• free radical;
  • hydroxyl radical;
  • proteasome;
  • protein aggregation;
  • reactive oxygen species;
  • superoxide.


  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

The brain and nervous system are prone to oxidative stress, and are inadequately equipped with antioxidant defense systems to prevent ‘ongoing’ oxidative damage, let alone the extra oxidative damage imposed by the neurodegenerative diseases. Indeed, increased oxidative damage, mitochondrial dysfunction, accumulation of oxidized aggregated proteins, inflammation, and defects in protein clearance constitute complex intertwined pathologies that conspire to kill neurons. After a long lag period, therapeutic and other interventions based on a knowledge of redox biology are on the horizon for at least some of the neurodegenerative diseases.

Abbreviations used

amyloid peptide


Alzheimer's disease

AGE product

advanced glycation end-product


α-ketoglutarate dehydrogenase


amyotrophic lateral sclerosis


Creutzfeldt–Jakob disease


copper- and zinc-containing superoxide dismutase


cytochrome P450


docosahexaenoic acid




familial ALS


glutathione peroxidase


reduced glutathione


oxidized glutathione




heat-shock protein


heme oxygenase


inhibitor of NF-κB


inducible nitric oxide synthase




lipid hydroperoxide


mild cognitive impairment


manganese-containing superoxide dismutase


1-methyl-4-phenylpyridinium ion


methionine sulfoxide reductase


neuronal nitric oxide synthase


NADPH oxidase






polyADP-ribose polymerase 1


Parkinson's disease




phosphatase and tensin homolog


polyunsaturated fatty acid


reactive bromine species


reactive chlorine species


reactive nitrogen species


reactive oxygen species


reactive species


superoxide dismutase




transient receptor potential melastatin-related


ubiquitin C-terminal hydrolase L1

In 1992, I wrote a review in this journal entitled ‘reactive oxygen species and the central nervous system’ (Halliwell 1992). It is interesting to look back over the past 13 years or so and see how much has changed. In some areas, not a lot (indeed, I can re-use about 10% of the original text!). In others, there have been enormous advances in our knowledge and understanding of molecular mechanisms of oxidative damage. Sadly, this knowledge has not yet translated into improved therapies for the neurodegenerative diseases, although that may soon change.

Attacking the nervous system

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

Humans need oxygen to survive, but hyperoxia produces toxicity, including neurotoxicity (Chavko et al. 2003). Indeed, as we have learned to measure oxidative damage accurately, we find that it happens in all our tissues all the time, even under 21% O2; there is always a basal level of oxidative damage to DNA, lipids and proteins (reviewed by Halliwell and Whiteman 2004; Halliwell and Gutteridge 2006). Thus systems that repair and replace oxidized biomolecules are essential, and failures in them contribute to neurodegeneration, as we see below.

In healthy aerobes, there is a balance between the production of various reactive species (RS; defined in Table 1) and antioxidant defenses – some species are not immediately removed because they perform important biological roles (reviewed by Halliwell and Gutteridge 2006). Indeed, RS participate directly in defense against infection and also are important coordinators of the inflammatory response. Infectious disease was a powerful driver of natural selection in early human civilizations – those able to mount a robust immune response (producing lots of RS and other antibacterial/viral agents) tended to survive to pass on their genes to the next generation. But too many RS can injure human tissues and, if this happens often enough, chronic disease results. However, this usually happens later in life, in the post-reproductive years, so was not selected against during evolution (Kirkwood 2005). Repair mechanisms tend to fail in old age probably for the same reason; evolution has selected to invest scarce energy resources in them only until reproduction is complete and the offspring can survive alone. An extreme example of this is the Pacific salmon (Oncorhynchus genus). They grow, reproduce once, and then deteriorate fast and die – they virtually ‘fall apart’. The chances of a Pacific salmon surviving to reproduce again are small, and evolution as usual did not favor ‘holding back resources’ for survival after reproduction. To quote Kirkwood (2005): ‘The rapid deterioration of Pacific salmon after mating is a natural byproduct of a life history that has been geared by natural selection to stake everything on a single bout of reproduction. As soon as the signal to reproduce is triggered, a massive effort is made to mobilize all available resources to maximize reproductive success, even if this leaves the adult so severely depleted or damaged that death ensues’. Post-reproductive humans do not fail so drastically, but as we age the incidence of disease increases, raising the ugly scenario that if we managed to live long enough we would all suffer cancer and neurodegenerative diseases.

Table 1.   Some reactive species
Free radicalsNon-radicals
  • ‘ROS’ is a collective term that includes both oxygen radicals and certain non-radicals that are oxidizing agents and/or are easily converted into radicals (HOCl, HOBr, O3, ONOO, 1O2, H2O2). All oxygen radicals are ROS, but not all ROS are oxygen radicals. Peroxynitrite and H2O2 are frequently erroneously described in the literature as free radicals, for example. ‘RNS’ is a similar collective term that includes NO and inline image as well as non-radicals such as HNO2 and N2O4. ‘Reactive’ is not always an appropriate term: H2O2, NO and inline image react fast with few molecules, whereas OH reacts fast with almost everything. Species such as inline imageinline image, RO, HOCl, HOBr, inline image, inline image, inline image, ONOO, NO2+ and O3 have intermediate reactivities.

  • *

    HOBr and BrCl could also be regarded as RBS.

  • HOCl and HOBr are often included as ROS, although HOCl is also a RCS.

  • ‡ONOO, O2NOO and ONOOH are often included as ROS, but are also classifiable as RNS.

  • §NO2Cl can also be regarded as a RNS.

  • ¶Oxidizing species formed in polluted air that are toxic to plants and animals. inline image is also produced in vivo by myeloperoxidase and from ONOO (Augusto et al. 2002). Ozone might also be produced in vivo, although the chemistry involved is unclear (Wentworth et al. 2003).

 Superoxide, inline image Hydrogen peroxide, H2O2
 Hydroxyl, OH Hypobromous acid, HOBr*
 Hydroperoxyl, inline image (protonated superoxide) Hypochlorous acid, HOCl
 Carbonate, inline image Ozone, O3
 Peroxyl, inline image Singlet O21Δg
 Alkoxyl, RO Organic peroxides, ROOH
 Carbon dioxide radical, inline image Peroxynitrite, ONOO
 Singlet O21Σg + Peroxynitrate, O2NOO
 Peroxynitrous acid, ONOOH
 Peroxomonocarbonate, HOOCO2
 Nitrosoperoxycarbonate, ONOOCO2
 Atomic chlorine, Cl Hypochlorous acid, HOCl
 Nitryl chloride, NO2Cl§
 Chlorine gas (Cl2)
 Bromine chloride (BrCl)*
 Chlorine dioxide (ClO2)
 Atomic bromine, Br Hypobromous acid (HOBr)
 Bromine gas (Br2)
 Bromine chloride (BrCl)
 Nitric oxide, NO Nitrous acid, HNO2
 Nitrogen dioxide, inline image Nitrosyl cation, NO+
 Nitrate, inline image Nitroxyl anion, NO
 Dinitrogen tetroxide, N2O4
 Dinitrogen trioxide, N2O3
 Peroxynitrite, ONOO
 Peroxynitrate, O2NOO
 Peroxynitrous acid, ONOOH
 Nitronium cation, NO2+
 Alkyl peroxynitrites, ROONO
 Alkyl peroxynitrates, RO2ONO
 Nitryl chloride, NO2Cl
 Peroxyacetyl nitrate, CH3C(O)OONO2

With these concepts in mind, let us look back at the basics and then forward to the future.

Radicals and other reactive species

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

Electrons within atoms and molecules occupy regions of space known as orbitals. Each orbital can hold a maximum of two electrons. For example, the two electrons that form a covalent bond occupy the same orbital, but have opposite spins. If an orbital contains only one electron, that electron is said to be unpaired. A free radical is defined as any species capable of independent existence (hence the term ‘free’) that contains one or more unpaired electrons (Halliwell and Gutteridge 2006). This broad definition encompasses a wide range of species (Table 1). For example, nitric oxide, a key player in the nervous system (Bredt 1999) is a free radical; indeed, the radical properties of NO explain many of its biological effects (Halliwell et al. 1999). The simplest free radical is atomic hydrogen – because it has only one electron, it must be unpaired.

Many free radicals exist in vivo, but I will focus here on the oxygen radicals. Indeed, the diatomic oxygen molecule (O2) qualifies as a radical because it has two unpaired electrons, each located in a different π* antibonding orbital, but both with the same spin quantum number. This parallel spin is one reason for the poor reactivity of O2, despite its powerful oxidizing nature. According to thermodynamics, the complex organic compounds of the human body should immediately combust in the O2 of the air. However, if O2 attempts to oxidize a molecule directly by accepting a pair of electrons from it, both of these electrons must have spins opposite to that of the unpaired electrons in the O2 in order to fit into the vacant spaces in the orbitals. A pair of electrons in the same orbital (e.g. a covalent bond) would not meet this criterion because they have spins opposite to each other. This spin restriction is one factor that slows down the reaction of O2 with non-radicals. Oxygen much prefers to react with radicals, accepting electrons one at a time. Fortunately, non-radical biomolecules way outnumber radical biomolecules in the human body, so O2 combusts us only incrementally and slowly. More reactive forms of O2, the singlet oxygens, can be generated by an input of energy that inverts the spin of one of the unpaired electrons to allow direct reaction with covalent bonds (Foote et al. 1984). The commonest form is singlet O21ΔgO2, in which both unpaired electrons have been moved into one of the π* orbitals. Most singlet O2 in biology is produced by photosensitization reactions (Foote et al. 1984; Girotti and Kriska 2004), not really much of a problem in the dark recesses of the skull. However, singlet O2 can form during lipid peroxidation and contribute to damage because of its ability to readily oxidize lipids, proteins and DNA (Girotti and Kriska 2004; Davies 2005).

More than 80% of the O2 taken up by the human body is used by mitochondrial cytochrome oxidase, which adds four electrons on to each O2 molecule to generate two molecules of water:

  • image

 This enzyme, like most others that use O2, has transition metal ions (in this case iron and copper) at its active sites. Transition metals such as iron, chromium, nickel, manganese, vanadium, copper, cobalt and titanium have variable oxidation states; changing between these states allows them to transfer single electrons and so to facilitate oxidation–reduction reactions (reviewed by Halliwell and Gutteridge 2006). Thus ‘uncaged’ transition metal ions are dangerous because of their powerful ability to promote free radical reactions (Table 2). By contrast, reduction of O2 to 2H2O by cytochrome oxidase proceeds in a stepwise fashion, with various partially reduced forms of oxygen held firmly bound to metal ions within the enzyme and not released into free solution. Thus, cytochrome oxidase does not release reactive oxygen radicals into its surroundings (Babcock 1999).

Table 2.   Role of transition metal ions in converting less reactive to more reactive species
Starting agentMore reactive species produced on addition of metal ionsMetal involvedComment
  1. Adapted from Halliwell and Gutteridge (2006) with the permission of Oxford University Press.

H2O2OH (and possibly reactive oxo-metal species)Fe/Cu/Co/Ni/Cr/VFenton chemistry (Fe); Fenton-type chemistry (the other metals). Manganese (II) cannot directly convert H2O2 to OH
HOClOHFeFe2+ reacts with HOCl to form OH Fe2+ + HOCl [RIGHTWARDS ARROW]Fe(III) + OH + Cl
Lipid peroxidesPeroxyl radicals, alkoxyl radicals, cytotoxic aldehydesFe/Cu and othersSee text
Thiols (R-SH)inline image, H2O2, RS, OH, oxysulfur radicalsFe/Cu and othersOxidation of thiols produces thiyl, oxysulfur and oxygen radicals. Such reactions may contribute to the neurotoxicity of excess cysteine
NAD(P)HNAD(P), inline image, OHFe/Cu and othersNAD(P) radicals reduce O2 to give inline image; copper is especially good at promoting NAD(P)H oxidation
Ascorbic acidOH, possibly inline image, semidehydroascorbate radicalCu (especially good) and othersOxidation of ascorbate produces cytotoxic species
Alloxan, epinephrine, DOPA, dopamine, tetrahydrofolates, 6-hydroxydopamine, other ‘autoxidizable’ compounds (including some plant polyphenols)OH, inline image, carbon-centred or other radicals derived from the toxinFe, Cu, Mn, often others‘Autoxidations’ depend on the presence of transition metal ions
Peroxynitrite (ONOO)NO2+Fe/CuTransition metal ions and some metalloproteins (including CuZnSOD and MnSOD) accelerate nitration of aromatic compounds, by facilitating conversion of ONOO to NO2+, nitronium ion, a potent nitrating agent


If a single electron is supplied to O2, it enters one of the π* antibonding orbitals to form an electron pair there. The product is superoxide radical, inline image, full name superoxide radical anion. (The superscript dot denotes a free radical, as we have aready used for NO. Thus the oxygen molecule should be written inline image but nobody ever does; one of many anomalies in the free radical literature). With only one unpaired electron, superoxide is less ‘radical’ than O2 (with two unpaired electrons), despite its ‘super’ name. Addition of another electron to inline image will give O22–, the peroxide ion, a non-radical (no unpaired electrons left). Addition of two more electrons to O22– eliminates the oxygen–oxygen bond entirely, giving two O2– (oxide ions). In vivo, the two-electron reduction product of O2 is hydrogen peroxide, and the four-electron product is water.

inline image

Superoxide is formed in vivo in a variety of ways. A major source is the activity of electron transport chains in endoplasmic reticulum and, especially, in mitochondria (Fridovich 1989, 1999; Turrens 2003). Some of the electrons passing through these chains ‘leak’ directly from intermediate electron carriers on to O2. Because O2 accepts electrons one at a time, O2•– is formed. The rate of leakage at physiological O2 concentrations is probably < 5% of total electron flow through the chains, but it rises as the O2 concentration is increased (Turrens 2003). Hence, the toxicity of excess O2 is thought to be due (Fridovich 1995) to increased formation of inline image. Faster electron leakage is one source; another is accelerated autoxidations.

Several non-radicals ‘autoxidize’ on exposure to air; examples are epinephrine, norepinephrine, 3,4-dihydroxyphenylalanine (DOPA), dopamine, 6-hydroxydopamine, and thiols such as cysteine. These compounds react slowly with O2 to produce inline image, which then reacts with more of the compound to accelerate the process. Rates of ‘autoxidations’in vitro depend on the amount of transition metal ions in the reaction mixture to catalyze the initial reaction with O2, and it may be that ‘autoxidizable’ compounds would not oxidize at all if metal ion contamination could be removed completely (an impossible task!). For example, manganese ions accelerate the oxidation of catecholamines to produce quinones, semiquinones and oxygen radicals; this has been suggested to explain the degeneration of catecholaminergic neurons that has been reported in miners of manganese-containing ores (Halliwell 1984; Martin 2006), producing the locura manganica syndrome. Autoxidations are also faster at raised O2 concentrations.

Questions of terminology

Many different terms are used in the literature to describe oxygen radicals and related (non-radical) species such as 1O2 and H2O2. I prefer the term reactive oxygen species (ROS), a collective descriptor that includes not only the oxygen radicals but also some non-radical derivatives of O2 (Table 1), such as H2O2 and hypochlorous acid (HOCl). Hence all oxygen radicals are ROS, but not all ROS are oxygen radicals (Table 1). ‘Reactive’ is a relative term; inline imageand H2O2 are highly selective in their reactions with biological molecules, leaving most of them unscathed, whereas OH is vicious enough to attack everything around it, as we shall soon see. Hence some authors prefer the term oxygen-derived species to ROS. Strictly, however, water is an O2-derived species but it is not much of a threat to us unless we drown in it. Another popular collective term is oxidants. However, this can be misleading because inline image, H2O2 and some other ROS can act as both oxidizing and reducing agents in different systems. The term RS has been expanded to include reactive nitrogen species (RNS), reactive chlorine species (RCS), reactive bromine species (RBS) (Table 1) and reactive sulfur species (not considered here).

What damage can RS do?

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

If two free radicals meet, they can join their unpaired electrons to form a covalent bond, e.g. NO reacts very fast with inline image to form a non-radical product, peroxynitrite

  • image

 At physiological pH, ONOO rapidly protonates to peroxynitrous acid, ONOOH, also a non-radical but nevertheless a very reactive agent, able to directly oxidize and nitrate proteins, lipids and DNA (Alvarez and Radi 2003). Peroxynitrous acid can cause additional damage by undergoing homolytic fission to hydroxyl radical

  • image

 Peroxynitrite also reacts with CO2 (Greenacre and Ischiropoulos 2001; Alvarez and Radi 2003; Ischiropoulos and Beckman 2003)

  • image
  • image

 Both inline image (nitrogen dioxide) and inline image (carbonate radical) are powerfully oxidizing radicals – not as bad as OH but still not pleasant (Augusto et al. 2002). Thus any system producing NO and inline image at levels sufficient to allow their reaction can cause biological damage, and this chemistry is known to occur at sites of tissue injury in many human diseases, including neurodegenerative disorders (Halliwell et al. 1999; Greenacre and Ischiropoulos 2001; Ischiropoulos and Beckman 2003; Table 3). A more-important source of OHin vivo is the Fenton reaction (Halliwell and Gutteridge 1984, 1992, 2006)

  • image

 But what about when a free radical meets a non-radical? If the radical is reactive enough it attacks the non-radical, a new radical results and chain reactions may occur. There are several types of reaction.

Table 3.   Neurodegenerative diseases – what they have in common
 PDADALSFreidreich's ataxiaHuntington's diseasePrion diseases
  1. *Dorfin is an E3 ubiquitin ligase, also found in Lewy bodies. CJD Creutzfeldt–Jakob disease. 8OHG, 8-hydroxyguanosine; Nitrotyrosine is formed by attack of several reactive nitrogen species, especially ONOO, on tyrosine residues in proteins (Halliwell et al. 1999). Modified from Halliwell and Gutteridge (2006) with the permission of Oxford University Press.

Mitochondrial dysfunctionComplex I[DOWNWARDS ARROW], αKGDH[DOWNWARDS ARROW]Complex IV[DOWNWARDS ARROW] (some studies), αKGDH[DOWNWARDS ARROW], pyruvate dehydrogenase[DOWNWARDS ARROW]. Hyperphosphorylated tau may damage mitochondria (Reddy and Beal 2005)Complexes I and IV[DOWNWARDS ARROW] (varying reports). Mitochondrial dysfunction very obvious in transgenic mouse models expressing mutant SODs related to FALSFrataxin is a mitochondrial Fe–S protein; levels of complexes I, II, III and aconitase decreasedComplexes II, III[DOWNWARDS ARROW], αKGDH, aconitase[DOWNWARDS ARROW], Mutant huntingtin may bind to and damage mitochondriaMitochondrial defects reported in brains of scrapie-infected mice
Proteasome dysfunctionSpecific genetic defects in this pathway cause inherited PD (Fig. 1). Proteasome proteolytic activities subnormal in sporadic PDProteasome proteolytic activities subnormal. One reason may be the presence of UBB+1, a mutant ubiquitin carrying a 19-amino acid C-terminal extension that is found in affected neurons in AD and Down's syndrome, apparently generated by errors during transcription. This mutant ubiquitin cannot attach itself to an expanding poly- ubiquitin chain, and also appears to inhibit the proteasome (Ciechanover and Brundin 2003)Proteasome activity may be decreased by aggregates of mutant SODs in FALSNo data as yetAggregates (see below) include proteasome subunits and may impair proteasome functionAccumulation of ubiquitinated proteins observed in animal models and CJD brain suggestive of proteasome dysfunction
Abnormal protein aggregatesLewy bodyAmyloid plaques, diffuse amyloid deposits, neurofibrillary tranglesA range of aggregates described, often containing ubiquitin, neurofilaments, dorfin*, etc. in motor cortex and in spinal motor neurons CuZnSOD is a major component of aggregates in FALS caused by SOD1 mutationsFrataxin aggregates in nucleusAggregates containing huntingtin, ubiquitin, HSPS and proteasome subunitsAbnormal protein aggregates. Fragments of PrPc are toxic to neurons in culture
Changes in iron metabolismMore iron in substantia nigraIron accumulates in plaquesIron deposition in dying motor neurons‘Catalytic’ Fe levels might be raised owing to abnormal frataxin, although this has not been shown experimentallyIron deposited in lesionsIron levels reported as raised in affected areas
Oxidative and nitrative damageSubstantia nigra shows decreased GSH and increased HNE, lipid peroxides, isofurans, 8OHdG, 8OHG (a product of oxidative RNA damage), protein carbonyls, 3-nitrotyrosine, cysteinyl-DOPA and cysteinyl-dopamine. CuZnSOD is one protein oxidized in PD brain, another is UCHL1. Decreases in GSHRises in 8OHdG and other base oxidation products in both nuclear and mitochondrial DNA, also rises in 8OHG, protein carbonyls, glutamic and aminoadipic semialdehydes (protein oxidation products), methionine sulfoxide, nitrotyrosine, acrolein, HNE, HO-1 and IPs in brain tissue. Plaques and paired helical filaments contain oxidized, glycated and nitrated proteins, and CSF from AD patients is also reported to contain raised levels of oxidized, nitrated and glycated proteins. One oxidized protein associated with plaques is CuZnSOD. Proteomics reveals that the oxidative protein damage in AD is not random, but highly selective and affects enzymes involved in protein turnover, energy metabolism and control of excitotoxicity. (Sultana et al. 2006). Heavily oxidized proteins in human AD brain include UCHL1 (also oxidized in PD) (Choi et al. 2004), creatine kinase, glutamine synthetase, CuZnSOD and α-enolase. ‘Total’ protein carbonyl levels in various regions of the brain in old mice correlate with impairment of cognitive and motor functions. Similarly, only certain proteins become heavily nitrated in AD brain, including α-enolase, β-actin and triosephosphate isomerase. Remarkably, mRNA oxidation in AD is also selective; only certain mRNAs contain 8OHG and others are unaffected (Shan et al. 2003). Even protein modification by HNE is not random: a glial glutamate uptake transporter is one of the specific proteins attackedRises in 8OHdG, protein carbonyls, HNE, glycoxidation products and 3-nitrotyrosine in affected spinal cordIncreased urinary 8OHdG.Idebenone (a coenzyme Q analogue with anti- oxidant properties) has some therapeutic benefit, as does coenzyme Q plus α-tocopherol (see text)Increased 8OHdG and F2-IPs in CSF. Rises in 8OHdG, increased peroxidation, nitro- tyrosine formation and mitochondrial disfunction are observed in mice overexpressing huntingtinBrains of scrapie-infected mice show raised levels of HNE, F2-IPs and nitro- tyrosine. Levels of F2-IPs in CSF and HNE in brain are elevated in human CJD patients (von Bohlen und Halbach et al. 2004)
  • (i) 
    A radical can add to a non-radical. The adduct must still have an unpaired electron. For example, OH adds to position 8 in the ring structure of guanine in DNA; the initial product is an 8-hydroxy-2′-deoxyguanosine radical. This can have various fates, including undergoing oxidation to the mutagenic lesion 8-hydroxy-2′-deoxyguanosine (8OHdG) (Kasai 2002). Hydroxyl radical also attacks other bases and the deoxyribose sugar in DNA, producing massive damage (Evans et al. 2004).
  • (ii) 
    A radical may be a reducing agent, donating a single electron. Thus the toxicity of paraquat (PQ) to brain and other animal tissues involves its one-electron reduction to paraquat radical cation, by cellular electron transport systems (e.g. cytosolic enzymes and the electron transport chains of mitochondria and endoplasmic reticulum)
    • image
    • image
    • image
     Thus the paraquat ‘catalyzes’inline image formation from O2, swamping the cell with inline image to cause damage (Przedborski and Ischiropoulos 2005).
  • (iii) 
    A radical may be an oxidizing agent, taking a single electron from a non-radical. The non-radical must then have an unpaired electron left behind. For example, OH oxidizes carbonate ion to carbonate radical (Augusto et al. 2002).
    • image
  •  (iv) 
    A radical may abstract a hydrogen atom from a C–H bond. As H has only one electron, an unpaired electron must be left on the carbon. For example, OH can abstract H from a hydrocarbon side-chain of a polyunsaturated fatty acid (PUFA) residue in a membrane.

inline image

Carbon-centred radicals react fast with O2 (like many radicals do, as mentioned earlier) to make peroxyl radicals

inline image

which are reactive enough to oxidize membrane proteins and cholesterol as well as to attack adjacent PUFA side-chains, propagating a chain reaction

inline image

A new C radical is formed to continue the chain, and a lipid hydroperoxide (LOOH) forms. If the PUFA chain is long enough, the peroxyl radicals can whip round and abstract H from the same PUFA, giving cyclic peroxides (Fam and Morrow 2003).

A single initiation event thus has the potential to generate multiple peroxide molecules in a membrane by a chain reaction. The initial H abstraction from a PUFA can occur at different points on the carbon chain. Thus peroxidation of arachidonic acid gives six lipid hydroperoxides as well as cyclic peroxides and other products (reviewed by Halliwell and Gutteridge 2006) including the isoprostanes (IPs) (Fam and Morrow 2003; Morrow 2005). Similarly, eicosapentaenoic acid can give eight hydroperoxides and several cyclic peroxides, whereas docosahexaenoic acid, an important PUFA in brain lipids, can give 10 hydroperoxides plus a large number of cyclic peroxides. If enough peroxyl radicals accumulate, they can self-react to make 1O2, which causes further oxidation.

inline image

A tetraoxide intermediate (inline image) may be involved (Miyamoto et al. 2003).

The overall effects of lipid peroxidation are to decrease membrane fluidity, make it easier for phospholipids to exchange between the two halves of the bilayer, increase the ‘leakiness’ of the membrane to substances that do not normally cross it other than through specific channels (e.g. K+, Ca2+) and damage membrane proteins, inactivating receptors, enzymes and ion channels (reviewed by Halliwell and Gutteridge 2006). Increases in Ca2+ induced by oxidative stress can activate phospholipase A2, which releases arachidonic acid from membrane phospholipids. The free arachidonic acid can then both undergo lipid peroxidation (Farooqui et al. 2001) and act as a substrate for eicosanoid synthesis. Prostaglandin synthesis is intimately linked to lipid peroxidation, because low levels of peroxides accelerate cyclooxygenase action on PUFAs (Smith 2005). The lysophospholipids left behind by the action of phospholipase A2 have mild detergent-like effects and can contribute to membrane disorganization if they accumulate in large amounts. Phospholipase A2 can also cleave oxidized arachidonic acid residues from membranes and may show some degree of selectivity for them, although its degree of ‘preference’ is uncertain (Nigam and Schewe 2000).

Continued oxidation of fatty acid side-chains and released PUFAs, and the fragmentation of peroxides to produce aldehydes and hydrocarbons (see below), eventually leads to loss of membrane integrity. For example, rupture of lysosomal membranes spills hydrolytic enzymes into the rest of the cell. In addition, some end-products of lipid peroxidation have direct damaging effects. These include the IPs, measurement of which is probably the best currently available assay of lipid peroxidation (Morrow 2005). IPs are prostaglandin-like cyclic peroxides formed from PUFAs with at least three double bonds, including linolenic acid, arachidonic acid (F2-IPs), eicosapentaenoic acid (F3-IPs) and docosahexaenoic acid (DHA) (F4-IPs, sometimes called neuroprostanes). Increased formation of IPs is observed in many human (including neurodegenerative) diseases and after exposure of animals to a range of toxins (Fam and Morrow 2003).

Iron and copper ions accelerate lipid peroxidation, by two mechanisms (Halliwell and Gutteridge 1984). First, they convert H2O2 to OH in the Fenton reaction by splitting the O–O bond. In an analogous reaction, they can split lipid hydroperoxides

  • image
  • image

giving alkoxyl (LO) and more peroxyl (LOO) radicals, both of which can abstract H to keep the chain reaction going.

Metal-catalyzed decomposition of peroxides also generates a complex mixture of products, including hydrocarbon gases (e.g. ethane, pentane) and toxic products such as epoxides and aldehydes. An example is the unsaturated aldehyde 4-hydroxy-2-trans-nonenal (HNE), which binds avidly to membrane proteins, inactivating enzymes, ion channels and receptors. It also attaches to DNA, producing mutagenic lesions (Esterbauer et al. 1991; Mark et al. 1997).

Problems of the brain

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

All aerobic cells suffer oxidative damage, yet the mammalian brain is often said to be especially sensitive (Halliwell 1992, 2001). One reason is its high O2 consumption; in adult humans, the brain accounts for only a few percent of body weight, but about 20% of basal O2 consumption. Hence it processes a lot of O2 per unit tissue mass. The discrepancy is even more striking in young children, who have much smaller bodies but not proportionately smaller brains. A major reason for the high O2 uptake is the vast amounts of ATP needed to maintain neuronal intracellular ion homoeostasis in the face of all the openings and closings of ion channels associated with propagation of action potentials and neurosecretion. Thus interrupting mitochondrial function in neurons by toxins, or failing to supply O2 or substrates for energy production, produces rapid damage. In particular, the high Ca2+ traffic across neuronal membranes means that interference with Ca2+ sequestration (e.g. by oxidative stress-dependent damage to plasma membrane Ca2+ exporters, or to Ca2+ pumps in the endoplasmic reticulum) and/or disruption of the ATP supply produces especially rapid rises in intracellular free Ca2+. In addition, some neurons and glia contain transient receptor potential melastatin-related (TRPM)2 cation channels, which rapidly allow Ca2+ to enter when ROS such as H2O2 are present (Fonfria et al. 2005). Rises in Ca2+ can activate neuronal nitric oxide synthase (nNOS) (in neurons containing this enzyme, although, once made, NO can diffuse several cell lengths in the brain), phospholipase A2 and calpains, a family of Ca2+-stimulated proteinases that can attack the cytoskeleton. In some animals calpains can convert xanthine dehydrogenase to the enzyme xanthine oxidase, which can generate inline image and H2O2, but this is probably unimportant in human brain (Linder et al. 1999). Rises in Ca2+ interfere with mitochondrial function, increasing mitochondrial inline image formation. The excess inline image can react with NO to form ONOO. The rise in arachidonic acid coupled with increased lipid peroxidation (Smith 2005) can promote eicosanoid formation (Farooqui et al. 2001; Phillis and O'Regan 2004) and, if prostaglandins are not quickly removed, they can undergo conversion to neurotoxic agents, the cyclopentenone prostaglandins and the levuglandins (Kondo et al. 2002; Musiek et al. 2005; Salomon 2005).

The brain has other problems.

  • (i) 
    The presence of excitotoxic amino acids. Concentrations of glutamate in brain extracellular fluids are normally low (< 1 μm). The death of cells or collapse of normal ion gradients (e.g. owing to severe energy depletion) in neurons can cause massive glutamate release. This binds to receptors on adjacent neurons, leading to excessive and prolonged increases in intracellular free Ca2+ and Na+ within them. Neurons treated with excess glutamate or other excitotoxins swell rapidly and die, usually by necrosis. Oxidative stress can damage neurons and promote the release of excitatory amino acids, generating a ‘vicious cycle’ of events (Mailly et al. 1999). Other relevant events may be the ability of several RS (including ONOO) to decrease glutamate uptake by glial cells and to inactivate glutamine synthetase (Aksenov et al. 1997), preventing conversion of glutamate to glutamine. Hydroxynonenal can readily damage glutamate transporters, slowing glutamate clearance (Mattson and Chan 2003).
  • (ii) 
    Neuronal mitochondria generate inline image (mostly from complex I; Kudin et al. 2005). Levels of 8OHdG, mutations and deletions increase with age in brain mitochondrial DNA.
  • (iii) 
    Several neurotransmitters (not glycine or glutamate) are autoxidizable. Dopamine, its precursor l-DOPA, serotonin and norepinephrine can react with O2 to generate not only inline image, but also quinones/semiquinones that can deplete reduced glutathione (GSH) and bind to protein SH groups (Spencer et al. 1998). Oxidation can be catalyzed by transition metal ions, as mentioned above, but if excess inline image is present it can react with norepinephrine, dopamine and serotonin (Wrona and Dryhurst 1998) to initiate their oxidation, which then continues with production of more ROS, quinones, etc. Dopamine–GSH conjugates are degraded by peptidase enzymes to produce dopamine-cysteine conjugates (e.g. 5S-cysteinyldopamine), which can be detected in several brain regions; levels are raised in Parkinson's disease (PD) (Spencer et al. 1998).
  • (iv) 
    Iron is found throughout the brain (Burdo and Connor 2003; Zecca et al. 2004). Important iron-containing proteins include cytochromes, ferritin, aconitases, non-heme iron proteins in the mitochondrial electron transport chain, cytochromes P450, and tyrosine and tryptophan hydroxylases. There are about 60 mg of non-heme iron in the ‘average’ adult human brain. Several brain areas (e.g. substantia nigra, caudate nucleus, putamen, globus pallidus) have a high iron content, which can be detected by magnetic resonance imaging (Schenck and Zimmerman 2004). Much iron in healthy brain is in ferritin, with some in hemosiderin. The iron content of the brain is low at birth but rapidly increases in early life to reach a maximum at about 30 years. If insufficient dietary iron is available to babies and children there may be permanent impairment of brain function. Transferrin delivers most of the required iron across the blood–brain barrier, utilizing receptors located on the brain microvasculature (Burdo and Connor 2003). However, there is a problem; damage to brain readily releases iron (and copper) ions in forms capable of catalyzing free radical reactions (Halliwell 2001; Table 2). ‘Catalytic’ iron released by brain damage can persist because CSF has little or no iron-binding capacity (Gutteridge 1992).
  • (v) 
    Neuronal membrane lipids are rich in highly polyunsaturated fatty acid side-chains, especially DHA (C22 : 6) residues. Homogenization of isolated brain tissue causes rapid lipid peroxidation, which can be largely inhibited by iron-chelating agents such as desferrioxamine (Halliwell and Gutteridge 1997). In addition, products of lipid peroxidation can injure the brain. 4-Hydroxynonenal is especially cytotoxic to neurons, increasing Ca2+ levels, inactivating glutamate transporters and damaging neurofilament proteins (Mark et al. 1997; Ong et al. 2000b). It can also inactivate α-ketoglutarate dehydrogenase (αKGDH), a key enzyme of the tricarboxylic acid cycle (Sheu and Blass 1999). IPs may act as vasoconstrictive agents in brain (Hou et al. 2004) and can damage developing oligodendrocytes in premature babies (Back et al. 2005). Other products of the IP pathway may also be neurotoxic, e.g. by damaging the proteasome (Halliwell 2006).
  • (vi) 
    Brain metabolism generates a lot of H2O2, not only via superoxide dismutases (SODs) (see below) but also by other enzymes. Especially important are monoamine oxidases A and B, flavoprotein enzymes located in the outer mitochondrial membranes of neurons and glia. They catalyze the reaction
    • image
    and generate substantial H2O2 in the brain (Gal et al. 2005). The ammonia is disposed of by several mechanisms, including its use by glutamine synthetase.
  • (vii) 
    Antioxidant defenses are modest. In particular, catalase levels are low in most brain regions; levels are somewhat higher in hypothalamus and substantia nigra than in cortex or cerebellum (reviewed by Halliwell 2001). Brain catalase is located in small peroxisomes (microperoxisomes) and its activity in rat or mouse brain is rapidly inhibited if aminotriazole is administered to the animals. This agent acts only on catalase complex I, confirming that brain generates H2O2in vivo and that at least some of it reaches catalase (Sinet et al. 1980). The catalase probably cannot deal with H2O2 generated in other subcellular compartments.
  • (viii) 
    Some glia are microglia, resident macrophage-type cells that arise from monocytes entering the brain during embryonic development. Normally, they help clear cellular debris (including apoptotic cells) and are alert for threats to neurons (Nimmerjahn et al. 2005). However, microglia can become activated to produce inline image, H2O2 and cytokines such as interleukin-1, interleukin-6 and tumor necrosis factor α. In turn, such cytokines can cause microglia to generate more ROS upon activation and to produce inducible nitric oxide synthase (iNOS) and hence excess NO. Cytokines can additionally be produced by activated astrocytes, which may again respond to them by iNOS induction. Thus microglia and astrocytes are major players in inflammatory processes in the brain (Duncan and Heales 2005).
  • (ix) 
    Cytochromes P450 (CYPs) are present in some brain regions (Miksys and Tyndale 2004). For example, CYP46 metabolizes cholesterol, and CYP2D6 is present in several human brain regions (Miksys et al. 2002). CYP2E1 is also found. Because CYP2E1 leaks electrons readily during its catalytic cycle, it produces more ROS than most other CYPs (Gonzalez 2005). It is thus another potential source of oxidative stress. The magnitude of this may be small in normal brain because CYP2E1 levels are low. However, CYP2E1 metabolizes ethanol, acetone, halothane, related anesthetics and organic solvents such as CCl4 and CHCl3, and its levels may be increased in human brain by ethanol and smoking. Thus it could contribute to solvent neurotoxicity (Howard et al. 2003). Brain CYP2D6 levels are also raised in human alcoholics (Miksys et al. 2002).
  • (x) 
    RS, both directly (e.g. by decreasing the synthesis of proteins involved in tight junctions between cells; Krizbai et al. 2005) and/or by activation of matrix metalloproteinases, can contribute to ‘opening up’ the blood–brain barrier, allowing neurotoxins, endotoxin and inflammatory cells to enter the brain (Kim et al. 2003; Savaraj et al. 2005).
  • (xi) 
    Like many other cells, neurons contain polyADP-ribose polymerase 1 (PARP-1), an enzyme that responds to DNA damage by cleaving NAD+ and attaching ADP-ribose residues to nuclear proteins to facilitate DNA repair. Overactivation of PARP-1 can kill cells by depleting NAD+, preventing energy production (Koh et al. 2005) and is involved in opening TRPM2 Ca2+ channels. Indeed, NAD+ is neuroprotective; NAD+ added to neurons can slow axonal degeneration, an effect that seems to require the sirtuin, silent information regulator-like protein 1 (SIRT1), (Araki et al. 2004). Sirtuins are NAD+-dependent protein deacetylase enzymes, intimately involved in the regulation of gene expression and of lifespan (Guarente 2005).
  • (xii) 
    Loss of trophic support can lead to oxidative stress and apoptosis in neurons, in part by activation of neuronal NADPH oxidase enzymes. NADPH oxidase (NOX) enzymes were first detected in phagocytes, but are now known to be widespread in animal tissues, seemingly producing inline image for defense and/or signaling purposes (Krause 2004). Neuronal NOX enzymes may promote necessary apoptosis during development of the nervous system but, if trophic support is lost in the developed brain, they may be activated inappropriately, leading to neuronal death (Sanchez-Carbente et al. 2005).
  • (xiii) 
    Hemoglobin is neurotoxic. This protein is normally safely transported in erythrocytes, which are rich in antioxidant defense enzymes. However, isolated hemoglobin is degraded on exposure to excess H2O2, with release of pro-oxidant iron ions from the heme ring (Gutteridge 1986; Puppo and Halliwell 1988). Heme can also be released and is a powerful promoter of lipid peroxidation (Chiu et al. 1996) by decomposing peroxides to peroxyl and alkoxyl radicals (Phumala et al. 2003). In addition, hemoglobin reacts with H2O2 and other peroxides to form oxidizing species (heme ferryl and various amino acid radicals) capable of stimulating lipid peroxidation. Hemoglobin also binds NO avidly, producing vasoconstriction (Alayash 2004). Both NO binding and oxidative damage are important in the vasoconstriction that can sometimes follow bleeding in the brain. For example, IP levels were higher in the CSF of patients with subarachonoid hemorrhage who suffered vasospasm that in those who did not (Asaeda et al. 2005).

Defending the brain

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

One important defense may be to keep brain oxygen levels as low as possible consistent with normal function. Low O2 decreases autoxidation reactions, mitochondrial inline image production and oxidase activities. Of course, keeping pO2 low has a disadvantage; it renders the brain sensitive to interruptions of the O2 supply. The presence in neurons of a monomeric O2-binding heme protein (neuroglobin) has been hypothesized to assist correct O2 delivery to mitochondria (Burmester and Hankeln 2004). This protein seems far less pro-oxidant in the presence of H2O2 than is hemoglobin (Herold et al. 2004).

SODs and peroxide-removing enzymes

All parts of the nervous system contain SODs, enzymes that remove inline image by catalyzing its dismutation, one inline image being reduced to H2O2 and another oxidized to O2 (Fridovich 1989, 1995; Halliwell 2001; Liochev and Fridovich 2005)

  • image

 Animals have SODs containing active-site manganese (MnSOD) in the mitochondrial matrix, plus SODs with copper and zinc (CuZnSOD) in the mitochondrial intermembrane space and in the rest of the cell. Interestingly, nNOS-containing neurons are fairly resistant to excitotoxicity; the mechanisms underlying resistance are uncertain, but one of them may be a high level of MnSOD (Gonzalez-Zulueta et al. 1998).

SODs must work together with enzymes that remove H2O2. We have already mentioned catalases, but they are not very important to the brain, and are not present in brain mitochondria, where much inline image is generated (Turrens 2003). Until recently, it was thought that the most important H2O2-removing enzymes in brain and other animal tissues are the glutathione peroxidases (GPx), a family (Brigelius-Flohe 1999) of selenium-containing enzymes. They remove H2O2 by coupling its reduction to H2O2 with the oxidation of GSH, a thiol-containing tripeptide (Glu-Cys-Gly).

  • image

 The product, oxidized glutathione (GSSG), consists of two GSH linked by a disulfide bridge, and can be converted back to GSH by glutathione reductase enzymes. Glutathione peroxidases can additionally act on peroxides other than H2O2, e.g. they catalyze GSH-dependent reduction of fatty acid hydroperoxides to alcohols (Brigelius-Flohe 1999).

  • image

In recent years, it has been realized that peroxiredoxins may be the most important H2O2-removal systems in animals (Rhee et al. 2005). They are a family of peroxidases that reduce H2O2 and organic peroxides. They are homodimers and contain no prosthetic groups: the redox reactions are dependent on cysteine at the active sites. There are at least three classes in animals; the typical 2-Cys (the most common), the atypical 2-Cys and the one-Cys peroxiredoxins. In all cases, the peroxide oxidizes an SH group on the peroxiredoxin to a sulfenic acid, Cys-SOH. In the 2-Cys enzymes, this sulfenic acid reacts with another SH on the protein to give a disulfide that is then reduced by thioredoxin, a small protein with two SH groups. In typical 2-Cys enzymes, the second thiol is on the other subunit, in atypical 2-Cys on the same subunit. In 1-Cys peroxiredoxins, it is not yet clear which cellular reductant regenerates the SH group; it is not thioredoxin. Thioredoxin plays many other metabolic roles, including helping to repair oxidatively damaged proteins (see below).

Peroxiredoxins are generally slower at catalyzing H2O2 removal than glutathione peroxidases, although the large amounts of peroxiredoxins present (up to 0.8% of total soluble protein in some animal cells) and their low Km for H2O2 (< 20 μm) can compensate for this, as can their presence in all subcellular organelles and in the cytosol (Rhee et al. 2005). One seemingly odd (at first sight) aspect of peroxiredoxins is that they are readily inactivated by H2O2, the essential cysteine being oxidized to a sulfinic acid, Cys-SO2H. Apparently even more odd, the eukaryotic peroxiredoxins are more susceptible to oxidation than bacterial ones (Georgiou and Masip 2003). This is probably because H2O2 is involved in many intercellular and intracellular signaling processes (reviewed by Rhee et al. 2005; Halliwell and Gutteridge 2006). At low (physiological) H2O2 concentrations, peroxiredoxins seem to dispose of much of the H2O2 generated in animal cells. However, when the cell is exposed to extra H2O2 (e.g. from phagocytes at a site of injury or inflammation, or when NOX enzymes in other cells are activated), the peroxiredoxins are partially inactivated to allow signaling and cellular responses. The cell then quickly makes more peroxiredoxin, and reactivates the inactive form, so that the extra H2O2 can be removed after it has done its job (Rhee et al. 2005; Georgiou and Masip 2003). However, too much peroxide can cause excessive peroxiredoxin inactivation and damage neurons. Damage can involve direct effects of H2O2 in inactivating sensitive enzymes as well as its reaction with iron ions to form OH (Halliwell 2001).

Animals require selenium for the catalytic function of both GPx enzymes and thioredoxin reductase, an NADPH-dependent enzyme that converts oxidized thioredoxin back to the dithiol form. The plasma protein selenoprotein P is important in delivering selenium to the brain; mice lacking this protein die with severe degeneration of motor neurons unless their diet is enriched with selenium (Valentine et al. 2005). Brain is a ‘privileged organ’ with respect to selenium, efficiently retaining it and maintaining GPx and thioredoxin reductase activities even under conditions of severe dietary selenium deficiency (Valentine et al. 2005).

All parts of the brain and nervous system contain CuZnSOD, MnSOD, GSH (at millimolar levels), glutathione peroxidases and the thioredoxin/peroxiredoxin system (Halliwell 2001; Patenaude et al. 2005). It is generally thought that neuronal GSH levels are lower than in glia, and that glia may assist neurons by supplying them with cysteinyl-glycine (as a GSH precursor) and efficiently degrading extracellular H2O2. For example, astrocytes may release GSH, which is degraded by γ-glutamyltransferase on their cell surface to produce cysteinyl-glycine, which neurons then further cleave to release cysteine for uptake and use in GSH synthesis (Dringen et al. 2005). In general, glia appear less susceptible to RS (including H2O2 and ONOO), tend to have higher GSH levels and are more able to accelerate GSH synthesis under stress than neurons. However, many of these conclusions are based on studies of isolated cells, and the cell culture and isolation processes are themselves a ‘stress’ (Halliwell 2003) that could deplete neural GSH but enhance glial GSH levels. Indeed, in a study on rat brain using antibodies to detect total glutathione, in vivo levels appeared higher in neurons than in glia. Injection of the excitotoxin kainate depleted levels in neurons while raising those in astrocytes (Ong et al. 2000a). There appears to be a close link between glutathione metabolism and the regulation of anxiety, at least in mice (Hovatta et al. 2005).

Scavengers of RS

The brain is enriched in several low molecular mass antioxidants in addition to GSH, especially ascorbate (vitamin C). Ascorbate concentrations in human CSF are higher than in plasma, and neurons and glia have transport systems that concentrate ascorbate even more, to millimolar intracellular levels (Rice 2000). Neurons readily take up ascorbate, whereas astrocytes take up dehydroascorbate and convert it to ascorbate intracellularly (Rice 2000). Indeed, it has been proposed that neurons may release dehydroascorbate for ‘recycling’ back to ascorbate by astrocytes (Swanson et al. 2004). Levels of ascorbate in CSF and brain remain high even when the plasma level decreases, indicative of the importance of ascorbate to CNS function. Key roles include its involvement in dopamine hydroxylation, collagen synthesis and formation of the myelin sheath (Passage et al. 2004). Mice lacking sodium–vitamin C transporter 2 show severely decreased ascorbate levels in the blood, brain and other tissues, indicating that this may be the most important transporter for brain ascorbate uptake. These mice die within a few minutes of birth with respiratory failure and brain hemorrhage (Sotiriou et al. 2002), indicating that ascorbate is essential in the lung and brain to cope with ‘birth hyperoxia’ (sudden exposure to 21% O2, much higher than intrauterine O2 levels).

However, if iron or copper ions become available, e.g. in damaged brain, ascorbate could conceivably stimulate oxidative damage by reducing Fe(III) and Cu2+ ions to the Fe2+ and Cu+ forms, which are more active in making OH from H2O2 and in decomposing lipid peroxides (Halliwell and Gutteridge 1990). By contrast, ascorbate can inhibit damage by heme protein/peroxide mixtures by reacting with and removing ferryl and amino acid radicals (Rice-Evans et al. 1989). Hence the net effect of ascorbate at sites of CNS injury is hard to predict. Administration of dehydroascorbate to mice decreased neuronal damage in one stroke model, suggesting that overall (at least in this model) ascorbate is beneficial (Huang et al. 2001).

‘Vitamin E’ describes a family of four tocopherols and four tocotrienols that inhibit lipid peroxidation by scavenging peroxyl radicals using a phenolic OH group (Parks and Traber 2000)

  • image

 The Toc-O radical is poorly reactive, and may be removed using ascorbate (although it has been hard to demonstrate this in vivo).

  • image

 In the brain, α-tocopherol seems to be the main or only form of vitamin E present (Muller and Goss-Sampson 1990; Roy et al. 2002; Hayton and Muller 2004), although γ-tocopherol has been detected in human CSF (Vatassery et al. 2004).

Much brain α-tocopherol is derived from plasma high-density lipoprotein. Severe and prolonged deprivation of α-tocopherol, as occurs in some inborn errors of metabolism that affect fat absorption (and hence uptake of fat-soluble vitamins such as E from the diet) or processing of α-tocopherol, produces neurological damage. In general, the CNS is more severely affected than the peripheral nervous system, and sensory axons more than motor ones. It may take a year on a tocopherol-deficient diet to decrease brain α-tocopherol levels to < 3% of control in animals. Similarly, it takes many weeks to increase the α-tocopherol content of brain in adult mammals supplemented with this vitamin (Muller and Goss-Sampson 1990). Hence brain α-tocopherol levels seem tightly regulated. This point is relevant when assessing the therapeutic effects of α-tocopherol supplementation in clinical trials: large doses need to be given for long periods to raise brain levels significantly. Levels of α-tocopherol appear normal in the brains of centenarians, patients with Alzheimer's disease (AD) or PD, and in fetuses with Down's syndrome. High doses of α-tocopherol have been reported in some studies to improve cognitive function in aged rodents (Martin et al. 2002) and in patients with AD, although the extent of the benefit is uncertain and, sadly, there was no beneficial effect (Blacker 2005) in patients with mild cognitive impairment (MCI). Cellular effects of α-tocopherol are not necessarily related to its antioxidant ability; it has other metabolic actions (Zingg and Azzi 2004).

Coenzyme Q may also be important (Beal and Shults 2003). Defects in the ability to make it cause encephalopathy (Van Maldergem et al. 2002) and oral administration of coenzyme Q to rats was reported to increase its levels in the brain and to protect against striatal lesions induced by malonate or 3-nitropropionic acid (inhibitors of complex II in mitochondria). Coenzyme Q also protects against 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP) toxicity in mice and may have some therapeutic benefit in Huntington's disease and Freidreich's ataxia (Beal and Shults 2003; Hart et al. 2005). Neuroprotective effects could involve its electron transport action in mitochondria and/or the antioxidant action of ubiquinol (CoQH2), which can scavenge peroxyl radicals and also convert Toc-O back to Toc-OH. In one study, the learning ability of old mice was improved by feeding a mixture of coenzyme Q and α-tocopherol, but not by either agent alone (McDonald et al. 2005).

Despite much discussion in the nutritional literature about possible brain-protecting effects of carotenoids and flavonoids, there is limited evidence for their importance as antioxidants in the CNS. No carotenoids were detectable in monkey, human or rat brain tissues even after dietary supplementation (Serbinova et al. 1992; Stahl et al. 1992; Leung et al. 2001). However, there is evidence that some phenolics can cross the blood–brain barrier, with several animal studies claiming that monophenols, for example ferulic acid, and flavonoids, such as epigallocatechin gallate, hesperitin and naringenin, can enter (Zbarsky et al. 2005; Spencer et al. 2004; Mandel et al. 2005). Some (but not all) reports suggest a protective effect of orally administered Ginkgo biloba extracts against the development of dementia. Such effects (if real) are often attributed to the flavonoids present, which include myricetin and quercetin. Mice whose diets were supplemented with Gingko showed changes in neuronal gene expression, suggesting that something had entered the brain (Watanabe et al. 2001). However, polyphenols are not only powerful antioxidants but can exert many other biological effects (Halliwell et al. 2005). These other actions, rather than antioxidant activity, may account for some or all of their neuroprotective actions (Mandel et al. 2005).

Transition metals and the brain: diminishing harm

The copper-containing protein ceruloplasmin catalyzes oxidation of Fe2+ to Fe(III) without release of any ROS, and is often regarded as an antioxidant protein (Gutteridge 1983). Although largely synthesized in the liver, ceruloplasmin is additionally found in brain, largely in astrocytes, being anchored to glycosylphosphatidylinositol on the membrane. It plays an important role in brain iron metabolism (Jeong and David 2003); the inborn disease aceruloplasminemia is associated with degeneration of the retina and basal ganglia and increased lipid peroxidation (Miyajima 2003). Prion protein may be involved in brain copper metabolism, possibly helping to bind copper in a non-redox-active form (Jones et al. 2005).

Haptoglobin is present in CSF, and may bind some hemoglobin released as a result of bleeding in a ‘safe’, non-redox-active, form (Gutteridge and Smith 1988). Brain also contains metallothioneins, small proteins rich in cysteine residues, including a ‘brain-specific’ metallothionein III isoform. Metallothioneins are major stores of intracellular zinc and also contain some copper. Transgenic mice lacking metallothioneins I and II (but oddly, not mice lacking metallothionein III) showed impaired repair of brain damage after cortical injury (Carrasco et al. 2003). Excess zinc is neurotoxic, and its release in excess (from metallothioneins and elsewhere) may contribute to damage during stroke and in excitotoxicity generally (Capasso et al. 2005). Finally, various histidine-containing dipeptides (e.g. carnosine) are present at high levels in brain and have been postulated to exert antioxidant effects (Marchis et al. 2000), for example by chelating transition metal ions and binding cytotoxic aldehydes produced during lipid peroxidation (Aruoma et al. 1989; Decker et al. 2000).

Heme oxygenase (HO), an enzyme that degrades heme with production of CO and biliverdin (which is then converted to bilirubin) is widespread in the brain: both constitutive (HO-2) and inducible (HO-1) forms have been detected. Indeed, CO may be a neurotransmitter. The levels of HO-1 increase after ischemia–reperfusion, bleeding in the brain, and in AD and some other neurodegenerative diseases (Calabrese et al. 2005; Chang et al. 2005). Is HO good or bad? Degradation of heme removes one potential pro-oxidant but causes release of another, iron ions. Overall, HO seems good; transgenic mice overexpressing HO-1 were less sensitive to ischemic brain injury, an effect that involved increased ferritin synthesis secondary to the increased HO. Bilirubin or biliverdin produced by heme degradation might have some antioxidant activities (Van Bergen et al. 1999), although bilirubin breakdown by ROS might produce vasoconstricting compounds (Pyne-Geithman et al. 2005) and too much bilirubin is neurotoxic (Lin et al. 2005).

Coping with oxidative damage in the brain

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

Enzymes that can repair damaged (including oxidatively damaged) DNA are present in neuronal nuclei and mitochondria (Englander and Ma 2006), and seem essential for normal function. For example, inhibition of uracil-DNA glycosylase in cultured rat hippocampal neurons increased levels of DNA damage and led to p53-dependent apoptosis (Kruman et al. 2004). Indeed, the disease spinocerebellar ataxia with axonal neuropathy-1 may result from defects in the repair of single-strand breaks induced by ROS or other agents (El-Khamisy et al. 2005). Oxidatively damaged lipids can be cleaved from membranes by phospholipase A2 and destroyed (Nigam and Schewe 2000).

Several mechanisms are available to deal with oxidatively damaged proteins. If methionine residues have been oxidized by RS to methionine sulfoxide (a frequent fate of this amino acid), the damage can be ‘repaired’ by methionine sulfoxide reductase (Msr) enzymes, which convert methionine sulfoxide in peptides or proteins back to methionine and thus reactivate proteins inhibited by methionine oxidation. Methionine sulfoxide has two stereoisomers, and reductases specific for each form (R or S) have been identified; MsrA (sometimes called MsrS) acts on the S-isomer and MsrB (MsrR) is specific for the R-isomer. The reducing power is supplied by thioredoxin. MsrA has an essential cysteine residue at its active site. In some animal MsrB enzymes, a selenocysteine plays the same role (Moskovitz and Stadtman 2003). Both MrsA and B are found in brain; MsrA-deficient mice show a peculiar neurological phenotype, ‘tip-toe walking’ (Moskovitz and Stadtman 2003).

However, often proteins are damaged by RS not only at methionine but at multiple sites, and are then ‘marked’ for proteolytic removal. This is necessary because accumulation of proteins with incorrect conformation can lead to cell death (Bence et al. 2001), especially if it occurs in the endoplasmic reticulum (Rao et al. 2006). By contrast, heavily oxidized proteins may resist proteolytic attack and form aggregates (Grune et al. 2004). Aggregation and precipitation may sometimes decrease their toxicity by sequestering them in insoluble clumps (Bence et al. 2001; Orr 2004).

There are several ways to destroy unwanted proteins. One is to dispatch them to the lysosomes, which contain hydrolytic enzymes, such as cathepsins. Lysosomes degrade proteins taken into cells by endocytosis, and also take up some cytoplasmic proteins and organelles (e.g. mitochondria) for destruction (autophagy). The devastating pathology of the neuronal ceroid lipofuscinoses, inborn defects in lysosomal function, illustrates the importance of this pathway to the nervous system (Ezaki and Kominami 2004). Another system is the Lon proteinase, an ATP-stimulated mitochondrial proteinase that degrades aconitase and several other mitochondrial proteins after they have become oxidized. It recognizes hydrophobic patches that appear on the surface of aconitase after oxidation (Bota and Davies 2002; Higgins et al. 2004).

However, most attention has been paid to the proteasome, a cytoplasmic and nuclear system in eukaryotic cells that removes unwanted proteins, including iKB, hypoxia-inducible factor-1α, phosphorylated catalase, cyclins, Keap1 and oxidatively damaged proteins (McNaught et al. 2001; Ciechanover and Brundin 2003; Keller et al. 2004; Stefanis and Keller 2006). The core of the system is the 20S proteasome, named for its sedimentation coefficient when spun in an ultracentrifuge. It is a cylindrical structure made of a stack of four rings, each containing seven protein subunits. The cylinder has three internal cavities: the central cavity contains proteolytic sites and access to it is controlled by gates from two outer cavities, too narrow to admit folded proteins. The chymotrypsin-like proteolytic sites cleave proteins adjacent to hydrophobic amino acid residues (tyrosine or phenylalanine), the trypsin-like sites after basic residues (arginine, lysine) and the peptidylglutamyl sites after acidic residues (aspartate, glutamate). In vivo, however, the 20S proteasome is often assembled into a larger structure, the 26S proteasome; the 20S cylinder plus additional groups of proteins (19S cap complexes) attached at each end (reviewed in Stefanis and Keller 2006).

The 26S proteasome recognizes many of its targets (e.g. cyclins and iKB) because they have been marked for degradation by attachment of ubiquitin, a 76-amino acid heat-shock protein (Hsp). Ubiquitination (attachment of multiple ubiquitin molecules) is ATP dependent. It is catalyzed by a group of enzymes that first attach ubiquitin to lysine residues on the target and then add more ubiquitins to lysines on the ubiquitin itself (Fig. 1). A chain of at least four ubiquitins is usually needed to ‘mark’ a protein for proteasomal degradation. Ubiquitin-activating enzymes (E1) use ATP to activate ubiquitin, which is transferred to several ubiquitin-conjugating proteins (E2). These cooperate with hundreds of ubiquitin ligases (E3), each of which recognizes ‘destroy me’ signals on its target protein and attaches ubiquitin. These E3 enzymes are largely responsible for the specificity with which proteins are targeted for degradation by the proteasome. Various de-ubiquitinating enzymes can remove ubiquitin residues and rescue proteins. If this does not happen, the cap structures recognize the polyubiquitinated protein, unfold it and feed it slowly into the proteolytic core using energy from ATP hydrolysis. Once inside, it is hydrolyzed into short peptides. Simultaneously, polyubiquitin is removed, disassembled back into monomers by enzymes, including ubiquitin C-terminal hydrolase L1 (UCHL1), and the monomeric ubiquitin made available for the cycle to repeat. The 26S proteasome thus catalyzes an ATP-dependent degradation of ubiquitinated proteins (reviewed in Ciechanover and Brundin 2003; Stefanis and Keller 2006). The 20S proteasome will not attack ubiquitinated proteins; recognition requires the cap complexes.


Figure 1.  Defects in the ubiquitin–proteasome system in PD. The blue section shows the normal ATP-dependent identification and labeling of unwanted proteins with ubiquitin molecules (ubiquitination) as a signal for ATP-dependent degradation by the 26S proteasome complex (proteolysis; red section). The green section shows recovery and recycling of ubiquitin molecules that are released from proteins. Also depicted are ways in which potential defects in the system cause PD. UCHL1, one of the most abundant proteins in brain (up to 2% of total brain protein), releases free ubiquitin and allows the cycle to continue (von Bohlen und Halbach et al. 2004). A variety of deletion and point mutations in the parkin gene can lead to PD. From McNaught et al. (2001) with the permission of Professor Peter Jenner and Nature publishers.

Download figure to PowerPoint

Several inhibitors of the proteasome have been described. For example the Streptomyces metabolite lactacystin inactivates the proteasome and is cytotoxic to dividing cells by interfering with cyclin-dependent regulation of the cell cycle. Other inhibitors include epoxomycin, aclacinomycin A and MG132 (a synthetic inhibitor). Proteasome inhibitors such as bortezomib (Velcade®) are under test for cancer chemotherapy (Mitsiades et al. 2005). Proteasome inhibitors induce oxidative stress in cells (Lee et al. 2001; Demasi and Davies 2003; Ding et al. 2004), and can cause apoptosis. Paradoxically, however, adding low levels of proteasome inhibitors can sometimes produce cytoprotection by increasing cellular production of HSPS and sometimes of the proteasome itself (Yew et al. 2005; Stefanis and Keller 2006).

Lowering proteasomal activity (by chemical inhibitors or antisense oligonucleotides) increases levels of oxidized proteins in cells, suggesting that their degradation proceeds via this system (Lee et al. 2001; Grune et al. 2003, 2004). However, the role of ubiquitin is uncertain; the 20S proteasome alone seems able to recognize and degrade at least some oxidized proteins, but it is unclear how this happens. It possibly recognizes partial unfolding of oxidized proteins to expose hydrophobic surface patches that somehow ‘open’ the gates and allow the oxidized proteins to enter in the absence of ubiquitin (Grune et al. 2003, 2004). Another possibility is that HSPS such as Hsp90 recognize oxidized proteins and deliver them to the 20S proteasome (Whittier et al. 2004).

Proteasome activity may decline with age, favoring accumulation of oxidized proteins (Grune et al. 2003, 2004; Stefanis and Keller 2006). In addition, some oxidatively damaged proteins appear to inhibit proteasome function rather than being degraded by it. These include HNE-modified proteins (Grune and Davies 2003) and proteins modified by isoketals (products of the IP pathway) or cyclopentenone prostaglandins; a single isoketal adduct slowed degradation by about 50% (Davies et al. 2002). Thus these end-products of lipid oxidation could lead to further cell dysfunction via impaired proteasome activity.

The neurotoxicity of superoxide

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

It was noted earlier that many believe that excess production of inline image causes O2 toxicity. Indeed, transgenic animal experiments show that SODs (especially MnSOD) are very important enzymes in animals. In one study (Lebovitz et al. 1996), most mice lacking MnSOD died within 10 days after birth, with cardiac abnormalities, fat accumulation in liver and skeletal muscle, metabolic acidosis and severe mitochondrial damage in heart and, to a lesser extent, in other tissues. There were decreases in complex I, succinate dehydrogenase and aconitase activities, and increases in oxidative damage to mitochondrial DNA (Hinerfeld et al. 2004). Even MnSOD+/– heterozygous mice, which at first seem normal, show increased mitochondrial oxidative damage as they age, as well as more nuclear oxidative DNA damage and increased cancer rates (Van Remmen et al. 2003). Mice completely lacking MnSOD that manage to survive longer than 10 days soon succumb to a variety of pathologies, including severe anemia, retinal defects and neurodegeneration. Treatment of MnSOD– mice with low molecular mass scavengers of inline image keeps them alive for longer, whereupon they suffer even more retinal damage and brain degeneration, perhaps because the scavengers do not enter the brain (Melov et al. 1998).

Removal of mitochondrial inline image is therefore essential in the brain and elsewhere. Why? Superoxide in aqueous solution does not react with many biomolecules, but the few that it does attack are very important. First, its reaction with NO to give ONOO is very fast. Second, inline image inactivates several enzymes important in energy production and amino acid metabolism (Imlay 2003; Liang and Patel 2004). Some of these enzymes have iron–sulfur clusters and their inactivation is caused by oxidation of the cluster, leading to release of iron to promote Fenton chemistry and lipid peroxidation

  • image
  • image
  • image

 The oxidized enzymes can be ‘repaired’in vivo by reassembling the iron clusters. Superoxide can also release iron from ferritins (Paul 2000), and we have already seen that degradation of heme proteins by H2O2 can release iron (Gutteridge 1986). In addition, peroxynitrite can displace iron from iron–sulfur proteins and copper from some copper-containing proteins such as ceruloplasmin (Swain et al. 1994).

Oxidative stress and neurodegenerative diseases: some general concepts

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

The various neurodegenerative diseases have different symptoms, affect different parts of the brain, and have different causes. Many of them are examined extensively in this volume and others in excellent reviews elsewhere, so I will not re-present them in detail. Instead, I will focus on what they have in common (Table 3); impaired mitochondrial function (Reddy and Beal 2005; Zeevalk et al. 2005), increased oxidative damage (Halliwell 1992, 2001; Jenner 2003), defects in the ubiquitin–proteasome system (Stefanis and Keller 2006; McNaught et al. 2001; Ciechanover and Brundin 2003; Keller et al. 2005), the presence of abnormal, aggregated proteins (Bence et al. 2001), changes in iron metabolism, and some involvement of excitotoxicity and of inflammation. Oxidative damage is manifested as increases in lipid peroxidation end-products, DNA (and often RNA) base oxidation products and oxidative protein damage (Halliwell 2001, 2002; Moreira et al. 2005; Sultana et al. 2006). The protein aggregates frequently contain proteins that are nitrated (Ischiropoulos and Beckman 2003), bear carbonyl residues, have attached aldehydes such as HNE or acrolein and, sometimes, carry advanced glycation end-products (AGE products) (Table 3).

I believe (Halliwell 2001, 2002, 2006) that all these events constitute a vicious cycle, and any one of them could initiate neuronal cell death, rapidly recruiting the others to its evil purpose (Fig. 2). The effects of mutations in the ubiquitin–proteasome system, together with the finding that UCHL1 levels are decreased even in sporadic PD (Choi et al. 2004), suggest that all the events shown in Fig. 1 are important. Indeed, this fall in UCHL1 activity involves oxidative damage, because the protein shows raised levels of carbonyls and methionine sulfoxide (Choi et al. 2004). Mice lacking UCHL1 show widespread neurodegeneration, formation of protein aggregates and increased oxidative damage (Castegna et al. 2004). Damage to mitochondria [e.g. by neurotoxins such as 1-methyl-4-phenylpyridinium ion (MPP+) or rotenone that target them] generates more ROS from the electron transport chain and causes oxidative damage that modifies proteins and other biomolecules. Thus, in some studies, treatment of rats or monkeys with low-dose rotenone over long periods produces PD-like symptoms and neurodegeneration accompanied by oxidative damage, nitrotyrosine formation, and generation of protein aggregates containing α-synuclein (Moore et al. 2005). Unlike MPP+, rotenone does not concentrate in dopaminergic neurons in this region, yet it can still induce fairly selective neurodegeneration in the substantia nigra. It follows that the neurons here may be especially sensitive to inhibition of complex I. Another clue pointing to a key role for mitochondria is provided by the observation that early-onset PD can be caused by mutations in the nuclear gene encoding a mitochondrial protein, PINK1, a protein kinase that is somehow able to protect cells against apoptosis induced by proteasome inhibition (Moore et al. 2005).


Figure 2.  Interplay of mitochondria, oxidative damage and the proteasome in neurodegeneration. Low-level proteasome inhibition can cause transient neuroprotection (e.g. by induction of HSPS) (Yew et al. 2005; Stefanis and Keller 2006). Adapted from Halliwell and Gutteridge (2006) with the permission of Oxford University Press.

Download figure to PowerPoint

Defects in mitochondria occur in the other common neurodegenerative disorders as well (Table 3). Indeed, in AD the toxicity of amyloid peptides (Aβs) can involve direct mitochondrial damage (Yan and Stern 2005). In addition, aggregating Aβs raise intracellular Ca2+, increase NOX activity in astrocytes and directly produce ROS (Barnham et al. 2004; Abramov and Duchen 2005; Caspersen et al. 2005; Sultana et al. 2006). If mitochondrial damage significantly depletes the ATP supply, this will interfere with removal of proteins by the ubiquitin-proteasome system (ATP dependent), and it may even cause cells to increase their rates of Aβ production (Velliquette et al. 2005).

Oxidized and nitrated proteins are usually removed by the proteasome; its inhibition allows abnormal proteins to accumulate and produces oxidative stress (Fig. 2). Exactly how this oxidative stress arises is unclear. Potential mechanisms include increased mitochondrial ROS production (Sullivan et al. 2004) and increases in nNOS activity, producing more NO (Lee et al. 2001). Formation of abnormal proteins resulting from gene mutations or of excessive amounts of normal proteins (e.g. α-synuclein, CuZnSOD) due to gene duplications or triplications (Table 4) could overload the proteasome; its activity tends to decrease with age in the brain in any case. Indeed, injecting the proteasome inhibitor lactacystin into mouse or rat substantia nigra produced neurodegeneration, movement disorders and protein aggregates (Zhang et al. 2005). Protein aggregates may stimulate RS formation from neurons, and activate microglia. Both proteasome and complex I inhibitors are widespread in nature (for example rotenone and lactacystin are natural products) and, if you were unlucky enough to consume too much of either or both (provided that they can cross the blood–brain barrier), sporadic PD or other neurodegenerative diseases might result (Halliwell 2006). Finally, RS-producing agents could initiate neurodegeneration, because RS damage mitochondria, and may inhibit proteasome function directly (e.g. by oxidative or nitrative inactivation of proteasome subunits) or indirectly (e.g. by interfering with ubiquitination, or overloading the proteasome by creating more oxidatively modified proteins such as those with HNE or isoketals attached; Halliwell 2006). Dopamine oxidation products (which accumulate in PD; Spencer et al. 1998) can both damage mitochondria and inactivate the proteasome (Keller et al. 2000).

Table 4.   Processes generating abnormal proteins in neurodegenerative diseases
  1. Adapted from Halliwell and Gutteridge (2006) with permission of Oxford University Press.

  2. *The enzyme myeloperoxidase, which uses H2O2 to oxidize Cl to HOCl, is not normally present in brain but has been reported to appear there in both patients with AD and those with PD, and can lead to protein chlorination (e.g. Green et al. 2004).

  3. †Glycoxidation involves both glycation and oxidation of proteins, forming AGE products that impair protein function and can be cytotoxic (reviewed by Halliwell and Gutteridge 2006).

Overexpression of a normal gene, causing too much normal protein to accumulate (e.g. triplication of the synuclein gene in some rare familial cases of PD)
Gene mutations, producing an abnormal protein
Aberrant splicing of mRNA, producing an abnormal protein
Faulty post-translational modification, producing an abnormal protein
Oxidation of amino acid residues by ROS
Modification by quinones/semiquinones, e.g. arising from oxidation of l-DOPA or dopamine
Nitration and/or oxidation of amino acid residues by RNS
Halogenation and/or oxidation of amino acid residues by RCS or RBS*
Spontaneous deamidation or deamination
Modification by end-products of lipid peroxidation such as HNE, other aldehydes and isoketals
Modification by end-products of the cyclooxygenase pathway, e.g. cyclopentenone prostaglandins, levuglandins

Is iron relevant?

Where does iron fit into the story? A common view [e.g. for PD and amyotrophic lateral sclerosis (ALS)] is that its deposition is a late stage in tissue injury, with the implication that it is unimportant. However, the pathology of subarachnoid hemorrhage (Asaeda et al. 2005) or ceruloplasmin deficiency (Miyajima 2003) in humans reveals the potential of heme and/or iron to cause neuronal damage. Similarly, a inborn error in the gene encoding ferritin L-chains is associated with abundant iron deposition in the basal ganglia accompanied by neurodegeneration (Schenck and Zimmerman 2004). Hallevorden–Spatz syndrome is associated with iron deposition in the globus pallidus and substantia nigra; the defective gene encodes pantothenate kinase, an enzyme involved in the biosynthesis of coenzyme A. Its absence causes cysteine accumulation, and it has been hypothesized that a pro-oxidant mixture of iron and cysteine contributes to neurodegeneration (Zhou et al. 2001). The iron content of most brain areas increases with age. Iron, copper and some other metals promote the aggregation of several proteins including α-synucleins and Aβ (Barnham et al. 2004). Indeed, the iron chelator desferrioxamine could protect against lactacystin neurotoxicity in rats (Zhang et al. 2005).

How do neurons die in these various diseases? Sometimes largely by necrosis, for example in excitotoxicity. Sometimes probably by apoptosis. As more studies are done, however, the role of ‘intermediate’ types of cell death, with features of both, is becoming more prominent (Tang et al. 2004). Some studies suggest that in mature neurons one action of oxidative stress (at a certain level) is to up-regulate expression of genes encoding cell cycle proteins, as if the neuron were about to enter the cell cycle. Most neurons cannot divide, however, and apoptosis results (Kruman 2004). This cell cycle activation seems to be associated with a response to increased levels of DNA damage.

There is intense debate about whether inclusion bodies or other abnormal proteins and their aggregates (Tables 3 and 4) are toxic to neurons. In general, it may be the early stages of aggregate formation (e.g. of huntingtin or Aβ) that are toxic rather than the final insoluble complexes; in fact, formation of the latter may be beneficial if it helps convert toxic oligomers to an insoluble form (Orr 2004).

Conclusion: prospects for therapeutic intervention with antioxidants

  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References

Increased oxidative damage occurs in all the human neurodegenerative diseases and seems especially important in AD and, perhaps to a lesser extent, in PD (Halliwell 2001; Halliwell and Gutteridge 2006). Indeed, lipid peroxidation, measured as F2-IPs in brain tissue or CSF, is already increased in patients with MCI. Neuroprostane levels are also raised in AD brain, and levels tend to rise further as dementia progresses (Markesbery et al. 2005). Brain protein carbonyls (a marker of oxidative protein damage; Halliwell and Whiteman 2004) are also increased in MCI (Keller et al. 2005) and probably oxidative DNA damage is too. High doses of α-tocopherol (2000 units per day) were reported to produce a significant delay in the deterioration of patients with AD, although sadly the same dose did not affect the progression of patients with MCI to AD (Blacker 2005). Given that α-tocopherol enters the brain only slowly, is very poor at decreasing lipid peroxidation in humans and is not depleted in AD brain, its apparent effect in AD gives hope that better antioxidants designed to target the brain could have a significant therapeutic impact (Halliwell and Gutteridge 2006). Indeed, in mice overexpressing mutant human amyloid precursor protein, brain IP formation preceded amyloid plaque deposition (Pratico et al. 2001) and inhibition of F2-IP formation by α-tocopherol administration delayed plaque formation and cognitive impairment (Sung et al. 2004). Unfortunately, α-tocopherol works better against lipid peroxidation, neurodegeneration and atherosclerosis in transgenic mice than it does in humans (Halliwell and Gutteridge 2006). Thus the development of novel antioxidants for AD, PD and other neurodegenerative diseases is a major research area (Halliwell 2001; Moosmann and Behl 2002; Mandel et al. 2005).

By contrast, none of the antioxidants tested so far for ALS has given much benefit in patients, despite the fact that several increased the lifespan or delayed the onset of symptoms in mice overexpressing familial ALS (FALS)-related mutant CuZnSODs (Halliwell 2001; Orrell et al. 2005). Nevertheless, the antioxidant idebenone (Rustin et al. 2004) or a mixture of coenzyme Q and α-tocopherol (Hart et al. 2005) have limited therapeutic benefit in Freidreich's ataxia, consistent with a role for oxidative damage. Indeed, idebenone decreased the raised urinary excretion of 8OHdG in these patients (Schulz et al. 2000).

What about stroke? Again, many agents (antioxidant and otherwise) have worked in animal models but not in humans (reviewed by Halliwell and Gutteridge 2006). However the low molecular mass ‘mimic’ of GPx activity, Ebselen (Klotz and Sies 2003) and the modified spin trap NXY-059 (Cerovive®) look more promising. Several clinical studies have suggested a benefit of Ebselen in subarachnoid hemorrhage and stroke (reviewed by Margaill et al. 2005) and Cerovive® seems to improve outcome in primate stroke models even when administered as late as 4 h after the ischemia. Studies in humans look promising (Maples et al. 2004; Margaill et al. 2005; del Zoppo 2006). However, Ebselen and Cerovive® may be working by effects additional to (Ebselen) or instead of (NXY-059) antioxidant ones. Thus Ebselen can inhibit lipoxygenases, NOS and phagocyte RS production, at least in vitro. Cerovive® is quite a poor scavenger of RS in vitro (Walther et al. 1999; Porciuncula et al. 2003) (Maples et al. 2001) but may act instead as a powerful anti-inflammatory agent (Maples et al. 2004).

Conversely, some agents beneficial in PD may exert part of their effects in vivo by antioxidant action. Thus the catechol-O-methyltransferase inhibitors tolcapone and nitecapone can act as chain-breaking antioxidants in vitro and might do so in vivo. However, their oxidation by ROS to produce potentially toxic semiquinones and quinones may complicate the picture (Smith et al. 2003). Selegiline and its metabolites might increase the levels of neuronal antioxidant defenses, as might pergolide and ropinirole (reviewed by Olanow 2004; Halliwell and Gutteridge 2006).

The importance of lifestyle in decreasing the risk of developing AD has recently become apparent. Diets rich in fruits and vegetables, and low in fat seem protective to some extent, as is regular physical exercise and mental activity (Salerno-Kennedy and Cashman 2005; Mattson et al. 2004; Marx 2005). Hypercholesterolemia, long known as a risk factor for vascular dementia and other aspects of atherosclerosis, now appears to be a significant risk factor for AD (Eckert et al. 2005). High cholesterol promotes lipid peroxidation (Morrow 2005) and may damage the blood–brain barrier, allowing excess iron to enter the brain (Ong and Halliwell 2004). In addition, too much cholesterol may encourage Aβ formation (Eckert et al. 2005).

So we have some way to go before finding effective antioxidant therapeutics for neurodegenerative diseases, but we have come a long way since 1992 (Halliwell 1992). Hopefully my next review (if I escape dementia for that long) will be an account of how the target was reached.


  1. Top of page
  2. Abstract
  3. Attacking the nervous system
  4. Radicals and other reactive species
  5. What damage can RS do?
  6. Problems of the brain
  7. Defending the brain
  8. Coping with oxidative damage in the brain
  9. The neurotoxicity of superoxide
  10. Oxidative stress and neurodegenerative diseases: some general concepts
  11. Conclusion: prospects for therapeutic intervention with antioxidants
  12. References
  • Abramov A. Y. and Duchen M. R. (2005) The role of an astrocytic NADPH oxidase in the neurotoxicity of amyloid β peptides. Philos. Trans. R. Soc. Lond. B Biol. Sci. 360, 23092314.
  • Aksenov M. Y., Aksenova M. V., Carney J. M. and Butterfield D. A. (1997) Oxidative modification of glutamine synthetase by amyloid β peptide. Free Radic. Res. 27, 267281.
  • Alayash A. I. (2004) Oxygen therapeutics: can we tame haemoglobin? Nat. Rev. Drug Discov. 3, 152159.
  • Alvarez B. and Radi R. (2003) Peroxynitrite reactivity with amino acids and proteins. Amino Acids 25, 295311.
  • Araki T., Sasaki Y. and Milbrandt J. (2004) Increased nuclear NAD biosynthesis and SIRT1 activation prevent axonal degeneration. Science 305, 10101013.
  • Aruoma O. I., Laughton M. J. and Halliwell B. (1989) Carnosine, homocarnosine and anserine: could they act as antioxidants in vivo? Biochem. J. 264, 863869.
  • Asaeda M., Sakamoto M., Kurosaki M., Tabuchi S., Kamitani H., Yokota M. and Watanabe T. (2005) A non-enzymatic derived arachidonyl peroxide, 8-iso-prostaglandin F2 alpha, in cerebrospinal fluid of patients with aneurysmal subarachnoid hemorrhage participates in the pathogenesis of delayed cerebral vasospasm. Neurosci. Lett. 373, 222225.
  • Augusto O., Bonini M. G., Amanso A. M., Linares E., Santos C. C. and De Menezes S. L. (2002) Nitrogen dioxide and carbonate radical anion: two emerging radicals in biology. Free Radic. Biol. Med. 32, 841859.
  • Babcock G. T. (1999) How oxygen is activated and reduced in respiration. Proc. Natl Acad. Sci. USA 96, 12 97112 973.
  • Back S. A., Luo N. L., Mallinson R. A. et al. (2005) Selective vulnerability of preterm white matter to oxidative damage defined by F2-isoprostanes. Ann. Neurol. 58, 108120.
  • Barnham K. J., Masters C. L. and Bush A. I. (2004) Neurodegenerative diseases and oxidative stress. Nat. Rev. Drug Discov. 3, 205214.
  • Beal M. F. and Shults C. W. (2003) Effects of Coenzyme Q10 in Huntington's disease and early Parkinson's disease. Biofactors 18, 153161.
  • Bence N. F., Sampat R. M. and Kopito R. R. (2001) Impairment of the ubiquitin–proteasome system by protein aggregation. Science 292, 15521555.
  • Blacker D. (2005) Mild cognitive impairment – no benefit from vitamin E, little from donepezil. N. Engl. J. Med. 352, 24392441.
  • Von Bohlen und Halbach O., Schober A. and Krieglstein K. (2004) Genes, proteins, and neurotoxins involved in Parkinson's disease. Prog. Neurobiol. 73, 151177.
  • Bota D. A. and Davies K. J. (2002) Lon protease preferentially degrades oxidized mitochondrial aconitase by an ATP-stimulated mechanism. Nat. Cell Biol. 4, 674680.
  • Bredt D. S. (1999) Endogenous nitric oxide synthesis: biological functions and pathophysiology. Free Radic. Res. 31, 577596.
  • Brigelius-Flohe R. (1999) Tissue-specific functions of individual glutathione peroxidases. Free Radic. Biol. Med. 27, 951965.
  • Burdo J. R. and Connor J. R. (2003) Brain iron uptake and homeostatic mechanisms: an overview. Biometals 16, 6375.
  • Burmester T. and Hankeln T. (2004) Neuroglobin: a respiratory protein of the nervous system. News Physiol. Sci. 19, 110113.
  • Calabrese V., Lodi R., Tonon C., D'Agata V., Sapienza M., Scapagnini G., Mangiameli A., Pennisi G., Stella A. M. and Butterfield D. A. (2005) Oxidative stress, mitochondrial dysfunction and cellular stress response in Friedreich's ataxia. J. Neurol. Sci. 233, 145162.
  • Capasso M., Jeng J. M., Malavolta M., Mocchegiani E. and Sensi S. L. (2005) Zinc dyshomeostasis: a key modulator of neuronal injury. J. Alzheimers Dis. 8, 93108.
  • Carrasco J., Penkowa M., Giralt M., Camats J., Molinero A., Campbell I. L., Palmiter R. D. and Hidalgo J. (2003) Role of metallothionein-III following central nervous system damage. Neurobiol. Dis. 13, 2236.
  • Caspersen C., Wang N., Yao J., Sosunov A., Chen X., Lustbader J. W., Xu H. W., Stern D., McKhann G. and Yan S. D. (2005) Mitochondrial Abeta: a potential focal point for neuronal metabolic dysfunction in Alzheimer's disease. FASEB J. 19, 20402041.
  • Castegna A., Thongboonkerd V., Klein J., Lynn B. C., Wang Y. L., Osaka H., Wada K. and Butterfield D. A. (2004) Proteomic analysis of brain proteins in the gracile axonal dystrophy (gad) mouse, a syndrome that emanates from dysfunctional ubiquitin carboxyl-terminal hydrolase L-1, reveals oxidation of key proteins. J. Neurochem. 88, 15401546.
  • Chang E. F., Claus C. P., Vreman H. J., Wong R. J. and Noble-Haeusslein L. J. (2005) Heme regulation in traumatic brain injury: relevance to the adult and developing brain. J. Cereb. Blood Flow Metab. 25, 14011417.
  • Chavko M., Auker C. R. and McCarron R. M. (2003) Relationship between protein nitration and oxidation and development of hyperoxic seizures. Nitric Oxide 9, 1823.
  • Chiu D. T., Van Den Berg J., Kuypers F. A., Hung I. J., Wei J. S. and Liu T. Z. (1996) Correlation of membrane lipid peroxidation with oxidation of hemoglobin variants: possibly related to the rates of hemin release. Free Radic. Biol. Med. 21, 8995.
  • Choi J., Levey A. I., Weintraub S. T., Rees H. D., Gearing M., Chin L. S. and Li L. (2004) Oxidative modifications and down-regulation of ubiquitin carboxyl-terminal hydrolase L1 associated with idiopathic Parkinson's and Alzheimer's diseases. J. Biol. Chem. 279, 13 25613 264.
  • Ciechanover A. and Brundin P. (2003) The ubiquitin proteasome system in neurodegenerative diseases: sometimes the chicken, sometimes the egg. Neuron 40, 427446.
  • Davies M. J. (2005) The oxidative environment and protein damage. Biochim. Biophys. Acta 1703, 93109.
  • Davies S. S., Amarnath V., Montine K. S., Bernoud-Hubac N., Boutaud O., Montine T. J. and Roberts L. J. II (2002) Effects of reactive gamma-ketoaldehydes formed by the isoprostane pathway (isoketals) and cyclooxygenase pathway (levuglandins) on proteasome function. FASEB J. 16, 715717.
  • Decker E. A., Livisay S. A. and Zhou S. (2000) A re-evaluation of the antioxidant activity of purified carnosine. Biochemistry (Mosc.) 65, 766770.
  • Demasi M. and Davies K. J. (2003) Proteasome inhibitors induce intracellular protein aggregation and cell death by an oxygen-dependent mechanism. FEBS Lett. 542, 8994.
  • Ding Q., Dimayuga E., Markesbery W. R. and Keller J. N. (2004) Proteasome inhibition increases DNA and RNA oxidation in astrocyte and neuron cultures. J. Neurochem. 91, 12111218.
  • Dringen R., Pawlowski P. G. and Hirrlinger J. (2005) Peroxide detoxification by brain cells. J. Neurosci. Res. 79, 157165.
  • Duncan A. J. and Heales S. J. (2005) Nitric oxide and neurological disorders. Mol. Aspects Med. 26, 6796.
  • Eckert G. P., Wood W. G. and Muller W. E. (2005) Statins: drugs for Alzheimer's disease? J. Neural Transm. 112, 10571071.
  • El-Khamisy S. F., Saifi G. M., Weinfeld M., Johansson F., Helleday T., Lupski J. R. and Caldecott K. W. (2005) Defective DNA single-strand break repair in spinocerebellar ataxia with axonal neuropathy-1. Nature 434, 108113.
  • Englander E. W. and Ma H. (2006) Differential modulation of base excision repair activities during brain ontogeny: implications for repair of transcribed DNA. Mech. Ageing Dev. 127, 6469.
  • Esterbauer H., Schaur R. J. and Zollner H. (1991) Chemistry and biochemistry of 4-hydroxynonenal, malonaldehyde and related aldehydes. Free Radic. Biol. Med. 11, 81128.
  • Evans M. D., Dizdaroglu M. and Cooke M. S. (2004) Oxidative DNA damage and disease: induction, repair and significance. Mutat. Res. 567, 161.
  • Ezaki J. and Kominami E. (2004) The intracellular location and function of proteins of neuronal ceroid lipofuscinoses. Brain Pathol. 14, 7785.
  • Fam S. S. and Morrow J. D. (2003) The isoprostanes: unique products of arachidonic acid oxidation – a review. Curr. Med. Chem. 10, 17231740.
  • Farooqui A. A., Ong W. Y., Lu X. R., Halliwell B. and Horrocks L. A. (2001) Neurochemical consequences of kainate-induced toxicity in brain: involvement of arachidonic acid release and prevention of toxicity by phospholipase A(2) inhibitors. Brain Res. Brain Res. Rev. 38, 6178.
  • Fonfria E., Marshall I. C., Boyfield I., Skaper S. D., Hughes J. P., Owen D. E., Zhang W., Miller B. A., Benham C. D. and McNulty S. (2005) Amyloid β-peptide (1–42) and hydrogen peroxide-induced toxicity are mediated by TRPM2 in rat primary striatal cultures. J. Neurochem. 95, 715723.
  • Foote C. S., Shook F. C. and Abakerli R. B. (1984) Characterization of singlet oxygen. Methods Enzymol. 105, 3647.
  • Fridovich I. (1989) Superoxide dismutases. An adaptation to a paramagnetic gas. J. Biol. Chem. 264, 77617764.
  • Fridovich I. (1995) Superoxide radical and superoxide dismutases. Annu. Rev. Biochem. 64, 97112.
  • Fridovich I. (1999) Fundamental aspects of reactive oxygen species, or what's the matter with oxygen? Ann. N. Y. Acad. Sci. 893, 1318.
  • Gal S., Zheng H., Fridkin M. and Youdim M. B. (2005) Novel multifunctional neuroprotective iron chelator–monoamine oxidase inhibitor drugs for neurodegenerative diseases. In vivo selective brain monoamine oxidase inhibition and prevention of MPTP-induced striatal dopamine depletion. J. Neurochem. 95, 7988.
  • Georgiou G. and Masip L. (2003) Biochemistry. An overoxidation journey with a return ticket. Science 300, 592594.
  • Girotti A. W. and Kriska T. (2004) Role of lipid hydroperoxides in photo-oxidative stress signaling. Antioxid. Redox Signal. 6, 301310.
  • Gonzalez F. J. (2005) Role of cytochromes P450 in chemical toxicity and oxidative stress: studies with CYP2E1. Mutat. Res. 569, 101110.
  • Gonzalez-Zulueta M., Ensz L. M., Mukhina G., Lebovitz R. M., Zwacka R. M., Engelhardt J. F., Oberley L. W., Dawson V. L. and Dawson T. M. (1998) Manganese superoxide dismutase protects nNOS neurons from NMDA and nitric oxide-mediated neurotoxicity. J. Neurosci. 18, 20402055.
  • Green P. S., Mendez A. J., Jacob J. S., Crowley J. R., Growdon W., Hyman B. T. and Heinecke J. W. (2004) Neuronal expression of myeloperoxidase is increased in Alzheimer's disease. J. Neurochem. 90, 724733.
  • Greenacre S. A. and Ischiropoulos H. (2001) Tyrosine nitration: localisation, quantification, consequences for protein function and signal transduction. Free Radic. Res. 34, 541581.
  • Grune T. and Davies K. J. (2003) The proteasomal system and HNE-modified proteins. Mol. Aspects Med. 24, 195204.
  • Grune T., Merker K., Sandig G. and Davies K. J. (2003) Selective degradation of oxidatively modified protein substrates by the proteasome. Biochem. Biophys. Res. Commun. 305, 709718.
  • Grune T., Jung T., Merker K. and Davies K. J. (2004) Decreased proteolysis caused by protein aggregates, inclusion bodies, plaques, lipofuscin, ceroid, and ‘aggresomes’ during oxidative stress, aging, and disease. Int. J. Biochem. Cell Biol. 36, 25192530.
  • Guarente L. (2005) Calorie restriction and SIR2 genes – towards a mechanism. Mech. Ageing Dev. 126, 923928.
  • Gutteridge J. M. (1983) Antioxidant properties of caeruloplasmin towards iron- and copper-dependent oxygen radical formation. FEBS Lett. 157, 3740.
  • Gutteridge J. M. (1986) Iron promoters of the Fenton reaction and lipid peroxidation can be released from haemoglobin by peroxides. FEBS Lett. 201, 291295.
  • Gutteridge J. M. (1992) Iron and oxygen radicals in brain. Ann. Neurol. 32, S16S21.
  • Gutteridge J. M. and Smith A. (1988) Antioxidant protection by haemopexin of haem-stimulated lipid peroxidation. Biochem. J. 256, 861865.
  • Halliwell B. (1984) Manganese ions, oxidation reactions and the superoxide radical. Neurotoxicology 5, 113117.
  • Halliwell B. (1992) Reactive oxygen species and the central nervous system. J. Neurochem. 59, 16091623.
  • Halliwell B. (2001) Role of free radicals in the neurodegenerative diseases: therapeutic implications for antioxidant treatment. Drugs Aging 18, 685716.
  • Halliwell B. (2002) Hypothesis: proteasomal dysfunction: a primary event in neurodegeneration that leads to nitrative and oxidative stress and subsequent cell death. Ann. N. Y. Acad. Sci. 962, 182194.
  • Halliwell B. (2003) Oxidative stress in cell culture: an under-appreciated problem? FEBS Lett. 540, 36.
  • Halliwell B. (2006) Proteasomal dysfunction: a common feature of all neurodegenerative diseases? Implications for the environmental origins of neurodegeneration. Antiox. Redox Signal in press:.
  • Halliwell B. and Gutteridge J. M. (1984) Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem. J. 219, 114.
  • Halliwell B. and Gutteridge J. M. (1990) Role of free radicals and catalytic metal ions in human disease: an overview. Methods Enzymol. 186, 185.
  • Halliwell B. and Gutteridge J. M. (1992) Biologically relevant metal ion-dependent hydroxyl radical generation. An update. FEBS Lett. 307, 108112.
  • Halliwell B. and Gutteridge J. M. (1997) Lipid peroxidation in brain homogenates: the role of iron and hydroxyl radicals. J. Neurochem. 69, 13301331.
  • Halliwell B. and Gutteridge J. M. C. (2006) Free Radicals in Biology and Medicine, 4th edn, Oxford University Press, Oxford.
  • Halliwell B. and Whiteman M. (2004) Measuring reactive species and oxidative damage in vivo and in cell culture: how should you do it and what do the results mean? Br. J. Pharmacol. 142, 231255.
  • Halliwell B., Zhao K. and Whiteman M. (1999) Nitric oxide and peroxynitrite. The ugly, the uglier and the not so good: a personal view of recent controversies. Free Radic. Res. 31, 651669.
  • Halliwell B., Rafter J. and Jenner A. (2005) Health promotion by flavonoids, tocopherols, tocotrienols, and other phenols: direct or indirect effects? Antioxidant or not? Am. J. Clin. Nutr. 81, 268S276S.
  • Hart P. E., Lodi R., Rajagopalan B. et al. (2005) Antioxidant treatment of patients with Friedreich ataxia: four-year follow-up. Arch. Neurol. 62, 621626.
  • Hayton S. M. and Muller D. P. (2004) Vitamin E in neural and visual function. Ann. N. Y. Acad. Sci. 1031, 263270.
  • Herold S., Fago A., Weber R. E., Dewilde S. and Moens L. (2004) Reactivity studies of the Fe(III) and Fe(II) NO forms of human neuroglobin reveal a potential role against oxidative stress. J. Biol. Chem. 279, 22 84122 847.
  • Higgins J. J., Pucilowska J., Lombardi R. Q. and Rooney J. P. (2004) A mutation in a novel ATP-dependent Lon protease gene in a kindred with mild mental retardation. Neurology 63, 19271931.
  • Hinerfeld D., Traini M. D., Weinberger R. P., Cochran B., Doctrow S. R., Harry J. and Melov S. (2004) Endogenous mitochondrial oxidative stress: neurodegeneration, proteomic analysis, specific respiratory chain defects, and efficacious antioxidant therapy in superoxide dismutase 2 null mice. J. Neurochem. 88, 657667.
  • Hou X., Roberts L. J. II, Gobeil F. Jr et al. (2004) Isomer-specific contractile effects of a series of synthetic F2-isoprostanes on retinal and cerebral microvasculature. Free Radic. Biol. Med. 36, 163172.
  • Hovatta I., Tennant R. S., Helton R. et al. (2005) Glyoxalase 1 and glutathione reductase 1 regulate anxiety in mice. Nature 438, 662666.
  • Howard L. A., Miksys S., Hoffmann E., Mash D. and Tyndale R. F. (2003) Brain CYP2E1 is induced by nicotine and ethanol in rat and is higher in smokers and alcoholics. Br. J. Pharmacol. 138, 13761386.
  • Huang J., Agus D. B., Winfree C. J. et al. (2001) Dehydroascorbic acid, a blood–brain barrier transportable form of vitamin C, mediates potent cerebroprotection in experimental stroke. Proc. Natl Acad. Sci. USA 98, 11 72011 724.
  • Imlay J. A. (2003) Pathways of oxidative damage. Annu. Rev. Microbiol. 57, 395418.
  • Ischiropoulos H. and Beckman J. S. (2003) Oxidative stress and nitration in neurodegeneration: cause, effect, or association? J. Clin. Invest. 111, 163169.
  • Jenner P. (2003) Oxidative stress in Parkinson's disease. Ann. Neurol. 53, S26S36.
  • Jeong S. Y. and David S. (2003) Glycosylphosphatidylinositol-anchored ceruloplasmin is required for iron efflux from cells in the central nervous system. J. Biol. Chem. 278, 27 14427 148.
  • Jones S., Batchelor M., Bhelt D., Clarke A. R., Collinge J. and Jackson G. S. (2005) Recombinant prion protein does not possess SOD-1 activity. Biochem. J. 392, 309312.
  • Kasai H. (2002) Chemistry-based studies on oxidative DNA damage: formation, repair, and mutagenesis. Free Radic. Biol. Med. 33, 450456.
  • Keller J. N., Huang F. F., Dimayuga E. R. and Maragos W. F. (2000) Dopamine induces proteasome inhibition in neural PC12 cell line. Free Radic. Biol. Med. 29, 10371042.
  • Keller J. N., Dimayuga E., Chen Q., Thorpe J., Gee J. and Ding Q. (2004) Autophagy, proteasomes, lipofuscin, and oxidative stress in the aging brain. Int. J. Biochem. Cell Biol. 36, 23762391.
  • Keller J. N., Schmitt F. A., Scheff S. W., Ding Q., Chen Q., Butterfield D. A. and Markesbery W. R. (2005) Evidence of increased oxidative damage in subjects with mild cognitive impairment. Neurology 64, 11521156.
  • Kim G. W., Gasche Y., Grzeschik S., Copin J. C., Maier C. M. and Chan P. H. (2003) Neurodegeneration in striatum induced by the mitochondrial toxin 3-nitropropionic acid: role of matrix metalloproteinase-9 in early blood–brain barrier disruption? J. Neurosci. 23, 87338742.
  • Kirkwood T. B. (2005) Understanding the odd science of aging. Cell 120, 437447.
  • Klotz L. O. and Sies H. (2003) Defenses against peroxynitrite: selenocompounds and flavonoids. Toxicol. Lett. 140–141, 125132.
  • Koh D. W., Dawson T. M. and Dawson V. L. (2005) Mediation of cell death by poly(ADP-ribose) polymerase-1. Pharmacol. Res. 52, 514.
  • Kondo M., Shibata T., Kumagai T., Osawa T., Shibata N., Kobayashi M., Sasaki S., Iwata M., Noguchi N. and Uchida K. (2002) 15-Deoxy-delta(12,14)-prostaglandin J(2): the endogenous electrophile that induces neuronal apoptosis. Proc. Natl Acad. Sci. USA 99, 73677372.
  • Krause K. H. (2004) Tissue distribution and putative physiological function of NOX family NADPH oxidases. Jpn J. Infect. Dis. 57, S28S29.
  • Krizbai I. A., Bauer H., Bresgen N., Eckl P. M., Farkas A., Szatmari E., Traweger A., Wejksza K. and Bauer H. C. (2005) Effect of oxidative stress on the junctional proteins of cultured cerebral endothelial cells. Cell. Mol. Neurobiol. 25, 129139.
  • Kruman I. I. (2004) Why do neurons enter the cell cycle? Cell Cycle 3, 769773.
  • Kruman I. I., Schwartz E., Kruman Y., Cutler R. G., Zhu X., Greig N. H. and Mattson M. P. (2004) Suppression of uracil-DNA glycosylase induces neuronal apoptosis. J. Biol. Chem. 279, 43 95243 960.
  • Kudin A. P., Debska-Vielhaber G. and Kunz W. S. (2005) Characterization of superoxide production sites in isolated rat brain and skeletal muscle mitochondria. Biomed. Pharmacother. 59, 163168.
  • Lebovitz R. M., Zhang H., Vogel H., Cartwright J. Jr, Dionne L., Lu N., Huang S. and Matzuk M. M. (1996) Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice. Proc. Natl Acad. Sci. USA 93, 97829787.
  • Lee M. H., Hyun D. H., Jenner P. and Halliwell B. (2001) Effect of proteasome inhibition on cellular oxidative damage, antioxidant defences and nitric oxide production. J. Neurochem. 78, 3241.
  • Leung I., Tso M., Li W. and Lam T. (2001) Absorption and tissue distribution of zeaxanthin and lutein in rhesus monkeys after taking Fructus lycii (Gou Qi Zi) extract. Invest. Ophthalmol. Vis. Sci. 42, 466471.
  • Liang L. P. and Patel M. (2004) Iron–sulfur enzyme mediated mitochondrial superoxide toxicity in experimental Parkinson's disease. J. Neurochem. 90, 10761084.
  • Lin S., Wei X., Bales K. R., Paul A. B., Ma Z., Yan G., Paul S. M. and Du Y. (2005) Minocycline blocks bilirubin neurotoxicity and prevents hyperbilirubinemia-induced cerebellar hypoplasia in the Gunn rat. Eur. J. Neurosci. 22, 2127.
  • Linder N., Rapola J. and Raivio K. O. (1999) Cellular expression of xanthine oxidoreductase protein in normal human tissues. Lab. Invest. 79, 967974.
  • Liochev S. I. and Fridovich I. (2005) Cross-compartment protection by SOD1. Free Radic. Biol. Med. 38, 146147.
  • Mailly F., Marin P., Israel M., Glowinski J. and Premont J. (1999) Increase in external glutamate and NMDA receptor activation contribute to H2O2-induced neuronal apoptosis. J. Neurochem. 73, 11811188.
  • Mandel S. A., Avramovich-Tirosh Y., Reznichenko L., Zheng H., Weinreb O., Amit T. and Youdim M. B. (2005) Multifunctional activities of green tea catechins in neuroprotection. Modulation of cell survival genes, iron-dependent oxidative stress and PKC signaling pathway. Neurosignals 14, 4660.
  • Maples K. R., Ma F. and Zhang Y. K. (2001) Comparison of the radical trapping ability of PBN, S-PPBN and NXY-059. Free Radic. Res. 34, 417426.
  • Maples K. R., Green A. R. and Floyd R. A. (2004) Nitrone-related therapeutics: potential of NXY-059 for the treatment of acute ischaemic stroke. CNS Drugs 18, 10711084.
  • Marchis S. D., Modena C., Peretto P., Migheli A., Margolis F. L. and Fasolo A. (2000) Carnosine-related dipeptides in neurons and glia. Biochemistry (Mosc.) 65, 824833.
  • Margaill I., Plotkine M. and Lerouet D. (2005) Antioxidant strategies in the treatment of stroke. Free Radic. Biol. Med. 39, 429443.
  • Mark R. J., Lovell M. A., Markesbery W. R., Uchida K. and Mattson M. P. (1997) A role for 4-hydroxynonenal, an aldehydic product of lipid peroxidation, in disruption of ion homeostasis and neuronal death induced by amyloid β-peptide. J. Neurochem. 68, 255264.
  • Markesbery W. R., Kryscio R. J., Lovell M. A. and Morrow J. D. (2005) Lipid peroxidation is an early event in the brain in amnestic mild cognitive impairment. Ann. Neurol. 58, 730735.
  • Martin C. J. (2006) Manganese neurotoxicity: Connecting the dots along the continuum of dysfunction. Neurotoxicology in press:.
  • Martin A., Youdim K., Szprengiel A., Shukitt-Hale B. and Joseph J. (2002) Roles of vitamins E and C on neurodegenerative diseases and cognitive performance. Nutr. Rev. 60, 308326.
  • Marx J. (2005) Neuroscience. Preventing Alzheimer's: a lifelong commitment? Science 309, 864866.
  • Mattson M. P. and Chan S. L. (2003) Neuronal and glial calcium signaling in Alzheimer's disease. Cell Calcium 34, 385397.
  • Mattson M. P., Duan W., Wan R. and Guo Z. (2004) Prophylactic activation of neuroprotective stress response pathways by dietary and behavioral manipulations. Neurorx 1, 111116.
  • McDonald S. R., Sohal R. S. and Forster M. J. (2005) Concurrent administration of coenzyme Q10 and alpha-tocopherol improves learning in aged mice. Free Radic. Biol. Med. 38, 729736.
  • McNaught K. S., Olanow C. W., Halliwell B., Isacson O. and Jenner P. (2001) Failure of the ubiquitin–proteasome system in Parkinson's disease. Nat. Rev. Neurosci. 2, 589594.
  • Melov S., Schneider J. A., Day B. J., Hinerfeld D., Coskun P., Mirra S. S., Crapo J. D. and Wallace D. C. (1998) A novel neurological phenotype in mice lacking mitochondrial manganese superoxide dismutase. Nat. Genet. 18, 159163.
  • Miksys S. and Tyndale R. F. (2004) The unique regulation of brain cytochrome P450 2 (CYP2) family enzymes by drugs and genetics. Drug Metab. Rev. 36, 313333.
  • Miksys S., Rao Y., Hoffmann E., Mash D. C. and Tyndale R. F. (2002) Regional and cellular expression of CYP2D6 in human brain: higher levels in alcoholics. J. Neurochem. 82, 13761387.
  • Mitsiades C. S., Mitsiades N., Hideshima T., Richardson P. G. and Anderson K. C. (2005) Proteasome inhibitors as therapeutics. Essays Biochem. 41, 205218.
  • Miyajima H. (2003) Aceruloplasminemia, an iron metabolic disorder. Neuropathology 23, 345350.
  • Miyamoto S., Martinez G. R., Medeiros M. H. and Di Mascio P. (2003) Singlet molecular oxygen generated from lipid hydroperoxides by the Russell mechanism: studies using 18(O)-labeled linoleic acid hydroperoxide and monomol light emission measurements. J. Am. Chem. Soc. 125, 61726179.
  • Moore D. J., West A. B., Dawson V. L. and Dawson T. M. (2005) Molecular pathophysiology of Parkinson's disease. Annu. Rev. Neurosci. 28, 5787.
  • Moosmann B. and Behl C. (2002) Antioxidants as treatment for neurodegenerative disorders. Expert Opin. Invest. Drugs 11, 14071435.
  • Moreira P. I., Honda K., Liu Q., Santos M. S., Oliveira C. R., Aliev G., Nunomura A., Zhu X., Smith M. A. and Perry G. (2005) Oxidative stress: the old enemy in Alzheimer's disease pathophysiology. Curr. Alzheimer Res. 2, 403408.
  • Morrow J. D. (2005) Quantification of isoprostanes as indices of oxidant stress and the risk of atherosclerosis in humans. Arterioscler. Thromb. Vasc. Biol. 25, 279286.
  • Moskovitz J. and Stadtman E. R. (2003) Selenium-deficient diet enhances protein oxidation and affects methionine sulfoxide reductase (MsrB) protein level in certain mouse tissues. Proc. Natl Acad. Sci. USA 100, 74867490.
  • Muller D. P. and Goss-Sampson M. A. (1990) Neurochemical, neurophysiological, and neuropathological studies in vitamin E deficiency. Crit. Rev. Neurobiol. 5, 239263.
  • Musiek E. S., Milne G. L., McLaughlin B. and Morrow J. D. (2005) Cyclopentenone eicosanoids as mediators of neurodegeneration: a pathogenic mechanism of oxidative stress-mediated and cyclooxygenase-mediated neurotoxicity. Brain Pathol. 15, 149158.
  • Nigam S. and Schewe T. (2000) Phospholipase A(2)s and lipid peroxidation. Biochim. Biophys. Acta 1488, 167181.
  • Nimmerjahn A., Kirchhoff F. and Helmchen F. (2005) Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science 308, 13141318.
  • Olanow C. W. (2004) The scientific basis for the current treatment of Parkinson's disease. Annu. Rev. Med. 55, 4160.
  • Ong W. Y. and Halliwell B. (2004) Iron, atherosclerosis, and neurodegeneration: a key role for cholesterol in promoting iron-dependent oxidative damage? Ann. N. Y. Acad. Sci. 1012, 5164.
  • Ong W. Y., Hu C. Y., Hjelle O. P., Ottersen O. P. and Halliwell B. (2000a) Changes in glutathione in the hippocampus of rats injected with kainate: depletion in neurons and upregulation in glia. Exp. Brain Res. 132, 510516.
  • Ong W. Y., Lu X. R., Hu C. Y. and Halliwell B. (2000b) Distribution of hydroxynonenal-modified proteins in the kainate-lesioned rat hippocampus: evidence that hydroxynonenal formation precedes neuronal cell death. Free Radic. Biol. Med. 28, 12141221.
  • Orr H. T. (2004) Neurodegenerative disease: neuron protection agency. Nature 431, 747748.
  • Orrell R. W., Lane R. J. and Ross M. (2005) Antioxidant treatment for amyotrophic lateral sclerosis/motor neuron disease. Cochrane Database Syst. Rev., CD002829.
  • Parks E. and Traber M. G. (2000) Mechanisms of vitamin E regulation: research over the past decade and focus on the future. Antioxid. Redox Signal. 2, 405412.
  • Passage E., Norreel J. C., Noack-Fraissignes P., Sanguedolce V., Pizant J., Thirion X., Robaglia-Schlupp A., Pellissier J. F. and Fontes M. (2004) Ascorbic acid treatment corrects the phenotype of a mouse model of Charcot–Marie–Tooth disease. Nat. Med. 10, 396401.
  • Patenaude A., Murthy M. R. and Mirault M. E. (2005) Emerging roles of thioredoxin cycle enzymes in the central nervous system. Cell. Mol. Life Sci. 62, 10631080.
  • Paul T. (2000) Effect of a prolonged superoxide flux on transferrin and ferritin. Arch. Biochem. Biophys. 382, 253261.
  • Phillis J. W. and O'Regan M. H. (2004) A potentially critical role of phospholipases in central nervous system ischemic, traumatic, and neurodegenerative disorders. Brain Res. Brain Res. Rev. 44, 1347.
  • Phumala N., Porasuphatana S., Unchern S., Pootrakul P., Fucharoen S. and Chantharaksri U. (2003) Hemin: a possible cause of oxidative stress in blood circulation of beta-thalassemia/hemoglobin E disease. Free Radic. Res. 37, 129135.
  • Porciuncula L. O., Rocha J. B., Cimarosti H., Vinade L., Ghisleni G., Salbego C. G. and Souza D. O. (2003) Neuroprotective effect of ebselen on rat hippocampal slices submitted to oxygen-glucose deprivation: correlation with immunocontent of inducible nitric oxide synthase. Neurosci. Lett. 346, 101104.
  • Pratico D., Uryu K., Leight S., Trojanoswki J. Q. and Lee V. M. (2001) Increased lipid peroxidation precedes amyloid plaque formation in an animal model of Alzheimer amyloidosis. J. Neurosci. 21, 41834187.
  • Przedborski S. and Ischiropoulos H. (2005) Reactive oxygen and nitrogen species: weapons of neuronal destruction in models of Parkinson's disease. Antioxid. Redox Signal. 7, 685693.
  • Puppo A. and Halliwell B. (1988) Formation of hydroxyl radicals from hydrogen peroxide in the presence of iron. Is haemoglobin a biological Fenton reagent? Biochem. J. 249, 185190.
  • Pyne-Geithman G. J., Morgan C. J., Wagner K., Dulaney E. M., Carrozzella J., Kanter D. S., Zuccarello M. and Clark J. F. (2005) Bilirubin production and oxidation in CSF of patients with cerebral vasospasm after subarachnoid hemorrhage. J. Cereb. Blood Flow Metab. 25, 10701077.
  • Rao R. V., Niazi K., Mollahan P., Mao X., Crippen D., Poksay K. S., Chen S. and Bredesen D. E. (2006) Coupling endoplasmic reticulum stress to the cell-death program: a novel HSP90-independent role for the small chaperone protein p23. Cell Death Differ. 13, 415425.
  • Reddy P. H. and Beal M. F. (2005) Are mitochondria critical in the pathogenesis of Alzheimer's disease? Brain. Res. Rev. 49, 618632.
  • Rhee S. G., Chae H. Z. and Kim K. (2005) Peroxiredoxins: a historical overview and speculative preview of novel mechanisms and emerging concepts in cell signaling. Free Radic. Biol. Med. 38, 15431552.
  • Rice M. E. (2000) Ascorbate regulation and its neuroprotective role in the brain. Trends Neurosci. 23, 209216.
  • Rice-Evans C., Okunade G. and Khan R. (1989) The suppression of iron release from activated myoglobin by physiological electron donors and by desferrioxamine. Free Radic. Res. Commun. 7, 4554.
  • Roy S., Lado B. H., Khanna S. and Sen C. K. (2002) Vitamin E sensitive genes in the developing rat fetal brain: a high-density oligonucleotide microarray analysis. FEBS Lett. 530, 1723.
  • Rustin P., Bonnet D., Rotig A., Munnich A. and Sidi D. (2004) Idebenone treatment in Friedreich patients: one-year-long randomized placebo-controlled trial. Neurology 62, 524525.
  • Salerno-Kennedy R. and Cashman K. D. (2005) Relationship between dementia and nutrition-related factors and disorders: an overview. Int. J. Vitam. Nutr. Res. 75, 8395.
  • Salomon R. G. (2005) Distinguishing levuglandins produced through the cyclooxygenase and isoprostane pathways. Chem. Phys. Lipids 134, 120.
  • Sanchez-Carbente M. R., Castro-Obregon S., Covarrubias L. and Narvaez V. (2005) Motoneuronal death during spinal cord development is mediated by oxidative stress. Cell Death Differ. 12, 279291.
  • Savaraj N., Wei Y., Unate H., Liu P. M., Wu C. J., Wangpaichitr M., Xia D., Xu H. J., Hu S. X. and Tien Kuo M. (2005) Redox regulation of matrix metalloproteinase gene family in small cell lung cancer cells. Free Radic. Res. 39, 373381.
  • Schenck J. F. and Zimmerman E. A. (2004) High-field magnetic resonance imaging of brain iron: birth of a biomarker? NMR Biomed. 17, 433445.
  • Schulz J. B., Dehmer T., Schols L. et al. (2000) Oxidative stress in patients with Friedreich ataxia. Neurology 55, 17191721.
  • Serbinova E., Choo M. and Packer L. (1992) Distribution and antioxidant activity of a palm oil carotene fraction in rats. Biochem. Int. 28, 881886.
  • Sheu K. F. and Blass J. P. (1999) The alpha-ketoglutarate dehydrogenase complex. Ann. N. Y. Acad. Sci. 893, 6178.
  • Sinet P. M., Heikkila R. E. and Cohen G. (1980) Hydrogen peroxide production by rat brain in vivo. J. Neurochem. 34, 14211428.
  • Smith K. S., Smith P. L., Heady T. N., Trugman J. M., Harman W. D. and Macdonald T. L. (2003) In vitro metabolism of tolcapone to reactive intermediates: relevance to tolcapone liver toxicity. Chem. Res. Toxicol. 16, 123128.
  • Smith W. L. (2005) Cyclooxygenases, peroxide tone and the allure of fish oil. Curr. Opin. Cell Biol. 17, 174182.
  • Sotiriou S., Gispert S., Cheng J. et al. (2002) Ascorbic-acid transporter Slc23a1 is essential for vitamin C transport into the brain and for perinatal survival. Nat. Med. 8, 514517.
  • Spencer J. P., Jenner P., Daniel S. E., Lees A. J., Marsden D. C. and Halliwell B. (1998) Conjugates of catecholamines with cysteine and GSH in Parkinson's disease: possible mechanisms of formation involving reactive oxygen species. J. Neurochem. 71, 21122122.
  • Spencer J. P., Abd-el-Mohsen M. M. and Rice-Evans C. (2004) Cellular uptake and metabolism of flavonoids and their metabolites: implications for their bioactivity. Arch. Biochem. Biophys. 423, 148161.
  • Stahl W., Schwarz W., Sundquist A. R. and Sies H. (1992) Cis-trans isomers of lycopene and beta-carotene in human serum and tissues. Arch. Biochem. Biophys. 294, 173177.
  • Stefanis L. and Keller J. N. (2006) The Proteasome in Neurodegeneration. Springer, New York.
  • Sullivan P. G., Dragicevic N. B., Deng J. H., Bai Y., Dimayuga E., Ding Q., Chen Q., Bruce-Keller A. J. and Keller J. N. (2004) Proteasome inhibition alters neural mitochondrial homeostasis and mitochondria turnover. J. Biol. Chem. 279, 20 69920 707.
  • Sultana R., Perluigi M. and Butterfield D. A. (2006) Redox proteomics identification of oxidatively modified proteins in Alzheimer's disease brain and in vivo and in vitro models of AD centered around Abeta (1–42). J. Chromatogr. B. 833, 311.
  • Sung S., Yao Y., Uryu K., Yang H., Lee V. M., Trojanowski J. Q. and Pratico D. (2004) Early vitamin E supplementation in young but not aged mice reduces Abeta levels and amyloid deposition in a transgenic model of Alzheimer's disease. FASEB J. 18, 323325.
  • Swain J. A., Darley-Usmar V. and Gutteridge J. M. (1994) Peroxynitrite releases copper from caeruloplasmin: implications for atherosclerosis. FEBS Lett. 342, 4952.
  • Swanson R. A., Ying W. and Kauppinen T. M. (2004) Astrocyte influences on ischemic neuronal death. Curr. Mol. Med. 4, 193205.
  • Tang S. Y., Whiteman M., Jenner A., Peng Z. F. and Halliwell B. (2004) Mechanism of cell death induced by an antioxidant extract of Cratoxylum cochinchinense (YCT) in Jurkat T cells: the role of reactive oxygen species and calcium. Free Radic. Biol. Med. 36, 15881611.
  • Turrens J. F. (2003) Mitochondrial formation of reactive oxygen species. J. Physiol. 552, 335344.
  • Valentine W. M., Hill K. E., Austin L. M., Valentine H. L., Goldowitz D. and Burk R. F. (2005) Brainstem axonal degeneration in mice with deletion of selenoprotein P. Toxicol. Pathol. 33, 570576.
  • Van Bergen P., Rauhala P., Spooner C. M. and Chiueh C. C. (1999) Hemoglobin and iron-evoked oxidative stress in the brain: protection by bile pigments, manganese and S-nitrosoglutathione. Free Radic. Res. 31, 631640.
  • Van Maldergem L., Trijbels F., DiMauro S., Sindelar P. J., Musumeci O., Janssen A., Delberghe X., Martin J. J. and Gillerot Y. (2002) Coenzyme Q-responsive Leigh's encephalopathy in two sisters. Ann. Neurol. 52, 750754.
  • Van Remmen H., Ikeno Y., Hamilton M. et al. (2003) Life-long reduction in MnSOD activity results in increased DNA damage and higher incidence of cancer but does not accelerate aging. Physiol. Genomics 16, 2937.
  • Vatassery G. T., Adityanjee Quach H. T., Smith W. E., Kuskowski M. A. and Melnyk D. (2004) α and γ tocopherols in cerebrospinal fluid and serum from older, male, human subjects. J. Am. Coll. Nutr. 23, 233238.
  • Velliquette R. A., O'Connor T. and Vassar R. (2005) Energy inhibition elevates β-secretase levels and activity and is potentially amyloidogenic in APP transgenic mice: possible early events in Alzheimer's disease pathogenesis. J. Neurosci. 25, 1087410883.
  • Walther M., Holzhutter H. G., Kuban R. J., Wiesner R., Rathmann J. and Kuhn H. (1999) The inhibition of mammalian 15-lipoxygenases by the anti-inflammatory drug ebselen: dual-type mechanism involving covalent linkage and alteration of the iron ligand sphere. Mol. Pharmacol. 56, 196203.
  • Watanabe C. M., Wolffram S., Ader P., Rimbach G., Packer L., Maguire J. J., Schultz P. G. and Gohil K. (2001) The in vivo neuromodulatory effects of the herbal medicine Ginkgo biloba. Proc. Natl Acad. Sci. USA 98, 65776580.
  • Wentworth P. Jr, Nieva J., Takeuchi C. et al. (2003) Evidence for ozone formation in human atherosclerotic arteries. Science 302, 10531056.
  • Whittier J. E., Xiong Y., Rechsteiner M. C. and Squier T. C. (2004) Hsp90 enhances degradation of oxidized calmodulin by the 20 S proteasome. J. Biol. Chem. 279, 46 13546 142.
  • Wrona M. Z. and Dryhurst G. (1998) Oxidation of serotonin by superoxide radical: implications to neurodegenerative brain disorders. Chem. Res. Toxicol. 11, 639650.
  • Yan S. D. and Stern D. M. (2005) Mitochondrial dysfunction and Alzheimer's disease: role of amyloid-β peptide alcohol dehydrogenase (ABAD). Int. J. Exp. Pathol. 86, 161171.
  • Yew E. H., Cheung N. S., Choy M. S. et al. (2005) Proteasome inhibition by lactacystin in primary neuronal cells induces both potentially neuroprotective and pro-apoptotic transcriptional responses: a microarray analysis. J. Neurochem. 94, 943956.
  • Zbarsky V., Datla K. P., Parkar S., Rai D. K., Aruoma O. I. and Dexter D. T. (2005) Neuroprotective properties of the natural phenolic antioxidants curcumin and naringenin but not quercetin and fisetin in a 6-OHDA model of Parkinson's disease. Free Radic. Res. 39, 11191125.
  • Zecca L., Youdim M. B., Riederer P., Connor J. R. and Crichton R. R. (2004) Iron, brain ageing and neurodegenerative disorders. Nat. Rev. Neurosci. 5, 863873.
  • Zeevalk G. D., Bernard L. P., Song C., Gluck M. and Ehrhart J. (2005) Mitochondrial inhibition and oxidative stress: reciprocating players in neurodegeneration. Antioxid. Redox Signal. 7, 11171139.
  • Zhang X., Xie W., Qu S., Pan T., Wang X. and Le W. (2005) Neuroprotection by iron chelator against proteasome inhibitor-induced nigral degeneration. Biochem. Biophys. Res. Commun. 333, 544549.
  • Zhou B., Westaway S. K., Levinson B., Johnson M. A., Gitschier J. and Hayflick S. J. (2001) A novel pantothenate kinase gene (PANK2) is defective in Hallervorden–Spatz syndrome. Nat. Genet. 28, 345349.
  • Zingg J. M. and Azzi A. (2004) Non-antioxidant activities of vitamin E. Curr. Med. Chem. 11, 11131133.
  • Del Zoppo G. P. (2006) Stroke and neurovascular protection. N. Engl. J. Med. 354, 553555.