ACKNOWEDGEMENTS I would like to thank my advisor, Stuart Thompson, for his help and encouragement. I would also like to thank William Gilly for his patience, answers, and the use of much of his equipment. Thank you also to George Mackey for the methylene blue. Mark Denny deserves many thanks for the use of his tools and the Fisher bathroom key. Bruce Hopkins was very helpful in setting up the Aquatron that made the whole-animal preparation possible. My labmates, John, Shannon, and Simone should be canonized for putting up with me through many frustrating hours. Thanks to the friendly people at Hopkins in general, who were always willing to take time out and lend a hand. A sigh of thanks to my roommate and best friend, Terri, who bore the brunt of my weary, edgy demeanor. And finally, last but far from least, hearty thanks to my TAs, Sam Wang and Jennifer Levitt, each of whom rescued me one way or another throughout the quarter. ABSTRACT Some of the behaviors of the marine nudibranch Melibe leonina have been described, such as feeding, "galloping", the crumple response, and its fascinating swimming behavior (Hurst, 1968: Thompson). However, a new behavior, the shrug response, has been noticed and is described here. It was noted via visual observations that this behavior habituated quickly and was inhibited during swimming and locomotion. The purpose of this study was to observe and describe the properties of this behavior, trace it neuroanatomically, and find its relation to other motor behaviors. Results of behavioral manipulations and electrophysiological recordings showed that the shrug response is part of a behavioral hierarchy and, because it is mediated through one, central pathway, would be good for studies of habituation. INTRODUCTION Melibe leonina is a nudibranch mollusc belonging to the family Tethyidae. It is found along the west coast of the United States, from Alaska to Mexico (McDonald, et al, 1980). Melibe leonina is so named (leonina="lion-like") because of its large oral hood complete with two rows of oral tentacles. On the dorsal surface of the oral hood are two rhinophores, which are chemoreceptive organs (see Figures 1, 2). Melibe were collected by snorkelers and SCUBA divers off Del Monte Beach on Monterey Bay and were kept at Hopkins Marine Station in tanks of flowing seawater filled with kelp. An additional tank was kept at the Monterey Bay Aquarium in order to ensure that many of the Melibe would stay alive in case problems arose with the tanks at Hopkins. Melibe leonina is an ideal organism for behavioral and neurophysiological study for many reasons. For one, unlike other nudibranchs, Melibe has a wide variety of active motor behaviors. including feeding, copulation, locomotion, defense responses, and swimming. In addition, it has a nearly translucent body wall, which facilitates dissection. Melibe has a "brain" consisting of three paired ganglia that are centrally located, and its large, round neurons are amenable to electrophysiological recordings. Swimming behavior is controlled by a Central Pattern Generator (Thompson). The swimming consists of rhythmic side-to¬ side contractions (see Figure 3) that occur when the foot becomes dislodged from the substratum and ceases when the foot comes in contact with the substratum again. This paper describes a new response, referred to as the shrug response (described in detail below), which cannot be elicited during swimming. This inhibition during swimming raises some interest questions. Is there active inhibition of the response during swimming? Is this centrally mediated? Or is the shrug a local retlex? Is the shrug inhibited during swimming because there are muscles common to the two behaviors, or for some other reason? If some other reason, what could it be? The present studies addressed these and other questions about the nature of the shrug response in Melibe leonina, MATERIALS AND METHODS Behavioral Methods A precondition for accurate neuroethological studies is knowledge of the experimental animal's behavioral repertoire in its natural context (Ewert, 1980). Thus an aquarium was set up for behavioral observations. Three animals were placed in the tank simultaneously (animals were rotated) and their behaviors observed and manipulated. Feeding, copulation, locomotion, and swimming were observed naturally. The aquarium was supplied with flowing seawater (11-15 C), and the tank was guarded from harsh sunlight. In addition to observing behaviors in natural context, behaviors were also manipulated. A glass rod was used to test the crumple response (Thompson), a defensive response to tactile stimulation. Tracing the Neuroanatomy In an attempt to isolate the nerve responsible for mediating the shrug response, the nerve roots from the brain innervating the oral hood were mapped. This was accomplished by careful dissection and methylene blue staining. Since the nerves in Melibe appear appreciably more grainy than the muscles under the dissecting scope, it was possible to trace the nerve roots out into the hood via careful dissection. In order to reinforce the dissections, specimens were stained with methylene blue (dilution ratio = 3 drops/20mL dH2O) for approximately 50 hours. This helped make the nerve roots stand out clearly and was especially helpful in getting a depth perspective on the roots as they dove out of the brain region and into the surrounding tissue. Once the particular nerve responsible for mediating the shrug response was isolated by electrophysiological methods (see below), it was traced thoroughly into the hood (see Figure 4). Electrophysiology In order to correlate behavioral activity with neural activity, a whole-animal preparation was used, similar to that used by Willows with Tritonia (Willows, 1967). With the brain suspended from a platform, the animal was able to behave in the tank while recordings trom its nerve cells were obtained (see Figure 5). Using focal extracellular electrodes, recordings were made from the anterior nerve roots in an attempt to isolate the nerve responsible for mediating the shrug response. In order to accomplish this, a portion of the nerve was sucked up into the tip of the electrode, spontaneous activity was noted, and any changes in firing when the shrug was elicited were noted. Is the shrug centrally commanded? The fact that a behavior can be commanded centrally does not rule out the possibility that there is a peripheral pathway sufficient to maintain the response in the absence of central excitation. This is the case with the Aplysja gill withdrawal (Peretz, et al, 1976). In order to address this, the C2 nerve (see Figures 4, 6) was cut both unilaterally and bilaterally and the effects on behavior noted. A control, in which each of the other anterior nerve roots was cut, tested for effects of trauma. As mentioned, the shrug appears to be inhibited during swimming. Therefore the activity of C2 was recorded while the animal was swimming versus when its foot was attached to the kelp. It the motor commands for the shrug response are being sent to the periphery via C2, stimulation of C2 should elicit (at least some component of) the behavior. Electrical stimulation was applied to C2, first intact and then to the cut end. Stimulation frequency was 5.5 Hz, with a delay of 2msec, and a duration of 120 msec. Voltage was the parameter varied. Since C2 attaches to the dorsal surface of the cerebral ganglion, the search for the shrug motoneurons began there. The method employed was the extracellular hunting technique. Recordings were made from various areas in the paired cerebral ganglia, the origin/termination site of C2 nerve root. It was difficult to reliably identify neurons, since sizes, shapes, and coloration varied from brain to brain. However, the cerebral ganglion has a few (3-6) readily identifiable giant cells, which are thought to be important hood motoneurons (Hurst, 1968), and recordings were made from these cells and the cells in this area (see Figure 6). Recordings included neural firing during swimming, while the animal was on the kelp, and while the shrug was elicited. RESULTS Behavioral Results While mapping the receptive fields for the crumple response, the shrug response was first noticed. When the animal was touched on the dorsal surface of the oral hood between the rhinophores, the hood (and the hood ALONE) was retracted toward the body and flattened (see Figure 7). The fact that the shrug is localized to the hood distinguishes it from the crumple, which is a full-body response. The shrug habituates rapidly. It habituated more rapidly with a smaller interstimulus interval (see Figure 8), and a 30¬ second interval yielded full recovery of the response. There are two components to this habituation: (1) amount of contraction/withdrawal and (2) time to re-extend the hood. The latter was reduced more rapidly. The shrug response was not tested for the other remaining eight characteristics necessary for true habituation (Shabb, 1984; Peeke and Herz, 1973). Other features of this response include the fact that a crumple response can be elicited if the novel stimulus is severe, and that this response is inhibited during swimming and during locomotion. Previous authors have noted that the hood is in a "neutral", semi¬ contracted state during swimming (Hurst, 1968; Thompson). Perhaps this posture is not only characteristic but very important to swimming, such that the shrug is actively inhibited during swimming Neuroanatomical Results Four anterior nerve roots were clearly identifiable from specimen to specimen (see Figure 6). These have been labelled as nerve roots C1-C4 (Hurst, 1968), although 1 found that C1 actually attaches to the tentacular lobe, and C4 appears to come from the ventrolateral surface of the pleural ganglion. In tracing the nerves from the cerebral ganglion out into the oral hood, it was found that the C2 nerve branched twice, which is in agreement with Hurst's results (Hurst, 1967). The terminal points of the first branch were found to be located in the region between the rhinophores, while other terminal points were in the longitudinal muscles of the hood (see Figure 4). Electrophysiological Results From recordings made from the anterior nerve roots, C2 was the only one that showed bursts when the shrug was elicited. It appeared to be somehow involved in mediating the shrug response. while the other nerve roots did not. C2 was identified as the nerve which mediates the shrug response by ablation experiments. Unilateral ablation of C2 resulted in a loss of the shrug. It was clear that this loss was not simply due to sensory information being cut off, because: 1) the rhinophores still withdrew when a stimulus was applied, and 2) the crumple was clearly elicited when the strength of the stimulus was increased. The control experiment showed that the loss of the shrug was not due to trauma. When each of the other nerve roots was cut, the shrug remained. It was only when C2 was cut that the shrug was abolished. Bilateral ablation of C2 resulted in loss of the crumple in response to mechanical stimulation between the rhinophores, while it still occurred in response to mechanical stimulation of the body wall. It appeared, then, that while the sensory information carried in C2 unilaterally was enough to elicit the crumple, C2 was needed bilaterally in order for the shrug to occur. Recordings from intact C2 showed clear bursts (with a mean duration of 890 msec and 10.3 spikes; see Figure 9) when the shrug was elicited. It was clear that these were not just the sensory neurons from the dorsal surface of the oral hood, because there were no bursts when stimulation did not produce the shrug (e.g., during swimming). Recordings from the cerebral giant area revealed some cells that appeared to burst when the shrug was elicited. The cells in this area also fired when hood and tentacular movements were made, in response to touch or pinch of the oral tentacles, and during feeding¬ type movements. A rather spectacular finding was that all of the cells surveyed seemed to have decreased firing when the animal was swimming as compared to when it was on the kelp. DISCUSSION The fact that the shrug response is inhibited during swimming brings up two possibilities for further study. The first is the mechanism by which the output of a Central Pattern Generator inhibits other motor behaviors. Because the shrug is inhibited during swimming , and because all of the cells surveyed in the region of the giant cells in the cerebral ganglion showed decreased firing during swim (consistent with results of Thompson), I arrived at the hypothesis that some neurotransmitter is acting more like a hormone by being dumped locally into the ganglion instead of being active at distinct synapses. This type of mechanism has been worked out for suppression of behaviors in both Aplysia (Kandel, 1976) and Pleurobranchaea (Davis, et al, 1974). This general inhibition of hood motor activities may be important if the hood must remain in its "neutral", semi-contracted state (State 2 of Figure 9) for swimming to be efficient. Another possibility for further study is the investigation of the behavioral hierarchy of Melibe. Diagrams detailing the interrelationships of behaviors have been constructed (and reconstructed) for many mollusks (Willows, 1985; Davis, et al. 1974; Gillette and Davis, 1974). The present study indicates that swimming has a fairly central place in the priority sequence of Melibe (see Figure 10 for preliminary hierarchy). Swimming inhibits both the shrug response and feeding movements, and it occurs whenever the animal becomes dislodged from the substratum. A more detailed study of the behavioral hierarchy of Melibe would reveal how homologous it is to those of other molluscs and might give valuable insight into the larger issue of behavioral chojce. which is common to a wide range of organisms in the animal kingdom. Another property common to many organisms is learning. Studies of training, conditioning, and habituation have served as models of basic, or simple, learning. Many researchers (Pinsker, et al, 1969, Kupferman, et al, 1969, and Castellucci, et al, 1969) have used molluscan preparations for studies of habituation. Howeyer, it has been shown that such learning studies have involved neural pathways that are not as straightforward as was once thought. In the most common preparation, Aplysia californica, three different sites for habituation of the gill withdrawal reflex have been found. each of which is sufficient to mediate habituation (Peretz, et al. 1976). The present study confirms that the shrug response in Melibe leonina is not mediated via a local reflex circuit (i.e., peripheral pathway, as in Aplysia). Thus, this organism and behavior might prove useful for the study of simple learning mechanisms, such as habituation. Behavioral studies revealed that the shrug response habituates rapidly, while studies of habituation on the cellular level have yet to be done. LITERATURE CITED Davis, W. J., Mpitos, G. J., Siegler, M. V. S., Pinneo, J. M., and Davis, K. B. (1974). Neuronal Substrates of Behavioral Hierarchies and Associative Learning in Pleurobranchaea. Ameriçan Zoologist. 14: 1037-1050. Ewert, Jorg-Peter. Neuroethology. Berlin: Springer-Verlag, 1980. Gillette, R. and Davis, W. J. (1976). The Role of the Metacerebral Giant Neuron in the Feeding Behavior of Pleurobranchaea. Journal of Comparative Physiology. 116: 129-159. Hurst, Anne. (1968). The Feeding Mechanism and Behaviour of the Opisthobranch Melibe leonina. Symp. Zool Soc. Lond. 22: 151-166. Kandel, E. R. Cellular Basis of Behavion. San Francisco: W. H. Freeman and Company, 1976. Kandel, E. R. and Schwartz, J. H. Principles of Neural Science. New York: Elsevier Science Publishing Co., 1985. Kennedy, D. (1975). Comparative Strategies in the Investigation of Neural Networks. Journal of Experimental Zoology. 194: 35-50. Peeke, Harman V. S. and Herz, Michael J., ed. Habituation.. (Two volumes). New York: Academic Press, 1973. Peretz, Bertram, Jacklet, Jon W., and Lukowiak, Kenneth. (1976) Habituation of Reflexes in Aplysia: Contribution of the Peripheral and Central Nervous Systems. Science. 191: 396-399. Shabb, Sam Richard. Habituation of the Gill Withdrawal Response and Mapping of Associated Neurons in the Nudibranch, Doriopsilla albopunctata. (Master's thesis completed at Hopkins Marine Station). Thompson, S. H. Personal communication. Willows, A. O. D. (1967). Behavioral Acts Elicited by Stimulation of Single Identifiable Nerve Cells. Science, 157: 520-524. Willows, A.O.D., ed. The Mollusca (vol. 8): Neurobiology and Behavior. Part 1. Orlando, Florida: Academic Press, Inc., 1985 LIST OF FIGURES 1. Melibe leonina in its natural environment 2. Dorsal view of Melibe with basic anatomy labelled 3. Melibe swimming pattern 4. Sketch of C2 nerve out into the oral hood 5. Whole-animal preparation setup 6. Sketch of central ganglia with anterior nerve roots and cerebral giant cells labelled 7. Sketch of the shrug response 8. Behavioral habituation, comparing 5- and 15-second interstimulus intervals (ISl's, aka epochs) 9. Burst of action potentials in C2 nerve during shrug; N = 23, for means 10. Preliminary behavioral hierarchy ethogram Fiqure 1 ORAL HOo— — RHINOPHRE — ORAL TENTACLE CERATA DORSAL VIEN or Helibe S CENTRAL GANGUA - Fiqure 2 FiG. 3. Three successive stages in swimming, taken from a moving film of elibe. (reprinted from Hurst, 168) Fiqure 3 ANTERIOR 2 ....... POSTERIOR Fiqure 4 D minophore oral hood S platfern recordi eleetrede 90 OTRW INFL CEICTERED SEAMATEE AT 10-158c) Figure 5 Eve- PTATOCVST c3 Ci C2 ANTERIOR Cerebral gont cells CEResent gaNGUDN e Pedt GANGUDN Prekle GAUGUON POSTÉRIOR Fiqure 6 C4 +-- I 30 — 1. fiqure 7 101 9+ 8+ 6 5+ 41 3+ 2 1+ BEHAVIORAL HABITUATION = 15-second epochs - - = 5-second epochs 0 -0-0-0-0— — 1 2 3 4 5 6 7 8 9 10 Trials Fqure 8 + + ktatatatatattt Shrug Response C2nerveroot X duaton - 810 2120 wsec spikes - 10.3 * 1. 3 Fiqure 7 CRUMPL E SWIMMING SHRUG Figure 10 E —