Elenbecestat and Compound 89 Potently Inhibit BACE1 but Not BACE2 When Subchronically Dosed in Non‐Human Primates

The β-secretase BACE1 (β-site amyloid precursor (APP) cleaving enzyme 1) is a major drug target for Alzheimer's disease (AD), as it catalyzes the first step in amyloid β (Aβ) generation, but has additional substrates and functions, in particular in the brain. Several advanced clinical trials with BACE1 inhibitors were stopped because of an adverse event, a mild cognitive worsening. The underlying mechanism is not yet known but may result from co-inhibition of the BACE1-homolog BACE2. While a cerebrospinal fluid (CSF) biomarker for measuring BACE2 activity is not yet established, VCAM-1 has been suggested as such a biomarker, but has not yet been tested upon prolonged dosing in vivo. Using CSF pharmacoproteomics and a subchronic dosing paradigm in non-human primates, we demonstrate that compound 89, a BACE inhibitor not yet tested in humans, and the clinically tested drug elenbecestat inhibit BACE1 in vivo, with little or no effect on BACE2, as seen with a reduction of substrates of BACE1, but not of the BACE2 substrate VCAM-1. As a control, verubecestat, which inhibits both BACE2 and BACE1, reduced CSF abundance of BACE1 substrates as well as of VCAM-1. This study demonstrates the suitability of VCAM-1 as a pharmacodynamic biomarker for measuring BACE2 target engagement in CSF.