Abstracts

Corticosteroids are frequently administered to pregnant women during the last third of pregnancy to decrease mortality of prematurely born neonates. In humans, weeks 24-34 of pregnancy coincide with the period of development of the hippocampus, a structure sensitive to corticosteroid effects. Thus, exposure to corticosteroids during this period may reprogram the hippocampus in terms of structural changes leading ofto postnatal functional alterations. As the hippocampus is one of the brain regions critically involved in seizure initiation, this may lead to changes in seizure susceptibility. Pregnant rats were injected with two doses of corticosteroids i.p. on day 15 of gestation (E15), the period of major development of CA hippocampal zones. Synthetic betamethasone was administered because this is the most frequently used corticosteroid in pregnant women. Additional pregnant rats were injected with hydrocortisone to create stress-like corticosteroid surge. Control pregnant rats received equivalent volumes of saline. Offspring were marked for identification and litters of corticosteroid-treated and control females were culled to ten. On postnatal day 15 (P15), offspring were tested for seizure susceptibility in the limbic system using different seizure models (kainic acid, N-methyl-D-aspartic acid, and dentate gyrus kindling). We found that in P15 rats treated with hydrocortisone on E15, automatisms and status epilepticus induced by kainic acid occurred significantly earlier than in P15 controls, i.e., prenatal hydrocortisone had proconvulsant features. In contrast, there was no difference in susceptibility to kainic acid-induced automatisms and status epilepticus between the rats prenatally treated with betamethasone and controls. Similarly, prenatal betamethasone treatment had no effect on susceptibility to NMDA seizures and in fact, decreased susceptibility to dentate gyrus kindling in P15-16 rats. The data indicate that prenatal corticosteroids may have differential effects on postnatal seizure susceptibility. Natural corticosteroids had proconvulsant effects consistent with a previous report showing proconvulsant effects of prenatal stress in kainic acid seizures. However, synthetic corticosteroids had no effects or even decreased seizure susceptibility. The mechanisms for these effects are unclear as prenatal corticosteroids may affect several target molecules in the hippocampus such as BDNF, CRH and NPY. It is thus possible that these molecules may be differentially affected by different corticosteroids. Our finding supports the preferential use of synthetic corticosteroids, if any, in obstetrics. (Supported by the grant NS-41366 from NIH.)