NMR and CD studies on the interaction of Alzheimer beta-amyloid peptide (12-28) with beta-cyclodextrin

Polymerization of the amyloid beta-peptide (Abeta) has been identified as a major feature of the pathogenesis of Alzheimer's disease (AD). Inhibition of the formation of these toxic polymers of Abeta has emerged as an approach for developing therapeutics for AD. NMR and CD spectra were used to investigate the interaction between cyclodextrin and Abeta(12-28) peptide, which was reported to be an important region for forming amyloid fibrils. CD spectral analyses show that of the alpha-, beta- and gamma-cyclodextrins only beta-cyclodextrin inhibits the aggregation of Abeta(12-28) at pH 5.0. Analysis of the one-dimensional proton NMR spectra of Abeta(12-28) and the mixture of Abeta(12-28) with beta-cyclodextrin clearly indicates that there are chemical shift changes in the aromatic ring of Phe19 and the methyl groups of Val18 in the peptide. The NOESY spectra show cross-peaks between H-3 and H-5 of beta-cyclodextrin and the aromatic protons of Phe19 and Phe20. These chemical shift differences and NOEs demonstrate that there is an interaction between Abeta(12-28) and beta-cyclodextrin. Analysis of the cross-peak intensity in the NOESY spectra reveals that the aromatic rings of Phe19 and 20 are generally inserted into beta-cyclodextrin at the broad side and are oriented toward the narrow side of the cavity.     

Mechanism of membrane depolarization caused by the Alzheimer Abeta1-42 peptide

We report a novel observation that the neurotoxic Alzheimer peptide Abeta1-42, when pre-incubated, causes a dramatic and lasting membrane depolarization in differentiated human hNT neuronal cells and in rodent PC12 cells in a concentration-dependent manner. This phenomenon involves activation of the metabotropic glutamate receptor, mGluR(1). Abeta-induced membrane depolarization in PC12 cells is sensitive to mGluR(1) antagonists and to pertussis and cholera toxins, indicating the involvement of particular G-proteins. The effect is different from the known ability of aggregated Abeta1-42 to cause a calcium influx. Since mGluR(1) agonists mimic the Abeta effect, we deduce that in this cell system glutamate can control the membrane potential and thereby the excitability of its target neurons. We propose that Abeta-induced membrane depolarization described here leads in Alzheimer's disease to hyperexcitability of affected neurons and is a crucially important molecular mechanism for beta-amyloid toxicity and cognitive dysfunction in the disease.     

Mitochondria morphology and DNA content upon sublethal exposure to beta-amyloid(1-42) peptide

Brains affected by Alzheimer's disease (AD) show a large spectrum of mitochondrial alterations at both morphological and genetic level. The causal link between amyloid beta peptides (AP) and mitochondrial dysfunction has been established in cellular models of AD using Abeta concentrations capable of triggering massive neuronal death. However, mitochondrial changes related to sublethal exposure to Abeta are less known. Here we show that subtoxic, 1 microM Abeta(1-42) exposure does not change the mitochondrial shape of living cells, as visualized upon the uptake of the non-potentiometric fluorescent probe Mitotracker Green and enhanced yellow fluorescent protein (EYFP)-tagged cytochrome c oxidase expression. Immunolocalization of oxidative adducts 8-hydroxy-2'-deoxyguanosine, 8-hydroxyguanine and 8-hydroxyguanosine demonstrates that one-micromolar concentration of Abeta(1-42) is also not sufficient to elicit dramatic qualitative changes in the RNA/DNA oxidative products. However, in comparison with controls, semi-quantitative analysis of the overall mitochondrial mass by integrated fluorescence intensity reveals an ongoing down-regulation in mitochondrial biosynthesis or, conversely, an enhanced autophagic demise of Abeta treated cells. Furthermore, a significant increase of the full-length mitochondrial DNA (mtDNA) from Abeta-treated versus control cells is found, as measured by long range polymerase chain reaction (PCR). Such up-regulation is accompanied by extensive fragmentation of the unamplified mtDNA, probably due to the detrimental effect of Abeta. We interpret these results as a sequence of compensatory responses induced by mtDNA damage, which are devoted to repression of oxidative burst. In conclusion, our findings suggest that early therapeutic interventions aimed at prevention of mitochondrial oxidative damage may delay AD progression and help in treating AD patients.     

Structure of the Alzheimer beta-amyloid peptide (25-35) and its interaction with negatively charged phospholipid vesicles.

The secondary structure of amyloid betaAP(25-35) peptide was studied in pure form and in the presence of different phospholipid vesicles, by using Fourier transform infrared spectroscopy (FT-IR). Pure peptide aggregated with time, forming fibrils with beta-structure. Phospholipid vesicles formed by negatively charged phospholipids such as 1,2-dimyristoyl-sn-glycerol-3-phospho-L-serine (Myr2PtdSer), 1,2-dimyristoyl-sn-glycerol-3-phospho-rac-1-glycerol (Myr2PtdGro) and 1,2-dimyristoyl-sn-glycerol 3-phosphate (Myr2PtdH), greatly accelerated the aggregation of the peptide. However, the presence of vesicles formed by the zwitterionic phospholipid, 1, 2-dimyristoyl-sn-glycerol-3-phosphocholine (Myr2PtdCho), slowed down the aggregation process. Differential scanning calorimetry (DSC) measurements showed that the effect of betaAP(25-35) on the gel to crystal liquid phase transition was small at neutral pH for negatively charged phospholipids and practically nil for Myr2PtdCho. In the case of Myr2PtdSer the effect was also zero at pH 9 but the effect was large at pH 3. The effect on Myr2PtdH was not, however, very dependent on pH. These results were fully confirmed by the observation through FT-IR of the change with temperature of the CH2 antisymmetric stretching vibration. The case of Myr2PtdGro was special as this phospholipid presents polymorphism giving solid quasicrystalline phases when it is not sufficiently hydrated, and it is remarkable that betaAP(25-35) was able to induce the formation of crystalline phases in samples prepared through a method which ensure a good hydration of phospholipid. These results show that the interaction of amyloid betaAP(25-35) peptide with phospholipids is based on electrostatic interactions, that these interactions favour the aggregation of the peptides, and that the presence of the aggregates may disturb the lipid-water interphase of the membrane.     

Membrane destabilization induced by beta-amyloid peptide 29-42: importance of the amino-terminus

Increasing evidence implicates interactions between Aβ-peptides and membrane lipids in Alzheimer's disease. To gain insight into the potential role of the free amino group of the N-terminus of Aβ29-42 fragment in these processes, we have investigated the ability of Aβ29-42 unprotected and Aβ29-42 N-protected to interact with negatively-charged liposomes and have calculated the interaction with membrane lipids by conformational analysis. Using vesicles mimicking the composition of neuronal membranes, we show that both peptides have a similar capacity to induce membrane fusion and permeabilization. The fusogenic effect is related to the appearance of non-bilayer structures where isotropic motions occur as shown by ³¹P and ²H NMR studies. The molecular modeling calculations confirm the experimental observations and suggest that lipid destabilization could be due to the ability of both peptides to adopt metastable positions in the presence of lipids. In conclusion, the presence of a free or protected (acetylated) amino group in the N-terminus of Aβ29-42 is therefore probably not crucial for destabilizing properties of the C-terminal fragment of Aβ peptides.     

Differential physiologic responses of alpha7 nicotinic acetylcholine receptors to beta-amyloid1-40 and beta-amyloid1-42

The beta-amyloid peptides (Abeta), Abeta(1-40) and Abeta(1-42), have been implicated in Alzheimer's disease (AD) pathology. Although Abeta(1-42) is generally considered to be the pathological peptide in AD, both Abeta(1-40) and Abeta(1-42) have been used in a variety of experimental models without discrimination. Here we show that monomeric or oligomeric forms of the two Abeta peptides, when interact with the neuronal cation channel, alpha7 nicotinic acetylcholine receptors (alpha7nAChR), would result in distinct physiologic responses as measured by acetylcholine release and calcium influx experiments. While Abeta(1-42) effectively attenuated these alpha7nAChR-dependent physiology to an extent that was apparently irreversible, Abeta(1-40) showed a lower inhibitory activity that could be restored upon washings with physiologic buffers or treatment with alpha7nAChR antagonists. Our data suggest a clear pharmacological distinction between Abeta(1-40) and Abeta(1-42).     

Kinetics of adsorption of beta-amyloid peptide Abeta(1-40) to lipid bilayers

The Alzheimer's disease-related peptide beta-amyloid (Abeta) is toxic to neurons. The toxicity of the peptide appears to require conversion of the monomeric form to an aggregated fibrillar species. The interaction of Abeta with cell membranes has attracted interest as one plausible mechanism by which the peptide exerts its toxic activity. We developed two methods to measure the adsorption of fresh (monomeric) and aged (aggregated) Abeta to lipid bilayers. In one method, the kinetics of Abeta adsorption and desorption to liposomes deposited onto a dextran-coated surface was measured using surface plasmon resonance. In the other method, Abeta was contacted with liposome-coated magnetic beads; adsorbed Abeta was separated from solution-phase peptide by use of a magnetic field. Monomeric Abeta adsorbed quickly but reversibly to lipid bilayers with low affinity, while aggregated Abeta adsorbed slowly but irreversibly. These two methods provide complementary means of quantifying the adsorption of aggregating proteins to membranes. The results correlate strongly with previous observations that fibrillar, but not monomeric, Abeta restricts the motion of acyl tails in phospholipid bilayers. The methods should be useful for further elucidation of the role of membrane adsorption in mediating Abeta toxicity, and in the search for inhibitors of toxicity.     

Structural studies of soluble oligomers of the Alzheimer beta-amyloid peptide

Recent studies have suggested that non-fibrillar soluble forms of Abeta peptides possess neurotoxic properties and may therefore play a role in the molecular pathogenesis of Alzheimer's disease. We have identified solution conditions under which two types of soluble oligomers of Abeta40 could be trapped and stabilized for an extended period of time. The first type of oligomers comprises a mixture of dimers/tetramers which are stable at neutral pH and low micromolar concentration, for a period of at least four weeks. The second type of oligomer comprises a narrow distribution of particles that are spherical when examined by electron microscopy and atomic force microscopy. The number average molecular mass of this distribution of particles is 0.94 MDa, and they are are stable at pH 3 for at least four weeks. Circular dichroism studies indicate that the dimers/tetramers possess irregular secondary structure that is not alpha-helix or beta-structure, while the 0.94 MDa particles contain beta-structure. Fluorescence resonance energy transfer experiments indicate that Abeta40 moieties in amyloid fibrils or protofibrils are more similar in structure to those in the 0.94 MDa particles than those in the dimers/tetramers. These findings indicate that soluble oligomeric forms of Abeta peptides can be trapped for extended periods of time, enabling their study by high resolution techniques that would not otherwise be possible.     

Eliminating membrane depolarization caused by the Alzheimer peptide Abeta(1-42, aggr.)

A high-throughput screen found compounds that eliminate the dramatic membrane depolarization caused by the aggregated Alzheimer Abeta1-42 peptide, which activates mGluR1 receptors. The library was composed of known biologically active compounds; the cell-based assay measured the changes of membrane potential with a slow-acting voltage-sensitive dye. We found 10 potentially useful compounds, some of which reduce the Abeta-induced membrane depolarization up to 96%. Interestingly, the active compounds include specific tyrosine kinase inhibitors and inhibitors of certain chloride channels. We deduce that mGluR1 receptors, activated by Abeta1-42 or otherwise, can control the membrane potential via downstream activation of certain tyrosine kinases and certain ion channels. Dopaminergic and serotonergic agonists that emerged from the screen presumably compensate for the Abeta-induced membrane depolarization by themselves causing a hyperpolarization. The hit compounds, whose pharmacokinetics are known, show promise for the restoration of cognitive function in the treatment of early and mid-stage Alzheimer's disease.     

Duration and specificity of humoral immune responses in mice vaccinated with the Alzheimer's disease-associated beta-amyloid 1-42 peptide.

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by overproduction of beta-amyloid (Abeta), which is formed from amyloid precursor protein (APP), with the subsequent pathologic deposition of Abeta in regions of the brain important for memory and cognition. Recently, vaccination of murine models of AD that exhibit Abeta deposition has halted or delayed the usual progression of the pathology of AD. Our group has demonstrated that vaccination of a doubly transgenic mouse model (expressing mutant APP and presenilin-1) with the Abeta 1-42 peptide protects these mice from the memory deficits they would ordinarily develop. This report further characterizes the Abeta 1-42 peptide vaccine in mice. Anti-Abeta response time course analysis indicated that at least three vaccinations (each 100 microg) were necessary to elicit a significant anti-Abeta titer. Subsequent vaccinations resulted in half-maximal antibody titers of at least 10,000, and these titers were maintained for at least 5 months after the final boost. Peptide binding competition studies indicated that the highest humoral responses are generated against the N terminus of the Abeta peptide. Also, measurement of specific murine Ig isotypes in Abeta-vaccinated mice demonstrated a predominant IgG(1) and IgG(2b) response, suggesting a type 2 (Th2) T-helper cell immune response, which drives humoral immunity. Finally, lymphocyte proliferation assay experiments using Abeta peptides and splenocytes from vaccinated mice demonstrated that the vaccine specifically stimulates T-cell epitopes present within the Abeta peptide.     

Elevation of beta-amyloid peptide 2-42 in sporadic and familial Alzheimer's disease and its generation in PS1 knockout cells

Urea-based beta-amyloid (Abeta) SDS-polyacrylamide gel electrophoresis and immunoblots were used to analyze the generation of Abeta peptides in conditioned medium from primary mouse neurons and a neuroglioma cell line, as well as in human cerebrospinal fluid. A comparable and highly conserved pattern of Abeta peptides, namely, 1-40/42 and carboxyl-terminal-truncated 1-37, 1-38, and 1-39, was found. Besides Abeta1-42, we also observed a consistent elevation of amino-terminal-truncated Abeta2-42 in a detergent-soluble pool in brains of subjects with Alzheimer's disease. Abeta2-42 was also specifically elevated in cerebrospinal fluid samples of Alzheimer's disease patients. To decipher the contribution of potential different gamma-secretases (presenilins (PSs)) in generating the amino-terminal- and carboxyl-terminal-truncated Abeta peptides, we overexpressed beta-amyloid precursor protein (APP)-trafficking mutants in PS1+/+ and PS1-/- neurons. As compared with APP-WT (primary neurons from control or PS1-deficient mice infected with Semliki Forest virus), PS1-/- neurons and PS1+/+ neurons overexpressing APP-Deltact (a slow-internalizing mutant) show a decrease of all secreted Abeta peptide species, as expected, because this mutant is processed mainly by alpha-secretase. This drop is even more pronounced for the APP-KK construct (APP mutant carrying an endoplasmic reticulum retention motif). Surprisingly, Abeta2-42 is significantly less affected in PS1-/- neurons and in neurons transfected with the endocytosis-deficient APP-Deltact construct. Our data confirm that PS1 is closely involved in the production of Abeta1-40/42 and the carboxyl-terminal-truncated Abeta1-37, Abeta1-38, and Abeta1-39, but the amino-terminal-truncated and carboxyl-terminal-elongated Abeta2-42 seems to be less affected by PS1 deficiency. Moreover, our results indicate that the latter Abeta peptide species could be generated by a beta(Asp/Ala)-secretase activity.     

Poly-L-lysine dissolves fibrillar aggregation of the Alzheimer beta-amyloid peptide in vitro

beta-Amyloid peptide (beta A) is a major fibrillar component of neuritic plaques in Alzheimer's disease (AD) brains and is related to the pathogenesis of the disease. In this study, using electron microscopy, we describe herein the results concerning the efficacy of compounds that can dissolve preformed beta A fibrils in vitro. For such a purpose, two hydrosoluble and biocompatible polymers such as polyethylene glycol and poly-L-lysine were used. The poly-L-lysine appears as a potent dissolver of preformed beta A fibrils in vitro. Its efficiency is instantaneous. Poly-L-lysine can be used as a universal dissolver of all types of oligomeric beta-sheet conformation, precursor of the fibrils. This finding provides the basis for future investigation of the therapeutic potential of poly-L-lysine in terms of preventing and/or retarding amyloidogenesis in AD and other types of amyloid-related disorders.     

IL-4-induced selective clearance of oligomeric beta-amyloid peptide(1-42) by rat primary type 2 microglia

A hallmark of immunopathology associated with Alzheimer's disease is the presence of activated microglia (MG) surrounding senile plaque deposition of beta-amyloid (Abeta) peptides. Abeta peptides are believed to be potent activators of MG, which leads to Alzheimer's disease pathology, but the role of MG subtypes in Abeta clearance still remains unclear. In this study, we found that IL-4 treatment of rat primary-type 2 MG enhanced uptake and degradation of oligomeric Abeta(1-42) (o-Abeta(1-42)). IL-4 treatment induced significant expression of the scavenger receptor CD36 and the Abeta-degrading enzymes neprilysin (NEP) and insulin-degrading enzyme (IDE) but reduced expression of certain other scavenger receptors. Of cytokines and stimulants tested, the anti-inflammatory cytokines IL-4 and IL-13 effectively enhanced CD36, NEP, and IDE. We demonstrated the CD36 contribution to IL-4-induced Abeta clearance: Chinese hamster ovary cells overexpressing CD36 exhibited marked, dose-dependent degradation of (125)I-labeled o-Abeta(1-42) compared with controls, the degradation being blocked by anti-CD36 Ab. Also, we found IL-4-induced clearance of o-Abeta(1-42) in type 2 MG from CD36-expressing WKY/NCrj rats but not in cells from SHR/NCrj rats with dysfunctional CD36 expression. NEP and IDE also contributed to IL-4-induced degradation of Abeta(1-42), because their inhibitors, thiorphan and insulin, respectively, significantly suppressed this activity. IL-4-stimulated uptake and degradation of o-Abeta(1-42) were selectively enhanced in type 2, but not type 1 MG that express CD40, which suggests that the two MG types may play different neuroimmunomodulating roles in the Abeta-overproducing brain. Thus, selective o-Abeta(1-42) clearance, which is induced by IL-4, may provide an additional focus for developing strategies to prevent and treat Alzheimer's disease.     

Plasma beta carotene in Alzheimer's disease. Association with cerebrospinal fluid beta-amyloid 1-40, (Abeta40), beta-amyloid 1-42 (Abeta42) and total Tau

We studied the plasma beta carotene concentrations in 40 Alzheimer's disease patients and the association with cerebrospinal fluid beta-amyloid 1-40, (Abeta40), cerebrospinal fluid beta-amyloid 1-42 (Abeta42) and cerebrospinal fluid total Tau. We found that patients with plasma beta carotene levels below the 25th percentile had 55% reduced ratios of Abeta40/Tau and 51% reduced ratios of Abeta 40/Abeta 42 compared with patients in the highest quartile. Mean Tau concentrations in the lowest quartile of plasma beta-carotene levels were 74% higher compared with the highest quartile of plasma beta-carotene levels. Thus, we could demonstrate an statistically significant association between beta carotene levels in plasma and neurochemical markers in the cerebrospinal fluid of Alzheimer's disease patients.     

Amyloidogenic processing of the Alzheimer beta-amyloid precursor protein depends on lipid rafts

Formation of senile plaques containing the beta-amyloid peptide (A beta) derived from the amyloid precursor protein (APP) is an invariant feature of Alzheimer's disease (AD). APP is cleaved either by beta-secretase or by alpha-secretase to initiate amyloidogenic (release of A beta) or nonamyloidogenic processing of APP, respectively. A key to understanding AD is to unravel how access of these enzymes to APP is regulated. Here, we demonstrate that lipid rafts are critically involved in regulating A beta generation. Reducing cholesterol levels in N2a cells decreased A beta production. APP and the beta-site APP cleavage enzyme (BACE1) could be induced to copatch at the plasma membrane upon cross-linking with antibodies and to segregate away from nonraft markers. Antibody cross-linking dramatically increased production of A beta in a cholesterol-dependent manner. A beta generation was dependent on endocytosis and was reduced after expression of the dynamin mutant K44A and the Rab5 GTPase-activating protein, RN-tre. This inhibition could be overcome by antibody cross-linking. These observations suggest the existence of two APP pools. Although APP inside raft clusters seems to be cleaved by beta-secretase, APP outside rafts undergoes cleavage by alpha-secretase. Thus, access of alpha- and beta-secretase to APP, and therefore A beta generation, may be determined by dynamic interactions of APP with lipid rafts.     

The interaction of beta-amyloid protein fragment (12-28) with lipid environments.

The neurotoxicity of beta-amyloid protein (beta AP) fragments may be a result of their solution conformation, which is very sensitive to solution conditions. In this work we describe NMR and CD studies of the conformation of beta AP(12-28) in lipid (micelle) environments as a function of pH and lipid type. The interaction of beta AP(12-28) with zwitterionic dodecylphosphocholine (DPC) micelles is weak and alters the conformation when compared to water solution alone. By contrast, the interaction of the peptide with anionic sodium dodecylsulfate (SDS) micelles is strong: beta AP(12-28) is mostly bound, is alpha-helical from K16 to V24, and aggregates slowly. The pH-dependent conformation changes of beta AP(12-28) in solution occur in the pH range at which the side-chain groups of E22, D23, H13, and H14 are deprotonated (pKas ca. 4 and 6.5); the interaction of beta AP(12-28) with SDS micelles alters the pH-dependent conformational transitions of the peptide whereas the weak interaction with DPC micelles causes little change.     

Effect of sterols on beta-amyloid peptide (AbetaP 1-40) channel formation and their properties in planar lipid membranes

We investigate the role played by membrane composition on the interaction and self-assembly of beta-amyloid peptide (AbetaP1-40) during pore formation in planar lipid membranes (PLMs). Incorporation studies showed that AbetaP does not interact with zwitterionic membranes made up of phosphatidylcholine, whereas the addition of cholesterol or ergosterol to the membranes leads to channel formation. Among the PLMs used, a higher propensity of AbetaP to form channels at low applied potential (+/-20 mV) was observed in 7-dehydrocholesterol and in oxidized cholesterol PLMs. These channels present long lifetimes, high-occurrence frequencies, and are voltage dependent. In particular, the AbetaP channel in oxidized cholesterol showed anion selectivity. Thus cholesterol (and sterols in general) could be considered as targets for AbetaP, which prevents the fibrillation process by increasing incorporation into membranes. Furthermore, by switching the channel selectivity versus anions, cholesterol helps to reduce the imbalance of the cellular ions, calcium included, induced by membrane depolarization, which could be one of the factors responsible for cytotoxicity in Alzheimer's disease.     

Alternate aggregation pathways of the Alzheimer beta-amyloid peptide: Abeta association kinetics at endosomal pH

The deposition of beta-amyloid peptide (Abeta) fibrils around neurons is an invariable feature of Alzheimer's disease and there is increasing evidence that fibrillar deposits and/or prefibrillar intermediates play a central role in the observed neurodegeneration. One site of Abeta generation is the endosomes, and we have investigated the kinetics of Abeta association at endosomal pH over physiologically relevant time frames. We have identified three distinct Abeta association phases that occur at rates comparable to endosomal transit times. Rapid formation of burst phase aggregates, larger than 200nm, was observed within 15 seconds. Two slower association phases were detected by fluorescence resonance energy transfer and termed phase 1 and phase 2 aggregation reactions. At 20 microM Abeta, pH 6, the half lives of the phase 1 and phase 2 aggregation phases were 3.15 minutes and 17.66 minutes, respectively. Atomic force microscopy and dynamic light scattering studies indicate that the burst phase aggregate is large and amorphous, while phase 1 and 2 aggregates are spherical with hydrodynamic radii around 30 nm. There is an apparent equilibrium, potentially mediated through a soluble Abeta intermediate, between the large burst phase aggregates and phase 1 and 2 spherical particles. The large burst phase aggregates form quickly, however, they disappear as the equilibrium shifts toward the spherical aggregates. These aggregated species do not contain alpha-helical or beta-structure as determined by circular dichroism spectroscopy. However, after two weeks beta-structure is observed and is attributable to the insoluble portion of the sample. After two months, mature amyloid fibrils appear and the spherical aggregates are significantly diminished.     

Alzheimer beta-amyloid peptides: normal and abnormal localization

Alzheimer's disease (AD) neuropathology is characterized by accumulation of "senile" plaques (SPs) and neurofibrillary tangles (NFTs) in vulnerable brain regions. SPs are principally composed of aggregates of up to 42/43 amino acid beta-amyloid (A beta) peptides. The discovery of familial AD (FAD) mutations in the genes for the amyloid precursor protein (APP) and presenilins (PSs), all of which increase A beta42 production, support the view that A beta is centrally involved in the pathogenesis of AD. A beta42 aggregates readily, and is thought to seed the formation of fibrils, which then act as templates for plaque formation. A beta is generated by the sequential intracellular cleavage of APP by beta-secretase to generate the N-terminal end of A beta, and intramembranous cleavage by gamma-secretase to generate the C-terminal end. Cell biological studies have demonstrated that A beta is generated in the ER, Golgi, and endosomal/lysosomal system. A central question involving the role of A beta in AD concerns how A beta causes disease and whether it is extracellular A beta deposition and/or intracellular A beta accumulation that initiates the disease process. The most prevalent view is that SPs are composed of extracellular deposits of secreted A beta and that A beta causes toxicity to surrounding neurons as extracellular SP. The recent emphasis on the intracellular biology of APP and A beta has led some investigators to consider the possibility that intraneuronal A beta may directly cause toxicity. In this review we will outline current knowledge of the localization of both intracellular and extracellular A beta.