Tobias Bast et al- The ventral hippocampus and fear conditioning in rats: Different anterograde...

8/3/2019 Tobias Bast et al- The ventral hippocampus and fear conditioning in rats: Different anterograde amnesias of fear af

1/14

Abstract Studies on the involvement of the rat hippo-campus in classical fear conditioning have focused main-ly on the dorsal hippocampus and conditioning to a con-

text. However, the ventral hippocampus has intimateconnections with the amygdala and the nucleus accumb-ens, which are involved in classical fear conditioning toexplicit and contextual cues. Consistently, a few recentlesion studies have indicated a role for the ventral hippo-campus in classical fear conditioning to explicit and con-textual cues. The present study examined whether neuro-nal activity within the ventral hippocampus is importantfor the formation of fear memory to explicit and contex-tual cues by classical fear conditioning. Tetrodotoxin(TTX; 10 ng/side), which completely blocks neuronalactivity, or muscimol (1 g/side), which increases GABAAreceptor-mediated inhibition, were bilaterally infused in-to the ventral hippocampus of Wistar rats before the con-ditioning session of a classical fear-conditioning experi-ment. Conditioning to a tone and the context were as-sessed using freezing as a measure of conditioned fear.TTX blocked fear conditioning to both tone and context.Muscimol only blocked fear conditioning to the context.The data of the present study indicate that activity ofneurons in the ventral hippocampus is necessary for theformation of fear memory to both explicit and contextualcues and that neurons in the ventral hippocampus thatbear the GABAA receptor are important for the forma-tion of fear conditioning to a context. In addition, bothbilateral muscimol (0.5 g/side and 1 g/side) and TTX(5 ng/side and 10 ng/side) infusion into the ventral hip-pocampus dose-dependently decreased locomotor activi-ty in an open-field experiment.

Keywords Classical fear conditioning Locomotoractivity Muscimol Tetrodotoxin Ventral hippocampus Rat

Introduction

In rats as well as primates, the hippocampus is widelybelieved to play an important role in learning and memo-ry, when complex representations, assembled from ele-mental stimuli and their spatial and/or temporal rela-tions, are involved (Eichenbaum et al. 1999; Jarrard1993; OKeefe and Nadel 1978; Rudy and Sutherland1995; Wallenstein et al. 1998). The hippocampus isthought to be involved in classical conditioning mainlywhen contextual information is utilized or when the con-

ditioned stimulus (CS) is a contextual one (Fanselow2000; Holland and Bouton 1999; Jarrard 1993; Marenand Holt 2000).

In classical fear conditioning, a neutral CS such as atone (explicit CS) or the environment (contextual CS) ispaired with an aversive unconditioned stimulus (US),such as an electrical foot shock. After a few pairings, theCS elicits conditioned fear responses, including freezing(Fanselow 1984). There is a consensus that neuronal pro-cessing in the amygdala is important for classical fearconditioning to contextual as well as explicit CS. Where-as some consider the amygdala to be the site wherethe fear memory is formed and stored (Fanselow andLeDoux 1999; Fendt and Fanselow 1999), others holdthe view that amygdaloid processes modulate the forma-tion and storage of the memory in other brain regions(Cahill et al. 1999). In addition, the nucleus accumbens,in particular its dopaminergic innervation from the mid-brain, seems to be involved in classical aversive condi-tioning, including fear conditioning (Haralambous andWestbrook 1999; Murphy et al. 2000; Parkinson et al.1999; Pezze et al. 2000; Riedel et al. 1997; Schwartingand Carey 1985; Westbrook et al. 1997; Wilkinson et al.1998; Young et al. 1993; A.L. Jongen-Rlo, S. Kauf-mann and J. Feldon, unpublished findings). As to the

T. Bast W.-N. Zhang J. Feldon ()Behavioral Neurobiology Laboratory,The Swiss Federal Institute of Technology Zurich,Schorenstrasse 16, CH 8603 Schwerzenbach, Switzerlande-mail: [email protected].: +41-1-6557448, Fax: +41-1-6557203

Exp Brain Res (2001) 139:3952DOI 10.1007/s002210100746

R E S E A R C H A R T I C L E

Tobias Bast Wei-Ning Zhang Joram Feldon

The ventral hippocampus and fear conditioning in rats

Different anterograde amnesias of fear after tetrodotoxin inactivation

and infusion of the GABAA agonist muscimol

Received: 27 October 2000 / Accepted: 8 March 2001 / Published online: 16 May 2001 Springer-Verlag 2001


2/14

hippocampal involvement in classical fear conditioning,studies examining the conditioned freezing response af-ter hippocampal lesions have led to the view that thedorsal hippocampus is specifically involved in fear con-ditioning to a contextual CS. Consistent with the widelyheld notions of hippocampus-dependent learning, this isattributed to the dorsal hippocampus being essential forcontextual learning (Fanselow 2000; Fendt and Fanselow

1999; Maren et al. 1998). However, it is also suggestedthat hippocampal lesions might impair the expression ofconditioned fear in the form of freezing, rather than con-textual memory (Gewirtz et al. 2000; Richmond et al.1999).

The ventral hippocampus possesses connections link-ing it with the basic neuronal processes underlying clas-sical fear conditioning. It has intimate reciprocal connec-tions with the amygdala and strong projections to the nu-cleus accumbens (Groenewegen et al. 1987; Pitknenet al. 2000; Swanson and Cowan 1977). Moreover, activ-ity of the ventral hippocampus regulates synaptic plastic-ity in the basolateral amygdala (Maren and Fanselow

1995) and extracellular dopamine levels in the nucleusaccumbens (Blaha et al. 1997; Brudzynski and Gibson1997; Legault and Wise 1999; Legault et al. 2000;Mitchell et al. 2000). Thus, in addition to the involve-ment of the dorsal hippocampus in classical fear condi-tioning to contextual CS, classical fear conditioningmight depend on activity of the ventral hippocampus.

Indeed, some behavioral studies have suggested thatnormal function of the ventral hippocampus is necessaryfor classical aversive conditioning, including fear condi-tioning. Formation and retrieval of memory in the pas-sive-avoidance paradigm are impaired following tempo-rary inactivation of the ventral hippocampus by tetrodo-

toxin (TTX; Ambrogi Lorenzini et al. 1997). As for clas-sical fear conditioning, electrolytic or excitotoxic lesionsof the ventral hippocampus, prior to or 1 day after condi-tioning, impair conditioned freezing to both explicit andcontextual CS (Maren 1999; Maren and Fanselow 1995;Richmond et al. 1999). Thus, the ventral hippocampuscould be important for the formation (acquisition andconsolidation), retrieval, or expression of conditionedfear.

The present study examined whether normal activityof the ventral hippocampus is necessary for the forma-tion of classical fear conditioning to explicit and contex-tual CS. For that purpose, 10 ng TTX/0.5 l per side or1 g muscimol/0.5 l per side were bilaterally infusedinto the ventral hippocampus of Wistar rats before theconditioning session of a classical fear-conditioning ex-periment, with freezing being utilized to assess condi-tioned fear. TTX, blocking voltage-dependent sodiumchannels, should temporarily induce the condition exist-ing after electrolytic lesions, i.e., inactivation of the ven-tral hippocampus as well as fibers of passage (AmbrogiLorenzini et al. 1999). Muscimol should temporarily in-hibit neuronal activity within the ventral hippocampus,because muscimol is an agonist at the GABAA receptor,the Cl channel receptor mediating inhibition of the hip-

pocampal excitatory network by GABA-releasing inter-neurons (Buhl et al. 1994). Activity of the ventral hippo-campus is important for the regulation of locomotor ac-tivity (Bardgett and Henry 1999; Bast et al. 2001, 2001;Brudzynski and Gibson 1997; Legault and Wise 1999;Yang and Mogenson 1987). Moreover, observations dur-ing the fear-conditioning experiment indicated that infu-sion of TTX and muscimol into the ventral hippocampus

markedly decreased activity of the rats. Therefore, in anopen-field experiment, we measured locomotor activityfollowing infusion of TTX (10 ng/0.5 l per side or5 ng/0.5 l per side) or muscimol (1 g/0.5 l per side or0.5 g/0.5 l per side) into the ventral hippocampus.

Materials and methods

Animals

Sixty-six male, adult Wistar rats [Zur:Wist(HanIbm); ResearchUnit Schwerzenbach, Schwerzenbach, Switzerland], weighingabout 250300 g at the time of surgery, were used in this study.The animals were housed in groups of four per cage under a re-versed light-dark cycle (lights on 19:0007:00) in a temperature(211C)- and humidity (555%)-controlled room. All rats wereallowed free access to food and water. Twelve rats were left unop-erated (UNOP) and 54 rats received bilateral implantation of infu-sion guide cannulae aimed at the ventral hippocampus. After sur-gery, all 66 rats were individually caged. Beginning 3 days beforesurgery and throughout the studies, all rats were handled daily. Allexperimental procedures were carried out in the dark phase of thecycle. Principles of laboratory animal care (NIH publication no.86-23, revised 1985) and Swiss regulations for animal experimen-tation were followed.

Implantation of guide cannulae for intracerebral infusion

Rats were anesthetized with 1 ml of Nembutal (50 mg pentobarbi-

tal sodium/ml; Abbott Labs, North Chicago, Ill.) per kilogrambody weight and their head was placed in a stereotaxic frame(Kopf Instruments, Tujunga, Calif.). After application of a localanesthetic (lidocaine), the scalp was incised to expose the skull,and bregma and lambda were aligned in the same horizontal plane.Guide cannulae (9 mm, 26-gauge stainless steel) were implantedbilaterally through small holes (diameter 1.5 mm) drilled on eachside of the skull. The tips of the guide cannulae aimed at the fol-lowing coordinates above the ventral hippocampus: 5.2 mm poste-rior and 5 mm lateral to bregma, and 5 mm ventral to dura. Theguide cannulae were fixed with dental cement for which threesmall, stainless steel screws, previously screwed into the skull,served as anchors. Stainless steel stylets (34-gauge), which ex-tended 0.5 mm beyond the tips of the guide cannulae, were placedinside the guide cannulae to prevent occlusion. After surgery, theexperimenters gave the rats daily health checks and gentle han-

dling, and replaced missing stylets. The behavioral experimentscommenced 5 days after surgery.

Intracerebral microinfusion and drugs

For microinfusions into the ventral hippocampus, rats were manu-ally restrained and the stylets removed from the guide cannulae.Then, infusion cannulae (34-gauge) were inserted into the guidecannulae. The infusion cannulae were connected to 10-l Hamil-ton microsyringes mounted on a microinfusion pump (KD scien-tific or WPI sp200i). The tips of the infusion cannulae protrudedinto the ventral hippocampus 1.6 mm beyond the tip of the guidecannulae, thus aiming at a final dorsoventral coordinate of 6.6 mm

40


3/14

below the dura. The rats were bilaterally infused with TTX (5 ngor 10 ng) or muscimol (0.5 g or 1 g) in 0.5 l vehicle (0.9% sa-line) or with 0.5 l vehicle only. The infusion speed was0.5 l/min. To allow for absorption of the infusion bolus by thebrain tissue, the infusion cannulae were left in the brain for 60 safter infusion before being replaced by the stylets. Muscimol[C4H6N2O2(1/2 H2O); Tocris, Bristol, UK] was dissolved in 0.9%saline at a concentration of 2 g/l or 1 g/l on the day of infu-sion. TTX (C11H17N3O8; Tocris, Bristol, UK) was stored at 40Cin aliquots containing 40 ng/l in 0.9% saline. On the day of infu-

sion, these aliquots were thawed and diluted with 0.9% saline toobtain solutions with a concentration of 20 ng/l or 10 ng/l.Muscimol was infused immediately and TTX 20 min before thebehavioral sessions. Accordingly, half of the rats infused with ve-hicle received infusion immediately before the behavioral sessionsand the other half, 20 min before the behavioral sessions. Dosesand time points for the infusion of TTX and muscimol were cho-sen on the basis of the literature (Ambrogi Lorenzini et al. 1997,1999; Givens and Olton 1990; Mao and Robinson 1998; Martin1991; Zhuravin and Bures 1991). Since, to our knowledge, no ad-verse long-term effects have been reported after intracerebral infu-sion of these two drugs, we used comparatively high doses to en-sure a drug effect.

Apparatus for behavioral testing

Fear conditioning

Eight operant test boxes (Habitest; Coulbourn Instruments, Allen-town, Pa.) were used. Conditioning and the context test sessionstook place in four shock boxes fitted with a parallel-grid shockfloor (16 parallel bars; E1010RF; Coulbourn Instruments),through which scrambled shocks could be delivered. The shockboxes were placed in light- and sound-attenuating chambers mea-suring 55 cm 40 cm 55 cm. These chambers had two sidewalls of aluminum and a rear and front wall of clear Perspex. Awhite waste tray was situated below the grid floor. The tone testsessions were conducted in four no-shock boxes fitted with a lat-tice grid (E1018NS; Coulbourn Instruments) and placed in light-and sound-attenuating chambers measuring 72 cm 45 cm 45 cm. The no-shock boxes had three black walls and a front wall

of clear Perspex. A brown waste tray was situated below the lat-tice grid. The four shock and the four no-shock boxes were placedin two different rooms. Presentation of the auditory CS and deliv-ery of electric foot shock were controlled by a PC with dedicatedsoftware (S. Frank, Psychology Department, University of TelAviv, Israel) connected to a Coulbourn Universal Environment In-terface (E9112) with Coulbourn Universal Environment Port(L9112). The auditory CS [85 dB(A)] was produced by a2.9-kHz tone module (E1202) fixed at one wall of the operantchamber. Shocks were delivered with a Coulbourn precision ani-mal shocker (E1312), which generated bipolar, rectangular 10-mscurrent pulses with a frequency of 10 Hz. Background noise wasprovided by a ventilation fan affixed to the light- and sound-atten-uating chambers during all sessions. A monochrome minivideocamera with a wide-angle (100) 2.5-mm lens (VPC-465B; CES,Zurich, Switzerland) was attached to the center of the ceiling of

each operant chamber. Four infrared (875 nm) light-emitting di-odes (HSDL-4220; Hewlett Packard) positioned in the ceiling ofeach operant chamber provided light sufficient for camerafunction. Throughout all sessions, images from each of the fourshock or no-shock boxes, respectively, were provided by thesecameras, integrated into a four-quarter single image by a multi-plexer (DX216CE; Sony), and recorded by a videotape recorder(SVT1000; Sony). The images were transferred to a computer(7600/120 Power Macintosh) equipped with an analysis program(Image; http://rsb.info.nih.gov/nih-image) and a macroprogram (P.Schmid, Behavioral Neurobiology Laboratory, Swiss Federal In-stitute of Technology, Zurich). The proportion of time that wasspent in total immobility except for respiratory movements by ratsplaced in the boxes was measured by this system comparing adja-

cent 1-s frames of the recorded images. These measures of immo-bility correspond to measures of freezing behavior as obtained byuse of visual criteria commonly employed. The validation andprinciple of the automated analysis of freezing behavior have beendescribed in detail in previous publications (Murphy et al. 2000;Pryce et al. 1999; Richmond et al. 1998, 1999).

Open-field locomotor activity

Locomotor activity was measured in four closed square arenas(76.5 cm 76.5 cm 49 cm) made of dark gray plastic and placedin a room dimly illuminated (200.5 lx provided by two halogenlights). Behavior in the arenas was recorded by a video camerasuspended from the ceiling and relayed to a monitor and a videotracking, motion analysis and behavior recognition system (Etho-vision; Noldus, Wageningen, The Netherlands). For testing ofopen-field locomotor activity, the rats were placed in the center ofan arena and the total distance the rat moved in centimetersthroughout the complete arena was calculated by the Noldussystem for each 10-min block of testing.

Experimental design and procedures for behavioral testing

Fear-conditioning experiments

All 66 rats were used in one of two fear-conditioning experiments(I and II) to test the effects of muscimol or TTX infusion into theventral hippocampus on the formation of fear conditioning to anexplicit CS (auditory CS) and a contextual CS (environment).Freezing, i.e., total immobility except for respiratory movements,was used to assess conditioned fear and measured by the automat-ed system described above. Originally, fear-conditioning experi-ment I, which was conducted with 32 rats (26 cannulated, sixUNOP), was planned to assess the effects of muscimol and TTXinfusion on conditioning to both an explicit and a contextual CS.However, the procedure of experiment I did not result in signifi-cant conditioning to the contextual CS in the control groups. Forthat reason, a second fear-conditioning experiment with another34 naive rats (28 cannulated, six UNOP) was conducted using aprocedure to yield robust conditioning to the contextual CS. The

timing of treatments for the different groups, and the conditioningand test procedures in fear-conditioning experiments I and II aresummarized in Table 1.

In both experiments, cannulated rats were allocated to one ofthree infusion groups to receive bilateral infusions into the ventralhippocampus before conditioning on day 1: TTX10 (n=7), 10 ngTTX/0.5 l per side 20 min before the conditioning session;MUS1 (n=7), 1 g muscimol/0.5 l per side, immediately beforethe conditioning session; VEH (n=12 in experiment I and n=14 inexperiment II), 0.5 l vehicle per side 20 min before (n=6 or 7) orimmediately (n=6 or 7) before the conditioning session. In addi-tion, in each fear-conditioning experiment, UNOP rats (n=6)served to test for unspecific effects of cannulae implantation andinfusion procedure. UNOP rats were handled the same way as thecannulated rats but were not subjected to the surgery proceduresbefore the fear-conditioning experiment. Furthermore, they re-

ceived no treatment instead of the infusion before conditioning.The conditioning and test procedures during fear-conditioning ex-periments I and II were the same for all groups.

Fear-conditioning experiment I. For conditioning, rats were put inthe shock boxes for a total of 27 min and were exposed to ten pair-ings of a 30-s auditory CS [85 dB(A), 2.9 kHz] and a 1-s footshock (current pulse amplitude 0.5 mA) separated by 2-min blocksbetween an initial and a final 2-min block. The 1-s foot shock wascontiguous with the final second of the CS. The proportion of timespent immobile was calculated for the eleven 2-min blocks pre-ceding and following the 30-s CS and for the duration of each 30-sCS. One day after conditioning, in the context test session, ratswere tested for freezing to the contextual CS, i.e., the environment

41


4/14

in which they had received the foot shocks. For that purpose, ratswere placed in the shock box for 8 min without being exposed tothe auditory CS or the foot shock. Two days after conditioning, inthe tone test session, rats were tested for freezing to the auditoryCS. For that purpose, they were put in the no-shock box for a totalof 11 min. After 3 min, the auditory CS was presented for the re-maining 8 min. During context and tone test sessions, the propor-tion of time spent immobile was calculated for each 1-min block.

Fear-conditioning experiment II. The conditioning procedure offear-conditioning experiment I did not result in marked freezingduring the context test session. Therefore, we conducted a secondfear-conditioning experiment to test the effects of TTX and musci-mol infusion into the ventral hippocampus on the formation offear conditioning to a contextual CS. In the second fear-condition-ing experiment, conditioning was conducted with an unsignaled

shock, i.e., without an auditory CS, to achieve a strong associationbetween the shock and the contextual CS, i.e., the environment inwhich the rats received the foot shock (compare Odling-Smee1978). For conditioning, rats were put in the shock boxes for a to-tal of 30 min and 5 s and were exposed to five 1-s foot shocks(current pulse amplitude 0.5 mA) separated by 5-min blocks be-tween an initial and a final 5-min block. Five-minute blocks werechosen to provide sufficient time in particular before the firstshock for the rats to build up a context representation that couldbe associated with the shock (compare Fanselow 1986, 2000).The proportion of time spent immobile was calculated for the six5-min blocks preceding and following the 1-s foot shock. One dayafter conditioning, in the context test session, rats were tested forfreezing to the contextual CS, i.e., the shock boxes. For that pur-

pose, rats were placed in the shock box for 8 min, and the propor-tion of time spent immobile was calculated for each 1-min block.

Verification of intact shock reaction. TTX and muscimol infusioninduced hypoactivity, and the TTX10 and MUS1 rats exhibitedquite high levels of immobility before the first shock of the condi-tioning session, especially in fear-conditioning experiment II (seeResults). This indicated that the immobility in the TTX10 andMUS1 rats measured during the conditioning sessions of the twofear-conditioning experiments was not only reflective of a condi-tioned postshock response but partly due to hypoactivity inducedby the infusions preceding conditioning. Therefore, we verifiedthat all rats were really exhibiting an intact postshock response.For that purpose, we compared the proportion of immobility dur-ing the 30-s periods preceding and the 30-s periods immediatelyfollowing all 1-s foot shocks (pre-S and post-S period) delivered

after the first foot shock. The intact reaction of a rat to a footshock includes an immediate postshock activity burst of severalseconds, which is an unconditioned shock response. With repeatedshock presentations in an enclosure, this unconditioned shock re-sponse is followed by the conditioned component of freezing.Thus, an intact shock response should be revealed by reduced im-mobility during the post-S periods as compared to the pre-S peri-ods (Fanselow 1982).

Open-field experiment

Twenty-five cannulated rats from fear-conditioning experiment Iwere subjected to an open-field experiment 7 days after the fear-

42

Table 1 Summary of the procedures in fear-condtitioning experiments I and II

Day 1 Day 2 Day 3Context test Tone test

Infusion Conditioning

Fear-conditioning TTX10 group: 10 ng All groups: 27 min in shock All groups: 8 min in shock All groups: 11 min inexperiment I tetrodotoxin/side 20 min box; after 2 min, ten tone box to test for freezing no-shock box; tone on after

before conditioning (30 s)-shock (1 s, 0.5 mA) in the context of the foot 3 min to test for freezing topairings separated by 2-min shocks the auditory CS

MUS1 group: 1 g blocks; last 1 s of tone andmuscimol/side, foot shock contiguousimmediately beforeconditioning

VEH group: vehicleinfusion, either 20 minbefore or immediatelybefore conditioning

UNOP group:no treatment

Fear-conditioning TTX10 group: 20 min All groups: 30 min and All groups: 8 min in shock No tone testexperiment II 10 ng tetrodotoxin/side in shock box; after 5 min, box to test for freezing in

before conditioning 5 s five 1-s foot shocks the context of the foot

(0.5 mA) separated by shocksMUS1 group: 5-min blocks1 g muscimol/side,immediately beforeconditioning

VEH group: vehicleinfusion, either 20 minbefore or immediatelybefore conditioning

UNOP group:no treatment


5/14

conditioning experiment. The open-field experiment was designedto test the effects of muscimol or TTX infusion into the ventralhippocampus on locomotor activity in the open field. It was basedon observations during the fear-conditioning experiment indicat-ing that bilateral infusion of both muscimol (1 g/0.5 l per side)and TTX (10 ng/0.5 l per side) into the ventral hippocampus re-duced activity. To get an impression of dose-response relations,we split the TTX10 group of the fear-conditioning experiment intoa TTX10 (n=4) and a TTX5 (n=3) group to receive 10 ng or 5 ngTTX/0.5 l per side, respectively, into the ventral hippocampus

and the MUS1 group into a MUS1 (n=4) and a MUS0.5 (n=3)group to receive 1 g or 0.5 g muscimol/0.5 l per side, respec-tively, into the ventral hippocampus. In addition, there was a VEHgroup (n=11) to receive vehicle infusion into the ventral hippo-campus. We did not include unoperated animals, since in previousexperiments we did not find a difference between unoperated ratsand rats infused into the hippocampus with vehicle concerning ac-tivity in the open field (Bast et al. 2001; Zhang et al. 2000). Onday 1 of the open-field experiment, animals were tested for base-line activity in a 30-min session. Based on baseline activity ofday 1, the rats were allocated to the subgroups of the TTX, MUS,and VEH groups so that all groups had matched baseline activity.On day 2, rats received the following bilateral infusions into theventral hippocampus before being subjected to an open-field test-ing of 60 min: TTX10 and TTX5, 10 ng or 5 ng TTX/0.5 l perside 20 min before open-field testing; MUS1 and MUS0.5, 1 g or

0.5 g/0.5 l per side immediately before open-field testing; VEH,0.5 l vehicle/side 20 min (n=5) or immediately (n=6) beforeopen-field testing. We did not include a habituation phase beforeinfusion, because we expected a decrease in locomotor activity byour drug infusions. This could have been masked by a floor effectif the animals had already had a low level of activity due to habit-uation. On day 3 of the open-field experiment, rats were subjectedto 30 min of open-field testing to test for possible long-term ef-fects of the drug infusions.

Histology

After completion of the behavioral experiments, the cannulatedrats were deeply anesthetized with an overdose of 2.5 ml/kg Nem-butal (50 mg pentobarbital sodium/ml i.p.) and transcardially per-

fused with 0.9% NaCl solution to rinse out the blood, followed by250 ml of 4% formalin (4C) to fix the brain tissue. After extrac-tion from the skull, the brains were postfixed in 4% formalin solu-tion and subsequently cut into 40-m coronal sections on a freez-ing microtome. For the verification of the infusion sites, everyfifth section through the ventral hippocampus was mounted on agelatin-treated slide and stained with cresyl violet. After staining,the sections were dehydrated and coverslipped. Subsequently, theywere examined light-microscopically to verify that the tips of theinfusion cannulae were placed in the ventral hippocampus and todraw their approximate locations onto plates taken from the atlasof Paxinos and Watson (1998).

Data analysis

Statistical analysis was conducted with the Statview softwaresystem. Data were first subjected to analysis of variance(ANOVA). Groups were taken as between-subjects factor and thedifferent time blocks of testing as repeated measures. If indicatedby a significant outcome of the ANOVA, post hoc comparisonswere conducted using Fishers protected least significant differ-ence test. Significant differences were accepted at P0.25) or duringthe eleven 2-min blocks preceding and following the CSpresentations (F3, 28=2.18, P>0.1; Fig. 2A, B). ANOVA

of the percentage of time spent immobile yielded only asignificant effect of the ten CS presentations (F9, 252=8.36, P


6/14

increased in the MUS1 rats during the first 2-min blockpreceding the first foot shock in experiment I, visual ob-servation during and after conditioning revealed also de-creased activity in this group. These observationsprompted us to conduct the open-field experiment to ex-amine the effects of TTX and muscimol infusion into theventral hippocampus on locomotor activity.

44

Fig. 1A, B Infusion sites in the ventral hippocampus. A Photomi-crograph of a coronal brain section with the tracks of the guidecannulae and beneath them the infusion sites visible in both hemi-spheres. B Approximate location of the tips of the infusion cannu-lae depicted on plates of coronal sections through the rat brain(Paxinos and Watson 1998). Open circles represent sites of vehicleinfusion (26 rats), open squares represent sites of muscimol infu-sion (14 rats), and black circles represent sites of TTX infusion(14 rats). Values on the right represent distance from bregma.

(CA1, CA2, and CA3 CA1, CA2, and CA3 field of the hippocam-pus,DG dentate gyrus,Ententorhinal cortex, S subiculum)

Fig. 2AD Freezing during the three sessions of fear-conditioningexperiment I. There were four experimental groups: unoperated

rats (UNOP, n=6); rats which received bilateral infusion of vehicle(VEH, n=12) 20 min (n=6) or immediately (n=6) before the condi-tioning session; rats which received bilateral infusion of 1 g mu-scimol/side, immediately before the conditioning session (MUS1,n=7); rats which received bilateral infusion of 10 ng tetrodotox-in/side 20 min before the conditioning session (TTX10, n=7). Thevalues are means. The inset bar represents 1 SE as derived fromANOVA. A Freezing during the ten 30-s blocks of conditionedstimulus (CS) presentation in the conditioning session. B Freezingduring the eleven 2-min blocks preceding and following the CSpresentations during the conditioning session. C Freezing duringthe eight 1-min blocks of the context test. D Freezing during thethree 1-min blocks preceding the CS presentation and the subse-quent eight 1-min blocks of CS presentation during the tone test


7/14

Context test

No group exhibited marked freezing throughout the eight1-min blocks of the context test session (Fig. 2C). Thepercentage of time spent freezing, averaged over 8 min,varied from 2% in the TTX10 to 13% in the UNOPgroup. ANOVA of the percentage of time spent immo-bile yielded neither an effect of group (F3, 28=1.94,

P>0.1) or the eight 1-min blocks (F7, 196=1.75, P>0.1)nor an interaction of group and 1-min blocks (F21, 196=1.29, P>0.1), reflecting that all groups exhibited a simi-lar low level of freezing throughout the whole contexttest session.

Tone test

The TTX10 rats showed virtually no conditioned fearthroughout the whole tone test session, whereas UNOP,VEH, and MUS1 rats exhibited marked conditionedfreezing to the tone CS (Fig. 2D). During the 3 min pre-

ceding the CS presentation, the proportion of time spentimmobile was very low (less than 5%) and did not differbetween the groups (F3, 28=1.50, P>0.2). However, dur-ing the 8 min of CS presentation, all groups except theTTX10 group exhibited marked freezing, i.e., condi-tioned fear. ANOVA of the percentage of time spent im-mobile during these 8 min yielded a main effect of group(F3, 28=3.23, P


8/14

and 2, as depicted in Figs. 2B and 3A, cannot be com-pared directly, since the shock was delivered after 2 minin the first and after 5 min in the second experiment.However, separate analysis of the first 2 min of the con-ditioning session of experiment 2 yielded a similar pic-ture to that of the analysis of the first 2-min block of ex-periment 1, with immobility levels of only the TTX10but not the MUS1 group being increased. For this period,

ANOVA yielded a significant group effect (F3, 30=14.0,P


9/14

30 min, only the MUS rats were less active than the oth-er groups. Moreover, VEH and MUS0.5 rats exhibitedhabituation, which was most pronounced during the first30 min, whereas the other groups virtually maintained a

similar low level of activity throughout the entire ses-sion. Although separate ANOVA of both the first andsecond 30 min of testing revealed a significant effect ofgroups on the total distance moved (F4, 20=19.01,P


10/14

pus. In contrast, the hyperactivity following lesions ofthe ventral hippocampus is a long-term effect, possiblyreflecting compensatory changes in brain circuits regu-lating locomotor activity. Moreover, excitotoxic lesionsof the ventral hippocampus have recently been demon-strated to result in widespread changes in various extra-hippocampal brain regions (Halim and Swerdlow 2000).The effects of acute stimulation (Bardgett and Henry

1999; Bast et al. 2001, 2001; Brudzynski and Gibson1997; Legault and Wise 1999; Wu and Brudzynski 1995;Yang and Mogenson 1987) and inhibition or inactivationof the ventral hippocampus (present study) on open-fieldactivity indicate that the hippocampus does not in gener-al act as part of a behavioral inhibition system, as hasbeen suggested based on the evidence of lesion studies(Gray 1995; Kimble 1968). Stimulation of the ventralhippocampus probably increases locomotor activity byincreasing dopaminergic transmission within the nucleusaccumbens (Bardgett and Henry 1999; Bast et al. 2001,2001; Brudzynski and Gibson 1997; Legault and Wise1999; Wu and Brudzynski 1995). Thus, we suggest that

blockade or inhibition of neuronal activity of the ventralhippocampus cause hypoactivity by decreasing dopami-nergic transmission in the nucleus accumbens. This issuecan be approached by microdialysis studies.

The results of our two fear-conditioning experimentsare in line with recent studies that suggest a role for theventral hippocampus in classical fear conditioning tocontextual and explicit cues. These studies have demon-strated reduced freezing to aversively conditioned ex-plicit and contextual cues following lesions of the ven-tral hippocampus before and after conditioning (Maren1999; Maren and Fanselow 1995; Richmond et al. 1999).Richmond et al. (1999) have proposed that these freezing

deficits might result from lesions of the ventral hippo-campus interfering with the expression of conditionedfear in the form of freezing, as they induce hyperactivity.In contrast to lesions of the ventral hippocampus, musci-mol and TTX infusion into the ventral hippocampus in-duced hypoactivity (for possible explanations of this dif-ference, see previous paragraph) facilitating the expres-sion of freezing behavior. This explains why freezingduring the conditioning sessions tended to be increasedfollowing infusion of muscimol and TTX into the ventralhippocampus, whereas freezing was decreased duringconditioning following lesions of the ventral hippocam-pus (Richmond et al. 1999). In contrast to lesions, ma-nipulations of the ventral hippocampus induced by mu-scimol and TTX applied before conditioning were tem-porary and did not affect basal activity of the ventral hip-pocampus on the test days of the fear-conditioning ex-periment. Thus, an involvement of whatever kind ofthe ventral hippocampus in the expression of freezingcannot account for the decreased freezing observed onthe test days in the present studies. Our results indicatethat the specific associative processes underlying the for-mation of classical fear conditioning are dependent onnormal activity of the ventral hippocampus. However,one has also to consider alternative interpretations of our

results. TTX and muscimol infusion into the ventral hip-pocampus might interfere with fear conditioning by dis-rupting the detection of the aversive unconditioned stim-ulus, i.e., the foot shock. This can virtually be excludedfor the muscimol infusion, because its effect was restrict-ed to conditioning to the contextual cue. In addition, de-creased freezing during the 30-s periods immediatelyfollowing the shocks (post-S periods) relative to the 30-s

periods preceding the shocks (pre-S periods), which isreflective of the postshock activity burst (Fanselow1982), was observed in all experimental groups. This in-dicates an intact shock reaction in all groups. In theMUS1 and the TTX10 rats, the ventral hippocampus wasnot in the same drug state during conditioning and testsessions. Therefore, the virtual absence of conditionedfear in these groups during the test sessions might reflectstate dependency (Overton 1964) of fear learning. Usingfreezing to assess conditioned fear, it is difficult to testfor this, as the appropriate control groups would have toreceive TTX or muscimol infusion into the ventral hip-pocampus before the test sessions. Similar to the situa-

tion during the conditioning sessions of the two fear-con-ditioning experiments in the present study, this would re-sult in high levels of freezing due to the hypoactivity in-duced by these treatments and therefore make the detec-tion of reduced conditioned fear difficult. Studies onthe involvement of the amygdala in fear conditioning(Guarraci et al. 2000; Helmstetter and Bellgowan 1994;Muller et al. 1997), however, suggest that local microin-fusions into single brain areas do not interfere withlearning of fear merely on the basis of state dependency.Finally, decreased exploratory locomotor activity duringthe conditioning session due to the muscimol and TTXinfusions might interfere with building up a neuronal

representation of the conditioning context and thereby beresponsible for the anterograde deficits in contextual fearconditioning observed in the MUS1 and TTX10 rats.However, it is not very likely that the decreased explor-atory locomotor activity following muscimol and TTXinfusion into the ventral hippocampus can effectivelyprevent the registration of those components that aremost characteristic of the conditioning context, i.e., thegrid floor and the dark enclosure. Thus, we assume thatthe results of our fear-conditioning experiments reflectthat activity of the ventral hippocampus is important forthe specific associative processes underlying the forma-tion of fear memory to explicit and contextual stimuli.

Formation of long-lasting memory comprises the ini-tial acquisition, a short-term consolidation, believed tobe completed within seconds to tens of minutes, and thesubsequent long-term consolidation of an association(McGaugh 2000; Nadel and Moscovitch 1997). In ourfear-conditioning experiments, the infusions of TTX andmuscimol occurred before conditioning, i.e., acquisitionand short-term consolidation, but the drugs might havebeen active in the ventral hippocampus beyond the con-ditioning sessions, i.e., throughout the early phase of thelong-term consolidation. Our open-field data demon-strate that the muscimol infusion affected behavior for at

48


11/14

least 60 min after the beginning of the test sessions. Al-though the effects of TTX on locomotor activity wereevident no longer than 30 min following the start of theopen-field session, this does not necessarily mean thatthe drugs effect in the ventral hippocampus did not ex-tend beyond this period. Thus, principally, both musci-mol and TTX might have disrupted activity of the ven-tral hippocampus necessary for the acquisition and short-

term consolidation as well as the early long-term consol-idation of fear conditioning. It is important to note thatthe unimpaired freezing following TTX or muscimol in-fusion during the conditioning session does not meanthat the acquisition and short-term consolidation of con-ditioned fear were unimpaired. Since both TTX and mu-scimol infusion facilitated freezing by inducing hypoac-tivity, freezing throughout the conditioning may havebeen high despite decreased conditioned fear. An in-volvement of the ventral hippocampus in acquisition andshort-term consolidation as well as early long-term con-solidation of classical aversive conditioning has beensuggested by a previous study, reporting TTX infusion

into the ventral hippocampus (10 ng/1 l per side) 1 hbefore, immediately after, and 15 min after conditioningto interfere with memory formation in the passive avoid-ance paradigm (Ambrogi Lorenzini et al. 1997). Thus, inour fear-conditioning experiments, infusion of TTX andmuscimol into the ventral hippocampus may have im-paired the formation of fear memory by interfering withacquisition, short-term consolidation, early long-termconsolidation, or all of these processes.

The effects of the TTX and muscimol infusions onfear conditioning were different. While TTX (10 ng/side)infusion into the ventral hippocampus caused completeanterograde amnesia of fear, muscimol (1 g/side) infu-

sion into the ventral hippocampus only blocked forma-tion of fear memory to a contextual CS. The specific an-terograde amnesia for contextual fear following musci-mol infusion might merely be due to the fact that thedose used was only sufficient to block the weaker condi-tioning to the context but not the stronger conditioning tothe tone. However, one has to consider that the mecha-nisms by which TTX and muscimol take effect are dif-ferent. TTX leads to temporary functional inactivation ofall neurons and traversing fibers in the infusion area(Ambrogi Lorenzini et al. 1999), whereas muscimol onlyincreases inhibition of neurons bearing the GABAA re-ceptor. Thus, even given that both substances are appliedat doses to induce their maximal physiological effect, thebehavioral effects of TTX and muscimol may differ. Infact, the patterns of behavioral effects observed in thepresent study after infusion of muscimol and TTX intothe ventral hippocampus differ. While the hypoactivity inthe open field was rather more pronounced following in-fusion of muscimol (1 g/side) than following infusionof TTX (10 ng/side; Fig. 5), and disruption of contextualconditioning was similar following the infusion of bothsubstances (Fig. 3B), only the infusion of TTX led to adisruption of fear conditioning to a tone (Fig. 2D). Thus,with respect to fear conditioning, the effects of muscimol

in the ventral hippocampus might in particular disturbconditioning to a context, whereas the effects of TTX af-fect conditioning to a tone and to a context to a similarextent. The fibers and cell bodies in the ventral hippo-campus that are most likely to be involved in classicalfear conditioning are those connecting neurons of theventral hippocampus with the amygdala and the nucleusaccumbens (Groenewegen et al. 1987; Pitknen et al.

2000; Swanson and Cowan 1977). The general antero-grade amnesia of fear after TTX infusion might be due tothe fact that these connections are completely inactivatedby TTX. Since muscimol only increases inhibition ofneurons in the ventral hippocampus that are subject toinhibition via GABAA receptors, it may have more spe-cific effects on fear conditioning than TTX. Neverthe-less, formation of fear conditioning to a context was dis-rupted by muscimol infusion into the ventral hippocam-pus in the present study, indicating that neurons withinthe ventral hippocampus bearing the GABAA receptorare at least important for fear conditioning to a context.There is evidence for the hippocampal network of

GABA-releasing inhibitory interneurons being importantfor memory formation (Buzski and Chrobak 1995). TheGABAA agonist muscimol interferes with this process-ing. Therefore, it is possible that the impaired formationof fear memory to a context induced by muscimol in theventral hippocampus does not primarily result from aninhibition of neuronal activity per se. Rather, the antero-grade amnesia of fear to a context caused by muscimolin the ventral hippocampus might be due to the fact thatundisturbed signaling in the local network of GABA-releasing interneurons is necessary for the specific pro-cessing of contextual information which is required forforming a memory of fear to a context.

A prominent notion of the hippocampuss role in clas-sical conditioning, and in particular in fear conditioning,is that the hippocampus processes contextual informationnecessary to form or utilize particular CS-US associa-tions which are formed and stored elsewhere in the brain(Fanselow 2000; Fendt and Fanselow 1999; Holland andBouton 1999; Jarrard 1993; Maren and Fanselow 1995;Maren and Holt 2000). In fact, disrupting activity of thedorsal hippocampus by lesions or infusions has mainlybeen reported to impair fear conditioning to a contextualCS but not to a tone (for review, see Fanselow 2000).Our present results and previous lesion studies (Maren1999; Richmond et al. 1999), however, suggest that nor-mal activity of the ventral hippocampus is important forthe formation of fear conditioning not only to a contextbut also to a tone.

TTX and muscimol infusion into the ventral hippo-campus might disrupt the specific associative processesunderlying the formation of fear conditioning by inter-fering with the sending or receiving of neuronal signalsby the ventral hippocampus. For example, deficient ac-tivity of the ventral hippocampus during fear condition-ing might prevent memory formation by disrupting sig-naling to the amygdala or the nucleus accumbens, or bydisrupting the reception of inputs from the amygdala, or

49


12/14

by disrupting reciprocal signaling between dorsal andventral hippocampus. At the moment, it is not possible todecide which interactions are important. Disconnectionlesions (Floresco et al. 1997) might help to identify thecritical connections. Moreover, to understand the interac-tions in this network, future studies are necessary whichfurther characterize or start to examine the neurochemi-cal and neurophysiological changes in the nucleus

accumbens (Murphy et al. 2000; Pezze et al. 2000;Wilkinson et al. 1998; Young et al. 1993), the amygdala,and the hippocampus (Nail-Boucherie et al. 2000) duringfear conditioning, and which clarify how neurochemicaland neurophysiological changes in these structures arerelated to each other. Experiments to address these issuesare currently being run or are in preparation in our labo-ratory.

Finally, the present study did not address the issue ofwhether the importance of the ventral hippocampus forfear memory in a classical conditioning paradigm is tem-porary, i.e., restricted to the formation (acquisition andconsolidation) of the memory, or whether the ventral

hippocampus is permanently necessary for the retrievalof the memory. Currently, there is a debate as to thequestion of whether the hippocampuss relevance for dif-ferent forms of memory is time limited or not (Knowltonand Fanselow 1998; Nadel and Moscovich 1997; Nadelet al. 2000). The role of the dorsal hippocampus in clas-sical fear conditioning to a contextual CS has recentlybeen suggested to be time limited (Anagnostaras et al.1999). In the ventral hippocampus this issue is still unex-plored. TTX infusion into the ventral and dorsal hippo-campus has been reported to impair retrieval of the pas-sive avoidance response memory, with the impairmentbeing more pronounced after TTX infusion into the dor-

sal hippocampus (Ambrogi Lorenzini et al. 1996, 1997).Using freezing as a measure of conditioned fear, the ef-fects of inactivation of the ventral hippocampus by le-sions or inactivating infusions on memory retrieval in aclassical fear-conditioning paradigm will be difficult toexamine, since both manipulations influence activity andthus affect freezing.

In conclusion, the results of the present study indi-cate that neuronal activity within the ventral hippocam-pus is necessary for the formation of fear memory to ex-plicit and contextual cues by classical conditioning. Thecomplete anterograde amnesia of fear after TTX inacti-vation of the ventral hippocampus suggests that the ven-tral hippocampus contains neurons that are importantfor the formation of fear memory to contextual as wellas to explicit cues. The anterograde amnesia of fear to acontext after infusion of muscimol into the ventral hip-pocampus indicates that neurons bearing the GABAA re-ceptor are among those neurons in the ventral hippo-campus which are involved in the formation of fear con-ditioning to a context. In addition, the marked hypoac-tivity following muscimol and TTX infusion into theventral hippocampus further corroborates that neuronalactivity within the ventral hippocampus drives locomo-tor activity.

Acknowledgements This work was supported by grants from theSwiss Federal Institute of Technology, Zurich. The authors thankthe Animal Facility team for their care of the animals, Ms. LizWeber for her histological preparations, Mr. Peter Schmid for thesetup and maintenance of the computerized systems for behavioralanalysis, and Ms. Bonnie Strehler for her assistance with manu-script preparation.

References

Ambrogi Lorenzini C, Baldi E, Bucherelli C, Sacchetti B, TassoniG (1996) Role of the dorsal hippocampus in acquisition, con-solidation and retrieval of rats passive avoidance response: atetrodotoxin functional inactivation study. Brain Res 730:3239

Ambrogi Lorenzini CG, Baldi E, Bucherelli C, Sacchetti B,Tassoni G (1997) Role of the ventral hippocampus in acquisi-tion, consolidation and retrieval of rats passive avoidance re-sponse memory trace. Brain Res 768:242248

Ambrogi Lorenzini CG, Baldi E, Bucherelli C, Sacchetti B,Tassoni G (1999) Neural topography and chronology of mem-ory consolidation: a review of functional inactivation findings.Neurobiol Learn Mem 71:118

Anagnostaras SG, Maren S, Fanselow MS (1999) Temporallygraded retrograde amnesia of contextual fear after hippocam-

pal damage in rats: within-subjects examination. J Neurosci19:11061114

Bardgett ME, Henry JD (1999) Locomotor activity and accumb-ens Fos expression driven by ventral hippocampal stimulationrequire D1 and D2 receptors. Neuroscience 94:5970

Bast T, Zhang W-N, Feldon J (2001) Hyperactivity, decreasedstartle reactivity, and disrupted prepulse inhibition followingdisinhibition of the rat ventral hippocampus by the GABAA re-ceptor antagonist picrotoxin. Psychopharmacology (in press)

Bast T, Zhang W-N, Heidbreder C, Feldon J (2001) Hyperactivityand disruption of prepulse inhibition induced by N-methyl-D-aspartate stimulation of the ventral hippocampus and the ef-fects of pretreatment with haloperidol and clozapine. Neuro-science 103:325335

Blaha CD, Yang CR, Floresco SB, Barr A, Phillips AG (1997)Stimulation of the ventral subiculum of the hippocampus

evokes gluatamate receptor-mediated changes in dopamine ef-flux in the rat nucleus accumbens. Eur J Neurosci 9:902911

Brudzynski SM, Gibson CJ (1997) Release of dopamine in the nu-cleus accumbens caused by stimulation of the subiculum infreely moving rats. Brain Res Bull 42:303308

Buhl EH, Halasy K, Somogyi P (1994) Diverse sources of hippo-campal unitary inhibitory postsynaptic potentials and the num-ber of synaptic release sites. Nature 368:823828

Buzski G, Chrobak JJ (1995) Temporal structure in spatially or-ganized neuronal ensembles: a role for interneuronal net-works. Curr Opin Neurobiol 5:504510

Cahill L, Weinberger NM, Roozendaal B, McGaugh JM (1999) Isthe amygdala a locus of conditioned fear? Some questionsand caveats. Neuron 23:227228

Eichenbaum H, Dudchenko P, Wood E, Shapiro M, Tanila H(1999) The hippocampus, memory, and place cells: is it spatial

memory or a memory space? Neuron 23:206226Fanselow M (1982) The postshock activity burst. Anim Learn

Behav 10:448454Fanselow M (1984) What is conditioned fear? Trends Neurosci

7:460462Fanselow M (1986) Associative vs topographical accounts of the

immediate shock-freezing deficit in rats: implications for theresponse selection rules governing species-specific defensivereactions. Learn Motiv 17:1639

Fanselow M (2000) Contextual fear, gestalt memories, and thehippocampus. Behav Brain Res 110:7381

Fanselow MS, LeDoux JE (1999) Why we think plasticity under-lying Pavlovian fear conditioning occurs in the basolateralamygdala. Neuron 23:229232

50


13/14

Fendt M, Fanselow MS (1999) The neuroanatomical and neuro-chemical basis of conditioned fear. Neurosci Biobehav Rev23:743760

Floresco SB, Seamans JK, Phillips AG (1997) Selective roles forhippocampal, prefrontal cortical, and ventral striatal circuits inradial-arm maze tasks with or without delay. J Neurosci17:18801890

Gewirtz JC, McNish KA, Davis M (2000) Is the hippocampusnecessary for contextual fear conditioning? Behav Brain Res110:8395

Givens BS, Olton DS (1990) Cholinergic and GABAergic modu-lation of medial septal area effects on working memory.Behav Neurosci 104:849855

Gray JA (1995) A model of the limbic system and basal ganglia:Applications to anxiety and schizophrenia. In: Gazzaniga MS(ed). The cognitive neurosciences. MIT Press, Cambridge,MA, pp 11651176

Groenewegen HJ, Vermeulen-Van der Zee E, Te Kortschot A,Witter MP (1987) Organization of the projections from thesubiculum to the ventral striatum in the rat: a study using an-terograde transport of Phaseolus vulgaris-leucoagglutinin.Neuroscience 23:103120

Guarraci FA, Frohardt RJ, Falls WA, Kapp BS (2000) The effectsof intra-amygdaloid infusions of a D2 dopamine receptorantagonist on Pavlovian fear conditioning. Behav Neurosci114:647651

Halim ND, Swerdlow NR (2000) Distributed neurodegenerativechanges 228 days after ventral hippocampal excitotoxic le-sions in rats. Brain Res 873:6074

Haralambous T, Westbrook RF (1999) An infusion of bupivacaineinto the nucleus accumbens disrupts the acquisition but not theexpression of contextual fear conditioning. Behav Neurosci113:925940

Helmstetter FJ, Bellgowan PS (1994) Effects of muscimol appliedto the basolateral amygdala on acquisition and expression ofcontextual fear condtioning in rats. Behav Neurosci 108:10051009

Holland PC, Bouton ME (1999) Hippocampus and context in clas-sical conditioning. Curr Opin Neurobiol 9:195202

Jarrard LE (1993) On the role of the hippocampus in learning andmemory in the rat. Behav Neural Biol 60:926

Kimble DP (1968) Hippocampus and internal inhibition. Psychol

Bull 70:285295Knowlton BJ, Fanselow MS (1998) The hippocampus, consolida-tion and on-line memory. Curr Opin Neurobiol 8:293296

Legault M, Wise RA (1999) Injections ofN-methyl-D-aspartate in-to the ventral hippocampus inrease extracellular dopamine inthe ventral tegmental area and the nucleus accumbens. Syn-apse 31:241249

Legault M, Rompr P-P, Wise RA (2000) Chemical stimulation ofthe ventral hippocampus elevates nucleus accumbens dopa-mine by activating dopaminergic neurons of the ventral teg-mental area. J Neurosci 20:16351642

Lipska BK, Jaskiw GE, Chrapust S, Karoum F, WeinbergerDR (1992) Ibotenic acid lesion of the ventral hippocampusdifferentially affects dopamine and its metabolites in thenucleus accumbens and prefrontal cortex of rats. Brain Res585:16

Mao J-B, Robinson JK (1998) Microinjection of GABA-A agonistmuscimol into the dorsal but not the ventral hippocampusimpairs non-mnemonic measures of delayed non-matching-to-position performance in rats. Brain Res 784:139147

Maren S (1999) Neurotoxic or electrolytic lesions of the ventralsubiculum produce deficits in the acquisition and expressionof Pavlovian fear conditioning in rats. Behav Neurosci 113:283290

Maren S, Fanselow MS (1995) Synaptic plasticity in the basolater-al amygdala induced by hippocampal formation stimulation invivo. J Neurosci 15:75487564

Maren S, Holt W (2000) The hippocampus and contextual memo-ry retrieval in Pavlovian conditioning. Behav Brain Res 110:97108

Maren S, Anagnostaras SG, Fanselow MS (1998) The startled sea-horse: is the hippocampus necessary for contextual fear condi-tioning? Trends Cogn Sci 2:3942

Martin JH (1991) Autoradiographic estimation of the extent of re-versible inactivation produced by microinjection of lidocaineand muscimol in the rat. Neurosci Lett 127:160164

McGaugh JL (2000) Memory a century of consolidation. Sci-ence 287:248251

Mitchell SN, Yee BK, Feldon J, Gray JA, Rawlins JN (2000) Acti-vation of the retrohippocampal region in the rat causes dopa-

mine release in the nucleus accumbens: disruption by fornixsection. Eur J Pharmacol 407:131138Muller J, Corodimas KP, Fridel Z, LeDoux JE (1997) Functional

inactivation of the lateral and basal nuclei of the amygdala bymuscimol infusion prevents fear conditioning to an explicitconditioned stimulus and to contextual stimuli. Behav Neuro-sci 111:683691

Murphy CA, Pezze MA, Feldon J, Heidbreder C (2000) Differen-tial involvement of dopamine in the shell and core of thenucleus accumbens in the expression of latent inhibitionto an aversively-conditioned stimulus. Neuroscience 97:469477

Nadel L, Moscovitch M (1997) Memory consolidation, retrogradeamnesia and the hippocampal complex. Curr Opin Neurobiol7:217227

Nadel L, Samsonovich A, Ryan L, Moscovich M (2000) Multiple

trace theory of human memory: computational, neuroimaging,and neuropsychological results. Hippocampus 10:352368Nail-Boucherie K, Dourmap N, Jaffard R, Costenin J (2000) Con-

textual fear conditioning is associated with an increase of ace-tylcholine release in the hippocampus of rat. Cogn Brain Res9:193197

OKeefe J, Nadel L (1978) The hippocampus as a cognitive map.Oxford University Press, Oxford

Odling-Smee FJ (1978) The overshadowing of background stimuliby an informative CS in aversive Pavlovian conditioning inrats. Anim Learn Behav 6:4351

Overton DA (1964) State-dependent or dissociated learningproduced with pentobarbital. J Comp Physiol Psychol 57:312

Parkinson JA, Robbins TW, Everitt BJ (1999) Selective excito-toxic lesions of the nucleus accumbens core and shell differen-

tially affect aversive Pavlovian conditioning to discrete andcontextual cues. Psychobiology 27:256266Paxinos G, Watson C (1998) The rat brain in stereotaxic coordi-

nates, 4th edn. Academic, San DiegoPezze M-A, Heidbreder C, Feldon J, Murphy C (2000) Selective

responding of nucleus accumbens core and shell dopamine toaversively conditioned contextual and discrete stimuli (ab-stract). Eur J Neurosci [Suppl 11] 12:87

Pitknen A, Pikkarainen M, Nurminen N, Ylinen A (2000) Recip-rocal connections between the amygdala and the hippocampalformation, perirhinal cortex, and postrhinal cortex in rats. Areview. Ann NY Acad Sci 911:369391

Pryce CR, Lehmann J, Feldon J (1999) Effect of sex on fear con-ditioning is similar for context and discrete CS in Wistar,Lewis and Fischer rat strains. Pharmacol Biochem Behav64:753759

Richmond MA, Murphy CA, Pouzet B, Schmid P, Rawlins JNP,Feldon J (1998) A computer controlled analysis of freezingbehaviour. J Neurosci Methods 86:9199

Richmond MA, Yee BK, Pouzet B, Veenman L, Rawlins JNP,Feldon J, Bannerman DM (1999) Dissociating context andspace within the hippocampus: effects of complete, dorsal andventral excitotoxic lesions on conditioned freezing and spatiallearning. Behav Neurosci 113:11891203

Riedel G, Harrington NR, Hall G, MacPhail EM (1997) Nucleusaccumbens lesions impair context, but not cue, conditioning inrats. Neuroreport 8:24772481

Rudy JW, Sutherland RJ (1995) Configural association theory andthe hippocampal formation: an appraisal and reconfiguration.Hippocampus 5:375389

51


14/14

Schwarting R, Carey RJ (1985) Deficits in inhibitory avoidanceafter neurotoxic lesions of the ventral striatum are neurochem-ically and behaviorally selective. Behav Brain Res 18:279283

Swanson LW, Cowan WM (1977) An autoradiographic study ofthe organization of the efferent connections of the hippocam-pal formation in the rat. J Comp Neurol 172:4984

Wallenstein GV, Eichenbaum HW, Hasselmo ME (1998) The hip-pocampus as an associator of discontiguous events. TrendsNeurosci 21:317323

Westbrook RF, Good AJ, Kiernan M (1997) Microinjection ofmorphine into the nucleus accumbens impairs contextuallearning in rats. Behav Neurosci 111:9961013

Wilkinson LS, Humby T, Killcross AS, Torres EM, Everitt BJ,Robbins TW (1998) Dissociations in dopamine release in me-dial prefrontal cortex and ventral striatum during the acquisi-tion and extinction of classical aversive conditioning in the rat.Eur J Neurosci 10:10191026

Wu M, Brudzynski SM (1995) Mesolimbic dopamine terminalsand locomotor activity induced from the subiculum. Neuro-report 6:16011604

Yang CR, Mogenson GJ (1987) Hippocampal signal transmissionto the pedunculopontine nucleus and its regulation by dopa-mine D2 receptors in the nucleus accumbens: an electrophysi-ological and behavioral study. Neuroscience 23:10411055

Young AMJ, Joseph MH, Gray JA (1993) Latent inhibition ofconditioned dopamine release in rat nucleus accumbens.Neuroscience 54:59

Zhang W-N, Bast T, Feldon J (2000) Microinfusion of the non-competitiveN-methyl-D-aspartate antagonist MK-801 (dizocil-pine) into the dorsal hippocampus of Wistar rats does not af-fect latent inhibition and prepulse inhibition, but increasesstartle reaction and locomotor activity. Neuroscience 101:589599

Zhuravin IA, Bures J (1991) Extent of tetrodotoxin induced block-ade examined by pupillary paralysis elicited by intracerebralinjection of the drug. Exp Brain Res 83:687690

52

Tobias Bast et al- The ventral hippocampus and fear conditioning in rats: Different anterograde...

Documents

Transcript of Tobias Bast et al- The ventral hippocampus and fear conditioning in rats: Different anterograde...