European Journal of Neuroscience, Vol. 18, pp. 2722±2730, 2003
ß Federation of European Neuroscience Societies
Speci®c g-hydroxybutyrate-binding sites but loss of
pharmacological effects of g-hydroxybutyrate in
GABAB(1)-de®cient mice
Klemens Kaupmann,1 John F. Cryan,1 Petrine Wellendorph,2 Cedric Mombereau,1 Gilles Sansig,1 Klaus Klebs,1
Markus Schmutz,1 Wolfgang Froestl,1 Herman van der Putten,1 Johannes Mosbacher,1 Hans BraÈuner-Osborne,2
Peter Waldmeier1 and Bernhard Bettler3
1
Novartis Institutes for BioMedical Research, WKL-125.7.42, Novartis Pharma AG, CH-4002 Basel, Switzerland
Department of Medicinal Chemistry, Danish University of Pharmaceutical Sciences, Copenhagen, Denmark
3
Pharmazentrum, Department of Clinical-Biological Sciences, Klingelbergstr. 50-70, CH-4056 Basel, Switzerland
2
Keywords: delta waves, dopamine, drug of abuse, hypolocomotion, hypothermia
Abstract
g-Hydroxybutyrate (GHB), a metabolite of g-aminobutyric acid (GABA), is proposed to function as a neurotransmitter or neuromodulator. g-Hydroxybutyrate and its prodrug, g-butyrolactone (GBL), recently received increased public attention as they emerged as
popular drugs of abuse. The actions of GHB/GBL are believed to be mediated by GABAB and/or speci®c GHB receptors, the latter
corresponding to high-af®nity [3H]GHB-binding sites coupled to G-proteins. To investigate the contribution of GABAB receptors to GHB
actions we studied the effects of GHB in GABAB(1) / mice, which lack functional GABAB receptors. Autoradiography reveals a similar
spatial distribution of [3H]GHB-binding sites in brains of GABAB(1) / and wild-type mice. The maximal number of binding sites and the
KD values for the putative GHB antagonist [3H]6,7,8,9-tetrahydro-5-hydroxy-5H-benzocyclohept-6-ylidene acetic acid (NCS-382)
appear unchanged in GABAB(1) / compared with wild-type mice, demonstrating that GHB- are distinct from GABAB-binding sites. In
the presence of the GABAB receptor positive modulator 2,6-di-tert-butyl-4-(3-hydroxy-2,2-dimethyl-propyl)-phenol GHB induced
functional GTPg[35S] responses in brain membrane preparations from wild-type but not GABAB(1) / mice. The GTPg[35S] responses
in wild-type mice were blocked by the GABAB antagonist [3-[[1-(S)-(3,4dichlorophenyl)ethyl]amino]-2-(S)-hydroxy-propyl]-cyclohexylmethyl phosphinic acid hydrochloride (CGP54626) but not by NCS-382. Altogether, these ®ndings suggest that the GHB-induced
GTPg[35S] responses are mediated by GABAB receptors. Following GHB or GBL application, GABAB(1) / mice showed neither the
hypolocomotion, hypothermia, increase in striatal dopamine synthesis nor electroencephalogram delta-wave induction seen in wildtype mice. It, therefore, appears that all studied GHB effects are GABAB receptor dependent. The molecular nature and the signalling
properties of the speci®c [3H]GHB-binding sites remain elusive.
Introduction
g-Hydroxybutyrate (GHB), a metabolite of g-aminobutyric acid
(GABA), is present at micromolar concentration in the brain. While
the physiological role of GHB is unclear, it is proposed to function as
a neurotransmitter or neuromodulator (Cash, 1994; Maitre, 1997;
Bernasconi et al., 1999; Maitre et al., 2000). Patients suffering from
GHB aciduria, a congenital enzyme defect causing GHB accumulation, exhibit psychomotor retardation, delayed or absent speech,
hypotonia, ataxia, hypore¯exia, seizures and electroencephalogram
(EEG) abnormalities (Hogema et al., 2001). g-Hydroxybutyrate is
clinically used as an anaesthetic adjuvant (Kleinschmidt et al., 1999)
and was recently approved by the Food and Drug Administration of the
USA for the treatment of narcolepsy (Tunnicliff & Raess, 2002).
Furthermore, bene®cial effects of GHB were described in the treatment of alcoholism (Addolorato et al., 1998) and opiate dependency
(Gallimberti et al., 1994, 2000). On the other hand, the mood-elevating, sedative, relaxing and anabolic properties of GHB lead to its illicit
Correspondence: Dr Klemens Kaupmann, as above.
E-mail: [email protected]
Received 18 July 2003, revised 3 September 2003, accepted 10 September 2003
doi:10.1046/j.1460-9568.2003.03013.x
use and abuse (Nicholson & Balster, 2001). In view of these therapeutic prospects and public health concerns, it is important to understand the mechanism of action of GHB.
The receptor interactions of GHB are a matter of much debate, in
particular the relation of the putative GHB receptor with the metabotropic GABAB receptor. Several lines of evidence support the idea that
native [3H]GHB-binding sites and GABAB receptors are distinct.
Firstly, the distribution and ontogenesis of [3H]GHB-binding sites
and GABAB receptors are different (Snead, 1994; Castelli et al., 2000).
Secondly, the putative GHB receptor antagonist 6,7,8,9-tetrahydro-5hydroxy-5H-benzocyclohept-6-ylidene acetic acid (NCS-382) has no
af®nity for GABAB receptors (Maitre et al., 1990; Mehta et al., 2001).
It was argued that GABAB receptor-mediated effects of GHB may be
secondary to a (i) conversion of GHB into GABA or (ii) GHB-induced
stimulation of GABA release (Hechler et al., 1997). These properties
of GHB actions would increase synaptic levels of GABA which, in
turn, would activate GABAB receptors. There is increasing evidence
that GABAB receptors mediate at least some effects of exogenously
applied GHB or its prodrug g-butyrolactone (GBL) (Waldmeier, 1991;
Xie & Smart, 1992; Williams et al., 1995; Nissbrandt & Engberg,
1996; Colombo et al., 1998; Erhardt et al., 1998; Madden & Johnson,
Loss of g-hydroxybutyrate action in GABAB(1) /
1998; Carai et al., 2001; Jensen & Mody, 2001). On the other hand,
distinct high-af®nity GHB receptors, which are probably related to
brain [3H]GHB-binding sites, are expected to mediate the signalling of
endogenous GHB. There are reports to suggest that these [3H]GHBbinding sites are coupled to G-proteins (Ratomponirina et al., 1995;
Snead et al., 2000).
Functional GABAB receptors assemble from two subunits,
GABAB(1) and GABAB(2) (Marshall et al., 1999). Accordingly,
GABAB(1) / mice lack any detectable pre- or postsynaptic GABAB
responses (Prosser et al., 2001; Schuler et al., 2001). GABAB(1) /
mice, therefore, provide the opportunity to study the effects of GHB in
the absence of coincident GABAB receptor responses. Here we
analysed the biochemical and behavioural effects of GHB in
GABAB(1) / mice in order to clarify which effects can be speci®cally attributed to GHB receptors.
Materials and methods
Animals
The BALB/c GABAB(1) knockout mice were described previously
(Schuler et al., 2001). Mice were used at an age of 10±17 weeks.
Additional male BALB/c mice (23±26 g) were obtained from Iffo
Credo (France). Housing was at room temperature, in a 12-h light/dark
cycle with lights on at 6 a.m. Food pellets and tap water were available
ad libitum. All behavioural experiments were conducted during the
light cycle. All animal experiments were subject to institutional review
and conducted in accordance with the Veterinary Authority of BaselStadt, Switzerland.
Membrane preparations
Mice were decapitated and the brains removed. The cortex was
dissected and immediately frozen in liquid nitrogen. For membrane
preparations frozen tissues were thawed in ice-cold 0.32 M sucrose and
homogenized using an UltraTurrax homogenizer. Crude membranes
were prepared according to the methods of Ransom & Stec (1988) or
Urwyler et al. (2001) and stored at 80 8C until use.
Compounds and radioligands
g-Hydroxybutyrate sodium salt and GBL were purchased from
Sigma (St Louis, MO, USA) and NCS-382 was from Tocris (Bristol,
UK). 2,6-Di-tert-butyl-4-(3-hydroxy-2,2-dimethyl-propyl)-phenol
(CGP7930) and [3-[[1-(S)-(3,4dichlorophenyl)ethyl]amino]-2-(S)hydroxy-propyl]-cyclohexylmethyl phosphinic acid hydrochloride
(CGP54626) were synthesized at Novartis. [3H]NCS-382 (20 Ci/
mmol) and [3H]GHB (20 Ci/mmol; g-hydroxybutyric acid, sodium
salt) were obtained from ARC (St Louis, MO, USA). GTPg[35S]
(c. 1000 Ci/mmol) was purchased from Amersham Biosciences
(Freiburg, Germany).
Ligand-binding assay
The [3H]NCS-382-binding assay described by Mehta et al. (2001) was
adapted to a 96-well ®ltration assay format. Potassium phosphate
(50 mM; pH 6.0) was used as binding buffer (50±70 mg of protein were
used per data point). The membranes were thawed, homogenized in 40
volumes of buffer and centrifuged at 48 000 g for 15 min at 4 8C. The
resulting pellet was homogenized in buffer and centrifuged at 48 000 g
for 15 min. This washing step was repeated three times and the ®nal
pellet was resuspended in binding buffer. For saturation binding,
[3H]NCS-382 concentrations ranged between 1 and 2000 nM whereas
for competition studies, 16 nM [3H]NCS-382 was used. Aliquots
(200 mL), in triplicate, were incubated for 1 h at 0±4 8C. The binding
reaction was terminated by rapid ®ltration through GF/C uni®lters,
mice
2723
using a 96-well FilterMate cell harvester (Packard), followed by three
washes with ice-cold buffer (200 mL). Nonspeci®c binding was determined in the presence of 1 mM GHB. The amount of [3H]NCS-382
bound to membranes was determined using a TOPCOUNT microplate
scintillation counter (Packard). Data were analysed by a nonlinear
regression curve-®tting procedure using the computer program PRISM
(Graphpad Prism 3.0; GraphPad Software Inc., San Diego, CA, USA).
Autoradiography
Cryostat sagittal brain sections (10 mm) were mounted on poly-Llysine slides (Electron Microscopy Sciences, Fort Washington, PA,
USA). The sections were preincubated for 30 min at 4 8C in 100 mM
KH2PO4, pH 6, followed by an incubation with 30 nM [3H]GHB in
buffer for 30 min at 4 8C. Nonspeci®c binding was determined in the
presence of 10 mM GHB. Slides were washed three times for 10 s in
ice-cold buffer, rinsed in H2O and air-dried. X-ray ®lms (hyper®lm
RPN535B; Amersham Life Sciences, Freiburg, Germany) were
exposed for 3±6 weeks. For better visualization gray values were
converted into pseudocolours. Bound radioactivity was calculated
from optical densities of gray values using [3H]microscales (Amersham) as calibration markers.
GTPg[35S] binding
The assay mixtures contained 50 mM Tris-HCl buffer, pH 7.7, 10 mM
MgCl2, 1.8 mM CaCl2, 100 mM NaCl, 30 mM GDP (Sigma), 20 mg of
membrane protein, 0.2 nM GTPg[35S] and test compounds. An alternative buffer system, as described by Snead (2000), was also investigated. Pico-plates (96-well, 300 mL volume; Packard) were used.
Nonspeci®c binding was measured in the presence of unlabelled
guanosine 50 -O-(3-thiotriphosphate) (10 mM; Sigma). The reagents
were incubated for 40 min at room temperature and subsequently
®ltered (uni®lter-GF/C; Packard) using a cell harvester (Filtermate;
Packard). After two washes with ice-cold assay buffer the plates were
dried for 1 h at 50 8C, 50 mL scintillation solution (Microscint 20;
Packard) were added and the plates counted (Topcount NXT; Packard).
For data analysis, nonspeci®c binding was subtracted from all of the
other values and the compound effects were expressed relative to basal
activity. Prism 3.0 software (GraphPad Software Inc.) was used for all
data calculations.
Measurement of locomotor activity
Sixty minutes after the administration of GHB or vehicle, animals
were placed in automated locomotor activity cages (31 19 16 cm;
TSE, Bad Homburg, Germany) and the distanced traveled was measured by the number of horizontal beam-breaks as previously
described (Spooren et al., 2000). Test sessions were for 60 min and
data were collected in 10-min intervals.
Measurement of core body temperature
Rectal temperature was measured to the nearest 0.1 8C by a thermometer (model DM 852; ELLAB Instruments; Copenhagen, Denmark)
by inserting a lubricated thermistor probe (model PRA-22002-A,
2.2 mm diameter; ELLAB Instruments) 20 mm into the rectum. The
mouse was hand held at the base of the tail and the thermistor probe
was left in place for 15 s. Statistical analysis was carried out using
analysis of variance (repeated measures) followed by Dunnett's tests or
Fisher's LSD tests where appropriate.
g-Butyrolactone-induced striatal dopamine synthesis
Mice were treated with 750 mg/kg i.p. GBL or 20 mg/kg i.p. baclofen
(p-chloro-beta-phenyl-GABA) followed 5 min later by 100 mg/kg i.p.
3-hydroxy-benzylhydrazine dihydrochloride (NSD1015; Sandev Ltd,
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
2724 K. Kaupmann et al.
Harlow, Essex, UK). The animals were killed 45 min after NSD1015
application by decapitation and striata were dissected out. Accumulation of 3,4-dihydroxyphenylacetic acid (DOPA) after central decarboxylase inhibition by NSD1015 45 min prior to killing of the animals
was determined by HPLC with electrochemical detection. Sample
preparation and HPLC conditions were as described by Waldmeier
(1991), except that electrochemical detection was amperometric as
described by Waldmeier et al. (1984). a-Methyl-DOPA (100 ng/
extract) was used as an internal standard. The retention times of
DOPA and a-methyl-DOPA were 6.0 and 8.5 min, respectively. The
accumulation of DOPA in the mouse striata was found beforehand to
be linear for at least 60 min after injection of 100 mg/kg NSD1015.
2001). Cortex membrane preparations derived from this tissue were,
therefore, used to explore the binding characteristics of the GHB
antagonist [3H]NCS-382 (Fig. 1B and C). g-Hydroxybutyrate and
NCS-382 displaced [3H]NCS-382 radioligand binding with similar
potency in wild-type and GABAB(1) / mice membranes. Saturation
isotherms with [3H]NCS-382 further corroborated that GABAB(1) does
not contribute signi®cantly to [3H]GHB binding in brain. Similar KD
values and maximal number of binding sites (Bmax) levels were determined using membranes prepared from wild-type and GABAB(1) /
mice (Fig. 1C). In summary, the regional distribution and pharmacological characteristics of [3H]GHB- as well as [3H]NCS-382-binding
sites were strikingly similar in wild-type and GABAB(1) / mice.
Electroencephalogram measurements
A three-pole socket was implanted over the cortex and embedded in
dental cement under anaesthesia (3 mL/kg i.p. Hypnorm, 5 mg/kg i.p.
diazepam, 30 mg/kg; s.c. buprenorphine). Bipolar leads from the mice
were recorded via cables connected to a slip-ring system. The behaviour of the animals, which were housed singly in wooden observation
cages, was observed over a closed-circuit TV system starting 21 days
after the operation. The EEGs were ampli®ed (EEG-2104; Spectralab),
recorded on a thermo recorder (MTK95; Astromed), and collected on a
personal computer.
Results
[3H]g-Hydroxybutyrate- and [3H]NCS-382-binding sites are
retained in GABAB(1) / mice
We used autoradiography and ligand-binding assays on cerebral cortex
membranes to study the distribution of [3H]GHB- and [3H]NCS-382binding sites in wild-type and GABAB(1) / mice (Fig. 1). Two
binding sites for [3H]GHB have been described in the brain, one of
high (KD, 30±90 nM) and one of low af®nity (KD, 2±16 mM) (Maitre
et al., 2000). Receptor autoradiography using 30 nM [3H]GHB probably reveals only the high-af®nity binding sites. High levels of speci®c
binding were detected in cortical layers, hippocampus and, to a lesser
extent, midbrain regions (Fig. 1A). In agreement with previous studies,
cerebellum and brain stem showed no speci®c GHB binding (Snead,
1994; Maitre, 1997; Castelli et al., 2000). The overall spatial distribution of GHB-binding sites appeared similar in wild-type, GABAB(1)/±
and GABAB(1) / mice, although differences in the abundance of
binding sites in speci®c brain regions cannot be excluded. As
[3H]GHB-binding sites are present in brains of GABAB(1) / mice
we concluded that [3H]GHB-binding sites are distinct from and
independent of GABAB(1).
The cerebral cortex was previously shown to contain a high density
of GHB-binding sites (Castelli et al., 2000; Snead, 2000; Mehta et al.,
Fig. 1. Pharmacological analysis of [3H]g-hydroxybutyrate (GHB)- and
[3H]6,7,8,9-tetrahydro-5-hydroxy-5H-benzocyclohept-6-ylidene acetic acid
(NCS-382)-binding sites in brains of wild-type and GABAB(1) / mice. (A)
Autoradiograms of [3H]GHB binding (30 nM) to sagittal brain sections of wildtype (/), heterozygous (/±) and GABAB(1) knockout mice ( / ). Nonspeci®c binding (nsb) was determined in the presence of 10 mM GHB; the
colour calibration indicates fmol/mg protein. (B) Displacement of [3H]NCS382 binding to cerebral cortex membranes from wild-type (/, &, ) and
GABAB(1) knockout mice ( / , &, *) by GHB and NCS-382. Data are means
SD from single experiments performed in triplicate, lines are hill equations
®tted to the data. The Ki values are 1.9 0.2 (/) and 2.4 0.1 mM ( / ) for
GHB and 0.27 0.04 (/) and 0.27 0.06 mM ( / ) for NCS-382 (n 2).
(C) Representative saturation isotherms of [3H]NCS-382 binding to cerebral
cortex membranes of wild-type (/, ) and GABAB(1) knockout mice ( / ,
*). The KD values calculated from three individual experiments are 0.36 0.06
(/) and 0.38 0.06 mM ( / ) and the Bmax values 26.7 1.9 (/) and
29.9 2.4 ( / ) pmol/mg protein (n 3).
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
Loss of g-hydroxybutyrate action in GABAB(1) /
mice
2725
Lack of g-hydroxybutyrate-induced GTPg[35S] binding in
GABAB(1) / mice
g-Hydroxybutyrate was described as stimulating GTPg[35S] binding
in rat brain slices, suggesting the involvement of G-proteins in its
signal transduction pathway. Therefore, it was proposed that the
GHB receptor belongs to the family of G-protein-coupled receptors
(Ratomponirina et al., 1995; Snead, 2000). We analysed GHB-induced
GTPg[35S] binding using cerebral cortex membranes of wild-type and
GABAB(1) / mice (Fig. 2). Using a rat brain slice preparation, Snead
(2000) reported GHB-induced GTPg[35S] binding at low (micromolar)
concentrations. We investigated a broad concentration range of GHB
(0.01±20 mM) and did not observe signi®cant GHB-induced GTPg[35S]
binding in wild-type mice, using both our own experimental conditions
(Fig. 2A; Urwyler et al., 2001) and those described by Snead (2000;
data not shown). However, in several experiments a small but nonsigni®cant increase of GTPg[35S] binding was observed at high GHB
concentrations (>1 mM). As millimolar concentrations of GHB were
previously shown to activate heterologously expressed GABAB(1,2)
receptors (Lingenhoehl et al., 1999), we addressed whether GHB was
able to generate a GTPg[35S] binding signal via the activation of
GABAB receptors. To this end, GHB was coapplied with the GABAB
receptor positive modulator CGP7930 (Urwyler et al., 2001). In this
context it is of note that CGP7930 does not stimulate GTPg[35S]
binding on its own but is only active when a GABAB receptor agonist is
present (Urwyler et al., 2001). In the presence of CGP7930, GHB
induced a substantial GTPg[35S] binding signal at concentrations of
1 mM and above (Fig. 2B). The stimulatory effect of GHB was not
observed in the presence of the GABAB receptor antagonist CGP54626
and NCS-382 failed to antagonize GHB-induced GTPg[35S] binding
(Fig. 2B). In the presence of CGP7930, millimolar GHB concentrations also stimulated GTPg[35S] binding at recombinantly expressed
heteromeric GABAB(1b,2) receptors (data not shown). In view of the
pharmacological data described above we concluded that high GHB
concentrations directly activate GABAB receptors. In further support
of this we did not observe any GHB-induced GTPg[35S] binding using
cerebral cortex membranes from GABAB(1) / mice, either in the
absence or presence of GABAB receptor positive modulators (Fig. 2C).
In summary, our experiments contrast with previous ®ndings (Snead,
2000) and do not support the hypothesis that high-af®nity [3H]GHBbinding sites re¯ect functional G-protein-coupled receptor ligandbinding sites.
A
B
C
g-Hydroxybutyrate-induced hypolocomotor effects are
absent in GABAB(1) / mice
Administration of GHB to rodents induces a decrease in locomotor
activity (Nissbrandt & Engberg, 1996). To explore the hypolocomotor
effects of GHB, a suitable dose of GHB was ®rst de®ned using BALB/c
mice, the mouse strain that was used to generate GABAB(1) / mice
(Schuler et al., 2001). Different oral doses of GHB were applied and
the spontaneous horizontal locomotor activity as well as the total
distance traveled was recorded over a 1-h observation period (Fig. 3A).
g-Hydroxybutyrate induced a signi®cant dose-related reduction in
horizontal activity (F4,41 60.37, P < 0.001) with oral doses of
300 mg/kg GHB signi®cantly reducing the total horizontal locomotor activity. A long-lasting sedative effect over the 1-h recording period
was seen after oral application of 1 g/kg GHB. This dose was, therefore, selected to study potential differential effects of GHB in wildtype, heterozygous (GABAB(1)/±) and GABAB(1) / mice. gHydroxybutyrate at 1 g/kg almost completely abolished locomotor
activity of wild-type and heterozygous mice as measured by the total
distance traveled over the 1-h recording period (F1,66 13.15,
P < 0.001). In sharp contrast, administration of 1 g/kg GHB to
Fig. 2. Effect of g-hydroxybutyrate (GHB) on GTPg[35S] binding using cerebral cortex membranes from wild-type and GABAB(1) / mice. (A) Effect of
different concentrations of GHB on GTPg[35S] binding using wild-type membranes. The basal level and the stimulation obtained with 1 mM g-aminobutyric
acid (GABA) are also shown for reference. In most experiments a marginal,
nonsigni®cant stimulatory effect of GHB was observed at a high concentration
(> 1 mM GHB). (B) Effect of different concentration of GHB on GTPg[35S]
binding in the presence of the GABAB receptor positive modulator 2,6-di-tertbutyl-4-(3-hydroxy-2,2-dimethyl-propyl)-phenol (CGP7930, 30 mM; Urwyler
et al. 2001). The stimulatory effect of GHB was not observed in the presence of
the GABAB receptor antagonist CGP54626 and was not antagonized by the
putative GHB antagonist 6,7,8,9-tetrahydro-5-hydroxy-5H-benzocyclohept-6ylidene acetic acid (NCS-382) (100 mM). Membranes from wild-type mice were
used. (C) Effect of GHB on GTPg[35S] binding using membranes from
GABAB(1) / mice. No signi®cant stimulation was observed either in the
absence or presence of CGP7930. In some experiments 10 and 20 mM GHB
slightly reduced GTPg[35S] binding compared with basal levels. Small inhibitory effects at high (millimolar) drug concentrations were occasionally
observed with various ligands and different membrane preparations and probably represent an unspeci®c inhibition. Dotted lines in B and C denote the basal
level of stimulation.
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
2726 K. Kaupmann et al.
A
A
B
B
Fig. 3. Analysis of the sedative effects of g-hydroxybutyrate (GHB) in wildtype (/), heterozygous (/±) and GABAB(1) / mice. (A) In wild-type
BALB/c mice application of GHB dose-dependently induced sedation as
measured by a reduction of horizontal locomotor activity compared with vehicle
(data are means SEM, n 9±10). Groups that differed signi®cantly from
vehicle-treated mice (P < 0.05, Dunnett's posthoc tests). (B) Application of 1 g/
kg GHB reduced the horizontal locomotor activity in wild-type (/) and
heterozygous (/±) but not in GABAB(1) / mice (data are means SEM,
n 10±14). Groups that differed signi®cantly from vehicle-treated mice;
#groups that differed signi®cantly from genotype control (P < 0.05, Fisher's
posthoc tests).
GABAB(1) / mice failed to induce any locomotor impairment
(Fig. 3B; F(2,66 4.60, P < 0.05). In vehicle-treated mice, locomotor
behaviour was similar in the three different genotypes with some
indication of hyperactivity of GABAB(1) / mice as previously
described by Schuler et al. (2001).
g-Hydroxybutyrate induces hypothermia in wild-type but not
GABAB(1) / mice
g-Hydroxybutyrate induces hypothermia in rodents (Snead, 1990)
and in humans (Chin et al., 1998). To explore the effects of GHB on
body temperature, a broad dose range of GHB (10±1500 mg/kg p.o.)
was ®rst investigated in BALB/c mice. g-Hydroxybutyrate induced a
marked dose-related hypothermia (F6,63 222.87, P < 0.001) which
changed over time (F6,63 14.21, P < 0.001). Posthoc analysis
demonstrated that GHB did not alter body temperature at doses
below 300 mg/kg p.o. whereas doses of 300 mg/kg signi®cantly
decreased the body temperature (Fig. 4A). Doses of 1 and 1.5 g/kg
induced a prolonged signi®cant hypothermia that lasted 3 and 4 h,
respectively (Fig. 4A). In a direct comparison a dose of 1 g/kg was
Fig. 4. Core body temperature after g-hydroxybutyrate (GHB) application in
wild-type and GABAB(1) / mice. (A) Body temperature after application of
different oral doses of GHB to BALB/c mice (data are means SEM, n 10).
Groups that differed signi®cantly from vehicle-treated mice (P < 0.05, Dunnett's posthoc tests). (B) Body temperature after application of 1 g/kg GHB
(p.o.) to wild-type (/) (n 10) and GABAB(1) / mice (n 7). The arrow
denotes the time of compound application. #Groups that differed signi®cantly
from genotype control (P < 0.05, Fisher's posthoc tests).
applied to GABAB(1) / mice and to wild-type controls (Fig. 4B).
Analysis of variance (repeated measures) revealed that there was a
signi®cant difference in temperature responses to GHB between
both genotypes (F1,15 15.05, P < 0.001) and a genotype±time
interaction (F5,75 69.40, P < 0.001). g-Hydroxybutyrate induced
a marked ( 6 8C) hypothermia in wild-type animals. However, there
was no signi®cant effect of GHB on temperature in GABAB(1) /
mice over the 3-h recording period after GHB application (Fig. 4B).
Posthoc analysis revealed that there was a slightly, but signi®cantly,
lower basal temperature in GABAB(1) / mice compared with
wild-type mice at both time points prior to GHB administration
(Fig. 4B).
Lack of g-butyrolactone-induced increase in dopamine synthesis
in GABAB(1) / mice
g-Hydroxybutyrate and GBL, a prodrug of GHB, cause an almost
instantaneous and marked increase in dopamine synthesis which
subsides rapidly and is followed by a decrease in synthesis compared
with controls after about 1 h and later (Gessa et al., 1966; Walters &
Roth, 1972; Nowycky & Roth, 1978; Waldmeier, 1991). Baclofen
shares these effects with GHB and GBL, and GABAB antagonists such
as CGP35348 prevent the stimulatory effect of baclofen and GHB on
dopamine synthesis (Waldmeier, 1991).
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
Loss of g-hydroxybutyrate action in GABAB(1) /
mice
2727
®cant trend towards a higher number of seizures after GBL application
observed. Most seizures in GABAB(1) / mice lasted for about 1 min
and were of a mild clonic type (forepaw clonus). As described
previously by Schuler et al. (2001), in some cases seizures included
absence-like phenomena such as immobility and 3±6 Hz spike±wave
complexes in the EEG or developed from clonic into tonic seizures.
Two hours after drug treatment the EEG pattern in wild-type mice
reverted to normal whereas that in GABAB(1) / mice remained
unchanged.
Discussion
Fig. 5. g-Butyrolactone (GBL) and baclofen increase striatal dopamine synthesis in wild-type (/) and heterozygous (/±) but not in GABAB(1) / mice.
Mice (sex-matched) were treated with 750 mg/kg i.p. GBL, 20 mg/kg i.p.
baclofen or vehicle 5 min before 100 mg/kg i.p. of the aromatic amino acid
decarboxylase inhibitor 3-hydroxy-benzylhydrazine dihydrochloride was applied. The mice were killed by decapitation 45 min thereafter and L-3,4-dihydroxyphenylacetic acid (DOPA) accumulation was determined by HPLC with
electrochemical detection. Data are means SEM (n 5±7). In / and /±
mice the effects of GBL and baclofen are signi®cantly different from vehicle
(P < 0.01), whereas there is no signi®cant difference for / mice (Dunnett's test).
We measured striatal dopamine synthesis after GBL administration
(750 mg/kg i.p.) to wild-type, heterozygous and GABAB(1) / mice
(Fig. 5). To measure the physiological responses produced by GHB,
GBL was used instead of GHB. g-Butyrolactone is more readily
absorbed and is converted rapidly and irreversibly to GHB after i.p.
administration (Hu et al., 2001). The accumulation of L-DOPA after
central L-aromatic amino acid decarboxylase inhibition (NSD1015,
100 mg/kg, i.p.) was measured 45 min after drug application (Waldmeier, 1991). g-Butyrolactone, like the GABAB receptor agonist
baclofen, markedly increased L-DOPA levels in wild-type and heterozygous mice (P < 0.01, Dunnett's test). However, no such increase in
L-DOPA levels was observed in GABAB(1) / mice (Fig. 5).
g-Butyrolactone induces delta wave electroencephalogram
alterations in wild-type but not GABAB(1) / mice
We investigated whether GBL application induces differential effects
on the EEG pattern in freely moving wild-type and GABAB(1) /
mice (Fig. 6). None of the wild-type BALB/c mice showed EEG
abnormalities prior to GBL application; 10 and 20 min after i.p.
administration of 100 mg/kg GBL the occurrence of delta waves with
a main frequency of 1±2 Hz was signi®cantly increased in wild-type
but not GABAB(1) / mice (Fig. 6A±C). This correlated with a
signi®cant decrease in theta, alpha and beta2 waves in wild-type mice
only. The behaviour associated with GBL application to wild-type
mice consisted of reduced locomotor activity. In rodents, GHB and
GBL have been described as inducing absence-like seizures characterized by the occurrence of spike-and-waves discharges at a frequency
of 5±6 Hz (Snead, 1992; Aizawa et al., 1997; Hu et al., 2000, 2001).
We did not observe `spike-and-wave discharges' after GBL application
to the wild-type BALB/c mice but spikes in the EEG were observed in
seven of eight mice.
g-Butyrolactone application to GABAB(1) / mice did not signi®cantly change the relative contributions to the EEG frequencies of
delta, theta, alpha and beta2 waves (Fig. 6A). On the other hand,
spontaneously recurring spikes and seizures were seen before and after
administration of GBL in several of the GABAB(1) / mice examined
(Fig. 6D and E). Within the 3-h period prior to GBL application three
of six GABAB(1) / mice exhibited seizures (®ve seizures in total).
After GBL application four of six mice had seizures (nine seizures, 3-h
period). Only for one of the six GABAB(1) / mice was a nonsigni-
Given both the therapeutic effects and emerging public-health issues
related to the use of GHB, it is essential to characterize the receptors
that mediate its effects. Our current study demonstrates that GABAB
receptors are involved in all of the well-characterized GHB-elicited
responses tested for.
There is increasing evidence that a number of biochemical, physiological and pharmacological responses that follow GHB application
are mediated by GABAB receptors (Aizawa et al., 1997; Colombo
et al., 1998; Erhardt et al., 1998; Madden & Johnson, 1998; Jensen &
Mody, 2001). De®ning the role of GABAB receptors in complex
biochemical and behavioural responses requires tools that are both
complete in action and speci®c in nature. The current studies take
advantage of recently generated mice lacking the GABAB(1) receptor
subunit. GABAB(1) / mice do not exhibit any residual pre- and
postsynaptic GABAB receptor responses (Schuler et al., 2001; Prosser
et al., 2001). Thus, mice lacking the GABAB(1) subunit provide an
excellent means to study the effects of GHB, in the absence of
interfering GABAB responses.
Receptor autoradiography and competition binding experiments
demonstrated speci®c binding sites in the brains of GABAB(1) /
mice for [3H]GHB and the GHB binding-site antagonist [3H]NCS-382
(Fig. 1). Our ®ndings support the previous notion that GABAB receptors do not signi®cantly contribute to GHB-binding sites in the brain
(Bernasconi et al., 1999). This had already been suggested based on the
differential expression pro®le and ontogeny of GABAB and speci®c
GHB-binding sites (Snead, 1994). As native GABAB receptors are
heteromeric assemblies of two subunits, GABAB(1) and GABAB(2), it
could be speculated that the GABAB(2) subunit contributes to the GHBbinding sites seen in GABAB(1) / mice. The expression pattern in
the brain of GABAB(1) and GABAB(2) does not fully overlap (Kulik
et al., 2002). Therefore, it is conceivable that the individual GABAB
receptor subunits ful®l additional functions independent of heteromeric GABAB(1,2) receptors. Ablation of the GABAB(1) subunit in
mice leads to a drastic down-regulation of the GABAB(2) subunit
(Prosser et al., 2001; Schuler et al., 2001). However, the maximal
number of binding sites for the GHB antagonist [3H]NCS-382 is
similar in GABAB(1) / and wild-type mice (Fig. 1), which renders
a contribution of GABAB(2) to brain GHB-binding sites unlikely.
Moreover, the lack of evolutionary conservation of the putative
ligand-binding domain suggests that the GABAB(2) subunit is not
involved in the binding of a natural ligand (Kniazeff et al., 2002).
Micromolar GHB concentrations failed to induce signi®cant
responses in functional GTPg[35S]-binding experiments (Fig. 2). In
the presence of a GABAB receptor positive modulator high concentrations of GHB (1 mM) stimulated GTPg[35S] binding, stimulation
which was not observed in the presence of a GABAB receptor
antagonist. These data are in accordance with our previous ®ndings
showing that GHB is a low potency partial agonist at recombinantly
expressed GABAB(1,2) receptors (EC50, 5 mM; Lingenhoehl et al.
1999). The endogenous levels of GHB in the brain do not exceed
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
2728 K. Kaupmann et al.
A
B
C
D
E
Fig. 6. g-Butyrolactone (GBL) induces delta waves in the electroencephalogram (EEG) in wild-type but not GABAB(1) / mice. Female mice (17 weeks) were
operated as described in Materials and methods and used in the experiment 3 weeks later. (A) Power spectra of EEG patterns in wild-type and GABAB(1) / mice.
The relative contributions to the EEG frequencies of delta, theta, alpha and beta2 waves 10 min before ( 10) and at indicated time points (10 min recordings) after the
application of 100 mg/kg i.p. GBL are shown. In wild-type BALB/c mice GBL signi®cantly increased delta wave power 10 and 20 min after drug application and
decreased the relative contributions of theta, alpha and beta2 waves (P < 0.01, paired t-test). No signi®cant EEG changes were recorded after GBL application to
GABAB(1) / mice. (B±E) EEG traces from individual wild-type BALB/c and GABAB(1) / mice before and after GBL application (100 mg/kg i.p). (B and C)
GBL application to wild-type mice induced delta wave patterns with a main frequency of 1±2 Hz, occasionally spikes in the EEG were observed. (D and E) Selected
EEG traces to show seizures (D) and spikes (E) in the EEG of GABAB(1) / mice before and after GBL application. In periods devoid of seizure activity the EEG
pattern of wild-type and GABAB(1) / mice was similar (Schuler et al., 2001). n 8, wild-type; n 6, GABAB(1) / .
micromolar concentrations and, therefore, GABAB(1,2) receptors are
unlikely to mediate the effects of endogenous GHB if it is not
synaptically released. Ratomponirina et al. (1995) observed that
GTP or its analogues decreases the af®nity of brain [3H]GHB-binding
sites, suggesting that GHB receptors are G-protein-coupled to Gai/otype G-proteins. In contrast, Snead (1996) reported an increase in
af®nity of [3H]GHB-binding sites in the frontal cortex after GTP or
pertussis toxin treatment. While the absence of detectable GHBinduced GTPg[35S] binding in GABAB(1) / mice does not exclude
a G-protein-coupled GHB receptor our data clearly do not further
corroborate this.
In order to dissect the in vivo effects of GHB from GABAB receptor
pathways we investigated, in GABAB(1) / mice, several welldescribed effects of GHB or its prodrug GBL. Like baclofen, GHB
induces sedation in rodents (Nissbrandt & Engberg, 1996). GABAB
receptor antagonists block the sedative effects of GHB, suggesting a
direct involvement of GABAB receptors. In support of this, GHB did
not show any sedative effects in GABAB(1) / mice (Fig. 3). gHydroxybutyrate had marked hypothermic effects that were absent
in GABAB(1) / mice (Fig. 4). Interestingly, GABAB(1) / mice
have a slight, yet signi®cant, reduction in basal temperature which
indicates that GABAB receptors play a key role in the maintenance of
thermoregulatory homeostasis. In experiments investigating striatal
dopamine synthesis and changes in the EEG after application of GBL,
a complete lack of effect was observed in GABAB(1) / mice, in
sharp contrast to wild-type mice (Figs 5 and 6).
g-Hydroxybutyrate/GBL has been described to induce 3±6 Hz
spike-and-wave discharges characteristic of absence-type seizures in
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
Loss of g-hydroxybutyrate action in GABAB(1) /
rats and mice (Hu et al., 2000; Snead et al., 2000). In contrast, we
observed GBL-induced spikes but no spike-and-wave discharges in the
BALB/c mice used in our study (Fig. 6). The most prominent EEG
change after GBL application was the appearance of delta waves,
similar to the effects of baclofen in this strain of mice (Schuler et al.,
2001). Our results, therefore, contrast with previous data reporting the
induction of absence-type seizures after application of GBL or baclofen to CD-1/129svj hybrid mice (Snead et al., 2000). Strain differences
may account for the apparent discrepancy. Of note, similar to our
observations in mice, delta wave induction after GHB application to
humans has been reported, concomitant with a decrease in alpha and
beta power (Entholzner et al., 1995).
In all of the above in vivo experiments we only recorded signi®cant
effects in wild-type mice after application of relatively high doses of
GHB or GBL. g-Hydroxybutyrate easily penetrates the blood±brain
barrier. A concentration in the brain of approximately 250 mM has been
determined 60 min after i.p. application of 200 mg/kg GHB to rats
(Kaufman et al., 1990). It is, therefore, expected that the effective
doses used in our studies led to brain concentrations high enough to
allow signi®cant activation of GABAB receptors.
It is important to note that compensatory changes which may
obliterate GHB effects in GABAB(1) / mice cannot be excluded.
However, many effects of GHB are antagonized by GABAB receptor
antagonists which renders such an explanation unlikely (Waldmeier,
1991; Xie & Smart, 1992; Williams et al., 1995; Nissbrandt &
Engberg, 1996; Colombo et al., 1998; Erhardt et al., 1998; Madden
& Johnson, 1998; Carai et al., 2001; Jensen & Mody, 2001). In
summary, it appears most likely that all of the investigated pharmacological effects of GHB and GBL are directly mediated by the
activation of GABAB receptors.
In view of the agonistic properties of GHB at GABAB receptors it is
interesting to compare the in vivo effects of GHB with those of the
prototypic GABAB agonist, baclofen. When overdosed, both drugs can
induce coma and respiratory depression (Ingels et al., 2000; Chapple
et al., 2001). Both baclofen and GHB have bene®cial effects in certain
conditions of drug abuse, such as alcoholism and opiate and heroin
withdrawal (Gallimberti et al., 1994, 2000; Addolorato et al., 2002).
However, GHB itself may cause physical dependence (Nicholson &
Balster, 2001) whereas no abuse potential has been reported for
baclofen after more than 30 years of clinical use. g-Hydroxybutyrate
is utilized as an adjuvant in anaesthesia whereas baclofen is not used
for similar purposes but has some analgesic properties. g-Hydroxybutyrate was recently approved by the Food and Drug Administration
as medication for the treatment of narcolepsy (Tunnicliff & Raess,
2002). In contrast, baclofen has never been described to be effective in
this indication. Altogether, these ®ndings suggest that additional effector systems account for the differential effects of GHB/GBL and baclofen. Interestingly, both GHB and baclofen can enhance brain and
plasma levels of certain GABAA receptor active neurosteroids such as
allopregnanolone and allotetrahydrodeoxy-corticosterone (Barbaccia
et al., 2002). Again, antagonist studies indicate that this effect of GHB
is mediated through GABAB receptors (Barbaccia et al., 2002).
The presence of speci®c high-af®nity binding sites for [3H]GHB in
brains of GABAB(1) / mice is intriguing. These binding sites may
represent an as yet uncharacterized GHB receptor and/or binding
protein and may be involved in the abovementioned effects of GHB
which are not mimicked by GABAB agonists.
Acknowledgements
We would like to thank Nicole Reymann, Jakob Heid, Rita Meyerhofer and
Hugo Buerki for expert technical assistance. B.B. is supported by grant 3100-
mice
2729
067100.01 of the Swiss Science Foundation. P.W. and H.B-O. are supported by
the Danish Medical Research Council.
Abbreviations
Baclofen, p-chloro-beta-phenyl-GABA; CGP54626, [3-[[1-(S)-(3,4-dichlorophenyl)ethyl]amino]-2-(S)-hydroxy-propyl]-cyclohexylmethyl-phosphinic
acid hydrochloride; DOPA, 3,4-dihydroxyphenylacetic acid; EEG, electroencephalogram; GABA, g-aminobutyric acid; GABAB(1), GABAB receptor subunit 1; GABAB(2), GABAB receptor subunit 2; GBL, g-butyrolactone; GHB, ghydroxybutyrate; NCS-382, 6,7,8,9-tetrahydro-5-hydroxy-5H-benzocyclohept6-ylidene acetic acid; NSD1015, 3-hydroxy-benzylhydrazine dihydrochloride.
References
Addolorato, G., Cibin, M., Capristo, E., Beghe, F., Gessa, G.L., Stefanini,
G.F. & Gasbarrini, G. (1998) Maintaining abstinence from alcohol with
g-hydroxybutyric acid. Lancet, 351, 38.
Addolorato, G., Caputo, F., Capristo, E., Janiri, L., Bernardi, M., Agabio, R.,
Colombo, G., Gessa, G.L. & Gasbarrini, G. (2002) Rapid suppression of
alcohol withdrawal syndrome by baclofen. Am. J. Med., 112, 226±229.
Aizawa, M., Ito, Y. & Fukuda, H. (1997) Roles of g-aminobutyric acidB
(GABAB) and g-hydroxybutyric acid receptors in hippocampal long-term
potentiation and pathogenesis of absence seizures. Biol. Pharm. Bull., 20,
1066±1070.
Barbaccia, M.L., Colombo, G., Affricano, D., Carai, M.A., Vacca, G., Melis, S.,
Purdy, R.H. & Gessa, G.L. (2002) GABAB receptor-mediated increase of
neurosteroids by gamma-hydroxybutyric acid. Neuropharmacology, 42,
782±791.
Bernasconi, R., Mathivet, P., Bischoff, S. & Marescaux, C. (1999) g-hydroxybutyric acid: an endogenous neuromodulator with abuse potential? Trends
Pharmacol. Sci., 20, 135±141.
Carai, M.A., Colombo, G., Brunetti, G., Melis, S., Serra, S., Vacca, G., Mastinu,
S., Pistuddi, A.M., Solinas, C., Cignarella, G., Minardi, G. & Gessa, G.L.
(2001) Role of GABAB receptors in the sedative/hypnotic effect of ghydroxybutyric acid. Eur. J. Pharmacol., 428, 315±321.
Cash, C.D. (1994) g-hydroxybutyrate: an overview of the pros and cons for
it being a neurotransmitter and/or a useful therapeutic agent. Neurosci.
Biobehav. Rev., 18, 291±304.
Castelli, M.P., Mocci, I., Langlois, X., Gommeren, W., Luyten, W.H.M.L.,
Leysen, J.E. & Gessa, G.L. (2000) Quantitative autoradiographic distribution
of g-hydroxybutyric acid binding sites in human and monkey brain. Brain
Res. Mol. Brain Res., 78, 91±99.
Chapple, D., Johnson, D. & Connors, R. (2001) Baclofen overdose in two
siblings. Pediatr. Emerg. Care, 17, 110±112.
Chin, R.L., Sporer, K.A., Cullison, B., Dyer, J.E. & Wu, T.D. (1998) Clinical
course of g-hydroxybutyrate overdose. Ann. Emerg. Med., 31, 716±722.
Colombo, G., Agabio, R., Lobina, C., Reali, R. & Gessa, G.L. (1998) Involvement of GABAA and GABAB receptors in the mediation of discriminative
stimulus effects of g-hydroxybutyric acid. Physiol. Behav., 64, 293±302.
Entholzner, E., Mielke, L., Pichlmeier, R., Weber, F. & Schneck, H. (1995)
EEG changes during sedation with g-hydroxybutyric acid. Anaesthetist, 44,
345±350.
Erhardt, S., Andersson, B., Nissbrandt, H. & Engberg, G. (1998) Inhibition of
®ring rate and changes in the ®ring pattern of nigral dopamine neurons by ghydroxybutyric acid (GHBA) are speci®cally induced by activation of
GABAB receptors. Naunyn Schmiedeberg's Arch. Pharmacol., 357, 611±
619.
Gallimberti, L., Schifano, F., Forza, G., Miconi, L. & Ferrara, S.D. (1994)
Clinical ef®cacy of g-hydroxybutyric acid in treatment of opiate withdrawal.
Eur. Arch. Psychiat. Clin. Neurosci., 244, 113±114.
Gallimberti, L., Spella, M.R., Soncini, C.A. & Gessa, G.L. (2000) g-hydroxybutyric acid in the treatment of alcohol and heroin dependence. Alcohol,
20, 257±262.
Gessa, G.L., Vargiu, L., Crabai, F., Boero, G.C., Caboni, F. & Camba, R. (1966)
Selective increase of brain dopamine induced by g-hydroxybutyrate. Life
Sci., 5, 1921±1930.
Hechler, V., Ratomponirina, C. & Maitre, M. (1997) g-Hydroxybutyrate
conversion into GABA induces displacement of GABAB binding that
is blocked by valproate and ethosuximide. J. Pharmacol. Exp. Ther., 281,
753±760.
Hogema, B.M., Gupta, M., Senephansiri, H., Burlingame, T.G., Taylor, M.,
Jakobs, C., Schutgens, R.B.H., Froestl, W., Snead, O.C., Diaz-Arrastia, R.,
Bottiglieri, T., Grompe, M. & Gibson, K.M. (2001) Pharmacologic rescue of
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730
2730 K. Kaupmann et al.
lethal seizures in mice de®cient in succinate semialdehyde dehydrogenase.
Nature Genet., 29, 212±216.
Hu, R.Q., Banerjee, P.K. & Snead, O.C. 3rd (2000) Regulation of g-aminobutyric acid (GABA) release in cerebral cortex in the g-hydroxybutyric acid
(GHB) model of absence seizures in rat. Neuropharmacology, 39, 427±439.
Hu, R.Q., Cortez, M.A., Man, H.Y., Roder, J., Jia, Z., Wang, Y.T. & Snead, O.C.
3rd (2001) g-hydroxybutyric acid-induced absence seizures in GluR2 null
mutant mice. Brain Res., 897, 27±35.
Ingels, M., Rangan, C., Bellezzo, J. & Clark, R.F. (2000) Coma and respiratory
depression following the ingestion of GHB and its precursors: three cases.
J. Emerg. Med., 19, 47±50.
Jensen, K. & Mody, I. (2001) GHB depresses fast excitatory and inhibitory
synaptic transmission via GABAB receptors in mouse neocortical neurons.
Cereb. Cortex, 11, 424±429.
Kaufman, E.E., Porrino, L.J. & Nelson, T. (1990) Pyretic action of low doses of
g-hydroxybutyrate in rats. Biochem. Pharmacol., 40, 2637±2240.
Kleinschmidt, S., Schellhase, C. & Mertzlufft, F. (1999) Continuous sedation
during spinal anaesthesia: g-hydroxybutyrate vs. propofol. Eur. J. Anaesthesiol., 16, 23±30.
Kniazeff, J., Galvez, T., Labesse, G. & Pin, J.P. (2002) No ligand binding in the
GB2 subunit of the GABAB receptor is required for activation and allosteric
interaction between the subunits. J. Neurosci., 22, 7352±7361.
Kulik, A., Nakadate, K., Nyiri, G., Notomi, T., Malitschek, B., Bettler, B. &
Shigemoto, R. (2002) Distinct localization of GABAB receptors relative to
synaptic sites in the rat cerebellum and ventrobasal thalamus. Eur. J. Neurosci., 15, 291±307.
Lingenhoehl, K., Brom, R., Heid, J., Beck, P., Froestl, W., Kaupmann, K.,
Bettler, B. & Mosbacher, J. (1999) g-hydroxybutyrate is a weak agonist at
recombinant GABAB receptors. Neuropharmacology, 38, 1667±1673.
Madden, T.E. & Johnson, S.W. (1998) g-hydroxybutyrate is a GABAB receptor
agonist that increases a potassium conductance in rat ventral tegmental
dopamine neurons. J. Pharmacol. Exp. Ther., 287, 261±265.
Maitre, M. (1997) The g-hydroxybutyric acid signalling system in the brain:
organization and functional implications. Prog. Neurobiol., 51, 337±361.
Maitre, M., Hechler, V., Vayer, P., Gobaille, S., Cash, C.D., Schmitt, M. &
Bourguignon, J.J. (1990) A speci®c g-hydroxybutyrate receptor ligand
possesses both antagonistic and anticonvulsant properties. J. Pharmacol.
Exp. Ther., 255, 657±663.
Maitre, M., Andriamampandry, C., Kemmel, V., Schmidt, C., Hode, Y., Hechler,
V. & Gobaille, S. (2000) g-hydroxybutyric acid as a signaling molecule in
brain. Alcohol, 20, 277±283.
Marshall, F.H., Jones, K.A., Kaupmann, K. & Bettler, B. (1999) GABAB receptors Ð the ®rst 7TM heterodimers. Trends Pharmacol. Sci., 20, 396±399.
Mehta, A.K., Muschaweck, N.M., Maeda, D.Y., Coop, A. & Ticku, M.K. (2001)
Binding characteristics of the g-hydroxybutyric acid receptor antagonist
[3H](2E)-(5-hydroxy-5,7,8,9-tetrahydro-6H- benzo[a][7]annulen-6-ylidene)
ethanoic acid in the rat brain. J. Pharmacol. Exp. Ther., 299, 1148±1153.
Nicholson, K.L. & Balster, R.L. (2001) GHB: a new and novel drug of abuse.
Drug Alcohol. Depend., 63, 1±22.
Nissbrandt, H. & Engberg, G. (1996) The GABAB-receptor antagonist, CGP
35348, antagonises g-hydroxybutyrate- and baclofen-induced alterations in
locomotor activity and forebrain dopamine levels in mice. J. Neural Transm.,
103, 1255±1263.
Nowycky, M.C. & Roth, R.H. (1978) Dopaminergic neurons: role of presynaptic receptors in the regulation of transmitter biosynthesis. Progr. NeuroPsychopharmacol. Biol. Psychiat., 2, 139±158.
Prosser, H.M., Gill, C.H., Hirst, W.D., Grau, E., Robbins, M., Calver, A., Sof®n,
E.M., Farmer, C.E., Lanneau, C., Gray, J., Schenck, E., Warmerdam, B.S.,
Clapham, C., Reavill, C., Rogers, D.C., Stean, T., Upton, N., Humphreys, K.,
Randall, A., Geppert, M., Davies, C.H. & Pangalos, M.N. (2001) Epileptogenesis and enhanced prepulse inhibition in GABAB(1) de®cient mice. Mol.
Cell. Neurosci., 17, 1059±1070.
Ransom, R.W. & Stec, N.L. (1988) Cooperative modulation of [3H]MK-801
binding to the N-methyl-D-aspartate receptor-ion channel complex by
L-glutamate, glycine, and polyamines. J. Neurochem., 51, 830±836.
Ratomponirina, C., Hode, Y., Hechler, V. & Maitre, M. (1995) g-Hydroxybutyrate receptor binding in rat brain is inhibited by guanyl nucleotides and
pertussis toxin. Neurosci. Lett., 189, 51±53.
Schuler, V., LuÈscher, C., Blanchet, C., Klix, N., Sansig, G., Klebs, K., Schmutz,
M., Heid, J., Gentry, C., Urban, L., Fox, A., Spooren, W., Jaton, A.L.,
Vigouret, J.M., Pozza, M., Kelly, P.H., Mosbacher, J., Froestl, W., Kaslin, E.,
Korn, R., Bischoff, S., Kaupmann, K., van der Putten, H. & Bettler, B. (2001)
Epilepsy, hyperalgesia, impaired memory, and loss of pre- and postsynaptic
GABAB responses in mice lacking GABAB(1). Neuron, 31, 47±58.
Snead, O.C. 3rd (1990) g-Hydroxybutyric acid-induced seizures bear no
relation to core temperature. Epilepsia, 31, 253±258.
Snead, O.C. 3rd (1992) Pharmacological models of generalized absence
seizures in rodents. J. Neural Transm. Suppl., 35, 7±19.
Snead, O.C. 3rd (1994) The ontogeny of [3H] g-hydroxybutyrate and [3H]
GABAB binding sites: relation to the development of experimental absence
seizures. Brain Res., 659, 147±156.
Snead, O.C. 3rd (1996) Relation of the [3H] g-hydroxybutyric acid (GHB)
binding site to the g-aminobutyric acidB (GABAB) receptor in rat brain.
Biochem. Pharmacol., 52, 1235±1243.
Snead, O.C. 3rd (2000) Evidence for a G protein-coupled g-hydroxybutyric acid
receptor. J. Neurochem., 75, 1986±1996.
Snead, O.C. 3rd, Banerjee, P.K., Burnham, M. & Hampson, D. (2000)
Modulation of absence seizures by the GABAA receptor: a critical
role for metabotropic glutamate receptor 4 (mGluR4). J. Neurosci., 20,
6218±6224.
Spooren, W.P., Gasparini, F., van der Putten, H., Koller, M., Nakanishi, S. &
Kuhn, R. (2000) Lack of effect of LY314582 (a group 2 metabotropic
glutamate receptor agonist) on phencyclidine-induced locomotor activity
in metabotropic glutamate receptor 2 knockout mice. Eur. J. Pharmacol., 26,
397R1±2.
Tunnicliff, G. & Raess, B.U. (2002) g-hydroxybutyrate (orphan medical). Curr.
Opin. Invest. Drugs, 3, 278±283.
Urwyler, S., Mosbacher, J., Lingenhoehl, K., Heid, J., Hofstetter, K., Froestl,
W., Bettler, B. & Kaupmann, K. (2001) Positive allosteric modulation
of native and recombinant g-aminobutyric acid(B) receptors by 2,6-ditert-butyl-4-(3-hydroxy-2,2-dimethyl-propyl)-phenol (CGP7930) and its
aldehyde analog CGP13501. Mol. Pharmacol., 60, 963±971.
Waldmeier, P.C. (1991) The GABAB antagonist, CGP 35348, antagonizes the
effects of baclofen, g-butyrolactone and HA 966 on rat striatal dopamine
synthesis. Naunyn Schmiedeberg's Arch. Pharmacol., 343, 173±178.
Waldmeier, P.C., Baumann, P.A., Fehr, B., De Herdt, P. & Maitre, L. (1984)
Carbamazepine decreases catecholamine turnover in the rat brain. J. Pharmacol. Exp, Ther., 231, 166±172.
Walters, J.R. & Roth, R.H. (1972) Effect of g-hydroxybutyrate on dopamine
and dopamine metabolism in the rat striatum. Biochem. Pharmacol., 21,
2111±2121.
Williams, S.R., Turner, J.P. & Crunelli, V. (1995) g-hydroxybutyrate promotes
oscillatory activity of rat and cat thalamocortical neurons by a tonic GABAB,
receptor-mediated hyperpolarization. Neuroscience, 66, 133±141.
Xie, X. & Smart, T.G. (1992) g-hydroxybutyrate hyperpolarizes hippocampal
neurones by activating GABAB receptors. Eur. J. Pharmacol., 212, 291±294.
ß 2003 Federation of European Neuroscience Societies, European Journal of Neuroscience, 18, 2722±2730