Annals of the Rheumatic Diseases 1995; 54: 505-5 10

Oxygen radicals, nitric oxide and humaninflammatory joint disease

Barry Halliwell

Interest in the role of free radicals inrheumatoid arthritis (RA) stems from theseminal work of McCord,' who noted thedecreased viscosity of synovial fluid in RApatients and showed that a similar decreasecould be produced by exposing synovial fluid,or solutions of hyaluronic acid, to a systemgenerating superoxide radical, 02-. McCord'sobservations led to interest in the use of intra-articular injections of the antioxidant enzymesuperoxide dismutase as a treatment in RA.However, the clinical data presented did notconvince many rheumatologists2 3 and over-enthusiastic interpretations of the data mayhave led to unwarranted scepticism about thereal role of free radicals in RA. Let us reviewour current knowledge.

Basic definitionsElectrons in atoms and molecules occupyregions of space termed orbitals, each of whichholds a maximum of two electrons. A freeradical is any species capable of independentexistence that contains one or more unpairedelectrons-that is, electrons alone in an orbital.Table 1 gives examples.

Radicals react with other molecules in anumber of ways.4 If two radicals meet, theycan combine their unpaired electrons andjoin to form a covalent bond (a shared pairof electrons). An important example is thefast reaction of 02- with nitric oxide (also afree radical, NO') to form the non-radicalperoxynitrite:5

O2-- +NO- ONOO- (1)A free radical might donate its unpaired

electron to another molecule. Thus °2- reducesferric (Fe3") cytochrome c to ferrous (Fe2+ )cytochrome c, a reaction often used to assay02- production by activated phagocytes:6

cyt c (Fe3+) + 02-- ' cyt c (Fe2+) + 02 (2)

A free radical might take an electron fromanother molecule, thus oxidising it. Forexample, 02 oxidises ascorbic acid, a processbelieved to occur in RA:7

ascorbate + 02 + H+- ascorbate radical+ H202 (3)

Left to itself, 02- undergoes the dismutationreaction:

02*- + 02-- +2H+ - H202 +02 (4)

Hydrogen peroxide, H2 02, is not a freeradical (no unpaired electrons are present).The global term reactive oxygen species is oftenused to include the oxygen radicals (02- and

OH") and important non-radical derivatives ofoxygen such as H202 and hypochlorous acid(HOCI).

Reactive oxygen species in vivoThe chemical reactivity of oxygen radicalsvaries (table 1). The most reactive is hydroxylradical (OH"), which reacts very fast withalmost all molecules in vivo. When OH" isformed, it damages whatever it is generatednext to; it cannot migrate within the cell.4

NITRIC OXIDE

Whereas OH is probably always harmful,other (less reactive) free radicals may be usefulin vivo. For example, NO is synthesised fromL-arginine by many cell types, includingchondrocytes.8 However, although humanphagocytes can make NO,9 it is not yet clearhow often they do so in vivo; much of the workwith NO has investigated rats and mice, bothof which species have phagocytes that makeNO much more readily.

SUPEROXIDE

Superoxide is produced by phagocytes as akilling mechanism.6 Lesser amounts of extra-cellular O2* may be generated, perhaps as anintercellular signal molecule, by several othercell types, including vascular endothelial cells,osteoclasts, chondrocytes, lymphocytes, andfibroblasts.""'3 For example, treatment ofhuman fibroblasts with RA synovial fluidcauses 02- secretion. 3

In addition to this 'deliberate' 02 -genera-tion, some 02-- iS produced within cells bymitochondria and endoplasmic reticulum,apparently by the unavoidable 'leakage' ofelectrons onto oxygen from their correct pathsin electron transfer chains and by chemical'autoxidation' reactions.4

HYPOCHLOROUS ACID

Another killing mechanism used by neutro-phils (but not macrophages) is the enzymemyeloperoxidase,14 which uses H202 to oxidisechloride ions into hypochlorous acid (HOCI), apowerful oxidising and chlorinating agent:

H202 +Cl-- HOCI + OH- (5)

Superoxide-nitric oxide interactionsWhat happens if both 02- and NO areproduced at the same site, for example by

Pharmacology Group,King's College,University ofLondon,London SW3 6LX,United KingdomB Halliwell

505

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from

Halliwell

Table 1 Examples offree radicals

Name Formula Comments

Hydrogen atom *H. The simplest free radical known.Trichloromethyl CC13 A carbon centred radical (the unpaired electron resides

on carbon). CC3l is formed during metabolism ofcarbon tetrachloride (CC14) in the liver andcontributes to the toxic effects of this solvent.

Superoxide 02- An oxygen centred radical. Reacts quickly with a fewmolecules (such as nitric oxide) but generally not veryreactive.

Hydroxyl OH An oxygen centred radical. The most highly reactiveoxygen radical known. When generated in vivo, reactsat its site of formation.

Thiyl RS' General name for a group of radicals with an unpairedelectron residing on sulphur. Reactivity varies; oftenreact with oxygen to give damaging oxysulphurradicals.

Peroxyl, alkoxyl RO2., RO Oxygen centred radicals formed during the breakdown oforganic peroxides.

Oxides of nitrogen NO, NO2. Both are free radicals. NO is formed in vivo from theamino acid L-arginine. N02 is made when NO reactswith oxygen, and is found in polluted air and smokefrom burning organic materials, such as cigarettesmoke.

*A superscript dot is used to denote free radical species.

activated macrophages, or when neutrophilsadhering to endothelium (a source of NO0)generate 02"-? One possibility is that 02-and NO antagonise each other's bio-logical actions. Inappropriate antagonism ofNO, by excess 02-, has been suggestedto contribute to impaired endotheliummediated vasodilatation, for example inhypertension. 15The interaction of 02- and NO can also

be dangerous.5 The product, peroxynitrite(equation (2)) is not only directly toxic, forexample by oxidising methionine and protein-SH groups, but also it breaks down togenerate multiple toxic products (figure), in-cluding nitrogen dioxide gas (NO2'), OH' andnitronium ion (NO24).5 16 Some of thesespecies will nitrate aromatic amino acids, sothat formation of nitroaromatics (especiallynitrotyrosine) is thought to be a 'marker' ofperoxynitrite generation.5 17

0; + NO

ONOO- Peroxynitrite

H+

ONOOH Peroxynitrous acid

NO- NO+

Nitrate ion Nitroniumion

NO. Nitrogen dioxide2 radical

OH- 'OH Hydroxyl radicalFormation and decomposition ofperoxynitrite.

Transition metals and hydrogen peroxideMany transition metals have variable oxidationnumbers: for example iron as Fe2+ or Fe3+, andcopper as Cu+ or Cu24. Changing betweenthese oxidation states transfers single electrons,for example:

Fe3+ + e-= Fe2+

Also, for titanium salts:

Ti4+ + e-= Ti3+

(6)

(7)Thus transition metal ions are good

promoters of free radical reactions. Forexample, copper, iron and titanium ions reactwith H202 to form OH" radicals:4

Cu4+ H202- Cu2+ +OH' + OH- (8)

Ti3+ + H202 -_ Ti4+ + OH + OH- (9)

H202 is produced in vivo by dismutation ofO2- and by several oxidase enzymes, includingxanthine oxidase.4 Like 02', H202 can beuseful in vivo; for example, it is a substrate fora thyroid peroxidase enzyme that helps makethyroid hormones."8 It can also regulate geneexpression, for example by activating thecytoplasmic gene transcription factor NF-KB.'9H202 is very diffusible within and betweencells,4 but if it comes into contact withtransition metal ions, OH will be generated atthat site and cause immediate damage.

Antioxidant defencesENZYMESLiving organisms have evolved multipleantioxidant defence systems. Superoxidedismutase (SOD) enzymes remove 02'- byaccelerating its dismutation by about fourorders of magnitude. Human cells have anSOD enzyme containing active site manganese(MnSOD) in mitochondria, whereas cytosolcontains a copper and zinc containing SOD(CuZnSOD).4 H202 can be destroyed bycatalases, but the most important H2 02removing enzymes in human cells are gluta-thione peroxidases, which remove H202 by usingit to oxidise reduced glutathione (GSH) tooxidised glutathione (GSSG):

2GSH + H202- GSSG + 2H20 (10)

METAL ION SEQUESTRATIONAnother important antioxidant defence is thatiron and copper ions are kept safely proteinbound whenever possible, so that OHformation is largely prevented. This isparticularly important in extracellular fluids,including synovial fluid, because their levels ofantioxidant defence enzymes are low.7 20 21The value of this sequestration of metal ions

is illustrated by an inspection of the severepathology suffered by patients with metaloverload diseases. For example, in patientswith iron overload secondary to idiopathichaemochromatosis, transferrin is iron satu-rated and iron ions 'catalytic' for free radicalreactions circulate in the blood.22 Among manyother problems, these patients can suffer joint

506

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from

Oxygen radicals, nitric oxide and human inflammatory joint disease

inflammation,23 illustrating the well knownrelationship between chronic inflammationand disordered iron metabolism.24

CI-TOCOPHEROLThis fat soluble vitamin functions as a chainbreaking antioxidant in membranes andlipoproteins.25 When peroxyl radicals aregenerated during lipid peroxidation (table 1),they abstract hydrogen preferentially from thephenolic -OH group of tocopherol:

-CO2- +TOH---CO2H+TO' (11)This stops peroxyl radicals from attacking anadjacent fatty acid side chain or protein andterminates the chain reaction, hence the namechain breaking antioxidant. The oa-tocopherolradical, tocopherol-O, is poorly reactive andis widely believed to migrate to the surfaceof membranes or lipoproteins for conversionback to a-tocopherol by reaction with ascorbicacid.

REPAIR SYSTEMSThe antioxidant defences of the human bodyare not 100% efficient, so that some freeradical damage occurs in the human bodyand repair systems are needed. Thus cellshave enzymes that can repair oxidisedDNA, degrade free radical damaged proteins,and remove lipid hydroperoxides frommembranes.4

Oxidative stressOxidative stress is said to result when reactiveoxygen species are generated in excess in thehuman body.26 This can occur if antioxidantconcentrations are too low (severe mal-nutrition, for example, can deplete levels ofa-tocopherol and vitamin C) or if free radicalformation is increased-for example by theaction of certain toxins.4

Cells can tolerate mild oxidative stress, andoften respond to it by increased synthesis ofantioxidant defence enzymes and other protec-tive proteins. However, severe oxidative stresscan cause cell injury or even death; oxidativedamage to DNA, proteins and lipids candamage or destroy cells. Cell death induced byoxidative stress can occur by necrosis or apop-tosis.27 Oxidative stress causes increased intra-cellular free Ca2" 28 and may release intracellulariron to catalyse OH' generation.29

Reactive oxygen species in RAGENERAL PRINCIPLESWhat is the role played by reactive oxygenspecies in human disease? Some diseases maybe caused by oxidative stress, for example thesequelae of overexposure to ionising radiation4and the neurodegeneration produced bychronic a-tocopherol deficiency.30For most human diseases, however, the

oxidative stress is secondary to the primarydisease process.29 For example, tissue injuryrecruits and activates neutrophils, which

produce 02'-, H202, HOCI, and possibly NO';in excess, these cause damage. Tissue injuryreleases iron and copper ions and haemproteins (haemoglobin and myoglobin), bothcatalytic for free radical reactions, from theirnormal intracellular storage sites;4 29 it alsodisrupts electron transport chains in mito-chondria and endoplasmic reticulum, so thatmore electrons leak to oxygen to form 02'- 31

For such secondary oxidative stress, the keyquestion is 'does it contribute significantly todisease pathology?'

THE CASE OF RA

Following the work of McCord,' I showed thathyaluronic acid depolymerisation by 02*-generating systems in vitro is caused by irondependent formation of OH' from 02- andH202. 32 The hydroxyl radical causes randomfragmentation of hyaluronate, eventuallyproducing oligosaccharides.33 How could OH'arise in the RA joint? 'Catalytic' copper ionswere not detected in fresh synovial fluid,34 but'catalytic' iron can be measured by thebleomycin assay in about 40% of synovialfluids aspirated from inflamed RA kneejoints,35 and this iron has been directlydemonstrated to stimulate lipid peroxidation.36In addition, aspiration of synovial fluid fromsome RA patients into a solution of phenyl-alanine produces a pattern of hydroxylationproducts characteristic of OH' attack upon thearomatic ring,37 suggesting that constituents ofRA synovial fluid can lead to OH' formation.The catalytic iron could arise by release from

dead cells, by H202 mediated degradation ofhaemoglobin38 (released by traumatic micro-bleeding in the joint), or by the action of 02'-on synovial fluid ferritin.39 Release of iron uponexposure of synovial fluid to 02'-, especially atacidic pH, has been demonstrated.40 The chem-ical pattern of damage to hyaluronate in RAsynovial fluids (as demonstrated using nuclearmagnetic resonance33) is consistent with OH'attack, though hyaluronate may additionally besecreted as abnormally short chains in RA.4'

Reaction of 02' with NO" is anotherpotential source of OH', as the synthesis notonly of 02'- but also of NO' 42 appears to beincreased in RA patients. Demonstration of thepresence of nitrotyrosines in patients withactive RA43 is consistent with formation ofperoxynitrite (equation (1)) in vivo.5A third source of OH' is reaction of 02'

with HOC1,44 both produced by activatedphagocytes:

02- + HOCI - 2 + OH- + Cl- (12)

SOURCES OF OXYGEN DERIVED SPECIES IN RA

The most discussed source is activatedphagocytes. Activated neutrophils liberate02'-, H202, elastase, HOCI, and eicosanoids,and synovial fluid IgG aggregates may activateneutrophils.45 Hypochlorous acid and 02-both react with ascorbate, which may help toexplain the low levels of ascorbate in RA bodyfluids.7 46 Hypochlorous acid inactivates

507

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from

Halliwell

a5-antiproteinase, an important inhibitor ofserpins (such as elastase), and the amount ofactive al-antiproteinase is decreased in RA;47HOCI also fragments collagen.48 The pannuscontains many macrophage-like cells, pre-sumably secreting 02*- H202 49 and, possibly,NO' (it is not yet clear if neutrophils or macro-phages in the RA joint make NO").

It has also been proposed that the inflamedrheumatoid joint, upon movement and rest,undergoes a hypoxia-reperfusion cycle, whichmay result in free radical generation by severalmechanisms.50 It is interesting to note that oneof these mechanisms is xanthine oxidase,51 52relating back nicely to the original work ofMcCord.'

DRUG-DERIVED RADICALSA few anti-inflammatory drugs may scavengereactive oxygen species in vivo, but this abilityis not widespread.53 Indeed, the reverse can betrue: several drugs used in the treatment ofRAmight themselves be converted into freeradicals in vivo. Thus they could suppress thesigns of RA whilst aggravating oxidativedamage. For example, radicals derived frompenicillamine, phenylbutazone, some fenamicacids, and the aminosalicylate component ofsulphasalazine, might inactivate a1-anti-proteinase, deplete ascorbic acid and acceler-ate lipid peroxidation.5355

Consequences of oxidative stress in RAThere is no doubt that oxidative stress occursin RA patients (table 2).51-71 Lunec et a167 haveargued that oxidative damage to IgG generatesprotein aggregates that can activate neutrophilsand set up a 'vicious cycle' of free radicalproduction.Does oxidative stress contribute to cartilage

and bone destruction in RA? The answer is

Table 2 Evidence consistent with oxidative stress in rheumatoid disease

Observation Reference Comment

Increased lipid peroxidation 56 Decreased a-tocopherol (per unit lipid) inproducts in serum and synovial fluid57 is consistent with increasedsynovial fluid lipid peroxidation, as are reports of 'foam cells'

containing oxidised low density lipoprotein inrheumatoid synovium"8 and increased levels of4-hydroxy-2-nonenal, a cytotoxic productgenerated by the decomposition of lipidperoxides, in RA.59

Depletion of ascorbate in See text Presumably results from oxidation of ascorbateserum and synovial fluid during its antioxidant action. Activated

neutrophils also take up oxidised ascorbaterapidly.60

Increased exhalation of 61 A putative endproduct of lipid peroxidation,4pentane although its validity as an assay is debated.62

Increased concentrations of 63 Products measured appear to be endproducts ofuric acid oxidation free radical attack upon uric acid.products

Formation of 2,3-dihydroxy- 65 2,3-DHB appears to be a product of attack ofbenzoate (2,3-DHB) from OH' upon salicylate in patients takingsalicylate in increased aspirin.'6amounts

Degradation of hyaluronic - See text.acid by free radicalmechanisms

Formation of 'fluorescent' 67,68 Fluorescence probably caused by oxidativeproteins damage to amino acid residues in proteins.

Increased steady state levels 69,70 80HdG is a major product of oxidative damage(in cellular DNA) and to DNA.4increased urinary excretionof 8-hydroxy-deoxyguano-sine (8OHdG)

Increased levels of 'protein 71 Protein carbonyls are an endproduct of oxidativecarbonyls' in synovial fluid damage to proteins.

unclear. Hydroxyl radicals degrade isolatedproteoglycans' 32 33 and HOC1 fragmentscollagen,48 but their effects on intact cartilageare probably limited. However, H202 is verydiffusible and inhibits cartilage proteoglycansynthesis,72 for example by interfering withATP synthesis.73 Indeed, intra-articular in-jection of H2 02 generating systems causessevere joint damage in animals.74 Henceinhibition of cartilage repair systems couldaggravate the effects of proteolytic and freeradical mediated cartilage degradation. HOC1can also activate latent forms of neutrophilcollagenases and gelatinase, though the extentto which this happens in vivo is uncertain.75Chondrocytes are damaged by H202,76 and ithas been suggested that low concentrationsof H2 02, 02' , or both, accelerate boneresorption by osteoclasts,77 78 whereas NO'inhibits it.79 In addition, ascorbate is essentialfor cartilage function80 and the low concen-trations found in RA synovial fluid mightimpair cartilage metabolism. The currentinterest in the role of tumour necrosis factor ct(TNFot) in RA8' may relate to observationsthat TNFot causes oxidative stress.82 83

In summary, we do not as yet know the exactcontribution made by reactive oxygen speciesto joint damage in RA. The development ofimproved assays of oxidative damage thatare applicable to humans29 should help toaddress this point and allow a rational selectionof antioxidants for possible therapeuticapplication.84

I thank the Arthritis and Rheumatism Council for researchsupport.

1 McCord J M. Free radicals and inflammation. Science 1974;185: 529-31.

2 Greenwald R A. Oxygen radicals, inflammation, andarthritis. Semin Arthritis Rheum 199 1; 20: 219-40.

3 Dowling E J, Chander C L, Claxson A W, Lillie C,Blake D R. Assessment of a human recombinantmanganese superoxide dismutase in models of inflam-mation. Free Radic Res Commun 1993; 18: 291-8.

4 Halliwell B, Gutteridge J M C. Free radicals in biology andmedicine. Oxford: Clarendon Press, 1989.

5 Beckman J S, Chen J, Ischiropoulos H, Crow J P. Oxidativechemistry of peroxynitrite. Methods Enzymol 1994; 233:229-40.

6 Cumutte J T, Babior B M. Chronic granulomatous disease.Adv Hum Genet 1987; 16: 229-45.

7 Blake D R, Hall N D, Treby D A, Halliwell B,Gutteridge J M C. Protection against superoxide andhydrogen peroxide in synovial fluid from rheumatoidpatients. Clin Sci 1981; 61: 483-6.

8 Palmer R M J, Hickery M S, Charles I G, Moncada S,Bayliss M T. Induction of nitric oxide synthase in humanchondrocytes. Biochem Biophys Res Commun 1993; 193:398-405.

9 Carreras M C, Pargament G A, Catz S D, Poderoso J J,Boveris A. Kinetics of nitric oxide and hydrogen peroxideproduction and formation of peroxynitrite during therespiratory burst of human neutrophils. FEBS Lett 1994;341: 65-8.

10 Rathakrishnan C, Tiku K, Raghavan A, Tiku M L. Releaseofoxygen radicals by articular chondrocytes. J Bone MinerRes 1992; 7: 1139-48.

11 Tiku M L, Liesch J B, Roberston F M. Production ofhydrogen peroxide by rabbit articular chondrocytes.Jlmmunol 1990; 145: 690-6.

12 Key L L Jr, Ries W L, Glasscock H, Rodriguiz R, Jaffe H.Osteoclastic superoxide generation. Int J Tissue React1992; XIV: 295-8.

13 Meier B, Radeke H H, Selle S, et al. Human fibroblastsrelease reactive oxygen species in response to treatmentwith synovial fluids from patients suffering from arthritis.Free Radic Res Commun 1990; 8: 149-60.

14 Weiss S J. Tissue destruction by neutrophils. N Engl J' Med1989; 320: 365-76.

15 Nakazono K, Watanabe N, Matsumo K, et al. Doessuperoxide underly the pathogenesis ofhypertension? ProcNatlAcad Sci USA 1991; 88: 10045-8.

508

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from

Oxygen radicals, nitric oxide and human inflammatory joint disease

16 Van der Vliet A, O'Neill C A, Halliwell B, Cross C E,Kaur H. Aromatic hydroxylation and nitration of phenyl-alanine and tyrosine by peroxynitrite. FEBS Len 1994;339: 89-92.

17 van der Vliet A, Smith D, O'Neill C A, et al. Interactionsof peroxynitrite with human plasma and its constituents:oxidative damage and antioxidant depletion. Biochem J71994; 303: 295-301.

18 Dupuy C, Virion A, Ohayon R, et al. Mechanism ofhydrogen peroxide formation catalyzed by NADPHoxidase in thyroid plasma membrane. Jf Biol Chem 1991;266: 3739-43.

19 Schreck R, Albermann K, Baeuerle P A. Nuclear factorkappa B: an oxidative stress-responsive transcriptionfactor of eukaryotic cells (a review). Free Radic ResCommun 1992; 17: 221-37.

20 Halliwell B, Gutteridge J M C. The antioxidants ofhuman extracellular fluids. Arch Biochem Biophys 1990;280: 1-8.

21 Biemond P, Swaak A J G, Koster J F. Protective factorsagainst oxygen free radicals and hydrogen peroxide inrheumatoid arthritis synovial fluid. Arthritis Rheum 1984;27: 760-5.

22 Gutteridge J M C, Rowley D A, Griffiths E, Halliwell B.Low-molecular-weight iron complexes and oxygen radicalreactions in idiopathic haemochromatosis. Clin Sci 1985;68: 463-7.

23 McLaren G D, Muir W A, Kellermeyer R W. Iron over-load disorders. CRC Crit Rev Clin Lab Sci 1983; 19:205-66.

24 Means R T Jr, Krantz S B. Progress in understanding thepathogenesis of the anemia of chronic disease. Blood1992; 80: 1639-47.

25 Burton G W, Traber M G. Vitamin E: antioxidant activity,biokinetics and bioavailability. Ann Rev Nutr 1990; 10:357-82.

26 Sies H. Oxidative stress II. Oxidants and antioxidants. NewYork and London: Academic Press, 1989.

27 Sarafian T A, Bredesen D E. Is apoptosis mediated byreactive oxygen species? Free Radic Res 1994; 21: 1-8.

28 Orrenius S, McConkey D J, Bellomo G, Nicotera P. Roleof Ca"+ in toxic cell killing. Trends Pharm Sci 1989; 10:281-5.

29 Halliwell B, Gutteridge J M C, Cross C E. Free radicals,antioxidants and human disease: where are we now?JLabClin Med 1992; 119: 598-620.

30 Muller D P R, Goss-Sampson M A. Neurochemical,neurophysiological and neuropathological studies invitamin E deficiency. Crit Rev Neurobiol 1990; 5: 239-65.

31 Ambrosio G, Zweier J L, Duilio C, et al. Evidencethat mitochondrial respiration is a source of potentiallytoxic oxygen free radicals in intact rabbit hearts sub-jected to ischemia and reflow. J Biol Chem 1993; 268:18532-41.

32 Halliwell B. Superoxide-induced generation of hydroxylradicals in the presence of iron salts. Its role indegradation of hyaluronic acid by a superoxide-generatingsystem. FEBS Lett 1978; 96: 238-42.

33 Grootveld M, Henderson E B, Farrell A, Blake D R,Parkes H G, Haycock P. Oxidative damage tohyaluronate and glucose in synovial fluid during exerciseof the inflamed rheumatoid joint. Biochem Jf 1991; 273:459-67.

34 Wmyard P G, Pall H, Lunec J, Blake D R. Non-caeruloplasmin copper (phenanthroline-copper) is notpresent in fresh serum or synovial fluid from patients withrheumatoid arthritis. BiochemJ 1987; 247: 245-7.

35 Gutteridge J M C. Bleomycin-detectable iron in knee-jointsynovial fluid from arthritic patients and its relationshipto the extracellular antioxidant activities of caerulo-plasmin, transferrin and lactoferrin. Biochem_J 1987; 245:415-21.

36 Gutteridge J M C, Rowley D A, Halliwell B. Superoxide-dependent formation of hydroxyl radicals and lipidperoxidation in the presence of iron salts. Bioche,n Jf 1982;206: 605-9.

37 Kaur H, Fagerheim I, Grootveld M, Puppo A, Halliwell B.Aromatic hydroxylation of phenylalanine as an assay forhydroxyl radicals. Anal Biochem 1988; 172: 360-7.

38 Gutteridge J M C. Iron promoters of the Fenton reactionand lipid peroxidation can be released from haemoglobinbyperoxides. FEBS Lett 1986; 201: 291-5.

39 Biemond P, Swaak A J G, Van Eijk H G, Koster J F.Intraarticular ferritin-bound iron in rheumatoid arthritis.Arthritis Rheum 1986; 29: 1187-93.

40 Kawasaki N, Tanimoto T, Tanaka A, Hayakawa T,Miyasaka N. Determination of non-protein-boundiron in human synovial fluid by high-performanceliquid chromatography with electrochemical detection.JChromatogr 1994; B656: 436-40.

41 Henderson E B, Grootveld M, Farrell A, Smith E C,Thompson P W, Blake D R. A pathological role fordamaged hyaluronan in synovitis. Ann Rheum Dis 1991;50: 196-200.

42 Farrell A J, Blake D R, Palmer R M J, Moncada S. Increasedconcentrations of nitrite in synovial fluid and serumsamples suggest increased nitric oxide synthesis inrheumatic diseases. Ann Rheum Dis 1992; 51: 1219-22.

43 Kaur H, Halliwell B. Evidence for nitric oxide-mediatedoxidative damage in chronic inflammation. FEBS Lett1994; 350: 9-12.

44 Candeias L P, Patel K B, StratfordM R L, Wardman P. Freehydroxyl radicals are formed on reaction between theneutrophil-derived species superoxide anion and hypo-chlorous acid. FEBS Lett 1993; 333: 151-3.

45 Robinson J, Watson F, Bucknall R C, Edwards S W.Activation of neutrophil reactive-oxidant production bysynovial fluid from patients with inflammatory jointdisease. BiochemJ 1992; 286: 345-5 1.

46 Halliwell B, Wasil M, Grootveld M. Biologically-significantscavenging of the myeloperoxidase-derived oxidanthypochlorous acid by ascorbic acid. FEBS Lett 1987; 213:15-8.

47 Chidwick K, WWmyard P G, Zhang Z, Farrell A J,Blake D R. Inactivation of the elastase inhibitory capacityof oa,-antitrypsin in fresh samples of synovial fluid frompatients with rheumatoid arthritis. Ann Rheum Dis 1991;50: 915-6.

48 Davies J M S, Horwitz D A, Davies K J A. Potential rolesof hypochlorous acid and N-chloroamines in collagenbreakdown by phagocytic cells in synovitis. Free Radic BiolMed 1993; 15: 637-43.

49 Hoffstein S T, Gennaro D E, Meunier P C. Cytochemicaldemonstration of constitutive H2 02 production bymacrophages in synovial tissue from rats with adjuvantarthritis. Am J Pathol 1988; 130: 120-5.

50 Merry P, Grootveld M, Lunec J, Blake D R. Oxidativedamage to lipids within the inflamed human jointprovides evidence of radical-mediated hypoxic-reperfusion injury. Am J7 Clin Nutr 1991; 56: 362-9S.

51 Stevens C R, Benboubetra M, Harrison R, Sahinoglu T,Smith E C, Blake D R. Localization of xanthine oxidaseto synovial endothelium. Ann Rheum Dis 1991; 50:760-2.

52 Miesel R, Zuber M. Elevated levels of xanthine oxidasein serum of patients with inflammatory and auto-immune rheumatic diseases. Inflammation 1993; 17:551-61.

53 Halliwell B, Evans P J, Kaur H, Chirico S. Drug-derivedradicals: mediators of the side effects of anti-inflammatorydrugs?Ann Rheum Dis 1992; 51: 1261-3.

54 Grisham M B, Ware K, Marshall S, Yamada T, Sandhu I S.Pro-oxidant properties of 5-aminosalicylic acid. Dig DisSci 1992; 37: 1383-9.

55 Evans P J, Akanmu D, Halliwell B. Promotion of oxidativedamage to arachidonic acid and ac,-antiproteinase by anti-inflammatory drugs in the presence of the haem proteinsmyoglobin and cytochrome c. Biochem Pharmacol 1994;48: 2173-9.

56 Rowley D A, Gutteridge J M C, Blake D R, Farr M,Halliwell B. Lipid peroxidation in rheumatoid arthritis.Clin Sci 1984; 66: 691-5.

57 Fairburn K, Grootveld M, Ward R J, et al. cx-Tocopherol,lipids and lipoproteins in knee-joint synovial fluid andserum from patients with inflammatory joint disease. ClinSci 1992; 83: 657-64.

58 Wmyard P G, Tatzber F, Esterbauer H, Kus M L,Blake D R, Morris C J. Presence of foam cells containingoxidized low density lipoprotein in the synovial mem-brane from patients with rheumatoid arthritis. Ann RheumDis 1993; 52: 677-80.

59 Selley M L, Boume D J, Bartlett M R, et al. Occurrenceof (E)-4-hydroxy-2-nonenal in plasma and synovial fluidof patients with rheumatoid arthritis and osteoarthritis.Ann Rheum Dis 1992; 51: 481-4.

60 Washko P, Rotrosen D, Levine M. Ascorbic acid transportand accumulation in human neutrophils.

_Biol Chem

1989; 264: 18996-9002.61 Humad S, Zarling E, Clapper M, Skosey J K. Breath

pentane excretion as marker of disease activity inrheumatoid arthritis. Free Radic Res Commun 1988; 5:101-6.

62 Phillips M, Greenburg J, Sabas M. Alveolar gradient ofpentane in normal human breath. Free Radic Res 1994; 20:333-7.

63 Grootveld M, Halliwell B. Measurement of allantoin anduric acid in human body fluids. Biochem J 1987; 243:803-8.

64 Kaur H, Halliwell B. Action of biologically-relevantoxidizing species upon uric acid. Chem Biol Interact 1990;73:235-47.

65 Grootveld M, Halliwell B. Aromatic hydroxylation as apotential measure of hydroxyl radical formation in vivo.BiochemJ7 1986; 237: 499-504.

66 Ingelman-Sundberg M, Kaur H, Terelius Y, Persson J 0,Halliwell B. Hydroxylation of salicylate by microsomalfractions and cytochrome P-450. Biochem 7 1991; 276:753-7.

67 Lunec J, Blake D R, McCleary S J, Brailsford S,Bacon P A. Self-perpetuating mechanisms of immuno-globulin G aggregation in rheumatoid inflammation.JClin Invest 1985; 76: 2084-90.

68 Jasin H E. Oxidative cross-linking of immune complexes byhuman polymorphonuclear leukocytes. J Clin Invest 1988;81: 6-15.

69 Bashir S, Harris G, Denman M A, Blake D R,Wmyard P G. Oxidative DNA damage and cellularsensitivity to oxidative stress in human autoimmunedisease. Ann Rheum Dis 1993; 52: 659-66.

70 Lunec J, Herbert K, Blount 5, Griffiths H R, Emery P8-Hydroxydeoxyguanosine. A marker of oxidative DNAdamage in systemic lupus erythematosus. FEBS Lett1994;348: 131-8.

71 Chapman M L, Rubin B R, Gracy R W. Increased carbonylcontent of proteins in synovial fluid from patients withrheumatoid arthritis. J Rheumatol 1989; 16: 15-18.

72 Bates E J, Johnson C C, Lowther D A. Inhibition ofproteoglycan synthesis by hydrogen peroxide in culturedbovine articular cartilage. Biochim Biophys Acta 1985;838: 221-8.

509

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from

Halliwell

73 Baker M S, Bolis S, Lowther D A. Oxidation of articularcartilage glyceraldehyde-3-phosphate dehydrogenase(G3PDH) occurs in vivo during carrageenin-inducedarthritis. Agents Actions 1991; 32: 299-304.

74 Schalkwijk J, Van den Berg W B, Van de Putte L B A,Joosten L B A. An experimental model for hydrogenperoxide-induced tissue damage. Arthitis Rheum 1986;29: 532-8.

75 Michaelis J, Vissers M C, Wimterbourn C C. Differenteffects of hypochlorous acid on human neutrophilmetalloproteinases. Arch Biochem Biophys 1992; 292:555-62.

76 Vincent F, Corral M, Defer N, Adolphe M. Effect ofoxygenfree radicals on articular chondrocytes in culture. Exp CellRes 1991; 192: 333-9.

77 Garrett I R, Boyce B F, Oreffo R 0 C, Bonewald L,Poser J, Mundy G R. Oxygen-derived free radicalsstimulate osteoclastic bone resorption in rodent bonein vitro and in vivo. Clin Invest 1990; 85: 632-39.

78 Bax B E, Alam A S M T, Banerji B, et al. Stimulation ofosteoclastic bone resorption by hydrogen peroxide.Biochem Biophys Res Commun 1992; 183: 1153-8.

79 MacIntyre I, Zaidi M, Alam A S M, et al. Osteo-clastic inhibition: An action of nitric oxide not mediatedby cyclic GMP. Proc Nail Acad Sci USA 1991; 88:2936-40.

80 Shapiro I M, Leboy P S, Tokuoka T, et al. Ascorbic acidregulates multiple metabolic activities of cartilage cells.Am Clin Nutr 199 1; 54: 1209-13S.

81 Williams R 0, Feldmann M, Maini R N. Anti-tumornecrosis factor ameliorates joint disease in murinecollagen-induced arthritis. Proc Natl Acad Sci USA 1992;89: 9784-8.

82 Zimmerman R J, Chan A, Leadon S C. Oxidative damagein murine tumor cells treated in vitro by recombinanthuman tumor necrosis factor. Cancer Res 1989; 49:1644-8.

83 Schulze-Osthoff K, Bakker A C, Vanhaesebroeck B, et al.Cytotoxic activity of tumor necrosis factor is mediated byearly damage ofmitochondrial functions. Evidence for theinvolvement of mitochondrial radical generation. BiolChemn 1992; 267: 5317-23.

84 Halliwell B. Drug antioxidant effects. A basis for drugselection? Drugs 1991; 42: 569-605.

510

on March 25, 2021 by guest. P

rotected by copyright.http://ard.bm

j.com/

Ann R

heum D

is: first published as 10.1136/ard.54.6.505 on 1 June 1995. Dow

nloaded from