarate which is precursor for [4-13C]glu-

tamate. In GABAergic neurons [4-13C]glutamate may be converted

to [2-13C]GABA. [4-13C]Glutamine may be synthesized from

[4-13C]glutamate directly in the astrocytes or generated from

[4-13C]glutamate initially labeled in the neuronal compartment

following neurotransmitter release and astrocytic uptake. By

inhibiting astrocytic TCA cycle activity it was shown that 40% of

glutamine synthesis relies on glutamate-glutamine cycle activity

(Hassel et al. 1997) suggesting that labeling in [4-13C]glutamine

partly reflects neuronal metabolism. In extracts of cerebral cortex

analyzed by NMR spectroscopy, a TCA cycling ratio, reflecting the

activity of TCA cycle metabolism (Walls et al. 2010), was

calculated based on the labeling in glutamate from [1-13C]glucose,

that is, (2*([3-13C]glutamate – [1,2-13C]glutamate))/[4-13C]gluta-

mate.

Data analysisData are presented as means ± SEM. Statistical differences

between the effects of kainate in homozygous GAD65)/) and

GAD65+/+ mice were determined using a one-way ANOVA followed

by a least significant difference post hoc test. Statistical analysis of

differences between the effect of different concentrations of

isoguvacine and genotypes, that is, GAD65)/) and GAD65+/+

mice was performed by two-way ANOVA followed by Tukey posthoc test. A value of p < 0.05 was taken to indicate a statistically

significant difference.

Results

SED measurements in the cortical wedge preparationThe effect of GAD65 deletion on SEDs was tested in thecortical wedge preparation (Fig. 1), a model suitable forinvestigating extrasynaptic GABAA receptor activity (Ebertet al. 2002; Storustovu and Ebert 2003). GABAA receptorexpression is assumed to be normal in GAD65)/) mice asreceptor density and binding affinity for muscimol wereunchanged compared to GAD65+/+ mice (Kash et al. 1999).Diazepam, which is a positive modulator of the synapticGABAA receptor, had no impact on the evolution of SEDsneither in GAD65)/) nor in GAD65+/+ mice. This is in linewith the suggestion that the GABAA receptor mediatedactivity monitored in the cortical wedge preparation repre-sents extrasynaptic GABAA receptor function (Ebert et al.2002), and is consequently considered as the controlcondition. In this model the GABAA receptor agonist

� 2010 The AuthorsJournal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408

GAD65 synthesizes GABA crucial for tonic inhibition | 1401

isoguvacine has previously been shown to exhibit highpotency at extrasynaptically located GABAA receptors (Ebertet al. 2002). In brain slices from both GAD65)/) andGAD65+/+ mice, the addition of isoguvacine led to aconcentration-dependent inhibition of neuronal activitymeasured as a decrease in the frequency of the SEDs.However, the inhibition was significantly (p < 0.001) lesspronounced in brain slices from GAD65)/) mice comparedto GAD65+/+ mice (Fig. 1).

Dose-finding for kainate in GAD65+/+ and GAD65)/)animalsKainate was employed to pharmacologically increase tonicinhibition concomitant with induction of a mild seizureactivity. The tested dose of 25 mg/kg kainate injected i.p.was found to induce stage 4–5 seizures on a modified Racinescale (Racine 1972; Ben-Ari 1985) within 20 min post-injection, that is, a severe effect which was similar to thatobserved by Zeng et al. (2007) using 30 mg/kg kainate. Incontrast, the dose of 15 mg/kg kainate led to head noddingand face washing corresponding to stage 2 seizures on amodified Racine scale (Racine 1972; Ben-Ari 1985) within5 min after injection. As the mice remained in this seizurestage throughout the next hour, this dose was taken to beadequate for the subsequent metabolic studies. This wasconfirmed by initial metabolic experiments demonstratingthat kainate treated animals exhibited significant hypome-tabolism compared to controls injected with saline (resultsnot shown) indicating a metabolic impact of an elevated tonicinhibition. Another dose finding procedure was required inGAD65)/) mice exhibiting an increased susceptibility toseizures (Asada et al. 1996; Kash et al. 1997). Treatment ofGAD65)/) animals with 15 mg/kg or 7.5 mg/kg kainate ledto death caused by seizures within 25 min, that is, within thetime course of the experiment. In contrast, treatment ofGAD65)/) mice with 3.75 mg/kg kainate led to headnodding and face washing, corresponding to stage 2 seizureson the modified Racine scale (Racine 1972; Ben-Ari 1985).Thus, to induce similar behavioral responses in GAD65)/)and GAD65+/+ mice two different doses of kainate wererequired, that is, 3.75 mg/kg and 15 mg/kg in GAD65)/)and GAD65+/+ mice, respectively, and these doses wereemployed in the metabolic studies.

In the metabolic study, all GAD65)/) mice treated with3.75 mg/kg kainate and all GAD65+/+ mice treated with15 mg/kg kainate exhibited seizure activity corresponding tostage 2 on the modified Racine scale (Racine 1972; Ben-Ari1985) and this was critical for inclusion in the study.Moreover,besides the group of GAD65+/+ animals exhibiting a similarbehavioral response to kainate as the GAD65)/) mice wewanted a group of GAD65+/+ mice receiving the same lowdose, that is, 3.75 mg/kg, of kainate, which was a subconvul-sive dose in GAD65+/+ mice. Within the groups tested nodifferences between genders were observed.

Amino acid metabolism in GAD65)/) miceThe cortical GABA content was significantly decreased inGAD65)/) mice compared to GAD65+/+ mice (Fig. 2a). Incontrast, the cortical levels of glutamate and glutamine wereunaltered in GAD65 deficient mice (Fig. 2a). The labeling in[4-13C]glutamate, [4-13C]glutamine and [2-13C]GABA wassignificantly lower in GAD65)/) mice compared toGAD65+/+ mice (Fig. 2b). The synthesis of [1,2-13C]GABAfrom [1,2-13C]acetate was decreased in GAD65)/) mice

Controlconditions 2 µM IGU

0.2 mV

10 min

5 µM IGU 20 µM IGU

GAD65+/+

0.1 mV

10 min

Controlconditions 2 µM IGU 5 µM IGU 20 µM IGU

GAD65–/–

80

100

120*

* *

*

20

40

60

*

0Control 2 µM IGU 5 µM IGU 20 µM IGU

Sp

on

tan

eou

s ep

ilep

tifo

rmd

isch

arg

es (

%)

(a)

(b)

(c)

Fig. 1 Spontaneous epileptiform discharges (SEDs) were monitored

in the absence or presence of isoguvacine (IGU; 2, 5 and 20 lM) in the

cortical wedge preparation. Representative electrophysiological

recordings obtained from brain slices of GAD65+/+ and GAD65)/)mice are shown in (a) and (b), respectively. Following termination of

drug application, the response tended to recover faster in GAD65)/)mice than in GAD65+/+ mice. Panel (c) shows the responses from

GAD65+/+ (white bars) and GAD65)/) (black bars) mice normalized

(%) to the frequency of the SEDs immediately before the drug appli-

cation was initiated. Data are presented as averages ± SEM (n = 4–6

from four different animals). All measurements were made in the

presence of 10 lM diazepam (control conditions). Cumulative appli-

cation of isoguvacine concentration-dependently inhibited the SEDs in

both genotypes. However, the inhibition was significantly less pro-

nounced in brain slices from GAD65)/) compared to GAD65+/+ mice

as indicated by asterisks (p = 0.001).

Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

1402 | A. B. Walls et al.

compared to GAD65+/+ mice while [4,5-13C]glutamate and[4,5-13C]glutamine were similar in the two genotypes(Fig. 2c). Moreover, the TCA cycling ratio calculated onthe basis of glutamate isotopomers labeled from [1-13C]glu-cose demonstrated a decrease of � 35% in GAD65)/)compared to GAD65+/+ mice. Together these findingsindicate a general attenuation of neuronal [1-13C]glucosemetabolism in GAD65)/) mice while astrocytic metabolismappears to be unaffected by the lack of GAD65. Although theexpression of GAD67 is unchanged in GAD65)/) mice(Kash et al. 1997) this isozyme might undergo adaptivecompensatory changes which could interfere with theinterpretation of the obtained results. Moreover, prior seizureactivity known to be associated with deletion of the GAD65gene, may induce neural circuitry reorganization (Kim et al.2007). The metabolic differences observed betweenGAD65+/+ and GAD65)/) mice impose the requirementfor a control group of each genotype when evaluating theeffect of kainate treatment.

In hippocampal extracts, the amount of GABA as well asthe amounts of both [1,2-13C]GABA and [2-13C]GABAweresignificantly reduced in GAD65)/) compared to GAD65+/+mice while the amounts of and labeling in glutamate andglutamine were unaltered in GAD65)/) mice (See Figure S1for details).

Amino acid metabolism in GAD65+/+ mice treated withkainateTreatment of GAD65+/+mice with the low dose (3.75 mg/kg)of kainate had no impact on the cortical contents of glutamate,glutamine and GABA (Fig. 3a, gray bars). Moreover, admi-

nistration of the low dose of kainate to GAD65+/+mice had noimpact on the amount of [4-13C]glutamate, [4-13C]glutamineand [2-13C]GABA labeled from [1-13C]glucose (Fig. 3b, greybars). In contrast, a decrease in the amount of [4,5-13C]glu-tamine labeled from [1,2-13C]acetate was observed aftertreatment of GAD65+/+ mice with the low dose of kainatewhile the amounts of [4,5-13C]glutamate and [1,2-13C]GABAwere unaltered (Fig. 3c, grey bars). Compatible with theunaltered 13C enrichment in glutamate labeled from[1-13C]glucose, the TCA cycling ratio in glutamatergicneurons calculated on the basis of glutamate labeling revealedno significant change after treatment with the low dose ofkainate (Fig. 3d, grey bars).

Administration of the high dose (15 mg/kg) of kainate toGAD65+/+ mice led to a reduction in the contents of bothglutamate, glutamine and GABA (Fig. 3a, black bars).GAD65+/+ mice treated with the high dose of kainateexhibited decreases in the amounts of glutamate, glutamineand GABA labeled from [1-13C]glucose (Fig. 3b, black bars)and in the amounts of glutamate and glutamine labeled from[1,2-13C]acetate (Fig. 3c, black bars). Moreover, the corticalTCA cycling ratio calculated on the basis of glutamatelabeling was significantly reduced by � 30% after treatmentwith the high dose of kainate (Fig. 3d, black bars).

Similar results were obtained from analysis of hippocam-pal extracts (See Figure S2 for details).

Amino acid metabolism in GAD65)/) mice treated withkainateTreatment of GAD65)/) mice with kainate (3.75 mg/kg)led to a decrease of approximately 20% in the cortical

10121416

100012001400

*

*2468

Am

ou

nt

(µm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

200400600800

**

0 0

700

300

400

500

600

*

0.5

0.6

0.4

0.3

0

100

200

* TC

A c

yclin

g r

atio

0.2

0.0

0.1

Glutamate Glutamine GABA [4-13C]Glutamate

[4-13C]Glutamine

[2-13C]GABA

[4,5-13C]Glutamate

[4,5-13C]Glutamine

[1,2-13C]GABA

(a) (b)

(c) (d)

Fig. 2 Amino acid metabolism in cerebral cortex of GAD65+/+ mice

(white bars) and GAD65)/) mice (black bars). (a) The amounts of

glutamate, glutamine and GABA in extracts from cerebral cortex were

determined by 1H-NMR spectroscopy (Results from Walls et al. 2010).

(b) The amounts of [4-13C]glutamate, [4-13C]glutamine and

[2-13C]GABA labeled from [1-13C]glucose (Recalculated from Walls

et al. 2010). (c) The amounts of [4,5-13C]glutamate, [4,5-13C]gluta-

mine and [1,2-13C]GABA labeled from [1,2-13C]acetate (Recalculated

from Walls et al. 2010). (d) TCA cycling ratio calculated on the basis of

glutamate isotopomers labeled from [1-13C]glucose (see text for de-

tails, results from Walls et al. 2010). Results are averages ± SEM

(n = 5–6) and the asterisk indicates a statistically significant difference

between results obtained from GAD65+/+ mice and GAD65)/) mice.

NaCl and kainate treated animals (p < 0.05).

� 2010 The AuthorsJournal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408

GAD65 synthesizes GABA crucial for tonic inhibition | 1403

glutamine content while the contents of glutamate andGABA were unaltered (Fig. 4a). When GAD65)/) micewere treated with kainate, the amount of 13C labeledglutamine in cerebral cortex was decreased regardless ofthe origin of the label, that is, from [1-13C]glucose or[1,2-13C]acetate (Fig. 4b and c). In contrast, the amountsof glutamate and GABA labeled from [1,2-13C]acetate or

[1-13C]glucose were unaltered after treatment with kainate(Fig. 4b and c). The TCA cycling ratio in cerebral cortexcalculated on the basis of glutamate labeling was signif-icantly increased by � 25% after treatment with kainate(Fig. 4d).

Similar results were obtained from analysis of hippocam-pal extracts (See Figure S3 for details).

101214161820

100012001400

*

**

*

* *

** *

*

2468

Am

ou

nt

(µm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

200400600800

0 0

700

300400500600 0.5

0.6

0.4

0.3

0100200

TC

A c

yclin

g r

atio

0.2

0.0

0.1

Glutamate Glutamine GABA [4-13C]Glutamate

[4-13C]Glutamine

[2-13C]GABA

[4,5-13C]Glutamate

[4,5-13C]Glutamine

[1,2-13C]GABA

(a) (b)

(c)(d)

Fig. 3 Amino acid metabolism in cerebral cortex of GAD65+/+ mice

(white bars), GAD65+/+ mice treated with 3.75 mg/kg kainate (grey

bars) and GAD65+/+ mice treated with 15 mg/kg kainate (black bars).

(a) The amounts of glutamate, glutamine and GABA in extracts from

cerebral cortex were determined by 1H-NMR spectroscopy. (b) The

amounts of [4-13C]glutamate, [4-13C]glutamine and [2-13C]GABA

labeled from [1-13C]glucose. (c) The amounts of [4,5-13C]glutamate,

[4,5-13C]glutamine and [1,2-13C]GABA labeled from [1,2-13C]acetate.

(d) TCA cycling ratio calculated on the basis of glutamate isotopomers

labeled from [1-13C]glucose (see text for details). Results are aver-

ages ± SEM (n = 3–8) and the asterisk indicates a statistically sig-

nificant difference between NaCl and kainate treated animals

(p < 0.05).

10121416 1000

900

*

*

*

2468

Am

ou

nt

(µm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

Am

ou

nt

(nm

ol/g

tis

sue)

200100

300400500600700800

0 0

700

300400500600

0.5

0.4

0.3

0100200

TC

A c

yclin

g r

atio

0.2

0.0

0.1

Glutamate Glutamine GABA [4-13C]Glutamate

[4-13C]Glutamine

[2-13C]GABA

[4,5-13C]Glutamate

[4,5-13C]Glutamine

[1,2-13C]GABA

(a) (b)

(c) (d)

Fig. 4 Amino acid metabolism in cerebral cortex of GAD65)/) mice

(white bars) and GAD65)/) mice treated with 3.75 mg/kg kainate

(black bars). (a) The amounts of glutamate, glutamine and GABA in

extracts from cerebral cortex were determined by 1H-NMR spectros-

copy. (b) The amounts of [4-13C]glutamate, [4-13C]glutamine and

[2-13C]GABA labeled from [1-13C]glucose. (c) The amounts of

[4,5-13C]glutamate, [4,5-13C]glutamine and [1,2-13C]GABA labeled

from [1,2-13C]acetate. (d) TCA cycling ratio calculated on the basis of

glutamate isotopomers labeled from [1-13C]glucose (see text for de-

tails). Results are averages ± SEM (n = 5–6) and the asterisk indi-

cates a statistically significant difference between NaCl and kainate

treated animals (p < 0.05).

Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

1404 | A. B. Walls et al.

Discussion

Synthesis of GABA from glutamate is catalyzed by the twoisoforms of GAD, that is, GAD65 and GAD67. Theseisoforms differ with regard to intracellular localization andbinding of the co-factor pyridoxal phosphate and thus havebeen assigned distinct functional roles (Esclapez et al. 1994).The lower GABA content observed in GAD65)/) mice is anobvious consequence of the lack of GAD65 in these mice andis in agreement with previous observations (Stork et al. 2000;Walls et al. 2010). Moreover, the fact that the decrease in[2-13C]GABA in GAD65)/) mice labeled from [1-13C]glu-cose is similar to that of its precursors [4-13C]glutamate and[4-13C]glutamine implies general decrements in neuronalglucose metabolism in GAD65)/) mice. In contrast, astro-cytic metabolism is unaffected in GAD65)/) mice asobserved by similar amounts of [4,5-13C]glutamine and[4,5-13C]glutamate labeled from the astrocyte specific sub-strate [1,2-13C]acetate. However, the conversion of astrocytic[4,5-13C]glutamine to [1,2-13C]GABA is significantly re-duced in GAD65)/) mice (Walls et al. 2010). This indicatesthat GAD65 is particularly important for synthesis of GABAfrom astrocytic glutamine, a process likely confined to thesynaptic area (Behar and Rothman 2001, Lebon et al. 2002).Also, in line with the synaptic localization of GAD65, severalstudies have provided evidence for a significant importance ofthis isoform for the maintenance of sustained synaptic, phasictransmission (Hensch et al. 1998; Tian et al. 1999; Choiet al. 2002). However, the present study was designed toinvestigate the functional importance of GAD65 for mainte-nance of GABA involved in the tonic components ofinhibition mediated extrasynaptically. The frequency of SEDsmeasured in the cortical wedge preparation was used tomonitor the role of GAD65 in the synthesis of GABAdestined for tonic inhibition. In line with the findings of Kuboet al. (2009), we find a decreased tonic inhibition inGAD65)/) mice as the inhibitory action of isoguvacine onSEDs in the cortical wedge preparation was diminished,likely because of a reduced amount of endogenous GABAacting extrasynaptically. This emphasizes the concept thatGABA mediating extrasynaptic tonic inhibition originatesfrom synaptic spillover presupposing the synaptic localizationof GAD65. Moreover, it verifies the hypothesis proposed byStork et al. (2000) that GABA synthesized via GAD65 maybe crucial for tonic inhibition. Hence, in addition to a role ofGAD65 in synthesis of GABA during sustained high intensitystimulation as demonstrated repeatedly (Hensch et al. 1998;Tian et al. 1999; Choi et al. 2002), GAD65 is important formaintenance of tonic inhibition, which has been estimated toconstitute a prominent part of GABAergic transmission(Brickley et al. 1996; Walker and Semyanov 2008).

Disturbances in tonic inhibition are likely associated withepileptogenesis as it has been demonstrated that regulation ofthe extrasynaptic availability of GABA is important for

seizure activity (Galvan et al. 2005; Keros and Hablitz 2005;Madsen et al. 2009, 2010). This is supported by anaugmented tonic inhibition observed following exposure ofhippocampal slices to kainate at concentrations giving rise toseizures in vivo (Ben-Ari and Cossart 2000; Semyanov2004). The increase in tonic inhibition has been proposed tobe a result of intensified synaptic GABA release and synapticspillover induced by activation of pre-synaptic kainatereceptors (Ben-Ari and Cossart 2000). The differentialdistribution of kainate receptors located pre- and post-synaptically throughout a heterogeneous synaptic networkcomprising both excitatory and inhibitory neurons suggests acomplex mechanism of action of kainate involving multiplemodifications. Such complexity of action is underlined by thedifferent effects of kainate observed in vitro, varyingaccording to brain region, properties of the synapse andmeasured parameters (Ben-Ari and Cossart 2000; Lerma2003; Semyanov 2004). In the present study, seizure activitywas induced after administration of kainate. However, thesimultaneous observation of a 30% reduction in the calcu-lated TCA cycle activity in GAD65+/+ mice, that is,hypometabolism, indicates that inhibition is a major part ofthe mechanism of action of kainate in vivo. Such rationale isbased upon the correlation between oxidative glucosemetabolism and the level of neuronal activity (Sibson et al.1998a,b; Dienel and Hertz 2001; Patel et al. 2004; Hyderet al. 2006; Gjedde 2007; Riera et al. 2008). The finding thatkainate exposure is associated with an augmented tonicinhibition (Ben-Ari and Cossart 2000; Semyanov 2004) maysuggest that this form of inhibition mediates the observedneuronal hypometabolism. Interestingly, kainate administra-tion did not induce hypometabolism in GAD65)/) mice,displaying deficits in tonic inhibition. Thus, we suggest thatthe hypometabolism observed in GAD65+/+ mice is med-iated via an augmented tonic inhibition induced by GABAsynthesized via GAD65. Actually, the calculated neuronalTCA cycle activity was increased by approximately 30% inthe GAD65)/) mice upon injection of kainate which islikely because of a lower amount of GABA acting extrasy-naptically and impaired tonic inhibition in GAD65)/) mice.Such impaired tonic inhibition is in line with a highersusceptibility to seizures observed in these mice (Asada et al.1996; Kash et al. 1997) and validate the importance ofGAD65-mediated GABA synthesis for tonic inhibition andseizure activity (Stork et al. 2000). Moreover, these findingsemphasize the association between tonic inhibition andmetabolic activity. It should, however, be taken into consid-eration that the use of null mutant mice may be hampered bycompensatory mechanism masking the functional importanceof the deleted gene and its transcriptional product. This maybe particularly important in the present context as epilepticseizures are known to induce major changes in brain circuitryand hence could have an effect on metabolic pathwaysrelevant for this.

� 2010 The AuthorsJournal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408

GAD65 synthesizes GABA crucial for tonic inhibition | 1405

In addition to the distribution of a-amino-3-hydroxy-5-methylisoxazole-4-propionate and kainate receptors through-out the neuronal network, these receptor types are alsoexpressed on astrocytes (Porter and McCarthy 1997; Lovicket al. 2005). This suggests that astrocytic metabolism may beaffected by kainate independent of neuronal metabolism andthat the effect of kainate on astrocytic metabolism ismediated directly via receptors located in the astrocyticmembrane. Such mechanism of action is supported bystudies in cultured astrocytes having demonstrated alterationsin ion homeostasis following exposure to kainate (MacVicaret al. 1988; Rose and Ransom 1996).

Uptake of [1,2-13C]acetate and glutamine synthesis areconfined to the astrocytic compartment and the reducedamount of [4,5-13C]glutamine labeled from [1,2-13C]acetatetherefore reflects impairment of astrocytic metabolism (Son-newald et al. 1993a; Waniewski and Martin 1998). Thisfinding is in line with a significant decrease in glutaminesynthetase activity previously reported following stereotaxicapplication of kainate into rat hippocampus and afterexposure of rat coronal slices to kainate (Waniewski andMcFarland 1990; McBean et al. 1995). In contrast to theneuronal hypometabolism observed in GAD65+/+ mice onlyafter treatment with the high dose of kainate, astrocyticmetabolism was affected also after injection of the low doseof kainate and the impaired synthesis of [4,5-13C]glutaminewas not further aggravated when the mice were treated withthe high dose of kainate. Moreover, a lower amount of[4,5-13C]glutamine was also observed in GAD65)/) miceafter treatment with kainate. Hence, hypometabolism med-iated by tonic inhibition seems to be limited to the neuronalcompartment as labeling in [4,5-13C]glutamine and accord-ingly astrocytic metabolism is affected to the same extent inGAD65)/) and GAD65+/+ mice.

In conclusion, we have provided evidence that tonicinhibition is significantly decreased in GAD65)/) mice.Although kainate gives rise to seizure activity in GAD65+/+mice we find that hypometabolism is evoked, which is likelymediated by an augmented tonic inhibition induced byGABA synthesized via GAD65. This is substantiated by theobservation that kainate elicited hypermetabolism inGAD65)/) mice which may be because of the inherentimpaired tonic inhibition leading to increased seizuresusceptibility. Altogether, we have demonstrated a significantrole of GAD65 in the synthesis of GABA destined forextrasynaptic tonic inhibition and hence, for control ofepileptiform activity.

Acknowledgements

This study was supported by grants from Manufacturer Vilhelm

Pedersen and Wife Memorial Legacy, a support granted upon

recommendation from the Novo Nordisk Foundation, as well as the

Novo Nordisk, the Hørslev, and the Lundbeck Foundations. Dr.

Bjarke Ebert, H. Lundbeck & Comp. LTD, is cordially thanked for

fruitful discussions related to the cortical wedge data. The technical

assistance of Ms. Gunilla Steven is gratefully acknowledged. A

special thanks to Claus Rolin Larsen for writing and validating the

data sampling program. The authors declare no conflict of interests.

Supporting information

Additional Supporting information may be found in the online

version of this article:

Figure S1. Amino acid metabolism in hippocampus of

GAD65+/+ mice (white bars) and GAD65)/) mice (black bars).

Figure S2. Amino acid metabolism in hippocampus of

GAD65+/+ mice (white bars), GAD65+/+ mice treated with

3.75 mg/kg kainate (grey bars) and GAD65+/+ mice treated with

15 mg/kg kainate (black bars).

Figure S3. Amino acid metabolism in hippocampus of

GAD65)/) mice (white bars) and GAD65)/) mice treated with

3.75 mg/kg kainate (black bars).

As a service to our authors, and readers, this journal provides

supporting information supplied by the authors. Such materials are

peer-reviewed and may be re-organized for online delivery, but are

not copy-edited or typeset. Technical support issues arising from

supporting information (other than missing files) should be

addressed to the authors.

References

Asada H., Kawamura Y., Maruyama K. et al. (1996) Mice lacking the65 kDa isoform of glutamic acid decarboxylase (GAD65) maintainnormal levels of GAD67 and GABA in their brains but are sus-ceptible to seizures. Biochem. Biophys. Res. Commun. 229, 891–895.

Battaglioli G. and Martin D. L. (1990) Stimulation of synaptosomalgamma-aminobutyric acid synthesis by glutamate and glutamine.J. Neurochem. 54, 1179–1187.

Battaglioli G. and Martin D. L. (1991) GABA synthesis in brain slices isdependent on glutamine produced in astrocytes. Neurochem. Res.16, 151–156.

Behar K. L. and Rothman D. L. (2001) In vivo nuclear magnetic reso-nance studies of glutamategamma-aminobutyric acid-glutaminecycling in rodent and human cortex: the central role of glutamine.J. Nutr. 131, 2498S–2504S.

Ben-Ari Y. (1985) Limbic seizure and brain damage produced by kainicacid: mechanisms and relevance to human temporal lobe epilepsy.Neuroscience 14, 375–403.

Ben-Ari Y. and Cossart R. (2000) Kainate, a double agent that generatesseizures: two decades of progress. Trends Neurosci. 23, 580–587.

Ben-Ari Y., Tremblay E., Riche D., Ghilini G. and Naquet R. (1981)Electrographic, clinical and pathological alterations followingsystemic administration of kainic acid, bicuculline or pentetrazole:metabolic mapping using the deoxyglucose method with specialreference to the pathology of epilepsy. Neuroscience 6, 1361–1391.

Bergado-Acosta J. R., Sangha S., Narayanan R. T., Obata K., Pape H. C.and Stork O. (2008) Critical role of the 65-kDa isoform of glutamicacid decarboxylase in consolidation and generalization of Pav-lovian fear memory. Learn. Mem. 15, 163–171.

Bradford H. F. (1995) Glutamate, GABA and epilepsy. Prog. Neurobiol.47, 477–511.

Bradford H. F., Ward H. K. and Thomas A. J. (1978) Glutamine – amajor substrate for nerve endings. J. Neurochem. 30, 1453–1459.

Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

1406 | A. B. Walls et al.

Brickley S. G., Cull-Candy S. G. and Farrant M. (1996) Development ofa tonic form of synaptic inhibition in rat cerebellar granule cellsresulting from persistent activation of GABAA receptors.J. Physiol. 497 (Pt 3), 753–759.

Choi S. Y., Morales B., Lee H. K. and Kirkwood A. (2002) Absence oflong-term depression in the visual cortex of glutamic Acid decar-boxylase-65 knock-out mice. J. Neurosci. 22, 5271–5276.

Chowdhury G. M., Gupta M., Gibson K. M., Patel A. B. and Behar K. L.(2007) Altered cerebral glucose and acetate metabolism in succinicsemialdehyde dehydrogenase-deficient mice: evidence for glialdysfunction and reduced glutamate/glutamine cycling. J. Neuro-chem. 103, 2077–2091.

Dienel G. A. and Hertz L. (2001) Glucose and lactate metabolism duringbrain activation. J. Neurosci. Res. 66, 824–838.

Ebert B., Storustovu S., Mortensen M. and Frølund B. (2002) Charac-terization of GABAA receptor ligands in the rat cortical wedgepreparation: evidence for action at extrasynaptic receptors? Br. J.Pharmacol. 137, 1–8.

Erlander M. G. and Tobin A. J. (1991) The structural and functionalheterogeneity of glutamic acid decarboxylase: a review. Neuro-chem. Res. 16, 215–226.

Erlander M. G., Tillakaratne N. J., Feldblum S., Patel N. and Tobin A. J.(1991) Two genes encode distinct glutamate decarboxylases.Neuron 7, 91–100.

Esclapez M., Tillakaratne N. J., Kaufman D. L., Tobin A. J. and HouserC. R. (1994) Comparative localization of two forms of glutamicacid decarboxylase and their mRNAs in rat brain supports theconcept of functional differences between the forms. J. Neurosci.14, 1834–1855.

Farrant M. and Nusser Z. (2005) Variations on an inhibitory theme:phasic and tonic activation of GABAA receptors. Nat. Rev. Neu-rosci. 6, 215–229.

Galvan A., Villalba R. M., West S. M., Maidment N. T., Ackerson L. C.,Smith Y. and Wichmann T. (2005) GABAergic modulation of theactivity of globus pallidus neurons in primates: in vivo analysis ofthe functions of GABA receptors and GABA transporters.J. Neurophysiol. 94, 990–1000.

Gjedde A. (2007) Coupling of brain function to metabolism: evaluationof energy requirements, in Handbook of Neurochemistry andMolecular Biology, 3rd Edn (Lajtha A., ed.), pp. 343–400.Springer, Berlin.

Harrison N. L. and Simmonds M. A. (1985) Quantitative studies onsome antagonists of N-methyl D-aspartate in slices of rat cerebralcortex. Br. J. Pharmacol. 84, 381–391.

Hassel B., Bachelard H., Jones P., Fonnum F. and Sonnewald U. (1997)Trafficking of amino acids between neurons and glia in vivo. Ef-fects of inhibition of glial metabolism by fluoroacetate. J. Cereb.Blood Flow Metab. 17, 1230–1238.

Hensch T. K., Fagiolini M., Mataga N., Stryker M. P., Baekkeskov S.and Kash S. F. (1998) Local GABA circuit control of experience-dependent plasticity in developing visual cortex. Science 282,1504–1508.

Hyder F., Patel A. B., Gjedde A., Rothman D. L., Behar K. L. andShulman R. G. (2006) Neuronal-glial glucose oxidation andglutamatergic-GABAergic function. J. Cereb. Blood Flow Metab.26, 865–877.

Kash S. F., Johnson R. S., Tecott L. H., Noebels J. L., Mayfield R. D.,Hanahan D. and Baekkeskov S. (1997) Epilepsy in mice deficientin the 65-kDa isoform of glutamic acid decarboxylase. Proc. NatlAcad. Sci. USA 94, 14060–14065.

Kash S. F., Tecott L. H., Hodge C. and Baekkeskov S. (1999) Increasedanxiety and altered responses to anxiolytics in mice deficient in the65-kDa isoform of glutamic acid decarboxylase. Proc. Natl Acad.Sci. USA 96, 1698–1703.

Kaufman D. L., Houser C. R. and Tobin A. J. (1991) Two forms of thegamma-aminobutyric acid synthetic enzyme glutamate decarbox-ylase have distinct intraneuronal distributions and cofactor inter-actions. J. Neurochem. 56, 720–723.

Keros S. and Hablitz J. J. (2005) Subtype-specific GABA transporterantagonists synergistically modulate phasic and tonic GABAA

conductances in rat neocortex. J. Neurophysiol. 94, 2073–2085.Kim J.-S., Kondratyev A., Tomita Y. and Gale K.. (2007) Neurodevel-

opmental impact of antiepileptic drugs and seizures in the imma-ture brain. Epilepsia 48 (Suppl. 5), 19–26.

Kubo K., Nishikawa K., Ishizeki J., Hardy-Yamada M., Yanagawa Y.and Saito S. (2009) Thermal hyperalgesia via supraspinal mecha-nisms in mice lacking glutamate decarboxylase 65. J. Pharmacol.Exp. Ther. 331, 162–169.

Kvamme E., Torgner I. A. and Roberg B. (2001) Kinetics and locali-zation of brain phosphate activated glutaminase. J. Neurosci. Res.66, 951–958.

Laake J. H., Takumi Y., Eidet J., Torgner I. A., Roberg B., Kvamme E.and Ottersen O. P. (1999) Postembedding immunogold labellingreveals subcellular localization and pathway-specific enrichment ofphosphate activated glutaminase in rat cerebellum. Neuroscience88, 1137–1151.

Lebon V., Petersen K. F., Cline G. W., Shen J., Mason G. F., Dufour S.,Behar K. L., Shulman G. I. and Rothman D. L. (2002) Astroglialcontribution to brain energy metabolism in humans revealed by 13Cnuclear magnetic resonance spectroscopy: elucidation of thedominant pathway for neurotransmitter glutamate repletion andmeasurement of astrocytic oxidative metabolism. J. Neurosci. 22,1523–1531.

Lerma J. (2003) Roles and rules of kainate receptors in synaptic trans-mission. Nat. Rev. Neurosci. 4, 481–495.

Lothman E. W. and Collins R. C. (1981) Kainic acid induced limbicseizures: metabolic, behavioral, electroencephalographic andneuropathological correlates. Brain Res. 218, 299–318.

Lovick T. A., Brown L. A. and Key B. J. (2005) Neuronal activity-related coupling in cortical arterioles: involvement of astrocyte-derived factors. Exp. Physiol. 90, 131–140.

MacVicar B. A., Baker K. and Crichton S. A. (1988) Kainic acidevokes a potassium efflux from astrocytes. Neuroscience 25,721–725.

Madsen K. K., Clausen R. P., Larsson O. M., Krogsgaard-Larsen P.,Schousboe A. and White H. S. (2009) Synaptic and extrasynapticGABA transporters as targets for anti-epileptic drugs. J. Neuro-chem. 109 Suppl 1, 139–144.

Madsen K. K., White H. S. and Schousboe A. (2010) Neuronal and non-neuronal GABA transporters as targets for antiepileptic drugs.Pharmacol. Ther. 125, 394–401.

Mawhinney T. P., Robinett R. S., Atalay A. and Madson M. A. (1986)Gas-liquid chromatography and mass spectral analysis of mono-,di- and tricarboxylates as their tert.-butyldimethylsilyl derivatives.J. Chromatogr. 361, 117–130.

McBean G. J., Doorty K. B., Tipton K. F. and Kollegger H. (1995)Alteration in the glial cell metabolism of glutamate by kainate andN-methyl-D-aspartate. Toxicon 33, 569–576.

Patel A. B., de Graaf R. A., Mason G. F., Kanamatsu T., Rothman D. L.,Shulman R. G. and Behar K. L. (2004) Glutamatergic neuro-transmission and neuronal glucose oxidation are coupled duringintense neuronal activation. J. Cereb. Blood Flow Metab. 24, 972–985.

Paulsen R. E., Odden E. and Fonnum F. (1988) Importance of glutaminefor gamma-aminobutyric acid synthesis in rat neostriatum in vivo.J. Neurochem. 51, 1294–1299.

Pinheiro P. and Mulle C. (2006) Kainate receptors. Cell Tissue Res. 326,457–482.

� 2010 The AuthorsJournal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408

GAD65 synthesizes GABA crucial for tonic inhibition | 1407

Porter J. T. and McCarthy K. D. (1997) Astrocytic neurotransmitterreceptors in situ and in vivo. Prog. Neurobiol. 51, 439–455.

Qu H., Haberg A., Haraldseth O., Unsgard G. and Sonnewald U. (2000)13C MR spectroscopy study of lactate as substrate for rat brain.Dev. Neurosci. 22, 429–436.

Racine R. J. (1972) Modification of seizure activity by electrical stim-ulation II. Motor seizure. Electroencephalogr. Clin. Neurophysiol.32, 281–294.

Rae C., Hare N., Bubb W. A., McEwan S. R., Broer A., McQuillan J. A.,Balcar V. J., Conigrave A. D. and Broer S. (2003) Inhibition ofglutamine transport depletes glutamate and GABA neurotransmit-ter pools: further evidence for metabolic compartmentation.J. Neurochem. 85, 503–514.

Richerson G. B. (2004) Looking for GABA in all the wrong places: therelevance of extrasynaptic GABAA receptors to epilepsy. EpilepsyCurr. 4, 239–242.

Riera J. J., Schousboe A., Waagepetersen H. S., Howarth C. and HyderF. (2008) The micro-architecture of the cerebral cortex: Functionalneuroimaging models and metabolism. Neuroimage 40, 1436–1459.

Rose C. R. and Ransom B. R. (1996) Mechanisms of H+ and Na+

changes induced by glutamate, kainate, and D-aspartate in rathippocampal astrocytes. J. Neurosci. 16, 5393–5404.

Rossi D. J. and Hamann M. (1998) Spillover-mediated transmission atinhibitory synapses promoted by high affinity alpha6 subunitGABAA receptors and glomerular geometry. Neuron 20, 783–795.

Schousboe A. and Waagepetersen H. S. (2006) Glial modulation ofGABAergic and glutamatergic neurotransmission. Curr. Top. Med.Chem. 6, 929–934.

Schousboe A. and White H. S. (2009) Glial modulation of excitabilityvia glutamate and GABA transporters, in Encyclopedia of BasicEpileptic Research (Schwartzkroin P. A., ed.), pp. 397–401. Aca-demic Press, Oxford.

Schousboe A., Sarup A., Bak L. K., Waagepetersen H. S. and Larsson O.M. (2004) Role of astrocytic transport processes in glutamatergicand GABAergic neurotransmission. Neurochem. Int. 45, 521–527.

Semyanov A. V. (2004) The effects of activation of kainate receptors ontonic and phasic gabaergic inhibition in interneurons in field CA1of guinea pig hippocampus slices. Neurosci. Behav. Physiol. 34,123–130.

Shen J., Petersen K. F., Behar K. L., Brown P., Nixon T. W., MasonG. F., Petroff O. A., Shulman G. I., Shulman R. G. and Rothman D.L. (1999) Determination of the rate of the glutamate/glutaminecycle in the human brain by in vivo 13C NMR. Proc. Natl Acad.Sci. USA 96, 8235–8240.

Sibson N. R., Dhankhar A., Mason G. F., Rothman D. L., Behar K. L.and Shulman R. G. (1998a) Stoichiometric coupling of brain glu-cose metabolism and glutamatergic neuronal activity. Proc. NatlAcad. Sci. USA 95, 316–321.

Sibson N. R., Shen J., Mason G. F., Rothman D. L., Behar K. L. andShulman R. G. (1998b) Functional energy metabolism: in vivo

13C-NMR spectroscopy evidence for coupling of cerebral glucoseconsumption and glutamatergic neuronalactivity. Dev. Neurosci.20, 321–330.

Sonnewald U., Westergaard N., Petersen S. B., Unsgard G. and Scho-usboe A. (1993a) Metabolism of [U-13C]glutamate in astrocytesstudied by 13C NMR spectroscopy: incorporation of more labelinto lactate than into glutamine demonstrates the importance of thetricarboxylic acid cycle. J. Neurochem. 61, 1179–1182.

Sonnewald U., Westergaard N., Schousboe A., Svendsen J. S., UnsgardG. and Petersen S. B. (1993b) Direct demonstration by [13C]NMRspectroscopy that glutamine from astrocytes is a precursor forGABA synthesis in neurons. Neurochem. Int. 22, 19–29.

Stork O., Ji F. Y., Kaneko K., Stork S., Yoshinobu Y., Moriya T., ShibataS. and Obata K. (2000) Postnatal development of a GABA deficitand disturbance of neural functions in mice lacking GAD65. BrainRes. 865, 45–58.

Storustovu S. and Ebert B. (2003) Gaboxadol: in vitro interaction studieswith benzodiazepines and ethanol suggest functional selectivity.Eur. J. Pharmacol. 467, 49–56.

Tian N., Petersen C., Kash S., Baekkeskov S., Copenhagen D. and NicollR. (1999) The role of the synthetic enzyme GAD65 in the controlof neuronal gamma-aminobutyric acid release. Proc. Natl Acad.Sci. USA 96, 12911–12916.

Waagepetersen H. S., Sonnewald U. and Schousboe A. (1999) TheGABA paradox: multiple roles as metabolite, neurotransmitter, andneurodifferentiative agent. J. Neurochem. 73, 1335–1342.

Waagepetersen H. S., Sonnewald U., Gegelashvili G., Larsson O. M. andSchousboe A. (2001) Metabolic distinction between vesicular andcytosolic GABA in cultured GABAergic neurons using 13C mag-netic resonance spectroscopy. J. Neurosci. Res. 63, 347–355.

Walker M. C. and Semyanov A. (2008) Regulation of excitability byextrasynaptic GABAA receptors. Results Probl. Cell Differ. 44, 29–48.

Walls A. B., Eyjolfsson E. M., Smeland O. B., Nilsen L. H., SchousboeI., Schousboe A., Sonnewald U. and Waagepetersen H. S. (2010)Knockout of GAD65 has major impact on synaptic GABA syn-thesized from astrocyte-derived glutamine. J. Cereb. Blood FlowMetab. Epub ahead of print.

Waniewski R. A. and Martin D. L. (1998) Preferential utilization ofacetate by astrocytes is attributable to transport. J. Neurosci. 18,5225–5233.

Waniewski R. A. and McFarland D. (1990) Intrahippocampal kainic acidreduces glutamine synthetase. Neuroscience 34, 305–310.

Wu Y., Wang W., ez-Sampedro A. and Richerson G. B. (2007) Non-vesicular inhibitory neurotransmission via reversal of the GABAtransporter GAT-1. Neuron 56, 851–865.

Zeng L. H., Xu L., Rensing N. R., Sinatra P. M., Rothman S. M. andWong M. (2007) Kainate seizures cause acute dendritic injury andactin depolymerization in vivo. J. Neurosci. 27, 11604–11613.

Journal of Neurochemistry � 2010 International Society for Neurochemistry, J. Neurochem. (2010) 115, 1398–1408� 2010 The Authors

1408 | A. B. Walls et al.