Abstracts

Rationale: Recently, brain inflammation has been recognized as a key factor in the pathology of various types of epilepsy, including temporal lobe epilepsy (TLE). However, its ambiguous role is not completely understood. The aim of the present study was to determine the spatial and temporal profile of microglial activation and cell loss during the disease ontogenesis and the chronic phase and to link this profile to the occurrence of spontaneous recurrent seizures (SRS) in the kainic acid-induced status epilepticus model (KASE). Methods: SE was induced in male Wistar rats by low-dose subcutaneous injection of kainic acid (n= 29) while controls received saline (n= 21). Different cohorts were sacrificed at 2d, 7d, 2w and 12w after SE (n=6-9 per treatment group). Microglial activation (Ox42) was determined in cortex, hippocampus, amygdala, piriform cortex, cerebellum and pons.Cell loss (cresyl violet) was scored in the hippocampus. Chronic epileptic animals (12w post SE) underwent 100 h of continuous video-EEG to quantify SRS. Results: At 2d post SE, the severity of the SE correlated with cell loss in the hippocampus (P<0.05). Significant hippocampal cell loss was also seen at 7d, 2w and 12w post SE, with a maximum of cell loss at 2w. At all time points investigated, KASE animals displayed Ox42 immunoreactivity. A significant upregulation of activated microglia peaked at the early phase of epileptogenesis (7d & 2w post SE) in the entire hippocampus, amygdala and piriform cortex. The Ox42 immunoreactivity was attenuated by 12w post SE compared to 2w post SE, but still significantly higher than controls in CA1 (P<0.05), the amygdala (P<0.01) and piriform cortex (P<0.05). During the chronic phase the animals showed SRS and interictal spikes. A positive correlation between the latency of the last SRS to sacrification and the Ox42 immunoreactivity score in CA1 was found (P<0.05). Conclusions: Microglial activation is present during both disease ontogenesis and the chronic phase in regions that are important in seizure generation. Firstly,we showed that acute cell loss was a consequence of the primary insult, namely SE. Secondly, we hypothesize that these acute neurodegenerative events activate scanning microglia in the area. In turn, these activated microglia probably attract other microglia from the vicinity to the affected regions with a peak in microglia activation and cell clearance 7d to 2w. This process probably needs time and could therefore explain the delay in microglial activation. During the chronic phase cell loss and microglial activation was less pronounced, which may indicate that once microglia cleared the neuronal debris they returned to inactivated status. Finally, the current Ox42 results showed that more brain inflammation in CA1 was associated with a longer latency since the last seizure, which might suggest a neuroprotective effect of microglia during the chronic epileptic phase. Nevertheless, this immature hypothesis needs confirmation with more data.