392 THE MOLECULAR BIOLOGY OF MEMORY STORAGE: A DIALOG BETWEEN GENES AND SYNAPSES Nobel Lecture, December 8, 2000 by Eric R. Kanpet, Howard Hughes Medical Institute, Columbia University, College of Physicians and Surgeons, New York, USA. INTRODUCTION One of the most remarkable aspects of an animal's behavior is the ability to modify that behavior by learning, an ability that reaches its highest form in human beings. For me, learning and memory have proven to be endlessly fas- cinating mental processes because they address one of the fundamental fea- tures of human activity: our ability to acquire new ideas from experience and to retain these ideas in memory. In fact, most of the ideas we have about the world and our civilization we have learned so that we are who we are in good measure because of what we have learned and what we remember. However, not all learning experiences are positive. Many psychological and emotional problems also result at least in part from our experiences. In addition, spe- cific disorders of learning and memory haunt both the infant and the adult Down syndrome, fragile X mental retardation, age-related loss of memory, and the devastation of Alzheimer’s disease are only the familiar examples of a large number of disorders that affect memo Throughout my career I have been interested in the biology of learning. I have been curious to know: What changes occur in the brain when we learn? And, once something is learned, how is that information retained in the brain as memory? I have tried to address these questions by developing a reduc- tionist approach that would allow me to investigate the biological mecha- nisms underlying elementary forms of learning and memory at a cell and molecular level — as specific molecular activities within specific identified nerve cells. For a biologist like myself, interested in mental processes, the study of learning has the further appeal that, unlike other mental processes such as thought, language and consciousness, learning is relatively accessible to a cel- lular and molecular analysis. Elementary forms of learning and memory have been well characterized by classical psychology since the work of Ivan Pavloy and Edgar Thorndike in the first half of the 20th century, and these forms of learning are the most clearly delineated and, for the experimenter, most easily controlled of any mental process. T first became interested in the study of memory in 1950, as a result of myEric R. Kandel 393, readings in psychoanalysis while still an undergraduate at Harvard College. Later, during medical training, I found the psychoanalytic approach limiting because it tended to treat the brain, the organ that generates behavior, as a black box. In the mid 1950s, while still in medical school, I began to appre- ciate that during my generation the black box of the brain would be opened and that the problems of memory storage, once the exclusive domain of psy- chologists and psychoanalysts, could be investigated with the methods of modern biology. As a result, my interest in memory shifted from a psycho- analytic to a biological approach. As a postdoctoral fellow at the National Institutes of Health (NIH) in Bethesda from 1957 to 1960, I focused on learn- ing more about the biology of the brain, and became interested in knowing how learning produces changes in the neural networks of the brain and how a transient shortterm memory is converted to an enduring long-term me- mory. From the beginning, my purpose in translating questions about the psy- chology of learning into the empirical language of biology was not to replace the logic of psychology or psychoanalysis with the logic of cell and molecular biology, but to try to join these ewo disciplines and to contribute to a new syn- thesis that would combine the mentalistic psychology of memory storage with the biology of neuronal signaling. As I thought more concretely about the neural mechanisms of memory storage, I hoped further that the biological analysis of memory might carry with it an extra bonus, that the study of me- mory storage might reveal new aspects of neuronal signaling. Indeed, this has proven true. Time and again, the molecular study of memory has revealed novel aspects of more general biological processes. DEVISING A RADICAL REDUCTIONIST STRATEGY TO LEARNING AND MEMORY At first thought, someone interested in learning and memory might be tempted to tackle the problem in its most complex and interesting form. This was the approach that my colleague Alden Spencer and I originally had in 1958 when, at the start of our scientific careers, we joined forces at the NIH to study the cellular properties of the hippocampus, the part of the mammalian brain thought to be most directly involved in aspects of complex memory (1) We were initially interested in a simple question: Are the electrophysiological properties of the pyramidal cells of the hippocampus, which were thought to be the key hippocampal cells involved in memory storage, fundamentally dif- ferent from other neurons in the brain, such as the well-studied motor neu- rons in the spinal cord involved in simple movement? In the course of study- ing the pyramidal cells of the hippocampus, it became clear to us that all nerve cells have similar signaling properties. Therefore, the intrinsic signal- ing properties of neurons would themselves not give us key insights into memory storage (17) Thus, the unique functions of the hippocampus had to arise not so much from the intrinsic properties of pyramidal neurons but from the pattern of304 Physiology or Medicine 2000 functional interconnections of these cells, and how those interconnections are affected by learning. To tackle that problem we needed to know how sen- sory information about a learning task reaches the hippocampus, and how in- formation processed by the hippocampus influences behavioral output. This was a formidable challenge, since the hippocampus has a large number of neurons and an immense number of interconnections. It seemed unlikely that we would be able to work out in any reasonable period of time how the neural networks, in which the hippocampus was embedded, participate in be- havior and how those networks are affected by learning. rhus, to bring the power of modern biology to bear on the study of learn- ing, it seemed necessary to take a very different approach ~ a radically reduc- tionist approach. Instead of studying the most complex cases, we needed to study the simplest instances of memory storage, and to study them in the ele- mentary reflex behavior of those animals that were most experimentally tractable. To do this we needed to find experimental systems in which a sim- ple behavioral act that was controlled by a small number of large and acces- sible nerve cells could be modified by learning. Only in this way could we cor- relate changes in the overt behavior of the animals with molecular events in identifiable neurons and examine how sensory processing in the brain is modified by learning to give rise to memories. Such a reductionist approach was hardly new in 20th century biology. One need only think of the use of Drosophila in genetics, of bacteria and bacterio- phages in molecular biology, and of the squid giant axon in the study of the conduction of nerve impulses. Nevertheless, when it came to the study of be- havior, many investigators were reluctant to use a reductionist strategy. In the 1950s and 1960s many biologists and psychologists believed that behavior was the one area of biology in which the use of simple animal models, particular- ly invertebrate ones, was least likely to succeed. They argued that only higher animals exhibit interesting forms of learning and that these forms require neuronal organizations and neuronal mechanisms qualitatively different from those found in simple animals. As a result, an approach to learning based on simple invertebrates was bound to fail because it would lack rele- vance to mammalian and particularly to human behavior. It was my belief at the outset, however, that concerns about the use of a simple experimental system to study learning were misplaced. The question was and is not whether there is something special about the human brain; there clearly is. Rather, the question is whether the human brain and human behavior have anything at all in common with the nervous system and be- havior of simpler animals. If so, these fundamental, common principles of neuronal organization might well be studied more profitably in simple ani- mals. The answer to this second question, about commonality, was clear. By 1960, work by students of comparative behavior such as Konrad Lorenz, Niko Tinbergen, and Karl von Frisch had shown that humans share many beha- vioral patterns and even simple forms of learning with simple animals (re- viewed in 2). That the evolution of behavior and learning is conservativeEric R. Kandel 395 should not be surprising, since the evolution of other biological functions is also conservative. There are, for example, no fundamental functional or bio- chemical differences between the nerve cells and synapses of humans and those of a snail, a worm or a fly. Since behavior and learning is an expression of nerve cell activity, it would be surprising if the learning capability of people did not have some elementary features in common with the learning of snails, worms, or flies. And, if elementary forms of learning are common to all ani- mals with an evolved nervous system, there must be conserved features in the mechanisms of learning at the cell and molecular level, that can be studied ef fectively even in simple invertebrate animals. A SIMPLE INVERTEBRATE SYSTEM THAT LENDS ITSELF TO A REDUCTIONIST APPROACH After an extensive search for a suitable experimental animal, I settled on the giant marine snail Aplysia (Figure 1), because it offers three important ex- perimental advantages: its nervous system is made up of a small number of nerve cells; many of these are very large; and (as became evident to me later) many are uniquely identifiable (3,13). Whereas the mammalian brain has a million-million central nerve cells, Aplysia has only 20,000. These cells are clustered in 10 anatomical units called ganglia, each of which contains about 2,000 cells (Figure 2). An individual ganglion, such as the abdominal gang- lion, mediates not one but a family of behaviors. Thus, the simplest behaviors that can be modified by learning may involve less than 100 cells. This nume- rical simplification made it possible to identify the specific contribution of in- dividual neurons to the behavior in which they participate (13). In addition to being few in number, these cells are the largest nerve cells in the animal kingdom, reaching up to 1,000 um in diameter, large enough to be seen with the naked eye (Figure 3). Because of their extraordinary size and their di- stinctive pigmentation, it is possible to recognize many of the cells as unique individuals. One can record from these large cells for many hours without any difficulty, and the same cell can be returned to and recorded from over a pe- Figure 1. The giant marine snail Aplysia californica below hobranch subclass of the gastropod molluscs. pecies may grow to be 30 cm in length an weigh 1 kilogram.396 Physiology or Medicine 2000 Human Brain Aplysia Brain 10"? Neurons 2x 104 Neurons Buccal Cerebral Pleural ~ Pedal Abdominal Figure 2. The human brain has a million-million neurons while the brain of Aplysia has 20,000 nerve cells, Apisia nerve cells are clustered together in five major bilateral pairs of ganglia, each ganglion containing about 2,000 nerve cells. riod of days. The cells can easily be dissected out for biochemical studies, so that from a single cell one can obtain sufficient mRNA to make a CDNA library. Finally, these identified cells can readily be injected with labeled com- pounds, antibodies, or genetic construct procedures which opened up the molecular study of signal transduction within individual nerve cells. REQUIREMENT FOR A CELL-BIOLOGICAL STUDY OF MEMORY STORAGE Given a technically advantageous experimental system, how does a cell biolo- gist begin to address the problem of learning? The strategy that my col- leagues and I developed involved four sequential steps: (1) We first wanted to define a simple behavior that can be modified by learning and that gives rise to memory storage. (2) We next wanted to identify the cells that make up the neural circuit of that behavior. (3) Within that neural circuit we then wanted to locate the critical neurons and interconnections that had been modified by learning and that store memory. (4) Finally, we wanted to analyze the changes that occur at those sites in response to learning and memory storage, first on Figure 3. A photomicrograph of an abdominal ganglion of Apsia shows the distinctive pig- mentation and positions of its cells. The largest limeter in diameter and can be se- mm. (See cells area en with the naked eye. Scale bar also Fig. 6)Eric R. Kandel 397 the cellular and then on the molecular level (13). I follow this outline in the discussion below. DELINEATING A BEHAVIOR IN APLYSIA THAT IS CAPABLE OF BEING MODIFIED BY LEARNING Irving Kupfermann and I first wanted to study the simplest possible behavior of Aplysia (4). We examined the animal's behavioral capabilities and deli- neated a very simple defensive reflex: the withdrawal of the gill upon stimu- lation of the siphon, an action that is like the quick withdrawal of a hand from a hot object. In Aplysia, the gill is a respiratory organ that lies exposed in the mantle cavity. When the animal is in a normal, relaxed state, the gill is par~ tially covered by a sheet of skin (the mantle shelf), which ends in a fleshy spout, the siphon (Figure 1). When a weak tactile stimulus is applied to the siphon, both the siphon and gill are withdrawn into the mantle cavity and for protection under the mantle shelf (Figure 4); ref. 5) Gillwithdrawel reflex Sensitization stimulus Figure 4. A do the mantle cavity and is normally covered pai siphon with a fine probe causes the siphon to ¢ cavity under the protection of the mantle sh for a better view of the gill. Sensitization of the gillwitharawal refle noxious stimulus to another part of the body, such as the tail, leads to an enh withdrawal reflex of both the siphon and the gi yy the mantle shelf. A light touch to the ‘Land the gill to withdraw into the mantle the mantle shelf is shown to be retracted produced by applying a pent of the Kupfermann, Harold Pinsker, and later Tom Carew, Robert Hawkins and I found that this simple reflex could be modified by three different forms of learning: habituation, sens in, and classical conditioning (5,8,9). As we examined these three forms of learning, we were struck by the resemblance each had to corresponding forms of memory storage in higher vertebrates and humans. As with vertebrate learning, memory storage for each type of learning in Aplysia has two phases: a transient memory that lasts minutes and an enduring memory that lasts days. Conversion of shortterm to long-term memory storage requires spaced repetition — practice makes perfect even in snails (8,9,10).398 Physiology or Medicine 2000 We focused initially on one type of learning, sensitization, a form of learned fear in which a person or an experimental animal learns to respond strongly to an otherwise neutral stimulus (5,8,10). For example, if somone is sudden- ly exposed to an aversive stimulus, such as a gunshot going off nearby, that person will be sensitized by the unexpected noise. As a result, that person will be frightened and will now startle to an otherwise innocuous stimulus like a tap on the shoulder. Similarly, on receiving an aversive shock to another part of the body such as the tail (or head), an Aphsia recognizes the stimulus as aversive and learns to enhance its defensive reflex responses to a variety of subsequent stimuli applied to the siphon, even innocuous stimuli (Figure 4; 12). The animal now remembers the shock, and the duration of this memory is a function of the number of repetitions of the noxious experience. A single shock gives rise to a memory lasting only minutes; this short-term memory does not require the synthesis of new protein. In contrast, four or five spaced shocks to the tail give rise to a memory lasting several days; this long-term memory does require the synthesis of new protein. Further taining gives rise to an even more enduring memory lasting weeks, which also requires new protein synthesis (Figure 5; 10,12). 1000: 4 trains / day _, for 4 days ie 4 single shocks Duration of withdrawal (% of control) Days after training Figure 5. Spaced repetition converts short-term memory into long-term memory in Aplysia. In the resting state, before sensitization training, a weak touch to the siphon causes only a weak and brief siphon and gill withdrawal reflex. Following a single noxious, sensitizing, shock to the tail, the siphon and gill withdrawal reflex repsonse to that same weak touch is enhanced and produces a much larger reflex withdrawel response. The memory following this single noxious stimulus lasts about one hour. After four or five single tail shocks, the animal's gill and siphon withdrawel re- flex to that same weak touch is enhanced further and now lasts more than a day. Ifthe animal re- ceives further training - four brief trains of single shocks a day over the course of four days ~ the siphon and gill reflex withdrawal to the same weak siphon touch is enhanced still further and the memory now lasts for several weeks. (Modified from 98.)Eric R. Kandel 399 Thus, just as in the complex learning in mammals (107,108), long-term sensitization differs from the short-term process in requiring the synthesis of new proteins. This was our first clear evidence for the conservation of bio- chemical learning mechanisms between Aplysia and vertebrates, and it rein- forced the hope that a detailed analysis of short-term memory and its transi- tion to long-term memory in Aplysia would reveal molecular mechanisms of general importance. DEFINING THE NEURAL CIRCUIT IN CELLULAR DETAIL To analyze the cellular mechanisms of sensitization, we needed to identify the neural circuit of the gill withdrawal reflex. Kupfermann and I quickly loc: ized the central neuronal machinery for the reflex behavior in the animal’s abdominal ganglion (4,6). Because we soon realized that many cells in Aplysia could be identified in every animal of the species (3,4,6,7), we were able to give the cells specific names and, most important, return to the same cell time and again - in both untrained and trained animals. In this way Kupfermann, Castellucci, Carew, Hawkins, John Byrne, and I were able to work out signifi- cant components of the neural circuit gill-withdrawal reflex in terms of indi- vidual cells and cell clusters. The circuit has 24 mechanoreceptor sensory neurons that innervate the siphon skin and make direct monosynaptic con- nections with six gill motor ceils (7,11,36). The sensory neurons also made in- direct connections with the motor cells through small groups of excitatory and inhibitory interneurons (18,19). In addition to being identifiable, indi- vidual cells also proved to have surprisingly large effects on behavior (Figure 6C; reviewed in 13,22,36) ‘As we examined the neural circuit of this reflex in detail, we were struck by its invariance ~ the cells that make up the circuit and their interconnections are always the same. In every animal we examined, each cell connected only to certain target cells and not to others (Figure 7). Carew, John Koester, Wayne Hening, and I also found this invariance in the neural circuitry of other behaviors in Aplysia — inking, control of the circulation, locomotion (reviewed in 13,15) - raising a key question in the celb-biological study of learning: How can learning occur in a neural circuit that is precisely wired? HOW DOES LEARNING AFFECT THE INVARIANT ELEMENTS OF THE. NEURAL CIRCUI In his Croonian Lecture to the Royal Society of 1894, Santiago Ramon y Cajal proposed a theory of memory storage: memory is stored in the growth of new connections (16). This prescient idea was neglected in good part for half a century as students of learning fought over newer competing ideas. First, Kar] Lashley, Ross Adey, Wolfgang Kohler, and a number of Gestalt psychologists proposed that learning leads to changes in electric fields or chemical gradi- ents, which they postulated surround neuronal populations and are pro- ity of cells recruited by the learning process.400 Physiology or Medicine 2000 L7 motorneuron LE sensory neurons Siphon vo We Noe HLL too A Re Figure 6. The neural circuit of the gill-withdrawal reflex can be delineated in terms of specific identified cells ‘A. This dorsal view of the abdominal ganglion shows some of the identifiable cells (compare with Fig. 3). The six identified motor cells to the gill are shaded brown; seven sensory neurons are shaded blue. In the figure a sensory neuron that synapses on gill motor neuron L7 is being stimulated electrically and a microelectrode in the motor neuron records the synaptic potential produced by the action potential in the sensory neuron (sce middle trace in part C). B. The physiological demonstration of the direct connections between the sensory neuron and motor neuron. The sensory neuron receives input from the siphon skin; the motor neuron makes direct connections onto the gill. The fact that the cells are large and identifiable allows for the mapping of connections between specific identified cells. This part shows the experimental arrangement for simultaneously recording in a pre-and postsynaptic cell, The sensory neuron makes a direct connection onto the motor neuron, as is evident in the middle trace in part C. C. Individual cells make significant contributions to the reflex. Stimulating a single motor neu ron (traces on the left) produces a detectable change in the gill, and stimulating a single senso- ry neuron produces a large synaptic potential in the motor neuron (traces in the middle) Repeated stimulation of a single sensory neuron increases the frequency of firing in the motor neuron, leading to a visible reflex contraction of the gill (traces on the right). A single tactile sti- mulus (0 the skin normally activates 6 to 8 of the 24 sensory neurons, causing each to fire 1 to 2 action potentials. The repetitive firing of 10 action potentials in a single sensory neuron, desig- ned to simulate the firing of the total population, in fact simulates reasonably well the reflex be- havioEric R. Kandel 401 Siphon ( 7 inh, Modulatory MN. interneurons | EXC. ©) ae : interneurons Figure 7. The neural circuit of the gilkwithdrawal reflex. The siphon is innervated by 24 sensory neurons that connect directly with the six motor neurons. The sensory neurons also connect to populations of excitatory and inhibitory interneurons that in turn connect with the motor neu- rons. Stimulating the tail activates modulatory interneurons that act on the terminals of the sen- sory neurons as well as on those of the excitatory interneurons. There are three classes of modu- latory neurons activated by tail stimuli: (1) neurons that release serotonin (5-HT), (2) neurons that release a peptide called the small cardioactive peptides (SCP), and (3) the L29 cells, which release an unidentified modulatory neurotransmitter. The serotonergic modulatory action is the most important. Blocking the action of these serotonergic cells blocks the effects of sensitizing stimuli, Second, Alexander Forbes and Lorente de N6 proposed that memory is stored dynamically by a selreexciting chain of neurons. This idea was later championed by Donald Hebb as a mechanism for short-term memory. Finally, Holger Hyden proposed that learning led to changes in the base composition of DNA or RNA. Even though there was much discussion about the merits of each of these ideas, there was no direct evidence to support any of them (re- viewed in 17). ‘We were now in a position to address these alternative ideas by confronting directly the question of how learning cans occur ina circuit with fixed neuro- nal elements. Kupfermann, Castellucci, Carew, Hawkins, and I examined the neural circuit of the gill-withdrawal reflex while the animal underwent neither sensitization or habituation, a form of learning in which the animal learns to ignore an innocuous stimulus to siphon when given with monoto- nous repetition. (We later also extended these studies to an examination of classical conditioning (20).) Our studies provided clear evidence for Cajal’s idea: learning results from changes in the strength of the synaptic connec- tions between precisely interconnected cells (6,7). Thus, while the organism's developmental program assures that the connections between cells are in-402 Physiology or Medicine 2000 variant, it does not specify their precise strength. Rather, the strength and ef- fectiveness of these preexisting chemical connections can be altered by expe~ rience. Seen in the perspective of these three forms of learning, synaptic pla- sticity emerged as a fundamental mechanism for information storage by the nervous system, a mechanism that is built into the very molecular architec- ture of chemical synapses (95). We soon appreciated that the synaptic strength of a given chemical synapse could be modified in two ways: homosynaptically and heterosynaptically. Homosynaptic changes in synaptic strength occur in a synapse because of ac- tivity in the presynaptic and postsynaptic neurons of that very synapse. During habituation, homosynaptic changes occur in the monosynaptic connections between the sensory neurons and the motor neurons of the gill-withdrawal re- flex. Heterosynaptic changes occur in a synapse where presynaptic and post- synaptic neurons are themselves not active but there is, instead, activity in one or more modulatory interneurons that act on the presynaptic neurons, on the postsynaptic neurons of the synapse, or on both, to modify the strength of their synaptic connections. During sensitization, heterosynaptic changes are induced in the monosynaptic connections between the sensory neurons and motor neurons of the gillwithdrawal reflex. Hawkins, Abrams, and I later found that these two types of regulation are recruited together in classical conditioning (20,21). Classical conditioning therefore illustrated that the el- ementary forms of homo- and heterosynaptic plasticity form an alphabet of basic mechanisms that can produce combinations of plasticity with novel properties. In all three forms of learning we found particularly large changes in the synaptic strength of the direct connections between the sensory and motor s of the reflex. We therefore focused on this one component of the re- flex and found several additional principles that have proven to be quite gen- eral. First, we found that the same synaptic connection can be modulated in opposite ways by different forms of learning, For example, habituation leads to a homosynaptic weakening of synaptic connections between the sensor neurons and their target cells, the motor neurons and interneurons, while sensitization leads to heterosynaptic strengthening of thes nections (7,13,33). Second, learning not only leads to changes in synaptic strength, it can also affect the excitability of neurons. In the case of sensitiza- tion the excitability of the sensory (presynaptic) neurons is increased (25). Third, the synaptic changes persist, thereby contributing to memory storage (10,33,94). Indeed, the same synaptic connection can store both short- and long-term memory (7,23,33,43,94,98). At a given synapse synaptic plasticity can either be short- or long-lived depending on the number of spaced repe- titions of the learning stimulus, and these parallel not only the behavioral changes of short-term memory but also those of long-term memory (7,43,93). Finally, long-term memory storage involves not only a change in synaptic strength, but also anatomical changes, changes in the number of synaptic connections (57,58,59).. The changes at the synapse between the sensory and the motor neuron are neuror same sets of con-Eric R. Kandel 403, only a part of the changes in the neural circuit of the gill-withdrawal reflex. Important changes occur elsewhere in the circuit, but we have studied them less. We have focused on the monosynaptic portion of the circuit in order to probe in depth the molecular mechanisms that contribute to learning and memory. INITIAL STEPS TOWARD A MOLECULAR CHARACTERIZATION OF SENSITIZATION What are the molecular mechanisms whereby short-term memory is estab- lished, and how is it converted to long-term memory? Initially, we focused on shortterm sensitization (Figure 8). In collaboration with James H. Schwartz, we found that the synaptic changes, like the short-term behavior, were expressed even when protein synthesis was inhibited. Since the short- term changes persisted for many minutes, it seemed unlikely that they in- volved a simple conformational change of one or more proteins. We there- fore proposed in 1971, that short-term memory might require a series of sequential reactions similar to that mediated by cAMP-mediated signaling (26). Our attention was drawn to cAMP because Sutherland, Rall and Krebs had found that various neurotransmitters could increase cAMP concentration in the brain and in other tissues, and in the liver cAMP activated the cAMP- dependent protein kinase (26). In 1972, Schwartz, Howard Cedar, and 1 found that stimulation of the modulatory pathways recruited during he- terosynaptic facilitation led to an increase in cAMP in the abdominal gang- lion (27). Cedar and Schwartz next applied a number of neurotransmitter candidates and found that serotonin and dopamine could increase levels of cAMP (29). This made us wonder whether the population of modulatory in- terneurons, which produces the heterosynaptic facilitation that gives rise to sensitization, contains serotonergic or dopaminergic cells. Later, Hawkins, Castellucci, David Glanzman, and I delineated the modu- latory system activated by a sensitizing stimulus to the tail (19,34,35), and con- firmed that it contains serotonergic interneurons. The system includes two other classes of heterosynaptic modulators, but the serotonergic neurons are most important for sensitization: blocking their actions alone blocked sensi- tization (34). A DISTINCTION BETWEEN MEDIATING AND MODULATORY CIRCUITRY In 1974, Castellucci and I next attempted to localize the change produced by habituation and sensitization to either the presynaptic or the postsynaptic component of the synapse. We therefore applied a quantal analysis to the synaptic connections between the sensory and motor cells and found that the short-term homosynaptic depression that accompanies habituation is presy- naptic, involving a reduction in the amount of transmitter released from the presynaptic sensory neuron (30). This transmitter, Nicholas Dale later404 Physiology or Medicine 2000 A Control Tail Shock Motor neuron Sensory 2mv neuron 20 mv 50 ms B Control Serotonin Motor neuron a a A. 2mv Benen ae 50 ms c Control Cyclic AMP Motor UNM neuron 72 hours) that is also restricted to the stimulated synapse (Figure 15). This long-lasting synapse-specific facilitation requires CREB and also leads to structural changes. Thus, despite recruitment of nuclear proces- ses, long-term changes in synaptic function and structure are confined only to those synapses stimulated by serotonin. A Initiation ® Capture 5 x SHT ye 1x5HT B___synapse-Specific Facilitation: Cell A sé EPSP amplitude change (%) 8 ° O14 12 24 48 72 C synaptic Capture: Cell B gz 160 3 E 100 g 2 60 E & 0 g & Time (h) Figure 15. Long-term facilitation is synapse-specific and can be captured at another branch by the stimulus that initiates the short-term process. A. Five puffs of serotonin are applied at the site of initiation (cell A) and produce a synapse- specific facilitation that is shown in part B. This synapse-specific facilitation is not evident at the synapse of cell B unless that synapse is itself primed with a single puff of serotonin. Under those circumstances the long-term facilitation at synapse A can be captured at synapse B in a reduced form, as illustrated in part C. (From 64.)