The uniqueness of using iPSC-differentiated neurons is obvious by the variation among the five lines that we tested. We found that 3D neuro-spheroids differentiated from subject AD1 did not respond to BACE1 inhibitor such that no reduction of Aβ42 was observed. This is an unexpected finding with implications for both the biology of APP processing and the clinical application of secretase inhibitors. If the same class of BACE1 inhibitors repetitively shows a lack of efficacy in blocking Aβ production in a number of subjects, we probably would not enroll these subjects for clinical trials for the BACE1 inhibitors. Analysis of these lines for SNP may yield highly useful pharmacogenetic markers for individualization of treatment strategies for amyloid reduction. The molecular mechanism behind this phenomenon is obscure. One possible explanation is that genetic variation, not necessary familial AD mutations like Swedish mutation at the BACE1 cleavage site [40], increases the expression levels of several genes like APP and BACE1. High expression of substrate and enzymatic activity may compensate for the reduction by BACE1 inhibitors. Similarly, levels of clathrin proteins were higher
