Abstracts

RATIONALE: Stereological methods and markers of apoptosis have provided evidence that repeated brief seizures are sufficient to induce neuronal death (Kotloski et al., Prog Brain Res 135: 95-110, 2002). Apoptotic damage is observed after a single evoked seizure in the dentate gyrus (Bengzon et al., 1997), and is detected throughout the hippocampus after repeated secondary generalized (Class V) seizures (Pretel et al., 1997; Zhang et al., 1998). Cumulative neuronal loss resembling hippocampal sclerosis is eventually observed. If GABA interneurons are among neurons undergoing seizure-induced loss, alterations in the balance of inhibition and excitation induced by poorly controlled seizures could increase susceptibility to additional seizures and potentially contribute to intractability. After 100 evoked Class V seizures, kindled rats develop spontaneous seizures, which are associated with loss of inhibition in the dentate gyrus, as demonstrated by loss of paired pulse inhibition and alterations in the kinetics of evoked granule cell IPSCs (Sayin et al., Soc Neurosci Abstr 24(1): 1205, 1998). In this study, the effect of repeated seizures on interneurons labeled by GAD65, the 65 kD isoform of glutamic acid decarboxylase and synthetic enzyme for GABA, was evaluated after 30 and 100 evoked Class V seizures evoked by kindling. Subclasses of interneurons labeled by GAT-1 (the neuronal GABA transporter) and CCK, which provide axo-axonic and axo-somatic inhibitory inputs, were also evaluated.
METHODS: Rats received olfactory bulb stimulation to evoke kindled seizures using standard techniques. Nonisotopic in situ hybridization with digoxygenin labeled riboprobes for GAD65 were used to label GABA interneurons. Profile counts of GAD65 labeled interneurons per unit area were obtained in the hilus of the dentate gyrus at a standard temporal location along the septotemporal axis, and were compared to batch processed age-matched controls. Conventional immunohistochemistry for GAT-1 and CCK and stereological methods were used to examine these interneuron subclasses.
RESULTS: Profile counts for GAD65 labeled interneurons were decreased by ~25% compared to controls after 30 Class V seizures, and by ~35% after 100 Class V seizures (p=0.0002, ANOVA) which indicated that seizures induce gradually progressive loss of GABA interneurons. After 100 Class V seizures, there was ~33% loss of CCK interneurons and extensive loss of GAT-1 expression in both the hippocampus and dentate gyrus, but loss of CCK or GAT-1 was not detected at earlier stages. Stereological assessment revealed that the CCK subclass was reduced along the entire septotemporal axis of the hippocampus.
CONCLUSIONS: Repeated brief seizures induce gradually progressive loss of GABA interneurons after as few as 30 seizures. The emergence of spontaneous seizures is associated with loss of the CCK and GAT-1 subclasses, which provide axo-axonic and axo-somatic inhibition that is an important control over propagation of synchronous activity. As GAT-1 and CCK subclasses are also reduced in the human epileptic hippocampus, seizure-induced loss of these interneuron subclasses appears to be a critical step in development of intractable temporal lobe epilepsy.
[Supported by: NINDS RO1-25020.]