Exp Brain Res (2003) 153: 400–402DOI 10.1007/s00221-003-1704-1

RESEARCH NOTE

César Quiroz . Isabel Martínez . Gina L. Quirarte .Teresa Morales . Sofía Díaz-Cintra .Roberto A. Prado-Alcalá

Enhanced inhibitory avoidance learning preventsthe memory-impairing effects of post-training hippocampalinactivationReceived: 10 February 2003 / Accepted: 2 September 2003 / Published online: 14 October 2003# Springer-Verlag 2003

Abstract Rats were trained on an inhibitory avoidancetask to study the effects of post-training administration oftetrodotoxin (TTX, which temporarily inactivates neuralactivity) on memory consolidation. During training,independent groups of rats received either a mild footshock (0.8 mA) or a stronger (1.0 mA) foot shock. TTXwas administered bilaterally into the dorsal hippocampusimmediately after training, and memory of the task wasmeasured 48 h later. We corroborated the typical amnesiceffect of intrahippocampal infusions of TTX in those ratstrained with the mild-intensity foot shock. More import-antly, with the stronger foot shock, the same treatment wasineffective in producing amnesia. These results suggestthat, after an enhanced learning experience, other brainregions are also activated, which may compensate for theamnesic effect of TTX infusions into the hippocampus.

Keywords Memory . Retention . Hippocampus .Tetrodotoxin . Inhibitory avoidance . Overlearning

Introduction

Memory is impaired when normal activity of any of anumber of cerebral structures is disrupted after a learningexperience. Growing evidence indicates, however, thatthere is no such interference when the learning experienceis significantly enhanced (Prado-Alcalá 1995). Forexample, it has been shown that intrastriatal administrationof drugs that produce retrograde amnesia of inhibitoryavoidance training becomes ineffective in animals trainedwith higher foot-shock intensities (Giordano and Prado-Alcalá 1986; Pérez-Ruiz and Prado-Alcalá 1989). Equiv-

alent protective effects of heightened inhibitory avoidancelearning have been found after interruption of synapticactivity of the amygdala (Parent et al. 1992, 1994) andsubstantia nigra (Cobos-Zapiaín et al. 1996).

Memory consolidation of inhibitory avoidance trainingalso involves activity of the hippocampus, as shown by thedeleterious effects of interference with hippocampalactivity on memory of this task (Izquierdo et al. 1992;Lorenzini et al. 1996; Stubley-Wheatherly et al. 1996;Martínez et al. 2002). The protective effect of enhancedtraining, however, has not been studied in relation to thehippocampus. We now report that a relatively smallincrement in the aversive stimulation used during inhib-itory avoidance training prevents the amnesic state that isusually produced by temporary inactivation of this struc-ture.

Material and methods

Naive male Wistar rats weighing between 250 and 350 g were used.They were individually housed and had free access to solid food andtap water in their home cages. Under Nembutal anesthesia (40 mg/kg), bilateral cannulae (23-gauge, stainless steel) were implanted,aimed at either the dorsal hippocampus (DH; AP=−4.1, L=±2.5, V=−3.1) or the parietal cortex, overlying the hippocampal placements(AP=−4.1, L=±2.5, V=−0.5; Paxinos and Watson 1982). All animalprocedures and the experimental work were approved by the ethicscommittee of the Institute of Neurobiology, National University ofMexico.The conditioning box, training, and testing procedures have been

described in detail elsewhere (Martínez et al. 2002). Briefly, the boxconsisted of two compartments of the same size (30×30×30 cm),separated by a guillotine door. The safe compartment was illumi-nated by a 10-W light bulb. The V-shaped lateral walls of the secondcompartment were stainless steel, and each wall was continuouswith half of the floor, which was connected in series to a square-pulse stimulator and constant current unit.During training, 1 week after surgery, each rat was placed inside

the safe compartment. Ten seconds later the door between thecompartments was opened, and the rat’s latency to enter the darker(electrifiable) compartment with all four paws was measured. Oncein this compartment, the door was closed and a mild (0.8 mA) or astronger (1.0 mA) foot shock was delivered; after 5 s the door wasreopened, allowing the rat to escape to the safe compartment and to

C. Quiroz . I. Martínez . G. L. Quirarte . T. Morales .S. Díaz-Cintra . R. A. Prado-Alcalá (*)Instituto de Neurobiología, Universidad Nacional Autónoma deMéxico,PO Box 70-228, 04510 Mexico D.F., Mexicoe-mail: prado@servidor.unam.mxTel.: +1-5255-56234047Fax: +1-5255-56234046

remain there for 30 s before drug treatment and being placed back inits home cage. The escape latency (i.e., latency to return to the safecompartment) was recorded. Retention testing was performed 48 hlater with procedures identical to those of the training session,except that no foot shock was delivered. Latency to enter the shockcompartment was used as a measure of retention. If a rat did notcross within 200 s to the shock compartment, the session was ended,and a score of 200 was assigned.A single, bilateral intrahippocampal infusion of TTX (10 ng;

Sigma Chemical) or of isotonic saline solution was administeredimmediately after training. The same dose of the toxin or the vehiclewere also bilaterally infused into the parietal cortex of control rats.The infusions (1 μl/hemisphere) were given at a rate of 1 μl/min.With the dose and volume used here, TTX suppresses neuronalactivity for about 24 h (Zhuravin and Bures 1991).Upon completion of the experiment, all rats were deeply

anesthetized with Nembutal, perfused with isotonic saline followedby 10% formalin, and their brains removed and kept in formalin forat least 1 week before frozen coronal sections (50 µm thick) werecut and stained (Nissl method) to determine the location of cannulaetips. Animals in which the cannulae tips were outside the targetareas were not further analyzed. The final sample size across groupsvaried from eight to ten rats. We used independent Mann-WhitneyU-tests to compare training, and escape and retention scores of eachof the TTX groups against its respective control (vehicle) group.Although several studies have reported the spread of TTX in the

area surrounding the nucleus of Edinger-Westphal rats (Zhuravinand Bures 1991) or in the cerebral cortex of raccoons (Boehnke andRasmusson 2001), they have not predicted the extent of spread ofTTX in the hippocampus of rats. Thus, the spreading of TTX awayfrom the cannula tip and its inactivating effects in the DH weredetermined at 30 and 90 min and at 48 h after microinjection, usingimmunohistochemistry to detect induced c-fos expression. As thebasal expression of this protein is not sufficient to allow clearevidence of inactivation (Herrera and Robertson 1996), the expres-sion of c-fos was increased using a kainic acid induction method(Willoughby et al. 1997). One week after surgery, the rats wereinjected with TTX (10 ng/μl per 1 min) into either the left or rightDH and with vehicle (1 μl/1 min) into the opposite DH; thus, eachanimal served as its own control. The rats were killed at 30 min,90 min, or 48 h after brain injections, but 10 min beforehand the ratswere intraperitoneally injected with kainic acid (50 mg/kg in PBS)to increase the expression of hippocampal c-fos.At the prescribed time of death, rats were deeply anesthetized

with chloral hydrate (350 mg/kg i.p.) and perfused via the ascendingaorta with saline followed by 4% paraformaldehyde (pH 9.5, 10°C).Brains were then post-fixed for 3 h and cryoprotected in 10%sucrose in 0.1 M phosphate buffer. Series of 30-μm-thick frozen

coronal sections throughout the length of the brain were collectedand stored in cryoprotectant (30% ethylene glycol and 20% glycerolin 0.05 M sodium phosphate buffer) at −20°C until histochemicalprocessing. Fos immunoreactivity was detected using a conven-tional, nickel-intensified, avidin-biotin-immunoperoxidase proce-dure (Vectastain Elite ABC kit; Vector Laboratories) to localize aprimary antiserum (1:5,000) raised against the N-terminal fragmentof the human Fos protein (Santa Cruz Biotechnology). Tissuepretreatments with H2O2 and NaBH4 were performed beforeincubation (4°C for 48 h) with the primary antiserum.

Results

There were no significant differences in training andescape latencies between the experimental and control

401

Fig. 1 Median retention scores and interquartile range of groups ofrats trained on 1-trial inhibitory avoidance and injected with eithervehicle (VEH; open bars) or tetrodotoxin (TTX; 10 ng; filled bars)into the parietal cortex (CX) or dorsal hippocampus (DH). TTXinfusion into the DH produced amnesia in the group trained with0.8 mA, but was innocuous when administered to the DH grouptrained with 1.0 mA or to the cortical group trained with 0.8 mA

Fig. 2A-C In each animal, TTX was infused into either the left orright hippocampus while isotonic saline was infused into theopposite hippocampus. TTX abolished kainic acid-induced c-fosexpression at 30 min (A) and 90 min (B) after infusion, and a totalrecovery was seen at 48 h after infusion (C). The upper insets aremagnifications (×200) of the corresponding side of the dentate gyrus(A), showing the intensity of the c-fos immunoreactivity. ×20

402

groups. When training was conducted with the lower foot-shock, infusion of TTX into the DH produced a significantretention deficit (P<0.02 versus low foot-shock, DH-vehicle group). However, administration of TTX into theDH of the group that had been trained with 1.0 mA causedno impairment of retention. The administration of thevehicle into the DH or of TTX into the parietal cortex didnot produce disturbances of retention, regardless of foot-shock intensity (Fig. 1).

Injections of TTX produced a profound inhibition ofkainic acid-induced c-fos expression in the DH at 30 and90 min after infusion of the toxin. The inactivationextended from −2 mm to −4.5 mm from Bregma across thelongitudinal axis of the DH. Forty-eight hours later, suchinhibition could not be detected (Fig. 2).

Discussion

The present findings show that post-training infusions ofTTX into the DH impair inhibitory avoidance retention.More importantly, increased foot-shock intensity duringinhibitory avoidance training prevents the amnesic effectsof the hippocampal inactivation. The present results alsoshow that the infusions of TTX into the DH impair neuralactivity, as measured by a decrease in kainic-acid inducedc-fos activation, for at least the first 90 min after injectionof TTX. c-fos expression returns to control levels within48 h after the TTX injection. Thus, it is likely that theimpaired retention seen in the animals that were trainedwith the lower foot-shock intensity was due to interferencewith the consolidation process. These data are consistentwith previous findings showing that interference withhippocampal functioning during the consolidation periodproduces retrograde amnesia (Lorenzini et al. 1996). Ourresults extend these previous findings by showing thatincreased foot-shock intensity completely prevents theretrograde amnesia. We have observed similar resultsfollowing post-training reversible inactivation of thestriatum (Pérez-Ruiz and Prado-Alcalá 1989). Thesefindings are also congruent with previous results showingthat inactivation of the amygdala basolateral complexeither immediately or 6 h post-training impairs inhibitoryavoidance when rats are trained with a low foot-shock.However, with a higher foot-shock intensity, amnesia isobserved only after the immediate post-training inactiva-tion (Parent and McGaugh 1994). These authors suggestedthat the higher foot-shock during training may acceleratethe consolidation process and decrease the duration ofinvolvement of the basolateral amygdala in consolidation.Applied to the present findings, this interpretation raisesthe possibility that consolidation may have been com-pleted by the time the TTX was infused into the DHimmediately after training with the higher foot-shock, orthat at least the acquisition of spatial/contextual informa-tion which is essential for storage of this information tookplace during this short interval. Finally, in agreement withLorenzini et al. (1996), our data suggest that the hippo-campus is not essential for long-term storage of strongemotional information, because after rapid consolidationthis information is probably conveyed to other brainstructures. The recruitment of these various structures

(Ambrogi Lorenzini et al. 1999) enables, therefore, theaccess to this information independently from the activa-tion state of the hippocampus.

Acknowledgements The authors thank Drs. Marise B. Parent andBenno Roozendaal for their helpful comments on the manuscript,and the technical assistance of Dorothy Pless, Angel Mendez,Azucena Aguilar, and M. V. Z. Martín García. Supported by PAPIIT-UNAM.

References

Ambrogi Lorenzini CG, Baldi E, Bucherelli C, Sacchetti B, TassoniG (1999) Neural topography and chronology of memoryconsolidation: a review of functional inactivation findings.Neurobiol Learn Mem 71:1−18

Boehnke S, Rasmusson D (2001) Time course and effective spreadof lidocaine and tetrodotoxin delivered via microdialysis: anelectrophysiological study in cerebral cortex. J NeurosciMethods 105:133–141

Cobos-Zapiaín GG, Salado-Castillo R, Sánchez-Alavez M, QuirarteGL, Roldán-Roldán G, Prado-Alcalá RA (1996) High level offoot shock during inhibitory avoidance training preventsamnesia induced by intranigral GABA antagonists. NeurobiolLearn Mem 65:202–206

Giordano M, Prado-Alcalá RA (1986) Retrograde amnesia inducedby post-trial injection of atropine into the caudate-putamen.Protective effect of the negative reinforcer. Pharmacol BiochemBehav 24:905–909

Herrera D, Robertson H (1996) Activation of c-fos in the brain.Progr Neurobiol 50:83–107

Izquierdo I, Da Cunha C, Rosat R, Jerusalinsky D, Ferreira MBC,Medina JH (1992) Neurotransmitter receptors involved inmemory processing by the amygdala, medial septum andhippocampus of rats. Behav Neural Biol 58:16–25

Lorenzini CA, Baldi E, Bucherelli C, Sacchetti B, Tassoni G (1996)Role of dorsal hippocampus in acquisition, consolidation andretrieval of rat’s passive avoidance response: a tetrodotoxinfunctional inactivation study. Brain Res 730:32–39

Martínez I, Quirarte GL, Díaz-Cintra S, Quiroz C, Prado-Alcalá RA(2002) Effects of lesions of hippocampal fields CA1 and CA3on acquisition of inhibitory avoidance. Neuropsychobiology46:97–103

Parent MB, McGaugh JL (1994) Posttraining infusion of lidocaineinto the amygdala basolateral complex impairs retention ifinhibitory avoidance training. Brain Res 661:97–103

Parent MB, Tomaz C, McGaugh JL (1992) Increased training in anaversively motivated task attenuates the memory-impairingeffects of posttraining N-methyl-D-aspartate-induced amygdalalesions. Behav Neurosci 106:789–797

Parent MB, West M, McGaugh JL (1994) Memory of rats withamygdala lesions induced 30 days after foot shock-motivatedescape training reflects degree of original training. BehavNeurosci 108:1080–1087

Paxinos G, Watson C (1982) The rat brain in stereotaxic coordinates.Academic, New York

Pérez-Ruiz C, Prado-Alcalá R A (1989) Retrograde amnesiainduced by lidocaine injection into the striatum: protectiveeffect of the negative reinforcer. Brain Res Bull 22:599–603

Prado-Alcalá RA (1995) Serial and parallel processing duringmemory consolidation. In: McGaugh JL, Bermúdez-Rattoni F,Prado-Alcalá RA (eds) Plasticity in the central nervous system.Learning and memory. Lawrence Erlbaum, N.J., pp 57–65

Stubley-Wheatherly L, Harding JW, Wrigh JW (1996) Effects ofdiscrete kainic acid-induced hippocampal lesions on spatial andcontextual learning and memory in rats. Brain Res 716:29–38

Willoughby J, Mackenzie L, Medvedev A, Hiscock J (1997) Fosinduction following systemic kainic acid: early expression inhippocampus and later widespread expression correlated withseizure. Neuroscience 77:379–392

Zhuravin Y, Bures J (1991) Extent of the tetrodotoxin inducedblockade examined by pupillary paralysis elicited by intrace-rebral injection of the drug. Exp Brain Res 83:687–690