Simple invertebrate organisms have provided a wealth of information concerning the molecular and cellular basis of learning and memory. Research conducted in one such simple system, the nematode Caenorhabditis elegans (C. elegans), has made several important contributions to the field.
Advantages of Caenorhabditis elegans as a Model System
C. elegans is a small (about one millimeter), hermaphroditic, soil-dwelling nematode (see Figure 1). In the laboratory, C. elegans spends its short life (about twenty days) swimming on agar-filled petri dishes laying eggs and feeding on E. coli bacteria. Its hermaphroditic mode of reproduction makes the maintenance of strains easy, thus allowing for large quantities of inexpensive, readily available animals for testing. Its nervous system is simple, consisting of 302 neurons with 5,000 electrical and 10,000 chemical synapses, all of which have been identified and mapped at the electron microscope level (White, Southgate, and Durbin, 1988; White, Southgate, Thomson, and Brenner, 1986). Complete developmental lineages of all cells are also known. Furthermore, the worm is transparent, so that with the appropriate microscopic power, its simple nervous system is easily visible. Single neurons can be easily ablated with a laser, while leaving the rest of the nervous system intact. Such an approach can elucidate the role of specific neurons in behavioral processes.
Consistent with the simplicity of its nervous system, the C. elegans genome is small and has been fully sequenced, with 8 × 107 nucleotide pairs arranged on six haploid chromosomes. Both classic genetic techniques and modern genetic engineering have produced a large number of mutant strains, which have provided the opportunity for investigating the role of single genes in behavioral processes. Worms are able to survive freezing, and thus mutant strains, once acquired, can be maintained indefinitely.
The behavioral repertoire of C. elegans is complex enough to offer a number of interesting behaviors to study. In the laboratory worms move forward along the surface of agar-filled petri dishes, using rhythmic, coordinated contractions of dorsal and ventral muscle groups, resulting in smooth sinusoidal waves of forward locomotion. Worms will respond to a variety of stimuli by changing direction and by swimming backward. Stimuli that produce reversals include touch, heat probes, some chemical compounds, and vibrations caused by the force of a mechanical tapper applied to the side of the dish. The response to a mechanical tap has been termed the tap withdrawal response (TWR) by Rankin, Beck, and Chiba (1990) and has proved to be important for studies of both short-and long-term memory. Using laser ablation techniques, Wicks and Rankin (1995) determined that the TWR consists of six sensory neurons, ten interneurons, and approximately sixty-nine motor neurons.
The simplest forms of learning are habituation, dishabituation, and sensitization. These are nonassociative forms of learning. Habituation is a decrease in the rate or amplitude or both, of responding due to repeated stimulus presentation (Groves and Thompson, 1970). Dishabituation is the facilitation of a decremented or habituated response following presentation of a novel, usually aversive, stimulus. Sensitization is the facilitation of a nondecremented response resulting from presentation of an aversive stimulus. The TWR of C. elegans shows each of these three nonassociative types of learning (Rankin, Beck, and Chiba, 1990).
The response habituation observed in the TWR with repeated presentation of the tap stimulus is not due to fatigue of the system, because presentation of a novel stimulus to a habituated animal immediately causes a return of the behavior to prehabituation levels—that is, dishabituation (Rankin, Beck, and Chiba, 1990). Consistent with the rules of habituation outlined by Groves and Thompson (1970), the rate of habituation of the TWR in C. elegans is sensitive to the frequency of stimulation. In wild-type worms habituation occurs more rapidly at short interstimulus intervals (ISIs) (i.e., intervals of two or ten seconds) compared to long ISIs (i.e., intervals of sixty seconds); subsequent recovery from habituation is also affected by ISI, with animals recovering more rapidly from habituation with short ISIs than with long ISIs (Rankin and Broster, 1992; see Figure 2). Thus short-term memory for habituation training lasts less than fifteen minutes when trained with short ISIs and can last for one to two hours when trained with long ISIs.
As C. elegans follows the generally agreed upon rules of habituation, it provides an effective model in which to study the role of genes underlying habituation.
[Image not available for copyright reasons]
There are mutant strains of worms that show differences in short-term habituation, one of which is eat-4. Studies of eat-4 and its mammalian homologues suggest that it regulates the amount of glutamate in neuron terminals (Bellocchio, Reimer, Fremeau, and Edwards, 2000; Lee et al., 1999). The eat-4 gene product is expressed on the sensory neurons of the TWR circuit (Lee et al., 1999). Rankin and Wicks (2000) showed that while responding normally to a single tap, and displaying the usual pattern of ISI-dependent habituation, eat-4 worms had more rapid overall rates of habituation with depressed asymptotic levels compared to wild-type worms. Furthermore, their recovery from habituation was slower compared to wild-type worms. Rankin and Wicks (2000) hypothesized that with repeated stimulation glutamate becomes rapidly depleted, resulting in more rapid response decrements in eat-4. However, the fact that the usual effects of ISI on the rate of habituation were preserved in eat-4 indicates that it is more than simple neurotransmitter depletion underlying habituation. The eat-4 strain did not show dishabituation. Since dishabituation does not occur in the eat-4 strain, dishabituation does require the presence of an intact eat-4 gene product. The cellular processes underlying habituation and dishabituation are likely not the same, and each involves eat-4 to a different extent.
Although for many years researchers believed that the response decrements seen in habituation were due solely to repeated stimulus presentation, it has now been demonstrated that organisms can make associations during habituation training that affect future performance. The most common of such associations is context conditioning, where some aspect of the training environment is encoded by the organism, which then influences future responses to the original training stimulus. Thus the long-held view that habituation is a purely nonassociative form of learning may not be warranted. Rankin (2000) showed that C. elegans is capable of context conditioning during habituation training. Worms were habituated to thirty-tap stimuli in the presence or absence of a distinctive chemosensory cue (sodium acetate). When tested an hour later, worms trained in the presence of sodium acetate showed greater retention of training when tested in the presence of sodium acetate than when tested on plain agar plates. Placing worms on sodium acetate plates for the hour between training and testing produced extinction of the context effect.
Classical conditioning has also been demonstrated in C. elegans using both appetitive and aversive conditioning paradigms. Conditioned worms showed clear postconditioning preferences for distinctive tastes that were paired with food during training (Wen et al., 1997). Similarly, in an aversive-conditioning paradigm, worms learned to avoid tastes that were paired with an aversive stimulus during training. Similar results have been obtained using olfactory rather than taste cues, where worms learned to avoid a previously attractive odor after it was paired with an aversive acetic acid solution (Morrison, Wen, Runciman, and van der Kooy, 1999). Two mutant strains of worms, lrn-1 and lrn-2, were generated that were not able to form taste or olfactory associations.
C. elegans is also capable of retention of habituation training for at least twenty-four hours. This phenomenon, long-term habituation, has been used as an effective model for the study of long-term memory (defined as retention for twenty-four hours or more). Distributed or spaced habituation training of the TWR (three to four blocks of twenty stimuli each at a sixty-second ISI, separated by one-hour periods) resulted in retention of the habituated response twenty-four hours later when tested with a series of twenty taps (Beck and Rankin, 1997; Rose, Chen, Kaun, and Rankin, 2001). This effect was observed only when a sixty-second ISI was used during training (i.e., it was not seen with a ten-second ISI) and when a distributed training procedure was used. Consistent with what has been observed in other species, such as Aplysia and Drosophila, massed training (one block of sixty taps), with either a ten-second or a sixty-second ISI, did not produce long-term habituation in C. elegans. In addition, Beck and Rankin (1995) showed that long-term habituation was protein-synthesis dependent. When worms were exposed to as few as fifteen minutes of heat shock, which disrupts protein synthesis, during the early part of the rest period in a distributed-training paradigm, long-term memory was significantly reduced, suggesting that the time immediately after training is most vulnerable to heat shock and thus may be a critical phase in the formation of long-term memory. Heat shock applied during the one-hour rest period did not, however, have any effect on the accumulation of short-term habituation seen over successive blocks of training, suggesting that short-and long-term memory recruit different cellular and molecular processes.
As with short-term memory, C. elegans also provides an effective model with which to study the role of genes underlying long-term memory for habituation training. For example, glr-1, a gene coding for a homologue of mammalian kainate/AMPA-type glutamate receptor, has been identified and cloned in C. elegans. Worms missing functional glr-1 showed no long-term retention of habituation training twenty-four hours later (Rose, Chen, Kaun, and Rankin, 2001). These results suggest that stimulation of kainate/AMPA-type glutamate receptors on the interneurons is required for long-term habituation to tap. The same mutation has also been shown to prevent olfactory associative learning (Morrison and van der Kooy, 2001).
C. elegans has proved to be a useful model system for the study of learning and memory processes. Unlike other animals used in this area of research, C. elegans is an inexpensive, easily obtained and maintained organism, with a large amount of information available about its nervous system and genome. Knowledge of the neural circuitry underlying behavior combined with knowledge of the genome has allowed for the investigation of genetic factors involved in both short-term and long-term memory. That is, if a certain gene is expressed on a component of a neural circuit known to underlie a behavior, then that gene is a prime candidate for playing a role in either the behavior itself or its plasticity. This kind of an approach has proved fruitful in C. elegans by elucidating the importance in learning and memory of glutamate, a neurotransmitter that has also been determined to play a major role in mammalian learning and memory processes. The fact that the same results are obtained in C. elegans attests to its validity as a model system. The rules of habituation outlined by Groves and Thompson (1970) apply to C. elegans, making it an appropriate model system for the study of short-term memory as well. Research using C. elegans has made it clear that habituation is not a unitary process but is rather a set of processes, differentially recruited by short and long ISIs that cannot be neatly tucked away into the category of "simple," nonassociative learning.
Beck, C. D. O., and Rankin, C. H. (1992). Caenorhabditis elegans : A simple systems approach to the genetics of behavior. In D. Goldowitz, D. Wahlstein, and R. E. Wimer, eds., Techniques for the genetic analysis of brain and behavior: Focus on the mouse, 445-463. Amsterdam: Elsevier.
—— (1995). Heat-shock disrupts long-term memory consolidation in Caenorhabditis elegans. Learning and Memory 2, 161-177.
—— (1997). Long-term habituation is produced by distributed training at long ISIs and not by massed training or short ISIs in Caenorhabditis elegans. Animal Learning and Behavior 25 (4), 446-457.
Bellocchio, E. E., Reimer, R. J., Fremeau, R. T., and Edwards, R. H. (2000). Uptake of glutamate into synaptic vesicles by an inorganic phosphate transporter. Science 289, 957-960.
Groves, P. M., and Thompson, R. F. (1970). Habituation: A dual-process theory. Psychological Reviews 77, 419-450.
Lee, R. Y. N., Sawin, E. R., Chalfie, M., Horvitz, H. R., and Avery, L. (1999). Eat-4, a homolog of a mammalian sodium-dependent inorganic phosphate co-transporter, is necessary for glutamatergic neurotransmission in Caenorhabditis elegans. Journal of Neuroscience 19, 159-167.
Morrison, G. E., and van der Kooy, D. (2001). A mutation in the AMPA-type glutamate receptor, glr-1, blocks olfactory associative and nonassociative learning in Caenorhabditis elegans. Behavioral Neuroscience 115 (3), 640-649.
Morrison, G. E., Wen, J. Y. M., Runciman, S., and van der Kooy, D. (1999). Olfactory associative learning in Caenorhabditis elegans is impaired in lrn-1 and lrn-2 mutants. Behavioral Neuroscience 113 (2), 358-367.
Rankin, C. H. (2000). Context conditioning in habituation in the nematode Caenorhabditis elegans. Behavioral Neuroscience 114, 496-505.
Rankin, C. H., Beck, C. D. O., and Chiba, C. M. (1990). Caenorhabditis elegans : A new model system for the study of learning and memory. Behavioral Brain Research 37, 89-92.
Rankin, C. H., and Broster, B. S. (1992). Factors affecting habituation in the nematode Caenorhabditis elegans. Behavioral Neuroscience 106, 239-242.
Rankin, C. H., and Wicks, S. R. (2000). Mutations of the Caenorhabditis elegans brain-specific inorganic phosphate transporter eat-4 affect habituation of the tap-withdrawal response without affecting the response itself. Journal of Neuroscience 20, 4,337-4,344.
Rose, J., Chen, S., Kaun, K., and Rankin, C. H. (2001). Glutamate and AMPA-receptor function are necessary for long-term memory in C. elegans. Society for Neuroscience Abstracts 27, 45.
Wen, J. Y. M., Kumar, N., Morrison, G., Rambaldini, G., Runciman, S., Rousseau, J., and van der Kooy, D. (1997). Mutations that prevent associative learning in C. elegans. Behavioral Neuroscience 111 (2), 354-368.
White, J. E., Southgate, E., and Durbin, R. (1988). Appendix 2: Neuroanatomy. In W. B. Wood, ed., The nematode Caenorhabditis elegans, pp. 433-455. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
White, J. E., Southgate, E., Thompson, J. N., and Brenner, S. (1986). The structure of the nervous system of the nematode Caenorhabditis elegans. Philosophical Transactions of the Royal Society of London. Series B: Biological Sciences 314, 1-340.
Wicks, S. R., and Rankin, C. H. (1995). Integration of mechanosensory stimuli in Caenorhabditis elegans. Journal of Neuroscience 15, 2,434-2,444.