Download Supplementary Methods

Supplementary Information
Supplementary materials and methods
Subjects
Male Sprague-Dawley rats (280-300 g) were used. Rats were single housed under a 12 hr
reverse dark/light cycle (on 20h00, off 8h00). Temperature (22 ± 1°C) and humidity (60 ± 5%)
were also controlled.
Drugs
Cocaine HCl (Coopération Pharmaceutique Française, Bordeaux, France) and Gentamicine
(Gentalline®) were dissolved in 0.9% NaCl. Picrotoxin and LY 379268 were from SIGMA (St.
Quentin Fallavier, France) and Tocris (Bristol, UK), respectively. DNQX and AP-5 were from
Tocris or Ascent Scientific (Bristol, UK).
Self-administration studies
Surgery
A silastic catheter (internal diameter = 0.28 mm; external diameter = 0.61 mm; dead volume
= 12μl) was implanted in the right jugular vein under ketamine (100 mg/kg) + xylazine (1
mg/kg) anaesthesia. Rats were allowed to recover for 5 to 7 days after surgery. During the
first 4 days following surgery, rats received an antibiotic treatment (gentamicine 1 mg/kg i.p.).
After surgery, catheters were flushed daily with a saline solution containing unfractionated
heparin (100 IU/ml).
Intravenous Self-administration (SA)
Self-administration apparatus. The SA setup consisted of 48 SA chambers made of Plexiglas
and metal (Imetronic, Pessac, France). Each chamber (40 cm long x 30 cm width x 52 cm
high) was located within an opaque box equipped with exhaust fans that assured air renewal
and masked background noise. Briefly, each rat was placed daily in a SA chamber and the
intracardiac catheter connected to a pump-driven syringe (infusion speed: 20 μl/sec). Two
holes, located at opposite sides of the SA chamber placed at 5 cm from the grid floor, were
used to record instrumental responding. A white house light at the top of the chamber
allowed its complete illumination. A white cue light (1.8 cm in diameter) was located 9.5 cm
above the active hole. A green cue light (1.8 cm in diameter) was located 10 cm right to the
white cue light. A blue cue light (1.8 cm in diameter) was located on the wall opposite to the
one containing the active hole at 33 cm from the floor on the left side. Experimental
contingencies were controlled and data collected with a PC Windows-compatible software
(Imetronic, Pessac, France).
1
Intravenous self-administration (SA) procedures
Basal training protocol. The daily SA session was composed of three drug components (40
min each) separated by two 15 min periods during which responding in the active hole had
no scheduled consequences (no drug periods). Drug periods were signaled by the blue cue
light, while the no drug periods were signaled by illumination of the entire SA box and
extinction of the blue cue light. During the drug periods, introduction of the rat's nose into one
hole (active) turned on the white cue light located above it and then, 1 sec later, switched on
the infusion pump. The cue light remained on for a total of 4 sec. Nose-pokes in the other
hole (inactive) had no scheduled consequences. The self-infusion volume was 40 μL (2 sec
infusion) and contained 0.8 mg/kg of cocaine. Each infusion was followed by a time-out
period of 40 sec. During the first 6 days, a Fixed Ratio 3 (FR3) schedule of reinforcement
was applied. Then, the FR was increased to 5 for the rest of the experiment. Criterion for
acquisition of cocaine SA was defined by a stable number of self-infusions over at least three
consecutive SA sessions (± 10%).
Addiction-like criteria. Cocaine use severity is assessed using three procedures described
below and resembling one or more of the seven DSM-IV criteria for addiction1: (i) high
motivation for the drug, measured by a progressive ratio schedule of reinforcement, (ii)
persistence in drug seeking even if the drug is not available, measured by active responding
during periods of signaled drug non-availability and (iii) drug use despite negative
consequences, measured by resistance to footshock-induced punishment during cocaine
SA.
Progressive-ratio schedule. The ratio of responses per infusion was increased after each
infusion according to the following progression (10, 20, 30, 45, 65, 85, 115, 145, 185, 225,
275, 325, 385, 445, 515, 585, 665, 745, 835, 925, 1025, 1125, 1235, 1345, 1465, 1585), the
last ratio completed, called the breakpoint, was used to measure motivation for cocaine. The
session ceased after either 5 hours or when a period of 1 hour had elapsed since the
previously earned infusion.
Persistence of drug seeking in the absence of cocaine. This test allows assessing the
subject's difficulty to control drug seeking and drug use. It was assessed daily by measuring
the responses in the active hole during the two no drug periods of the basal training protocol.
Active responses during the two daily no drug periods were summed. For analysis, the mean
responses over two to three consecutive sessions were used as dependent variables.
Resistance to punishment: During this session rats were placed for 40 min in the SA
chamber. The blue cue light signaling drug availability was on. The schedule was the
2
following: The first response led to the illumination of the green cue light signaling the
presence of the shock. After 3 additional responses, rats received an electric footshock (0.2
mA, 1 sec), and after the 5th response, rats received both an electric footshock (0.2 mA, 1
sec) and a cocaine infusion (0.8 mg/kg), associated with the corresponding conditioned
stimulus (white cue light). Then the green cue light was turned off. The schedule reinitiated at
the end of the time-out period, i.e. 40 sec after the infusion. If, within a minute, rats did not
complete response requirements leading to shock and shock plus infusion respectively, the
green cue light turned off and the sequence was reinitiated, i.e. the following response turned
on the green cue light. The number of active responses earned during this test was
expressed in percentage of the number of active responses performed, over the same period
of time, during the basal training session conducted the day before.
Evaluation of cocaine use severity based on the three addiction-like criteria
Establishment of Addict-like and non Addict-like groups
Analysis of the distributions of each of the three addiction-like behaviors revealed that the
distribution of resistance to punishment might be bimodal, composed of a first log-normal
distribution and a second normal sub-distribution2 while distribution of the motivation for the
drug and the persistence of drug seeking are best fitted by a log-normal regression2
(Supplementary Figure S1). The bimodal quality of distributions is highly depending upon the
challenging nature of the test. In this context, it is pertinent that only resistance to
punishment might distribute bi-modally and can also be log-normal (see Supplementary
Figure S1) as it can be influenced by factors such as the time at which the test is performed
and whether rats have experienced the test after early training or not. The latest the test is
conducted, as well as previous experience with it, favor the two subpopulations to diverge.
This is confirmed by the progression in the three scores from early to late training in the two
groups.3 Anyway, this bimodal distribution provides us an objective criterion to determine a
threshold in the population in order to carry out a categorical strategy to identify animals that
show addiction-like behavior, i.e., 30-40% highest part of the population.
Based on these observations, our approach consists then in ranking rats for their scores for
each of the three addiction-like tests and considering a rat positive for an addiction-like
criterion when its score for this behavior was in the 35% highest percentile of the distribution.
Four groups of rats are isolated depending on the number of positive criteria met (0crit, 1crit,
2crit and 3crit). 0 and 3 criteria rats are defined as Non Addict-like and Addict-like animals
respectively. 1crit and 2crit rats either represent specific stable cocaine “use-related troubles”
or intermediate steps toward addiction.
3
Importantly, as previously discussed,3, 4,the scores in the three addiction-like behaviors are
linearly related to the number of addiction-like criteria met (Supplementary Figure S5). In
addition, an arbitrary change of the selection threshold from 25% to 40% had very little effect
on the addiction-like endophenotype and the linear distribution of the scores as a function of
the number of criteria met (data not shown). The validity of our categorical approach (number
of positive criteria) is then strongly supported by the dimensional approach (scores for each
criterion).
Addiction score
It was calculated as the algebraic sum of standardized scores of each of the three addictionlike criteria. Standardization consisted in subtracting the mean of the group to each individual
score and then dividing this number by the standard deviation. This procedure yields scores
which have a mean of 0 and a standard deviation of 1. The addiction score is thus distributed
along a scale from -3 to 3.2, 4
Evaluation of vulnerability to addiction-like behavior
Our research goal is to study the mechanisms underlying transition to addiction. It requires to
conduct longitudinal studies from early (before addiction occurs) to late (when addiction
expresses) cocaine use. In case of invasive techniques, such as ex vivo electrophysiology,
obviously rats cannot be followed from early to late training. In this context one of our
objectives has been to find early behavioral features (after 15 to 20 days of SA) that allow
predicting transition in a reliable and consistent manner.
We focused on the addiction criterion that we measure during every basal SA session, i.e.
persistence in active responses during the no drug periods. We identified A meta-analysis
was performed on two previous experiments conducted on 71 rats. The protocol applied in
these experiments was similar to the one described by Deroche-Gamonet and coworkers3.
This analysis, previously detailed5, allows a behavioral identification of future Addict-like rats
(future 3crit) as early as 17 days of SA, based on responding during the no drug periods. 5
After early training, in average, 3crit rats show a particular pattern of responding in the active
hole during the no drug periods. From session 1 to session 17, they tend to progressively
increase responding during these periods; the increase tending to be more pronounced from
the early to the late minutes of the periods, as if they were progressively anticipating the next
drug period (Supplementary Figure S2). Four variables characterizing responding during the
early no drug periods have been identified which, when combined, allow predicting transition
to addiction. These variables are: 1. the number of active responses during the last 9
4
minutes averaged over sessions 15 to 17 (Last 9min D15-D17); 2. The number of active
responses during the first 6 minutes averaged over sessions 15 to 17 (First 6min D15-D17);
3. The difference between the first and the second variables (Last9min D15 to D17 First6min D15 to D17); 4. The difference between the number of active responses during the
last 9 minutes averaged over sessions 15 to 17 and the number of active responses during
the last 9 minutes averaged over sessions 1 to 5 (Last9min D15 to D17 - Last 9min D1 to
D5). Rats were ranked for these four variables. Rats with a score in the 20th highest
percentile of the population were defined as positive for this predictive criterion. Rats with 4
positive criteria are called 3crit predicted rats.
Importantly, as shown in Supplementary Figure S3a, the prediction accuracy of this strategy
was the best we could expect, i.e. 100% of rats classified as 3crit predicted become addicts.
However, as shown in Supplementary Figure S3b, the prediction potency, i.e. the ability of
this strategy to identify all future addicts, was of 53%. 47% of addicted rats did not show this
early extreme pattern of responses during the non-drug periods. Interestingly enough,
however, is the fact that the 3crit predicted rats exhibit a higher addiction score than the 3crit
that were not predicted (Supplementary Figure S3c). The 3crit rats identified by our strategy
are therefore probably the most vulnerable and the earliest to reveal. For what concerns 0crit
rats, prediction accuracy was not as potent as for 3crit rats, but was still close to 60%.
Indeed, 58 % of the 0c predicted were effectively diagnosed as non addicts after 80 days of
self-administration (Supplementary Figure S3a). For what concerns prediction potency, it
was of only 30%. 70% of the 0crit rats were not early identified by our strategy
(Supplementary Figure S3b). Importantly, whether they were predicted or not, 0crit rats did
not differ by the addiction score (Supplementary Figure S3c).
Active responses during the no drug periods of Resistant and Vulnerable rats tested in the
present study are represented on Supplementary Figure S4. As expected, Vulnerable rats
progressively increased active responses during the no drug periods from sessions 1 to 17
(Supplementary Figure S4a) while Resistant rats did not. This increase was due to a
progressive increase in responses as the 15 minutes of the no drug periods elapse, as
shown for the 17th session on Supplementary Figure S4b.
Electrophysiological Studies
Slice preparation. These methods were described in details previously.5, 6 Briefly, rats were
anesthetized with a mixture of ketamine (100 mg/kg) + xylacine (1 mg/kg) and decapitated.
The brain was sliced (300 μm) in the coronal plane using a vibratome (Integraslice, Campden
Instruments, Loughborough, UK). Slices were maintained in a sucrose based physiological
5
solution at 4° C (in mM: 87 NaCl, 75 sucrose, 25 glucose, 5 KCl, 21 MgCl2, 0.5 CaCl2 and
1.25 NaH2PO4). Immediately after cutting, slices were stored for 40 min at 32°C in an
artificial cerebrospinal fluid (ACSF) (in mM): 130 NaCl, 11 Glucose, 2.5 KCl, 2.4 MgCl2, 1.2
CaCl2, 23 NaHCO3, 1.2 NaH2PO4, and were equilibrated with 95% O2/5% CO2. Slices
were then stored in ACSF at room temperature until recording. Recordings were made
between 1 and 6 hours after cutting. Slices were placed in the recording chamber and
superfused (1.5 - 2 ml/min) with ACSF and left to equilibrate for 10 minutes before recording.
All experiments were performed at ~30°C. The superfusion medium contained picrotoxin
(100 μM) to block GABAA receptors. All drugs were added at the final concentration to the
superfusion medium.
Patch Clamp and Extracellular Recording Procedures: Whole cell patch-clamp
recordings were performed from visualized deep layer pyramidal neurons located in the
prelimbic area of the PFC. Glass electrodes (resistance 4-6 MOhms) were filled with either
Cesium Methane-Sulfonate or K+Gluconate based solutions, as follows (mM): 128 Cesium
Methane-Sulfonate or K+Gluconate, 20 NaCl, 1 MgCl2, 1 EGTA, 0.3 CaCl2, 2 Na2+ATP, 0.3
Na+ GTP, buffered with 10 Hepes, pH 7.3, osmolarity 290-300 mOsm. In order to evaluate
the access resistance (Ra), a 2 mV hyperpolarizing pulse was applied before each EPSC.
Ra was not compensated and cells were rejected if Ra was >25 MOhms or changed >20%
during the experiment. The potential reference of the amplifier was adjusted to zero prior to
breaking into the cell. Data was recorded with Axopatch-1D or Axopatch-200B amplifiers
(Molecular Devices, Sunnyvale, USA), filtered at 1-2 kHz, digitized at 10 kHz on a DigiData
1332A interface (Molecular Devices, Sunnyvale, USA) and collected on a PC using Clampex
(9.2 or 10). Analysis was performed using Clampfit (9.2 or 10)(Molecular Devices,
Sunnyvale, USA). For field excitatory post-synaptic potential (fEPSP), extracellular recording
electrodes were filled with ACSF. To evoke synaptic currents, stimuli (100 μsec duration)
were delivered at 0.1 Hz with a glass electrode filled with ACSF and placed in layer 2/3 as
previously described.7
LTD recordings: After 10 minutes of stable baseline EPSC recording, eCB-LTD was
induced stimulating layer 2/3 afferents during 10 minutes at 10 Hz 7. The magnitude of eCBLTD was estimated from EPSC recorded after 20-30 minutes LTD induction as percentage of
baseline EPSC amplitudes. mGluR2/3-LTD was induced, after a stable baseline recording of
fEPSP, by applying the specific mGluR 2/3 agonist LY 379268 (100 nM) for 10 minutes. The
magnitude of mGlu2/3-LTD was estimated from fEPSP recorded after 25-30 minutes of
washing out the drug. At the end of the experiments, the glutamatergic nature of the fEPSP
was confirmed from its block with the AMPA receptor antagonist DNQX (20 M).
6
AMPAR/NMDAR ratio: EPSC were evoked while holding cells at +40 mV. The AMPARmediated EPSC was isolated after bath application of the NMDAR antagonist D-2-amino-5phosphonovaleric acid (D-APV, 100 M). The NMDAR-mediated EPSC was obtained by
digital subtraction of the AMPAR EPSC from the dual (AMPAR +NMDAR-mediated) EPSCs.
Immunoblotting studies
Protein extraction from rat prefrontal cortex (PFC): Protein extracts from rat dorsomedial PFC
(including prelimbic, cingulate and the medial part of the secondary motor cortices)
containing protease and phosphatase inhibitors (Sigma, USA) were prepared using the
homogenizer Precellys 24 (Bertin Technologies, France) as previously described 8. Briefly,
the homogenizing protocol in RIPA buffer also containing protease and phosphatase
inhibitors was 5000 rpm 2x30 sec + 10 sec break using ceramic CK14 beads (Bertin
Technologies, France).
Immunoblotting analysis: Proteins suspended in Laemmli buffer were separated by SDSPAGE (10% gels) and transferred onto PVDF membranes (Immobilon P, #IPVH00010,
Millipore, USA). The proteins were revealed as previously described8, 9 with affinity purified
rabbit anti-mGluR2/3 antibody (#PPS078; 1/1000 dilution, R&D Systems, Inc., USA). The
Neuronal Class III -tubulin (III-tubulin) monoclonal antibody (#MMS-435P; 1/1000 dilution)
used as a loading control was purchased from Eurogentec (Seraing, Belgium). The X-Ray
films (Kodak, USA) were then quantified by densitometry using a GS-800 scanner (in
transmission mode) and the associated Quantity One software (Bio-Rad, CA, USA)
according to the manufacturer’s instructions.
Experimental timeline
First experiment: Differences between Addict-like and non Addict-like rats. Five to 7 days
after surgery, rats were either tested for cocaine SA or left undisturbed (Controls) in the
animal house. SA sessions were performed between 10:00 and 16:00 hrs. Two independent
groups of animals were used, the first to perform electrophysiological studies (cocaine SA
N=96; controls N=11) and the second to carry out western blot analysis (cocaine SA N=96;
Controls N=5).
A progressive ratio was conducted on sessions 40 and resistance to punishment was tested
on sessions 45. The mean of the total active responses during the non-drug periods of
sessions 37 to 39 were also considered. Cocaine use severity was evaluated based on the
scores to these three addiction-like criteria.
For electrophysiological studies, five additional basal training sessions were performed
7
before the animals started to be tested for the expression of synaptic plasticity.
Electrophysiological experiments were carried out 24 hours after their last SA session. Non
Addict- (N=8), Addict-like (N=8) and age-matched control rats were alternatively tested (one
rat per day). Basal SA training was maintained for all animals until they were tested for
electrophysiology. Self-administering rats (Addict- and Non Addict-like) were tested for LTD
between the 50th and 72nd self-administration sessions. Persistence in responding during
the no drug periods was followed on a daily basis to control for the maintenance of the non
Addict/Addict-like status. Over these 22 days, 8 0crit, 8 3crit and 6 control rats were
alternated. To minimize differences in cocaine exposure between the first and the last tested
self-administering rats, the 5 additional controls were tested for LTD, before or after the first
and last self-administering rats were tested.
For western blot analysis, 5 Addict- and 8 Non Addict-like rats were selected. Their addiction
score were 2.5 ± 1.03 and -1.99 ± 0.29 respectively, i.e. comparable to the scores of Addictlike and Non Addict-like rats tested for synaptic plasticity.
Second experiment: Rats were trained for cocaine SA during 17 sessions according to the
basal training protocol. They were studied for synaptic plasticity 24 hours after the last SA
session and after being classified as Vulnerable or Resistant to addiction-like behavior. This
procedure was repeated twice in two independent group of rats (N=44 in each). mGluR2/3LTD was studied in Vulnerable (N=6) and Resistant (N=6) rats from the first group.
AMPA/NMDA ratio and eCB-LTD was analyzed in Vulnerable (N=6) and Resistant (N=5) rats
from the second group. Rats of matching age and purchase, left undisturbed in the animal
house, were used as Controls (N=14). In order to obtain the same time point within groups
the start of the SA sessions of each animal was scheduled appropriately.
Statistics
All values are given as mean ± S.E.M. For SA data, one-way or repeated measures analysis
of variance (ANOVA) was used to determine possible group effects and interactions
[experimental groups (Controls, Addict-like, Non Addict-like, Vulnerable and Resistant to
addiction-like behavior) were used as between-subject factors]. For electrophysiological
experiments, N corresponds to the number of individual cells analyzed, with at least 5 rats
included in each condition. To determine whether LTD was successfully induced in any given
group, a paired Wilcoxon signed rank test was performed comparing the mean EPSC or
fEPSP amplitude during baseline and 20-30 minutes after LTD induction. To evaluate group
differences in LTD induction, LTD was expressed by the post-induction response in % of
baseline response. For both LTD and AMPA/NMDA ratio, group differences were assessed
8
with a one-way ANOVA. The Newman Keuls post hoc test was used for pair wise
comparison of means in both SA and electrophysiological experiments. Pearson’s correlation
analysis was used to assess the relationship between behavior and synaptic markers.
Statistical tests were performed with GraphPad Prism (GraphPad Software Inc., La Jolla, CA,
USA) for the electrophysiological data and Statistica 6.0© (StatSoft Inc, Tulsa, OK, USA) for
the SA data and the correlation analysis. A critical probability of p < 0.05 was applied.
Legends to Supplementary Figures
Supplementary Figure S1. Distribution of each addiction-like criterion. Data are
represented as number of observations for each range of scores. The whole population
(n=81) was log-normal distributed for motivation for cocaine (measured by the breakpoint
during a progressive ratio session) (K-S, d=0.09858, p=n.s., Khi2=2.50519, dl=4,
p=0.64371), inability to refrain from cocaine seeking when it was not available and signalled
as so (non reinforced active nose-pokes) (K-S, d=0.12703, p=n.s., Khi2=1.12696, dl=2,
p=0.56922) and resistance to punishment (maintenance of drug use and drug seeking
despite contingent electric foot-shocks as measured by active nose-pokes as a percentage
of baseline) (K-S, d=0.09268, p=n.s., Khi2=1.94839, dl=3, p=0.58319).
Supplementary Figure S2. Examples of the pattern of responding in the no drug
periods from sessions 1 to 29 in future non Addict- and Addict-like rats. Results are
expressed as mean responding per minute over the two daily no drug periods. Rats were
trained for cocaine SA over 60-70 SA sessions and classified as 3crit and 0crit. Persistence
in responding was retrospectively analyzed. After early training (starting around sessions 1113), future 3crit rats tend to progressively increase responding during the last part of the no
drug periods. Animals were then classified in Vulnerable and Resistant for addiction-like
behavior on the basis of four variables measuring the pattern of responding during the no
drug periods between day 15 and day 17 of SA.
Supplementary Figure S3. Accuracy and potency of the prediction of vulnerability and
resistance to addiction-like behavior. a. Prediction accuracy of the strategy to early
identify future Addict-like rats. All the rats predicted as 3crit rats became Addict-like. For 0crit
rats, the prediction accuracy was 58%. b. Prediction potency. Our strategy identified 53% of
all future Addict-like rats, whereas 30% of all future Non Addict-like was isolated. c. Addiction
score calculated after 50-70 cocaine SA sessions. 3crit-predicted rats exhibit a higher
addiction score than the 3crit rats that had not been predicted (*p<0.05). Conversely,
predicted or not 0crit rats did not differ by the addiction score.
9
Supplementary Figure S4. Early identification of future Addict-like rats. a. Persistence
in responding although the drug is signalled as unavailable from session 1 to 17 in rats
identified as Vulnerable and Resistant to transition to addiction-like behavior. Data are
expressed as the average total active responses over the two no drug periods of the basal
training protocol. As expected, Vulnerable rats significantly differ from Resistant rats during
sessions 15-17. b. Persistence in responding during the last session (SA day 17): Time
course over the 15 minutes (5 min time blocks). Data are expressed as the average total
active responses over the two no drug periods of the last SA session. Vulnerable rats show a
typical progressive increase in responding from the early to the late phase of the no drug
period.
Supplementary Figure S5. Scores in the three addiction-like criteria in Non Addict-like
(0crit) and Addict-like (3crit) rats. As compared to 0crit rats, 3crit rats (n=8) displayed a
higher motivation to take the drug. The breakpoint during a progressive ratio session was
used as an index of motivation to take the drug. They also show a high level of active nosepokes than 0crit rats (n=16) when the drug is signalled as not available (non drug periods).
They maintained cocaine taking and seeking despite its adverse consequences at higher
level. This DSM-IV criterion was measured by the level of responses maintained (resistance
to punishment) by rats when an electric shock is delivered contingently to the fourth and fifth
lever press of the FR5 instrumental sequence.
Supplementary Figure S6. In the pPFC, short-term facilitation is not affected by
cocaine. Paired-pulse facilitation (PPF) is a form of short-term synaptic plasticity of
presynaptic origin thought to correlate with the basal probability of neurotransmitter release.
We compared synaptic responses in deep layers pyramidal pPFC neurons to two
consecutive stimuli delivered at 20 Hz. The PPF was computed as the ratio between the
amplitudes of the second EPSC over the first EPSC. We found that PPF was similar between
Controls (1.66 ± 0.11, N=15), Non Addict-like (1.74 ± 0.06, N=7) and Addict-like rats (1.98 ±
0.27, N=8). This suggests that the basal presynaptic properties of excitatory synapses in the
PFC are unchanged during the transition to cocaine addiction.
References
1.
American Psychiatric Association. Diagnostic and Statistical Manual of Mental Disorders.
4th ed. revised edn. American Psychiatric Press: Washington, DC, 2000.
2.
Belin D, Berson N, Balado E, Piazza PV, Deroche-Gamonet V. High-novelty-preference rats
are predisposed to compulsive cocaine self-administration. Neuropsychopharmacology
10
2011; 36(3): 569-579.
3.
Deroche-Gamonet V, Belin D, Piazza PV. Evidence for addiction-like behavior in the rat.
Science 2004; 305(5686): 1014-1017.
4.
Belin D, Balado E, Piazza PV, Deroche-Gamonet V. Pattern of intake and drug craving
predict the development of cocaine addiction-like behavior in rats. Biol Psychiatry 2009;
65(10): 863-868.
5.
Kasanetz F, Deroche-Gamonet V, Berson N, Balado E, Lafourcade M, Manzoni O et al.
Transition to addiction is associated with a persistent impairment in synaptic plasticity.
Science 2010; 328(5986): 1709-1712.
6.
Kasanetz F, Manzoni OJ. Maturation of excitatory synaptic transmission of the rat nucleus
accumbens from juvenile to adult. J Neurophysiol 2009; 101(5): 2516-2527.
7.
Lafourcade M, Elezgarai I, Mato S, Bakiri Y, Grandes P, Manzoni OJ. Molecular components
and functions of the endocannabinoid system in mouse prefrontal cortex. PLoS One 2007;
2(8): e709.
8.
Revest JM, Kaouane N, Mondin M, Le Roux A, Rouge-Pont F, Vallee M et al. The
enhancement of stress-related memory by glucocorticoids depends on synapsin-Ia/Ib. Mol
Psychiatry; 15(12): 1125, 1140-1151.
9.
Revest JM, Di Blasi F, Kitchener P, Rouge-Pont F, Desmedt A, Turiault M et al. The MAPK
pathway and Egr-1 mediate stress-related behavioral effects of glucocorticoids. Nat
Neurosci 2005; 8(5): 664-672.
11