ARCHIVES

OF BIOCHEMISTRY

Vol. 298, No. 2, November

AND

BIOPHYSICS

1, pp. 43%445,1992

Kinetics of Superoxide Nitration of Phenolics Joseph S. Beckman, Jun Chen,* Joseph

**l Harry Harrison,$

Dismutase- and Iron-Catalyzed by Peroxynitrite

Ischiropoulos ,* Ling Zhu,* Mark James C. Martin,* and Michael

van der Woerd,?

Departments of *Anesthesiology and SPhysics and TThe Center for Macromolecular The Uniuersity of Alnbama at Birminghnm, Birmingham, Alabama, 35294

Received

April

2, 1992, and in revised

form

June

Craig Smith,t

Tsai$ Crystallography,

15, 1992

mutase catalyzes the nitration of tyrosine residue 108 by Superoxide dismutase and Fe’+EDTA catalyzed the nitration by peroxynitrite (ONOO-) of a wide range of phenolics including tyrosine in proteins. Nitration was not mediated by a free radical mechanism because hydroxyl radical scavengers did not reduce either superoxide dismutase or Fe3+EDTA-catalyzed nitration and nitrogen dioxide was not a signiticant product from either catalyst. Rather, metal ions appear to catalyze the heterolytic cleavage of peroxynitrite to form a nitroniumlike species (NOz+). The calculated energy for separating peroxynitrous acid into hydroxide ion and nitronium ion is 13 kcal - mall’ at pH 7.0. Fe’+EDTA catalyzed nitration with an activation energy of 12 kcal . mall’ at a rate of 5700 M-‘.s-’ at 37°C and pH 7.5. The reaction rate of peroxynitrite with bovine Cu,Zn superoxide dismutase was 10’ M-l es- 1 at low superoxide dismutase concentrations, but the rate of nitration became independent of superoxide dismutase concentration above 10 pM with only 9% of added peroxynitrite yielding nitrophenol. We propose that peroxynitrite anion is more stable in the cis conformation, whereas only a higher energy species in the trans conformation can fit in the active site of Cu,Zn superoxide dismutase. At high superoxide dismutase concentrations, phenolic nitration may be limited by the rate of isomerization from the cis to trans conformations of peroxynitrite as well as by competing pathways for peroxynitrite decomposition. In contrast, Fe3+EDTA appears to react directly with the cis anion, resulting in greater nitration yields. 0 1992 Academic PWJS, IDC.

peroxynitrite

on

a second

molecule

of

superoxide

dis-

mutase (1,2). In the present study, we found that superoxide

dismutase

as well

as low

molecular

weight

transition

metals like Fe3+EDTA also catalyze nitration of a wide range of phenolics including tyrosine residues of other proteins. Phenolic nitration catalyzed by superoxide dismutase was most rapid at neutral pH and the kinetics were remarkably complex. Peroxynitrite is a strongly oxidizing agent capable of reacting by multiple mechanisms at neutral pH (3). For example, the yield of hydroxyl radical

products

from

the

oxidation

of

dimethylsulfoxide

(DMS0)2 and deoxyribose decreases at more alkaline pH, exhibiting an apparent pK, of 7.5-7.8 (4), while lipid peroxidation

and

the

rate

of sulfhydryl

oxidation

were

max-

imal at pH 7.5 (5, 6). This complex behavior at neutral pH suggeststhat the conformation of peroxynitrite may be an important determinant of its reactivity (4). Peroxynitrite is apparently most stable in the cis form (79), but appears to be more reactive in the trans configuration (3). In the present study, the kinetics and activation energies of peroxynitrite-mediated nitration of phenols by superoxide dismutase and by Fe3+EDTA were characterized. We propose that the isomerization of cis to trans peroxynitrite

may

be an

important

rate-limiting

step

determin-

ing the kinetics of superoxide dismutase-catalyzed phenolic nitration, but not for Fe3+EDTA which appears to react

directly

MATERIALS

with

AND

cis peroxynitrite.

METHODS

Peroxynitrite anion reacts with the copper in the active site of Cu,Zn superoxide dismutase to form a strong nitrating agent. In phosphate buffer alone, superoxide dis-

Preparation of peroxynitrite, H,O,-phenylglyoxal-modified dismutase and &-free superoxide dismutase was described (1). High quality deionized water was used for all solutions.

superoxide previously For exper-

1 To whom correspondence should be addressed at Department Anesthesiology, University of Alabama at Birmingham, 619 South Street, Birmingham, AL. Fax: (205) 934-7437.

’ Abbreviations used: DMSO, dimethyl sulfoxide; DTPA, triaminepentaacetic acid; NO,-HPA, 4.hydroxy-3.nitrophenylacetic 4-HPA, 4-hydroxyphenylacetic acid.

diethyleneacid;

of 19th

438 All

Copyright 0 1992 rights of reproduction

ooo3-sS61/92 $5.00 by Academic Press, Inc. in any form reserved.

NITRATION

OF

PHENOLICS

BY

439

PEROXYNITRITE

iments involving trace metal catalysis, water was passed through layers consisting of a Dowex l-X8 anion-exchange resin, a 5OW-X8 cationexchange resin, Chelex 100 resin (Sigma), charcoal and then hydroxyquinolate glass beads (Pierce Chem. Co). Phosphate buffers were passed through Chelex 100 resin and hydroxyquinolate glass beads. After chromatography, each solution was made to 0.1 mM diethylenetriaminepentaacetic acid (DTPA) and stored in plastic. A l.l-fold excess of EDTA, DTPA, or desferrioxamine was added to 100 mM Fe3+NH,S0, (Aldrich) dissolved in 0.1 M HCl to prepare solutions of Fe3+ chelators. Nitration assays. Reactions were initiated by adding a small drop of the alkaline stock peroxynitrite solution with a Hamilton syringe on the test tube wall above the assay solution and then rapidly mixing with the solution by vortexing. After completion of the reaction, the pH was measured to compensate for changes due to the alkaline stock of peroxynitrite. The pH was then adjusted to 10.0-10.6 with 3 M NaOH and the absorbance at 430 nm recorded at each pH. The yield of 4-hydroxy3-nitrophenylacetic acid (NO,-HPA) was calculated from 6 = 4400 Me’ *cm-‘. Results are reported for the mean + STD of at least four determinations from two or more separate experiments. Reversed phase HPLC showed that NO,-HPA was the principal product from the superoxide dismutase reaction (1). Nitric oxide and nitrogen dioxide were measured in l-ml sample volumes in sealed test tubes, which were rapidly bubbled with a stream of helium. The head space was connected to an Antek Model 705D nitric oxide detector (Houston, TX), which detects nitric oxide in the gas phase by chemiluminescence (10). Nitrogen dioxide from the sample was converted to nitric oxide in the gas stream by heating a 20-cm length of steel tubing in a 550°C oven placed immediately before the inlet to the nitric oxide detector. Analysis of kinetics. Stopped-flow measurements were performed on a Hi-Tech Scientific instrument (Birmingham, England) with a mixing time of under 2 ms. The decomposition of peroxynitrite was followed at 302 nm, while the nitration of phenol, 4-hydroxyphenylacetic acid (4-HPA), and glycyl-tyrosine was followed at 412, 432, and 420 nm, respectively. Peroxynitrite does not absorb significantly at these wavelengths. Data were fitted by nonlinear regression to a single exponential function with a nonzero offset using software supplied with the stopped

0.0 ’

I

I

I

I

1

4

5

6

7

8

9

PH FIG. 1. The pH optima for nitration of 0.5 mM 4-HPA in the presence (N) and absence (0) of 10 pM superoxide dismutase. The buffers were 50 mM potassium phosphate, 100 FM DTPA at pH 7.4 and 37°C with 0.5 mM peroxynitrite. Results are means + STD of at least four determinations.

1

OK 0

0.25

0.5

Peroxynitrite

0.75

(mM)

FIG. 2. Nitration of 4-HPA in the presence (m) and absence (0) of 10 @M superoxide dismutase as a function of initial peroxynitrite concentration. The buffer was 50 mM potassium phosphate, pH 7.4, plus 100 (.tM DTPA at 37°C.

flow instrument. A small but consistent deviation from first-order kinetics that occurred during the first l-2 s was deleted from the analysis. Other curves were fitted by nonlinear regression using Kale&graph on a Macintosh.

RESULTS

Both Fe3+EDTA and superoxide dismutase catalyzed the peroxynitrite-dependent nitration of phenol, 4-HPA, and glycyl-tyrosine as well as tyrosine residues in lysozyme and histone. The yield of nitrated products followed a consistent pattern as illustrated with 4-HPA. The pH optima for both spontaneous and superoxide dismutasemediated nitration occurred at pH 7.5 (Fig. 1). The yield decreased at both more acidic and more basic pHs with apparent p&s at 6.8 and at 7.9. The total yield of NOPHPA increased in direct proportion to the initial peroxynitrite concentration of the superoxide dismutase-catalyzed reaction at pH 7.5 (Fig. 2). However, only 9.0% of added peroxynitrite was recovered as NO,-HPA with 10 PM superoxide dismutase, while a 6.9% yield was observed in the absence of superoxide dismutase. Nitration yield was maximal with 4-HPA concentrations greater than 2 mM with 50% of maximum occurring at 0.17 mM 4-HPA with superoxide dismutase and 0.24 mM 4-HPA in its absence (Fig. 3). The yield of NO,-HPA also reached a maximum when superoxide dismutase concentrations were greater than 10 pM (Fig. 4). The reaction reached 50% saturation with 1.6 PM superoxide dismutase. Cu,Zn superoxide dismutase that was first allowed to react with peroxynitrite in the absence of 4-HPA was subsequently able to nitrate 4HPA to the same extent as unmodified superoxide dismutase. Peroxynitrite-modified Cu,Zn superoxide dismutase also retained its superoxide scavenging activity

440

BECKMAN

ET

AL.

6or

15or

04

I 0

0.5

1

2

1.5

”

4-HPA

0

(m&l)

FIG. 3. Nitration yield as a function of I-HPA concentration in the presence (a) and absence (0) of 10 pM superoxide dismutase. The buffer was 50 mM potassium phosphate, pH 7.5, plus 100 pM DTPA at 37°C and the concentration of peroxynitrite was 0.5 mM.

Thus, the limited yield was not due to inactivation of superoxide dismutase by peroxynitrite. Transition metal catalysis. In contrast to superoxide dismutase, transition metals such as Fe3+EDTA or CuSOl catalyzed nitration of 4-HPA when added in approximately stoichiometric amounts (Fig. 5). The yield or percentage of added peroxynitrite recovered as nitrated phenolic depended upon the type of phenol (Table I) and was 90 sfr1.4% with phenol but only 40 + 3.4% with 4-HPA. The yield was independent of pH over a range from 6 to

0.25

0.5

0.75

1

Fe+3-chelate(mM) FIG. 6. Nitration catalyzed by FeS+ chelates. The final concentration of peroxynitrite was 0.5 mM. The buffer was 50 mu potassium phosphate, pH 7.5, plus 0.5 mM 4-HPA at 37°C. Fe3+EDTA (+), Fe3+DTPA, (m) and Fe3+-desferrioxiamine (A).

(1).

T

9, when sufficient concentrations of Fe3+EDTA were added to out-compete spontaneous decomposition. The reaction was still catalytic, because multiple equimolar additions of peroxynitrite to Fe3+EDTA resulted in the same amount of nitration with each addition (Fig. 6). The reaction rate was strictly linear with Fe3+EDTA concentration. The second-order rate constant for peroxynitrite and Fe3+EDTA was 5700 M-l. s-l at 37°C (Fig. 7a). The apparent activation energy for Fe3+EDTA-catalyzed nitration of 4-HPA was 12 kcal *mol-l (Fig. 7b). In contrast, spontaneous nitration required an apparent activation energy of 20 kcal * mol-‘. Unlike Fe3+EDTA, Fe3+DTPA did not catalyze nitration by peroxynitrite (Fig. 5). Desferrioxamine bound to Fe3+in a 1.1 to 1 ratio significantly inhibited nitration (Fig. 5), but this appears to be due to

f

0

TABLE Yield

of Phenolic

Nitration

(PM

2oL-----0

4

8

12

16

Superoxide dismutase(PM) FIG. 4. The effects of increasing superoxide dismutase on NO,-HPA yield. The buffer was 50 mM potassium 7.5, plus 100 pM DTPA and 0.5 mM I-HPA at 37°C.

concentration phosphate, pH

p-Hydroxyphenylacetic Glycyl-tyrosine Phenol p-Cresol Benzoate Salicylate

acid

I Catalyzed

Yield + SD)

199 f 17 17Ozk 8 448+ 7 118+ 8 0 44+ 4

by Fe3+EDTA

% of added 40 34 90 24

peroxynitrite

+ k + + 0 8.8 +

3.4 1.6 1.4 1.6 0.8

Note. The concentration of the phenolics was 5 mM in 50 mM potassium phosphate, pH 7.5, plus 2.5 mM Fe3+EDTA at 37°C. The final concentration of peroxynitrite was 0.5 mM.

NITRATION

OF

PHENOLICS

”

0

0.5

1

Peroxynitrite

1.5

2

(mM)

FIG. 6. Catalytic action of Fe3+EDTA-mediated nitration by peroxynitrite. Sequential additions of (A) 0.125 mM and (0) 0.25 mM peroxynitrite were made to 0.25 mM Fe3+EDTA and 1 mM I-HPA in 100 mM potassium phosphate, pH 7.5, and at 37’C. The yield of NO,-HPA is reported as a function of total peroxynitrite concentration.

the slight excess of unbound desferrioxamine, which was also inhibitory at low concentrations (data not shown). We have previously observed desferrioxamine to directly inhibit DMSO and deoxyribose oxidation by peroxynitrite (4). Addition of 1 mM peroxynitrite to 1 mM Fe3+EDTA produced only 2.5 I.LM detectable nitrogen dioxide, compared to 5 PM in buffer alone. The hydroxyl radical scav-

BY

441

PEROXYNITRITE

engers acetate and ethanol (100 mM) had no significant effect on Fe3+EDTA-mediated nitration of 1 mM 4-HPA, while DMSO and mannitol (100 mM) reduced nitration by 24 and 12%, respectively (Fig. 8). Kinetics of superoxide dismutase nitration. The kinetics of superoxide dismutase-catalyzed nitration were studied at pH 8.0 rather than at 7.5 because catalyzed nitration was faster than spontaneous nitration. In addition, NO,-HPA is nearly completely ionized at pH 8, which increases its apparent extinction coefficient at 432 nm compared to pH 7.5. At pH 8 and 37”C, the rate of nitration was 1 X lo5 M-l * s-l at low superoxide dismutase concentrations (Fig. 9). However, the rate reached a maximum when superoxide dismutase concentrations were above 10 PM, similar to the results observed with total yield (Fig. 2). The apparent zero-order rate constant with saturating superoxide dismutase concentrations was 0.65 s-l. The rate of spontaneous nitration of 2 mM 4-HPA was 0.36 s-l, which was approximately the same as the spontaneous rate of peroxynitrite decomposition of 0.34 s-l followed at 302 nm in the absence of 4-HPA. Spontaneous nitration of 4-HPA by peroxynitrite was not due to trace metal catalysis in the phosphate buffers, because it was only slightly decreased when buffers were treated with 100 ~.LM DTPA. Furthermore, DTPA had no effect on the yield of nitration products after phosphate buffers were first passed over metal chelation columns. DISCUSSION

Transition metals as well as superoxide dismutase catalyze the nitration of a wide range of phenolics, including

10.00

a

1.00

0.10

0.01 0.5

1

Fee3EDTA (mM)

1.5

2

0.003

0.0031

0.0032

0.0033

0.0034

0.0035

0.0036

1 / Temp (“K)

FIG. 7. a. The rate of 4-HPA nitration versus Fe3+EDTA concentration using 0.5 mM peroxynitrite. The buffer contained 2 mM 4-HPA, 50 mM potassium phosphate, pH 7.5, plus 100 pM DTPA. The temperatures were 10°C (+), 2O’C (A), 30°C (m), 40°C (O), and 50°C (Cl). b. Arrhenius plot for spontaneous (+) and Fe3+EDTA catalyzed nitration (W). The rates are derived from the slopes for the Fe3+EDTA-catalyzed reaction in Fig. 7a, and from the y-intercepts for spontaneous nitration.

BECKMAN

ET

AL.

macrophages and potentially many other cell types can produce peroxynitrite (21), metal-catalyzed nitration could be a major pathological mechanism of tissue injury. Recently, we have calculated that the energy required for heterolytic cleavage of peroxynitrite into nitronium ion and hydroxyl ion to be about 13 kcal . mol-’ in water at pH 7.0 (3), ONOOH

Control

Acetate

DMSO

Ethanol

Mannitol

FIG. 8. Effect of hydroxyl radical scavengers (100 mM) on 1.25 mM Fe’+EDTA-catalyzed nitration of 0.5 mM I-HPA by 0.3 mM peroxynitrite in 100 mM potassium phosphate, 100 pM DPTA, pH 7.5, and 37°C. Statistical significance (*P < 6.05) was determined by a one-way analysis of variance using the least-significant-difference post-hoc test using the JMP program from SAS (Cary, NC).

tyrosine residues in most proteins. We presented results principally for 4-HPA because of its high solubility compared to tyrosine and low toxicity compared to phenol, but the overall pattern of reactivity was similar with other phenolics. Nitration of 4-HPA catalyzed by low superoxide dismutase concentrations proceeded at a reaction rate of 1 X 10’ M-l s-l at 37°C. This reaction is about 40 times faster than the fastest reaction with peroxynitrite we have determined to date, the oxidation of cysteine (5). Thus, the reaction of superoxide dismutase with peroxynitrite could be pathologically relevant in biological systems producing both superoxide and nitric oxide and could contribute to the apparent toxicity of superoxide dismutase in high dosages to ischemic heart (l&12). Metalcatalyzed nitration by peroxynitrite also provides an alternative explanation to the Haber-Weiss reaction for the role of transition metals in oxidative tissue injury. The rate of peroxynitrite reaction with Fe3+EDTA was 5700 M-l * s-l, which is in the same range as the rate of hydrogen peroxide reacting with Fe2+EDTA [7000 M-l. s-l; Ref. (13)]. Moreover, the reaction with peroxynitrite does not require that iron first be reduced by superoxide or another reductant to be toxic. Nitration of tyrosine residues by tetranitromethane is well known to alter protein function, including cytochrome P450 (14), cu-thrombin (15), and mitochondrial ATPase (16). Treatment with tetranitromethane inactivates the complement subcomponent Clq binding capacity of human IgG (17), abolishes the inhibitory activity of human a-1-proteinase inhibitor for elastase (18), and inhibits the binding of human high density lipoprotein to liver plasma membranes (19). Phosphorylation of tyrosines plays a critical role in cell regulation and is an important target damaged by nitration (20). Since activated l

j

NOi

+ OH-

13 f 2 kcal . mol-’

[l]

This compares favorably with the 12 kcal * mol-’ activation energy measured for Fe3+EDTA-catalyzed nitration (Fig. 7b). Thermodynamic calculations also suggest that heterolytic cleavage would not occur spontaneously because the energy required to produce an initial charge separation into N026+’ ’ “-OH in water would require approximately 45 kcal * mol-’ (3). Thus, a substantial activation barrier exists for this pathway unless the reaction is catalyzed. Transition metals like Fe3+EDTA could catalyze the formation of a nitronium ion-like species from peroxynitrite by the following reactions: ONOO-

+ Fe3+EDTA

phenol + NOi+ -

=+ NOi’

‘-0 -

‘-0 -

Fe2+EDTA

N02-phenol 2H+ + -0 -

-

Fe3+EDTA

*

+ H+ + -O-Fe2+EDTA

Fe2+EDTA

*

[2]

[3]

Hz0 + Fe3+EDTA

[4]

In the first step, the negatively charged peroxynitrite anion will be electrostatically attracted to Fe3+EDTA to form an intermediate complex. Fe3+EDTA has a seventh coordination position available on the iron (22), which al-

“.LJ”

0

4

8

12

16

20

Superoxide dismutase (PM) FIG. 9. Rate of 4-HPA nitration as a function of supsroxide dismutase concentration. The buffer was 50 mM potassium phosphate, 100 PM DTPA, plus 2 mu 4-HPA. The reaction was followed at 430 nm at 37’C.

NITRATION

OF

PHENOLICS

lows it to be reduced by superoxide as well as to react with peroxynitrite. On the other hand, all coordination sites in Fe3+DTPA are occupied, explaining why it does not catalyze phenolic nitration by peroxynitrite or participate in the iron-catalyzed Haber-Weiss reaction. Transition metals are far more electropositive than hydrogen. Therefore, electron density in the peroxynitriteFe3+EDTA complex will be pulled away from the nitrogen toward the iron, favoring heterolytic cleavage to give a nitronium-like species that attacks phenols. However, it seems unlikely that nitronium ion is actually released from the Fe3+EDTA complex as a separate species, because the high nitration yield from added peroxynitrite with only millimolar phenolic concentrations. Water reacts rapidly with nitronium ion to give nitrate and we found that direct addition of nitronium salts of tetrafluoroborate to 2 mM 4-HPA gave nitration yields of less than 0.5%. Thus, the peroxynitrite-Fe3+EDTA complex may directly react with phenol without nitronium ion being physically separated from the complex. The -O-Fe2+EDTA will rapidly add two hydrogen ions from the solvent to release water and regenerate Fe3+EDTA (Eq. [4] ). The addition of hydrogen ions could easily occur as part of the nitration reaction rather than as a separate step as written. The kinetics were consistent with the rate-limiting step of 4-HPA nitration being the reaction of peroxynitrite with Fe3+EDTA (Eq. [2] ), because the reaction rate was first order with respect to both of their concentrations. The rate of nitration became zero-order with 4-HPA concentrations greater than 2 mM, indicating that the actual nitration step (Eq. [3]) was not rate-limiting. Although the second-order rate constant for Fe3+EDTA catalyzing nitration of 4-HPA is moderately fast, high concentrations of Fe3+EDTA (-1 mM) were required to achieve 100% yields. This appears to be a consequence of competition with spontaneous decomposition, which occurs at 0.65 s-l at 37°C. Yet Fe3’EDTA was clearly acting as a catalyst because the same extent of nitration was achieved after multiple equimolar additions of peroxynitrite (Fig. 6). Phenolic nitration can also be mediated by free radical mechanisms involving addition of nitrogen dioxide to phenoxyl radicals (23). Peroxynitrous acid decomposes to form a hydroxyl radical-like oxidant, capable of oxidizing phenolics to radicals. It also produces nitrogen dioxide in the presence of hydroxyl radical scavengers (24). Indeed, the decomposition of peroxynitrous acid at pH 2 results in the nitration as well as hydroxylation and dimerization of phenolics (25, 26). However, the yield from such a pathway is relatively low, since only a maximum of 2530% of peroxynitrous acid gives a detectable hydroxyl radical-like oxidant (4, 27), and multiple hydroxylation and dimerization products would also be formed. At pH 5, we have observed the formation of catechol and hydroquinone from the reaction of phenol plus peroxynitrite (H. Ischiropoulos, in preparation). However, only nitrated

BY

PEROXYNITRITE

443

derivatives were found by HPLC with catalysis by either Fe3+EDTA or superoxide dismutase. It is unlikely that Fe3’EDTA catalyzed phenolic nitration through a free radical mechanism, though such a mechanism is possible. Peroxynitrite itself is a strong oxidant (E’, = 1.4 V at pH 7) and thus might oxidize Fe3+EDTA to a ferry1 intermediate (Fe4+ = 0), releasing nitrogen dioxide. The ferry1 radical so formed can readily oxidize phenolics to phenoxyl radicals, which then react with nitrogen dioxide to give nitrophenols (23). However, the yield of nitrogen dioxide from peroxynitrite plus Fe3+EDTA was only 0.2% even in the absence of 4-HPA, which was half of that formed in phosphate buffer alone. Furthermore, addition of 100 mM acetate, ethanol, DMSO, or mannitol to 0.5 mM 4-HPA (a 200-fold excess of scavenger) either had no effect or only slightly reduced the yield of NO,-HPA. Thus, a radical mechanism is unlikely to account for phenolic nitration catalyzed by Fe3+EDTA. The differences between Fe3+EDTA and superoxide dismutase-catalyzed nitration of phenol were striking. Only 9% of added peroxynitrite reacted with superoxide dismutase to nitrate 4-HPA compared to 90% with phenol and 40% with 4-HPA with Fe3+EDTA. The reaction rate and yield were maximal with >lO pM superoxide dismutase whereas the rate with Fe3+EDTA continuously increased without evidence of saturation (Fig. 7a). The maximum rate with superoxide dismutase was still lofold slower than the pseudo-first-order rate constant observed with 1 mM Fe3+EDTA. Furthermore, the yield was maximal at pH 7.5 with superoxide dismutase, while the yield with Fe3+EDTA was independent of pH over a wide range. Thus, some rate-limiting step involving peroxynitrite appears to be necessary before peroxynitrite can react with superoxide dismutase. To account for the different kinetics between Fe3+EDTA and superoxide dismutase, we propose a working model postulating a potentially rate-limiting step involving the isomerization of peroxynitrous acid from the cis to the trans geometry (Scheme 1). The difference in nitration kinetics would arise from the active site of superoxide dismutase accommodating only peroxynitrite anion in the higher energy tram configuration (l), whereas Fe3+EDTA may react directly with the more stable cis peroxynitrite. The copper of Cu,Zn superoxide dismutase sits at the base of a deep hydrophobic pocket shaped to accommodate superoxide (28). Thus, peroxynitrite in the cis conformation would be sterically hindered from reaching the copper, whereas the bans configuration could fit into the active site with the O-O moiety occupying the pocket and the -N=O remaining exposed to solvent (1). No such steric interference would be expected for cis-peroxynitrite reacting with a small ion like Fe3+EDTA. Both 15N NMR and laser Raman spectroscopic studies indicate that peroxynitrite in alkaline solution is present in only one geometry, which appears to be the cis con-

444

BECKMAN

tram

r=O -0-o

SOD

Phenolic Nitration

e

cis Rate-limiting Isomerization

0 “N + Hoso/ e

pK, 1.9

N=O Ho-d )

-0-o’

“\ o,N=o+N03-

“HO. + *NO,”

pK, 6.8

OQN

H

1

Fe+3-EDTA *

Phenolic Nitration

SCHEME 1. Proposed role of conformation of peroxynitrite on reactivity. Peroxynitrite anion is most stable in the cis conformation, and is proposed to react directly with Fe3+EDTA to form an efficient nitrating agent. Calculations indicate that the barrier to isomerization is substantially lower for peroxynitrous acid and therefore is assumed here to be the principal route of isomerization. The trots geometry appears to be slightly higher in energy than the cis form, which implies that the rate of truns-to-& isomerization must be faster than the forward reaction. The truns geometry may also be more reactive because of reduced stabilization between the terminal peroxide oxygen and the N = 0 group. For example, bending of the N - 0 - 0 angle coupled with stretching of the 0 - 0 bond yields a high energy intermediate, which appears to react like hydroxyl radical (3). Such an intermediate derived from the truns geometry may be responsible for spontaneous nitration. On the other hand, ionization to trans-peroxynitrite anion would allow it to react with superoxide dismutase. A slightly higher PK., of trans-peroxynitrite would account for the greater yield of nitration observed at pH 7.5, while the rate of isomerization could account for the rate-limited step at high superoxide dismutase and 4-HPA concentrations.

formation (9). Quantum mechanical calculations indicate that the cis conformation is lower in energy because of interactions between the terminal peroxide oxygen with the distal oxygen (J. Harrison, M. van der Woerd, and J. S. Beckman, unpublished observations). In the tram anion, these interactions are weaker because the terminal oxygens cannot interact. Peroxynitrous acid (ONOOH), the conjugate acid of peroxynitrite, is also more stable in the cis form by l-2 kcal . mol-’ (7,8). Calculations further indicate that the barrier for rotation about the 0 - 0 bond may be substantially smaller for peroxynitrous acid compared to the anion because the negative charge on the anion limiting rotation is neutralized by addition of the hydrogen. However, the tram geometry may be more reactive than the cis geometry. For example, the reaction of peroxynitrous acid to give a hydroxyl radical-like oxidant and nitrogen dioxide appears to be mediated by a high energy intermediate derived from tram peroxynitrous acid [Scheme 1, Ref. (3)]. Such an intermediate may be important for the spontaneous nitration of phenols. Under the proposed model shown in Scheme 1, cis peroxynitrite would first have to be protonated to isomerize to tram peroxynitrous acid and then ionize to form trans peroxynitrite anion before it could react with superoxide dismutase. Both spontaneous nitration and superoxide

ET

AL.

dismutase-catalyzed nitration were maximal at pH 7.5 and decrease with apparent pK,s of 6.8 and 7.9 (Fig. 1). The pK= of 6.8 agrees with the apparent pK, derived previously from the rate of peroxynitrite decomposition (3, 5, 29), which we propose corresponds to the pK, for cis peroxynitrite. The higher pK,, of 7.9 may result from ionization of tram peroxynitrous acid, which would require the energy difference between the cis and tram anions to be slightly higher than the difference between cis and trams peroxynitrous acid. Because of its higher pK,, tram peroxynitrous acid would be relatively stabilized at neutral pH with respect to cis peroxynitrous acid. If the attacking speciesis derived from trans peroxynitrous acid, the slight difference in pK,s between the cis and the trams geometries would explain the increase in nitration (Fig. 1) and lipid peroxidation (6) observed at pH 7.5. ACKNOWLEDGMENTS We are grateful for many helpful discussions with Drs. W. Koppenol (Louisiana State University) and Rafael Radi (University of the R.epublic, Montevideo, Uruguay). Drs. H. Cheung and S. Lin kindly allowed us frequent usage of their stopped flow spectrophotometer and provided frequent advice and assistance. This work was supported by NIH Grant HL-46407 and a Grant-in-Aid from the American Heart Association. J. S. Beckman is an Established Investigator of the American Heart Association.

REFERENCES 1. Ischiropoulos, H., Zhu, L., Chen, J., Tsai, H. M., Martin, J. C., Smith, C. D., and Beckman, J. S. (1992) Arch. B&hem. Biophys. 298,431-437. 2. Smith, C. D., Carson, M., Van der Woerd, H., and Beckman, J. S. (1992) Submitted 3. Koppenol, W. H., Moreno, and Beckman, J. S. (1992) 4. Beckman, Freeman,

J. J., Pryor, W. A., Ischiropoulos, Chem. Res. Toxicol., in press.

J. S., Beckman, T. W., Chen, J., Marshall, B. A. (1990) Proc. Natl. Acad. Sci. USA 87,

5. Radi, R., Beckman, J. S., Bush, J. Biol. Chem. 266,4244-4250. 6. Radi, Arch.

M., Chen, J., Ischiropoulos, for publication. H.,

P. M., and 1620-1624.

K. M., and Freeman,

B. A. (1991)

R., Beckman, J. S., Bush, K. M., and Freeman, Biochem. Biophys. 288,481-487.

B. A. (1991)

7. Cheng, B. M., Lee, J. W., andLee, 2814-2817.

Y. P. (1991)

8. McGrath, M. P., Francl, M. M., Rowland, (1988) J. Phys. Chem. 92,5352-5357. 9. Tsai, M. (1991) thesis, University

A. J., and Ronco,

11. Omar, B. A., Gad, N. M., Jordan, W. J., Downey, J. M., and McCord, Med. 9,465-471.

Chem.

F. S., and Hehre,

Raman Spectra of Peroxynitrite of Alabama at Birmingham.

10. Fontijn, A., Sadadell, 42,575-578.

J. Phys.

Anion.

R. J. (1970)

Anal.

95, W. J.

Masters Chem.

M. C., Striplin, S. P., Russell, J. M. (1990) Free Radical Biol.

12. Omar, B. A., and McCord, J. M. (1990) Free Radical Biol. Med. 9, 413-418. Z., and Koppenol, W. H. (1990) in Methods 13. Rush, J. D., Maskos, in Enzymology (Packer, L., and Glazer, A., Eds.), Vol. 186, pp. 14% 156, Academic Press, San Diego. 14. Janing, G. R., Kraft, R., Blanck, J., Rabe, H., and Ruckpaul, K. (1987) B&him. Biophys. Acta 916,512-523.

NITRATION

OF

PHENOLICS

15. Lundblad, R. L., Noyes, C. M., Featherstone, G. L., Harrison, and Jenzano, J. W. (1988) J. Biol. Chem. 263,3729-3734. 16. Guerrieri,

F., Yagi, T., and Papa,

S. (1984)

J. Bioenerg.

J. H.,

Biomembr.

16,251-262. 17. McCall, M. N., and Easterbrook-Smith, S. B. (1989) Biochem. J. 257,845-851. 18. Mierzwa, S., and Chan, S. K. (1987) Biochem. J. 246, 37-42. 19. Chacko, G. K. (1985) J. Lipid Res. 26,745-754. 20. Martin, B. L., Wu, D., Jakes, S., and Graves, D. J. (1990) J. Biol. Chcm. 265.7108-7111. 21. Ischiropoulos, H., Zhu, L., and Beckman, J. S. (1992) Arch. Biochem. Biophys. 298,446-451. 22. Stezowski, J. J., Countryman, R., and Hoard, J. L. (1973) Znorg. Ckm. 12,1749-1754.

BY

445

PEROXYNITRITE

23. Priitz, W. A., Miinig, H., Butler, Biochem. Biophys. 243,125-134. 24. Zhu, L., Gunn, 298.452-457. 25. Halfpenny,

C., and Beckman,

J., and Land, J. (1992)

Arch.

E. J. (1985) Biochem.

Arch.

Biophys.

E., and Robinson,

P. L. (1952)

J. Chem.

Sot.

1952,

E., and Robinson,

P. L. (1952)

J. Chem.

Sot.

1962,

928-938. 26. Halfpenny,

939-946. 27. Mahoney,

L. R. (1970)

J. Am. Chem.

Sot. 92, 5262-5263.

28. Getzoff, E. D., Tainer, J. A., Weiner, P. K., Kollman, P. A., Richardson, J. S., and Richardson, D. C. (1983) Nature (London) 306, 287-290. 29. Keith,

W. G., and Powell,

P. E. (1969)

J. Chem.

Sot. A,

1969,453.