Structure-Based Drug Design of Selective BACE1 and BACE2 Inhibitors for the Treatments of Alzheimer's Disease and Type II Diabetes

Yu-Chen Yen, Purdue University

Abstract

The BACE (Beta-site Amyloid precursor protein Cleaving Enzyme) protein family contains two members, BACE1 and BACE2. BACE1 has been demonstrated to be the major β-secretase involved in the production of amyloid beta (Aβ) peptide. The Aβ cascade hypothesis suggests that the accumulation of Aβ peptides in the brain would lead to the formation of plaques and tangles, which are the neuropathological hallmarks of Alzheimer disease (AD). Based upon this, inhibition of the enzymatic activity of BACE1 to prevent the production of Aβ peptides has been proposed as a novel therapeutic approach for the treatment of AD. Recently, the other member of this family, BACE2, has been identified as a potential target for the treatment of Type II diabetes (TIID). In 2011, BACE2 was shown to be a major sheddase of TMEM27, which is a transmembrane protein located on the pancreatic β cells. The full-length TMEM27 on pancreatic β cells would promote the proliferation of pancreatic β cells as well as stimulate the secretion of insulin. However, it loses its mitogenic activity once it is cleaved by BACE2. Therefore, the inhibition of the enzymatic activity of BACE2 would likely maintain the function of TMEM27 on pancreatic β cells and thus benefit TIID patients. This hypothesis has been tested using a BACE2 inactive mouse model and a diabetes mouse model treated with a BACE2 inhibitor. Increased insulin level and improved control of glucose homeostasis have been observed in both models, supporting BACE2 as a promising target for the treatment of TIID. Although both BACE1 and BACE2 have been proposed as promising therapeutic targets for the treatments of AD and TIID, selective inhibitors against either BACE1 or BACE2 are needed. From an animal study, it has been shown that BACE1/2 double knockout mice exhibit a higher mortality rate and a smaller body size compared to wild type mice. Therefore, the design of selective inhibitors is needed to prevent any undesired side effects. However, it is challenging to develop selective inhibitors against only BACE1 or BACE2 since they share high sequence similarity (68%). In this dissertation, the ultimate goal is to develop selective BACE1/2 inhibitors by structure-based drug design approaches. Both BACE1 and BACE2 are expressed in the insoluble inclusion body fractions when using E.coli as the expression system. Although there are a few reported protocols for production and purification of both BACE1 and BACE2 from E.coli in the literature, the ones that have reported methods are either time-consuming or costly. Here, the protein production and purification protocols have been optimized for both BACE1 and BACE2 from E.coli inclusion bodies. BACE1 and BACE2 can both be simply refolded in nanopure water without adding any refolding adjuvants, and it takes only 4 to 5 days to refold each enzyme to reach the maximum activity. After refolding, milligram quantities of BACE1 and BACE2 can be purified via chromatography. The purity of the resulting BACE1 and BACE2 purified proteins is greater than 95 % as judged by SDS-PAGE and their resulting specific activities are 126 and 330 µM/min/mg. The steady-state kinetic parameters of BACE1 and BACE2 were determined using a fluorogenic FRET (fluorescence resonance energy transfer)-based peptide substrate derived from the β-secretase cleavage site of the Swedish APP mutation. The purified BACE1 enzyme has a catalytic efficiency (k cat /Km) of 0.58 min –1 µM–1 (k cat = 8.0 ± 0.4 min–1, K m = 18.3 ± 1.6 µM ). The catalytic efficiency ( kcat / Km) of purified BACE2 is 0.23 min–1 µM–1 ( kcat = 5.7 ± 0.5 min–1, Km = 24.8 ± 4.7 µM). The crystallization conditions for both enzymes were also optimized for structural characterization and drug design. In collaboration with Prof. Arun Ghosh’s lab in the chemistry department at Purdue University, structure activity relationship (SAR) studies were performed to identify potent and selective inhibitors against either BACE1 or BACE2. Since the inhibitory constant (Ki) values ranged from subnanomolar to micromolar for each individual compound tested, different kinetic models were applied to determine the Ki values of compounds For the classical inhibitors, the Michaelis - Menten model was used to fit the kinetic data to obtain individual K i value of each compound. The Morrison equation was utilized for potent inhibitors since the assays were performed under the tight binding condition ([E] ≤ [I]). The detailed steps of applying the Morrison equation for curve fitting are also reported here. First and foremost, an accurate active enzyme concentration has to be determined using a titration exper- iment in order to obtain accurate Ki values with the Morrison equation. (Abstract shortened by ProQuest.)

Degree

Ph.D.

Advisors

Mesecar, Purdue University.

Subject Area

Biology|Pharmacology

COinS