126b New Twists to the Amyloid Folding Problem Revealed by Solid State NMR and Electron Microscopy of Alzheimer's Beta Amyloid Fibrils

Anant K. Paravastu1, Isam Qahwash2, Wa-Ming Yao3, Richard D. Leapman4, Steven C. Meredith2, and Robert Tycko3. (1) Chemical and Biomedical Engineering, Florida A&M University and Florida State University, 2525 Pottsdamer St., Tallahassee, FL 32310, (2) Biochemistry and Molecular Biology, University of Chicago, 5841 S. Maryland Ave., Chicago, IL 60637, (3) Laboratory of Chemical Physics, National Institute of Diabetes & Digestive & Kidney Diseases, NIH, 9000 Rockville Pike, Building 5, Room 111, Bethesda, MD 20892, (4) National Institute of Biomedical Imaging and Bioengineering, NIH, 9000 Rockville Pike, Bethesda, MD 20892

Self-assembly of peptides into amyloid fibrils has received considerable attention because of the pathological conditions associated with amyloid deposition in vivo. The amyloid folding problem is especially interesting because of the variety of amino acid sequences that can self-assemble into the characteristic cross-beta structural motif. Recently, it was discovered that self-assembly of a single peptide could be driven towards different amyloid fibrils with distinct morphologies and underlying molecular structures (Petkova, A.T., Leapman, R.D., Tycko, R., et. al. Science 307, 262 (2005).). For the 40-residue Alzheimer's beta amyloid peptide (ABeta), we have characterized the range of possible fibril molecular structures by preparing fibrils using a variety of experimental protocols. Solid state NMR was used to assess secondary structures, structural order, and site-specific inter-atomic proximities. Electron microscopy was used to measure fibril dimensions, morphology, and mass. Together, these results are all consistent with folded beta strand molecular conformations within the cross-beta structural motif. Observed structural variations between different ABeta fibrils will be discussed in terms of interfaces between beta sheets within the fibril core, conformations of non-beta-strand regions, and two classes of overall fibril symmetry. This presentation will culminate in the description of a novel 3-fold symmetric ABeta fibril structure, and preliminary results on new ABeta fibrils seeded with human Alzheimer's brain material. The results indicate the importance of multiple distinct but energetically similar pathways for amyloid fibril formation.