O2-03-04: Aph1B gamma-secretase generates long Abeta peptides and genetic ablation improves Alzheimer's disease phenotypes without affecting notch signaling in the mouse

Abstract

Gamma-secretase cleavage releases the Abeta peptides of Alzheimer's disease (AD), but is also involved in regulated intramembrane proteolysis of Notch and other important signaling pathways. Epsilon-cleavage releases intracellular domains and is physiologically essential, while gamma-cleavages process the remnant peptide until it is soluble and is released. Abeta is heterogeneous as a consequence of gamma-activity and the ratio of long versus short Abeta is central to AD pathogenesis. Gamma-secretase activity is mediated by a multiprotein complex, minimally consisting of Presenilin 1 or 2 (Psen), Aph1A or B/C, Pen2 and Nicastrin. Differential association of Psen or Aph1 isoforms results in structural heterogeneity in the gamma-secretase complex in vivo. This structural diversity raises questions with regard to specificity and biological activity of the different complexes. In the current work, we analyze the activity in reconstituted knock-out cell lines, in purified complexes from mouse and human brain and test the in vivo consequences of this functional heterogeneity in a murine AD model. In contrast with recent in vitro overexpression studies, we find significant differences in the profiles of the Abeta peptides generated by complexes with different Aph1 components. Furthermore, when we specifically inactivate the Aph1B/C gamma-secretase complex in a mouse AD model, we find remarkable improvements of disease-relevant phenotypes without serious Notch side effects. We show that gamma-secretase exists as structurally heterogeneous complexes and that the Aph1 subunit affects gamma-cleavage through a shift in the longer versus shorter Abeta species, but does not affect the release of intracellular domains. This allows dissociation of toxicity and therapeutic effects because specific inactivation of Aph1B(C) in an AD model results in decreased amyloid plaque, Abeta and Abeta oligomer accumulation, and rescue of AD-related phenotypes without significantly altering Notch signalling. Our work suggests that the different Aph1 gamma-secretases contribute differentially to biological and pathological functions and opens a new avenue for more specific and less toxic AD therapy.