Inhibition of firing rate and changes in the firing pattern of nigral dopamine neurons by γ-hydroxybutyric acid (GHBA) are specifically induced by activation of GABA(B) receptors

Article abstract of DOI:10.1007/PL00005215

Previous studies have shown that administration of γ-hydroxybutyric acid (GHBA) or the GABA(B) receptor agonist baclofen are associated with a decrease in firing rate, a regularisation of firing pattern and a decrease in burst activity of midbrain dopamine (DA) neurons in the substantia nigra (SN). In the present study we compared the ability of the novel GABA(B) receptor antagonist SCH 50911 and the selective antagonist of GHBA binding sites, NCS-382, to antagonise the effects of baclofen or GHBA, respectively, on the neuronal activity of BA neurons in anaesthetised rats. SCH 50911 (75 mg/kg, i.v.) was found to antagonise the decrease in firing rate, the regularisation of firing rhythm and the decrease of burst activity in DA cells, induced by baclofen (1-32 mg/kg, i.v.) or GHBA (12.5-1600 mg/kg, i.v.). NCS-382 (100 mg/kg, i.v.) did not affect the baclofen-induced changes in neuronal activity. Neither was the drug able to influence the GHBA-induced alterations in firing rate or in burst activity, although NCS-382 to some extent antagonised the regularisation of the firing pattern observed following low doses of GHBA (<100 mg/kg). The results of the present study give further support for the notion that the GHBA-induced changes in neuronal activity of nigral dopamine neurons are mediated by stimulation of GABA(B) receptors.

Full text of DOI:10.1007/PL00005215

Naunyn-Schmiedeberg’s Arch Pharmacol (1998) 357:611–619
ORIGINAL ARTICLE
© Springer-Verlag 1998
Sophie Erhardt · Bengt Andersson
Hans Nissbrandt · Göran Engberg
Inhibition of firing rate and changes in the firing pattern
of nigral dopamine neurons by γ-hydroxybutyric acid (GHBA)
are specifically induced by activation of GABAB receptors
Received: 13 October 1997 / Accepted: 10 March 1998
Abstract Previous studies have shown that administration
of γ-hydroxybutyric acid (GHBA) or the GABAB receptor
Introduction
agonist baclofen are associated with a decrease in firing
rate, a regularisation of firing pattern and a decrease in burst γ-Hydroxybutyric acid (GHBA) is an endogenous com-
activity of midbrain dopamine (DA) neurons in the substan- pound thought to act as a neurotransmitter in mammalian
tia nigra (SN).
brain (Roth and Giarman 1970; Vayer et al. 1987). GHBA,
In the present study we compared the ability of the novel which is synthesised from GABA (Snead et al. 1989), dis-
GABAB receptor antagonist SCH 50911 and the selective plays a discrete, uneven distribution with high levels in the
antagonist of GHBA binding sites, NCS-382, to antagonise substantia nigra (SN; Vayer et al. 1988). Administration of
the effects of baclofen or GHBA, respectively, on the neu- GHBA, or its prodrug γ-butyrolactone (GBL), produces se-
ronal activity of DA neurons in anaesthetised rats. SCH dation and anaesthesia in humans, although its low potency
50911 (75 mg/kg, i.v.) was found to antagonise the decrease in this regard has limited its use as an anaesthetic (Snead
in firing rate, the regularisation of firing rhythm and the de- 1977; Tunnicliff 1992). In animals, the drug has been re-
crease of burst activity in DA cells, induced by baclofen ported to suppress the spontaneous firing rate of midbrain
(1–32 mg/kg, i.v.) or GHBA (12.5–1600 mg/kg, i.v.). NCS- dopamine (DA) neurons (Roth et al. 1973), to decrease lo-
382 (100 mg/kg, i.v.) did not affect the baclofen-induced comotor activity, and to increase brain DA levels, the latter
changes in neuronal activity. Neither was the drug able to effects being attributed to a reduction in dopamine release
influence the GHBA-induced alterations in firing rate or in (Gessa et al. 1968; Walters and Roth 1972; Stock et al.
burst activity, although NCS-382 to some extent anta- 1973; Nissbrandt and Engberg 1996). Many effects of
gonised the regularisation of the firing pattern observed fol- GHBA on the central nervous system show a remarkable
lowing low doses of GHBA (≤100 mg/kg).
similarity to the effects produced by the GABAB-receptor
The results of the present study give further support for the agonist baclofen. In this sense, the mode of action by which
notion that the GHBA-induced changes in neuronal activity GHBA produces its central effects remains a matter of con-
of nigral dopamine neurons are mediated by stimulation of troversy. According to previous radioligand binding stud-
GABAB receptors.
ies, some actions of GHBA have been suggested to be me-
diated via activation of highly specific membrane-binding
sites without affinity for GABA, GABAergic agonists or
GBL (Benavides et al. 1982; Maitre et al. 1983; see Maitre
1997). Furthermore, these binding sites appear to have a
different regional, anatomic distribution in rat brain com-
pared to GABAB-binding sites (Snead 1994). On the other
hand, many effects produced by GHBA are shown to be at-
tenuated by administration of GABAB-receptor antagonists.
Previous studies in our laboratory have shown that system-
ic administration of GHBA, like baclofen, produces several
effects on the neuronal activity of DA neurons in the SN, in-
cluding a decreased firing rate and a regularisation of the
firing rhythm with reduction of burst activity (Engberg and
Nissbrandt 1993). These actions of GHBA were attenuated
by the specific GABAB-receptor antagonist CGP 35348, in-
dicating a GABAB-receptor-mediated action of GHBA
Key words Substantia nigra · γ-Hydroxybutyrate ·
Baclofen · SCH 50911 · NCS-382 · Dopamine · Burst
firing · GABAB receptors
S. Erhardt ( ) · G. Engberg
✉
Karolinska Institute, Department of Physiology and Pharmacology,
Division of Pharmacology, SE 171 77 Stockholm, Sweden
B. Andersson · H. Nissbrandt
Department of Pharmacology, Göteborg University, P.O. Box 431,
SE 413 90 Göteborg, Sweden

612
guage called “G” (Lab VIEW, National Instruments, Austin, Tex.,
USA). The software was designed to sample and analyse the intervals
of an arbitrary number of TTL pulses (corresponding to spikes pass-
ing through the discriminating filter) using a time resolution of 1 µs.
An interspike interval was designated as the time (in ms) elapsed be-
tween the rising edge of two sequential TTL pulses. In order to avoid
artefacts in the sampling procedure, time intervals below 20 ms were
ignored by the spike analyser. The onset of a burst was determined as
an interspike time interval shorter than 80 ms and termination of a
burst by the next interval longer than 160 ms (Grace and Bunney
(Engberg and Nissbrandt 1993). Accordingly, CGP 35348
has been shown to antagonise in vivo the effects of bacl-
ofen, GHBA or GBL on striatal DA synthesis/release
(Waldmeier 1991; Nissbrandt and Engberg 1996). More-
over, GHBA has been reported to displace [3H]baclofen and
the GABAB-receptor antagonist [3H]CGP 27492 from
GABAB sites in cortical and thalamic homogenates
(Bernasconi et al. 1992).
In the present study we utilised SCH 50911, a novel se- 1984). Cells were considered to be bursting if at least 1 interspike
time interval of 100 recorded spikes was below 80 ms. The intervals
were analysed with regard to number of bursts that occurred during
each 100-spike sampling period along with a calculation of the per-
centage of spikes in bursts.
lective and potent GABAB-receptor antagonist (Bolser et al.
1995) and NCS-382, a novel compound suggested to act as
a specific antagonist of GHBA sites (Maitre et al. 1990), in
order to further characterise the receptor involved in the ef-
fects of GHBA on the firing of nigrostriatal DA neurons.
In addition, the software programme sorted the intervals and di-
vided them into 3-ms bins and displayed the results as an interspike
time interval histogram (ISH) with regard to the number of intervals
corresponding to each bin. The analysis also included calculation of
the variation coefficient, which was used as a measure of the regular-
ity of firing (Werner and Mountcastle 1963). Variation coefficient, ex-
pressed as the median of at least three consecutive ISHs, was calculat-
ed as the ratio between the standard deviation and the mean interval of
an ISH. Data collected from the spike analyser obtained less than 1
min after drug administration were not included in the analysis in or-
der to exclude drug wear-on effects.
Materials and methods
Animals
The experiments were performed on male BKL:SD rats (B&K Uni-
versal AB, Sollentuna, Sweden; weighing between 180 and 250 g).
The animals were housed in groups of five and kept under constant
temperature (25°C), maintained on a 12-h light/dark cycle (lights on
at 06.00 h) with ad libitum access to food and water.
Chemistry
The preparation of NCS-382 was mainly performed as described pre-
viously (Maitre et al. 1990), however, with some modifications given
below.
Animal surgery
Rats were anaesthetised (chloral hydrate 400 mg/kg, intraperitoneal-
ly) and mounted in a stereotaxic frame (David Kopf Instr.,Tujunga,
Calif., USA). Additional anaesthesia and drugs were given through a
lateral tail vein. Throughout the experiments the body temperature of
the animals was maintained at 37°C by means of a heating pad. The
skull surface was exposed and a 3-mm burr hole was drilled with its
centre located approximately 2 mm anterior to lambda and 2 mm lat-
eral to the midline. The dura was carefully removed and a single-bar-
rel micropipette was lowered about 7 mm into the region of the SN.
5-Oxo-5,7,8,9-tetrahydrobenzocycloheptene acetic acid
Commercially available1-benzosuberone (12.5 g, 78.1 mmol) was re-
acted under argon atmosphere with glyoxylic acid as previously de-
scribed. After the reaction was completed, the mixture was poured in-
to crushed ice. The crystals were filtered off, washed with water and
dissolved in diluted NaOH solution (1.5 equivalent NaOH). The aque-
ous layer was extracted with CH2Cl2 and separated. The aqueous so-
lution was poured into diluted HCl (40 ml conc. HCl in 200 g ice-wa-
ter) under intense stirring. The crystals were filtered off and carefully
washed with water and dried at 50°C, yielding 13.3 g (81%) of pure
product.
Single unit recording
A glass microelectrode with a tip diameter of approximately 1–2 µm
(filled with 2 M sodiumacetate saturated with Pontamine Sky Blue)
was lowered by means of a hydraulic microdrive (David Kopf Instr.,
Tujunga, Calif., USA) into the zona compacta-SN. The in vitro im-
pedance of the electrodes was generally 6–9 MΩ, measured in 0.9%
saline at 135 Hz. Single unit potentials were passed through a high in-
put-impedance amplifier and filters. Nerve impulses were discrimi-
nated from background noise and fed into a computer, and simulta-
neously displayed on a digital storage oscilloscope, monitored on an
audio monitor and on a strip chart recorder (Gould). The position of
the electrode was marked at the end of each experiment by ionto-
phoretic ejection of Pontamine Sky Blue. The brains were then re-
moved, buffered in 10% formaldehyde solution and subsequently
sliced in 50-µm-thick sections by a microtome and stained with neu-
tral red. Only DA cells within the SN were included in this study and
only one neuron per animal was studied.
5-Hydroxy-5,7,8,9-tetrahydrobenzocycloheptene acetic acid
To a solution of 5-oxo-5,7,8,9-tetrahydrobenzocycloheptene acetic
acid (1 g, 4.6 mmol) inMeOH (30 ml) fresh NaBH4 (1 g, 26.4 mmol)
was added. The solution was refluxed for 4 h. Additional NaBH4 (0.5
g, 13.2 mmol) was added and the reaction mixture was refluxed over
night. The reaction mixture was poured into H2O (200 ml) and the pH
was adjusted to 1–2 with diluted aqueous HCl and extracted with
CH2Cl2. The organic layer was washed with water, separated, dried
(Na2CO3) and filtered, yielding 0.5 g (50%) as crude product. Rec-
rystallisation from trichlorethylene afforded pure compound as crys-
1
2
3
tals. H NMR (300 MHz, DMSO) 1.5–1.9 (m, H), 2.7–3.1 (m, H),
3.4–3.6 (m, 1H), 5.3 (s, 1H), 5.8 (s, br, 1H), 5.9 (s, 1H), 7.1–7.3 (m, 3H),
7.5 (d, 1H). 13C NMR (75.4 MHz, DMSO) 27.83, 29.83, 33.91, 75.69,
114.27, 125.29, 126.37, 126.98, 129.32, 139.79, 141.78, 162.78,
167.85.
Data analysis
The product was converted to the sodium salt (NCS-382) as previ-
ously described (Maitre et al. 1990).However, we were unable to use
isopropyl alcohol as a tritration agent, since the compound dissolved
in this solvent. Instead, diisopropyl ether was used. The compound
was finally dried under vacuum at room temperature.
The temporal distribution of spikes was analysed on-line utilising a
Macintosh computer. The software used for the analysis of firing was
written in-house using a high level object-oriented programming lan-

613
Fig. 1a–d Extracellular record-
ing from a spontaneously burst-
ing dopamine (DA) neuron in
the substantia nigra (SN) follow-
ing intravenous administration
of GHBA. a Interspike time in-
terval histogram (ISH) before
administration of GHBA. b,c
ISHs after the administration of
GHBA. d Cumulative rate histo-
gram depicting the action of
GHBA (GHBA, 12.5+12.5+25+
50+100+200+400+800 mg/kg,
at arrows) on the firing rate of
the same neuron. Horizontal
bars indicate the time periods
where the three ISHs were
recorded
Fig. 2a–d Extracellular recording from a spontaneously bursting DA
neuron in the SN following intravenous administration of baclofen. a
ISH before administration of baclofen. b,c ISHs after the administra-
tion of baclofen. d Cumulative rate histogram depicting the action of
baclofen (baclofen, 1+1+2+4+8+16 mg/kg, at arrows) on the firing
rate of the same neuron. Horizontal bars indicate the time periods
where the three ISHs were recorded

614
Fig. 3a–e Extracellular recording from a spontaneously bursting DA
neuron in the SN following intravenous administration of NCS-382
and GHBA. a ISH before drug administration. b ISH after the admin-
istration of NCS-382. c,d ISHs after the administration of GHBA. e
Cumulative rate histogram depicting the action of GHBA (GHBA,
12.5+12.5+25+50+100+200+400 mg/kg, at arrows) on the firing rate
of the same neuron in a rat treated with NCS-382 (100 mg/kg, i.v.).
Horizontal bars indicate the time periods where the four ISHs were
recorded
Fig. 4a–e Extracellular recording from a DA neuron in the SN fol-
lowing intravenous administration of NCS-382 and baclofen. a ISH
before drug administration. b ISH after the administration of NCS-
382. c,d ISHs after the administration of baclofen. e Cumulative rate
histogram depicting the action of baclofen (baclofen, 1+1+2+4+8+16
mg/kg, at arrows) on the firing rate of the same neuron in a rat treated
with NCS-382 (100 mg/kg, i.v.). Horizontal bars indicate the time
periods where the four ISHs were recorded

615
Fig. 5a–e Extracellular recording from a DA neuron in the SN fol-
lowing intravenous administration of SCH 50911 and GHBA. a ISH
before drug administration. b ISH after the administration of SCH
50911. c,d ISHs after the administration of GHBA. e Cumulative rate
histogram depicting the action of GHBA (GHBA, 12.5+12.5+25+50
+100+200+400+800 mg/kg, at arrows) on the firing rate of the same
neuron in a rat treated with SCH 50911 (75 mg/kg, i.v.). Horizontal
bars indicate the time periods where the four 15 Hs were recorded
Fig. 6a–e Extracellular recording from a DA neuron in the SN fol-
lowing intravenous administration of SCH 50911 and baclofen. a ISH
before drug administration. b ISH after the administration of SCH
50911. c,d ISHs after the administration of baclofen. e Cumulative
rate histogram depicting the action of baclofen (baclofen, 1+1+2+4+
8+16+32 mg/kg, at arrows) on the firing rate of the same neuron in a
rat treated with SCH 50911 (75 mg/kg, i.v.). Horizontal bars indicate
the time periods where the four ISHs were recorded
dent decrease in the firing rate of nigral DA neurons (Figs.
1, 2, 7). Furthermore, both drugs dose-dependently induced
a regularisation of the firing pattern, expressed as the varia-
tion coefficient of an ISH, with a maximal effect (90%
decrease with GHBA and 91% decrease with baclofen) ob-
tained at 800 mg/kg or 32 mg/kg, respectively (Figs. 1, 2,
8). In all spontaneously bursting cells tested, both GHBA
and baclofen also significantly reduced the percentage of
burst firing (Figs. 1, 2, 9a, 10a).
Drugs
Chloral hydrate (Merck, Darmstadt, Germany); D-L-baclofen and
SCH 50911 (kindly donated by Dr. Kreutner, Schering-Plough, Ke-
ningworth, N.J., USA); γ-hydroxybutyric acid (GHBA; Sigma, St.
Louis, Mo.,USA).
Results
Administration of NCS-382 (100 mg/kg, i.v.), suggested
to act as a specific antagonist of GHBA sites, did not affect
Intravenous administration of GHBA (12.5–1600 mg/kg, the regularity of firing or the firing rate of the DA neurons
i.v.) or baclofen (1–32 mg/kg, i.v.) induced a dose-depen- when given alone. The percentage of spikes occurring in

616
Fig. 7a,b Dose response curves illustrating the action of intravenous- Fig. 8a,b Dose response curves for the change in regularity of firing
ly administered GHBA (a) or baclofen (b) on the firing rate of nigral of nigral DA neurons, induced by GHBA (a) or baclofen (b), ex-
DA neurons in controls, in rats treated with SCH 50911 (75 mg/kg, pressed as the variation coefficient, in controls, in rats treated with
i.v., 3–8 min before GHBA or baclofen) and in rats treated with NCS- SCH 50911 (75 mg/kg, i.v., 3–8 min before GHBA or baclofen) and
382 (100 mg/kg, i.v., 3–8 min before GHBA or baclofen). Each point in rats treated with NCS-382 (100 mg/kg, i.v., 3–8 min before GHBA
represents the mean ± SEM obtained from 6–9 rats. Statistics: or baclofen). Each point represents the mean ± SEM obtained from
*P<0.05,**P<0.01 vs controls (Mann-Whitney U-test)
6–9 rats. Statistics: *P<0.05,*P<0.01 vs controls (Mann-Whitney
U-test)
bursts was not affected by the drug (data not shown). Fur-
thermore, pretreatment with NCS-382 (100 mg/kg, i.v., 3–8
min) was not found to significantly antagonise the GHBA-
Discussion
induced reduction in firing rate or burst activity, although The results of the present study confirm the previous report-
the drug to some extent antagonised the regularisation of ed action of GHBA to induce an inhibition in firing rate and
the firing rhythm observed following low doses of GHBA a regularised firing pattern with a lack of burst activity of
(≤100 mg/kg; Figs. 3, 7a, 8a, 9b). NCS-382 was not nigral DA neurons (Engberg and Nissbrandt 1993; Niss-
able to influence the dose-dependent decrease in variation brandt et al. 1994). In this sense, the actions of GHBA show
coefficient, inhibition of firing rate or reduction in burst ac- a remarkable similarity to the effects induced by the specif-
tivity induced by baclofen (Figs. 4, 7b, 8b, 10b).
ic GABAB-receptor agonist baclofen on these neurons
Pretreatment with the novel, potent, and selective (Engberg et al. 1993). All these electrophysiological effects
GABAB-receptor antagonist, SCH 50911 (75 mg/kg, i.v., of baclofen on nigrostriatal DA neurons are suggested to be
3–8 min) antagonised all effects, i.e., inhibition of firing the consequence of an activation of a somatodendritic
rate, regularisation of firing rhythm and burst activity in- GABAB-receptor-mediated input to the SN, which mainly
duced by both GHBA and baclofen (Figs. 5–8, 9c, 10c). serves to regularise the firing pattern and to prevent burst
Systemic administration of SCH 50911 (75 mg/kg, i.v.) activity of nigral DA neurons. Functionally, activation of
alone induced a slight increase in burst activity in spontane- this GABAB-receptor-mediated input decreases DA termi-
ously bursting nigral DA cells, and a vast majority of non- nal efflux in the striatum (Nissbrandt et al. 1994). Further-
bursting cells were converted to a bursting firing pattern by more, recent studies from our laboratory demonstrate that
the drug (Figs. 5, 6).
this input to the SN constitutes the final link in a striatoni-
gral inhibitory control system (Engberg et al. 1997). As dis-
cussed below, the results of the present investigation dem-

617
Fig. 9 a Effect of GHBA on burst activity of nigral DA neurons in Fig. 10 a Effect of baclofen on burst activity of nigral DA neurons in
controls. Each bar represents the mean ± SEM of 6 spontaneously controls. Each bar represents the mean ± SEM of 4 spontaneously
bursting neurons. b Effect of GHBA on burst activity of nigral DA bursting neurons. b Effect of baclofen on burst activity of nigral DA
neurons in rats pretreated with NCS-382 (100 mg/kg, i.v., 3–8 min be- neurons in rats pretreated with NCS-382 (100 mg/kg, i.v., 3–8 min be-
fore the first injection of GHBA). Each bar represents the mean ± fore the first injection of baclofen). Each bar represents the mean ±
SEM of 9 bursting cells. c Effect of GHBA on burst activity of nigral SEM of 7 bursting cells. c Effect of baclofen on burst activity of 12
DA neurons in rats pretreated with SCH 50911 (75 mg/kg, i.v., 3–8 nigral DA neurons in rats pretreated with SCH 50911 (75 mg/kg, i.v.,
min before the first injection of GHBA). Each bar represents the mean 3–8 min before the first injection of baclofen). Each bar represents the
± SEM of 6 bursting cells
mean ± SEM of 5 bursting cells
onstrate that GHBA, like baclofen, produces its actions on a previous study the effects of baclofen, GHBA and GBL
nigral DA neurons by affecting the above-described on the firing of nigral DA neurons were antagonised by pre-
GABAB-receptor-mediated input to the SN.
treatment with the specific GABAB-receptor antagonist
The major issue of the present study was to evaluate CGP 35348 (Engberg et al. 1993) which is devoid of affini-
whether the effects of GHBA on the neuronal activity of ty for GHBA binding sites (Snead 1996). In addition, in the
nigral DA neurons is mediated via GABAB receptors or via present study a novel GABAB-receptor antagonist, SCH
affinity sites for GHBA. With regard to GHBA´s central ef- 50911 (Bolser et al. 1995), was found to antagonise the ef-
fects in general, previous studies have formed a picture full fects of both baclofen and GHBA. SCH 50911 is reported
of contradictions (cf. Introduction). Based on the present to act as a potent and highly specific GABAB-antagonist
investigation, there are several reasons to believe that with no affinity for a variety of other receptors including rat
GHBA produces its effects on the firing of nigral DA neu- GABAA-, histamine H1-, DA-, D2-, D3-, or GHBA receptors
rons specifically via activation of GABAB receptors and not (Bolser et al. 1995; Snead 1996). The effectiveness of SCH
via stimulation of GHBA binding sites: (1) the fact that 50911 in inhibiting the actions of GHBA on the firing of
GHBA and baclofen [which is reported devoid of affinity to nigral DA neurons principally indicates that GHBA in this
GHBA binding sites (Snead 1996)] produces almost identi- regard activates GABAB receptors. Thirdly, the effects of
cal effects on the firing of nigral DA neurons indicates per GHBA and baclofen on nigral DA activity were not reliably
se that GHBA in this regard displays a similar mode of ac- antagonised by NCS-382. This drug is reported to selective-
tion as baclofen, i.e. stimulation of GABAB receptors; (2) in ly block GHBA sites with no antagonistic action at GABAB

618
receptors (Maitre et al. 1990; Snead 1996) and was in the
Notably, SCH 50911appeared more potent than CGP
present study administered (100 mg/kg, i.v.) at a dose twice 35348 in antagonising the electrophysiological actions of
as high as that required to fully antagonise behavioural ef- GHBA or baclofen of nigral DA neurons. Thus, pretreat-
fects of GHBA (Colombo et al. 1995). NCS-382 was not ment with SCH 50911 (75 mg/kg, i.v.) almost totally pre-
able to antagonise the decrease in firing rate as produced by vented the inhibition of firing rate induced by baclofen.
GHBA, nor the GHBA-induced reduction in burst firing. When rats were pretreated with CGP 35348 (200 mg/kg,
Furthermore, the baclofen-induced alterations in the neu- i.v.), baclofen given in high doses (32 mg/kg, i.v.) still pro-
ronal activity of nigral DA neurons, i.e. reduction in firing duced about 60% inhibition of firing rate (Engberg et al.
rate, variation coefficient of ISHs and burst firing, was not 1993).This difference in potency between the two GABAB-
influenced by pretreatment with NCS-382. Among all elec- receptor antagonists is also in agreement with previous in
trophysiological parameters measured on nigral DA neu- vivo studies comparing their ability to antagonise inhibition
rons, the only significant effect observed with NCS-382 of the antitussive or the respiratory depressant effect of
was a slight antagonism of the decrease in variation coeffi- baclofen (Bolser et al. 1995).
cient of ISHs following doses of 100 mg/kg or less of
Altogether, the results of the present study clearly show
GHBA. This action of NCS-382 is obscure and cannot eas- that GHBA and the GABAB-receptor agonist baclofen sim-
ily be explained in terms of GHBA- or GABAB-receptor ilarly produce their actions on the firing rate and burst ac-
mechanisms.
tivity of nigral DA neurons by specifically inhibiting nigral
In all, our results show that GHBA affects DA neuronal DA neuronal activity via activation of GABAB receptors.
activity via activation of GABAB receptors. A stumbling
Acknowledgements This study was supported by the Swedish Med-
ical Research Council (no. 7484), Åhlén-stiftelsen, Loo och Hans
block with this idea is the lack of reliable binding data
showing high affinity of GHBA to GABAB-receptor sites.
Ostermans stiftelse, and the Karolinska Institute. The expert technical
Rather, some studies show a lack of affinity of GHBA to
assistance of Ms. Carolina Hurtado is gratefully acknowledged.
GABAB-receptor sites (Benavides et al. 1982; Snead and
Liu 1984; see Maitre 1997). One may suggest that GHBA
could influence the neuronal DA activity indirectly, e.g. by
References
increasing concentrations of GABA, either by a formation
of GABA from GHBA or, less likely though, via a release
of the transmitter (see Maitre1997). If so, however, then
GABAA receptors, in addition to GABAB receptors, should
be indirectly affected by GHBA. Previous studies have
shown that systemic administration of muscimol, a specific
GABAA-receptor agonist, is associated with effects on neu-
ronal DA activity opposite to those produced by GHBA (i.e.
increased firing rate and burst activity), making this hy-
pothesis less attractive. Furthermore, whereas the actions of
GHBA on neuronal DA activity are almost instantaneous in
onset, no convincing picture from previous studies has
emerged that central endogenous levels of GABA are im-
mediately increased; whereas several studies fail to show
increased GABA concentration following GHBA adminis-
tration (Mitoma and Neubauer 1968; Margolis 1969;
Doherty et al. 1975; Möhler et al. 1976; Doherty et al.
1978), others report a slight GABA increase, however,
more than 60 min after injection of GHBA (Della Pietra et
al. 1966; De Feudis and Collier 1970). In addition, a puta-
tive releasing action of GABA by GHBA as induced via
GHBA sites is unlikely to occur since such an effect would
be blocked by NCS. Consequently, an indirect action of
GHBA in affecting neuronal DA activity appears unlikely
judging from our results. Rather, a direct action on recep-
tors sensitive to GHBA, i.e. GABAB receptors, is likely to
account for the effects of GHBA in this regard. Such a
mode of action of the drug on central mechanisms is also
strongly supported by biochemical and behavioural data
(Nissbrandt and Engberg 1996), where the GABAB-recep-
tor antagonist CGP 35348 was shown to antagonise the
similar action of GHBA and baclofen on striatal DA release
and locomotor activity, respectively.
Benavides J, Rumigny JF, Bourguignon JJ, Cash C, Wermuth C, Man-
del P, Vincendon G, Maitre M (1982) High affinity binding sites
for γ-hydroxybutyric acid in rat brain. Life Sci 30:953–961
Bernasconi R, Lauber J, Marescaux C, Vergnes M, Martin P, Rubio V,
Leonhardt T, Reymann N, Bittiger H (1992) Experimental ab-
sence seizures: potential role of γ-hydroxybutyric acid and
GABAB receptors. J Neural Transm Suppl 35:155–177
Bolser DC, Blythin DJ, Chapman RW, Egan RW, Hey JA, Rizzo C,
Kuo S-C, Kreutner W (1995) The pharmacology of SCH 50911: a
novel, orally active GABAB receptor antagonist. J Pharmacol Exp
Ther 274:1393–1398
Colombo G, Agabio R, Bourguignon J, Fadda F, Lobina C, Maitre M,
Reali R, Schmitt M, Gessa GL (1995) Blockade of the discrimina-
tive stimulus effects of γ-hydroxybutyric acid (GHB) by the GHB
receptor antagonist NCS-382. Physiol Behav 58:587–590
De Feudis FV, Collier B (1970) Conversion of γ-hydroxybutyrate to γ-
aminobutyrate by mouse brain in vivo. Experientia 26:1072–1073
Della Pietra G, Illiano G, Capano V, Rava R (1966) In vivo conversion
of γ-hydroxybutyrate into γ-aminobutyrate. Nature 210:733–734
Doherty JD, Stout RW, Roth RH (1975) Metabolism of [1–14 C]γ-hy-
droxybutyric acid by rat brain after intraventricular injection.
Biochem Pharmacol 24:469–474
Doherty JD, Hattox SE, Snead OC, Roth RH (1978) Identification of
endogenous γ-hydroxybutyrate in human and bovine brain and its
regional distribution in human, guinea pig and rhesus monkey
brain. J Pharmacol Exp Ther 207:130–139
Engberg G, Nissbrandt H (1993) γ-hydroxybutyric acid (GHBA) in-
duces pacemaker activity and inhibition of SN dopamine neurons
by activating GABAB-receptors. Naunyn Schmiedebergs Arch
Pharmacol 348:491–497
Engberg G, Kling-PetersenT, Nissbrandt H (1993) GABAB-receptors
activation alters the firing pattern of dopamine neurons in the rat
substantia nigra. Synapse 15:229–238
Engberg G, Elverfors A, Jonason J, Nissbrandt H (1997) Inhibition of
dopamine re-uptake: significance for nigral dopamine neuron ac-
tivity. Synapse 25:215–226
Gessa GL, Crabai F, Vargiu L, Spano PF (1968) Selective increase of
brain dopamine induced by γ-hydroxybutyrate: study of the mech-
anism of action. J Neurochem 15:377–381

619
Snead OC (1994) The ontogeny of [3H]γ-butyrate and [3H]GABAB-
binding sites: relation to the development of experimental absence
seizures. Brain Res 659:147–156
Grace AA, Bunney BS (1984) The control of firing pattern in nigral
dopamine neurons: burst firing. J Neurosci 4:2877–2890
Maitre M (1997) The gamma-hydroxybutyrate signalling system in
brain: organization and functional implications. Prog Neurobiol
51:337–361
Maitre M, Rumigny JF, Cash CD, Mandel P (1983) Subcellular distri-
bution of γ-hydroxybutyrate binding sites in rat brain. Principal lo-
calization in the synaptosomal fraction. Biochem Biophys Res
Commun 110:262–265
Snead OC (1996) Relation of the [3H]γ-hydroxybutyric acid (GHB)
binding site to the γ-aminobutyric acidB (GABAB) receptor in the
rat brain. Biochem Pharmacol 52:1235–1243
Snead OC, Liu CC (1984) Gamma-hydroxybutyric acid binding sites
in rat and human brain synaptosomal membranes. Biochem Phar-
macol 33:2587–2590
Maitre M, Hechler V, Vayer P, Gobaille S, Cash CD, Schmitt M, Bo-
urguignon JJ (1990) A specific γ-hydroxybutyrate receptor ligand
possesses both antagonistic and anticonvulsant properties. J Phar-
macol Exp Ther 255:657–663
Snead OC III, Furner R, Liu CC (1989) In vivo conversion of γ-ami-
nobutyric acid and 1,4-butanediol to γ-hydroxybutyric acid in rat
brain: studies using stable isotopes. Biochem Pharmacol
38:4375–4380
Margolis RK (1969) The effect of γ-hydroxybutyric acid on amino-ac-
id levels in brain. Biochem Pharmacol 18:1243–1246
Mitoma C, Neubauer SE (1968) Gamma-hydroxybutyric acid and
sleep. Experientia 24:12–13
Stock G, Magnusson T, Andén N-E (1973) Increase in brain dopam-
ine after axotomy or treatment with γ-hydroxybutyric acid due to
elimination of the nerve impulse flow. Naunyn Schmiedebergs
Arch Pharmacol 278:347–361
Möhler H, Patel AJ, Balazs R (1976) Gamma-hydroxybutyrate degra-
dation in the brain in vivo: negligible direct conversion to GABA.
J Neurochem 27:253–258
Nissbrandt H, Engberg G (1996) The GABAB-receptor antagonist
CGP 35348 antagonises γ-hydroxybutyrate- and baclofen-induced
alterations in locomotor activity and forebrain dopamine levels in
mice. J Neural Transm 103:1255–1263
Tunnicliff G (1992) Significance of γ-hydroxybutyric acid in the
brain. Gen Pharmacol 23:1027–1034
Vayer P, Mandel P, Maitre M (1987) Gamma-hydroxybutyrate, a pos-
sible neurotransmitter. Life Sci 41:1547–1557
Vayer P, Ehrhardt J-D, Gobaille S, Mandel P, Maitre M (1988) γ-
hydroxybutyrate distribution and turnover rates in discrete brain
regions of the rat. Neurochem Int 12:53–59
Nissbrandt H, Elverfors A, Engberg G (1994) Pharmacologically in-
duced cessation of burst activity in nigral dopamine neurons: sig-
nificance for the terminal dopamine efflux. Synapse 17:217–224
Roth RH, Giarman NJ (1970) Natural occurrence of gamma-hydrox-
ybutyric acid in mammalian brain. Biochem Pharmacol
21:2111–2121
Roth RH, Walters JR, Aghajanian GK (1973) Effect of impulse flow
on the release and synthesis of dopamine in the rat striatum. In:
Snyder SH, Usdin E (eds) Frontiers in catecholamine research.
Pergamon Press, New York, pp 567–574
Waldmeier PC (1991) The GABAB-antagonist, CGP 35348, antagoni-
ses the effects of baclofen, γ-butyrolactone and HA 966 on rat stri-
atal dopamine synthesis. Naunyn Schmiedebergs Arch Pharmacol
343:173–178
Walters JR, Roth RH (1972) Effect of gamma-hydroxybutyrate on
dopamine and dopamine metabolites in the rat striatum. Biochem
Pharmacol 21:2111–2121
Werner G, Mountcastle VB (1963) The variability of central neural
activity in a sensory system and its implication for the central re-
flection of sensory events. J Neurophysiol 26:958–977
Snead OC (1977) Minireview: gamma-hydroxybutyrate. Life Sci
20:1935–1943