Unsupervised functional neurocartography of the Aplysia buccal ganglion

A major limitation of large-scale neuronal recordings is the difficulty in locating the same neuron in different subjects, referred to as the "correspondence" issue. This issue stems, at least in part, from the lack of a unique feature that unequivocally identifies each neuron. One promising approach to this problem is the functional neurocartography framework developed by Frady et al. (2016), in which neurons are identified by a semi-supervised machine learning algorithm using a combination of multiple selected features. Here, the framework was adapted to the buccal ganglia of Aplysia. Multiple features were derived from neuronal activity during motor pattern generation, responses to peripheral nerve stimulation, and the spatial properties of each cell. The feature set was optimized based on its potential usefulness in discriminating neurons from each other, and then used to match putatively homologous neurons across subjects with the functional neurocartography software. A matching method was developed based on a cyclic matching algorithm that allows for unsupervised extraction of groups of neurons, thereby enhancing scalability of the analysis. Cyclic matching was also used to automate the selection of high-quality matches, which allowed for unsupervised implementation of the machine learning algorithm. This study paves the way for investigating the roles of both well-characterized and previously uncharacterized neurons in Aplysia, as well as helps to adapt this framework to other systems.

Release of a single neurotransmitter from an identified interneuron coherently affects motor output on multiple time scales

Neurotransmitters can have diverse effects that occur over multiple time scales often making the consequences of neurotransmission difficult to predict. To explore the consequences of this diversity, we used the buccal ganglion of Aplysia to examine the effects of GABA release by a single interneuron, B40, on the intrinsic properties and motor output of the radula closure neuron B8. B40 induces a picrotoxin-sensitive fast IPSP lasting milliseconds in B8 and a slow EPSP lasting seconds. We found that the excitatory effects of this slow EPSP are also mediated by GABA. Together, these two GABAergic actions structure B8 firing in a pattern characteristic of ingestive programs. Furthermore, we found that repeated B40 stimulation induces a persistent increase in B8 excitability that was occluded in the presence of the GABA B receptor agonist baclofen, suggesting that GABA affects B8 excitability over multiple time scales. The phasing of B8 activity during the feeding motor programs determines the nature of the behavior elicited during that motor program. The persistent increase in B8 excitability induced by B40 biased the activity of B8 during feeding motor programs causing the motor programs to become more ingestive in nature. Thus, a single transmitter released from a single interneuron can have consequences for motor output that are expressed over multiple time scales. Importantly, despite the differences in their signs and temporal characteristics, the three actions of B40 are coherent in that they promote B8 firing patterns that are characteristic of ingestive motor outputs.

Feeding CPG in Aplysia Directly Controls Two Distinct Outputs of a Compartmentalized Interneuron That Functions as a CPG Element

In the context of motor program generation in Aplysia, we characterize several functional aspects of intraneuronal compartmentalization in an interganglionic interneuron, CBI-5/6. CBI-5/6 was shown previously to have a cerebral compartment (CC) that includes a soma that does not generate full-size action potentials and a buccal compartment (BC) that does. We find that the synaptic connections made by the BC of CBI-5/6 in the buccal ganglion counter the activity of protraction-phase neurons and reinforce the activity of retraction-phase neurons. In buccal motor programs, the BC of CBI-5/6 fires phasically, and its premature activation can phase advance protraction termination and retraction initiation. Thus the BC of CBI-5/6 can act as an element of the central pattern generator (CPG). During protraction, the CC of CBI-5/6 receives direct excitatory inputs from the CPG elements, B34 and B63, and during retraction, it receives antidromically propagating action potentials that originate in the BC of CBI-5/6. Consequently, in its CC, CBI-5/6 receives depolarizing inputs during both protraction and retraction, and these depolarizations can be transmitted via electrical coupling to other neurons. In contrast, in its BC, CBI-5/6 uses spike-dependent synaptic transmission. Thus the CPG directly and differentially controls the program phases in which the two compartments of CBI-5/6 may transmit information to its targets.

Nitric Oxide Potentiates cAMP-Gated Cation Current in Feeding Neurons of Pleurobranchaea californica Independent of cAMP and cGMP Signaling Pathways

Critical roles for nitric oxide (NO) in regulating cell and tissue physiology are broadly appreciated, but aspects remain to be explored. In the mollusk Pleurobranchaea, NO synthase activity is high in CNS ganglia containing motor networks for feeding and locomotion, where a cAMP-gated cation current ( INa,cAMP) is also prominent in many neurons. We examined effects of NO on INa,cAMP using voltage-clamp methods developed to analyze cAMP signaling in the live neuron, focusing on the identified metacerebral giant neuron of the feeding network. NO donors enhanced the INa,cAMP response to injected cAMP by an averaged 85%. In dose-response measures, NO increased the current stimulated by cAMP injection without altering either apparent cAMP binding affinity or cooperativity of current activation. NO did not detectably alter levels of native cAMP or synthesis or degradation rates as observable in both current saturation and decay rate of INa,cAMP responses to cAMP injection. NO actions were not exerted by cGMP signaling, as they were not mimicked by cGMP analogue nor blocked by inhibitors of guanylate cyclase and protein kinase G. NO potentiation of INa,cAMP was broadly distributed among many other neurons of the feeding motor network in the buccal ganglion. However, NO did not affect a second type of INa,cAMP found in locomotor neurons of the pedal ganglia. These results suggest that NO acts through a novel mechanism to regulate the gain of cAMP-dependent neuromodulatory pathways that activate INa,cAMP and may thereby affect the set points of feeding network excitability and reactivity to exogenous input.

Transforming Tonic Firing Into a Rhythmic Output in the Aplysia Feeding System: Presynaptic Inhibition of a Command-Like Neuron by a CPG Element

Tonic stimuli can elicit rhythmic responses. The neural circuit underlying Aplysia californica consummatory feeding was used to examine how a maintained stimulus elicits repetitive, rhythmic movements. The command-like cerebral-buccal interneuron 2 (CBI-2) is excited by tonic food stimuli but initiates rhythmic consummatory responses by exciting only protraction-phase neurons, which then excite retraction-phase neurons after a delay. CBI-2 is inhibited during retraction, generally preventing it from exciting protraction-phase neurons during retraction. We have found that depolarizing CBI-2 during retraction overcomes the inhibition and causes CBI-2 to fire, potentially leading CBI-2 to excite protraction-phase neurons during retraction. However, CBI-2 synaptic outputs to protraction-phase neurons were blocked during retraction, thereby preventing excitation during retraction. The block was caused by presynaptic inhibition of CBI-2 by a key buccal ganglion retraction-phase interneuron, B64, which also causes postsynaptic inhibition of protraction-phase neurons. Pre- and postsynaptic inhibition could be separated. First, only presynaptic inhibition affected facilitation of excitatory postsynaptic potentials (EPSPs) from CBI-2 to its followers. Second, a newly identified neuron, B54, produced postsynaptic inhibition similar to that of B64 but did not cause presynaptic inhibition. Third, in some target neurons B64 produced only presynaptic but not postsynaptic inhibition. Blocking CBI-2 transmitter release in the buccal ganglia during retraction functions to prevent CBI-2 from driving protraction-phase neurons during retraction and regulates the facilitation of the CBI-2 induced EPSPs in protraction-phase neurons.

A Newly Identified Buccal Interneuron Initiates and Modulates Feeding Motor Programs in Aplysia

Despite considerable progress in characterizing the feeding central pattern generator (CPG) in Aplysia, the full complement of neurons that generate feeding motor programs has not yet been identified. The distribution of neuropeptide-containing neurons in the buccal and cerebral ganglia can be used as a tool to identify additional elements of the feeding circuitry by providing distinctions between otherwise morphologically indistinct neurons. For example, our recent study revealed a unique and potentially interesting unpaired PRQFVamide (PRQFVa)-containing neuron in the buccal ganglion. In this study, we describe the morphological and electrophysiological characterization of this novel neuron, which we designate as B50. We found that activation of B50 is capable of producing organized rhythmic output of the feeding CPG. The motor programs elicited by B50 exhibit some similarities as well as differences to motor programs elicited by the command-like cerebral-to-buccal interneuron CBI-2. In addition to activating the feeding CPG, B50 may act as a program modulator.

Diverse Synaptic Connections Between Peptidergic Radula Mechanoafferent Neurons and Neurons in the Feeding System ofAplysia

The buccal ganglion of Aplysia contains a heterogeneous population of peptidergic, radula mechanoafferent (RM) neurons. To investigate their function, two of the larger RM cells (B21, B22) were identified by morphological and electrophysiological criteria. Both are low-threshold, rapidly adapting, mechanoafferent neurons that responded to touch of the radula, the structure that grasps food during ingestive and egestive feeding movements. Sensory responses of the cells consisted of spike bursts at frequencies of 8–35 Hz. Each cell was found to make chemical, electrical, or combined synapses with other sensory neurons, motor neurons and interneurons involved in radula closure and/or protraction-retraction movements of the odontophore. Motor neurons receiving input included the following: B8a/b, B15, and B16, which innervate muscles contributing to radula closing; and B82, a newly identified neuron that innervates the anterodorsal region of the I1/I3 muscles of the buccal mass. B21 and B22 can affect buccal motor programs by way of their connections to interneurons such as B19 and B64. Fast, chemical, excitatory postsynaptic potentials (EPSPs) produced by RM neurons, such as B21, exhibited strong, frequency-dependent facilitation, a form of homosynaptic plasticity. Firing B21 also produced a slow EPSP in B15 that increased the excitability of the cell. Thus a sensory neuron mediating a behavioral response may have modulatory effects. The data suggest multiple functions for RM neurons including 1) triggering of phase transitions in rhythmic motor programs, 2) adjusting the force of radula closure, 3) switching from biting to swallowing or swallowing to rejection, and 4) enhancing food-induced arousal.

C-PR Neuron of Aplysia Has Differential Effects on “Feeding” Cerebral Interneurons, Including Myomodulin-Positive CBI-12

C-PR neuron of Aplysia has differential effects on “feeding” cerebral interneurons, including myomodulin-positive CBI-12. Head lifting and other aspects of the appetitive central motive state that precedes consummatory feeding movements in Aplysia is promoted by excitation of the C-PR neuron. Food stimuli activate C-PR as well as a small population of cerebral-buccal interneurons (CBIs). We wished to determine if firing of C-PR produced differential effects on the various CBIs or perhaps affected all the CBIs uniformly as might be expected for a neuron involved in producing a broad undifferentiated arousal state. We found that when C-PR was fired, it produced a wide variety of effects on various CBIs. Firing of C-PR evoked excitatory input to a newly identified CBI (CBI-12) the soma of which is located in the M cluster near the previously identified CBI-2. CBI-12 shares certain properties with CBI-2, including a similar morphology and a capacity to drive rhythmic activity of the buccal-ganglion. Unlike CBI-2, CBI-12 exhibits myomodulin immunoreactivity. Furthermore when C-PR is fired, CBI-12 receives a polysynaptic voltage-dependent slow excitation, whereas, CBI-2 receives relatively little input. C-PR also polysynaptically excites other CBIs including CBI-1 and CBI-8/9 but produces inhibition in CBI-3. In addition, firing of C-PR inhibits plateau potentials in CBI-5/6. The data suggest that activity of C-PR may promote the activity of one subset of cerebral-buccal interneurons, perhaps those involved in ingestive behaviors that occur during the head-up posture. C-PR also inhibits some cerebral-buccal interneurons that may be involved in behaviors in which C-PR activity is not required or may even interfere with other feeding behaviors such as rejection or grazing, that occur with the head down.

Behavior-Dependent Activities of a Central Pattern Generator in Freely Behaving Lymnaea stagnalis

Jansen, Rene F., Anton W. Pieneman, and Andries ter Maat. Behavior-dependent activities of a central pattern generator in freely behaving Lymnaea stagnalis. J. Neurophysiol. 78: 3415–3427, 1997. Cyclic or repeated movements are thought to be driven by networks of neurons (central pattern generators) that are dynamic in their connectivity. During two unrelated behaviors (feeding and egg laying), we investigated the behavioral output of the buccal pattern generator as well as the electrical activity of a pair of identified interneurons that have been shown to be involved in setting the level of activity of this pattern generator (PG). Analysis of the quantile plots of the parameters that describe the behavior (movements of the buccal mass) reveals that during egg laying, the behavioral output of the PG is different compared with that during feeding. Comparison of the average durations of the different parts of the buccal movements showed that during egg laying, the duration of one specific part of buccal movement is increased. Correlated with these changes in the behavioral output of the PG were changes in the firing rate of the cerebral giant neurons (CGC), a pair of interneurons that have been shown to modulate the activity of the PG by means of multiple synaptic contacts with neurons in the buccal ganglion. Interval- and autocorrelation histograms of the behavioral output and CGC spiking show that both the PG output and the spiking properties of the CGCs are different when comparing egg-laying animals with feeding animals. Analysis of the timing relations between the CGCs and the behavioral output of the PG showed that both during feeding and egg laying, the electrical activity of the CGCs is largely in phase with the PG output, although small changes occur. We discuss how these results lead to specific predictions about the kinds of changes that are likely to occur when the animal switches the PG from feeding to egg laying and how the hormones that cause egg laying are likely to be involved.

Behavioral Function of Glutamatergic Interneurons in the Feeding System of Lymnaea: Plateauing Properties and Synaptic Connections with Motor Neurons

Brierley, Matthew J., Kevin Staras, and Paul R. Benjamin. Behavioral function of glutamatergic interneurons in the feeding system of Lymnaea: plateauing properties and synaptic connections with motor neurons. J. Neurophysiol. 78: 3386–3395, 1997. Intracellular recording techniques were used to examine the electrical properties and behavioral function of a novel type of retraction phase interneuron, the N2 ventral (N2v) cells in the feeding network of the snail Lymnaea. The N2vs were compared with the previously identified N2 cells that now are renamed the N2 dorsal (N2d) cells. The N2vs are a bilaterally symmetrical pair of electrotonically coupled plateauing interneurons that are located on the ventral surfaces of the buccal ganglia. Their main axons project to the opposite buccal ganglion, but they have an additional fine process in the postbuccal nerve. N2v plateaus that outlast the duration of the stimulus can be triggered by depolarizing current pulses and prematurely terminated by applied hyperpolarizing pulses. Gradually increasing the amplitude of depolarizing pulses reveals a clear threshold for plateau initiation. N2v plateauing persists in a high Mg2+/nominally zero Ca2+ saline that blocks chemical synaptic connections, suggesting an endogenous mechanism for plateau generation. The N2vs fire sustained bursts of action potentials throughout the N2/rasp phase of the fictive feeding cycle and control the retraction phase feeding motor neurons. The N2vs excite the B3 and B9 feeding motor neurons to fire during the rasp phase of the feeding cycle. They also inhibit the B7 and B8 feeding motor neurons. The B8 cells recover from inhibition and fire during the following swallowing phase. These synaptic connections appear to be monosynaptic as they persist in high Mg2+/high Ca2+ (HiDi) saline that blocks polysynaptic pathways. Strong current-induced plateaus in the N2vs generate brief inhibitory postsynaptic responses in the B4CL rasp phase motor neurons, but this was due to the indirect N2v → N2d → B4CL pathway. The N2vs are coupled electrotonically to the N2d cells, and triggering plateau in a N2v usually induced one or two spikes in a N2d. Previous experiments showed that the N2ds generate plateau potentials during a fictive feeding cycle. Here we show that the main component of the “plateauing” waveform is due to the electrotonic coupling with the N2v cells. The differential synaptic connections of the N2v and N2d cells with retraction phase motor neurons results in a sequence of motor neuron burst activity B9 → B4CL → B8 that produces the full retraction (rasp → swallow) movements of the feeding apparatus (buccal mass). We conclude that the N2v cells are an essential component of the interneuronal network required to produce feeding motor neuron activity.