The entorhinal cortex (EC) is an area of the temporal lobe that is particularly susceptible to epileptogenesis. We have begun to study changes in synaptic function in the EC which accompany induction of a chronic epileptic state. In the present experiments we have analysed the characteristics of spontaneous inhibitory (sIPSCs) and excitatory (sEPSCs) postsynaptic currents recorded under voltage-clamp conditions in deep and superficial layers of the EC.

Male Wistar rats (100 g) were injected with a mixture of pilocarpine and picrotoxin (150 and 1.5 mg kg^-1 I.P., respectively). Over a period of 60 min, they developed a spectrum of epileptiform behaviours leading to spontaneous tonic-clonic seizure episodes that were allowed to continue uninterrupted for 2 h before being terminated with midazolam (5 mg kg^-1 I.P.). Following a 6-8 week silent period, rats began to exhibit spontaneous recurrent seizures and characteristic epileptiform behaviours (wet-dog shakes, nodding, excessive grooming). At this point, animals were deeply anaesthetised with ketamine/xylazine (8 and 120 mg kg^-1 I.M., respectively) prior to decapitation and removal of the brain. Slices (400 µm) containing the EC were prepared from rats demonstrating epileptiform behaviours, and from non-treated control animals. sPSCs were recorded from visually identified neurones in layers II and V using the whole-cell patch-clamp technique, with recording conditions appropriate for monitoring either sEPSCs or sIPSCs. Measurements (200 PSCs from each cell) were pooled and plotted as cumulative probability curves and were statistically validated using the Kolgomorov-Smirnov (KS) test at the 95 % confidence level (P < 0.05). All mean values quoted are ± S.E.M.

sEPSCs had a similar mean amplitude, 10-90 % rise time, and event decay time in layer V neurones in control and epileptic animals. However, in slices from epileptic rats the mean frequency was increased from 2.54 ± 0.11 to 8.00 ± 0.29 Hz (P < 0.05, KS, n = 11). In layer II, sEPSCs again showed little change in amplitude or kinetics in epileptic animals, but mean frequency increased slightly from 2.74 ± 0.08 to 4.13 ± 0.22 Hz (P < 0.05, KS, n = 3). In the case of sIPSCs, increases in frequency in the two layers were also detected, but these were accompanied by altered decay kinetics with no change in rise time or amplitude. In layer II, mean sIPSC frequency was 22.01 ± 0.54 Hz in epileptic animals and 10.42 ± 0.07 Hz in controls (P < 0.05, KS, n = 8), and the mean decay time decreased from 6.94 ± 0.17 to 4.57 ± 0.05 ms (P < 0.05, KS). In layer V there was a modest increase in mean sIPSC frequency (control 3.09 ± 0.19 Hz, epileptic 4.44 ± 0.32 Hz, P < 0.05, KS, n = 9), and again, mean decay time was reduced from 9.99 ± 0.17 to 7.85 ± 0.11 ms (P < 0.05, KS).

These data suggest that in this model chronic epileptogenesis is accompanied by alterations in both spontaneous excitatory and inhibitory synaptic transmission in the entorhinal cortex, with the more pronounced effects at excitatory synapses in layer V, and inhibitory synapses in layer II. These changes may alter the focus of epileptogenesis in the EC, and have implications for spread of activity to other areas of the temporal lobe.

We thank The Wellcome Trust, the Taberner Trust and the MRC for financial support.