Toxicity of Amyloid β Peptide: Tales of Calcium, Mitochondria ...

478648.qxd

1/22/04

6:01 PM

Page 637

Neurochemical Research, Vol. 29, No. 3, March 2004 (© 2004), pp. 637–650

Toxicity of Amyloid Peptide: Tales of Calcium, Mitochondria, and Oxidative Stress* Laura Canevari,1,2 Andrey Y. Abramov,1 and Michael R. Duchen1 (Accepted October 14, 2003)

Alzheimer’s disease (AD) is characterized by the accumulation of amyloid- (A) peptides. Although the disease undoubtedly reflects the interaction of complex multifactorial processes, A itself is toxic to neurons in vitro and the load of A in vivo correlates well with the degree of cognitive impairment. There has therefore been considerable interest in the mechanism(s) of A neurotoxicity. We here review the basic biology of A processing and consider some of the major areas of focus of this research. It is clear that both AD and A toxicity are characterized by oxidative stress, alterations in the activity of enzymes of intermediary metabolism, and mitochondrial dysfunction, especially impaired activity of cytochrome c oxidase. Studies in vitro also show alterations in cellular calcium signaling. We consider the mechanisms proposed to mediate cell injury and explore evidence to indicate which of these many changes in function are primary and which secondary. KEY WORDS: Amyloid peptide; intracellular calcium; oxidative stress; mitochondria; NADPH oxidase.

INTRODUCTION

obscure. Nevertheless, accumulation of A appears to be sufficient to cause dementia. This is supported by the fact that patients with Down’s syndrome, who have an extra copy of the chromosome carrying the gene for the A precursor protein (APP), develop plaques and dementia at an early age. Furthermore, all the known mutations associated with familial AD are in the genes for APP or for the presenilins, enzymes involved in the processing of APP—all of which result in A accumulation. The other pathological feature of AD, the NFT, are composed of hyperphosphorylated tau protein, appears to be a consequence of A pathology; tau mutations have been described that result in NFT formation but no deposition of amyloid (1) and Down’s syndrome patients develop A deposits before NFT (2). The “amyloid hypothesis” has nevertheless been challenged in view of the fact that APP transgenic mice show signs of electrophysiological and behavioral changes before any significant A deposition (3) and that the plaque number does not correlate clearly with cognitive

Alzheimer’s disease (AD) is the most common form of dementia, and it is by definition characterized by the accumulation in the brain of extracellular neuritic plaques, together with the presence of intraneuronal neurofibrillary tangles (NFT). The plaques are mainly composed of amyloid- (A), a 39–43 amino acid peptide, forming large insoluble fibrillary aggregates, and are surrounded by dystrophic neurites and activated glial cells. The peptide is normally present in the brain in small amounts, although its possible physiological role remains * Special issue dedicated to Professor John B. Clark. 1 Division of Neurochemistry, Institute of Neurology, London, United Kingdom and Mitochondrial Biology Group, Department of Physiology, UCL, London, United Kingdom. 2 Address reprint requests to: L. Canevari, Division of Neurochemistry, Institute of Neurology, Queen Square, London WC1N 3BG, United Kingdom. Tel: 44-(0)20-7837-3611, ext. 4153; Fax: 44-(0)207833-1016; E-mail: [email protected]

637 0364-3190/04/0300–0637/0 © 2004 Plenum Publishing Corporation

478648.qxd

1/22/04

6:01 PM

Page 638

638

Canevari, Abramov, and Duchen

impairment. Indeed, in mouse models showing excessive amyloid deposition, the neuronal loss is modest, even though the relative amyloid burden may be far greater than that found in patients with AD (4). Some cognitively intact elderly individuals also have been described in whom a considerable plaque load has been found at postmortem examination (5). One likely explanation of this apparent conflict is that the pathogenic form of A is not the large, relatively unreactive, static plaque, but also perhaps that a combination of amyloid deposition and tau pathology is required to promote the full AD phenotype. In addition, in AD, it seems to be the total A load, rather than the plaque load, that correlates with the degree of cognitive impairment (6). In a recent study, a transgenic mouse was generated containing mutations for A, PS-1, and tau (7). The mice develop defects in synaptic plasticity manifest as a failure of long-term potentiation (LTP) before the appearance of plaques and correlating rather with the intraneuronal accumulation of A. This raises fascinating questions about the action of intraneuronal A and the mechanism whereby its accumulation interferes with synaptic physiology. AD is most certainly a complex and multifactorial disorder, for which aging is the main identified risk factor. However, there are many other interacting components contributing to its occurrence, some of which we are just starting to identify. In this review, we will focus on A and the various hypotheses on the mechanisms of toxicity. We will highlight the importance of mitochondria in AD and other neurodegenerative diseases and the links with oxidative stress and the dynamics of intracellular calcium handling.

DISCUSSION APP and A Generation APP is a 695-770 amino acid membrane-spanning glycoprotein expressed in brain and other tissues. The Nterminal is found exposed to the extracellular surface or toward the lumen of intracellular vesicles such as the endoplasmic reticulum, trans-Golgi network and endocytotic vesicles. The normal function of APP remains unclear, although a role has been proposed in extracellular signal transduction, via binding to the G0 protein (8), in axonal transport via its interaction with kinesin-I (9), and for its Kunitz-type protease inhibitor domain, in addition to neurotrophic and neuroprotective activities (for a review, see 10). However, deletion of the APP gene in mutant mice does not change life expectancy or produce any major pathology except some gliosis in the adults (11), suggesting that homologous proteins may be expressed. APP Processing. APP is processed via at least two proteolytic pathways (Fig. 1). One involves cleavage by enzymes with -secretase activity, which cut in the middle of the A sequence, thereby generating an extracellular, soluble sAPP, and a smaller C-terminal fragment known as p3. Alternatively, APP can be processed by an amyloidogenic route, which generates A and a soluble sAPP, by the sequential cleavage by - and -secretases. The -secretase is an enzymatic complex comprising the presenilins (PS1 or PS2) together with nicastrin, aph-1, and pen-2 (12), mostly associated with the endoplasmic

Fig. 1. Schematic diagram of APP processing.

478648.qxd

1/22/04

6:01 PM

Page 639

Mechanism(s) of A Toxicity reticulum. It is part of a recently described family of aspartate proteases that cut within the membrane-spanning domain of A and of other protein substrates such as Notch (13), therefore possibly involved in neural development. The -secretases (BACE1 and BACE2) are also aspartic proteases, present in the Golgi apparatus and endosomes, whose physiological function is not known. It has been suggested that the expression of BACE may be increased in AD (14) but BACE knockout mice have a normal phenotype (15). Both secretases are the object of intense research as potential therapeutic targets, but so far it has proved problematic to find -secretase inhibitors that do not affect Notch signaling or -secretase inhibitors sufficiently small and lipophilic to reach their target. A certain heterogeneity is found in the A peptides in AD brains, both at the N-terminal, where several amino acids can be truncated, and at the C-terminal, where the -secretase can cleave at amino acid 40 or 42. The 1–40 form is normally the most abundant, but the 1–42 is the form that is most increased during AD. Little is known about the factors causing the accumulation of A—resulting from either increased production or decreased turnover—over and above the minimal amounts found in normal aging brains, to the levels associated with AD. The production of A, in particular A1–42 is increased by mutations in the gene for APP, such as the “Swedish” mutation (16), but familial AD cases are only a very small minority compared to the number of sporadic AD patients. Accumulation resulting from altered endosomal function is another intriguing possibility (17). The presence of the 4 isoform of apolipoprotein E (ApoE4) genotype is associated with higher incidence of AD (18) and higher amounts of A deposits in the brain (19). ApoE3 binds A more strongly than ApoE4 (20) and is protective against A toxicity in vitro (21). A Aggregation and Toxicity A is toxic to cells in culture and in vivo (22,23). Early in vitro studies correlated A toxicity to neuronal cultures with its state of aggregation (22), with the more toxic 1–42 aggregating more rapidly than the 1–40 peptide (24). A molecules, mainly with random coil structure, can spontaneously aggregate to form, in sequence, low molecular weight oligomers (3–6 units, and up to 24 units), then short, flexible protofibrils (200 nm) and long, rigid fibrils—all in a dynamic equilibrium with each other. Fibrils have a high percentage of -sheet conformation, which is the basis of its positive staining with Congo red and thioflavin T and have classically been considered the toxic form of A. However, it has been demonstrated that oligomers and protofibrils are also

639 toxic, possibly the main toxic form in AD (25–27). This has obvious implications for the validity of solubilization of fibrillary A or immunization with fibrillary antigens, which are being investigated as possible therapeutic approaches. Other factors known to promote the aggregation of soluble A include: high peptide concentration; acidic pH (28); the presence of metal ions such as iron, zinc, and aluminium (29); the presence of certain phospholipid metabolites (30); and oxidative stress (31). A Localization It appears that the critical process of A oligomerization begins intracellularly, particularly in neurons (32). It has been proposed that intracellular accumulation of A leads to neuronal lysis and dispersion of the cell contents into the surrounding parenchyma. According to this hypothesis, the aggregated A from the dead neurons would be released extracellularly (33) and possibly act as a nidus for nucleation and aggregation of the soluble extracellular pool of A, to form plaques; Meyer-Luehmann et al. (34) recently demonstrated the importance of the “amplification” role of extracellular A. Intracellular A, especially 1–42, co-localizes with NFTs and is associated with signs of cytoskeletal degeneration (35). Furthermore, the mRNA found in senile plaques in AD brain is predominantly of neuronal origin (36). Intraneuronal A accumulation has been directly observed (37,38) in both human brain tissue and transgenic mice models and has been associated with synaptic pathology and dysfunction (39,40). A can be generated from APP present on the plasma membrane and simultaneously secreted in the extracellular space. However, three intracellular generation pathways have also been described: (i) via the endosomal-lysosomal system, which degrades APP reinternalized from the cell surface; (ii) via the Golgi apparatus, which produces mainly the 40 amino acid form of A (A1–40); and (iii) via the endoplasmic reticulum (ER), which mainly yields the 42 amino acid A (A1–42); this is retained within the cells, predominantly neurons (41–46). Mechanism of Toxicity The exact sequence of events leading to neuronal damage in AD is not known. Several potentially damaging pathways appear to be activated, some of which are also found as a consequence of A treatment of in vitro preparations. For example, it seems clear that the effects of A are associated with oxidative stress, mitochondrial dysfunction, disturbances of Ca2 homeostasis, NO

478648.qxd

1/22/04

6:01 PM

Page 640

640 generation, microglial activation, and many others, but the causal relationship or the exact sequence of these events remain controversial. Several findings suggest a prevalence of necrotic death in AD, for example, the presence of necrotic but not apoptotic morphology in AD brain (47), the evidence for necrotic neuronal death in the hippocampus of PS-1 knockin mice (48) and the reported necrotic effect of A (49). However, a large volume of literature describes evidence for the prevalence of apoptosis in AD, such as DNA fragmentation (50), alteration of the balance of proapoptotic and antiapoptotic molecules (51,52) and products of caspase cleavage (53). Treatment of cells in vitro with A has been shown by others to induce apoptosis (54–56). A number of caspases have been implicated in A toxicity, especially 2, 3, 8, and 9 (reviewed in 57), but there is no consensus on what relevance these may have in the pathogenesis of AD. One possibility is that caspase activation may play a role that is not necessarily related to apoptosis; both APP and presenilins are cleaved by caspase-3, increasing the generation of A (58,59). Essentially, a conclusive verdict on the mechanism of cell death in AD remains to be reached. The conditions associated with the disease such as oxidative stress, calcium overload, etc., are known to induce either a necrotic or an apoptotic type of death. In all likelihood, these two mechanisms of cell death might be largely overlapping, possibly depending on the intensity of the stimuli and on the bioenergetic status of the cell. In any case, it must be remembered that neuronal loss, which certainly occurs in AD, may not necessarily be at the basis of dementia. Rather, neuronal dysfunction and synapse loss, which precede extensive cell death, are thought to be the functional correlates to cognitive impairment. Regeur and et al. (60) found no difference in neuronal numbers between control and demented patients. Although it seems clear that there is substantial cell loss late in the disease, the principle that early-stage disease may reflect altered function rather than cell death has profound implications for potential therapies for early stages of AD, where it may be possible to recover dysfunctional neurones before the irreversible cell loss. Mitochondria as Targets and Mediators of Toxicity Mitochondria are essential for neuronal function, because the limited glycolytic capacity of these cells makes them highly dependent on aerobic oxidative phosphorylation for their energetic needs. A decrease in

Canevari, Abramov, and Duchen cerebral glucose utilization has been observed in AD, even when corrected for brain atrophy (61). Mitochondrial abnormalities, namely a decrease in mitochondrial mass and reduced mtDNA content, have been shown to be a very early pathological sign in AD, preceding the appearance of NFT, specifically in the same neurons vulnerable to degeneration (62,63). Recently, a most intriguing study has shown that APP may also be mitochondrially targeted in neurons in a transgenic mouse model. The full-length molecule interacts with the mitochondrial protein import motors, but transport is not complete, causing mitochondrial dysfunction and inhibition of ATP synthesis (64). Another convincing corroboration that mitochondria are involved in AD pathology comes from the finding that cybrid cells containing mitochondrial DNA from sporadic AD patients show characteristics compatible with the disease, including increased secretion and intracellular accumulation of A, decreased mitochondrial membrane potential, oxidative stress, decreased cytochrome c oxidase (COX; complex IV of the respiratory chain) activity, and altered mitochondrial morphology (65,66). However, so far, convincing evidence of mtDNA mutations being linked to AD is still lacking. Lin et al. (67) found an accumulation of point mutations in mtDNA with aging and AD, correlated with COX activity, but no significant difference between elderly controls and AD patients. The activity of key mitochondrial enzymes such as -ketoglutarate dehydrogenase (-KGDH), pyruvate dehydrogenase (PDH), and COX is decreased in AD (68–71). In particular, the literature showing COX deficiency in AD is abundant, showing reductions in COX mRNA (72) and protein expression (73) and decreased COX activity even in peripheral tissues (74). APP overexpressed in cultured cells leads to decreased COX activity (75), abnormal mitochondrial morphology and decreased mitochondrial membrane potential (76). We and others have found that exposure of isolated rat brain mitochondria to A causes a decrease in mitochondrial enzyme activity, respiration, and membrane potential (77–79). It has been shown that A induces cytochrome c release and caspase activation only in cells with a functional respiratory chain (80), suggesting that functional mitochondria play a significant part as targets or mediators of A toxicity. A potentiated calcium-induced opening of cyclosporine A–dependent pores in the inner mitochondrial membrane (the mitochondrial permeability transition or mPTP) and mitochondrial swelling (81). In addition to being sensitive targets of oxidative damage, dysfunctional mitochondria are also a source of reactive oxygen species (ROS) (82 and see below), and mitochondrial inhibition,

478648.qxd

1/22/04

6:01 PM

Page 641

Mechanism(s) of A Toxicity which increases ROS generation, has also been shown to increase the amyloidogenic processing of APP (83), creating the conditions for further cell damage. Amyloid Toxicity and Oxidative Stress It has been suggested for many years that the pathogenesis of the neurodegeneration in AD involves oxidative stress in some form. Evidence for oxidative damage to proteins and DNA and increased levels of lipid peroxidation have all been described in postmortem tissue from patients with AD. Widespread peroxynitrite-induced nitration of tyrosine residues has been described (84), and A has also been shown using a variety of assays to cause increased production of ROS or impaired antioxidant defences (by definition, oxidative stress) in a number of model systems. Similarly, considerable evidence suggests that antioxidant strategies may protect cell model systems from A-induced toxicity (e.g., for review, see 85). The key questions then are the relative importance of oxidative stress as a contributor to A-induced neurodegeneration, the mechanism(s) by which A causes oxidative stress and the mechanism(s) whereby oxidative stress leads to neurodegeneration. As described above, the activity of a number of enzymes involved in intermediary metabolism is decreased in AD brains and in cells exposed to A— notably the activity of glutamine synthetase, creatine kinase, -KGDH, PDH, and aconitase (see 86 and 87). These are all enzymes that are very highly sensitive to oxidative modification and are altered by exposure to a range of pro-oxidants (e.g., see 88). Another system shown to be vulnerable to oxidative modification is the glutamate transporter on astrocytes (see 89), and it has been suggested that the failure of the transport pathway might lead to extracellular glutamate accumulation and therefore glutamate excitotoxic neurodegeneration. In this respect, the data are not straightforward, and Abe and Misawa (90) have suggested that exposure of astrocytes to A in fact upregulates the activity of the glutamate transporter, enhancing the rate of clearance of glutamate from the extracellular space and thus altering synaptic processing. A and Glutathione Depletion. One major index of oxidative stress is the level of reduced glutathione (GSH; critically, it is the redox state of GSH, or GSH/GSSG ratio that matters), and oxidation of GSH has been described repeatedly in a variety of model systems in AD brains, other tissues of patients with familial forms of AD, and models in which cells have been exposed to A (91–93). Glutathione (GSH) is one of the major antioxidant systems in the CNS. Astrocytes

641 have a substantially higher GSH content than neurons and appear responsible for the maintenance of GSH levels in neurons by releasing GSH, which is broken down by an ectoenzyme into a dipeptide that can then be used by neurons to maintain their own levels of GSH (94,95). Using confocal imaging of [GSH] in mixed hippocampal cultures we found that exposure to A for 24 h caused depletion of GSH in both astrocytes and in neurons, identifiable separately within the co-culture in the imaging experiments (Fig. 2; see 93 and 94). GSH depletion was Ca2 dependent, suggesting that alterations in Ca2 homeostasis might lie upstream in the cascade of injury caused by A and that the oxidative injury might itself be Ca2 dependent (Fig. 3). The protection of cells by GSH precursors (A. Y. Abramov, L. Canevari and M. R. Duchen, unpublished observations) or upregulation of GSH synthesis (96), however, strongly suggests that GSH depletion plays a major role in the progression towards cell death. Mechanisms of A-Induced ROS Generation. All these data point to a major role of oxidative stress in response to A and indicate multiple potential targets of oxidative damage. Several mechanisms have been proposed whereby A may increase ROS generation. A may apparently generate oxygen radicals directly in solution, but may also interact with a number of biological systems to increase the rate of radical production through modification or stimulation of intrinsic pathways. The most obvious of these are probably the activation of endogenous radical generating systems in microglia and possibly other cell types in the CNS by activation of the flavoprotein linked enzyme system NADPH oxidase. As indicated above, it has also been suggested that A may increase the production of ROS from mitochondria by causing damage to the mitochondrial respiratory chain (97). A, as a small soluble aggregate, appears able to insert into the lipid membrane and to generate ROS directly. The disruption of cellular homeostasis may simply follow as an inevitable consequence of the peroxidation of lipids, causing damage to cell membranes, alterations in the activity of ion channels and transport systems and so to a loss of regulated cell function and ultimately to neurodegeneration. It has been suggested that the decreasing antioxidant capacity of the aging brain might account for the greater susceptibility to A-induced damage with age. Given other data discussed in the present review, further amplification processes might involve amyloidinduced calcium influx, (which can enhance mitochondrial ROS generation), and enhanced ROS generation resulting from damage to the mitochondrial respiratory chain.

478648.qxd

1/22/04

6:01 PM

Page 642

642


Fig. 2. Confocal imaging of monochlorobimane fluorescence can be used to differentiate GSH levels in astrocytes and neurons in co-culture. Hippocampal neurons were stained with MCB until the signal reached a steady state (abut 30 min). The MCB reacts with GSH in a reaction catalyzed by glutathione S-transferase producing a fluorescent adduct excited in the UV with an emission at about 450 nm. In the images shown, the preparation still bathed in standard physiological saline was placed on the stage of a Zeiss 510 UV-vis confocal imaging system. When images were obtained in the Z-plane, images of astrocytes that grow on the glass coverslip showed a very bright signal. Focusing up to the neurons, which tend to grow on the astrocyte layer, revealed a much darker signal indicating much lower levels of GSH (typically about 3-fold darker than the astrocytes (see Keelan et al., 2001(94)). This is shown in panel C, in which the intensity of signal with space is shown in 3 dimensions, showing a bright layer of astrocytes above which sits the much darker neuron.

A alone in solution—without the presence of any other biological or cellular material—can apparently promote ROS generation directly (for a review, see 85), possibly through a reaction of A with trace levels of Cu2. However, the truncated peptide 25–35 also causes ROS generation, even though it appears to lack the Cu2 binding site. Trace levels of iron also have been suggested to interact with A, but this appears not to be a consistent observation. A peptide binds to metal ions with a selectivity Cu2 Fe3 Zn2 (98), all of which promote aggregation. Cu2 can undergo redox cycling and generate ROS, while Zn2 is not redox active but competes with Cu2 for binding and therefore inhibits the oxidant properties of metal-bound A (99). Trace

amounts of metals may also promote A aggregation in an -helix conformation, while a high concentration of Zn2 (and to a lesser extent, Cu2) promotes -sheet fibrillar aggregation, which is classically associated with A toxicity. Clioquinol, a chelator of Cu2, Zn2, and Fe2, prevents aggregation and resolubilizes A, and also has been shown to have a beneficial effect in mouse models of AD (100,101). It has been shown also to abolish some of the effects of A and to be neuroprotective on cells in culture (93). A further possible mechanism whereby A could increase ROS generation involves the action of A on mitochondria. We have discussed above that A may cause mitochondrial dysfunction through a variety of

478648.qxd

1/22/04

6:01 PM

Page 643

Mechanism(s) of A Toxicity

643

Fig. 3. A causes Ca2-dependent depletion of GSH in hippocampal astrocytes and neurons. MCB fluorescence was measured in cultures of hippocampal neurons and astrocytes before and after exposure to A 25–35 (50 M) for 24 h in culture medium at 37°C the presence (gray bars) and absence (dark bars) of extracellular calcium. Mean values of the fluorescence intensity are indicated and show that GSH was dramatically reduced in astrocytes and also significantly reduced in neurons. The loss of GSH was clearly Ca2 dependent, suggesting that the change in GSH is secondary to changes in Ca2.

mechanisms, including the failure of substrate supply through effects on -KGDH and other enzymes of intermediary metabolism, but may also cause impaired activity of COX. The former will simply cause a loss of mitochondrial potential but will not increase mitochondrial ROS generation. However, damage to COX may increase mitochondrial ROS generation as the intermediary complexes become fully reduced and transfer electrons to available oxygen if downstream carriers cannot accept them. This seems unlikely to represent a primary mechanism of cell injury in A toxicity—after all, the mitochondrial damage must come first and is likely to be a reflection of oxidative stress generated through another pathway—but it will provide an amplifying mechanism that may compound oxidative injury initiated elsewhere. A and NADPH Oxidase. The NADPH oxidase in an enzyme principally associated with phagocytic immune cells—neutrophils, macrophages, and microglia. The enzyme generates superoxide radicals through an electron transfer reaction between NADPH and flavins,

a process thought to be fundamental in the bactericidal activity of neutrophils. A number of studies have suggested that A can increase the rate of ROS generation in immunocompetent cells, although the models vary considerably. A has been reported to activate neutrophils directly, to enhance neutrophil activation by other stimuli (such as PMA, which activates the enzyme through activation of PKC, 103) and to cause enhanced ROS generation in response to activation by P2x receptors (which raise [Ca2]c) of neutrophils from patients with familial AD (104). It remains unclear to what extent these effects relate to neurodegeneration in AD patients. Certainly, neuritic plaques are enriched in microglia and microglial activation seems a more likely contributor to neurodegeneration than the activation of peripheral neutrophils, although of course alterations in the periphery may reflect interesting changes in cell biology in these patients. Several studies have suggested that the effects of A on immunocompetent cells may be mediated through the activation of the so-called scavenger receptor CD36,

478648.qxd

1/22/04

6:01 PM

Page 644

644 which is apparently expressed on microglia and possibly on astrocytes, and the AGE receptors (RAGE, receptors for advanced glycation end products). Interestingly, microglia adhere to surfaces coated with fibrillar A, and that adhesion appears to be mediated by CD36. Husemann and Silverstein (105) suggested that the expression of CD36 receptors is upregulated in AD brains, although El-Khoury et al. (106) find expression in normal brains also. The latter group have shown that microglia from CD36-null mice show reduced activation in response to A and so suggest that the production of cytokines and of ROS by the A activated microglia may be CD36 dependent. RAGE have also been shown to activate the NADPH oxidase (107), and their expression is enhanced during AD in neurons, microglia, and astrocytes (108). A small cohort of papers have also shown expression of the NADPH oxidase or of CD36 receptors in astrocytes and on an immortalized astrocyte cell line, the C6 glioma cell line. Most interestingly, increased ROS generation was demonstrated in the C6 cells in response to A (109). That ROS generation was suppressed by expression of a dominant negative form of rac1, a small GTPase that plays an important role in the activation of the NADPH oxidase, even though the evidence for the expression and function of the enzyme in astrocytes is minimal. This raises fascinating questions about the mechanisms of ROS formation in astrocytes and the mechanisms of cell activation. In our own recent study (93), we found that A caused a loss of GSH in astrocytes, suggesting that astrocytes might represent the primary source of ROS and of oxidative stress. We also recently found that A induced an increase in ROS production in astrocytes that was blocked by inhibitors of the NADPH oxidase, diphenylene iodonium, and apocynin. These drugs also prevented A-induced GSH depletion and protected neurons from cell death, strongly implicating an astrocytic NADPH oxidase in the pathway to oxidative stress and cell death (Abramov, Canevari and Duchen, in press). The question also arises whether microglial activation triggers neurodegeneration “simply” through ROS generation or whether other aspects of microglial activation are more important. This is not the appropriate place to discuss details of the consequences of microglial activation, which has been extensively reviewed recently (e.g., see 110). However, although we have already alluded to the signs of oxidative injury that seem so prominent in all aspects of Alzheimer’s pathology, it is important to remember that activated microglia (and indeed, activated astrocytes) also release a range of cytokines and the cathepsins (cysteine endopeptidases),

Canevari, Abramov, and Duchen which may play critical roles as triggers to neurodegeneration. It remains possible, perhaps even likely, that these mechanisms act in concert and that the oxidative stress and GSH depletion, rather than being themselves mechanisms of lethality, render cells vulnerable to other normally innocuous stimuli. A and [Ca2]c Signaling Although effects of amyloid on [Ca2]c signaling have been extensively studied by a number of groups, these studies have thrown up a range of observations without any clear consensus on mechanism or the relationship between alterations in calcium signaling and the pathogenesis of the disease. Essentially, these studies seem to fall into the following several categories: i. Alterations in the amplitude or time course of neuronal calcium signals that could cause alterations in neuronal signal processing and thus directly to alterations in CNS function ii. A sensitization of neurons to calcium signals mediated by glutamate that might increase neuronal vulnerability to excitotoxicity and therefore increase neuronal cell death in response to modest levels of glutamate iii. Alterations in calcium signals in microglia associated with their activation that therefore generate an inflammatory response and thus initiate pathological changes in neighboring neurons iv. Alterations in calcium signals in astrocytes that might initiate alterations in the microenvironment of the neurons and so alter their function Calcium signals—changes in the concentration of intracellular Ca2 ([Ca2]c)—play a fundamental role in both neurons and glia as coordinators and integrators of signaling pathways and thus represent an obvious candidate as a mediator or manifestation of pathological process in the CNS. Amyloid peptides have been described as increasing [Ca2]c, decreasing [Ca2]c, altering the dynamics of [Ca2]c signals, having no effect at all. From such a confusing literature, it is almost impossible to extract an intelligible picture from which to draw conclusions. One problem probably arises from the variety of preparations used. A number of groups have studied the effects of A on cell lines and on nonCNS cells such as fibroblasts. Although AD is primarily a disease of the CNS, one of the striking and curious features of the disease is the expression of abnormalities of cell or mitochondrial function in peripheral tissues. The significance of these alterations has never been clear. They do not appear to result in manifest pathology

478648.qxd

1/22/04

6:01 PM

Page 645

Mechanism(s) of A Toxicity in other tissues, but, if A has effects on fibroblasts, this might inform us about some fundamental action of the peptide on cell membranes or aspects of cell function. However, it does not follow clearly that we can extrapolate directly to understand the basis of disease in the CNS. To complicate matters further, it is not clear whether changes in [Ca2]c signaling are a primary or secondary phenomenon. Do these changes represent a cause of the disease or are they rather a manifestation of some other pathological process? Given these limitations, we will focus here on studies that have explored the effects of amyloid peptides on [Ca2]c signals. What, then, is the evidence that A causes alterations in [Ca2]c signaling that might in turn exacerbate if not trigger pathological changes in the disease? [Ca2]c signaling may be affected in many ways— all mechanisms involved in cellular [Ca2]c regulation are potentially subject to modulation. Thus amyloid may interfere with calcium influx pathways, primarily represented in neurons by voltage-gated calcium channels and ligand-gated channels and in astrocytes as capacitative influx channels. It may interfere with the release or storage of calcium from intracellular pools. It may interfere with the clearance mechanisms required to restore [Ca2]c to resting levels after a signal, the Ca2-H ATPase or possibly the Na/Ca2 exchanger, either directly or indirectly by altering cellular metabolic state. The observations published so far make all these mechanisms candidates for alteration by A. A has been shown to alter the activity of voltage-gated Ca2 channels in a variety of cell types. In neuroblastoma cells (111), A seems to increase the amplitude of current through L-type Ca2 channels, and, indeed, the authors showed that cells were protected from A-induced cell death by blockade of L-type channels. The view in this system was clearly that the toxicity of A was mediated by a direct augmentation of Ca2 influx, and [Ca2]c overload, which was then responsible for triggering oxidative stress and cell injury. A study by Rovira et al. (112) also showed that A may increase the Ca2 current in hippocampal neurons in culture, but found that although the effect of A 25–35 was attributable to activation of L-type channels, A1–40 seemed rather to open other channel types, possibly including N-type channels associated with transmitter release, leading the authors to suggest that these might be neurotoxic by enhancing glutamate release and so contribute to excitotoxic cell death. Green and Peers (113) also found that A can increase the amplitude of L-type calcium currents in PC12 cells but they found that A 25–35 and 1–40 seemed to act in the same way. In contrast, Kasparova et al. (114), showed that chronic exposure of

645 neuroblastoma cells to A 1–40 reduced the Ca2 current carried through N-type channels and suggested that this effect might underlie a failure of neurotransmitter release and altered synaptic activity and that this could account in some measure for loss of cognition. These studies suggested that the primary action of A is to alter [Ca2]c signaling. In contrast, in a study of cortical neurons in culture, it was suggested that an A-induced increase in Ca2 was attenuated by antioxidants that were also cytoprotective, suggesting that the rise in [Ca2]c is secondary to the oxidant stress induced by A (115). This group also found that the A-induced Ca2 influx seen in these cells was not attributable to influx through any readily identifiable pathway, suggesting that A induces Ca2 entry through a pore or by some form of alteration in membrane leakiness to Ca2 (see below). Changes in the kinetics and the characteristics of [Ca2]c signals in several cell types have been described; for example, in cortical neurons in culture, metabotropic [Ca2]c signals in response to ACh were attenuated by A and this effect was reversed by antioxidants, again suggesting that the observed change in [Ca2]c signaling is downstream of the primary effector mechanism of A but may contribute to altered information processing or cell death (116). Remarkably, a further study found that chronic depolarization of neurons by exposure to high potassium was neuroprotective (117), and these authors suggested that high [Ca2]c induced by the depolarization might activate enzyme pathways or initiate new protein synthesis that serves to protect the cells. Yet another group found that an A-induced increase in [Ca2]c in response to glutamate was mediated by activation of non-NMDA glutamate receptors (118). We found that A had no apparent effect at all on [Ca2]c in hippocampal neurons over the course of an hour or so of exposure, but that it did cause the appearance of complex [Ca2]c signals in astrocytes attributable to Ca2 influx through an A induced channel (93). Amyloid as Ca2 Channel Former. In the early 1990s, it was shown that amyloid- peptides can form channels when incorporated into artificial planar lipid membranes. These channels were essentially cationic with a selectivity Cs Li Ca2 K, and there is substantial data to suggest that these channels may mediate calcium influx into cells in response to A. The channels show sporadic activity and seem capable of generating a number of different conductance states (119). The peptide apparently inserts into the membranes of lipid vesicles, where it can be visualized through immunofluorescence (120) and where it forms a calcium-permeable channel. The channel is apparently

478648.qxd

1/22/04

6:01 PM

Page 646

646 blocked by zinc, although the action of zinc is confusing, because it can also complex with A, prevent aggregation, and so prevent insertion into the membrane or prevent pore formation rather than blocking the formed channel (see 93, 121, and 122). As in all the experiments described here, the channels are formed by A 25–35 or 1–40, but not by the reverse peptide 35–25, confirming that this is specific to the toxic peptides and is not some nonspecific general disruption of membrane structure by addition of peptides. Lin et al. (123) demonstrated multimeric channel structures in artificial membranes by atomic force microscopy and have shown that A can induce sporadic transient changes in Ca2 in endothelial cells, which is entirely due to Ca2 influx (120). In our recent study of hippocampal cells in culture, we also found that A 25–35 and 1–42 but not 35–25 caused sporadic changes in [Ca2]c in astrocytes that were entirely dependent on external Ca2 (93). Using the quench of the Ca2 indicator, fura-2, by Mn2 influx we were able to show that each Ca2 transient was due to a pulse of Ca2 influx and that the signals were independent of Ca2 release from ER-stored Ca2. Confocal imaging revealed focal points of [Ca2]c elevation in astrocytes, followed by the slower diffusional spread of Ca2 through the cell. The signals were prevented by zinc, but once established, could not be blocked by zinc, suggesting that zinc acts to prevent aggregation rather than as a channel blocking agent. These data provide further evidence to suggest that A may form cationic calciumpermeable channels in cell membranes and mediate changes in [Ca2]c signaling (93). What was perhaps most surprising about this study was the selectivity whereby A caused dramatic changes in Ca2 signals in astrocytes while having negligible effects on [Ca2]c signals in adjacent neurons. Thus, if A acts simply by inserting into lipid bilayers and forming a channel, why should there be a difference in different cells? The answer may lie in observations on the importance of lipid composition in the pore forming activity of the peptide. Recent studies have emphasized the importance of the cholesterol content of lipid membranes for A channel formation. Thus the pore-forming activity of A in bilayers is inversely related to the cholesterol content of the lipid mixture. Similarly, depletion of cholesterol content in cells by treatment with cyclodextrin or inhibition of cholesterol synthesis in PC12 cells increased A toxicity (125). Similarly, Kawahara and Kuroda (126) showed that the [Ca2]c increase in cells was attenuated by pretreament of the cells to increase membrane cholesterol content. We have no data at present about the differences in cho-

Canevari, Abramov, and Duchen lesterol content of different cell types in the CNS, but this seems a simple mechanism that might account for differences in the vulnerability of different cell types to A toxicity. What was also surprising in our own study (93) was that, although the [Ca2]c signals induced by A were entirely confined to the astrocytes, the cells that died later were predominantly the neurons (Fig. 4). The astrocyte [Ca2]c signals were not influenced at all by antioxidants. GSH was, however, depleted in both astrocytes and neurons, and the GSH depletion was calcium dependent, strongly suggesting a sequence of events whereby A forms a channel in the astrocyte membrane, promoting calcium influx into astrocytes as a primary phenomenon. This then generates oxidative stress within the astrocytes, possibly through activation of the NADPH oxidase (see above) causing GSH depletion, as both the rate of increase of ROS generation and the depletion of GSH induced by A was calcium dependent. It seems then that astrocyte GSH export can no longer sustain the requirements of neurons for GSH precursors and so the neurons also become GSH depleted. It is the experience of many investigators that depletion of neuronal GSH may alone be sufficient to cause neuronal death, because these cells seem much more vulnerable to endogenous pro-oxidants than the astrocytes, which appear far more robust. This model sees changes in [Ca2]c as primary and oxidative stress as a consequence, although cell death

Fig. 4. A causes Ca2-dependent cell death in hippocampal neurons and astrocytes. Cell viability was assessed before and after exposure to A (25–35 at 50 M for 24 h with (gray bars) and without (black bars) Ca2, using Hoechst 33342 to stain all nuclei plus propidium iodide (PI) to stain only the nuclei of dead cells, and signals are expressed as the percentage of Hoechst-positive nuclei stained with PI. Again, A caused cell death in both astrocytes and neurons and in both cases; the cell death was dependent on external calcium.

478648.qxd

1/22/04

6:01 PM

Page 647

Mechanism(s) of A Toxicity seems to ensue as a consequence of oxidative stress, not as a direct consequence of “calcium overload.”

CONCLUSION The literature on the toxicity of A is vast and without doubt profoundly confusing. We have here attempted to draw as much as possible together from a literature that is often contradictory. What is apparent are the recurring themes of oxidative stress, disordered mitochondrial function, altered activity of enzymes involved in intermediary metabolism, and altered calcium signaling. Our own data support a model in which A acts as a Ca2 channel pore former in astrocytes in which it also activates the NADPH oxidase. The resultant oxidative stress causes glutathione depletion both in the astrocytes and in adjacent neurons that are highly vulnerable to oxidative stress and eventually die. This suggests that a disturbance in the communication between glial cells and the neurons that they support plays a central role in the pathogenesis of the disease. Perhaps a key difficulty in all these experiments lies in the extent to which one can ever hope to mimic a disease that might take 70 years to develop in a tissue culture dish in an afternoon’s experiment, and new approaches in generating transgenic animal models may ultimately provide the clues that we need so badly if we are to understand this appalling disease.

ACKNOWLEDGMENTS Work in our laboratories is supported by the Wellcome Trust, the Medical Research Council and the Miriam Marks Foundation.

REFERENCES 1. Hardy, J., Duff, K., Hardy, K. G., Perez-Tur, J., and Hutton, M. 1998. Genetic dissection of Alzheimer’s disease and related dementias: Amyloid and its relationship to tau. Nat. Neurosci. 1:355–358. 2. Lemere, C. A., Blusztajn, J. K., Yamaguchi, H., Wisniewski, T., Saido, T. C., and Selkoe, D. J. 1996. Sequence of deposition of heterogeneous amyloid -peptides and APO E in Down syndrome: Implications for initial events in amyloid plaque formation. Neurobiol. Dis. 3:16–32. 3. Chapman, P. F., White, G. L., Jones, M. W., Cooper-Blacketer, D., Marshall, V. J., Irizarry, M., Younkin, L., Good, M. A., Bliss, T. V. P., Hyman, B. T., Younkin, S. G., and Hsiao, K. K. 1999. Impaired synaptic plasticity and learning in aged amyloid precursor protein transgenic mice. Nat. Neurosci. 2:271–276. 4. Phinney, A. L., Calhoun, M. E., Wolfer, D. P., Lipp, H. P., Zheng, H., and Jucker, M. 1999. No hippocampal neuron or synaptic bouton loss in learning-impaired aged beta-amyloid precursor protein-null mice. Neuroscience 90:1207–1216. 5. Snowdon, D. A. 1997. Aging and Alzheimer’s disease: Lessons from the Nun study. Gerontologist 37:150–156.

647 6. Näslund, J., Haroutunian, V., Mohs, R., Davis, K.-L., Davies, P., Greengard, P., and Buxbaum, J.-D. 2000. Correlation between elevated levels of amyloid -peptide in the brain and cognitive decline. JAMA 283:1571–1577. 7. Oddo, S., Caccamo, A., Shepherd, J. D., Murphy, M. P., Golde, T. E., Kayed, R., Metherate, R., Mattson, M. P., Akbari, Y., and LaFerla, F. M. 2003. Triple-transgenic model of Alzheimer’s disease with plaques and tangles: Intracellular Abeta and synaptic dysfunction. Neuron 39:409–421. 8. Nishimoto, I., Okamoto, T., Matsuura, Y., Takahashi, S., Okamoto, T., Murayama, Y., and Ogata, E. 1993. Alzheimer amyloid protein precursor complexes with brain GTP-binding protein Go. Nature 362:75–79. 9. Kamal, A., Stokin, G. B., Yang, Z., Xia, C. H., and Goldstein, L. S. 2000. Axonal transport of amyloid precursor protein is mediated by direct binding to the kinesin light chain subunit of kinesin-I. Neuron 28:449–459. 10. Mattson, M. P. 1997. Cellular actions of -amyloid precursor protein and its soluble and fibrillogenic derivatives. Physiol. Rev. 77:1081–1132. 11. Zheng, H., Jiang, M., Trumbauer, M. E., Sirinathsinghji, D. J., Hopkins, R., Smith, D. W., Heavens, R. P., Dawson, G. R., Boyce, S., and Conner, M. W. 1995. -Amyloid precursor protein-deficient mice show reactive gliosis and decreased locomotor activity. Cell 81:525–531. 12. Kimberly, W. T., LaVoie, M. J., Ostaszewski, B. L., Ye, W., Wolfe, M. S., and Selkoe, D. J. 2003. -Secretase is a membrane protein complex comprised of presenilin, nicastrin, Aph-1, and Pen-2. Proc. Natl. Acad. Sci. USA 100:6382–6387. 13. Wolfe, M. S., De Los Angeles, J., Miller, D. D., Xia, W., and Selkoe, D. J. 1999. Are presenilins intramembrane-cleaving proteases? Implications for the molecular mechanism of Alzheimer’s disease. Biochemistry 38:11223–11230. 14. Holsinger, R. M., McLean, C. A., Beyreuther, K., Masters, C. L., and Evin, G. 2002. Increased expression of the amyloid precursor -secretase in Alzheimer’s disease. Ann. Neurol. 51:783–786. 15. Roberds, S. L., Anderson, J., Basi, G., Bienkowski, M. J., Branstetter, D. G., Chen, K. S., Freedman, S. B., Frigon, N. L., Games, D., Hu, K., Johnson-Wood, K., Kappenman, K. E., Kawabe, T. T., Kola, I., Kuehn, R., Lee, M., Liu, W., Motter, R., Nichols, N. F., Power, M., Robertson, D. W., Schenk, D., Schoor, M., Shopp, G. M., Shuck, M. E., Sinha, S., Svensson, K. A., Tatsuno, G., Tintrup, H., Wijsman, J., Wright, S., and McConlogue, L. 2001. BACE knockout mice are healthy despite lacking the primary -secretase activity in brain: Implications for Alzheimer’s disease therapeutics. Hum. Mol. Genet. 10:1317–1324. 16. Tanzi, R. E., Kovacs, D. M., Kim, T. W., Moir, R. D., Guenette, S. Y., and Wasco, W. 1996. The gene defects responsible for familial Alzheimer’s disease. Neurobiol. Dis. 3:159–168. 17. Cataldo, A. M., Peterhoff, C. M., Troncoso, J. C., Gomez-Isla, T., Hyman, B. T., and Nixon, R. A. 2000. Endocytic pathway abnormalities precede amyloid deposition in sporadic Alzheimer’s disease and Down syndrome: Differential effects of APOE genotype and presenilin mutations. Am. J. Pathol. 157:277–286. 18. Strittmatter, W. J., Saunders, A. M., Schmechel, D., Pericak-Vance, M., Enghild, J., Salvesen, G. S., and Roses, A. D. 1993. Apolipoprotein E: High-avidity binding to -amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc. Natl. Acad. Sci. USA 90:1977–1981. 19. Schmechel, D. E., Saunders, A. M., Strittmatter, W. J., Crain, B. J., Hulette, C. M., Joo, S. H., Pericak-Vance, M. A., Goldgaber, D., and Roses, A. D. 1993. Increased amyloid -peptide deposition in cerebral cortex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc. Natl. Acad. Sci. USA 90: 9649–9653. 20. Yang, D. S., Smith, J. D., Zhou, Z., Gandy, S. E., and Martins, R. N. 1997. Characterization of the binding of amyloid- peptide to cell culture-derived native apolipoprotein E2, E3, and E4 isoforms and to isoforms from human plasma. J. Neurochem. 68:721–725.

478648.qxd

1/22/04

6:01 PM

Page 648

648 21. Puttfarcken, P. S., Manelli, A. M., Falduto, M. T., Getz, G. S., and LaDu, M. J. 1997. Effect of apolipoprotein E on neurite outgrowth and -amyloid-induced toxicity in developing rat primary hippocampal cultures. J. Neurochem. 68:760–769. 22. Pike, C. J., Burdick, D., Walencewicz, A. J., Glabe, C. G., and Cotman, C. W. 1993. Neurodegeneration induced by -amyloid peptides in vitro: The role of peptide assembly state. J. Neurosci. 13:1676–1687. 23. Geula, C., Wu, C. K., Saroff, D., Lorenzo, A., Yuan, M., and Yankner, B. A. 1998. Aging renders the brain vulnerable to amyloid -protein neurotoxicity. Nat. Med. 4:827–831. 24. Snyder, S. W., Ladror, U. S., Wade, W. S., Wang, G. T., Barrett, L. W., Matayoshi, E. D., Huffaker, H. J., Krafft, G. A., and Holzman, T. F. 1994. Amyloid- aggregation: selective inhibition of aggregation in mixtures of amyloid with different chain lengths. Biophys. J. 67:1216–1228. 25. Lambert, M. P., Barlow, A. K., Chromy, B. A., Edwards, C., Freed, R., Liosatos, M., Morgan, T. F., Rozovsky, I., Trommer, B., Viola, K. L., Wals, P., Zhang, C., Finch, C. E., Krafft, G. A., and Klein, W. L. 1998. Diffusible, nonfibrillar ligands derived from A 1–42 are potent central nervous system neurotoxins. Proc. Natl. Acad. Sci. USA 95:6448–6453. 26. Walsh, D. M., Hartley, D. M., Kusumoto, Y., Fezoui, Y., Condron, M. M., Lomakin, A., Benedek, G. B., Selkoe, D. J., and Teplow, D. B. 1999. Amyloid -protein fibrillogenesis: Structure and biological activity of protofibrillar intermediates. J. Biol. Chem. 274:25945–25952. 27. Klein, W. L., Krafft, G. A., and Finch, C. E. 2001. Targeting small A oligomers: The solution to an Alzheimer’s disease conundrum? Trends Neurosci. 24:219–224. 28. Burdick, D., Soreghan, B., Kwon, M., Kosmoski, J., Knauer, M., Henschen, A., Yates, J., Cotman, C., and Glabe, C. 1992. Assembly and aggregation properties of synthetic Alzheimer’s A4/ amyloid peptide analogs. J. Biol. Chem. 267:546–554. 29. Mantyh, P. W., Ghilardi, J. R., Rogers, S., DeMaster, E., Allen, C. J., Stimson, E. R., and Maggio, J. E. 1993. Aluminum, iron, and zinc ions promote aggregation of physiological concentrations of amyloid peptide. J. Neurochem. 61:1171–1174. 30. Klunk, W. E., Xu, C. J., McClure, R. J., Panchalingam, K., Stanley, J. A., and Pettegrew, J. W. 1997. Aggregation of -amyloid peptide is promoted by membrane phospholipid metabolites elevated in Alzheimer’s disease brain. J. Neurochem. 69: 266–272. 31. Dyrks, T., Dyrks, E., Hartmann, T., Masters, C., and Beyreuther, K. 1992. Amyloidogenicity of A4 and A4-bearing amyloid protein precursor fragments by metal-catalyzed oxidation J. Biol. Chem. 267:18210–18217. 32. Walsh, D. M., Tseng, B. P., Rydel, R. E., Podlisny, M. B., and Selkoe, D. J. 2000. The oligomerization of amyloid -protein begins intracellularly in cells derived from human brain. Biochemistry 39:10831–10839. 33. D’Andrea, M. R., Nagele, R. G., Wang, H. Y., Peterson, P. A., and Lee, D. H. 2001. Evidence that neurones accumulating amyloid can undergo lysis to form amyloid plaques in Alzheimer’s disease. Histopathology 38:120–134. 34. Meyer-Luehmann, M., Stalder, M., Herzig, M. C., Kaeser, S. A., Kohler, E., Pfeifer, M. D., Boncristiano, S., Mathews, P. M., Mercken, M., Abramowski, D., Staufenbiel, M., and Jucker, M. 2003. Extracellular amyloid formation and associated pathology in neural grafts. Nat. Neurosci. 6:370–377. 35. Grundke-Iqbal, I., Iqbal, K., George, L., Tung, Y. C., Kim, K. S., and Wisniewski, H. M. 1989. Amyloid protein and neurofibrillary tangles coexist in the same neuron in Alzheimer disease. Proc. Natl. Acad. Sci. USA 86:2853–2857. 36. Ginsberg, S. D., Crino, P. B., Hemby, S. E., Weingarten, J. A., Lee, V. M., Eberwine, J., and Trojanowski, J. Q. 1999. Predominance of neuronal mRNAs in individual Alzheimer’s disease senile plaques. Ann. Neurol. 45:174–181. 37. Li, Q. X., Maynard, C., Cappai, R., McLean, C. A., Cherny, R. A., Lynch, T., Culvenor, J. G., Trevaskis, J., Tanner, J. E., Bailey, K. A.,


38.

39.

40.

41.

42.

43.

44.

45.

46. 47.

48.

49. 50. 51.

52. 53.

Czech, C., Bush, A., Beyreuther, K., and Masters, C. L. 1999. Intracellular accumulation of detergent-soluble amyloidogenic A fragment of Alzheimer’s disease precursor protein in the hippocampus of aged transgenic mice. J. Neurochem. 72:2479–2487. Gouras, G. K., Tsai, J., Näslund, J., Vincent, B., Edgar, M., Checler, F., Greenfield, J. P., Haroutunian, V., Buxbaum, J. D., Xu, H., Greengard, P., and Relkin, N. R. 2000. Intraneuronal A42 accumulation in human brain. Am. J. Pathol. 156:15–20. Takahashi, R. H., Milner, T. A., Li, F., Nam, E. E., Edgar, M. A., Yamaguchi, H., Beal, M. F., Xu, H., Greengard, P., and Gouras, G. K. 2002. Intraneuronal Alzheimer A42 accumulates in multivesicular bodies and is associated with synaptic pathology. Am. J. Pathol. 161:1869–1879. Oddo, S., Caccamo, A., Shepherd, J. D., Murphy, M. P., Golde, T. E., Kayed, Metherate, R., Mattson, M. P., Akbari, Y., and LaFerla, F. M. 2003. Triple-transgenic model of Alzheimer’s disease with plaques and tangles: Intracellular A and synaptic dysfunction. Neuron 39:409–421. Maruyama, K., Tagawa, K., Kawamura, Y., Asada, H., Ishiura, S., and Obata, K. 1995. Secretion of Alzheimer A4 protein (1–40) and intracellular retention of A4 protein (1–42) in transfected COS cells. Biochem. Biophys. Res. Comm. 207:971–917. Turner, R. S., Suzuki, N., Chyung, A. S., Younkin, S. G., and Lee, V. M. 1996. Amyloids 40 and 42 are generated intracellularly in cultured human neurons and their secretion increases with maturation. J. Biol. Chem. 271:8966–8970. Wild-Bode, C., Yamazaki, T., Capell, A., Leimer, U., Steiner, H., Ihara, Y., and Haass, C. 1997. Intracellular generation and accumulation of amyloid -peptide terminating at amino acid 42. J. Biol. Chem. 272:16085–16088. Tienari, P. J., Ida, N., Ikonen, E., Simons, M., Weidemann, A., Multhaup, G., Masters, C. L., Dotti, C. G., and Beyreuther, K. 1997. Intracellular and secreted Alzheimer -amyloid species are generated by distinct mechanisms in cultured hippocampal neurons. Proc. Natl. Acad. Sci. USA 94:4125–4130. Greenfield, J. P., Tsai, J., Gouras, G. K., Hai, B., Thinakaran, G., Checler, F., Sisodia, S. S., Greengard, P., and Xu, H. 1999. Endoplasmic reticulum and trans-Golgi network generate distinct populations of Alzheimer -amyloid peptides. Proc. Natl. Acad. Sci. USA 96:742–747. Wilson, C. A., Doms, R. W., and Lee, V. M. 1999. Intracellular APP processing and A production in Alzheimer disease. J. Neuropathol. Exp. Neurol. 58:787–794. Lucassen, P. J., Chung, W. C., Kamphorst, W., and Swaab, D. F. 1997. DNA damage distribution in the human brain as shown by in situ end labeling: Area-specific differences in aging and Alzheimer disease in the absence of apoptotic morphology. J. Neuropathol. Exp. Neurol. 56:887–900. Guo, Q., Fu, W., Sopher, B. L., Miller, M. W., Ware, C. B., Martin, G. M., and Mattson, M. P. 1999. Increased vulnerability of hippocampal neurons to excitotoxic necrosis in presenilin-1 mutant knock-in mice. Nat. Med. 5:101–106. Behl, C., Davis, J. B., Klier, F. G., and Schubert, D. 1994. Amyloid peptide induces necrosis rather than apoptosis. Brain Res. 645:253–264. Su, J. H., Anderson, A. J., Cummings, B. J., and Cotman, C. W. 1994. Immunohistochemical evidence for apoptosis in Alzheimer’s disease. Neuroreport 5:2529–2533. Kitamura, Y., Shimohama, S., Kamoshima, W., Ota, T., Matsuoka, Y., Nomura, Y., Smith, M. A., Perry, G., Whitehouse, P. J., and Taniguchi, T. 1998. Alteration of proteins regulating apoptosis, Bcl-2, Bcl-x, Bax, Bak, Bad, ICH-1 and CPP32, in Alzheimer’s disease. Brain Res. 780:260–269. Engidawork, E., Gulesserian, T., Seidl, R., Cairns, N., and Lubec, G. 2001. Expression of apoptosis related proteins in brains of patients with Alzheimer’s disease. Neurosci. Lett. 303:79–82. Yang, F., Sun, X., Beech, W., Teter, B., Wu, S., Sigel, J., Vinters, H. V., Frautschy, S. A., and Cole, G. M. 1998. Antibody to caspase-cleaved actin detects apoptosis in differentiated neurob-

478648.qxd

1/22/04

6:01 PM

Page 649

Mechanism(s) of A Toxicity

54.

55.

56.

57. 58.

59.

60.

61.

62. 63.

64.

65.

66.

67. 68.

69.

lastoma and plaque-associated neurons and microglia in Alzheimer’s disease. Am. J. Pathol. 152:379–389. Estus, S., Tucker, H. M., van Rooyen, C., Wright, S., Brigham, E. F., Wogulis, M., and Rydel, R. E. 1997. Aggregated amyloid- protein induces cortical neuronal apoptosis and concomitant “apoptotic” pattern of gene induction. J. Neurosci. 17:7736–7745. Kienlen-Campard, P., Miolet, S., Tasiaux, B., and Octave, J. N. 2002. Intracellular amyloid- 1–42, but not extracellular soluble amyloid- peptides, induces neuronal apoptosis. J. Biol. Chem. 277:15666–15670. Tamagno, E., Parola, M., Guglielmotto, M., Santoro, G., Bardini, P., Marra, L., Tabaton, M., and Danni, Q. 2003. Multiple signaling events in amyloid -induced, oxidative stress-dependent neuronal apoptosis. Free Rad. Biol. Med. 35:45–58. Roth, K. A. 2001. Caspases, apoptosis, and Alzheimer disease: Causation, correlation, and confusion. J. Neuropathol. Exp. Neurol. 60:829–838. Kim., T. W., Pettingell, W. H., Jung, Y. K., Kovacs, D. M., and Tanzi, R. E. 1997. Alternative cleavage of Alzheimer-associated presenilins during apoptosis by a caspase-3 family protease. Science 277:373–376. Gervais, F. G., Xu, D., Robertson, G. S., Vaillancourt, J. P., Zhu, Y., Huang, J., LeBlanc, A., Smith, D., Rigby, M., Shearman, M. S., Clarke, E. E., Zheng, H., Van Der Ploeg, L. H., Ruffolo, S. C., Thornberry, N. A., Xanthoudakis, S., Zamboni, R. J., Roy, S., and Nicholson, D. W. 1999. Involvement of caspases in proteolytic cleavage of Alzheimer’s amyloid- precursor protein and amyloidogenic A peptide formation. Cell 97:395–406. Regeur, L., Jensen, G. B., Pakkenberg, H., Evans, S. M., and Pakkenberg, B. 1994. No global neocortical nerve cell loss in brains from patients with senile dementia of Alzheimer’s type. Neurobiol. Aging 15:347–352. Ibanez. V., Pietrini, P., Alexander, G. E., Furey, M. L., Teichberg, D., Rajapakse, J. C., Rapoport, S. I., Schapiro, M. B., and Horwitz, B. 1998. Regional glucose metabolic abnormalities are not the result of atrophy in Alzheimer’s disease. Neurology 50:1585–1593. de la Monte, S. M., Luong, T., Neely, T. R., Robinson, D., and Wands, J. R. 2000. Mitochondrial DNA damage as a mechanism of cell loss in Alzheimer’s disease. Lab. Invest. 80:1323–1335. Hirai, K., Aliev, G., Nunomura, A., Fujioka, H., Russell, R. L., Atwood, C. S., Johnson, A. B., Kress, Y., Vinters, H. V., Tabaton, M., Shimohama, S., Cash, A. D., Siedlak, S. L., Harris, P. L., Jones, P. K., Peterson, R. B., Perry, G., and Smith, M. A. 2001. Mitochondrial abnormalities in Alzheimer’s disease. J. Neurosci. 21:3017–3023. Anandatheerthavarada, H. K., Biswas, G., Robin, M. A., and Avadhani, N. G. 2003. Mitochondrial targeting and a novel transmembrane arrest of Alzheimer’s amyloid precursor protein impairs mitochondrial function in neuronal cells. J. Cell Biol. 161:41–54. Khan, S. M., Cassarino, D. S., Abramova, N. N., Keeney, P. M., Borland, M. K., Trimmer, P. A., Krebs, C. T., Bennett, J. C., Parks, J. K., Swerdlow, R. H., Parker, W. D. Jr., and Bennett, J. P. Jr. 2000. Alzheimer’s disease cybrids replicate -amyloid abnormalities through cell death pathways. Ann. Neurol. 48:148–155. Trimmer, P. A., Swerdlow, R. H., Parks, J. K., Keeney, P., Bennett, J. P. Jr., Miller, S. W., Davis, R. E., and Parker, W. D., Jr. 2000. Abnormal mitochondrial morphology in sporadic Parkinson’s and Alzheimer’s disease cybrid cell lines. Exp. Neurol. 162:37–50. Lin, M. T., Simon, D. K., Ahn, C. H., Kim, L. M., and Beal, M. F. 2002. High aggregate burden of somatic mtDNA point mutations in aging and Alzheimer’s disease brain. Hum. Mol. Genet. 11:133–145. Gibson, G. E., Zhang, H., Sheu, K. F., Bogdanovich, N., Lindsay, J. G., Lannfelt, L., Vestling, M., and Cowburn, R. F. 1998. -Ketoglutarate dehydrogenase in Alzheimer brains bearing the APP670/671 mutation. Ann. Neurol. 44:676–681. Sheu, K. F., Kim, Y. T., Blass, J. P., and Weksler, M. E. 1985. An immunochemical study of the pyruvate dehydrogenase deficit in Alzheimer’s disease brain. Ann. Neurol. 17:444–449.

649 70. Kish, S. J., Bergeron, C., Rajput, A., Dozic, S., Mastrogiacomo, F., Chang, L. J., Wilson, J. M., DiStefano, L. M., and Nobrega, J. N. 1992. Brain cytochrome oxidase in Alzheimer’s disease. J. Neurochem. 59:776–779. 71. Mutisya, E. M., Bowling, A. C., and Beal, M. F. 1994. Cortical cytochrome oxidase activity is reduced in Alzheimer’s disease. J. Neurochem. 63:2179–2184. 72. Chandrasekaran, K., Hatanpaa, K., Brady, D. R., and Rapoport, S. I. 1996. Evidence for physiological down-regulation of brain oxidative phosphorylation in Alzheimer’s disease. Exp. Neurol. 142:80–88. 73. Bedetti, C. D. 1985. Immunocytochemical demonstration of cytochrome c oxidase with an immunoperoxidase method: A specific stain for mitochondria in formalin-fixed and paraffin-embedded human tissues. J. Histochem. Cytochem. 33:446–452. 74. Parker, W. D. Jr., Mahr, N. J., Filley, C. M., Parks, J. K., Hughes, D., Young, D. A., and Cullum, C. M. 1994. Reduced platelet cytochrome c oxidase activity in Alzheimer’s disease. Neurology. 44:1086–1090. 75. Askanas, V., McFerrin, J., Baqué, S., Alvarez, R. B., Sarkozi, E., and Engel, W. K. 1996. Transfer of -amyloid precursor protein gene using adenovirus vector causes mitochondrial abnormalities in cultured normal human muscle. Proc. Natl. Acad. Sci. USA 93:1314–1319. 76. Grant, S. M., Shankar, S. L., Chalmers-Redman, R. M., Tatton, W. G., Szyf, M., and Cuello, A. C. 1999. Mitochondrial abnormalities in neuroectodermal cells stably expressing human amyloid precursor protein (hAPP751). Neuroreport 10:41–46. 77. Pereira, C., Santos, M. S., and Oliveira, C. 1998. Mitochondrial function impairment induced by amyloid -peptide on PC12 cells. Neuroreport 9:1749–1755. 78. Canevari, L., Clark, J. B., and Bates, T. E. 1999. -Amyloid fragment 25–35 selectively decreases complex IV activity in isolated mitochondria. FEBS Lett. 457:131–134. 79. Casley, C. S., Canevari, L., Land, J. M., Clark, J. B., and Sharpe, M. A. 2002. -Amyloid inhibits integrated mitochondrial respiration and key enzyme activities. J. Neurochem. 80:91–100. 80. Cardoso, S. M., Santos, S., Swerdlow, R. H., and Oliveira, C. R. 2001. Functional mitochondria are required for amyloid -mediated neurotoxicity. FASEB J. 15:1439–1441. 81. Shevtzova, E. F., Kireeva, E. G., and Bachurin, S. O. 2001. Effect of -amyloid peptide fragment 25-35 on nonselective permeability of mitochondria. Bull. Exp. Biol. Med. 132:1173–1176. 82. Cadenas, E., Boveris, A., Ragan, C. I., and Stoppani, A. O. M. 1977. Production of superoxide radicals and hydrogen peroxide by NADH-ubiquinone reductase and ubiquinol-cytochrome c reductase from beef-heart mitochondria. Arch. Biochem. Biophys. 180: 248–257. 83. Gabuzda, D., Busciglio, J., Chen, L. B., Matsudaira, P., and Yankner, B. A. 1994. Inhibition of energy metabolism alters the processing of amyloid precursor protein and induces a potentially amyloidogenic derivative. J. Biol. Chem. 269: 13623–13628. 84. Smith, M. A., Richey Harris, P. L., Sayre, L. M., Beckman, J. S., and Perry, G. 1997. Widespread peroxynitrite-mediated damage in Alzheimer’s disease. J. Neurosci. 17:2653–2657. 85. Varadarajan, S., Yatin, S., Aksenova, M., and Butterfield, D. A. 2000. Alzheimer’s amyloid -peptide-associated free radical oxidative stress and neurotoxicity. J. Struct. Biol. 30:184–208. 86. Kish, S. J. 1997. Brain energy metabolizing enzymes in Alzheimer’s disease: -Ketoglutarate dehydrogenase complex and cytochrome oxidase: Ann. NY Acad. Sci., 826:218–228. 87. Gibson, G. E., Park, L. C., Sheu, K. F., Blass, J. P., and Calingasan, N. Y. 2000. The -ketoglutarate dehydrogenase complex in neurodegeneration. Neurochem. Int. 36:97–112. 88. Tretter, L., and Adam-Vizi, V. 2000. Inhibition of Krebs cycle enzymes by hydrogen peroxide: A key role of -ketoglutarate dehydrogenase in limiting NADH production under oxidative stress. J. Neurosci. 20:8972–8979.

478648.qxd

1/22/04

6:01 PM

Page 650

650 89. Harris, M. E., Wang, Y., Pedigo, N. W. Jr., Hensley, K., Butterfield, D. A., and Carney, J. M. 1996. Amyloid peptide (25–35) inhibits Na-dependent glutamate uptake in rat hippocampal astrocyte cultures. J Neurochem. 67:277–286. 90. Abe, K. and Misawa, M. 2003. Amyloid protein enhances the clearance of extracellular L-glutamate by cultured rat cortical astrocytes. Neurosci. Res. 45:25–31. 91. Cecchi, C., Latorraca, S., Sorbi, S., lantomasi, T., Favilli, F., Vincenzini, M. T., and Liguri, G. 1999. Gluthatione level is altered in lymphoblasts from patients with familial Alzheimer’s disease. Neurosci. Lett. 275:152–154. 92. Muller, W. E., Romero, F. J., Perovic, S., Pergande, G., and Pialoglou, P. 1997. Protection of flupirtine on -amyloidinduced apoptosis in neuronal cells in vitro: Prevention of amyloid-induced glutathione depletion. J. Neurochem. 68:71–77. 93. Abramov, A. Y., Canevari, L., and Duchen, M. R. 2003. Changes in intracellular calcium and glutathione in astrocytes as the primary mechanism of amyloid neurotoxicity. J. Neurosci. 23:5088–5095. 94. Keelan, J., Allen, N. J., Antcliffe, D., Pal, S., and Duchen, M. R. 2001. Quantitative imaging of glutathione in hippocampal neurons and glia in culture using monochlorobimane. J. Neurosci. Res. 66:873–884. 95. Dringen, R. and Hirrlinger, J. 2003. Glutathione pathways in the brain. Biol. Chem. 384:505–516. 96. Barkats, M., Millecamps, S., Abrioux, P., Geoffroy, M. C., and Mallet, J. 2000. Overexpression of glutathione peroxidase increases the resistance of neuronal cells to A-mediated neurotoxicity. J. Neurochem. 75:1438–1446. 97. Benzi, G. and Moretti, A. 1995. Are reactive oxygen species involved in Alzheimer’s disease? Neurobiol. Aging 16:661–674. 98. Atwood, C. S., Moir, R. D., Huang, X., Scarpa, R. C., Bacarra, N. M., Romano, D. M., Hartshorn, M. A., Tanzi, R. E., and Bush, A. I. 1998. Dramatic aggregation of Alzheimer A by Cu(II) is induced by conditions representing physiological acidosis. J. Biol. Chem. 273:12817–12826. 99. Cuajungco, M. P., Goldstein, L. E., Nunomura, A., Smith, M. A., Lim, J. T., Atwood, C. S., Huang, X., Farrag, Y. W., Perry, G., and Bush, A. I. 2000. Evidence that the -amyloid plaques of Alzheimer’s disease represent the redox-silencing and entombment of A by zinc. J. Biol. Chem. 275:19439–19442. 100. Cherny, R. A., Atwood, C. S., Xilinas, M. E., Gray, D. N., Jones, W. D., McLean, C. A., Barnham, K. J., Volitakis, I., Fraser, F. W., Kim, Y., Huang, X., Goldstein, L. E., Moir, R. D., Lim, J. I., Beyreuther, K., Zheng, H., Tanzi, R. E., Masters, C. L., and Bush, A. I. 2001. Treatment with a copper-zinc chelator markedly and rapidly inhibits -amyloid accumulation in Alzheimer’s disease transgenic mice. Neuron 30:665–676. 101. Melov, S. 2002. ‘ . . . and C is for Clioquinol’: The ACs of Alzheimer’s disease. Trends Neurosci. 25:121–123. 102. Sheehan, J. P., Swerdlow, R. H., Miller, S. W., Davis, R. E., Parks, J. K., Parker, W. D., and Tuttle, J. B. 1997. Calcium homeostasis and reactive oxygen species production in cells transformed by mitochondria from individuals with sporadic Alzheimer’s disease. J. Neurosci. 17:4612–4622. 103. Meda, L., Bonaiuto, C., Baron, P., Otvos, L. Jr., Rossi, F., and Cassatella, M. A. 1996. Priming of monocyte respiratory burst by -amyloid fragment (25–35). Neurosci. Lett. 219:91–94. 104. Parvathenani, L. K., Tertyshnikova, S., Greco, C. R., Roberts, S. B., Robertson, B., and Posmantur, R. 2003. P2X7 mediates superoxide production in primary microglia and is up-regulated in a transgenic mouse model of Alzheimer’s disease. J. Biol. Chem. 278:13309–13317. 105. Husemann, J. and Silverstein, S. C. 2001. Expression of scavenger receptor class B, type I, by astrocytes and vascular smooth muscle cells in normal adult mouse and human brain and in Alzheimer’s disease brain. Am. J. Pathol. 158:825–832. 106. El Khoury, J. B., Moore, K. J., Means, T. K., Leung, J., Terada, K., Toft, M., Freeman, M. W., and Luster, A. D. 2003. CD36 mediates the innate host response to -amyloid. J. Exp. Med. 197:1657–1666.

Canevari, Abramov, and Duchen 107. Wautier, M. P., Chappey, O., Corda, S., Stern, D. M., Schmidt, A. M., and Wautier, J. L. 2001. Activation of NADPH oxidase by AGE links oxidant stress to altered gene expression via RAGE. Am. J. Physiol. Endocrinol. Metab. 280:E685–694. 108. Yan, S. D., Chen, X., Fu, J., Chen, M., Zhu, H., Roher, A., Slattery, T., Zhao, L., Nagashima, M., Morser, J., Migheli, A., Nawroth, P., Stern, D., and Schmidt, A. M. 1996. RAGE and amyloid- peptide neurotoxicity in Alzheimer’s disease. Nature 382:685–691. 109. Lee, M., You, H. J., Cho, S. H., Woo, C. H., Yoo, M. H., Joe, E. H., and Kim, J. H. 2002. Implication of the small GTPase Rac1 in the generation of reactive oxygen species in response to -amyloid in C6 astroglioma cells. Biochem. J. 366:937–943. 110. Pocock, J. M. and Liddle, A. C. 2001. Microglial signalling cascades in neurodegenerative disease. Prog. Brain Res. 132:555–565. 111. Ho, R., Ortiz, D., and Shea, T. B. 2001. Amyloid- promotes calcium influx and neurodegeneration via stimulation of L voltagesensitive calcium channels rather than NMDA channels in cultured neurons. J. Alzheimers Dis. 3:479–483. 112. Rovira, C., Arbez, N., and Mariani, J. 2002. A(25–35) and A(1–40) act on different calcium channels in CA1 hippocampal neurons. Biochem. Biophys. Res. Commun. 296:1317–1321. 113. Green, K. N. and Peers, C. 2001. Amyloid peptides mediate hypoxic augmentation of Ca2 channels. J. Neurochem. 77: 953–956. 114. Kasparova, J., Lisa, V., Tuˇcek, S., and Dolezal, V. 2001. Chronic exposure of NG108-15 cells to amyloid peptide (A1–42) abolishes calcium influx via N-type calcium channels. Neurochem. Res. 26:1079–1084. 115. Huang, H. M., Ou, H. C., and Hsieh, S. J. 2000. Antioxidants prevent amyloid peptide-induced apoptosis and alteration of calcium homeostasis in cultured cortical neurons. Life Sci. 66:1879–1892. 116. Kelly, J. F., Furukawa, K., Barger, S. W., Rengen, M. R., Mark, R. J., Blanc, E. M., Roth, G. S., and Mattson, M. P. 1996. Amyloid -peptide disrupts carbachol-induced muscarinic cholinergic signal transduction in cortical neurons. Proc. Natl. Acad. Sci. USA 93:6753–6758. 117. Pike, C. J., Balazs, R., and Cotman, C. W. 1996. Attenuation of -amyloid neurotoxicity in vitro by potassium-induced depolarization. J. Neurochem. 67:1774–1777. 118. Blanchard, B. J., Konopka, G., Russell, M., and Ingram, V. M. 1997. Mechanism and prevention of neurotoxicity caused by -amyloid peptides: Relation to Alzheimer’s disease. Brain Res. 776:40–50. 119. Arispe, N., Pollard, H. B., and Rojas, E. 1993. Giant multilevel cation channels formed by Alzheimer disease amyloid -protein [AP-(1–40)] in bilayer membranes. Proc. Natl. Acad. Sci. USA 90:10573–10577. 120. Rhee, S. K., Quist, A. P., and Lal, R. 1998. Amyloid protein(1–42) forms calcium-permeable, Zn2-sensitive channel. J. Biol. Chem. 273:13379–13382. 121. Arispe, N., Pollard, H. B., and Rojas, E. 1996. Zn2 interaction with Alzheimer amyloid protein calcium channels. Proc. Natl. Acad. Sci. USA 93:1710–1715. 122. Bush, A. J. 2003. The metallobiology of Alzheimer’s disease. Trends Neurosci. 26:207–214. 123. Lin, H., Bhatia, R., and Lal, R. 2001. Amyloid protein forms ion channels: Implications for Alzheimer’s disease pathophysiology. FASEB J. 15:2433–2444. 124. Bhatia, R., Lin, H., and Lal, R. 2000. Fresh and globular amyloid protein (1–42) induces rapid cellular degeneration: Evidence for AP channel-mediated cellular toxicity. FASEB J. 14:1233–1243. 125. Arispe, N. and Doh, M. 2002. Plasma membrane cholesterol controls the cytotoxicity of Alzheimer’s disease AP (1–40) and (1–42) peptides FASEB J. 16:1526–1536. 126. Kawahara, M. and Kuroda, Y. 2001. Intracellular calcium changes in neuronal cells induced by Alzheimer’s -amyloid protein are blocked by estradiol and cholesterol. Cell. Mol. Neurobiol. 21:1–13.