Department of Pharmacology and Experimental Therapeutics, and the
Neuroscience Graduate Program, Loyola University Chicago, Stritch
School of Medicine, Maywood, Illinois
The present study explored the possibility that excitatory amino acid
(EAA) sensitivity within the ventral pallidum (VP) is altered by
long-term removal of dopamine (DA). Electrophysiological experiments
were conducted in chloral hydrate-anesthetized rats 21 to 28 days after
they received unilateral substantia nigra injections of the
dopaminergic toxin 6-hydroxydopamine (6-OHDA). VP neurons increased
firing at low microiontophoretic ejection currents of the EAA agonists
N-methyl-D-aspartate (NMDA) and
-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA); however,
high currents decreased action potential amplitude and rapidly caused
cessation of neuronal firing. These responses likely reflected the
induction of depolarization block for they were reversed by
coiontophoresis of the hyperpolarizing transmitter 
-aminobutyric
acid (GABA) at ejection current levels that normally suppressed firing.
The ability of NMDA and AMPA to induce such inactivation was greater in
the VP of 6-OHDA-lesioned hemispheres, but unchanged in reserpinized
rats, verifying that the alterations in responding to NMDA were the
result of chronic, rather than acute, DA removal. The adaptations do
not appear to be the consequence of a diminished GABAergic tone for the
ability of bicuculline to increase firing (due to blocking a tonic
GABAergic input) was not changed. However, low ejection currents of
GABA that were insufficient to alter firing rate greatly attenuated the
ability of NMDA to induce an apparent depolarization inactivation when
coiontophoresed with NMDA onto VP neurons of the lesioned, but not the
unlesioned, hemisphere. These studies show that chronic DA removal
altered the EAA-induced amplitude-decreasing (i.e., the apparent
depolarization inactivation) effects in VP neurons in the absence of a
decrease in GABAergic tone.
 |
Introduction |
EAAs
provide the major excitatory drive within the basal ganglia and limbic
systems of the brain. Popular circuitry models for basal ganglia
adaptations that accompany chronic near-total lesions of the ascending
dopaminergic system incorporate the wealth of evidence for an
up-regulation of EAA transmission (for review, see Ossowska, 1994
).
Moreover, suppression of EAA transmission by EAA antagonists attenuates
motor dysfunction and augments the beneficial effects of levodopa in
animal models of PD (Ossowska, 1994). The neuroanatomical substrates
engaged by the up-regulation of EAA transmission in the DA-deafferented
brain include those influenced by the enhancement in activity of STN
EAA-containing projections, e.g., the GPi (Brotchie et al., 1991;
Filion and Tremblay, 1991) and the substantia nigra zona reticulata
(Rohlfs et al., 1997).
The VP receives direct innervation of EAA-containing afferents from
limbic regions (e.g., amygdala; Russchen and Price, 1984; Maslowski-Cobuzzi and Napier, 1994) and the basal ganglia (e.g., STN;
Groenewegen and Berendse, 1990; Turner et al., 2001). These inputs act
on two major subtypes of ionotropic EAA receptors in the VP, NMDA, and
non-NMDA (Monaghan and Cotman, 1985; Standaert et al., 1994). Agonists
to each subtype (i.e., NMDA and AMPA, respectively) induce robust
alterations in VP neuronal activity, with low synaptic concentrations
enhancing firing and higher concentrations producing an excessive
stimulation of the receptor to the point of inducing a state of
inactivation (Turner et al., 2001). As shown in several other brain
regions, when extreme, acute overexcitation can lead to long-term
dysregulation of neuronal function and even excitotoxic cell death (for
review, see Choi, 1992). Thus, if the STN projection to the VP is
up-regulated after chronic DA depletion, this may increase the
propensity for VP neurons to enter a persistent dysfunctional state.
The present study tested this hypothesis by monitoring enhancements in
spiking rate, as well as the apparent depolarization-induced
attenuation of spiking, seen with microiontophoretically applied NMDA
and AMPA in VP neurons after long-term destruction of the ascending
dopaminergic neurons.
In the dorsal striatal-pallidal system, the adaptations in striatal
GABAergic transmission that occur after chronic DA depletion greatly
influence pallidal structures (Smith et al., 1998). The VP receives a
large GABAergic projection from the nucleus accumbens (Walaas and
Fonnum, 1979; Groenewegen and Russchen, 1984), and it is likely that
GABA- and EAA-containing projections converge onto the same VP neurons
(Zaborszky et al., 1991; Johnson and Napier, 1997a). Thus, the
possibility that DA depletion may alter the interactive effects of
these two amino acid transmitter groups on VP neuron spiking also was explored.
| |
Materials and Methods |
Adult, male Sprague-Dawley rats (Harlan Bioproducts for Science,
Indianapolis, IN) were allowed at least 1 week to acclimate to the
local vivarium conditions before experimentation was initiated. They
were housed in pairs, with a 12-h light/dark cycle and ad libitum
access to food and water. All procedures were approved by the Loyola
University Chicago Institutional Animal Care and Use Committee and were
in accord with the Guide for the Care and Use of Laboratory Animals
published by the National Research Council.
DA Depletion Treatments.
One group of rats was unilaterally
treated with the dopaminergic toxin 6-OHDA to deplete DA in one
hemisphere. To do so, rats (250-300 g) were pretreated with
desipramine HCl (30 mg/kg as the salt, administered i.p. in saline, 60 min before 6-OHDA injection; Sigma-Aldrich, St. Louis, MO) and
pargyline (50 mg/kg i.p. in saline, 30 min before 6-OHDA injection;
Sigma-Aldrich), and then anesthetized with pentobarbital (50 mg/kg
i.p.; Cardinal Health, Chicago, IL). A stereotaxically located hole was
drilled over the substantia nigra (5.3 mm anterior to lambda, 2.1 mm
lateral to midline, 7.2 mm ventral to dura). By using 33-gauge
injectors connected to an infusion pump, 6-OHDA HBr (8 µg/4 µl as
base; Sigma-Aldrich) or its 0.3% ascorbic acid vehicle was injected into the nigra of each hemisphere at a rate of 0.5 µl/min. The skull
hole was filled with bone wax, the overlying skin was sutured, and the
animals were allowed to recover from the anesthesia on a warm heating
pad in a dimly lit environment. The rats were fed a high-calorie diet
[sometimes also requiring tube feeding of warmed Similac (Ross
Pediatrics, Columbus, OH) or i.p. injections of warmed saline] for a
few days after surgery; thereafter, they were given standard laboratory
rat chow and water ad libitum. To provide a functional indicator of the
lesion extent, 14 days after the 6-OHDA treatments, the rats were given
an injection of apomorphine (1 mg/kg i.p.; Sigma-Aldrich) and their
circling behavior was quantified. Rats that circled at a rate of nine
turns or more per minute were used for the electrophysiological
experiments 7 to 14 days later (21 to 28 days postlesion).
To allow comparisons between the effects of chronic and acute removal
of DA, another group of rats was acutely depleted of monoamines with
reserpine [5 mg/kg i.p. (Sigma-Aldrich); dissolved using glacial
acetic acid then diluted with 0.3% tartaric acid in 1.5% ethanol
solution and adjusted to pH 4.0 by using 0.01 M NaOH].
Electrophysiological studies were conducted 6 to 8 h later.
The lesion extent was verified for both 6-OHDA- and reserpine-treated
rats. To do so, at the conclusion of the electrophysiological experiments, the striatum and olfactory tubercles were dissected. The
tissue was assayed for monoamine content by HPLC-EC, as used previously
(Napier and Potter, 1989). The HPLC-EC data, presented as nanograms per
milligram of wet tissue ± S.E.M., were compared using ANOVA with
post hoc Newman-Keuls (if significant) with P < 0.05 set for significance.
Electrophysiology and Microiontophoresis.
Rats (280-320 g)
were anesthetized with chloral hydrate (400 mg/kg i.p. as the salt, in
0.9% NaCl; Sigma-Aldrich) and mounted in a stereotaxic apparatus. A
tail vein was cannulated to allow anesthetic supplements as needed, and
the animal's body temperature was maintained at 36°C throughout the
experiment. A glass microelectrode/microiontophoretic pipette assembly
(for construction, see Turner et al., 2001) was lowered into the VP
(7.0-8.5 mm from dura) through a burr hole in the skull (0.4-0.7 mm
posterior to bregma, 2.1-2.4 mm lateral to the midline).
The recording microelectrodes were filled with a 2% pontamine sky
blue/0.5 M sodium acetate solution (tip diameter, 2-3 µm; 4-6 M
at 100 Hz in vitro; Winston Electronics Company, Millbray, CA). Action
potentials from individual VP neurons were amplified, isolated, and
quantified with standard extracellular recording techniques as used
previously (Turner et al., 2001). To normalize the data collection, the
window discriminator (Fintronics, Orange, CT) was set such that the
minimal voltage level that an action potential must achieve to be
counted (i.e., converted to a d.c. signal for the PC) was 50% of the
initial spike amplitude (Figs. 1, B and
C, and 4A). The action potentials were categorized as biphasic or
triphasic and by the initial direction of the voltage deflection
(positive or negative). Amplitude (µV) and duration (ms) also were
determined. Neuronal firing was characterized by rate, ISI analysis of
firing pattern (mean/mode ISI ratio), and the number of spontaneously
firing neurons encountered per pass of the electrode through the VP
(cells per tract). ANOVA was used to compare these profiles among the
various treatment groups, with P < 0.05 set for
significance.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 1.
Rate histograms illustrating the response of a VP
neuron to NMDA. The 10-s applications of NMDA are indicated by the
horizontal bars over the histogram. A, four panels of continuous
recordings of a single VP neuron illustrating the approach used to
generate ejection current-response curves for both the rate enhancement
and apparent depolarization inactivation. The responses, indicated by
the B (at 35 nA in panel 1) and C (at 110 nA in panel 4), are enlarged
in the bottom histograms. B, typical rate enhancement seen at lower
ejection currents. Shown in the inset is an example of an action
potential that occurred during drug retention at the time indicated by
the arrow. Also indicated is the discriminating voltage level
(horizontal line crossing action potential), standardized to 50% of
peak amplitude, used to detect the spike. C, typical responses seen at
higher ejection currents. The detected spiking initially increased and
after a few seconds of application it decreased as action potential
amplitude was reduced until it could no longer be discriminated from
background (apparent depolarization inactivation, e.g., right action
potential inset). Note both the rate enhancement and apparent
depolarization inactivation induced by NMDA outlast the period of drug
ejection. Moreover, these effects are reversible, repeatable, and the
response magnitude correlates with the ejection current magnitude.
|
|
Attached to the recording microelectrode, a five-barrel pipette (for
construction description, see Turner et al., 2001) was used for
microiontophoretic application of the following drugs: AMPA (10 mM in
150 mM NaCl, pH 8.0; Sigma-Aldrich), NMDA (50 mM in 200 mM NaCl, pH
8.0; Sigma-Aldrich), GABA (100 mM in 200 mM NaCl, pH 7.0;
Sigma-Aldrich) and bicuculline (5 mM in H2O, pH 5.6; Sigma-Aldrich). In some cases, the side barrels were filled with
the EAA antagonists CNQX and AP-5 instead of GABA and
bicuculline. Results obtained with the EAA antagonists are reported in
Turner et al. (2001). The center barrel was filled with the 2%
pontamine sky blue/0.5 M sodium acetate solution and it was used to
balance the net charge at the electrode tip. Electrode impedance was 25 to 60 M for the drug-containing barrels and 20 to 45 M for the balance barrel. AMPA and NMDA were ejected with various anionic currents and retained in the pipette by 10 nA of cationic current. Bicuculline was ejected with cationic currents 5 to 120 nA and retained
with a 10-nA anionic current. To eject GABA, the current was
incremented by 2-nA steps in the positive direction from the holding/retention current until the neuron stopped firing. GABA was
retained in the pipette with a 10- to 20-nA anionic current. Retention
currents were always applied in between drug application periods,
including when baseline spiking was assessed.
To determine relative efficacy (i.e., maximal change in firing rate,
Emax), and potency (i.e., the ejection
current required to achieve 50% of this firing rate,
ECur50), responses of VP neurons to
microiontophoretically applied ligands were characterized over a wide
range of ejection currents (typically 5-120 nA, in epochs of 10-s
ejection: 30-s retention). Occasionally, it was necessary to extend the
retention period to ensure that the neuronal firing rate and action
potential shape returned to baseline in between ejection periods. This
pulsatile paradigm produced repeatable, noncumulative, and reversible
changes in VP neuronal firing as well as consistent interejection
(baseline) spiking, even after repeated applications for several hours.
Applying 5 to 120 nA to pipette solutions of 100 to 200 mM NaCl at a pH
ranging 5.6 to 8.0 did not change neuronal firing rate (data not
shown), indicating the effects seen during current application to
drug-containing pipettes did not result from nonselective changes in
local electrolyte concentrations or current application. To test the
ability of GABA (via its hyperpolarizing effects) to reverse the
apparent depolarization block induced by NMDA, constant ejection
currents of NMDA were applied for 1 to 4 min. The NMDA ejection current was fine-tuned so that the action potential shape and amplitude remained stable for the duration of NMDA applications. Then GABA was
coiontophoresed in 10- to 30-s ejection periods. For all treatment assessments, firing rate was quantified as spikes per 1-s bin.
At the end of the experiment, the rat was overdosed with chloral
hydrate. The recording site was marked by depositing pontamine sky blue
at the tip of the microelectrode. The brain was removed and the
olfactory tubercles and dorsal striatum were dissected and quick-frozen
on dry ice for HPLC-EC analysis. The remaining forebrain, which
contained the VP, was cut into 50-µm coronal sections and stained
with cresyl violet. The stereotaxically determined location of
recording sites was verified histologically by two independent observers.
Analysis of Electrophysiological Data.
It has been
demonstrated in several brain regions that after excessive neuronal
activation, the firing rate progressively increases until action
potentials show a decrement in amplitude and this often is followed by
a cessation in spiking (Bunney and Grace, 1978; Grace and Bunney, 1986;
Henry et al., 1992; Hollerman et al., 1992; White et al., 1995; Hu and
White, 1996). When firing rate is quantified by discriminating
individual spikes based on a minimum voltage requirement, this yields
an inverted "U" or bimodal ejection current-response or
dose-response curve. The cessation in spiking can be reversed when the
membrane is hyperpolarized either electrically (Grace and Bunney, 1986)
or by hyperpolarizing agents (Bunney and Grace, 1978; Grace and Bunney,
1986; Henry et al., 1992; Hollerman et al., 1992; White et al., 1995).
By using similar electrophysiological approaches, these phenomena were
assessed in the present study for EAA-induced effects in the VP.
Consistent with aforementioned reports, the increased spiking
that occurred in the present study with lower EAA ejection currents is
hereafter referred to as rate enhancements (Fig. 1, A and B). The
decrease in action potential amplitude and eventual cessation of
spiking that occurred with higher ejection currents is termed apparent
depolarization inactivation (Fig. 1, A-C). EAA-induced rate
enhancements were considered to be significant if firing in any one of
the 10 one-s bins was greater than 20% above the mean spontaneous rate
that immediately preceded EAA iontophoresis, and if this occurred for
three consecutive applications. As ejection current was increased, the
peak rate-enhancing response generally was not maintained due to
encroaching apparent depolarization inactivation (Fig. 1). Thus, for
analyses of the treatment-effect relationship of the EAA-induced rate
enhancement, the ejection current range evaluated for each neuron began
at 5 nA and ended at the ejection current level where firing in three
subsequent bins was decreased from the peak excitatory response level
by at least 20%. To quantify rate enhancement for each 5-nA interval, the peak spikes per 1-s bin was used. Two criteria also were used to
define apparent depolarization inactivation: 1) During EAA application,
the action potential decreased by more than 50% of the baseline
amplitude (i.e., during drug retention). 2) Spike amplitude was
restored to baseline levels during EAA retention. This continuous,
on-line assessment of spike amplitude as well as using a drug
ejection/drug retention amplitude ratio assured that our measures of
the drug-induced amplitude changes were not caused by nonspecific
variations in the status of the recording (e.g., the microelectrode
moving away from the neuron). The ejection current range used for
evaluating the treatment-effect relationship of apparent depolarization
inactivation started from the current level were apparent
depolarization inactivation began (i.e., peak firing per bin changed by
20%), and ended at the maximal ejection current tested (i.e., 120 nA).
As shown in Fig. 1C, maximal apparent depolarization inactivation
generally occurred 8 to 11 s after initiating ejection of
relatively high currents of an EAA agonist (the effects of the agonist
often remained for a few seconds after ending the 10-s ejection
period). Due to the rapidly changing nature of the response, averaging
across several bins dampened down the measured effect rather than
producing the desired effect of minimizing variability. Thus, the
minimum number of spikes per 1-s bin in the time frame where the
amplitude-decreasing effect was the greatest (8-11 s from the
initiation of the 10-s EAA ejection period) was used for data analysis.
Linear regression analysis of both the ascending (i.e., rate-enhancing)
and descending (inactivation) limbs of the population ejection
current-firing rate response relationship was used to ascertain whether
the magnitude of the ejection current of microiontophoretically applied
ligands was related to the response magnitude of VP neurons. Treatment-effect associations were considered significant if the slope
was significantly different from zero (P < 0.05). To
optimize the accuracy of single point descriptors of this relationship (e.g., Emax and
ECur50), the ascending and descending limbs of the treatment-effect data obtained for each neuron were first fitted to
a third order polynomial. This typically yielded an r2 > 0.9, and only curves with an
r2 
0.7 were used for further
analysis. For neurons exhibiting highly fluctuating spiking
(n = 36 of the 204 neurons analyzed) and
correspondingly low r2, an average of
the response at a given current with the response to 5 nA higher and 5 nA lower was used. From the polynomials, the
Emax and ECur50
for both the ascending and descending limbs of the treatment-effect
curves were calculated (GraphPad Prism; GraphPad Software, San Diego,
CA). If the calculated Emax or
ECur50 was ± 3 S.E.M. from the mean, it was
considered to be an outlier (and thus an erroneous estimate of the
event) and the neuron was excluded from analysis. Using this criterion,
no more than two cells were omitted per treatment condition.
To contrast the mean Emax and
ECur50 obtained from the various
iontophoretically applied ligands in lesioned and unlesioned conditions, two sets of comparisons were performed. The first set had
three conditions: control (pooled untreated rats + contralateral to
lesioned hemisphere), 6-OHDA-treated, and reserpine-treated rats. The
second set had four conditions: NMDA alone in pooled control, NMDA
alone in 6-OHDA-treated hemispheres, NMDA + GABA in hemispheres
contralateral to the 6-OHDA treatment, and NMDA + GABA in
6-OHDA-treated hemispheres. If ANOVA proved significant then
Student-Newman-Keuls was used for pairwise comparisons. For analysis of
both sets of NMDA-evoked responses, the same lesion and control groups
were used. Because of these multiple comparisons, a more conservative
of 0.025 was used for determining statistical significance for
ANOVA. For AMPA, only control and 6-OHDA-treated conditions were
examined and the results compared using Student's t test
with P < 0.05. These data are reported as mean spikes
per second ± S.E.M. Categorical comparisons were made using
Pearson's chi square (
2) test of association,
with P < 0.05. Degrees of freedom are indicated in
subscript following the variables for F, t, and
2.
| |
Results |
Electrophysiological Profile of VP Recordings.
Neurons (303)
were histologically verified within the rostral-caudal and
medial-lateral extent of the subcommissural VP and rostral
sublentular substantia innominata (refer to Turner et al., 2001
for details of VP borders and stereotaxic maps of recording sites).
These sites were equally sampled from three groups of rats: untreated
rats (n = 68), rats that received unilateral 6-OHDA microinjections into the substantia nigra to destroy DA-containing neurons (n = 20), and rats acutely depleted of
monoamines by reserpine (n = 9).
Pallidal neuronal subpopulations have been identified by their
electrophysiological characteristics that demonstrate distinct pharmacological profiles to DA (Napier et al., 1991) and to opioids (Mitrovic and Napier, 1995). This possibility was also explored in the
present study. Consistent with prior reports (Napier et al., 1991;
Mitrovic and Napier, 1995; Johnson and Napier, 1997b; Turner et al.,
2001), the most frequently encountered action potential profile was a
biphasic, initially negative waveform. This profile comprised 80% of
neurons (137/172) recorded from intact rats, 95% of neurons (40/42)
from hemispheres contralateral to 6-OHDA-injections), 72% (44/61) of
neurons recorded from the 6-OHDA-lesioned hemisphere, and 100% of
neurons (40/40) from reserpine-treated rats (Table 1). Triphasic, initially positive
waveforms also were encountered, but less frequently (refer to
cells/track for intact controls in Table 1;
t198 = 7.7, P < 0.001). The triphasic action potentials displayed a larger action
potential amplitude (t157 = 4.6, P < 0.001) and a longer duration
(t157 = 9.6, P < 0.001) than did the biphasic action potentials (Table 1, intact
controls). Qualitative histological assessments clearly revealed that
the probability to encounter either neuronal profile was not correlated
with the anatomical locale of the recording site within the VP. Because these two types of neuronal profiles did not distinguish themselves with regard to lesion state or to EAA-induced effects (see below), results obtained from recordings with these two profiles were pooled.
View this table:
[in this window]
[in a new window]
|
TABLE 1
Action potential and spiking characteristics from initially negative
biphasic and initially positive triphasic action potentials recorded in
the VP: comparisons among the various control and treatment conditions
Data are presented as mean ± S.E.M. Action potential calibrations
are as follows: horizontal bar, 1 ms; vertical bar for biphasic action
potential, 100 µV; vertical bar for triphasic action potential, 125 µV. See Results for comparisons between the biphasic and triphasic
profiles. Triphasic action potentials, Student's t test between
treatment conditions, t24 = 0.0 to 1.8, N.S.
|
|
EAA-Induced Responding in VP of Controls.
NMDA induced rate
enhancements (Fig. 1) in 12 of the 20 neurons tested in the hemisphere
contralateral to the 6-OHDA infusions and in eight of eight tested in
the intact rats (2 = 2.7; N.S.). AMPA increased firing
in 100% of the neurons tested in both control conditions (two were
tested in the contralateral hemispheres and nine were tested in intact
rats). The Emax and ECur50 were not different for NMDA between the
two control conditions (t14 = 0.2-0.6, P = 0.57-0.84) (n = 2 for
AMPA-tested cells in contralateral controls and this is too few for an
accurate comparison of means). Thus, the data were pooled for these two
conditions and presented hereafter as "control".
As previously shown in intact controls (Turner et al., 2001), at low
ejection currents for EAA, sensitive VP neurons only responded with an
increase in firing. Increases in the ejection current magnitude
incremented the rate enhancements between 5 and 60 nA for AMPA (linear
regression: slope 0.67 ± 0.04 spikes/s/nA, r2 = 0.96;
F1,10 = 306, P < 0.001; Fig. 2A) and 5 to 70 nA for NMDA (slope 0.51 ± 0.02, r2 = 0.97;
F1,13 = 496, P < 0.001; Fig. 3A). Our prior work
demonstrated that, in these current ranges, the AMPA-induced effects
are attenuated by the AMPA receptor-selective antagonist CNQX (and not
by the NMDA receptor-selective antagonist AP-5), whereas NMDA is
blocked by AP-5 and not by CNQX (Turner et al., 2001). This verifies
that with the ejection current ranges tested, AMPA and NMDA are acting selectively at the non-NMDA and NMDA receptors, respectively.
View larger version (31K):
[in this window]
[in a new window]
|
Fig. 2.
Response of ventral pallidal neurons to AMPA in
control and 6-OHDA-treated conditions. Rate-enhancing (A and B) and
apparent depolarization inactivation (C and D) effects are shown.
Neurons were included in this analysis only if 50 nA (or less) of AMPA
reproducibly increased firing by more than 20%, and if the
goodness-of-fit (r2) for a third order
polynomial of their ejection current-response relationship was >0.7. A
and C, population average ejection current-response curves. B and D,
Emax and ECur50 and as
determined from the individual ejection current-neuronal response
curves and then averaged (see Materials and Methods).
 , Emax for the rate enhancement induced
by AMPA was significantly different between the control and the
6-OHDA-treated hemisphere (t21 = 3.6, P < 0.05).
|
|
View larger version (39K):
[in this window]
[in a new window]
|
Fig. 3.
Responses of ventral pallidal neurons to NMDA was
enhanced after 6-OHDA-induced lesions; however, acute depletion of
monoamines did not alter the response. Rate enhancement (A and B) and
apparent depolarization inactivation (C and D) are shown. The criteria
for inclusion in these analyses and the data presentations are the same
as those defined in legend for Fig. 2. B, ANOVA for NMDA-induced rate
enhancement Emax:
F2,53 = 4.3, P < 0.025. a, Emax for the
NMDA-induced rate enhancement differed between control and the
6-OHDA-treated hemisphere (Newman-Keuls; q = 3.8, P < 0.05). b,
Emax differed between the reserpine-treated
rats and the 6-OHDA-treated hemisphere (q = 3.4, P < 0.05). D, ANOVA for NMDA-induced apparent
depolarization inactivation Emax:
F2,65 = 4.0, P < 0.025. c, Emax from 6-OHDA
and reserpine conditions differed (q = 4.0, P < 0.05). ANOVA for ECur50 of
NMDA-induced apparent depolarization inactivation:
F2,67 = 5.1, P < 0.025. d, ECur50 differed between control
and the 6-OHDA-treated hemisphere (q = 4.3, P < 0.05). e, significant
difference between the 6-OHDA-treated and the reserpine condition
(q = 3.5, P < 0.05). All other
comparisons were nonsignificant.
|
|
Effects of Chronic DA Depletion on Responses of VP Neurons to
EAA.
HPLC-EC evaluations of DA and its metabolites were used to
verify the extent of the 6-OHDA-induced lesions in the striatum, a
termination site for dopaminergic projections from the substantia nigra, and the olfactory tubercles, a termination of ventral tegmental dopaminergic neurons. In intact controls, DA content in the striatum was 99.9 ± 6.1 (all monoamine content data are in nanograms per milligram of wet tissue), 3,4-dihydroxyphenylacetic acid was 6.7 ± 0.4, and homovanillic acid was 2.5 ± 0.1. After 6-OHDA
treatments, DA and its metabolites were reduced by 99%. Norepinephrine
(intact control levels, 5.2 ± 0.3) was reduced by 98%. Serotonin
(2.5 ± 0.2 in intact controls) remained unchanged in lesioned
striata, and consistent with a compensatory increase in serotonin
turnover, its metabolite 5-hydroxyindoleacetic acid (1.9 ± 0.1)
was increased by 74%. DA levels in the olfactory tubercle of intact
controls were 46.8 ± 2.4, and despite receiving part of its DA
innervation from the ventral tegmental area, it demonstrated a profound
reduction (by 97%) of DA after the nigrally targeted 6-OHDA
treatments. Therefore, on the day when electrophysiological experiments
were conducted (i.e., 21-28 days post 6-OHDA treatment), near-total depletions of DA and its major metabolites remained for both major ascending dopaminergic projections.
The 6-OHDA treatment did not alter several of the measured
electrophysiological parameters. Baseline firing rate, cells per tract,
and action potential characteristics were unchanged (Table 1). The
proportion of neurons that responded to EAA iontophoresis was similar
to controls; AMPA enhanced firing in 94% (16/17) of neurons in the
lesioned hemisphere and in 100% (11/11) tested in pooled controls, and
NMDA increased firing in 88% (28/32) and 71% (20/28) of VP neurons
tested in lesioned hemispheres and controls, respectively
(2 = 0.05-2.01; N.S. for AMPA and NMDA). The potency
(ECur50) of AMPA (Fig. 2, A and B) and NMDA (Fig.
3, A and B) to induce rate increases was unchanged after 6-OHDA
treatments. In contrast, the Emax for
AMPA- and NMDA-induced rate enhancements was decreased by
6-OHDA-induced lesions (Figs. 2, A and B, and 3, A and B). But rather
than reflecting a decrease in the efficacy of the agonist to enhance
firing per se, this observation may reflect the ability of the EAA
agonists to induce apparent depolarization inactivation at lower firing
rates in the lesioned state.
Inactivation of spiking ensued within a few seconds of EAA application,
even when using moderate ejection currents that initially increased
firing rate. This biphasic response to a single ejection current level
is shown in Fig. 1, A and C. The phenomenon also is illustrated by the
fact that the right-most extreme of the EAA ejection current-response
curves (Figs. 2A and 3A), which show the peak rate enhancement usually
obtained in the first few seconds of EAA application (illustrated in
Fig. 1, A and B), has a greater spikes per second than the left-most
extreme of the apparent depolarization inactivation curves (Figs. 2C
and 3C, showing data collected 8-11 s after the onset of EAA ejection, as seen in Fig. 1C). The Emax and
Ecur50 for AMPA-induced apparent depolarization
inactivation were not altered by long-term DA deafferentation (Fig.
2D). This is in stark contrast to the apparent depolarization inactivation seen during NMDA application where
ECur50 was reduced in the lesioned condition
(Fig. 3D). The results suggest that after chronic lesions of
DA-containing neurons there is an enhancement in the potency for NMDA
(but not AMPA) to induce apparent depolarization inactivation effects
in VP neurons. Thus, NMDA was used for further characterization of VP
neuronal responses after DA deafferentation.
Effects of Acute DA Depletion on Responses of VP Neurons to
NMDA.
The change in the response of VP neurons to the application
of NMDA could be due to the loss of the direct dopaminergic input, or
to adaptations in the circuitry that includes the VP as the system
attempts to compensate for the loss. To examine whether similar changes
occur with acute DA removal, the sensitivity of VP neurons to NMDA was
examined within hours after rats were treated with reserpine. As
indicated by HPLC-EC assays of tissue harvested at the conclusion of
the electrophysiological evaluations, the reserpine treatment
essentially depleted DA (98% reduction) in the striatum, but did not
alter levels of the DA metabolites 3,4-dihydroxyphenylacetic acid or
homovanillic acid. Reserpine depleted striatal norepinephrine (by
100%) and serotonin (by 92%), but 5-hydroxyindoleacetic acid remained
similar to controls. Olfactory tubercle monoamine content displayed a
similar pattern after reserpine treatment. The data show that the
reserpine treatment protocol provided an acute removal of brain
monoamine content that was similar in extent to that obtained with the
6-OHDA treatment. However, in contrast to the 6-OHDA-induced lesions,
the reserpine treatment did not result in compensatory alterations in
DA or serotonin turnover (suggested by the stable levels of DA and
serotonin metabolites).
The firing rate and action potential duration of VP neurons from
reserpine-treated rats were greater than in controls; effects that were
not seen in 6-OHDA-treated rats (Table 1). Eighty-nine percent of the
37 VP neurons recorded from reserpinized rats showed rate increases to
NMDA. This response distribution was comparable to that seen 21 to 28 days after 6-OHDA treatments (2 = 0.03, N.S.) and to
controls (2 = 1.6, N.S.). The
Emax for the rate-enhancing effects of
NMDA in reserpinized rats differed from that seen in the 6-OHDA state but not that from controls (Fig. 3, A and B). Following the same trend,
the Emax and
ECur50 for the apparent depolarization
inactivation observed in reserpinized rats were different from the
chronic lesion state but not from controls (Fig. 3, C and D). These
findings indicate that acute removal of monoaminergic inputs enhances
baseline firing rate but does not alter responses to NMDA, whereas
chronic removal 1) makes VP neurons shift more readily into apparent
depolarization inactivation upon exposure to NMDA so that the peak
rate-enhancing effects can not be obtained, and 2) renders NMDA more
potent in inducing this inactivation.
Effects of GABA on Ability of NMDA to Produce Rate Enhancement and
Apparent Depolarization Inactivation.
To further evaluate the
ability of NMDA to evoke the above-described responding patterns, we
conducted two sets of experiments. The first experiment tested the
hypothesis that if VP neurons were unable to generate action potentials
because of a persistent depolarization induced by NMDA then the
addition of a hyperpolarizing agent such as GABA would be able to
reinstate spiking (Grace and Bunney, 1986; Henry et al., 1992;
Hollerman et al., 1992; White et al., 1995). To make this
determination, four VP neurons recorded from three intact rats were
tested. All of these neurons demonstrated a robust decrease in
spontaneous firing rate with GABA and an increase in firing with NMDA
(Fig. 4A). Additionally, in all four neurons, suprathreshold ejection currents of GABA (i.e., those that
decreased spontaneous firing) reversed the decreases in spike amplitude
and number of detected spikes caused by excessive NMDA receptor
stimulation (Fig. 4). Thus, consistent with prior literature (Grace and
Bunney, 1986; Henry et al., 1992; Hollerman et al., 1992; White et al.,
1995), the fact that GABA was able to enhance VP neuronal firing rate
when coapplied with NMDA, rather than inhibiting it, provides strong
evidence that the observed NMDA-induced rate decreases were due to
depolarization inactivation.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 4.
GABA reverses the apparent depolarization
inactivation induced by NMDA. Rate histograms from three different
neurons are shown (A-C). Horizontal lines above the histograms
illustrate the time of drug application by using the ejection currents
indicated. The time scale in A and B is the same as that shown in C. A,
action potentials obtained at different points in the experiment are
shown under the histogram with the minimal spike voltage required for
detection indicated by the horizontal line. Although GABA alone
decreased firing rate, the action potential remained intact. NMDA (40 nA) initially increased firing rate, but with sustained application
this current level was sufficient to induce a substantial decrease in
the action potential height and a concomitant decrease in the
detectable spikes per second. Coapplication of GABA, at an ejection
current that when used alone suppressed firing (15 nA), was able to
attenuate the effects of NMDA both on the action potential and the
firing rate. Termination of the ejection current to both NMDA and GABA
allowed the action potential to return to its baseline configuration. B
and C, two additional ventral pallidal neurons demonstrating the same
phenomenon, i.e., GABA coiontophoresis was able to reverse the apparent
depolarization inactivation caused by NMDA.
|
|
To determine whether the enhanced potency of NMDA to induce this
apparent depolarization inactivation after chronic DA removal was
paralleled by an enhancement in the GABA reinstatement profile, a
second experiment was performed in 14 unilaterally lesioned rats. In
each of the neurons that were sensitive to both GABA and NMDA
(n = 14), an ejection current of GABA below that
necessary to induce a significant change in firing rate (<20%) was
determined. This "subthreshold" ejection current was used for
further testing to avoid the possible confound of GABA-induced rate
decreases on the effects seen when it was coiontophoresed with NMDA.
Indeed, in the contralateral control, the rate-enhancing (Fig.
5, A and B) and apparent depolarization
inactivation (Fig. 6, A and B) effects of
NMDA in VP neurons were unchanged by coiontophoresis with subthreshold
GABA. In contrast, GABA generally had a restorative effect on
NMDA-induced rate enhancement and apparent depolarization inactivation
in the lesioned hemisphere (Figs. 5 and 6, A and B), making the
responses more like those seen in controls. Thus, the
ECur50 of NMDA-induced apparent depolarization
inactivation in the lesioned hemisphere when coiontophoresed with GABA
was not different from NMDA alone in control (Fig. 6, B versus D; Emax). (Note: the ECur50
of NMDA done in the lesioned hemisphere was lower than
control; see Fig. 3D). There also was a trend to alter the NMDA
efficacy of the apparent depolarization inactivation response with GABA
coapplication in the lesioned state; however, this trend did not reach
statistical significance at = 0.025 (see Materials and
Methods; Fig. 6D). These data indicate that the increase in
apparent depolarization inactivation seen with NMDA in the lesioned
hemisphere is attenuated by a hyperpolarizing influence on these
neurons.
View larger version (33K):
[in this window]
[in a new window]
|
Fig. 5.
Effects of subthreshold applications of GABA on the
rate-enhancing effects of NMDA in the control condition (A and B), and
in the 6-OHDA-treated hemisphere (C and D). The criteria for inclusion
in these analyses and the data presentations are defined in the legend
for Fig. 2. Comparing NMBA-induced responses with and without GABA in
controls and in 6-OHDA-treated hemispheres for both
Emax and ECur50 revealed that
the ECur50 () for the NMDA rate-enhancing effect in the
lesioned hemisphere was significantly increased after coiontophoresis
with GABA (F3,44 = 3.5, P < 0.025; q = 4.2, P < 0.05). Remaining comparisons were not
significant.
|
|
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 6.
Subthreshold applications of GABA on the apparent
depolarization inactivation seen with NMDA. A and B, NMDA-induced
apparent depolarization inactivation in the control condition. C and D,
NMDA-induced apparent depolarization inactivation in the 6-OHDA-treated
condition. The criteria for inclusion in these analyses and the data
presentation are defined in the legend for Fig. 2. Comparing
NMBA-induced responses with and without GABA in controls and in
6-OHDA-treated hemispheres for both Emax and
ECur50 revealed that the ECur50 for the
NMDA-induced apparent depolarization inactivation in the lesioned
hemisphere was significantly increased after coiontophoresis with GABA
(F3,44 = 6.8, P < 0.025; q = 5.0, P < 0.05).
Other comparisons among these four treatment conditions were not
different.
|
|
Effect of Chronic Removal of DA on Function of Tonically Active
Endogenous GABA Transmission.
To investigate whether chronic DA
depletion changes the sensitivity of VP neurons to GABA or the tonic
influence of endogenous GABA on VP neurons, two lines of investigation
were pursued. The response of VP neurons to iontophoretic GABA was
determined, and the effect of blocking endogenous GABA with the
GABAA receptor antagonist bicuculline was measured.
GABA decreased the firing rate in all of the VP neurons tested in both
the lesioned hemispheres (n = 10 cells) and in the contralateral control hemispheres (n = 6), often
resulting in near-complete cessation of firing with the higher ejection
currents. As shown in Fig. 7A, the
magnitude of the GABA ejection current was directly related to the
magnitude of the response (control: slope 
0.47 ± 0.01 spikes/s/nA, F1,14 = 1219, P < 0.05; lesion: slope 0.43 ± 0.03. F1,12 = 175, P < 0.05) and the treatment-effect curves in the lesioned and unlesioned
conditions were superimposable. Thus, chronic removal of dopaminergic
neurons did not alter the ability of GABA to decrease VP cell firing.
View larger version (29K):
[in this window]
[in a new window]
|
Fig. 7.
6-OHDA treatments did not alter the ability of GABA
to suppress, or bicuculline to enhance, ventral pallidal neuron firing.
The criteria for inclusion in these analyses and the data presentation
are defined in the legend for Fig. 2. For these evaluations, the
control condition was the hemisphere contralateral to the
6-OHDA-injected hemisphere. A, GABA ejection current-firing rate
relationships for the control and 6-OHDA-treated conditions were
superimposable. B, blockade of the GABAA receptor with
bicuculline produced an equivalent increase in firing rate in the
control and lesioned states. This observation indicates that there is
no change in tonic GABA release in the VP of the DA-depleted
hemisphere.
|
|
Blockade of GABAA receptors with bicuculline
resulted in an increase in neuronal firing rate, indicating that there
is a tonic level of GABA released on these VP neurons that acts to
decrease neuronal firing rate. However, the response to bicuculline by neurons in the VP of the lesioned hemisphere was not different from
that obtained in controls (Fig. 7B). Thus, it appears that the level of
tonic GABA release in the lesioned hemisphere remains functionally unaltered.
| |
Discussion |
This report revealed that a consequence of chronic lesions of the
brain's ascending dopaminergic neuronal system is an enhancement in
the effects of activating NMDA receptors within the VP. This included a
substantial increase in the ability of the agonist to disrupt normal
action potential generation in VP neurons. When considered in the
context of the adaptive processes that may be relevant to PD, there are
several salient features that warrant close examination.
None of the action potential characteristics often used to identify
neuronal subpopulations (e.g., action potential waveform, duration, or
amplitude) or indices of basal spiking activity (i.e., frequency of
encountering spontaneously firing neurons, firing rate, or firing
pattern) were altered in the VP of 6-OHDA-treated hemispheres. These
findings distinguish this more limbically oriented VP from the
adaptations that occur after chronic DA depletion in either one of the
pallidal structures that serve as basal ganglia outputs, i.e., the
GPi/entopeduncular nucleus or the GPe/globus pallidus, which,
respectively, show rate increases (Filion and Tremblay, 1991) and
decreases (Pan and Walters, 1988; Filion and Tremblay, 1991). One
mechanism posed for these differential adaptations is a divergence in
the effects of chronic DA removal on striatal GABAergic outputs. That
is, the GABAergic projection to the GPi/entopeduncular nucleus is
thought to be diminished, whereas an enhancement occurs in GABAergic
influences on the GPe/globus pallidus (Gerfen et al., 1990). However,
we did not observe any difference in the effects of tonically released
GABA on VP neurons in the lesioned and unlesioned conditions (evidenced
by similar responding to bicuculline). Because the nucleus accumbens
provides the major GABAergic input to the VP (Walaas and Fonnum, 1979;
Groenewegen and Russchen, 1984), these findings concur with reports
showing a lack of changes in nucleus accumbens firing rate after
lesions of midbrain dopaminergic neurons (Debonnel and De Montigny,
1988). Thus, it may be that adaptations in the limbic system differ
from those observed in the basal ganglia.
Acute depletion of brain monoamines (by reserpine treatment) also did
not alter the number of spontaneously firing cells, but in contrast to
VP effects of chronic lesions, it did lead to an enhancement in
spontaneous firing rate. One possible mechanism for this finding may
involve interactions among local (VP) neurotransmitters. The VP is
directly innervated by DA-containing terminals (Voorn et al., 1986),
and local applications of DA induce robust changes in VP neuronal
firing rate (Napier and Potter, 1989; Napier et al., 1991;
Maslowski-Cobuzzi and Napier, 1994). DA also normally regulates the
effects of several VP transmitters, including inhibiting glutamate-induced excitations (Johnson and Napier, 1997a). Subthreshold levels (so that spontaneous firing is not altered) of exogenously applied or endogenously released DA are capable of attenuating the
effects of exogenously applied or endogenously released glutamate (Maslowski-Cobuzzi and Napier, 1994; Johnson and Napier, 1997a), including the EAA released upon activation of the amygdala
(Maslowski-Cobuzzi and Napier, 1994), a source of EAA to the VP. The VP
also receives EAA-containing inputs from the STN (Groenewegen and
Berendse, 1990; Turner et al., 2001), a brain region that is known to
demonstrate increases in activity and changes in firing pattern in the
DA-deafferented state (Bergman et al., 1994). This enhanced drive to
the VP would no longer be modulated by DA and such a scenario may
render VP neurons more susceptible to various toxic effects of
increased EAA input. Thus, it is plausible that acute removal of DA
modulatory influences allowed for a larger excitatory drive to be
exerted onto those neurons that are innervated by EAA-containing
afferents. However, because the increase in firing rate is no longer
seen 3 weeks after 6-OHDA treatments, it appears that the engaged
adaptive processes, which do not involve GABA transmission, are
sufficient to counter the effects of acute DA removal.
Even though indices of basal firing were unchanged in the VP after
chronic DA-deafferentation, VP responding to EAA was greatly altered.
The most profound effect occurred in the ability of EAAs, especially
NMDA, to diminish spike amplitude to the point of rendering it
indistinguishable from background activity. This response profile is
consistent with the theory of a depolarization blockade, a phenomenon
well characterized in other brain regions of intact rats (Bunney and
Grace, 1978; Grace and Bunney, 1986; Henry et al., 1992; Hollerman et
al., 1992; White et al., 1995; Hu and White, 1996). This theory
maintains that with excessive exposure to rate-enhancing agents,
neurons enter a depolarized, but inactive state. The depolarized nature
of the membrane during this inactive state is verified by demonstrating
that spiking can be restored with hyperpolarizing agents (Bunney and
Grace, 1978; Grace and Bunney, 1986; Henry et al., 1992; Hollerman et
al., 1992; White et al., 1995). Concurring with this dogma, we
demonstrated that when VP spiking was no longer detected during
excessive microiontophoretic application of NMDA, coapplication of GABA
at levels that normally suppress firing, reinstated spiking.
The mechanisms underlying the apparent depolarization inactivation
induced acutely by EAA are not yet known, but alterations in
Na+ and/or Ca2+ conductance
are likely candidates (Herrling et al., 1983; Lambert et al., 1989;
Cepeda et al., 1991; Ciardo and Meldolesi, 1991). With intracellular
recordings of striatal neurons, microiontophoretic applications of NMDA
initiates spiking within a few hundred milliseconds of ejection current
onset, and this is followed by a depolarization plateau upon which
spikes with a decreased amplitude occur (Herrling et al., 1983; Cepeda
et al., 1991). The initial burst of action potentials is likely
mediated by Na+ and the longer lasting (several
seconds) depolarized plateau is dependent on Ca2+
currents (Herrling et al., 1983; Lambert et al., 1989; Cepeda et al.,
1991; Ciardo and Meldolesi, 1991). In the present study, an apparent
depolarization blockade was also observed with AMPA. Some AMPA receptor
channels are permeable to Ca2+ and AMPA receptor
activation promotes the opening of voltage-dependent Ca2+ channels (Choi, 1992); thus, such effects
would be expected. An involvement of long-lasting
Ca2+ plateaus with an acute depolarization
inactivation is consistent with the response profile seen with EAA
application on VP neurons. That is, with either continued EAA exposure
or the use of higher ejection currents, more and more
Ca2+ channels may be recruited to increasingly
summate the plateau current. This would result in the observed response
continuum of an initial decrease in spike amplitude to a complete
blockade where spiking no longer occurred.
Persistent alterations in such membrane parameters and/or intracellular
handling of the influx of cations may underlie the enhanced propensity
of EAA to cause an apparent depolarization inactivation in VP neurons
in the 6-OHDA condition. Grace and Bunney (1986) demonstrated in rats
chronically treated with neuroleptics that dopaminergic neurons
recorded in vivo exist in a depolarized state, and these neurons more
readily enter into a depolarization block upon subsequent neuroleptic
administration. Likewise, Tseng et al. (2001) have shown in vivo in
rats chronically depleted of DA, that spontaneously active striatal
medium spiny neurons spend less time in the "down" state
(hyperpolarized) relative to unlesioned controls, both the "up" and
down states are relatively depolarized, and neurons in the up state are
much more likely to fire action potentials (although such effects were
not seen in vitro by Calabresi et al., 1993). If VP neurons are
depolarized at rest in the chronic DA-deafferented condition, this
would make them more susceptible to effects of EAAs. But a persistent
depolarization alone may not always be sufficient or necessary for the
observed effects. For example, Mulder et al. (1996) saw profound
changes in responses of nucleus accumbens neurons to iontophoretically applied glutamate in vivo despite finding no changes in intrinsic membrane properties (determined in vitro). It also is possible that
alterations in the ability of VP neurons to process intracellular Ca2+ may be relevant to the enhanced effects seen
with EAA in the lesioned state. Calcium-binding proteins, like
calbindin, calretinin, and parvalbumin are important buffers that are
critical to maintaining appropriate intracellular concentrations of
Ca2+ (for review, see Heizmann and Braun, 1992).
Although still controversial, there is evidence that forebrain levels
of these proteins are decreased in PD (Iacopino and Christakos, 1990;
Heizmann and Braun, 1992), and that calbindin may be neuroprotective in
humans with PD and animal models of this disease (German et al., 1992).
If the capacity of VP neurons to buffer Ca2+ is
diminished in the 6-OHDA-lesioned state, the effects of the Ca2+ influx caused by excessive EAA receptor
activation would be enhanced. Moreover, as observed in the present
study for measures of basal spontaneous firing in the VP of the
chronically lesioned hemisphere, the enhancement in EAA-induced effects
could occur in the absence of changes in basal levels of neuronal activity.
It has been demonstrated in several other brain regions that a
consequence of excessive EAA is excitotoxic damage and cell death
(Choi, 1992); however, this remains to be fully investigated for the
VP. The possibility that there is a physiological change, or neuronal
damage or loss, in the VP is supported by reports showing alterations
in cognitive tasks (Rammsayer and Classen, 1997) and affective states
(Cummings, 1992) in PD patients, for these functions are influenced by
the VP (learning, Chrobak et al., 1991; motivation, Mogenson and Yang,
1991; and reward, McAlonan et al., 1993; Panagis et al., 1995; Gong et
al., 1997). Such effects may explain part of the limbic circuitry
changes underlying the affective and cognitive symptoms of PD.
We thank Dr. William Wolf for the use of HPLC-EC equipment.
This work was supported by U.S. Public health Service Grant
MH11607 (to M.S.T. and T.C.N.), and by the M.D./Ph.D. and Neuroscience graduate programs at Loyola University Chicago.
EAA, excitatory amino acid;
PD, Parkinson's
Disease;
DA, dopamine;
STN, subthalamic nucleus;
GPi, internal segment
of globus pallidus;
VP, ventral pallidum;
NMDA, N-methyl-D-aspartate;
AMPA, -amino-3-hydroxy-5-methylisoxazole-4-propionic acid;
GABA, -aminobutyric acid;
6-OHDA, 6-hydroxydopamine;
HPLC-EC, high-performance liquid chromatography with electrochemical detection;
ANOVA, analysis of variance;
ISI, interspike interval;
CNQX, 6-cyano-2,3-dihydroxy-7-nitroquinoxaline;
AP-5, 2-amino-5-phosphopentanoic acid;
GPe, external segment of globus
pallidus.
![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Top
Abstract
Introduction
and kappa opioid receptors in the ventral pallidum.
J Pharmacol Exp Ther
272:
1260-1270
