CHANGES IN INHIBITORY NEUROTRANSMISSION IN THE CA1 REGION AND DENTATEGYRUS IN A CHRONIC MODEL OF TEMPORAL-LOBE EPILEPSY

1. In this report we compare changes in inhibitory neurotransmission w ithin the CA1 region and the dentate gyrus (DG) in a model of chronic temporal lobe epilepsy (TLE). Extracellular and intracellular recordin gs were obtained in combined hippocampalparahippocampal slices greater than or equal to 1 mo after a period of self-sustaining limbic status epilepticus (SSLSE) induced by continuous hippocampal stimulation. 2. Polysynaptic inhibitory postsynaptic potentials (IPSPs) were induced by positioning electrodes to activate specific afferent pathways and e voking responses in the absence of glutamate receptor antagonists [D(- )-2-amino-5-phosphonovaleric acid (APV) and 6-cyano-7-nitroquinoxaline -2,3-dione (CNQX)]. Polysynaptic IPSPs were evoked in CA1 pyramidal ce lls from electrodes positioned in stratum radiatum and in stratum lacu nosum/moleculare. Polysynaptic IPSPs were evoked in DG granule cells f rom electrodes positioned over the perforant path located in the subic ulum. Monosynaptic IPSPs were induced by positioning electrodes within 200 mu m of the intracellular recording electrode (near site stimulat ion) and stimulating in the presence of APV and CNQX to block ionotrop ic glutamate receptors. Monosynaptic IPSPs were evoked in CA1 pyramida l cells with electrodes positioned in the stratum lacunosum/moleculare and stratum pyramidale. Monosynaptic IPSPs were evoked in DG granule cells with electrodes positioned in the stratum moleculare. 3. Populat ion spike (PS) amplitudes were employed to assure that a full range of stimulus strengths, from subthreshold for action potentials to an int ensity giving maximal-amplitude PSs, was used to elicit polysynaptic I PSPs in CA1 pyramidal cells in both post-SSLSE and control slices. In control tissue, polysynaptic IPSPs were biphasic, composed of early an d late events. In post-SSLSE tissue, polysynaptic IPSPs were markedly diminished. The diminution of polysynaptic IPSPs was detected at all l evels of stimulus intensity. Both early IPSPs [mediated by gamma-amino butyric acid-A (GABA(A)) receptors] and late IPSPs (mediated by GABA(B ) receptors) were diminished. Polysynaptic IPSPs were diminished with both stratum radiatum and with stratum lacunosum/moleculare stimulatio n. 4. Reversal potentials for either polysynaptic early or polysynapti c late IPSPs evoked in CA1 pyramidal cells by stratum radiatum stimula tion were not different in slices from post-SSLSE animals as compared with control animals. Likewise, reversal potentials for either polysyn aptic early or polysynaptic late IPSPs evoked by stratum lacunosum/mol eculare stimulation did not differ in the two groups. These findings e xcluded changes in driving force as an explanation for the diminished amplitude of IPSPs in CA1 pyramidal cells in the post-SSLSE model. 5. Temporal aspects of polysynaptic IPSPs were determined for maximal-amp litude responses evoked in CA1 pyramidal cells in the two study groups . Latencies to peaks of polysynaptic early IPSPs and latencies to peak s of polysynaptic late IPSPs did not differ between the control and po st-SSLSE tissue, for either stratum radiatum or stratum lacunosum/mole culare stimulation. Times of half-decay from peaks of IPSPs in post-SS LSE tissue were significantly shorter than those in control tissue. Th is can be attributed to a preferential deterioration of late, GABA(B)- receptor-mediated LPSPs. 6. Only a late, GABA(B)-receptor-mediated com ponent was detected in polysynaptic IPSPs in DG granule cells, even wi th depolarization of the membrane potential well away from the chlorid e reversal potential, a procedure that enhanced early, GABA-receptor-m ediated IPSPs in CA1 pyramidal cells. In contrast to the situation in CA1, no reduction in the amplitude of polysynaptic IPSPs was found in DG granule cells. In fact, polysynaptic IPSPs elicited in post-SSLSE c ells were larger than those seen in control cells. However, this diffe rence did not reach statistical significance. Times to peak and times for half-decay for IPSPs in control and in post-SSLSE tissue were not different. 7. Reversal potentials for polysynaptic IPSPs in DG granule cells obtained in post-SSLSE tissue were not different from those obt ained in control tissue, both being at the expected potassium reversal potential. 8. Monosynaptic IPSPs evoked in CA1 pyramidal cells in con trol tissue during blockade of ionotropic glutamatergic receptors disp layed early (GABA(A)-receptor mediated) and late (GABA(B)-receptor med iated) IPSPs, both with stratum pyramidale and with stratum lacunosum/ moleculare stimulation. The morphologies of monosynaptic IPSPs evoked in CA1 in post-SSLSE tissue were more variable than those in control t issue. Typically, monosynaptic late LPSPs were absent, although they w ere detected in some cases but were of small amplitude; this was the c ase for both stimulus sites. On the other hand, monosynaptic early IPS Ps in post-SSLSE tissue did not differ appreciably from those in contr ol tissue with respect to amplitude, latency to peak, or reversal pote ntial; this was the case for both stimulus sites. Concomitant with the loss of the monosynaptic late IPSP, times for half-decay from the IPS P peaks were significantly shorter in the post-SSLSE tissue than in co ntrol tissue. 9. Monosynaptic IPSPs evoked in DG granule cells had bot h early and late components. The early component had a reversal potent ial near the expected chloride potential and was blocked by antagonist s effective at GABA(A) receptors. The late component had a reversal po tential near the expected potassium potential and was blocked by antag onists effective at GABA(B) receptors. 10. We conclude that area CA1 a nd DG respond fundamentally differently in the post-SSLSE chronic mode l of TLE with respect to GABAergic inhibition. In area CA1, polysynapt ic IPSPs are reduced or eliminated, whereas they are retained in DG. D ata presented in this report indicate that the defect in GABAergic neu rotransmission in CA1 has two mechanisms. One deficit is a functional ''disconnection'' of inhibitory interneurons from excitatory drive tha t may result from either a defect in excitatory terminals impinging on the interneurons or a reduction in the ability of interneurons themse lves to respond to normal activity of excitatory terminals. This disco nnection appears to extend to at least two types of GABAergic cells in CA1, basket cells and interneurons in stratum lacunosum/moleculare. T he second deficit is a preferential diminution in the potenc