Oxidation of threonylmethionine by peroxynitrite. Quantification of the one-electron transfer pathway by comparison to one-electron photooxidation

Article abstract of DOI:10.1021/ja964031z

Peroxynitrite can modify methionine by one- and two-electron oxidation pathways. Here, we have quantified the extent of one-electron oxidation of threonylmethionine (Thr-Met) by peroxynitrite using a characteristic reaction according to which Thr-Met sulfur radical cations decompose via fragmentation of the Thr side chain, yielding acetaldehyde. The efficiencies, f(acet, photo) for the formation of acetaldehyde from Thr-Met sulfur radical cations were obtained by means of one-electron photooxidation using triplet 4- carboxybenzophenone. Exact quantum yields for the formation of Thr-Met sulfur radical cations by triplet 4-carboxybenzophenone were obtained by laser flash photolysis and time-resolved UV spectroscopy. Acetaldehyde yields were measured for the reaction of peroxynitrite with Thr-Met, and division of these acetaldehyde yields by f(acet, photo) yielded the extents to which peroxynitrite reacted with Thr-Met via the one-electron transfer pathway. There was little one-electron oxidation of Thr-Met by peroxynitrite at pH 7.4, i.e., 1.5%, 1.8%, and 5.3% based on the total chemical conversion of Thr-Met for Thr-Met concentrations of 1 x 10^-3, 5 x 10^-4, and 1.75 x 10^{- 4} M, respectively. In all cases the major reaction product was the two- electron oxidation product threonylmethionine sulfoxide. However, at pH 6.0, one-electron oxidation of Thr-Met showed a significantly higher efficiency of 14% for [Thr-Met] = 1.75 x 10^-4 M. Under all experimental conditions the extent of one-electron oxidation increased with decreasing peptide concentration in agreement with a recently established mechanism according to which the one-electron oxidation of Met by peroxynitrite requires a unimolecular transformation of peroxynitrous acid to an excited species which is the ultimate one-electron oxidant.

Full text of DOI:10.1021/ja964031z

J. Am. Chem. Soc. 1997, 119, 4749-4757
4749
Oxidation of Threonylmethionine by Peroxynitrite.
Quantification of the One-Electron Transfer Pathway by
Comparison to One-Electron Photooxidation
†
†
‡
§
Jana L. Jensen, Brian L. Miller, Xiaoping Zhang, Gordon L. Hug, and
Christian Sch o1 neich*,†
Contribution from the Department of Pharmaceutical Chemistry, UniVersity of Kansas,
2
095 Constant AVenue, Lawrence, Kansas 66047, Department of Chemistry, UniVersity of
Kansas, Lawrence, Kansas 66045, and Radiation Laboratory, UniVersity of Notre Dame,
Notre Dame, Indiana 46556
X
ReceiVed NoVember 21, 1996
Abstract: Peroxynitrite can modify methionine by one- and two-electron oxidation pathways. Here, we have quantified
the extent of one-electron oxidation of threonylmethionine (Thr-Met) by peroxynitrite using a characteristic reaction
according to which Thr-Met sulfur radical cations decompose via fragmentation of the Thr side chain, yielding
acetaldehyde. The efficiencies, facet,photo, for the formation of acetaldehyde from Thr-Met sulfur radical cations were
obtained by means of one-electron photooxidation using triplet 4-carboxybenzophenone. Exact quantum yields for
the formation of Thr-Met sulfur radical cations by triplet 4-carboxybenzophenone were obtained by laser flash
photolysis and time-resolved UV spectroscopy. Acetaldehyde yields were measured for the reaction of peroxynitrite
with Thr-Met, and division of these acetaldehyde yields by facet,photo yielded the extents to which peroxynitrite reacted
with Thr-Met via the one-electron transfer pathway. There was little one-electron oxidation of Thr-Met by peroxynitrite
at pH 7.4, i.e., 1.5%, 1.8%, and 5.3% based on the total chemical conversion of Thr-Met for Thr-Met concentrations
-
3
-4
-4
of 1 × 10 , 5 × 10 , and 1.75 × 10 M, respectively. In all cases the major reaction product was the two-
electron oxidation product threonylmethionine sulfoxide. However, at pH 6.0, one-electron oxidation of Thr-Met
-
4
showed a significantly higher efficiency of 14% for [Thr-Met] ) 1.75 × 10 M. Under all experimental conditions
the extent of one-electron oxidation increased with decreasing peptide concentration in agreement with a recently
established mechanism according to which the one-electron oxidation of Met by peroxynitrite requires a unimolecular
transformation of peroxynitrous acid to an excited species which is the ultimate one-electron oxidant.
Introduction
Peroxynitrite, ONOO , is a biologically important oxidant,
1e pathway generated an intermediary methionine sulfide radical
cation, MetS (1), decomposing via reactions 1 and 2.
•
+
-
1
•
+
-
+
-
generated through the reaction of nitric oxide, NO, with
CH S CH CH CH(CO )NH (1) + HO f
3
2
2
2
3
•
- 2,3
4
superoxide, O2
,
in tissues exposed to oxidative stress where
0
.5CH SSCH + CH dCH + HNdCHCO H + H O (1)
5
6
7
3 3 2 2 2 2
peroxynitrite can modify tyrosine, tryptophan, cysteine, and
methionine (Met)8 residues. Through Met oxidation, peroxy-
,9
HNdCHCO H + H O f NH + CHOCO H
(2)
8
2
2
3
2
nitrite inhibits the R-1-proteinase inhibitor and calmodulin
stimulation of neuronal NO synthase.10 Recently, Pryor et al.
reported that peroxynitrite can modify Met via a one-electron
However, on the basis of the established chemistry for 1, we
do not expect reaction 1 to be very efficient. For example, when
(
1e) and a two-electron (2e) pathway, yielding ethylene and
1
was generated at neutral pH by pulse radiolysis, 51% of 1
9
methionine sulfoxide, respectively. It was proposed that the
underwent direct decarboxylation (reaction 3), yielding 1-amino-
11
3
-(methylthio)prop-1-yl radicals 2. In complementary experi-
*
To whom correspondence should be addressed. Phone: (913) 864-
ments 1 was generated by 1e photooxidation of Met by triplet
4
880. FAX: (913) 864-5736. E-mail: schoneich@smissman.hbc.ukans.edu.
†
3
Department of Pharmaceutical Chemistry, University of Kansas.
Department of Chemistry, University of Kansas.
University of Notre Dame.
4-carboxybenzophenone, CB*, where 60% of 1 underwent
‡
12
direct decarboxylation. These respective efficiencies of 50-
§
X
60% decarboxylation do not necessarily mean that the residual
fractions of 1 directly yield ethylene since other competing
pathways such as deprotonation in the R-position to the sulfur
Abstract published in AdVance ACS Abstracts, May 1, 1997.
1) Koppenol, W. H.; Moreno, J. J.; Pryor, W. A.; Ischiropoulos, H.;
Beckman, J. S. Chem. Res. Toxicol. 1992, 5, 834.
(
(2) Huie, R. E.; Padmaja, S. Free Radical Res. Commun. 1993, 18, 195.
(3) Goldstein, S; Czapski, G. Free Radical Biol. Med. 1995, 19, 505.
(4) Beckman, J. S.; Ye, Y. Z.; Anderson, P. G.; Chen, J.; Accavitti, M.
(reaction 4) must also be considered.
+
•
A.; Tarpey, M. M.; White, C. R. Biol. Chem. Hoppe-Seyler 1994, 375, 81.
5) Ischiropoulos, H.; Zhu, L.; Chen, J.; Tsai, M.; Martin, J. C.; Smith,
C. D.; Beckman, J. S. Arch. Biochem. Biophys. 1992, 298, 431.
6) Alvarez, B.; Rubbo, H.; Kirk, M.; Barnes, S.; Freeman, B. A.; Radi,
R. Chem. Res. Toxicol. 1996, 9, 390.
7) Radi, R.; Beckman, J. S.; Bush, K. M.; Freeman, B. A. J. Biol. Chem.
991, 266, 4244.
1 f CO
2
+ H + CH
3
SCH
2
CH
2
C HNH
2
(2)
(3)
(
•
-
+
1
+ B f CH S CHCH CH(CO )NH /
(
3
2
2
3
•
-
+
+
CH SCH CH CH(CO )NH + BH (4)
(
2
2
2
2
3
1
(
(
8) Moreno, J. J.; Pryor, W. A. Chem. Res. Toxicol. 1992, 5, 425.
9) (a) Pryor, W. A.; Jin, X.; Squadrito, G. L. Proc. Natl. Acad. Sci.
Thus, any quantification of the 1e pathway between peroxy-
nitrite and Met must take into account all of these competing
U.S.A. 1994, 91, 11173. (b) Pryor, W. A.; Squadrito, G. L. Am. J. Physiol.
995, 268, L699.
10) H u¨ hmer, A. F. R.; Gerber, N. C.; Ortiz de Montellano, P. R.;
Sch o¨ neich, Ch. Chem. Res. Toxicol. 1996, 9, 484.
1
(11) Hiller, K.-O.; Asmus, K.-D. Int. J. Radiat. Biol. 1981, 40, 583.
(12) Bobrowski, K.; Marciniak, B.; Hug, G. L. J. Am. Chem. Soc. 1992,
114, 10279.
(
S0002-7863(96)04031-0 CCC: $14.00 © 1997 American Chemical Society

4750 J. Am. Chem. Soc., Vol. 119, No. 20, 1997
Jensen et al.
Scheme 1
•
pathways. Another complication arises from the fact that 2
furanyl radical TM(S OH) (5) (reactions 8 and 9). Species 5
undergoes a coupled proton/electron transfer with the protonated
N-terminal amino group of Thr-Met (reaction 10; k10 > 2.2 ×
reduces O213 [E°(O2/O2 ) ) -0.16 V ], ultimately yielding
•-
14
3
6
-(methylthio)propionaldehyde (methional) via reactions 5 and
In fact, methional was detected as a major product
6
-1 17
.
10 s
) yielding the three-electron S∴ N-bonded intermediate
3
15
subsequent to the 1e oxidation of Met by CB*.
6. This pathway has been characterized by pulse radiolysis
coupled to time-resolved UV spectroscopy where 5 was
generated specifically by the reaction of Thr-Met with hydroxyl
•
-
+
2
+ O f O + H + CH SCH CH CHdNH (5)
2 2 3 2 2
1
7
radicals (reaction 13). Species 6 exists in equilibrium 11 with
the N-centered radical cation 7 which suffers heterolytic
cleavage of the CR-Câ bond of the Thr residue (reaction 12),
yielding protonated acetaldehyde and the capto-dative stabilized
radical 8. The latter has been detected by time-resolved EPR
CH SCH CH CHdNH + H O f
3
2
2
2
NH + CH SCH CH CHdO (6)
3
3
2
2
Methional itself has been shown to yield ethylene as a result
17
16
spectroscopy. Mechanistically, it is important to note that the
radical cation 3 itself does not directly react with the protonated
of oxidation by a variety of oxidants. Thus, one possible route
to ethylene in the peroxynitrite/Met system would be the
conversion of Met to methional, followed by a conversion of
methional to ethylene, i.e., a multistep mechanism.
1
8
N-terminal amino group (pK ≈ 8.2 ). An alternative pathway
for the formation of 6 would require proton transfer to a base
from the protonated N-terminal amino group of 3 (reaction 14)
or 4 (reaction 15) to allow bond formation between the radical
cationic sulfur center and the electron lone pair of the depro-
tonated N-terminal amino group (reaction 16). The base
required for the proton transfer reactions in both pathways will
essentially be water and, for more alkaline pH values, hydroxide
ion.
In this paper we have measured the efficiency of acetaldehyde
formation from 3 by means of 1e photooxidation experiments
with CB* and used this efficiency to quantify the extent of the
1e pathway for the reaction of peroxynitrite with the thioether
moiety of Thr-Met. All necessary information about quantum
yields for the formation of 3 during the reaction of CB* with
In this paper we report a chemical system which permits the
direct quantification of 1e oxidation of a Met residue in a model
peptide by peroxynitrite, based on a previously established
mechanism by which sulfide radical cations from Thr-Met, TM-
•
+
(
S ) (3), specifically generated by reaction of Thr-Met with
•
-
SO4 , intramolecularly form acetaldehyde via Thr side chain
cleavage, displayed in Scheme 1 (reactions 7-16).
17
The underlying mechanism, displayed in Scheme 1, is
primarily based on the reaction of water and a base with TM-
3
•
+
(
S ) (3) or its three-electron-bonded dimeric sulfide radical
+
cation (S∴ S) (4). In one potential pathway the addition of
water to 3 or 4 yields the hydrated sulfur radical cations
>S ‚‚‚OH2 from 3 or [(S∴ S)OH2] from 4) which subse-
quently transfer a proton to a base, generating the hydroxysul-
3
•
+
+
(
Thr-Met was obtained by complementary time-resolved laser
photolysis experiments. These experiments with Thr-Met do
not only provide an alternative method of quantitation of the
e oxidation but also an example for 1e oxidation of a Met
residue embedded in a peptide for comparison to that of the
free amino acid Met. A quantitative understanding of the 1e
(
(
(
(
13) Hiller, K.-O.; Asmus, K.-D. J. Phys. Chem. 1983, 87, 3682.
14) Sawyer, D. T.; Valentine, J. S. Acc. Chem. Res. 1981, 14, 393.
15) Cohen, S. G.; Ojanpera, S. J. Am. Chem. Soc. 1975, 97, 5633.
16) (a) Yang, S. F.; Ku, H. S.; Pratt, H. K. J. Biol. Chem. 1967, 242,
1
5
274. (b) Pryor, W. A.; Tang, R. H. Biochem. Biophys. Res. Commun. 1978,
8
1, 498.
(
17) Sch o¨ neich, Ch.; Zhao, F.; Madden, K. P.; Bobrowski, K. J. Am.
(18) Handbook of Biochemistry; Long, C., Ed.; E.&F.N. SPON Ltd.:
London, 1968; pp 45-52.
Chem. Soc. 1994, 116, 4641.

Oxidation of Threonylmethionine by Peroxynitrite
J. Am. Chem. Soc., Vol. 119, No. 20, 1997 4751
Scheme 2
spectrum of all transients in solution at this specific time,
generated via reactions 17-21, displayed in Scheme 2.
The initial product immediately after (and during) the laser
3
pulse is CB* (reaction 17) which rapidly reacts with Thr-Met
by formation of a charge transfer complex 9 with the sulfide
9
-1 -1 20
moiety of Thr-Met (reaction 18; k18 ≈ 2 × 10 M
s
).
3
The initial yields of CB* after the laser flash were indepen-
-
5
dently determined in aqueous solution, pH 11.5, as 1.98 × 10
3
-1
M, taking λmax( CB*) ) 535 nm and ꢀ535 ) 6250 M
-
1 21-23
-2
cm
. For 1 × 10 M Thr-Met, reaction 18 is nearly
complete at 200 ns after the pulse (see below) so that there are
3
only minor residual amounts of CB* in the solution at this
time which will not contribute significantly to the composite
spectrum. Generally, radical cation 3 exists in equilibrium 7
+
with (S∴ S) (4) where equilibrium 7 will be located virtually
completely on the side of the dimer at peptide concentrations
-
2
24
+
of 10 M. For Thr-Met (S∴ S) is characterized by an optical
1
7
absorption with λmax ) 480 nm. The original value of ꢀ480 )
-1
-1 17
-1
6
540 M cm
has recently been corrected to ꢀ ) 8690 M
-
1 25
cm
which was taken as a reference value in this paper. The
spectral characteristics of the other reaction products in solution
Figure 1. Absorption spectra recorded 200 ns after laser flash
•-
are well established; i.e., for the ketyl radical anion (CB ) λmax
-
3
photolysis of an Ar-saturated aqueous solution containing 2 × 10
CB, 1 × 10 M Thr-Met, and 2 × 10 M sodium phosphate, pH
.43. The closed circles represent the experimental spectrum. The
experimental spectrum was deconvoluted into the individual spectra
M
-1
-1 21,22
)
(
660 nm and ꢀ660 ) 7660 M cm ,
for the ketyl radical
-
2
-2
•
-1
-1 21,22
CBH ) λmax ) 570 nm and ꢀ570 ) 5200 M cm ,
and for
7
species 10 and 11 λmax ≈ 280 nm and ꢀ280 ≈ (3000 ( 600)
cm . Thus, at 200 ns after the laser flash the composite
spectrum in the range between 360 and 720 nm essentially
-1
-1 26
3
•-
•
+
M
of its components, CB*, CB , CBH , and (S∴ S) , and a computer
fit of the composite spectrum (solid line) obtained by summation of
the absorbances of the individual components at their respective
concentrations.
•
-
•
+
contains contributions from CB , CBH , and (S∴ S) . On the
basis of their known individual spectra and extinction coef-
ficients, the composite spectrum can be resolved into the spectra
oxidation by peroxynitrite is important in view of its biological
relevance: in tissue protein methionine sulfoxide residues can
be enzymatically reduced to Met by methionine sulfoxide
27
of the individual components by a linear regression technique
of the form of eq I where ∆A(λj) is the observed absorbance
1
9
(20) Marciniak, B.; Hug, G. L.; Bobrowski, K.; Kozubek, H. J. Phys.
Chem. 1995, 99, 13560.
reductase whereas 1e oxidation often leads to an irreversible
decomposition of Met residues.
(
21) Inbar, S.; Linshitz, H.; Cohen, S. G. J. Am. Chem. Soc. 1981, 103,
323.
(22) Hurley, J. K.; Linshitz, H.; Treinin, A. J. Phys. Chem. 1988, 92,
151.
(23) Hurley, J. K.; Sinai, N.; Linshitz, H. Photochem. Photobiol. 1983,
8, 9.
24) Bobrowski, K.; Sch o¨ neich, Ch.; Holcman, J.; Asmus, K.-D. J. Chem.
7
5
3
Results
Photooxidation of Thr-Met. Laser Photolysis. Figure 1,
closed circles, shows the optical spectrum recorded at 200 ns
after the laser flash photolysis of an Ar-saturated aqueous
(
Soc., Perkin Trans. 2 1991, 353.
(25) Bobrowski, K.; Pogocki, D. Manuscript in preparation.
(26) Hiller, K.-O.; Asmus, K.-D. Int. J. Radiat. Biol. 1981, 40, 597.
-
2
solution containing 2 × 10 M phosphate buffer, pH 7.43, 2
-
3
-2
×
10 M CB, and 1 × 10 M Thr-Met. It is the composite
(27) Bevington, P. R. Data Reduction and Error Analysis for the Physical
(
19) Brot, N.; Weissbach, H. BioFactors 1991, 3, 91.
Sciences; McGraw-Hill: New York, 1969.

4752 J. Am. Chem. Soc., Vol. 119, No. 20, 1997
Jensen et al.
Figure 2. Concentration vs time profiles for intermediates obtained
after laser flash photolysis of an Ar-saturated aqueous solution,
-
3
-2
-2
containing 2 × 10 M CB, 1 × 10 M Thr-Met, and 2 × 10
M
Figure 3. Concentration vs time profiles for intermediates obtained
sodium phosphate, pH 7.43.
after laser flash photolysis of an Ar-saturated aqueous solution,
-
3
-2
-2
containing 2 × 10 M CB, 1 × 10 M Thr-Met, and 2 × 10
M
change of the composite spectrum, and ꢀi(λj) is the molar
extinction coefficient of the ith species at the jth wavelength of
observation.
sodium phosphate, pH 5.96.
Thr-Met. Photochemical quantum yields can be calculated as
3
3
Φ(product) ) [product]/[ CB*]initial where [ CB*]initial ) 1.98
n
-
5
•-
•
×
10 M (see above). We derive Φ(CB + CBH ) ) 0.4,
∆
A(λ ) ) ꢀ (λ )a
(I)
3
j
∑
i
j
i
indicating that 40% of the reaction of CB* with Thr-Met
proceeds via reactions 20 and 21, and 60% afford physical
quenching (reaction 19), yielding ground state CB. An inspec-
tion of the first 200 ns of the concentration vs time profile
i)1
The linear regression coefficients correspond to cil where ci
is the concentration of the ith transient and l is the optical path
•
-
+
reveals a 1:1 stoichiometry of CB and (S∴ S) , reflecting the
length of the monitoring light. Further details of this method
have been described elsewhere.28 Figure 1 displays a quantita-
1e oxidation of Thr-Met according to reaction 20. The quantum
+
+
yield for (S∴ S) amounts to Φ[(S∴ S) ] ) 0.20, indicating
that 50% of the chemical quenching is accounted for by electron
transfer (reaction 20). The quantum yield for the hydrogen
tive calculation of the spectra of the individual components
contributing to the composite spectrum (closed circles), and the
solid line represents a simulated composite spectrum based on
the calculated concentrations of the individual components. By
application of this linear regression technique to composite
spectra recorded at various times after the laser flash, the
concentration vs time profiles of the individual components can
be monitored, as displayed in Figure 2 for an Ar-saturated
•
•-
transfer reaction 21 is then calculated as Φ(CBH ) ) Φ(CB
•
+
+
CBH ) - Φ[(S∴ S) ] ) 0.20. Figure 3 displays a concentra-
tion vs time profile for all transients at pH 5.96. At this pH
•
-
the protonation of CB is faster and equilibrium 24 is nearly
completely located on the left hand side. The photochemical
quantum yields at pH 5.96 are Φ(CB + CBH ) ) 0.40,
Φ(CBH ) ) 0.20, and Φ[(S∴ S) ] ) 0.20.
•
-
•
-
2
aqueous solution containing 2 × 10 M phosphate buffer, pH
•
+
-
3
-2
7
.43, 2 × 10 M CB, and 1 × 10 M Thr-Met.
Within 300 ns after the laser flash there is a complete decay
An important detail of Figures 2 and 3 is that the protonation
3
+
•-
•-
+
of CB* paralleled by the formation of (S∴ S) , CB , and
CBH . The dimeric sulfide radical cation (S∴ S) remains rather
stable over a time period of 1 µs. However, the formation of
CB is superimposed by a subsequent decay, paralleled by the
formation of additional yields of CBH . This process reflects
the protonation equilibrium 24 (pK24 ) 8.2 ) which is nearly
completely established at ca. 1.1 µs after the laser flash.
of CB does not afford deprotonation of (S∴ S) (cf. reactions
22 and 23; Scheme 2) but rather involves proton transfer from
the solvent or the buffer.
•
+
•
-
Steady-State Photolysis. Steady-state photolysis experi-
•
-4
ments in N -saturated aqueous solutions containing 2.5 × 10
2
21
-2
M CB, 2 × 10 M phosphate buffer, and various concentrations
of Thr-Met were carried out to determine the efficiency of
photochemical acetaldehyde formation, facet,photo, per equivalent
of initial radical cation 3. The laser photolysis experiments had
•
•-
+
CBH h CB + H
(24)
3
shown that chemical quenching of CB* by Thr-Met in
•
-
On the basis of pK24 ) 8.2, we expect that at pH 7.43 ([CB ]/
phosphate buffer proceeds via hydrogen transfer (50%) and
electron transfer (50%). Under N2 each electron and hydrogen
transfer process is associated with the formation of 1 equiv of
•
[
CBH ])theor ) 0.17, in reasonable agreement with our experi-
•
-
•
mental value of ([CB ]/[CBH ])exp ) 0.14 at 1.1 µs after the
•
-
•
laser flash. The combined yields of CB and CBH at 1.1 µs
•-
•
CB /CBH which, at pH < 8, reacts bimolecularly to yield
,2-dihydroxy-1,2-bis(4′-carboxyphenyl)-1,2-diphenylethane. Of
3
after the laser flash are representative of the fraction of CB*
which reacts via chemical quenching (reactions 20 and 21) with
1
this loss of CB (∆CB), 50% is due to the electron transfer
reaction 20. The yields of acetaldehyde and ∆CB were
measured for various concentrations of Thr-Met at pH 6.0 and
(
28) Marciniak, B.; Bobrowski, K.; Hug, G. L. J. Phys. Chem. 1993,
9
7, 11937.

Oxidation of Threonylmethionine by Peroxynitrite
Table 1. Product Yields from the Oxidation of Thr-Met by CB* and Peroxynitritea
J. Am. Chem. Soc., Vol. 119, No. 20, 1997 4753
peroxynitrite (ONOO ) + Thr-Metc
3
3CB* + Thr-Met
-
conditions
entry [Thr-Met], 10-
3
M
pH
atmosphere
f
b
[ONOO ], 10
-
-4
M
f
f
f
∑
d
f
1e,PN
e
acet,photo
acet,PN
Thr-Met(O),PN
NH3,PN
prod,PN
1
2
3
4
5
6
7
1.0
0.5
0.175
1.0
0.5
0.175
1.0
7.4
7.4
7.4
7.4 air
7.4 air
7.4 air
7.4 2.5 × 10
N
N
N
2
2
2
0.78
0.80
0.58
f
f
f
f
5.0
0.012
0.014
0.031
0.014
0.028
0.035
0.050
0.96
0.72
0.44
0.73
0.51
0.35
0.12
0.028
0.033
0.023
0.06
0.07
0.11
1.0
0.015
0.018
0.053
0.018
0.035
0.060
0.064
2.55
0.875
5.0
2.5
0.875
5.0
0.77
0.50
0.81
0.61
0.50
0.19
-
2
M
0.02
-
HCO
3
, air
8
9
0
1
2
3
1.0
0.5
0.175
1.0
0.5
6.0
6.0
6.0
6.0 air
6.0 air
6.0 air
N
N
N
2
2
2
0.18
0.17
0.17
f
f
f
g
g
g
g
g
g
g
g
g
0.76
0.68
0.50
g
g
g
g
g
g
g
g
g
g
g
g
g
g
g
1
1
1
1
5.0
2.5
0.875
0.008
0.015
0.024
0.044
0.088
0.140
0.175
a
b
•+
c
All yields are reported with an error limit of (20%.
f
f
acet,photo ) [acetaldehyde]/[TM(S )] ) [acetaldehyde]/0.5∆CB.
f
product/PN ) [product]/
d
prod,PN ) facet,PN + fThr-Met(O),PN + fNH3,PN. e
f
3
[
loss of TM].
CB + O . Not determined.
2
∑
2
1e,PN ) facet,PN/facet,photo. Not determined because of competitive reaction CB* + O f
1
g
7
.4, respectively, and facet,photo calculated according to eq II. The
derivation of the mathematical expressions for kobs. In alkaline
aqueous solution, peroxynitrite exists predominantly as cis-
results for facet,photo are displayed in Table 1 (column 5).
-
31
15
ONOO , as concluded from Raman spectroscopy and N-
3
2
[
acetaldehyde]
NMR. Peroxynitrite anion is quite stable toward decompo-
facet,photo
)
(II)
-1
1
sition into nitrate with kobs ≈ 0.01 s for pH 8.6 at 25 °C (via
protonation of peroxynitrite to peroxynitrous acid, ONOOH).
Peroxynitrous acid converts rapidly into nitrate with k ) 1.3
0.5[∆CB]
As expected, facet,photo is higher at pH 7.4 due to a more
efficient formation of 6 at higher pH. In control experiments
-
1
1
32
s
at pH < 6.0 and 25 °C. Theoretical calculations predict
-
-
4
an equilibrium between cis-ONOO and the cis,cis isomer of
ONOOH (reaction 25) with pK25 ) 6.8, and the pH-rate profile
for the unimolecular decomposition of peroxynitrite showed a
we photooxidized mixtures of 5 × 10 M Thr-Leu and 5 ×
-
4
1
0
M Gly-Met under conditions similar to those of Thr-Met.
They showed no significant yields of acetaldehyde at both pH
.0 and 7.4, respectively. The design of these control experi-
1
,9b
sigmoidal behavior with pKa at 6.8, indicating that ONOOH
is an important intermediate in the overall conversion of cis-
6
3
ments should be rationalized briefly. Generally, CB* reacts
-
-
9
-1 -1 20
3
ONOO to NO (reactions 25 and 26).
ca. 10 times faster (k ≈ 2 × 10 M
s
) with the sulfide
moiety of Met as compared to a deprotonated N-terminal amino
-
+
8
-1 -1 29
group (e.g., of Ala at alkaline pH, k ≈ 2 × 10 M
s
).
cis,cis-ONOOH h cis-ONOO + H
cis,cis-ONOOH ff NO3- + H+
(25)
(26)
One apparently obvious control experiment for the importance
of Met oxidation in Thr-Met for intramolecular acetaldehyde
formation would, therefore, be the photooxidation of Thr-Leu.
1
8
However, at pH 7.44 ca.15% of Thr-Leu (pKa ≈ 8.2 ) will
We note that, though pKa ) 6.8 appears to be characteristic
-
32
exist in the N-terminal deprotonated form. In the absence of
for the protonation of cis-ONOO to cis,cis-ONOOH (reaction
25), it is difficult to experimentally measure the preferred
solution conformation of ONOOH. In fact, additional conform-
ers of ONOOH might form in the course of the isomerization
to nitrate (reaction 26) such as trans-perp-ONOOH32 and/or
several possible excited states (see the Discussion).
3
any sulfide function, electron transfer between CB* and the
fraction of the N-terminal deprotonated amino group would
become a possibility, directly yielding the N-centered radical
cation 7 (see Scheme 1), an efficient precursor for acetaldehyde.
In a 1:1 mixture of Thr-Leu and Gly-Met at pH 6.0 or 7.4,
3
CB* will predominantly react with the sulfide moiety of Gly-
In general, ONOOH is a much better oxidant as compared
1
9
Met with ca. 60% of this chemical quenching, yielding Gly-
Met(S ). The fact that no acetaldehyde was formed in such
systems shows (i) that intermolecular electron transfer between
Gly-Met(S ) and Thr-Leu does not occur, suggesting that
intermolecular electron transfer between Thr-Met(S ) and the
N-terminal amino group of a second Thr-Met molecule is
to peroxynitrite anion, and earlier results with Met suggest
•
+
20
-
that any oxidation of Met by ONOO may be neglected at pH
e 7.4 due to the fast equilibrium 25 (i.e., for ONOO + Met,
k ) 0.2 M s , and for ONOOH + Met, k ) 2060 M
-
•
+
-1 -1
-1 -1 33
s
).
•
+
Thus, we expect that the oxidation of Thr-Met and Gly-Met at
pH 7.4 exclusively involves ONOOH. As will be discussed in
detail below, ONOOH can directly react with the X-Met
peptides (reaction 27; X ) Thr or Gly), yielding the respective
•
+
unlikely, (ii) that acetaldehyde formation from X-Met(S )
requires X ) Thr, and (iii) that a direct reaction of CB* with
3
the fraction of the deprotonated N-terminus of Thr-Met at both
pH 6.0 and 7.4 can be considered negligible.
+
-
ONOOH + X-Met f X-Met(O) + H + NO2 (27)
Oxidation of Thr-Met by Peroxynitrite. Stopped-Flow
Experiments. By stopped-flow rapid scan UV spectroscopy
we measured the decomposition kinetics of the peroxynitrite
sulfoxides, X-Met(O), or unimolecularly convert into an excited
(
31) Tsai, J.-H. M.; Harrison, J. G.; Martin, J. C.; Hamilton, T. P.; Van
der Woerd, M.; Jablonski, M. J.; Beckman, J. S. J. Am. Chem. Soc. 1994,
116, 4115.
(32) Tsai, H.-H.; Hamilton, T. P.; Tsai, J.-H. M.; Van der Woerd, M.;
Harrison, J.G.; Jablonski, M. J.; Beckman, J. S.; Koppenol, W. H. J. Phys.
Chem. 1996, 100, 15087.
(33) Padmaja, S.; Squadrito, G. L.; Lemercier, J.-N.; Cueto, R.; Pryor,
W. A. Free Radical Biol. Med. 1996, 21, 317.
-
30
anion (ONOO ; λmax ) 302 nm ) at pH 7.4 (25 °C) in the
absence and presence of various concentrations of Thr-Met and,
for comparison, Gly-Met. Some details are important for the
(
29) Bobrowski, K.; Hug, G. L.; Marciniak, B.; Kozubek, H. J. Phys.
Chem. 1994, 98, 537.
30) Hughes, M. N.; Nicklin, H. G. J. Chem. Soc. A 1968, 450.
(

4754 J. Am. Chem. Soc., Vol. 119, No. 20, 1997
Jensen et al.
-
under pseudo-first-order conditions ([ONOO ] , [X-Met]), that
d[ONOO-]
-
) kobs,pH 7.4[ONOO-]
(III)
(IV)
dt
kobs,pH 7.4 ) k26,pH 7.4 + k27,pH 7.4[peptide]
+
[
H ]
kobs,pH 74 ) (k + k [peptide])
(V)
26
27
+
(
)
[
H ] + K25
ONOOH is the key intermediate for unimolecular decomposition
1
9
of peroxynitrite into nitrate and for oxidation of X-Met, that
reaction 27 represents the predominant pathway of X-Met
oxidation at pH 7.4, and that equilibrium 25 is rapid compared
1
,9
to reactions 26 and 27.
The slopes of the straight lines yield k27,pH 7.4 ) 283 ( 3
M-
1
s
-1
for Thr-Met and k27,pH 7.4 ) 280 ( 2 M
-1 -1
s
for Gly-
Met. These values are comparable to the rate constants for the
bimolecular reaction of peroxynitrite with 4-methylthiobutanoic
-
1
-1
9a
33
acid (kpH 7.4 )283 M s ) and Met [kpH 7.4 ) 181 -416
-
1
-1
-7
-1 1,9b
M
s ]. On the basis of K25 ) 1.58 × 10
M
and, at
+
-8
pH 7.4, [H ] ) 4 × 10 M, we derive that for Thr-Met k27 )
-1
-
1
-1
1
400 ( 15 M
s
and for Gly-Met k27 ) 1386 ( 10 M
-
1
s . The value for the unimolecular decomposition of perox-
ynitrite in the absence of peptide corresponds to k26,pH 7.4 ) 0.27
-
1
(
0.01 s (see above).
-
4
Product Yields. When 5 × 10 M peroxynitrite reacted
-3
with 1.0 × 10 M Thr-Met in N2-saturated aqueous phosphate
-
2
-4
Figure 4. kobs vs concentration of (a) Thr-Met and (b) Gly-Met for
the reaction of peroxynitrite with both peptides in aqueous solution
containing 0.1 M sodium phosphate buffer, pH 7.4. The rate constants
were obtained by following the disappearence of the 302 nm absorbance
of peroxynitrite anion using stopped-flow rapid scan UV spectroscopy.
buffer (2 × 10 M, pH 7.4), (2.5 ( 0.5) × 10 M Thr-Met
was lost, but only very small yields of acetaldehyde [(3.0 (
-6
0
.3) × 10 M] were formed together with small yields of
-6
ammonia, NH3 [(7 ( 1) × 10 M], and major yields of
-
4
threonylmethionine sulfoxide, Thr-Met(O) [(2.4 ( 0.5) × 10
-
4
M]. In a first approximation, a loss of ca. 2.5 × 10 M Thr-
Met under these experimental conditions is not unexpected on
the basis of our experimentally derived rate constants for the
reaction of peroxynitrite with Thr-Met at pH 7.4. Taking
species which subsequently reacts with the peptides via 1e
oxidation. At pH 7.4, we observed very little 1e oxidation of
-
4
the peptides for [Thr-Met] > 1.75 × 10 M (see below),
indicating that reactions 25-27 are sufficient for a kinetic
analysis of our stopped-flow experiments.
-1
-1 -1
k26,pH 7.4 ) 0.27 s and k27,pH 7.4 ) 283 M
s
for Thr-Met,
-3
competition kinetics predict that in the presence of 1 × 10
The kinetics of peroxynitrite decomposition were first-order
-4
M Thr-Met ca. 51% (corresponding to 2.55 × 10 M) of the
-1
in the absence of peptides with kobs ) 0.27 ( 0.01 s , in good
added peroxynitrite should react with Thr-Met and the residual
agreement with earlier measurements at pH 7.4 which yielded
49% of peroxynitrite should suffer unimolecular decomposition.
-
1 1
-1 9b
kobs ) 0.25 ( 0.04 s
and 0.26 ( 0.01 s , respectively. In
The efficiency of peroxynitrite-mediated product formation
general, the concentration vs time profiles of peroxynitrite anion
could be well fitted with a single exponential function. Only
at very early time points a biexponential fit was slightly better.
This can be rationalized by the fact that mixing of the alkaline
can be calculated as fprod,PN ) [product]/[lost Thr-Met]. Thus,
the efficiencies of product formation are facet,PN ) 0.012 for
acetaldehyde, fNH3,PN ) 0.028 for ammonia, and fThr-Met(O),PN
)
0.96 for threonylmethionine sulfoxide, as displayed in Table
, entry 1. All three products accounted for 100% of the
(pH 12) stock solution of peroxynitrite with the phosphate-
1
buffered stock solution of the peptide (pH 7.4) causes a rapid
equilibration of peroxynitrite with peroxynitrous acid according
to reaction 25. However, the fact that equilibrium 25 is fast as
compared to the subsequent processes (i.e., reactions 25 and
converted Thr-Met, i.e., ∑prod,PN ) facet,PN + fNH3,PN + fThr-Met(O),PN
)
1.0 (∑prod,PN < 1.0 for lower concentrations of Thr-Met; see
17
below). As expected, the oxidation of Ala-Met or Thr-Leu
by peroxynitrite did not yield acetaldehyde. The absence or
presence of O2 had little influence on the yields of acetaldehyde
but apparently some small influence on the yields of Thr-Met-
2
6) is evident from the result that, except for very early time
points, the concentration vs time profile for peroxynitrite anion
can be well fitted with a single exponential function. A similar
kinetic scheme was derived for the radiative and nonradiative
decay of protonated and nonprotonated triplet benzophenone
in very acidic solution.34 In the presence of peptides the kinetics
changed to pseudo-first-order where kobs ()(ln 2)/t1/2) increased
with increasing peptide concentrations. Plots of kobs vs peptide
concentration were linear, as displayed in Figure 4.
(O) (compare entries 1-3 with entries 4-6). However, it
remains to be shown whether this effect may be caused by
different levels of CO2 in the oxygenated and the N2-saturated
solutions, respectively (see also below). There was no influence
-4
III
3-
of added 5.0 × 10
M [Fe (CN)6]
on the yields of
acetaldehyde and Thr-Met(O), indicating that the additional
presence of a potent electron acceptor did not affect the product
yields. This result is important with regard to the fact that
peroxynitrite itself (in particular ONOOH) is a potent electron
Equations III-V represent mathematical expressions for
kobs,pH 7.4, derived on the basis that the kinetics were measured
acceptor with E°′,
red,pH 7.4
(ONOO , 2 H /NO2,aq) ) 1.4 V. With
increasing concentrations of Thr-Met at constant ratios of [Thr-
-
+
1
(
34) Rayner, D. M.; Wyatt, P. A. H. J. Chem. Soc., Faraday Trans. 2
1
974, 70, 945.

Oxidation of Threonylmethionine by Peroxynitrite
J. Am. Chem. Soc., Vol. 119, No. 20, 1997 4755
-
Met]:[peroxynitrite] ) 2:1, facet,PN decreased whereas fThr-Met(O),PN
increased. This feature of facet,PN is different from the efficiency
of photochemical acetaldehyde formation, facet,photo, where
increasing concentrations of Thr-Met promoted no (at pH 6.0)
or even a slight increase (at pH 7.4) of facet,photo (see Table 1,
column 5), and bears important mechanistic information (see
the Discussion). We note that at concentrations of [Thr-Met]
formation of an adduct between ONOO and CO240 (present
-
+
through the equilibrium HCO3 + H h H2O + CO2) in
competition with the reaction of peroxynitrite with Thr-Met.
Potential structures of such adducts between peroxynitrite and
40,41
CO2 have been proposed
but not experimentally confirmed.
Formation of Methional from Met and Thr-Met. As
displayed in reactions 3, 5, and 6, methional is a stable molecular
product originating from the 1e oxidation of Met. One potential
mechanistic problem associated with the formation of ethylene
from Met is the fact that ethylene may result not only from a
direct decomposition of a Met-derived intermediate but also
from further oxidation of methional, a product of the decom-
position of 1. Here, we have included experimental evidence
that methional is, in fact, a product from the reaction of
peroxynitrite with Met. We reacted different concentrations of
-
3
<
1.0 × 10 M, ∑prod,PN < 1.0. This suggests that at lower
concentrations of Thr-Met peroxynitrite may react via additional
channels with Thr-Met, possibly via hydrogen abstraction. In
fact, MALDI-TOF mass spectrometric analysis of our reactions
indicated products with molecular masses corresponding to
either MW(TM) + MW(NO) or MW(TM) + MW(O2) - 2MW-
(H) and MW(TM) + MW(NO3) (TM ) Thr-Met) which remain
to be characterized. However, this detail does not affect our
calculations of the efficiency of 1e oxidation of Thr-Met by
peroxynitrite, determined by measurement of acetaldehyde, since
we have provided a reference value for every concentration of
Thr-Met by measurement of the photochemical efficiency of
acetaldehyde formation, facet,photo (see also the Discussion). In
control experiments we confirmed that acetaldehyde was not a
substrate for peroxynitrite under our reaction conditions. In a
-
3
-4
-4
Met (1 × 10 , 5 × 10 , 1.75 × 10 M) with peroxynitrite
-2
at ratios of [Met]:[peroxynitrite] ) 2:1 in 2 × 10 M phosphate
buffer, pH 7.4, similarly to the experimental system described
-
3
for Thr-Met (see above). The exposure of 1 × 10 M Met to
-
4
-4
5 × 10 M peroxynitrite resulted in the loss of 3.3 × 10
M
-5
Met, accompanied by the formation of 2.0 × 10 M methional.
Thus, the efficiency for methional formation by peroxynitrite,
fmethional,PN ) [methional]/[loss of Met] corresponds to fmethional,PN
) 0.062. With decreasing concentrations of Met, fmethional,PN
-
5
typical experiment the reaction of 8.75 × 10 M peroxynitrite
with 1.75 × 10 M Thr-Met yielded 1.4 × 10 M acetalde-
hyde at pH 7.4. When a solution containing 1.75 × 10
Thr-Met and 9 × 10 M acetaldehyde was reacted with 8.75
10 M peroxynitrite, the final concentration of acetaldehyde
-
4
-6
-
4
-4
M
increased to fmethional,PN ) 0.15 for 5 × 10 M Met and
-
6
-4
fmethional,PN ) 0.31 for 1.75 × 10 M Met. Thus, in particular
-5
×
at lower Met concentrations methional accounts for a significant
fraction (up to 31%) of the products formed during the oxidation
of Met by peroxynitrite. When we exposed Thr-Met to
peroxynitrite, there was no formation of methional at all
-
5
after completion of the reaction was 1.04 × 10 M, i.e., an
amount corresponding to the sum of initially added acetaldehyde
and the yield of acetaldehyde expected on the basis of an
exclusive reaction of peroxynitrite with Thr-Met even in the
presence of acetaldehyde.
-
4
-4
concentrations of Thr-Met, 1.75 × 10 , 5 × 10 , and 1 ×
-
3
1
0
M.
When 1.75 × 10-4 M Thr-Met was oxidized by 8.75 × 10
-5
-
3
-2
Discussion
M peroxynitrite in the presence of 3.5 × 10 or 1.75 × 10
M methanol, there was only a 10% or 20% decrease of facet,PN,
Quantification of the 1e Oxidation of Thr-Met by Peroxy-
nitrite. Scheme 1 displays two potential pathways according
to which sulfide radical cations from Thr-Met produce acetal-
dehyde, (i) the formation of 5 (reactions 8 and 9) with
subsequent conversion to 6 (reaction 10) or (ii) the deprotonation
of the N-terminal amino group of 3 or 4 (reactions 14 and 15)
to allow direct formation of 6 (reaction 16). We note, however,
that acetaldehyde formation is only one possible pathway of
the decomposition of threonylmethionine sulfide radical cations.
Competing pathways include deprotonation in the R-position
to the sulfur such as shown in reactions 22 and 23 (Scheme 2).
In general, these deprotonation pathways are more efficient at
respectively. This result indicates that acetaldehyde is not the
•
product of a reaction of free hydroxyl radicals (HO ) with Thr-
Met in the peroxynitrite system. Initially, it had been suggested
that hydroxyl radicals may be generated by homolytic cleavage
of peroxynitrite (reaction 28),35 but both theoretical and
1
•
•
ONOOH f HO + NO
(28)
2
experimental36,37 evidence against such a reaction has now been
presented. If free hydroxyl radicals would have been responsible
for acetaldehyde formation from Thr-Met in our systems,
4
2
-
3
higher pH and lower sulfide concentrations. This fact may
rationalize our result that photochemical acetaldehyde formation
methanol would have reduced facet,PN by 66% (3.5 × 10
methanol) and 91% (1.75 × 10 M methanol), respectively,
on the basis of k(HO +Thr-Met) ≈ 9.8 × 10 M and
k(HO +CH3OH) ) 9.7 × 10 M s .
In the presence of a physiological concentration of 2.5 ×
M HCO3 , the product yields were significantly changed
Table 1, entry 7), as we observed significantly higher values
M
-
2
-
4
•
9
-1 -1 38
at pH 7.4 was less efficient for 1.75 × 10 M Thr-Met as
s
-4 -3
•
8
-1 -1 39
compared to 5 × 10 and 1 × 10 M Thr-Met. However,
the pathways leading to acetaldehyde also benefit from higher
pH, rationalizing why acetaldehyde formation was generally
more efficient at pH 7.4 as compared to pH 6.0. A quantitative
prediction of acetaldehyde formation as a function of pH and
Thr-Met concentration is currently not possible as not all rate
constants for the individual (competing) processes are known.
However, a semiempirical approach allows this prediction. We
have measured the primary yields of sulfide radical cations,
-
2
-
1
0
(
for facet,PN but significantly lower values for fThr-Met(O),PN.
Mechanistically, these features can be rationalized by the rapid
(
35) Beckman, J. S.; Beckman, T. W.; Chen, J.; Marshall, P. A.; Freeman,
B. A. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1620-1624.
36) (a) Lemercier, J.-N.; Squadrito, G. L.; Pryor, W. A. Arch. Biochem.
(
Biophys. 1995, 321, 31. (b) Pryor, W. A.; Jin, X.; Squadrito, G. L. J. Am.
Chem. Soc. 1996, 118, 3125.
(40) Lymar, S. V.; Hurst, J. K. J. Am. Chem. Soc. 1995, 117, 8867.
Lymar, S. V.; Jiang, Q.; Hurst, J. K. Biochemistry 1996, 35, 7855. Uppu,
R. M.; Squadrito, G. L.; Pryor, W. A. Arch. Biochem. Biophys. 1996, 327,
335.
(41) Houk, K. N.; Condroski, K. R.; Pryor, W. A. J. Am. Chem. Soc.
1996, 118, 13002.
(
37) Goldstein, S.; Squadrito, G. L.; Pryor, W. A.; Czapski, G. Free
Radical Biol. Med. 1996, 21, 965.
38) Sch o¨ neich, Ch.; Yang, J. J. Chem. Soc., Perkin Trans 2 1996,
15.
39) Buxton, G. V.; Greenstock, C. L.; Helman, W. P.; Ross, A. B. J.
Phys. Chem. Ref. Data 1988, 17, 513.
(
9
(
(42) M o¨ nig, J.; Goslich, R.; Asmus, K.-D. Ber. Bunsen-Ges. Phys. Chem.
1986, 90, 115.

4756 J. Am. Chem. Soc., Vol. 119, No. 20, 1997
Jensen et al.
+
•
-
-
2
represented by (S∴ S) (4), during the oxidation of Thr-Met
2
CBH f ( O CPh)(Ph)C(OH)C(OH)(Ph)(PhCO ) (30)
2
3
by CB* using laser photolysis. The only other primary reaction
products formed during laser photolysis are the R-(alkylthio)-
d[CBH•]
3
d[ CB*]
Φ(CB•- + CBH ) - k [CBH ] (VII)
•
• 2
alkyl radicals 10 and 11, but these do not cause acetaldehyde
)
30
formation.17 Sulfide radical cation dimers can be conveniently
dt
dt
observed within 1.1 µs after the laser flash but earlier pulse
3
the quantum yield for chemical quenching of CB* according
to reactions 20 and 21 (Scheme 2), measured immediately after
the laser flash (i.e., compare Figures 2 and 3). Subsequently,
radiolytic measurements have shown that such species decay
-
5
-3
with t1/2 e 1 × 10 s at peptide concentrations of e1 × 10
43
M, i.e., conditions representative for our steady-state photolysis
and peroxynitrite experiments. For (S∴ S) (4), this subsequent
•
-
•
on a longer time scale, CB accepts a proton to yield CBH .
+
decay yields acetaldehyde, 8, 10, and 11. (Some additonal
•
d[CBH ]/dt ) 0
(VIII)
(IX)
decarboxylation has been observed for hydroxyl radical-initiated
24
3
oxidations of X-Met peptides but was negligible for CB*-
initiated oxidations,20 suggesting that branching leading to
decarboxylation may not necessarily involve dimeric sulfide
radical cations.) By measurement of the efficiency of photo-
chemical acetaldehyde formation per threonylmethionine sulfide
radical cation, facet,photo, for each concentration of Thr-Met at
pH 6.0 and 7.4, we obtained reference values which can be used
for the exact calculation of the 1e oxidation efficiency of
peroxynitrite by measurement of the yields of acetaldehyde in
the peroxynitrite/Thr-Met systems. For example, the efficiency
of acetaldehyde formation during the anaerobic oxidation of 1
3
•-
•
1/2
(
d[ CB*]/dt)Φ(CB +CBH )
•
[CBH ] )
s
[
]
k30
In our steady-state photolysis experiments we measured ∆CB
as a function of time which is related to the chemical quenching
of CB* through eq X. Combining eqs X and IX yields eq XI.
3
3
∆CB
dt
d[ CB*]
•-
•
)
Φ(CB +CBH )
(X)
dt
•
1/2
-
3
-4
×
10 M Thr-Met by 5 × 10 M peroxynitrite was facet,PN )
[CBH ] ) [(∆CB/dt)/k ]
(XI)
s
30
0
.012. Since under such conditions the decomposition of
threonylmethionine sulfide radical cations yields 0.78 mol of
acetaldehyde/mol of sulfide radical cation, determined photo-
chemically (i.e., facet,photo ) 0.78), the application of eq VI reveals
that a fraction of f1e,PN ) 0.015 (i.e., 1.5%) of the peroxynitrite-
induced oxidation of Thr-Met must have proceeded via 1e
oxdiation.
For our steady-state photolysis conditions we determined that
-
7
-1
•
-8
∆CB/dt ) 7.4 × 10 M s so that [CBH ]s ) 2.9 × 10 M.
-
3
+
For [Thr-Met] e 1 × 10 M, the overall decay of (S∴ S)
(reaction 31) follows first-order kinetics with t1/2 e 1 × 10
-
5
4
-1
s, corresponding to k31 g 6.9 × 10 s .
+
(
S∴ S) f products
(31)
f1e,PN ) facet,PN/facet,photo
(VI)
The rate for reaction 31 can be simply expressed as V31 )
+
+
k31[(S∴ S) ] and that for reaction 29 as V29 ) k29[(S∴ S) ]‚
In a similar way the efficiencies for 1e oxidation of Thr-Met
by peroxynitrite were calculated for all experimental conditions,
as displayed in the last column of the Table 1. In principle,
the value of facet,photo is not expected to be influenced by the
absence or presence of oxygen. However, the photochemical
measurement of facet,photo in oxygenated solutions is difficult as
•
10
-1 -1
[
CBH ]s. If we approximate that k29 e 10
M
s , it follows
that V31/V29 g 238. Thus, reaction 29 will not contribute to the
+
disappearence of (S∴ S) under steady-state photolysis condi-
tions.
The Mechanism. On the basis of the values for f1e,PN, we
conclude that the efficiency for 1e oxidation of Thr-Met by
peroxynitrite increases with decreasing concentrations of Thr-
Met. These results are in line with a mechanism displayed in
3
•-
•
oxygen reacts rapidly with CB*, CB , and CBH , producing
singlet oxygen and superoxide, respectively, which may com-
+
44
petitively react with Thr-Met or (S∴ S) . Therefore, the values
facet,photo obtained photochemically in N2-saturated solutions were
used for the calculation of f1e,PN for the peroxynitrite-mediated
processes in oxygenated solutions.
9
Scheme 3, analogous to reactions proposed by Pryor et al. for
the oxidation of Met.
-
In general ground state ONOO should be able to oxidize
Thr-Met (reaction 32), although k32 is expected to be small.9
However, by analogy to earlier results with Met it appears that,
at pH e 7.4, it will be predominantly ONOOH which oxidizes
Thr-Met to Thr-Met(O) via oxygen transfer in a bimolecular
,33
One additonal feature should be discussed briefly. We had
•
-
shown that protonation of any CB , initially formed during
the photochemical 1e oxidation of Thr-Met (reaction 20), does
+
not affect the lifetime of (S∴ S) . However, for an exact
9
,33
reaction (reaction 33).
At this point we cannot define to what
quantification of facet,photo we have to exclude that, at a later
32
•
+
extent the individual conformers of ONOOH, cis-cis, cis-perp,
or trans-perp would contribute to reaction 33. In competition
with these bimolecular pathways, ONOOH transforms unimo-
lecularly into nitrate according to two independent pathways.
The first pathway proceeds via a metastable reactive intermedi-
ate, ONOOH* (reaction 34), which constitutes a precursor not
only for the subsequent unimolecular rearrangement into nitrate
stage, CBH does not react significantly with (S∴ S) under the
conditions of steady-state photolysis (i.e., via reaction 29),
+
thereby reducing the yields of (S∴ S) , effectively available for
acetaldehyde formation.
•
+
+
CBH + (S∴ S) (4) f CB + H + 2Thr-Met (29)
(
reaction 35) but also for the 1e oxidation of Thr-Met (reaction
Equations VII-XI allow the calculation of an upper limit of
•
•
36) or, potentially, for hydrogen transfer reactions (reaction 37).
the steady-state concentration of CBH , [CBH ]s, assuming that
37
•
In the second pathway, recently established, ONOOH directly
rearranges into nitrate apparently without the intermediacy of
an excited state (reaction 38). Recently, two potential structures
of ONOOH* have been located using density functional theory
methods (the Becke3LYP functional and 6-31G* basis set).
They are essentially hydrogen-bonded radical pairs ( OH‚‚‚
CBH decays solely according to reaction 30 with 2k30 ) 1.8
9
-1 -1 21
•-
•
×
10 M s . In eq VII, Φ(CB +CBH ) corresponds to
(
43) By analogy to Gly-Met: Bobrowski, K.; Holcman, J. Int. J. Radiat.
Biol. 1987, 52, 139.
44) Miller, B. L.; Williams, T. D.; Sch o¨ neich, Ch. J. Am. Chem. Soc.
996, 118, 11014.
41
(
•
1

Oxidation of Threonylmethionine by Peroxynitrite
J. Am. Chem. Soc., Vol. 119, No. 20, 1997 4757
-
3
Scheme 3
yields significant amounts of methional (e.g., 6.2% for 1 × 10
M Met), another 1e oxidation product of Met. At pH 6.0 there
was a higher extent of 1e oxidation for the reaction of
peroxynitrite with Thr-Met where, e.g., f1e,PN ) 0.14 for 1.75
-
4
×
10 M Thr-Met (i.e., 14% of the reaction of peroxynitrite
proceeded via 1e-oxidation). At present, we cannot explain this
higher efficiency at pH 6.0 as compared to pH 7.4 but note
that maximum 1e oxidation yields were also observed for the
reaction of peroxynitrite with dimethyl sulfoxide and 2,2′-
azinobis(3-ethyl-1,2-dihydrobenzothiazoline-6-sulfonate) (ABTS)
4
6
at pH 6.0.
One interesting question to pursue in future experiments
would be whether ONOOH* can also perform 2e oxidations
(sulfoxide formation) in its reaction with Met residues. For
example, one could envisage two consecutive 1e oxidation reac-
tions: the first 1e oxidation process between ONOOH* and R2S
•
would lead to a sulfide radical cation and NO2 which, within
•
ON O) with different OH‚‚‚O bond lengths and show diradical
a solvent cage, could react to yield a sulfide dication (reactions
39 and 40). The latter would react with water to form sulfoxide
(reaction 41). Charge stripping experiments in the gas phase
have, in fact, demonstrated the possibility of removal of an
character with degenerate triplet and singlet states. Depending
on the level of theory, these hydrogen-bonded complexes are
5.6 or 15.0 kcal/mol higher in energy as compared to ground
state cis,cis-ONOOH. Interestingly, Tsai et al. also located
a triplet instability for cis-ONOO which is 7.9 kcal/mol (MP2/
-311+(d)) or 14.1 kcal/mol (Becke3LYP) higher in energy than
singlet cis-ONOO . In addition, they located the triplet state
of trans-ONOO , being 13.0 kcal/mol (MP2/6-311+(d)) or 9.4
kcal/mol (Becke3LYP) higher in energy as compared to singlet
1
41
32
•+ 47
electron from R S .
2
-
•
+ •
-
R S + ONOOH* f [R S / NO /HO ]
(39)
6
2
2
2
cage
-
[R S / NO /HO ] ff R2S2+ + NO2 + HO (40)
•
+ •
-
-
-
-
2
2
cage
2
+
+
-
R2S + H O f R SO + 2H
(41)
cis-ONOO . Thus, theoretically both the hydrogen-bonded
2
2
radical pair41 and the triplet states (which do not necessarily
have to be identical) could account for the 1e oxidation of Thr-
Met by peroxynitrite. In addition, the location of these excited
states helps to rationalize potential hydrogen abstraction reac-
tions of ONOOH (peroxynitrite was reported to initiate lipid
32
We conclude that the reactions of peroxynitrite with Met
residues might show quite different results depending on whether
Met is present as a free amino acid or embedded in a peptide
or protein. Comparable observations have been made when Met
and Met-containing peptides were subjected to the oxidation
4
5
peroxidation ) and may serve to explain the fact that we
observed additional channels of the reaction of peroxynitrite
with Thr-Met at lower Thr-Met concentrations which led to the
formation of species with molecular weights of MW(Thr-Met
NO) or MW(Thr-Met + O2 - 2H) and MW(Thr-Met + NO3)
see above). These products potentially form after hydrogen
24,48
3
12,20
by hydroxyl radicals
and CB*.
Future studies must now
show whether and under what conditions peroxynitrite will react
with protein-bound Met residues via 1e-oxidation and whether
such reactions occur in ViVo.
+
(
Experimental Section
abstraction from any C-H bond of Thr-Met (reaction 37) and
subsequent net addition of NOx or O2.
See the Supporting Information.
The fact that facet,PN is higher for lower concentrations of Thr-
Met is simply a result of an increasing fraction of unimolecular
formation of ONOOH* (reaction 34) at the expense of the
bimolecular reaction 33, the rate of which depends on the
concentration of Thr-Met.
Acknowledgment. This work was supported by the NIH
Grant PO1AG12993-02) and the American Heart Association
Grant KS-96-GB-63) (to Ch.S.), by an Allen M. and Lila Self
Fellowship (to B.L.M.), and by the Office of Basic Energy
Sciences of the U.S. Department of Energy. This is Document
No. NDRL-3984 from the Notre Dame Radiation Laboratory.
(
(
An important result is that we are able to exactly quantify
the extent of 1e oxidation of Thr-Met by means of measuring
acetaldehyde formation. Thus, at pH 7.4 only 1.5% and 1.8%
of the reaction of peroxynitrite with Thr-Met proceeded via 1e
oxidation in anaerobic and aerobic atmospheres, respectively.
This result is quite different from earlier findings with the free
Supporting Information Available: Experimental section
(4 pages). See any current masthead page for ordering and
Internet access instructions.
JA964031Z
9
amino acid Met where, on the basis of ethylene measurements,
(46) Crow, J. P.; Spruell, C.; Chen, J.; Gunn, C.; Ischiropoulos, H.; Tsai,
at least 8% of the reaction of Met with peroxynitrite afforded
-
3
M.; Smith, C. D.; Radi, R.; Koppenol, W. H.; Beckman, J. S. Free Radical
Biol. Med. 1994, 16, 331.
(47) Drewello, T.; Lebrilla, C. B.; Asmus, K.-D.; Schwarz, H. Angew.
Chem. 1989, 101, 1247; Angew. Chem., Int. Ed. Engl. 1989, 28, 1275.
(48) Hiller, K.-O.; Masloch, B.; G o¨ bl, M.; Asmus, K.-D. J. Am. Chem.
Soc. 1981, 103, 2743.
1
e oxidation at a substrate concentration of [Met] ) 1 × 10
M. In addition to ethylene, the oxidation of Met by peroxynitrite
(45) Radi, R.; Beckman, J. S.; Bush, K. M.; Freeman, B. A. Arch.
Biochem. Biophys. 1991, 288, 481.