Search

which express either the 4-repeat tau domain with the FTDP-17 mutation Delta-K280 (TauRD/DeltaK280 “pro-aggregation mutant”) or the 4-repeat tau domain with Delta-K280 deletion and two proline mutations in the hexapeptide motifs (TauRD/DeltaK280/I277P/I308 “antiaggregation mutant”). Results: The DeltaK280 mutation accelerates the aggregation of tau, but the inserted proline residues inhibit the tau aggregation in vitro and in cell models. Inducible transgene expression in mice was driven by a forebrain-specific CaMKII promoter in a Tet-Off system and can be suppressed by doxycycline. The pro-aggregation mutant showed aggregated tau in sarkosyl insoluble fractions and Gallyas silver stained neurofibrillary tangles from 3 months onwards, even though the level of the human tau protein was lower than endogenous mouse tau. Tau preparations from pro-aggregation mutant mice revealed PHFs by electron microscopy. Consistent with the tau pathology the neuronal loss was age-dependent and visible in the dentate gyrus as early as 5 months. The immunohistochemisty results showed phosphorylated tau at S262 missorted into the somatodendritic compartment of cortical and hippocampal neurons. The anti-aggregation mutant with a similar expression level as the pro-aggregation mutant did not show aggregated tau or neuronal loss, but missorting into the somatodendritic compartment. The level of synaptophysin, a presynaptic marker and the number of spine-synapses were reduced in the stratum radiatum of the pro-aggregation mutant, but not of the antiaggregation mutant. Six weeks of switching off the tau transgene in the pro-aggregation mutant lead to 90% reduction in the level of soluble human tau protein, to complete reversal of the pathological somatodendritic localization and of phosphorylation in the repeat domain. Remarkably, the aggregation was only partly reversed. The remaining aggregates did not contain the exogenous human tau, but the endogenous mouse tau. Conclusions: The results argue that a toxic species of tau, once introduced into neurons, can “poison” endogenous tau and propagate its aggregation for extended time periods. Supported by MPG and DFG.