An Alzheimer's disease-linked PS1 variant rescues the developmental abnormalities of PS1-deficient embryos

Mutations in presenilin 1 (PS1) cosegregate with similar to 25% of early onset familial Alzheimer's disease (FAD) pedigrees. A variety of in vitro and in vivo paradigms have established that one mechanism by which PS1 variants cause AD is by elevating the production of highly amyloidogenic A beta 1-42/43 peptides. PS1 is homologous to sel-12, a C. elegans protein that facilitates signaling mediated by the Notch/lin-12 family of receptors. Wild-type human PS1 complements an egg-laying defect in C. elegans lacking sel-12, while FAD-linked PS1 variants exhibit reduced rescue activity. These data suggested that mutant PS1 may cause disease as a result of reduction in PS1 function. To test the function of FAD-linked PS1 in mammals, we examined the ability of the A246E PS1 variant to complement the embryonic lethality and axial skeletal defects in mice lacking PS1. Finally, to examine the influence of reduced PS1 levels on A beta production, we quantified A beta 1-42/43 peptide levels in PS1 heterozygous null mice (PS1(+/-) mice). We now report that both human wild-type and A246E PS1 efficiently rescue the phenotypes observed in PS1(-/-) embryos, findings consistent with the view that FAD-linked PSI mutants retain sufficient normal function during mammalian embryonic development. Moreover, the levels of A beta 1-42/43 and A beta 1-40 peptides between PS1(+/-) and control mice are indistinguishable. Collectively, these data lead us to conclude that mutant PS1 causes AD not by loss of normal PS1 function but by influencing amyloid precursor protein (APP) processing in a manner that elevates A beta 1-42/43 production.