No abstract

Against a background of existing models relating perforated synapses to synaptic plasticity, the numerical density and frequency of perforated synapses in rat neocortex have been assessed from 1 d to 22 mo of age using the disector procedure, and changes in their morphology were assessed using 3-D computer reconstructions. The data point toward perforated and nonperforated synapses being separate synaptic populations from early in development, and with perforated synapses playing a part in the maintenance of neuronal postsynaptic density surface area from mid-adulthood onwards. This suggests that they play a crucial role in synaptic plasticity, although its nature may be different from that postulated by most recent workers.

Over recent years attention has been focused on perforated synapses (PSs), on account of their possible involvement in synaptic plasticity in the nervous system. The mechanism invoked has involved the conversion of conventional (non-perforated) synapses into perforated ones, with the latter subsequently splitting to form new daughter synapses. Data challenging this mechanism have emerged from studies using unbiased stereological and three-dimensional reconstruction procedures in early synaptogenesis and also in early- to mid-adulthood. The present study is an attempt to complement these earlier studies by concentrating on the latter part of synaptogenesis and early adulthood in rats (7-60 days postnatal). Using the disector method, the frequency of perforated synapses increases with age from 12% at seven days to 33% at 60 days. Reconstruction of the perforated synapses has led to the description of three sub-types, while quantitation of the reconstructions has highlighted major differences between perforated and non-perforated synapses over this developmental period. For instance, the postsynaptic density of perforated synapses increases in size with age, but remains static in the non-perforated variety; in addition, it is 2-3 times larger in perforated synapses. The proportion of the synaptic contact zone occupied by the postsynaptic density increases in perforated synapses with increasing age, but is static in non-perforated synapses. No evidence has been found to support a perforated synapse splitting model, since perforated synapses are present from early in synaptogenesis, the frequency of non-perforated synapses reaches a peak prior to, but not following, that of perforated synapses, and the size of the postsynaptic density of non-perforated synapses remains constant throughout the period of the study.

It has been known for some time that perforated synapses increase in number and size with increasing age. Initially, these trends were used to support the concept that nonperforated synapses enlarge until an optimal size is reached, at which point they perforate and may subsequently split. More recent stereological and three-dimensional reconstruction investigations, however, suggest that this may not be the case and that perforated and nonperforated synapses constitute separate synaptic populations. In order to test the separate population hypothesis, synapses have been studied ultrastructurally in the parietal cortex of rats aged 19 and 20 days gestation, and 1 and 4 days postnatal. By examining synapses serially, and also by studying three-dimensional reconstructions, it has been demonstrated that perforated synapses are present at each of these ages. Some are relatively simple in organization, resembling previously described perforated synapses at 14 days of age, although others appear to consist of two or more separate PSD components. These findings demonstrate that perforated synapses are present from early on in synaptogenesis and that developing perforated synapses may have distinct characteristics that cast light on their developmental course.