Transient and enduring morphological correlates of synaptic activity and efficacy change in the rat hippocampal slice.

This study examined anatomical correlates of: (1) long-term potentiation (LTP); (2) equivalent low frequency synaptic activity; (3) continuous high frequency synaptic activation which did not produce LTP; and (4) synaptic inactivation by high Mg2+/low Ca2+ incubation in hippocampal subfield CA1 in the in vitro slice, and examined the persistence of changes at 10-15 min, 2 h and 8 h after stimulation. After potentiating stimulation (6 trains at 100 Hz for 1 s or 200 Hz for 0.5 s), compared to an equivalent number of low frequency stimuli (1 Hz for 600 s), there were increases in numbers of shaft and sessile spine synapses (synapses on stubby, headless spines). This suggested an increase in the number of shaft synapses onto inhibitory interneurons and/or an enhancement of synapse formation on pyramidal neurons possibly involving initial formation of shaft synapses and a transition from shaft, to sessile spine, to full grown spine synapses. Postsynaptic spine heads also assumed a rounder shape, as indicated by decreases in spine perimeter to area ratios, contact lengths, and the percentage of 'cup' shaped spines. There was no effect of potentiating stimulation on bouton or spine areas. After continuous high frequency synaptic activation (40 Hz or 100 Hz for 10 min), which produced no apparent LTP, there were no changes in synapse numbers or spine head shape parameters. However, in contrast to effects of LTP, there was an increase in bouton mitochondrial area and a marginal increase in bouton area compared to the low frequency condition. Inactivation did not affect any of these measures. LTP-associated increases in numbers of shaft and sessile spine synapses persisted over an 8 h incubation period, while the effect on spine shape disappeared after 2 h. Physiologically-demonstrable LTP persisted over the 8 h period. Effects of continuous high level activation on mitochondrial and bouton areas were even more transient, disappearing 2 h after stimulation. These findings: (1) confirm previously reported effects of potentiating stimulation on synapse numbers and spine shape; (2) indicate that spine shape changes are not necessary for the maintenance of LTP; and (3) indicate that continuous high frequency activation which does not produce potentiation has different and non-persisting effects from potentiating stimulation.