Development of terminal arbors of retino-geniculate axons in the kitten--II. Electron microscopical observations

Individual retino-geniculate axon arbors from kittens of 1 to 8 postnatal weeks were densely filled with horseradish peroxidase. The light-microscopical appearance of various types of immature growing tips was correlated with their synaptic relations as seen in the electron-microscope. During the first 2 postnatal weeks, the broad irregular bases of spray-like terminal extensions as well as foliate growth-cone-like terminals contain the round synaptic vesicles and pale mitochondria characteristic of adult retinal terminals. They form elementary glomerular arrangements in which they are the central profile synapsing with dendritic profiles. Many postsynaptic profiles contain large electron-lucent vesicles characteristic of growing neurites, whereas the immature retinal terminations do not. The more numerous finger-like growing tips also contain round synaptic vesicles and make short en passage contacts onto dendrites and outgrowths of perikarya of geniculate cells. The latter type of contact is not seen in the adult. After 3 weeks, retinal terminals are larger and make contacts in distinctly glomerular arrangements. Glial sheaths are evident for the first time. By 6-8 weeks, the terminals which appear crenulated in the light-microscope are always the central profile in a mature glomerulus. Smaller rounder terminals make simple axo-dendritic contacts as in the adult. The few immature terminal structures seen at 8 weeks also contain synaptic vesicles, but their contacts are not adult-like, nor are their synaptic arrangements entirely surrounded by glia. These findings, along with those of the previous paper, demonstrate that (a) many retino-geniculate terminals establish synaptic contacts long before attaining their adult form; (b) during the period optimal for induction of translaminar axon sprouting (1-2 weeks postnatal), only a few immature terminal forms participate in elementary glomeruli; (c) the decreasing malleability of these axons after 3 weeks is accompanied by an increase in crenulated terminals that enter glomeruli surrounded by glial sheaths. It is suggested that ensheathing of synaptic arrangements by astrocytic glia may be one factor which subsequently impedes axon sprouting in older animals.