Evaluation of NTF1836 as an inhibitor of the mycothiol biosynthetic enzyme MshC in growing and non-replicating Mycobacterium tuberculosis

Article abstract of DOI:10.1016/j.bmc.2011.05.028

The mycothiol biosynthesis enzyme MshC catalyzes the ligation of cysteine with the pseudodisaccharide GlcN-Ins and has been identified as an essential enzyme in Mycobacterium tuberculosis. We now report on the development of NTF1836 as a micromolar inhibi

Full text of DOI:10.1016/j.bmc.2011.05.028

Bioorganic & Medicinal Chemistry 19 (2011) 3956–3964
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Evaluation of NTF1836 as an inhibitor of the mycothiol biosynthetic enzyme
MshC in growing and non-replicating Mycobacterium tuberculosis
⇑
Gerald L. Newton, Nancy Buchmeier, James J. La Clair, Robert C. Fahey
Department of Chemistry and Biochemistry, University of California, San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0314, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The mycothiol biosynthesis enzyme MshC catalyzes the ligation of cysteine with the pseudodisaccharide
GlcN-Ins and has been identified as an essential enzyme in Mycobacterium tuberculosis. We now report on
the development of NTF1836 as a micromolar inhibitor of MshC. Using commercial libraries, we con-
ducted preliminary structure–activity relationship (SAR) studies on NTF1836. Based on this data,
NTF1836 and five structurally related compounds showed similar activity towards clinical strains of M.
tuberculosis. A gram scale synthesis was developed to provide ample material for biological studies. Using
this material, we determined that inhibition of M. tuberculosis growth by NTF1836 was accompanied by a
fall in mycothiol and an increase in GlcN-Ins consistent with the targeting of MshC. We also determined
that NTF1836 kills non-replicating M. tuberculosis in the carbon starvation model of latency.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 11 April 2011
Revised 11 May 2011
Accepted 17 May 2011
Available online 24 May 2011
Keywords:
Redox thiols
Mycothiol
Biosynthesis
Tuberculosis
Antibiotic discovery
1. Introduction
produce a different low-molecular-weight thiol, mycothiol (MSH,
1, Fig. 1). In M. tuberculosis, MSH (1) acts in a similar manner as
Tuberculosis (TB), resulting from infection by Mycobacterium
tuberculosis, continues to be a serious health concern. In 2009,
there were an estimated 9.4 million cases worldwide, resulting in
an estimated 1.7 million deaths.1 Current estimates suggest that
one-third of the world population harbors latent TB that emerges
upon immune system compromise. TB continues to be a major
cause of death among HIV-positive individuals.2 Currently, the
six to nine month multidrug protocol used in the treatment of TB
is highly effective with drug-susceptible TB but deviant protocols
and poor patient compliance promotes development of drug-
resistance.3 The emergence of multidrug-resistant and extensively
drug-resistant TB has exacerbated the problem of tuberculosis
control.4 There is a recognized need for new faster acting drugs
that are effective against drug-resistant TB while remaining
compatible with other TB and HIV drugs, as they are often used
in combination.5 Ideally, there is an immediate need for drugs that
can terminate the non-replicating bacilli of latent TB.6 After a long
lull in the introduction of new TB candidates a number of promis-
ing leads are now in clinical trials.5,6 However, the drug pipeline
needs continual expansion in order to successfully meet the chal-
lenges of treating drug-resistant and latent infections.5
GSH.7,8 The biosynthesis of MSH arises from UDP-GlcNAc and
1L-myo-inositol-1-phosphate (Ins-P) by five enzymes as outlined
in Scheme 1.9 Analogous to GSH, MSH is involved in detoxification.
It is responsible for the removal of select antibiotics (RX, Scheme 1)
through the production of mycothiol-S-conjugates (MSR) that are
cleaved by mycothiol-S-conjugate amidase (Mca). This process
generates a mercapturic acid (AcCySR, Scheme 1) that can be
excreted. The resulting GlcN-Ins is used to regenerate MSH.10 The
mycothiol deacetylase (MshB),11 and Mca,10,12 are homologous
enzymes and have overlapping activities.
MshB was the first mycothiol biosynthesis enzyme identified.13
A mshB gene knockout in M. tuberculosis produced MSH, albeit at a
lower level than wild-type cells during exponential growth, and
exhibited increased sensitivity to rifampin and cumene hydroper-
oxide.14 Since Mca also displays deacetylase activity, it may
contribute to MSH production in the absence of MshB.12 Mycobac-
terium smegmatis mutants defective in MshC have been isolated
that are devoid in MSH production,15 however, attempts to
produce a mshC knockout in M. tuberculosis were unsuccessful
unless a second copy of the gene was present.16 This evidence
suggested that MshC was essential in M. tuberculosis.
Metabolic pathways unique to bacteria, and their associated en-
zymes, offer excellent targets for drug development. Mycobacteria
are unusual in that they do not synthesize glutathione (GSH) but
Based on this evidence, MSH biosynthesis has been suggested as
a target for therapeutic development for TB. Compounds that inhi-
bit both MshB and Mca have been synthesized and display activi-
ties in the low micromolar range,17 but have so far failed to
ablate MSH biosynthesis. No inhibitors of MshA or MshD have been
reported and the phosphatase remaining enzyme, MshA2, has yet
⇑
Corresponding author.
E-mail address: rcfahey@ucsd.edu (R.C. Fahey).
0968-0896/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2011.05.028

G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
3957
and then partially-proteolyzed with trypsin, crystallization was
possible. In 2008, a 1.6 Å crystal structure was reported indicating
that residues 289–293 containing the conserved KMSKS motif had
been proteolyzed.22 The structure identifies interactions of CSA
with the protein, including binding of the thiolate group to the ac-
tive site zinc residue.
The first MSH biosynthesis inhibitor identified, NTF1836 (2),
was found in a screen for inhibitors of the native M. tuberculosis
MshC.18 Recently, a screen using M. tuberculosis MPB-MshC identi-
fied two quinaldinium derivatives 3 and 4 (Fig. 1) as micromolar
inhibitors of MshC.23 We now report on the optimization, synthesis
and biological evaluation of 2. These studies validate MshC as the
primary target of 2 in M. tuberculosis as well as demonstrate effi-
cacy of 2 against non-replicating M. tuberculosis using the phos-
phate buffered saline (PBS) carbon starvation model of
dormancy.24
OH
O
HO
HO
mycothiol
HN
O
OH
OH
MSH (1)
HO
OH
OH
HS
O
NHAc
O
S
O
NTF1836
N
(2)
O
N
NH
Cl
Cl-
Cl-
+
+
3
N
N
8
2. Results and discussion
H2N
NH2
2.1. Structure–activity relationship (SAR) studies
NO2
H
N
Our studies began by examining the structure–activity relation-
ship (SAR) requirements of 2. This began by obtaining a panel of 39
analogs, compounds 5–43 (Tables 1 and 2) that contained modifi-
cations within three positions R1, R2 and X.
O
H
N
N+
Cl-
4
N
+N
H
Cl-
NO2
Two motifs were examined. The first (motif A, Table 1) focused
on variation with the R1 and X groups. The presence of a tri-substi-
tuted amine at the terminus of R1 played a key role in the activity.
Replacement of the amine with a sulfide as given in the compari-
son of 2 with 6 or 7 with 8 (Table 1) led to a complete loss of mea-
surable activity. All of the compounds exhibiting measurable
inhibition contained this amine terminus in R1. The oxidation state
of the sulfur atom X was also critical to activity. Sulfoxides (X = SO)
were 5- to 15-fold better inhibitors than their corresponding sulf-
ones (X = SO2) as evident in the comparison of 7 with 9, 10 with 11,
12 with 13 or 14 with 16 (Table 1). All of the sulfoxides employed
in this study were racemic mixtures of the two sulfoxide enantio-
mers and the reported IC50 values are composites of the contribu-
tion by each individual enantiomer. Only a limited number of the
corresponding sulfides (X = S) were sufficiently soluble for study,
however those that were, were less active than their corresponding
sulfoxides (compare 5 with 2, 15 with 14 or 18 with 17, Table 1). In
contrast, there was only a minor modification by the choice of the
terminal alkyl groups in R1. For instance, the activity of compounds
containing common 3 carbon linked amine at R1 and sulfoxide at X
including 7, 12, 17, 19, and 21 displayed IC50 values that were
within a factor of 1.5 from that for 2. Compound 10 was less effec-
tive, possibly reflecting steric limit of the motif. Finally, compound
14 was similarly ineffective which may reflect the effect of the
ether oxygen on the ionization state of the amino group.
Figure 1. Structures of mycothiol (MSH, 1) and MshC inhibitors NTF1836 (2),
quinaldinium salt 3 and quinaldinium salt 4.
to be identified. On the other hand, MshC has been the most stud-
ied enzyme involved in MSH biosynthesis. Based on this, we fo-
cused our efforts to identify inhibitors of MshC.
Efforts to study M. tuberculosis MshC have been hindered by dif-
ficulties in efficient purification of the active cloned protein via
conventional His6 affinity chromatography. The native untagged
M. tuberculosis protein was successfully cloned and expressed in
M. smegmatis but the purification was tedious.18 Successful expres-
sion and affinity purification of three fusion protein constructs of
M. tuberculosis MshC in M. smegmatis has been reported.19 To date,
the best results have been obtained using a maltose binding pro-
tein (MBP) fusion to generate low milligram quantities of the
tagged M. tuberculosis MBP-MshC. Expression and purification of
active His6-tagged M. smegmatis MshC in Escherichia coli is less
troublesome and high milligram amounts of this protein have been
generated.20
Steady state kinetic studies20 and positional isotopic exchange
analysis21 of M. smegmatis MshC were consistent with a Bi-Uni
Uni-Bi ping-pong mechanism in which GlcN-Ins binds after
formation of the bound cysteinyl-adenylate intermediate. Crystal-
lization efforts with M. smegmatis MshC have so far been unsuc-
cessful.22 However, when the protein was treated with the
The effect of the side chain group R2 was also examined within
the second motif (motif B, Table 2). Removal of the chloro-
substituent from R2 increased the IC50 value about fourfold in the
potent inhibitor 50-O-[N-(
L-(cysteinyl)-sulfamoyl]adenosine (CSA)
Scheme 1. Biosynthesis of mycothiol (AcCys-GlyN-Ins, MSH) arises through five steps catalyzed by the biosynthetic enzymes: MshA, MshA2 MshB, MshC and MshD. Once
formed, MSH can react with select antibiotics (RX) to form mycothiol-S-conjugates (MSR). Cleavage of MSR by mycothiol-S-conjugate amidase (Mca) releases GlnN-Ins, which
is recycled, and an equivalent of antibiotic modified as a mercapturic acid (AcCysSR).

3958
G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
Table 1
Table 2
Variation of IC50 values for inhibition of MshC with modifications at X and R1 in Motif
A
Variation of IC50 values for inhibition of MshC with modifications at X, R1, and R2 in
motif B
X
X
R1
R1
N
N
R2
O
O
O
O
Cl
Motif B
Motif A
No.
R1
R2
X
IC50 (mM)
No.
R1
X
IC50 (mM)
H
N
N
N
N
H
N
2
Cl
SO
0.085 0.010
N
N
S
2
SO
0.085 0.010
H
N
H
N
25
26
27
28
SO
0.3 0.1
0.5 0.1
1.7 0.4
>3
5
6
S
0.28 0.03
H
N
H
N
SO2
SO2
SO
SO
P1a
H
N
H
N
7
8
9
SO
0.11 0.02
>3a
N
S
N
S
H
SO
N
H
N
H
SO2
1.7 0.3
N
N
10
11
SO
0.50 0.06
2.3 0.7
H
N
N
N
H
N
29
30
31
SO
SO
SO
>3
N
SO2
H
H
N
P3
>3
N
N
N
O
H
N
12
13
14
15
16
SO
SO2
SO
S
0.08 0.02
1.1 0.3
0.5 0.1
0.8 0.1
ꢀ3
H
N
N
N
H
N
H
N
N
32
33
34
35
36
37
SO
0.34 0.02
0.33 0.04
0.25 0.10
0.71 0.02
>3
O
O
O
H
N
N
Cl
H
H
N
N
N
N
N
SO2
SO2
SO
H
N
SO2
Cl
Cl
F
H
H
N
17
18
SO
S
0.10 0.03
0.13 0.03
N
N
N
H
N
N
H
N
19
20
SO
SO
0.15 0.03
0.10 0.02
N
H
N
N
H
N
N
H
N
N
N
H
N
21
22
SO
S
0.10 0.05
0.32 0.05
N
SO
F
H
N
N
H
N
N
SO
>3
23
24
SO2
SO2
>3
F
NH
H
N
F
F
2.5 0.5
H
N
N
38
39
SO2
SO
0.8 0.1
>3
a
From measurements at 50 lM and 100 lM.
H
N
N
N
sulfoxide series (compare 2 with 25, Table 2) while comparable ef-
fects were not seen in the corresponding sulfones (compare 25
with 26, Table 1). This observation suggests that the effects of X,
R1 and R2 may not be cumulative.
While the trends observed in Table 1 were maintained, the sec-
ond panel (Table 2) provided additional SAR data. For instance, re-
moval of the aromatic ring in R2 as in compounds 29–31 led to
complete loss in activity. The set with R1 = 4-chlorobenzyl 32–34
displayed very similar activities, being about three- to fourfold less
effective than 2. Here again, we observed a non-cumulative effect,
as analogs 33–34 were screened as sulfones (X = SO2). From
H
N
40
41
42
SO
ꢀ2
H
N
N
N
SO
>3
H
N
SO2
1.4 0.2

G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
3959
OMe
O
Table 2 (continued)
O
OMe
F
a
No.
R1
R2
X
IC50 (mM)
>3
S
HS
+
O
NO2
O
O
H
N
NO2
46
OMe
45
N
43
SO2
OMe
44
b
Table 1, one would expect that these materials would be 5- to
15-fold less active than 32, however this was not the case
(compounds 32–34 shared comparable activity).
S
O
OMe
c,d
O
S
N
H
Further screening, indicated that the 4-fluorobenzyl group in 35
was less active (fivefold) than the corresponding 3-chlorobenzyl
derivative 19. A larger (30-fold) loss was obtained when comparing
derivatives 36 and 21. Substitution of 3-chlorobenzyl by 2-fluorob-
enzyl in 37 resulted in a >35-fold loss in activity when compared to
2. Incorporation of the 2,5-dimethylbenzyl group as R1 in 41 and 43
resulted in no measurable inhibition whereas for 40 and 42 it led to
just detectable inhibition.
While we have yet to identify a substantially more active analog
than 2, these SAR studies indicate a clear reliance on the inclusion
of a sulfoxide (X = SO) and presence of a amine functionality at the
terminus of R1 and an aryl group in R2. Fortuitously, our initial
screening efforts returned an analog whose activity was well de-
fined. While these studies do not preclude the development of
more active analogs, via more refined analog screening efforts,
the observed SAR indication provide strong support for further
evaluating NTF1836 (2). We turned our efforts to critically evaluate
our ability to prepare 2, validate its mode of action and examine its
potential at treating resistant and non-replicating TB infections.
O
N
O
NH2
47
OMe
N
48
e
S
N
S
g
O
O
O
OMe
N
H
O
Cl
OR
f
51
49 R=H
50 R=Me
h,i
O
S
S
N
O
O
N
O
j
O
NH
NH
N
Cl
Cl
N
2.2. Chemical synthesis of NTF1836 (2)
52
2
While our SAR studies indicated a clear motif, two concerns re-
mained. First, we had not independently verified the structure of 2
(compounds were commercially obtained and screened) and there-
fore we wanted to prepare synthetic material in house as a means
of structure validation. Second, our planned biological studies re-
quired access to gram quantities of 2. Based on these concerns,
we developed an effective synthetic route to NTF1836 (2). After
examining several approaches, the route depicted in Scheme 2 pro-
vided facile access to gram quantities of 2 in nine steps from com-
mercial materials.
The synthesis began with a SNAr nucleophilic substitution to
couple thiol 44 with arylfluoride 45 and provides yellow needles
of mercaptan 46 after recrystallization. Reduction of 46 provided
amine 47 in quantitative yield, which was, without purification,
converted to thiazepinone 48 through a three-step process that be-
gan with hydrolysis of two methyl esters. Treatment of the result-
ing diacid with 1,10-carbonyldiimidazole resulted in cyclization to
acylimidazole intermediate 48, which upon workup with 4-
dimethylaminopyridine (DMAP) in dry MeOH afforded methyl es-
ter 50 with traces of acid 49. After optimization, purification using
a combination of Dry Column Vacuum Chromatography (DCVC)
and recrystallization provided gram quantities of 50 as a pure
white powder. As of note, acid 49 obtained in this process could
be converted to 50 by treatment with trimethylsilyldiazomethane.
At this stage ester 50 was ready for incorporation of the R2
group. The protection of carboxylic acid 49 as its methyl ester 50
was key to facilitating this transformation. First, this conversion
provided an effective means of purification as compound 50 was
readily purified by recrystallization (the corresponding acid 49
could not be recrystallized). Second, it provided a more cost effec-
tive conversion to 2 by reducing the amount of 3-chlorobenzylbro-
Scheme 2. Gram-scale synthesis of NTF1836 (2). Reagents and conditions: (a)
Cs2CO3, DMF, 87%; (b) PtO2, Pd–C, EtOAc, MeOH, 98%; (c) LiOH, THF, H2O, rt; (d) 1,1-
carbonydimidazole, DMF; (e) DMAP, MeOH, 83% over three steps from 47; (f)
trimethylsilyldiazomethane, THF, rt, 1 h; (g) NaH, 3-chloro-benzylbromide, DMF,
0 °C to rt, 79%; (h) LiOH, THF, H2O, rt; (i) N-1-cyclohexyl-N-1-methylpropane-1,3-
diamine, HATU, EtN(iPr)2, DMF, rt, 16 h; 81% over two steps from 49; (j) m-CPBA,
NaHCO3, CH2Cl2, 65%.
After screening, we identified an effective three-step procedure
to install R1 and R2. This began by treating 50 with NaH and 3-chlo-
robenzylbromide. The resulting methyl ester 51 was then hydro-
lyzed, treated with HATU, and coupled to N-1-cyclohexyl-N-1-
methylpropane-1,3-diamine to deliver sulfide 52. The latter pro-
cess was optimally conducted in one operation. After screening a
variety of conditions, the oxidation of sulfide 51 to the desired sulf-
oxide 2 was most effectively conducted by treatment with NaHCO3
buffered m-CPBA. Using this procedure, we were able to provide
gram quantities of 2 (detailed synthetic procedures are provided
in Section 4.2) with a 29% overall yield from 45 using a process that
required the use of one DCVC column and two Flash chromato-
graphic purifications.
2.3. NTF1836 (2) kills growing M. tuberculosis
The sensitivity of actively growing M. tuberculosis to 2 was ini-
tially tested with the Erdman strain. A culture of mid-log phase M.
tuberculosis Erdman was diluted to an OD600 of 0.08 and aliquots
were incubated at 37 °C for 7 d with dilutions of 2. In three exper-
iments, 2 inhibited the growth of M. tuberculosis in a dose depen-
dent manner as measured by both change in optical density and
by plating for viable bacteria (Fig. 2). Inhibition of growth was
mide
and
N-1-cyclohexyl-N-1-methylpropane-1,3-diamine
required in subsequent steps.
apparent at 20 lM 2 and increased at higher concentrations.

3960
G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
of GlcN-Ins (Fig. 3). Since 2 was toxic to M. tuberculosis in this con-
centration range (Fig. 2), the decrease in MSH level might be due to
cell death rather than inhibition of MshC. However, the accumula-
tion of GlcN-Ins was expected for MshC inhibition and was unex-
pected in killed cells. The pattern of GlcN-Ins and MSH changes
with increasing concentration of 2 was consistent with inhibition
at the MshC step of mycothiol biosynthesis and validates MshC
as a target for 2 (Scheme 1).
2.5. NTF1836 (2) kills nonreplicating M. tuberculosis
In order to examine the effect of 2 on nonreplicating M. tubercu-
losis, we used a nutrient starvation model to generate M. tuberculo-
sis in the nonreplicating state.24 Bacteria were cultured for 6 weeks
in PBS at which time dilutions of 2 were added and the cultures
were incubated for an additional 15 d. Samples were then assessed
for viable bacteria by plating. There was a 99% loss in viability at a
Figure 2. Inhibition of exponentially growing M. tuberculosis Erdman as monitored
by a 7-day change in optical density at 600 nm (d; initial value 0.08) and change in
viability (cfu day 7/cfu day 0; s) as a function of the concentration of 2. Error bars
show mean and standard deviation of three determinations.
dose of approximately 90 lM 2, indicating that 2 was capable of
killing nonreplicating M. tuberculosis (Fig. 4). Two other experi-
ments produced similar degrees of killing of the nonreplicating
bacteria.
A greater sensitivity to 2 was noted in the plating results with 50%
2.6. Mycothiol biosynthesis as a drug target in M. tuberculosis
loss of viability occurring at 20
OD600 occurred between 30 M and 50
tions of 2 (50 or 75 M) the plating results demonstrated a 97%
loss in viability compared with the zero drug control, the viable
cells amounting to only one-third of those present in the day 0
inoculum.
l
M 2 whereas a 50% decline in
l
l
M 2. At higher concentra-
The essentiality of MSH and MSH biosynthesis in M. tuberculosis
is controversial and likely depends on the specific growth condi-
tions. The earliest data was obtained with saturating transposon
mutagenesis, which indicated that mshC and mtr (Rv2855, mycoth-
iol disulfide reductase) are essential for in vitro growth of M. tuber-
culosis H37Rv,26 indicating that MSH itself was essential. We have
constructed two MSH biosynthesis mutants in M. tuberculosis
l
Using a 96 well plate format, the capacity of 2 to inhibit patient
isolates of M. tuberculosis was analyzed using four patient strains. A
95% inhibition was observed with a 110 lM concentration of 2 (Ta-
ble 3). The activity of five other analogs of 2 was also tested and 4
of the 5 analogs (12, 18, 19 and 20 inhibited the growth M. tuber-
culosis at similar dosages to that of 2 (Table 3). Toxicity toward
Vero cells was determined as an indicator of potential human tox-
icity (Table 3). The results indicate that these compounds were as
toxic to mammalian cells as to M. tuberculosis. The MIC95 values for
these compounds were more than two orders of magnitude higher
than that of rifampin.
2.4. NTF1836 (2) inhibits MSH biosynthesis in M. tuberculosis
If NTF1836 (2) functions by inhibiting MshC, addition of 2 to M.
tuberculosis should prevent the production of Cys-GlcN-Ins and
MSH, and lead to the accumulation of GlcN-Ins in cells (Scheme 1).
In order to test this, samples were prepared from M. tuberculosis
Erdman cultured for 7 d with varying dilutions of 2 and the cells
were analyzed for MSH and GlcN-Ins content. The cellular level
of Cys-GlcN-Ins was below detectable limits in M. tuberculosis
Erdman25 and was not quantitated. In two independent experi-
Figure 3. Concentration of MSH (d) and GlcN-Ins (s) following 7 day culture of M.
tuberculosis Erdman in Middlebrook 7H9 medium containing varying concentra-
tions of 2. Error bars show mean and range of duplicate samples.
ments, we observed that at 20–75 lM 2 resulted in a fourfold de-
cline in MSH level accompanied by a ninefold increase in the level
Table 3
Concentration (lM) of compounds producing 95% inhibition (MIC95) of M. tuberculosis human isolates, or 90% (MIC90) Vero cell toxicity
M. tuberculosis isolate
2
12
17
18
19
20
Rifampin
CDC1551
110
110
110
110
28–55
50
32
110
55
55
55
220
220
220
220
55
110
110
110
110
55
55
55
55
55
55
100
110
110
110
110
28
—
—
—
—
0.24
—
VA1215a
UCSD14768Ja
VA1345a
Erdman (control strain)
Vero Cell toxicityb (MIC90
)
100
50
50
100
a
Two passages in vitro after isolation from patient.
Vero Cell toxicity determined at 48 h using the neutral red assay.
b

G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
3961
general synthetic route described can provide the basis for produc-
tion of a second generation of MshC inhibitors based upon 2 and
structural modeling based upon the MshC structure22 may provide
the basis for design of new MshC inhibitors.
4. Experimental
4.1. General methods
Samples (1–10 mg) of compounds 2 and 6–43 were purchased
from ChemDiv Inc. for screening purposes. Conventional condi-
tions were used for the chemical synthesis of 2. Unless stated
otherwise, reactions were performed in flame-dried glassware un-
der a positive pressure of Ar using freshly distilled solvents. Tetra-
hydrofuran (THF), CH2Cl2, MeOH, and toluene were freshly purified
using a Seca Solvent System (GlassContour, Laguna Beach, CA) and
transferred under an Ar atmosphere. EtN(Ni-Pr)2 and Et3N were
distilled from ninhydrin, dried with Na2SO4, redistilled from Na,
and stored under Ar over 4 Å mol sieves. Thin-layer chromatogra-
phy (TLC) was accomplished using EMD 60 F254 pre-coated plates
(0.25 mm). Visualization was performed by UV absorbance (254 or
365 nm), iodine on silica gel, or by charring after treatment with p-
anisaldehyde or phosphomolybdic acid. All solvent mixtures were
reported as (v/v) unless noted otherwise. Flash chromatography
and dry column vacuum chromatography (DCVC) were performed
Figure 4. Viability of nonreplicating, persistent M. tuberculosis Erdman generated
by 6 week incubation in PBS followed by addition of 2 to various concentrations,
subsequent incubation for an additional 15 d, and determination of viability by
plating. Error bars show mean and range of duplicate samples.
Erdman, both of which were very growth inhibited unless exoge-
nous catalase was supplied.14,25 The mshB and mshD mutants did
produce limited amounts of MSH or related compounds and were
not complete MSH knockout mutants. Studies of M. smegmatis
indicated that mshA and mshC mutants would be complete MSH
knockout mutants if produced in M. tuberculosis. We were unable
to produce an mshC mutant in M. tuberculosis Erdman unless a sec-
ond copy of the gene was present.16 More recently Jacobs and col-
leagues have produced an mshA mutant in M. tuberculosis which is
reported to produce no MSH and requires exogenous catalase for
growth, indicating a sensitivity to oxygen.27 Paradoxically, despite
the in vitro growth fault this mshA mutant was not attenuated for
growth in mice. Recently, a MshA mutant in M. smegmatis was
found to produce very large amounts of organic hydroperoxide
reductase (Ohr) and also to produce ꢀ20-fold higher levels of ergo-
thioneine.28 A similar overproduction of ergothioneine was found
in other MSH mutants in M. smegmatis. Although Ohr is not en-
coded by the M. tuberculosis genome the ergothioneine biosynthe-
sis genes are present and enhanced production of ergothioneine
could partially compensate for the loss of MSH in an M. tuberculosis
mshA mutant.
This study provides chemical support for the essentiality of
MshC and mycothiol in M. tuberculosis. MshC inhibitor 2 was found
to kill actively growing and non-replicating M. tuberculosis in the
same concentration range as the inhibition of MshC. Inhibitor 2
was also shown to specifically inhibit MshC in a concentration
dependent manner in actively growing M. tuberculosis although
other essential targets for 2 cannot be ruled out. This study indi-
cates that MshC is a druggable and essential protein target for M.
tuberculosis.
using EMD silica gel with 40–63
lm particle size, 40 or 60 Å pore
size and 25–40 m particle size, 60 Å pore size, respectively.
l
NMR spectra were recorded on JOEL ECA 500 or Varian VX
500 MHz instruments. Chemical shifts were reported in parts per
million (ppm) on d scale relative to solvent CDCl3 (d 7.26 ppm),
CD3OD (d 4.89 ppm), or D2O (d 4.80 ppm) for 1H and CDCl3 (d
77.00 ppm) or CD3OD (d 49.15 ppm) for 13C. Data for 1H NMR were
reported as follows: chemical shift, multiplicity (s = singlet,
d = doublet, t = triplet, q = quartet, m = multiplet, br = broad), inte-
gration, and coupling constants. All 13C NMR spectra were recorded
with complete proton decoupling. When applicable, 1D spectra
were recorded using the standard protocols with proton integra-
tions measured at decreased 30° pulse angles and increased 5 s
relaxation times. Spectral assignments were confirmed by homo-
nuclear COSY, gradient gCOSY and heteronuclear gradient HSQC
experiments. Spectra were displayed and plotted using MestraNo-
va (MestreLab Research). Low-resolution mass spectral analyses
were conducted by LC–MS on a Waters model 600 HPLC with a
Waters model 486 UV–vis detector and a Waters-Micromass
ZMD monoquad mass detector. High-resolution mass spectral
analyses were conducted on a ThermoFinnigan MAT900XL with
reference to perfluorokerosene. Biological methods and materials
were denoted within each study.
4.2. Chemical synthesis of NTF1836 (2)
4.2.1. Methyl-4-((2-(methoxycarbonyl)phenyl)thio)-3-nitro-
benzoate (46)
3. Conclusion
A mixture of methyl 4-fluoro-3-nitrobenzoate 45 (14. 1 g,
0.0706 mol) and methyl 2-mercaptobenzoate 44 (10 mL,
0.0727 mol) was dissolved in 120 mL of anhydrous DMF. Finely
ground Cs2CO3 (30.9 g, 0.950 mol) was added at rt. After mixing
for 30 min, the suspension was stirred at 40 °C for 2 h. After cooling
to rt, the mixture was diluted with CH2Cl2 (200 mL) and washed
with water (50 mL) and brine (50 mL) and dried (Na2SO4). The
crude product was recrystallized from EtOAc/hexanes to afford
21.42 g (87%) of long yellow needles 46, mp 177–178 °C; 1H NMR
(500 MHz, CDCl3) d 9.22 (s, 1H); 8.12 (dt, J = 3.8, 5.2 Hz, 1H); 8.09
(d, J = 10.6 Hz, 1H); 7.71–6.59 (m, 3H), 6.81 (d, J = 10.7 Hz, 1H),
3.08 (s, 3H), 2.89 (s, 3H); 13C NMR (125 MHz, CDCl3) d 189.1,
A survey of compounds structurally related to 2 identified sev-
eral features important for the activity of 2 to inhibit MshC but did
not identify inhibitors with substantially greater potency. A chem-
ical synthesis of 2 was reported which provided material for more
detailed testing. The present study establishes that 2 was lethal to
both growing and non-replicating persistent M. tuberculosis. Cell
killing by 2 was accompanied by a fall in cellular MSH level and
an increase in the cellular level of GlcN-Ins, the substrate of MshC,
supporting the proposed mechanism of action. While the present
results support the validity of MshC as a target, a more potent
and less toxic inhibitor of MshC than 2 was clearly required. The

3962
G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
187.0, 162.1, 161.3, 152.8, 150.9, 147.5, 147.0, 145.1, 144.4, 143.9,
142.7, 140.1, 139.5, 41.0, 40.9; HREIMS [M+Na]+ m/z: 370.0389
(calcd for C16H13NNaO6S, 370.0361).
mide (12.79 g, 0.0622 mmol) in DMF (40 mL) was added drop wise
over 30 min. The reaction was then warmed over 2 h to 0 °C, stirred
at 0 °C for 3 h, warmed to rt over 2 h, then stirred at rt for 12 h.
Care was taken to stir vigorously and warm the reaction slowly.
The reaction was cooled to 0 °C and terminated by the addition
of 5% aq MeOH (100 mL). The mixture was then filtered with EtOAc
(800 mL) through DCVC column charged with 100 g of Celite
(packed at top of column) and 150 g of silica gel (packed at bottom
of column). The eluted material was concentrated on a rotary evap-
orator and recrystallization from toluene to afford 16.21 g (79%) of
51 as a white powder. Mp 168–171 °C, 1H NMR (500 MHz, CDCl3) d
8.00 (d, J = 1.7 Hz, 1H); 7.76 (dd, J = 1.7, 7.6 Hz, 1H); 7.73 (ddd,
J = 0.6, 1.8, 8.1 Hz, 1H), 7.63 (d, J = 8.1 Hz, 1H), 7.44 (m, 1H), 7.41
(br s, 1H), 7.35 (dddd, J = 0.6, 1.6, 5.2, 12.8 Hz, 1H), 7.32 (dddd,
J = 0.6, 1.6, 7.5, 14.8 Hz, 1H), 7.27 (m, 2H), 7.20 (m, 2H), 5.80 (d,
J = 15.2 Hz, 1H), 4.91 (d, J = 15.2 Hz, 1H), 3.89 (s, 3H); 13C NMR
(125 MHz, CDCl3) d 169.9, 165.8, 143.5, 141.8, 138.9, 138.3,
137.7, 134.6, 133.6, 131.8, 131.7, 131.4, 131.4, 129.9, 129.2,
128.3, 127.8, 127.5, 126.8, 126.3, 54.2, 52.7; HREIMS [M+Na]+ m/
z: 432.0527 (calcd for C22H16ClNNaO3S, 432.0437).
4.2.2. Methyl 3-amino-4-((2-(methoxycarbonyl)phenyl)thio)-
benzoate (47)
PtO2 (0.3 g) and Pd on carbon (5.0 g) were suspended in a mix-
ture of compound 46 (21.42 g, 0.0669 mmol) in EtOAc (300 mL)
and MeOH (300 mL). The flask was degassed, charged with H2
and stirred under a constant atmosphere of H2 (ballon of H2). After
12 h, the reaction was complete as evident by TLC analyses. The
reaction degassed, filled with N2 and filtered through a 200 g plug
of silica gel eluting with 1:10 MeOH/EtOAc (600 mL). The elutant
was concentrated on a rotary evaporator to afford 19.12 g (98%)
of an off-white powder 47, mp 92–95 °C (dec), 1H NMR
(500 MHz, CDCl3)
d 8.19 (dd, J = 1.9, 9.8 Hz, 1H); 7.56 (d,
J = 10.0 Hz, 1H); 7.51 (d, J = 2.1 Hz, 1H), 7.42 (dd, J = 2.2, 1.0 Hz,
1H), 7.24 (m, 1H), 7.10 (dd, J = 5.1, 14.2 Hz, 1H), 6.55 (d,
J = 10.0 Hz, 1H), 3.70 (br s, 2H, NH), 3.12 (s, 3H), 3.06 (s, 3H); 13C
NMR (125 MHz, CDCl3) d 189.5, 189.4, 167.4, 156.3, 153.2, 147.1,
146.8, 145.3, 139.0, 138.3, 136.7, 130.1, 129.7, 125.9, 46.3; HREIMS
[M+H]+ m/z: 318.0870 (calcd for C16H16NO4S, 318.0800).
4.2.5. 10-(3-Chlorobenzyl)-N-(3-(cyclohexyl(methyl)amino)-
propyl)-11-oxo-10,11 dihydrodibenzo[b,f][1,4]thiazepine-8-
carboxamide (52)
4.2.3. Methyl 11-oxo-10,11-dihydrodibenzo[b,f][1,4]-thiaze-
pine-8-carboxylate (50)
Ester 51 (16.12 g, 0.0393 mol) was dissolved in THF (170 mL)
containing H2O (30 mL). After cooling to 0 °C, LiOHꢁH2O (4.95 g,
0.117 mol) was added in one portion. After 2 h at 0 °C, the reaction
was warmed to rt. HPLC or TLC indicated that hydrolysis was com-
plete after 23 h at which point the pH was adjusted to 7.0 by the
addition of 2 M HCl. The remaining mixture was dried by sequen-
tial rotary evaporation followed by lyophilization with MeOH
(6 ꢂ 100 mL) resulting in yellow wax. The wax was further dried
by rotary evaporation of toluene (3 ꢂ 120 mL). The resulting mate-
rial was suspended in EtN(iPr)2 (20.55 mL, 0.117 mmol) in DMF
(150 mL). After cooling to 0 °C, HATU (22.42 g, 0.0590 mol) was
in three portions over 15 min. The reaction was warmed to rt over
1 h. After an additional 2 h at rt, N-1-cyclohexyl-N-1-methylpro-
pane-1,3-diamine (7.00 g, 0.0411 mmol) was added in 10 mL
DMF. After 18 h of vigorous stirring at rt, the mixture was diluted
in EtOAc (200 mL) and filtered with a mixture of 1:5 MeOH/EtOAc
(1 L) through a DCVC column charged with 100 g of Celite (packed
at top of column) and 100 g of silica gel (packed at bottom of col-
umn). The elutant was concentrated on a rotary evaporator. Flash
chromatography (2:1 Hx/EtOAc to 10:1 MeOH/EtOAc) afforded
5.68 g (81%) of clear wax 52. 1H NMR (500 MHz, CD3OD) d 7.96
(d, J = 1.9 Hz, 1H), 7.65 (m, 2H); 7.56 (dd, J = 1.9, 8.1 Hz, 1H), 7.48
(m, 1H), 7.40 (m, 1H), 7.38 (m, 2H), 7.29 (dd, J = 1.5, 7.4 Hz, 1H),
7.22 (t, J = 7.6 Hz, 1H), 7.18 (m, 1H) 5.84 (d, J = 15.5 Hz, 1H), 4.95
(d, J = 15.5 Hz, 1H), 3.44 (td, J = 1.7, 6.4 Hz, 2H), 3.20 (ddt, J = 3.4,
11.9, 14.3 Hz, 2H), 3.13 (t, J = 7.5 Hz, 2H), 2.76 (s, 3H), 1.98 (m,
4H), 1.88 (m, 2H), 1.67 (m, 1H), 1.47 (dq, J = 2.8, 12.9 Hz, 2H),
1.34 (tq, J = 2.8, 12.1 Hz, 2H), 1.17 (tq, J = 3.6, 13.0 Hz, 2H 1H);
13C NMR (125 MHz, CD3OD) d 171.1, 169.1, 144.4, 141.7, 140.5,
139.7, 138.9, 137.0, 135.4, 134.6, 132.8, 132.5, 132.4, 131.1,
130.4, 129.2, 128.7, 127.5, 126.6, 126.4, 66.5, 54.6, 52.5, 38.9,
37.1, 36.6, 31.8, 27.8, 26.3, 26.2, 26.1; HREIMS [M+Na]+ m/z:
570.1917 (calcd for C31H34ClN3NaO2S, 570.1958).
Amine 47 (19.12 g, 0.0602 mol) was dissolved in THF (425 mL)
containing H2O (75 mL). The mixture was cooled to 0 °C and
LiOHꢁH2O (7.58 g, 0.181 mol) was added in one portion. After 1 h,
the reaction was warmed to rt. HPLC or TLC monitoring indicated
that hydrolysis was complete after 36 h at rt. The pH was adjusted
to 4.0 by the addition of 2 M HCl. The remaining mixture was dried
to an off white powder by rotary evaporation followed by lyophi-
lization with MeOH (6 ꢂ 300 mL). The resulting powder was fur-
ther dried by rotary evaporation with toluene (3 ꢂ 300 mL) and
suspended in DMF (400 mL). After cooling to ꢃ20 °C, 1,10-car-
bonyl-diimidazole (48.84 g, 0.301 mol) was added in portions over
3 h. The flask was equipped with a drying tube filled Drierite. The
reaction was warmed to rt over 2 h. After 24 h of vigorous stirring
at rt, reaction was cooled to 0 °C and methanol (200 mL) was
added. The reaction was warmed over 1 h to rt and stirred for an
addition 12 h. The contents of the flask were then diluted in EtOAc
(200 mL) and filtered with 1:10 MeOH/EtOAc (2 L) through a DCVC
column charged with 250 g of Celite (packed at top of column) and
500 g of silica gel (packed at bottom of column). The elutant was
concentrated on a rotary evaporator. Recrystallization from a min-
imal amount hot toluene followed by slow cooling to rt and storage
at 4 °C for 24 h provided 14.21 g (83%) of methyl ester 50 as a
white powder. The remaining mother liquor contained a mixture
of 50 and its corresponding acid 49 that could be converted to 50
(1.23 g, 6% yield) by treatment with 0.5 M TMSCH2N2 in THF. This
recycled material was not tabulated in the yield. Mp 179–181 °C,
1H NMR (500 MHz, CDCl3) d 8.48 (s, 1H, NH), 7.85 (dd, J = 1.9,
7.4 Hz, 1H); 7.79 (d, J = 1.7 Hz, 1H); 7.77 (dd, J = 1.7, 8.0 Hz, 1H),
7.63 (d, J = 8.0 Hz, 1H), 7.49 (dd, J = 1.5, 7.5 Hz, 1H), 7.40 (ddd,
J = 1.6, 8.3, 11.4 Hz, 1H), 7.39 (ddd, J = 1.6, 8.4, 8.7 Hz, 1H), 3.89
(s, 3H); 13C NMR (125 MHz, CDCl3) d 170.2. 166.3, 140.1, 137.2,
136.8, 136.0, 133.5, 133.0, 132.5, 132.5, 132.0, 129.6, 127.3,
124.2, 53.1; HREIMS [M+Na]+ m/z: 308.0489 (calcd for
C
15H11NNaO3S, 308.0357).
4.2.6. N-[3-(Cyclohexylmethylamino)propyl]-10-[(3-chloro-
phenyl)methyl]-10,11-dihydro-11-oxo-dibenzo-[b,f][1,4]-
thiazepine-8-carboxamide-5-oxide (2)
Sulfide 52 (2.25 g, 4.09 mmol) was dissolved in CH2Cl2 (200 mL)
containing NaHCO3 (1.72 g, 20.43 mmol). m-CPBA was purified by
dissolving commercial material in CH2Cl2 and washing repetitively
with a 10% w/v solution of NaHCO3, brine, and then drying by ro-
tary evaporation. A sample of freshly purified m-CPBA (0.78 g,
4.2.4. Methyl 10-(3-chlorobenzyl)-11-oxo-10,11-dihydrodi-
benzo[b,f][1,4]thiazepine-8-carboxylate (51)
NaH washed free of oil (1.49 g, 0.0622 mol) was added carefully
in portions to ester 50 (14.21, 0.0498 mol) dissolved in DMF
(200 mL) under Ar at ꢃ20 °C. After warming to 5 °C over 1 h, the
resulting suspension was cooled to ꢃ20 °C and 3-chlorobenzylbro-

G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
3963
4.50 mmol) was added as a solid to the cooled mixture of 50 and
stirred for 2 h at ꢃ78 °C. The solution was then warmed over 2 h
to 0 °C at which point 0.05 M NaOH (100 mL) was added, and the
mixture was extracted with CH2Cl2 (3 ꢂ 200 mL). The combined
organic layers were then washed with 0.05 M NaOH (50 mL) and
brine (100 mL), dried with Na2SO4, and concentrated by rotary
evaporation. Flash chromatography (2:1 Hx/EtOAc to 10:1 MeOH/
EtOAc) afforded 1.50 g (65%) as an off-white powder. Mp 141–
143 °C, 1H NMR (500 MHz, CD3OD) d 8.05 (m, 1H); 7.85 (dt,
J = 1.7, 8.1 Hz, 1H); 7.81 (d, J = 7.6 Hz, 1H); 7.75–7.66 (m, 3H),
7.57 (m, 1H), 7.40 (br s, 1H), 7.33–7.26 (m, 3H), 5.86 (d,
J = 15.3 Hz, 1H), 4.93 (d, J = 15.5 Hz, 1H), 3.47 (m, 2H), 3.24 (m,
2H), 3.08 (m, 1H), 2.78 (s, 3H), 2.05 (m, 4H), 1.89 (d, J = 13.0 Hz
2H), 1.69 (d, J = 14.1, Hz, 1H), 1.55–1.34 (m, 3H), 1.19 (m, 2H);
13C NMR (125 MHz, CD3OD) d 168.7, 167.2, 149.4, 149.0, 140.3,
138.9, 136.9, 135.6, 133.9, 132.5, 132.3, 131.5, 132.42, 129.7,
129.6, 129.2, 128.0, 127.8, 126.8, 122.4, 120.4, 66.5, 53.7, 52.5,
38.1, 36.5, 36.5, 28.8, 26.7, 26.2, 26.1, 26.0; HREIMS [M+Na]+ m/z:
586.1929 (calcd for C31H34ClN3NaO3S, 586.1907).
Dilutions of drug were incubated with 200
l
L of bacteria that
had been adjusted to an OD600 of 0.100. The plate was incubated
at 37 °C until the wells containing the medium plus DMSO control
were completely confluent with bacteria, usually 8–10 d. The dilu-
tion of drug that inhibited approximately 95% growth as deter-
mined visually was noted.
4.3.5. Inhibition of non-replicating M. tuberculosis by 2
In order to produce non-replicating bacteria, M. tuberculosis
were cultured under nutrient starvation conditions as described
by Betts24 with the addition of 0.05% Tween 80 to the PBS to limit
the sticking and clumping of bacteria. A culture of M. tuberculosis
Erdman was grown in Middlebrook 7H9 plus ADS to mid-log phase
and was then washed twice with 0.05% Tween 80 in PBS. After
washing, the bacteria were re-suspended in their original volume
with 0.05% Tween 80 in PBS (OD600 = 0.6–0.8) and 15 mL aliquots
were incubated at 37 °C in 50 mL polypropylene tubes (Fisher)
with the cap tightened. The tubes were opened weekly for cell via-
bility counts. After 6 weeks, dilutions of 2 were added to the tubes,
which were incubated for an additional 15 d. The number of viable
bacteria was determined by plating serial dilutions on Middle-
brook 7H9 agar medium with OADC (Difco).
4.3. Biological evaluation
4.3.1. Measurement of MSH and GlcN-Ins levels
MSH from M. tuberculosis cell extracts was determined by reac-
tion with monobromobimane (Invitrogen) and quantitation by
HPLC analysis with fluorescent detection as previously described.14
GlcN-Ins was determined in M. tuberculosis cell extracts using the
AccQ Tag amine reagent (Waters) as previously described.14
4.3.6. Cytotoxicity of compounds to Vero cells
The toxicity of the compounds in Table 3 was tested using the
neutral red assay for Vero cell cytotoxicity as previously de-
scribed30 with minor modifications. African green monkey Vero-
E6 cells were seeded at 104 cells in 200
l
L of DMEM medium with
10% fetal calf serum. Compounds were added to the medium at 25,
50, 100, and 200 M (DMSO <2%) and the cells incubated for 2 d at
37 °C. The medium was removed and replaced with medium con-
taining 40 g/mL neutral red and then incubated for 3 h for uptake
4.3.2. Determination of IC50 values for MshC
The MshC activity was assayed by the production of Cys-GlcN-
Ins from 100 lM Cys, 100 lM ATP, 50 lM GlcN-Ins and recombi-
nant M. tuberculosis MshC. The Cys-GlcN-Ins was quantitated by
HPLC with fluorescence detection after reaction with monobromo-
bimane, as previously described.29 Inhibitors were dissolved in
l
l
of the dye. The medium was removed and the cells quickly washed
with 0.5% formaldehyde with 1% CaCl2. The wells were filled with
200 lL of 1% HOAc in 50% aq EtOH. After 10 min, the microtiter
plate was read at 540 nm. The results were expressed as the test
compound concentration required to reduce the cell number by
1 log (MIC90).
DMSO and were added at 50, 100, 200 and 400 lM to the substrate
mix prior to the addition of MshC. The DMSO level was limited to
5% below which the activity of MshC was unaffected. Assays with
visible inhibitor precipitation were discarded. The IC50 values were
determined from a linear fit of the ratio of MshC activity without
inhibitor to that with inhibitor.
Acknowledgments
This work was funded by National Institutes of Health Grant
AI072133. We thank Cyrille Bonhomme and Michael Buchmeier
for assistance with Vero cell assays.
4.3.3. Mycobacterium strains and culture
M. tuberculosis Erdman (ATCC 35801) and M. tuberculosis
CDC1551 were obtained from the American Type Culture Collec-
tion (ATCC) and the Centers for Disease Control (CDC), respectively.
Human isolates of M. tuberculosis were obtained from UC San Diego
Medical Center and the Veterans Administration Hospital in La Jol-
la, CA via Don Guiney. Unless otherwise stated M. tuberculosis
strains were cultured in Middlebrook 7H9 media with 0.05% Tween
80 and ADS supplement as described previously.14
Supplementary data
Supplementary data (copies of NMR spectra on NTF1836 (2) and
select synthetic intermediates have been provided) associated with
this article can be found, in the online version, at doi:10.1016/
j.bmc.2011.05.028.
4.3.4. Inhibition of M. tuberculosis growth by 2
References and notes
M. tuberculosis Erdman and patient isolates were grown in Mid-
dlebrook 7H9 + 10% ADS (albumin–dextrose–saline) to mid-log
phase and the bacteria were adjusted to an OD600 of 0.08 in fresh
1. WHO. Global Tuberculosis Control 2010. World Health Organization, 2010.
2. Young, D. B.; Perkins, M. D.; Duncan, K.; Barry, C. E., 3rd. J. Clin. Invest. 2008, 118,
1255.
3. Ginsberg, A. M.; Spigelman, M. Nat. Med. 2007, 13, 290.
4. Migliori, G. B.; Dheda, K.; Centis, R.; Mwaba, P.; Bates, M.; O’Grady, J.;
Hoelscher, M.; Zumla, A. Trop. Med. Int. Health 2010, 15, 1052.
5. Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. Nature 2011, 469,
483.
6. Chao, M. C.; Rubin, E. J. Annu. Rev. Microbiol. 2010, 64, 293.
7. Jothivasan, V. K.; Hamilton, C. J. Nat. Prod. Rep. 2008, 25, 1091.
8. Newton, G. L.; Buchmeier, N.; Fahey, R. C. Microbiol. Mol. Biol. Rev. 2008, 72, 471.
9. Newton, G. L.; Ta, P.; Bzymek, K. P.; Fahey, R. C. J. Biol. Chem. 2006, 281, 33910.
10. Newton, G. L.; Av-Gay, Y.; Fahey, R. C. Biochemistry 2000, 39, 10739.
11. Newton, G. L.; Ko, M.; Ta, P.; Av-Gay, Y.; Fahey, R. C. Protein Expr. Purif. 2006, 47,
542.
7H9 + ADS.14 Varying amounts of compound 2 from 10
l
M to
75
lM from a 23 mM stock solution in DMSO were added to
8 mL aliquots of the bacteria that were then incubated in vented
T-25 flasks for 7 d at 37 °C. Duplicate samples were prepared
including control samples that contained an equivalent amount
of DMSO as was present in the 75 lM sample. After 7 d, samples
were taken for measurement of growth by OD600, of viability by
plating on Middlebrook 7H9 + OADC agar plates and quantitation
of MSH and GlcN-Ins as described above.14 The effect of each com-
pound on growth was also analyzed using a 96 well plate assay.

3964
G. L. Newton et al. / Bioorg. Med. Chem. 19 (2011) 3956–3964
12. Steffek, M.; Newton, G. L.; Av-Gay, Y.; Fahey, R. C. Biochemistry 2003, 42, 12067.
13. Newton, G. L.; Av-Gay, Y.; Fahey, R. C. J. Bacteriol. 2000, 182, 6958.
14. Buchmeier, N. A.; Newton, G. L.; Koledin, T.; Fahey, R. C. Mol. Microbiol. 2003,
47, 1723.
22. Tremblay, L. W.; Fan, F.; Vetting, M. W.; Blanchard, J. S. Biochemistry 2008, 47,
13326.
23. Gutierrez-Lugo, M. T.; Baker, H.; Shiloach, J.; Boshoff, H.; Bewley, C. A. J. Biomol.
Screening 2009, 14, 643.
15. Rawat, M.; Newton, G. L.; Ko, M.; Martinez, G. J.; Fahey, R. C.; Av-Gay, Y.
Antimicrob. Agents Chemother. 2002, 46, 3348.
24. Betts, J. C.; Lukey, P. T.; Robb, L. C.; McAdam, R. A.; Duncan, K. Mol. Microbiol.
2002, 43, 717.
16. Sareen, D.; Newton, G. L.; Fahey, R. C.; Buchmeier, N. A. J. Bacteriol. 2003, 185,
6736.
25. Buchmeier, N. A.; Newton, G. L.; Fahey, R. C. J. Bacteriol. 2006, 188, 6245.
26. Sassetti, C. M.; Boyd, D. H.; Rubin, E. J. Mol. Microbiol. 2003, 48, 77.
27. Vilcheze, C.; Av-Gay, Y.; Attarian, R.; Liu, Z.; Hazbon, M. H.; Colangeli, R.; Chen,
B.; Liu, W.; Alland, D.; Sacchettini, J. C.; Jacobs, W. R. Mol. Microbiol. 2008, 69,
1316.
17. Metaferia, B. B.; Fetterolf, B. J.; Shazad-Ul-Hussan, S.; Moravec, M.; Smith, J. A.;
Ray, S.; Gutierrez-Lugo, M. T.; Bewley, C. A. J. Med. Chem. 2007, 50, 6326.
18. Newton, G. L.; Ta, P.; Sareen, D.; Fahey, R. C. Anal. Biochem. 2006, 353, 167.
19. Gutierrez-Lugo, M. T.; Newton, G. L.; Fahey, R. C.; Bewley, C. A. Protein Expr.
Purif. 2006, 50, 128.
28. Ta, P.; Buchmeier, N. A.; Newton, G. L.; Rawat, M.; Fahey, R. C. J. Bacteriol. 2011,
193, 1981.
20. Fan, F.; Luxenburger, A.; Painter, G. F.; Blanchard, J. S. Biochemistry 2007, 46,
11421.
21. Williams, L.; Fan, F.; Blanchard, J. S.; Raushel, F. M. Biochemistry 2008, 47, 4843.
29. Sareen, D.; Steffek, M.; Newton, G. L.; Fahey, R. C. Biochemistry 2002, 41, 6885.
30. Valdivieso-Garcia, A.; Clarke, R. C.; Rahn, K.; Durette, A.; Macleod, D. L.; Gyles,
C. L. Appl. Environ. Microbiol. 1993, 59, 1981.