A Golgi-electron microscopic study of goldfish optic tectum. II. Quantitative aspects of synaptic organization

The size, density, and number of the synaptic contacts of three types of interneurons (types I, III, and XIV of Meek and Schellart, '78) and three types of efferent neurons (types VI, XII, and XIII) of the goldfish optic tectum were quantified by means of a quantitative stereological study of Golgi-EM serial sections. Furthermore, an estimation was made of the percentage of optic terminals on these six cell types and of the ratio between terminals with pleomorphic and terminals with round vesicles. The mean density of contacts per receptive component (i.e. the cell body and the different parts of the dendritic tree) varies from 0 to 100 per 100/micrometer2 surface, corresponding to 0-8% receptive surface. Each cell type has a characteristic average density as well as a characteristic density distribution along the distinct components. This suggests that the receptive components of the tectal cell types investigated have a predetermined density and that a morphological classification of tectal cells has functional relevance. The mean length of the contact zones in the ultrathin sections varies from 213 to 332 nm for identified postsynaptic elements and from 188 to 293 nm for identified presynaptic elements. The size of the contacts on the distinct receptive components appears to be primarily related to the tectal lamination pattern. Distinct types of axons, however, have characteristic mean sizes of contacts. This might suggest that the size of the contacts, contrary to their density, is primarily determined by the presynaptic elements. The mean number of synaptic contacts calculated per cell type is as follows: type XIV, 200; type III, 450; type VI, 1,400; type I, 2,100; type XII, 4,200; and type XIII, 5,400. Multiplication of these numbers with the number of cells per tectal half shows that the population of type XIV cells has by far the most synaptic contacts, since their low number of synaptic contacts is clearly overruled by their high frequency of occurrence. Optic terminals, identified by their characteristic mitochondria and large round vesicles, appear to contribute to about 10-20% of the contacts on identified post-synaptic elements in layer 5. The ratio between presynaptic elements with pleomorphic vesicles and those with round vesicles shows a slight tendency to increase when the distance to the origin of the axon decreases. It is concluded that a combination of the Golgi-EM technique with quantitative stereological methods appears well suited to the study of the synaptic organization of brain centers, and that combination of quantitative Golgi-EM with neuronal tracing methods (degeneration, HRP, autoradiography) offers good prospects for detailed investigations of neuronal connectivity.