Early postnatal development of the monkey visual system. II. Elimination of retinogeniculate synapses

Profiles of retinal terminals, and of their synaptic and non-synaptic contacts, were measured in electron micrographs from magnocellular and parvocellular laminae of the dorsal lateral geniculate nucleus (LGNd) in newborn, 1-,4-,8- and 17-week-old rhesus monkeys. Morphologic criteria, i.e., the presence of pale mitochondria and large round vesicles, were used to identify the profile of retinal origin. Size-frequency histograms were stereologically reconstructed and used to calculate the density of retinal boutons and synaptic and non-synaptic plaques. The density values were adjusted for laminar growth to yield estimates of total numbers of these elements. Numerical estimates indicate bouton proliferation during the first week, followed by substantial reductions in bouton number accompanied by profound decreases in synapse number and cumulative synaptic area. In magnocellular layers, the reduction in synapse number is more pronounced after the eighth week, whereas the decrements in both features in the parvocellular laminae occur before this time. This synapse elimination process may be due entirely to retinal bouton retraction in parvocellular layers, but involves additional retinal synapse loss in the magnocellular segment. The parvocellular division shows a further size contraction of the remaining synapses. Immature synapses predominate in the LGNd throughout the 4-month period, and no quantitative evidence for direct transformation of immature to mature contacts is obtained. Non-synaptic junctions are stable in number but of increasing size in magnocellular layers, whereas substantial increases in number and area are found in parvocellular laminae. The preceding modifications in synaptic organization of the monkey LGNd occurring during the initial postnatal period may provide morphologic bases for the physiological and behavioral changes observed in this species during the same interval. Our data underscore the conclusion that synaptic reorganization occurs over a prolonged period, probably extending beyond 4 months, and involving the process of synapse elimination.