Structure-based design of a new series of d-glutamic acid based inhibitors of bacterial UDP-N-acetylmuramoyl-l-alanine: D-glutamate ligase (MurD)

Article abstract of DOI:10.1021/jm2002525

MurD ligase is one of the key enzymes participating in the intracellular steps of peptidoglycan biosynthesis and constitutes a viable target in the search for novel antibacterial drugs to combat bacterial drug-resistance. We have designed, synthesized, and evaluated a new series of d-glutamic acid-based Escherichia coli MurD inhibitors incorporating the 5-benzylidenethiazolidin-4- one scaffold. The crystal structure of 16 in the MurD active site has provided a good starting point for the design of structurally optimized inhibitors 73-75 endowed with improved MurD inhibitory potency (IC₅₀ between 3 and 7 μM). Inhibitors 74 and 75 showed weak activity against Gram-positive Staphylococcus aureus and Enterococcus faecalis. Compounds 73-75, with IC ₅₀ values in the low micromolar range, represent the most potent d-Glu-based MurD inhibitors reported to date.

Full text of DOI:10.1021/jm2002525

ARTICLE
pubs.acs.org/jmc
Structure-Based Design of a New Series of D-Glutamic Acid Based
Inhibitors of Bacterial UDP-N-acetylmuramoyl-L-alanine:D-glutamate
Ligase (MurD)†
‡
‡
‡
‡
§
||,^,#
ꢀ
Tihomir Tomaꢀsiꢁc, Nace Zidar, Roman Sink, Andreja Kovaꢀc, Didier Blanot, Carlos Contreras-Martel,
Andrꢁea Dessen,||,^,# Manica M€uller-Premru,¥ Anamarija Zega,‡ Stanislav Gobec,‡ Danijel Kikelj,‡ and
Lucija Peterlin Maꢀsiꢀc*,‡
‡Faculty of Pharmacy, University of Ljubljana, 1000 Ljubljana, Slovenia
§Enveloppes Bactꢁeriennes et Antibiotiques, IBBMC, UMR 8619 CNRS, Universitꢁe Paris-Sud, 91405 Orsay, France
Bacterial Pathogenesis Group, ^Commissariat a l’Energie Atomique (CEA), and #Centre National pour la Recherche Scientifique
(CNRS), Institut de Biologie Structurale, Universitꢁe Grenoble I, Rue Jules Horowitz, 38027 Grenoble, France
¥Medical Faculty, Institute of Microbiology and Immunology, University of Ljubljana, 1105 Ljubljana, Slovenia
S Supporting Information
b
ABSTRACT:
MurD ligase is one of the key enzymes participating in the intracellular steps of peptidoglycan biosynthesis and constitutes a viable
target in the search for novel antibacterial drugs to combat bacterial drug-resistance. We have designed, synthesized, and evaluated a
new series of D-glutamic acid-based Escherichia coli MurD inhibitors incorporating the 5-benzylidenethiazolidin-4-one scaffold. The
crystal structure of 16 in the MurD active site has provided a good starting point for the design of structurally optimized inhibitors
73ꢀ75 endowed with improved MurD inhibitory potency (IC50 between 3 and 7 μM). Inhibitors 74 and 75 showed weak activity
against Gram-positive Staphylococcus aureus and Enterococcus faecalis. Compounds 73ꢀ75, with IC50 values in the low micromolar
range, represent the most potent D-Glu-based MurD inhibitors reported to date.
ATP-dependent Mur ligases (MurC to MurF) catalyze a series
of reactions leading to UDP-MurNAc-pentapeptide by sequen-
tially adding L-Ala (MurC), D-Glu (MurD), L-Lys or meso-diami-
nopimelic acid (MurE), and D-Ala-D-Ala dipeptide (MurF) to the
starting MurC substrate UDP-MurNAc.6,7 The fact that Mur
ligases are vital for the survival of bacteria makes them attractive
targets for antibacterial drug discovery.7,8
MurD ligase, like other Mur ligases, catalyzes the formation of
a peptide bond between the carboxyl group of the UDP substrate
and the amino group of the condensing D-glutamic acid. Mur
ligases operate through a similar chemical mechanism9,10 and, as
shown for MurC and MurF, an ordered kinetic mechanism.11,12
In the initial step, ATP is bound to the free enzyme, followed by
1. INTRODUCTION
The escalating emergence of bacterial strains resistant to
most of the currently available antibiotics has compromised the
treatment of bacterial infections and led to increased morbidity
and mortality worldwide.1 Development of effective antibacter-
ial drugs, necessitating novel mechanisms of action and more
chemical diversity than current drugs, is therefore essential to
combat bacterial drug-resistance.2ꢀ4 Peptidoglycan is an essen-
tial cell wall polymer unique to prokaryotic cells that preserves
cell integrity by withstanding high internal osmotic pressure
and maintaining a defined cell shape.5 The biosynthesis of
peptidoglycan (a viable source of targets for development of
novel antibacterials) is a multistep process comprising intracel-
lular assembly of the UDP-MurNAc-pentapeptide, which is
subsequently translocated through the cytoplasmic membrane
and incorporated into the growing peptidoglycan layer.
Received: March 3, 2011
Published: May 18, 2011
r
2011 American Chemical Society
4600
dx.doi.org/10.1021/jm2002525 J. Med. Chem. 2011, 54, 4600–4610
|

Journal of Medicinal Chemistry
ARTICLE
the corresponding UDP substrate. The terminal carboxyl group
of the UDP substrate is then activated by ATP-promoted
phosphorylation, generating an acyl phosphate intermediate that
is attacked by the amino group of the incoming amino acid or
dipeptide. The resulting tetrahedral, high-energy intermediate
breaks down with elimination of inorganic phosphate and
concomitant formation of a peptide or amide bond. The crystal
structures show MurD ligase to have a three-domain topology,
with the N-terminal and central domains binding the UDP
precursor and ATP, respectively, while the C-terminal domain
binds the condensing amino acid or dipeptide.9,13ꢀ16
Figure 1. Thiazolidine-2,4-dione- and rhodanine-based MurD inhibi-
tors I and II.28
Since MurD catalyzes the formation of UDP-MurNAc-L-Ala-
D-Glu (UMAG) by addition of D-Glu to the substrate UDP-
MurNAc-L-Ala (UMA), the D-Glu moiety is often an important
part of MurD inhibitors. For instance, phosphinates17,18 and
sulfonamides15,16 bearing a D-Glu moiety have been designed as
tetrahedral transition state or product analogue inhibitors.
MurD inhibitors have also been identified by virtual screen-
ing,19,20 by de novo structure-based drug design,21 by screening
phage display libraries,22 and by a new high-throughput fluori-
metric assay.23 Although several inhibitors of Mur ligases have
been reported by other research groups, many of them lack
antibacterial activity and none have been introduced to the
clinic.7,8,24
MurD. Moreover, the crystal structures of the representative
glutamic acid-based inhibitors I and 16, incorporating thiazoli-
dine-2,4-dione and rhodanine moieties, in complex with MurD,
support the specific action of this structural type of MurD
inhibitors, since 5-benzylidenerhodanines have recently been
presented as pan assay interference compounds (PAINS), espe-
cially in screening-based research. They were recognized as
frequent hitters that can covalently modify proteins by acting
as Michael acceptors and often represent false positive hits in a
variety of assays.30ꢀ32
2. RESULTS AND DISCUSSION
Recently, we reported several series of Mur inhibitors, includ-
ing multitarget inhibitors of Mur ligases25,26 and a series of
MurD-selective inhibitors represented by the most potent com-
pounds I and II (Figure 1).27,28 The design of MurD inhibitors
combining the D-Glu moiety and the thiazolidin-4-one ring was
guided by the highest ranked quinazoline-based compound in
virtual screening.27 Using structure-based design, through several
structure optimization cycles, we increased the MurD inhibitory
potency of this type of inhibitor up to 45 μM (compound II,
Figure 1).28 The crystal structure of the MurDꢀI complex
revealed that these inhibitors act as MurD product analogues,
since the D-Glu moiety is seen to occupy the D-Glu-binding
pocket and the thiazolidine-2,4-dione ring binds into the uracil-
binding pocket.28
In the present study, starting from I and II, we report the
design, synthesis, biological evaluation, and binding mode studies
of a new series of structurally optimized, D-Glu-based potent
MurD inhibitors containing the thiazolidin-4-one ring or its
surrogates. On the basis of the crystal structure of the MurDꢀI
complex and supported by ligand docking studies using GOLD
4.1,29 we examined the effects on MurD inhibitory activity of
different structural modifications of the parent compounds I and
II, including (i) addition of a fluorine or hydroxyl group to ring A
(compounds 15ꢀ17), (ii) replacement of the thiazolidine-2,
4-dione or rhodanine ring by pseudothiohydantoin, hydantoin,
cinnamic, and acetic acid moieties (compounds 28, 29, 34, 40,
45, and 54), (iii) replacement of the methyleneamino linker
between rings A and B by aminomethylene (compound 61), and
(iv) introduction of substituents with acidic character on the
linker amino group (compounds 70ꢀ75) in both the rhodanine-
and thiazolidine-2,4-dione-based analogues. These efforts resul-
ted in novel MurD inhibitors with improved potency, some of
which displayed weak antibacterial activity. In addition, we
present the crystal structure of a novel inhibitor 16 in complex
with MurD that reveals additional information on the binding
mode of the 5-benzylidenethiazolidin-4-one class of MurD
inhibitors and served in this work as a foundation for structure-
based design of compounds with improved potency against
2.1. Design. The crystal structure of MurD in complex with
thiazolidine-2,4-dione-based inhibitor I shows that the D-Glu
moiety that occupies the D-Glu-binding pocket, and the hetero-
cyclic ring that binds into the uracil-binding pocket form the
majority of hydrogen bonds with MurD active site residues, while
the linker between these two structural elements is involved
mainly in hydrophobic interactions and hydrogen-bonds to the
protein through crystal water molecules.28 The design of the new
series of MurD inhibitors was guided by information obtained
from this crystal structure and was supported by ligand docking,
using GOLD 4.1,29 into the MurD active site.
Given the importance of the D-Glu moiety for MurD inhibi-
tion, we left this part of the molecule unchanged and, in the first
step, modified phenyl ring A by introducing a fluorine or
hydroxyl group in order to achieve additional interactions with
the protein. Larger substituents on phenyl ring A have not been
considered, since our docking calculations showed that incor-
poration of larger groups on ring A resulted in a different binding
mode than that of the parent compounds. Candidates for the
synthesis (compounds 15ꢀ17, Scheme 1) had similar binding
modes and gave comparable or better scores according to
GOLDscore33 than the parent compounds. Although the experi-
mentally determined binding of 16 in the MurD active site
showed a similar binding mode to that of the parent compound II
(described below), the MurD inhibitory activity was weaker.
Since introduction of additional groups in ring A resulted in
different positioning of the part of the molecule between the
D-Glu and the rhodanine ring (which was apparently the cause of
the weaker activity against MurD), we decided to retain the
unsubstituted 1,3-phenylene ring A, present in parent com-
pounds I and II, as a building block for the newly designed
compounds.
In the next step, we studied the replacement of the
rhodanine or thiazolidine-2,4-dione ring by the pseudothio-
hydantoin and hydantoin rings (compounds 28 and 29,
Scheme 2) that have similar potential for forming the hydro-
gen bond between Oγ of Thr36 and the NH group at position
4601
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
Scheme 1. Synthesis of 5-Benzylidenerhodanine and 5-Benzylidenethiazolidine-2,4-dione Derivatives 15ꢀ17 Substituted on
Ring Aa
a Reagents and conditions: (a) TBTU, Et3N, CH2Cl2, room temp, 3 h; (b) 4-nitrobenzaldehyde, piperidine, AcOH, EtOH, microwave 150 ꢀC, 20 min;
(c) SnCl2 2H2O, EtOH, reflux, 2 h; (d) 4 or 5, NaCNBH3, AcOH, MeOH, room temp, 24 h; (e) 2 M LiOH, MeOH/H2O, room temp, 16 h.
3
Scheme 2. Synthesis of 5-Benzylidenepseudothiohydantoin and 5-Benzylidenehydantoin Derivatives 28 and 29a
a Reagents and conditions: (a) TBTU, Et3N, CH2Cl2, room temp, 3 h. (b) For the preparation of 22: 4-nitrobenzaldehyde, NaOAc, AcOH, reflux, 20 h.
For the preparation of 23: 4-nitrobenzaldehyde, NH4OAc, AcOH, reflux, 20 h. (c) For the preparation of 24: SnCl2 2H2O, EtOH, reflux, 2 h. For the
3
preparation of 25: H2, Pd/C, THF/DMF, room temp, 20 h. (d) For the preparation of 26: 19, Na(OAc)3BH, CF3COOH, CH2Cl2, room temp, 20 h.
For the preparation of 27: 19, NaCNBH3, AcOH, MeOH, room temp, 20 h. (e) 2 M LiOH, MeOH/H2O, room temp, 16 h.
3 of the thiazolidine-2,4-dione ring observed in the crystal
structure of the MurDꢀI complex.28 Further, as an acyclic
mimetic of the thiazolidin-4-one ring, the carboxylic acid
moiety was considered (cf. cinnamic acid derivative 34 and
phenoxyacetic acid analogues 40 and 45, Scheme 3). While
cinnamic acid derivative 34 is conformationally rigid, since it
retains conjugation between the phenyl ring and the carbonyl
of the carboxylate group through the double bond, the
phenoxyacetic acid derivatives 40 and 45 can be treated as
flexible unconjugated analogues of 34. A similar approach was
used in the case of 5-benzylthiazolidine-2,4-dione-based
PPARγ ligands, such as rosiglitazone, where the thiazoli-
dine-2,4-dione ring was replaced by the 2-hydroxy-3-phenyl-
propanoic acid moiety, e.g., in aleglitazar.34,35 The retained
formation of the hydrogen bond, observed between the NH
group of the thiazolidine-2,4-dione ring or the carboxylate
group of the propanoic acid moiety and the OH group
of Tyr473 in PPARγ-rosiglitazone and PPARγ-aleglitazar
complexes, respectively, makes acyclic carboxylic acids inter-
esting thiazolidin-4-one ring surrogates. Since, with introduc-
tion of the phenoxyacetic acid moiety in our MurD inhibitors,
the number of rotatable bonds is increased, the central part of
the inhibitor was rigidified by cyclization to the indole ring
(compound 54, Scheme 4). In addition to this change, in the
case of the indole-based compound 54, we have also replaced
the methyleneamino linker between phenyl rings A and B by
an aminomethylene group, which was also considered in
compound 61 (Scheme 5). Docking the designed compounds
28, 29, 34, 40, 45, 54, and 61 in the MurD active site
(Supporting Information Figure S2) showed the expected
binding modes, similar to that of the parent compound I,
and supported the decision to synthesize these potential
MurD inhibitors.
The diphosphate-MurNAc part of the UMAG molecule forms
several interactions with MurD active site residues.9 However,
superposition of the UMAG molecule and the thiazolidin-4-one-
4602
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
Scheme 3. Synthesis of Cinnamic and Acetic Acid Derivatives 34, 40, and 45a
a Reagents and conditions: (a) SOCl2, MeOH, room temp, 40 h; (b) SnCl2 2H2O, EtOH, reflux, 2 h; (c) 19, Na(OAc)3BH, AcOH, THF, room temp,
3
20 h; (d) 2 M LiOH, MeOH/H2O, room temp, 16 h; (e) (Boc)2O, 1,4-dioxane, room temp, 16 h; (f) ethyl bromoacetate, K2CO3, CH3CN, 60 ꢀC, 18 h;
(g) CF3COOH, CH2Cl2, room temp, 1 h; (h) H2, Pd/C, THF/EtOH, room temp, 4 h.
Scheme 4. Synthesis of Indole Core Containing Acetic Acid Derivative 54a
a Reagents and conditions: (a) ethyl bromoacetate, K2CO3, CH3CN, 60 ꢀC, 18 h; (b) SOCl2, DMF, CH2Cl2, 0 ꢀC, 45 min; (c) benzyl bromide, DBU,
DMF, room temp, 15 h; (d) 48, NaH, DMF, room temp, 18 h; (e) H2, Pd/C, THF, room temp, 1 h; (f) H-D-Glu(OMe)-OMe HCl, TBTU, Et3N,
3
CH2Cl2, room temp, 5 h; (g) 2 M LiOH, MeOH/H2O, room temp, 16 h.
Scheme 5. Synthesis of Derivative 61 with an Aminomethylene Linker between Rings A and Ba
a Reagents and conditions: (a) ethylene glycol, p-toluenesulfonic acid, benzene, reflux, 24 h; (b) 4-aminobenzoic acid, NaCNBH3, MeOH, room temp,
16 h; (c) 1 M HCl, THF/H2O, room temp, 18 h; (d) thiazolidine-2,4-dione, piperidine, AcOH, EtOH, reflux, 24 h; (e) H-D-Glu(OMe)-OMe HCl,
3
TBTU, Et3N, CH2Cl2, room temp, 5 h; (f) 2 M LiOH, MeOH/H2O, room temp, 16 h.
based inhibitors in the MurD active site reveals that the dipho-
sphate-MurNAc binding site is not occupied by the inhibitors,
thus providing the potential for additional interactions with the
active site residues that could improve the inhibitory potency of
the parent compounds I and II. Because of the acidic character of
the diphosphate moiety of UMAG, we have introduced nega-
tively charged oxalyl, malonyl, and succinyl moieties on the
amino group between rings A and B (compounds 70ꢀ75,
4603
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
Scheme 6. Synthesis of N-Acylated Compounds 70ꢀ75a
a Reagents and conditions: (a) 19, NaCNBH3, AcOH, MeOH, room temp, 20 h. (b) For the preparation of 64 and 67: ethyl oxalyl chloride, Et3N, DMF,
room temp, 24 h. For the preparation of 65 and 68: methyl malonyl chloride, Et3N, DMF, room temp, 24 h. For the preparation of 66 and 69: methyl
succinyl chloride, Et3N, DMF, room temp, 24 h. (c) 2 M LiOH, MeOH/H2O, room temp, 16 h.
Scheme 6) to explore the diphosphate-binding pocket and to try
to improve MurD inhibition by gaining additional interactions
with the active site residues.
amine 32. Compound 33 was obtained by sodium triacetoxybor-
ohydride promoted reductive amination of 32 with aldehyde 19
and, in the following step, hydrolyzed with aqueous lithium
hydroxide to yield target compound 34. Acetic acid derivatives
40 and 45 were synthesized using slightly modified synthetic
procedures because of the different starting materials. In the
case of 40, the amino group of 3-aminophenol (35) was first
protected as tert-butyl carbamate 36, which was then alkylated
using ethyl bromoacetate and potassium carbonate to yield 37.
The latter was deprotected with acidolysis using trifluoroacetic
acid to yield 38 which, after reductive amination with aldehyde
19 followed by alkaline hydrolysis of the obtained 39, gave target
compound 40. For the synthesis of 45, 4-nitrophenol (41) was
first alkylated with ethyl bromoacetate and the resulting inter-
mediate 42 hydrogenated to aromatic amine 43. Target com-
pound 45 was then obtained from 43 using the same synthetic
strategy as described for 40. Indole derivative 54 was synthesized
starting from commercially available indole-6-carboxylic acid
(49) that was first protected as benzyl ester 50 using benzyl
bromide and DBU. N-Alkylation of indole 50 with benzyl
chloride 48, using sodium hydride as a base, gave 51. Compound
48 had been obtained from 4-hydroxybenzyl alcohol (46) after
alkylation with ethyl bromoacetate, followed by nucleophilic
substitution of the hydroxyl group in 47 by chlorine using thionyl
chloride and a catalytic amount of N,N-dimethylformamide.39
Catalytic hydrogenation of 51 resulted in the deprotection of the
benzyl ester moiety. The resulting carboxylic acid 52 was
converted to D-Glu derivative 53 using TBTU-promoted cou-
pling with the dimethyl ester of D-Glu (1). The final compound
54 was obtained by alkaline hydrolysis of 53 with aqueous lithium
hydroxide.
Compound 61, which possesses an aminomethylene linker
between phenyl rings A and B, was synthesized as described in
Scheme 5. In the first step, terephthalaldehyde (55) was mono-
protected with ethylene glycol, using p-toluenesulfonic acid as
catalyst, to give the aldehyde 56. Reductive amination of
aldehyde 56 with 4-aminobenzoic acid afforded intermediate
57 which, after deprotection using 1 M hydrochloric acid, gave
aldehyde 58. 5-Benzylidenethiazolidine-2,4-dione derivative 59
was then prepared via a Knoevenagel condensation between
aldehyde 58 and thiazolidine-2,4-dione (6). The presence of only
one signal for the methyne proton at 7.77 ppm in the 1H NMR
spectrum again suggested the formation of only the Z isomer.
The target compound 61 was obtained after alkaline hydrolysis of
2.2. Chemistry. The syntheses of target compounds 15ꢀ17
with fluorine or a hydroxyl group on the phenyl ring A are outlined
in Scheme 1. In the first step, 5-(4-nitrobenzylidene)thiazolidine-
2,4-dione (8) and 5-(4-nitrobenzylidene)rhodanine (9) were
synthesized via Knoevenagel condensation between 4-nitroben-
zaldehyde and thiazolidine-2,4-dione (6) or rhodanine (7), using
microwave irradiation with glacial acetic acid and piperidine as
catalysts. The presence of only one signal in the 1H NMR spectra
for the methyne proton at 7.90 and 7.74 ppm, respectively,
suggested, according to previous studies, exclusive formation of
the thermodynamically stable Z-isomer.36,37 Reduction of the
nitro group in 8 and 9 with tin(II) chloride dihydrate gave amines
10 and 11 which, after reductive amination with benzaldehydes 4
or 5 using sodium cyanoborohydride, produced secondary amines
12ꢀ14. Benzaldehydes 4 and 5 had been synthesized using
TBTU-promoted coupling of D-glutamic acid dimethyl ester (1)
with benzaldehyde 2 or 3. Finally, target compounds 15ꢀ17 were
obtained after alkaline hydrolysis of dimethyl esters 12ꢀ14 with
aqueous lithium hydroxide.
The synthesis of compounds 28 and 29 incorporating pseu-
dothiohydantoin (20) and hydantoin (21) rings instead of the
thiazolidine-2,4-dione or rhodanine heterocycle is shown in
Scheme 2. First, 5-(4-nitrobenzylidene)pseudothiohydantoin
(22) and 5-(4-nitrobenzylidene)hydantoin (23) were obtained
via Knoevenagel condensation of 4-nitrobenzaldehyde with 20
or 21 using sodium or ammonium acetate as catalyst. Aromatic
amines 24 and 25 were synthesized by reduction of 22 with
tin(II) chloride dihydrate and by catalytic hydrogenation of 23,
respectively. In the next step, reductive amination of 24 or 25
with benzaldehyde 19, synthesized from 1 and 18, gave com-
pounds 26 and 27. Because of the very poor reactivity of 24, an
alternative method for reductive amination was necessary to
obtain 26. In this case, sodium triacetoxyborohydride in trifluor-
oacetic acid was used to yield 26 in reasonable yield.38 Pseu-
dothiohydantoin- and hydantoin-based analogues 28 and 29
were obtained after alkaline hydrolysis of methyl esters 26 and
27, using an aqueous solution of lithium hydroxide.
The syntheses of compounds possessing cinnamic and acetic
acid moieties (34, 40, 45, and 54) are outlined in Schemes 3 and
4. 4-Nitrocinnamic acid (30) was first protected as methyl ester
31 and then reduced using tin(II) chloride dihydrate to afford
4604
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
Table 1. Inhibitory Activities of Compounds 15ꢀ17, 28, 29, 34, 40, 45, 54, 61, and 70ꢀ75 against MurD ligase from E. coli
MurD RAa or IC50
compd
substitution
Z
MurD RA,a %
compd
X
n
MurD IC50,b μM
b
compd
X
Y
R
15
16
17
28
29
O
S
F
71%
34
40
45
54
61
1,4
1,3
1,4
CHdCH
OCH2
96
89
87
94
88
70
71
72
73
74
75
O
O
O
S
0
1
2
0
1
2
40 ( 1
15 ( 2
105 ( 4
5 ( 0.4
3 ( 0.3
7 ( 0.2
S
S
F
253 ( 27 μM
124 ( 3 μM
85%
S
S
OH
H
OCH2
NH
O
S
NH
H
99%
S
S
a Residual activity (in %) of the enzyme at 500 μM tested compound. Results are the mean of two independent experiments. Standard deviations were
within (10% of the mean. b Concentration of the inhibitor, where residual activity of the enzyme is 50%.
60, which was prepared via TBTU-promoted amide bond
formation between 59 and D-glutamic acid dimethyl ester
hydrochloride (1).
rhodanine and thiazolidine-2,4-dione surrogates introduced are
not favorable for binding to the uracil-binding pocket of the
MurD active site. The lack of activity of compounds 40, 45, and
54 could be attributed to greater flexibility of the phenoxyacetic
acid moiety of the molecules, compared to the 5-benzylide-
nethiazolidin-4-one moiety and/or to the loss of conjugation,
which has been observed for reduced analogues of I and II and
for compounds with flexible alkyl linkers replacing the benzyl-
amine moiety.28 Thiazolidine-2,4-dione 61 with 1,4-substitution
on the two phenyl rings also proved to be inactive against MurD
ligase, which further supports the necessity of the 1,3-substitu-
tion pattern in ring A for achieving potent MurD inhibition.28
Finally, introduction of carboxylate-containing acyl substituents,
such as oxalyl, malonyl, and succinyl groups, on the amino group
between rings A and B resulted in the most potent series of MurD
inhibitors, with IC50 values down to 3 μM. In general, rhodanine-
based compounds 73ꢀ75 were an order of magnitude better
than the thiazolidine-2,4-dione-based compounds 70ꢀ72. This
observation is consistent with our previous results showing that
the rhodanine-based compound II and its analogues were usually
more potent than their thiazolidine-2,4-dione-based counter-
parts, such as I.28 MurD inhibition by thiazolidine-2,4-diones
70ꢀ72 was dependent mainly on the chain length of the N-acyl
substituent. Compound 71, with a malonyl group, was the most
potent in the series, with an IC50 of 15 μM, followed by oxalyl-
substituted compound 70 with an IC50 of 40 μM. The weakest
inhibitor was compound 72, with a succinyl moiety, which had an
IC50 of 105 μM. The same trend is also seen in the rhodanine
series that exhibited IC50 values of 5 μM (73), 3 μM (74) and
7 μM (75), confirming the malonyl moiety as the optimal chain
length.
N-Acyl derivatives 70ꢀ75 were synthesized as outlined in
Scheme 6. Compounds 62 and 63, prepared from 10 or 11 and
19 as reported previously,28 were acylated using an excess of ethyl
oxalyl chloride, methyl malonyl chloride or methyl succinyl chlor-
ide and triethylamine, to yield 64ꢀ69. Finally, alkaline hydrolysis of
64ꢀ69 with aqueous lithium hydroxide gave compounds 70ꢀ75.
2.3. Biological Activity. Target compounds 15ꢀ17, 28, 29,
34, 40, 45, 54, 61, and 70ꢀ75 were assayed for inhibition of
MurD ligase from E. coli (Table 1). The Malachite green assay,40
which detects orthophosphate generated during the enzymatic
reaction, was used. All compounds were tested in the presence of
detergent (0.01% Triton X-100) to avoid nonspecific inhibition
due to aggregate formation.41 The results are presented as
residual activities (RAs) of the enzyme in the presence of 500 μM
of each compound and as IC50 values for the most active
compounds. They show that incorporation of the fluorine or
hydroxyl group in ring A reduces MurD inhibition. The thiazo-
lidine-2,4-dione-based compound 15 (RA = 71%) had almost
none of the MurD inhibitory activity of the parent compound I
(IC50 = 85 μM); that of the rhodanine-based compound 16, with
an IC50 of 253 μM, was more than 5-fold weaker than that of
compound II (IC50 = 45 μM). A comparison of the position of
ring A in the MurD active site of the crystal structures of
MurDꢀI and MurDꢀ16 complexes shows that the introduction
of fluorine results in different orientations of ring A (see below).
The hydroxyl-substituted compound 17 with IC50 of 124 μM
was almost 3-fold less potent than II, but the effect of the
hydroxyl group was less detrimental to the MurD inhibitory
activity than that of the fluorine atom. Pseudothiohydantoin-
(28), hydantoin- (29), cinnamic acid- (34), acetic acid- (40 and
45), and the rigidified indole- (54) based compounds were all
devoid of MurD inhibitory activity, which indicates that the
MurD ligase inhibitors 74 and 75 were tested for their
antibacterial activity against two Gram-positive (S. aureus
ATCC 29213, E. faecalis ATCC 29212) and two Gram-negative
(E. coli ATCC 25922, Pseudomonas aeruginosa ATCC 27853)
4605
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
enzyme active site residues results in a different position of
phenyl ring A, which is rotated by 48ꢀ relative to ring A of I in the
MurDꢀI complex. In the MurDꢀ16 complex, the uracil-binding
pocket is occupied by the rhodanine ring of the inhibitor, the
rhodanine ring participating in interplane stacking, with a salt
bridge between Asp35 and Arg37. The ring nitrogen forms a
hydrogen bond with the Oγ of Thr36. However, the hydrogen
bond between Thr36 and the inhibitor oxo group, as formed in
the MurDꢀI complex, is not observed in the MurDꢀ16 com-
plex, which may, together with the different ring A position,
account for the weaker MurD inhibitory potency of 16 with
respect to I. The crystal structure of compound 16 in complex
with MurD also provides evidence that the exocyclic double
bond exists in the Z-configuration.
2.5. Molecular Docking. Plausible binding modes of target
compounds 15, 17, 28, 29, 34, 40, 45, 54, 61, and 70ꢀ75 in the
active site of E. coli MurD ligase were determined by ligand
docking of the compounds using GOLD 4.1.29 In the crystal
structure of MurDꢀI complex (PDB entry 2X5O)28 the active
site was defined as the area within 10 Å from inhibitor I. First,
the system was validated by redocking I in the defined active
site. The program, in combination with the GOLDscore33
scoring function, was able to reproduce the experimentally
determined binding conformation of I with an all heavy atom
rmsd of only 0.72 Å. After the cocrystal structure of MurDꢀ16
(PDB entry 2Y68) was obtained, a similar experiment was
carried out by redocking 16 in the active site. The calculated
rmsd value between the experimental and calculated binding
conformation of 16 was 1.03 Å. These results confirm GOLD
4.1 with the use of GOLDscore as a suitable docking program
for predicting the binding modes of the new compounds
designed to target MurD.
Docking inhibitor 17 in the MurD active site shows a binding
mode similar to that of 16 and I (Figure 3a). The calculated
conformation of ring A in 17 is comparable to its conformation
in 16. The hydroxyl group on ring A is within hydrogen bonding
distance of the His183 side chain, which may account for the
better binding affinity of 17 than of 16, while the observed
conformation of the middle part of the inhibitor may be less
favorable than in I. Docked conformations of oxalyl-substituted
inhibitors 70 (Figure S2) and 73 (Figure 3b) show that the
oxalyl group is oriented away from the diphosphate-binding
pocket and forms a plausible hydrogen bond with the side chain
of Asn138. This additional interaction may account for the
better binding affinity of 70 and 73 than that of the parent
compounds I and II. Compounds 71 (Figure S2) and 74
(Figure 3c), with the malonyl substituent on the amino group
of the linker, exhibited the best inhibitory potencies in the two
series of N-acyl derivatives, which may be attributed to the
possible interaction of the carboxyl end of the malonyl moiety
with the guanidine group of Arg37, a residue that interacts with
the phosphate group of the UMAG molecule. However, the
weaker inhibitory activity of 72 than of I on the one hand and
the more potent activity of 75 (Figure 3d) than of II on the
other cannot be rationalized by the docking results, since
the terminal carboxylate of the succinyl group of both 72
(Figure S2) and 75 is within hydrogen bonding distance of
the Thr16 side chain hydroxyl group. The predicted additional
hydrogen bond with Thr16, which is also part of the dipho-
sphate-binding pocket of UMAG, results in the stronger
inhibition of MurD by 75, but the less potent MurD inhibition
by 72 may be attributed to conformational changes in other
Figure 2. X-ray binding mode of inhibitor 16 (in magenta) in the active
site of MurD from E. coli (PDB entry 2Y68). Hydrogen bonds between
16 and MurD active site residues (in yellow) and crystal water molecules
(in red) are presented as dashed green lines.
bacteria. Compound 74 was found to inhibit E. faecalis growth
weakly, with an MIC of 128 μg/mL, while compound 75
inhibited the growth of both Gram-positive bacteria with an
MIC of 128 μg/mL. Additionally, esters 68 and 69 were tested
as their potential prodrugs with better potential for entry into
the bacterial cell but were found to be inactive against all four
bacterial strains. All compounds were inactive against both
Gram-negative bacteria.
2.4. X-ray Crystallography. The binding mode of inhibitor
16 in the MurD active site was determined by solving the crystal
structure of the MurDꢀ16 complex (PDB entry 2Y68). Apo-
MurD13 was employed as a search model in a molecular
replacement experiment (Figure 2 and Figure S1 in the
Supporting Information). The crystal structure, solved to 1.5 Å,
shows that the inhibitor occupies the binding site of the MurD
product UMAG, which has already been observed in the case of
a previously reported thiazolidine-2,4-dione-based inhibitor I.28
The D-Glu moiety of the inhibitor in the MurDꢀ16 complex
occupies the same site as the D-Glu residue of the product
UMAG.9 The R-carboxyl group of the D-Glu moiety of inhibitor
16 forms a charge-based interaction with Nξ of Lys348 and is
additionally hydrogen-bonded to a conserved water molecule
W33 that is further hydrogen-bonded to Oγ of Thr321 and to the
carboxyl group of Asp182. The carboxyl group of the D-Glu side
chain is held in place by hydrogen bonds with Oγ and NR of
Ser415 and also with the backbone nitrogen of Phe422 and
W292; the latter further interacts, via W91, with the side chain
carboxyl group of Glu423. Water molecule W292 also forms a
hydrogen bond with water molecule W437, which is in direct
contact with the amide nitrogen of the D-Glu moiety of the
inhibitor, forming an extended ring-type system. The carbonyl
oxygen in the amide bond between D-Glu and ring A is hydrogen-
bonded to W332 which, through W426, is in contact with the
R-carboxyl group of the inhibitor. Additional interactions be-
tween the amino group in the para position of the 5-benzylide-
nerhodanine ring and W118, the latter being hydrogen-bonded
to Oγ of Thr16, also contribute to the recognition within the
active site. The aminobenzylidene moiety is held in place by
hydrophobic interactions with Gly73, while the phenyl ring A is
similarly held in place by hydrophobic interactions with Leu416
and by πꢀπ interactions with Phe161. Introduction of fluorine
that, contrary to our expectations, does not interact with the
4606
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
Figure 3. Superposition of inhibitor I (in magenta) and the best ranked GOLD-calculated conformation of inhibitor (in green): (a) 17; (b) 73; (c) 74;
(d) 75. For clarity, only the amino acid residues interacting with inhibitor I (according to the crystal structure) and additional residues forming plausible
hydrogen bonds (presented as black dashed lines) with oxalyl, malonyl, or succinyl carboxylate groups of 73ꢀ75 are shown.
parts of the molecule of 72. Moreover, the difference in the
potencies of thiazolidine-2,4-diones 70ꢀ72 compared to those
of rhodanines 73ꢀ75 can also not be interpreted by the
docking results.
4. EXPERIMENTAL SECTION
4.1. Molecular Docking. 4.1.1. Ligand and Protein Preparation.
Three-dimensional models of target compounds were built from a
standard fragment library, and their geometries were optimized using
the CHARMm force field42 with MMFF9443 partial atomic charges. The
Smart Minimizer algorithm as available in Accelrys Discovery Studio
2.5,44 running on a workstation with Intel Core i7 860 CPU processor,
8 GB RAM, two 750 GB hard drives and a Nvidia GT220 GPU graphic
card, running Centos 5.5, was used for energy minimization until the
gradient value was smaller than 0.001 kcal/(mol Å). Molecular docking
calculations were performed using GOLD program, version 4.1.29
5-Benzylidenethiazolidine-2,4-dione-based inhibitor I and water mol-
ecules were deleted from the crystal structure of MurDꢀI complex
(PDB code 2X5O), and hydrogen atoms were added to the protein using
GOLD. The amino acid residues within a radius of 10 Å around the
inhibitor were defined as the active site.
4.1.2. Docking Validation and Ligand Docking. In order to validate
GOLD 4.1 as a suitable docking program, compound I was docked into
the MurD active site (PDB code 2X5O)28 in 100 independent genetic
algorithm (GA) runs. The best ranked GOLD-calculated conformation
of I, using GOLDscore as a scoring function, had an all heavy atom rmsd
value of 0.72 Å compared to the experimentally determined conforma-
tion of I in MurD active site.
3. CONCLUSION
We have designed, synthesized, and evaluated a series of novel
D-glutamic acid-based inhibitors of E. coli MurD ligase that
incorporate the 5-benzylidenethiazolidin-4-one moiety. They
exhibit more potent inhibition than the parent compounds I
and II. The crystal structure of MurD ligase in complex with 16
shows a binding mode slightly different from the previously
determined binding for I in the MurD active site. Both crystal
structures have served as a foundation for the structure-based
design of the most potent series of MurD inhibitors 73ꢀ75,
which have acidic moieties attached to the amino group of the
linker between the two phenyl rings. They inhibited MurD with
IC50 values between 3 and 7 μM and showed weak antibacterial
activity. Inhibition of MurD ligase in the low micromolar range
makes compounds 73ꢀ75, to the best of our knowledge, the
most potent D-glutamic acid based MurD inhibitors reported
to date.
4607
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
The target compounds were docked in 100 independent GA runs.
The GA parameters were set as suggested by GOLD 4.1. Ligands with
rmsd value less than 1.5 Å were joined in clusters. Early termination was
allowed if the top 10 solutions were within 1.0 Å of the rmsd value.
GOLDscore33 was used as scoring function. The 10 best ranked docking
solutions were inspected visually, and the best ranked GOLD-calculated
conformation was used for analysis and representation. The figures were
prepared by Pymol.45
for C29H30N3O9S2, 628.1423; found, 628.1420. Anal. (C29H29N3O9S2)
C, H, N.
(R,Z)-Dimethyl 2-(3-((3-Methoxy-3-oxo-N-(4-((4-oxo-2-thioxothiazolidin-
5-ylidene)methyl)phenyl)propanamido)methyl)benzamido)pentanedioate
(68). Yield: 28%; orange crystals; mp 80ꢀ81 ꢀC; [R]2D0 þ18.1ꢀ (c 0.12,
MeOH). IR (KBr): ν = 3422, 2952, 2849, 1736, 1654, 1597, 1508, 1438,
1330, 1230, 1179, 1009, 840 cmꢀ1. 1H NMR (DMSO-d6): δ 1.96ꢀ2.18
(m, 2H, CHCH2CH2), 2.44 (t, 2H, J = 7.5 Hz, CHCH2CH2), 3.38 (s,
2H, COCH2CO), 3.57 (s, 3H, CH3), 3.58 (s, 3H, CH3), 3.64 (s, 3H,
CH3), 4.41ꢀ4.48 (m, 1H, CHCH2CH2), 5.01 (s, 2H, CH2N),
7.40ꢀ7.43 (m, 4H, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH), 7.60ꢀ7.63 (m, 3H,
CHdC, 2 ꢁ ArꢀH0), 7.70ꢀ7.77 (m, 2H, 2 ꢁ ArꢀH), 8.70 (d, 1H,
J = 7.5 Hz, CONH), 13.88 (br s, 1H, CONHCS) ppm. MS (ESIþ): m/z
(%) = 628 ([M þ H]þ, 100). HRMS (ESIþ): m/z [M þ H]þ calcd for
4.2. Chemistry. Chemicals were obtained from Acros, Aldrich
Chemical Co., and Fluka and used without further purification. Analytical
thin-layer chromatography was performed on silica gel Merck 60 F254
precoated plates (0.25 mm), using visualization with ultraviolet light,
ninhydrin, and 2,4-dinitrophenylhydrazine. Flash column chromatogra-
phy was carried out on silica gel 60 (particle size 0.040ꢀ0.063 mm;
Merck, Germany). HPLC analyses were performed on an Agilent
Technologies HP 1100 instrument with a G1365B UVꢀvis detector, a
G1316A thermostat, and a G1313A autosampler, using a Phenomenex
Luna C18 column (4.6 mm ꢁ 250 mm) at a flow rate of 1 mL/min. The
eluant was a mixture of 0.1% TFA in water (A) and methanol (B).
Gradient was 10% B to 80% B in 20 min. Melting points were determined
on a Reichert hot stage microscope and are uncorrected. 1H NMR and
13C NMR spectra were recorded at 300 and 75 MHz on a Bruker
AVANCE DPX300 spectrometer in CD3OD, CDCl3, or DMSO-d6
solution, with TMS as the internal standard. Spectra were assigned using
C29H30N3O9S2, 628.1423; found, 628.1438. Anal. (C29H29N3O9S2
0.6H2O) C, H, N.
3
(R,Z)-Dimethyl 2-(3-((4-Methoxy-4-oxo-N-(4-((4-oxo-2-thioxothiazolidin-
5-ylidene)methyl)phenyl)butanamido)methyl)benzamido)pentanedioate
(69). Yield: 43%; red oil; [R]2D0 þ8.4 (c 0.16, MeOH). IR (NaCl):
ν = 3650, 3340, 2945, 2832, 2577, 1831, 1741, 1602, 1361, 1268, 1196,
1023, 846, 740, 705 cmꢀ1. 1H NMR (DMSO-d6): δ 1.95ꢀ2.17 (m, 2H,
CHCH2CH2), 2.43 (t, 2H, J = 7.5 Hz, CHCH2CH2), 3.56 (s, 3H, CH3),
3.57 (s, 3H, CH3), 3.59 (s, 2H, COCH2CH2CO), 3.61 (s, 2H,
COCH2CH2CO), 3.63 (s, 3H, CH3), 4.40ꢀ4.48 (m, 1H, CHCH2CH2),
4.88ꢀ5.10 (m, 2H, CH2N), 7.37ꢀ7.47 (m, 4H, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH),
7.61ꢀ7.77 (m, 5H, 2 ꢁ ArꢀH0, CHdC, 2 ꢁ ArꢀH), 8.68 (d, 1H,
J = 7.5 Hz, CONH), 13.86 (br s, 1H, CONHCS) ppm. 13C NMR
(CDCl3): δ 27.1, 27.4, 30.3, 33.2, 35.5, 52.0, 52.3, 52.6, 53.2, 126.7, 126.9,
129.1, 129.7, 130.9, 131.9, 133.3, 133.7, 137.2, 142.9, 167.1, 168.1, 168.7,
172.0, 172.4, 172.8, 173.6, 176.2, 193.1 ppm. MS (ESIþ): m/z (%) = 642
([M þ H]þ, 100). HRMS (ESIþ): m/z [M þ H]þ calcd for
C30H32N3O9S2, 642.1580; found, 642.1604.
Synthesis of Compounds 73ꢀ75. To a stirred solution of dimethyl
ester 67ꢀ69 (0.20 mmol) in MeOH/water (1:1) (10 mL), 2 M LiOH
(1.40 mmol) was added, and the reaction mixture stirred overnight at
room temperature. The solution was neutralized with 1 M HCl and
concentrated under reduced pressure. The residual aqueous solution
was acidified with 1 M HCl to pH 2 and the product extracted with ethyl
acetate (3 ꢁ 15 mL). The combined organic phases were washed with
brine (2 ꢁ 15 mL), dried over Na2SO4, filtered, and the solvent was
evaporated under reduced pressure.
1
gradient COSY, HSQC, and H-coupled 13C NMR experiments. IR
spectrawererecorded ona Perkin-Elmer1600 FT-IR spectrometer. Mass
spectra were obtained using a VG-Analytical Autospec Q mass spectro-
meter. Optical rotations were measured on a Perkin-Elmer 241 MC
polarimeter. The reported values for specific rotation are average values
of 10 successive measurements, using an integration time of 5 s. Since
compounds 16, 17, 61ꢀ63, 73, and 74 showed ellipticity and circular
dichroism, their specific rotation values are not reported. Microanalyses
were performed on a Perkin-Elmer C, H, N analyzer 240 C. Analyses
indicated by the symbols of the elements were within 0.4% of the
theoretical values. The purity of the tested compound was established
to be g95%. Microwave-assisted reactions were performed using a
focused microwave reactor (Discover, CEM Corporation, Matthews,
NC). Reactions were carried out in septum-sealed glass vials (10 mL)
which enable high-pressure reaction conditions (max 20 bar). The
temperature of the reaction mixture was monitored using a calibrated
infrared temperature controller mounted under the reaction vessel.
Synthesis of Compounds 67ꢀ69. Compound 6328 (1 mmol) was dis-
solved in anhydrous tetrahydrofuran (10 mL) and triethylamine (10 mmol).
The solution was cooled to 0 ꢀC, and ethyl oxalyl chloride, methyl malonyl
chloride, or methyl succinyl chloride (5 mmol) was added. The reaction
mixture was stirred at room temperature for 18 h, and then the solvent was
evaporated under reduced pressure. The oily residue was dissolved in ethyl
acetate (30 mL) and successively washed with 10% aqueous citric acid
solution (2 ꢁ 30 mL), saturated aqueous NaHCO3 solution (2 ꢁ 30 mL),
and brine (30 mL), dried over Na2SO4, filtered and the solvent evaporated
under reduced pressure. The crude product was purified with flash column
chromatography using dichloromethane/methanol (20:1) as eluent.
(R,Z)-Dimethyl 2-(3-((2-Ethoxy-2-oxo-N-(4-((4-oxo-2-thioxothiazolidin-
5-ylidene)methyl)phenyl)acetamido)methyl)benzamido)pentanedioate
(67). Yield 58%; red solid; mp 70ꢀ71 ꢀC; [R]2D0 þ5.3ꢀ (c 0.25, MeOH).
IR (NaCl): ν = 3630, 3374, 3054, 2946, 2831, 1685, 1360, 1266, 1194,
1070, 1017, 964, 848, 742 cmꢀ1. 1H NMR (DMSO-d6): δ 0.93 (t, 3H,
J = 7.0 Hz, CH2CH3), 1.93ꢀ2.18 (m, 2H, CHCH2CH2), 2.44 (t, 2H,
J = 7.4 Hz, CHCH2CH2), 3.58 (s, 3H, CH3), 3.64 (s, 3H, CH3), 4.03 (q,
2H, J = 7.0 Hz, CH2CH3), 4.41ꢀ4.48 (m, 1H, CHCH2CH2), 5.10 (s,
2H, CH2N), 7.38ꢀ7.47 (m, 4H, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH), 7.61ꢀ7.63
(m, 3H, CHdC, 2 ꢁ ArꢀH0), 7.73ꢀ7.79 (m, 2H, 2 ꢁ ArꢀH), 8.75 (d,
1H, J = 7.4 Hz, CONH), 13.88 (br s, 1H, CONHCS) ppm. MS (ESIþ):
m/z (%) = 628 ([M þ H]þ, 100). HRMS (ESIþ): m/z [M þ H]þ calcd
(R,Z)-2-(3-((1-Carboxy-N-(4-((4-oxo-2-thioxothiazolidin-5-ylidene)-
methyl)phenyl)formamido)methyl)benzamido)pentanedioic Acid (73).
The crude product was purified with flash column chromatography using
dichloromethane/methanol/glacial acetic acid (12:1:1) as eluent. Yield:
51%; red solid; mp 140ꢀ142 ꢀC. IR (KBr): ν = 3422, 1718, 1638, 1430,
1285, 1232, 1180, 1066, 1008, 901, 811 cmꢀ1. 1H NMR (DMSO-d6): δ
1.97ꢀ2.12 (m, 2H, CHCH2CH2), 2.35 (t, 2H, J = 6.9 Hz, CHCH2CH2),
4.34ꢀ4.41 (m, 1H, CHCH2CH2), 5.08 (s, 2H, CH2N), 7.36ꢀ7.45 (m,
4H, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH), 7.56ꢀ7.59 (m, 3H, CHdC, 2 ꢁ ArꢀH0),
7.74ꢀ7.77 (m, 2H, 2 ꢁ ArꢀH), 8.60 (d, 1H, J = 7.6 Hz, CONH), 12.23
(br s, 4H, CONHCS, 3 ꢁ COOH) ppm. 13C NMR (CD3OD): δ 20.8,
27.5, 31.5, 53.8, 128.0, 128.5, 128.6, 129.4, 130.0, 131.2, 132.5, 132.9, 134.4,
134.7, 135.7, 138.1, 170.1, 170.9, 174.9, 175.0, 175.3, 176.7, 196.3 ppm. MS
(ESIþ): m/z (%) = 572 ([M þ H]þ, 100), 380 (57). HRMS (ESIþ):
m/z [M þ H]þ calcd for C25H22N3O9S2, 572.0797; found, 572.0790.
Anal. (C25H21N3O9S2 1.5H2O 0.5CH3COOH) C, H, N.
3
3
(R,Z)-2-(3-((2-Carboxy-N-(4-((4-oxo-2-thioxothiazolidin-5-ylidene)-
methyl)phenyl)acetamido)methyl)benzamido)pentanedioic Acid (74).
The crude product was purified with flash column chromatography using
dichloromethane/methanol/glacial acetic acid (12:1:0.5) as eluent. Yield
88%; orange crystals; mp 104ꢀ106 ꢀC. IR (KBr): ν = 3420, 2927, 2076,
1
1718, 1636, 1596, 1508, 1418, 1231, 1180, 1116, 1011, 840 cmꢀ1. H
NMR (DMSO-d6): δ 1.91ꢀ2.10 (m, 2H, CHCH2CH2), 2.34 (t, 2H,
4608
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
J = 7.1 Hz, CHCH2CH2), 3.28 (s, 2H, COCH2CO), 4.38ꢀ4.42 (m, 1H,
CHCH2CH2), 5.02 (s, 2H, CH2N), 7.26ꢀ7.44 (m, 4H, 2 ꢁ ArꢀH0, 2 ꢁ
ArꢀH), 7.60ꢀ7.74 (m, 5H, CHdC, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH), 8.55
(d, 1H, J= 7.6Hz, CONH), 12.50(brs, 3H, 3 ꢁ COOH), 13.86 (brs, 1H,
CONHCS) ppm. 13C NMR (CD3OD): δ 20.8, 27.5, 31.5, 53.8, 68.2,
127.9, 128.4, 128.8, 129.6, 129.9, 130.4, 131.1, 132.3, 132.9, 135.5, 138.6,
144.3, 170.1, 170.2, 170.8, 171.0, 175.0, 176.6, 196.3 ppm. MS (ESIþ):
m/z (%) = 586 ([M þ H]þ, 100). HRMS (ESIþ): m/z [M þ H]þ calcd
for C26H24N3O9S2: 586.0954, found: 586.0947. HPLC tR = 12.695 min
(95.20% at 220 nm, 95.88% at 254 nm).
’ REFERENCES
(1) Livermore, D. M. Bacterial resistance: origins, epidemiology, and
impact. Clin. Infect. Dis. 2003, 36, 11–23.
(2) Chopra, I.; Schofield, C.; Everett, M.; O’Neill, A.; Miller, K.;
Wilcox, M.; Frere, J. M.; Dawson, M.; Czaplewski, L.; Urleb, U.;
Courvalin, P. Treatment of health-care-associated infections caused by
Gram-negative bacteria: a consensus statement. Lancet Infect. Dis. 2008,
8, 133–139.
(3) Rice, L. B. Unmet medical needs in antibacterial therapy.
Biochem. Pharmacol. 2006, 71, 991–5.
(4) Nordmann, P.; Naas, T.; Fortineau, N.; Poirel, L. Superbugs in
the coming new decade; multidrug resistance and prospects for treat-
ment of Staphylococcus aureus, Enterococcus spp. and Pseudomonas
aeruginosa in 2010. Curr. Opin. Microbiol. 2007, 10, 436–440.
(5) Vollmer, W.; Blanot, D.; de Pedro, M. A. Peptidoglycan structure
and architecture. FEMS Microbiol. Rev. 2008, 32, 149–167.
(6) van Heijenoort, J. Recent advances in the formation of the
bacterial peptidoglycan monomer unit. Nat. Prod. Rep. 2001, 18,
503–519.
(7) Barreteau, H.; Kovaꢀc, A.; Boniface, A.; Sova, M.; Gobec, S.;
Blanot, D. Cytoplasmic steps of peptidoglycan biosynthesis. FEMS
Microbiol. Rev. 2008, 32, 168–207.
(8) El Zoeiby, A.; Sanschagrin, F.; Levesque, R. C. Structure and
function of the Mur enzymes: development of novel inhibitors. Mol.
Microbiol. 2003, 47, 1–12.
(R,Z)-2-(3-((3-Carboxy-N-(4-((4-oxo-2-thioxothiazolidin-5-ylidene)-
methyl)phenyl)propanamido)methyl)benzamido)pentanedioic Acid (75).
The crude product was purified with flash column chromatography using
dichloromethane/methanol/glacial acetic acid (12:1:1) as eluent. Yield: 23%;
orange solid; mp 133ꢀ135 ꢀC; [R]2D0 ꢀ15.0ꢀ (c 0.15, DMF). IR (KBr):
ν = 3422, 1718, 1635, 1406, 1229, 1178, 1010, 837, 747 cmꢀ1. 1H NMR
(DMSO-d6): δ 1.91ꢀ2.12 (m, 2H, CHCH2CH2), 2.24ꢀ2.36 (m, 4H,
CHCH2CH2, NCOCH2CH2CO), 2.58ꢀ2.86 (m, 2H, NCOCH2CH2CO),
4.34ꢀ4.41 (m, 1H, CHCH2CH2), 4.86ꢀ5.12 (m, 2H, CH2N), 7.37ꢀ7.48
(m, 4H, 2 ꢁ ArꢀH0, 2 ꢁ ArꢀH), 7.51ꢀ7.77 (m, 4H, 2 ꢁ ArꢀH0, CHdC,
2 ꢁ ArꢀH), 8.52 (d, 1H, J = 7.7 Hz, CONH), 12.18 (br s, 4H,
CONHCS, 3 ꢁ COOH) ppm. 13C NMR (CD3OD): δ 20.8, 27.8, 28.4,
31.6, 34.3, 54.1, 127.9, 128.2, 128.9, 129.9, 131.0, 131.1, 132.9, 133.8, 135.1,
135.6, 138.7, 144.1, 170.1, 170.6, 171.0, 175.7, 175.9, 176.8, 196.5 ppm. MS
(ESIþ): m/z (%) = 600 ([M þ H]þ, 100). HRMS (ESIþ): m/z [M þ
(9) Bertrand, J. A.; Auger, G.; Martin, L.; Fanchon, E.; Blanot, D.; Le
Beller, D.; van Heijenoort, J.; Dideberg, O. Determination of the MurD
mechanism through crystallographic analysis of enzyme complexes.
J. Mol. Biol. 1999, 289, 579–590.
H]þ calcd for C27H26N3O9S2, 600.1110; found, 600.1099. HPLC tR
=
11.833 min (96.06% at 220 nm, 96.63% at 254 nm).
(10) Bouhss, A.; Dementin, S.; van Heijenoort, J.; Parquet, C.;
Blanot, D. MurC and MurD synthetases of peptidoglycan biosynthesis:
borohydride trapping of acyl-phosphate intermediates. Methods Enzy-
mol. 2002, 354, 189–196.
(11) Emanuele, J. J.; Jin, H. Y.; Yanchunas, J.; Villafranca, J. J. Evalua-
tion of the kinetic mechanism of Escherichia coli uridine diphosphate-N-
acetylmuramate:L-alanine ligase. Biochemistry 1997, 36, 7264–7271.
(12) Anderson, M. S.; Eveland, S. S.; Onishi, H. R.; Pompliano, D. L.
Kinetic mechanism of the Escherichia coli UDPMurNAc-tripeptide D-
alanyl-D-alanine-adding enzyme: use of a glutathione S-transferase
fusion. Biochemistry 1996, 35, 16264–16269.
’ ASSOCIATED CONTENT
S
Supporting Information. Experimental procedures and
b
characterization of intermediate and target compounds, results
from microanalyses and HRMS, description of crystallization,
preparation of the inhibitor complex, data collection, crystal
structure solution, and model refinement, description of enzyme
assay and determination of antibacterial activity. This material is
available free of charge via the Internet at http://pubs.acs.org.
Accession Codes
†The PDB code for the Murꢀ16 complex is 2Y68.
(13) Bertrand, J. A.; Auger, G.; Fanchon, E.; Martin, L.; Blanot, D.;
van Heijenoort, J.; Dideberg, O. Crystal structure of UDP-N-acetylmur-
amoyl-L-alanine:D-glutamate ligase from Escherichia coli. EMBO J. 1997,
16, 3416–3425.
(14) Bertrand, J. A.; Fanchon, E.; Martin, L.; Chantalat, L.; Auger,
G.; Blanot, D.; van Heijenoort, J.; Dideberg, O. “Open” structures of
MurD: domain movements and structural similarities with folylpolyglu-
tamate synthetase. J. Mol. Biol. 2000, 301, 1257–1266.
’ AUTHOR INFORMATION
Corresponding Author
*Phone: þ386-1-4769635. Fax: þ386-1-4258031. E-mail: lucija.
peterlin@ffa.uni-lj.si.
(15) Kotnik, M.; Humljan, J.; Contreras-Martel, C.; Oblak, M.;
Kristan, K.; Hervꢁe, M.; Blanot, D.; Urleb, U.; Gobec, S.; Dessen, A.;
ꢀ
Solmajer, T. Structural and functional characterization of enantiomeric
’ ACKNOWLEDGMENT
glutamic acid derivatives as potential transition state analogue inhibitors
of MurD ligase. J. Mol. Biol. 2007, 370, 107–115.
This work was supported by the EU FP6 Integrated Project
EUR-INTAFAR (Project No. LSHM-CT-2004-512138), by the
Slovenian Research Agency (Grant No. P1-0208), and by the
World Federation of Scientists. The authors thank Professor
Roger Pain for critical reading of the manuscript.
(16) Humljan, J.; Kotnik, M.; Contreras-Martel, C.; Blanot, D.;
ꢀ
Urleb, U.; Dessen, A.; Solmajer, T.; Gobec, S. Novel naphthalene-N-
sulfonyl-D-glutamic acid derivatives as inhibitors of MurD, a key
peptidoglycan biosynthesis enzyme. J. Med. Chem. 2008, 51, 7486–7494.
(17) Tanner, M. E.; Vaganay, S.; van Heijenoort, J.; Blanot, D.
Phosphinate Inhibitors of the D-glutamic acid-adding enzyme of pepti-
doglycan biosynthesis. J. Org. Chem. 1996, 61, 1756–1760.
’ ABBREVIATIONS USED
ꢀ
(18) Strancar, K.; Blanot, D.; Gobec, S. Design, synthesis and
UDP, uridine 50-diphosphate; UMA, UDP-N-acetylmuramoyl-L-
Ala; UMAG, UDP-N-acetylmuramoyl-L-Ala-D-Glu; MurC, UDP-
N-acetylmuramate:L-Ala ligase; MurD, UDP-N-acetylmuramoyl-
L-Ala:D-Glu ligase; MurE, UDP-N-acetylmuramoyl-L-Ala-D-Glu:
meso-diaminopimelate ligase; MurF, UDP-N-acetylmuramoyl-L-
Ala-γ-D-Glu-meso-diaminopimelate:D-Ala-D-Ala ligase
structure-activity relationships of new phosphinate inhibitors of MurD.
Bioorg. Med. Chem. Lett. 2006, 16, 343–348.
(19) Perdih, A.; Kovaꢀc, A.; Wolber, G.; Blanot, D.; Gobec, S.;
Solmajer, T. Discovery of novel benzene 1,3-dicarboxylic acid inhibitors
ꢀ
of bacterial MurD and MurE ligases by structure-based virtual screening
approach. Bioorg. Med. Chem. Lett. 2009, 19, 2668–2673.
4609
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610

Journal of Medicinal Chemistry
ARTICLE
(20) Turk, S.; Kovaꢀc, A.; Boniface, A.; Bostock, J. M.; Chopra, I.;
Blanot, D.; Gobec, S. Discovery of new inhibitors of the bacterial
peptidoglycan biosynthesis enzymes MurD and MurF by structure-
based virtual screening. Bioorg. Med. Chem. 2009, 17, 1884–1889.
(21) Horton, J. R.; Bostock, J. M.; Chopra, I.; Hesse, L.; Phillips,
S. E.; Adams, D. J.; Johnson, A. P.; Fishwick, C. W. Macrocyclic
inhibitors of the bacterial cell wall biosynthesis enzyme MurD. Bioorg.
Med. Chem. Lett. 2003, 13, 1557–1560.
(22) Paradis-Bleau, C.; Beaumont, M.; Boudreault, L.; Lloyd, A.;
Sanschagrin, F.; Bugg, T. D.; Levesque, R. C. Selection of peptide
inhibitors against the Pseudomonas aeruginosa MurD cell wall enzyme.
Peptides 2006, 27, 1693–1700.
(38) Boros, E. E.; Thompson, J. B.; Katamreddy, S. R.; Carpenter,
A. J. Facile reductive amination of aldehydes with electron-deficient
anilines by acyloxyborohydrides in TFA: application to a diazaindoline
scale-up. J. Org. Chem. 2009, 74, 3587–3590.
(39) Percec, V.; Imam, M. R.; Peterca, M.; Wilson, D. A.; Heiney,
P. A. Self-assembly of dendritic crowns into chiral supramolecular
spheres. J. Am. Chem. Soc. 2009, 131, 1294–1304.
(40) Lanzetta, P. A.; Alvarez, L. J.; Reinach, P. S.; Candia, O. A.
Improved assay for nanomole amounts of inorganic-phosphate. Anal.
Biochem. 1979, 100, 95–97.
(41) McGovern, S. L.; Helfand, B. T.; Feng, B.; Shoichet, B. K. A
specific mechanism of nonspecific inhibition. J. Med. Chem. 2003, 46,
4265–4272.
(23) Kristan, K.; Kotnik, M.; Oblak, M.; Urleb, U. New high-
throughput fluorimetric assay for discovering inhibitors of UDP-N-
acetylmuramyl-L-alanine:D-glutamate (MurD) ligase. J. Biomol. Screening
2009, 14, 412–418.
(42) Brooks, B. R.; Bruccoleri, R. E.; Olafson, B. D.; States, D. J.;
Swaminathan, S.; Karplus, M. Charmm: a program for macromolecular
energy, minimization, and dynamics calculations. J. Comput. Chem. 1983,
4, 187–217.
ꢀ
(24) Kotnik, M.; Stefaniꢀc Anderluh, P.; Preꢀzelj, A. Development of
novel inhibitors targeting intracellular steps of peptidoglycan biosynth-
(43) Halgren, T. A. Merck molecular force field. 1. Basis, form,
scope, parameterization, and performance of MMFF94. J. Comput.
Chem. 1996, 17, 490–519.
esis. Curr. Pharm. Des. 2007, 13, 2283–2309.
ꢀ
(25) Sink, R.; Kovaꢀc, A.; Tomaꢀsiꢁc, T.; Rupnik, V.; Boniface, A.;
Bostock, J.; Chopra, I.; Blanot, D.; Peterlin Maꢀsiꢀc, L.; Gobec, S.; Zega, A.
Synthesis and biological evaluation of N-acylhydrazones as inhibitors of
MurC and MurD ligases. ChemMedChem 2008, 3, 1362–1370.
(26) Tomaꢀsiꢁc, T.; Zidar, N.; Kovaꢀc, A.; Turk, S.; Simꢀciꢀc, M.; Blanot,
D.; M€uller-Premru, M.; Filipiꢀc, M.; Grdadolnik, S. G.; Zega, A.;
Anderluh, M.; Gobec, S.; Kikelj, D.; Peterlin Maꢀsiꢀc, L. 5-Benzylide-
nethiazolidin-4-ones as Multitarget Inhibitors of Bacterial Mur Ligases.
ChemMedChem 2010, 5, 286–295.
(44) Accelrys Discovery Studio is available from Accelrys Inc., San
Diego, CA 92121, U.S.
(45) Pymol is available from Delano Scientific LLC, San Francisco,
CA; http://pymol.sourceforge.net.
(27) Tomaꢀsiꢁc, T.; Zidar, N.; Rupnik, V.; Kovaꢀc, A.; Blanot, D.;
Gobec, S.; Kikelj, D.; Peterlin Maꢀsiꢀc, L. Synthesis and biological
evaluation of new glutamic acid-based inhibitors of MurD ligase. Bioorg.
Med. Chem. Lett. 2009, 19, 153–157.
ꢀ
(28) Zidar, N.; Tomaꢀsiꢁc, T.; Sink, R.; Rupnik, V.; Kovaꢀc, A.; Turk, S.;
Patin, D.; Blanot, D.; Contreras Martel, C.; Dessen, A.; M€uller Premru,
M.; Zega, A.; Gobec, S.; Peterlin Maꢀsiꢀc, L.; Kikelj, D. Discovery of novel
5-benzylidenerhodanine and 5-benzylidenethiazolidine-2,4-dione inhi-
bitors of MurD ligase. J. Med. Chem. 2010, 53, 6584–6594.
(29) GOLD, version 4.1, is available from The Cambridge Crystal-
lographic Data Centre, 12 Union Road, Cambridge, CB2 1EZ, U.K.;
www.ccdc.cam.ac.uk.
(30) Baell, J. B.; Holloway, G. A. New substructure filters for removal
of pan assay interference compounds (PAINS) from screening libraries
and for their exclusion in bioassays. J. Med. Chem. 2010, 53, 2719–2740.
(31) Baell, J. B. Observations on screening-based research and some
concerning trends in the literature. Future Med. Chem. 2010, 2,
1529–1546.
ꢀ
(32) Sink, R.; Gobec, S.; Peꢀcar, S.; Zega, A. False positives in the early
stages of drug discovery. Curr. Med. Chem. 2010, 17, 4231–4255.
(33) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible docking.
J. Mol. Biol. 1997, 267, 727–748.
(34) Nolte, R. T.; Wisely, G. B.; Westin, S.; Cobb, J. E.; Lambert,
M. H.; Kurokawa, R.; Rosenfeld, M. G.; Willson, T. M.; Glass, C. K.;
Milburn, M. V. Ligand binding and co-activator assembly of the peroxi-
some proliferator-activated receptor-gamma. Nature 1998, 395, 137–143.
(35) Benardeau, A.; Benz, J.; Binggeli, A.; Blum, D.; Boehringer, M.;
Grether, U.; Hilpert, H.; Kuhn, B.; Marki, H. P.; Meyer, M.; Puntener,
K.; Raab, S.; Ruf, A.; Schlatter, D.; Mohr, P. Aleglitazar, a new, potent,
and balanced dual PPARalpha/gamma agonist for the treatment of type
II diabetes. Bioorg. Med. Chem. Lett. 2009, 19, 2468–2473.
(36) Ishida, T.; In, Y.; Inoue, M.; Ueno, Y.; Tanaka, C.; Hamanaka,
N. Structural elucidation of epalrestat(ono-2235), a potent aldose
reductase inhibitor, and isomerization of its double-bonds. Tetrahedron
Lett. 1989, 30, 959–962.
(37) Zidar, N.; Kladnik, J.; Kikelj, D. A convenient synthesis of 4-benzyl-
2-(2-(4-oxo-2-thioxothiazolidin-5-ylidene)ethyl)-2H-1,4-benzoxazin-
3(4H)-ones and 5-(2-(4-benzyl-3-oxo-3,4-dihydro-2H-1,4-benzoxazin-2-
yl)ethylidene)thiazolidine-2,4-diones. Acta Chim. Slov. 2009, 56, 635–642.
4610
dx.doi.org/10.1021/jm2002525 |J. Med. Chem. 2011, 54, 4600–4610