N co
A n
T de
Ni , Ma
M
Ba euro
me itory
(Ce
Me exce
alc hole
GA ed in
of
Re info
gra onde
GA t of i
act
Co sign
tar uron
ne pend
Ke
for
A
pre
consequences, compulsive alcohol-seeking behavior, loss of con-
tro
alc
sta
pro
inv
str
ute
wi
see
ex
me
dra
co
po
are
he
rel
pe
Ne
dan
spe
am
be
Recently, it hasbeen shown thatNPY suppresses alcohol drinking in
Fro
Ad
Rec
00
do
l limiting intake, andanegative emotional state in theabsenceof
ohol (1). Amajor goal of basic research on alcoholism is to under-
nd the neural underpinnings of alcohol use and the pathological
gression to alcohol dependence (2).
The transition from casual drinking to alcohol dependence
olves numerous neuroadaptive changes in brain reward and
ess systems. The recruitment of brain stress systems contrib-
s to the aversive aspects of alcohol withdrawal and, along
th dysregulation of brain reward systems, promotes alcohol-
king behaviors and relapse (3). Brain stress systems in the
tended amygdala (4) appear to be especially important in
diating motivational deficits associated with alcohol with-
wal and excessive alcohol drinking. The extended amygdala
ntains the central nucleus of the amygdala (CeA), the lateral
rtion of the bed nucleus of the stria terminalis, and a transition
a in the nucleus accumbens shell. These three regions are
avily interconnected and have high concentrations of stress-
ated neuropeptides.
rats (10) via its actions in CeA (11–14). More specifically, NPYmicro-
injection into the CeA exhibits an enhanced ability to suppress
alcohol drinking in certain subpopulations of drinkers, including
rats that are made dependent on alcohol via vapor inhalation.
In vitro electrophysiological studies have revealed that acute
alcohol facilitates spontaneous and evoked GABAergic transmis-
sion in the CeA via presynaptic and postsynaptic mechanisms (15).
In CeA neurons from rats exposed to chronic alcohol vapor, base-
line GABAergic transmission is increased and CeA neurons do not
exhibit tolerance to the facilitatory effects of acute alcohol (16).
Recent data have highlighted the effects of various neuropeptides
onalcohol-induced facilitationof inhibitory transmission in theCeA
of rats. For example, corticotropin-releasing factor (CRF) facilitates
(17) and nociceptin opposes (18) GABAergic transmission in the
CeA.
In thepresent study,we tested thehypothesis thatNPYhas a key
role in the development of excessive alcohol drinking during the
transition to alcohol dependence. We also examined the effects of
NPY on baseline GABA type A (GABAA)-mediated inhibitory trans-
mission in the CeA and the interactions of NPY and acute alcohol in
alcohol-naive and alcohol-dependent rats, as well as the synaptic
mechanisms and NPY receptor subtypes responsible for these ef-
fects.We report thatNPYadministrationblocks thedevelopmentof
excessive alcohol drinking associated with the transition to alcohol
dependence and that NPY opposes alcohol effects on GABAergic
transmission inCeA, likely via activationofpresynaptic Y2 receptors.
m the Committee on the Neurobiology of Addictive Disorders, The
Scripps Research Institute, La Jolla, California.
dress correspondence to Nicholas W. Gilpin, Ph.D., The Scripps Research
Institute, Committee on Neurobiology of Addictive Disorders, SP30-
2400, 10550 N Torrey Pines Road, La Jolla, CA 92037; E-mail: nickg@
scripps.edu.
eived Nov 29, 2010; revised Feb 4, 2011; accepted Feb 7, 2011.
BIOL PSYCHIATRY 2011;69:1091–109906-3223/$36.00
i:10.1016/j.biopsych.2011.02.004 © 2011 Society of Biological Psychiatry
europeptide Y Opposes Al
minobutyric Acid Release i
ransition to Alcohol Depen
cholas W. Gilpin, Kaushik Misra, Melissa A. Herman
arisa Roberto
ckground: During the transition to alcohol and drug addiction, n
diate aspects of withdrawal and relapse via convergence on inhib
A).
thods: This study investigated the role of neuropeptide Y (NPY) in
ohol vapor inhalation. This study also utilized intracellular and w
BAergic inhibitory transmission in CeA, synapticmechanisms involv
alcohol-naive and alcohol-dependent rats.
sults: Chronic NPY treatment blocked excessive operant alcohol-re
dual increases in alcohol responding by intermittently tested n
BAergic transmission and reversed alcohol-induced enhancemen
ions at presynaptic Y2 receptors.
nclusions: These results highlight NPY modulation of GABAergic
get for the treatment of alcoholism. Gamma-aminobutyric acid ne
uromodulator systems recruited during the transition to alcohol de
y Words: BIBP3226, BIIE0246, central amygdala, negative rein-
cement, Y1 receptor, Y2 receptor
lcoholism, or alcohol dependence, is a progressive and
chronically relapsing disorder. The development of alcohol-
ism is characterized by frequent episodes of intoxication,
occupation with alcohol and the use of alcohol despite adverse
hol Effects on Gamma-
Amygdala and Blocks the
nce
ureen T. Cruz, George F. Koob, and
modulator systems in the extended amygdala are recruited to
gamma-aminobutyric acid (GABA) neurons in central amygdala
ssive alcohol drinking by making rats dependent on alcohol via
-cell recording techniques to determine the effects of NPY on
these NPY effects, and NPY interactions with alcohol in the CeA
rced responding associated with alcohol dependence, as well as
pendent control animals. Neuropeptide Y decreased baseline
nhibitory transmission in CeA by suppressing GABA release via
aling in central amygdala as a promising pharmacotherapeutic
s in the CeA likely constitute a major point of convergence for
ence.
Several neuropeptides have prominent roles in the aversive as-
cts of alcohol withdrawal and relapse via their actions in the CeA.
uropeptide Y (NPY) is an inhibitory peptide produced in abun-
ce in the hypothalamus and phylogenetically conserved across
cies (5). Neuropeptide Y is highly co-localized with gamma-
inobutyric acid (GABA) in the amygdala (6), which is important
cause NPY reduces anxiety (7) via actions in the amygdala (8,9).
Methods andMaterials
Be
Kin
sur
pla
ho
an
sio
gu
Use
bo
(19
rat
an
lev
ing
ad
com
res
lize
can
me
mi
inf
an
[aC
(no
veh
ven
de
No
im
(21
tim
rec
(Ha
the
lef
ad
(PE
Ha
de
the
mu
ho
gro
at
vap
occ
rat
Lo
tes
we
10
tha
dri
consumption (grams ethanol per kilograms body weight), and eth-
anol preference (ethanol consumed/total fluid consumed) using
thr
wh
po
wit
11,
alc
ual
fes
ho
sta
Ele
or
pre
(12
cha
cut
no
30
fus
aC
po
sul
ate
sol
tia
ne
8h
sha
ing
mV
pC
ing
me
sup
din
ph
nis
par
toc
the
ten
1 to
at 5
firs
fro
tud
sur
sup
mi
on
Po
ron
fer
Ca
ter
ron
1092 BIOL PSYCHIATRY 2011;69:1091–1099 N.W. Gilpin et al.
ww
havioral Studies
Animals. We used 42 adult male Wistar rats (Charles River,
gston, New York). The average body weight before vapor expo-
e was 541.7 � 21.4 g. Animals were single-housed in standard
stic cages with wood chip bedding under a 12-hour light/12-
ur dark cycle (lights off at 8:00 AM) with ad libitum access to food
d water throughout except during experimental drinking ses-
ns. All procedures were conducted in the dark cycle andmet the
idelines of theNational Institutes of HealthGuide for the Care and
of Laboratory Animals (National Research Council, 1996).
Operant Alcohol Self-Administration Training. Operant
xes and training procedures have been described previously
). Briefly, we trainedWistar rats to respond on a continuous fixed
io-1 schedule for .1 mL deliveries of supersaccharin (3% glucose
d .125% saccharin; see [20]) versus water in a concurrent, two-
er, free-choice contingency during 30-minute sessions. Follow-
these operant training sessions, 10% (wt/vol) ethanol was
ded and sweeteners gradually removed from the solution. Upon
pletion of this fading procedure, we allowed rats 25 operant
ponding sessions for 10% (wt/vol) ethanol versus water to stabi-
operant responding. Rats were stereotaxically implanted with
nulae and subsequently divided into four groups based on
an intakes during four postsurgery prevapor operant self-ad-
nistration sessions: 1) chronic alcohol vapor exposure and NPY
usion (dependent-NPY,n� 7); 2) chronic alcohol vapor exposure
d vehicle infusion (dependent-artificial cerebrospinal fluid
SF]; n � 8); 3) chronic air vapor exposure and NPY infusion
ndependent-NPY; n� 7); and 4) chronic air vapor exposure and
icle infusion (nondependent-aCSF; n� 8).
Stereotaxic Surgeries. We surgically implanted intracerebro-
tricular (ICV) cannulae using aseptic procedures as previously
scribed (11), after isoflurane anesthesia (Abbott Laboratories,
rth Chicago, Illinois). A guide cannula (22 gauge)was unilaterally
planted according to the appropriate stereotaxic coordinates
), with a dummy cannula (28 gauge) in the guide cannula at all
es except during infusions. We monitored rats for a 1-week
overyperiod todetermine that animals resumednormal activity.
Microinfusions. We used a Harvard 33 microinfusion pump
rvard Apparatus, Holliston, Massachusetts) for all infusions into
lateral ventricles (ICV) at a rate of 2.5 �L/min for 2 minutes and
t the injection cannula in place 1 additional minute to allow for
equate diffusion.We delivered infusions via polyethylene tubing
20; Becton Dickinson & Co, Glencoe, Maryland) connected to a
milton 25 �L syringe (Hamilton Company, Reno, Nevada).
Operant Tests During Alcohol Vapor Exposure. We infused
pendent and nondependent rats with NPY or vehicle and tested
m for operant behavior at the 6-hour withdrawal time point on
ltiple days during the first 15 days of chronic intermittent alco-
l vapor exposure (CIE) (Supplement 1). Rats in the dependent
ups were exposed to alcohol vapor 14 hours per day (vapor off
8:00 AM); rats in the nondependent groups were exposed to air
or 24 hours per day. Operant tests and NPY infusions never
urred on the same day. On days 2, 4, 6, 8, 10, 12, and 14 of CIE,
s received either .0 �g or 10.0 �g NPY ICV (Sigma-Aldrich, St.
uis, Missouri) in 5.0 �L aCSF. On days 3, 7, 11, and 15 of CIE, we
ted rats for operant alcohol responding. On day 16 of CIE, rats
re sacrificed and cannula placements histologically verified. The
.0 �g NPY dose was chosen based on prior studies that showed
t acute ICV infusion of this dose reliably suppresses ethanol
nking in rats (10,12).
Statistical Analysis. We analyzed operant responses, ethanol
w.sobp.org/journal
ee-way repeated-measures (RM) analyses of variance (ANOVAs),
ere between-subjects factors were vapor exposure (alcohol va-
r vs. air vapor) and NPY dose (.0 �g vs. 10.0 �g) and day was the
hin-subjects factor (baseline vs. operant tests on vapor days 3, 7,
and 15). Because a priori differences were not expected early in
ohol vapor exposure,we also analyzedoperant test days individ-
ly to determine when vapor effects and/or NPY effects mani-
ted during the transition to alcohol dependence. Wemade post
c comparisons using the Student-Newman-Keuls test and set
tistical significance at p� .05.
ctrophysiological Studies
Slice Preparation. Following �2 weeks of CIE (14 hours/day)
air vapor exposure (Supplement 1), we prepared CeA slices as
viously described (14,15). Briefly, male Sprague-Dawley rats
0–300 g; 4–7 weeks old, n � 57) were removed from vapor
mbers, anesthetized with halothane (3%), and decapitated. We
400-�m coronal slices on a Vibratome Series 3000 (Leica, Ban-
ckburn, Illinois), incubated them in an interface configuration for
minutes, then completely submerged and continuously super-
ed (flow rate of 2–4 mL/min) them with warm (31°C), gassed
SFof the following composition inmmol/L: sodiumchloride, 130;
tassium chloride (KCl), 3.5; sodium phosphate, 1.25; magnesium
fate heptahydrate, 1.5; calcium chloride, 2.0; sodium bicarbon-
, 24; glucose, 10. Drugs were added to the aCSF from stock
utions to obtain known concentrations in the superfusate.
Intracellular Recording of Inhibitory Postsynaptic Poten-
ls in CeANeurons. We recorded in alcohol-free aCSF from CeA
urons (n� 64) of alcohol-naive or alcohol-dependent rats for 2 to
ours after cutting, aspreviouslydescribed (15).We recordedwith
rpmicropipettes (3mol/L KCl) using current-clampmode, hold-
potentials near resting membrane potential (mean Vm � �77
). Data were acquired with an Axoclamp-2A preamplifier and
LAMP software (Axon Instruments). We used a bipolar stimulat-
electrode to evokepharmacologically isolatedGABAA receptor-
diated inhibitory postsynaptic potentials (IPSPs), while
erfusing the slices with the glutamate receptor blockers 6,7-
itroquinoxaline-2,3-dione (10 �mol/L) and DL-2-amino-5-phos-
onovalerate (30 �mol/L) and the GABA type B receptor antago-
t CGP 55845A (1 �mol/L). To determine the IPSP response
ameters for each cell, we performed an input-output (I/O) pro-
ol (14,15). The stimulus strengths were maintained throughout
duration of the experiment. We normalized three stimulus in-
sities of equal steps (threshold, half-maximal, and maximal) as
3X.
We examined paired-pulse facilitation (PPF) using paired stimuli
0-millisecond interstimulus intervals (15). The amplitude of the
t IPSP was 50% of themaximal stimulus strength, as determined
m the I/O relationship. We calculated the PPF ratio as the ampli-
e of the second IPSP divided by that of the first IPSP. All mea-
es were taken before the first drug superfusion (control), during
erfusion (15–20 min per drug), and following washout (�15
n). To avoid tachyphylaxis, we superfused each drug only once
to a single cell.
Whole-Cell Patch-Clamp Recording of Miniature Inhibitory
stsynaptic Currents in CeA Neurons. We visualized CeA neu-
s in brain slices (350–400 �m) using infrared differential inter-
ence contrast and CCD camera (EXi Aqua, QImaging, Surrey, BC,
nada).Weused aw60water immersionobjective (Olympus, Cen-
Valley, Pennsylvania) for identifying and approaching CeA neu-
s. Whole-cell voltage-clamp recordings were made with a Mul-
ticlamp 700B amplifier (Molecular Devices, Sunnyvale, California),
low
lec
10
pu
ne
L):
2; H
5’-t
uo
con
pe
�m
(1
mi
cla
tw
pro
.05
err
cor
a lo
we
pa
sen
Re
Be
tra
(i.e
fro
vap
a g
alc
att
tio
CIE
rat
an
CIE
ho
ma
wa
an
inf
15
do
an
F (1
(p
eff
Tw
tes
op
5.8
NP
.07
for
exp
alc
thr
tha
F (1
chr
do
rev
in a
Ele
(15
no
de
cre
fig
to
cou
no
IPS
�m
Fig
van
hol
Chr
ind
also
(n �
wit
(ind
on
sup
vap
arti
of N
pen
N.W. Gilpin et al. BIOL PSYCHIATRY 2011;69:1091–1099 1093
-pass filtered at 2 kHz to 5 kHz, digitized (Digidata 1440A; Mo-
ular Devices), and stored on a personal computer using pCLAMP
software (Axon Instruments). Patch pipettes (4-8M�) were
lled from borosilicate glass (Warner Instruments, Hamden, Con-
cticut) andfilledwith an internal solution composedof (inmmol/
KCl, 145; ethyleneglycol tetraacetic acid, .5;magnesiumchloride,
EPES, 10; adenosine 5’-triphosphate disodium salt, 2; guanosine
riphosphate sodiumsalt, .2. Series resistance (�10M�)wascontin-
usly monitored with a 10 mV hyperpolarizing pulse. Drugs were
stituted inaCSFandappliedbybath superfusion. Recordingswere
rformed in the presence of 6,7-dinitroquinoxaline-2,3-dione (10
ol/L), DL-2-amino-5-phosphonovalerate (30�mol/L), CGP 55845A
�mol/L), and tetrodotoxin (1 �mol/L) to isolate GABAA receptor
niature inhibitory postsynaptic currents (mIPSCs). All cells were
mped at�60mV for the duration of the recording.
Statistical Analysis. Data were analyzed with one-way and
o-way between-subjects ANOVA or RM ANOVA and, when ap-
priate, with the Student-Newman-Keuls post hoc test, with p�
considered statistically significant. Because there were unequal
or variances betweengroups and therewas a significant positive
relation betweenmeans and variance of I/O data, we performed
g-transformation on all I/O data before analysis. In some cases,
used independent-samples or paired-samples t tests for com-
ring individual pairs of means. Means reported in figures repre-
t measurements at the end of particular drug infusion periods.
sults
havioral Studies
Effects of Alcohol Vapor Exposure onOperant Behavior. To
ck the effects of alcohol vapor on operant behavior over time
., the transition to alcohol dependence), we analyzed data only
m vapor-exposed and air-exposed aCSF-infused rats. Two-way
or history�dayRMANOVAs indicated that over time, therewas
lobal increase in alcohol responding, F (3,57)� 3.62, p� .05, and
ohol consumption (g/kg), F (3,57) � 4.32, p � .01, which was
ributable to the gradual increase in responding and consump-
n by CIE vapor-exposed rats (Table S1 in Supplement 1). The
-exposed rats consumed more ethanol (g/kg) than air-exposed
s across days, F (1,19)� 4.45, p� .05. The ethanol response rates
d consumption quantities observed following 11 and 15 days of
exposure (in aCSF control animals) reliably produce blood-alco-
l levels of approximately 100 mg/dL in alcohol-dependent ani-
ls (19). There were no effects of dependence history or day on
ter responding in aCSF-treated rats (p .05).
Effects of NPY on Operant Behavior. Figure 1A shows oper-
t alcohol responding by CIE-exposed and air-exposed rats
used chronically with NPY or aCSF and tested across the first
days of vapor exposure. A three-way (vapor exposure � NPY
se� day) RM ANOVA revealed a significant suppression of oper-
t alcohol responding across days by chronic NPY infusions,
,38)� 6.23, p� .017. There was no three-way interaction effect
.05), but therewas a significant dependence� day interaction
ect on operant alcohol responding, F (3,114) � 3.60, p � .016.
o-way (vapor exposure � NPY dose) ANOVAs for each of the 4
t days (vapor days 3, 7, 11, 15) revealed that NPY suppressed
erant alcohol responding on day 15 of vapor exposure, F (1,38)�
4, p� .05, and that there was a tendency toward a suppressive
Y effect on alcohol responding on days 3 (p� .06) and 11 (p�
) of vapor exposure.
A separate two-way (vapor exposure � NPY dose) RM ANOVA
cumulative alcohol responding on days 11 and 15 of vapor
osure (Figure 1B) indicated a significant suppression of operant
ohol responding by NPY, F (1,38) � 5.84, p � .021. Finally, a
ee-way (vapor exposure�NPYdose�day) RMANOVA revealed
t alcohol-dependent rats responded less for water across days,
,38)� 7.82, p� .008, but water responding was not affected by
onic NPY infusions (p .05). Two-way (vapor exposure � NPY
se) ANOVAs for each of the 4 test days (vapor days 3, 7, 11, 15)
ealed no significant effects of NPY onwater responding (p .05
ll cases).
ctrophysiology Studies
NPY and Ethanol Effects on IPSPs. As previously described
), baseline GABAA-IPSP input-output curves were higher (data
t shown) and basal PPF ratio of IPSPs was lower in slices from
pendent rats relative to naive control rats, suggesting an in-
ased GABAergic tone via GABA release (statistics below and in
ure captions).
We first applied ethanol alone, then concomitantly applied NPY
CeA slices from vapor-exposed rats and naive control rats (time
rse in Figure 2A). As previously demonstrated (14), acute etha-
l (44mmol/L) produced robust 35%to45% (p� .001) increases in
P amplitudes (Figure 3A,B). Subsequent application of NPY (.5
ol/L) returned IPSPs to baseline levels (p� .025 difference from
ure 1. Chronic neuropeptide Y (NPY) administration in a clinically rele-
t treatment regimen blocks gradual and cumulative elevations in alco-
drinking similarly in alcohol-dependent and nondependent rats. (A)
onic NPY treatment blocks the development of alcohol dependence-
uced increases in alcohol responding (n�11/vapor-exposedgroup) and
blocks moderate increases in alcohol drinking by nondependent rats
10/air-exposed group) over time. Rats were infused in the ventricles
h NPY (10 �g) or vehicle on even-numbered days of vapor exposure
icated by arrows) and tested for responding at 6- to 8-hour withdrawal
days 3, 7, 11, and 15 of vapor exposure. (B) Chronic NPY treatment
presses cumulative alcohol responding across operant test sessions on
or days 11 and 15. *Significant (p� .05) suppression by NPY relative to
ficial cerebrospinal fluid vehicle regardless of vapor history (main