Advertisement

Copper stabilizes antiparallel β-sheet fibrils of the amyloid β40 (Aβ40)-Iowa variant

Open AccessPublished:May 06, 2020DOI:https://doi.org/10.1074/jbc.RA119.011955
      Cerebral amyloid angiopathy (CAA) is a vascular disorder that primarily involves deposition of the 40-residue–long β-amyloid peptide (Aβ40) in and along small blood vessels of the brain. CAA is often associated with Alzheimer's disease (AD), which is characterized by amyloid plaques in the brain parenchyma enriched in the Aβ42 peptide. Several recent studies have suggested a structural origin that underlies the differences between the vascular amyloid deposits in CAA and the parenchymal plaques in AD. We previously have found that amyloid fibrils in vascular amyloid contain antiparallel β-sheet, whereas previous studies by other researchers have reported parallel β-sheet in fibrils from parenchymal amyloid. Using X-ray fluorescence microscopy, here we found that copper strongly co-localizes with vascular amyloid in human sporadic CAA and familial Iowa-type CAA brains compared with control brain blood vessels lacking amyloid deposits. We show that binding of Cu(II) ions to antiparallel fibrils can block the conversion of these fibrils to the more stable parallel, in-register conformation and enhances their ability to serve as templates for seeded growth. These results provide an explanation for how thermodynamically less stable antiparallel fibrils may form amyloid in or on cerebral vessels by using Cu(II) as a structural cofactor.

      Introduction

      β-Amyloid peptide (Aβ) fibrillar assembly and deposition in brain parenchymal plaques is a hallmark of Alzheimer's disease (AD), a prevalent neurodegenerative disorder in the elderly population (
      • Selkoe D.J.
      • Hardy J.
      The amyloid hypothesis of Alzheimer's disease at 25 years.
      ). In addition to AD, cerebral amyloid angiopathy (CAA) is also characterized by amyloid deposition in the brain, but in this case in and around the cerebral vasculature (
      • Thal D.R.
      • Griffin W.S.
      • de Vos R.A.
      • Ghebremedhin E.
      Cerebral amyloid angiopathy and its relationship to Alzheimer's disease.
      ). Although AD is associated with neuronal cell death, vascular amyloid deposits in CAA result in the loss of vessel wall integrity and can lead to hemorrhage and stroke.
      The Aβ peptide, a major component of the amyloid deposits in both parenchymal and vascular plaques, is generated by proteolytic cleavage of the transmembrane region of the amyloid precursor protein (APP). APP first undergoes cleavage by either α- or β-secretase. α-Cleavage is nonamyloidogenic, as the target site is within the Aβ domain of APP. β-Cleavage, however, occurs 16 residues upstream from the α-site at the N-terminal end of the Aβ domain. Subsequent cleavage by γ-secretase at the C-terminal end of the Aβ domain of APP releases the Aβ peptide. Aβ peptides ranging from 38 to 50 amino acids in length have been identified (
      • Qi-Takahara Y.
      • Morishima-Kawashima M.
      • Tanimura Y.
      • Dolios G.
      • Hirotani N.
      • Horikoshi Y.
      • Kametani F.
      • Maeda M.
      • Saido T.C.
      • Wang R.
      • Ihara Y.
      Longer forms of amyloid β protein: implications for the mechanism of intramembrane cleavage by γ-secretase.
      ). The 40-residue Aβ40 peptide is the predominant cleavage product representing ∼90% of the total secreted Aβ, although the more toxic 42-residue Aβ42 peptide represents ∼5–10% (
      • Selkoe D.J.
      Alzheimer disease: mechanistic understanding predicts novel therapies.
      ).
      Notable differences between AD-specific parenchymal plaques and CAA-specific vascular deposits have appeared in recent years. First, parenchymal amyloid deposits are preferentially composed of Aβ42 (5), whereas vascular deposits are enriched in Aβ40 (
      • Castaño E.M.
      • Prelli F.
      • Soto C.
      • Beavis R.
      • Matsubara E.
      • Shoji M.
      • Frangione B.
      The length of amyloid-β in hereditary cerebral hemorrhage with amyloidosis, Dutch type: implications for the role of amyloid-β 1-42 in Alzheimer's disease.
      ,
      • Roher A.E.
      • Lowenson J.D.
      • Clarke S.
      • Woods A.S.
      • Cotter R.J.
      • Gowing E.
      • Ball M.J.
      β-Amyloid-(1-42) is a major component of cerebrovascular amyloid deposits: implications for the pathology of Alzheimer-disease.
      ). Second, preferential binding by antibodies (
      • Rutgers K.S.
      • van Remoortere A.
      • van Buchem M.A.
      • Verrips C.T.
      • Greenberg S.M.
      • Bacskai B.J.
      • Frosch M.P.
      • van Duinen S.G.
      • Maat-Schieman M.L.
      • van der Maarel S.M.
      Differential recognition of vascular and parenchymal β amyloid deposition.
      ) and resorufin analogs (
      • Han B.H.
      • Zhou M.L.
      • Vellimana A.K.
      • Milner E.
      • Kim D.H.
      • Greenberg J.K.
      • Chu W.H.
      • Mach R.H.
      • Zipfel G.J.
      Resorufin analogs preferentially bind cerebrovascular amyloid: potential use as imaging ligands for cerebral amyloid angiopathy.
      ) to vascular amyloid has been observed. Moreover, several naturally-occurring familial mutations have been reported to enhance early-onset vascular deposition and cognitive impairment. Two prominent mutations are the Iowa-type (D23N) and Dutch-type (E22Q) mutations (
      • Grabowski T.J.
      • Cho H.S.
      • Vonsattel J.P.
      • Rebeck G.W.
      • Greenberg S.M.
      Novel amyloid precursor protein mutation in an Iowa family with dementia and severe cerebral amyloid angiopathy.
      ,
      • Van Broeckhoven C.
      • Haan J.
      • Bakker E.
      • Hardy J.A.
      • Van Hul W.
      • Wehnert A.
      • Vegter-Van der Vlis M.
      • Roos R.A.
      Amyloid β protein precursor gene and hereditary cerebral hemorrhage with amyloidosis (Dutch).
      ). Both mutant peptides have an acidic residue substituted with an uncharged amine that may lead to structural differences originating from electrostatic interactions. Structure-sensitive dyes have exhibited differences in binding to amyloid deposited from WT and familial mutant Aβ peptides (
      • Condello C.
      • Lemmin T.
      • Stöhr J.
      • Nick M.
      • Wu Y.
      • Maxwell A.M.
      • Watts J.C.
      • Caro C.D.
      • Oehler A.
      • Keene C.D.
      • Bird T.D.
      • van Duinen S.G.
      • Lannfelt L.
      • Ingelsson M.
      • Graff C.
      • et al.
      Structural heterogeneity and intersubject variability of Aβ in familial and sporadic Alzheimer's disease.
      ). Together these findings point toward a structural difference between parenchymal and vascular amyloid.
      Structural studies on WT Aβ40 have found the thermodynamically stable form of the fibril adopts a parallel, in-register cross β-sheet conformation (
      • Tycko R.
      Solid-state NMR studies of amyloid fibril structure.
      ). However, amyloid fibrils are highly polymorphic and can adopt an array of distinct fibril structures (
      • Tycko R.
      Amyloid polymorphism: structural basis and neurobiological relevance.
      ). The ΔE22 mutant of Aβ40 was found to rapidly aggregate, but does not form parallel, in-register structure (
      • Cloe A.L.
      • Orgel J.P.
      • Sachleben J.R.
      • Tycko R.
      • Meredith S.C.
      The Japanese mutant Aβ (ΔE22-Aβ1-39) forms fibrils instantaneously, with low-thioflavin T fluorescence: Seeding of wild-type Aβ1-40 into atypical fibrils by ΔE22-Aβ1-39.
      ). The Aβ40-Iowa mutant was found to adopt a metastable anti-parallel β-sheet architecture (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). The anti-parallel structure was also observed in fibrils seeded from vascular amyloid isolated from transgenic mouse (
      • Xu F.
      • Fu Z.
      • Dass S.
      • Kotarba A.E.
      • Davis J.
      • Smith S.O.
      • Van Nostrand W.E.
      Cerebral vascular amyloid seeds drive amyloid β-protein fibril assembly with a distinct anti-parallel structure.
      ) and rat (
      • Davis J.
      • Xu F.
      • Hatfield J.
      • Lee H.
      • Hoos M.D.
      • Popescu D.
      • Crooks E.
      • Kim R.
      • Smith S.O.
      • Robinson J.K.
      • Benveniste H.
      • Van Nostrand W.E.
      A novel transgenic rat model of robust cerebral microvascular amyloid with prominent vasculopathy.
      ) models of CAA. The findings that Aβ40-Iowa fibrils preferentially form on cerebral blood vessels (
      • Tomidokoro Y.
      • Rostagno A.
      • Neubert T.A.
      • Lu Y.
      • Rebeck G.W.
      • Frangione B.
      • Greenberg S.M.
      • Ghiso J.
      Iowa variant of familial Alzheimer's disease accumulation of posttranslationally modified Aβ D23N in parenchymal and cerebrovascular amyloid deposits.
      ), are linked to severe CAA symptoms in humans (
      • Grabowski T.J.
      • Cho H.S.
      • Vonsattel J.P.
      • Rebeck G.W.
      • Greenberg S.M.
      Novel amyloid precursor protein mutation in an Iowa family with dementia and severe cerebral amyloid angiopathy.
      ), and can form a novel anti-parallel structure (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) suggest a causal relationship between Aβ fibril structure and clinical pathology.
      In parallel with the structural studies on the Aβ peptides, a wide range of work has investigated the propensity of Aβ fibrils to bind metal ions both in vitro and in vivo (
      • Curtain C.C.
      • Ali F.E.
      • Smith D.G.
      • Bush A.I.
      • Masters C.L.
      • Barnham K.J.
      Metal ions, pH, and cholesterol regulate the interactions of Alzheimer's disease Aβ peptide with membrane lipid.
      ,
      • Faller P.
      • Hureau C.
      • La Penna G.
      Metal ions and intrinsically disordered proteins and peptides: from Cu/Zn amyloid-β to general principles.
      ,
      • Faller P.
      Copper and zinc binding to amyloid-β: coordination, dynamics, aggregation, reactivity and metal-ion transfer.
      ,
      • Karr J.W.
      • Szalai V.A.
      Cu(II) binding to monomeric, oligomeric, and fibrillar forms of the Alzheimer's disease amyloid-β peptide.
      ,
      • Atwood C.S.
      • Huang X.
      • Khatri A.
      • Scarpa R.C.
      • Kim Y.S.
      • Moir R.D.
      • Tanzi R.E.
      • Roher A.E.
      • Bush A.I.
      Copper catalyzed oxidation of Alzheimer Aβ.
      ,
      • Schrag M.
      • Crofton A.
      • Zabel M.
      • Jiffry A.
      • Kirsch D.
      • Dickson A.
      • Mao X.W.
      • Vinters H.V.
      • Domaille D.W.
      • Chang C.J.
      • Kirsch W.
      Effect of cerebral amyloid angiopathy on brain iron, copper, and zinc in Alzheimer's disease.
      ,
      • Leskovjan A.C.
      • Lanzirotti A.
      • Miller L.M.
      Amyloid plaques in PSAPP mice bind less metal than plaques in human Alzheimer's disease.
      ,
      • Miller L.M.
      • Wang Q.
      • Telivala T.P.
      • Smith R.J.
      • Lanzirotti A.
      • Miklossy J.
      Synchrotron-based infrared and X-ray imaging shows focalized accumulation of Cu and Zn co-localized with β-amyloid deposits in Alzheimer's disease.
      ). We have recently shown that copper preferentially accumulates in amyloid-containing brain blood vessels compared with parenchymal amyloid plaques in human AD (
      • Zhu X.
      • Victor T.W.
      • Ambi A.
      • Sullivan J.K.
      • Hatfield J.
      • Xu F.
      • Miller L.M.
      • Van Nostrand W.E.
      Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
      ). This observation agrees with previous studies in which copper in the monovalent state was found to be primarily associated with the cerebral vasculature and only weakly bound to Aβ-containing plaques in human AD tissues exhibiting CAA (
      • Schrag M.
      • Crofton A.
      • Zabel M.
      • Jiffry A.
      • Kirsch D.
      • Dickson A.
      • Mao X.W.
      • Vinters H.V.
      • Domaille D.W.
      • Chang C.J.
      • Kirsch W.
      Effect of cerebral amyloid angiopathy on brain iron, copper, and zinc in Alzheimer's disease.
      ). Moreover in a transgenic mouse model (Tg2576) of AD, treatment with the copper chelator tetrathiomolybdate (TTM) was found to significantly reduce both CAA and parenchymal plaque load (
      • Zhu X.
      • Victor T.W.
      • Ambi A.
      • Sullivan J.K.
      • Hatfield J.
      • Xu F.
      • Miller L.M.
      • Van Nostrand W.E.
      Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
      ). Importantly, although the chelator reduced the copper content in parenchymal plaque, it did not change the copper levels in vascular amyloid. These results suggest that copper may be bound more tightly to vascular amyloid, which in turn may reflect a structural difference between parenchymal and vascular Aβ fibrils.
      In the current study, we explore the role of copper–Aβ interactions in the stability of anti-parallel Aβ40-Iowa fibrils. The focus on Aβ40-Iowa stems from its propensity to form metastable anti-parallel fibrils and its ability to enhance early-onset vascular amyloid deposition and cognitive impairment.
      We first address using synchrotron X-ray fluorescence microscopy (XFM) whether copper accumulates in Iowa-type familial CAA as has been observed for sporadic CAA associated with AD (
      • Zhu X.
      • Victor T.W.
      • Ambi A.
      • Sullivan J.K.
      • Hatfield J.
      • Xu F.
      • Miller L.M.
      • Van Nostrand W.E.
      Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
      ). XFM has previously been used to measure zinc and copper localization in parenchymal plaques of both human brains (
      • Miller L.M.
      • Wang Q.
      • Telivala T.P.
      • Smith R.J.
      • Lanzirotti A.
      • Miklossy J.
      Synchrotron-based infrared and X-ray imaging shows focalized accumulation of Cu and Zn co-localized with β-amyloid deposits in Alzheimer's disease.
      ) and mouse models of AD (
      • Leskovjan A.C.
      • Lanzirotti A.
      • Miller L.M.
      Amyloid plaques in PSAPP mice bind less metal than plaques in human Alzheimer's disease.
      ). Here, we find a marked increase in the amount of copper co-localized with Aβ in blood vessels from brain originating from patients diagnosed with Iowa-type familial CAA compared with controls. The level of copper is comparable with that found in sporadic CAA.
      We next address whether Cu(II) can bind to and stabilize anti-parallel Aβ40-Iowa fibrils. Both FTIR and NMR approaches have previously been used to distinguish parallel and anti-parallel β-sheet structure (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ,
      • Xu F.
      • Fu Z.
      • Dass S.
      • Kotarba A.E.
      • Davis J.
      • Smith S.O.
      • Van Nostrand W.E.
      Cerebral vascular amyloid seeds drive amyloid β-protein fibril assembly with a distinct anti-parallel structure.
      ,
      • Tang T.C.
      • Hu Y.
      • Kienlen-Campard P.
      • El Haylani L.
      • Decock M.
      • Van Hees J.
      • Fu Z.
      • Octave J.N.
      • Constantinescu S.N.
      • Smith S.O.
      Conformational changes induced by the A21G Flemish mutation in the amyloid precursor protein lead to increased Aβ production.
      ). We show that binding of Cu(II) can stabilize anti-parallel Aβ40-Iowa fibrils in vitro and block the transition to the parallel, in-register conformation. In addition, we find that the presence of Cu(II) enhances the ability for anti-parallel Aβ40-Iowa fibrils to seed the growth of new fibrils by the rapid addition of monomeric Aβ. These results raise the possibility that metal ions play an important role in influencing the structure and stability of Aβ fibrils that occur in amyloid deposits in the human brain.

      Results

      Copper binding to human Aβ40-Iowa and sporadic vascular amyloid

      Using XFM we first compare the copper content in vascular amyloid in human sporadic CAA and Iowa-type familial CAA relative to control vessels lacking amyloid (Fig. 1). Copper has previously been shown to accumulate in vascular amyloid associated with sporadic CAA formed primarily by Aβ40-WT (
      • Schrag M.
      • Crofton A.
      • Zabel M.
      • Jiffry A.
      • Kirsch D.
      • Dickson A.
      • Mao X.W.
      • Vinters H.V.
      • Domaille D.W.
      • Chang C.J.
      • Kirsch W.
      Effect of cerebral amyloid angiopathy on brain iron, copper, and zinc in Alzheimer's disease.
      ,
      • Zhu X.
      • Victor T.W.
      • Ambi A.
      • Sullivan J.K.
      • Hatfield J.
      • Xu F.
      • Miller L.M.
      • Van Nostrand W.E.
      Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
      ). The studies below focus on Aβ40-Iowa because it has an enhanced ability to form vascular amyloid in human brain, which may be related to its unique anti-parallel fibril structure that is known to form in solution. The presence of amyloid in brain sections was detected by thioflavin S binding (Fig. 1, A–C). The relative amount of copper bound to Aβ was then determined by normalizing to the protein content in the vessels, which was separately estimated by FTIR microspectroscopy (Fig. 1G). Vessels from control cases without amyloid show very little copper present. However, in the CAA vessels, the copper is significantly elevated (p < 0.05) compared with the controls. These results show that copper accumulates in both sporadic and Iowa-type CAA. Moreover, because there are other proteins in brain tissue with high affinity copper-binding sites (
      • Barritt J.D.
      • Viles J.H.
      Truncated amyloid-β(11-40/42) from Alzheimer disease binds Cu2+ with a femtomolar affinity and influences fiber assembly.
      ), the observation of increased copper binding to vascular amyloid (relative to the control vessels) reflects the ability of these Aβ deposits to bind copper tightly.
      Figure thumbnail gr1
      Figure 1X-ray fluorescence microscopy of copper localization in vascular amyloid. Thioflavin S staining (A–C) and copper XFM images (D–F) of vessels in human control, sporadic CAA, and Iowa-type familial CAA. Scale bars are 20 μm. The relative copper content (G) was determined by normalizing to the protein content at each pixel and shows that copper is significantly elevated (*, p < 0.05) over the controls using a Student's paired t test with a two-tailed distribution. The protein content at each pixel was determined by integrating the amide II protein peak using FTIR microspectroscopy (see “Materials and methods”).

      Rapid fibril formation of Aβ40-Iowa

      Aβ40-Iowa is unusual for being able to form anti-parallel fibrils (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) as well as its rapid fibrillization kinetics (
      • Hatami A.
      • Monjazeb S.
      • Milton S.
      • Glabe C.G.
      Familial Alzheimer's disease mutations within the amyloid precursor protein alter the aggregation and conformation of the amyloid- β peptide.
      ). Thioflavin T fluorescence is a common method for monitoring fibril formation of Aβ peptides. Fig. 2A compares the fluorescence increase due to fibril formation of Aβ40-Iowa with that of Aβ40-WT at 24 °C and a 100 μm Aβ concentration. For Aβ40-WT there is a typically a lag phase before the increase in fibril formation, which is associated with the formation of small oligomers and protofibrils. Under the conditions used, Aβ40-WT does not exhibit fibril formation for several days, whereas Aβ40-Iowa shows a strong immediate increase in fluorescence after warming monomeric peptide to 24 °C.
      Figure thumbnail gr2
      Figure 2Rapid fibril formation of Aβ40-Iowa. A, thioflavin T fluorescence of Aβ40-Iowa (100 μm) in the presence and absence of Cu(II) at 24 °C. Copper was added to monomeric Aβ40-Iowa at Cu(II):Aβ molar ratios of 1:10, 1:5, and 1:2. The thioflavin T fluorescence of Aβ40-WT in the absence of copper under the same conditions is shown for comparison (dashed line). B–D, single touch AFM images of the Aβ40-Iowa peptides after incubation of 10 (B), 30 (C), and 60 (D) min at room temperature reveal oligomers associating into protofibrils and fibrils. The scale bars for the AFM figures are 50 (B), 50 (C), and 400 nm (D).
      Intermediates along the pathway to mature fibrils can be imaged using atomic force microscopy (
      • Mastrangelo I.A.
      • Ahmed M.
      • Sato T.
      • Liu W.
      • Wang C.
      • Hough P.
      • Smith S.O.
      High-resolution atomic force microscopy of soluble Aβ42 oligomers.
      ,
      • Fu Z.
      • Aucoin D.
      • Davis J.
      • Van Nostrand W.E.
      • Smith S.O.
      Mechanism of nucleated conformational conversion of Aβ42.
      ). Within minutes Aβ40-Iowa monomers associate to form small oligomers that have an elongated appearance. Protofibrils are observed after 30 min (Fig. 2C) and distinct fibrils are observed within an hour (Fig. 2D). As with WT Aβ peptides, the protofibrils appear to result from association of oligomers, and fibrils reflect maturation of the protofibrils as cross-β-sheet structure forms (
      • Fu Z.
      • Aucoin D.
      • Davis J.
      • Van Nostrand W.E.
      • Smith S.O.
      Mechanism of nucleated conformational conversion of Aβ42.
      ).
      For the studies below, anti-parallel fibrils were formed by allowing the peptides to associate in a quiescent fashion at room temperature for a 24-h period prior to assessing the influence of copper on the structure of the fibrils. For Aβ40-WT, if copper is added at the monomer stage, the peptides aggregate into nonfibrillar species in a Cu(II)-dependent fashion (
      • Jun S.M.
      • Saxena S.
      The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
      ,
      • Pedersen J.T.
      • Østergaard J.
      • Rozlosnik N.
      • Gammelgaard B.
      • Heegaard N.H.
      Cu(II) mediates kinetically distinct, non-amyloidogenic aggregation of amyloid-ß peptides.
      ,
      • Chen W.T.
      • Liao Y.H.
      • Yu H.M.
      • Cheng I.H.
      • Chen Y.R.
      Distinct effects of Zn2+, Cu2+, Fe3+, and Al3+ on amyloid-β stability, oligomerization, and aggregation: amyloid-β destabilization promotes annular protofibril formation.
      ). Fibrils are able to form at low Cu(II):Aβ ratios, whereas aggregates form at high ratios, with a transition point at a molar ratio of ∼1:2 Cu(II):Aβ (
      • Jun S.M.
      • Saxena S.
      The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
      ,
      • Chen W.T.
      • Liao Y.H.
      • Yu H.M.
      • Cheng I.H.
      • Chen Y.R.
      Distinct effects of Zn2+, Cu2+, Fe3+, and Al3+ on amyloid-β stability, oligomerization, and aggregation: amyloid-β destabilization promotes annular protofibril formation.
      ). For Aβ40-Iowa (Fig. 2A), the Cu(II)-induced aggregation of Aβ monomers is reflected in decreased thioflavin T fluorescence as the Cu(II):Aβ ratio is increased. Interestingly, the initial rates of thioflavin T increase are all faster with Cu(II) than without Cu(II). We show below that this increase in rate results from Aβ40-Iowa fibril seeds that form rapidly and are stabilized by bound Cu(II). As a consequence, the studies below target the influence of copper binding on fibril stability rather than on the pathway to fibril formation.

      Distinguishing parallel and anti-parallel Aβ fibril structure using FTIR and NMR

      Because the anti-parallel structure is not commonly observed for mature Aβ fibrils, two approaches were taken to distinguish anti-parallel and parallel, in-register fibril conformations. First, we introduce a novel FTIR method to enhance the difference between parallel and anti-parallel β-sheet structures in FTIR spectra. Second, we use solid-state NMR spectroscopy in combination with specific 13C labeling to distinguish these two conformations of Aβ. Together these methods allow us to follow the global (FTIR) and local (NMR) influence of Cu(II) on Aβ40-Iowa fibril structure and stability.
      The amide I region (1600-1700 cm−1) of the FTIR spectrum has previously been used to assess the presence of parallel and anti-parallel β-sheet structures in Aβ peptides (
      • Fu Z.
      • Aucoin D.
      • Davis J.
      • Van Nostrand W.E.
      • Smith S.O.
      Mechanism of nucleated conformational conversion of Aβ42.
      ,
      • Paul C.
      • Wang J.
      • Wimley W.C.
      • Hochstrasser R.M.
      • Axelsen P.H.
      Vibrational coupling, isotopic editing, and β-sheet structure in a membrane-bound polypeptide.
      ,
      • Petty S.A.
      • Decatur S.M.
      Experimental evidence for the reorganization of β-strands within aggregates of the Aβ(16-22) peptide.
      ) and in APP (
      • Tang T.C.
      • Hu Y.
      • Kienlen-Campard P.
      • El Haylani L.
      • Decock M.
      • Van Hees J.
      • Fu Z.
      • Octave J.N.
      • Constantinescu S.N.
      • Smith S.O.
      Conformational changes induced by the A21G Flemish mutation in the amyloid precursor protein lead to increased Aβ production.
      ). These studies generally made use of a single 13C=O label in the Aβ sequence that results in an isotope-induced splitting into lower and higher frequency bands. For β-sheet secondary structure, a diagnostic band at ∼1626-1630 cm−1 splits into bands at ∼1630-1636 and ∼1605-1615 cm−1. The intensity of the lower frequency band is characteristically greater for the anti-parallel β-sheet than for parallel β-sheet (
      • Fu Z.
      • Aucoin D.
      • Davis J.
      • Van Nostrand W.E.
      • Smith S.O.
      Mechanism of nucleated conformational conversion of Aβ42.
      ,
      • Paul C.
      • Wang J.
      • Wimley W.C.
      • Hochstrasser R.M.
      • Axelsen P.H.
      Vibrational coupling, isotopic editing, and β-sheet structure in a membrane-bound polypeptide.
      ,
      • Petty S.A.
      • Decatur S.M.
      Experimental evidence for the reorganization of β-strands within aggregates of the Aβ(16-22) peptide.
      ). In the current study, 13C=O labels have been incorporated at four positions to enhance the intensity difference between parallel and anti-parallel β-sheets. Because the Aβ peptide has two distinct hydrophobic stretches that form β-strands, we introduced two 13C labels in the N-terminal LVFFA sequence at Leu-17 and Ala-21 and two 13C labels in the C-terminal GLMVG sequence at Gly-33 and Gly-37. We refer to this peptide as LAGG-labeled Aβ40.
      To illustrate the effect of 13C labeling, we prepared Aβ40-WT and Aβ40-Iowa fibrils having parallel or anti-parallel β-sheet structures, respectively. Aβ40-WT fibrils having a parallel, in-register orientation are readily formed at 37 °C under agitation (200 rpm). The amide I region of Aβ40-WT fibrils without isotopic labeling exhibits a strong β-sheet band at ∼1626 cm−1 (Fig. 3A, red trace). In LAGG-labeled Aβ40-WT, the 1626 cm−1 band shifts to slightly higher frequency (1630 cm−1), and a new resonance appears at ∼1607 cm−1 (Fig. 3A, black trace). The intensity of the 1607 cm−1 band is roughly proportional to the number of 13C=O labeled carbons having β-sheet structure.
      Figure thumbnail gr3
      Figure 3Distinguishing anti-parallel and parallel fibrils. A, amide I and II regions of Aβ40-WT fibrils with (black) and without (red) LAGG 13C labeling after 3 weeks of incubation at 37 °C with strong agitation (200 rpm). The amide I region (1600-1700 cm−1) is sensitive to the secondary structure with the observed band between ∼1626 and 1634 cm−1 being characteristic of β-sheet. B, amide I and II regions of Aβ40-Iowa fibrils with (black) and without (red) LAGG 13C labeling formed at room temperature for 24 h under quiescent conditions.
      In contrast, the influence of 13C labeling on anti-parallel Aβ40-Iowa fibrils is markedly different (Fig. 3B). In this case, anti-parallel nuclei were formed at low temperature (6 °C) to favor the anti-parallel structure as described by Qiang et al. (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) (see “Materials and methods”). The sample was then incubated at 22 °C and fibrils were allowed to elongate for 24 h under quiescent conditions. The spectrum of unlabeled anti-parallel Aβ40-Iowa fibrils exhibits a β-sheet band at 1634 cm−1 (red trace). However, the incorporation of 13C=O labels using the LAGG labeling scheme increases the splitting of the β-sheet band into a high frequency component at 1640 cm−1 and a lower frequency component at 1605 cm−1 (black trace). With the LAGG labeling scheme, the intensity of the lower frequency band is substantially greater than with a single 13C label used in earlier studies. Moreover, in the anti-parallel Aβ40-Iowa fibrils the 1605 cm−1 band is much more intense than the 1640 cm−1 band, despite the fact that only four backbone C=O groups are 13C-labeled. In the studies below, FTIR provides a rapid method to follow the time evolution of fibril formation and assess changes in the contribution of parallel and anti-parallel structure as mature fibrils form.
      The second approach for distinguishing anti-parallel and parallel in-register fibril conformations involves two-dimensional solid-state NMR spectroscopy, which probes internuclear 13C distances that are specific to parallel or anti-parallel structure. The experiments described below were designed to assess the presence of both parallel, in-register and anti-parallel structures in the same sample.
      For these experiments, two Aβ40-WT or two Aβ40-Iowa peptides containing different 13C labels are co-mixed. The first peptide contains [2-13C]Ala-30 and [1-13C]Val-36, which together provide a probe of close distances between β-strands on adjacent peptides in anti-parallel fibrils. In the 2D 13C NMR experiment, the NMR resonances along the diagonal of the 2D spectrum match those observed in a 1D 13C NMR spectrum, whereas the off-diagonal cross-peaks arise from 13C dipolar couplings of 13C sites that are within ∼6 Å of each other. The anti-parallel fibril structure determined of Aβ40-Iowa using solid-state NMR shows that the 13C labels described above on Ala-30 and Val-36 are closely packed between neighboring strands in the fibril (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). The second peptide contains [3-13C]Ala-30. In parallel, in-register fibrils, Ala-30 on one peptide packs against Ala-30 on the neighboring peptide resulting in short distances between [2-13C]Ala-30 and [3-13C]Ala-30 when the two peptides are mixed as monomers prior to fibril formation. That is, the 13C labeling scheme has the potential to reveal inter-strand Ala-30–Val-36 cross-peaks characteristic of anti-parallel fibrils, and Ala-30–Ala-30 cross-peaks characteristic of parallel, in-register fibrils when the two peptides are mixed in one sample.
      Fig. 4A presents the 2D 13C NMR spectrum of Aβ40-WT fibrils. These fibrils were prepared in the same fashion as those in Fig. 3A and have a parallel, in-register structure. The region boxed in blue exhibits cross-peaks for dipolar couplings between [2-13C]Ala-30 and [3-13C]Ala-30 if these 13C sites are within 6 Å. In parallel fibril structures these 13C labels are typically ∼4-5 Å apart, whereas in the anti-parallel fibril structure of Aβ40-Iowa (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) these 13C labels are >10 Å apart. The observation of an intense cross-peak between these 13C labels in the 2D 13C NMR spectrum is consistent with the parallel, in-register fibril structure. In Fig. 4C rows are taken through the diagonal resonances of [2-13C]Ala-30 (red) and [3-13C]Ala-30 (blue). The rows better illustrate the cross-peaks characteristic of parallel, in-register structure (blue) and the absence of anti-parallel structure (red). It is instructive to note that these experiments with single 13C labels within specific amino acids provide an advantage in allowing unambiguous identification and assignments of the cross-peaks to specific atoms. NMR spectra using U-13C-labeled amino acids yield intense cross-peaks from directly bonded 13C atoms, but often are crowded making assignments more challenging.
      Figure thumbnail gr4
      Figure 4Distinguishing antiparallel and parallel fibrils. A, region of the 2D NMR spectrum of parallel Aβ40-WT fibrils. The blue box indicates the position of a cross-peak expected to arise between the [2-13C]Ala-30 and [3-13C]Ala-30 diagonal resonances. Cross-peaks marked with an asterisk are artifacts arising from MAS side bands. B, region of the 2D NMR spectrum of anti-parallel Aβ40-Iowa fibrils. The red box indicates the position of a cross-peak expected to arise between the [1-13C]Val-36 and [2-13C]Ala-30 diagonal resonances. C, rows are shown through the diagonal resonances as in A. The regions shown correspond to the colored boxes. D, rows are shown through the diagonal resonances as in B. E, integration of the cross-peaks diagnostic of parallel and anti-parallel structure. The total intensity for the regions of containing the parallel and anti-parallel cross-peaks was normalized to 1. The intensities of the parallel cross-peak (n = 2) in C, and the anti-parallel cross-peak (n = 3) in D were signficant (p < 0.05) relative to the noise using a Student's paired t test with a two-tailed distribution.
      In a fashion similar to the FTIR experiments, the Aβ40-Iowa fibrils were prepared under low temperature (∼6 °C) and low salt (10 mm sodium phosphate buffer) conditions that favor anti-parallel structures and then incubated for 24 h at room temperature under quiescent conditions. This sample was then lyophilized for solid-state NMR measurements. The 2D 13C NMR spectrum of 24 h Aβ40-Iowa fibrils is shown Fig. 4B. The boxed region in red exhibits cross-peaks for dipolar couplings between [2-13C]Ala-30 and [1-13C]Val-36. The appearance of a cross-peak in the 2D spectrum between these labels indicates that the 13C-sites are within 6 Å, consistent with the anti-parallel fibril structure of Aβ40-Iowa (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) where the 13C labels are ∼4-5 Å apart. In fibril structures where the β-strands have a parallel, in-register orientation, these labels are >10 Å apart. Rows through the region of the anti-parallel and parallel cross-peaks are shown in Fig. 4D. Integration of the cross-peak intensities for samples that are predominantly parallel and anti-parallel indicates that the parallel cross-peaks are roughly twice as intense as the anti-parallel cross-peaks (Fig. 4E). In the experiments below, the 13C probes allow us to make comparative measurements of the contribution of anti-parallel and parallel, in-register structures in fibrils with or without bound Cu(II).

      Influence of copper on the stability of anti-parallel Aβ40-Iowa fibrils

      In this section we describe the use of both FTIR and NMR spectroscopy to determine the ability of Cu(II) to stabilize anti-parallel Aβ40-Iowa fibrils. Anti-parallel Aβ40-Iowa fibrils prepared at low temperature (6 °C) using a filtering protocol were previously found to be metastable (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). In this earlier study, the thermodynamically preferred structure corresponding to parallel, in-register fibrils was prepared using a sonication-incubation protocol at 6 °C. For our experiments, anti-parallel Aβ40-Iowa fibrils were prepared at low temperature as described above and under “Materials and methods.” However, an increase in temperature to 37 °C along with shaking was then introduced to induce the conversion to the parallel, in-register conformation.
      In Fig. 5, A and B, the FTIR spectrum of the anti-parallel Aβ40-Iowa fibrils containing the [13C]LAGG labeling is shown in black. These anti-parallel fibrils were then incubated at 37 °C with shaking and FTIR spectra were obtained as a function of the incubation time. The spectra reveal a gain in intensity and a frequency shift of the 1640 cm−1 band to 1636 cm−1 (Fig. 5, A and B). The gain of intensity relative to the 1606 cm−1 band, which remains constant, suggests a small conversion to the parallel fibril structure (Fig. 5, A and B).
      Figure thumbnail gr5
      Figure 5FTIR spectroscopy and TEM of fibril formation of Aβ40-Iowa. A, evolution of Aβ40-Iowa FTIR spectra as a function of time and temperature. B, expansion of the A. The amide I region of Aβ40-Iowa (black) is shown after incubation for 24 h at room temperature under quiescent conditions. The temperature was then increased to 37 °C and the sample was shaken at 170 rpm. FTIR spectra were obtained after 24 (orange), 48 (red), 72 (green), and 96 h (blue). These time points are designated 24 + 24 h, 24 + 48 h, 24 + 72 h, and 24 + 96 h, respectively. B is an expansion of A. C, influence of Cu(II) on the evolution of Aβ40-Iowa at 37 °C. D is an expansion of the C. Cu(II) was added after 24 h of room temperature incubation and the sample was then further incubated at 37 °C with shaking. FTIR spectra were obtained after Cu(II) addition (black), 24 + 24 h (orange), 24 + 48 h (red), 24 + 72 h (green), and 24 + 96 h (blue). TEM images obtained after 24 h of room temperature incubation (E), after 24 + 96 h without copper (F), and after 24 + 96 h with Cu(II) (G and H) show the presence of fibrils. The scale bars in (E–H) are 100 nm.
      To observe the influence of copper on the FTIR spectra of [13C]LAGG-labeled Aβ40-Iowa, we added Cu(II) in a 1:2 Cu(II):Aβ ratio to the solution after the 24-h period of incubation at room temperature. The difference between these studies and previous studies on the influence of copper on Aβ40-WT fibrillization is that copper is added to pre-formed fibrils rather than to monomeric Aβ. Here, we test the ability of Cu(II) to stabilize the preformed anti-parallel structure rather than to induce a specific conformation. The FTIR spectrum obtained just before the sample was moved to higher temperature shows a slight broadening of the 1640-1644 cm−1 band and a loss of intensity in the 1605 cm−1 band (Fig. 5, C and D). The broadening may be associated with Cu(II)-induced aggregation of the Aβ peptides (
      • Jun S.M.
      • Saxena S.
      The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
      ,
      • Smith D.P.
      • Ciccotosto G.D.
      • Tew D.J.
      • Fodero-Tavoletti M.T.
      • Johanssen T.
      • Masters C.L.
      • Barnham K.J.
      • Cappai R.
      Concentration dependent Cu2+ induced aggregation and dityrosine formation of the Alzheimer's disease amyloid-beta peptide.
      ). However, after shifting the incubation temperature to 37 °C, the 1605 and 1636 cm−1 bands narrow as a function of incubation time indicative of a transition to a more homogeneous fibril conformation. In contrast to the changes in the FTIR spectra without copper added, the 1605 cm−1 band increases in intensity relative to the 1636 cm−1 band. This increase suggests that the anti-parallel structure is being stabilized by copper binding.
      The FTIR time series highlights the advantages and disadvantages of the FTIR approach. The advantage is that it provides a rapid assay of global structure, whereas the disadvantage is that it is only semi-quantitative.
      To characterize the influence of Cu(II) binding on fibrillar morphology, we imaged the solution with negative stain transmission EM (TEM) after 24 h of room temperature incubation without copper, after 24 + 96 h of incubation without copper, and after 24 + 96 h of incubation with Cu(II) (Fig. 5, E–H). TEM images show the presence of fibrils at all three time points. After 24 h of incubation without copper, the fibrils have a curvilinear appearance and some protofibrillar fragments are visible (Fig. 5E). After 24 + 96 h, the fibrils are longer but have the same appearance as at 24 h (Fig. 5F). With the addition of Cu(II), straight, branched fibrils are observed (Fig. 5, G and H). In addition, no low molecular weight amyloid species are observed in the supernatant after centrifugation of the copper-bound fibrils by UV-visible spectroscopy (see below). Together, TEM and UV-visible spectroscopy indicate that a majority of the Aβ has converted to fibrils and that the anti-parallel character is not due to monomeric or oligomeric species remaining in solution.
      Solid-state NMR measurements provide a way to directly compare the contribution of parallel, in-register and anti-parallel fibrils after incubation of Aβ40-Iowa. We first obtained the 2D NMR spectrum of Aβ40-Iowa after the initial 24 h incubation at room temperature. Rows from the 2D NMR spectrum (Fig. 6C) reveal only anti-parallel fibrils are present, similar to the data presented in Fig. 4, A and C. After incubation of this sample for 96 h (Fig. 6, A and D), cross-peaks are observed that are characteristic of the anti-parallel structure (red) as well as parallel, in-register structure (blue). The appearance of the cross-peak reporting on parallel structure is consistent with the studies of Qiang et al. (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) showing that the anti-parallel fibrils are metastable and can convert to parallel, in-register fibrils (
      • Sgourakis N.G.
      • Yau W.M.
      • Qiang W.
      Modeling an in-register, parallel “Iowa“ Aβ fibril structure using solid-state NMR data from labeled samples with rosetta.
      ). In our experiments, we found that a key determinant in the transition from anti-parallel to parallel, in-register fibrils was the strength of agitation. Here we employ rotary shaking of 170-200 rpm to induce conversion to parallel fibrils. Anti-parallel Aβ40-Iowa fibrils under quiescent or gentle shaking conditions (<100 rpm) remain anti-parallel for weeks even at 37 °C.
      Figure thumbnail gr6
      Figure 6Two-dimensional 13C solid-state NMR of Aβ40-Iowa. Two-dimensional 13C DARR NMR spectrum of Aβ40-Iowa fibrils obtained after 96 h of incubation at room temperature under quiescent conditions without Cu(II) (A) and with Cu(II) (B). The fibrils were formed from an equimolar mixture of two 13C-labeled peptides; one containing [2-13C]Ala-30 and [1-13C]Val-36, and one containing [3-13C]Ala-30. The blue boxes indicate the position of cross-peaks expected to arise between the [2-13C]Ala-30 and [3-13C]Ala-30 diagonal resonances, diagnostic of parallel, in-register fibril structure, whereas the red boxes indicate the position of cross-peaks expected to arise between the [1-13C]Val-36 and [2-13C]Ala-30 diagonal resonances, diagnostic of anti-parallel fibril structure. Cross-peaks marked with an asterisk are artifacts arising from MAS side bands. C, rows are shown through the diagonal resonances of [2-13C]Ala-30 (red) and [3-13C]Ala-30 (blue) of the 2D NMR spectrum after the initial 24 h of incubation without Cu(II). These rows are duplicate experiments of those shown in D indicating the reproducibility of these measurements. The regions shown correspond to the colored boxes. D, rows are shown through the diagonal resonances as in A after incubation of the Aβ40-Iowa sample for 24 + 96 h. E, rows are shown through the diagonal resonances as in B after incubation of the Aβ40-Iowa sample for 24 + 96 h with Cu(II) added after the first 24 h incubation step at room temperature. F, integration of the cross-peaks for parallel and anti-parallel structure. The total intensity for the regions of containing the parallel and anti-parallel cross-peaks was normalized to 1. The intensities of the anti-parallel cross-peak (n = 3) in C, the anti-parallel and parallel cross-peaks (n = 3) in D, and the anti-parallel cross-peak (n = 2) in E were significant (p < 0.05) relative to the noise using a Student's paired t test with a two-tailed distribution.
      To assess the dependence of Cu(II) on the anti-parallel to parallel transition, copper was added at a 1:2 Cu(II):Aβ ratio to fibrils incubated for 24 h. This sample was then incubated for an additional 96 h before lyophilization and solid-state NMR measurements. In this case (Fig. 6E), only the anti-parallel cross-peak is observed (red row). The absence of a parallel-specific cross-peak (blue row) shows that Cu(II) ions stabilize the anti-parallel fibril structure.

      Influence of Cu(II) on Aβ40-Iowa fibril growth

      The observation that Cu(II) can stabilize the anti-parallel structure raises the question of whether it can influence the rate of fibril growth. To address this question, we incubated Aβ40-Iowa monomer in the presence or absence of Cu(II) ions and used these fibrils as seeds for templated fibril growth. Aβ40-Iowa fibrils were prepared as described above, either with or without addition of 1:2 Cu(II):Aβ at the 24-h time point, and then allowed to incubate at 37 °C for 2 weeks. After incubation, fibrils were sonicated to develop fibril seeds prior to the addition of fresh Aβ40-Iowa monomer. Aβ40-Iowa peptide monomers either alone or with the addition of 5, 10, 15, or 20% (w/w) Aβ40-Iowa seeds were incubated at room temperature under quiescent conditions. Fibril elongation was followed by thioflavin-T fluorescence (Fig. 7).
      Figure thumbnail gr7
      Figure 7Influence of Cu(II) on Aβ fibril growth from Aβ40-Iowa seeds with Aβ40-Iowa monomer. A and B, thioflavin T fluorescence of Aβ40-Iowa monomer (100 μm) in the presence of Aβ40-Iowa seeds with (A) or without (B) bound Cu(II). The seeds were added at levels of 5, 10, 15, or 20% (w/w) relative to the monomer concentration and the mixtures was incubated at room temperature under quiescent conditions. Seeded growth was observed under all conditions in A, but only with 15 and 20% seeds in B.
      In both the Cu(II)-free and Cu(II)-containing samples, Aβ40-Iowa fibrillization is influenced by the presence of Aβ40-Iowa fibril seeds compared with monomer alone (dark blue trace) (Fig. 7, A and B). In both cases, the fluorescence intensity increased as a function of incubation time. However, the fluorescence traces corresponding to the Cu(II) and Cu(II)-free seeded samples are different. Fluorescence traces of samples with the addition of Aβ40-Iowa fibril seeds in the presence of Cu(II) (Fig. 7A) exhibit an exponential increase in fluorescence intensity followed by a plateau. The rate of increase of the seeded samples is comparable suggesting that the available Aβ40-Iowa monomers are rapidly incorporated into fibrils as a function of incubation time and the monomer population is depleted after ∼6 h.
      Fluorescence traces corresponding to Cu(II)-free samples show an increase in fluorescence with Aβ40-Iowa monomer added to 15 or 20% (w/w) Aβ40-Iowa fibril seeds compared with Aβ40-Iowa monomer alone (dark blue trace) (Fig. 7B). Seeded samples corresponding to 15 and 20% (w/w) Aβ40-Iowa fibrils in the absence of Cu(II) have a visible increase in fluorescence followed by a plateau at ∼24 h. Seeded samples with 5 and 10% (w/w) the Aβ40-Iowa fibrils in the absence of Cu(II) exhibit an initial linear increase in fluorescence similar to what is observed for the Aβ40-Iowa monomer alone and do not plateau within 24 h indicating that the fibrils are continuing to form. An overall comparison of the two fluorescence traces shows that the Aβ40-Iowa seeds that lack Cu(II) do not template fibril growth as well as those with Cu(II).

      Interaction of His-6, His-13, and His-14 with Cu(II) in anti-parallel Aβ40-Iowa fibrils

      The observation that Cu(II) binding stabilizes anti-parallel structure raises the question of whether there is a specific Cu(II)-binding site. In this section, we probe the interaction of Cu(II) with His-6, His-13, and His-14 using specifically 13C-labeled Aβ. These three histidine residues have been shown to coordinate Cu(II) in both Aβ40 (
      • Shin B.K.
      • Saxena S.
      Direct evidence that all three histidine residues coordinate to Cu(II) in amyloid-β(1-16).
      ,
      • Hou L.
      • Zagorski M.G.
      NMR reveals anomalous copper(II) binding to the amyloid A beta peptide of Alzheimer's disease.
      ,
      • Parthasarathy S.
      • Long F.
      • Miller Y.
      • Xiao Y.
      • McElheny D.
      • Thurber K.
      • Ma B.
      • Nussinov R.
      • Ishii Y.
      Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
      ) and Aβ42 (
      • Karr J.W.
      • Szalai V.A.
      Cu(II) binding to monomeric, oligomeric, and fibrillar forms of the Alzheimer's disease amyloid-β peptide.
      ,
      • Sarell C.J.
      • Syme C.D.
      • Rigby S.E.
      • Viles J.H.
      Copper(II) binding to amyloid-β fibrils of Alzheimer's disease reveals a picomolar affinity: stoichiometry and coordination geometry are independent of Aβ oligomeric form.
      ). Measurements of His-Cu(II) interactions were made by adding Cu(II) at a 1:2 molar ratio of Cu(II):Aβ after 24 h of room temperature incubation and allowing the fibrils to incubate an additional 96 h at 37 °C. Under these conditions, Cu(II) binds quantitatively to the Aβ fibrils (Fig. 8A). A comparison was made to parallel Aβ40-Iowa fibrils (see “Materials and methods”). There was no significant difference in Cu(II) binding consistent with the observations on Aβ40-WT that Cu(II) binds strongly to parallel Aβ fibrils (
      • Parthasarathy S.
      • Long F.
      • Miller Y.
      • Xiao Y.
      • McElheny D.
      • Thurber K.
      • Ma B.
      • Nussinov R.
      • Ishii Y.
      Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
      ). FTIR spectra of Aβ40-Iowa peptides containing [13C]LAGG labeling illustrate the difference in intensity of the 1606 cm−1 peak for anti-parallel and parallel fibrils (Fig. 8D).
      Figure thumbnail gr8
      Figure 8Copper binding to anti-parallel and parallel fibrils of Aβ40-Iowa. A, binding of Cu(II) by antiparallel Aβ40-Iowa fibrils. The copper-Aβ solution was prepared by adding Cu(II) in a 1:2 Cu:Aβ ratio to Aβ40-Iowa incubating at room temperature for 24 h, and then incubating the solution at 37 °C for an additional 96 h. Fibrils were pelleted by centrifugation. The amount of Cu(II) in the pellet and supernatant was determined using a photometric assay with TETD. The amount of Aβ40-Iowa peptide was determined by absorption at 280 nm. B, binding of Cu(II) by parallel Aβ40-Iowa fibrils, prepared as described under “Materials and methods” and analyzed as described above. C, control experiment showing copper distribution between the supernatant and pellet in the absence of amyloid peptide. Cu(II) ions partition equally between the supernatant (53.6%) and pellet (46.4%). Pairwise comparison was achieved via a unidirectional Wilcoxon rank sum test. Asterisk (*) indicates p < 0.05. n.s. indicates p > 0.05. D, FTIR spectra of anti-parallel (red) and parallel (black) Aβ40-Iowa fibrils having LAGG 13C labeling.
      The Cu(II) ion is paramagnetic and results in broadening of 13C NMR resonances when in close proximity. The Cu(II)-binding sites are consequently identified by loss of 13C intensity. Fig. 9A presents solid-state MAS NMR spectra of anti-parallel fibrils of Aβ40-Iowa in the region of the histidine backbone 13C=O and side chain 13Cε1, 13Cγ, and 13Cδ2 resonances with (red) and without (black) Cu(II) added. To highlight the differences between the Cu(II) binding at each histidine site, we normalized the intensity to the 13C=O resonance on the basis of the expectation that Cu(II) will coordinate with the unprotonated side chain imidazole nitrogen and not the backbone C=O. The largest intensity changes are observed for the 13Cε1 and 13Cδ2 resonances of His-6 and His-14.
      Figure thumbnail gr9
      Figure 9Copper coordination by histidine in anti-parallel Aβ40-Iowa fibrils. Solid-state 13C MAS NMR spectra are shown for Aβ40-Iowa uniform-13C-labeled at His-6, His-13, and His-14 with (red) and without (black) Cu(II). Anti-parallel fibrils (A) were formed by incubation at room temperature for 24 h. Copper was added after 24 h of room temperature incubation at a molar ratio of 1:2 copper:Aβ and the buffered (10 mm sodium phosphate buffer) solution was then incubated at 37 °C for an additional 96 h with shaking (50 rpm) before lyophilizing for NMR measurements. B, copper coordination by histidine in parallel Aβ40-Iowa fibrils. Parallel fibrils of Aβ40-Iowa were formed by incubation at 37 °C with 200 rpm shaking for 2 weeks (10 mm sodium phosphate buffer, 50 mm NaCl). Cu(II) was added after the 2-week period and then incubated for an additional 24 h. C, structure of the anti-parallel fibril of Aβ40-Iowa constructed by adding the N-terminal residues Asp-1–His-14 to the solid-state NMR coordinates (PDB code 2LNQ) from Gln-15–Val-40. The structure shows the relative locations of His-13 and His-14 on one monomer, and Glu-22 and Val-24 on the adjacent antiparallel monomer, after energy minimization and MD simulations. The β-strand between Lys-16 and Glu-22 continues through to at least His-13 and His-14. This geometry places His-14 next to Glu-22 on the adjacent monomer in a solvent-exposed position, whereas His-13 is oriented toward the fibril interior. The flexible N terminus allows His-6 to form a copper-binding site with His-14 and Glu-22.
      In contrast, measurements of Cu(II) binding to Aβ40-Iowa fibrils with β-strands in a parallel and in-register orientation exhibited a different pattern of Cu(II) binding to His-13 and His-14 (Fig. 9B). In these parallel, in-register fibrils there was a loss of intensity for the His-13 13Cγ resonance and an increase of intensity for the His-14 13Cε1 resonance compared with the fibrils with an anti-parallel geometry. The intensity changes of His-6 are comparable for the two Aβ40-Iowa structures. These observations indicate that Cu(II) coordinates with histidine in both the anti-parallel and parallel, in-register conformations of Aβ40-Iowa, but that the details of the coordination are different consistent with their different structures.
      Reduction of Cu(II) via oxidation of histidine, methionine, and tyrosine in the Aβ peptide has also been implicated in the generation of reactive oxygen species (
      • Atwood C.S.
      • Huang X.
      • Khatri A.
      • Scarpa R.C.
      • Kim Y.S.
      • Moir R.D.
      • Tanzi R.E.
      • Roher A.E.
      • Bush A.I.
      Copper catalyzed oxidation of Alzheimer Aβ.
      ,
      • Ali F.E.
      • Separovic F.
      • Barrow C.J.
      • Cherny R.A.
      • Fraser F.
      • Bush A.I.
      • Masters C.L.
      • Barnham K.J.
      Methionine regulates copper/hydrogen peroxide oxidation products of Aβ.
      ,
      • Nakamura M.
      • Shishido N.
      • Nunomura A.
      • Smith M.A.
      • Perry G.
      • Hayashi Y.
      • Nakayama K.
      • Hayashi T.
      Three histidine residues of amyloid-β peptide control the redox activity of copper and iron.
      ). To assess the oxidation of the Aβ peptide under our conditions, we probed chemical shift and intensity changes in the [5-13C]Met-35 resonance with or without Cu(II) addition. The absence of significant changes indicated that the Met35 –S-CH3 group had not been oxidized and that this group is not in close proximity to bound Cu(II) (data not shown). A similar observation was made in the studies of Ishii and co-workers for WT Aβ40 (
      • Parthasarathy S.
      • Long F.
      • Miller Y.
      • Xiao Y.
      • McElheny D.
      • Thurber K.
      • Ma B.
      • Nussinov R.
      • Ishii Y.
      Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
      ).
      To understand the differences in binding of Cu(II) to His-6, His-13, and His-14, we constructed a model of Aβ40-Iowa by adding an unstructured N terminus to the NMR structure of anti-parallel Aβ40-Iowa fibrils (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) and performing energy minimization and restrained MD simulations (Fig. 9C). The NMR structure is limited to residues 15-40, and consequently does not include the 3 histidine residues on the N terminus. A number of insights were gleaned from this computational model. First, when the Aβ sequence terminating at Gln-15 is extended and the structure is energy minimized, the β-strand structure from Gln-15–Asn-23 is largely maintained at His-13 and His-14. As a result, His-14 is generally oriented toward the solvent, whereas His-13 tends to be oriented toward the fibril interior. Orientation of the His-13–His-14 pair would be consistent with the observation of greater Cu(II) interactions with His-14 than His-13. A second observation from the computational model is that Val-24 is positioned between His-14 residues on an adjacent monomer and potentially blocks the formation of a His-14–Cu(II) –His-14–binding site. However, a small change in the position of Val-24 may generate a Cu(II)-binding site involving His-14 and Glu-22. The computational model shows that His-14 on one monomer typically hydrogen bonds with Glu-22 on the adjacent monomer.

      Discussion

      AD is associated with the formation of parenchymal amyloid plaques primarily composed of the Aβ42 peptide (
      • Iwatsubo T.
      • Odaka A.
      • Suzuki N.
      • Mizusawa H.
      • Nukina N.
      • Ihara Y.
      Visualization of Aβ42(43) and Aβ40 in senile plaques with end-specific Aβ monoclonals: Evidence that an initially deposited species is Aβ42(43).
      ). In contrast, CAA is a distinct disease, which often occurs in conjunction with AD, where the amyloid deposits form in and around the cerebral blood vessels and are primarily composed of Aβ40 (6). Here, we first measured the relative copper content in vascular amyloid formed on blood vessels in cases of human sporadic CAA and Iowa-type familial CAA. We then provide evidence using FTIR and NMR spectroscopy that Cu(II) is able to bind to anti-parallel fibrils in vitro and block the conversion of these fibrils to the more thermodynamically stable parallel, in-register structure. Finally, we found that Cu(II)-induced stabilization of anti-parallel Aβ40-Iowa fibrils results in an increased seeding efficiency of Aβ40 monomers. Together with previous studies on transgenic mouse and rat models showing that vascular amyloid contains fibrillar Aβ with anti-parallel structure (
      • Xu F.
      • Fu Z.
      • Dass S.
      • Kotarba A.E.
      • Davis J.
      • Smith S.O.
      • Van Nostrand W.E.
      Cerebral vascular amyloid seeds drive amyloid β-protein fibril assembly with a distinct anti-parallel structure.
      ,
      • Davis J.
      • Xu F.
      • Hatfield J.
      • Lee H.
      • Hoos M.D.
      • Popescu D.
      • Crooks E.
      • Kim R.
      • Smith S.O.
      • Robinson J.K.
      • Benveniste H.
      • Van Nostrand W.E.
      A novel transgenic rat model of robust cerebral microvascular amyloid with prominent vasculopathy.
      ), these observations suggest that Cu(II) binding may stabilize the anti-parallel Aβ fibrils in vascular amyloid and accelerate amyloid deposition.

      Cu(II) binding to vascular amyloid in human CAA

      There is considerable evidence for the dysregulation of copper and zinc ion concentrations in AD brain (
      • Sensi S.L.
      • Granzotto A.
      • Siotto M.
      • Squitti R.
      Copper and zinc dysregulation in Alzheimer's disease.
      ,
      • Savelieff M.G.
      • Nam G.
      • Kang J.
      • Lee H.J.
      • Lee M.
      • Lim M.H.
      Development of multifunctional molecules as potential therapeutic candidates for Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis in the last decade.
      ), and a general consensus for their accumulation in parenchymal plaques (
      • Miller L.M.
      • Wang Q.
      • Telivala T.P.
      • Smith R.J.
      • Lanzirotti A.
      • Miklossy J.
      Synchrotron-based infrared and X-ray imaging shows focalized accumulation of Cu and Zn co-localized with β-amyloid deposits in Alzheimer's disease.
      ,
      • Lovell M.A.
      • Robertson J.D.
      • Teesdale W.J.
      • Campbell J.L.
      • Markesbery W.R.
      Copper, iron and zinc in Alzheimer's disease senile plaques.
      ). Copper is a redox-active cofactor involved in a range of enzymatic reactions. In AD, Cu(I)/Cu(II) redox reactions involving the Aβ peptide have been implicated in the generation of reactive oxygen species leading to oxidative stress, which is associated with neuronal degeneration (
      • Cheignon C.
      • Tomas M.
      • Bonnefont-Rousselot D.
      • Faller P.
      • Hureau C.
      • Collin F.
      Oxidative stress and the amyloid β peptide in Alzheimer's disease.
      ). However, few studies have focused on the presence of copper and its potential role in CAA.
      Using X-ray fluorescence microscopy, we find that the relative copper content of vascular amyloid in sporadic CAA and in Iowa-type familial CAA is markedly higher than in the control vessels lacking amyloid, i.e. vascular amyloid in both sporadic and familial CAA accumulates copper. This result, obtained with XFM, agrees with previous studies of Schrag et al. (
      • Schrag M.
      • Crofton A.
      • Zabel M.
      • Jiffry A.
      • Kirsch D.
      • Dickson A.
      • Mao X.W.
      • Vinters H.V.
      • Domaille D.W.
      • Chang C.J.
      • Kirsch W.
      Effect of cerebral amyloid angiopathy on brain iron, copper, and zinc in Alzheimer's disease.
      ) that copper is strongly associated with vascular amyloid in CAA cases.
      In concert with the recognition that metal ion dysregulation may contribute to the progression of AD, efforts have been made to develop multifunctional small molecules that can bind both the Aβ peptide and selective metal ions. For example, Lim and co-workers (
      • Savelieff M.G.
      • Nam G.
      • Kang J.
      • Lee H.J.
      • Lee M.
      • Lim M.H.
      Development of multifunctional molecules as potential therapeutic candidates for Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis in the last decade.
      ) have developed small molecules that target and react with Cu(II)–Aβ over Zn(II)–Aβ, and have the potential to modulate metal-induced aggregation and neurotoxicity (
      • Choi J.S.
      • Braymer J.J.
      • Nanga R.P.
      • Ramamoorthy A.
      • Lim M.H.
      Design of small molecules that target metal-Aβ species and regulate metal-induced Aβ aggregation and neurotoxicity.
      ). Recently, we have shown the Cu(II) chelator TTM reduces both CAA and parenchymal plaque load in Tg2576 mice (
      • Zhu X.
      • Victor T.W.
      • Ambi A.
      • Sullivan J.K.
      • Hatfield J.
      • Xu F.
      • Miller L.M.
      • Van Nostrand W.E.
      Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
      ). This observation would be consistent with a potential role of Cu(II) in accelerating amyloid deposition.

      Anti-parallel β-sheet structure in fibrils associated with vascular amyloid

      Anti-parallel fibril structure is unusual for the Aβ peptides. The observation that Cu(II) stabilizes anti-parallel fibrils of Aβ40-Iowa provides a possible explanation for why the familial Iowa mutant of Aβ40 enhances the formation of vascular amyloid in CAA. The Aβ40-WT and Aβ42-WT peptides can rapidly form anti-parallel β-hairpins during the conversion of monomers to fibril nuclei in solution (
      • Fu Z.
      • Aucoin D.
      • Davis J.
      • Van Nostrand W.E.
      • Smith S.O.
      Mechanism of nucleated conformational conversion of Aβ42.
      ), but mature fibrils typically adopt parallel, in-register cross-β sheet. The structure of the Aβ40-Iowa fibrils (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ) reveals that complementary electrostatic interactions within the hydrophobic core of the fibril stabilize the anti-parallel conformation. These interactions are likely enhanced when the fibrils are embedded in vascular cell membranes. It is well known that even for the WT Aβ peptides, exposed hydrophobic residues lead to partitioning on and into cell membranes. For example in Aβ40-WT, the binding of the hydrophobic dye 8-anilino-1-naphthalenesulfonic acid (ANS) and 15N NMR exchange experiments indicate that shielding the N terminus and exposure of the hydrophobic sequence between residues 17 and 28 drive membrane interactions and are correlated with toxicity (
      • Ahmed R.
      • Akcan M.
      • Khondker A.
      • Rheinstädter M.C.
      • Bozelli Jr., J.C.
      • Epand R.M.
      • Huynh V.
      • Wylie R.G.
      • Boulton S.
      • Huang J.
      • Verschoor C.P.
      • Melacini G.
      Atomic resolution map of the soluble amyloid β assembly toxic surfaces.
      ). As a result, even though the anti-parallel Aβ40-Iowa conformation is metastable in solution, charge matching within the hydrophobic core of the fibrils would allow for its accumulation on vascular membranes. Further stabilization of nascent anti-parallel fibrils by Cu(II) may accelerate fibril growth as observed in Fig. 7A.
      The model for stabilization envisaged involves Cu(II) bridging Aβ monomers within the fibril. The net positive charge added to the fibril core by binding of Cu(II) would also enhance binding to negatively charged membrane bilayers. Moreover, it was found computationally that Zn(II) binding to preexisting polymorphic forms of Aβ peptides can shift the population equilibrium from parallel to anti-parallel β-sheet (
      • Miller Y.
      • Ma B.
      • Nussinov R.
      Zinc ions promote Alzheimer Aβ aggregation via population shift of polymorphic states.
      ). Together these studies suggest that metal ions may be important co-factors in determining the specific structures of Aβ fibrils in the human brain.
      Finally, the Aβ40-Iowa and Aβ40-Dutch peptides are much more toxic to human smooth muscle cells than Aβ40-WT (
      • Melchor J.P.
      • Van Nostrand W.E.
      Fibrillar amyloid β-protein mediates the pathologic accumulation of its secreted precursor in human cerebrovascular smooth muscle cells.
      ,
      • Melchor J.P.
      • McVoy L.
      • Van Nostrand W.E.
      Charge alterations of E22 enhance the pathogenic properties of the amyloid β-protein.
      ,
      • Van Nostrand W.E.
      • Melchor J.P.
      • Cho H.S.
      • Greenberg S.M.
      • Rebeck G.W.
      Pathogenic effects of D23N Iowa mutant amyloid β-protein.
      ). In cell culture, cell surface fibril assembly of Aβ40-Dutch is required for inducing downstream pathologic responses in human cerebral-vascular smooth muscle cells, including cell death (
      • Melchor J.P.
      • Van Nostrand W.E.
      Fibrillar amyloid β-protein mediates the pathologic accumulation of its secreted precursor in human cerebrovascular smooth muscle cells.
      ), whereas Aβ40-WT is relatively nontoxic. In contrast, Aβ40 oligomers are thought to mediate neuronal toxicity in AD (
      • Vivoli Vega M..
      • Cascella R.
      • Chen S.W.
      • Fusco G.
      • De Simone A.
      • Dobson C.M.
      • Cecchi C.
      • Chiti F.
      The toxicity of misfolded protein oligomers Is independent of their secondary structure.
      ). The differences in the influence of Aβ40-WT on human cerebral-vascular smooth muscle cells and neurons may be linked to the differences in membrane composition, including differences in gangliosides (
      • Yamamoto N.
      • Van Nostrand W.E.
      • Yanagisawa K.
      Further evidence of local ganglioside-dependent amyloid β-protein assembly in brain.
      ,
      • Yamamoto N.
      • Hirabayashi Y.
      • Amari M.
      • Yamaguchi H.
      • Romanov G.
      • Van Nostrand W.E.
      • Yanagisawa K.
      Assembly of hereditary amyloid β-protein variants in the presence of favorable gangliosides.
      ).

      Copper ion binding in CAA

      Most neurodegenerative disorders increase with age and often appear to be linked to oxidative stress. The Aβ peptides have metallo-binding sites for Zn(II), Cu(II), and Fe(III), and metals accumulate in brain with age. Three histidine residues (His-6, His-13, and His-14) located in the hydrophilic N-terminal part of the Aβ peptide have widely been shown to contribute to Cu(II) and Zn(II) binding (
      • Hou L.
      • Zagorski M.G.
      NMR reveals anomalous copper(II) binding to the amyloid A beta peptide of Alzheimer's disease.
      ,
      • Lim K.H.
      • Kim Y.K.
      • Chang Y.T.
      Investigations of the molecular mechanism of metal-induced Aβ (1-40) amyloidogenesis.
      ,
      • Shin B.K.
      • Saxena S.K.
      ESR spectroscopy suggests unequal contributions of the three histidine residues to Cu(II) binding in amyloid-β at physiological pH.
      ). The number of histidine and other coordination ligands depends on the pH and the Aβ sequence or structure (
      • Summers K.L.
      • Schilling K.M.
      • Roseman G.
      • Markham K.A.
      • Dolgova N.V.
      • Kroll T.
      • Sokaras D.
      • Millhauser G.L.
      • Pickering I.J.
      • George G.N.
      X-ray absorption spectroscopy investigations of copper(II) coordination in the human amyloid β peptide.
      ). Other coordination ligands are thought to include the backbone NH and C=O of Asp-1 or Ala-2, a glutamate residue or the C-terminal carboxyl group on Val-40 (
      • Drew S.C.
      • Barnham K.J.
      The heterogeneous nature of Cu2+ interactions with Alzheimer's amyloid-β peptide.
      ).
      Cu(II) can bind to Aβ monomers, oligomers, or fibrils (
      • Karr J.W.
      • Szalai V.A.
      Cu(II) binding to monomeric, oligomeric, and fibrillar forms of the Alzheimer's disease amyloid-β peptide.
      ,
      • Lim K.H.
      • Kim Y.K.
      • Chang Y.T.
      Investigations of the molecular mechanism of metal-induced Aβ (1-40) amyloidogenesis.
      ). In addition to Aβ40-WT and Aβ42-WT, copper binding has been measured for a variety of mutant and truncated Aβ peptides. For example, measurements on the E22Q Dutch variant that forms CAA reveal that Cu(II) can displace Zn(II) (
      • Clements A.
      • Allsop D.
      • Walsh D.M.
      • Williams C.H.
      Aggregation and metal-binding properties of mutant forms of the amyloid Aβ peptide of Alzheimer's disease.
      ). In our studies, we added Cu(II) ions after fibrils were formed to avoid issues with nonspecific aggregation. It is known that at substoichiometric concentrations of Cu(II), Aβ40 forms fibrils (
      • Jun S.M.
      • Saxena S.
      The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
      ,
      • Chen W.T.
      • Liao Y.H.
      • Yu H.M.
      • Cheng I.H.
      • Chen Y.R.
      Distinct effects of Zn2+, Cu2+, Fe3+, and Al3+ on amyloid-β stability, oligomerization, and aggregation: amyloid-β destabilization promotes annular protofibril formation.
      ,
      • Jun S.
      • Gillespie J.R.
      • Shin B.K.
      • Saxena S.
      The second Cu(II)-binding site in a proton-rich environment interferes with the aggregation of amylold-β(1-40) into amyloid fibrils.
      ), and that under these concentrations the Cu(II) is coordinated by histidine (
      • Jun S.
      • Gillespie J.R.
      • Shin B.K.
      • Saxena S.
      The second Cu(II)-binding site in a proton-rich environment interferes with the aggregation of amylold-β(1-40) into amyloid fibrils.
      ). At higher Cu(II) concentrations, Aβ40 forms amorphous aggregates instead of amyloid fibrils (
      • Jun S.M.
      • Saxena S.
      The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
      ,
      • Smith D.P.
      • Ciccotosto G.D.
      • Tew D.J.
      • Fodero-Tavoletti M.T.
      • Johanssen T.
      • Masters C.L.
      • Barnham K.J.
      • Cappai R.
      Concentration dependent Cu2+ induced aggregation and dityrosine formation of the Alzheimer's disease amyloid-beta peptide.
      ). In this regard, we observe a broadening in the FTIR spectra when Cu(II) is initially added that we attribute to nonspecific aggregation (Fig. 5, C and D). The changes observed by FTIR following Cu(II) addition, however, are consistent with the formation of more homogeneous anti-parallel fibrils.
      Copper co-localization with cerebral blood vessels and Aβ imply an interaction that may influence the Aβ fibrillization. As such, we found that the presence of Cu(II) in vitro increased the seeding efficiency of Aβ40-Iowa fibrils. In vivo, it is possible that Aβ40-Iowa fibrils are stabilized by Cu(II) that then can act as a structural scaffold for both WT and mutant Aβ40 monomers.
      We specifically investigated the role of His-6, His-13, and His-14 in coordinating Cu(II) in anti-parallel Aβ40-Iowa fibrils. His-13 and His-14 are located just outside of the hydrophobic core of these fibrils. However, as mentioned above, the fibril core is charge-matched with complementary electrostatic interactions on neighboring peptides within the fibril between Lys-16 and Glu-22 and between Lys-28 and the C terminus (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). The observation that His-14 exhibits greater loss of NMR signal intensity than His-13 was surprising based on the previous studies of Aβ40-WT (
      • Parthasarathy S.
      • Long F.
      • Miller Y.
      • Xiao Y.
      • McElheny D.
      • Thurber K.
      • Ma B.
      • Nussinov R.
      • Ishii Y.
      Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
      ) and on the structure of the Aβ40-Iowa fibrils where Val-24 lies between the His-14 residues (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). However, in anti-parallel fibrils (Fig. 9C), His-14 was found to be oriented away from the fibril core and hydrogen-bonded to Glu-22 on an adjacent monomer within the fibril. One possibility is that Cu(II) bridges these two residues. In this regard, glutamate was found to be a Cu(II) ligand in solid-state NMR studies on Aβ40-WT fibrils (
      • Ishii Y.
      A solid-state NMR study reveals structure and dynamics in copper(II)-binding to Alzheimer's β-amyloid fibrils.
      ), although the glutamate involved (Glu-3, Glu-11, or Glu-22) was not assigned.
      Finally, the functional role of copper binding in AD is associated with the Fenton-type reaction in which copper reduction reactions are linked to the generation of reactive oxygen species (
      • Huang X.
      • Atwood C.S.
      • Hartshorn M.A.
      • Multhaup G.
      • Goldstein L.E.
      • Scarpa R.C.
      • Cuajungco M.P.
      • Gray D.N.
      • Lim J.
      • Moir R.D.
      • Tanzi R.E.
      • Bush A.I.
      The Aβ peptide of Alzheimer's disease directly produces hydrogen peroxide through metal ion reduction.
      ). In the case of CAA, we propose an alternative role for copper, namely that Cu(II) stabilizes the anti-parallel structure as fibrils form on the surface of cerebral vascular cells. Membrane-embedded fibrils then have the potential to disrupt cerebral blood vessels leading to hemorrhaging and stroke.

      Materials and methods

      Peptide synthesis

      Aβ peptides were synthesized using N-t-Boc chemistry (ERI-amyloid, Waterbury, CT) and purified by HPLC. The mass of the purified peptide was measured using matrix-assisted laser desorption or electrospray ionization MS, and was consistent with the calculated mass for the peptide. On the basis of analytical reverse phase HPLC and MS, the purity of the peptides was generally >98%.
      The 13C-labeled N-t-Boc protected amino acids were purchased from Cambridge Isotope Laboratory (Tewksbury, MA) except for [U-13C]histidine. Protected histidine was prepared from l-histidine•HCl•H2O (13C6,15N3). Briefly, His (0.502 g, 2.30 mmol) was reacted with methanol in the presence of acetyl chloride to give His-OMe•2HCl (0.540 g). His-OMe•2HCl (0.540 g) was reacted with di-tert-butyl dicarbonate in the presence of N,N-diisopropylethylamine in methanol to give Boc-His(Boc)-OMe (0.727 g). Boc-His(Boc)-OMe (0.720 g) was reacted with benzyl chloromethyl ether in dichloromethane to give Boc-His(Bom)-OMe•HCl (0.401 g). Boc-His(Bom)-OMe•HCl was hydrolyzed with sodium hydroxide in methanol to produce Boc-His(Bom)-OH (0.374 g, 0.973 mmol).

      Sample preparation

      Monomeric Aβ peptide was prepared by first dissolving purified peptides in hexafluoro-2-propanol and freeze-drying under a 25 mTorr vacuum overnight. Lyophilized Aβ peptides were dissolved in 50 mm NaOH for 1 h, diluted in 10 mm phosphate buffer at 4 °C to a concentration of 100 μm, and then filtered with 0.2-μm filters before use. To form anti-parallel fibrils, the solution was incubated for 24 h at room temperature, typically 21-23 °C, under quiescent conditions. To drive the transition to fibrils with parallel, in-register orientation, the solution was heated to 37 °C and placed under moderately strong shaking conditions (170 rpm).
      For FTIR, NMR, and TEM studies with Cu(II), the copper was added after 24 h of room temperature incubation at a molar ratio of 1:2 copper:Aβ in the form of copper glycinate (0.5 mm CuCl2, 3 mm glycine) (
      • Smith D.P.
      • Smith D.G.
      • Curtain C.C.
      • Boas J.F.
      • Pilbrow J.R.
      • Ciccotosto G.D.
      • Lau T.L.
      • Tew D.J.
      • Perez K.
      • Wade J.D.
      • Bush A.I.
      • Drew S.C.
      • Separovic F.
      • Masters C.L.
      • Cappai R.
      • Barnham K.J.
      Copper-mediated amyloid-β toxicity is associated with an intermolecular histidine bridge.
      ). Glycine was present to prevent coordination of Cu(II) by phosphate. The solution was then placed in 37 °C for an additional 96 h with shaking (170 rpm) before lyophilizing for NMR measurements.
      Parallel fibrils of Aβ40-Iowa for the copper binding and AFM studies were formed by first allowing Aβ40-Iowa monomer (100 μm) to fibrillize under shaking conditions (100 rpm) at 37 °C in 10 mm NaCl, 10 mm sodium phosphate buffer (pH 7.3) for 2 weeks. The resulting fibrils were then subjected to mild sonication to produce fibril fragments that were used to seed a second generation of fibrils with the addition of newly prepared Aβ40-Iowa monomer. The re-seeding process was repeated 5 times to yield mature parallel, in-register Aβ40-Iowa fibrils.

      Copper-binding assay

      The copper-binding assay was adapted from previous studies (
      • Parthasarathy S.
      • Long F.
      • Miller Y.
      • Xiao Y.
      • McElheny D.
      • Thurber K.
      • Ma B.
      • Nussinov R.
      • Ishii Y.
      Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
      ). The target solution was centrifuged at 18,000 × g for 1 h. The top 35% of the supernatant (upper) was carefully separated from the bottom 65% (lower) of the solution. 2 m HCl (0.02%, v/v) was added and the solution was allowed to sit for 15 min to reach a pH of 1.5-2. The absorbance at 270 nm was monitored on an Agilent Model V spectrophotometer to quantify Aβ content. N,N,N‘,N‘-Tetraethylthiuram disulfide (TETD) was used as a photometric indicator of unbound Cu(II). We prepared the photometric reagent by dissolving 0.04% TETD in a 4:1 acetone:water ratio. The TETD solution was added to the amyloid sample (0.08%, v/v) and allowed to sit 1.5 h, which we found to be the optimal incubation time for maximal absorbance at 420 nm. The copper-binding experiments were repeated three times (n = 3). Values were normalized within each condition using the sum of the absorbance in the upper and lower solutions as 100%. For clarity, amyloid-specific absorbance values for the amyloid-free control sample were not normalized as they were essentially null. Error bars show the standard deviation.

      X-ray fluorescence microscopy

      Brain tissue samples of autopsy cases of control (n = 3) and neuropathologically confirmed AD (n = 5), sporadic CAA (n = 3), and Iowa D694N APP mutation familial CAA (n = 1) were used for this study. For each case, 3–10 vessels were examined. These studies were approved by the Stony Brook University Institutional Review Board and abide by the Declaration of Helsinki principles.
      Samples from the hippocampal and frontal regions were snap-frozen in liquid nitrogen and stored at −80 °C. The tissues were embedded in OCT and cryo-sectioned at −12 °C to a thickness of 30 μm. The tissue sections were mounted on 4-µm thick Ultralene film (SPEX Certiprep, Metuchen, NJ), an X-ray transparent and trace-metal-free substrate. The Ultralene film was previously stretched and glued to a Delrin ring using 3M white/gray epoxy (3M Scotch-Weld, St. Paul, MN) to provide structural support for the Ultralene film. The localization of amyloid was determined by staining the tissue sections with thioflavin S.
      Prior to XFM data collection, the protein density in the vessels was determined using FTIR microspectroscopy (FTIRM). Because the vessels can be denser than the surrounding tissue, the protein density was used to normalize the XFM data when quantifying the metal content, thus avoiding an overestimate of the metal content within the vessels. FTIRM spectra were acquired using a Spectrum Spotlight FTIR microscope with 8 cm−1 spectral resolution over the mid-IR spectral region (800 to 4000 cm−1). A 20 μm aperture was used with 64 scans per point. The relative protein content at each pixel was determined by integrating the amide II protein peak from 1490 to 1580 cm−1 with a linear baseline from 1480 to 1800 cm−1. The area under this peak is directly proportional to the amount of protein in the specimen. The relative protein density was calculated as the amide II area on/off the vessels.
      The copper concentration within the vessels was determined using synchrotron XFM at beamline 13-ID-E at the Advanced Photon Source, Argonne National Laboratory, and beamline 5-ID at the National Synchrotron Light Source II, Brookhaven National Laboratory. The energy of the incident X-ray beam was 10 keV. The X-ray beam was focused to a 3 × 3 μm spot using Kirkpatrick-Baez focusing mirrors. The specimens were placed at a 45° angle with respect to the incident X-ray beam, and the X-ray fluorescence was detected by a Si-drift detector oriented at a 90° angle from the incident beam. Energy dispersive spectra were collected at every pixel while raster scanning across the specimen at 0.5 s per point with a 2-μm step size. The copper concentration in the vessels was normalized to the relative protein content to account for any increased density within the vessels. Statistical significance between groups was determined using a two-tailed t test.

      Solid-state NMR spectroscopy

      Solid-state NMR 13C experiments were performed at frequency of 125 MHz on a Bruker Avance spectrometer. The MAS frequency was set to 9-11 KHz (±5 Hz). Ramped amplitude cross-polarization was used with a contact time of 2 ms. Two-pulse phase-modulated decoupling was used during the evolution and acquisition periods with a radiofrequency field strength of 80 kHz. Internuclear 13C distance constraints were obtained from 2D dipolar-assisted rotational resonance (DARR) NMR experiments using a mixing time of 600 ms. For these experiments, the recycle delay was set to 2 s with 64 increments in the F1 dimension. The total acquisition time for each 2D 13C spectrum was typically 4 days. In contrast, the total acquisition time for each 1D 13C NMR spectrum for the [U-13C]histidine-labeled Aβ samples was typically 1 day.
      The 13C MAS NMR spectra were externally referenced to the 13C resonance of neat TMS at 0 ppm at room temperature. Using TMS as the external reference, we calibrated the carbonyl resonance of solid glycine at 176.46 ppm. The chemical shift difference between 13C of DSS in D2O relative to neat TMS is 2.01 ppm.

      Thioflavin T fluorescence spectroscopy

      Fluorescent measurements were taken using a Spectra Max iD3 spectrometer (Molecular Devices, San Jose, CA). Peptide solutions corresponding to 100 μm total Aβ were used for kinetic fluorescent studies. A final concentration of 37.5 μm thioflavin-T was used with an excitation wavelength of 440 nm and an emission wavelength of 490 nm. 200 μl of the Aβ and thioflavin-T solution was added to each corresponding well in a 96-well clear (Greiner) microplate. Measurements were taken from the bottom of the plate every 10-min for 24-48 h with 4 s slow orbital shaking between reads.

      Atomic force microscopy

      AFM images were obtained using a MultiMode microscope (Digital Instruments, Santa Barbara, CA) with a custom-built controller (LifeAFM, Port Jefferson, NY) that allows one low force contact of the AFM tip to the sample surface per pixel. Super-sharp silicon probes with a tip width of 3-5 nm (at a height of 2 nm) were modified for magnetic retraction by attachment of samarium cobalt particles. Samples for AFM were diluted to a concentration of 0.5 μm deposited onto freshly cleaved ruby mica (S & J Trading, Glen Oaks, NY) and imaged under hydrated conditions.

      FTIR spectroscopy

      FTIR measurements were made with a Bruker Vertex 70v spectrometer with a room temperature detector and attenuated total reflectance (ATR) accessory. Samples were layered on a 2-mm germanium ATR plate (Pike Technologies) by drying 50-100 μl of peptide solution on the Ge surface with air. The spectral resolution was 4 cm−1. Spectra were normalized to the amide II band.

      Transmission EM

      Samples were diluted, deposited onto carbon-coated copper mesh grids, rinsed with water, and negatively stained with 2% (w/v) uranyl acetate. The excess stain was wicked away, and the sample grids were allowed to air dry. The samples were viewed with a FEI Tecnai 12 BioTwin 80 kV transmission electron microscope, and digital images were taken with an Advanced Microscopy Techniques camera.

      MD simulations

      The initial model was generated using model 1 of PDB ID 2LNQ (
      • Qiang W.
      • Yau W.-M.
      • Luo Y.
      • Mattson M.P.
      • Tycko R.
      Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
      ). The 14 N-terminal residues with missing density on each chain were modeled in an extended conformation using the addaa tool in UCSF Chimera (
      • Pettersen E.F.
      • Goddard T.D.
      • Huang C.C.
      • Couch G.S.
      • Greenblatt D.M.
      • Meng E.C.
      • Ferrin T.E.
      UCSF chimera: a visualization system for exploratory research and analysis.
      ). Manual rotation around backbone dihedrals was performed in Chimera to relieve steric clashes and direct the added residues outside the β-sheet region. The last two peptide chains were discarded, leaving a model with 6 peptide hairpins. Simulation setup was performed using the Leap module of Amber version 18 (
      • Case D.A.
      • Cheatham T.E.
      • Darden T.
      • Gohlke H.
      • Luo R.
      • Merz K.M.
      • Onufriev A.
      • Simmerling C.
      • Wang B.
      • Woods R.J.
      The Amber biomolecular simulation programs.
      ), using the ff14SB force field (
      • Maier J.A.
      • Martinez C.
      • Kasavajhala K.
      • Wickstrom L.
      • Hauser K.E.
      • Simmerling C.
      ff14SB: Improving the accuracy of protein side chain and backbone parameters from ff99SB.
      ) and GBneck implicit water model. Minimization and equilibration were performed with restraints on β-strand residues 16-40 and 56-80 in each chain. Minimization of the initial model was performed for 10,000 steps with restraints applied to heavy atoms in the β-strand residues listed above, with a force constant of 100 kcal/mol. The same restraints were maintained during subsequent heating to 298 K over 1 ns, and an additional 1 ns at 298 K. The restraint force constant was reduced to 10 kcal/mol for an additional 1 ns MD. The restraints on side chains were then removed, followed by 10,000 steps minimization, 1 ns MD at 298 K, and 1 ns MD at 298 K with the restraint force constant reduced to 1.0 kcal/mol, then reduced to 0.1 kcal/mol for an additional 1 ns at 298 K, and 1 ns at 298 K with no restraints. Production MD was carried out for 100 ns at 298 K.

      Data availability

      All data described in the manuscript are contained within the manuscript. Additional data are available upon request, Steven O. Smith, Department of Biochemistry and Cell Biology, Stony Brook University, Stony Brook, NY 11794. Tel: 631-632-1210; Fax: 631-632-8575; E-mail: [email protected]

      Acknowledgments

      We thank Dr. Matthew Frosch, Director, MA Alzheimer's Disease Research Center supported by Grant P50 AG005134 for providing human brain specimens of Iowa-type CAA and the Neuropathology Core at University of California, Irvine, CA, for providing human brain specimens of AD and sporadic CAA. Portions of this work were performed at GeoSoilEnviroCARS (The University of Chicago, Sector 13), Advanced Photon Source (APS), and Argonne National Laboratory. This research used resources of the Advanced Photon Source, a United States Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under Contract number DE-AC02-06CH11357. This research also used beamline 5-ID (SRX) of the National Synchrotron Light Source II, a United States Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Brookhaven National Laboratory under contract DE-SC0012704.

      References

        • Selkoe D.J.
        • Hardy J.
        The amyloid hypothesis of Alzheimer's disease at 25 years.
        EMBO Mol. Med. 2016; 8 (27025652): 595-608
        • Thal D.R.
        • Griffin W.S.
        • de Vos R.A.
        • Ghebremedhin E.
        Cerebral amyloid angiopathy and its relationship to Alzheimer's disease.
        Acta Neuropathol. 2008; 115 (18369648): 599-609
        • Qi-Takahara Y.
        • Morishima-Kawashima M.
        • Tanimura Y.
        • Dolios G.
        • Hirotani N.
        • Horikoshi Y.
        • Kametani F.
        • Maeda M.
        • Saido T.C.
        • Wang R.
        • Ihara Y.
        Longer forms of amyloid β protein: implications for the mechanism of intramembrane cleavage by γ-secretase.
        J. Neurosci. 2005; 25 (15647487): 436-445
        • Selkoe D.J.
        Alzheimer disease: mechanistic understanding predicts novel therapies.
        Ann. Internal Med. 2004; 140: 627-638
        • Castaño E.M.
        • Prelli F.
        • Soto C.
        • Beavis R.
        • Matsubara E.
        • Shoji M.
        • Frangione B.
        The length of amyloid-β in hereditary cerebral hemorrhage with amyloidosis, Dutch type: implications for the role of amyloid-β 1-42 in Alzheimer's disease.
        J. Biol. Chem. 1996; 271 (8943274): 32185-32191
        • Roher A.E.
        • Lowenson J.D.
        • Clarke S.
        • Woods A.S.
        • Cotter R.J.
        • Gowing E.
        • Ball M.J.
        β-Amyloid-(1-42) is a major component of cerebrovascular amyloid deposits: implications for the pathology of Alzheimer-disease.
        Proc. Natl. Acad. Sci. U.S.A. 1993; 90 (8248178): 10836-10840
        • Rutgers K.S.
        • van Remoortere A.
        • van Buchem M.A.
        • Verrips C.T.
        • Greenberg S.M.
        • Bacskai B.J.
        • Frosch M.P.
        • van Duinen S.G.
        • Maat-Schieman M.L.
        • van der Maarel S.M.
        Differential recognition of vascular and parenchymal β amyloid deposition.
        Neuro. Aging. 2011; 32: 1774-1783
        • Han B.H.
        • Zhou M.L.
        • Vellimana A.K.
        • Milner E.
        • Kim D.H.
        • Greenberg J.K.
        • Chu W.H.
        • Mach R.H.
        • Zipfel G.J.
        Resorufin analogs preferentially bind cerebrovascular amyloid: potential use as imaging ligands for cerebral amyloid angiopathy.
        Mol. Neurodegen. 2011; 6: 86
        • Grabowski T.J.
        • Cho H.S.
        • Vonsattel J.P.
        • Rebeck G.W.
        • Greenberg S.M.
        Novel amyloid precursor protein mutation in an Iowa family with dementia and severe cerebral amyloid angiopathy.
        Ann. Neurol. 2001; 49 (11409420): 697-705
        • Van Broeckhoven C.
        • Haan J.
        • Bakker E.
        • Hardy J.A.
        • Van Hul W.
        • Wehnert A.
        • Vegter-Van der Vlis M.
        • Roos R.A.
        Amyloid β protein precursor gene and hereditary cerebral hemorrhage with amyloidosis (Dutch).
        Science. 1990; 248: 1120-1122
        • Condello C.
        • Lemmin T.
        • Stöhr J.
        • Nick M.
        • Wu Y.
        • Maxwell A.M.
        • Watts J.C.
        • Caro C.D.
        • Oehler A.
        • Keene C.D.
        • Bird T.D.
        • van Duinen S.G.
        • Lannfelt L.
        • Ingelsson M.
        • Graff C.
        • et al.
        Structural heterogeneity and intersubject variability of Aβ in familial and sporadic Alzheimer's disease.
        Proc. Natl. Acad. Sci. U.S.A. 2018; 115 (29311311): E782-E791
        • Tycko R.
        Solid-state NMR studies of amyloid fibril structure.
        Annu. Rev. Phys. Chem. 2011; 62: 279-299
        • Tycko R.
        Amyloid polymorphism: structural basis and neurobiological relevance.
        Neuron. 2015; 86 (25950632): 632-645
        • Cloe A.L.
        • Orgel J.P.
        • Sachleben J.R.
        • Tycko R.
        • Meredith S.C.
        The Japanese mutant Aβ (ΔE22-Aβ1-39) forms fibrils instantaneously, with low-thioflavin T fluorescence: Seeding of wild-type Aβ1-40 into atypical fibrils by ΔE22-Aβ1-39.
        Biochemistry. 2011; 50 (21291268): 2026-2039
        • Qiang W.
        • Yau W.-M.
        • Luo Y.
        • Mattson M.P.
        • Tycko R.
        Antiparallel β-sheet architecture in Iowa-mutant β-amyloid fibrils.
        Proc. Natl. Acad. Sci. U.S.A. 2012; 109 (22403062): 4443-4448
        • Xu F.
        • Fu Z.
        • Dass S.
        • Kotarba A.E.
        • Davis J.
        • Smith S.O.
        • Van Nostrand W.E.
        Cerebral vascular amyloid seeds drive amyloid β-protein fibril assembly with a distinct anti-parallel structure.
        Nat. Commun. 2016; 7 (27869115): 13527
        • Davis J.
        • Xu F.
        • Hatfield J.
        • Lee H.
        • Hoos M.D.
        • Popescu D.
        • Crooks E.
        • Kim R.
        • Smith S.O.
        • Robinson J.K.
        • Benveniste H.
        • Van Nostrand W.E.
        A novel transgenic rat model of robust cerebral microvascular amyloid with prominent vasculopathy.
        Am. J. Pathol. 2018; 188 (30446159): 2877-2889
        • Tomidokoro Y.
        • Rostagno A.
        • Neubert T.A.
        • Lu Y.
        • Rebeck G.W.
        • Frangione B.
        • Greenberg S.M.
        • Ghiso J.
        Iowa variant of familial Alzheimer's disease accumulation of posttranslationally modified Aβ D23N in parenchymal and cerebrovascular amyloid deposits.
        Am. J. Pathol. 2010; 176 (20228223): 1841-1854
        • Curtain C.C.
        • Ali F.E.
        • Smith D.G.
        • Bush A.I.
        • Masters C.L.
        • Barnham K.J.
        Metal ions, pH, and cholesterol regulate the interactions of Alzheimer's disease Aβ peptide with membrane lipid.
        J. Biol. Chem. 2003; 278 (12435742): 2977-2982
        • Faller P.
        • Hureau C.
        • La Penna G.
        Metal ions and intrinsically disordered proteins and peptides: from Cu/Zn amyloid-β to general principles.
        Acct. Chem. Res. 2014; 47: 2252-2259
        • Faller P.
        Copper and zinc binding to amyloid-β: coordination, dynamics, aggregation, reactivity and metal-ion transfer.
        ChemBioChem. 2009; 10 (19877000): 2837-2845
        • Karr J.W.
        • Szalai V.A.
        Cu(II) binding to monomeric, oligomeric, and fibrillar forms of the Alzheimer's disease amyloid-β peptide.
        Biochemistry. 2008; 47 (18393444): 5006-5016
        • Atwood C.S.
        • Huang X.
        • Khatri A.
        • Scarpa R.C.
        • Kim Y.S.
        • Moir R.D.
        • Tanzi R.E.
        • Roher A.E.
        • Bush A.I.
        Copper catalyzed oxidation of Alzheimer Aβ.
        Cell. Mol. Biol. 2000; 46 (10875439): 777-783
        • Schrag M.
        • Crofton A.
        • Zabel M.
        • Jiffry A.
        • Kirsch D.
        • Dickson A.
        • Mao X.W.
        • Vinters H.V.
        • Domaille D.W.
        • Chang C.J.
        • Kirsch W.
        Effect of cerebral amyloid angiopathy on brain iron, copper, and zinc in Alzheimer's disease.
        J. Alzheimers Dis. 2011; 24 (21187585): 137-149
        • Leskovjan A.C.
        • Lanzirotti A.
        • Miller L.M.
        Amyloid plaques in PSAPP mice bind less metal than plaques in human Alzheimer's disease.
        NeuroImage. 2009; 47 (19481608): 1215-1220
        • Miller L.M.
        • Wang Q.
        • Telivala T.P.
        • Smith R.J.
        • Lanzirotti A.
        • Miklossy J.
        Synchrotron-based infrared and X-ray imaging shows focalized accumulation of Cu and Zn co-localized with β-amyloid deposits in Alzheimer's disease.
        J. Struct. Biol. 2006; 155 (16325427): 30-37
        • Zhu X.
        • Victor T.W.
        • Ambi A.
        • Sullivan J.K.
        • Hatfield J.
        • Xu F.
        • Miller L.M.
        • Van Nostrand W.E.
        Copper accumulation and the effect of chelation treatment on cerebral amyloid angiopathy compared to parenchymal amyloid plaques.
        Metallomics. 2020; 12: 539-546
        • Tang T.C.
        • Hu Y.
        • Kienlen-Campard P.
        • El Haylani L.
        • Decock M.
        • Van Hees J.
        • Fu Z.
        • Octave J.N.
        • Constantinescu S.N.
        • Smith S.O.
        Conformational changes induced by the A21G Flemish mutation in the amyloid precursor protein lead to increased Aβ production.
        Structure. 2014; 22 (24462250): 387-396
        • Barritt J.D.
        • Viles J.H.
        Truncated amyloid-β(11-40/42) from Alzheimer disease binds Cu2+ with a femtomolar affinity and influences fiber assembly.
        J. Biol. Chem. 2015; 290 (26408196): 27791-27802
        • Hatami A.
        • Monjazeb S.
        • Milton S.
        • Glabe C.G.
        Familial Alzheimer's disease mutations within the amyloid precursor protein alter the aggregation and conformation of the amyloid- β peptide.
        J. Biol. Chem. 2017; 292 (28049728): 3172-3185
        • Mastrangelo I.A.
        • Ahmed M.
        • Sato T.
        • Liu W.
        • Wang C.
        • Hough P.
        • Smith S.O.
        High-resolution atomic force microscopy of soluble Aβ42 oligomers.
        J. Mol. Biol. 2006; 358 (16499926): 106-119
        • Fu Z.
        • Aucoin D.
        • Davis J.
        • Van Nostrand W.E.
        • Smith S.O.
        Mechanism of nucleated conformational conversion of Aβ42.
        Biochemistry. 2015; 54 (26069943): 4197-4207
        • Jun S.M.
        • Saxena S.
        The aggregated state of amyloid-β peptide in vitro depends on Cu2+ ion concentration.
        Angew. Chem.-Int. Edit. 2007; 46: 3959-3961
        • Pedersen J.T.
        • Østergaard J.
        • Rozlosnik N.
        • Gammelgaard B.
        • Heegaard N.H.
        Cu(II) mediates kinetically distinct, non-amyloidogenic aggregation of amyloid-ß peptides.
        J. Biol. Chem. 2011; 286 (21642429): 26952-26963
        • Chen W.T.
        • Liao Y.H.
        • Yu H.M.
        • Cheng I.H.
        • Chen Y.R.
        Distinct effects of Zn2+, Cu2+, Fe3+, and Al3+ on amyloid-β stability, oligomerization, and aggregation: amyloid-β destabilization promotes annular protofibril formation.
        J. Biol. Chem. 2011; 286 (21216965): 9646-9656
        • Paul C.
        • Wang J.
        • Wimley W.C.
        • Hochstrasser R.M.
        • Axelsen P.H.
        Vibrational coupling, isotopic editing, and β-sheet structure in a membrane-bound polypeptide.
        J. Am. Chem. Soc. 2004; 126 (15125676): 5843-5850
        • Petty S.A.
        • Decatur S.M.
        Experimental evidence for the reorganization of β-strands within aggregates of the Aβ(16-22) peptide.
        J. Am. Chem. Soc. 2005; 127 (16190699): 13488-13489
        • Smith D.P.
        • Ciccotosto G.D.
        • Tew D.J.
        • Fodero-Tavoletti M.T.
        • Johanssen T.
        • Masters C.L.
        • Barnham K.J.
        • Cappai R.
        Concentration dependent Cu2+ induced aggregation and dityrosine formation of the Alzheimer's disease amyloid-beta peptide.
        Biochemistry. 2007; 46 (17297919): 2881-2891
        • Sgourakis N.G.
        • Yau W.M.
        • Qiang W.
        Modeling an in-register, parallel “Iowa“ Aβ fibril structure using solid-state NMR data from labeled samples with rosetta.
        Structure. 2015; 23 (25543257): 216-227
        • Shin B.K.
        • Saxena S.
        Direct evidence that all three histidine residues coordinate to Cu(II) in amyloid-β(1-16).
        Biochemistry. 2008; 47 (18690709): 9117-9123
        • Hou L.
        • Zagorski M.G.
        NMR reveals anomalous copper(II) binding to the amyloid A beta peptide of Alzheimer's disease.
        J. Am. Chem. Soc. 2006; 128 (16848423): 9260-9261
        • Parthasarathy S.
        • Long F.
        • Miller Y.
        • Xiao Y.
        • McElheny D.
        • Thurber K.
        • Ma B.
        • Nussinov R.
        • Ishii Y.
        Molecular-level examination of Cu2+ binding structure for amyloid fibrils of 40-residue Alzheimer's β by solid-state NMR spectroscopy.
        J. Am. Chem. Soc. 2011; 133 (21341665): 3390-3400
        • Sarell C.J.
        • Syme C.D.
        • Rigby S.E.
        • Viles J.H.
        Copper(II) binding to amyloid-β fibrils of Alzheimer's disease reveals a picomolar affinity: stoichiometry and coordination geometry are independent of Aβ oligomeric form.
        Biochemistry. 2009; 48 (19338344): 4388-4402
        • Ali F.E.
        • Separovic F.
        • Barrow C.J.
        • Cherny R.A.
        • Fraser F.
        • Bush A.I.
        • Masters C.L.
        • Barnham K.J.
        Methionine regulates copper/hydrogen peroxide oxidation products of Aβ.
        J. Pept. Sci. 2005; 11 (15635660): 353-360
        • Nakamura M.
        • Shishido N.
        • Nunomura A.
        • Smith M.A.
        • Perry G.
        • Hayashi Y.
        • Nakayama K.
        • Hayashi T.
        Three histidine residues of amyloid-β peptide control the redox activity of copper and iron.
        Biochemistry. 2007; 46 (17929832): 12737-12743
        • Iwatsubo T.
        • Odaka A.
        • Suzuki N.
        • Mizusawa H.
        • Nukina N.
        • Ihara Y.
        Visualization of Aβ42(43) and Aβ40 in senile plaques with end-specific Aβ monoclonals: Evidence that an initially deposited species is Aβ42(43).
        Neuron. 1994; 13 (8043280): 45-53
        • Sensi S.L.
        • Granzotto A.
        • Siotto M.
        • Squitti R.
        Copper and zinc dysregulation in Alzheimer's disease.
        Trends Pharmacol. Sci. 2018; 39 (30352697): 1049-1063
        • Savelieff M.G.
        • Nam G.
        • Kang J.
        • Lee H.J.
        • Lee M.
        • Lim M.H.
        Development of multifunctional molecules as potential therapeutic candidates for Alzheimer's disease, Parkinson's disease, and amyotrophic lateral sclerosis in the last decade.
        Chem. Rev. 2019; 119 (30095897): 1221-1322
        • Lovell M.A.
        • Robertson J.D.
        • Teesdale W.J.
        • Campbell J.L.
        • Markesbery W.R.
        Copper, iron and zinc in Alzheimer's disease senile plaques.
        J. Neurol. Sci. 1998; 158 (9667777): 47-52
        • Cheignon C.
        • Tomas M.
        • Bonnefont-Rousselot D.
        • Faller P.
        • Hureau C.
        • Collin F.
        Oxidative stress and the amyloid β peptide in Alzheimer's disease.
        Redox Biol. 2018; 14 (29080524): 450-464
        • Choi J.S.
        • Braymer J.J.
        • Nanga R.P.
        • Ramamoorthy A.
        • Lim M.H.
        Design of small molecules that target metal-Aβ species and regulate metal-induced Aβ aggregation and neurotoxicity.
        Proc. Natl. Acad. Sci. U.S.A. 2010; 107 (21131570): 21990-21995
        • Ahmed R.
        • Akcan M.
        • Khondker A.
        • Rheinstädter M.C.
        • Bozelli Jr., J.C.
        • Epand R.M.
        • Huynh V.
        • Wylie R.G.
        • Boulton S.
        • Huang J.
        • Verschoor C.P.
        • Melacini G.
        Atomic resolution map of the soluble amyloid β assembly toxic surfaces.
        Chem. Sci. 2019; 10 (31360412): 6072-6082
        • Miller Y.
        • Ma B.
        • Nussinov R.
        Zinc ions promote Alzheimer Aβ aggregation via population shift of polymorphic states.
        Proc. Natl. Acad. Sci. U.S.A. 2010; 107 (20448202): 9490-9495
        • Melchor J.P.
        • Van Nostrand W.E.
        Fibrillar amyloid β-protein mediates the pathologic accumulation of its secreted precursor in human cerebrovascular smooth muscle cells.
        J. Biol. Chem. 2000; 275 (10734132): 9782-9791
        • Melchor J.P.
        • McVoy L.
        • Van Nostrand W.E.
        Charge alterations of E22 enhance the pathogenic properties of the amyloid β-protein.
        J. Neurochem. 2000; 74 (10800967): 2209-2212
        • Van Nostrand W.E.
        • Melchor J.P.
        • Cho H.S.
        • Greenberg S.M.
        • Rebeck G.W.
        Pathogenic effects of D23N Iowa mutant amyloid β-protein.
        J. Biol. Chem. 2001; 276 (11441013): 32860-32866
        • Vivoli Vega M..
        • Cascella R.
        • Chen S.W.
        • Fusco G.
        • De Simone A.
        • Dobson C.M.
        • Cecchi C.
        • Chiti F.
        The toxicity of misfolded protein oligomers Is independent of their secondary structure.
        ACS Chem. Biol. 2019; 14 (31074957): 1593-1600
        • Yamamoto N.
        • Van Nostrand W.E.
        • Yanagisawa K.
        Further evidence of local ganglioside-dependent amyloid β-protein assembly in brain.
        Neuroreport. 2006; 17 (17047463): 1735-1737
        • Yamamoto N.
        • Hirabayashi Y.
        • Amari M.
        • Yamaguchi H.
        • Romanov G.
        • Van Nostrand W.E.
        • Yanagisawa K.
        Assembly of hereditary amyloid β-protein variants in the presence of favorable gangliosides.
        FEBS Lett. 2005; 579 (15811339): 2185-2190
        • Lim K.H.
        • Kim Y.K.
        • Chang Y.T.
        Investigations of the molecular mechanism of metal-induced Aβ (1-40) amyloidogenesis.
        Biochemistry. 2007; 46 (17973493): 13523-13532
        • Shin B.K.
        • Saxena S.K.
        ESR spectroscopy suggests unequal contributions of the three histidine residues to Cu(II) binding in amyloid-β at physiological pH.
        Biophys. J. 2011; 100: 145
        • Summers K.L.
        • Schilling K.M.
        • Roseman G.
        • Markham K.A.
        • Dolgova N.V.
        • Kroll T.
        • Sokaras D.
        • Millhauser G.L.
        • Pickering I.J.
        • George G.N.
        X-ray absorption spectroscopy investigations of copper(II) coordination in the human amyloid β peptide.
        Inorg. Chem. 2019; 58 (31013069): 6294-6311
        • Drew S.C.
        • Barnham K.J.
        The heterogeneous nature of Cu2+ interactions with Alzheimer's amyloid-β peptide.
        Acc. Chem. Res. 2011; 44: 1146-1155
        • Clements A.
        • Allsop D.
        • Walsh D.M.
        • Williams C.H.
        Aggregation and metal-binding properties of mutant forms of the amyloid Aβ peptide of Alzheimer's disease.
        J. Neurochem. 1996; 66 (8592147): 740-747
        • Jun S.
        • Gillespie J.R.
        • Shin B.K.
        • Saxena S.
        The second Cu(II)-binding site in a proton-rich environment interferes with the aggregation of amylold-β(1-40) into amyloid fibrils.
        Biochemistry. 2009; 48 (19824649): 10724-10732
        • Ishii Y.
        A solid-state NMR study reveals structure and dynamics in copper(II)-binding to Alzheimer's β-amyloid fibrils.
        Biophys. J. 2010; 98: 457A-458A
        • Huang X.
        • Atwood C.S.
        • Hartshorn M.A.
        • Multhaup G.
        • Goldstein L.E.
        • Scarpa R.C.
        • Cuajungco M.P.
        • Gray D.N.
        • Lim J.
        • Moir R.D.
        • Tanzi R.E.
        • Bush A.I.
        The Aβ peptide of Alzheimer's disease directly produces hydrogen peroxide through metal ion reduction.
        Biochemistry. 1999; 38 (10386999): 7609-7616
        • Smith D.P.
        • Smith D.G.
        • Curtain C.C.
        • Boas J.F.
        • Pilbrow J.R.
        • Ciccotosto G.D.
        • Lau T.L.
        • Tew D.J.
        • Perez K.
        • Wade J.D.
        • Bush A.I.
        • Drew S.C.
        • Separovic F.
        • Masters C.L.
        • Cappai R.
        • Barnham K.J.
        Copper-mediated amyloid-β toxicity is associated with an intermolecular histidine bridge.
        J. Biol. Chem. 2006; 281 (16595673): 15145-15154
        • Pettersen E.F.
        • Goddard T.D.
        • Huang C.C.
        • Couch G.S.
        • Greenblatt D.M.
        • Meng E.C.
        • Ferrin T.E.
        UCSF chimera: a visualization system for exploratory research and analysis.
        J. Comput. Chem. 2004; 25 (15264254): 1605-1612
        • Case D.A.
        • Cheatham T.E.
        • Darden T.
        • Gohlke H.
        • Luo R.
        • Merz K.M.
        • Onufriev A.
        • Simmerling C.
        • Wang B.
        • Woods R.J.
        The Amber biomolecular simulation programs.
        J. Comput. Chem. 2005; 26 (16200636): 1668-1688
        • Maier J.A.
        • Martinez C.
        • Kasavajhala K.
        • Wickstrom L.
        • Hauser K.E.
        • Simmerling C.
        ff14SB: Improving the accuracy of protein side chain and backbone parameters from ff99SB.
        J. Chem. Theory Comput. 2015; 11 (26574453): 3696-3713