You are on page 1of 10

Behavioural Brain Research 239 (2013) 94103

Contents lists available at SciVerse ScienceDirect


Behavioural Brain Research
j our nal home page: www. el sevi er . com/ l ocat e/ bbr
Research report
Dorsolateral striatum and dorsal hippocampus: A serial contribution
to acquisition of cue-reward associations in rats
M. Jacquet
a,1
, L. Lecourtier
b,1
, R. Cassel
b
, M. Loureiro
b
, B. Cosquer
b
, G. Escofer
a
, M. Migliorati
a
,
J.-C. Cassel
b
, F.S. Roman
a
, E. Marchetti
a,
a
Aix-Marseille University, CNRS, NICN, UMR7259, 13344 Marseille, France
b
Laboratoire dImagerie et de Neurosciences Cognitives, UMR7237 Universit de Strasbourg-CNRS, Facult de Psychologie, 12 rue Goethe, F-67000 Strasbourg, France
h i g h l i g h t s
Rats with dorsolateral striatum lesions were impaired in procedural and declarative-like memories.
Rats with dorsal hippocampal lesions were impaired in declarative-like memory.
Rats with dorsal hippocampal lesions were not impaired procedural memory.
Establishment of a procedural may be a prerequisite for a declarative-like memory.
a r t i c l e i n f o
Article history:
Received 6 August 2012
Received in revised form25 October 2012
Accepted 29 October 2012
Available online 9 November 2012
Keywords:
Dorsolateral striatum
Dorsal hippocampus
Procedural memory
Reference memory
Learning
a b s t r a c t
In laboratory rodents, procedural and declarative-like memory processes are often considered operat-
ing in dual, sometimes even competing with each other. There is evidence that the initial approach
of a repetitive task rst engages a hippocampus-dependent declarative-like memory system acquiring
knowledge. Over repetition, there is a gradual shift towards a striatum-dependent response memory
system. In the current experiment, Long-Evans male rats with bilateral, ber-sparing ibotenic acid-
induced lesions of the dorsolateral striatum or the dorsal hippocampus were trained in an olfactory
associative task requiring the acquisition of both a procedural and a declarative-like memory. Rats with
dorsolateral striatum lesions, and thus an intact hippocampus, were impaired on both sub-categories of
memory performance. Rats with dorsal hippocampal lesions exhibited a substantial decit in learning the
declarative-like cue-reward associations, while the acquisition of the procedural memory component of
the task was not affected. These data suggest that the dorsolateral striatum is required to acquire the task
rule while the dorsal hippocampus is required to acquire the association between a given stimulus and
its associated outcome. The nding is that the dorsolateral striatum and the dorsal hippocampus most
probably contribute to successful learning of cue-reward associations in a sequential (from procedural
to declarative-like memory) order using this olfactory associative learning task.
2012 Published by Elsevier B.V.
1. Introduction
Damage to hippocampal and retrohippocampal structures in
rodents causes selective impairments ina variety of spatial andnon
spatial declarative-like learning tasks [14]. Such damage, how-
ever, was also shown to have little or no effects on other types of
learning [5,6], including olfactory associative learning [710]. An
example is provided by the effects of such lesions in the successive

Corresponding author at: Aix-Marseille Universit, CNRS, UMR7259, NICN, Cen-


tre St Charles-3, Place Victor Hugo, 13331 Marseille Cedex 03, France.
Tel.: +33 4 13 55 08 53; fax: +33 4 13 55 08 58.
E-mail address: evelyne.marchetti-gauthier@univ-amu.fr (E. Marchetti).
1
Contributed equivalently to this work.
go/no go olfactory discrimination task [11]. In this task, animals
are successively presented with individual odor stimuli, which are
paired (i.e., indicative of go trials) or not (i.e., indicative of no go
trials) with reinforcement. The prerequisite for successful learn-
ing of go/no go tasks after hippocampal damage is that the cues
are presented with a short inter-trial interval. Indeed, when rats
withhippocampal damage have tomanage the successive odor pre-
sentations with long inter-trial intervals (>15s), the odor-reward
association is not acquired [9,10]. These observations conrmed
earlier ndings in both rats and monkeys subjected to fornix or
hippocampal lesions [12]. Surprisingly, there are only a fewstudies
which aimed at identifying the brain regions involved in this type
of hippocampus-independent learning [13].
One candidate structure is the striatum, the rodent equivalent of
the caudateputamen in primates. In rats, lesions of the striatum,
0166-4328/$ see front matter 2012 Published by Elsevier B.V.
http://dx.doi.org/10.1016/j.bbr.2012.10.061
M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103 95
Table 1
Coordinates and injection volumes for ibotenate lesions of the dorsolateral striatumor the dorsal hippocampus. Coordinates are given in mmanterior and lateral to Bregma
(Paxinos and Watson, 1998) and fromthe skull surface in depth.
Structure Anteriority Laterality Depth Volume (L)
Dorsolateral striatum +0.5 3.9 5.2 0.4
+1.2 3.6 5.2 0.4
Dorsal hippocampus 2.6 1.0 3.8 0.06
3.1 3.0 2.8 0.06
3.0 1.4 3.2; 2.4 0.06 each
3.8 3.7 2.8 0.06
3.8 2.6 3.0; 2.5 0.06 each
4.1 4.0 3.2 0.08
4.1 2.5 2.5 0.08
and more specically of its dorsal regions, impair the acquisition
or/and retrieval of several operant tasks that are not or only poorly
affected by hippocampal lesions [2]. These include tasks based
on avoidance responses [1419], brightness discrimination [20],
processing of visual cues in the Morris water maze [21,22], and
visual/olfactory conditioning of emotional responses [23].
Our associative olfactory task [9] enables separate assessment
of performance linked to procedural or declarative-like memory
contributions. Using this task, a rst experiment assessed per-
formance after ber-sparing lesions of the dorsolateral striatum.
Given the outcome of this experiment (i.e., the impairment of
procedural learning prevented acquisition of the hippocampus-
dependent declarative-like component of the task), we performed
a complementary experiment in which the effects of dorsal hip-
pocampus lesions were assessed according to exactly the same
protocol. Following hippocampal lesions, we expected normal pro-
cedural learningbut noacquisitionof thehippocampus-dependent,
declarative-like component of the task. Our results suggest a serial
implication of the striatum and the hippocampus in cue-reward
associations learning.
2. Material and methods
2.1. Animals
Fifty two young adult Long-Evans rats (R. Janvier, France) weighing 250275g
at the time of surgery were used. They were individually housed and kept on a 24h
lightdark cycle (lights onat 7:00 amandoff at 7:00 pm) ina roomheldat a constant
temperature (22

). After surgery and before testing, each animal was rst handled
(10 min) daily for three consecutive days. Then it was deprived of water 48h before
the beginning of training. During training of the following days, the rats were given
water ad libitumfor 30min per day at 6:30 pm. European guidelines on procedures
for animal experimentation were followed, and all efforts were made to minimize
the animals suffering and to reduce the number of animals used while complying
with statistical constraints.
2.2. Lesion surgeries
All rats were anaesthetized with an intraperitoneal (ip) administration of pen-
tobarbital (60mg/kg, i.p.). Once anaesthetized, they were given an intramuscular
injection of an antibiotic (Extencilline; 0.3mL/kg) and were subsequently placed
on a stereotaxic apparatus. Their scalp was opened longitudinally with a scalpel
and little holes were drilled in the skull at locations where a needle had to be
introduced into the brain to perform the ibotenate infusions. The ibotenate solu-
tion (5g/L) was infused at the coordinates indicated in Table 1 at a speed of
0.25L/min. After each infusion, the needle was left in place for 4min before being
slowly retracted. In the sham-operated rats, the needle was descended into the
brain at each of the coordinates used for the lesion, but no infusion was performed.
After surgery, the scalp was sutured and all rats were placed in a heated cage until
complete recovery from anesthesia. Behavioral training was started 4 weeks after
surgery.
2.3. Apparatus and training procedure
The olfactory training apparatus was a rectangular box made of wire mesh
(30cm30cm50cm). A conical odor port (1.5cm in diameter, 0.5cm above the
oor) was drilled horizontally through a triangular wedge of Plexiglas, mounted
in one corner of the cage. A circular (1cm diameter) water port in the shape of a
well was placed directly above the odor port. Responses to the odor presentation
Fig. 1. Representative photographs illustrating lesions following NeuN staining in the groups of SHAM(A, C, E) and lesioned (B, D, F) rats. B and D showtypical examples of
lesions in the dorsolateral striatumand in the dorsal hippocampus, respectively. Dashed lines indicate the limits of the damaged area. E and F indicate that rostral regions of
the hippocampus were not damaged after lesions of the dorsal hippocampus (F) as compared to a sham-operated rat (E). Scale bars at the bottomright of photos B, D and F
correspond to 1M.
96 M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103
Fig. 2. Schematic representation of the smallest (light gray) and largest (dark gray) extent of dorsal hippocampal (left) and dorsolateral striatal (right) lesions in the rats kept
for statistical analyses. Numbers above the drawings indicate the distance (in mm) of each section fromBregma (according to Paxinos and Watson, 1998).
were monitored by a photoelectric circuit. Two ashlight bulbs which could be
turned on and off, as the testing conditions required, were placed outside the
cage, one on each side of the odor and water ports, 10cm above the oor. Indi-
vidual odors were delivered by forcing clean air (0.7 bars) through one of two
1 L Erlenmeyer asks that contained 500mL of water mixed with 0.2% of chemi-
cal odorants (jasmine vs. strawberry) from Sigma Inc. Non-odorized air, between
each odor presentation, was delivered by sending air through a ask that contained
only water. Odorized and clean air streams were passed individually through tubes
that were put through the back of the sound-attenuating chamber and attached to
the odor port. Water was delivered using a gravity-fed system, and passed through
a valve which, when opened, allowed 0.1mL of water to be released into the water
port.
All experiments were conducted simultaneously in four olfactory cages to
ensure that, in each experiment, representatives from each group were trained at
the same time, and thus under identical conditions [10,24]. Animals were trained
to make two odor-reward associations. Each odor had to be associated with a spe-
cic reward, one arbitrarily designated as positive and the other as negative. One
session was made of 60 trials using a successive Go or No-Go paradigm. Individual
trials (S
+
or S

) were run in a quasi-random fashion (no more than 3 consecutive


trials with the same valence). When the odor (S
+
) was delivered into the cage, if rat
M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103 97
responded by going to the water port a reward of 0.l mL of water was obtained. The
same response to delivery of the other odor (S

) resulted in no water and activation


of an error light. The maximum duration of odor presentation (S
+
or S

) was 10s.
Correct responses, therefore, corresponded to Go before the 10s presentation for
one odor had elapsed (S
+
) and No-Go to the other (S

). Once responded or not to


the odor presentation (the trial) a xed inter-trial interval (ITI) of 15seconds with
clean air was started. If a response was given during the last second of the ITI, the
next trial was delayed by 10s and delayed by additional 10s fractions as long as the
rat was still present during the last second of each fraction in the corner where the
reward is delivered. Consequently, the duration between two trials could last longer
than 15s and constituted the cumulative time to the minimum ITI. Animals were
tested every day between 08:00 am and 02:00 pm. Then, all animals were trained
to make associations with the odor pair for six sessions at a pace of one session per
day.
2.4. Histological verications
Upon termination of the behavioral experiments, all rats were deeply anes-
thetizedwithanoverdoseof pentobarbital (150mg/kgi.p.). Theyweretranscardially
perfusedwith60mL of phosphate-buffered4%paraformaldehyde (pH7.4; 4

C) over
5 min. The brains were extracted, postxed for 2h in the same xative (4

C), and
transferred into a 0.1M phosphate-buffered 20% sucrose solution for about 48h
(4

C). Brains were subsequently frozen in isopentane (30

C) and 40m-thick sec-


tions were cut in the coronal plane at 21

C using a freezing microtome and placed


in phosphate buffer saline. NeuN-staining was performed using a mouse polyclonal
antibody directed against NeuN (1/2000; Millipore, MAB377) as primary antibody
and a horse anti-mouse biotinylatedantibody (1/500; Vector, BA2001) as secondary
antibody.
2.5. Data analysis
Correct responses were characterized as Go for the positive odor and No-Go
for the negative one. Incorrect responses were Go for the negative odor and No-
Go for the positive one. The number of correct responses for both positive and
negative odors was expressed as a percentage of the total number of odor pre-
sentations (60 per session), thereby providing a global estimate of performance for
all groups (global memory). Training continued until a criterion of 805% correct
responses was reached by the control groups. This level of performance is required
to ensure that all animals have learned both associations. Latencies for positive (S
+
)
and negative (S

) odors were recorded and represented the time elapsed between


the beginning of a trial and its end (10s), when the rat responded to the odor
(reference memory, which is presumed to correspond by some aspects to declar-
ative memory in humans; [24]. When a rat did not respond, a latency of 10s was
scored. The mediancumulativetime (accountingfor the efciencyof some aspects of
procedural memory; [24]) was the number of seconds above the xed 15s ITI,
divided by the number of ITIs, which in this experiment amounted to 59. All beha-
vorial events were controlled and recorded by a computer.
Statistical analyses were performed with SPSS/PC+ statistics 11.0 software mar-
ketedbySPSS, Inc. All dataarepresentedas means SEM. Performancewas analyzed
using a repeated-measure MANOVA. Then, subsequent ANOVAs for each session
were computed. The threshold for signicance was set at p0.05. (on graphs:
*p 0.05; **p0.01; p*** 0.001). Percentage of correct responses in comparison
to chance level (50%) was analyzed using a one-sample t-test.
3. Results
3.1. Histological analysis
Typical examples of lesion extents and placements are shown
in Fig. 1. Lesion extent and location are further illustrated in Fig. 2.
Histological verications showed that 6 rats which had received
ibotenate into the dorsolateral striatum (DL STRIATUM; Experi-
ment 1) and 7 rats which had received ibotenate into the dorsal
hippocampus (D HIPP; Experiment 2) had lesions that were either
partly misplaced (e.g., encroaching too much onto the dorsome-
dial striatum, being slightly too dorsal, or showing asymmetry) or
of insufcient extent. These rats were withdrawn from statistical
analyses. Consequently, rats with dorsal hippocampus lesions con-
stituted a group of 9, those with dorsolateral striatum lesions a
group of 10. The respective groups of sham-operated rats (SHAM)
comprised 10 subjects.
3.2. Behavioral studies
3.2.1. Experiment 1: lesions of the dorsolateral striatum
Analyses of the percentage of correct responses (Fig. 3A)
showed that only rats of the SHAM group improved their overall
1 2 3 4 5
0
30
40
50
60
70
80
90
SHAM
DL STRIATUM
*
*
*
*
*
P
E
R
C
E
N
T
A
G
E

O
F

C
O
R
R
E
C
T

R
E
S
P
O
N
S
E
S
0 1 2 3 4 5
0
1
2
3
4
5
6
7
8
9
10
SHAM S+
SHAM S-
DL STRIATUM S+
DL STRIATUM S-
*
*
*
*
*
*
SESSION (Day)
L
A
T
E
N
C
Y

(
S
e
c
.
)
0 1 2 3 4 5
0
1
2
3
4
5
6
7
8
9
10
11
12
SHAM
DL STRIATUM
*
C
U
M
U
L
A
T
I
V
E

T
I
M
E

(
S
e
c
.
)
A
C
B
Fig. 3. Mean performance (SEM) obtained across the ve sessions of 60 trials
in control (SHAM) (N=10), and bilaterally dorsolateral STRIATUM lesioned rats
(DL STRIATUM) (N=10). (A) Mean percentage of correct responses. The difference
betweenthe SHAMandDL STRIATUMgroups was signicant fromthe fourthsession
onward. (B) Meancumulative time (inseconds). Signicant statistical difference was
observed between the two groups on session 5. (C) Mean latencies (in seconds), S
+
are the latencies for the positive odor and S

the latencies for the negative odor


one. Only the SHAM group started to make signicant associations from session 4
onward.
performance across the ve sessions (MANOVA, GroupSession
interaction: [F(4,68) =6.09; p<0.001]; the overall Group effect was
signicant [F(1,17) =9.18 p<0.01]. Subsequent ANOVAs revealed
a clear-cut difference between the two groups from the fourth
sessiononwards [F(1,17) =6.84p<0.01]. IntheDL STRIATUMgroup,
the percent correct responses was never different from chance
level, whatever the session (t 1.10, NS).
The cumulative time (Fig. 3B) decreased across sessions in
the SHAM group, whereas the DL STRIATUM group performed
at a constant level throughout all ve training sessions. Con-
sequently, a signicant difference was observed between the
performance of these two groups across the ve successive ses-
sions [F(1,17) =10.79 p<0.01]; the GroupSession interaction was
98 M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103
S1
0 1 2 3 4 5 6
0
5
10
15
20
C
U
M
U
L
A
T
I
V
E

T
I
M
E

(
s
e
c
.
)
S2
0 1 2 3 4 5 6
0
2
4
6
8
10
S3
0 1 2 3 4 5 6
0
2
4
6
8
10
S4
0 1 2 3 4 5 6
0
2
4
6
8
10
S5
0 1 2 3 4 5 6
0
2
4
6
*
*
*
*
*
*
*
BLOCK OF TRIALS
DL STRIATUM
SHAM
S=SESSION
Fig. 4. Mean cumulative time by block of 10 trials (SEM) obtained during sessions S1S5. Only the SHAMgroup showed a time that decreased over sessions, inversely to
DL STRIATUMgroup. A signicant difference was observed between SHAMand STRIATUMgroups in session 5. See also legend to Fig. 3.
signicant (F(4,68) =6.02 p<0.001), most probably because of the
substantial inter-group difference found on session 5 [(ANOVA,
F(1,17) =5.52 p<0.05].
Training performance analyzed in terms of S
+
and S

laten-
cies (Fig. 3C) showed a dramatic impairment in the DL STRIATUM
group. Indeed, converselytotheir SHAMcounterparts, lesionedrats
were unable to make correct associations on S
+
and S

stimuli.
In the SHAM group, correct associations were observed from the
fourth session onwards [F(1,18) 15.39; p<0.001]. The DL STRIA-
TUMgroup also showed a slight gradual decrease in the time taken
to respond to both stimuli across the ve sessions. SHAM rats
showed a continuous decrease in latencies of response to S
+
stimuli
throughout the training session while for S

stimuli a decrease was


observedduringsession2and3, but thentherats begantowithhold
their responses (and therefore, started to exhibit correct behav-
ior). Consequently, a signicant difference between the groups was
observed on S
+
[F(1,17) =10.05; p<0.01, but not across the ve
sessions [F(4.68) =0.65; NS], and not on S

stimuli [F(1,17) =0.26;


NS], but with a signicant difference across sessions [F(4,68) =3.97;
p<0.01].
We additionally performed a detailed analysis considering per-
formance in blocks of 10 trials within each of the 5 sessions (S1-S5;
Fig. 4). From S1 to S3 a slight difference between groups was
observedoncumulative time. This difference didnot reachthe level
of signicance. On S4, the difference was more pronounced, yet
not signicant. On S5, however, the difference was dramatic and
signicant [F(1,17) 5.55; p0.05]. Similarly, in terms of laten-
cies (Fig. 5), a decit between the two groups appeared on S4 and
S5 (Fig. 5) and only the SHAM group made constant correct asso-
ciations between S
+
and S

latencies on all blocks on S4 and S5


[F(1,18) 4.62; p0.05]; lesioned rats showed correct associations
only during one single block (block3) on session 5, [F(1,16) =4.73;
p<0.05].
3.2.2. Experiment 2: lesions of the dorsal hippocampus
The performance of SHAM rats improved across sessions,
reaching a correct response score of 805% on session 5. This
improvement was not observed in D HIPP rats (Fig. 6A). MANOVAs
indicated a group difference across the ve sessions [F(1,17) =11.2;
p<0.01] and even over sessions (groupsession interaction,
[F(4,68) =28.42; p<0.001]. Subsequent ANOVAs revealed that this
difference appearedfromthe thirdsessiononwards [F(1,17) 5.64;
p0.05]. Only on session 5, D HIPP rats performed above chance
level (t =3.57; p<0.01).
The cumulative time decreased over the sessions
[F(4,68) =11.89; p<0.001] but similarly [F(4,68) =0.52; NS] in
both groups (Fig. 6B), and there was no signicant group effect
[F(1,17) =0.51; NS].
Training performance analyzed by S
+
and S

latencies for the


two groups is presented in Fig. 6C. While correct associations on
M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103 99
S1
0 1 2 3 4 5 6
0
2
4
6
8
10
L
A
T
E
N
C
Y

(
s
e
c
.
)
S2
0 1 2 3 4 5 6
0
2
4
6
8
10
S3
0 1 2 3 4 5 6
0
2
4
6
8
10
S4
0 1 2 3 4 5 6
0
2
4
6
8
10
*
*
*
*
*
*
*
*
*
*
L
A
T
S5
0 1 2 3 4 5 6
0
2
4
6
8
10
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
BLOCK OF TRIALS
DL STRIATUM S+
DL STRIATUM S-
SHAM S +
SHAM S -
S=SESSION
Fig. 5. Mean latencies by block of 10 trials (SEM) recorded during sessions S1 to S5. S
+
are the latencies for the positive odor and S

the latencies for the negative one. SHAM


rats started to make constant signicant odor-reward associations fromthe rst block of the fourth session onwards. Only a transient signicant association was observed
for STRIATUMrats during the third block of the fth session, indicating absence of consistency. See also legend to Fig. 3.
S
+
and S

stimuli started to be signicant from the third session


onwards in the SHAM group [F(1,18) 4.74 p0.05], D HIPP rats
responded to S
+
or S

trials without discernment [F(1,16) 1.6;


NS]. This difference came as well over the sessions from the
S
+
[F(4,68) =14.87; p<0.001] as the S

stimuli [F(4,68) =11.89;


p<0.001]
Amore detailed analysis considered changes incumulative time
over blocks of 10 trials within each of the 5 sessions (S1S5; Fig. 7).
It showed that no scores, even during one single block, were sig-
nicantly different between SHAM and D HIPP rats. Inversely, on
latencies, a profound difference between the two groups appeared
fromthe fourth session onwards (Fig. 8), where a signicant odor-
reward association appeared fromthe rst block of S4 in the SHAM
group [F(1,18) 9.52 p0.01]. In D HIPP rats, such an effect was
never observed whatever the session [F(1,16) 1.5; NS].
4. Discussion
In this operant task, when a response is given during the last
second of the ITI, the start of the next trial is delayed by 10s, and
additional 10s delay are added as long as, on the last second of an
ongoing delay, the rat is still present inthe corner where the reward
is delivered. Thus, the duration between two trials could last longer
than 15s and constituted the cumulative time added to the 15s of
the xed ITI time. A behavioral inhibition decit, once the rats have
learned that they can get the positive reward in the corner, should
be followed by a substantial increase of the cumulative time in all
sessions. In fact, such behavior is observed only during the rst
and the second sessions (Figs. 4 and 7) for both sham groups and
hippocampal lesioned rats. Afterwards, one observes a signicant
decrease. On the contrary, the rats with lesions of the dorsolateral
striatumdid not increase or even decreased their cumulative time
across sessions; they responded randomly during the last second
of the ITI and occasionally delayed the next trial. Consequently, the
mean of the cumulative time was constant throughout the sessions
and did not seemto be related to an alteration of behavioral inhibi-
tion capacities. An alternative explanation could be that the lesions
of the dorsolateral structure disrupted the timing ability, in agree-
ment with the scalar timing theory developed by Meck and Church
[25]. In fact, as demonstrated previously in this task [26], improved
performance is observed despite altered timing ability. Indeed, the
disruptionof theabilitytomanagethetimingas seenafter lesions of
thedorsomedial prefrontal cortexledrats torespondonlywhenthe
olfactory cue was presented, thereby facilitating the odor-reward
association due to the absence of extra time added to the xed ITI.
A third possibility would be that the dorsolateral striatum lesions
100 M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103
1 2 3 4 5
0
30
40
50
60
70
80
90
*
*
*
*
*
*
*
SHAM
D HIPP
*
*
*
*
*
P
E
R
C
E
N
T
A
G
E

O
F
C
O
R
R
E
C
T

R
E
S
P
O
N
S
E
S
0 1 2 3 4 5
0
1
2
3
4
5
6
7
8
9
10
SHAM S +
SHAM S -
D HIPP S+
D HIPP S -
*
*
*
*
*
*
*
SESSION (Day)
L
A
T
E
N
C
Y

(
S
e
c
.
)
0 1 2 3 4 5
0
1
2
3
4
5
6
7
8
9
10
SHAM
D HIPP
C
U
M
U
L
A
T
I
V
E

T
I
M
E

(
S
e
c
.
)
A
C
B
Fig. 6. Mean performance (SEM) obtained across the ve sessions of 60 trials in
control rats (SHAM) (N=10) and rats with bilateral dorsal hippocampus lesions (D
HIPP) (N=9). (A) Mean percentage of correct responses. The difference between the
SHAMand D HIPP groups was signicant fromthe third session onwards. (B) Mean
cumulative time (in seconds); no statistical difference was observed between the
two groups. (C) Meanlatencies (inseconds); S
+
are the latencies for the positive odor
and S

the latencies for the negative odor. Only the SHAM group started to make
signicant associations fromsession 3 onwards.
disrupted the establishment of stimulus-response associations
[13,27] or altered the habit-based memory system [28]. How-
ever, in agreement with these authors, dorsal striatum lesions
marginally impair the initial learning of odor discrimination. Fur-
thermore, before a decit became apparent, the presentation of
three successive discrimination problems was required. Comp-
ton [29] suggested that the dorsal striatum is necessary for the
mediation of stimulus-response learning while the hippocampus is
necessary to mediate expression of place learning (in their experi-
ments); these processes can occur simultaneously and in parallel.
In our current experiment, dorsolateral striatumlesioned rats were
both impaired in cumulative time and in odor-reward associa-
tions, whereas hippocampal lesioned rats were able to withhold
a response during the ITI but showed unable to make correct
stimulus-response associations. To master the timing in this task
a stimulus-elicited response is not required and the dorsolateral
striatum rats were impaired by responding randomly during the
ITI. So, in addition to stimulus-response impairments, it seems that
striatumlesioned rats did not gure out howto use the task to get
only the positive reward and to get it rapidly. This decit might be
interpreted as consequent to an impairment of suppressing auto-
matic responding.
Inrodents, lesions altering the medial part of the dorsal striatum
usually impair goal-directed learning. Damage to the dorsolateral
striatum is described as impacting habit learning (for a review,
see [30]). Research on skill learning and cognitive control is often
guidedbyamultiplememorysystems framework. This framework
usually contrasts declarative memory, which provides exible and
explicite access to semantic and episodic content and requires con-
scious awareness, with memory for howto implement procedures
(i.e., how to perform a sequence of actions). These procedures
can become automatic and subconscious [31]. Based on the recent
reviewby Liljeholmand ODoherty [30], the impairments reported
herein after lesions of the dorsolateral striatumcould be the result
of impaired procedure learning and consequently reect a proce-
dural memory dysfunction.
Althoughseveral studies showa dissociationbetweenthe forms
of memory related to the striatum (procedural) and those related
to the hippocampus (declarative), there is also evidence pointing
to the involvement of the caudate nucleus in declarative memory
[32,33]. For instance, Ben Yakov and Dudai [34] performed fMRI
experiments in humans in which they measured brain activity
time-locked to the offset of short narrative audiovisual episodes.
They found that a set of brain regions, most prominently the
hippocampus and the caudate nucleus, both bilaterally exhibited
memory-predictive activity that was time-locked to the stimu-
lus offset. The hippocampus and caudate nucleus are involved in
registeringintegratedepisodes tomemoryas cohesiveinternal rep-
resentations. It cannot be excluded that in our rats the lesions of
the dorsolateral striatum disabled a similar process and thus pre-
vented learning by affecting also the declarative-like component of
the task.
We would nevertheless privilege an alternative explanation.
The fact that procedural memory decits prevented the acquisi-
tion of the association, and thus a declarative-like item, could be
questioned from a view considering encoding in the SPI system
proposed by Tulving in humans [35]. This systemplaces procedural
memory at the lowest hierarchical level, semantic memory above,
the highest level being that of episodic memory. From this func-
tional hierarchy of encoding processes, it can be derived that, in
tasks requiring a serial engagement of procedural and declarative-
like memory processes, as seems to be the case for the task used
in the current study, a decit affecting the lowest system should
have consequences on the efciency of the higher one. This is one
possible interpretation of our observations in rats with lesions of
the dorsolateral striatum. In our study both the hippocampus and
striatumwere involved in learning and memorizing the meaning of
each olfactory cue. It is thus possible, given the type of task which
weused, that thedorsolateral striatumandthedorsal hippocampus
had to contribute to task acquisition in a sequential order. First the
dorsolateral striatummight have been required to master the pro-
cedural component of the task, mainly during the two rst sessions.
Then, thehippocampus might havetakenover theacquisitionof the
odor-reward associations, and thus the declarative-like items. In
other words, before being able to make correct associations engag-
ing the hippocampus, rats had to learn the timing sequence of
the task, for which the engagement of the dorsolateral striatum
might be crucial. Indeed, learning to respond only when an olfac-
tory cue is presented and not to respond during the inter-trial delay
(when only a neutral air is on) required a tuned control of the loco-
motor neural system to engage into going or refrain from going
(engage into No-Go behavior). Then, in a second phase, a choice
has to be made in order to respond only to the positively-rewarded
M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103 101
S1
0 1 2 3 4 5 6
0
5
10
15
20
C
U
M
U
L
A
T
I
V
E

T
I
M
E

(
s
e
c
.
)
S2
0 1 2 3 4 5 6
0
2
4
6
8
10
S3
0 1 2 3 4 5 6
0
2
4
6
8
10
S4
0 1 2 3 4 5 6
0
2
4
6
8
10
S5
0 1 2 3 4 5 6
0
2
4
6
BLOCK OF TRIALS
D HIPP
SHAM
S=SESSION
Fig. 7. Mean cumulative time in blocks of 10 trials (SE) recorded during sessions S1 to S5. Both groups showed a time decreasing over sessions. No signicant difference
was observed between SHAMand D HIPP groups. See also legend to Fig. 6.
cue. Whenthe striatum-dependent learningis preventedbystriatal
lesions, the acquisition of the hippocampus-dependent selection
of the correct odor to engage in an appropriate response might
become impossible. Previous results using this olfactory associa-
tive task showed that 5-HT4 receptors are substantially involved
in the modulation of learning and memory processes [24]. Prior to
the rst training session, the injection of a selective 5-HT4 receptor
antagonist (RS 67532) was followed by an increase in cumulative
time, demonstrating that treated rats were more active at gur-
ing out how to get the reward and delayed the upcoming trial
by responding during the ITIs on the rst session. This behavior
allowed themto learn the protocol timing faster and consequently,
once the procedural aspect of the task was acquired, treated rats
startedto make consistent cuerewardassociations during the sec-
ond session while control rats did not start until the third one. The
injections of the partial agonist (RS 67333) prior to the rst session
had no signicant effect, whether on the rst or subsequent ses-
sions, and only a learning decit was observed. One interesting
point, in this study, was that injections administered prior to the
rst or thirdtrainingsessiongaverisetooppositeeffects. Afacilitat-
ing effect on learning and memory was found on session 3 when RS
67333hadbeeninjectedprior to this session, andanopposite effect
when the injection occurred prior to the rst session. The same
procedure with antagonist RS 67532 injections was followed by a
learning and memory performance decit on session 4 when the
injection occurred before the third session, and a facilitatory effect
on session 2 when it occurred before the rst session. The opposite
effects obtained in these experiments with the partial agonist and
antagonist canbe relatedto the sub-categories of memory involved
inthis task. Decreasing the activationof the hippocampal systemby
the use of a 5-HT4 receptor antagonist at the onset of training facili-
tates procedural memory, andthe inverse seems true for the partial
agonist. By contrast, once the timing component of the task (proce-
dural memory) has been learned (after the rst two sessions), the
injection of the partial agonist facilitates hippocampal processing
andconsequentlyimproves memory for the associations (reference
memory, related to declarative memory in humans); the opposite
reasoning applies to the antagonist. As a whole, fromthese exper-
iments, it seems that procedural learning comes rst in this task
and is crucial for acquiring subsequent declarative-like learning.
When declarative-like and procedural memory processes are
investigated in laboratory rodents, they are often considered as
operating in dual and sometimes even competing memory sys-
tems. In rats trained for response learning in the cross-maze task,
Packard and McGaugh [36] showed that the initial approach of the
task engaged a hippocampus-dependent declarative-like memory
systemacquiring knowledge about place. Only over task repetition
was there a gradual shift towards a striatum-dependent response
memory system. Approaches and tasks such as the ones used by
Packard and McGaugh posit the declarative-like memory as a pre-
requisite for response memory to be established over behavior
repeats (see also [37,38]). In the test situation used in our current
experiments, the cognitive operations required for task acquisi-
tion seem to occur according to an inverse schedule: only when
102 M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103
S1
0 1 2 3 4 5 6
0
1
2
3
4
5
6
7
8
9
10
L
A
T
E
N
C
Y

(
s
e
c
.
)
S2
0 1 2 3 4 5 6
0
1
2
3
4
5
6
7
8
9
10
S3
0 1 2 3 4 5 6
0
1
2
3
4
5
6
7
8
9
10
S4
0 1 2 3 4 5 6
0
1
2
3
4
5
6
7
8
9
10
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
*
S5
0 1 2 3 4 5 6
0
1
2
3
4
5
6
7
8
9
10 ***
BLOCK OF TRIALS
D HIPP S +
D HIPP S -
SHAM S +
SHAM S-
S=SESSION
Fig. 8. Mean latencies in block of 10 trials (SE) recorded during sessions S1 to S5. S
+
are the latencies for the positive odor and S

the latencies for the negative one. SHAM


rats started to make constant signicant odor-rewards associations fromthe rst block of the fourth session onwards. No signicant associations were observed in D HIPP
rats across the ve sessions. See also legend to Fig. 6.
the dorsolateral striatum-dependent component of the task is
mastered can the declarative-like, hippocampus-dependent com-
ponent of the task be learned and memorized, as demonstrated
by the consequences of our permanent lesions. This, along with
recently published data [37,38], is another example showing that
systems driving declarative-like and procedural memories may
cooperate for adapting goal-directed behaviors.
5. Conclusion
The original contribution of our current study resides in the
demonstration that, at least in the olfactory association task we
used, the establishment of a procedural memory may be a prereq-
uisite for that of a declarative-like one.
Acknowledgements
We are grateful for the expert secretarial assistance of Valerie
Demare.
References
[1] Hirsh R. The hippocampus and contextual retrieval of information frommem-
ory: a theory. Behavioral Biology 1974;12:42144.
[2] OKeefe J, Nadel L. The hippocampus as a cognitive map. London: Oxford Uni-
versity Press; 1978.
[3] OltonDS, Werz MA. Hippocampal functionandbehavior: spatial discrimination
and response inhibition. Physiology and Behavior 1978;20:597605.
[4] Jeltsch H, Bertrand F, Lazarus C, Cassel JC. Cognitive performances and loco-
motor activity following dentate granule cell damage in rats: role of lesion
extent and type of memory tested. Neurobiology of Learning and Memory
2001;76:81105.
[5] Silveira JM, Kimble DP. Brightness discrimination and reversal in
hippocampally-lesioned rats. Physiology and Behavior 1968;3:62530.
[6] Harley CW. Hippocampal lesions and two cue discrimination in the rat. Physi-
ology and Behavior 1972;9:3438.
[7] Eichenbaum H, Fagan A, Cohen NJ. Normal olfactory discrimination learning
set and facilitation of reversal learning after medial-temporal damage in rats:
implications for an account of preserved learning abilities in amnesia. Journal
of Neuroscience 1986;6:187684.
[8] Roman FS, Simonetto I, Soumireu-Mourat B. Learning and memory of odor-
reward association: selective impairment following horizontal diagonal band
lesions. Behavioral Neuroscience 1993;107:7281.
[9] Chaillan FA, Devigne C, Diabira D, Khrestchatisky M, Roman FS, Ben-Ari Y, et al.
Neonatal gamma-ray irradiation impairs learning and memory of an olfactory
associativetaskinadult rats. EuropeanJournal of Neuroscience1997;9:88494.
[10] Marchetti E, Jacquet M, JeltschH, Migliorati M, Nivet E, Cassel JC, et al. Complete
recovery of olfactory associative learning by activation of 5-HT4 receptors after
M. Jacquet et al. / Behavioural Brain Research 239 (2013) 94103 103
dentate granule cell damage in rats. Neurobiology of Learning and Memory
2008;90:18591.
[11] Eichenbaum H, Fagan A, Mathews P, Cohen NJ. Hippocampal system
dysfunction and odor discrimination learning in rats: impairment or facil-
itation depending on representational demands. Behavioral Neuroscience
1988;102:3319.
[12] Rawlins JN, Tsaltas E. The hippocampus, time and working memory.
Behavioural Brain Research 1983;10:23362.
[13] Packard MG, Hirsh R, White NM. Differential effects of fornix and caudate
nucleus lesions on two radial maze tasks: evidence for multiple memory sys-
tems. Journal of Neuroscience 1989;9:146572.
[14] Neill DB, Grossman SP. Behavioral effects of lesions or cholinergic blockade of
the dorsal and ventral caudate of rats. Journal of Comparative and Physiological
Psychology 1970;71:3117.
[15] Mitcham JC, Thomas Jr RK. Effects of substantia nigra and caudate nucleus
lesions on avoidance learning in rats. Journal of Comparative and Physiological
Psychology 1972;81:1017.
[16] Allen JD, Davison CS. Effects of caudate lesions on signaled and nonsignaled
Sidman avoidance in the rat. Behavioral Biology 1973;8:23950.
[17] Kirkby RJ, Polgar S. Active avoidance in the laboratory rat following lesions
of the dorsal or ventral caudate nucleus. Physiological Psychology 1974;2(3-
A):3016.
[18] Winocur G. Functional dissociation within the caudate nucleus of rats. Journal
of Comparative and Physiological Psychology 1974;86:4329.
[19] Prado-Alcala RA, Grinberg ZJ, Arditti ZL, Garcia MM, Prieto HG, Brust-Carmona
H. Learning decits produced by chronic and reversible lesions of the corpus
striatumin rats. Physiology and Behavior 1975;15:2837.
[20] SchwartzbaumJS, Donovick PJ. Discrimination reversal and spatial alternation
associated with septal and caudate dysfunction in rats. Journal of Comparative
and Physiological Psychology 1968;65:8392.
[21] Whishaw IQ, Zaborowski JA, Kolb B. Postsurgical enrichment aids adult
hemidecorticaterats onaspatial navigationtask. Behavioral andNeural Biology
1984;42:18390.
[22] Whishaw IQ, Mittleman G, Bunch ST, Dunnett SB. Impairments in the acqui-
sition, retention and selection of spatial navigation strategies after medial
caudate-putamen lesions in rats. Behavioural Brain Research 1987;24:12538.
[23] Viaud MD, White NM. Dissociation of visual and olfactory conditioning in the
neostriatumof rats. Behavioural Brain Research 1989;32:3142.
[24] Marchetti E, Chaillan FA, Dumuis A, Bockaert J, Soumireu-Mourat B, Roman FS.
Modulation of memory processes and cellular excitability in the dentate gyrus
of freely moving rats by a 5-HT4 receptors partial agonist, and an antagonist.
Neuropharmacology 2004;47:102135.
[25] Meck WH, Church RM. Simultaneous temporal processing. Journal of Experi-
mental Psychology: Animal Behavior Processes 1984;10:129.
[26] Chaillan FA, Marchetti E, Delfosse F, Roman FS, Soumireu-Mourat B. Opposite
effects depending on learning and memory demands in dorsomedial prefrontal
cortex lesioned rats performing an olfactory task. Behavioural Brain Research
1997;82:20312.
[27] Mishkin M, Malamut B, Bachelier J. Memories and habits: two neural systems.
In: Lynch G, McGaugh JL, Weinberger NM, editors. Neurobiology of learning
and memory. NewYork: Guilford; 1984. p. 6577.
[28] Broadbent NJ, Squire LR, Clark RE. Rats depend on habit memory for discrimi-
nation learning and retention. Learning and Memory 2007;14:14551.
[29] Compton DM. Behavior strategy learning in rat: effects of lesions of the dorsal
striatumor dorsal hippocampus. Behavioural Processes 2004;67:33542.
[30] Liljeholm M, ODoherty JP. Contributions of the striatum to learning, motiva-
tion, and performance: an associative account. Trends in Cognitive Sciences
2012;16:46775.
[31] Poldrack RA, Foerde K. Category learning and the memory systems debate.
Neuroscience and Biobehavioral Reviews 2008;32:197205.
[32] Schott BH, Seidenbecher CI, Fenker DB, Lauer CJ, Bunzeck N, Bernstein HG,
et al. The dopaminergic midbrain participates in human episodic mem-
ory formation: evidence from genetic imaging. Journal of Neuroscience
2006;26:140717.
[33] Blumenfeld RS, Parks CM, Yonelinas AP, Ranganath C. Putting the pieces
together: the role of dorsolateral prefrontal cortex inrelational memory encod-
ing. Journal of Cognitive Neuroscience 2010;23:25765.
[34] Ben-Yakov A, Dudai Y. Constructing realistic engrams: poststimulus activity
of hippocampus and dorsal striatum predicts subsequent episodic memory.
Journal of Neuroscience 2011;31:903242.
[35] Tulving E. How many memory systems are there? American Psychologist
1985;40:38598.
[36] Packard MG, McGaughJL. Inactivationof hippocampus or caudate nucleus with
lidocaine differentially affects expression of place and response learning. Neu-
robiology of Learning and Memory 1996;65:6572.
[37] Cassel R, Kelche C, Lecourtier L, Cassel JC. The match/mismatch of visuo-spatial
cues between acquisition and retrieval contexts inuences the expres-
sion of response vs place memory in rats. Behavioural Brain Research
2012;230:33342.
[38] Pol-Bodetto S, Jeltsch-David H, Lecourtier L, Rusnac N, Mam-Lam-Fook C, Cos-
quer B, et al. The double-H maze test, a novel, simple, water-escape memory
task: acquisition, recall of recent and remote memory, and effects of systemic
muscarinic or NMDA receptor blockade during training. Behavioural Brain
Research 2010;218:13851.

You might also like