The effect of neuronal activity on the competitive elimination of neuromuscular junctions in tissue culture.

The number of functional contacts between neurones of explanted ciliary ganglia and cultured muscle cells of the chicken, was determined electrophysiologically. After one week in culture the average muscle cell in the vicinity of a ganglion appeared to have contact with 2.5 neurones. This innervation remained unchanged during the next week. Electrophysiological measurements and a morphological study showed that the initial polyneuronal innervation was not confined to one site on the muscle cell, junctions were often found at a distance of 100 micron from one another. 'Phasic' stimulation (intermittent 2-s trains of 60 stimuli) of the neurones of a ganglion during one day or more caused most muscle cells to become mononeuronally innervated. Some muscle cells lost all functional contacts with the neurones. 'Tonic' stimulation, however (3 or 6 Hz continuously), even during one week and with the same total number of stimuli as the 'phasic' stimulation, produced no alteration in the distribution of the contacts. If two ganglia innervated the same muscle cells, after one week in culture twice as many neuronal contacts were found per muscle cell, equally divided over both ganglia. 'Phasic' stimulation of one of a pair of two ganglia eliminated nearly all functional contacts with the non-stimulated ganglion, whereas the innervation from the active ganglion became mononeuronal on most muscle cells. It is concluded that in tissue culture activity related mechanisms have a decisive influence upon the development of either polyneuronal or mononeuronal innervation of muscle cells. The results demonstrate that the signal for the elimination of supernumerary synapses in tissue culture is mediated by the postsynaptic cell, a mechanism which may also act in vivo. It is suggested that this signal may be the local depolarization around the surviving synapse.