O3‐06‐03: A new drug target for treating Alzheimer's disease (AD): Structural images and models of Abeta‐42 channels and their blockade by small molecule drugs, synthetic peptides, and metal cations

O3-06-03 A NEW DRUG TARGET FOR TREATING ALZHEIMER’S DISEASE (AD): STRUCTURAL IMAGES AND MODELS OF ABETA-42 CHANNELS AND THEIR BLOCKADE BY SMALL MOLECULE DRUGS, SYNTHETIC PEPTIDES, AND METAL CATIONS H. Robert Guy, Yinon Shafrir, Stewart R. Durell, Nelson Arispe, Harvey B. Pollard, National Cancer Institute, NIH, Bethesda, MD, USA; Uniformed Services University School of Medicine, USUHS, Bethesda, MD, USA. Contact e-mail: bg4y@nih.gov Background: Amyloid beta (Abeta) peptides form Ca2þ permeant channels in membranes. Abeta channels can be inhibited by various metals, synthetic peptides, and novel synthetic small molecules, which also inhibit Abeta-induced apoptosis (Diaz et al PNAS(USA) 2009). These agents are considered candidate drugs for AD. We have also obtained freeze-fracture microscopy images of membrane-bound Abeta assemblies with diameters of w7, 11, and 13-14 nm. Here we present new structural models consistent with these results. Methods: Preliminary structural models developed using the PSSHOW program were minimized using CHARMM. Molecular dynamics simulations were then performed using GROMACS for Abeta assemblies embedded in a POPE lipid bilayer. Results: Atomic force microscopy studies have identified large Abeta membrane assemblies in which six peaks surround a central cavity. In our models, each peak corresponds to a hexamer in which the hydrophobic C-terminus third of six Abeta42 monomers comprise an antiparallel beta-barrel located on the membrane surface. Parallel beta-barrels formed by residues 1-13 from all six hexamers form a central pore. We have modeled an insertion process in which the assembly diameters decrease from w15 nm to 7 nm as the 6-stranded barrels reorient from the membrane surface to span the membrane and then merge to form a 36-stranded beta-barrel. The model most consistent with blockage by the small molecule drugs MRS2481 and MRS2485 have two N-terminus 12-stranded parallel beta-barrels (residues 1-13). The C-termini of these barrels meet in the center of the membrane, where 48 imidazol rings of His13 and His14 side chains line the pore. Channel blocking drugs, Zn2þ ions, and histidine-containing synthetic peptides were modeled to bind here to Glu11, His13, and His14 side chains. This central region is surrounded by a 24-stranded antiparallel beta-barrel (residues 17-21). The transmembrane outer layer formed by hydrophobic residues 29-42 initially comprise six hydrophobic 6-stranded beta-barrels; however, these may merge to form a 36-stranded beta-barrel. The latter model is exceptionally stable during MD simulations, and is consistent with all of our other modeling criteria. Conclusions: Drugs that block Aeta channels may be effective in treating Alzheimer’s disease. Our models should be useful in the rational design of such candidate AD drugs.