GSK-3beta activation mediates Abeta induced tau pathology in multiple transgenic mice

Abstract

Alzheimer's disease (AD) is a complex disorder neuropathologically characterized by extra- and/or intracellular aggregation of toxic Abeta peptide and tau protein species, neuroinflammation, neuronal and synaptic loss. Toxic Abeta species are considered to be upstream in the etiopathological mechanism. We could mimic all neuropathological characteristics except for the tau pathology in mutant V717I amyloid precursor protein (APP-V717I) overexpressing mice. This mutation causes a familial form of AD in patients. In order to have a more complete animal model we overexpressed APP-V717I and tau-P301L protein in mice. Tau-P301L causes hereditary frontotemporal dementia (FTD) in patients. Whereas the single transgenic APP-V717 and Tau-P301L transgenic mice developed amyloid and tau pathology (intraneuronal aggregation of tau protein), both pathologies were combined in the bigenic (biAT) mice. However, tau pathology was redistributed from the spinal cord and hindbrain to the forebrain in the bigenic mice as compared to the tau-P301L mice. Aggregated tau is invariably hyperphosphorylated. To investigate the contribution of tau phosphorylation in the development of tau pathology tau-P301L and the most important tau kinase, GSK-3beta, were co-overexpressed in mice (biGT mice). Strikingly, the tau pathology in the biGT mice was redistributed in a similar fashion in the biGT mice as in the biAT mice compared to the tau-P301L mice. It could be demonstrated as well that GSK-3beta activity in brain is increased as a consequence of APP overexpression. In a cellular model, tau expressing yeast, we could demonstrate that increased phosphorylation of tau causes a conformational change that starts tau aggregation. Together the data imply that toxic Abeta species cause kinase activation, increased tau phosphorylation, which subsequently leads to tau pathology. The mechanism by which Abeta causes activation of GSK-3beta is unknown. Currently we are investigating the possibility whether neuroinflammation is involved.