Amyloid beta, neural lipids, cholesterol & Alzheimer's disease

Abstract

AMYLOID BETA, NEURAL LIPIDS, CHOLESTEROL & ALZHEIMER'S DISEASE

Natalia V. Koudinova1, Anatol Kontush2,3, Temirbolat T. Berezov1, Alexei R. Koudinov1§ CA

1 Russian Academy Med Sci, Moscow, Russian Federation
2 INSERM Unit 551, Hospital Pitie, Paris, France
3 Inst Med Biochem, U Hospital Eppendorf, Hamburg, Germany

To date great number of articles were devoted to cholesterol (chol) but only few articles studied the role for chol in neuron function/degeneration. For decades this molecule had been known to be important for atherosclerosis and heart disease. First indication of the involvement of chol in Alzheimer disease (AD), however, come from the mid 1990s. At that time it was shown that heart disease patients develop brain deposits of amyloid beta (Ab), a major dogmatic molecule of AD; that apoE (a chol transport apolipoprotein) allele e4 is a major genetic risk factor for AD; and that lab animals fed a chol diet express brain amyloid.1 These days it turns out that Ab, long thought to be exclusively a pathologic protein, is a normal and functional apolipoprotein constituent of high density lipoproteins in plasma and CSF. Thus, we and others showed that Ab modulates chol and phospholipid synthesis, and affects chol esterification.1 Protection of lipoproteins and other biomolecules from oxidation may represent another important function of Ab.2 We also discovered that neuronal chol homeostasis failure and the lack of chol supply to neurons by means of lipoprotein transport causes AD features, such as the failure of the neurotransmission and synaptic plasticity, degeneration of neuronal cell processes, and tau protein pathology.3, 4

Neurobiol. Lipids Vol.1, 6 (2003).


Posing the problem


Since 1966 more than one hundred thousand research papers were devoted to cholesterol and only few articles dealt with the role of cholesterol in neuronal function, synaptic plasticity and neurodegeneration (according to the PubMed and HighWire Press search engines). This mysterious molecule is accused in atherosclerosis and heart disease, but was largely understudied in relation to brain function and neural structural and functional (i.e. activity dependent) plasticity. Living cells (including neurons and glial cells) produce their own cholesterol and can receive or donate cholesterol via lipoprotein transit, an attested body lipid transportation system. This system operates a number of vehicle classes, including well-known (and considered to be “bad”) LDLs and “good“ HDLs [1, 2]

Apart from this whole-body-system stands distinct brain lipid transport authority that uses different subtypes of HDLs and normally does not maintain LDLs [3]. This vital service is in charge of cholesterol redistribution inside the brain and cholesterol export out of the brain border to liver for excretion. There is no reported cholesterol import to the brain, an issue that makes brain cholesterol availability entirely dependent on local manufacturing. Brain lipid transportation must have good management and operating capacity, because the quantity of cholesterol in the brain is much higher then anywhere else in the rest of the body. Thus, having just two percent of the body weight, brain has a quarter of cholesterol present in the whole individual.

Conceivable, the break in any element of the harmonized system of brain/neuronal cholesterol transport (caused by genetic defects of one of the enzyme or receptor associated with cholesterol turnover; by pharmacological modulation or environmentally) may result in abnormal homeostasis of cholesterol in the brain and impair fine tuning of synaptic function (see online Refs. 2, 3, 4 for instant access to detailed bibliography).


Synaptic function basic

Cholesterol (and phospholipids) is a building block of any cell membrane (the nervous system wrapping material, where neuronal information in the form of electric activity is generated and propagated) and specialized membrane structures, lipid rafts and synaptic vesicles. In the brain the information (coded as nerve impulse) is transmitted from presynaptic neuron-transmitter to a postsynaptic neuron-receiver at the tiny gap between cells called synapse. The first neuron output nerve ending (called axon) releases synaptic vesicles containing chemicals called neurotransmitters. The neurotransmitter molecules then bind to receptors of the neuron-receiver input processes, located in postsynaptic membrane functional domains, called lipid rafts. After the minute interaction with a receptor on the neuron-receiver, neurotransmitter molecules normally return back to the nerve ending (from where they were released) for recycling in order to be ready for the next act of neurotransmission. This transient neurotransmitter-receptor interaction represents the quantum of neurotransmission and synaptic function. It launches a number of chemical changes inside the postsynaptic neuron-receiver, essential for the nerve signal processing, synaptic amplification or modulation (also called synaptic plasticity), and for the formation of memory [2].


Cholesterol and Alzheimer's disease background

CHOLESTEROL AND ALZHEIMER’S DISEASE BACKGROUND


The turn of the century was marked by several reports that demonstrated Alzheimer’s diseases features in neuronal cells, brain slices and laboratory animals that model cholesterol pathology (see Refs. 2, 4, 5 for details). Our related research during past decade included the study of the interaction of lipid (particularly cholesterol) metabolism and amyloid b (Ab) protein (insoluble brain deposits of Ab are widely believed to be Alzheimer's characteristic feature and the cause of the disease, see our another presentation at the 32nd Society for Neuroscience Annual Meeting 2002) as a "missing link in Alzheimer's puzzle" (for brief review also see Ref. 1). This relation became especially important recently because more then dozen studies showed that cholesterol modulation affects generation of Ab and the processing of its precursor [4, 5, 6].

In the past, however, just few reports implicated cholesterol in basic synaptic function, particularly in trafficking and recycling of synaptic vesicles, in receptor function, activity of accessory synaptic proteins, and in modulation of membrane biophysical properties (see bibliography in Refs. 2, 4).


Cholesterol, phospholipids, synaptic plasticity & neurodegeneration



In our recent study published in June 2001 as research article in The FASEB Journal and in March 2002 as letter in Science magazine [2, 7, fr1, fr2, fr3] we attempted to gain basic knowledge on the role of cholesterol and phospholipids in neurotransmission, synaptic plasticity and nerve cells degeneration. To this end we prepared slices from the laboratory rat hippocampus, a part of the brain essential for learning and memory storage and having a lot of lipoprotein receptor molecules handling neuronal uptake of lipoproteins carrying cholesterol. Very thin slices (less then half a millimeter) of the hippocampus, retaining the hippocampal integrity and complex neurochemical and neuronal network architecture, were maintained live in a test tube. Then we set out to model cholesterol disbalance by acute increase of cholesterol turnover and inability of the hippocampus to redistribute cholesterol from one cell (astrocytes in particular) to another (neurons and their projections) via lipoprotein transport. To achieve this outcome the slices were subjected to biochemical increase of cholesterol removal with model and natural chemicals, methyl-b-cyclodextrin or normal human CSF HDL. Simultaneously or immediately thereafter we evoked and recorded two different brain waveforms: so called paired pulse facilitation (PPF) and long term potentiation (LTP), indicative of neurotransmission and synaptic plasticity, respectively. We found that the lack of cholesterol supply of neurons via lipoprotein transport impaired both PPF and LTP indicative of the neurotransmission and synaptic plasticity impairment. Moreover, we traced lipid synthesis with radioactive acetate (a precursor molecule which gets incorporated into all cellular lipids) after the generation of the LTP in the hippocampal slices. Based on the increase of the radioactive label incorporation into newly synthesized lipids specifically at the site of the LTP recording we concluded that this form of synaptic plasticity is associated with the increase of lipid synthesis. Such activity dependent lipid synthesis may represent the mechanism essential for structural plasticity of synapses at late stages of the LTP after the immediate access to lipids via lipoprotein transport during the LTP early stages.

In addition to the recording of neural waves we analyzed slices by immunofluorescence. We found that cholesterol disbalance causes neurofilament degeneration and the appearance of PHF-like phosphorylation of tau protein (confirmed in many studies by others, see Ref. 2, 4 for citations) in hippocampal axonal mossy fibers [2]. Exessive phosphorylation of tau is one of the key features of Alzheimer’s disease that is also present in cells under genetic cholesterol homeostasis disorder called Niemann-Pick type C disease [2, 4].


Cholesterol is the possible cause of neurodegeneration in AD


In a separate set of experiments by modifying rat cholesterol status with the diet containing 2 % cholesterol for several months we generated animal model, that expressed brain amyloid similar to amyloid deposition of Alzheimer's disease brain specimens [4, 6, fr3, fr4, fr5]. We then prepared slices from the hippocampus, maintained them live in vitro, and subjected to the study of cholesterol synthesis by labeling with radioactive acetate. After labeling the lipids were extracted from slices, separated by thin layer chromatography, and quantitated by radioactivity counting. This methodology allowed us to quantitate the higher rate of cholesterol and phospholipids synthesis in rats fed a cholesterol diet, and to conclude that brain cholesterol synthesis upregulation is a possible cause (but NOT a consequence) of Alzheimer's-like brain amyloid. Most important, we also performed electrophysiological analysis of slices. We found that cholesterol-fed rats lack hippocampal LTP and thus have impaired synaptic plasticity [4, 6] It is notable that impaired LTP could be reversed by the reversal of a 2% cholesterol diet to a regular diet for an extended period of time. It is in accord with the notion that the rate of cholesterol turnover in the brain is very low, and thus requires long time for a disturbance (or a correction in case of the opposite order of events) to yield appreciable change. Our functional analysis data significantly extended several earlier histochemical reports that demonstrated amyloid buildup in rabbits and in amyloid precursor protein transgenic mice fed a cholesterol diet (see Refs. 4, 5, 6 for details).

CHOLESTEROL, AMYLOID b AND ALZHEIMER’S CSF-HDL

Our other study [3] showed that Alzheimer’s patients have unique pattern of HDL distribution in the CSF, the lipoprotein fraction that is especially important in cholesterol transport in tissue environment and across the body, and that has soluble form of Ab as an apolipoprotein constituent [1]. Particularly we observed an increase in content of soluble Ab and selected apolipoproteins in the HDL subfraction called HDL1. Remembering a special role for HDL1 in reverse cholesterol transport on one hand, and the role for Ab in cholesterol esterification (that causes HDL size change and the formation of HDL1, Ref.3) we interpreted our data as an additional piece in Alzheimer’s cholesterol puzzle [3, 4].


Amyloid b is a potential physiological antioxidant...

AMYLOID b IS A POTENTIAL PHYSIOLOGICAL ANTIOXIDANT FOR LIPOPROTEINS IN CEREBROSPINAL FLUID AND PLASMA

Increased oxidative stress is related to the Alzheimer's development, and amyloid beta protein (Ab) is considered to be an important prooxidant in this process [8]. To induce oxidation, however, Ab must be present at high concentrations, typically in a micromolar range [9]. In addition, an Ab preparation must be 'aged' to yield Ab aggregates and fibrils [10]. In vitro, Ab is readily aggregated by transition metal ions [11]; in contrast, in the absence of metals Ab is monomeric. The presence of transition metals is not only required for Ab aggregation but also for its prooxidative activity [12]. Therefore, Ab toxicity is mediated by a direct interaction between Ab and transition metals with subsequent generation of reactive oxygen species (ROS).

The requirement of fibrillation and transition metals for the prooxidative activity of Ab can be understood taking into account its redox properties. In order to function as a prooxidant, Ab must first bind metals to its metal-binding site(s) at His residues [13] and then reduce them in its metal-reducing site at Met35 residue [14] in order to produce ROS.

However, metals bind to the N-terminal hydrophilic part of Ab, whereas metal reduction occurs at its C-terminal part. Since metals must be placed in the vicinity of the reductant to be reduced, fibrillation is likely to fulfil this task by forming complexes where metal atoms bound to the N-terminal part of one molecule of Ab, at the same time might be available for the reductive Met35 residues belonging to other Ab molecules. The resulting reduced transition metal ions can participate in further redox reactions, generating various free radical species. Due to relatively slow reduction of metals by Ab, the above mechanism can only be operative at high (micromolar) concentrations of the Ab peptide.

In contrast to prooxidative properties, an antioxidative activity of Ab peptides (that contradicts the dogmatic view on Ab as toxic) has been barely studied. We have shown that at low-nanomolar concentrations (i.e., those of soluble Ab in CSF and plasma), exogenously added Ab inhibits metal-catalyzed oxidation of lipoproteins of human CSF and plasma [15, 16]. The effect is observed at the peptide concentration reported for biological fluids (0.1-1.0 nM); at higher concentration of Ab its antioxidant action is abolished. In contrast, all Ab peptides are unable to considerably influence metal-independent lipoprotein oxidation, suggesting that the antioxidative activity of Ab is mainly mediated by chelating transition metal ions. Endogenous Ab present in CSF can also act as an antioxidant, as is suggested by the positive correlation between CSF resistance to oxidation and the CSF level of Ab [17].

Our data were confirmed by Zou et al. [18] who reported potent antioxidant activity of Ab in neuronal cells in the presence of transition metals. Ab protected neurons against toxic action of copper and iron; the effect strictly depended on the aggregation state of the peptide. Monomeric Ab was protective even at micromolar concentrations, whereas aggregated Ab lost its antioxidant properties. These results are consistent with earlier observations of Whitson et al., Yankner et al. and Koo et al. who showed that at low-nanomolar concentrations Ab is monomeric, non-toxic and exerts beneficial effect on neuron survival, axonal length and neurite outgrowth [19, 20, 21]. We propose that all these activities may be related to antioxidative properties of the peptide.

Mechanistically, antioxidative activity of Ab can be related to the fact that in lipoproteins, a metal-binding region of Ab expresses greater hydrophilic properties and extends into the outer aqueous phase where it can bind transition metals and inhibit metal-catalysed oxidation. In this regard it is important to note that neuronal cell cultures secrete a high molecular weight product, presumably a lipoprotein complex, that possesses an antioxidative activity [22].

As soon as Ab has antioxidative properties on one hand, and is secreted by cells as a part of lipoprotein complexes [1, 3, 23] on another, it is well possible that Ab is secreted by cells to serve as a natural antioxidant for lipoproteins. Ab can bind transition metal ions in inactive form and prevents them from catalyzing oxidation of lipoproteins and other biomolecular complexes. Amphiphilic properties of Ab may allow extracellular chelation of metal ions that escape binding by hydrophilic chelators.

AMYLOID b RESTORES HIPPOCAMPAL LONG TERM POTENTIATION: A CENTRAL ROLE FOR CHOLESTEROL
In our most recent study [24] we attempted to dissect out the role for Ab in the synaptic plasticity in brain slices from adult male rat hippocampus under the condition [2] that we characterised previously with regard to cholesterol and phospholipid synthesis. The prolonged maintanance of slices in a test tube for more then twenty hours in our experimental setup preserved synaptic function (input/output curve, a basic measure of synaptic function, for example) but abrogated synaptic plasticity (LTP). Ab protein of the 1-40 aminoacids’ molecule length (representing the major form of soluble Ab) rescued LTP while cholesterol synthesis inhibition with a statin abolished the LTP restoration by the peptide.

Our observation implies an intriguing perspective that Ab protein is a functional player in an activity-dependent cholesterol neurochemical pathways and in synaptic structure-functional plasticity [2, 3, 4, 24, 25]. The finding also supports our proposed hypothesis that the change in Ab biochemistry in Alzheimer's disease and related disorders is a functional (but NOT pathologic [25, 26, fr10]) compensatory phenomenon aiming to counterbalance impaired cholesterol dynamics and associated neurotransmission and synaptic plasticity [3, 4, 5, 25]. Such cholesterol mediated failure of synaptic function and neural degeneration [2, 4] in our view may represent the cause of the major sporadic form of Alzheimer's disease.


Scheme 1

Scheme 1 SCHEMATIC REPRESENTATION OF THE CASCADE OF THE COMMON SPORADIC FORMS OF ALZHEIMER'S DISEASE


(see Ref.4 for the extended bibliography for the scheme)

Neuronal cholesterol dynamics misregulation causes the key Alzheimer's disease (AD) feature of learning and memory failure as a result of the impairment of neuronal function, neurotransmission and synaptic plasticity through the mechanisms precise molecular nature which remains to be identified.

Cholesterol-mediated change in neurochemistry of amyloid beta, tau phosphorylation, neuronal cytoskeleton rearrangements and the modulation of physiological equilibrium of oxidative stress reactions could provide physiological transitory mechanisms aiming to compensate impaired brain cholesterol dynamics and neurotransmission and synaptic plasticity.

The break in neuronal cholesterol homeostasis may require very long (i.e. chronic) onset time frame due to the physiologically slow turnover of the central nervous system (CNS) cholesterol. Such condition may be genetically set (right top) and be assisted environmentally by the long term dietary habits. While during the past 30 years the concept of healthy food has become synonymous with avoiding dietary cholesterol, the question of how this avoidance and its compensation affects brain cholesterol chemistry, learning and memory remained non-addressed for many years. Several basic reports, however, documented that brain cholesterol is a delicate substance very sensitive to many influences, ranging from lipid preparation diets and chemical delivery systems for drugs and food additives (cyclodextrins, for example) to learning process itself. It is thus possible that antifat lifestyle “soft science” doctrine contributed to the increase of dementia and Alzheimer's prevalence in industrialised countries during 1970s and 1980s.

The indicated physiological compensatory changes may slowly invert when neuronal cholesterol dynamics is recovering slow to the initial physiological level. Such reversibility was proved experimentally (see Ref. 4) and certified by nature as an important mechanism of the CNS plasticity, as exampled by high expression of PHF-phosphorylated tau during an ontogenic period of cholesterol-demanding intense neuritic outgrowth. General compensatory nature of amyloid and tau neurochemistry modulation was proposed previously and is illustrated by its change observed under related to cholesterol (but different from AD) cardiovascular and Niemann-Pick type C pathologies, as well as in normal cases and during aging.

When neuronal cholesterol dynamics is not recovering compensatory mechanisms fail yielding (yet possible reversible) the development of conventional Alzheimer's disease hallmarks (right). These hallmarks, however, are not causative for the sporadic AD, and thus unlikely represent the proper target for the efficient AD therapy, as supported by the cognitive decline and dementia in AD patients without detectable lesions. Of these disease markers demonized amyloid beta pathology is the key enemy for the amyloid cascade hypothesis.

Plaque amyloid may itself impair (dotted arrows) synaptic plasticity and learning, neural networks, protein phosphorylation and oxidative stress status. Therefore it may have separate pathogenetic significance for the familial forms of AD, caused by the mutations in amyloid precursor protein and presenilins genes.

Similarly, oxidative stress independently disrupts synaptic plasticity and thus may have separate pathogenic value for the Down syndrome (characterized by upregulation of the reactions of oxidative stress due to the possible overexpression of the enzyme Cu/Zn-superoxide dismutase (SOD1), a chromosome 21 gene product) and for the pre-plaque stages of AD. The hallmarks trigger third order events of microglia activation, astrocytosis, cytokine/acute-phase protein release and cell death (not shown). This may convert physiological compensation into the pathological final and lock the cascade and the disease irreversibility.


Supplement

Supplement 1

ALZHEIMER’S DISEASE, CHOLESTEROL & AMYLOID b: DOGMA PREVAILS

Up to now, cholesterols’ role in Alzheimer's disease was mainly explained in terms of the dogmatic view that a reduction of amyloid burden by lowering cholesterol is beneficial [5, fr8], fr9, fr10]. This viewpoint (that become questioned in October 2002 Neurology article by Fassbender et al. [fr9]). was based on the in vitro data on the importance of cholesterol in amyloid precursor protein processing and Ab generation [4, 5]. Two recent articles further showed that cellular generation of Ab is modulated by cholesterol compartmentation and intracellular cholesteryl-ester levels [fr8].

The biochemical relation of cholesterol and Ab, however, is bidirectional (See Ref.4 for detailed bibliography).

Moreover, the modulation of neuronal cholesterol dynamics by Ab may have important functional consequences.

Particularly, Ab modulates neuronal cholesterol esterification, influx, efflux, and thus may regulate neural cholesterol intracellular compartmentation and extracellular trafficking [4]. Ab also modulates neuronal physical property of membrane fluidity important for receptor function, and it is well possible that this effect is mediated by the peptide antioxidant properties (see next section). Additionally, Ab increases neural lipid synthesis, in contrast to the peptide inhibitory effect, observed in human hepatic HepG2 and in HEK293 cells, in fetal rat liver and in neuronal tissue under the condition of potassium-evoked depolarization and under oxidative stress. The latter results highlight the importance of developmental, tissue and neuronal functional specificity of Ab-cholesterol biochemical relation, which may vary in different brain regions and be of special importance in determining Alzheimer's specific areas of neurodegeneration. The latter data also suggest that Ab may serve a molecular messanger function and manage the crosstalk of hepatic, systemic and brain cholesterol, and thus maintain the tissue-specific coordinate regulation of cholesterol biosynthesis. Taken together, the above functional consideration and recent data on the importance of cholesterol compartmentation for Ab generation indicate feedback mechanism between cholesterol and Ab homeostasis, additionally supported by a dependency of amyloid precursor protein processing and Ab production on the site 2 processing of SREBP and associated inability of cells to upregulate the expression of several enzymes and proteins involved in cholesterol synthesis and turnover [4].


Footnotes

CACorresponding author: Alexei Koudinov, M.D., Ph.D., P.O.Box 1665, Rehovot 76100, Israel ; e-mail: koudin@med.pfu.edu.ru ; URL: http://anzwers.org/free/neurology
#Footnote 1: Prior to the 32nd Society for Neuroscience Meeting (Orlando, Nov. 2-7, 2002) Neurobiology of Lipids published the editorial (2002, Vol.1, 5) that compiled the abstracts on the subject of the journal scope. Neurobiology of Lipids editors were invited to select the most noteworthy abstracts from the list of more then two hundered presentations. The short list yielded eighteen presentations. The authors of these presentations were invited to publish with us proceedings articles (matching the content of the SFN 2002 presentations) without additional peer review. This article represents the one of eight bronze choice presentations. The acceptance date represents the day of the article appearence at the 32nd SFN Annual Meeting in Orlando, Florida, Nov 2-7, 2002.

To review other neurobiology of lipids SFN Annual Meeting abstracts 2002 and the editors' choice shortlist please click here.

§Footnote 2: The presented proceedings article is based on a History of Neuroscience session of the 32nd Society for Neuroscience Annual Meeting 2002, and represents the research by two groups. Dr. Kontush and Dr. Arlt author the presentation section entitled "Amyloid b is a potential physiological antioxidant for lipoproteins in cerebrospinal fluid and plasma". Dr. Koudinova, Dr. Koudinov and Dr. Berezov contributed to the rest of the article. Dr. Arlt (Hamburg, Germany) decided that he did not contribute sufficiently to justify his authorship of this proceedings article.


References


1. Koudinov AR, Berezov TT, Koudinova NV. Alzheimer's amyloid beta and lipid metabolism: a missing link? FASEB J. 12, 1097-99 (1998)

2. Koudinov AR, Koudinova NV. Essential role for cholesterol in synaptic plasticity and neuronal degeneration. FASEB J. 15, 1858-60 (2001). Originally published online June 27, 2001, 10.1096/ fj.00-0815fje 

3. Koudinov AR, Berezov TT, Koudinova NV. The levels of soluble Ab in different HDL subfractions distinguish Alzheimer's/ normal aging CSF: implication for brain cholesterol pathology? Neurosci Lett. 314, 115-118 (2001)

4. Koudinov AR, Koudinova NV. Brain cholesterol pathology is the cause of Alzheimer's disease. Clin Med Health Res. Published online November 27, 2001, clinmed/2001100005

5. Koudinov AR and Koudinova NV. Alzheimer’s anti-amyloid vaccination and statins: two approaches, one dogma. The time for change. British Med J. Published online 20 March 2002

6. Koudinov AR, Koudinova NV. Modulation of cholesterol metabolism initiates Alzheimer’s amyloid deposition and neuronal dysfunction. Soc Neurosci Abstr. 26, 497 (2000)

7. Koudinov AR, Koudinova NV. Cholesterols' role in synapse formation. Science. 295, 2213 (2002)

8. Markesbery WR. Oxidative stress hypothesis in Alzheimer’sdisease. Free Radic. Biol. Med.23: 134–47 (1997) 

9. Varadarajan S, Yatin S, Aksenova M, Butterfield DA. Review: Alzheimer’s amyloid beta-peptide-associated free radical
oxidative stress and neurotoxicity. J Struct Biol.130, 184–208 (2000)

10. Iversen LL, Mortishire-Smith RJ, Pollack SJ, Shearman MS. The toxicity in vitro of beta-amyloid protein. Biochem J.
311, 1–16 (1995)

11. Atwood CS, Moir RD, Huang X, et al. Dramatic aggregation of Alzheimer Abeta by Cu(II) is induced by conditions representing physiological acidosis. J Biol Chem.273, 12817–26 (1998)

12. Rottkamp CA, Raina AK, Zhu X, et al. Redox-active iron mediates amyloid-beta toxicity. Free Radic Biol Med. 30, 447–50 (2001) 

13. Miura T,  Suzuki K, Kohata N, Takeuchi H. Metal binding modes of Alzheimer’s amyloid beta-peptide in insoluble aggregates and soluble complexes. Biochemistry39, 7024–31 (2000)

14. Huang X, Atwood CS, Hartshorn MA, et al. The Abeta peptide of Alzheimer’s disease directly produces hydrogen peroxide
through metal ion reduction. Biochemistry 38, 7609–16 (1999) 

15. Kontush A. Amyloid-beta: an antioxidant that becomes a pro-oxidant and critically contributes to Alzheimer's disease. Free Radic Biol Med. 31, 1120-31 (2001)

16. Kontush A, Berndt C, Weber W, et al. Amyloid-beta is an antioxidant for lipoproteins in cerebrospinal fluid  and plasma. Free Radic Biol Med. 30, 119-28 (2001) 

17. Kontush A, Donarski N, Beisiegel U. Resistance of human cerebrospinal fluid to in vitro oxidation is directly related to its amyloid-beta content. Free Radic Res. 35, 507-17 (2001)18. Zou K, Gong JS, Yanagisawa K, Michikawa M. A novel function of monomeric amyloid beta-protein serving as an antioxidant molecule against metal-induced oxidative damage. J Neurosci. 22, 4833-41 (2002) 

19. Whitson JS, Glabe CG, Shintani E, Abcar A, Cotman CW. Beta-amyloid protein promotes neuritic branching  in hippocampal cultures. Neurosci Lett.110, 319-24 (1990)

20. Yankner BA, Duffy LK, Kirschner DA. Neurotrophic and neurotoxic effects of amyloid beta protein: reversal by tachykinin neuropeptides. Science 250, 279-82 (1990)

21. Koo EH, Park L, Selkoe DJ. Amyloid beta-protein as a substrate interacts with extracellular matrix to promote neurite outgrowth. Proc Natl Acad Sci USA. 90, 4748-52 (1993)

22. Berndt C, Kontush A, Beisiegel U. Neuronal cell cultures protect low density lipoprotein from oxidation. Neurobiol Aging 19, S284 (1998) 

23. Koudinov AR, Koudinova NV. Soluble amyloid beta protein is secreted by HepG2 cells as an apolipoprotein. Cell Biol Inter. 25, 265-71 (1997)

24. Koudinov AR, Koudinova NV. Amyloid beta protein restores hippocampal long term potentiation: a central role for cholesterol? Soc Neurosci Abstr online. Program No.884.1 Published online 23 September 2002.

25. Koudinov AR, Smith MA, Perry G, Koudinova NV. Alzheimer's disease and amyloid beta protein. Science. Published online 25 June 2002.

26. Koudinov AR. Amyloid was never clearly implicated in Alzheimer's disease, so look at Abeta from a different angle. BMJ. Published  online 1 Dec 2002


http://www.biology-online.org/articles/amyloid_beta_neural_lipids/abstract.html