Dengue virus replication inhibition by dibenzothiepin derivatives

Article abstract of DOI:10.1007/s00044-018-02286-1

The presented research uses both a target-based drug design strategy focused on dengue virus (DENV) helicase, and the repurposing of a known scaffold, the dibenzo[b,e]thiepine moiety, extensively used in antidepressants drugs. A series of dihydrodibenzo[b,e]thiepin derivatives were synthesized and tested at 10 μg/mL in HEK293 cells infected with DENV2. The replication inhibitory effect was average and depends on the chemical structure. The best antiviral effect was recorded for compounds, (E)-(2-methyl-6,11-dihydrodibenzo[b,e]thiepin-11-ylidene)amino butanoate (TM3) and (E)-(2-methyl-6,11-dihydrodibenzo[b,e]thiepin-11-ylidene)amino 3-fluorobenzoate (TM24); the concentrations resulting in a 90% (1 log) inhibiton of viral titers (IC₉₀) being calculated at 10 μM for TM3 and 0.25 μM for TM24. A molecular docking study has been conducted in order to predict the binding affinity of the tested compounds to DENV2 NS3 helicase and also on dopamine D4 receptor and to establish an in silico–in vitro correlation. The results obtained indicate that the antiviral mechanisms are complex and differ significantly depending on the structure. The majority of compounds appear to inhibit only the viral helicase, some of them both helicase and D4 receptors, and in the case of one compound the mechanism is elusive. We also observed that a 2-methyl substitution and S-oxidation on the dibenzo[b,e]thiepin scaffold significantly improves the inhibition of the viral replication.

Full text of DOI:10.1007/s00044-018-02286-1

MEDICINAL
CHEMISTRY
RESEARCH
Medicinal Chemistry Research
https://doi.org/10.1007/s00044-018-02286-1
ORIGINAL RESEARCH
Dengue virus replication inhibition by dibenzothiepin derivatives
1
Dragos P. Mihai1 George M. Nitulescu
Jessica L. Smith2 Alec J. Hirsch2 Camelia E. Stecoza1
●
●
●
●
Received: 26 June 2018 / Accepted: 26 December 2018
© Springer Science+Business Media, LLC, part of Springer Nature 2019
Abstract
The presented research uses both a target-based drug design strategy focused on dengue virus (DENV) helicase, and the
repurposing of a known scaffold, the dibenzo[b,e]thiepine moiety, extensively used in antidepressants drugs. A series of
dihydrodibenzo[b,e]thiepin derivatives were synthesized and tested at 10 µg/mL in HEK293 cells infected with DENV2. The
replication inhibitory effect was average and depends on the chemical structure. The best antiviral effect was recorded for
compounds, (E)-(2-methyl-6,11-dihydrodibenzo[b,e]thiepin-11-ylidene)amino butanoate (TM3) and (E)-(2-methyl-6,11-
dihydrodibenzo[b,e]thiepin-11-ylidene)amino 3-fluorobenzoate (TM24); the concentrations resulting in a 90% (1 log)
inhibiton of viral titers (IC90) being calculated at 10 µM for TM3 and 0.25 µM for TM24. A molecular docking study has
been conducted in order to predict the binding affinity of the tested compounds to DENV2 NS3 helicase and also on
dopamine D4 receptor and to establish an in silico–in vitro correlation. The results obtained indicate that the antiviral
mechanisms are complex and differ significantly depending on the structure. The majority of compounds appear to inhibit
only the viral helicase, some of them both helicase and D4 receptors, and in the case of one compound the mechanism is
elusive. We also observed that a 2-methyl substitution and S-oxidation on the dibenzo[b,e]thiepin scaffold significantly
improves the inhibition of the viral replication.
●
●
●
●
●
Keywords Dengue fever NS3 helicase inhibitor D4 dopamine receptor antagonists Drug repurposing Dibenzothiepin
Molecular docking
Introduction
hemorrhagic fever, or shock syndrome, occur in approxi-
mately 5% of cases and are characterized by high fever,
hemorrhagic manifestations, and thrombocytopenia
(Rodriguez-Roche and Gould 2013). Dengue is ubiquitous
throughout the tropics, being present in more than 100
countries, with ~390 million infections per year, of which
96 million manifest clinically (Bhatt et al. 2013). There are
currently no licensed effective antiviral drugs to treat den-
gue infection and treatment is only supportive. Recently, a
tetravalent vaccine was developed and approved in several
endemic countries, but its efficacy is limited, especially for
DENV1 and 2 (Low et al. 2017).
Even though most clinical studies are focused on pro-
phylactic vaccines, the development of small-molecule
inhibitors for dengue is increasing based on recent struc-
tural information and validation of several targets, both
structural and nonstructural proteins of DENV (Lim et al.
2013). Most antiviral research is focused on targeting one or
more of the seven nonstructural (NS) proteins, especially
NS3, NS4B, and NS5 (El Sahili and Lescar 2017). NS3
contains a serine protease domain and a RNA helicase
domain, NS5 has methyl-transferase and RNA-dependent
Dengue fever (DF) is a viral mosquito-borne disease caused
by the infection of humans by one of the four dengue virus
serotypes (DENV1, DENV2, DENV3, and DENV4),
members of the Flaviviridae family, genus flavivirus (Bäck
and Lundkvist 2013). DENV infection is associated with a
variety of clinical outcomes. In most instances, these range
from asymptomatic to DF, characterized by high fever,
headache, myalgia, arthralgia, weakness, and cutaneous
rash (Carabali et al. 2015). Most patients recover without
complications around ten days after the onset of illness.
Severe, life-threatening forms of the disease, dengue
* George M. Nitulescu
nitulescu_mihai@yahoo.com
1
Faculty of Pharmacy, “Carol Davila” University of Medicine and
Pharmacy, Traian Vuia 6, Bucharest 020956, Romania
2
Vaccine and Gene Therapy Institute, Oregon Health and Sciences
University, Beaverton, OR, USA

Medicinal Chemistry Research
RNA polymerase activities, while NS4B engages in a
complex network of protein–protein interactions essential
for viral replication and evasion of host immune responses
(de Almeida et al. 2013; Xie et al. 2015).
prochlorperazine, and ketotifen are some of the most
important examples of this drug development strategy
(Simanjuntak et al. 2015; Klug et al. 2016; Low et al. 2017).
The research presented in this article uses both a target-
based drug design strategy focused on DENV helicase and
the repurposing of a known promiscuous scaffold, the
dibenzothiepin moiety. The study is based on the structural
similarities of prochlorperazine, ketotifen, norclozapine,
ST-610, and dibenzo[b,e]thiepin derivatives previously
designed as antidepressants. The antiviral activity of the
isomer dibenzo[b,f]thiepin scaffold was demonstrated by
the identification of 7-fluoro-11-(4-methylpiperazin-1-yl)-
10,11-dihydrodibenzo[b,f]thiepin-2-ol, referred as SKI-
417616 (Smith et al. 2014), a very close derivative of
norclozapine (Fig. 1). Notably, the antiviral activity of SKI-
417616 is primarily mediated through its inhibition of the
D4 dopamine receptor pathway and downstream phos-
phorylation of EGFR-related kinase (ERK) (Smith et al.
2014). We therefore focused our in silico molecular docking
analysis on both the DENV helicase and the D4 dopamine
receptor.
The bipartite NS3 protease-NS3 helicase is crucial for the
replication of dengue and other flaviviruses and is a very
attractive target for antiviral development. Viral proteases
are confirmed antiviral targets, with several inhibitors
widely used clinically to treat human immunodeficiency
virus or hepatitis C virus infections. However, protease
inhibitors are associated with problematic metabolic adverse
events caused by peripheral insulin resistance, including
hyperlipidemia, lipodystrophy, and impaired glucose toler-
ance (Flint et al. 2009; Antonelli and Turriziani 2018).
Considering this problem, the helicase active site of NS3
emerged as a target for developing DENV replication
inhibitors (García et al. 2017). Despite these efforts, only a
few inhibitors of NS3 helicase have been reported, includ-
ing the antihelminthic drugs ivermectin and suramin, the
benzoxazole derivative ST-610, and the primuline deriva-
tive ML283 (Sweeney et al. 2015). The lead ST-610 was
well tolerated in strain 129 mice and AG129 mice after
intraperitoneal administration with no visible adverse
effects. ST-610 treatment reduced virus yield by five-fold
compared to the control group (Byrd et al. 2013). Further,
this drug design strategy is impaired by the structural
similarity of the flavivirus NS3 helicase and human DDX3
proteins (Fang et al. 2016).
Experimental methods
Synthesis of the compounds
A set of 20 dibenzo[b,e]thiepin derivatives were selected
based on their molecular diversity and were prepared in a
multistep synthesis as previously described using as starting
materials 6,11-dihydrodibenzo[b,e]thiepin-11-ones 1, con-
verted to the corresponding 11-hydroximino-6,11-dihy-
drodibenzo[b,e]thiepin (OXH) and 11-hydroximino-2-
methyl-6,11-dihydrodibenzo[b,e]thiepin (OXM) by treating
with hydroxylamine hydrochloride. The dibenzothiepin
Drug repurposing has emerged as an alternative approach
for rapid identification of effective therapeutics to treat
challenging infectious diseases, like DENV (Zheng et al.
2017). Several well-known drugs from various therapeutic
classes have been identified as potential candidates for
treatment of DENV. Chloroquine, prednisolone, balapiravir,
modipafant, lovastatin, mycophenolic acid, norclozapine,
A
B
CI
C
F
D
Fig. 1 The chemical structures
of dengue replication inhibitors
prochlorperazine a, norclozapine
b, SKI-417616 c, and ST-610 d
O
N
H
N
S
N
S
N
O
S
OH
CI
N
N
N
N
N
O
NH2
N
N
N
H
S
S
S
NH2OH.HCl
pyridine, reflux, 24h
R-COCl
benzene; pyridine
reflux; 2h
Y
Y
Y
O
N
N
OH
O
1
TH (Y=H)
TM (Y=CH3)
OXH (Y=H)
OXM (Y=CH3)
R
O
Fig. 2 Synthesis of dibenzothiepin derivatives of the TH and TM series

Medicinal Chemistry Research
derivatives of the TH and TM series were synthesized by
acylation of the oximes with various acid chlorides (Fig. 2).
The dibenzothiepin-5,5-dioxides code as SH and SM
were prepared using two alternative synthesis pathways.
One assumes the oxidation with hydrogen peroxide of the
corresponding TH and TM compounds (Fig. 3). The other
requires the following steps: the aforementioned ketones 1
were converted to the corresponding 5,5-dioxides 2 by
oxidation with hydrogen peroxide and subsequently to the
corresponding oximes 3. The compounds from SH and SM
series were prepared by acylation of the oximes with var-
ious acid chlorides (Fig. 4).
All reagents and solvents were purchased from Sigma-
Aldrich (St. Louis, Missouri, USA). The nuclear magnetic
resonance (NMR) spectra were recorded on a Gemini 300BB
instrument (Varian, USA), using DMSO-d6 as solvent;
infrared spectra were recorded on a JASCO FT/IR-
4200 spectrometer (JASCO, Japan); elemental analysis (using
Perkin Elmer Series II 2400 CHNS/O Analyser, USA) and
TLC were performed to assess the compounds’ purity.
molecular complexity (Cplx), non-hydrogen atoms (nH),
non-carbon/hydrogen atoms (nCH), and molecular flex-
ibility (Flx). DataWarrior application was used for several
structure activity relationships (SAR) analyses. All the
statistical analyses were performed with SPSS software
(version 18.0 for Windows; SPSS Inc., Chicago, IL).
Molecular docking
A molecular docking study has been conducted in order to
predict the binding affinity of the tested compounds to
DENV2 NS3 helicase (PDB code: 2BMF) and dopamine D4
receptor (PDB code: 5WIU). Crystal structures of the proteins
have been retrieved from RCSB Protein Data Bank
(https://www.rcsb.org/pdb/home/home.do), followed by the
removal of solvent molecules and the addition of polar
hydrogen atoms. AutoDock Tools 1.5.6 graphical user inter-
face (Scripps Research Institute, San Diego, CA, USA) was
used to compute Kollman and Gasteiger charges (Morris et al.
2009). The grid box was selected in order to include the
known active sites and grid parameters were calculated using
AutoGrid 4.0. Ligands were docked using AutoDock Vina
1.1.2 (Scripps Research Institute, San Diego, CA, USA)
(Trott and Olson 2010) and protein–ligand interactions have
been visualized with Discovery Studio® Visualizer (Accelrys
Software Inc., San Diego, CA, USA). The molecular docking
experiment was carried out on the synthetized dibenzothiepin
derivatives used in their E conformation and known NS3
helicase inhibitors (ST-610) and dopamine D4 receptor
antagonists (norclozapine, nemonapride, and SKI-417616)
(Durcan et al. 1995; Smith et al. 2014; Wang et al. 2017).
Structure activity relationships
A series of structural descriptors were calculated using
DataWarrior application (http://www.openmolecules.org):
molecular weight (MW), octanol/water partition coefficient
(clogP), hydrogen bond donor (HD) and acceptor (HA)
counts, number of rotatable bonds (RB), total surface area
(TSA), polar surface area (PSA), shape index (ShI),
O
S
S
+ H2O2
O
Inhibition of DENV2 replication
Y
CH3COOH
reflux, 2h
N
R
Y
O
N
R
O
Compounds were dissolved in 100% DMSO at 10 mg/mL.
HEK293 cells were treated with 10 μg/mL final concentra-
tion of individual compounds and infected with DENV2
(NGC) at a multiplicity of 0.5 pfu/cell. At 72 h pi, culture
(Y=H)
(Y=H)
SH
SM
TH
TM
O
O
(Y=CH )
(Y=CH )
3
3
Fig. 3 Synthesis of dibenzothiepin-5,5-dioxides code as SH and SM
(pathway 1)
O
S
O
O
S
S
+ H2O2
NH2OH.HCl
O
CH3COOH
reflux, 2h
pyridine, reflux, 24h
Y
Y
Y
N
O
O
OH
1
2
3
O
S
O
R-COCl
benzene; pyridine
reflux; 2h
Y
N
R
O
SH (Y=H)
SM (Y=CH3)
O
Fig. 4 Synthesis of dibenzothiepin-5,5-dioxides code as SH and SM (pathway 2)

Medicinal Chemistry Research
Table 1 Compound structure
and % inhibition (I%) on
Code
I%
Series
Y
R
Reference
HEK293 cells infected with
DENV2 after treatment with 10
μg/mL of individual compound
TH3
0
S
H
C3H7
Stecoza and Missir (1999)
Stecoza and Missir (1998)
Stecoza et al. (2008)
TH6
17
0
S
H
2-furyl
TH25
TH29
TH43
TM3
TM5
TM8
TM9
TM24
TM26
TM29
SH5
S
H
4F-C6H4
4Cl-C6H4
4C3H7-C6H4
C3H7
33
0
S
H
Ilie et al. (2009)
S
H
Stecoza et al. (2013a)
Stecoza and Drăghici (2005)
Stecoza et al. (2013b)
Stecoza and Drăghici (2005)
Stecoza and Missir (2000)
Stecoza and Missir (2001)
Stecoza et al. (2013b)
Stecoza and Missir (2001)
Stecoza et al. (2009)
92
35
62
20
89
41
44
73
51
69
72
75
55
82
0
S
CH3
CH3
CH3
CH3
CH3
CH3
CH3
H
S
2-thienyl
C6H5
S
S
2CH3-C6H4
3F-C6H4
4CF3-C6H4
4Cl-C6H4
2-thienyl
4Cl-C6H4
4Br-C6H4
2-furyl
S
S
S
SO2
SO2
SO2
SO2
SO2
SO2
SO2
S
SH29
SH35
SM6
H
Ilie et al. (2009)
H
Stecoza and Missir (2001)
Stecoza et al. (2013b)
Stecoza et al. (2013b)
Stecoza and Missir (2001)
Stecoza et al. (2013b)
Stecoza and Missir (1998)
CH3
CH3
CH3
CH3
H
SM7
1-methyl-pyrazol-4-yl
3CH3O-C6H4
4CF3-C6H4
–
SM16
SM26
OXH
supernatants were collected and virus quantified by focus-
forming assay (Smith et al. 2014). % inhibition (I%) is
calculated as focus-forming units/mL for treated samples
divided by control (DMSO)-treated samples. Compounds
that showed inhibitory activity were re-examined using
three-fold serial dilutions of compound to determine inhi-
bitory concentrations of 90% (1 log) of virus production
(IC90)
Results and discussion
Synthesis of the compounds
Fig. 5 Dose response of DENV2 to compounds TM3 and TM24.
HEK293 cells were treated with three-fold serial dilutions of com-
pounds. Culture supernatants were collected at 72 h pi and viral titers
quantified by focus-forming assay. Concentrations resulting in a 90%
(1 log) inhibition of viral titers (IC90) are 10 µM for TM3 and 0.25 µM
for TM24
The structures and purity of the target compounds were
confirmed by common analytical techniques, NMR, infra-
red, and elemental analysis, and were consistent with the
literature data (Table 1). NMR spectral data of C-6 atom
and the ester carbonyl showed the presence of only one of
the stereoisomers in all synthesized compounds, except for
compounds TH43 and TM24, which were obtained in ratio
Z/E 1/4.5 and 1/3.6, respectively.
presented in Table 1. Most compounds presented an
average replication inhibitory effect. Two compounds,
TM3 and TM24, resulted in a ≥1 log reduction in virus
titer. HEK293 cells were treated with three-fold serial
dilutions of TM3 and TM24 and viral titers were quan-
tified by focus-forming assay at 72 h pi. Concentrations
resulting in a 90% (1 log) inhibition of viral titers (IC90)
were calculated at 10 µM for TM3, and 0.25 µM for
TM24 (Fig. 5).
Inhibition of DENV2 replication
HEK293 cells were treated with 10 µg/mL of new
dibenzo[b,e]thiepin derivatives and infected with DENV2
at a multiplicity of 0.5 pfu/cell. One percent was calcu-
lated by comparing with the negative control and are

Medicinal Chemistry Research
Fig. 6 Structure activity (I%) relationships using the most central ring method
Structure activity relationships
weight, the number of hydrogen bonds acceptors, and polar
surface area.
The analysis of I% values using the most central ring sys-
tem for the new tested compounds revealed smaller effects
for the unsubstituted dibenzothiepins and significant better
inhibition for the 2-methyl-dibenzothiepins. The transfor-
mation in sulfones also increases the replication inhibition
(Fig. 6).
The results of the central ring system SAR were con-
firmed using an independent samples t-test analysis of I%
values, using as grouping variable the presence of the 2-
methyl substitution, or of the sulfone group, respectively.
The group of 2-methyl-6,11-dihydro-dibenzo[b,e]thiepin
derivatives produced an average inhibition of 60.64% (n =
11) compared with 27.00% for the unsubstituted group (n
= 9, p < 0.05) indicating the usefulness of this substitution
(Y=CH3). The tested 6,11-dihydro-dibenzo[b,e]thiepin-5,5-
dioxide derivatives have an average I% of 68.14 (n = 7),
significantly higher (p < 0.05) than 33.31 for the 6,11-
dihydro-dibenzo[b,e]thiepin group (n = 13).
For all the compounds a series of significant descriptors
were computed (Table 2) and the correlation with I% were
tested using the Pearson coefficient (P). Statistical correla-
tion was obtained for Cplx (P = 0.645), nCH (P = 0.570),
HA (P = 0.524), MW (P = 0.473), PSA (P = 0.454), ShI
(P = −0.619). This can be translated that a smaller shape
index and a higher complexity favor the inhibition of viral
replication. The dibenzothiepin scaffold could be improved
by adding more heteroatoms to increase the molecular
Protein–ligand interactions and binding affinities
A molecular docking study was performed in order to
predict the molecular mechanism of action for the tested
compounds and to establish an in silico–in vitro correlation.
The hypothesis of the viral inhibition of these compounds
based on the structural similarity with ST-610, NS3 helicase
inhibitor, and SKI-417616, D4 dopamine receptor antago-
nist, was tested using a docking study on both targets. The
relative free binding energies (ΔG) of the ligands, estimated
inhibitory constants (Ki), and residues involved in hydrogen
bond formation are shown in Table 3. The inhibitory con-
stant was calculated using the following formula:
ΔG
e K Â T
Ki ¼
ðMÞ;
where R (gas constant) = 1.9858775 × 10−3 kcal K−1 and
T = 298.15 K
Relative binding energies varied between −6.7 (OXH,
TH3) and −9.5 kcal/mol (SM26) for DENV NS3 helicase
and between −6.9 (TH3) and −10.7 kcal/mol (SM16) for
dopamine D4 receptor. A linear regression analysis of I%
values over the binding energies predicted on NS3 helicase
resulted in a small correlation (R = 0.404) and pointed out
the existence of several outliers. When the compounds
TM3, TM24, SH5, TH25, and TH43 were removed from

Medicinal Chemistry Research
Table 2 Structural descriptors
for the tested dibenzothiepins
CODE
MW
cLogP
HA
HD
TSA
PSA
63.96
nCH
RB
ShI
Flx
Cplx
TH3
311.40
335.38
363.41
379.87
387.50
325.43
365.48
359.49
373.48
377.44
427.45
393.89
383.45
411.86
456.32
381.41
395.44
421.47
459.44
241.31
4.854
4.577
5.490
5.995
6.603
5.198
5.599
5.733
6.077
5.833
6.581
6.339
4.422
5.161
5.280
4.088
2.961
4.829
5.747
3.685
3
4
3
3
3
3
3
3
3
3
3
3
5
5
5
6
7
6
5
2
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
241.27
250.94
267.6
3
5
5
5
4
4
5
4
4
5
7
5
7
7
7
7
8
7
9
3
4
3
3
3
5
4
3
3
3
3
4
3
3
3
3
3
3
4
4
0
0.500
0.458
0.500
0.500
0.536
0.478
0.440
0.462
0.444
0.444
0.467
0.481
0.423
0.464
0.464
0.407
0.429
0.433
0.437
0.471
0.248
0.191
0.212
0.212
0.262
0.245
0.198
0.205
0.199
0.210
0.2640
0.210
0.207
0.218
0.218
0.198
0.228
0.235
0.267
0.186
0.842
0.854
0.853
0.836
0.858
0.858
0.869
0.865
0.874
0.872
0.873
0.853
0.905
0.882
0.882
0.918
0.922
0.924
0.921
0.814
TH6
77.10
63.96
63.96
63.96
63.96
92.20
63.96
63.96
63.96
63.96
63.96
109.42
81.18
81.18
94.32
99.00
90.41
81.18
57.89
TH25
TH29
TH43
TM3
TM5
TM8
TM9
TM24
TM26
TM29
SH5
276.67
301.03
253.53
269.45
273.51
285.77
279.86
302.97
288.93
268.28
287.76
290.97
274.29
284.82
306.86
314.06
180.09
SH29
SH35
SM6
SM7
SM16
SM26
OXH
Table 3 Predicted ligand
Code
NS3 helicase
Dopamine D4 receptor
affinities of the new compounds
on DENV2 NS3 helicase (PDB:
2BMF) and dopamine D4
receptor (PDB: 5WIU)
ΔG (kcal/mol) Ki (µM) H-bonding residues ΔG (kcal/mol) Ki (µM) H-bonding residues
TH3
−6.7
−7.5
−8.3
−8.8
−7.6
−6.8
−7.6
−8.6
−7.6
−8.4
−8.8
−7.9
−7.3
−8.4
−8.4
−8.3
−8.5
−8.8
0.6897 ASP290
−8.1
−8.9
−10
1.1447
0.2964
0.0462
0.0648
0.2114
8.6873
0.5825
2.6634
0.0168
0.0330
0.4155
0.0909
0.0909
0.0462
0.4155
0.1786
0.3510
0.0142
SER94
TH6
0.2114 ASP290
SER94
TH25
TH29
TH43
TM3
TM5
TM8
TM9
TM24
TM26
TM29
SH5
0.1076 GLY320, HIS487
SER94
0.3510
2.6634
–
–
–
−9.8
−9.1
−6.9
−8.5
–
THR434
10.2857
–
2.6634 VAL544
ASP115, THR434
0.4920 ARG387, LYS388 −7.6
–
2.6634
0.6897
–
−10.6
−10.2
−8.7
−9.6
−9.6
−10
LEU187
LEU187
–
–
0.0909 LYS388, ASP603
1.6047
4.4207
–
–
–
ASP115
HIS414, THR434
GLN417
LEU187
LEU187
SH29
SH35
SM6
0.6897 LYS388
0.6897 LYS388
0.8166 LYS388
0.5825 ARG599
−8.7
−9.2
−8.8
−10.7
SM7
SM16
0.2114
–
ASP115, SER196,
HIS414
SM26
−9.5
−6.7
−8.2
−7.1
–
0.1076 LYS388, ASP603
1.6047
−10.5
−7.9
−10.5
−8.9
−9.3
−9.5
0.0199
1.6047
0.0199
0.2964
0.1508
0.1076
ASP115
CYS185
CYS185
ARG186
ASP115
ASP115
OXH
-
ST−610
0.9668 LEU542
6.1971 LYS388
SKI-417616
Norclozapine
Nemonapride
–
–
–
–
–
the set, the correlation of I% values with the docking
energies for NS3 helicase was statistically significant (R =
0.832, n = 15). This result suggest the inhibition of the viral
helicase as the mechanism for the majority of the
compounds, and the existence of other mechanisms for
compounds TM3, SH5, and TM24 (Fig. 7). In the case
of D4 dopamine receptor, the docking results were
not correlated with I% values (R = 0.025). One major

Medicinal Chemistry Research
limitation of the study is the use of I% values for the SAR
analyses.
the case of the third outlier, compound TM3, a very low
affinity is predicted for the D4 receptor; the high viral
inhibition being produced by neither mechanisms and will
be the subject of future research. On the other hand, com-
pounds TH25 and TH43 have a lower inhibitory effect than
predicted by the docking on helicase, and also on D4
receptor. The causes of this mismatch can be multiple, the
most probable being the low solubility or a high cellular
metabolism.
In the case of TM3, SH5, and TM24, I% is higher than
predicted by the helicase inhibition and is probably due to
other mechanism. The docking on D4 receptor indicate that
TM24 and SH5 have a dual mechanism, antagonism on D4
dopamine receptor, and a helicase inhibition. However, in
Looking at docking results on both targets, the com-
pounds SH5, SM6, and SM7 appear to have also a dual
mechanism, I% values being a sum of helicase inhibition
and D4 antagonism. Chemically, they share a common
penta-atomic heterocyclic acyl scaffold.
The docking results are confirmed in the case of the
control compounds, ST-610, as helicase inhibitor, and SKI-
417616, norclozapine, and nemonapride as D4 antagonists.
In both methods, OXH and TH3 have low affinity for the
target proteins, in accord with the low antiviral effect.
The results of the docking on helicase protein indicate
SM26 as the lead for development of dibenzothiepins as
dual helicase/D4 inhibitors. In silico modeling suggests that
SM26 forms three hydrogen bonds with the NS3 helicase
and attaches to the nucleic acid binding site by interacting
with several residues that were previously shown to bind the
viral RNA strand (Xu et al. 2005): two hydrogen bonds with
LYS388, pi-cation interaction with ARG387, alkyl inter-
action with LEU443, pi-alkyl interaction with ARG599 and
van der Waals interaction with ILE365 (Fig. 8). Moreover,
Fig. 7 Linear correlation of I% values for the new dibenzothiepin
derivatives and the calculated binding energy on DENV2 NS3 helicase
(PDB code: 2BMF). Outliers are boxed
Fig. 8 a 3D conformation of helicase-SM26 complex. b 2D diagram of helicase–SM26 complex interactions

Medicinal Chemistry Research
Fig. 9 a 3D conformation of D4-SM26 complex. b 2D diagram of D4-SM26 complex interactions
the sulfone and methyl moieties of the dibenzothiepin
nucleus may play a major role in the antiviral activity of
SM26, in agreement with the SAR results.
lipophile moieties capable of accepting hydrogen bonds
(e.g. –O–CH3, –S–CH3, –CF3).
SM26 interacts with several dopamine D4 receptor
residues that bind nemonapride in the crystal structure of the
nemonapride–D4 receptor complex (Wang et al. 2017) and
showed a five-fold lower inhibitory constant. Thus, SM26
forms pi-cation and pi-sulfur interactions with HIS414,
carbon–hydrogen bond and halogen bond with SER196, pi-
sigma interaction with LEU187 and van der Waals inter-
actions with PHE91 (Fig. 9).
Conclusions
A series of dibenzo[b,e]thiepin derivatives were synthe-
sized and tested at a concentration of 10 µg/mL in
HEK293 infected with DENV2. Two compounds, (E)-(2-
methyl-6,11-dihydrodibenzo[b,e]thiepin-11-ylidene)
amino butanoate (TM3) and (E)-(2-methyl-6,11-dihy-
Judging by the docked poses of the ligand in the active
sites, dibenzothiepin derivative SM26 might be able to
inhibit NS3 helicase and dopamine D4 receptor in an
orthosteric manner.
drodibenzo[b,e]thiepin-11-ylidene)amino
3-
fluorobenzoate (TM24), resulted in a ≥1 log reduction in
virus titer. Structural activity relationship analysis indi-
cated that a 2-methyl substitution on the 6,11-dihydro-
dibenzo[b,e]thiepin scaffold significantly improves the
inhibition of the viral replication.
The results indicated that despite the high degree of
chemical similarity within the analyzed series of new
compounds, the antiviral mechanisms are complex and
differ significantly depending on the structure. The majority
of compounds appear to inhibit only the viral helicase, some
of them both helicase and D4 receptors, and in the case of
TM3 the mechanism is elusive.
On the basis of both molecular docking results and SAR
analysis, it can be stated that the presence of sulfone and
methyl moieties on dibenzothiepin central ring contributes
to a higher affinity for NS3 helicase and higher antiviral
activity. In most cases, the sulfone group acted as a
hydrogen bond acceptor, while the methyl group was
involved in alkyl interactions with residues present in the
helicase active site. The lack of sulfone moiety from both
TM3 and TM24 suggests that the methyl radical is an
essential pharmacophore responsible for a more efficient
inhibition of viral replication. Thus, a strategy for designing
new dibenzothiepins with NS3 helicase inhibitor activity
may consist in replacing the methyl group with other
Considering the extensive use of the dibenzo[b,e]thiepin
scaffold in antidepressants drugs, the proposed strategy of
repurposing them as future therapies in dengue infection is
promising and could be successful.

Medicinal Chemistry Research
Compliance with ethical standards
Morris GM, Huey R, Lindstrom W et al. (2009) AutoDock4 and
AutoDockTools4: automated docking with selective receptor
flexibility. J Comput Chem 30:2785–2791. https://doi.org/10.
1002/jcc.21256
Rodriguez-Roche R, Gould EA (2013) Understanding the dengue
viruses and progress towards their control. Biomed Res Int
2013:690835. https://doi.org/10.1155/2013/690835
Conflict of interest The authors declare that they have no conflict of
interest.
Publisher’s note: Springer Nature remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Simanjuntak Y, Liang JJ, Lee YL, Lin YL (2015) Repurposing of
prochlorperazine for use against dengue virus infection. J Infect
Dis 211:394–404. https://doi.org/10.1093/infdis/jiu377
Smith JL, Stein DA, Shum D et al. (2014) Inhibition of dengue virus
replication by a class of small-molecule compounds that antag-
onize dopamine receptor D4 and downstream mitogen-activated
protein kinase signaling. J Virol 88:5533–5542. https://doi.org/
10.1128/JVI.00365-14
Stecoza CE, Caproiu MT, Draghici C et al. (2008) Synthesis and
characterization of new dibenzo[b, e]thiepine derivatives. I. Rev
Chim 59:1348–1351
Stecoza CE, Caproiu MT, Draghici C et al. (2009) Synthesis, char-
acterization and antimicrobial activity evaluation of some new
derivatives of 6,11-dihydrodibenzo[b,e]thiepin 5,5-dioxide. Rev
Chim 60:137–141
Stecoza CE, Drăghici C (2005) Noi derivaţi ai 6,11-dihidro-dibenzo[b,
e]tiepinei. Nota V Farm 53:49–58
Stecoza CE, Ilie C, Caproiu MT (2013a) New O-acyl-oximino-
dibenzo[b,e]thiepines and their sulfones synthesis and character-
ization. Rev Chim 64:529–533
Stecoza CE, Majekova M, Majek P et al. (2013b) Novel diben-
zothiepins with antibiofilm activity demonstrated by micro-
biological assays and molecular modeling. Curr Org Chem
17:113–124. https://doi.org/10.2174/1385272811317020006
Stecoza CE, Missir A-V (1999) Noi derivaţi ai dibenzo[b,e]tiepinei.
Nota II. Farmacia 47:47–51
References
Antonelli G, Turriziani O (2018) Antiviral therapy: old and current
issues. Int J Antimicrob Agents 40:95–102. https://doi.org/10.
1016/j.ijantimicag.2012.04.005
Bäck ATLundkvist Å, (2013) Dengue viruses—an overview Infect
Ecol Epidemiol 3:. 10.3402/iee.v3i0.19839
Bhatt S, Gething PW, Brady OJ et al. (2013) The global distribution
and burden of dengue. Nature 496:504–507. https://doi.org/10.
1038/nature12060
Byrd CM, Grosenbach DW, Berhanu A et al. (2013) Novel benzox-
azole inhibitor of dengue virus replication that targets the NS3
helicase. Antimicrob Agents Chemother 57:1902–1912. https://
doi.org/10.1128/AAC.02251-12
Carabali M, Hernandez LM, Arauz MJ et al. (2015) Why are people
with dengue dying? A scoping review of determinants for dengue
mortality. BMC Infect Dis 15:301. https://doi.org/10.1186/
s12879-015-1058-x
de Almeida H, Bastos IMD, Ribeiro BM et al. (2013) New binding site
conformations of the dengue virus NS3 protease accessed by
molecular dynamics simulation. PLoS ONE 8:e72402. https://doi.
org/10.1371/journal.pone.0072402
Durcan MJ, Rigdon GC, Norman MH, Morgan PF (1995) Is clozapine
selective for the dopamine D4 receptor? Life Sci 57:PL275–
PL283. https://doi.org/10.1016/0024-3205(95)02151-8
El Sahili A, Lescar J (2017) Dengue virus non-structural protein 5.
Viruses 9:91. https://doi.org/10.3390/v9040091
Fang J, Li H, Kong D et al. (2016) Structure-based discovery of two
antiviral inhibitors targeting the NS3 helicase of Japanese ence-
phalitis virus. Sci Rep 6:34550. https://doi.org/10.1038/
srep34550
Flint OP, Noor MA, Hruz PW et al. (2009) The Role of Protease
Inhibitors in the Pathogenesis of HIV-Associated Lipodystrophy:
Cellular Mechanisms and Clinical Implications. Toxicol Pathol
37:65–77. https://doi.org/10.1177/0192623308327119
García LL, Padilla L, Castaño JC (2017) Inhibitors compounds of the
flavivirus replication process. Virol J 14:95. https://doi.org/10.
1186/s12985-017-0761-1
Ilie C, Stecoza CE, Caproiu MT et al. (2009) Synthesis and char-
acterization of new dibenzo[b,e]thiepine derivatives. II. Rev
Chim 60:588–591
Klug DM, Gelb MH, Pollastri MP (2016) Repurposing strategies for
tropical disease drug discovery. Bioorg Med Chem Lett 26:2569–
2576. https://doi.org/10.1016/j.bmcl.2016.03.103
Lim SP, Wang Q-Y, Noble CG et al. (2013) Ten years of dengue drug
discovery: progress and prospects. Antivir Res 100:500–519.
https://doi.org/10.1016/j.antiviral.2013.09.013
Stecoza CE, Missir A-V (1998) Noi derivaţi ai dibenzo[b,e]tiepinei.
Nota I Farm 46:61–66
Stecoza CE, Missir A-V (2000) Noi derivaţi ai 6,11-dihidro-dibenzo[b,
e]tiepinei. Nota III. Farmacia 48:51–59
Stecoza CE, Missir A-V (2001) Noi derivaţi ai 6,11-dihidro-dibenzo[b,
e]tiepinei. Nota IV. Farmacia 49:18–23
Sweeney NL, Hanson AM, Mukherjee S et al. (2015) Benzothiazole
and pyrrolone flavivirus inhibitors targeting the viral helicase.
ACS Infect Dis 1:140–148. https://doi.org/10.1021/id5000458
Trott O, Olson AJ (2010) AutoDock Vina: improving the speed and
accuracy of docking with a new scoring function, efficient opti-
mization and multithreading. J Comput Chem 31:455–461.
https://doi.org/10.1002/jcc.21334
Wang S, Wacker D, Levit A et al. (2017) D(4) dopamine receptor high
resolution structures enable the discovery of selective agonists.
Science 358:381–386. https://doi.org/10.1126/science.aan5468
Xie X, Zou J, Wang Q-Y, Shi P-Y (2015) Targeting dengue virus
NS4B protein for drug discovery. Antivir Res 118:39–45. https://
doi.org/10.1016/j.antiviral.2015.03.007
Xu T, Sampath A, Chao A et al. (2005) Structure of the dengue virus
helicase/nucleoside triphosphatase catalytic domain at a resolu-
tion of 2.4 Å. J Virol 79:10278–10288. https://doi.org/10.1128/
JVI.79.16.10278-10288.2005
Zheng W, Sun W, Simeonov A (2017) Drug repurposing screens and
synergistic drug-combinations for infectious diseases. Br J
Pharmacol 175:181–191
Low JGH, Ooi EE, Vasudevan SG (2017) Current status of dengue
therapeutics research and development. J Infect Dis 215:S96–
S102. https://doi.org/10.1093/infdis/jiw423