Báo cáo khoa học: A superoxide dismutase–human hemoglobin fusion protein showing enhanced antioxidative properties pptx

A superoxide dismutase–human hemoglobin fusion
protein showing enhanced antioxidative properties
Marie Grey
1
, Sakda Yainoy
1,2
, Virapong Prachayasittikul
2
and Leif Bu
¨
low
1
1 Department of Pure and Applied Biochemistry, Centre for Chemistry and Chemical Engineering, Lund University, Sweden
2 Department of Clinical Microbiology, Faculty of Medical Technology, Mahidol University, Bangkok, Thailand
Introduction
The toxicity of Hb outside of its natural protective red
blood cell environment has been linked partly to the
redox activity of the Hb molecule. Consequently, Hb
can react with, and generate, reactive oxygen species
such as the superoxide anion. Hb, both inside and out-
side of the red blood cell, readily undergoes autoxida-
tion, but the degree of oxidation is normally restricted
to approximately 3% of the total Hb by the metHb
reductase system. However, if damage occurs to the red
blood cells, Hb may be released and, as the normal pro-
tection systems, involving also superoxide dismutase
(SOD) and catalase, are no longer associated with the
Hb, Hb is exposed to oxidative damage. Autoxidation
leads to nonfunctional metHb (HbFe
3+
) and superox-

ide ions (Eqns 1–4). Furthermore, reoxygenation of a
tissue previously deprived of oxygen will produce super-
oxide, a phenomenon known as reperfusion injury [1,2].
HbFe
2þ
O
2
þ H
2
O
2
! HbFe
4þ
@O þ H
2
O þ O
2
ð1Þ
HbFe
2þ
O
2
! HbFe
3þ
þ O
À
2
ð2Þ
HbFe
3þ

þ H
2
O
2
!

HbFe
4þ
@ O þ H
2
O ð3Þ
HbFe
4þ
@ O þ H
2
O
2
! HbFe
3þ
þ O
À
2
þ H
2
O !
heme degradation ð4Þ
Superoxide can be reduced either enzymatically by
SOD, or spontaneously by dismutation to H
2
O

2
.
Both superoxide and H
2
O
2
increase the rate of
Keywords
antioxidation; fusion protein; hemoglobin;
superoxide dismutase
Correspondence
L. Bu
¨
low, Department of Pure and Applied
Biochemistry, Centre for Chemistry and
Chemical Engineering, Lund University,
P.O. Box 124, SE-211 00, Lund, Sweden
Fax: +46 46 222 4611
Tel: +46 46 222 9594
E-mail:
(Received 13 May 2009, revised 27 July
2009, accepted 24 August 2009)
doi:10.1111/j.1742-4658.2009.07323.x
Much of the toxicity of Hb has been linked to its redox activity; Hb may
generate reactive oxygen species, such as the superoxide anion. Superoxide
is intrinsically toxic, and superoxide dismutase (SOD) provides important
cellular protection. However, if the Hb molecule is located outside the red
blood cell, the normal protection systems involving SOD and catalase are
no longer closely associated with it, exposing Hb and its cellular surround-
ings to oxidative damage. In order to produce less toxic Hb molecules, we

have explored gene fusion to obtain homogeneous SOD–Hb conjugates.
The chimeric protein was generated by coexpressing the human Hb
a-chain ⁄ manganese SOD gene together with the b-chain gene in Escherichia
coli. We show that the engineered SOD–Hb fusion protein retains the
oxygen-binding capacity and, moreover, decreases cytotoxic ferrylHb
(HbFe
4+
) formation when challenged with superoxide radicals. The SOD–
Hb fusion protein also exhibits a 44% lower autoxidation rate and higher
thermal stability than Hb alone.
Abbreviations
DSC, differential scanning calorimetry; ROS, reactive oxygen species; SOD, superoxide dismutase.
FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS 6195
metHb formation [3] and eventually lead to heme
degradation. The catalase present intracellularly may
be sufficient to remove H
2
O
2
by conversion to water
[4]; however, there is not sufficient SOD to eliminate
the superoxide. Chang [2], Alagic [4] and Tarasov [5]
have characterized Hb chemically conjugated to SOD
alone or SOD combined with catalase, i.e. Hb–SOD
and polyHb–SOD–catalase, respectively. The latter
Hb conjugate has been shown to have antioxidative
properties and offer protection against reperfusion
injury [6].
In order to create proximity between SOD and Hb,
we consequently explored gene fusion. Such closeness

generates a favorable microenvironment in which the
formed superoxide ions can be directly taken care of.
The SOD–Hb protein was prepared by linking the
human Hb a-chain gene with the human manganese
SOD gene. Manganese SOD was chosen because it
shows lower product inhibition by H
2
O
2
and has a
much longer half-life in serum, 5–6 h as compared to
6–10 min for copper and zinc SOD [7,8]. This
approach enabled us to produce the SOD–Hb in
Escherichia coli and to obtain a homogeneous product.
We characterized this novel enzyme by analyzing the
stability, activity and antioxidative properties of the
fused SOD–Hb, and we then compared it with human
Hb, also produced in E. coli.
Results
Protein expression and purification
It is notoriously difficult to produce human Hb in
microorganisms, because the a-chains and b-chains
must be produced in equal amounts to give functional
Hb molecules. The a-chains are especially prone to
precipitation and degradation if expression is unbal-
anced [9]. By linking the SOD gene to the a-chain
gene, we could stabilize fusion protein expression. The
proteins were purified using affinity chromatography;
details can be found in Doc. S1. After optimization of
the expression protocol, SOD–Hb was expressed well,

with a yield of 8 mgÆL
)1
, which is four times higher
than the yields obtained with other Hb fusion proteins
[10]. The expression level was lower for SOD–Hb than
for wild-type Hb, but may be further optimized by
introducing compensatory mutations to increase
expression and improve heme transport into the cell
[11]. Denaturation SDS ⁄ PAGE analysis (Fig. 1)
showed bands of the expected sizes: 16 kDa for
b-chain Hb, and 38 kDa for SOD–a-chain Hb. In
addition, SOD–Hb showed the characteristic red color
of Hb.
Molecular mass determination
Molecular masses were determined by gel filtration on
a Superdex 200 prep grade column, which gave four
components. Two of these, corresponding to almost
50% of the peak area (Fig. 2), were complexes larger
than 160 kDa. The remaining peaks corresponded to
SOD–Hb in the monomer and dimer forms at equal
concentrations (where the Hb b-chain is included;
Fig. 2).
Activity measurements
The SOD activity assay showed a SOD activity of
1.3 · 10
5
UÆlmol
)1
for SOD–Hb, which means that
the fusion protein retained more than half (52%)

of the native SOD activity. This corresponds closely
to the amount of high molecular mass SOD–Hb. As
the native SOD is functional as a tetramer, it is possi-
ble that the dimer and monomer fractions of SOD–Hb
exhibit much lower activity, or may even be nonfunc-
tional [7,12].
In the CO and O
2
binding assays, the protein sam-
ples were reduced with sodium dithionite and then
gently bubbled with CO or O
2
(Fig. 3). SOD–Hb
showed the typical absorption spectrum characteristics
of human Hb, with peaks located at wavelengths of
417, 536 and 566 nm, which are almost identical to
those of HbCO (417, 537 and 567 nm). An O
2
spec-
trum was also recorded, and SOD–Hb again showed
the typical peaks at 414, 538 and 573 nm, similar to
HbO
2
(413, 539 and 574 nm). In particular, the O
2
peaks of the fusion protein in the visible region
appear to be slightly broader than those of Hb, and
there also appears to be a minor peak at 630 nm on
the CO spectrum. These could be explained by very
low amounts of ferric protein in both the SOD–Hb

spectra [13].
Fig. 1. The results of SDS ⁄ PAGE on crude extract and purified
sample. Lane 1: molecular mass marker. Lane 2: crude extract of
SOD–Hb. Lane 3: purified SOD–Hb.
An Hb fusion protein with antioxidative properties M. Grey et al.
6196 FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS
Autoxidation studies were performed at room tem-
perature (20–22 °C) for 48 h, using an Hb concentration
of 8 lm, and followed using visible spectrophotometry.
The rate constant for Hb (0.18 h
)1
) was almost twice as
high as that for SOD–Hb (0.10 h
)1
).
Stability
Thermal stability was investigated using differential
scanning calorimetry (DSC). The asymmetric shape of
the DSC curve (Fig. 4) suggests a complex denatur-
ation path, which is irreversible under the present con-
ditions. Thermal denaturation of Hb begins with the
dissociation of the tetramer into monomers [14], and
ends with a certain degree of aggregation. The
Fig. 2. Schematic representation of SOD–Hb
showing possible conformations resulting
in (from top to bottom): the high molecular
mass complex, an SOD–Hb dimer, and finally
an SOD–Hb monomer.
A
B

Fig. 3. Reduced SOD–Hb and Hb were bubbled with O
2
(A) or CO
(B). SOD–Hb shows the absorption spectrum characteristics typical
of human Hb for both O
2
and CO. Critical regions of the protein
spectra have been enlarged. The scale on the y-axis is offset for
clarity.
Fig. 4. DSC was performed on 60 lM Hb (dashed line) or 30 lM
SOD–Hb (solid line). Cp, heat capacity.
M. Grey et al. An Hb fusion protein with antioxidative properties
FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS 6197
broader, less well-defined peak of SOD–Hb indicates
contributions from more than one thermal process,
and thus a more intricate mixture of complexes (e.g.
higher-order ‘aggregates’), in agreement with the
results of the gel filtration experiments showing several
species. In addition, both Hb and SOD are normally
tetramers, which could account for the numerous ther-
mal processes as they dissociate into dimers and mono-
mers. In addition to dissociation into dimers, the
complexity of the DSC curve is increased by the ther-
mal inactivation of the SOD enzymatic activity, result-
ing in smaller conformational changes [15], which
probably contribute to the broadness of the peak.
Apparent T
m
values were calculated using the Microcal
software origin, and found to be 48.7 °C (Hb) and

55.1 °C (SOD–Hb), corresponding to the T
m
value of
53.2 °C reported by Olsen [16] (HbACO). SOD–Hb
thus exhibits greater heat stability than Hb, although
lower stability than SOD alone (T
m
of 68–94 °C)
[12,17].
Heme reactivity
Heme loss is dependent on the geometry of the protein
and, possibly, also water shielding. Oxidation of the
heme to the met form (Fe
3+
) increases the probability
of heme loss as the fifth coordination bond to the
proximal histidine is weakened [18]. The fast phase has
been associated with heme loss from the b-chains,
whereas the slow phase is attributed to the a-chains
[19]. SOD–Hb showed a heme loss rate of
0.19 ± 0.05 min
)1
for the fast phase, whereas the loss
rate of Hb was almost three times lower
(0.069 ± 0.008 min
)1
). Although the b-chains do not
take part directly in the fusion, the effect is more
profound here, suggesting that the b-chains are less
protected during fusion than Hb. As expected, for the

slow phase both Hb and SOD–Hb a-chains exhi-
bited lower heme loss rates (0.011 ± 0.001 and
0.017 ± 0.003 min
)1
, respectively), indicative of higher
heme affinity.
SOD–Hb and Hb were also incubated together with
a xanthine ⁄ xanthine oxidase superoxide-generating sys-
tem. After 30 min of incubation, sodium sulfide, which
reacts with ferrylHb to form sulfHb, was added. As
shown in Fig. 5, SOD–Hb was significantly more effec-
tive in reducing the amount of ferrylHb formed
(P = 0.0018) than Hb.
Heme degradation of Hb and SOD–Hb by H
2
O
2
was measured by monitoring the fluorescence of the
degradation products (Fig. 6). For all H
2
O
2
concentra-
tions used (1, 2 and 4 mm), the amount of heme degra-
dation for SOD–Hb was much lower than for Hb
alone. For both proteins, increasing the H
2
O
2
concen-

tration increased the denaturation, as indicated by flu-
orescence measurements. Blank measurements without
H
2
O
2
gave no contribution to the fluorescence signal
(data not shown).
Discussion
The proximity between two proteins is often impor-
tant. For instance, the physical closeness between
two or several enzymes catalyzing sequential reac-
tions is a feature of many metabolic pathways. The
frequently observed improved kinetic behavior of
such associated proteins has been explained by the
formation of a favorable ‘microenvironment’ in
Fig. 5. FerrylHb formation was measured with 10 lM Hb or SOD–
Hb. *Significantly different from Hb (P = 0.0018, Student’s t-test).
Fig. 6. The formation of fluorescent heme degradation products
was measured at a fixed concentration of oxyHb (15 l
M) and vary-
ing concentrations of H
2
O
2
: e, Hb with 4 mM H
2
O
2
; D, Hb with

2m
M H
2
O
2
; h, Hb with 1 mM H
2
O
2
; ¤, SOD–Hb with 4 mM
H
2
O
2
; , SOD–Hb with 2 mM H
2
O
2
; , SOD–Hb with 1 mM H
2
O
2
.
An Hb fusion protein with antioxidative properties M. Grey et al.
6198 FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS
which the local concentration of intermediates is
higher. Similarly, a second protein in an aggregate
can provide protection against a toxic compound or
intermediate.
There are several ways of creating proximity

between proteins. A frequently used technique involves
chemical cross-linking of the relevant proteins. This
normally generates protein conjugates with random
linking and random orientation of the active centers.
Proximity between two or more proteins may also be
achieved by performing in-frame gene fusion of the
corresponding structural genes. The chimeric gene then
encodes a polypeptide chain carrying two or more
active centers. This strategy resembles the proposed
model for the evolution of naturally occurring multi-
functional proteins [20]. Through comparison of DNA
sequences from different species, it has been demon-
strated that such proteins probably evolved through
gene translocation followed by gene fusion. Extensive
studies of such naturally occurring protein aggregates
have enabled the three-dimensional structures to be
determined for some of them, notably phosphoribosyl
anthranilate isomerase ⁄ indole glycerol phosphate syn-
thase and tryptophan synthase. Besides generating
homogeneous conjugates, this method also allows for
in vivo testing of the fusion proteins, as opposed to
chemical approaches.
Superoxide is intrinsically toxic, and SOD provides
important cellular protection. Superoxide radicals
and ⁄ or impaired SOD functionality have been impli-
cated in diabetes, cancer and neurodegenerative dis-
eases, in addition to cellular damage such as lipid
peroxidation, protein oxidation, and DNA damage
[21,22]. Hb may generate reactive oxygen species,
which can be particularly deleterious in clinical appli-

cations, e.g. in blood substitutes. When a gene fusion
approach is used, the link between the SOD and Hb
is always present between the same amino acid resi-
dues, which precludes heterogeneity. Conversely, the
chemically linked Hb–SOD products described by
Alagic [4] and Tarasov [5] have cross-links introduced
at different residues, as well as reactive groups, which
must be blocked in a separate step. In these chemical
linking approaches, usually both intra-tetramer and
inter-tetramer bonds are introduced, with the involve-
ment of multiple side chains [23]. In addition, it has
been shown that shorter cross-links may provide
better stabilized Hb [24]. The SOD–Hb was designed
with a very short linker of only one alanine, in order
to promote proximity, as well as to reduce proteo-
lytic degradation during production. However, our
SOD–Hb shows the presence of different aggregation
forms, with large complexes as well as dimers and
monomers. The gel filtration step described could,
however, be used to remove these fractions. As it is
likely that the monomer and dimer forms are inac-
tive, this simple step could increase the homogeneity
of the product as well as removing the less active
fractions.
SOD–Hb retains about half of the specific activity
of native SOD. This lower activity may be due to the
steric difficulties encountered in forming the necessary
SOD tetramers [7,12]. The tetrameric conformation is
important for the formation of an active site suitable
for Mn ligation [12]. An alternative strategy could be

to coexpress native SOD with our fusion protein,
which may facilitate tetramer formation. Similarly, we
can coexpress native Hb a-chains together with SOD–
Hb to generate a free N-terminal end of this polypep-
tide chain. It is therefore essential to consider subunit
interactions when engineering fusion proteins, particu-
larly when producing larger Hb conjugates. As indi-
cated earlier, catalase is a key component when
generating functional Hb-based blood substitutes. We
have prepared an SOD–catalase fusion protein that
can be coexpressed with SOD–Hb in E. coli.By
exploring the natural SOD subunit interactions, we
can then form a SOD–catalase–Hb protein. The poly-
Hb–SOD–catalase developed by Chang [2] also showed
lower activity, 85–90% of that of native SOD, depend-
ing on the conjugation ratio, implying that this prepa-
ration also suffers from steric hindrance. However, the
chemically linked Hb–SOD [4] showed identical activ-
ity for both conjugated and free SOD, probably owing
to locking of the SOD in the tetrameric form during
conjugation.
The genetic fusion of Hb and SOD does not seem to
perturb the globin structure noticeably, as both the O
2
and CO spectra are essentially identical to those of
Hb. All peaks in the Soret region correspond well to
that of Hb, indicative of full functionality. The autoxi-
dation rate constants were determined to be 0.18 h
)1
for Hb and 0.10 h

)1
for SOD–Hb at room tempera-
ture. The rate constant for SOD–Hb is thus approxi-
mately half of that for Hb, although higher than the
values reported by Vandegriff (0.007–0.0021 h
)1
) [18].
This may partly be explained by the residual catalase
activity, as the Hb in their study was derived from
outdated donated blood.
An in vitro test was performed to evaluate the
antioxidative properties of SOD–Hb. SOD–Hb and
Hb were incubated together with a superoxide ⁄ H
2
O
2
-
generating system. The formation of ferrylhemoglobin
with SOD–Hb after 30 min was significantly lower
(P = 0.0018) than with Hb (Fig. 5). FerrylHb is
cytotoxic and has been implicated in oxidative stress
M. Grey et al. An Hb fusion protein with antioxidative properties
FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS 6199
situations in a range of diseases [25,26]. The ability to
reduce the formation of this toxic species means that
SOD–Hb has considerable promise for the future
development of a clinical product. Interestingly, extra-
cellular Hb polymers with SOD activity, with a molec-
ular mass of over 3800 kDa, occur naturally in the
blood of earthworms [32,33].

Heme degradation of Hb and SOD–Hb by H
2
O
2
was measured by monitoring the fluorescence of the
degradation products. The oxyHb first reacts with one
molecule of H
2
O
2
, forming oxyferrylHb (HbFe
4+
=O).
When a second molecule of H
2
O
2
reacts with the oxy-
ferrylHb, metHb and a superoxide radical (Eqns 2–4)
are produced, initiating the degradation process. The
damage is irreversible, and leads to a cascade of reac-
tions that ultimately result in iron release and fluores-
cent degradation products. The superoxide radical,
which has a lifetime of 0.2 s [27], is perfectly located to
react with the heme group. Moreover, the superoxide
exhibits higher reactivity, owing to the heme pocket
environment [28,29]. As can be seen in Fig. 6, heme
degradation in SOD–Hb was much lower than in Hb,
and both showed an increase in denaturation when the
H

2
O
2
concentration was increased. The combination
of SOD and Hb thus protects the heme molecule from
denaturation during oxidative stress in vitro, probably
via a proximity effect.
The heme degradation is dependent on the lifetime
of ferrylHb and the susceptibility of the heme to super-
oxide-induced damage. The close association of SOD
with Hb can allow removal of the superoxide radicals
quickly, leading to less degradation, as shown in this
study. Additional modifications of SOD–Hb, e.g. by
introducing mutations affecting ferryl reduction kinet-
ics, may further reduce the Hb toxicity. We have previ-
ously shown that either introducing or removing
suitably located tyrosines affects ferryl reduction kinet-
ics in human Hb [26]. Introducing the same mutations
in SOD–Hb could result in a SOD–Hb molecule that
is even more suitable for practical use. Additionally, to
allow further in vivo studies of our construct, suitable
protein surface protection, such as pegylation or
encapsulation, needs to be developed.
In conclusion, the engineered SOD–Hb exhibits a
lower autoxidation rate and higher thermal stability
than Hb alone, and the process creates a homogeneous
link between the Hb and the SOD, which chemical
conjugation does not. Additionally, in vitro tests show
that cytotoxic ferrylHb formation is significantly
decreased in the presence of superoxide radicals. Con-

sequently, the combination of SOD and Hb protects
the Hb molecule from denaturation during oxidative
stress.
Experimental procedures
Construction of human manganese SOD and
human SOD–Hb
Please see the Supporting information. Primers used for
site-directed mutagenesis or cloning can be found in
Table S1.
Protein expression and purification
Protein expression and purification were performed essen-
tially as described previously [26]. Details can be found in
Doc. S1.
Molecular mass determination
Gel filtration chromatography was used for molecular mass
determination on an A
¨
KTA purifier system controlled by
unicorn software (GE Healthcare, Uppsala, Sweden).
Highly purified protein was loaded on a HiLoad 16 ⁄ 60
Superdex 200 column (also GE Healthcare), equilibrated
with 50 mm phosphate buffer (pH 7.2) containing 0.15 m
EDTA, and eluted with the same buffer at a flow rate of
1.0 mLÆmin
)1
. Standard proteins with molecular masses
ranging from 6.5 to 158.0 kDa (GE Healthcare) were used
to produce standard curves.
SOD activity assay
The assay used to determine SOD activity was a slightly

modified version of that described by Ewing [30]. In prin-
ciple, the assay is based on the ability of SOD to inhibit
nitroblue tetrazolium reduction by an aerobic mixture of
NADH and prenazine methosulfate, which produces
superoxide at nonacidic pH. Details can be found in
Doc. S1.
DSC
DSC measurements were performed with a Microcal differ-
ential scanning calorimeter (Microcal, Northampton, MA,
USA) with a cell volume of 0.5072 mL. All samples were
degassed for 15 min at room temperature prior to scanning.
Baseline scans were obtained with buffer in both the refer-
ence and sample cells, and these were later subtracted from
sample scans. Protein samples (60 lm Hb or 30 lm SOD–
Hb) in 70 mm sodium phosphate (pH 7.2) were scanned in
the temperature range 20–90 °C at a rate of 60 °CÆh
)1
.
Heme loss rates
The heme exchange rate between metHb and human serum
albumin was determined as described by Benesch [19] and
An Hb fusion protein with antioxidative properties M. Grey et al.
6200 FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS
Jeong [31], with slight modifications. Potassium ferricyanide
was added in excess in order to oxidize the proteins to the
ferric form, and this was followed by removal of ferricya-
nide and ferrocyanide with a Sephadex G-25 column (GE
Healthcare) equilibrated with 0.05 m bis-Tris and 0.1 m
NaCl (pH 7.5). Each Hb sample was then mixed with
human serum albumin (5 lm final concentration of each),

and 0.5 m Tris buffer (pH 9.05) was added to a total vol-
ume of 2 mL. The absorbance (A) at 578 and 620 nm was
then recorded every minute for 90 min, using a Beckman
Coulter DU-800 spectrophotometer. The amounts of
metHb and methemalbumin were calculated using the
following two equations:
metHb½¼146:03 Â A
578
À 134:48 Â A
620
and
methemalbumin½¼À61:95  A
578
þ 220:01 Â A
620
The data were then fitted to a double-exponential decay
curve, essentially as described by Vandegriff [18]:
Y ¼ A Â e
Àk
fast
Ât
þ e
Àk
slow
Ât
ÀÁ
þ C:
Autoxidation
The autoxidation analysis was performed as described by
Jeong [31], but with a lower Hb concentration of 8 lm.CO

was removed by shining light on the sample, which was
kept on ice, while gently oxygenating the solution using a
stream of oxygen gas. The reaction was carried out in
0.1 m sodium phosphate (pH 7.0) at room temperature
(20–22 °C). Spectra from 400 to 700 nm were collected at
specific times, using a Beckman Coulter DU-800 spectro-
photometer. The baseline was adjusted by setting the absor-
bance at 700 nm to zero. The experimental data were then
fitted to a single-exponential equation of the form [18]:
Y ¼ DY
max
 1 À e
kt
ÀÁ
þ Y
0
where Y is the relative metHb concentration (%), DY
max
is
the total relative change in metHb at the end of the reac-
tion, k is the rate constant, t is time, and Y
0
is the relative
metHb concentration at t =0.
Measurement of ferrylHb formation
FerrylHb formation was measured using xanthine ⁄ xanthine
oxidase as the oxidation system, based on a method
described by D’Agnillo [2,3]. Ten micromolar Hb or
SOD–Hb in 70 mm sodium phosphate (pH 7.2) was reacted
with 100 l m xanthine and 10 mUÆmL

)1
xanthine oxidase,
both from Sigma-Aldrich (Stockholm, Sweden). The Hb
concentration was calculated on the basis of heme, using
the relation: e
523
= 7.12 mm
)1
Æcm
)1
[32]. At given times, an
excess of catalase (Roche, Mannheim, Germany) was added
to remove residual H
2
O
2
, after which 2 mm sodium sulfide
was added; finally, the absorbance at 620 nm (A
620 nm
) was
measured.
Measurement of heme degradation by
fluorescence
Fluorescence measurements originally described by
Nagababu [33] were slightly modified. OxyHb (15 lm)
was incubated with H
2
O
2
(1, 2 or 4 mm)in50mm potas-

sium phosphate buffer (pH 7.4) in a total volume of
2mL at 25°C. Reagent-grade H
2
O
2
(30% v ⁄ v) was
obtained from Sigma-Aldrich, and standardized using an
extinction coefficient of 72.8 m
)1
Æcm
)1
at 230 nm [34].
The fluorescence signal was recorded for 30 min at excita-
tion and emission wavelengths of 460 and 525 nm, respec-
tively, using a fluorimeter (PTI Photon Technology
International, London, Canada).
Acknowledgements
This study was supported by European Union Frame-
work VI Eurobloodsubstitutes project and the Staff
Development Project by the Ministry of Education,
Thailand (SY).
References
1 Alayash AI (2004) Oxygen therapeutics: can we tame
haemoglobin? Nat Rev Drug Discov 3, 152–159.
2 D’Agnillo F & Chang TMS (1998) Polyhemoglobin–
superoxide dismutase–catalase as a blood substitute
with antioxidant properties. Nat Biotechnol 16, 667–671.
3 D’Agnillo F & Chang TMS (1998) Absence of hemo-
protein-associated free radical events following oxidant
challenge of crosslinked hemoglobin-superoxide dismu-

tase catalase – absence of increased free radical genera-
tion following administration in a rabbit model of renal
ischemia and reperfusion. Free Radic Biol Med 24,
906–912.
4 Alagic A, Koprianiuk A & Kluger R (2005) Hemoglo-
bin–superoxide dismutase – chemical linkages that
create a dual-function protein. J Am Chem Soc 127,
8036–8043.
5 Tarasov E, Blaszak MM, LaMarre JM & Olsen KW
(2007) Synthesis of a hemoglobin polymer containing
antioxidant enzymes using complementary chemistry
of maleimides and sulfhydryls. Artif Cells Blood Substit
Biotechnol 35, 31–43.
6 Powanda DD & Chang TM (2002) Cross-linked poly-
hemoglobin–superoxide dismutase–catalase supplies
oxygen without causing blood–brain barrier disruption
or brain edema in a rat model of transient global brain
M. Grey et al. An Hb fusion protein with antioxidative properties
FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS 6201
ischemia–reperfusion. Artif Cells Blood Substit Immobil
Biotechnol 30, 23–37.
7 Borgstahl GEO, Parge HE, Hickey MJ, Beyer WF,
Hallewell RA & Tainer JA (1992) The structure of
human mitochondrial manganese superoxide dismutase
reveals a novel tetrameric interface of two 4-helix
bundles. Cell 71, 107–118.
8 Gorecki MBY, Hartman JR, Fischer M, Weiss L,
Tochner Z, Slavin S & Nimrod A (1991) Recombinant
human superoxide dismutases: production and potential
therapeutical uses. Free Radic Res Commun 12–13, 401–

410.
9 Hoffman SJ, Looker DL, Roehrich JM, Cozart PE,
Durfee SL, Tedesco JL & Stetler GL (1990) Expres-
sion of fully functional tetrameric human hemoglobin
in Escherichia coli. Proc Natl Acad Sci USA 87, 8521–
8525.
10 Panetta G, Arcovito A, Morea V, Bellelli A & Miele
AE (2008) Hb([alpha][alpha],[beta][beta]): a novel
fusion construct for a dimeric, four-domain hemoglo-
bin. Biochim Biophys Acta Proteins Proteomics 1784,
1462–1470.
11 Graves PE, Henderson DP, Horstman MJ, Solomon BJ
& Olson JS (2008) Enhancing stability and expression
of recombinant human hemoglobin in E. coli: progress
in the development of a recombinant HBOC source.
Biochim Biophys Acta Proteins Proteomics 1784, 1471–
1479.
12 Borgstahl GEO, Parge HE, Hickey MJ, Johnson MJ,
Boissinot M, Hallewell RA, Lepock JR, Cabelli DE &
Tainer JA (1996) Human mitochondrial manganese
superoxide dismutase polymorphic variant Ile58Thr
reduces activity by destabilizing the tetrameric interface.
Biochemistry 35, 4287–4297.
13 Antonini E & Brunori M(1971) Hemoglobin and
Myoglobin in their Reactions with Ligands, Frontiers of
Biology, Vol. 21, North Holland Publishing Com-
pany,Amsterdam.
14 Michnik A, Drzazga Z, Kluczewska A & Michalik K
(2005) Differential scanning microcalorimetry study of
the thermal denaturation of haemoglobin. Biophys

Chem 118, 93–101.
15 Hearn AS, Fan L, Lepock JR, Luba JP, Greenleaf WB,
Cabelli DE, Tainer JA, Nick HS & Silverman DN
(2004) Amino acid substitution at the dimeric interface
of human manganese superoxide dismutase. J Biol
Chem 279, 5861–5866.
16 Olsen KW (1994) Thermal-denaturation procedures for
hemoglobin. Methods Enzymol 231, 514–524.
17 Quint PS, Domsic JF, Cabelli DE, McKenna R &
Silverman DN (2008) Role of a glutamate bridge
spanning the dimeric interface of human manganese
superoxide dismutase. Biochemistry 47, 4621–4628.
18 Vandegriff KD, Malavalli A, Minn C, Jiang E, Lohman
J, Young MA, Samaja M & Winslow RM (2006)
Oxidation and haem loss kinetics of poly(ethylene
glycol)-conjugated haemoglobin (MP4): dissociation
between in vitro and in vivo oxidation rates. Biochem J
399, 463–471.
19 Benesch RE (1994) The stability of the heme–globin
linkage – measurement of heme exchange. Methods
Enzymol 231, 496–502.
20 Chen B, Markillie LM, Xiong Y, Mayer MU & Squier
TC (2007) Increased catalytic efficiency following gene
fusion of bifunctional methionine sulfoxide reductase
enzymes from Shewanella oneidensis. Biochemistry 46,
14153–14161.
21 Culotta VC, Yang M & O’Halloran TV (2006) Activa-
tion of superoxide dismutases: putting the metal to the
pedal. Biochim Biophys Acta Mol Cell Res 1763, 747–
758.

22 McCord JM & Edeas MA (2005) SOD, oxidative stress
and human pathologies: a brief history and a future
vision. Biomed Pharmacother 59, 139–142.
23 Ronda L, Bruno S, Abbruzzetti S, Viappiani C &
Bettati S (2008) Ligand reactivity and allosteric regula-
tion of hemoglobin-based oxygen carriers. Biochim
Biophys Acta Proteins Proteomics 1784, 1365–1377.
24 Bobofchak KM, Tarasov E & Olsen KW (2008) Effect
of cross-linker length on the stability of hemoglobin.
Biochim Biophys Acta Proteins Proteomics 1784, 1410–
1414.
25 Alayash AI, Patel RP & Cashon RE (2001) Redox
reactions of hemoglobin and myoglobin: biological and
toxicological implications. Antioxid Redox Signal 3,
313–327.
26 Reeder BJ, Grey M, Silaghi-Dumitrescu R-L, Svistu-
nenko DA, Bulow L, Cooper CE & Wilson MT (2008)
Tyrosine residues as redox cofactors in human hemo-
globin: implications for engineering non toxic blood
substitutes. J Biol Chem 283, 30780–30787.
27 Zang LY & Misra HP (1992) EPR kinetic studies of
superoxide radicals generated during the autoxidation
of 1-methyl-4-phenyl-2,3-dihydropyridinium, a bioacti-
vated intermediate of parkinsonian-inducing neurotoxin
1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine. J Biol
Chem 267, 23601–23608.
28 Nagababu E & Rifkind JM (2000) Reaction of
hydrogen peroxide with ferrylhemoglobin: superoxide
production and heme degradation. Biochemistry 39,
12503–12511.

29 Jia Y & Alayash AI (2002) Stopped-flow fluorescence
method for the detection of heme degradation products
in solutions of chemically modified hemoglobins and
peroxide. Anal Biochem 308, 186–188.
30 Ewing JF & Janero DR (1995) Microplate superoxide
dismutase assay employing a nonenzymatic superoxide
generator. Anal Biochem 232, 243–248.
31 Jeong ST, Ho NT, Hendrich MP & Ho C (1999)
Recombinant hemoglobin(alpha 29leucine -> phenylal-
An Hb fusion protein with antioxidative properties M. Grey et al.
6202 FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS
anine, alpha 96valine -> tryptophan, beta 108aspara-
gine -> lysine) exhibits low oxygen affinity and high
cooperativity combined with resistance to autoxidation.
Biochemistry 38, 13433–13442.
32 Snell SM & Marini MA (1988) A convenient spectro-
scopic method for the estimation of hemoglobin concen-
trations in cell-free solutions. J Biochem Biophys
Methods 17, 25–34.
33 Nagababu E & Rifkind JM (1998) Formation of fluo-
rescent heme degradation products during the oxidation
of hemoglobin by hydrogen peroxide. Biochem Biophys
Res Commun 247, 592–596.
34 George P (1953) The chemical nature of the second
hydrogen peroxide compound formed by cytochrome c
peroxidase and horseradish peroxidase. I. Titration with
reducing agents. Biochem J 54, 267–276.
Supporting information
The following supplementary material is available:
Doc. S1. Experimental procedures.

Table S1. Primer sequences.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and
may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
M. Grey et al. An Hb fusion protein with antioxidative properties
FEBS Journal 276 (2009) 6195–6203 ª 2009 The Authors Journal compilation ª 2009 FEBS 6203