Abstracts

Rationale: Temporal lobe epilepsy is common and difficult to treat effectively. Seizures initiate in mesial temporal lobe structures, which include dentate gyrus. Dentate granule cells receive excitatory synaptic input from layer II neurons in entorhinal cortex and hilar mossy cells. Both populations are reduced in epileptic tissue. However, it remains unclear how excitatory synaptic input to granule cells changes during epileptogenesis. Methods: Male rats (33-64 d old) were treated systemically with pilocarpine. Status epilepticus was suppressed with diazepam after 2 h. Rats were perfused 5 d post-status (n = 3) or later (123 ± 27 d post-status) after development of chronic, spontaneous, recurrent seizures verified by video-monitoring (n = 5). Controls (n = 6) included age-matched naive and pilocarpine-treated rats that did not experience status. An electron microscopy-adapted optical fractionator method was used to estimate total number of excitatory synapses in dentate gyrus molecular layer. A random and systematic 1-in-12 series of transverse hippocampal sections was processed for EM with post-embedding GABA-immunoreactivity. Entorhinal cortical neurons synapse with granule cells in the outer 2/3rds of the molecular layer; whereas, mossy cells synapse in the inner 1/3. Areas of these regions were measured using Neurolucida. In each hippocampus, 6 sample sites were randomly and systematically chosen in each region. Each sample site consisted of 100 serial 50-nm-thick sections in which an aligned counting frame (3.7 x 5.1 μm) was analyzed at 48,400X final magnification. Excitatory synapses were identified by thick, asymmetric postsynaptic densities and concentration of synaptic vesicles with lack of GABA-immunoreactivity in presynaptic boutons. Synapses were counted, postsynaptic targets identified, synapse lengths measured, and synapse areas calculated. Results: In both inner and outer molecular layer, total number of excitatory synapses per hippocampus was reduced 5 d post-status but recovered substantially in chronically epileptic animals. In inner molecular layer, excitatory synapses were reduced to 33% of controls 5 d post-status and then recovered to 87% of controls in chronically epileptic rats. In outer molecular layer, excitatory synapses were reduced to 49% and then recovered to 92%. In both regions, the vast majority of excitatory synapses were with GABA-negative spines. At 5 d post-status, however, there was a shift toward more excitatory synapses with dendritic shafts, which continued to a smaller but significant extent in chronically epileptic rats. Perforated synapses became significantly more prevalent in the inner molecular layer of chronically epileptic rats. Excitatory synapse area increased progressively and significantly in both regions. Conclusions: These findings suggest excitatory synaptic input from entorhinal cortex and mossy cells to granule cells decreases dramatically after an epileptogenic injury. Over time, most lost excitatory inputs are replaced by larger and presumably stronger synapses, which might contribute to epileptogenesis.