Synthetic calanolides with bactericidal activity against replicating and nonreplicating mycobacterium tuberculosis

Article abstract of DOI:10.1021/jm4019228

It is urgent to introduce new drugs for tuberculosis to shorten the prolonged course of treatment and control drug-resistant Mycobacterium tuberculosis (Mtb). One strategy toward this goal is to develop antibiotics that eradicate both replicating (R) and

Full text of DOI:10.1021/jm4019228

Article
pubs.acs.org/jmc
Synthetic Calanolides with Bactericidal Activity against Replicating
and Nonreplicating Mycobacterium tuberculosis
Purong Zheng,†,# Selin Somersan-Karakaya,∥,# Shichao Lu,†,‡ Julia Roberts,§ Maneesh Pingle,§
Thulasi Warrier,§ David Little,§ Xiaoyong Guo,†,‡ Steven J. Brickner,⊥ Carl F. Nathan,*,§ Ben Gold,*,§
and Gang Liu*,†,‡
‡
†Tsinghua-Peking Center for Life Sciences and Department of Pharmacology and Pharmaceutical Sciences, School of Medicine,
Tsinghua University, Haidian Dist., Beijing 100084, P. R. China
§Department of Microbiology and Immunology, Weill Cornell Medical College, New York, New York 10065, United States
∥Department of Medicine, Weill Cornell Medical College, New York, New York 10065, United States
⊥SJ Brickner Consulting, LLC, Ledyard, Connecticut 06339, United States
S
* Supporting Information
ABSTRACT: It is urgent to introduce new drugs for tuberculosis to shorten
the prolonged course of treatment and control drug-resistant Mycobacterium
tuberculosis (Mtb). One strategy toward this goal is to develop antibiotics that
eradicate both replicating (R) and nonreplicating (NR) Mtb. Naturally
occurring (+)-calanolide A was active against R-Mtb. The present report details
the design, synthesis, antimycobacterial activities, and structure−activity
relationships of synthetic calanolides. We identified potent dual-active nitro-
containing calanolides with minimal in vitro toxicity that were cidal to axenic
Mtb and Mtb in human macrophages, while sparing Gram-positive and -negative
bacteria and yeast. Two of the nitrobenzofuran-containing lead compounds
were found to be genotoxic to mammalian cells. Although genotoxicity
precluded clinical progression, the profound, selective mycobactericidal activity
of these calanolides will be useful in identifying pathways for killing both R- and
NR-Mtb, as well as in further structure-based design of more effective and drug-like antimycobacterial agents.
(LD50 = 7.6 μg/mL), with an unacceptable selectivity index
(SI) of 2.4.7 Although a small number of pyranocoumarins
have also been shown to possess antimycobacterial activity,8
as a class, calanolides have not been thoroughly explored
as antimycobacterial agents. To do this, we initiated a high-
throughput screening of a chemical library enriched for
calanolides synthesized in our laboratory and discovered several
12-oxo-calanolides10a with activity against both R- and NR-Mtb.
A further SAR campaign identified more potent synthetic
calanolides with decreased toxicity, including three compounds
with activity against Mtb infecting human macrophages.
INTRODUCTION
■
Tuberculosis (TB), resulting from Mycobacterium tuberculosis
(Mtb), is the leading cause of death by a single bacterial agent.
Fueled in part by pandemics of HIV infection and diabetes,
TB remains prevalent. Mtb can persist for prolonged periods in
a state that is non- or slowly replicating, here called non-
replicating (NR) for simplicity. An estimated one-third of the
world’s population is latently infected with Mtb.1,2 If Mtb does
not contain genetic mutations associated with resistance to a
given drug, the drug can generally kill Mtb rapidly in vitro
under conditions where the bacterium is otherwise replicating
(R). However, most TB drugs are far less effective against the
same strains of Mtb under NR conditions.3 In patients with
active, uncomplicated TB, the presence of NR subpopulations
of Mtb may account for the need to extend treatment for
6 months to achieve acceptable cure rates.4 A variety of NR
models, such as carbon starvation or hypoxia, have been
employed to identify compounds that kill NR-Mtb.5 An addi-
tional NR model uses a combination of four physiologic con-
ditions: hypoxia, mild acidity, nitrosative stress, and restriction
of the major carbon source to a fatty acid.5b
CHEMISTRY
■
Round 1 of Design and Synthesis: Structural Diversifi-
cations of Ring D. The chiral ring D in (+)-calanolide A
possesses three asymmetric carbons at the C10-, C11-, and
C12-positions that complicate its synthesis.9 Our previous work
revealed that 10-chloromethyl-11-demethyl-12-oxo-calanolide
had higher potency toward drug-susceptible and drug-resistant
HIV-1 and afforded a superior safety profile compared to the
parent compound, (+)-calanolide A.10 For these reasons, we
In a previous report, the HIV-1-active6 natural product
(+)-calanolide A was active against R-Mtb (MIC = 3.1 μg/mL),
but the compound was almost equally toxic to Vero cells
Received: December 16, 2013
Published: April 2, 2014
© 2014 American Chemical Society
3755
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Figure 1. Summary of SARs from round 1 of design and synthesis.
a
Scheme 1. Synthesis of Compounds from Round 1
a
Reagents and conditions: (a) Tf2O, lutidine, CH2Cl2, 0 °C, 30 min, 73%; (b) secondary amine, Et3N, DMF, 90 °C, 12 h, 25−85%; (c) For 4a:
Me2SO4, K2CO3, acetone, rt, 12 h, 92%; For 4b: acetic anhydride, pyridine, DMAP, CH2Cl2, rt, 12 h, 84%; For 4c: benzoyl chloride, pyridine,
DMAP, CH2Cl2, rt, 12 h, 93%; For 4d: cinnamic acid, DCC, DMAP, CH2Cl2, rt, 12 h, 54%; For 4e−4q: R3OH, PPh3, DEAD, THF, rt, 12 h, 38−
95%; For 4r: 2-chloro-N-phenethylacetamide, K2CO3, 100 °C, 24 h, DMF, 30%; (d) Tf2O, pyridine, CH2Cl2, 0 °C, 30 min, 79%; (e) For 6a−6f:
acetylene, PdCl2(PPh3)2, CuI, Ar, Et3N, THF, reflux, 12 h, 48−92%; For 7a−7d: Primary amine, Pd(OAc)2, BINAP, Cs2CO3, Ar, dioxane, 100 °C
5 h, 68−87%; For 8a−8d: secondary amine, Et3N, DMF, 90 °C, 12 h, 37−76%; (f) TFA, CH2Cl2, rt, 30 min, 89%; (g) R4-NCO or
isopentanoyl chloride, Et3N, THF, reflux, 60−93%; (h) i, diphenylmethanimine, Pd(OAc)2, BINAP, Cs2CO3, Ar, dioxane, 100 °C, 5 h, 55%; ii,
2 M HCl, THF, rt, 12 h, 73%; (i) For 12a: acetic anhydride, pyridine, DMAP, THF rt, 12 h, 64%; For 12b: benzoyl chloride, pyridine, THF, rt,
12 h, 71%; For 12c: 4-fluorophenyl-NCO, THF, reflux, 3 h, 69%, 12d:1-naphthyl-NCO, THF, reflux, 3 h, 52%. For 12e: pyrazine-2-
carboxylic acid, HOBt, BOP, DMF, rt, 3 h, 78%; (j) iodobenzene diacetate, TBAI, AcONa, CH3CN, rt, 5 h, 35−41%; (k) sat. aq NaHCO3, THF, rt,
72h, 55%; (l) H2O, DBU, THF, reflux, 5 h, 83%; (m) morpholine, DBU, THF, −15 °C to rt, 5 h, 91%; (n) i, NO2CH2Br, K2CO3, acetone, rt, 3 h; ii,
acetic anhydride, 100 °C, 1 h, 92%.
explored ring D diversifications in our first round of compound
design.
Intermediates 1a−1d were prepared as described.11 Inter-
mediate 1a was reacted with triflic anhydride to prepare
trifluoromethanesulfonate 2. Employing SNAr reactions of 2,
compounds 3a−3g (Table 1) were attained using nucleophilic
secondary amines (Scheme 1). A propiophenone at the C10
position of these compounds was placed to mimic the chroman-
4-one unit of the previously published compound F18.10b
To develop analogues exploring a wide range of substitutions
at the C9 position of the ABC scaffold (Table 1), intermediate
On the basis of the reported active calanolides,8 we anti-
cipated that the ABC scaffold with appropriate side chains at
the C9 and/or C10 position might show improved activity
against Mtb (Figure 1). To test this hypothesis, we syn-
thesized compounds belonging to the ABC scaffold with
diversified side chains at C9 and/or C10 positions (Scheme 1,
Table 1).
3756
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Table 1. In Vitro Activity of the Analogues in Round 1 of Design and Synthesis
3757
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Table 1. continued
*
**
These compounds gained activity when exposed to NR conditions for longer (7 days) with indicated MICs in parentheses. The activity was
confirmed in WT Mtb H37Rv, MICs in the table are indicative of WT Mtb H37Rv bacteria. See Scheme 1 for structures of compounds 13−18. SI,
̂
selectivity index was determined by LD50 for HepG2 cells/MIC90 for R-Mtb. ND, not determined. CN, cannot be calculated.
3758
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
1b was reacted alternatively with Me2SO4, acetic anhydride,
benzoyl chloride, cinnamic acid, or 2-chloro-N-phenethyl-
acetamide to produce compounds 4a−4d and 4r, respectively.
Reaction of 1b with a series of commercially available alcohols
under Mitsunobu reaction conditions produced compounds
4e−4q. Trifluoromethanesulfonate 5 was synthesized from 1b.
Intermediate 5 allowed us to construct a C−C bond as well as a
C−N bond side chains at the C9 position by taking advantage
of palladium-catalyzed cross-coupling reactions. For example, a
Sonogashira cross-coupling reaction between 5 and commer-
cially available acetylenes gave C9 alkynyl-substituted com-
pounds 6a−6f. When 5 underwent Buchwald−Hartwig cross-
coupling with benzylamines and aniline, analogues 7a−7d were
obtained. Benefiting from SNAr reactions with secondary
amines, 5 was also used to produce compounds 8a−8c in
moderate yield. Potentially resulting from the pyranone’s
lactone, the reaction between 5 and secondary amines such
as dimethylamine led to multiple additional side reactions when
the reaction temperature was increased or the reaction time
was prolonged. When tert-butylpiperazine-1-carboxylate was
reacted with 5, 8d was produced. The Boc group of 8d was
cleaved in TFA to generate compound 9. Derivatization of 9
with commercially available isocyanates and acetyl chloride
in the presence of triethylamine yielded compounds 10a−10l.
By using diphenylmethanimine as a coupling partner with 5
followed by hydrolysis, 5 was converted into compound 11,
which was treated with either acetic anhydride, benzoyl
chloride, 4-fluorophenylisocyanate, 1-isocyanatonaphthalene, or
pyrazine-2-carboxylic acid, respectively, to afford compounds
12a−12e (Scheme 1).
in the presence of K2CO3, followed by dehydration in acetic
anhydride to afford the anticipated 2-nitrobenzofuran 18
(Scheme 1).17
Table 1 illustrates the activity of compounds against Mtb
in round 1 of design and synthesis. Compound 14 inhibited
NR-Mtb with MIC90 of 6 μg/mL, and compounds 4f, 4k, and
8c inhibited NR-Mtb when incubated for extended periods
(7 days), with MIC90 values of 7, 2, and 2 μg/mL, respectively.
These compounds showed no effect on R-Mtb. Compounds 4e,
4g, 7a, and 9 showed R-Mtb inhibitory activity with MIC90
values of 20, 19, 19, and 23 μg/mL, respectively, but exhibited
no effects on NR-Mtb.
When ring D was replaced with a benzofuran-3-one ring,
compound 14 showed only activity against NR-Mtb. However,
replacement of ring D with the 2-nitrobenzofuran pharmaco-
phore (18) led to potent inhibition of both R-Mtb and NR-
Mtb, with MIC90 values of 0.3 and 0.6 μg/mL, respectively.
Figure 1 summarizes the SAR at this round. As the most potent
analogue in the first round of design and synthesis with
submicrogram per milliliter MIC90 value and improved SI com-
pared to (+)-calanolide A, compound 18 was selected for
further SAR studies.
Round 2 of Design and Synthesis: Synthesis of 2-
Nitrobenzofuran 18 Analogues. Criteria for the design and
synthesis of 18 analogues (Table 2) were the following: (1) to
explore various double bonds and to improve the clogP value
of the target compound (22a−22g, 26a−26c, Scheme 2);
(2) to increase the molecular flexibility, including saturating
the C−C double bond (30, 32, and 33, Scheme 2), as well
as removing ring C (34−40, Scheme 3); (3) to define the role
of the nitro group (42, Scheme 4); and (4) to explore various
nitrofurans (46, 49, 52, 56, Schemes 5−8).
Using the same procedure as for the synthesis of 18,
compounds 22a−22g bearing various substituents at C3 and
C4 positions were synthesized through intermediates 19−21.
The reaction of intermediate 20j and (E)-(3,3-diethoxyprop-
1-enyl)benzene10a provided intermediate 23, which led to
target compound 24 containing a phenyl substitution at the
C6 position in place of the dimethyl substitution found on 18
(Scheme 2).
We next tried to improve the water solubility of the
compounds. Hydrophilic groups were introduced by 4-methyl
substitution in 22a using intermediate 21i with chloromethyl
substitution at the C4 position, which reacted with either
morpholine, thiomorpholine, or N-methylpiperazine, to
produce intermediates 25a−25c, respectively. When these
O-hydroxybenzaldehydes (25a−25c) were treated with bromo-
nitromethane as was done for 18 (Scheme 1, reaction condition
n, (ii)), mixed products were observed in the dehydration step in
acetic anhydride. The optimal condition in the presence of PPh3
and DEAD was finally employed, resulting in 26a, 26b, and 26c
(Scheme 2).18
To investigate the effect of ring C modifications, ring
C-saturated analogues 30a and 30b were synthesized. Hydro-
genation of 1d with Pd/C catalyst gave a complicated product
mixture due to an aldehyde moiety. Protecting 1d’s aldehyde as
the diacetyl acetal also acetylated the hydroxyl group to give
27a, which was hydrogenated with Pd/C catalyst to afford 28a.
When treated with K2CO3 in methanol, 28a was deprotected
to 29a. Then 29a was reacted with BrCH2NO2 in the presence
of K2CO3 followed by dehydration with acetic anhydride to
give the target compound 30a. Using the same method, the
Compounds 13 and 14 were designed to explore replacement
of ring D with the benzofuran-3-one fragment, found in the TB
drug rifampicin. In the presence of iodobenzene diacetate,
tetrabutylammonium iodide, and AcONa, 1a and 1c yielded the
target compounds 13 and 14, respectively.12 Using compound
14, we also attained compounds 15, 16, and 17 by the method
reported recently by our laboratory (Scheme 1).13
Although considered a major structural alert due to potential
mutagenicity, 5-nitrofuran and 2-nitrobenzofuran moieties have
found utility in some antibacterials.14 A precedent for species-
specific nitroreductases that confer specificity at the drug
activation stage led us to explore nitrofurano analogues of
calanolides as antimycobacterial agents. For example, reduction
of a nitrofuran-containing molecule by an endogenous bacterial
nitro-reductase caused DNA lesions in E. coli.14a The
mechanism of action of nitrofurans remains to be elucidated.
One possible mechanism is through cleavage of the nitrofuran
ring to form the specific tail group with the potential to
covalently bind biological molecules.14a In human cells, recent
research demonstrated that aldehyde dehydrogenase (ALDH)
2 was involved in off-target toxic side effects of 5-nitrofurans.
Combining 5-nitrofurans with ALDH2 inhibitors to block off-
target 5-nitrofuran biological activity may alleviate clinical side
effects caused by 5-nitrofurans.15 Reduction of a nitro group to
a hydroxylamino group may generate reactive intermediates
that covalently interact with DNA.16 This reduction is expected
to occur upon exposure of the nitro compound to the huge
diversity of organisms comprising the human gut microbiome
and can be evaluated during in vitro toxicity and genotoxicity
assays of nitrofuran-containing compounds.
To examine the potential anti-Mtb activity of nitrofurano
analogues of calanolides, we replaced ring D with a nitrofurano
moiety by reacting intermediate 1d with bromonitromethane
3759
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Table 2. In Vitro Activity of the Analogues in Round 2 of Design and Synthesis
compd
**
MIC90 (μg/mL) R, NR
LD50, SI
compd
MIC90 (μg/mL) R, NR
LD50, SI
ND
22a
0.08, 0.2
2, >21
1, 16
22b
22d
22f
24
2, 19
22c
ND
<0.1, 0.2
19, 19
>50, >510
ND
**
22e
0.6, 1
>25, >40
ND
22g
26a
<0.2, 2
20, 6
3, 0.1
<0.2, <0.2
0.08, 0.3
0.8, 2
1, >5
3, 31
26b
<0.3, <0.3
0.6, 1
1, >5
**
**
26c
30a
>50, >80
ND
30b
33a
33c
35
>50, >64
3, >11
25, 64
ND
32
>18, >18
<0.1, <0.1
0.2, 0.4
<0.2, <0.2
0.6, 3
<0.2, <0.2
0.4, 0.8
33b
33d
36
13, > 128
25, 128
0.3, >2
>100, >160
1, >5
>14, >14
0.7, 18
**
38
ND
39
**
40a
0.08, 0.2
>100, >100
>18, >18
0.2, > 100
10, 125
ND
40b
46
<0.2, <0.2
>50, 6
42
ND
49
>18, CN
3, 16
52
>20, >20
0.4, 50
ND
56a
56b
3, 6
*
**
See Scheme 2 for R6 and R7 substitutions. The activity was confirmed in WT Mtb H37Rv; MICs in the table are indicative of WT Mtb H37Rv
bacteria. ND, not determined. CN, cannot be calculated SI, selectivity index determined by LD50 for HepG2 cells/MIC90 for R-Mtb.
3-fluoro-4-methyl analogue with saturated ring C was syn-
thesized, resulting in 30b (Scheme 2).
Several attempts to synthesize 11-dehydroxylated dihydroni-
trofurano analogue of 33a as well as the ring D saturated
analogue of 18 failed.
Synthesis of the target compound 32 permitted study of the
antitubercular activity of the dihydrocoumarin analogue. One
aspect of interest leading to the preparation of this com-
pound was to explore possible removal of the potential Michael
acceptor at this site of the coumarin. 3,4-Dihydro-5,7-
dihydroxycoumarin 31 was synthesized by hydrogenation of
5,7-dihydroxycoumarin 19j, employing Pd/C as catalyst aided
by ultrasound. Bis-phenol 31 was then converted to target
compound 32 in three steps, as was done for 18 (Scheme 2).
trans-11-Hydroxy-10,11-dihydro-10-nitrobenzofuran 33a, the
synthetic precursor of compound 18, was also prepared for
SAR studies. The relative trans configuration was confirmed by
Analogues 34−40 which lack ring C were designed to reduce
planarity of the target compounds. Synthesis of compound 34
started from 20j. The 5-hydroxy of 20j was selectively reacted
with chloromethyl methyl ether followed by the same method
that was used to make 18 from 1d to yield 34. Removal of
the protective group of 34 in acid conditions produced 35.
Reaction of 35 with either dimethyl sulfate or 1-bromomethyl-
4-fluorobenzene produced the anticipated compounds 36
and 39, respectively. Compound 35 was also converted to
triflate 37, which was used to prepare compound 38 employing
the Suzuki cross-coupling reaction. Mitsunobu reaction of 35
with commercially available alcohols 2-morpholinoethanol and
2-(pyrrolidin-1-yl)ethanol afforded target compounds 40a and
40b, respectively (Scheme 3).
19
1
H10−H11 coupling in H NMR. For further SAR evalua-
tion of the dihydronitrobenzofuran, 33b−33d were synthe-
sized using the same method as was done for 33a (Scheme 2).
3760
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
a
Scheme 2. Synthesis of Compounds 22a−22g, 24, 26a−26c, 30a, 30b, 32, and 33a−33d
a
Reagents and conditions: (a) POCl3, DMF, CH2Cl2, rt, 24 h; (b) (CH3)2CCHCH(OEt)2, pyridine, toluene, reflux, 3 h, 24−72% in two steps;
(c) i, NO2CH2Br, K2CO3, acetone, rt, 3 h; ii, acetic anhydride, 100 °C, 1 h, 40−51%; (d) (E)-(3,3-diethoxyprop-1-enyl)benzene, pyridine, toluene,
reflux, 3 h, 50%; (e) amine, DIPEA, DMF, rt, 12 h, 58−69%; (f) i, NO2CH2Br, K2CO3, acetone, rt, 3 h; ii, PPh3, DEAD, THF, rt, 1 h, 25−34%; (g)
acetic anhydride, Amberlite IR-120, rt, 1 h, 72−81%; (h) H2, Pd/C, CH3CH2OH, rt, 1 h, 90−96%; (i) K2CO3, CH3OH, rt, 1 h, 91−92%;
(j) H2, Pd/C, CH3CH2OH, ultrasound, rt,10 h, 60%; (k) BrCH2NO2, K2CO3, acetone, rt, 3 h, 51−99%.
Compound 42, a 10-denitro analogue of 18, was designed
to determine if the nitro group of compound 18 was essential
for antitubercular activity. To obtain 42, compound 41 was
prepared from 1b in a reaction with N-iodosuccinimide, followed by
a Sonogashira cross-coupling reaction with trimethylsilylacetylene,
intramolecular cyclization, and removal of the trimethylsilyl group to
provide target compound 42 (Scheme 4).20
As shown by compound 46, the 5-nitrofurano moiety was
incorporated at the ring C position, whereas the ring D was
2-methyl-2,3-dihydro-4-pyranone. This compound was useful to
study the impact of having a nitrofurano moiety in ring C. Similar
to intermediate 41, 44 was prepared from 4310a by reaction with
N-iodosuccinimide, followed by a Sonogashira cross-coupling
reaction with trimethylsilylacetylene, intramolecular cyclization,
and nitration to yield target compound 46 (Scheme 5).21
To further study the impact of the nitrofuran group, a
5-nitrofuran-2yl moiety was directly linked at the C9 position as
shown by compound 49 (Scheme 6). Its synthesis benefited
from triflate 47, which was synthesized from coumarin 19a using
the same procedure as that for triflate 5. Triflate 47 subjected to a
Suzuki coupling reaction gave compound 48, which was then
nitrated with HNO3 to produce compound 49.22
For SAR studies, 5-nitrofuran was also linked at the C9
position as shown by amide 52 (Scheme 7). Amine 50 was
synthesized in a similar method as 11. The reaction of 50 with
5-nitrofuran-2-carbonyl chloride 51 resulted in compound 52.
To clarify the role of the 2-nitrobenzofuran unit in antitubercular
activity, 2-nitrobenzofuran 56a and 2-nitro-2,3-dihydrobenzofuran
56b were also synthesized by the method recently reported by
our laboratory (Scheme 8).23 All of the analogues obtained in the
second round of design and synthesis were screened for their
activity against R-Mtb and NR-Mtb (Table 2), and the SAR
analyses are summarized in the subsequent sections.
3761
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
a
Scheme 3. Synthesis of Compounds 34−40 Belonging to the ABD Scaffold
a
Reagents and conditions: (a) i, CH3OCH2Cl, K2CO3, acetone, rt, 5 h; ii, NO2CH2Br, K2CO3, acetone, rt, 3 h; iii, acetic anhydride, 100 °C, 1 h,
41%; (b) 2 M HCl, THF, rt, 1 h, 91%; (c) Me2SO4, K2CO3, acetone, rt, 12 h, 95%; (d) Tf2O, 2,6-di-tert-butyl-4-methylpyridine, CH2Cl2, 0 °C,
30 min, 51%; (e) 4-fluorophenylboronic acid, PdCl2(PPh3)2, Na2CO3, H2O, DME, Ar, 55 °C, 24 h, 49%; (f) 1-(bromomethyl)-4-fluorobenzene,
K2CO3, acetone, reflux, 3 h, 51%; (g) For 40a: 2-morpholinoethanol, PPh3, DEAD, THF, rt, 12 h, 56%; For 40b: 2-(pyrrolidin-1-yl)ethanol, PPh3,
DEAD, THF, rt, 12 h, 17%.
a
a
Scheme 4. Synthesis of Compound 42
Scheme 6. Synthesis of Target Compound 49
a
a
Reagents and conditions: (a) furan-2-boronic acid, PdCl2(PPh3)2,
Na2CO3, H2O, DME, Ar, 55 °C, 24h, 73%; (b) HNO3, acetic anhydride,
20 °C, 2h, 31%.
Reagents and conditions: (a) NIS, CH2Cl2, rt, 3 h, 55%; (b) i,
Me3SiCCH, PdCl2(PPh3)2, CuI, THF, reflux, 12 h; ii, TFA, CH2Cl2, rt,
2 h, 37% for two steps.
a
Scheme 7. Synthesis of Target Compound 52
Modification of Ring B (compounds 22a−22g, 26a−26c,
32, Table 2). Compound 22a, in which C4 was methyl
substituted, markedly improved anti-Mtb activity, but its SI did
not improve compared to compound 18. C4-phenyl analogue
22b showed reduced potency against R-Mtb and NR-Mtb.
Introduction of benzyl at the C3 position of compound 22a,
making compound 22c, decreased R-Mtb potency and
abrogated activity against NR-Mtb. Introduction of hydrophilic
groups at the C4-methyl, such as 4-morpholinomethyl- (26a),
4-thiomorpholinomethyl- (26b), and 4-(4-methylpiperazino-
methyl)- (26c) gave rise to comparable potencies and SIs to 22a.
Introduction of fluoro substitution at the C3 position of compound
22a resulted in compound 22d and generated comparable potency
against R-Mtb and NR-Mtb with a markedly improved SI,
demonstrating the positive effect of a fluoro substitution at the
a
Reagents and conditions: CH2Cl2, rt, 3 h, 78%.
C3 position on SI. Compounds 22e and 22f, as C4 ethyl and C4
propyl analogues of 22d, respectively, had decreased potency,
which suggested that the substitution effect was n-propyl < ethyl <
methyl. The analogue 22g, in which C3−C4 had formed a six-
membered ring, was less active against NR-Mtb compared to the
a
Scheme 5. Synthesis of Target Compound 46
a
Reagents and conditions: (a) NIS, CH2Cl2, rt, 3 h, 84%; (b) Me3SiCCH, PdCl2(PPh3)2, CuI, THF, reflux, 12 h; (c) HNO3, Ac2O, rt, 2 h, 36% in 2 steps.
3762
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
a
Scheme 8. Synthesis of Target Compounds 56a, 56b
a
Reagents and conditions: (a) AcONH4, CH3NO2, 120 °C, 70 min, 70%; (b) NaBH4, CH3OH, 0 °C, 30 min, 82%; (c) DIB, TBAI, AcONa,
CH3CN, rt, 5 h, 56a (60%), 56b (10%).
Figure 2. Summary of SARs from round 2 of design and synthesis.
parent compound 18. Compound 32, a 3,4-dihydro derivative of
18, had an R and NR MIC90 > 18 μg/mL, demonstrating the
importance of the coumarin unit for anti-Mtb activity. This loss of
activity may either be due to the lower stability of the six-membered
lactone than that of the coumarin structure or simply reflect the
necessity for the α,β-unsaturated lactone moiety as a component
involved in its mode of action or multiple modes of action.
In summary, the SAR studies from these analogues revealed
that (a) the unsaturated coumarin structure (ring B) must
be maintained; (b) introduction of fluoro substitution at C-3
position improved LD50; (c) introduction of a C3 benzyl or
C3−C4 fused six-membered ring was detrimental to potency;
(d) the sequence of positive effects of C4 substitution on
R-Mtb potency was phenyl < n-propyl < ethyl < methyl. When
methyl was substituted with 4-morpholinomethyl-, 4-thiomor-
pholinomethyl-, and 4-(4-methylpiperazinomethyl)-, the result-
ing compounds generally maintained potency.
stitution at the C6 position of 18 with phenyl monosubstitution
gave 24 with decreased potency and SI. Compound 30a, formed
by 7,8-dihydrogenation of 18, reduced potency by 2-fold but
improved SI >10-fold due to improved LD50. Compound 30b,
formed by 7,8-dihydrogenation of 22d, displayed reduced potency,
which resulted in a reduced SI despite similar LD50 values. These
data suggested that saturation of ring C reduced potency with
improved or stable LD50 values. Compounds 35, 36, and 38−40
were generated as analogues of the ABD scaffold with various C5
substitutions. The 5-hydroxyl analogue (compound 35) was
inactive and 5-(4-fluorophenyl) analogue (38) lost most NR-Mtb
activity. However, the 5-OMe analogue (36), 5-(4-fluorobenzyloxy)
analogue (39), 5-(2-morpholinoethoxy) analogue (40a), and
5-(2-(pyrrolidin-1-yl)ethoxy) analogue (40b) all exhibited potent
activity against Mtb. In particular, 40a had a 16-fold improved
SI value compared to 18. SAR around ring C modifications can
be summarized as follows: (a) saturation of ring C reduced
potency with improved or stable LD50 values; (b) monophenyl
at the C6 position was inferior to dimethyl for both potency and
Modification of Ring C (Compounds 24, 30, 35, 36, 38,
39, 40a, 40b, Table 2). Replacement of the dimethyl sub-
3763
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
a
Scheme 9. Synthesis of Compounds 58, 59, 62, 65−67
a
Reagents and conditions: (a) PPh3, DEAD, THF, rt, 12 h, for 57a: 77%, for 57b−57e: crude product were obtained; (b) BrCH2NO2, K2CO3,
acetone, rt, 3 h, for ( )58a: 59%, for 58b−58e: 42−63% (from 20j or 20g); (c) CH3CHBrNO2, K2CO3, acetone, rt, 3 h, 55%; (d) DPPA, DIPEA,
DMF, rt, 2 h, 74%; (e) i. CH3NO2, K2CO3, rt, 1 h, ii. Ac2O, pyridine, rt, 1 h, 67%; (f) xylene, reflux, 3 h, 52%; (g) AcONH4, CH3NO2, 120 °C,
70 min, 92%; (h) for 64a: isopropylmagnesium chloride, THF, −78 °C, 1 h; for 64b: methylmagnesium bromide, THF, −78 °C, 1 h; (i) TBAI,
H2O2, CH3CN, rt, 6 h, for 65a, 49% (from 63); 67, 50% (from 63); (j) iodobenzene diacetate, TBAI, Et3N, CH3CN, 35 °C, 1 h, for 66a: 63% (from
63), for 66b: 65% (from 63); (k) HCl/CH3OH, 75 °C, 2 h, 83%.
SI; and (c) introduction of an alkoxyl substitution for R8 such
as 2-morpholinoethoxy produced a more potent analogue with
improved SI.
furans had stable SIs and slightly improved potency when
compared to nitrobenzofurans. Compound 42, a 10-denitro
analogue of 18 lacked activity against R-Mtb and NR-Mtb,
demonstrating the high significance of 18’s nitro group for its
antitubercular activity. Analogue 46, with a 5-nitrofurano on ring
C, had significantly reduced activity, emphasizing the importance
of the location of the nitro group. Compounds 49, a ring D
acyclic analogue with a C9 5-nitrofuran-2-yl substituent, and 52,
another ring D acyclic analogue with a 5-nitrofuran-
2-carboxamide side chain at the C9 position, were inactive.
These analogues were consistent with 2-nitrobenzofuran being
the pharmacophore of the potent compounds. Therefore,
Modification of Ring D (Compounds 33a−33d, 42, 46, 49,
52, 56a, and 56b, Table 2). During the synthesis of compound
18, intermediate compound 33a was prepared as an inseparable
mixture of trans isomers. This mixture exhibited a comparable
MIC and SI to compound 18. Compound 33b, as the synthetic
intermediate of compound 30b, had improved potency and SI
compared to 30b. Other dihydronitrobenzofurans, such as 33c
and 33d had comparable potency and SI to 33b. In conclusion,
this class of trans-3-hydroxy-2-nitro-2,3-dihydronitrobenzo-
3764
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Table 3. In Vitro Activity of the Analogues in Round 3 of Design and Synthesis
compd
**
MIC90 (μg/mL) R, NR
LD50, SI
compd
MIC90 (μg/mL) R, NR
LD50, SI
**
(+)-58a
58b
58d
59
0.08, 0.08
<1, <1
0.8, 10
<1, CN
0.4, >0.3
>100, >1
>22, >1
22, <1
(−)-58a
58c
0.08, 0.08
<1, <1
2, 20
<1, CN
0.4, >0.3
25, <0.5
>21, CN
<0.3, <0.3
<1, <1
58e
<1, <1
100, 100
22, >22
>22, >22
100, 100
62
>50, >50
>21, >21
1, 3
65a
66a
67
65b
66b
ND
**
The activity was confirmed in WT Mtb H37Rv; MICs in the table were indicative of WT Mtb H37Rv bacteria. ND, not determined. CN, cannot
be calculated. SI, selectivity index was determined by LD50 for HepG2 cells/MIC90 for R-Mtb.
Figure 3. Summary of SARs from round 3 of design and synthesis.
2-nitrobenzofuran 56a and 2-nitro-2,3-dihydrobenzofuran 56b
were evaluated for anti-Mtb activity. These two compounds exhi-
bited potent R-Mtb inhibitory activity but lacked NR-Mtb
inhibitory activity. This result demonstrated the significance of
the coumarin unit in the synthetic calanolide nitrobenzofurans
with potent dual activity toward both R-Mtb and NR-Mtb, such as
18, 30a, and 40a.
2-nitrobenzofuran and coumarin moieties were essential for
potent dual inhibitory activity against R-Mtb and NR-Mtb.
Round 3 of Design and Synthesis: Synthesis of 2-Nitro-
benzofuran 40a Analogues. Analysis of the SAR studies based
on the 2-nitrobenzofuran 18 showed that the 2-nitrobenzofuran
40a had improved potency and SI values (Table 2). Addi-
tionally, it had satisfactory skeletal flexibility, because the rigid ring
C was replaced with a morpholinoethoxyl substitution. Therefore,
we next carried out modifications based on 40a.
In summary (Figure 2), the SAR studies of analogues exploring
modifications of rings B, C, and D informed us that both the
3765
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Based on our conclusion from the second round of design
and synthesis that trans-3-hydroxy-2-nitro-2,3-dihydronitroben-
zofurans maintained SI with slightly improved potency
compared to nitrobenzofurans, the synthetic precursor of
compound 40a was used to obtain trans-enantiomeric ( )-58a.
To determine the MIC and SI of the enantiomeric isomer
respectively, (+)- and (−)-58a were separated by chiral HPLC
and their absolute configurations were proposed on the basis of
the CD spectra (see Supporting Information). Using a similar
procedure, racemic 58b with pyrazine-2-carboxylethoxyl at C5
position, racemic 58c with 2-hydroxylethoxyl at C5 position, and
racemic 58d and 58e with R1 and R2 fused as a six-membered ring
were also prepared for SARs studies (Scheme 9).
We desired to explore whether, under certain conditions
(for example, in the antitubercular assays), compounds such as
58a, having a 3-hydroxy-2-nitro-2,3-dihydrobenzofuran moiety
in place of the 2-nitrobenzofuran, could eliminate water in situ,
to give back a 2-nitrobenzofuran. Thus, additional compounds
blocking this elimination were prepared, such as compound 59,
where a methyl was introduced to the nitro-substituted carbon
atom of 58a to block elimination of the hydroxyl. 1-Bromo-
1-nitroethane was employed to obtain compound 59 as a
stereoisomeric mixture (Scheme 9).
3-isopropyl (or methyl)-2-nitro-2,3-dihydrobenzofuran 65a, 65b
showed no significant potency. These results may provide
circumstantial evidence that the unblocked 3-hydroxy-2-nitro-2,3-
dihydrobenzofuran analogues may eliminate water to give back
a nitrobenzofuran, although no experimental evidence has been
obtained to confirm such elimination (Figure 3).
Additional Aspects of Biological Activity of Synthetic
Calanolides. Select Calanolides Are Cidal against Replicat-
ing and Nonreplicating Mtb. The synthetic calanolides with
potent R and NR activity and acceptable selectivity indices were
chosen for further bactericidal evaluation. A colony-forming unit
(CFU) assay was employed to determine the impact of several
growth-inhibitory nitrofurano analogues of calanolides on wild-
type Mtb. Compounds 18 and 30a were bactericidal against both
R- and NR-Mtb. Compound 18 reduced survival of R-Mtb by 3.1
log10 at 0.6 μg/mL (2× MIC) and NR-Mtb by 2.6−4.5 log10 at
0.6 μg/mL (1× MIC) (Figure 4a,b). Similarly, compound 30a
On the basis of the structure of 40a, 2-nitroindole was
explored to replace the 2-nitrobenzofuran moiety yielding
compound 62. To do this, compound 57a was treated with
diphenyl azidophosphate to produce azide 60.24 To the best
of our knowledge, this method for azide preparation was first
applied to O-formyl phenols, such as compound 57a. Sub-
sequently, condensation between 60 and nitromethane
produced compound 61, which cyclized to nitroindole 62 in
refluxing xylene (Scheme 9).25
To evaluate the impact of the hydroxyl group in compounds
58a, compounds 65a/65b were synthesized to substitute the
hydroxyl with a methyl (65b) or isopropyl group (65a). Catalyzed
by ammonium acetate, compound 57a and nitromethane were
condensed to 63, which underwent Michael additions with
methylmagnesium bromide or isopropylmagnesium chloride to
produce 64a/64b, respectively. Upon treatment with hydrogen
peroxide and tetrabutylammonium iodide, 64a/64b produced
65a/65b (Scheme 9).23
To further substitute the furan ring in 40a, 64a/64b were
treated with iodobenzene diacetate and tetrabutylammonium
iodide to produce 66a/66b with isopropyl and acetoxylmethyl
substitutions,23 respectively, which may sterically block the reduc-
tion of the nitro group. Compound 66b underwent deacetyla-
tion under hydrogen chloride treatment to afford 67, having a
hydroxymethyl on the nitrobenzofuran ring (Scheme 9).
Figure 4. Nitrofurano analogues of calanolides kill R and NR-Mtb.
Compounds 18 (a, b), 30a (c, d) or DMSO were coincubated with
Mtb in R conditions (200 μL of 3.4 × 106/mL bacteria) (a, c) or NR
conditions (200 μL of 3 × 107/mL bacteria) (b, d). After 7−8 days,
bacteria were enumerated on 7H11 agar plates supplemented with
OADC and glycerol and incubated an additional ∼3 weeks, after which
colonies were counted. Results are means SD of triplicates of one
experiment representative of two.
Analogues obtained in the third round of design and syn-
thesis were assayed for their ability to kill R- and NR-Mtb
(Table 3). These data suggested that (1) the nitroindole
analogue, and analogues with substitution of isopropyl, acetoxy-
methyl, or hydroxymethyl at the C3 position of 2-nitro-
benzofuran such as compounds 62, 66a, 66b, and 67, showed
markedly reduced potency or no potency; (2) trans-3-hydroxy-
2-nitro-2,3-dihydrobenzofuran analogues 58a−58e had MICs <
1 μg/mL with LD50’s < 2. Among these compounds, activities of
the enantiomers (+)-58a and (−)-58a were surprisingly
equipotent. R6 as H, R7 as n-propyl (58a, 58b, 58c) and R6, R7
forming a six-membered ring (58d, 58e) had similar effects on
MIC and SI. When R9 was morpholino, hydroxyl, or pyrazine-
2-carboxyl, the resulting analogues had comparable MICs and
SIs; (3) 3-hydroxy-2-methyl-2-nitro-2,3-dihydrobenzofuran 59 and
reduced R-Mtb by 2.6 log10, at 0.6 μg/mL (1× MIC), and NR
bacilli by 2.1 log10, at 0.6 μg/mL (0.5−1× MIC) (Figure 4c,d),
when compared to the inoculum plated.
Calanolides Spare Gram-Positive/-Negative Bacteria and
Yeast. The activity of nitrofurano analogues of calanolide 30a
was determined against a panel of Gram-positive bacteria,
Gram-negative bacteria, and yeast. Despite its potent activity against
Mtb, 30a failed to kill Escherichia coli, Staphylococcus aureus,
Pseudomonas aeruginosa or Candida albicans at concentrations
≤25 μg/mL. Compound 30a was weakly active against a
nonpathogenic mycobacterial strain, Mycobacterium smegmatis
(∼20−70% inhibition of growth at 10 μg/mL) (Figure 5).
Calanolides Kill Mtb within Primary Human Macro-
phages. The majority of the intracellular Mtb in the infected
3766
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
Figure 5. Nitrofurano analogues of calanolides are selectively active against Mtb. Compound 30a, DMSO vehicle, or control antibiotics
(moxifloxacin, Meropenem, and rifampicin) were coincubated with log-phase Gram positive (a) or negative bacteria (b, c) or yeast (d) for 6−18 h.
At the end of the incubation, OD600 was measured. Results are means SD for triplicates, representative of two similar experiments. Mycobacterium
smegmatis (e) was coincubated with compounds indicated (30a and rifampicin) for 24−48 h. At the end of the incubation, OD580 was measured.
Results are means SD of triplicates, representative of two similar experiments.
host resides within macrophages.26 We evaluated the activity of
select synthetic calanolides for their ability to kill wild-type Mtb
infecting primary monocyte-derived human macrophages.27
Compound 18 reduced the number of Mtb in human
macrophages by ∼1.8 log10 at 1 μg/mL [3× MIC]. Compound
30a, which had minimal toxicity to HepG2 cells, reduced the
number of viable intracellular Mtb by 2−4 log10 (Figure 6). The
extent of bactericidal activity of these calanolides was greater
than that of the first-line TB drug, rifampicin (1−1.5 log10 killing
at 30× MIC, data not shown). Although compounds 39 and 40a
decreased Mtb viability in human macrophages by 1−2 log10 of
killing (data not shown), 30a was the most potent calanolide
tested against Mtb in the human macrophage model.
Certain Calanolides Are Genotoxic. The nitrofuran moiety
has been associated with DNA damage and is regarded as a major
structural alert in medicinal chemistry.14a Although compound 30a
and other select nitrofurano analogues of calanolides were either
weakly toxic or nontoxic to HepG2 cells, our concern regarding the
nitrofurano moiety prompted us to evaluate the DNA damaging
capacity of these compounds. Thus, compounds 30a and 40a were
subjected to genotoxicity testing in a mouse lymphoma assay
(kindly performed by GlaxoSmithKline) and, not surprisingly,
were found to be genotoxic (data not shown), precluding their
progression as lead compounds. However, the potent activity of
select synthetic calanolides against both R- and NR-Mtb, narrow
microbicidal spectrum, and potency against Mtb residing in human
macrophages make them attractive candidates as tool compounds
for target identification.
SUMMARY
■
Newly synthesized ring-D-modified analogues led to identi-
fication of several series of novel synthetic calanolides with
selective activity against R-Mtb and/or NR-Mtb. In particular,
analogues bearing 2-nitrofurano at the ring D position had
markedly improved in vitro efficacy and reduced mammalian
cell toxicity compared with the parent compound, (+)-calano-
lide A. For example, compound 39 had MIC values of 0.6 μg/mL
and 3 μg/mL against R-Mtb and NR-Mtb, respectively,
3767
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
spectra were recorded at 25 °C on 300, 400, or 500 MHz
spectrometers. 13C NMR spectra were at 25 °C on 500 MHz,
or 600 MHz spectrometers. Unless otherwise noted below, all
compounds were >95% pure by analytical HPLC.
6,6-Dimethyl-10-nitro-4-propyl-2H-5H-furo[2,3-h]-pyrano-
[2,3-f ]chromen-2-one (18). 5-Hydroxy-2,2-dimethyl-10-propyl-
8-oxo-2H,8H-pyrano[2,3-f]chromen-8-one-6-carbaldehyde
(1d) was synthesized by the reported procedure:11b 1H NMR
(300 MHz, DMSO) δ: 12.66 (s, 1H), 10.16 (s, 1H), 6.52
(d, J = 10.1, 1H), 6.11 (s, 1H), 5.78 (d, J = 10.1, 1H), 2.92−
2.70 (m, 2H), 1.63−1.42 (m, 8H), 0.97 (t, J = 7.3, 3H). 13C
NMR (75 MHz, DMSO) δ: 192.56, 160.58, 158.56, 158.41,
158.21, 157.97, 128.37, 114.60, 111.91, 105.25, 104.16, 102.80,
81.00, 38.29, 28.41, 23.54, 14.41. ESI-MS (m/z): 315 [M + H]+
(MW = 314.1).
Figure 6. Compound 30a kills wild-type Mtb infecting human
macrophages. Primary human macrophages were differentiated,
activated by IFN-gamma, and infected with wild-type Mtb at a
multiplicity of infection (MOI) of 0.1. After 24 h of infection,
compound or vehicle (DMSO) was added for an additional 7 days.
Macrophages were evaluated by microscopy for evidence of toxicity,
after which they were washed twice with PBS to remove residual
compound. Intracellular bacteria were released by lysing the
macrophages and enumerated by plating serial dilutions on 7H11
agar plates supplemented with OADC and glycerol. Colonies were
To a solution of the 1d (53 mg, 0.17 mmol) in 1.5 mL
anhydrous acetone was added K2CO3 (24 mg, 0.17 mmol) and
BrCH2NO2 (15 μL, 0.21 mmol) successively at room
temperature, and the reaction mixture was stirred for 3 h.
Upon completion as shown by TLC, the reaction mixture was
filtered and concentrated in vacuum. The residue was treated
with 1 mL of acetic anhydride and stirred for 1 h at 100 °C.
Upon completion as shown by TLC, the reaction mixture was
cooled, concentrated, and purified by column chromatography
on silica gel (petroleum ether/ethyl acetate) to provide the
compound 18: yellow solid (55 mg, 92% yield); mp =183−185 °C.
1H NMR (500 MHz, DMSO) δ: 8.35 (s, 1H), 6.75 (d, J = 9.9,
1H), 6.32 (s, 1H), 5.99 (d, J = 9.9, 1H), 3.05−2.77 (m, 2H), 1.70−
1.58 (m, 2H), 1.56 (s, 6H), 1.03 (t, J = 7.2, 3H). 13C NMR (126
MHz, DMSO) δ: 158.69, 157.97, 152.91, 150.46, 150.08, 130.73,
114.04, 113.99, 109.72, 107.67, 107.26, 103.00, 80.33, 38.10, 27.77,
23.38, 14.21. ESI-MS (m/z): 356 [M + H]+ (MW = 355.1).
HRMS (m/z): ([M + H]+) calcd 356.1134; found, 356.1142.
6,6-Dimethyl-10-nitro-4-propyl-7,8-dihydro-2H-5H-furo-
[2,3-h]-pyrano[2,3-f ]chromen-2-one (30a). 5-Acetoxy-2,2-
dimethyl-10-propyl-8-oxo-2H,8H-pyrano[2,3-f]chromen-8-one-
6-methylene diacetate (27a). To a solution of the 1d (300 mg,
0.95 mmol) in 3 mL of acetic anhydride was added 150 mg
of Amberlite IR-120 at room temperature, and the reaction
mixture was stirred for 1 h. Upon completion as shown by
TLC, the reaction mixture was filtered and concentrated in
vacuum. The residue was purified by column chromatography
on silica gel (petroleum ether/ethyl acetate) to provide the
counted after ∼3 weeks incubation. Results are means
triplicates of one of three similar experiments.
SD of
with LD50 values> 100 μg/mL to HepG2 cells. Similarly, com-
pound 40a was improved over (+)-calanolide A both in potency
(39-fold against R-Mtb; MIC90 = 0.08 μg/mL) and SI
(60-fold). Compound 22d had an MIC of 0.08 μg/mL (R-Mtb)
(a 40-fold improvement over calanolide A) and was nontoxic
to HepG2 cells, with an LD50 value of >50 μg/mL. Several of the
active analogues (40a, 30a) had potent bactericidal activity
against Mtb H37Rv residing in primary human macrophages.
Their eukaryotic genotoxicities preclude them from clinical
progression. However, the remarkable bactericidal properties of
these synthetic calanolides should enable their use as tool com-
pounds in identifying pathways essential for Mtb during replica-
tion, nonreplication, and infection of human macrophages. In
addition, these potent calanolides will aid further structure-based
design of more effective and drug-like antimycobacterial agents.
EXPERIMENTAL SECTION
■
General Chemical Methods. Details on the synthesis of all
chemical entities described in this article are in the Supporting
Information. All chemicals were purchased as reagent grade
and used without further purification unless otherwise noted.
Reactions were monitored by analytical thin-layer chromatog-
raphy on silica gel GF254 precoated on glass plates (10−40 μm,
Yantai, China). Spots were detected under UV (254 nm).
Solvents were evaporated under reduced pressure and below
45 °C (water bath). Column chromatography was performed
on silica gel (200−300 mesh, Qingdao, China). The automatic
LC-MS analysis was also performed on a Thermo Finnigan LCQ
Advantage mass spectrometer equipped with an Agilent HPLC
system and an eluent splitter (5% eluent was split into the MS
system). A Kromasil C18 column (4.6 μm, 4.6 mm × 50 mm)
from DIKMA was employed. The eluent was a mixture of
acetonitrile and water containing 0.05% formic acid with a linear
gradient from 5:95 v/v acetonitrile−H2O to 95:5% acetonitrile−
H2O within 10 min at 1 mL/min. The UV detection wavelength
was 254 nm. Mass spectra were recorded in either positive or
negative-ion mode using electrospray ionization. High-resolution
mass spectrometry (HRMS) was obtained using the time-of-flight
method. The ion source is electrospray ionization (ESI). 1H NMR
1
compound as a white solid (355 mg, 81% yield). H NMR
(300 MHz, CDCl3) δ: 8.09 (s, 1H), 6.21 (d, J = 9.9, 1H), 6.12
(s, 1H), 5.66 (d, J = 10.0, 1H), 2.93−2.87 (m, 2H), 2.42
(s, 3H), 2.08 (s, 6H), 1.70−1.62 (m, 2H), 1.51 (s, 6H), 1.04
(t, J = 7.2, 3H). ESI-MS (m/z): 459 [M + H]+ (MW = 458.2).
5-Acetoxy-3,4-dihydro-2,2-dimethyl-10-propyl-8-oxo-2H,8H-
pyrano[2,3-f]chromen-8-one-6-methylene diacetate (28a).
To the solution of 0.30 g 27a in 5 mL of ethanol was added
0.15 g of Pd/C, and the resulting mixture was stirred under
atmospheric pressure of H2 at room temperature for 1 h. Upon
completion as shown by TLC, the reaction mixture was filtered.
The filtrate was concentrated under reduced pressure, and the
residue was purified by column chromatography on silica gel
(petroleum ether/ethyl acetate) to provide the compound to
1
give a white solid (289 mg, 96% yield). H NMR (300 MHz,
DMSO) δ: 8.02 (s, 1H), 6.20 (s, 1H), 2.97−2.74 (m, 2H), 2.37
(s, 3H), 2.02 (s, 6H), 1.86−1.73 (m, 2H), 1.66−1.46 (m, 2H),
1.42−1.16 (m, 8H), 0.97 (t, J = 7.3, 3H). ESI-MS (m/z): 461
[M + H]+ (MW = 460.2).
3768
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
5-Hydroxy-3,4-dihydro-2,2-dimethyl-10-propyl-8-oxo-
2H,8H-pyrano[2,3-f]chromen-8-one-6-carbaldehyde (29a). To
a solution of 50 mg (0.16 mmol) of 28a in 1.5 mL of methanol
was added K2CO3 (66 mg, 0.478 mmol) at room temperature,
and the reaction mixture was stirred for 1 h. Upon completion
as shown by TLC, the reaction mixture was filtered and con-
centrated in vacuum. The residue was purified by column
chromatography on silica gel (petroleum ether/ethyl acetate)
to provide the compound as a white solid (32 mg, 91% yield).
1H NMR (300 MHz, CDCl3) δ: 12.93 (s, 1H), 10.40 (s, 1H),
6.01 (s, 1H), 2.98−2.79 (m, 2H), 2.70 (t, J = 6.8, 2H), 1.85
(t, J = 6.8, 2H), 1.73−1.48 (m, 2H), 1.44 (s, 6H), 1.03 (t, J =
7.3, 3H). ESI-MS (m/z): 317 [M + H]+ (MW = 316.1).
To a solution of 29a (25 mg, 0.08 mmol) in 1.5 mL of
anhydrous acetone was added K2CO3 (11 mg, 0.08 mmol) and
BrCH2NO2 (7 μL, 0.10 mmol) successively at room temper-
ature, and the reaction mixture was stirred for 3 h. Upon
completion as shown by TLC, the reaction mixture was filtered
and concentrated in vacuum. The residue was treated with
1 mL of acetic anhydride and stirred for 1 h at 100 °C. Upon
completion as shown by TLC, the reaction mixture was cooled,
concentrated, and purified by column chromatography on silica
gel (petroleum ether/ethyl acetate) to provide compound
30a, a yellow solid (12 mg, 45% yield); mp = 212−213 °C.
1H NMR (300 MHz, CDCl3) δ: 7.90 (s, 1H), 6.17 (s, 1H), 3.00
(m, 4H), 1.94 (t, J = 6.7, 2H), 1.67 (m, 2H), 1.48 (s, 6H), 1.05
(t, J = 7.3, 3H). 13C NMR (126 MHz, DMSO) δ: 158.95,
158.55, 154.31, 153.66, 152.64, 149.47, 113.81, 108.67, 107.80,
107.47, 102.33, 78.54, 38.56, 30.25, 26.63, 23.48, 16.43, 14.14.
ESI-MS (m/z): 358 [M + H]+ (MW = 357.1). HRMS (m/z):
([M + H]+) calcd 358.1291; found, 358.1292.
10,11-trans-6,6-Dimethyl-11-hydroxyl-10-nitro-4-propyl-
10,11-dihydro-2H-5H-furo[2,3-h]-pyrano[2,3-f ]chromen-2-
one (33a). To a solution of 1d (63 mg, 0.20 mmol) in 1.5 mL
of anhydrous acetone was added K2CO3 (28 mg, 0.20 mmol)
and BrCH2NO2 (22 μL, 0.30 mmol) successively at room
temperature, and the reaction mixture was stirred for 3 h. Upon
completion as shown by TLC, the reaction mixture was filtered
and concentrated in vacuum. The residue was triturated with
ethyl acetate to break up lumps, and the suspension was
filtered. The solid thus obtained was washed with ethyl acetate,
air-dried, and then dried under reduced pressure overnight to
give compound 33a: white solid (74 mg, 99% yield); mp =
145−150 °C. 1H NMR (300 MHz, CDCl3) δ: 6.38 (d, J = 9.9,
1H), 6.02 (s, 1H), 5.77 (s, 1H), 5.48 (m, 2H), 2.70 (t, J = 7.2,
2H), 1.38 (s, 3H), 1.33 (s, 3H), 0.85 (t, J = 7.1, 3H). ESI-MS
(m/z): 374 [M + H]+ (MW = 373.1). HRMS (m/z): ([M +
H]+) calcd 374.1240; found, 374.1239.
(m, 4H), 1.63 (m, 2H), 0.98 (t, J = 7.2, 3H). ESI-MS (m/z):
403 [M + H]+ (MW = 402.1). HRMS (m/z): ([M + H]+)
calcd 403.1505; found, 403.1511. 40b: red solid (7 mg, 17%
1
yield). H NMR (300 MHz, CDCl3) δ: 7.90 (s, 1H), 7.01
(s, 1H), 6.22 (s, 1H), 4.39 (s, 2H), 3.16 (s, 2H), 3.06−2.95
(m, 2H), 2.79 (s, 4H), 1.93 (s, 4H), 1.67 (m, 2H), 1.04 (t, J =
7.3, 3H). ESI-MS (m/z): 387 [M + H]+ (MW = 386.2). HRMS
(m/z): ([M + H]+) calcd 387.1556; found, 387.1558.
(8R,9R)-9-Hydroxy-5-(2-morpholinoethoxy)-8-nitro-4-
propyl-8,9-dihydrofuro[2,3-h]chromen-2-one ((+)-58a) and
(8S,9S)-9-Hydroxy-5-(2-morpholinoethoxy)-8-nitro-4-propyl-
8,9-dihydrofuro[2,3-h]chromen-2-one ((−)-58a). The com-
pound ( )-58a was synthesized as a white solid (69 mg, 59%
yield) from compound 57a (100 mg, 0.28 mmol) as done for
compound 33a from 1d. 1H NMR (500 MHz, acetone) δ: 6.89
(s, 1H), 6.54 (br, 1H), 6.04 (s, 1H), 5.77 (s, 1H), 4.56−4.19
(m, 2H), 3.68−3.66 (m, 4H), 3.13−3.01 (m, 2H), 2.92−2.90
(m, 2H), 2.57 (s, 4H), 1.74−1.70 (m, 2H), 1.05 (t, J = 7.4 Hz,
3H). ESI-MS (m/z): 421 [M + H]+ (MW = 420.2). HRMS
m/z: ([M + H]+) calcd 421.1611; found, 421.1619.
Sino-Chiral AD 0P10003-A was used to separate the
enantiomeric mixture of the enantiomers to provide (+)-58a
and (−)-58a. Chromatography condition: mobile phase =
isopropanol/n-hexane; flow velocity = 3 mL/min. Two fraction
peaks were observed eluting at 28.3 min (compound (−)-58a
[α]D20 = −16.7°) and 39.2 min (compound (+)-58a [α]D20
=
+14.3°). On the basis of the calculated and experimental CD
spectrum, the absolute configuration of compound (−)-58a
was proposed to be (8R, 9R) and compound (+)-58a was
proposed to be (8S, 9S) (see the Supporting Information for
the calculated and experimental CD spectrum).
Bacterial Strains and Growth Conditions. Mtb H37Rv
was grown in Middlebrook 7H9 supplemented with 0.2%
glycerol, 0.02% tyloxapol, and 10% ADN (albumin, dextrose,
NaCl) supplement. An attenuated Mtb strain mc26220
ΔpanCDΔlysA was used for HTS assays and grown in Middle-
brook 7H9 supplemented with 0.5% glycerol, 0.02% tyloxapol,
10% OADC (oleic acid, albumin, dextrose, catalase), 0.05%
casein hydrolysate (CAS) amino acids, 240 μg/mL L-lysine, and
24 μg/mL pantothenate.
For replicating (R) conditions, Mtb was propagated at 37 °C
with 20% O2 and 5% CO2. For nonreplicating (NR) conditions,
bacteria were incubated at 37 °C with 1% O2 and 5% CO2 in a
Sauton’s based medium (“NR medium”) consisting of 0.5 g of
MgSO4, 0.05 g of ferric ammonium citrate, 0.5 g of KH2PO4,
0.5% BSA, 0.085% NaCl, 0.02% tyloxapol, 50 μM butyrate, and
0.5 mM NaNO2 at pH 5.0. Strain Mtb mc26220 ΔpanCDΔlysA
(a kind gift of W. Jacobs, Jr., Albert Einstein College of
Medicine, New York) was grown with an additional 240 μg/mL
L-lysine and 24 μg/mL pantothenate.
Other bacterial strains were growth in Mycobacterium
smegmatis (Middlebrook 7H9 supplemented with 0.2% glycerol,
0.02% tyloxapol), Candida albicans (YM), uropathogenic
Escherichia coli (Luria Broth), Pseudomonas aeruginosa PAO1
(Luria Broth), and Staphyloccus aureus ATCC 29213 (Mueller-
Hinton broth). All of these strains were grown in a shaking
incubator at 37 °C at 200 rpm except for Candida albicans,
which was grown at 30 °C.
High-Throughput Screening. High throughput screening
(HTS) was performed as described previously with BSL2+-
rated attenuated strain of Mtb, mc26220 ΔpanCDΔlysA.5b
Briefly, to assay R activity, bacteria in mid log growth were
diluted to an OD580 of 0.01 and dispensed into 384-well plates
5-(2-Morpholinoethoxy)-8-nitro-4-propyl-2H-furo[2,3-h]-
chromen-2-one (40a) and 8-Nitro-4-propyl-5-(2-(pyrrolidin-1-
yl)ethoxy)-2H-furo[2,3-h]chromen-2-one (40b). To a solution
of 35 (30 mg, 0.10 mmol) and 2-morpholinoethanol (14 mg,
0.11 mmol) or 2-(pyrrolidin-1-yl)ethanol (13 mg, 0.11 mmol)
in 1.5 mL of THF was added PPh3 (35 mg, 0.13 mmol) and
diethyl diazenedicarboxylate (23 μL, 0.15 mmol) successively at
room temperature, and the reaction mixture was stirred for
12 h. Upon completion as shown by TLC, the reaction mixture
was concentrated in vacuum and purified by column
chromatography on silica gel (petroleum ether/ethyl acetate)
to provide the compound 40a as a brown solid (23 mg, 56%
1
yield); mp = 204−206 °C. H NMR (300 MHz, DMSO)
δ: 8.34 (s, 1H), 7.50 (s, 1H), 6.30 (s, 1H), 4.32 (m, 2H), 3.60
(m, 4H), 3.02 (t, J = 7.3, 2H), 2.80 (m, 2H), 2.50−2.35
3769
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
at 50 μL/well. Test compound (500 nL) in DMSO was added
to each well. After 7 days of co-incubation, the optical density
was determined at 580 nm. For NR activity, log phase bacteria
were washed 2× with PBS/0.02% tylaxopol and resuspended in
NR medium. Bacteria were diluted to an OD580 of 0.1, and
15 μL/well was dispensed into 384-well plates, and 150 nL of
test compound was added per well. After 3 days of co-incubation
at 37 °C with 1% O2 and 5% CO2, an outgrowth stage was
initiated by addition of 60 μL of R medium (effectively diluting
the test compound 5-fold) and incubation at 37 °C with 20% O2
and 5% CO2. The OD580 was read after 7 days of outgrowth.
MIC90 (minimal inhibitory concentration) is the measure of
activity of compound against Mtb, defined as 90% inhibition of
growth when compared to vehicle control (DMSO).
for viability. If drug-treated macrophages looked healthy, they
were lysed with 0.55% Triton-X for enumeration of intracellular
bacteria by plating on Middlebrook 7H11 agar plates con-
taining OADC supplement.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details for preparation of all new compounds are available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Authors
*E-mail: (C.F.N.) cnathan@med.cornell.edu. Tel.: 212-746-
2985.
■
Antibacterial Activity. For select compounds, as indicated
in the tables, activity was confirmed against wild-type Mtb
H37Rv in 96 well plates. The setup for this was similar to HTS
as described above, except the final volume was 200 μL per well
and during the outgrowth stage, instead of 1:5 dilution, a 1:21
dilution was performed into a new 96-well plate using 7H9-
ADN medium. To test selectivity against the panel of Gram-
negative and Gram-positive microorganisms, log phase bacteria
were diluted to OD 0.01 in their respective media, and 200 μL
of cells were dispensed per well of 96-well tissue-culture plates.
Two microliters of compound or DMSO vehicle control was
added. After drug addition, bacteria were incubated for 8−48 h
at 37 °C with shaking, after which the OD600 was determined.
Colony-forming units (CFU) assays were performed by plating
wild-type Mtb (200 μL at OD 0.01 or 0.1 for R and NR
conditions, respectively of a single-cell suspension in 96-well
plates and co-incubating with compound). After 7 days, bacteria
were enumerated by serially diluting in 7H9-AND and plating
onto Middlebrook 7H11 agar plates containing 0.5% glycerol
and 10% OADC.
Toxicity Assays. Human hepatoma HepG2 cells were
grown in Dulbecco’s Modified Eagle Medium (DMEM)
containing 10% fetal bovine serum (FBS) supplemented with
1 mM pyruvate, 2 mm glutamine, 10 mM HEPES, and 1%
nonessential amino acids. When the cell monolayer became
confluent, HepG2 cells were removed from the flask and
seeded at 3000 cells per well in a tissue-culture treated 384-well
Greiner plate. The mixture was incubated for 2 days after the
compound was added while keeping the final DMSO
concentration at 1%. The CellTiter-Glo kit from Promega was
used to measure ATP content of the cell represented by a
luminescent signal as an indicator of viability. LD50 was defined as
the concentration of compound that caused a 50% decrease in the
ATP signal compared to the DMSO control. The selectivity index
(SI) was calculated as LD50 divided by the MIC90.
Macrophage Assays. Human macrophages were generated
as described previously27 under an IRB-approved protocol.
Briefly, monocytes were isolated from healthy donors by CD14
positive selection using magnetic beads. For the assay, 100 000
cells/well were differentiated in 200 μL/well in 40% human
plasma containing 0.5 ng/mL human GM-CSF and 0.5 ng/mL
human TNF-alpha at 10% O2, 5% CO2. After 2 weeks, cells
were stimulated with 5 ng/mL IFN-gamma overnight
[ACTIMMUNE] and infected with wild-type Mtb at multi-
plicity of infection (MOI) of 0.1. After 24 h of infection, either
DMSO vehicle control or test compound was added,
maintaining the final DMSO concentration at 1%. Infected
human macrophages containing DMSO or drug were then
coincubated for 7 days. Macrophages were visually inspected
*E-mail: (B.G.) bsg2001@med.cornell.edu. Tel.: 646-962-6209.
*E-mail: (G.L.) gangliu27@tsinghua.edu.cn. Tel.: +86 10
62797740.
Author Contributions
#P.Z. and S.S.K. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported by the Bill & Melinda Gates
Foundation, TB Drug Accelerator Phase II: Hit generation for
tuberculosis drug discovery, Global Health Grant OPP1024029,
and by the Milstein Program in Chemical Biology and
Translational Medicine. The Department of Microbiology
and Immunology is supported by the William Randolph Hearst
Foundation. We thank Dr. Li Li (Institute of Materia Medica,
Peking Union Medical College) for insightful discussions about
CD results of compounds (+)-58a and (−)-58a. S.S.K. was
partially supported by the training grant T32 AI007613. We
thank Amy Cunningham Bussel (Weill Cornell Medical
College) for sharing her expertise in preparation of human
macrophages. We thank Khisimuzi Mdluli (Global Alliance for
TB Drug Discovery) and Sharon Robinson, Elenea Jimenez-
Navarro, and Alfonso Mendoza-Losana (GlaxoSmithKline) for
their generosity and assistance with the genotoxicity assays.
ABBREVIATIONS
■
AIDS, acquired immunodeficiency syndrome; BINAP, 2,2′-
bis(diphenylphosphino)-1,1′- binaphthalene; Boc, tert-butoxy-
carbonyl; BOP, (benzotriazol-1-yloxy)tris(dimethylamino)-
phosphonium hexafluorophosphate; DBU, 1,8-diazabicyclo-
[5.4.0]undec-7-ene; DCC, dicyclohexylcarbodiimide; DEAD,
diethyl azodicarboxylate; DIB, iodobenzene diacetate; DMAP,
4-dimethylaminopyridine; DME, 1,2-dimethoxyethane; DMF,
dimethylformamide; HIV, human immunodeficiency virus;
HOBt, hydroxybenzotriazole; INH, isoniazid; LD50, median lethal
dose; MDR, multidrug-resistant; MIC, minimum inhibitory
concentration; NIS, N-iodosuccinimide; NMR, nuclear magnetic
resonance; NR-Mtb, nonreplicating Mycobacterium tuberculosis;
RIF, rifampicin; R-Mtb, replicating Mycobacterium tuberculosis;
SAR, structure−activity relationship; SI, selective index; TB,
tuberculosis; TBAI, tetrabutylammonium iodide; TFA, trifluoro-
acetic acid; THF, tetrahydrofuran
REFERENCES
■
(1) World Health Organization. Global tuberculosis control: WHO
report 2010. WHO Press: Geneva, Switzerland, 2010.
3770
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
(2) For recent reviews on TB: (a) Gandhi, N. R.; Nunn, P.; Dheda,
K.; Schaaf, H. S.; Zignol, M.; van Soolingen, D.; Jensen, P.; Bayona, J.
Multidrug-resistant and extensively drug-resistant tuberculosis: a threat
to global control of tuberculosis. Lancet 2010, 375, 1830−1843.
(b) Koul, A.; Arnoult, E.; Lounis, N.; Guillemont, J.; Andries, K. The
challenge of new drug discovery for tuberculosis. Nature 2011, 469,
483−490. (c) Ma, Z.; Lienhardt, C.; McIlleron, H.; Nunn, A. J.; Wang,
X. Global tuberculosis drug development pipeline: the need and the
reality. Lancet 2010, 375, 2100−2109. (d) Gutierrez-Lugo, M.; Bewley,
C. A. Natural products, small molecules, and genetics in tuberculosis
drug development. J. Med. Chem. 2008, 51, 2606−2612. (e) Froth-
ingham, R.; Stout, J. E.; Hamilton, C. D. Current issues in global
tuberculosis control. Int. J. Infect. Dis. 2005, 9, 297−311.
(3) (a) Xie, Z.; Siddiqi, N.; Rubin, E. J. Differential antibiotic
susceptibilities of starved Mycobacterium tuberculosis isolates. Anti-
microb. Agents Chemother. 2005, 49, 4778−4780. (b) Nathan, C. Fresh
approaches to anti-infective therapies. Sci. Transl. Med. 2012, 4, 140sr2.
(4) Nathan, C.; Gold, B.; Lin, G.; Stegman, M.; de Carvalho, L. P.;
Vandal, O.; Venugopal, A.; Bryk, R. A philosophy of anti-infectives as a
guide in the search for new drugs for tuberculosis. Tuberculosis 2008,
88, S25−S33.
(5) (a) Bryk, R.; Gold, B.; Venugopal, A.; Singh, J.; Samy, R.; Pupek,
K.; Cao, H.; Popescu, C.; Gurney, M.; Hotha, S.; Cherian, J.; Rhee, K.;
Ly, L.; Converse, P. J.; Ehrt, S.; Vandal, O.; Jiang, X.; Schneider, J.; Lin,
G.; Nathan, C. Selective killing of nonreplicating Mycobacteria. Cell
Host Microbe 2008, 3, 137−145. (b) Gold, B.; Pingle, M.; Brickner, S.
J.; Shah, N.; Roberts, J.; Rundell, M.; Bracken, W. C.; Warrier, T.;
Somersan, S.; Venugopal, A.; Darby, C.; Jiang, X.; Warren, J. D.;
Fernandez, J.; Ouerfelli, O.; Nuermberger, E. L.; Cunningham-Bussel,
A.; Rath, P.; Chidawanyika, T.; Deng, H.; Realubit, R.; Glickman, J. F.;
Nathan, C. Nonsteroidal anti-inflammatory drug sensitizes Mycobacte-
rium tuberculosis to endogenous and exogenous antimicrobials. Proc.
Natl. Acad. Sci. U.S.A. 2012, 109, 16004−16011. (c) Mak, P. A.; Rao, S.
P.; Ping Tan, M.; Lin, X.; Chyba, J.; Tay, J.; Ng, S. H.; Tan, B. H.;
Cherian, J.; Duraiswamy, J.; Bifani, P.; Lim, V.; Lee, B. H.; Ling Ma,
N.; Beer, D.; Thayalan, P.; Kuhen, K.; Chatterjee, A.; Supek, F.;
Glynne, R.; Zheng, J.; Boshoff, H. I.; Barry, C. E., 3rd; Dick, T.; Pethe,
K.; Camacho, L. R. A high-throughput screen to identify inhibitors of
ATP homeostasis in non-replicating Mycobacterium tuberculosis. ACS
Chem. Biol. 2012, 7, 1190−1197.
(6) (a) Kashman, Y.; Gustafson, K. R.; Fuller, R. W.; Cardellina, J. H.,
II; McMahon, J. B.; Currens, M. J.; Buckheit, R. W., Jr; Hughes, S. H.;
Cragg, G. M.; Boyd, M. R. The calanolides, a novel HIV-inhibitory
class of coumarin derivatives from the tropical rainforest tree,
calophyllum lanigerum. J. Med. Chem. 1992, 35, 2735−2743.
(b) Eiznhamer, D. A.; Creagh, T.; Ruckle, J. L.; Tolbert, D. T.;
Giltner, J.; Dutta, B.; Flavin, M. T.; Jenta, T.; Xu, Z. Safety and
pharmacokinetic profile of multiple escalating doses of (+)-calanolide
A, a naturally occurring nonnucleoside reverse transcriptase inhibitor,
in healthy HIV-negative volunteers. HIV Clin. Trials 2002, 3, 435−
450.
active (+)-inophyllum B and (+)-calanolide A by application of
(−)-quinine-catalyzed intramolecular oxo-Michael addition. J. Org.
Chem. 2004, 69, 2760−2767. (e) Khilevich, A.; Mar, A.; Flavin, M. T.;
Rizzo, J. D.; Dzekhtser, S.; Brankovic, D.; Lin, L.; Zhang, H.; Chen,
W.; Liao, S.; Zembower, D. E.; Xu, Z. Q. Synthesis of (+)-calanolide A,
an anti-HIV agent, via enzyme-catalyzed resolution of the Aldol
products. Tetrahedron: Asymmetry 1996, 7, 3315−3326.
(10) (a) Ma, T.; Liu, L.; Xue, H.; Li, L.; Han, C.; Wang, L.; Chen, Z.;
Liu, G. The chemical library and structure−activity relationship of 11-
demethyl-12-oxo calanolide A analogues against HIV-1. J. Med. Chem.
2008, 51, 1432−1446. (b) Xue, H.; Lu, X.; Zheng, P.; Liu, L.; Han, C.;
Hu, J.; Liu, Z.; Ma, T.; Li, Y.; Wang, L.; Chen, Z.; Liu, G. Highly
suppressing wild-type HIV-1 and Y181C mutant HIV-1 strains by 10-
chloromethyl-11-demethyl-12-oxo-calanolide A with druggable profile.
J. Med. Chem. 2010, 53, 1397−1401. (c) Lu, X.; Liu, L.; Zhang, X.;
Lau, T. C.; Tsui, S. K.; Kang, Y.; Zheng, P.; Zheng, B.; Liu, G.; Chen,
Z. F18, a novel small molecule nonnucleoside reverse transcriptase
inhibitor, inhibits HIV-1 replication using distinct binding motifs as
demonstrated by resistance selection and docking analysis. Antimicrob.
Agents Chemother. 2012, 56, 341−351.
(11) (a) Zhang, J.; Kirchhoff, E. W.; Zembower, D. E.; Jimenez, N.;
Sen, P.; Xu, Z. Q.; Flavin, M. T. Automated process research: an
example of accelerated optimization of the Friedel−Crafts acylation
reaction, a key step for the synthesis of anti-HIV (+)-calanolide A. Org.
Process Res. Dev. 2000, 4, 577−580. (b) Deshpande, P. P.; Tagliaferri,
F.; Victory, S. F.; Yan, S.; Baker, D. C. Synthesis of optically active
calanolide A and B. J. Org. Chem. 1995, 60, 2964−2965. (c) Fox, M.
E.; Lennon, I. C.; Meek, G. A novel synthesis of 5-hydroxy-2,2-
dimethyl-10-propyl-2H-pyrano[2,3-f ]chromen-8-one. Tetrahedron
Lett. 2002, 43, 2899−2902.
(12) Fan, R.; Sun, Y.; Ye, Y. Iodine(III)-mediated tandem
acetoxylation−cyclization of o-acyl phenols for the facile construction
of α-acetoxybenzofuranones. Org. Lett. 2009, 11, 5174−5177.
(13) Zheng, P.; Lu, S.; Liu, G. An unexpected C−C cleavage
reaction: new and mild access to o-OH and o-NH−Ts benzoic acids or
benzoamides. Mol. Diversity 2011, 15, 971−977.
(14) (a) Vass, M.; Hruska, K.; Franek, M. Nitrofuran antibiotics: a
review on the application, prohibition and residual analysis. Vet. Med.
2008, 53, 469−500. (b) Genevieve, L.; Pierre, D.; Rene, R.; Philippe,
G.; Jacky, F. Synthesis and microorganism-inhibiting effects of 3-
methoxy-2-nitrobenzofuran having a monomethoxy group on the
benzene ring. Eur. J. Med. Chem. (Chimie therapeutique) 1986, 21,
379−383.
(15) Zhou, L.; Ishizaki, H.; Spitzer, M.; Taylor, K. L.; Temperley, N.
D.; Johnson, S. L.; Brear, P.; Gautier, P.; Zeng, Z.; Mitchell, A.;
Narayan, V.; McNeil, E. M.; Melton, D. W.; Smith, T. K.; Tyers, M.;
Westwood, N. J.; Patton, E. E. ALDH2 mediates 5-nitrofuran activity
in multiple species. Chem. Biol. 2012, 19, 883−892.
(16) Debnath, A. K.; Lopez de Compadre, R. L.; Debnath, G.;
Shusterman, A. J.; Hansch, C. Structure-activity relationship of
mutagenic aromatic and heteroaromatic nitro compounds. Correlation
with molecular orbital energies and hydrophobicity. J. Med. Chem.
1991, 34, 786−797.
(17) Royer, R. R.; Demerseman, P. G.; Rene, L. J.; Cavier, R. E. 2-
Nitro-5 or 6 or 7-methoxy benzofuran. U.S. Patent No. 3,838,173 A,
1974.
(7) Xu, Z. Q.; Barrow, W. W.; Suling, W. J.; Westbrook, L.; Barrow,
E.; Lin, Y. M.; Flavin, M. T. Anti-HIV natural product (+)-calanolide A
is active against both drug-susceptible and drug-resistant strains of
Mycobacterium tuberculosis. Bioorg. Med. Chem. 2004, 12, 1199−1207.
(8) Xu, Z. Q.; Pupek, K.; Suling, W. J.; Enache, L.; Flavin, M. T.
Pyranocoumarin, a novel anti-TB pharmacophore: synthesis and
biological evaluation against Mycobacterium tuberculosis. Bioorg. Med.
Chem. 2006, 14, 4610−4626.
(9) (a) Chenera, B.; West, M. L.; Finkelstein, J. A.; Dreyer, G. B.
Total synthesis of (+)-calanolide A, a non-nucleoside inhibitor of HIV-
1 reverse transcriptase. J. Org. Chem. 1993, 58, 5605−5606.
(b) Deshpande, P. P.; Tagliaferri, F.; Victory, S. F.; Yan, S.; Baker,
D. C. Synthesis of optically active calanolides A and B. J. Org. Chem.
1995, 60, 2964−2965. (c) Trost, B. M.; Toste, F. D. A catalytic
enantioselective approach to chromans and chromanols. A total
synthesis of (−)-calanolide A and B and the vitamin E nucleus. J. Am.
Chem. Soc. 1998, 120, 9074−9075. (d) Sekino, E.; Kumamoto, T.;
Tanaka, T.; Ikeda, T.; Ishikawa, T. Concise synthesis of anti-HIV-1
(18) Plantan, I.; Selic, L.; Mesar, T.; Anderluh, P. S.; Oblak, M.;
Prezelj, A.; Hesse, L.; Andrejasic, M.; Vilar, M.; Turk, D.; Kocijan, A.;
Prevec, T.; Vilfan, G.; Darko, K.; Copar, A.; Urleb, U.; Solmajer, T. 4-
Substituted trinems as broad spectrum β-lactamase inhibitors:
structure-based design, synthesis, and biological activity. J. Med.
Chem. 2007, 50, 4113−4121.
(19) Butts, C. P.; Eberson, L.; Gonzals-Luque, R.; Hartshorn, C. M.;
Hartshorn, M. P.; Merchan, M.; Robinson, W. T.; Roos, B. O.;
Vallance, C.; Wood, B. R. Photochemical nitration by tetranitro-
methane. Part XL. Regiochemistry of trinitromethyl attachment in the
photolysis of benzofuran with tetranitromethane. Acta Chem. Scand.
1997, 51, 984−999.
3771
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772

Journal of Medicinal Chemistry
Article
(20) Chang, C. W.; Chein, R. J. Absolute configuration of anti-HIV-1
agent (−)-Concentricolide: total synthesis of (+)-(R)-Concentricolide.
J. Org. Chem. 2011, 76, 4154−4157.
(21) Koyanagi, J.; Yamamoto, K.; Nakayama, K.; Tanaka, A. A new
synthetic route to 2-substituted naphtho[2,3-b]furan-4,9-dione. J.
Heterocycl. Chem. 1997, 34, 407−412.
(22) Padwa, A.; Dimitroff, M.; Waterson, A. G.; Wu, T. Diels-Alder
reaction of 2-amino-substituted furans as a method for preparing
substituted anilines. J. Org. Chem. 1997, 62, 4088−4096.
(23) Lu, S.; Zheng, P.; Liu, G. Iodine(III)-mediated tandem oxidative
cyclization for construction of 2-nitrobenzo[b]furans. J. Org. Chem.
2012, 77, 7711−7717.
(24) Hisato, T.; Shoji, E. Synthesis of novel carbo- and heteropoly-
cycles. Part 9. Benzylic azide functionalization of tautomeric o-
acylbenzoic acid, -benzamide, and -nicotinic acid derivatives, and
thermal decomposition of the derived azides. J. Chem. Soc., Perkin
Trans. 1 (1972−1999) 1988, 8, 2149−2154.
(25) Pelkey, E. T.; Gribble, G. W. Synthesis of 2-nitroindoles via the
Sundberg indole synthesis. Tetrahedron Lett. 1997, 38, 5603−5606.
(26) Glickman, M. S.; Jacobs, W. R., Jr. Microbial pathogenesis of
Mycobacterium tuberculosis: dawn of a discipline. Cell 2001, 104, 477−
485.
(27) Vogt, G.; Nathan, C. In vitro differentiation of human
macrophages with enhanced antimycobacterial activity. J. Clin. Invest.
2011, 121, 3889−3901.
3772
dx.doi.org/10.1021/jm4019228 | J. Med. Chem. 2014, 57, 3755−3772