Exploring the water-binding pocket of the type II dehydroquinase enzyme in the structure-based design of inhibitors

Article abstract of DOI:10.1021/jm500175z

Structural and computational studies to explore the WAT1 binding pocket in the structure-based design of inhibitors against the type II dehydroquinase (DHQ2) enzyme are reported. The crystal structures of DHQ2 from M. tuberculosis in complex with four of

Full text of DOI:10.1021/jm500175z

Article
pubs.acs.org/jmc
Exploring the Water-Binding Pocket of the Type II Dehydroquinase
Enzyme in the Structure-Based Design of Inhibitors
Beatriz Blanco,†,⊥ Antía Sedes,†,⊥ Antonio Peon,† Jose M. Otero,‡ Mark J. van Raaij,§ Paul Thompson,∥
́
́
Alastair R. Hawkins,∥ and Concepcion Gonzalez-Bello*,†
́
́
†Centro Singular de Investigacion
15782 Santiago de Compostela, Spain
‡Departamento de Bioquímica y Biología Molecular, Centro Singular de Investigacion
(CIQUS), Universidad de Santiago de Compostela, 15782 Santiago de Compostela, Spain
§Departamento de Estructura de Macromolec
ulas, Centro Nacional de Biotecnología (CSIC), Campus Cantoblanco, 28049 Madrid,
Spain
́
en Química Biolog
́
ica y Materiales Moleculares (CIQUS), Universidad de Santiago de Compostela,
́
́
en Química Biologica y Materiales Moleculares
́
∥Institute of Cell and Molecular Biosciences, Medical School, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH,
U.K.
S
* Supporting Information
ABSTRACT: Structural and computational studies to explore
the WAT1 binding pocket in the structure-based design of
inhibitors against the type II dehydroquinase (DHQ2) enzyme
are reported. The crystal structures of DHQ2 from M.
tuberculosis in complex with four of the reported compounds
are described. The electrostatic interaction observed between
the guanidinium group of the essential arginine and the
carboxylate group of one of the inhibitors in the reported
crystal structures supports the recently suggested role of this
arginine as the residue that triggers the release of the product
from the active site. The results of the structural and molecular
dynamics simulation studies revealed that the inhibitory
potency is favored by promoting interactions with WAT1
and the residues located within this pocket and, more importantly, by avoiding situations where the ligands occupy the WAT1
binding pocket. The new insights can be used to advantage in the structure-based design of inhibitors.
been focused on the inhibition of the enzymes involved in the
shikimic acid pathway, in which chorismic acid is synthesized.
These enzymes are recognized to be attractive targets for the
development of new antibacterial agents because they are
essential in important pathogenic bacteria but are absent in
mammals.9 Special attention has been paid to the inhibition of
the third enzyme of the pathway, namely, type II
dehydroquinase (3-dehydroquinate dehydratase, EC 4.2.1.10,
DHQ2), which is one of the two known DHQ enzymes.
DHQ2 is essential in Mycobacterium tuberculosis (aroD gene,
Mt-DHQ2),10 which is responsible for causing tuberculosis, and
Helicobacter pylori (aroD/aroQ gene, Hp-DHQ2),10 which is
the causative agent of gastric and duodenal ulcers and has been
classified as a type I carcinogen.
INTRODUCTION
■
The increasing development and spread of resistance to current
antibiotics have made infectious diseases the second most
common cause of death worldwide.1 Despite this alarming
trend, only two new classes of antibiotics have been brought to
the market in the past 3 decades and there are only a few
antibacterial agents under development to face the challenge of
multidrug resistance.2−5 Hence, there is a great deal of interest
in the development of alternative therapies, including new
strategies for the treatment of infections that are resistant to
current antibiotics, in particular drugs with new mechanisms of
action to combat the growth of antibiotic-resistant bacteria.6
One strategy to combat bacterial infections is based on the
disruption of the growth cycle by preventing the synthesis and
assembly of key components of bacterial processes.7,8 Many
drugs that are highly successful in human clinical use are
mimetics of the substrate or intermediates of key enzymatic
reactions. For instance, sulfonamide-based drugs are analogues
of p-aminobenzoate, the natural substrate of dihydropteroate
synthase, an essential enzyme in the folic acid biosynthesis
pathway. During the past few years, a great deal of effort has
DHQ2 catalyzes the reversible dehydration of 3-dehydro-
quinic acid (1) to form 3-dehydroshikimic acid (2). The anti
elimination of water involves the loss of the more acidic pro-S
hydrogen from C2 of 1 (Scheme 1). After substrate binding, a
Received: February 1, 2014
Published: April 1, 2014
© 2014 American Chemical Society
3494
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
Scheme 1. Reaction Mechanism for the Enzymatic
Conversion of 3-Dehydroquinic Acid (1) to 3-
Dehydroshikimic Acid (2) Catalyzed by the DHQ2 Enzyme
a
a
The reaction proceeds via enolate intermediate 3. Relevant residues
are indicated. For the Hp-DHQ2 enzyme, the generation of the
catalytic tyrosinate takes place with the assistance of a water molecule
(not shown), while for Mt-DHQ2, the tyrosine is directly
deprotonated by the aspartate residue.
flexible loop forms a lid that closes the active site. Three
residues, an arginine (Arg17/Arg19 in Hp-DHQ2 and Mt-
DHQ2, respectively), a tyrosine (Tyr22/Tyr24 in Hp-DHQ2
and Mt-DHQ2, respectively) and an aspartate (Asp89*/
Asp88* in Hp-DHQ2 and Mt-DHQ2, respectively) from the
neighboring enzyme subunit (residues from this subunit will be
marked with an asterisk), were identified by chemical
modification and site-directed mutagenesis studies as essential
for enzyme activity.11−13 The elimination proceeds by a
stepwise E1CB mechanism that involves an enolate inter-
mediate 3.14,13 The reaction is initiated by the essential
Asp89*/Asp88* (Figure 1A). This residue deprotonates the
essential Tyr22/Tyr24, which is located in the loop, to afford
the catalytic tyrosinate, which in turn triggers the enzymatic
process.13 For the Hp-DHQ2 enzyme, this reaction takes place
with the assistance of a water molecule, while for Mt-DHQ2,
the tyrosine is directly deprotonated by the aspartate residue.13
The guanidinium group of Arg17/Arg19 must be close to the
essential tyrosine side chain for catalysis because in this
arrangement it controls the appropriate orientation of the
tyrosine, which is roughly perpendicular to the cyclohexane ring
of the natural substrate, through a cation−π interaction and
contributes to decreasing the pKa of the phenol group.15 The
tyrosinate removes the axial hydrogen from C2 of 1 to give an
enolate intermediate 3 (Figure 1B), which is stabilized by the
positively charged side chain of the essential Arg17/Arg19 and
by a hydrogen bond between the neutral aspartate Asp89*/
Asp88* and a conserved water molecule (henceforth WAT1).
The final step is the acid-catalyzed elimination of the C1
hydroxyl group, a reaction mediated by the conserved histidine
His102/His101 (in Hp-DHQ2 and Mt-DHQ2, respectively)
acting as a proton donor.
Figure 1. View of the Mt-DHQ2 Michaelis complex (A) and enolate
intermediate 3 geometries (B) obtained by MD and QM/MM studies.
Note how Asp88* from a neighboring enzyme subunit is responsible
for the generation of the catalytic form of the essential tyrosine
(tyrosinate) and that the enolate intermediate 3 is stabilized by the
guanidinium group of Arg19 and hydrogen bonding with WAT1 and
the neutral Asp88*. Relevant residues and water molecules are
indicated and labeled. Hydrogen bonds (red) and key distances
(black) are shown.
the main-chain amide of a glycine or an alanine (Ala79/Gly78
in Hp-DHQ2 and Mt-DHQ2, respectively). Initially, it was
proposed that this water molecule would be involved in the
protonation of the enolate intermediate 3, assuming that the
side chain of a conserved asparagine (Asn10/Asn12 in Hp-
DHQ2 and Mt-DHQ2, respectively) would be deprotonated by
WAT1.16 However, considering the high energy required to
deprotonate the side chain of an asparagine and on the basis of
results from quantum mechanical (QM) calculations, this
possibility was then discarded.17 Recently, the results from
computational studies of the catalytic mechanism (Hp-DHQ2
and Mt-DHQ2) in atomic detail by means of a hybrid QM/
MM approach suggested that WAT1 would be mainly involved
in the stabilization of the substrate and enol intermediate 3.17,13
Furthermore, the crystal structure of Mt-DHQ2 in complex
with a nanomolar reversible competitive inhibitor, compound
The presence of WAT1 in many crystal structures and its
role in the enzymatic mechanism have been widely discussed.
This water molecule interacts through hydrogen bonding with a
conserved asparagine (Asn10/Asn12 in Hp-DHQ2 and Mt-
DHQ2, respectively), the carbonyl group of a conserved proline
(Pro9/Pro11 in Hp-DHQ2 and Mt-DHQ2, respectively), and
3495
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
4c, solved at 1.5 Å shows that the oxygen atom of the
methylenoxy spacer of 4c is located 3.1 Å away from WAT1
(Figure 2).18 The aromatic moiety of these compounds blocks
ligand side chains. Moreover, molecular dynamics (MD)
simulation studies were carried out to study the binding
modes of ligands 9 and 10 and to gain a better understanding
of the binding determinants that enhance the inhibition
potency.
RESULTS AND DISCUSSION
■
Synthesis of Compounds 5 and 6. The synthesis of the
target compound 5 was performed from the previously
reported allylic bromide 12,25 which can be prepared in five
steps from commercially available (−)-quinic acid (11)
(Scheme 2). First, nucleophilic substitution of allylic bromide
a
Scheme 2. Synthesis of Compounds 5 and 6
Figure 2. Selected examples of 3-methoxyaryl derivatives 4 that are
DHQ2 competitive inhibitors and new target compounds 5−10. A
selected view of the crystal structure of the Mt-DHQ2/4c binary
complex (PDB entry 2Y71, 1.5 Å) is shown along with the Ki values of
4 against Mt-DHQ2.
the entry of the essential arginine side chain into the active site
and causes a significant change in the conformation and
flexibility of the loop, which is essential for the enzymatic
activity. Considering that the replacement of the oxygen atom
of the methylenoxy spacer of 4 by a carbon atom (ethylene
spacer) leads to a decrease in the inhibition potency of up to
20-fold and that both spacers provide the necessary conforma-
tional flexibility to the aryl moiety to accommodate it in the
active site in an extended conformation,19 we assume that an
important contribution of the high potency of the inhibitor,
which has Ki values of 42 18 and 130 nM 20 against Mt-DHQ2
and Hp-DHQ2, respectively, is due to the interaction between
the oxygen atom of the methylenoxy spacer with WAT1.
Structural water molecules in enzyme active sites are of
interest to medicinal chemists because they can increase the
binding affinity of a ligand to a protein and can provide new
directions for the design of more potent inhibitors.21 For
instance, this approach has been successfully used in the
development of potent HIV protease inhibitors such as a cyclic
urea reported by Lam et al.22 or the pyrones described by
Romines et al.23 or Vara Prasad et al.24 Taking into account the
important stabilizing contribution that the interaction between
WAT1 and either the natural substrate, the enolate
intermediate 3, or enolate mimetics like compound 4c seem
to have in their binding with the DHQ2 enzyme, we decided to
explore the WAT1 binding pocket in the structure-based design
of inhibitors of these enzymes. The possible enhanced binding
affinity of mimetics of the enolate intermediate 3 was studied
by promoting interactions with WAT1 or by displacement of
WAT1, causing in the latter case a direct interaction with the
active site residues. To this end, we report here the synthesis of
new reversible competitive inhibitors of DHQ2 from H. pylori
and M. tuberculosis, namely, compounds 5−10 (Figure 2), along
with the results of inhibition studies (Figure 2). The crystal
structures of DHQ2 from M. tuberculosis in complex with
compounds 5−8 are also described. These structures show in
detail the binding modes of these inhibitors and highlight the
importance of the stereochemistry of the chiral center of the
a
Reagents and conditions: (a) ref 26; (b) NaOAc, DMF, rt; (c) TFA/
H2O (20:1), 0 °C; (d) (1) LiOH (aq), THF, rt; (2) Amberlite IR-120
(H+), rt; (e) ref 21; (f) vinylboronic acid pinacol ester, Pd(PPh3)4,
K2CO3 (aq), dioxane, Δ; (g) (1) BH3−THF, 0 °C; (2) NaBO3 (aq);
(h) TBAF, THF, rt.
12 with sodium acetate in DMF gave acetate 13. Removal of
the diacetal protecting group by treatment with aqueous
trifluoroacetic acid afforded triol 14, which was converted into
the desired acid 5 by basic hydrolysis of the methyl ester 14 and
subsequent protonation with Amberlite IR-120 (H+) ion-
exchange resin.
The synthesis of compound 6 was achieved by a Suzuki
cross-coupling reaction between our previously reported vinyl
triflate 1526 and vinylboronic acid pinacol ester as the key step,
followed by selective hydroboration−oxidation of the external
double bond (Scheme 2). The cross-coupling reaction between
vinyl triflate 15 and vinylboronic acid pinacol ester was
performed using Pd(PPh3)4 as the catalyst in the presence of
aqueous potassium carbonate to give the cross-coupling
product 16 in excellent yield. Treatment of diene 16 with
borane tetrahydrofuran complex and subsequent reaction with
sodium perborate gave alcohol 17. Removal of the TBS groups
with TBAF gave the triol 18, which was converted to the
3496
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
desired acid 6 by basic hydrolysis and protonation with an ion-
exchange resin.
ketones 19 and 20 gave a chromatographically separable
mixture of alcohols in which the R-isomer was the major
diastereoisomer in both cases. The stereochemistry of the
reduction was confirmed by NOE experiments. Finally, the
resulting lactones 23 and 24 were converted to the desired
acids 7 and 8, respectively, in the same way as compound 6
from hydroxyl lactone 18.
The desired chloro derivatives 9 and 10 were prepared from
the corresponding R-hydroxyl and S-hydroxyl derivatives 22
and 21, respectively (Scheme 3). Treatment of alcohol 22 with
triphenylphosphine and N-chlorosuccinimide afforded chloride
27, which was converted to the desired acid 9 by deprotection
of the TBS groups with TBAF followed by basic hydrolysis of
the corresponding lactone 28 and protonation with an ion-
exchange resin. Finally, compound 21 was converted to the
desired acid 10 in the same way as acid 9 from hydroxyl lactone
22.
Synthesis of Compounds 7−10. The synthesis of the
target compounds 7−10 involved the preparation of the α,β-
unsaturated ketone 19, which was obtained by a carbonylative
Suzuki cross-coupling reaction between vinyl triflate 1526 and
commercially available B-benzyl-9-BBN using Pd(dppf)Cl2 as
catalyst and K3PO4 as base (Scheme 3). The desired hydroxyl
a
Scheme 3. Synthesis of Compounds 7−10
Inhibition Assay Results. The inhibitory properties of
compounds 5−10 against Mt-DHQ2 and Hp-DHQ2 were
tested. These compounds proved to be reversible competitive
inhibitors of both enzymes. The inhibition data (Ki), which
were obtained from Dixon plots (1/v vs [I]), are summarized in
Table 1.
In general, with the exception of chloride 10 (Table 1, entry
6), all of the compounds proved to be modest competitive
inhibitors of both enzymes with Ki values in the micromolar
range. Compounds 5 and 6 showed Ki values against Mt-DHQ2
and Hp-DHQ2 in the range 11−15 and 188−282 μM,
respectively. The increased length of the C3 hydroxyl chain
in the latter inhibitors has a more pronounced effect on the
inhibitory potency against the Hp-DHQ2 enzyme, while only
slight differences were observed for the Mt-DHQ2 enzyme
(Table 1, entries 1 and 2). The two hydroxyl derivatives 7 and
8 showed similar Ki values in the range 27−37 μM against both
enzymes (Table 1, entries 3 and 4). Compound 8, which has an
R configuration in the C3 side chain, proved to be slightly more
potent than its epimer. The most pronounced differences were
found with the chlorides 9 and 10, which showed Ki values
against Mt-DHQ2 and Hp-DHQ2 in the range 14.5−0.48 and
40.5−10.5 μM, respectively (Table 1, entries 5 and 6). The
a
Reagents and conditions: (a) CO, B-benzyl-9-BBN, Pd(dppf)Cl2
(cat.), K3PO4 (aq.), THF, 55 °C; (b) TBAF, THF, rt; (c)
NaBH(OAc)3, CeCl3, MeOH, THF, 0 °C; (d) (1) LiOH (aq),
THF, rt; (2) Amberlite IR-120 (H+), rt; (e) NCS, PPh3, THF, rt.
acids 7 and 8 were obtained by 1,2-reduction of α,β-
unsaturated ketone 19 or 20, the latter being obtained by
treatment of 19 with TBAF. A difference in the diastereose-
lectivity of the 1,2-reduction was not observed on using either
TBS-protected α,β-unsaturated ketone 19 or the corresponding
deprotected ketone 20. Reduction of the two α,β-unsaturated
Table 1. Ki (μM) Values for Compounds 5−10 against Hp- and Mt-DHQ2 Enzymes
a
b
Assay conditions: pH 7.0, 25 °C, 50 mM Tris-HCl; Km (1) = 444 μM. Assay conditions: pH 7.0, 25 °C, 50 mM Tris-HOAc; Km (1) = 15 μM.
3497
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
3498
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
Figure 3. Binding interactions of inhibitors 5 (yellow), 6 (pink), 7 (green), and 8 (orange) in the binary complexes Mt-DHQ2/5 (A, PDB entry
4CIV), Mt-DHQ2/6 (B, PDB entry 4CIW), Mt-DHQ2/7 (C, PDB entry 4CIX), and Mt-DHQ2/8 (D, PDB entry 4CIY). Comparison of the binary
complexes Mt-DHQ2/5 (yellow) vs Mt-DHQ2/6 (pink) and Mt-DHQ2/8 (orange) vs Mt-DHQ2/4c (PDB entry 2Y71, cyan) are also shown in
(E) and (F), respectively. A maximum-likelihood weighted 2Fo − Fc map contoured at 1σ is shown up to 1.6 Å around the WAT1 molecule (red).
For better comparison, water molecules in PDB entry 2Y71 are shown in yellow. Hydrogen-bonding interactions and distance differences between
WAT2004 and WAT2001 are shown as red and blue dashes, respectively. Relevant residues are shown and labeled. For (a)−(d), the Asp88* residue
from a symmetry-related neighboring enzyme subunit is shown in cyan.
3499
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
results show that the replacement of the hydroxyl group in
compound 8 by a chloride increases the inhibitory potency
against Mt-DHQ2 and Hp-DHQ2 by more than 50- and 3-fold,
respectively. Chloride 10, which has an R configuration on the
C3 side chain, proved to be the most potent inhibitor of the
series, with Ki values of 0.48 and 10.5 μM against Mt-DHQ2
and Hp-DHQ2, respectively. It is worth highlighting that a
recent comparative binding energy (COMBINE) analysis
carried out with a series of competitive reversible inhibitors
of Hp-DHQ2 and Mt-DHQ2 enzymes allowed us to quantify
the interactions that contribute most significantly to explaining
the key differences between the two enzymes and to quantify
the importance of ligand interactions at the interface pocket
between chains, which is close to the active site.27 The latter
pocket is involved in ligand efficiency and reveals significant
differences between both enzymes (see Figures S7 and S8 of
the Supporting Information). We believe that the significant
differences in the inhibitory potency against both enzymes
shown by the most potent inhibitor here reported, chloride 10,
are due to the differences in the binding interactions with the
residues involved in this pocket for both enzymes.
In an effort to gain further insight into the binding mode of
these inhibitors and to identify the key binding interactions
responsible for the experimentally obtained inhibition data, the
crystal structures of Mt-DHQ2 in complex with compounds 5−
8 were solved. The binding modes of compounds 9 and 10
with Mt-DHQ2 and the dynamic behavior of the key
interactions identified in the binary complexes reported here
were studied by MD simulation studies, the results of which are
discussed below.
contain the conserved Arg108, Glu109, and Arg112 and 126−
130. As a result of the favorable electrostatic interaction
between the Arg19 side chain and the carboxylate group of 5,
the usual strong hydrogen bonding interactions involving the
side chain NH2 group of Asn75 and the main chain NH group
of Ser103 is lost as a consequence of the carboxylate rotation of
∼51° (Figure 3A). This crystal structure also highlights the
unusual position of WAT1 (WAT2001), which is displaced by
∼1.3 Å relative to its corresponding position in the binary
complex Mt-DHQ2/6 (see below) toward the side chain of
Asp88* of a symmetry-related neighboring enzyme subunit
(Figure 3E). As a result of this displacement, the usual
hydrogen bonding interaction with the carbonyl group of Pro11
residue is lost.
Mt-DHQ2/6 Binary Complex. In this crystal structure, Arg19
is pointing toward the active site and its guanidinium group
interacts with the phenolic group of Tyr24 and the C3 side
chain hydroxyl group of ligand 6 through two bridging water
molecules (WAT2010 and WAT2081) (Figure 3B). Three
water molecules join together the side chains of Arg19, Tyr24,
and Asp88* residues. All of these attractive interactions
facilitate the stabilization of the loop in the closed configuration
and allow its description. In one of the possible arrangements of
Asp88*, the side chain forms the usual salt bridge with the side
chain of the conserved Arg112, and in the other this salt bridge
is partially broken because the carboxylate group is located
close to WAT1 (WAT2004). In this arrangement, WAT2004 is
engaged in four hydrogen bonds, namely, with the side chains
of Asp88* and Asn10, the main chain amide of Gly78, and the
carbonyl group of Pro11.
Structural Studies. The crystal structures of Mt-DHQ2
with the inhibitors 5−8 were obtained by soaking apo-Mt-
DHQ2 crystals that were determined at 2.9, 2.2, 2.9, and 2.1 Å,
respectively. All structures were determined by molecular
replacement, using the previously described structure of Mt-
DHQ2 bound to 3-hydroxyiminoquinic acid (PDB entry
1H0S)28 as a search model, and the structures were refined
(Table 2). All binary complexes Mt-DHQ2/5−8 contain a
single Mt-DHQ2 molecule in the crystallographic asymmetric
unit (Figure 3). The structures were refined with good
geometric parameters, and clear electron density is visible for
all amino acids with the exception of certain residues of the
flexible substrate-covering loop, a situation that is not surprising
given that this flexibility is essential for the function of the
enzyme (Figure S1 of the Supporting Information).
Mt-DHQ2/5 Binary Complex. The most remarkable features
of this crystal structure are the interactions of the guanidinium
group of the essential Arg19 with the C1 carboxylate group of 5
(electrostatic) and with the C3 side chain hydroxyl group of 5
(hydrogen bonding) (Figure 3A).29 This arrangement of the
side chain of Arg19 near the carboxylate group of a ligand has
not been seen before in any of the available DHQ2 crystal
structures, and it supports the recently suggested additional role
of the essential arginine as the residue that triggers the release
of the product from the active site. Thus, MD simulation
studies carried out in our group on the binary complex Mt-
DHQ2/2 indicated that the electrostatic interaction of the
guanidinium group of Arg19 pulling on the carboxylate group
in 2 is responsible for breaking the favorable interactions of the
reaction product with the active site residues, specifically with
Asn75, Ile102, and Ser103 residues.13 As a consequence, the
product is expelled from the active site after a large
conformational change involving residues 101−112, which
Mt-DHQ2/7 Binary Complex. The benzyl group of inhibitor
7 seems to block the approach of the catalytic tyrosine to
Arg108, thus preventing the formation of the hydrogen-
bonding interaction required to close the active site upon
substrate binding (Figure 3C). The position of this group
seems to be controlled by a cation−π interaction between the
phenyl moiety and the guanidinium group of Arg18, which is
located in the loop. Furthermore, the hydroxyl group of the
inhibitor side chain establishes a good hydrogen bond with the
carboxylate group of Asp88* of the neighboring subunit (2.7
Å). Notably and in contrast to the other crystal structures
reported here, electron density was not observed for the
structural water molecule. This finding is not surprising
considering that a water molecule in the same position would
sterically clash with the C3 side chain hydroxyl group of the
ligand. As discussed below, the results of MD simulation studies
also suggest that the structural water molecule is not
compatible with the presence of the ligand.
Mt-DHQ2/8 Binary Complex. Although the electron density
is not conclusive, we modeled the side chain of Tyr24 to point
inward with the phenol group interacting with Arg108 through
hydrogen-bonding with a bridging water molecule (WAT2010)
(Figure 3D). As found in several binary complexes containing
ligands with a C3 flexible chain substituted with an aromatic
moiety, the side chain of Arg19 is probably displaced outside
the active site by the benzyl group of 8. In contrast to its
epimer, compound 7 (see above), high electron density for
WAT1 (WAT2004) is observed. Moreover, a comparison of
this crystal structure with that of the Mt-DHQ2/4c binary
complex (PDB entry 2Y7118) indicates that both structures are
virtually identical, with rms differences of 0.155 Å after
superposition of 124 Cα-atom pairs (Figure 3F). In particular,
the relative positions of their aromatic moieties and the oxygen
3500
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
Figure 4. Comparison of the binding mode of the inhibitors 7, 8, 9, and 10 in the binary complexes Mt-DHQ2/7 (A, B), Mt-DHQ2/8 (E), Mt-
DHQ2/9 (C, D), and Mt-DHQ2/10 (F) after minimization and prior simulation (gray) and after 10 ns of dynamic simulation (yellow). For Mt-
DHQ2/7 and Mt-DHQ2/9 binary complexes, MD simulations were carried out with WAT1 (A, C) and without WAT1 (B, D), respectively. Note
how, for the Mt-DHQ2/9 binary complex, WAT1 is expelled from the WAT1 binding pocket during simulation while it remains in place for Mt-
DHQ2/10. Relevant side chain residues are shown and labeled. Relevant hydrogen-bonding (red), halogen-bonding (blue), and CH···Cl interactions
(black) are shown. For Mt-DHQ2/7 binary complex (A), distance differences between WAT1 after minimization and prior simulation and after 10
ns of dynamic simulation are shown as dashed lines (magenta).
3501
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
atom of their C3 side chains with respect to WAT1 are very
similar. Despite the structural similarities between these two
binary complexes, compound 8 proved to be much less potent
than compound 4c. However, we must take into account the
important differences in the aromatic moieties of the two
inhibitors (phenyl vs 5-methylbenzo[b]thiophenyl) and, there-
fore, their different interactions with the interface pocket close
to the active site. Indeed, the results of our recent QSAR
studies suggest that the inhibitory potency of the ligands against
the Mt-DHQ2 enzyme can be enhanced by establishing
favorable van der Waals interactions with Glu92* of the
neighboring subunit.27 The larger size of the benzo[b]-
thiophenyl ring compared with the phenyl one favors these
types of interactions (see Figure S2 of the Supporting
Information). Finally, the R-configuration of the hydroxyl
group in the C3 side chain of 8 not only allows it to establish a
strong hydrogen-bonding interaction with WAT1 but also
enables the appropriate positioning of the phenyl ring for a π-
stacking interaction with the phenolic ring of the essential
Tyr24, and an extended conformation of the flexible side chain
is displayed.
Molecular Dynamics Simulation Studies. In order to
analyze the relevance of the electrostatic interaction between
the guanidinium group of Arg19 and the C1 carboxylate group
in 5 and the hydrogen-bonding interactions with WAT1 in the
reported complexes, the four binary complexes described here
were subjected to 10 ns of dynamic simulation. Moreover, in
order to gain further information concerning the inhibition
results obtained on replacing the hydroxyl group in the C3 side
chains in 7 and 8 by a chloride, the binding modes of
compounds 9 and 10 in the active site of the M. tuberculosis
enzyme were also studied by MD simulations. These
simulations were carried out on a catalytically active trimer30
in aqueous solution using the molecular mechanics force field
AMBER.31 Unsolved residues were modeled using the Web-
based ModLoop server.32 For Mt-DHQ2/5 and Mt-DHQ2/7,8
binary complexes, the calculated loops locate the side chain of
Tyr24 and Arg19 in a disposition in which they point outside
the active site in both cases. Simulations with Mt-DHQ2/9,10
binary complexes were carried out using the enzyme
coordinates found in the crystal structures of Mt-DHQ2 in
complex with their corresponding hydroxyl derivatives 7 and 8,
with manual replacement of ligands 7 and 8 by 9 and 10,
respectively.
The MD simulation studies performed with the Mt-DHQ2/5
binary complex showed that the relative position of ligand 5
and Arg19 in the active site does not change significantly during
the whole simulation (10 ns) (Figure S3 of the Supporting
Information). The guanidinium group of Arg19 and the C1
carboxylate group of 5 interact directly during most of the
simulation, and when they do not, the two residues are linked
by a bridging water molecule. These results highlight that the
interactions of the guanidinium group of Arg19 with the C1
carboxylate group (electrostatic) and the primary hydroxyl
group of 5 (hydrogen bonding) are very strong. More
importantly, these results provide new evidence of the high
mobility of the Arg19 side chain and thus the flexible loop in
which it is located. This mobility not only enables control of
the appropriate orientation of the essential tyrosine side chain
by a cation−π interaction for the enolate generation but also
enables it to interact with the carboxylate group of the ligand.
The latter interaction supports the hypothesis that the essential
Arg19 is also involved in the release of the reaction product
from the active site.13 On the other hand, significant changes
were not observed for the Mt-DHQ/6 binary complex during
the whole simulation (Figure S3 of the Supporting
Information).
The computational studies carried out with Mt-DHQ2/7,8
binary complexes revealed the importance of the stereo-
chemistry of the chiral center of the C3 side chain in 7 and 8
for their interaction with the WAT1 binding pocket (Figure 4).
Thus, for the Mt-DHQ2/7 binary complex after 10 ns of
simulation, WAT1 is displaced by ∼1.5 Å relative to its usual
position by the hydroxyl group of the C3 side chain in 7
(Figure 4A). In fact, when the simulation studies were
performed after manual removal of WAT1, the binding
interactions of ligand 7 in the Mt-DHQ2/7 binary complex
proved to be much more stable (Figure 4B). These findings are
more clearly shown in the MD simulation studies conducted
with ligand 9, in which the hydroxyl group has been replaced by
a chloride. This simulation will be discussed below. In contrast,
significant changes were not observed in the WAT1 binding
pocket during the simulation on the Mt-DHQ2/8 binary
complex, which shows once again the relevance of the
hydrogen-bonding interaction of the ligand with WAT1 in
the binding mode of ligand 8 (Figure 4E).
The binding modes of chlorides 9 and 10 were studied using
the enzyme geometries found in the crystal structures of Mt-
DHQ2 in complex with compounds 7 and 8. For the Mt-
DHQ2/9 binary complex, similar behavior was observed
compared to that in the simulation studies conducted with
the hydroxyl derivative 7 (see above), although for ligand 9 the
observed displacement of WAT1 from the WAT1 binding
pocket by the chloride is clearly more pronounced. The MD
simulation studies conducted with Mt-DHQ2/9 containing
WAT1 showed that WAT1 is rapidly expelled from its usual
position to be occupied by the chloride atom of ligand 9
(Figure 4C). Large movements in the bottom of the interface
pocket close to the active site were observed, and this is mainly
because WAT1 controls the correct positioning of the Asn12
side chain, which is also establishing a strong hydrogen bond
with the side chain of His63* from the neighboring enzyme
subunit. These results highlight the critical structural role of
WAT1 in stabilizing the region of the active site between helix
310 H1 and α3′, which contain residues 10−12 and 77−79,
respectively. In fact, MD simulations conducted with the apo-
form of the Mt-DHQ2 enzyme, i.e., in the absence of ligand,
showed that WAT1 is not exchanged with the bulk solvent and
does not undergo significant changes during the whole
simulation (Figure S4 of the Supporting Information).
Moreover, when the simulations of the Mt-DHQ2/9 binary
complex were performed in the absence of WAT1, the binding
interactions of the ligand proved to be very stable and the
position of the chloride is anchored by two hydrogen-bonding
interactions, one with the main chain amide of Gly78 and the
other with the side chain of the conserved Asn12, and a halogen
bond interaction with the carbonyl oxygen of Pro11 (Figure
4D).33 These results suggest that when the ligand occupies the
position of WAT1, part of this stabilization is lost because the
positions of residues Asn12, Pro11, and Gly78 do not seem to
be as efficiently “frozen” as they are when a water molecule is
involved.
In the case of the Mt-DHQ2/8 binary complex, the binding
mode of inhibitor 10 in the active site of Mt-DHQ2 obtained
by MD simulation studies showed that the substituent of the
C3 side chain (chloride) in 10 interacts with WAT1 by
3502
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
hydrogen-bonding (Figure 4F). However, unlike ligand 8, the
chloride in 10 establishes two extra contacts, namely, a halogen
bond with the carbonyl oxygen of Pro11 and a CH···Cl
interaction with the conserved Leu13. All of these interactions
remained very stable during the whole simulation, which
suggests that these interactions all contribute greatly to its
binding with the enzyme and might explain the more than 50-
fold higher inhibitory potency of 10 in comparison to 8. The
calculated binding free energies for both ligands also support
these differences. The calculated binding free energies of
compounds 9 and 10 in the active site of Mt-DHQ2 for the
MD simulation studies conducted, using the molecular
mechanics Poisson−Boltzmann surface area (MM/PBSA)34
approach in explicit water (GB) as implemented in Amber, also
support the higher affinity of chloride 10 vs 9 as found
experimentally (Table S1 of the Supporting Information). In
addition, to examine temporal changes in the enzyme structure
during the 10 ns of MD simulation, the root-mean-square
deviation (rmsd) with respect to the starting structure was
calculated for all complexes (Figure S5 of the Supporting
Information). The obtained rmsd values are between 1.3 and
2.7 Å and are stabilized after about 0.2 ns. To determine the
most deviated residues during the simulation, the rms
fluctuation (rmsf) values of the Cα atoms of the Mt-DHQ2
enzyme were calculated and plotted as a function of the residue
number for the three chains (Figure S6 of the Supporting
Information). As expected, the residues located in the substrate-
covering loop, in which flexibility is essential for the enzymatic
activity, have the greatest rmsf values.
respectively. The structures of these complexes show that the
hydroxyl group of the C3 side chain of compounds 5−6 and 8
does not occupy the WAT1 binding pocket. For the Mt-
DHQ2/8 binary complex, this hydroxyl group interacts by
hydrogen-bonding with WAT1, while for the Mt-DHQ2/7
binary complex, displacement of the structural water molecule
and direct interaction with the enzyme were observed.
Moreover, the solved crystal structure of the Mt-DHQ2/5
binary complex showed an electrostatic interaction between the
guanidinium group of the essential Arg19 and the carboxylate
group of inhibitor 5 in the Mt-DHQ2/5 binary complex. This
interaction has not been seen before in any DHQ2 crystal
structure. This finding supports the recently proposed addi-
tional role of Arg19 as the residue that triggers the release of
the product from the active site.13 The results of the structural
and MD simulation studies reveal important differences in the
binding mode interactions with WAT1 and the WAT1 binding
pocket, and these differences might explain the higher
inhibitory potency of the analogues with R configuration on
the C3 side chain, i.e., compounds 8 and 10, in comparison to
their epimers, compounds 7 and 9, which also bind to the active
site with an extended conformation of this chain. This effect is
particularly remarkable on replacing the OH group with a
chloride. The results of MD simulation studies suggest that the
OH/Cl groups of the C3 side chain in 7 and 9 respectively
occupy the position of WAT1 in the active site. In contrast, for
their epimers, compounds 8 and 10, the OH/Cl groups interact
by hydrogen-bonding with WAT1 without expelling this
molecule from the WAT1 binding pocket. In view of the
structural and computational studies reported here, the
inhibition potency against the DHQ2 enzyme would be
favored by promoting interactions with WAT1, by avoiding
the displacement of WAT1, and by promoting extra
interactions with the residues located in the WAT1 binding
pocket. In addition, those groups, which would be incorporated
in the C3 side chain of the enolate mimetic with this purpose,
must also allow an extended conformation of the C3 side chain
in order to enable close contact of the aromatic moiety with the
interface pocket between chains, which is highly involved in
ligand efficiency.27 In view of our previous results with
compounds 4,18 the replacement of the phenyl group in 10
by a larger size ring (naphthyl or benzo[b]thiophenyl) would
enhance binding affinity of the ligand to the enzyme. Further
chemical modifications of the reported acids are required to
improve internalization into Mycobacterium tuberculosis cell.
This is currently being carried out in our laboratories through a
propyl ester prodrug approach as with previously reported O-
alkyl derivatives 4.18 These computational studies also
highlighted the key role of WAT1 in the stabilization of the
bottom of the interface pocket close to the active site and
provide guidance for the future design of inhibitors against a
recognized attractive target for the development of new
antibacterial agents.
When considered together, the results described above
suggest that the binding affinity for the DHQ2 enzyme is
enhanced for those ligands that can interact with WAT1,
without causing changes in the WAT1 binding pocket, and
promote extra interactions with the residues located in this
pocket. The replacement of the hydroxyl group in 8 by a
chloride not only enhanced the hydrogen-bonding interaction
with WAT1 but also favored extra contacts with the carbonyl
group of Pro11 and the side chain of Leu13, a situation that
may explain its markedly higher inhibitory potency.
CONCLUSIONS AND FINAL REMARKS
■
Several enol mimetics, compounds 5−10, of the reaction
catalyzed by DHQ2, the third enzyme of the shikimic acid
pathway, have been synthesized and tested as part of a
structure-based design of inhibitors of DHQ2. These
compounds were specifically designed to explore the WAT1
binding pocket, which has been suggested to play an important
role in the stabilization of the natural substrate and the enolate
intermediate 3 during the catalysis process. These compounds
were synthesized by nucleophilic substitution or Suzuki cross-
coupling from previously reported allylic bromide 12 or vinyl
triflate 15. The reported compounds proved to be reversible
competitive inhibitors of Mt-DHQ2 and Hp-DHQ2 with Ki
values in the micromolar range. Chloride 10, which has an R-
configuration in the C3 side chain, proved to be the most
potent inhibitor of the series with a Ki value of 0.48 and 10.5
μM against Mt-DHQ2 and Hp-DHQ2, respectively. The results
show that the replacement of the hydroxyl group (R
configuration) of the C3 side chain in 8 by a chloride increases
the inhibitory potency against Mt-DHQ2 and Hp-DHQ2 by
more than 50- and 3-fold, respectively.
EXPERIMENTAL SECTION
General Procedures. All starting materials and reagents were
commercially available and were used without further purification. H
■
1
NMR spectra (250, 300, 400, and 500 MHz) and 13C NMR spectra
(63, 75, 100, and 125 MHz) were measured in deuterated solvents. J
values are given in hertz. NMR assignments were carried out by a
combination of 1D, COSY, and DEPT-135 experiments. FT-IR
spectra were recorded as NaCl plates or KBr discs. [α]2D0 values are
given in 10−1 deg cm2 g−1. UV spectra and enzyme assays were
performed at 25 °C using 1 cm path quartz cells in conjunction with a
The crystal structures of Mt-DHQ2 in complex with
compounds 5−8 were solved at 2.9, 2.2, 2.9, and 2.1 Å,
3503
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
thermostatically controlled water bath. All procedures involving the
use of ion-exchange resins were carried out at room temperature using
Milli-Q deionized water. Amberlite IR-120 (H+) (cation exchanger)
was washed alternately with water, 10% NaOH, water, 10% HCl, and
finally water before use.
reduced pressure to yield an oil, which was purified by flash
chromatography on silica gel, eluting with (25:75) dichloromethane/
hexanes to give diene 16 (318 mg, 82%) as a colorless oil. [α]D20
−193.7° (c 1.6, CHCl3). 1H NMR (300 MHz, CDCl3) δ: 6.14 (dd, J =
17.6 and 11.0 Hz, 1H, CHCH2), 5.98 (d, J = 1.4 Hz, 1H, H2), 5.37
(dd, J = 17.6 and 0.9 Hz, 1H, CHCHH), 5.17 (dd, J = 11.0 and 0.9
Hz, 1H, CHCHH), 4.56 (dd, J = 5.8 and 3.4 Hz, 1H, H5), 4.37 (d, J
= 3.4 Hz, 1H, H4), 2.44 (d, J = 10.6 Hz, 1H, H6ax), 2.35 (ddd, J =
10.6, 5.8, and 1.9 Hz, 1H, H6eq), 0.93 (s, 9H, C(CH3)3), 0.89 (s, 9H,
C(CH3)3), 0.19 (s, 3H, CH3), 0.16 (s, 6H, 2 × CH3), and 0.15 (s, 3H,
CH3) ppm. 13C NMR (75 MHz, CDCl3) δ: 175.6 (C), 136.8 (C),
134.7 (CH), 133.8 (CH), 116.3 (CH2), 75.9 (CH), 75.1 (C), 66.0
(CH), 37.1 (CH2), 25.8 (C(CH3)3), 25.8 (C(CH3)3), 18.2 (2 ×
C(CH3)3), −2.9 (2 × CH3), −3.9 (CH3), and −4.4 (CH3) ppm. IR
(film) υ: 1801 (CO) cm−1. MS (ESI) m/z = 433 (MNa+). HRMS
calcd for C21H38O4Si2Na (MNa+), 433.2201; found, 433.2178.
Methyl (1R,3S,4S,6R,9R)-7-Acetoxymethyl-3,4-dimethoxy-
3,4-dimethyl-9-trimethylsilyloxy-2,5-dioxabicyclo[4.4.0]dec-7-
en-9-carboxylate (13). A stirred solution of the bromide 1225 (300
mg, 0.64 mmol) in dry DMF (4.3 mL) was treated with sodium
acetate (160 mg, 1.9 mmol). The mixture was stirred for 24 h at room
temperature, and diethyl ether and water were added. The organic
layer was separated, and the aqueous phase was extracted with diethyl
ether (×2). The combined organic extracts were washed with brine.
The organic extract was dried (anhydrous Na2SO4), filtered, and
concentrated under reduced pressure. The residue was purified by
flash chromatography on silica gel, eluting with (40:60) diethyl ether/
hexanes to give acetate 13 (104 mg, 36%) as a colorless oil. Starting
material (164 mg, 55%) was also recovered. Corrected yield: 46%.
[α]2D0 +106.8° (c 1.0, CHCl3). 1H NMR (250 MHz, CDCl3) δ: 5.72 (s,
1H, H8), 4.74 (d, J = 13.8 Hz, 1H, CHHO), 4.66 (d, J = 13.8 Hz, 1H,
CHHO), 4.23 (br d, J = 10.4 Hz, 1H, H6), 4.09−3.99 (m, 1H, H1),
3.74 (s, 3H, OCH3), 3.27 (s, 3H, OCH3), 3.26 (s, 3H, OCH3), 2.21−
2.09 (m, 2H, CH2-10), 2.06 (s, 3H, CH3), 1.31 (s, 3H, CH3), 1.30 (s,
3H, CH3), and 0.12 (s, 9H, Si(CH3)3) ppm. 13C NMR (63 MHz,
CDCl3) δ: 173.8 (C), 170.5 (C), 136.0 (C), 126.0 (CH), 100.5 (C),
100.0 (C), 75.8 (C), 69.4 (CH), 65.6 (CH), 62.6 (CH2), 52.7
(OCH3), 48.1 (OCH3), 48.0 (OCH3), 37.7 (CH2), 21.0 (CH3), 18.0
(CH3), 17.9 (CH3), and 2.0 (3 × CH3) ppm. IR (film) υ: 1744 (CO)
cm−1. MS (ESI) m/z = 469 (MNa+). HRMS calcd for C20H34O9SiNa
(MNa+), 469.1864; found, 469.1870.
(1R,4S,5R)-1,4-Di(tert-butyldimethylsilyloxy)-3-(2-hydroxy)-
ethylcyclohex-2-en-1,5-carbolactone (17). A stirred solution of
diene 16 (50 mg, 0.12 mmol) in dry THF (3 mL), under argon at 0
°C, was treated with a BH3−THF solution (1.22 mL, 1.22 mmol). The
mixture was stirred for 1 h. An aqueous solution of NaBO3 (3 mL, 1.22
mmol, 0.4 M) was added, and the resulting solution was stirred for 2 h
at this temperature. The mixture was warmed to room temperature,
and ethyl acetate and water were added. The organic layer was
separated, and the aqueous layer was extracted with ethyl acetate (×2).
The combined organic extracts were dried (anhydrous Na2SO4),
filtered, and concentrated under reduced pressure. The residue was
purified by flash chromatography on silica gel, eluting with (50:50)
diethyl ether/hexanes to give alcohol 17 (18 mg, 35%) as a colorless
oil. [α]2D0 +132.8° (c 1.8, CHCl3). H NMR (300 MHz, CDCl3) δ:
1
Methyl (1R,4S,5R)-3-Acetoxymethyl-1,4,5-trihydroxycyclo-
hex-2-en-1-carboxylate (14). A solution of the acetal 13 (74 mg,
0.16 mmol) in a 20:1 (v/v) mixture of TFA/H2O (0.3 mL) was stirred
at 0 °C for 15 min. Solvents were removed under reduced pressure
and the crude residue was purified by flash chromatography on silica
gel, eluting with (40:60) ethyl acetate/hexanes to give triol 14 (28 mg,
5.84 (br d, J = 1.1 Hz, 1H, H2), 4.48 (m, 1H, H5), 4.06 (d, J = 3.1 Hz,
1H, H4), 3.72 (td, J = 6.5 and 1.7 Hz, 2H, CH2O), 2.33−2.31 (m, 4H,
2 × CH2), 1.57 (s, 1H, OH), 0.92 (s, 9H, C(CH3)3), 0.91 (s, 9H,
C(CH3)3), 0.18 (s, 3H, CH3), 0.16 (s, 3H, CH3), 0.15 (s, 3H, CH3),
and 0.14 (s, 3H, CH3) ppm. 13C NMR (75 MHz, CDCl3) δ: 176.1
(C), 136.1 (C), 133.1 (CH), 76.0 (CH), 74.8 (C), 68.2 (CH), 61.0
(CH2), 37.2 (CH2), 35.2 (CH2), 25.8 (C(CH3)3), 25.7 (C(CH3)3),
18.1 (C(CH3)3), 18.0 (C(CH3)3), −2.9 (CH3), −2.9 (CH3), −4.5
(CH3), and −4.5 (CH3) ppm. IR (film) υ: 3421 (OH) and 1794 (CO)
cm−1. MS (ESI) m/z = 451 (MNa+). HRMS calcd for C21H40O5Si2Na
(MNa+), 451.2306; found, 451.2305.
68%) as a colorless oil. [α]2D0 +49.3° (c 1.9, CH3OH). H NMR (250
1
MHz, CD3OD) δ: 5.73 (br s, 1H, H2), 4.78 (d, J = 13.7 Hz, 1H,
CHHO), 4.61 (d, J = 13.7 Hz, 1H, CHHO), 4.00−3.85 (m, 2H, H4 +
H5), 3.74 (s, 3H, OCH3), 2.08 (s, 3H, CH3), and 2.02 (m, 2H, CH2-
6) ppm. 13C NMR (63 MHz, CD3OD) δ: 176.1 (C), 172.4 (C), 140.5
(C), 126.1 (CH), 74.0 (C), 73.9 (CH), 70.6 (CH), 64.9 (OCH2), 53.1
(OCH3), 40.5 (CH2), and 20.8 (CH3) ppm. IR (film) υ: 3419 (OH)
and 1732 (CO) cm−1. MS (ESI) m/z = 283 (MNa+). HRMS calcd for
C11H16O7Na (MNa+), 283.0788; found, 283.0789.
(1R,4S,5R)-1,4-Dihydroxy-3-(2-hydroxy)ethylcyclohex-2-en-
1,5-carbolactone (18). Tetrabutylammonium fluoride (0.5 mL, 0.47
mmol, ∼1.0 M in THF) was added to a stirred solution of the silyl
ether 17 (82 mg, 0.19 mmol) in dry THF (2.7 mL) under argon at 0
°C. The mixture was stirred for 2 h, and water and ethyl acetate were
added. The aqueous layer was acidified with dilute HCl (10%), and the
organic layer was separated. The aqueous phase was extracted with
ethyl acetate (×2). The combined organic extracts were dried
(anhydrous Na2SO4), filtered, and concentrated under reduced
pressure. The residue was purified by flash chromatography on silica
gel, eluting with (80:20) ethyl acetate−hexanes to yield triol 18 (24
(1R,4R,5R)-1,4,5-Trihydroxy-3-hydroxymethylcyclohex-2-en-
1-carboxylic Acid (5). A solution of the ester 14 (19 mg, 0.07 mmol)
in THF (0.6 mL) was treated at room temperature with an aqueous
solution of LiOH (0.07 mL, 0.17 mmol, 2.5 M). The mixture was
stirred for 1 h. The THF was removed under reduced pressure, and
water was added. The aqueous layer was washed with ethyl acetate
(×3) and then treated with Amberlite IR-120 (H+) until pH 6 was
obtained. The resin was filtered and washed with Milli-Q water. The
filtrate and the washings were lyophilized to give acid 5 (16 mg, 99%)
mg, 63%) as a colorless oil. [α]2D0 −232° (c 1.5, MeOH). H NMR
1
(300 MHz, CD3OD) δ: 5.83 (br s, 1H, H2), 4.59 (ddd, J = 5.2, 3.2 and
0.8 Hz, 1H, H5), 4.03 (d, J = 3.2 Hz, 1H, H4), 3.68 (dd, J = 7.0 and
2.2 Hz, 1H, OCHH), 3.66 (dd, J = 7.0 and 2.8 Hz, 1H, OCHH), and
2.44−2.23 (m, 4H, 2 × CH2) ppm. 13C NMR (63 MHz, CD3OD) δ:
179.1 (C), 138.7 (C), 132.3 (CH), 77.9 (CH), 74.0 (C), 68.0 (CH),
61.3 (CH2), 37.4 (CH2), and 36.4 (CH2) ppm. IR (film) υ: 3366
(OH) and 1778 (CO) cm−1. MS (ESI) m/z = 223 (MNa+). HRMS
calcd for C9H12O5Na (MNa+), 223.0577; found, 223.0571.
as a yellow oil. [α]2D0 −80.7° (c 1.2, CH3OH). H NMR (250 MHz,
1
CD3OD) δ: 5.68 (s, 1H, H2), 4.18 (s, 2H, OCH2), 3.99 (d, J = 6.9 Hz,
1H, H4), 3.89 (m, 1H, H5), and 2.07 (m, 2H, CH2-6) ppm. 13C NMR
(63 MHz, CD3OD) δ: 179.3 (C), 143.7 (C), 125.2 (CH), 74.1 (C),
73.3 (CH), 71.1 (CH), 63.3 (OCH2), and 39.8 (CH2) ppm. IR (film)
υ: 3434 (OH) and 1730 (CO) cm−1. MS (ESI) m/z = 203 (M − H).
HRMS calcd for C8H11O6 (M − H), 203.0561; found, 203.0568.
(1R,4S,5R)-1,4-Di(tert-butyldimethylsilyloxy)-3-vinylcyclo-
hex-2-en-1,5-carbolactone (16). A Schlenk tube was charged with
triflate 1526 (500 mg, 0.94 mmol), Pd(PPh3)4 (115 mg, 0.1 mmol),
K2CO3 (2.54 mL, 2.79 mmol, 1.1 M), and vinylboronic acid pinacol
ester (24 μL, 1.4 mmol) in dry dioxane (11.6 mL). The resultant
suspension was deoxygenated and heated at 100 °C for 2 h under
argon. The mixture was allowed to cool to room temperature. The
suspension was filtered over Celite, and the residue was washed with
diethyl ether. The filtrate and washings were concentrated under
(1R,4R,5R)-1,4,5-Trihydroxy-3-(2-hydroxy)ethylcyclohex-2-
ene-1-carboxylic Acid (6). A solution of the lactone 18 (15 mg, 0.08
mmol) in THF (0.8 mL) and aqueous lithium hydroxide (0.4 mL, 0.2
mmol, 0.5 M) was stirred at room temperature for 1 h. Water was
added, and the THF was removed under reduced pressure. The
resulting aqueous solution was washed with ethyl acetate (×2), and the
aqueous extract was treated with Amberlite IR-120 (H+) until pH 6
was obtained. The resin was filtered off and washed with Milli-Q water.
The filtrate and the washings were lyophilized to give acid 6 (12 mg,
3504
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
75%) as a colorless oil. [α]2D0 −63° (c 1.2, H2O). 1H NMR (300 MHz,
D2O) δ: 5.60 (br s, 1H, H2), 4.05 (d, J = 8.0 Hz, 1H, H4), 3.92 (m,
1H, H5), 3.75 (t, J = 6.5 Hz, 2H, OCH2), 2.55 (m, 1H, CHH), 2.34
(m, 1H, CHH), and 2.18−2.07 (m, 2H, CH2) ppm. 13C NMR (63
MHz, D2O) δ: 178.8 (C), 142.2 (C), 124.9 (CH), 73.9 (C), 73.6
(CH), 69.9 (CH), 60.2 (CH2), 39.4 (CH2), and 35.1 (CH2) ppm. IR
(film) υ: 3339 (OH) and 1708 (CO) cm−1. MS (ESI) m/z = 217 (M
− H). HRMS calcd for C9H13O6 (M − H), 217.0708; found,
217.0714.
135.7 (C), 130.6 (2 × CH), 129.5 (2 × CH), 127.8 (CH), 77.4 (CH),
74.4 (C), 64.2 (CH), 45.6 (CH2), and 36.7 (CH2) ppm. IR (KBr) υ:
3457 (OH), 3420 (OH), 1766 (CO) and 1680 (CO) cm−1. MS (ESI)
m/z: 297 (MNa+). HRMS calcd for C15H14O5Na (MNa+), 297.0733;
found, 297.0736.
Reduction of α,β-Unsaturated 19. A stirred solution of the
ketone 19 (100 mg, 0.20 mmol) in dry THF (2 mL), under argon at 0
°C, was treated with cerium(III) chloride heptahydrate (127 mg, 0.60
mmol), sodium borohydride (3.2 mg, 0.08 mmol), and methanol (5
drops). The mixture was stirred for 1 h, and water and diethyl ether
were added. The organic layer was separated, and the aqueous layer
was extracted with diethyl ether (×3). The combined organic extracts
were dried (anhydrous Na2SO4), filtered, and concentrated under
reduced pressure. The crude product was purified by flash
chromatography on silica gel, eluting with (5:5:90) diethyl ether/
dichloromethane/hexanes to give alcohol 21 (28 mg, 28%) and
alcohol 22 (51 mg, 51%).
Carbonylation of Triflate 15. A stirred solution of triflate 1522
(813 mg, 1.52 mmol) and Pd(dppf)Cl2 (86 mg, 0.11 mmol) in dry
THF (150 mL) saturated with a continuous flow of carbon monoxide
was treated with B-benzyl-9-BBN (8.25 mL, 4.13 mmol, ∼0.5 M in
THF) and an aqueous solution of K3PO4 (4 mL, 4 mmol, 1 M). The
resulting mixture was heated at 55 °C for 4 h. The mixture was allowed
to cool to room temperature. The organic solvent was removed under
reduced pressure, and the mixture was diluted with diethyl ether and
water. The organic layer was separated, and the organic phase was
extracted with diethyl ether (×2). The combined organic extracts were
dried (anhydrous Na2SO4), filtered, and concentrated under reduced
pressure. The residue was purified by flash chromatography, eluting
with (5:95) diethyl ether/hexanes to give ketone 19 (385 mg, 51%) as
a white solid. The corresponding cross-coupling product was also
obtained in 13% yield (93 mg).
Data for (1R,4R,5R)-1,4-di(tert-butyldimethylsilyloxy)-3-[(1S)-1-
hydroxy-2-phenyl]ethylcyclohex-2-en-1,5-carbolactone (21). White
1
solid. [α]2D0 − 12.3° (c 1.0, CHCl3). H NMR (300 MHz, CDCl3)
δ: 7.34−7.24 (m, 3H, 3 × ArH), 7.14 (m, 1H, ArH), 6.03 (s, 1H, H2),
4.47 (q, J = 3.2 Hz, 1H, H5), 4.43 (m, 1H, CHOH), 4.09 (d, J = 3.3
Hz, 1H, H4), 2.94 (dd, J = 13.6 and 5.0 Hz, 1H, CHHPh), 2.79 (dd, J
= 13.6 and 7.9 Hz, 1H, CHHPh), 2.33 (d, J = 2.6 Hz, 2H, CH2-6),
0.92 (s, 9H, C(CH3)3), 0.91 (s, 9H, C(CH3)3), 0.18 (s, 3H, CH3),
Data for 19. White solid. Mp: 80−81 °C. [α]2D0 −15.8° (c 1.0,
CHCl3). 1H NMR (400 MHz, CDCl3) δ: 7.30 (m, 2H, 2 × ArH), 7.24
(m, 1H, ArH), 7.17 (d, J = 7.3 Hz, 2H, 2 × ArH), 6.88 (d, J = 0.9 Hz,
1H, H2), 4.68 (d, J = 3.4 Hz, 1H, H4), 4.55 (dd, J = 5.7 and 3.5 Hz,
1H, H5), 3.92 (br d, J = 15.7 Hz, 1H, CHHPh), 3.89 (br d, J = 15.7
Hz, 1H, CHHPh), 2.48 (d, J = 10.9 Hz, 1H, H6ax), 2.38 (m, 1H,
H6eq), 0.94 (s, 9H, C(CH3)3), 0.83 (s, 9H, C(CH3)3), 0.17 (s, 3H,
CH3), 0.17 (s, 3H, CH3), 0.12 (s, 3H, CH3), and 0.06 (s, 3H, CH3)
ppm. 13C NMR (75 MHz, CDCl3) δ: 196.2 (C), 174.4 (C), 144.9
(CH), 138.5 (C), 134.0 (C), 129.4 (2 × CH), 128.9 (2 × CH), 127.1
(CH), 75.7 (CH), 75.2 (C), 63.5 (CH), 45.0 (CH2), 36.4 (CH2), 25.8
(C(CH3)3), 25.7 (C(CH3)3), 18.1 (C(CH3)3), 18.1 (C(CH3)3), −3.0
(2 × CH3), −4.8 (CH3), and −5.1 (CH3) ppm. IR (KBr) υ: 1801
(CO) and 1684 (CO) cm−1. MS (ESI) m/z = 525 (MNa+). HRMS
calcd for C27H42O5Si2Na (MNa+), 525.2463; found, 525.2449.
Data for cross-coupling product [(1R,4R,5R)-1,4-di(tert-butyldime-
thylsilyloxy)-3-benzylcyclohex-2-en-1,5-carbolactone]. Colorless oil.
0.16 (s, 3H, CH3), 0.14 (s, 3H, CH3), and 0.12 (s, 3H, CH3) ppm. 13
C
NMR (63 MHz, CDCl3) δ: 175.7 (C), 141.0 (C), 136.9 (C), 131.6
(CH), 129.7 (2 × CH), 128.8 (2 × CH), 127.0 (CH), 75.9 (CH), 74.9
(C), 70.9 (CH), 66.6 (CH), 43.0 (CH2), 37.1 (CH2), 25.8 (C(CH3)3),
25.8 (C(CH3)3), 18.2 (C(CH3)3), 18.0 (C(CH3)3), −2.9 (2 × CH3),
and −4.3 (2 × CH3) ppm. IR (KBr) υ: 3421 (OH) and 1792 (CO)
cm−1. MS (ESI) m/z = 597 (MNa+). HRMS calcd for C27H44O5Si2Na
(MNa+), 527.2619; found, 527.2624.
Data for (1R,4R,5R)-1,4-di(tert-butyldimethylsilyloxy)-3-[(1R)-1-
hydroxy-2-phenyl]ethylcyclohex-2-en-1,5-carbolactone (22). Colorless
oil. [α]2D0 −37.2° (c 1.0, CHCl3). H NMR (300 MHz, CDCl3) δ:
1
7.34−7.23 (m, 3H, 3 × ArH), 7.18 (m, 2H, 2 × ArH), 5.95 (s, 1H,
H2), 4.51 (q, J = 3.1 Hz, 1H, H5), 4.42 (d, J = 3.2 Hz, 1H, H4), 4.33
(m, 1H, CHOH), 3.02 (dd, J = 13.7 and 4.5 Hz, 1H, CHHPh), 2.83
(dd, J = 13.7 and 8.7 Hz, 1H, CHHPh), 2.35 (d, J = 2.6 Hz, 2H, CH2-
6), 0.93 (s, 9H, C(CH3)3), 0.91 (s, 9H, C(CH3)3), 0.19 (s, 3H, CH3),
0.18 (s, 3H, CH3), 0.15 (s, 3H, CH3), and 0.12 (s, 3H, CH3) ppm. 13
C
[α]2D0 −16.2° (c 1.0, CHCl3). H NMR (300 MHz, CDCl3) δ: 7.29
1
NMR (63 MHz, CDCl3) δ: 175.5 (C), 139.5 (C), 137.7 (C), 133.2
(CH), 129.5 (2 × CH), 128.8 (2 × CH), 126.8 (CH), 75.9 (CH), 74.8
(C), 72.3 (CH), 66.6 (CH), 42.1 (CH2), 37.1 (CH2), 25.9 (C(CH3)3),
25.7 (C(CH3)3), 18.2 (C(CH3)3), 18.0 (C(CH3)3), −2.9 (2 × CH3),
−4.3 (CH3), and −4.4 (CH3) ppm. IR (film) υ: 3417 (OH) and 1791
(CO) cm−1. MS (ESI) m/z = 527 (MNa+). HRMS calcd for
C27H44O5Si2Na (MNa+), 527.2619; found, 527.2625.
Reduction of α,β-Unsaturated Ketone 20. A stirred solution of
the ketone 20 (43 mg, 0.16 mmol) in dry THF (4 mL), under argon at
0 °C, was treated with cerium(III) chloride heptahydrate (59 mg, 0.16
mmol), sodium triacetoxyborohydride (110 mg, 0.52 mmol), and
methanol (5 drops). The mixture was stirred for 1 h, and water and
ethyl acetate were added. The organic layer was separated, and the
aqueous layer was extracted with ethyl acetate (×3). The combined
organic extracts were dried (anhydrous Na2SO4), filtered, and
concentrated under reduced pressure. The crude product was purified
by HPLC on a Luna Phenomenex C18 column, eluting with
acetonitrile/water (48:52) at 7 mL·min−1 to give triols 23 (10 mg,
26%) and 24 (22 mg, 51%), both as white solids.
(m, 2H, 2 × ArH), 7.21 (m, 1H, ArH), 7.09 (d, J = 7.1 Hz, 2H, 2 ×
ArH), 5.54 (s, 1H, H2), 4.48 (m, 1H, H5), 4.02 (d, J = 3.0 Hz, 1H,
H4), 3.46 (d, J = 16.1 Hz, 1H, CHHPh), 3.28 (d, J = 16.1 Hz, 1H,
CHHPh), 2.34 (m, 2H, CH2-6), 0.93 (s, 9H, C(CH3)3), 0.87 (s, 9H,
C(CH3)3), 0.10 (s, 6H, 2 × CH3), 0.09 (s, 3H, CH3), and 0.06 (s, 3H,
CH3) ppm. 13C NMR (100 MHz, CDCl3) δ: 176.0 (C), 138.8 (C),
137.3 (C), 132.7 (CH), 129.2 (2 × CH), 128.5 (2 × CH), 126.5
(CH), 75.9 (CH), 74.8 (C), 67.5 (CH), 38.2 (CH2), 37.4 (CH2), 25.7
(C(CH3)3), 25.6 (C(CH3)3), 17.9 (C(CH3)3), 17.9 (C(CH3)3), −3.1
(2 × CH3), −4.5 (CH3), and −4.7 (CH3) ppm. IR (film) υ: 1805
(CO) cm−1. MS (ESI) m/z = 497 (MNa+). HRMS calcd for
C26H42O4Si2Na (MNa+), 497.2514; found, 497.2523.
(1R,4R,5R)-1,4-Dihydroxy-3-(2-benzylacetyl)cyclohex-2-en-
1,5-carbolactone (20). A stirred solution of the silyl ether 19 (385
mg, 0.77 mmol) in dry THF (7.7 mL), under argon at 0 °C, was
treated with tetrabutylammonium fluoride (1.92 mL, 1.92 mmol, ∼1.0
M in THF). The mixture was stirred for 30 min. Saturated NH4Cl was
added, and the organic layer was extracted with diethyl ether (×3).
The combined organic extracts were dried (anhydrous Na2SO4),
filtered, and concentrated under reduced pressure. The crude product
was purified by flash chromatography on silica gel, eluting with (1:1:1)
diethyl ether/acetone/hexanes to give diol 20 (60 mg, 29%) as a white
solid. [α]2D0 −15.0° (c 1.0, CH3OH). Mp: 172−173 °C. 1H NMR (300
MHz, CD3OD) δ: 7.29 (m, 2H, 2 × ArH), 7.23 (m, 1H, ArH), 7.18
(m, 2H, 2 × ArH), 7.13 (br s, 1H, H2), 4.69 (m, 1H, H5), 4.59 (d, J =
3.4 Hz, 1H, H4), 4.07 (d, J = 16.2 Hz, 1H, CHHPh), 4.00 (d, J = 16.2
Hz, 1H, CHHPh), and 2.41 (m, 2H, CH2-6) ppm. 13C NMR (100
MHz, CH3OD) δ: 198.9 (C), 177.1 (C), 145.5 (CH), 139.6 (C),
Data for (1R,4R,5R)-1,4-dihydroxy-3-[(1S)-1-hydroxy-2-
phenylacetyl]cyclohex-2-en-1,5-carbolactone (23). White solid. [α]D20
1
−18.1° (c 1.0, CH3OH). Mp: 170−171 °C. H NMR (250 MHz,
CD3OD) δ: 7.28−7.14 (m, 5H, 5 × ArH), 5.90 (s, 1H, H2), 4.61 (q, J
= 3.2 1H, H5), 4.36 (m, 2H, H4+CHOH), 2.97 (dd, J = 13.7 and 5.2
Hz, 1H, CHHPh), 2.84 (d, J = 13.7 and 8.3 Hz, 1H, CHHPh), and
2.32 (m, 2H, CH2-6) ppm. 13C NMR (63 MHz, CH3OD) δ: 178.6
(C), 142.2 (C), 139.9 (C), 132.0 (CH), 130.5 (2 × CH), 129.2 (2 ×
CH), 127.2 (CH), 77.8 (CH), 74.0 (C), 73.9 (CH), 66.6 (CH), 43.0
3505
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
(CH2), and 37.5 (CH2) ppm. IR (KBr) υ: 3402 (OH), 3418 (OH),
and 1765 (CO) cm−1. MS (ESI) m/z = 299 (MNa+). HRMS calcd for
C15H16O5Na (MNa+), 299.0890; found, 299.0891.
9H,C(CH3)3), 0.91 (s, 9H,C(CH3)3), 0.18 (s, 3H, CH3), 0.16 (s, 6H,
2 × CH3), and 0.09 (s, 3H, CH3) ppm. 13C NMR (63 MHz, CDCl3)
δ: 175.1 (C), 138.5 (C), 137.6 (C), 133.4 (CH), 129.1 (2 × CH),
128.6 (2 × CH), 127.0 (CH), 75.7 (CH), 75.1 (C), 66.1 (CH), 57.9
(CH), 40.8 (CH2), 37.4 (CH2), 25.8 (C(CH3)3), 25.7 (C(CH3)3),
18.2 (C(CH3)3), 18.0 (C(CH3)3), −3.0 (2 × CH3), −4.5 (CH3), and
−4.6 (CH3) ppm. IR (film): 1805 (CO) cm−1. MS (ESI) m/z (%) 545
(MNa+). HRMS calcd for C27H43O4ClNa (MNa+), 545.2281; found,
545.2250.
(1R,4R,5R)-1,4-Dihydroxy-3-[(1R)-1-chloro-2-phenyl]-
ethylcyclohex-2-en-1,5-carbolactone (26). A stirred solution of
the silyl ether 25 (32 mg, 0.06 mmol) in dry THF (1.0 mL), under
argon at 0 °C, was treated with tetrabutylammonium fluoride (0.15
mL, 0.05 mmol, ∼1.0 M in THF). The mixture was stirred for 15 min.
Saturated NH4Cl was added, and the organic layer was extracted with
diethyl ether (×3). The combined organic extracts were dried
(anhydrous Na2SO4), filtered, and concentrated under reduced
pressure. The crude product was purified by flash chromatography
on silica gel, eluting with (1:1:2) diethyl ether/acetone/hexanes to
give diol 26 (14 mg, 79%) as a white foam. [α]2D0 −6.9 (c 1.0, CHCl3).
1H NMR (300 MHz, CDCl3) δ: 7.36−7.28 (m, 4H, 4 × ArH), 7.18
(m, 1H, ArH), 6.08 (s, 1H, H2), 4.75 (m, 1H, CHCl), 4.70 (m, 1H,
H5), 4.57 (d, J = 3.3 Hz, 1H, H4), 3.29 (dd, J = 14.0 and 6.4 Hz, 1H,
CHHPh), 3.20 (dd, J = 14.0 and 8.3 Hz, 1H, CHHPh), 2.49 (d, J =
11.2 Hz, 1H, H6ax), and 2.38 (ddd, J = 11.2, 5.8, and 1.8 Hz, 1H, H6eq)
ppm. 13C NMR (63 MHz, CD3OD) δ: 178.1 (C), 141.0 (C), 139.2
(C), 133.4 (CH), 130.3 (2 × CH), 129.4 (2 × CH), 127.8 (CH), 77.7
(CH), 74.2 (C), 66.3 (CH), 60.3 (CH), 42.3 (CH2), and 37.5 (CH2)
ppm. IR (film): 3422 (OH) and 1782 (CO) cm−1. MS (ESI) m/z (%)
317 (MNa+). HRMS calcd for C15H15O4ClNa (MNa+), 317.0551;
found, 317.0575.
Data for (1R,4R,5R)-1,4-dihydroxy-3-[(1R)-1-hydroxy-2-
phenylacetyl]cyclohex-2-en-1,5-carbolactone (24). White solid. [α]D20
−7.8° (c 1.0, CH3OH). Mp: 138−139 °C. 1H NMR (250 MHz,
CD3OD) δ: 7.27−7.15 (m, 5H, 5 × ArH), 5.96 (s, 1H, H2), 4.57 (m,
1H, H5), 4.50 (m, 1H, CHOH), 4.02 (d, J = 2.7 Hz, 1H, H4), 2.97
(dd, J = 13.5 and 5.2 Hz, 1H, CHHPh), 2.79 (dd, J = 13.5 and 7.3 Hz,
1H, CHHPh), and 2.28 (m, 2H, CH2-6) ppm. 13C NMR (63 MHz,
CH3OD) δ: 178.8 (C), 143.7 (C), 139.4 (C), 130.9 (CH), 130.7 (2 ×
CH), 129.2 (2 × CH), 127.2 (CH), 78.0 (CH), 74.1 (C), 72.1 (CH),
66.4 (CH), 43.8 (CH2), and 37.3 (CH2) ppm. IR (KBr) υ: 3416 (OH)
and 1781 (CO) cm−1. MS (ESI) m/z = 299 (MNa+). HRMS calcd for
C15H16O5Na (MNa+), 299.0890; found, 299.0882.
(1R,4R,5R)-1,4,5-Trihydroxy-3-[(1S)-1-hydroxy-2-
phenylethyl]cyclohex-2-en-1-carboxylic Acid (7). A solution of
carbolactone 23 (10.5 mg, 0.04 mmol) in THF (0.4 mL) and aqueous
lithium hydroxide (0.1 mL, 0.10 mmol, 1 M) was stirred at room
temperature for 30 min. Water was added, and the THF was removed
under reduced pressure. The resulting aqueous solution was washed
with diethyl ether (×3), and the aqueous extract was treated with
Amberlite IR-120 (H+) until pH 6 was obtained. The resin was filtered
off and washed with Milli-Q water. The filtrate and the washings were
lyophilized to give acid 7 (10 mg, 90%) as a white solid. [α]2D0 −6.0° (c
1
1.0, MeOH). Mp: 132−133 °C. H NMR (250 MHz, D2O) δ: 7.43−
7.29 (m, 5H, 5 × ArH), 5.78 (s, 1H, H2), 4.59 (dd, J = 9.2 and 4.5 Hz,
1H, CHOH), 4.27 (d, J = 7.5 Hz, 1H, H4), 3.97 (q, J = 7.5 Hz, 1H,
H5), 3.12 (dd, J = 13.9 and 4.5 Hz, 1H, CHHPh), 2.90 (dd, J = 13.9
and 9.2 Hz 1H, CHHPh), and 2.11 (d, J = 7.5 Hz, 2H, CH2-6) ppm.
13C NMR (63 MHz, D2O) δ: 179.4 (C), 144.3 (C), 139.2 (C), 129.8
(2 × CH), 129.0 (2 × CH), 126.9 (CH), 125.7 (CH), 73.7 (C), 72.5
(2 × CH), 69.9 (CH), 41.7 (CH2), and 38.6 (CH2) ppm. IR (KBr) υ:
3367 (OH) and 1721 (CO) cm−1. MS (ESI) m/z = 293 (M − H).
HRMS calcd for C15H17O6 (M − H), 293.1031; found, 293.1029.
(1R,4R,5R)-1,4,5-Trihydroxy-3-[(1R)-1-hydroxy-2-
phenylethyl]cyclohex-2-en-1-carboxylic Acid (8). A solution of
carbolactone 24 (7 mg, 0.025 mmol) in THF (0.4 mL) and aqueous
lithium hydroxide (0.09 mL, 0.09 mmol, 1 M) was stirred at room
temperature for 30 min. Water was added, and the THF was removed
under reduced pressure. The resulting aqueous solution was washed
with diethyl ether (×3), and the aqueous extract was treated with
Amberlite IR-120 (H+) until pH 6 was obtained. The resin was filtered
off and washed with Milli-Q water. The filtrate and the washings were
lyophilized to give acid 8 (5 mg, 67%) as a white solid. [α]2D0 −1.1° (c
1.0, MeOH). Mp: 95−96 °C. NMR (300 MHz, D2O) δ: 7.39 (m, 2H,
2 × ArH), 7.32 (m, 23H, 3 × ArH), 5.70 (s, 1H, H2), 4.08 (d, J = 7.8
Hz, 1H, H4), 3.95 (m, 1H, H5), 3.61 (m, 1H, CHOH), 3.16 (dd, J =
14.0 and 4.0 Hz, 1H, CHHPh), 2.74 (dd, J = 14.0 and 8.4 Hz 1H,
CHHPh), and 2.14 (m, 2H, CH2-6) ppm. 13C NMR (63 MHz, D2O)
δ: 178.5 (C), 146.0 (C), 138.6 (C), 130.0 (2 × CH), 128.9 (2 × CH),
127.0 (CH), 122.4 (CH), 73.6 (C), 72.6 (CH), 70.8 (CH), 69.9
(CH), 41.7 (CH2), and 38.6 (CH2) ppm. IR (KBr) υ: 3430 (OH) and
1721 (CO) cm−1. MS (ESI) m/z = 293 (M − H). HRMS calcd for
C15H17O6 (M − H), 293.1031; found, 293.1029.
(1R,4R,5R)-1,4,5-Trihydroxy-3-[(1R)-1-chloro-2-phenyl]-
ethylcyclohex-2-en-1-carboxylic Acid (10). A solution of
carbolactone 26 (14 mg, 0.05 mmol) in THF (3.0 mL) at 0 °C was
treated with aqueous lithium hydroxide (0.06 mL, 0.06 mmol, 1 M),
and the mixture was stirred for 2 h. Water was added, and the THF
was removed under reduced pressure. The resulting aqueous solution
was washed with diethyl ether (×3), and the aqueous extract was
treated with Amberlite IR-120 (H+) until pH 6 was obtained. The
resin was filtered off and washed with Milli-Q water. The filtrate and
the washings were lyophilized to give acid 10 (2.9 mg, 23%) as a white
solid. [α]2D0 −1.7 (c 1.0, MeOH). H NMR (500 MHz, D2O) δ: 7.40
1
(m, 2H, 2 × ArH), 7.34 (m, 3H, 3 × ArH), 5.79 (s, 1H, H2), 4.59 (dd,
J = 9.1 and 4.4 Hz, 1H, CHCl), 4.27 (d, J = 7.4 Hz,1H, H4), 3.97 (m,
1H, H5), 3.13 (dd, J = 14.0 and 4.4 Hz, 1H, CHHPh), 2.90 (dd, J =
14.0 and 9.1 Hz, 1H, CHHPh), and 2.11 (m, 2H, CH2-6) ppm. 13C
NMR (63 MHz, D2O) δ: 162.0 (C), 145.1 (C), 139.3 (C), 130.0 (2 ×
CH), 129.2 (2 × CH), 127.2 (CH), 125.1 (CH), 73.6 (C), 72.9 (CH),
72.8 (CH), 69.9 (CH), 41.9 (CH2), and 38.8 (CH2) ppm. IR (film):
3399 (OH) and 1685 (CO) cm−1. MS (ESI) m/z (%) 311 (M − H).
HRMS calcd for C15H17O5Cl (M − H), 311.1085; found, 311.1087.
(1R,4R,5R)-1,4-Di(tert-butyldimethylsilyloxy)-3-[(1S)-1-
chloro-2-phenyl]ethylcyclohex-2-en-1,5-carbolactone (27). A
stirred solution of the alcohol 22 (143 mg, 0.29 mmol) in dry THF
(10 mL), at room temperature and under argon, was treated with N-
chlorosuccinimide (77 mg, 0.58 mmol) and triphenylphosphine (114
mg, 0.44 mmol). The reaction mixture was stirred for 1 h, and the
solvent was removed under reduced pressure. The crude product was
purified by flash chromatography on silica gel, eluting with diethyl
ether/hexanes (5:95) to give chloride 27 (66.5 mg, 44%) as a white
(1R,4R,5R)-1,4-Di(tert-butyldimethylsilyloxy)-3-[(1R)-1-
chloro-2-phenyl]ethylcyclohex-2-en-1,5-carbolactone (25). A
stirred solution of the alcohol 21 (112 mg, 0.22 mmol) in dry THF
(10 mL), at room temperature and under argon, was treated with N-
chlorosuccinimide (59 mg, 0.44 mmol) and triphenylphosphine (87
mg, 0.33 mmol). The reaction mixture was stirred for 1 h, and the
solvent was removed under reduced pressure. The crude product was
purified by flash chromatography on silica gel, eluting with diethyl
ether/hexanes (5:95) to give chloride 25 (48 mg, 42%) as a white
foam. [α]2D0 −13.15 (c 1.0, CHCl3). H NMR (300 MHz, CDCl3) δ:
1
7.37−7.24 (m, 4H, 4 × ArH), 7.13 (dd, J = 6.5 and 1.6 Hz,1H, ArH),
6.00 (s, 1H, H2), 4.56−4.49 (m, 1H, CHCl), 4.43 (m, 1H, H5), 4.00
(d, J = 3.3 Hz, 1H, H4), 3.20 (m, 2H, CH2Ph), 2.30 (m, 2H, CH2-6),
0.94 (s, 9H,C(CH3)3), 0.90 (s, 9H,C(CH3)3), 0.20 (s, 3H, CH3), 0.14
(s, 3H, CH3), 0.13 (s, 3H, CH3), and 0.10 (s, 3H, CH3) ppm. 13C
NMR (63 MHz, CDCl3) δ: 175.2 (C), 137.9 (C), 136.8 (C), 135.3
(CH), 129.8 (2 × CH), 127.3 (2 × CH), 127.0 (CH), 75.7 (CH), 74.8
(C), 67.2 (CH), 61.0 (CH), 45.5 (CH2), 37.0 (CH2), 25.9 (C(CH3)3),
25.7 (C(CH3)3), 18.1 (C(CH3)3), 18.1 (C(CH3)3), −3.0 (2 × CH3),
foam. [α]2D0 −41.7 (c 1.0, CHCl3). H NMR (250 MHz, CDCl3) δ:
1
7.35−7.18 (m, 5H, 5 × ArH), 6.10 (s, 1H, H2), 4.66 (dd, J = 9.3 and
5.3 Hz, 1H, CHCl), 4.53 (dd, J = 5.6 and 3.2 Hz,1H, H5), 4.44 (d, J =
3.2 Hz, 1H, H4), 3.30 (dd, J = 14.3 and 5.3 Hz, 1H, CHHPh), 3.15
(dd, J = 14.3 and 9.3 Hz, 1H, CHHPh), 2.37 (ddd, J = 10.6, 5.6 and
1.5 Hz, 1H, H6eq), 2.27 (d, J = 10.6 Hz, 1H, H6ax), 0.92 (s,
3506
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
−4.1 (CH3), and −4.4 (CH3) ppm. IR (film): 1802 (CO) cm−1. MS
(ESI) m/z (%) 545 (MNa+). HRMS calcd for C27H23O4ClNa
(MNa+), 545.2281; found, 545.2290.
Structure Determination of Binary Complexes. Crystals were
harvested in reservoir solution supplemented with 20% (v/v) glycerol,
mounted into cryoloops, and flash-frozen by rapid immersion in liquid
nitrogen. X-ray diffraction data for the Mt-DHQ2/5 complex and Mt-
DHQ2/7,8 complexes were collected on beamlines ID23-2 and ID29
(ESRF, Grenoble, France), respectively, from crystals maintained at
100 K. Data for the Mt-DHQ2/6 complex were collected on beamline
BL13-Xaloc (Alba Synchrotron, Barcelona, Spain) from crystals
maintained at 100 K. The data were processed, scaled, and corrected
for absorption effects, and the crystal unit-cell parameters were
calculated by global refinement using XDS,38 SCALA,39 and other
programs from the CCP4 software suite.40 The structures of Mt-
DHQ2/5−8 binary complexes were solved by molecular replacement
using the program MOLREP41 with a search model generated from
PDB entry 1H0S.28
(1R,4R,5R)-1,4-Dihydroxy-3-[(1S)-1-chloro-2-phenyl]-
ethylcyclohex-2-en-1,5-carbolactone (28). A stirred solution of
the silyl ether 27 (66.5 mg, 0.13 mmol) in dry THF (3.0 mL), under
argon at 0 °C, was treated with tetrabutylammonium fluoride (0.3 mL,
0.3 mmol, ∼1.0 M in THF). The mixture was stirred for 15 min.
Saturated NH4Cl was added, and the organic layer was extracted with
diethyl ether (×3). The combined organic extracts were dried
(anhydrous Na2SO4), filtered, and concentrated under reduced
pressure. The crude product was purified by flash chromatography
on silica gel, eluting with (1:1:2) diethyl ether/acetone/hexanes to
give diol 28 (30 mg, 81%) as a white foam. [α]2D0 −3.2 (c 1.0, MeOH).
1H NMR (250 MHz, CD3OD) δ: 7.30−7.15 (m, 5H, 5 × ArH), 6.05
The structures and geometrical restraints of inhibitors were
generated with the PRODRG2 server42 and were manually placed
during the model building, which was performed with COOT.43
(s, 1H, H2), 4.79 (t, J = 6.8 Hz, 1H, CHCl), 4.57 (m, 1H, H5), 4.10
(d, J = 3.3 Hz, 1H, H4), 3.29 (dd, J = 13.8 and 6.6 Hz, 1H, CHHPh),
3.17 (dd, J = 13.8 and 7.3 Hz, 1H, CHHPh), and 2.30 (m, 1H, CH2-6)
ppm. 13C NMR (63 MHz, CD3OD) δ: 178.2 (C), 140.2 (C), 138.4
(C), 134.7 (CH), 130.8 (2 × CH), 129.3 (2 × CH), 127.8 (CH), 77.8
(CH), 74.1 (C), 66.6 (CH), 62.9 (CH), 45.6 (CH2), and 37.2 (CH2)
ppm. IR (film): 3409 (OH) and 1777 (CO) cm−1. MS (ESI) m/z (%)
317 (MNa+). HRMS calcd for C15H15O4ClNa (MNa+), 317.0551;
found, 317.0543.
44
Reflections for calculating Rfree were selected randomly. Refinement
of the models was performed with REFMAC,45 and final structure
validation was performed with MOLPROBITY.46 The data collection,
refinement, and model statistics are summarized in Table S1 of the
Supporting Information. Figures depicting structures were prepared
using PYMOL.47
Molecular Dynamics Simulations. Ligand Minimization.
Ligand geometries were optimized using a restricted Hartree−Fock
(RHF) method and a 6-31G(d,p) basis set, as implemented in the ab
initio program Gaussian 09.48 The charge distribution for each ligand
studied was obtained by fitting the quantum mechanically calculated
(B3LYP/cc-pVTZ and IEF-PCM as solvation model) molecular
electrostatic potential (MEP) of the geometry-optimized molecule to
a point charge model, as implemented in the assisted model building
with energy refinement (AMBER)49 suite of programs. The missing
bonded and nonbonded parameters were assigned by analogy or
through interpolation from those already present in the AMBER
database (GAFF).50
Generation and Minimization of the DHQ2−Ligand Complexes.
For Mt-DHQ2/5−8 binary complexes, simulations were carried out
using the enzyme geometries found in their respective crystal
structures (PDB entries 4CIV, 4CIW, 4CIX, and 4CIY, respectively).
For Mt-DHQ2/9,10 binary complexes, the enzyme geometries found
in the crystal structures of Mt-DHQ2 in complex with ligands 7 (PDB
entry 4CIX) and 8 (PDB entry 4CIY) were used after manual
replacement of ligands 7 and 8 by 9 and 10, respectively. Given that
the minimum catalytic unit of DHQ2 was experimentally determined
to be a trimer,30 the latter was used for modeling. For Mt-DHQ2/5
binary complex, unsolved residues were modeled using the Web-based
ModLoop server.32 For Mt-DHQ2/7,8 binary complexes, the missing
residues were modeled as PDB entry 2XB8.15 Hydrogens were added
to the protein using the Web-based PROPKA3.1 server,51 which
assigned protonation states to all titratable residues at the chosen pH
of 7.0. However, δ and/or ε protonation was manually corrected for
His102 (dual) of the active site because of mechanistic considerations
and on the basis of results from preliminary MD simulations.
Molecular mechanics parameters from the ff12SB and GAFF force
fields respectively were assigned to the protein and the ligands using
the LEaP module of AMBER 12.52 All terminal hydrogens were first
minimized in vacuum (2000 steps, half of them steepest descent, the
other half conjugate gradient). Energy minimization was carried out in
two stages using the implicit solvent GB model; first protein side
chains (2000 steps, idem) and second the entire complex (1000 steps,
idem). Thereafter, each molecular system was immersed in a truncated
octahedron containing TIP3P53 water molecules (10 Å radius) and
Na+ ions54 to achieve electroneutrality.
(1R,4R,5R)-1,4,5-Trihydroxy-3-[(1S)-1-chloro-2-phenyl]-
ethylcyclohex-2-en-1-carboxylic Acid (9). A solution of carbo-
lactone 28 (28 mg, 0.08 mmol) in THF (3.0 mL) at 0 °C was treated
with aqueous lithium hydroxide (0.10 mL, 0.10 mmol, 1 M), and the
mixture was stirred for 2 h. Water was added, and the THF was
removed under reduced pressure. The resulting aqueous solution was
washed with diethyl ether (×3), and the aqueous extract was treated
with Amberlite IR-120 (H+) until pH 6 was obtained. The resin was
filtered off and washed with Milli-Q water. The filtrate and the
washings were lyophilized to give acid 9 (3.0 mg, 11%) as a white
solid. [α]2D0 −5.8 (c 1.0, MeOH). 1H NMR (300 MHz, D2O) δ: 7.41−
7.36 (m, 5H, 5 × ArH), 5.79 (s, 1H, H2), 4.59 (dd, J = 8.3 and 4.2 Hz,
1H, CHCl), 4.28 (d, J = 7.3 Hz, 1H, H4), 3.97 (q, J = 7.3 Hz, 1H, H5),
3.13 (dd, J = 13.6 and 4.2 Hz, 1H, CHHPh), 2.90 (dd, J = 13.6 and 8.3
Hz, 1H, CHHPh), and 2.12 (d, J = 7.3 Hz, 2H, CH2-6) ppm. 13C
NMR (63 MHz, D2O) δ: 163.4 (C), 145.1 (C), 139.3 (C), 130.0 (2 ×
CH), 129.2 (2 × CH), 127.2 (CH), 125.1 (CH), 73.5 (C), 72.9 (CH),
72.8 (CH), 69.9 (CH), 41.9 (CH2), and 38.8 (CH2) ppm. IR (film):
3417 (OH) and 1640 (CO) cm−1. MS (ESI) m/z (%) 311 (M − H).
HRMS calcd for C15H17O5Cl (M − H), 311.1085; found, 311.1052.
Dehydroquinase Assays. Both enzymes were purified as
described previously.35,36 Concentrated solutions of Hp-DHQ2 (6.4
mg mL−1) or Mt-DHQ2 (2.4 mg mL−1) were stored in potassium
phosphate buffer (50 mM, pH 7.2), DTT (1 mM), and NaCl (150
mM). When required for assays, aliquots of the enzyme stock solutions
were diluted in water and buffer and stored on ice. Dehydroquinase
was assayed in the forward direction by monitoring the increase in
absorbance at 234 nm in the UV spectrum due to the absorbance of
the enone−carboxylate chromophore of 3-dehydroshikimic acid (2) (ε
(M−1 cm−1) of 12 000). Standard assay conditions were pH 7.0 at 25
°C in Tris-HCl (50 mM) for Hp-DHQ2 and in Tris-HOAc (50 mM)
for Mt-DHQ2. Each assay was initiated by the addition of the
substrate. Solutions of 3-dehydroquinic acid (1) were calibrated by
equilibration with DHQ2 and measurement of the change in the UV
absorbance at 234 nm due to the formation of the enone−carboxylate
chromophore of 3-dehydroshikimic acid (2). The Ki values of acids 5−
10 against DHQ2 were obtained from Dixon plots (1/v vs [I]) of
assay data. The initial rates at fixed enzyme and substrate
concentrations (0.25−1.4 Km) were measured in the absence and in
the presence of various inhibitor concentrations.
Crystallization of Mt-DHQ2/5−8 Binary Complexes. Apo-Mt-
DHQ2 crystals37 were soaked in 10 mM solutions of inhibitors 5−8 in
the crystallization mixture [32% (v/v) 2-methyl-2,4-pentanediol, 0.3 M
ammonium sulfate and 0.1 M 4-(2-hydroxyethyl)piperazine-1-
ethanesulfonic acid sodium salt (HEPES sodium salt), pH 7.5] for 6
h (inhibitors 5, 7, and 8) and 20 h (inhibitor 6).
Simulations. MD simulations were performed using the AMBER 11
suite of programs and Amber ff12SB force field. Periodic boundary
conditions were applied, and electrostatic interactions were treated
using the smooth particle mesh Ewald method (PME)55 with a grid
spacing of 1 Å. The cutoff distance for the nonbonded interactions was
9 Å. The SHAKE algorithm56 was applied to all bonds containing
3507
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
hydrogen, using a tolerance of 10−5 Å and an integration step of 2.0 fs.
Minimization was carried out in three steps, starting with the
octahedron water hydrogens, followed by solvent molecules and
sodium counterions and finally the entire system. The minimized
system was heated at 300 K (1 atm, 25 ps, a positional restraint force
constant of 50 kcal mol−1 Å−2). These initial harmonic restraints were
gradually reduced to 5 kcal mol−1 Å−2 (10 steps), and the resulting
systems were allowed to equilibrate further. MD simulations were
carried out for 10 ns. System coordinates were collected every 20 ps
for further analysis.
(3) Fischbach, M. A.; Walsh, C. T. Antibiotics for emerging
pathogens. Science 2009, 325, 1089−1093.
(4) Lewis, K. Platforms for antibiotic discovery. Nat. Rev. Drug
Discovery 2013, 12, 371−387.
(5) Bush, K.; Courvalin, P.; Dantas, G.; Davies, J.; Eisenstein, B.;
Huovinen, P.; Jacoby, G. A.; Kishony, R.; Kreiswirth, B. N.; Kutter, E.;
Lerner, S. A.; Levy, S.; Lewis, K.; Lomovskaya, O.; Miller, J. H.;
Mobashery, S.; Piddock, L. J. V.; Projan, S.; Thomas, C. M.; Tomasz,
A.; Tulkens, P. M.; Walsh, T. R.; Watson, J. D.; Witkowski, J.; Witte,
W.; Wright, G.; Yeh, P.; Zgurskaya, H. I. Tackling antibiotic resistance.
Nat. Rev. Microbiol. 2011, 9, 894−896.
(6) Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. The
challenge of new drug discovery for tuberculosis. Nature 2011, 469,
483−490.
(7) Butler, D. New fronts in an old war. Nature 2000, 406, 670−672.
(8) Walsh, C. Antibiotics: Actions, Origins, Resistance; ASM Press:
Washington, DC, 2003.
(9) Lamichhane, G.; Freundlich, J. S.; Ekins, S.; Wickramaratne, N.;
Nolan, S. T.; Bisha, W. R. Essential metabolites of Mycobacterium
tuberculosis and their mimics. mBio 2011, 2, e00301−e00310.
(10) DEG: Database of Essential Genes. www.essentialgene.org.
(11) Krell, T.; Pitt, A. R.; Coggins, J. R. The use of electrospray mass
spectrometry to identify an essential arginine residue in type II
dehydroquinases. FEBS Lett. 1995, 360, 93−96.
(12) Krell, T.; Horsburgh, M. J.; Cooper, A.; Kelly, S. M.; Coggins, J.
R. Localization of the active site of type II dehydroquinase.
Identification of a common arginine-containing motif in the two
classes of dehydroquinases. J. Biol. Chem. 1996, 271, 24492−24497.
ASSOCIATED CONTENT
* Supporting Information
■
S
Figures showing the reported crystal structures of Mt-DHQ2/
5−8 binary complexes, amino acid sequence alignments for
DHQ2 enzymes from different sources, and results from MD
simulation studies; 1H NMR, 13C NMR, and DEPT spectra and
Dixon plots for compounds 5−10. This material is available
free of charge via the Internet at http://pubs.acs.org.
Accession Codes
Coordinates and structure factors are available from the Protein
Data Bank with accession codes 4CIV, 4CIW, 4CIX, and 4CIY
for Mt-DHQ2/5, Mt-DHQ2/6, Mt-DHQ2/7, and Mt-DHQ2/
8 binary complexes, respectively.
AUTHOR INFORMATION
Corresponding Author
*Phone: 34 881 815726. E-mail: concepcion.gonzalez.bello@
usc.es.
■
(13) Coderch, C.; Lence, E.; Peon
́
, A.; Lamb, H.; Hawkins, A. R.;
Gago, F.; Gonzalez-Bello, C. Mechanistic insight into the reaction
́
catalyzed by bacterial type II dehydroquinases. Biochem. J. 2014, 458,
547−557.
́
(14) Harris, J.; Gonzalez-Bello, C.; Kleanthous, C.; Coggins, J. R.;
Author Contributions
⊥B.B. and A.S. contributed equally.
Hawkins, A. R.; Abell, C. Evidence from kinetic isotope studies for an
enolate intermediate in the mechanism of type II dehydroquinases.
Biochem. J. 1996, 319, 333−336.
Notes
The authors declare no competing financial interest.
́ ́
(15) Peon, A.; Otero, J. M.; Tizon, L.; Prazeres, V. F. V.; Llamas-Saiz,
A. L.; Fox, G. C.; van Raaij, M. J.; Lamb, H.; Hawkins, A. R.; Gago, F.;
Castedo, L.; Gonzalez-Bello, C. Understanding the key factors that
ACKNOWLEDGMENTS
́
■
control the inhibition of type II dehydroquinase by (2R)-2-benzyl-3-
dehydroquinic acids. ChemMedChem 2010, 5, 1726−1733.
We thank ESRF, beamlines ID23-2 and ID29 (Grenoble,
France), and the ALBA, synchrotron beamline BL13-XALOC
(Barcelona, Spain), for the provision of beam time. We are also
(16) Roszak, A. W.; Robinson, D. A.; Krell, T.; Hunter, I. S.;
Frederickson, M.; Abell, C.; Coggins, J. R.; Lapthorn, A. J. The
structure and mechanism of the type II dehydroquinase from
Streptomyces coelicolor. Structure 2002, 10, 493−503.
́
grateful to the Centro de Supercomputacion de Galicia
(CESGA) for use of the Finis Terrae computer. Financial
support from the Spanish Ministry of Science and Innovation
(Grant SAF2010-15076) and the Xunta de Galicia (Grant
GRC2013/041) is gratefully acknowledged. B.B. and A.P. thank
the Spanish Ministry of Education for their respective FPU
fellowships. A.S. thanks the Spanish Ministry of Economy and
Competitiveness for her FPI fellowship. J.M.O. thanks the
Xunta de Galicia for a Plan I2C postdoctoral fellowship.
(17) Blomberg, L. M.; Mangold, M.; Mitchell, J. B. O.; Blumberger, J.
Theoretical study of the reaction mechanism of Streptomyces coelicolor
type II dehydroquinase. J. Chem. Theory Comput. 2009, 5, 1284−1294.
́
(18) Tizon, L.; Otero, J. M.; Prazeres, V. F. V.; Llamas-Saíz, A. L.; van
Raaij, M. J.; Lamb, H.; Hawkins, A. R.; Ainsa, J. A.; Castedo, L.;
Gonzalez-Bello, C. A prodrug approach for improving anti-tuberculosis
́
activity of potent Mycobacterium tuberculosis type II dehydroquinase
inhibitors. J. Med. Chem. 2011, 54, 6063−6084.
(19) Blanco, B.; Sedes, A.; Peon
́
, A.; Lamb, H.; Hawkins, A. R.;
ABBREVIATIONS USED
■
Castedo, L.; Gonzalez-Bello, C. Synthesis of 3-alkyl enol mimics
́
inhibitors of type II dehydroquinase: factors influencing their
inhibition potency. Org. Biomol. Chem. 2012, 10, 3662−3676.
DHQ2, type II dehydroquinase; Hp-DHQ2, type II dehy-
droquinase from Helicobacter pylori; Mt-DHQ2, type II
dehydroquinase from Mycobacterium tuberculosis; QM, quantum
mechanics; MD, molecular dynamics; MM, molecular
mechanics; WAT1, conserved water molecule; MM-PBSA,
molecular mechanics Poisson−Boltzmann surface area
́
(20) Prazeres, V. F. V.; Tizon, L.; Otero, J. M.; Guardado-Calvo, P.;
Llamas-Saiz, A. L.; van Raaij, M. J.; Castedo, L.; Lamb, H.; Hawkins, A.
R.; Gonzalez-Bello, C. Synthesis and biological evaluation of new
́
nanomolar competitive inhibitors of Helicobacter pylori type II
dehydroquinase. Structural details of the role of the aromatic moieties
with essential residues. J. Med. Chem. 2010, 53, 191−200.
(21) Babine, R. E.; Blender, S. L. Molecular recognition of protein−
ligand complexes: applications to drug design. Chem. Rev. 1997, 97,
1359−1469.
(22) Lam, P. Y.; Jadhav, P. K.; Eyermann, C. J.; Hodge, C. N.; Ru, Y.;
Bacheler, L. T.; Meek, J. L.; Otto, M. J.; Rayner, M. M.; Wong, Y. N.;
Chang, C.-H.; Weber, P.; Jackson, D. A.; Sharpe, T. R.; Erickson-
REFERENCES
■
(1) Clatworthy, A. E.; Pierson, E.; Hung, D. T. Targeting virulence: a
new paradigm for antimicrobial therapy. Nat. Chem. Biol. 2007, 3,
541−548.
(2) Mckenna, M. Antibiotic resistance: the last resort. Nature 2013,
499, 394−396.
3508
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
Viitanen, S. Rational design of potent, bioavailable, nonpeptide cyclic
ureas as HIV protease inhibitors. Science 1994, 263, 380−384.
(23) Romines, K. R.; Watenpaugh, K. D.; Howe, W. J.; Tomich, P.
K.; Lovasz, K. D.; Morris, J. K.; Janakiraman, M. N.; Lynn, J. C.;
Horng, M.-M.; Chong, K.-T.; Hinshaw, R. R.; Dolak, L. A. Structure-
based design of nonpeptidic HIV protease inhibitors from a
cyclooctylpyranone lead structure. J. Med. Chem. 1995, 38, 4463−
4473.
(24) Vara Prasad, J. V. N.; Lunney, E. A.; Ferguson, D.; Tummino, P.
J.; Rubin, J. R.; Reyner, E. L.; Stewart, B. H.; Guttendorf, R. J.;
Domagala, J. M.; Suvorov, L. I.; Gulnik, S. V.; Topol, I. A.; Bhat, T. N.;
Erickson, J. W. HIV protease inhibitors possessing a novel, high-
affinity, and achiral P1′/P2/ligand with a unique pattern of in vitro
resistance. Importance of a conformationally-restricted template in the
design of enzyme inhibitors. J. Am. Chem. Soc. 1995, 117, 11070−
11074.
(38) Kabsch, W. XDS. Acta Crystallogr. 2010, D66, 125−132.
(39) Evans, P. Scaling and assessment of data quality. Acta Crystallogr.
2006, D62, 72−82.
(40) Winn, M. D. An overview of the CCP4 project in protein
crystallography: an example of a collaborative project. J. Synchrotron
Radiat. 2003, 10, 23−25.
(41) Vagin, A.; Teplyakov, A. MOLREP: an automated program for
molecular replacement. J. Appl. Crystallogr. 1997, 30, 1022−1025.
(42) Schuttelkopf, A. W.; van Aalten, D. M. F. PRODRG: a tool for
̈
high-throughput crystallography of protein−ligand complexes. Acta
Crystallogr. 2004, D60, 1355−1363.
(43) Emsley, P.; Cowtan, K. Coot: model-building tools for
molecular graphics. Acta Crystallogr. 2004, D60, 2126−2132.
(44) Brunger, A. T. Free R value: cross-validation in crystallography.
̈
Methods Enzymol. 1997, 277, 366−396.
(45) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of
macromolecular structures by the maximum-likelihood method. Acta
Crystallogr. 1997, D53, 240−255.
(25) Montchamp, J.-L.; Frost, J. W. Cyclohexenyl and cyclo-
hexylidene inhibitors of 3-dehydroquinate synthase: active site
interactions relevant to enzyme mechanism and inhibitor design. J.
Am. Chem. Soc. 1997, 119, 7645−7653.
(46) Davis, I. W.; Leaver-Fay, A.; Chen, V. B.; Block, J. N.; Kapral, G.
J.; Wang, X.; Murray, L. W.; Arendall, W. B., 3rd; Snoeyink, J.;
Richardson, J. S.; Richardson, D. C. MolProbity: all-atom contacts and
structure validation for proteins and nucleic acids. Nucleic Acids Res.
2007, 35, W375−W383.
(26) San
Hawkins, A. R.; Gonzal
and biological evaluation of inhibitors of Mycobacterium tuberculosis
type II dehydroquinase. J. Med. Chem. 2005, 48, 4871−4881.
́ ́
(27) Peon, A.; Coderch, C.; Gago, F.; Gonzalez-Bello, C.
́
chez-Sixto, C.; Prazeres, V. F. V.; Castedo, L.; Lamb, H.;
ez-Bello, C. Structure-based design, synthesis
́
(47) DeLano, W. L. The PyMOL Molecular Graphics System; DeLano
Scientific LLC: Palo Alto, CA, 2008; http://www.pymol.org.
(48) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Comparative binding energy (COMBINE) analysis for understanding
the binding determinants of type II dehydroquinase inhibitors.
ChemMedChem 2013, 8, 740−747.
(28) The X-ray crystal structure is available from the Protein Data
Bank (PBD code 1H0S): Robinson, D. A.; Roszak, A. W.;
Frederickson, M.; Abell, C.; Coggins, J. R.; Lapthorn, A. J. Structural
basis for specificity of oxime based inhibitors towards type II
dehydroquinase from M. tuberculosis. Unpublished material.
(29) For Mt-DHQ2/5 (PDB entry 4CIV): rms differences of 0.249
and 0.225 Å after superposition of 125 and 126 Cα-atom pairs of PDB
entries 2XB8 and 2Y71, respectively. For Mt-DHQ2/6 (PDB entry
4CIW): rms differences of 0.161 and 0.120 Å after superposition of
118 and 115 Cα-atom pairs of PDB entries 2XB8 and 2Y71,
respectively.
̈
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
(30) Price, N. C.; Boam, D. J.; Kelly, S. M.; Duncan, D.; Krell, T.;
Gourley, D. G.; Coggins, J. R.; Virden, V.; Hawkins, A. R. The folding
and assembly of the dodecameric type II dehydroquinases. Biochem. J.
1999, 338, 195−202.
(31) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K.
M.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. A second generation force field for the simulation of
proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 1995,
117, 5179−5197.
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09,
revision A.2; Gaussian, Inc.: Wallingford, CT, 2009.
(49) Case, D. A.; Cheatham, T. E.; Darden, T.; Gohlke, H.; Luo, R.;
Merz, K. M.; Onufriev, O.; Simmerling, C.; Wang, B.; Woods, R. J.
The AMBER biomolecular simulation program. J. Comput. Chem.
2005, 26, 1668−1688.
(50) Wang, J.; Wang, W.; Kollman, P. A.; Case, D. A. Automatic
atom type and bond type perception in molecular mechanical
calculations. J. Mol. Graphics Modell. 2006, 25, 247−260.
(51) Olsson, M. H. M.; Søndergard, C. R.; Rostkowski, M.; Jensen, J.
H. PROPKA3: consistent treatment of internal and surface residues in
empirical pKa predictions. J. Chem. Theory Comput. 2011, 7, 525−537.
(52) AMBER tools 13 and AMBER 12: Case, D. A.; Darden, T. A.;
Cheatham, T. E., III; Simmerling, C. L.; Wang, J.; Duke, R. E.; Luo, R.;
Walker, R. C.; Zhang, W.; Merz, K. M.; Roberts, B.; Hayik, S.;
(32) Fiser, A.; Sali, A. ModLoop: automated modeling of loops in
protein structures. Bioinformatics 2003, 19, 2500−2501.
(33) Bissantz, C.; Kuhn, B.; Stahl, M. A medicinal chemist’s guide to
molecular interactions. J. Med. Chem. 2010, 53, 5061−5084.
(34) Miller, B. R., III; McGee, T. D., Jr.; Swails, J. M.; Homeyer, N.;
Gohlke, H.; Roitberg, A. E. MMPBSA.py: an efficient program for end-
state free energy calculations. J. Chem. Theory Comput. 2012, 8, 3314−
3321.
́
Roitberg, A.; Seabra, G.; Swails, J.; Goetz, A. W.; Kolossvary, I.; Wong,
K. F.; Paesani, F.; Vanicek, J.; Wolf, R. M.; Liu, J.; Wu, X.; Brozell, S.
R.; Steinbrecher, T.; Gohlke, H.; Cai, Q.; Ye, X.; Wang, J.; Hsieh, M.-
J.; Cui, G.; Roe, D. R.; Mathews, D. H.; Seetin, M. G.; Salomon-Ferrer,
R.; Sagui, C.; Babin, V.; Luchko, T.; Gusarov, S.; Kovalenko, A.;
Kollman, P. A. AMBER 12; University of California: San Francisco,
CA, 2012.
(35) Prazeres, V. F. V.; San
́
chez-Sixto, C.; Castedo, L.; Shuh, S. W.;
ez-Barbero, J.;
́
ez-Bello, C. Competitive inhibitors of Helicobacter pylori type II
Lamb, H.; Hawkins, A. R.; Canada, F. J.; Jimen
Gonzal
́
̃
dehydroquinase: synthesis, biological evaluation, and NMR studies.
ChemMedChem 2008, 3, 756−770.
(36) Gourley, D. G.; Coggins, J. R.; Isaacs, N. W.; Moore, J. D.;
Charles, I. G.; Hawkins, A. R. Crystallization of a type II
dehydroquinase from Mycobacterium tuberculosis. J. Mol. Biol. 1994,
241, 488−491.
(37) Gourley, D. G.; Shrive, A. K.; Polikarpov, I.; Krell, T.; Coggins,
J. R.; Hawkins, A. R.; Isaacs, N. W.; Sawyer, L. The two types of 3-
dehydroquinase have distinct structures but catalyse the same overall
reaction. Nat. Struct. Biol. 1999, 6, 521−525.
(53) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D. Temperature
and size dependence for Monte Carlo simulations of TIP4P water. J.
Chem. Phys. 1983, 79, 926−935.
(54) Aqvist, J. Ion−water interaction potentials derived from free
energy perturbation simulations. J. Phys. Chem. 1990, 94, 8021−8024.
(55) Darden, T. A.; York, D.; Pedersen, L. G. Particle mesh Ewald:
An W log(N) method for Ewald sums in large systems. J. Chem. Phys.
1993, 98, 10089−10092.
3509
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510

Journal of Medicinal Chemistry
Article
(56) Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical
integration of the cartesian equations of motion of a system with
constraints: molecular dynamics of n-alkanes. J. Comput. Phys. 1977,
23, 327−341.
3510
dx.doi.org/10.1021/jm500175z | J. Med. Chem. 2014, 57, 3494−3510