Functional profile of the giant metacerebral neuron of Helix aspersa: temporal and spatial dynamics of electrical activity in situ

The present understanding of how individual nerve cells process information is incomplete. The classical model of the neuron (Rall, 1967) is one based on early studies of vertebrate spinal motoneurons (Coombs et al. 1957; Eccles, 1957; Fatt, 1957a,b) and the formal description developed by Rall using the simplifying assumption that dendrites behave as passive electrical elements. This model is currently being replaced by a more detailed description that involves complex, non-linear, active properties of the dendritic membrane (Wong et al. 1979; Llinas & Sugimori, 1980; Lasser-Ross & Ross, 1992; Stuart & Sakmann, 1994; De Schutter & Bower, 1994; Traub et al. 1994; Spruston et al. 1995; Chen et al. 1997; Cash & Yuste, 1999; Gongyu et al. 1999). An important consequence of active dendrites is that regional electrical properties of branching neuronal processes and the functional organization of individual neurons will be extraordinarily complex, dynamic and, in general, impossible to predict using a model derived in the absence of detailed measurements. To obtain such measurements, one would ideally like to be able to monitor, at multiple sites, subthreshold events as they propagate from the sites of origin on neuronal processes and summate at particular locations to influence action potential initiation. It is important to be able to perform these measurements in at least partially intact neuronal structures (isolated invertebrate ganglia or tissue slices of vertebrate CNS) to ensure that highly specific regional electrical properties of individual neurons (Tauc & Hughes, 1963; Llinas & Sugimori, 1980; Stuart & Sakmann, 1994; Stuart & Hausser, 1994) and characteristic synaptic connections, largely lost in dissociated primary cultures, are preserved. The development of recording methods that would approximate these ideal requirements has been slow. Direct electrical measurements of the detailed spatial distribution and dynamics of voltage transients from neuronal processes are not possible due to size considerations. A true multi-site recording might, however, be achieved by using voltage-sensitive dyes (Cohen & Salzberg, 1978; Wu & Cohen, 1993). Recently, the sensitivity of intracellular voltage-sensitive dye techniques for monitoring voltage transients from neuronal processes in situ, introduced by Grinvald et al. (1987), has been improved 100-fold (Antic & Zecevic, 1995). We have used this approach previously to analyse the metacerebral cell from the terrestrial snail, Helix aspersa, and found multiple spike trigger zones for action potentials evoked by electrical stimulation of the soma. It was possible to determine the precise position of one of these sites (Zecevic, 1996). The metacerebral neuron is a bilaterally paired serotonergic modulatory interneuron whose activity accounts for some aspects of food-induced arousal in Helix, Aplysia and other molluscs (Kupferman & Weiss, 1982; Yeoman et al. 1996). The metacerebral neurons innervate the muscles controlling biting and serve to modulate muscle contractions resulting from the firing of motoneurons during feeding. It has been proposed that the metacerebral cell has gating/enabling function in feeding behaviour because a minimum level of tonic firing was required to support feeding in fine-wire recording experiments in the intact animal (Yeoman et al. 1994). The metacerebral cell fires action potentials in response to food stimuli applied to the lips of the animal. It has been postulated that excitatory inputs to metacerebral cells are mediated through interneurons (Horn et al. 1999). However, the exact nature of these synaptic inputs is unknown. Here, we analysed the initiation and propagation of action potentials evoked synaptically. First, the location of two trigger zones for synaptically evoked spikes was determined. The position of the trigger zones was stable and similar for both synaptically evoked action potentials and spikes elicited by electrical stimulation of the soma. Furthermore, we found that different sets of synaptic inputs activate different trigger zones. Also, we showed that a spike initiated at a remote axonal site did not always invade all parts of the neuron; the conduction of the axonal impulse was regularly blocked at particular locations and the failure of propagation was monitored directly. Finally, the propagating spikes in some axonal branches consistently reversed direction at certain branch points, a phenomenon known as reflection.