Báo cáo khoa học: X-ray crystallographic and enzymatic analyses of shikimate dehydrogenase from Staphylococcus epidermidis pot

X-ray crystallographic and enzymatic analyses of shikimate
dehydrogenase from Staphylococcus epidermidis
Implications for substrate binding and conformational change
Cong Han
1,
*, Tiancen Hu
2,
*, Dalei Wu
2
,SuQu
1
, Jiahai Zhou
3
, Jianping Ding
4
, Xu Shen
2
,DiQu
1
and Hualiang Jiang
2
1 Institutes of Biomedical Sciences and Key Laboratory of Medical Molecular Virology, Institute of Medical Microbiology, Shanghai Medical
College, Fudan University, China
2 Drug Discovery and Design Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Chinese Academy of
Sciences, China
3 Shanghai Institute of Organic Chemistry, Chinese Academy of Sciences, China
4 State Key Laboratory of Molecular Biology and Research Center for Structural Biology, Institute of Biochemistry and Cell Biology, Shanghai
Institutes for Biological Sciences, Chinese Academy of Sciences, China
Keywords
crystal structure; shikimate dehydrogenase;
shikimate pathway; site-directed

mutagenesis; Staphylococcus epidermidis
Correspondence
D. Qu, Institutes of Biomedical Sciences and
Key Laboratory of Medical Molecular Virology,
Institute of Medical Microbiology, Shanghai
Medical College, Fudan University, Shanghai
200032, China
Fax: +86 21 54237603
Tel: +86 21 54237524
E-mail:
X. Shen, Drug Discovery and Design Center,
State Key Laboratory of Drug Research,
Shanghai Institute of Materia Medica, Chinese
Academy of Sciences, Shanghai 201203, China
Fax ⁄ Tel: +86 21 50806918
E-mail:
*These authors contributed equally to this work
Database
Coordinate and structure factor files for SeSDH
and SeSDH in complex with shikimate have
been deposited in the Protein Data Bank under
the accession numbers 3DON and 3DOO
(Received 28 August 2008, revised
9 November 2008, accepted 12 December
2008)
doi:10.1111/j.1742-4658.2008.06856.x
Shikimate dehydrogenase (SDH) catalyzes the NADPH-dependent
reduction of 3-dehydroshikimate to shikimate in the shikimate pathway. In
this study, we determined the kinetic properties and crystal structures of
Staphylococcus epidermidis SDH (SeSDH) both in its ligand-free form and

in complex with shikimate. SeSDH has a k
cat
of 22.8 s
)1
and a K
m
of
73 lm towards shikimate, and a K
m
of 100 lm towards NADP. The overall
folding of SeSDH comprises the N-terminal a ⁄ b domain for substrate bind-
ing and the C-terminal Rossmann fold for NADP binding. The active site
is within a large groove between the two domains. Residue Tyr211, nor-
mally regarded as important for substrate binding, does not interact with
shikimate in the binary SeSDH–shikimate complex structure. However, the
Y211F mutation leads to a significant decrease in k
cat
and a minor increase
in the K
m
for shikimate. The results indicate that the main function of
Tyr211 may be to stabilize the catalytic intermediate during catalysis. The
NADP-binding domain of SeSDH is less conserved. The usually long helix
specifically recognizing the adenine ribose phosphate is substituted with a
short 3
10
helix in the NADP-binding domain. Moreover, the interdomain
angle of SeSDH is the widest among all known SDH structures, indicating
an inactive ‘open’ state of the SeSDH structure. Thus, a ‘closing’ process
might occur upon NADP binding to bring the cofactor close to the

substrate for catalysis.
Abbreviations
AaSDH, Aquifex aeolicus shikimate dehydrogenase; Af, Archaeoglobus fulgidus; AtSDH, Arabidopsis thaliana shikimate dehydrogenase;
EcSDH, Escherichia coli shikimate dehydrogenase; Gk, Geobacillus kaustophilus; HiSDH, Haemophilus influenzae shikimate dehydrogenase;
HpSDH, Helicobacter pylori shikimate dehydrogenase; IPTG, isopropyl thio-b-
D-galactoside; MAD, multiple-wavelength anomalous diffraction;
MjSDH, Methanococcus jannaschii shikimate dehydrogenase; MtSDH, Mycobacterium tuberculosis shikimate dehydrogenase; SDH,
shikimate dehydrogenase; SDHL, shikimate dehydrogenase-like enzyme; SeMet-SeSDH, selenomethionine-substituted SeSDH; SeSDH,
Staphylococcus epidermidis shikimate dehydrogenase; TtSDH, Thermus thermophilus shikimate dehydrogenase.
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1125
Staphylococcus epidermidis is the most common causal
microorganism responsible for infections of implanted
medical devices such as central venous catheters,
cardiac pacemakers, artificial lenses and prosthetic
joints [1]. The pathogenesis of S. epidermidis-mediated
infections is mainly attributed to the adherence and
subsequent formation of a multilayered biofilm of
S. epidermidis on biomaterials. Bacterial cells within
the biofilm are dramatically less susceptible to antibi-
otic treatment and attacks by the immune system than
planktonic cells. Moreover, a biofilm may continuously
release bacteria into the bloodstream on a chronic
basis, resulting in bacteremia. Therefore, the formation
of a biofilm of pathogenic bacteria often results in the
removal of implanted medical devices, thus leading to
substantial morbidity and mortality [2]. Moreover, the
appearance of multiresistant and vancomycin-
resistant S. epidermidis strains may impair the effi-
cacy of antibiotic treatment regimens. The pressing
need to control S. epidermidis-mediated infection is

creating an urgent challenge to discover novel anti-
bacterial agents that are active against new bacterial
targets.
In bacteria, seven enzymes involved in the shiki-
mate pathway catalyze the sequential conversion
from erythrose 4-phosphate and phosphoenolpyruvate
via shikimate to chorismate, which serves as a pre-
cursor for the synthesis of essential metabolites such
as aromatic amino acids, folic acid and ubiquinone
[3]. The shikimate pathway is crucial in algae, higher
plants, bacteria, apicomplexan parasites and fungi, but
is absent in mammals, making the enzymes involved in
this pathway potential targets for the development of
nontoxic antimicrobial agents, herbicides and anti-para-
site drugs [4]. Shikimate dehydrogenase (SDH, EC
1.1.1.25) catalyzes the fourth reaction of the shikimate
pathway, an NADPH-dependent reduction of 3-dehyd-
roshikimate to shikimate. The inhibitors targeting Heli-
cobacter pylori SDH can block growth of the bacteria
in vitro, demonstrating that SDH is a promising target
for antimicrobial agents [5]. Shikimate dehydrogenase
belongs to the superfamily of NAD(P)H-dependent
oxidoreductases. In plant, SDH is coupled with 3-dehy-
droquinate dehydratase to form a bifunctional enzyme
[6]. In fungi and yeast, SDH serves as a component of
the penta-functional AROM enzyme complex that
catalyzes steps 2–6 within the shikimate pathway [7].
There are three SDH orthologues – AroE, YdiB and
SDH-like enzyme (SDHL) – in bacteria. AroE has been
identified as a single monofunctional enzyme that is

strictly specific for the NADPH-dependent reduction of
3-dehydroshikimate to shikimate in most bacteria.
YdiB, found in Escherichia coli, Salmonella typhi-
murium, Streptococcus pneumoniae and Haemophilus
influenzae, is characterized as a quinate ⁄ shikimate
dehydrogenase that not only retains the function of
AroE but also reversibly reduces dehydroquinate to
quinate using either NADH or NADPH as a cofactor.
It plays a more important role in the quinate pathway
than in the shikimate pathway. The SDHL, in a small
group of species such as Pseudomonas, only catalyzes
the NADPH-dependent reduction of 3-dehydro-
shikimate to shikimate but with a dramatically lower
catalytic rate than that of AroE [8]. However, the
complete genome sequence of S. epidermidis has
revealed the presence of only AroE in the shikimate
biosynthetic route.
A total of 20 crystal structures of SDH have been
determined so far covering all the three orthologues
of SDH mentioned above, including AroE from
E. coli (PDB code: 1NYT [9]), H. influenzae (1P74
and 1P77 [10]), Methanococcus jannaschii (1NVT
[11]), Aquifex aeolicus (2HK7, 2HK8 and 2HK9
[12]), Thermus thermophilus (1WXD, 2CY0, 2D5C
and 2EV9 [13]), Arabidopsis thaliana (2GPT [14],
2O7Q and 2O7S [15]) and Geobacillus kaustophilus
(2EGG); YdiB from E. coli (1O9B [9], 1NPD [16]
and 1VI2) and Corynebacterium glutamicum (2NLO
[17]); and SDHL from H. influenzae (1NPY [8]).
These structures comprise the following diverse con-

formations (a) apo-enzyme (1NPY, 1P74, 1WXD,
2EGG, 2HK7, 2HK8 and 2NLO), (b) binary com-
plex bound with either cofactor (1NPD, 1NVT,
1NYT, 1O9B, 1P77, 1VI2 and 2CY0) or substrate
(2D5C, 2GPT and 2O7Q) and (c) inactive (2HK9_A
and 2EV9) and active (2HK9_D and 2O7S) ternary
complexes. Analysis of these different conformations
reveals the binding information of substrate and co-
factor, the structural basis underlying the cofactor
specificity of AroE and YdiB [9], and the putative
catalytic mechanism of SDH [12]. Notably, the rela-
tive positions of the two domains responsible for
substrate and cofactor binding, respectively, are
different among these structures, representing two dis-
tinct states of SDH, namely the open form and the
closed form. Only the closed form is believed to be
competent for catalysis [9,12,13]. In addition to the
unliganded enzyme structures, the crystallographic
analysis of SDH in complex with cofactor, and of even
ternary enzyme–cofactor–substrate complexes, sheds-
new light on the catalytic mechanism and provides
clues for the rational design of anti-infective com-
pounds. Although the 3D structures of SDH have
offered much detailed structural information, few
reported SDH structure originates from pathogenic
bacteria, particularly gram-positive bacteria.
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1126 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
In this article, we report the crystal structures of
S. epidermidis SDH (SeSDH), in both ligand-free form

and in complex with shikimate, and the enzymatic
characterization of SeSDH. Our structure represents
the first SDH structure from gram-positive bacteria.
The overall folding of SeSDH is similar to that of
other SDH structures, constituted by the N-terminal
a ⁄ b domain for substrate binding and the C-terminal
Rossmann fold for NADP binding. The active site is
present within a large groove between these two
domains. The N-terminal domain and the shikimate-
binding residues of SeSDH are highly conserved
among SDH enzymes, except that the tyrosine residue
(Tyr211), normally regarded as important for substrate
binding, does not interact with shikimate in the crystal
structure of SeSDH in complex with shikimate. On the
basis of the enzymatic data of the Y211F mutant, we
suggest that Tyr211 plays a crucial role to stabilize the
catalytic intermediate during catalysis. The NADP-
binding domain of SeSDH is less conserved. A long
helix specifically recognizing the adenine ribose phos-
phate is substituted with a short 3
10
helix in this
domain. Moreover, the interdomain angle of SeSDH is
the widest among all known SDH structures. Extensive
comparison with other SDH structures indicates an
inactive ‘open’ state of our structure and implies that a
‘closing’ process might occur upon NADP binding to
bring the cofactor close to the substrate for catalysis.
Our study is expected to enhance the understanding of
SDH features and provide useful information for the

rational drug design of novel antimicrobial agents
targeting SeSDH.
Results
Biochemical characterization of SeSDH and
its Y211F mutant
After one-step purification of nickel-affinity chroma-
tography, the recombinant SeSDH, coupled to a
C-terminus six-histidine tag, was purified to apparent
homogeneity. The LC-ESI-MS spectral data gave a
molecular mass of 31 011 Da for the recombinant
SeSDH, which is in good agreement with the theoreti-
cal molecular mass of 31 013 Da calculated from the
amino acid sequence. Similarly, the substitution of
Y211F was corroborated by MS. The predicted and
observed molecular mass values were 30 996 and
30 995 Da, respectively. In the gel-filtration experi-
ments, the size-exclusion chromatography of SeSDH
showed only one peak. The elution volume of SeSDH
was larger than that of ovalbumin (43.0 kDa) and
smaller than that of chymotrypsinogen A (25.0 kDa)
(Fig. 1). Considering that the molecular mass of
SeSDH is equal to 31 011 Da, we conclude that
SeSDH might exist as a monomer in solution state.
Next, we investigated the catalytic properties of
SeSDH, as well as its Y211F mutant, and the effects
of pH on catalysis. The kinetic parameters K
m
and
V
max

were calculated from the slope and intercept val-
ues of the linear fit in a Lineweaver–Burke plot. For
example, the Lineweaver–Burke plot for the NADP-
dependent oxidation of shikimate to 3-dehydroshiki-
mate is shown in Fig. 2. In comparison with the
kinetic parameters of SDH enzymes from other bacte-
ria shown in Table 1, the K
m
and k
cat
values of SeSDH
were comparable with those of A. aeolicus SDH
(AaSDH) [12] and H. pylori SDH (HpSDH) [5]. As
illustrated in Table 1, the Y211F mutation resulted in
a 345-fold decrease in the k
cat
value, a three-fold
increase in the K
m
value and a 1073-fold decrease in
the k
cat
⁄ K
m
value for shikimate, which indicates that
Tyr211 plays a major role in the catalytic process and
a minor role in the initial substrate binding. Similarly
to HpSDH [5] and Archaeoglobus fulgidus SDH (Af-
Fig. 1. Gel-filtration study of SeSDH. (A) Size-exclusion chromato-
graphy of low-molecular-mass standards. The elution points of

molecular mass standards [BSA (67.0 kDa), ovalbumin (43.0 kDa),
chymotrypsinogen A (25.0 kDa) and ribonuclease A (13.7 kDa)] are
shown for reference. (B) Size-exclusion chromatography of SeSDH.
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1127
SDH) [18], SeSDH can also oxidize shikimate using
NAD as a cofactor, yielding a k
cat
of 87 ± 20 s
)1
,a
K
m
of 10.6 ± 2.6 mm and a k
cat
⁄ K
m
of
8.2 · 10
3
m
)1
Æs
)1
towards NAD. SeSDH showed a K
m
for NAD that was almost 100 times higher than that
for NADP at the saturation of shikimate, suggesting
that NADP is the preferred cofactor of SeSDH. We
also tested whether SeSDH could utilize quinate as a

substrate. Even in the presence of a high concentration
of quinate (4 mm), SeSDH displayed no activity using
either NADP or NAD as a cofactor.
The pH can dramatically affect enzyme activity in a
number of ways. As shown in Fig. 3, SeSDH is active
within a wide pH range of 7–12, with the highest activ-
ity occurring at around pH 11. It was reported that
the pH optimum of HpSDH is 8–10 [5], and the pH
optimum of AfSDH is 7–7.5 [18]. By contrast, SeSDH
exhibits very high activity at an extremely basic pH
range of 10–12, similar to Mycobacterium tuberculosis
SDH (MtSDH) [19]. Thus, it can be speculated that
the active site of SDH might involve several acidic ⁄
basic amino acid residues that play crucial roles in the
substrate-binding and catalytic processes.
Overall 3D structure of SeSDH
The overall structure of apo-SeSDH is basically iden-
tical to that of the shikimate–SeSDH complex, with
an rmsd value of 0.51 A
˚
from aligning 246 pairs of
Ca atoms. Both structures contain one molecule per
asymmetric unit, indicating that SeSDH might func-
tion as a monomer, which is also in accordance with
gel-filtration analysis. However, M. jannaschii SDH
(MjSDH) and T. thermophilus SDH (TtSDH) are con-
sidered to be dimers under physiological conditions
[11,13].
The statistics of the apo-SeSDH and shikimate-
bound SeSDH structures are summarized in Table 2.

The apo-SeSDH structure contains 264 amino acids
(residues 1–271, the loop containing residues 185–191
is disordered) and 223 water molecules. The binary
shikimate–SeSDH structure contains 258 amino acids
Fig. 2. The Lineweaver–Burke plot for the NADP-dependent oxida-
tion of shikimate to 3-dehydroshikimate.
Table 1. Comparison of kinetic parameters of SDHs from various bacteria.
SDH species pH k
cat
(s
)1
) (shikimate) K
m
(lM) (shikimate) K
m
(lM) (NADP) k
cat
⁄ K
m
(M
)1
Æs
)1
) (shikimate) k
cat
⁄ K
m
(M
)1
Æs

)1
) (NADP)
SeSDH 8.0 22.8 ± 1.5 73 ± 3 100 ± 10 3.12 · 10
5
2.28 · 10
5
Y211F mutant 8.0 0.066 ± 0.017 227 ± 15 279 ± 55 291 236
AaSDH
a
9.0 55.5 ± 1.5 42.4 ± 1.6 42.4 ± 0.9 1.31 · 10
6
1.31 · 10
6
HpSDH
b
8.0 7.7 ± 0.9 148 ± 28 182 ± 27 5.2 · 10
4
3.9 · 10
4
AfSDH
c
7.3 361 170 ± 30 190 ± 10 2.12 · 10
6
1.9 · 10
6
MtSDH
d
9.0 399 30 63 1.33 · 10
7
6.33 · 10

6
EcSDH
e
9.0 237 65 56 3.65 · 10
6
4.23 · 10
6
a
Kinetic parameters for A. aeolicus SDH are from [12].
b
Kinetic parameters for H. pylori SDH are from [5].
c
Kinetic parameters for A. fulgi-
dus SDH are from [18].
d
Kinetic parameters for M. tuberculosis SDH are from [19].
e
Kinetic parameters for E. coli SDH are from [9].
Fig. 3. pH profile of SeSDH enzyme activity.
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1128 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
(residues 1–267, the loop containing residues 185–193
is disordered), one shikimate molecule and 151 water
molecules.
As illustrated in Fig. 4A, similarly to other SDH
structures, SeSDH comprises two domains. The N-ter-
minal substrate-binding domain contains amino acid
residues 1–100 and 233–267. It is formed by a central
six-strand mixed b sheet (b2, b1, b3, b5, b6 and b4; b5
is antiparallel to the others) flanked by three a-helices

(a1, a9 and a8) on the inner side and by two a-helices
(a2 and a3) and two 3
10
helices (g1 and g2) on the
outer side. Helices a8 and a9 are in the most C-termi-
nal region of the polypeptide, which folds back into
the N-terminal domain. The C-terminal NADP-bind-
ing domain comprises two parallel Rossmann folds.
The mostly parallel six-stranded b-sheet (b9, b8, b7,
b10, b11 and b12) at the core of the domain is flanked
by three helices (a4, a5 and g4) on the inner side and
by two helices (g5 and a6) on the outer side. The two
domains are connected by two a-helices (a4 and a8) in
the middle of the molecule, creating a deep groove
where the catalysis occurs.
Substrate-binding domain
Superposition reveals that the overall folding of the
SeSDH substrate-binding domain is highly similar to
that of other SDH structures, with rmsd values rang-
ing from 0.68 to 1.6 A
˚
. The only difference is that the
a2 helix in SeSDH is packed more towards the central
b-sheet, and the orientation of the C-terminal helix a9
is divergent among these enzymes (Fig. 4B).
The substrate shikimate is unambiguously positioned
in the well-defined annealed omit map of the complex
Table 2. Data collection, phasing, and refinement statistics.
Apo-SeSDH Shikimate-SeSDH Apo-SeMet-SeSDH
Data collection

Space group P2
1
2
1
2
1
P2
1
P2
1
P2
1
P2
1
Cell dimensions
a, b, c (A
˚
) 52.88, 54.15, 102.72 45.19, 52.53, 56.78 45.15, 52.36, 56.71 45.18, 52.38, 56.74 45.20, 52.40, 56.77
a, b, c (°) 90, 90, 90 90, 96.14, 90 90, 95.77, 90 90, 95.77, 90 90, 95.72, 90
Peak Edge RemoteH
Wavelength (A
˚
) 1.5418 1.5418 0.97901 0.97953 0.96429
Resolution (A
˚
)
a
50–2.1 (2.18–2.10) 30–2.2 (2.28–2.20) 20–2.5 (2.56–2.50) 20–2.5 (2.56–2.50) 20–2.5 (2.56–2.50)
R
sym

b
0.100 (0.334) 0.122 (0.380) 0.036 (0.055) 0.037 (0.059) 0.040 (0.064)
I ⁄ rI 5.9 (2.0) 5.2 (1.9) 22.4 (16.4) 21.9 (15.7) 20.8 (14.8)
Completeness (%) 97.9 (97.9) 99.2 (99.2) 95.8 (74.7) 95.7 (74.7) 95.7 (75.6)
Redundancy 5.57 (5.82) 3.52 (3.45) 2.89 (2.54) 2.90 (2.55) 2.90 (2.55)
Refinement
Resolution (A
˚
) 15–2.1 15–2.2
No. reflections 16 488 12 793
R
work
⁄ R
free
c
0.183 ⁄ 0.260 0.188 ⁄ 0.264
No. of atoms
Protein 2091 2048
Water 223 151
Substrate – 12
B-factors
Protein 22.3 28.1
Water 31.0 31.5
Substrate – 25.5
rmsd (A
˚
) 0.02 0.03
Bond angles (°) 1.72 2.08
Ramachandran plot (%)
Most favored 93.6 89.0

Allowed 6.4 11.0
Generously allowed 0 0
Disallowed 0 0
a
Values in parentheses are for the highest-resolution shell.
b
R
sym
=
P
h
P
i
I
hi
À I
h
hijj
=
P
h
P
i
I
hi
, where I
hi
and I
h
hi

are the ith and mean
measurement of the intensity of reflection h, respectively.
c
R
work
⁄ R
free
=
P
h
F
o:h
À F
c:h
jj=
P
h
F
o:h
, where F
o.h
and F
c.h
are the observed and
calculated structure factor amplitudes, respectively.
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1129
structure contoured at 1.0r (Fig. 4C, inset). The
bound molecule adopts a half-chair conformation, and
the group bonded to C3 is orthogonal to the ring

system. As illustrated by the scheme (Fig. 4C) and the
structure-based sequence alignment (Fig. 5), the shiki-
mate is hydrogen bonded to several highly conserved
residues. In detail, the carboxylate group of shikimate
is recognized by the hydroxyl groups of Ser13 and
Ser15, as well as by the backbone amide of Leu14 via
a water molecule. The C5-hydroxyl group of shikimate
forms hydrogen bonds with the side-chain amides of
Asn58 and Asn85. The C4-hydroxyl group of shiki-
mate interacts with the carboxylate group of Asp100
and with the side-chain amides of Asn85 and Lys64.
The C3-hydroxyl group forms extensive hydrogen
bonds with the side chains of Lys64, Asp100, Thr60
and Gln239. The absolutely conserved residues Lys64
and Asp100 are also believed to be catalytically active
and responsible for the deprotonation of the
C3-hydroxyl group during the catalysis [12]. In brief,
AB
CD
Fig. 4. The overall structure and the substrate-binding domain of SeSDH in complex with shikimate. (A) The overall structure of SeSDH is
shown as a cartoon. The N-terminal substrate-binding domain is colored in orange and green, and the C-terminal NADP-binding domain is
colored in blue. The bound shikimate molecule and its binding residues are shown as stick and lines, respectively. (B) Superposition of the
N-terminal domains from all SDH structures. The structure of SeSDH is colored in blue. (C) Schematic diagram of the substrate-binding site
of SeSDH. Dotted lines represent hydrogen bonds. The asterisk beside the C3-hydroxyl group of shikimate indicates the proton to be deliv-
ered to the bulk solvent during catalysis. The proton-conducting route is represented by the arrows. The inset is the annealed omit map
around shikimate, contoured at the 1.0r level. (D) Superposition of the substrate-binding residues from all SDH structure. The structure of
SeSDH is colored in cyan, and the distance between the side chain of Tyr211 and the carboxyl of shikimate is colored in red.
Fig. 5. Structure-based sequence alignment of various SDHs. The secondary structures are from SeSDH. a-helices are represented as squig-
gles, 3
10

helices are marked with g, b-strands are rendered as arrows and b-turns are shown as TT. The blue and green numbers beneath
the alignment indicate substrate-binding residues and NADP-binding residues, respectively. The parenthesized 211 indicates that the con-
served residue Tyr211 does not interact with the substrate in the SeSDH structure. The figure was prepared using the program ESPript. The
sequence alignment was created using the following sequences from the Protein Data Bank: SeSDH(3DON), GkSDH(2EGG), MjSDH(1NVT),
EcSDH(1NYT), TtSDH(1WXD), AaSDH (2HK8), HiSDH(1P74), AtSDH(2GPT), EcYdiB(1NPD), CgYdiB(2NLO) and HiSDHL(1NPY).
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1130 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1131
Lys64 serves as a general base to abstract a proton
from the C3-hydroxyl group and transfer it to the
adjacent Asp100, which subsequently delivers the pro-
ton to the bulk solvent via structurally conserved water
molecules to refresh the enzyme. The proton-conduct-
ing route is represented by the arrows in Fig. 4C.
The conserved residue Tyr211 participates in sub-
strate binding in the other structures of SDH in com-
plex with shikimate, and is believed to be the ionizable
group with a pK
a
of 9.7 that must be protonated for
catalysis [12]. The distances between the conserved
residue, Tyr211, and shikimate in the active sites of
various SDH structures are shown in Table 3.
Remarkably, the phenol hydroxyl of Tyr211 is not
within hydrogen bond distance from the carboxyl
oxygen of shikimate (5.7 A
˚
) in the binary SeSDH–
shikimate complex structure (Fig. 4D). However, the

Y211F mutation results in a remarkable reduction in
enzyme activity, which indicates that the phenol
hydroxyl of Tyr211 still interacts with shikimate to
stabilize the catalytic intermediate, playing an essential
role in the catalytic process.
NADP-binding domain
The C-terminal NADP-binding domain of SeSDH is
less conserved than the substrate-binding domain. The
rmsd values obtained from superposing various SDH
NADP-binding domains ranged from 0.91 to 5.84 A
˚
.
Figure 6A shows the structural superposition of the
NADP-binding domains from SeSDH and AaSDH.
The latter represents the common fold of NADP-bind-
ing domain in SDH structures. There are three obvious
differences observed from the superposition. First, the
long helix interacting with the adenine ribose 2-phos-
phate of NADP in AaSDH is substituted with a short
3
10
helix (g4) in SeSDH. Remarkably, the short helix
has a high temperature factor in both ligand-free
(46.3 A
˚
2
versus the overall 22.3 A
˚
2
) and shikimate-

complexed (36.7 A
˚
2
versus the overall 28.1 A
˚
2
) struc-
tures, indicating the flexibility of the region. The ‘basic
patch’ for binding the adenine ribose phosphate
(NRTXXR ⁄ K motif, residue 148–153), which endows
SDH with the specificity for NADP over NAD [9], is
located at the helix g4 and at the nearby b8–g4 loop.
Second, a helix packing at the outer side of the central
b-sheet in AaSDH is replaced with two short helices
(a6 and g5) in SeSDH. Third, the part in SeSDH cor-
responding to the b10–b11 loop of AaSDH is disor-
dered. This region in AaSDH is found immediately
after the residue that helps to sandwich the NADP
adenine, and thus the flexibility of this part in SeSDH
is probably caused by the absence of NADP. However,
despite these differences, the key NADP-binding
motifs are still structurally conserved in all reported
SDH structures (Fig. 6A, inset), including the basic
patch, the adenine sandwich, the nicotinamide-binding
residues and the glycine-rich loop (GAGGA motif, res-
idue 124–128) recognizing the pyrophosphate and the
adenine ribose 3¢-hydroxyl of NADP.
Based on this observation, we modeled the NADP
molecule from the superposed AaSDH structure into
the C-terminal domain of SeSDH to check whether the

conformation of this domain is appropriate for binding
NADP. As shown in Fig. 6B, the glycine-rich loop and
the nicotinamide-binding residues, as well as Asn148
and Thr150 within the basic patch, are in the correct
positions to form hydrogen bonds with their binding
partners of NADP. However, the basic patch residue
Arg149 deviates away from the adenine ribose phos-
phate, whereas Arg153 collides with it. Furthermore,
the two residues Arg149 and Pro184 flanking the ade-
nine are too far away from each other to form a sand-
wich structure. As the nicotinamide nucleotide and the
pyrophosphate of NADP are properly anchored in the
C-terminal domain of SeSDH, the adenosine moiety is
unlikely to adopt a different orientation, implying that
the basic patch and the adenosine sandwich structures
of SeSDH might undergo conformational change upon
NADP binding. Actually, the loop following residue
Pro184 is completely disordered in both ligand-free
and shikimate-complexed SeSDH structures, further
indicating the flexibility of the region. Similarly,
NADP-induced conformational changes are indicated
Table 3. The distance (A
˚
) between the conserved tyrosine residue
and shikimate in the active sites of various SDH structures.
Tyrosine
numbering
Tyr-Ca –
shikimate-C1
Tyr-OH –

the nearest
carboxyl
oxygen of
shikimate
Binary complex (shikimate)
SeSDH 211 12.7 5.7
AtSDH (2GPT_A) 550 9.1 2.5
AtSDH (2O7Q_A) 550 9.1 2.5
TtSDH (2D5C_A) 207 10.6 2.8
TtSDH (2D5C_B) 207 8.7 6.4
a
Ternary complex
AaSDH (2HK9_D, active) 216 8.1 2.5
AaSDH (2HK9_A, inactive) 216 9.3 2.8
AtSDH (2O7S_A, active) 550 8.9 2.7
TtSDH (2EV9_A, inactive) 207 11.2 3.7
a
The tyrosine has flipped its side chain to interact with the 3-hydro-
xyl of shikimate via a water molecule. Its Ca atom still remains
close to the carboxyl of shikimate.
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1132 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
by the structures of apo-H. influenzae shikimate dehy-
drogenase (HiSDH) and its complex with NADPH
(Fig. 6B, inset) [10].
Open and closed conformational change
Two overall structural states of the SDH structure –
the open form and the closed form – have been
reported [9,12,13]. Table 4 summarizes the key features
of various SDH structures.

The ‘openness’ of the molecule could be evaluated
by the interdomain angle, which is defined as the angle
among the centroids of the two domains and the Ca
atom of the conserved hinge residue aspartate (Asp100
in SeSDH). As shown in Table 4, the interdomain
angle of SeSDH is the widest among all reported SDH
structures, indicating that the structure represents the
most ‘open’ form of SDH. There are two distinct
structural features related to such ‘openness’ of
SeSDH.
First, the Ca distance between the central glycine of
the conserved NADP-binding motif GAGGA (Gly126
in SeSDH) and the catalytic lysine residue (Lys64 in
SeSDH) is $ 14 A
˚
, much larger than the correspond-
ing distances in the active ternary complexes ($ 8A
˚
)
and the reported ‘closed’ structures (< 11.2 A
˚
). It is
A
B
Fig. 6. The NADP-binding domain of SeSDH. (A) The superposition of the NADP-binding domains of SeSDH (cyan) and AaSDH ternary com-
plex (orange). The NADP molecule from AaSDH is shown as a stick. The inset represents the superposition of the NADP-binding domains
from all SDH structures. SeSDH is colored in cyan. The conserved NADP-binding motifs are colored in blue. (B) The NADP molecule from
the superposed AaSDH is modeled into the NADP-binding domain of SeSDH. The residues colored in cyan are not in the correct positions
to interact with NADP. The inset is the superposition of apo-(yellow) and NADP-bound (cyan) HiSDH structures. The putative conformational
change occurring upon NADP binding is indicated by the arrow.

C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1133
also larger than those in most of the ‘open’ and native
forms of SDH. The feature indicates that the NADP-
binding motif is far away from the shikimate-binding
site in the SeSDH structure, representing an unfavor-
able state for the hydride transfer between NADP and
shikimate. Thus, we conclude that a closing process
between the two domains of SeSDH might occur upon
NADP approaching shikimate during catalysis.
Second, the Ca distances between the conserved
tyrosine residue (Tyr211 in SeSDH) and the two
serine residues interacting with the carboxyl of
shikimate (Ser13 and Ser15 in SeSDH) are larger
than those in other SDH structures, indicating that
Tyr211 of SeSDH does not interact with substrate in
the structure.
To investigate in greater detail the open–close mech-
anism of SeSDH, we superposed the substrate-binding
domain of the SeSDH structure with that of the active
AaSDH ternary complex structure. As shown in
Fig. 7, the NADP-binding domain of SeSDH is
located further away from the substrate-binding
domain than that of AaSDH. The deviation starts
from the conserved hinge residue Asp100 at the begin-
ning of helix a4 in SeSDH, which coincides with the
centroid of the molecule (Fig. 7A). Superposition of
the closed and open forms of AaSDH by their sub-
strate-binding domains reveals that the departure of
their NADP-binding domains also begins from the

corresponding conserved residue Asp106 (Fig. 7A,
inset). Detailed inspection of the two superposed
NADP-binding domains shows that the key NADP-
binding motifs in SeSDH deviate from those of
AaSDH in the same direction (Fig. 7B–D), which
implies that these motifs of SeSDH might be able to
‘pivot’ to their competent NADP-binding positions
upon the closure of the molecule via the hinge residue.
B-factor analysis of all the 14 SDH structures listed in
Table 4 also indicated that the temperature factor of
the NADP-binding domain is usually higher than that
of the substrate-binding domain, especially at the heli-
ces containing the NADP-binding motifs, suggesting
that the NADP-binding domain of SDH might be
more prone to conformational changes than the
substrate-binding domain.
Discussion
To date, many studies of SDH have revealed various
intriguing parts of the medically important target
enzyme [8–16], including 20 crystal structures of vari-
ous SDHs in their ligand-free form and in complex
with either substrate or cofactor or both, enzymatic
kinetics for catalysis, a detailed catalytic mechanism of
SDH underlying the hydride transfer between NADP
and substrate coupled with the deprotonation of the
3-hydroxyl group of shikimate and the transfer route
of the abstracted proton to the bulk solvent [12], and
the relationship between the catalytic activity and the
open–closed conformations of SDH. The catalytic
properties and substrate specificity of SeSDH demon-

strate that SeSDH belongs to the AroE enzyme family
and can utilize NAD as a cofactor in vitro. The pH
profile of SeSDH indicates that the basic condition
may be suitable for the ionization of the catalytic
Table 4. Key features of various SDH structures.
Interdomain
angle (°)
Gly126 fi
Lys64
a
Tyr211 fi
Ser13
Tyr211 fi
Ser15 B
CD-ND
b
⁄ B
overall
(%)
SeSDH Native 157.8 14.2 13.3 14.3 30.1
SeSDH Shikimate 164.7 14.4 13.9 14.9 32.3
AaSDH (2HK9_D)
c
Ternary (active) closed
d
127.3 8.5 9.3 10.9 21.9
AaSDH (2HK9_A) Ternary (inactive)
e
undefined 139.3 9.8 9.5 11.3 15.5
AaSDH (2HK8_F) Native, open 149.4 12.1 11.3 12.5 7.2

AtSDH (2O7S) Ternary (active) undefined 120.9 8.2 9.1 11.4 42.6
TtSDH (2EV9_A) Ternary (inactive) closed 134.5 11.2 11.3 13.1 )29.3
TtSDH (2CY0_B) Native, open 142.2 12.5 11.7 13.6 14.0
EcSDH (1NYT_A) NADP, closed 146.8 8.4 10.7 12.7 16.2
EcSDH (1NYT_C) NADP, open 139.9 12.2 13.0 14.5 0.6
HiSDH (1P77_A) NADP, undefined 153.8 12.2 11.6 13.1 0.0
HiSDH (1P74_A) Native, undefined 144.2 14.6 13.6 15.0 )0.8
GkSDH (2EGG_A) Native, undefined 147.9 13.7 13.3 14.7 24.3
MjSDH (1NVT_A) Native, undefined 132.9 13.5 12.6 13.8 41.5
a
The distance (A
˚
) between the Ca atoms of the two residues in SeSDH or their counterparts in the other structures.
b
The difference
between the B factor of the N-terminal domain and that of the C-terminal domain.
c
The PDB code and chain name.
d
The reported
closed ⁄ open state.
e
The structure contains a disordered NADP nicotinamide and thus is considered inactive.
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1134 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
residues, for the binding affinity of substrate and for
the stability of the enzyme. The crystal structures of
the ligand-free and shikimate-complexed SeSDH offer
incremental knowledge for understanding the catalytic
mechanism of SDH, especially the role of the

conserved residue Tyr211 formerly regarded important
for substrate binding, the conformational change that
might occur upon NADP binding and the requirement
A
b
B
CD
Fig. 7. The structure superposition of SeSDH complexed with shikimate (cyan) and the AaSDH ternary complex (orange) by their N-terminal
domains. (A) View from the back of the superposed a4 helices. The hinge residue Asp100 is shown as a stick. The centroids are repre-
sented as spheres. SKM represents shikimate. The inset shows the superposition of the open (magenta) and closed (orange) forms of
AaSDH. (B–D) The comparison of the key NADP-binding motifs from SeSDH and AaSDH.
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1135
of a ‘closing’ process during catalysis as well as its
putative underlying mechanism.
The overall folding of SeSDH is similar to that
of other SDH structures, comprising an N-terminal
substrate-binding domain and a C-terminal NADP-
binding domain. On the basis of gel-filtration and
crystallographic analyses, we conclude that SeSDH
exists as a monomer in both crystal and solution
states.
The substrate-binding domain of SeSDH is struc-
turally conserved in all three classes of SDH. The
substrate forms hydrogen bonds with several polar
residues in the domain. Structure-based sequence
alignment (Fig. 5) shows that most hydroxyl-binding
residues are conserved among all three SDH classes,
including Thr60, Asn85, Gln239 and the two catalytic
active residues Lys64 and Asp100. The carboxyl-bind-

ing residues Ser13, Ser15 and Tyr211 are conserved in
both AroE and YdiB, but not in SDHL. The most
noteworthy residue is Tyr211. As shown in Table 3, in
all structures of SDH complexed with shikimate
(except for SeSDH), the tyrosine residue is hydrogen
bonded to the carboxyl oxygen of shikimate. By con-
trast, in the current structure of SeSDH complexed
with shikimate, the tyrosine residue is too far away
from the substrate to form a hydrogen bond. The
Y211F mutation leads to a significant decrease in the
k
cat
value by a factor of 345 but a minor increase in
the K
m
value for shikimate by a factor of 3. The above
results taken together indicate that the main function
of Tyr211 may be to stabilize the catalytic intermediate
during catalysis. Thus, a ‘closing’ process of SeSDH is
expected to occur upon NADP binding, which might
lead to the binding of the Tyr211 residue to shikimate.
In fact, we may find that the Ca distances between the
tyrosine residue and the other two shikimate-carboxyl-
binding serine residues ($ 9 and $ 11 A
˚
, respectively),
in the two confirmed active conformations of SDH
[AaSDH and A. thaliana shikimate dehydrogenase
(AtSDH)], are substantially smaller than those in other
SDH structures (Table 4), which further supports our

speculation.
The NADP-binding domain of SeSDH is less struc-
turally conserved than the substrate-binding domain,
especially in view of the fact that a common long helix
bearing a basic patch for binding the adenine ribose
phosphate is substituted with a small 3
10
helix g4in
SeSDH. Moreover, compared with the NADP-bound
C-terminal domain of AaSDH, we found that the
domain of SeSDH will be appropriate for binding
NADP if some small conformational changes occur.
These changes included the side-chain swing of the two
arginine residues constituting the basic patch and the
closing-up of the adenine sandwich. Actually, such
NADP-induced conformational changes are indicated
by the HiSDH structures [10], and are also suggested
by homology modeling and CD of MtSDH [20]. The
high temperature factor values of the SeSDH
NADP-binding domain also imply their predisposition
to conformational changes. By contrast, the structure
of SeSDH complexed with shikimate is almost the
same with its apo form, indicating that no conforma-
tional change could be induced by substrate alone.
It has been widely accepted that SDH has two over-
all conformations, namely the open form and the
closed form, and the closed form is necessary for catal-
ysis [9,12,13]. Michel et al. suggested that E. coli shiki-
mate dehydrogenase (EcSDH) switches from the open
to the closed conformation upon substrate binding [9].

Bagautdinov et al. reported that both forms of TtSDH
exist in solution, and the cofactor only binds to the
closed form while the substrate binds to both forms.
Thus, they suggested a random Bi–Bi reaction mecha-
nism of TtSDH [13].
The structure of SeSDH represents the most open
form of all known SDH structures, as indicated by its
largest interdomain angle (Table 4). Accordingly, the
distance ($ 14 A
˚
) between the NADP binding ‘GAG-
GA’ motif and the catalytically active residue Lys64 in
SeSDH is the largest among all SDH structures,
almost twice as large as those in the active ternary
complexes ($ 8A
˚
). Such a distance is directly propor-
tional to the interdomain angle of the enzyme. Thus,
we could expect the occurrence of a ‘closing’ process
of SeSDH, probably via the hinge residue, Asp100, to
bring NADP close to the substrate for catalysis. The
binding of shikimate alone does not cause any signifi-
cant structural change of SeSDH, and the C-terminal
domain is more prone to rearrangement than the
N-terminal domain in SeSDH (as reflected by the
temperature factor). Thus, we believe that the ‘closing’
process might be induced by NADP binding and
stabilized by shikimate binding.
In summary, we determined the crystal structures
and enzymatic properties of SeSDH. The combina-

tion of the crystallographic and enzymatic analysis
sheds more light on the role of key amino acid resi-
dues and the catalytic mechanism. Notably, the
Y211F mutation proves that Tyr211 plays a vital role
for the oxidation of shikimate, although it has no
interaction with shikimate in the crystal structure of
the SeSDH–shikimate complex. We speculate that the
open to closed conformational change of SeSDH
might occur upon NADP binding, which needs to be
identified by analysis of the ternary complex structure
of SeSDH–shikimate–NADP(H).
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1136 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
Experimental procedures
Materials
S. epidermidis strain RP62A was obtained from the Ameri-
can Type Culture Collection. All chemicals and reagents
were of reagent grade or ultrapure quality, and are
commercially available.
Preparation of wild-type and mutant SeSDH
Based on the genome sequence of S. epidermidis strain
RP62A (GenBank accession number: NC_002976), two
PCR primers (forward: 5¢-GCGCATC
CATATGAAATTT
GCAGTAATTGG-3¢ and reverse: 5¢-CCG
CTCGAGTAA
TTCTCCTTTCAATTTTTG-3¢) were designed to amplify
the aroE gene on the chromosome of S. epidermidis strain
RP62A. The PCR products were digested using restriction
endonucleases NdeI and XhoI (Fermentas, Burlington,

Ontario, Canada) and then cloned into a prokaryotic
expression vector pET22b (Novagen, Madison, WI, USA)
to produce the recombinant plasmid pET22b–SeSDH,
containing a C-terminal six-histidine tag, for purification
purposes. The recombinant clone pET22b–SeSDH was
verified by DNA sequencing.
The recombinant clone pET22b–SeSDH was transformed
into E. coli strain BL21(DE3) (Stratagene, La Jolla, CA,
USA) and cultured at 37 °C in Luria–Bertani (LB) medium
supplemented with 100 lgÆmL
)1
of ampicillin. When the
A
600
reached 0.6, the culture was induced by 0.4 mm iso-
propyl thio-b-d-galactoside (IPTG) and incubated at 25 °C
for an additional 6 h. The cells were harvested by centrifu-
gation and suspended in buffer A (20 mm Tris ⁄ HCl, pH
8.0, 500 mm NaCl, 10 mm imidazole). After sonication on
ice, the mixture was centrifuged to yield a clear superna-
tant, which was loaded onto a column containing Ni-nitril-
otriacetic acid resin (Qiagen, Hilden, Germany)
pre-equilibrated in buffer A. The column was washed with
buffer B (20 mm Tris ⁄ HCl, pH 8.0, 500 mm NaCl, 20 mm
imidazole) for several times and eluted with buffer C
(20 mm Tris ⁄ HCl, pH 8.0, 500 mm NaCl, 200 mm imidaz-
ole), then the eluted fractions were pooled and dialyzed
against buffer D (20 mm Tris ⁄ HCl, pH 8.0, 150 mm NaCl)
for enzymatic assay and crystallization trials. The samples
were pooled and concentrated by ultrafiltration using an

Amicon centrifugal filter device. All purification, dialysis
and concentration procedures were performed at 4 °C. Pro-
tein concentration was determined by the Bradford assay
using BSA as standard.
The oligomeric state was studied by gel-filtration chroma-
tography using an AKTA purifier system (GE Healthcare,
Uppsala, Sweden). The sample of protein was chromato-
graphed on a HiLoad 16 ⁄ 60 Superdex 75 prepacked grade
column with buffer D at a flow rate of 1 mL Æ min
)1
. The
Superdex 75 column was precalibrated with the following
low-molecular-mass standards: BSA (67.0 kDa), ovalbumin
(43.0 kDa), chymotrypsinogen A (25.0 kDa) and ribonucle-
ase A (13.7 kDa) (GE Healthcare, Uppsala, Sweden).
Site-directed mutagenesis was performed with the
QuikChange site-directed mutagenesis kit (Stratagene) using
the plasmid pET22b–SeSDH as template. The primers for
the Y211F mutant were: (sense) 5¢-GTAAGTGATATTGT
TTTTAATCCATATAAAACACC-3¢ and (antisense) 5¢-G
GTGTTTTATATGGATTAAAAACAATATCACTTAC-3¢.
The mutant plasmid was confirmed by DNA sequencing
before transformation into E. coli strain BL21(DE3). The
expression and purification of the Y211F mutant were the
same as that of the wild-type enzyme.
Preparation of selenomethionyl SeSDH
Selenomethionine-substituted SeSDH (SeMet-SeSDH) was
generated in the methionine auxotrophic E. coli strain
B834(DE3) (Novagen). The strain transformed with
pET22b–SeSDH was grown at 37 °C in M9 minimal med-

ium, which contained 40 mgÆL
)1
of all of the amino acids
except methionine. A selenomethionine (Acros, Geel,
Belgium) stock solution of 10 mgÆmL
)1
in water was
added to the medium to a concentration of 40 mgÆL
)1
.
The SeMet-SeSDH protein was induced with 0.4 mm
IPTG for about 12 h, and purified using the same method
as for the native protein. Finally, the SeMet-SeSDH frac-
tions were pooled and dialyzed against buffer E (20 mm
Tris ⁄ HCl, pH 8.0, 150 mm NaCl, 5 mm dithiothreitol and
0.2 mm EDTA) for crystallization trials. ES-MS analysis
of the SeMet-SeSDH indicated substitution of all six
methionine residues.
Enzymatic activity assay
The enzymatic activities of SeSDH and its Y211F mutant
were assayed at 25 °C by monitoring the reduction of
NADP (or NAD) at 340 nm (e
340
= 6180 m
)1
Æcm
)1
) in the
presence of shikimate. All assays were conducted in a
96-well microplate spectrophotometer (Beckman Coulter

DTX880, Fullerton, CA, USA). The assay mixture
contained shikimate and NADP (or NAD) at the desired
concentrations in 100 mm Tris–HCl (pH 8.0). The K
m
and
V
max
values for substrate and cofactor were determined by
varying the concentrations of one ligand while keeping the
other at saturation. In the experiment where shikimate was
varied (0.0625, 0.125, 0.25, 0.5, 1 and 2 mm), the concentra-
tion of NADP was maintained at 2 mm, whereas the con-
centration of shikimate was fixed at 2 mm when NADP
was varied (0.0625, 0.125, 0.25, 0.5, 1 and 2 mm). The assay
reaction was initiated by the addition of diluted enzyme
solution. To measure the kinetic parameters for NAD, the
concentration of shikimate was fixed at 2 mm when NAD
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1137
was varied (0.5, 1, 2, 4, 8 and 16 mm). The kinetic parame-
ters K
m
and V
max
were calculated from the slope and inter-
cept values of the linear fit in a Lineweaver–Burke plot. To
test the enzymatic activity of SeSDH in the presence of qui-
nate, the assay solution consisted of 100 mm Tris ⁄ HCl (pH
8.0), 4 mm quinate and 2 mm NADP (or NAD). Each
measure was taken in triplicate.

The effects of pH on the enzymatic activity of SeSDH
were determined using the above assay method. All the
assay solutions contained 2 mm shikimate and 2 mm
NADP. The enzymatic kinetics of SeSDH was measured in
buffers of different pH values (50 mm Bistris ⁄ HCl for pH
5.0–7.0, Tris ⁄ HCl for pH 8.0–9.0 and Caps ⁄ NaOH for pH
10–12). All assays were conducted three times.
Crystallization and data collection
All crystallization trials were performed at 277 K using the
hanging-drop vapor-diffusion method. Initial crystals of
SeSDH were obtained using commercially available crystal-
lization screen kits (Hampton Research Crystal Screens 1
and 2, Aliso Viejo, CA, USA). After several rounds of opti-
mization, crystals of both native SeSDH and SeMet-SeSDH
were grown in drops containing equal volumes of the pro-
tein solution (20 mgÆmL
)1
) and the crystallization solution
(0.1 m tri-sodium citrate dihydrate at pH 5.8, 24%
PEG4000, 0.2 m ammonium acetate) to the maximal size
after 30 days. The crystals were harvested and cryopro-
tected with the reservoir solution containing up to 10%
(v ⁄ v) glycerol. To generate the SeSDH–shikimate complex,
protein samples were incubated with 5 mm shikimate at
277 K for 12 h prior to crystallization. Crystals of SeSDH
complexed with shikimate were obtained using a different
crystallization buffer (0.1 m sodium cacodylate at pH 6.5,
23% PEG 8000, 0.2 m ammonium sulfate). The crystals of
SeSDH in complex with shikimate were cryoprotected by
including 30% (v ⁄ v) glycerol in the mother liquor. Each

crystal was picked up with a nylon loop and flash-cooled in
liquid nitrogen before data collection. The crystal
diffraction data for the crystals of both apo-SeSDH and
SeSDH complexed with shikimate were collected at 100 K
on a Rigaku rotating-anode X-ray generator operated at
50 kV and 100 mA (k = 1.5418 A
˚
) with a Rigaku R-AXIS
IV++ imaging-plate detector. The data were processed
using the crystalclear program [21]. For the crystal of
SeMet-SeSDH, a multiple-wavelength anomalous diffrac-
tion data set was collected on beamline BL-6A of the Pho-
ton Factory (Tsukuba, Japan) with an ADSC Quantum 4R
CCD detector using three wavelengths (peak = 0.97901 A
˚
;
edge = 0.97953 A
˚
; and high-energy remote = 0.96429 A
˚
).
The three wavelengths were determined by a fluorescence
scan for selenium. The multiple-wavelength anomalous
diffraction data were processed using the HKL-2000 suite
of programs [22]. The statistics of data collection are
summarized in Table 2.
Structure solution and refinement
For phase determination, SOLVE was used to locate all of
the six selenium sites and to calculate the phases [23].
RESOLVE was used for density modification and initial

model building [24]. The complete model was built using
COOT [25]. Refinement of the structure was carried out
using CNS [26] and Refmac from the CCP4 suite [27].
procheck was used to check the stereochemical quality of
the final structural model [28]. All figures and superpositions
were made using pymol [29]. Centroids of the structures
were calculated using PDBSET (http://www.
ccp4.ac.uk/html/pdbset.html). The crystal structures of
apo-SeSDH and of SeSDH in complex with shikimate were
solved by molecular replacement using the SeMet-SeSDH
structure as a search model. The refinement statistics are
summarized in Table 2. The atomic coordinates and
structure factors have been deposited in the Protein
Data Bank under the accession numbers 3DON and 3DOO.
Acknowledgements
We thank the staff members at the Photon Factory of
Japan for technical support in diffraction data collec-
tion, Dr Yanhui Xu for assistance with gel-filtration
experiments and Dr Xinhua Ji for critical reading
of the manuscript. This work was supported by the
State Key Program of Basic Research of China
(grant 2002CB512803), the International Science and
Technology Cooperation Program of China (grant
2006DFA32760), the National Natural Science
Foundation of China (grant 30800036), and China
Postdoctoral Science Foundation Funded Project
(grants 200801172 and 20060400574).
References
1 Vuong C & Otto M (2002) Staphylococcus epidermidis
infections. Microbes Infect 4, 481–489.

2 Costerton JW, Stewart PS & Greenberg EP (1999)
Bacterial biofilms: a common cause of persistent
infections. Science 284, 1318–1322.
3 Parish T & Stoker NG (2002) The common aromatic
amino acid biosynthesis pathway is essential in Myco-
bacterium tuberculosis. Microbiology 148, 3069–3077.
4 Coggins JR, Abell C, Evans LB, Frederickson M, Rob-
inson DA, Roszak AW & Lapthorn AP (2003) Experi-
ences with the shikimate-pathway enzymes as targets
for rational drug design. Biochem Soc Trans 31, 548–
552.
5 Han C, Wang L, Yu K, Chen L, Hu L, Chen K, Jiang
H & Shen X (2006) Biochemical characterization and
inhibitor discovery of shikimate dehydrogenase from
Helicobacter pylori. FEBS J 273, 4682–4692.
Shikimate dehydrogenase from S. epidermidis C. Han et al.
1138 FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS
6 Bonner CA & Jensen RA (1994) Cloning of cDNA
encoding the bifunctional dehydroquinase.shikimate
dehydrogenase of aromatic-amino-acid biosynthesis in
Nicotiana tabacum. Biochem J 302 , 11–14.
7 Duncan K, Edwards RM & Coggins JR (1987) The
pentafunctional arom enzyme of Saccharomyces cerevi-
siae is a mosaic of monofunctional domains. Biochem J
246, 375–386.
8 Singh S, Korolev S, Koroleva O, Zarembinski T, Coll-
art F, Joachimiak A & Christendat D (2005) Crystal
structure of a novel shikimate dehydrogenase from Hae-
mophilus influenzae. J Biol Chem 280, 17101–17108.
9 Michel G, Roszak AW, Sauve V, Maclean J, Matte A,

Coggins JR, Cygler M & Lapthorn AJ (2003) Structures
of shikimate dehydrogenase AroE and its paralog YdiB.
J Biol Chem 278, 19463–19472.
10 Ye S, Von Delft F, Brooun A, Knuth MW, Swanson
RV & McRee DE (2003) The crystal structure of shiki-
mate dehydrogenase (AroE) reveals a unique NADPH
binding mode. J Bacteriol 185, 4144–4151.
11 Padyana AK & Burley SK (2003) Crystal structure of
shikimate 5-dehydrogenase (SDH) bound to NADP:
insights into function and evolution. Structure 11, 1005–
1013.
12 Gan J, Wu Y, Prabakaran P, Gu Y, Li Y, Andrykov-
itch M, Liu H, Gong Y, Yan H & Ji X (2007)
Structural and biochemical analyses of shikimate
dehydrogenase AroE from Aquifex aeolicus:
implications for the catalytic mechanism. Biochemistry
46, 9513–9522.
13 Bagautdinov B & Kunishima N (2007) Crystal struc-
tures of shikimate dehydrogenase AroE from Thermus
thermophilus HB8 and its cofactor and substrate com-
plexes: insights into the enzymatic mechanism. J Mol
Biol 373, 424–438.
14 Singh SA & Christendat D (2006) Structure of Arabid-
opsis dehydroquinate dehydratase-shikimate dehydroge-
nase and implications for metabolic channeling in the
shikimate pathway. Biochemistry 45, 7787–7796.
15 Singh SA & Christendat D (2007) The DHQ-dehydros-
hikimate-SDH-shikimate-NADP(H) complex: insights
into metabolite transfer in the shikimate pathway. Cryst
Growth Des 7, 2153–2160.

16 Benach J, Lee I, Edstrom W, Kuzin AP, Chiang Y,
Acton TB, Montelione GT & Hunt JF (2003) The
2.3-A
˚
crystal structure of the shikimate 5-dehydrogenase
orthologue YdiB from Escherichia coli suggests a novel
catalytic environment for an NAD-dependent
dehydrogenase. J Biol Chem 278, 19176–19182.
17 Schoepe J, Niefind K & Schomburg D (2008) 1.6 A
˚
structure of an NAD(+)-dependent quinate dehydroge-
nase from Corynebacterium glutamicum. Acta Crystal-
logr D Biol Crystallogr 64, 803–809.
18 Lim S, Schroder I & Monbouquette HG (2004) A ther-
mostable shikimate 5-dehydrogenase from the archaeon
Archaeoglobus fulgidus. FEMS Microbiol Lett 238, 101–
106.
19 Zhang X, Zhang S, Hao F, Lai X, Yu H, Huang Y &
Wang H (2005) Expression, purification and properties
of shikimate dehydrogenase from Mycobacterium
tuberculosis. J Biochem Mol Biol 38, 624–631.
20 Arcuri HA, Borges JC, Fonseca IO, Pereira JH, Neto
JR, Basso LA, Santos DS & de Azevedo WF Jr
(2008) Structural studies of shikimate 5-dehydrogenase
from Mycobacterium tuberculosis. Proteins 72, 720–
730.
21 Pflugrath JW (1999) The finer things in X-ray diffrac-
tion data collection. Acta Crystallogr D Biol Crystallogr
55, 1718–1725.
22 Otwinowski Z & Minor W (1997) Processing of X-ray

diffraction data collected in oscillation mode. Methods
Enzymol 276, 307–326.
23 Terwilliger TC & Berendzen J (1999) Evaluation of
macromolecular electron-density map quality using the
correlation of local r.m.s. density. Acta Crystallogr D
Biol Crystallogr 55, 1872–1877.
24 Terwilliger TC (2000) Maximum-likelihood density
modification. Acta Crystallogr D Biol Crystallogr 56,
965–972.
25 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D Biol
Crystallogr 60, 2126–2132.
26 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 Crys-
tallogr 54, 905–921.
27 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the
maximum-likelihood method. Acta Crystallogr D Biol
Crystallogr 53, 240–255.
28 Laskowski RA, Macarthur MW, Moss DS & Thornton
JM (1993) Procheck: a program to check the stereo-
chemical quality of protein structures. J Appl Crystal-
logr 26, 283–291.
29 DeLano WL (2002) The PyMOL Molecular Graphics
System. DeLano Scientific, Palo Alto, CA.
C. Han et al. Shikimate dehydrogenase from S. epidermidis
FEBS Journal 276 (2009) 1125–1139 ª 2009 The Authors Journal compilation ª 2009 FEBS 1139