In this study, we first confirmed previous reports that the expression of human P301L tau in the JNPL3 mouse CNS leads to the formation of PHF-tau aggregates and elevated levels of insoluble GSK-3β in a pattern that correlates with regional degeneration in this model. High levels of PHF-tau and GSK-3β are observed in spinal cord and cortex, whereas lower levels of GSK-3β and PHF-tau are observed in the cerebellum. In parallel with the increase of GSK-3β in affected CNS regions, we report the novel observation that total β-catenin levels are elevated in the same pattern in affected CNS regions early in the disease course, prior to overt pathology. In addition, we show that this elevation of β-catenin specifically reflects activated Wnt signaling, which induces Tcf transcriptional activity. This may seem paradoxical, since GSK-3β activity is expected to diminish β-catenin levels. In turn, a reduction of β-catenin would be expected to lead to reduced Wnt signaling/Tcf transcriptional activity. However, we also find that during this early phase of disease, GSK-3β becomes co-localized with pathologically phosphorylated tau in the insoluble protein fraction, which we hypothesized would lead to a reduced availability of GSK-3β in the cytosol, causing elevated β-catenin due to decreased phosphorylation of β-catenin by GSK-β. We observed this elevation of β-catenin at three months in an early stage of the disease prior to overt neurodegeneration.
These findings are in accordance with an earlier invertebrate study, in which we evaluated the function of β-catenin/Tcf/Lef in both over expression and loss-of-function studies in a Drosophila
eye model of human tau over expression. When homologs of β-catenin and the nuclear co-transcription factor Tcf were increased in Drosophila
, the neurodegeneration phenotype was enhanced [12
]. In contrast to PHF formation, which was observed only in adult flies, the exacerbation of degeneration caused by GSK-3β occurred at all developmental stages, - larval, pupal and adult. Thus, the tangle formation took place later in the fly than the initiation of cell death. Feany and colleagues also observed cell death without NFT’s in drosophila overexpressing human tau wild type and several mutants [28
Our previous fly data suggested that increased Wnt signaling via β-catenin could be an early feature in tau related neurodegeneration. Here we tested this hypothesis in a mouse model of FTD. The present study is the first in vivo report providing evidence in mammals, that total and activated β-catenin, and canonical Wnt signaling via β-catenin/Tcf/Lef complex are elevated many months prior to overt pathology with neuronal loss. Since we observed that β-catenin is elevated only early in the disease process in the JNPL3 mouse, our results suggest that β-catenin may play a role in the initial phase of degenerative changes.
These findings are particularly interesting in light of previous reports in mice with conditional over expression of GSK-3β. Aspects of tau-induced neuropathology, such as tau hyperphosphorylation, reactive astrocytosis, and neuronal death [27
], and spatial learning deficit [11
] were reported when GSK-3β was over expressed. When GSK-3β over expression was inhibited, normal GSK-3β activity, reduced levels of phosphorylated tau, and diminished neuronal death were shown [8
], supporting a more complex role of GSK-3β in neurodegeneration than solely its role in tau-phosphorylation.
Thus, complementary to observations in the fly [12
], the present evidence that β-catenin is increased early in the disease course before NFT formation peaks in the mouse supports the hypothesis that cellular processes leading to neuronal degeneration may not be initiated by tau hyperphosphorylation itself. We hypothesize that tau hyperphophorylation is a product of increased interaction between GSK-3β and tau, whereas the effect of this interaction on the Wnt pathway may be a separate consequence of the tau/GSK-3β interaction, both, leading eventually to neurodegeneration. This hypothesis is supported by previous reports that the link of tau phosphorylation and GSK-3β activity to neurodegeneration is not consistent in studies of transgenic mice [19
]. Therefore, based on the present data, we conclude that tau toxicity is not solely attributable to altered microtubule binding and its aggregation, but that tau can also indirectly affect cellular pathways causing neurodegeneration. The role of the Wnt pathway, and its connection to other processes that are implicated in neurodegeneration, such as cell cycle activation, and its interaction with other pro-degenerative factors, such as β-amyloid, as well as Wnt-pathway inhibitors such as proteins of the Dickkopf family that have been implicated in Alzheimer’s disease [6
], are deserving of further investigation.