Non-chelating p-phenylidene-linked bis-imidazoline analogs of known influenza virus endonuclease inhibitors: Synthesis and anti-influenza activity

Article abstract of DOI:10.1016/j.ejmech.2018.10.063

A novel chemotype topologically similar to known influenza virus PA endonuclease inhibitors has been designed. It was aimed to reproduce the extended topology of the known metal-chelating ligands with a p-phenylidene-linked bis-imidazoline scaffold. It wa

Full text of DOI:10.1016/j.ejmech.2018.10.063

European Journal of Medicinal Chemistry 161 (2019) 526e532
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Short communication
Non-chelating p-phenylidene-linked bis-imidazoline analogs of
known influenza virus endonuclease inhibitors: Synthesis and
anti-influenza activity
Dmitry Dar'in a, Vladimir Zarubaev b, Anastasia Galochkina b, Maxim Gureev c,
,
*
Mikhail Krasavin a
a Saint Petersburg State University, Saint Petersburg, 199034, Russian Federation
b Pasteur Institute of Epidemiology and Microbiology, 14 Mira Street, Saint Petersburg, 197101, Russian Federation
c I.M. Sechenov First Moscow State Medical University, Moscow, 119991, Russian Federation
a r t i c l e i n f o
a b s t r a c t
Article history:
A novel chemotype topologically similar to known influenza virus PA endonuclease inhibitors has been
designed. It was aimed to reproduce the extended topology of the known metal-chelating ligands with a
p-phenylidene-linked bis-imidazoline scaffold. It was envisioned that aromatic groups introduced to this
scaffolds via metal-catalyzed N-arylation (Buchwald-Hartwig or Chan-Evans-Lam) would contribute to
lipophilic binding to the target and one of the imidazoline nitrogen atoms would ensure non-chelating
coordination to the prosthetic divalent metal ion. The compounds displayed appreciable anti-influenza
activity in vitro and substantial concentration window from the general cytotoxicity range. Docking
analysis of low-energy poses of the most active compound (as well as their comparison to the binding of
an inactive compound) revealed that these compounds reproduced similar binding components to a
known PA endonuclease inhibitor and displayed similar binding pose and desired monodentate metal
coordination, as was initially envisioned. These findings warrant further investigation of the mechanism
of action of the newly discovered series.
Received 25 September 2018
Received in revised form
24 October 2018
Accepted 28 October 2018
Available online 29 October 2018
Keywords:
Influenza virus
N-aryl imidazoline
Bis-imidazoline
Metal-catalyzed N-Arylation
In silico modeling
© 2018 Elsevier Masson SAS. All rights reserved.
1. Introduction
Moreover, it is the high variability within each segment itself that
leads to the virus developing resistance to the existing anti-
Influenza is perhaps the most widespread infectious disease
causing up to 500,000 deaths every year from either the disease
itself or its complications [1]. Prevention of influenza by vaccina-
tion presents a formidable challenge as it can only be effective if the
specific strains of the flu virus that will circulate in the forthcoming
epidemic season are predicted correctly. This is particularly difficult
to achieve as the variability of the virus is very high due to such
factors as aerosol dissemination mechanism, existence of natural
reservoirs of its circulation (birds, pigs, etc.) exacerbated by the
possibility of interspecies transmission [2]. In addition, the
segmented nature of viral genome allows the process of re-
assortment, which results in the emergence of novel gene con-
stellations with new antigenic and pathogenic properties.
influenza drugs in the current clinical use (such as neuraminidase
inhibitors, M2 channel blockers [3]). Thus, developing novel anti-
influenza drugs is of a paramount importance. Thus, it is impor-
tant that new and effective drug candidates, preferably with a novel
mechanism of action are constantly progressed through the
development pipeline.
The influenza virus polymerase acidic (PA) endonuclease is a
bridged dinuclear metalloenzyme that plays a crucial role in initi-
ating viral replication. Since it is an enzyme that is essential for the
viral lifecycle, it can be considered a valid drug target for antiviral
therapy development [4]. To-date, there are no anti-flu therapies
with this mechanism of action with worldwide approval. However,
the recent advancement of two PA endonuclease inhibitors (AL-
794, S-033188, baloxavir marboxil) into the clinical studies speaks
for the validity of this novel target approach [5]. The results of
phase III clinical trials of baloxavir marboxil [6] confirmed it to
possess superior efficacy, likely due to the novel mechanism of
* Corresponding author. Institute of Chemistry, Saint Petersburg State University,
26 Universitetsky Prospect, Peterhof, 198504, Russian Federation.
E-mail address: m.krasavin@spbu.ru (M. Krasavin).
https://doi.org/10.1016/j.ejmech.2018.10.063
0223-5234/© 2018 Elsevier Masson SAS. All rights reserved.

D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
527
action in contrast to other clinically available antivirals [7]. This led
to the approval of baloxavir marboxil (trate name Xofluza™) in
Japan in the early 2018 [8]. The majority of known PA endonuclease
inhibitors incorporate a metal-chelating moiety which ensures the
small molecule's affinity to the protein target containing prosthetic
Mn2þ or Mg2þ ions [9]. Particularly relevant to the present study are
3-hydroxy 2-pyridone 1 [10], 5-hydroxy 4-pyrimidone 2 [11] and 3-
hydroxy-4-pyridone 3 [12] compounds decorated with a p-phe-
nylidene-linked NH-tetrazole moiety. Likewise, the clinical front-
runner baloxavir marboxil (4) [6] also belongs to this chemical
class. Being a prodrug activated in vivo, it possesses a metal-
chelating moiety capped with a hydrolytically prone methyl car-
bonate moiety. At the same time, a number of non-chelating
compounds (e. g., THC-19 (5) [13] and PA-30 (6) [14]) have sur-
faced from phenotypical screening of large compound libraries that
exhibited activity against influenza virus and were shown to inhibit
PA endonuclease. Although the binding mode for these compounds
has not been established [4], the absence of obvious chelating
motifs in their structure suggests a possibility of alternative non-
chelating binding to the prosthetic divalent metals. This, in turn,
inspired us to try and mimic the extended, linear arrangement of
two heterocyclic motifs present in compounds 1e3 with p-phe-
nylidene-linked bis-imidazoline scaffold 7 [15]. We reasoned that,
while one of the imidazoline moieties could replace the NH-tetra-
zolyl motif, the other could also coordinate to the metal ion.
Moreover, considering our recent developments in the area of
metal-catalyzed imidazoline N-arylation [16,17], various aromatic
and heteroaromatic substituents could be installed at the two
imidazoline moieties in 6, thus leading to additional hydrophobic
and/or hydrogen-bonding interactions with the target (Fig. 1).
Herein, we present the results of reducing this idea to practice,
which resulted in identifying compounds endowed with pro-
nounced anti-influenza activity and low cytotoxicity.
Scheme 1. Preparation of bis- (8a-e) and mono-arylated (9a-c) bis-imidazolines via the
Chan-Evans-Lam arylation of 7.
While introduction of 2-pyridyl and 2-pyrazinyl groups via the
Chan-Evans-Lam arylation protocol would not be feasible due to
low stability of the respective boronic acids and their tendency to
de-borylation under the reaction conditions, the complimentary
approach [16] involving Pd-catalyzed Buchwald-Hartwig-type N-
arylation with respective, highly reactive heteroaryl halides, was
considered. Indeed, using a 2.5-fold excess of respective chlor-
oazines, the bis-arylation reaction gave the desired compounds 8f-
h in moderate yield (Scheme 2).
2.2. Biological evaluation
The anti-influenza activity of compounds synthesized was
evaluated in MDCK cells infected with A/Puerto Rico/8/34 (H1N1)
strain of influenza virus. Additionally, the compounds were
2. Results and discussion
2.1. Synthesis
The starting p-phenylidene-linked bis-imidazoline 7 was syn-
thesized in 90% yield from terephthalonitrile and ethylene diamine
via a CS2-catalyzed reaction, as described in the literature [18]. Bis-
arylation of 7 with 4 equiv. of aryl boronic acids under the Chan-
Evans-Lam conditions [17] resulted in the formation of the
desired N,Nʹ-bis-aryl compounds 8a-e isolated chromatographically
in moderate yield. In three cases, respective mono-arylated ver-
sions 9a-c were also isolated (Scheme 1).
Scheme 2. Preparation of bis-heteroaryl bis-imidazolines 8f-h via the Buschwald-
Hartwig arylation.
Fig. 1. Metal-chelating (1e4, critical portion highlighted), known non-chelating (5e6) influenza virus PA endonuclease inhibitors and the p-phenylidene-linked bis-imidazoline
scaffold 7 explored in this work.

528
D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
evaluated for their general cytotoxicity as the ability to cause death
of uninfected MDCK cells. The resulting data expressed as IC50 and
2.3. Docking studies
CC50
(
m
mol/L) are presented in Table 1.
The crystal structure of influenza virus PA endonuclease com-
plex with compound 1 (PDB index 4M4Q) was employed in docking
simulation of the binding of the most potent compounds from the
mono- (9a) and bis-aryl series (8b and 8d) in comparison with
compound 8g which is completely devoid of the desired activity.
First, the key interactions of 1 with the protein were examined to
reveal the importance of the p-fluorophenyl substituent interation
within the hydrophobic pocket lined with Ile38/Ala37/Tyr24/
Met21/Ala20 residues. Additionally, electrostatic interaction of the
tetrazolyl moiety with Lys34/Arg124 diad is crucial as well as
interaction with the divalent metal ions present in the enzyme's
Generally, both the mono- and bis-aryl series demonstrated
tendency to suppress viral titer in lower concentration range than
that causing reduction of mammalian MDCK cell count (see
Experimental section). Preliminarily, in the bis-aryl series, the
antiviral activity appears to be enhanced by introducing sub-
stitutions in the para-position of the phenyl ring (cf. 8a vs. 8b-d)
and lowered by nitrogen heteroaromatics in lieu of phenyl coun-
terparts (cf. 8e-h). Particularly detrimental to the antiviral potency
was the introduction of the second nitrogen atom (as bis-pyrazinyl
analog 8g). Also notable is the lower cytotoxicity of the mono-aryl
compounds which warrants further investigation of this subclass of
p-phenylidene-linked bis-imidazolines. Although these com-
pounds (9a-c) are somewhat weaker compared to clinically used
antiviral drug rimantadine, their cytotoxicity is much lower. Thus,
further optimization may lower the IC50 values while keeping the
compounds’ profile within the ample selectivity window already
observed for the initial leads.
active site. The latter principal component consists of
p-cation
interaction of the p-fluorophenyl group with the Mg2þ ion and the
chelation of one of the three Mn2þ ions by the hydroxypyridone
motif (Fig. 2).
Compound 1 was re-docked into the active site of PA endonu-
clease along with compounds 8b, 8d and 9a (most active) as well as
8g (inactive) studied in this work. To our delight, this led to
comparably favorable energy characteristics for the active com-
pounds (1, 8b, 8d and 9a) and a markedly lower calculated binding
energy for compound 8g. Moreover, this is reflected in the principal
components (lipophilic, van-der-Waals and Coulomb) contributing
to the overall binding (Table 2).
Table 1
Close examination of the orientation of active compounds 8b, 8d
and 9a revealed strikingly similar contribution from all three
principal components: i. lipophilic pocket filling with an N-aryli-
Anti-influenza activity (IC50), cytotoxicity (CC50) and calculated selectivity indices
(SI) of compounds 8a-h and 9a-c synthesized in this work.
Compound
Structure
CC50
,
m
M
IC50
,
m
M
SI
midazoline moiety (with a notable
p-stacking interaction of that
aryl moiety with Tyr24), ii. electrostatic interation of the other
imidazoline moiety with Lys134/Lys137 diad (possibly supported
by an additional p-stacking interaction with Tyr130 for 8b and 8d),
iii. non-chelating coordination of the first (lower) N-aryl imidazo-
line moiety to the active site Mg2þ ion via the unsubstituted ni-
trogen atom (Fig. 3).
In contrast to the above findings, inactive compound 8g docked
into PA active site displayed no stable binding mode and poor
overlay with reference ligand 1. The binding pose with the best
GlideScore and Emodel values was characterized exclusively by a set
8a
8b
8c
8d
8e
>817
144
361
9
77
12
16
5
>11
12
23
of hydrophobic interactions with no p-stacking or metal coordina-
tion contacts (Fig. 4). This situation differs from that observed for
compound 8b whose overall binding energy was only ~0.4 kcal/mol
lower compared to 8g; compound 8b, however, displayed rather
stable docking poses and good overlay with reference ligand 1.
2
>813
68
>12
8f
127
35
4
8g
>809
>809
e
8h
455
103
4
9a
>1031
>971
224
30
65
44
11
>34
>15
5
9b
9c
Fig. 2. Known metal-chelating PA endonuclease inhibitor 1: two-dimensional ligand-
interaction diagram (LID) showing critical interactions with the protein target.
Rimantadine
62
6

D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
529
Table 2
Results of molecular docking of compounds 1, 8b, 8d, 8g and 9a into the crystal structure of influenza virus PA endonuclease complex with 1 (PDP index 4M4Q).
Compound
GlideScore
Emodel
Lipo
VdW
Coulomb
DG (kcal/mol)
1
ꢁ5.517
ꢁ5.678
ꢁ5.078
ꢁ5.350
ꢁ5.917
ꢁ83.725
ꢁ80.884
ꢁ76.581
ꢁ72.198
ꢁ71.219
ꢁ0.654
ꢁ0.686
ꢁ0.555
ꢁ0.687
ꢁ0.008
ꢁ17.369
ꢁ11.644
ꢁ11.047
ꢁ16.073
ꢁ4.829
ꢁ49.449
ꢁ52.125
ꢁ49.495
ꢁ46.920
ꢁ38.469
ꢁ8.010
ꢁ8.060
ꢁ7.691
ꢁ8.079
ꢁ7.330
9a
8b
8d
8g
3. Conclusion
diluted with EtOAc (80 mL) containing Et3N (2 mL), stirred for
10 min and filtered through a pad of celite. The filtrate was washed
with water (20 mL) and brine (15 mL), dried over anhydrous Na2SO4
and concentrated in vacuo. The resulting crude material was sub-
jected to column chromatography on silica gel to afford compounds
8a-e and 9a-c. Compounds 8a, 9a and 8d were purified by elution
with EtOAc/MeOH/Et3N system (from 94:5:1 to 90:9:1), com-
pounds 8b, 9b, 8c, 9c and 8e - with EtOAc/MeOH/Et3N system (from
90:9:1 to 80:16:4).
We described bis- and mono-arylated p-phenylidene-linked
bis-imidazolines as a novel, non-chelating mimic of known influ-
enza virus PA endonuclease inhibitors with similarly extended
linear arrangement of heterocyclic moieties. The antiviral activity of
these compounds was evaluated in vitro to reveal appreciable po-
tency and significantly lower general cytotoxicity. Docking simu-
lation of binding of the most active compounds to the target - and
comparison of their binding to that of the known PA endonuclease
inhibitors - revealed similar contribution from lipophilic and elec-
trostatic components as well as non-chelating binding to the
prosthetic Mg2þ ion of one of the imidazoline nitrogen atoms,
confirming the initial design idea (preliminarily suggesting that
these compounds might exert their activity through such an inhi-
bition mechanism). This finding further extends the utility of N-aryl
imidazoline moiety as an emerging privileged motif in drug design
[19]. Further investigation of the mechanism of action of the newly
discovered series of anti-influenza compounds is required and will
be reported on in due course.
4.1.2.1. 1,4-Bis(1-phenyl-4,5-dihydro-1H-imidazol-2-yl)benzene (8a).
Prepared according to GP1. Yield 61 mg (24%) as colorless solid;
m.p.: 221e223 ꢀC. 1H NMR (400 MHz, CDCl3):
d
7.43 (s, 4H),
7.19e7.12 (m, 4H), 6.99 (t, J ¼ 7.4 Hz, 2H), 6.83e6.68 (m, 4H), 4.06 (s,
8H). 13C NMR (101 MHz, CDCl3):
d 162.2, 142.7, 132.5, 128.8, 128.6,
123.6, 122.6, 54.0, 53.0. HRMS m/z [MþH]þ calcd for C24H23N4
367.1917, found 367.1913.
4.1.2.2. 2-(4-(4,5-Dihydro-1H-imidazol-2-yl)phenyl)-1-phenyl-4,5-
dihydro-1H-imidazole (9a). Prepared according to GP1. Yield:
16 mg (8%) as colorless solid; m.p.: 182e184 ꢀC. 1H NMR (400 MHz,
4. Experimental section
DMSO‑d6):
d
7.83 (d, J ¼ 8.3 Hz, 2H), 7.51 (d, J ¼ 8.3 Hz, 2H), 7.19 (t,
J ¼ 7.9 Hz, 2H), 7.00 (t, J ¼ 7.4 Hz, 1H), 6.81 (d, J ¼ 7.6 Hz, 2H),
4.14e3.86 (m, 4H), 3.71 (s, 4H). 13C NMR (101 MHz, DMSO‑d6):
4.1. Chemistry
d
163.9, 161.2, 143.5, 134.7, 129.5, 129.3, 128.8, 127.9, 123.8, 123.0,
54.3, 53.4, 48.5. HRMS m/z [MþH]þ calcd for C18H19N4 291.1604,
NMR spectroscopic data were recorded with Bruker Avance 400
spectrometer (400.13 MHz for 1H and 100.61 MHz for 13C) in
DMSO‑d6 and in CDCl3 and were referenced to residual solvent
proton signals (dH ¼ 7.26 and 2.50 ppm, respectively) and solvent
carbon signals (dC ¼ 77.0 and 39.5 ppm, respectively). Melting
points were determined with a Stuart SMP50 instrument in open
capillary tubes. Mass spectra were recorded with a Bruker Maxis
HRMS-ESI-qTOF spectrometer (electrospray ionization mode).
Toluene was distilled from P2O5 and stored over molecular sieves
4 Å.
found 291.1613.
4.1.2.3. 1,4-Bis(1-(4-fluorophenyl)-4,5-dihydro-1H-imidazol-2-yl)
benzene (8b). Prepared according to GP1. Yield 68 mg (24%),
colorless solid, mp 213e215 ꢀC. 1H NMR (400 MHz, CDCl3):
d
7.40 (s,
4H), 6.86 (t, J ¼ 8.6 Hz, 4H), 6.76 (dd, J ¼ 9.0, 4.7 Hz, 4H), 4.11e4.03
(m, 4H), 4.03e3.94 (m, 4H). 13C NMR (101 MHz, CDCl3):
162.4,
d
159.4 (d, J ¼ 244.4 Hz), 139.1 (d, J ¼ 2.9 Hz),132.3, 128.6, 124.9 (d,
J ¼ 8.2 Hz), 115.7 (d, J ¼ 22.6 Hz), 54.6, 53.1. HRMS m/z [MþH]þ calcd
for C24H21F2N4 403.1729, found 403.1742.
4.1.1. Synthesis of 1,4-bis(4,5-dihydro-1H-imidazol-2-yl)benzene
(7)
4.1.2.4. 2-(4-(4,5-Dihydro-1H-imidazol-2-yl)phenyl)-1-(4-
fluorophenyl)-4,5-dihydro-1H-imidazole (9b). Prepared according
to GP1. Yield: 23 mg (11%) as colorless solid; m.p.: 196e198 ꢀC. 1H
To a suspension of terephthalonitrile (2.56 g, 0.02 mol) in
ethylene diamine (10 mL) 5 drops of carbon disulfide were added
and the mixture was stirred at room temperature for 10 min. Then
the mixture was stirred at 130 ꢀC for 3 h. Upon cooling water
(50 mL) was added, the precipitate was filtered off and stirred with
DMSO (60 mL) at 110 ꢀC for 5 min. After cooling to ambient tem-
perature crystals were filtered, washed with water and hot aceto-
nitrile, and dried at 80 ꢀC for overnight to afford 3.88 g (90%),
colorless solid, mp 292e294 ꢀC, lit. 312e314 ꢀC [17]. 1H NMR
NMR (400 MHz, CDCl3):
d
7.76 (d, J ¼ 8.2 Hz, 2H), 7.41 (d, J ¼ 8.2 Hz,
2H), 6.75 (t, J ¼ 8.6 Hz, 2H), 6.67 (dd, J ¼ 8.9, 4.8 Hz, 2H), 4.02e3.92
(m, 2H), 3.91e3.83 (m, 2H), 3.73 (s, 4H). 13C NMR (101 MHz, CDCl3):
d
164.3, 162.0, 159.2 (d, J ¼ 244.1 Hz), 139.2 (d, J ¼ 2.8 Hz), 134.2,
128.8, 128.8, 127.6, 124.9 (d, J ¼ 8.2 Hz), 115.6 (d, J ¼ 22.6 Hz), 54.6,
53.4, 48.3. HRMS m/z [MþH]þ calcd for C18H18FN4 309.1510, found
309.1517.
(400 MHz, DMSO‑d6):
d 7.85 (s, 4H), 6.91 (br.s, 2H), 3.62 (s, 8H).
4.1.2.5. 1,4-Bis(1-(4-methoxyphenyl)-4,5-dihydro-1H-imidazol-2-yl)
4.1.2. General procedure for the preparation of compounds 8a-e
and 9a-c (GP1)
benzene (8c). Prepared according to GP1. Yield: 83 mg (28%) as
colorless solid; m.p.: 211e213 ꢀC. 1H NMR (400 MHz, CDCl3):
d 7.38
A mixture of compound 7 (0.7 mmol), ArB(OH)2 (2.8 mmol),
Cu(OAc)2 (2.0 mmol), K2CO3 (3.5 mmol) and DMSO (3 mL) was
stirred in an open flask at ambient temperature for 3e5 days (re-
action progress was controlled by TLC). Reaction mixture was
(s, 4H), 6.75 (d, J ¼ 9.0 Hz, 4H), 6.69 (d, J ¼ 9.0 Hz, 4H), 4.07e4.00 (m,
4H), 3.97e3.91 (m, 4H), 3.75 (s, 6H). 13C NMR (101 MHz, CDCl3):
d
163.1, 156.5, 136.3, 132.3, 128.5, 125.2, 114.2, 55.4, 55.0, 53.1. HRMS
m/z [MþH]þ calcd for C26H27N4O2 427.2129, found 427.2136.

530
D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
Fig. 3. Overlay of docking poses of compounds 8b (A), 8d (B) and 9a (C) with compound 1 in its complex with influenza virus PA endonuclease and two-dimensional LIDs.

D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
531
Fig. 4. Overlay of docked compound 8g with ligand 1 in the active site of influenza virus PA endonuclease and the respective two-dimensional LID.
4.1.2.6. 2-(4-(4,5-Dihydro-1H-imidazol-2-yl)phenyl)-1-(4-
methoxyphenyl)-4,5-dihydro-1H-imidazole (9c). Prepared accord-
ing to GP1. Yield: 24 mg (11%) as colorless solid; m.p.: 191e193 ꢀC.
7.40 (s, 4H), 6.89 (ddd, J ¼ 7.3, 4.9, 0.7 Hz, 2H), 6.57e6.42 (m, 2H),
4.14e4.09 (m, 4H), 3.96e3.91 (m, 4H). 13C NMR (101 MHz,
DMSO‑d6): d 160.0, 154.5, 148.1, 137.7, 133.9, 128.3, 117.7, 114.0, 53.3,
1H NMR (400 MHz, CDCl3):
d
7.70 (d, J ¼ 8.4 Hz, 2H), 7.50 (d,
J ¼ 8.4 Hz, 2H), 6.80 (d, J ¼ 9.0 Hz, 2H), 6.72 (d, J ¼ 9.0 Hz, 2H),
4.11e4.02 (m, 2H), 4.02e3.92 (m, 2H), 3.77 (s, 4H), 3.75 (s, 3H). 13
51.6. HRMS m/z [MþH]þ calcd for C22H21N6 369.1822, found
369.1811.
C
NMR (101 MHz, CDCl3):
d 164.3, 163.0, 156.6, 136.4, 133.4, 131.2,
4.1.3.2. 1,4-Bis(1-(pyrazin-2-yl)-4,5-dihydro-1H-imidazol-2-yl)ben-
zene (8g). Prepared according to GP2. Yield: 70 mg (27%) as color-
less solid; m.p.: 229e231 ꢀC (EtOAc). 1H NMR (400 MHz, DMSO‑d6):
128.8, 126.8, 125.3, 114.3, 55.4, 55.1, 53.3, 50.1. HRMS m/z [MþH]þ
calcd for C19H21N4O 321.1710, found 321.1711.
d
8.15 (dd, J ¼ 2.5, 1.5 Hz, 2H), 8.09 (d, J ¼ 2.6 Hz, 2H), 7.91 (d,
4.1.2.7. 1,4-Bis(1-(4-chlorophenyl)-4,5-dihydro-1H-imidazol-2-yl)
J ¼ 1.5 Hz, 2H), 7.47 (s, 4H), 4.22e4.11 (m, 4H), 4.04e3.99 (m, 4H).
13C NMR (101 MHz, DMSO‑d6):
d 159.2, 150.9, 142.1, 137.3, 136.1,
benzene (8d). Prepared according to GP1. Yield: 97 mg (32%) as
colorless solid; m.p.: 222e224 ꢀC. 1H NMR (400 MHz, CDCl3):
d
7.40
133.9, 128.4, 53.8, 50.9. HRMS m/z [MþH]þ calcd for C20H19N8
(s, 4H), 7.14 (d, J ¼ 8.6 Hz, 4H), 6.71 (d, J ¼ 8.7 Hz, 4H), 4.14e4.01 (m,
371.1727, found 371.1737.
8H). 13C NMR (101 MHz, CDCl3):
d 162.2, 140.5, 131.9, 129.6, 129.0,
124.5, 114.0, 53.8, 52.2. HRMS m/z [MþH]þ calcd for C24H21Cl2N4
4.1.3.3. 1,4-Bis(1-(5-(trifluoromethyl)pyridin-2-yl)-4,5-dihydro-1H-
imidazol-2-yl)benzene (8h). Prepared according to GP2. Yield:
156 mg (31%) as colorless solid; m.p.: 212e214 ꢀC (EtOAc). 1H NMR
435.1138, found 435.1149.
4.1.2.8. 1,4-Bis(1-(pyridin-3-yl)-4,5-dihydro-1H-imidazol-2-yl)ben-
zene (8e). Prepared according to GP1. Yield: 66 mg (25%) as color-
less solid; m.p.: 225e227 ꢀC. 1H NMR (400 MHz, CDCl3):
(400 MHz, DMSO‑d6):
2.6 Hz, 2H), 7.48 (s, 4H), 6.63 (d, J ¼ 8.8 Hz, 2H), 4.22e4.18 (m, 4H),
4.02e3.97 (m, 4H). 13C NMR (101 MHz, DMSO‑d6):
159.0, 156.5,
d
8.50 (d, J ¼ 2.4 Hz, 2H), 7.91 (dd, J ¼ 8.8,
d
d
8.30e8.16 (m, 4H), 7.46 (s, 4H), 7.11e7.02 (m, 2H), 6.94 (d,
145.4, 134.8, 133.7, 128.5, 124.6 (q, J ¼ 271.0 Hz), 118.0 (q,
J ¼ 8.5 Hz, 2H), 4.19e4.02 (m, 8H). 13C NMR (101 MHz, CDCl3):
J ¼ 32.9 Hz), 113.0, 53.5, 51.2. HRMS m/z [MþH]þ calcd for
d
161.3, 144.6, 143.7, 139.3, 132.4, 129.1, 128.8, 123.2, 53.5, 53.5.
C24H19F6N6 505.1570, found 505.1588.
HRMS m/z [MþH]þ calcd for C22H21N6 369.1822, found 369.1826.
4.2. Biology
4.1.3. General procedure for the preparation of compounds 8f-h
(GP2)
4.2.1. Evaluation of the anti-influenza virus activity
A solution of catalyst (prepared by short-time stirring at 110 ꢀC
of mixture of Pd(OAc)2 (23 mg, 0.1 mmol) and BINAP (125 mg,
0.2 mmol) in dry toluene (5 mL)) was poured into a screw-capped
vessel with a suspension of compound 7 (214 mg, 1.0 mmol),
corresponding chloroazine (2.5 mmol), Cs2CO3 (652 mg,
2.0 mmol) in the mixture of dry toluene (5 mL) and dry DMF
(5 mL), filled with argon. The reaction mixture was stirred at
110 ꢀC for 20 h. After cooling to ambient temperature mixture was
diluted with EtOAc (40 mL), filtered through a pad of celite and
evaporated to dryness. The resulting crude material was recrys-
tallized from EtOAc.
100 mL of the compounds dissolved in MEM with 1 mg/mL
trypsin was added into wells with MDCK cells and the plates were
incubated for 1 h at 36 ꢀC at 5% CO2. Cells were infected with
100 mL of influenza virus A/Puerto Rico/8/34 (H1N1) (m.o.i. 0.01)
for 1 h. Cells were washed twice with MEM and the fresh medium
containing the compounds at the same concentrations was added.
The plates were kept for 24 h at 36 ꢀC at 5% CO2. The culture me-
dium was used for the preparation of the series of 10-fold di-
lutions, fresh cells were infected with the dilutions, and the plates
were incubated for 48 h at 36 ꢀC at 5% CO2. After 48 h 100
culture fluid was transferred into the wells of round-bottom
plates. 100 L per well of 1% suspension of chicken erythrocytes
mL of
m
4.1.3.1. 1,4-Bis(1-(pyridin-2-yl)-4,5-dihydro-1H-imidazol-2-yl)ben-
zene (8f). Prepared according to GP2. Yield: 66 mg (18%) as color-
less solid; m.p.: 220e222 ꢀC (EtOAc). 1H NMR (400 MHz, DMSO‑d6):
in saline was added and the results were checked after 40 min
incubation at room temperature. The infectious titer of the virus
was considered as a reciprocal to the maximum virus dilution that
caused complete erythrocytes agglutination. The decrease in the
d
8.14 (ddd, J ¼ 4.9,1.9, 0.7 Hz, 2H), 7.55 (ddd, J ¼ 8.3, 7.3, 2.0 Hz, 2H),

532
D. Dar'in et al. / European Journal of Medicinal Chemistry 161 (2019) 526e532
infectious titer of the virus indicated the antiviral activity of
compounds. The data obtained were used for the calculation of
the 50% effective concentration of the compound, or the sub-
stance concentration that caused the two-fold decrease in the
virus titer (IC50), and then the selectivity index was calculated,
SI]CC50/IC50. Each concentration of the compounds was tested in
triplicate. The IC50 values for each compound were calculated by
GraphPad Prism software using four-parameter logistic curve
model.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.ejmech.2018.10.063.
References
[1] C. Cianci, S.W. Gerritz, C. Deminie, M. Krystal, Influenza nucleoprotein:
promising target for antiviral chemotherapy, Antiviral Chem. Chemother. 23
(2013) 77e91.
[2] M. Tonelli, E. Cichero, Fight against H1N1 influenza a virus: recent insights
towards the development of druggable compounds, Curr. Med. Chem. 23
(2016) 1802e1817.
[3] A.E. Fiore, A. Fry, D. Shay, L. Gubareva, J.S. Bresee, T.M. Uyeki, Antiviral agents
for the treatment and chemoprophylaxis of influenza, MMWR Recomm. Rep.
(Morb. Mortal. Wkly. Rep.) 60 (2011) 1e24.
[4] H. Ju, J. Zhang, B. Huang, D. Kang, B. Huang, X. Liu, P. Zhan, Inhibitors of
influenza virus polymerase acidic (PA) endonuclease: contemporary de-
velopments and perspectives, J. Med. Chem. 60 (2017) 3533e3551.
[5] X. Wu, X. Wu, Q. Sun, C. Zhang, S. Yang, L. Li, Z. Jia, Progress of small molecular
inhibitors in the development of anti-influenza virus agents, Theranostics 7
(2017) 826e845.
4.2.2. Cytotoxicity assay
MDCK cells were seeded into 96-well plates and incubated for
24 h at 36 ꢀC at 5% CO2 until confluent monolayer is formed.
Three-fold dilutions (300 - 4 mM) were prepared on Eagle's
minimal essential medium (MEM) from the compounds under
investigation, added to the cells and incubated for 24 h at 36 ꢀC at
5% CO2. The cell monolayer was washed twice with saline (0.9%
NaCl) and 100
2,5-diphenyltetrazolium bromide], 0.5
m
L of MTT solution [3-(4,5-dimethylthiazole-2)-
g/mL in MEM, were
[6] S. Portsmouth, K. Kawaguchi, M. Arai, K. Tsuchiya, T. Uehara, Cap-dependent
endonuclease inhibitor S-033188 for the treatment of influenza: results from
a phase 3, randomized, double-blind, placebo- and active-controlled study in
otherwise healthy adolescents and adults with seasonal influenza, Open
Forum Infect. Dis. 4 (Suppl. 1) (2017) S734.
m
added into each well. The plates were incubated for 1 h at 36 ꢀC,
then the medium was removed and formazan pellets were dis-
solved in dimethyl sulfoxide (0.1 mL per well). The optical den-
sity in the wells was measured on a spectrophotometer Thermo
Multiskan FC at the wavelength of 540 nm. The results obtained
were used for calculating the concentration of the compound
resulting in death of 50% cells in the culture (CC50) using
GraphPad Prism software employing the four-parameter logistic
curve model.
ꢀ
[7] A. Dias, D. Bouvier, T. Crepin, A.A. McCarthy, D.J. Hart, F. Baudin, S. Cusack,
R.W.H. Ruigrok, The cap-snatching endonuclease of influenza virus polymer-
ase resides in the PA subunit, Nature 458 (2009) 914e918.
[8] Y.A. Heo, Baloxavir: first global approval, Drugs 78 (2018) 693e697.
[9] M. Carcelli, D. Rogolino, A. Gatti, L. De Luca, M. Sechi, G. Kumar, S.W. White,
A. Stevaert, L. Naesens, N-acylhydrazone inhibitors of influenza virus PA
endonuclease with versatile metal binding modes, Sci. Rep. 6 (2016) 31500.
[10] J.D. Baumann, D. Patel, S.F. Baker, R.S.K. Vijayan, A. Xiang, A.K. Parhi,
L. Martinez-Sobrido, E.J. Lavoie, K. Das, E. Arnold, Crystallographic fragment
screening and structure-based optimization yields a new class of influenza
endonuclease inhibitors, ACS Chem. Biol. 8 (2013) 2501e2508.
4.3. In silico modeling
[11] H.Y. Sagong, J.D. Bauman, D. Patel, K. Das, E. Arnold, E.J. LaVoie, Phenyl
substituted 4-hydroxypyridazin-3(2H)-ones and 5-hydroxypyrimidin-4(3H)-
ones: inhibitors of influenza
8086e8098.
[12] C.V. Credille, Y. Chen, S.M. Cohen, Fragment-based identification of influenza
endonuclease inhibitors, J. Med. Chem. 59 (2016) 6444e6454.
[13] K. Yamada, H. Koyama, K. Hagiwara, A. Ueda, Y. Sasaki, S.N. Kanesashi,
R. Ueno, H.K. Nakamura, K. Kuwata, K. Shimizu, M. Suzuki, Y. Aida, Identifi-
cation of a novel compound with antiviral activity against influenza A virus
depending on PA subunit of viral RNA polymerase, Microb. Infect. 14 (2012)
740e747.
a endonuclease, J. Med. Chem. 57 (2014)
The structure of influenza H1N1 virus PA endonuclease (PDP
€
index 4M4Q) was prepared using Schrodinger Protein Preparation
Wizard to eliminate possible errors in the structure of protein-
ligand complex [20]. All ligands were prepared with use of
€
Schrodinger LigPrep module [21]. Three-dimensional structure of
all compounds was generated in OPLS3e force field [22]. The re-
€
ceptor grid was prepared using Schrodinger Glide grid generator.
[14] S. Yuan, H. Chu, K. Singh, H. Zhao, K. Zhang, R.Y. Kao, B. Chow, J. Zhou,
B.J. Zheng, A novel small-molecule inhibitor of influenza A virus acts by
suppressing PA endonuclease activity of the viral polymerase, Sci. Rep. 6
(2016) 22880.
[15] V. Mirkhani, M. Moghadam, S. Tangestaninejad, H. Kargar, Rapid and efficient
synthesis of 2-imidazolines and bis-imidazolines under ultrasonic irradiation,
Tetrahedron Lett. 47 (2006) 2129e2132.
[16] M. Krasavin, Novel diversely substituted 1-heteroaryl-2-imidazolines for
fragment-based drug discovery, Tetrahedron Lett. 53 (2012) 2876e2880.
[17] D. Dar'in, M. Krasavin, The Chan-Evans-Lam N-Arylation Of 2-Imidazolines,
J. Org. Chem. 81 (2016) 12514e12519.
[18] M. Nasr-Esfahani, M. Montazerozohori, S. Mehrizi, Efficient and one-pot cat-
alytic synthesis of 2-imidazolines and bis-imidazolines with p-toluenesulfonic
acid under solvent free conditions, J. Heterocycl. Chem. 48 (2011) 249e254.
[19] M. Krasavin, N-(Hetero)aryl 2-imidazolines: an emerging privileged motif for
contemporary drug design, Chem. Heterocycl. Comp. 53 (2017) 240e255.
[20] G.M. Sastry, M. Adzhigirey, R. Annabhimoju, W. Sherman, Protein and ligand
preparation: parameters, protocols, and influence on virtual screening en-
richments, J. Comput. Aided Mol. Des. 27 (2013) 221e234.
The grid dimensions were similar to those of ligand (12.38 Å, in XYZ
axes) with an 8 Å buffer zone. The total grid side size was 20 Å,
centered on the ligand midpoint. Softening potential (VdW scaling
factor) applied was 0.8 with a partial charge cutoff of 0.25. During
the grid generation, metal coordination constraints were imposed.
The initial docking was conducted in standard precision mode with
standard conditions (VdW scaling factor, RMS deviation). The
number of output poses per ligand was set at 20. For each ligand,
the QM charges were computed using semi-empirical method
(Mulliken charges). The best binding poses were selected based on
RMSD from reference ligand [23]. The best docking solutions
included in ligand-protein complex with endonuclease (4M4Q)
were analyzed with MM/GBSA method [24]. All images of bound
€
ligands were generated using Schrodinger Maestro 11.6 and Ligand
Interaction Diagram Generator.
[21] M.L. Verdonk, V. Berdini, M.J. Hartshorn, W.T. Mooij, C.W. Murray, R.D. Taylor,
P. Watson, Virtual screening using protein-ligand docking: avoiding artificial
enrichment, J. Chem. Inf. Comput. Sci. 44 (2004) 793e806.
[22] E. Harder, W. Damm, J. Maple, C. Wu, M. Reboul, J.Y. Xiang, L. Wang,
D. Lupyan, M.K. Dahlgren, J.L. Knight, J.W. Kaus, D.S. Cerutti, G. Krilov,
W.L. Jorgensen, R. Abel, R.A. Friesner, OPLS3: a force field providing broad
coverage of drug-like small molecules and proteins, J. Chem. Theor. Comput.
12 (2016) 281e296.
[23] A.E. Cho, D. Rinaldo, Extension of QM/MM docking and its applications to
metalloproteins, J. Comput. Chem. 30 (2016) 2609e2616.
[24] A. Suenaga, N. Okimoto, Y. Hirano, K. Fukui, An efficient computational
method for calculating ligand binding affinities, PLoS One 7 (2012), e42846.
Acknowledgement
This research was supported by the Russian Foundation for Basic
Research (project grant 14-50-00069). NMR and mass spectrom-
etry studies were performed at the Research Centre for Magnetic
Resonance and the Centre for Chemical Analysis and Materials
Research of Saint Petersburg State University Research Park.