-ig a independent set of NMR experiments that confirm cross-b pattern that is observed in diffraction studies of amyloids40 1–42and has been shown to be toxic to neuron cell cultures (Kayed et al., 2003; Petkova et al., 2005). The Ab1–40 and Ab1–42 segments are derived from the enzymatic cleavage of larger amyloid (Bertini et al., 2011; Lu et al., 2013; Luhrs et al., 2005; Paravastu et al., 2008; Petkova et al., 2002, 2006). These full-length structures have been proven nonamenable to X-ray diffractionprecursor proteins (APPs) (Barrett et al., 2012; O’Brien and
Wong, 2011), and single point mutations of the APP locus are usually associated with familial, early-onset AD (Karran et al., studies, presumably due to the presence of flexible N-terminal and loop segments (residues 1–15 and 23–29) that limit the formation of a well-ordered crystal lattice. Solid-state nuclearkey features.
The deposition of amyloid fibrils is a crucial clinical hallmark of a variety of fatal neurodegenerative diseases, including Alzheimer’s disease (AD), Huntington’s, and Parkinson’s disease (Chiti and Dobson, 2006; Selkoe, 1991; Siepe et al., 2012). Specifically for AD, formation of b amyloid (Ab) fibrils and various oligomers consisting of the 40- or 42-residue Ab peptides (Ab1– or Ab , respectively), correlates with disease progression (Sunde et al., 1997). An early study by Lansbury et al. (1995) demonstrated that the C-terminal segment of Ab (named Ab34– 42) produces amyloid fibrils with an antiparallel b sheet structure.
X-ray crystallography conducted on a number of short Ab segments (6–8 residues) revealed a polymorphic group of primarily antiparallel b sheet structures, and further highlighted the importance of a ‘‘steric zipper’’ motif, formed by intercalated side chains of hydrophobic residues, in stabilizing the core of the fibril structure (Colletier et al., 2011; Sawaya et al., 2007). However, the longer Ab10–35 segment may also form parallel, in-register b sheet structures (Benzinger et al., 1998). Several studies focusing on Ab1–40 or Ab1–42 have suggested that the parallel in-register structures are the main species of full-length Ab fibrils‘‘Iowa’’ mutant (D23N) at high resolution (1.2A˚ backbone rmsd). The final models are validated using an gate along themain axis of the fibril, resulting in the characteristicModeling an In-Register, Pa
Fibril Structure Using Solid from Labeled Samples with
Nikolaos G. Sgourakis,1 Wai-Ming Yau,1 and Wei Qiang1,2,* 1Laboratory of Chemical Physics, National Institute of Diabetes and D
MD 20892, USA 2Department of Chemistry, Binghamton University, 4400 Vestal Parkw *Correspondence: email@example.com http://dx.doi.org/10.1016/j.str.2014.10.022
Determining the structures of amyloid fibrils is an important first step toward understanding the molecular basis of neurodegenerative diseases. For b-amyloid (Ab) fibrils, conventional solid-state NMR structure determination using uniform labeling is limited by extensive peak overlap. We describe the characterization of a distinct structural polymorph of Ab using solid-state NMR, transmission electron microscopy (TEM), andRosettamodel building. First, the overall fibril arrangement is established using mass-per-length measurements from TEM. Then, the fibril backbone arrangement, stacking registry, and ‘‘steric zipper’’ core interactions are determined using a number of solid-state NMR techniques on sparsely 13C-labeled samples. Finally, we perform
Rosetta structure calculations with an explicitly symmetric representation of the system. We demonstrate the power of the hybrid Rosetta/NMR approach by modeling the in-register, parallel216 Structure 23, 216–227, January 6, 2015 ª2015 Elsevier Ltd All rigStructure
Resource rallel ‘‘Iowa’’ Ab
State NMR Data
Rosetta estive and Kidney Diseases, National Institutes of Health, Bethesda, y East, Binghamton, NY 13902, USA 2011). Characterization of the high-resolution structures of Ab fibrils has two important impacts to the field of AD. First, it provides crucial information on the molecular mechanism of Ab amyloid formation process, which is believed to disrupt normal neuronal functions and elicit toxicity (Mason et al., 1996, 1999;
Peters et al., 2009; Widenbrant et al., 2006). Second, atomic models could serve as templates for the development of molecules targeting fibril structures, which is one of the clinically tested therapeutic strategies to combat AD (Ladiwala et al., 2012; Petrassi et al., 2000; Sievers et al., 2011).
Biochemical and biophysical characterizations of Ab fibril structures have been performed extensively during the past two decades (Antzutkin et al., 2000, 2002; Balbach et al., 2000;
Benzinger et al., 1998; Bertini et al., 2011; Lansbury et al., 1995; Lu et al., 2013; Paravastu et al., 2008; Petkova et al., 2002, 2006; Qiang et al., 2012; Tycko et al., 2009). These studies typically use aqueous buffers to mimic physiological conditions (pH, temperature, and salt concentration) and therefore serve as good in vitro model systems for Ab fibrils formed around neurons. The sequences of Ab1–40 and Ab1–42 contain two ‘‘amyloidogenic’’ regions at residues 10–22 and 30–40(42) that may form parallel or antiparallel b sheet structures due to their high hydrophobicity (Tycko, 2011). These b sheet structures propa-hts reserved magnetic resonance (NMR) spectroscopy provides a powerful alternative to crystallography for these challenging systems, and several structural models of Ab1–40 fibrils have been determined using a variety of solid-state NMR experiments (Tycko, 2011). These studies revealed a highly diverse range of fibril structures, suggesting that the amyloid state is not uniquely defined by the amyloidogenic sequence, but it is dependent on the precise fibril growth conditions. The types of models that are consistent with the solid-state NMR data include different numbers of cross-b subunits, packing arrangements, and b-stand variations (Lu et al., 2013; Paravastu et al., 2008;