Alzheimer's presenilin 1 modulates sorting of APP and its carboxyl-terminal fragments in cerebral neurons in vivo.

Sam Gandy,Yun-wu Zhang,Annat Ikin,Stephen D Schmidt,Alexey Bogush,Efrat Levy,Roxanne Sheffield,Ralph A Nixon,Francesca-Fang Liao,Paul M Mathews,Huaxi Xu,Michelle E Ehrlich
DOI: https://doi.org/10.1111/j.1471-4159.2007.04587.x
2007-01-01
Journal of Neurochemistry
Abstract:Studies in continuously cultured cells have established that familial Alzheimer's disease (FAD) mutant presenilin 1 (PS1) delays exit of the amyloid precursor protein (APP) from the trans-Golgi network (TGN). Here we report the first description of PS1-regulated APP trafficking in cerebral neurons in culture and in vivo. Using neurons from transgenic mice or a cell-free APP transport vesicle biogenesis system derived from the TGN of those neurons, we demonstrated that knocking-in an FAD-associated mutant PS1 transgene was associated with delayed kinetics of APP arrival at the cell surface. Apparently, this delay was at least partially attributable to impaired exit of APP from the TGN, which was documented in the cell-free APP transport vesicle biogenesis assay. To extend the study to APP and carboxyl terminal fragment (CTF) trafficking to cerebral neurons in vivo, we performed subcellular fractionation of brains from APP transgenic mice, some of which carried a second transgene encoding an FAD-associated mutant form of PS1. The presence of the FAD mutant PS1 was associated with a slight shift in the subcellular localization of both holoAPP and APP CTFs toward iodixanol density gradient fractions that were enriched in a marker for the TGN. In a parallel set of experiments, we used an APP : furin chimeric protein strategy to test the effect of artificially forcing TGN concentration of an APP : furin chimera that could be a substrate for beta- and gamma-cleavage. This chimeric substrate generated excess Abeta42 when compared with wildtype APP. These data indicate that the presence of an FAD-associated mutant human PS1 transgene is associated with redistribution of the APP and APP CTFs in brain neurons toward TGN-enriched fractions. The chimera experiment suggests that TGN-enrichment of a beta-/gamma-secretase substrate may play an integral role in the action of mutant PS1 to elevate brain levels of Abeta42.
What problem does this paper attempt to address?