1. Central neuronal mechanisms of siphon withdrawal in Aplysia were studied for the first time in intact, freely behaving animals by means of population recordings from implanted whole-nerve cuff electrodes. Intracellular follow-up studies were then conducted when the same animal was reduced to a semi-intact preparation. 2. Background spontaneous activity in the siphon nerve consisted of low-frequency firing of a population of efferent units containing identified siphon motoneurons. 3. Spontaneous patterned bursts of efferent activity occurred irregularly and were associated with all-or-nothing contractions of the parapodia, gill, and siphon. Spontaneous bursts were due to centrally generated activity in the interneuron II (INT II) network, an oscillatory network with endogenous pacemaker properties. 4. In intact animals, even weak tactile stimuli to the siphon typically triggered an INTII burst shortly after the stimulus-locked efferent activity. Thus, the stimulus can phase-advance the INT II oscillator. In semi-intact preparations, short-latency INT II bursts were triggered less less frequently and required more intense stimuli. 5. With weak to moderate-intensity stimuli in intact animals, the presence of short-latency triggered INT II bursts largely determined the duration of the siphon component and amplitude of the gill component of the withdrawal reflex. 6. When stimuli were repeated over a range of interstimulus intervals (from 60 to 1 min), the likelihood of triggering a short-latency INT II burst die not change systematically. Thus, the ability of the siphon stimulus to stably entrain the all-or-none INT II component over a wide range of intervals will interact behaviorally with the decrement of the monosynaptic component of the reflex with repetition.

No abstract

No abstract

No abstract

Two neurons with cell bodies symmetrically located in the abdominal ganglion and giant axons in the left (L1) and right (R1) pleurovisceral connectives of Aplysia californica were examined in vivo and in vitro. Direct stimulation of R1 and L1 in the intact animal does not elicit any observable behavior, suggesting that they are neither motoneurons nor command neurons. These cells respond in vivo to sudden onset mechanical stimulation of widespread regions of the body. R1 and L1 spikes are initiated in at least three different loci: (1) the peripheral axon in the foot, (2) the neuropil of the pleural and/or pedal ganglion, and (3) the neuropil of the abdominal ganglion. Furthermore, R1 and L1 probably have two different mechanisms for spike initiation: (1) sensory (foot), and (2) synaptic (abdominal and/or head ganglia). The different loci for spike initiation account for the bidirectional conduction of R1 and L1 spikes. As sensory (mechanoreceptor) neurons, R1 and L1 have peripheral axons in the ipsilateral posterior pedal nerve, show low threshold responses to stimulation of the ipsilateral posterior foot, they are rapidly adapting their responses do not decrease with repetition, and they are not blocked by high Mg++/low Ca++ solutions. As synaptically-driven neurons, R1 and L1 have widespread bilateral responsiveness, their responses decrease with repetition and their inputs are blocked with high Mg++/low Ca++ solutions. These neurons integrate sensory and synaptic inputs and conduct bidirectionally, however, their output connections must be specified before their behavioral function can be understood.