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
Trang 1dehydrogenase from Staphylococcus epidermidis
Implications for substrate binding and conformational change
Cong Han1,*, Tiancen Hu2,*, Dalei Wu2, Su Qu1, Jiahai Zhou3, Jianping Ding4, Xu Shen2, Di Qu1 and Hualiang Jiang2
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: dqu@shmu.edu.cn
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: xshen@mail.shcnc.ac.cn
*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 epidermidisSDH (SeSDH) both in its ligand-free form and
in complex with shikimate SeSDH has a kcat of 22.8 s)1 and a Km of
73 lm towards shikimate, and a Kmof 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 bindbind-ing 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 kcatand a minor increase
in the Km 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 310 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.
Trang 2Staphylococcus 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 anti-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 3-dehydro-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
Trang 3In this article, we report the crystal structures of
S epidermidisSDH (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 310 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 theoreti-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 Km and
Vmaxwere 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 Kmand kcatvalues 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 kcat value, a three-fold increase in the Km value and a 1073-fold decrease in the kcat⁄ Km 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.
Trang 4SDH) [18], SeSDH can also oxidize shikimate using
NAD as a cofactor, yielding a kcat of 87 ± 20 s)1, a
Km of 10.6 ± 2.6 mm and a kcat⁄ Km of
8.2· 103m)1Æs)1towards NAD SeSDH showed a Km
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 kcat(s)1) (shikimate) Km(l M ) (shikimate) Km(l M ) (NADP) kcat⁄ K m ( M )1Æs)1) (shikimate) k
cat ⁄ K m ( M )1Æs)1) (NADP)
4.23 · 10 6 a
Kinetic parameters for A aeolicus SDH are from [12].bKinetic parameters for H pylori SDH are from [5].cKinetic 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.
Trang 5(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 310 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.
Data collection
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
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)
Refinement
R work ⁄ R freec 0.183 ⁄ 0.260 0.188 ⁄ 0.264
No of atoms
B-factors
Ramachandran plot (%)
a Values in parentheses are for the highest-resolution shell b Rsym= P
h
P
i j I hi I h i h j= P
h
P
i I hi , where Ihiand I h i are the ith and mean h
measurement of the intensity of reflection h, respectively.cR work ⁄ R free = P
h j F o:h F c:h j= P
h F o:h , where F o.h and F c.h are the observed and calculated structure factor amplitudes, respectively.
Trang 6structure 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,
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, 310helices 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).
Trang 8Lys64 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 pKa 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
310helix (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)
Ternary complex
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.
Trang 9by 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 ( 8 A˚) 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.
Trang 10also 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-NADP-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
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 Lys64a
Tyr211 fi Ser13
Tyr211 fi Ser15 B CD-NDb⁄ B overall (%)
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. cThe PDB code and chain name. dThe reported closed ⁄ open state e
The structure contains a disordered NADP nicotinamide and thus is considered inactive.