Search

Tauopathies differ in terms of the brain regions that are affected. In Alzheimer's disease, basal forebrain and hippocampus are mainly involved, while frontotemporal lobar degeneration affects the frontal and temporal lobes and subcortical nuclei including striatum. Over 90% of human cases of tauopathies are sporadic, although the majority of established tau-transgenic mice have had mutations. This prompted us to establish transgenic mice expressing wild-type human tau (Tg601). Old (>14 months old) Tg601 mice displayed decreased anxiety in the elevated plus maze test and impaired place learning in the Morris water maze test. Immunoblotting of brain tissue identified that soluble tau multimer was increased with aging even though insoluble tau was not observed. In the striatum of old Tg601, the level of AT8- or AT180-positive tau was decreased compared with that of other regions, while PHF-1-positive tau levels remained equal. Phosphorylated tau-positive axonal dilations were present mainly in layers V and VI of the prefrontal cortex. Loss of synaptic dendritic spine and decreased immunohistochemical level of synaptic markers were observed in the nucleus accumbens. In vivo 2-[18F]fluoro-2-deoxy-d-glucose positron emission tomography analysis also showed decreased activity exclusively in the nucleus accumbens of living Tg601 mice. In Tg601 mice, the axonal transport defect in the prefrontal cortex–nucleus accumbens pathway may lead to decreased anxiety behavior. Differential distribution of hyperphosphorylated tau may cause region-specific neurodegeneration.

Activation of GSK-3beta is presumed to be involved in various neurodegenerative diseases, including Alzheimer's disease (AD), which is characterized by memory disturbances during early stages of the disease. The normal function of GSK-3beta in adult brain is not well understood. Here, we analyzed the ability of heterozygote GSK-3beta knockout (GSK+/-) mice to form memories. In the Morris water maze (MWM), learning and memory performance of GSK+/- mice was no different from that of wild-type (WT) mice for the first 3 days of training. With continued learning on subsequent days, however, retrograde amnesia was induced in GSK+/- mice, suggesting that GSK+/- mice might be impaired in their ability to form long-term memories. In contextual fear conditioning (CFC), context memory was normally consolidated in GSK+/- mice, but once the original memory was reactivated, they showed reduced freezing, suggesting that GSK+/- mice had impaired memory reconsolidation. Biochemical analysis showed that GSK-3beta was activated after memory reactivation in WT mice. Intraperitoneal injection of a GSK-3 inhibitor before memory reactivation impaired memory reconsolidation in WT mice. These results suggest that memory reconsolidation requires activation of GSK-3beta in the adult brain.

Advanced age and mutations in the genes encoding amyloid precursor protein (APP) and presenilin (PS1) are two serious risk factors for Alzheimer's disease (AD). Finding common pathogenic changes originating from these risks may lead to a new therapeutic strategy. We observed a decline in memory performance and reduction in hippocampal long-term potentiation (LTP) in both mature adult (9-15 months) transgenic APP/PS1 mice and old (19-25 months) non-transgenic (nonTg) mice. By contrast, in the presence of bicuculline, a GABA(A) receptor antagonist, LTP in adult APP/PS1 mice and old nonTg mice was larger than that in adult nonTg mice. The increased LTP levels in bicuculline-treated slices suggested that GABA(A) receptor-mediated inhibition in adult APP/PS1 and old nonTg mice was upregulated. Assuming that enhanced inhibition of LTP mediates memory decline in APP/PS1 mice, we rescued memory deficits in adult APP/PS1 mice by treating them with another GABA(A) receptor antagonist, picrotoxin (PTX), at a non-epileptic dose for 10 days. Among the saline vehicle-treated groups, substantially higher levels of synaptic proteins such as GABA(A) receptor alpha1 subunit, PSD95, and NR2B were observed in APP/PS1 mice than in nonTg control mice. This difference was insignificant among PTX-treated groups, suggesting that memory decline in APP/PS1 mice may result from changes in synaptic protein levels through homeostatic mechanisms. Several independent studies reported previously in aged rodents both an increased level of GABA(A) receptor alpha1 subunit and improvement of cognitive functions by long term GABA(A) receptor antagonist treatment. Therefore, reduced LTP linked to enhanced GABA(A) receptor-mediated inhibition may be triggered by aging and may be accelerated by familial AD-linked gene products like Abeta and mutant PS1, leading to cognitive decline that is pharmacologically treatable at least at this stage of disease progression in mice.