Mitosis-specific phosphorylation of amyloid precursor protein at Threonine 668 leads to its altered processing and association with centrosomes

BACKGROUND: Atypical expression of cell cycle regulatory proteins has been implicated in Alzheimer's disease (AD), but the molecular mechanisms by which they induce neurodegeneration are not well understood. We examined transgenic mice expressing human amyloid precursor protein (APP) and presenilin 1 (PS1) for changes in cell cycle regulatory proteins to determine whether there is a correlation between cell cycle activation and pathology development in AD. RESULTS: Our studies in the AD transgenic mice show significantly higher levels of cyclin E, cyclin D1, E2F1, and P-cdc2 in the cells in the vicinity of the plaques where maximum levels of Threonine 668 (Thr668)-phosphorylated APP accumulation was observed. This suggests that the cell cycle regulatory proteins might be influencing plaque pathology by affecting APP phosphorylation. Using neuroglioma cells overexpressing APP we demonstrate that phosphorylation of APP at Thr668 is mitosis-specific. Cells undergoing mitosis show altered cellular distribution and localization of P-APP at the centrosomes. Also, Thr668 phosphorylation in mitosis correlates with increased processing of APP to generate Aβ and the C-terminal fragment of APP, which is prevented by pharmacological inhibitors of the G1/S transition. CONCLUSIONS: The data presented here suggests that cell cycle-dependent phosphorylation of APP may affect its normal cellular function. For example, association of P-APP with the centrosome may affect spindle assembly and cell cycle progression, further contributing to the development of pathology in AD. The experiments with G1/S inhibitors suggest that cell cycle inhibition may impede the development of Alzheimer's pathology by suppressing modification of βAPP, and thus may represent a novel approach to AD treatment. Finally, the cell cycle regulated phosphorylation and processing of APP into Aβ and the C-terminal fragment suggest that these proteins may have a normal function during mitosis.