Forum: Oxidative Stress Status - CNCcnc.cj.uc.pt/BEB/private/pdfs/ToxNitr/SM 3.pdfThe biological...

Forum: Oxidative Stress Status

UNRAVELING PEROXYNITRITE FORMATION IN BIOLOGICAL SYSTEMS

RAFAEL RADI,* GONZALO PELUFFO,* M ARı́A NOEL ALVAREZ,* M ERCEDESNAVILIAT ,† and ALFONSO CAYOTA‡

Departamentos de *Bioquı´mica, †Reumatologı´a, y ‡Medicina, Facultad de Medicina, Universidad de la Repu´blica, Montevideo,Uruguay

(Received2 May 2000;Revised29 June2000;Accepted29 June2000)

Abstract—Peroxynitrite promotes oxidative damage and is implicated in the pathophysiology of various diseases thatinvolve accelerated rates of nitric oxide and superoxide formation. The unambiguous detection of peroxynitrite inbiological systems is, however, difficult due to the combination of a short biological half-life, limited diffusion, multipletarget molecule reactions, and participation of alternative oxidation/nitration pathways. In this review, we provide theconceptual framework and a comprehensive analysis of the current experimental strategies that can serve to unequiv-ocally define the existence and quantitation of peroxynitrite in biological systems of different levels of organization andcomplexity. © 2001 Elsevier Science Inc.

Keywords—Nitric oxide, Superoxide, Peroxynitrite, Hydroxyl radical, Nitrogen dioxide, 3-nitrotyrosine, Free radicals,Detection, Methods

INTRODUCTION

Peroxynitrite,1 the product of the combination reactionbetween nitric oxide (zNO) and superoxide (O2

•2), is a

reactive and short-lived species that promotes oxidativemolecular and tissue damage [1–7]. In addition to thegeneration of a pro-oxidant species, the formation ofperoxynitrite results in decreased bioavailability of•NO,therefore diminishing both its salutary physiologicalfunctions [8–10] and its strong antioxidant actions overfree radical and metal-mediated processes [11–13]. Per-oxynitrite formation and reactions are proposed to con-tribute to the pathogenesis of a series of diseases includ-ing acute and chronic inflammatory processes, sepsis,ischemia-reperfusion, and neurodegenerative disorders,among others [14–27].

The detection of peroxynitrite in biological systemshas been a challenge over the past decade because of the(i) elusive nature of peroxynitrite which precludes itsdirect isolation and detection, (ii) necessity to find de-tector molecules that can efficiently outcompete the mul-tiple reactions that peroxynitrite can undergo, (iii) non-existence of footprints totally specific of peroxynitritereactions, and (iv) the difficulty to discriminate between

Rafael Radi, M.D., Ph.D., obtained his doctoral degree at the Uni-versidad de la Repu´blica, Montevideo, Uruguay in 1989 and performedposdoctoral studies at the University of Alabama at Birmingham(1989–1991). He returned to Uruguay in 1992 to a faculty position atthe Departamento de Bioquı´mica, Facultad de Medicina, Universidadde la Repu´blica, where he initiated a research group that investigatesthe biochemistry and cell biology of nitric oxide and peroxynitrite. Heis at present a Professor of Biochemistry, an International ResearchScholar of the Howard Hughes Medical Institute, and the currentSecretary General of the Oxygen Society.

Gonzalo Peluffo, M.D., and Marı´a Noel Alvarez, M.S., obtainedtheir degrees at the Universidad de la Repu´blica in 1998 and arecurrently performing Ph.D. studies on aspects referred to the biologicalformation, detection, and diffusion of oxygen radicals, nitric oxide, andperoxynitrite in biochemical and cell systems (MNA) or humans (GP).They are both Assistant Professors of Biochemistry at the Facultad deMedicina, Universidad de la Repu´blica.

Mercedes Naviliat, M.D., obtained her degree at the Universidad dela República in 1985 and has just completed a Ph.D. thesis on the roleof nitric oxide and peroxynitrite in inflammatory disease. She is anAssistant Professor of Rheumatology at the Facultad de Medicina,Universidad de la Repu´blica.

Alfonso Cayota, M.D., Ph.D., obtained his M.D. degree at Universidadde la Repu´blica in 1986 and performed Ph.D. studies at the Universite´Paris VI-Pasteur Institute, Paris, France from 1990–1995. He is currentlyan Associate Professor of Medicine at the Facultad de Medicina, Univer-sidad de la Repu´blica, where he investigates the role of reactive oxygenand nitrogen species during normal and pathological immune responses.

Address correspondence to: Dr. Rafael Radi, Departamento de Bio-quı́mica, Facultad de Medicina, Avda. General Flores 2125, 11800

Montevideo, Uruguay; Fax: 1598 (2) 9249563; E-Mail:[email protected] term peroxynitrite refers to the sum of peroxynitrite anion(ONOO2) and peroxynitrous acid (ONOOH). IUPAC recommendednames for peroxynitrite anion, peroxynitrous acid, nitroso-peroxocar-boxylate (ONOOCO2

2), and nitric oxide are oxoperoxynitrate (12),hydrogen oxoperoxynitrate, 1-carboxylato-2-nitrosodioxidane, and ni-trogen monoxide, respectively.

Free Radical Biology & Medicine, Vol. 30, No. 5, pp. 463–488, 2001Copyright © 2001 Elsevier Science Inc.Printed in the USA. All rights reserved

0891-5849/01/$–see front matter

PII S0891-5849(00)00373-7

463

the biological effects of peroxynitrite versus that of itsprecursors,•NO and O2

•2, and other•NO-derived oxi-dants.

In spite of the fact that the biological formation andreactions of peroxynitrite are kinetically and thermody-namically favored, the importance of the peroxynitritepathway in biology has been occasionally questioned[28–30], in part due to the difficulties related to itsdetection. Even though there is solid evidence supportingthe formation of peroxynitrite in vivo and its contributionto biomolecular damage and cell and tissue pathology,unambiguous detection and quantitation is not trivial andrequires a subtle knowledge of its biological chemistryand a multifaceted approach. In this work, we will (i)provide biochemical and physico-chemical foundationneeded to search for peroxynitrite, (ii) analyze currentmethodologies used in the detection of peroxynitrite and(iii) establish criteria that must be fulfilled to unravel theformation of peroxynitrite in biological systems of dif-ferent levels of organization and complexity.

PEROXYNITRITE BIOCHEMISTRY

Formation reactions

The biological formation of peroxynitrite anion(ONOO2) is mainly due to the fast reaction between•NOand O2

•2 (Eqn. 1). This radical-radical combination reac-tion undergoes with a second order rate constant that hasbeen independently determined as 4.3, 6.7, and 193 109

M21 s21 [31–33], and therefore one can safely assume avalue of ;1010 M21s21, which indicates a diffusion-controlled reaction.

•NO 1 O2•23 ONOO2 v 5 k@•NO][O2

•2](1)

Since both precursor radical species,•NO and O2•2,

are transient in nature, the biological formation of per-oxynitrite requires the simultaneous generation of bothradicals which, in addition, must approach and reactwithin the same compartment. However, while•NO hasa biological half-life in the range of seconds and readilydiffuses across membranes [34,35], O2

•2 lasts less thanmilliseconds and permeates membranes only via anionchannels [36]. Thus, due to both the greater half-life andfacile diffusion of •NO compared to O2

•2, peroxynitriteformation will predominantly occur nearer to the O2

•2

formation sites.Peroxynitrite anion exists in protonation equilibrium

with peroxynitrous acid (ONOOH, pKa5 6.8) [2]. Thus,under biological conditions both ONOO2 and ONOOHwill be present, the ratio depending on local pH (i.e., atpH 7.4, 80% of peroxynitrite will be in the anionic form).This is relevant because both species have different re-activities and diffusional properties [7].

In addition to the•NO plus O2•2 reactions, other pro-

cesses such as the reaction of nitroxyl anion with molec-ular oxygen [37,38],•NO oxidation of oxy-, hemo-, andmyoglobin [39], and turnover ofL-arginine-depleted ni-tric oxide synthase [40] may also contribute to the bio-logical formation of peroxynitrite.

Reactivity

Peroxynitrite promotes biological effects via differenttypes of reactions (Fig. 1), which could be classified in

Fig. 1. Peroxynitrite reaction pathways. Numbers I to V indicate possible fates of peroxynitrite: direct reactions include the one-electronoxidation of transition metal centers (Fe, Mn, Cu) (I); the two-electron oxidation with a target substrate (RH) (II) and the formationof nitroso-peroxocarboxylate (III), that rapidly decomposes to secondary radicals in 35% yield. Peroxynitrous acid undergoeshomolysis at 0.9 s21 to yield free radicals in 30% yields (IV) or rearrange to nitrate (V).

464 R. RADI et al.

three main groups:

1) direct redox reactions (I and II)2) reaction with carbon dioxide (III)3) homolytic cleavage of ONOOH (IV)

Recognizing the different reactivities of peroxynitriteis critical for planning adequate experimental strategiesdirected to assess its biological formation and quantita-tion.

In the direct reactions (e.g., interactions with metalcenters, thiol oxidation), peroxynitrite can promote one-or two-electron oxidation reactions with second orderrate constants in the order of 103 (e.g., thiols) to 106 M21

s21 (e.g., metal centers) [7,41]. The reaction with carbondioxide [42] is fast (k5 5.7 3 104 M21 s21) [43] andyields a short-lived intermediate (estimated half-life,1ms), nitroso-peroxocarboxylate (ONOOCO2

2), which ho-molyses to carbonate radical (CO3

•2) and nitrogen diox-ide (•NO2) in ;35% yields [44–46]. Carbonate radical isa relatively strong one-electron oxidant and•NO2 is amore moderate oxidant and also a nitrating agent; there-fore the radical products arising from ONOOCO2

2 de-composition promote secondary oxidation events. Fi-nally, ONOOH can undergo homolysis to•OH and•NO2,with a first order rate constant of 0.9 s21 at pH 7.4 and37°C in;30% yields, while the rest of ONOOH isomer-izes directly to nitrate (NO3

2) [1,47,48] (Fig. 1, route V).Hydroxyl radical is a more powerful oxidant than CO3

•2

and •NO2, but it is significantly less selective in targetmolecule reactions and addition reactions predominateover one-electron abstractions. In the absence of targets,the proton- or carbon dioxide-catalyzed decompositionof peroxynitrite mainly yields nitrate, due to recombina-tion of the radical intermediates arising from homolysis.However, in the presence of targets, most peroxynitriteyields nitrite (NO2

2).A wide variety of biomolecules can be oxidized by

peroxynitrite in vitro either by direct reactions or by thesecondary radicals (CO3

•2, •NO2,•OH). However, in bi-

ological systems where multiple direct targets of per-oxynitrite are present, the•OH-pathway is slow in com-parison with the direct bimolecular reactions ofperoxynitrite and a relatively small number of reactionspredominate [41]. In fact, in vivo, more than 95% of allperoxynitrite formed will be consumed by direct reac-tions, with less than 5% evolving to•OH and •NO2.Unlike •OH that typically reacts with most biomoleculeswith rate constants in the order of 109 M21s21, peroxyni-trite is a much more selective oxidant and reacts atsignificantly slower rates. The rate constants and prefer-ential pathways of peroxynitrite decomposition in biol-ogy have been reviewed elsewhere [7,41], as the contri-bution of a biomolecule to the overall fate ofperoxynitrite will be a function of both rate constant and

target concentration, reactions with sulfhydryls, transi-tion metal centers (Fe, Cu, and Mn), and carbon dioxiderepresent major initial pathways accounting for the bio-logical effects of peroxynitrite.

Nitration by peroxynitrite. Peroxynitrite promotes nitra-tion (incorporation of a nitro2NO2 group) of aromaticand aliphatic residues. Most notably, protein tyrosineresidues constitute key targets for peroxynitrite-mediatednitrations and the presence of 3-nitrotyrosine in proteinsrepresents a usual modification introduced by the biolog-ical formation of peroxynitrite [49]. Peroxynitrite doesnot react at appreciable rates with tyrosine [50] andtherefore tyrosine nitration by peroxynitrite requires theintermediate formation of secondary species. The nitra-tion process involves free radical mechanisms in whichone-electron oxidants derived from peroxynitrite attackthe aromatic ring leading to the formation of tyrosylradical, which then rapidly combines with•NO2 to yield3-nitrotyrosine [51,52] (Fig. 2).

In addition to 3-nitrotyrosine, the reactions of per-oxynitrite-derived oxidants with tyrosine, also yield 3,39-dityrosine [53] (Fig. 2). Small quantities of 3-hydroxy-tyrosine may be formed, following the addition reactionof •OH with tyrosine [54] (Fig. 2).

Peroxynitrite diffusion

Due to target molecule reactions, the biological half-life of peroxynitrite is estimated to be less than 100 ms[41,55]. This half-life is long enough for peroxynitrite topotentially travel some distances (e.g., 5–20mm) acrossextra- and/or intracellular compartments. However, inaddition to the estimated diffusion in aqueous environ-ments, the biological effects and detection of peroxyni-trite will be influenced by its ability to permeate cellmembranes. In this regard, both ONOO2 and ONOOHcan cross biological membranes, via anion channels andpassive diffusion, respectively [56,57].

The diffusion distances of peroxynitrite will criticallyinfluence the distribution of oxidative modificationswithin a tissue and the reaction yields with reportermolecules.

AFFIRMING THE BIOLOGICAL FORMATION OF

PEROXYNITRITE

General considerations

The detection of peroxynitrite relies on either (i) mod-ification of exogenously added probes, or (ii) footprint-ing reactions on endogenous molecules. However, theseare not straightforward procedures; at present there areno totally specific modifications of either probe or bi-

465Unraveling peroxynitrite

omolecules that can directly and unambiguously assurethe formation of peroxynitrite. Probe modification and/orfootprinting reactions require additional criteria to con-stitute sufficient evidence for affirming peroxynitrite for-mation. Some of these additional criteria involve phar-macological modulation, while others rely on theappreciation of the unique biochemistry of peroxynitriteversus that of•NO and other•NO-derived species such asnitrogen dioxide (•NO2), S-nitrosothiols (RSNO), and thespecies arising from catalytic action of hemeperoxidasesin the presence of hydrogen peroxide (H2O2) and nitrite(NO2

2).

Differential reactivities of•NO and ONOO2

An important aspect to unravel the formation of per-oxynitrite in biology is to recognize the differential re-activities (and effects) of•NO and peroxynitrite over celland tissues. In particular,•NO is neither a strong oxidantnor a nitrating agent; it mostly participates in reversibleinteractions with iron centers, radical-radical combina-tion reactions (e.g., with lipid radicals to terminate lipidoxidation chain reactions, with O2

•2 to form peroxyni-trite) and nitrosylation reactions via intermediate forma-tion of dinitrogen trioxide (N2O3). On the other hand,peroxynitrite is a strong oxidant and nitrating agent anda poor nitrosylating agent. Thus, several oxidation andnitration reactions measured in probes or biomoleculessecondary to•NO formation reflect the presence of per-oxynitrite.

It has become apparent [6,58–61] that a series of

“ •NO-dependent” biological processes, are notdirectlymediated by•NO, but they rather depend on the forma-tion of secondary species. For instance, the•NO-depen-dent inactivation of mitochondrial electron transportcomplexes I and II cannot be readily accomplished by•NO itself, but is efficiently mediated by peroxynitrite.Thus, in this case protection from inactivation by block-ing •NO production infers that secondary species derivedfrom •NO are participating in the process, and then orientfor the search of peroxynitrite as the proximal oxidant.

Probe oxidation

Oxidizable probes can be conveniently used to mon-itor peroxynitrite formation in biochemical and cell sys-tems. For quantitation purposes, probes would be ideallylocated close to the peroxynitrite forming sites and out-compete peroxynitrite reactions with various other bio-logical targets. Also, the selected probe should be readilyoxidized by peroxynitrite, but not by•NO and O2

•2.There are caveats in the use of probes which are to be

defined in each case: (i) intra- versus extracellular dis-tribution of the probe and detection of reactive interme-diates, (ii) contribution of competing reactions to overalloxidation yields, (iii) potential redox-cycling and/or sec-ondary reactions of probe radical intermediates, whichmay artifactually modify oxidation yields, (iv) limitedknowledge of the mechanisms of probe oxidation byperoxynitrite, which restrains quantitative interpretationof the data.

Fig. 2. Tyrosine oxidation pathways by peroxynitrite. Represented compounds are tyrosine (I), tyrosyl radical (II), tyrosine-hydroxylradical adduct (III), 3-nitrotyrosine (IV), 3-39-dityrosine (V), 3-hydroxytyrosine (VI), all of which can be yielded during reactions ofperoxynitrite-derived oxidants with tyrosine. Both CO3

•2 and•NO2 can perform a one-electron abstraction of tyrosine to yield tyrosylradical, while•OH predominantly (.95%) leads to the formation of a radical adduct that can either decay by water elimination totyrosyl radical, or be oxidized to 3-hydroxy-tyrosine, among other reactions. o-Tyrosine (VII) and 3-chlorotyrosine (VIII), products ofthe reaction of•OH or HOCl with tyrosine, respectively, are also shown. Final product distribution will largely depend on type and fluxof the different radical/oxidant species present as well as the amount of available free and protein-bound tyrosine.

466 R. RADI et al.

Footprinting

Peroxynitrite formation and reactions can be evi-denced through the detection of oxidative modificationsthat peroxynitrite promotes in target biomolecules. Oxi-dative modifications can be performed by different oxi-dants and/or may be readily reversed by appropriaterepair systems, therefore ideal modifications to measurewould be those that are (i) more specific for peroxynitriteversus other oxidant systems, and (ii) relatively perma-nent and stable. These modifications involve oxidationreactions in proteins, DNA or lipids, and most notably,protein tyrosine nitration.

Protein tyrosine nitration deserves special discussion.As •NO is incapable of directly promoting nitration re-actions, nitration was initially thought to constitute aspecific footprint of peroxynitrite reactions in biology[17]. However, it has become apparent that other mech-anisms of biological nitration may exist in addition toperoxynitrite, including (i) the H2O2-NO2

2-hemeperoxi-dase [62] (ii) the reactions of•NO2 (formed from theaerobic oxidation of•NO) with tyrosine residues [63],and (iii) the oxidation of unstable nitrosotyrosine [64].Importantly, however, these alternative mechanisms ofnitration appear to be more restricted than the nitrationmediated by peroxynitrite. Indeed, the peroxidase mech-anism requires the presence of a specific enzyme (e.g.,myeloperoxidase or eosinophil peroxidase) in the site offormation of oxygen radicals and•NO. This may belimited to those tissue regions or compartments andprocesses involving an important participation of acti-vated inflammatory cells. The•NO2 mechanism appearsto be of minor relevance because under normal or lowtissue oxygen tensions, the oxidation of•NO to •NO2 isa rather slow reaction, and also, because the initial reac-tion of •NO2 with tyrosine is not too fast and otherreactions such as•NO2-mediated thiol oxidation will bekinetically favored. Finally, the rapid reaction of tyrosylradical with •NO transiently yields nitrosotyrosine, in areaction that appears to be reversible. In the case of free

tyrosine, most nitrosotyrosine dissociates back to tyrosylradical and•NO, to ultimately predominantly yield 3,39-dityrosine [52]. However, in some proteins containingadjacent redox centers (e.g., prostaglandin synthase)[64], protein-bound nitrosotyrosine may be further oxi-dized to 3-nitrotyrosine by a “site-specific” one-electronoxidation to iminoxyl radical, followed by a secondone-electron oxidation to nitro-tyrosine. The mechanismof 3-nitrotyrosine formation involving the transient for-mation of nitrosotyrosine would be limited to a smallnumber of proteins.

Thus, nitration of biomolecules reveals the participa-tion of •NO-derived oxidants during oxidative damage totissues. In this context, peroxynitrite is a central contrib-utor to protein tyrosine nitration.

Peroxynitrite pharmacology

Pharmacological strategies are available to unravelperoxynitrite formation. Indeed, pharmacological modu-lation by NOS inhibitors and•NO scavengers, inhibitorsof O2

•2 formation and SOD/SOD-mimics and peroxyni-trite decomposition catalysts and scavengers is directedto attenuate peroxynitrite-mediated oxidative modifica-tions in biomolecules and probes as well as biologicaleffects such as cell death and tissue injury (Fig. 3).

Compounds used to interfere on the peroxynitritepathway and their mechanism of action will be discussedin more detail at the end of the chapter.

METHODOLOGIES FOR PEROXYNITRITE DETECTION

In this section we will describe current methodologiesfor peroxynitrite detection. The analysis will concentrateon those techniques that have been more widely used andvalidated. We will provide the biochemical backgroundfor each methodology, its potency and limitations. Thefield is in progress and awaits further application anddevelopment.

Fig. 3. Peroxynitrite pharmacology. Potential sites of pharmacological intervention to diminish peroxynitrite formation and reactionsare indicated and aimed to inhibit probe oxidation/nitration and footprinting reactions. NOS, NOX and XO are nitric oxide synthase,NADPH oxidase and xanthine oxidase, respectively.


Probe oxidation/nitration

Probes must ideally combine simplicity, sensitivity,and specificity for their application. In this regard, aseries of compounds have been successfully used todetect peroxynitrite. None of them being totally specific,knowledge of their reactivity with different oxidants andproper use of scavengers and inhibitors is required.

Oxidation of fluorescent probes

Two fluorescent probes, dichlorofluorescin (DCFH)and dihydrorhodamine (DHR), have been frequentlyused for assaying the formation of cell and tissue-derivedoxidants. Indeed, the two-electron oxidation of dichlo-rofluorescin to dichlorofluorescein (DCF;lex 5 502 nm,lem 5 523 nm) and dihydrorhodamine to rhodamine(RH; lex 5 500 nm, lem 5 536 nm), results in theformation of highly fluorescent products and is promotedby strong oxidants such as•OH, oxo-iron complexes, andperoxynitrite [65–69].

This method allows the detection of submicromolarlevels (e.g., 50 nM) of peroxynitrite. Additionally, sinceDCF and RH are strong chromophores at 500 nm («500559,500 M21 cm21 and 78,800 M21 cm21, respectively),larger amounts of peroxynitrite can be also detectedspectrophotometrically. Importantly, neither•NO norO2

•2 and only to a marginal extent•NO2, are able tooxidize either probe at significant yields.

The oxidation of DCFH or DHR by peroxynitriteresults in;35 and 42% oxidation yields (molecules ofoxidized probe per molecules of peroxynitrite) at pH 7.4,37°C, in potassium phosphate buffer, respectively [69].Oxidation yields are highly influenced by environmentalconditions including pH and buffer components as wellas by the presence of molecules that compete for per-oxynitrite including bicarbonate, thiols, and urate thatmay inhibit probe oxidation (Table 1). Changes in pHmay not only affect oxidation efficiency by peroxynitrite(which is the highest in the pH 7.4 region for bothprobes) [66,67], but it may also cause subtle effects onfluorescent intensity yields. Thus, it is critical to measureprobe oxidation yields with authentic peroxynitrite underthe conditions of the assay, as this will permit a moreaccurate determination of the levels produced.

The mechanisms of DCFH and DHR oxidation byperoxynitrite are largely unknown; this remains as alimitation for interpreting the probe oxidation data underconditions in which other biological targets for peroxyni-trite exist, since the probes used in the biological assaysare never present in large excess (typical concentration;10 mM), due to their high cost and relatively lowsolubility. For both DCFH and DHR, hydroxyl radicalscavengers are unable to inhibit oxidation, suggesting

that peroxynitrite itself may be mediating the oxidationprocess (Fig. 4). However, evidence for direct reactionsbetween peroxynitrite and the probes is at the momentlacking.

A separate note deserves the influence of excess•NOor O2

•2 on probe oxidation. Nitric oxide/superoxide for-mation ratios different to one will result in peroxynitriteformation but also in remaining quantities of one of theprecursor radical species. Both excess•NO and O2

•2 maydecrease probe oxidation yields [68,70] due to reactionswith radical intermediates arising from one-electron ox-idation of the probes (Fig. 4). Thus, excess radical for-mation may result in underestimation of peroxynitrite

Table 1. Probe Oxidation Yields and Influence of MediaComponents

Probea Yield

Influence of components

CO2 ThiolsUricacid Desferrioxamine

DCFH [67,69] 35 2 22 2 2DHR [66,69] 42 2 22 22 2Tyrosine [4,43,200] 6 1 22 22 2Cyt c21 [43,119,

120].50 — 2 — —

Luminolb [78,83,133]

(;20%)c 11 22 22 22

a Data provided were obtained at pH 7.4, 37°C in phosphate bufferand from bolus addition of authentic peroxynitrite. Symbols indicate:2, partial inhibition;22, total inhibition; —, without influence;1,moderate enhancement;11; large enhancement.

b Referred as light emission.c Non applicable, as luminol oxidation yield is not utilized for

detection/quantitation purposes.DCFH 5 dichlorofluorescin; DHR5 dihydrorhodamine.

Fig. 4. Dichlorofluorescin oxidation by peroxynitrite. Dichlorofluores-cin (DCFH) is postulated to be oxidized by one-electron by peroxyni-trite (I) to yield the corresponding radical anion (DCF•2), which iseither dismutated (II) or further oxidized (III) to yield dichlorofluores-cein (DCF). Secondary relevant reactions include the reaction ofDCF•2 with molecular oxygen to yield O2

•2 (IV), or with •NO to yieldnonfluorescent products (P) (V). Analogous reactions are expected tooccur with dihydrorhodamine.

468 R. RADI et al.

formation rates due to secondary,•NO/O2•2-mediated ter-

mination reactions.It has been recently pointed out that the one-electron

oxidation product of DCFH, the DCF semiquinone rad-ical (DCF•2) may subsequently react with molecularoxygen to yield O2

•2 and DCF [71,72] (Fig. 4) and thenartifactually enhance probe oxidation yields. This mech-anism of probe-dependent O2

•2 formation may also beoperative for other fluorescent compounds of similarstructure to that of DCFH, such as DHR. Indeed, DCF•2-dependent O2

•2 formation could, in principle, contributeto generating peroxynitrite in•NO-generating systems.This phenomenon requires study but may not be criticalduring peroxynitrite-mediated DCFH oxidation, sinceexcess•NO over O2

•2 leads toinhibition instead of aug-mentation of probe oxidation [68], indicating that, if any,O2

•2 formation by DCF•2 does not significantly contrib-ute to further formation/detection of peroxynitrite. In anyevent, experiments designed to assess (i) cell O2

•2 for-mation in the presence of DCFH, and (ii) DCFH oxida-tion with and without NOS activity, will provide valu-able information to account for or rule out the detectionof potentially high values of peroxynitrite.

DCFH and DHR oxidation may serve to indicateeither intra- or extracellular formation of peroxynitrite.Both probes are able to enter cells and participate inintracellular oxidation reactions [65]. Thus, extracellularformation of peroxynitrite may be preferentially studiedby adding the probes right before the beginning of theexperiment. On the other hand, preincubation of cellswith DCFH or DHR for;1 h leads to significant probeaccumulation intracellularly, and both intra- and extra-cellular-associated fluorescence can be studied. How-ever, there are potential pitfalls in the interpretation ofthe data for the following reasons: probes oxidized in acompartment, for example, extracellular, may diffuse tointracellular compartments and vice versa, with the ex-ception of the extracellular diffusion of RH; indeed, RHis lipophilic, positively charged, and concentrates in mi-tochondria and other negatively charged cell compart-ments, and will mostly remain inside the cell as long asmitochondria maintain electrochemical potential. How-ever, peroxynitrite itself may lead to mitochondrial dys-function and therefore RH leak to extracellular milieu.Also, cell washings after incubation with probes result inextracellular diffusion of intracellular DCFH, DCH, andDHR [65,69]. Thus, the dynamics of parent and oxidizedprobe permeation through cell membranes must be spe-cifically studied. In any event, fluorescence of cell su-pernatants, cell homogenates, and even fluorescence mi-croscopy has been applied successfully for peroxynitritedetection with DCFH and DHR [73–77].

Chemiluminescence probes

Luminol (5-amino-2,3-dihydro-1,4-phthalazinedione)is a chemiluminescent probe widely use for detection ofoxygen radical intermediates produced in biological sys-tems. The general mechanism of chemiexcitation in-cludes a two-electron oxidation process and the produc-tion of an unstable endoperoxide intermediatedecomposes yielding excited 3-aminophtalate, which de-cays to ground state aminophtalate with the emission oflight with a maxima at 425 nm [78].

Peroxynitrite is capable of inducing luminol chemilu-minescence (LCL), especially in the presence of bicar-bonate-carbon dioxide, which results in high quantumyields [78]. Indeed, this technique has been successfullyused in a number of studies for peroxynitrite detection atthe cellular and organ level [79–82]. The mechanism ofchemiexcitation of luminol by peroxynitrite in the pres-ence of bicarbonate-carbon dioxide, involves the forma-tion of CO3

•2 from the homolytic decomposition ofONOOCO2

2, which readily performs a one-electron ox-idation of luminol, initiating the pathway for light emis-sion (Fig. 5). Peroxynitrite-mediated light emission isinhibited by thiols and urate (Table 1), and quantumyields increase with pH, especially because only decom-position of luminol monoanion (pKa5 8.2) leads to thelight-emitting route. Chemiluminescence makes it possi-ble to follow the time course of peroxynitrite formation,and with highly sensitive photon counting techniques[83], peroxynitrite fluxes down to;1 nM/min can beaccurately detected.

LCL can be induced by other biologically relevantoxidants such as•OH and O2

•2, which may operate se-quentially during luminol oxidation; however,•NO is notcapable of inducing LCL, and the potential contributionof •NO2 is, at most, marginal [83]. Thus, inhibition ofLCL by NOS inhibitors is compatible with the formationof peroxynitrite. Nevertheless, it is important to note thatexcess•NO may lead to inhibition of luminol chemilu-minescence, most likely via termination reactions withluminol radical intermediates [83] (Fig. 5). In this regard,quantum yields from authentic peroxynitrite are some-what higher than those obtained with equimolar fluxes of•NO and O2

•2.In addition to luminol, recently another highly sensi-

tive probe, coelenterazine (2-(4-hydroxybenzyl-)-6-(4-hydroxyphenyl)-8-benzyl-3,7-dihydroimidazo[1,2-}]py-razin-3-one), has been shown to efficiently produce lightemission during its reaction with reactive oxygen speciesand peroxynitrite [84]. Further studies characterizingperoxynitrite-induced coelenterazine chemiluminescencein chemical and biological systems are being carried outin our laboratories.

While O2•2-producing systems also lead to the che-


miexcitation of another widely used probe, lucigenin(10,10-dimethyl-bis-9,9-bisacridinium nitrate) via a re-ductive dioxygenation process [85], lucigenin is unableto yield light by reactions with oxidants. Thus, peroxyni-trite is unableto promote lucigenin chemiluminescenceas described in an early report [78]. Then, while•NOmay enhance luminol and coelenterazine-dependentchemiluminescence in O2

•2-producing systems throughthe formation of peroxynitrite, it will inhibit lucigeninchemiluminescence. Combined utilization of luminoland lucigenin may provide further insights with regard tothe relative contributions of oxygen radicals and per-oxynitrite to observed chemiluminescence.

Nitration of phenolic compounds

Peroxynitrite promotes nitration and oxidation of phe-nolic compounds such as tyrosine or p-hydroxyphenyl-acetic acid (p-HPA). Nitration leads to the formation of3-nitrotyrosine or 3-nitro-pHPA, while oxidation resultsin dimerization and hydroxylation reactions, which canalso be accomplished by other oxidant systems such as•OH and oxo-iron complexes. Thus, nitration of phenolicprobes has been utilized as a more specific modificationinduced by peroxynitrite and potentially other•NO-de-rived oxidants.

Nitration of the aromatic compound p-hydroxyphenyl-

acetic acid (p-HPA) in the extracellular milieu was firstused to detect peroxynitrite release from activated mac-rophages [86] and more recently for the detection ofnitrating species generated by activated neutrophils [62].In addition, free tyrosine has been used to trap nitratingspecies arising from peroxynitrite formed from chemical/biochemical fluxes of•NO and O2

•2 [52,87].3-Nitro-pHPA and 3-nitrotyrosine are stable products

and can be directly assessed spectrophotometrically,since they have a characteristic absorbance in the 350–450 nm region that is strongly pH-dependent [88]. Whileat low pH (,5.5) 3-nitrotyrosine has a peak absorbanceat 360 nm (« 5 2,790 M21 cm21), at alkaline pHmaximum absorption is at 430 nm (« 5 4,400 M21

cm21). Since dissociation of the aromatic hydroxylgroup in 3-nitrotyrosine is responsible for the shift inabsorbance to 430 nm (pKa5 7.5), samples can bequantitated at 430 nm by the difference between absor-bance values at pH,5.5 and pH;10. The limit ofsensitivity is;1 mM and potentially subject to interfer-ence from other compounds that may absorb in the sameregion, therefore it is only useful for relatively concen-trated and pure samples.

More sensitive and specific detection of 3-nitro-PHPAor 3-nitrotyrosine typically requires high performanceliquid chromatography (HPLC) or gas chromatography(GC) separation and different alternative methods for

Fig. 5. Luminol chemiluminescence by peroxynitrite. Carbonate radicals arising from the decomposition of ONOOCO22 promote a

one-electron oxidation to luminol radical, which then reacts with O2•2 to follow the light emitting pathway. Under excess•NO, luminol

radicals yield intermediates and products (P) that led to a dark route.

470 R. RADI et al.

detection, including UV/VIS, electrochemical (EC), flu-orescent, and mass spectrometry (MS), depending onrequired detection limits and equipment availability andexpertise (Table 2). This methodology also allows theconcomitant detection of the parent aromatic compound(i.e., p-HPA and tyrosine) and other oxidized formsincluding dimeric and hydroxylated products. SeveralHPLC methods have been developed to determine theseanalytes, the vast majority relying in reverse phaseHPLC with different mobile phase setups [21,86–93](Table 2). UV-VIS detection can be accomplished by theuse of acidified samples and simultaneous detection at280 and 360 nm. The A280/A360 ratio for 3-nitrotyrosineis 2.3, while tyrosine and other oxidation products suchas 3-hydroxytyrosine and 3,39-dityrosine do not absorb at360 nm. Standards of 3-nitrotyrosine are commerciallyavailable. Electrochemical detection can achieve over100-fold higher sensitivity than UV-VIS (see Table 2)and various EC methods have been developed. An in-trinsic problem of EC-detection of 3-nitrotyrosine relieson the high voltage required for a response (;800 mV),which results in instability of baseline values and de-creased specificity, since other compounds that may coe-lute with 3-nitrotyrosine and even components of the

mobile phase may also provide an electrochemical sig-nal. Various approaches have been used to overcome thisproblem, including the use of array detector systemsand/or reduction of 3-nitrotyrosine to 3-aminotyrosine, atreatment that substantially reduces the potential for ox-idation (;70–100 mV). 3-Nitrotyrosine is resistant toreduction by common reductants such as ascorbate anddithiothreitol, but can be readily reduced by addition ofsodium dithionite under alkaline conditions. 3-Aminoty-rosine is colorless and fluoresces at pH5 3.0–3.5 withlexc 5 277 nm andlem of 308 and 350 nm, which canserve for fluorescence detection purposes. Since tyrosinealso fluoresces at 308 nm, thelem at 350 nm is the oneof choice, when assessing 3-aminotyrosine by fluores-cent detection methods. 3-Aminotyrosine typically has apoor retention in reverse-phase HPLC columns, thuseither derivatization (e.g., acetylation) [89,94] or use ofion pairing mobile phases [90] have been designed toshift 3-aminotyrosine retention to a convenient elutiontime. Commercial standards of 3-aminotyrosine areavailable; however, the reduction of 3-nitrotyrosine to3-aminotyrosine is in many cases not complete and3-aminotyrosine can progressively be air-oxidized backto 3-nitrotyrosine.

Table 2. Selected Methods for Detection of 3-Nitrotyrosine

Method AnalyteSensitivity

(pmol) Comments Reference

HPLC/UV-VIS 3-Nitrotyrosine 10 Widespread availability and simple. [88]3-39-Dityrosine No derivatization required.Tyrosine Low sensitivity.

HPLC/EC N-acetyl-3-aminotyrosine

0.02 Improved signal to noise ratio bydetection of 3-aminotyrosine.Alternative method for detection oftyrosine needed.

[89]

HPLC/EC 3-Nitrotyrosine 10 Allows detection of multiple [91]array system 3-Aminotyrosine 1.0a analytes in a single run.

3,39-Dityrosine Increase direct specificity by3-Hydroxytyrosine EC signature.Tyrosine Detection system not readily

available.

HPLC/Fluo 3-Aminotyrosine More sensitive than UV-Vis. [95]3-39-Dityrosine – Minimally explored and

validated.

GC/MSb 3-Nitrotyrosine 0.050 Highly sensitive and specific. [97](NICI) 3-Aminotyrosine 0.0004a,c Requires preparative steps and

3,39-Dityrosine specialized equipment.3-HydroxytyrosineTyrosine

a Sensitivity values for 3-nitrotyrosine and 3-aminotyrosine are independently indicated.b Chromatographed and measured as n-propyl heptafluorobutyryl-derivatives.c The limit of detection of the n-propyl heptaflurobutyryl-derivative of 3-aminotyrosine is 400 amol.HPLC5 high performance liquid chromatography; EC5 electrochemical; GC5 gas chromatography; MS5 mass spectrometry; NICI5 negative

ion chemical ionization.


Recent studies have also used gas chromatographyseparation methods followed by mass spectrometry [95–100] (Table 2). Of these, isotope dilution GC-MSmethod in the negative ion chemical ionization (NICI)mode for the n-propyl, heptaflurobutyryl-derivative of3-nitrotyrosine [97] results in a limit of detection (signalto noise ratio.10) of ;50 fmol and reduction to 3-ami-notyrosine of the derivatized amino acid allows detectiondown to ;400 amol (attomol) (Table 2). This methodprovides specific structural information on the analyte,minimizing potential confounding effects of compoundsthat coelute with the target analyte. It represents a highlysensitive technique, which has been successfully appliedfor the detection of 3-nitrotyrosine from protein hydro-lyzates of tissue samples [96] (Table 2). Recently, an-other highly sensitive method for detection of plasmafree 3-nitrotyrosine was developed [99]. In this, free3-nitrotyrosine was separated from protein-bound 3-ni-trotyrosine by ultrafiltration, then excluded from nitrateand nitrite by HPLC and recovered by solid phase ex-traction, derivatized, and analyzed by GC-tandem MS inNICI mode. By this several steps methodology, the au-thors obtain an averagebasal concentration of plasmafree 3-nitrotyrosine of 2.8 nM and indicate an overallrecovery of 3-nitrotyrosine of 50% and a limit of detec-tion as low as 4 amol.

As seen in the previous paragraphs, several poten-tially useful methods for nitro-PHPA and 3-nitrotyrosinedetection are available. Each laboratory must adapt ordevelop the appropriate methodology depending on theexperimental models under study. Given that the utilizedmethod will be specific for 3-nitrotyrosine, the decisionshould be made based on expected detection levels, theintrinsic complexity of the method, and the availabilityof special equipment such as multielectrode array detec-tor systems or GS-MS technology, not readily accessiblein all laboratories.

Nitration yields by peroxynitrite in biological systemsare typically low due to the intrinsic reaction chemistryand competing reactions (Table 1). Some compounds canenhance and catalyze nitration, including bicarbonate-carbon dioxide [43], Fe-edta, hemeperoxidases, or inac-tivated SOD [4,49]. However, the modulation of biolog-ical nitration by these compounds has been marginallyexplored and requires further specific investigation.

Specificity of tyrosine nitration as a footprint of per-oxynitrite. Nitration is not totally specific of peroxyni-trite, since other nitrating species may be released fromcells and lead phenolic nitration as well, as recentlyindicated for the case of activated human neutrophils[62] and monocytes [101], which can promote extracel-lular nitration via a myeloperoxidase mechanism. Thus,specific pharmacological experiments to define whether

the peroxynitrite or the nitrite-hemeperoxidase pathwaylead to phenolic nitration may be necessary to perform insystems in which both mechanisms may coexist, such asin the case of neutrophils, monocytes, or eosinophils (inthis latter case due to eosinophil peroxidase, (EPO)[102]. In addition, myeloperoxidase- and eosinophil per-oxidase-dependent mechanisms of oxidation lead to theformation of halogenated derivatives of tyrosine [102–104] (e.g., 3-chlorotyrosine, 3-bromotyrosine), whichcan also be detected by some of the techniques describedabove.

Nitration yields by•NO plus O2•2 fluxes. An important

issue is whether fluxes of•NO and O2•2 would result in

comparable nitration yields to that of authentic peroxyni-trite [4,52,87,105]. It is becoming apparent that nitrationyields by fluxes of•NO and O2

•2 are highly dependent onfree radical formation rates and ratios, with lower nitra-tion yields obtained at low•NO plus O2

•2 fluxes [87].Dimerization of tyrosyl radicals to 3,39-dityrosine pre-dominate at low fluxes, while the tyrosyl radical combi-nation reaction with•NO2 to yield 3-nitrotyrosine be-comes more relevant at higher fluxes. Thus, the ratio ofoxidized over nitrated products (Fig. 2) may vary de-pending on the rate of formation of peroxynitrite andeven the concentration of reporter molecules. The rele-vance of these findings to nitration (and oxidation) yieldsin reporter molecules by cell or tissue-derived•NO andO2

•2 remains to be established.

Other methods

Conceptually, any molecule that can be oxidized ornitrated by peroxynitrite leading to a stable and measur-able change could potentially be used for peroxynitritedetection. In this regard, other methods relying on oxi-dation reactions have been used, including aromatic hy-droxylation, EPR-spin trapping detection, and oxidationor formation of chromophores.

Aromatic hydroxylation. Aromatic compounds such asbenzoate, phenol, phenyalanine, and salicylate becomehydroxylated and nitrated by peroxynitrite via•OH and•NO2 reactions formed from ONOOH homolysis, andcan be used to follow the formation of peroxynitrite[43,54,106,107]. Benzoate hydroxylation has been con-veniently used in simple biochemical systems by directfluorescence spectroscopy (lexc 5 300 nm,lem 5 410nm) for competition assays [43]. Phenylalanine reactionswith peroxynitrite leads to the formation of p-, m- ando-tyrosine, specific products of•OH radical attack, aswell as nitrated products, including possibly 4-nitrophe-nylalanine among other products [106]. Salicylate reacts

472 R. RADI et al.

with peroxynitrite-derived species leading to the forma-tion of various products including 2,3-dihydroxybenzo-ate, 2,5-dihydroxybenzoate, and 2-hydroxy-5-benzoate[54]. The two hydroxylated derivatives are typical of•OH reaction with salicylate and the nitro-derivative isspecific for•NO-derived oxidants. The combined detec-tion of hydroxylated and nitrated products arising fromperoxynitrite reactions with phenylalanine and/or salic-ylate, could be applied for the in vitro and even in vivodetection of peroxynitrite. In vivo, however, free 3-ni-trotyrosine formation may be a more sensitive markerthan salicylate because the yield of peroxynitrite reactionwith salicylate is much smaller than with tyrosine [54].This is supported by the fact that increased levels of free3-nitrotyrosine were not accompanied with elevated•OHtrapping by exogenous addition of salicylate, in spinalcord and brain stem of a mice model of amyotrophiclateral sclerosis, a neurodegenerative disorder in whichperoxynitrite is proposed to play a pathogenic role [108].

EPR-spin trapping of peroxynitrite-derived oxidants.The EPR-spin trapping technique has been proved tobe useful to study peroxynitrite decomposition path-ways and target molecule reactions and mechanisms[46,47,109 –112]. Spin traps also have some potentialfor detection of peroxynitrite derived-oxidants or sec-ondary radicals formed in biochemical/cellular sys-tems. The formation of•OH from peroxynitrite decom-position can be detected using different spin traps suchas the cell permeable hydrophilic DMPO (5,5-dimeth-ylpyroline N-oxide), that yield stable spin trap-OHadducts [47,109]. The reaction mechanism of per-oxynitrite with DMPO involves the formation of•OHafter homolysis of peroxynitrous acid, and therefore itis unable to compete efficiently under conditions inwhich other targets that react with second order kinet-ics with peroxynitrite are present. Other frequentlyused spin traps such as the cell permeable (lipophilic)PBN (C-phenyl N-tert-butylnitrone) and (hydrophilic)POBN (}-4-pyridiyl-1-oxide N-tert-butylnitrone) pro-vide more stable signals with carbon-centered radicalsand can be used as long as a•OH scavenger such asethanol or DMSO is added in excess to the system, inwhich case the radical character is transferred to yielda-hydroxy-ethyl or methyl radicals that then form spinadducts with the traps [47].

Other compounds have been proposed for peroxyni-trite detection by EPR, such as the cell-permeable TEM-PONE (1-hydroxy-2,2,6,6-tetramethyl-4-oxo-piperidine)and CP-H (1-hydroxy-3-carboxy-2,2,5,5-tetramethylpir-rolidine) or non-cell-permeable PP-H (1-hydroxy-4-phosphono-oxy-2,2,6,6-tetramethyl pyridine) [113–115].These compounds are not spin traps, because they do notparticipate in addition (trapping) reactions with radical

intermediates, and therefore do not provide structuralinformation on the attacking reactive species; they are,rather, prone to one-electron oxidation reactions by sev-eral oxidants including O2

•2 or •OH derived from homol-ysis of ONOOH, to yield stable nitroxides.

These methods could potentially be used for biochem-ical as well as more complex biological systems but havebeen only minimally explored. In particular, their use incell (and even animal) systems remains to be exploredand validated. A major problem with several spin traps isthat spin adducts may decay in the presence of excessperoxynitrite or•NO2 (e.g., DMPO-OH), and then thesignal may be formed and “silenced” due to the sameoxidants. Low concentrations of thiols such as GSH(e.g., 1 mM) may help to protect the signal from excessoxidant, but small amounts of thiyl radical may beformed. Another problem can arise from the reduction ofthe spin adducts by cellular reductants such as glutathi-one and ascorbate [109]. Thus, all these variables mustbe considered when using and validating spin traps forperoxynitrite studies.

Current EPR-spin trapping methodology does not pro-vide precise quantitation of peroxynitrite formed sincethe radicals trapped are always a small fraction of thoseformed and spin adducts can decay, be reduced, or fur-ther oxidized. However, this method may serve to estab-lish qualitative differences among different experimentalconditions and the use of cell-permeable or nonperme-able probes may provide data in regards to peroxynitriteformation sites, all of which needs to be established.

Oxidation or formation of chromophores. Use of com-pounds that undergo significant absorbance changes dur-ing oxidation reactions could be potentially applied tofollow peroxynitrite formation in biochemical systems.Among these, compounds such as carminic acid, gallo-cyanine, pyrogallol red [116], 2,29-azinobis(3-ethylben-zothiazoline-6-sulfonate) (ABTS) [117] and ferrocya-nide [118], react with the•OH radical arising fromdecomposition of ONOOH. Others such as iodide, Ni21

cyclam [118], and reduced cytochrome c (cytochromec21) [119] react directly at variable rates with peroxyni-trite. Many of these compounds have been applied incompetition kinetic studies and their use as peroxynitritedetectors require further characterization, as it may behampered by the fact that some will be toxic to cells andothers can react and/or undergo redox cycles with inter-mediates other than peroxynitrite.

To illustrate some of the complexities and potentialpitfalls behind the use of oxidizable chromophores forperoxynitrite detection, we will briefly describe theuse of cytochrome c21, which we have successfullyapplied to follow peroxynitrite formation by biochem-ical systems [119,120], and also to determine second


order rate constants of targets via competition kinetics[121,122].

Cytochrome c21 oxidation. While the reduction of oxi-dized cytochrome c (cytochrome c31) is widely em-ployed to follow O2

•2 formation by biochemical andcellular systems [123,124], the cogeneration of•NO andO2

•2 usually limits O2•2 detection by this technique since

the formation of peroxynitrite outcompetes the O2•2-de-

pendent cytochrome c31 reduction. Moreover, cyto-chrome c21 is readily oxidized by peroxynitrite at pH 7.4and 37°C (k5 2.5 3 104 M21s21) and at;50 mMresults in.50% oxidation yields [119]. The method hasthe following advantages: (i) fast-second order-kineticsof cytochrome c21 with peroxynitrite, (ii) cytochromec21 can be conveniently prepared in concentrated (mM)stock solutions and added in excess (e.g., 50–100mM) tothe reaction to outcompete most other reactions of per-oxynitrite, and (iii) oxidation to cytochrome c31 can bereadily followed spectrophotometrically («550 5 21,000M21 cm21), either as a continuous or as end-point assay.The main limitations with the method are: (i) air oxida-tion to yield O2

•2, (ii) cytochrome c21 oxidation byexcess hydrogen peroxide (k5 ;2 M21 s21) [125], (iii)reduction of cytochrome c31 by O2

•2 (k 5 3 3 106 M21

s21) [126], (iv) reactions of excess•NO with eitherreduced (k5 ;10 M21 s21) or oxidized (k;103 M21

s21) cytochrome c [12]. Air oxidation is slow and can bediscounted by using appropriate controls. The addition ofcatalase minimizes the potential contribution of H2O2-dependent oxidation; the reduction reaction of cyto-chrome c31 back to cytochrome c21 becomes relevantonly if the oxidized form is accumulated significantlyand under conditions of excess O2

•2 over •NO. The reac-tions of •NO with cytochrome c, especially cytochromec21, are slow, and its potential participation can bediagnosed by the formation of cytochrome c-nitrosylcomplexes, of characteristic absorption spectra. Cautioususe of this method has been successfully applied in vitroboth for measuring peroxynitrite fluxes formed fromchemical/biochemical fluxes of•NO plus O2

•2, and fordefining rate constants and reaction mechanisms of com-peting targets. Importantly, due to its fast reaction rate,peroxynitrite-mediated cytochrome c21 oxidation is noteasily affected by low molecular weight scavengers suchas uric acid that affect many peroxynitrite-mediated ox-idative processes [119,120] (Table 1).

Indirect methods

These methods rely on the independent detection of•NO (directly, end products, or by bioassays) and O2

•2.Direct detection of•NO (either by electrochemical de-

tection or by EPR techniques) or its bioactivity (i.e.,vasodilation, inhibition of platelet aggregation, cGMPaccumulation) can be performed and evaluate whethersignificant interactions with O2

•2 exist. Indeed, if thiswere the case SOD or SOD mimics should enhance theconcentration of•NO and its bioactivity [127,128]. Acommonly used method for the detection of•NO inbiology is the quantitation of NO2

2 plus NO32 (NOx

2). Inbiochemical and cellular systems, most•NO preferen-tially decays to NO2

2, with NO32 being typically less

abundant. The interactions of•NO with O2•2 increase the

formation of NO32, due to the proton-catalyzed peroxy-

nitrous acid (Fig. 1). Thus, at the same output of•NO andthen total amount of NOx

2, the formation of O2•2 leading

to peroxynitrite increases the nitrate/nitrite ratio [86]. Onthe other hand, O2

•2 detection by methods such as theSOD-inhibitable reduction of cytochrome c21 under con-ditions in which peroxynitrite is being formed may bemarginal, but will be greatly enhanced by the addition ofNOS inhibitors [86].

Reaction yields and quantitation

In order to quantitate peroxynitrite, it is critical toknow the reaction yield of the probe used (i.e., DHR,DCF, cytochrome c21, p-HPA) with authentic peroxyni-trite (Table 1). However, it is important to recognize thatbolus addition of peroxynitrite to probes does not mimicthe low flux formation of peroxynitrite that may occurunder biologically relevant conditions. In addition, sincemost reactions involving probe oxidation by peroxyni-trite occur through free radical intermediates, and the fateof free radicals (including probe radicals) is typicallydependent on steady-state concentrations, oxidationyields obtained with acute addition of peroxynitrite canbe, at most, an approximation to what is expected tooccur in biology. In particular, phenolic nitration byperoxynitrite infusion over an extended period of time orindependent fluxes of•NO and O2

•2 results in lowernitration yields than those observed during bolus addi-tion of peroxynitrite; however, peroxynitrite fluxes maybe slightly more efficient in other assays such as DHR(unpublished observations).

Peroxynitrite quantitation is affected by various otherfactors including (i) competing reactions that peroxyni-trite may have with cell-tissue components, (ii)•NO andO2

•2 reactions with probe radical intermediates, (iii) re-dox reactions of probe radical intermediates with molec-ular oxygen or reductants, among other reactions. Not-withstanding, the independent detection of•NO and O2

•2

provides critical information regarding the existence ofpotential confounding factors during peroxynitrite detec-tion and help to obtain accurate values in both biochem-ical and biological (cell-tissue-organ level) systems.

474 R. RADI et al.

Potential pitfalls and artifacts

The detection of peroxynitrite by oxidizable probes ishighly affected by conditions such as pH, temperature,metals, buffer composition, presence of carbon dioxide-bicarbonate, and biomolecules such as thiols.

The reaction with carbon dioxide. The reaction of peroxyni-trite with carbon dioxide is critical to consider becausecarbon dioxide is ubiquitous in biological systems and alsobecause sometimes carbon dioxide levels are poorly con-trolled in reaction systems. Indeed, variable levels of carbondioxide are present on air-equilibrated solutions and buffersystems, depending on composition, ionic strength, andtemperature. In air-equilibrated 50 mM phosphate buffer,pH 7.4 a typical value of contaminating carbon dioxide is inthe order of 5–10mM, unless special measures are taken forpreparation of the solutions [129]. Moreover, stock solu-tions of alkaline peroxynitrite may provide significantamounts of carbon dioxide, arising from equilibration oftrapped carbonate with neutral pH buffers. Experiments canbe specifically designed in the presence of known amountsof carbon dioxide-bicarbonate. In addition, cells and tissuesactively produce carbon dioxide and cell cultures are typi-cally equilibrated with 5% carbon dioxide that translates toa dissolved concentration of;1.0–1.5 mM carbon dioxidein equilibrium with bicarbonate at pH 7.4. Carbon dioxidehas the following net consequences on peroxynitrite bio-chemistry: (i) decreases peroxynitrite half-life and thereforeits distance of diffusion, (ii) decreases one-electron oxida-tions and hydroxylation of probes, (iii) enhances nitration ofphenolic probes. Therefore, in biochemical, cell, tissue, andorgan levels, the presence of carbon plays a critical role inmodulating peroxynitrite biochemistry and affecting reac-tion yields of the different probes.

Buffer systems, media components, and metals. Phos-phate buffer does not interfere in peroxynitrite reactions,but other commonly used buffer systems including Trisand Hepes react with peroxynitrite inhibiting oxidativeprocesses [130] and may secondarily lead to the forma-tion of radical intermediates and•NO-donors [131,132].Similarly, the presence of thiols, glucose, proteins,and/or plasma will decrease probe oxidation yields.

Transition metals also react with peroxynitrite andmay affect probe oxidation yields. In biochemical sys-tems, the use of metal chelators such as dtpa (diethyl-enetriamine-N-N-N9-N0-pentaacetate) allows a moreprecise detection of peroxynitrite, minimizing poorlycontrolled metal-catalyzed oxidation reactions. How-ever, dtpa reacts with•OH and CO3

•2, and therefore largeconcentrations (.100 mM) should not be used. Impor-tantly, desferrioxamine is not recommended because ithas been shown to inhibit peroxynitrite-mediated oxida-

tive processes, independently of its capacity to bindtransition metals [3,133]. In a separate approach, addi-tion of known amounts of transition metal-complexesmay serve to significantly increase nitration yields.

Footprinting

The reactions of peroxynitrite with cell and tissuecomponents leave oxidative modifications that may serveas “footprints” of peroxynitrite formations and reactions.Of these, the most frequently used is the nitration ofprotein tyrosine residues. Protein nitration is a ratherstable chemical modification and metabolic systems ableto repair or reduce nitration are either nonexistent oroperate at low rates [134–136]. Indeed, at present thereis no solid evidence for the presence of specific meta-bolic systems that remove or reduce 3-nitrotyrosine.However, proteolysis of oxidized proteins may contrib-ute to 3-nitrotyrosine turnover [137], and can affect3-nitrotyrosine stability in stored samples over extendedperiods of time or in samples that undergo repeatedfreeze-thawing cycles. Detection and quantitation of3-nitrotyrosine in biological samples (cells and tissues)have been detected by one of two main strategies: im-munochemical detection of nitrated proteins, and quan-titation of 3-nitrotyrosine after protein hydrolysis. Al-though detection of protein 3-nitrotyrosine can increaseseveral-fold in diverse pathophysiological situations, ni-trated proteins exist under physiological conditions, sup-porting the concept of a low flux of oxidants beingformed and causing molecular damage, even under basal/normal conditions.

Immunochemistry for 3-nitrotyrosine

Immunochemical-based methods rely on the use ofanti-nitrotyrosine antibodies [138]. Polyclonal andmonoclonal antibodies have been raised and purified bydifferent laboratories and used as immunological probesin different experimental methods [138–140]. Immuno-chemical methods are useful tools for studying biologicalsamples and the only to be applied to whole tissues. Theyhave the advantage of being simple and they require onlyroutine laboratory equipment, and therefore can be per-formed in most research and clinical laboratories.

The preparation and characterization of antinitroty-rosine antibodies have been described recently else-where [138,139]. As there are no controlled studiescomparing both specificity and sensitivity of the avail-able antibodies, precise characterization must be per-formed before utilization. The specificity of these an-tibodies for nitrated proteins in the samples understudy have to be confirmed by two types of experi-


ments. First, antibody binding must be displaced byeither free 3-nitrotyrosine or, more efficiently, bysmaller amounts of 3-nitrotyrosine-containing pep-tides such as Gly-NO2Tyr-Ala or Gly-Gly-NO2-Tyr-Ala. Secondly, samples must be reduced by sodiumdithionite to 3-aminotyrosine, therefore preventing an-tibody binding. While monoclonal antibodies are morespecific than the polyclonal, they are less sensitive fordetection of protein-bound 3-nitrotyrosine.

Enzyme-linked immunosorbent assays (ELISA)

ELISA is a very versatile method to detect and mea-sure nitrated proteins. In a simple ELISA, a standardnitrated protein (e.g., nitrated bovine serum albumin) iscoated on the plate and the antinitrotyrosine antibody isused as the first antibody. When ELISA assays are per-formed by using secondary antibodies conjugated tohorseradish peroxidase (HRP) and o-Phenylenediaminedihydrochloride as chromogen, levels of 3-nitrotyrosinedown to 200 fmol can be detected using 0.5mg/ml ofantinitrotyrosine polyclonal antibody.

In order to detect 3-nitrotyrosine present in biologicalfluids, and as nitrated proteins can be poorly represented inthese samples, two different approaches can be performed:(i) competitive ELISA, where the antibody binding to thestandard nitrated protein coated on the ELISA plate isinhibited by nitrated proteins contained in the sample [141],or (ii) a capture ELISA, in which the nitrated protein is firstcaptured by the anti-nitrotyrosine antibody coated on theplate and then alternatively detected by secondary antibod-ies directed against (i) the native protein, (ii) the nitroty-rosine epitope with either antinitrotyrosine antibodies raisedin a different animal species than that used for coating or thesame antibody used for coating conjugated to HRP on analternative detection system [140]. In the competitiveELISA, detection and quantitation is determined as a func-tion of the percentage of antibody binding. In the captureELISA, measurement requires comparison with standardnitrated proteins.

However, useful for detection purposes, results ob-tained with ELISA assays are not fully extrapolable inquantitative terms, and different factors may be takeninto account for their analysis. For example, the antibodyavidity and specificity may change related to the place ofthe 3-nitrotyrosine epitope in different proteins, the an-tibody may not detect some internal 3-nitrotyrosine moi-eties, and peroxynitrite may not nitrate the same tyrosineresidues in vitro and in vivo.

Western-blot and dot-blot assays

Western blot is a descriptive and semiquantitativemethod that can be used to detect immunoreactivity of

proteins carrying 3-nitrotyrosine residues with anti-nitrotyrosine antibodies including isolated proteins[57], and protein extracts from cells or tissue samples[138,139,142]. Immunoreactive proteins can be fur-ther characterized by their molecular weight, specificrecognition by a second western blot developed withantibodies directed against specific proteins or, pref-erably, after immunoprecipitation [143]. Additionally,western-blot assays can be used for competition andspecificity analysis of antinitrotyrosine antibodies[138,139] and for semiquantitative purposes in con-junction with digital imaging analysis.

As a more simple and faster alternative to westernblots to assess total nitrated proteins, irrespective ofspecific nitration profiles, and also for antibody charac-terization, dot-blot analysis can be performed. Detectionlimits must be defined in each particular system; forinstance, concentrations as low as 10 ng/ml of nitratedkeyhole limpet hemocyanin representing 430 pmol/ml3-nitrotyrosine, can be detected by dot blot using 0.1mg/ml of polyclonal antinitrotyrosine antibody associ-ated to a chemiluminescent detection system.

Recently, it has been shown that during sampleboiling under reducing conditions (b-mercaptoetha-nol) an important fraction of 3-nitrotyrosine can bereduced to 3-aminotyrosine, if hemeproteins arepresent [136]. This results in a significant decrease andeven disappearance of the 3-nitrotyrosine signal andmay explain why western-blot methodology has beenless used and successful than immunohistochemistryfor detection of nitrated proteins in complex biologicalsamples. In future studies using western blot analysisof nitrated proteins it will be critical to assess theinfluence of sample reduction on 3-nitrotyrosine yieldsto avoid artifactually low signals.

Immunohistochemistry and immunocytochemistry

Peroxynitrite-mediated nitration has been shown tooccur in different human and animal conditions by im-munohistochemistry, including human atherosclerosis[17], acute lung injury [14,15], intestinal inflammation[144], and physiological immune responses [139,145],among several others (for a review see [49]). Indeed,immunohistochemistry has been widely used to detect3-nitrotyrosine from tissues and the only one that permitslocalization of 3-nitrotyrosine residues in specific areas,with the maintenance of the histoarchitecture. Immuno-histochemistry has the advantage of permitting the rec-ognition of cells as well as subcellular structures orcompartments implicated in peroxynitrite and•NO-de-rived oxidants-induced nitration (Fig. 6). As peroxyni-trite can diffuse in tissues up to;20 mm from theformation site, a diffuse pattern of anti-nitrotyrosine

476 R. RADI et al.

binding is frequently observed; that should not be con-sidered as unspecific binding.

It is recommended to evaluate successive slices toperform the appropriate controls. Again, the technique isrelatively simple to perform and can be carried out withparaffin-embedded or frozen tissues, as has been recentlyreviewed elsewhere [146].

3-Nitrotyrosine in cultured cells can be detected byimmunofluorescence techniques [146]. This techniquehas been successfully applied, for instance, to the studyof cell nitration during degeneration of motoneurons inculture by either growth-factor deprivation [26] or theaction of intracellular Zn-deficient SODs [27].

Flow cytometry

Flow cytometry is an analytical method widelyused in cell biology that has not been significantlyapplied for the detection of nitrated proteins. Initialwork in our laboratory was performed in order toassess its validity to detect nitrated proteins in periph-eral blood mononuclear cells (PBMC). We associatedour polyclonal antinitrotyrosine antibodies [139] to

phycoerythrine (PE) or fluorescein isothyocyanate(FITC)-conjugated secondary antibody as fluorescentprobes and evaluated the fluorescence intensity ofintact (membrane-associated) versus permeabilizedPBMC (membrane- and intracellular-associated) afterperoxynitrite exposure (Fig. 7).

Flow Cytometry can be also performed by two- orthree-color analysis using different fluorochromes.Figure 8 shows a two-color cytometry analysis thatwas performed with anexin V conjugated to FITC toidentify apoptotic cells and with anti-nitrotyrosine an-tibodies associated to a PE-secondary antibody to de-tect cells bearingsurface-associated nitrated proteins.These results indicate that levels of protein nitration can besuccessfully detected in cells by flow cytometry using PE-or FITC-conjugated secondary antibodies as fluorescentprobes.

Protein hydrolysis and 3-nitrotyrosine quantitation

Tyrosine is a relatively abundant amino acid inproteins, with most mammalian proteins containing anaverage of;3.5 mol% tyrosine residues, having a

Fig. 6. Immunocytochemistry of protein-bound 3-nitrotyrosine. Nitrotyrosine staining of a human lymph node undergoing nonspecificimmune activation is shown. Human lymph nodes were obtained from surgical resections for colonic cancer. Lymph nodes were fixedin 10% formalin, paraffin embedded and 5mm sections were mounted in silanized microscope slides. Tissue sections were probed withantinitrotyrosine polyclonal antibodies (working dilution 50mg/ml) and developed with a secondary antibody coupled to biotin usinga streptavidin-peroxidase kit (Sigma) and diaminobenzidine as chromogen. Histological sections were counterstained with hemalum.Lymph nodes were free of metastasic invasion and showed mild to strong reactive follicular hyperplasia and sinusoidal histiocytosis.The sinuses of the medullar zone with histiocytosis show a diffuse staining (asterisk). A weak reactivity is seen at the interstitial level,between lymphocytes (arrow). The strongest 3-nitrotyrosine immunoreactivity is observed in macrophages (arrowhead), (1003).Insert: a macrophage with strong intracytoplasmic staining, (10003). See also [139].


range of 1– 8 mol%. Protein 3-nitrotyrosine can bequantitated after total protein hydrolysis of tissue orcell homogenate and biological fluids. Usual methodsused for hydrolysis have been either vapor phase acidhydrolysis or protease treatment (e.g., pronase) [92,94], with both methods resulting in high 3-nitroty-rosine recovery yields. During acid hydrolysis sub-stantial care must be taken to minimize artifactualnitration (6 M HCl, 120°C) due to contaminating NO2

2

and NO32 present in the sample. Extensive wash and/or

dialysis is critical to eliminate NOx2 before protein.

Also, addition of 1% phenol to the samples prior tohydrolysis helps to diagnose whether NO2

2 and NO32

contribute to artifactual nitration and also trap signif-icantly (but not totally) the reactive nitrogen interme-diates that may arise during sample workup and anal-ysis. The hydrolyzed samples are dried under vacuumand neutralized before initiating the detection of 3-ni-trotyrosine. Pronase treatment reduces the potentialinterference of undesirable secondary nitrations, butproteolysis of the sample may not be complete and

3-nitrotyrosine may not be recovered at pronase-resis-tant sites while, on the other hand, autoproteolysismay contribute to enhance tyrosine levels of the sam-ples. Finally, a third alternative has been just pub-lished [100] involving alkaline hydrolysis, which mayhelp to circumvent some of the problems observedwith the other two methods.

Detection and quantitation of 3-nitrotyrosine. The result-ing free 3-nitrotyrosine is separated from the rest of theamino acids using reverse phase-based HPLC methodol-ogy [92] and quantitated by one of the various methodsshown in Table 2. Initial studies detected 3-nitrotyrosineby amino acid analysis after derivatization of all aminoacids with phenyl isothiocyanate (PITC) [147], usingappropriate mobile phase gradient and the PITC-derivat-ized amino acids in hydrolyzates detected by opticalabsorption at 254 nm. 3-Nitrotyrosine elutes betweenleucine and phenyalanine and can be detected at levels inthe range of;4 pmol. Controls can be performed byprereducing the sample with dithionite, which results in

Fig. 7. Nitrotyrosine detection by flow cytometry. Nitrated proteins analyzed in normal peripheral blood mononuclear cells (PBMC)by monoparametric flow cytometry analysis. After peroxynitrite treatment to 53 106 cells/ml in isotonic phosphate buffer salinecontaining 100 mM potassium phosphate, pH 7.4, cells were either directly incubated with anti-nitrotyrosine antibodies followed bya secondary FITC-conjugated secondary antibody (nonpermeabilized cells) or fixed (permeabilized cells) prior to antibody incubationswith 2% paraformaldehyde in PBS and permeabilized with Triton X-100 at 0.5% in PBS.

478 R. RADI et al.

the conversion of 3-nitro to 3-aminotyrosine. Thismethod is useful to simultaneously follow changes indifferent amino acids but has important limitations: (i)low sensitivity, (ii) difficulty for obtaining accurate 3-ni-trotyrosine values, since 3-nitrotyrosine is usually repre-sented as a small shoulder in the chromatogram betweenpeaks of significantly more abundant vicinal elutingamino acids, and (iii) limited specificity in complexbiological samples.

The current trend is to assess 3-nitrotyrosine in conjunc-tion with tyrosine using more sensitive and specific meth-ods, such as those described before in this chapter. Quan-titation of 3-nitrotyrosine can be expressed on a molar basisrespect to total tyrosine or per mg protein, thus normalizingfor the variable protein content in different samples. How-ever, it is recommended to express the 3-nitrotyrosine valueas a function of total protein-bound tyrosine, as this ratiotakes into account not only protein content, but also effi-ciency of protein hydrolysis [92]. For instance, 3-nitroty-rosine in protein hydrolyzates of RAW 264.7 macrophagesunder control (no stimulation) condition using EC-detectiontechniques [89] yielded values of;one 3-nitrotyrosine (asN-acetyl-3-aminotyrosine) per 106 tyrosines (as N-acetyl-tyrosine) and increased;8-fold after treatment with zymo-san and IFN-g.

Free 3-nitrotyrosine has been detected and used as amarker of •NO-derived oxidant formation in biologicalfluids (i.e., plasma, synovial fluid, cerebrospinal fluid,bronchoalveolar lavage fluid) under basal [99,100,148]and a variety of pathological conditions [18,19,149,150].It is not defined at present whether free 3-nitrotyrosine in

biological fluids represents the nitration of free tyrosineor is a result of endogenous proteolysis of nitrated pro-teins.

Detection of nitrated peptides. Tyrosine nitration of pep-tides and proteins can be also studied by LC-MS tech-niques [151–153]. Nitration will increase the molecularweight of the control peptide by 45 Da; if more than onetyrosine residue is susceptible to nitration di-, tri-, orpoly-nitrations can be observed. Importantly, whenchanges other than nitration occur, increases in molecu-lar weights may not be a straightforward method foraffirming nitration. As an additional or alternative ap-proach, peptides can be followed by UV/VIS absorbancetaking advantage of the characteristic absorption proper-ties of 3-nitrotyrosine.

Partial proteolysis of nitrated proteins with trypsin orother proteases has been successfully used for peptidemapping of 3-nitrotyrosine residues in proteins [153].

Compartmentalization and determinants in proteinnitration

Peroxynitrite formed extracellularly could react witheither extra- or intracellular proteins, due to its capacityto diffuse and permeate membranes. Thus, potentially, atarget cell protein can be nitrated from peroxynitritearising from another source cell or compartment. Forinstance, inflammatory cells forming peroxynitrite maycause nitration in membrane or intracellular compart-

Fig. 8. Flow cytometry detection of nitrated and apoptotic cells. Two-color cytometric analysis of normal PBMC (53 106 cells/ml)exposed to authentic peroxynitrite (250mM) in isotonic phosphate buffer saline containing 100 mM potassium phosphate, pH 7.4. Afterperoxynitrite, cells were incubated with anti-nitrotyrosine antibodies followed by a secondary PE-conjugated secondary antibody. Afterthis, cells were further stained with Anexin V-FITC to identify apoptotic cells.


ments of target cells. On the other hand, intracellularlyformed peroxynitrite tends to react within the same or anadjacent cell compartment. Thus, the pattern of tissue orcell protein nitration observed in immunohistologicaland western blots studies can only suggest the site(s)where peroxynitrite had been formed. For instance, ac-tivated macrophages release most of peroxynitrite to theextracelullar milieu, but become nitrated themselves in-tracellularly [86,89, see also Fig. 9 inset]. In a tissue,nitration is preferentially localized around the areas inwhich peroxynitrite-producing cells are more abundant,which is consistent with the concept that peroxynitritecan diffuse one to two cell diameters; indeed, strong3-nitrotyrosine immunoreactivity in inflammatory pro-cesses is principally observed in the macrophage or neu-trophil-rich areas [15,139,154].

Peroxynitrite reactions in a compartment results innitration of a limited number of proteins. Selectivity ofprotein nitration depends on a series of factors [153],some of which are still unknown or partially defined, andinclude protein concentration, abundance and localiza-tion of tyrosine residues within a protein, presence oftransition metal centers close to nitration sites, and in-fluence of neighboring amino acids and secondary struc-ture.

Other modifications

Various other oxidative modifications may be used forperoxynitrite footprinting in biological systems. In pro-teins, peroxynitrite may lead to the formation of lowlevels of o-tyrosine and 3,39-dityrosine [96]. Formationof protein carbonyls in cells have been proposed to arisefrom peroxynitrite reactions with proteins [155,156], butperoxynitrite reactions with pure proteins do not signif-icantly yield carbonyls [157], thus cell formation ofprotein carbonyls may depend on secondary oxidativeprocesses. Other modifications that may contribute tounravel peroxynitrite which have been minimally ex-plored are: (i) lipid oxidation and formation of nitrosy-lated and nitrated lipid derivatives [11,158]. These com-pounds are unstable, and methods for improvement ofextraction techniques and stabilization are being devel-oped; they are of particular interest in LDL and athero-sclerotic plaques, where peroxynitrite may play a criticalprooxidant role. (ii) DNA strand breaks and base modi-fication. Peroxynitrite is able to cause DNA strand breaksand oxidation and nitration of DNA bases such as gua-nine to oxo- and 8-nitroguanine [159–161]. The forma-tion of these modified bases by the actions of peroxyni-trite in vivo awaits future research. (iii) modifications ofubiquitous low molecular weight compounds, may alsoserve as biomarkers of peroxynitrite formation and reac-tions. Some of these include recently characterized mod-

ifications, including the formation on nitrated-derivativesof glucose [162], the potential formation of nitratedthiols (such as nitroglutathione) [163] and decompositionproducts of uric acid reaction with peroxynitrite [164,165]. It remains to be established whether these modifi-cations are quantitatively sufficient and measurable to beused as potential footprints of peroxynitrite.

It is important to recognize that all of the modifica-tions described in this section are not a consequence ofdirect bimolecular reactions of peroxynitrite with thetarget molecules, but rather of free radical intermediatesarising from secondary reactions, including CO3

•2 radi-cals, and•NO2. Moreover, in complex biological sys-tems, as yet undescribed modifications may arise, prod-ucts of free radical processes initiated by peroxynitrite.

Importantly, in spite of all the potential oxidativemodifications indicated, there is no available singlechemical modification 100% specific to peroxynitrite.Thus, unraveling peroxynitrite with the highest degree ofcertainty may require the detection of a panel of charac-teristic footprints, such as, for example, the simultaneousdetection of 3-nitrotyrosine and 3,39-dityrosine in bio-logical samples. At the same time, the lack of detectionof other biomarkers of oxidative damage such as o-tyrosine and chlorotyrosine, which most likely reflectreactions of•OH and MPO-dependent reactions, respec-tively, will provide even stronger evidence of peroxyni-trite-mediated oxidations.

PHARMACOLOGY TO UNRAVEL PEROXYNITRITE

Due to the lack of techniques and chemical modifi-cations completely specific for peroxynitrite, additionalexperimental evidence is obtained by the use of: (i) drugsthat decrease•NO and O2

•2 levels, (ii) treatments thatmodify the concentration of biomolecules that criticallyinfluence peroxynitrite reactivity, (iii) peroxynitrite de-composition catalysts and scavengers, and (iv) com-pounds that interfere with alternative oxidation/nitrationpathways. Finally, in some models it may be of interestto promote or increase peroxynitrite formation/detectionby pharmacological modulation.

Drugs that decrease•NO and O2•2 levels—inhibition of

•NO and O2•2 formation

Nitric oxide. The inhibition of•NO formation is obtainedwith the use of a wide variety of NOS inhibitors ofvarious degrees of selectivity and potency over the threedifferent isoforms of NOS (nNOS, iNOS, and eNOS)[166,167]. Also, NOS-deficient cells and knockout mod-els of one or more than one isoforms of NOS [21,168,169] are potential tools to define the contribution of•NOand•NO-derived oxidants.

480 R. RADI et al.

Superoxide. Pharmacological inhibition of O2•2 forma-

tion is more complex than that of•NO, because thesources of O2

•2 are variable depending on cell types andmetabolic status, and also because typically more thanone cellular source contributes to overall cell O2

•2 for-mation. Thus, inhibition of O2

•2 formation requires aprevious definition of the main cell or tissue sources ofO2

•2 under the specific conditions of the study. In themore straightforward case of O2

•2 formation from acti-vated macrophages, neutrophils, and other cells of theimmune system via NADPH oxidase activation, inhibi-tion of this enzyme may prove effective to inhibit per-oxynitrite formation. However, there is a relatively shortgroup of NADPH oxidase inhibitors, most of which arerelative unspecific or their mechanism of action largelyunknown. For instance, the commonly used inhibitor ofNADPH oxidase, diphenyliodonium [170], is a flavopro-tein inhibitor that can also influence NOS activity, de-pending on the concentrations used [171]. In systems inwhich O2

•2 is principally formed due to the catalyticaction of xanthine oxidase, enzyme inhibitors such asallopurinol or oxypurinol may be useful [172,173].NADPH oxidase deficient cells and knockout models arealso available, and may be of great help to define thecontribution of O2

•2 (and therefore peroxynitrite) to•NO-mediated oxidations. Interestingly, double knockoutmodels for NADPH oxidase and iNOS have been devel-oped recently [168,169,174].

Enhanced consumption of•NO and O2•2

Nitric oxide. The consumption of•NO may be promotedby the use of•NO scavengers, such as derivatives ofphenyl-4,4,5,5,-tetramethylimidazoline-1-oxyl 3-oxide(PTIO) [25]. PTIO reacts stoichometrically with•NO toyield •NO2 and 2-phenyl-4,4,5,5,-tetramethylimidazo-line-1-oxyl (PTI), as seen in Eqn. 2:

PTIO 1 •NO3 •NO2 1 PTI (2)

This reaction rapidly consumes•NO and may partiallycompete with O2

•2.

Superoxide. Enhanced elimination of O2•2 is obtained by

overexpression of SOD (cytosolic or mitochondrial)[175–177] or supplementation with cell-permeable SODmimics (e.g., manganese-complexes) [178,179], or withliposomal entrapped SOD [180]; these manipulationsresult in fast consumption of O2

•2 by pathways that donot yield peroxynitrite.

It is important to appreciate that since the combinationreaction between•NO and O2

•2 is diffusion-controlled,only those traps that react fast enough with either•NO or

O2•2 and/or can approach relatively high concentrations

in the sites where peroxynitrite is being produced, will beable to prevent peroxynitrite formation.

Endogenous components that modulate peroxynitritereactivity

Modulation of cellular thiol levels. Glutathione and thi-ols are critical endogenous intracellular antioxidantsagainst peroxynitrite and derived species, and thereforechanges in thiol and glutathione levels will have a directinfluence on peroxynitrite detection and biological out-come. Cellular glutathione depletion by treatment withbuthionine sulfoximine has been shown to increase thedetection (e.g., intracellular nitration) and biological ef-fects of peroxynitrite [142]. Enhancement of intracellularthiol levels by supplementation with thiol or thiolestersmay also decrease the detection and effects of peroxyni-trite.

Modulation of carbon dioxide levels. The presence ofcarbon dioxide significantly affects peroxynitrite reactiv-ity and diffusion. Thus, variation of carbon dioxide lev-els in biochemical systems, cells, and tissues could havea significant effect of detection reactions (e.g., carbondioxide inhibits oxidations but promotes nitrations).

Use of peroxynitrite decomposition catalysts andscavengers

Peroxynitrite or its secondary products can be readilydecomposed by the use of decomposition catalysts andscavengers, aimed to inhibit biological oxidations and/ornitrations and therefore protect from peroxynitrite-in-duced oxidative damage. However, to be effective incomplex biological systems, both peroxynitrite decom-position catalysts and scavengers must be able to out-compete peroxynitrite target molecule reactions. There-fore, it is important to appreciate how much and howefficiently these molecules can be delivered to the siteswhere peroxynitrite is being formed and/or react wheninterpreting protection and detection data.

Among the first group of compounds are includedthose that catalyze peroxynitrite isomerization (e.g., ironporphyrins) [181] and those that catalytically reduceperoxynitrite by redox cycles that consume endogenousreductants such as thiols, ascorbate, or urate (e.g., man-ganese-porphyrins, selenocompounds) [182–187] (Fig.9). Various organo-transition metal complexes have beenlately used as peroxynitrite decomposition catalysts inbiochemical, cell, and animal models and shown to pro-tect against the effects of exogenous and endogenousperoxynitrite [188,189] and inhibit protein nitration [26,


139]. Some compounds such as the manganese porphy-rins (Mn-tmpyp) can potentially have a dual action byacting both as SOD mimics as well as peroxynitritedecomposition catalysts. Still, although promising, cur-rent compounds could have secondary effects [190] andshort- and long-term toxicity in cell and animal modelsrequires further investigation.

There are some caveats as regards the use of per-oxynitrite decomposition catalysts as a way to confirmthe participation of peroxynitrite by using probe oxida-tion or footprinting, because these compounds can un-dergo secondary nondesirable redox reactions under par-ticular conditions. For instance, during the catalyticdecomposition of peroxynitrite by manganese-porphy-rins, cellular reductants such as glutathione are con-sumed. Therefore, the inhibitory effects on peroxynitritedetection will be highly dependent on the glutathione aswell as other reductants concentrations in the cell. More-over, the compounds may not be inhibitory of peroxyni-trite-mediated processes or may even enhance oxidativemodifications under conditions of cellular glutathionedepletion, since the oxo-manganese complexes formedduring the catalytic cycle may themselves promote oxi-dation and/or nitration of cellular targets [191].

The scavengers include a

Forum: Oxidative Stress Status - CNCcnc.cj.uc.pt/BEB/private/pdfs/ToxNitr/SM 3.pdfThe biological...

Documents

Transcript of Forum: Oxidative Stress Status - CNCcnc.cj.uc.pt/BEB/private/pdfs/ToxNitr/SM 3.pdfThe biological...