De novo design approaches targeting an envelope protein pocket to identify small molecules against dengue virus

Article abstract of DOI:10.1016/j.ejmech.2019.111628

Dengue fever is a mosquito-borne viral disease that has become a major public health concern worldwide. This disease presents with a wide range of clinical manifestations, from a mild cold-like illness to the more serious hemorrhagic dengue fever and dengue shock syndrome. Currently, neither an approved drug nor an effective vaccine for the treatment are available to fight the disease. The envelope protein (E) is a major component of the virion surface. This protein plays a key role during the viral entry process, constituting an attractive target for the development of antiviral drugs. The crystal structure of the E protein reveals the existence of a hydrophobic pocket occupied by the detergent n-octyl-β-d-glucoside (β-OG). This pocket lies at the hinge region between domains I and II and is important for the low pH-triggered conformational rearrangement required for the fusion of the virion with the host's cell. Aiming at the design of novel molecules which bind to E and act as virus entry inhibitors, we undertook a de novo design approach by “growing” molecules inside the hydrophobic site (β-OG). From more than 240000 small-molecules generated, the 2,4 pyrimidine scaffold was selected as the best candidate, from which one synthesized compound displayed micromolar activity. Molecular dynamics-based optimization was performed on this hit, and thirty derivatives were designed in silico, synthesized and evaluated on their capacity to inhibit dengue virus entry into the host cell. Four compounds were found to be potent antiviral compounds in the low-micromolar range. The assessment of drug-like physicochemical and in vitro pharmacokinetic properties revealed that compounds 3e and 3h presented acceptable solubility values and were stable in mouse plasma, simulated gastric fluid, simulated intestinal fluid, and phosphate buffered saline solution.

Full text of DOI:10.1016/j.ejmech.2019.111628

Journal Pre-proof
De novo design approaches targeting an envelope protein pocket to identify small
molecules against dengue virus
Emilse S. Leal, Natalia S. Adler, Gabriela A. Fernández, Leopoldo G. Gebhard,
Leandro Battini, Maria G. Aucar, Mariela Videla, María Eugenia Monge, Alejandro
Hernández de los Ríos, John Alejandro Acosta Dávila, María L. Morell, Sandra M.
Cordo, Cybele C. García, Andrea V. Gamarnik, Claudio N. Cavasotto, Mariela Bollini
PII:
S0223-5234(19)30762-7
DOI:
https://doi.org/10.1016/j.ejmech.2019.111628
EJMECH 111628
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 16 May 2019
Revised Date: 3 August 2019
Accepted Date: 14 August 2019
Please cite this article as: E.S. Leal, N.S. Adler, G.A. Fernández, L.G. Gebhard, L. Battini, M.G. Aucar,
M. Videla, Marí.Eugenia. Monge, A. Hernández de los Ríos, J.A. Acosta Dávila, Marí.L. Morell, S.M.
Cordo, C.C. García, A.V. Gamarnik, C.N. Cavasotto, M. Bollini, De novo design approaches targeting an
envelope protein pocket to identify small molecules against dengue virus, European Journal of Medicinal
Chemistry (2019), doi: https://doi.org/10.1016/j.ejmech.2019.111628.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier Masson SAS.

De novo design approaches targeting an envelope protein pocket to identify small
molecules against dengue virus
Emilse S. Leala, Natalia S. Adlera,c, Gabriela A. Fernándeza, Leopoldo G. Gebhardb,
Leandro Battinia, Maria G. Aucarc, Mariela Videlaa, María Eugenia Mongea, Alejandro
Hernández de los Ríosf, John Alejandro Acosta Dávilaf, María L. Morellf, Sandra M.
Cordof, Cybele C. Garcíaf, Andrea V. Gamarnikg, Claudio N. Cavasottoc,d,e, and Mariela
Bollinia,#.
aCentro de Investigaciones en Bionanociencias (CIBION), Consejo Nacional de
Investigaciones Científicas y Técnicas (CONICET), Godoy Cruz, 2390 Ciudad
Autónoma de Buenos Aires, Argentina.
bCONICET-Departamento de Ciencia y Tecnología, Universidad Nacional de Quilmes,
Roque Sáenz Peña 352, B1876 Bernal, Buenos Aires, Argentina.
cComputational Drug Design and Molecular Informatics Laboratory, Translational
Medicine Research Institute (IIMT), CONICET-Universidad Austral, Pilar-Derqui,
Buenos Aires, Argentina.
dFacultad de Ciencias Biomédicas, y Facultad de Ingeniería, Universidad Austral, Pilar-
Derqui, Buenos Aires, Argentina
eAustral Institute for Artificial Intelligence, Universidad Austral, Pilar-Derqui, Buenos
Aires, Argentina.
fUniversidad de Buenos Aires, Facultad de Ciencias Exactas y Naturales, Departamento
de Química Biológica, Laboratorio de Estrategias Antivirales. CONICET, Instituto de
Química Biológica (IQUIBICEN), Buenos Aires, Argentina.
1

gFundación Instituto Leloir-CONICET, Av. Patricias Argentinas 435, Ciudad Autónoma
de Buenos Aires, Argentina Buenos Aires, Argentina.
#Correponding author: mariela.bollini@conicet.cibion.
2

Abstract
Dengue fever is a mosquito-borne viral disease that has become a major public health
concern worldwide. This disease presents with a wide range of clinical manifestations,
from a mild cold-like illness to the more serious hemorrhagic dengue fever and dengue
shock syndrome. Currently, neither an approved drug nor an effective vaccine for the
treatment are available to fight the disease.
The envelope protein (E) is a major component of the virion surface. This protein
plays a key role during the viral entry process, constituting an attractive target for the
development of antiviral drugs. The crystal structure of the E protein reveals the
existence of a hydrophobic pocket occupied by the detergent n-octyl-β-d-glucoside
(β-OG). This pocket lies at the hinge region between domains I and II and is
important for the low pH-triggered conformational rearrangement required for the
fusion of the virion with the host’s cell. Aiming at the design of novel molecules
which bind to E and act as virus entry inhibitors, we undertook a de novo design
approach by “growing” molecules inside the hydrophobic site (β-OG). From more
than 240000 small-molecules generated, the 2,4 pyrimidine scaffold was selected as
the best candidate, from which one synthesized compound displayed micromolar
activity. Molecular dynamics-based optimization was performed on this hit, and thirty
derivatives were designed in silico, synthesized and evaluated on their capacity to
inhibit dengue virus entry into the host cell. Four compounds were found to be potent
antiviral compounds in the low-micromolar range. The assessment of drug-like
physicochemical and in vitro pharmacokinetic properties revealed that compounds 3e
and 3h presented acceptable solubility values and were stable in mouse plasma,
simulated gastric fluid, simulated intestinal fluid, and phosphate buffered saline
solution.
3

Keywords De novo design, molecular dynamics, anti-dengue virus compounds,
pharmacokinetics in vitro properties
Highlights
•
De novo design approach was used to identify new compounds against DENV E
protein
•
•
•
•
Lead Optimization was performed via molecular dynamics
Several compounds displayed excellent potency against DENV
Compounds 3e and 3h were potent compounds of all four sero-types of DENV
3e and 3h showed acceptable in vitro pharmacokinetics profile.
Graphical Abstract
4

1. Introduction
Dengue fever is the most prevalent mosquito-borne viral disease and has become a
major public health concern worldwide in recent years. Dengue virus (DENV) belongs
to the genus Flavivirus (Flaviviridae), which are positive-sense single-stranded RNA
genome viruses, as several other human pathogens such as West Nile, Japanese
encephalitis and Zika viruses [1–3]. To date, four serotypes of DENV have been
described, i.e., DENV1 to DENV4. At present, DENV is endemic in more than 100
countries worldwide [4]. The Americas, South-East Asia and Western Pacific are the
most seriously affected regions. During 2016 serious outbreaks occurred worldwide,
with more than 2.3 million cases being reported in the Americas [5,6].
Controlling DENV epidemic outbreaks remains difficult. Dengue vaccine (Dengvaxia)
developed by Sanofi Pasteur has been listed in more than 10 countries to prevent dengue
fever [7–9]. However, vaccination requires a three-dose regimen that is limited to
individuals aged 9–45 years old, and in seronegative individuals provides modest
protection[10]. Currently, there is still no specific antiviral to fight the disease Although
some drugs have been proposed for different DENV proteins, such as the viral
polymerase and protease, none of them have been approved [10,11].
The entry of viruses into the host cell is an early and specific stage of infection of which
different viral or cellular targets are sensitive to therapeutic intervention. The entry of
DENV is initiated by the binding of the viral particle to receptors on the host cell’s
plasma membrane, followed by internalization via receptor-mediated endocytosis of the
virion into the cytosol. The acidic conditions of the endosome compartment trigger a
conformational rearrangement of the viral envelope proteins (E) which induces the
fusion of the viral and endosomal membranes creating a pore through which the viral
genome is released into the host cell’s cytoplasm [12,13]. The crystal structure of
5

DENV-2 E protein reveals a hydrophobic pocket occupied by the detergent n-octyl-β-D-
glucoside (β-OG). This pocket lies at the hinge region between domains I and II of the E
protein. The β-OG pocket is the region where the major conformational changes occur
during membrane fusion. Several authors have proposed that small molecules
interacting with the E protein can block viral entry. [13–15] The best pharmacological
evidence for the pocket as a suitable target was published by Wispelaere et al. The
authors demonstrated that pyrimidines analogs can bind extracellularly to the E protein,
and prevent the infection by blocking E-mediated membrane fusion during viral entry.
A resistance mutation, M196V, located adjacent to the β-OG pocket, reduces the
affinity and viral infectivity of the evaluated compound [14]. The aim of this work was
to identify new molecules that would bind to the β-OG pocket and inhibit DENV entry.
To this end, a de novo design strategy was employed and the selected compound
showed viral inhibition at micromolar concentrations. Through a lead optimization
process, several optimized derivatives were obtained which displayed high
improvement in the antiviral activity and pharmacokinetics in vitro properties
(solubility, logP, and stability).
2. Results and discussion
2.1. Chemistry
On the basis of the de novo design considerations (section 2.3), we prepared a range of
2,4-disubstituted pyrimidine analogs (Scheme 1). The synthesis of pyrimidine analogs
was based on previously described procedures [16]. The desired compounds were
obtained through a SNAr reaction of anilines and amines with heteroaryl chlorides [17].
Compounds 2a-f and 4a-d were obtained in high yield by selective substitution in
position 4 with commercially available or synthesized anilines and benzylamines in the
presence of DIPEA in n-butanol (2a-f) or ethanol (4a-d). Finally, the desired
6

compounds (3a-v and 5a-e) were prepared through a SNAr from monosubstituted
pyrimidine, (2a-f and 4a-d) and the corresponding alkyl amines. Due to the lack of
reactivity of the 2-position of N-4-pyrimidine derivatives, the reaction was carried out in
sealed tube at reflux with a low yield.
Compound 8, 11 and 12 were obtained as shown in scheme 2. The intermediate
hydrazone (7) was obtained by the condensation of 4-hydrazino-2-chloro pyrimidine (6)
with 2,4-dichlorobenzaldehyde. Compound 10 was prepared by the condensation of 5-
(pyridin-2-yl)thiophene-2-carboxylic acid and 2-chloropyrimidin-4-amine, in the
presence of EDCI.HCl and HOBt in NMP. Compound 12 was prepared via a SNAr
reaction of the compound (2c) with a sodium alkoxide in THF.
7

Scheme 1. General procedure for synthesis of compounds 3a-v and 5a-e. Reagents and conditions (a) Amines, DIPEA, n-butanol, 110°C,
overnight. (b) Amines, DIPEA, n-butanol, 190°C, sealed vial. (c) Benzylamines, DIPEA, ethanol, 90°C, overnight.
8

Scheme 2. Synthesis of compounds 8, 11 and 12. Reagents and conditions (a) NH2NH2,
TEA, methanol, rt, 2 h. (b) 2,4-Dichlorobenzaldehyde, ethanol, reflux, 1 h. (c) 4-(2-
Aminoethyl)morpholine, DIPEA, 190°C, sealed vial. (d) NH4OH, 90°C, 5 h. (e) 5-
(Pyridin-2-yl)thiophene-2-carboxylic acid, EDCI·HCl, HOBt, NMP, rt, overnight, (f) 4-
(2-Hydroxyethyl)morpholine, NaH, THF, 80°C, overnight.
2.2. Antiviral activity
The antiviral activity was evaluated for each newly designed compounds monitoring the
expression of the luciferase from a recombinant fully functional reporter infectious
DENV-2 bearing a luciferase coding sequence. [18,19] After infection of cells in culture
9

with the reporter DENV, the luciferase activity peaks at 8 h as a result of the translation
of the incoming viral genome. Thus, luciferase activity at this time point reflects entry
of the reporter virus. After 24 h post-infection, the genome amplification by the viral
polymerase occurs, leading to an increase in the luciferase activity, which accounts for
the synthesis of viral RNA. The luciferase activity was measured at 8 h post-infection
with the reporter DENV in either untreated control and in cells treated with each of the
thirty compounds synthesized (Tables 1 and 2). Two of the active inhibitors, that block
an early infection step, were further evaluated for their antiviral activity against fully
DENV1-4 infectious viral particles by a virus yield inhibition assay [20]. Cells were
infected at a multiplicity of infection (m.o.i.) of 0.1 plaque formation units per cell
(PFU/cell) and further incubated in the presence of different compound concentrations
during 24 and 48ꢀh. Virus replication was inhibited by both compounds in a dose-
dependent manner attaining a reduction higher than 50% in virus titres in the range of
concentrations 0.3-1.5ꢀµM for all DENV serotypes (Table 3).
In parallel to antiviral activity, cell viability was measured after the treatment with
different concentrations of each compound using cristal violet and the 3-(4,5-
dimethylthiazol-2yl)-2,5-diphenyl tetrazolium bromide (MTT) assay, as described
previously[21].
2.3. De novo design of DENV E protein inhibitors
Taking into account the crystal structure of the binding site occupied by β-OG reported
by Modis et al. [22], we designed new small organic molecules with the capacity to
bind the E protein like the β-OG ligand. The β-OG pocket entrance contains some
hydrophilic amino acid residues that could establish electrostatically favorable
interactions with ligands (for example, Thr48, Glu49, Gln200, Gln271, and Thr280).
Besides, several polar atoms present at the pocket entrance are suitable partners for
10

energetically favorable hydrogen bond interactions. The hydrophobic pocket, which is
occupied by the n-octyl chain of the β-OG ligand in the crystal structure, is large
enough to accommodate the hydrophobic groups present in small molecule inhibitors
[14,23–27].
Based on the chemical features and critical interactions with the β -OG binding site of
the active compounds previously reported [27,28], we generated a virtual chemical
library using the BOMB program [29]. The library was grown starting with NH3 as the
core, which was positioned to form a hydrogen bond with the Thr48 carbonyl group. All
templates were designed to fulfill these key requirements: (i) to deliver a hydrophobic
group into the hydrophobic pocket, as the β-OG octyl chain; (ii) to incorporate an NH to
the hydrogen bond with Thr48, and (iii) to incorporate polar substituents into the
channel entrance. An extensive conformational search was performed for each ligand;
each conformer was optimally positioned and the one with lowest energy was scored as
output. The structure optimization was performed with the OPLS-AA force field[30].
The prime parameters for the analysis were the protein-ligand potential energy (EPL),
the desolvation energy and surface-area burial for the ligand, and the drug-like predicted
properties including solubility and cell permeability [31]. Around 240,000 molecules
were then built to form the Phx‐NH‐Het‐U motif, where NH is a core, U is a
solubilizing group, Het is a heterocyclic ring, and Phx is a substituted (X) phenyl ring
(Ph). The top-25 scoring analogs are dominated by quinazoline and pyrimidine as Het.
The top 5-analogs have chlorine at 4- and 5- phenyl position (PhX) and 2-
morpholinoethanamine, and piperazine as solubilizing group (U). In the end, we
selected compound 3a based on the high EPL value, adequate drug-like predicted
properties and synthetic feasibility compared to the rest of the candidates.
11

Next, we docked 3a within the binding site and performed a 100 ns molecular dynamics
(MD) simulation to determine the stability of the predicted conformations by BOMB
and to explore the putative binding mode of the selected ligands with the E protein.
During the first 40 ns of the simulation, 3a established a hydrogen bond interaction
between the N-H at position 4 of the pyrimidine structure and the OH of the Thr280
(Figure 1A). The interatomic distance and the typical hydrogen bond angle were nearly
2ꢀÅ and 160°, respectively (Figure S1). After establishing this interaction, the ligand
moved slightly from the initial conformation to another conformation that remained
stable for the rest of the simulation. No hydrogen bond interactions were detected
during that period.
Compound 3a was synthesized and cytotoxicity was evaluated in A549 cell line. Also,
antiviral activity was tested by the luciferase-reporter DENV assay mentioned above
[18,19]. This ligand displayed antiviral activity against DENV-2 (EC50 23.6 µM) and no
cytotoxicity was observed at 50 µM. Although the antiviral activity was moderate, it
was considered a good starting point for lead optimization.
2.4. Lead optimization
Initial analogs of 3a. A computer-aided drug optimization was carried out in order to
improve the antiviral activity of 3a. For this purpose, we used a combination of design
using the BOMB software followed by docking/MD simulations. In the first step, the
determination of the optimal substitution pattern (X) of the Ph scaffold was performed
by BOMB. The best EPL scores were obtained by the introduction of voluminous
substituents such as X= Ph group. Analogues were the docked within the binding site,
followed by 100 ns MD simulations. We were thus able to confirm that the close
contacts made by all the ligands and the E protein became stronger and more stable as
the size of the X chain increased. Furthermore, the more voluminous ligands also
12

established stronger and more stable H bond interactions. In particular, the MD
simulations of compound 3c and 3d revealed that the biphenyl ring group fitted very
well in the hydrophobic pocket establishing an energetically favorable hydrophobic
contact with Thr48, Val130, Leu135, Met196, Ile270, and Phe279 side chains. In
particular, a strong hydrophobic interaction was observed between the terminal aromatic
ring of the biphenyl group and Phe193 at a distance of 5 Å. These ligands established
stable hydrogen bond interactions with the E protein. Both of them presented a
moderate hydrogen bond between the N-H at position 4 of the pyrimidine with the
carbonyl O of the Thr48, (interatomic distances of 2.2 Å and a bond angle of 150o)
(Figures S2 and S3). At the end of the simulation performed with 3d, this ligand was
found to move slightly from the initial conformation to another conformation that
remained stable for the rest of the measurement period, and a new hydrogen bond was
formed between the O of the Ala50 and the N-H at position 2 of the pyrimidine. The
interatomic distances and H bond angle were 1.9ꢀÅ and 160°, respectively (Figure S4).
During the second half of the simulation, another H bond was formed between the N3 of
3c and 3d and the backbone amide H of Ala50 (interatomic distance of 2.2 Å and a
bond angle of 160°) (Figures S5 and S6). For the entire simulation period, the
morpholine group remained exposed to the solvent (Figure 1B). The MD simulations
revealed a common pattern of interaction with the β-OG pocket in the E protein, which
was in agreement with the interactions predicted by the de novo design.
13

Figure 1. Predicted interaction of active compounds 3a (A) and 3c (B) within the E
protein binding site (PDB ID 1OKE) obtained by the molecular dynamics simulations.
Based on the outcome of the MD simulations, compounds 3c and 3d were synthesized
and evaluated in the early infection steps dengue antivirals using an assay that allows
discriminating entry/translation and genome amplification. As expected, both
compounds were 3- and 13-fold more active than the predecessor 3a, respectively. The
influence of other substituents, at position 4 of the pyrimidine ring on the biological
activity was also investigated. Compounds 8 and 11, whose aromatic rings were found
to fit well in the hydrophobic pocket were more active than the reference compound 3a.
(Table 1, Figure S7). In addition, a pairwise comparison of the activities of compounds
3c and 12 (EC50 = 8.6 vs. 39.7 µM) revealed the crucial role of the NH at position 2 of
the pyrimidine for the biological activity.
14

Table 1. Anti-DENV activity and cytotoxicity of compounds 3a-d, 8, 11 and 12.
EC50
CC50
Core structure
Compound
R1
SId
4
(µM) a
(µM)b
23.6±1.2 89.6±2.3
3a
21.3±1.6
>50
>2
7
3b
3c
8.6±1.1 59.0±1.8
1.8±1.0 10.2±0.9
5
3d
10.6±1.0 76.5±1.8
3.3±1.8 10.5±1.1
8
4
8
11
12
39.7±1.1
NDc
ND
aEffective concentration 50%: concentration required to achieve a two-fold reduction of
luciferase activity compared to control. Results represent the mean from at least two
independent experiments analyzed by GraphPad Prism software.
bCytotoxic concentration 50 is defined as a concentration required to reduce the viability of
A549 cells by 50%, as determined by MTT method. Results represent the mean from at least
two independent experiments.
cND: Not determined.
dSI: selectivity index (SIꢀ=ꢀCC50/EC50).
15

The next step in the study was to evaluate the influence of the terminal heterocyclic ring
on the antiviral activity. Therefore, we replaced the ethylmorpholine scaffold of 3a by
ethylpiperidine (3e), tetramethylpiperidine (3h), methylpyridine (3m), methylpyridine
N-oxide (3p), and 4-methyl and N-ethanolpiperazine (3q and 3v, respectively). In
general, all the compounds were active within the low micromolar range. These results
demonstrate that the activity is more dependent on the size of the X chain than on the
heterocyclic ring exposed to the solvent. In particular, 3e and 3h exhibited higher
activity against DENV, as compared to the rest of compounds of the series (3c, 3h, 3m,
3p, 3q and 3v). (Table 2).
Analogs of 3h. Several 3h derivatives (EC50=0.8 µM) were synthesized and evaluated
against DENV. In general, all derivatives exhibited activity in a low micromolar range,
including compound 3g, which bears a p-chlorophenyl group at position 4 of the
pyrimidine ring (EC50=1.9 µM). Compounds 3i and 3j, bearing a chlorine atom in p-
and m-position of the biphenyl group, were the most potent compounds of this series,
with EC50 values of 0.1 µM and 0.5 µM, respectively. These results were also
corroborated by measuring intracellular viral RNA level after treatment with the
compounds for 48 h post infection (p.i.) by qRT-PCR.
Taking into account the activity of 3g, the X chain appears to have less influence on the
biological activity, as compared to the rest of the series (Tables 1 and 2). Further MD
simulations revealed the existence of a stable electrostatic interaction between the NH
charged group of the tetramethylpiperidine scaffold and Glu49 (Figure 2). The
contribution of this residue to the calculated binding free energy was notably higher
than that of the other residues, which explains why the influence of the X chain on the
EC50 values is diminished.
16

Table 2. Anti-DENV activity and cytotoxicity of compounds 3e-v and 5a-e.
EC50
CC50
Core
Compound
R1
SId
23
(µM) a
(µM)b
0.8±0.2
18.1±1.0
14.3±0.8
3e
1.1±0.3
13
10
22
3f
3g
3h
1.90±1.0 18.6±1.9
0.8±0.2
0.18±0.1
0.54±0.2
17.3±0.8
9.8±1.1
5.1±1.2
54
10
8
3i
3j
1.51±1.0 12.0±0.7
3k
1.8±1.1
21.1±1.3
12
7
5b
5c
5d
1.98±1.1 12.8±0.2
4.90±1.0 15.2±0.5
3
5e
2
12.5±0.9 18.3±1.2
17

EC50
CC50
Core
Compound
R1
SId
ND
(µM) a
(µM)b
24.0±1.1
NDc
3l
7.4±1.0
40±1.2
5
6
3m
3n
4.1±1.1 25.7±1.2
3.4±1.1 18.5±1.9
5
3o
5a
23.7±1.6
ND
ND
9.1±1.0 30.2±1.5
3
3p
4.6±0.9 25.1±1.2
3.3±0.9 23.5±1.2
5
5
3q
3r
2.9 ±1.1 15.3±0.8
4.9±1.2 27.7±1.0
5
6
3s
3t
18

EC50
CC50
Core
Compound
R1
SId
ND
(µM) a
(µM)b
24.0±1.1
ND
3u
19.9±1.3
ND
ND
3v
aEffective concentration 50%: concentration required to achieve a two-fold reduction of
luciferase activity compared to control. Results represent the mean from at least two
independent experiments analyzed by GraphPad Prism software.
bCytotoxic concentration 50 is defined as a concentration required to reduce the viability of
A549 cells by 50%, as determined by MTT method. Results represent the mean from at least
two independent experiments.
cND: Not determined.
dSI: selectivity index (SIꢀ=ꢀCC50/EC50).
19

Figure 2. Predicted interaction of compound 3h within the E protein β-OG binding site,
obtained by molecular dynamics simulation. The electrostatic interaction and hydrogen bond are
shown as violet and green dashed lines respectively.
Since all DENV serotypes co-circulate simultaneously in several tropical and
subtropical areas, the antiviral activity against infectious virus particles of serotypes 1-4
was further evaluated. Based on the high conservation of the residues involved in the
interaction between protein E and ligands, we expected active compounds to display
antiviral activity against the four serotypes of DENV. Compounds 3e and 3h were
selected for evaluation because they presented high potency and better calculated
physicochemical properties (solubility, clogP). As expected, both compounds were
found to be active against the four DENV serotypes, with slight differences between
serotypes (Table 3).
20

Table 3. Antiviral activity profile of 3e and 3h against DENV serotypes.
Compound
EC50 (µM)a
DENV-1
0.87±0.1
0.58±0.1
DENV-2
DENV-3
0.56±0.2
0.39±0.8
DENV-4
2.5±1.1
0.85±0.2
0.81±0.2
3e
0.87±0.3
3h
aEffective concentration 50% (EC50) was calculated as the concentration required to reduce
virus yield by 50% in the compound‐treated cell cultures compared with untreated ones at 48 h
p.i. All determinations were performed in triplicate. Values represent the meanꢀ±ꢀstandard
deviation from triplicate independent tests.
2.5 Physicochemical and in vitro pharmacokinetic properties of the selected compounds
The physicochemical properties of the most promising compounds (3e and 3h) were
evaluated in vitro in terms of solubility and stability. The solubility was tested at three
different pH values (1.2, 6.8, and 7.4) employing the shake-flask method [32].
Solubility data are summarized in Table 3. All compounds presented acceptable
solubility values, which were within the range normally observed for oral drugs, i.e., 4-
4000 µg/mL (corresponding to S values ranging from 10-2 to 10-5 M for compounds
with a molecular weight of 400) [33]. The clogP values for the listed compounds ranged
from 3 to 5, which are also within the range for oral drugs (0-5) [34]. Stability studies
were conducted in PBS (pH 7.4), simulated intestinal fluid (SIF, pH 6.8), simulated
gastric fluid (SGF, pH 1.2), and mouse plasma and both compounds were proven to be
stable under all evaluated conditions.
21

Table 4. Physicochemical and pharmacokinetics in vitro properties for 3e and 3h.
Cmpd
Solubility (µg/mL)a
bcLogP
t1/2 (min)c
SGF
SIF
PBS
SGF
SIF
PBS
Plasma
(pH 1.2) (pH 6.8) (pH 7.4)
(pH 1.2) (pH 6.8) (pH 7.4)
506±17
76±10
324±29
580±49
32±6
93±6
4.7
5.0
> 120
> 120
> 120
> 120
> 120
> 120
> 120
> 120
3e
3h
aExperimental solubility at pH 1.2, 6.4 and 7.4. Values are expressed as the mean ± standard
deviation of two independent experiments run in triplicate.
bcLog P: Predicted by the Qikprop software. [35]
c SGF, SIF, PBS and mouse plasma stability
3. Conclusions
On the basis of our hit identification and lead optimization strategy, thirty new
molecules were synthesized. Four of the new derivatives displayed significant anti-
DENV activity (EC50 <1 µM) targeting the early steps of viral infection. Two of the
most promising compounds (3e and 3h) proved to be active against the four DENV
serotypes. Both compounds showed adequate solubility and high stability profiles in
different media. Taking all this into account we consider that the pyrimidine analogs are
a good starting point for the development of novel antivirals against DENV infection.
4. Experimental section
4.1.Chemistry
Column chromatography was carried out employing Merck silica gel (Kieselgel 60, 63-
200 µm). Precoated silica gel plates F-254 were used for thin-layer analytical
22

chromatography. NMR spectra were recorded on Bruker Biospin 600 MHz AVIII600,
Bruker advance II 500 MHz and Bruker 300 MHz spectrometers at room temperature.
Chemical shifts (δ) are reported in ppm and coupling constants (J) in Hertz. The mass
spectrometer utilized was a Xevo G2S QTOF (Waters Corporation, Manchester, UK)
with an electrospray ionization (ESI) source. The mass spectrometer was operated in
positive and negative ion modes with probe capillary voltages of 2.5 and 2.3 kV,
respectively. The sampling cone voltage was 30 V. The source and desolvation gas
temperatures were set to 120 and 350 °C, respectively. The nitrogen gas desolvation
flow rate was 600 L h−1 and the cone desolvation flow rate was 10 L h−1. The mass
spectrometer was calibrated across the range of m/z 50−1200 using a 0.5 mM sodium
formate solution prepared in 2- propanol/water (90:10 v/v). Data were drift corrected
during acquisition using a leucine encephalin reference spray (LockSpray) infused at 2
µL min−1. Data were acquired in the range of m/z 50−1200, and the scan time was set
to 1 s. Data acquisition and processing were carried out using MassLynx, version 4.1
(Waters Corp., Milford, MA, USA).
4.1.1. General procedure for synthesis of compounds 2a-f. To a pressure round-bottom
flask containing the mixture of 2,4-dichloropyrimidine (1) (2.0 mmol) and the
corresponding amine (2.0 mmol) in n-butanol (5 mL), DIPEA was added (10.0 mmol)
o
slowly to the solution. The mixture was stirred at 110 C overnight. The products were
obtained as solids, which were filtrated from the reaction mixture and recrystallized
from ethanol.
1
4.1.1.1. 2-chloro-N-(4-chlorophenyl)pyrimidin-4-amine (2a). Yield: 0.224 g, 46%. H
NMR (600 MHz, CDCl3) δ 8.16 (d, J = 5.9 Hz, 1H), 7.40 – 7.36 (m, 2H), 7.31 – 7.27
(m, 2H), 6.93 (br s, 1H), 6.55 (d, J = 5.9 Hz, 1H).
23

1
4.1.1.2. 2-chloro-N-(p-tolyl)pyrimidin-4-amine (2b). Yield: 0.380 g, 87%. H NMR
(300 MHz, CDCl3) δ 8.11 (d, J = 5.9 Hz, 1H), 7.25 – 7.16 (m, 4H), 6.95 (s, 1H), 6.54
(d, J = 5.9 Hz, 1H), 2.29 (s, 3H).
4.1.1.3 N-((1,1'-biphenyl)-4-yl)-2-chloropyrimidin-4-amine (2c). Yield: 0.350 g, 53%.
1H NMR (600 MHz, DMSO-d6) δ 10.14 (s, 1H), 8.19 (d, J = 5.9 Hz, 1H), 7.73 – 7.64
(m, 6H), 7.46 (t, J = 7.7 Hz, 2H), 7.34 (t, J = 7.4 Hz, 1H), 6.81 (d, J = 5.9 Hz, 1H).
4.1.1.4. 2-chloro-N-(4'-chloro-[1,1'-biphenyl]-4-yl)pyrimidin-4-amine (2d). Yield: 0.082
1
g, 40%). H NMR (600 MHz, CDCl3) δ 8.16 (d, J = 5.9 Hz, 1H), 7.62 – 7.57 (m, 2H),
7.53 – 7.49 (m, 2H), 7.44 – 7.39 (m, 4H), 7.01 (s, 1H), 6.65 (d, J = 5.9 Hz, 1H).
4.1.1.5. 2-chloro-N-(3'-chloro-[1,1'-biphenyl]-4-yl)pyrimidin-4-amine (2e). Yield: 0.206
1
g, 86%. H NMR (500 MHz, CDCl3) δ 8.17 (d, J = 5.9 Hz, 1H), 7.60 (d, J = 8.7 Hz,
2H), 7.47 – 7.41 (m, 4H), 7.40 – 7.36 (m, 2H), 7.13 (s, 1H), 6.67 (d, J = 5.9 Hz, 1H).
4.1.1.6. 2-chloro-N-(4'-methyl-[1,1'-biphenyl]-4-yl)pyrimidin-4-amine (2f). Yield: 0.069
1
g, 21%. H NMR (500 MHz, CDCl3) δ 8.15 (d, J = 5.8 Hz, 1H), 7.64 – 7.59 (m, 2H),
7.50 – 7.47 (m, 2H), 7.39 – 7.35 (m, 2H), 7.29 – 7.25 (m, 2H, signal overlapping with
CDCl3), 7.00 (s, 1H), 6.65 (d, J = 5.9 Hz, 1H), 2.41 (s, 3H).
4.1.2. General procedure for synthesis of compounds 3a-v. To a solution of alkyl amine
(1.0 mmol) and DIPEA (1.6 mmol) in n-butanol (1 mL), the corresponding
monosubtituted pyrimidine (0.4 mmol) was added. The mixture was stirred and heated
o
at 190 C in a sealed vial, and the reaction progress was monitored by TLC. The
reaction solution was poured into water and extracted with dichloromethane. The
organic layer was sequentially washed with brine, dried over anhydrous Na2SO4 and
24

concentrated in vacuo. The crude product was purified by column chromatography
(SiO2, dichloromethane/methanol).
4.1.2.1. N4-(4-chlorophenyl)-N2-(2-(morpholin-4-yl)ethyl)pyrimidine-2,4-diamine (3a).
Yield: 0.028 g, 25%. 1H NMR (600 MHz, CDCl3) δ 7.94 (d, J = 5.7 Hz, 1H), 7.36 (d, J
= 8.7 Hz, 2H), 7.32 – 7.28 (m, 2H), 6.56 (br s, 1H), 5.98 (d, J = 5.7 Hz, 1H), 5.36 (br s,
1H), 3.74 – 3.70 (m, 4H), 3.49 (q, J = 5.7 Hz, 2H), 2.59 (t, J = 6.1 Hz, 2H), 2.49 (br s,
4H). 13C NMR (151 MHz, CDCl3) δ 162.0, 161.0, 157.0, 137.4, 129.2, 128.9, 122.7,
95.2, 67.0, 57.3, 53.4, 37.7. HR-MS (ES) calcd for C16H21ClN5O [M+1]+ 334.1435,
found 334.1434.
4.1.2.2. N2-(2-morpholinoethyl)-N4-(p-tolyl)pyrimidine-2,4-diamine (3b). Yield: 0.013
1
g, 21%. H NMR (500 MHz, CDCl3) δ 7.89 (d, J = 5.8 Hz, 1H), 7.22 (d, J = 8.1 Hz,
2H), 7.15 (d, J = 8.1 Hz, 2H), 6.61 (br s, 1H), 5.99 (d, J = 5.8 Hz, 1H), 5.55 (br s, 1H),
3.71 (t, J = 4.6 Hz, 4H), 3.48 (q, J = 5.8 Hz, 2H), 2.57 (t, J = 6.2 Hz, 2H), 2.48 (br s,
4H), 2.33 (s, 3H). 13C NMR (126 MHz, CDCl3) δ 162.1, 161.7, 156.7, 135.9, 134.2,
129.8, 122.5, 94.6, 67.0, 57.4, 53.4, 37.7, 20.9. HR-MS (ES) calcd for C17H24N5O
[M+1]+ 314.1981, found 314.1984.
4.1.2.3. N4-((1,1'-biphenyl)-4-yl)-N2-(2-(morpholin-4-yl)ethyl)pyrimidine-2,4-diamine
(3c). Yield: 0.020 g, 26%. 1H NMR (600 MHz, CDCl3) δ 7.97 (d, J = 5.8 Hz, 1H), 7.60
– 7.56 (m, 4H), 7.47 – 7.42 (m, 4H), 7.36 – 7.31 (m, 1H), 6.58 (s, 1H), 6.08 (d, J = 5.8
Hz, 1H), 5.46 (br s, 1H), 3.72 (t, J = 4.6 Hz, 4H), 3.52 (q, J = 5.8 Hz, 2H), 2.60 (t, J =
13
6.2 Hz, 2H), 2.50 (m, 4H). C NMR (151 MHz, CDCl3) δ 162.4, 161.3, 157.5, 140.6,
138.3, 129.0, 128.0, 127.2, 126.9, 121.9, 67.1, 57.5, 53.6, 37.9. HR-MS (ES) calcd for
C22H26N5O [M+1]+ 376.2137, found 376.2140.
4.1.2.4.
N4-(4'-methyl-[1,1'-biphenyl]-4-yl)-N2-(2-morpholinoethyl)pyrimidine-2,4-
1
diamine (3d). Yield: 0.0145 g, 56 H NMR (600 MHz, CDCl3) δ 7.94 (d, J = 5.8 Hz,
25

1H), 7.56 (d, J = 8.2 Hz, 2H), 7.48 (d, J = 7.8 Hz, 2H), 7.43 (d, J = 8.1 Hz, 2H), 7.25 (d,
J = 7.6 Hz, 2H), 6.71 (s, 1H), 6.07 (d, J = 5.8 Hz, 1H), 5.61 (br s, 1H), 3.72 (t, J = 4.6
Hz, 4H), 3.51 (q, J = 5.8 Hz, 2H), 2.60 (t, J = 6.1 Hz, 2H), 2.49 (t, J = 4.4 Hz, 4H), 2.39
(s, 3H). 13C NMR (151 MHz, CDCl3) δ 162.2, 161.4, 157.0, 137.9, 137.7, 137.1, 137.0,
129.7, 127.7, 126.8, 122.1, 95.2, 67.1, 57.5, 53.6, 37.9, 21.2. HR-MS (ES) calcd for
C23H27N5O [M+H]+ 390.2294, found 390.2302; [M+Na]+: 412.2113, found: 412.2113.
4.1.2.5.
N4-([1,1'-biphenyl]-4-yl)-N2-(2-(piperidin-1-yl)ethyl)pyrimidine-2,4-diamine
(3e). Yield: 0.025 g, 36%. 1H NMR (600 MHz, CDCl3) δ 7.95 (d, J = 5.6 Hz, 1H), 7.60
– 7.55 (m, 4H), 7.48 – 7.41 (m, 4H), 7.33 (t, J = 7.3 Hz, 1H), 6.82 (br s, 1H), 6.09 (d, J
= 5.7 Hz, 1H), 5.64 (br s, 1H), 3.55 (q, J = 5.6 Hz, 2H), 2.63 (t, J = 6.1 Hz, 2H), 2.51
13
(br s, 4H), 1.66 – 1.61 (m, 4H), 1.46 (br s, 2H). C NMR (151 MHz, CDCl3) δ 162.3,
161.3, 157.2, 140.6, 138.4, 136.8, 128.9, 127.9, 127.2, 126.9, 121.9, 95.3, 57.7, 54.5,
38.2, 25.7, 24.3. HR-MS (ES) calcd for C23H28N5 [M+1]+ 374.2345, found 374.2322.
4.1.2.6. N4-(4'-chloro-[1,1'-biphenyl]-4-yl)-N2-(2-(piperidin-1-yl)ethyl)pyrimidine-2,4-
diamine (3f). Yield: 0.014 g, 34%.1H NMR (600 MHz, CDCl3) δ 7.96 (d, J = 5.8 Hz,
1H), 7.53 (d, J = 8.6 Hz, 2H), 7.50 (d, J = 8.5 Hz, 2H), 7.47 (d, J = 8.3 Hz, 2H), 7.40 (d,
J = 8.5 Hz, 2H), 6.68 (s, 1H), 6.06 (d, J = 5.8 Hz, 1H), 5.65 (s, 1H), 3.52 (q, J = 5.8 Hz,
2H), 2.60 (t, J = 6.2 Hz, 2H), 2.47 (s, 4H), 1.61 (p, J = 5.6 Hz, 4H), 1.45 (br s, 2H). 13C
NMR (151 MHz, CDCl3) δ 162.3, 161.2, 157.3, 139.1, 138.7, 135.4, 133.3, 129.1,
128.1, 127.8, 121.8, 57.7, 54.5, 38.3, 25.9, 24.4. HR-MS (ES) calcd for C23H26ClN5
[M+H]+ 408.1955, found: 408,1955.
4.1.2.7.
N4-(4-chlorophenyl)-N2-(2,2,6,6-tetramethylpiperidin-4-yl)pyrimidine-2,4-
diamine (3g). Yield: 0.0304 g, 41%. 1H NMR (500 MHz, CDCl3) δ 7.92 (d, J = 5.6 Hz,
1H), 7.44 (d, J = 8.3 Hz, 2H), 7.27 (m, 1H), 7.25 (m, 1H), 6.51 (s, 1H), 5.93 (d, J = 5.7
Hz, 1H), 4.81 (br s, 1H), 4.35 – 4.25 (m, 1H), 2.03 (dd, J = 12.7, 3.7 Hz, 2H), 1.31 (s,
26

6H), 1.16 (s, 6H), 0.98 (t, J = 12.2 Hz, 2H). 13C NMR (151 MHz, CDCl3) δ 161.8,
161.0, 157.5, 137.9, 129.2, 128.6, 122.2, 95.7, 51.5, 45.8, 43.9, 35.0, 28.8. HR-MS (ES)
calcd for C19H26ClN5 [M+H]+ 360.1955, found 360.1951.
4.1.2.8. N4-([1,1'-biphenyl]-4-yl)-N2-(2,2,6,6-tetramethylpiperidin-4-yl)pyrimidine-2,4-
1
diamine (3h). Yield: 0.026 g, 16%. H NMR (600 MHz, CDCl3) δ 7.94 (d, J = 5.6 Hz,
1H), 7.57 – 7.55 (m, 6H), 7.44 (t, J = 7.9 Hz, 2H), 7.33 (t, J = 7.4 Hz, 1H), 6.70 (br s,
1H), 6.02 (d, J = 5.6 Hz, 1H), 4.94 (br s, 1H), 4.40 – 4.31 (m, 1H), 2.06 (dd, J = 12.6,
3.7 Hz, 2H), 1.34 (s, 6H), 1.18 (s, 6H), 1.05 (d, J = 12.0 Hz, 2H). 13C NMR (151 MHz,
CDCl3) δ 161.8, 161.2, 157.2, 140.7, 138.5, 136.6, 129.0, 127.8, 127.2, 126.9, 121.5,
95.7, 51.7, 45.6, 43.8, 34.9, 28.6. HR-MS (ES) calcd for C25H31N5 [M+H]+ 402.2658,
found 402.2658.
4.1.2.9.
N4-(4'-chloro-[1,1'-biphenyl]-4-yl)-N2-(2,2,6,6-tetramethylpiperidin-4-
1
yl)pyrimidine-2,4-diamine (3i). Yield: 0.0035 g, 6%. H NMR (600 MHz, CDCl3) δ
7.90 (d, J = 5.4 Hz, 1H), 7.55 (d, J = 8.5 Hz, 2H), 7.50 (d, J = 8.3 Hz, 2H), 7.48 (d, J =
8.4 Hz, 2H), 7.41 – 7.39 (m, 2H), 6.75 (br s, 1H), 6.01 (d, J = 5.7 Hz, 1H), 4.39-4.33
(m, 1H), 2.05 (d, J = 3.7 Hz, 1H), 2.03 (s, 2H), 1.38 (s, 6H), 1.23 (s, 6H), 1.16 (t, J =
13
12.2 Hz, 2H). C NMR (151 MHz, CDCl3) δ 161.3, 161.2, 139.0, 138.6, 135.4, 133.3,
129.1, 128.1, 127.7, 121.5, 95.8, 52.5, 45.0, 43.6, 34.2, 29.8, 28.1. HR-MS (ES) calcd
for C25H30ClN5 [M+H]+: 436.2268, found 436.2266; [M+Na]+: 458.2087, found:
458.2091
4.1.2.10.
N4-(3'-chloro-[1,1'-biphenyl]-4-yl)-N2-(2,2,6,6-tetramethylpiperidin-4-
1
yl)pyrimidine-2,4-diamine (3j). Yield: 0.0150 g, 11%. H NMR (600 MHz, CDCl3) δ
7.93 (d, J = 5.4 Hz, 1H), 7.57 – 7.50 (m, 5H), 7.45 – 7.41 (m, 1H), 7.36 (t, J = 7.8 Hz,
1H), 7.31 (dq, J = 8.0, 1.0 Hz, 1H), 6.66 (br s, 1H), 6.02 (d, J = 5.5 Hz, 1H), 5.21 (br s,
1H), 4.38 – 4.33 (m, 1H), 2.08 – 2.03 (m, 2H), 1.39 (s, 6H), 1.25 (s, 6H), 1.18 – 1.14
27

(m, 2H). 13C NMR (151 MHz, CDCl3) δ 161.5, 161.1, 142.5, 139.0, 135.1, 134.9, 130.2,
127.9, 127.2, 127.0, 125.0, 123.1, 121.4, 95.8, 45.0, 43.6, 34.2, 29.8, 28.1. HR-MS (ES)
calcd for C25H31ClN5 [M+H]+ 436.2268, found 436.2279.
4.1.2.11.
N4-(4'-methyl-[1,1'-biphenyl]-4-yl)-N2-(2,2,6,6-tetramethylpiperidin-4-
1
yl)pyrimidine-2,4-diamine (3k). Yield: 0.0062 g, 11%. H NMR (600 MHz, CDCl3) δ
7.91 (d, J = 5.6 Hz, 1H), 7.54-7.52 (m, 4H), 7.46 (d, J = 7.8 Hz, 2H), 7.25 (d, J = 7.8
Hz, 2H), 6.60 (br s, 1H), 6.01 (d, J = 5.7 Hz, 1H), 5.14 (br s, 1H), 4.38-4.32 (m, 1H),
2.40 (s, 3H), 2.06 (d, J = 3.7 Hz, 1H), 2.04 (d, J = 3.5 Hz, 2H), 1.34 (s, 6H), 1.26 (s,
6H), 1.05 (t, J = 12.0 Hz, 2H). 13C NMR. HR-MS (ES) calcd for C26H33N5 [M+H]+
416.2814, found 416.2818.
4.1.2.12. N4-(4-chlorophenyl)-N2-((pyridin-3-yl)methyl)pyrimidine-2,4-diamine (3l).
1
Yield: 0.016 g, 26%. H NMR (600 MHz, CDCl3) δ 8.62 (s, 1H), 8.52 (d, J = 4.0 Hz,
1H), 7.97 (d, J = 5.7 Hz, 1H), 7.68 (d, J = 7.8 Hz, 1H), 7.29 – 7.24 (m, 5H, signal
overlapping with CDCl3), 6.44 (s, 1H), 6.01 (d, J = 5.8 Hz, 1H), 5.35 (br s, 1H), 4.64 (d,
J = 6.0 Hz, 2H). 13C NMR (151 MHz, CDCl3) δ 162.2, 161.1, 157.6, 149.2, 148.8,
137.4, 135.2, 135.1, 129.3, 125.8, 123.6, 122.9, 43.1. HR-MS (ES) calcd for C16H15ClN5
[M+1]+ 312.1016, found 312.1018.
4.1.2.13.
N4-([1,1'-biphenyl]-4-yl)-N2-(pyridin-3-ylmethyl)pyrimidine-2,4-diamine
1
(3m). Yield: 0.024 g, 27%. H NMR (500 MHz, DMSO-d6) δ 9.29 (s, 1H), 8.57 (s, 1H),
8.42 (dd, J = 4.70, 1.5 Hz, 1H), 7.88 (d, J = 5.7 Hz, 1H), 7.74 (d, J = 7.8 Hz, 1H), 7.63 –
7.61 (m, 3H), 7.53 (d, , J = 6.5 Hz , 2H), 7.45 – 7.42 (m, 3H), 7.34 – 7.29 (m, 2H), 6.05
(d, J = 5.6 Hz, 1H), 4.51 (d, J = 6.2 Hz, 2H). 13C NMR (126 MHz, CDCl3) δ 158.3,
154.3, 151.6, 140.9, 140.5, 135.1, 133.4, 128.9, 128.8, 128.7, 127.6, 127.4, 127.2,
126.9, 126.7, 121.4, 104.9, 43.8. HR-MS (ES) calcd for C22H20N5 [M+1]+ 354.1719,
found 354.1722.
28

4.1.2.14.
N4-(4'-chloro-[1,1'-biphenyl]-4-yl)-N2-(pyridin-3-ylmethyl)pyrimidine-2,4-
diamine (3n). Yield: 0.010 g, 21%. 1H NMR (500 MHz, DMSO-d6) δ 9.33 (s, 1H), 8.58
(s, 1H), 8.43 (dd, J = 4.7, 1.6 Hz, 1H), 7.89 (d, J = 5.7 Hz, 1H), 7.77 – 7.73 (m, 1H),
7.68 – 7.65 (m, 2H), 7.58 – 7.44 (m, 5 H), 7.36 – 7.33 (m, 1H), 6.06 (d, J = 5.6 Hz, 1H),
4.52 (d, J = 6.2 Hz, 2H). 13C NMR (151 MHz, DMSO-d6) δ 161.7, 160.4, 160.4, 156.2,
148.6, 147.6, 140.2, 138.7, 138.6, 136.1, 134.6, 131.6, 131.4, 128.7, 127.7, 126.6,
123.3, 123.2, 119.5, 41.9. HR-MS (ES) calcd for C22H18ClN5 [M+H]+ 388.1329, found
388.1325
4.1.2.15.
N4-(3'-chloro-[1,1'-biphenyl]-4-yl)-N2-(pyridin-3-ylmethyl)pyrimidine-2,4-
diamine (3o). Yield: 0.013 g, 21%. 1H NMR (600 MHz, CDCl3) δ 8.64 (br s, 1H), 8.52
(d, J = 4.6 Hz, 1H), 7.99 (d, J = 5.8 Hz, 1H), 7.71 (d, J = 7.8 Hz, 1H), 7.57 – 7.54 (m,
1H), 7.52 (d, J = 8.4 Hz, 2H), 7.45 (d, J = 7.9 Hz, 1H), 7.42 (d, J = 8.5 Hz, 2H), 7.36 (t,
J = 7.9 Hz, 1H), 7.30 (d, J = 7.9 Hz, 1H), 7.28 – 7.24 (m, 1H, signal overlapping with
CDCl3), 6.58 (s, 1H), 6.11 (d, J = 5.7 Hz, 1H), 5.40 (br s, 1H), 4.66 (d, J = 6.1 Hz, 2H).
13C NMR (151 MHz, CDCl3) δ 171.8, 161.1, 157.3, 149.1, 148.6, 142.3, 138.5, 135.3,
135.1, 135.0, 134.7, 130.0, 127.8, 127.1, 126.9, 124.9, 123.4, 122.0, 121.7, 43.0. HR-
MS (ES) calcd for C22H19ClN5 [M+1]+ 388.1329, found 388.1329.
4.1.2.16. 3-(((4-([1,1'-biphenyl]-4-ylamino)pyrimidin-2-yl)amino)methyl)pyridine 1-
oxide (3p). Yield: 0.0071 g, 26%. 1H NMR (600 MHz, CDCl3) δ 8.31 (s, 1H), 8.11 (d, J
= 6.3 Hz, 1H), 7.93 (d, J = 5.8 Hz, 1H), 7.59 – 7.53 (m, 4H), 7.44 (t, J = 7.6 Hz, 2H),
7.39 (d, J = 8.2 Hz, 2H), 7.34 (t, J = 7.4 Hz, 1H), 7.28 (d, J = 8.0 Hz, 1H), 7.22 (t, J =
6.4 Hz, 1H), 6.84 (s, 1H), 6.14 (d, J = 5.8 Hz, 1H), 5.65 (br s, 1H), 4.60 (d, J = 6.2 Hz,
2H). 13C NMR (151 MHz, CDCl3) δ 162.0, 161.9, 161.4, 157.2, 140.5, 140.0, 138.5,
137.9, 137.8, 137.2, 129.0, 127.9, 127.3, 127.0, 125.8, 125.2, 122.2, 42.4. HR-MS (ES)
calcd for C22H20N5O [M+1]+ 370.1668, found 370.1679.
29