Tài liệu Báo cáo khoa học: Antioxidant protein 2 prevents methemoglobin formation in erythrocyte hemolysates doc

Antioxidant protein 2 prevents methemoglobin formation
in erythrocyte hemolysates
Karl M. Stuhlmeier
1
, Janet J. Kao
2
, Pia Wallbrandt
3
, Maria Lindberg
3
, Barbro Hammarstro¨m
3
,
Hans Broell
1
and Beverly Paigen
4
1
Ludwig Boltzmann Institute for Rheumatology and Balneology, Vienna, Austria;
2
Maimonides Medical Center, Brooklyn, NY,
USA;
3
Department of Molecular Biology, AstraZeneca, Umea
˚
, Sweden; and
4
The Jackson Laboratory, Bar Harbor, ME, USA
Antioxidant protein 2 (AOP2) is a member of a family of
thiol-specific antioxidants, recently renamed peroxiredoxins,
that evolved as part of an elaborate system to counteract and

control detrimental effects of oxygen radicals. AOP2 is
found in endothelial cells, erythrocytes, monocytes, T and B
cells, but not in granulocytes. AOP2 was found solely in the
cytoplasm and was not associated with the nuclear or
membrane fractions; neither was it detectable in plasma.
Further experiments focused on the function of AOP2 in
erythrocytes where it is closely associated with the hemo-
globin complex, particularly with the heme. An investigation
of the mechanism of this interaction demonstrated that the
conserved cysteine-47 in AOP2 seems to play a role in
AOP2-heme interactions. Recombinant AOP2 prevented
induced as well as noninduced methemoglobin formation in
erythrocyte hemolysates, indicating its antioxidant proper-
ties. We conclude that AOP2 is part of a sophisticated system
developed to protect and support erythrocytes in their many
physiological functions.
Keywords: hemoglobin; erythrocytes; reactive oxygen species;
antioxidant protein 2.
Evolving antioxidant defence systems to protect against O
2
toxicity has been a prerequisite for an organism’s use of O
2
for efficient energy production. To benefit from O
2
as an
energy source, multicellular organisms had to develop a
system to distribute O
2
. In mammals this function is carried
out by red blood cells (RBC), which utilize hemoglobin to

distribute O
2
to cells. Not only are RBC highly specialized
O
2
and CO
2
carriers, they also serve an additional important
function, namely acting as a sink for reactive oxygen species
(ROS) [1]. Erythrocytes can take up O
2
-radicals as well as
H
2
O
2
in plasma to protect the organism from damage by
such compounds [2–4]. These tasks make erythrocytes
especially vulnerable to damage by ROS. Furthermore,
carrying high concentrations of O
2
andhighlevelsof
potentially pro-oxidant heme protein inside a membrane
rich in polyunsaturated fatty acid side chains cause
additional problems. RBCs are therefore exposed to a
constant flow of hemoglobin auto-oxidation, as approxi-
mately 3% of the hemoglobin undergoes oxidation to
methemoglobin (metHb) every day. Moreover RBC are
also exposed to repeated physical stress through deforma-
tion. More importantly, RBC have low metabolic activities

with no ability to synthesize new proteins or lipids to replace
damaged molecules [1]. Due to these properties, RBC need
to be equipped with a series of enzymes that can protect cells
from damage by free radicals; such enzymes include Cu-Zn-
superoxide dismutase, catalase, glutathione peroxidase,
metHb reductase, and glucose 6-phosphate dehydrogenase.
Recently, a new type of antioxidant protein has been
reported to be present in RBC [5,6], the thiol-specific
antioxidant proteins, which are members of a large family of
more than 40 proteins found in prokaryotes as well as
eukaryotes [7–10]. The peroxiredoxin proteins show no
significant homology with previously identified antioxidant
proteins. The nomenclature of these proteins is still confu-
sing, as these molecules were originally described under
several names e.g. rehydrins, thioredoxin-dependent per-
oxide reductases, but this family has been renamed as
peroxiredoxins [11,12]. Peroxiredoxins are grouped into
1-Cys proteins with a conserved cysteine at amino acid
position 47 and 2-Cys proteins with a second conserved
amino acid at position 170 (relative to yeast peroxiredoxin).
They usually exist as homodimers. The substrates are alkyl
hydroperoxides [9], peroxynitrates [13] and hydrogen per-
oxides [14], and they detoxify these substrates by oxidation
of the Cys at amino acid 47 [9,15]. These proteins
enzymatically detoxify hydroxyradicals using reducing
equivalents from thiol-containing molecules such as thio-
redoxins and glutathione. As a major function of these
proteins is to regulate ROS levels, they not only protect
Correspondence to K. M. Stuhlmeier, Ludwig Boltzmann
Institute for Rheumatology and Balneology, Kurbadstrasse 10,

PO Box 78, A-1107 Vienna, Austria.
Fax: + 43 1 68009 9234, Tel.: + 43 1 68009 9237,
E-mail:
Abbreviations: AOP2, antioxidant protein 2; metHb, methemoglobin;
MNCs, mononuclear cells; PMNs, polymorphonuclear cells;
RBC, red blood cells; ROS, reactive oxygen species.
Note: The nomenclature of antioxidant protein 2 is currently under-
going reconsideration. This protein is currently named antioxidant
protein 2 in humans and peroxiredoxin 5 in mice. However, peroxi-
redoxin 5 in humans refers to a different protein (named peroxiredoxin
6 in mouse). As the same protein is supposed to have the same name
in different species, we will use the old name of antioxidant protein 2
until this nomenclature issue is resolved by the human and mouse
nomenclature committees.
(Received 11 June 2002, revised 13 October 2002,
accepted 26 November 2002)
Eur. J. Biochem. 270, 334–341 (2003) Ó FEBS 2003 doi:10.1046/j.1432-1033.2003.03393.x
macromolecules from oxidation, but also may be involved
in signal transduction as well because ROS are implicated in
physiological signaling [16,17].
We investigated the distribution and function of AOP2,
which is a 1-Cys member of the peroxiredoxin protein
family. Its presence has been described for mouse, pig and
human cells. Murine AOP2 was first isolated from the liver
and kidney as a cDNA corresponding to a protein variant
that differs between the C57BL/6J and DBA/2J strains of
mice [18]. Subsequently, the genomic structure of AOP2 in
mice was determined [19] and the crystal structure of the
human protein determined [20]. An analysis of the EST
database suggested that AOP2 may be ubiquitously

expressed [21]. The encoded protein is 224 amino acids in
length with a predicted size of 25 kDa. However, the protein
detected in cells of mouse, human, rat and pig origin, as well
as the recombinant form of AOP2, migrates in PAGE as a
protein of approximately 32 kDa. To date, the protein
expression pattern of the native AOP2 has been reported
only for rat, which demonstrated AOP2 exclusively in the
lung [22]. However, we screened organs and different cell
types in mice with an antibody raised against a unique
peptide from the AOP2 protein and found high levels of this
protein in essentially all organs (unpublished results).
Endothelial cells, erythrocytes and white blood cells are
targets as well as sources of many forms of ROS [1]. We
therefore investigated the distribution and function of this
novel antioxidant protein in these cells. Herein we report
our findings on AOP2 distribution and its protective effects
in RBC, as well as the mechanism of interactions.
Experimental procedures
Materials
Pyrrolidine dithiocarbamate, cysteine, serine, alanine, cross-
linked hemoglobin, hemin, pig serum, Igepal CA-630, and
globin were obtained from Sigma (St. Louis, MO) or Sigma
(Vienna, Austria).
L
-Glutamine was obtained from PAA
Labor- und ForschungsgesmbH. (Vienna, Austria), penicil-
lin, streptomycin, cell culture medium and fetal bovine
serum from Gibco BRL, Life Technologies (Vienna,
Austria). Tumor necrosis factor alpha was obtained from
R & D Research (Minneapolis, MN).

Generation of AOP2 antibodies
One synthetic peptide (FPKGVFTKELPSGKKYLRYC)
corresponding to amino acids 202–220 of murine AOP2 was
generated. The peptide sequence was chosen based on an
antigenicity calculation (Jamesson–Wolf, based on hydro-
philicity, surface probability, flexibility and secondary
structure predictions) and had been shown by Kang et al.
[23], to be an antigen determinant. An additional amino
acid, cysteine, was added to the C-terminal end of the
synthetic peptide to obtain specific coupling of the peptide
to BSA and affinity gel matrices. Rabbits were immunized
with BSA-conjugated peptides and repeatedly injected to
increase the antibody response. The IgG antibody fraction
was purified from the antisera obtained after IV immuni-
zations using Protein G Sepharose 4 Fast Flow (Pharmacia
LKB Biotechnology, Uppsala, Sweden). Polyclonal AOP2
antibodies were further purified on Sulfolink (Pierce,
Rockford, IL, USA) affinity columns to which the synthetic
peptides had been coupled by their cysteine residues.
Affinity-purified antibodies were characterized by ELISA
and Western blot analysis. In immunoblotting, the AOP2
antibody recognized a polypeptide with a relative molecular
mass of approximately 32 kDa in human and mouse liver
homogenates. N-terminal amino acid sequencing identified
the 32 kDa polypeptide as AOP2. In addition, the AOP2
antibody recognized recombinant AOP2D124A and the
recombinant protein encoded by the Aop2-related sequence 1
(data not shown).
Recombinant AOP2
Production and purification of recombinant AOP2 in

Escherichia coli were performed mainly as described [23].
Full-length Aop2 cDNA was amplified by the polymerase
chain reaction (PCR) and subsequently used for cloning.
PCR was performed with forward primer (5¢-CGGCA
TATGCCCGGAGGGTTGCTTCTC-3¢), which contains
nucleotides 1–21 of the mouse Aop2 sequence, the initiation
codon and a NdeI cleavage site, and reverse primer (5¢-
CGCGAATTCTTATTAAGGCTGGGGTGTATAACG
G-3¢), which contains nucleotides 653–671 of the mouse
Aop2 sequence, two stop codons and an EcoRI cleavage
site. The resulting PCR products were cloned into pGEM-
T. An NdeI–EcoRI fragment from pGEM-T, containing the
cDNA encoding AOP2, was subcloned into pET-17b
generating pETAop2D123A. The nucleotide sequence of
pETAop2D124A was verified by DNA sequencing.
E. coli cells (BL21(DE3)pLysS) were transformed with
pETAop2D124A. Cells were grown overnight in a small
volume of liquid broth supplemented with carbencillin
and chloramphenicol and thereafter transferred to new
medium. Production of recombinant protein was induced
by addition of isopropyl-b-
D
-thiogalactopyranoside. Cells
were harvested, disrupted and recombinant protein was
purified from the soluble fraction as described [23]. The
purification method includes streptomycin sulfate precipi-
tation, ammonium sulfate precipitation, hydrophobic
chromatography and anion exchange chromatography.
The sample was applied to a Q-Sepharose column
(Pharmacia LKB Biotechnology, Uppsala, Sweden), equil-

ibrated with 20 m
M
Tris/HCl (pH 8.0), 2 m
M
dithio-
threitol and 1 m
M
EDTA. At this pH recombinant AOP2
was bound to the column matrix. Thereafter, bound
material was eluted with a linear gradient of 0–0.5
M
NaCl in 20 m
M
Tris/HCl (pH 8.0), 2 m
M
dithiothreitol
and 1 m
M
EDTA.
During purification recombinant AOP2 was detected by
immunoblot analysis using specific polyclonal antibodies.
The purified protein was more than 95% pure and the yield
was  50%. The N-terminal amino acid sequence, purity
and accurate molecular mass of recombinant AOP2 were
verified by N-terminal amino acid sequencing and electro-
spray mass spectroscopy (data not shown).
Preparation of cell lysate
Blood from human or mouse was diluted in phosphate
buffered saline (NaCl/P
i

), and RBC separated from white
Ó FEBS 2003 Distribution and function of antioxidant protein 2 (Eur. J. Biochem. 270) 335
blood cells by density gradient centrifugation on a 67% (w/v)
Percoll (Amersham Pharmacia Biotech, Uppsala, Sweden)
gradient. RBC were washed three times with NaCl/P
i
,and
then hemolyzed with three volumes of distilled water for
15 s. Afterwards, 10 · NaCl/P
i
wasusedtoadjustthe
salinity and pH to physiological conditions. Hemolysate
was used for metHb measurements, Western blots and pre-
cipitation studies following centrifugation at 20 800 g for
15 min at 4 °C. RBC membrane extract was prepared as
follows: ghosts (pellets after hemolysis) were washed three
times with NaCl/P
i
, afterwards, membranes were dissolved
in lysis buffer containing 0.32
M
sucrose, 3 m
M
CaCl
2
,
2m
M
magnesium acetate, 0.1 m
M

dithiothreitol, 0.5 m
M
phenylmethylsulfonyl fluoride, and 0.5% Igepal CA-630. In
some cases, RBC were lysed directly in such a buffer and the
resulting protein solution used for Western blot experi-
ments. Endothelial cell cytoplasm and nuclear extract were
prepared as described [24]. Polymorphonuclear cells were
separated from mononuclear cells on a density gradient (67
and 55% Percoll, respectively), washed three times and cell
extracts prepared as described above. B and T cell extracts
were obtained from whole cell lysates of human lymphoma
cells, BJAB (Santa Cruz Biotechnology, Santa Cruz, CA)
and mouse T cell lymphoma cells, CTLL-2 (Santa Cruz
Biotechnology, Santa Cruz, CA).
Cell culture
Porcine aortic endothelial cells were isolated and cultured as
previously described [24]. Briefly, endothelial cells were
cultured in Dulbecco’s modified Eagle’s medium with
4.5 gÆL
)1
glucose and supplemented with 10% heat inacti-
vated fetal bovine serum,
L
-glutamine and 50 unitsÆmL
)1
penicillin/streptomycin.
Methemoglobin measurement
Hemolysate was placed in disposable cuvettes (polymethyl
methacrylate, 2.5 mL capacity) with a light path of 1 cm.
Lysate and indicated amounts of reagents were incubated at

room temperature without shaking. The metHb formed was
measured at time intervals ranging from 0–200 h. A
spectrophotometer (Spectronic Genesys 5, Milton Roy,
Rochester, NY) was used in survey scan mode to measure
absorbance between 400 and 700 nm.
Immunoprecipitation
Cross-linked hemoglobin, hemin and globin were obtained
from Sigma (Vienna, Austria). Aliquots of beads were
washed three times in NaCl/P
i
. Red blood cell lysate,
antibodies to AOP2, recombinant AOP2 protein, bovine
serum albumin, or a combination of the above were
incubated at 4 °C on a rotating device. We used either
protein A/G PLUS-Agarose (Santa Cruz Biotechnology,
Santa Cruz, CA) or hemoglobin (bovine) cross-linked with
cyanogen bromide to beaded agarose for these precipitation
studies. To ensure high specificity of protein–protein
interactions, after 1 h beads were collected by centrifugation
and washed six times in high stringency buffer (0.1%
Tween 20, 0.1% SDS, 1% Igepal CA-630, in NaCl/P
i
).
After the last wash solution was aspirated, 25 lLSDS
sample buffer was added, heated to 95 °C for 5 min, and
immunoprecipitates collected for SDS/PAGE.
Western blot analysis
Cytosolic and nuclear extracts from endothelial cells,
mononuclear cells (MNCs) and polymorphonuclear cells
(PMNs) were prepared as described [25]. Equal amounts of

protein (immunoprecipitates, nuclear and/or cytoplasm
extract) were separated by SDS/PAGE (10 or 12%),
transferred to an Immobilon-P poly(vinylidene difluoride)
(PVDF) membrane using a semidry transfer cell (Bio-Rad
Laboratories, Hercules, CA), and probed with a rabbit
polyclonal antibody no. 5 against AOP2. Bands were
visualized using horseradish peroxidase, conjugated donkey
antirabbit IgG, and the Enhanced ChemiLuminescence
assay (Amersham Life Science Inc., Arlington Heights, IL)
according to the manufacturer’s instructions.
Results
Expression of AOP2
Using Western blots, we determined the expression of
AOP2 in various cell types. High levels of AOP2 occurred in
the cytoplasm of aortic endothelial cells but not in the
nuclear extract (Fig. 1A). Furthermore, stimulating endo-
thelial cells (50 l
M
of pyrolidine dithiocarbamate or
5ngÆmL
)1
of tumor necrosis factor) for 10 min, 2, 8 or
Fig. 1. Expression of AOP2 in various tissues. Arrow indicates AOP2. (A) Western blots demonstrated that AOP2 was expressed in RBC and
MNCs but not in PMNs. Furthermore, in endothelial cells (EC), AOP2 is found in the cytoplasm (EC Cyto) but not in the nucleus (EC Nuc). AOP2
was expressed similarly in mouse cells (data not shown). (B) Whole cell lysate of resting B and T cells were separated by SDS/PAGE and stained
with a specific anti-AOP2 Ig as described in experimental procedures section. (C) RBC from two mice (A and B) were lysed and separated into
membrane and cytoplasm (stroma) fractions. Equal amounts of protein were separated by SDS/PAGE. Membranes were exposed to films for times
ranging from 30 s to 15 min. The ÔMembraneÕ blot was exposed approximately 10 times longer than was the ÔStromaÕ blot.
336 K. M. Stuhlmeier et al.(Eur. J. Biochem. 270) Ó FEBS 2003
24 h did not lead to changes in AOP2 levels (data not

shown). Next, we tested whether polymorphonuclear cells
(PMNs), mononuclear cells (MNCs), or erythrocytes
express AOP2. Human and mouse blood was collected,
and cells separated immediately on a Percoll gradient.
AOP2wasdetectedinRBCandMNCs,butnotinPMNs
(Fig.1B).EvengelsloadedwithmaximalamountsofPMN
whole cell extract (50 lgproteinÆlane
)1
) and deliberately
overexposed exhibited only a weak band of AOP2 (data not
shown), which may have resulted from RBCs contamin-
ating the PMN fraction. A representative Western blot
(Fig. 1B) indicates that both B cells (human BJAB cells) and
T cells (mouse lymphocyte cell line CTLL-2) contained
equal amounts of AOP2 and showed no difference in its
molecular size. Similarly, AOP2 of RBC from mouse and
human origin had identical staining patterns (data not
shown).
AOP2 is located in the cytoplasm of RBC but is not
membrane bound (Fig. 1C). We collected blood of A/J and
C57BL/6 strains, separated the plasma from the solid
components, fractioned the RBCs, and electrophoresed the
membrane and cytoplasm portions on a 12% gel. The gels
were immunostained, and blots were exposed to films for
30 s to an hour. Even when exposed for an hour, blots
revealed that AOP2 was present only in RBC cytoplasm
(Fig. 1C).
AOP2 was not present in mouse, pig, or human plasma.
By using SDS/PAGE (5 lL plasma per lane) and over-
exposing Western blots so that concentrations of recom-

binant AOP2 smaller than 4 ng per lane could easily be
detected (Fig. 2), we determined that AOP2 was absent
from porcine, murine, and human plasma (data not shown).
To gain insight into how AOP2 interacts with itself or
other proteins, RBC stroma proteins were examined using
native PAGE. Under native conditions, AOP2 migrates as a
molecule of approximately 130 kDa in contrast to the
32 kDa observed under reducing conditions (Fig. 3A).
Because adding either reducing SDS or native, nonreducing
sample buffer precipitated hemoglobin, we vortexed the
sample before loading the gel to ensure uniform protein
distribution. We used this precipitation of hemoglobin to
study AOP2–hemoglobin interactions. When RBC stoma
protein was incubated in nonreducing native sample buffer,
and supernatant and hemoglobin pellets collected separately,
electrophoresis on a native gel showed no AOP2 in the
supernatant. This fact suggested that AOP2 coprecipitates
with the hemoglobin fraction. Hemoglobin in its native state
forms a 64-kDa tetramer consisting of four 16 kDa
subunits; this hemoglobin complex is visible on the blots
as white areas (Fig. 3A), because the large quantities of
hemoglobin blocked nonspecific binding of antibodies
during the immunostaining procedure.
AOP2 binds to hemoglobin
As AOP2 coprecipitated with hemoglobin, we tested
whether AOP2 actually binds to the hemoglobin complex.
Hemoglobin cross-linked to agarose beads were prepared
and used as described; special care was taken to remove
unbound proteins through extensive washing steps under
high stringency conditions. As AOP2 could be detected in

liver, heart, testis, and brain (data not shown), we used
whole cell extracts of RBC, liver and heart as well as
recombinant AOP2 for these experiments. Recombinant
AOP2, as well as AOP2 isolated from liver and heart cells,
binds to hemoglobin. Interestingly, when red blood cell
extract, which contains high levels of AOP2 (as shown in
Fig. 1), were used, no AOP2 bound to hemoglobin cross-
linked to beads. The controls, which have NaCl/P
i
,dem-
onstrate that purchased hemagarose beads are free of AOP2
protein. Furthermore, AOP2 binding to hemagarose beads
is specific, as it could not be blocked by addition of bovine
serum albumin.
Next, RBC, heart and liver cell extracts were incubated
on a rolling platform for 1 h at 4 °C together with the anti-
AOP2 IgG and protein A/G cross-linked to agarose beads.
As expected AOP2 was immunoprecipitated from the cell
extracts (Fig. 4B). Interestingly, large amounts of hemo-
globin coprecipitated with the AOP2 from RBC extracts
Fig. 2. Demonstration of sensitivity and linearity of the Western blot
system used. AOP2 could be detected at 4 ng. Recombinant AOP2 was
separated by SDS/PAGE under reducing conditions at increasing ng
quantities. At 400 and 1000 ng, the formation of dimers can be
observed where the amount of AOP2 added exceeds the reducing
capacity of 2-mercaptoethanol.
Fig. 3. AOP2 binds to and coprecipitates with hemoglobin. (A). Under
nonreducing conditions, AOP2 migrates as a molecule complex of
approximately 130 kDa. Hemoglobin (Hb) tetramers can be observed
at 64 kDa as a light unstained band. These samples were RBC lysates

separated on a 10% PAGE gel. (B). Under native conditions, AOP2
precipitated with hemoglobin. This Western blot was prepared after
RBClysateinnativesamplebufferwaswarmedtoprecipitatehemo-
globin. After centrifuging the lysate and the precipitated hemoglobin,
10 and 20 lL aliquots of supernatant and the washed and resuspended
hemoglobin pellet were electrophoresed by native-PAGE on a 10%
gel.
Ó FEBS 2003 Distribution and function of antioxidant protein 2 (Eur. J. Biochem. 270) 337
(Fig. 4B), supporting the hypothesis that AOP2 binds
tightly to hemoglobin complexes. Hemoglobin is visible
as the unstained (lighter) areas in Western blots and as
reddish-colored dots on PVDF membranes following
protein transfer. Hemoglobin monomers (16 kDa), dimers
(32 kDa) and tetramers (64 kDa) are readily recognizable
on the blots because the large amounts of hemoglobin could
not be reduced completely by the 2-mercaptoethanol
present in sample buffer. These experiments indicated that
AOP2 is tightly bound to the hemoglobin complex. The
failure to precipitate any AOP2 with hemagarose (Fig. 4A)
may indicate that all AOP2 in RBC is tightly bound to
hemoglobin so that no free AOP2 protein is available to
react with added hemagarose.
To determine if AOP2 binds to the globin or the heme
portion of hemoglobin, we incubated hemin, globulin and
hemoglobin cross-linked to agarose beads with red blood
cell lysate for 1 h at 4 °C. Beads were collected by
centrifugation and extensively washed with high stringency
buffer. Western blots revealed that AOP2 in red blood cell
lysate bound only to hemin but not to globin (Fig. 4C). It is
somewhat puzzling that AOP2 in red blood cell lysate

bound to heme alone, but not to heme in hemoglobin. This
might result if AOP2 has a higher affinity for isolated heme
so that heme on agarose beads can successfully compete for
AOP2 already bound to the endogenous hemoglobin in
RBC lysate.
Cysteine 47 is essential for AOP2-heme binding
The binding of AOP2 and heme led us to investigate the
mechanism by which AOP2 binds to heme. All peroxire-
doxin proteins have a conserved cysteine residue corres-
ponding to Cys47 in yeast peroxiredoxin [26]. To determine
if the sulfur group of this cysteine bound to the iron in heme,
heme cross-linked to agarose beads were preincubated for
30 min at 4 °Cwith25m
M
cysteine, 25 m
M
alanine,
25 m
M
serine, or NaCl/P
i
(as control). Subsequently, equal
amounts of recombinant AOP2 were added and the mixture
incubated for an additional hour at 4 °Conarotary
platform. The supernatants were aspirated after a final wash
and the pellets resuspended in 40 lL of reducing sample
buffer. Aliquots of 20 lL were loaded per lane and
separated by PAGE. Quantitation of the immunoblot
reveals that hemagarose beads preincubated with cysteine
inhibited the subsequent binding of AOP2 by 86%, while

preincubation with alanine or serine had no effect (Fig. 5),
indicating that AOP2 does bind to heme by its conserved
cysteine.
AOP2 prevents induced and spontaneous
methemoglobin formation
ROS oxidize hemoglobin to methemoglobin (metHb),
which is unable to deliver oxygen to tissues. MetHb can
form either spontaneously or be induced to form by many
substances, including ascorbic acid at high doses [1,27]. To
determine if the tight binding of AOP2 to hemoglobin could
prevent MetHb from forming spontaneously, we compared
the amount of MetHb in fresh hemoglobin to that formed
after 72 h in three solutions: untreated hemoglobin, hemo-
globin treated with 4 lgÆmL
)1
AOP2, and hemoglobin
treated with 5 m
M
ascorbic acid. After comparing the
absorbance of these three solutions to that of MetHb
(characteristic peak at 620–640 nm) [28] and unoxidized
fresh Hb (576 nm), we concluded that treatment with AOP2
preventedtheoxidationofHbtometHb(Fig.6).
We then determined whether AOP2 could prevent
ascorbic acid-induced metHb from forming and whether
such an effect would be dose-dependent. We prepared two
samples of fresh hemoglobin combined with 5 m
M
ascorbic
acid and containing either 2 lgor7lg of recombinant

AOP2 per ml. MetHb formation was measured in these two
samples after they had incubated for 48 and 120 h at 25 °C.
MetHb was calculated as a ratio of absorbance at 575 and
626 nm according to published methods [28], with the slight
modification of using the second peak of the hemoglobin
spectrum instead the first at around 546 nm. As Table 1
shows, AOP2 did indeed prevent ascorbic acid-induced
metHb from forming, and it did so in a dose-dependent
manner.
Fig. 4. AOP2 binds to hemoglobin. (A) Recombinant AOP2, and
cytoplasm extracts from heart, liver, and RBC were incubated with
hemoglobin bound to agarose beads. High stringency washing con-
ditions were used to remove unbound proteins. Aliquots of beads were
added to reducing SDS sample buffer, and the solutions loaded on a
10% gel. Recombinant AOP2, as well as AOP2 in heart and liver cells
bound to hemoglobin, while the AOP2 present in RBC did not. BSA
was added to some samples to block nonspecific binding sites on
hemoglobinagarosebeads.(B)Lysatesofliver,heart,orRBCwere
incubated with anti-AOP2 Ig and protein A/G cross linked to agarose
beads. The anti-AOP2 Ig not only precipitates AOP2 in red blood cell
lysate but also pulls down hemoglobin (indicated as lighter area in the
Western blot) bound to AOP2. This is a further indicator of the tight
interactions of AOP2 and hemoglobin in RBC. (C) Red blood cell
lysate was incubated for 1 h at 4 °C on a rotating platform with
agarose beads cross linked to hemin, globin, or hemoglobin. AOP2
bound to the heme molecule but not to globin or hemoglobin.
Fig. 5. Cysteine is essential for AOP2 heme binding. Hemin, cross-
linked to agarose beads, was used to study the involvement of cysteine
in AOP2–heme interactions. AOP2 was incubated for 1 h at 4 °Cwith
hemin-beads preincubated with NaCl/P

i
or (25 m
M
) cysteine, serine or
alanine for 30 min at 4 °C. Unbound AOP2 was removed by exces-
sive washing. Cysteine blocks the binding of AOP2 to hemin while
serine and alanine had no effect.
338 K. M. Stuhlmeier et al.(Eur. J. Biochem. 270) Ó FEBS 2003
Discussion
We report on the distribution and function of AOP2 in
the four cell types tested: granulocytes, lymphocytes,
endothelial cells, and RBCs. All these cells are either
exposed to or can release ROS and would therefore
appear to benefit from mechanisms to protect themselves
from ROS. RBCs are especially vulnerable because they
have a high iron content, are exposed to high oxygen
pressure [29], and cannot repair proteins or other
molecules damaged by ROS. We demonstrated that
AOP2 is a cytoplasmic protein widely present in lympho-
cytes, endothelial cells, and RBC, but not in granulocytes.
It was not present in plasma or in RBC membranes or
nucleus.
Aop2 belongs to a multigene family, the members of
which each have multiple bands detectable by Southern
blot [21,30,31]. It is unknown if these proteins function
similarly in different cell types. Our data indicate that
AOP2, in a dose-dependent manner, significantly protects
RBCs from ROS by binding to hemoglobin, thus
preventing both its induced and spontaneous oxidation
to metHb. Another member of the peroxiredoxin protein

family, the 25-kDa protector protein [32], has also been
shown to suppress metHb formation and membrane lipid
peroxidation.
The alignment of the peroxiredoxin family members
reveals that many of them have two highly conserved
cysteine residues corresponding to the Cys47 and Cys170
in yeast peroxiredoxin. The N-terminal cysteine is con-
served in all family members and the C-terminal is
conserved in all except six members. AOP2 is among the
family members with only one conserved cysteine. The
eight other amino acids in the cysteine region of mamma-
lian peroxiredoxin are 100% conserved, indicating an
essential role in peroxiredoxin function. Furthermore, there
is an additional motif of six conserved amino acids out of
11, as well as a region with six out of 15 conserved amino
acids [21]. Both Cys47 and Cys170 were shown to be
necessary to maintain peroxiredoxin polymers, but only
the Cys47 region seems to be essential for antioxidant
activity [33].
Because cysteine is often present to help bind heme
groups to substrates, we investigated whether it was
essential for binding the heme iron to AOP2. We found
that heme preincubated with cysteine did not bind to
AOP2, whereas heme preincubated with several other
unrelated amino acids freely bound to AOP2. This result
contrasted somewhat to those of others showing that yeast
peroxiredoxin with mutant Cys47 cannot prevent the
inactivation of glutamine synthetase induced by dithio-
threitol/Fe
3+

/O
2
, indicating that Cys47 has an antioxidant
rather than a binding function [33]. Nevertheless, the same
study found that Cys47 is essential to maintain peroxire-
doxin dimers, indicating that it does have a binding
function. This was supported by our findings that recom-
binant AOP2, not disassociated by a reducing agent,
separated into large complexes of dimers and tetramers
(data not shown). Because cysteine is almost always
present to help bind heme groups to substrates, we
investigated whether it was essential for binding the heme
iron to AOP2. Through immunoprecipitation studies, we
found that heme preincubated with cysteine did not bind
to AOP2, whereas heme preincubated with several other
unrelated amino acids freely bound to AOP2. Another
member of the peroxiredoxin family also uses cysteine for
binding to heme [34].
We addressed the functional importance of AOP2 in
erythrocytes and found that, as can the well-known
superoxide dismutase and catalase antioxidants, it was
able to prevent hemoglobin oxidation (both induced and
spontaneous). Hemoglobin is oxidized only at its iron
atoms, which are sheltered in its hydrophobic pocket. Our
Table 1. AOP2 prevents spontaneous as well as induced methemoglobin formation. The indicated reagents were added to aliquots of a freshly
prepared hemoglobin solution. After 48 and 120 hours respectively at 25 °C, absorbency readings were recorded and the 570/626 nm ratio
calculated. AOP2 was added at a molar ratio of 1 : 1060 (AOP2 (2 lgÆml
)1
) + ascorbic acid (5 m
M

)) and 1 : 303 (AOP2 (7 lgÆml
)1
)+ascorbic
acid (5 m
M
)) respectively.
Reagent added to Hb solution Ratio 570/626 nm (48 h) Ratio 570/626 nm (120 h)
None 7.6 6.0
Ascorbic acid (5 m
M
) 4.1 3.7
AOP2 (2 lgÆmL
)1
) 8.7 6.7
AOP2 (7 lgÆmL
)1
) 10.9 9.1
AOP2 (2 lgÆmL
)1
) + ascorbic acid (5 m
M
) 6.6 6.2
AOP2 (7 lgÆmL
)1
) + ascorbic acid (5 m
M
) 7.5 8.3
Fig. 6. Methemoglobin formation is prevented by AOP2. AOP2 pre-
vented metHb from forming spontaneously over a 72-h period. MetHb
(peak absorbance around 622 nm) is nearly absent from fresh hemo-

globin. Hemoglobin (peak absorbance at 576 nm) is partially oxidized
over 72 h and oxidized more by ascorbic acid. AOP2 prevents Hb from
being spontaneously oxidized to metHb.
Ó FEBS 2003 Distribution and function of antioxidant protein 2 (Eur. J. Biochem. 270) 339
immunoprecipitation studies indicated that AOP2 bound
with high affinity to the heme complex, supporting the
hypothesis that it protects or restores hemoglobin to its
active form. This hypothesis has yet to be confirmed.
Because metHb is so undesirable, organisms need a
redundant system to reduce its formation. RBCs, lacking
the ability to replace damaged molecules, depend more
than do other cells on efficient detoxifying systems: a lack
of protective enzymes in RBCs would result in a surge of
oxidation. Abnormal metHb formation has not been
found in human acatalasemics [1], a further indicator
that other molecules such as AOP2 might compensate
for the missing catalase in these patients. AOP2 may well
be a major antioxidant in RBCs, and its role in other cells
is being investigated in our laboratories. Heme is the
prosthetic group of several proteins and enzymes (myo-
globin, cytochrome c, cytochrome P450, ubiquinol-cyto-
chrome c reductase, cytochrome c oxidase, tryptophan
pyrrolase, and NO synthase), and AOP2 may protect them
also. This would explain its nearly ubiquitous presence and
provide further evidence for its importance as a major
protective protein.
Acknowledgements
The authors thank Ray Lambert for his editing skills. This work
was supported in part by a grant from AstraZeneca, by the
Austrian Ministry of Education, Science and Culture, Austrian

Ministry of Social Security and Generations GZ.236.065/6-VI/B/10/01
(GZ:236.065/7-VIII/A/6/00), and the City of Vienna.
References
1. Halliwell, B. & Gutteridge, J.M. (1998) Free Radicals in Biology
and Medicine, 3rd edn. Oxford University Press Inc, New York,
USA.
2. Denicola, A., Souza, J.M. & Radi, R. (1998) Diffusion of peroxy-
nitrite across erythrocyte membranes. Proc.NatlAcad.Sci.USA
95, 3566–3571.
3. Lynch, R.E. & Fridovich, I. (1978) Permeation of the erythrocyte
stroma by superoxide radical. J. Biol. Chem. 253, 4697–4699.
4. Matsuoka, I., Fujino, Y., Ogita, S. & Inoue, M. (1995) Impact of
erythrocytes on mouse embryonal development in vitro. FEBS
Lett. 371, 297–299.
5. Cha, M.K., Kim, H.K. & Kim, I.H. (1995) Thioredoxin-linked
Ôthiol peroxidaseÕ from periplasmic space of Escherichia coli.
J. Biol. Chem. 270, 28635–28641.
6. Cha, M.K., Yun, C.H. & Kim, I.H. (2000) Interaction of human
thiol-specific antioxidant protein 1 with erythrocyte plasma
membrane. Biochemistry 39, 6944–6950.
7. Kim, K., Kim, I.H., Lee, K.Y., Rhee, S.G. & Stadtman, E.R.
(1988) The isolation and purification of a specific ÔprotectorÕ
protein which inhibits enzyme inactivation by a thiol/Fe
(III)/O
2
mixed-function oxidation system. J. Biol. Chem. 263,
4704–4711.
8. Chae, H.Z. & Rhee, S.G. (1994) A thiol-specific antioxidant and
sequence homology to various proteins of unknown function.
Biofactors 4, 177–180.

9. McGonigle, S., Dalton, J.P. & James, E.R. (1998) Peroxidoxins: a
new antioxidant family. Parasitol. Today 14, 139–145.
10. Hofmann, B., Hecht, H.J. & Flohe, L. (2002) Peroxiredoxins. Biol.
Chem. 383, 347–364.
11. Chae, H.Z., Kim, H.J., Kang, S.W. & Rhee, S.G. (1999) Char-
acterization of three isoforms of mammalian peroxiredoxin that
reduce peroxides in the presence of thioredoxin. Diabetes Res.
Clin. Pract. 45, 101–112.
12. Chae, H.Z., Robison, K., Poole, L.B., Church, G., Storz, G. &
Rhee, S.G. (1994) Cloning and sequencing of thiol-specific anti-
oxidant from mammalian brain: alkyl hydroperoxide reductase
and thiol-specific antioxidant define a large family of antioxidant
enzymes. Proc. Natl Acad. Sci. USA 91, 7017–7021.
13. Bryk, R., Griffin, P. & Nathan, C. (2000) Peroxynitrite reductase
activity of bacterial peroxiredoxins. Nature 407, 211–215.
14. Barr, S.D. & Gedamu, L. (2001) Cloning and characterization of
three differentially expressed peroxidoxin genes from Leishmania
chagasi. Evidence for an enzymatic detoxification of hydroxyl
radicals. J. Biol. Chem. 276, 34279–34287.
15. Peshenko, I.V. & Shichi, H. (2001) Oxidation of active center
cysteine of bovine 1-Cys peroxiredoxin to the cysteine sulfenic acid
form by peroxide and peroxynitrite. Free Radic. Biol. Med. 31,
292–303.
16. Seo,M.S.,Kang,S.W.,Kim,K.,Baines,I.C.,Lee,T.H.&Rhee,
S.G. (2000) Identification of a new type of mammalian peroxi-
redoxin that forms an intramolecular disulfide as a reaction
intermediate. J. Biol. Chem. 275, 20346–20354.
17. Zhou, Y., Kok, K.H., Chun, A.C., Wong, C.M., Wu, H.W., Lin,
M.C., Fung, P.C., Kung, H. & Jin, D.Y. (2000) Mouse peroxir-
edoxin V is a thioredoxin peroxidase that inhibits p53-induced

apoptosis. Biochem. Biophys. Res. Commun. 268, 921–927.
18. Iakoubova, O.A., Pacella, L.A., Her, H. & Beier, D.R. (1997)
LTW4 protein on mouse chromosome 1 is a member of a family of
antioxidant proteins. Genomics 42, 474–478.
19. Phelan, S.A., Johnson, K.A., Beier, D.R. & Paigen, B. (1998)
Characterization of the murine gene encoding Aop2 (antioxidant
protein 2) and identification of two highly related genes. Genomics
54, 132–139.
20. Declercq, J.P., Evrard, C., Clippe, A., Stricht, D.V., Bernard, A. &
Knoops, B. (2001) Crystal structure of human peroxiredoxin 5, a
novel type of mammalian peroxiredoxin at 1.5 A resolution.
J. Mol. Biol. 311, 751–759.
21. Phelan, S.A. (1999) AOP2 (Antioxidant Protein 2): Structure and
function. Antiox. Redox. Signal. 1, 1–14.
22. Kim, T.S., Dodia, C., Chen, X., Hennigan, B.B., Jain, M.,
Feinstein, S.I. & Fisher, A.B. (1998) Cloning and expression of rat
lung acidic Ca(2+) -independent PLA2 and its organ distribution.
Am. J. Physiol. 274, L750–L761.
23. Kang, S.W., Baines, I.C. & Rhee, S.G. (1998) Characterization of
a mammalian peroxiredoxin that contains one conserved cysteine.
J. Biol. Chem. 273, 6303–6311.
24. Stuhlmeier,K.M.,Csizmadia,V.,Cheng,Q.,Winkler,H.&Bach,
F.H. (1994) Selective inhibition of E-selectin, ICAM-1, and
VCAM in endothelial cells. Eur. J. Immunol. 24, 2186–2190.
25. Dyer, R.B. & Herzog, N.K. (1995) Isolation of intact nuclei for
nuclear extract preparation from a fragile B-lymphocyte cell line.
Biotechniques 19, 192–195.
26. Chae, H.Z., Kim, I.H., Kim, K. & Rhee, S.G. (1993) Cloning,
sequencing, and mutation of thiol-specific antioxidant gene of
Saccharomyces cerevisiae. JBiolChem.268, 16815–16821.

27. Carr, A. & Frei, B. (1999) Does vitamin C act as a pro-oxidant
under physiological conditions? Faseb J. 13, 1007–1024.
28. Salvati, A.M. & Tentori, L. (1981) Determination of aberrant
hemoglobin derivatives in human blood. In Methods in
Enzymology (Colowick, S.P. & Kaplan, N.O., eds), pp. 715–731.
Academic Press Inc., New York, USA.
29. Mazor, D., Golan, E., Philip, V., Katz, M., Jafe, A., Ben-Zvi, Z. &
Meyerstein,N.(1996)Redbloodcellpermeabilitytothiolcom-
pounds following oxidative stress. Eur. J. Haematol. 57, 241–246.
30. Munz, B., Frank, S., Hubner, G., Olsen, E. & Werner, S. (1997) A
novel type of glutathione peroxidase: expression and regulation
during wound repair. Biochem. J. 326, 579–585.
340 K. M. Stuhlmeier et al.(Eur. J. Biochem. 270) Ó FEBS 2003
31. Frank, S., Munz, B. & Werner, S. (1997) The human homologue
of a bovine non-selenium glutathione peroxidase is a novel kera-
tinocyte growth factor-regulated gene. Oncogene 14, 915–921.
32. Lim, Y.S., Cha, M.K., Yun, C.H., Kim, H.K., Kim, K. & Kim,
I.H. (1994) Purification and characterization of thiol-specific
antioxidant protein from human red blood cell: a new type of
antioxidant protein. Biochem. Biophys. Res. Commun. 199, 199–206.
33. Chae, H.Z., Uhm, T.B. & Rhee, S.G. (1994) Dimerization of thiol-
specific antioxidant and the essential role of cysteine 47. Proc. Natl
Acad. Sci. USA 91, 7022–7026.
34. Hirotsu,S.,Abe,Y.,Nagahara,N.,Hori,H.,Nishino,T.,Okada,
K. & Hakoshima, T. (1999) Crystallographic characterization of a
stress-induced multifunctional protein, rat HBP-23. J. Struct. Biol.
126, 80–83.
Ó FEBS 2003 Distribution and function of antioxidant protein 2 (Eur. J. Biochem. 270) 341