A new glucocorticoid hypothesis of brain aging: implications for Alzheimer's disease.

The original glucocorticoid (GC) hypothesis of brain aging and Alzheimer's disease proposed that chronic exposure to GCs promotes hippocampal aging and AD. This proposition arose from a study correlating increasing plasma corticosterone with hippocampal astrocyte reactivity in aging rats. Numerous subsequent studies have found evidence consistent with this hypothesis, in animal models and in humans. However, several results emerged that were inconsistent with the hypothesis, highlighting the need for a more definitive test with a broader panel of biomarkers. We used microarray analyses to identify a panel of hippocampal gene expression changes that were aging-dependent, and also corticosterone-dependent. These data enabled us to test a key prediction of the GC hypothesis, namely, that the expression of most target biomarkers of brain aging should be regulated in the same direction (increased or decreased) by both GCs and aging. This prediction was decisively contradicted, as a majority of biomarker genes were regulated in opposite directions by aging and GCs, particularly inflammatory and astrocyte-specific genes. Thus, the initial hypothesis of simple positive cooperativity between GCs and aging must be rejected. Instead, our microarray data suggest that in the brain GCs and aging interact in more complex ways that depend on the cell type. Therefore, we propose a new version of the GC-brain aging hypothesis; its main premise is that aging selectively increases GC efficacy in some cell types (e.g., neurons), enhancing catabolic processes, whereas aging selectively decreases GC efficacy in other cell types (e.g., astrocytes), weakening GC anti-inflammatory activity. We also propose that changes in GC efficacy might be mediated in part by cell type specific shifts in the antagonistic balance between GC and insulin actions, which may be of relevance for Alzheimer's disease pathogenesis.