sign in sign up
<<
Home , Ivan Izquierdo, James McGaugh, Lynn Nadel

BrazilianNeurobiology Journal of oflearning Medical and and Biological Research (2000) 33: 993-1002 993 ISSN 0100-879X Virtual Symposium

The brain decade in debate: I. Neurobiology of and memory

A. Baddeley1, 1Department of , University of Bristol, Bristol, UK O. Bueno2, 2Departamento de Psicobiologia, Universidade Federal de São Paulo, L. Cahill3, São Paulo, SP, Brasil J.M. Fuster4, 3Center for Neurobiology of Learning and Memory and Department of Neurobiology I. Izquierdo5, and Behavior, University of California, Irvine, CA, USA J.L. McGaugh3, 4Neuropsychiatric Institute, University of California, Los Angeles, CA, USA R.G.M. Morris6, 5Centro de Memória, Departamento de Bioquímica, Instituto de Biociências, L. Nadel7, Universidade Federal do , , RS, Brasil A. Routtenberg8, 6Department of , The University of Edinburgh, Edinburgh, UK G. Xavier9 and 7Department of Psychology, University of Arizona, Tucson, AZ, USA C. Da Cunha10 8Northwestern University, Evanston, IL, USA 9Departamento de Fisiologia, Instituto de Biociências, Universidade de São Paulo, São Paulo, SP, Brasil 10Laboratório 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 some active researchers were invited by the Brazilian Society for · Memory Laboratório de Fisiologia e Neuroscience and Behavior (SBNeC) to discuss the last decade’s · Learning Farmacologia do SNC advances in neurobiology of learning and memory. The way different · Departamento de Farmacologia, parts of the brain are recruited during the of different kinds of · Prefrontal cortex UFPR · memory (e.g., short-term vs long-term memory, declarative vs proce- Caixa Postal 19.031 · Memory systems 81531-990 Curitiba, PR dural memory) and even the property of these divisions were dis- Brasil cussed. It was pointed out that the brain does not really store memo- E-mail: [email protected] ries, but stores traces of information that are later used to create , not always expressing a completely veridical picture of the Publication supported by FAPESP. past experienced reality. To perform this process different parts of the The names of the authors are brain act as important nodes of the neural network that encode, store presented in alphabetical order, except for the last one that and retrieve the information that will be used to create memories. corresponds to the symposium Some of the brain regions are recognizably active during the activation organizer. of short-term (e.g., prefrontal cortex), or the storage of information retrieved as long-term explicit memories (e.g., hippo- campus and related cortical areas) or the modulation of the storage of memories related to emotional events (e.g., amygdala). This does not Received May 16, 2000 mean that there is a separate neural structure completely supporting Accepted June 21, 2000 the storage of each kind of memory but means that these memories critically depend on the functioning of these neural structures. The current view is that there is no sense in talking about hippocampus- based or amygdala-based memory since this implies that there is a one- to-one correspondence. The present question to be solved is how systems interact in memory. The pertinence of attributing a critical role to cellular processes like synaptic tagging and protein kinase A activation to explain the memory storage processes at the cellular level was also discussed.

Braz J Med Biol Res 33(9) 2000 994 A. Baddeley et al.

The last decade of this millenium was ever, why is the failure of short-term memory considered as “The Brain Decade” because not a retrieval one? research in this field increased exponen- : Better than that article tially. Among the many subjects under study is a recent review which came out in Behav- in neuroscience, the Neurobiology of Learn- ioural Brain Research (3) a few months ago. ing and Memory is considered by Steven The evidence that short-term memory fail- Rose as the brain’s Rosetta stone (1). The ure is not a retrieval failure is that the re- Rosetta stone was the key to the decoding of trieval of short-term memory is specifically the Egyptian hieroglyphs, because it con- affected by few treatments given, say, 10 tained a text written in this and two other min before the test, whereas the same treat- languages, Demotic Egyptian and Greek. As ments (protein kinase A (PKA) inhibitors, the Rosetta stone, the memory code, if it PD098058, etc.) given 0, 22 or 45 min before really exists, will be the key to the decoding liquidate short-term memory. of all brain functions. In order to discuss the : I haven’t read the Nature scientific advances achieved by science in paper, but the idea that short-term and long- this field, the Brazilian Society of Neurosci- term memory might be separate is well es- ence and Behavior (SBNeC) organized a tablished. Ivan’s latest comment points out virtual symposium held on March 10, 2000 two ways this could happen. I don’t, how- in a chat site provided by the Conselho Na- ever, find it very useful in the declarative/ cional de Desenvolvimento Científico e procedural distinction, so perhaps someone Tecnológico (CNPq). The present article was could help me see what that adds. written based on the transcription of this Ivan Izquierdo: Lynn, you may be right. symposium. Bures recently presented a talk at a small symposium in his country called “How de- Gilberto Xavier: At the system level, clarative can a rat’s memory be?”, in which how is the information encoded into and he also argued that it is simply impossible to maintained in the different memory systems? measure any memory in rodents or even Are there differences in the representation bigger animals without measuring proce- code for these systems? Are the retrieval dures. mechanisms different? Lynn Nadel: Ivan - What I think is im- Ivan Izquierdo: There are many differ- portant is to find principled ways of distin- ences among memories at the system level. guishing memory systems, ones that go be- First, and perhaps foremost, is the difference yond the procedures used to measure them. I between short- and long-term processes, prefer to use the terms explicit and implicit which are handled by different structures or because Schacter and others who coined the same structures with different timing. them were clear that these were only opera- Second, working memory is a separate entity tional definitions. It is important to detect from all the others. Third, and very impor- differences such as “flexibility of use” and tant, is declarative and , the like, in distinguishing memory systems. which involves different systems. But I’m Elizabeth Warrington (4) showed about 30 afraid this doesn’t answer your question, years ago that you could get long-term does it? memory in the absence of short-term memory, Aryeh Routtemberg: We were just re- and that was the basis for her arguments viewing in class a short paper published in against ’s (5) consolidation Nature (2). I think it is intriguing if short- view of hippocampal function. term and long-term memory are separate and Aryeh Routtemberg: Lynn - What I the latter not dependent on the former. How- should have made clear is that short-term

Braz J Med Biol Res 33(9) 2000 Neurobiology of learning and memory 995 and long-term memory are independent and del) and short-term memory for semantically that Ivan showed... and there is not much on processed material. Short-term memory for this as yet... that you can observe long-term speech and visuo-spatial material is very memory in the absence of short-term memory. short in duration (seconds) and usually as- This is intriguing and gets to the heart of a sessed by a span test (digit span, blocks of new way of looking at multiple memory Corsi) (5). But I’m not sure about the present systems. status of short-term memory (or recent Ok, Lynn and Ivan... assuming that you memory, whatever name one gives to it) for have long-term memory in the absence of meaningful material in the cognitive psy- short-term memory demonstrated in rats and chology literature. Anyway, I feel as a chal- humans, what kind of neural/molecular mech- lenging matter for the next years a closer anism do you envision operating to produce integration among , that consequence? Are there cases in the neuropsychology and neurobiology of mem- ‘normal’ individual where this occurs? ory. James McGaugh: Aryeh - I thought this Aryeh Routtemberg: Ok, guys, so any was what Warrington and Shallice reported thoughts on how short-term memory is liqui- years ago (4). dated (Ivan’s term) while long-term memory Aryeh Routtemberg: Jim and Lynn - remains solvent? Were these in normal subjects? Gilberto Xavier: Aryeh, could James McGaugh: Aryeh - No, in a pa- be in the control of the process of sending tient. Not an often cited paper. information either to short-term memory or Lynn Nadel: Good question, Aryeh! Jim, long-term memory. This would help to un- what Warrington and Shallice (4) reported derstand how information gets into implicit had to do with conduction aphasics, who memory systems as well. What do you think? indeed had poor short-term memory and good Lynn Nadel: Aryeh - The problem of long-term memory. But the trick is they were poor short-term memory/good long-term talking about different “contents” ( that memory is only a problem if you think (a la their work predated the multiple memory Hebb) that the former is the necessary gate- systems idea). way to the latter. If they are handled by Ivan Izquierdo: Depressed short-term different systems, or even within the same memory in the presence of good long-term system but by mechanisms that are not caus- memory is characteristic of patients with ally linked, then the result poses no particu- delirium, and, to some extent, of depressive lar problem. patients. James McGaugh: Unfortunately, Hebb Lynn Nadel: That is very interesting - I was wrong about the causal link between would appreciate getting citations to those short-term memory and long-term memory. data later on. How many of you are aware of Lynn Nadel: Jim - What exactly do you the recent report of decreased hippocampal mean by saying “short-term memory uses volumes in chronically depressed people? long-term memory”? Ivan Izquierdo: What I just said is from James McGaugh: Lynn - It seems obvi- psychiatry textbooks. At least that’s where ous. Short-term memory must be about some- I’ve seen it mentioned. thing. And it is always about things we al- Orlando Bueno: I believe that in hu- ready know about. Try your digit span. Try mans it is necessary to distinguish between reading this. All in short-term memory im- short-term memory of different contents mediately and all uses well-established long- (speech, visuo-spatial, as the phonological term memory. and visuo-spatial loops in Baddeley’s mo- Gilberto Xavier: If we adopt the con-

Braz J Med Biol Res 33(9) 2000 996 A. Baddeley et al.

cept of working memory as proposed by showing that there are parallels between scan- Baddeley, the difficulties to understand the ning results and those obtained on the basis interaction between short-term and long-term of cognitive studies and lesion approaches. memory diminish; apparently it becomes an We are now, I think, in a position to begin attentional question, doesn’t it? using scanning as a means of answering Claudio Da Cunha: Ester Palacios, from carefully targeted questions. Vitoria, Brazil, e-mailed us a message ask- Challenges for the future? One major ing about the participation of the prefrontal challenge is to ensure that we combine good cortex in the working memory processes. questions with advancing scanning tech- Richard Morris: I’m not an appropriate niques - there is a temptation often to take a person to comment on the role of the pre- particular paradigm and simply label it with frontal cortex in working memory - maybe the name of some cognitive function. For Alan Baddeley can say a word or two. Or example, in my own area, working memory else check out Paulesu’s (6) review articles. is often regarded as synonymous with the n- Also Jonides (7). back task - a useful but complicated para- Larry Cahill: Regarding the e-mailed digm that is difficult to interpret theoreti- question about the prefrontal cortex, I think cally. I think scanning and other electro- the work of Petrides (8) and others has made physiological methods offer a very useful it clear that while it is involved in various tool for tackling the important but difficult “manipulation” aspects of memory, it can- issues of executive control. In the longer not be essential for memory per se as lesions term, I have high hopes for our improved have little or no effect. When a task involves capacity to label neurotransmitters. As cog- some sort of manipulation of information, nitive psychologists, we will need to be bet- prefrontal cortex lesion effects seem to fall ter at dealing with situations in which the out, but not when simple retention of infor- modulation of a whole system is changed, mation per se is required. rather than one in which a particular sub- Alan Baddeley (message e-mailed dur- component is removed, for example by le- ing the chat): As a cognitive psychologist I sion. I expect that diseases that reflect defi- am pleased with the extent to which our cits of this kind, perhaps schizophrenia and analysis of human memory into separate but Alzheimer’s disease for example, will prove interrelated systems has proceeded and con- theoretically as well as practically impor- solidated. There remain many controversies, tant. Finally, I hope that my many colleagues but broadly speaking, most people would working on the neurobiological basis of accept the usefulness of a distinction be- memory will converge on some more well- tween long-term and working memory, and established interpretations that can be within working memory into separable visuo- mapped onto what we know of the cognitive spatial, verbal and executive components. In psychology of memory. Sorry I am not able long-term memory there is now abundant to take a more active part in what I am sure evidence for separable implicit learning and will be a lively and stimulating session. Best memory systems, which differ from episodic wishes, Alan Baddeley. memory and from , al- Joaquin Fuster: Dear Alan, I agree with though the relationship between these last you, the identification of different systems two remains controversial. In terms of links for memory has been an unquestionable ad- with neurobiology, I think the development vance. I am afraid, however, that this has led of neuroradiological methods has been very to misconceptions on the part of many cog- productive - initially much of the work was nitive neuroscientists. There is no greater concerned with validating the methods and problem in this field now than the tendency

Braz J Med Biol Res 33(9) 2000 Neurobiology of learning and memory 997 to allocate separate cognitive functions to The big problem in memory is not whether separate neural structures - though each struc- it is short or long (it is both) but how infor- ture may participate differently and to a dif- mation is integrated in such a way that it can ferent degree in different functions. That be retrieved and affect flexible performance. tendency is largely the result of misinterpre- No use to store something if it cannot be tations of lesion results and a hypertrophy of adequately accessed. We have focused on the concept of “module” which originated in the storage problem to a large degree be- sensory physiology. Both have led us to a cause we have no hypotheses about how new phrenology more or less legitimized by behavioral responses use stored information the scientific method. It is appropriate to in a flexible way. Or does anyone have such dissect cognitive functions by experimental hypotheses? and cognitive psychology, as Aristotle would Lynn Nadel: Richard - Here’s a thought have undoubtedly approved. But our suc- I’ve had about a possible difference between cess in separating cognitive functions or con- frontal and hippocampal contributions to tents (, working memory, episodic memory. Imagine for a moment that the “where”, syntax of language, retrieval...) the hippocampus rapidly establishes a “tem- does not imply that there is a separate neural plate” or trace that serves to link together structure for each. We should certainly avoid various neocortical/limbic representations of the usage of relative differences (in tasks, all the things that happen in an episode. areas, functional magnetic neuroimage acti- What is missing from this “static” ensemble vation - i.e., subtractions), as justifications is information about the sequence in which for the categorical allocation of a particular these things went together during the epi- cognitive function to a particular piece of sode. This, I think, is what the frontal contri- cortex or cluster of cells. I address this com- bution could be, and why it seems so crucial ment to you, Alan, fully aware that I am in “working memory”, which is nothing but preaching to the converted, for you have a record of the sequence of recently experi- shown repeatedly serious misgivings about enced things. what I, not very originally, call the new Richard Morris: OK Lynn. But let’s phrenology. This gives me also the opportu- focus on your ‘rapidly’. I’d like to add ‘auto- nity to share with you, who is the father of matically’ to that, rather than ‘effortfully’. the concept of working memory, something That is, the hippocampus is clocking away, quite amusing. After discovering “working- recording events-in-scenes that are attended memory cells” in the prefrontal cortex (9), I to, regardless of where a subject is directing feel condemned to spend the rest of my life his efforts with respect to . I’m convincing people that working memory is attracted to the idea of a distinction between not the only function of the prefrontal cortex ‘automatic’ components of episodic encod- and the prefrontal cortex is not the only ing (hippocampal) and ‘effortful’ compo- cortex involved in working memory. Funny nents (the classical Tulving (10) list learning world! Cheers, Joaquin. experiment). James McGaugh: Are there some James McGaugh: Richard and Lynn - broader issues to discuss? I think it is now Do we know that the hippocampus proper is extremely clear that short-term and long- doing specific things in memory? Things, term memory are independent processes. So, that is, that are not also done by other brain what are the implications? We need systems regions? and mechanisms for both and both must be Lynn Nadel: As to some special role for able to interact in microseconds. After all, the hippocampus - the latest data from re- short-term memory uses long-term memory. mote memory studies seem to show that

Braz J Med Biol Res 33(9) 2000 998 A. Baddeley et al.

there are always subtle differences after hip- tween sequences and causal judgments... pocampal damage (11-13). There are also Lynn Nadel: This forum gives me a strong data indicating a necessary role for chance to ask Larry and Jim something - the hippocampus in some kinds of naviga- what are your current views on the impact of tional situations, but I’m not sure that is a arousal/ on memory - do you think that “memory” function per se. it is always positive, or is it U-shaped, or Richard Morris: Only connect!!! Every- does it depend on what kind of memory one thing is combinatorial. Soon we will have an is talking about (e.g., amygdala-based, hip- electronic journal called “Associative Mo- pocampal-based, etc.)? lecular Biology”!!!! Mind you, if it is Larry Cahill: Lynn - I see no reason to “Transgenic Associative Molecular Biology” think any brain structure or system operates maybe Aryeh won’t want to be on the Edito- in a non-inverted U fashion in memory. rial Board!! Sorry to lower the tone, folks - Ivan Izquierdo: Maybe it’s late to raise back to serious stuff!! such a potentially explosive point. But, what James McGaugh: Richard (and Aryeh) - are the grounds to talk about “amygdala- Aryeh might want to be Senior Editor of based” as opposed to “hippocampal-based” such a journal. Why not? memory? I, for one, don’t see any. Aryeh Routtemberg: Settle down, boys. James McGaugh: I agree with Ivan’s Lynn Nadel: Richard - OK. I suppose comment. one could see the frontal role as relating to Ivan Izquierdo: Jim - Thank you. It is a effortful encoding, as distinct from the auto- very basic point. matic nature of hippocampal action, but Lynn Nadel: Ivan - Excellent point - and which part of the system solves the sequence we don’t have enough time left to deal with problem that I raised? This also applies to it. I was hoping to get away with it. retrieval (Jim) - when we retrieve an epi- Richard Morris: Amygdala-based mem- sodic memory, how do we get the bits and ory. I’m rather disturbed about ’s pieces in the right temporal order? exclusion of the amygdala-based memory Richard Morris: Lynn - sequences are from the class of representational types of tough. I’m still not sure how good animals memory (declarative in his terms) (14). The are at sequences. The latest purely behavior- exclusion is based on the failure of amygdala al primate data on this question reveal lesions to affect delayed non-matching - but competences, but not much more than that. It that is perhaps not the best of rocks on which is an important issue, I concede, but is the to rest any theory just now. Surely we can ability to remember things accurately in se- ‘declare’ our emotional states? What is the quence likely to be something for which basis for excluding amygdala-dependent there would have been a great deal of selec- memory from representational memory? tive pressure in early mammalian evolution Larry Cahill: Richard - The failure of when the hippocampus and prefrontal lobe amygdala lesions to affect the delayed-non- came along? matching-to-sample task (a memory task used Lynn Nadel: Richard - A good point, but mainly for monkeys) fits nicely with the think about it more implicitly - even lowly view that its primary function is to influence animals need to get sequences right in order long-term for stress- to get causality right - it matters if the condi- hormone releasing events. tioned stimulus precedes or follows the un- James McGaugh: Richard - Amygdala conditioned stimulus, and animals are ca- dependent, in my view, does not mean emo- pable of remembering this sort of thing. I am tional content (15,16). Amygdala activity particularly intrigued by the connection be- modulates memory of what to do in a water

Braz J Med Biol Res 33(9) 2000 Neurobiology of learning and memory 999 maze and where to go. It modifies memory one memory or other mixed with bits of of the size of a reward, it modifies memory other, usually older memories. The memo- of a story, etc. etc. I think it is a mistake to ries of many old people seem to be built like identify the amygdala with emotion. Emo- old cities (Jerusalem, Rome), tie over tie of tion activates this region but the consequences older and older memories; you drill in, take need not be and usually are not memories of up some material and the final result is often emotions. false. My mother, who otherwise had a good Orlando Bueno: I think Jim has touched memory, confused events of my youth with the point. The important thing is to find out those of her brother’s; two persons she liked how the structures and memory systems work and who had a few things in common. together and so we would find that the amyg- Lynn Nadel: Ivan - I agree that separat- dala participates in some way in declarative ing the short-lasting “consolidation” events memory as an enhancer of memory (17,18). from the longer-lasting (many years) things James McGaugh: Lynn - It is time for all is crucial. Another topic for the next ses- of us to stop talking about “amygdala based” sion... or “hippocampus based” as that implies that Claudio Da Cunha: I agree with Ivan, there is a one-to-one correspondence which, but how are these traces of information of course, is nonsense. What we need to coded? In my opinion, if we compare our know is how systems interact in memory. knowledge about the neurobiology of learn- We have shown very clearly that amygdala ing and memory with the knowledge in ge- activity regulates memory that involves hip- netics we are in a pre-double-helix stage: we pocampal activity (i.e., ). cannot decode a brain state as traces of infor- Thus, the amygdala is very clearly part of the mation useful to create memories. We can- explicit memory system as are, no doubt, not decode the information files! My guess many other brain systems. There are no “hip- is that with the refinement of the resolution pocampal memories” and there are no “amyg- of functional neuroimages this will be pos- dala memories”. If you think there are, direct sible in the future. me to the references supporting that view. Joaquin Fuster: In answer to Claudio’s Lynn Nadel: Jim - Of course I agree with question, I suggest we should, at least in you - in fact I think we should dump the principle, accept the concept of network rep- whole term “memory”. What we store is resentation in the cortex and the distributed information, and we later use that to “create” nature of all cognitive functions and types of memories. memory (9), however difficult our experi- Richard Morris: Lynn - exactly! Traces ments may thus become. I can see no better are stored. Memories are created! source of progress now than that position of Ivan Izquierdo: Lynn - as to “creating” principle. memories out of the information we store. Lynn Nadel: Claudio - I agree with your Yes, maybe that is related to Squire’s con- assessment of the problem, but not with your cept that consolidation takes years (14). I optimism about the resolution coming from would, like most, restrict use of the word the kind of functional neuroimaging avail- consolidation to the 6 h or so period in which able now (or in the foreseeable future). I memories are really set down after encoding. think Joaquin touched on some of the prob- But I, again like most, would agree with Jim lems with this approach. that we do form new memories on top of old Claudio Da Cunha: Indeed my opti- ones, eventually modifying them and mak- mism that we (or most probably the next ing up veritable lies. Every time we retrieve generations) will discover a kind of key for we probe into the complex and draw bits of decoding these traces of information that the

Braz J Med Biol Res 33(9) 2000 1000 A. Baddeley et al.

brain uses to compose trace memories is not James McGaugh: Claudio - Evidence of based only on functional neuroimages. Now, CREB involvement in memory is interesting as Dr. Baddeley pointed out, this technique but not very informative as it is upstream is becoming more and more powerful in from a large number of genes. Not a very resolution and proper to answer more specif- precise tool, in my view (though, as you ic questions that we could not solve with know we have worked on it here - as has lesion/pharmacological approaches. I believe Ivan’s group). that someday we will find a way of using this Richard Morris: Ah Jim - but the speci- or other techniques to decode a brain state in ficity need not be at the level of the nucleus. a sensible information useful to compose a That’s what the synaptic tagging (23) idea is memory trace. partly about. You can have a nuclear signal Lynn Nadel: Richard - I would like to that is turned on by a variety of things, and hear your current thoughts about whether can trigger via CREB (or other transcrip- properties evident at one level of analysis tional activators) all sorts of genes. The pro- (e.g., cellular/molecular) show up at another teins then synthesized will only be seques- level (e.g., systems) (19). An example might tered at activity-regulated sites that have the be the claim that the rapid plasticity evident appropriate tags. I really like Harry Bito’s in hippocampal long-term potentiation (LTP) (24) description of this problem as the ‘many- (20) is also evident in the one-trial learning to-one’ problem (synapses to the nucleus) this system seems to be capable of (21,22). and the ‘one-to-many’ problem (nucleus to Do you (and others) think this parallelism is the synapses). an accident, or it will turn out to be a general Aryeh Routtemberg: Claudio - We have feature? something like 100,000 genes, of which Aryeh Routtemberg: Lynn - I grant you roughly 50% are regulatory. Of those 50,000 that conceptually, but given that there are genes, one of them is CREB. We still need to many molecular mechanisms out there, what know what the other 49,999 are doing. I am would be a plausible molecular/cellular sce- deeply concerned about the focus on CREB nario? at different levels, but the most important Richard Morris: Lynn - I have been one is that transcription factors never work intrigued by the possibility that the one-shot in isolation but in a combinatorial way bind- characteristic of LTP induction may be a ing to recognition sites on the promoter and physiological substrate of episodic or epi- forming complex protein-DNA-protein com- sodic-like memory. A twist is that I don’t plexes that lead to the initiation of transcrip- think LTP will, in general, last very long tion. The focus on CREB undermines any unless there is relevant protein sequestration attempt to understand the cell biology of and tag-protein interactions to stabilize these memory. one-shot changes in synaptic efficacy that James McGaugh: Richard and Aryeh - I would underlie a ‘record of experience’ (23). see no conflict in views. CREB clearly gives A further twist is I don’t understand the no synaptic specificity. But, Richard, how differential contributions of the hippocam- do we get a flexible behavioral read out from pus and prefrontal cortex to episodic memory synaptic specificity? This is the silent ques- - at encoding. tion. Claudio Da Cunha: I received an e-mail Ivan Izquierdo: As judged by the effects from Edson Cardia, from Brazil asking about of PKA (25) and mitogen-activated protein the participation of cAMP response element kinase (MAPK) inhibitors (26), both sets of binding (CREB) in the process of learning enzymes act at different moments and on and memory. Would someone like to reply? different substrates to create short-term and

Braz J Med Biol Res 33(9) 2000 Neurobiology of learning and memory 1001 long-term memory in the hippocampus. That cross-talk between short-term memory and would be a nice starting point: to look for long-term memory, as there must be be- those substrates. The bad news is that now tween spatial and contextual components, it’s known that PKA and the extracellular declarative and procedural, visual and olfac- signal-regulated kinase system are truly en- tory, etc. The thing is: who does the cross- meshed, cross-talk a lot with each other, and talking and where? For short-term and long- cross-talk with protein kinase C (PKC) a lot term memory we do have indirect hints that (27). it may take place at the PKA-MAPK-PKC Aryeh Routtemberg: That’s it! Parallel mesh. In the hippocampus at least; perhaps signal transduction processes, one leading to also in the entorhinal cortex. Now, a major short-term memory retrieval the other kick- point is, who cross-talks the very early mo- ing long-term memory into operation. Dis- ments of memory, those in which, according ruption of the former leads to the kind of to Richard, synapses are tagged, with the result you have shown. Perhaps cross-talk is ensuing moments? necessary, under normal conditions, for link- ing both signal transduction streams to long- Acknowledgments term memory. James McGaugh: Aryeh - Reasonable The SBNeC organizers are grateful to resolution. Independent and parallel in de- Hidevaldo Machado and Roberto Andreatini velopment. But, again, short-term memory for the help with the organization of the chat must rapidly retrieve long-term memory in and to Simone Alencar for technical assis- order to work. tance during the chat. The CNPq-Prossiga is Ivan Izquierdo: Back to a point raised also acknowledged for providing the chat by Aryeh. Yes, I suppose there must be web room.

References

1. Rose SPR (1991). Memory. The brain’s Science, 283: 1657-1661. task. Psychobiology, 27: 313-330. Rosetta stone? Concepts in Neuroscien- 8. Petrides M (1997). Visuo-motor condi- 14. Squire LR (1992). Memory and the hippo- ces, 2: 43-64. tional associative learning after frontal and campus; a synthesis from findings with 2. Izquierdo I, Barros DM, Mello e Souza T, temporal lesions in the human brain. rats, monkeys and humans. Psychological De Souza MM, Izquierdo LA & Medina JH Neuropsychologia, 35: 989-997. Review, 99: 195-221. (1998). Mechanisms for memory types 9. Fuster JM (1995). Memory in the Cere- 15. Cahill L & McGaugh JL (1998). Mechan- differ. Nature, 393: 635-636. bral Cortex. Massachusetts Institute of isms of emotional arousal and lasting de- 3. Izquierdo I, Medina JH, Vianna MRM, Technology Press, Cambridge, MA. clarative memory. Trends in Neuroscien- Izquierdo LA & Barros DM (1999). Sepa- 10. Tulving E (1983). Elements of Episodic ces, 11: 294-299. rate mechanisms for short- and long-term Memory. Clarendon Press, Oxford. 16. McGaugh JL (2000). Memory - A century memory. Behavioural Brain Research, 11. Scoville WB & Milner B (1957). Loss of of consolidation. Science, 287: 248-251. 103: 1-11. recent memory after bilateral hippocam- 17. Izquierdo I, Da Cunha C, Rosat R, Ferreira 4. Warrington EK & Shallice T (1969). The pus lesions. Journal of , Neuro- MBC, Jerusalinsky D & Medina JH (1992). selective impairment of auditory verbal surgery and Psychiatry, 20: 11-21. Neurotransmitter receptors involved in short-term memory. Brain, 92: 885-896. 12. Moscovitch M, Yaschyshyn T, Ziegler M memory processing by the amygdala, me- 5. Milner B (1971). Interhemispheric differ- & Nadel L (1999). Remote episodic mem- dial septum and hippocampus of rats. Be- ences in the localization of psychological ory and retrograde : Was Endel havioral and Neural Biology, 58: 16-25. processes in man. British Medical Bulle- Tulving right all along? In: Tulving E (Edi- 18. Da Cunha C, Levi De Stein ML, Wolfman tin, 27: 272-277. tor), Memory, Consciousness and the C, Koya R, Izquierdo I & Medina JH (1992). 6. McGuire PK, Paulesu E, Frackowiak RS & Brain: The Tallinn Conference. Psychol- Effect of various training procedures in an Frith CD (1996). Brain activity during ogy Press, Philadelphia, 331-345. elevated plus maze: Possible relation with stimulus independent thought. Neuro- 13. Kubie JL, Sutherland RJ & Muller RU brain regional levels of - Report, 7: 2095-2099. (1999). Hippocampal lesions produce a like molecules. , Biochem- 7. Smith EE & Jonides J (1999). Storage and temporally graded on istry and Behavior, 43: 677-682. executive processes in the frontal lobes. a dry version of the Morris swimming 19. Morris RGM, Anderson E, Lynch GS &

Braz J Med Biol Res 33(9) 2000 1002 A. Baddeley et al.

Baudry M (1986). Selective impairment of 23. Frey U & Morris RGM (1998). Synaptic sively-motivated task in rats. Proceedings learning and blockade of long-term poten- tagging: Implications for late maintenance of the National Academy of Sciences, tiation by an N-methyl-D-aspartate recep- of hippocampal long-term potentiation. USA, 94: 1041-1046. tor antagonist, AP5. Nature, 297: 681-683. Trends in , 21: 181-187. 26. Atkins CM, Selcher JC, Petraitis JJ, 20. Malenka RC & Nicoll RA (1999). Long- 24. Bito H (1998). The role of calcium in activ- Trzaskos JM & Sweatt JD (1998). The term potentiation - a decade of progress? ity-dependent neuronal gene regulation. MAPK cascade is required for mammalian Science, 285: 1870-1874. Cell Calcium, 23: 143-150. associative learning. Nature Neuroscien- 21. Carew TJ (1996). Molecular enhancement 25. Bernabeu R, Bevilaqua L, Ardenghi P, ces, 1: 602-609. of memory formation. Neuron, 16: 5-8. Bromberg E, Schmitz PK, Bianchin M, 27. Micheau J & Riedel G (1999). Protein ki- 22. Xu L, Anwyl R & Rowan MJ (1998). Spa- Izquierdo I & Medina JH (1997). Involve- nases: which is the memory molecule? tial exploration induces a persistent rever- ment of hippocampal D1/D5 - Cellular and Molecular Life Sciences, 55: sal of long-term potentiation in rat hippo- cAMP signaling pathways in a late 534-548. campus. Nature, 394: 891-894. memory consolidation phase of an aver-

Braz J Med Biol Res 33(9) 2000