Abstracts

Temporal lobe epilepsy (TLE) is one of the most common types of epilepsy in adult humans. TLE patients often show a characteristic pattern of neurodegeneration defined as hippocampus sclerosis. The most vulnerable region in the hippocampus seems to be the dentate hilus. The role of this neurodegeneration for the development of epilepsy is still discussed. It is described that the various types of neurons in the hilus differ in their susceptibility to neurodegeneration.
The kindling model is a widely used model of TLE. If the kindling process is extended to more than 150 electrically induced generalized seizures, some of the rats develop spontaneous recurrent seizures (SRS) while in others SRS can not be detected.
We recently showed that the total number of neurons in the hilus in these two subgroups of rats do not differ from each other nor from control rats. This seem to indicate that a loss of neurons in the hilus is not required for the development of SRS (Brandt et al., Epilepsy Res., 62; 135-56: 2004). However, there might be alterations in subpopulations of hilar neurons. In the present study we investigated whether there is a change of parvalbumin- (PV) and somatostatin- (SOM) expressing neurons in the hilus of rats with and without SRS. We therefore electrically stimulated female Wistar rats 235 times (range 174 to 283) over a period of about one year. Rats were video- and EEG-monitored every 8 to 10 weeks for 1 to 4 weeks to check if they developed SRS. At the end of the stimulation period we had a group of rats with SRS and one without SRS. We further had age-matched naive control rats, sham-kindled rats and conventionally-kindled rats. All rats were perfused with 4% paraformaldehyde, the brains were cut on a freezing microtome in 6 series of 40 [mu]m thick slices. The slices were immunhistochemically processed for PV and SOM. The total number of labeled neurons were counted in the hilus by an investigator unaware of the treatment groups. The number of PV-labeled neurons was significantly higher in the hilus of rats with SRS compared to rats without SRS. The increased number of PV-labeled neurons in SRS rats can probably be linked to epileptogenesis and not to the high number of induced seizures since rats without SRS had an equal number of induced seizures. SOM-labeled neurons were significantly increased in the hilus in rats with and without SRS compared to control rats. In this case, the seizure activity itself is probably the cause of the increased number of SOM-labeled neurons. Thus, for both PV- and SOM-labeled GABAergic interneurons an increase and not a loss could be detected in epileptic rats. This increase of GABAergic neurons could be attributed either to the development of new neurons as an attempt to counteract the seizure activity or to an increase of the expression of PV and SOM in hilar neurons.