Entorhinal cortex-subiculum interactions in an experimental model of mesial temporal lobe epilepsy

Mesial temporal lobe epilepsy (MTLE) patients present with seizures involving the limbic system and with a pattern of brain damage characterized by neuronal loss in CA1/CA3 areas, dentate hilus, and entorhinal cortex (EC), layer III (Houser CR. Adv Neurol 1999;79:743–61). Similar findings are seen in laboratory animals following pilocarpine injection (Turski WA, et al. Behav Brain Res 1983;9:315–35). This procedure induces an initial convulsive response, which is followed within 2–3 weeks by recurrent seizures. Limbic network hyperexcitability in MTLE and in animal models results from seizure-induced brain damage leading to (a) synaptic reorganization (Cavazos JE, et al. J Neurosci 1991;11:2795– 803; Houser CR. Adv Neurol 1999;79:743–61) and (b) changes inGABA receptor–mediated inhibition (Buhl EH, et al. Science 1996;271:369–7; Doherty J, Dingledine R. J Neurosci 2001;21:2048–57. However, it is unclear how these changes lead to a chronic epileptic condition. CA3-driven interictal activity induced in normal brain tissue by epileptogenic stimuli inhibits the EC from generating ictal discharges (Barbarosie M, Avoli M. J Neurosci 1997;17:9308–14), suggesting that CA3 damage causes a decrease of hippocampal output activity that would release EC ictogenesis and establish a chronic epileptic condition. Accordingly, slices obtained from pilocarpine-treated epileptic mice respond to 4-aminopyridine (4AP) application by generating (a) CA3- driven interictal activity that is less frequent than in nonepileptic control (NEC) tissue, and (b) ictal discharges that do not disappear over time and propagate to the CA1-subiculum via the temporoammonic path (D’Antuono M, et al. J Neurophysiol 2002;87:634–9). From these findings, we predicted that limbic seizures result from EC–subiculum interactions. Using brain slices obtained from pilocarpine-treated, epileptic rats, we found that decreased CA3 output function, along with reverberation between EC and subiculum networks, lead to in vitro epileptogenesis. First, intense activation of EC and subiculum was identified with intrinsic optical signal (IOS) recordings in pilocarpine-treated, but not in NEC slices. Second, using field potential recordings during 4AP application, we established that CA3-driven interictal activity occurs at lower frequency in pilocarpine-treated slices and that disconnection of the EC from the subiculum attenuates 4AP-induced ictal discharges in pilocarpine-treated, but not in NEC slices. Third, the distribution of FosB/FosB-related proteins in epileptic tissue demonstrated distinct patterns overlapping those seen with IOS recordings, with the highest intensity in layer III of the lateral EC. In conclusion, our data show that hippocampal damage in epileptic rats, and perhaps in MTLE patients, hampers the ability of CA3 output activity to control ictogenesis in the EC. Such a process is reinforced by interactions between subiculum and EC networks.