Electrophysiological studies of mouse cerebellum neurons in culture.

Investigation of synaptic connections in a group of neurons in culture from the peduncular region of the pontine tegmentum (hereinafter referred to as brain stem (BS) neurons) was carried out by means of two simultaneous single unit recordings. Primarily extracellular records were obtained, and correlations between the spike trains were determined by cross correlation methods (Perkel et al, 1967b) as well as various modifications of time series analysis adapted to the BS culture system. Correlations obtained between the spike trains were then reformulated in terms of possible synaptic mechanisms which could give the observed correlation. This resulted in a multiplicity of possible explanations, and thus a smaller number of simultaneous extracellular/intracellular recordings of two different neurons were made so that synaptic activity could be directly observed rather than inferred as from the purely extracellular records. Because of the low rates of spontaneous activity present in the BS neurons, the majority of cell pairs were recorded using double barreled electrodes from which iontophoretically applied glutamate (glu) could be utilized to increase the rate of activity of the neurons. Most BS neurons were consistently excited by glu, although some anomalous responses to glu were observed. A large degree of correlated activity was observed in the recorded neurones. Of 92 extracellular cell pairs, 64% exhibited a definite correlation, 12% could possibly have had an interaction, but were ambiguous, and only 24% were definitely non-correlated. Of the 59 correlated extracellular cell pairs, three principal interaction categories could be distinguished: correlated bursting, directional excitation, and a small number of miscellaneous interactions. Correlated bursting was the numerically greatest interaction with 46 (78%) of the correlated pairs exhibiting some form of correlated bursting. Directional excitation was observed in 5 (8%) of the cell pairs, and may represent a direct excitatory synaptic connection from one neuron to the other, although conclusive evidence of this was not obtained in any of the-extracellular data. 15 good cell pairs comprised the intracellular data, and here again correlated bursting was the largest interaction type (4 cases). Intracellular records revealed that common bursting was due to the occurrence of a large depolarization of variable size and duration. The variability from one intracellular recorded burst to another implied that the large depolarization was the sum of numerous nearly synchronous EPSP's. In one extra/intra cell pair, a large consistently evoked EPSP (from the extracellular cell) was observed. When other intracellular records were analyzed by the techniques of extracellular spike triggered intracellular averaging, an additional 2 EPSP's were observed as well as 3 nondirectional interactions which consisted of a depolarizing potential extending backwards and forwards from the time of the extracellular spike. The intracellular records provided evidence for predominantly excitatory interactions among the BS neurons, although two cases of common inhibition were also found in the intracellular data. Intracellular recordings revealed a more extensive interaction among the neurons than shown in the extracellular data. With the excitatory nature of the BS to BS connections established, the data were examined in order to see if the type of culture had any influence on the interactions. Two types of BS cultures may be distinguished; BS - Cx cultures in which cerebellar cortex and deep cerebellar nuclei are included, and pure BS cultures in which almost all of the cerebellar elements have been eliminated. No difference could be found between the frequency of occurrence of the various interaction types and the type of culture. Only when the cell pairs were grouped by the electrode-configuration used; one glu and one single barrelled, or two glu electrodes, was a difference apparent. The interactions in the cultures recorded with 2 glu electrodes were almost exclusively (92%) correlated bursting, whereas this category made up only 57% of interactions in cell pairs recorded using one glu electrode. A model which has been proposed to explain this data is that the BS neurons consist of a network of interconnected neurons with excitatory interactions in which most synapses are weak. When glu is applied at one site, a portion of the network is activated, but the activity rate is not sufficiently high to obscure interactions (non-bursting type) which may be occurring with the cell on the other electrode. With two glu electrodes, a greater number of neurons are activated, and when proper synchrony of spikes occurs, unlimited positive feedback causes synchronous firing in all cells in the network until accumulated refractoriness ends the cycle. Such an event would correspond to a correlated burst