[p2–066]: resilience mechanism of the amyloid precursor protein (app) pathway revealed by systems pharmacology modeling following β‐ and gamma‐secretase inhibition

According to the amyloid hypothesis, proteolytic processing of APP to form the β-amyloid (Aβ) peptides plays a central role in the pathophysiology of Alzheimer's Disease (AD). One of the main therapeutic strategies for AD aims at Aβ reduction through either inhibition of Aβ production or enhancing of Aβ clearance. Due to the complexity of the underlying biochemical network, the effects of these interventions on individual attributes of the APP processing pathways are difficult to predict. The objective of the investigation was to extend a systems pharmacology model of the APP processing pathway (van Maanen et al. (2016)), to describe in a strictly quantitative manner drug effects on the various attributes. In a 4-way crossover study in cisterna-magna-ported rhesus monkeys the effects of β-secretase (BACE1) (MBi-5; 30, 125 mg/kg) and γ-secretase (GS) (L675; 240 mg/kg) inhibitors on CSF concentrations of six biomarkers (sAPPβ, Aβ40, Aβ42, Aβ38, Aβ oligomers (AβO), sAPPα) were determined. AβO concentrations were quantified using a two-site ELISA (Savage et al. (2014)). A systems pharmacology model was used to analyze biomarker responses to BACE1 and GS inhibition on the basis of the underlying biological processes (Figure 1). The APP systems model quantified the response of all six biomarkers to BACE1 and GS inhibition. The systems analysis of the decrease of CSF sAPPβ and increase of sAPPα in response to GS inhibition let to the identification of a homeostatic feedback loop regulated by C99: The increase in C99 following GS inhibition stimulated α-secretase processing of APP. A difference in the ratio Aβ42:Aβ40:Aβ38 following BACE1 versus GS inhibition was found, which was explained by a stepwise successive cleavage of C99 by GS, wherein part of Aβ38 is converted from Aβ42. Both BACE1 and GS inhibition resulted in similar profiles for reduction of the AβOs and monomeric Aβ. Both BACE1 and GS inhibition reduce the putatively neurotoxic Aβ oligomer pool. The model suggest that GS inhibition may enhance non-amyloidogenic processing of APP via homeostatic feedback exerted by C99. The APP systems pharmacology model can bring us closer to optimizing therapeutic intervention to reduce AβO burden. Schematic of systems model of APP processing. The model comprised fourtheen compartments: Eight biomarker compartments in brain (yellow circles) and six transit compartments from brain to CSF (white circles). Six biomarkers were measured in CSF (sAPPα, sAPPβ Aβ40, Aβ42, Aβ38 and AβO), indicated by the blue boxes. The drug effect of the BACE1 inhibitor (BACEi EFF) inhibited Rinβ. The drug effect of the GS inhibitor (GSi EFF) inhibited Kin40, Kin40, Kin38 and Kin382. As driver of biomarker response Ctarget was used, which was derived from the PK models of the BACE1 inhibitor (Van Maanen et al. (2016)) and GS inhibitor (unpublished), respectively. The red arrow indicates the homeostatic feedback on α-secretase through the action of C99. APP: Aβ-precursor protein; Aβ: amyloid-β-peptide; Ctarget: drug concentration target site; Kin38: Aβ38 formation rate from C99; kin382: Aβ38 formation rate from Aβ42; Kin40: Aβ40 formation rate; Kin42: Aβ42 formation rate; Kout: Aβ38, Aβ40 and Aβ42 degradation rate; Krev: Oligomer dissociation rate; KtAP: transit rate sAPPα and sAPPβ from brain to CSF; Kpl: Oligomerization rate; KtAB: transit rate Aβ from brain to CSF; KtABO: transit rate AβO from brain to CSF; RinAPP: source of APP; Rinβ: sAPPβ formation rate; Rinα: sAPPα formation rate; Rout: sAPPβ degradation rate; Routα: sAPPα degradation rate.