AbstractAmyloid-β peptide (Aβ) isoforms of different lengths and aggregation propensities coexist in vivo. These different isoforms are able to nucleate or frustrate the assembly of each other. N-terminally truncated Aβ(11–40) and Aβ(11–42) make up one fifth of plaque load yet nothing is known about their interaction with full-length Aβ(1–40/42). We show that in contrast to C-terminally truncated isoforms, which do not co-fibrillize, deletions of ten residues from the N terminus of Aβ have little impact on its ability to co-fibrillize with the full-length counterpart. As a consequence, N-terminally truncated Aβ will accelerate fiber formation and co-assemble into short rod-shaped fibers with its full-length Aβ counterpart. This has implications for the assembly kinetics, morphology, and toxicity of all Aβ isoforms.

AbstractAmyloid-β peptide (Aβ) isoforms of different lengths and aggregation propensities coexist in vivo. These different isoforms are able to nucleate or frustrate the assembly of each other. N-terminally truncated Aβ(11–40) and Aβ(11–42) make up one fifth of plaque load yet nothing is known about their interaction with full-length Aβ(1–40/42). We show that in contrast to C-terminally truncated isoforms, which do not co-fibrillize, deletions of ten residues from the N terminus of Aβ have little impact on its ability to co-fibrillize with the full-length counterpart. As a consequence, N-terminally truncated Aβ will accelerate fiber formation and co-assemble into short rod-shaped fibers with its full-length Aβ counterpart. This has implications for the assembly kinetics, morphology, and toxicity of all Aβ isoforms.

Amyloid fiber formation is a key event in many misfolding disorders. The ability to monitor the kinetics of fiber formation and other prefibrillar assemblies is therefore crucial for understanding these diseases. Here we compare three fluorescent probes for their ability to monitor fiber formation, ANS (1-anilinonaphthalene-8-sulfonic acid) and bis-ANS (4,4′-dianilino-1,1′-binaphthyl-5,5′-disulfonic acid) along with the more widely used thioflavin T (ThT). For this, we have used two highly amyloidogenic peptides: amyloid-β (Aβ) from Alzheimer’s disease and islet amyloid polypeptide (IAPP) associated with type II diabetes. Using a well-plate reader, we show all three fluorophores can report the kinetics of fiber formation. Indeed, bis-ANS is markedly more sensitive to fiber detection than ThT and has a submicromolar affinity for Aβ fibers. Furthermore, we show that fluorescence detection is very sensitive to the presence of excess fluorophore. In particular, beyond a 1:1 stoichiometry these probes demonstrate marked fluorescence quenching, for both Aβ and IAPP. Indeed, the fiber-associated fluorescence signal is almost completely quenched in the presence of excess ThT. There is also intense interest in the detection of prefibrillar amyloid assemblies, as oligomers and protofibrils are believed to be highly cytotoxic. We generate stable, fiber-free, prefibrillar assemblies of Aβ and survey their fluorescence with ANS and bis-ANS. Fluorescence from ANS has often been used as a marker for oligomers; however, we show ANS can fluoresce more strongly in the presence of fibers and should therefore be used as a probe for oligomers with caution.

The cellular prion protein (PrPC) binds to Cu2+ ions in vivo, and a misfolded form of PrPC is responsible for a range of transmissible spongiform encephalopathies. Recently, disruption of Cu2+ homeostasis in mice has been shown to impart resistance to scrapie infection. Using full-length PrPC and model peptide fragments, we monitor the sequential loading of Cu2+ ions onto PrPC using visible circular dichroism. We show the N-terminal amino group of PrPC is not the principal binding site for Cu2+; however, surprisingly, it has an affinity for Cu2+ tighter than that of the individual octarepeat binding sites present within PrPC. We re-evaluate what is understood about the sequential loading of Cu2+ onto the full-length protein and show for the first time that Cu2+ loads onto the N-terminal amino group before the single octarepeat binding sites.

•The related neuropeptides S1 and S2 both act as muscle relaxants in starfish.•S2 (SGPYSFNSGLTFamide) is ~ 10 × more potent than S1 (GFNSALMFamide).•S2, unlike S1, adopts a well-defined conformation in aqueous solution.•The N-terminal SGPY motif facilitates self-association of S2 at high concentrations.•Structural properties of S2 may have relevance to its biosynthesis and bioactivity.The neuropeptides S1 (GFNSALMFamide) and S2 (SGPYSFNSGLTFamide), which share sequence similarity, were discovered in the starfish Asterias rubens and are prototypical members of the SALMFamide family of neuropeptides in echinoderms. SALMFamide neuropeptides act as muscle relaxants and both S1 and S2 cause relaxation of cardiac stomach and tube foot preparations in vitro but S2 is an order of magnitude more potent than S1. Here we investigated a structural basis for this difference in potency using spectroscopic techniques. Circular dichroism spectroscopy showed that S1 does not have a defined structure in aqueous solution and this was supported by 2D nuclear magnetic resonance experiments. In contrast, we found that S2 has a well-defined conformation in aqueous solution. However, the conformation of S2 was concentration dependent, with increasing concentration inducing a transition from an unstructured to a structured conformation. Interestingly, this property of S2 was not observed in an N-terminally truncated analogue of S2 (short S2 or SS2; SFNSGLTFamide). Collectively, the data obtained indicate that the N-terminal region of S2 facilitates peptide self-association at high concentrations, which may have relevance to the biosynthesis and/or bioactivity of S2 in vivo.

There are a group of diseases associated with protein misfolding and accumulation into amyloid fibers. Many of these diseases have a major impact on human health, in particular, Alzheimer's (AD), Parkinson's (PD) and prion diseases. The focus of this review is to highlight how metal ions influence amyloid formation in a number of neurodegenerative diseases. Firstly, the various mechanisms by which metal ions might influence the kinetics of amyloid fiber formation are surveyed. The coordination of metal ions to a number of amyloidogenic proteins, with an emphasis on metal binding to intact fibers is reviewed. The kinetics of amyloid formation and the influence Cu2+, Zn2+, Fe3+ and Ca2+ have on amyloid-beta peptide (Aβ) fiber formation in AD is described in detail. The effect of metal ions on fibril formation for other amyloidogenic proteins, in particular Cu2+ binding to α-synuclein (αSyn) and the prion protein (PrP), are also reviewed. The mechanism by which metal ions might influence neurotoxicity of amyloids is also discussed. Levels of metal ions found at the synapse are described and related to the affinity of metal ions for Aβ, PrP and αSyn. In vivo evidence for a link between metal ions in these common neurodegenerative diseases, and the interplay between Aβ the prion protein and copper are reported. Finally, the possibility of a shared mechanism by which metal ions might influence amyloidosis is discussed.Graphical abstractHighlights► Metal ions influence amyloid formation in a number of neurodegenerative diseases. ► Cu2+ accelerates amyloid-beta peptide fiber formation and enhances cytotoxicity. ► Mechanisms by which metal ions accelerate the kinetics of fiber formation are discussed. ► Metal ions at the synapse, and the interplay between Aβ, the prion protein and copper are discussed.

The prion protein (PrP) is a cell-surface Cu2+ binding glycoprotein which can bind six copper ions. The role of Cu2+ in PrP function, misfolding, and prion disease has generated much interest; however, the field has been hampered by a lack of consensus with regard to the affinity of Cu2+ for PrPC. Here we build on our understanding of the appearance of visible CD spectra for full-length PrP and fragments to determine the affinity of Cu2+ for four different binding modes, with dissociation constants ranging between 13 and 66 nM at pH 7.4.

Cu2+ ions are found concentrated within senile plaques of Alzheimer’s disease patients directly bound to amyloid-β peptide (Aβ) and are linked to the neurotoxicity and self-association of Aβ. The affinity of Cu2+ for monomeric Aβ is highly disputed, and there have been no reports of affinity of Cu2+ for fibrillar Aβ. We therefore measured the affinity of Cu2+ for both monomeric and fibrillar Aβ(1−42) using two independent methods: fluorescence quenching and circular dichroism. The binding curves were almost identical for both fibrillar and monomeric forms. Competition studies with free glycine, l-histidine, and nitrilotriacetic acid (NTA) indicate an apparent (conditional) dissociation constant of 10−11 M, at pH 7.4. Previous studies of Cu-Aβ have typically found the affinity 2 or more orders of magnitude weaker, largely because the affinity of competing ligands or buffers has been underestimated. Aβ fibers are able to bind a full stoichiometric complement of Cu2+ ions with little change in their secondary structure and have coordination geometry identical to that of monomeric Aβ. Electron paramagnetic resonance studies (EPR) with Aβ His/Ala analogues suggest a dynamic view of the tetragonal Cu2+ complex, with axial as well as equatorial coordination of imidazole nitrogens creating an ensemble of coordination geometries in exchange between each other. Furthermore, the N-terminal amino group is essential for the formation of high-pH complex II. The Aβ(1−28) fragment binds an additional Cu2+ ion compared to full-length Aβ, with appreciable affinity. This second binding site is revealed in Aβ(1−42) upon addition of methanol, indicating hydrophobic interactions block the formation of this weaker carboxylate-rich complex. A Cu2+ affinity for Aβ of 1011 M−1 supports a modified amyloid cascade hypothesis in which Cu2+ is central to Aβ neurotoxicity.

Oxidative stress plays a key role in Alzheimer’s disease (AD). In addition, the abnormally high Cu2+ ion concentrations present in senile plaques has provoked a substantial interest in the relationship between the amyloid β peptide (Aβ) found within plaques and redox-active copper ions. There have been a number of studies monitoring reactive oxygen species (ROS) generation by copper and ascorbate that suggest that Aβ acts as a prooxidant producing H2O2. However, others have indicated Aβ acts as an antioxidant, but to date most cell-free studies directly monitoring ROS have not supported this hypothesis. We therefore chose to look again at ROS generation by both monomeric and fibrillar forms of Aβ under aerobic conditions in the presence of Cu2+ with/without the biological reductant ascorbate in a cell-free system. We used a variety of fluorescence and absorption based assays to monitor the production of ROS, as well as Cu2+ reduction. In contrast to previous studies, we show here that Aβ does not generate any more ROS than controls of Cu2+ and ascorbate. Aβ does not silence the redox activity of Cu2+/+ via chelation, but rather hydroxyl radicals produced as a result of Fenton−Haber Weiss reactions of ascorbate and Cu2+ rapidly react with Aβ; thus the potentially harmful radicals are quenched. In support of this, chemical modification of the Aβ peptide was examined using 1H NMR, and specific oxidation sites within the peptide were identified at the histidine and methionine residues. Our studies add significant weight to a modified amyloid cascade hypothesis in which sporadic AD is the result of Aβ being upregulated as a response to oxidative stress. However, our results do not preclude the possibility that Aβ in an oligomeric form may concentrate the redox-active copper at neuronal membranes and so cause lipid peroxidation.

The prion protein (PrPC) is a copper binding cell surface glycoprotein which when misfolded causes transmissible spongiform encephalopathies. The cooperative binding of Cu2+ to an unstructured octarepeat sequence within PrPC causes profound folding of this region. The use of NMR to determine the solution structure of the octarepeat region of PrP with Cu2+ bound has been hampered by the paramagnetic nature of the Cu2+ ions. Using NMR we have investigated the binding of candidate diamagnetic replacement ions, to the octarepeat region of PrP. We show that Pd2+ forms diamagnetic complexes with the peptides HGGG, HGGGW and QPHGGGWGQ with 1 : 1 stoichiometry. The 1H NMR spectra indicate that these peptides are in slow-exchange between free and bound Pd2+ on the chemical-shift time-scale. We demonstrate that the Pd–peptide complex forms slowly with a time taken to reach half-maximal signal of 3 hours. Other candidate metal ions, Ni2+, Pt2+ and Au3+, were investigated but only the Pd2+ complexes gave resolvable 1H NMR spectra. We have determined the solution structure of the QPHGGGWGQ–Pd 1 : 1 complex using 71 NOE distance restraints. A backbone RMSD of 0.30 Å was observed over residues 3 to 7 in the final ensemble. The co-ordinating ligands consist of the histidine imidazole side chain Nε, the amide N of the second and third glycines with possibly H2O as the fourth ligand. The co-ordination geometry differs markedly from that of the HGGGW–Cu crystal structure. This survey of potential replacement metal ions to Cu2+ provides insight into the metal specificity and co-ordination chemistry of the metal bound octarepeats.

A natively unfolded region of the prion protein, PrP(90–126) binds Cu2+ ions and is vital for prion propagation. Pentapeptides, acyl-GGGTH92–96 and acyl-TNMKH107–111, represent the minimum motif for this Cu2+ binding region. EPR and 1H NMR suggests that the coordination geometry for the two binding sites is very similar. However, the visible CD spectra of the two sites are very different, producing almost mirror image spectra. We have used a series of analogues of the pentapeptides containing His96 and His111 to rationalise these differences in the visible CD spectra. Using simple histidine-containing tri-peptides we have formulated a set of empirical rules that can predict the appearance of Cu2+ visible CD spectra involving histidine and amide main-chain coordination.

•Assembly of Amyloid-β (Aβ) peptide is fundamental to Alzheimer's disease.•As little as 0.01 mol equivalent of Zn2 + (100 nM) dramatically influences Aβ misfolding.•Zn2 + impacts misfolding even with Cu2 + and glutamate present.•A single Zn2 + ion influences assembly of many Aβ peptides through dynamic exchange.•Synaptic Zn2 +, despite its weak affinity for Aβ, can impact misfolding and assembly.The misfolding and self-assembly of amyloid-β (Aβ) into oligomers and fibres is fundamental to Alzheimer's disease pathology. Alzheimer's disease is a multifaceted disease. One factor that is thought to have a significant role in disease aetiology is Zn2 + homeostasis, which is disrupted in the brains of Alzheimer's disease sufferers and has been shown to modulate Alzheimer's symptoms in animal models. Here, we investigate how the kinetics of Aβ fibre growth are affected at a range of Zn2 + concentrations and we use transmission electron microscopy to characterise the aggregate assemblies formed. We demonstrate that for Aβ(1–40), and Aβ(1–42), as little as 0.01 mol equivalent of Zn2 + (100 nM) is sufficient to greatly perturb the formation of amyloid fibres irreversibly. Instead, Aβ(1–40) assembles into short, rod-like structures that pack tightly together into ordered stacks, whereas Aβ(1–42) forms short, crooked assemblies that knit together to form a mesh of disordered tangles. Our data suggest that a small number of Zn2 + ions are able to influence a great many Aβ molecules through the rapid exchange of Zn2 + between Aβ peptides. Surprisingly, although Cu2 + binds to Aβ 10,000 times tighter than Zn2 +, the effect of Zn2 + on Aβ assembly dominates in Cu2 +/Zn2 + mixtures, suggesting that trace levels of Zn2 + must have a profound effect on extracellular Aβ accumulation. Trace Zn2 + levels profoundly influence Aβ assembly even at concentrations weaker than its affinity for Aβ. These observations indicate that inhibitors of fibre assembly do not necessarily have to be at high concentration and affinity to have a profound impact.Download high-res image (536KB)Download full-size image

The cellular isoform of the prion protein PrPC is a Cu2+-binding cell surface glycoprotein that, when misfolded, is responsible for a range of transmissible spongiform encephalopathies. As changes in PrPC conformation are intimately linked with disease pathogenesis, the effect of Cu2+ ions on the structure and stability of the protein has been investigated. Urea unfolding studies indicate that Cu2+ ions destabilise the native fold of PrPC. The midpoint of the unfolding transition is reduced by 0.73 ± 0.07 M urea in the presence of 1 mol equiv of Cu2+. This equates to an appreciable difference in free energy of unfolding (2.02 ± 0.05 kJ mol− 1 at the midpoint of unfolding). We relate Cu2+-induced changes in secondary structure for full-length PrP(23–231) to smaller Cu2+ binding fragments. In particular, Cu2+-induced structural changes can directly be attributed to Cu2+ binding to the octarepeat region of PrPC. Furthermore, a β-sheet-like transition that is observed when Cu ions are bound to the amyloidogenic fragment of PrP (residues 90–126) is due only to local Cu2+ coordination to the individual binding sites centred at His95 and His110. Cu2+ binding does not directly generate a β-sheet conformation within PrPC; however, Cu2+ ions do destabilise the native fold of PrPC and may make the transition to a misfolded state more favourable.Research Highlights► PrP is a Cu2+-binding glycoprotein. ► Misfolding of PrP (PrPC) causes prion diseases. ► Urea unfolding studies indicate that Cu ions destabilise the native fold of PrPC. ► Cu2+-induced structural changes can directly be attributed to Cu2+ binding to the octarepeat region of PrPC. ► Cu2+ may make the transition to a misfolded state more favourable.

Oxidative stress is believed to play a central role in the pathogenesis of prion diseases, a group of fatal neurodegenerative disorders associated with a conformational change in the prion protein (PrPC). The precise physiological function of PrPC remains uncertain; however, Cu2+ binds to PrPC in vivo, suggesting a role for PrPC in copper homeostasis. Here we examine the oxidative processes associated with PrPC and Cu2+. 1H NMR was used to monitor chemical modifications of PrP fragments. Incubation of PrP fragments with ascorbate and CuCl2 showed specific metal-catalyzed oxidation of histidine residues, His96/111, and the methionine residues, Met109/112. The octarepeat region protects His96/111 and Met109/112 from oxidation, suggesting that PrP(90–231) might be more prone to chemical modification. We show that Cu2+/+ redox cycling is not ‘silenced’ by Cu2+ binding to PrP, as indicated by H2O2 production for full-length PrP. Surprisingly, although detection of Cu+ indicates that the octarepeat region of PrP is capable of reducing Cu2+ even in the absence of ascorbate, H2O2 is not generated unless ascorbate is present. Full-length PrP and fragments cause a dramatic reduction in detectable hydroxyl radicals in an ascorbate/Cu2+/O2 system; however, levels of H2O2 production are unaffected. This suggests that PrP does not affect levels of hydroxyl radical production via Fentons cycling, but the radicals cause highly localized chemical modification of PrPC.