Pinnix, Inga

ABSTRACT An aspartyl protease, Beta‐Site APP cleaving enzyme (BACE), was identified as the β‐secretase responsible for generating the Amyloid β protein that is believed to cause Alzheimer's disease. BACE has a short propeptide domain that is absent in the mature enzyme because of proteolytic cleavage after the sequence RLPR. This sequence is a predicted substrate for proprotein convertases such as furin. To determine the role of furin and other proprotein convertases, we expressed proBACE in a furin‐deficient mutant Chinese hamster ovary (CHO‐K1) line, RPE.40. ProBACE signal was higher in RPE.40 than in the CHO‐K1 parent, which confirmed that furin plays a role in propeptide removal. However, two independent approaches showed that proBACE is cleaved to mature BACE in RPE.40: proBACE was rapidly turned over in RPE.40 although total BACE was stable, and decanoyl‐RVKR‐chloromethylketone, an inhibitor of the proprotein convertase family, substantially increased proBACE levels in both RPE40 and CHO‐K1. Transient transfection shows that furin, PACE4, PC5/6, and PC7 mediate BACE cleavage in vivo , at least when overexpressed. RPE.40 is proficient in BACE activity despite its furin deficiency. Therefore, our finding that proBACE is cleaved in this mutant leaves open the possibility that maturation is an important regulatory step and a therapeutic target.

ABSTRACT The abnormal accumulation of the amyloid β protein (Aβ) has been implicated as an early and critical event in the etiology and pathogenesis of Alzheimer's disease (AD). Compounds that reduce Aβ accumulation may therefore be useful therapeutically. In cell‐based screens we detected a significant reduction in Aβ concentration after treatment with the phosphatidylinositol kinase inhibitors wortmannin and LY294002. To determine the effect of this class of compounds on in vivo Aβ accumulation, we administered wortmannin to the Tg2576 mouse model of AD. Oral administration of wortmannin over four months resulted in a significant, non‐overlapping 40%–50% reduction in the number of senile plaques, one of the pathological hallmarks of AD. Sandwich ELISA analysis of formic acid extractable Aβ in the brain of treated animals indicates that both Aβ40 and the longer, more amyloidogenic form of the peptide, Aβ42, were significantly reduced. These data provide the first direct evidence that compounds identified by their ability to reduce Aβ concentration in vitro can reduce Aβ accumulation and deposition in the brain, thus establishing a basic paradigm for the identification and evaluation of additional compounds that lower Aβ accumulation.

ABSTRACT The amyloid β‐protein (Aβ) deposited in Alzheimer's disease (AD) is a normally secreted proteolytic product of the amyloid β‐protein precursor (APP). Generation of Aβ from the APP requires two sequential proteolytic events: an initial β‐secretase cleavage at the amino terminus of the Aβ sequence followed by γ ‐secretase cleavage at the carboxyl terminus of Aβ. We describe the development of a robust in vitro assay for γ ‐secretase cleavage by showing de novo Aβ production in vitro and establish that this assay monitors authentic gamma‐secretase activity by documenting the production of a cognate γ ‐CTF, confirming the size of the Aβ produced by mass spectrometry, and inhibiting cleavage in this system with multiple inhibitors that alter γ‐ secretase activity in living cells. Using this assay, we demonstrate that the γ‐secretase activity 1 ) is tightly associated with the membrane, 2 ) can be solubilized, 3 ) has a pH optimum of 6.8 but is active from pH 6.0 to pH >8.4, and 4 ) ascertain that activities of the γ ‐40 and γ ‐42 are indeed pharmacologically distinct. These studies should facilitate the purification of the protease or proteases that are responsible for this unusual activity, which is a major therapeutic target for the treatment of AD.