Abstract
The nudibranch Melibe leonina swims by rhythmically flexing its body from side to side at a frequency of 1 cycle every 2–5 sec. Melibe swim spontaneously, when they are dislodged from the substrate, or when they come in contact with predatory seastars, such as Pycnopodia helianthoides. Intracellular recordings obtained from semi-intact swimming Melibe reveal a population of ∼15 swim motoneurons (SMNs) in each pedal ganglion. In general, SMNs in one pedal ganglion fire out-of-phase with SMNs in the opposite pedal ganglion, resulting in rhythmic side-to-side bending movements. In isolated brains, recordings from SMNs yield similar results, indicating the existence of a swim central pattern generator (CPG). There is no evidence for synaptic interactions between SMNs and either inhibiting or exciting SMNs has no impact on the swim pattern. The SMNs are driven by a CPG consisting of 4 interneurons; 2 in the cerebropleural ganglia and 1 in each pedal ganglion. Appropriate bursting activity in the swim interneurons is necessary for swimming to occur. Either hyperpolarization or depolarization of any of the 4 CPG interneurons disrupts the normal swim pattern. Swimming behavior, and the fictive swim motor program expressed by the isolated brain, are inhibited by light and nitric oxide donors. NADPH-diaphorase staining and nitric oxide synthase (NOS) immunocytochemistry of Melibe brains suggests the source of nitric oxide might be a pair of bilaterally symmetrical cells located in the cerebropleural ganglia.
INTRODUCTION
Crawling is the most common form of locomotion in opisthobranch molluscs, but at least 47 different species also swim, using five general kinds of swimming movements (Farmer, 1970). Of these five types of swimming, three are by far the most common: parapodial flapping, dorsal-ventral undulation, and lateral-bending.
The neural basis of opisthobranch swimming behavior has been investigated in several different opisthobranchs for decades because they have very large, identifiable, neurons that can be studied in both semi-intact, moving preparations and inisolated brains. Of the four main species that have been examined, two swim by flapping their “wings” or parapodia (Clione limacina and Aplysia brasiliana) and two use dorsal-ventral flexions (Tritonia diomedea and Pleurobranchaea californica). Recently, we have been investigating the neuroethology of swimming in Melibe leonina, which swims using rhythmic lateral-bending movements. This manuscript is meant to provide a short review of previous studies concerning this subject and a summary of the results of ongoing investigations in our laboratory.
Although the neural basis of swimming in Melibe has received little attention until recently, a number of short descriptions of the behavior have appeared in the literature, going back to 1919 (Agersborg, 1919, 1921). These early papers provided descriptions of Melibe swimming, crawling, and feeding, as well as discussions of the adaptive significance of different forms of Melibe locomotion. In 1968 Ann Hurst published an excellent paper about various Melibe behaviors and she provided a good map of the brain along with accounts of some recordings obtained from a variety of neurons in the CNS. Recently, members of our laboratory have been carrying out further investigations of Melibe neuroethology, beginning with feeding behavior (Trimarchi and Watson, 1992; Watson and Trimarchi, 1992; Watson and Chester, 1993) and, more recently, extending to swimming behavior (Lawrence, 1997; Watson, 1997; Newcomb and Watson, 2001). This new information makes it possible to compare and contrast the neural basis of swimming in a lateral-bending species with opisthobranchs that swim using parapodial flapping and dorsal-ventral flexions.
One of the major motivations for investigating the swimming behavior of opisthobranchs is that once we understand the neural circuits responsible for producing the behavior, we can then investigate how these circuits are modified by various stimuli or past experiences. We can also attempt to determine how these swim circuits overlap or interact with circuits for other behaviors, such as feeding and mating. In this Symposium there are a number of good examples of this type of analysis, ranging from the influence of serotonin on swimming in Tritonia and Clione, to the overlap between feeding and swimming circuits in Pleurobranchaea. Currently, we are studying how light and nitric oxide influence Melibe swimming and the underlying neural networks. In this short review we will provide some of this new evidence and discuss how the two processes might be related.
MATERIALS AND METHODS
Animals
Melibe were collected in the Puget Sound, WA, near the University of Washington's Friday Harbor Laboratories (FHL), on San Juan Island. Some experiments were carried out at FHL, while others were performed in the Zoology Department facilities at the University of New Hampshire.
Behavior
Long-term patterns of Melibe swimming activity were determined by analysis of time-lapse videotapes. Experiments took place in a 20-liter aquarium continuously perfused with natural seawater, at 8–12°C. Individual Melibe (n = 12) were exposed to a natural light regime of 14L:10D, as well as constant darkness and constant light conditions. During constant darkness the tank was illuminated with a 40-watt ceramic coated red light and constant light was provided by a 30-watt light.
Experiments ran for 40–72 hr and were monitored by a low-light video camera connected to a Panasonic AG-RT600P time-lapse VCR. Videotapes were analyzed to determine the number and duration of spontaneous bouts of swimming.
Semi-intact and isolated brain preparations
Experiments with semi-intact preparations were carried out using the same chamber and methods employed by Willows to study swimming in Tritonia (Willows, 1967; Willows et al., 1973). Each Melibe was suspended by hooks from its dorsal integument in a chamber continuously perfused with natural seawater. A small opening was made in the integument just over the brain, and the CNS was immobilized by pinning it to a small wax-covered platform. Typically, this procedure stabilized the brain enough so that intracellular recordings could be obtained for up to an hour while the Melibe exhibited intermittent bouts of swimming.
In other experiments brains were removed from animals by cutting all nerve roots, except for the pedal-pedal connectives. Isolated brains were pinned in a 2-ml recording dish embedded in an aluminum plate. Coolant or ambient seawater was circulated through the aluminum plate to keep the recording chamber at 8–12°C.
Intracellular recording
Intracellular recordings were obtained using 20–40 megΩ microelectrodes filled with 2 M potassium acetate. Certain neurons were filled with lucifer yellow (5% dissolved in 0.15 M LiCl2) by passing hyperpolarizing current pulses (5–10 nA, 50% duty cycle). After neurons were filled the ganglia were fixed overnight in 4% paraformaldehyde in filtered seawater, and then dehydrated, cleared and mounted in DPX. Permanent images were obtained with a Nikon Optiphot-2 epifluorescence microscope or a Bio-Rad MRC-600 confocal microscope.
Nitric oxide synthase histology
Putative nitric oxide synthase (NOS) containing neurons were localized using both immunocytochemical and NADPH-diaphorase histochemical staining methods. The immunocytochemical methods have been described in detail in Watson and Willows (1992). Isolated brains were fixed, rinsed in 0.1 M phosphate buffer (PB) and then treated with 0.1% trypsin for 15 min and 4% Triton X-100 in PB for 1 hr. They were then rinsed in PB containing 0.4% Triton X-100 and 0.1% sodium azide (PTA) for 24 hr and incubated overnight in 6% goat serum in PTA (PTA-GS). This was followed by incubation in primary antibodies (anti-universal NOS, Affinity Bioreagents, Golden, CO., 1:100 in PTA-GS) for 48 hr, a 24 hr rinse in PTA, incubation in goat anti-rabbit secondary antibodies conjugated to fluorescein (1:100 in PTA-GS, 24 hr), and then a final rinse in PB. Brains were then dehydrated, cleared in methyl salicylate, mounted in DPX and viewed as described above in the lucifer methods.
For NADPH-diaphorase staining brains were removed from animals and fixed in 4% paraformaldehyde in 0.1 M PB for 15 min at 4°C. They were then washed 3 times in 5-min changes of 0.5 M TrisHCl (pH = 8.0). Staining was carried out in the dark, overnight, at 4°C in the following solution: 2 mM NADPH, 0.5 mM Nitro Blue Tetrazolium, 0.01 M dicumarol and 0.25% Triton X-100 in 0.5 M TrisHCl (pH = 8.0). The following day, brains were washed twice, for 5 min each, in 0.5 M TrisHCl and then post-fixed for 1 hr in 4% paraformaldehyde in methanol. They were then dehydrated in an alcohol series to 100% EtOH, cleared in methyl salicylate, and mounted in Permount for viewing and photography.
RESULTS
Swimming behavior
Melibe leonina are normally found in eelgrass or kelp beds. They use their foot to hold onto blades of eelgrass or kelp and they extend their oral hood to capture food in the water column (Agersborg, 1921; Hurst, 1968; Watson and Trimachi, 1992; Watson and Chester, 1993). They are rarely observed swimming spontaneously during the daytime. If they are dislodged, they will swim until they come in contact with another plant, or in the laboratory, the side of the aquarium. When they stop they usually attach first with the anterior portion of their foot. Typically, a bout of swimming lasts for 30–90 sec. Video clips of actual swimming can be viewed at the following website: http://zoology.unh.edu/faculty/win/Melibe/melibeswimming.htm.
Melibe swim spontaneously and in response to a number of stimuli (Lawrence, 1997). The most effective natural stimulus is contact with the tube feet of a predatory seastar, such as Pycnopodia helianthoides. In fact, a brief (<1 sec) touch with a single tube foot is usually sufficient to elicit a bout of swimming. Pinching with forceps will also cause animals to swim, but it is much less effective. Finally, application of 0.5 ml of 1 M KCl is a very effective, although unnatural, stimulus.
Melibe swimming has previously been described by several authors, and a more detailed account will appear in a manuscript that was recently submitted (Lawrence and Watson). After closing their oral veil, and releasing from the substrate, they curl their foot, extend their cerrata and flatten themselves in the lateral direction to create a larger surface for lateral swimming. Then they rhythmically bend their entire body from side-to-side, forming a shape like the letter “C” with each lateral flexion. They also curl their foot a bit with each flexion, so that it pushes water behind them a bit, rather that just to the side. This sculling motion tends to move them ventrally, in the direction of their foot. They usually have their head toward the surface when they swim, so moving in the direction of their foot tends to carry them horizontally (Lawrence, 1997; see videos at previously cited website).
Melibe swim fairly slowly, with each complete cycle lasting 2–5 sec. A typical swim episode lasts for 30–90 sec, but they can last for up to 20 min. During these prolonged swimming bouts they often pause and float near the surface, and then resume swimming. In fact, Melibe are often found floating on the surface of water in the Puget Sound from November–May (Mills, 1994; personal observation).
Influence of light on Melibe swimming
Long-term time-lapse video analyses reveal that Melibe often swim spontaneously and that the probability of swimming is higher at night than during the day (Fig. 1). At night, they swim >20 time/hr, while during the day they rarely swim (<1/hr). This pattern appears to be due to the combined influence of light and an endogenous clock on the swim central pattern generator (Watson, Newcomb and Lawrence, in preparation).
Intracellular recordings from semi-intact, swimming Melibe
In order to identify neurons that might be involved in the swimming behavior of Melibe, intracellular recordings were obtained from putative swim motoneurons in partially restrained, swimming animals. These cells are located in both pedal ganglia (Fig. 2) and their rhythmic activity correlates very well with lateral swimming movements. Swim motoneurons (SMNs) in the right pedal ganglion fire just prior to, and during, flexions to the right, and visa versa (Fig. 3). When animals stop swimming, SMN bursting stops and they either become quiet, or fire tonically, depending on the position of the animal. When animals start swimming again, rhythmic activity in the SMNs resumes. Finally, when animals are not swimming, stimulation of many of these SMNs results in contraction of the body wall in the ipsilateral direction.
Four interneurons that have a very strong influence on swimming behavior have been identified in the brains of semi-intact animals. One symmetrical pair (Swim Interneuron I, SiI) is located in the cerebropleural ganglia and the other pair (Swim Interneuron II, SiII), have cell bodies in the middle of the dorsal surface of each pedal ganglion (Fig. 2). Hyperpolarization of any one of these interneurons, sufficient to inhibit bursting, also causes animals to stop swimming. Thus, these neurons appear to be an integral part of the swim central pattern generator (CPG).
Expression of the swim motor program by isolated brains
The swim motor program is spontaneously expressed by isolated Melibe brains, indicating that swimming is a fixed action pattern produced by a CPG in the brain that does not require sensory input to create the rhythm. We have recorded from SMNs, SiIs, and SiIIs in isolated brains and their firing patterns are identical to those obtained in intact animals (Fig. 4).
The swimming rhythm in Melibe appears to be produced by 4 interneurons. Each side of the brain has two interneurons (Fig. 2) that are electrically coupled to each other, fire in phase with each other (Fig. 4) and reciprocally inhibit the pair on the other side of the brain. They also inhibit the SMNs on the contralateral side of the brain and excite the SMNs on the same side of the brain. A pool of ∼15 SMNs is located on the dorsal surface of each pedal ganglion (Fig. 2) and they tend to burst in phase with each other because they receive common synaptic input. In intact animals, firing of the SMNs on a given side of the brain causes the animal to flex in that direction (Fig. 3). The SMNs are not involved in generating the swim motor program and either hyperpolarizing or depolarizing the SMNs has no influence on expression of the swim pattern in the isolated brain.
Influence of light on expression of the swim motor program
Although the isolated brain spontaneously expresses a normal swim motor program, light reduces the probability that the swim pattern will be produced (Fig. 5). In most cases, shining a bright light on the brain will completely inhibit swimming. After 30 min in complete darkness even very low light levels are sufficient to inhibit swimming activity. When the lights are turned off, after a period of illumination, cells in the swim circuit depolarize and quickly begin to express the swim motor program (Fig. 5). The effects of light are apparently mediated by the eyes, which are attached directly to the brain via the optic lobe (Fig. 2). If the eyes are removed, the brain no longer responds to light.
Nitric oxide in Melibe
Staining of the Melibe brain with the NADPH-diaphorase histochemical technique reveals a pair of symmetrical neurons on either side of the cerebropleural ganglion (Fig. 6; Newcomb and Watson, 2001). The putative nitrergic neurons project past the eye to the pedal ganglia. Control experiments, in which NADPH is replaced with NAD+, yield no staining in the brain. Immunocytochemical staining with antibodies directed against a conserved region of NOS also reveals the same pair of neurons, confirming the results obtained with the diaphorase method (Fig. 6).
In addition to the pair of NOS-positive neurons in the brain, the diaphorase method stains processes in the tentacles, neurons in small ganglia at the base of the tentacles, and the neuropil in the cerebral ganglia. It appears as if all of the staining in the cerebral ganglia is due to processes emanating from cells in the ganglia at the base of the tentacles. But the function of these neurons is unknown.
Application of the NO-donors sodium nitroprusside (SNP, 1.0 mM) and S-nitroso-N-acetyl-penicillamine (SNAP, 1.0 mM) to isolated brains that are expressing the swim motor program, causes a dramatic slowing of the swim rhythm (Fig. 7). In isolated brains, the rate of bursting of swim interneurons and motoneurons goes from 1 cycle every 2–5 sec to 1 cycle every 20–30 sec. Preliminary experiments with semi-intact Melibe indicate that NO-donors slow down the rate of swimming, rather than activating a different type of behavior, such as feeding or crawling. The role of NO in Melibe and the function of the nitrergic neurons in the brain and tentacles are currently under investigation.
DISCUSSION
Melibe swimming behavior provides an excellent model system for investigating the production, control and modulation of a rhythmic behavior. Melibe swim in response to a distinct natural stimulus (seastar tube foot); it is possible to record from both swim motoneurons and interneurons in semi-intact, swimming animals; the isolated brain spontaneously expresses the swim motor program; the CPG circuit is relatively simple and accessible for manipulation; and the behavior is modulated by a variety of different inputs. In the coming years, along with other comparative studies of opisthobranch swimming, this system is likely to contribute a great deal to our understanding of the general principles of motor control and modulation.
Possible reasons for swimming in opisthobranchs
It is generally accepted that a number of opisthobranchs swim to avoid predators, such as the seastar Pycnopodia helianthoides, although this is rarely observed in their natural habitat. In fact, Ajeska and Nybakken (1976) and Bickell-Page (1991) have reported that Pycnopodia and 3 other seastar species all exhibited aversive responses to contact with Melibe, presumably in response to secretions from the repugnatorial glands (Ayer and Andersen, 1983; Bickell-Page, 1991). The only seastar that has been observed eating a Melibe is Crossaster paposus (Mauzey et al., 1968; Bickell-Page, 1991). Therefore, it seems odd that Melibe so readily swim in response to seastar tube feet, given the low probability of actually being attacked by seastars. In contrast, Ajeska and Nybakken (1976) and Mauzey et al. (1968) have observed Pugettia producta attacking Melibe in the field and Bickell-Page (1991) has reported that Pugettia and two other crab species capture and eat Melibe in captivity. Melibe will swim in response to pinches and when their cerata are pinched they autotomize them (Bickell-Page, 1989). This combination of responses may help them escape from predatory crabs that are not deterred by chemical defenses.
Melibe also swims spontaneously, as do Clione and Aplysia brasiliana. Clione swims almost continuously in order to remain at a certain depth in the water column (Satterlie et al., 1985), and swimming is intimately involved in their feeding behavior (Norekian, 1995). Aplysia brasiliana is capable of directed movement (Hamilton and Ambrose, 1975), and Melibe swimming also appears to be somewhat directional (Lawrence, 1997), although it is not clear how and why they use this directed movement. Our current hypothesis is that they swim spontaneously in order to move from one blade of eelgrass, or kelp, to another; either to find more food or to find a potential mate.
Whether opisthobranchs swim spontaneously, or not, is a function of the underlying neural circuits that control swimming behavior. Tritonia and Pleurobranchaea require stimulation to elicit swimming, both in intact animals and in isolated brains. While intact A. brasiliana swim spontaneously, the isolated CNS requires stimulation to elicit the swim motor program (McPherson and Blankenship, 1991; Gamkreledze et al., 1995). In contrast, the isolated CNS of Clione and Melibe continuously expresses the swim motor program. The Clione system is clearly designed for almost continuous swimming due to its pelagic mode of existence. On the other hand, while the Melibe CPG spontaneously expresses the swimming motor program, it is normally under almost continuous inhibition due to input from the foot when it is in contact with a substrate. When that inhibition is removed, in the intact animal, or in the isolated brain, swimming is expressed. Whether the swim is elicited or spontaneous it only lasts 30–90 sec, so the mechanisms that dictate the duration of a swimming episode might be similar in Melibe, Tritonia, and Pleurobranchaea.
Another factor that strongly influences Melibe swimming is light. This is a topic that has received little attention in the other swimming molluscs, except for Aplysia brasiliana (Hamilton and Russell, 1982). While a great deal is known about the circadian rhythm of neuronal discharge emanating from the eyes of Aplysia californica (Jacklet, 1969) and Bulla gouldiana (Block and Wallace, 1982), this issue has not been thoroughly studied in the swimming molluscs. Our studies indicate that light probably provides one, of many, inputs modulating the swim CPG. The eyes of Melibe are directly attached to the brain via the optic lobe, and because the skin of Melibe is fairly transparent, light easily impinges upon the eyes. The primary effect of light on swimming is inhibitory, both in the intact animal (Fig. 1) and in the isolated brain (Fig. 5). This reduces the probability that Melibe will swim during the daytime, perhaps because this is when they are most susceptible to visual predators. At the neural level, when the lights are turned off, the swim CPG interneurons depolarize and when lights are turned on, they return to a more hyperpolarized level. In some animals this hyperpolarization is sufficient to completely inhibit swimming, while in others it merely reduces the intensity of bursting. Clearly, light is not the only factor dictating when animals will swim, but it does affect the probability that animals will swim, both spontaneously and in response to predators. There is also a weak endogenous rhythm of locomotion in Melibe that might influence the expression of swimming, and the presence of food inhibits both crawling and swimming (Watson and Newcomb, unpublished observations). Further studies with this preparation are likely to reveal how the combination of an animal's behavioral state, and sensory inputs, controls the expression of swimming.
Comparisons between neural circuits underlying lateral-bending swimming vs. other types of opisthobranch swimming
It would appear, based on our current knowledge of the Melibe swim circuit, that it is relatively simple in comparison with the other known opisthobranch swim circuits. The CPG consists of only 4 interneurons, which, due to electrical coupling, function as a pair that reciprocally inhibit each other. In this respect, the Melibe circuit is somewhat similar to the Clione circuit. Both Tritona and Pleurobranchaea, which swim with dorsal-ventral flexions, have a much more intricate CPG, involving more neurons and more complex synaptic interactions (Getting, 1980; Getting et al., 1980; Lennard et al., 1980; Jing and Gillette, 1995, 1999; Katz, et al., 2001; Frost et al., 2001, Gillette and Jing, 2001). In Clione, an almost continuous parapodial swimmer, the basic CPG is fairly simple (Satterlie, 1985; Arshavsky et al., 1985). But because the circuit is also involved in other types of behaviors, and can change speeds, the overall swim circuit is more complex (Satterlie, 1991; Norekian, 1995; Satterlie and Norekian, 2001; Norekian and Satterlie, 2001).
Melibe leonina is the only lateral-bending swimming mollusc that has been investigated at the neural level. Given our current state of understanding, it appears as if the neural circuit for swimming in Melibe has some similarities, in terms of the overall organization, to the circuit used by the dorsal-ventral swimmers, such as Tritonia and Pleurobranchaea. First, the motoneurons play little, if any, role in producing the swim pattern. Second, most of the motoneurons are located in the pedal ganglia, although they are segregated in a different manner for side-to-side vs. dorsal-ventral locomotion. Third, reciprocal inhibition is a critical element involved in generating the alternating pattern of movement. Finally, some of the key interneurons are located in approximately the same location in all three species (see Katz et al. [2001] for more on this issue). However, a very key interneuron in Melibe resides in the pedal ganglion, as is the case in both of the parapodial flappers, Aplysia and Clione (Satterlie, 1985; Gamkrelidze et al., 1995), but neither of the dorsal-ventral swimmers. In the future, knowledge about the neural basis of swimming in some additional lateral-bending species, such as Dendronotus festivus, would be very helpful in determining the relative importance of mode of swimming vs. phylogenetic relationships in shaping the neural circuits underlying swimming.
The putative role of nitric oxide in Melibe swimming
Nitric oxide synthase has been identified in all of the species discussed in this Symposium (Moroz and Gillette, 1996; Moroz et al., 1996; Satterlie, personal communication), except Aplysia brasiliana (although it is present in Aplysia californica [Jacklet and Gruhn, 1994]). However, in contrast with serotonin (see Norekian and Satterlie, 2001; Katz et al., 2001), the role of NO in opisthobranch swimming is poorly understood. In Melibe, diaphorase staining reveals a pair of neurons in the cerebropleural ganglion, as well as dense staining of the cerebral neuropil and neurons in the tentacles and associated ganglia. Immunocytochemical staining confirms the presence of NOS in the cerebropleural neurons. Application of NO-donors slows the swim pattern from a rate of 1 cycle/2–5 sec, to 1 cycle/20–30 sec. These data, taken together with the known influence of NO on feeding in several molluscs (Moroz et al., 1993; Elphick et al., 1995; Teyke 1996), suggests a role of NO in feeding in Melibe. However, our results in studies with intact animals indicates that the slow pattern of activity induced by NO in pedal neurons is simply a very slow expression of the swim CPG. In intact animals, which are usually not swimming, this pharmacological effect of NO probably manifests itself as a reduction in the probability that animals will swim; which is also how both light and food influence swimming. Thus, our current hypothesis is that NO is involved in mediating either the inhibitory effects of light, or feeding stimuli, on the swim CPG.
Fig. 1. Influence of light on Melibe swimming. A. Nocturnal pattern of swimming in a natural light : dark cycle. Each data point represents the average number of swimming episodes expressed by 12 different Melibe. Error bars are not included because, although the pattern was similar, the hour by hour variability was large. B. When data from the same experiment are averaged, there is a statistically significant difference between the number of swim episodes/hour observed in the night and during the day (paired T-test)
Fig. 2. Anatomy of Melibe brain showing approximate locations of swim circuit neurons. The two fused ganglia that constitute most of the brain are the cerebropleural ganglia and they are connected by a large commissure. Near that commissure are the paired tentacular lobes. The left and right pedal ganglia are lateral to the cerebropleural ganglia. Normally the pedal-pedal connectives (PC) connect both pedal ganglia. In this figure the black dots in the left pedal ganglion indicate the approximate locations of some, but not all, of the SMNs. On the right side of the brain are illustrations showing the positions and projections of swim interneurons I and II (SiI, SiII). SiI projects to the ipsilateral pedal ganglion and SiII projects to the contralateral pedal ganglion via the PC. Note that the eye is located in very close association with the brain so that in isolated brain preparations the eyes continue to function. Abbreviations: PD=pedal, P=pleural, C=cerebral, S=statocyst
Fig. 3. Simultaneous recordings of Melibe movements and SMN activity. The top trace shows the output of a Photonic optical movement detector that provided a minimum voltage when animals were flexed to the right. Note that each time the right SMN fired, the animal slowly moved to the right. A video, showing this experiment, can be observed at the following website: http://zoology.unh.edu/faculty/win/Melibe/neuralswim.htm
Fig. 4. Spontaneous expression of the swim motor program in isolated brains. A. Intracellular recordings from left and right SMNs. Notice how they are out of phase with each other. B. Recordings from 2 of the 4 swim interneurons (SiI and SiII), in a different preparation from A. Both interneurons are on the same side of the brain, and due to electrical coupling and synaptic interactions, they fire in phase with each other
Fig. 5. The influence of light on expression of the swim motor program. With the lights off, this isolated Melibe brain produced a normal swim motor program, manifested in this illustration by rhythmic bursting in one of the SMNs located in the pedal ganglia. When the lights illuminating the preparation were turned on, fictive swimming ceased. When the lights were turned off again, swimming quickly resumed
Fig. 6. Cells containing NOS in the Melibe CNS. A. NADPH-diaphorase staining reveals a pair of small, bilaterally symmetrical neurons in the cerebropleural ganglia. These neurons project to the pedal ganglia. There is also intense staining in the cerebral neuropil, but this does not appear to originate from the NOS cells in the brain. B. Immunohistochemical staining of a Melibe brain with antibodies directed against a conserved region of NOS also stain what appear to be the identical neurons. The scale bar = 100 μm in both pictures
Fig. 7. Influence of a NO donor on expression of the swim motor program by an isolated Melibe brain. In A. 1 min after addition of 1.0 mM SNP, the normal swim rhythm expressed by one of the SMNs was disrupted and a new, slower rhythm was produced. B. This slow rhythm became more consistent and lasted for more than 20 min despite continuous washing. After 1 hr the pattern returned to normal
From the Symposium Swimming in Opisthobranch Mollusks: Contributions to Control of Motor Behavior presented at the Annual Meeting of the Society for Integrative and Comparative Biology, 4–8 January 2000, at Atlanta, Georgia.
E-mail: win@unh.edu
Current address: Highline Community College, 2400 S. 240th St, P.O. Box 98000, Des Moines, Washington, 98198-9800
Current address: Department of Biology, Georgia State University, P.O. Box 4010, Atlanta, Georgia 30302-4010
We would like to thank all the people at the Friday Harbor Laboratories (FHL), particularly Dennis Willows, David Duggins and Craig Staude for being wonderful hosts, solving problems, and providing a great atmosphere for research. We also appreciate all the FHL students who contributed insights about Melibe neuroethology throughout the years. We would also like to thank Cy, Janis, Celia and Sooz for collecting animals and helping with video analyses, Noel Carlson and Steve Jury for sharing their underwater observations, and Dan O'Grady for help maintaining animals at UNH. Finally, we would like to express our appreciation to the anonymous reviewer whose input improved this manuscript. This study was supported by Center for Marine Biology Grants and Summer Teaching Fellowships to K.A.L. through the University of New Hampshire, Society for Integrative and Comparative Biology and Center for Marine Biology grants to J.M.N. and a NIH grant to W.H.W. It is contribution number #377 of the Center for Marine Biology/Jackson Estuarine Laboratory series. The Symposium was supported by National Science Foundation grant IBN 990 5990.
