SUPEROXIDE DISMUTASE AND CATALASE ARE REQUIRED TO DETECTSNO FROM
BOTH COUPLED AND UNCOUPLED NEURONAL NO SYNTHASE
A. REIF,*,1Z. V. SHUTENKO,y,1M. FEELISCH,zand H.H.H.W. SCHMIDT§
*Department of Psychiatry and Psychotherapy, Julius-Maximilians-University Wqrzburg, D-97078 Wqrzburg, Germany,yDepartment of
Pharmacology and Toxicology, Julius-Maximilians-University, D-97078 Wqrzburg, Germany,zDepartment of Molecular and Cellular Physiology,
Louisiana State University Health Sciences Center, Shreveport, LA 71130, USA, and the§Rudolf-Buchheim-Institute for Pharmacology,
Justus-Liebig-University Giegen, D-35392 Giegen, Germany
(Received 23 January 2004; Revised 20 May 2004; Accepted 1 July 2004)
Available online 22 July 2004
Abstract—Despite numerous approaches to measuring nitric oxide (SNO) formation from purified NO synthase (NOS),
it is still not clear whetherSNO is a direct or indirect product of the NO synthase reaction. The direct detection of
catalytically formedSNO is complicated by side reactions with reactive oxide species like H2O2and superoxide. The aim
of the present study was therefore to reinvestigate these reactions both electrochemically and by chemiluminescence
detection with particular emphasis on the requirement for cofactors and their interference withSNO detection. Flavins
were found to generate large amounts of H2O2and were therefore excluded from subsequent incubations. Under
conditions of both coupled and uncoupled catalysis, SOD was absolutely required to detectSNO from NOS. H2O2
formation took place also in the presence of SOD and gave a smaller yet significant interfering signal. Similar data were
obtained when the proposed intermediate NN-hydroxy-l-arginine was utilized as substrate. In conclusion, standard Clark-
typeSNO electrodes are cross-sensitive to H2O2and therefore both SOD and catalase are absolutely required to
specifically detectSNO from NOS.
D 2004 Elsevier Inc. All rights reserved.
Keywords—Nitric oxide, NOS, Superoxide, Catalase, SOD, Flavins, NO electrode, Free radicals
Nitric oxide synthases (NOS) catalyze the oxidation of
a terminal guanidino nitrogen of l-arginine . All
three known NOS isoforms represent multifunctional
oxidoreductases that require tetrahydrobiopterin
(H4Bip), Ca2+/calmodulin, FAD, and FMN as cofactors
or stabilizing molecules and l-arginine, molecular
oxygen, and NADPH as substrates [2,3]. For activity
measurements, the by-product of the NOS reaction, l-
citrulline, is commonly determined utilizing cation-
exchange chromatography. In contrast, to dateSNO
formation has mostly been measured indirectly (e.g., via
the oxyhemoglobin assay or the Griess reaction ), as
the direct measurement of NOS-derivedSNO is difficult
and the correct conditions for its detection remain
controversial [5–7]. There is considerable evidence that
in vivo NOS is not saturated with l-arginine and
H4Bip. Under these conditions, the electron flow within
the enzyme is uncoupled and neuronal NOS (NOS-I)
also catalyzes the generation of hydrogen peroxide
(H2O2) . Furthermore, especially l-arginine levels are
thought to regulate superoxide production from NOS-I
[9–13]. Autoxidation of reagent H4Bip  was
suggested as an alternative mechanism of O2S?
formation. Superoxide, of either origin, could rapidly
inactivateSNO to peroxynitrite and hence complicate
directSNO measurements, especially when the short
half-life ofSNO (b5 s) is taken into account.
In this context, intracellular H4Bip seems to play an
ambivalent role in NOS catalysis; on the one hand, it is
apparently required as an activator and stabilizer of NOS;
at the same time, it was suggested to be a source for
O2S?, which immediately scavenges NOS-derivedSNO
Address correspondence to: A. Reif, Molecular and Clinical Psycho-
biology, Department of Psychiatry and Psychotherapy, Julius-Max-
imilians-University Wqrzburg, Fqchsleinstrasse 15, D-97080 Wqrzburg,
Germany; Fax: +49 931 201 77550; E-mail: firstname.lastname@example.org.
1These authors contributed equally to this work.
Free Radical Biology & Medicine, Vol. 37, No. 7, pp. 988–997, 2004
Copyright D 2004 Elsevier Inc.
Printed in the USA. All rights reserved
0891-5849/$-see front matter
at diffusion-limited rates . However, other groups did
not detect relevant amounts of O2S?from H4Bip by
electron paramagnetic resonance (EPR) [11,15] and no
autoxidation products from H4Bip could be found in
HPLC measurements, but it seemed to scavenge reactive
nitrogen species . Therefore, rather than H4Bip
autoxidation , flavin-derived reactive oxygen species
(ROS) may prevent the detection of NOS-derivedSNO in
Methods to directly quantifySNO formation from
purified NOS include electrochemical detection using a
Clark-type electrode with reversed polarity [14,17],
chemiluminometric detection (CLD) , and EPR
spectroscopy by scavengingSNO with spin trap com-
pounds such as N-methyl-d-glucamine dithiocarbonate
(MGD) and subsequent formation of a paramagnetic
complex, NO–Fe–MGD . Previous studies [7,14,17]
demonstrated thatSNO could not be detected directly
unless high concentrations of superoxide dismutase
(SOD; 5 kU ml?1) were also present. It therefore was
suggested thatSNO may not be the primary NOS
product. One possible candidate, or at least model
compound, for an alternative product would be nitroxyl
(HNO). SOD catalyzes the conversion of HNO toSNO
[19,20] and this reaction could well account for the SOD
dependence of NOS-derivedSNO signals . Con-
versely, in a different study  using Fe–MGD EPR
spectroscopy, a NOS-derived signal could be observed
in the absence of both SOD and flavins (FAD and
FMN). However, recent evidence suggested that this
method is not selective forSNO and therefore not
applicable to discriminate betweenSNO and related N-
oxides, especially nitroxyl [18,21], although this was
The present study was therefore conducted to further
elaborate the experimental conditions for the electro-
chemical detection ofSNO during uncoupled as well as
coupled NOS-I catalysis and to investigate the possible
interference of H4Bip or NOS-derived ROS withSNO
NN-Hydroxy-l-arginine (OH-Arg) was obtained from
Calbiochem (La Jolla, CA, USA); CPTIO and spermine
NONOate were from Cayman Chemical Company (Ann
Arbor, MI, USA); FAD, GSH, SOD, and catalase were
from Boehringer Mannheim (Mannheim, Germany);
FMN was from Fluka (Buchs, Switzerland); H4Bip was
from Dr. Schirks Laboratories (Jona, Switzerland). All
other reagents were of the highest analytical grade
available and from either Sigma (Deisenhofen, Germany)
or Merck (Darmstadt, Germany).
Expression and purification of human recombinant
NOS-I and purification of native porcine cerebellum
Human recombinant NOS-I was expressed in a
baculovirus/Sf9 cell system and purified to apparent
homogeneity by subsequent 2V ,5V -ADP–Sepharose and
calmodulin affinity chromatography . The purity of
the preparation was determined densitometrically from
Coomassie-stained SDS/PAGE gels. The protein concen-
tration was determined according to the method of
Bradford using bovine serum albumin as a standard
. Two enzyme preparations with similar specific
activities (400 nmol mg?1min?1) and purity (z95%)
were used. For comparison, native porcine cerebellum
NOS-I, purified as published , was used.
The electrochemical detection was performed as
previously described . The reversed polarity Clark-
type electrode detectsSNO, but is also sensitive for
H2O2in a ratio of 1:5 . In order to quantify both
SNO and H2O2, the electrode was calibrated with
authenticSNO and H2O2. The detection limits for
SNO and H2O2were 10 and 50 nM, respectively, based
on a signal-to-noise ratio of 3:1. All experiments were
carried out at room temperature. Each electrode reading
given in the figures is representative of three to six
Nonenzymatic electrode signals
To determine whether the observed electrode signals
were of nonenzymatic origin, i.e., derived from flavins or
other assay components, some experiments were per-
formed by incubating FAD or FMN in a total volume of
300 Al TrisSHCl (50 mM, pH 7.4) containing 1 mM
NADPH. The reaction was started by addition of 5 AM
FAD or 5 AM FMN. In some of the experiments, 10 AM
H4Bip, 5 kU ml?1SOD, and 5.2 kU ml?1catalase were
also included in the incubation mixture.
Measurements of NOS-derived electrochemical sig-
nals (0.14 AM NOS) were performed in a total volume
of 300 Al TrisSHCl (50 mM, pH 7.4) containing 1 mM
NADPH, 1 mM l-arginine or OH-Arg, and 0.05 or
1 AM CaM for uncoupled and coupled conditions,
respectively. The enzymatic reaction was started by
addition of 1 mM CaCl2. Where indicated, 10 AM H4Bip,
5 kU ml?1SOD, up to 3.9 kU ml?1catalase, and 100 or
300 AM CPTIO were also included in the incubation
mixture. In all experiments with NOS-I, 15 Ag of enzyme
NOS-derived NO signals require both SOD989
Chemiluminometric NO detection
A microreaction chamber containing 50 mM Mops
buffer (pH 7.0), 5 AM FAD, 10 AM FMN, 1 mM
NADPH, 50 nM CaM, 7 mM GSH, 1 mM CaCl2, and
500 AM either l-arginine or OH-Arg was connected to a
CLD 780 TR chemiluminescence measurement device
(eco Physics, Dqrnten, Switzerland). The reaction mix-
ture, which also contained 5 kU ml?1SOD for some
experiments, was equilibrated with NOx-free air passed
over the continuously stirred sample. The reaction was
started by the addition of NOS (10 Ag) and carried out at
room temperature in a total volume of 1.0 ml. Calibra-
tion, measurements, and data analysis were performed as
described . The detection limit for NO was 5 pmol
min?1based on a signal-to-noise ratio of 3:1. Neither
H2O2nor O2S?could be detected with this method, as
verified by authentic H2O2or a superoxide-generating
system (xanthine oxidase/hypoxanthine; not shown).
Results are expressed as means ± SEM of at least three
SNO reacts with both O2S?and H2O2leading to a
considerable quenching of theSNO electrode signal.
Both ROS are supposed to be formed during uncoupled,
enzymatic l-arginine turnover but also nonenzymatically
from assay components. In order to optimize the assay
conditions for the electrochemical NO detection, we
investigated possible nonenzymatic sources of O2S?and
Both FAD and FMN generate electrode signals due to
The presence of FAD (5 AM) in Trisd HCl buffer (pH
7.4) resulted in a strong NADPH-dependent electro-
chemical signal (Fig. 1a) not affected by up to 5 kU ml?1
SOD (Fig. 1c). However, the presence of catalase (5.2 kU
ml?1) completely abolished this signal (Fig. 1e), suggest-
ing that the source of this signal was H2O2. Because only
uncharged molecules, and not O2S?, can be detected by
the electrode used, SOD should have increased the signal
if concomitant O2S?formation took place, as it catalyzes
the dismutation of O2S?to H2O2(compare Fig. 1a with
1c). Therefore concomitant superoxide formation, as
suggested by a recent study , seems to be minor
compared to H2O2generation. Similar data were obtained
with FMN (5 AM) (Fig. 1b). Again, the electrochemical
signal was not altered upon addition of SOD (Fig. 1d), but
was abolished by catalase (5.2 kU ml?1) (Fig. 1f). These
data suggest that predominantly H2O2is formed from
flavins. However, the simultaneous formation of small
amounts of O2S?remains a possibility .
No electrochemical signals can be obtained from H4Bip
It has been postulated that H4Bip produces O2S?upon
autoxidation . In contrast, we did not detect any
electrochemical signal from 10 AM H4Bip, neither in the
absence nor in the presence of 1 mM NADPH or SOD (5
kU ml?1; data not shown). In agreement with this, a
recent study utilizing EPR technology  revealed that
Fig. 1. Reagent flavins generate H2O2 during their nonenzymatic
reduction by 1 mM NADPH. The reaction was started (E) by addition
of 5 AM FAD (a, c, e) or FMN (b, d, f); thereafter, electrochemical
measurements were performed as described under Experimental
Procedures. In the absence of SOD or catalase (a, b), a prominent
signal was recorded that was not affected by the presence of 5 kU ml?1
SOD (c, d). The simultaneous application of 5 kU ml?1SOD and 5.2
kU ml?1catalase (e, f) totally abolished the signals, indicating
predominant H2O2formation from both flavins during their reduction
by NADPH. Original recordings are representative of three independent
experiments yielding similar results.
A. Reif et al.
H4Bip (10 AM), in aqueous oxygenated solution, did not
generate O2S?, which was confirmed by another study
examining the influence of cofactors on superoxide
production from NOS .
Thus, the only quantitatively relevant and nonenzy-
matic source of ROS under the investigated incubation
conditions were flavins, which in the presence of
NADPH formed H2O2, but no or only trace amounts of
O2S?. As NOS is a flavoprotein, containing tightly
bound endogenous FAD and FMN, it is only moderately
or not at all dependent on exogenous flavins. This could
be confirmed by measurement of the enzymatic con-
Fig. 2. NOS-I generates simultaneously anSNO precursor and hydrogen peroxide. As described under Experimental Procedures, 15 Ag
NOS-I was incubated in the presence of 1 mM l-arginine, 1 mM NADPH, and 0.05 (solid lines) or 1 AM (dashed lines) CaM and (a, c, e,
g) in the absence and (b, d, f, h, i) presence of 10 AM H4Bip. Reactions were started by addition of 1 mM CaCl2(E). In the absence of
both SOD and catalase, an electrochemical signal could be observed which was of similar height in the absence (a) and in the presence (b)
of H4Bip. It could be totally abolished when 3.9 kU ml?1catalase was present from the start on (c, d), indicating that the nature of this
signal was H2O2. In the presence of 5 kU ml?1SOD, but in the absence of catalase (e), a prominent signal was observed, which could be
enlarged by addition of 10 AM H4Bip (f, solid line) and even more so when CaM-levels were saturating (f, dashed line). NOS under these
conditions most likelyproducedSNO,generated by SODfrom a yet unidentifiedSNO precursormolecule,andH2O2: when3.9 kUml?1
catalase was added (. in e and f), the signal was reduced, but this reduction was partly reversible. When both catalase and SOD were
presentfromthe start of the reactionon(g, h), the observedsignalsweremoreprominent(i), most likelyduetoinitial H2O2inactivationby
catalase. TheSNO scavenger CPTIO (o, 100 AM; 5, 300 AM in f and h) abolished the signals dose-dependently and at higher
concentrations completely. Original recordings are representative of nine independent experiments yielding similar results. (i) The bar
graph represents means F SEM (n = 9); the asterisk indicates a significant difference between both values (0.038 F 0.005 and 0.06 F
0.004 mV, respectively; p b .05; paired two-tailed t test).
NOS-derived NO signals require both SOD991
version of l-arginine to l-citrulline (not shown). We
therefore excluded flavins from the incubation mixture to
minimize nonenzymatic H2O2formation and its interfer-
ence with enzymatic N-oxides.
Another source of ROS is uncoupled NOS catalysis
. There are two modes to induce uncoupling of NOS:
submaximal saturation of either the l-arginine/H4Bip or
the calmodulin binding sites. As maximal levels of both
l-arginine and H4Bip are necessary to stabilize the NOS
dimer during catalysis , uncoupling was induced in
the present study by incubating NOS in the presence of
reduced amounts of CaM. Under these conditions,
electron flow to the oxygenase domain is reduced, which
is thought to result in uncoupled superoxide or H2O2
formation from the flavins within the reductase domain.
Uncoupled NOS catalysis then was compared to the
coupled reaction, which was measured with enzyme fully
saturated with calmodulin.
The specific detection of NOS-derived NO requires both
SOD and catalase
Using these modified assay conditions, i.e., in the
absence of both FAD and FMN, we examined NOS-
derived electrode signals both in the absence (Fig. 2a)
and in the presence (Fig. 2b) of 10 AM H4Bip during l-
arginine turnover. In both cases, the observed signal
height was similar and the signals had similar kinetics.
Moreover, NOS-derived electrode signals in the presence
of H4Bip but absence of SOD were similar for both
coupled and uncoupled conditions of catalysis, i.e.,
saturating and subsaturating CaM concentrations (data
not shown). In contrast, no electrode signal could be
observed when catalase (3.9 kU ml?1) was present in the
incubation mixture from the beginning of the reaction,
independent of the presence of H4Bip (Figs. 2c and 2d).
This remained unchanged by increasing CaM concen-
tration (up to 1 AM) in the presence of H4Bip. Thus, any
electrode signal observed in the absence of SOD was
entirely due to H2O2.
When SOD (5 kU/ml) was present from the beginning
of experiment, an electrode signal could be detected
during l-arginine turnover in the absence of both H4Bip
and catalase (Fig. 2e), which had kinetics different from
that in the absence of SOD. The addition of exogenous
H4Bip now markedly increased the signal height (Fig. 2f,
solid line) and even more so when CaM levels were
saturating, i.e., under conditions of coupled NOS
catalysis (Fig. 2f, dashed line).
To differentiate the source of this signal (SNO, H2O2,
or both), catalase was applied additionally 15 min after
the start of the reaction (indicated by the filled circle in
Figs. 2e and 2f). Almost immediately the signal was
partially reduced to a value significantly above basal (n =
6; paired two-tailed t test; Figs. 2e and 2f; solid lines).
Again, the signal had similar characteristics, but was
more prominent when CaM levels were maximal (Fig.
2f, dashed line).
When catalase was present from the beginning of the
experiment together with SOD, but in the absence of
H4Bip, the electrode signal from uncoupled NOS was
lower than in the respective controls (Fig. 2g; compare
with 2a and 2e). Again, when exogenous H4Bip was
included in the reaction mixture containing a subsaturat-
ing calmodulin concentration (0.05 AM CaM vs. 0.14
AM NOS), the electrode signal was increased to a small
extent (Fig. 2h, solid line). Under fully coupled
conditions (1 AM CaM), the electrode signals were more
prominent but had similar kinetics (Fig. 2h, dashed line).
The time point at which catalase was added seemed to be
crucial: when catalase was added after 15 min of
incubation, the height of the signal was about half of
the signal measured upon the initial presence of catalase
(Fig. 2i; compare 2h with 2f).
To examine whether the NOS-derived electrode signal
was indeedSNO, CPTIO, an irreversible and specific
SNO scavenger, was applied at concentrations of 100 and
300 AM. Independent of whether catalase was present
from the beginning of experiment or added later, the
addition of CPTIO completely abolished the signal (Figs.
2f and 2h). Moreover, when 100 AM CPTIO was present
together with SOD and catalase from the beginning of
the reaction, no signal was observed (data not shown).
Therefore, signals observed in the presence of both SOD
and catalase were considered to beSNO.
Importantly, throughout all experiments the detection
of authenticSNO was absolutely dependent on both SOD
and catalase: in the absence of SOD, the signals were
entirely due to H2O2, in its presence, bothSNO and H2O2
were detected simultaneously (cf. Table 2).
Table 2. NOS-Derived Signals and Their Characteristics
Signal height (mV)
at t = 15 min
SOD + catalase
0.035 F 0.002a
0.06 F 0.004
0.038 F 0.005a
SNO precursor + H2O2
SNO precursor + H2O
NOS (15 Ag) was incubated in the absence of flavins, in the presence
of 0.05 AM CaM (i.e., partially uncoupled conditions), as well as 10 AM
H4Bip, and otherwise under the conditions described under Exper-
imental Procedures. When neither catalase nor SOD was present, NOS
produced H2O2 as well as an undetectable N-oxide. SOD was
absolutely required to detect specificSNO signals, whereas catalase
is important to eliminate H2O2and thus protectsSNO from reacting
with this ROS.SNO precursor = currently not identifiedSNO precursor
molecule; n.d. = no signal detectable.
aPlease note that in the absence of any additive, the signal is
increasing constantly, whereas in the presence of SOD and catalase, it
reaches a steady-state level at 15 min (cf. Fig. 2b with 2h).
A. Reif et al.
Catalase is a reversibleSNO scavenger
In the previous set of experiments we noted that
catalase slightly, albeit reversibly, lowered the SOD-
dependent NOS-derivedSNO signal (Figs. 2f and 2h).
SNO donor compound spermine NONOate (0.167 AM)
were indeed attenuated by the addition of catalase (up to
10.4 kU ml?1). However, similar to NOS-derived signals,
this interaction was only transient and fully reversible
within a few minutes. This effect could also be observed
when catalase was added repetitively (Fig. 3a). Likewise,
when SOD and catalase were present from the beginning
of the experiment to ensure that onlySNO and not H2O2
was recorded, repetitive addition of catalase (up to 3.9 kU
ml?1) had a transient effect on NOS-derived electrode
signals (Fig. 3b). This indicated a reversible reaction of
catalase with both NOS/SOD-derived and reagentSNO.
OH-Arg as a substrate behaves similar to l-arginine
In addition to l-arginine, the intermediate or by-
product of NOS catalysis, OH-Arg, can be converted to
l-citrulline by NOS. We therefore studied OH-Arg as an
alternative substrate in equimolar concentrations as l-
arginine. Under all investigated conditions, the observed
signals had the same characteristics as those with l-
arginine. Interestingly, the addition of H4Bip to the
reaction mixture increased the signal by a factor of 2.2;
when the specific pterin antagonist PHS-32  was
applied, this could be reversed (data not shown). Hence
H4Bip seems to be important also for the second step of
NOS catalysis, despite the fact that the generation ofSNO
from OH-Arg is not specific for NOS [27,28]. However,
when catalase was added, either at the beginning of the
experiment or later, the signal was more prominent than
the corresponding signals utilizing l-arginine as a
substrate (compare Figs. 4a, 4b, and 4c to Figs. 2f, 2h,
and 2i). Again, catalase had a reversibleSNO-scavenging
Fig. 3. Catalase reversibly scavengesSNO.SNO was generated either
(a) with spermine NONOate (0.167 AM) or (b) with NOS (in the
presence of 1 mM l-arginine, 1 mM NADPH, 0.05 (solid line) or 1 AM
(dashed line) CaM, 10 AM H4Bip, SOD 5 kU/ml) and thereafter
determined electrochemically as described above.SNO derived from
spermine NONOate could be reversibly scavenged with 10.4 kU
catalase ml?1; similarly, NOS/SOD-derivedSNO could be scavenged
with 3.9 kU ml?1catalase under both coupled (dashed line) and
uncoupled (solid line) conditions. o indicates the addition of spermine
NONOate (a), E the start of the reaction by addition of 1 mM CaCl2
(b), and. addition of catalase. Original recordings are representative of
three to six independent experiments yielding similar results.
Fig. 4. Catalase has a reversible scavenging effect on NOS-derived
signals with OH-Arg as substrate. Signals were registered electro-
chemically in the presence of 15 Ag NOS-I, 1 mM OH-Arg, 10 AM
H4Bip, 5 kU ml?1SOD, and otherwise as described. 3.9 kU ml?1
catalase was either (a) added after 15 min of incubation or (b) present
from the onset of the incubation. Again, catalase reversibly attenuated
the signal; notably, (c) when catalase was included from the start, the
subsequently observed signal was higher than when catalase was added
at t = 15 min. E indicates the start of the reaction by addition of CaCl2,
. the addition of catalase. Original recordings are representative of
three to six independent experiments yielding similar results. The bar
graph (c) represents means F SEM (n = 6); the asterisk indicates a
significant difference between both values (0.06 F 0.006 and 0.109 F
0.011 mV, respectively; p b .05, unpaired two-tailed t test).
NOS-derived NO signals require both SOD993
effect, as additional amounts of catalase (up to 3.9 kU
ml?1) transiently attenuated the electrode signal (Fig. 3b).
ChemoluminometricSNO detection from both l-arginine
and OH-Arg also requires the presence of SOD
As the chemiluminometric detection ofSNO is
insensitive to both H2O2 and O2S?, we applied this
method to monitor NOS-derivedSNO formation from
either l-arginine or OH-Arg. For either substrate, the
presence of SOD (5 kU ml?1) was absolutely required to
gain significantSNO signals (Table 1). In agreement with
previously published data , this suggested that the
primary NOS product has similar characteristics inde-
pendent of whether l-arginine or OH-Arg is used as
substrate. In contrast to electrochemicalSNO measure-
ments, the NOS/SOD-derived CLD signals were less
pronounced when OH-Arg was the substrate (Table 1),
possibly due toSNO/OH-Arg adduct formation which
may escape the gas-phase detection for chemilumines-
cence but not the electrochemical measurements .
Again, the CLD signals from both substrates could be
increased about 2.5-fold upon addition of 10 AM H4Bip
and blocked by either arginine- or pterin-based inhibitors
Only few reports claim to provide direct evidence for
the formation ofSNO as the main or sole product of l-
arginine turnover by NOS, under both coupled and—
presumably more physiological—uncoupled conditions.
However, none of theSNO assays, including spin
trapping, is direct. The lack of detectability of NOS-
derivedSNO is frequently attributed to the presence of
ROS under the commonly used in vitro assay conditions.
Moreover, ROS may derive from uncoupled NOS
In this study, we demonstrate that the main non-
enzymatic sources of ROS in the NOS incubation
mixture are NADPH-dependently reduced flavins. Fla-
vin-derived signals originated most likely from H2O2, as
they were completely abolished by catalase, which is in
accordance with previously published data . It is
unlikely that O2S?was the prevailing product of flavin
reduction, as SOD, which converts O2S?to H2O2, had no
effect on the signal height.
H4Bip-derived O2S?has been suggested to be another
significant source of ROS, thus reducingSNO recovery
. Surprisingly, no electrode signals could be detected
from H4Bip in the absence or in the presence of NADPH
and/or SOD. Consistent with EPR spin trap data [5,11],
our data exclude an interference of H4Bip with the
electrochemical signal. Based on these data, we con-
sequently excluded exogenous FAD and FMN, but not
H4Bip, from all NOS incubations. Using this exper-
imental protocol, all measured signals were considered to
be of enzymatic origin, i.e., NOS-dependent.
SNO rapidly reacts with ROS such as O2S?, with
intermediates in the Fenton or peroxidase mechanisms,
and with iron-containing proteins. All of these reactions
are relatively fast and lead to the formation of more or less
stable intermediates, possibly causing subsequent oxida-
tive cellular and protein damage [31–33]. Autoxidation of
H4Bip was proposed to generate O2S?, which, by
instantly reacting withSNO to peroxynitrite [32,35],
should account for the fast loss ofSNO from NOS
incubation mixtures . However, under NOS assay
conditions we found no H4Bip autoxidation, neither
electrochemically in the present study nor by HPLC
measurements . Thus, O2S?formation from H4Bip in
significant amounts seems unlikely and therefore cannot
account for the absolute SOD dependence of NOS-
derivedSNO signals. Rather H4Bip may react directly
withSNO . A recent study demonstrated by EPR
measurements that NOS itself does not produce signifi-
cant amounts of superoxide  (although there has been
considerable controversy about this topic afterward
[22,36–38]). This rules out superoxide production from
fully coupled NOS, as used in the present study.
Furthermore, no nonenzymatic ROS source was present.
Why cannot NOS-derivedSNO then be measured—both
electrochemically and by CLD—under these circum-
stances, as it should be ifSNO is indeed the primary
reaction product of NOS?
In a recent study using EPR spectroscopy  a
prominent EPR signal of NO–Fe–MGD was detected in
the absence of SOD during NOS-I catalysis. This
suggested that NOS-I, unlike our and others suggestions,
is able to convert at least some l-arginine directly to
SNO, avoiding any intermediate nitrogen species like
HNO/NO?. Importantly, Angeli’s salt was used as an
HNO donor and gave no NO–Fe–MGD EPR signal. This
is somehow surprising, as it implies a unique specificity
of this assay forSNO over HNO, considering that it has
Table 1. SOD Is Required to DetectSNO Formation
from both l-Arginine and OH-Arg
SNO formation (nmol mg?1min?1)
ControlSOD (5 kU ml?1)
SNO from NOS (10 Ag) was determined chemiluminometrically as
described under Experimental Procedures in the absence or presence of
SOD (5 kU ml?1). For both substrates (500 AM), the presence of SOD
was absolutely required to detectSNO.
aNot significantly different from baseline.
bSignificantly different from baseline.
0.61 F 0.30a
0.64 F 0.25a
7.81 F 0.22b
4.12 F 0.61b
A. Reif et al.
otherwise been shown for this iron complex to give an
EPR signal with the NO+donor sodium nitroprusside and
even to oxidize hydroxyurea to give an NO–Fe–MGD
signal . Furthermore, in these experiments no
positive control was performed to show that Angeli’s
salt, a hygroscopic compound that is highly susceptible
to hydrolyzing to NO2
control experiments with SOD were reported. We have
recently reinvestigated the specificity of this assay and, in
contrast to Zweier et al., show that Fe–MGD is fully
sensitive to intact Angeli’s salt and thus not suitable for
discriminating betweenSNO and HNO or related N-
oxides . The appropriate control experiments, how-
ever, were performed in a later report of the Zweier
group, so that the question whether the spin trap used is
selective forSNO still remains a matter of debate [18,22].
Furthermore, because HNO only very slowly deproto-
nates, it is especially prone to interfering with other
protonation–deprotonation processes. Specific experi-
mental conditions, e.g., differing pH values and buffers
as well as varying Angeli’s salt preparations of different
purity and stability, should be taken into account as
potential sources of error, probably explaining conflict-
ing results in previous studies.
As there is little evidence for directSNO formation by
NOS in the absence of SOD, it was hypothesized that
NOS catalyzes the formation of a nitrogen oxide distinct
fromSNO  that can be converted toSNO by SOD. In
one study, using chemiluminescence detection , some
SNO was detected from NOS-I and NOS-II during l-
arginine turnover. However, SOD (10 kU ml?1)
enhanced the generation ofSNO without affecting the
recovery of reagentSNO. Therefore, a second role of
SOD distinct from O2S?dismutation was suggested.
Under aerobic conditions, the capability of SOD to
enhance the conversion of nitrogen oxide species, such
as the model compound nitroxyl (HNO), toSNO is one
possible explanation [7,19,20,40,41]; recently, evidence
was provided that nitroxyl serves as a substrate for SOD
even under anaerobic conditions .
We have previously shown  that, even in the
absence of flavins, direct electrochemical and CLD
detection ofSNO from native porcine NOS-I is not
possible unless high concentrations of SOD (10 kU
ml?1) are present. In the present study, we could measure
a significant NOS-dependent signal in the absence of l-
arginine and SOD that was completely abolished by
catalase (Table 2). In the presence of SOD, the electrode
signals were diminished when catalase was added to the
incubation mixture 15 min after the start of the reaction
regardless of the other incubation conditions, i.e., with
and without H4Bip, with saturating and subsaturating
CaM levels. Thus, recombinant human NOS-I evidently
produced H2O2under both uncoupled and, surprisingly,
?, was still intact. Additionally, no
coupled conditions. This is in line with findings from a
recent study employing both H4Bip-free and H4Bip-
saturated NOS-I, which argued that H4Bip plays a pivotal
role in shifting superoxide production (in pterin-free
enzyme) to hydrogen peroxide formation (when pterin is
bound to NOS) .
Also when catalase was added from the start of the
reaction on and concomitant H2O2formation was there-
fore abolished, SOD was absolutely required to measure
any signals from NOS under the investigated conditions
(Table 2). As we excluded nonenzymatic sources of ROS
from the incubation mixture, it seems likely that NOS
primarily produces a yet to be identified N-oxide distinct
fromSNO (Table 2). Alternatively, it recently has been
proposed that NOS in fact primarily produces nitro-
soarginine, which then degrades to citrulline and nitrite
, which, however, has not been replicated to date by
other groups. Importantly, the latter study was conducted
in the absence of both SOD and catalase, but in the
presence of exogenously added flavins. Thus, it seems
likely that (1) both superoxide and H2O2were formed,
considering the results of the present study, and that (2)
this in turn leads to the formation of nitrosoarginine.
AsSNO is the only N-oxide able to stimulate the
soluble guanylyl cyclase , it is intriguing to speculate
that NOS might produce different molecules regulated by
coexpressed enzymes:SNO, which depends on the
presence of both SOD and catalase, as a intercellular
messenger, e.g., in neurons, and a yet to be identified
SNO precursor molecule, serving as a toxic agent, e.g., in
stimulated macrophages, as it is known that nitroxyl is
able to damage biological micro-  and macro-
molecules by oxidation [31,46]. Also the potentially
harmful production of H2O2by NOS deserves further
investigation; matters are even more complicated when
the ambiguous role ofSNO in cell protection, especially
from H2O2-caused damage, is taken into account [47,48].
Whether such a bswitchingQ of principal NOS products
occurs physiologically remains to be established and is
surely worth in-depth research.
High concentrations of catalase reversibly bindSNO
and, in the presence of H2O2, causeSNO degradation
[26,49]. Furthermore,SNO binding inhibits catalase
activity, presumably by competing with H2O2 .
Consistent with this, high concentrations of catalase
reversibly scavengedSNO, generated by theSNO-donor
spermine NONOate. The same was observed for NOS/
SOD-derivedSNO. Thus catalase seems to play a dual
role in the NO pathway: it prevents H2O2formation and
scavengesSNO. A recent study proposed that the NOS
oxygenase domain itself displays catalase activity .
From our data this can be neither rejected nor proven; if
such a catalase activity of NOS indeed occurs, it seems to
play only a minor role in the NOS reaction mechanism as
NOS-derived NO signals require both SOD995
exogenous catalase was necessary to fully suppress H2O2
signals under all conditions examined. Consistent with
this, in the aforementioned study a nonphysiological ratio
of NOS/H2O2was necessary to achieve a catalase-like
activity of NOS. Taken together, a specific catalase
activity of NOS seems unlikely and the observed effects
might be ascribed to the general ability of heme proteins
to catalyze redox reactions.
In conclusion, all studies investigating the direct
chemical detection of a NOS-derived N-oxide [5,17,52]
suggest at least the simultaneous appearance of HNO,
NO?, or related precursors as well asSNO. SOD, in this
case, facilitates the transformation of the intermediates to
SNO. Catalase is required to protect already formedSNO
from reacting with both NOS-derived and exogenous
H2O2. Both enzymes are therefore absolutely necessary to
specifically quantify theSNO release from NO synthase.
Acknowledgments—We thank M. Bernhard for excellent technical
assistance. This study was supported by the Deutsche Forschungsge-
meinschaft (SFB 547/C7).
 Nathan, C.; Xie, Q.-W. Nitric oxide synthases: roles, tolls, and
controls. Cell 78:915–918; 1994.
 Abu-Soud, H. M.; Yoho, L. L.; Stuehr, D. J. Calmodulin controls
neuronal nitric-oxide synthase by a dual mechanism: activation of
intra- and interdomain electron transfer. J. Biol. Chem.
 Marletta, M. A. Nitric oxide synthase structure and mechanism. J.
Biol. Chem. 268:12231–12234; 1993.
 Schmidt, H. H. H. W. Determination of nitric oxide via
measurement of nitrite and nitrate in culture media. Biochem.
Inf. 96:20–21; 1995.
 Xia, Y.; Zweier, J. Direct measurement of nitric oxide gene-
ration from nitric oxide synthase. Proc. Natl. Acad. Sci. USA 94:
 Schmidt, H. H. H. W.; Pollock, J. S.; Nakane, M.; Gorsky, L. D.;
Ffrstermann, U.; Murad, F. Purification of a soluble isoform of
guanylyl cyclase-activating-factor synthase. Proc. Natl. Acad. Sci.
USA 88:365–369; 1991.
 Hobbs, A. J.; Fukuto, J. M.; Ignarro, L. J. Formation of free nitric
oxide from l-arginine by nitric oxide synthase: direct enhance-
ment of generation by superoxide dismutase. Proc. Natl. Acad.
Sci. USA 91:10992–10996; 1994.
 Kotsonis, P.; Frey, A.; Frfhlich, L.; Hofmann, H.; Reif, A.; Wink,
M.; Feelisch, M.; Schmidt, H. H. H. W. Autoinhibition of neuronal
NO synthase: distinct effects of reactive nitrogen and oxygen
species on enzyme activity. Biochem. J. 340:745–752; 1999.
 Pou, S.; Pou, W. S.; Bredt, D. S.; Snyder, S. S.; Rosen, G. M.
Generation of superoxide by purified brain nitric oxide synthase.
J. Biol. Chem. 267:24173–24176; 1992.
 Pou, S.; Keaton, L.; Surichamorn, W.; Rosen, G. M. Mechanism
of superoxide generation by neuronal nitric-oxide synthase. J.
Biol. Chem. 274:9573–9580; 1999.
 Vasquez-Vivar, J.; Kalyanaraman, B.; Martasek, P.; Hogg, N.;
Masters, B. S. S.; Karoui, H.; Tordo, P.; Pritchard, K. A.
Superoxide generation by endothelial nitric oxide synthase: the
influence of cofactors. Proc. Natl. Acad. Sci. USA 95:9220–9225;
 Rosen, G. M.; Tsai, P.; Weaver, J.; Porasuphatana, S.; Roman, L. J.;
Starkov, A. A.; Fiskum, G.; Pou, S. The role of tetrahydrobiopterin
in the regulation of neuronal nitric-oxide synthase-generated
superoxide. J. Biol. Chem. 277:40275–40280; 2002.
 Song, Y.; Cardounel, A. J.; Zweier, J. L.; Xia, Y. Inhibition of
superoxide generation from neuronal nitric oxide synthase by heat
shock protein 90: implications in NOS regulation. Biochemistry
 Mayer, B.; Klatt, P.; Werner, E. R.; Schmidt, K. Kinetics and
mechanism of tetrahydrobiopterin-induced oxidation of nitric
oxide. J. Biol. Chem. 270:655–659; 1995.
 Xia, Y.; Zweier, J. Superoxide and peroxynitrite generation from
inducible nitric oxide synthase in macrophages. Proc. Natl. Acad.
Sci. USA 94:6954–6958; 1997.
 Reif, A.; Frfhlich, L. G.; Kotsonis, P.; Frey, A.; Bfmmel, H. M.;
Wink, D. A.; Pfleiderer, W.; Schmidt, H. H. H. W. Tetrahydro-
inneuronalNOsynthase. J. Biol. Chem. 274:24921–24929; 1999.
Cunningham, D. D.; Feelisch, M. NoSDNO from NO synthase.
Proc. Natl. Acad. Sci. USA 93:14492–14497; 1996.
 Komarov, A. M.; Reif, A.; Schmidt, H. H. In vitro detection of
nitric oxide and nitroxyl by electron paramagnetic resonance.
Methods Enzymol. 359:18–27; 2002.
 Murphy, M. E.; Sies, H. Reversible conversion of nitroxyl anion to
nitric oxide by superoxide dismutase. Proc. Natl. Acad. Sci. USA
 Bahr, N.; Gqller, R.; Reymond, J. L.; Lerner, A. L. A nitroxyl
synthase catalytic antibody. J. Am. Chem. Soc. 118:3550–3555;
 Komarov, A. M.; Wink, D. A.; Feelisch, M.; Schmidt, H. H. H. W.
Electron-paramagnetic resonance spectroscopy using N-methyl-d-
glucosamine dithiocarbamate cannot discriminate between nitric
oxide and nitroxyl: implications for the detection of reaction
products from NO synthase. Free Radic. Biol. Med. 28:739–742;
 Xia, Y.; Cardounel, A. J.; Vanin, A. F.; Zweier, J. L. Electron
paramagnetic resonance spectroscopy with N-methyl-d-glucamine
dithiocarbamate iron complexes distinguishes nitric oxide and
nitroxyl anion in a redox-dependent manner: applications in
identifying nitrogen monoxide products from nitric oxide syn-
thase. Free Radic. Biol. Med. 29:793–797; 2000.
 Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of
protein-dye binding. Anal. Biochem. 72:248–254; 1976.
 Nagase, S.; Ohkoshi, N.; Ueda, A.; Aoyagi, K.; Koyama, A.
Hydrogen peroxide interferes with detection of nitric oxide by an
electrochemical method. Clin. Chem. 43:1246; 1997.
 Bfmmel, H. M.; Reif, A.; Frfhlich, L. G.; Frey, A.; Hofmann, H.;
Marecak, D. M.; Groehn, V.; Kotsonis, P.; La, M.; Kfster, S.;
Meinecke, M.; Bernhardt, M.; Weeger, M.; Ghisla, S.; Prestwich,
G. D.; Pfleiderer, W.; Schmidt, H. H. H. W. Anti-pterins as tools to
characterize the function of tetrahydrobiopterin in NO synthase. J.
Biol. Chem. 273:33142–33149; 1998.
 Brown, G. C. Reversible binding and inhibition of catalase by
nitric oxide. Eur. J. Biochem. 232:188–191; 1995.
 Boucher, J. L.; Genet, A.; Vadon, S.; Delaforge, M.; Mansuy, D.
Formation of nitrogen oxides and citrulline upon oxidation of NN-
hydroxy-l-arginine by hemeproteins. Biochem. Biophys. Res.
Commun. 184:1158–1164; 1992.
 Boucher, J. L.; Genet, A.; Vadon, S.; Delaforge, M.; Henry, Y.;
Mansuy, D. Cytochrome P450 catalyzes the oxidation of NN-
hydroxy-l-arginine by NADPH and O2 to nitric oxide and
citrulline. Biochem. Biophys. Res. Commun. 187:880–886; 1992.
 Zembowicz, A.; Hecker, M.; Macarthur, H.; Sessa, W. C.; Vane,
J. R. Nitric oxide and another potent vasodilator are formed from
A. Reif et al.
NN-hydroxy-l-arginine by cultured endothelial cells. Proc. Natl. Download full-text
Acad. Sci. USA 88:11172–11176; 1991.
 Massey, V. Activation of molecular oxygen by flavins and
flavoproteins. J. Biol. Chem. 269:22459–22462; 1994.
 Wink, D. A.; Kasprzak, K. S.; Maragos, C. M.; Elespuru, R. K.;
Misra, M.; Dunams, T. M.; Cebula, T. A.; Koch, W. H.; Allen,
J. S.; Keefer, L. K. DNA deamination ability and genotoxicity
of nitric oxide and its progenitors. Science 254:1001–1003;
 Beckman, J. S.; Beckman, T. W.; Chen, J.; Marshall, P. A.;
Freeman, B. A. Apparent hydroxyl radical production by
peroxynitrite: implications for endothelial injury from nitric oxide
and superoxide. Proc. Natl. Acad. Sci. USA 87:1620–1624; 1990.
 Radi, R.; Beckman, J. S.; Bush, K. M.; Freeman, B. A.
Peroxynitrite-induced membrane lipid peroxidation: the cytotoxic
potential of superoxide and nitric oxide. Arch. Biochem. Biophys.
 Schmidt, K.; Klatt, P.; Mayer, B. Reaction of peroxynitrite with
oxyhaemoglobin: interference with photometrical determination of
nitric oxide. Biochem. J. 301:645–647; 1994.
 Koppenol, W. H.; Moreno, J. J.; Pryor, W. A.; Ischiropoulos, H.;
Beckman, J. S. Peroxynitrite, a cloaked oxidant formed by nitric
oxide and superoxide. Chem. Res. Toxicol. 5:834–842; 1992.
 Xu, K. Y. Does nitric oxide synthase catalyze the synthesis of
superoxide?. FEBS Lett. 23619:1–2; 2000.
 Vasquez-Vivar, J.; Kalyanaraman, B. Generation of superoxide
from nitric oxide synthase. FEBS Lett. 481:305–306; 2000.
 Xu, K. Y. A key negative control experiment provides evidence
that nitric oxide synthase does not catalyze superoxide formation.
FEBS Lett. 481:306–307; 2000.
 Tsuchiya, K.; Corbett, J.; Mason, R. P. Reaction mechanisms of
water-soluble nitric oxide spin-trapping agent, MGD. Free Radic.
Biol. Chem. 25 (Suppl. 1):191; 1998.
 Moncada, S.; Palmer, R. M. J.; Higgs, A. E. Nitric oxide:
physiology, pathophysiology, and pharmacology. Pharmacol. Rev.
 Liochev, S. I.; Fridovich, I. Copper, zinc superoxide dismutase as
a univalent NO(?) oxidoreductase and as a dichlorofluorescein
peroxidase. J. Biol. Chem. 276:35253–35257; 2001.
 Liochev, S. I.; Fridovich, I. Nitroxyl (NO?): a substrate for
superoxide dismutase. Arch. Biochem. Biophys. 402:166–171;
 Meulemans, A. A brain nitric oxide synthase study in the rat:
production of a nitroso-compound NA and absence of nitric oxide
synthesis. Neurosci. Lett. 321:115–119; 2002.
 Dierks, E. A.; Burstyn, J. N. Nitric oxide (NO), the only nitrogen
monoxide redox form capable of activating soluble guanylyl
cyclase. Biochem. Pharmacol. 51:1593–1600; 1996.
 Reif,A.;Zecca, L.;Riederer, P.;Feelisch,M.;Schmidt,H. H.H.W.
Nitroxyl oxidizes NADPH in a superoxide dismutase inhibitable
manner. Free Radic. Biol. Med. 30:803–808; 2000.
 Wink, D.; Feelisch, M.; Fukuto, J.; Chistodoulou, D.; Jourd’heuil
D.; Grisham, M.; Vodovotz, Y.; Cook, J.;Krishna, M.; DeGraff,
W.; Kim, S.; Gamson, J.; Mitchell, J. The cytotoxic mechanism of
nitroxyl: possible implications for the pathophysiological role of
NO. Arch. Biochem. Biophys. 351:66–74; 1998.
 Wink, D. A.; Cook, J. A.; Pacelli, R.; DeGraff, W.; Gamson, J.;
Liebmann, J.; Krishna, M. C.; Mitchell, J. B. The effect of various
nitric oxide-donor agents on hydrogen peroxide-mediated toxicity:
a direct correlation between nitric oxide formation and protection.
Arch. Biochem. Biophys. 331:241–248; 1996.
 Schmidt, H. H. H. W.; Warner, T. D.; Murad, F. Double-edged role
of endogenous nitric oxide. Lancet 339:986; 1992.
 Deisseroth, A.; Dounce, A. L. Nature of the change produced by
catalase by lyophilization. Arch. Biochem. Biophys. 120:671–692;
 Malinski, T.; Taha, Z. Nitric oxide release from a single cell
measured in situ by a porphyrinic-based microsensor. Nature
 Adhikari, S.; Ray, S.; Gachhui, R. Catalase activity of oxygenase
domain of rat neuronal nitric oxide synthase: evidence for product
formation from l-arginine. FEBS Lett. 23619:1–4; 2000.
 Noronha-Dutra, A. A.; Epperlein, M. M.; Woolf, N. Reaction of
nitric oxide with hydrogen peroxide to produce potentially
cytotoxic singlet oxygen as a model for nitric oxide-mediated
killing. FEBS Lett. 321:59–62.
CPTIO—carboxy-PTIO (1H-imidazol-1-yloxy, 2-(4-
EPR—electron paramagnetic resonance
PAGE—polyacrylamide gel electrophoresis
ROS—reactive oxygen species
SDS—sodium dodecyl sulfate
NOS-derived NO signals require both SOD997