Abstracts

Rationale: In animal models of status epilepticus (SE) induced by the chemoconvulsants pilocarpine and kainic acid, the administration of human recombinant erythropoietin (rhEPO) has a neuroprotective effect and reduces seizures. However, due to the inherent variability associated with the use of chemoconvulsants, the true extent of this neuroprotection has been difficult to assess. An animal model of controlled perforant pathway stimulation in awake-rats was previously shown to reliably produce neuronal injury and hippocampal-onset epilepsy, with minimal mortality, more restricted neuronal injury, and less variability between animals. Thus, utilizing this improved model of SE, we studied the neuroprotective pattern of rhEPO. We hypothesized that rhEPO will reduce neuronal injury in this model, and mediate such reduction through altering inflammatory responses subsequent to seizures.Methods: Male Sprague Dawley rats (400-450g) were implanted with bilateral stimulating and recording electrodes in the angular bundle of the perforant pathway and granule cell layers, respectively. After a 2-week recovery period from microelectrode implantation, SE was induced by bilateral electrical stimulation of the perforant pathway for 3 hours. After 3 hours of confirmed hippocampal and behavioral SE, isoflurane and a sub-anesthetic dose of urethane (0.8 g/kg s.c.) was administered to terminate and prevent the reoccurrence of SE. Stimulated animals were divided into two experimental groups rhEPO-treated (n=8) and untreated (n=6). Treated animals received two different dosing paradigms. In the first paradigm, rats (n=3) received a dose of rhEPO (5000 IU/kg, i.p.) on 3 consecutive days. In the second paradigm, rats (n=5) received a total of 7 doses. On day 8 post-stimulation, animals were perfused with 4% paraformaldehyde. The fixed brains were sectioned and stained with cresyl violet, Fluoro-Jade B, anti-GFAP, anti-NG2, and anti-NeuN.Results: Nissl and FluoroJade B staining revealed reduced neuronal loss and injury in the hilus and CA3 field of the hippocampus of treated animals compared to the untreated stimulated animals. Anti-NeuN confirmed the pattern of neuroprotection in treated rats. Anti-NG2 and anti-GFAP staining showed a correlation of GFAP- and NG2-immunopositive cells with regions of neuronal loss. Microarray analysis of gene expression suggests a decrease in expression of inflammatory factors in treated animals.Conclusions: Our data suggest that therapeutically administered rhEPO has a neuroprotective effect in the hippocampus after SE in this model. It is likely that the decreased expression of inflammatory factors in treated animals is associated with this neuroprotection. The increased expression of GFAP-positive cells in injured areas may be a source of such factors. Expression of NG2-immunopositive cells, which are a population of excitable cells, may be responsible for the long-term development of recurrent seizure activity in this model.