JNJ-40418677

γ-Secretase is responsible for the proteolysis of amyloid precursor protein (APP) into amyloid-beta (Aβ) peptides, which are centrally implicated in the pathogenesis of Alzheimer's disease (AD). The biochemical mechanism of how processing by γ-secretase is regulated, especially as regards the interaction between enzyme and substrate, remains largely unknown. Here, mutagenesis reveals that the hydrophilic loop-1 (HL-1) of presenilin-1 (PS1) is critical for both γ-secretase step-wise cleavages (processivity) and its allosteric modulation by heterocyclic γ-modulatory compounds. Systematic mutagenesis of HL-1, including all of its familial AD mutations and additional engineered variants, and quantification of the resultant Aβ products show that HL-1 is necessary for proper sequential γ-secretase processivity. We identify Y106, L113, and Y115 in HL-1 as key targets for heterocyclic γ-secretase modulators (GSMs) to stimulate processing of pathogenic Aβ peptides. Further, we confirm that the GxxxG domain in the APP transmembrane region functions as a critical substrate motif for γ-secretase processivity: a G29A substitution in APP-C99 mimics the beneficial effects of GSMs. Together, these findings provide a molecular basis for the structural regulation of γ-processivity by enzyme and substrate, facilitating the rational design of new GSMs that lower AD-initiating amyloidogenic Aβ peptides.

In the process of drug development, in vitro studies do not always adequately predict human-specific drug responsiveness in clinical trials. Here, we applied the advantage of human iPSC-derived neurons, which offer human-specific drug responsiveness, to screen and evaluate therapeutic candidates for Alzheimer's disease (AD). Using AD patient neurons with nearly 100% purity from iPSCs, we established a robust and reproducible assay for amyloid β peptide (Aβ), a pathogenic molecule in AD, and screened a pharmaceutical compound library. We acquired 27 Aβ-lowering screen hits, prioritized hits by chemical structure-based clustering, and selected 6 leading compounds. Next, to maximize the anti-Aβ effect, we selected a synergistic combination of bromocriptine, cromolyn, and topiramate as an anti-Aβ cocktail. Finally, using neurons from familial and sporadic AD patients, we found that the cocktail showed a significant and potent anti-Aβ effect on patient cells. This human iPSC-based platform promises to be useful for AD drug development.