Design And Synthesis Of Two-face Bim BH3α-helix Mimetics Based On A Quinazolone Scaffold

Bim is a nonselective BH3-only protein which could interact with all members of the anti-apoptotic proteins, such as Mcl-1/Bcl-2, which is an important mimicking target for inhibitors of Bcl-2proteins. The Bim BH3domain formed a two-face a-helical fold when bound to the anti-apoptotic proteins. So this study focused on the design and synthesis of the efficient structure of two-face a-helix mimetics and obtained compounds which could simutaneously occupy p3and p4pockets of anti-apoptotic proteins on basis of the new scaffold.Tricyclic anthraquinone derivatives had been verified to be two-faced Bim BH3α-helix mimetics. However, other reports have indicated that similar anthraquinones may have potential hepatotoxicity and allergic reaction in animal studies which limiting their druggability. We then simplified the anthraquinone scaffold by maintaining the effective structure. Based on structure-activity relathionship (SAR), we opened the tricyclic anthraquinone and derived monocyclic (rhodanine and pyrogallol) and bicyclic (phthalimide and quinazolone) scaffolds. The binding affinities Ki of these compounds were evaluated using fluorescence polarization assays (FPA) that measured their affinities toward Mcl-1/Bcl-2. The outcome indicated that bicyclic quinazolone was the efficient one to mimic the two-faced binding mode of a-helix and it was worthy to be optimized. Computer-aided docking studies and SAR studies inferred that monocyclic scaffolds lacked sufficient rigidity and the weak hydrophobic and hydrogen bond interactions led to weak affinities toward proteins; the bicyclic phthalimide failed to stablize the scaffold binding mode to mimic the two-faced a-helix; the quinazolon scaffold utilized the bicyclic to mimic the width of the a-helix, and three adjacent hydroxyl groups to form strong hydrogen bonds which maintained its binding mode with proteins. It demonstraed that the quinazolone scaffold was the efficient one to mimic the two-faced a-helix.Previous studies have proved that F69is one of hot spots of Bim which corresponds to p4pocket of Mel-1. However known inhibitors of Bcl-2family proteins failed to occupy the p4pocket of Mcl-1. We expected to explore new a-helix mimetics which could simutaneously occupy p3and p4pockets of Mcl-1/Bcl-2proteins by optimizing the bicyclic scaffold. At the same time, the2-position and3-position of quinazolone scaffold pointed to the two pockets, so we designed the dual substituted compounds9d [6.7,8-trihydroxy-3-(2-hydroxy-5-methylbenzyl)-2-phenylquinazolin-4(3H)-one] and9e [3-(2-hydroxy-5-isopropylbenzyl) -6,7,8-trihydroxy-2-phenylquinazolin-4(3H)-one]. The FP assays indicated that these com-pounds showed better affinities toward Bcl-2and Mcl-1proteins. Especially, the Ki value of9d was123nM toward Mcl-1, and179nM toward Bcl-2which was a promising dual inhibitor. The docking studies showed that the adjacent three hydroxyl groups of9d mimicing D67of Bim a-helix formed hydrogen bonds with R263/N260of Mcl-1. The phenyl group in2-position mimicing F69of Bim a-helix occupied p4pocket of Mcl-1and the benzyl group in3-position mimicing165of Bim a-helix expanded to the p3pocket, besides an additional hydrogen bond was formed between the hydroxyl of benzyl group of9d with T266of Mcl-1. The solubility of compound9d was30mg/mL in water. These results could provide some fundamental insights into the future design and development of Bim BH3a-helix mimetics.