Copyright © 1999 Massachusetts Institute of Technology
How the brain codes, stores, and retrieves memories is among the most important and baffling questions in science. The uniqueness of each human being is due largely to the MEMORY store -- the biological residue of memory from a lifetime of experience. The cellular basis of this ability to learn can be traced to simpler organisms. In the past generation, it has become clear that various forms and aspects of LEARNING and memory involve particular systems, networks, and circuits in the brain, and it now appears possible we will identify these circuits, localize the sites of memory storage, and ultimately analyze the cellular and molecular mechanism of memory.
All aspects of learning share a common thrust. As Rescorla (1988) has stressed, basic associative learning is the way organisms, including humans, learn about causal relationships in the world. It results from exposure to relations among events in the world. For both modern Pavlovian and cognitive views of learning and memory, the individual learns a representation of the causal structure of the world and adjusts this representation through experience to bring it in tune with the real causal structure of the world, striving to reduce any discrepancies or errors between its internal representation and external reality.
Most has been learned about the simplest forms of learning: nonassociative processes of habituation and sensitization, and basic associative learning and memory. Here we focus on CONDITIONING in the mammalian brain. We em-phasize classical or Pavlovian conditioning because far more is known about brain substrates of this form of learning than about more complex instrumental learning. Pavlovian conditioning involves pairing a "neutral" stimulus, for example, a sound- or light-conditioned stimulus (CS) with an unconditioned stimulus (US) that elicits a response, the unconditioned response (UR). As a result of repeated pairings, with the CS onset preceding the US onset by some brief period of time, the CS comes to elicit a conditioned response (CR). Conditioning may be the way organisms, including humans, first learn about the causal structure of the world. Contemporary views of Pavlovian conditioning emphasize the predictive relations between the CS and the US, consistent with cognitive views of learning and memory. The key factor is the contingencies among events in the organism's environment.
When animals, including humans, are faced with an aversive or threatening situation, at least two complementary processes of learning occur. Learned fear or arousal develops very rapidly, often in one trial. Subsequently, the organism learns to make the most adaptive behavioral motor responses to deal with the situation. These observations led to theories of "two-process" learning: an initial learned fear or arousal, followed by slower learning of discrete, adaptive behavioral responses (Rescorla and Solomon 1967). As the latter learning develops, fear subsides. We now think that at least in mammals a third process of "declarative" memory for the events and their relations also typically develops (cf. EPISODIC VS. SEMANTIC MEMORY).
Learned fear develops rapidly, often in one trial, and involves changes in autonomic responses (heart rate, blood pressure, pupillary dilation) and nonspecific skeletal res-ponses (freezing, startle). The afferent limb of the conditioned fear circuit involves projections from sensory relay nuclei via thalamic projections to the AMYGDALA. Although lesions of the appropriate regions of the amygdala can abolish all signs of learned fear, lesions of the efferent targets of the amygdala can have selective effects, for example, lateral hypothalamic lesions abolish cardiovascular signs of learned fear but not behavioral signs (e.g., freezing), whereas lesions of the periqueductal gray abolish learned freezing but not the autonomic signs of learned fear (see, for example, Le Doux et al. 1988). This double disassociation of conditioned responses stresses the key role of the amygdala in learned fear, as do studies involving recording of neuronal activity and electrical stimulation (Davis 1992). The amygdala is critically involved in unlearned fear responses as well. The structures most involved in generating the appropriate responses in basic associative learning and memory seem also to be the most likely sites of memory storage (see below).
Higher brain structures also become critically engaged in learned fear under certain circumstances. Thus when an organism experiences strong shock in a particular environment, reexperiencing that environment elicits learned fear. This context-dependent learned fear involves both the amygdala and the HIPPOCAMPUS for a time-limited period after the experience, a temporal property characteristic of more cognitive aspects of declarative memory (Kim and Fanselow 1992).
A vast amount of research has been done using Pavlovian conditioning of the eye blink response in humans and other mammals (Gormezano, Kehoe, and Marshall-Goodell 1983). The eye blink response exhibits all the basic laws and properties of Pavlovian conditioning equally in humans and other mammals. The basic procedure is to present a neutral CS such as a tone or a light followed a quarter of a second or so later by a puff of air to the eye or a periorbital (around-the-eye) shock (US), the two stimuli terminating together. This is termed the delay procedure. If a period of no stimuli intervenes between CS offset and US onset, it is termed the trace procedure, which is much more difficult to learn than the delay procedure. Initially, there is no response to the CS and a reflex eye blink to the US. After a number of such trials, the eyelid begins to close in response to the CS before the US occurs, and in a well-trained subject, the eyelid closure CR becomes very precisely timed so that the eyelid is maximally closed about the time that the air puff or shock US onset occurs. This very adaptive timing of the eye blink CR develops over the range of CS-US onset intervals where learning occurs, about 100 milliseconds to 1 second. Thus the conditioned eye blink response is a very precisely timed elementary learned motor skill. The same is true of other discrete behavioral responses learned to deal with aversive stimuli (e.g., the forelimb or hindlimb flexion response, head turn, etc.).
Two brain systems become massively engaged in eye blink conditioning, hippocampus and CEREBELLUM (Thompson and Kim 1996). If the US is sufficiently aversive, learned fear also occurs, involving the amygdala, as noted above. Neuronal unit activity in the hippocampus increases in paired (tone CS - corneal air puff US) training trials very rapidly, shifts forward in time as learning develops, and forms a predictive "temporal model" of the learned behavioral response, both within and over the training trials. The growth of this hippocampal neuronal unit response is, under normal conditions, an invariable and strongly predictive concomitant of subsequent behavioral learning (Berger, Berry, and Thompson 1986).
Interestingly, in the basic delay procedure, hippocampal lesions do not impair the eye blink CR, although if the more difficult trace procedure is used, the hippocampal lesions massively impair learning of the CR and, in trained animals, impair memory in a time-limited manner. These results are strikingly consistent with the literature concerned with declarative memory deficit following damage to the hippocampal system in humans and monkeys, as is the hippocampus-dependent contextual fear discussed above. So even in "simple" learning tasks like eye blink and fear conditioning, hippocampus-dependent "declarative" memory processes develop.
The cerebellum has long been a favored structure for modeling a neuronal learning system, in part because of the extraordinary architecture of the cerebellar cortex, where each Purkinje neuron receives 100,000+ excitatory synapses from mossy-parallel fibers but only one climbing fiber from the inferior olive (see below). The reflex eye blink response pathways activated by the US (corneal air puff or periorbital shock) involve direct and indirect relays through the brain stem from the sensory (trigeminal) nucleus to the relevant motor nuclei (largely the seventh and accessory sixth). The CS (e.g., tone) pathway projects to the forebrain and also, via mossy fibers, to the cerebellum. The US (e.g., corneal air puff) pathway projects from the trigeminal nuclei to the forebrain and also, via the inferior olive, as climbing fibers to the cerebellum. These two projection systems converge on localized regions of the cerebellum, where the memory traces appear to be formed. The CR pathway projects from the cerebellar cortex and nuclei (interpositus nucleus) via the red nucleus to the motor nuclei generating the eye blink response. (The cerebellum does not participate in the reflex eye blink response.) A wide range of evidence, including electrophysiological recording, lesions, electrical stimulation, and reversible inactivation during training, has demonstrated conclusively that the cerebellum is necessary for this form of learning (both delay and trace) and that the cerebellum and its associated circuitry form the essential (necessary and sufficient) circuitry for this learning. Moreover, the evidence strongly suggests that the essential memory traces are formed and stored in the localized regions in the cerebellum (see Thompson and Krupa 1994; Lavond, Kim, and Thompson 1993; Yeo 1991).
These results constitute an extraordinary confirmation of the much earlier theories of the cerebellum as a neuronal learning system, first advanced in the classic papers of MARR (1969) and Albus (1971) and elaborated by Eccles (1977) and Ito (1984). These theories proposed that mossy-parallel fibers conveyed information about stimuli and movement contexts (CSs here) and the climbing fibers conveyed information about specific movement errors and aversive events (USs here) and they converged (e.g., on Purkinje neurons in cerebellar cortex and interpositus nucleus neurons).
The cerebellar system essential for a basic form of learning and memory constitutes the clearest example to date of localizing memory traces to particular sites in the brain (i.e., in the cerebellum).
Albus, J. S. (1971). A theory of cerebellar function. Mathematical Bioscience 10:25-61.
Berger, T. W., S. D. Berry, and R. F. Thompson. (1986). Role of the hippocampus in classical conditioning of aversive and appetitive behaviors. In R. L. Isaacson and K. H. Pribram, Eds., The Hippocampus. New York: Plenum Press, pp. 203-239.
Davis, M. (1992). The role of the amygdala in fear and anxiety. Annual Review of Neuroscience 15:353-375.
Eccles, J. C. (1977). An instruction-selection theory of learning in the cerebellar cortex. Brain Research 127:327-352.
Gormezano, I., E. J. Kehoe, and B. S. Marshall-Goodell. (1983). Twenty years of classical conditioning research with the rabbit. In J. M. Sprague and A. N. Epstein, Eds., Progress in Physiological Psychology. New York: Academic Press, pp. 197-275.
Ito, M. (1984). The Cerebellum and Neural Control. New York: Raven.
Kim, J. J., and M. S. Fanselow. (1992). Modality-specific retrograde amnesia of fear. Science 256:675-677.
Kim, J. J., and R. F. Thompson. (1997). Cerebellar circuits and synaptic mechanisms involved in classical eyeblink conditioning. Trends in Neurosciences 20:177-181.
Krupa, D. J., J. K. Thompson, and R. F. Thompson. (1993). Localization of a memory trace in the mammalian brain. Science 260:989-991.
Lavond, D. G., J. J. Kim, and R. F. Thompson. (1993). Mammalian brain substrates of aversive classical conditioning. Annual Review of Psychology 44:317-342.
LeDoux, J. E., J. Iwata, P. Cicchetti, and D. J. Reis. (1988). Different projections of the central amygdaloid nucleus mediate autonomic and behavioral correlates of conditioned fear. Journal of Neuroscience 8:2517-2529.
Marr, D. (1969). A theory of cerebellar cortex. Journal of Physiology (London) 202:437-470.
McGaugh, J. L. (1989). Involvement of hormonal and neuromodulatory systems in the regulation of memory storage. Annual Review of Neuroscience 12:255-287.
Rescorla, R. A. (1988). Behavioral studies of Pavlovian conditioning. Annual Review of Neuroscience 11:329-352.
Rescorla, R. A., and R. L. Solomon. (1967). Two-process learning theory: Relationships between Pavlovian conditioning and instrumental learning. Psychological Review 74:151-182.
Thompson, R. F., and J. J. Kim. (1996). Memory systems in the brain and localization of a memory. Proceedings of the National Academy of Sciences 93:13438-13444.
Thompson, R. F., and D. J. Krupa. (1994). Organization of memory traces in the mammalian brain. Annual Review of Neuroscience 17:519-549.
Yeo, C. H. (1991). Cerebellum and classical conditioning of motor response. Annals of the New York Academy of Sciences 627:292-304.
Zola-Morgan, S., and L. R. Squire. (1993). Neuroanatomy of memory. Annual Review of Neuroscience 16:547-563.
Copyright © 1999 Massachusetts Institute of Technology