Alzheimer mutations stabilize synaptotoxic γ-secretase-substrate complexes

Alzheimers disease is characterized pathologically by cerebral deposition of 42-residue amyloid {beta}-peptide (A{beta}42), proteolytically produced from amyloid precursor protein (APP) by {beta}- and {gamma}-secretases.1 Although mutations in APP and presenilin, the catalytic component of {gamma}-secretase, cause familial Alzheimers disease (FAD), a role for A{beta}42 as the primary disease driver has not been clearly established and remains controversial.2,3 Here we show through comprehensive analysis of the multi-step proteolysis of APP substrate C99 by {gamma}-secretase that FAD mutations are consistently deficient in early proteolytic events, not later events that produce secreted A{beta} peptides. Cryo-electron microscopy revealed that a substrate mimetic traps {gamma}-secretase at the transition state for intramembrane proteolysis, and this structure closely aligns with activated enzyme-substrate complex captured by molecular dynamics simulations. In silico simulations and fluorescence lifetime imaging microscopy in cultured cells support stabilization by FAD mutations of enzyme-substrate and/or enzyme-intermediate complexes. Neuronal expression of C99 and/or presenilin-1 in Caenorabditis elegans led to age-dependent synaptic loss only when one of the transgenes carried an FAD mutation. Designed mutations that stabilize the enzyme-substrate complex and block proteolysis likewise led to synaptic loss. Collectively, these findings implicate the stalled process--not the released products--of {gamma}-secretase cleavage of substrates in FAD pathogenesis.