J. Janson(1), S. Eketjäll(2), H. Yan(1), K. Tunblad(1), F. Jeppsson(2), S. Briem(3), C. Dahlqvist(2), A.-C. Radesäter(2), J. Fälting(4), S. Visser(5)
(1)Modeling & Simulation, (2)Neuroscience, (3)Global DMPK, (4)Project Management, (5)Global DMPK Centre of Excellence, Innovative Medicines CNSP AstraZeneca Södertälje, Sweden
Objectives: The transmembrane aspartic acid protease BACE1 cleaves amyloid precursor protein (APP) to generate soluble APPβ. The remainder of APP in the cell membrane is in turn cleaved by γ-secretase to form beta amyloid (Aβ). In vivo, pharmacologically induced Aβ reduction is dependent on compound- and system-parameters, i.e. potency and Aβ clearance respectively. Preclinical PKPD modeling was applied 1) to gain insight in the time-course between oral dose, plasma and brain exposure and inhibitory effect on Aβ in brain and CSF of novel BACE1 inhibitors as well as system constants, 2) to quantify the in vivo potency using population pharmacokinetic and pharmacodynamic (PKPD) modeling, and 3) to investigate the correlation between in vitro and in vivo potency.
Methods: BACE1 inhibitors were characterized in vitro in human SH-SY5Y cells, mouse N2A cells, and in mouse and guinea pig primary cortical neurons (PCN). The PKPD properties of 28 compounds were evaluated in vivo using female C57BL/6 mice in single dose, dose- and/or time-response studies (10 compounds). Four compounds were studied in male Dunkin-Hartley guinea pigs. Plasma exposure was converted to free brain exposure using brain exposure, plasma protein and brain binding. Free brain concentrations were used as input to inhibition of brain Aβ production rate. For both species, population modeling of all in vivo data was performed using an indirect response model with inhibition on the Aβ production rate to estimate the unbound brain concentration giving 20% inhibition from baseline (IC20%).
Results: In vitro potency in human SH-SY5Y cells was correlated to potency in mouse PCN with a 4.5-fold lower IC50 value than in SH-SY5Y cells. In vivo BACE1 inhibitors exhibited concentration- and time-dependent lowering of plasma, CSF (guinea pig) and brain Aβ levels. In mouse the population estimate of turnover of Aβ40 in brain was 26 minutes. In guinea pig this population turnover rate was estimated at 1 hour.
Conclusions: A good correlation between mouse and guinea pig in vitro and in vivo potency and an excellent correlation between mouse PCN and human SH-SY5Y, increased the confidence in using human cell lines for screening and optimization for effect of novel BACE1 inhibitors. The established in vitro-in vivo correlations could thereby reduce the number and change design of preclinical in vivo effect studies.
Reference: PAGE 21 () Abstr 2340 [www.page-meeting.org/?abstract=2340]
Poster: CNS