Neural circuits underlying feeding in Aplysia

Feeding behavior of Aplysia is a useful model system with which to study the neural control of relatively complex and adaptive behaviors. Key aspects of feeding behavior involve rhythmic movements of structures in the foregut, such as the radula (the toothed grasping surface). The buccal ganglia contain a central pattern generator (CPG) that mediates rhythmic movements of the foregut during feeding. The behavior and the underlying pattern of neural activity (i.e., buccal motor patterns; BMPs) can be divided into two basic phases: the protraction and retraction of the radula. Different behaviors (e.g., ingestion vs rejection) are determined by the timing of other movements, such as closure of the radula. Thus, the CPG generates at least two types of BMPs: one that mediates ingestion (iBMP) and one that mediates rejection (rBMP). Although many cells and synaptic connections within the CPG have been identified, an analysis of its overall function is still incomplete. The constructionist methods of mathematical modeling can help develop insights to the functions of neural circuits. Previously, we developed a conductance-based model of the individual neurons and their synaptic connections. The simulations were performed with the neurosimulator SNNAP. The model successfully simulated the observed properties of neurons and synaptic connections within the CPG. The model circuit contained neurons B31/32, B34, B35, B63, B4/5, B8, B51, B64, B52 and an unidentified cell (termed Z). The Z cell was necessary to mediate the transition between protraction and retraction and represents an important prediction of the model. The present study extended our previous model by incorporating a more detailed description of B31/B32 (see Hurwitz et al. 2008; Saada et al. 2009). B31/32 respond after a delay to stimuli and play a key role in the decision to initiate a BMP. The currents underlying the decision making process have been characterized and consist of one inward and three outward currents. The inward current is modulated by muscarinic synaptic transmission. We modeled this current and its modulation and we re-evaluated the mechanisms underlying the genesis of BMPs. The network model produced robust rhythmic activity similar to rBMPs. In addition, the model faithfully reproduced the delay of activity following stimulation. The present study illustrates that a ten-cell network can reproduce some patterns of activity that underlie feeding in Aplysia. However, the model does not represent all of the identified elements of the CPG. For example, the model is currently be extended to include cells B20 and B65, which can initiate BMPs, in part via their actions on B31/B32. Thus, the model provides a quantitative and modifiable framework with which to investigate how additional elements may contribute to the overall function of the CPG. In addition to providing a tool with which to investigate the CPG, the model can be expanded to include higher-order cells (e.g., the command-like neurons) and modulatory processes. The continual expansion and development of the model will provide a useful tool for analyses of the neuronal mechanisms underlying rhythmic behaviors and their plasticity.