Structural Studies of Alzheimer's beta-Amyloid Fibrils

2012财年在以下方面取得了进展： 1.来自脑组织的淀粉样纤维结构：我们之前已经开发了一种方案，用于从尸检中获得的脑组织中部分纯化淀粉样蛋白，并将这种材料用作从合成的同位素标记的多肽中生长纤维的种子。有了这个方案，我们可以在一个单一的纤维生长步骤中创建1毫克的纤维样本，适合于固态核磁共振和电子显微镜研究，从1克脑组织开始。将这一方案应用于阿尔茨海默病患者的额颞叶和枕叶组织，我们发现该组织中有单一的纤维结构，这是一个令人惊讶的结果。我们已经对这些脑种子纤维进行了大量的固体核磁共振测量和电子显微镜测量，在特定的位置进行了同位素标记，并进行了统一的同位素标记。从这些数据中，我们开发了脑源性纤维的完整结构模型，这是第一个此类模型。脑源性纤维在结构上类似于我们在前几年人工创造和表征的纤维，但有一些独特的特征(例如，整个40个残基的多肽序列在结构上是有序的)。我们还发现，其他阿尔茨海默病患者的脑组织中有不同的纤维结构，反映在我们测量的核磁共振化学位移上的显著差异，以及电子显微镜图像中略有不同的形态。这些结果正在撰写中，准备在2013财年发表。 2.突变的β-淀粉样纤维中的反平行β-折叠结构：我们最近与S.C.Meredith合作，证明了人类β-淀粉样蛋白的Asp23-to-ASN突变体(D23N突变体，或爱荷华州突变体)能够形成含有反平行β-折叠的淀粉样纤维。这是第一次证明全长的多肽或蛋白质可以形成含有反平行(而不是平行)的β-折叠的纤维。2012财年，我们完成并发表了反平行D23N-Abeta纤维的全分子结构模型，揭示了类似的疏水相互作用如何稳定平行或反平行结构。在与NIA的Mattson团队的合作中，我们已经证明了平行和反平行结构在细胞培养中都是神经毒性的。我们已经证明了反平行结构在最终转变为平行结构时是亚稳态的。 3.纤维结构与疾病进展的相关性：使用#1中描述的方案，我们已经开始研究β-淀粉样蛋白纤维结构(如固体核磁共振和电子显微镜所揭示的)与临床病史/病理之间的可能的相关性。在这些实验中，我们与伦敦大学学院的约翰·科林奇教授建立了合作关系，他的团队为我们提供了一组来自有不同病史的AD患者的脑组织样本。过去几个月获得的初步数据表明，这种相关性确实存在，这反过来又表明，淀粉样原纤维结构的细节可能影响疾病的进展。这些实验将在2013财年继续进行。

Progress in FY2012 has been in the following areas: 1. AMYLOID FIBRIL STRUCTURES DERIVED FROM BRAIN TISSUE: We have previously developed a protocol for partial purification of amyloid from brain tissue obtained at autopsy, and for using this material as a "seed" for growing fibrils from synthetic, isotopically-labeled peptide. With this protocol, we can create 1 mg fibril samples suitable for solid state NMR and electron microscopy studies, starting with 1 g of brain tissue, in a single fibril growth step. Applying this protocol to fronto-temporal lobe and occipital lobe tissue from a Alzheimer's disease patient, we found that there is a single fibril structure in this tissue, a surprising result. We have performed numerous solid state NMR measurements and electron microscopy measurements on these brain-seeded fibrils, with isotopic labeling at specific sites and with uniform isotopic labeling. From these data, we have developed a full structural model for brain-derived fibrils, the first of its kind. The brain-derived fibrils are structurally similar to fibrils that we have created synthetically and characterized in previous years, but have some unique features (e.g., the entire 40-residue peptide sequence is structurally ordered). We have also found that other Alzheimer's disease patients have different fibril structures in their brain tissue, reflected in significant differences in NMR chemical shifts in our measurements and somewhat different morphologies in electron microscopy images. These results are being written up for publication in FY2013. 2. ANTIPARALLEL BETA-SHEET STRUCTURE IN MUTANT BETA-AMYLOID FIBRILS: In collaboration with S.C. Meredith, we have recently shown that the Asp23-to-Asn mutant of human beta-amyloid (D23N mutant, or Iowa mutant) is capable of forming amyloid fibrils that contain antiparallel beta-sheets. This is the first demonstration that a full-length peptide or protein could form fibrils that contain antiparallel (rather than parallel) beta-sheets. In FY2012, we completed and published a full molecular structural model for antiparallel D23N-Abeta fibrils, revealing how similar sets of hydrophobic interactions can stabilize either parallel or antiparallel structures. In collaboration with Mattson's group in NIA, we have shown that both parallel and antiparallel structures are neurotoxic in cell cultures. We have shown that antiparallel structures are metastable with respect to eventual conversion to parallel structures. 3. CORRELATIONS BETWEEN FIBRIL STRUCTURE AND DISEASE PROGRESSION: Using the protocols described in #1, we have begun investigating possible correlations between beta-amyloid fibril structure (as revealed by solid state NMR and electron microscopy) and clinical history/pathology. For these experiments, we have established a collaboration with Prof. John Collinge of University College London, whose group has provided us with a set of brain tissue samples from AD patients with distinct histories. Preliminary data, obtained in the past several months, suggest that correlations do exist, which in turn suggests that the details of amyloid fibril structure may influence disease progression. These experiments will continue in FY2013.