Brazilian Journal of Medical and Biological Research (2001) 34: 283-293 Neurobiology of emotion 283 ISSN 0100-879X

The brain decade in debate: III. Neurobiology of emotion

C. Blanchard1, Departments of 1Neurobiology and 2Psychology, University of Hawaii, R. Blanchard2, Honolulu, HI, USA J.-M. Fellous3, 3Computational Neurobiology Laboratory, The Salk Institute, Biological Studies, F.S. Guimarães4, La Jolla, CA, USA W. Irwin5, 4Departamento de Fisiologia, Faculdade de Medicina de Ribeirão Preto, J.E. LeDoux6, Universidade de São Paulo, Ribeirão Preto, SP, Brasil J.L. McGaugh7, 5Department of Psychology, University of Wisconsin, Madison, WI, USA J.B. Rosen8, 6Center for Neural Science, New York University, New York, NY, USA L.C. Schenberg9, 7Center for Neurobiology of Learning and Memory and E. Volchan10 and Department of Neurobiology and Behavior, University of California, Irvine, CA, USA C. Da Cunha11 8University of Delaware, Neuroscience Program, Newark, DE, USA 9Laboratório de Neurobiologia, Departamento de Ciências Fisiológicas, Centro de Biomedicina, Universidade Federal do Espírito Santo, Vitória, ES, Brasil 10Laboratório de Neurobiologia 2, Instituto de Biofísica Carlos Chagas Filho, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil 11Laboratório de Fisiologia e Farmacologia do Sistema Nervoso Central, Departamento de Farmacologia, Universidade Federal do Paraná, Curitiba, PR, Brasil

Abstract

Correspondence This article is a transcription of an electronic symposium in which Key words C. Da Cunha active researchers were invited by the Brazilian Society of Neurosci- · Emotion Laboratório de Fisiologia e ence and Behavior (SBNeC) to discuss the advances of the last decade · Fear Farmacologia do SNC · in the neurobiology of emotion. Four basic questions were debated: 1) Anxiety Departamento de Farmacologia · Learning What are the most critical issues/questions in the neurobiology of UFPR · Memory Caixa Postal 19.031 emotion? 2) What do we know for certain about brain processes · Amygdala 81531-990 Curitiba, PR involved in emotion and what is controversial? 3) What kinds of · Nucleus accumbens Brasil research are needed to resolve these controversial issues? 4) What is · Periaqueductal gray matter E-mail: [email protected] the relationship between learning, memory and emotion? The focus was on the existence of different neural systems for different emotions Publication supported by FAPESP. and the nature of the neural coding for the emotional states. Is emotion

The names of the authors are the result of the interaction of different brain regions such as the presented in alphabetic order, amygdala, the nucleus accumbens, or the periaqueductal gray matter except for the last who was the or is it an emergent property of the whole brain neural network? The symposium organizer. relationship between unlearned and learned emotions was also dis- cussed. Are the circuits of the former the underpinnings of the latter? It was pointed out that much of what we know about emotions refers to aversively motivated behaviors, like fear and anxiety. Appetitive Received August 21, 2000 Accepted January 30, 2001 emotions should attract much interest in the future. The learning and memory relationship with emotions was also discussed in terms of conditioned and unconditioned stimuli, innate and learned fear, con- textual cues inducing emotional states, implicit memory and the property of using this term for animal memories. In a general way it could be said that learning modifies the neural circuits through which emotional responses are expressed.

Braz J Med Biol Res 34(3) 2001 284 C. Blanchard et al.

Introduction first set is conceptual: Few areas of science have a conceptual framework as loose and Emotion has long being recognized as an imprecise and so influenced by common important drive for behaviors important to sense as the field of emotion. What are emo- the survival of animal species. In humans it tions? What are the differences between contributes to welfare and suffering, with an emotion, mood, drive, temperament and per- important role in psychiatric diseases. The sonality? Are drives and emotions distinct efforts of psychologists and neuroscientists entities? How many emotions do we have? to create a scientific theory of emotion can What are the differences between sadness, be viewed in the theories and studies of anxiety, fear, panic, happiness, love, sexual Charles Darwin, William James, Sigmund arousal, hunger, pleasure, aversion, etc.? Are Freud, James Papez, and Paul McLean, there basic emotions? Is pain a sensation or among others. The advances introduced by an emotion? What is suffering? What do we these scientists were followed by a long mean by stress? How do these entities relate period of time in the middle of the 20th to each other? The second set of problems is century during which both neuroscience and experimental and deals with the specific con- cognitive science neglected the emotional tribution of any neural system to the genesis nature of the mind. Hopefully, in the so- of emotions. Conversely, it also addresses called “Decade of the Brain”, a step forward the question if one neural system can express was made by the attempts to join the progress two or more kinds of emotions depending on of the cognitive neuroscience approach and its mode of operation. For instance, are anxi- emotion in a unified theory of mind (see, for ety and panic generated by different systems example, the books of Damasio (1) and or by the different functioning of a single LeDoux (2)). Antonio Damasio (3) called neural system? Hopefully, the solution of this movement “a second chance for emo- the second set of problems will help us an- tion”, in spite of the fact that today there are swer the above theoretical questions. more disagreements than fully accepted theo- Jean-Marc Fellous: Is there one emo- ries for emotions (4). Nevertheless, some tional system, or are there several emotional consensus has been reached in particular systems? Is there an animal model of emo- issues, like the role attributed to the amygda- tion (or of a particular emotion)? la in emotion. In order to discuss the scien- Claudio Da Cunha: In my opinion the tific advances and views achieved in this most critical questions in the neurobiology field over the last decade, the Brazilian Soci- of emotion are: 1) What is emotion in a ety of Neuroscience and Behavior (SBNeC) neurobiological perspective? 2) Which are organized a virtual symposium held on April the basic neurobiological systems involved 24, 2000. The chat site was provided by the in specific kinds of emotion? Which brain Conselho Nacional de Desenvolvimento structures participate in the expression of Científico e Tecnológico (CNPq). The pre- different types of emotion and how do they sent article is based on the transcription of do so? Is an emotion encoded in a specific this symposium. brain region or is it a property of the interac- tion of the whole brain neural network? What What are the most critical questions is the participation of the different emotional in the neurobiology of emotion? brain systems in psychiatric diseases? Joseph LeDoux: I will start with two Luiz Carlos Schenberg: From my point obvious questions. First, are there different of view there are two sets of problems that neural systems for different emotions? Sec- should be worked out in the next years. The ond, how do we get from our understanding

Braz J Med Biol Res 34(3) 2001 Neurobiology of emotion 285 of emotional reactions in the brain to emo- there is growing evidence for differential tional feelings? involvement of different aspects of (systems Francisco Guimarães: What is the in- within) fear/defense in different psychiatric teraction between emotion and memory? conditions. How do emotional experiences induce last- James McGaugh: I think that we need to ing behavioral changes? What is the role of express some caution about concluding that the hippocampal formation in emotion? What we have good models for studying any emo- about the neurobiology of other emotional tion in animals. We have to consider both the states (since a large part of the research has autonomic expression and other aspects of been done with a specific emotional state: expression of emotions. fear)? And what is the neurochemistry of William Irwin: In part related to Joseph’s emotion, particularly the role of neuropep- comment, at least at the human level, it is tides found in huge concentration in areas important to distinguish between the pro- related to emotion? cesses underlying the perception of emo- Robert & Caroline Blanchard: We’re tionally salient stimuli from the neural sub- at home with only one computer so our strates that “generate” emotion. names will appear together. One critical is- Luiz Carlos Schenberg: Jeffrey - At sue is - What are the emotions? Taking the least in the periaqueductal gray matter there example of fear, we don’t know if this in- seems to be a great overlapping of systems, if volves one or several systems and if the latter not in the periaqueductal gray matter itself, how to differentiate and classify them. It systems elsewhere in the brain that are trig- probably will be possible to create a unified gered by activation of a single site within the view of emotions, their evolutionary func- periaqueductal gray (5,6). tioning and all, but to obtain a solid result Jeffrey Rosen: Yes, the periaqueductal this is going to have to rest on adequate gray matter seems to be divided at least into analyses for at least some focal systems. active (escape) and passive (freezing, immo- Jeffrey Rosen: Are there common neu- bility) behavioral coordinating areas; how- roanatomical substrates for different emo- ever, these may be activated by different tions or are there separate systems? For ex- forebrain neural systems, one with projec- ample, the amygdala is important for fear tions from the hypothalamus and the other conditioning, but evidence suggests it is also from the amygdala, respectively. involved in appetitive conditioning. Are these William Irwin: In part the route for pre- overlapping systems using the same circuits served responses will be sense modality spe- or separate parallel circuits? cific. Joseph LeDoux: We all seem to be on Luiz Carlos Schenberg: William, I agree the same track. The question then is what is with you but I would add to your comment the best way to figure this out. We have that we should also differentiate processes decent models of emotional behavior for that generate emotions from those which are studying fear, but less good models for other merely executive of hard-wired behaviors emotions. Appetitive conditioning, for ex- (releasing mechanisms). ample, doesn’t have the same face validity James McGaugh: I agree with Irwin’s for human emotion the way fear condition- comment. ing does. William Irwin: Getting back to the issue Robert & Caroline Blanchard: We can of different neural systems for different emo- muster a case for good models of aggression, tions, if you take the perceptual processes (in as well as fear, the latter possibly involving humans) with regard to facial expressions, more than one system anyway. Certainly then the amygdala appears not to be in-

Braz J Med Biol Res 34(3) 2001 286 C. Blanchard et al.

volved with certain emotional states (e.g., Marc’s comment, in our last chat (on neuro- anger). biology of learning and memory, see Ref. 8), James McGaugh: I think it is important Joaquin Fuster said that he believes we are to distinguish between the learning of an living times of “a new phrenology more or emotion (emotional response) and the influ- less legitimized by the scientific method”. Is ence of emotion on learning and memory. this the case for the neurobiology of emotion? This is a critical distinction. I think it highly Jeffrey Rosen: In reply to Jean-Marc’s unlikely that an emotion is encoded in any earlier comment, I agree that the brain prob- specific brain region. Interactions among ably did not develop as emotion systems per brain regions seem to be required. se, but as systems for rapid appraisal (neural Claudio Da Cunha: I agree with you, but computation) of incoming stimuli and gen- sometimes it is not what it seems to people eration of appropriate responses. Emotion that are not working in this area but who are may be an emergent property of this neural interested in reading the papers. activity. For learning or rapid evaluation of William Irwin: Indeed, McGaugh’s dis- already learned fear, the amygdala is per- tinction is important. Relatedly, Davis’ group forming many of these computations. For (7) have reported that there are different unlearned or innate fear I think it is not clear neural systems (e.g., amygdala) that are re- what the role of the amygdala is. Much of the sponsible for conditioned vs unconditioned appraisal of unconditioned emotional stimuli fear responses. may be taking place in the sensory organs, Claudio Da Cunha: Dr. LeDoux, do you thalamus or cortex. agree with Dr. McGaugh’s opinion? Joseph LeDoux: Jean-Marc, it could eas- Joseph LeDoux: Yes, I do agree with ily be the case for emotion if you only look at Jim McGaugh about the difference between human work, imaging, and lesions. I think it brain mechanisms of emotion and emotional may be less of a criticism of the animal work influences on memory mechanisms. that looks at mechanisms within and be- James McGaugh: I agree with Irwin that tween cells as opposed to brain areas. it is important to distinguish between learned James McGaugh: I think we need to and unlearned emotional responses. consider much more seriously the intercon- nectedness of brain systems mediating emo- What do we know for certain about tional responses and emotional influences. brain processes involved in emotion Several recent papers have reported that fear and what is controversial? can be and is expressed in rats lacking an amygdala. Also, Joseph, you have shown Jean-Marc Fellous: One possible ap- that different aspects of fear are mediated by proach is to focus on a particular emotion different amygdala nuclei (9). and study it exhaustively (from neurochem- Joseph LeDoux: Jim is right on target. istry to behavior). This is the most conserva- We are very good at focusing on small pieces tive/traditional way and it brought about al- of the puzzle in neuroscience (on memory or ready plenty of information (fear condition- perception or emotion or attention) and have ing, for example). But I am wondering if done less well in understanding how all of such an approach will not bias a priori the these things come together to make us who scope of research... What if there were no we are. emotional centers/circuits per se, but rather Jean-Marc Fellous: It seems to me that emotion is an emergent property of existing in addition to asking ‘what are the neurons/ computations... centers involved in emotion?’, we should Claudio Da Cunha: With respect to Jean- also ask ‘how a particular set of neurons

Braz J Med Biol Res 34(3) 2001 Neurobiology of emotion 287 computes during an apparent emotional state’ the basis for analyses of the neurobehavioral (i.e., fast, slow, using synaptic changes, re- systems controlling at least one type of emo- cruiting more/less cells...). tion. James McGaugh: Another important Jeffrey Rosen: In regard to learned and controversy is that of whether the same pro- unlearned emotional responses, I think the cesses underlay learned and unlearned emo- circuits may be different. We have lesion tional responses. What do we know about data showing that the lateral, basolateral and that? central nuclei of the amygdala are important Joseph LeDoux: Almost everything is for conditioned fear (freezing), but not for controversial. Regarding learned vs unlearned freezing to a predator odor. We’re not yet mechanisms, my own view is that learning sure what the systems are for freezing to modifies the circuits through which unlearned predator odor. responses are expressed. Luiz Carlos Schenberg: Jeffrey, may be William Irwin: Regarding LeDoux’s last there is an involvement of the medial amyg- statement about learning modifying a cir- dala. Unpublished data from our laboratory cuit, can you say a bit more about the nature suggest an important role of the medial amyg- of that modification? Is it more like a param- dala in fear-like behaviors (immobility, trot- eter adjustment? Do certain structures go ting). The Blanchards have convincing data “off line” and do certain structures go “on as well (11,12). line” (i.e., a transfer of function)? Eliane Volchan: Joseph, going back to Joseph LeDoux: I had in mind the switch- the issue of learning modifying unlearned ing of inputs. Natural stimuli turn the system circuits: How does a learning fear experi- on and through learning novel stimuli gain ence (which involves the amygdala) modify access. I base this in part on Blanchard’s a circuit of unlearned fear, if the amygdala early work with rat fear of cats (10). But Jeff was not involved in the first place? Rosen’s new data question this to some ex- Joseph LeDoux: The way I see it is that tent. We need more work on what “uncondi- the emotional responses elicited by danger- tioned” fear is. ous stimuli are hard-wired outputs of the Eliane Volchan: If an unlearned visual amygdala. These outputs are engaged when fear response like that for looming does not the amygdala is turned on. Some things that involve the amygdala, it is difficult to imag- turn on the amygdala do so because of evolu- ine the switching of inputs needed for condi- tionary wiring. Others do so because of learn- tioning to novel stimuli. Shouldn’t the same ing. The natural or unlearned or evolution- basic circuit be involved in order to be modi- ary stimuli are prewired to access the hard- fied by the learning experience? wired response circuits. Other things can James McGaugh: I heartily agree that enter those circuits if they activate the cir- we need to know more about the nature of cuits at the same time that a natural stimulus unlearned emotions (including but not re- does. Electric shock, being a form of pain stricted to fear). elicited by tissue damage, naturally activates Robert & Caroline Blanchard: While the amygdala. Amygdala cells are respon- the learned-unlearned fear distinction is im- sive to shocks and also to tones. And if the portant (go to it, guys), we sense a disturbing shock comes on while the tone is activating lack of attention to the specifics of the neuro- the amygdala, the tone acquires the ability to behavioral systems that regulate defense per activate the amygdala as well. This is a stan- se (never mind emotion, until you get the dard Hebbian interpretation of plasticity. basics down). These are beautifully complex Eliane Volchan: Joseph, in that case, the sensorimotor systems and should provide same basic circuit should be involved to be

Braz J Med Biol Res 34(3) 2001 288 C. Blanchard et al.

modified by the learning experience. What William Irwin: Could McGaugh com- happens if the amygdala is not involved in ment further on this point: We know that the first place? “emotion enhances memory” but does the Joseph LeDoux: I think I sort of an- magnitude of the emotion correlate with the swered this above. But basically, for any part degree of memory enhancement? of the brain to learn an association between James McGaugh: The strength of memo- two stimuli, those two stimuli have to have ry varies with the intensity of stimulation some impact on that part of the brain. In this used to induce the memory and with the sense the amygdala has to be involved in the degree of release of catecholamines - par- first place (which I assume means it re- ticularly norepinephrine in the amygdala. sponds to the unconditioned stimulus) if it is Francisco Guimarães: Related to Wil- to be involved in the second place (if it is to liam’s question another issue is: Could a learn). negative emotional experience introduce a Claudio Da Cunha: Talking about con- bias in memory (for example, facilitating troversy - why do medial amygdala nuclei memories of other negative events)? There participate in fear behavior but not modulate are some studies with depressive patients memory? Are these two processes correlated suggesting this. or not? William Irwin: At the human level, it is Joseph LeDoux: I am not sure of the role important for us to systematically examine of the medial amygdala in fear. Damage has the relation between “self-reported” emo- no effect on fear conditioning, at least when tion (that is, at a neural level it is important we have done it, but the medial amygdala is what people say/feel about their emotional connected with the paraventricular nucleus states) and brain systems. We are examining and may be involved in stress hormone re- this with neuroimaging techniques. Also, lease. note the obvious statement that this can’t be Luiz Carlos Schenberg: Stimulation of done in non-human animals. the medial amygdala elicits responses re- Joseph LeDoux: It is true you can’t use sembling those of the periaqueductal gray “self-reporting” in animals but it’s important such as immobility and trotting and Blan- to also remember that just because you can chard and colleagues (11) showed taming of use “self-report” in humans doesn’t mean wild rats by lesioning the medial amygdala. that’s the key to emotions or other aspects of James McGaugh: Have brain imaging mental life (other than the aspects that are studies revealed anything critical about dif- accessible to self-report). I am sure that’s ferential brain systems involved in different what you meant but I just wanted to make it types of emotion? clear. Claudio Da Cunha: From what I have William Irwin: Yes, thank you for that read, I agree with the opinion of Dr. Baddeley clarification. To further add (and related to in the other chat (8): at present brain imaging your comments about learning modifying a has just proved what we already knew about neural circuit), the (human) process of feel- these processes. Now it is time to use it to ings may be an important component that answer new questions. modifies the underlying neural substrata (of Joseph LeDoux: Jim, there are some course, an obvious territory would be the imaging studies that show different activa- prefrontal cortex). tions to different emotional stimuli (faces) James McGaugh: But, we can do a rea- (13). I am not sure if this really addresses sonably good job of getting at something like your question but I think that may be all that “self-reporting” in animals by asking them a exists. lot of questions that they can answer behav-

Braz J Med Biol Res 34(3) 2001 Neurobiology of emotion 289 iorally (i.e., by testing them in various ways). with MPTP (1-methyl-4-phenyl-1,2,3,6- There are a lot of ways, for example to find tetrahydropyridine) lesions in the substantia out what animals like and whether they know nigra pars compacta. It seems to me that how to find what they like and remember some “emotional” aspect may also be en- where it is. This is a bit like self-reporting but compassed by the prefrontal cortex/striate probably more believable! circuitry (unpublished data). Francisco Guimarães: We found that What kinds of research are needed poststress facilitation of 5-HT1A (19) or to resolve these controversial issues? blockade of NMDA-mediated neurotrans- mission in the dorsal hippocampus inhibits Claudio Da Cunha: It would be a good the decrease in open arm exploration found idea if each of us stated in a few words what 24 h after a 2-h period of restraint stress. we are doing in our laboratories to address Luiz Carlos Schenberg: I do believe these controversial issues on neurobiology that anxiety and panic are triggered by phe- of emotion. nomena within the amygdala and periaque- James McGaugh: We find that lesions ductal gray, respectively. Similar to the of the basolateral amygdala block the blunted neuroendocrine response of panic memory-modulating (enhancing or impair- attacks in humans (20,21), our recent results ing) effects of post-training drug and hor- showed that neither plasma ACTH nor pro- mone infusions into other brain regions - lactin levels change following a 1-min full- including, but not restricted to, the dorsal blown defense reaction induced by stimula- hippocampus. The same effects are found tion of the dorsal periaqueductal gray mat- with beta-adrenergic antagonists infused se- ter. Moreover, the dorsal periaqueductal gray lectively into the basolateral amygdala (14). matter-induced defense reaction was attenu- Similar effects have been reported by sev- ated by 21-day administration of clomipra- eral other laboratories studying basolateral mine (22). amygdala influences on dentate gyrus long- Francisco Guimarães: An interesting term potentiation. Basolateral amygdala le- finding with clomipramine. What do you sions block the long-term potentiation and think about the initial worsening that this basolateral amygdala stimulation enhances drug might cause in panic patients? it (15,16). Luiz Carlos Schenberg: There was no Jeffrey Rosen: In our laboratory we are “worsening”, i.e., threshold reduction, in in- studying immediate early gene expression in termediate sessions. In turn, fluoxetine (1 the lateral nucleus of the amygdala with fear mg kg-1 day-1) which is a less efficient conditioning. Early growth response gene 1 panicolytic but a selective 5-HT reuptake (EGR-1; also called zif268) mRNA increases inhibitor, attenuated galloping only. with conditioning, but not during retrieval Robert & Caroline Blanchard: Was that (17). Does that mean the amygdala plays acute fluoxetine? In our laboratory acute and different roles in acquisition and retrieval of chronic fluoxetine have opposite and strong fear? effects on flight, so your “galloping” is very Claudio Da Cunha: In my laboratory we interesting. did some experiments showing that “anx- Luiz Carlos Schenberg: Chronic fluox- ious” and “non-anxious” rats present corre- etine test trials were performed before the lated behavior in learning and memory tasks daily dose so as to avoid the acute effect. (18). I suppose this is in agreement with Robert, it is interesting that neither acute Jean-Marc’s view of an interconnected net- diazepam and buspirone nor a 10-day treat- work. More recently I started to study rats ment with buspirone caused any threshold

Braz J Med Biol Res 34(3) 2001 290 C. Blanchard et al.

increase in periaqueductal gray matter-in- Jean-Marc Fellous: Claudio - It does go duced defensive responses. in the direction of emotion as an interacting Robert & Caroline Blanchard: Which circuit. But I would like to push further... periaqueductal gray matter-induced defen- The fact that two regions are communicating sive responses? during an emotional state is interesting and Luiz Carlos Schenberg: Immobility, trot- important to know. What I was trying to say ting, galloping, micturition, defecation, jump- is that very little research is done on param- ing, and exophthalmus. etrizing the communication (is it fast, slow, Robert & Caroline Blanchard: Related how many neurons are involved, what is the to panic, we have been examining results timing of spikes, what neuromodulators are from a wide array of panic “models” and one involved?) Perhaps this level of description possible interpretation of these is that anti- could be useful to characterize/understand panic and panicolytic drugs work very selec- the emotional state... For example, the amyg- tively on flight, but that when flight is elic- dala is involved in fear conditioning to tone ited lower and lower in the system (i.e., and the hippocampus is involved in condi- experimentally vs stimulation of dorsome- tioning to context: Are stimuli routed with dial hypothalamus vs periaqueductal gray respect to their complexity? Can this be matter) some variability creeps into the sys- changed/modulated in cases where ‘fear’ is tem, suggesting that these drugs may act not involved? This surely would help a com- along the route, having less effect near the putational neurobiologist like me...! end (23). James McGaugh: Although we know William Irwin: I think there are different that the amygdala is involved in fear condi- (likely overlapping) neural substrata that tioning the meaning of “involved” needs govern “preferences” (as you describe) ver- serious consideration. Our findings (24,25) sus “feelings” (self-reported). Note that if and those of many other recent studies (26,27) we are going to be honest with ourselves, a clearly indicate that the amygdala has a modu- truly comprehensive neural theory of emo- latory role in influencing storage in other tion is necessarily going to have to deal with brain systems. We find no evidence that the “feelings” (whatever they may be). amygdala is a site of storage for any kind of James McGaugh: Much of the focus of fear conditioning (28). emotional research has, of course, been on Jean-Marc Fellous: I do think that modu- aversively induced emotional states. How lation of activity, rather than activity itself, is are we to study positive emotional states? where a lot can be learned about emotion... Any guess about the brain processes in- Very much in your line of thought... volved? Jeffrey Rosen: For positive emotions, I Claudio Da Cunha: Talking about posi- think the nucleus accumbens is probably tive reinforcers, could the nucleus accum- involved. Interestingly, LeDoux and others bens play a more important role than the (9) are finding that learning of active avoid- amygdala in this case? ance behavior seems to involve the basal James McGaugh: The nucleus accum- nucleus of the amygdala, but not the central bens is certainly involved. Nucleus accum- nucleus. Because the basal nucleus projects bens lesions produce effects like those of to the nucleus accumbens, it is possible that stria terminalis and basolateral amygdala le- the nucleus accumbens plays a role in active sions in blocking the effects of stress hor- avoidance. Since avoidance is a coping re- mones on memory - even if the stress hor- sponse to an aversive event there are prob- mone agonists are infused into the dorsal ably positive emotions associated with it. If hippocampus. the nucleus accumbens is part of an avoid-

Braz J Med Biol Res 34(3) 2001 Neurobiology of emotion 291 ance circuit, positive emotions should emerge Luiz Carlos Schenberg: Interestingly, during avoidance. stimulation of the central nucleus of the Claudio Da Cunha: Related to Dr. amygdala produces only immobility (unpub- McGaugh’s proposition, a long time ago I lished data). read some papers suggesting that the amyg- Claudio Da Cunha: Does this immobil- dala and some processes may be equally ity mean an emotional state? Is it related to involved in both positive (reward) and aver- the memory processes modulated by the ba- sive behaviors (29). solateral nucleus of the amygdala? Joseph LeDoux: The work of Everitt Luiz Carlos Schenberg: I do not know if and Robbins (30) has helped clarify the role this is tense immobility such as in freezing, of the amygdala and nucleus accumbens. not quiescent immobility. The amygdala seems important for learning Jeffrey Rosen: So how do you see the about incentives and the nucleus accumbens role of the amygdala and periaqueductal gray is where dopamine amplifies the incentive matter in fear (anxiety) vs panic? value of the stimulus coming from the amyg- Luiz Carlos Schenberg: Flight is the dala, if I have got their ideas right. typical response of periaqueductal gray mat- William Irwin: With regard to the amyg- ter, not of the amygdala, and is the response dala/accumbens interactions, I believe it is most sensitive to panicolytics (22). the case that while there are no amygdalofugal projections to the ventral striatum, there are What is the relationship between projections in the reverse. It’s not clear ex- learning, memory and emotion? actly how to interpret these findings, but I think they are important to consider. Robert & Caroline Blanchard: Related Claudio Da Cunha: Other evidence for to learning, we think some attention to an the participation of the amygdala in both additional distinction might be profitable. aversive and reward tasks is that the actions Typical paradigms explore the relationship of benzodiazepines are mediated by the ba- between pain and contextual or point stimuli. solateral amygdala and they block the ex- What about nonpain stimuli such as preda- tinction of an aversive conditioning caused tors as the unconditioned stimulus in context by the learning of a secondary task (31). or other types of learning? Predator-context What do you think about it? associations constitute a significant para- James McGaugh: As benzodiazepines digm in the real world and have the advan- induced amnesia in humans (32) - for all tage of enabling us to systematically add kinds of learning - it seems reasonable to pain activation to analyses. conclude that they affect positive as well as Jeffrey Rosen: Predator odor does not negative learning. seem to act as a good unconditioned stimu- Jeffrey Rosen: Do benzodiazepines lus. We don’t get context conditioning to a change thresholds in the amygdala? predator odor, although the predator odor Luiz Carlos Schenberg: I did not test it elicits robust freezing (35). on thresholds, but it was shown long ago that James McGaugh: Have you tried many microinjection releases punished behavior kinds of odors and intense odors? (33). Jeffrey Rosen: Jim, fox odor (trimethyl- James McGaugh: Benzodiazepines do thiazoline) elicits robust freezing for long not affect memory in animals with basolat- periods of time (at least 15 min). We also get eral amygdala lesions (34). Thus, the benzo- a very nice dose-response curve with small diazepine receptors in the amygdala are criti- amounts of fox odor. We have tried to elicit cal for memory effects. freezing with butyric acid, caproic acid and

Braz J Med Biol Res 34(3) 2001 292 C. Blanchard et al.

isoamyl acetate at amounts 100 times more and explicit memory in animals - can we than the fox odor. Only isoamyl acetate gives specify those prior to conducting any par- us moderate levels of freezing when used in ticular experiment? In the case of human very large amounts (35). We have tested two subjects the distinction is clear on the basis other animal odors (besides caproic acid). of priming experiments. But do we have the Fox urine elicited freezing, while mouse urine same clear evidence for animal studies? did not. Joseph LeDoux: I tend to think of most Luiz Carlos Schenberg: Odors are im- memory in animals and humans as implicit. plicit memory, we can recognize them but In humans we have some sense that explicit not recall, may be this is the point. memory is what happens when information Joseph LeDoux: Most things that hu- that was encoded by the hippocampus is mans are emotional about are things we have retrieved and placed in working memory. At learned about in one way or another. Memory most, then, we should think of explicit is therefore very important in emotion. Both memory in animals as this kind of hippocam- explicit and implicit memory, and vice-versa. pal based memory. But it is also possible that Emotions, once aroused, strongly influence other animals have little or no explicit memory, but Jim should talk about that. memory (at least the kind of explicit memory James McGaugh: We can certainly learn that we talk about in humans). So for me to be emotional in the presence of cues that most memory in animals is implicit. This previously induced the emotion. We all agree terminology is sort of backwards since im- on that. But emotion can also, as discussed plicit can only be defined by contrast to above, modulate the degree of memory of explicit, and it is likely that explicit came the specific events (i.e., declarative memory) later in evolution. that elicited the emotional arousal. And, we Eliane Volchan: Does anyone know the know that this involves the release of stress role of the superior colliculus in the circuit of hormones, the activation of the basolateral positive and negative emotions? For the vi- amygdala (release of norepinephrine) and sual system it is known that the superior activation of other brain regions including colliculus feeds the pulvinar on the way to the insular cortex and hippocampus. the amygdala. William Irwin: Are the emotion-enhanc- Luiz Carlos Schenberg: Michael Davis ing effects on explicit/declarative types of (36) showed that the superior colliculus is memory the same for implicit types of memory? involved in fear potentiation of acoustic James McGaugh: Does anyone know startle by visual cues. how to study implicit memory in animals? We would need to know that in order to Acknowledgments answer Irwin’s question. Joseph LeDoux: Jim, did you mean im- The SBNeC organizers are grateful to plicit or explicit memory in your question Hidevaldo Machado and Roberto Andreatini about animals? for help with the organization of the chat, and James McGaugh: I meant “implicit” but to Simone Alencar for technical assistance I said that because it is not clear to me how during the chat. CNPq-Prossiga is also ac- we can clearly distinguish between implicit knowledged for providing the chat web room. References

1. Damasio A (1999). The Feeling of What 2. LeDoux JE (1996). The Emotional Brain. emotion. In: Lane RD & Nadel L (Editors), Happens. Harcourt Brace & Company, Simon and Schuster, New York. Cognitive Neuroscience of Emotion. Ox- New York. 3. Damasio A (2000). A second chance for ford University Press, New York, 12-23.

Braz J Med Biol Res 34(3) 2001 Neurobiology of emotion 293

4. Lane RD, Nadel L, Allen JJB & Kaszniak 15. Ikegaya Y, Saito H & Abe K (1995). High- nitive memory of contextual fear condi- AW (2000). The study of emotion from frequency stimulation of the basolateral tioning. Proceedings of the National Acad- the perspective of cognitive neuro- amygdala facilitates the induction of long- emy of Sciences, USA, 95: 15003-15007. science. In: Lane RD & Nadel L (Editors), term potentiation in the dentate gyrus in 26. Lehmann H, Treit D & Parent MB (2000). Cognitive Neuroscience of Emotion. Ox- vivo. Neuroscience Research, 22: 203- Amygdala lesions do not impair shock- ford University Press, New York, 3-11. 207. probe avoidance retention performance. 5. Schenberg LC, Marçal LPA, Seeberger F, 16. Jones MW, French PJ, Bliss TVP & Behavioral Neuroscience, 114: 107-116. Barros MR & Sudré ECM (2000). L-type Rosenblum J (1999). Molecular mechan- 27. Cahill L, Vazdarjanova A & Setlow B calcium channels selectively control the isms of long-term potentiation in the insu- (2000). The basolateral amygdala complex defensive behaviors induced by electrical lar cortex in vivo. Journal of Neurosci- is involved with, but not necessary for stimulation of dorsal periaqueductal gray ence, 19: A1-A8. rapid acquisition of Pavlovian “fear condi- and overlying collicular layers. Behaviour- 17. Malkani S & Rosen JB (2000). Specific tioning”. European Journal of Neurosci- al Brain Research, 111: 175-185. induction of early growth response gene ence, 12: 3044-3050. 6. Vargas LC, Azevedo TM & Schenberg LC 1 (EGR-1) in the lateral nucleus of the 28. McGaugh JL, Ferry B, Vazdarjanova A & (2000). Micturition and defensive behav- amygdala following contextual fear condi- Roozendaal B (2000). Amygdala: Role in iors are controlled by distinct neural net- tioning in rats. Neuroscience, 97: 693-702. modulation of memory storage. In: Aggle- works within the dorsal periaqueductal 18. Ribeiro RL, Andreatini R, Wolfman C, Viola ton JP (Editor), The Amygdala: A Func- gray and deep gray layer of the superior H, Medina JH & Da Cunha C (1999). The tional Analysis. Oxford University Press, colliculus of the rat. Neuroscience Let- “anxiety state” and its relation with rat London, 391-423. ters, 280: 45-48. models of memory and habituation. Neu- 29. Cador M, Robbins TW & Everitt BJ (1989). 7. Walker DL & Davis M (1997). Double dis- robiology of Learning and Memory, 72: Involvement of the amygdala in stimulus- sociation between the involvement of the 78-94. reward associations: Interaction with the bed nucleus of the stria terminalis and the 19. Mendonça Netto S & Guimarães FS ventral striatum. Neuroscience, 30: 77- central nucleus of the amygdala in startle (1996). Role of hippocampal 5-HT1A re- 86. increases produced by conditioned ver- ceptors on elevated plus maze explora- 30. Everitt BJ & Robbins TW (1992). Amygda- sus unconditioned fear. Journal of Neuro- tion after a single restraint experience. la-ventral striatal interactions and reward- science, 17: 9375-9383. Behavioural Brain Research, 77: 215-218. related processes. In: Aggleton JP (Edi- 8. Baddley A, Bueno O, Cahill L, Fuster JM, 20. Liebowitz MR, Gorman JM, Fyer AJ, tor), The Amygdala: Neurobiological As- Izquierdo I, McGaugh JL, Morris RGM, Levitt M, Dillon D, Levy G, Appleby IL, pects of Emotion, Memory, and Mental Nadel L, Routtenberg A, Xavier G & Da Anderson S, Palij M & Davies SO (1985). Dysfunction. Wiley-Liss, Inc., New York. Cunha C (2000). The brain decade in de- Lactate provocation of panic attacks. II. 31. Pereira ME, Dalmaz C, Rosat RM & bate: I. Neurobiology of learning and Biochemical and physiological findings. Izquierdo I (1988). Diazepam blocks the memory. Brazilian Journal of Medical and Archives of General Psychiatry, 42: 709- interfering effect of post-training behav- Biological Research, 33: 993-1002. 719. ioral manipulations on retention of a 9. Amorapanth P, LeDoux JE & Nader K 21. Kellner M, Knaudt K, Jahn H, Holsboer F shuttle avoidance task. Psychopharmacol- (2000). Different lateral amygdala outputs & Wiedemann K (1998). Atrial natriuretic ogy, 94: 402-404. mediate reactions and actions elicited by hormone in lactate-induced panic attacks: 32. Mintzer MZ & Griffiths RR (1999). Triazo- a fear-arousing stimulus. Nature Neuro- Mode of release and endocrine and patho- lam and zolpidem: effects on human science, 3: 74-79. physiological consequences. Journal of memory and attentional processes. Psy- 10. Blanchard DC & Blanchard RJ (1972). In- Comparative and Physiological Psychol- chopharmacology, 144: 8-19. nate and conditioned reactions of rats ogy, 32: 37-48. 33. Kruger JS & Petersen EN (1982). Anticon- with amygdaloid lesions. Journal of Com- 22. Schenberg LC, Vargas LC, Medeiros RL, flict effect of the benzoadiazepines medi- parative and Physiological Psychology, 8: Marques TA, Silva SR, Fraipont P & Tufik ated by a gabaergic mechanism in the 281-290. S (1998). Evaluation of the rat defense amygdala. European Journal of Pharma- 11. Kemble ED, Blanchard DC, Blanchard RJ reaction as a model of panic attack in cology, 82: 115-116. & Takushi R (1984). Taming in wild rats man: Pharmacological and neuroendo- 34. Tomaz C, Dickinson-Anson H & McGaugh following medial amygdaloid lesions. crine evidence. International Journal of JL (1992). Basolateral amygdala lesions Physiology and Behavior, 32: 131-134. Psychophysiology, 30: 127-128. block diazepam-induced anterograde am- 12. Kemble ED, Blanchard DC & Blanchard 23. Blanchard DC, Griebel G, Rodgers RJ & nesia in an inhibitory avoidance task. Pro- RJ (1990). Effects of regional amygdaloid Blanchard RJ (1998). Benzodiazepine and ceedings of the National Academy of Sci- lesions on flight and defensive behaviors serotonergic modulation of antipredator ences, USA, 89: 3615-3619. of wild black rats (Rattus rattus). Physiolo- and conspecific defense. Neuroscience 35. Wallace KJ & Rosen JB (2000). Predator gy and Behavior, 48: 1-5. and Biobehavioral Reviews, 22: 597-612. odor as an unconditioned fear stimulus in 13. Breiter HC, Etcoff NL, Whalen PJ, 24. McGaugh JL, Cahill L & Roozendaal B rats: Elicitation of freezing by trimethylthi- Kennedy WA, Rauch SL, Buckner RL, (1996). Involvement of the amygdala in azoline, a component of fox feces. Behav- Strauss MM, Hyman SE & Rosen BR memory storage: interaction with other ioral Neuroscience, 114: 912-922. (1996). Response and habituation of the brain systems. Proceedings of the Na- 36. Tischler MD & Davis M (1983). A visual human amygdala during visual processing tional Academy of Sciences, USA, 93: pathway that mediates fear-conditioned of facial expression. Neuron, 17: 875-887. 13508-13514. enhancement of acoustic startle. Brain 14. McGaugh JL (2000). Memory - A century 25. Vazdarjanova A & McGaugh JL (1998). Research, 276: 55-71. of consolidation. Science, 287: 248-251. Basolateral amygdala is not critical for cog-

Braz J Med Biol Res 34(3) 2001