A gamma-secretase inhibitor decreases amyloid-beta production in the central nervous system

Abstract

Objective: Accumulation of amyloid-beta (Abeta) by overproduction or underclearance in the central nervous system (CNS) is hypothesized to be a necessary event in the pathogenesis of Alzheimer's disease. However, previously, there has not been a method to determine drug effects on Abeta production or clearance in the human CNS. The objective of this study was to determine the effects of a gamma-secretase inhibitor on the production of Abeta in the human CNS.

Methods: We utilized a recently developed method of stable-isotope labeling combined with cerebrospinal fluid sampling to directly measure Abeta production during treatment of a gamma-secretase inhibitor, LY450139. We assessed whether this drug could decrease CNS Abeta production in healthy men (age range, 21-50 years) at single oral doses of 100, 140, or 280mg (n = 5 per group).

Results: LY450139 significantly decreased the production of CNS Abeta in a dose-dependent fashion, with inhibition of Abeta generation of 47, 52, and 84% over a 12-hour period with doses of 100, 140, and 280mg, respectively. There was no difference in Abeta clearance.

Interpretation: Stable isotope labeling of CNS proteins can be utilized to assess the effects of drugs on the production and clearance rates of proteins targeted as potential disease-modifying treatments for Alzheimer's disease and other CNS disorders. Results from this approach can assist in making decisions about drug dosing and frequency in the design of larger and longer clinical trials for diseases such as Alzheimer's disease, and may accelerate effective drug validation. Ann Neurol 2009.

Figure 1. Diagram of Central Nervous System Stable Isotope Labeled Kinetics (CNS-SILK)

1) A stable isotope labeled amino acid is infused into the blood stream and is transported to the brain. 2) The labeled amino acid is incorporated into newly synthesized proteins (e.g. APP in neurons). 3) Labeled APP is cut by β- and γ-secretases to produced labeled Aβ, or in the presence of a gamma-secretase inhibitor (4) labeled Aβ production is inhibited. 5) Labeled and unlabeled Aβ is transported and cleared through the cerebral-spinal fluid which is in direct communication with the extracellular space of the brain, where sampling occurs over time to measure the production and clearance of Aβ.

Figure 2. Concentration of LY450139 in cerebral-spinal fluid during production and clearance of Aβ during and after labeling

A) There are dose-dependent increases in the concentration of LY450139 in cerebral-spinal fluid with the majority of drug present in the central nervous system during the first twelve hours. B) The absolute amount of newly generated Aβ (labeled) was measured during exposure to placebo or the γ-secretase inhibitor LY450139. The production phase (hours 1–12) and clearance phase (hours 24–36) indicate the time period of analysis for production and clearance of labeled Aβ.

Figure 3. Area under the curve (AUC) analysis of newly generated Aβ

A) A dose-response reduction in newly generated Aβ is present during hours 1–12. B) There was similar clearance of Aβ between hour 24–36. C). Individual AUC values for hours 1–12 (AUC_1–12) of newly generated Aβ_1–x for each participant is plotted according to dose as well as mean +/− 95% confidence intervals. There is a dose-dependent decrease in the production of Aβ. D) AUC for hours 24–36 (AUC_24–36) of newly generated Aβ for each participant as well as mean +/− 95% confidence intervals are shown. There is no significant difference in Aβ clearance between groups.

Figure 4. Analysis of pharmacokinetic-pharmacodynamic (PK-PD) relationship of LY450139

The AUC_1–12 of newly generated Aβ is plotted as a function of AUC_1–12 of LY450139 in the cerebral-spinal fluid. There is a significant (p=0.002) correlation of drug exposure and decreased Aβ generation.

Figure 5. Normalized cerebral-spinal fluid Aβ levels measured by ELISA

A. The average Aβ levels normalized to baseline values at hour 0 is shown for each group. There is a trend for increasing Aβ levels over time in the placebo and treated groups. The average standard deviation is 48 percent in the placebo group, and error bars are not displayed for readability. B. The average Aβ levels for each group were subtracted from the placebo group values at each hour. Although fluctuations are present, there is a trend towards decrease in the Aβ_1–x AUC_1–12, with only LY450139 280mg dose being significant (p<0.05).