Ebselen as a peroxynitrite scavenger in vitro and ex vivo

Article abstract of DOI:10.1016/S0006-2952(99)00309-3

We have previously shown that peroxynitrite (PN) selectively impaired prostacyclin (PGI₂)-dependent vasorelaxation by tyrosine nitration of PGI₂ synthase in an in situ model (Zou MH, Jendral M and Ullrich V, Br J Pharmacol 126: 1283-

Full text of DOI:10.1016/S0006-2952(99)00309-3

Biochemical Pharmacology, Vol. 59, pp. 153–160, 2000.
1999 Elsevier Science Inc. All rights reserved.
ISSN 0006-2952/00/$–see front matter
PII S0006-2952(99)00309-3
©
Ebselen as a Peroxynitrite Scavenger In Vitro and
Ex Vivo
Andreas Daiber, Ming-Hui Zou, Markus Bachschmid and Volker Ullrich*
FACULTY OF BIOLOGY, UNIVERSITY OF KONSTANZ, 78457 KONSTANZ, GERMANY
ABSTRACT. We have previously shown that peroxynitrite (PN) selectively impaired prostacyclin (PGI )-
2
dependent vasorelaxation by tyrosine nitration of PGI synthase in an in situ model (Zou MH, Jendral M and
2
Ullrich V, Br J Pharmacol 126: 1283–1292, 1999). By using this established model, we tested whether or not
ebselen (2-phenyl-1,2-benzisoselenazol-3(2H)-one), which reacts rapidly with the anionic form of PN, affected
PN inhibition of PGI synthase. Administration of ebselen (1 to 50 ␮M) to bovine coronary strips 5 min prior
2
to PN (1 ␮M) treatment neither prevented PN-triggered vasoconstriction nor the inhibition of PGI release. In
2
1
4
line with these results, ebselen affected neither PN inhibition of the conversion of [ C]-PGH into
2
6
-keto-PGF1␣ nor the nitration of PGI synthase in bovine aortic microsomes. Following the hypothesis that a
2
reaction of ebselen with cellular thiols could have caused the inefficiency of ebselen, we observed that free
ebselen quickly reacted with thiols in both coronary strips and in aortic microsomes to form two metabolites, one
of which was identified as the ebselen–glutathione adduct, whereas the other had a similar retention time to that
of the ebselen–cysteine adduct. The nitration of phenol by PN in a metal-free solution could be blocked more
efficiently in the presence of ebselen or glutathione alone than in the presence of both, indicating that like
selenomethionine and other selenocompounds, ebselen–thiol adducts were less reactive towards PN than ebselen
itself. Further evidence came from the results that ebselen became effective in preventing the inhibition and
nitration of PGI synthase after thiol groups of microsomal proteins were previously oxidized with Ellman’s
2
reagent. We conclude that in cellular systems ebselen is present as thiol adducts and thus loses its high reactivity
towards PN, which is required to compete with the nitration of PGI synthase. BIOCHEM PHARMACOL 59;2:
2
1
53–160, 2000. © 1999 Elsevier Science Inc.
KEY WORDS. prostacyclin synthase; tyrosine nitration; glutathione adduct; selenocompounds; antioxidants
NO† has been established as a potent messenger in biolog-
ical signaling pathways which are mostly linked to a
cules such as DNA [5], lipids [6], or proteins [7] have been
reported (for review see [8, 9]). Typical reactions of proteins
include thiol and methionine oxidations [10, 11] or ty-
rosine nitrations [12, 13], which have been observed under
pathological conditions such as tissue graft rejection [14],
ischemic events [15], or neurodegenerative diseases [9].
Therefore, efforts have been undertaken to block PN
reactions either by preventing its formation or by quench-
ing it with antioxidants [16–18]. Indeed, several reports on
efficient scavengers for PN suggest that antioxidants are
promising tools for suppressing PN-mediated oxidations
and could be potential drugs against its presumed damaging
effects in tissues [19].
down-regulation of cellular activity [1, 2]. Due to its almost
Ϫ
2
diffusion-limited reaction with superoxide (O ) [3, 4], this
radical becomes an efficient antagonist of NO. Whereas
each radical alone has rather weak oxidizing properties, the
resulting product, PN, has been described as a strong
oxidizing species either by one-electron acceptance or as an
oxene donor equivalent to a two-electron acceptor. It is
mainly the chemical structure of the reactant that will
determine the mechanism by which PN will oxidize a given
substrate. Since PN can be formed in biological systems, its
reactivity with cellular constituents has often been inves-
tigated, and many destructive reactions with macromole-
We recently reported a very sensitive inhibition by PN of
PGI synthase, which was paralleled by a tyrosine nitration
2
at the active site [13, 20]. Using intact cells and tissues, this
modification could also be observed ex vivo [21], which at
first sight seemed incompatible with the fact that the high
cellular antioxidant potential should have eliminated such
low PN concentrations already effective as 0.1 ␮M bolus
additions. A closer investigation into the chemical mech-
*
Corresponding author: Prof. Volker Ullrich, Universit a¨ t Konstanz,
Fakult a¨ t f u¨ r Biologie, Fach X 911—Sonnenb u¨ hl, D 78457 Konstanz,
Germany. Tel. ϩ49 7531 88 2287; FAX ϩ49 7531 88 4084; E-mail:
volker.ullrich@uni-konstanz.de
†
Abbreviations: PN, peroxynitrite; PGI , prostacyclin; PCS, prostacy-
2
clin synthase; NO, nitric oxide; DTNB, 5,5Ј-dithiobis-(2-nitrobenzoic
acid) (Ellman’s reagent); PGH , 15-hydroxy-11 alpha, 9 alpha-(ep-
oxymethano)prosta-5, 13-dienoic acid; A II, angiotensin II; PMSF, phe-
nylmethanesulfonyl fluoride; ECL, enhanced chemiluminescence; and
LPS, lipopolysaccharide.
2
anism of tyrosine nitration of PGI synthase revealed a
2
catalytic action of its heme active site, allowing the PN
anion to directly and rapidly form the nitrating species [22]
Received 26 April 1999; accepted 1 July 1999.

1
54
A. Daiber et al.
and not be converted to its excited protonated state, which
is considered to be the active form of PN in the absence of
transition metals [3, 23]. This could also explain the failure
of antioxidants such as ascorbate, glutathione, vitamin E, or
␮M) was incubated with 0.134 g dry weight (first sample)
and 0.077 g (second sample) of bovine coronary tissue in 10
mL PBS buffer and aliquots of 50 ␮L were taken after 1, 15,
30, and 35 min and injected on HPLC. Ebselen and
GSH–ebselen were identified by external standards. (2)
Five hundred microliters of bovine aortic microsomes (1
mg/mL protein) was incubated with 500 ␮L of 1 mM
DTNB for 10 min in 0.2 M K-phosphate buffer pH 7.5. The
solution turned yellow, indicating the reaction of DTNB
with free thiol groups. After centrifugation, the supernatant
was removed and 1 mL fresh K-phosphate buffer was added.
The sample was vortexed and then 20 or 50 ␮M ebselen
was added and incubated for 5 min. Fifty microliters of this
solution was injected on HPLC. The same procedure was
performed for the non-DTNB-treated microsomes. Con-
centrations of ebselen were quantified by external stan-
dards.
uric acid to block the inhibition of PGI synthase by PN in
2
rat mesangial cells [21]. As a result of such findings, we set
out to look for inhibitors that directly attack and quench
the PN anion in order to compete with PGI synthase. An
2
ideal candidate should be the drug ebselen (2-phenyl-1,2-
benzisoselenazol-3(2H)-one), since its second-order reac-
6
Ϫ1 Ϫ1
tion constant with PN was found to be 2*10 M
s
[24]
and it should therefore be able to effectively block the
inhibition of PGI synthase in vitro and in vivo. However,
2
we report herein that such expectations were not fulfilled in
an ex vivo model which, on the other hand, provided
interesting insights into the mechanism of ebselen action.
MATERIALS AND METHODS
Materials
Protective Effects of Ebselen, GSH, and Ebselen-GSH
on Phenol and BSA Nitration
All chemicals were purchased from Sigma or Fluka at the
highest purity available. Ebselen was a gift from Prof.
Wendel (University of Konstanz). Bovine aortic micro-
somes were prepared as described in [25]. PN was prepared
Solutions of 5 mM phenol in 0.1 M K-phosphate buffer at
pH 7 or 9 were supplemented with ebselen, GSH, or both,
and were reacted with 1 mM PN. The yield of nitrophenols
(pH 7) or 4-nitrosophenol (pH 9) was determined from the
absorption at 400 nm (after addition of NaOH(aq)). PN
and ebselen decomposition kinetics were recorded at 302
against 370 nm. At this wavelength, both PN and ebselen
absorb, but not ebselenoxide, and the ebselen–GSH adduct
has a very low absorption. PN was always added last to the
stirred solution in the cuvette by a syringe through a
septicum. The decomposition of 100 ␮M PN and 50 ␮M
ebselen with 0, 20, or 40 ␮M BSA in 0.2M K-phosphate
buffer at pH 10 was recorded as was the decomposition of
from NO and KO according to Koppenol and Kissner [4].
2
Two HPLC systems were used: 1) Jasco PU-980 pump,
UV-975 variable-wavelength detector, and LG-980-02
low-pressure gradient mixer; and 2) LKB 2150 pump,
Spectra Physics spectra focus detector, and SP4290 integra-
tor. In both systems solvents were gassed with helium.
Kinetics were recorded on an Aminco DW-2 UV/Vis
spectrophotometer in the dual-wavelength mode, equipped
with a magnetic stirrer.
2
4
50 or 500 ␮M PN and 200 ␮M ebselen with 0, 200, and
00 ␮M GSH in 0.2 M K-phosphate buffer pH 8, 9, and 10.
Separation of Ebselen, GSH–Ebselen, and Ebselenoxide
System I was used with a Nucleosil 300-5 C column
4
2
50*4.6 (Macherey–Nagel), and the mobile phase con-
Ebselen Effects on PN-Treated Bovine Coronary
Arteries
sisted of 80% 0.1 M K-phosphate buffer pH 4.5 and 20%
acetonitril. The flow was 1.2 mL/min and the products were
detected at 270 nm. Spectra were taken from each peak.
Solutions of 500 ␮M ebselen (from a 10-mM stock solution
of ebselen in acetonitril) in 0.1 M K-phosphate buffer at pH
Bovine hearts were obtained from the local slaughterhouse.
The epicardial coronary arteries of the left ventricle were
quickly dissected, cleaned of adhering fat and connective
tissues, and placed in an ice-cold Krebs-bicarbonate solu-
7
and 9 were treated with different concentrations of GSH
Ϫ1
(0.05–5 mM) and were then reacted with 500 ␮M PN. Fifty
tion consisting of (in mmolL ): NaCl 118; KCl 4.7; CaCl2
microliters of each sample were injected on the HPLC.
2.5; MgSO 1.2; KH PO 1.2; NaHCO 25; EDTA 0.016;
4
2
4
3
glucose 11.1. The tissues were kept in ice-cold Krebs buffer
for up to 24 hr with frequent buffer changes. For the
isometric measurement of tension in bovine coronary
arteries, a spiral (20 ϫ 5 mm) cut from bovine coronary
arteries was suspended in a tissue bath filled with 15 mL of
Krebs buffer, gassed with 95% O , 5% CO (37°; pH 7.4)
Determination of Free Ebselen Concentrations in Bovine
Aortic Microsomes and Tissue-containing Solutions
System II was used with a Nucleosil 300-5 C column
4
2
50*4.6 (Macherey–Nagel), with the mobile phase consist-
2
2
ing of 80% 0.1 M K-phosphate buffer pH 4.5 and 20%
acetonitril for the tissue-samples (1) and 75% water and
and attached to a force–displacement TFT6V5 transducer
coupled to a polygraph for the measurement of isometric
tension. Passive tension was adjusted to approximately 5 g
over a 30-min equilibrating period and then coronary
arteries were constricted by addition of the thromboxane
2
5% acetonitril with 0.1% added trifluoroacetic acid in the
microsomal experiments (2). The flow was 1.2 mL/min and
the products were detected at 270 nm. (1) Ebselen (4.3

Ebselen as a Peroxynitrite Scavenger
155
Ϫ1
mimetic U46619 (0.001 ϳ 0.01 ␮molL ). The vasorelax-
Activity Assay of PGI Synthase
2
Ϫ1
ation after acetylcholine (1 ␮molL ) was used to demon-
Bovine aortic microsomes were treated with the indicated
strate a functional endothelium. Throughout the experi-
ment, care was taken to avoid any injury to the endothe-
lium. After rinsing the tissue with fresh buffer and allowing
the mechanical tension to obtain a steady-state value, a
reference response of vasocontraction–relaxation was ob-
tained for each spiral by addition of A II (50 nmolL ).
Thirty minutes after A II stimulation, the medium was
collected for prostanoid analysis and the tissue was treated
with ONOO as follows: the vessel was placed into a 2-mL
Eppendorff tube containing 0.9 mL prewarmed Krebs buffer
with 0.1–50 mM ebselen mixed thoroughly with 0.1 mL
ONOO (0.1 to 10 mM) to give the indicated concentra-
tion of ONOO , and was quickly and thoroughly mixed.
Ϫ
concentrations of ONOO and were quickly and thor-
oughly mixed. Two minutes after treatment, 100
Ϫ1 14
␮
molL [ C]-PGH2 was added and the samples were
incubated for an additional 3 min. The reaction was
stopped by acidification with 1 N HCl to pH 3.5. The
incubation media were extracted with 3 volumes of ethyl
acetate and after centrifugation, the organic phases were
evaporated to dryness under nitrogen. Samples were then
resuspended in 60 ␮L of ethyl acetate and subsequently
separated by TLC (ethyl acetate/water/isooctane/acetic
acid; 90:100:50:20). Prostanoids were quantified with a
phosphor imager system (Image Quant, Molecular Dynam-
ics) as previously described.
Ϫ1
Ϫ
Ϫ
Ϫ
Control vessels were treated with the same volume of
alkaline Krebs buffer medium or decomposed ONOO (24
Ϫ
hr at room temperature). Two minutes after treatment, the
tissue was resuspended in the organ bath without adjusting
the tension. Where indicated, the pharmacological agents
indomethacin, SQ29548, CGS 13080, and U51605 were
added to the organ bath immediately after treatment. After
Immunoprecipitation of 3-Nitrotyrosine-containing
Proteins
Bovine aortic microsomes (5 mg/mL) were diluted (1 to 1)
Ϫ1
Ϫ1
in 50 mmolL Tris buffer containing 1 mmolL PMSF, 5
Ϫ1
Ϫ1
mmolL EDTA, 150 mmolL NaCl, and 0.5% Nonidet
P-40 (pH 8.0) and for 30 sec in a Branson Sonifier 250
6
0 min of re-equilibration, the tissue was stimulated with
the same concentration of A II for another 30 min. At the
end of the experiment, vessels were collected for immuno-
precipitation and incubation media were collected and
stored at Ϫ20° for prostaglandin and thromboxane analysis.
The percentage of vasoconstriction/relaxation was calcu-
lated as the percentage of relaxation in the secondary
stimulation with A II relative to that in the first stimula-
tion. Similarly, the amounts of prostanoids were expressed
as the percentage of prostanoids formed by the second
stimulation compared to those after the first stimulation
(
Schw a¨ bisch Gm u¨ nd), setting 5, at 50% duty cycle and
centrifuged (14,000 g) at 4° for 5 min to remove cellular
debris. Protein concentrations were determined using the
Bradford assay. Solubilized proteins (3 mg) were precleared
by addition of 40 ␮L of protein A-Sepharose CL-4B and
the supernatant was incubated (18 hr, 4°) with 10 ␮g of
monoclonal anti-nitrotyrosine antibody (Upstate Biotech-
nology). Immune complexes were precipitated with 30 ␮L
of protein A CL-4B and washed with 0.5 mL SNNTE
Ϫ1
Ϫ
(0.5% sucrose, 1% Nonidet P-40, 0.5 molL NaCl, 50
with A II. In order to evaluate the effects of ONOO on
Ϫ1
Ϫ1
mmolL Tris, and 5 mmolL EDTA, pH 7.4). Resus-
pended protein pellets were then precipitated by centrifu-
gation (14,000 g, 1 min) and resuspended in 40 ␮L
Laemmli sample buffer/␤-mercaptoethanol (9:1), heated at
95° for 5 min and kept on ice.
the receptor-dependent or receptor-independent vasocon-
striction mediated by U46619, PGH , or KCl, a reference
2
contractile response for each compound was initially ob-
tained for each spiral. Subsequently, the tissues were treated
Ϫ
with ONOO or alkaline Krebs solution as described above
and then resuspended in the organ bath containing fresh
Krebs buffer. Once the tension had returned to baseline,
the vessels were treated by addition of the same concentra-
Western Blots
Ϫ8
Ϫ1
Ϫ6
Ϫ1
tion of U46619 (5 ϫ 10 molL ), PGH (10 molL ),
Protein precipitates were separated by 7.5% SDS–PAGE
(30 mA, 1 hr) and blotted for 1 hr with a constant current
of 200 mA onto a nitrocellulose membrane in a semidry
blot procedure (48 mmolL Tris/ 39 mmolL glycine/
20% methanol/ 0.037% SDS). Proteins were visualized
with a 0.1% Ponceau S solution in 5% acetic acid to check
transfer efficiency. After destaining, the membrane was
blocked with 5% milk powder in PBS/ 0.1% Tween 20 for
2 hr at room temperature and incubated with a polyclonal
2
Ϫ4
Ϫ1
or KCl (6 ϫ 10 molL ). Agonist-induced vasconstric-
tion of coronary arteries was calculated as the percentage of
vasoconstriction in the secondary stimulation compared to
the reference contractile response.
Ϫ1
Ϫ1
Extraction and Analysis of Prostanoids
After acidification with 1 N HCl to pH 3.5, the collected
medium was extracted with 3 volumes of ethyl acetate. The
organic phases were collected and evaporated to dryness
under nitrogen. Samples were resuspended in PBS contain-
antibody directed against PGI -synthase [13, 20] (1 ␮g/mL)
2
overnight at 4°. After washing several times with PBS/0.1%
Tween 20, the membrane was further incubated with a goat
anti-rabbit antibody at a dilution of 1:7500 for 45 min.
Antibody binding was visualized by the ECL technique,
according to the instructions of the supplier (Amersham).
ing 0.1% BSA. Prostanoids and thromboxane B were
2
analyzed by enzyme-linked immunoassay kits, according to
the instructions provided by the supplier.

1
56
A. Daiber et al.
FIG. 1. (A) Time-dependent loss of 4.3 ␮M “free” ebselen (white bars) in a solution containing 0.134 g (dry weight) bovine coronary
tissue and parallel formation of the ebselen–GSH adduct (grey, area/1000) and an unknown product (black, area/1000) (similar
retention time as the ebselen–cysteine adduct). Thirty minutes after addition of 5 ␮M GSH, all the ebselen and the unknown product
were converted to the ebselen–GSH adduct. Separation and quantification was done by HPLC. This diagram shows the results of one
out of two almost identical experiments. (B) UV/Vis spectra of 100 ␮M ebselen incubated for 5 min with 0 (2), 50 (3), 100 (4), 200
(5), and 400 ␮M (6) GSH in 0.2 M K-phosphate buffer pH 7. (1)-buffer without ebselen and GSH.
The blots which had been stained with a polyclonal
antibody against PGI2 synthase were then stripped by
with sulfhydryl groups in the tissue during the preincuba-
tion phase, as was the case for albumin [28]. This was tested
by setting up an HPLC system for the detection of ebselen
and its possible reaction products.
Ϫ1
incubating the membrane in stripping buffer (100 mmolL
Ϫ1
2
6
-mercaptoethanol, 2% SDS, 62.5 mmolL Tris–HCl, pH
.7) at 50° for 30 min with agitation. The membranes were
washed, blocked, and incubated overnight at 4° with a
monoclonal antibody against 3-nitrotyrosine at a dilution
of 1 ␮g/mL and with a goat anti-mouse antibody for 45 min
at a dilution of 1:7500. Antibody binding was visualized as
described above.
The Fate of Ebselen in Tissues
The fact that ebselen forms adducts with thiols, especially
glutathione (GSH) [29], could have been responsible for
the lack of the expected ebselen effect. Indeed, ebselen
disappeared from the bath solution time dependently and
two metabolite peaks were formed concomitantly, as seen
in Fig. 1A. According to the optical spectra, the band with
a retention time of 7 min proved to be identical to the
spectra of the ebselen–GSH adduct as seen from a titration
of ebselen with GSH (Fig. 1B). The second metabolite was
not identified, but it did have a similar retention time to
that of the ebselen–cysteine adduct and disappeared after
excess addition of GSH to the bath solution. Not only GSH
but also reactive thiols in proteins can bind ebselen, as had
been shown for albumin and lipoate [28, 29]. Using a
microsomal fraction from aortic rings, it was possible by
HPLC to observe a decrease in “free” ebselen after addition
of this membrane fraction (Expt 1:50 mM to 13; Expt 2:20
mM to 0). However, if the protein-bound thiols had been
oxidized with Ellman’s reagent (DTNB), more ebselen was
recovered (Expt 1:50 mM to 38; Expt 2:20 mM to 4 mM).
RESULTS
Effects of Ebselen on PN Action in Bovine Aortic Rings
PGI is an effective mediator and relaxant of vascular tone
2
and can be studied via its pharmacological actions in
bovine aortic rings [26]. Using this model, we showed a
sensitive inhibition by PN of the relaxation phase after
angiotensin II stimulation of this tissue, paralleled by a
decrease in PGI formation [27]. In this ex vivo model of
2
PGI action, it was also possible to measure decreased PGI
2
2
levels as a consequence of PGI synthase inhibition by PN
2
and by parallel immunoprecipitations to quantitate the
amount of nitrated enzyme. In a similar series of experi-
ments, we used LPS as a presumed trigger of PN and indeed
were able to mimic the effects of PN in this in situ model of
vascular tone (unpublished result). When looking for a
pharmacological tool to inhibit PN action, we knew that
the selenocompound ebeselen had been reported to react
Since this microsomal fraction also contained PGI syn-
2
6
Ϫ1
with PN at a second-order rate constant of 2.*10 M
thase, it was possible to conduct an experiment with PN
and ebselen with and without previous treatment with
DTNB. Additions of 1 and 10 ␮M PN caused the expected
inhibitions of 6-keto PGF formation in the presence of
Ϫ1
sec [24], suggesting that it might be an efficient scaven-
ger. However, additions of up to 10 ␮M ebselen to aortic
rings failed to show any effect (results not shown). On the
basis of the reported kinetic constants, 10 ␮M ebselen
should have been sufficient to quench the inhibitory action
of PN at the submicromolar level. A possible explanation
for the lack of effect of ebselen could have been its reaction
1
␣
1
4
[ C]-PGH , with DTNB having no significant modulatory
2
effect. However, when 10 ␮M ebselen was added before
PN, the inhibition was largely prevented (Table 1). A
qualitative confirmation of this result was obtained by

Ebselen as a Peroxynitrite Scavenger
157
TABLE 1. Protection of prostacyclin synthase from inhibition
by peroxynitrite in DTNB-treated and non-treated bovine
aortic microsomes
%
inhibition of % inhibition of
Microsomes (PCS)
PCS (control)
PCS (DTNB)
ϩ 1 ␮M PN
57.3 Ϯ 7.1
73.6 Ϯ 6.9
55.4 Ϯ 5.6
74.1 Ϯ 6.9
50.4 Ϯ 6.4
64.2 Ϯ 8.7
11.3 Ϯ 4.7
19.9 Ϯ 5.6
ϩ 10 ␮M PN
ϩ 1 ␮M PN ϩ ebselen
ϩ 10 ␮M PN ϩ ebselen
Effect of ebselen on peroxynitrite inhibition of the conversion of [14C]-PGH
-keto-PGF1␣ in bovine aortic microsomes with and without DTNB treatment. One
into
2
6
hundred micrograms of bovine aortic microsomes was treated with the indicated
FIG. 3. Chromatograms of the reaction mixtures of (1) 500 ␮M
ebselen and 500 ␮M PN at pH 7, (2) 500 ␮M ebselen and 500
␮M PN at pH 9, (3) 500 ␮M ebselen, 500 ␮M GSH, and 500
␮M PN at pH 7, and (4) 500 ␮M ebselen, 500 ␮M GSH, and
500 ␮M PN at pH 9. Incubation times were 5 min for all
reactions. Ebselen peak retention times: tR ؍ 16 min; ebselen–
GSH adduct, tR ؍ 6.6 min; and ebselenoxide, tR ؍ 4.5 min.
concentrations of PN 10 min after the addition of ebselen or ebselen plus DTNB.
Five minutes after PN treatment, 100 ␮M[14C]-PGH2 was added and further
incubated for 3 min. The reaction was stopped by acidification to pH 3.5. Prostanoids
were extracted and analyzed as described under Methods. Results are means Ϯ SEM
of seven separate experiments.
immunoprecipitation experiments with a monoclonal anti-
nitrotyrosine antibody [20, 27]. Without DTNB pretreat-
ment of aortic microsomes, 1 and 10 ␮M PN caused a
massive nitration not influenced by ebselen (Fig. 2B)
whereas the DTNB-mediated modification of microsomes
allowed ebselen to partially trap PN, leading to a dimin-
tion was carried out at pH 9, the stability of PN was high
enough to lead to an almost quantitative conversion of
ebselen to its selenoxide after equimolar mixing with PN.
Only a minor amount of ebselen remained, due to some
isomerization of PN to nitrate [3] or its dismutation to
nitrite and dioxygen [31] (Fig. 3). At pH 7, only about 50%
of the ebselen was converted to the oxide. When the same
experiment was carried out when GSH was added at a slight
molar excess over ebselen 1 min before PN, only a small
amount of ebselen oxide appeared at pH 7 and a larger
amount at pH 9 (Fig. 3). The peak of the ebselen–GSH
adduct formed under these conditions suggested that this
adduct does not undergo a rapid reaction with PN to its
corresponding selenoxide. At pH 9, a small peak with a
lower retention time than the ebselen–GSH adduct could
be consistent with such a selenoxide, confirming the better
stability of PN at pH 9, which allowed formation of this
product at pH 9 but not at pH 7. Such results point to a
slower reaction of PN with the ebselen–GSH adduct than
with ebselen itself. Direct kinetic measurements are diffi-
cult to perform since the absorbancies of PN and that of the
adduct are in the same region, i.e. between 290 and 340
nm. However, when choosing 320/370 nm as a wavelength
pair and pH 10 when the stability of PN is high enough,
one can observe kinetics of PN decomposition with ebselen
ished nitration of PGI synthase (Fig. 2A). It is interesting
2
to note that control microsomes already contained a certain
amount of nitrated enzyme, as was consistently observed in
previous experiments [13, 20, 21].
The Reaction of PN with Ebselen in the Absence and
Presence of Thiols
Ebselen and the PN anion rapidly react to give the
corresponding selenoxide and nitrite [30]. When this reac-
(
Fig. 4). At the same wavelength pair, a preincubated
equimolar mixture of ebselen and GSH decomposes PN
with slower kinetics, although the maximum absorbance
change is somewhat larger (Fig. 4A). This is at least a
qualitative confirmation of the slower reaction of the
adduct. As a control, a secondary addition of PN resulted
only in a minor change due to the absorbance of PN at 302
nm [32]. A corresponding experiment was performed with
the ebselen–albumin adduct (Fig. 4B). Additions of 20 and
FIG. 2. Western blot analysis of the peroxynitrite-mediated
nitration of prostacyclin synthase in bovine aortic microsomes.
Bovine aortic microsomes were treated with the indicated
concentrations of PN in the presence (A) or absence (B) of
4
0 ␮M albumin to 100 ␮M ebselen at pH 10 slowed down
DTNB. Nitrated PGI synthase was immunoprecipitated with a
2
the kinetics of the ebselen reaction. Again, a second
addition of PN did not alter the absorbance difference
significantly, proving that in the initial reaction the disap-
monoclonal antibody against 3-nitrotyrosine and immunoblotted
with polyclonal antibodies against 3-nitrotyrosine or against
PGI synthase. Results are representative of four experiments.
2

1
58
A. Daiber et al.
FIG. 4. Kinetics of the decomposition of peroxynitrite and disappearance of ebselen. (A) The system consisted of: (1) 400 ␮M ebselen
and 500 ␮M PN in 0.2M K-phosphate buffer at pH 10; (2) second addition of 500 ␮M PN to the reaction solution of (1); (3) 400
␮M ebselen, 200 ␮M GSH, and 500 ␮M PH; and (4) second addition of 500 ␮M PN to the reaction solution of (3). (B) The system
consisted of: (1) 100 ␮M ebselen and 100 ␮M PN in 0.2 M K-phosphate buffer at pH 10; (2) second addition of 100 ␮M PN to the reaction
solution of (1); (3) 100 ␮M ebselen, 20 ␮M BSA, and 100 ␮M PN; and (4) 100 ␮M ebselen, 40 ␮M BSA, and 100 ␮M PN.
pearance of the adduct through oxidation by PN had been
monitored. A weaker scavenging effect of the ebselen–
GSH adduct than with GSH alone could also be observed
during the nitration of 5 mM phenol with 650 ␮M PN. The
corresponding hydroxylated and nitrated phenol metabo-
lites were separated by HPLC as described [23]. In this
system, 100 ␮M ebselen caused 21 Ϯ 2% inhibition, 100
many antioxidants, as well as ebselen or even its thiol
adducts, can act as scavengers.
DISCUSSION
There are several aspects to the results presented. First, the
unexpected lack of protection by ebselen of PGI synthase
2
␮
M GSH gave 88 Ϯ 1%, and the combination of 100 ␮M
ebselen/100 ␮M GSH yielded only 83 Ϯ 3% inhibition
mean values of five measurements). With tyrosine, the
nitration and inhibition is consistent with the high reac-
tivity of ebselen with thiol compounds [28, 29]. Second,
given to in vitro preparations of aortic rings, ebselen is
present almost exclusively in the form of its adducts to
thiols. Due to its high concentration of about 5 mM in
cells, GSH is a favorite reactant, but some protein-bound
thiols/thiolates can be even more nucleophilic and there-
fore could also form adducts. Third, ebselen adducts can
scavenge PN in a slower process as may occur in methio-
nine or sulfhydryl oxidations.
(
corresponding differences were even more distinct (results
not shown).
The Effect of Ebselen on PN-mediated Inactivation and
Nitration of PGI Synthase
2
According to these data, ebselen failed to inhibit the
nitration of PGI synthase, because the cellular content of
GSH and protein thiols favored its conversion to adducts
which reacted too slowly to scavenge PN for the obviously
Our results can be summarized by Scheme 1, which also
includes the known complex chemistry that ebselen can
perform under physiological conditions. The numerous
reactions and equilibria that take place in this system are
also the reason why we could not determine a kinetic
constant for the reaction of PN with ebselen–GSH adduct.
There are several species absorbing in the same wavelength
region and thus complicating such kinetic measurements,
but at least one can qualitatively see a much slower reaction
compared with ebselen itself (Fig. 4, A and B). As shown in
Scheme 1, GSH also reacts with PN (6), as already
described [33, 36, 37]. Consequently, it is also impossible to
distinguish between kinetic effects which originate from
ebselen, GSH, or ebselen–GSH adduct. We could not
determine whether the reactivity towards PN in an ebselen-
and GSH-containing system results from the small amount
of “free” ebselen (1) which is formed by a slow equilibrium
from the adduct (3) (if so it must be very slow, or else one
would not get a peak for the adduct in HPLC), or if it comes
directly from the adduct. The latter would mean that a
2
rapid reaction of PN with PGI synthase. This may be
2
different for the free radical pathway of PN oxidations and
nitrations, which are thought to involve a protonated
excited state of PN [3, 23]. A reaction of this species with
GSH, ascorbate, uric acid, or other antioxidants has been
reported [16–18, 33]. Using phenol as a model substrate,
one can test this one-electron oxidation pathway by fol-
lowing the formation of nitrophenols and phenolic dimer-
ization products [23, 34, 35]. In this system at pH 7, the
absorbance increase due to the formation of 2- and 4-ni-
trophenol could be effectively blocked by ebselen, even if
its concentration was only 50% of that of PN. An equimo-
lar mixture of ebselen and GSH, which would lead to
adduct formation, was also quenching although to a some-
what lower extent (not shown). Thus, if PN performs
nitrations or oxidations by a mechanism either different or
slower than the heme-catalyzed nitration of PGI synthase,
2

Ebselen as a Peroxynitrite Scavenger
159
SCHEME 1. Chemistry of the system ebselen, glutathione and peroxynitrite.
reaction from the adduct (3) to the ebselenoxide (2) is
possible, but then the question arises as to what happens to
the ebselen-bound GSH. Ebselenoxide (2) can be con-
verted back to ebselen (1) by two molecules of GSH [38].
Another possibility is the conversion of the adduct (3) to
the corresponding selenol (4) by one molecule of GSH
followed by the reaction with another molecule of ebselen
to give the diselenid (5) [29]. Of course (4) and (5) may
also react with PN [19].
In summary, the mechanism of ebselen inactivation
towards PN is complex, but this scheme fits the observa-
tions made in this study: ebselen binds to thiols and forms
isolatable adducts. These systems show a decreased reactiv-
ity towards PN anion at least comparable to that of other
tested selenocompounds such as Se–methionine [39].
Therefore, the loss of reactivity of ebselen towards per-
oxynitrite in biological systems seems to be due to this thiol
binding. Hence, in the absence of any thiols as shown after
preincubation with Ellman’s reagent, ebselen can strongly
This work was supported by the Deutsche Forschungsgemeinschaft,
Forschergruppe “Endogene Gewebszerst o¨ rung—Mechanismen der Au-
todestruktion”, and the Graduiertenkolleg “Biochemische Pharmakolo-
gie”.
References
1
. Moncada S, Palmer RMJ and Higgs EA, Nitric oxide: Phys-
iology, pathophysiology and pharmacology. Pharmacol Rev 43:
109–142, 1991.
2
3
4
. Nathan C, Nitric oxide as a secretory product of mammalial
cells. FASEB J 6: 3051–3064, 1992.
. Beckman JS, Oxidative damage and tyrosine nitration from
peroxynitrite. Chem Res Toxicol 9: 836–844, 1996.
. Kissner R, Nauser T, Bugnon P, Lye PG and Koppenol WH,
Formation and properties of peroxynitrite as studied by laser
flash photolysis, high-pressure stopped-flow technique, and
pulse radiolysis. Chem Res Toxicol 10: 1285–1292, 1997.
. Epe B, Ballmaier D, Roussyn I, Briviba K and Sies H, DNA
damage by peroxynitrite characterized with DNA repair
enzymes. Nucleic Acids Res 24: 4105–4110, 1996.
5
compete with microsomal PGI synthase for PN. Although
6. Heinecke JW, Mechanisms of oxidative damage of low
density lipoprotein in human atherosclerosis. Curr Opin
Lipidol 8: 268–274, 1997.
2
the ebselen adducts reacted too slowly to protect PGI2
synthase, they did interact with PN and therefore at least
could function as antioxidants in cells. Therefore, the free
radical nitration of phenol could be inhibited. In the
7
. Ischiropoulos H and Al-Mehdi AB, Peroxynitrite-mediated
oxidative protein modifications. FEBS Lett 364: 279–282,
1
995.
context of PN nitration of PGI synthase, the results lend
8. Packer L (Ed.), Nitric Oxide: Part B Physiological and
2
pathological processes. Methods Enzymol 269: 1–375, 1996.
support to our mechanistic proposal that PN reacts, as does
its anion, directly at the ferric heme and that the nitrated
tyrosine must be at the active site, as judged from the
protection against nitration by quasi-substrates blocking
the substrate binding site [20]. Thus, for kinetic and steric
reasons, none of the known antioxidants was able to
interfere with this process. In the case of an initially
assumed inhibitory action of ebselen, it might have been
9
. Beckman JS and Koppenol WH, Nitric oxide, superoxide and
peroxynitrite: The good, the bad and the ugly. Am J Physiol
271: C1424–C1437, 1996.
1
0. Crow JP, Beckman JS and McCord JM, Sensitivity of the
essential zinc-thiolate moiety of yeast alcohol dehydrogenase
to hypochlorite and peroxynitrite. Biochemistry 34: 3544–
3
552, 1995.
1
1. Vogt W, Oxidation of methionyl residues in proteins: Tools,
targets, and reversal. Free Radic Biol Med 18: 93–105, 1995.
possible to decide whether the nitration of PGI synthase,
2
12. Crow JP, Ye YZ, Strong M, Kirk M, Barnes S and Beckman
JS, Superoxide dismutase catalyzes nitration of tyrosine by
peroxynitrite in the rod and head domains of neurofila-
ment-L. J Neurochem 69: 1945–1953, 1997.
which now has been identified to occur during atheroscle-
rosis [40] and endotoxin exposure of aortic rings [41], is only
a pathological and damaging process or whether the simul-
1
3. Zou MH and Ullrich V, Peroxynitrite formed by simultaneous
generation of nitric oxide and superoxide selectively inhibits
bovine aortic prostacyclin synthase. FEBS Lett 382: 101–104,
1996.
taneous trapping of NO and PGI by superoxide generation
2
is a physiological reaction required for host defense and
cellular activation.

1
60
A. Daiber et al.
1
4. MacMillan-Crow LA, Crow JP, Kerby JD, Beckman JS and
Thompson A, Nitration and inactivation of manganese su-
peroxide dismutase in chronic rejection of human renal
allografts. Proc Natl Acad Sci USA 93: 11853–11858, 1996.
5. Wang P and Zweier JL, Measurement of nitric oxide and
peroxynitrite generation in the postischemic heart. J Biol
Chem 271: 29223–29230, 1996.
6. Christen AA, Woodall MK, Southwell-Keely PT, Duncan W
and Ames BN, Reactions of peroxynitrite with alpha-toco-
pherol. Proc Natl Acad Sci USA 94: 3218–3223, 1997.
7. Bartlett D, Church DF, Bounds PL and Koppenol WH, The
kinetics of the oxidation of L-ascorbic acid by peroxynitrite.
Free Radic Biol Med 18: 85–92, 1995.
28. Ullrich V, Weber P, Meisch F and von Appen F, Ebselen-
binding equilibria between plasma and target proteins. Bio-
chem Pharmacol 52: 15–19, 1996.
29. Haenen G, De Roorj B, Vermeulen N and Bast A, Mecha-
nism of the reaction of ebselen with endogenous thiols. Mol
Pharmacol 37: 412–422, 1989.
1
1
1
1
30. Masumoto H and Sies H, The reaction with ebselen with
peroxynitrite. Chem Res Toxicol 9: 262–267, 1996.
3
3
3
1. Pfeiffer S, Gorren ACF, Schmidt K, Werner ER, Hansert B,
Bohle DS and Mayer B, Metabolic fate of peroxynitrite in
aqueous solution. J Biol Chem 272: 3465–3470, 1997.
2. Hughes MN and Niklin HG, The chemistry of pernitrides. I.
Kinetics of decomposition of pernitrous acid. J Am Soc (Part
A): 450–462, 1968.
3. Scorza G and Minetti M, One-electron oxidation pathway of
thiols by peroxynitrite in biological fluids: Bicarbonate and
ascorbate promote the formation of albumin disulfide dimers
in human blood plasma. Biochem J 329: 405–413, 1998.
4. Ramezanian MS, Padmaja S and Koppenol WH, Nitration
and hydroxylation of phenolic compounds by peroxynitrite.
Chem Res Toxicol 9: 232–240, 1996.
8. Squadrito GL, Jin X and Pryor WA, Stopped-flow kinetic
study of the reaction of ascorbic acid with peroxynitrite. Arch
Biochem Biophys 322: 53–59, 1995.
1
2
9. Arteel GE, Briviba K and Sies H, Protection against peroxyni-
trite. FEBS Lett 445: 226–230, 1999.
0. Zou MH, Martin C and Ullrich V, Tyrosine nitration as a
mechanism of selective inactivation of prostacyclin synthase
by peroxynitrite. Biol Chem 378: 707–713, 1997.
3
3
2
1. Zou MH, Klein T, Pasquet JP and Ullrich V, Interleukin
5. Lemercier JN, Padmaja S, Cueto R, Squadrito GL, Uppu RM
and Pryor WA, Carbon dioxide modulation of hydroxylation
and nitration of phenol by peroxynitrite. Arch Biochem Bio-
phys 345: 160–170, 1997.
1
beta decreases prostacyclin synthase activity in rat mesangial
cells via endogenous peroxynitrite formation. Biochem J 336:
5
07–512, 1998.
2
2
2
2
2. Mehl M, Daiber A, Herold S, Shoun H and Ullrich V,
3
3
3
3
6. Radi R, Beckman JS, Bush KM and Freeman BA, Peroxyni-
trite oxidations of sulfhydryls. J Biol Chem 266: 4244–4250,
Peroxynitrite reaction with heme proteins. Nitric Oxide 3:
1
42–152, 1999.
1
991.
3. Daiber A, Mehl M and Ullrich V, New aspects in the reaction
mechanism of phenol with peroxynitrite: The role of phenoxy
radicals. Nitric Oxide 2: 259–269, 1998.
4. Masumoto H, Kissner R, Koppenol WH and Sies H, Kinetic
study of the reaction of ebselen with peroxynitrite. FEBS Lett
7. Gow AJ, Buerk DG and Ischiropoulos H, A novel reaction
mechanism for the formation of S-nitrosothiol in vivo. J Biol
Chem 272: 2841–2845, 1997.
8. Masumoto H and Sies H, The reaction of 2-(methylseleno)-
benzanilide with peroxynitrite. Chem Res Toxicol 9: 1057–
398: 179–182, 1996.
1
062, 1996.
5. Hara S, Miyata A, Yokoyama C, Inoue H, Brugger R,
Lottspeich F, Ullrich V and Tanabe T, Isolation and molec-
ular cloning of prostacyclin synthase from bovine endothelial
cells. J Biol Chem 31: 19897–19903, 1994.
9. Briviba K, Roussyn I, Sharov VS and Sies H, Attenuation of
oxidation and nitration reactions of peroxynitrite by sel-
enomethionine, selenocysteine and ebselen. Biochem J 319:
1
3–15, 1996.
2
6. Bunting S, Gryglewski R, Moncada S and Vane JR, Arterial
walls generate from prostaglandin endoperoxides a substance
4
0. Zou MH, Leist M and Ullrich V, Selective nitration of
prostacyclin synthase and defective vasorelaxation in athero-
sclerotic bovine coronary arteries. Am J Pathol 154: 1354–
1365, 1999.
(
prostaglandin X) which relaxes strips of mesenteric and
coeliac arteries and inhibits platelet aggregation. Prostaglan-
dins 12: 897–913, 1976.
2
7. Zou MH, Jendral M and Ullrich V, Peroxynitrite induces
prostaglandin endoperoxide dependent-vasospasm in bovine
coronary arteries via nitration of prostacyclin synthase. Br J
Pharmacol 126: 1283–1292, 1999.
41. Szabo C, Salzman AL and Ischiropoulos H, Endotoxin triggers
the expression of an inducible isoform of nitric oxide synthase
and the formation of peroxynitrite in the rat aortae in vivo.
FEBS Lett 363: 235–238, 1995.