Peroxynitrite-induced luminol chemiluminescence by iasiatube


									Biochem. J. (1993) 290, 51-57 (Printed in Great Britain)                                                                                          51

Peroxynitrite-induced luminol chemiluminescence
Rafael RADI,*ttThomas P. COSGROVE,* Joseph S. BECKMAN* and Bruce A. FREEMAN*t§
Departments of * Anesthesiology, t Biochemistry and § Pediatrics, University of Alabama at Birmingham, Birmingham, AL 35233-6810, U.S.A.

Vascular endothelial cells, smooth muscle cells, macrophages,                       plex with luminol, yielding luminol radical and 02'-. Luminol
neutrophils, Kupffer cells and other diverse cell types generate                    radical reacts with 02'- to form the unstable luminol
superoxide (02'-) and nitric oxide ('NO), which can react to                        endoperoxide, which follows the light-emitting pathway.
form the potent oxidant peroxynitrite anion (ONOO-).                                Neither 'NO nor O2;- alone were capable of directly inducing
Peroxynitrite reacted with luminol to yield chemiluminescence                       significant luminol chemiluminescence in our assay systems.
which was greatly enhanced by bicarbonate. The quantum                              These results suggest that ONOO- can be a critical unrecog-
chemiluminescence yield of the ONOO- reaction with luminol                          nized mediator of cell-derived luminol chemiluminescence re-
in bicarbonate was approx. 10-3. Chemiluminescence was                              ported in previous studies. In addition, it is shown that bicarb-
superoxide dismutase-inhibitable, indicating that 02- was a key                     onate can participate in secondary oxidation reactions after
intermediate for chemiexcitation. 02'- appears to be formed                         reacting with ONOO-.
secondarily to the reaction of a bicarbonate-peroxynitrite com-

INTRODUCTION                                                                        protein thiols [14], deoxyribose [11] and membrane phospholipids
Vascular endothelial cells, smooth muscle cells, macrophages,                       [15]. In addition to its role in oxidation reactions, ONOO- can
neutrophils, Kupffer cells and a growing list of other cell types                   also nitrate free or protein-associated tyrosine [16].
generate superoxide (02'-) [1] and nitric oxide ('NO) [2]. Different                   Luminol chemiluminescence has been widely used to detect the
mediators stimulate the simultaneous production of 'NO and                          production of reactive oxygen species (i.e. 02-- H202 and 'OH)
 02--' including interferon y [3], calcium ionophores [2,4],                        from enzyme, cell and organ systems [17-19] and has been useful
lipopolysaccharide [5,6] and phorbol esters [4,7]. Thus con-                        for examining the kinetics and reaction mechanisms of oxygen-
comitant 'NO and 02'- generation may be enhanced in a variety                       radical processes [17,20]. In order to yield light, luminol has
of pathophysiological situations such as ischaemia-reperfusion,                     to undergo a two-electron oxidation and form an unstable
acute inflammatory processes, atherosclerosis, bacterial infec-                     endoperoxide. This luminol endoperoxide decomposes to an
tions and sepsis.                                                                   excited state, 3-aminophthalic acid, which relaxes to the ground
   Both 'NO and 02- are free-radical species that rapidly react                     state by emitting photons [18,21]. In most cases of luminol
with each other in aqueous solution at pH 7.4, yielding per-                        chemiexcitation in biological systems, 02 -is a key intermediate
oxynitrite anion (ONOO-) (k2 = 3.7 x 107 M-1 s-1) [8]. Several                      [22,23], but alternative pathways ofchemiexcitation not requiring
observations support the in vivo formation of ONOO-. First, the                     02- have been described [17,24].
half-life of endothelial-derived 'NO is doubled in the presence of                     Luminol chemiluminescence is observed during the respiratory
superoxide dismutase (SOD), suggesting that 02- is involved in                      burst of macrophages and neutrophils, which has been commonly
its degradation [9]. Second, inhibitors of 'NO synthase increase                    attributed to the production of 02- and H202 [18]. Nevertheless,
detectable 02-- release from macrophages [10]. Third, decompo-                      there are instances where more potent oxidation reactions are
sition of the syndominine, SIN-1, to products including 'NO and                     needed to explain observed chemiluminescence yields [25]. For
02'- yields a species with 'OH-like reactivity, inferring the                       example, myeloperoxidase liberated from macrophage granules
intermediate formation of ONOO- [11,12]. Finally, ONOO- has                         greatly enhances the reactivity of H202 towards luminol by
been directly detected as a product of phorbol myristate acetate-                   forming an oxohaem oxidant species [18]. Also, 'NO-derived
activated rat lung alveolar macrophages, which produce                              intermediates contribute to luminol chemiluminescence in
100 pmol of ONOO-/min per 106 cells [13].                                           phorbol ester-activated Kupffer cells and may react via ONOO-
   Peroxynitrite (PKa = 6.8 [14]) is an unstable species at                         [7].
physiological pH (t1 < 1 s), protonating to peroxynitrous acid                         Herein we report that ONOO-, a newly described biological
(ONOOH) which spontaneously decomposes to 'NO2 and 'OH                              oxidant of emerging significance [26-28] induced SOD-
in 20-30 % yield:                                                                   inhibitable luminol chemiluminescence. Furthermore, it is shown
                                                                                    that bicarbonate can participate in secondary oxidative reactions
                                                                                    after reacting with ONOO-.
            *NO+02'-           ONOO-          ONOOH        -*'NO2+*OH      (1)
The remaining ONOOH will directly isomerize to nitrate (NO3-,
[1 1]).                                                                             Materials
   Peroxynitrite may be an important mediator of free-radical-                      Cysteine, cystine, uric acid, mannitol, dimethyl sulphoxide
dependent toxicity [1 1-14] because of its strong oxidizing proper-                 (DMSO), horseradish peroxidase type VI and tetranitromethane
ties towards different biomolecules, including protein and non-                     were obtained from Sigma. Potassium superoxide, cis-dicyclo-

  Abbreviations used: SOD, superoxide dismutase; DMSO, dimethyl sulphoxide.
  t Permanent address: Department of Biochemistry, Faculty of Medicine, University of the Republic, Montevideo, Uruguay CP11800.
  1 To whom correspondence should be addressed.
52          R. Radi and others

hexano- 1 8-crown-6 and tetrafluoroborate nitronium were from        bicarbonate buffer at pH 10.5 (Figure 1). The proportion of
Aldrich. H202 and luminol (5-amino-2,3-dihydro-1 ,4-phthal-          chemiluminescence yield in phosphate buffer, with respect to
azinedione) were obtained from Fluka and Cu/Zn SOD was               bicarbonate buffer, was increased by lowering pH. At pH 7.5,
generously provided by Grunenthal GmBH.                              the luminol chemiluminescence yield of ONOO- reaction in
  Peroxynitrite was synthesized in a quenched-flow reactor as        phosphate buffer became 15 % of that in bicarbonate buffer. The
previously described [11,14,15]. Fresh solutions of potassium        addition of H202 (10, 50 and 200,#M) to reaction mixtures
superoxide (0.1 M) were prepared daily in DMSO and 18-crown-         containing 400 ,tM luminol and 100 ,tM ONOO- in either
6 as described elsewhere [29]. A 1.7 mM 'NO solution was             bicarbonate or phosphate buffers at pH 10.5 did not have a
prepared by extensive bubbling of 'NO gas in anaerobic deionized     significant effect on chemiluminescence yield or the time course
water contained in a gas sampling tube.                              of light emission (results not shown).
                                                                       The quantum yield of ONOO--induced chemiluminescence
                                                                     (quanta/molecules of ONOO- consumed) was determined at
Inactivation of Cu/Zn SOD
SOD inactivation by H202 [30] was performed by incubating
2.0 ,ug/ml SOD (2100 units/mg) with 1.0 mM H202 in 50 mM                         12                                                              120
sodium pyrophosphate, pH 10.0, at 25 °C for 30 min, with
residual H202 removed by Sephadex G-25 chromatography.                                                                                                 0)
SOD activity was determined by measuring the inhibition of                       10
cytochrome c3+ reduction by xanthine plus xanthine oxidase [31].            0
                                                                            I                                                                          E
SOD was approximately 86 % inactivated by this method.                      Z     8-                                                                   ._

Chemiluminescence measurement                                               C

Chemiluminescence studies were performed in a SLM DMX-
1000 fluorimeter by closing the excitation window and maximally              X 4-                                                                      U)
                                                                            x                                                                          0)
opening the emission slit width to 16 nm. Reactions were initiated          0
by injection of ONOO- directly to cuvettes with continuous
stirring through a syringe adapted to the instrument. Light                       2
emission was simultaneously recorded by computer interface.
Reported chemiluminescence is the integrated light emission
from a 5 s time interval, unless otherwise stated. When studying                  0-
                                                                                  300                  400           50(
the effect of different oxygen tensions on chemiluminescence                                            Wavelength (nm)
yield, reactions were performed with continuous oxygen, air or       Figure 1 Chemiluminescence spectra for luminoiONOO- reactions
nitrogen bubbling at 1-2 ml/min in a 3 ml cuvette volume.
                                                                     Peroxynitrite (100 ,uM) was added to reaction mixtures containing 400 ,uM luminol in 50 mM
                                                                     bicarbonate (O or phosphate (0), pH 10.5, at 25 °C. Maximal light intensities obtained at
                                                                     different emission wavelengths are reported. Results are means+S.E.M.
Spectrophotometric analysis
Absorbance measurements and spectra were performed with
a Gilford Response spectrophotometer. Luminol oxidation                          25
was measured by absorbance decrease at 350 nm (e =
7200 M-1' cm-') and 300 nm (e = 6800 M-1 * cm-'). For ONOO-
concentrations utilized in this paper, there was no significant                  20-
interference of the ONOO- decomposition products N03-
and NO2- on luminol oxidation determination, because of
the extremely low absorption coefficients of these species in                    O15-
the 300-350 nm region compared with luminol (NO2-, £350 =
23 M-lcm-' and 6300 = 10 M-l cm-l; NO3-, £350 = 0 and
6300 = 9 M-l cm-').
   Data reported herein are the average of three independent                 x 10
determinations performed in a one-day experiment. Each ex-
periment was repeated on a minimum of 4 different days and
similar results were obtained with a maximum 10 % variation.                      5

RESULTS                                                                           0               40           80               120            160
Luminol chemiluminescence spectra and quantum yield                                                          Time (s)
Addition of peroxynitrite stimulated luminol chemiluminescence       Figure 2 Luminol chemiluminescence as a function of ONOO- concentrafton
in both bicarbonate and phosphate buffer. Emission spectra were      Reactions were started by addition of 10 ,uM (A), 25 /uM (A), 50 ,uM (0) or 100 ,uM (0)
similar for both buffers, with Aem maximum = 425 nm (Figure 1),      ONOO- to reaction mixtures containing 400 ,sM luminol in 50 mM bicarbonate, pH 10.5, at
which was used as the Aem for further studies. Maximal light-        25 °C. kd values were 0.020 s-1, 0.024 s-1, 0.022 s-1 and 0.021 s-1 for 10, 25, 50 and
emission intensity in phosphate buffer was 0.3-0.5 % of that in      100 ,tM ONOO- respectively. Results are means+ S.E.M.
                                                                                                                         Peroxynitrite and chemiluminescence                    53

          500                                                                                     and was quantified as the first-order rate constant kd [17,20].
                                                                                                  Chemiluminescence intensity was proportional to ONOO- con-
                                                                                                  centration, with the calculated kd independent of ONOO-. Light-
                                                                                                  emission decay in bicarbonate buffer was zero order with respect
          400-                                                                                    to luminol in concentrations up to 1 mM. No further in-
                                                                                                  crease in 100 ,uM ONOO--induced chemiluminescence yield was
                                                                                                  obtained with concentrations of luminol greater than 400 ,uM.
          300-                                                                                    Alternatively, chemiluminescence was strongly influenced by pH,
                                                                                                  with kd increasing at lower pH (Figure 3). Addition of a second
                                                                                                  portion of ONOO- to reaction mixtures after chemiluminescence
      .2' 200-                                                                                    ceased resulted in similar chemiluminescence yields. Thus
                                                                                                  quenching by ONOO- and luminol reaction byproducts did not
                                                                                                  artifactually influence the ONOO--induced luminol chemilumi-
                                                                                                     Chemiluminescence decay rates and peak intensities depended
                                                                                                  on bicarbonate concentration, with kd directly proportional to
                                                                                                  bicarbonate (Figure 4a). Total chemiluminescence yields (area
            0                     10                  20                  30                      below the curve) were similar for all bicarbonate concentrations
                                         Time (s)                                                 at a particular pH (Figure 4b). In contrast, the potassium
Figure 3 Luminol chemiluminescence as a function of pH
                                                                                                  phosphate concentration of phosphate-buffered systems did not
                                                                                                  influence the rate of luminol chemiluminescence decay. No
Reactions were started by addition of 100 uM ONOO- to a reaction mixture containing 400 IM        chemiluminescence was observed after addition of tetranitro-
luminol and 50 mM bicarbonate, pH 10.0 (-), 9.5 (A), 8.5 (0) and 7.5 (+) at 25 °C. kd             methane (200 ,uM), nitronium tetrafluoroborate (200 ,uM),
values were 0.17 s-, 0.45 s-1, 1.11 s-1 and 1.20 s-1 respectively. Chemiluminescence was          nitrate (100 ,uM), nitrite (100 1sM), nitric oxide (up to 200 ,uM) or
integrated every 0.2 s.                                                                           potassium superoxide (up to 2.5 mM) to 400 ,M luminol in
                                                                                                  50 mM bicarbonate, pH 10.5, at 25 'C. Luminol chemilumi-
                                                                                                  nescence from O2;- only occurs when 02-- reacts with luminol
pH 10.5 with luminol in excess of ONOO- (400,uM luminol,                                          radical, the formation of which requires prior reaction with
100 ,M ONOO-). Quantum yields were approx. 10' in bi-                                             stronger oxidants.
carbonate buffer and < 10-s in phosphate buffer. A secondary                                         Net luminol oxidation was measured spectrophotometrically
standard of persulphate/H202 was used as our reference system                                     at 350 and 305 nm and was ONOO- concentration-dependent
[32].                                                                                             (Table 1), with about 5-8 and 3-6 mol of ONOO- consumed per
                                                                                                  mol of oxidized luminol at pH 7.5 and 10.5 respectively. Luminol
                                                                                                  oxidation yields were greater at pH 10.5 than at pH 7.5, were
Time course of luminol chemlluminescence                                                          greater at lower ONOO- concentrations and were similar for
Luminol chemiluminescence induced by ONOO- reached peak                                           bicarbonate and phosphate buffers. There was no effect of
intensity by 5 s at pH 10.5 and was followed by an exponential                                    bicarbonate concentration (50-400 mM) on oxidation yield
decay (Figure 2). The emission decay followed first-order kinetics                                (results not shown).


                          21                                                                 cn


                          0                                                          500                                   80
                                                 [HCO3-1 (mM)                                                            Time (s)

Figure 4 kd as a function of bicarbonate concentration
(a) Assay conditions were: 100 ,uM ONOO-, 400 ,uM luminol and 50-400 mM bicarbonate at pH 10.5 (A) or 9.5 (0) at 25 °C. kd was determined as the tirst-order constant of exponential
decay for each buffer concentration. (b) Chemiluminescence records were obtained in 50 mM (0) and 300 mM (A\) bicarbonate, pH 10.5. Results are means+S.E.M.
54               R. Radi and others

Table 1 Luminol oxidation as a function of ONOO- concentration                                 Table 2 Effect of antioxidants on luminol oxidation
Peroxynitrite (1 00-400 ,uM) was added to 400 ,uM luminol in 50 mM bicarbonate, pH 10.5        Assay conditions were as described in Figure 6. Incubations were performed for 10 min at
or 7.5, at 25 OC and incubated for 10 min. Data represent means+ S.D. (n = 4).                 25 °C. Data represent means+ S.D. (n = 3). Superscripts denote signiticance of difference
                                                                                               from control (a) and ONOO- (b)-treated (no addition) condition after ANOVA and analysis by
                                         Luminol (uM)                                          Duncan's multiple-range test (P < 0.05).

                 [ONOO-] (,uM)           pH 10.5                   pH 7.5                                         Condition                                 Luminol    (IuM)
                    0                    400 + 5                   400 + 5                                        Control                                   396 + 2
                  100                    366 + 8                   379 + 6                                        + ONOO-
                  250                    340 + 8                   364 + 5                                           No addition                            371 + 2"
                  500                    307 + 7                   333 +12                                            +100 mM mannnitol                     372 + 3a
                  750                    274 + 5                   305 + 20                                           +100 mM DMSO                          372 + 3a
                 1000                    243 + 4                   274 + 20                                           +3 mM cystine                         371 + 4
                                                                                                                      +3 mM cysteine                        395 +3
                                                                                                                      +1 mM urate                           397 + 4
                                                                                                                      + 2.3 units/ml SOD                    378 + 4a
                                                                                                                      + 23.0 units/ml SOD                   386 + 3a


                                             10                                                             20
       15                                -

                                              0       25      50      75      100
      o10                                                  Time (s)                                         0)~~~
  x                                                                                                    C10

                                                                                                      o                              0
                                                                                                             5                              0

                                                                                                                       A           A~U                  AU               A
        0                  25                 50                75                  100                      0                25            50              75                 100
                                          Time (s)                                                                                       Time (s)
Figure 5 Effect of Cu/Zn SOD and oxygen on luminol chemiluminescence                           Figure 6 Effect of antioxidants on luminol chemiluminescence
Assay conditions were: 100,uM ONOO-, 400 uM luminol with no SOD (O), 0.23 units/ml             Assay conditions were: 100 ,tM ONOO-, 400 ,#M luminol with no additions (M), 100 mM
native SOD, (A, 3.75 nM), 2.3 units/ml native SOD (+, 37.5 nM), 0.27 units/ml inactivated      DMSO (0), 100 mM mannitol (C]), 3 mM cysteine (0), 1 mM urate (A) and 0.375 PuM SOD
SOD (0, protein equivalent 37.5 nM) and 2.7 units/ml inactivated SOD (El, protein equivalent   in 50 mM bicarbonate, pH 10.5, at 25 °C. Where indicated, results are means+ S.E.M.
2.7 units/ml) in 50 mM bicarbonate, pH 10.5, at 25 OC. The inset shows the effect of oxygen
on chemiluminescence. Assay conditions were as above with no SOD present. Reactions were
saturated with air (@), nitrogen (V) or oxygen (O).
                                                                                               duration by about 10-20 %, whereas nitrogen saturation only
                                                                                               slightly inhibited chemiluminescence (Figure 5, inset). Similar
                                                                                               extents of SOD inhibition of ONOO--induced luminol chemi-
Effect of Cu/Zn SOD                                                                            luminescence occurred in oxygen, air or nitrogen-saturated
SOD caused a dose-dependent inhibition of luminol chemi-
luminescence and a small increase in kd (Figure 5). When H202-
inactivated SOD replaced native SOD, the inhibition of light                                   Effect of antioxidants
emission corresponded to that expected for the residual SOD                                    Uric acid (1 mM) and cysteine (3.0 mM) completely inhibited
activity remaining in the inactivated enzyme preparation (Figure                               both luminol chemiluminescence and luminol oxidation (Figure
5). A significantly greater SOD activity than that used to inhibit                             6, Table 2). The OH scavengers mannitol (100 mM) and DMSO
chemiluminescence only partially inhibited luminol oxidation                                   (100 mM), and the disulphide cystine (3.0 mM), did not affect
(Table 2). With the experimental conditions of Figure 5, cupric                                luminol chemiluminescence and luminol oxidation (Table 2).
sulphate (1 and 20 1tM) inhibited luminol chemiluminescence by
70% and 95 % respectively. The copper chelate Cu/EDTA
(1: 1.1, mol/mol; up to 200 ,#M) did not inhibit light emission.                               Secondary products from ONOO-ireactlon with luminol
   Reactions conducted in 100 % oxygen-saturated bicarbonate                                   After luminol reacted with ONOO-, a pronounced 425 nm yellow
buffer, pH 10.5, increased both chemiluminescence yield and                                    absorbance was found in samples which progressively disap-
                                                                                                                   Peroxynitrite and chemiluminescence          55

                                                                                             there was no direct source of H202 in our reaction mixtures.
                                                                                             Exogenous addition of H202 had no significant effect on light
                                                                                             yield as well. Formation of a nitrated luminol endoperoxide by
                                                                                             ONOO- adduct formation with the diazaquinone, followed by
                                                                                             internal rearrangement, also does not account for light emission
                                                                                             because nitrophthalate is not chemiluminescent (G. Merenyi and
                                                                                             J. Lind, personal communication).
                                                                                                The inhibitory effects of SOD and free Cu2+ on chemilumi-
   A2                                                                                        nescence imply the participation of 02- in chemiexcitation. This
                                                                                             concept is reinforced by the observation that Cu/EDTA, which
                                                                                             lacks SOD activity [35], did not inhibit chemiluminescence.
                                                                                             Different mechanisms could operate to generate 02O- during
                                                                                             ONOO- reaction with luminol. 02- could have been produced
                                                                                             by a luminol radical-dependent univalent reduction of molecular
                                                                                             oxygen [21]. This mechanism was probably not the main source
                                                                                             of 02@- since saturation with oxygen or nitrogen had a marginal
                                     380       420                                           effect on chemiluminescence yield (Figure 5, inset). Moreover,
                                    Wavelength (nm)                                          SOD inhibited light emission in both aerobic and anaerobic
Figure 7 Absorption spectra of luminol after oxidation by ONOO-                              conditions. The direct formation of 02-- (plus 'NO) from
                                                                                             ONOO- did not occur for thermodynamic and kinetic
Luminol (400 uM) reacted with 100 ,uM ONOO- in 50 mM bicarbonate, pH 10.5 at 25 OC.          considerations. The AGO' for formation of ONOO- from O2.-
Spectra were recorded immediately before (----) and 3 min after (        ) ONOO- addition.   and 'NO is about -92 kJ/mol (-22 kcal/mol), thus Keq =
A decrease in A350 and the appearance of a peak in the 425 nm region was observed. Whereas
A350 did not change with time, the 425 nm absorbance was transient and decayed (inset).      5 x 101' M-l at 25 °C [25]. With the k2 for formation of ONOO-
                                                                                             -being at least 3.4 x 107 M-1 * s-' [8], it can be estimated that the
                                                                                             reverse re.action (formation of O2 - from ONOO-) will proceed
                                                                                             slowly at 10-8 s-1. Thus O2- must be generated after direct
peared over 15-25 min at both pH 10.5 and 7.5 in either 50 mM                                reaction of ONOO- with luminol and by a bicarbonate-
bicarbonate or phosphate buffers (Figure 7 and inset). This                                  stimulated mechanism (Figures 1 and 4). When bicarbonate was
425 nm absorbance was pH insensitive. Addition of uric acid                                  absent, luminol could still be oxidized by ONOO- (Table 1), but
(1 mM) and cysteine (3 mM) but not DMSO (100 mM) or                                          the presence of bicarbonate led the oxidation process to a light-
mannitol (100 mM) to ONOO--luminol reactions inhibited the                                   emitting pathway, presumably by favouring O2- production,
appearance of the 425 nm absorbance. Luminol reaction with                                   thus supporting a greater chemiluminescence quantum yield. In
tetranitromethane (100 ,uM) or much higher concentrations of                                 the presence of high bicarbonate concentrations, the rate of
nitronium tetrafluoroborate (5 mM) resulted in a stable yellow                               light decay was independent of both ONOO- and luminol
product with a broad absorbance at 400-500 nm rather than the                                concentration and was pseudo-first-order with respect to
sharper 425 nm peak after luminol reaction with ONOO-.                                       bicarbonate. Thus, the formation of the key oxidant might
Oxidation of luminol with horseradish peroxidase plus H202 did                               precede the reaction of ONOO- with luminol. A mechanism
not generate this 425 nm-absorbing species (results not shown).                              consistent with these observations can involve a peroxynitrite-
                                                                                             bicarbonate intermediate:
DISCUSSION                                                                                         ONOO- + HC03- + H+ -÷ ONOOC(O)0- + H20                      (4)
Peroxynitrite-induced luminol chemiluminescence was greatly                                        ONOOC()0-+ LH- - L- +02--+ NO- + CO2 + H+ (5)
enhanced by HCO3- (Figure 1). Spectral analysis of luminol
chemiluminescence (Figure 1) indicated that excited amino-                                          L--+ 02 -- light + aminophthalate                           (6)
phthalate was the emitting species [33]. Since formation of                                  With phosphate-buffered systems, a similar but minor reaction
aminophthalate depends on decomposition of luminol endo-                                     could be responsible for some light generation, with the principal
peroxide, ONOO- reaction with luminol must yield this unstable                               light emission being due to the reduction of ONOO- to NO2 and
intermediate.                                                                                 H20 by luminol. The hallmark of eqns. (4)8(6) is that the
  Peroxynitrite is an effective one-electron oxidant (E'ONOO/-NO2) =                          reduction of ONOO- can yield 02n-' with the efficiency of this
+ 1.4 V [34]), thus oxidation of luminol by ONOO- is thermo-                                  reaction increasing after ONOO- reaction with bicarbonate.
dynamically feasible. The first oxidation by ONOO- or its                                       Other mechanisms might also participate in ONOO--depen-
conjugate acid would lead to the formation of luminol radical                                 dent bicarbonate-mediated luminol-oxidation reactions. For
(L*-) as follows:                                                                             example, formation of a bicarbonate radical from ONCO-
                                                                                              reaction may occur, with the EL'(HC03/HC03-) being + 1.5 V [36].
          ONOO- + LH- -*'NO2 + L- + OH-                           (2)                         ONOO-/'NO2 and HCO3*/HCO3- couples have similar one-
If NO2 is in close proximity to luminol radical, an immediate                                 electron standard redox potentials. Thus, under conditions
second oxidation will take place to yield a diazaquinone (L), with                            where ONOO- reacts with excess HCO3-, one-electron oxidation
the overall process a two-electron oxidation according to:                                    of bicarbonate by ONOO- is thermodynamically favourable
          ONOO-+ LH- -- NO2-+OH-+L                                (3)                        according to:
Alternatively, luminol radical can be oxidized to the diazaquinone                                     ONOO-+HCO3-+ H+-HCO3             NO2 + OH-
                                                                                                                                           +             (7)
by direct reaction with a second molecule of ONOO-. If the
diazaquinone reacts with H202, the luminol endoperoxide in-                                  Bicarbonate radical is known to oxidize luminol [37] and other
termediate is formed [21]. A diazaquinone-H202 reaction was                                  aromatic and heterocyclic molecules with second-order rate
not responsible for ONOO--induced chemiluminescence, since                                   constants ranging from 5 x 105 to 5 x I07 M-1 s-s         [38].
56          R. Radi and others

Alternatively, ONOO- may peroxidize bicarbonate to                    potassium superoxide and 200 ,uM 'NO. This          concurs   with
peroxybicarbonate [39], another strong oxidizing species. Thus,       previous observations of poor reactivity between O2-- and
there are several potential mechanisms by which bicarbonate           luminol [38]. When 02'- participates in luminol chemilumi-
could influence the overall reactivity of ONOO- in both model         nescence reactions, it does so by reacting with luminol radical
and biological conditions. Interestingly, bicarbonate is expected     rather than luminol. Luminol radical can be formed through a
to accumulate in tissues undergoing ischaemia-reperfusion, a          one-electron oxidation of luminol by 'OH, ferryl ion (FeO2+) or
condition for which peroxynitrite is proposed to contribute           by other oxidants of similar reactivity [18,20,39]. When 02'- is
significantly to net oxidant stress [11,14,15,28].                    formed after reaction of ONOO- with HCO3-, it is possible that
   Smaller yields of ONOO--induced luminol chemiluminescence          H202 (formed after dismutation of 02'-) can contribute to
were found at lower pH (Figure 3). This could be due to the fact      chemiluminescence by reacting with the diazaquinone formed
that only the decomposition of luminol monoanion (apparent            after reaction with a second molecule of ONOO- (eqns. (2) and
pK. = 8.2) leads to the light-emitting route [21]. In addition,       (3)].
proton-catalysed decomposition of ONOO- becomes a more                   Peroxynitrite is, in addition to being an oxidant, a potent
efficient competing reaction by decreasing pH [11]. Consistent        nitrating species [16,40]. Luminol nitration occurring con-
with this concept, peroxynitrite oxidized luminol more efficiently    comitantly with oxidation by ONOO- is suggested by the
at pH 10.5 than at 7.5 (Table 2) and there was a shorter half-life    observed transient 425 nm absorbance. This by-product was not
of luminol chemiluminescence at lower pH (Figure 3). Luminol-         formed when other oxidants such as H202 attacked luminol.
oxidation studies were conducted at alkaline pH in order to           The mechanism of luminol nitration by ONOO- is expected to
define mechanisms of luminol reaction with ONOO- and limit            be different from other nitrating agents such as tetranitro-
the proton-catalysed decomposition of ONOO-, so that in-              methane and nitronium tetrafluoroborate. Nitration of aromatic
formation could be more precisely obtained about the kinetics         molecules by ONOO- occurs in concert with the 'OH-like
and mechanism of luminol oxidation. There were no differences         reaction of ONOO- with double bonds [40]. This mechanism
in the mechanisms of luminol oxidation at physiological versus        requires the formation of a complex with ONOOH, and attack
alkaline pH, thus all results relate to biological events.            of the peroxynitrous acid hydroxyl with the aromatic molecule to
   Cysteine and uric acid inhibited luminol chemiluminescence.        form an aromatic radical species which then combines with the
Cysteine reacts with ONOO- with an apparent second-order rate         remaining NO2 of peroxynitrous acid to yield the nitrated product
constant of 30-40 M-1 * s-1 at pH 10.5 [14]. Cysteine can therefore   [16]. In agreement with the concept of a concerted nitration-
inhibit luminol chemiluminescence by both direct reaction with        oxidation reaction, we observed that uric acid and cysteine
ONOO- and by reacting with bicarbonate-derived oxidizing              inhibited both luminol oxidation and the appearance of the
species [38]. Cystine, which does not react at significant rates      425 nm absorbance peak. The final nitrated adduct was probably
with ONOO- [14], did not have any effect on luminol chemi-            a derivative of luminol oxidation products, since exposure of
luminescence or oxidation. Uric acid inhibits xanthine/xanthine       luminol to nitrating agents did not result in the characteristic
oxidase and H202/cytochrome c-induced luminol chemilumi-              425 nm peak. The transient nature of the nitrated adduct was not
nescence by scavenging OH and iron complexes having 'OH-like          related to chemiluminescence, since the 425 nm absorbance was
reactivity [20,24]. However, the OH radical scavengers mannitol       longer lived than light emission (Figure 7).
and DMSO did not affect light emission, ruling out a primary             Superoxide is frequently invoked as being responsible for
role for 'OH in chemiexcitation. We propose that uric acid            biologically produced luminol chemiluminescence due to the
reacted directly with ONOO- and/or bicarbonate-derived oxidi-         inhibitory effects of SOD.. Still, oxidants stronger than 02' and
zing species, for which we have supporting preliminary h.p.l.c.-      even H202 may be needed to explain the observed light yields
based product analyses. Uric acid and cysteine might also directly    [25]. Since macrophages and neutrophils will simultaneously
reduce luminol radical to its ground state, preventing chemi-         produce 'NO and 02'- [2] and we have recently observed the
excitation [20]. Scavenging of reactive species rather than           formation of ONOO- by macrophages [13], it will be important
quenching of excited species (resulting in non-radiative pathways     to assess what proportion of luminol chemiluminescence induced
of relaxation) was the mechanism of cysteine and urate inhibition     by activated macrophages and neutrophils depends on ONOO-.
of luminol chemiluminescence, because luminol oxidation was           For example, 'NO- and 0 ;--derived products from macrophage-
prevented by these compounds (Table 2).                               like liver Kupffer cells synergistically contribute to luminol
   Whereas urate and cysteine inhibited both luminol                  chemiluminescence. This observation strongly supports the role
chemiexcitation and oxidation to similar extents, SOD                 of ONOO- as a source of hepatic oxidant stress [7]. If ONOO-
significantly inhibited only chemiluminescence. SOD activity, ten     is a critical mediator of cell-derived luminol clhemiluminescence,
times greater (23.0 units/ml) than that required to almost            the inhibitory effects of SOD would depend on at least three
completely inhibit light emission (2.3 units/ml, Figure 5), only      different mechanisms: (1) inhibition of ONOO- formation,
partially inhibited luminol oxidation (Table 2). SOD effects on       preventing the reaction of 02-- with 'NO, (2) inhibition of 02'--
light yields are explained by the scavenging of O2e, whereas the      mediated oxidation of luminol radical [eqn. (6)] and (3) catalysis
influence on oxidative yields occurring with higher SOD activities    of the direct decomposition of ONOO-. Mechanisms (2) and (3)
were probably due to a direct reaction between ONOO- and the          would complement mechanism (1) because SOD is unable to
enzyme. Indeed, Cu/Zn SOD catalyses decomposition of                  totally inhibit ONOO- production by activated macropLages,
ONOO- to a species with a reactivity similar to nitronium ion         since a fraction of ONOO- was apparently formed in sites where
[16]. Thus SOD can directly inhibit ONOO--mediated luminol-           SOD does not have access [11].
oxidation reactions by also scavenging ONOO-. Since nitronium            Cell and tissue toxicity from excess production of peroxynitrite
ion could then attack luminol, we added nitronium                     may occur because of its strong oxidizing properties. In addition,
tetrafluoroborate to luminol and saw that it did not yield            our data suggest that ONOO- will react with bicarbonate,
chemiluminescence.                                                    forming secondary bicarbonate-derived oxidizing species.
   Unlike ONOO-, its biological precursors 02- and NO did not         Extracellular and intravascular compartments ean have up to
stimulate chemiexcitation of luminol. We observed no direct           25 mM bicarbonate, inferring that products of the reaction of
oxidation of luminol or light emission with up to 2.5 mM              ONOO- with bicarbonate will be an important potential mech-
                                                                                                                        Peroxynitrite and chemiluminescence                     57

anism of oxidant tissue injury. The formation of bicarbonate                                 13 Ischiropoulos, H., Zhu, L. and Beckman, J. S. (1992) Arch. Biochem. Biophys. 298,
radicals greatly increases 02--induced lysis of erythrocytes [41],                              446-451
stimulates enzyme inactivation and enhances luminol chemi-                                   14 Radi, R., Beckman, J. S., Bush, K. and Freeman, B. A. (1991) J. Biol. Chem. 266,
luminescence [37]. We envisage that, after ONOO- production                                  15 Radi, R., Beckman, J. S., Bush, K. and Freeman, B. A. (1991) Arch. Biochem.
in vivo, several competing pathways will contribute to ONOO-                                    Biophys. 288, 481-487
reactions and decomposition. Some of these pathways will include                             16 Ischiropoulos, H., Zhu, L., Chen, J., Tsai, H. M., Martin, J. C., Smith, C. D. and
(a) proton-catalysed ONOO- decomposition to potent secondary                                    Beckman, J. S. (1992) Arch. Biochem. Biophys. 298, 431-437
oxidants, (b) direct reaction of ONOO- with biomolecules such                                17 Radi, R., Rubbo, H. and Prodanov, E. (1989) Biochim. Biophys. Acta 994, 89-93
as methionine [42] or thiols and (c) reaction of ONOO- with                                  18 Allen, R. C. (1986) Methods Enzymol. 133, 449-493
                                                                                             19 Archer, S. L., Nelson, D. P. and Weir, K. E. (1989) J. Appl. Physiol. 67, 1903-1911
bicarbonate to form toxic secondary reactive species.                                        20 Radi, R., Rubbo, H., Thomson, L. and Prodanov, E. (1990) J. Free Rad. Biol. Med. 8,
This work was supported by NIH grants NS24275, HL48676 and HL40458 (B.A.F.)                  21 Merenyi, G., Lind, J. and Eriksen, T. E. (1990) J. Biolum. Chemilum. 5, 53-56
and HL46407 (J.S.B.). J.S.B. is also an Established Investigator of the American             22 Hodgson, E. K. and Fridovich, I. (1973) Photochem. Photobiol. 18, 451-455
Heart Association. We thank Dr. Gabor Merenyi and Dr. Johan Lind from the Royal              23 Miller, E. K. and Fridovich, I. (1986) J. Free Rad. Biol. Med. 2,107-110
Institute of Technology in Stockholm, Sweden, for suggestions regarding the                  24 Radi, R., Thomson, L., Rubbo, H. and Prodanov, E. (1991) Arch. Biochem. Biophys.
mechanism of luminol chemiexcitation by peroxynitrite. We also thank Dr. Harry                  288, 112-117
Ischiropoulos for discussions throughout the course of this work and Yvonne Lambott          25 Merenyi, G., Lind, J. and Eriksen, T. E. (1985) Photochem. Photobiol. 41, 203-208
for manuscript preparation.                                                                  26 Mulligan, M. S., Hevel, J. M., Marletta, M. A. and Ward, P. A. (1991) Proc. Natl.
                                                                                                Acad. Sci. U.S.A. 88, 6338-6342
REFERENCES                                                                                   27 Mulligan, M. S., Warren, J. S., Smith, C. W., Anderson, D. C., Yeh, C. G., Rudolph,
                                                                                                A. R. and Ward, P. A. (1992) J. Immunol. 148, 3086-3092
 1   Rosen, G. M. and Freeman, B. A. (1984) Proc. Natl. Acad. Sci. U.S.A. 81,                28 Matheis, G., Sherman, M. P., Buckberg, G. D., Haybron, D. M., Young, H. H. and
     7268-7273                                                                                  Ignarro, L. J. (1992) Am. J. Physiol. 262, H616-620
 2   Marletta, M. A. (1989) Trends Biochem. Sci. 14, 488-492                                 29 Valentine, J. S., Miksztal, A. R. and Sawyer, D. (1984) Methods Enzymol. 105,
 3   Ding, A. H., Nathan, C. F. and Stuehr, D. J. (1988) J. Immunol. 141, 2407-2412             71-81
 4   Matsubara, T. and Ziff, M. (1986) J. Cell. Physiol. 127, 207-210                        30 Hodgson, E. K. and Fridovich, I. (1975) Biochemistry 14, 5294-5299
 5   Stuehr, D. and Marletta, M. T. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 7738-7742       31 Flohe, L. and Oiting, F. (1984) Methods Enzymol. 105, 93-104
 6   Wagner, D. A., Young, V. R. and Tannenbaum, S. R. (1983) Proc. Nati. Acad. Sci.         32 Rauhut, M. M., Semsel, A. M. and Roberts, B. G. (1966) J. Org. Chem. 31,
     U.S.A. 80, 4518-4521                                                                       2431-2436
 7   Wang, J. F., Komarov, P., Sies, H. and de Groot, H. (1991) Biochem. J. 279,             33 Thorpe, G. H. G. and Kricka, L. J. (1986) Methods Enzymol. 133, 331-353
     311-314                                                                                 34 Koppenol, W. H., Moreno, J. J., Pryor, W. A., Ischiropoulos, H. and Beckman, J. S.
 8   Saran, M., Michel, C. and Bors, W. (1990) Free Radical Res. Commun. 10, 221-226            (1992) Chem. Res. Toxicol. 5, 834-842
 9   Gryglewski, R. J., Palmer, R. M. J. and Moncada, S. (1986) Nature (London) 320,         35 Beyer, W. F. Jr. and Fridovich, I. (1988) Anal. Biochem. 173, 160-165
     454-456                                                                                 36 Koppenol, W. and Rush, J. D. (1987) J. Phys. Chem. 91, 4430-4431
10   Albina, J. E., Mills, C. D., Henry, W. L. and Caldwell, M. D. (1989) J. Immunol. 143,   37 Michelson, A. M. and Maral, J. (1983) Biochimie 65, 95-104
     3641-3646                                                                               38 Chen, S. N. and Hoffman, M. Z. (1973) Radiat. Res. 56, 40-47
11   Beckman, J. S., Beckman, T. W., Chen, J., Marshall, P. A. and Freeman, B. A. (1990)     39 Keith, W. G. and Powell, R. E. (1969) J. Chem. Soc. (A), 90
     Proc. Natl. Acad. Sci. U.S.A. 87, 1620-1624                                             40 Halfpenny, E. and Robinson, P. L. (1952) J. Chem. Soc., 939-946
12   Hogg, N., Darley-Usmar, V. M., Wilson, M. T. and Moncada, S. (1992) Biochem. J.         41 Michelson, A. M. and Durosay, P. (1977) Photochem. Photobiol. 25, 55-63
     28, 419-424                                                                             42 Moreno, J. J. and Pryor, W. A. (1992) Chem. Res. Toxicol. 5, 425-431

Received 3 August 1992; accepted 27 August 1992

To top