Distributed processing by visual interneurons of crayfish brain. I. Response characteristics and synaptic interactions.

1. The visual responses and synaptic interactions of a small population of crayfish interneurons are described. 2. The discharge of optic nerve sustaining fibers (tonic on-cells) appears in the brain prior to the onset of the light-evoked discharge of any of the higher order, descending visual interneurons. Direct depolarization of impaled sustaining fibers elicits impulse responses in a large number of descending interneurons. These results indicate that the sustaining fibers provide the visual input to higher order interneurons. 3. Four classes of descending interneurons can be distinguished. All arise in the brain and have axons in the circumesophageal connectives. The response forms vary from tonic to phasic. Two classes of tonic cells are distinguished by response latency and two classes of phasic neurons are distinguished by the rate of response adaptation. The phasic neurons exhibit the most rapid habituation, the largest receptive fields, and the most potent nonvisual inputs. 4. Synaptic interactions are studied by cross-correlation of impulse trains and direct observation of synaptic potentials. About 84% of the cells examined reveal evidence of functional connections to other descending visual interneurons. 5. Cross-correlograms derived from impulses of parallel interneurons exhibit a mean time lage to peak of 6.6 +/- 2.8 ms (SD). The measured delay from EPSP onset to spike onset is 6.0 +/- 4.0 ms. Thus a substantial proportion of the correlogram's time lag to peak is associated with postsynaptic integration time. 6. Direct depolarization of impaled tonic on-cells elicits impulse activity at a fixed delay in other descending interneurons. 7. Synaptic potentials in descending visual interneurons are correlated 1:1 with axon spikes of other descending interneurons. 8. A third of the 80 interactions examined were reciprocal and many cells were implicated in multiple interactions. 9. The results suggest that the descending visual interneurons are organized in a complex network, which can cordinate the discharge of various subpopulations of the ensemble. It is proposed that the coordination of impulses in parallel interneurons may be a mechanism for coding and information transfer in the crayfish nervous system.

Distributed processing by visual interneurons of crayfish brain. II. Network organization and stimulus modulation of synaptic efficacy.

1. Multiple interactions were examined between five or six visual neurons simultaneously monitored in the circumesophageal connective. 2. A single neuron can make divergent connections to at least five other visual interneurons. 3. Conversely, a single cell may receive convergent inputs from up to four visual interneurons. 4. The convergent interactions are sufficiently intense so that 80--90% of a postsynaptic cell's visual activity can be attributed to observed network interactions. 5. Connectivity diagrams suggest that the descending interneurons, which arise in the visual neuropil of the brain, are organized into three interconnected layers: a) neurons that receive input from the optic nerve and project to other visual interneurons, b) neurons that both receive input and project to other descending interneurons in the brain--these cells exhibit a preponderance of reciprocal interactions, c) neurons that receive input from both the first and second network layers and project exclusively to the more caudal ganglia of the ventral nerve cord. 6. The network is systematically organized with respect to visual and nonvisual responsiveness. The cells of the first layer exhibit the strongest visual responses. The cells of the third layer exhibit spontaneous activity and the strongest tactile and/or proprioceptive responses. 7. The intensity of the network interactions is under stimulus control. The synaptic efficacy of a presynaptic spike can vary by over 100-fold as a consequence of stimulus presentation and/or location. The expressed organization of the network thus exhibits a dynamic, stimulus-dependent, plasticity. 8. The results indicate that the descending visual interneurons of the brain rather than forming a parallel tract actually constitute a complex distributed network. Furthermore, the results indicate the feasibility of population neural coding based on stimulus-dependent inpulse coordination in an array of neurons.