Isolated neurons from adult central ganglia of the snail Helisoma were cultured in vitro in modified Liebowitz L-15 medium.

We describe a means of visualizing the same neuron in the superior cervical ganglion of young adult mice over intervals of up to 3 months. The dendrites of these neurons change during this interval; some branches retract, others elongate, and still others appear to form de novo. Thus, neuronal dendrites in this part of the nervous system are subject to continual change beyond what is usually considered the developmental period. The remodeling of postsynaptic processes further implies that the synaptic connections made onto these cells undergo substantial rearrangement well into adulthood.

Our previous experiments have suggested the hypothesis that conjoint active neuronal outgrowth may be necessary for formation of new electrical synapses between identified neurons of adult Helisoma buccal ganglia. This growth dependence hypothesis now has been tested by examining the responses of individual pairs of neurons in isolation from the influences of the ganglionic environment. Isolated cell culture of identified neurons (neuron 5) showed that: (i) neurons growing in cell culture undergo a predictable sequence of morphological changes culminating in a stable morphological state (i.e., growth stops); (ii) contact between actively growing neurons in cell culture results in the formation of electrical connections, just as in ganglia; and (iii) when an actively growing neuron encounters a neuron that is morphologically stable, electrical connections do not form or are very weak, even though strong connections are made between pairs of actively growing neurons in the same culture. These results establish that processes closely associated with growth are required for formation of electrical synapses between these neurons.

The electrically coupled buccal ganglion neurons 4R and 4L of the snail, Helisoma, display predictable plasticity. The strength of the electronic synapse between them increases significantly following axotomy. Synaptic strength was assayed by measurements of electrical coupling coefficients and by assessment of dye coupling (passage of dye into the uninjected neuron) following injection of Lucifer Yellow CH into one neuron. Within 3 to 5 days, axotomy induced an increase in electrical coupling coefficients between neurons 4R and 4L from 0.54 +/- 0.11 (n = 13) in normal preparations to 0.72 +/- 0.14 (n = 24). A parallel axotomy-induced increase in the probability of dye coupling occurred. Only 27% (n = 27) of normal neuron 4 pairs were dye coupled, compared with 87% (n = 15) of axotomized neuronal pairs. Irrespective of treatment, electrical and dye measurements in the same neuron 4 pairs showed a consistent correlation between the magnitude of the electrical coupling coefficients and the probability of detectable dye coupling. No dye coupling was observed in neuronal pairs with electrical coupling coefficients less than 0.50. Dye coupling always occurred when coupling coefficients were greater than 0.70. Sixty-seven percent of neuronal pairs with intermediate coupling (0.50 to 0.70) coefficients displayed dye coupling. The results show that axotomy evokes a predictable enhancement of communication at an "identified" electronic synapse and suggests that electrical and dye coupling are mediated by similar mechanisms.

Identified neurons of the snail, Helisoma, undergo extensive remodeling in response to axotomy, including the formation of specific sets of novel electrical connections. This communication addresses the question of why, under the conditions employed, some neurons readily form new connections with a single "test" neurons, whereas others do not. The present experiments are a test of the hypothesis that, for these adult neurons, competence to form electrical connections is restricted to pairs of neurons with interacting regions of active outgrowth. Morphological observations demonstrated profuse overlapping outgrowth from neurons which formed electrical connections, whereas neurons which did not connect displayed no simultaneous new outgrowth, although there could be regions of physical overlap or proximity. The causal relationship between growth and the ability to form new connections was tested more directly by two means: (1) Previously nonconnecting neurons were recruited into the connectivity pattern by axotomy-induced growth. (2) Previously connecting neurons did not connect when they were not induced to grow. Thus, growth or lack of growth is an effective discriminator for determining specific sets of interconnected neurons.