Alzheimer's disease is a progressive dementia characterized by pronounced degeneration of certain populations of neurons in the hippocampus and cerebral cortex of the brain. One theory is that glutamate receptor-mediated toxicity plays a role in cell loss associated with Alzheimer's disease. We used in situ hybridization to examine GluR1, GluR2, and GluR3 messengerRNAs (encoding α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid/kainate receptor subunits) in sections of autopsy samples of Alzheimer's disease brains and age-, sex-, and post-mortem delay-matched brains from non-demented (control) subjects. GluR1 and GluR2 exhibited a heterogeneous distribution in control brain. GluR1 was expressed in granule cells of the dentate gyrus, in pyramidal cells of the CA1 and CA3 hippocampal subfields and in neurons of the subiculum and entorhinal cortex. GluR2 mRNA was at high density in the dentate gyrus and in CA3, but was at low density in CA1, subiculum, and entorhinal cortex. GluR3 hybridization was at very low levels but selectively localized to the dentate gyrus and CA3. In cerebellum, GluR1 was found in granule and Purkinje cell layers. In sections from Alzheimer's disease brain, a high degree of intersubject variability was observed: some samples showed markedly reduced GluR1 mRNA levels in dentate gyrus, CA1 and CA3 relative to controls; others showed no changes. Microscopic observation of emulsion-dipped sections revealed that the reduction of GluR1 seen in the dentate gyrus and CA3 of some Alzheimer's disease subjects was not due to cell loss. Due to generally low hybridization signals obtained for GluR2 and GluR3, only GluR1 could be reliably quantitated. When hybridization densities were correlated with a number of pre- and post-mortem variables, a significant negative correlation (r = - 0.68, P = < 0.01) was found between the storage duration of brain samples and GluR1 messengerRNA levels in dentate gyrus. Our results show that GluR1 messengerRNA is not changed in Alzheimer's disease hippocampus; relative to storage interval-matched control hippocampus; all of the differences in mean expression levels between Alzheimer and control brains could be ascribed to greater post-mortem storage time of tissue samples. The feasibility of studies of glutamate receptor messengerRNA levels in neurological diseases of humans is demonstrated, as is the importance of controlling for variations in handling of tissues.
ASJC Scopus subject areas