Superoxide reaction with tyrosyl radicals generates para -hydroperoxy and para-hydroxy derivatives of tyrosine

Article abstract of DOI:10.1021/ja307215z

Tyrosine-derived hydroperoxides are formed in peptides and proteins exposed to enzymatic or cellular sources of superoxide and oxidizing species as a result of the nearly diffusion-limited reaction between tyrosyl radical and superoxide. However, the structure of these products, which informs their reactivity in biology, has not been unequivocally established. We report here the complete characterization of the products formed in the addition of superoxide, generated from xanthine oxidase, to several peptide-derived tyrosyl radicals, formed from horseradish peroxidase. RP-HPLC, LC-MS, and NMR experiments indicate that the primary stable products of superoxide addition to tyrosyl radical are para-hydroperoxide derivatives (para relative to the position of the OH in tyrosine) that can be reduced to the corresponding para-alcohol. In the case of glycyl-tyrosine, a stable 3-(1-hydroperoxy-4- oxocyclohexa-2,5-dien-1-yl)-l-alanine was formed. In tyrosyl-glycine and Leu-enkephalin, which have N-terminal tyrosines, bicyclic indolic para-hydroperoxide derivatives were formed ((2S,3aR,7aR)-3a-hydroperoxy-6-oxo-2, 3,3a,6,7,7a-hexahydro-1H-indole-2-carboxylic acid) by the conjugate addition of the free amine to the cyclohexadienone. It was also found that significant amounts of the para-OH derivative were generated from the hydroxyl radical, formed on exposure of tyrosine-containing peptides to Fenton conditions. The para-OOH and para-OH derivatives are much more reactive than other tyrosine oxidation products and may play important roles in physiology and disease.

Full text of DOI:10.1021/ja307215z

Article
pubs.acs.org/JACS
Superoxide Reaction with Tyrosyl Radicals Generates para-
Hydroperoxy and para-Hydroxy Derivatives of Tyrosine
Matías N. Moller,†,‡ Duane M. Hatch,† Hye-Young H. Kim,† and Ned A. Porter*,†
̈
†Department of Chemistry and Vanderbilt Institute of Chemical Biology, Vanderbilt University, Nashville, Tennessee 37235, United
States
‡
Instituto de Química Biolog
́
ica, Facultad de Ciencias, and Center for Free Radical and Biomedical Research, Universidad de la
Republica, Montevideo, Uruguay
́
S
* Supporting Information
ABSTRACT: Tyrosine-derived hydroperoxides are formed in
peptides and proteins exposed to enzymatic or cellular sources of
superoxide and oxidizing species as a result of the nearly
diffusion-limited reaction between tyrosyl radical and superoxide.
However, the structure of these products, which informs their
reactivity in biology, has not been unequivocally established. We
report here the complete characterization of the products formed
in the addition of superoxide, generated from xanthine oxidase, to
several peptide-derived tyrosyl radicals, formed from horseradish
peroxidase. RP-HPLC, LC-MS, and NMR experiments indicate
that the primary stable products of superoxide addition to tyrosyl
radical are para-hydroperoxide derivatives (para relative to the position of the OH in tyrosine) that can be reduced to the
corresponding para-alcohol. In the case of glycyl-tyrosine, a stable 3-(1-hydroperoxy-4-oxocyclohexa-2,5-dien-1-yl)-L-alanine was
formed. In tyrosyl-glycine and Leu-enkephalin, which have N-terminal tyrosines, bicyclic indolic para-hydroperoxide derivatives
were formed ((2S,3aR,7aR)-3a-hydroperoxy-6-oxo-2,3,3a,6,7,7a-hexahydro-1H-indole-2-carboxylic acid) by the conjugate
addition of the free amine to the cyclohexadienone. It was also found that significant amounts of the para-OH derivative
were generated from the hydroxyl radical, formed on exposure of tyrosine-containing peptides to Fenton conditions. The para-
OOH and para-OH derivatives are much more reactive than other tyrosine oxidation products and may play important roles in
physiology and disease.
INTRODUCTION
■
Oxidation of tyrosine by free radicals and other reactive species
occurs under physiological conditions and is usually augmented
in many pathologies as a consequence of increased production
of reactive species.1−3 Reported oxidative modifications of
tyrosine generally involve substitutions at the ortho position of
the tyrosine phenol (relative to the OH), such as the case for o-
nitrotyrosine, o,o′-dityrosine, and o-chlorotyrosine (Figure
1).1−7 These stable products are thermodynamically favored
by re-aromatization of the phenolic ring subsequent to
substitution. The para position is usually considered much
less reactive because of steric hindrance toward substitution
along with the loss of the rearomatization stabilization.
However, there are some reactions that modify tyrosine in
the para position, including the reaction between tyrosyl radical
and lipid peroxyl radicals8 and the reaction between tyrosine
and singlet oxygen.9,10
Figure 1. Tyrosine structure showing the different ring positions and
different products of tyrosine oxidation involving modifications at the
ortho and para positions.
Superoxide addition to tyrosyl radical was proposed to occur
at both ortho and para positions,11 but there is still some debate
about the structure of the products formed from this reaction.
In the first study of this reaction, Jin, Leitich, and von Sonntag
isolated two products, a mixture of diasteroisomers of a bicyclic
p-hydroxy derivative of tyrosine (assigned to (2S,3aR,7aS)- and
(2S,3aS,7aR)-3a-hydroxy-6-oxo-2,3,3a,6,7,7a-hexa-hydro-1H-in-
dole-2-carboxylic acids, HOHICA, Figure 1).11 These products
Received: July 23, 2012
Published: September 18, 2012
© 2012 American Chemical Society
16773
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
Scheme 1
was added 140 nM HRP and flushed with oxygen for 3 min. Then, 1
mM acetaldehyde was added, and then XO was added, mixed, and left
to react for 30 min at room temperature (23 °C), with mechanical
agitation and protected from light. Each condition was done in
triplicate. The production of superoxide by XO was measured previous
to the assay by cytochrome c reduction15 and then was added to the
were proposed to derive from intermediate hydroperoxides
formed at both ortho and para positions that could undergo
intramolecular conjugate addition by the free amine followed
by hydrolysis.11
More recent studies have observed the formation of tyrosine
hydroperoxides from the addition of superoxide to tyrosyl
radical in many tyrosine-containing peptides12−14 and in
myoglobin15 using LC-MS and hydroperoxide assays. Some
of these studies included xanthine oxidase (XO) as an
enzymatic source of superoxide and hydrogen peroxide12,13,15
and horseradish peroxidase (HRP) to catalyze the formation of
tyrosyl radicals.12,13 Interestingly, the same products were
observed after exposing the peptides to activated neutrophils,14
supporting the plausibility of these reactions in vivo. Character-
ization by mass spectrometry confirmed that hydroperoxide
and alcohol derivatives were formed in the tyrosine residues,
but the structures were not unambiguously established.12−16
Nonetheless, mechanisms and structures show products from
exclusive addition of superoxide to the ortho position, see
Scheme 1.12,14−18
Recent studies in our laboratory demonstrated that lipid
peroxyl radicals undergo reaction exclusively at the tyrosyl para
position, an observation that led us to compare this result with
the ortho−para selectivity of tyrosyl-superoxide adduction.8
We report here that tyrosine-containing peptides oxidized
with xanthine oxidase−horseradish peroxidase give para-
coupled adducts as the major products. The formation of
para-substituted cyclohexadienones from tyrosine oxidation is
of interest due to its general novelty as well as to challenge the
“steric hindrance” and “rearomatization stabilization” argument
invoked to explain the reactions of other reactive species with
tyrosine. Furthermore, the cyclohexadienone product is a
bifunctional electrophile that potentially reacts with one or two
nucleophiles to generate novel Michael-type adducts to
promote protein covalent cross-links and trigger electrophile
response signaling events. The hydroperoxide and alcohol−
peptide derivatives of this reaction were examined by LC-MS
and HPLC-UV and compared to various oxidative systems.
Authentic standards were prepared that allowed the complete
structural characterization by NMR as well as quantification of
the products formed.
• −
reaction mixture the quantity necessary to generate 3.8 μM O2
min−1. Typical conditions involved a total 1000-fold dilution of the
stock XO. In some assays, 10 μg/mL SOD was included. After 30 min,
the samples were treated with catalase (10 μg/mL) for 10 min in the
dark to remove remaining H2O2. The samples were then filtered
through a 0.22 μm nylon filter (Costar) and taken to 0.2 M HCl with
4 M HCl to prevent degradation of the hydroxylation products. This
sample was split into two. One was analyzed directly by HPLC, while
the other was reduced with 40 mM DMS (from 2.0 M DMS in
acetonitrile) for 1 h at room temperature, with mechanical agitation
and protected from light.
Hydroxylation of Peptides by Hydroxyl Radical. Oxidation of
the peptides with H2O2 in the presence of Fe(II):EDTA (1:2) was
performed in 5 mM sodium phosphate at pH 7.4, using 2 mM peptide
and either 0.2 mM H2O2 and 0.2 mM Fe(II):EDTA (1:2) or 4 mM
H2O2 and 4 mM Fe(II):EDTA (1:2). The Fe(II):EDTA (1:2)
solution was prepared fresh from mixing equal volumes of 20 mM
FeSO4 and 40 mM EDTA (pH 7.4). After mixing with H2O2, the
reaction was allowed to proceed for 5 min, and then the sample was
taken to 0.2 M HCl using 4 M HCl and kept on ice to prevent
degradation of the hydroxylated products. For LC-MS analysis, the
samples were concentrated 5−10-fold in a speedvac at room
temperature (1 h).
Synthesis of o-OH Derivatives. o-OH derivatives of the tyrosine-
containing peptides were synthesized using mushroom tyrosinase in
the presence of ascorbate to prevent further oxidation, as described.20
o-OH-tyrosine-containing peptides were isolated by preparative HPLC
(see HPLC conditions below). The peptides were confirmed by UV,
MS, and NMR and quantified by UV spectrophotometry, ε280 = 2700
M−1 cm−1.20
Synthesis of p-Hydroperoxide- and p-Hydroxytyrosine
Derivatives. p-OOH derivatives were prepared by singlet oxygen
1
(1O2, molecular oxygen in its Δg state) mediated oxidation of the
peptide. A solution of 10 mM peptide and 30 μM methylene blue in
D2O at pH 7 was exposed to visible light (two 200 W white light
bulbs) with constant oxygen bubbling and stirring in an ice bath for 1
h. Under these conditions, the p-OOH was formed as a major product,
while some of the p-OH derivative was also observed. No o-OH
derivative or dimers were observed under these conditions. The p-OH
derivative was obtained by reduction with 0.1 M DMS in the presence
of acetonitrile (30% of the total volume) for 1 h at room temperature
and protected from light. The reaction mixture was then dried by
rotary evaporation, redissolved in water, and purified by preparative
HPLC. The purified p-OH derivatives were characterized by MS and
NMR.
EXPERIMENTAL SECTION
■
Materials. Glycyl-tyrosine was from TCI (Portland, OR), tyrosyl-
glycine was from Bachem (Torrance, CA), and Leu-enkephalin (Leu-
enk) was from American Peptide Company (Sunnyvale, CA).
Mushroom tyrosinase was from Pfaltz & Bauer (Waterbury, CT),
xanthine oxidase (XO, grade I), horseradish peroxidase (HRP, Type
VI), superoxide dismutase from bovine erythrocytes (SOD), and
dimethyl sulfide (DMS) were from Sigma (St Louis, MO). Tyrosine-
HPLC Conditions. HPLC experiments were done on an Alliance
HPLC equipped with diode array UV detection (Waters, Milford,
MA). Preparative HPLC was done using a Discovery C18 column
(250 × 21.2 mm, 5 μm, Supelco), while analytical HPLC was done
using a Microsorb-MV 100-5 C18 column (250 × 4.6 mm, 5 μm,
Varian). The mobile phase consisted of 0.1% trifluoroacetic acid in (A)
water and (B) acetonitrile. HPLC condition 1: to separate small
peptides, 100% A was used for 10 min, and then a gradient with B was
set. HPLC condition 2: Leu-enk derivatives were separated starting
with 85% A for 10 min and then increasing B with time. The products
containing peptides were quantified by UV spectrophotometry, ε275
=
1470 M−1 cm−1.19
Oxidation of Peptides with Xanthine Oxidase and Horse-
radish Peroxidase. Oxidation of tyrosine-containing peptides with
this system was based on the work of Winterbourn et al.12,13 Briefly, to
200 μM peptide in 50 mM phosphate buffer, 0.1 mM DTPA at pH 7.4,
16774
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
from the oxidation of peptides by XO and HRP were identified by
comparing the retention times and UV spectra with 1O2-oxidized
peptide
DMS, with NMR quantified p-OH derivative + starting
peptide, with isolated o-OH derivatives (synthesized using tyrosinase),
and with starting peptide. Response factors for the starting peptide and
the isolated o-OH derivative were determined by measuring the area of
the peaks at 280 nm of UV spectrophotometry quantified samples.
The response factor for the p-OH derivatives was obtained as follows:
A known amount of the unmodified peptide was added to the isolated
p-OH derivative, and the relative concentrations were quantified by 1H
integration in NMR. This mixture was then analyzed by HPLC-UV.
The concentration of unmodified peptide in this mixture was
quantified, allowing for the calculation of the concentration of p-OH
and the response factor at different wavelengths. The response factor
obtained for the p-OH derivatives was also used to quantify the p-
OOH derivatives because they share the same chromophore.
LC-ESI Mass Spectrometry. HPLC−MS/MS analysis was
performed on a Waters Aquity UPLC system (Waters, Milford,
MA) equipped with a Thermo LTQ ion trap mass spectrometer
(Thermo Fisher Scientific, Waltham, MA), operating in the ESI
positive ion mode. The separation was performed on a 150 × 4.6 mm
Luna 3 μ C18(2) 100 Å column (Phenomenex, Torrance, CA) with
solvents consisting of 1% formic acid in (A) water and (B) acetonitrile
with flow rate 350 μL/min. Gradient elution of the peptides was
achieved starting at 100% A for 10 min for glycyl-tyrosine (GY) and
tyrosyl-glycine (YG) derivatives or at 85% A for Leu-enk. MS analysis
was performed in a full scan followed by five data-dependent MS/MS
scans.
NMR Spectroscopy. All reported NMR spectra were acquired
using a 14.1 T Bruker magnet equipped with a Bruker DRX
spectrometer operating at proton Larmor frequency of 600.13 MHz.
1H spectra were acquired in 2.5 mm NMR tubes using a Bruker 5 mm
cryogenically cooled NMR probe. NMR samples were prepared in
2
D2O, which also served as the H lock solvent, and chemical shifts
1
were referenced internally to HDO (4.79 ppm). H, 13C, 2D COSY,
HSQC, and HMBC were obtained for all the adducts for structural
assignment. The stereochemistry of some of the adducts was
confirmed by nuclear Overhauser enhancement (NOE) experiments.
RESULTS
■
Figure 2. Oxidation of GY. (A) GY and GY-oxidation products. (B)
Extracted ion chromatograms of GY hydroperoxide (m/z = 271) and
the alcohol derivatives (m/z = 255) after exposure to different
oxidizing systems. XO/HRP: 200 μM GY was incubated with XO and
1 mM acetaldehyde that generated 3.8 μM O2• −/min and 140 nM
HRP in 50 mM sodium phosphate, 0.1 mM DTPA pH 7.4 for 30 min
at room temperature protected from light. The reaction was stopped
by addition of 10 μg/mL catalase, filtration, and acidification to 0.2 M
HCl and then analyzed by LC-MS. A hydroxyl radical producing
system, H2O2 and Fe(II):EDTA, was used to produce a mixture
enriched in o-OH 4. The p-OOH 2 and p-OH 3 derivatives were
prepared with 1O2. By comparison with these products, it was
concluded that the XO/HRP system produced all three p-OOH 2, p-
OH 3, and o-OH 4. (C) Product ions of GY and hydroxylated
derivatives: (a) GY; (b) p-OOH 2; (c) p-OH 3; (d) o-OH 4. p-OOH
2 showed a complex fragmentation involving loss of O2 and H2O2,
indicating the presence of a hydroperoxide group, while the
fragmentation of both p-OH 3 and o-OH 4 was more similar to that
of the original peptide. The most significant difference between the
fragmentation of p-OH 3 and o-OH 4 is the loss of a second molecule
of water by p-OH 3.
LC-MS Analysis of GY Oxidation Products. GY (1) was
incubated with XO and HRP, and the oxidation progress was
monitored by LC-MS (Figure 2). Two hydroxylated derivatives
of GY at retention times 7.8 and 18 min (m/z = 255) and one
hydroperoxide derivative (m/z = 271) were formed. The peak
at 7.8 min (m/z = 255) had the same retention time and
1
fragmentation pattern as that generated by O2 (Figure 2B),
which corresponded to the p-OH derivative of GY (3). The
most distinct product ion was m/z = 180, which corresponded
to the simultaneous loss of water and glycine (Figure 2C,c).
The peak at 10.5 min (m/z = 271) has the same retention time
1
and fragmentation pattern as that generated by O2 (Figure
2B), and corresponded to p-OOH 2. The major fragment m/z
= 214 corresponded to the loss of glycine, with the two atoms
of oxygen remaining attached to the tyrosine. Other fragments
indicated the loss of H2O2 and O2, consistent with the presence
of a hydroperoxide in the tyrosine. The minor product at 18
min (m/z = 255) had the same retention time and
fragmentation pattern as the main product from the reaction
with hydroxyl radical (H2O2 + Fe(II):EDTA), which
corresponded to the o-OH derivative of tyrosine (4). The
fragmentation pattern of o-OH 4 was similar to that of GY 1,
where loss of CH2O2 and loss of glycine predominated. Loss of
water was also observed for both molecules (Figure 2C,d). The
main difference between the mass spectra of o-OH 4 and p-OH
3, was the loss of a second molecule of water for p-OH 3, a
pattern that may serve as a diagnostic fragmentation for p-OH
derivatives of tyrosine.
NMR Structural Characterization of Products. Un-
ambiguous structural assignments were done by NMR studies.
Distinctive chemical shifts and splitting patterns of p-OH 3
revealed obvious differences relative to GY 1 and o-OH 4. As
16775
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
shown in Figure 3, GY 1 has symmetric aromatic protons H2
and H6 at 7.2 ppm and H3 and H5 at 6.9 ppm. Further, the
Figure 3. 1H NMR spectra in the aromatic−vinylic region of GY
derivatives. (A) GY 1 shows two doublets of doublets corresponding
to the equivalent protons at positions 2 and 6 (7.17 ppm) and 3 and 5
(6.86 ppm). (B) o-OH 4 shows three nonequivalent protons at
positions 5 (6.87), 2 (6.81), and 6 (6.73 ppm). (C) p-OH 3 shows
four distinctive diastereotopic protons with one bond and long-range
couplings corresponding to the positions 2/6 (7.08)/(7.04) and 3/5
(6.36)/(6.24) ppm. 1H assignments were based on the distinctive
chemical shifts, splitting patterns, and 2D-NMR experiments, including
COSY, HSQC, and HMBC. All the spectra were acquired in D2O at
298 K.
Figure 4. Quantification of the products from GY oxidation with XO
and HRP. (A) HPLC separation of products of GY oxidation by XO
and HRP monitored at 227 and 280 nm. (a) Authentic o-OH 4
produced with mushroom tyrosinase; (b) 1O2-oxidized GY 1, showing
assignments were confirmed by HMBC to phenolic carbon at
154.1 ppm (Table S1). In contrast, all four separate peaks in
the aromatic−vinylic region for p-OH 3 (Figure 4C) indicated
that all protons are in different environments. 13C shift at 188.3
ppm (Table S1) is consistent with the formation of a carbonyl,8
and its HMBC correlations to protons H2 and H6 confirmed
the para-substituted cyclohexadienone. Similar NMR assign-
ments were reported for the p-OOH and p-OH cyclo-
hexadienone derivatives of 3-(4-hydroxyphenyl)-propionic
1
p-OOH 2 and p-OH 3; (c) O2-oxidized GY 1 reduced with DMS,
showing p-OH 3 only; (d) GY 1 oxidized with XO and HRP using the
conditions described in Figure 2, which generated p-OOH 2, p-OH 3,
and o-OH 4; (e) GY 1 oxidized with XO and HRP followed by DMS
reduction. The p-OOH 2 was completely reduced to the p-OH 3,
while no change occurred in o-OH 4. (B) Quantification of the
products showed that the primary products were p-OOH 2, followed
by p-OH 3 and then o-OH 4. Omission of HRP or addition of SOD to
the oxidizing system led to a significant decrease in the formation of
both p-OOH 2 and p-OH 3, confirming the involvement of both
superoxide and tyrosyl radical in the formation of these products.
Treatment with DMS led to the complete reduction of the p-OOH 2
and formation of p-OH 3, with no change in o-OH 4 levels.
9
acid and of the peptide Gly-Tyr-Gly by Wright et al. The o-
OH product 4, which was generated using mushroom
tyrosinase,20 showed an absorption peak at 280 nm in the
UV and three different protons in the aromatic region in NMR
that also distinctively differ from GY 1 and p-OH 3 but are
consistent with the 3,4-dihydroxyphenylalanine structure
(Figure 3).
Quantification of GY Oxidation Products with XO and
HRP. The products of GY oxidation by XO and HRP were
quantified by HPLC-UV. Under HPLC condition 1, p-OH 3
and p-OOH 2 eluted at 5.8 and 7.9 min, respectively, with
maximal absorption at 227 nm, while o-OH 4 eluted at 22.2
min (Figure 4) with an absorption peak at 280 nm. The analysis
shows that the products of the reaction between superoxide and
tyrosyl radical were the hydroperoxide p-OOH 2 (1.2 μM),
followed by p-OH 3 (0.35 μM) and o-OH 4 (0.2 μM) (Figure
4). After reduction with DMS, p-OOH 2 was converted
stoichiometrically into p-OH 3 (1.7 μM), whereas the yield of
o-OH 4 (0.2 μM) was not affected by DMS. We conclude that
para addition dominates the reaction of tyrosyl and superoxide,
with the ortho addition product being formed at about 12−15%
of the para products.
The involvement of superoxide and tyrosyl radical in the
generation of these products was evidenced by the significant
decrease in p-OOH 2 in the absence of HRP or after the
addition of SOD (Figure 4B). The production of p-OOH 2 in
the absence of HRP is probably mediated by hydroperoxyl
radical, the protonation product of superoxide (pKa = 4.8), that
is theoretically capable of oxidizing tyrosine.21,22 This oxidation
was very sensitive to SOD, which reduced to zero the
production of p-OOH 2 (not shown).
LC-MS Analysis of YG Oxidation Products. Oxidation of
YG with XO and HRP gave two hydroperoxides, p-OOH 6 and
7 (m/z = 271), and two alcohol derivatives, p-OH 8 and 9 (m/
z = 255), of the dipeptide (Figure 5). These structures have
been confirmed by comparison with standards generated using
1O2 oxidation and subsequent analyses. Reduction with DMS
converted both p-OOH 6 and 7 to the corresponding p-OH 8
16776
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
6 and 7 to the alcohol derivatives p-OH 8 and 9. No significant
o-OH derivative of YG was observed under these conditions.
• −
The involvement of both O2
and tyrosyl radical in the
formation of these products is evidenced by the 20-fold
decrease in the formation of p-OOH derivatives in the absence
of HRP or in the presence of SOD (Figure 5D). The yield of p-
OOH derivatives from YG 5 was 12 times higher than the one
observed with GY 1. This effect has been observed before and
has been attributed to free amines favoring superoxide
adduction rather than reduction to the N-terminal tyrosyl
radical.12,13,16,17
NMR Structural Characterization of Products. Further
characterization of the hydroxylated products by NMR
confirmed that p-OH 8 and 9 were bicyclic indolic
diastereoisomers (see in Tables S2 and S3). NOE NMR
experiments enabled the assignment of configuration for p-OH
8 as (2S,3aR,7aR)-3a-hydroxy-6-oxo-2,3,3a,6,7,7a-hexahydro-
1H-indole-2-carboxylic and p-OH 9 to (2S,3aR,7aS)-3a-
hydroxy-6-oxo-2,3,3a,6,7,7a-hexahydro-1H-indole-2-carboxylic
acid, respectively, as shown in Figure 6 (more details in Section
Figure 5. Oxidation of YG with XO and HRP. (A) Structures of YG
and its derivatives. (B) Extracted ion chromatograms of YG exposed to
different oxidizing systems. YG was incubated with XO, acetaldehyde,
and HRP as described in Figure 2 (details in the Experimental
section). The primary products consisted of p-OOH 6 and 7 (m/z =
271) that were reduced to alcohols p-OH 8 and 9 with DMS (m/z =
255) and that had the same retention time and fragmentation pattern
1
as those produced by O2 (not shown). No detectable amount of o-
OH derivative was observed in this reaction by comparison with
mushroom-tyrosinase generated standard. Hydroxylation by Fe-
(II):EDTA and hydrogen peroxide led to the formation of both o-
and p-OH derivatives of YG. (C) RP-HPLC separation of the products
of YG oxidation by XO and HRP monitored at 220 and 280 nm: (a)
authentic o-OH YG derivative produced with mushroom tyrosinase;
Figure 6. Stereochemistry of p-OH 8 and 9 assigned based on NOE
NMR experiments. Key cross-peaks found in each diastereomer are
shown by the arrows. The colors and thickness of the arrows indicate
the strength of the NOE: red, strong; orange, medium; green, weak.
(A) The oxidized tyrosine in p-OH 8 has the same configuration as
(2S,3aR,7aR)-HOHICA: missing cross-peaks between H7a, H3α, and
H2 indicates that H7a is in an opposite face to H2 and therefore
should be 7aR. The difference in NOE between H3α and H3β to H4
indicates that is a cis-fused ring conformation and not a trans-fused
ring, so it follows that 3a is R configuration. (B) The oxidized tyrosine
in p-OH 9 has the same configuration as (2S,3aR,7aS)-HOHICA:
observed cross-peaks between H7a, H3α, and H2 indicates they are on
the same face of the molecule, that is 7aS. The similar NOE intensity
between H3α and H3β to H4 suggests it is a trans-fused ring
conformation, characteristic of 3aR,7aS. More detailed analysis and
crude data are provided in Section S4, SI.
1
(b) YG oxidized with O2, showing p-OOH 6 and 7 and p-OH 8 and
9; (c) YG oxidized with 1O2 and then reduced with DMS, showing p-
OH 8 and 9; (d) YG oxidized with XO and HRP; (e) YG oxidized
with XO and HRP and then reduced with DMS. (D) Quantification of
the different products showed that in the presence of both XO and
HRP, p-OOH 6 was preferentially formed, followed by p-OOH 7,
which were converted stoichiometrically to p-OH 8 and 9 by reduction
with DMS. No o-OH derivative was observed under these conditions.
The formation of these products was reduced more than 20 times in
the absence of HRP or after the addition of SOD, indicating the
• −
involvement of O2 and tyrosyl radical in the formation of these
products.
and 9. Interestingly, hydroxylation of YG by Fe(II):EDTA and
H2O2 produced both o- and p-OH derivatives, in contrast to
tyrosinase that produced exclusively o-OH derivative (Figure
5B,C). The collision induced dissociation of YG 5 and its o-OH
derivative showed a similar pattern, where the main fragments
arose from the loss of NH3 and the loss of the glycine and CO
(Figure S1). In contrast, loss of H2O was observed in both p-
OH peaks, in addition to the loss of glycine and CO (Figure
S1). The p-OOH derivatives showed a more complex
fragmentation, involving loss of H2O2, in addition to the loss
of H2O, glycine, and CO (Figure S1).
Quantification of YG Oxidation Products. The major
products from the addition of superoxide to the tyrosyl radical
of YG, produced using XO and HRP, were p-OOH 6 and 7,
with only trace amounts of p-OH 8 and 9 (Figure 5D).
Subsequent reduction with DMS completely converted p-OOH
S4, SI). These structures are consistent with a Michael-type
cyclization to give both cis- and trans-fused bicyclic products;
the cis-fused p-OOH 6 being favored over the trans-fused p-
OOH 7 by about 3:1. These bicyclic derivatives of tyrosine
have the same core structure as HOHICA.9,11 The stereo-
chemical configurations determined here for the main product
p-OH 8 are different from that originally reported by Jin et al.11
but are in agreement with the recent report by Wright et al.9
LC-MS Analysis of Leu-Enkephalin Oxidation Prod-
ucts. Leu-enk oxidation by XO and HRP also led to the
formation of bicyclic indolic compounds p-OOH 11 and 12
(m/z 588) and p-OH 13 and 14 (m/z 572) (Figure 7). No
evidence for an o-OH derivative was found. Collision induced
dissociation of all of the indolic compounds gave predom-
inantly amide bond cleavage over side chain fragmentation. The
mass shifts in b and y ions in the oxidation products of Leu-enk
16777
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
catalyzed decomposition of H2O2 led to hydroxylation at both
ortho and para positions of tyrosine. The formation of GY and
YG p-OH derivatives is evident by LC-MS matching m/z = 255
and retention time with the same fragmentation patterns as
1
those generated by O2 or XO/HRP oxidation (Figures 2 and
5). No hydroperoxide derivatives were observed under these
conditions. Quantification of the hydroxylated products showed
that para hydroxylation was a minor process relative to ortho
hydroxylation: p-OH 3 accounted for 12% relative o-OH 4 and
p-OH 9 accounted for 6% relative to the o-OH derivative
(Table 1).
Table 1. o- and p-Hydroxylation of Tyrosine in GY and YG
a
by Fenton-Generated Hydroxyl Radical
products
concentration (μM)
30
GY
YG
o-OH 4
p−OH 3
o-OH
3
3.8 0.3
12.7 0.9
0.8 0.1
p−OH 9
a
2.0 mM GY or YG exposed to 0.4 mM FeSO4, 0.8 mM EDTA, and
0.4 mM H2O2 in 5 mM sodium phosphate at pH 7.4 for 5 min at room
temperature. Sample was then acidified to 0.2 M HCl and analyzed by
HPLC-UV. Experiments were done in triplicate, and results are
expressed as mean SD.
Figure 7. Oxidation of Leu-enk with XO and HRP. (A) Structures of
Leu-enk and its derivatives. (B) Extracted ion chromatograms of Leu-
enk exposed to different oxidizing systems. Leu-enk was incubated
with XO, acetaldehyde, and HRP in order to generate superoxide and
tyrosyl radical simultaneously (details in Figure 2). Both the
hydroperoxide (m/z = 588) and the alcohol derivatives (m/z =
572) were monitored. The primary products consist of p-OOH 11 and
12 (m/z = 588) that were reduced to the alcohol derivatives p-OH 13
and 14 with DMS (m/z = 572). No o-OH derivative was observed in
this reaction by comparison with authentic standard. (C) RP-HPLC
chromatograms of Leu-enk oxidized with XO and HRP monitored at
230 nm: (a) authentic o-OH derivative of Leu-enk; (b) Leu-enk
oxidized with 1O2; (c) Leu-enk standard; (d) Leu-enk oxidized by XO
and HRP; (e) Leu-enk oxidized by XO and HRP and then reduced
with DMS. (D) Quantification of the different products showed that in
the presence of both XO and HRP, p-OOH 11 was preferentially
formed, followed by p-OOH 12. Addition of DMS converted the p-
OOH derivatives quantitatively to p-OH 13 and 14. In the absence of
HRP, or after adding SOD, the formation of these products was
reduced more than 20 times, evidencing the involvement of superoxide
and tyrosyl radical in the formation of the p- and p-OH derivatives.
Little oxidation of Leu-enk occurred under the Fenton
conditions used for reaction of GY and YG, but when Leu-enk
was reacted with 10-fold more H2O2 and Fe(II):EDTA,
oxidation products were observed by LC-MS. At least five
different peaks with an m/z consistent with the addition of one
atom of oxygen were found (m/z = 572, Figure S3). According
to their fragmentation patterns, the peaks 1 and 2 were
identified as phenylalanine hydroxylation products (Figure S4).
Peaks 3−5 had the same retention time and fragmentation
pattern as those of p-OH 13 and 14 and the o-OH derivative
(Figures S3−S5). We also found a hydroperoxide on the
leucine (m/z = 588), based on the shifts in b and y ions (Figure
S5). Leucine hydroperoxides have been observed before in
proteins exposed to peroxyl and alkoxy radicals.23
DISCUSSION
■
indicated that the p-OOH and p-OH modifications are on the
terminal tyrosine (Figure S2).
Superoxide and tyrosyl radicals are two of the most abundant
free radicals in biology. Superoxide can be produced by many
pathways including its direct formation by the enzymes
NADPH oxidase and xanthine oxidase and as a result of
electron leakage in the mitochondrial electron transport chain.
Furthermore, many pathological conditions including diabetes,
cardiovascular disease, and cancer are associated with an
increased production of superoxide from these sources.24−26
The tyrosyl radical can also be formed by multiple pathways,
including oxidation of tyrosine by oxygen radicals, nitrogen
dioxide, carbonate radical, and hemeperoxidases.1,2,27−29 o-
Nitrotyrosine and o,o′-dityrosine are detected in vivo in normal
conditions and are usually increased in pathological conditions,
providing evidence of tyrosyl radical formation in vivo.1,28−31
The reaction between superoxide and tyrosyl radicals occurs at
nearly diffusion-controlled rates (k = 1.5 × 109 M−1 s−1),11,16
suggesting that this reaction may be important in vivo. There
are a large number of alternative possible reactions for these
radicals in vivo that will affect the amount of p-OOH formed.
Lancaster built a kinetic model that takes into account the most
important reactions of oxygen and nitrogen reactive species
Quantification of Leu-Enkephalin Oxidation Products.
When Leu-enk was oxidized with XO and HRP, the p-OOH
derivatives 11 and 12 predominated, being approximately six
times more abundant than the corresponding p-OH derivatives
13 and 14 (Figure 7D). After treatment with DMS, both p-
OOH compounds were reduced to the p-OH derivatives
quantitatively, with a product ratio of 3:1 p-OH 13:14.
NMR Structural Characterization. The structures of p-
OH 13 and 14 were confirmed by 1H, COSY, HSQC, HMBC,
and NOE NMR experiments (Tables S4 and S5). The two
diastereomeric p-OH 13 and 14 compounds did not show
much difference in 1H NMR (Table S5). Distinctive differences
found in NOE indicated that p-OH 13 had a cis-fused ring
conformation, while p-OH 14 had the trans-fused ring (see a
detailed explanation in in Section S4, SI). As was the case for
YG, the nucleophilic addition for Leu-enk favors the cis-fused
bicyclic system in a ratio of 3:1 p-OOH 11:12.
p-Hydroxylation by Hydroxyl Radical. Exposing the
peptides to hydroxyl radical generated from Fe(II):EDTA-
16778
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
with other radicals and low molecular weight and enzymatic
antioxidants under biological conditions.32 One of the
conclusions of this work was that the main reactions of tyrosyl
radical were the radical swap with glutathione and the reaction
with superoxide, the latter being significantly more important
than the reaction with nitrogen dioxide.32 All these results
suggest that p-OOH and p-OH derivatives of tyrosine should
be significantly more abundant than o-nitrotyrosine. We are
presently developing methods to quantify p-OH-tyr to clarify
this matter.
In most of the recent work on superoxide addition to tyrosyl
radicals the ortho−para selectivity of adduction was not
addressed and the structure of products formed was left
unresolved.12,14−18 Our results demonstrate that tyrosyl radicals
of different peptides react with superoxide to produce
hydroperoxides predominately in the para position of tyrosine.
This conclusion was evident by LC-MS and HPLC-UV when
the products of XO/HRP oxidation were compared to the o-
OH products generated from enzymatic oxidation with
mushroom tyrosinase and to the p-OH and p-OOH products
generated from singlet oxygen. Furthermore, reduction of the
XO/HRP or singlet oxygen-derived hydroperoxides with DMS
led exclusively to the p-OH derivatives, confirmed by LC-MS,
HPLC-UV, and NMR. In contrast, the dopa-like o-OH tyrosine
derivatives could only be observed in very low amounts from
GY, and these products were unaffected by DMS, indicating
that no stable hydroperoxides are precursors to these
compounds.
In the case of YG and Leu-enk, where the tyrosine is N-
terminal, the oxidized tyrosine formed a bicyclic indolic alcohol
(HOHICA) after reduction with DMS. Hydroperoxides were
not studied directly by NMR, but their reduction with DMS led
to a single alcohol derivative, indicating that the formation of
the bicyclic structure occurred at the hydroperoxide stage. Two
major diasteroisomers were observed in both YG and Leu-enk,
that were assigned to two different configurations of HOHICA,
the cis-fused ring 2S,3aR,7aR and the trans-fused ring
2S,3aR,7aS (Figure 7). These bicyclic structures are formed
by amine conjugate attack to the α,β-carbonyl of the
intermediate noncyclized cyclohexadienone p-OOH, and the
addition occurs preferentially from the face of the ring opposite
to the hydroperoxide, giving the cis-fused ring 2S,3aR,7aR
compound as the major product.
In the case of GY, where the tyrosyl is C-terminal, the p-OH
derivative (3) did not undergo cyclization. Other peptides with
N-protected tyrosine, such as N-acetyl-Tyr-Ala methyl ester,8
[Tyr5]bradykinin (to be published), and Gly-Tyr-Gly,9 formed
stable p-OH products after oxidation, suggesting that
cyclization occurs readily only in N-terminal tyrosine
containing peptides.
this could explain the experimentally observed preference for
the para position: The negative charge at the ortho positions is
acting as a kinetic barrier that slows down the reaction with the
negatively charged superoxide at that site.
The N-terminal tyrosine peptides studied here gave 10-fold
more superoxide adduction under identical reaction conditions
than those peptides having an internal tyrosine. This propensity
for N-terminal tyrosines to give more oxidation products in
reactions with XO/HRP has been observed before and has
been attributed the selective preference for adduction rather
than reduction of superoxide to the N-terminal tyrosyl
radical.12,13,16,17
The reaction of YG and GY under Fenton conditions shown
in Table 1 confirm earlier reports that tyrosine hydroxylation by
hydroxyl radical from different sources gives dopa-like o-OH
derivatives as the main products.33−36 The work by Solar et al.
on the reaction of •OH with tyrosine by pulse radiolysis
•
suggested indirectly that para and ipso addition of OH were
formed, but no products were isolated and characterized.37
Hydroxyl radical has been shown to give approximately 10% of
the p-OH product in its reaction with p-cresol,38,39 but our
studies with YG and GY appear to be the first time that p-OH
products have been unambiguously shown to be a part of the
Fenton-tyrosine product mixture.
Recent studies identified 64 endogenous sites of tyrosine
hydroxylation in rat brain and heart proteins, most of which
locate in the mitochondria,6 and 9 endogenous sites of tyrosine
hydroxylation in human cells (HeLa) mitochondria.40 Taking
into account the many ways through which p-OH derivatives of
tyrosine can be generated and that these p-OH derivatives
ionize better by electrospray than dopa-like derivatives, it is
very likely that some of the identified sites of hydroxylation
correspond to p-OH derivatives, rather than the assigned o-
OH.6,40 Caution is recommended when interpreting endoge-
nous tyrosine hydroxylation, and p-OH tyrosine should be
considered as an additional relevant product.
There are at least four routes that generate para-substituted
tyrosines: (1) the reaction between tyrosyl radical and lipid
peroxyl radicals;8 (2) the reaction between tyrosyl radical and
superoxide; (3) the hydroxylation of tyrosine via hydroxyl
radical (presented here); and (4) the reaction of tyrosine with
1O2. These para-substituted tyrosines are powerful electrophiles
and can form adducts with thiols and other nucleophiles. Using
peptides having tyrosines at sites other than the N-terminus, we
observed the formation of Michael-type adducts with a number
of nucleophiles of biological interest, including the addition of
two cysteines to a single oxidized tyrosine (manuscript in
preparation). While preparing this manuscript, a report by
Nagy et al. appeared that showed that N-terminal tyrosine-
containing peptides formed monoadducts with glutathione
when oxidized with XO and HRP.18 It seems clear from our
work that the products reported are the result of tyrosine para
oxidation and Michael addition to the indolic bicycles. It also
follows that N-terminal tyrosines will give adduction of only
one nucleophile while oxidation of tyrosines at sites other than
the N-terminus can lead to mono- or diadduction of
nucleophiles (see Scheme 2 for an example of adduction of
two thiol nucleophiles) with potentially important biological
consequences. As has been observed for other electrophiles,
such as 4-hydroxynonenal, it is possible that at low
concentrations these electrophilic para-substituted tyrosines
trigger adaptive responses, like heat shock response or
antioxidant response to prevent further damage.41,42 On the
The selective addition of superoxide to the para position of
tyrosyl radicals is intriguing. In the tyrosyl radical both the ortho
and the para positions are spin-rich, and the ortho site is
preferred for the addition of many radicals, including nitrogen
dioxide and a second tyrosyl.1,17,28,30 There are of course two
available ortho positions, which are less sterically hindered than
the one para site, and the addition of radicals ortho allows
ultimately for rearomatization. Quantum calculations indicate
that the formation of the product of addition of superoxide to
tyrosyl radical in the ortho position is thermodynamically more
favorable over the reaction at the para position.17 The same
study showed that the charge distribution in the tyrosyl radical
placed a significant negative charge at the ortho positions,17 and
16779
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780

Journal of the American Chemical Society
Article
(8) Shchepin, R.; Moller, M. N.; Kim, H. Y.; Hatch, D. M.;
Bartesaghi, S.; Kalyanaraman, B.; Radi, R.; Porter, N. A. J. Am. Chem.
Soc. 2010, 132, 17490.
̈
Scheme 2
(9) Wright, A.; Bubb, W. A.; Hawkins, C. L.; Davies, M. J. Photochem.
Photobiol. 2002, 76, 35.
(10) Jin, F.; Leitich, J.; von Sonntag, C. J. Photochem. Photobiol A
1995, 92, 147.
(11) Jin, F.; Leitich, J.; von Sonntag, C. J. Chem. Soc., Perkin Trans.
1993, 2, 1583.
(12) Nagy, P.; Kettle, A. J.; Winterbourn, C. C. J. Biol. Chem. 2009,
284, 14723.
(13) Winterbourn, C. C.; Parsons-Mair, H. N.; Gebicki, S.; Gebicki, J.
M.; Davies, M. J. Biochem. J. 2004, 381, 241.
(14) Nagy, P.; Kettle, A. J.; Winterbourn, C. C. Free Radical Biol.
Med. 2010, 49, 792.
other hand, higher concentrations of these modified tyrosines
could contribute to protein aggregation, such as that associated
with neurodegenerative diseases,43,44 promote cellular damage,
and ultimately lead to cell death. Such potential biological roles
warrant further investigation of these tyrosine-derived electro-
philes.
(15) Das, A. B.; Nagy, P.; Abbott, H. F.; Winterbourn, C. C.; Kettle,
A. J. Free Radical Biol. Med. 2010, 48, 1540.
(16) Mozziconacci, O.; Mirkowski, J.; Rusconi, F.; Pernot, P.;
Bobrowski, K.; Houee-Levin, C. Free Radical Biol. Med. 2007, 43, 229.
(17) Field, S. M.; Villamena, F. A. Chem. Res. Toxicol. 2008, 21, 1923.
(18) Nagy, P.; Lechte, T. P.; Das, A. B.; Winterbourn, C. C. J. Biol.
Chem. 2012, 287, 26068.
CONCLUSIONS
■
The reaction between tyrosyl radicals and superoxide yields
para-hydroperoxide derivatives that can cyclize to mostly a cis-
fused indolic derivative, 2S,3aR,7aR-HOHICA, when the
tyrosine is N-terminal, or remain as an acyclic p-OOH
cyclohexadienone when the tyrosine is elsewhere in the
peptide. The hydroperoxide can hydrolyze spontaneously or
be reduced to yield the corresponding p-alcohol. Both products
contain α,β-unsaturated carbonyls, which make them good
electrophiles, and could confer on them interesting biological
properties. We are presently assessing the formation of these
electrophilic tyrosine-derivatives in vivo as well as their
biological effects.
(19) Edelhoch, H. Biochemistry 1967, 6, 1948.
(20) Marumo, K.; Waite, J. H. Biochim. Biophys. Acta 1986, 872, 98.
(21) Buettner, G. R. Arch. Biochem. Biophys. 1993, 300, 535.
(22) Folkes, L. K.; Trujillo, M.; Bartesaghi, S.; Radi, R.; Wardman, P.
Arch. Biochem. Biophys. 2011, 506, 242.
(23) Steinmann, D.; Ji, J. A.; Wang, Y. J.; Schoneich, C. Mol. Pharm.
2012, 9, 803.
(24) Guzik, T. J.; Mussa, S.; Gastaldi, D.; Sadowski, J.; Ratnatunga,
C.; Pillai, R.; Channon, K. M. Circulation 2002, 105, 1656.
(25) Kelkar, A.; Kuo, A.; Frishman, W. H. Cardiol. Rev. 2011, 19, 265.
(26) Dikalov, S. Free Radical Biol. Med. 2011, 51, 1289.
(27) Augusto, O.; Bonini, M. G.; Amanso, A. M.; Linares, E.; Santos,
C. C.; De Menezes, S. L. Free Radical Biol. Med. 2002, 32, 841.
(28) Heinecke, J. W.; Li, W.; Francis, G. A.; Goldstein, J. A. J. Clin.
Invest. 1993, 91, 2866.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Tables of H and 13C NMR chemical shifts, 1D and 2D NMR
spectra, HPLC-MS data, and mass spectra of all the
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
(29) Heinecke, J. W.; Li, W.; Daehnke, H. L., III; Goldstein, J. A. J.
Biol. Chem. 1993, 268, 4069.
(30) DiMarco, T.; Giulivi, C. Mass Spectrom. Rev. 2007, 26, 108.
(31) Pennathur, S.; Bergt, C.; Shao, B.; Byun, J.; Kassim, S. Y.; Singh,
P.; Green, P. S.; McDonald, T. O.; Brunzell, J.; Chait, A.; Oram, J. F.;
O’Brien, K.; Geary, R. L.; Heinecke, J. W. J. Biol. Chem. 2004, 279,
42977.
AUTHOR INFORMATION
Corresponding Author
n.porter@vanderbilt.edu
■
(32) Lancaster, J. R., Jr. Chem. Res. Toxicol. 2006, 19, 1160.
(33) Lynn, K.; Purdie, J. Int. J. Radiat. Phys. Chem. 1976, 8, 685.
(34) Boguta, G.; Dancewicz, A. Int. J. Radiat. Biol. 1981, 39, 163.
(35) Getoff, N. Amino Acids 1992, 2, 195.
(36) Maskos, Z.; Rush, J. D.; Koppenol, W. H. Arch. Biochem.
Biophys. 1992, 296, 521.
(37) Solar, S.; Solar, W.; Getoff, N. J. Phys. Chem. 1984, 88, 2091.
(38) Schuler, R. H.; Albarran, G.; Zajicek, J.; George, M. V.;
Fessenden, R. W.; Carmichael, I. J. Phys. Chem. A 2002, 106, 12178.
(39) Albarran, G.; Schuler, R. H. J. Phys. Chem. A 2005, 109, 9363.
(40) Lee, S.; Chen, Y.; Luo, H.; Wu, A. A.; Wilde, M.; Schumacker, P.
T.; Zhao, Y. J. Proteome Res. 2010, 9, 5705.
(41) Jacobs, A. T.; Marnett, L. J. Acc. Chem. Res. 2010, 43, 673.
(42) Iles, K. E.; Dickinson, D. A.; Wigley, A. F.; Welty, N. E.; Blank,
V.; Forman, H. J. Free Radical Biol. Med. 2005, 39, 355.
(43) Giasson, B. I.; Duda, J. E.; Murray, I. V.; Chen, Q.; Souza, J. M.;
Hurtig, H. I.; Ischiropoulos, H.; Trojanowski, J. Q.; Lee, V. M. Science
2000, 290, 985.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
N.P. thanks the National Science Foundation for support of
this work. We thank Professor Rafael Radi from the
■
Universidad de la Republica, Montevideo, Uruguay, Yoel
́
́
́
Garcia-Diaz of Vanderbilt University for helpful discussions,
and Donald Stec for assistance in NMR experiments.
REFERENCES
■
(1) Radi, R. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 4003.
(2) Ischiropoulos, H. Arch. Biochem. Biophys. 1998, 356, 1.
(3) Ischiropoulos, H. Arch. Biochem. Biophys. 2009, 484, 117.
(4) Shao, B.; Oda, M. N.; Oram, J. F.; Heinecke, J. W. Chem. Res.
Toxicol. 2010, 23, 447.
(5) Pennathur, S.; Wagner, J. D.; Leeuwenburgh, C.; Litwak, K. N.;
Heinecke, J. W. J. Clin. Invest. 2001, 107, 853.
(44) Stefani, M.; Dobson, C. M. J. Mol. Med. (Berl) 2003, 81, 678.
(6) Zhang, X.; Monroe, M. E.; Chen, B.; Chin, M. H.; Heibeck, T.
H.; Schepmoes, A. A.; Yang, F.; Petritis, B. O.; Camp, D. G., 2nd;
Pounds, J. G.; Jacobs, J. M.; Smith, D. J.; Bigelow, D. J.; Smith, R. D.;
Qian, W. J. Mol. Cell. Proteomics 2010, 9, 1199.
(7) Fu, S.; Davies, M. J.; Stocker, R.; Dean, R. T. Biochem. J. 1998,
333 (Pt 3), 519.
16780
dx.doi.org/10.1021/ja307215z | J. Am. Chem. Soc. 2012, 134, 16773−16780