Development of Processes Underlying Learning

In the 1980s exciting progress was made in understanding a variety of developmental processes, ranging from principles governing the birth, differentiation, and migration of nerve cells to the mechanisms underlying the functional assembly of complex neural circuits. In addition to the intrinsic interest in development as a fundamental field of inquiry, the analysis of development has a secondary gain: By affording the experimental opportunity of examining early-emerging processes in functional isolation from later-emerging ones, development can serve as a powerful analytic tool with which to dissect and examine specific behavioral, cellular, and molecular processes as they are expressed and integrated during ontogeny.

A developmental strategy such as that described above has been useful in furthering the analysis of learning and memory in the marine mollusk Aplysia, a preparation that has proved to be quite powerful for cellular and molecular studies of several forms of learning. Specifically, using this strategy, it has been possible to identify and dissociate multiple components of nonassociative learning on both behavioral and cellular levels. This type of analysis has also revealed previously unappreciated behavioral and cellular processes in Aplysia. Moreover, the developmental dissociation of different components of learning in juvenile Aplysia prompted a similar analysis in adult animals, where the same clearly dissociable components of learning were identified. Thus an analysis of the developmental assembly of learning has provided important insights into the final phenotypic expression of learning in the adult.

The Development of Aplysia

The life cycle of Aplysia can be divided into five phases:

1. an embryonic phase (lasting about ten days, from fertilization to hatching)
2. a planktonic larval phase (lasting about thirty-five days)
3. a metamorphic phase (lasting only two to three days)
4. a juvenile phase (lasting at least ninety days)
5. the adult phase, defined as the onset of reproductive maturity.

These five phases can be further divided into thirteen discrete stages, each defined by a specific set of morphological criteria (Kriegstein, 1977). In the analysis of learning that will be described here, most work has focused on the juvenile phase of development (stages nine through twelve), since it is during this time that many of the behavioral systems of interest emerge.

Forms of Learning and Developmental Timetables

In adult Aplysia the siphon withdrawal reflex exhibits both nonassociative and associative forms of learning. The developmental analysis thus far carried out in Aplysia has focused primarily on nonassociative learning in that reflex. The three most common forms of nonassociative learning are habituation, dishabituation, and sensitization. Habituation refers to a decrease in response magnitude occurring as a function of repeated stimulation to a single site; dishabituation describes the facilitation of a habituated response by the presentation of a strong or novel stimulus, usually to another site; sensitization refers to the facilitation of nondecremented responses by a similar strong or novel stimulus. Using a behavioral preparation that permitted quantification of siphon withdrawal throughout juvenile development, Rankin and Carew (1987, 1988) found that these three forms of nonassociative learning emerged according to different developmental timetables. Habituation of siphon withdrawal was present very early (in stage nine) and progressively matured across all juvenile stages in terms of its interstimulus interval (ISI) function: In young animals, extremely short ISIs were necessary to produce habituation, whereas in older animals, progressively longer ISIs could be used to produce habituation. Dishabituation (produced by tail shock) emerged soon after habituation, in a distinct and later stage (stage ten). However, sensitization (also produced by tail shock) did not emerge until surprisingly late in juvenile development (stage twelve), at least sixty days after the emergence of dishabituation (see Figure 1).

The observation that dishabituation and sensitization can be developmentally dissociated raises important theoretical questions for a complete explanation of nonassociative learning. For example, until recently a commonly held view was that nonassociative learning could be accounted for by a dual process theory involving two opposing processes: a single decrementing process that gives rise to habituation, and a single facilitatory process that underlies both dishabituation and sensitization (Carew, Castellucci, and Kandel, 1971; Groves and Thompson, 1970). A key prediction of this view is that dishabituation and sensitization should always occur together. However, the developmental dissociation of these processes, together with recent behavioral and cellular evidence in adult Aplysia, suggests that a dual-process view is inadequate to account for nonassociative learning in Aplysia.

The emergence of sensitization in stage twelve is not confined to the siphon withdrawal reflex in Aplysia. Stopfer and Carew (1988) examined another response system, escape locomotion, and found that sensitization in that system also emerges in stage twelve. Thus sensitization is expressed in two different systems, one a graded reflex and the other a centrally programmed cyclical behavior, at the same time in development. This raises the interesting hypothesis that one or more developmental signals may switch on the general process of sensitization in stage twelve, not only in individual response systems but in the whole animal.

Cellular Analogues of Learning and Behavioral Learning

The developmental separation of different learning processes described above provides the opportunity to examine the unique contribution of specific cellular mechanisms to each form of learning. An important step in such a cellular analysis is to show that the cellular analogue of each form of learning can be identified in the central nervous system of juvenile Aplysia and that these analogues exhibit a developmental time course parallel to the behavioral expression of the learning. The identified motor neuron can serve as a reliable cellular monitor of plasticity in the afferent input to the siphon-withdrawal reflex.

The developmental emergence of the cellular analogue of habituation (synaptic decrement) and of dishabituation (facilitation of decremented synaptic potentials) was first examined by Rayport and Camardo (1984). They found that synaptic decrement could be observed in neuron R2 as early as stage nine, and that facilitation of depressed synaptic potentials emerged in stage ten. Nolen and Carew (1988) then examined the emergence of the cellular analogue of sensitization (facilitation of nondecremented synaptic potentials) in R2. They found that the analogue of sensitization emerged between early and middle stage twelve, many weeks after the emergence of the analogue of dishabituation. Taken collectively, these results illustrate two important points. First, the cellular analogue of each form of learning emerges in close temporal register with its respective behavioral form (see Figure 1). Second, just as dishabituation and sensitization can be developmentally dissociated on a behavioral level, so can their cellular analogues be developmentally dissociated.

A Novel Inhibitory Process

When the effects of sensitization training (i.e., the effects of tail shock on nondecremented reflex responses) in different developmental stages were examined, an unexpected effect of tail shock was discovered: Prior to the emergence of sensitization in stage twelve, tail shock had an inhibitory effect on reflex responsiveness (Rankin and Carew, 1988). The properties of this inhibitory process in juvenile Aplysia have been studied by Rankin and Carew (1989), who found that tail shock-induced inhibition of siphon withdrawal can be detected in two ways: 1. by reduction of reflex responsiveness; and 2. by the apparent competition of the inhibitory process with the facilitatory process of dishabituation. Specifically, they found that as levels of tail shock were increased, progressively more inhibition resulted and, concomitantly, progressively less dishabituation was produced, suggesting the hypothesis that the tail shock-induced inhibition could significantly retard the expression of dishabituation in early developmental stages. Finally, as the process of sensitization matured, there was a clear transition from the inhibitory effect of tail shock to reflex facilitation between early and late stage twelve.

In parallel with the behavioral reflection of inhibition described above, Nolen and Carew (1988) identified a clear cellular analogue of this inhibitory process. Specifically, they found that prior to the emergence of the cellular analogue of sensitization in mid to late stage twelve, activation of the pathway from the tail produced significant inhibition of non-decremented synaptic responses in neuron R2 (see Figure 1). As with the behavior, there was a clear transition from inhibition to facilitation in mid to late stage twelve.

The inhibitory process first identified in juvenile Aplysia has received considerable attention in the adult. Several laboratories have observed behavioral tail shock-induced inhibition of the siphon withdrawal reflex (Krontiris-Litowitz, Erikson, and Walters, 1987; Mackey et al., 1987; Marcus et al., 1988), and important progress has been made in studying the cellular mechanisms underlying the inhibitory process. For example, Mackey et al. (1987) found that tail shock produced presynaptic inhibition of the transmission from siphon sensory neurons. Wright, Marcus, and Carew (1991) found that polysynaptic input to the siphon motor neurons plays an important role in mediating tail shock-induced inhibition, and Wright and Carew (1990) found that a single identified inhibitory interneuron in the abdominal ganglion, cell L16, can account for most, if not all, of the inhibition of siphon withdrawal following tail shock. Finally, Fitzgerald and Carew (1991) found that serotonin, a known facilitatory neuromodulator in Aplysia, can also mimic the inhibitory effects of tail shock. It will be of considerable interest to study the development of these inhibitory mechanisms and examine the way in which they are integrated with facilitatory forms of behavioral plasticity.

Behavioral Dissociation of Dishabituation, Sensitization, and Inhibition in Adults

The developmental studies described above show that dishabituation, sensitization, and a novel inhibitory process, as well as their respective cellular analogues, can each be dissociated in juvenile animals. It is possible, however, that these processes, although separable during ontogeny, are not distinct in the final adult phenotype. Thus an important question arose as to whether the same forms of behavioral plasticity could be identified and separated in adult animals. Marcus et al. (1988) addressed this issue by examining, in adult Aplysia, the effects of a wide range of tail-shock intensities, at several times after tail shock, on both habituated and nonhabituated siphon withdrawal responses. They found that dishabituation and sensitization could be clearly dissociated in adult animals in two ways.

First was time of onset. When tested soon (ninety seconds) after tail shock, dishabituation was evident at a variety of stimulus intensities, whereas, in this early test, sensitization was not exhibited at any stimulus intensity. In fact, examining nondecremented responses revealed that tail shock produced inhibition of reflex amplitude. Although no sensitization was evident in the ninety-second test, in subsequent tests (twenty to thirty minutes after tail shock) significant sensitization was observed. Thus, dishabituation has an early onset (within ninety seconds), whereas sensitization has a very delayed onset (twenty to thirty minutes) after tail shock. Juvenile Aplysia also exhibit delayed-onset sensitization that emerges in early stage twelve, at least thirty days after the emergence of dishabituation.

Second was stimulus intensity. When a range of stimulus intensities to the tail was examined, maximal dishabituation was produced by relatively weak stimuli, whereas maximal sensitization was produced by stronger stimuli. Moreover, the stimulus intensity that was most effective in producing dishabituation produced no sensitization, and the intensity that was most effective in producing sensitization produced no significant dishabituation. Thus, as in juvenile Aplysia (Rankin and Carew, 1989), the processes of dishabituation, sensitization, and inhibition can be behaviorally dissociated in adult animals.

The behavioral observations described above raise important questions about the cellular processes underlying the dissociation of dishabituation and sensitization. One possibility is that these two forms of learning reflect different underlying cellular mechanisms. Alternatively, the same or related mechanisms may be involved in both forms of learning, and the dissociation observed could be due to differential interaction of the inhibitory process with dishabituation and sensitization. Behavioral results alone cannot distinguish between these possibilities. However, progress has been made in elucidating the cellular mechanisms underlying dishabituation and sensitization (Hochner et al., 1986) as well as inhibition (Bellardetti, Kandel, and Siegelbaum, 1987; Mackey et al., 1987; Wright and Carew, 1990; Wright, Marcus, and Carew, 1991). Thus it will be important to determine the degree to which these different cellular processes can account for the behavioral dissociations that are observed in both developing and adult Aplysia.

Conclusion

A developmental analysis in Aplysia has shown that different forms of learning, as well as their cellular analogues at central synapses, emerge according to very different developmental timetables. These studies have allowed the dissociation of four behavioral processes (see Figure 1): two decrementing (habituation and inhibition) and two facilitatory (dishabituation and sensitization). Whether these dissociations are produced by different facilitatory mechanisms, by differential interactions of inhibition with decremented and nondecremented responses, or by some combination of these alternatives, results suggest that a dual-process view of nonassociative learning, which postulates a single decremental and a single incremental process, requires revision, and that a multiple-process view, which includes the possibility of inhibitory as well as facilitatory interactions, is necessary to account adequately for the mechanisms underlying nonassociative learning.

The developmental studies discussed in this brief review have focused only on nonassociative learning in Aplysia, and only on short-term learning, which is retained for a relatively brief time (minutes to hours). However, Aplysia is also capable of exhibiting a variety of forms of associative learning. Moreover, in addition to short-term forms, both nonassociative and associative learning in Aplysia can exist in long-term forms lasting days to weeks. It will be of interest to examine the development of these additional processes in order to gain insights into theoretically important questions such as the relationships between nonassociative and associative learning and between short-term and long-term memory. As a step in this direction, Carew, Wright, and McCance (1989) have established that long-term memory for sensitization emerges in exactly the same stage as short-term memory, stage twelve. This observation lends support to the notion that short-and long-term memory may be mechanistically interrelated in Aplysia (Golet et al., 1986). By analyzing the development of these diverse processes at synaptic, biophysical, and molecular levels, it may be possible to gain unique insights into the substrates underlying learning and memory by examining their developmental assembly.

See also:APLYSIA: CLASSICAL CONDITIONING AND OPERANT CONDITIONING; APLYSIA: MOLECULAR BASIS OF LONG-TERM SENSITIZATION

Bibliography

Bellardetti, F., Kandel, E. R., and Siegelbaum, S. (1987). Neuronal inhibition by the peptide FMRFamide involves opening of S K⁺ channels. Nature 325, 153-156.

Carew, T. J., Castellucci, V. F., and Kandel, E. R. (1971). Analysis of dishabituation and sensitization of the gill withdrawal reflex in Aplysia. International Journal of Neuroscience 2, 79-98.

Carew, T. J., Wright, W. G., and McCance, E. F. (1989). Development of long-term memory in Aplysia : Long-term sensitization is present when short-term sensitization first emerges. Society for Neuroscience Abstracts 15, 1,285.

Fitzgerald, K., and Carew, T. J. (1991). Serotonin mimics tail shock in producing transient inhibition in the siphon withdrawal reflex of Aplysia. Journal of Neuroscience 11, 2,510-2,518.

Golet, P., Castellucci, V. F., Schacher, S., and Kandel, E. R. (1986). The long and short of long-term memory—a molecular framework. Nature 322, 419-422.

Groves, P. M., and Thompson, R. F. (1970). Habituation: A dual process theory. Psychological Review 77, 419-450.

Hochner, B., Klein, M., Schacher, S., and Kandel, E. R. (1986). Additional component in the cellular mechanism of presynaptic facilitation contributes to behavioral dishabituation in Aplysia. Proceedings of the National Academy of Sciences of the United States of America 83, 8,794-8,798.

Kriegstein, A. R. (1977). Stages in post-hatching development of Aplysia californica. Journal of Experimental Zoology 199, 275-288.

Krontiris-Litowitz, J. K., Erikson, M. T., and Walters, E. T. (1987). Central suppression of defensive reflexes in Aplysia by noxious stimulation and by factors released from body wall. Society for Neuroscience Abstracts 13, 815.

Mackey, S. L., Glanzman, D. L., Small, S. A., Dyke, A. M., Kandel, E. R., and Hawkins, R. D. (1987). Tail shock produces inhibition as well as sensitization of the siphon-withdrawal reflex of Aplysia : Possible behavioral role for presynaptic inhibition mediated by the peptide Phe-Met-Arg-Phe-NH². Proceedings of the National Academy of Sciences of the United States of America 84, 8,730-8,734.

Marcus, E. A., Nolen, T. G., Rankin, C. H., and Carew, T. J. (1988). Behavioral dissociation of dishabituation, sensitization, and inhibition in Aplysia. Science 241, 210-213.

Nolen, T. G., and Carew, T. J. (1988). A cellular analog of sensitization emerges at the same time in development as behavioral sensitization in Aplysia. Journal of Neuroscience 8, 212-222.

Rankin, C. H., and Carew, T. J. (1987). Development of learning and memory in Aplysia : II. Habituation and dishabituation. Journal of Neuroscience 7, 133-144.

—— (1988). Dishabituation and sensitization emerge as separate processes during development in Aplysia. Journal of Neuroscience 8, 197-211.

—— (1989). Developmental analysis in Aplysia reveals inhibitory as well as facilitatory effects of tail shock. Behavior and Neuroscience 103, 334-344.

Rayport, S. G., and Camardo, J. S. (1984). Differential emergence of cellular mechanisms mediating habituation and sensitization in the developing Aplysia nervous system. Journal of Neuroscience 4, 2,528-2,532.

Stopfer, M., and Carew, T. J. (1988). Development of sensitization of escape locomotion in Aplysia. Journal of Neuroscience 8, 223-230.

Wright, W. G., and Carew, T. J. (1990). Contributions of interneurons to tail-shock induced inhibition of the siphon withdrawal reflex in Aplysia. Society for Neuroscience Abstracts 16, 20.

Wright, W. G., Marcus, E. A., and Carew, T. J. (1991). A cellular analysis of inhibition in the siphon withdrawal reflex of Aplysia. Journal of Neuroscience 11, 2,498-2,509.

Thomas J.Carew