Oxidative Mechanisms in β-Amyloid Cytotoxicity

NEURODEGENERATION, Vol. 5, pp 441–444 (1996)

Oxidative Mechanisms in β-Amyloid Cytotoxicity John B. Davis Department of Molecular Neuropathology, SmithKline Beecham Pharmaceuticals, Third Avenue, Harlow, CM19 5AW Amyloid β-peptide has been demonstrated to be toxic for primary and clonal neuronal cell lines in vitro. Oxidative mechanisms have been implicated in this pathway at several points, including the aggregation of β-amyloid necessary for cytotoxic activity, generation of radicals by the peptide itself, and intracellularly in response to toxic β-amyloid peptides. Supporting an oxidative hypothesis are the observations that cells mount a stress response to β-amyloid similar to that seen in response to oxidative stress and that they may be rescued from cytotoxicity by antioxidants, inhibitors of oxidative enzyme metabolism, and overexpression of antioxidant enzymes. Although the source(s) of the oxygen radicals has not yet been identified, altered antioxidant enzyme levels and oxidative by-products in Alzheimer’s disease brain samples relate the in vitro studies to the human disease. © 1996 Academic Press Limited

Key words: beta-amyloid, hydrogen peroxide, neurotoxicity, oxidation

Introduction

The cellular response to βA4

THE AMYLOID OR SENILE plaques are the neuropathological hallmark of Alzheimer’s disease and, following the extraction and sequencing of beta-amyloid (βΑ4 ) from these plaques (Masters et al., 1985) and the linkage of familial Alzheimer’s disease loci to the precursor gene from which βA4 is processed (Goate et al., 1991), much effort has focused on the biological properties of βA4. These reports and the neuronal cell loss evident in Alzheimer’s disease (Terry et al., 1994), has led to the development of much interest in the cytotoxic properties of βA4. This review discusses the evidence supporting the hypothesis that the cytotoxic properties of βA4 are mediated via a pathway involving reactive oxygen species (ROS). A large body of additional research supports this field but is outside the scope of this review; for example, the detailed analysis of the structure and biophysics of βA4 in relation to itself and chaperone molecules is highly relevant, on account of the oligomerization of βA4 necessary for its cytotoxic properties. For a recent and excellent review that includes these broader elements see Iversen et al. (1995).

Study of the mechanisms of βA4 induced neurotoxicity were initiated in 1989 by the work of Yankner and colleagues who showed that βA4 1–40, and fragments 25–35 and 1–28, are toxic for primary hippocampal neurons (Yankner et al., 1990). The direct toxicity of βA4 peptides was confirmed using clonal neuronal cell lines (Behl et al., 1992). Several hypotheses for mechanisms of βA4 toxicity have been proposed, including a potentiation of disturbances to ion homeostasis (Mattson et al., 1992), ion pore formation (Arispe et al., 1993), and blockade of a K1 channel (Etcheberrigary et al., 1994). Also reported is a βA4 effect upon Na/K ATPase activity (Mark et al., 1995). This latter effect is thought to be mediated via a ROS intermediate and it is possible that many other ion homeostasis destabilising effects of βA4 are similarly mediated by ROS dependent redox changes or signal transductions. The first indication that a pathway involving ROS might be involved in βA4 cytotoxicity was the observation that vitamin E and other lipophilic antioxidants rescue neuronal cell lines from βA4-induced cytotoxicity (Behl et al., 1992). A build up of ROS is detectable in cells exposed to βA4 and antioxidants that prevent the intracellular increase in ROS rescue the cells from cell death (Behl et al., 1994). These pharmacological observations supporting a ROS-mediated mechanism

Correspondence to: John B. Davis, MNR (H26), New Frontiers Science Park North. SmithKline Beecham, Third Avenue, Harlow CM19 5AW. e-mail: [email protected] © 1996 Academic Press Limited 1055-8330/96/040441 1 4 $25.00/0

441

442

J.B. Davis

have been replicated in primary cortical neurones (Behl et al., 1994) and hippocampal cultures (Mattson & Goodman 1995). The view that ROS are involved in βA4 cytotoxicity is strengthened by the observation that cells resistant to βA4 toxicity have elevated levels of the antioxidant enzymes catalase and glutathione peroxidase (Sagara et al., 1996). These adaptations also confer resistance to toxicity by exogenous hydrogen peroxide and by other compounds that are toxic by virtue of oxidative mechanisms. Also, transfection of catalase or glutathione peroxidase into PC12 cells rescues them from the cytotoxic activity of βA4 (Sagara et al., 1996), suggesting hydrogen peroxide as a key ROS intermediate.

The origin of ROS in βA4 cytotoxicity One of the earliest apparent cellular responses to βA4 exposure is an alteration in the cellular reduction of MTT (Shearman et al., 1994) and, consequently, some attention has focused upon the redox enzymes capable of reducing MTT as a possible source of the ROS induced by βA4. Hydrogen peroxide causes a similar decrease in cellular MTT reduction to that observed with βA4, indicating that this metabolic effect may be downstream of the production of intracellular ROS (Behl et al., 1994). In addition, cells exposed to low concentrations of βA4 have a reduced MTT reductase activity but survive indefinitely (JBD unpublished observations), suggesting that MTT reduction is sensitive to elevated ROS but that the inhibition is not sufficient for cell death. These observations raise two points: (1) that the MTT viability assay does not monitor cell death in these cell systems and (2) that βA4 elicits dramatic metabolic change in cells without concomitant cell death. Such metabolic changes may be sufficient to impair neuronal function and lead to cognitive impairment in the absence of neuronal cell loss. Increased oxidative stress may arise by two basic mechanisms: (1) a stimulation of ROS production or (2) a reduction in the cellular ability to remove constitutively generated ROS. Indirectly supporting the latter, diphenyliodonium rescues cortical neurones (Behl et al., 1994), whereas nordihydroguaiaretic acid rescues hippocampal neurones (Goodman et al., 1994), indicating that the source of ROS may be different enzymes in each cell type. βA4 might possibly induce ROS production via several enzymes or, alternatively, exert its effect by inhibiting a mechanism for antagonizing ROS production that is common to several cell types. The brain is particularly vulnerable to oxidative

stress, having low levels of catalase. A small increase in superoxide or superoxide dismutase activity may result in an increase in hydrogen peroxide. Hydrogen peroxide must then be detoxified, in the absence of catalase, by glutathione peroxidase or it may be converted via the metal dependent Fenton reaction to the highly toxic hydroxyl radical. Although chronic exposure to βA4 is known to alter antioxidant enzyme activities (Sagara et al., 1996), the possibility that acute exposure to βA4 may increase intracellular levels of ROS by altering turnover through superoxide dismutase or glutathione has yet to be fully explored. Hydrogen peroxide is less toxic than other ROS and the observation that iron facilitates βA4 toxicity is consistent with the conversion of hydrogen peroxide into highly toxic hydroxyl radicals (Schubert & Chevion, 1995). A very different hypothesis for the production of ROS is based on the observation, by electron paramagnetic resonance, that free radicals are produced by the amyloid peptides themselves and suggests that these radicals cause neurodegenerative damage directly (Hensley et al., 1994). Not all amyloid peptides which produce radicals are cytotoxic, suggesting that additional properties are necessary. As radicals are shortlived, one such property may be binding of the peptides to the cell surface in order to bring the radicals into contact with the cell. Peptides less prone to aggregation, such as scrambled peptides, may not acquire the correct conformation for binding to the cell surface, thus rationalizing their ability to produce radicals but lack of toxicity. These observations raise some interesting questions. Firstly, it is now well established that the toxic form of βA4 is not monomeric but an illdefined oligomer (Pike et al., 1993). In-vitro translation and aggregation experiments suggest that oxidative steps might also be involved in the aggregation of βA4 to form toxic oligomers (Dyrks et al., 1992). Given that most cell culture experiments are conducted over a time course of many hours or days, at least a portion of the antioxidants claimed to have protective properties against βA4 toxicity may do so by inhibiting oxidation dependent oligomerization, as opposed to ROS production subsequent to the binding of peptide to the cell surface.

The mechanism of cell death: apoptosis vs. necrosis Inhibition of hydrogen peroxide accumulation and prevention of cell death by antioxidants and enzyme

Oxidation and β-amyloid toxicity

inhibitors provides compelling evidence that ROS are involved in βA4 cytotoxicity, yet the downstream mechanisms of cell death have not been identified. A first step is the distinction between pathways akin to necrosis or apoptosis. βA4 has been shown to produce either necrotic cell death or apoptosis in different neuronal cell types (Gschwind & Huber, 1995) and apoptosis in transgenic mice overexpressing βA4 (LaFerla et al., 1995), indicating that the radicals and stimuli induced by βA4 are finely balanced in determining the fate of the cell. TUNEL staining has also identified apoptotic cells in Alzheimer’s disease brain but it is unclear whether the death of these cells is a major contributor to the disease process (Su et al., 1994). The initial oxidative stress induced by βA4 may be catastrophic and lead to necrosis. Alternatively, transcriptional activation, together with redox regulation of ion channels and metabolic pathways that are activated by βA4 induced ROS, may lead to responses that either rescue the cells from the oxidative stress or activate apoptosis. In support of βA4 induced ROS being able to transduce signals playing a role in apoptosis, exposure to βA4 results in activation of the transcription factor NFκB, which is also known to be activated in response to hydrogen peroxide (Behl et al., 1994). In addition, the antioxidant N-acetylcysteine can protect PC12 cell from apoptosis arising from the overexpression of mutant amyloid precursor protein (Zhao et al., 1995; Yamatsuji et al., 1996). A second gene, in which mutations leading to familial Alzheimer’s disease occur, has been linked to apoptotic mechanisms (Vito et al., 1996). A C-terminal portion of presenilin-2 is able to rescue a T cell hybridoma line cell line from apoptosis. A second presenilin gene, presenilin-1, is structurally very similar to presenilin-2 and is also linked with familial AD. Although a role for presenilins, in neuronal apoptosis, has yet to be demonstrated, it is tempting to speculate as to how presenilin biology and βA4 toxicity might be related.

Receptor for βA4 Despite the evidence for the cytotoxicity of βA4, little is known about the nature of the interaction between the peptide and the cell surface. The adhesiveness and aggregating properties of βA4 have made the identification of a receptor responsible for mediating the toxic effects of βA4 problematic. Studies showing differential susceptibility to βA4 cytotoxicity in a number of cell lines (Behl et al., 1992; Gschwind & Huber, 1995) suggest expression of a cell type specific recep-

443 tor. In contrast, several amphipathic peptides known to form amyloids exert a cytotoxic effect by a common oxidative mechanism (Schubert et al., 1995), and βA4 25–35 synthesised from all D-amino acids, and so having a structure that is a mirror image of the native peptide, has the same aggregation and toxicity profile as the L-isomer (Frey et al., 1994). These findings suggest that a sequence-specific receptor is unlikely. It is possible, however, especially in the light of the findings of Hensley et al., that conformation-specific binding, leading to adsorption to the cell surface and internalization, or membrane insertion may be critical steps. Interaction of βA4 with several receptors has been proposed, e.g. serpin-enzyme complex receptor, the lowdensity lipoprotein receptor related protein, or tachykinin receptors, but evidence that these receptors are involved in mediating the oxidative or cytotoxic responses to βA4 is lacking. More recently, it has been reported that the receptor for advanced-glycated-end products (RAGE) might be the receptor responsible for mediating many of the cellular responses to βA4. Although confirmation of these results is required, soluble RAGE and antibodies to RAGE are capable of blocking the oxidative responses to βA4 (Yan et al., 1996), and identification of a cell surface protein necessary for mediating amyloid toxicity would represent an important step forward.

Summary The in vitro evidence linking an oxidative response to the neurotoxicity of βA4 is compelling but still isolated by the realization that we know little about the interaction between βA4 and the cell surface, the origin of the ROS, or what mechanistic steps lie downstream from the oxidative stress. Human Alzheimer’s disease tissue contains altered levels of antioxidant enzymes and hallmarks of oxidative damage (Furuta et al., 1995; Lovell et al., 1995), which combined with the in vitro evidence above, raises some hope that an antioxidant approach may provide a useful therapy for Alzheimer’s disease.

Acknowledgements The author is very grateful to David Howlett for constructive criticism of the manuscript.

References Arispe N, Rojas E, Pollard HB (1993) Alzheimer’s disease amyloid β protein forms calcium channels in bilayer membranes: block-

444 ade by tromethamine and aluminium. Proc Natl Acad Sci USA 90:567–571 Behl C, Davis JB, Cole GM, Schubert D (1992) Vitamin E protects nerve cells from amyloid β protein toxicity. Biochem Biophys Res Comm 186:944–950 Behl C, Davis JB, Lesley R, Schubert D (1994) Hydrogen peroxide mediates amyloid β protein toxicity. Cell 77:817–827 Dyrks T, Dyrks E, Hartmann T, Masters C, Beyreuther K (1992) Amyloidogenicity of βA4 and βA4-bearing amyloid protein precursor fragments by metal-catalyzed oxidation. J Biol Chem 267:18210–18217 Etcheberrigary R, Ito E, Kim CS, Alkon DL (1994) Soluble β-amyloid induction of Alzheimer’s phenotype for human fibroblast K1 channels. Science 264:276–279 Frey P, Lis M, Ristig D, Stauss U, Amstutz R, Swoboda R, Staufenbiel M (1994) Studies on the mechanism of beta A4 toxicity. Neurobiol Aging 15 (Suppl. 1) S18 Furuta A, Price DL, Pardo CA, Troncoso JC, Xu ZS, Taniguchi N, Martin LJ (1995) Localization of superoxide dismutases in Alzheimer’s disease and Down’s syndrome neocortex and hippocampus. Am J Pathol 146:357–367 Goate A, Chartier-Harlin M-C, Mullan M, Brown J, Crawford F, Fidani L, Giuffra L, Haynes A, Irving N, James L, Mant R, Newton P, Rooke K, Roques P, Talbot C, Pericak-Vance M, Roses A, Williamson R, Rossor M, Owen M, Hardy J (1991) Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer’s disease. Nature 349:704–706 Goodman Y, Steiner MR, Steiner SM, Mattson MP (1994) Nordihydroguaiaretic acid protects hippocampal neurons against amyloid β-peptide toxicity, and attenuates free radical and calcium accumulation. Brain Res 654:171–176 Gschwind M, Huber G (1995) Apoptotic cell death induced by betaamyloid 1–42 peptide is cell type dependent. J Neurochem 65:292–300 Hensley K, Carney JM, Mattson MP, Aksenova M, Harris M, Wu JF, Floyd RA, Butterfield DA (1994) A model for β-amyloid aggregation and neurotoxicity based on free radical generation by the peptide: Relevance to Alzheimer’s disease. Proc Natl Acad Sci USA 91:3270–3274 Iversen LL, Mortishire-Smith RJ, Pollack SJ, Shearman MS (1995) The toxicity in vitro of β-amyloid protein. Biochem J 311:1–16 LaFerla FM, Tinkle BT, Bieberich CJ, Haudenschild CC, Jay G (1995) The Alzheimer’s Aβ peptide induces neurodegeneration and apoptotic cell death in transgenic mice. Nat Genetics 9:21–29 Lovell MA, Ehmann WD, Butler SM, Markesbery WR (1995) Elevated thiobarbituric acid-reactive substances and antioxidant enzyme activity in the brain in Alzheimer’s disease. Neurology 45:1594–1601 Mark RJ, Hensley K, Butterfield DA, Mattson MP (1995) Amyloid beta-peptide impairs ion-motive ATPase activities: evidence for a role in loss of neuronal Ca21 homeostasis and cell death. J Neurosci 15:6239–6249

J.B. Davis Masters CL, Simms G, Weinman NA, Multhaup G, McDonald BL, Beyreuther K (1985) Amyloid plaque core protein in Alzheimer disease and Down syndrome. Proc Natl Acad Sci USA 82:4245–4249 Mattson MP, Cheng B, Davis D, Bryant K, Lieberburg I, Rydel RE (1992) β-Amyloid peptides destabilise calcium homeostasis and render human cortical neurons vulnerable to excitotoxicity. J Neurosci 12:376–389 Mattson MP, Goodman Y (1995) Different amyloidogenic peptides share a similar mechanism of neurotoxicity involving reactive oxygen species and calcium. Brain Res 676:219–224 Pike CJ, Burdick D, Walencewicz AJ, Glabe CG, Cotman CW (1993) Neurodegeneration induced by β-amyloid peptides in vitro: the role of peptide assembly state. J Neurosci 13:1676–1687 Sagara Y, Dargusch R, Klier FG, Schubert D, Behl C (1996) Increased antioxidant enzyme activity in amyloid beta protein-resistant cells. J Neurosci 16:497–505 Schubert D, Chevion M (1995) The role of iron in beta amyloid toxicity. Biochem Biophys Res Commun 216:702–707 Schubert D, Behl C, Lesley R, Brack A, Dargusch R, Sagara Y, Kimura H (1995) Amyloid peptides are toxic via a common oxidative mechanism. Proc Natl Acad Sci USA 92:1989–1993 Shearman MS, Ragan IC, Iversen LL (1994) Inhibition of PC12 cell redox activity is a specific, early indicator of the mechanism of β-amyloid-mediated cell death. Proc Natl Acad Sci USA 91:1470–1474 Su JH, Anderson AJ, Cummings BJ, Cotman CW (1994) Immunohistochemical evidence for apoptosis in Alzheimer’s disease. Neuroreport 5:2529–2533 Terry RD, Katzman R, Bick KL (eds) (1994) Alzheimer’s disease, Raven Press, New York Vito P, Lacana E, D’Adamio L (1996) Interfering with apoptosis: Ca21-binding protein ALG-2 and Alzheimer’s disease gene ALG-3. Science 271:521–525 Yamatsuji T, Matsui T, Okamoto T, Komatsuzaki K, Takeda S, Fukumoto H, Iwatsubo T, Suzuki N, Asami-Odaka A, Ireland S, Kinane B, Giambarella U, Nishimoto I (1996) G protein-mediated neuronal DNA fragmentation induced by familial Alzheimer’s disease-associated mutants of APP. Science 272:1349–1352 Yan SD, Chen X, Fu J, Chen M, Godman G, Stern D, Schmidt A-M (1996) RAGE: A receptor upregulated in Alzheimer’s disease in neurons, microglia and cerebrovascular endothelium that binds amyloid-β peptide and mediates induction of oxidant stress. Neurology 46:S23.005 Yankner BA, Duffey LK, Kirschner DA (1990) Neurotrophic and neurotoxic effects of amyloid β protein. Reversal by tachykinin neuropeptides. Science 250:279–281 Zhao B, Sisodia SS, Kusiak JW (1995) Possible mechanisms of apoptosis induced by expression of mutant amyloid precursor protein in PC12 cells. Neurosci Abstr 21:672.10