A derivatization assay using gaschromatography/negative chemical ionization tandem mass spectrometry to quantify 3-nitrotyrosine in human plasma

Article abstract of DOI:10.1002/jms.543

Endogenous free or protein-associated 3-nitrotyrosine (3-NT) has been proposed as a biomarker of in vivo oxidative damage caused by nitrating agents. Isotopic dilution assay gaschromatographic/mass spectrometric (GC/MS) techniques have been employed to measure endogenous 3-NT levels. However, the quantitative normal plasma values reported so far are inconsistent. The results vary between the assays; they may have been influenced by in vitro artifactual nitration of tyrosine to 3-NT. In this study, a simple and artifact-free derivatization method for quantifying the endogenous 3-NT content of biological samples by GC/negative chemical ionization MS/MS is presented. The method is based on reduction of the nitro group of the molecule by dithionite, heptafluorobutyric acylation and subsequent methyl derivatization, di-O-methyldi-N-heptafluorobutyryl being the major derivative. The results showed excellent GC and MS properties, such as low background and a favorable fragmentation pattern. Endogenous 3-NT was unequivocally quantified using collision-induced dissociation in the selected reaction monitoring mode, whereas co-elution of unknown compounds interfered in the selected-ion monitoring mode. We found that tyrosine was nitrated in the presence of nitrate anions and heptafluorobutyric anhydride, but the product appeared as a di-O-methylmono-N-heptafluorobutyryl derivative. Therefore, artifactually formed 3-NT did not contribute to the measured endogenous 3-NT level owing to its different derivative structure. The method was applied to determine endogenous 3-NT in human plasma and plasma proteins. A detection limit of 0.03 nm for ¹³C₆-labeled 3-NT in plasma samples was established and the response was linear over a concentration range of 0-50 nM (R² > 0.999). The endogenous free 3-NT level (mean ± SD) in ultrafiltered plasma samples from 12 healthy adults was 0.74 ± 0.30 nM. The mean concentration of 3-NT in their plasma total proteins was 0.60 ± 0.40 pmol mg^-1. Hence, the described method is selective, eliminates the problem of artifactual nitration and is feasible for the quantification of free and protein-associated 3-NT in biological samples such as plasma. Copyright

Full text of DOI:10.1002/jms.543

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2003; 38: 1187–1196
Published online 3 November 2003 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.543
A derivatization assay using gaschromatography/
negative chemical ionization tandem mass
spectrometry to quantify 3-nitrotyrosine in
human plasma
1
2
1†
3
3
¨
¨
´
Ann-Sofi Soderling, Henrik Ryberg, Anders Gabrielsson, Mona Larstad, Kjell Toren,
Sohbat Niari1 and Kenneth Caidahl1∗
1
Department of Clinical Physiology, Go¨ teborg University, Sahlgrenska University Hospital, SE-413 45 Go¨ teborg, Sweden
Institute of Clinical Neuroscience, Go¨ teborg University, Sahlgrenska University Hospital, SE-413 45 Go¨ teborg, Sweden
Department of Occupational Medicine, Go¨ teborg University, Sahlgrenska University Hospital, SE-413 45 Go¨ teborg, Sweden
2
3
Received 25 March 2003; Accepted 4 September 2003
Endogenous free or protein-associated 3-nitrotyrosine (3-NT) has been proposed as a biomarker of
in vivo oxidative damage caused by nitrating agents. Isotopic dilution assay gaschromatographic/mass
spectrometric (GC/MS) techniques have been employed to measure endogenous 3-NT levels. However, the
quantitative normal plasma values reported so far are inconsistent. The results vary between the assays;
they may have been influenced by in vitro artifactual nitration of tyrosine to 3-NT. In this study, a simple
and artifact-free derivatization method for quantifying the endogenous 3-NT content of biological samples
by GC/negative chemical ionization MS/MS is presented. The method is based on reduction of the nitro
group of the molecule by dithionite, heptafluorobutyric acylation and subsequent methyl derivatization,
di-O-methyldi-N-heptafluorobutyryl being the major derivative. The results showed excellent GC and
MS properties, such as low background and a favorable fragmentation pattern. Endogenous 3-NT was
unequivocally quantified using collision-induced dissociation in the selected reaction monitoring mode,
whereas co-elution of unknown compounds interfered in the selected-ion monitoring mode. We found that
tyrosine was nitrated in the presence of nitrate anions and heptafluorobutyric anhydride, but the product
appeared as a di-O-methylmono-N-heptafluorobutyryl derivative. Therefore, artifactually formed 3-NT
did not contribute to the measured endogenous 3-NT level owing to its different derivative structure. The
method was applied to determine endogenous 3-NT in human plasma and plasma proteins. A detection
limit of 0.03 nM for 13C6-labeled 3-NT in plasma samples was established and the response was linear over a
concentration range of 0–50 nM (R2 > 0.999). The endogenous free 3-NT level (mean SD) in ultrafiltered
plasma samples from 12 healthy adults was 0.74 0.30 nM. The mean concentration of 3-NT in their
plasma total proteins was 0.60 0.40 pmol mg−1. Hence, the described method is selective, eliminates the
problem of artifactual nitration and is feasible for the quantification of free and protein-associated 3-NT
in biological samples such as plasma. Copyright  2003 John Wiley & Sons, Ltd.
KEYWORDS: gaschromatography/mass spectrometry; 3-nitrotyrosine; nitration; protein; plasma
of enzymatic activity or interference with regulatory
functions.1–3 Consequently, the post-translational nitration
of tyrosine residues that occurs during oxidative stress
or inflammation may be deleterious for cellular function.4
The generation of 3-nitrotyrosine (3-NT) can be attributed
to several reactive nitrogen species.5–7 However, little is
known about these biological processes and the mechanisms
that trigger nitration of tyrosine. This is partly due to the
problems encountered in the quantification of low concen-
trations of endogenous biomarkers in complex biological
matrices. Several analytical methods have been developed
to detect and quantify 3-NT and its metabolites in biological
fluids,6,8–10 but there is as yet no established method. This
has led to inconsistencies in the reported values of circulating
INTRODUCTION
Oxidative damage of amino acids in proteins is known
to cause structural alterations manifested e.g., by loss
ŁCorrespondence to: Kenneth Caidahl, Department of Clinical
Physiology, Sahlgrenska University Hospital, SE 413 45 Go¨teborg,
Sweden. E-mail: caidahl@clinphys.gu.se
†Present address: Department of Chemistry, Queen Mary College,
University of London, Mile End Road, London E1 4NS, UK.
Contract/grant sponsor: Swedish Research Council; Contract/grant
number: K 2002-71X-14 231-01A; K99-04RM-13 192-01.
Contract/grant sponsor: Swedish Heart and Lung Foundation.
Contract/grant sponsor: Va¨stra Go¨taland region.
Contract/grant sponsor: Go¨teborg University.
Contract/grant sponsor: Jubileumsfonden.
Contract/grant sponsor: Foundation for Neurological Research.
Copyright  2003 John Wiley & Sons, Ltd.

1188
A.-S. So¨derling et al.
free 3-NT in normal individuals.11 This inconsistency may
be attributable to methodological factors. For instance, the
specificity and sensitivity of the methods employed may
vary. Moreover, under acidic conditions the presence of
nitrate and/or nitrite in biological samples can cause in vitro
nitration of aromatic amino acids, particularly tyrosine, prior
to the derivatization procedure.8,12–15 Thus, artifactual 3-NT
may add to the quantitative values.
Assays involving gaschromatographic/mass spectromet-
ric (GC/MS) techniques have been reported to be sensitive
and selective when the influence of artifactual nitration can
be ruled out.9,16 This may be achieved by pre-clearance of
nitrate before using a suitable derivatization method. Thus,
artifactual nitration is proposed to be minimized by using
high-performance liquid chromatographic fractionation and
purification procedures prior to derivatization12 or employ-
ing an alkaline environment.8 In spite of this, reported normal
basal plasma values of 3-NT range from <117 to 64 nM.8
Furthermore, our previous experiments with the n-propyl
perfluorobutyryl derivative resulted in an unstable product,
high background noise and rapid degradation of the GC col-
umn. These drawbacks led us to develop a new derivatization
method to be used in GC/negative chemical ionization (NCI)
MS/MS analyses of 3-NT in biological samples. The possible
impact of false nitration was evaluated. We used the method
in analyses of basal levels of endogenous circulating 3-NT
in plasma from healthy volunteers and the corresponding
protein content of 3-NT after enzymatic degradation.
Nitration of [13C6]tyrosine as internal standard and
nitration of BSA
Isotopically labeled 3-nitro[13C6]tyrosine was prepared from
[13C6]tyrosine and the nitrating agent TNM as described
previously.18 The sample content of 3-nitro[13C6]tyrosine was
concentrated by anion-exchange chromatography.9 The final
yield of 3-nitro[13C6]tyrosine was determined, in the selected
reaction monitoring mode (SRM) mode after derivatization,
by comparison to peak areas of known amounts of authentic
3-NT. BSA was nitrated with TNM in the same way as for
the labeled tyrosine. After the nitration procedure, BSA was
separated from the reactants by precipitation with methanol.
Ultra centrifugation
Ultrafiltered samples were prepared by adding the internal
standard (10 µl of 0.5 µM 3-nitro[13C6]tyrosine) to 1 ml
of EDTA plasma. These samples were ultrafiltered in a
microfuge with a Vivaspin 500 cartridge (30 min, 12 000 rpm,
ambient temperature). The ultrafiltered plasma samples were
subsequently derivatized (100 µl aliquots).
Derivatization procedure for 3-NT
The first step (1) in the derivatization procedure (Fig. 1)
was the reduction of 3-NT to 3-aminotyrosine,19 obtained
by adding 20 µl of 50 mM dithionite solution (dissolved in
50 mM di-sodium hydrogenphosphate buffer, pH 9.0). The
mixture was briefly stirred and subsequently evaporated to
dryness under a stream of nitrogen. In the second step (2),
the residue was dissolved in 300 µl of acetonitrile containing
10% (v/v) HFBA and sonicated for 30 min. Excess solvent
and reagent were removed under nitrogen. In the third
step (3), a liquid–liquid extraction, 4 ml of dichloromethane
and 1 ml of 0.5 M HCl were added to each sample and the
tubes were rotated for 30 min. After centrifugation (1 min,
3000 rpm), the upper phase was discarded and the remaining
organic phase was evaporated to dryness. The residue was
reconstituted in 100 µl of a methanol–toluene solution (4 : 1,
v/v) and, in the fourth step (4), methylation was performed
by addition of 20 µl of TMSD. The reaction was stopped after
10 min by the start of evaporation. All derivatization steps
were carried out at ambient temperature. The derivative was
dissolved in acetonitrile and stored in a refrigerator pending
analysis.
EXPERIMENTAL
Chemicals
Deionized water (Millipore, Carrigtwohill, Ireland) was
used to prepare buffers and solutions. All organic solvents
were of high-purity grade (Merck, Darmstadt, Germany).
13C6 ring-labeled L-tyrosine ([13C6]tyrosine, 99.0%) was
obtained from Cambridge Isotope Laboratories (Andover,
MA, USA). Trizma base (tris(hydroxymethyl)aminomethane
hydrochloride, Tris–HCl, 99.9%), tyrosine, 3-NT, dithionite
(sodium hydrosulfite, Na2S2O4, 85%), bovine serum albu-
min (BSA, 98%), trimethylsilyldiazomethane (TMSD, 2
M
in hexane), heptafluorobutyric anhydride (HFBA, >98%),
tetranitromethane (TNM) and pronase (Streptomyces griseus
EC 3.4.24.31) were purchased from Sigma-Aldrich (St.
Louis, MO, USA). Sodium hydroxide (NaOH, 99.0%), urea
(>99.5%), Triton X-100, sodium acetate (>99.5%) and dis-
odium hydrogenphosphate (>99%) were supplied by Merck.
High-purity grade 15N-labeled potassium nitrate (K15NO3,
99.3%) was obtained from Larodan (Lund, Sweden). A
Vivaspin 500 µl ultrafiltration cartridge (cut-off 5 kDa) was
purchased from Vivascience (Lincoln, UK).
Artifactual nitration study
We conducted several experiments to examine if artifactual
nitration of tyrosine occurs during the different steps of the
derivatization procedure.
Initially, aqueous samples were used for three model
experiments to study full-scan operation GC/NCI-MS of the
reaction products. For these experiments, stock solutions
of tyrosine at a concentration of 298 µM and 15NO3 at a
ꢀ
Blood collection
concentration of 533 µM dissolved in deionized water were
prepared. To each of four samples, 40 µl of tyrosine and
Venous blood was collected from 12 healthy, non-smoking,
30–57-year-old volunteers (six men, six women) in tubes
containing ethylenediaminetetraacetic acid (EDTA). The
EDTA tubes were pre-chilled and the blood was instantly
20 µl of 15NO3 were added. The derivatization of two of
ꢀ
these samples started as described above with the addition
of dithionite (step 1 in Fig. 1) and another two samples
started with acylation (step 2). Further, duplicate samples
containing only tyrosine were derivatized starting from step
°
centrifuged (10 min, 3000 rpm, 4 C). The plasma was stored
°
at ꢀ80 C until used.
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

GC/NCI-MS/MS quantification of 3-nitrotyrosine in plasma
1189
to samples containing a constant amount of 5 mg mlꢀ1 of
BSA. The proteins were precipitated by mixing 200 µl of the
mixture with 600 µl of methanol and 200 µl of water. The
digestion of the proteins and the quantification of 3-NT were
carried out as described for plasma proteins.
OH
H2N
O
3-nitrotyrosine
NO2
The linearity and the limit of detection were estimated
from a calibration curve prepared with known amounts of 3-
nitro[13C6]tyrosine in EDTA plasma to final concentrations of
0.0, 0.05, 0.10, 0.15, 0.20 and 0.25 nM. After centrifugation with
the Vivaspin 500, 100 µl of the ultrafiltrate were derivatized.
OH
(1)
Reduction
O
O
OH
O
H2N
NH
O
O
C-C3F7
F7C
3-C
Enzymatic digestion of plasma protein
O
HFBA
(2)
To precipitate plasma proteins, 200 µl of the plasma samples
were mixed with 600 µl of methanol and 200 µl of deionized
water. The samples were vortex mixed and centrifuged
C-C3F7
NH2
NH
O
OH
3-aminotyrosine
O
F7C3-C
Tetra-N-heptafluorobutyryl
°
(10 min, 3000 rpm, 4 C).
The protein pellets were dissolved in urea buffer (20 mM,
Tris–HCl 9 M urea and 0.35 mM SDS) and the total protein
content was determined using bicinchoninic acid according
to the manufacturer’s instructions (BCA Protein Assay
Reagent; Pierce Chemical, Rockford, IL, USA). An amount
corresponding to 1.25 mg of protein was transferred to a
microcentrifuge tube and precipitated by addition of 1 ml
of methanol, followed by centrifugation. The methanol was
removed and the pellet was washed with sodium acetate
buffer (10 mM sodium acetate 0.1% Triton X-100). To the
remaining pellet were added 225 µl of the sodium acetate
buffer and 25 µl of a stock solution (10 mg mlꢀ1 pronase in
20 mM sodium acetate) of pronase enzyme from S. griseus.20
Iiquid-liquid
extraction
(3)
CH3
O
O
NH
OH
NH
O
O
F7C3-C
F7C3-C
O
O
O
O
TMSD
(4)
C-C3F7
C-C3F7
NH
NH
OH
H3C
Di-O-methyl-di-N-heptafluorobutyryl
derivative
°
The samples were heated in a block heater at 50 C with
Figure 1. Schematic representation of the presumed
repeated vortex mixing for 1 h or until all the protein was
dissolved and were thereafter incubated and agitated in
a block heater for 18 h. Of the final digest, 50 µl were
used together with 10 µl (5 nM) of the internal standard
for derivatization and quantification of 3-NT. An additional
experiment was undertaken to study the contribution of
3-NT from pronase itself when excluding the target protein.
derivatization pathways.
ꢀ
1, but 15NO3 of the same concentration dissolved in 0.5 M
HCl was added in step 3. A tyrosine sample without addition
ꢀ
of 15NO3 was used as a reference. After derivatization, the
samples were diluted with 500 µl of acetonitrile.
In another set of three experiments, artificial nitration was
studied at biological concentrations of endogenous tyrosine
GC/MS analysis
ꢀ
and NO3 using human plasma samples. The first experi-
GC/MS analysis was carried out using a model 3400
gaschromatograph (Varian, Palo Alto, CA, USA) with an
on-column, septum-equipped programmable injector (SPI;
Varian) and a DB5-MS column (30 m ð 0.25 mm i.d., 0.25 µm
film thickness (J & W scientific, Folsom, CA, USA)). The GC
instrument was coupled to a TSQ 7000 triple-quadrupole
mass spectrometer (Thermo Finnigan, San Jose, CA, USA).
Helium was used as the carrier gas at a head pressure of
41 kPa. The GC SPI injector temperature program was from
ment was the derivatization of a blank sample consisting of
300 µl of human plasma and 100 µl of deionized water. In
a second experiment, 100 µl of [13C6]tyrosine (220 µM) was
added to 300 µl human plasma before the ultrafiltration pro-
cedure. In a third experiment, the plasma sample (300 µl) was
spiked with 100 µl of 3-nitro[13C6]tyrosine (20 nM). Complete
derivatization procedures were carried out for all samples.
After derivatization, the samples were dissolved in 100 µl of
acetonitrile. All samples were screened by the SRM technique
to establish the absence or presence of artifactual 3-NT.
ꢀ1
°
°
150 to 280 C at a rate of 100 C min without any delay
°
time. The GC column temperature was kept at 100 C for
°
Calibration curves for 3-NT and detection limit of
3-nitro[13C6]tyrosine
the first 2 min and was then increased to 280 C at a rate of
30 C minꢀ1. An amount of 1 µl of sample was loaded into
°
the GC injector in the splitless mode for all experiments.
The temperatures of the transfer line and the ion source
AcalibrationcurvewasgeneratedfrompooledEDTAplasma
samples. We added 200 µl of plasma to each of five tubes
containing known amounts of authentic 3-NT (0, 0.5, 1.0, 5.0
and 10.0 pmol) and a fixed amount of 3-nitro[13C6]tyrosine
(1 pmol).
°
housing were pre-set at 260 and 180 C, respectively. The
°
effect of ion source temperature (100, 150 and 180 C) on
the fragmentation pattern was evaluated. Methane was used
as the chemical ionization gas at pressures of 0.27–1.1 kPa.
The MS instrument was pre-set in the NCI mode at an
A second calibration curve was generated from nitrated
BSA. Nitrated BSA (0, 2, 4 and 10 µg mlꢀ1ꢀ was added
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

1190
A.-S. So¨derling et al.
ionization energy of 115 eV and 200 µA trap current. Data
acquisition and instrument control were carried out using
Xcalibur software (Thermo Finnigan).
*
8.42
100
50
A
Three different MS techniques (I–III) were employed.
The derivatization products of 3-NT samples and of the
artifactual nitration studies were recorded in (I) full-scan
mode (mass range 100–800 at a scan rate of 1 s per ion)
to study the fragmentation pattern. In (II) the selected-ion
monitoring (SIM) mode, four different pre-selected mass
center values were recorded at a scan rate of 0.1 s per ion.
Further, the fragment ion spectra of (III) collision-induced
dissociation (CID) of pre-selected parent ions were recorded
in either the SRM or full-scan operation mode. The CID full-
scan mass fragment ion spectrum was recorded in the mass
range m/z 100–700, a scan rate of 1 s per ion and a collision
energy of 10 eV using argon at collision gas pressure of
0.2 Pa. In the SRM mode, two ions with different mass center
values were pre-selected as precursor ions, and two product
ions were recorded as fragment ions at a scan rate of 0.1 s
per ion at the same collision energy.
8.55
8.64
RtlivAbudance
5.73
6.13
8.05
6.65
7.18
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Time (min)
100
B
CH3
O
NH O
O
582
576
F7C3-C
O
C-C3F7
NH
O
50
Statistical analyses and validation of the method
Data are expressed as mean š SD from representative exper-
iments. To investigate the possible relationship between the
healthy volunteers’ plasma concentrations and age, we used
Spearman’s non-parametric correlation. Limit of detection
was calculated as Ymin D YB C 3SB, where YB is the inter-
cept and SB is the standard deviation of the intercept, and
the limit of quantitation as 10 times the limit of detection.21
The corresponding concentration was calculated from the
calibration curve equation. The within-day precision was
obtained from a plasma pool divided in 10 tubes, each
spiked with 2.5 pmol of internal standard and ultrafiltered
prior to derivatization (repeatability). The between-day pre-
cision was obtained from four different 1 ml aliquots of
plasma samples each spiked with 2.5 pmol of internal stan-
dard and frozen. On two consecutive days the samples were
thawed and ultra-filtered (reproducibility) prior to deriva-
tization and the coefficient of variation (CV D 100ð SD
(mean value)ꢀ1) was calculated. To determine the recovery
and accuracy of the method, spiked pooled plasma samples
were prepared in duplicate, adding authentic 3-NT at three
levels (corresponding to an increase of 0.5, 1.0 and 1.5 nM
or 50–150% of the expected target concentration found in
plasma, 1 nM). Accuracy was calculated as 100ð ((measured
value—expected value)/expected value) and recovery as
100ð (measured value/expected value). A linear regression
model was used to estimate linearity.
H3C
RtlivAbudance
596
545
551
283
602
299
212
C
337
200
300
400
500
600
700
m/z
545
100
50
212
e
P
576
364
517
283 293
300
413
400
487
100
200
500
600
m/z
Figure 2. (A) Full-scan operational GC/NCI-MS TIC
chromatogram of the derivatization reaction product of
equimolar (0.5 µM) authentic 3-NT and 3-nitro[13C6]tyrosine.
(B) Averaged mass spectrum of the peak at tR 8.42 min. The
molecular structure of the di-O-methyldi-N-heptafluorobutyryl
derivative is depicted. (C) CID full-scan fragment ion averaged
mass spectrum of the precursor ion at m/z 576. The asterisk
denotes the analyte and Pꢀ the parent ion.
RESULTS
GC/NCI-MS/MS of authentic and labelled 3-NT
derivatives
was identified as the derivatized 3-NT product, while the
peak at tR 8.55 was a partially O-methylated-O-silylated
compound and the peak at 8.64 min was a bis-N-methylated
by-product with two CH3 groups occupying the same amino
group. The peaks at tR 6.65, 7.18 and 8.05 min were due
to impurities in the reagents. The quantitative amount of
the by-products, based on peak areas, was calculated to
Figure 2(A) shows the TIC chromatogram for GC/NCI-
MS applied to a sample of the derivatization reaction
product obtained from a mixture of authentic 3-NT and
3-nitro[13C6]tyrosine at a concentration of 0.5 pmol µlꢀ1 of
either compound. Several peaks were observed at different
retention times (tRꢀ. The predominant peak at tR 8.42 min
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

GC/NCI-MS/MS quantification of 3-nitrotyrosine in plasma
1191
be 2% of the total 3-NT content of the sample. The 3-
NT derivative was stable as the results after >6 months
in a freezer were comparable to those after storage for
24 h. The averaged mass spectrum of the peak at tR 8.42
min is also shown (Fig. 2(B)). The peak corresponding to
[M-H]ꢀ, at m/z 615, was not visible in the spectrum. The
fragment ions at m/z 596 correspond to [M-HF]ꢀ, the intense
peak at m/z 576 to [M-2HF]ꢀ and the peak at m/z 545
to [M-2HF-CH3O]ꢀ. The same fragmentation pattern was
observed for the derivative of 3-nitro[13C6]tyrosine; [M-H]ꢀ
at m/z 621 and all its fragment ions shifted by 6 Da. The results
suggest a di-O-methyldi-N-heptafluorobutyryl derivative to
be the major derivatization product of 3-NT. It should be
noted that unreduced 3-NT could not be detected under
the given GC/NCI-MS conditions. The effect of varying the
chemical ionization gas pressure with constant ionization
energy (e.g. 115 eV) was examined. This experiment (data
not shown) indicated that the fragmentation pathway was
unaffected, whereas the TIC indication of the sensitivity
reached a maximum at about 0.87 kPa. Thus, all the applied
experimental conditions resulted in the fragment ion at m/z
576 being the dominant peak in the spectrum. In addition, the
fragmentation pattern was the same at different ion source
temperatures (data not shown).
*
8.44
100
A
50
8.00
7.99
7.29
5.98
6.32
6.65
100
50
B
7.62
7.31
7.19
8.42
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Time (min)
Figure 2(C) shows the CID fragment ion averaged mass
spectrum of the precursor ion at m/z 576. The spectrum was
dominated by an intense peak at m/z 545 corresponding to
[M ꢀ CH3O]ꢀ, as confirmed from the CID mass spectrum
of the parent peak at m/z 582 corresponding to the 3-
nitro[13C6]tyrosine derivative. The peak at m/z 212 was
common for labeled and unlabeled 3-NT derivatives and
might correspond to the fragment [C3F7CONH]ꢀ.
Figure 3. (A) GC/NCI-MS SIM mode TIC chromatogram of
3-nitro[13C6]tyrosine in non-ultrafiltered aqueous solution (12.5
nM). (B) SIM mode of 3-NT of ultrafiltered human plasma
sample spiked with 3-nitro[13C6]tyrosine (12.5 nM). Volume of
100 µl of samples were derivatized and dissolved in 100 µl of
acetonitrile and 1 µl was loaded on the column. The ions at m/z
582, 576, 551 and 545 were recorded consecutively for all runs.
Evaluation of the selectivity in SIM and SRM
modes for the nitrotyrosine derivative
ꢀ
presence of dithionite and addition of 15NO3 in step
The TIC chromatograms (SIM) of a derivatized aqueous
sample (Fig. 3(A)) and of a human plasma sample (Fig. 3(B))
are illustrated. For SIM analysis, four fragment ions at m/z
582, 576, 551 and 545 were preselected and recorded. Several
peaks of different intensity and tR can be seen in Fig. 3(A)
and (B). The intense peak at tR 8.44 min in Fig. 3(A) is in
good agreement with the expected tR for the derivative. The
low-intensity peak at tR 8.42 min in Fig. 3(B) could represent
the derivative. The results strongly suggest co-elution of
unknown peaks.
The TIC chromatograms of the same aqueous and plasma
samples are also shown in the SRM mode (Fig. 4(A) and (B)).
The precursor fragment ions at m/z 582 and 576 were pre-set
and the TIC chromatogram was recorded using m/z 551 and
545 fragment ions. The m/z 551 and 545 mass ion extraction
and their quantitative mass areas are shown in the same
figures. Both figures show intense peaks at tR 8.4 min, well
in accordance with the tR for the derivative (Fig. 2(A)).
3 of the derivatization procedure (i.e. the liquid–liquid
extraction). Figure 5(B) shows the TIC chromatogram of
the derivatization products of the tyrosine sample in the
presence of dithionite and 15NO3ꢀ. Figure 5(C) represents
the TIC chromatogram of the derivatization products of the
tyrosine sample in the presence of 15NO3ꢀ and absence of
dithionite.
The full-scan averaged mass spectrum of the peak at
tR 7.93 min from Fig. 5(A) is shown in Fig. 6(A). The peak
was identified as a tyrosine derivative. The molecular ion
of the derivative was not observed, the peak at m/z 443 is
[M ꢀ HF]ꢀ and the intense fragment ion peak at m/z 371 is
[M ꢀ HF ꢀ SiCHꢁCH3ꢀ2]ꢀ, which is followed by consecutive
losses of HF groups at m/z 351 and 331, respectively.
The averaged mass spectrum of the peak at tR 9.14
min in Fig. 5(C) is shown in Fig. 6(B). The peak at m/z ž451
corresponds to the radical anion of the molecular ion Mꢀ of
nitrated tyrosine, which has been observed for compounds
with a stabilized aromatic ring, such as nitrobenzene.22 The
fragmentation pathway shows consecutive losses of HF
from the molecular ion at m/z 431 and 411. The peak at
m/z 434 represents [M ꢀ OH]ꢀ and the peak at m/z 402
represents [M ꢀ OH ꢀ CH3OH]ꢀ. The results suggest that
Artifactual nitration study
The derivatization reaction products of the artifactual
nitration study as analyzed by full-scan GC/NCI-MS are
shown in Fig. 5. Figure 5(A) is the TIC chromatogram of
the derivatization products of the tyrosine sample in the
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

1192
A.-S. So¨derling et al.
7.93
*
100
8.43
A
100
A
8.43 545 m/z
MA: 3029
551 m/z
MA: 245009
50
50
9.0
8.0
7.16
7.60
7.44
RltivAbudance
7.75
6.50
6.94
7.92
100
B
8.44
100
B
8.44 545 m/z
MA: 8797
551 m/z
MA: 95329
50
9.14
50
8.0
9.0
7.60
7.45
RltivAbudance
7.78
7.55
7.26
8.69
6.47
6.76
7.20
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Time (min)
9.14
100
C
Figure 4. GC/NCI-MS/MS SRM mode TIC chromatogram of
the selected precursor ions at m/z 582 and 576 of the same
samples as in Fig. 3. The fragment ions of m/z 551 and 545
were recorded consecutively for all runs. (A) Aqueous sample;
(B) ultrafiltered human plasma sample.
50
nitration of tyrosine to 3-NT occurs in the presence of nitrate
during step 2 (i.e. HFBA acylation). A comparison of all the
chromatograms illustrated in Fig. 5(A)–(C) reveals that the
peak at tR9.14 representing the nitrated tyrosine derivative is
absent in Fig. 5(A), present in Fig. 5(B) and predominates
the chromatogram in Fig. 5(C). Hence the nitration of
tyrosine seems to be facilitated by HFBA, but inhibited by
dithionite. The effects of different dithionite concentrations
on the artifactual 3-NT formation at fixed tyrosine and
15NOꢀ3 concentrations were investigated. The results (data
not shown) demonstrated decreasing 3-NT formation with
increasing dithionite concentration. Artifactual 3-NT could
not be detected under the given GC/NCI-MS conditions.
The absence of the di-O-methyldi-N-heptafluorobutyryl
derivative in all samples was verified by the SRM technique.
Figure 7 demonstrates the results of GC/NCI-MS/MS
(SRM) analyses of the derivatization products from the
artifact study. Figure 7(A) represents an ordinary plasma
sample. The peak at tR 8.40 min is the endogenous 3-NT
derivative (i.e. m/z 576 ! 545). The peak representing 3-
nitro[3C6]tyrosine (i.e m/z 582 ! 551) is completely absent.
Figure 7(B) is from a plasma sample with addition of
[13C6]tyrosine. No peak corresponding to the di-O-methyldi-
N-heptafluorobutyryl derivative originating from artifactual
7.93
5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 9.5 10.0
Time (min)
Figure 5. GC/NCI-MS Full-scan operational TIC
chromatograms of derivatization reaction products from an
artifactual nitration stꢀudy: (A) tyrosine in the presence of
dithionite with 15NO3 added at step 3; (B) the products of
tyrosine, diꢀthionite and 15NO3ꢀ; (C) the products of tyrosine
and 15NO3
.
nitration of [13C6]tyrosine could be observed. Figure 7(C)
shows the analysis of a plasma sample spiked with 3-
nitro[13C6]tyrosine for comparison.
Calibration curves, limit of detection and
performance over time
The calibration curve generated from authentic 3-NT and
added 3-nitro[13C6]tyrosine (internal standard) in plasma
samples showed good linearity (y D 1.00x C 1.15, R2
>
0.999). The limit of detection was calculated as 0.03 nM (30
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

GC/NCI-MS/MS quantification of 3-nitrotyrosine in plasma
1193
371
545 m/z
MA: 1628
S/N: 5
A
B
C
100
A
100
O
F7C3
NH
O
8.40
CH3
O
O
H3C
H3C
Si
CH3
100
50
551 m/z
372
333
351
233
212
283
443
331
311
O
257
219 245
359
291
390
100
100
545 m/z
MA: 2095
S/N: 5
451
100
B
F7C3
NH
CH3
8.40
O
O
NO2
H3C
O
551 m/z
50
434
431
452
402
403
411
283
264
100
100
374
545 m/z
MA: 2643
S/N: 11
200
240
280
320
360
400
440
480
m/z
8.40
8.40
Figure 6. Averaged mass spectra of peaks from Fig. 5 and the
molecular structures of the derivatives: (A) spectrum of the
peak at tR 7.93 min; (B) spectrum of the peak at tR 9.14 min.
551 m/z
MA: 15404
S/N: 117
amol injected) at a signal-to-noise ratio of 4, corresponding to
a limit of quantitation of 0.3 nM. The coefficient of variation for
3-NT measurements was 16% for between-run assays (n D 4)
and 6% for within-run assays (n D 10) of pooled individually
ultrafiltered plasma samples. The corresponding figures
obtained in aqueous samples were 6 and 3%. The calibration
curve for protein showed good linearity (R2 D 0.978).
Furthermore, we evaluated the performance over time in
terms of linearity, recovery and accuracy (Table 1). The
recovery and accuracy were good, with a minor decline
by day 3.
8.30
8.40
8.50
8.60
Time (min)
Figure 7. TIC chromatogram of SRM analysis of an artifact
nitration study: (A) the reference plasma sample; (B) the
addition of [13C6]tyrosine to the plasma sample; (C) the plasma
sample spiked with 3-nitro[13C6]tyrosine (5 nM).
Determination of basal 3-NT levels in plasma and
plasma proteins of healthy individuals
DISCUSSION
Table 2 summarizes the human basal content of free 3-NT. All
samples were run in duplicate and the amount contributed
by 3-NT from the internal standard (0.71 š 0.14% (n D 6))
was considered. We found a weak relation between age
and the plasma level of free 3-NT (r D 0.54, p < 0.07),
whereas protein-associated 3-NT was totally unrelated to
age (r D ꢀ0.04, p D 0.91). We detected no 3-NT when we
measured the proteolytic enzyme in buffer in comparison
with buffer only, both spiked with 3-nitro[13C6]tyrosine.
We have described a simple and sensitive GC/NCI-MS/MS
method for quantifying free and protein-associated 3-NT.
The method is based on the conversion of the nitro to
an amino group by dithionite and heptafluoroacylation of
amino and hydroxyl groups with HFBA. A liquid–liquid
extraction method was used to remove selectively the hep-
tafluorobutyryl from the hydroxyl groups. The final deriva-
tization consisted of O-methylation of the free hydroxyl
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

1194
A.-S. So¨derling et al.
of m/z 545 (Fig. 2(C)) (i.e. m/z 551 for 3-nitro[13C6]tyrosine),
which minimized the sensitivity loss upon collision with
argon in the collision cell. The fragment ion at m/z 545
(i.e. m/z 551 for 3-nitro[13C6]tyrosine) was selected for SRM
analyses of the derivative in biological samples. In com-
parison with the results presented by SIM (Fig. 3(B)), SRM
showed a better signal-to-noise ratio (Fig. 4(B)), which has
also been suggested by others.12,15 In addition, we found that
the derivative was stable and no rapid degradation of the
GC column was observed.
Table 1. Performance of the GC/NCI-MS/MS method in
quantifying 3-nitrotyrosine in human plasma samples
Day 1
Day 2
Day 3
Linearity (R2)
Basal 3-NT (nM)
Recovery (%)a
Accuracy (%)b
0.5 nM
0.99
0.99
0.99
0.82 š 0.007
99.4 š 0.08
0.83 š 0.005
96.6 š 1.8
0.85 š 0.003
90.3 š 0.04
3.0
ꢀ1.5
ꢀ5.3
ꢀ2.8
ꢀ5.2
ꢀ12.8
ꢀ10.6
1.0 nM
1.5 nM
ꢀ2.7
Whether applying derivatization techniques for GC/MS
orLC/MS, itisessentialtoconsiderartifactualnitration.12,15,25
The present study of artifact formation indicated that arti-
factual nitration of tyrosine occurs after the reduction step
(step 1, Fig. 1). Thus, the artifactually formed 3-NT has a di-
O-methylmono-N-heptafluorobutyryl derivative structure,
which gives a different mass and a different chromatographic
retention time compared with the target derivative (i.e. di-
O-methyldi-N-heptafluorobutyryl). Hence the interference
of artifactually formed 3-NT and its quantitative contribu-
tion were eliminated. We found that nitration of tyrosine
decreased in the presence of dithionite (Fig. 4(B)). Although
the reaction between dithionite and nitrate ions is not fully
understood, this diminished nitration may be due to a reduc-
tion of nitrate to gaseous species such as NO. Decreasing the
nitrate concentration in the sample using dithionite decreases
the likelihood of artifact formation. It should be noted that a
peak representing the di-O-methyldi-N-heptafluorobutyryl
derivative was completely absent, demonstrating absence
of 3-NT formation in step 1. This was verified by the SRM
technique at the limit of detection of our instrument. Fur-
thermore, the formation of artifactual 3-NT was examined
under biological conditions using [13C6]tyrosine and plasma
samples. The SRM results (Fig. 7) demonstrated no forma-
tion of the di-O-methyldi-N-heptafluorobutyryl derivative.
Hence we believe that the quantified basal endogenous 3-NT
reported in this work (Table 2) has not been influenced by
an artifact.
We did not attempt to measure plasma tyrosine. As also
reported by others,12,15 nitration of tyrosine can occur in the
presence of the nitrate anion. It might influence tyrosine
quantification without a previous nitrate pre-clearance
of plasma.8,9,15,25 Furthermore, the plasma concentration
of tyrosine is of the order of micromolar compared to
with nanomolar for 3-NT, and using isotopically labeled
tyrosine as an internal standard is an economical issue.
Since our intention was to develop a simple and artifact-
free high-throughput method for the quantification of
endogenous 3-NT, we suggest that the tyrosine content can
be easily and accurately measured by cheaper HPLC when
appropriate.
ꢀ3.0
a Mean š SD, n D 2, on adding authentic 3-NT corresponding to
increased concentrations of 0.5, 1.0 and 1.5 n M.
b Mean, n D 2.
Table 2. Characteristics of healthy non-smoking voluntary
subjects and their basal endogenous 3-NT content in
ultra-filtered plasma and total plasma protein as determined by
GC/NCI-MS/MS (SRM mode)
3-NT in
Blood
pressure
(mm Hg)
total plasma
protein
(pmol mgꢀ1
ꢀ
Age
3-NT in
plasma (nM)
(years)
Gender
40
57
44
52
35
36
36
53
51
49
30
30
Mean
SD
M
F
125/75
125/85
125/85
115/80
145/80
140/90
110/70
110/70
110/70
100/65
115/70
110/70
0.51
1.11
0.37
0.65
0.63
0.75
0.62
1.41
0.93
0.70
0.31
0.92
0.74
0.31
1.88
0.42
0.49
0.43
0.42
0.55
0.52
0.71
0.60
0.30
0.49
0.35
0.60
0.42
M
M
M
M
F
F
F
F
M
F
43
9
and carboxyl groups with TMSD. All the derivatization
steps were carried out under mild experimental condi-
tions. The major derivatization product was identified as
a di-O-methyldi-N-heptafluorobutyryl derivative by full-
scan GC/NCI-MS. In comparison with the n-propyl per-
fluorobutyryl derivative of 3-NT,9 the di-O-methyldi-N-
heptafluorobutyryl derivative showed improved GC/NCI-
MS properties.
The fragmentation of the di-O-methyldi-N-heptafluoro-
butyryl derivative under the ionization conditions of the
present study is demonstrated in Fig. 2(B), where the promi-
nent fragment ion at m/z 576 corresponds to [M ꢀ 2HF]ꢀ
for authentic 3-NT and at m/z 582 for 3-nitro[13C6]tyrosine.
The molecular ion [M ꢀ H]ꢀ was of low abundance. Instead,
we found that the most characteristic ions were [M ꢀ HF]ꢀ
and [M ꢀ 2HF]ꢀ, which is in accord with previous studies of
steroids and aromatic amino acids.23,24 The CID of m/z 576 as
a precursor ion remarkably yielded a prominent fragment ion
Our results show that free circulating 3-NT was present
in basal human blood samples from healthy, 30–57-year-old
volunteers. The levels were low in these normal subjects,
which is in agreement with findings by Schwedhelm
et al.,12 whose values are also in the lower range of those
reported.11,12,15 In fact, we obtained plasma values in healthy
volunteers (0.7 nM) identical with recently published data
from the Hannover group.17 There are few previous reports
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

GC/NCI-MS/MS quantification of 3-nitrotyrosine in plasma
8. Frost MT, Halliwell B, Moore KP. Analysis of free and
1195
on 3-NT levels in total human plasma proteins. One
study using MS reported these levels to be 2.7 ng mgꢀ1
protein (12 pmol mgꢀ1ꢀ.8 This should be compared with our
level of 0.6 pmol mgꢀ1. Tsikas et al.17 found the protein-
associated 3-NT level in human serum albumin (HSA) from
plasma to be 0.42 3-NT per 106 tyrosines ꢁ0.1 pmol mgꢀ1ꢀ,
which is fairly close to our reported total plasma protein
levels. The 3-NT level in HSA is not necessarily the same
as in total plasma proteins but logically of the same
magnitude since HSA is the most abundant protein in human
plasma.
Although it is possible also to apply LC/MS/MS for
the measurement of 3-NT, it seems likely that GC/MS/MS
is more sensitive, as the detection limit appears to be
of the order of 1.5 nM when applying the former tech-
nique, allowing the detection of 3-NT in only 50% of
rat plasma samples.26 Age-related changes in different
blood compartments have been described for some amino
acids.27 In our healthy volunteers, we found a weak (non-
significant) relation between plasma concentrations of 3-NT
and age, which might have been significant in a larger
number of subjects. However, for protein-associated 3-
NT after enzymatic digestion, we found no relationship
with age.
protein-bound nitrotyrosine in human plasma by
a gas
chromatography/mass spectrometry method that avoids
nitration artifacts. Biochem. J. 2000; 345(Pt 3): 453.
9. Crowley JR, Yarasheski K, Leeuwenburgh C, Turk J, Hei-
necke JW. Isotope dilution mass spectrometric quantification of
3-nitrotyrosine in proteins and tissues is facilitated by reduction
to 3-aminotyrosine. Anal. Biochem. 1998; 259: 127.
10. Leeuwenburgh C, Hardy MM, Hazen SL, Wagner P, Oh-ishi S,
Steinbrecher UP, Heinecke JW. Reactive nitrogen intermediates
promote low density lipoprotein oxidation in human
atherosclerotic intima. J. Biol. Chem. 1997; 272: 1433.
11. Tsikas D, Schwedhelm E, Frolich JC. Methodological consider-
ations on the detection of 3-nitrotyrosine in the cardiovascular
system. Circ. Res. 2002; 90: E70.
12. Schwedhelm E, Tsikas D, Gutzki FM, Frolich JC. Gas
chromatographic–tandem mass spectrometric quantification of
free 3-nitrotyrosine in human plasma at the basal state. Anal.
Biochem. 1999; 276: 195.
13. Smythe GA, Matanovic G, Yi D, Duncan MW. Trifluoroacetic
anhydride-catalyzed nitration of toluene as an approach to
the specific analysis of nitrate by gas chromatography–mass
spectrometry. Nitric Oxide 1999; 3: 67.
14. Tesch JW, Rehg WR, Sievers RE. Microdetermination of nitrates
and nitrites in saliva, blood, water, and suspended particulates
in air by gas chromatography. J. Chromatogr. 1976; 126: 743.
15. Yi D, Ingelse BA, Duncan MW, Smythe GA. Quantification of
3-nitrotyrosine in biological tissues and fluids: generating valid
results by eliminating artifactual formation. J. Am. Soc. Mass
Spectrom. 2000; 11: 578.
16. Gaut JP, Byun J, Tran HD, Heinecke JW. Artifact-free
quantification of free 3-chlorotyrosine, 3-bromotyrosine, and
3-nitrotyrosine in human plasma by electron capture-negative
chemical ionization gas chromatography mass spectrometry and
liquid chromatography–electrospray ionization tandem mass
spectrometry. Anal. Biochem. 2002; 300: 252.
On the basis of the present experiments, we plan to inves-
tigate circulating levels of nitrotyrosine in plasma and other
biological fluids from patients with various cardiovascular
disorders and risk factors.
Acknowledgements
17. Tsikas D, Schwedhelm E, Stutzer FK, Gutzki FM, Rode I,
Mehls C, Frolich JC. Accurate quantification of basal plasma
levels of 3-nitrotyrosine and 3-nitrotyrosinoalbumin by gas
chromatography–tandem mass spectrometry. J. Chromatogr. B
2003; 784: 77.
This study was supported by the Swedish Research Council (K 2002-
71X-14231-01A and K99-04RM-13192-01), the Swedish Heart and
Lung Foundation, the Va¨stra Go¨taland region, Go¨teborg University,
Jubileumsfonden and the Foundation for Neurological Research,
Go¨teborg, Sweden. The discussion of the project with Professor Peter
Roepstorff, Department of Biochemistry and Molecular Biology,
Odense University, Denmark, is greatly appreciated. We are grateful
to Mrs Ingrid Larsson for language revision.
18. Sokolovsky M, Riordan JF, Vallee BL. Tetranitromethane.
A
reagent for the nitration of tyrosyl residues in proteins.
Biochemistry 1966; 5: 3582.
19. Sokolovsky M, Riordan JF, Vallee BL. Conversion of 3-
nitrotyrosine to 3-aminotyrosine in peptides and proteins.
Biochem. Biophys. Res. Commun. 1967; 27: 20.
20. Hensley K, Maidt ML, Pye QN, Stewart CA, Wack M,
Tabatabaie T, Floyd RA. Quantitation of protein-bound 3-
nitrotyrosine and 3,4-dihydroxyphenylalanine by high-
performance liquid chromatography with electrochemical array
detection. Anal. Biochem. 1997; 251: 187.
REFERENCES
1. Ischiropoulos H. Biological tyrosine nitration: a pathophysiolog-
ical function of nitric oxide and reactive oxygen species. Arch.
Biochem. Biophys. 1998; 356: 1.
2. MacMillan-Crow LA, Crow JP, Thompson JA. Peroxynitrite-
mediated inactivation of manganese superoxide dismutase
involves nitration and oxidation of critical tyrosine residues.
Biochemistry 1998; 37: 1613.
21. Miller JN, Miller JC. Statistics and Chemometrics for Analytical
Chemistry. (4th edn). Ellis Horwood: Chichester, 2000.
22. Laramee JA, Kocher CA, Deinzer ML. Application of
a
3. Kaur H, Halliwell B. Evidence for nitric oxide-mediated
oxidative damage in chronic inflammation. Nitrotyrosine in
serum and synovial fluid from rheumatoid patients. FEBS Lett.
1994; 350: 9.
4. Aulak KS, Miyagi M, Yan L, West KA, Massillon D, Crabb JW,
Stuehr DJ. Proteomic method identifies proteins nitrated in vivo
during inflammatory challenge. Proc. Natl. Acad. Sci. USA 2001;
98: 12 056.
5. Ischiropoulos H, al-Mehdi AB. Peroxynitrite-mediated oxida-
tive protein modifications. FEBS Lett. 1995; 364: 279.
6. Herce-Pagliai C, Kotecha S, Shuker DE. Analytical methods for
3-nitrotyrosine as a marker of exposure to reactive nitrogen
species: a review. Nitric Oxide 1998; 2: 324.
7. Halliwell B, Zhao K, Whiteman M. Nitric oxide and
peroxynitrite. The ugly, the uglier and the not so good: a personal
view of recent controversies. Free Rad. Res. 1999; 31: 651.
trochoidal electron monochromator/mass spectrometer system
to the study of environmental chemicals. Anal. Chem. 1992; 64:
2316.
23. Choi MH, Chung BC, Lee W, Lee UC, Kim Y. Determination
of anabolic steroids by gas chromatography/negative-
ion chemical ionization mass spectrometry and gas
chromatography/negative- ion chemical ionization tandem
mass spectrometry with heptafluorobutyric anhydride
derivatization. Rapid Commun. Mass Spectrom. 1999; 13: 376.
24. Hazen SL, Crowley JR, Mueller DM, Heinecke JW. Mass
spectrometric quantification of 3-chlorotyrosine in human
tissues with attomole sensitivity:
a sensitive and specific
marker for myeloperoxidase-catalyzed chlorination at sites of
inflammation. Free Rad. Biol. Med. 1997; 23: 909.
25. Delatour T, Richoz J, Vouros P, Turesky RJ. Simultaneous
determination of 3-nitrotyrosine and tyrosine in plasma proteins
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196

1196
A.-S. So¨derling et al.
of rats and assessment of artifactual tyrosine nitration. J.
Chromatogr. B 2002; 779: 189.
26. Delatour T, Guy PA, Stadler RH, Turesky RJ. 3-Nitrotyrosine
butyl ester: a novel derivative to assess tyrosine nitration in rat
plasma by liquid chromatography–tandem mass spectrometry
detection. Anal. Biochem. 2002; 302: 10.
27. Proenza AM, Palou A, Roca P. Amino acid distribution in
human blood. A significant pool of amino acids is adsorbed
onto blood cell membranes. Biochem. Mol. Biol. Int. 1994; 34: 971.
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 1187–1196