Structure-guided design of novel thiazolidine inhibitors of O -Acetyl serine sulfhydrylase from mycobacterium tuberculosis

Article abstract of DOI:10.1021/jm400710k

The cysteine biosynthetic pathway is absent in humans but essential in microbial pathogens, suggesting that it provides potential targets for the development of novel antibacterial compounds. CysK1 is a pyridoxalphosphate- dependent O-acetyl sulfhydrylase

Full text of DOI:10.1021/jm400710k

Article
pubs.acs.org/jmc
Structure-Guided Design of Novel Thiazolidine Inhibitors of O‑Acetyl
Serine Sulfhydrylase from Mycobacterium tuberculosis
†,∥
Omer Poyraz, Variam Ullas Jeankumar,‡,∥ Shalini Saxena,‡ Robert Schnell,† Martin Haraldsson,†
̈
Perumal Yogeeswari,‡ Dharmarajan Sriram,*,‡ and Gunter Schneider*,†
†Division of Molecular Structural Biology, Department of Medical Biochemistry & Biophysics, Karolinska Institutet, S-171 77
Stockholm, Sweden
‡Drug Discovery Research Laboratory, Department of Pharmacy, Birla Institute of Technology & SciencePilani, Hyderabad
Campus, Shameerpet, R.R. District, Hyderabad-500078, Andhra Pradesh, India
§Chemical Biology Consortium Sweden, Science for Life Laboratory, Department of Medical Biochemistry & Biophysics, Karolinska
Institutet, S-171 77 Stockholm, Sweden
S
* Supporting Information
ABSTRACT: The cysteine biosynthetic pathway is absent in
humans but essential in microbial pathogens, suggesting that it
provides potential targets for the development of novel antibacterial
compounds. CysK1 is a pyridoxalphosphate-dependent O-acetyl
sulfhydrylase, which catalyzes the formation of L-cysteine from O-
acetyl serine and hydrogen sulfide. Here we report nanomolar
thiazolidine inhibitors of Mycobacterium tuberculosis CysK1
developed by rational inhibitor design. The thiazolidine compounds
were discovered using the crystal structure of a CysK1−peptide
inhibitor complex as template. Pharmacophore modeling and
subsequent in vitro screening resulted in an initial hit compound 2
(IC50 of 103.8 nM), which was subsequently optimized by a
combination of protein crystallography, modeling, and synthetic
chemistry. Hit expansion of 2 by chemical synthesis led to improved thiazolidine inhibitors with an IC50 value of 19 nM for the
best compound, a 150-fold higher potency than the natural peptide inhibitor (IC50 2.9 μM).
nitrosative stress.9 In view of these findings, enzymes involved
in sulfur metabolism and cysteine biosynthesis have been
INTRODUCTION
■
Mycobacterium tuberculosis, the causative agent of tuberculosis
(TB), is one of the most devastating human pathogens
worldwide. Treatment of this disease is becoming increasingly
complicated due to the alarming occurrence of multidrug-
proposed as potential targets for the development of novel
antimycobacterial compounds.10,11
The classical CysK1 dependent pathway to produce cysteine
in M. tuberculosis uses hydrogen sulfide as sulfur source, which
resistant (MDR) or extremely drug-resistant (XDR) strains of
is derived from the APS−PAPS pathway.12 CysK1 from M.
tuberculosis binds specifically to a four residue long peptide
derived from the C-terminus of the serine acetyl transferase
CysE. This tetrapeptide, with the amino acid sequence DFSI
(1, Figure 1), binds in the active site cleft of the enzyme and is
a competitive inhibitor of CysK1 with a Ki of 5 μM.12
Previously, we reported the identification of novel inhibitors of
CysK1 from M. tuberculosis based on virtual high-throughput
screening.13 In the present study, we utilized the crystal
structure of the CysK1−DFSI complex as template in energy-
based pharmacophore (e-pharmacophore) modeling and in
silico docking to identify nonpeptidic compounds as putative
active site ligands. One thiazolidine compound, 3-({5-[2-
(carboxymethoxy)benzylidene]-3-methyl-4-oxo-1,3-thiazolidin-
M. tuberculosis.1,2 Further complications arise from the life cycle
of this bacterium, resulting in approximately one-third of the
world’s population carrying M. tuberculosis as latent infection,
which is difficult to treat with currently available antibiotics.
Latent TB is characterized by dormant bacteria with low
metabolic activity that can survive and replicate in macrophages
for years.3 The environment of dormant M. tuberculosis in
infected macrophages of the host is characterized by oxygen
depletion, nutrient depletion, and oxidative stress.4 To
inactivate damaging reactive oxygen species generated by
macrophages and hence to maintain redox homeostasis, M.
tuberculosis uses mycothiol as principal low-molecular-weight
antioxidant. This metabolite is present in millimolar concen-
tration in the cytoplasm and is derived from L-cysteine.5 Genes
involved in the sulfate reduction pathway and cysteine
biosynthesis have been found to be up-regulated in different
latency models using nutrient or oxygen depletion6−8 or
Received: May 13, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Structure of the inhibitory peptide DFSI used as template (left) and structure of the best hit 2, 3-({5-[2-(carboxymethoxy)benzylidene]-3-
methyl-4-oxo-1,3-thiazolidin-2-ylidene}amino)benzoic acid (right), from e-pharmacophore modeling used for further inhibitor improvement and
further SAR studies.
2-ylidene}amino)benzoic acid (2), within this set was identified
as a strong inhibitor of CysK1 with an IC50 in the nM range
(Figure 1) and chosen for improvement and further SAR
studies.
Compounds retrieved by the e-pharmacophore model using
phase with a fit value above 1.5 were regarded as potential hits
and were carried forward for high-throughput virtual screening.
Top compounds from this screen resulting in a score of ≥ −5.0
kcal mol−1 and docked with three or more hydrogen bonds
were subjected to another round of docking by Glide XP. The
GlideXP combines accurate, physics-based scoring terms and
thorough sampling, and the results gave scores ranging from
−9.38 to −7.10 kcal mol−1. Final shortlisting of possible hit
compounds was based on visual inspection of the important
amino acid residues in the active site cleft involved in binding
that included hydrogen bonds to Thr71, Gln144, Ser72, and
Lys215 and hydrophobic interactions with Phe145. The
selected hits retrieved from the Asinex database were
experimentally screened at a compound concentration of 2
mM against M. tuberculosis CysK1 using an assay based on the
spectrophotometric determination of the reaction product L-
cysteine adapted to 96-well plate format described previously.12
Twenty-eight compounds displayed >80% inhibition, and these
were rescreened at 100 μM inhibitor concentration. Seven out
of those showed >80% inhibition under these conditions
(Supporting Information Figure 2), with the thiazolidine
compound 2 (Figure 1) being the most potent inhibitor with
an IC50 of 103 nM.
To provide a structural basis for further inhibitor improve-
ment, we determined the crystal structure of the complex of
CysK1 with 2 to a resolution of 2.0 Å (Figure 3 and Supporting
Information Table 1). The overall structure of the enzyme in
the complex is very similar to the structure in the enzyme−
peptide complex (root-mean-square deviation value after
structural superposition of 0.3 Å for 300 equivalent Cα-
atoms). The electron density map showed that 2 was bound in
the active site cleft (Figure 3A). The experimentally observed
binding mode of the ligand agrees well with the predicted
binding from the e-pharmacophore modeling, with an rmsd of
0.8 Å after superimposition. However, there are significant
differences of 2 to the DFSI binding mode (Figure 3B). While
the DFSI peptide extends from the active site to the protein
surface, 2 is, with the exception of the 2-carboxymethylether
headgroup, bound entirely in the cleft between the two
domains. The benzoic acid is buried in the interior of the active
site with its carboxyl group occupying the same position as the
carboxyl groups of the C-terminal isoleucine residue of the
DFSI peptide and aminoacrylate reaction intermediate.12 One
of the carboxyl oxygen atoms of the benzoic acid moiety
interacts with the backbone amide nitrogen atom of Ser72 from
the conserved 71TSTGNT75 serine loop (Figure 4). The second
oxygen atom is within hydrogen bond distance to the side
chains of Thr71 and Gln144 and a water molecule. The phenyl
RESULTS AND DISCUSSION
■
In the first step, the e-pharmacophore approach, which
combines aspects of structure-based and ligand-based techni-
ques, was explored to identify small molecules that bind to the
active site of CysK1 using the crystal structure of the CysK1−
peptide inhibitor complex (PDB entry 2Q3C)12 as template.
The pharmacophore hypotheses based on mapping of the
energetic terms from the extra precision Glide scoring function
(Glide XP) (Glide, version 5.7, Schrodinger, LLC, New York,
̈
NY, 2011)14−16 onto atom centers was employed to derive
pharmacophore sites. These were based on the structural and
energy information between the protein and the ligand using
phase (Phase, v3.3, Schrodinger, LLC, New York, NY). The
̈
maximum number of pharmacophore sites derived for the DFSI
peptide was five with two acceptors (A), two negative ionizable
groups (N), and one aromatic ring (A) (Figure 2). The five-
point e-pharmacophore was then utilized for the virtual
screening of a commercial database (Asinex) using a protocol
described in more detail in the Experimental Section of the
Supporting Information and summarized in Figure 2.
Figure 2. Workflow for the identification of initial hits of
Mycobacterium tuberculosis CysK1 inhibitors by e-pharmacophore
modeling and in silico docking. The protocol started with the
structure of the enzyme with the bound peptide inhibitor, DFSI.
B
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
of the thiazolidine ring and the para-position of the benzylidine
moiety, respectively, that potentially could be explored for
modifications of the inhibitor scaffold (Figure 4B).
The promising results of the virtual screening hits and the 2.0
Å structure of the enzyme−hit complex encouraged us to
design a library with the goal of obtaining a lead series with
tractable SAR and potencies better than the identified initial
screening hits. Investigation of thiazolidinone and thiazolidine-
dione pharmacophores has recently generated significant
interest from a medicinal chemistry point of view,17−20 and
their synthesis has been well explored in the literature.21,22 The
synthetic pathway used to achieve the lead modifications is
delineated in Scheme 1 and described in more detail in the
Supporting Information.
On the basis of the input from protein−ligand interactions
observed in the structure of CysK1 with 2, the following
modifications (and combinations thereof) were explored in a
first ligand expansion step: (i) extending the alkyl chain
attached to the nitrogen (N-3) of the thiazolidinone core (−N-
methyl (R1)) with an ethyl, propyl, isopropyl, allyl, or butyl
group and also modifying the OCH2COOH (R3) substitution
at the ortho position of the phenyl ring in the hit molecule with
−OH, OCH2COOCH3 (9−17) and (ii) shifting these groups
from the ortho to the para position (18−48) to identify the
ideal sites for introducing chemical diversity and to have a
clearer understanding of the role of these substituents as
determinants of inhibitory potency.
Figure 3. (Upper panel) Stereo view of the unbiased Fo − Fc
difference electron density map at 3.0 σ for the bound ligand 2. The
electron density map is shown in blue, with the bound ligand shown as
a stick model. Loops constituting the active site cleft are colored in
magenta (N-terminal domain) and orange (C-terminal domain).
(Lower panel) Superposition of the peptide inhibitor DFSI (green)
with the thiazolidine compound 2 (yellow) bound to the active site of
CysK1.
A library of 40 derivatives (sublibrary 1) was synthesized
(compounds 9−48, Tables 1 and 2) and evaluated for their
ability to inhibit CysK1 as steps toward the derivation of
structure−activity relationships (SAR) and hit optimization.
The inhibition of M. tuberculosis CysK1 for compounds 9−17
(Table 1) at 500 nM was not encouraging because all
substitutions led to less potent molecules compared to the
initial hit 2. However, substitutions at the para-position of the
benzylidine ring highlighted this position as an important
determinant of inhibitory potency (18−48) (Table 2). Whereas
the analogues of OCH2COOH/OCH2COOCH3 modifications
in the para-position showed little inhibition of CysK1, the
corresponding hydroxy derivatives (18, 19, 22, 25, 28, and 31)
resulted in molecules with similar or improved IC50 values
compared to 2. Among these derivatives (Table 2), the order of
activity with respect to the N-substitution (R1) was methyl >
propyl > ethyl > allyl > isopropyl > butyl substitution, i.e.,
suggests a preference for a short alkyl side chain at this position.
Finally, replacement of the carboxylic acid group in the
secondary phenyl ring by an acetyl moiety (34−48) was
ring occupies the same position as the isoleucine side chain of
the peptide and packs against Phe145, Gly178, Gly222, and
Ala225. These interactions directly match those observed for
the DFSI C-terminal isoleucine residue. The thiazolidine core
mimics the phenyl ring of the DFSI peptide and occupies an
equivalent position within the hydrophobic pocket formed by
Met122, Phe145, Ala225, and Phe227. Its ring plane is almost
perpendicular to that of the benzoic acid moiety. The keto
group in ring position 2 points toward the protein surface and
forms hydrogen bonds with the side chain of Lys215 and a
water molecule. Distinct to the DFSI peptide, whose N-
terminal aspartic acid residue extends to the protein surface, the
benzylidene moiety of 2 occupies a hydrophobic pocket inside
the cleft (Figures 3 and 4). This pocket is located close to the
protein surface and is surrounded by Pro70, Ile126, Phe145,
and Phe227. Finally, the 2-carboxymethoxy substitution in ring
position 10 extends toward the solvent and neither the ether
oxygen atom nor the carboxyl headgroup are involved in any
hydrogen bonding. In addition the structure of the CysK1−2
complex revealed two pockets adjacent to the N3-methyl group
Figure 4. (A) Stereo view of the active site in the CysK1−2 complex. The amino acid side chains (green) and the bound thiazolidine inhibitor 2
(yellow) are shown as sticks and water molecules are shown as red spheres. Hydrogen bonds are indicated by dashed lines. (B) Surface
representation of the ligand binding site in the CysK1−2 complex highlighting pocket 1 (green) proximal to the thiazolidine core and pocket 2
(blue) adjacent to the benzylidine moiety of the ligand.
C
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
a
Scheme 1. Synthetic Protocol to Thiazolidine Analogues
a
(a) (C2H5)3N, CHCl3, 0 °C to rt, 12 h; (b) BrCH2COOH, CH3COONa, C2H5OH, reflux; (c) substituted aldehydes, piperidine, ethanol reflux.
molecules, with the order of activity fluoro (87) > O-benzyl
(57) > O-methyl (51) > N,N-dimethylamino (93). Other
substitutions like bromo, chloro, isopropyl, nitro, or methyl
were less active as inhibitors of CysK1.
Table 1. Activities of Synthesized Compounds (9−17)
Multiple substitutions were also studied including 2,4-
dichloro (63, 64), 3,4-dichloro (89, 90), 3,4-dimethoxy (83,
84), 3-methoxy, 4-hydroxy (81, 82), and 3,4,5-trimethoxy (95,
96) with N-methyl or ethyl groups. Among these, 3-methoxy,
2-hydroxy (81), 3,4-dimethoxy (83, 84), and 3,4,4-trimethoxy
(95) analogues were found to be very active and their IC50
values were determined (Table 3). Efforts were also made to
replace the primary benzyl ring by heteroaryl ring systems, and
10 compounds were synthesized (Table 4). Replacement of the
benzyl ring by furan (97, 98), pyrrole (99, 100), pyridine (101,
102), 4-trifluoromethyl, 3-pyridyl (103, 104), or indole (105,
106) moieties resulted in significant reduction in activity,
compared to the best inhibitors from the previous modification
step, thus emphasizing a favorably substituted benzyl ring as the
primary aryl ring. Among the thiazolidine analogues studied in
this work, the 4-fluoro derivative (87) emerged as the most
potent analogue, with an IC50 of 19 nM, five times more active
than the initial hit compound (2) and 150 times more potent
than the natural peptide DFSI (1), the starting point of the
inhibitor design in this study.
Overall, the SAR exploring pocket 2 adjacent to the
benzylidine ring agreed well with the crystal structure of the
complex of CysK1 with 2. Modeling of the substitutions at the
p-position of the benzylidine ring of 2, which led to the best
inhibitors, suggests that, in particular, interactions with side
chains of residues Pro70, Ile126, Ala129, Val141, and Gln143
could contribute to the increase in binding strength.
Substitutions at the N atom of the thiazolidine ring did,
however, contrary to expectations, not lead to significant
increases in inhibitory activity.
compd
R1
R2
% inhibition at 500 nM
9
methyl
methyl
ethyl
OH
62.7 9.3
56.3 4.5
53.2 6.3
31.0 10.9
15.9 7.7
38.6 7.8
25.4 5.6
55.7 9.0
18.5 7.0
10
11
12
13
14
15
16
17
OCH2C(O)OCH3
OH
ethyl
OCH2C(O)OCH3
OCH2C(O)OH
OCH2C(O)OCH3
OCH2C(O)OH
OCH2C(O)OCH3
OCH2C(O)OH
ethyl
allyl
allyl
isopropyl
isopropyl
detrimental to bioactivity as it resulted in inactive molecules
(Table 2). This observation highlights the importance of a
carboxyl group at this position and is consistent with the tight
interactions through hydrogen bonding observed in the
complexes of CysK1 with the reaction intermediate, the
peptide inhibitor,12 the initial hit 2 (Figure 2A) as well as
CysK−peptide complexes from other species.23−25
On the basis of these observations, a new sublibrary was
designed and synthesized that incorporated the importance of a
small alkyl chain at the R1 position and further explored pocket
2 by a series of substitutions on the phenyl ring. This sublibrary
of 60 molecules (49−106) was designed and synthesized by
modifying various parts of the ligand as shown in Tables 3 and
4 using commercially available substituted benzaldehydes and
heterocyclic aldehydes and employing a similar protocol as
outlined in Scheme 1. In vitro characterization of these
compounds showed that for substitutions at the C-2 position
only the methoxy derivative with N-methyl substitution 67 gave
a potent inhibitor (IC50 160 nM), whereas substitutions with
bromo, chloro, O-benzyl, or methyl groups showed significantly
less inhibitory activity. At the C-3 position trifluoromethyl and
methyl substitutions were studied (59, 60, 77, 78) and found to
be much less active than the parent compound 2. Substitutions
at the C-4 position resulted in several strongly inhibitory
Efforts to obtain well-diffracting crystals of CysK1 with the
most potent inhibitors were not successful. For compound 22,
we obtained crystals of the enzyme−ligand complex that
however only diffracted to 3.8 Å resolution. The electron
density map revealed binding of the inhibitor in the active site
cleft of CysK1, but the limited resolution did not allow an
unambiguous fit of the inhibitor molecule to the density map.
As an alternative method to ascertain that the inhibitors bind in
the active site cleft adjacent to the PLP cofactor, we employed a
fluorescence-based method whereby the change in PLP
D
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 2. Activities of Synthesized Compounds (18−48)
% inhibition at
% inhibition at
500 nM
IC50
compd
R1
R2
500 nM
IC50 (nM)
compd
R1
ethyl
R2
(nM)
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
methyl
ethyl
OH
OH
92.6 0.9
90.2 3.3
9.9 8.6
62.6 12.5
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
OH
3.4 9.4
8.2 5.1
10.0 1.5
10.7 9.4
0.0 3.2
0.0 8.1
28.9 5.0
10.9 3.7
0.0 8.3
8.2 5.0
7.4 6.5
10.2 8.4
25.1 4.7
0.0 5.5
24.1 2.1
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
101.4 6.7
ethyl
OCH2C(O)OCH3
OCH2C(O)OH
OH
ethyl
OCH2C(O)OCH3
OCH2C(O)OH
OH
nd
ethyl
ethyl
36.9 7.0
95.1 15.5
nd
propyl
propyl
propyl
allyl
propyl
propyl
propyl
allyl
95.1 15.5
OCH2C(O)OCH3
OCH2C(O)OH
OH
OCH2C(O)OCH3
OCH2C(O)OH
OH
0
0
7.7
5.6
nd
nd
92.3 0.9
19.9 9.7
1.4 8.6
90.3 0.5
110.4 24.8
allyl
OCH2C(O)OCH3
OCH2C(O)OH
OH
allyl
OCH2C(O)OCH3
OCH2C(O)OH
OH
nd
allyl
allyl
nd
isopropyl
isopropyl
isopropyl
butyl
isopropyl
isopropyl
isopropyl
butyl
140.6 13.5
OCH2C(O)OCH3
OCH2C(O)OH
OH
OCH2C(O)OCH3
OCH2C(O)OH
OH
0
4.7
nd
nd
nd
nd
nd
12.6 6.6
78.5 1.5
16.4 14.0
butyl
OCH2C(O)OCH3
butyl
OCH2C(O)OCH3
butyl
OCH2C(O)OH
butyl
OCH2C(O)OH
0
8.6
fluorescence in response to ligand binding is monitored.26 The
peptide inhibitor DFSI (1) and inhibitor 2 were used as
positive controls as their binding in the proximity of the
cofactor was verified by crystal structure analysis. The tested
compounds 18, 19, 22, 25, 28, 51, and 57 showed the expected
spectral changes in PLP fluorescence as the control consistent
with binding of these inhibitors in the active site cleft of CysK1
(Supporting Information Figure 2).
comprehensive SAR was obtained by synthesizing a library of
98 compounds to understand the importance of substitutions in
the phenyl ring system and the nitrogen of the thiazolidine
nucleus. These findings were used to develop inhibitor 87 (IC50
19 nM), the most potent CysK1 inhibitor described so far, with
a 150 times higher potency than the original DFSI tetrapeptide
(IC50 2.9 μM). This inhibitor, as well as other compounds in
this series, have favorable log P values (2.7−5.3), and the
ADME values predicted using the Schrodinger software are in
̈
acceptable ranges. The study provides the basis for further
chemical optimization of these potent thiazolidine inhibitors.
CONCLUSIONS
■
Previous efforts to develop inhibitors against CysK1 from
pathogenic organisms were directed toward improvements of
natural peptide inhibitors.25,27 Given the unfavorable drug-like
properties of peptides, alternative approaches also involved
virtual screening13,28 and chemical synthesis29 to identify
nonpeptidic inhibitors of CysK. These efforts led to small-
molecule inhibitors of moderate potency, with IC50 or KD
values in the low μM range. In this study, a combination of
ligand-based and structure-based e-pharmacophore modeling
has been employed to identify structurally diverse small-
molecule ligands of M. tuberculosis CysK1 based on the crystal
structure of the enzyme with a peptide inhibitor. The
thiazolidine scaffold 2 was identified as a potent inhibitor of
CysK1 with an IC50 of 103 nM. We determined the crystal
structure of the complex of the enzyme with this inhibitor, the
first CysK1−nonpeptide inhibitor. The structure revealed
hitherto unknown binding pockets and inhibitor binding
modes distinct from the peptide binding sites which could be
exploited successfully in inhibitor development. Our initial SAR
data validated the presence of the carboxyl group of the
secondary phenyl ring of the scaffold as essential for strong
binding of the inhibitor, consistent with previous structural
data.12,23−25 Further the para-position of the benzylidene group
was sensitive to changes and crucial for CysK1 inhibition. A
EXPERIMENTAL PROCEDURES
Computational Details. PHASE 3.3 implemented in the Maestro
■
9.2 software package (Schrodinger, LLC) was used to derive the e-
̈
pharmacophore. Glide energy grids were generated for each of the
prepared complexes. The binding site was defined by a rectangular box
surrounding the ligand in the X-ray structure. Ligands were refined
using the “Refine” option in Glide, and the Glide XP (extra precision)
descriptor was chosen (Glide v5.7, Schrodinger, LLC, New York, NY).
Default settings were used for the refinement and scoring.
Chemistry. All commercially available chemicals and solvents were
used without further purification. TLC experiments were performed
on alumina-backed silica gel 40 F254 plates (Merck, Darmstadt,
Germany). The homogeneity of the compounds was monitored by
thin layer chromatography (TLC) on silica gel 40 F254 coated on
aluminum plates, visualized by UV light and KMnO4 treatment. All 1H
and 13C NMR spectra were recorded on a Bruker AM-300 (300.12
MHz, 75.12 MHz) NMR spectrometer, BrukerBioSpin Corp.,
Germany. Molecular weights of the synthesized compounds were
checked by LCMS 6100B series Agilent Technology. Chemical shifts
are reported in ppm (δ) with reference to the internal standard TMS.
The signals are designated as follows: s, singlet; d, doublet; dd, doublet
of doublets; t, triplet; m, multiplet. Elemental analyses were carried out
on an automatic Flash EA 1112 series, CHN Analyzer (Thermo). The
purity of the final compounds was examined by HPLC (Shimadzu,
Japan, (on Phenomenex C8 (150 mm × 4.6 mm, 5 μm, 100 Å) double
E
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Table 3. Activities of Compounds 49−96
compd
R1
R2
R3
R4
R5
% inhibition at 500 nM
IC50 (nM)
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
methyl
ethyl
bromo
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
50.5 5.8
54.1 9.9
88.6 1.4
56.2 5.8
59.7 3.3
54.4 15.2
70.5 3.7
60.9 12.8
91.5 2.2
81.1 3.1
26.1 10.9
26.0 16.0
54.6 3.0
43.3 11.6
57.4 7.6
45.4 8.3
44.8 5.9
59.3 0.8
96.9 2.0
59.5 13.8
54.0 7.2
49.6 8.7
51.8 10.2
32.7 8.8
52.2 14.6
39.3 9.4
57.4 14.6
47.0 14.2
49.6 1.9
38.5 6.7
45.6 9.6
41.7 15.2
87.5 2.1
75.6 0.7
89.7 1.0
90.9 1.5
72.1 8.1
50.8 6.5
89.5 1.8
68.3 7.4
60.0 5.9
37.8 14.5
56.8 5.6
52.8 6.9
93.4 2.4
52.4 7.8
89.4 0.6
37.4 9.0
nd
bromo
H
nd
methyl
ethyl
H
methoxy
methoxy
chloro
chloro
bromo
bromo
33.9 3.6
H
nd
methyl
ethyl
H
nd
H
nd
methyl
ethyl
H
nd
H
nd
methyl
ethyl
H
benzyloxy
25.6 2.9
H
benzyloxy
nd
methyl
ethyl
H
trifluoro-methyl
H
nd
H
trifluoro-methyl
H
nd
methyl
ethyl
chloro
H
H
nd
chloro
H
H
nd
methyl
ethyl
chloro
H
chloro
nd
chloro
H
chloro
nd
methyl
ethyl
benzyloxy
H
H
nd
benzyloxy
H
H
nd
methyl
ethyl
methoxy
H
H
160.3 20.3
methoxy
H
H
nd
methyl
ethyl
H
H
isopropyl
nd
H
H
isopropyl
nd
methyl
ethyl
H
H
nitro
nd
H
H
nitro
nd
methyl
ethyl
H
H
H
nd
H
H
H
nd
methyl
ethyl
methyl
methyl
H
H
H
nd
H
H
nd
methyl
ethyl
methyl
methyl
H
H
nd
H
H
nd
methyl
ethyl
H
methyl
nd
H
H
methyl
nd
methyl
ethyl
H
methoxy
methoxy
methoxy
methoxy
H
hydroxy
hydroxy
methoxy
methoxy
trifluoromethyl
trifluoromethyl
fluoro
60.0 18.3
H
nd
methyl
ethyl
H
54.3 9.9
H
56.0 4.4
methyl
ethyl
H
nd
H
H
nd
methyl
ethyl
H
H
19.0 1.1
H
H
fluoro
nd
methyl
ethyl
H
chloro
chloro
hydroxy
hydroxy
H
chloro
nd
H
chloro
nd
methyl
ethyl
H
H
nd
nd
H
H
methyl
ethyl
H
N,N-dimethylamine
N,N-dimethylamine
methoxy
methoxy
37.0 3.9
nd
H
H
methyl
ethyl
H
methoxy
methoxy
methoxy
methoxy
69.9 1.2
nd
H
end-capped RP-HPLC column)) and was >95%. Compounds 2−9
were commercially available and were purchased. The experimental
details of the final, most potent analogues 18, 19, 22, 25, 28, 51, 57,
67, 81, 83, 84, 87, 93, and 95 are described below. A full description of
the synthetic protocol and spectroscopic analysis of all other
compounds can be found in the Supporting Information.
General Procedure for the Synthesis of 18, 19, 22, 25, 28,
51, 57, 67, 81, 83, 84, 87, 93, and 95. To a well-stirred solution of
F
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
8.07 (m, 4H), 10.23 (br s, 1H). 13C NMR (DMSO-d6): δc 166.3, 163,
159.8, 157.2, 148.4, 141.4, 131.1, 130.3, 129.4, 127.6, 126.8, 1227,
115.7, 115.4, 43.5, 20, 11.2. EI-MS m/z (calcd for C20H18N2O4S,
382.43); found, 381.09 (M − H)−. Anal. Calcd for C20H18N2O4S: C,
62.81; H, 4.74; N, 7.33. Found: C, 62.80; H, 4.71; N, 7.34.
Table 4. Activities of Synthesized Compounds 97−106
3-((Z)-((Z)-5-(4-Hydroxybenzylidene)-3-allyl-4-oxothiazolidin-2-
ylidene)amino)benzoic Acid (25). The compound was synthesized
according to the general procedure using (Z)-3-((3-allyl-4-oxothiazo-
lidin-2-ylidene)amino)benzoic acid (0.25 g, 0.9 mmol), piperidine
(0.023 g, 0.27 mmol), and 4-hydroxybenzaldehyde (0.122 g, 0.99
1
mmol) to afford 25 as buff-colored solid (0.23 g, 68%). H NMR
compd
R1
R
% inhibition at 500 nM
(DMSO-d6): δH 4.51 (d, 2H, J = 5.3 Hz), 5.14−5.23 (m, 2H), 5.73−
5.89 (m, 1H), 6.83 (d, 2H, J = 8.4 Hz), 7.34−7.58 (m, 3H), 7.68−8.06
(m, 4H), 10.35 (br s, 1H). 13C NMR (DMSO-d6): δc 167, 163.2,
160.3, 157.3, 148.5, 141.3, 132.3, 131.2, 130.3, 129.6, 127.7, 127, 123,
117.1, 115.8, 115.5, 42.4. EI-MS m/z (calcd for C20H16N2O4S,
380.42); found, 379.06 (M − H)−. Anal. Calcd for C20H16N2O4S: C,
63.14; H, 4.24; N, 7.36. Found: C, 63.11; H, 4.26; N, 7.34.
3-((Z)-((Z)-5-(4-Hydroxybenzylidene)-3-isopropyl-4-oxothiazoli-
din-2-ylidene)amino)benzoic Acid (28). The compound was
synthesized according to the general procedure using (Z)-3-((3-
isopropyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 0.9
mmol), piperidine (0.023 g, 0.27 mmol), and 4-hydroxybenzaldehyde
(0.122 g, 0.99 mmol) to afford 28 as pale-yellow solid (0.28 g, 82%).
1H NMR (DMSO-d6): δH 1.47 (d, 6H, J = 7.01 Hz,), 4.46 (sep, 1H, J
97
methyl
ethyl
furyl
30.1 9.4
29.2 1.9
64.4 4.5
38.7 11.4
58.4 0.0
32.5 10.1
34.4 10.3
46.0 10.9
73.5 6.5
61.9 3.1
98
furyl
99
methyl
ethyl
2-pyridyl
2-pyridyl
2-pyrrole
2-pyrrole
100
101
102
103
104
105
106
methyl
ethyl
methyl
ethyl
4-trifluoromethyl-3-pyridyl
4-trifluoromethyl-3-pyridyl
3-indolyl
methyl
ethyl
3-indolyl
the 3-((3-alkyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (1
mmol) in ethanol (6 mL) was added dropwise piperidine (0.3
mmol) and the corresponding aldehyde (1.1 mmol). The reaction
mixture was refluxed for 6 h (monitored by TLC and LCMS for
completion) and evaporated to dryness. The residue was diluted with
water (1 × 10 mL) and acidified with 2N HCl to pH 2.
Dichloromethane (1 × 20 mL) was added to the reaction mixture,
and the layers were separated. The aqueous layer was again extracted
with dichloromethane (1 × 15 mL). The combined organic layer was
washed with water (1 × 10 mL), brine (1 × 10 mL), dried over
anhydrous sodium sulfate, and evaporated to dryness. The residue was
then purified by column chromatography using hexane:ethyl acetate as
eluent to give the desired product in good yield as described below.
3-((Z)-((Z)-5-(4-Hydroxybenzylidene)-3-methyl-4-oxothiazolidin-
2ylidene)amino)benzoic Acid (18). The compound was synthesized
according to the general procedure using (Z)-3-((3-methyl-4-
oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1 mmol),
piperidine (0.026 g, 0.3 mmol), and 4-hydroxybenzaldehyde (0.134
g, 1.1 mmol) to afford 18 as pale-yellow solid (0.22 g, 63%). 1H NMR
(DMSO-d6): δH 3.33 (s, 3H), 6.82 (d, 2H, J = 8.5 Hz), 7.32−7.56 (m,
3H), 7.64−8.06 (m, 4H), 10.31 (br s, 1H). 13C NMR (DMSO-d6): δc
167.1, 163.1, 160.1, 157.3, 148.6, 141.5, 131.3, 130.3, 129.7, 127.6,
127.1, 123, 115.8, 115.6, 30.2. EI-MS m/z (calcd for C18H14N2O4S,
354.38); found, 353.05 (M − H)−. Anal. Calcd for C18H14N2O4S: C,
61.01; H, 3.98; N, 7.90. Found: C, 60.03; H, 3.98; N, 7.88.
3-((Z)-((Z)-5-(4-Hydroxybenzylidene)-3-ethyl-4-oxothiazolidin-2-
ylidene)amino)benzoic Acid (19). The compound was synthesized
according to the general procedure using (Z)-3-((3-ethyl-4-oxothia-
zolidin-2-ylidene)amino)benzoic acid (0.25 g, 0.95 mmol), piperidine
(0.024 g, 0.28 mmol), and 4-hydroxybenzaldehyde (0.127 g, 1.04
mmol) to afford 19 as off-white solid (0.25 g, 72%). 1H NMR
(DMSO-d6): δH 1.26 (t, 3H, J = 7.11 Hz), 3.83 (q, 2H, J = 7.0 Hz),
6.84 (d, 2H, J = 8.5 Hz), 7.34−7.57 (m, 3H), 7.63−8.03 (m, 4H),
10.27 (br s, 1H). 13C NMR (DMSO-d6): δc 166.5, 162.9, 160, 157.1,
148.4, 141.6, 131.2, 130.3, 129.6, 127.6, 126.9, 122.8, 115.7, 115.4,
35.5, 12.1. EI-MS m/z (calcd for C19H16N2O4S, 368.41); found,
367.27 (M − H)−. Anal. Calcd for C19H16N2O4S: C, 61.94; H, 4.38; N,
7.60. Found: C, 61.92; H, 4.34; N, 7.62.
= 7.02 Hz), 6.83 (d, 2H, J = 8.4 Hz,), 7.33−7.57 (m, 3H), 7.64−8.04
(m, 4H), 10.18 (br s, 1H). 13C NMR (DMSO-d6): δc 166.1, 162.6,
159.7, 156.9, 148.1, 141.3, 131, 130.4, 129.7, 127.5, 126.7, 122.4, 115.6,
115.3, 51.4, 20.1. EI-MS m/z (calcd for C20H18N2O4S, 382.43); found,
381.12 (M − H)−. Anal. Calcd for C20H18N2O4S: C, 62.81; H, 4.74; N,
7.33. Found: C, 62.79; H, 4.75; N, 7.33.
3-((Z)-((Z)-5-(4-Methoxybenzylidene)-3-methyl-4-oxothiazolidin-
2-ylidene)amino)benzoic Acid (51). The title compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 4-methoxybenzaldehyde
1
(0.15 g, 1.1 mmol) to afford 51 as off-white solid (0.29g, 78%). H
NMR (DMSO-d6): δH 3.34 (s, 3H), 3.81 (s, 3H), 6.96 (d, 2H, J = 8.9
Hz), 7.34−8.07 (m, 7H). 13C NMR (DMSO-d6): δc 166.6, 163.1,
160.3, 159.1, 148.2, 141.6, 131.2, 129.9, 129.6, 127.1, 126.9, 125.6,
122.9, 115.8, 113.8, 55.3, 30.2. EI-MS m/z (calcd for C19H16N2O4S,
368.41); found, 367.05 (M − H)−. Anal. Calcd for C19H16N2O4S: C,
61.94; H, 4.38; N, 7.60. Found: C, 61.95; H, 4.37; N, 7.60.
3-((Z)-((Z)-5-(4-(Benzyloxy)benzylidene)-3-methyl-4-oxothiazoli-
din-2-ylidene)amino)benzoic Acid (57). The compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 4-(benzyloxy)-
benzaldehyde (0.23 g, 1.1 mmol) to afford 57 as white solid (0.38
1
g, 86%). H NMR (DMSO-d6): δH 3.32 (s, 3H), 5.21 (s, 2H), 7.03−
7.42 (m, 6H), 7.53−8.06 (m, 8H). 13C NMR (DMSO-d6): δc 167,
163.2, 160.2, 157.8, 148.4, 141.6, 136.1, 131.1, 129.8, 129.5, 128.4,
127.1, 126.8, 126.4, 125.6, 122.7, 115.7, 113.6, 70.2, 30.4. EI-MS m/z
(calcd for C25H20N2O4S, 444.50); found, 443.12 (M − H)−. Anal.
Calcd for C25H20N2O4S: C, 67.55; H, 4.54; N, 6.30. Found: C, 67.52;
H, 4.55; N, 6.29.
3-((Z)-((Z)-5-(2-Methoxybenzylidene)-3-methyl-4-oxothiazolidin-
2-ylidene)amino)benzoic Acid (67). The compound was synthesized
according to the general procedure using (Z)-3-((3-methyl-4-
oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1 mmol),
piperidine (0.026 g, 0.3 mmol), and 2-methoxybenzaldehyde (0.15
1
g, 1.1 mmol) to afford 67 as off-white solid (0.29 g, 78%). H NMR
3-((Z)-((Z)-5-(4-Hydroxybenzylidene)-3-propyl-4-oxothiazolidin-
2ylidene)amino)benzoic Acid (22). The compound was synthesized
according to the general procedure using (Z)-3-((3-propyl-4-
oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 0.9 mmol),
piperidine (0.023 g, 0.27 mmol), and 4-hydroxybenzaldehyde (0.121
(DMSO-d6) δH 3.35 (s, 3H), 3.83 (s, 3H), 6.98−7.03 (m, 2H), 7.23−
7.31 (m, 2H), 7.64−8.08 (m, 5H). 13C NMR (DMSO-d6): δc 167.1,
163.4, 160.1, 158.8, 148.3, 141.5, 131.1, 129.4, 128.5, 127.7, 126.8,
125.8, 122.8, 120.2, 115.7, 114.3, 113.6, 55.9, 30.5. EI-MS m/z (calcd
for C19H16N2O4S, 368.41); found, 367.05 (M − H)−. Anal. Calcd for
C19H16N2O4S: C, 61.94; H, 4.38; N, 7.60. Found: C, 61.97; H, 4.38;
N, 7.62.
1
g, 0.99 mmol) to afford 22 as pale-yellow solid (0.26 g, 76%). H
NMR (DMSO-d6): δH 1.01 (t, 3H, J = 7.2 Hz), 1.76 (m, 2H), 3.89 (t,
1H, J = 7.5 Hz), 6.86 (d, 2H, J = 8.3 Hz), 7.36−7.56 (m, 3H), 7.65−
G
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
1
3-((Z)-((Z)-5-(4-Hydroxy-3-methoxybenzylidene)-3-methyl-4-oxo-
thiazolidin-2-ylidene)amino)benzoic Acid (81). The compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 4-hydroxy-3-methox-
ybenzaldehyde (0.17 g, 1.1 mmol) to afford 81 as pale-yellow solid
84%). H NMR (DMSO-d6): δH 3.37 (s, 3H), 3.83 (s, 3H), 3.87 (s,
3H), 6.81 (s, 2H), 7.32−8.13 (m, 5H). 13C NMR (DMSO-d6): δc
166.9, 163.4, 160.3, 152.7, 148.6, 141.4, 138, 131.2, 129.6, 129, 127.1,
125.5, 123, 115.6, 103.4, 60.5, 55.8, 30.6. EI-MS m/z (calcd for
C21H21N3O3S, 428.46); found, 427.11 (M − H)−. Anal. Calcd for
C21H21N3O3S: C, C, 58.87; H, 4.70; N, 6.54. Found: C, 58.85; H, 4.68;
N, 6.52.
1
(0.32 g, 84%). H NMR (DMSO-d6) δH 3.32 (s, 3H), 3.86 (s, 3H),
6.99−7.16 (m, 3H), 7.32−8.06 (m, 5H), 9.94 (br s, 1H). 13C NMR
(DMSO-d6): δc 166.8, 163.2, 160, 148.8, 148.6, 147.5, 141.6, 131.1,
129.6, 128.4, 127.1, 125.6, 123, 122.5, 116.4, 115.7, 111.5, 55.8, 30.4.
EI-MS m/z (calcd for C19H16N2O5S, 384.41); found, 383.05 (M −
H)−. Anal. Calcd for C19H16N2O5S: C, 59.37; H, 4.20; N, 7.29. Found:
C, 59.35; H, 4.19; N, 7.29.
CysK1 Hit Identification and Validation. Recombinant CysK1
was produced and purified as described previously.12 Compounds were
screened against M. tuberculosis CysK1 using an assay based on the
spectrophotometric determination of the reaction product L-cysteine
by ninhydrine30 adapted to 96-well plate format described
previously.12 Each enzyme reaction was carried out at room
temperature in a total volume of 50 μL containing 100 mM MOPS
(pH 7.0), 2 mM O-acetyl-L-serine, 1 mM sodium sulfide, 5 pmol
CysK1, and 2 mM, 500 μM, or 100 μM of the tested compounds. One
μL of compound solution was transferred to all assay plates before 44
μL of reaction mix was added. The reactions were started by addition
of 5 μL of sodium sulfide solution. After 20 min of incubation at 22
°C, the reactions were stopped by addition of 10 μL of trichloroacetic
acid (final concentration, 16.6%). Subsequently, 60 μL of acid−
ninhydrin reagent was added and the plates were heated at 96 °C for
10 min. Reactions were cooled and transferred into new 96-well plates,
diluted 1:1 in ethanol (95%), and the absorbance of each sample was
measured at 560 nm in a BioTek Synergy HT-1 plate reader. Each
reaction was carried out in triplicate. Positive controls (100%
inhibition) contained the assay mix with the inhibitory DFSI
tetrapeptide (final concentration, 100 μM), negative controls (0%
inhibition) did not contain any inhibitory compounds.
3-((Z)-((Z)-5-(3,4-Dimethoxybenzylidene)-3-methyl-4-oxothiazo-
lidin-2-ylidene)amino)benzoic Acid (83). The compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 3,4-dimethoxybenzalde-
hyde (0.18 g, 1.1 mmol) to afford 83 as light-brown solid (0.31 g,
78%). 1H NMR (DMSO-d6) δH 3.31 (s, 3H), 3.85 (s, 6H), 6.96−7.31
(m, 4H), 7.66−8.09 (m, 4H). 13C NMR (DMSO-d6): δc 166.7, 163.2,
160.1, 149.4, 148.6, 148.4, 141.6, 131.1, 129.7, 128.1, 127, 125.6, 122.9,
122.1, 115.8, 111.3, 110.9, 55.7, 30.3. EI-MS m/z (calcd for
C20H18N2O5S, 398.43); found, 397.06 (M − H)−. Anal. Calcd for
C20H18N2O5S: C, 60.29; H, 4.55; N, 7.03. Found: C, 60.31 H, 4.54; N,
7.02.
3-((Z)-((Z)-5-(3,4-Dimethoxybenzylidene)-3-ethyl-4-oxothiazoli-
din-2-ylidene)amino)benzoic Acid (84).). The compound was
synthesized according to the general procedure using (Z)-3-((3-
ethyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 0.95
mmol), piperidine (0.024 g, 0.28 mmol), and 3,4-dimethoxybenzalde-
hyde (0.173 g, 1.04 mmol) to afford 84 as light-brown solid. 1H NMR
(DMSO-d6) δH 1.24 (t, 3H, J = 7.05 Hz), 3.78 (q, 2H, J = 6.94 Hz),
3.84 (s, 3H), 3.86 (s, 3H), 6.94−7.31 (m, 3H), 7.61−8.05 (m, 5H).
13C NMR (DMSO-d6): δc 166.2, 163.1, 159.8, 149.3, 148.5, 148.4,
Compounds were designated as initial hits if the inhibition was
>95%, and the IC50 values for these were derived from dose−response
curves. Serial 1:3 compound dilutions were first prepared in 50%
dimethyl sulfoxide in 100 μL (final dimethyl sulfoxide concentration in
assay: 5%) and transferred into a 96-deep well block. Then 890 μL of
reaction mix was added (final concentrations: 100 mM MOPS [pH
7.0], 2 mM O-acetyl-L-serine, 1.25 μg CysK1) and the reactions started
by addition of 10 μL of sodium sulfide (final concentration, 1 mM).
The KM for O-acetyl-L-serine is 5 mM and for Na2S < 300 μM, hence
the chosen substrate concentrations in the assay (2 mM OAS and 1
mM Na2S) will favor identification of competitive inhibitors.31 This
approach was validated by a Michaelis−Menten analysis of lead
compound 2, which indeed shows a competitive inhibition pattern
(Supporting Information Figure S3). The Ki value of 111.2 nM for
compound 2 also agrees well with the IC50 value of 103.8 nM. At
various time intervals (1, 2, 4, 6, 8, 10, 15, and 20 min), 100 μL
aliquots were removed and assayed for L-cysteine formation as
described above. This protocol ensures that the initial velocities are
derived from the linear phase of the reaction. The initial rates were
normalized to the initial rate for the CysK1-catalyzed reaction in the
absence of inhibitors. The PI (percent inhibition) was calculated as
100% × (1 − normalized activity). The derived PI values were plotted
against log[inhibitor concentration]. To determine the IC50 values, a
four-parameter nonlinear regression fit was performed with the
program Origin (OriginLab) using equation: PI = Bottom + ((Top
141.7, 131, 129.6, 127.9, 126.8, 125.5, 123, 122, 115.9, 111.3, 110.8,
55.6, 35.4, 12.2. EI-MS m/z (calcd for C21H20N2O5S, 412.46); found,
411.11 (M − H)−. Anal. Calcd for C21H20N2O5S: C, 61.15; H, 4.89; N,
6.79. Found: C, 61.17; H, 4.87; N, 6.81.
3-((Z)-((Z)-5-(4-Fluorobenzylidene)-3-methyl-4-oxothiazolidin-2-
ylidene)amino)benzoic Acid (87). The compound was synthesized
according to the general procedure using (Z)-3-((3-methyl-4-
oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1 mmol),
piperidine (0.026 g, 0.3 mmol), and 4-fluorobenzaldehyde (0.136 g,
1
1.1 mmol) to afford 87 as pale-yellow solid. (0.33 g, 81%). H NMR
(DMSO-d6): δH 3.31 (s, 3H), 7.20−7.34 (m, 3H), 7.68 −8.08 (m,
6H). 13C NMR (DMSO-d6): δc 166.8, 163, 161.7, 160.1, 148.3, 141.4,
131.1, 130.5, 130.1, 129.6, 127.1, 125.6, 123, 116.8, 115, 30.3. EI-MS
m/z (calcd for C18H13FN2O3S, 356.37); found, 355.05 (M − H)−.
Anal. Calcd for C18H13FN2O3S: C, 60.67; H, 3.68; N, 7.86. Found: C,
60.63; H, 3.70; N, 7.84.
3-((Z)-((Z)-5-(4-(Dimethylamino)benzylidene)-3-methyl-4-oxo-
thiazolidin-2-ylidene)amino)benzoic Acid (93). The compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 4-(dimethylamino)-
benzaldehyde (0.164 g, 1.1 mmol) to afford 93 as brownish-red solid
− Bottom)/(1 + 10((pXC −log[hit compound])×p)), where bottom =
50
minimal % inhibition, top = maximal % inhibition, p = Hill coefficient,
and pXC50 = −log(IC50). As a positive control, the IC50 value for the
inhibitory peptide DFSI was determined under identical conditions.
The resulting IC50 for this peptide, 2.9 μM, compares well with the Ki
of 5 μM determined previously.12
1
(0.22 g, 58%). H NMR (DMSO-d6): δH 3.03 (s, 3H), 3.37 (s, 3H),
6.74 (d, 2H, J = 8.3 Hz), 7.34−7.47 (m, 3H), 7.66 −8.12 (m, 4H). 13C
NMR (DMSO-d6): δc 166.6, 163.1, 160.2, 149.9, 148.5, 141.4, 131,
129.5, 129.3, 127.1, 125.6, 124.4, 122.9, 115.7, 111.4, 39.9, 30.3. EI-MS
m/z (calcd for C20H19N3O3S, 381.45); found, 380.1 (M − H)−. Anal.
Calcd for C20H19N3O3S: C, 62.97; H, 5.02; N, 11.02. Found: C, 62.95;
H, 5.04; N, 11.03.
ASSOCIATED CONTENT
■
S
* Supporting Information
3-((Z)-((Z)-5-(3,4,5-Trimethoxybenzylidene)-3-methyl-4-oxothia-
zolidin-2-ylidene)amino)benzoic Acid (95). The compound was
synthesized according to the general procedure using (Z)-3-((3-
methyl-4-oxothiazolidin-2-ylidene)amino)benzoic acid (0.25 g, 1
mmol), piperidine (0.026 g, 0.3 mmol), and 3,4,5-trimethoxybenzal-
dehyde (0.22 g, 1.1 mmol) to afford 95 as off-white solid (0.36 g,
Additional experimental details concerning e-pharmacophore
modeling, library synthesis, docking, characterization of all
synthesized compounds, spectroscopic data, and structural
studies. This material is available free of charge via the Internet
at http://pubs.acs.org.
H
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
alpha-aminoacrylate intermediate and an enzyme−inhibitor complex. J.
Biol. Chem. 2007, 282, 23473−23481.
(13) Kumar, V. U. J.; Poyraz, O.; Saxena, S.; Schnell, R.; Yogeeswari,
P.; Schneider, G.; Sriram, D. Discovery of novel inhibitors targeting
the Mycobacterium tuberculosis O-acetylserine sulfhydrylase (CysK1)
using virtual high-throughput screening. Bioorg. Med. Chem. Lett. 2013,
23, 1182−1186.
(14) Friesner, R. A.; Murphy, R. B.; Repasky, M. P.; Frye, L. L.;
Greenwood, J. R.; Halgren, T. A.; Sanschagrin, P. C.; Mainz, D. T.
Extra Precision Glide: docking and scoring incorporating a model of
hydrophobic enclosure for protein−ligand complexes. J. Med. Chem.
2006, 49, 6177−6196.
Accession Codes
The PDB code for the structure of the complex of M.
tuberculosis CysK1 with 2 is 3ZEI.
AUTHOR INFORMATION
Corresponding Author
*For D.S.: E-mail, drdsriram@yahoo.com. For G.S.: phone,
■
+4652487675; fax, +468327626; E-mail, Gunter.Schneider@ki.
se.
Author Contributions
∥
̈
O.P. and V.U.J. contributed equally to this work.
(15) Friesner, R. A.; Banks, J. L.; Murphy, R. B.; Halgren, T. A.;
Klicic, J. J.; Mainz, D. T.; Repasky, M. P.; Knoll, E. H.; Shelley, M.;
Perry, J. K.; Shaw, D. E.; Francis, P.; Shenkin, P. S. Glide: a new
approach for rapid, accurate docking and scoring. 1. Method and
assessment of docking accuracy. J. Med. Chem. 2004, 47, 1739−1749.
(16) Eldridge, M. D.; Murray, C. W.; Auton, T. R.; Paolini, G. V.;
Mee, R. P. Empirical scoring functions: I. The development of a fast
empirical scoring function to estimate the binding affinity of ligands in
receptor complexes. J. Comput.-Aided. Mol. Des. 1997, 11, 425−445.
(17) Dayam, R.; Aiello, F.; Deng, J.; Wu, Y.; Garofalo, A.; Chen, X.;
Neamati, N. Discovery of Small Molecule Integrin αVβ3 Antagonists
as Novel Anticancer Agents. J. Med. Chem. 2006, 49, 4526−4534.
(18) Vicini, P.; Geronikaki, A.; Incerti, M.; Zani, F.; Dearden, J.;
Hewittc, M. 2-Heteroarylimino-5-benzylidene-4-thiazolidinones ana-
logues of 2-thiazolylimino-5-benzylidene-4-thiazolidinones with anti-
microbial activity: synthesis and structure−activity relationship. Bioorg.
Med. Chem. 2008, 16, 3714−3724.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge access to synchrotron radiation at
the ESRF, France, BESSY, Germany, and MAX IV laboratory,
Sweden. This work was supported by the Swedish Govern-
mental Agency for Innovation Systems (VINNOVA) and the
Department of Biotechnology (DBT), India.
ABBREVIATIONS USED
CysK1, O-acetylserine sulfhydrylase; rt, room temperature
■
REFERENCES
■
(1) Dye, C.; Williams, B. G. The Population Dynamics and Control
of Tuberculosis. Science 2010, 328, 856−861.
(19) Verma, A.; Saraf, K. S. 4-Thiazolidinone A biologically active
scaffold. Eur. J. Med. Chem. 2008, 43, 897−905.
(2) Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. The
challenge of new drug discovery for tuberculosis. Nature 2011, 469,
483−490.
(20) Jain, K. A.; Vaidya, A.; Ravichandran, V.; Kashaw, S. K.; Agrawal,
K. R. Recent developments and biological activities of thiazolidinone
derivatives: A review. Bioorg. Med. Chem. 2012, 20, 3378−3395.
(21) Maccari, R.; Ottana, R.; Curinga, C.; Vigorita, M. G.; Rakowitz,
D.; Steindlb, T.; Langer, T. Structure−activity relationships and
molecular modelling of 5-arylidene-2,4-thiazolidinediones active as
aldose reductase inhibitors. Bioorg. Med. Chem. 2005, 13, 2809−2823.
(22) Kosma, P.; Selzer, E.; Mereiter, K. 4-({(Z)-5-[(Z)-3-Ethoxy-4-
hydroxybenzylidene]-3-methyl-4-oxo-1,3-thiazolidin-2-ylidene}-
amino)benzoic acid dimethylformamide monosolvate. Acta Crystal-
logr., Sect. E: Struct. Rep. Online 2012, E68, o3417−o3418.
(23) Huang, B.; Vetting, M. W.; Roderick, S. L. The active site of O-
acetylserine sulfhydrylase is the anchor point for bienzyme complex
formation with serine acetyltransferase. J. Bacteriol. 2005, 187, 3201−
3205.
(3) Gengenbacher, M.; Kaufmann, S. H. E. Mycobacterium tuber-
culosis: success through dormancy. FEMS Microbiol. Rev. 2012, 36,
514−532.
(4) Ehrt, S.; Schnappinger, D. Mycobacterial survival strategies in the
phagosome: defence against host stresses. Cell. Microbiol. 2009, 11,
1170−1178.
(5) Jothivasan, V. K.; Hamilton, C. J. Mycothiol: synthesis,
biosynthesis and biological functions of the major low molecular
molecular weight thiol in actinomycetes. Nat. Prod. Rep. 2008, 25,
1091−1117.
(6) Hampshire, T.; Soneji, S.; Bacon, J.; James, B. W.; Hinds, J.;
Laing, K.; Stabler, R. A.; Marsh, P. D.; Butcher, P. D. Stationary phase
gene expression of Mycobacterium tuberculosis following a progressive
nutrient depletion: a model for persistent organisms. Tuberculosis
2004, 84, 228−238.
(7) Betts, J. C.; Lukey, P. T.; Robb, L. C.; McAdam, R. A.; Duncan,
K. Evaluation of a nutrient starvation model of Mycobacterium
tuberculosis persistence by gene and protein expression profiling.
Mol. Microbiol. 2002, 43, 717−731.
(8) Voskuil, M. I.; Visconti, K. C.; Schoolnik, G. K. Mycobacterium
tuberculosis gene expression during adaptation to stationary phase and
low-oxygen dormancy. Tuberculosis 2004, 84, 218−227.
(9) Schnappinger, D.; Ehrt, S.; Voskuil, M. I.; Liu, Y.; Mangan, J. A.;
Monahan, I. M.; Dolganov, G.; Efron, B.; Butcher, P. D.; Nathan, C.;
Schoolnik, G. K. Transcriptional adaptation of Mycobacterium
tuberculosis within macrophages: Insights into the phagosomal
environment. J. Exp. Med. 2003, 198, 693−704.
(10) Bhave, D. P.; Muse, W. B., III; Carroll, K. S. Drug targets in
mycobacterial sulfur metabolism. Infect. Disord. Drug Targets 2007, 7,
140−158.
(11) Hatzios, S. K.; Bertozzi, C. R. The Regulation of Sulfur
Metabolism in Mycobacterium tuberculosis. PLoS Pathogens 2011, 7,
e1002036 DOI: 10.1371/journal.ppat.1002036.
(12) Schnell, R.; Oehlmann, W.; Singh, M.; Schneider, G. Structural
insights into catalysis and inhibition of O-acetylserine sulfhydrylase
from Mycobacterium tuberculosiscrystal structures of the enzyme
(24) Francois, J. A.; Kumaran, S.; Jez, J. M. Structural basis for
interaction of O-acetylserine sulfhydrylase and serine acetyltransferase
in the Arabidopsis cysteine synthase complex. Plant Cell 2006, 18,
3647−3655.
(25) Salsi, E.; Bayden, A. S.; Spyrakis, F.; Amadasi, A.; Campanini, B.;
Bettati, S.; Dodatko, T.; Cozzini, P.; Kellogg, G. E.; Cook, P. F.;
Roderick, S. L.; Mozzarelli, A. Design of O-acetylserine sulfhydrylase
inhibitors by mimicking nature. J. Med. Chem. 2010, 53, 345−356.
(26) Campanini, B.; Speroni, F.; Salsi, E.; Cook, P. F.; Roderick, S. L.;
Huang, B.; Bettati, S.; Mozzarelli, A. Interaction of serine
acetyltransferase with O-acetylserine sulfhydrylase active site: evidence
from fluorescence spectroscopy. Protein Sci. 2005, 14, 2115−2124.
(27) Spyrakis, F.; Felici, P.; Bayden, A. S.; Salsi, E.; Miggiano, R.;
Kellogg, G. E.; Cozzini, P.; Cook, P. F.; Mozzarelli, A.; Campanini, B.
Fine tuning of the active site modulates specificity in the interaction of
O-acetylserine sulfhydrylase isozymes with serine acetyltransferase.
Biochim. Biophys. Acta 2013, 1834, 169−181.
(28) Nagpal, I.; Raj, I.; Subbarao, N.; Gourinath, S. Virtual screening,
identification and in vitro testing of novel inhibitors of O-acetyl-L-
serine sulfhydrylase of Entamoeba histolytica. PLoS One 2012, 7,
e30305 DOI: 10.1371/journal.pone.0030305.
(29) Amori, L.; Katkevica, S.; Bruno, A.; Campanini, B.; Fekici, P.;
Moxxareilli, A.; Costantino, G. Design and synthesis of trans-2-
I
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
substituted-cyclopropane-1-carboxylic acids as the first non-natural
small molecule inhibitors of O-acetylserine sulfhydrylase. Med. Chem.
Commun. 2012, 3, 1111−1116.
(30) Gaitonde, M. K. A spectrophotometric method for the direct
determination of cysteine in the presence of other naturally occurring
amino acids. Biochem. J. 1967, 104, 627−633.
(31) Wu, G.; Yuan, Y.; Hodge, C. N. Determining appropriate
substrate conversion for enzymatic assays in high-throughput screen-
ing. J. Biomol. Screening 2003, 8, 694−700.
J
dx.doi.org/10.1021/jm400710k | J. Med. Chem. XXXX, XXX, XXX−XXX