Regionally specific and rapid increases in brain-derived neurotrophic factor messenger RNA in the adult rat brain following seizures induced by systemic administration of kainic acid.

In situ hybridization techniques were used to analyse the spatiotemporal pattern of brain-derived neurotrophic factor messenger RNA elevation associated with kainic acid-induced seizure activity in the rat. Pronounced increases in hippocampal brain-derived neurotrophic factor messenger RNA levels were observed as early as 30 min following the onset of behavioral seizures. The greatest increase (10-fold) occurred in the dentate granule cell layer, while pyramidal layers CA1, CA3, and CA4 exhibited increases of two- to six-fold. Peak elevation of brain-derived neurotrophic factor messenger RNA in CA1 hippocampal region was evident at 4 h in CA3, and in the dentate granule layer at 30 min postseizure. Elevations persisted in the dentate and hilar regions to four days, while the increases in CA1 and CA3 returned to control levels by 16 h following seizure. Significant increases in brain-derived neurotrophic factor messenger RNA were also observed in the superficial layers of cortex (II and III) and in the piriform cortex which reached peak elevations by 8 h. No detectable changes were observed in the dorsomedial thalamus. Although histologically defined pyramidal and granule cell layers displayed relatively uniform increases in brain-derived neurotrophic factor messenger RNA in response to kainate, a closer examination of the labeling patterns using emulsion autoradiography revealed discrete areas of high grain densities overlapping uniform, moderate hybridization densities in the dentate granule cell layer and CA3, suggesting that the capacity to upregulate brain-derived neurotrophic factor messenger RNA in these regions may differ among individual neurons. In conclusion, our studies revealed that brain-derived neurotrophic factor messenger RNA induction in response to systemic kainate administration differs in hippocampal and cortical areas, in magnitude, time of onset and duration. The observed temperospatial pattern does not correspond in a simple way to increases in metabolic or electrical activity associated with seizures or neuronal vulnerability coincident with the seizures.