Tài liệu Báo cáo khoa học: Crystal structure of the cambialistic superoxide dismutase from Aeropyrum pernix K1 – insights into the enzyme mechanism and stability pdf

Crystal structure of the cambialistic superoxide dismutase
from Aeropyrum pernix K1 – insights into the enzyme
mechanism and stability
Tsutomu Nakamura
1
, Kasumi Torikai
1,2
, Koichi Uegaki
1
, Junji Morita
2
, Kodai Machida
3,4
,
Atsushi Suzuki
5
and Yasushi Kawata
3,4
1 National Institute of Advanced Industrial Science and Technology, Ikeda, Osaka, Japan
2 Department of Food Science and Nutrition, Faculty of Human Life and Science, Doshisha Women’s College of Liberal Arts, Kyoto, Japan
3 Department of Chemistry and Biotechnology, Graduate School of Engineering, Tottori University, Japan
4 Department of Biomedical Science, Institute of Regenerative Medicine and Biofunction, Graduate School of Medical Science,
Tottori University, Japan
5 Power Train Material Engineering Division, Toyota Motor Corporation, Aichi, Japan
Keywords
Aeropyrum pernix; cambialistic; metal
coordination; stability; superoxide dismutase
Correspondence
T. Nakamura, National Institute of Advanced
Industrial Science and Technology, 1-8-31
Midorigaoka, Ikeda, Osaka 563-8577, Japan

Fax: +81 72 751 8370
Tel: +81 72 751 9272
E-mail:
(Received 2 August 2010, revised 28
October 2010, accepted 2 December 2010)
doi:10.1111/j.1742-4658.2010.07977.x
Aeropyrum pernix K1, an aerobic hyperthermophilic archaeon, produces a
cambialistic superoxide dismutase that is active in the presence of either of
Mn or Fe. The crystal structures of the superoxide dismutase from A. per-
nix in the apo, Mn-bound and Fe-bound forms were determined at resolu-
tions of 1.56, 1.35 and 1.48 A
˚
, respectively. The overall structure consisted
of a compact homotetramer. Analytical ultracentrifugation was used to
confirm the tetrameric association in solution. In the Mn-bound form, the
metal was in trigonal bipyramidal coordination with five ligands: four side
chain atoms and a water oxygen. One aspartate and two histidine side
chains ligated to the central metal on the equatorial plane. In the Fe-bound
form, an additional water molecule was observed between the two histidines
on the equatorial plane and the metal was in octahedral coordination with
six ligands. The additional water occupied the postulated superoxide bind-
ing site. The thermal stability of the enzyme was compared with superoxide
dismutase from Thermus thermophilus, a thermophilic bacterium, which
contained fewer ion pairs. In aqueous solution, the stabilities of the two
enzymes were almost identical but, when the solution contained ethylene
glycol or ethanol, the A. pernix enzyme had significantly higher thermal sta-
bility than the enzyme from T. thermophilus. This suggests that dominant
ion pairs make A. pernix superoxide dismutase tolerant to organic media.
Database
Structural data have been deposited in the Protein Data Bank under the accession numbers

3AK1 (apo-form), 3AK2 (Mn-bound form) and 3AK3 (Fe-bound form)
Structured digital abstract
l
MINT-8075688: Superoxide dismutase (uniprotkb:Q9Y8H8) and Superoxide dismutase
(uniprotkb:
Q9Y8H8) bind (MI:0407)bycosedimentation in solution (MI:0028)
l
MINT-8075667: Superoxide dismutase (uniprotkb:Q9Y8H8) and Superoxide dismutase
(uniprotkb:
Q9Y8H8) bind (MI:0407)byx-ray crystallography (MI:0114)
l
MINT-8075678: Superoxide dismutase (uniprotkb:Q9Y8H8) and Superoxide dismutase
(uniprotkb:
Q9Y8H8) bind (MI:0407)bymolecular sieving (MI:0071)
Abbreviations
ApeSOD, superoxide dismutase from Aeropyrum pernix; SOD, superoxide dismutase; TthSOD, superoxide dismutase from
Thermus thermophilus.
598 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
Introduction
Superoxide dismutases (SODs; EC 1.15.1.1) play a pro-
tective role against oxidative stress by catalyzing dis-
proportionation of the superoxide anion radical (O
2
Æ
)
)
to hydrogen peroxide (H
2
O
2

) and dioxygen (O
2
). The
SOD-catalyzed reaction proceeds through a redox
cycle of metal ions as described by the equations [1]:
Enz À M
ðnþ1Þ
þO
2
Á
À
! Enz À M
nþ
þO
2
Enz À M
nþ
þO
2
Á
À
þ2H
þ
! Enz À M
ðnþ1Þ
þH
2
O
2
where Enz and M represent the enzyme and the metal

cofactor, respectively. SODs are grouped into four
classes according to their metal cofactors: copper and
zinc-containing SOD (Cu,Zn-SOD), iron-containing
SOD (Fe-SOD), manganese-containing SOD (Mn-
SOD) and nickel-containing SOD (Ni-SOD). These
four types of SOD are divided into three groups based
on amino acid sequence homology; Fe- and Mn-SODs
are homologous [2].
Although Mn-SOD and Fe-SOD are closely related
in amino acid sequence and tertiary structure, they are
generally active only in the presence of their specific
metals. For example, although the Fe-SOD and
Mn-SOD of Escherichia coli have 45% sequence identity
and can bind each other’s metals, they are inactive
when the wrong metal is incorporated at the active site
[3]. However, several SODs are active in the presence
of either Fe or Mn. These types of SODs are referred
as to cambialistic SODs. In addition to the tertiary
structures of metal-specific SODs [4], crystal structures
of several cambialistic SODs have been reported,
including those from Porphyromonas gingivalis [5] and
Propionibacterium shermanii [6]. The metal-specificity
of cambialistic SODs can be suppressed by mutagene-
sis at a site 11 A
˚
away from the reaction center, as
reported for the P. gingivalis SOD [7]. This supports
the hypothesis that cambialism is a consequence of
multiple factors rather than the result of a unique type
of active site structure [8].

Aeropyrum pernix K1 is a strictly aerobic hyper-
thermophilic archaeon [9,10] that has been the target
of several studies investigating primitive antioxidation
mechanisms in aerobic life [11–15]. A. pernix K1 pro-
duces a hyperthermophilic, cambialistic SOD [16] that
exhibits more activity when Mn, rather than Fe, is the
cofactor. Different metals affect not only the catalytic
activity itself, but also sensitivity to inhibitors such as
sodium azide, sodium fluoride and hydrogen peroxide:
The Fe-bound SOD of A. pernix K1 (ApeSOD) is
more sensitive to these inhibitors than Mn-bound
ApeSOD [16].
Despite significant characterizations of ApeSOD, the
enzymological features of this enzyme have not been
explained from a structural perspective because the ter-
tiary structure of ApeSOD has not been elucidated.
In the present study, for the first time, we describe the
crystal structure of ApeSOD. In particular, we focus
on the coordination of the metal cofactor in the active
site as well as the changes it experiences in response to
different metal cofactors. Finally, by comparing Ape-
SOD with the SOD from the thermophilic bacterium
Thermus thermophilus, we evaluate the relationship
between electrostatic interaction and the protein’s
stability in an organic medium.
Results
Protein preparation and incorporation
of metal ions
E. coli cells harboring the expression plasmid for Ape-
SOD were grown in LB medium and the enzyme was

purified in the presence of EDTA. An assay of the
ApeSOD preparation revealed that the enzyme (referred
to as the apo-enzyme) had low activity (Table 1). This
was attributed to the Mn or Fe ions incorporated into
the enzyme as it accumulated in the E. coli cells. Indeed,
the activity of the apo-enzyme was significantly lower
than the metal-containing enzyme.
Metal cofactors were incorporated into the enzyme;
when the growth medium contained MnSO
4
or FeSO
4
,
the activity of the obtained enzymes was 20-fold or
six-fold higher, respectively, than that of the apo-
enzyme (Table 1), indicating that the metals had suc-
cessfully been incorporated during bacterial expression.
When the metal cofactors were added to the purified
apo-enzyme and incubated at 70 °C, the enzyme
became significantly more active (Table 1). Incubation
with the metal at 37 °C did not raise the activity of
Table 1. SOD activity.
Enzyme
Activity
(unitÆmg
)1
)
a
Mn
(molÆmol

)1
)
Fe
(molÆmol
)1
)
Apo 27.5 < 0.01 < 0.01
Mn-med
b
550 (5100) 0.11 < 0.01
Mn-rec
c
2700 (4600) 0.66 < 0.01
Fe-med
b
160 (240) < 0.01 0.65
Fe-rec
c
230 (250) < 0.01 0.92
a
Values in parentheses are the calculated activities per mg of
metal containing enzyme.
b
Metal ions were added to the medium
during bacterial expression.
c
The enzymes were incubated at
70 °C for 1 h in the presence of metal ions.
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 599

ApeSOD (data not shown); similar temperature depen-
dence has been reported for the cambialistic SOD from
Pyrobaculum aerophilium [17]. This indicates that pro-
teins in solution need to have structural flexibility to
successfully incorporate metal cofactors. Because the
metal contents of the reconstituted enzymes were
higher (Table 1), the crystallographic studies were per-
formed on apo- and metal-reconstituted ApeSODs.
Crystallization and determination of structure
The crystals of ApeSOD were grown for 2–3 days in
the presence of polyethylene glycol. The crystals
belonged to the space group P2
1
, with four polypep-
tides in the asymmetric unit. After collection of diffrac-
tion data, the crystal structures were refined to 1.56,
1.35 and 1.45 A
˚
resolutions for apo, Mn-bound and
Fe-bound ApeSODs, respectively. Data collection and
refinement statistics are summarized in Table 2.
All three ApeSODs had essentially the same struc-
ture (Fig. 1A). The polypeptides consisted of seven
a-helices, a three-stranded antiparallel b-sheet and loops
connecting these secondary structure elements. The con-
tents of the a-helix and b-strand were 50% and 11%,
respectively. The monomer structure comprised two
domains: the rod-shaped N-terminal domain consisting
Table 2. Data collection and refinement statistics.
Protein Data Bank code 3AK1 3AK2 3AK3

Apo Mn Fe
Data collection
X-ray source BL44XU, SPring-8
Wavelength (A
˚
) 0.9
Space group P2
1
Unit cell (A
˚
, °) a = 69.26
b = 72.25
c = 76.65
b = 90.99
a = 69.06
b = 71.78
c = 76.85
b = 91.81
a = 69.06
b = 71.76
c = 76.40
b = 91.72
Resolution range (A
˚
)
a
50.0–1.56 (1.62–1.56) 50.0–1.35 (1.40–1.35) 50.0–1.48 (1.53–1.48)
Z 888
V
M

(A
˚
3
ÆDa
)1
) 1.95 1.94 1.92
R
merge
(%)
a,b
7.3 (39.6) 4.8 (36.1) 7.7 (38.3)
Completeness (%)
a
98.0 (92.4) 98.0 (97.0) 95.9 (88.9)
Total reflections 454833 1179512 732659
Unique reflections 106085 164370 124631
Redundancy
a
4.4 (2.7) 7.3 (6.9) 6.1 (5.2)
I ⁄ r(I)
a
18.1 (2.9) 15.9 (5.8) 16.6 (3.6)
B-factors of data from Wilson plot (A
˚
2
) 19.8 13.8 18.3
Refinement
Resolution range (A
˚
)

a
36.13–1.57 (1.61–1.57) 31.86–1.35 (1.38–1.35) 49.75–1.48 (1.52–1.48)
Number of reflections
a
98731 (6589) 152991 (10838) 113539 (7965)
R
cryst
(%) ⁄ R
free
(%)
a,c,d
19.8 (33.7) ⁄ 23.6 (35.9) 18.8 (23.9) ⁄ 20.3 (25.4) 22.6 (29.5) ⁄ 25.6 (32.7)
rmsd bond length (A
˚
) 0.008 0.008 0.012
rmsd bond angle (°) 1.135 1.170 1.382
Protein atoms 6984 6823 6855
Metal atoms 0 4 4
Ethylene glycol molecules 29 18 10
Water molecules 630 734 417
Average B-factor
Protein atoms 24.5 14.8 21.0
Solvent atoms 39.1 27.6 28.4
Metal atoms 10.1 16.0
Ramachandran plot (%)
e
Favored 91.8 92.9 92.2
Allowed 8.0 7.1 7.8
a
Values in parentheses are for the highest resolution shell.

b
R
merge
¼
P
hkl
P
j
jI
hkl;j
À<I
hkl
>j=
P
hkl
P
i
I
Ihkl;j
, where I
hkl,j
is the intensity of
observation I
hkl,j
and <I
hkl
> is the average of symmetry-related observations of a unique reflection.
c
R
cryst

=
P
||F
o
| ) |F
c
|| ⁄
P
|F
o
|, where F
o
and F
c
are observed and calculated structure factor amplitudes, respectively.
d
R
free
was calculated using a randomly-selected 5% of the
dataset that was omitted from all stages of refinement.
e
Ramachandran plots were prepared for all residues other than Gly and Pro.
Crystal structure of SOD from A. pernix K1 T. Nakamura et al.
600 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
of the N-terminal extended region and the following
two a-helices, and the globular, (a + b)-type C-termi-
nal domain.
Oligomeric structure
The A ⁄ B and C ⁄ D chains formed dimers in the crystal
packing. This dimerization buried 27% of the accessi-

ble surface area of each monomer. These two dimers
associated to form a tetramer in the asymmetric unit
(Fig. 1B). The tetramerization buried 13% of the
accessible surface area of each dimer. Neighboring tet-
ramers came into loose contact with each other in the
crystal packing. Similar molecular arrangements have
been found in the crystal structures of SODs from sev-
eral sources, such as Sulfolobus solfataricus [18], Myco-
bacterium tuberculosis [19], Aquifex pyrophilus [20] and
P. shermanii [6]. These SODs are assumed to be tetra-
meric in solution. Figure 1C illustrates the superimpo-
sition of ApeSOD with the cambialistic SOD from
P. shermanii. The overall structures of these enzymes
were similar; the rmsd of the 747 Ca atoms was
0.695 A
˚
.
ApeSOD eluted from the gel-filtration column with
the elution volume for a molecular mass of 57 kDa
(Fig. 2A). Because the calculated molecular mass of
the ApeSOD monomer is 24 577 Da, the gel filtration
results suggested a dimeric association; similar results
have previously been reported for the same protein
[16,21]. A second gel filtration through a Superdex 200
column also suggested that ApeSOD has a dimeric
structure in solution (data not shown). These findings
are in contrast to the results obtained in the crystallo-
graphic study (described above), which demonstrated
that ApeSOD has a tetrameric structure. To deter-
mine whether ApeSOD polypeptide associations were

90°
90°
A
B
C
Fig. 1. Crystal structure of ApeSOD.
(A) Monomer structures of apo (green),
Mn-bound (magenta) and Fe-bound (cyan)
ApeSODs are superimposed and shown as
a stereo view. The metal cofactor of the
Mn-bound form is indicated by a ball.
(B) Tetramer structures of ApeSOD viewed
from two directions. Chains A, B, C and D
are shown in red, yellow, green and blue,
respectively. (C) ApeSOD (red)
superimposed with the SOD of P. shermanii
(blue; Protein Data Bank code: 1AR5) [6].
The structure of ApeSOD is viewed from
the same directions as in (B). The Mn-bound
form is shown as the representative of
ApeSOD in (B) and (C). Prepared with
PYMOL
[44].
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 601
different in crystals and in solution, we used ultracen-
trifugation to accurately measure the molecular mass
of ApeSOD in solution. The result (96 265 Da) clearly
showed tetrameric assembly of ApeSOD in solution
(Fig. 2B). Because ultracentrifugation is a direct, accu-

rate measurement, we conclude that ApeSOD exists as
a tetramer in solution.
Active site geometry
Each monomer had an independent metal-binding site
at the interface of the two domains, which consists
of four side chains: two (His31 and His79) from the
N-terminal domain and two (Asp165 and His169) from
the C-terminal domain (Fig. 3). In Mn-reconstituted
ApeSOD, the metal ion was five-coordinate in trigonal
bipyramidal geometry (Fig. 3A). Three of the ligands,
OD2 of Asp165, NE2 of His79 and NE2 of His169,
formed an equatorial plane. The other protein ligand,
NE2 of His31, bound to the metal, in the company of
a water oxygen, from the apical positions. The manga-
nese was only 0.06 A
˚
out of the equatorial plane
(Table 3). The angles around the metal cofactor sug-
gested that the ligation form of Mn in ApeSOD is tri-
gonal bipyramidal.
In Fe-reconstituted ApeSOD, the metal was coordi-
nated with six ligands: the five same ligands in the
Mn-reconstituted enzyme and an additional water oxy-
gen, which, together with the OD2 of Asp165, the
NE2 of His79 and the NE2 of His169, formed an
equatorial plane (Fig. 3B). The metal ion and the addi-
tional water oxygen were only 0.03 and 0.04 A
˚
, respec-
tively, out of the equatorial plane defined by the other

three atoms (Table 3). The angles around the metal
ion indicated that Fe-bound ApeSOD contains
distorted octahedral coordination around the metal
cofactor.
The absence of an anomalous Fourier map demon-
strated that the active site in apo-ApeSOD did not
contain a metal cofactor (Fig. 3C). However, the side
chains and apical water coordinating the metal center
had the same configurations as those in the metal-
bound ApeSOD. This implies that the conformation
around the active site of ApeSOD is independent of
the presence of a metal cofactor.
Figure 3D shows the superimposition of the active
site structures of apo, Mn-bound and Fe-bound Ape-
SODs. The most significant difference among them
was related to the OH of Tyr39, which shifted 1.1 A
˚
toward the apical water molecule upon Fe binding
(Fig. 3D). The shift upon Mn binding was negligible,
and no significant differences were observed in other
3
4
5
6
7
8
9
100
10
2

MW (kDa)
131211109
Elution volume (mL)
252015
10
5
0
Elution volume (mL)
1.5
–0.05
0.00
0.05
Residuals
0.5
1.0
A
280
A
280
6.956.85 6.90 7.107.057.00
0.0
Radius (cm)
A
B
Fig. 2. Assembly of ApeSOD chains in solution. (A) Representative
gel filtration chromatogram of ApeSOD and calibration curve (inset).
The standard proteins are albumin (1), ovalbumin (2), chymotrypsin-
ogen A (3) and ribonuclease A (4). The estimated molecular mass
of ApeSOD is 57.0 kDa. (B) Sedimentation equilibrium distribution
of ApeSOD. The line reflects the best fit of the data and indicates

that the apparent molecular mass is 96 265 Da. The deviation
between empirical data and the fitted line is plotted in the upper
panel.
Fig. 3. Active site structure of ApeSOD. Active site structures of Mn-bound (A), Fe-bound (B) and apo (C) ApeSODs. (A–C) The orange map
represents the r
A
-weighted F
o
)F
c
electron density map at the 3 r level, where the indicated residues are excluded from the calculation of
the structure factor. The blue map represents the anomalous difference map contoured at 10 r. Schematic representations of coordinated
atoms are shown on the right in (A–C), where oxygen, nitrogen and metal atoms are represented by red, blue and black balls, respectively.
(C) The anomalous difference map was not seen even when the sigma level was set to 3 (data not shown). (D) Apo (green), Mn-bound
(magenta) and Fe-bound (cyan) structures are superimposed, and the residues around the active site are shown via stick models. The red,
magenta and cyan balls represent the active site water molecules in apo, Mn-bound and Fe-bound ApeSODs, respectively. Balls labeled
‘Metal’ are Mn or Fe. (E) Showing the superimposition of the active site structures of the Mn- (magenta) and Fe-bound (cyan) forms of
P. shermanii SOD. Water molecules and metal atoms are shown in the same way as in (D). Prepared with
PYMOL [44].
Crystal structure of SOD from A. pernix K1 T. Nakamura et al.
602 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
D
E
A
B
C
H169 (NE2)
D165 (OD2)
H31 (NE2)
H79 (NE2)

WAT
H169 (NE2)
D165 (OD2)
WAT
H31 (NE2)
H79 (NE2)
WAT
H31 (NE2)
H79 (NE2) WAT
H169 (NE2)
D165 (OD2)
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 603
active site residues between Mn- and Fe-bound ApeS-
ODs, although slight changes in His31, His79 and
Asp165 were observed between the metal-bound and
apo ApeSODs. The shift of the conserved Tyr residue
depending on the metal cofactor does not appear to be
a common feature among cambialistic SODs because,
in the case of the cambialistic SOD from P. shermanii,
no significant difference was found between the active
sites of the Mn- and Fe-bound forms (Fig. 3E).
Stability in organic medium
After we elucidated the tertiary structure of ApeSOD,
we were able to compare the number of ion pairs
among thermophilic SODs. Table 4 summarizes the
data obtained from two species each of archaea and
bacteria. A single ApeSOD polypeptide was found to
have seven intrasubunit ion pairs, whereas 24 intersub-
unit ion pairs were found in the ApeSOD tetrameric

assembly. This is significantly higher than the number
of ion pairs found in the SOD from the thermophilic
bacterium T. thermophilus (TthSOD) [22]. Because
electrostatic interactions in proteins are more domi-
nant when the solvent polarity is lower, we hypo-
thesized that ApeSOD would be more stable than
TthSOD in an organic solvent.
When we compared the stabilities of ApeSOD and
TthSOD, we found that the two enzymes were similar
in aqueous solution at temperatures up to 85 °C
(Fig. 4A). However, when the solvent contained 40%
ethylene glycol, TthSOD was inactivated by incubation
Table 4. Interactions in SODs. ApSOD, SOD from A. pyrophilus;
SsoSOD, SOD from S. solfataricus.
ApeSOD SsoSOD ApSOD TthSOD
Number of intrasubunit
ion pairs ⁄ monomer
7 5 14 6
Number of intrasubunit
ion pairs ⁄ residue
0.033 0.024 0.066 0.030
Number of intersubunit
ion pairs ⁄ tetramer
24 24 22 4
Number of intersubunit
ion pairs ⁄ residue
0.028 0.029 0.026 0.0049
Table 3. Distances and angles around the metal ions. For the iden-
tity of the atoms, see Fig. 3.
Mn Fe

Distance (A
˚
)
Metal – equatorial plane 0.06 ± 0.01 0.03 ± 0.02
Wat
a
– equatorial plane 0.04 ± 0.03
Angle (°)
D165-M-H79 104.8 ± 1.3 97.2 ± 2.9
H79-M-H169 136.1 ± 0.8 151.5 ± 2.1
H79-M-wat
a
77.1 ± 1.9
Wat
a
-M-H169 74.5 ± 0.6
H169-M-D165 118.8 ± 0.8 111.1 ± 1.5
H31-M-D165 82.7 ± 0.6 82.1 ± 1.2
H31-M-wat
b
170.1 ± 1.5 169.5 ± 1.5
a
Water on the equatorial plane.
b
Water at the apical position.
A
B
C
Fig. 4. Thermal stability of ApeSOD and TthSOD. Residual activi-
ties of ApeSOD (closed circles) and TthSOD (open circles) after

incubation at 85 °C in the aqueous solution (A) and the solution
containing 40% ethylene glycol (B) are plotted against incubation
time. (C) ApeSOD (closed symbols) and TthSOD (open symbols)
were incubated for 1 h at 60 °C (circles), 70 °C (squares) or 80 °C
(triangles) in a solution containing various concentrations of ethanol.
The residual activities after the incubation are plotted against the
ethanol concentration.
Crystal structure of SOD from A. pernix K1 T. Nakamura et al.
604 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
at 85 °C, whereas ApeSOD remained active (Fig. 4B).
Inclusion of ethanol in the solution destabilized both
ApeSOD and TthSOD. However, the extent of the
destabilization was more significant in TthSOD than in
ApeSOD (Fig. 4C). In other words, ApeSOD was
more stable than TthSOD under organic conditions.
This was not unexpected given the number of ion pairs
in ApeSOD and TthSOD. The larger number of elec-
trostatic interactions in ApeSOD contributed to its
stability in a low-polarity solvent.
Discussion
In the present study, we crystallized ApeSOD and
determined its tertiary structure. The asymmetric unit
contained four polypeptides that consisted of two
homodimers. Although both crystallographic study and
analytical ultracentrifugation indicated that ApeSOD
was a tetrameric protein, gel filtration suggested that
the protein was dimeric (Fig. 2A); in other words, the
molecular mass estimated by gel filtration was lower
than the true value. Similar results have also been
reported for other homologous SODs, such as those

from S. solfataricus [23] and P. shermanii [24]. Ursby
et al. [18] reported that the estimated molecular mass
of the S. solfataricus SOD increased when the column
was recalibrated with thermophilic proteins from the
same source. In cases such as these, analytical ultra-
centrifugation, rather than gel filtration, is a powerful
tool for accurately determining the oligomeric struc-
ture of SOD enzymes.
We observed five-coordinate and six-coordinate struc-
tures of metal ions in Mn-bound and Fe-bound Ape-
SOD crystal structures, respectively. In the six-
coordinated Fe-bound ApeSOD, an additional water
molecule was found in the equatorial plane (Fig. 3B).
Although trigonal bipyramidal 5-coordinate structures
are often observed around bound metals of SODs,
the octahedral six-coordinate crystal structures are
detected only in exceptional cases; for example, in the
cryo-trapped form of E. coli Mn-SOD [25], in Fe-substi-
tuted Mn-SOD of E. coli [3], in peroxide-soaked Mn-
SOD of E. coli [26] and in TthSOD complexed with
azide, a SOD inhibitor [27]. It remains unclear why the
six-coordinated structure was observed in Fe-bound
ApeSOD, but not in the Mn-bound form, because the
crystal contained neither superoxide substrate, nor anio-
nic inhibitors. The answer to this question will shed light
on the reaction mechanism of this cambialistic SOD.
Several enzymological properties known for Ape-
SOD [16] can be related to the structural characteris-
tics in the active site. Cambialistic SODs can be
divided into two groups: one with almost the same

activity in the Mn- and Fe-forms, and the other exhib-
iting low activity in the Fe-form and high activity in
Mn-form. ApeSOD belongs to the latter group [16].
This was also confirmed by the activity assay, which
demonstrated that ApeSOD was approximately 20-fold
more active in its Mn-bound form than in its
Fe-bound form (Table 1). We propose two possible
explanations for this finding. One is related to inhibi-
tion of the binding of the superoxide substrate and the
other to product inhibition by hydrogen peroxide.
There are two mechanisms proposed for the reaction
cycle of Mn ⁄ Fe-SOD: the 5-6-5 mechanism [27] and
the associative displacement mechanism [28]. In the
former model, the metal is five-coordinated in the rest-
ing state and six-coordinated when the substrate super-
oxide is bound. In the latter model, the association of
superoxide is concomitant with displacement of one of
the oxygen ligands. In both mechanisms, the superox-
ide substrate coordinates to the metal center from the
equatorial plane, and the coordination site is the same
as that of the additional water molecule in the
Fe-bound ApeSOD octahedral structure. Thus, it can
be assumed that the coordinated water in Fe-bound
ApeSOD inhibits superoxide binding.
Although the polypeptide conformations of the
Mn- and Fe-bound ApeSODs were almost the same
(Fig. 1A), a slight difference was observed in the side
chain of Tyr39 (Fig. 3D). This Tyr residue is conserved
in Mn- and Fe-SODs, constitutes the outer sphere of
the active site and has been shown by mutational stud-

ies to play a critical role in catalysis [29,30]. The OH
of Tyr39 was found to have shifted toward the apical
water molecule upon Fe binding (Fig. 3D). This shift
was analogous to that of Tyr34 in E. coli Mn-SOD
upon binding of hydrogen peroxide to the central
metal [26]. Peroxide-bound E. coli Mn-SOD represents
a product-inhibited form of the reduction step from
superoxide to hydrogen peroxide [26]. These findings
lead to the hypothesis that Fe-bound ApeSOD mimics
the product-inhibited form and the shift of Tyr39 sup-
presses the release of the peroxide product. This may
be one of the reasons why ApeSOD is less active in its
Fe-bound form. It is noteworthy that this shift of the
Tyr residue is not observed in the cambialistic SOD
from P. shermanii (Fig. 3E), which exhibits almost the
same activity in the presence of Mn and Fe as cofac-
tors [24].
Thermophilic bacteria, as well as thermophilic
archaea, produce thermophilic enzymes. Comparison
of two thermophilic SODs, ApeSOD (an archaeon)
and TthSOD (a bacterium) revealed that ApeSOD
contained more ion pairs, especially intersubunit ion
pairs, than TthSOD (Table 4). Because electrostatic
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 605
interactions are dominant under organic conditions,
we expected that the difference in number of ion pairs
would be reflected in a corresponding difference in the
stability of the proteins in organic solvents. Indeed,
although the stabilities of the two SODs were indistin-

guishable in aqueous solution, ApeSOD was more sta-
ble than TthSOD in a solution containing ethylene
glycol or ethanol (Fig. 4). It is reasonable to conclude
that the dominant electrostatic interaction in ApeSOD
contributed not only to heat tolerance, but also to
tolerance of the enzyme to organic environment.
For enzymes to be used in industrial applications,
it is preferable that they have both structural and func-
tional stability when placed under severe conditions.
Thus, because of its thermal stability and tolerance to
an organic medium, ApeSOD shows promise as a
potentially applicable enzyme.
Experimental procedures
Protein expression and purification
PCR was used to amplify the ORF of ApeSOD
(APE_0741) from genomic DNA of A. pernix K1 (NBRC
100138, provided by National Institute of Technology and
Evaluation, Chiba, Japan). The PCR product was cloned
into an expression plasmid vector pET11 (Novagen,
Darmstadt, Germany). E. coli Rosetta (DE3) cells harboring
the expression plasmid were cultivated in LB medium con-
taining 0.1 mgÆmL
)1
ampicillin at 37 °C and protein
expression was induced by the addition of 1 mm isopropyl
thio-b-d-galactoside (final concentration). The E. coli cells
were disrupted by sonication and the soluble proteins were
subjected to streptomycin and heat treatments, as described
previously [31]. The resulting solution was dialyzed in
20 mm sodium acetate buffer (pH 4.8) and applied onto

a cation exchange column (HiTrap SP; GE Healthcare,
Piscataway, NJ, USA). The protein was eluted by a linear
gradient of 0–1 m NaCl in the same buffer. The fractions
containing ApeSOD were collected, concentrated and gel-
filtered with a Superdex 75 column equilibrated with
20 mm Tris-HCl (pH 8.1), with 150 mm NaCl as the final
step. The purified protein was dissolved in 20 mm Tris-HCl
(pH 8.1). The protein concentration was determined from
its absorbance at 280 nm [32].
Gel filtration
Analytical gel filtration chromatography was performed
using a Superdex75GL (10 ⁄ 30) column (GE Healthcare)
with a buffer containing 20 mm Tris-HCl (pH 8.1) and
150 mm NaCl. The flow rate was 0.8 mLÆmin
)1
. The
column was calibrated using a Gel Filtration Calibration
Kit LMW (GE Healthcare).
Ultracentrifugation
ApeSOD solution in 20 mm Tris-HCl (pH 8.1) containing
150 mm NaCl was subjected to a sedimentation equilibrium
analysis using a Beckman Optima XL-A analytical ultracen-
trifuge with an An-60 Ti rotor (Beckman Coulter, Fullerton,
CA, USA). Samples were centrifuged at 9500 g. for 41 h at
20 °C. The molecular mass of the protein was calculated
from the sedimentation equilibrium plot.
Incorporation of metals
Metal cofactors were incorporated into the enzyme by one
of two procedures. To incorporate metals during growth
of bacterial host cells, 1 mm MnSO

4
(for Mn-bound Ape-
SOD) or 1 mm FeSO
4
(for Fe-bound ApeSOD) was added
to the medium [33]. The metal-containing enzymes were
purified using EDTA-free buffers. Alternatively, the metals
could be incorporated into the purified enzyme. For this
technique, approximately 1 mm ApeSOD in 20 mm Tris-
HCl (pH 8.1) was incubated with 10 mm MnSO
4
or
10 mm FeSO
4
at 70 °C for 1 h. The excess metals were
removed by dialysis and the enzyme was further purified by
gel filtration with a Superdex75 column equilibrated with
20 mm Tris-HCl (pH 8.1) and 150 mm NaCl. The metal
content of the enzyme was measured by inductively coupled
plasma atomic emission spectrometry using a Perkin-Elmer
Optima3300DV spectrometer (Perkin-Elmer, Waltham,
MA, USA).
Activity assay
The activity of SOD was assayed with a SOD Assay Kit-
WST (Dojin, Kumamoto, Japan) at 37 °C via the xanthine
oxidase-WST-1 method [34]. In accordance with the
manufacturer’s instructions, one unit was defined as the
amount of activity that caused 50% inhibition of WST-1
reduction.
Crystallization, data collection and processing

ApeSOD was crystallized by hanging drop vapor diffu-
sion, with a reservoir solution containing 100 mm Hepes-
NaOH (pH 7.5) 10% (w ⁄ v) poly(ethylene glycol) 6000 and
8% (v ⁄ v) ethylene glycol at 20 °C. The crystal was cryo-
protected with modified reservoir solution containing
22.5% ethylene glycol, cooled in a nitrogen gas stream
(100 K) and subjected to X-ray diffraction measurements
with synchrotron radiation at SPring-8 (Harima, Japan).
The collected data were integrated and scaled with
hkl2000 [35]. The initial phase was determined by molec-
ular replacement with molrep in the ccp4 suite [36,37].
Fe-SOD from S. solfataricus [18] (Protein Data Bank
code: 1WB8) was used as the search model. The resulting
structure was subjected to simulated annealing using cns
Crystal structure of SOD from A. pernix K1 T. Nakamura et al.
606 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
[38], followed by further refinement with refmac in the
ccp4 suite [37,39].
The secondary structure of proteins was assigned by dssp
[40]. Stereochemical analysis was performed using pro-
check [41]. Ion pairs were analyzed using contact in the
ccp4 suite [37] and confirmed visually using coot [42]. Ion-
pair interactions were identified using distances < 4 A
˚
.
When we counted the interactions, we excluded all residues
involved in the binding of the metal cofactor.
ApeSODs in different forms were superimposed with the
least square fit of Ca atoms of the residues in the range
10–200. ApeSOD and P. shermanii SOD were superimposed

with secondary-structure matching [43].
Thermal stability
Heat treatment of ApeSOD (Mn-bound form) and TthSOD
were performed in a buffer containing 50 mm Hepes-KOH
(pH 7.0) and 0.5 mm MnSO
4
, with a protein concentration
of 0.8 lm. When indicated, the buffer contained 40% ethyl-
ene glycol or various concentrations of ethanol. Thermal
stability was evaluated by measuring enzyme activity after
incubation.
Acknowledgements
Diffraction data were collected at the Osaka University
beamline BL44XU at SPring-8, equipped with
MX225-HE (Rayonix), which is financially supported
by Academia Sinica and National Synchrotron
Radiation Research Center (Taiwan, ROC). We
thank Ms M.Sakai (Osaka University) for performing
ultracentrifugation analysis and Mr K. Mieda and
M. Sakata (Tottori University) for technical assistance
in enzyme assay. This study was supported by a
Grant-in-Aid for Scientific Research (21510237) from
Japan Society for the Promotion of Sciences (JSPS).
References
1 Miller AF (2004) Superoxide dismutases: active sites
that save, but a protein that kills. Curr Opin Chem Biol
8, 162–168.
2 Smith MW & Doolittle RF (1992) A comparison of
evolutionary rates of the two major kinds of superoxide
dismutase. J Mol Evol 34, 175–184.

3 Edward RA, Whittaker MM, Whittaker JW, Jameson
GB & Baker EN (1998) Distinct metal environment in
Fe-substituted manganese superoxide dismutase pro-
vides a structural basis of metal specificity. J Am Chem
Soc 120, 9684–9685.
4 Perry JJ, Shin DS, Getzoff ED & Tainer JA (2010) The
structural biochemistry of the superoxide dismutases.
Biochim Biophys Acta 1804, 245–262.
5 Sugio S, Hiraoka BY & Yamakura F (2000) Crystal
structure of cambialistic superoxide dismutase from
Porphyromonas gingivalis. Eur J Biochem 267, 3487–
3495.
6 Schmidt M, Meier B & Parak F (1996) X-ray structure
of the cambialistic superoxide dismutase from Propioni-
bacterium shermanii active with Fe or Mn. J Biol Inorg
Chem 1, 532–541.
7 Yamakura F, Sugio S, Hiraoka BY, Ohmori D &
Yokota T (2003) Pronounced conversion of the
metal-specific activity of superoxide dismutase from
Porphyromonas gingivalis by the mutation of a single
amino acid (Gly155Thr) located apart from the active
site. Biochemistry 42, 10790–10799.
8 Tabares LC, Gatjens J & Un S (2010) Understanding
the influence of the protein environment on the Mn(II)
centers in superoxide dismutases using high-field elec-
tron paramagnetic resonance. Biochim Biophys Acta
1804, 308–317.
9 Sako Y, Nomura N, Uchida A, Ishida Y, Morii H,
Koga Y, Hoaki T & Maruyama T (1996) Aeropyrum
pernix gen. nov., sp. nov., a novel aerobic hyperthermo-

philic archaeon growing at temperatures up to 100°C.
Int J Syst Bacteriol 46, 1070–1077.
10 Kawarabayasi Y, Hino Y, Horikawa H, Yamazaki S,
Haikawa Y, Jin-no K, Takahashi M, Sekine M, Baba S,
Ankai A et al. (1999) Complete genome sequence of an
aerobic hyper-thermophilic crenarchaeon, Aeropyrum
pernix K1. DNA Res 6, 83–101.
11 Nakamura T, Yamamoto T, Inoue T, Matsumura H,
Kobayashi A, Hagihara Y, Uegaki K, Ataka M, Kai Y
& Ishikawa K (2006) Crystal structure of thioredoxin
peroxidase from aerobic hyperthermophilic archaeon
Aeropyrum pernix K1. Proteins 62, 822–826.
12 Nakamura T, Yamamoto T, Abe M, Matsumura H,
Hagihara Y, Goto T, Yamaguchi T & Inoue T. (2008)
Oxidation of archaeal peroxiredoxin involves a hyperva-
lent sulfur intermediate. Proc Natl Acad Sci USA 105,
6238–6242.
13 Nakamura T, Kado Y, Yamaguchi T, Matsumura H,
Ishikawa K & Inoue T (2010) Crystal structure of
peroxiredoxin from Aeropyrum pernix K1 complexed
with its substrate, hydrogen peroxide. J Biochem 147,
109–115.
14 Mizohata E, Sakai H, Fusatomi E, Terada T, Muray-
ama K, Shirouzu M & Yokoyama S (2005) Crystal
structure of an archaeal peroxiredoxin from the aerobic
hyperthermophilic crenarchaeon Aeropyrum pernix K1.
J Mol Biol 354, 317–329.
15 Jeon SJ & Ishikawa K (2002) Identification and charac-
terization of thioredoxin and thioredoxin reductase
from Aeropyrum pernix K1. Eur J Biochem 269, 5423–

5430.
16 Yamano S, Sako Y, Nomura N & Maruyama T (1999)
A cambialistic SOD in a strictly aerobic hyperthermo-
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 607
philic archaeon, Aeropyrum pernix. J Biochem 126,
218–225.
17 Whittaker MM & Whittaker JW (2000) Recombinant
superoxide dismutase from a hyperthermophilic archa-
eon, Pyrobaculum aerophilium. J Biol Inorg Chem 5,
402–408.
18 Ursby T, Adinolfi BS, Al-Karadaghi S, De Vendittis E
& Bocchini V (1999) Iron superoxide dismutase from
the archaeon Sulfolobus solfataricus: analysis of
structure and thermostability. J Mol Biol 286, 189–
205.
19 Cooper JB, McIntyre K, Badasso MO, Wood SP,
Zhang Y, Garbe TR & Young D (1995) X-ray structure
analysis of the iron-dependent superoxide dismutase
from Mycobacterium tuberculosis at 2.0 A
˚
resolution
reveals novel dimer-dimer interactions. J Mol Biol 246,
531–544.
20 Lim JH, Yu YG, Han YS, Cho S, Ahn BY, Kim SH &
Cho Y (1997) The crystal structure of an Fe-superoxide
dismutase from the hyperthermophile Aquifex pyrophi-
lus at 1.9 A
˚
resolution: structural basis for thermostabil-

ity. J Mol Biol 270, 259–274.
21 Lee HJ, Kwon HW, Koh JU, Lee DK, Moon JY &
Kong KH (2010) An efficient method for the expression
and reconstitution of thermostable Mn ⁄ Fe superoxide
dismutase from Aeropyrum pernix K1. J Microbiol
Biotechnol 20, 727–731.
22 Ludwig ML, Metzger AL, Pattridge KA & Stallings
WC (1991) Manganese superoxide dismutase
from Thermus thermophilus. A structural model
refined at 1.8 A
˚
resolution. J Mol Biol 219, 335–
358.
23 Dello Russo A, Rullo R, Nitti G, Masullo M &
Bocchini V (1997) Iron superoxide dismutase from the
archaeon Sulfolobus solfataricus: average hydrophobic-
ity and amino acid weight are involved in the adapta-
tion of proteins to extreme environments. Biochim
Biophys Acta 1343, 23–30.
24 Meier B, Barra D, Bossa F, Calabrese L & Rotilio G
(1982) Synthesis of either Fe- or Mn-superoxide dismu-
tase with an apparently identical protein moiety by an
anaerobic bacterium dependent on the metal supplied.
J Biol Chem 257, 13977–13980.
25 Borgstahl GE, Pokross M, Chehab R, Sekher A & Snell
EH (2000) Cryo-trapping the six-coordinate, distorted-
octahedral active site of manganese superoxide dismu-
tase. J Mol Biol 296, 951–959.
26 Porta J, Vahedi-Faridi A & Borgstahl GE (2010) Struc-
tural analysis of peroxide-soaked MnSOD crystals

reveals side-on binding of peroxide to active-site manga-
nese. J Mol Biol 399, 377–384.
27 Lah MS, Dixon MM, Pattridge KA, Stallings WC, Fee
JA & Ludwig ML (1995) Structure-function in Escheri-
chia coli iron superoxide dismutase: comparisons with
the manganese enzyme from Thermus thermophilus
.
Biochemistry 34, 1646–1660.
28 Whittaker MM & Whittaker JW (1996) Low-tempera-
ture thermochromism marks a change in coordination
for the metal ion in manganese superoxide dismutase.
Biochemistry 35, 6762–6770.
29 Edwards RA, Whittaker MM, Whittaker JW, Baker
EN & Jameson GB (2001) Outer sphere mutations per-
turb metal reactivity in manganese superoxide dismu-
tase. Biochemistry 40, 15–27.
30 Perry JJ, Hearn AS, Cabelli DE, Nick HS, Tainer JA &
Silverman DN (2009) Contribution of human manga-
nese superoxide dismutase tyrosine 34 to structure and
catalysis. Biochemistry 48, 3417–3424.
31 Nakamura T, Matsumura H, Inoue T, Kai Y, Uegaki
K, Hagihara Y, Ataka M & Ishikawa K (2005)
Crystallization and preliminary X-ray diffraction
analysis of thioredoxin peroxidase from the aerobic
hyperthermophilic archaeon Aeropyrum pernix K1.
Acta Crystallogr Sect F: Struct Biol Cryst Commun
61, 323–325.
32 Edelhoch H (1967) Spectroscopic determination of
tryptophan and tyrosine in proteins. Biochemistry 6,
1948–1954.

33 Gabbianelli R, Battistoni A, Polizio F, Carri MT,
De Martino A, Meier B, Desideri A & Rotilio G
(1995) Metal uptake of recombinant cambialistic
superoxide dismutase from Propionibacterium
shermanii is affected by growth conditions of host
Escherichia coli cells. Biochem Biophys Res Commun
216, 841–847.
34 Ukeda H, Kawana D, Maeda S & Sawamura M (1999)
Spectrophotometric assay for superoxide dismutase
based on the reduction of highly water-soluble tetrazo-
lium salts by xanthine-xanthine oxidase. Biosci Biotech-
nol Biochem 63, 485–488.
35 Otwinowski Z & Minor W (1997) Processing of X-ray
diffraction data collected in oscillation mode. Methods
Enzymol 276 , 307–326.
36 Vagin A & Teplyakov A (1997) MOLREP: an auto-
mated program for molecular replacement. J Appl
Crystallogr 30, 1022–1025.
37 Collaborative Computational Project, Number 4
(1994) The CCP4 suite: programs for protein crystal-
lography. Acta Crystallogr D: Biol Crystallogr 50,
760–763.
38 Brunger AT, Adams PD, Clore GM, DeLano WL,
Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J,
Nilges M, Pannu NS et al. (1998) Crystallography &
NMR system: a new software suite for macromolecular
structure determination. Acta Crystallogr D Biol
Crystallogr 54, 905–921.
39 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the

Crystal structure of SOD from A. pernix K1 T. Nakamura et al.
608 FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS
maximum-likelihood method. Acta Crystallogr D: Biol
Crystallogr 53, 240–255.
40 Kabsch W & Sander C (1983) Dictionary of protein
secondary structure: pattern recognition of hydrogen-
bonded and geometrical features. Biopolymers 22,
2577–2637.
41 Laskowski RA, MacArthur MW, Moss DS & Thornton
JM (1993) PROCHECK – a program to check the
stereochemical quality of protein structures. J Appl
Crystallogr 26, 283–291.
42 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D: Biol
Crystallogr 60, 2126–2132.
43 Krissinel E & Henrick K (2004) Secondary-structure
matching (SSM), a new tool for fast protein structure
alignment in three dimensions. Acta Crystallogr D: Biol
Crystallogr 60, 2256–2268.
44 DeLano WL (2002) The PyMol Molecular
Graphics System. DeLano Scientific, San Carlos, CA,
USA.
T. Nakamura et al. Crystal structure of SOD from A. pernix K1
FEBS Journal 278 (2011) 598–609 ª 2010 The Authors Journal compilation ª 2010 FEBS 609