Antioxidants in Health and Disease J Clin Pathol: First Published As 10.1136/Jcp.54.3.176 on 1 March 2001

Home , Antioxidant

176 J Clin Pathol 2001;54:176–186 Antioxidants in health and disease J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from

I S Young, J V Woodside

Abstract induced tissue damage, and the function of Free radical production occurs continu- antioxidant defence systems in health and dis- ously in all cells as part of normal cellular ease. function. However, excess free radical production originating from endogenous Free radicals and their chemical or exogenous sources might play a role in reactions many diseases. Antioxidants prevent free A free radical can be defined as any molecular radical induced tissue damage by prevent- species capable of independent existence that ing the formation of radicals, scavenging contains an unpaired electron in an atomic them, or by promoting their decomposi- orbital.2 The presence of an unpaired electron tion. This article reviews the basic chem- results in certain common properties that are istry of free radical formation in the body, shared by most radicals. Radicals are weakly the consequences of free radical induced attracted to a magnetic field and are said to be tissue damage, and the function of anti- paramagnetic. Many radicals are highly reac- oxidant defence systems, with particular tive and can either donate an electron to or reference to the development of athero- extract an electron from other molecules, sclerosis. therefore behaving as oxidants or reductants. Department of Clinical (J Clin Pathol 2001;54:176–186) As a result of this high reactivity, most radicals Biochemistry, Institute have a very short half life (10−6 seconds or less) of Clinical Science, Keywords: free radicals; antioxidants; oxidative stress; coronary heart disease; atherosclerosis in biological systems, although some species Grosvenor Road, may survive for much longer.2 The most Belfast, Northern Ireland, BT12 6BJ, UK important free radicals in many disease states I S Young An antioxidant can be defined as: “any are oxygen derivatives, particularly superoxide substance that, when present in low concentra- and the hydroxyl radical. Radical formation in Department of tions compared to that of an oxidisable the body occurs by several mechanisms, Surgery, Royal Free involving both endogenous and environmental substrate, significantly delays or inhibits the and University College factors (fig 1). London Medical oxidation of that substrate”.1 The physiological Superoxide (O −.) is produced by the addi- School, 67–73 Riding role of antioxidants, as this definition suggests, 2 tion of a single electron to oxygen, and several House Street, London, is to prevent damage to cellular components W1P 7LD, UK mechanisms exist by which superoxide can be arising as a consequence of chemical reactions 3

J V Woodside produced in vivo. Several molecules, including http://jcp.bmj.com/ involving free radicals. In recent years, a adrenaline, ﬂavine nucleotides, thiol com- Correspondence to: substantial body of evidence has developed Professor Young, pounds, and glucose, can oxidise in the Department of Clinical supporting a key role for free radicals in many presence of oxygen to produce superoxide, and Biochemistry, Institute of fundamental cellular reactions and suggesting these reactions are greatly accelerated by the Clinical Science, Royal that oxidative stress might be important in the Group of Hospitals, presence of transition metals such as iron or Grosvenor Road, Belfast pathophysiology of common diseases including copper. The electron transport chain in the BT12 6BJ, UK atherosclerosis, chronic renal failure, and inner mitochondrial membrane performs the [email protected] diabetes mellitus. The aim of this review is to reduction of oxygen to water. During this on October 1, 2021 by guest. Protected copyright. Accepted for publication consider mechanisms of free radical formation process free radical intermediates are gener- 5 June 2000 in the body, the consequences of free radical ated, which are generally tightly bound to the components of the transport chain. However, Endogenous sources Environmental sources Free radical production there is a constant leak of a few electrons into •mitochondrial leak •cigarette smoke the mitochondrial matrix and this results in the •respiratory burst •pollutants 4 •enzyme reactions •UV light formation of superoxide. The activity of O –., H O •autooxidation reactions 2 2 2 •ionising radiation several other enzymes, such as cytochrome •xenobiotics p450 oxidase in the liver and enzymes involved Transition in the synthesis of adrenal hormones, also metals results in the leakage of a few electrons into the Fe2+, Cu+ surrounding cytoplasm and hence superoxide formation. There might also be continuous OH. production of superoxide by vascular endothelium to neutralise nitric oxide,56production of superoxide by other cells to regulate cell growth and diVerentiation,7 and the produc- Lipid peroxidationModified DNA bases Protein damage tion of superoxide by phagocytic cells during the respiratory burst.8 Any biological system generating superoxide will also produce hydrogen peroxide as a result Tissue damage of a spontaneous dismutation reaction. In Figure 1 Major sources of free radicals in the body and the consequences of free radical addition, several enzymatic reactions, includ- damage. ing those catalysed by glycolate oxidase and

www.jclinpath.com Antioxidants in health and disease 177

D-amino acid oxidase, might produce hydro- The net result of the reaction sequence illus- gen peroxide directly.9 Hydrogen peroxide is trated above is known as the Haber-Weiss reac- not a free radical itself, but is usually included tion. Although most iron and copper in the J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from under the general heading of reactive oxygen body are sequestered in forms that are not species (ROS). It is a weak oxidising agent that available to catalyse this reaction sequence, it is might directly damage proteins and enzymes still of importance as a mechanism for the for- containing reactive thiol groups. However, its mation of the hydroxyl radical in vivo. The most vital property is the ability to cross cell actual reactions, however, may be somewhat membranes freely, which superoxide generally more complex than those described above and cannot do.10 Therefore, hydrogen peroxide it is possible that other reactive intermediates such as the ferryl and perferryl radicals might formed in one location might diVuse a consid- 12 erable distance before decomposing to yield also be formed. the highly reactive hydroxyl radical, which is Approximately 4.5 g of iron can be found in the average adult man, most of which is likely to mediate most of the toxic eVects contained in the haemoglobin molecule and ascribed to hydrogen peroxide. Therefore, other haem containing proteins. Dietary iron is hydrogen peroxide acts as a conduit to transmit absorbed preferentially from the proximal part free radical induced damage across cell com- of the small intestine in the divalent form and is partments and between cells. In the presence of transferred to the circulation in which it is car- hydrogen peroxide, myeloperoxidase will gen- ried by transferrin.14 Under most circum- erate hypochlorous acid and singlet oxygen, a stances iron remains tightly bound to one of reaction that plays an important role in the 11 several proteins, including transferrin, lactofer- killing of bacteria by phagocytes. rin, haem proteins, ferritin, or haemosiderin. In . The hydroxyl radical (OH ), or a closely addition, however, it seems likely that a small related species, is probably the ﬁnal mediator of iron pool will be maintained as complexes with 12 most free radical induced tissue damage. All a variety of small molecules, such as nucleo- of the reactive oxygen species described above tides and citrate within the cytoplasm and exert most of their pathological eVects by subcellular organelles.14 This pool is probably giving rise to hydroxyl radical formation. The capable of catalysing an iron driven Fenton reason for this is that the hydroxyl radical reaction in vivo. Certainly, these complexes can reacts, with extremely high rate constants, with promote hydroxyl radical formation in vitro.15 almost every type of molecule found in living Redox reactive iron can be measured using the cells including sugars, amino acids, lipids, and bleomycin iron assay,16 although it remains nucleotides. Although hydroxyl radical forma- unclear to what extent iron detected by this tion can occur in several ways, by far the most assay correlates with any discrete anatomical or important mechanism in vivo is likely to be the physiological pool. In normal circumstances, transition metal catalysed decomposition of no bleomycin reactive iron is detectable in superoxide and hydrogen peroxide.13 plasma from healthy subjects, implying that

All elements in the first row of the d-block of transferrin or ferritin bound iron is not http://jcp.bmj.com/ the periodic table are classified as transition available to drive hydroxyl radical production.17 metals. In general, they contain one or more However, transferrin will release its iron at an unpaired electrons and are therefore them- acidic pH, particularly in the presence of small selves radicals when in the elemental state. molecular weight chelating agents such as 15 However, their key property from the point of ADP, ATP, and citrate. Such conditions are view of free radical biology is their variable found in areas of active inflammation and during ischaemia reperfusion injury,18 and it is

valency, which allows them to undergo reac- on October 1, 2021 by guest. Protected copyright. tions involving the transfer of a single electron. therefore likely that hydroxyl radicals contrib- The most important transition metals in ute to tissue damage in these settings. Iron is human disease are iron and copper. These ele- released from ferritin by reducing agents including ascorbate and superoxide itself,19 20 ments play a key role in the production of and hydrogen peroxide can release iron from a hydroxyl radicals in vivo.13 Hydrogen peroxide range of haem proteins.21 Therefore, although can react with iron II (or copper I) to generate the iron binding proteins eVectively chelate the hydroxyl radical, a reaction ﬁrst described iron and prevent any appreciable redox eVects by Fenton in 1894: under normal physiological conditions, this Fe2+ +HO →Fe3+ +OH. +OH− 2 2 protection can break down in disease states. This reaction can occur in vivo, but the situ- The role of copper is analogous to that ation is complicated by the fact that superoxide described above for iron.22 23 (the major source of hydrogen peroxide in vivo) Although free radical production occurs as a will normally also be present. Superoxide and consequence of the endogenous reactions hydrogen peroxide can react together directly described above and plays an important role in to produce the hydroxyl radical, but the rate normal cellular function, it is important to constant for this reaction in aqueous solution is remember that exogenous environmental fac- virtually zero. However, if transition metal ions tors can also promote radical formation. are present a reaction sequence is established Ultraviolet light will lead to the formation of that can proceed at a rapid rate: singlet oxygen and other reactive oxygen 3+ − → 2+ 24 Fe +O2 Fe +O2 species in the skin. Atmospheric pollutants 2+ → 3+ . − Fe +H2O2 Fe +OH +OH such as ozone and nitrogen dioxide lead to net result: radical formation and antioxidant depletion in − → − . O2 +H2O2 OH +OH +O2 the bronchoalveolar lining ﬂuid, and this may

www.jclinpath.com 178 Young, Woodside

Free radical production greatest activity is present in liver and erythro- Enzyme antioxidants cytes but some catalase is found in all tissues.

•Superoxide dismutases Metal binding J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from •Catalase –. proteins O2 , H2O2 Glutathione peroxidases and glutathione reductase •Glutathione peroxidase •Transferrin Glutathione peroxidases catalyse the oxidation •Caeruloplasmin •Ferritin Transition of glutathione at the expense of a hydroperox- •Lactoferrin metals ide, which might be hydrogen peroxide or Fe2+, Cu+ another species such as a lipid hydroperoxide34: → Chain breaking antioxidants ROOH + 2GSH GSSG + H2O+ROH •Directly scavenge free radicals Other peroxides, including lipid hydroperox- •Consumed during scavenging process ides, can also act as substrates for these enzymes, which might therefore play a role in Lipid phase Aqueous phase repairing damage resulting from lipid peroxi- •Tocopherols •Ascorbate OH. •Ubiquinol •Urate dation. Glutathione peroxidases require sele- •Carotenoids •Glutathione and other thiols nium at the active site, and deficiency might •Flavonoids occur in the presence of severe selenium deficiency.35 Several glutathione peroxidase Tissue damage enzymes are encoded by discrete genes.36 The plasma form of glutathione peroxidase is Repair mechanisms believed to be synthesised mainly in the Figure 2 Antioxidant defences against free radical attack. Antioxidant enzymes catalyse kidney.37 Within cells, the highest concentra- the breakdown of free radical species, usually in the intracellular environment. Transition tions are found in liver although glutathione metal binding proteins prevent the interaction of transition metals such as iron and copper with hydrogen peroxide and superoxide producing highly reactive hydroxyl radicals. Chain peroxidase is widely distributed in almost all breaking antioxidants are powerful electron donors and react preferentially with free radicals tissues. The predominant subcellular distribu- before important target molecules are damaged. In doing so, the antioxidant is oxidised and tion is in the cytosol and mitochondria, must be regenerated or replaced. By definition, the antioxidant radical is relatively unreactive and unable to attack further molecules. suggesting that glutathione peroxidase is the main scavenger of hydrogen peroxide in these exacerbate respiratory disease.25–27 Cigarette subcellular compartments. The activity of the smoke contains millimolar amounts of free enzyme is dependent on the constant availabil- radicals, along with other toxins.28 ity of reduced glutathione.38 The ratio of Various xenobiotics also cause tissue damage reduced to oxidised glutathione is usually kept as a consequence of free radical generation, very high as a result of the activity of the including paraquat,29 paracetamol,30 bleomy- enzyme glutathione reductase: cin,31 and anthracyclines.32 GSSG + NADPH + H+ → 2GSH + NADP+ The NADPH required by this enzyme to replenish the supply of reduced glutathione is Antioxidant defence systems provided by the pentose phosphate pathway. Because radicals have the capacity to react in Any competing pathway that utilises NADPH an indiscriminate manner leading to damage to (such as the aldose reductase pathway) might http://jcp.bmj.com/ almost any cellular component, an extensive lead to a deficiency of reduced glutathione and range of antioxidant defences, both endog- hence impair the action of glutathione peroxi- enous and exogenous, are present to protect dase. Glutathione reductase is a flavine nucle- cellular components from free radical induced otide dependent enzyme and has a similar damage. These can be divided into three main tissue distribution to glutathione peroxidase.39 groups: antioxidant enzymes, chain breaking antioxidants, and transition metal binding pro- Superoxide dismutase

2 on October 1, 2021 by guest. Protected copyright. teins (fig 2). The superoxide dismutases catalyse the dismutation of superoxide to hydrogen peroxide: − − → THE ANTIOXIDANT ENZYMES O2 +O2 + 2H+ H2O2 +O2 Catalase The hydrogen peroxide must then be Catalase was the first antioxidant enzyme to be removed by catalase or glutathione peroxidase, characterised and catalyses the two stage as described above. There are three forms of conversion of hydrogen peroxide to water and superoxide dismutase in mammalian tissues, oxygen: each with a specific subcellular location and → catalase–Fe(III) + H2O2 compound I diVerent tissue distribution. → compoundI+H2O2 catalase–Fe(III) + (1) Copper zinc superoxide dismutase (CuZn-

2H2O+O2 SOD): CuZnSOD is found in the cyto- Catalase consists of four protein subunits, plasm and organelles of virtually all mam- each containing a haem group and a molecule malian cells.40 It has a molecular mass of of NADPH.33 The rate constant for the approximately 32 000 kDa and has two reactions described above is extremely high protein subunits, each containing a cata- (∼107 M/sec), implying that it is virtually lytically active copper and zinc atom. impossible to saturate the enzyme in vivo. (2) Manganese superoxide dismutase Catalase is largely located within cells in (MnSOD): MnSOD is found in the mito- peroxisomes, which also contain most of the chondria of almost all cells and has a enzymes capable of generating hydrogen per- molecular mass of 40 000 kDa.41 It con- oxide. The amount of catalase in cytoplasm sists of four protein subunits, each prob- and other subcellular compartments remains ably containing a single manganese atom. unclear, because peroxisomes are easily rup- The amino acid sequence of MnSOD is tured during the manipulation of cells. The entirely dissimilar to that of CuZnSOD

www.jclinpath.com Antioxidants in health and disease 179

and it is not inhibited by cyanide, allowing peripheral neuropathy that occurs in abetalipo- MnSOD activity to be distinguished from proteinaemia.51 that of CuZnSOD in mixtures of the two The absorption, transport, and regulation of J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from enzymes. plasma concentrations of vitamin E in humans (3) Extracellular superoxide dismutase (EC- has been reviewed by Kayden and Traber,52 SOD): EC-SOD was described by Mark- although in general the metabolism of vitamin lund in 198242 and is a secretory copper E is not well described. In cell membranes and and zinc containing SOD distinct from the lipoproteins the essential antioxidant function CuZnSOD described above. EC-SOD is of vitamin E is to trap peroxyl radicals and to synthesised by only a few cell types, break the chain reaction of lipid peroxidation.53 including fibroblasts and endothelial cells, Vitamin E will not prevent the initial formation and is expressed on the cell surface where of carbon centred radicals in a lipid rich it is bound to heparan sulphates. EC-SOD environment, but does minimise the formation is the major SOD detectable in extracellu- of secondary radicals. á-Tocopherol is the most lar fluids and is released into the circula- potent antioxidant of the tocopherols and is tion from the surface of vascular endothe- also the most abundant in humans. It quickly 43 lium following the injection of heparin. reacts with a peroxyl radical to form a relatively EC-SOD might play a role in the regula- stable tocopheroxyl radical, with the excess tion of vascular tone, because endothelial charge associated with the extra electron being derived relaxing factor (nitric oxide or a dispersed across the chromanol ring. This closely related compound) is neutralised in 44 resonance stabilised radical might subse- the plasma by superoxide. quently react in one of several ways. á-Tocopherol might be regenerated by reaction 54 THE CHAIN BREAKING ANTIOXIDANTS at the aqueous interface with ascorbate or Whenever a free radical interacts with another another aqueous phase chain breaking antioxi- molecule, secondary radicals may be generated dant, such as reduced glutathione or urate.55 that can then react with other targets to Alternatively, two á-tocopheroxyl radicals produce yet more radical species. The classic might combine to form a stable dimer, or the example of such a chain reaction is lipid radical may be completely oxidised to form peroxidation, and the reaction will continue to tocopherol quinone. propagate until two radicals combine to form a The carotenoids are a group of lipid soluble stable product or the radicals are neutralised by antioxidants based around an isoprenoid car- a chain breaking antioxidant.45 Chain breaking bon skeleton.56 The most important of these is antioxidants are small molecules that can â-carotene, although at least 20 others may be receive an electron from a radical or donate an present in membranes and lipoproteins. They electron to a radical with the formation of are particularly eYcient scavengers of singlet stable byproducts.46 In general, the charge oxygen,57 but can also trap peroxyl radicals at associated with the presence of an unpaired low oxygen pressure with an eYciency at least http://jcp.bmj.com/ electron becomes dissociated over the scaven- as great as that of á-tocopherol. Because these ger and the resulting product will not readily conditions prevail in many biological tissues, accept an electron from or donate an electron the carotenoids might play a role in preventing to another molecule, preventing the further in vivo lipid peroxidation.58 The other impor- propagation of the chain reaction. Such tant role of certain carotenoids is as precursors antioxidants can be conveniently divided into of vitamin A (retinol). Vitamin A also has anti- aqueous phase and lipid phase antioxidants. oxidant properties,59 which do not, however, show any dependency on oxygen concentra- on October 1, 2021 by guest. Protected copyright. Lipid phase chain breaking antioxidants tion. These antioxidants scavenge radicals in mem- Flavonoids are a large group of polyphenolic branes and lipoprotein particles and are crucial antioxidants found in many fruits, vegetables, in preventing lipid peroxidation. The most and beverages such as tea and wine.60–62 Over important lipid phase antioxidant is probably 4000 flavonoids have been identified and they vitamin E.47 Vitamin E occurs in nature in eight are divided into several groups according to diVerent forms, which diVer greatly in their their chemical structure, including flavonols degree of biological activity. The tocopherols (quercetin and kaempherol), flavanols (the cat- (á, â, ã, and ä) have a chromanol ring and a echins), flavones (apigenin), and isoflavones phytyl tail, and diVer in the number and (genistein). Epidemiological studies suggest an position of the methyl groups on the ring. The inverse relation between flavonoid intake and tocotrienols (á, â, ã, and ä) are structurally incidence of chronic diseases such as coronary similar but have unsaturated tails. Both classes heart disease (CHD).63–65 However, little is cur- of compounds are lipid soluble and have rently known about the absorption and me- pronounced antioxidant properties.48 They tabolism of flavonoids and epidemiological react more rapidly than polyunsaturated fatty associations might be a consequence of con- acids with peroxyl radicals and hence act to founding by other factors. Available evidence break the chain reaction of lipid peroxidation. suggests that the bioavailability of many In addition to its antioxidant role, vitamin E flavonoids is poor,66–68 and plasma values very might also have a structural role in stabilising low, although there is some evidence that aug- membranes.49 Frank vitamin E deficiency is menting the intake of flavonoids might improve rare in humans, although it might cause biochemical indices of oxidative damage.68 69 haemolysis50 and might contribute to the Apart from flavonoids, other dietary phenolic

www.jclinpath.com 180 Young, Woodside

compounds might also make a small contribu- be particularly important in providing protection to total antioxidant capacity.70 tion against certain oxidising agents, such as Ubiquinol-10, the reduced form of coen- ozone.82 Indeed, it has been suggested that the J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from zyme Q10, is also an eVective lipid soluble increase in life span that has occurred during chain breaking antioxidant.71 Although present human evolution might be partly explained by in lower concentrations than á-tocopherol, it the protective action provided by uric acid in can scavenge lipid peroxyl radicals with higher human plasma.83 Part of the antioxidant eVect eYciency than either á-tocopherol or the caro- of urate might be attributable to the formation tenoids, and can also regenerate membrane of stable non-reactive complexes with iron, but bound á-tocopherol from the tocopheryl radi- it is also a direct free radical scavenger. cal.72 Indeed, whenever plasma or isolated low Albumin bound bilirubin is also an eYcient density lipoprotein (LDL) cholesterol is ex- radical scavenger,84 and it has been suggested posed to radicals generated in the lipid phase, that it might play a particularly crucial role in ubiquinol-10 is the first antioxidant to be con- protecting the neonate from oxidative dam- sumed, suggesting that it might be of particular age,85 because deficiency of other chain break- importance in preventing the propagation of ing antioxidants is common in the newborn. lipid peroxidation.73 However, work to clarify The other major chain breaking antioxidants further its role has been hampered by the ease in plasma are the protein bound thiol groups. with which ubinquinol-10 becomes oxidised The sulphydryl groups present on plasma pro- during sample handling or analysis. teins can function as chain breaking antioxidants by donating an electron to neutralise a Aqueous phase chain breaking antioxidants free radical, with the resultant formation of a These antioxidants will directly scavenge protein thiyl radical. Albumin is the predomi- radicals present in the aqueous compartment. nant plasma protein and makes the major con- Qualitatively the most important antioxidant of tribution to plasma sulphydryl groups, al- this type is vitamin C (ascorbate).74 In humans, though it also has several other antioxidant 86 ascorbate acts as an essential cofactor for properties. Albumin contains 17 disulphide several enzymes catalysing hydroxylation reac- bridges and has a single remaining cysteine tions. In most cases, it provides electrons for residue, and it is this residue that is responsible enzymes that require prosthetic metal ions in a for the capacity of albumin to react with and 87 reduced form to achieve full enzymatic activity. neutralise peroxyl radicals. This property is Its best known role is as a cofactor for prolyl important in view of the role albumin plays in and lysyl oxidases in the synthesis of collagen. transporting free fatty acids in the blood. In However, in addition to these well defined addition, albumin has the capacity to bind actions, several other biochemical pathways copper ions and will inhibit copper dependent depend upon the presence of ascorbate.75 In lipid peroxidation and hydroxyl radical forma- addition to its role as an enzyme cofactor, the tion. It is also a powerful scavenger of the phagocytic product hypochlorous acid, and other major function of ascorbate is as a key http://jcp.bmj.com/ chain breaking antioxidant in the aqueous provides the main plasma defence against this phase.76 Ascorbate has been shown to scavenge oxidant.88 superoxide, hydrogen peroxide, the hydroxyl Because albumin itself is damaged when it radical, hypochlorous acid, aqueous peroxyl acts as an antioxidant, it has been viewed as a radicals, and singlet oxygen. During its antioxi- sacrificial molecule that prevents damage dant action, ascorbate undergoes a two elec- occurring to more vital species.86 The high tron reduction, initially to the semidehy- plasma concentration of albumin and a rela- droascorbyl radical and subsequently to tively short half life mean that any damage suf- on October 1, 2021 by guest. Protected copyright. dehydroascorbate. The semidehydroascorbyl fered is unlikely to be of biological importance. radical is relatively stable owing to dispersion of However, in vitro work has shown that protein the charge associated with the presence of a thiyl radicals can themselves act as a potential single electron over the three oxygen atoms, source of reactive oxidants. The thiyl radical and it can be readily detected by electron spin can abstract an electron from a polyunsatu- resonance in body fluids in the presence of rated fatty acid to initiate the process of lipid increased free radical production.77 Dehy- peroxidation,89 a reaction that can be inhibited droascorbate is relatively unstable and hydro- by ascorbate and retinol. The antioxidant lyses readily to diketogulonic acid, which is eVects of albumin and other proteins have been subsequently broken down to oxalic acid. Two shown to decrease at high concentrations and it mechanisms have been described by which has been suggested that this is because thiyl dehydroascorbate can be reduced back to radicals can oxidatively damage other mol- ascorbate; one is mediated by the selenoen- ecules. The importance of these findings to the zyme thioredoxin reductase78 and the other is a antioxidant role of albumin in vivo remains non-enzyme mediated reaction that uses re- unclear. duced glutathione.79 Dehydroascorbate in Reduced glutathione (GSH) is a major plasma is probably rapidly taken up by red source of thiol groups in the intracellular com- blood cells before recycling, so that very little, partment but is of little importance in the if any, dehydroascorbate is present in plasma.80 extracellular space.90 GSH might function Apart from ascorbate, other antioxidants are directly as an antioxidant, scavenging a variety present in plasma in high concentrations. Uric of radical species, as well as acting as an essen- acid eYciently scavenges radicals, being con- tial factor for glutathione peroxidase (discussed verted in the process to allantoin.81 Urate might above). Thioredoxin might also function as a

www.jclinpath.com Antioxidants in health and disease 181

key intracellular antioxidant, particularly in rise to a variety of products including short redox induced activation of transcription fac- chain aldehydes such as malondialdehyde or tors.91 4-hydroxynonenal, alkanes, and alkenes, conju- J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from gated dienes, and a variety of hydroxides and 45 INTERACTIONS BETWEEN CHAIN BREAKING hydroperoxides. Many of these products can ANTIOXIDANTS be measured as markers of lipid peroxidation. Although the actions of chain breaking antioxi- Detailed discussion of this complex issue is dants have been considered separately above, it outside the scope of this review, but the is vital to remember that in vivo complex inter- measurement of isoprostanes by gas chroma- actions between antioxidants are likely to tography mass spectroscopy is probably the occur. For instance, it is likely that ascorbate most specific marker of free radical damage to will recycle the tocopheryl radical at the lipids.95 Oxidative damage to proteins and aqueous–lipid interface, so regenerating toco- nucleic acids similarly gives rise to a variety of pherol.54 This might be crucial in ensuring that specific damage products as a result of modifi- tocopherol concentrations are maintained in cations of amino acids or nucleotides.45 Such lipoproteins and membranes. In a similar man- oxidative damage might also lead to cellular ner, glutathione can regenerate ascorbate from dysfunction, and it is this that might contribute dehydroascorbate. A complex interplay is to the pathophysiology of a wide variety of dis- therefore likely to exist between antioxidants, eases. making it diYcult to predict how antioxidants will function in vivo. It therefore becomes meaningless to ask which antioxidant is most important: the answer will depend on the Oxidative stress and disease circumstances existing in a particular microen- A role for oxidative stress has been postulated vironment at a specific time, and on the nature in many conditions, including atherosclerosis, of the oxidant injury taking place. inflammatory conditions,96 certain cancers,97 A second important property of chain and the process of aging.98 In many cases, this breaking antioxidants is their ability to act as follows the observation of increased amounts pro-oxidants. In certain circumstances, the of free radical damage products, particularly presence of an antioxidant might paradoxically markers of lipid peroxidation, in body fluids. It lead to increased oxidative damage. For is important to remember, however, that lipid instance, it has been reported that the adminis- peroxidation is an inevitable accompaniment of tration of vitamin C can sometimes lead to an cell death from any cause. In most cases increase in oxidative damage, particularly if peroxidation is a secondary phenomenon, and iron is also administered.92 Similarly, it has this does not therefore directly indicate an been clearly shown in vitro that tocopherol important role for oxidative stress in the disease might promote LDL oxidation in the absence concerned. If a primary role for oxidative stress of an aqueous phase antioxidant such as ascor- in a particular setting is to be sustained, there 93

bate. Whether these reactions are important should be a plausible mechanism by which http://jcp.bmj.com/ in vivo is as yet unclear. However, the increased free radical production or a decrease possibility that antioxidants may have proin antioxidant defences might occur. In addi- oxidant eVects in vivo must be considered tion, evidence of oxidative stress should be when designing and interpreting the results of detectable before the onset of tissue damage clinical trials of antioxidant supplementation. and augmentation of antioxidant status at an early stage should either prevent or greatly THE TRANSITION METAL BINDING PROTEINS reduce tissue damage. As discussed above, transition metal binding Atherosclerosis can be taken as an example on October 1, 2021 by guest. Protected copyright. proteins (ferritin, transferrin, lactoferrin, and of a process for which there is substantial caeruloplasmin) act as a crucial component of evidence of a role for oxidative stress. Hyperc- the antioxidant defence system by sequestering holesterolaemia is universally accepted as a iron and copper so that they are not available to major risk factor for atherosclerosis. However, drive the formation of the hydroxyl radical. at any given concentration of plasma choles- The main copper binding protein, caeruloplas- terol, there is still great variability in the occur- min, might also function as an antioxidant rence of cardiovascular events. One of the enzyme that can catalyse the oxidation of diva- major breakthroughs in atherogenesis research lent iron.94 has been the realisation that oxidative modiﬁ- 2+ + → 3+ 4Fe +O2 +4H 4Fe +2H2O cation of LDL might be a crucially important Fe2+ is the form of iron that drives the Fenton step in the development of the atherosclerotic reaction and the rapid oxidation of Fe2+ to the plaque.99 100 The formation of foam cells from less reactive Fe3+ form is therefore an antioxi- monocyte derived macrophages in early dant eVect. atherosclerotic lesions is not caused by native LDL but only after the modiﬁcation of LDL by Consequences of oxidative damage various chemical reactions such as oxidation. Oxidative stress, arising as a result of an imbal- Oxidation of LDL is a process initiated and ance between free radical production and anti- propagated by free radicals or by one of several oxidant defences, is associated with damage to enzymes,101 and is believed to occur mainly in a wide range of molecular species including the arterial wall in a microenvironment where lipids, proteins, and nucleic acids. Lipoprotein antioxidants may become depleted. All the cells particles or membranes characteristically un- of the vessel wall—endothelial cells, smooth dergo the process of lipid peroxidation, giving muscle cells, macrophages, and lymphocytes—

www.jclinpath.com 182 Young, Woodside

can modify LDL in vitro.102–104 Several mecha- healthy subjects or patients with CHD can nisms are likely to be involved, including tran- reduce levels of free radical damage products sition metal ion mediated generation of and protect LDL against oxidation.120 121 Vita- J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from hydroxyl radicals, the production of reactive min E appears to be the most eVective antioxi- oxygen species by enzymes such as myeloper- dant; both â-carotene and vitamin C have pro- oxidase and lipoxygenase, and direct modifica- duced extensions in lag time to oxidation only tion by reactive nitrogen species. Because in a few studies, although it remains possible oxidation of LDL is primarily a free radical that they might have a beneficial eVect in indi- mediated process that is inhibited by antioxi- viduals with poor baseline status. Third, large dants, antioxidant depletion might be a risk scale epidemiological studies generally show factor for cardiovascular disease (CVD). that low intakes of antioxidants are associated Evidence for LDL oxidation in vivo is now with increased cardiovascular risk after correct- well established. In immunocytochemical stud- ing for other risk factors.122–125 The epidemio- ies, antibodies against oxidised LDL stain logical evidence is strongest in the case of vita- atherosclerotic lesions but not normal arterial min E. In particular, two large longitudinal 105 tissue. LDL extracted from animal and studies in the USA examined the association human lesions has been shown to be oxidised between antioxidant intake and the risk of and is rapidly taken up by macrophage scaven- 106 CHD. In a group of 39 910 male health ger receptors. In young survivors of myocar- professionals, men who took vitamin E supple- dial infarction (MI), an association has been ments in doses of at least 100 IU/day for over demonstrated between increased susceptibility two years had a 37% lower relative risk of CHD of LDL to oxidation and the degree of 107 than those who did not take vitamin E supple- coronary atherosclerosis, whereas the pres- ments, after adjustment for age, coronary risk ence of ceroid, a product of lipid peroxidation, factors, and intake of vitamin C and has been shown in advanced atherosclerotic â-carotene.126 In the nurses’ health study of plaques.108 87 245 female nurses, women who took Oxidised LDL has several properties that vitamin E supplements for more than two years promote atherogenesis, apart from its rapid had a 41% lower relative risk of major coronary uptake into macrophages via the scavenger disease.127 This eVect persisted after adjust- receptor. Oxidised forms of LDL are chemotactic for circulating macrophages and smooth ment for age, smoking, obesity, exercise, blood muscle cells and facilitate monocyte adhesion pressure, cholesterol, and the use of postmeno- to the endothelium and entry into the suben- pausal oestrogen replacement, aspirin, vitamin dothelial space.109 Oxidised LDL is also C, and â-carotene. High vitamin E intakes cytotoxic towards arterial endothelial cells110 from dietary sources were not associated with a and inhibits the release of nitric oxide and the significant decrease in risk, although even the resulting endothelium dependent vasodila- highest dietary vitamin E intakes were far lower tion.111 Therefore, there is a potential role for than intakes among supplement users. The evidence linking the water soluble vita- oxidised LDL in altering vasomotor responses, http://jcp.bmj.com/ perhaps contributing to vasospasm in diseased min C with CVD is less strong than for vitamin vessels. In addition, oxidised LDL is immuno- E. In the physicians’ follow-up study, a high genic; autoantibodies against various epitopes intake of vitamin C was not associated with a of oxidised LDL have been found in human lower risk of CHD in men, whereas in women serum.112 113 and immunoglobulin (IgG) spe- from the nurses’ health survey, an initial eVect cific for epitopes of oxidised LDL can be found was attenuated after adjustment for multivita- in lesions.114 Oxidised LDL can induce arterial min use. Only one prospective study involving wall cells to produce chemotactic factors, 11 348 adults demonstrated an inverse relation on October 1, 2021 by guest. Protected copyright. adhesion molecules, cytokines, and growth between vitamin C intake and overall cardio- factors that have a role to play in the develop- vascular mortality.128 This eVect resulted ment of the plaque.115 116 largely from the use of vitamin C in supple- Apart from the atherogenic consequences of ments and might have been caused by other LDL oxidation, it is increasingly recognised antioxidant vitamins in multivitamin prepara- that reactive oxygen and nitrogen species tions. A prospective population study of 1605 directly interact with signalling mechanisms in healthy men aged 42, 48, 54, or 60 years in the arterial wall to regulate vascular function.117 Finland has recently shown that men who had The activities of oxidant generating enzymes in vitamin C deficiency had a relative risk of MI of the arterial wall are regulated by both receptor 2.5 compared with men with higher plasma activation and by non-receptor mediated path- vitamin C concentrations, after adjustment for ways. The eVects of antioxidants on these other risk factors.129 There is also some indica- processes are complex but provide alternative tion that increased dietary intake of â-carotene mechanisms by which antioxidant supplemen- is associated with reduced risk of CHD, tation might ameliorate vascular pathology, for although again the evidence is less convincing instance by improving endothelial function. than that for vitamin E. In the prospective Evidence that antioxidant micronutrients nurses’ health survey, consumption of vitamins potentially reduce the risk of CHD comes from A and â-carotene in food and supplements four major sources. First, studies of antioxidant weakly predicted the incidence of CHD; supplementation in animal models of athero- Gaziano and Hennekens calculated a 22% risk sclerosis have generally shown a reduction in reduction for women in the highest quintile of disease.118 119 Second, many studies have now â-carotene compared with those in the low- shown that antioxidant supplementation in est.130

www.jclinpath.com Antioxidants in health and disease 183

Thus, there is a plausible case supported by harm (or benefit) among the 11% of partici- experimental studies, animal experiments, and pants who were current smokers at baseline, epidemiology linking oxidative stress and although small eVects could not be ruled out. J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from atherosclerosis. The key test of such a hypoth- Greenberg et al studied the eVect of esis is whether increased antioxidant intake can â-carotene supplementation (50 mg/day) in be shown to prevent the clinical manifestations 1720 male and female subjects for a median of atherosclerosis in humans. Several published period of 4.3 years with a median follow up of randomised studies have now considered this 8.2 years.136 Subjects whose plasma values of issue, and others are currently ongoing. Early â-carotene were in the highest quartile at the results have not been encouraging. beginning of the study had the lowest risk of The á-tocopherol, â-carotene cancer preven- death from all causes compared with those in tion trial (ATBC), conducted among 29 133 the lowest quartile. However, supplementation male heavy smokers in Finland, found no had no eVect on either all cause or cardiovas- reduction in CHD morbidity or mortality dur- cular mortality. Thus for â-carotene supple- ing five to eight years of treatment with vitamin mentation, it would appear that there are no 131 E (50 mg daily) or â-carotene (20 mg daily). overall benefits among those individuals with a Those assigned vitamin E had no significant good nutritional status who are at low or aver- decrease in deaths from ischaemic heart age risk of developing CHD. The situation disease (IHD), but a 50% excess of deaths from might be diVerent, however, for those with a cerebral haemorrhage, whereas those assigned previous history of such disease. to â-carotene experienced an 11% increase in Hodis et al have shown a reduction in CAD deaths from IHD. In a further analysis, a progression (as measured angiographically) in subgroup of the original subjects who had suf- men given 100 IU vitamin E daily, although no 132 fered a previous MI were considered. The benefit was found for vitamin C.137 Singh et al endpoint of this substudy was the first major found that a combination of vitamins A, C, E, coronary event after randomisation. The pro- and â-carotene administered within a few portion of major coronary events did not hours after acute MI and continued for 28 days decrease with either á-tocopherol or led to significantly fewer cardiac events and a â-carotene supplements. In fact, â-carotene lower incidence of angina pectoris in the conferred an excess of fatal IHD (75% increase supplemented group.138 The Cambridge heart in risk). There was a beneficial eVect of vitamin antioxidant study (CHAOS), a trial of vitamin E on non-fatal MI with a risk reduction of E supplementation on 2002 patients with 38%. By contrast, in the Chinese cancer angiographic evidence of coronary disease, was prevention study conducted among 29 584 carried out with a mean treatment duration of poorly nourished residents of Linxian, China, 1.4 years.139 It was found that this short term those randomised for 5.25 years to a combined supplementation with á-tocopherol (268 or regimen of 15 mg/day â-carotene, 30 mg/day 537 mg/day) reduced CHD morbidity in pa- vitamin E, and 50 µg/day selenium had a tients, in that patients had a significantly (77%) http://jcp.bmj.com/ significant 9% reduction in total mortality, a decreased risk of subsequent non-fatal MI. significant 21% decrease in stomach cancer However, no benefit was found in terms of deaths, and a non-significant 10% decrease in cardiovascular mortality, with a non-significant cerebrovascular mortality.133 However, the wis- excess among vitamin E allocated participants. dom of generalising these findings to well The GISSI-P study randomised 11 324 men nourished populations remains uncertain. surviving a myocardial infarction to 300 mg The â-carotene and retinol eYcacy trial (CARET), designed to test the eVects of a vitamin E, 1 g n-3 polyunsaturated fatty acids on October 1, 2021 by guest. Protected copyright. (PUFAs), both, or neither in a randomised, combined supplement of 30 mg â-carotene 140 and 25 000 IU retinol daily among 18 314 placebo controlled trial. Results suggested a cigarette smokers and individuals with occupa- beneficial eVect of n-3 PUFAs but no benefit tional asbestos exposure, was ended early when with vitamin E (p = 0.07). However, further researchers recognised a raised risk of death analysis of secondary endpoints suggested from lung cancer in those receiving â-carotene some beneficial eVects of vitamin E. In and, again, no beneficial eVect on CVD was addition, the eVect of vitamin E might have found.134 For CVD mortality, there was a non- been ameliorated by the Mediterranean diet of significant 26% increase in the treated group the subjects. Neither of these qualifications 141 (p = 0.06). holds true for the HOPE study, which The physicians’ health study followed more recruited over 9000 subjects likely to be eating than 22 000 US male doctors treated with a typical northern European diet, who were at 50 mg â-carotene or placebo every other day high risk for cardiovascular events because they for an average of 12 years. The trial appears to had CVD or diabetes in addition to one other have been conducted meticulously and its risk factor. Subjects were randomly assigned results seriously question any beneficial eVect according to a two by two factorial design to with such supplementation on CVD in well receive either 400 IU of vitamin E daily from nourished populations. There were no signifi- natural sources or matching placebo, and either cant eVects on individual outcomes, or on a an angiotensin converting enzyme inhibitor combined endpoint of non-fatal MI, non-fatal (ramipril) or matching placebo for a mean of stroke, and cardiovascular death, for which the 4.5 years. Vitamin E supplementation had no relative risk was 1.0 (95% confidence interval, eVect on primary or secondary cardiovascular 0.91 to 1.09).135 There was also no evidence of endpoints.

www.jclinpath.com 184 Young, Woodside

Thus, for vitamin E in Western populations, of antioxidant defence systems plays a key role the only available trial data in primary preven- in protecting against oxidative damage. These tion are from the ATBC trial, which show no processes appear to be disordered in many J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from eVect. In secondary prevention, the accumulat- conditions, and a plausible hypothesis may be ing trial data for vitamin E are less consistent, constructed implicating oxidative stress as a although not particularly encouraging. The cause of tissue damage. However, as illustrated CHAOS study was positive, although it suVers by the example of CHD, attempts to intervene from design limitations. The GISSI-P study therapeutically by using antioxidant supple- gave a borderline result, whereas the HOPE ments have so far been largely unsuccessful. A study was unequivocally negative. more complete understanding of the biochemi- How should we interpret the discordance cal events occurring at a cellular level to influ- between data from cohort studies and the ence oxidative damage is required to guide results so far available from clinical trials? In future therapeutic advances. general, it might be that the duration of clinical trials is too short to show a benefit, and that 1 Halliwell B, Gutteridge JC. The definition and measure- antioxidant intake over many years is required ment of antioxidants in biological systems. Free Radic Biol to prevent atherosclerosis. Thought needs to be Med 1995;18:125–6. 2 Halliwell B; Gutteridge JM. Free radicals in biology and medi- given to trial design, with dose, duration of cine, 2nd ed. Oxford: Clarendon Press, 1989. treatment and follow up period, initial antioxi- 3 Halliwell B, Gutteridge JC. Biologically relevant metal ion- dependent hydroxyl radical generation—an update. FEBS dant values and dietary intake, and extent and Lett 1992;307:108–12. distribution of existing atherosclerosis being 4 Becker LB, Vanden Hoek TL, Shao ZH, et al. Generation of superoxide in cardiomyocytes during ischemia before taken into consideration. Animal models have reperfusion. Am J Physiol 1999;277:H2240–6. nearly always tested the eVects of antioxidants 5 Barbacanne MA, Margeat E, Arnal JF, et al. Superoxide release by confluent endothelial cells, an electron spin reso- on the early atherosclerotic lesions. Whether or nance (ESR) study. J Chim Phys Phys Chim Biol not antioxidants have inhibitory eVects on the 1999;96:85–92. 6 Tsao PS, Heidary S, Wang A, et al. Protein kinase C-epsilon later stages remains to be seen. In addition, the mediates glucose-induced superoxide production and complex mixture of antioxidant micronutrients MCP-1 expression in endothelial cells. FASEB J 1998;12: 512. found in a diet high in fruit and vegetable 7 Masters CJ. Cellular signalling: the role of the peroxisome. intake might be more eVective than large doses Cell Signal 1996;8:197–208. 8 Curnutte JT, Babior BM. Chronic granulomatous disease. of a small number of antioxidant vitamins. It Adv Hum Genet 1987;16:229–45. could be that several of these compounds work 9 Chance B, Sies H, Boveris A. Hydroperoxide metabolism in mammalian organs. Physiol Rev 1979;59:527–605. together but have no eVect individually, or that 10 Halliwell B, Gutteridge JMC. Role of free radicals and cata- other dietary components (such as trace lytic metal ions in human disease: an overview. Methods Enzymol 1990;186:1–85. elements) might be eVectors of antioxidant 11 Tatsuzawa H, Maruyama T, Hori K, et al. Singlet oxygen as action. The trial evidence available so far the principal oxidant in myeloperoxidase-mediated bacte- rial killing in neutrophil phagosomes. Biochim Biophys Res relates only to á-tocopherol and â-carotene. Commun 1999;262:647–50. Although eVective at protecting against lipid 12 Lloyd RV, Hanna PM, Mason RP. The origin of the hydroxyl radical oxygen in the Fenton reaction. Free Radic peroxidation, these antioxidants have little Biol Med 1997;22:885–8.

eVect on arterial endothelial function. Ascor- 13 Stohs SJ, Bagchi D. Oxidative mechanisms in the toxicity of http://jcp.bmj.com/ metal ions. Free Radic Biol Med 1995;18:321–36. bate, in contrast, seems more eVective in 14 Goldenberg HA. Regulation of mammalian iron improving endothelial function, although there metabolism: current state and need for further knowledge. Crit Rev Clin Lab Sci 1997;34:529–72. is less epidemiological support for a protective 15 Aruoma OI, Halliwell B. Superoxide-dependent and eVect of ascorbate. ascorbate-dependent formation of hydroxyl radicals from Alternatively, the signiﬁcant results linking hydrogen peroxide in the presence of iron. Biochem J 1987; 241:273–8. antioxidant intake with CHD risk observed in 16 Gutteridge JM, Rowley DA, Halliwell B. Superoxide- cohort studies might be the result of confound- dependent formation of hydroxyl radicals in the presence of iron salts. Biochem J 1981;199:263–5. on October 1, 2021 by guest. Protected copyright. ing with other lifestyle behaviours. Slattery et al 17 Gutteridge JC, Quinlan GJ, Evans TW. Transient iron over- examined dietary antioxidants and plasma lip- load with bleomycin detectable iron in the plasma of patients with adult respiratory distress syndrome. Thorax ids in the coronary artery risk development in 1994;49:707–10. young adults (CARDIA) study and found that 18 Pepper JR, Mumby S, Gutteridge JC. Sequential oxidative damage, and changes in iron binding and iron oxidizing a higher intake of antioxidants was associated plasma antioxidants during cardiopulmonary bypass sur- with other lifestyle factors such as physical gery. Free Radic Res 1994;21:377–85. 142 19 Gutteridge JMC, Quinlan GJ. Malondialdehyde formation activity and non-smoking. Plasma concen- from lipid hydroperoxides in the thiobarbituric acid test: trations of antioxidants are linked with social the role of lipid radicals, iron salts and metal chelators. J Appl Biochem 1983;5:293–9. class, being higher in more aZuent groups. 20 O’Connell M, Halliwell B, Moorhouse C, et al. Formation Although these variables can be individually of hydroxyl radicals in the presence of ferritin and haemosiderin. Is haemosiderin formation a biological protective controlled for in analyses, it might be that a mechanism? Biochem J 1986;234:727–31. complex lifelong behaviour pattern needs to be 21 Benatti U, Morelli A, Guida L, et al. The production of acti- vated oxygen species by an interaction of methemoglobin studied before conclusions regarding antioxi- with ascorbate. Biochem Biophys Res Commun 1983;111: dants and CHD can be made. For example, 980–7. 22 Swain J, Gutteridge JM. Prooxidant iron and copper, with passive smoking has recently been shown to ferroxidase and xanthine oxidase activities in human have an atherogenic eVect on LDL,143 yet atherosclerotic material. FEBS Lett 1995;368:513–15. 23 Koizumi M, Fujii J, Suzuki K, et al. A marked increase in exposure to smoke is a diYcult lifestyle variable free copper levels in the plasma and liver of LEC rats: an to control for in cohort analyses. animal model for Wilson disease and liver cancer. Free Radic Res 1998;28:441–50. 24 McCaughan JS. Photodynamic therapy—a review. Drugs Conclusions Aging 1999;15:49–68. 25 Cross CE, Van der Vielt A, O’Neill C, et al. Oxidants, anti- There is overwhelming evidence that oxidative oxidants, and respiratory tract lining ﬂuids. Environ Health stress occurs in cells as a consequence of Perspect 1994;102:185–91. normal physiological processes and environ- 26 Kelly FJ, Mudway I, Krishna MT, et al. The free radical basis of air pollution focus on ozone. Respir Med mental interactions, and that the complex web 1995;89:647–56.

www.jclinpath.com Antioxidants in health and disease 185

27 Kelly FJ, Tetley TD. Nitrogen dioxide depletes uric acid and 60 Rice-Evans CA, Miller NJ, Paganga G. Structure– ascorbic acid but not glutathione from lung lining fluid. antioxidant activity relationships of flavonoids and phenolic Biochem J 1997;325:95–9. acids. Free Radic Biol Med 1996;20:933–56. 28 Pourcelot S, Faure H, Firoozi F, et al. Urinary 8-oxo-7,8- 61 Hertog MGL, Hollman PCH, Katan MB. Content of J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from dihydro-2'-deoxyguanosine and 5- (hydroxymethyl) uracil potentially anticarcinogenic flavonoids of 28 vegetables and in smokers. Free Radic Res 1999;30:173–80. 9 fruits commonly consumed in the Netherlands. J Agric 29 Wilhelm J, Frydrychova M, Vizek M. Hydrogen peroxide in Food Chem 1992;40:2379–83. the breath of rats: the eVects of hypoxia and paraquat. 62 Hertog MGL, Hollman PCH, Putte B. Content of Physiol Res 1999;48:445–9. potentially anticarcinogenic flavonoids of tea infusions, 30 Kapiotis S, Sengoelge G, Hermann M, et al. Paracetamol wines, and fruit juices. J Agric Food Chem 1993;41:1242–6. catalyzes myeloperoxidase-initiated lipid oxidation in LDL. 63 Hertog MGL, Feskens EJM, Hollman PCH, et al. Dietary Arterioscler Thromb Vasc Biol 1997;17:2855–60. antioxidant flavonoids and risk of coronary heart disease: 31 Weijl NI, Cleton FJ, Osanto S. Free radicals and the Zutphen elderly study. Lancet 1993;342:1007–11. antioxidants in chemotherapy-induced toxicity. Cancer 64 Hertog MGL, Kromhout D, Aravanis C, et al. Flavonoid Treat Rev 1997;23:209–40. intake and long-term risk of coronary-heart-disease and 32 Gewirtz DA. A critical evaluation of the mechanisms of cancer in the 7 countries study. Arch Intern Med 1995;155: action proposed for the antitumor eVects of the anthracy- 381–6. cline antibiotics adriamycin and daunorubicin. Biochem 65 Rimm EB, Katan MB, Ascerio A, et al. Relation between Pharmacol 1999;57:727–41. intake of flavonoids and risk of coronary heart disease in 33 Kirkman HN, Galiano S, Gaetani GF. The function of male health professionals. Ann Intern Med 1996;125:384–9. catalase-bound NADPH. J Biol Chem 1987;262:660–5. 66 Hollman PCH, Vries JHM, Leeuwen SD, et al. Absorption 34 Takahashi K, Cohen HJ. Selenium-dependent glutathione of dietary quercetin glycosides and quecetin in healthy peroxidase protein and activity: immunological investiga- ileostomy volunteers. Am J Clin Nutr 1995;62:1276–82. tions on cellular and plasma enzymes. Blood 1986;68:640– 67 McAnlis GT, McEneny J, Pearce J, et al. Absorption and 6. antioxidant eVects of quercetin from onions in man. Eur J 35 Nakane T, Asayama K, Kodera K, et al.EVect of selenium Clin Nutr 1999;53:92–6. deficiency on cellular and extracellular glutathione 68 Serafini M, Ghiselli A, Ferro-Luzzi A. In vivo antioxidant peroxidases: immunochemical detection and mRNA analy- eVect of green and black tea in man. Eur J Clin Nutr 1996; sis in rat kidney and serum. Free Radic Biol Med 50:28–32. 1998;25:504–11. 69 Stein JH, Keevil JG, Wiebe DA, et al. Purple grape juice 36 Brigelius-Flohe R. Tissue-specific functions of individual improves endothelial function and reduces the susceptibil- glutathione peroxidases. Free Radic Biol Med 1999;27:951– ity of LDL cholesterol to oxidation in patients with coron- 65. ary artery disease. Circulation 1999;100:1050–5. 37 Roxborough HE, Mercer C, McMaster D, et al. Plasma glu- 70 Day AP, Kemp HJ, Bolton C, et al.EVect of concentrated tathione peroxidase activity is reduced in haemodialysis red grape juice consumption on serum antioxidant capacity patients. Nephron 1999;81:278–83. and low-density lipoprotein oxidation. Ann Nutr Metab 38 Holben DH, Smith AM. The diverse role of selenium within 1997;41:353–7. selenoproteins: a review. J Am Diet Assoc 1999;99:836–43. 71 Shi HL, Noguchi N, Niki E. Comparative study on dynam- 39 Gibson DG, Hawrylko J, McCay PB. GSH-dependent inhi- ics of antioxidative action of alpha-tocopheryl hydroqui- bition of lipid peroxidation: properties of a potent cytosolic none, ubiquinol, and alpha-tocopherol, against lipid system which protects cell membranes. Lipids 1985;20: peroxidation. Free Radic Biol Med 1999;27:334–46. 704–10. 72 Lass A, Sohal RS. Electron transport-linked ubiquinone- 40 Liou W, Chang L-Y, Geuze HJ, et al. Distribution of CuZn dependent recycling of alpha-tocopherol inhibits autooxi- superoxide dismutase in rat liver. Free Rad Biol Med 1993; dation of mitochondrial membranes. Arch Biochem Biophys 14:201–7. 1998;352:229–36. 41 Weisiger RA, Fridovich I. Mitochondrial superoxide 73 Thomas SR, Neuzil J, Stocker R. Cosupplementation with dismutase: site of synthesis and intramitochondrial loca- coenzyme Q prevents the prooxidant eVect of alpha- tion. J Biol Chem 1973;248:4793–6. tocopherol and increases the resistance of LDL to 42 Marklund S. Human copper-containing superoxide dis- transition metal-dependent oxidation initiation. Arterioscler mutase of high molecular weight. Proc Natl Acad Sci U S A Thromb Vasc Biol 1996;16:687–96. 1982;79:7634–8. 74 Levine M, Rumsey SC, Daruwala R, et al. Criteria and rec- 43 Karlsson K, Sandstrom J, Edlund A, et al. Pharmacokinetics ommendations for vitamin C intake. JAMA 1999;281: of extracellular superoxide dismutase in the vascular 1415–23. system. Free Radic Biol Med 1993;14:185–90. 75 Levine M. New concepts in the biology and biochemistry of 44 McIntyre M, Bohr DF, Dominiczak AF. Endothelial ascorbic acid. N Engl J Med 1986;314:892–902. function hypertension—the role of superoxide anion. 76 Jialal I, Vega GL, Grundy SM. Physiologic levels of Hypertension 1999;34:539–45. ascorbate inhibit the oxidative modification of low density http://jcp.bmj.com/ 45 De Zwart LL, Meerman JN, Commandeur JM, et al. lipoprotein. Atherosclerosis 1990;82:185–91. Biomarkers of free radical damage applications in experi- 77 Pietri S, Seguin JR, Darbigny P, et al. Ascorbyl free- mental animals and in humans. Free Radic Biol Med 1999; radical—a noninvasive marker of oxidative stress in human 26:202–26. open heart surgery. Free Radic Biol Med 1994;16:523–8. 46 Halliwell B. How to characterize an antioxidant: an update. 78 May JM, Cobb CE, Mendiratta S, et al. Reduction of the Biochem Soc Symp 1995;61:73–101. ascorbyl free radical to ascorbate by thioredoxin reductase. 47 Esterbauer H, Dieber-Rotheneder M, Striegl G, et al. Role J Biol Chem 1998;273:23039–45. of vitamin E in preventing the oxidation of low-density 79 May JM, Qu ZC, Whitesell RR, et al. Ascorbate recycling in lipoprotein. Am J Clin Nutr 1991;53:314s–21s. human erythrocytes: role of GSH in reducing dehy- 48 Horwitt MH. Data supporting supplementation of humans droascorbate. Free Radic Biol Med 1996;20:543–51.

with vitamin E. J Nutr 1991;121:424–9. 80 Koshiishi I, Mamura Y, Liu J, et al. Evaluation of an acidic on October 1, 2021 by guest. Protected copyright. 49 Urano S, Inomori Y, Sugawara T, et al. Vitamin-E— deproteinization for the measurement of ascorbate and inhibition of retinol-induced hemolysis and membrane- dehydroascorbate in plasma samples. Clin Chem 1998;44: stabilizing behavior. J Biol Chem 1992;267:18365–70. 863–8. 50 Swann IL, Kendra JR. Anaemia, vitamin E deficiency and 81 Grootveld M, Halliwell B. Measurement of allantoin and failure to thrive in an infant. Clin Lab Haematol uric acid in human body fluids. A potential index of 1998;20:61–3. free-radical reactions in vivo? Biochem J 1987;243:803–8. 51 Sokol RJ. Vitamin E deficiency and neurologic disease. 82 Cross CE, Reznick AZ, Packer L, et al. Oxidative damage to Annu Rev Nutr 1988;8:351–73. human plasma proteins by ozone. Free Radic Res Commun 52 Kayden HJ, Traber MG. Absorption, lipoprotein transport, 1992;15:347–52. and regulation of plasma concentrations of vitamin E in 83 Ames BN, Cathcart R, Schwiers E, et al. Uric acid provides humans. J Lipid Res 1993;34:343–58. an antioxidant defense in humans against oxidant- and 53 Burton GW, Ingold KU. Vitamin E: applications of the radical-caused aging and cancer. Proc Natl Acad Sci U S A principles of physical organic chemistry to the exploration 1981;78:6858–62. of its structure and function. Acc Chem Res 1986;19:194– 84 Frei B, Stocker R, Ames BN. Antioxidant defences and lipid 201. peroxidation in human blood plasma. Proc Natl Acad Sci U 54 May JM, Qu ZC, Mendiratta S. Protection and recycling of SA1988;85:9748–52 alpha-tocopherol in human erythrocytes by intracellular 85 Gopinathan V, Miller NJ, Milner AD, et al. Bilirubin and ascorbic acid. Arch Biochem Biophys 1998;349:281–9. ascorbate antioxidant activity in neonatal plasma. FEBS 55 Kagan VE, Tyurina YY. Recycling and redox cycling of phe- Lett 1994;349:97–200. nolic antioxidants. Towards prolongation of the healthy life 86 Halliwell B. Albumin—an important extracellular antioxi- span. AnnNYAcadSci1998;854:425–34. dant? Biochem Pharmacol 1988;37:569–71. 56 Cooper DA, Eldridge AL, Peters JC. Dietary carotenoids 87 Stocker R, Frei B. Endogenous antioxidant defences in and certain cancers, heart disease, and age-related macular human blood plasma. In: Rice-Evans C, ed. Oxidative stress: degeneration: a review of recent research. Nutr Rev oxidants and antioxidants. London: Academic Press, 1991: 1999;57:201–14. 213–43. 57 Fukuzawa K, Inokami Y, Tokumura A, et al. Rate constants 88 Hu ML, Louie S, Cross CE, et al. Antioxidant protection for quenching singlet oxygen and activities for inhibiting against hypochlorous acid in human plasma. J Lab Clin lipid peroxidation of carotenoids and alpha-tocopherol in Med 1993;121:257–62. liposomes. Lipids 1998;33:751–6. 89 Riedl A, Shamsi Z, Anderton M. DiVering features of 58 Chaudiere J, Ferrari-Iliou R. Intracellular antioxidants: from proteins in membranes may result in antioxidant or prooxi- chemical to biochemical mechanisms. Food Chem Toxicol dant action: opposite eVects on lipid peroxidation of 1999;37:949–62. alcohol dehydrogenase and albumin in liposomal systems. 59 Keys SA, Zimmerman WF. Antioxidant activity of retinol, Redox Rep 1996;2:35–40. glutathione, and taurine in bovine photoreceptor cell mem- 90 Sastre J, Pallardo FV, Vina J. Glutathione, oxidative stress branes. Exp Eye Res 1999;68:693–702. and aging. Age 1996;19:129–39.

www.jclinpath.com 186 Young, Woodside

91 Arrigo AP. Gene expression and the thiol redox state. Free 119 Verlangieri AJ, Bush MJ. EVects of d-alpha-tocopherol Radic Biol Med 1999;27:936–44. supplementation on experimentally induced primate 92 Suh J, Zhu BZ, Frei B. Anti- and pro-oxidant eVects of atherosclerosis. J Am Coll Nutr 1992;11:131–8. ascorbate on iron-mediated oxidative damage to bovine 120 Stocker R, Bowry VW, Frei B. Ubiquinol-10 protects J Clin Pathol: first published as 10.1136/jcp.54.3.176 on 1 March 2001. Downloaded from serum albumin. Free Radic Biol Med 1999;27:305s1. human low-density lipoprotein more eYciently against 93 Neuzil J, Thomas SR, Stocker R. Requirement for, promo- lipid peroxidation than does alpha-tocopherol. Proc Natl tion, or inhibition by alpha-tocopherol of radical-induced Acad Sci U S A 1991;88:1646–50. initiation of plasma lipoprotein lipid peroxidation. Free 121 Esterbauer H, Puhl H, Dieber-Rotheneder M, et al.EVect Radic Biol Med 1997;22:57–71. of antioxidants on oxidative modification of LDL. Ann Med 94 Atanasiu RL, Stea D, Mateescu MA, et al. Direct evidence 1991;23:573–81. of caeruloplasmin antioxidant properties. Mol Cell Biochem 122 Reaven PD, Khouw A, Beltz WF, et al.EVect of dietary 1998;189:127–35. antioxidant combinations in humans. Protection of LDL 95 Liu T, Stern A, Roberts LJ, et al. The isoprostanes: novel by vitamin E but not by beta-carotene. Arterioscler Thromb prostaglandin-like products of the free radical-catalyzed 1993;13:590–600. peroxidation of arachidonic acid. J Biomed Sci 1999;6:226– 123 Gey KG, Puska P, Jordan P, et al. Inverse correlation 35. between vitamin E and mortality from ischaemic heart dis- 96 Rosenfeld ME. Inflammation, lipids, and free radicals: ease in cross-cultural epidemiology. Am J Clin Nutr lessons learned from the atherogenic process. Semin Reprod 1991;53:S326–34. Endocrinol 1998;16:249–61. 124 Riemersma RA, Wood DA, MacIntyre CCA, et al. Risk of 97 Hecht SS. Tobacco smoke carcinogens and lung cancer. J angina pectoris and plasma concentrations of vitamins A, C Natl Cancer Inst 1999;91:1194–210. and E and carotene. Lancet 1991;337:1–5. 98 Ashok BT, Ali R. The aging paradox: free radical theory of 125 Salonen JT, Salonen R, Seppanen K, et al. Relationship of aging. Exp Gerontol 1999;34:293–303. serum selenium and antioxidants to plasma lipoproteins, 99 Witztum JL, Steinberg D. Role of oxidized LDL in platelet aggregability and prevalent ischaemic heart disease atherogenesis. J Clin Invest 1991;88:1785–92. in eastern Finnish men. Atherosclerosis 1988;70:155–60. 100 Witztum JL, Horkko S. The role of oxidized LDL in 126 Rimm EB, Stampfer MJ, Ascherio A, et al. Vitamin E con- atherogenesis: immunological response and anti- sumption and the risk of coronary heart disease in men. N phospholipid antibodies. AnnNYAcadSci1997;811:88– Engl J Med 1993;328:1450–6. 99. 127 Stampfer MJ, Hennekens CH, Manson JE, et al. Vitamin E 101 Heinecke JW. Mechanisms of oxidative damage of low consumption and the risk of coronary heart disease in density lipoprotein in human atherosclerosis. Curr Opin women. N Engl J Med 1993;328:1444–9. Lipidol 1997;8:268–74. 128 Enstrom JE, Kanim LE, Klein MA. Vitamin C intake and 102 Morel DW, DiCorleto PE, Chisholm GM. Endothelial and mortality among a sample of the United States population. smooth muscle cells alter low density lipoprotein in vitro by Epidemiology 1992;3:194–202. free radical oxidation. Arteriosclerosis 1984;4:357–64. 129 Nyyssonen K, Parviainen MT, Salonen R, et al. Vitamin C 103 Henriksen T, Mahoney EM, Steinberg D. Enhanced mac- deficiency and risk of myocardial infarction: prospective rophage degradation of low density lipoprotein previously population study of men from eastern Finland. BMJ 1997; incubated with cultured endothelial cells: recognition by 314:634–8. the receptor for acetylated low density lipoproteins. Proc 130 Gaziano JM, Hennekens CH. The role of beta-carotene in Natl Acad Sci U S A 1981;78:6499–503. the prevention of cardiovascular disease. AnnNYAcadSci 104 Folcik VA, Nivar-Aristy RA, Krajewski LP, et al. Lipoxyge- 1993;691:148–55. nase contributes to the oxidation of lipids in human 131 Alpha-Tocopherol, Beta Carotene Cancer Prevention atherosclerotic plaques. J Clin Invest 1995; 96:504–10. Study Group. The eVect of vitamin E and beta carotene on 105 Itabe H, Yamamoto H, Imanaka T, et al. Sensitive the incidence of lung cancer and other cancers in male detection of oxidatively modified low-density lipoprotein smokers. N Engl J Med 1994;330:1029–35. using a monoclonal antibody. J Lipid Res 1996;37:45–53. 132 Rapola JM, Virtamo J, Ripatti S, et al. Randomised trial of 106 Yla-Herttuala S, Palinski W, Rosenfeld ME, et al. Evidence á-tocopherol and â-carotene supplements on incidence of for the presence of oxidatively modified low density major coronary events in men with previous myocardial lipoprotein in atherosclerotic lesions of rabbit and man. J infarction. Lancet 1997;349:1715–20. Clin Invest 1989;84:1086–95. 133 Blot WJ, Li JY, Taylor PR, et al. Nutrition intervention 107 Regnstrom J, Nilsson J, Tornvall P, et al. Susceptibility to trials in Linxian, China: supplementation with specific low-density lipoprotein oxidation and coronary atheroscle- vitamin/mineral combinations, cancer incidence, and rosis in man. Lancet 1992;339:1183–6. disease-specific mortality in the general population. JNatl 108 Ball RY, Carpenter KL, Mitchinson MJ. What is the Cancer Inst 1993;85:1483–92. significance of ceroid in human atherosclerosis? Arch Pathol 134 Omenn GS, Goodman GE, Thornquist MD, et al.EVects Lab Med 1987;111:1134–40. of a combination of beta carotene and vitamin A on lung 109 Navab M, Imes S, Hama S, et al. Monocyte transmigration cancer and cardiovascular disease. N Engl J Med 1996;334:

induced by modiﬁcations of LDL in cocultures of human 1150–5. http://jcp.bmj.com/ aortic wall cells is due to induction of monocyte chemotac- 135 Hennekens CH, Buring JE, Manson JE, et al. Lack of eVect tic protein I synthesis and is abolished by HDL. J Clin of long-term supplementation with beta carotene on the Invest 1991;88:2039–46. incidence of malignant neoplasms and cardiovascular 110 Hessler JR, Robertson AL, Chisolm GM. LDL cytotoxic- disease. N Engl J Med 1996;334:1145–9. ity and its inhibition by HDL in human vascular smooth 136 Greenberg ER, Baron JA, Karagas MR, et al. Mortality muscle and endothelial cell culture. Atherosclerosis 1979;32: associated with low plasma concentration of beta-carotene 213. and the eVect of oral supplementation. JAMA 1996;275: 111 Kugiyama K, Kerns SA, Morrisett JD, et al. Impairment of 699–703. endothelial-dependent arterial relaxation by lysolecithin in 137 Hodis HN, Mack WJ, LaBrec L, et al. Serial coronary modiﬁed low-density lipoproteins. Nature 1990;344:160–2. angiographic evidence that antioxidant vitamin intake 112 Salonen JT, Yla-Herttuala S, Yamamoto R, . Autoanti- reduced progression of coronary artery atherosclerosis.

et al on October 1, 2021 by guest. Protected copyright. body against oxidised LDL and progression of carotid JAMA 1995;273:1849–54. atherosclerosis. Lancet 1992;339:883–7. 138 Singh RB, Niaz MA, Rastogi SS, et al. Usefulness of anti- 113 Bergmark C, Wu R, de Faire U, et al. Patients with oxidant vitamins in suspected acute myocardial infarction. early-onset peripheral vascular disease have increased levels Am J Cardiol 1996;77:232–6. of autoantibodies against oxidized LDL. Arterioscler 139 Stephens NG, Parsons A, Schoﬁeld PM, et al. Randomised Thromb Vasc Biol 1995;15:441–5. controlled trial of vitamin E in patients with coronary 114 Libby P, Hansson GK. Involvement of the immune system disease: Cambridge heart antioxidant study (CHAOS). in human atherogenesis: current knowledge and unan- Lancet 1996;347:781–6. swered questions. Lab Invest 1991;64:5–15. 140 Valagussa F, Franzosi MG, Geraci E, et al. Dietary supple- 115 Kita T, Kume N, Ishii K, et al. Oxidized LDL and expres- mentation with n-3 polyunsaturated fatty acids and vitamin sion of monocyte adhesion molecules. Diabetes Res Clin E after myocardial infarction: results of the GISSI- Pract 1999;45:123–6. Prevenzione trial. Lancet 1999;354:447–55. 116 Huang YH, Schafer-Elinder L, Wu R, et al. Lysophosphati- 141 Yusuf S, Phil D, Dagenais G, et al. Vitamin E supplemen- dylcholine (LPC) induces proinﬂammatory cytokines by a tation and cardiovascular events in high- risk patients. platelet-activating factor (PAF) receptor-dependent mech- N Engl J Med 2000;342:154–60. anism. Clin Exp Immunol 1999;116:326–31. 142 Slattery ML, Jacobs DR, Dyer A, et al. Dietary antioxi- 117 Wolin MS. Interactions of oxidants with vascular signaling dants and plasma lipids: the CARDIA study. J Am Coll systems. Arterioscler Thromb Vasc Biol 2000;20:1430–42. Nutr 1995;14:635–42. 118 Bjorkhem I, Henrikssonfreyschuss A, Breuer O, et al. The 143 Valkonen M, Kuusi T. Passive smoking induces athero- antioxidant butylated hydroxytoluene protects against genic changes in low-density lipoprotein. Circulation 1998; atherosclerosis. Arterioscler Thromb 1991;11:15–22. 97:2012–16.