S-Nitrosoglutathione covalently modifies cysteine residues of human carbonyl reductase 1 and affects its activity

Article abstract of DOI:10.1016/j.cbi.2012.12.011

Carbonyl reductase 1 (CBR1 or SDR21C1) is a ubiquitously-expressed, cytosolic, monomeric, and NADPH-dependent enzyme. CBR1 participates in apoptosis, carcinogenesis and drug resistance, and has a protective role in oxidative stress, cancer and neurodegeneration. S-Nitrosoglutathione (GSNO) represents the newest addition to its diverse substrate spectrum, which includes a wide range of xenobiotics and endogenous substances. GSNO has also been shown to covalently modify and inhibit CBR1. The aim of the present study was to quantify and characterize the resulting modifications. Of five candidate cysteines for modification by 2 mM GSNO (positions 26, 122, 150, 226, 227), the last four were analyzed using MALDI-TOF/TOF mass spectrometry and then quantified using the Selected Reaction Monitoring Approach on hyphenated HPLC with a triple quadrupole mass spectrometer. The analysis confirmed GSNO concentration-dependent S-glutathionylation of cysteines at positions 122, 150, 226, 227 which was 2-700 times higher compared to wild-type CBR1 (WT-CBR1). Moreover, a disulfide bond between neighboring Cys-226 and Cys-227 was detected. We suggest a role of these two cysteines as a redox-sensitive cysteine pair. The catalytic properties of wild-type and enzyme modified with 2 mM GSNO were also investigated by steady state kinetic experiments with various substrates. GSNO treatment of CBR1 resulted in a 2-5-fold decrease in k_cat with menadione, 4-benzoylpyridine, 2,3-hexanedione, daunorubicin and 1,4-naphthoquinone. In contrast, the same treatment increased k_cat for substrates containing a 1,2-diketo group in a ring structure (1,2-naphthoquinone, 9,10-phenanthrenequinone, isatin). Except for 9,10-phenanthrenequinone, all changes in k_cat were at least in part compensated for by a similar change in K_m, overall yielding no drastic changes in catalytic efficiency. The findings indicate that GSNO-induced covalent modification of cysteine residues affects the kinetic mechanism of CBR1 both in terms of substrate binding and turnover rate, probably by covalent modification of Cys-226 and/or Cys-227.

Full text of DOI:10.1016/j.cbi.2012.12.011

Chemico-Biological Interactions 202 (2013) 136–145
Contents lists available at SciVerse ScienceDirect
Chemico-Biological Interactions
journal homepage: www.elsevier.com/locate/chembioint
S-Nitrosoglutathione covalently modifies cysteine residues of human
carbonyl reductase 1 and affects its activity
a
b
c
Tereza Hartmanová , Vojtech Tambor , Juraj Lenco , Claudia A. Staab-Weijnitz d,1, Edmund Maser d,
ˇ
ˇ
Vladimír Wsól a,
⇑
a Department of Biochemical Sciences, Faculty of Pharmacy, Charles University, Heyrovského 1203, 50005 Hradec Králové, Czech Republic
b Biomedical Research Center, University Hospital Hradec Králové, Sokolská 581, 500 05 Hradec Králové, Czech Republic
c Institute of Molecular Pathology, Faculty of Military Health Sciences, University of Defence, Hradec Králové, Czech Republic
d Institute of Toxicology and Pharmacology for Natural Scientists, University Medical School Schleswig-Holstein, Campus Kiel, Brunswiker Str. 10, 24105 Kiel, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Available online 5 January 2013
Carbonyl reductase 1 (CBR1 or SDR21C1) is a ubiquitously-expressed, cytosolic, monomeric, and NADPH-
dependent enzyme. CBR1 participates in apoptosis, carcinogenesis and drug resistance, and has a protec-
tive role in oxidative stress, cancer and neurodegeneration. S-Nitrosoglutathione (GSNO) represents the
newest addition to its diverse substrate spectrum, which includes a wide range of xenobiotics and endog-
enous substances. GSNO has also been shown to covalently modify and inhibit CBR1. The aim of the pres-
ent study was to quantify and characterize the resulting modifications. Of five candidate cysteines for
modification by 2 mM GSNO (positions 26, 122, 150, 226, 227), the last four were analyzed using
MALDI-TOF/TOF mass spectrometry and then quantified using the Selected Reaction Monitoring
Approach on hyphenated HPLC with a triple quadrupole mass spectrometer. The analysis confirmed
GSNO concentration-dependent S-glutathionylation of cysteines at positions 122, 150, 226, 227 which
was 2–700 times higher compared to wild-type CBR1 (WT-CBR1). Moreover, a disulfide bond between
neighboring Cys-226 and Cys-227 was detected. We suggest a role of these two cysteines as a redox-sen-
sitive cysteine pair. The catalytic properties of wild-type and enzyme modified with 2 mM GSNO were
also investigated by steady state kinetic experiments with various substrates. GSNO treatment of CBR1
resulted in a 2–5-fold decrease in kcat with menadione, 4-benzoylpyridine, 2,3-hexanedione, daunorubi-
cin and 1,4-naphthoquinone. In contrast, the same treatment increased kcat for substrates containing a
1,2-diketo group in a ring structure (1,2-naphthoquinone, 9,10-phenanthrenequinone, isatin). Except
for 9,10-phenanthrenequinone, all changes in kcat were at least in part compensated for by a similar
change in Km, overall yielding no drastic changes in catalytic efficiency. The findings indicate that
GSNO-induced covalent modification of cysteine residues affects the kinetic mechanism of CBR1 both
in terms of substrate binding and turnover rate, probably by covalent modification of Cys-226 and/or
Cys-227.
Keywords:
Carbonyl reductase
SDR21C1
GSNO
S-glutathionylation
Quinone
Ó 2013 Elsevier Ireland Ltd. All rights reserved.
1. Introduction
diverse substrates, mostly carbonyls. The best known xenobiotic
substrates include quinones (such as the vitamin K precursor men-
Human CBR1 (carbonyl reductase 1, EC 1.1.1.184), or according
to the new nomenclature SDR21C1 [1], is a ubiquitously-ex-
pressed, monomeric, cytosolic enzyme belonging to the short-
chain dehydrogenase/reductase (SDR) superfamily [2]. CBR1 cata-
lyzes the NADPH-dependent reduction of a variety of structurally
adione and 9,10-phenanthrenequinone, both often used as model
substrates), anthracyclines, ketoaldehydes, aromatic aldehydes,
and NNK (4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone), the
carcinogenic nitrosamine of tobacco smoke [3,4]. CBR1 also re-
duces a range of endogenous substances including prostaglandins,
Abbreviations: CBR, carbonyl reductase; SDR, short-chain dehydrogenase/reductase; NNK, 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone; GSNO, S-nitrosoglutathione;
GSNOR, GSNO reductase; GSH, glutathione; DTT, dithiothreitol; IPTG, isopropyl b-D-1-thiogalactopyranoside; IAM, iodoacetamide; DMSO, dimethyl sulfoxide; ACN,
acetonitrile; TFA, trifluoroacetic acid; MALDI-TOF/TOF, matrix assisted laser desorption ionization - time of flight/time of flight; MS, mass spectrometry; LC-SRM, liquid
chromatography-selected reaction monitoring.
⇑
Corresponding author. Tel.: +420 739488218.
E-mail addresses: wsol@faf.cuni.cz, vladimir.wsol@faf.cuni.cz (V. Wsól).
Present address: Institute for Clinical Radiology, Ludwig Maximilians University, 80336 Munich, Germany.
1
0009-2797/$ - see front matter Ó 2013 Elsevier Ireland Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.cbi.2012.12.011

T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
137
steroids and other aliphatic aldehydes and ketones along with the
endogenous indole isatin and S-nitrosothiol GSNO (S-nitrosogluta-
thione) [5,6]. Moreover, CBR1 plays a protective role in oxidative
stress, tumor metastasis, neurodegeneration and apoptosis [7,8].
The underlying molecular mechanisms are poorly understood.
One study has shown that CBR1 inactivates the lipid aldehyde 4-
oxo-2-nonenal [9], a lipid peroxidation product formed during oxi-
dative stress. Hence CBR1 might protect from oxidative stress by
eliminating reactive oxygen species [10].
Before the discovery of GSNO as CBR1 substrate in 2008 [6],
CBR1 had only been known to convert carbonyl groups to alcohols.
GSNO reduction thus represents a new mechanism, where an NO
bond is reduced. The kinetic parameters of CBR1 and GSNO are
comparable to those of model CBR1 substrates as isatin or menadi-
one [6]. CBR1 seems to be specific for GSNO, because S-nitrosocys-
teine is not reduced by CBR1. Moreover, CBR1-dependent GSNO
reduction has been described in A549 lung adenocarcinoma cell ly-
sates. This indicates that CBR1 also acts as GSNO reductase in vivo.
GSNO is a key endogenous S-nitrosothiol, which serves as a res-
ervoir and donor of NO in organisms [11–13]. In humans, GSNO is
physiologically present up to micromolar levels [11], plays a role in
apoptosis [14], has a neuroprotective role [15], inhibits platelet
activation [16], and has a strong bronchodilatation effect in asthma
[17]. Until 2008, only GSNOR (GSNO-reductase, also termed alco-
hol dehydrogenase 3) was known to reduce GSNO, resulting in
no NO release but in NO signaling termination [18].
The first indication that CBR1 can contain a glutathione binding
site had been suggested already by Wermuth in 1981. He had
found that the glutathione adduct of prostaglandin PGA1 is re-
duced by CBR1 while free PGA is not [2]. Another fact supporting
this hypothesis was that CBR1 from human placenta was inhibited
by oxidized GSH adducts [20]. Later, the same research group sug-
gested that a cysteine residue could play an important role in glu-
tathione binding and enzyme activity [21]. In 2008, the X-ray
crystal structure of the GSH-binding site was found in close struc-
tural proximity to the active site of CBR1 [7].
Recent studies have revealed that GSNO causes covalent modi-
fication of CBR1, which results in loss of enzyme activity at a con-
centration around 100 lM GSNO [19]. The fact that treatment with
dithiothreitol (DTT) restored the enzyme activity, while incubation
with ascorbic acid did not, indicated that S-glutathionylation was
the mechanism responsible for the enzyme inhibition. Further
indirect evidence pointed towards Cys-227 as the glutathionylated
residue and, in agreement with a previous study that has provided
strong support for Cys-227 as the reactive residue [20], Cys-227
was hypothesized to be subject to glutathionylation [19].
In the present study, we sought to identify the cysteine residues
in CBR1 that are modified by GSNO and to characterize and quan-
tify the modifications in a mass spectrometry approach. CBR1 con-
tains 5 cysteine residues in its sequence (positions 26, 122, 150,
226, 227) (Fig. 1), of which mutation of two (Cys-226, Cys-227)
is known to affect activities for substrates like menadione, 4-ben-
zoylpyridine and daunorubicin [20].
To detect these modifications, we used MALDI-TOF/TOF mass
spectrometry followed by a selected reaction monitoring (SRM) ap-
proach on hyphenated liquid chromatography with triple quadru-
pole mass spectrometer. Next, we investigated the influence of
these modifications upon other substrates of CBR1 by spectropho-
tometric and HPLC kinetic studies.
2. Materials and methods
2.1. Cloning, protein overexpression and GSNO preparation
Cloning, overexpression and purification of CBR1 were done as
described in [19] with the exceptions that PCR primers, including
restriction sites, were synthesized by Generi Biotech (Hradec Kra-
love, Czech Republic), and the primer sequences were as follows:
forward primer 50-GGA TTC CAT ATG TCG TCC GGC ATC CA-30, re-
verse primer 50-CGC CTC GAG TCA CCA CTG TTC AAC TC-30. Next
NdeI and XhoI restriction enzymes and commercially available
pET-28b(+) vector (Novagen, Darmstadt, Germany) were used.
The correct sequence insertion was verified by sequencing at Gen-
eri Biotech. GSNO was synthesized according to Hart [19,21] and
the purity was checked as described by Staab et al. [19] before each
experiment.
2.2. Initial MALDI-TOF/TOF analysis of CBR1
2.2.1. CBR1 treatment with GSNO and trypsin digestion
CBR1 (10 lg) in 0.1 M potassium phosphate buffer (pH 7.4) and
50% glycerol was incubated either in water (control CBR1), in 2 mM
GSNO, or in 200 mM GSNO at room temperature for 2 h. Thiol
groups were blocked by addition of 60 mM iodoacetamide (Sig-
ma–Aldrich, St. Louis, MO) immediately after incubation with
GSNO. Blocking was performed at room temperature for 60 min
in darkness. Low molecular compounds were removed and buffer
was exchanged using 0.5 ml Zeba Spin Desalting Columns (7K
MWCO; Thermo Scientific, Milford, MA). CBR1 in 120 ll of
50 mM ammonium bicarbonate was digested using sequencing
grade trypsin (Promega, Madison, WI, USA). After overnight diges-
tion in a thermomixer at 37 °C, the resulting peptides were de-
salted on 4 mm/1 ml Empore C18-SD solid phase extraction (SPE)
cartridges (Sigma–Aldrich).
2.2.2. MALDI-TOF/TOF analysis and spectra evaluation
Peptides eluted from SPE columns using 80% ACN, 0.1% TFA
were dried in vacuo and subsequently redissolved in 25
ACN, 0.1% TFA. Five microliters from each sample were then mixed
with 5 mg/ml of -cyano 4-hydroxycinnamic acid solution (Laser-
ll of 5%
a
Bio Labs, Sophia-Antipolis Cedex, France) in 50% ACN, 0.1% TFA in a
1:1 ratio (v/v) and spotted onto a MALDI sample plate (AB Sciex,
Foster City, CA). The mass spectra and tandem mass spectra were
recorded on an ABI 4800 MALDI-TOF/TOF mass analyzer (AB Sciex)
and evaluated in a Peak Explorer (AB Sciex) and using mMass soft-
ware [22]. Tryptic peptides resulting from CBR1 treated with
200 mM GSNO were additionally resolved on a nano-HPLC system
Fig. 1. The crystal structure of CBR1 (1WMA) created in PyMOL showing all
cysteines in the sequence. Only Cys-227 points towards the enzyme catalytic
center.

138
T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
and spotted in time-resolved fractions onto a MALDI sample plate
using a Probot fraction collector (Dionex, Sunnyvale, CA) as de-
scribed earlier [23]. MS/MS spectra were evaluated using the MAS-
COT search engine (Matrix Science, London, UK).
tion. The remaining two qualifier transitions were used to confirm
the identity of the compound of interest.
2.3.3. LC-SRM analysis
The LC-SRM analyses were performed on an Agilent 1260 LC
system coupled to an Agilent 6490 Triple Quadrupole mass spec-
trometer (both from Agilent, Palo Alto, CA), controlled by Mas-
sHunter acquisition software (version B.04.01) (Agilent).
Dry peptides were reconstituted in 5% ACN, 0.1% formic acid
and injected onto a reversed phase column (Poroshell 120 SB-
2.3. Quantitative LC-SRM analysis of CBR1 modifications
2.3.1. CBR1 treatment with GSNO and trypsin digestion
CBR1 (100
l
g) in 0.1 M potassium phosphate buffer (pH 7.4)
M, 100 lM
and 50% glycerol was incubated in triplicate with 10
l
or 2 mM GSNO and H2O2 as well as with water (control CBR1).
The incubation conditions and sample preparation were the same
as described in Section 2.2.1. After buffer exchange, the protein
concentration was assessed using the Bradford method. Ten micro-
grams from each sample was digested with sequencing grade tryp-
sin overnight in a thermomixer at 37 °C. Subsequently, the samples
were acidified using TFA and heated for 5 min at 95 °C to inactivate
trypsin, dried in vacuo, and stored at ꢀ80 °C until analysis.
C18, 100 ꢁ 2.1 mm I.D., 2.7
was maintained at 40 °C. Peptides were separated by a linear gra-
dient from 5% to 40% ACN, 0.1% formic acid over 5 min at a 300 l/
lm core–shell particles; Agilent) which
l
min flow rate and electrosprayed using a JetStream ion source into
the mass spectrometer. The acquisition method used the following
parameters: drying gas flow of 15 l/min at 200 °C, nebulizing gas
flow at 30 psi, sheath gas flow of 11 l/min at 250 °C, 4000 V capil-
lary voltage, 300 V nozzle voltage, 380 V fragmentor voltage, 4 V
cell accelerator voltage, and MS operating pressure of 5 ꢁ 10ꢀ5
-
Torr. All transitions were acquired in positive ion mode with a
dwell time of 10 ms and with both Q1 and Q3 set to unit resolution
(0.7 FWHM).
2.3.2. LC-SRM assay development
The SRM assays were developed, managed and evaluated using
Skyline software (MacCoss Lab, University of Washington, Seattle,
WA) [24]. The recombinant CBR1 sequence (Fig. 2) was imported
into the application and digested with trypsin in silico to obtain
individual tryptic peptides. The S-glutathionylation and carbami-
domethylation modifications as well as a dehydro-modification
on cysteinyl residues 226 and 227 (disulfide bond between neigh-
boring cysteine residues) were set as variable cysteine modifica-
For assay development experiments, approximately 1 lg of the
digest was injected onto the column, whereas for the final quanti-
fication experiments, 300 ng from each sample was analyzed. All
analyses for the final quantification experiments were performed
in three technical replicates. All acquired data were imported into
the Skyline application and evaluated. All integrated peaks were
inspected to confirm proper automatic peak selection and accurate
integration.
tions. In addition to the cysteinyl peptides,
9 non-cysteinyl
peptides were also selected for purposes of sample processing
and assay quality control.
For all peptides, doubly- and triply-charged precursors and sin-
gly-charged y and b fragment ions were included in the initial list,
which was subsequently used as a starting point to identify the
best responding transitions. In addition, quadruply- and quintu-
ply-charged precursors and doubly- and triply-charged fragments
were added for peptides carrying glutathionylated cysteine
residue.
Transitions for carbamidomethylated peptides and for non-cys-
teinyl peptides were optimized using CBR1 sample incubated with
water and blocked with 60 mM IAM, whereas transitions for pep-
tides carrying either a glutathionylated cysteine residue alone or
in combination with carbamidomethylation as well as transitions
for the dehydro-modified peptide were optimized using CBR1 sam-
ple incubated with 200 mM GSNO followed by 60 mM IAM.
From the initial transition list, the three most intensive transi-
tions were selected per precursor for use in the final assay. For each
but one precursor, the most intensive transition with fragment ion
mass higher than precursor mass was used as a quantifier transi-
2.4. Incubation of CBR1 with GSNO and kinetics
CBR1 was treated with 2 mM GSNO for 2 h at room temperature
in darkness. The enzyme was then rebuffered in 0.1 M potassium
phosphate buffer (pH 7.4) using PD-10 desalting columns (GE
Healthcare, Piscataway, NJ), and the concentration was determined
by Bradford method with BSA as a standard.
2.4.1. Spectrophotometric assays
Catalytic properties of WT-CBR1 and CBR1 ‘‘modified’’ were
measured spectrophotometrically at 340 nm (Cary 100 scan pho-
tometer, Varian, CA). The reaction temperature was held constant
at 25 °C, and the total volume of reaction mixture was 1 ml. The
reaction conditions were 0.1 mM NADPH, 1 mM DTPA (diethylene
triamine pentaacetic acid), 0.1 M potassium phosphate buffer
(pH 7.4), enzyme (WT-CBR1 or CBR1 ‘‘modified’’) dissolved in
0.1 M potassium phosphate buffer (pH 7.4) and the substrate. Dau-
norubicin (5–500
dissolved in water. Menadione (5–200
quinone (2.5–75 M), 1,2-naphthoquinone (10–750
naphthoquinone (5–300 M), isatin (5–200 M) and 4-benzoylpir-
idine (10–1000 M) were dissolved in DMSO. For those substrates,
l
M) and 2,3-hexanedione (5–1000
M), 9,10-phenanthrene-
M), 1,4-
lM) were
l
l
l
l
l
l
the final concentration of DMSO in the cuvette was 10% (v/v). A ref-
erence cuvette was composed of the reaction solution without the
enzyme.
The activity was measured using at least seven different con-
centrations of each substrate, with each measurement being done
in triplicate. The reaction velocity was calculated in mol/min per
mg, corresponding to the amount of NADPH consumed in the reac-
tion by 1 mg of enzyme per 1 min, using the NADPH extinction
coefficient 6,220 Mꢀ1 cmꢀ1. The results were processed using
GraphPad Prism software.
Fig. 2. Sequence of CBR1. Within frames are shown analyzed peptides (DLCR,
DVCTELLPLIKPQGR, SCSPELQQK, ILLNACCPGWVR) after digestion with trypsin. The
peptide DLCR is too short for MALDI-TOF/TOF and LC-SRM analysis.

T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
139
2.4.2. HPLC assays
in CBR1 treated with 2 mM GSNO than in the control sample.
Although MALDI-TOF/TOF MS analysis is not regarded as quantita-
tive, we hypothesized that GSNO might promote formation of a
disulfide bond between Cys-226 and Cys-227. While such an effect
of GSNO was not expected, it is nevertheless still in line with pre-
vious experiments with DTT that restored CBR1 activity [19] and
with the theory, that as soon as the mixed-disulfide bond is formed
at one cysteine, it can undergo a nucleophilic attack by the neigh-
boring cysteine. Since IAM was added to the sample right after
GSNO incubation, we excluded the contribution of artificial oxida-
tion of cysteine residues during sample preparation for MALDI-
TOF/TOF analysis. Nevertheless, we cannot eliminate the possibil-
ity, that in small amount the disulfide bond is formed during the
enzyme preparation.
We detected no peak at m/z = 1954.8 that would represent the
ILLNACCPGWVR peptide with both cysteine residues modified by
S-glutathionylation among those peptides resulting from CBR1
treated with 2 mM GSNO. This fact does not necessarily mean that
such modification is not probable. The absence of such peak might
rather be associated with the most likely low abundance or low
ionization efficiency of the ILLNACCPGWVR peptide with such
modifications as compared to its unmodified counterpart.
For reasons given above, we incubated CBR1 protein with extre-
mely high concentration of GSNO (200 mM, a concentration with-
out physiological relevance) to check if anticipated higher
abundance of the ILLNACCPGWVR peptide modified by glutathione
might overcome the issues associated with its detection in MALDI-
TOF/TOF MS spectra. Overlay of MALDI-TOF/TOF MS spectra from
control CBR1, CBR1 treated with 2 mM GSNO, and CBR1 treated
with 200 mM GSNO showed a specific peak at m/z = 1954.8 in
the latter sample (Fig. 3). Notably, the correlation of peak intensity
at m/z = 1342.7 with increasing concentration of GSNO supports
our initial hypothesis. To restore the reduced state of the ILL-
NACCPGWVR peptide, we treated the sample with tris(2-carboxy-
ethyl) phosphine hydrochloride (TCEP) which effectively reduced
both the disulfide bond as well as the glutathionylated cysteine
residues.
Kinetic properties with oracin and NNK (4-(methylnitrosami-
no)-1-(3-pyridyl)-1-butanone) were measured using HPLC based
on production of their reduced metabolites NNAL (4-(methylnit-
rosamino)-1-(3-pyridyl)-1-butanol) and DHO (11-dihydrooracin).
The enzyme samples were first preincubated for 5 min at 37 °C
in an NADPH regeneration system (0.8 mM NADP+, 6 mM glu-
cose-6-phosphate, 0.35 units of glucose-6-phosphate dehydroge-
nase, 3 mM MgCl2) in a 10 mM sodium phosphate buffer (pH
7.4). The reaction was started by adding the substrate (oracin
0.05–1.5 mM, NNK 1–20 mM). After 30 min of incubation at
37 °C, the reaction was stopped by adding 40
the samples were kept on ice. The reduced metabolites were ex-
tracted three times with 300 l of ethyl acetate; the upper layers
were combined and the organic solvent was evaporated under vac-
uum. The samples were then dissolved in 200 l mobile phase and
used for HPLC analysis. For both substrates, the LiChroCARTÒ 250–
LiChrospherÒ 100 RP-18 (5
m) column (Merck Millipore,
ll of 25% NH3 and
l
l
4
l
Darmstadt, Germany) was used. The composition of the mobile
phase was kept constant during the analysis. For NNK and NNAL
determination, the mobile phase consisted of 10 mM phosphate
buffer (pH 7.4)/ACN (82:18), and the peaks were monitored at
230 nm using a UV detector. The flow rate was 1 ml/min. The com-
position of the mobile phase used for oracin and DHO determina-
tion consisted of 10 mM hexanesulfonic acid and 0.1 M
triethylamine adjusted to pH 3.27 with H3PO4/ACN (75:25). The
peaks were monitored using a fluorescence detector at 340 nm
excitation and 418 nm emission wavelengths. The flow rate was
1.5 ml/min.
All samples were measured in triplicate. Concentrations of
NNAL and DHO were calculated with reference to corresponding
calibration curves. Specific activities were expressed as
NNAL or DHO/min per mg of protein. Results were analyzed using
GraphPad Prism software.
lmol of
3. Results and discussion
The ILLNACCPGWVR peptide carries two cysteine residues,
meaning that in addition to both residues being glutathionylated,
GSNO incubation could also result in an ILLNACCPGWVR peptide
having one glutathionylated and one carbamidomethylated cys-
teine. Such a peptide would correspond to m/z = 1706.8. We ob-
served no such peak in the MALDI-TOF/TOF MS spectrum. As
mentioned above, however, abundance might not be the only rea-
son why a fragment at m/z = 1706.8 remained undetected. Thus,
we performed fast nano-HPLC fractionation of the peptides directly
on a MALDI plate using Probot fraction collector. Using Peak Ex-
plorer software, we extracted m/z = 1706.8 from the LC-MALDI
chromatogram. Subsequent MALDI-TOF/TOF MS/MS analysis and
MASCOT search confirmed this peak to carry glutathione as well
as a carbamidomethyl group.
Encouraged by successful detection of the partially modified
ILLNACCPGWVR peptide, we searched LC-MALDI chromatograms
for the remaining predicted tryptic peptides containing cysteine:
DVCTELLPLIKPQGR, SCSPELQQK and DLCR. While the latter peptide
is too short for MALDI-TOF/TOF analysis, peaks corresponding to
both DVCTELLPLIKPQGR and SCSPELQQK were found and identified
as being modified by S-glutathionylation.
3.1. MALDI-TOF/TOF analysis of CBR1 protein
Tryptic digestion of CBR1 produces 4 cysteine-containing pep-
tides, 3 carrying one cysteinyl residue each (positions 26, 122,
and 150) and 1 peptide with two cysteines (position 226 and
227) (Fig. 2).
To verify the earlier hypothesis regarding modification of cys-
teine residues 226 and 227 in CBR1 by S-glutathionylation upon
incubation with GSNO [19], we performed MALDI-TOF/TOF MS
and MS/MS analyses of tryptic peptides resulting from CBR1 pro-
tein treated with GSNO. In silico digestion of CBR1 showed that cys-
teine residues 226 and 227 comprise part of the tryptic peptide
ILLNACCPGWVR with m/z = 1344.7 (cysteine residues in reduced
form) or m/z = 1458.7 (both cysteine residues carbamidomethylat-
ed with IAM). A gain of two glutathione residues would result in a
peak at m/z = 1954.8.
We observed no substantial difference in MS spectra when
comparing control CBR1 to CBR1 treated with 2 mM GSNO
(Fig. 1, supplementary data). We found a high intensity peak at
m/z = 1458.7 in both spectra that was identified based on MS/MS
analysis and MASCOT search as the ILLNACCPGWVR peptide with
both cysteine residues carbamidomethylated.
3.2. LC-SRM assay development
We furthermore observed a peak at m/z = 1342.7 in MALDI-TOF/
TOF MS spectra, a value that agrees with an ILLNACCPGWVR pep-
tide where the two neighboring cysteine residues form a disulfide
bond. This was confirmed by MALDI-TOF/TOF MS/MS analysis and
subsequent MASCOT search with dehydro-cysteine modification
enabled. Intensity of the peak at m/z = 1342.7 was ꢂ10-fold higher
MALDI-TOF/TOF MS is valuable for initial qualitative analyses,
but, due to its non-quantitative nature, it cannot easily be used
for relative assessment as to the abundance of various peptide
modifications. Therefore, we adopted an LC-SRM MS method to
quantify the effect of GSNO treatment on the cysteine modifica-

140
T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
Fig. 3. MALDI-TOF/TOF mass spectrometry spectra of tryptic peptides resulting from CBR1 protein with a zoomed peak at m/z = 1342.7. Bottom spectrum is control CBR1,
mid-spectrum was obtained from CBR1 protein treated with 2 mM GSNO, upper spectrum was obtained from CBR1 protein treated with 200 mM GSNO. Spectra are offset by
200 a.i. units.
tions that were observed in the MALDI-TOF/TOF MS and MS/MS
analyses.
5) –C[ꢀ1]C[ꢀ1]–. While the analysis of variants 1, 4 and 5 is rela-
tively straightforward, as individual forms differ in molecular
weight and have a distinct retention on the stationary phase, the
discrimination of variants 2 and 3 poses a challenge, since not only
do both isoforms have an identical precursor mass and retention
time, but also the masses of the majority of fragments are identical.
In fact, only the y6 and/or b6 precursor fragments can distinguish
between the two variants. All the aforementioned variants, includ-
ing the discriminating fragments, were successfully detected and,
along with the cysteinyl peptides, the developed LC-SRM assay also
covers all 9 selected non-cysteinyl peptides.
In order to quantify changes in CBR1 modification upon incuba-
tion with GSNO, we developed an LC-SRM assay targeting 4 out of
the total 5 cysteines in the CBR1 sequence. In addition, we selected
9 non-cysteinyl peptides for use in sample processing and assay
quality control (Table 1).
We successfully detected 3 tryptic peptides carrying 4 cysteine
residues (DVCTELLPLIKPQGR, SCSPELQQK, ILLNACCPGWVR). Those
peptides containing a single cysteine residue can be modified
either by S-glutathionylation (adding 305 Da) or by carbamidome-
thylation (adding 57 Da). In addition, the cysteines in the ILL-
NACCPGWVR peptide were shown by MALDI-TOF/TOF analysis to
be modified in a number of combinations: 1) –C[+57]C[+57]–; 2)
–C[+57]C[+305]–; 3) –C[+305]C[+57]–; 4) -C[+305]C[+305]– and a
variant with a disulfide bond between the neighboring peptides,
3.3. LC-SRM analysis
Using the developed LC-SRM assay, we targeted 4 out of 5 cys-
teines in CBR1 in order to quantify changes in S-glutathionylation
Table 1
The developed LC-SRM assay.
Peptide sequence
Peptide
MW
Precursor
m/z
Precursor
charge
Collision energy
(V)
Quantifier
m/z
Ion
Qualifier
m/z
Ion
Qualifier
m/z
Ion
LFSGDVVLTAR
GQAAVQQLQAEGLSPR
FHQLDIDDLQSIR
EYGGLDVLVNNAGIAFK
VADPTPFHIQAEVTMK
TNFFGTR
VVNVSSIMSVR
SETITEEELVGLMNK
EGWPSSAYGVTK
DVC[+57]TELLPLIKPQGR
DVC[+305]TELLPLIKPQGR
SC[+57]SPELQQK
1177.7
1652.9
1599.8
1779.9
1783.9
842.4
1190.7
1692.9
1281.6
1739.0
1987.0
1076.5
1324.6
1458.7
1706.8
1706.8
589.3
551.6
533.9
594.0
595.3
421.7
595.8
565.0
641.3
580.3
497.5
538.8
442.2
729.9
569.6
569.6
489.5
671.8
++
15
11
10
12
12
6
15
11
17
12
9
12
7
22
11
11
8
917.5
729.4
641.3
834.4
806.4
627.3
779.4
661.4
909.5
908.6
454.8
829.4
516.3
934.4
1022.4
774.4
576.9
818.3
y9+
y7+
b5+
y8+
y7+
y5+
y7+
y6+
y9+
y8+
y8++
y7+
y4+
y7+
y6+
y6+
830.5
529.3
503.3
734.3
379.2
480.3
992.5
562.3
812.4
457.3
567.9
742.4
371.7
774.4
740.8
740.8
y8+
y5+
y4+
b7+
y3+
y4+
y9+
y5+
y8+
y4+
y10++ 511.3
y6+
y6++
y6+
y10++ 684.3
y10++ 684.3
658.4
658.4
754.4
720.4
478.3
333.2
692.4
919.4
725.4
585.3
y6+
y6+
b6+
y7+
y4+
y3+
y6+
b8+
y7+
y5+
y9++
y4+
y7++
y5+
y9++
y9++
y9+++
y8+
+++
+++
+++
+++
++
++
+++
++
+++
++++
++
+++
++
+++
+++
++++
++
516.3
415.2
614.3
SC[+305]SPELQQK
ILLNAC[+57]C[+57]PGWVR
ILLNAC[+57]C[+305]PGWVR
ILLNAC[+305]C[+57]PGWVR
ILLNAC[+305]C[+305]PGWVR 1954.8
ILLNAC[-1]C[-1]PGWVR 1342.7
y10+++ 614.3
y7+ 1003.4
y5+
y9+
539.2
889.4
19
The peptide sequence is shown including individual modifications (+57 for carbamidomethylation,+305 for S-glutathionylation, ꢀ1 for dehydro-modification), peptide
molecular weight (MW), precursor mass-to-charge ratio (m/z), precursor charge and respective collision energies in volts. Moreover, both quantifier and qualifier precursor
fragments are shown along with respective ion types and charges.

T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
141
as a function of GSNO concentrations during incubation. CBR1
incubated with water was used as an untreated control. Moreover,
we sought using our assay to verify the presence of a disulfide
bridge between cysteine residues 226 and 227. In addition, we
incubated CBR1 with H2O2 to see whether the disulfide bond
would be formed as well and thus, if this formation can occur dur-
ing oxidative stress when the redox-state conditions are altered.
We were able to confirm our initial MALDI-TOF/TOF hypothesis,
and using LC-SRM the presence of this disulfide bond was detected.
Moreover, in accordance with the suggestion that increased levels
of GSNO would promote the disulfide bond formation, we observed
a positive correlation between the GSNO concentration used dur-
ing incubation and the levels of the ILLNACCPGWVR peptide with
dehydro-cysteine modification. The effect when GSNO causes
disulfide bridge formation has been described, for example, for
the thioredoxin system, where the reduced form of the redox ac-
tive disulfide –Cys–Gly–Pro–Cys– was oxidized after addition
in Fig. 2 in supplementary data. Interestingly, the variant of the ILL-
NACCPGWVR peptide with both cysteines blocked by carbami-
domethylation can be observed at comparable levels in the
control and in the samples incubated at 10 lM and 100 lM GSNO
while its levels drop about threefold in the sample incubated with
2 mM GSNO. It is most likely that formation of the disulfide bridge
between the neighboring cysteines and/or modification of the res-
idues with S-glutathionylation underlies this drop. The results are
summarized in Table 2, where the modified form is always com-
pared to the standard, represented by WT-CBR1.
The incubation with 10
sulted in disulfide bond formation. The greatest increase of this for-
mation was detected at 100 M H2O2, followed by 2 mM H2O2
lM, 100 lM and 2 mM H2O2 also re-
l
(Table 2). We explain this effect by the hypothesis that further oxi-
dation of cysteines into sulfenic, sulfinic, and sulfonic acids at high-
er H2O2 concentrations might happen. During oxidative stress, the
level of 100 lM H2O2 is easily reached in vivo. Thus, disulfide bond
10
lM GSNO [25]. Due to the significantly higher sensitivity of
formation and the associated enzyme activity alteration are
the triple quadrupole instrument in the targeted analyses, we con-
fidently detected the ILLNACCPGWVR peptide with a disulfide
bridge not only at all examined concentrations of GSNO but even
in the untreated control sample represented by WT-CBR1, which
suggests the presence of the disulfide bond in a fraction of our re-
combinant protein sample. Our MALDI-TOF/TOF-based experi-
ments confirmed the presence of an ILLNACCPGWVR variant with
just one cysteine modified with glutathione and the other blocked
by carbamidomethylation. We were able to confirm the presence
of this peptide variant and, in addition, we successfully detected
specific peptide fragments which, as described above, enable the
discrimination of the two isoforms. We show that increasing con-
centration of GSNO during incubation positively correlates with in-
creased levels of both -C[+57]C[+305]– and –C[+305]C[+57]–
isoforms. Our results show that no form is preferred for S-glutath-
ionylation, as the increase in the abundance is comparable for both
isoforms with increasing concentration of GSNO (Table 2).
Although we detected the ILLNACCPGWVR peptide with both
cysteine residues glutathionylated in our MALDI-TOF/TOF analy-
ses, we were only able to observe this variant in CBR1 samples
incubated with 200 mM GSNO. The LC-SRM analysis allowed us
to detect levels of this particular ILLNACCPGWVR variant after
incubation with a 100-fold lower GSNO level, but even LC-SRM
failed to reliably detect this modification at lower concentrations
of GSNO incubation. We speculate that either this variant arises
only with increased levels of GSNO and is thus present at extre-
mely low levels with lower GSNO concentrations or the two gluta-
thione residues inhibit the ionization efficiency of the molecule
and thus impair its detection at low concentrations. The LC-SRM
chromatograms of the ILLNACCPGWVR peptide variants are shown
feasible.
While confirmation of the MALDI-TOF/TOF-based findings
regarding the DVCTELLPLIKPQGR and SCSPELQQK variants carrying
glutathionylated cysteines was not surprising, the fact that we de-
tected these modifications even in the control sample certainly
was. Improper sample handling and accidental contact of the
CBR1 sample with GSNO in the control sample and carryover were
prevented during the LC-SRM analysis.
3.4. Physiological relevance of the confirmed cysteine modifications
The formation of S-glutathionylated proteins (Fig. 4) is well de-
scribed and often serves as a posttranslational regulatory mecha-
nism. It is a consequence of oxidative stress, wherein it protects
the protein thiol group from irreversible oxidation or plays a role
in redox state regulation [26–28].
To date, no such effect has been described for CBR1. As the cys-
teine at position 227 had been described as the reactive residue in-
volved in glutathione binding [19,20], we focused on modifications
at this cysteine. We confirmed formation of the disulfide bond be-
tween adjacent cysteines 226 and 227. The disulfide bond forma-
tion between the two vicinal cysteines changes the 3D structure
of the protein and forces the two vicinal cysteines into a distorted
trans conformation [29]. It can be hypothesized, therefore, that the
disulfide bond formation may change CBR1 conformation in a way
that favors the binding of 1,2-diketo substrates into the active site.
Generally, the disulfide bond formation between two vicinal cyste-
ines is rare and has been described for only a small number of en-
zymes [29]. The function of disulfides is mostly conformational or
the cysteines have a regulatory function as thiol-based redox-
Table 2
Relative extent of cysteine modifications after incubation with GSNO and H2O2.
Modification
Fold change compared to control/AUC
Fragments of CBR1 after trypsin digestion
10
l
M GSNO
100
0.98
47
61
0.72
Not detected
52
Detected
1.8
l
M GSNO
2 mM GSNO
EYGGLDVLVNNAGIAFK
DVC[SSG]TELLPLIKPQGR
SC[SSG]SPELQQK
0.98
1.8
2.8
0.85
Not detected
3.5
112500 2200
630 76
370 82
94600 7000
112300 2400
16300 4500
7900 270
0.92
710
530
0.22
Detected
170
105600 2500
247500 39100
68800 3000
24100 2200
8000 600
ILLNAC[IAM]C[IAM]PGWVR
ILLNAC[SSG]C[SSG]PGWVR
ILLNAC[SSG]C[IAM]PGWVR
ILLNAC[IAM]C[SSG]PGWVR
ILLNAC[Dehydro]C[Dehydro]PGWVR (-S-S-)
80600 2000
140 29
2100 110
6700 510
Detected
0.90
12
4
83
5
Detected
5.0
730
9
11200 670
22600 1000
64500 4800
10
1.7
lM H2O2
100
4.9
lM H2O2
2 mM H2O2
3.4
ILLNAC[Dehydro]C[Dehydro]PGWVR (-S-S-)
22900 1000
66200 2600
46200 5200
The ratio of modification and the AUCs measured for one non-cysteinyl peptide and 4 analyzed cysteines in CBR1 after treatment with different concentrations of GSNO and
H2O2. The ratio defines the fold increase in cysteine modification for ‘‘modified’’ CBR1 (treated with varying GSNO and H2O2 concentrations) compared to the standard, WT-
CBR1. Where the modification was not found in WT-CBR1 is indicated as ‘‘not detected’’, and where quantification was not reliable is indicated as ‘‘detected’’.

142
T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
To ensure we have an enzyme in active form, we measured the
kinetics with GSNO and compared the data so obtained to the re-
sults published by Staab et al. [19]. These were almost identical
(data not shown). For menadione, isatin and 9,10-phenanthrene-
quinone our recombinant WT-CBR1 revealed kinetics constants
comparable to those in the literature, albeit with some differences.
The Km constants obtained in our experiments are similar to the
ones published by other researchers [10,20,35–37]. For menadione,
we obtained the Km value 17.8
lM, for isatin 8 lM and for 9,10-
phenanthrenequinone 31.7 M. This indicates similar recognition
l
of the substrates by the enzyme. In contrast, we noticed some dif-
ferences in kcat, which were usually lower than the published val-
ues (2–10-fold). Kinetics with 1,2- and 1,4-naphthoquinone were
measured spectrophotometrically as described in the materials
and methods in comparison to the modified method used by Pilka
et al. [10]. As both substrates are unstable, they were dissolved in
DMSO and the kinetics studies were performed immediately to
avoid substrate decomposition. Kinetics constants for all substrates
are summarized in Table 3.
The differences in catalytic properties can be explained by addi-
tional factors. It is known that CBR1 occurs in 3 multiple enzyme
forms with differences in size and charge [2]. These differences
are probably caused by various posttranslational modifications
[36,38,39] and Staab et al. [19] had proposed that S-glutathionylat-
Fig. 4. GSNO structure and major mechanisms of protein S-glutathionylation
inspired by Zaffagnini et al. [42]. The physiological levels of GSH in living cells are
up to 10 mM [43] GSH is used as a cofactor for glutathione peroxidase to reduce
toxic peroxides. This reaction results in the oxidized form of glutathione, GSSG,
which is reduced back to GSH by glutathione reductase. In cells, the ratio of
GSH:GSSG is about 100:1 [44].
Table 3
Kinetics constants for WT-CBR1 and CBR1 ‘‘modified’’.
Enzyme form/substrate
Km
(
l
M)
kcat
kcat/Km
(( )
M.min)ꢀ1
switches [29–31]. According to Carugo et al. [29], the second pos-
sibility is more likely for bonds between vicinal cysteines. Those
authors described several cases where the enzyme function is to-
tally abolished when the naturally occurring disulfide bond is re-
duced [29,32,33]. The opposite effect was observed in another
study, where formation of a disulfide bond inactivated the enzyme
[34].
The unexpected detection of the disulfide bond in untreated
WT-CBR1 and its increase after incubation with GSNO and H2O2
may indicate a role for these two cysteines in the regulation of en-
zyme activity and during oxidative stress. Hitherto, no redox-sen-
sitive cysteine has been described for CBR1 and this hypothesis
provides a new viewpoint on the regulation of CBR1 activity. On
the other hand, despite indications as to the cysteine residue in-
volved, we cannot be sure that disulfide bond formation is the only
reason behind the changes in the activity and therefore more
investigation is needed.
(minꢀ1
)
l
WT-CBR1
70 12
208 12
402
3.0
3.4
6.8
1,2-naphthoquinone
CBR1 ‘‘modified’’
1,2-naphthoquinone
WT-CBR1
9,10-phenanthrenequinone
CBR1 ‘‘modified’’
9,10-phenanthrenequinone
WT-CBR1
117 12
6
32
18
8.0
17
5
2
1
3
214 19
586 44
10 0.3
32
1.3
isatin
CBR1 ‘‘modified’’
isatin
18
1
1.1
WT-CBR1
250 38
72 10
0.29
0.94
4.4
4-benzoylpyridine
CBR1 ‘‘modified’’
4-benzoylpyridine
WT-CBR1
1,4-naphthoquinone
CBR1 ‘‘modified‘‘
1,4-naphthoquinone
WT-CBR1
16
7.3
32
18
14
7
2
9
2
3
15
33
17
1
1
2
0.52
2.3
41 0.4
19
308 24
3.5. Kinetic assays and properties of WT-CBR1
menadione
CBR1 ‘‘modified’’
menadione
2
1.3
In 2011, Staab et al. described inhibition of CBR1 with GSNO and
suggested S-glutathionylation of Cys-227 as the reason behind the
enzyme inhibition [19]. Based on the results obtained by Staab
et al. together with the results from our MALDI-TOF/TOF and LC-
SRM MS analysis, we decided to investigate the differences in cat-
alytic properties of the WT-CBR1 and CBR1 ‘‘modified’’ forms. We
incubated CBR1 with 2 mM GSNO for 2 h and then measured the
catalytic properties with various substrates spectrophotometri-
cally at 340 nm and by HPLC. Those results were always compared
to the catalytic properties of WT-CBR1. The quinone substrates
were represented by menadione, 9,10-phenanthrenequinone, 1,2-
and 1,4-naphthoquinone. Other substrates used included isatin
and the cytostatic drug daunorubicin. The ketone substrates were
represented by 4-benzoylpyridine and 2,3-hexanedione, the latter
of which was chosen as a ketone substrate without aromatic struc-
ture. The HPLC method was used for measurements with NNK and
oracin.
WT-CBR1
159 20
1.9
2,3-hexanedione
CBR1 ‘‘modified’’
2,3-hexanedione
WT-CBR1
daunorubicin
CBR1 ‘‘modified’’
daunorubicin
WT-CBR1
oracin
CBR1 ‘‘modified’’
oracin
WT-CBR1
NNK
60
67
43
96
34
5
7
6
9
5
91
76
2
1
1
1
1
2
3
1.5
1.1
29
0.67
0.73
0.68
0.015
0.010
70
23
7840 690
5850 300
121
60
CBR1 ‘‘modified’’
NNK
CBR1 ‘‘modified’’ stands for CBR1 treated with 2 mM GSNO. Kinetics were measured
spectrophotometrically at 340 nm and by HPLC (for oracin, NNK). The errors for kcat
Km were less than 20%.
/

T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
143
Fig. 5. Steady state kinetics measured spectrophotometrically and by HPLC. Mod.: kinetics for CBR1 ‘‘modified’’ (treated with 2 mM GSNO), wt: kinetics for WT-CBR1. The
increased activity of the ‘‘modified’’ form can be seen in the cases of 1,2-naphthoquinone, 9,10-phenanthrenequinone and isatin. The lines represent the fit to the Michaelis–
Menten equation.
ed CBR1 might correspond to one of the three known isoforms of
CBR1 [2,19]. As we have confirmed the disulfide bond in WT-
CBR1, we suggest that this modification can occur in one of those
forms as well. Also, the properties of the recombinant enzyme
and the enzyme purified from human tissue can differ. Next, the
production of recombinant protein and the methods and chemicals
used can alter the catalytic properties. Other aspects which can
influence the enzyme activity is the buffer composition used for
kinetics assays, the type and volume of dissolving agent for the
substrate, the maximal substrate concentration used, and the sub-
strate itself. In any case, it is not unusual to find different kinetic
values published for a given substrate by different authors. As
the same enzyme was used to study the differences in catalytic
properties of the WT-CBR1 and CBR1 ‘‘modified,’’ the results can
be easily compared.
with both WT-CBR1 and CBR1 ‘‘modified’’ using various substrates.
The activity decrease caused by GSNO-mediated S-glutathionyla-
tion was previously described for thioredoxin (Trx) [40], which
led us to the idea that CBR1 might be influenced in a similar man-
ner. The results we obtained were surprising, and especially so
with regard to the substrate structure (Table 3, Fig. 5).
We found that, for most of the substrates investigated, kcat val-
ues decreased 2–5-fold with modified CBR1. In contrast, with 1,2-
naphthoquinone, 9,10-phenanthrenequinone and isatin, kcat was
increased by 2–3 times for the modified form. Interestingly, these
three substrates all contain a 1,2-diketo group in their structures
and are also metabolized by CBR3 (carbonyl reductase 3,
SDR21C2), an enzyme which shares 72% identity with CBR1,
including the cysteine pair at positions 226 and 227, but which dis-
plays a much more narrow substrate spectrum [10].
The observed changes in kcat indicate that the modification af-
fects the rate-limiting step of the reaction. The fact that CBR1 has
different kcat values for the various substrates suggests that release
of NADP+ is not a predominant rate-limiting step. Hence, the mod-
ification most probably affects hydride transfer or release of the
hydroxyl product. Moreover, it does so differently for substrates
3.6. Comparison of catalytic properties of WT-CBR1and CBR1
‘‘modified’’
To explore whether the modification of CBR1 caused by GSNO
has any effect on its activity, we performed kinetics measurements

144
T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
containing two vicinal carbonyl groups in a ring structure (like 1,2-
naphthoquinone, 9,10-phenanthrenequinone, and isatin) and the
other substrates, as kcat is increased for the first substrate type,
but decreased for all others.
analyzed 4 out of 5 cysteines of CBR1 and evaluated their modifi-
cation upon incubation with GSNO. Using MALDI TOF/TOF and
LC-SRM analyses, we confirmed the formation of GSH-mixed disul-
fides on all those cysteines analyzed, as well as the formation of a
disulfide bond between the adjacent cysteines 226 and 227. Stea-
dy-state kinetics with 10 substrates showed that the modifications
affect enzyme activity in terms of turnover rate and substrate bind-
ing. The direction of these changes depended on the location of the
substrate carbonyl groups. For substrates with vicinal carbonyl
groups in a ring structure (1,2-naphthoquinone, 9,10-phenan-
threnequinone and isatin) we observed increased catalytic turn-
over rates, in contrast to decreased rates for all other substrates
tested.
For CBR1, this work provides the first evidence for two vicinal
cysteines that might function as a redox switch representing a no-
vel regulatory mechanism of CBR1 activity. As we have detected
the disulfide bond formation upon incubation with H2O2, a role
as a redox-sensitive cysteine in oxidative stress is also probable.
In conclusion, further investigation is needed on this topic, as the
redox-sensitive cysteine pair would represent a new mode of the
regulation of CBR1 activity [19].
The effect of the modification(s) on catalytic efficiency defined
by kcat/Km, was not as obvious as that for kcat values. For most sub-
strates, a decrease in kcat was totally or in part compensated for by
a decrease in Km. Hence it appears that the modification affects a
kinetic step that is common to kcat and kcat/Km, such as hydride
transfer. Notable changes in catalytic efficiency were only observed
for 4-benzoylpyridine, (kcat decreased 5-fold, kcat/Km increased 3-
fold), for 1,4-naphthoquinone (kcat decreased 2-fold, kcat/Km de-
creased 8-fold), and for 9,10-phenanthrenequinone (kcat increased
3-fold, kcat/Km increased 5-fold). Based on these results, we hypoth-
esize that modification of Cys 226 and/or 227 by GSNO changes
CBR1 activity by affecting the catalytic turnover rate or the nature
of the rate-limiting step as well as by altering substrate binding.
Another point to consider in the interpretation of the kinetic data
is the fact that one would expect to observe negative cooperativity
if the ‘‘modified’’ CBR1 were heterogeneous. However, the data fit
the Michaelis–Menten model well (Fig. 5). This suggests that either
most modified forms show similar kinetic behavior or, what is
more likely, the activity of one differently modified form domi-
nates, accounting for most of the changes in kinetic mechanism
and masking the other less active or less frequent forms.
5. Conflicts of interest
There are no competing interests.
As described in the LC-SRM analysis, we have confirmed forma-
tion of GSH-mixed disulfides with all cysteines analyzed, along
with the disulfide bridge between the cysteines at positions 226
and 227. Based on these findings, we cannot be certain as to which
of the aforementioned modifications could cause the change in the
CBR1 activity. Nevertheless, cysteines at positions 26 and 150 are
apparently distant from the active site and presumably not in-
volved in changes in activity. In case, when the cysteines 226
and 227 would be inactivated by the modification, the possible
explanation is that Cys-122 may play a role in activity changes,
as it is greatly modified and not so far from the active site as cys-
teines 26 and 150. However, Cys-227 has already been identified as
the most reactive residue involved in glutathione binding [19,20],
we suggest that modifications of the vicinal cysteines will be
responsible for the activity changes. Although Staab et al. did not
consider the possibility of the disulfide bond formation, their re-
sults with the enzyme reactivation by DTT are still in accordance
with our findings, as the disulfide bridge as well as the GSH-mixed
disulfide would be reduced by DTT. Also, the facts that the CBR1
Cys227Ser mutant showed a higher Km for GSNO reduction than
did WT-CBR1 together with loss of the enzyme inhibition lends
support to the previous conclusions [19]. This hypothesis is also
supported by the data that the Cys227Ser mutant showed 7 times
higher kcat and 20 times higher kcat/Km values with 9,10-phenan-
threnequinone [19].
Acknowledgements
Our thanks to Dr. Hans-Jörg Martin for interesting discussions
and his help in any situation.
This project was supported by the Grant Agency of Charles Uni-
versity (Grant No. 347211/C/2011), by Charles University Project
SVV 265 004, by the Institutional program of the University Hospi-
tal Hradec Králové, by Postdoctoral fellowship project No. CZ.1.07/
2.3.00/30.0012. The publication is co-financed by the European So-
cial Fund and the state budget of the Czech Republic, TEAB, project
no. CZ.1.07/2.3.00/20.0235. The project was also supported by the
Deutsche Forschungsgemeinschaft (MA 1704/5-2).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cbi.2012.12.011.
References
[1] B. Persson, Y. Kallberg, J.E. Bray, E. Bruford, S.L. Dellaporta, A.D. Favia, R.G.
Duarte, H. Jornvall, K.L. Kavanagh, N. Kedishvili, M. Kisiela, E. Maser, R.
Mindnich, S. Orchard, T.M. Penning, J.M. Thornton, J. Adamski, U. Oppermann,
The SDR (short-chain dehydrogenase/reductase and related enzymes)
nomenclature initiative, Chem. Biol. Interact. 178 (1–3) (2009) 94–98.
[2] B. Wermuth, Purification and properties of an NADPH-dependent carbonyl
reductase from human brain. Relationship to prostaglandin 9-ketoreductase
and xenobiotic ketone reductase, J. Biol. Chem. 256 (3) (1981) 1206–1213.
[3] A. Atalla, E. Maser, Characterization of enzymes participating in carbonyl
reduction of 4-methylnitrosamino-1-(3-pyridyl)-1-butanone (NNK) in human
placenta, Chem. Biol. Interact. 130–132 (1–3) (2001) 737–748.
[4] C. Finckh, A. Atalla, G. Nagel, B. Stinner, E. Maser, Expression and NNK reducing
activities of carbonyl reductase and 11beta-hydroxysteroid dehydrogenase
type 1 in human lung, Chem. Biol. Interact. 130–132 (1–3) (2001) 761–773.
[5] G.L. Forrest, B. Gonzalez, Carbonyl reductase, Chem. Biol. Interact. 129 (1–2)
(2000) 21–40.
One important point to mention in this context is that treat-
ment of CBR1 with GSNO was performed in absence of cofactor,
which does not reflect the in vivo situation where NADPH is avail-
able in micromolar amounts. Future work will also have to eluci-
date whether NADPH binding affects any quantitative and
qualitative aspects of the modification. However, previous work
has provided indirect evidence for the modification in presence
of NADPH, as we have observed inactivation of the enzyme in stea-
dy-state kinetics of GSNO reduction at GSNO concentrations >5 Km.
[6] R.L. Bateman, D. Rauh, B. Tavshanjian, K.M. Shokat, Human carbonyl reductase
1 is an S-nitrosoglutathione reductase, J. Biol. Chem. 283 (51) (2008) 35756–
35762.
[7] E. Ismail, F. Al-Mulla, S. Tsuchida, K. Suto, P. Motley, P.R. Harrison, G.D. Birnie,
Carbonyl reductase: a novel metastasis-modulating function, Cancer Res. 60
(5) (2000) 1173–1176.
[8] J.A. Botella, J.K. Ulschmid, C. Gruenewald, C. Moehle, D. Kretzschmar, K. Becker,
S. Schneuwly, The Drosophila carbonyl reductase sniffer prevents oxidative
stress-induced neurodegeneration, Curr. Biol. 14 (9) (2004) 782–786.
4. Conclusion
Human CBR1 is a well characterized enzyme with broad sub-
strate specificity and is known to play a role in many pathophysi-
ological conditions [41]. Inspired by earlier studies [19,20], we

T. Hartmanová et al. / Chemico-Biological Interactions 202 (2013) 136–145
145
[9] J.A. Doorn, E. Maser, A. Blum, D.J. Claffey, D.R. Petersen, Human carbonyl
reductase catalyzes reduction of 4-oxonon-2-enal, Biochemistry 43 (41) (2004)
13106–13114.
[26] P. Klatt, S. Lamas, Regulation of protein function by S-glutathiolation in
response to oxidative and nitrosative stress, Eur. J. Biochem. 267 (16) (2000)
4928–4944.
[10] E.S. Pilka, F.H. Niesen, W.H. Lee, Y. El-Hawari, J.E. Dunford, G. Kochan, V. Wsol,
H.J. Martin, E. Maser, U. Oppermann, Structural basis for substrate specificity
in human monomeric carbonyl reductases, PLoS ONE 4 (10) (2009) e7113.
[11] B. Gaston, J. Reilly, J.M. Drazen, J. Fackler, P. Ramdev, D. Arnelle, M.E. Mullins,
D.J. Sugarbaker, C. Chee, D.J. Singel, et al., Endogenous nitrogen oxides and
bronchodilator S-nitrosothiols in human airways, Proc. Natl. Acad. Sci. USA 90
(23) (1993) 10957–10961.
[12] J.S. Stamler, O. Jaraki, J. Osborne, D.I. Simon, J. Keaney, J. Vita, D. Singel, C.R.
Valeri, J. Loscalzo, Nitric oxide circulates in mammalian plasma primarily as an
S-nitroso adduct of serum albumin, Proc. Natl. Acad. Sci. USA 89 (16) (1992)
7674–7677.
[13] C. Lindermayr, G. Saalbach, J. Durner, Proteomic identification of S-nitrosylated
proteins in Arabidopsis, Plant Physiol. 137 (3) (2005) 921–930.
[14] K. Sandau, B. Brune, The dual role of S-nitrosoglutathione (GSNO) during
thymocyte apoptosis, Cell. Signal. 8 (3) (1996) 173–177.
[15] K. Inoue, T. Akaike, Y. Miyamoto, T. Okamoto, T. Sawa, M. Otagiri, S. Suzuki, T.
Yoshimura, H. Maeda, Nitrosothiol formation catalyzed by ceruloplasmin.
Implication for cytoprotective mechanism in vivo, J. Biol. Chem. 274 (38)
(1999) 27069–27075.
[16] A.J. de Belder, R. MacAllister, M.W. Radomski, S. Moncada, P.J. Vallance, Effects
of S-nitroso-glutathione in the human forearm circulation: evidence for
selective inhibition of platelet activation, Cardiovasc. Res. 28 (5) (1994) 691–
694.
[17] H. Wu, I. Romieu, J.J. Sienra-Monge, B. Estela Del Rio-Navarro, D.M. Anderson,
C.A. Jenchura, H. Li, M. Ramirez-Aguilar, I. Del Carmen Lara-Sanchez, S.J.
London, Genetic variation in S-nitrosoglutathione reductase (GSNOR) and
childhood asthma, J. Allergy Clin. Immunol. 120 (2) (2007) 322–328.
[18] C.A. Staab, M. Hellgren, J.O. Hoog, Medium- and short-chain dehydrogenase/
reductase gene and protein families: dual functions of alcohol dehydrogenase
3: implications with focus on formaldehyde dehydrogenase and S-
nitrosoglutathione reductase activities, Cell. Mol. Life Sci. 65 (24) (2008)
3950–3960.
[27] I. Dalle-Donne, R. Rossi, D. Giustarini, R. Colombo, A. Milzani, S-
glutathionylation in protein redox regulation, Free Radic. Biol. Med. 43 (6)
(2007) 883–898.
[28] I. Dalle-Donne, A. Milzani, N. Gagliano, R. Colombo, D. Giustarini, R. Rossi,
Molecular mechanisms and potential clinical significance of S-
glutathionylation, Antioxid. Redox. Signal. 10 (3) (2008) 445–473.
[29] O. Carugo, M. Cemazar, S. Zahariev, I. Hudaky, Z. Gaspari, A. Perczel, S. Pongor,
Vicinal disulfide turns, Protein Eng. 16 (9) (2003) 637–639.
[30] M.A. Wouters, R.A. George, N.L. Haworth, ‘‘Forbidden’’ disulfides: their role as
redox switches, Curr. Protein Pept. Sci. 8 (5) (2007) 484–495.
[31] D. Fass, Disulfide bonding in protein biophysics, Annu. Rev. Biophys. 41 (2012)
63–79.
[32] X. Wang, M. Connor, R. Smith, M.W. Maciejewski, M.E. Howden, G.M.
Nicholson, M.J. Christie, G.F. King, Discovery and characterization of a family
of insecticidal neurotoxins with a rare vicinal disulfide bridge, Nat. Struct. Biol.
7 (6) (2000) 505–513.
[33] C.C. Blake, M. Ghosh, K. Harlos, A. Avezoux, C. Anthony, The active site of
methanol dehydrogenase contains
a disulphide bridge between adjacent
cysteine residues, Nat. Struct. Biol. 1 (2) (1994) 102–105.
[34] S.M. Miller, M.J. Moore, V. Massey, C.H. Williams Jr., M.D. Distefano, D.P. Ballou,
C.T. Walsh, Evidence for the participation of Cys558 and Cys559 at the active
site of mercuric reductase, Biochemistry 28 (3) (1989) 1194–1205.
[35] N. Usami, K. Kitahara, S. Ishikura, M. Nagano, S. Sakai, A. Hara, Characterization
of a major form of human isatin reductase and the reduced metabolite, Eur. J.
Biochem. 268 (22) (2001) 5755–5763.
[36] K.M. Bohren, B. Wermuth, D. Harrison, D. Ringe, G.A. Petsko, K.H. Gabbay,
Expression, crystallization and preliminary crystallographic analysis of human
carbonyl reductase, J. Mol. Biol. 244 (5) (1994) 659–664.
[37] Y. El-Hawari, A.D. Favia, E.S. Pilka, M. Kisiela, U. Oppermann, H.J. Martin, E.
Maser, Analysis of the substrate-binding site of human carbonyl reductases
CBR1 and CBR3 by site-directed mutagenesis, Chem. Biol. Interact. 178 (1–3)
(2009) 234–241.
[19] C.A. Staab, T. Hartmanova, Y. El-Hawari, B. Ebert, M. Kisiela, V. Wsol, H.J.
Martin, E. Maser, Studies on reduction of S-nitrosoglutathione by human
carbonyl reductases 1 and 3, Chem. Biol. Interact. 191 (1–3) (2011) 95–103.
[20] J.N. Tinguely, B. Wermuth, Identification of the reactive cysteine residue
(Cys227) in human carbonyl reductase, Eur. J. Biochem. 260 (1) (1999) 9–14.
[21] T.W. Hart, Some observations concerning the S-nitroso and S-phenylsulphonyl
derivatives of L-cysteine and glutathione, Tetrahedron Lett. 26 (16) (1985)
2013–2016.
[38] G.L. Forrest, S. Akman, S. Krutzik, R.J. Paxton, R.S. Sparkes, J. Doroshow, R.L.
Felsted, C.J. Glover, T. Mohandas, N.R. Bachur, Induction of a human carbonyl
reductase gene located on chromosome 21, Biochim. Biophys. Acta 1048 (2–3)
(1990) 149–155.
[39] M. Krook, D. Ghosh, R. Stromberg, M. Carlquist, H. Jornvall, Carboxyethyl lysine
in
a protein: native carbonyl reductase/NADP(+)-dependent prostaglandin
dehydrogenase, Proc. Natl. Acad. Sci. USA 90 (2) (1993) 502–506.
[40] S. Casagrande, V. Bonetto, M. Fratelli, E. Gianazza, I. Eberini, T. Massignan, M.
Salmona, G. Chang, A. Holmgren, P. Ghezzi, Glutathionylation of human
thioredoxin: a possible crosstalk between the glutathione and thioredoxin
systems, Proc. Natl. Acad. Sci. USA 99 (15) (2002) 9745–9749.
[22] M. Strohalm, D. Kavan, P. Novak, M. Volny, V. Havlicek, MMass 3: a cross-
platform software environment for precise analysis of mass spectrometric
data, Anal. Chem. 82 (11) (2010) 4648–4651.
[23] J. Lenco, M. Link, V. Tambor, J. Zakova, L. Cerveny, A.J. Stulik, ITRAQ
quantitative analysis of Francisella tularensis ssp. holarctica live vaccine
strain and Francisella tularensis ssp. tularensis SCHU S4 response to different
temperatures and stationary phases of growth, Proteomics 9 (10) (2009)
2875–2882.
[24] B. MacLean, D.M. Tomazela, N. Shulman, M. Chambers, G.L. Finney, B. Frewen,
R. Kern, D.L. Tabb, D.C. Liebler, M.J. MacCoss, Skyline: an open source
document editor for creating and analyzing targeted proteomics
experiments, Bioinformatics 26 (7) (2010) 966–968.
[41] P. Malatkova, E. Maser, V. Wsol, Human carbonyl reductases, Curr. Drug Metab.
11 (8) (2010) 639–658.
[42] M. Zaffagnini, M. Bedhomme, S.D. Lemaire, P. Trost, The emerging roles of
protein glutathionylation in chloroplasts, Plant Sci. 185–186 (2012) 86–96.
[43] A. Meister, Glutathione metabolism and its selective modification, J. Biol.
Chem. 263 (33) (1988) 17205–17208.
[44] A. Pompella, A. Visvikis, A. Paolicchi, V. De Tata, A.F. Casini, The changing faces
of glutathione,
1499–1503.
a cellular protagonist, Biochem. Pharmacol. 66 (8) (2003)
[25] D. Nikitovic, A. Holmgren, S-nitrosoglutathione is cleaved by the thioredoxin
system with liberation of glutathione and redox regulating nitric oxide, J. Biol.
Chem. 271 (32) (1996) 19180–19185.