2-N-Arylthiazole inhibitors of Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.bmcl.2017.07.067

To develop agents for the treatment of infections caused by Mycobacterium tuberculosis, a novel phenotypic screen was undertaken that identified a series of 2-N-aryl thiazole-based inhibitors of intracellular Mycobacterium tuberculosis. Analogs were optim

Full text of DOI:10.1016/j.bmcl.2017.07.067

Accepted Manuscript
2-N-Arylthiazole Inhibitors of Mycobacterium tuberculosis
Michael P Clark, Tiansheng Wang, Emanuele Perola, David D. Deininger,
Harmon J. Zuccola, Steven M. Jones, Hong Gao, Brian C. VanderVen, David
G. Russell, Carolyn M. Shoen, Michael H. Cynamon, John A. Thomson,
Christopher P. Locher
PII:
S0960-894X(17)30773-4
DOI:
http://dx.doi.org/10.1016/j.bmcl.2017.07.067
BMCL 25180
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
26 May 2017
23 July 2017
25 July 2017
Please cite this article as: Clark, M.P., Wang, T., Perola, E., Deininger, D.D., Zuccola, H.J., Jones, S.M., Gao, H.,
VanderVen, B.C., Russell, D.G., Shoen, C.M., Cynamon, M.H., Thomson, J.A., Locher, C.P., 2-N-Arylthiazole
Inhibitors of Mycobacterium tuberculosis, Bioorganic & Medicinal Chemistry Letters (2017), doi: http://dx.doi.org/
10.1016/j.bmcl.2017.07.067
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

2-N-Arylthiazole Inhibitors of Mycobacterium tuberculosis.
Michael P Clark*a, Tiansheng Wang*a, Emanuele Perolaa, David D. Deiningera, Harmon J. Zuccolaa, Steven M.
Jonesb, Hong Gaoa, Brian C. VanderVenc, David G. Russellc, Carolyn M. Shoend, Michael H. Cynamond, John A.
Thomsone, and Christopher P. Locher,f
a Vertex Pharmaceuticals Incorporated, 50 Northern Avenue, Boston, Massachusetts, 02210, USA.
b Contrafect Corporation, Yonkers NY 10701, USA.
c Dept of Microbiology and Immunology, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853,
USA.
d Central New York Research Corporation, Syracuse, New York 13210, USA.
e P.O. Box 2241, Acton, MA 01720, USA.
f Versatope Therapeutics, Inc., Boston, MA 02210, USA.
Abstract: To develop agents for the treatment of infections caused by Mycobacterium tuberculosis, a novel phenotypic
screen was undertaken that identified a series of 2-N-aryl thiazole-based inhibitors of intracellular Mycobacterium
tuberculosis. Analogs were optimized to improve potency against an attenuated BSL2 H37Ra laboratory strain cultivated
in human macrophage cells in vitro. The insertion of a carboxylic acid functionality resulted in compounds that retained
potency and greatly improved microsomal stability. However, the strong potency trends we observed in the attenuated
H37Ra strain were inconsistent with the potency observed for virulent strains in vitro and in vivo.
Tuberculosis (TB) is a chronic infectious disease caused by Mycobacterium tuberculosis (Mtb). It has been estimated that
as much as one third of the world’s population is infected with latent Mtb.1,2 Approximately one in ten latent infections
progresses into active tuberculosis disease.1,2 Mtb is classified as a Gram positive bacteria that has a cell wall, but lacks a
phospholipid outer membrane. Its unique biology of surviving in an intracellular state within the human macrophage
combined with its ability to reduce its metabolic rate and enter dormancy in response to drug treatment makes treating
with standard antibiotics very difficult. Without an effective vaccine, the current standard of care relies on a long regimen
of four anti-TB drugs selected from isoniazid, rifampicin, pyrazinamide, ethambutol and streptomycin for two months
followed by two of these drugs for an additional four months. This burdensome regimen is made more difficult by lack of
patient compliance and oversight. The failure of patient compliance has contributed to the rapid increase in multidrug-
resistant (MDR-TB) and extensively drug-resistant (XDR-TB) strains that have become resistant to many of the known
TB therapies.3 XDR-TB has been reported by 84 countries. An important characteristic of this infection is the formation
of granulomatous lesions, consisting of clusters of infected macrophages, T lymphocytes, B lymphocytes and fibroblasts.
It has been suggested that Mtb utilizes these granulomas to avoid destruction by the immune system of the host.4 Co-
infection with HIV as a major driver for the disease along with the rapid spread of deadly MDR and XDR strains of
Mtb reinforces the need for new anti-TB therapies.

Our objective was to utilize a unique phenotypic approach to identify screening hits with an advantage over more
traditional Mtb screening approaches, specifically to identify small molecules that kill M. tuberculosis inside its host
cell. An initial screening set of 340,000 compounds was screened in murine J774 macrophages infected with Mtb
engineered to fluoresce (mCherry)5,6 and multiple series of inhibitors were identified. By virtue of this screening
approach, identified hits offered the potential advantage of not only possessing anti-mycobacterial activity, but also
the ability to enter the macrophage and penetrate the Mtb bacterial membrane. Compound 1 was identified as a
compound of high interest due to its strong potency (IC50 < 40 nM) in the Mtb-infected macrophage assay.
Additionally, the low molecular weight (MW = 246) and structural simplicity were appealing as a starting point for
medicinal chemistry lead optimization. Interestingly, the original screening hit was thought to be thiazole
carboxylic acid 2. However, hit validation analysis of the sample used for the phenotypic screen showed that
thiazole 2 had undergone a decarboxylation over time in storage to give 4-methylthiazole 1. Both compounds were
re-synthesized and shown to be potent with a slight edge in potency for the methyl thiazole analog 1. Interestingly,
other examples of N-aryl-thiazole inhibitors of Mtb have been published recently,7,8 highlighted by compounds 3
and 4, which reinforces the utility of these thiazole-based inhibitors.
H
H
N
N
N
N
Me
S
OH
S
O
1
2
H
H
N
N
N
N
F
OMe
S
S
F
4
3
Initial SAR exploration focused on the 4-methylthiazoles derived from 1 (Table 1). An early goal for this series was
to reduce lipophilicity (cLogP for 1 = 5.7) and improve upon the poor microsomal stability of 1. Metabolite ID
studies (data not shown) with compound 1 identified the butyl side chain as the primary site for metabolism.
Compounds were designed, synthesized and then tested in H37Ra Mtb-infected macrophages as our primary assay.9
Table 1 shows our exploratory efforts to replace this side chain while retaining potency and reducing oxidative
metabolism. Initial analogs focused on various alkyl side chains. Strong potency was retained with sec-butyl,
isopropyl and tert-butyl side chains (compounds 5, 6 and 9). A reduction in potency was observed with side chains
smaller than isopropyl, exemplified by the ethyl and methyl substitutions (compounds 7 and 8). Extending the butyl
side chain to > 4 carbons retained potency (10, 11) but offered no additional advantages. Limited exploration of the

meta position of the phenyl ring was explored. Moving the n-butyl side chain to the meta-position dropped potency
more than 70-fold (macrophage IC50 = 360 nM) relative to compound 1. A more moderate 7-fold drop in potency
was observed with the ethyl analog (macrophage IC50 = 840 nM) relative to compound 7. Moving the ethyl
functionality position resulted in complete loss of potency (macrophage IC50 > 20 M). For this reason, we focused
our efforts on the para-position. Introduction of polar functionalities designed to reduce the lipophilicity of the alkyl
side chain showed good tolerance for ethers and alcohols. Ethers (13, 14, 19) showed consistently good potency,
while alcohol 12 retained potency and the 2-hydroxybutyl analog (16) only lost 10-fold in potency. The terminal
alcohol (compound 15) showed a 100-fold reduction in potency. Side chains containing a basic amine (17, 18) were
not tolerated. Despite success in identifying additional polar side chains that conferred good potency, stability as
measured in rat liver microsomes was generally poor. The one exception to this trend is compound 15, which is the
least lipophilic (cLogP = 2.0) relative to the remaining compounds with RLM data (cLogP range = 3.1 to 5.6).
Table 1. 4-methylthiazole analogs.
H
N
N
Me
S
R
RLMb
stability (%
unchanged
at 30 min)
Compd
R
Macrophagea RLMb
IC50 (M)
stability (%
unchanged
at 30 min)
Compd
R
Macrophagea
IC50 (M)
HO
<0.005
<0.040
1
0
0.060
0.150
12
4
1
5
12
13
O
F3C
HO
O
0.010
0.120
0
0
<0.040
3.19
1
6
7
14
15
73
O
OH
Me
> 20
ND
0.400
>20
1
0
8
9
16
H
N
<0.040
0.040
0
3
17
18
N
>20
ND
10
10
11
0.070
O
0.050
ND
19

aA reporter strain of Mtb expressing firefly luciferase was used to infect THP-1 cells and measure bacterial growth at 50% inhibitory
concentration at five days post-infection; brat liver microsome stability measured as % unchanged at 30 minutes (1 µM); ND = not determined
We next turned our attention to the thiazole carboxylic acids related to 2 (Table 2). Our hypothesis was the methyl
group at the 4-position of the electron rich thiazole ring might also be an easy target for CYP-mediated metabolism
and the acetic acid functionality could serve to stabilize the methyl thiazole ring system. The SAR shown in Table 2
demonstrates the high tolerance for variable substitutions at the thiazole 4-position (R3) of the thiazole ring in this
sub-series. The acetic acid analogs at the 4-position of the thiazole ring (2, 21, 22) showed similar potency in the
macrophage assay. Replacing the carboxylic acid functionality with a primary carboxamide (20) resulted in a 5-fold
decrease in potency while attaching the carboxylate directly to the 4-position of the thiazole (23) resulted in total
loss of potency. As shown in Table 2, analogs containing the carboxylic acid functionality generally showed lower
clearance and good oral exposure in mice, highlighted by acid analogs 2, 21, 22 and 24-27. Attempts to incorporate
an ether functionality into the phenyl side chain generally resulted in a loss of potency (data not shown). The best
alkyl ether side chain was compound 28, which showed good exposure in mouse, but the potency was less than
desirable.
Table 2. N-aryl-2-aminothiazole carboxylate analogs with mouse pharmacokinetic parameters
H
N
N
R3
S
R1
R2
R
R
M
Mouse PKb
1
3
acr
op
ha
gea
IC
iv Clc, T1/2 (h)
AUCd, Cmaxe
50
(
M)
OH
n-Bu
n-Bu
H
H
0.15
0.86
5, 3.1
4.9, 4.8
--
2
O
O
NH2
--
20
OH
n-Bu
H
0.14
4, 0.9
4.8, 4.5
21
O

OH
n-Bu
n-Bu
t-Bu,
H
H
H
0.14
> 20
0.07
18, 5
--
1.4, 1.8
--
22
23
24
O
O
OH
OH
4, 1.1
24.1, 4.0
O
OH
OH
n-Bu
n-Bu
Me
Cl
0.10
0.07
7, 1.1
5, 1.1
4.8, 4.1
6, 2.9
25
26
O
O
(+/-)-2-
Bu
OH
OH
H
H
0.07
0.27
3, 2.8
14.4, 6.6
15, 4.0
27
28
O
O
1.5, 1.6
aA reporter strain of Mtb expressing firefly luciferase was used to infect THP-1 cells and measure bacterial growth at 50% inhibitory
concentration at five days post-infection ; bmouse PO at 3 mg/kg, mouse IV at 1 mg/kg; cCl measured (mL/min/kg); cAUC measured (g*h/mL);
d
The thiazole analogs were prepared as shown in Scheme 1. All compounds were prepared in a similar fashion.
Starting with 4-butylaniline, conversion to the primary thiourea, 29, is followed by alkylation and subsequent
cyclocondensation with ethyl 4-chloro-3-oxobutanoate to afford the thiazole intermediate, 30. For compound 2, the final
step was ester hydrolysis under basic conditions.
H
NH2
N
NH2
a
b
S
29
H
N
H
N
N
N
c
S
OEt
S
OH
O
O
30
2
Scheme 1. Synthesis of thiazole acid 2. Reagents and conditions: (a) NH4SCN, HCl, 105 oC; (b) ethyl 4-chloro-3-
oxobutanoate, EtOH, 90 oC; (c) LiOH, MeOH.
Given the strong potency in the Mtb macrophage assay coupled with good plasma exposure in mice, we advanced
compounds 2, 24, 26 and 27 as well as the methyl thiazole analog, 1, to be evaluated in a murine model of Mtb
infection. Intranasal infections of BALB/c mice with 106 cfu of the Mtb Erdman (BSL3) isolate (six mice per group)

were performed as previously described.10 An early control (EC) group was euthanized at the initiation of therapy to
determine the infection load and a late control (LC) group was used to determine the infection load at the end of an
experiment and to confirm virulence. Mice were treated daily (5 days/week) for four weeks (100 mg/kg) and colony
forming units (cfu) were determined from Mtb recovered from lung homogenates. Rifampin (10 mg/kg) was used as
a positive control antibiotic.
Unfortunately, none of the five analogs showed a statistically significant reduction of Mtb cfu in lung homogenates
from infected mice. The lack of efficacy in the mouse model led us to further explore the lack of in-vitro to in-vivo
translation. Comparison between BSL2 broth and macrophage assays shows a strong correlation in activity (Table
3). However, upon testing these analogs in multiple BSL3 Mtb strains, we were surprised to see a substantial
reduction in potency. First, a significant drop in potency was observed for compounds 24, 26 and 27 in the BSL3
H37Rv-7H9 assay with an average reduction of 120-fold. In this case, the H37Rv-7H9 assay utilizes the same
medium as the macrophage assay with glucose and glycerol as the carbon source. An additional drop in activity
was observed when these analogs were tested in the H37Rv-7H12, Erdman 7H12 and H37Rv low oxygen recovery
(LORA) BSL3 assays. The low oxygen recovery assay duplicates the H37Rv-7H12 assay but under anaerobic
conditions.11 All three of these assays utilize a different medium with palmitic acid as the source of carbon,
suggesting that the source of carbon as well as the strain (BSL2 vs BSL3) both play a significant role in the lack of
in vitro translation. The inability of these compounds to show efficacy in the mouse model was consistent with the
limited potency observed against the virulent BSL3 strains.
Table 3. Potency of thiazole analogs in Mtb BSL2 and BSL3 strains
Compd
H37Ra
brotha
BSL-2
H37Ra
H37Rv 7H9b
BSL-3
Erdman 7H9c
BSL-3
H37Rv 7H12b
BSL-3
H37Rv LORAb
BSL-3
Erdman 7H12b
BSL-3
macrophageb
BSL2
IC50 (M)
MIC (M)
IC50 (M)
IC50 (M)
IC50 (M)
MIC (M)
IC50 (M)
0.04
0.16
0.04
0.15
ND
ND
7
47
91
12
>100
56
1
2
7.8
17.8
0.16
0.39
0.07
0.07
9.4
6.3
6.3
>100
>100
>100
5.6
63
54
24
26
12.5

0.16
0.08
0.07
0.03
9.4
--
6.3
--
>100
--
96
--
>100
--
27
rifampin
a Concentration (M) of test compound resulting in 70% inhibition of growth compared to control wells was reported as MIC in the BSL2 H37Ra
broth assay; b Reporter strain of Mtb expressing firefly luciferase was used to infect THP-1 cells and measure bacterial growth at 50% inhibitory
concentration at five days post-infection; c Concentration (M) of test compound resulting in 90% inhibition of growth compared to control
wells was reported as MIC; ND = not determined
While it has been established that H37Ra is a useful surrogate for identifying and optimizing anti-Mtb activity, it has
also been shown that false positives can sometimes be identified.10 Franzblau and Collins observed a good
correlation between the activities against H37Ra and H37Rv strains across a panel of 30 known anti-Mtb agents.12
One of the few agents in their study that did not show translation from H37Ra to H37Rv was thiacetazone 31
(TAC). Given that the SAR of the N-aryl-2-aminothiazole series suggests the 2-aminothiazole as the core scaffold
responsible for the anti-Mtb activity, we hypothesized that this series of inhibitors might undergo sulfur oxidation to
induce formation of a bioactive metabolite, similar to historic Mtb treatments, such as isoxyl (ISO), 31, and
thiacetazone (TAC), 32. Both of these compounds are known to inhibit mycolic acid biosynthesis after
bioactivation.13,14 While our aminothiazoles are not free thiocarbamides, as in 31 and 32, the thiazole functionality
does contain an embedded thiourea. In 2003, it was shown that the anti-Mtb effect of isoxyl could be partially
reversed by the addition of oleic acid.15 We carried out similar oleic acid reversal experiments with both isoxyl and
our N-aryl-2-aminothiazole inhibitors (compounds 1, 2 and 27) and confirmed partial loss of activity (4 fold for
compounds 2 and 27 and 8 fold for compound 1). In this experiment, oleic acid was able to diminish compound
inhibition at concentrations typically around 25 M. It is possible that the monooxygenase (EthA) that is
responsible for bioactivation of ISO and TAC may also be an agent for activation of our 2-aminothiazole Mtb
inhibitors.
H
N
H
H
N
N
N
S
S
O
O
1
31
H
N
NH2
O
N
S
N
H
32

Starting from a phenotypic screening effort using an intracellular (macrophage) H37Ra Mtb assay, a class of potent
2-N-arylaminothiazoles was identified. It was found that the 4-methylthiazole ring system could be stabilized
against oxidative metabolism by the introduction of an acetic acid functionality with minimal loss in potency. The
strong anti-mycobacterial activity observed in our H37Ra broth and macrophage assays did not consistently
correlate with activity against the more virulent BSL3 strains (e.g., H37Rv strain) in vitro and did not afford
protection in a mouse model when evaluated in vivo. Our data showed that the diminished potency of these
thiazoles against virulent BSL3 strains was dependent on the strain used in our in vitro assays and also the carbon
source. Preliminary experiments were conducted that showed a possible overlap of mechanism of bioactivation with
historical anti-Mtb treatments such as isoxyl and thiacetazone.
Acknowledgements.
We thank Yuzhou Xu and Jie Wang of Shanghai ChemPartner for biological assay support, Xun Tao and Min Guo of
Shanghai ChemPartner for PK assay support, and Scott Franzblau and Baojie Wang of the University of Illinois Chicago for
evaluations of MIC compounds in vitro.
References
1) WHO: http://www.who.int/mediacentre/factsheets/fs104/en/.
2) CDC: http://www.cdc.gov/tb/statistics/default.htm.
3) CDC: http://www.cdc.gov/tb/publications/factsheets/drtb/xdrtb.htm.
4) Russell, D. G., Barry, C. E., Flynn, J. L.; Tuberculosis: What we don’t know can, and does, hurt us.
Science 2010, 328, 852-856.
5) VanderVen, B.C., Fahey, R.J., Lee, W., Liu, Y. Abramovitch, R, Memmott, C, Crowe, A.M., Eltis, L.D.,
Perola, E., Deininger, D.D., Wang, T.S., Locher, C.P., Russell, D.G. Novel inhibitors of cholesterol
degradation in Mycobacterium tuberculosis reveal how the bacterium’s metabolism is constrained by the
intracellular environment. PLOS Pathog., 2015, 11, 1-20.
6) mCherry fluorescence measured in live bacteria. Several reporters were screened in Mtb-infected
macrophages, mCherry showed highest sensitivity. Fluorescent compounds were shown to not interfere
with signal in this assay.
7) Pieroni, M., Wan, B., Cho, S., Franzblau, S. G., Costantino, G.; Design, synthesis and investigation on the
structure-activity relationships of N-substituted 2-aminothiazole derivatives as antitubercular agents. Eur. J.
Med. Chem., 2014, 72, 26-34.
8) Roy, K. K., Singh, S., Sharma, S. K., Srivastava, R., Chaturvedi, V., Saxena, A.; Synthesis and biological
evaluation of substituted 4-arylthiazol-2-amino derivatives as potent growth inhibitors of replicating
Mycobacterium tuberculosis H37Rv. Bioorg. Med. Chem. Lett., 2011, 21, 5589-5593.
9) Locher, C. P., Jones, S. M., Hanzelka, B. L., Perola, E., Shoen, C. M., Cynamon, M. H., Ngwane, A. H.,
Wild, I. J., vanHelden, P. D., Betoudji, F., Nuermberger, E. L.; A novel inhibitor of Gyrase B is a potent
drug candidate for treatment of tuberculosis and nontuberculosis Mycobacterial infections. Antimicrob.
Agents Ch., 2015, 59, 1455-1465.
10) Shoen, C. M., DeStefano, M. S., Sklaney, M. R., Monica, B. J., Slee, A. M., Cynamon, M. H.; Short-course
treatment regimen to identify potential antituberculosis agents in a murine model of tuberculosis. J.
Antimicrobial Chemo., 2004, 53, 641-645.

11) Cho, S. H., Warit, S., Wan, B., Hwang, C. H., Pauli, G. F., Franzblau, S. G.; Low oxygen-recovery assay
for high-throughput screening of compounds against nonreplicating Mycobacterium tuberculosis.
Antimicrob. Agents Ch., 2007, 53, 1380-1385.
12) Collins, L. A., Franzblau, S. G.; Microplate alamar blue assay versus BACTEC 460 system for high-
throughput screening of compounds against Mycobacterium tuberculosis and Mycobacterium avium.
Antimicrob. Agents Ch., 1997, 41, 1004-1009.
13) Alahari, A., Trivelli, X., Guerardel, Y., Dover, L. G., Besra, G. S., Sacchettini, J. C., Reynolds, R. C.,
Coxon, G. D., Kremer, L.; Thioacetazone, an antitubercular drug that inhibits cyclopropanation of cell wall
mycolic acids in Mycobacteria. PLoS ONE 2007, 2, e1303.
14) Dover, L. G., Alahari, A., Gratraud, P., Gomes, J. M., Bhowruth, V., Reynolds, R. C., Besra, G. S.,
Kremer, L.; EthA, a common activator of thiocarbamide-containing drugs acting on different
Mycobacterial targets. Antimicrob. Agents Ch., 2007, 51, 1055-1063.
15) Phetsuksiri, B., Jackson, M., Scherman, H., McNeil, M., Besra, G. S., Baulard, A. R., Slayden, R. A.,
Debarber, A. E., Barry III, C. E., Baird, M. S., Crick, D. C., Brennan, P. J.; Unique mechanism of action of
the thiourea drug Isoxyl on Mycobacterium tuberculosis. J. Biol. Chem. 2003, 278, 53123-53130.

H
N
H
N
N
N
Me
S
S
OH
O
(24)
(1)