Localization in the amygdala of the amnestic action of diazepam on emotional memory

Behavioural Brain Research, 58 (1993) 99-105 © 1993 Elsevier Science Publishers B.V. All rights reserved. 0166-4328/93/$06.00

99

BBR 1522

Localization in the amygdala of the amnestic action of diazepam on emotional memory C. Tomaz a,,, H. Dickinson-Anson b, J.L. McGaugh b, M.A. Souza-Silva a, M.B. Viana a, F.G. Graeff a aLaboratory of Psychobiology, FFCLRP and Center for Neuroscience and Behavior of the University of Sgw Paulo, Ribeirglo Preto, SP 14040-901 (Brazil), b Center for the Neurobiology of Learning and Memory and Department of Psychobiology, University of California, Irvine, CA (USA) (Received 20 December 1992) (Revised version received 6 August 1993) (Accepted 6 August 1993)

Key words: Memory; Amygdala; Anxiety; Diazepam; Aversive learning; Amnesia; Lesion; Microinjection; T maze

It is well known that systemically administered benzodiazepines (BZDs) induce anterograde amnesia in a variety of learning tasks. BZs effects are mediated through the GABA A complex by enhancing GABA-induced synaptic inhibition. As the GABAergic system in the amygdaloid complex (AC) is a site of action for the anxiolytic effects of BZs, such findings suggest that BZs may also influence memory through the amygdala. The present report summarizes a recent serie of experiments designed to examine this implication. In a first experiment rats received either sham or bilateral AC lesion using N-methyl-D-aspartic acid (NMDA). One week later, animals were trained on an inhibitory avoidance task and tested 48 h later. Diazepam (DZP; 1.0 and 2.0 mg/kg, i.p.) or vehicle was injected 30 min prior to acquisition. The results demonstrate that DZPinduced retention deficits was blocked in rats with AC lesions. In a second experiment, in an attempt to localize the site of BZDs amnestic action in the AC, we tested the effects of DZP in rats with bilateral ibotenic acid-induced lesions of central (CE), lateral (LAT) or basolateral (BL) amygdala nuclei. The results shown that retention was impaired in animals with CE and LAT lesions but not in animals with BL lesions. In a third experiment we tested the effects of DZP microinjections in different nuclei of the AC on retention performance of rats trained in an avoidance task. The results demonstrate that DZP microinjection prior training in the BL/LAT, but not CE nuclei produce anterograde amnesia. Taken together, these findings strongly suggest that the amnestic effects of DZP are mediate, at least in part, through influences involving the AC, in particular the BL amygdala nucleus. In addition, in a further study, some of us developed a new experimental model for measuring simultaneously drug effects on anxiety and memory: the elevated T maze. In rats placed in this T maze, consisting of an enclosed arm at a right angle with open arms elevated 50 cm above the ground, DZP (2 and 4 mg/kg, i.p.) abolished the delay of withdrawal from the enclosed arm towards the open arms in tests immediately after the training as well as 72 h later, suggesting both an anxiolytic and an amnestic effect in this inhibitory avoidance paradigm. In contrast, in the same subject, DZP did not affect the latency of withdrawal from one of the open arms towards the closed arm. Moreover, on the 72 h later test, the latency of this escape response decreased in all groups, indicating that memory of this task was resistant to DZP. These results further support the view that the anxiolytic and the amnestic effects of BZs are closely related.

INTRODUCTION

Several lines of evidence indicate that the amygdaloid complex (AC) is involved in the modulation of memory storage, particularly in emotional based memory7'irA8. Anatomical studies have shown that the AC receives projections from all sensory modalities, the lateral/basolateral nuclei being the major site of convergence, while efferent fibers depart from the central nucleus to the medial hypothalamus and the midbrain periaqueductal gray matter, as well as to the thalamus, neocortex, hippocampus, striatum, and cells groups of * Corresponding author. Fax: (55) (16) 633-5015.

the basal forebrain 1'z'8'12. This anatomical characteristic of the AC has led to the proposal that it is involved in linking information about the sensory aspects of the stimuli processed by the neocortex with the fundamental motivational drive mechanisms and emotions 2°. Thus, the influence of the AC on learning and memory processes seems to be close related to its role in emotional sphere. Studies examining the memory-modulating effects of drugs treatments have provided evidence that memory can be modulated by systemic as well as intra-amygdala GABAergic compounds. When administered shortly after training, GABAergic agonists (e.g. muscimol and baclofen) impair memory retention while GABAergic

100 antagonists (e.g. picrotoxin and bicuculline) enhance r e t e n t i o n 4-6,9.

On the other hand, there is extensive evidence indicating a key role for the GABA neurotransmission in the modulation of fearful/defensive behaviors. It has been found that systemic or intra-amygdala injections of GABA agonists reduces and GABA antagonists enhances experimental fear and anxiety ~4. Furthermore, lesions of the amygdala attenuate the anti-anxiety as well as the memory-modulating effects of GABAergic drugs 3,26. There is also extensive evidence that benzodiazepines (BZDs), the major class of anxiolytic drugs in therapeutic use, induce anterograde amnesia in both humans and a n i m a l s 19'27. The findings of several experiments suggest that the anxiolytic properties of BZDs involve influences mediated by the amygdala. Intra-amygdala injections of BZDs produce anxiolytic effects comparable to those induced by systemic injections 21'25. Furthermore, intra-amygdala injections of the BZD antagonist flumazenil attenuate the anxiolytic effects of systemically administered B Z D s 16. Recent findings by Izquierdo and colleagues 1°'17 suggest that BZDs impairment of memory involve GABAergic type A receptors in the amygdala. Posttraining intra-amygdala injection of flumazenil causes memory facilitation comparable to that found with systemic injections and systemic injection of flumazenil before training attenuate the amnestic effects of posttraining intra-amygdala injection of muscimol. The evidence summarized above suggest that both the anxiolytic and the amnestic effects of GABA/BZDs are mediated by the amygdala. If this is the case, lesions of the amygdala should block the amnestic effects of BZDs. The aim of this report is to review a recent series of experiments dealing with the effects of systemic and intra-amygdala injections of the BZD diazepam (DZP) on the acquisition and retention of an inhibitory avoidance response in control and amygdala-lesioned rats. In our lesion studies we have adopted the lesioning procedures by the infusion of excitotoxic amino acids into the amygdala in order to minimize damage to fiber bundles passing through it. Additionally, we report the results of a recent study dealing with the development of a new experimental model in an attempt to simultaneously measure drug effects on anxiety and memory.

learning and retention of a continuous multiple trial inhibitory avoidance (CMIA) in rats with bilateral lesions of the amygdala. Amygdala lesions were produced using N-methyl-D-aspartic acid (NMDA). N M D A (8 #/0.8 #1 phosphate buffer) was infused at at rate of 0.2 gl/min, with injection needle left in place 5 min following the infusion. Control animals received either a sham or no lesion at all. One week later, animals were trained to criterion on the CMIA task and tested 48 h later. The CMIA avoidance task has been described in details elsewhere3°. Briefly, the apparatus consisted of a trough-shaped alley with two compartments separated by a sliding door that opened by retracting into the floor. In this task, whenever a rat entered the dark compartment, a footshock was given (maximum of 5 s) until the rat escaped to the lighted safe starting compartment. The numbers of trials required before the rat remained in the safe compartment for 100 s in one trial was recorded. On the 48-h retention test, the rat was placed in the starting (safe) compartment, as on the training session, and the latency to step into the dark compartment (maximum 600 s) was recorded. Diazepam (1.0 or 2.0 mg/kg, i.p,) or vehicle (saline plus Tween 80) were injected 30 min prior to acquisition. The results showed that within each surgery treatment group (i.e. unoperated, sham and lesion) there was no effect of drug treatment on acquisition performance; all groups learned the task to the same criterion level. However, as is shown in Fig. 1, D Z P produced a dose dependent impairment of retention in unoperated and sham rats. These findings are consistent with other evidence indicating that BZDs produce anterograde amnesia without impairing acquisition performance 13. Such DZP-induced amnestic effects were not I ' ~ VEHICLE DZP1 OZPZ~

~

m ~ o A M Y G D A L A LESIONS BLOCK THE A M N E S T I C EFFECTS OF DZP

In the first study of this series w e 3° investigated the effects of pre-training systemic injections of D Z P on

s~m

LSm

Fig. 1. Diazepam effect, in N M D A amygdala-lesioned animals, on retention test conducted 48 h after inhibitory avoidance training using the continuous-trial criterion training. Values represent median (in, terquartile ranges) step-through latencies. Asterisks indicate difference from corresponding vehicle control group, * P < 0.05, • * P < 0.001, Mann-Whitney U-test. From Tomaz et al.3°.

101 observed in amygdala-lesioned groups suggesting that at least part of the amnestic effects of D Z P are mediated through influences involving the amygdala.

BASOLATERAL A M Y G D A L A LESIONS BLOCK DZP AMNESTIC EFFECTS

Several studies have documented that specific nuclei of the amygdala are involved in the anxiolytic properties of BZDs. Injections of the BZDs chlordiazepoxide and midazolam into the lateral (LAT)/basolateral (BL) amygdala induce anxiolytic effects, whereas injections of midazolam into the central (CE) amygdala are without e f f e c t 16'23'24. Such findings are consistent with the evidence that receptors for BZDs are heavily located in the LAT/BL amygdala nuclei22'3~. Taking in account the evidence summarized above that both the amnestic and the anxiolytic effects of BZDs are mediated by the amygdala, and that the LAT/BL nuclei are critically involved in the anxiolytic effects of BZDs, in the second study of this series w e 29 investigated the effects of D Z P treatment in control and amygdala nuclei cell body lesions rats. Bilateral lesions of the CE, LAT, or BL amygdala were made by infusing ibotenic acid (IBO, 0.25 ~tg/0.25 #1 phosphate buffer, over 3.5 min). Sham operations consisted the same general lesion surgery procedure except that the needle was lowered only to the level of the caudate/putamen and them removed immediately without injection of IBO. One week later, the rats were trained and tested as described for the first experiment. D Z P (2.0 mg/kg) or vehicle were systemically injected 30 min before training.

[~

VEHICLE

I"-1 o~ 2 , ~ g 6OO

500

E

4oo

'~

300

2oo 100

O.

m SHAM

CENTRAL

LATERAL 8ASOLATERAL

Fig. 2. Diazepam effect, in rats with IBO lesions of the nuclei of the amygdala, on the retention performance test conducted 48 h after training. Values represent median (interquartile ranges) step-through latencies. Asterisks indicate difference from corresponding vehicle control group (*P<0.05), Mann-Whitney U-tests. From Tomaz et al. 29.

The acquisition data showed that there were no differences among the groups in the number of trials required to reach the acquisition criterion. The retention test latencies are shown in Fig. 2. As can be see, D Z P impaired retention in the sham group, as well as the CE and LAT lesion groups, but did not affect retention of the BL lesion group. That is, lesions of the BL amygdala nucleus completely block the retention impairment produced by systemic injections of D Z P administered prior to training. The anterograde amnesia produced in sham, as well as CE- and LAT-lesioned animals, may be due to activation of BZD receptors in the BL nucleus of the amygdala. These data suggest, therefore, that as observed for the anxiolytic effects of BZDs, the BL amygdala appears to be important for the D Z P influences on memory.

EFFECTS OF DZP M I C R O I N J E C T I O N IN THE A M Y G D A L A NUCLEI

Although the results of the above mentioned lesion studies have suggested a clear role for the amygdala (in particular the BL nucleus) in mediating the amnestic effects of D Z P one could argue about the completeness of the lesions. An alternative approach to search for the brain areas related to D Z P amnestic effects is to test the effects of local microinjections of this drug in different nuclei of the amygdala. Thus, going one step further, in the third study of this series we (in preparation) investigated the effects of D Z P microinjections in different amygdala nuclei on the acquisition and retention of an inhibitory avoidance training using a continuous-trial step-down avoidance procedure. The apparatus consisted of a platform (15 x 11 x 6 cm) fixed in a corner of electrifiable grid floor of a rectangular box (50 x 50 x 37 cm). In this task, whenever the animal step-down the platform it received a scrambled footshock (0.8 mA) until it returned and stay for 200 continuous s on the platform. To examine the effects of D Z P microinjected in amygdaloid nuclei, rats were chronically implanted with stainless-steel guide cannulae into either the CE or LAT/BL amygdala. One week later, they were trained on the continuous multiple-trial step-down avoidance task and tested 48 h later (stepdown latencies). Fifteen minutes prior to training the rats received either a microinjection of D Z P (1.0 #g//~l) or vehicle in the CE or LAT/BL nuclei. Control animals were sham-injected. Fig. 3 shows a representative photomicrograph of the site of infusion in CE or LAT/BL nuclei of the amygdala. Pre-training microinjections of D Z P or vehicle in either amygdaloid nuclei had no effect on acquisition

102

:~

...... : :

Fig. 3. Representative photomicrograph indicating the localization of the injection sites in the lateral/basolateral (A) or central (Bt nuclei of tile amygdala. Injection volume of 0.2/d used in this study would produce average diffusion of about 0.5 ram. The arrows show the tip of the injection needle. 600. 500,

~ 4oo, "~

300,

(:3

"=" 200 z9¢

100 =II

o SI

Veh CENTRAL

DZP

SI

Veh

DZP

LAT/BL

Fig. 4. Effect of diazepam microinjection in different amygdala nuclei. Fifteen min before training animals received 0.2/~1 (0.1 #l/min) of either DZP (1.0 #g/#l) or vehicle, or were sham-injected. Values represent median (interquartile ranges) step-down latencies for the retention test conducted 48 h after training. Asterisk indicates difference from DZP-CE, SI-LAT/BL and Veh-LAT/BL (P<0.01; Mann-Whitney U-test). From M.A. Souza-Silva and C. Tomaz, unpublished findings.

performance; all animals learned the task. In contrast, a retention impairment was observed in animals that received D Z P into the LAT/BL nucleus but not in CE-microinjected animals (Fig. 4). These results strongly supports the previous evidence from our lesion study indicating that the DZP-induced anterograde amnesia for inhibitory avoidance training is mediated, at least in part, through influences involving the basolateral amygdala.

EFFECTS OF DZP ON THE ELEVATED T MAZE TEST

There is considerable experimental and clinical evidence that anxiety and memory are closely related. Thus, inhibitory avoidance, a paradigm widely used in memory studies, is also an anxiety model. The findings of many studies strongly suggest that the brain systems

thought to be involved in anxiety and in the modulation of memory are extensively overlapped 2s. For instance, the results of the above described studies suggest that the memory-modulating effects of BZDs as its anxiolytic effects, seem to involve the same or overlapping neuromechanisms in the amygdala. In spite of this, anxiety and memory have so far been generally studied apart. Taking this in account, some of us ~5 have developed a new experimental model in an attempt to simultaneously measure drug effects on anxiety and memory. The apparatus has been described in details elsewhere 15. Briefly, it consists of an enclosed arm at a right angle with two open arms elevated 50 cm above the ground, and the test exploits the rat's innate fear of height and openness. Thus, when the rat is placed at the end of the enclosed arm, it does not see the open arms until he pokes his head out of the walls of the closed arm. In this way inhibitory avoidance can be learned. On the other hand, when the rat is placed at the end of one of the open arms he can move towards the close arm, performing an escape response. In this experiment we investigate the effects of D Z P (1-4 mg/kg, i.p., 25 min before training) on acquisition and retention of the inhibitory and the escape response. The following measures were taken: (1) baseline latency, that is, the time from being placed at the end of the enclosed arm to withdrawal; (2) inhibitory avoidance latency, measured as the baseline at subsequent trials; (3) escape latency--time from being placed at the end of an open arm to withdrawal. During the training trial the baseline latency and, 30 s later, the inhibitory avoidance learning (AVOID 1) were measured under either drug or saline. Immediately afterwards, in the same rat, the escape latency (ESCAPE 1) was also measured. After 72 h, the retention of the inhibitory avoidance (AVOID

103

300

¢/J

240

Z

o 0

180

VIf A

I >-

Ow 120

L

I,IJ I--

6O

AVOID 2

AVOID 1

BASELINE

[ ] couT.oL [ ]

1 Ma/*Q [ ]

2 MaIK~ [] 4 MG/KG

Fig. 5. Effect of diazepam on acquisition and retention of the inhibitory avoidance using the elevated T maze test. Values represent median latencies and interquartile ranges for the inhibitory avoidance of the open arms. Asterisks indicate difference between AVOID 2 and AVOID 1, and circles difference between AVOID 1 and B A S E L I N E (P < 0.05, Wilcoxon's multiple comparison test). From Graeff et al. 15.

2) and the escape response (ESCAPE 2) was tested in the same training situation, but in the absence of drug. Fig. 5 illustrates the results for the acquisition and retention of the inhibitory avoidance response. In control rats and in rats treated with 1 mg/kg of DZP, the latency increased from the baseline to AVOID 1 and to AVOID 2, indicating acquisition and retention of the inhibitory avoidance response. The doses of 2 and 4 mg/kg of DZP abolished the influence of trials on this measure, thus evidencing both an anxiolytic drug effect, in the first (training) session, and an amnestic effect, in the second.

30 CO Z

0

2O

LU CO I

,z, lo < _J

t

ESCAPE

[ ] CONTROL[ ]

"1 MG/KG [ ]

ESCAPE 2 2 MG/KG [ ]

4 MG/KG

Fig. 6. Diazepam effect on acquisition and retention of the escape response using the elevated T maze test. Rats were first trained (ESCAPE 1) immediately after AVOID 1. ESCAPE 2 was measured 72 h later. Values represent median latencies (interquartile ranges) for escape from the open arm. Asterisks indicate difference between ESCAPE 2 and ESCAPE 1 latencies for corresponding treatments groups ( P < 0.05, Wilcoxon's test). From Graeff et a1.15.

The results for the escape response are illustrate in Fig. 6. On the first day, the escape latency (ESCAPE 1) was similar in every treatment group, indicating that DZP did not affect the escape performance. Moreover, the ESCAPE 2 latency (tested 72 h later) was shorter than the ESCAPE 1 in all treatment groups, indicating that memory was unaffected by DZP. These results support the view that the anxiolytic and the amnestic effects of BZDs are closely related.

DISCUSSION AND CONCLUSIONS

The results described above provide evidence that the amnestic effects of DZP are mediated through the amygdala, in particular the basolateral nucleus, strongly suggesting a overlapped neurosubstrate for the amnestic and the anxiolytic effects of the benzodiazepine DZP. The results from our two first studies showed that excitotoxic cell body lesions of the amygdala3°, and in particular the basolateral nucleus of the amygdala29 completely block the retention impairment produced by systemic injections of DZP administered prior to training. Supporting evidence for a key role of the basolateral amygdala was provided by the results of our third study (M.A. Souza-Silva and C. Tomaz, unpublished results) using microinjection of DZP in different amygdala nuclei: DZP microinjection in the basolateral/lateral amygdala, but not in the central amygdala, produces anterograde amnesia. These results suggest that, at least in part, the amnestic effects of diazepam are due to activation of BZD receptors within the basolateral/lateral amygdala. Since an intact BL/LAT amygdala appears not to be required for the acquisition of inhibitory avoidance, it seems likely that DZP-induced memory impairment is due to alteration of activity in brain regions influenced by projections from these nuclei. It is also possible, although it seems unlikely, that the DZP effects on memory result from activation of BZD receptors in other brain regions and that such effects require concurrent activation of the BL/LAT nuclei. Additional experiments are needed to examine this issue. Additional support for the view that the anxiolytic and the amnestic effects of BZDs are closely related, is provided by the results of our fourth study 15. Using responses seemingly motivated by the genetically prepared, species-typical fear of openness and elevation, systemic administration of DZP impaired inhibitory avoidance but not escape performance, in the same animal. It is possible that inhibitory avoidance in this experimental model can be representative of anticipatory anxiety and generalized anxiety disorders, while

104 the escape response can be representative of phobias and panic disorders, that are resistant to anxiotytic doses of BZD. In fact, DZP only impaired memory when it had decreased anxiety (inhibitory avoidance). If the amygdala also play a critical role in the mediation of these responses, is a question that remains open to the empirical investigation. In conclusion, the results of the above-described studies show that the site of the amnestic action of the benzodiazepine DZP is localized in the lateral/ basolateral amygdala. Additionally, the amnestic effects of DZP seems to be related to its anxiolytic effect. These findings strongly suggest that the anxiolytic and amnestic effects of DZP (and BZDs, in general) may involve the same, or overlapping neuromechanisms, being the amygdala a critical structure.

ACKNOWLEDGEMENTS

The researches from our laboratories reported in this paper were supported by grants from FAPESP 90/3474 and CNPq (C.T. and F.G.G.), and MH12526 from NIMH and NIDA and Office of Naval Research Contract N00014-90-J- 1626 (J.L.M.).

10

11

12

13

14

15

16

17

REFERENCES 18 1 Amaral, D.G., Memory: anatomical organization and candidate brain regions. In F. Plum (Ed.), Handbook of Physiology, Sect. 1: Neurophysiology, Vol. 5: Higher Functions of the Brain, American Physiological Society, Bethesda, 1987, pp. 211-294. 2 Amaral, D.G., Price, J.L., Pitkanen, A. and Carmichael, S.T., Anatomical organization of the primate amygdaloid complex. In J.P. Aggleton (Ed.), The Amygdala, Wiley-Liss, New York, 1992, pp. 1-66. 3 Ammassari-Teule, M., Pavone, F., Castellano, C. and McGaugh, J.L., Amygdala and dorsal hippocampus lesion block the effect of GABAergic drugs on memory storage, Brain Res., 551 (1991) 104-109. 4 Breen, R.A. and McGaugh, J.L., Facilitation of maze learning with posttrial injections of picrotoxin, J. Comp. Physiol. Psychol., 54 (1961) 498-501. 5 Brioni, J.D. and McGaugh, J.L., Posttraining administration of GABAergic antagonists enhance retention of aversively motivated tasks, Psychopharmacology, 96 (1988) 505-510. 6 Brioni, J.D., Nagahara, A.H. and McGaugh, J.L., Involvement of the amygdala GABAergic system in the modulation of memory storage, Brain Res., 487 (1989) 105-112. 7 Cahill, L. and McGaugh, J.L., Amygdaloid complex lesions differentially affect retention of tasks using appetitive and aversive reinforcement, Behav. Neurosci., 104 (1990) 523-543. 8 Carlsen, J., New perspectives on the functional anatomical organization of the basolateral amygdala, Acta Neurol. Scand., 79 (1989) 4-28. 9 Castellano, C., Brioni, J.D., Nagahara, A.H. and McGaugh, J.L., Posttraining systemic and intra-amygdala administration of the

19 20

21

22 23

24

25

26

27

GABA-B agonist baclofen impair retention, Behuv ~:eurJ/Bi~d., 52 (1989) 170-179. Da Cunha, C., Huang, C.H., Walz, R., Dias, M., Koya. R., Bianchin, M., Pereira, M.E., Izquierdo, I. and Medina, J.H., Memory facilitation by post-training intraperitoneal, intracerebral and intra-amygdala injection of Ro 5-4864, Brain Res., 544 (1991) 133-136. Davis, M., Role of the amygdala in conditioned fear. In J.P. Aggleton (Ed.), The Amygdala, Wiley-Liss, New York, 1992, pp. 255-306. De Olmos, J,. Alheid, G.F. and Beltramino, C.A., Amygdala. In G. Paxinos (Ed.), The Rat Nervous System, Vol. l. Forebrain and Midbrain, Academic Press, Sydney, 1985, pp. 223-334. Decker, M.W., Tran, T. and McGaugh, J.L., A comparison of the effects of scopolamine and diazepam on acquisition and retention of inhibitory avoidance in mice, Psychopharmacology, 100 (1990) 515-521. Graeff, F.G., Brain defense systems and anxiety. In G.D. Burrows, M. Roth and R. Noyers Jr: (Eds.), Handbook of Anxiety. Vol. 3: The Neurobiology of Anxiety, Elsevier Science Publishers BV, Amsterdam, 1990, pp. 307-354. Graeff, F.G., Viana, M.B. and Tomaz, C., The elevated T maze, a new experimental model of anxiety and memory: effect of diazepam, Braz. J. Med. Biol. Res., 26 (1993) 67-7(I. Hodges, H., Green, S. and Glenn, B., Evidence that the amygdala is involved in benzodiazepine and serotonergic effects on punished responding but not discrimination, Psyehopharmacology, 92 (1987) 491-504. lzquierdo, I., Da Cunha, C., Huang, C.H., Walz, R., Wolfman, C. and Medina, J.H., Posttraining down-regulation of memory consolidation by a GABA-A mechanism in the amygdala modulated by endogenous benzodiazepine, Behav. Neural BioL, 54 (1990) 105-109. LeDoux, J.E., Emotion and the amygdala. In J.P. Aggleton (Ed.), The Amygdala, Wiley-Liss, New York, 1992, pp. 339-351. Lister, R.G., The amnestic action of benzodiazepines in man, Neurosci. Biobehav. Rev., 9 (1985) 87-94. Mishkin, M. and Aggleton, J., Multiple functional contributions of the amygdala in the monkey. In Y. Ben-Aft (Ed.), The Amygdaloid Complex, Elsevier/North Holland, Amsterdam, I981, pp. 409-420. Nagy, J., Zambo, K. and Decsi, L., Anti-anxiety action of diazepam after intra-amygdaioid application in the rat, Neuropharmacology, 18 (1979) 573-576. Niehoff, D.L. and Kuhar, M.J., Benzodiazepine receptors: localization in rat amygdala, J. Neurosci., 10 (1983) 209t-2097. Petersen, E.N., Braestrup, C. and Schell-Kruger, J., Evidence that the anticonflict effect of midazolam in the amygdala is mediated by the specific benzodiazepine receptor, Neurosei. Lett., 53 (1985) 285-288. Schell-Kruger, J. and Petersen, E.N., Anticonflict effect of the benzodiazepines mediated by a GABAergic mechanism in the amygdala, Eur. J. Pharmacol., 82 (1982) 115-116. Shibata, K , Kataoka, Y., Gomita, Y. and Ueki, S., Localization of the site of the anticonflict action of benzodiazepines in the amygdaloid nucleus of rats, Brain Res., 234 (1982) 442-446. Shibata, S., Yamashita, K., Yamamoto, E., Ozaki, T. and Ueki, S., Effects of benzodiazepine and GABA antagonists on anticonflict effects of antianxiety drugs injected into the rat amygdala in a water-lick suppression test, Psychopharmacology, 98 (1989) 38-44. Thiebot, M., Some evidence for amnestic-like effects of benzodiazepine in animals, Neurosci. Biobehav. Rev., 9 (1985) 95-100.

105 28 Tomaz, C., Brand,to, M.L. and Garcia-Cairasco, N., Overlapping neural substrates underlying defense reactions, aversive memory, and convulsive behavior. In L.L. Butcher, M. Decker and E. Lewin (Eds.), Neurotransmitters Interactions and Cognitive Function, Birkhgmser Boston, Cambridge, 1992, pp. 240256. 29 Tomaz, C., Dickinson-Anson, H. and McGaugh, J.L., Basolateral amygdala lesions block diazepam-induced anterograde

amnesia in an inhibitory avoidance task, Proc. Natl. Acad. Sci. USA, 89 (1992) 3615-3619. 30 Tomaz, C., Dickinson-Anson,H. and McGaugh, J.L., Amygdala lesions block the amnestic effects of diazepam, Brain Res., 568 (1991) 85-91. 31 Young III, W.S. and Kuhar, M.J., Radiohistochemical localization of benzodiazepine receptors in the rat brain, J. Pharmacol. Exp. Ther., 212 (1980) 337-346.