In experimental animal models and in human surgical tissue from intractable epileptics, there is often sclerosis of the hippocampus. Here we demonstrate in a slice model, a semiquantitative method to compare the damage induced by three agents reported to induce seizure-like activity: bicuculline (Bic), a GABA_A receptor antagonist; 3-nitropropionic acid (3-NPA), an inhibitor of succinate dehydrogenase; and kainic acid (KA), a known excitotoxin.

Hippocampal slices (250 µm) from humanely killed neonate Wistar rats (P7) were maintained in vitro for 12-14 days before use (Stoppini et al. 1991). Slices were exposed to Bic (10 µM), 3-NPA (100 µM), or KA (25 µM) for 4 h, followed by 15 h recovery. A dual-stain fluorescent assay was used to assess viable (calcein-AM, 4 µM) and damaged cells (propidium iodide, 20 µM). Confocal images were collected (Bio-Rad) from four randomly selected areas within hippocampal subregions CA1 and CA3, using a X 60 water-immersion objective. Fluorescence area was determined using Simple-PCI (C-Imaging). All values are means ± S.E.M. from slices from four separate animals.

All three agents caused varying degrees of cell death with clear differences between hippocampal subregions. Slices exposed to Bic and 3-NPA showed similar patterns of damage: CA1 was 5-6 times more vulnerable than CA3. However, 3-NPA caused approximately twice the damage in both subregions compared with Bic (543.6 ± 148.8 and 108.6 ± 37.2 µm² with Bic, compared with 1218.5 ± 370.4 and 201.4 ± 89.7 µm² with 3-NPA in CA1 and CA3, respectively). In contrast, KA caused the most extensive neuronal damage in both CA1 and CA3 (2644.7 ± 154.6 and 2560.9 ± 191.6 µm², respectively, P < 0.001, n = 4, ANOVA). Regions from control slices showed minimal cell damage (47.5 ± 15.2 and 42.4 ± 13.2 µm² for CA1 and CA3).

Neither Bic nor 3-NPA altered the area of total calcein staining compared to controls. However, an increase in the number of isolated sites where neurones were no longer present (or ‘holes’), was seen in Bic-treated (2.0 ± 0.7 and 0.8 ± 0.6 for CA1 and CA3, respectively) and 3-NPA-treated slices (4.6 ± 1.1 and 1.3 ± 0.9 for CA1 and CA3) when compared with controls (0.3 ± 0.3 and 0.2 ± 0.1 for CA1 and CA3; P < 0.01, n = 4, ANOVA). KA, whilst further increasing the ‘holes’ (9.2 ± 2.2 and 6.7 ± 1.4 for CA1 and CA3), also resulted in increased calcein-stained area from swollen but surviving neurones (P < 0.05, n = 4, ANOVA).

This technique has identified a spectrum of hippocampal damage, ranging from mild to severe, at an early time point following treatment, with subregion differences. Our approach provides ways of assessing subtle neuronal damage that may be relevant to the development of intractable epilepsy, associated with hippocampal sclerosis.