RESEARCH I NEWS

pendent, this system would be expected to [2] HEEsch and J E Bums. Distance estimation work quite well, since the new recruit bees by foraging honeybees, The Journal ofExperi­ mental Biology, Vo1.199, pp. 155-162, 1996. tend to take the same route as the experi­ [3] M V Srinivasan, S W Zhang, M Lehrer, and T enced scout bee. What dOJhe ants do when S Collett, Honeybee Navigation en route to they cover similarly several kilometres on the goal: Visual flight control and odometry, foot to look for food? Preliminary evidence The Journal ofExperimental Biology, Vol. 199, pp. 237-244, 1996. indicates that they don't use an odometer but [4] M V Srinivasan, S W Zhang, M Altwein, and instead might count steps! J Tautz, Honeybee Navigation: Nature and Calibration of the Odometer, Science, Vol. Suggested Reading 287, pp. 851-853, 1996.

[1] Karl von Frisch, The dance language and OT­ Moushumi Sen Sarma, Centre for Ecological Sci­ ientation of bees, Harvard Univ. Press, Cam­ ences, Indian Institute of Science, Bangalore 560012, bridge, MA, USA, 1993. India. Email: [email protected]

Learning from a Sea Snail: modify future behaviour, then memory is the Eric Kandel form in which this information is stored. Together, they represent one of the most valuable and powerful adaptations ever to Rohini Balakrishnan have evolved in nervous systems, for they In the year 2000, Eric Kandel shared the allow the future to access the past, conferring Nobel Prize in Physiology and Medicinel flexibility to behaviour and improving the with two other neurobiologists: Arvid chances of survival in unpredictable, rapidly Carlsson and Paul Greengard. Whereas changing environments. Carlsson and Greengard were awarded for The scientific study oflearning began only at their work on dopamine, an important neu­ the end of the nineteenth century. Before rotransmitter in the brain, changes in the that, learning and behaviour in general, were level of which result in Parkinson's disease, attributed to vital forces rather than to mate­ Eric Kandel was honoured for his extensive rialistic processes. The latter half of the nine­ and invaluable contribution to our under­ teenth century saw an increasing interest in standing of the cellular and molecular mecha­ animal psychology, due in large part to nisms of learning and memory. Darwin's views: "The difference in mind If learning may be defined as the acquisition between man and the higher animals, great of information during the lifetime of an indi­ as it is, certainly is one of degree and not of vidual, which may be used to direct and kind" (Charles Darwin, 1871).

Resonance readers can look forward to another article on the same subject in a forthcoming issue.

-86------~~------RESONANCE I June 2001 RESEARCH I NEWS

In addition to these associative forms oflearn­ The objective study of observable behaviour ing, there are several non-associative forms, was greatly extended by two psychologists, two of the most ubiquitous being habitua­ Edward Thorndike (1898) and Ivan Pavlov tion and sensitization. Habituation is simply (1906) who developed two of the most widely the decline in response to a stimulus that is studied learning paradigms: instrumental repeatedly presented. Sensitization is the en­ conditioning (or learning by reinforcement) hancement of the response to a stimulus as a and classical conditioning. Thorndike placed result of presentation of another, often strong a hungry cat in a wooden cage, from which it or noxious stimulus: this increase in response could escape if it pulled a latch. By rewarding does not depend on pairing of the two stimuli the cat with food every time it performed the in time and is therefore non-associative in correct behaviour, he found that the cat soon nature. learned how to pull the latch. The time taken to learn this behaviour was shorter when a Psychologists had defined a number of dif­ reward (reinforcement) was provided than ferent kinds of learning based on rigorous when it was not, hence the term learning by behavioural experiments on laboratory ani­ reinforcement. The cat thus learned to per­ mals. Although it was generally accepted that form the specific behaviour that was associ­ the basis for the changes in behaviour that ated with a reward. Pavlov, on the other characterised the different learning para­ hand, showed that salivation in dogs, which digms must lie within the nervous system, is usually a reflex or unconditioned behaviour the exact physiological or structural changes that occurs at the smell of food (the uncondi­ were not understood. The idea that changes tioned stimulus), could be induced by a nor­ in the strength of connections (synapses) be­ mally ineffective stimulus (the conditioned tween neurons (nerve cells) were responsible stimulus) such as the sound of a bell, if the for the changes in behaviour was conceived latter was repeatedly presented just before as early as the 18908 but was further devel­ the food. In other words, the animal learned oped by two psychologists, Jerry Konorski to associate the ringing of a bell with the and Donald Hebb, in the 1940s. Hebb pro­ imminent appearance of food. Both instru­ posed that coincident activity in two neu­ qlental and classical conditioning are asso­ rons, which were connected to each other ciative, that is, the animal either associates a would result in the strengthening of the con­ particular behaviour that it carries out with a nection (synapse) between them. He also reward, or it learns to associate two stimuli emphasised the distinction between two that occur in quick succession, using the forms of memory: short term memory, last­ first, usually biologically irrelevant one as a ing minutes or hours, and long term memory, predictor of occurrence of the second, bio­ lasting for days. Over the last forty years, logically relevant stimulus. thanks in large part to the work of Eric

------~~------­ RESONANCE I June 2001 87 RESEARCH I NEWS

Kandel*, we now understand much more therefore possible to consistently perform about the cellular and molecular basis of the recordings of the activity of single, identified different forms of learning and memory. neurons in a number of different individuals and to work out the details of every neuronal Born in Vienna on November 7th, 1929 circuit to the last cell. It was this advantage Kandel emigrated to the United States, where that attracted Kandel to Aprysia. During the he graduated from Harvard College, major­ early years of his work, he used two ap­ ing in history and literature. He then ob­ proaches to develop Aplysia into a model tained a medical degree at the New York system for the study of the cellular basis of University School of Medicine. After a short learning. The first approach was behavioural: stint as a Resident in Psychiatry at the Kandel established that Aplysia was in fact Harvard Medical School (1960-1965) he capable of at least three of the four most moved to Columbia University, New York, widely studied types of learning: habitua­ where he has worked ~ver since. Casting tion, sensitization and classical condition­ around for a model organism in which to ing. And also the gill-withdrawal reflex study the neuronal basis of learning, Kandel behaviour (withdrawing the gill in response eschewed the typical rats, mice and pigeons to touch), could be habituated, sensitized or that psychologists had used for decades to classically conditioned. He thus showed that develop their theories of learning and settled a single behaviour could be modified in dif­ instead on an unfamiliar, unloved and un­ ferent ways. He then went on to painstak­ lovely organism: the humble sea snail Aprysia ingly elucidate the entire neuronal circuitry californica. This decision was to have far­ underlying the gill withdrawal reflex. By de­ reaching consequences, leading ultimately veloping a semi-intact preparation in which to the Nobel Prize. it was possible to simultaneously monitor At first sight,Aprysia hardly appears to be the both the behavioural response of gill with­ kind of organism one would choose for any drawal and the electrical activity of neurons behavioural study, leave alone learning. But, in a behaving animal, he was able to closely in common with a number of other molluscs examine the physiological changes in indi­ to whom we owe much of our current under­ vidual neurons of this circuit. He was able to standing of neurobiology, Aplysia has one pinpoint which of the-neurons in the circuit great asset: a relatively modest number of were modified as a result of learning and in large, individually identifiable neurons. It is what manner they were modified.

*For the sake of clarity of the text, I have avoided giving references to the hundreds of original papers on WhiCh. this article is based. When I refer to Kandtl, it goes without saying that the edifice of his work rests on the generatIons of students, technicians and postdocs that have passed through his laboratory in almost four decades.

______,AAnAA.~------88 v V VV V v RESONANCE I June 2001 RESEARCH I NEWS

Using this approach, he showed that habitu­ which is in turn responsible for the enhanced ation of the gill-withdrawal reflex was ac­ amoun t of neurotransmi tter released (increase companied by a decrease in the strength of in strength of the synapse). Thus, Kandel the synaptic input from sensory neurons showed that, in Aplysia, two different forms (which receive the tactile stimulus) to suc­ of learning, one associative and the other ceeding (postsynaptic) interneurons: there non-associative, shared a common molecular was a decrease in the amount of neurotrans­ mechanism! mitter released. This immediately suggested Further, since most forms of learning are a molecular mechanism for habituation: since reversible (much of what is learned is forgot­ the reserves of readily available (mobilized) neurotransmitter at the synapse are limited, ten!), it follows that memories have finite repeated stimulation would result in deple­ duration. On the basis of retention time, tion of this transmitter pool, decreasing the there are two major forms of memory which amount released on successive stimulation. were again first described by psychologists: This would result in a decreased excitation of short term memory (STM) lasting minutes or at most, hours, and long term memory the following neurons and hence the (LTM) lasting days, weeks or years. Whereas behaviour would no longer be elicited. ST M is induced by just one or a few trials, Kandel also went on to show that the physi­ repeated or prolonged presentation of stimuli ological and molecular basis of both sensiti­ is necessary for L T M to be induced. Very zation and classical conditioning were essen­ interestingly, studies in animals as diverse as tially the same: presynaptic facilitation, or Aplysia, birds and humans have shown that an increase in neurotransmitter release at memory retention is much longer if the train­ synapses, which were thus strengthened. The ing is distributed rather than massed: that is, central player in this process of presynaptic presenting stimuli a large number of times facilitation was a well-known molecule: within a short duration ('cramming') is es­ cAMP (cyclic adenosine mono phosphate), a sentially ineffective in inducing LTM! This cyclic nucleotide involved in a variety of is surely one of the strongest arguments cellular processes. The sensitizing or condi­ against memory-based, examination-oriented tioning stimulus typically causes an increase systems of education. in the level of cAMP which binds to the All forms of learning involve both short and regulatory subunit of an enzyme called pro­ long term memory, suggesting that STM may tein kinase A (PKA), releasing its 'active' or be the first stage of memory acquisition, catalytic subunit. This catalytic subunit of which is then followed by consolidation in PKA, via a biochemical cascade of events, the form ofLTM. It is well known that STM causes an enhancement in the amount of does not require the synthesis of new pro­ calcium entering the neuronal cell terminals, teins or nucleic acids, whereas LTM does.

------~~------~ RESONANCE I June 2001 89 RESEARCH I NEWS

Kandel showed that the physical basis (or that many of the molecular mechanisms elu­ the 'trace') of short term memory was in fact cidated in Aplysia also underlie learning in simply the elevated cAMP level and all its the mammalian hippocampus, although there contingent modifications of protein struc­ are some additional mechanisms, such as ture and function. It follows that the dura­ Hebbian learning, where the activity of the tion of short-term memory is determined by postsynaptic cell is important for the strength­ the turnover time of this cAMP. Having ex­ ening of the synapse. Nonetheless, Kandel plored the cellular and biochemical basis of established that cAMP has a central role in STM, Kandel then turned his attention to the induction of L TM in the mammalian the molecular basis of LTM in Aplysia. He hippocampus as well. showed that the crucial step in the transition Kandel's work showed that very different from STM to L TM is that, with multiple forms of learning may share common physi­ stimulation, some of the 'active' PKA cata­ ological and molecular mechanisms; further, lytic subunit induced by STM moves into the fundam~ntal molecular processes in­ the nucleus of the cell, where it interacts with volved in short and long-term memory ap­ a number of transcription factors (proteins pear to be conserved across a variety of di­ that regulate gene expression). The most verse phyla. Evolutionarily conserved mo­ important of these are CREBI and CREB2 lecular mechanisms of learning and memory (cAMP responsive element binding proteins): will surely stand as one of the most profound CREB2 is a repressor of L TM, whereas discoveries of the twentieth century, vindi­ CREBI induces it. The central role of PKA, cating Darwin's views on the essential unity then, is probably to derepress CREB2 and of all life forms. activate CREB 1: there is now also evidence for such dual control of memory induction Suggested Reading from studies in Drosophila. The final steps of L T M are the strengthening of the connec­ [1] E R Kandel, Cellular Basis ofBehavior, Freeman and tions by the formation of new synapses, thus Company, San Francisco, USA, 1976. [2] E R Kandel, J H Schwartz and T M JesselI, Principles consolidating the memory. ofNeural Science (Third Edition), Appleton & Lange, E Norwalk, Connecticut, USA, 1991. In recent years, Kandel has moved on from [3] C Darwin, The Descent of Man, and Selection in Aplysia to studies of the mammalian hippoc­ Relation to Sex, Appleton, New York, 1871. ampus, the region of the brain on which [4] C H Bailey, D Bartsch and E R Kandel, Toward a many of the studies of L T M in mammals molecular definition of long-term memory storage, Proc. Natl. Acad. Sci., USA Vol. 93, pp.13445-13452, have been focused. The hippocampus is 1996. known to mediate spatial learning, which is somewhat more complex than the simple Rohini Balakrishnan, Centre for Ecological Sciences, paradigms studied in Aplysia. Kandel found Indian Institute of Science, Bangalore 560012, India.

------~~------90 RESONANCE I June 2001