Novel Schiff-Base-Derived FabH Inhibitors with Dioxygenated Rings as Antibiotic Agents

Article abstract of DOI:10.1002/cmdc.201200587

Fatty acid biosynthesis plays a vital role in bacterial survival and several key enzymes involved in this biosynthetic pathway have been identified as attractive targets for the development of new antibacterial agents. Of these promising targets, β-ketoacyl-acyl carrier protein (ACP) synthaseIII (FabH) is the most attractive target that could trigger the initiation of fatty acid biosynthesis and is highly conserved among Gram-positive and -negative bacteria. Designing small molecules with FabH inhibitory activity displays great significance for developing antibiotic agents, which should be highly selective, nontoxic and broad-spectrum. In this manuscript, a series of novel Schiff base compounds were designed and synthesized, and their biological activities were evaluated as potential inhibitors. Among these 21 new compounds, (E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)hexadecan-1-amine (10) showed the most potent antibacterial activity with a MIC value of 3.89-7.81μM^-1 against the tested bacterial strains and exhibited the most potent E.coli FabH inhibitory activity with an IC₅₀ value of 1.6μM. Docking simulation was performed to position compound 10 into the E.coli FabH active site to determine the probable binding conformation.

Full text of DOI:10.1002/cmdc.201200587

CHEMMEDCHEM
FULL PAPERS
DOI: 10.1002/cmdc.201200587
Novel Schiff-Base-Derived FabH Inhibitors with
Dioxygenated Rings as Antibiotic Agents
Yang Zhou,[a] Qian-Ru Du,[a] Jian Sun,[a] Jing-Ran Li,[a] Fei Fang,[a] Dong-Dong Li,[a]
Yong Qian,[a] Hai-Bin Gong,*[b] Jing Zhao,[a] and Hai-Liang Zhu*[a]
Fatty acid biosynthesis plays a vital role in bacterial survival
and several key enzymes involved in this biosynthetic pathway
have been identified as attractive targets for the development
of new antibacterial agents. Of these promising targets, b-ke-
toacyl-acyl carrier protein (ACP) synthase III (FabH) is the most
attractive target that could trigger the initiation of fatty acid
biosynthesis and is highly conserved among Gram-positive and
-negative bacteria. Designing small molecules with FabH inhib-
itory activity displays great significance for developing antibi-
otic agents, which should be highly selective, nontoxic and
broad-spectrum. In this manuscript, a series of novel Schiff
base compounds were designed and synthesized, and their
biological activities were evaluated as potential inhibitors.
Among these 21 new compounds, (E)-N-((3,4-dihydro-2H-
benzo[b][1,4]dioxepin-7-yl)methylene)hexadecan-1-amine (10)
showed the most potent antibacterial activity with a MIC value
of 3.89–7.81 mmÀ1 against the tested bacterial strains and ex-
hibited the most potent E. coli FabH inhibitory activity with an
IC50 value of 1.6 mm. Docking simulation was performed to po-
sition compound 10 into the E. coli FabH active site to deter-
mine the probable binding conformation.
Introduction
Fatty acid synthesis (FAS) is an essential metabolic process for
the viability and growth of cells in prokaryotic organisms. Con-
trolling this pathway in some key enzymes or proteins is a val-
uable route for developing new antibacterial agents. Bacteria
generate fatty acids (components of phospholipids) through
type II fatty acid synthesis (FAS II), which generally includes
two stages: initiation and elongation (Scheme 1). This is in con-
trast to mammalian type I fatty acid synthase (FAS I), which is
a multi-enzyme complex process, with each of the reactions
performed by a separate enzyme. Thus, targeting one or two
dominant catalytic units in the FAS II pathway represents a rea-
sonable tactic in the development of new antibiotics. For ex-
ample, the two natural products, cerulenin and thiolactomycin
(Figure 1), are inhibitors of the condensing enzymes both FabB
and FabF, and both compounds show antibacterial activity.[1] In
addition, the bacterial FAS system and proteins bear little ho-
Scheme 1. FabH-catalyzed initiation reaction of fatty acid biosynthesis. Ab-
breviations: acyl carrier protein (ACP).
[a] Dr. Y. Zhou,+ Dr. Q.-R. Du,+ Dr. J. Sun, Dr. J.-R. Li, Dr. F. Fang, Dr. D.-D. Li,
Dr. Y. Qian, Prof. J. Zhao, Prof. H.-L. Zhu
State Key Laboratory of Pharmaceutical Biotechnology, Nanjing University
Hankou Road, Nanjing 210093 (P.R. China)
E-mail: zhuhl@nju.edu.cn
Figure 1. Several reported antibacterial agents targeting the fatty acid syn-
thesis (FAS) pathway: cerulenin,[1] thiolactomycin,[1] 2-tosylnaphthalene-1,4-
diol (2008),[6] benzylidene derivatives (e.g., 6a),[7] diethyl sulfonamide deriva-
tives (e.g., B82),[5] and [3-phenoxybenzoylamino]benzoic acid derivatives
(e.g., B83).[5]
[b] Dr. H.-B. Gong
Xuzhou Central Hospital, Xuzhou 221009 (P.R. China)
[+] These authors contributed equally to this work.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 433

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
mology to the human system, indicating that type II-selective
inhibitors could be identified with high probability.[1]
Among these enzymes, b-ketoacyl-acyl carrier protein (ACP)
synthase III (FabH) plays an essential role in the biosynthesis of
bacterial fatty acids, as it initiates the FAS cycle by catalyzing
the first condensation step between acetyl-CoA and malonyl-
ACP (Scheme 1). Moreover, FabH proteins from both Gram-
positive and -negative bacteria are highly conserved at the se-
quence and structural level. When considering bacteria resist-
ant to known therapies, FabH should become a more attrac-
tive target for the development
tification of protein active sites and analysis of receptor–ligand
interactions are the two important steps for generating phar-
macophore maps. These are mainly composed of sets of inter-
actions (chemical features or functionalities), aligned in three-
dimensional space, and include several features along with ex-
cluded volume regions derived from receptor atom posi-
tions.[17,18] In this manuscript, we firstly attempt to design four
novel Schiff bases (1–3, 5) derived from YKAs3003[19] that main-
tain the structure of the cyclohexylamine moiety. The design
strategy and experimental flowchart are shown in Scheme 2:
of new antibacterial agents than
other important FAS enzymes .
Several optimized FabH inhibi-
tors presented in Figure 1 have
been optimized for the enzyme–
inhibitor complex,[2,3] and ac-
cordingly, their structure–activity
relationships (SARs) have been
described over many years.[4–7]
We start our research based on
these related studies.
Recently, Brinster et al.[8] re-
ported that, when the culture
medium is supplemented with
fatty acids or human serum,
FAS II is not essential in Strepto-
coccus agalactiae (Lacetobacil-
Scheme 2. Design strategy: a lead compound (YKAs3003) initiates a series of novel antibiotic agents.
lales). This not only challenged
the general knowledge on tar-
geting FAS II in Gram-positive
pathogens but also triggered an explosive debate on whether
these observations with S. agalactiae could be reasonably gen-
eralized for all Gram-positive or, further, Gram-negative bacte-
ria. So far, the views of this issue are essentially divided into
two opposite opinions, and a majority tends to the conserva-
tive side, which is supported by the finding by Balemans
et al.[9] that fatty acids were unable to rescue Staphylococcus
aureus when treated with an enoyl-ACP reductase (FabI) inhibi-
tor. Moreover, Joshua et al.[10] point out that important details
on how Gram-positive bacteria uptake and use the exogenous
fatty acids are still unclear, and these could not explain various
reports on the efficacy of inhibitors of fatty acid synthesis
against S. aureus in mouse models.[11–13] Interestingly, they did
not construct an S. aureus FabH knockout model that was
a fatty acid auxotroph, and this might leave an unknown role
for FabH in the S. aureus strain when using exogenous fatty
acids. Therefore, much work towards drug development in the
the molecular docking of the five compounds (including
YKAs3003) was performed using the CDOCKER protocol,[20] and
receptor-oriented pharmacophore model based on Escherichia
coli FabH protein complex (PDB: 1HNJ) could be generated by
using the Discovery Studio software.[20] The combined analysis
of the docking study and pharmacophore screening shows
that compound 5 with the heptane ring is most suitable as
a lead compound for an optimized scaffold for the design of
new FabH inhibitors. The designed inhibitors on the basis of
compound 5 were synthesized and evaluated against two
Gram-negative bacterial strains, E. coli and Pseudomonas fluo-
rescens, and two Gram-positive bacterial strains, Bacillus subtilis
and S. aureus. In addition, several potent antibacterial agents
were subsequently assessed for the corresponding FabH inhib-
itory activity in vitro.
pathway of bacterial fatty acid synthesis remains. Our study is Results and Discussion
based on the presumption that disturbing the FAS cycle would
Molecular docking and development of a FabH
influence cell viability and growth in prokaryotic organisms.
Molecular docking and structure-based pharmacophore in
silico screening are two effective methods to identify novel
and specific ligands or fragments. Because the in silico screen-
ing process can systematically analyze all possible binding
modes between a compound and protein in the active site, it
is an applied methodology of rational drug design.[14–16] Iden-
pharmacophore model
The synthesis of the four kinds of benzaldehyde analogues
with dioxygenated rings has been described in our early
study,[21] all of which were promising building modules in the
assembly of the newly designed compounds. Moreover, Lee
et al.[19] reported YKAs3003 (Scheme 2) to be a potent inhibitor
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 434

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
of pathogenic KAS III, performing well against various bacteria.
Taken together, four novel compounds (1–3, and 5) were de-
signed, and subsequently together with YKAs3003, a docking
study was performed by using the CDOCKER algorithm.[20] As
default, each compound generally retained 10 preferential con-
formations, which would generate 50 poses in all. Next, we
ranked all 50 poses by CDOCKER interaction energy, and the
corresponding result is illustrated in Figure 2, showing the
Figure 3. The training set, previously described compounds 6a–g,[7] used by
HipHop.
Figure 2. The CDOCKER interaction energy of the five active-like com-
pounds: the lead compound YKAs3003 and four designed molecules (1–3,
5) docked into the active site of E. coli FabH receptor (PDB: 1HNJ)[22]; the
result shows that two compounds 3 and 5 (energy range generally from
À30 kcalmolÀ1 to À24 kcalmolÀ1) perform better than the other three com-
pounds.
values of the high active group were set to two, whereas the
medium active group was set to one. The next step was to
pick up and refine the feature mapping derived from analysis
of molecular docking conformation in the binding site. The in-
struction of common feature pharmacophore was implement-
ed and 10 kinds of b-ketoacyl-ACP synthase III that ranked by
RANK_score had been obtained, the first one of which was
46.712. As shown in Figure 4, this pharmacophore model, con-
taining one ring and two acceptor features, could be selected
as an effective tool to screen the docking poses of the five
molecules (1–3, 5, and YKAs3003) in the further study.
energy values of the five small molecules in the range of À30
to À20 kcalmolÀ1. Overall, compounds 3 and 5 gave better re-
sults than the other three compounds. Therefore, the dioxy-
genated rings that belonged to compound 1 and 2 were es-
sentially ruled out, and the eight- and seven-membered rings
located at the left position of the designed molecules were an-
alyzed further. In order to determine which of the two dioxy-
genated rings the left bulk of the designed molecules was de-
rived from, we attempted to build a FabH pharmacophore
model based on the E. coli FabH protein complex that binds
the substrate molecule malonyl-CoA.
In this construction process, seven small molecules
(Figure 3) and malonyl-CoA were selected as the training set to
generate the pharmacophore model with common feature.
Since the Catalyst software is integrated into the Discovery
Studio 3.1 platform,[20] the pharmacophore generation method
used could also be called “HipHop” derived from Catalyst. As
previously described, we accomplished the docking study of
these seven small compounds[5] and chose the appropriate
conformation before the pharmacophore model was generat-
ed. As to the training set, these compounds were classified
into two groups: high active (6c–e, and substrate malonyl-
COA) and medium active (6a,b,f,g) according to their antimi-
crobial activity. In the treatment with HipHop, a “Principal”
column that described the active level of the training set was
added into the attribute of the compounds and the “Principal”
Figure 4. E. coli FabH protein (PDB: 1HNJ)[22] receptor-oriented pharmaco-
phore model. The procedures of construction based on the training set in-
clude: molecular docking of the training set molecules, and common feature
pharmacophore generation by using the Discovery Studio 3.1.[20] This map
contains one aromatic ring and two acceptors.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 435

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
All 50 poses from the five compounds could be collected
into a small database, and clicking “Search, Screen and Profile”,
and inputting corresponding poses and finally the screening
result including twenty matching poses would be obtained.
The top five poses are listed in Table 1, and show that com-
tures of these compounds were fully characterized by spectro-
scopic methods and elemental analysis.
Biological activity
All synthesized compounds (5–25) were evaluated for antibac-
terial activity against the two Gram-negative bacterial strains,
E. coli and P. fluorescence, and the two Gram-positive bacterial
strains, B. subtilis and S. aureus, using an MTT (3-(4,5-dime-
thylhttp://en.wikipedia.org/wiki/Thiazole-2-yl)-2,5-dihttp://en.-
wikipedia.org/wiki/Phenyltetrazolium bromide) assay. The mini-
mum inhibitory concentrations (MIC) of these compounds
against these bacteria are presented in Table 2. The antibacteri-
al agent kanamycin B (positive control) was also screened
under the same condition, and the results revealed that most
of the small molecules exhibited a certain antibacterial activity.
All tested compounds were divided into three groups accord-
ing to their biological activity: high activity (8–10, 15, 23),
medium activity (5, 7, 12–14, 17, 19, 20, 21, 24, 25), and low
activity (6, 11, 16, 18, 22). All compounds in the high-activity
group demonstrated excellent inhibitory activity against Gram-
negative E. coli and P. fluorescence, and Gram-positive B. subtilis
and S. aureus. Especially, compound 10 showed comparable
activities to positive control kanamycin B (MIC=3.89, 7.81,
7.81, and 3.89 mm À1, respectively; Table 2). Most small mole-
cules of the medium-activity group showed modest antibacte-
rial activities against the four pathogenic strains, and the re-
maining compounds of the last group had little or no bioactiv-
ity for any strains.
Table 1. Docking screen by combination of the CDOCKER protocol and
b-ketoacyl-ACP synthase III mapping.
Compd
Fit value[a]
Energy
[kcalmolÀ1
[b]
]
YKAs3003
1.81484
1.57282
1.55467
1.46776
1.16376
À21.8319
À28.4496
À30.3326
À25.4376
À26.9283
5
3
2
1
[a] Results from screen of b-ketoacyl-ACP synthase III pharmacophore
model. [b] CDOCKER interaction energy; results of the docking analysis.
pound 3 was slightly better than compound 5 in the “Fit
value” evaluation, whereas 5 actually overmatched 3 in the in-
teraction energy. Considering the high flexibility of the eight-
membered ring derived from 3 and the high activity of FabH
inhibitor with a seven-membered ring,[21] compound 5 was se-
lected as the designed molecule template. A series of Schiff
bases (5–25) were designed, and their relative biological activi-
ties were investigated.
Chemistry
The synthetic route for prepar-
ing the targeted compounds 5–
25 is outlined in Scheme 3.
These compounds were synthe-
sized from intermediate 4, which
had been synthesized as previ-
ously described.[21] In order to
develop more effective antibiot-
ics agents and conveniently dis-
cuss the different trends in struc-
ture–activity relationships (SARs)
of the designed compounds,
substituents on the right-hand
part of compounds were re-
placed from previously used ali-
phatic amine groups (5–11) to
aromatic amine groups (12–25).
Thus, the designed compounds
were prepared by a dehydration
reaction of the intermediate al-
dehyde 4 with different amines
dissolved in ethanol. The crude
products were purified with
flash chromatography, and com-
pounds 5–25 were obtained
Scheme 3. General synthesis of the Schiff base derivatives 5–21. Reagents and conditions: a) Cs2CO3, DMF, 708C,
with 80–95% yields. The struc- overnight; b) RÀNH2, EtOH, RT, overnight.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 436

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
The E. coli FabH inhibitory potencies of the synthetic com-
pounds with potent antibacterial activities (6–10, 13, 15, 23–
25) were examined, and the results are listed in Table 3. All
Table 2. Antibacterial activity of synthesized compounds.
Compd[a]
AlogP[b]
MIC [mgmLÀ1
Gram-negative
]
Gram-positive
P. aeruginosa B. subtilis S. aureus
E. coli
5
6
7
8
3.556
3.026 >100
25
50
>100
50
>100
50
>100
>100
25
12.5
6.25
1.56
>100
Table 3. E. coli FabH inhibitory activity of synthesized compounds.
Compd[a]
E. coli FabH
IC50 [mm]
Hemolysis LC[b]
PSA[c]
[ꢁ2]
AlogP[c]
3.939
4.851
6.676
8.501
2.629
3.281
3.265
4.351
4.223
3.659 >100
2.463 50
3.274 >100
3.939
3.480
3.258
12.5
12.5
6.25
3.13
3.13
[mgmLÀ1
]
3.13
3.13
1.56
6.25
6.25
3.13
9
6
7
8
9
10
13
15
23
24
25
9.3
7.4
3.3
2.1
1.6
15.5
6.8
5.9
12.7
4.4
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
32.43
30.82
30.82
30.82
30.82
41.66
32.43
30.82
44.79
43.71
2.923
3.939
4.851
6.676
8.501
3.161
4.12
4.023
3.13
3.327
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
50
>100
>100
25
25
50
25
12.5
50
50
12.5
>100
25
>100
50
>100
25
25
25
50
12.5
6.25
12.5
>100
50
>100
>100
100
>100
12.5
50
100
12.5
50
50
12.5
25
50
50
50
>100
[a] The compounds tested for antibacterial activity are consistent with
the description in the Experimental Section. [b] Lytic concentration 30%.
[c] Molecular polar surface area (PSA); calculated with Discovery studio
3.1.[20]
2.257 >100
>100
>100
4.023
3.234
3.327
6.25
25
12.5
6.25
50
25
3.13
25
12.5
3.13
25
25
Kanamycin B À7.144
3.13
3.13
1.56
1.56
[a] The compounds tested for antibacterial activity are consistent with
the description in the Experimental Section. [b] Calculated with Discovery
studio 3.1.[20]
tested compounds showed potent E. coli FabH inhibition, and
particularly compounds 8–10 with aliphatic substituents
showed considerable inhibitory activities (IC50 <4 mm). With the
length of the aliphatic chain from compound 7 to 10, the in-
hibitory activities increased, and this result supports the
potent antibacterial activities of 10 being the most potent
FabH inhibitor. Compounds 6, 13 and 24 actually exhibited
moderate inhibitory activities but very poor antibacterial activi-
ties. This disparity could be explained by their AlogP values
(Table 3). Compared with the other tested compounds, 6, 13
and 24 have lower AlogP values, which could indicated these
small molecules do not easily penetrate the bacterial cell mem-
brane. The molecular polar surface areas (PSA) of these com-
pounds were in the range of 30–44 ꢁ2, and therefore, this mo-
lecular property cannot significantly account for the results ob-
tained. In summary, the results of the E. coli FabH inhibitory ac-
tivity of the test compounds described above generally corre-
sponded to the SARs of their antibacterial activities. This
demonstrates that the potent antibacterial activities of the se-
lected compounds are probably correlated to their FabH inhib-
itory activities.
In order to describe the SARs of these compounds, this
series was classified into two groups: compounds with aliphat-
ic amine side chains (5–11) and those with aromatic amine
substituents (12–25). Compared with the aromatic group, the
aliphatic group could contribute more to the antibiotic activity
of the compounds, maybe due to their hydrophobic proper-
ties. Although compounds 6 and 11 exhibited poor activity,
the reason for this may lie in the steric clash located in the
binding site of the target protein rather than in the hydropho-
bic nature of the aliphatic amine substituents. Moreover, in-
creasing the length of the aliphatic chain seems to enhance
the potency of compounds as seen in the MIC values from
compound 6 to 10 (see Table 2). Correspondingly, the predict-
ed lipophilicity (AlogP) values of these four molecules rise in
the same order. In the aromatic group, some compounds pos-
sessed effective antibacterial activities and others maintained
lower activities, which indicates that the two factors (substitu-
ents on the benzene ring, and length of the linker between
the benzene ring and N atom) affect their antibacterial activity.
As shown in Table 2, one- or two-carbon-atom gaps between
the benzene ring and N atom were favorable for the relative
biological activity when comparing compounds 12–15 with
17–21. However, compound 23 represents an exception, prob-
ably because the bromine substituent could make up for loss
of activity caused by the length of the linker. As to the sub-
stituent of the benzene ring, trifluoromethyl performed better
than other moieties (compound 15). The last noticeable dis-
covery from compounds 24 and 25 was that the heterocycle
replacing the benzene ring was another strategy to explore
better antibacterial agents.
With the aim to determine the interaction binding mode be-
tween the target protein and small molecules, molecular dock-
ing of the most active compound 10 and E. coli FabH was per-
formed on the binding model based on the E. coli FabH–CoA
complex structure (PDB: 1HNJ).[22] The corresponding result is
shown in Figure 5, which is composed of three interaction
maps. The docking study was performed using the CDOCKER
protocol.[20] The 3D optimal conformation and the 2D diagram
interacting with the FabH active site are presented in Figure 5a
and b, respectively. The two amino acid residues Arg151 and
Trp32 each form p–p interactions with compound 10, making
them vital for the stabilization of its binding mode. Figure 5c
shows the Ludi interaction map of the binding pocket inside
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 437

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
compound 10. In addition, the assay data of E. coli FabH inhibi-
tory activity also indicates that compound 10 is a potential in-
hibitor of E. coli FabH with potent antibacterial activity.
Conclusions
In this paper, 21 novel Schiff base derivatives (5–25) were syn-
thesized and for the first time evaluated for E. coli FabH inhibi-
tory activity as antibacterial agents. Within this research, sever-
al compounds with potent and selective anti-Gram-negative
bacteria activities were obtained. Compounds 8–10, and 25 ex-
hibited excellent activities against Gram-negative E. coli and
P. aeruginosa. Particularly, compound 10 showed the most
potent E. coli FabH inhibitory activity with an IC50 value of
1.6 mm, which was compared with the positive control, kana-
mycin B. Molecular docking simulation was performed to posi-
tion compound 10 into the E. coli FabH active site in order to
determine the probable binding conformation. Based on the
data obtained in this study, we conclude that compound 10 is
the E. coli FabH inhibitor most deserving of further research as
a potential antibiotic. Moreover, the development of similar
novel compounds based on the structure of compound 3 will
be investigated in future studies.
Experimental Section
Chemistry
General: All chemicals (reagent grade) were purchased from
Sigma–Aldrich. Separation of the compounds by column chroma-
tography was carried out with silica gel 60 (200–300 mesh ASTM,
Merck). The quantity of silica gel used was 50–100 times the
weight charged on the column. Then, the eluates were monitored
using thin-layer chromatography (TLC). Melting points (mp) were
determined on a XT4 MP apparatus (Taike Corp., Beijing, China).
ESI mass spectra (MS) were obtained on a Mariner System 5304
1
mass spectrometer, and H and 13C NMR spectra were recorded on
a Bruker DPX 400 spectrometer. Elemental analyses were per-
formed on a Heraeus CHN-O-Rapid instrument and were within
Æ0.4% of the theoretical values.
Figure 5. Three interaction maps of compound 10 derived from the docking
studies that display: a) the 3D diagram, b) the 2D diagram, c) the Ludi inter-
action map. In panels a and b, the models predict the formation of a p–p in-
teraction with residue Arg151 and a p–cation bond with residue Trp32, and
as such, the aliphatic chain of compound 10 is easily inserted into the hy-
drophobic domain in the active center, which could favor binding. The inter-
action map shown in panel c is directly involved in the pharmacophore anal-
ysis of the ligand–protein complex (PDB: 1HNJ), which could search three
basic features in the protein active pocket (hydrogen bond acceptor labeled
as red sticks, hydrogen bond donor labeled as blue sticks, and the hydro-
phobic site labeled as gray point). The figure shows a defined spatial net-
work in which compound 10 and the natural substrate malonyl-CoA can
both circumvent the hydrophobic points and match the corresponding
donor or accepter sites well.
3,4-dihydro-2H-benzo[b][1,4]dioxepine-7-carbaldehyde
(4):
The synthesis was performed as previously described.[21] 3,4-Dihy-
droxybenzaldehyde (0.716 g, 5.2 mmol) and Cs2CO3 (3.389 g,
10.4 mmol) were added to a 25 mL round flask. Dry DMF (8 mL),
and 1,3-dibromopropane (2.1 g, 10.4 mmol) was added, and the re-
action mixture was heated to 708C overnight. After cooling to RT
and filtration (Celite), the filtrate was concentrated in vacuo. The
crude product was purified by column chromatography (SiO2,
EtOAc/hexane), and compound 4 was obtained as a colorless oil
(814 mg, 88%): Rf =0.6 (EtOAc/hexane, 1:4); 1H NMR (400 MHz,
CDCl3): d=9.85 (s, 1H), 7.47 (dd, J=7.2, 1.8 Hz, 2H), 7.06 (m, 1H),
4.36 (m, 2H), 4.30 (t, J=5.9 Hz, 2H), 2.26 ppm (m, 2H); 13C NMR
(100 MHz, CDCl3): d=190.5, 156.4, 151.0, 131.9, 125.4, 122.7, 121.8,
70.3, 70.2, 30.7 ppm; MS (ESI): m/z 179.2 [M+H]+; Anal. calcd for
C10H10O3: C 67.41, H 5.66, O 26.94, found: C 67.32, H 5.68, O 26.96.
the protein receptor, mainly depicting the interaction environ-
ment around the existing substrate malonyl-CoA. Strikingly,
the molecular binding mode demonstrates that compound 10
could mimic the original crystal conformation of substrate ma-
lonyl-CoA. This supports the understanding of the most active
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)cy-
clohexanamine (5): Compound 4 (535 mg, 3 mmol) was dissolved
in MeOH (3 mL), and cyclohexanamine (3.6 mmol, 357 mg) was
added. The reaction mixture was stirred at RT overnight. The sol-
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 438

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
vent was removed in vacuo, and the crude product was purified
by column chromatography (Neutral Al2O3, EtOAc/hexane) to give
compound 5 as a yellow solid (700 mg, 90%): Rf =0.6 (EtOAc/
0.88 ppm (t, J=6.8 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=159.5,
153.1, 151.1, 132.0, 123.2, 121.6, 121.1, 70.4, 61.7, 31.9, 31.5, 31.0,
29.7, 29.6, 29.5, 29.4, 27.4, 22.7, 14.1 ppm; MS (ESI): m/z 402.6 [M+
H]+; Anal. calcd for C26H43NO2: C 77.75, H 10.79, N 3.49, found: C
77.38, H 10.62, N 3.84.
1
hexane, 1:4); mp: 68.0–68.98C; H NMR (400 MHz, CDCl3): d=8.19
(s, 1H), 7.36 (d, J=2.0 Hz, 1H), 7.28 (m, 1H), 6.97 (d, J=8.2 Hz, 1H),
4.23 (m, 4H), 3.15 (m, 1H), 2.20 (m, 2H), 1.82 (m, 2H), 1.68 (m, 2H),
1.56 (m, 2H), 1.31 ppm (m, 4H); 13C NMR (100 MHz, CDCl3): d=
157.5, 153.1, 151.1, 132.3, 123.3, 121.5, 121.1, 70.4, 70.4, 69.8, 34.4,
31.5, 25.7, 24.8 ppm; MS (ESI): m/z 260.3 [M+H]+; Anal. calcd for
C16H21NO2: C 74.10, H 8.16, N 5.40, found: C 74.10, H 8.16, N 5.40.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-2-
methylpropan-2-amine (11): Compound 11 was obtained as
a yellow oil (270 mg, 91%): Rf =0.6 (EtOAc/hexane, 1:6); 1H NMR
(400 MHz, CDCl3): d=8.15 (s, 1H), 7.35 (d, J=2.0 Hz, 1H), 7.28 (m,
1H), 6.98 (d, J=8.2 Hz, 1H), 4.24 (m, 4H), 2.21 (dd, J=11.4, 5.7 Hz,
2H), 1.35 ppm (s, 9H); 13C NMR (100 MHz, CDCl3): d=159.5, 153.1,
151.1, 132.0, 123.2, 121.6, 121.1, 70.4, 31.5 ppm; MS (ESI): m/z
334.1 [M+H]+; Anal. calcd for C14H19NO2: C 72.07, H 8.21, N 6.00,
found: C 72.07, H 8.21, N 6.00.
Compounds 6–25 were synthesized as described for compound 5
(see above) with the corresponding amine.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)bu-
tan-1-amine (6): Compound 6 was obtained as a yellow oil
(558 mg, 94%): Rf =0.6 (EtOAc/hexane, 1:5); 1H NMR (400 MHz,
CDCl3): d=8.15 (s, 1H), 7.35 (d, J=2.0 Hz, 1H), 7.28 (m, 1H), 6.98
(d, J=8.2 Hz, 1H), 4.24 (m, 4H), 3.57 (m, 2H), 2.21 (m, 2H), 1.66 (m,
2H), 1.37 (m, 2H), 0.94 ppm (t, J=7.4 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=159.6, 153.2, 151.1, 132.0, 123.2, 121.6, 121.1, 70.4, 61.3,
33.0, 31.5, 20.4, 13.9 ppm; MS (ESI): m/z 234.3 [M+H]+; Anal. calcd
for C14H19NO2: C 72.07, H 8.21, N 6.00, found: C 72.23, H 8.11, N
6.07.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-1-
phenylmethanamine (12): Compound 12 was obtained as
a yellow oil (454 mg, 86%): Rf =0.6 (EtOAc/hexane, 1:4); 1H NMR
(400 MHz, CDCl3): d=8.29 (s, 1H), 7.45 (t, J=5.4 Hz, 1H), 7.36 (m,
5H), 7.28 (dd, J=7.0, 2.5 Hz, 1H), 7.01 (dd, J=8.2, 1.6 Hz, 1H), 4.80
(s, 2H), 4.25 (m, 4H), 2.20 ppm (m, 2H); 13C NMR (100 MHz, CDCl3):
d=160.9, 153.5, 151.2, 139.5, 131.8, 128.5, 128.0, 127.0, 123.6,
121.6, 121.4, 70.4, 64.9, 31.5 ppm; MS (ESI): m/z 268.3 [M+H]+;
Anal. calcd for C17H17NO2: C 76.38, H 6.41, N 5.24, found: C 76.42, H
6.62, N 5.07.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-
hexan-1-amine (7): Compound 7 was obtained as a yellow oil
(408 mg, 92%): Rf =0.6 (EtOAc/hexane, 1:5); 1H NMR (400 MHz,
CDCl3): d=8.13 (s, 1H), 7.34 (d, J=2.0 Hz, 1H), 7.27 (m, 1H), 6.96
(d, J=8.2 Hz, 1H), 4.21 (m, 4H), 3.54 (dd, J=10.1, 3.9 Hz, 2H), 2.18
(m, 2H), 1.65 (m, 2H), 1.31 (t, J=7.1 Hz, 6H), 0.87 ppm (t, J=
6.7 Hz, 3H); 13C NMR (100 MHz, CDCl3): d=159.5, 153.1, 151.1,
132.0, 123.2, 121.5, 121.1, 70.4, 61.6, 31.7, 31.5, 30.9, 27.0, 22.6,
14.1 ppm; MS (ESI): m/z 262.3 [M+H]+; Anal. calcd for C16H23NO2:
C 73.53, H 8.87, N 5.36, found: C 73.92, H 9.07, N 4.93.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-1-
(4-methoxyphenyl)methanamine (13): Compound 13 was ob-
tained as a yellow solid (290 mg, 90%): Rf =0.7 (EtOAc/hexane,
1
1:4); mp:134.4–135.28C; H NMR (400 MHz, CDCl3): d=8.36 (s, 1H),
7.56 (s, 1H), 7.48 (dd, J=8.3, 1.9 Hz, 1H), 7.20 (d, J=8.1 Hz, 2H),
7.14 (d, J=8.1 Hz, 2H), 7.05 (d, J=8.2 Hz, 1H), 4.29 (m, 4H), 2.39 (s,
3H), 2.24 ppm (m, 4H); 13C NMR (100 MHz, CDCl3): d=158.4, 153.9,
151.2, 149.5, 135.6, 132.0, 129.8, 124.2, 121.8, 120.8, 115.3, 70.4,
70.4, 31.4, 21.0, 20.5 ppm; MS (ESI): m/z 298.4 [M+H]+; Anal. calcd
for C18H19NO3: C 72.71, H 6.44, N 4.71, found: C 72.53, H 6.79, N
4.47.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)oc-
tan-1-amine (8): Compound 8 was obtained as a yellow oil
(294 mg, 94%): Rf =0.7 (EtOAc/hexane, 1:8); 1H NMR (400 MHz,
CDCl3): d=8.13 (s, 1H), 7.34 (d, J=2.0 Hz, 1H), 7.27 (m, 1H), 6.96
(d, J=8.2 Hz, 1H), 4.22 (m, 4H), 3.55 (m, 2H), 2.19 (dd, J=11.4,
5.7 Hz, 2H), 1.66 (m, 2H), 1.29 (m, 10H), 0.87 ppm (t, J=6.8 Hz,
3H); 13C NMR (100 MHz, CDCl3): d=159.5, 153.1, 151.1, 132.0,
123.2, 121.5, 121.1, 70.4, 61.6, 31.9, 31.5, 31.0, 29.4, 29.3, 27.3, 22.7,
14.1 ppm; MS (ESI): m/z 290.4 [M+H]+; Anal. calcd for C18H27NO2:
C 74.70, H 9.40, N 4.84, found: C 74.61, H 9.48, N 4.92.
(E)-2-(4-bromophenyl)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-
7-yl)methylene)ethanamine (14): Compound 14 was obtained as
a white solid (522 mg, 91%): Rf =0.7 (EtOAc/hexane, 1:4); mp:
1
141.7–142.58C; H NMR (400 MHz, CDCl3): d=8.03 (s, 1H), 7.39 (m,
2H), 7.33 (d, J=2.0 Hz, 1H), 7.25 (m, 1H), 7.10 (d, J=8.4 Hz, 2H),
6.98 (d, J=8.2 Hz, 1H), 4.26 (m, 4H), 3.79 (m, 2H), 2.95 (t, J=
7.2 Hz, 2H), 2.21 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=
160.5, 153.3, 151.1, 139.0, 131.7, 131.4, 130.8, 123.3, 121.6, 121.1,
119.9, 70.4, 62.7, 36.9, 31.5 ppm; MS (ESI): m/z 361.3 [M+H]+; Anal.
calcd for C18H18BrNO2: C 60.01, H 5.04, N 3.89, found: C 60.33, H
5.36, N 4.01.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)do-
decan-1-amine (9): Compound 9 was obtained as a yellow oil
(331 mg, 93%): Rf =0.8 (EtOAc/hexane, 1:8); 1H NMR (400 MHz,
CDCl3): d=8.14 (s, 1H), 7.35 (d, J=2.0 Hz, 1H), 7.28 (m, 1H), 6.97
(d, J=8.2 Hz, 1H), 4.23 (m, 4H), 3.55 (m, 2H), 2.19 (m, 2H), 1.66 (m,
2H), 1.29 (m, 18H), 0.88 ppm (t, J=6.8 Hz, 3H); 13C NMR (100 MHz,
CDCl3): d=159.5, 153.1, 151.1, 132.0, 123.2, 121.5, 121.1, 70.4, 61.6,
31.9, 31.5, 31.0, 29.7, 29.6, 29.6, 29.5, 29.3, 27.3, 22.7, 14.1 ppm; MS
(ESI): m/z 346.5 [M+H]+; Anal. calcd for C22H35NO2: C 76.47, H
10.21, N 4.05, found: C 76.63, H 11.33, N 3.97.
N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-1-(4-
(trifluoromethyl)phenyl)methanamine (15): Compound 15 was
obtained as a yellow solid (546 mg, 89%): Rf =0.5 (EtOAc/hexane,
1
1:4); mp: 136.2–137.88C; H NMR (400 MHz, CDCl3): d=8.30 (s, 1H),
7.60 (d, J=8.1 Hz, 2H), 7.46 (m, 3H), 7.36 (dd, J=8.3, 2.0 Hz, 1H),
7.01 (d, J=8.2 Hz, 1H), 4.82 (s, 2H), 4.26 (m, 4H), 2.21 ppm (m,
2H); 13C NMR (100 MHz, CDCl3): d=161.6, 153.7, 151.2, 143.8,
131.5, 128.1, 125.4, 125.4, 125.3, 123.7, 121.7, 121.3, 70.4, 70.4, 64.2,
31.4 ppm; MS (ESI): m/z 336.3 [M+H]+; Anal. calcd for C18H16F3NO2:
C 64.47, H 4.81, N 4.18, found: C 64.62, H 4.52, N 4.39.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-
hexadecan-1-amine (10): Compound 10 was obtained as a pale
yellow solid (365 mg, 95%): Rf =0.7 (EtOAc/hexane, 1:10); mp:
1
51.0–52.08C; H NMR (400 MHz, CDCl3): d=8.15 (s, 1H), 7.35 (d, J=
2.0 Hz, 1H), 7.28 (m, 1H), 6.98 (d, J=8.2 Hz, 1H), 4.24 (m, 4H), 3.56
(m, 2H), 2.21 (dd, J=11.4, 5.7 Hz, 2H), 1.67 (m, 2H), 1.29 (m, 26H),
(R,E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-
1-phenylethanamine (16): Compound 16 was obtained as
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 439

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
a yellow oil (477 mg, 92%): Rf =0.6 (EtOAc/hexane, 1:4); 1H NMR
(400 MHz, CDCl3): d=8.26 (s, 1H), 7.43 (dd, J=9.1, 4.7 Hz, 3H), 7.35
(m, 3H), 7.25 (m, 1H), 6.98 (d, J=8.2 Hz, 1H), 4.51 (q, J=6.6 Hz,
1H), 4.25 (m, 4H), 2.21 (m, 2H), 1.58 ppm (t, J=5.2 Hz, 3H);
13C NMR (100 MHz, CDCl3): d=158.3, 153.3, 151.1, 145.4, 132.1,
128.6, 128.4, 126.8, 126.6, 123.6, 121.5, 121.2, 70.4, 69.6, 31.5,
24.9 ppm; MS (ESI): m/z 282.4 [M+H]+; Anal. calcd for C18H19NO2:
C 76.84, H 6.81, N 4.98, found: C 76.99, H 6.72, N 4.65.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-1H-
pyrazol-3-amine (22): Compound 22 was obtained as a white
solid (370 mg, 81%): Rf =0.3 (EtOAc/hexane, 1:4); mp: 89.4–90.88C;
1H NMR (400 MHz, CDCl3): d=8.61 (s, 2H), 8.55 (s, 1H), 7.44 (d, J=
2.0 Hz, 1H), 7.37 (dd, J=8.3, 2.0 Hz, 1H), 6.99 (d, J=8.3 Hz, 1H),
4.27 (m, 4H), 2.21 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=
156.3, 155.0, 151.2, 138.3, 126.6, 124.5, 122.1, 121.7, 70.4, 70.3,
31.0 ppm; MS (ESI): m/z 244.3 [M+H]+; Anal. calcd for C13H13N3O2:
C 64.19, H 5.39, N 17.27, Found: C 64.38, H 5.27, N 17.22.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-1-
(furan-2-yl)methanamine (17): Compound 17 was obtained as
a yellow solid (364 mg, 84%): Rf =0.5 (EtOAc/hexane, 1:4); mp:
(E)-4-bromo-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)me-
thylene)aniline (23): Compound 23 obtained as a yellow solid
(420 mg, 87%); Rf =0.5 (EtOAc/hexane, 1:4); mp: 102.3–103.48C;
1H NMR (400 MHz, CDCl3): d=8.29 (s, 1H), 7.53 (d, J=2.0 Hz, 1H),
7.46 (m, 3H), 7.05 (t, J=8.9 Hz, 3H), 4.28 (m, 4H), 2.23 ppm (m,
2H); 13C NMR (100 MHz, CDCl3): d=159.6, 154.2, 151.2, 151.0,
132.2, 131.5, 124.4, 122.6, 121.9, 121.8, 119.1, 70.4, 70.4, 31.3 ppm;
MS (ESI): m/z 333.2 [M+H]+; Anal. calcd for C16H14BrNO2: C 57.85,
H 4.25, N 4.22, found: C 57.85, H 4.25, N 4.22.
1
154.2–155.08C; H NMR (400 MHz, CDCl3): d=8.22 (s, 1H), 7.39 (dd,
J=10.0, 1.5 Hz, 2H), 7.33 (dd, J=8.3, 2.0 Hz, 1H), 6.98 (d, J=8.3 Hz,
1H), 6.34 (dd, J=3.0, 1.9 Hz, 1H), 6.25 (d, J=3.0 Hz, 1H), 4.73 (s,
2H), 4.23 (m, 4H), 2.19 ppm (m, 2H); 13C NMR (100 MHz, CDCl3):
d=161.9, 153.5, 152.5, 151.1, 142.1, 131.5, 123.6, 121.6, 121.4,
110.4, 107.4, 70.4, 57.1, 31.4 ppm; MS (ESI): m/z 258.3 [M+H]+;
Anal. calcd for C15H15NO3: C 70.02, H 5.88, N 5.44, found: C 70.12, H
5.94, N 54.21.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-2-
(3H-indol-2-yl)ethanamine (24): Compound 24 was obtained as
a white solid (450 mg, 82%): Rf =0.4 (EtOAc/hexane, 1:4); mp:
168.7–169.38C; 1H NMR (400 MHz, CDCl3): d=8.04 (d, J=12.8 Hz,
2H), 7.67 (d, J=7.7 Hz, 1H), 7.35 (d, J=6.3 Hz, 2H), 7.28 (d, J=
10.6 Hz, 1H), 7.20 (t, J=7.3 Hz, 1H), 7.13 (t, J=7.3 Hz, 1H), 6.99 (m,
2H), 4.25 (m, 4H), 3.91 (t, J=7.1 Hz, 2H), 3.16 (t, J=7.2 Hz, 2H),
2.21 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=160.4, 153.3,
151.2, 136.3, 131.9, 127.6, 123.4, 122.3, 121.9, 121.6, 121.2, 119.2,
119.0, 114.0, 111.2, 70.5, 61.9, 31.5, 27.0 ppm; MS (ESI): m/z 321.4
[M+H]+; Anal. calcd for C20H20N2O2: C 74.98, H 6.29, N 8.74, found:
C 75.04, H 6.33, N 8.51.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)ani-
line (18): Compound 18 was obtained as a yellow solid (548 mg,
91%): Rf =0.5 (EtOAc/hexane, 1:4); mp: 92.5–93.48C; 1H NMR
(400 MHz, CDCl3): d=8.29 (s, 1H), 7.45 (t, J=5.4 Hz, 1H), 7.36 (m,
5H), 7.28 (dd, J=7.0, 2.5 Hz, 1H), 7.01 (dd, J=8.2, 1.6 Hz, 1H), 4.25
(m, 4H), 2.20 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=160.9,
153.5, 151.2, 139.5, 131.8, 128.5, 128.0, 127.0, 123.6, 121.6, 121.4,
70.4, 31.5 ppm; MS (ESI): m/z 254.3 [M+H]+; Anal. calcd for
C16H15NO2: C 75.87, H 5.97, N 5.53, found: C 75.66, H 5.83, N 5.72.
(E)-4-chloro-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)me-
thylene)aniline (19): Compound 19 was obtained as a yellow solid
(390 mg, 88%): Rf =0.5 (EtOAc/hexane, 1:4); mp: 101.2–101.78C;
1H NMR (400 MHz, CDCl3): d=8.31 (s, 1H), 7.54 (d, J=2.0 Hz, 1H),
7.47 (dd, J=8.3, 2.0 Hz, 1H), 7.34 (m, 2H), 7.13 (m, 2H), 7.04 (d, J=
8.3 Hz, 1H), 4.29 (m, 4H), 2.24 ppm (m, 2H); 13C NMR (100 MHz,
CDCl3): d=159.5, 154.2, 151.2, 150.5, 131.3, 129.2, 124.4, 122.2,
121.9, 121.8, 116.2, 70.4, 70.4, 31.3 ppm; MS (ESI): m/z 288.7 [M+
H]+; Anal. calcd for C16H14ClNO2: C 66.79, H 4.90, N 4.87, found: C
66.88, H 4.97, N 4.56.
(E)-5-chloro-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)me-
thylene)pyridin-2-amine (25): Compound 25 was obtained as
a white solid (334 mg, 87%): Rf =0.4 (EtOAc/hexane, 1:4); mp:
1
121.0–121.48C; H NMR (400 MHz, CDCl3): d=9.43 (s, 1H), 7.60 (d,
J=1.8 Hz, 1H), 7.54 (m, 3H), 7.25 (m, 1H), 7.02 (d, J=8.3 Hz, 1H),
4.30 (m, 4H), 2.24 ppm (m, 2H); 13C NMR (100 MHz, CDCl3): d=
164.9, 155.6, 155.3, 151.1, 130.4, 125.7, 122.8, 122.6, 122.0, 115.0,
70.3, 70.3, 31.0 ppm; MS (ESI): m/z 289.7 [M+H]+; Anal. calcd for
C15H13ClN2O2: C 62.40, H 4.54, N 9.70, found: C 62.36, H 4.75, N
9.63.
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-4-
fluoroaniline (20): Compound 20 was obtained as a yellow solid
(382 mg, 85%): Rf =0.5 (EtOAc/hexane, 1:4); mp: 101.7–109.88C;
1H NMR (400 MHz, CDCl3): d=8.31 (s, 1H), 7.53 (d, J=2.0 Hz, 1H),
7.46 (dd, J=8.3, 2.0 Hz, 1H), 7.34 (m, 2H), 7.12 (m, 2H), 7.04 (d, J=
8.3 Hz, 1H), 4.29 (m, 4H), 2.24 ppm (p, J=5.7 Hz, 2H); 13C NMR
(100 MHz, CDCl3): d=159.5, 154.2, 151.2, 150.6, 131.5, 131.2, 129.2,
124.3, 122.2, 121.9, 121.8, 70.4, 70.4, 31.3 ppm; MS (ESI): m/z 272.3
[M+H]+; Anal. calcd for C16H14FNO2: C 70.84, H 5.20, N 5.16, found:
C 70.61, H 5.28, N 5.33.
Biology
Bacterial suppressive assay: The antibacterial activity of the synthe-
sized compounds was tested against E. coli ATCC 25922, P. aerugi-
nosa ATCC 27853, S. aureus ATCC 6538 and B. subtilis ATCC 530
(kindly provided by pro Chang-Hong Liu, State Key Laboratory of
Pharmaceutical Biotechnology, Nanjing University, Nanjing) using
Mueller-Hinton medium (MH medium: 17.5 g casein hydrolysate,
1.5 g soluble starch, 1000 mL beef extract). The minimum inhibito-
ry concentration (MIC) values of the tested compounds were deter-
mined by a colorimetric method using 3-(4,5-dimethylthiazol-2-yl)-
2,5-diphenyltetrazolium bromide (MTT). A stock solution of the test
compound (100 mgmLÀ1) in dimethyl sulfoxide (DMSO) was pre-
pared, and graded quantities were added to a specified volume of
sterilized liquid MH medium. A specified volume of the com-
pound-containing medium was then poured into microtiter plates.
A suspension of the microorganism was prepared to contain
approx. 105 cfumLÀ1 and applied to microtiter plates with serially
(E)-N-((3,4-dihydro-2H-benzo[b][1,4]dioxepin-7-yl)methylene)-4-
methoxyanilin (21): Compound 21 was obtained as a yellow solid
(440 mg, 90%): Rf =0.6 (EtOAc/hexane, 1:4); mp: 108.5–109.18C;
1H NMR (400 MHz, CDCl3): d=8.36 (s, 1H), 7.53 (d, J=2.0 Hz, 1H),
7.46 (dd, J=8.3, 2.0 Hz, 1H), 7.21 (m, 2H), 7.03 (d, J=8.3 Hz, 1H),
6.93 (m, 2H), 4.28 (m, 4H), 3.83 (s, 3H), 2.23 ppm (m, 2H); 13C NMR
(100 MHz, CDCl3): d=158.2, 157.3, 153.7, 151.2, 145.0, 132.1, 124.0,
122.1, 121.7, 121.6, 114.4, 70.4, 55.5, 31.4 ppm; MS (ESI): m/z 284.3
[M+H]+; Anal. calcd for C17H17NO3: C 72.07, H 6.05, N 4.94, found:
C 72.18, H 6.24, N 4.59.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 440

CHEMMEDCHEM
FULL PAPERS
www.chemmedchem.org
diluted compounds in DMSO to be tested and incubated at 378C
for 24 h. After the MIC values were visually determined on each of
the microtiter plates, phosphate buffered saline (PBS; 50 mL, 0.01m,
pH 7.4; 2.9 g Na2HPO4·12H2O, 0.2 g KH2PO4, 8.0 g NaCl, 0.2 g KCl,
1000 mL distilled H2O) containing MTT (2 mgmLÀ1) was added to
each well. Incubation was continued at RT for 4–5 h. The content
of each well was removed, and 100 mL of isopropanol containing
5% HCl (final concentration 1m) was added to extract the dye.
After 12 h of incubation at RT, the optical density (OD) was mea-
sured with a microplate reader at 550 nm.
Acknowledgements
This work was financed by theNational Natural Science Founda-
tion of China (grant no. J1103512).
Keywords: antibacterial agents · dioxygenated rings · FabH ·
inhibitors · b-ketoacyl-ACP synthase III
E. coli FabH inhibitory assay: Native E. coli FabH protein was overex-
pressed in E. coli DH10B cells using the pET30 vector (pET30 vector
was kindly supplied by the State Key Laboratory of Pharmaceutical
Biotechnology, Nanjing University) and purified to homogeneity in
three chromatographic steps (Q-Sepharose, MonoQ, and hydroxya-
patite) at 48C. The selenomethionine-substituted protein was ex-
pressed in E. coli BL21(DE3) cells and purified in a similar way. Har-
vested cells containing FabH were lysed by sonication in 20 mm
Tris (pH 7.6) containing 5 mm imidazole and 0.5m NaCl, and centri-
fuged (20000 rpm, 30 min, 48C). The supernatant was applied to
a Ni-NTA agarose column, washed, and eluted using a 5–500 mm
imidazole gradient over 20 column volumes. Eluted protein was di-
alyzed against 20 mm Tris (pH 7.6) containing 1 mm dithiothreitol
(DTT) and 100 mm NaCl. Purified FabHs were concentrated to
2 mgmLÀ1 and stored at À808C in 20 mm Tris (pH 7.6) containing
100 mm NaCl, 1 mm DTT, and 20% glycerol for enzymatic assay.
[1] R. A. Daines, I. Pendrak, K. Sham, G. S. Van. Aller, A. K. Konstantinidis,
J. T. Lonsdale, C. A. Janson, X. Qiu, M. Brandt, S. S. Khandekar, C. Silver-
man, M. S. Head, J. Med. Chem. 2003, 46, 5–8.
[2] Z. Nie, C. Perretta, J. Lu, Y. Su, S. Margosiak, K. S. Gajiwala, J. Cortez, V.
Nikulin, K. M. Yager, K. Appelt, S. Chu, J. Med. Chem. 2005, 48, 1596–
1609.
[3] K. S. Gajiwala, S. Margosiak, J. Lu, J. Cortez, Y. Su, Z. Nie, K. Appelt, FEBS
Lett. 2009, 583, 2939–2946.
[4] M. M. Alhamadsheh, N. C. Waters, S. Sachdeva, P. Lee, K. A. Reynolds,
Bioorg. Med. Chem. Lett. 2008, 18, 6402–6405.
[5] J. Lee, K. Jeong, S. Shin, J. Lee, Y. Kim, Eur. J. Med. Chem. 2012, 47, 261–
269.
[6] A. Ashek, S. J. Cho, Bioorg. Med. Chem. 2006, 14, 1474–1482.
[7] S. Singh, L. K. Soni, M. K. Gupta, Y. S. Prabhakar, S. G. Kaskhedikar, Eur. J.
Med. Chem. 2008, 43, 1071–1080.
[8] S. Brinster, G. Lamberet, B. Staels, P. Trieu-Cuot, A. Gruss, C. Poyart,
Nature 2009, 458, 83–86.
[9] W. Balemans, N. Lounis, R. Gilissen, J. Guillemont, K. Simmen, K. Andries,
A. Koul, Nature 2010, 463, E3.
In a final 20 mL reaction, 20 mm Na2HPO4 (pH 7.0) containing
0.5 mm DTT, 0.25 mm MgCl2, and 2.5 mm holo-ACP were mixed
with 1 nm FabH, and H2O was added to 15 mL. After 1 min incuba-
tion, a 2 mL mixture of 25 mm acetyl-CoA, 0.5 mm NADH, and
0.5 mm NADPH was added for FabH reaction for 25 min. The reac-
tion was stopped by adding 20 mL of ice-cold 50% trichloroacetic
acid (TCA), incubating for 5 min on ice, and centrifuging to pellet
the protein. The pellet was washed with 10% ice-cold TCA and re-
[10] J. B. Parsons, C. O. Rock, Curr. Opin. Microbiol. 2011, 14, 544–549.
[11] D. Payne, W. Miller, V. Berry, J. Brosky, W. Burgess, E. Chen, W. DeWolf, Jr.,
A. Fosberry, R. Greenwood, M. Head, D. Heerding, C. Janson, D. Jawor-
ski, P. Keller, P. Manley, T. Moore, K. Newlander, S. Pearson, B. Polizzi, X.
Qiu, S. Rittenhouse, C. Slater-Radosti, K. Salyers, M. Seefeld, M. Smyth, D.
Takata, I. Uzinskas, K. Vaidya, N. Wallis, S. Winram, C. Yuan, W. Huffman,
Antimicrob. Agents Chemother. 2002, 46, 3118–3124.
[12] W. H. Miller, M. A. Seefeld, K. A. Newlander, I. N. Uzinskas, W. J. Burgess,
D. A. Heerding, C. C. K. Yuan, M. S. Head, D. J. Payne, S. F. Rittenhouse,
T. D. Moore, S. C. Pearson, V. Berry, W. E. DeWolf, P. M. Keller, B. J. Polizzi,
X. Qiu, C. A. Janson, W. F. Huffman, J. Med. Chem. 2002, 45, 3246–3256.
[13] J. R. Miller et al., Proc. Natl. Acad. Sci. USA 2009, 106, 1737–1742.
[14] C. McInnes, Curr. Opin. Chem. Biol. 2007, 11, 494–502.
[15] J. Kirchmair, S. Ristic, K. Eder, P. Markt, G. Wolber, C. Laggner, T. Langer,
J. Chem. Inf. Model. 2007, 47, 2182–2196.
3
suspended in 0.5m NaOH (5 mL). The incorporation of the H signal
in the final product was read by liquid scintillation. When deter-
mining the IC50 values, inhibitors were added from a concentrated
DMSO stock such that the final concentration of DMSO did not
exceed 2%.
Docking study
[16] P. D. Kirchhoff, R. Brown, S. Kahn, M. Waldman, C. M. Venkatachalam, J.
Comput. Chem. 2001, 22, 993–1003.
[17] S. D. Pickett, J. S. Mason, I. M. McLay, J. Chem. Inf. Comput. Sci. 1996, 36,
1214–1223.
[18] M. Elhallaoui, M. Laguerre, A. Carpy, F. C. Ouazzani, J. Mol. Model. 2002,
8, 65–72.
Molecular docking of compound 10 into the three-dimensional X-
ray structure of E. coli FabH (PDB: 1HNJ)[22] was carried out using
Discovery Studio (v3.1) as implemented through the graphical user
interface DS-CDOCKER protocol.[20]
[19] J. Lee, K. Jeong, J. Lee, D. Kang, Y. Kim, Bioorg. Med. Chem. 2009, 17,
1506–1513.
[20] Discovery Studio 3.1, Accelrys Software Inc., San Diego, CA (USA), 2011.
[21] D. D. Li, F. Fang, J. R. Li, Q. R. Du, J. Sun, H. B. Gong, H. L. Zhu, Bioorg.
Med. Chem. Lett. 2012, 22, 5870–5875.
The 3D structures of the aforementioned compounds were con-
structed using Chem3D ultra 12.0 (Chemical Structure Drawing
Standard; Cambridge Soft corporation, USA) and energy minimized
by using MMFF94 with 5000 iterations and minimum RMS gradient
of 0.10. The crystal structures of E. coli FabH (PDB: 1HNJ)[22] com-
plex were retrieved from the RCSB Protein Data Bank (http://
www.rcsb.org/pdb/). All bound waters and ligands were eliminated
from the protein and the polar hydrogen was added to the pro-
teins.
[22] X. Qiu, C. A. Janson, W. W. Smith, M. Head, J. Lonsdale, A. K. Konstantini-
dis, J. Mol. Biol. 2001, 307, 341–356.
Received: December 25, 2012
Published online on February 7, 2013
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2013, 8, 433 – 441 441