New antiparasitic flexible triaryl diamidines, their prodrugs and aza analogues: Synthesis, in vitro and in vivo biological evaluation, and molecular modelling studies

Article abstract of DOI:10.1016/j.ejmech.2021.113625

Dicationic diamidines have been well established as potent antiparasitic agents with proven activity against tropical diseases like trypanosomiasis and malaria. This work presents the synthesis of new mono and diflexible triaryl amidines (6a-c, 13a,b and

Full text of DOI:10.1016/j.ejmech.2021.113625

European Journal of Medicinal Chemistry 222 (2021) 113625
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Research paper
New antiparasitic flexible triaryl diamidines, their prodrugs and aza
analogues: Synthesis, in vitro and in vivo biological evaluation, and
molecular modelling studies
,
b
,
,
,
e
,
, e
Reem K. Arafa a *, Mohamed A. Ismail c **, Tanja Wenzler d 1, Reto Brun d
,
Ananya Paul f, W. David Wilson f, Amira A. Alakhdar a b, David W. Boykin f
,
a Drug Design and Discovery Lab, Zewail City of Science and Technology, Cairo, 12578, Egypt
b Biomedical Sciences Program, University of Science and Technology, Zewail City of Science and Technology, Cairo, 12578, Egypt
c Department of Chemistry, Faculty of Science, Mansoura University, Mansoura, 35516, Egypt
d Swiss Tropical and Public Health Institute, 4002, Basel, Switzerland
e University of Basel, 4003, Basel, Switzerland
f Department of Chemistry, Georgia State University, Atlanta, GA, 30303, Georgia
a r t i c l e i n f o
a b s t r a c t
Article history:
Dicationic diamidines have been well established as potent antiparasitic agents with proven activity
against tropical diseases like trypanosomiasis and malaria. This work presents the synthesis of new
mono and diflexible triaryl amidines (6a-c, 13a,b and 17), their aza analogues (23 and 27) and respective
methoxyamidine prodrugs (5, 7, 12a,b, 22 and 26). All diamidines were assessed in vitro against Trypa-
nosoma brucei rhodesiense (T. b. r.) and Plasmodium falciparum (P. f.) where they displayed potent to
moderate activities at the nanomolar level with IC50s ¼ 11e378 nM for T. b. r. and 4e323 nM against P. f..
In vivo efficacy testing against T. b. r. STIB900 has shown the monoflexible diamidine 6c as the most
potent derivative in this study eliciting 4/4 cures of infected mice for a treatment period of >60 days
Received 26 March 2021
Received in revised form
2 June 2021
Accepted 2 June 2021
Available online 9 June 2021
Keywords:
Flexible diamidines
Triaryl dications
Antitrypanosomal
Antimalarial
Minor groove binders
Molecular dynamics
upon a 4 ꢀ 5 mg/kg dose i. p. treatment. Moreover, thermal melting analysis measurement
DTm for this
series of diamidines/poly (dA-dT) complexes fell between 0.5 and 19 ꢁC with 6c showing the highest
binding to the DNA minor groove. Finally, a 50 ns molecular dynamics study of an AT-rich DNA
dodecamer with compound 6c revealed a strong binding complex supported by vdW and electrostatic
interactions.
© 2021 Elsevier Masson SAS. All rights reserved.
1. Introduction
humans, viz. Trybanosoma brucei gambiense and Trybanosma brusei
rhodesiense (T. b. r.). HAT, also known as the sleeping sickness, can
Protozoal infection diseases transmitted to humans by insect
vectors such as human African trypanosomiasis (HAT) and malaria
have a limited number of FDA approved drugs. HAT, a sub-Saharan
Africa's neglected disease, is caused by the tsetse fly bite trans-
mitting the protozoan parasites of the Trypanosoma genus to
be lethal if not discovered early and treated inadequately as it can
progress from a systemic stage to a neurological stage which re-
quires complicated and risky treatment procedures. [1,2] Treatment
depends on the disease stage where drugs used in the first (sys-
temic) stage are safer and easier to administer than those for sec-
ond (neurological) stage. Pentamidine (Fig. 1), which is used for
treatment of first stage HAT is usually better tolerated by patients
than melarsoprol, the treatment of second stage HAT gambiense
infection, which is an arsenic derivative having many side effects
like the serious reactive encephalopathy. With 70 million people
being at risk of the disease, it was targeted by the WHO for elimi-
nation as a public health problem by 2020 [1,2].
* Corresponding author. Zewail City of Science and Technology, Ahmed Zewail
Road, October Gardens, Giza, 12578, Egypt.
** Corresponding author. Mansoura University, Faculty of Science, Department of
Chemistry, Mansoura, 35516, Egypt.
E-mail addresses: rkhidr@zewailcity.edu.eg (R.K. Arafa), mismail@mans.edu.eg
(M.A. Ismail).
Malaria is another vector-borne parasitic disease widespread
globally in 91 countries particularly within tropical and subtropical
1
Current address: University of Bern, Institute of Cell Biology, Bern, 3012,
Switzerland.
https://doi.org/10.1016/j.ejmech.2021.113625
0223-5234/© 2021 Elsevier Masson SAS. All rights reserved.

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Fig. 1. Structures of key dicationic antitrypanosomal and antimalarial compounds.
regions. Transmitted by the bite of the female Anopheles mosquito,
the two most prevalent species Plasmodium falciparum (P. f.) and
P. vivax, produce blood-stage parasites that colonize crucial organs
[3]. According to the WHO reports, malaria remains a global burden
with 435,000 deaths and 219 million cases in 2017 featuring no
significant advancement in reducing the total number of malaria
cases in the 2015e2017 as compared to the precedent year [4].
Those aforementioned parasites have mitochondrial kinetoplast
DNA (kDNA) that contains DNA minicircles with an AT content of
>75%. These AT rich DNA minicircles have been proven to
demonstrate strong binding regions and selective primary thera-
peutic target sites for heterocyclic dicationic compounds [5e7].
Pentamidine (Fig. 1), a parental drug approved by the FDA for
treatment of the first stage of HAT and as a second choice for
leishmaniasis, is a dicationic diamidine that delivers its effect by
binding to the minor groove of the parasitic DNA leading to inhi-
bition of DNA-dependent enzymes and transcription [8]. On the
otherhand, for developing an oral analogue of pentamidine, the
conformationally rigid furamidine (DB75) and its orally-active
prodrug, pafuramidine (DB289), were introduced displaying
lower pKa values [9,10]. Pafuramidine reached phase III clinical
trials for the treatment of early-stage HAT and phase II trials for
malaria, however, the trials were halted due to nephro- and
hepato-toxicity manifestations [11e13]. Various rigid structures
have been investigated as potential scaffolds for the bridging
moiety of the molecule between the dication heads, however,
linear scaffolds like CGP40215 (Fig. 1) and linear di- and tri-aryl
dications have exhibited strong binding to the minor groove of
the DNA [14e18]. Besides, introducing mono or bi-flexibilities to
the triaryl scaffold in guanidine dications showed favorable results
[19].
In continuation of our work on aryl cationic compounds and
efforts to explore new antiprotozoal agents [19e22], this study
investigated the effectiveness of mono- and di-flexible triaryl
scaffolds with diamidines as the dicationic termini with respect to
their DNA minor groove binding efficacy and antiprotozoal activity.
Therefore, seventeen novel diamidines, aza-analogues and their
respective prodrugs were synthesized and their DNA binding was
assessed using a thermal melting analysis. Moreover, the com-
pounds were in vitro screened against T. b. r. and P. f. to determine
their antiprotozoal activity, and administrated to T. b. r. STIB 900
infected mice to assess their in vivo efficacy in overcoming phar-
macokinetic barriers and eliciting the intended outcomes. Finally, a
molecular dynamics study was performed for the most active
compound to determine its interaction mode and binding stability
to AT-rich regions of the DNA minor groove.
2. Results and discussion
2.1. Chemistry
Synthesis of the target dicationic compounds is depicted in
Schemes 1e4. First, synthesis of the monoflexible diamidines 6a-c
and their prodrugs 5 and 7 is outlined in Scheme 1. Starting from 1-
bromo-3-(chloromethyl)benzene 2a, 1-bromo-4-(chloromethyl)
benzene 2b, or 2-chloro-5-(chloromethyl)pyridine 2c a Suzuki
coupling with (4-cyanophenyl)boronic acid 1 under tetrakis (tri-
phenylphosphine)palladium (0) metal catalysis furnished the cor-
responding key intermediate dinitriles 3a-c. Subsequent treatment
of the latter by hydroxylamine hydrochloride and potassium tert-
butoxide in DMSO provided the corresponding di-amidoximes 4a-
c. The amidoxime 4a was subjected to O-methylation using dime-
thylsulfate under basic conditions achieved by LiOH$H2O in DMF
furnishing the methoxyamidine prodrug 5. Diamidines 6a-c were
prepared from the corresponding di-amidoxime prodrugs 4a-c
employing acetylation followed by catalytic hydrogenation. The N-
(4-(dimethylamino)butanoyl)oxy)amidine prodrug 7 was synthe-
sized through 1,10-carbonyl-diimidazole catalyzed coupling of the
amidoxime 4a with 4-(dimethylamino)butyric acid hydrochloride
in the presence of a catalytic amount of triethylamine.
Scheme 2 illustrates the synthesis procedure adopted for pre-
paring the diflexible diamidines 13a,b and their prodrugs 12a,b.
Similar to Scheme 1, 4-(bromomethyl)benzonitrile 8 was cross-
coupled with 1,3- or 1,4-phenylenediboronic acids 9a,b under
Suzuki coupling conditions to give the corresponding diflexible
dinitriles 10a,b. The dinitrile compounds 10a,b were transformed
to the di-amidoxime counterparts 11a,b by reaction with hydrox-
ylamine. The corresponding diamidines 13a,b and O-methyl de-
rivatives 12a,b were synthesized using the same synthetic routes as
in Scheme 1.
Scheme 3 displays the synthesis of the diflexible furan di-
amidine 17. Compound 8 was reacted with 2,5-furandiylbis (trib-
utylstannane) 14 using Stille coupling to give the dinitrile com-
pound 15. The dinitrile was then converted to the corresponding
diamidine 17 through the amidoxime intermediate 16 using the
same procedure described with the previous diamidines 6a-c and
13a,b.
Scheme 4 depicts the synthesis of oxygenated mono- and di-
flexible diamidines 23 and 27 and their O-methyl prodrugs 22 and
26, respectively. For synthesizing the monoflexible derivatives, the
nucleophile 4-hydroxybenzonitrile 18 (monoequivalent), underwent
reactionwith2,5-dibromopyridine in the presence ofsodiumhydride
as a basic catalyst in DMSO to achieve selective mono-substitution of
2

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Scheme 1. Synthesis of the monoflexible amidines and aza analogues 6a-c and their prodrugs 5 and 7.
Scheme 2. Synthesis of the diflexible amidines 13a,b and their prodrugs 12a,b.
the 2-bromo atom and exclusively yield the bromo derivative 19.
Suzuki cross-coupling of 19 with 4-cyanophenylboronic acid in the
presence of tetrakis (triphenylphosphine)palladium (0) and the base
sodium carbonate resulted in the formation of the dinitrile 20. The
diamidine 23 was directly synthesized from the dinitrile 20 upon
reaction with lithium bis(trimethylsilyl)amide with subsequent
3

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Scheme 3. Synthesis of the diflexible furan derivative 17.
Scheme 4. Synthesis of mono and diflexible amidines 23 & 27, and their respective prodrugs 22 & 26.
treatment with ethanol/HCl (gas). On the other hand, the di-
amidoxime 21 was prepared from the dinitrile 20 using
hydroxylamine hydrochloride and potassium tert-butoxide in DMSO.
O-methylation of 21 was thenperformed using dimethylsulfate in the
4

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
same way as compound 5 (Scheme 1) to give the prodrug compound
22.
respectively), pyridine 27 is too twisted and has a low D
Tm (1.25 ꢁC).
The original Tm determinations were done at pH 6.2 (MES 10) and
The oxygen atom bridged diflexible derivatives were also pre-
pared starting from 18 (two equivalents) which was reacted with
2,6-dibromopyridine under catalysis of potassioum-tert-butoxide
to furnish the diflexible dinitrile 24. Synthesis of the dimethoxime
prodrug 26 and diamidine 27 was achieved from 24 by the same
chemical reaction conditions described above for the synthesis of
the dimethoxyamidine 12 and the diamidine 23, respectively.
pyridine, in the DNA anionic field, would be expected to be pro-
tonated under these conditions. Thus, the
DTm measurements for
both 6c and 23 at pH 8.5 (TRIS.HCl) were determined to evaluate
charge effects. Respective DTm values for 6c and 23 were found to be
5 ꢁC and 10 ꢁC indicating lower binding at higher pH as expected as
the pyridines are less protonated at higher pH.
To confirm selectivity of these derivatives to poly (dA-dT)n DNA
sequences, compounds from each structural class were also eval-
uated against poly (dG-dC)n DNA. With every compound the in-
2.2. DNA thermal melting analysis study
crease in
D
Tm was only between 0 and 1 ꢁC which is within the
Diamidines have been well reported as DNA minor groove
binders and antiprotozoal agents. They are believed to exert their
action via inhibiting DNA-dependent enzymes or through direct
inhibition of DNA transcription [23e25]. Thermal melting analysis
data as well as in vitro activity against T. b. r. STIB900 and P. f. K1 are
presented in Table 1.
experimental error range (data not shown).
2.3. Circular dichroism (CD): probing the binding mode
Circular dichroism spectroscopy is a valuable technique for
characterizing DNA minor groove interactions [26]. CD titration
experiments were performed for selected compounds (23 and 6c)
on the Poly (dA). Poly (dT) to monitor the induced CD as a function
of compound concentration. Obtained CD spectra presenting the
binding mode of the tested diamidines 23 and 6c are presented in
Fig. 2. Results showed that neither the hairpin DNA nor the free
compound exhibited CD signals in the compound absorption region
above 300 nm (spectral data for the CD of the free compounds are
not presented). Addition of minor groove binders to DNA resulted
in substantial positive induced CD signals on complex formation
with DNA at the wavelengths between 300 and 350 nm. Upon
addition of 23 or 6c to the Poly (dA). Poly (dT) sequence, strong
positive CD signals were observed in the compound absorption
region (300e380 nm) being indicative of a minor groove binding
mode (Fig. 2A and B, respectively).
Thermal melting analysis measurement
complex e Tm of free DNA) is a good and fast representation
method for reflecting the binding affinities of diamidines. Tm
DTm (Tm of DNA-ligand
D
values for this series of diamidines/poly (dA-dT)n complexes lie
between 0.5 and 19 ꢁC and that for the amidoxime prodrugs were
not tested as they are well known to be poor DNA binders. With
regards to the methylene monoflexible diamidine 6a and 6b, both
displayed moderate binding to the DNA minor groove with not
much difference between the 1,3- and 1,4- positional isomerism
(
D
Tm ¼ 8.5 and 9 ꢁC, respectively). For the aza analogue 6c with para
orientation of the substituents on the pyridine central ring, intro-
duction of a nitrogen atom at the central ring imparted a sub-
stantial increase in DNA binding as reflected by the
DTm data that
was found to be 19 ꢁC. On the other hand, introduction of another
methylene group to infer more flexibility to the diamidine de-
rivatives as in analogues 13a,b (respective diflexible counterparts of
diamidines 6a,b) led to a dramatic loss of the DNA binding affinity
2.4. Molecular dynamics
(
D
Tm ¼ 0.5, 1 ꢁC, respectively). This observation might in part be
The DNA-compound 6c complex was subjected to a 50 ns MD
simulation. Structural stability of the simulation trajectory of the
complex was analyzed where the root mean square deviation
(RMSD) of the offset distance of all the DNA atoms relative to their
initial position was taken as a demonstration of the system's sta-
bility along the simulation course. RMSD values fluctuated between
1 Å and 4 Å (Fig. 3, upper left panel), reflecting an acceptable range
for the DNA relative stability. Moreover, the DNA radius of gyration
values remained within 2 Å (Fig. 3, upper right panel) compared to
the initial structure. The results for both RMSD and radius of gy-
ration fluctuations confirms the integrity of the DNA structure
throughout the MD simulations. For the DNA-compound 6c com-
plex, the RMSDs of the center of masses of the heavy atoms of the
diamidine 6c relative to the helix atoms were also calculated. RMSD
values of the ligand remained below 1 nm along the majority of the
simulation time frame. They only surpassed that value in 4 minor
occasions indicating the displacement of compound 6c from the
DNA. However, compound 6c returned to the DNA minor groove
each time and remained tightly bound from the 40th ns to the end
of the simulation, and complex converged to a stable state (Fig. 3,
lower left panel). Moreover, the free binding energy calculations of
the complex over the last ten ns of the simulation showed favorable
values for van der Waals and electrostatic interactions between
compound 6c and the DNA (Table 2). van der Waals forces elicited
the chief contribution to the binding free energy between com-
pound 6c and the DNA compared to the electrostatic forces.
attributed to an unfavourable complementarity with the curvature
of the DNA minor groove. Additionally, there was not much favour
in replacing the central phenyl ring by a smaller sized furan ring as
seen for the diflexible amidines 17 (
oflexible aza derivative 23 displayed a great enhancement in DNA
minor groove binding affinity (
compliance with that observed for the aza counterpart 6c with
flexibility imparted by a monomethylene bridge. This affinity was
D
Tm ¼ 2.25 ꢁC). Further, mon-
D
Tm ¼ 17.75 ꢁC), a behaviour in
lost for the meta-diflexible aza analogue 27 where the DTm dropped
to 1.25 ꢁC, a behaviour that complies with what was observed for
the meta disubstituted methylene flexible counterpart 13a. The
D
Tm of the tightest binders among this series, 6c and 23 (19 and
17.75 ꢁC, respectively), was higher than the
DTm of pentamidine
(12.6 ꢁC) but lower than furamidine (25 ꢁC).
The above results reveal that there are two key factors in the
strong DNA binding of several of this series of dications. First the
bridge effect, where weak binding is clearly caused by having two
eCH2- or two -O- in a single compound. These findings are in line
with our previous findings during our investigation of dicationic
-O- flexible triaryl guanidines and imidamides [19]. This is appar-
ently because bibridging of the central aromatic ring to the ter-
minal amidine bearing rings produces highly twisted compounds
that can not effectively enter the minor groove. Stronger binding
occurs with only one eCH2- or -O- as a linker between the aryl rings
giving less twisted amidines. The less twisted molecular structure
can better fit into the minor groove.
The hydrogen bond analysis showed that hydrogen bond number
varied between 0 and 4 throughout the 50 ns (Fig. 3, lower right
panel). Hence, thetrajectoryframes between the 40thnsand the50th
ns were clustered using an RMSD cutoff value of 2 Å, and the four
The second factor is a charge on the central pyridine ring. There
are three pyridines derivatives in the class 6c, 23 and 27. While both
6c, 23 have the two highest
D
Tm values (19 and 17.75 ꢁC,
5

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Table 1
Antiprotozoal assessment and thermal melting analysis results of the new derivatives.
Compd#
X
R
Position
D
Tm poly dA-dTa T. b. r. IC50 (nM)b P. f. IC50 (nM)c Cytotox. IC50
(m
M)d SI/T. b. r.e SI/P. f.f In vivo T. b. r. (
mM)
Cures, days
5
6a
6b
6c
CH OMe
1,3- (meta) NTg
1,3- (meta) 8.5
16.6 K
35
73
>9.5 K
31
3
11.2
183
>9.5 K
>9.5 K
23
34.8
2.0
4.0
18.1
1.1
32.2
37.3
28.3
19.4
25.1
91.0
2.1
116.6
99.3
46
2.1
3.7
0/4, 13 dh
0/4, 22.75 di
0/4, 22 d
4/4 cures, >60 d
0/4, 7 d
0/4, 9.2 d
NT
0/4, 15.5 d
0/4, 16.25 d
0/4, 18.75 d
NT
2/4, >45.5 d
NT
NT
2/4, 20 d
1/4, 24 d
0/4, 11 d
CH
CH
N
H
H
H
57.1
54.8
441.4
10.5
1.7
0.29
1886
269.4
929.6
6.3
190.9
4.5
262.7
20,909
1422
e
64.5
1333
1616
6
>3.4
>3.9
1230
970
1394
9.7
525
12.3
307.4
991
413
e
1,4- (para)
9
2,5- (para) 19.0
41
7
CH OCO(CH2)3N(CH3)2 1,4- (para) NT
105
18.7 K
126 K
15
72
27
14.4 K
11
26.1 K
378
2.2
12a
12b
13a
13b
17
22
23
26
27
e
e
e
e
e
e
e
e
e
e
e
e
OMe
OMe
H
H
H
OMe
H
OMe
H
1,3- (meta) NT
1,4- (para) NT
1,3- (meta) 0.5
1,4- (para)
1
20
18
e
e
e
e
e
e
e
e
2.25
NT
17.75
NT
1.25
12.6
25
9.4 K
4
>9.5 K
323
46.4
15.5
e
PNTj
FMDj
Controlk
e
e
4.5
e
6.40
e
e
e
a
Thermal melting increase of poly (dA-dT)n.
b
c
d
e
f
STIB900 was the strain of T. b. r. (Trypanosoma brucei rhodesiense) used. Values are the average of duplicate determinations.
IC50 values obtained against the chloroquine resistant P. f. (Plasmodium falciparum) strain K1. Values are the average of duplicate determinations.
Cultured L6 rat myoblast cells were used for cytotoxicity assessment.
Selectivity Index for T. b. r. is the ratio: IC50 (L6)/IC50 (T. b. r.).
Selectivity Index for P. f. is the ratio: IC50 (L6)/IC50 (P. f.).
NT ¼ not tested.
g
h
i
In vivo efficacy was determined for prodrugs in T. b. r. (STIB900) infected mice at 4 ꢀ 25 mg/kg p. o. dose.
In vivo efficacy was determined for diamidines in T. b. r. (STIB900) infected mice at 4 ꢀ 5 mg/kg i. p. dose.
PNT ¼ pentamidine; FMD ¼ furamidine.
j
k
Control group animals receiving saline (drug free treatment).
resulting clusters were analyzed for their hydrophobic and hydrogen
bond interactions (Fig. 4). In each of the four clusters, the aromatic
rings of compound 6c are stalked in between the DNA backbone
deoxyribose sugar rings. Hydrogen bonds were not consistent among
the clusters where clusters 1, 2, and 3 showed three different in-
teractions, each with a different nitrogen atom of compound 6c. In
cluster 1, the N4 atom acted as a hydrogen bond donor for oxygen
from the phosphate group of residue dT-19 from chain B (N4 to O1P of
dT-19). Incluster2,theprotonatedNatomwithinthecentralaromatic
ring made an electrostatic contact with the oxygen atom within the
deoxyribose ring of residue dT-8 from chain A (N to O4 of dT-8). In
cluster 3, the N2atom of the ligand acted as a hydrogen bond donor to
the carbonyl group at position 2 of the thymine ring in residue dT-19
fromchainB(N2toO2ofdT-19). Meanwhile,cluster4containedthree
hydrogen bonds, the N4 atom shifted its hydrogen bond from chain B
to the oxygen from the phosphate group of residue dC-9 from chain A
(N4 to O1PofdC-9), and theN2maintained thesame bondascluster 3
andmadeasimilaronewiththecorrespondingcarbonylgroupwithin
residue dT-20 (N4 to O1P of dT-20). For further investigation of those
five different polarcontacts, distancecalculations were performed for
each bond at a time using VMD. The analysis showed that the bond
presentincluster1(N4toO1PofdT-19), andtheoneincluster4(N2to
O2 of dT-20) were maintained to a great extent through the last ten ns
of the simulation. The bond of cluster 3 (N2 to O2 of dT-19) appeared
only between the 46th and the 48th ns, while the remaining two
bonds in clusters 2 and 4 only appeared occasionally (Fig. 5).
The predicted binding mode of 6c, especially cluster 4, is
relatively comparable to that observed for the reference minor
groove binder 2-[5-(4-carbamimidoylphenyl)thiophen-2-yl]-3H-
benzimidazole-5-carboximidamide in its co-crystal complex with
6

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Fig. 2. Circular dichroism spectra of the titration of representative compounds, 23 and 6c, with a 20
mM Poly (dA). Poly (dT) sequence in the buffer (10 mM MES, 100 mM NaCl, 1 mM
EDTA) at 25 ꢁC. The black arrows indicate positive induced changes.
Fig. 3. Upper left panel: RMSD of the MD simulation of DNA with reference to its initial position. Upper right panel: DNA Radius of gyration during the MD simulation. Lower left
panel: RMSD of the ligand's heavy atoms relative to the DNA. Lower right panel: Alteration in the number of hydrogen bonds between 6c and DNA during MD.
Table 2
Binding free energy elements calculations for the diamidine 6c.
Van der Waals energy (kJ/mol)
Electrostatic energy (kJ/mol)
Polar solvation energy (kJ/mol)
109.785 27.423
SASA energy (kJ/mol)
Binding free energy (kJ/mol)
ꢂ153.150 14.653
ꢂ59.220 16.886
ꢂ18.047 1.717
ꢂ120.631 19.159
the DNA dodecamer d (CGCGAATTCGCG)2 and (PDB ID: 1VZK)
[27]. The diamidine 6c formed hydrogen bonds with more or less
the same nitrogenous bases recognized by the co-crystallized
ligand viz. dT-7, dT-8, dC-9 in chain A and with dA-18, dT-19 and
dT-20 in chain B of the DNA dodecamer (Figure 43S, supplemen-
tary material).
7

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Fig. 4. Cluster analysis of the last 10 ns of the MD study for DNA-compound 6c complex (6c carbons in cyan and DNA carbons in green, H bonds are presented as red dashed lines).
compounds 13a,b and 17 were in general slightly more active than
their monoflexible analogues 6a-c. Focusing on the monoflexible
derivatives 6a-c, it was observed that the 3-benzyl-1,10-biphenyl
derivative 6a was almost twice as active as its para positional
analogue 6b (IC50 ¼ 35 and 73 nM, respectively). Also, the 5-benzyl-
2-phenylpyridine derivative 6c was more potent than its non-aza
analogue 6b (IC50 ¼ 43 and 73 nM, respectively). Within the
diflexible series, it was found that curvature in the structure of the
backbone ring system enhanced activity whereas the 1,3-
dibenzylbenzene derivative 13a (IC50 ¼ 15 nM), was more active
than derivative 17 with a 2,5-diphenylfuran central ring system
(IC50 ¼ 27 nM), both showing higher potency than the compara-
tively more linear 1,4-dibenzylbenzene scaffold-based counterpart
13b (IC50 ¼ 72 nM). For those derivatives with one (23) or two (27)
oxygen atom bridge(s) between the triaryl system, it was found that
monoflexibility was much in favour of the antitrypanosomal ac-
tivity (IC50 ¼ 11 and 378 nM, respectively).
Considering the in vitro antimalarial screening results against
P. f., the new dicationic diamidines exhibited an IC50 values be-
tween 3 and 323 nM. With the exception of 27, all the new dia-
midines were more potent than the reference drug pentamidine
Fig. 5. Distance analysis of polar contacts from the cluster analysis of the DNA-
compound 6c complex.
(IC50 ¼ 46.4 nM) and both 6b (IC50 ¼ 3 nM) and 23 (IC50 ¼ 4 nM)
showed respective almost 5 and 4 fold higher potency than fur-
2.5. In vitro antiprotozoal activity and in vivo antitrypanosomal
activity
amidine (IC50 ¼ 15.5 nM). Within the monoflexible series 6a-c, the
3-benzyl-1,10-biphenyl derivative 6a was about ten fold less active
than their para positional analogue 6b (IC50 ¼ 31 and 3 nM,
respectively), while the aza counterpart 6c possessed intermediate
activity (IC50 ¼ 11.2 nM). With respect to the diflexible congeners
13a,b and 17, neither the relative position of the methylene bridges
nor the nature of the central ring (phenyl versus furan) had a
profound effect on efficacy (IC50 ¼ 23, 20 and 18 nM, respectively).
Regarding the oxy flexible compounds with a central pyridine ring
23 and 27, monoflexibility imparted an almost 80-fold more active
derivative than its diflexible analogue (IC50 ¼ 4 and 323 nM,
respectively).
Antiprotozoal assessment results for the new synthesized
compounds are displayed in Table 1 where the diamidines and their
respective prodrugs were tested against both T. b. r. and P. f. in vitro.
Concerning the in vitro antitrypanosomal activity, tested dia-
midines possessed IC50 values between 11 and 378 nM. While none
of these dications was more potent than the reference compounds
pentamidine and furamidine (IC50 ¼ 2.2 and 4.5 nM, respectively),
the most active ones were still potent at the low nanomolar range.
Structure-activity relationship (SAR) findings drawn from the anti-
T. b. r. results (Fig. 6) demonstrate that among the diamidines with
methylene bridge flexibility (6a-c, 13a,b and 17), the diflexible
Moreover, as expected prodrugs did not show any in vitro ac-
tivity against both tested parasites. An exception was the
8

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Fig. 6. Summary of SAR findings drawn from the anti-T. b. r. results.
dimethylaminoethoxycarbonyl ester prodrug 7 which surprisingly
showed a weak in vitro anti-trypanosomal and anti-malarial ac-
tivity (IC50 ¼ 105 and 183 nM, respectively).
3. Conclusion
This research work presents the synthesis of a new series of
mono and diflexible triaryl dicationic diamidines and their pro-
drugs. The dications were evaluated in vitro against T. b. r. and P. f.
and in vivo against T. b. r. STIB 900 mouse model. The tested dica-
tionic compounds were more selective towards T. b. r.
(SI ¼ 54.8e1886) and P. f. (SI ¼ 64.5e1394) than L6 rat myoblasts
reflecting a good therapeutic window and high safety profile. The
monoflexible derivative 6c was the most potent in this study
To assess the safety profile of these dicationic diamidines, their
cytotoxic profile was assessed against cultured L6 rat myoblast cells
where they demonstrated modest cytotoxicity displaying IC50
values of 2e116.6 mM. Selectivity indices of all compounds were
calculated and presented in Table 1. These cations were more se-
lective towards T. b. r. (SI ¼ 54.8e1886) and P. f. (SI ¼ 64.5e1394)
than L6 rat myoblasts reflecting a good therapeutic window and a
high safety profile fairly comparable to that of both reference
dications pentamidine and furamidine.
showing the best DNA binding (
DTm ¼ 19), with SI of 441.4 and 1616
against T. b. r. and P. f. respectively, and the best in vivo effect to-
wards T. b. r. infected mice giving 4/4 cures for a treatmet of
4 ꢀ 5 mg/kg p. o. dose over a period of 60 days. Additionally, a 50 ns
molecular dynamics study of 6c binding with an AT-rich DNA
dodecamer demonstrated a strong binding complex supported by
vdW and electrostatic interactions. Members of this series, espe-
cially compound 6c, have unraveled new understanding of the
limitations and requirements for effective DNA binding and
consequent anti-parasitic activity. The mono-flexible molecules
appear to have an advantage over the bi-flexible counterparts. Also,
the methylene monoflexible compounds are better than the oxo
flexible ones. These findings are in compliance with our previous
findings of di-oxofelxible triaryl guanidines and imidamides.
Additionally, introducing a nitrogen atom on one ring of the aro-
matic ring system seems to impart better activity to those com-
pounds. Again a finding that is in line with our work on other
diamidine frameworks like benzimidaoles and others. In conclu-
sion, 6c represents a good lead which merits further optimization.
Future investigations could be directed towards synthesis of new
aza bioisosteric analogues with a nitrogen atom on the amidine
bearing terminal aryl rings and studying the effect of alternation of
the position of the N atom on the activity of those monomethylene
flexible congeners of 6c.
We still do believe that a full understanding of all the underlying
molecular mechanisms associated with the biological effects of
these and similar dicationic compounds is far from being
completely acknowledged and there is still much room for inves-
tigation in this area.
2.6. In vivo antitrypanosomal activity
The in vivo antitrypanosomal activity assessment of the new
compounds was performed against a mouse model of T. b. r.
STIB900 infected mice and data are presented in Table 1. Prodrugs
were administered p. o. and diamidines were injected i. p. as
detailed in the experimental section and animals were observed for
cures for a period of 60 days. Best in vivo effect was displayed by the
monoflexible diamidine 6c having a methylene bridge and central
pyridine ring where treatment with which resulted in 4/4 cures
over the 60 days treatment course. Second best was the analogue
23 which was also monoflexible but with an oxygen bridge which
elicited 2/4 cures for a treatment period of >45.5 days. Those two
new diamidines were considerably more effective in vivo than the
reference dicationic minor groove binders pentamidine (2/4 cures
over 20 days) and furamidine (1/4 cures over 24 days). All other
in vivo tested diamidines and prodrugs gave 0/4 cures with modest
prolongation in animals’ survival during the treatment course
(7e22.75 days).
4. Experimental protocols
4.1. Chemical synthesis
The combined results of the in vitro and in vivo studies highlight
the better activity of the monoflexible amidines as opposed to the
diflexible ones. Additionally, those analogues with an azaheter-
ocyclic ring demonstrated better antiprotozoal activity than their
respective bioisosteres. Thus, future work would better focus on
investigating more monomethylene flexible aza derivatives.
Melting points were recorded on a Thomas-Hoover (Uni-Melt)
capillary melting point apparatus and were uncorrected. TLC
analysis was performed using silica gel 60 F254 precoated
aluminium sheets and was detected using a UV lamp. 1H and 13C
NMR spectra were recorded by means of a Varian GX400 or Varian
Unity Plus 300 spectrometer, and chemical shifts
(d) were
9

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
presented in ppm with reference to the internal standard TMS.
Mass spectra were obtained by a VG analytical 70-SE spectrometer.
Elemental analyses were performed at Atlantic Microlab Inc.
(Norcross, GA) and lie within 0.4 of the theoretical values. Com-
pounds reported in their salt forms analyzed correctly for fractional
moles of water and/or ethanol of solvation where in each case 1H
NMR affirmed the existence of the specified solvent(s). All chem-
icals and solvents were acquired from Aldrich Chemical Co., Fisher
Scientific, Frontier or Lancaster.
4.1.5. N0-Hydroxy-4'-(4-(N0-hydroxyamidino)benzyl)-[1,10-
biphenyl]-4-carboxamidine (4b)
Compound 4b was prepared employing the same methodology
adopted for preparation of compound 4a starting from the corre-
sponding dinitrile 3b. Yield % 94, mp 213e214.5 ꢁC. 1H NMR
(DMSO‑d6):
d 3.39 (s, 2H), 5.70 (s, 2H), 5.79 (s, 2H), 7.19e7.28 (m,
4H), 7.53e7.71 (m, 8H), 9.52 (s, 1H), 9.63 (s, 1H). 13C NMR
(DMSO‑d6):
d 150.7, 142.0, 137.3, 132.2, 131.4, 129.3, 128.5, 126.7,
126.1, 125.9, 125.6, 40.4. Mass (m/z, rel. int.): 361 (Mþþ1, 50), 271
(10), 243 (12), 226 (5), 181 (100).
4.1.1. 3'-(4-cyanobenzyl)-[1,10-biphenyl]-4-carbonitrile (3a)
4.1.6. 2-(4-(N-Hydroxyamidino)phenyl)-5-(400-N-
hydroxyamidinobenzyl)pyridine (4c)
To
a stirred solution of 1-bromo-3-(chloromethyl)benzene
(2.039 g, 10 mmol) and Pd(PPh3)4 (320 mg) in toluene (20 mL), an
aqueous solution of Na2CO3 (2 M, 10 mL) was added, followed by 4-
cyanophenylboronic acid (3.234 g, 22 mmol) in methanol (8 mL).
The stirred reaction mixture was heated at 80 ꢁC for overnight, after
this the reaction mixture was partitioned between an aqueous
solution containing concentrated ammonia (20 mL) and methylene
chloride (400 mL, 3x). The organic layer was evaporated to dryness
under reduced pressure and the solid was filtered off and recrys-
tallized from ethanol to furnish the biphenyl derivative 3a. Yield %
Compound 4c was prepared adopting the same methodology
used for preparation of compound 4a starting from the corre-
sponding dinitrile 3c. Yield % 99, mp 205e206 ꢁC. 1H NMR
(DMSO‑d6):
d 4.01 (s, 2H), 5.77 (s, 2H), 5.85 (s, 2H), 7.28 (d,
J ¼ 8.4 Hz, 2H), 7.60 (d, J ¼ 8.4 Hz, 2H), 7.68e7.78 (m, 3H), 7.90 (d,
J ¼ 8.1 Hz,1H), 8.05 (d, J ¼ 8.4 Hz, 2H), 8.59 (d, J ¼ 1.8 Hz,1H), 9.57 (s,
1H), 9.71 (s, 1H). 13C NMR (DMSO‑d6):
d 153.4, 150.7, 150.5, 149.6,
141.3, 138.8, 137.2, 135.6, 133.5, 131.3, 128.4, 126.0, 125.7, 125.6,
120.0, 37.4. MS (ESI) m/e (rel. int.): 362 (Mþ þ 1, 60), 181 (100).
75, mp 123e124 ꢁC; 1H NMR (DMSO‑d6):
d 4.12 (s, 2H), 7.30 (d,
J ¼ 6.9 Hz, 1H), 7.40e7.50 (m, 3H), 7.57 (d, J ¼ 6.9 Hz, 1H), 7.64 (s,
4.1.7. (N0-Methoxy-3'-(3-(N0-methoxyamidino)benzyl)-[1,10-
biphenyl]-4-carboxamidine hydrochloride salt (5)
1H), 7.73 (d, J ¼ 8.1 Hz, 2H), 7.81e7.94 (m, 4H). 13C NMR (DMSO‑d6):
d
147.1, 144.4, 141.1, 138.5, 133.0, 132.8, 132.4, 129.7, 129.4, 129.2,
A solution of the amidoxime 4a (0.55 gm, 1.52 mmol) in DMF
(10 mL) was treated with a 5 mL aqueous solution of LiOH$H2O
(0.76 gm, 18.31 mmol). Dimethylsulfate (1.45 mL, 15.26 mmol) was
added after 15 min and the reaction was kept stirring overnight at r.
t. The reaction mixture was poured onto ice/water and extracted
with ethyl acetate. The organic layer was separated, washed with
brine (3 ꢀ 10 mL) and evaporated to dryness under vacuum to give
the target methoxyamidine 5. Hydrochloride salt of 5 was pre-
pared by treating the free base with an alcoholic solution of HCl
(gas) and stirring at r. t. overnight. Dilution with ethyl ether
resulted in precipitation of 5 HCl salt which was filtered, washed
with diethylether and dried. Yield % 86, mp 252e254 ꢁC. 1H NMR
128.0, 127.5, 125.1, 118.9, 110.0, 108.9, 40.8. Mass (m/z, rel. int.): 294
(Mþ, 100).
4.1.2. 4'-(4-cyanobenzyl)-[1,10-biphenyl]-4-carbonitrile (3b)
Dinitrile compound 3b was prepared employing a Suzuki
coupling methodology used for 3a starting with 1-bromo-4-
(chloromethyl)benzene instead of 1-bromo-3-(chloromethyl)ben-
zene. Yield % 73, mp 144e145.5 ꢁC; 1H NMR (DMSO‑d6):
d
4.07 (s,
2H), 7.36e7.45 (m, 4H), 7.60e7.85 (m, 8H). 13C NMR (DMSO‑d6):
d
146.9, 144.3, 140.8, 136.3, 132.8, 132.4, 129.7, 129.6, 127.4, 127.3,
118.9, 118.8, 109.8, 108.9, 40.5. Mass (m/z, rel. int.): 294 (Mþ, 100).
(DMSO‑d6):
d 4.01 (s, 2H), 5.77 (s, 2H), 5.85 (s, 2H), 7.28 (d,
J ¼ 8.4 Hz, 2H), 7.60 (d, J ¼ 8.4 Hz, 2H), 7.68e7.78 (m, 3H), 7.90 (d,
4.1.3. 2-(40-cyanophenyl)-5-(400-cyanobenzyl)pyridine (3c)
J ¼ 8.1 Hz,1H), 8.05 (d, J ¼ 8.4 Hz, 2H), 8.59 (d, J ¼ 1.8 Hz,1H), 9.57 (s,
Dinitrile compound 3c was prepared employing a Suzuki
coupling methodology used for 3a starting with 2-chloro-5-
(chloromethyl)pyridine instead of 1-bromo-3-(chloromethyl)ben-
1H), 9.71 (s, 1H). 13C NMR (DMSO‑d6):
d 153.4, 150.7, 150.5, 149.6,
141.3, 138.8, 137.2, 135.6, 133.5, 131.3, 128.4, 126.0, 125.7, 125.6,
120.0, 37.4. MS (ESI) m/e (rel. int.): 362 (Mþ þ1, 60), 181 (100). Anal.
Calcd. for C23H24N4O2-2.0HCl-0.2H2O: C, 59.40; H, 5.72; N, 12.04.
Found: C, 59.21; H, 5.68; N, 11.96.
zene. Yield % 83, mp 187e187.5 ꢁC. 1H NMR (DMSO‑d6):
d 4.13 (s,
2H), 7.50 (d, J ¼ 8.4 Hz, 2H), 7.74e7.79 (m, 3H), 7.91 (d, J ¼ 8.4 Hz,
2H), 7.99 (d, J ¼ 7.8 Hz, 1H), 8.21e8.24 (m, 2H), 8.65 (s, 1H). 13C NMR
(DMSO‑d6):
d 152.2, 150.0, 146.2, 142.5, 137.6, 135.8, 132.7, 132.5,
4.1.8. 3'-(4-Amidinobenzyl)-[1,10-biphenyl]-4-carboxamidine
acetate salt (6a)
129.7, 127.0, 121.0, 118.7, 111.2, 109.2, 37.5. MS (ESI) m/e (rel. int.):
296 (Mþ þ 1, 100).
A solution of diamidoxime 4a (360 mg, 1 mmol) in glacial acetic
acid (10 mL) was slowly treated with acetic anhydride (0.35 mL).
After stirring overnight, TLC indicated complete acylation of the
starting material. The reaction mixture was poured onto ice-water
and the obtained precipitate was filtered off and washed with
water. 80 mg of palladium on carbon (10%.) was added to the
acetylated product dissolved in glacial acetic acid (13 mL), and
ethanol (20 mL). The mixture was placed on a Parr hydrogenator for
4 h under 50 psi at r. t. The mixture was filtered off through hyflo
and the filter pad was washed with water. The filtrate was evapo-
rated under vacuum and the precipitate was filtered and washed
with ether giving the diamidine 6a. Yield % 74, mp 245e247 ꢁC. 1H
4.1.4. N0-Hydroxy-3'-(4-(N0-hydroxyamidino)benzyl)-[1,10-
biphenyl]-4-carboxamidine (4a)
Potassium t-butoxide (1.68 g, 15 mmol, 10 equiv.) was added
portionwise to a mixture of hydroxylamine hydrochloride (1.042 g,
15 mmol, 10 equiv.) in anhydrous DMSO (12 mL) that was cooled to
5 ꢁC. The mixture was stirred for 30 min before adding the dinitrile
3a (441 mg, 1.5 mmol, 1 equiv.). The reaction mixture was kept
stirring overnight at r. t. after which it was poured slowly onto ice-
water. The obtained precipitate was filtered off and washed with
water followed by ethanol. Yield % 89, mp 195e196 ꢁC. 1H NMR
(DMSO‑d6):
4H), 7.49e7.64 (m, 5H), 7.69e7.77 (m, 3H), 9.46 (s, 1H), 9.58 (s, 1H).
13C NMR (DMSO‑d6):
150.7, 150.4, 142.0, 141.9, 140.4, 139.7, 132.3,
d
4.02 (s, 2H), 5.65 (s, 2H), 5.75 (s, 2H), 7.21e7.38 (m,
NMR (DMSO‑d6): d 1.77 (s, 2 x CH3), 4.13 (s, 2H), 7.29.21e7.38 (m,
4H), 7.55e7.73 (m, 4H), 7.79e7.90 (m, 4H). Mass (m/z, rel. int.): 329
(Mþ þ1, 28), 312 (100). HRMS Calc. for C21H21N4 MHþ 329.1766.
Obsd: 329.1761. Anal. Calcd. for C21H20N4-2.0AcOH-1.8H2O: C,
62.43; H, 6.62; N, 11.64. Found: C, 62.44; H, 6.29; N, 11.25.
d
131.2, 129.1, 128.4, 128.0, 127.0, 126.3, 125.8, 125.6, 124.3, 40.8. Mass
(m/z, rel. int.): 361 (Mþþ1, 100).
10

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
4.1.9. 4'-(4-Amidinobenzyl)-[1,10-biphenyl]-4-carboxamidine
acetate salt (6b)
4.1.14. N0-Hydroxy-4-(3-(4-N0-hydroxyamidino)benzyl)benzyl)
benzamidine (11a)
The same procedure described for the preparation of 6a was
used starting with 4b. to obtain the diacetate salt 6b. Yield % 83, mp
The method utilized for synthesis of diamidoxime compound 4a
was adopted for synthesis of compound 11a starting with dinitrile
263.5e265 ꢁC. 1H NMR (D2O/DMSO‑d6):
d
1.77 (s, 2 x CH3), 4.04 (s,
10a. Yield % 91, mp 160e161.5 ꢁC. 1H NMR (DMSO‑d6):
d
3.91 (s, 4H),
2H), 7.48e7.60 (m, 4H), 7.76e7.92 (m, 8H). Mass (m/z, rel. int.): 329
(Mþþ1, 80), 241 (10), 165 (100). Anal. Calcd. For C21H20N4-2.0AcOH-
1.9H2O: C, 62.20; H, 6.59; N, 11.60 Found C, 62.25; H, 6.21; N, 11.30.
6.21 (br s, 4H), 7.02e7.23 (m, 8H), 7.58 (d, J ¼ 8.4 Hz, 4H), 9.78 (br s,
2H); 13C NMR (DMSO‑d6):
d 152.1, 142.9, 141.1, 129.8, 129.1, 128.6,
128.5, 126.5, 125.9, 40.7. MS (ESI: m/z, rel. int.); 375 (Mþþ1, 100);
HRMS Calc. for C22H23N4O2 MH þ mass 375.1821, Obsd. 375.1823.
4.1.10. 2-(40-Amidinophenyl)-5-(400-amidinobenzyl)pyridine
acetate salt (6c)
4.1.15. N0-Hydroxy-3-(4-(4-N0-hydroxyamidino)benzyl)benzyl)
benzamidine (11b)
The procedure described for preparation of 6a was utilized
The method utilized for synthesis of amidoxime 4a was adopted
for synthesis of 11b starting with dinitrile 10b. Yield % 98, mp 218-
employing 4c. Yield % 80, mp 258e259 ꢁC. 1H NMR (D2O/DMSO‑d6):
20 ꢁC. 1H NMR (DMSO‑d6):
d 3.87 (s, 4H), 5.70 (br s, 4H), 7.12 (s, 4H),
d
1.80 (s, 3 x CH3), 4.12 (s, 2H), 7.46 (d, J ¼ 7.2 Hz, 2H), 7.72e7.78 (m,
7.18 (d, J ¼ 8.1 Hz, 4H), 7.55 (d, J ¼ 8.1 Hz, 4H), 9.52 (br s, 2H); 13C
3H), 7.87e7.98 (m, 3H), 8.18 (d, J ¼ 7.8 Hz, 2H), 8.64 (s, 1H). MS (ESI)
m/e (rel. int.): 330 (Mþ þ1, 50), 165 (100). Anal. Calc. for C20H19N5-
3.0CH3CO2H-0.45H2O): C% 60.32, H% 6.21, N% 13.53; Found C%
59.96, H% 6.24, N% 13.92.
NMR (DMSO‑d6):
40.4; HRMS Calc. for C22H23N4O2 MH
375.1824.
d
150.8, 142.2, 138.8, 1þ31.0, 128.8, 128.4, 125.5,
mass 375.1821, Obsd.
4.1.16. N0-Methoxy-3-(3-(4-(N0-methoxyamidino)benzyl)benzyl)
benzamidine hydrochloride salt (12a)
4.1.11. 4'-(N-(4-(Dimethylamino)butanoyl)oxy)-4-amidinobenzyl)-
[1,10-biphenyl]-4-(N'-(4-(dimethylamino)butanoyl)oxy)
A solution of the amidoxime 11a (0.39 gm, 1.05 mmol) in DMF
(8 mL) was treated with a 2 mL aqueous solution of LiOH$H2O
(0.1 gm, 2.30 mmol). After 15 min of stirring methyliodide (0.2 mL,
3.14 mmol) was added and the reaction mixture was kept stirring
overnight at r. t. The reaction mixture was poured onto ice/water
and extracted with ethyl acetate. The organic layer was separated,
washed with brine (3 ꢀ 10 mL) and evaporated to dryness under
vacuum to give the free base of methoxyamidine 12a. 1H NMR
carboxamidine) (7)
A solution of 4-(dimethylamino)butyric acid hydrochloride
(670 mg, 4 mmol), 1,10-carbonyl-diimidazole (810 mg, 5 mmol) in
DMF (10 mL) was heated at 65e70 ꢁC maintaining the heating for
30 min. Then the reaction mixture was allowed to cool down to
room temperature before adding the diamidoxime 4b (337 mg,
1 mmol), and Et3N (607 mg, 6 mmol). The reaction was kept stirring
for 24 h. Where after the reaction mixture was poured onto saline
solution and the solid was collected by filtration to furnish com-
pound 7 (Free base). Yield % 90 mp 149.5e151 ꢁC. 1H NMR
(DMSO‑d6):
d 3.70 (s, 2xOCH3), 3.89 (s, 4H), 5.85 (s, 2H), 5.98 (br s,
4H), 7.02 (d, J ¼ 7.2 Hz, 2H), 7.12 (s, 1H), 7.16e7.21 (m, 5H), 7.54 (d,
J ¼ 8.1 Hz, 4H). 13C NMR (DMSO‑d6):
d 150.9, 142.5, 141.1, 130.3,
(DMSO‑d6);
d 1.72e1.78 (m, 4H), 2.16 (s, 12H), 2.25e2.31 (m, 4H),
129.1, 128.6, 128.4, 126.5, 125.9, 60.5, 40.8. Hydrochloride salt of
12a was prepared by treating the free base with and alcoholic so-
lution of HCl and stirring at r. t. overnight. Dilution with dieth-
ylether resulted in precipitation of 12a HCl salt which was filtered,
washed with diethylether and dried. Yield % 81, mp 136e138 ꢁC. 1H
2.46e2.53 (m, 4H), 4.07 (s, 2H), 6.77 (s, 2H), 6.85 (s, 2H), 7.37 (d,
J ¼ 8.1 Hz, 4H), 7.67 (d, J ¼ 8.1 Hz, 4H), 7.74e7.84 (m, 4H). 13C NMR
(DMSO‑d6);
d 171.0, 156.5, 156.2, 143.7, 141.8, 140.8, 137.1, 130.4,
129.6, 129.4, 128.7, 127.2, 126.9, 126.8, 126.3, 58.3, 45.1, 30.3, 22.4.
MS (m/z, rel. int.); 587 (Mþþ1, 10), 474 (60), 343 (30), 163 (100).
HRMS Calc. for C33H43N6O4 MHþ 587.3346. Obsd: 587.3337. Anal.
Calcd. for C33H42N6O4-0.25H2O: C, 67.03; H, 7.24; N, 14.21. Found: C,
67.00; H, 7.10; N, 14.26.
NMR (DMSO‑d6):
d 3.97 (s, 2xOCH3), 4.08 (s, 4H), 5.85 (s, 2H), 7.05
(d, J ¼ 7.2 Hz, 2H), 7.18e7.23 (m, 3H), 7.37 (d, J ¼ 8.1 Hz, 2H), 7.65 (d,
J ¼ 8.1 Hz, 4H). HRMS Calc. for C24H27N4O2 MHþ 403.2134. Obsd:
403.2147. Anal. Calcd. for C24H26N4O2-2.0HCl-0.5H2O-0.5C2H5OH:
C, 59.17; H, 6.35; N, 11.04. Found: C, 59.20; H, 6.24; N, 10.77.
4.1.12. 4,4'-(1,3-Phenylenebis(methylene))dibenzonitrile (10a)
A solution of 4-(bromomethyl)benzonitrile (3.90 g, 20 mmol)
and 1,3-phenylenediboronic acid (1.66 gm, 10 mmol) was reacted
under the above mentioned Suzuki coupling conditions utilized for
synthesis of 3a to give the target dinitrile 10a. Yield % 70, mp
4.1.17. N0-Methoxy-3-(4-(4-(N0-methoxyamidino)benzyl)benzyl)
benzamidine (12b)
The same procedure described for the preparation of 12a was
used starting with 11b. Free base of 12b, 1H NMR (DMSO‑d6):
d 3.70
103e104 ꢁC; 1H NMR (DMSO‑d6):
d
3.98 (s, 4H), 7.06 (d, J ¼ 8.1 Hz,
(s, 2xOCH3), 3.88 (s, 4H), 5.90 (s, 4H), 7.12 (s, 4H), 7.19 (d, J ¼ 8.4 Hz,
4H), 7.54 (d, J ¼ 8.4 Hz, 4H). 13C NMR (DMSO‑d6):
d 151.0, 142.6,
2H), 7.15 (s, 1H), 7.19e7.24 (m, 1H), 7.40 (d, J ¼ 8.1 Hz, 4H), 7.73 (d,
J ¼ 8.1 Hz, 4H); 13C NMR (DMSO‑d6):
d 147.2, 140.3, 132.3, 129.6,
138.7, 130.3, 128.7, 128.4, 125.8, 60.5, 40.3. Hydrochloride salt of
12b was prepared by treating the free base with alcoholic solution
of hydrogen chloride and stirring at r. t. overnight. Dilution with
ethyl ether resulted in precipitation of 12b HCl salt which was
filtered, washed with diethyl ether and dried. Yield % 73, mp
129.3, 128.9, 126.8, 118.9, 108.8, 40.8. MS (ESI: m/z, rel. int.); 309
(Mþ þ1, 100); HRMS Calc. for C22H17N2 MHþ 309.1392. Obsd:
309.1391.
147e148 ꢁC. 1H NMR (D2O/DMSO‑d6):
d 3.83 (s, 2xOCH3), 3.96 (s,
4.1.13. 3-(4-(4-Cyanobenzyl)benzyl)benzonitrile (10b)
4H), 7.18 (s, 4H), 7.40 (d, J ¼ 8.4 Hz, 4H), 7.69 (d, J ¼ 8.4 Hz, 4H).
HRMS Calc. for C24H27N4O2 MHþ 403.2134. Obsd: 403.2140. Anal.
Calcd. for C24H26N4O2-2.0HCl-1.5H2O-0.25C2H5OH: C, 57.25; H,
6.37; N, 10.90. Found: C, 57.48; H, 6.34; N, 10.58.
A solution of 4-(bromomethyl)benzonitrile (3.92 gm, 20 mmol)
and 1,4-phenylenediboronic acid (1.66 gm, 10 mmol) was reacted
employing the aforementioned Suzuki coupling conditions
described for synthesis of 3a to obtain the target dinitrile 10b. Yield
% 76; mp 171e173 ꢁC; 1H NMR (DMSO‑d6):
d
3.97 (s, 4H), 7.15 (s, 4H),
4.1.18. 3-(4-(4-Amidinobenzyl)benzyl)benzamidine acetate salt
(13a)
Synthesis of the target diamidine 13a was conducted adopting
the procedure described for diamidine 6a starting with diamidoxime
7.39 (d, J ¼ 8.4 Hz, 4H), 7.71 (d, J ¼ 8.4 Hz, 4H); 13C NMR (DMSO‑d6):
d
147.3,138.1,132.6,129.7,129.0,118.9,108.8, 40.5. MS (m/z, rel. int.);
309 (Mþþ1, 10), 308 (40), 192 (100).
11

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
11a. Yield % 68, mp 254e256 ꢁC. 1H NMR (D2O/DMSO‑d6):
d
1.74 (s,
4.1.24. 4-(6-(4-Cyanophenoxy)pyridin-3-yl)benzonitrile (20)
The dinitrile derivative 20 was synthesized by reaction of 19
(1.75 gm, 6.36 mmol) with 4-cynaophenylboronic acid (1.02 gm,
7 mmol) under the Suzuki coupling conditions described for
compound 3a. Yield % 72, mp 193e195 ꢁC; 1H NMR (DMSO‑d6):
2xCH3), 3.99 (s, 4H), 7.06e7.21 (m, 4H), 7.37 (s, 4H), 7.70 (s, 4H);
HRMS Calc. for C22H23N4 MH þ mass 343.1923, Obsd. 343.1906; Anal.
Calc. for C22H22N4-2.0AcOH-0.5H2O-0.5EtOH: C, 65.56; H, 6.88; N,
11.32; Found C, 65.36; H, 6.56; N, 10.93.
d
7.28 (d, J ¼ 8.7 Hz, 1H), 7.37 (d, J ¼ 8.7 Hz, 2H), 7.89e7.96 (m, 6H),
4.1.19. 3-(4-(4-Amidinobenzyl)benzyl)benzamidine acetate salt
(13b)
8.31 (dd, J ¼ 2.4, 8.7 Hz, 1H), 8.60 (d, J ¼ 2.4 Hz, 1H); 13C NMR
(DMSO‑d6):
d 162.1, 157.5, 145.9, 140.9, 139.2, 134.2, 132.9, 130.2,
Synthesis of the target diamidine 13b was conducted adopting
the procedure described for diamidine 6a starting with diamidox-
127.4, 122.0, 118.7, 118.5, 112.4, 110.4, 107.1. MS (ESI) m/e (rel. int.):
298 (Mþ þ1, 100).
ime 11b. Yield % 75, mp 256-8 ꢁC. 1H NMR (D2O/DMSO‑d6):
d 1.79 (s,
2xCH3), 3.97 (s, 4H), 7.14 (s, 4H), 7.34 (d, J ¼ 8.1 Hz, 4H), 7.68 (d,
J ¼ 8.1 Hz, 4H); HRMS Calc. for C22H23N4 mass 343.1923, Obsd.
343.1939; Anal. Calc. for C22H22N4e2CH3CO2H-0.9H2O: C, 65.22; H,
6.69; N, 11.70; Found C, 65.43; H, 6.45; N, 11.32.
4.1.25. N0-Hydroxy-4-(6-(4-(N0-hydroxyamidino)phenoxy)pyridin-
3-yl)benzamidine (21)
Synthesis of 21 was performed adopting the method described
for diamidoxime 4a. Yield % 81, mp 198e200 ꢁC. 1H NMR
(DMSO‑d6):
d 5.82 (br s, 2H), 5.85 (br s, 2H), 7.12e7.17 (m, 3H)
4.1.20. 4,4'-(Furan-2,5-diylbis(methylene))dibenzonitrile (15)
7.67e7.78 (m, 6 H), 8.19 (dd, J ¼ 2.7 Hz, J ¼ 8.7 HZ, 1 H), 8.49 (d,
4-(Bromomethyl)benzonitrile (2.90 g, 20 mmol) and 2,5-bis(tri-
n-butylstannyl)furan (6.46 g, 10 mmol), Pd(PPh3)4 (320 mg) in
60 mL 1,4-dioxane heated at 80e90 ꢁC for 12 h the reaction mixture
was extracted with methylene chloride (400 mL, 3x). The organic
layer was evaporated to dryness under reduced pressure and the
solid was collected and recrystallized from ethanol to furnish the
dinitrile derivative 15. Yield % 72, mp 147e148 ꢁC. 1H NMR
J ¼ 2.7 Hz, 1 H), 9.61 (br. s, 1 H), 9.69 (br. s, 1H). 13C NMR (DMSO‑d6):
d
162.5, 154.4, 150.5, 150.4, 145.2, 138.4, 136.8, 132.6, 130.7, 129.8,
126.8, 126.1, 125.9, 120.7, 111.7. MS (ESI) m/e (rel. int.): 363 (Mþ,
100).
4.1.26. N0-Methoxy-4-(6-(4-(N0-methoxyamidino)phenoxy)pyridin-
3-yl)benzamidine hydrochloride salt (22)
(DMSO‑d6):
d
4.00 (s, 4H), 6.05 (s, 2H), 7.37 (d, J ¼ 8.1 Hz, 4H), 7.75
Synthesis of 22 was performed adopting the method described
for methoxyamidine compound 5. Free base of 22, yield % 66, mp
(d, J ¼ 8.1 Hz, 4H). 13C NMR (DMSO‑d6):
d 152.2, 144.2, 132.4, 129.5,
118.9, 109.3, 107.8, 33.5. MS (ESI) m/e (rel. int.): 299 (Mþ þ 1, 100).
148e150 ꢁC. 1H NMR (DMSO‑d6):
d 3.73 (s, 3H), 3.74 (s, 3H), 6.12 (br.
HRMS Calc. for C20H15N2O: 299.1184. Observed: 299.1193.
s, 2H), 6.14 (br. s, 2H), 7.13e7.18 (m, 3 H), 7.68e7.77 (m, 6H), 8.19
(dd, J ¼ 2.7 Hz, J ¼ 8.7 Hz, 1H), 8.49 (d, J ¼ 2.7 Hz, 1H). 13C NMR
4.1.21. N0-Hydroxy-4-((5-(4-(-N0-hydroxyamidino)benzyl)furan-2-
yl)methyl)benzamidine (16)
(DMSO‑d6): d 162.5, 154.8, 150.7, 150.6, 145.3, 138.5, 137.3, 131.8,
130.7, 128.9, 127.3, 126.4, 126.2, 120.8, 111.7, 60.6, 60.5. Hydro-
The method utilized for synthesis of diamidoxime compound 4a
was adopted for synthesis of compound 16 starting with dinitrile
chloride salt of 22. Yield % 82, mp 181e183 ꢁC. 1H NMR (DMSO‑d6):
d
3.85 (s, 3H), 3.86 (s, 3H), 7.26 (dd, J ¼ 1.2 Hz, J ¼ 8.7 Hz,1H), 7.37 (d,
15. Yield % 91, mp 170e171 ꢁC. 1H NMR (DMSO‑d6):
d
3.90 (s, 4H),
J ¼ 8.7 Hz, 2H), 7.83e7.89 (m, 6H), 8.31 (dd, J ¼ 1.2 Hz, J ¼ 8.7 Hz,
1H), 8.57 (s, 1H). HRMS Calc. for C21H22N5O3 MHþ 392.1723. Obsd:
392.1724. Anal. Calcd. for C21H21N5O3-2.0HCl-1.75H2O: C, 50.86; H,
5.38; N, 14.12. Found: C, 51.02; H, 5.44; N, 13.90.
5.75 (s, 4H), 5.99 (s, 2H), 7.18 (d, J ¼ 8.4 Hz, 4H), 7.57 (d, J ¼ 8.4 Hz,
4H), 9.56 (s, 1H). 13C NMR (DMSO‑d6):
d 153.0, 150.7, 139.1, 131.5,
128.2, 125.5, 107.1, 33.4. MS (ESI) m/e (rel. int.): 365 (Mþ þ1, 100).
4.1.22. 4,4'-(Furan-2,5-diylbis(methylene))dibenzamidine acetate
salt (17)
4.1.27. 4-(6-(4-Amidinophenoxy)pyridin-3-yl)benzamidine
hydrochloride salt (23)
Synthesis of the target diamidine 17 was conducted adopting
the procedure described for diamidine 6a starting with diamidox-
The dinitrile 20 (298 mg, 1 mmol) was suspended in freshly
distilled THF (5 mL) and treated with lithium trimethylsilylamide
(1 M solution in THF, 6 mL, 6 mmol). The reaction was kept stirring
overnight. The reaction mixture was then cooled to 0 ꢁC and then
15 mL of ethanol/HCl (gas) was added whereupon a precipitate
started forming. The reaction mixture was left to run overnight
where after it was diluted with ether and the precipitate was
collected by filtration. The diamidine was purified through
neutralization with 1 N NaOH followed by filtration of the resultant
solid and washing with water. Finally, the free base was stirred with
ethanol/HCl (gas) overnight, diluted with ether, and the solid
formed was filtered off and dried giving the diamidine hydrochlo-
ime 16. Yield % 75, mp 238e240 ꢁC. 1H NMR (D2O/DMSO‑d6):
d 1.79
(s, 2 x CH3), 4.02 (s, 4H), 6.07 (s, 2H), 7.39 (d, J ¼ 8.1 Hz, 4H), 7.78 (d,
J ¼ 8.1 Hz, 4H). 13C NMR (DMSO‑d6) (free base):
d 165.3, 152.4, 144.7,
128.9, 128.3, 126.0, 107.7, 33.4. MS (ESI) m/e (rel. int.): 333 (Mþþ1,
100). HRMS Calc. for C20H21N4O: 333.1715. Observed: 333.1717.
Anal. Calc. for C20H20N4O-2.0CH3CO2H-0.5H2O-0.75EtOH: C, 61.74;
H, 6.79; N, 11.29. Found C, 61.63; H, 6.66; N, 10.51.
4.1.23. 4-((5-Bromopyridin-2-yl)oxy)benzonitrile (19)
A solution of 4-hydroxybenzonitrile (1.43 gm, 12 mmol) in dry
DMF (10 mL) was treated with sodium hydride (0.29 gm, 10 mmol)
at r. t. and the reaction was stirred for 30 min. Then, a solution of
2,5-dibromopyridine (2.36 gm, 10 mmol) in dry DMF (5 mL) was
added portion-wise to the above reaction mixture. The reaction was
heated at 80 ꢁC for 48 h. The reaction mixture was cooled, poured
onto ice/water and the precipitated solid was filtered, washed with
water, dried and recrystallized from ethanol providing bromo
compound 19. Yield % 57, mp 110e112 ꢁC; 1H NMR (DMSO‑d6):
ride salt 23. Yield % 71, mp 201e203 ꢁC. 1H NMR (DMSO‑d6):
d 7.30
(d, J ¼ 8.4 Hz, 1H), 7.42 (d, J ¼ 8.7 Hz, 2H), 7.90e7.96 (m, 6H), 8.35
(dd, J ¼ 2.4 Hz, 8.4 Hz, 1H), 8.60 (d, J ¼ 2.4 Hz, 1H), 9.23 (br. s, 4H),
9.44 (br. s, 4H). 13C NMR (DMSO‑d6):
d 165.0, 164.9, 162.4, 158.1,
145.8, 141.5, 139.2, 130.2, 128.9, 127.1, 126.9, 124.1, 121.6, 112.3.
HRMS Calc. for C19H18N5O MHþ 332.1511. Obsd: 332.1523. Anal.
Calcd. for C19H17N5O-2.0HCl-1.5H2O-0.25C2H5OH: C, 52.88; H, 5.34;
N, 15.81. Found: C, 53.00; H, 5.23; N, 15.57.
d
7.16 (1H, J ¼ 8.7 Hz, 1H), 7.34 (d, J ¼ 8.7 Hz, 2H), 7.88 (d, J ¼ 8.7 Hz,
2H), 8.12 (dd, J ¼ 2.4, 8.7 Hz, 1H), 8.31 (d, J ¼ 2.4 Hz, 1H); 13C NMR
4.1.28. 3-((6-(4-Cyanophenoxy)pyridin-2-yl)oxy)benzonitrile (24)
A solution of p-hydroxybenzonitrile (2.86 gm, 24 mmol) in dry
DMSO was treated with KeO-tert-butoxide (2.69 gm, 24 mmol) at r.
t. and the reaction was kept stirring for 1 h. Thus after, 2,6-
(DMSO‑d6):
d 160.9, 157.4, 147.9, 143.0, 134.3, 121.9, 118.6, 114.6,
114.5, 107.2. MS (ESI) m/e (rel. int.): 277 (Mþ þ 2, 11), 276 (Mþ þ1,
94), 275 (Mþ, 79), 167 (100).
12

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
dibromopyridine (2.36 gm, 10 mmol) was added and the reaction
mixture was heated at 120 ꢁC for 48 h. The reaction solution was
poured onto ice/water and the precipitated solid was filtered off,
washed with water and dried. The pure cyano compound 24 was
obtained by column chromatography (EtOAc:hexanes, 60:40). Yield
adopting the exact conditions except for replacing MES 10 buffer
with TRIS. HCl (50 mM Tris-HCl, 1 mM EDTA, 100 mM NaCl; pH 8.5).
Another experiment was conducted on poly (dG-dC)n DNA in
MES 10 buffer employing a representative example from each se-
ries following the same experimental protocol to test the selectivity
of this class compounds to poly (dA-dT)n.
% 61, mp 179e180 ꢁC. 1H NMR (DMSO‑d6):
d
6.91 (d, J ¼ 8.1 Hz, 2H),
7.28 (d, J ¼ 9 Hz, 4H), 7.81 (d, J ¼ 9 Hz, 4H), 8.00 (t, J ¼ 8.1 Hz,1H). 13
C
NMR (DMSO‑d6):
d
160.1, 157.1, 144.3, 134.0, 121.5, 118.5, 107, 106.9.
4.3. Circular dichroism (CD)
4.1.29. N0-hydroxy-3-((6-(4-(-N0-hydroxyamidino)phenoxy)
pyridin-2-yl)oxy)benzamidine (25)
CD spectra were collected employing a Jasco J-810 spectrometer.
CD experiments were performed in a 1 cm quartz cell at 25 ꢁC. The
The amidoxime 25 was prepared according to the procedure
described for diamidoxime 4a employing the dinitrile derivative 24.
Poly (dA). Poly (dT) DNA (20 mM) sequence in MES 10 buffer (10 mM
MES, 100 mM NaCl, 1 mM EDTA; pH 6.2) was added to the cell prior
to the experiment and then the compound was added to the DNA
solution and incubated for 10 min to achieve the equilibrium
binding for the DNA complex. For each titration point four spectra
were averaged from 500 to 220 nm with scan speed 50 nm/min,
and a response time of 1 s. Buffer subtracted graphs were created
using the Kaleidagraph software.
Yield % quantitative, mp 165e167 ꢁC. 1H NMR (DMSO‑d6):
d 5.76 (br.
s, 4H), 6.68 (d, J ¼ 8.1 Hz, 2H), 7.12 (d, J ¼ 9 Hz, 4H), 7.64 (d, J ¼ 9 Hz,
4H), 7.83 (t, J ¼ 8.1 Hz, 1H), 9.61 (br. s, 2H). MS (ESI) m/e (rel. int.):
379 (Mþ, 4), 364 (42).
4.1.30. N0-Methoxy-3-((6-(4-(N0-methoxyamidino)phenoxy)
pyridin-2-yl)oxy)benzamidine hydrochloride salt (26)
The reaction was carried out utilizing the procedure described
for 12a and employing the amidoxime 25 (0.45 gm, 1.18 mmol) and
methlyiodide (0.22 mL, 3.55 mmol). Free base of 26, yield % 53, mp
4.4. Molecular dynamics
4.4.1. Molecular dynamics simulations
143e145 ꢁC. 1H NMR (DMSO‑d6):
d
3.72 (s, 2XCH3), 6.02 (br. s, 4H),
DNAecompound 6c complex was subjected to a 50 ns molecular
dynamics (MD) simulation to estimate the stability of the complex
structure and gain insights about the ligand binding to the minor
groove of the DNA by means of Gromacs Molecular Dynamics
package (version 5.1x) [28,29]. The MD simulations were started
from the X-ray crystal structure (PDB ID: 1VZK) of the DNA
dodecamer d (CGCGAATTCGCG)2 in complex with the minor groove
binder 2-[5-(4-carbamimidoylphenyl)thiophen-2-yl]-3H-benzimid-
azole-5-carboximidamide [27]. Compound 6c was built using MOE
version 2019.01, and energy-minimized employing the MMFF94x
forcefield. Then it was docked into the pre-processing of the crystal
structure of the complex. Placement of compound 6c was carried out
by means of the Triangle Matcher method, and 30 poses were
initially retained and ranked according to the London dG scoring
function, energy-minimized, then re-scored by the GBVI/WSA dG
scoring function. Solvent molecules were removed before the
docking. The best scoring pose was then selected for the MD simu-
lation. Compound 6c was parameterized according to the General
AMBER force field (GAFF) using the LEaP function implemented in
the Antechamber of Amber 18. The MD simulations were carried out
by the Gromacs Molecular Dynamics package (version 5.1x) using
the AMBER-OL15 forcefield and the TIP3P water model [28,29]. The
DNA-compound 6c complex was solvated in a dodecahedron water
boxþof SPC216 extending at least 15 Å from the edge of the complex.
Na counter ions were used to neutralize the negatively-charged
system. The system was then energy-minimized and equilibrated
by restraining the ligand then applying NVT and NPT for 100 ps using
the Verlet algorithm. Also, the ParrinelloeRahman barostat coupling
was used to maintain a pressure of 1 bar and a coupling constant of
2 ps? Berendsen coupling was employed to keep a constant tem-
perature of 300 K and a coupling constant of 0.1 ps [30]. The MD
simulation was completed for 50 ns with a time-step of 2 fs applying
periodic boundary conditions. Short-range van der Waals and elec-
trostatic interactions were measured at a cutoff of 12 Å, and long-
range electrostatics were calculated using Particle Mesh Ewald
Summation [31]. The LINCS algorithm was used for constraining H-
bonds [32].
6.68 (d, J ¼ 8.1 Hz, 2H), 7.11 (d, J ¼ 8.7 Hz, 4H), 7.64 (d, J ¼ 8.7 Hz,
4H), 7.88 (t, J ¼ 8.1 Hz,1H). 13C NMR (DMSO‑d6):
d 161.4,154.5,150.5,
143.6, 128.9, 127.2, 120.2, 105.5, 60.5. Hydrochloride salt of 26 was
prepared as previously described for methoxyamidine salt 5. Yield
% 94, mp 216e218 ꢁC. 1H NMR (D2O/DMSO‑d6):
d 3.83 (s, 2xCH3), 4.4
(br. s, 4H, NH2), 6.87 (d, J ¼ 8.1 Hz, 2H), 7.29 (d, J ¼ 8.7 Hz, 4H), 7.82
(d, J ¼ 8.7 Hz, 4H), 8.00 (t, J ¼ 8.1 Hz, 1H), 8.75 (br. s, 2H, HCl salt).
HRMS Calc. for C21H22N5O4 MHþ 408.1672. Obsd: 408.1659. Anal.
Calcd. for C21H21N5O4-2.0HCl-0.25H2O-0.5C2H5OH: C, 52.02; H,
5.25; N, 13.78. Found: C, 51.90; H, 5.23; N, 13.69.
4.1.31. 3-((6-(4-Amidinophenoxy)pyridin-2-yl)oxy)benzamidine
hydrochloride salt (27)
Synthesis of 27 was performed adopting the method described
for diamidine salt 23. Yield % 72, mp > 300 ꢁC. 1H NMR (DMSO‑d6):
d
6.92 (d, J ¼ 8.1 Hz, 2H), 7.36 (d, J ¼ 9 Hz, 4H), 7.92 (d, J ¼ 9 Hz, 4H),
8.03 (t, J ¼ 8.1 Hz, 1H), 9.17 (br. s, 4H), 9.43 (br. s, 4H). 13C NMR
(DMSO‑d6):
d 164.6, 160.3, 158.1, 144.4, 130.2, 123.4, 120.4, 107.4.
HRMS Calc. for C19H18N5O2 MHþ 348.1461. Obsd: 348.1470. Anal.
Calcd. for C19H17N5O2-2.0HCl: C, 54.29; H, 4.55; N, 16.66. Found: C,
54.32; H, 4.59; N, 16.47.
4.2. Absorbance spectroscopy and thermal melting (
experiments
DTm)
Absorbance and thermal melting (Tm) experiments were con-
ducted using a Cary 300 Bio spectrophotometer, and the software
supplied with the instrument. Experiments were performed in MES
10 buffer (10 mM MES, 1 mM EDTA, and 100 mM NaCl; pH 6.2). For
absorbance recording, 1 cm quartz cuvettes containing the buffer
were initially scanned from 400 to 250 nm, and then aliquots of
concentrated stock solutions of the test compounds were titrated
into the buffer and rescanned. Cuvettes were mounted in a thermal
block, and the solution temperatures were monitored by a therm-
istor in the reference cuvette. Temperatures were maintained by a
computer-controlled heating rate of 0.5 ꢁC/min. Experiments were
performed using a concentration of 2 ꢀ 10 ꢂ5 M base pair for poly
(dA-dT)n using DNA polymers-purchased from Pharmacia and
characterized by their melting curves. In experiments with com-
plexes, 0.3 compound per base pair ratio for poly (dA-dT)n was
used. Another experimental run was performed only for 6c and 27
4.4.2. Trajectory post-processing and analysis
The trjconv function of Gromacs was used for recentering and
rewrapping of the molecules within the unit cells of the obtained
trajectory. The radius of gyration, root-mean-square deviation
13

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
(RMSD), and root mean square fluctuation (RMSF) of the DNA with
reference to its initial position as per the X-ray crystallography
coordinates were calculated for at 10-ps intervals to check the DNA
integrity. The RMSD of the ligand heavy atoms was calculated with
reference to the DNA. The number of hydrogen bonds between the
diamidine 6c and the DNA was analyzed utilizing the hbond func-
tion in Gromacs. For analyzing the interactions between compound
6c and the DNA, the trajectory frames from the 40th ns to the 50th
ns were clustered using an RMSD cutoff of 0.2 nm for two structures
to be neighbors, and the hydrogen bonds found in the clustered
structures were investigated using distance calculations in VMD
[33].
4.6. In vivo T. B. r. STIB 900 mouse model
Groups of four mice were injected i. p. with 2 ꢀ 105 bloodstream
forms of T. b. r. STIB 900. On days 3, 4, 5, and 6 post-infection, the
experimental groups were treated with the test agents either i. p.
for the amidines or p. o. for their derivative prodrugs [36]. A pre-
toxicological experiment was used to determine the highest toler-
ated dose, which was subsequently administrated to the mice.
Parasitemia of the mice was investigated daily up to 14 days post-
infection, then twice per week up to 60 days. One mice group
was not treated and acted as a control. For relapsing mice, the day
of death was documented and the survival time was determined
[36].
4.4.3. MM/PBSA calculations
The binding free energy between for the DNA-diamidine 6c
complex was determined employing the Molecular Mechanics
PoissoneBoltzmann Surface Area (MM-PBSA) method using the
Gromacs compatible tool g_mmpbsa [33]. Different energy com-
ponents were calculated for the 50 ps interval 1001 snapshots
obtained over the 50 ns of the MD simulation. However, only the
Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
average of 201 snapshots was taken into account in calculating DG
Acknowledgement
starting from the 40th to 50th ns accounting for the convergence of
trajectory of the MD simulations to an equilibrium phase. The
binding free energy was calculated as earlier described [33].
This work was supported by an award from the Bill and Melinda
Gates Foundation (Award No. 38920) through the Consortium for
Parasitic Drug Development (R.B., W.D.W., D.W.B.) and by the US
National Institutes of Health Grant GM111749 (W.D.W. and D.W.B.).
4.5. In vitro biological assays
4.5.1. In vitro assay for T. b. r
References
Following the established protocol of Baltz et al. minimum
[1] P. Büscher, G. Cecchi, V. Jamonneau, G. Priotto, Human african trypanosomi-
asis, Lancet 390 (2017) 2397e2409.
[2] P.G.E. Kennedy, Update on human African trypanosomiasis (sleeping sick-
ness), J. Neurol. 266 (2019) 2334e2337.
[3] E. Ashley, A. Pyae Phyo, WoodrowC, J. Malaria. Lancet 391 (2018) 1608e1621.
[4] WHO, World Malaria Report, 2018.
[5] M. Soeiro, K. Werbovetz, D. Boykin, W. Wilson, M. Wang, A. Hemphill, Novel
amidines and analogues as promising agents against intracellular parasites: a
systematic review, Parasitology 140 (2013).
essential medium (MEM) (50 mL) was supplemented to each well of
a 96-well microtiter plate and 2-mercaptoethanol and 15% heat-
inactivated horse serum were added [34]. Then, seven 3-fold se-
rial dilutions of the test compounds covering a range from 90 to
0.123 m mL of T. b. r. STIB 900
g mLꢂ1 were added to the wells. 50
suspension was added to each well, and the plate was incubated at
37 ꢁC under a 5% CO2 atmosphere for 72 h. Subsequently, Alamar
[6] A.R. Chowdhury, R. Bakshi, J. Wang, G. Yildirir, B. Liu, V. Pappas-Brown,
G. Tolun, J.D. Griffith, T.A. Shapiro, R.E. Jensen, The killing of African try-
panosomes by ethidium bromide, PLoS Pathog. 6 (2010), e1001226.
[7] W.D. Wilson, F.A. Tanious, A. Mathis, D. Tevis, J.E. Hall, D.W. Boykin, Anti-
parasitic compounds that target DNA, Biochimie 90 (2008) 999e1014.
[8] 20 J. Ashley, H. Barber, A. Ewins, G. Newbery, A. Self, A chemotherapeutic
comparison of the trypanocidal action of some aromatic diamidines, J. Chem.
Soc. (1942) 103e116.
Blue (10 mL) was added to each well and incubation lasted for an
extra 2e4 h [35]. The microplate fluorometer system, Spectramax
Gemini by Molecular Devices, was used for reading the microplate
with an excitation wavelength of 536 nm and an emission wave-
length of 588 nm. Fluorescence development was expressed as a
percentage of the control, and IC50 values calculated [36].
[9] B.P. Das, D.W. Boykin, Synthesis and antiprotozoal activity of 2,5-bis(4-
guanylphenyl)furans, J. Med. Chem. 20 (1977) 531e536.
4.5.2. In vitro assay for P. f
[10] R.R. Tidwell, S.K. Jones, J.D. Geratz, K.A. Ohemeng, C.A. Bell, B.J. Berger, J.E. Hall,
Development of pentamidine analogues as new agents for the treatment of
Pneumocystis carinii pneumonia, Ann. N. Y. Acad. Sci. 616 (1990) 421e441.
[11] C. Yeates, DB-289 (immtech international), Idrugs 6 (2003) 1086.
[12] P. Yeramian, S.R. Meshnick, S. Krudsood, K. Chalermrut, U. Silachamroon,
N. Tangpukdee, J. Allen, R. Brun, J.J. Kwiek, R. Tidwell, S. Looareesuwan, Effi-
cacy of DB289 in Thai patients with Plasmodium vivax or acute, uncompli-
cated Plasmodium falciparum infections, J. Infect. Dis. 192 (2005) 319e322.
[13] W.D. Wilson, H.P. De Koningꢁ, C.A. Olson, G. Pohlig, Diamidines for human
African trypanosomiasis, Curr. Opin. Invest. Drugs 11 (2010) 876e883.
[14] R. Brun, C. Burri, C.W. Gichuki, The story of CGP 40 215: studies on its efficacy
and pharmacokinetics in African green monkey infected with Trypanosoma
brucei rhodesiense, Trop. Med. Int. Health 6 (2001) 362e368.
[15] M. Munde, M.A. Ismail, R. Arafa, P. Peixoto, C.J. Collar, Y. Liu, L. Hu, M.H. David-
Cordonnier, A. Lansiaux, C. Bailly, D.W. Boykin, W.D. Wilson, Design of DNA
minor groove binding diamidines that recognize GC base pair sequences: a
dimeric-hinge interaction motif, J. Am. Chem. Soc. 129 (2007) 13732e13743.
[16] M.A. Ismail, R.K. Arafa, R. Brun, T. Wenzler, Y. Miao, W.D. Wilson, C. Generaux,
A. Bridges, J.E. Hall, D.W. Boykin, Synthesis, DNA affinity, and antiprotozoal
activity of linear dications: terphenyl diamidines and analogues, J. Med. Chem.
49 (2006) 5324e5332.
The antimalarial activity was measured utilizing the of P. f. K1
strain which is resistant to chloroquine and pyrimethamine. The
modified [3H]-hypoxanthine incorporation assay was adopted
where infected human red blood cells in RPMI 1640 medium,
25 mM HEPES, Na2CO3, and glutamine supplemented with 5%
AlbuMAX II, 2% D-sucrose, 0.3% glutamine, and 150 mM hypoxan-
thine. Cultures were maintained under an atmosphere of 5% O2, 5%
CO2, and 90% N2. Wells were subjected to serial dilutions of the test
agents in microtiter plates for 48 h (a stock of 10 mg in screening
medium was used to prepare seven serial dilutions of a factor 1:2
with 100 mL of medium RPMI 1640 without hypoxanthine). Mea-
surements were validated by recording the incorporation of [3H]-
hypoxanthine using liquid scintillation counting 24 h after adding
the radiolabel. The counts were expressed as a percentage of the
control untreated cultures, and sigmoidal inhibition curves were
plotted from which IC50 values were determined [37,38].
[17] D.A. Patrick, S.A. Bakunov, S.M. Bakunova, S.K. Jones, T. Wenzler, T. Barszcz,
A. Kumar, D.W. Boykin, K.A. Werbovetz, R. Brun, R.R. Tidwell, Synthesis and
antiprotozoal activities of benzyl phenyl ether diamidine derivatives, Eur. J.
Med. Chem. 67 (2013) 310e324.
[18] L. Hu, R.K. Arafa, M.A. Ismail, A. Patel, M. Munde, W.D. Wilson, T. Wenzler,
R. Brun, D.W. Boykin, Synthesis and activity of azaterphenyl diamidines
against Trypanosoma brucei rhodesiense and Plasmodium falciparum, Bioorg.
4.5.3. Cytotoxicity studies
Cytotoxicity was assessed using cultured L-6 rat myoblast cells
following the same in vitro assay procedure described for T. b. r.
[19,34,35].
14

R.K. Arafa, M.A. Ismail, T. Wenzler et al.
European Journal of Medicinal Chemistry 222 (2021) 113625
Med. Chem. 17 (2009) 6651e6658.
UCHL-3 inhibitors via virtual screening, molecular dynamics and binding
mode analysis, J. Biomol. Struct. Dyn. (2020) 1e14.
[19] R.K. Arafa, T. Wenzler, R. Brun, Y. Chai, W.D. Wilson, Molecular modeling
study and synthesis of novel dicationic flexible triaryl guanidines and imi-
damides as antiprotozoal agents, Eur. J. Med. Chem. 46 (2011) 5852e5860.
[20] E. De Souza, P. Da Silva, A. Nefertiti, M. Ismail, R. Arafa, B. Tao, C. Nixon-Smith,
D. Boykin, M. Soeiro, Trypanocidal activity and selectivity in vitro of aromatic
amidine compounds upon bloodstream and intracellular forms of Trypano-
soma cruzi, Exp. Parasitol. 127 (2011) 429e435.
[29] M.S. Nafie, K. Arafa, N.K. Sedky, A.A. Alakhdar, R.K. Arafa, Triaryl dicationic
DNA minor-groove binders with antioxidant activity display cytotoxicity and
induce apoptosis in breast cancer, Chem. Biol. Interact. 324 (2020) 109087.
[30] J. Wang, W. Wang, P.A. Kollman, D.A. Case, Antechamber: an accessory soft-
ware package for molecular mechanical calculations, J. Am. Chem. Soc. 222
(2001) U403.
[21] M.A. Ismail, M.M. Youssef, R.K. Arafa, S.S. Al-Shihry, W.M. El-Sayed, Synthesis
and antiproliferative activity of monocationic arylthiophene derivatives, Eur.
J. Med. Chem. 126 (2017) 789e798.
[31] B. Hess, C. Kutzner, D. van der Spoel, E. Lindahl, Gromacs 4: algorithms for
highly efficient, load-balanced, and scalable molecular simulation, J. Chem.
Theor. Comput. 4 (2008) 435e447.
[22] M.M. Youssef, R.K. Arafa, M.A. Ismail, Synthesis, antimicrobial, and anti-
proliferative activities of substituted phenylfuranylnicotinamidines, Drug Des.
Dev. Ther. 10 (2016) 1133.
[32] R. Kumari, R. Kumar, C. Open, Source Drug Discovery, A. Lynn, g_mmpbsa–a
GROMACS tool for high-throughput MM-PBSA calculations, J. Chem. Inf.
Model. 54 (2014) 1951e1962.
[23] C.N. Generaux, G.R. Ainslie, A.S. Bridges, M.A. Ismail, D.W. Boykin, R.R. Tidwell,
D.R. Thakker, M.F. Paine, Compartmental and enzyme kinetic modeling to
[33] A.A. Alakhdar, A.H. Saleh, R.K. Arafa, Targeting homologous recombination
(HR) repair mechanism for cancer treatment: discovery of new potential
UCHL-3 inhibitors via virtual screening, molecular dynamics and binding
mode analysis, J. Biomol. Struct. Dyn. (2020) 1e14.
elucidate the biotransformation pathway of
a centrally acting anti-
trypanosomal prodrug, Drug Metab. Dispos. 41 (2013) 518e528.
[24] R.A. Hunt, M. Munde, A. Kumar, M.A. Ismail, A.A. Farahat, R.K. Arafa, M. Say,
A. Batista-Parra, D. Tevis, D.W. Boykin, W.D. Wilson, Induced topological
changes in DNA complexes: influence of DNA sequences and small molecule
structures, Nucleic Acids Res. 39 (2011) 4265e4274.
[34] T. Baltz, D. Baltz, C. Giroud, J. Crockett, T. equiperdum, T. evansi,
T. rhodesiense, T. gambiense, Cultivation in a semi-defined medium of animal
infective forms of Trypanosoma brucei, EMBO J. 4 (1985) 1273e1277.
€
[35] B. Raz, M. Iten, Y. Grether-Bühler, R. Kaminsky, R. Brun, The Alamar Blue®
[25] S. Depauw, M.l. Lambert, S. Jambon, A. Paul, P. Peixoto, R. Nhili, L. Marongiu,
M. Figeac, C. Dassi, C. Paul-Constant, Heterocyclic diamidine DNA ligands as
HOXA9 transcription factor inhibitors: design, molecular evaluation, and
cellular consequences in a HOXA9-dependant leukemia cell model, J. Med.
Chem. 62 (2019) 1306e1329.
assay to determine drug sensitivity of African trypanosomes (Tb rhodesiense
and Tb gambiense) in vitro, Acta Trop. 68 (1997) 139e147.
[36] M.A. Ismail, R. Brun, J.D. Easterbrook, F.A. Tanious, W.D. Wilson, D.W. Boykin,
Synthesis and antiprotozoal activity of aza-analogues of furamidine, J. Med.
Chem. 46 (2003) 4761e4769.
[26] B. Norden, M. Kubista, T. Kurucsev, Linear dichroism spectroscopy of nucleic
acids, Q. Rev. Biophys. 25 (1992) 51e170.
[27] S. Mallena, M.P. Lee, C. Bailly, S. Neidle, A. Kumar, D.W. Boykin, W.D. Wilson,
Thiophene-based diamidine forms a “super” at binding minor groove agent,
J. Am. Chem. Soc. 126 (2004) 13659e13669.
[37] H. Matile, J. Richard, L. Pink, Plasmodium falciparum malaria parasite cultures
and their use in immunology, in: Immunological Methods, Elsevier, 1990,
pp. 221e234.
[38] M.A. Ismail, A. Batista-Parra, Y. Miao, W.D. Wilson, T. Wenzler, R. Brun,
D.W. Boykin, Dicationic near-linear biphenyl benzimidazole derivatives as
DNA-targeted antiprotozoal agents, Bioorg. Med. Chem. 13 (2005)
6718e6726.
[28] A.A. Alakhdar, A.H. Saleh, R.K. Arafa, Targeting homologous recombination
(HR) repair mechanism for cancer treatment: discovery of new potential
15