Inhibitors of the cysteine synthase CysM with antibacterial potency against dormant mycobacterium tuberculosis

Article abstract of DOI:10.1021/acs.jmedchem.6b00674

Cysteine is an important amino acid in the redox defense of Mycobacterium tuberculosis, primarily as a building block of mycothiol. Genetic studies have implicated de novo cysteine biosynthesis in pathogen survival in infected macrophages, in particular f

Full text of DOI:10.1021/acs.jmedchem.6b00674

Subscriber access provided by La Trobe University Library
Article
Inhibitors of the cysteine synthase CysM with antibacterial
potency against dormant Mycobacterium tuberculosis
Katharina Brunner, Selma Maric, Rudraraju Srilakshmi Reshma, Helena Almqvist, Brinton
A. Seashore-Ludlow, Anna-Lena Gustavsson, Ömer Poyraz, Perumal Yogeeswari, Thomas
Lundbäck, Michaela Vallin, Dharmarajan Sriram, Robert Schnell, and Gunter Schneider
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00674 • Publication Date (Web): 05 Jul 2016
Downloaded from http://pubs.acs.org on July 6, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
Inhibitors of the cysteine synthase CysM with antibacterial potency against
8
dormant Mycobacterium tuberculosis
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Katharina Brunnera#, Selma Marica#, Rudraraju Srilakshmi Reshmab, Helena Almqvistc, Brinton
SeashoreꢀLudlowc, AnnaꢀLena Gustavssonc, Ömer Poyraza, Perumal Yogeeswarib, Thomas
Lundbäckc, Michaela Vallinc, Dharmarajan Sriramb, Robert Schnella
and Gunter Schneidera
a Division of Molecular Structural Biology, Department of Medical Biochemistry & Biophysics,
Karolinska Institutet, Sꢀ171 77 Stockholm, Sweden
b Drug Discovery Research Laboratory, Department of Pharmacy, Birla Institute of Technology
& ScienceꢀPilani, Hyderabad Campus, Shameerpet, R.R. District, Hyderabadꢀ500078, Andhra
Pradesh, India
c Chemical Biology Consortium Sweden, Science for Life Laboratory, Division of Translational
Medicine & Chemical Biology, Department of Medical Biochemistry & Biophysics, Karolinska
Institutet, Sꢀ171 65 Solna, Sweden
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 45
1
2
3
4
5
6
ABSTRACT
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Cysteine is an important amino acid in the redox defense of Mycobacterium tuberculosis,
primarily as a building block of mycothiol. Genetic studies have implicated de novo cysteine
biosynthesis in pathogen survival in infected macrophages, in particular for persistent M.
tuberculosis. Here, we report on the identification and characterization of potent inhibitors of
CysM, a critical enzyme in cysteine biosynthesis during dormancy. A screening campaign of
17312 compounds identified ligands that bind to the active site with µM affinity. These were
characterized in terms of their inhibitory potencies and structureꢀactivity relationships through
hit expansion guided by threeꢀdimensional structures of enzymeꢀinhibitor complexes. The top
compound binds to CysM with 300 nM affinity and displays selectivity over the mycobacterial
homologues CysK1 and CysK2. Notably, two inhibitors show significant potency in a nutrientꢀ
starvation model of dormancy of Mycobacterium tuberculosis, with little or no cytotoxicity
towards mammalian cells.
ACS Paragon Plus Environment

Page 3 of 45
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
Nearly oneꢀthird of the global population is infected with Mycobacterium tuberculosis, the
causative agent of tuberculosis (http://www.who.int/tb/publications/global_report/en/). In most
infected individuals, the bacilli reside in macrophages where they evade the immune defense
mechanisms of the host by adopting a dormant lifestyle.1 In this nonꢀcontagious phase of the
disease, the pathogen is contained in granuloma formed by macrophages, lymphocytes and
fibroblasts within the lung tissue.2 The large number of infected individuals represents a
reservoir of bacilli that upon relapse into the acute state of tuberculosis can further spread the
disease. Since dormant bacteria are less susceptible to standard antibiotic treatment new antiꢀ
mycobacterial agents are needed that are effective against persistent M. tuberculosis.3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Host macrophages use reactive oxygen (ROI) and nitrogen (RNI) intermediates to kill the
engulfed bacteria and redox defense mechanisms are therefore crucial for intracellular survival of
the pathogen.4 In mycobacteria, intracellular redox homeostasis is largely maintained by
mycothiol, the functional analog of glutathione.5,6 Mycothiol contains a cysteineꢀderived
sulfhydryl group and its biosynthesis is dependent on the availability of Lꢀcysteine.7 De novo
cysteine biosynthesis relies on the provision of reduced sulfur by the sulfate reduction pathway
(APS/PAPS pathway). It is therefore not surprising that metabolic pathways related to sulfur
assimilation and cysteine biosynthesis are upregulated in several different dormancy models.8,9
Enzymes from these pathways have therefore been suggested as potential targets for new antiꢀ
mycobacterial compounds10,11 and first potent inhibitors of adenosineꢀ5’ꢀphosphosulfate
reductase with bactericidal activity have been described.12
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 45
1
2
3
4
5
6
7
8
To date three pyridoxalꢀphosphate (PLP) dependent cysteine synthases have been characterized
in M. tuberculosis (Figure 1).11 CysK1, a bona fide Oꢀacetylserine sulfhydrylase (OASS), uses
Oꢀacetylserine and hydrogenꢀsulfide from the APS/PAPS pathway for the synthesis of
cysteine.13,14 CysM on the other hand utilizes Oꢀphosphoserine (OPS) together with a small
sulfur delivery protein CysO, which contains a thiocarboxylated carboxy terminus.15–17 A third
putative cysteine synthase, CysK2, also utilizes OPS as acceptor substrate and is able to produce
cysteine with hydrogenꢀsulfide as sulfur donor. However, this enzyme shows a preference for
thiosulfate, leading to Sꢀsulfocysteine as the primary product and is thought to be involved in
redoxꢀsignaling.18 CysM, cysO and cysK2 have all been shown to be upꢀregulated under
conditions simulating dormancy8,9,19,20 and randomized transposon mutagenesis studies
demonstrated profound phenotypic effects for cysM and cysO by gene disruption.21 These studies
showed attenuation of cysO and cysM mutants of M. tuberculosis in in vitro models of dormancy
and in a mouse model of infection22, indicating the CysM/CysO pathway as the dominant route
to cysteine in dormant bacteria. These findings are consistent with the observation that CysM
protects the oxidation sensitive αꢀaminoacrylate intermediate from redox stress induced by small
molecules such as hydrogen peroxide to a larger extent than CysK1.23 Inhibitors of CysM are
therefore expected to weaken the redox defense lines of M. tuberculosis and potentially allow
natural antibacterial actions of the host macrophages to eliminate the internalized mycobacteria.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Here we describe the identification and characterization of small molecules targeting M.
tuberculosis CysM. Active site binders were identified in a smallꢀmolecule screening campaign
based on ligandꢀinduced changes in fluorescence of the PLP cofactor, previously observed in
ACS Paragon Plus Environment

Page 5 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
other members of this enzyme family.24ꢀ26 The inhibitory effects of hit compounds were tested in
followꢀup enzyme activity assays, which were also employed for examination of structureꢀ
activity relationships following synthesis of closely related analogs. The binding modes of seven
of the strongest inhibitors were furthermore determined through highꢀresolution crystal structures
of proteinꢀligand complexes. The ligands were shown to bind to the CysM active site nearly
parallel with the PLPꢀcofactor, forming extensive interactions with the target both at the
substrate carboxylate binding pocket and at an adjacent cavity. This cavity is flexible such that it
adapts to variously sized substituents. The best inhibitors showed targetꢀspecific affinity in the
0.3ꢀ3 µM range and bactericidal activity in a nutrientꢀstarvation model of dormancy of
Mycobacterium tuberculosis. Prevention of the bactericidal effect of the inhibitors by the
addition of cysteine points towards perturbation of cysteine biosynthesis as a significant
component of their mode of action.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
RESULTS AND DISCUSSION
Small–molecule screening and hit selection: Binding of smallꢀmolecule ligands to the active
site of CysM from M. tuberculosis is accompanied by an increase in the fluorescence of the
covalently bound cofactor PLP (Figures 2a & b). We exploited this property in a 384ꢀwell
microplateꢀbased screening assay to identify new smallꢀmolecule binders of this enzyme. To this
end 17312 chemically diverse and leadꢀ to drugꢀlike compounds27 were screened against
recombinant CysM at a single concentration of 10 ꢀM at Chemical Biology Consortium Sweden
(www.cbcs.se). Compounds that resulted in an arbitrarily chosen fluorescence increase by more
than 30% were considered as primary hits that warranted further investigation at multiple
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 45
1
2
3
4
5
6
7
8
9
concentrations and counterꢀassays to remove autoꢀfluorescent compounds. This selection
criterion gave 431 initial hits for followꢀup experiments (Figure 2c, Supporting Figure S1).
Counterscreening at three compound concentrations in the absence of CysM effectively removed
a significant portion of these compounds as autoꢀfluorescent, leaving 20 compounds showing
concentrationꢀdependent binding to CysM with minimal interference from background
fluorescence of the compound itself (Supporting Figure S2a). This corresponds to a hit rate of
0.1% in our library. Seven of these confirmed hits demonstrated a saturating CysM dependent
fluorescence signal in the interval between 2.5 to 40 µM. After examination of both panꢀassay
interference compounds (PAINS)28; removal of unwanted reactive groups, (REOS filter)29 and
potential aggregator filters30 (Supporting Figure S2b and Supporting Table 1) one hit compound,
2‐(4‐phenylbenzoyl)benzoyl chloride (CBK241255) was filtered out. All other compounds
passed the filters and were regarded as having low to medium risk of being aggregators
(Supporting table 1). Among these six remaining compound structures, two were closely related
urea derivatives of which none had previously appeared as hits in other inꢀhouse screens. As
further analogs of this scaffold were easily accessible to us through analog searches of the CBCS
compound collection and from synthesis, we chose to focus our next efforts on this compound
series.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The changes in PLP fluorescence were next used to assess the apparent affinity of the two initial
urea hits in full concentrationꢀresponse curves (Figure 3a and Supporting Figure S3), giving
apparent Kd values of 0.95±0.09 ꢀM for 3ꢀ(3ꢀ(3,4ꢀdichlorophenyl)ureido)benzoic acid 1 and
1.7±0.2 µM for 3ꢀ(3ꢀ([1,1'ꢀbiphenyl]ꢀ3ꢀyl)ureido) benzoic acid 2 (Table 1), respectively.
Although the fluorescenceꢀbased assay was robust enough to support the screening campaign and
ACS Paragon Plus Environment

Page 7 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
hit confirmation, it is associated with significant limitations when it comes to the characterization
of potent binders. This is because a CysM concentration of 2 ꢁM is required to obtain a reliable
signal to background ratio in the assay, such that stoichiometry reasons require the addition of
corresponding concentrations of ligand to saturate all protein in the solution. Consequently,
affinities approaching the subꢀµM range cannot be accurately measured with this assay, and
measurements result in underestimations of their true affinities. For this reason, compounds with
potentially underestimated affinities (those with low µM apparent affinities in Figure 3d and
Supporting Figure S3) were reꢀevaluated whenever possible by isothermal titration calorimetry
(ITC), which is not affected by the same stoichiometric limitations. For compound 1, ITC gave a
Kd value of 0.32± 0.01 ꢀM (Table 1), whereas binding of compound 2 could not be evaluated by
ITC. We therefore regard the value of 1.7±0.2 µM from the PLP fluorescence experiment as an
upper apparent Kd for this compound.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Hit characterization and expansion: Hit expansion was carried out by several rounds of
compound selection/synthesis, where threeꢀdimensional structures of CysM−inhibitor complexes
(see below) in combination with biochemical data informed the design of the analogs. The
identified hits had urea as a linker bridging two flanking phenyl rings (Table 1). Based on this
scaffold, 17 analogs were selected from the inꢀhouse compound collection and small subꢀ
libraries comprising an additional 54 molecules were synthesized. The synthetic pathways for the
synthesis of these urea derivatives are delineated in Scheme 1. A total of 71 analogues were thus
characterized to derive an initial understanding of the structureꢀactivity relationships (SAR).
Data for the most potent inhibitors are included in Table 1, while Supplementary Table S2 gives
a summary of the activity data for all tested molecules. Collectively these span an affinity range
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 45
1
2
3
4
5
6
7
8
9
from 300 nM to completely inactive compounds. We next employed an orthogonal assay to
evaluate inhibition of the CysM enzymatic activity by monitoring the αꢀaminoacrylate
intermediate formed in the first halfꢀreaction (Supporting Figure S4). A significant reduction of
the apparent first order rate constants (kobs) was observed for compounds 1 (Figure 3) and 2
(Table 1), while the corresponding data for other representatives mirrored the apparent affinities
from the PLPꢀfluorescence assay (Table 1 and Supporting Table S3).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The second halfꢀreaction of the catalytic cycle of CysM (Supporting Figure S4) uses
thiocarboxylated CysO (CysOꢀSH) as sulfur donor, leading to the formation of a covalent CysOꢀ
cysteine adduct which can be detected by mass spectrometry.15,16 In order to assess the ability of
the best inhibitors of the first half reaction to completely inhibit CysM (i.e. no formation of the
covalent CysOꢀcysteine adduct) the reactions were carried out in the absence and presence of hit
compounds and analyzed by ESIꢀMS (Figure 3d, Supporting Figure S5). The only peaks detected
in the presence of compounds 1 and 2 were at ~9571 Da corresponding to the mass of
thiocarboxylated CysO (CysOꢀSH) showing that these compounds completely inhibit CysOꢀ
cysteine adduct (expected molecular weight of 9658 Da) formation, i.e. the overall reaction of
the enzyme. Compounds 3, 5–7 showed peaks at both 9571 Da and 9658 Da in different ratios
(Supporting Figure S5) indicating partial inhibition, while 17 showed a main peak at 9658 Da
demonstrating no or very limited inhibition of CysOꢀcysteine adduct formation in the CysM
catalyzed reaction compared to the top compounds. For the tightest binding compound 1 we also
determined the IC50 value for the overall reaction (Figure 3e), which with 0.48 ꢀM agrees well
with the Kd of 0.32 ꢀM determined using ITC.
ACS Paragon Plus Environment

Page 9 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Overall three-dimensional structures of the enzyme-ligand complexes. The crystal structures
of the complexes of CysM with seven inhibitors, 1-7, were determined by xꢀray crystallography
to 1.64 – 2.7 Å resolution (Supporting Figure S6 and tables S3 and S4). Overall the threeꢀ
dimensional structures of the enzyme complexes are similar to the structure of the unliganded
enzyme with r.m.s.d. values for the Cα atoms in the range of 0.7 ꢀ1.4 Å upon superimposition. In
chain A of the orthorhombic crystal form (inhibitors 2-4) and chains A, B, and C of the triclinic
space group (inhibitors 1,5-7) the ligand is bound to the enzyme in an open conformation
characterized by a disordered active site loop (residues 211 – 227) and a disordered Cꢀterminus,
residues (310ꢀ323). The disorder of these peptide segments results in an open active site, which
renders the inhibitor partially accessible from the solvent. In the B chains of the orthorhombic
crystals the active site loop is however well defined in electron density and has folded back over
the active site thus shielding the bound ligand from the solvent. Also in chain D of the triclinic
space group both active site loop and the Cꢀterminal residues are well defined in density and
contribute to ligand binding. The active site loop is folded over the ligandꢀbinding site and the Cꢀ
terminal residues 320ꢀ323 are inserted into the active site cleft. In the following we describe
enzymeꢀligand interactions for the closed conformations as in these structures ligand interactions
are fully developed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Enzyme-inhibitor interactions and emerging SAR: The confirmed smallꢀmolecule inhibitors
consist of three parts: a leftꢀhandꢀside (LHS) substituted phenyl group, a central linker part and a
rightꢀhandꢀside (RHS) aryl group (Table 1). In all seven coꢀcrystal structures, the compounds
bind in the enzyme active site parallel to the plane of the pyrimidine ring of the PLP cofactor,
which forms stacking interactions with the urea linker (Supporting Figure S6A).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 45
1
2
3
4
5
6
7
8
The LHS carboxylate group is consistently positioned in the substrate binding pocket in a similar
way as the carboxylate group of the aꢀaminoacrylate intermediate, as previously observed in
CysK1 from M. tuberculosis14, while the RHS binds into a hydrophobic pocket that extends from
the PLP site toward the enzyme surface. The recognition between the enzyme and the linker and
RHS, respectively, differs significantly between compounds carrying a RHS phenyl (denoted
type I inhibitors) or biphenyl group (type II inhibitors), respectively, (Figure 4) as outlined in
detail below.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The LHS of type I inhibitors 1, 2, 4−7 show similar interactions to residues in the carboxylate
binding pocket, i.e. hydrogen bonds to main chain amides of Thr78, Ser79, Asn81 and Thr82
located at the Nꢀterminal of an α–helix (Figure 4a & b). The interaction between the LHS
carboxylate substituent and the enzyme in the carboxylate binding pocket is found in all potent
inhibitors of CysM homologues described so far, i.e. CysK1 from M. tuberculosis14,25, OASS
from Haemophilus influenzae31,32 and Salmonella typhimurium26 and thus appears to be a
hallmark of this enzyme family. The exchange of the carboxylic acid group for the tetrazole
bioisostere further supports this notion, as this resulted in a complete loss of affinity (compounds
1 vs 8, Table 1). Further examples reveal that type I inhibitors showed little tolerance for
introduction of substituents at LHS. Modifications such as addition of a hydroxyl group and a
switch from mꢀ to pꢀcarboxylate led to loss of binding (compounds 9−11). However, type II
inhibitors with the corresponding LHS modifications retained binding affinity and showed
considerable inhibitory power (compounds 3−4 vs 9−10). This difference relates to
ACS Paragon Plus Environment

Page 11 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
conformational changes in the protein induced by the linker and the RHS of type II inhibitors as
discussed below.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
We next explored the role of the central linker region. In the crystal structures of CysM with type
I inhibitors the carbonyl group of the urea linker interacts with the side chain of Asn221 (Figure
4a). One of the amides interacts with the terminal carboxyl group of Ala323, inserted into the
active site and the second amide forms a hydrogen bond to an internal water molecule. In the
structures of type II inhibitors, the carbonyl oxygen of the linker instead forms a hydrogen bond
with Arg220 from the mobile active site loop and both amides interact with internal water
molecules (Figure 4b). To understand the SAR and explore the potential for further optimization
of the hit compounds, the urea motif was exchanged for thiourea, sulfonamide or amide linkers.
The linker was also extended by a oneꢀcarbon elongation, but all these modifications gave
inactive or poorly active compounds (compounds 12–15, Table 1), confirming the importance of
the urea hydrogen bonds observed. The difference in activity between urea and thiourea analogs
(e.g. compound 7 vs 12) may be due to the weaker hydrogen bonding ability of sulfur compared
to oxygen and the larger size of the sulfur atom that may affect the angle between the LHS and
RHS phenyl groups. Additional examples of thiourea analogs supporting these results are
provided in Supporting Table S2 (e.g. compound 20 and 21).
Significant differences between type I and II inhibitors were also found for the interactions of the
enzyme with the RHS. The crystal structures of CysM with 1, 5, 6 or 7 (type I) are all very
similar, with r.m.s.d. values for Cα atoms after superimposition of 0.25–0.3 Å and they also
display the same enzymeꢀligand interactions. Binding of the type I inhibitors induces a disorderꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 45
1
2
3
4
5
6
7
8
9
order transition involving the mobile active site loop (residues 211−227), which folds over the
inhibitor binding site, and the Cꢀterminal residues that insert into the active site and form
important interactions with the linker (Figure 4c). The RHS phenyl group is bound in a
hydrophobic pocket denoted pocket A (Figure 4d). The size of this pocket is too small to
accommodate inhibitors with a biphenyl substituent, and binding of type II compounds induces a
different conformation of the active site loop that increases the size of cavity A by pocket B. The
different conformation of the active site loop at the same time prevents insertion of the Cꢀ
terminus into the active site leading to a change in the interactions of the linker with enzyme
residues. The expansion of the binding pocket A with pocket B thus provides sufficient space for
the larger biphenyl substituent of compounds 2, 3 and 4 (Figure 4b and e).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The unsubstituted phenyl analog 16 showed an apparent Kd of 12.9 ꢀM, similar to derivatives
with the strongly electronꢀwithdrawing pꢀNO2 substituent and the electronꢀdonating pꢀOMe
group (compounds 17 and 18) as well as the pꢀMe analog 5. However, substitutions with
halogens (ꢀCl 7, ꢀBr 6) resulted in a stronger binding, indicating that pꢀhalogens are preferred.
Substitution with chloride in both the para and meta−position resulted in the best inhibitor 1 with
a 40 fold better Kd value than the parent compound 16. The electron withdrawing acetyl group
had a negative effect on the activity as compound 19 was a poor inhibitor.
The terminal phenyl group of compound 2 binds in pocket B (Figure 4e) and the crystal
structures of CysM with 2, 3 or 4 show that there is little additional space available in this pocket
(less than 4 Å in any direction) for further modifications of this phenyl ring. Consequently,
substitutions such as ꢀMe, ꢀCO2H or ꢀF all result in lower affinity or a complete loss of binding
(compounds 58, 66, and 69, Supporting Table S2). Exchanging the RHS terminal phenyl ring for
ACS Paragon Plus Environment

Page 13 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a nitrogen heterocycle (pyridine or pyrimidine) (compounds 63, 68) led to less active compounds
(1–2 orders of magnitude difference in Kd), most likely due to less favorable interactions of these
more polar ring systems with the hydrophobic binding pocket.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Anti-mycobacterial activity and cytotoxicity: The most potent CysM inhibitors were further
tested against a growing culture of M. tuberculosis and all showed modest bactericidal effects,
with MIC values ranging from 2 – 20 ꢀM (0.7ꢀ 6.2 ꢀg/ml) (Table 2). Although these observed
MIC values are in the low ꢀM range, they do not reach the potency of clinically used drugs such
as isoniazid and rifampicin. As the gene coding for CysM is predominantly expressed during
dormancy8,9,19 this is also not to be expected. We therefore used the nutrient starvation model19 to
test the antiꢀmycobacterial activity of these inhibitors at 10 µM concentration under conditions
simulating dormancy, including a comparison with the drugs isoniazid, rifampicin and
moxifloxacin (Figure 5A). As expected isoniazid showed only a marginal (~ 1 log) effect on
nonꢀreplicating cells, and compounds 6, 7, 17 and 18 displayed a similar lack of potency.
Compounds 1, 2, 3 and 5 performed considerably better, with 1 and 3 showing a 3ꢀlog decrease
in bacterial count, corresponding to similar or improved killing activities when compared to
rifampicin and moxifloxacin.
In an effort to initiate studies of the mechanism of action of these compounds also in vivo we
next examined the effect of cysteine complementation during the treatment of nutrientꢀstarved
cultures of M. tuberculosis with compounds 1 and 3 (Figure 5B). Addition of cysteine (0.2 mM)
led to a complete loss of the bactericidal effect of the two inhibitors. We could not detect any
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 45
1
2
3
4
5
6
7
8
9
degradation of compound 1 during incubation with cysteine for 24 hrs using massꢀspectrometry,
HPLC and spectroscopy, thus excluding inactivation of 1 by reaction with cysteine as the cause
for the loss of antiꢀmycobacterial activity. These data point towards cysteine biosynthesis as
important component of the action of compounds 1 and 3 also in vivo. Additional experiments
are required to further validate these data and to exclude contributions from other mechanisms.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To better understand whether these inhibitors are useful for later studies also in the context of
bacterial viability within mammalian cells we performed cell viability studies in a small panel of
human and mouseꢀderived cell lines. These cytotoxicity assays with the urea derivatives using
three different human cell lines and a mouse cell line did not reveal significant cytotoxic effects
at concentrations of 10 ꢀM (Supporting Figure S7 and S8).
Compound selectivity: De novo cysteine biosynthesis in M. tuberculosis is rather complex, with
three potential pathways leading to cysteine. Each of these routes contains a specific PLPꢀ
dependent cysteine synthase, CysK1, CysK2 and CysM (Figure 1).11 In order to assess the
selectivity of the ureaꢀderived inhibitors, we examined the binding and affinity of compound 1
and 2 to the mycobacterial homologues using the fluorescence−based binding assay. Compounds
1 and 2 bound to CysK1, albeit much weaker, returning Kd values of 241.1±25.3 µM and 96.8
±16.1 µM, respectively, showing an eighty to several hundred fold stronger binding to CysM
than to CysK1. CysK2 also displayed weaker interactions with 1 (30.1±3.4 µM) and 2 (22.6±2.4
µM), corresponding to a 15ꢀ100 fold preference of these compounds for CysM over CysK2.
ACS Paragon Plus Environment

Page 15 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As humans lack the de novo biosynthesis pathways to cysteine, these enzymes are particularly
attractive as potential targets for new antibiotics. A search of the human genome returned CBSA
(Cystathionine betaꢀsynthase, Uniprot AccNr: P35520, 29% sequence identity) that uses serine
and homocysteine as substrates as the closest sequence homologue of CysM. CBSA binds
compound 1 with a Kd of 94 µM, i.e. with approximately 300 times lower affinity compared to
the mycobacterial target protein and does not bind 2. The low affinity of the inhibitors for the
closest human homologue and their low cytotoxicity suggest that these compounds are suitable
starting points for further development.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSIONS
Activity assays of enzymes utilizing other postꢀtranslationally modified proteins, for instance
thiocarboxylated CysO, as (coꢀ)substrates can be challenging and are not easily adapted to high
throughput format. Here we monitored changes in the fluorescence emission of the PLP cofactor
in 384ꢀwell format and successfully screened a chemically diverse substance library. The screen
resulted in the identification of new smallꢀmolecule binders of CysM following removal of false
positives that were disregarded based on their inherent fluorescence. We note, however, that our
approach does not capture those compounds that bind at sites far from the PLP cofactor and thus
potential allosteric inhibitors may have been missed.
One of the identified compound classes is based on an urea scaffold. This finding is in line with
results from a previous phenotypic highꢀthroughput screen for inhibitors of M. tuberculosis
H37Rv that also included several ureaꢀbased hits with antiꢀmycobacterial activities.33 We
explored the urea hits from our targetꢀbased screen in more detail, which led to the discovery of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 45
1
2
3
4
5
6
7
8
9
potent CysM inhibitors that all displayed a similar binding mode to the enzyme active site.
However, active site loops of CysM showed a considerable flexibility in creating suitable
binding pockets to accommodate the differently substituted inhibitors. This conformational
flexibility of the enzyme is difficult to predict and illustrates present limitations in the scope of in
silico screening approaches.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A search of the Reaxys Medicinal Chemistry (www.reaxys.com) and the ChEMBL34 databases
using the structures of compounds 1, 2 and 3 regarding previously reported bioactivities
identified compound 1 in a patent35 claiming inhibitors of the mammalian kinase p38, whereas
no known bioactivities could be found for 2 and 3. Based on information from the coꢀcrystal
structures and the emerging SAR we next expanded the search to include compounds displaying
80 % structural similarity to our top hits, retaining the crucial mꢀcarboxylate on the LHS,
including the unsubstituted analog 16. This search identified six compounds, for which a total of
147 bioactivities have been reported in ChEMBL. None of these is on targets of mycobacterial
origin. Activities include the mammalian targets NiemannꢀPick C1, Rasꢀrelated protein Rabꢀ
9A, luciferin 4ꢀmonooxygenase, GLP1ꢀreceptor, microRNA21, ATPase family AAA domainꢀ
containing protein 5, nuclear factor erythroid 2ꢀrelated factor 2, Ataxinꢀ2, H4 and broadly on
Jurkat cells. Further optimization of our CysM top hits will have to address this issue, but the
lack of cytotoxicity of the top hits are encouraging to progress further.
In view of the large number of individuals infected with persistent M. tuberculosis worldwide
drugs active against the latent stage of this infectious disease are urgently needed. Indeed most of
the antiꢀmycobacterial drugs in clinical use are not effective in the dormant state of the disease.
Significantly, several of the top hits identified in this work display potent antibacterial activity,
ACS Paragon Plus Environment

Page 17 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with three compounds showing potency against nonꢀreplicating M. tuberculosis and thus are
promising leads for further development. First experiments towards the validation of CysM as
the in vivo target of the top hits showed that addition of cysteine during compound treatment led
to complete loss of bactericidal activity in a nutrientꢀstarvation model of M. tuberculosis,
suggesting perturbed cysteine biosynthesis in the pathogen as a result of compound exposure.
Although these results are promising we note that unambiguous target validation requires more
elaborate efforts to firmly establish the in vivo mode of action of these molecules. Thus, at
present, we cannot exclude additional offꢀtargets effects.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The genome of M. tuberculosis contains three genes that code for OASS homologues that have
been associated with three different de novo biosynthetic pathways to cysteine in this organism
(Figure 1). These enzymes belong to the same family, with sequence identities ranging from 26–
37%. While the precise role of these pathways and their regulation is not well understood, it
appears that cysteine synthesis under normal growth conditions relies predominantly on the
classical, CysK1ꢀdependent pathway, whereas CysM/CysO dependent cysteine biosynthesis
occurs during intraꢀmacrophage lifestyle and dormancy.21,22 Selective inhibitors of enzymes from
these pathways, for instance compound 1, which inhibits CysM more potently than other family
members, are useful chemical tools to elucidate the regulation of cysteine biosynthesis in M.
tuberculosis under different metabolic conditions.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 45
1
2
3
4
MATERIALS & METHODS
5
6
7
8
9
Protein production. Recombinant CysM, CysO, CysK1 and CysK2 were produced and purified
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
as described in the Supporting Information.
PLP-fluorescence-based screening assay. The CysM smallꢀmolecule screen was performed by
monitoring the fluorescence of the PLP cofactor. A library of 17312 compounds was screened at
a single concentration of 10 ꢁM in the presence of 2 µM CysM. Compounds that showed an
increase in fluorescence by ≥30% were identified as hits. For details about the small molecule
library, the screening assay and the followꢀup hit confirmation as well as the counter assay
experiments see Supporting Information.
Chemistry – General methods: The compounds in this study were either cherryꢀpicked from
the CBCS compound selection or synthesized. For synthesis, all commercially available
chemicals and solvents were used without further purification. Analytical thin layer
chromatography (TLC) was performed on aluminaꢀbacked silica gel 40 F254 plates (Merck,
Darmstadt, Germany) that were visualized with UV light and KMnO4 solution, and flash column
chromatography was carried out on 60 Å (40−63 ꢀm) silica gel. HPLCꢀMS analyses were run on
1
13
an Agilent series 6100B system with electrospray ionization (ESI+). H and C NMR spectra
were recorded at 400 MHz and 100 MHz respectively on a Bruker AM spectrometer. Chemical
shifts are reported in ppm (δ) with reference to TMS as internal standard, and coupling constants
J are quoted in Hertz. Elemental analyses were carried out using a Vario MICRO cube in CHN
mode. The experimental details of the most potent CysM inhibitors 1–7 are described below. A
ACS Paragon Plus Environment

Page 19 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
full description of the synthetic protocol and spectroscopic analysis of the other compounds is
found in the Supporting information. All compounds show a purity > 95% as determined by
HPLC.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The seven hit compounds were examined using the PAINS28 and REOS29 filters in the software
Canvas36. Also, the seven hits were evaluated as potential aggregators30, using structural
similarity to known aggregators and SlogP calculations in a Knime work flow
(http://www.myexperiment.org/workflows/4749.html) assessing the aggregating risk.
General procedure for the synthesis of 1, 5, 6 and 7: Substituted phenyl isocyanate (1 mmol)
was added dropwise to a stirred iceꢀcooled solution of 3ꢀamino benzoic acid, 4ꢀamino salicylic
acid, 5ꢀamino salicylic acid or 4ꢀaminoꢀ3ꢀhydroxy benzoic acid (1 mmol), respectively, and
trimethylamine (2 mmol) in dry ethanol. Reaction progress was monitored by TLC and HPLCꢀ
MS. After 24 h, solvent was evaporated under reduced pressure and the resulting residue was
diluted with aq. sodium bicarbonate and the mixture was filtered. The filtrate was acidified with
2N hydrochloric acid and the precipitate was filtered and recrystallized in ethanol to yield the
respective urea derivatives.
General procedure for the synthesis of 2,
3
and 4: Substituted 3ꢀ(3ꢀ(4ꢀ
bromophenyl)ureido)benzoic acid (1 mmol), the corresponding boronic acid (2 mmol), sodium
carbonate (2M solution) (4 mmol) and Pd(PPh3)4 (0.06 mmol) were taken in acetonitrile and
irradiated in microwave at 100 °C for 1 h (monitored by TLC and LCMS for completion). The
reaction mixture was passed through celite and washed with methanol. The filtrate was
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 45
1
2
3
4
5
6
concentrated and purified by column chromatography using hexane:ethyl acetate as eluent to get
the desired product in good yield.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3ꢀ(3ꢀ(3,4ꢀDichlorophenyl)ureido)benzoic acid (1). The compound was synthesized according to
the general procedure using 3ꢀamino benzoic acid (0.10g, 0.73 mmol), 3,4ꢀdichloro phenyl
isocyanate (0.14g, 0.73 mmol) and triethylamine (0.15g, 1.46 mmol) to afford 1 as white solid
1
(0.18g, 75% yield). M.P:119 – 121 °C. H NMR (DMSOꢀd6): δ 12.96 (br s, 1H), 9.30 (s, 1H),
9.25 (s, 1H), 8.14 (s, 1H), 7.90 (s, 1H), 7.59 (d, J = 8 Hz, 1H), 7.54 (d, J = 8.8 Hz, 1H), 7.44 (t, J
= 8 Hz, 1H), 7.37 (m, 2H). 13C NMR (DMSOꢀd6): δ 167.2, 152.3, 139.9, 139.6, 131.3, 131.0,
130.5, 129.0, 123.1, 123.0, 122.1, 119.3, 119.0, 118.3. ESIꢀMS m/z 324.01(MꢀH)+. Anal. Calcd
for C14H10Cl2N2O3; C, 51.72; H, 3.10; N, 8.62; Found C, 51.82; H, 3.09; N, 8.64.
3ꢀ(3ꢀ([1,1'ꢀBiphenyl]ꢀ3ꢀyl)ureido) benzoic acid (2). The compound was synthesized according
to the general procedure using 3ꢀ(3ꢀ(3ꢀbromophenyl)ureido)benzoic acid (0.1 g, 0.24 mmol),
phenyl boronic acid (0.057 g, 0.47 mmol), Na2CO3 (0.1g, 0.96 mmol) and Tetrakis (0.019, 0.016
mmol) to afford 2 as white solid (0.17 g, 79% yield). M.P. 229–231 °C.1H NMR (DMSOꢀd6): δ
12.8 (br s, 1H), 9.54 (s, 1H), 9.48 (s, 1H), 8.13 (s, 1H), 8.03 (s, 1H), 7.99 – 7.79 (m, 2H), 7.64 –
7.32 (m, 9H). 13C NMR (DMSOꢀd6): δ 167.2, 152.6, 144.6, 141.2, 140.6, 135.8, 130.2 (2C),
129.8, 129.0, 128.5, 127.9, 127.6 (2C), 126.4, 126.3, 124.5, 122.2, 121.8, 120.6. EIꢀMS m/z:
331.01 (MꢀH)+. Anal. Calcd for C20H16N2O3: C, 72.28; H, 4.85; N, 8.43. Found: C, 72.46; H,
4.86; N, 8.41.
ACS Paragon Plus Environment

Page 21 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4ꢀ(3ꢀ([1,1'ꢀBiphenyl]ꢀ3ꢀyl)ureido)ꢀ2ꢀhydroxybenzoic acid (3). The compound was synthesized
according to the general procedure using 4ꢀ(3ꢀ(3ꢀbromophenyl)ureido)ꢀ2ꢀhydroxybenzoic acid
(0.1 g, 0.28 mmol), phenyl boronic acid (0.069 g, 0.56 mmol), Na2CO3 (0.12g, 1.12 mmol) and
Tetrakis (0.021 g, 0.019 mmol) to afford 3 as white solid (0.042 g, 43% yield). MP: 210 – 212
°C. 1H NMR (DMSOꢀd6): δ 12.86 (br s, 1H), 9.68 (s, 1H), 9.57 (s, 1H), 7.87 (s, 1H), 7.66 (t, J =
8 Hz, 3H), 7.50 – 7.35 (m, 5H), 7.26 (d, J = 7.6 Hz, 1H), 7.03 (s, 1H), 6.84 (d, J = 7.2, 1H), 4.63
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
(s, 1H). C NMR (DMSOꢀd6): δ 172.6, 162.9, 152.6, 143.6, 140.7, 140.6, 140.4, 130.6, 129.3,
128.9 (2C), 127.4, 126.6 (2C), 120.0, 117.2, 116.4, 113.2, 107.1, 104.5. EIꢀMS m/z: 347.01 (Mꢀ
H)+. Anal. Calcd for C20H16N2O4: C, 68.96; H, 4.63; N, 8.04; Found C, 69.13; H, 4.62; N, 8.03.
5ꢀ(3ꢀ([1,1'ꢀBiphenyl]ꢀ3ꢀyl)ureido)ꢀ2ꢀhydroxybenzoic acid (4). The compound was synthesized
according to the general procedure using 5ꢀ(3ꢀ(3ꢀbromophenyl)ureido)ꢀ2ꢀhydroxybenzoic acid
(0.1 g, 0.28 mmol), phenylboronic acid (0.069 g, 0.56 mmol), Na2CO3 (0.12g, 1.12 mmol) and
Tetrakis (0.021 g, 0.019 mmol) to afford 4 as brown solid (0.048 g, 48% yield). MP: 208 – 210
1
°C. H NMR (DMSOꢀd6): δ 12.79 (br s, 1H), 9.55 (s, 1H), 9.49 (s, 1H), 8.12 (s, 1H), 8.05 (s,
13
1H), 7.93 (d, J = 8.2 Hz, 1H), 7.65 – 7.01 (m, 9H), 5.03 (s, 1H). C NMR (DMSOꢀd6): δ 172.3,
162.2, 152.5, 139.6, 139.0, 137.8, 137.5, 130.6, 130.2 (2C), 129.5, 128.3 (2C), 128.2, 124.6,
122.8, 121.6, 120.5, 118.7, 117.7. EIꢀMS m/z: 347.01 (MꢀH)+. Anal. Calcd for C20H16N2O4: C,
68.96; H, 4.63; N, 8.04; Found C, 69.10; H, 4.64; N, 8.02.
3ꢀ(3ꢀ(pꢀTolyl)ureido) benzoic acid (5). The compound was synthesized according to the general
procedure using 3ꢀamino benzoic acid (0.10g, 0.73 mmol), pꢀtolylisocyanate (0.09g, 0.73 mmol)
and triethylamine (0.15g, 1.46 mmol) to afford 5 as pale yellow solid (0.14g, 71% yield). M.P:
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 45
1
2
242 – 244 °C. 1H NMR (DMSOꢀd6): δ. 12.74 (br s, 1H), 9.32 (s, 1H), 8.91 (s, 1H), 8.23 (s, 1H),
7.78 (d, J = 8.2 Hz, 2H), 7.73 – 7.41 (m, 5H), 2.56 (s, 3H). 13C NMR (DMSOꢀd6): δ 167.5,
152.7, 142.7,138.9, 138.1, 130.7 (2C), 126.7, 126.3, 124.4, 123.9, 119.8 (2C), 118.2, 22.6. ESIꢀ
MS m/z 269.10(MꢀH)+. Anal. Calcd for C15H14N2O3; C, 66.66; H, 5.22; N, 10.36; Found C,
66.48; H, 5.24; N, 10.39.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3ꢀ(3ꢀ(4ꢀBromophenyl)ureido)benzoic acid (6). The compound was synthesized according to the
general procedure using 3ꢀamino benzoic acid (0.10g, 0.73 mmol), 4ꢀbromo phenyl isocyanate
(0.14g, 0.73 mmol) and triethylamine (0.15g, 1.46 mmol) to afford 6 as white solid (0.19g, 80%
1
yield). M.P: 275 – 277 °C. H NMR (DMSOꢀd6): δ 12.63 (br s, 1H), 9.00 (s, 1H), 8.93 (s, 1H),
8.12 (s, 1H), 7.65 (m, 1H), 7.57 (d, J = 8 Hz, 1H), 7.45 – 7.39 (m, 5H). 13C NMR (DMSOꢀd6): δ
167.3, 152.7, 145.8, 140.9, 133.8 (2C), 129.6, 128.4, 124.1, 122.8, 120.3, 119.6 (2C), 118.5. ESIꢀ
MS m/z 334.01 (MꢀH)+. Anal. Calcd for C14H11BrN2O3; C, 50.17; H, 3.31; N, 8.36; Found C,
50.32; H, 3.33; N, 8.38.
3ꢀ(3ꢀ(4ꢀChlorophenyl)ureido)benzoic acid (7). The compound was synthesized according to the
general procedure using 3ꢀamino benzoic acid (0.10g, 0.73 mmol), 4ꢀchloro phenyl isocyanate
(0.11g, 0.73 mmol) and triethylamine (0.15g, 1.46 mmol) to afford 7 as white solid (0.17g, 78%
1
yield). M.P: 279 – 281 °C. H NMR (DMSOꢀd6): δ 12.88 (br s, 1H), 8.98 (s, 1H), 8.89 (s, 1H),
8.15 (s, 1H), 7.88 (d, J = 8 Hz, 2H), 7.68 – 7.36 (m, 5H). 13C NMR (DMSOꢀd6): δ 168.0, 152.7,
141.9, 139.3, 135.8, 130.3 (2C), 129.6, 128.5, 127.1, 123.4, 121.6, 118.6 (2C). ESIꢀMS m/z
289.05 (MꢀH)+. Anal. Calcd for C14H11ClN2O3; C, 57.84; H, 3.81; N, 9.64; Found C, 57.67; H,
3.82; N, 9.62.
ACS Paragon Plus Environment

Page 23 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Enzyme assays. The catalytic cycle of CysM and the basis for the different assays used in this
study are shown in supporting Figure S4. The influence of the inhibitors on the first halfꢀreaction
of CysM was monitored spectrophotometrically at 463 nm, the characteristic wavelength of the
aminoacrylate intermediate.16 The apparent kobs values were derived from exponential fits to the
recorded absorbance data and compared to the nonꢀinhibited reaction.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Electrospray ionization mass spectrometry was used to assess the formation of the covalent
CysOꢀcysteine adduct in the second halfꢀreaction of the CysM catalytic cycle in the presence of
top HTS hits and their derivatives. Mass spectra of the reaction mixture products were compared
to the mass spectrum of the covalent CysOꢀcysteine adduct formed in the nonꢀinhibited reaction
(9658.7 Da) as well as to the mass spectrum of the substrate CysOꢀSH (9571.6 Da). Multiple
turnover enzyme activity assays for IC50 determination for CysM with the natural substrates OPS
and thiocarboxylated CysO (CysOꢀSH) were based on the spectrophotometric determination of
the stoichiometric amount of phosphate ions released using the Malachiteꢀgreen (MG) reagent.
For details of the assays, see Supporting Information.
ASSOCIATED CONTENT
Supporting Information.
Additional experimental details concerning isothermal titration calorimetry, compound synthesis,
crystallization and structure determination of enzymeꢀhit complexes, mammalian cell
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 45
1
2
3
4
5
6
cytotoxicity evaluation and antiꢀmycobacterial potency assays. This material is available free of
charge via the Internet.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Accession Codes
Crystallographic data for the enzymeꢀligand complexes have been deposited with the Protein
Data Bank under accession codes 5I7A (CysMꢀ1 complex), 5I7R (CysMꢀ2 complex), 5IWC
(CysMꢀ3 complex), 5IW8 (CysMꢀ4 complex), 5I6D (CysMꢀ5 complex), 5I7H (CysMꢀ6
complex) and 5I7O (CysMꢀ7 complex). The crystallographic data will be released upon article
publication.
AUTHOR INFORMATION
Corresponding Author
G. Schneider eꢀmail: Gunter.Schneider@ki.se; Phone:+4652487675 FAX:+468327626; R.
Schnell, eꢀmail: robert.schnell@ki.se; D. Sriram, email: dsriram@hyderabad.bitsꢀpilani.ac.in
Author contribution
#: equally contribution authors
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We gratefully acknowledge access to synchrotron radiation at the ESRF, France and BESSY,
Germany and to instrumentation at the Protein Science Facility, Karolinska Institutet. We also
ACS Paragon Plus Environment

Page 25 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
thank Jolanta Kopec, Structural Genomics Consortium (SGCꢀOxford) for provision of a sample
of CBSA. This work was supported by the Swedish Research Council and the Department of
Biotechnology (DBT), India. Chemical Biology Consortium Sweden is funded by the Karolinska
Institutet, SciLifeLab and the Swedish Research Council. DS acknowledges DBT for the Tata
Innovation Fellowship.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABBREVIATIONS USED
APS, adenosineꢀ5’ꢀphosphosulfate; CBSA cystathionineꢀbetaꢀsynthase; HTS highꢀthroughput
screening; INH isoniazid; ITC, isothermal titration calorimetry; MIC, minimal inhibitory
concentration; MPN, most probable number; MOXI, moxifloxacin; OAS, OꢀacetylꢀLꢀserine;
OASS, Oꢀacetylserine sulfhydrylase, OPS, OꢀphosphoꢀLꢀserine; RIF, rifampicin; RNI, reactive
nitrogen intermediates; PAPS, 3’ꢀphosphoadenosineꢀ5’ꢀphosphosulfate; PLP, pyridoxalꢀ
phosphate; PAINS, pan assay interference compounds; REOS, removal of unwanted reactive
groups; ROI, reactive oxygen intermediates; TMS, tetramethylsilane.
REFERENCES
(1) Gengenbacher, M.; Kaufmann, S.H. Mycobacterium tuberculosis: success through dormancy.
FEMS Microbiol. Rev. 2012, 36, 514ꢀ532.
(2) Guirado, E.; Schlesinger, L.S. Modeling the Mycobacterium tuberculosis granuloma ꢀ the
critical battlefield in host immunity and disease. Front. Immunol. 2013, 4, 98.
(3) Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. The challenge of new drug
discovery for tuberculosis. Nature 2011, 469, 483ꢀ490.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 45
1
2
3
4
5
6
7
8
9
(4) Dorhoi, A.; Reece, S.T.; Kaufmann S.H. For better or for worse: the immune response
against Mycobacterium tuberculosis balances pathology and protection. Immunol. Rev. 2011,
240, 235ꢀ251.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5) Sareen, D.; Newton, G.L.; Fahey, R.C.; Buchmeier, N.A. Mycothiol is essential for growth of
Mycobacterium tuberculosis. J. Bacteriol. 2003, 185, 6736–6740.
(6) Bhaskar, A.; Chawla, M.; Mehta, M.; Parikh, P.; Chandra, P.; Bhave, D.; Kumar, D.; Carroll,
K.S.; Singh, A. Reengineering redox sensitive GFP to measure mycothiol redox potential of
Mycobacterium tuberculosis during infection. PLoS Pathog. 2016, 10, e1003902.
(7) 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,
1091ꢀ1117.
(8) 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, 693ꢀ704.
(9) Voskuil, M. I.; Visconti, K. C.; Schoolnik, G. K. Mycobacterium tuberculosis gene
expression during adaptation to stationary phase and lowꢀoxygen dormancy. Tuberculosis 2004,
84, 218ꢀ227.
(10) Bhave, D. P.; Muse, W. B., 3rd; Carroll, K. S. Drug targets in mycobacterial sulfur
metabolism. Infect. Disord. Drug Targets 2007, 7, 140ꢀ158.
(11) Schnell, R.; Sriram, D.; Schneider, G. Pyridoxalꢀphosphate dependent mycobacterial
cysteine synthases: structure, mechanism and potential as drug targets. Biochem. Biophys. Acta
2014, 1854, 1175ꢀ1183
ACS Paragon Plus Environment

Page 27 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(12) Palde, P.B.; Bhaskar, A.; Pedró Rosa, L.E.; Madoux, F.; Chase, P.; Gupta, V.; Spicer, T.;
Scampavia, L.; Singh, A.; Carroll K.S. Firstꢀinꢀclass inhibitors of sulfur metabolism with
bactericidal activity against nonꢀreplicating M. tuberculosis. ACS Chem. Biol. 2016, 11, 172ꢀ184
(13) Schnell, R.; Sandalova, T.; Hellman, U.; Lindqvist, Y.; Schneider, G. Sirohemeꢀ and [Fe4ꢀ
S4]ꢀdependent NirA from Mycobacterium tuberculosis is a sulfite reductase with a covalent Cysꢀ
Tyr bond in the active site, J. Biol. Chem. 2005, 280, 27319ꢀ27328.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(14) Schnell, R.; Oehlmann, W.; Singh, M.; Schneider, G. Structural insights into catalysis and
inhibition of Oꢀacetylserine sulfhydrylase from Mycobacterium tuberculosis ꢀ crystal structures
of the enzyme alphaꢀaminoacrylate intermediate and an enzymeꢀinhibitor complex. J. Biol.
Chem. 2007, 282, 23473ꢀ23481.
(15) Burns, K.E.; Baumgart, S.; Dorrestein, P.C.; Zhai, H.; McLafferty, F.W.; Begley, T.P.
Reconstitution of a new cysteine biosynthetic pathway in Mycobacterium tuberculosis. J. Am.
Chem. Soc. 2005, 127, 11602ꢀ11603.
(16) Ågren D.; Schnell R.; Oehlmann D.W.; Singh M.; Schneider G. Cysteine synthase (CysM)
of Mycobacterium tuberculosis is an Oꢀphosphoserine sulfhydrylase: evidence for an alternative
cysteine biosynthesis pathway in mycobacteria. J. Biol. Chem. 2008, 283, 31567ꢀ31574.
(17) O'Leary, S.E.; Jurgenson, C.T.; Ealick, S.E.; Begley, T.P. OꢀphosphoꢀLꢀserine and the
thiocarboxylated sulfur carrier protein CysOꢀCOSH are substrates for CysM, a cysteine synthase
from Mycobacterium tuberculosis. Biochemistry 2008, 47, 11606ꢀ11615.
(18) Steiner, E.M.; Böth, D.; Lössl, P.; Vilaplana, F.; Schnell, R.; Schneider, G. CysK2 from
Mycobacterium tuberculosis is an OꢀphosphoꢀLꢀSerine dependent Sꢀsulfocysteine synthase. J
Bacteriol. 2014, 196, 3410ꢀ3420.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 45
1
2
3
4
5
6
7
8
9
(19) 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, 717ꢀ731.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(20) 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, 228ꢀ238.
(21) Rengarajan, J.; Bloom B.R.; Rubin, E.J. Genomeꢀwide requirements for Mycobacterium
tuberculosis adaptation and survival in macrophages. Proc. Natl. Acad. Sci. U. S. A. 2005, 102,
8327–8332.
(22) Sassetti, C. M.; Rubin E.J. Genetic requirements for mycobacterial survival during
infection. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 12989–12994.
(23) Ågren, D.; Schnell, R.; Schneider, G. The Cꢀterminal of CysM from Mycobacterium
tuberculosis protects the aminoacrylate intermediate and is involved in sulfur donor selectivity.
FEBS Lett. 2009, 583, 330ꢀ336.
(24) Campanini, B.; Speroni, F.; Salsi, E.; Cook, P.F.; Roderick, S.L.; Huang, B.; Bettati, S.;
Mozzarelli, A. Interaction of serine acetyltransferase with Oꢀacetylserine sulfhydrylase active
site: evidence from fluorescence spectroscopy. Protein Sci. 2005, 14, 2115ꢀ2124
(25) Poyraz, O.; Jeankumar, V.U.; Saxena, S.; Schnell, R.; Haraldsson, M.; Yogeeswari, P.;
Sriram, D.; Schneider, G. Structureꢀguided design of novel thiazolidine inhibitors of Oꢀacetyl
serine sulfhydrylase from Mycobacterium tuberculosis. J. Med. Chem. 2013, 56, 6457ꢀ6466.
(26) Pieroni, M.; Annunziato, G.; Beato, C.; Wouters, R.; Benoni, R.; Campanini, B.; Pertinhez,
T.; Bettati, S.; Mozzarelli, A.; Costantino, G. Rational Design, Synthesis and Preliminary
ACS Paragon Plus Environment

Page 29 of 45
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structureꢀactivity relationships of αꢀsubstitutedꢀ2ꢀphenylcyclopropane carboxylic acids as
inhibitors of Salmonella typhimurium Oꢀacetylserine sulfhydrylase. J. Med. Chem. 2016, DOI:
10.1021/acs/jmedchem.5bo1775
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(27) Lipinski, C.A. Leadꢀ and drugꢀlike compounds: the ruleꢀofꢀfive revolution. Drug Discovery
Today 2004, 4, 337ꢀ341
(28) 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
(29) Walters, W.P,; Ajay; Murcko, M.A Recognizing molecules with drugꢀlike properties. Curr.
Opin. Chem. Biol. 1999, 3: 384–387
(30) Irwin, J.J.; Duan, D.; Torosyan, H.; Doak, A.K.; Ziebart, K.T.; Sterling, T.; Tumanian, G.;
Shoichet, B.K. An aggregation advisor for ligand discovery. J. Med. Chem. 2015, 58, 7076ꢀ7087.
(31) 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 Oꢀacetylserine
sulfhydrylase inhibitors by mimicking nature. J. Med. Chem. 2010, 53, 345ꢀ356
(32) Amori, L.; Katkevica, S.; Bruno, A.; Campanini, B.; Felici, P.; Mozzarelli, A.; Costantino,
G. Design and synthesis of transꢀ2ꢀsubstitutedꢀcyclopropaneꢀ1ꢀcarboxylic acids as the first nonꢀ
natural small molecule inhibitors of Oꢀacetylserine sulfhydrylase. MedChemComm 2012, 3,
1111ꢀ1116
(33) Ananthan, S.; Faaleolea, E.R.M; Goldman, R.C.; Hobrath, J.V.; Kwong, C.D.; Laughon,
B.E.; Maddry, J.A.; Mehta, A.; Rasmussen, L.; Reynolds, R.C.; Secrist, J.A. 3rd; Shindo, N.;
Showe, D.N.; Sosa, M.I.; Suling, W.J.; White, E.L. Highꢀthroughput screening for inhibitors of
Mycobacterium tuberculosis H37Rv. Tuberculosis 2009, 89, 334ꢀ353.
ACS Paragon Plus Environment