1oksidative Stress Test

can initiate cellular tissue damage by modifyinglipids, proteins and DNA, which can seriouslycompromise cell health and viability or induce avariety of cellular responses through generationof secondary reactive species, leading, at last, tocell death by necrosis or apoptosis. Oxidativedamage of any of these biomolecules, ifunchecked, is probably responsible of disease de-velopment. However, definitive evidence for thisassociation is often lacking because of recog-nized shortcomings with methods available to as-sess oxidative stress status in vivo in humans3.

There are some exogenous sources of free rad-icals such as UV-photolysis, radiation, ozone,pollution, pharmacological agents, smoking, al-cohol, iron-overload, pesticides and mycotoxins4.Imbalance between production and eliminationof free radicals may cause oxidative stress.

Free radicals can be scavenged by several met-alloenzymes (e.g., glutathione peroxidase, cata-lase, superoxide dismutase) as well as by thenon-enzymatic antioxidant defence system (e.g.,tocopherol, -carotene, ubiquinol, vitamin C,glutathione, lipoic acid, uric acid, metalloth-ionein, bilirubin) which quench their activity.Therefore, much attention of nutritionists is nowfocused on the possible role of the enhancementof the defences against ROS5 (Table II).

Despite the harmful cellular damaging effects,free radical reactions are also involved into bene-ficial physiological response when produced inhigh levels mediating cytotoxicity of polymor-phonuclear leukocytes, macrophages and mono-cytes during the respiratory-burst6.

Moreover, low levels of ROS are involved inthe regulation of the tone of smooth muscle cells7and have been demonstrated to upregulate the re-dox sensitive transcription factors such as nu-clear factor-B and activator protein-18-10.

309

Abstract. Monitoring oxidative stressin humans is achieved by assaying products ofoxidative damage or by investigating the po-tential of an organism, tissue or body fluids towithstand further oxidation. Unfortunately,there is little consensus concerning the selec-tion of parameters of oxidative stress or an-tioxidant state to be determined in defined pa-tients or diseases. This is not only due to theuncertainty wheter or not a certain parameteris playing a causative role. Moreover, themethods of determination described in the lit-erature represent very different levels of ana-lytical practicability, costs, and quality. Gener-ally accepted reference ranges and interpreta-tions of pathological situations are lacking aswell as control materials. At present, the situa-tion is changing dramatically and sophisticat-ed methods like HPLC (High Performance Liq-uid Chromatography) and immunochemical de-terminations have become more and morecommon standard.

Key Words:

Oxidative stress, Free radicals, Antioxidants, Reac-tive oxygen species, Lipid peroxidation.

Introduction

The role of free radicals is gaining increasingworldwide attention since so many physiologicaland pathophysiological phenomena are related toredox status cell modification.

A free radical is, by definition, a chemicalspecies containing unpaired electrons and istherefore paramagnetic1. Most of the oxygen de-rived free radicals relevant to cell biology are un-stable, short-lived and highly reactive2 (Table I).For these reasons, reactive oygen species (ROS)

European Review for Medical and Pharmacological Sciences 2007; 11: 309-342

Oxidative stress tests: overview onreliability and usePart I

B. PALMIERI, V. SBLENDORIO

Department of General Surgery and Surgical Specialties, University of Modena and Reggio EmiliaMedical School, Surgical Clinic, Modena, Italy

Corresponding Author: Beniamino Palmieri, MD; e-mail: [email protected]

310

Highly specific analytical techniques are re-quired to monitor the biological significance offree radicals.

Increased oxidative/nitrosative stress generallydescribes a condition in which cellular antioxi-dant defences are unable to completely inactivate

the ROS11 and reactive nitrogen species (RNS)generated because of excessive production ofROS/RNS, loss of antioxidant defences, or both.

The localization and effects of oxidativestress, as well as information regarding the na-ture of the ROS/RNS, may be revealed from theanalysis of discrete biomarkers of oxidative/ni-trosative stress/damage isolated from tissues andbiological fluids. Biomarkers are quali-quantita-tive indicators of normal and pathological bio-chemical processes or of druginduced effect intherapeutic protocols. Several in vitro markers ofoxidative/nitrosative stress are available, includ-ing ROS/RNS themselves, but most are of limit-ed value in vivo because they lack sensitivityand/or specificity or require invasive methods.Although some ROS/RNS have been directly de-tected in vitro by electron spin resonance with orwithout spin trapping reagents or by chemilumi-nescence, these methods are not yet applicable inclinical practice because of the instability ofmany reactive species and the need for expensive

B. Palmieri, V. Sblendorio

Half-lives(s)Molecules Symbol at 37C

Molecular oxygen O2 > 102Lipid peroxide ROOH > 102Semiquinone radical Q > 102Hydrogen peroxide H2O2 10Peroxyl radical ROO 1 102Superoxide radical O2 1 106Singlet oxygen 1O2 1 106Alkoxyl radical RO 1 106Hydroxyl radical OH 1 109

Table I. Free radicals.

Preventive antioxidants:a) Non-radical decomposition of hydroperoxides and hydrogen peroxide:

Catalase: Decomposition of hydrogen peroxide: 2H2O2 2H2O + O2Glutathione peroxidase (cellular) Decomposition of hydrogen peroxide and free fatty acid

hydroperoxides:H2O2 + 2 GSH 2 H2O + GSSGLOOH + 2GS LOH + H2O + GSSG

Glutathione peroxidase (plasma) Decomposition of hydrogen peroxide and phospholipid hydroperoxides

Phospholipid hydroperoxide PLOOH + 2GSH PLOH + H2OGlutathione peroxidase GSSGGlutathione-S-transferase Decomposition of lipid hydroperoxidesThioredoxin Reduction of peroxides

b) Sequestration of metals by chelationTransferrin, lactoferrin: IronHaptoglobin HaemoglobinHemopexin Stabilisation of hemeCeruloplasmin, albumin Copper

c) Quenching of active oxygensSuperoxide dismutase (SOD) Disproportionation of superoxide: 2O2 + 2H + H2O2 + O2Carotenoids, vitamin E Quenching of singlet oxygen

Radicals-scavenging antioxidants: scavenge radicals to inhibit chain initiation and break chain propagationLipophilic: Vitamin E, ubiquinol, carotenoidsHydrophilic: Vitamin C, uric acid, bilirubin, albumin

Repair and de novo enzymes: repair the damage and reconstitute membranes:Lipase, protease, DNA repair enzymes, transferase

Adaptation: generate appropriate antioxidant enzymes and transfer them to appropriate site at the right time andconcentration

Table II. Antioxidant defence system.

equipment. Furthermore, ROS/RNS are usuallytoo reactive and/or have a half-life too short(even much shorter than seconds) to allow directmeasurements in cells/tissues or body fluids. Be-cause molecular products formed from the reac-tion of ROS/RNS with biomolecules are usuallyconsidered more stable than ROS/RNS them-selves, most commonly ROS/RNS have beentracked by measuring stable metabolites (e.g., ni-trate/nitrite) and/or concentrations of their oxida-tion target products, including lipid peroxidationend products and oxidized proteins12-16. Tech-niques for quantification of oxidative damagemarkers are often called fingerprinting methodsby which specific end products deriving from theinteraction of the ROS with biomolecules, suchas DNA, proteins, lipid and LMWA (low-molec-ular-weight antioxidant) are measured. The pres-ence of these end products serves as proof of theprior existence of ROS that left their footprints inthe cell. To function as suitable biomarkers ofoxidative modifications in relation to disease, itis critical that such oxidation products are stable,can accumulate to detectable concentrations, re-flect specific oxidation pathways and correlatewith disease gravity, so that they can be utilisedas diagnostic tools.

To demonstrate a role of ROS in a particulartype of tissue injury, evidence should be present-ed that:

1. ROS are detectable locally and the time-course of their formation is such that theycould play a role;

2. the chemical production of ROS producessimilar lesions;

3. compounds able to remove ROS protect fromthe injury.

Measuring Free Radicals in VivoThe increasing interest in the role of free radi-

cals in the pathogenesis of human disease has ledto widespread attempts to develop techniquessuitable to measure free radicals and their reac-tions in vivo, specifically, in clinical pathology.The first major problem to be faced is the quickreactivity of free radicals reaction close to theirbiochemical source. Consequently, free radicalsare not amenable to direct assay and free radicalactivity is usually assessed by indirect methodssuch as measurement of the various end productsof reactions with lipids, proteins and DNA17,18.However, many of these products are themselvesreactive, albeit orders of magnitude less than the

free radicals that begat them. The second majorproblem is that the most commonly available bi-ological fluid to be screened are blood, urine andexpired breath. Clinical biochemistry detectsusually abnormal metabolic products, recoveredfrom these sources, which are related to specificdiseases. On the contrary, reactive free radicalsas end products of intracellular metabolism fromdifferent tissues have a microseconds-measurablehalf life and they are not detectable in the bloodstream. In a very few special cases, the actualsite of free radical generation may be the bloodand direct (or semi-direct) detection of free radi-cal species may be possible, but generally speak-ing only secondary free radical products are de-tectable in a body fluid. A wide array of analyti-cal techniques has been developed to measurethese end products though not all of them aresuitable to detect clinical conditions samplingblood, urine and expired breath. Lipid peroxida-tion is the most intensively studied process andprovides a number of possibilities for assays.Protein and nucleic acid oxidation are presentlyvery appealing. The currently available tech-niques, however, are limited to semi-quantitativeassays of damage to broad classes of biomole-cules and there is an urgent need for more specif-ic and informative methods.

Electron Spin Resonance and Radical Trapping

The only analytical technique that directlymeasures free radicals is electron spin resonance(ESR) spectrometry. However, since it is rela-tively insensitive and requires steady-state con-centrations of free radicals in the micromolarrange its of very limited value for use in vivo.Whole-body ESR, analogous to whole-bodyNMR, has been investigated but not yet fully de-veloped. Nevertheless, ESR has been used to de-tect free radicals in human tissue obtained ex vi-vo: an example is the detection of a signal be-lieved to be that of lipid peroxyl radical in hu-man uterine cervix19. ESR spectrometry can usu-ally be applied to analysis of samples in vivo on-ly through the technique of spin trapping. Thisinvolves the addition to samples of a compoundknown as spin-trap, which reacts rapidly with thefree radicals to form radical-adducts that are verymuch more stable and longer-lived than the origi-nal species and can therefore build up to steady-

311

Oxidative stress tests: overview on reliability and use. Part I

312

state concentrations in the detactable range.Spin-traps have been used in experimental ani-mals to demonstrate the generation of free radi-cals in vivo, but as no effective spin trapspresently exist that can be administered to hu-mans, the technique is currently limited to sam-ples of blood mixed with the spin trap as soon aspossible after taking them. Despite the obviousshortcomings of this approach, valuable data hasbeen obtained, for example, relating to free radi-cal production during angioplasty20.

Other trapping procedure allow a radical to re-act with a detector molecule to yield a stableproduct that can be evaluated using a variety oftechniques, such as hydroxylation of salicyclicacid21, the deoxyribose assay22, 23, the cytochromec reduction assay for detection of superoxide rad-icals24, and detection of nitric oxide radicals bycolored end-product compounds25.

Thus, the attack of hydroxyl radicals on sali-cylic acid produces 2,3-dihydroxybenzoate(DHB) and on phenylalanine produces o-and m-tyrosines. These products are not produced enzy-matically in humans. Thus, the method can beused in vivo and detection of 2,3-DHB or the ty-rosines in body fluids can be taken as evidenceof hydroxyl radical generation26. As the trappingcompound has to complete with all other biomol-ecules for reaction with the radicals, this tech-nique, like ESR-spin-trapping, is unlikely to pro-vide more than semi-quantitative data.

Spin trapping is a powerful method that facili-tates the visualization of free radicals, includingthose formed in complex biological systems. Thespin trap is a diamagnetic compound that reactswith a reactive free radical to form a more stableradical adduct. Although detection through ESRspectroscopy offers some distinct advantages inits high sensitivity, and in some cases its speci-ficity toward some radical species, there are alsoseveral drawbacks to using this technique.

The technique was developed in the lates1960s by several laboratories27. Two groups ofcompounds are commonly utilized as spin-trap-ping agents: nitroso and nitrone compounds. Thenitrogen atom of the nitroso spin trap reacts di-rectly with the free radical species, giving dis-tinctive spectral features. Two nitroso com-pounds are currently used in biological investiga-tion: 2-methyl-2-nitroso propane (MNP) and 3,5-dibromo-4-nitrosobenzenesulfonate (DBNBS).

The nitrone spin traps each have a radicaladded to the carbon; the radical will be with re-spect to the nitroxide radical center. The lack of

spectral information about the trapped radical isthe major drawback of this class of spin trap.Three commonly used spin traps will be dis-cussed: phenyl-t-butyl nitrone (PBN), (4-pyridyl-1-oxide)-N-t-butyl nitrone (POBN), and5,5-dimethyl-1-pyrroline N-oxide (DMPO).

The technique has been associated with vari-ous cases of incorrect interpretations; these gen-erally can be attributed to:

1. changes in the spin trap (nonradical, chemical,photochemical, enzymatic reactions);

2. perturbation of the biological system by theprobe;

3. artifactual reporting associated with intrinsicproperties of the probe.

Spin trapping is a good example of the com-plex interaction between the model system andthe artificial addition of a probe, with a high pos-sibility of recording artifactual results.

The first that has proposed the term spin trap-ping has been Janzen27. Spin trapping in biologyis covered by various reviews28,29. An extensiveliterature survey has been carried out by Dodd30.Specifically devoted to examining the problemsassociated with the spin trapping of oxygen-cen-tered free radicals are the reviews of Finkelsteinet al1, Rosen and Rauckman31, Rosen et al32, andPou and Rosen33.

Invaluable help in disentangling the number ofspectra and attributions is given in the databasefor spin-trapping by Li and Chignell34, which hasbeen made feely available to all those interestedin the field.

Electron Paramagnetic Resonance(EPR)

Another technique for the measurement of theoxidative stress status in biological systems isbased on the X-band EPR (electron paramagneticresonance) detection of a persistent nitroxidegenerated under physiological or pseudo-physio-logical conditions by oxidation of a highlylipophylic hydroxylamine probe. The probe em-ployed is bis(1-hydroxy-2,2,6,6-tetramethyl-4-piperidinyl)-decandioate which is administratedas hydrochloride salt. This way of making OSstatus detectable involves the use of exogenousnitroxides as probe of the redox balance in a giv-en enviroment. This probe is able to give a fast


reaction with the most of radical species in-volved in the oxidative stress. The rate at whichthe nitroxide is reduced to the diamagnetic hy-droxylamine, which can be evaluated by EPR, isrelated to the reducing capacity of the organismand hence to its oxidative status35. Furthermore,it crosses cell membranes and distributes in a bi-ological environment without the need to alter ordestroy compartmentation. The method is there-fore suitable for quantitative measurements ofROS and can be applied to human tissues in realclinical settings36. It has been successfully em-ployed in systems of growing complexity and in-terest, ranging from subcellular fractions towhole animals and human liver. Liver diseasewas chosen as the prototype of a patology inwhich the involvement of inflammatory process-es has a relevant role in the evaluation of the dis-ease37. Thirty-two subjects, including 10 healthycontrols, were enrolled after giving informedconsent. Ten of the 22 patients had hepatitis C, 3had hepatitis B, while the remainder had a vari-ety of diseases characterized by an autoimmunenature which, for statistical purpose, were clus-tered in a group called nonviral liver diseases(NVLD). The method developed by Valgimigli etal38 was enough simple and only moderately in-vasive: 2-3 mg of liver biopsy (obtained by thefine needle technique) were weighted and incu-bated for 5 minutes at 37C with a physiologicalsolution of the hydroxylamine I (1 mM) contain-ing a metal chelating agent. After incubation, thesample was quickly frozen in liquid nitrogen todenaturate enzymes and stop any reaction, andsubsequently warmed at room temperature priorto the EPR measurement. For practical reasons,these researchers monitored the maximum con-centration of nitroxide instead of the full timeevolution. Diseased tissue provided a more oxi-dizing enviroment than healthy liver. Further-more, the nature of the disease affected the ox-idative status.

The effect of the various experimental condi-tions on the final result, including lenght of incu-bation, time from tissue extraction to addition ofthe probe and time from incubation to EPR mea-surement, were sistematically investigated in or-der to set the optimal standardized experimentalconditions. Interestingly, these results revealedthat omogenization of the tissue is unnecessarysince the signal measured immediately afteromogenization in the presence of the probe wasvery close to that obtained after 5 minutes incu-bation with the whole biopsy. After calibration of

the spectrometer response it was possible to ob-tain quantitative values for the oxidative stress.These results indicate that the OS level in dis-eased liver is several orders of magnitude higherthan in healthy controls and the differences werehighly significant.

An endogenous molecule might also functionas a trap, although it can be argued that measur-ing specific end products of the trapping of RS(reactive species) by endogenous molecules isthe same as measuring biomarkers. Ascorbatereaction with free radicals is one example; anoth-er is urate, which is readily oxidized by a rangeof RS39, including proxynitrite40. Several groupshave used urate as a selective scavenger ofONOO- in animal studies, neglecting the fact thatit reacts with many species41. One of urates oxi-dation products, allantoin, can be measured inhuman body fluids and its plasma levels are ele-vated in conditions associated with oxidativestress, such as chronic inflammation, diabetes,premature birth, iron overload, chronic hearthfailure and exercise42-45. Allantoin can also bemeasured in urine46 and cerebral microdialysisfluid47. Levels of allantoin rise in the humanmuscle during exhaustive exercise, presumiblydue to oxidation of urate by RS generated duringexercise48. Allantoin measurement may be one ofthe more promising techniques for human use,since human urate levels in vivo are high andurate reacts with a wide range of RS3.

Nuclear Magnetic Resonance (Nmr)Based Metabolomics/Metabonomics

Analysis of Biofluids

Metabolomics (also called metabonomics) isdefined as the quantitative measurement of thedynamic multiparametric metabolic response ofliving systems to pathophysiological stimuli orgenetic modification49. High resolution nuclear1H-magnetic resonance (NMR) spectroscopicanalysis of biofluids allows simultaneous detec-tion of hundreds of low molecular weight specieswithin a sample of body fluid, resulting in thegeneration of a metabolic profile or NMR fin-gerprint that is altered characteristically in re-sponse to physiological status50. Once NMRspectra are obtained, the highly complex spectraare analyzed using pattern recognition and multi-variate statistical methods to produce models forsamples classification51. This technology has

313


314

been widely applied to toxicology studies with arange of biological fluids, such as urine and plas-ma, in both experimental animals andhumans52,53. Statistical analysis of urine sampleshas been shown to result in inherent clusteringbehaviour for drugs and toxins acting on differ-ent organs, such as liver or kidney, or having dif-ferent toxic mechanisms. These cluster analysesare similar to those currently being developed forgene array expression analysis and proteomics,and have been demonstrated to classify toxins intest samples correctly. Metabolomic analysis,therefore, seems particularly suited to the analy-sis of biofluids from clinical/laboratory studieswith the potential to measure simultaneously arange of oxidative stress products and other in-flammatory markers. Analysis of such samplesmay lead to the identification of novel single bio-markers of interest for wider study in patientpopulations. Brindle and colleagues54 studiedserum metabolome obtained from coronary heartdisease and healthy individuals. Moreover, appli-cation of metabolomics allows the simultaneousanalysis of multiple end products and it may bethat these fingerprints characteristic of diseaseare a more powerful and robust means by whichto stratify disease severity, progression and to as-sess drug efficacy than the analysis of any singlemarker over a patient population.

Online Measurements of Oxidative Stress Biomarkers

Infrared laser spectroscopy is a promisingmethod for free radical research, enabling onlinemeasurement of oxidative stress biomarkers,such as lipid peroxidation products, with highsensitivity and efficiency. Murtz et al55 have de-veloped real-time analysis of volatile ethane frac-tions in exhaled breath (gaseous molecularspecies) using laser absorption spectroscopy. Thegroup monitored the ethane fraction exhaled by asmoker after smoking a cigarette every 30 minover a period of 4 h, and observed a strong in-crease and subsequent decay of the ethane frac-tion after smoking. This method is unique, withvery sensitivity and specificity for rapid and pre-cise breath testing. The detection limit is 300volume parts per trillion ethane in exhaled breathwith an integration time of 5 s. Another majoradvantage of this method is that it allows theanalysis of biomarkers without pre-concentration

or pre-treatment of exhaled breath. The develop-ment and introduction of this biosensor techniquefor immediate analysis of EBC (exhaled breathcondensate) has potential for undertaking real-time EBC monitoring of oxidative stress in ani-mal research and clinical practice. Newer tech-niques, such as online masurements using sensi-tive biosensors, are being developed for more re-producible measurement of hydrogen peroxide.For example, it is possible to detect hydrogenperoxide online (real-time) using a silver elec-trode or polymer with horseradish peroxidase56,57.A similar enzyme detector system also may bedeveloped for real-time monitoring of 8-iso-prostane.

Lipid Peroxidation

Lipid peroxidation is a complex processwhereby polyunsaturated fatty acids (PUFAs) inthe phospholipids of cellular membranes undergoreaction with oxygen to yield lipid hydroperox-ides (LOOH). The reaction occurs through a freeradical chain mechanism initiated by the abstrac-tion of a hydrogen atom from a PUFA by a reac-tive free radical, followed by a complex se-quence of propagative reactions.

The LOOH and conjugated dienes that areformed can decompose to form numerous otherproducts including alkanals, alkenals, hydrox-yalkenals, malondialdehyde (MDA) and volatilehydrocarbons58. Lipid peroxidation is often thefirst parameter to which researchers turn whenthey wish to prove the involvement of free radi-cals in cell damage. There are several reasonsfor this. First, lipid peroxidation is an extremelylikely consequence if a reactive free radical isformed in a biological tissue where PUFAs aregenerally abundant. Second, lipid peroxidationis a very important process in free radicalpathology as its so damaging to cells. Finally, avast array of analytical techniques has been de-veloped to measure lipid peroxidation, thoughnot all of them are applicable to the situation invivo59.

For all assays its important that artifactualchanges in lipid peroxidation products are min-imised both during and after sampling. Radical-scavenging antioxidants and metal-chelatingagents are added to prevent the further formationof lipid hydroperoxides and the breakdown ofexisting lipid hydroperoxides. Enzymic reactions


that may affect levels of products are inhibitedby mixing the sample with acid or organic sol-vents. It is generally advisable to assay samplesas quickly as possible after taking them, since atendency to increased lipid peroxidation on stor-age has been reported60,61. Conversely, lipid hy-droperoxides can deteriorate on storage62.

The lipid peroxidations reaction in biologicalmembranes causes impairment of membranefunctioning63,64, decreases fluibility, inactivationof membrane-bound receptors and enzymes andincreases non-specific permeability to ions suchas Ca2+. Additionally, lipid hydroperoxides de-compose upon exposure to iron or copper ions,simple chelates of these metal ions (e.g. withphosphate esters), haem, and some iron proteins,including haemoglobin and myoglobin. Productsof these complex decomposition reactions in-clude hydrocarbon gases (such as ethane andpentane), radicals that can abstract further hydro-gen atoms from fatty acid side chains and cyto-toxic carbonyl molecules, of which the mostharmful are the unsaturated aldehydes such as 4-hydroxy-2-trans-nonenal. Indeed, a major con-tributor to extracellular antioxidant defence inmammals is the existence in body fluids of pro-teins that bind copper ions (caeruloplasmin andalbumin), iron ions (transferrin), haem(haemopexin) or haem proteins (haptoglobins)and stop them from accelerating lipid peroxida-tion and other free radical reactions65,66.

Biomedical Lipid PeroxidationThe measurement of putative elevated end

products of lipid peroxidation in human samplesis probably the evidence most frequently quotedin support of the involvement of free radical re-actions in tissue damage by disease or toxins.Studies beginning in the 1950s provided goodevidence that several halogenated hydrocarbonsexert some, or all, of their toxic effects by stimu-lating lipid peroxidation in vivo. This is particu-larly true of carbon tetrachloride and probablytrue of bromobenzene.

This early choice of halogenated hydrocarbonsfor study was both casual (in that it gave earlyemphasis to the important biological role of freeradical reactions) but also unfortunate, since laterstudies have shown that most toxins stimulatingoxidative damage to cells do not appear to act byaccelerating the bulk peroxidation of cell mem-brane lipids67:

toxin lipid peroxidation cell damage

Rises in intracellular free Ca2+, with conse-quent activation of proteases and nucleases andformation of membrane blebs, oxidation ofcritical SH groups and DNA damage are oftenmore relevant toxic events than is the bulk perox-idation of membrane lipids68.

Lipid peroxidation is often (but by no meansalways) a late event, accompanying rather thancausing final cell death69. Indeed, cell and tissuedestruction (whether mediated by radicals orotherwise) can often lead to more lipid peroxi-dation because antioxidants are diluted out andtransition metal ions that can stimulate the per-oxidation process are released from disruptedcells.

This stimulation of lipid peroxidation as a con-sequence of tissue injury can sometimes make arelevant contribution to worsening the injury. Forexample, in atherosclerosis there is good evi-dence that lipid peroxidation occurs within theatherosclerotic lesion and leads to foam cell gen-eration and hence lesion growth70. In traumaticinjury to the brain and spinal cord, good evi-dence again exists that iron ion release into thesurrounding area, and consequent iron-stimulatedfree radical reactions, worsen the injury71.

It is equally likely that in some other dis-eases, the increased rates of free radical reac-tions induced as a result of tissue injury makeno significant contribution to the disease pathol-ogy. Each proposal that free radicals in general,or lipid peroxidation in particular, are importantcontributors to the pathology of a given diseasemust be carefully evaluated on its merits. Thisobviously requires accurate methodology formeasuring these processes in cells, tissues andwhole organisms.

Detection and Measurement of Lipid Peroxidation: General Principles

Oxidation of lipids can be measured at differ-ent stages, including:

1. losses of unsaturated fatty acids;2. measurement of primary peroxidation prod-

ucts;3. measurement of secondary carbonyls and hy-

drocarbon gases.

Between phases 1, 2 and 3 it is possible detectcarbon-and oxygen-centred radicals (by ESRcombined with the use of spin traps) and iden-tify these radicals by their ESR spectra72.

315


316

It should be noted that the chemical composi-tion of the end products of peroxidation will de-pend on the fatty acid composition of the lipidsubstrate used and upon what metal ions (if any)are present. Thus, copper and iron ions give dif-ferent end-product distributions and so the selec-tion of only a single test to monitor peroxidationcan give misleading results. Copper salts effi-ciently decompose peroxides, leading to lowconcentrations of detectable peroxides but highamounts of some carbonyl molecules containingamino groups to form fluorescent products. Themost accurate assays of lipid peroxidation are themost chemically sophisticated ones. They alsorequire the most sample preparation and greatcare (e.g. by working under nitrogen) has to betaken to ensure that peroxidation does not occurduring the handling of lipid material.

Measurement of Lipid HydroperoxidesLOOH are the major initial molecular prod-

ucts of lipid peroxidation and can be measured inplasma by a lot of techniques. A sensitive andspecific assay is based on the capacity of LOOHto initiate the cyclooxygenase reaction catalysedby activation of prostaglandin endoperoxide syn-thase and uses an oxygen electrode73. Anothersensitive, even if expensive, method to measureplasma LOOH uses gas chromatography-massspectrometry (GC/MS) and involves reduction ofLOOH to the hydroxy acids with triphenylphos-phine74.

Plasma LOOH can be measured using com-mercially-available assay kits. One such kit reliesupon the reaction of LOOH with a haem com-pound, concomitantly oxidising a precursor toproduce methylene blue which is measured spec-trophotometrically. Although very simple andquick, only total hydroperoxide concentrationsare measured and results do not relate well withsome other measures of lipid peroxidation75.

Some methods have been developed that candistinguish specific or different classes ofLOOH. These are based on separation accordingto lipid class of the various hydroperoxides in aFolch lipid extract of plasma by high perfor-mance liquid chromatography (HPLC) and mea-surement of the chemiluminescence producedduring their breakdown in the presence of eitherluminol76 or isoluminol77.

The pathogenic role of lipid peroxidation inthe reperfusion injury of the liver is still contro-versial. Caraceni et al78 wanted to determinewhether the damage caused by oxygen free radi-

cals during reoxygenation in perfused rat hepato-cytes is related to lipid peroxidation. Superoxideanion was detected by lucigenin-enhancedchemiluminescence. Lipid peroxidation and cellinjury were assessed by the release of malondi-aldehyde and lactic dehydrogenase. Upon reoxy-genation following 2.5 h of anoxia, isolated he-patocytes generated considerable amount of O2.Following O2 formation, a significant increasein malondialdehyde release was measured. Cellinjury was temporally delayed relative to O2generation, but preceded the occurrence of a sig-nificant lipid peroxidation. Treatment with Vita-min E abolished lipid peroxidation but had no ef-fect upon superoxide anion formation and cell in-jury. These results suggest that in perfused rathepatocytes non-peroxidative mechanisms aremore important than peroxidative mechanisms inthe pathogenesis of the early phases of reoxy-genation injury.

Gasbarrini et al79 wanted to determine whetherthe formation of oxygen free radicals occurs inmurine osteoblast-like cells (MC3T3-E1) exposedto anoxia and reoxygenation and to explore its re-lation to the reoxygenation injury. Cells were castin agarose and perfused with oxygenated Krebs-Henseleit bicarbonate buffer. Anoxia was ob-tained by shifting the gas phase of the media to95% N2-5% CO2. Oxygen free radicals were de-tected by enhanced chemiluminescence: anion su-peroxide or hydrogen peroxide was measured byadding lucigenin or luminol plus horseradish per-oxidase to the media, respectively. Cell injurywas assessed by the rate of lactate dehydrogenaserelease. During the control period, lucigenin andluminol plus horseradish chemiluminescenceswere 15 1 nA per chamber and 20 2 nA perchamber, respectively. and lactate dehydrogenaserelease was 10 1 mU per minute. During anox-ia, both chemiluminescences dropped to back-ground levels, although lactate dehydrogenase re-lease increased progressively to 38 7 mU perminute. During reoxygenation, O2 formation in-creased sharply to 45 6 nA and decreased tocontrol levels; H2O2 production increased slowly,reaching 42 7 nA at the end of the reoxygena-tion period; lactate dehydrogenase declined pro-gressively to control values. These data show thatosteoblastlike cells produce measurable amountsof superoxide and hydrogen peroxide radicalsduring reoxygenation. Because lactate dehydroge-nase release did not appear to relate to chemilu-minescence, oxyradical flux may serve as a signalfor other events that eventually lead to cell injury.


Ojetti et al80 evaluated the combined use ofchemiluminescence and gastroendoscopy tech-niques and to assess the real-time production offree radicals during ischemic damage of the gas-tric wall in an animal model. For the experiment,an optical junction was set up between a fibroen-doscope and a luminograph apparatus. Threepigs were submitted to gastrofibroendoscopy be-fore, during and after 30 min of clamping of thecoeliac artery. Under basal conditions, at the endof the ischemic phase and at the beginning ofreperfusion, 1 mM of lucigenin, a specific super-oxide enhancer, was injected in the left gastricartery of the animal. The endoscopic live imagesand chemiluminescence emission were recordedand successively superimposed to measure rateand spatial distribution of photon emission (pho-tons/s). Free radical production was not observedunder basal conditions or during the ischemicphase, but significantly increased during reperfu-sion reaching a maximum peak after 15 min (0.6 0.2 photons 10(5)/s) and decreased progres-sively thereafter. The superimposition of live andchemiluminescence images allowed the determi-nation of the regional production rate and distrib-ution of photons.

The LOOH are identified by comparison withauthentic standards: although hydroperoxides offree fatty acids are commercially available, thoseof others lipids (e.g. phospholipids, cholesterol es-ters) must be synthesised, which is both time-con-suming and inconvenient. The assays have pico-molar sensitivity and additionally, LOOH isachieved by monitoring conjugated dienes either at234 nm or by measuring the complete UV absorp-tion spectrum of the sample with a diode-array de-tector. Alternatively, treatment with the reducingagent sodium borohydride will eliminate thechemiluminescent signal. The assay is relativelyspecific for hydroperoxides although ubiquinols inhuman plasma produce a positive response.

Accurate measurement of LOOH is difficultdue to their rapid degradation in vitro. It is ex-tremely important to minimise this by the addi-tion of antioxidants and quick processing of sam-ples at 4C which is often not possible in a clini-cal situation. Also, these HPLC methods, al-though specific and sensitive, are time-consum-ing in their analysis and preparation of standardsand are best used only when information on indi-vidual hydroperoxides is required. Free fattyacid- and cholesterol-hydroperoxides have beendetected in patients with adult respiratory distresssyndrome81 or undergoing angioplasty.

Measurement of Conjugated DienesLOOH possess a conjugated diene structure

having a characteristic UV absorption around234 nm. Measurement of this absorbance hasbeen extremely useful as an index of peroxida-tion in pure lipid systems and in tissue prepara-tions from experimental animals. There are, how-ever, difficulties in measuring conjugated dienesin biological materials because many of the othersubstances present (e.g. haem proteins) absorbstrongly in the UV and create a high background.This is partly eliminated by extraction of conju-gated dienes into an organic solvent such as chlo-roform/metahnol. However, PUFA, themselves,and carbonyl compounds produced from thebreakdown of LOOH absorb UV light strongly atabout 210 nm so that the conjugated diene ab-sorbance appears as a shoulder on the PUFA ab-sorbance spectrum82.

Measurement is also complicated by the rela-tively low levels of conjugated dienes normallypresent in human plasma. A second derivativespectroscopy method83 allows greater sensitivity,since the conjugated diene shoulder that appearsin the ordinary spectrum translates into a sharpminimum peak that is more easily measurable.The increased resolution of this technique mayallow discrimination between the different conju-gated diene structures present. However, most(90%) of the conugated diene in human plasma isa non-oxygen-containing isomer of linoleic acid(9, 11-octadecadienoic acid) that can be assayedspecifically by HPLC84. This product is notfound in the plasma of animals subjected to ox-idative stress and may be of dietary origin or pro-duced by the metabolism of gut bacteria. Appli-cation of conjugated diene methods to humanbody fluids is thus probably not measuring lipidperoxidation and is not recommended for humanstudies.

Measurement of Thiobarbituric Acid Reacting Substances (TBARS) and Malondialdehyde (MDA)

The thiobarbituric acid (TBA) assay is themost common and easiest method used as an in-dicator of lipid peroxidation and free radical ac-tivity in biological samples. The assay is basedupon the reaction of two molecules of TBA withone of MDA, a physiologic ketoaldehyde pro-duced by peroxidative decomposition of unsatu-rated lipids as a byproduct of arachidonate me-tabolism. The excess MDA produced as a result

317


318

of tissue injury can combine with free aminogroups of proteins (MDA reacts mainly with Lysresidues by Michael addition), producing MDA-modified protein adducts. Modification of pro-teins by MDA could conceivably alter their bio-logical properties. Moreover, MDA-modifiedproteins are immunogenic, and autoantibodiesagainst MDA-modified Lys residues have beendetected in the sera of rabbits and humans. Somestudies have reported that the titer of these au-toantibodies is related to the burden of, and maypredict progression of, atherosclerosis and my-ocardial infarction. Higher titers of autoantibod-ies have also been correlated to coronary arterydisease11.

There are a lot of variations85 but basically thesample is heated with TBA under acidic condi-tions and the amount of pink-coloured MDA-TBA adduct produced is measured at 532 nm.For increased sensitivity, the complex can be ex-tracted into an organic solvent such as butanoland measured fluorometrically86. In a few experi-mental systems the TBA test has been demon-strated actually to be measuring MDA itself. Inuncharacterised systems it is usual to refer to theassay of TBA-reactive substances (TBARS) asthe test is not specific for MDA.

The test itself is very simple and quick but itsapplication to biological samples can be prob-lematic. The exact conditions of the test arevery important. Biological samples normallycontain only a small amount of free MDA andin tests where the unseparated sample is incu-bated for a prolonged time the majority of theMDA measured is formed by the decompositionof LOOH and further peroxidation during theheating stage of the assay itself. The widelyused Yagi test that utilise TBA87, are probablyassays of lipid hydroperoxides. Various biologi-cal compounds react with TBA and the fluores-cence method may be more selective than spec-trophotometry. Other factors that can markedlyaffect the apparent concentration of TBARS inplasma include the iron content of reagents usedin the analysis and the storage of samples at70C, although the addition of EDTA tochelate iron may reduce variability. To furtherminimise the problems related with the TBAtest, the MDA-TBA adduct may be measuredby HPLC and GC although this is time-consum-ing, involving complex sample preparation toremove contaminants or sample extraction intoorganic solvents to improve sensitivity and peakseparation.

Direct assessment of free MDA is most reli-ably done by HPLC but the technique requiresvery careful handling of the sample. However,MDA is a minor product of lipid peroxidationand is readily metabolised; it is therefore not apromising subject for the analysis of lipid peroxi-dation in vivo.

Plasma MDA concentrations are increased indiabetes mellitus and MDA can be found in theatherosclerotic plaques promoted by diabetes88.Increasead MDA concentrations have been foundin samples from women with preeclampsia89, inplasma and breath condensates from asthmatics90and in the brains of patients suffering fromParkinson disease (PD), whereas increasedTBARS have been observed in plasma of pa-tients with amyotrophic lateral sclerosis (ALS) aswell as in Alzheimers patients91.

Measurement of Aldehydes Other Than MDA

A great number of various aldehydes are pro-duced during lipid peroxidation and they differgreatly in their biological activity and capacity tocause further damage. The different classes ofaldehydic peroxidation products in biologicalsamples can be quantified by a method devel-oped by Esterbauer and Cheeseman86. The alde-hydes are derivatised with dinitrophenylhy-drazine (DNPH), the various classes of differentpolarity (e.g. alkanals, hydroxyalkenals, alke-nals) separated by TLC (Thin Layer Chromatog-raphy) and the individual aldehydes then re-solved by HPLC with UV detection. An alterna-tive procedure involves HPLC separation of thefluorescent cyclohexanedione (CHD) derivativesof the aldehydes92. These techniques are general-ly extremely time-consuming and quite expen-sive and are unlikely at present to be used as rou-tine measures of lipid peroxidation. They are on-ly likely to be used where it is necessary to knowthe full range of aldehydes produced in a particu-lar condition.

Hydroalkenals, such as 4-hydroxynonenal(HNE), are probably the most important endproducts of the lipid peroxidation process interms of cytotoxicity. HNE is a major and toxicaldehyde produced by free radical attack on -6polyunsaturated fatty acids (arachidonic, linoleicand linolenic acids)93 and is considered a secondtoxic messenger of oxygen free radicals94,95. HNEundergoes many reactions with proteins, pep-tides, phospholipides and nucleic acids; it there-fore has a high biological activity and exhibits


various cytotoxic, mutagenic, genotoxic and sig-nal effects, including inhibition of protein andDNA synthesis, inactivation of enzymes, stimu-lation of phospholipase C, reduction of gap-junc-tion communication, stimulation of neutrophilchemotaxis, modulation of platelet aggregationand modulation of the expression of somegenes96. Additionally, HNE may be an importantmediator of oxidative stress-induced apoptosis,cellular proliferation and signaling pathways97.HNE is permanently formed at basal concentra-tions under physiologic conditions, but its pro-duction is greatly enhanced in pathologic condi-tions associated to lipid peroxidation. Underphysiologic conditions, the cellular concentra-tions of HNE ranges from 0.1 to 3 mol/L. Un-der conditions of oxidative stress, HNE concen-trations are significantly increased in plasma,some organs and cell types98. During heavystress, e.g., in patients with severe rheumatologicdiseases such as rheumatoid arthritis, systemicsclerosis, lupus erythematosus, chronic lym-phedema or chronic renal failure, serum HNE isincreased to concentrations up to 3- to 10-foldhigher than physiologic concentrations99. HNEand acrolein, compound present in some environ-mental sources like cigarette smoke, are highlyreactive toward proteins (particularly, HNE ismuch more reactive to proteins than to DNA),forming stable covalent adducts with His, Lysand Cys residues through Michael addition; theseadducts are known as advanced lipoxidation endproducts (ALEs)100,101. This process introducescarbonyl groups into proteins.

Furthermore, concentrations of acrolein- andHNE-protein adducts are increased in cardiovas-cular disease102. Acrolein reacts with Lysresidues of apolipoprotein A-I (apoA-I), the ma-jor protein of HDL, which plays a relevant rolein mobilizing cholesterol from artery wallmacrophages. Acrolein adducts colocalize withapoA-I in human atherosclerotic lesions. More-over, the capacity of acrolein-modified apoA-I toremove cholesterol from cultured cells is im-paired, suggesting that carbonylation might inter-fere with the normal function of apoA-I in pro-moting cholesterol removal from artery wallcells, thus playing a critical role in atherogene-sis103.

Increased concentrations of HNE-proteinadducts have been reported in the lungs of smok-ers with and without chronic obstructive pul-monary disease (COPD). Notably, HNE concen-trations in the pulmonary epithelium, airway en-

dothelium and, particularly, neutrophils of COPDpatients were found to be inversely related tolung function104. COPD patients also had higherdiaphragm concentrations of both protein car-bonyls and HNE-protein adducts. Furthermore, anegative correlation was found between carbonylgroups and airway obstruction (i.e., concentra-tions of reactive carbonyls related to diseaseseverity) and between HNE-protein adducts andrespiratory muscle strength (i.e., HNE-proteinadduct formation associated to respiratory mus-cle function)105.

Because HNE is such an important, biological-ly active product it may be of interest to measurespecifically. HNE can be measured by HPLCwith UV detection or GC-MS which is more sen-sitive but expensive.

IsoprostanesF2-Isoprostanes (F2-IsoPs), isoprostanes con-

taining an F-type prostane ring, are a family of,theoretically, 64 prostaglandin F2-like moleculesproduced in vivo, primarily in situ, by nonenzy-matic free-radical-catalyzed peroxidation of es-terified arachidonic acid and then cleaved and re-leased into the circulation by phospholipases(s)before excretion in the urine as free isoprostanes.Reports have shown that F2-IsoPs are authentic,reliable biomarkers of lipid peroxidation and areuseful in vivo indicators of oxidative stress invarious clinical conditions, such as acute andchronic inflammation, ischemia/reperfusion in-jury, diabetes and atherosclerosis106-110. F2-IsoPshave also been used to assess in vivo oxidativeresponse to some drugs, antioxidants or dietaryinterventions for their free-radical-scavengingproperties. Various techniques for F2-IsoP quan-tification are available (GC-MS)111. Additionally,to being markers of oxidative stress and antago-nists of the action of prostaglandins, they may al-so exert unique biological effects.

A tissue that does not contain isoprostanes isyet to be reported. Isoprostanes have also beenfound in measurable quantities in most of the bi-ological fluids analyzed, including plasma, urine,synovial fluid, bronchoalveolar fluid, bile,lymph, microdialysis fluid from various organs,and amniotic, pericardial, and seminal fluid, evenif plasma and urine are the sample types that arecommonly analyzed, being the most convenientto obtain and the least invasive112.

At present, measurement of F2-IsoPs is regard-ed as one of the most reliable approaches for theassessment of oxidative status or free-radical-

319


320

mediated lipid peroxidation in vivo. Available da-ta indicate that quantification of F2-isoprostanesin either plasma or urine gives a highly preciseand accurate index of oxidative stress113. Where-as the biological validity of F2-IsoPs as biomark-ers of oxidative status is well established, it istechnically quite complicated to measure F2-IsoPs and their metabolites in body fluids andsome limits with respect to their measurementmust be taken into account. F2-IsoPs are chemi-cally stable in vivo and ex vivo, but once theyproduced and released into the circulation, theyare fastly metabolized (even if not as quickly oras extensively as prostaglandins) and eliminated.Their fast disappearance from plasma may pre-vent practical application. Current techniques:GC-MS, GC-tandem MS (GC-MS/MS), liquidchromatography (LC)-MS, LC-MS/MS, enzymeimmunoassays and radioimmuno assays (RIAs),are able to detect the steady-state concentrationsof F2-IsoPs in many tissues and body fluids, evenin the basal state, concentrations after any degreeof oxidant stress or lipid peroxidation in vivo.Different internal standards (18O- or 2H-labeledanalogs of specific isoprostane isomers) areavailable from commercial sources to quantifythe isoprostanes by MS techniques.

Alternative methods have also been developedto quantify F2-IsoPs by immunologic techniques(RIAs and enzyme immunoassays) and a few im-munoassay reagent sets are commercially avail-able. A potential drawback of these techniques isthat limited information is currently available re-garding their precision and accuracy. Moreover,few data exist comparing F2-IsoP concentrationsmeasured by immunoassays with MS results.Futhermore, the sensitivity and/or specificity ofthese assays may vary substantially among man-ufacturers. However, even if MS techniques ofF2-IsoP quantification are considered the goldstandard, immunoassays have expanded re-search in this area because of their low cost andrelative ease of use. Additionally to commercialimmunoassays, some researchers have generatedpolyclonal antibodies and have developed assaysfor F2-IsoP114. It appears that there is good corre-lation between these techniques and MS.

Various analytical methods are available forthe analysis of isoprostanes, the most sensitive,highly specific and reliable technique being GCwith negative-ion chemical ionization (NICI)MS. For quantification of lipid peroxidation,measurements of F2-IsoP have a clear advantageover currently available techniques such as as-

says for MDA, TBARS, lipid hydroperoxides orconjugated dienes, which are hampered by somemethodologic limits.

F2-IsoPs are very well suited as biomarkers ofoxidative stress for the following reasons:

1. The in vivo formation of isoprostanes increas-es as a function of oxidative stress115,116.

2. They can be measured accurately down to pi-comolar concentrations with analytical meth-ods such as GC-MS, GC-MS/MS, LC-MS,LC-MS/MS or RIA. The first 4 methods caneasily differentiate among the various types ofisoprostanes, but they require extensive prepa-ration of the material (e.g., phospholipid ex-traction and alkaline hydrolysis) and/or expen-sive instrumentation. RIAs are somewhat easi-er to perform and are widely available com-mercially. However, many of these are notable (or have not been shown to be able) todistinguish between the prostanoids and theisoprostanes, much less between the differenttypes of isoprostanes.

3. They are stable in isolated samples of bodyfluids, including urine and exhaled breath con-densates, providing an exceedingly noninva-sive route for their measurement.

4. Their measured values do not exhibit diurnalvariations and are not affected by lipid contentin the diet117,118. However, they do varymarkedly in clinical and experimental condi-tions characterized by oxidative stress andclosely parallel disease severity. Some diurnalvariation in urinary F2-IsoP excretion does oc-cur within individual humans, even if thisvariation is not present when F2-IsoPs areevaluated on a group level. Furthermore, evenif pooled urine samples are likely preferable,F2-IsoPs determined in urine collected in themorning or in several spot urine samples ade-quately represent the daily F2-IsoP excretion.

5. They are specific products of peroxidation.6. They are present in detectable amounts in all

healthy tissues and biological fluids, thus al-lowing definition of a reference interval.

Because of free-radical-catalyzed conversionof arachidonic acid to isoprostanes, precautionsmust be taken to avoid artifactual formation dur-ing sample storage and processing. Blood plasmasamples contain considerable amounts of arachi-donic acid, mainly esterified to membrane phos-pholipids. Storage of these samples at 80C andaddition of antioxidants (e.g., butylated hydroxy-


toluene and triphenylphosphine) during samplepreparation is therefore recommended. More-over, isoprostanes in blood samples may occur asfree fatty acids or esterified to phospholipids orlipoproteins. Thus, one has to distinguish be-tween the two fractions of isoprostanes in humanblood, i.e., free and total (free plus esterified).Analysis of the esterified molecules requires hy-drolysis to yield the free derivatives. Becauseurine samples have a very low lipid content, au-tooxidation is not a problem. Nevertheless, as aprecaution, samples should be supplementedwith EDTA and 4-hydroxy-2,2,6,6-tetram-ethylpiperidine 1-oyl (4-hydroxy-TEMPO) andstored at 20C.

Different diseases and experimental conditionshave been shown to be related to marked increas-es in urinary, plasma and tissue concentrations ofF2-IsoPs. However, it has been suggested theyshould be considered not just mere markers, butalso mediators of disease, as they evoke impor-tant biological responses in virtually every celltype found within the lung. Infact, the iso-prostanes may mediate many of the features ofthe disease states for which they are used as indi-cators. 8-iso-prostaglandinF2, the biologicallyactive component, is produced in great amount inotherwise normal individuals exposed to ciga-rette smoke, allergenes, ozone or hyperoxia andduring ventilated ischemia. It is also markedlyincreased, serving as a biomarker, in the bron-choalveolar lavage (BAL) fluid, plasma, urine orexhaled breath condensate (a noninvasive tech-nique for direct measurement of oxidative stressin the lungs) in some pulmonary diseases such asasthma, COPD, interstitial lung disease, cystic fi-brosis, pulmonary hypertension, acute chest syn-drome, sickle cell disease, acute lung injury (in-cluding acute respiratory distress syndrome,ARDS) and severe respiratory failure in infantsas well as in healthy chronic smokers119,120. Sys-temic and synovial fluid concentrations of 8-iso-PGF2 are higher in patients with rheumatoidarthritis, psoriatic arthritis, reactive arthritis andosteoarthritis than in healthy controls. Plasmaconcentrations are increased in patients with car-diovascular disease and it has been suggestedthat this may be a useful biomarker of risk121.Similarly, some cardiovascular conditions featuremarked increases in F2-IsoP concentrations, in-cluding during and after cardiopulmonary by-pass, renal, cerebral and myocardial ischemia-reperfusion injury, unstable angina, heart failure,coronary heart disease, acute ischemic stroke,

hypercholesterolemia and atherosclerosis. Uri-nary 8-iso-PGF2, measured by GC-MS/MS, wasfound to be a novel, sensitive and independentrisk marker in patients with coronary heart dis-ease, additionally to know risk factor of thispathology, i.e., diabetes mellitus, hypercholes-terolemia, hypertension, obesity and smoking122.Increased concentrations of 8-iso-PGF2 have al-so been found in plasma or urine samples frompatients with type 2 diabetes.

Lipid peroxidation in Human Material:Past and Future

Despite the problems that can occur with as-says such as the TBA test, diene conjugation andlight emission, they usually work adequatelywhen applied to measurements of peroxidation inliposomes, microsomes or other isolated mem-brane fractions, provided that one is alert to thevarious artefacts that can arise123. Much more se-rious problems occur when these assays are ap-plied to human body fluids or to tissue extracts.

By far the most misleading assay to use on hu-man material, especially plasma, is the TBA test.Plasma contains many compounds that react inthe TBA assay, including bile pigments, aminoacids and carbohydrates. Some of these sub-stances (e.g. bile salts) produce a different chro-mogen, and this interference can be overcome byseparating out the authentic (TBA)2-MDA adduct(e.g. by HPLC) before measurement. However,this solves only part of the problem becausesome molecules (especially amino acids and sug-ars) react in the assay to form an authentic(TBA)2-MDA adduct. The lack of specificity ofthe TBA assay when applied to plasma is dra-matically illustred by a simple experiment per-formed by Lands et al124. Using the cyclooxyge-nase assay these researchers measured the lipidperoxide amount of some human plasma samplesas about 0.5 M. Expressing the results of a TBAtest on the same samples as peroxide equiva-lentsgave a value of 38 M. When specificchemical assays are used, the authors and oth-ers125 find that human plasma, freshly taken fromhealthy subjects, has less than 0.1 M lipid per-oxide. This is perhaps not surprising, since evenif peroxides do form in vivo and enter the circu-lation, they can be quickly cleared. For example,although lipid peroxidation is thought to be rele-vant in atherosclerosis, it seems to be peroxida-tion in the arterial wall that matters, not peroxi-dation in the bulk plasma. Thus, some of the ear-lier suggestions that circulating lipid peroxides

321


322

kill vascular endothelial cells and initiate athero-sclerosis need to be re-evaluated.

In order to know as much as possible about thereal occurrence of lipid peroxidation in humanmaterial, it is important to use methods that givespecific chemical information about what is pre-sent. Indeed, food scientists have followed thisprinciple for years. Thus various groups are sepa-rating the different peroxidation products beforemeasuring them. This is often done by HPLC;for example, HPLC methods for measuring cyto-toxic aldehydes are available126. However, con-version of material into volatile derivatives, sep-aration by gas chromatography and identificationby mass spectrometry is likely to give more pre-cise chemical information when complex mix-tures are being studied127,128. Thus, derivatizationand mass spectrometry have been used to charac-terize peroxidized fatty acids and cholesterol oxi-dation products in human atherosclerotic le-sions129.

Specificity can also be achieved by the use ofantibody techniques, particularly monoclonal an-tibodies. Thus, antibodies directed against low-density lipoprotein that has undergone peroxida-tion or has been treated with 4-hydroxynonenalbind to rabbit atherosclerotic lesions. Additional-ly, low-density lipoproteins eluted from such le-sions can bind to antibody specific for MDA-treated low-density lipoproteins. Antibody-basedmethods can also be applied to plasma samples.Using such specific techniques, the exact roleplayed by lipid peroxidation in cell injury anddeath mediated by toxins and in human diseaseshould at last become clearer130.

Measurement of Protein Damage

Reactive free radicals can modify amino acidresidues of proteins and lead to cross-linking,changes in conformation and loss of function.Oxidatively damaged proteins are likely to be re-moved rapidly by proteases rather than accumu-late to readily-detectable levels.

Glutathione and S-Glutathionylated Proteins

Because blood glutathione concentrations mayreflect glutathione status in other, less accessibletissues, measurement of both reduced glutathione(GSH) and glutathione disulfide (GSSG) inblood has been considered relevant as an index

of whole-body GSH status and a useful indicatorof oxidative stress status in humans. Differenttechniques have been optimized to identify andquantify glutathione forms in human samples, in-cluding spectrophotometric, fluorometric and bi-oluminometric assays, often applied to HPLCanalysis, as well as the more recently developedGC-MS and HPLC-electrospray ionization-MStechniques131,132. Futhermore, a wide variety oftechniques have been introduced for the determi-nation of GSH and GSSG in human blood, themeasurement of which, particularly that ofGSSG, could be overstimated if samples are notproperly processed133,134. A specific warning hasto be addressed to correct sample manipulationand prevention of artifactual GSH oxidation. Theauthors have shown that the main artifact resultsfrom sample acidification (for protein separation)without prevention of artificial oxidation of SHgroups by blocking with alkylating agents. Actu-ally, many published articles reporting concentra-tions of GSH, GSSG and S-glutathionylated pro-teins in blood, both from healthy controls and pa-tients affected by various pathologies, are not ar-tifact free, which makes the conclusions reachedin these articles meaningless. Consequently, thenotion that some pathophysiologic conditionscan alter and/or be influenced by the GSH/GSSGhomeostasis of blood still needs to be confirmed.

It is well known that a decrease in GSH con-centration may be associated with ageing135 andthe pathogenesis of many diseases, includingrheumatoid arthritis, amyotrophic lateral sclero-sis, acquired immune deficiency syndrome,Alzheimers disease, alcoholic liver disease,cataract genesis, respiratory distress syndrome,cardiovascular disease and Werner syndrome.Furthermore, there is a drastic depletion in cyto-plasmic concentrations of GSH within the sub-stantia nigra of Parkinsons disease patients136.Depletion of total GSH (GSH + 2 GSSG + pro-tein-bound glutathione) and a decreasedGSH:GSSG ratio are indicators of oxidative/ni-trosative stress in ischemic brain disease137, car-diovascular diseases138 and cancer139, and de-creased concentrations of GSH are consistentlyobserved in both types of diabetes mellitus. LowGSH concentrations and a high GSSG:GSH ratiohave been measured in blood of patients withvarious diseases, including breast and lung can-cer, coronary heart surgery and preeclampsia140.The GSH system is also altered in lung inflam-matory conditions. For example, GSH concentra-tions are increased in the epithelial lining fluid of


chronic smokers, whereas they decrease fastly inpatients with mild asthma during an asthma ex-acerbation. Similarly, GSH concentrations in theepithelial lining fluid are decreased in idiopathicpulmonary fibrosis, asbestosis, acute respiratorydistress syndrome and in HIV-positivepatients141. Total GSH was markedly decreasedin older patients with chronic diseases142, thedeficit being attributable to lower GSH concen-trations and not to higher GSSG. These resultssuggested that the decrease in GSH might beused to monitor the severity and progress of thediseases. Conversely, total GSH concentrationsare high in the blood of elderly persons who arein excellent physical and mental breath143.

Tyrosine Oxidation,Nitration and Halogenation

The toxicity of NO is enhanced by its reactionwith a superoxide to form ONOO144. It or sec-ondary metabolites can cause tyrosine nitrationin protein, creating nitrotyrosine, a footprint de-tectable in vivo.

Analysis of 3-nitrotyrosine (NO2-Tyr), a stablemarker for NO derived oxidants, and halo-genated Tyr products such as 3-chlorotyrosine(Cl-Tyr) or 3-bromotyrosine has been performedin some diseases and different techniques havebeen developed for such measurements145-149. Thequantitative measurement of NO2-Tyr is hinderedby severe methodologic problems. The most ofthe data available on NO2-Tyr in tissues and flu-ids have been derived from antibody-based tech-niques, which however, are often not rigorouslyvalidated. Therefore, such immunologic tech-niques should be considered semiquantitativeand the results interpreted accordingly. HPLCwith ultraviolet detection does not provide ade-quate sensitivity or specificity for biological ma-terials. In contrast, HPLC with electrochemicaldetection, LC-MS/MS, electron capture-negativechemical ionization (EC-NCI) GC-MS and GC-MS/MS are able to quantify NO2-Tyr in biologi-cal materials and human plasma150-152.

At present, only MS/MS-based techniques,both GC-MS/MS and LC-MS, provide reliablevalues for circulating and excreted NO2-Tyr, withLC-MS/MS being at present considerably lesssensitive than GC-MS/MS and that the basalconcentrations obtained by this analytical ap-proach may serve as reference values.

Another methodologic problem is consider-able interference by coeluting molecules, whichcan eliminated only by use of MS/MS.

Increased concentrations of stable halogenatedTyr residues have been detected in proteins iso-lated from atherosclerotic plaques as well as inplasma and airway secretions of patients withasthma, ARDS and cystic fibrosis, and halo-genated Tyr residues are widely used as markersfor damage mediated by hypohalous acids (HO-CL and HOBr) in these diseases153-156. The majorproducts are Cl-Tyr and 3-bromotyrosine, but di-halogenated compounds (3,5-dichlorotyrosineand 3,5-dibromotyrosine) are formed with highexcesses of HOCL and HOBr. Dramatic selectiveenrichment in protein-bound NO2-Tyr and Cl-Tyramount within ApoA-I, the major protein con-stituent within HDL, recovered from humanplasma and atherosclerotic lesions has beendemonstrated by proteomic and MS methods.Analysis of serum also showed that protein-bound NO2-Tyr and Cl-Tyr concentrations inApoA-I are markedly higher in individuals withestablished coronary heart disease157,158. These re-sults suggest that increased concentrations of Cl-Tyr and NO2-Tyr in circulating HDL might rep-resent specific markers for clinically significantatherosclerosis.

Increased concentrations of nitrated plasmaproteins have been associated with predispositionto develop lung injury in premature infants aswell as with unfavorable outcome on develop-ment of lung injury159. The clinical relevance ofprotein Tyr nitration has been emphasized by theobservation of a strong association between pro-tein bound NO2-Tyr concentrations and coronaryartery disease risk. Circulating concentrations ofprotein-bound NO2-Tyr may serve as an indepen-dent biomarker to assess atherosclerosis risk,burden and incident cardiac events, as well as tomonitor the vasculoprotective action of drugssuch as statins (hydroxymethylglutaryl-CoA re-ductase inhibitors)160.

Patients with lung cancer have significantlyhigher serum concentrations of nitrated proteins,supporting the presence of oxidative and ni-trosative stress161,162. Specific locations and tar-gets of Tyr nitration in lung cancer have, recent-ly, been detailed163. Increased nitrotyrosine im-munostaining is limited mainly to the tumor andnot to surrounding healthy tissue or is weakly re-active in different regions of the lung from thesame patients with cancer, suggesting a uniqueenviroment inside the tumor that may contributeto the disease process. This was noted in squa-mous cell carcinoma as in the well-differentiatedadenocarcinoma. Using proteomic and genomic

323


324

approaches, authors have identified the proteintargets. Most of the nitrated proteins fall into 4categories: oxidant defense (such as manganesesuperoxide dismutase and carbonic anhydrase),energy production (many glycolitic enzymes),structure (such as -actin, - and -tubulin andvimentin) and those involved in apoptosis (an-nexins).

Tyrosine nitration is one of the earliest mark-ers found in brains from persons affected byAlzheimers disease, in the plaques of brainsfrom persons with multiple sclerosis and in de-generating upper and lower motor neurons inALS patients164. Nitrated -synuclein selectivelyaccumulates in Lewy bodies and protein inclu-sions in many pathologies (Alzheimers disease,Parkinsons disease, synucleinopathies andtauopathies). Nitrated proteins have been evi-denced in some inflammatory disease, chronicrenal failure, rheumatoid arthritis, type 1 andtype 2 diabetes and cystic fybrosis. On the otherhand, basal protein nitration has been detectedunder physiologic conditions in most tissues, in-cluding plasma and the human pituitary andsome of the nitrated proteins have been identi-fied. Two-dimensional Western blotting and LC-MS/MS analyses have been used to detect andcharacterize 4 nitrated proteins, including actin,in the healthy human pituitary, which partecipatein neurotransmission, cellular immunity, and cel-lular structure and motility165. These results areconsistent with the emerging perspective thatlow-level Tyr nitration may be a physiologic reg-ulator of a signaling pathway166.

Carbonylated ProteinsProtein carbonyls may be produced by the oxi-

dation of some amino acid side chains (e.g., inLys, Arg, Pro and Thr); by the formation ofMichael adducts between Lys, His and Cysresidues and ,-unsaturated aldehydes, formingALEs (Advanced Lipooxidation End Products);and by glycation/glycoxidation of Lys aminogroups, forming advanced glycation end AGEproducts167-169. The generation of carbonyl mole-cules is the most general and widely used markerof severe protein oxidation both in vitro and invivo, with different assays developed for thequantification of these species (170). The chemi-cal stability of protein carbonyls makes themsuitable targets for laboratory measurements andis also useful for their storage: their stability dur-ing storage for 10 years at 80C has beendemonstrated.

Assays of general oxidative damage to pro-teins, while an important index of oxidativestress occuring in vivo, need to be replaced withassays of oxidative damage to specific proteinshaving relevance to the lesion under considera-tion.

As a marker of oxidative damage to proteins,carbonyls have been shown to accumulate duringaging, ischemia/reperfusion injury, chronic in-flammation, cystic fibrosis and many of age-re-lated diseases in some organisms171.

Specific carbonylated proteins have been de-tected in both the brain tissue and plasma ofAlzheimers disease patients172. The observationof carbonylated proteins in plasma suggests thatthese oxidized species may be useful as diagnos-tic biomarkers for (possibly early) Alzheimersdisease.

The carbonyl content in plasma proteins(mainly albumin and -globulins) from childrenwith different forms of juvenile chronic arthri-tis was significantly higher than in healthy chil-dren, and more importantly, the carbonyls in-creased in parallel with the activity of the dis-ease.

Correlation between the carbonyl concentra-tion and the activity or the type of chronic juve-nile arthritis indicates that plasma protein car-bonyl groups are a good marker of inflammatoryprocess activity and may allow the use of car-bonyls as a clinical marker of antioxidant barrierimpairment in this group of patients as well asfor monitoring possible pharmacologic treat-ments173.

Plasma concentrations of protein carbonyls, aswell as free F2-IsoPs and protein reduced thiols,differ significantly between chronic kidney dis-ease patients and healthy people. Furthermore,such biomarkers of oxidative/nitrosative stressare significantly higher in patients with diabetesand hypercholesterolemia174.

Winterbourn et al175 determined that proteincarbonyl concentrations were increased in bothplasma and BAL (bronchoalveolar lavage) fluidof patients with severe sepsis or major trauma,which correlated well with measured concentra-tions of ALEs and with indices of neutrophiliaand neutrophil activation. Moreover, patientswith acute pancreatitis had significantly in-creased plasma concentrations of protein car-bonyls, which were related to disease severity,thus confirming that this protein modificationcould be a useful plasma marker of oxidativedamage.


Measurement of DNA Damage

Cellular DNA damage can be caused by ROSproduced under several conditions and differentmethods have been developed to measure theoxidatively modified nucleobases in DNA176,177.Oxidative DNA damage seems to relate to an in-creased risk of cancer development later inlife178. DNA subjected to attack by hydroxyl rad-ical produces a wide range of base and sugarmodification products. Amongst these, the majorreaction product of .OH with thymine is thymineglycol and with guanine, 8-hydroxy-guanine.These DNA products are eliminated by repairenzymes (excision enzyme and glycolases) andare excreted in the urine either as the free baseproducts or as the nucleoside derivatives, thymi-dine glycol (Tg) and 8-hydroxydeoxyguanosine.The latter products can be used as an index ofradical attack upon DNA in vivo and Cathcart179haved calculated from such measurements thatoxidative damage to mammalian DNA may totalabout 105 oxidative hits per cell per day179.Measurement of urine samples is based on chro-matographic pre-purification by normal, reverse-phase or immuno-affinity columns to prevent in-terference by many urinary compounds, fol-lowed by derivatization and analysis. Originally,HPLC combined with UV detection was used tomeasure Tg31 but great concentrations of sam-ples were required to obtain sufficient sensitivi-ty. Most current procedures are based on HPLC,GC-MS, LC-MS and antibody-based methods180.The advantages of artifacts produced duringmeasurement of 8OHdG are useful for visualiza-tion of damage, but they seem likely to be onlysemiquantitative.

DNA can also be damaged by RNS, undergo-ing mainly nitration and deamination of purines.Techniques for the measurement of DNA basenitration and deamination products have been de-veloped but may need more refinement and vali-dation before they can routinely applied to hu-man materials.

None of the analytical techniques mentionedabove identifies where the oxidative damage islocated. Another problem in studying damageto DNA by ROS/RNS is the limited availabilityof human tissues from which to obtain DNA.Most studies are performed on DNA isolatedfrom lymphocytes or total leukocytes from hu-man blood and it is assumed (possibly erro-neously) that changes here are reflected in oth-er tissues.

8-hydroxydeoxyguanosine (8-OHdG), an oxi-dized form of guanine, is a major oxidativeDNA-damage product that can produce mutation.This compound causes A:T to C:C or G:C to T:Atransversion mutations because of its base pair-ing with adenine as well as cytosine. Measure-ment of 8-OHdG in urine has been used to assesswhole-body oxidative DNA damage. This canbe achieved by HPLC and MS methods. Howev-er, 8-OHdG can arise from degradation of oxi-dized dGTP in the DNA precursor pool, not justfrom removal of oxidized guanine residues fromDNA by repair processes. Furthermore, there aremany other products of oxidative DNA damage.Hence, urinary 8-OHdG is a partial measure ofdamage to guanine residues in DNA and its nu-cleotide precursor pool, and 8-OHdG concentra-tions may not truly reflect rates of oxidativedamage to DNA.

Papa et al181 wanted to assess the production ofROS and 8-OHdG in gastric mucosa, according toH. pylori status and cytotoxic associated geneproduct A (CagA) and to determine the relation-ship between ROS production and amount of 8-OHdG. Gastric biopsy specimens were obtainedfrom 60 consecutive patients. ROS generation wasmeasured by luminol enhanced chemilumines-cence. 8-OHdG detection was performed by animmunoperoxidase method, using a specific anti8-OHdG monoclonal antibody. 40/60 patients(67%) were H. pylori-positive. ROS generationwas significantly higher in patients positive for H.pylori infection as compared to negative. 8-OHdGdetection was performed in 30 patients in whichCagA presence was also investigated. High ex-pression of 8-OHdG was detected in 14/20 (70%)H. pylori-positive patients (13 CagA+ and 1CagA) and in 2/10 (20%) H. pylori-negative pa-tients. A significant correlation was found be-tween ROS production and 8-OHdG content.

However, the recently completed Biomarkersof Oxidative Stress Study (BOSS), using acuteCCl4 poisoning in rodents as a model for oxida-tive stress, has demonstrated that 8-OHdG inurine is a potential candidate general biomarkerof oxidative stress, whereas neither leukocyteDNA-MDA adducts nor DNA-strand breaks re-sulted from CCl4 treatment.

Immunohistochemical accumulation of highlevels of 8-OHdG was reported to occur in vari-ous human tumors, like high-grade glioma, com-pared to adiacent, normal tissue or low-gradeglioma182. These studies suggested that oxidativestress may play a role in tumor progression.

325


326

As with other indices of whole-body oxidativestress, the measurement of products of oxidativeDNA damage is limited by some problems, in-cluding the obscurity of the tissue of origin of theproducts.

Another technique utilized to detect DNAadducts is the comet assay183. Other methods ex-ist to determine single- and double-strandbreaks184. Different oxidized adducts of DNA canbe determined. Examples are DNA-aldehydeadducts, such as deoxyguanosine-malondialde-hyde adducts185, or the end product of the reac-tion between DNA and 4-hydroxynonenal, thealdehyde formed following exposure to ROS186to produce N2-ethenodeoxyguanosine.

Measurement of Antioxidants

Different animal studies have shown that an-tioxidants delay or protect against the oxidativedamage produced by free radical reactions. Radi-cal-scavenging antioxidants are consumed duringthis process and antioxidant status is sometimesused indirectly to assess free radical activity. Thecommonly used TRAP assay (Total [peroxyl]Radical-trapping Antioxidant Parameter) is, in itbasic form, an empirical measurement of antioxi-dant activity in plasma187. Assessment of the rela-tive contribution of individual antioxidants(ascorbate, urate, -tocopherol, protein sul-phydryls) to the total antioxidant capacity re-quires separate specific assays. Measurements ofeither TRAP or the individual antioxidants arenot likely to be useful indices of free radical gen-eration as the latter would need to be extensivefor the steady-state concentrations of the antioxi-dants to be disturbed in vivo. However, antioxi-dant amounts are interesting parameters in them-selves, indicative of the propensity of the indi-vidual to oxidative stress.

Many methods exist for evaluating the activityand composition of the anti-oxidant enzymes,which, along with the LMWA, constitute the twomajor components of the anti-oxidative system.Some techniques that directly evaluate enzymaticactivity utilize spectroscopic measurements orgel-activity procedures; other methods employimmunocytochemistry. Assays of anti-oxidantenzymes may indicate prior exposure of the cellto oxidative stress, even if these enzymes are un-der regulation, and one might detect an increase,rather than a decrease, in their activity. ROS may

serve as a stimulating species for induction of an-tioxidant enzymes on the one hand and, on theother, may themselves damage the proteins; forexample, O2 might inactivate catalase. Determi-nation of the fate of LMWA may serve as a betterindicator for ROS, because the adduct is specificto these molecules.

Determination of the ratio between oxidantand reductant (e.g., ascorbate/dehydroascorbicacid or GSH/GSSG) may therefore serve as in-dicator of oxidative damage. One of the ap-proaches most commonly used is the measure-ment of the total anti-oxidant activity of a bio-logical site. Depletion of one anti-oxidant mol-ecules causes changes in the level of overallanti-oxidant molecules and may be evaluatedusing a variety of techniques including bio-chemical, immunohistochemical, spectroscopi-cal and electrochemical188.

The total-antioxidant-activity assay offersmany advantages and is considered a usefultool for detecting oxidative stress phenomenain bodily fluids and tissues. It may serve as anappropriate tool for the evaluation of anti-oxi-dant therapy. Determinations of total LMWArather than individual anti-oxidants are impor-tant, because LMWAs work in concert189, andmeasurement of only one or a few compoundsout of many present at a specific biological lo-cation might be misleading. Moreover, mea-surement of the total LMWA ensures a reliablepicture of the physiological situation. Now itsdont know the concentration of a specificcompound at a specific location at a given mo-ment. Sometimes the researchers try to detectcompounds that are not present in the site un-der investigation.

A few dozen LMWA exist, and usually only afew of them, such as vitamin E and ascorbic anduric acids, are revealed; thus, many compoundsthat can be present at the biological site are ne-glected. The measurement of the total LMWA isdesigned to overcome these problems. Numer-ous procedures allow measurement of the totalLMWA activity. These include indirect and di-rect methods for measuring total anti-oxidantactivity originating from the LMWA. Indirectmethods are those that measure consequentialfactors of redox capacity, such as oxidationproducts formed or concentrations of major re-dox couples in the biological enviroment, byfluorescent or spectrophotometric techniques. Inthis approach one assumes that a biological re-dox buffer exists in the form of a redox couple


that is sensitive to changes in the redox enviro-ment. Thus, it reflects changes in the reducingpower of the measured sample, which is in cor-relation with all of the LMWAs. Other indirecttechniques are inibition methods that involveadding a radical species to the sample togetherwith a scavenger that can be monitored withlaboratory instruments. The LMWAs present inthe sample under investigation can quench theradical and, therefore, interfere in its reactionwith the added scavenger. Examples of indirectmethods are:

1. measurement of electrochemical couples, suchas GSH/GSSG (glutathione/oxidized glu-tathione190);

2. NADH/NAD+ (reduced nicotinamide dinu-cleotide/nicotinamide dinucleotide), andascorbic acid/ascorbate191,192;

3. the Trolox equivalent-antioxidant capacity(TEAC) assay193;

4. the total radical-trapping potential (TRAP), anassay to define, for example, the stage of ath-erosclerosis194,195;

5. the chemiluminescence method for superoxidedetection;

6. the oxygen-radical absorbance capacity(ORAC) methodology.

Direct methods for measuring total LMWA arethose that utilize an external probe to measurethe reducing or oxidizing capacity of a system.An example is an electrode, in which the currentis proportional to the concentrations of the scav-enger or the redox couple under investigation.These direct methods can be classified into 2groups: chemical and electrochemical. Thechemical methods measure a known redox activecouple whose reduced and measured as a func-tion of concentration. For example, the ferric-re-ducing antioxidant power (FRAP) assay is basedon the reaction of the redox couple ferric/ferrouswith anti-oxidants in the sample and results inthe creation of a blue color that can be measuredat 593 nm196.

Several methods have been developed toassess the total antioxidant capacity (TAC);the molecules measured using these assays arereducants, able to reduce oxidant species andprotect oxidizable compounds197,198. However,the number of different anti-oxidants in serumor other biological samples makes it difficultto measure each element separately. Addition-ally, the possible interaction among different

anti-oxidants in vivo could make the measure-ment of any individual anti-oxidant unrepre-sentative of the overall anti-oxidant status.Moreover, because the measured TAC of a bi-ological samples depends on which procedureis used in the measurement199, the comparisonof different analytical methods represents acrucial factor in helping researchers to chooseand to understand the results obtained using aspecific method, due to the different princi-ples on which they are based200. For example,studies evaluating the single contribution ofpure plasma component to the total anti-oxi-dant activities of blood samples, indicate thatthe main contribution in the FRAP assay, butnot in others, is the acid uric. On the contrary,the anti-oxidant capacity of reduced glu-tathione is not detected by using the FRAP as-say but significantly contributes to the anti-oxidant capacity measured by utilizing othertests. Recent data showed differences betweenthe FRAP assay and other methods in TACmeasured in the same sample from normal in-dividuals. Thus, each method is sensitive tovarious anti-oxidants in a different mannerand consequently may also evidence, as com-pared to another method, different levels ofanti-oxidant capacity in the same sample.Moreover, the interaction between anti-oxi-dant components may complicate the evalua-tion of in vivo results.

Horoz et al201 aimed to measure total anti-oxidant response (TAR) using a novel auto-mated method in nonalcoholic steatohepatitis(NASH) subjects. As a reciprocal measure,they also aimed to determine total peroxidelevels in the same plasma samples. The ratiopercentage of the total plasma peroxide levelto the plasma TAR value was regarded as ox-idative stress index (OSI)202. Twenty-two sub-jects with biopsy proven NASH (19 male, 3 fe-male; mean age 37.7 8.8) and 22 healthycontrols (17 male, 5 female; mean age 34.6 9.3) were enrolled. The most important indica-tions for liver biopsy in those 22 subjects wereultrasonographically diagnosed fatty liver andelevation in alanine aminotransferase (ALT).The total anti-oxidant status of the plasma wasmeasured using a novel automated colorimet-ric measurement method for TAR developedby Erel203. In this method, the hydroxyl radical,the most potent biological radical, is producedby the Fenton reaction, and reacts with thecolourless substrate O-dianisidine to produce

327


328

the dianisyl radical, which is brigth yellowish-brown in colour. Upon the addition of a plasmasample, the oxidative reactions initiated by thehydroxyl radicals present in the reaction mixare suppressed by the anti-oxidant componentsof the plasma, preventing the colour changeand thereby providing an effective measure ofthe total anti-oxidant capacity of the plasma.The assay results are expressed as mmolTrolox eq./L, and the precision of this assay isexcellent, being lower than 3%. The total plas-ma peroxide concentrations were determinedusing the FOX2 method80 with minor modifi-cations. The FOX2 test sistem is based on theoxidation of ferrous iron to ferric iron by thevarious types of peroxides contained in theplasma samples, in the presence of xylenol or-ange which produces a coloured ferric-xylenolorange complex whose absorbance can bemeasured. Total antioxidant response of sub-jects with NASH was significantly lower thancontrols (p < 0.05), while mean total peroxidelevel and mean oxidative stress index werehigher (all p < 0.05). In subjects with NASH,fibrosis s

Date post:	05-Mar-2016
Category:	Documents
Upload:	dian-wijayanti
View:	230 times
Download:	0 times

1oksidative Stress Test

Documents