The amyloid-beta rise and gamma-secretase inhibitor potency depend on the level of substrate expression.

The amyloid-beta (Abeta) peptide, which likely plays a key role in Alzheimer disease, is derived from the amyloid-beta precursor protein (APP) through consecutive proteolytic cleavages by beta-site APP-cleaving enzyme and gamma-secretase. Unexpectedly gamma-secretase inhibitors can increase the secretion of Abeta peptides under some circumstances. This "Abeta rise" phenomenon, the same inhibitor causing an increase in Abeta at low concentrations but inhibition at higher concentrations, has been widely observed. Here we show that the Abeta rise depends on the beta-secretase-derived C-terminal fragment of APP (betaCTF) or C99 levels with low levels causing rises. In contrast, the N-terminally truncated form of Abeta, known as "p3," formed by alpha-secretase cleavage, did not exhibit a rise. In addition to the Abeta rise, low betaCTF or C99 expression decreased gamma-secretase inhibitor potency. This "potency shift" may be explained by the relatively high enzyme to substrate ratio under conditions of low substrate because increased concentrations of inhibitor would be necessary to affect substrate turnover. Consistent with this hypothesis, gamma-secretase inhibitor radioligand occupancy studies showed that a high level of occupancy was correlated with inhibition of Abeta under conditions of low substrate expression. The Abeta rise was also observed in rat brain after dosing with the gamma-secretase inhibitor BMS-299897. The Abeta rise and potency shift are therefore relevant factors in the development of gamma-secretase inhibitors and can be evaluated using appropriate choices of animal and cell culture models. Hypothetical mechanisms for the Abeta rise, including the "incomplete processing" and endocytic models, are discussed.

Ex vivo occupancy of gamma-secretase inhibitors correlates with brain beta-amyloid peptide reduction in Tg2576 mice.

Reduction of brain beta-amyloid peptide (Abeta) synthesis by gamma-secretase inhibitors is a promising approach for the treatment of Alzheimer's disease. However, measurement of central pharmacodynamic effects in the Alzheimer's disease patient will be a challenge. Determination of drug occupancy may facilitate the analysis of efficacy of gamma-secretase inhibitors in a clinical setting. In this study, the relationship of gamma-secretase site occupancy and brain Abeta40 reduction by gamma-secretase inhibitors was examined in Tg2576 mice. [3H](2R,3S)-2-Isobutyl-N1-((S)-1-methyl-2-oxo-5-phenyl-2,3-dihydro-1H-benzo[e][1,4]diazepin-3-yl)-3-propylsuccinamide (IN973) was used as a gamma-secretase radioligand, since it has been shown to bind to gamma-secretase in rat, rhesus, and human brains with high affinity and specificity. We extended these findings by showing that [3H]IN973 bound to gamma-secretase in Tg2576 brains with an affinity, specificity, and regional localization very similar to the other species. To quantify gamma-secretase occupancy by gamma-secretase inhibitors, an ex vivo binding assay was developed using [3H]IN973 and frozen brain sections from drug-treated mice. Gamma-secretase occupancy and brain Abeta40 reduction were found to be highly correlated in animals dosed with either 2-[(1R)-1-[[4-chlorophenyl)-sulfonyl](2,5-difluorophenyl)amino] ethyl]-5-fluoro-benzenepropanoic acid (BMS-299897) or (S)-2-((S)-2-(3,5-difluorophenyl)-2-hydroxyacetamido)-N-((S,Z)-3-methyl-4-oxo-4,5-dihydro-3H-benzo[d][1,2]diazepin-5-yl)propanamide (BMS-433796) over a wide range of doses and times postdose, with the exception of the earliest times postdose. This lag in Abeta40 response to gamma-secretase inhibition is probably related to the delayed clearance of previously produced Abeta40. The excellent correlation between brain Abeta40 and gamma-secretase occupancy suggests that a positron emission tomography ligand for gamma-secretase could be a valuable biomarker to determine whether gamma-secretase inhibitors bind to their target in humans.