Tritonia swim network
Tritonia is a gastropod mollusc that executes an escape swimming response to particular aversive stimuli, such as contact with the tube feet of the sea star, Pycnopodia helianthoides. The neural commands for this rhythmic behavior are generated by a central pattern generator (CPG) that has been well characterized. The Tritonia swim CPG has been a model system for studying rhythmic motor pattern generation, neuromodulation, and behavioral plasticity. There are several important older reviews of this system that represent the fundamental knowledge of its organization (Getting, 1989; Getting and Dekin, 1985b; Getting, 1983b). This article provides our most up-to-date summary of how the Tritonia swim network functions.
Tritonia's swim behavior (Fig. 1) consists of a series of 2-8 alternating whole body flexions. The animal bends at its midpoint along its length and touches the tips of its body at the anterior and posterior extremes first on the ventral side and then on the dorsal side. The initial ventral body flexion provides lift that gets the animal off the sea floor. The animal ends a bout of swimming on a dorsal flexion. This settles the foot of the animal back onto the substrate. Behavioral observations in natural habitats suggest that the swim is a rare behavior, performed primarily for predator avoidance (Wyeth and Willows, 2006).
The escape swim is generated by identified neurons in the central nervous system (see Fig. 3). By recording the activity of individual neurons in an intact animal preparation, it was first shown that the activity of neurons in the central ganglia could be correlated with particular behaviors (Willows, 1967). It was later shown that the organized bursting firing pattern that mediates the swim could be produced by the isolated nervous system, proving that the pattern of activity was centrally generated and not the result of a cycle-by-cycle reflex response (Dorsett et al., 1969; Willows and Hoyle, 1969).
Sensory activation of the swim motor pattern
The swim motor pattern (Fig. 2) is initiated by chemical stimuli associated with contact with the tube feet of certain predatory seastars. Mechanical stimulation is not sufficient to activate the behavior. In the laboratory, other stimuli will also elicit a swim response. These include concentrated NaCl solutions or salt crystals applied to the skin. Electrical stimulation of the skin will also initiate a swim response. In the isolated nervous system, brief electrical stimulation of a body wall nerve, such as Pedal Nerve 2 or 3 is used as a standard stimulus to initiate what is often termed the "fictive" motor pattern.Tri0002367) mediate the response to the starfish (Getting, 1976). These neurons are both mechanoreceptive and chemoreceptive. The S-cells have somata in the pleural ganglia and receptive fields on the body surface. An S-cell responds to mechanical stimulation in its receptive field with a brief burst of spikes. However, contact with the tube feet of a starfish causes more maintained firing. These neurons may be homologous to the Aplysia pleural ventral cluster sensory neurons.
A single S-cell is insufficient to activate the swim motor pattern. However, multiple S-cells firing simultaneously will trigger a swim motor pattern. The S-cells make excitatory glutamateric synapses (Megalou et al., 2009) onto two neurons that convey the excitation to the CPG: Tr1 (Tri0002513) and DRI (Tri0002471). These EPSPs are short in duration and therefore do not account for the duration of the swim motor pattern, which can last for more than a minute.
Command and Gating interneuronsThe sensory input from starfish touch is conveyed to interneurons Tr1 and DRI (Fig. 3), which initiate the motor pattern. Tr1 acts as a transient trigger-type command neuron (Frost et al., 2001), analogous to those described in leech swimming (Brodfuehrer et al., 1995), namely, it fires for a brief period of time but activates a longer-lasting rhythmic motor pattern. Depolarization of Tr1 is occasionally sufficient to trigger a swim motor pattern. Tr1 does not appear to synapse directly on the CPG neurons. Instead, it relays the excitation to DRI. However, unlike the rapid EPSPs produced by the S-cells on DRI, the synapse from Tr1 has a slow component as well, which could aid in the initiation phase of the motor pattern.
There are two inhibitory gating interneurons: Pl 9 (Tri0002525) and Pl 10 (Tri0002586). These neurons can prevent the initiation of the swim motor pattern by feeding back to the sensory neurons and inhibiting other points along the pathway.
DRI, through its potent fast excitatory EPSPs onto all 6 DSIs, acts as a command gating neuron for the swim motor pattern (Frost and Katz, 1996) again analogous to the function of neurons described in the pathway for leech swimming (Brodfuehrer et al., 1995). It fulfills the strict requirements for a “command neuron” (Kupfermann and Weiss, 1978), namely that is both necessary and sufficient to elicit the response. When DRI is hyperpolarized, stimulation of the nerve or sensory stimuli applied to the body wall will not evoke a swim motor pattern (Frost and Katz, 1996; Frost et al., 2001). Depolarization of a single DRI is sufficient to trigger and maintain a rhythmic swim motor pattern.
The Swim Central Pattern Generator
The swim CPG produces the pattern of rhythmic activity that drives the swim efferent neurons, located in the pedal ganglia, which in turn convey their pattern of activity to the muscles. The swim CPG neurons have been identified by their interconnections with each other, their ability to reset the phase of the swim rhythm, and their synaptic connections to the pedal efferent neurons.Tri0002380), three Dorsal Swim Interneurons (DSI-A, DSI-B, DSI-C: Tri0001043), and two Ventral Swim Interneurons (VSI-A and VSI-B: Tri0002406, Tri0002436). VSI-A has connections to C2 and DSI that are consistent with it being a member of the CPG (Getting et al., 1980; Getting, 1981), but it does not appear to exert a noticeable effect on the motor pattern when perturbed. The reason for this has not been explored. VSI-B was later identified as member of the CPG (Getting, 1983a) based on its firing during the swim rhythm, its connections with other CPG neurons and swim efferent neurons, and its ability to reset the phase of the rhythm when made to fire.
Of the three DSIs, two of them (DSI-B,C) are indistinguishable from one another. DSI-B and C are electrically coupled to each other, but not to DSI-A (Fig. 4). DSI-A also differs slightly from DSI-B,C in its synaptic connectivity (Getting, 1981). The DSIs make mutual excitatory chemical synapses onto one another. Generally, in publications, the type of DSI is not noted.
None of the CPG neurons has been observed to have intrinsic bursting properties. Rather, the rhythmic bursting during the swim motor pattern (Fig. 2) seems to arise through conventional synaptic interactions, making the CPG a rare example of a "Network Oscillator". A basic scenario for initiation and production of a swim motor pattern is as follows:
- Sensory input activates DRI.
- DRI excites DSI
- DSI excites C2
- C2 feeds back and excites DRI, further exciting DSI via a positive feedback loop.
- C2 excites VSI, which inhibits DSI and C2, thereby momentarily interrupting the positive feedback loop.
A positive feedback loopAs noted above, DRI acts as a gating-type command interneuron for the swim motor pattern. Most sensory input to the DSIs appears to be funneled through DRI, which bilaterally excites all of the DSIs with large, synchronous monosynaptic EPSPs. These are so strong that at low to medium firing rates DRI action potentials typically elicit one-for-one action potentials in the DSIs, making this the strongest known synapse in the Tritonia swim network. When the DSIs are activated individually, they recruit inhibition onto themselves through an unidentified inhibitory interneuron (the so-called I-cell) (Getting and Dekin, 1985a). The synchronous excitation of the DSIs allows them to overcome their mutual inhibition, leading to a strong burst of activity. The DSIs fire at about 40Hz at the onset of a swim motor program.
The DSIs directly excite C2, leading to its firing after DRI is activated. During each cycle of the swim motor pattern, DSI spiking activity precedes that of C2 (Figs. 2, 5). In addition to the synaptic activation of C2, the DSIs evoke neuromodulatory actions on C2, enhancing its excitability and increasing its synaptic strength (Katz et al., 1994; Katz and Frost, 1995b; Katz and Frost, 1997). Both the synaptic and neuromodulatory actions are mediated by serotonin (5-hydroxytryptamine, 5-HT) released from DSI (Katz and Frost, 1995a). The serotonergic neuromodulatory actions may be necessary for C2 to excite DRI; at rest C2 stimulation is often not sufficient to cause DRI spiking.
Early models (Getting and Dekin, 1985b; Getting, 1989) differed from this scenario because DRI had not yet been identified. Instead of having positive feedback from C2 to DRI, those models relied on a constantly decaying “Ramp” depolarization to DSI that was triggered by sensory input. This decaying ramp depolarization can be attributed to input from DRI, leading it to be named the "Dorsal Ramp Interneuron" (Frost and Katz, 1996). The initial models also included positive feedback, but it was among the DSIs, which have mutual excitatory synapses. Later results suggest that this component of the excitation is minor compared to DRI input to the DSIs.
The positive feedback loop (CPG to DRI to CPG) is interrupted periodically by inhibition from VSI. DSI inhibits VSI-B, serving as feedforward inhibition. But C2 provides delayed excitation of VSI-B. Once VSI-B escapes DSI, it feeds back inhibition, strongly inhibiting both C2 and DSI, preventing them from firing. Once C2 stops firing, it ceases to excite VSI-B. This causes VSI-B to fall silent, allowing the C2-DRI-DSI positive feedback between the CPG and DRI to continue.
During the VSI inhibition, DRI continues to fire. Its firing during the swim motor pattern is not bursting, but is more tonic with a slight increase in frequency timed to C2 bursts. Thus, DRI activity appears to be necessary to sustain excitation in DSI and thereby also C2. This suggests that bursting is a network function; sustained and recurrent excitation recruits inhibition back on itself.
Where is the oscillator?
Given the scenario above, the question arises as to whether DRI is a member of the CPG or is upstream from the CPG. Although DRI does not display large amplitude membrane potential oscillations, its firing frequency is modulated in time with the ongoing motor pattern. Furthermore, as already noted, DRI activity is necessary for activation and maintenance of rhythmic activity in response to sensory stimulation. However, rhythmic activity can be elicited by injecting current into one or more DSIs to cause them to fire tonically at 10-20 Hz (Fickbohm and Katz, 2000; Katz et al., 2004). When DRI is hyperpolarized, DSI stimulation can still elicit rhythmic activity (Frost et al., 2001). This suggests that DRI is important for maintaining excitatory input to the rhythmogenic component of the circuit, but is not part of the CPG that generates alternating bursting because downstream elements (DSI, C2, and VSI) can oscillate without DRI if provided with the sustained excitatory input that DRI normally provides.
The motor pattern elicited by tonic DSI firing does not continue past the end of the DSI spike train. Interestingly, this is not true when the serotonin content of the DSIs is enhanced by pretreatment with the serotonin precursor 5-hydroxytryptophan (5-HTP). Instead, bursting activity continues while the DSIs fall silent (Fickbohm and Katz, 2000), suggesting that the function of the DSIs in creating the oscillatory pattern is simply to deliver serotonin. Furthermore, with the enhanced serotonin content, brief depolarization of a DSI is sufficient to trigger prolonged rhythmic activity in which the DSIs are silent. Thus, with high serotonin content in the DSIs, they do not act as part of the oscillator.
As with DRI and DSI, depolarization of C2 can lead to rhythmic activity in the other CPG members (Taghert and Willows, 1978; Getting, 1977). However, unlike DSI and DRI, a brief depolarization of C2 can sometimes elicit a prolonged motor pattern in which C2 participates (Katz et al., 2004). It is not known why brief activation of C2, but not DRI or DSI is capable triggering a regenerative swim motor pattern. However, it suggests that activation of C2 is the key step for motor pattern generation.
Control of cycle period duration
During the course of the swim motor pattern, the cycle period gradually lengthens (Lennard et al., 1980). The lengthening of the cycle period appears to be caused by a delay in the excitation of C2 by DSI. It is associated with a gradual decrease in DSI firing frequency. Thus, the duration of the swim motor pattern might depend upon a certain level of sustained excitation within the circuit.
The initial cycle period of a swim motor pattern depends upon the strength of the stimulus, when the previous swim motor pattern was elicited, and the ambient temperature. It can be as short as 4 seconds or as long as 15 seconds (Lennard et al., 1980; Katz et al., 2004). The cycle period is not affected by changing the membrane potential of any of the CPG neurons (Katz et al., 2004). This suggests that the bursting is not produced through voltage-dependent properties, but could be a synaptically-driven oscillator. The periodicity may also be related to intracellular signaling cascades driven by G protein-coupled receptors; inhibiting adenylyl cyclase increased the period of the motor pattern before disrupting it completely (Clemens et al., 2007).
The DSIs release 5-HT to evoke neuromodulatory as well as synaptic actions within the CPG (Katz et al., 1994; Katz and Frost, 1995a; Fickbohm and Katz, 2000; Sakurai and Katz, 2003; McClellan et al., 1994). Because the DSIs are part of the CPG, this was termed “intrinsic neuromodulation” to contrast it with neuromodulatory input arising from other parts of the nervous system (Katz and Frost, 1996). DSI activity leads to a short-term increase in the strength of synapses made by C2 and VSI. It also alters the membrane properties of these two neurons (Katz and Frost, 1997; Sakurai et al., 2006).The DSI neuromodulatory actions on VSI synaptic strength are timing and state-dependent (Sakurai and Katz, 2009). DSI increases synaptic strength when its spikes precede the VSI spikes by less than 15 sec (Fig. 6 A,B), but if the VSI synapses are in a potentiated state, then DSI stimulation depotentiates them (Fig. 6 B, D). The synaptic reset occurs even if VSI follows DSI by more than 15 sec (Fig. 6D). This may play a role in reseting VSI synaptic strength after the end of a swim episode.
The swimming behavior is a rare, high threshold response that the animal engages in only in response to particular stimuli. The neuromodulatory actions of the DSI contribute to this switch from a non-swimming state to a swimming state. Computational modeling studies suggest that in the absence of the neuromodulation, the neurons and synapses are incapable of producing rhythmic bursting (Calin-Jageman et al., 2007). It is difficult to construct a model of the circuit with static parameters that will switch from non-bursting to bursting. In an examination of parameter space, it was found that of the models that exhibited bursting, 96% continued to burst indefinitely and more than 85% of them produced spontaneous bursting in the absence of an external trigger. This suggests that both oscillatory state the quiescent state are stable. Transition from one state to another requires changing a number of parameters in the swim CPG.
Output to Efferent Neurons
The CPG neurons are all interneurons; none have been found to have axons that leave the central ganglia. They synapse on a set of efferent neurons that project to the body musculature. The efferent neurons that are rhythmically active during a swim motor pattern have been broadly classified as Ventral Flexion Neurons (VFN) and Dorsal Flexion Neurons (Willows et l., 1973), which are comprised of identified neurons in the pedal ganglia (Hume et al., 1982a). The DFN class is further subdivided into DFN-A and DFN-B based on their bursting properties; the DFN-B neurons exhibited higher spike frequency and more defined bursts than DFN-A. Each of the CPG neurons synapses in a divergent fashion onto each class of flexion neuron (Hume et al., 1982b). The synapses are multicomponent consisting of both fast and slow components. These neurons are called efferents rather than motor neurons because it has been difficult to show that they have direct effects on the muscles. Instead, they may cause the muscles to contract through indirect means such as acting on peripheral neurons.
Some, if not all of the neurons in the swim CPG are multifunctional. That is, they have other functions when the animal is not swimming. For example, the DSIs excite neurons in the pedal ganglia that increase the speed of crawling (Popescu and Frost, 2002). Following a swim motor pattern the DSIs continue to fire at 1-5 Hz. This contributes to an increase in motility observed following a swim episode. In contrast to the DSIs, VSI activity inhibits crawling neurons (as well as the DSIs) and would presumably lead to a decrease in crawling.
The swimming behavior and the swim motor pattern produced by the isolated nervous system both exhibit behavioral plasticity in the form of habituation, sensitization, and pre-pulse inhibition.
Habituation is a form of non-associative learning in which an animal decreases its response to a repeated stimulus. The Tritonia swim motor pattern exhibits habituation in the form of a decrease in the number of body flexion cycles with repeated stimulation (Abraham and Willows, 1971; Brown et al., 1996; Frost et al., 1996; Mongeluzi and Frost, 2000). Habituation progressively increases the initial burst period and decreases number of cycles. Presentation of a swim-eliciting stimulus also sensitizes the animal for the presentation of the next stimulus. This is evidenced as a decrease in the response latency in the second presentation (Frost et al., 1998). After the second presentation of the stimulus, response latency gradually increases. The sites of action within the swim circuit that mediate these behavioral changes have not been fully elucidated.
The Tritonia swim response exhibits prepulse inhibition (PPI), which is the ability of a weak prestimulus to suppress the response to a closely following stronger stimulus (Mongeluzi et al., 1998). PPI has clinical relevance because deficits in PPI have been associated with human disorders, such as schizophrenia, that involve defective sensorimotor gating. In Tritonia, a brief (100ms) tactile stimulus to one part of the body suppresses the initiation of swimming in response to electrical stimulation of another part. The tactile stimulus causes PPI by recruiting a pair of inhibitory gating interneurons (Pl 9 and Pl 10) that suppress activation of the swim network at a number of sites (Frost et al, 2003). First, Pl 9 causes presynaptic inhibition of transmitter release from the sensory neurons, preventing them from activating Tr1 and DRI. Second, Pl 9 postsynaptically inhibits both Tr1 and DRI. Third, the other interneuron that mediates PPI, Pl 10, polysynaptically inhibits the DSIs. Thus, following a prepulse stimulus the two interneurons briefly deprive the CPG of inputs elicited by other skin stimuli, leading to PPI. Because these interneurons also provide inhibitory input to the CPG in response to the typical swim-eliciting stimulus, they may also play a role in the balance of excitation and inhibition to the CPG that mediates the animal's decision to swim.
Evolution of the swim circuit
The neurons in the Tritonia swim system have homologues in related opisthobranch molluscs. Neurons homologous to DSI and C2 have been identified in Pleurobranchea and play a similar role during that animal’s swim motor pattern (Jing and Gillette, 1999; Jing and Gillette, 2000). However, there are species that are more closely related to Tritonia than Pleurobranchaea is that either do not swim or they swim with side-to-side body flexions rather than dorsal-ventral body flexions (Katz et al., 2001). These other species have homologues of the DSIs (Newcomb and Katz, 2007). In Melibe leonine, which swims with lateral body flexions, the DSI homologues are not members of the swim CPG but instead play an extrinsic neuromodulatory role: they are sufficient to initiate the motor pattern, but are not necessary nor are they rhythmically active (Newcomb and Katz, 2008). This suggests that the swim CPG in Tritonia arose independently from that in Pleurobranchea. Furthermore, the evidence suggests that both animals arose from a common ancestor that did not swim and had non-oscillatory circuitry (Newcomb et al., 2012).
In Tritonia serotonergic neuromodulation by the DSIs enhances the synaptic strength of C2 (Katz et al, 1994). In Pleurobranchaea, the DSI homologues and serotonin also the synaptic strength of the C2 homologue (Lillvis and Katz, 2013). However, individual animals sometimes do not swim. The strength of serotonergic neuromodulation correlates with the number of swim cycles. Furthermore, this modulation is absent in Hermissenda crassicornis, which does not exhibit this dorsal/ventral swimming behavior (Lillvis and Katz, 2013).
- Abraham FD, Willows AOD (1971) Plasticity of a fixed action pattern in the sea slug Tritonia diomedia [sic]. Comm Behav Biol 6:271-280.
- Brodfuehrer PD, Debski EA, O'Gara BA, Friesen WO (1995) Neuronal control of leech swimming. J Neurobiol 27:403-418.
- Brown GD, Frost WN, Getting PA (1996) Habituation and iterative enhancement of multiple components of the Tritonia swim response. Behavioral Neuroscience 110:478-485.
- Brown GD, Yamada S, Sejnowski TJ (2001) Independent component analysis at the neural cocktail party. Trends Neurosci 24:54-63.
- Calin-Jageman RJ, Tunstall MJ, Mensh BD, Katz PS, Frost WN (2007) Parameter space analysis suggests multi-site plasticity contributes to motor pattern initiation in Tritonia. J Neurophysiol 98:2382-2398.
- Clemens S, Calin-Jageman R, Sakurai A, Katz PS (2007) Altering cAMP levels within a central pattern generator modifies or disrupts rhythmic motor output. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 193:1265-1271.
- Dorsett DA, Willows AOD, Hoyle G (1969) Centrally generated nerve impulse sequences determining swimming behavior in Tritonia. Nature 224:711-712.
- Fickbohm DJ, Katz PS (2000) Paradoxical actions of the serotonin precursor 5-hydroxytryptophan on the activity of identified serotonergic neurons in a simple motor circuit. J Neurosci 20:1622-1634.
- Frost WN, Brandon CL, Mongeluzi DL (1998) Sensitization of the Tritonia escape swim. Neurobiol Learning and Mem 69:126-135.
- Frost WN, Brown GD, Getting PA (1996) Parametric features of habituation of swim cycle number in the marine mollusc Tritonia diomedea. Neurobiol Learning and Mem 65:125-134.
- Frost WN, Hoppe TA, Wang J, Tian LM (2001) Swim initiation neurons in Tritonia diomedea. Am Zool 41:952-961.
- Frost WN, Katz PS (1996) Single neuron control over a complex motor program. Proc Natl Acad Sci U S A 93:422-426.
- Frost WN, Tian L-M, Hoppe, TA, Mongeluzi, DL, and Wang, J (2003) A cellular mechanism for prepulse inhibition. Neuron, 40:991-1001.
- Frost WN, Wang J, Brandon CJ (2007) A stereo-compound hybrid microscope for combined intracellular and optical recording of invertebrate neural network activity. J Neurosci Methods.
- Getting PA (1976) Afferent neurons mediating escape swimming of the marine mollusc, Tritonia. J Comp Physiol 110:271-286.
- Getting PA (1977) Neuronal organization of escape swimming in Tritonia. J Comp Physiol A 121:325-342.
- Getting PA (1981) Mechanisms of pattern generation underlying swimming in Tritonia. I. Neuronal network formed by monosynaptic connections. J Neurophysiol 46:65-79.
- Getting PA (1983a) Mechanisms of pattern generation underlying swimming in Tritonia. III. Intrinsic and synaptic mechanisms for delayed excitation. J Neurophysiol 49:1036-1050.
- Getting PA (1983b) Neural control of swimming in Tritonia. In: Symposia of the Society for Experimental Biology, No.37, Neural Origin of Rhythmic Movements (Roberts A, Roberts BL, eds), pp 89-128. New York: Cambridge Univ. Press.
- Getting PA (1989) A network oscillator underlying swimming in Tritonia. In: Neuronal and Cellular Oscillators (Jacklet JW, ed), pp 215-236. New York: Marcel Dekker, Inc.
- Getting PA, Dekin MS (1985a) Mechanisms of pattern generation underlying swimming in Tritonia. IV. Gating of central pattern generator. J Neurophysiol 53:466-480.
- Getting PA, Dekin MS (1985b) Tritonia swimming: a model system for integration within rhythmic motor systems. In: Model Neural Networks and Behavior (Selverston AI, ed), pp 3-20. New York: Plenum Press.
- Getting PA, Lennard PR, Hume RI (1980) Central pattern generator mediating swimming in Italic textTritonia. I. Identification and synaptic interactions. J Neurophysiol 44:151-164.
- Hume RI, Getting PA, Del Beccaro MA. (1982a)Motor organization of Tritonia swimming. I. Quantitative analysis of swim behavior and flexion neuron firing patterns. J.Neurophysiol. 47:60-74.
- Hume RI and Getting PA. (1982b) Motor organization of Tritonia swimming. II. Synaptic drive to flexion neurons from premotor interneurons. J.Neurophysiol. 47:75-90.
- Jing J, Gillette R (1999) Central pattern generator for escape swimming in the notaspid sea slug Pleurobranchaea californica. J Neurophysiol 81:654-667.
- Jing J, Gillette R (2000) Escape swim network interneurons have diverse roles in behavioral switching and putative arousal in Pleurobranchaea. J Neurophysiol 83:1346-1355.
- Katz PS, Fickbohm DJ, Lynn-Bullock CP (2001) Evidence that the swim central pattern generator of Tritonia arose from a non-rhythmic neuromodulatory arousal system: Implications for the evolution of specialized behavior. Am Zool 41:962-975.
- Katz PS, Frost WN (1995a) Intrinsic neuromodulation in the Tritonia swim CPG: Serotonin mediates both neuromodulation and neurotransmission by the dorsal swim interneurons. J Neurophysiol 74:2281-2294.
- Katz PS, Frost WN (1995b) Intrinsic neuromodulation in the Tritonia swim CPG: The serotonergic dorsal swim interneurons act presynaptically to enhance transmitter release from interneuron C2. J Neurosci 15:6035-6045.
- Katz PS, Frost WN (1996) Intrinsic neuromodulation: Altering neuronal circuits from within. Trends Neurosci 19:54-61.
- Katz PS, Frost WN (1997) Removal of spike frequency adaptation via neuromodulation intrinsic to the Tritonia escape swim central pattern generator. J Neurosci 17:7703-7713.
- Katz PS, Getting PA, Frost WN (1994) Dynamic neuromodulation of synaptic strength intrinsic to a central pattern generator circuit. Nature 367:729-731.
- Katz PS, Sakurai A, Clemens S, Davis D (2004) Cycle Period of a Network Oscillator Is Independent of Membrane Potential and Spiking Activity in Individual Central Pattern Generator Neurons. J Neurophysiol 92:1904-1917.
- Kupfermann I, Weiss KR (1978) The command neuron concept. Behav Brain Sci 1:3-39.
- Lennard PR, Getting PA, Hume RI (1980) Central pattern generator mediating swimming in Tritonia. II. Initiation, maintenance, and termination. J Neurophysiol 44:165-173.
- Lillvis JL and Katz PS (2013) Parallel evolution of serotonergic neuromodulation underlies independent evolution of rhythmic motor behavior. J.Neurosci. 33 (6):2709-2717.
- McClellan AD, Brown GD, Getting PA (1994) Modulation of swimming in Tritonia: Excitatory and inhibitory effects of serotonin. J Comp Physiol A 174:257-266.
- Megalou, EV, Brandon, CJ and Frost, WN (2009) Evidence that the Tritonia diomedea swim afferent neurons are glutamatergic. Biological Bulletin, 216: 103-112.
- Mongeluzi DL, Frost WN (2000) Dishabituation of the Tritonia escape swim. Learn Mem 7:43-47.
- Mongeluzi DL, Hoppe TA, Frost WN (1998) Prepulse Inhibition of the Tritonia Escape Swim. J Neurosci 18:8467-8472.
- Newcomb JM, Katz PS (2007) Homologues of serotonergic central pattern generator neurons in related nudibranch molluscs with divergent behaviors. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 193:425-443.
- Newcomb JM, Katz PS (2008) Different functions for homologous serotonergic interneurons and serotonin in species-specific rhythmic behaviours. Proc Biol Sci.
- Newcomb JM, Sakurai A, Lillvis JL, Gunaratne CA, and Katz PS (2012) Homology and homoplasy of swimming behaviors and neural circuits in Nudipleura molluscs. Proc.Natl.Acad.Sci.U.S A 109 Suppl 1:10669-10676.
- Popescu IR, Frost WN (2002) Highly dissimilar behaviors mediated by a multifunctional network in the marine mollusk Tritonia diomedea. J Neurosci 22:1985-1993.
- Sakurai A, Darghouth NR, Butera RJ, Katz PS (2006) Serotonergic enhancement of a 4-AP-sensitive current mediates the synaptic depression phase of spike timing-dependent neuromodulation. J Neurosci 26:2010-2021.
- Sakurai A, Katz PS (2003) Spike timing-dependent serotonergic neuromodulation of synaptic strength intrinsic to a central pattern generator circuit. J Neurosci 23:10745-10755.
- Sakurai A, Katz PS (2009) State-, timing-, and pattern-dependent neuromodulation of synaptic strength by a serotonergic interneuron. J Neurosci 29:268-279.
- Taghert PH, Willows AOD (1978) Control of a fixed action pattern by single, central neurons in the marine mollusk, Tritonia diomedea. J Comp Physiol 123:253-259.
- Willows AO (1967) Behavioral acts elicited by stimulation of single, identifiable brain cells. Science 157:570-574.
- Willows AOD, Dorsett DA, Hoyle G. (1973) The neuronal basis of behavior in Tritonia. III. Neuronal mechanism of a fixed action pattern. J.Neurobiol. 4:255-285.
- Willows AOD, Hoyle G (1969) Neuronal network triggering of fixed action pattern. Science 166:1549-1551.
- Wyeth RC, Willows AO (2006) Field behavior of the nudibranch mollusc Tritonia diomedea. Biol Bull 210:81-96.
- Tamas Freund and Szabolcs Kali (2008) Interneurons. Scholarpedia, 3(9):4720.
- James Newcomb (2008) Melibe. Scholarpedia, 3(5):3965.
- Peter Jonas and Gyorgy Buzsaki (2007) Neural inhibition. Scholarpedia, 2(9):3286.
- Rodolfo Llinas (2008) Neuron. Scholarpedia, 3(8):1490.
- Jeff Moehlis, Kresimir Josic, Eric T. Shea-Brown (2006) Periodic orbit. Scholarpedia, 1(7):1358.
- Jose-Manuel Alonso and Yao Chen (2009) Receptive field. Scholarpedia, 4(1):5393.
- Marco M Picchioni and Robin Murray (2008) Schizophrenia. Scholarpedia, 3(4):4132.
- Philip Holmes and Eric T. Shea-Brown (2006) Stability. Scholarpedia, 1(10):1838.
- Arkady Pikovsky and Michael Rosenblum (2007) Synchronization. Scholarpedia, 2(12):1459.
- Paul S. Katz (2007) Tritonia. Scholarpedia, 2(6):3504.