Using a method based on ESR spectroscopy and spin-trapping, we have shown that Aβ (amyloid β-peptide) (implicated in Alzheimer's disease), α-synuclein (implicated in Parkinson's disease), ABri (British dementia peptide) (responsible for familial British dementia), certain toxic fragments of the prion protein (implicated in the transmissible spongiform encephalopathies) and the amylin peptide (found in the pancreas in Type 2 diabetes mellitus) all have the common ability to generate H2O2 in vitro. Numerous controls (reverse, scrambled and non-toxic peptides) lacked this property. We have also noted a positive correlation between the ability of the various proteins tested to generate H2O2 and their toxic effects on cultured cells. In the case of Aβ and ABri, we have shown that H2O2 is generated as a short burst during the early stages of aggregation and is associated with the presence of protofibrils or oligomers, rather than mature fibrils. H2O2 is readily converted into the aggressive hydroxyl radical by Fenton chemistry, and this extremely reactive radical could be responsible for much of the oxidative damage seen in all of the above disorders. We suggest that the formation of a redox-active complex involving the relevant amyloidogenic protein and certain transition-metal ions could play an important role in the pathogenesis of several different protein misfolding disorders.
- hydrogen peroxide
- oxidative stress
- reactive oxygen species
Introduction to protein conformational diseases or ‘proteopathies’
The formation of extracellular deposits of amyloid and/or intracellular inclusions containing amyloid-like protein fibrils is a key pathological feature of many different human neurodegenerative diseases . These include AD (Alzheimer's disease), Parkinson's disease, dementia with Lewy bodies, FTLD (frontotemporal lobar degeneration), the transmissible spongiform encephalopathies or ‘prion’ diseases, HD (Huntington's disease) and related trinucleotide repeat disorders (the spinocerebellar ataxias), MND (motor neuron disease), and some rarer conditions such as the familial British and Danish forms of dementia (see Table 1).
AD is perhaps the best known example of a localized form of brain amyloidosis. The two main histopathological features of this disease are the extracellular senile plaques, which contain amyloid fibrils composed of the Aβ (amyloid β-peptide), and the intracellular NFTs (neurofibrillary tangles), which mainly consist of a highly phosphorylated form of the microtubule-associated protein tau. Historically, the temporal sequence of events leading to this type of characteristic brain pathology in AD has been the subject of considerable dispute. According to the ‘amyloid hypothesis’ of AD, the deposition of Aβ in the brain is the seminal process that initiates a pathological cascade of events leading to NFT formation, neurodegeneration, neuronal and synaptic loss and dementia . The discovery of pathogenic mutations in the gene encoding the precursor to Aβ (i.e. the APP gene, on chromosome 21) has provided strong support for this hypothesis .
In addition to the APP gene, pathogenic mutations have now been found in the ASYN, MAPT, PRNP, IT15, SOD, TARDBP and BRI2 genes, which together encode all of the aggregating proteins presented in Table 1. This leads to a more ‘generalized form’ of the amyloid hypothesis, where misfolded and aggregating proteins play a key role in the molecular pathology of a whole range of different neurodegenerative conditions. In this respect, it is informative to note that, in the systemic amyloidoses, where fibrillar deposits accumulate in numerous different tissues and organs outside of the brain, it is well recognized that pathogenic mutations in the genes encoding the relevant fibril-forming proteins are often the cause of inherited disease. In the case of the brain diseases presented in Table 1, until recently, the only exception to this general rule was the TAR DNA-binding protein 43 (TDP-43), which was initially reported as forming the inclusions present in one type of FTLD (named FTLD with ubiquitinated inclusions, or FTLD-U) and in MND [3,4]. However, several different pathogenic mutations have now been found in the TARDBP gene [5–7], suggesting that the formation of TDP-43 inclusions is the initial cause of some forms of neurodegenerative disease.
Toxic properties of protein oligomers
Further evidence for the central importance of the various fibril-forming proteins (Table 1) in the pathogenesis of brain disease has arisen from the fact that most of them have been shown to be toxic to cultured neuronal cells in vitro. This implies a possible direct link between protein aggregation in the brain and the degeneration and loss of nerve cells in vivo. The precise molecular mechanisms responsible for this toxic effect are not altogether clear, but cell damage almost invariably seems to be due to changes in membrane permeability, Ca2+ ion influx and the induction of oxidative damage, often followed by apopotosis . Earlier studies emphasized the role of fully formed amyloid protein fibrils in the toxic mechanism, but, more recently, attention has focused on the possibility that much earlier protein assemblies could be responsible (Figure 1). Various types of these assemblies have been described and referred to as protofibrils, annular protofibrils, soluble oligomers, globular neurotoxins or (in the case of Aβ) ADDLs (Aβ-derived diffusible ligands) [9–16]. It is not always clear how the protein aggregates described by different research groups relate to each other. Also, whether these structures are early (obligate) intermediates on the assembly pathway to ‘mature’ amyloid fibrils, or whether they are formed independently of amyloid fibrils is not very well established. However, it is clear that these types of small protein assembly can be derived from several different disease-associated proteins and they are now thought to be at least partially responsible for some of their well-established toxic effects on cultured cells. Previous reports suggest that this could involve a common (probably conformation-dependent) mechanism, based on the insertion of these oligomers into cell membranes, causing damage and resultant changes in membrane permeability . However, the detailed molecular mechanisms responsible for this toxic effect are not clear. One possibility is that oligomers can form ‘pores’ or ‘ion channels’ in the membrane . Another possibility is that toxic oligomers can insert into membranes and generate ROS (reactive oxygen species) in situ, which would inflict oxidative damage on any molecules in their immediate vicinity, including the lipid and protein components of the membranes themselves [17–28], and this might explain reported changes in membrane permeability.
Role of H2O2 and redox-active transition-metal ions
There is mounting evidence for a major contribution played by oxidative stress in the pathology of most, if not all, of the neurodegenerative diseases mentioned above . Oxidative stress is due to an imbalance between the production of ROS and the ability of antioxidant defences to cope with ROS production. Evidence for oxidative stress covers features such as increased levels of redox-active transition-metal ions and the detection in the brain of products of lipid peroxidation and DNA, RNA and protein oxidation. In the case of AD, some researchers have stressed the fact that this type of oxidative damage could precede and even precipitate the aggregation of Aβ . However, Aβ has been shown to generate H2O2, a key ROS, directly from molecular oxygen, through electron-transfer interactions involving bound redox-active copper and/or iron ions . H2O2 is readily converted into the very aggressive hydroxyl radical, by Fenton chemistry, and this highly reactive free radical could be responsible for much of the early oxidative damage seen in AD.
Our research group at Lancaster University have confirmed the ability of Aβ to generate H2O2 in vitro by using ESR spectroscopy in conjunction with the spin-trapping technique [18,22–28]. Any H2O2 formed during peptide incubation is converted into hydroxyl radicals, via the Fenton reaction, upon addition of Fe(II). The resulting radicals are then trapped by DMPO (5,5-dimethyl-1-pyrroline N-oxide) to form the DMPO hydroxyl radical adduct, which has a uniquely characteristic four-line ESR spectrum. We have also shown that, in addition to Aβ, α-synuclein, , ABri (British dementia peptide), , certain toxic fragments of the prion protein [22,23] and the human form of the ‘amylin’ peptide found in the islets of Langerhans of the pancreas in late-onset (Type 2) diabetes mellitus  can also generate H2O2. In the latter case, the ESR technique was supplemented by detection of H2O2 with the highly sensitive and quantitative Amplex Red dye fluorescence method, and H2O2 formation was greatly increased in the presence of Cu(II) ions. Numerous controls (e.g. reversed, scrambled and non-toxic peptides) lacked any significant ability to generate H2O2 in our experiments, so demonstrating the specificity of its formation. Examples of this type of control are the non-amyloidogenic and non-toxic rodent form of amylin, and the reverse peptide Aβ-(40–1), both of which failed to generate H2O2. We have also noted a positive correlation between the ability of the various proteins and peptides tested to generate H2O2 and their toxic effects on cultured cells.
In studies published previously , we followed the timedependence of H2O2 formation during aggregation of Aβ and ABri. In both cases, H2O2 was generated as a short ‘burst’ during the early stages of peptide aggregation. Atomic force microscopy revealed the presence of structures resembling protofibrils or oligomers during these early periods of H2O2 formation (Figure 1), whereas mature Aβ fibrils lacked the ability to generate H2O2. Overall, our results suggest that the generation of H2O2 is associated with the presence of early oligomers or protofibrils, rather than the monomer or mature amyloid fibrils. Alternatively, it is possible that H2O2 is generated as a ‘by-product’ of the aggregation process itself, since it accumulates during an early period of rapid aggregation. This would be consistent with other reports closely relating toxicity to nucleation-dependent protein aggregation . In either case, an interaction of the amyloidogenic protein concerned with redox-active transition-metal ions is likely to be involved.
Interaction with redox-active transition-metal ions
It has been reported that Aβ can bind both Cu(II) and Fe(III) ions with high affinity [33,34] and also that the N-terminal domain of PrP (prion-related protein) can bind up to six Cu(II) ions with remarkably high affinity . The N-terminal domain of PrP contains four tandem repeats of the octapeptide sequence PHGGGWGQ which, alongside two histidine residues at positions 96 and 111, contribute to the overall Cu(II)-binding properties of the protein. The small toxic peptide PrP-(106–126) also binds Cu(II) ions, and we have shown that this confers on it the ability to generate H2O2 . α-Synuclein also binds Cu(II) ions, which facilitate its aggregation , and this protein also generates H2O2 . Our own recent results suggest a significant interaction between human amylin, but not rodent amylin, and Cu(II) ions, and, interestingly, only the former peptide can generate H2O2 .
The precise mechanism by which H2O2 is generated through these interactions between amyloidogenic proteins and their associated metal ions is unknown, but is likely to involve the formation of a redox-active complex in which the transition-metal ion is bound to the protein/peptide which then generates H2O2 from molecular oxygen. This process results in a reduction in the oxidation state of the metal ion and, potentially, oxidation of the peptide/protein. In addition the former is set up in ideal conditions for reaction with H2O2 via the Fenton reaction. As a result, the highly reactive and potentially damaging hydroxyl radical is formed, which becomes available to attack further the protein/peptide itself, or, in an in vivo situation, other molecules in its immediate vicinity. These reactions are outlined in Figure 2. Significantly, we have found that sufficient H2O2 is self-generated during the early stages of aggregation of Aβ to produce detectable peptidyl radicals, upon addition of Fe(II) . This supports the hypothesis that oxidative damage to Aβ and surrounding molecules in the brain in AD could be due, at least in part, to the self-generation of ROS. Clearly, a similar mechanism could operate in other protein conformational disorders.
Results from in vivo models
Numerous key observations regarding local brain amyloidosis and the potentially toxic effects of the proteins concerned have been made using animal models of neurodegenerative disease. Transgenic mice expressing wild-type or mutant forms of the aggregating proteins mentioned above can exhibit most of the characteristic histopathological, neurochemical and behavioural features of their equivalent human disease. These animal models have therefore been pivotal in providing additional support for the primary role of amyloid proteins in disease pathogenesis. They have also provided key results in support of the role of toxic oligomers and have helped to elucidate the temporal sequence of events regarding senile plaque formation and neurodegeneration in AD.
Some cultured cells expressing human mutant APP secrete low-order oligomeric forms of Aβ (e.g. dimers, trimers or tetramers) into their CM (culture medium). Conditioned CM from these cells, containing very low (nanomolar) concentrations of Aβ, has been shown to inhibit LTP (long-term potentiation) and impair the memory of complex learned behaviours when injected into the lateral ventricles of rats . Imunodepletion of this CM with anti-Aβ antibodies was reported to block these effects, whereas pre-incubation of CM with insulin-degrading enzyme (which degrades Aβ monomer, but not oligomers) did not alter these effects. This work suggests that small assemblies of naturally secreted Aβ oligomers are very potent neurotoxins that are potentially capable of affecting memory and LTP. However, it is still not clear how these observations relate to the neurotoxic properties of protein aggregates towards cultured cells, which are usually carried out at higher concentrations and often involve poorly defined preparations (in terms of their exact state of protein assembly) derived from synthetic peptides or recombinant proteins. Furthermore, it is not yet clear how these observations relate to neurodegeneration in the human brain.
The temporal sequence of events relating Aβ deposition to neuritic alteration has now been clearly established in one particular transgenic mouse model of AD. In an important recent publication , longitudinal in vivo multiphoton microscopy was used to obtain repeat images of the brains of a transgenic mouse strain known to develop senile plaques containing Aβ deposits. Surprisingly, it was found that these lesions develop extraordinarily quickly (over the period of 1 day), but, perhaps more importantly, progressive neuritic changes were only seen a few days after the initial appearance of a newly developed senile plaque. This observation demonstrates clearly that amyloid deposition precedes neuronal damage, at least in this transgenic mouse model, so providing support for the amyloid hypothesis. However, it should be borne in mind that it is uncertain whether these local ‘neuritic changes’ in the vicinity of the senile plaque do inevitably lead to the more widespread neurofibrillary pathology, neurodegeneration and nerve cell loss that is typical of fully developed AD.
Arrasate et al.  have made some important observations in a mouse model of HD that appear, on the face of it, to contradict much of what has been stated above. These authors used an ingenious robotic microscope to track hundreds of individual neurons expressing the human mutant form of the huntingtin protein (with its polyglutamine-expanded tract) that is the direct cause of HD. Mutant huntingtin was tagged with green fluorescent protein to allow its distribution to be seen via fluorescent imaging, and the formation of inclusion bodies and the survival of nerve cells were then tracked in real time. Perhaps surprisingly, the cells containing the inclusion bodies were actually the ones that survived in this experiment, with the unhealthy-looking nerve cells having no huntingtin inclusions. This observation can be reconciled with the more ‘generalized version’ of the amyloid hypothesis mentioned above by considering that intracellular inclusion bodies are actually a beneficial coping response for sequestering damaging protein aggregates and thereby reducing levels of toxic oligomers elsewhere in the cell, so prolonging neuronal survival. If this is the case, then those cells without inclusions could actually be more exposed to the damaging effects of the toxic oligomers.
Small oligomeric assemblies, referred to under various terms, are likely to be responsible for many of the reported toxic effects of several different amyloidogenic proteins in vitro and could be responsible for neurodegeneration in vivo in a variety of different brain diseases and possibly cellular degeneration in other forms of amyloidosis, such as Type 2 diabetes mellitus. If so, they will provide a useful target for the development of novel treatments for these diseases [41,42]. An interaction between the proteins concerned and redox-active transition-metal ions, particularly those of iron and copper, and the consequent generation of ROS, could provide the key to understanding at least some of the pathological effects of these misfolded proteins.
Our current research in this area is funded by The Alzheimer's Society, U.K. We are also grateful for previous financial support from The Wellcome Trust and The Ford Foundation (U.S.A.).
Metal Metabolism: Transport, Development and Neurodegeneration: A Biochemical Society Focused Meeting held at Imperial College London, U.K., 9–10 July 2008. Organized and Edited by David Allsop (Lancaster, U.K.) and Harry McArdle (Rowett Research Institute, Aberdeen, U.K.).
Abbreviations: Aβ, amyloid β-peptide; ABri, British dementia peptide; AD, Alzheimer's disease; APP, amyloid precursor protein; CM, culture medium; DMPO, 5,5-dimethyl-1-pyrroline N-oxide; FTLD, frontotemporal lobar degeneration; HD, Huntington's disease; LTP, long-term potentiation; MND, motor neuron disease; NFT, neurofibrillary tangle; PrP, prion-related protein; ROS, reactive oxygen species; TDP-43, TAR DNA-binding protein 43
- © The Authors Journal compilation © 2008 Biochemical Society