Antiprotozoal activity and DNA binding of N-substituted N-phenylbenzamide and 1,3-diphenylurea bisguanidines

Article abstract of DOI:10.1016/j.ejmech.2014.04.083

Two series of N-alkyl, N-alkoxy, and N-hydroxy bisguanidines derived from the N-phenylbenzamide and 1,3-diphenylurea scaffolds were synthesised in three steps from the corresponding 4-amino-N-(4-aminophenyl)benzamide and 1,3-bis(4-aminophenyl)urea, respec

Full text of DOI:10.1016/j.ejmech.2014.04.083

Accepted Manuscript
Antiprotozoal activity and DNA binding of N-Substituted N-phenylbenzamide and 1,3-
Diphenylurea Bisguanidines
Carlos H. Ríos Martínez, Laura Lagartera, Marcel Kaiser, Christophe Dardonville
PII:
S0223-5234(14)00412-7
10.1016/j.ejmech.2014.04.083
EJMECH 6954
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 7 March 2014
Revised Date: 29 April 2014
Accepted Date: 30 April 2014
Please cite this article as: C.H. Ríos Martínez, L. Lagartera, M. Kaiser, C. Dardonville, Antiprotozoal
activity and DNA binding of N-Substituted N-phenylbenzamide and 1,3-Diphenylurea Bisguanidines,
European Journal of Medicinal Chemistry (2014), doi: 10.1016/j.ejmech.2014.04.083.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. 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.

*Graphical Abstract (for review)
Click here to download Graphical Abstract (forAreCvCiewE)P: TGrEaDphiMcaAl aNbsUtrSacCt.RepIsPT
H
N
NH2
AT-rich DNA
O
R
N
N
N
R
H
H N
N
2
H
T. b. rhodesiense
9 nM (SI > 13000)
11 nM (SI > 9000)
P. falciparum
R = Et
232 nM (SI = 500)
i
R = Pr
4
46 nM (SI = 230)
4/4 cures at 20 mg/kg ip

*Highlights (for review)
ACCEPTED MANUSCRIPT
Highlights



N-alkoxy, N-hydroxy and N-alkyl bisguanidine derivatives were synthesized.
The compounds were tested against T. b. rhodesiense and P. falciparum.
N-alkylated compounds 7b and 7c were 100% curative in the STIB900 mouse
model of HAT.


DNA binding affinity was measured by SPR with AATT, (AT) and (CG)4
4
hairpin duplexes.
7b and 7c bind selectively to AT-rich DNA.

*Revised Manuscript
Click here to view linked References
ACCEPTED MANUSCRIPT
Short communication
Antiprotozoal activity and DNA binding of N-
Substituted N-phenylbenzamide and 1,3-
Diphenylurea Bisguanidines
a
a
b,c
Carlos H. Ríos Martínez , Laura Lagartera , Marcel Kaiser and Christophe
a*
Dardonville
a
Instituto de Química Médica, IQM–CSIC, Juan de la Cierva 3, E–28006 Madrid,
Spain.
b
Swiss Tropical and Public Health Institute, Socinstrasse, 57, CH-4002 Basel,
Switzerland.
c
University of Basel, Basel, Switzerland.
*
Corresponding author. Tel.: +34 912587490; Fax.: +34 915644853; e-mail:
dardonville@iqm.csic.es
‡
‡
Abbreviations: BBB, blood-brain barrier; CNS, central nervous system; DIPEA,
diisopropylethylamine; EDAC, 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide; HAT,
human African trypanosomiasis; MW, microwave; NECT, nifurtimox-eflornithine
combination therapy; SAR, structure-activity relationships; SI, selectivity index; SPR,
surface plasmon resonance; TMSCl, trimethylsilylchloride.
1

ACCEPTED MANUSCRIPT
Abstract
Two series of N-alkyl, N-alkoxy, and N-hydroxy bisguanidines derived from the N-
phenylbenzamide and 1,3-diphenylurea scaffolds were synthesised in three steps from
the
corresponding
4-amino-N-(4-aminophenyl)benzamide
and
1,3-bis(4-
aminophenyl)urea, respectively. All of the new compounds were evaluated in vitro
against T. b. rhodesiense (STIB900) trypomastigotes and P. falciparum NF54 parasites
(erythrocytic stage). N-alkoxy and N-hydroxy derivatives showed weak micromolar
range IC50 values against T. b. rhodesiense and P. falciparum whereas the N-alkyl
analogues displayed submicromolar and low nanomolar IC50 values against P.
falciparum and T. brucei, respectively. Two compounds, 4-(2-ethylguanidino)-N-(4-(2-
ethylguanidino)phenyl)benzamide dihydrochloride (7b) and 4-(2-isopropylguanidino)-
N-(4-(2-isopropylguanidino)phenyl)benzamide dihydrochloride (7c), which showed
favourable drug-like properties and in vivo efficacy (100% cures) in the STIB900
mouse model of acute human African trypanosomiasis represent interesting leads for
further in vivo studies. The binding of these compounds to AT-rich DNA was
confirmed by biosensor-SPR experiments.
2

ACCEPTED MANUSCRIPT
Highlights



N-alkoxy, N-hydroxy and N-alkyl bisguanidine derivatives were synthesized.
The compounds were tested against T. b. rhodesiense and P. falciparum.
N-alkylated compounds 7b and 7c were 100% curative in the STIB900 mouse
model of HAT.


DNA binding affinity was measured by SPR with AATT, (AT) and (CG)4
4
hairpin duplexes.
7b and 7c bind selectively to AT-rich DNA.
Keywords: Trypanosoma brucei, Plasmodium falciparum, parasite chemotherapy,
guanidine, minor groove binder, biosensor-SPR.
3

ACCEPTED MANUSCRIPT
1
. Introduction
Human African trypanosomiasis (HAT) and malaria are parasitic diseases that are
endemic in sub-Saharian Africa where they cause great morbidity and high mortality
rates (especially for malaria). Two subspecies of trypanosomes (Trypanosoma brucei
gambiense and T. b. rhodesiense) lead to chronic and acute forms of HAT, respectively
[
1]. The apicomplexan parasite Plasmodium falciparum, which is the cause of the most
severe and deadly form of malaria, is mainly prelavent in Africa being responsible of
over 75% of cases in that continent [2].
New drugs are required to complete and/or improve the chemotherapeutic arsenal
available for both diseases. On the one hand, drugs for the treatment of HAT are scarce
(i.e. pentamidine, suramin, melarsoprol, eflornithine and nifurtimox-eflornithine
combination therapy – NECT –), stage and species specific, very toxic, and not orally
active [3, 4]. On the other hand, malaria chemotherapy is suffering from extended
resistance to conventional drugs (e.g. chloroquine) and the rapid development of
Plasmodium strains that are resistant to “newly” marketed drugs (e.g. artemisinin) [5,
6
]. In a word, new drugs are urgently needed to fight against these deadly diseases.
Dicationic compounds such as diamidines [7-10] and bis(2-aminoimidazolines) [11-13]
have a track record of success as antiprotozoan agents. These compounds bind strongly
to the DNA minor groove at AT-rich sites which is one of the known targets of these
molecules. Accordingly dicationic diamidines were shown to be localized to the DNA-
containing nucleus and/or kinetoplast of trypanosomes [14]. Bisguanidine related
molecules are also very active against T. brucei and P. falciparum [11-13, 15-19].
However, low nanomolar in vitro activities do not always result in potent in vivo
activity in murine models of HAT and malaria [16, 17]. Besides, this class of dibasic
4

ACCEPTED MANUSCRIPT
compounds is highly polar (i.e. dicationic at physiological pH) and, like pentamidine,
mostly fails in curing CNS-stage T. brucei infections.
In recent years, we have discovered a set of bisguanidinium hits that displayed excellent
activities in vitro against T. b. rhodesiense and P. falciparum [11]. Two hit compounds
derived from the N-phenylbenzamide scaffold (I) and the 1,3-diphenylurea scaffold (II)
that presented nanomolar IC s, adequate selectivities, and promising in vivo activity in
5
0
the STIB900 mouse model of acute HAT were chosen as templates for further
development (Figure 1). The choice of these scaffolds as antitrypanosomal templates
was also supported by the excellent results obtained with bis(2-aminoimidazoline)
analogues that proved to be 100% curative in vivo in the STIB900 mouse model [11].
Figure 1. Structure, In Vitro Activity and Selectivity Index of the N-Phenylbenzamide
(I) and 1,3-Diphenylurea (II) Bisguanidinium Leads [11]
In previous work, Arafa et al [15] showed that N-alkyl derivatives of diguanidino fused
ring systems had improved in vivo potency in the STIB900 mouse model of acute HAT
5

ACCEPTED MANUSCRIPT
compared with the unsubstituted diguanidino parent compounds. In the present study,
we tested a similar strategy with the hits I and II to check if the antiprotozoan activity
of these hits could be enhanced in that way. In fact, the addition of N-alkyl and N-
alkoxy substituents in the structure should increase the lipophilicity of the compounds
whereas N-alkoxy substituents should decrease their basicity with respect to I and II.
All of these modifications were expected to enhance the antiprotozoan activity of the
compounds as well as to improve their membrane permeation (e.g. blood-brain barrier).
Hence, six N-alkyl (7a-c, 8a-c), four N-alkoxy (7d-g), and one hydroxy derivatives of
the hits I and II were synthesised and tested against T. b. rhodesiense and P.
falciparum. The most active compounds in vitro were further evaluated in vivo in the
STIB900 mouse model of acute HAT. This allowed drawing preliminary structure-
activity relationships (SAR) for antimalarial and antitrypanosomal activity. Besides, the
DNA binding affinity of the compounds was assessed by biosensor–surface plasmon
resonance (SPR) experiments with three different hairpin oligonucleotides.
2
. Results
2
.1. Chemistry
N-substituted guanidines 7a-g and 8a-c were synthesized in three steps from diamines 1
and 2, respectively, following the Manimala and Anslyns’s methodology (Scheme 1)
[
20]. The synthesis started with the preparation of the bis(ethoxycarbonylthioureas 3
and 4 from 4-amino-N-(4-aminophenyl)benzamide and 1,3-bis(4-aminophenyl)urea,
respectively. These thioureas were reacted with an excess (4 equiv.) of N-alkyl (R = Me,
i
Et, Pr) and N-alkoxy amines (R = OMe, OEt, OBn, OTHP) in the presence of 1-ethyl-
3
-(3-dimethylaminopropyl)carbodiimide (EDAC) and N,N-diisopropylethylamine
6

ACCEPTED MANUSCRIPT
(DIPEA) to give the corresponding N-substituted bis(ethoxycarbonyl) guanidines 5a-g
and 6a-c. Two strategies were tried to remove the ethoxycarbonyl groups to obtain the
target guanidines 7a-h and 8a-c.
a
Scheme 1. Synthesis of N-alkyl and N-alkoxy guanidines
a
Reagents and conditions. (i) Ethoxycarbonyl isothiocyanate, CH Cl , rt; (ii) R–NH ,
2
2
2
CH Cl , DIPEA, EDAC, rt; (iii) KOH, MeOH, 80 ºC; (iv) HCl –saturated 1,4-dioxane
2
2
g
solution.* Yields for step (iii).
The first strategy using Lewis acids such as trimethylsilyl (TMSCl) chloride [21, 22]
and bromide (TMSBr) [20] to generate directly the guanidinium salts was mostly
unsuccessful in our hands, leading to complex mixtures of partially and fully
7

ACCEPTED MANUSCRIPT
deprotected products together with 2-alkylamino(3H)quinazolin-4-one by-products as
previously reported [22]. Apparently, the success of this deprotection protocol is very
dependent on the R group present on the guanidine moiety [e.g. compound 5c with R =
i
Pr was cleanly deprotected using TMSCl (10 equiv) / DMF / 80 ºC / MW / 60 min] and
was not applicable to most of our compounds. Optimal conditions for the removal of the
ethoxycarbonyl groups took place under basic conditions (0.1 M aqueous KOH in
MeOH at 80 ºC in a sealed tube) to yield 7a-g and 8a-c as free bases. Hydrochloride
salts of 7a-f, 8a-c, and 7h were obtained from the free base guanidines 7a-f, 8a-c, and
7
g with HCl –saturated 1,4-dioxane solution, respectively.
gas
2
.2. In Vitro and In Vivo Antiprotozoal Activity
The target compounds 7a-g and 8a-c were tested in vitro against T. brucei rhodesiense
STIB900 strain) by means of an Alamar blue-based assay [23]. The antimalarial
(
activity against erythrocytic forms of P. falciparum was evaluated with a
3
[
H]hypoxanthine incorporation assay [24] using the chloroquine-sensitive strain NF54.
The cytotoxicity of the compounds against mammalian L6-cells was also evaluated to
determine their selectivity index. The results are shown in Table 1.
All of the guanidines with N-alkyl substituents (7a-c and 8a-c) displayed
submicromolar range IC values against T. b. rhodesiense. Compounds with ethyl (7b,
5
0
8
b) and isopropyl groups (7c, 8c) were particularly effective with low nanomolar IC50
values, in the range of the reference drug melarsorprol (IC50 = 8 nM), and high
selectivities (SI > 1000). These values represent a 2- to 3-fold increase in
antitrypanosomal activity compared with the unsubstituted parent compounds II and I,
respectively. In contrast, N-hydroxy (7h) and N-alkoxy (7d, 7g) guanidines showed
8

ACCEPTED MANUSCRIPT
weak activity. As far as the linker of the diphenyl scaffold is concerned, the amide gave
5
- to 10-fold higher potencies than the urea linker. In general, most of these guanidines
presented low cytotoxicity (IC50 > 100 µM against L6-cells) even though some
differences could be observed between both series. Compound I derivatives (7a-d and
7
g-h) in particular were approximately 10-times less cytotoxic than the parent guanidine
I. On the contrary, derivatives 8a-c were somewhat more cytotoxic than II.
As with the anti-T. brucei activity, the best antimalarial compounds were the N-alkyl
derivatives 7b-c and 8a-c that displayed submicromolar IC values with selectivities
5
0
for the parasite > 200. Again, N-hydroxy and N-alkoxy derivatives were less potent with
IC s in the low micromolar range.
5
0
To evaluate the in vivo potential of these new N-alkyl guanidines, the best trypanocidal
compounds in vitro (7b, 7c, 8b, and 8c) were tested in a mouse model of acute (stage 1)
HAT. Groups of four mice infected with T. b. rhodesiense (strain STIB900) received a
four days treatment (days 3, 4, 5 and 6 post-infection) with the tested compounds (4×20
mg/kg ip). One control group (3 mice) remained untreated. The blood parasitemia was
checked periodically over the 60 days of the experiment and the mean day of relapse of
parasitemia was calculated. The mice that survived 60 days free of parasites were
considered as cured (Table 1).
The 1,3-diphenylurea analogues (8b, 8c) were poorly active in vivo and showed acute
toxicity at the dose tested. The lead compound I showed high efficacy in this model,
giving 4/4 cures by intraperitoneal (4×5 mg/kg ip) and oral administration (4×50 mg/kg
po). Unfortunately, it was inactive in the GVR35 model of late-stage disease (i.e. 5×20
mg/kg ip, data not shown), which is probably due to poor BBB permeation. The N-
phenylbenzamide derivatives 7b and 7c were 100% curative in the STIB900 model
9

ACCEPTED MANUSCRIPT
showing that the introduction of N-ethyl and N-isopropyl substituents is well tolerated
in vivo. Altogether, the results indicate that compounds 7b and 7c are promising
candidates for further in vivo studies (e.g. CNS stage).
Table 1. In Vitro and In Vivo Antitrypanosomal and Antimalarial Activity
a
b
c
STIB900 mouse model
Cmpd
R
T. b. rhodesiense P. falciparum Cytotoxicity
d
of acute HAT
Mean
e
Dosis
mg/kg)
IC (Selectivity index) µM
CC µM
Cured/infected day off
5
0
50
(
relapse
g
h
g
I
H
0.036 (319)
ND
11.5
4×5 ip
4/4
4/4
ND
>60
>60
ND
4
×50 po
g
g
II
H
Me
Et
0.187 (>1256)
ND
> 235
7
a
0.177 (977)
0.009 (13111)
0.011 (9364)
56.0
1.02 (169)
0.232 (509)
0.446 (231)
7.49
173
118
103
141
6.6
160
100
177
116
126
ND
4/4
4/4
ND
ND
ND
ND
ND
ND
>60
>60
ND
ND
ND
ND
ND
18.3
10
7
b
4×20 ip
4×20 ip
i
7
c
Pr
7
d
OMe
OBn
OTHP
OH
7
f
3.13
51.4
4.88
2.83
42.6
3.58
7
g
7
h
8
a
Me
0.264 (670)
0.090 (1289)
0.049 (2571)
0.192 (922)
0.397 (292)
0.437 (288)
i
8
b
Et
4×20 ip
4×20 ip
0/3
i
j
8
c
Pr
0/2
IC values reported are the average of two independent assays and vary less than ±
5
0
a
5
0
0%. T. b. rhodesiense STIB900 trypomastigotes; reference drug: melarsoprol, IC =
50
b
.008 µM. P. falciparum NF54, intraerythrocytic stage: this strain is susceptible to
c
chloroquine, IC50 = 0.006 µM. Rat skeletal myoblast L6-cells; reference drug:
d
podophyllotoxin, IC = 0.014 µM. Compounds were administrated intraperitoneally
5
0
e
f
at 20 mg/kg/day (4 days). SI = [CC /IC (parasite)]. Untreated control mice were all
50
50
g
positive at day 7. Data previously reported in reference [11] and shown here for
1
0

ACCEPTED MANUSCRIPT
h
i
j
comparative purpose. Not determined. One mouse died after first application. Two
th
mice died after 4 application.
2
.3. DNA Binding Affinity. Biosensor–SPR Experiments
Since N-phenylbenzamide and 1,3-diphenylurea guanidine and imidazoline derivatives
are strong DNA minor groove binders [11, 25-27], we were interested in quantitatively
evaluating the binding affinity of the new compounds with DNA. Binding to DNA and
resultant inhibition of transcription and activity of other DNA-dependent enzymes is
thought to contribute to the antiprotozoal activity of dicationic compounds [28]. This
was done by biosensor–SPR experiments with different DNA hairpin duplexes [i.e.,
AATT, (AT) , (CG) ] immobilized on a chip biosensor surface [29, 30]. All of the
4
4
compounds were first screened for binding to the three oligonucleotide sequences at a
single concentration (25 µM). This allowed the ranking of the compounds as
binders/non binders (data not shown). Then, the binding constants of the most
interesting compounds with the three different DNA hairpins were determined (Table
2
4
). Compounds 7b and 7c bound selectively to AT-rich DNAs (K values in the range
5
-1
.0 – 9.8 × 10 M ) with a slight preference for the AATT (approximately 1.5 – 2-fold)
over the (AT) sequence. The stoichiometry of binding was 2 for both AT-sequences
4
(Figure 2). These K values were approximately 7 to 9-fold lower (for AATT) and 3 to 4-
fold lower (for (AT) ) than the values reported for the unsubstituted bisguanidine I [25].
4
Hence, in this series, the alkylation of the guanidine moieties resulted in a decrease in
binding affinity to AT sequences. Besides, no important differences in binding were
i
observed between Pr and Et groups. These findings agree with the data reported earlier
with a series of linear triaryl bisguanidines [17].
1
1

ACCEPTED MANUSCRIPT
On the one hand, none of the compounds tested bound significantly to the (CG)4
sequence (Figure 2B) suggesting a specific AT-rich minor groove mode of binding for
7
b and 7c. On the other hand, binding of the urea derivative 8b was approximately 30-
5
-1
times weaker than that of 7b and 7c (K < 1× 10 M ) suggesting unspecific binding to
DNA.
Table 2. DNA Binding Constants Determined by SPR
5
-1
K (×10 M )
a
b
c
Compound
AATT
(AT)4
(CG)4
d
d
d
I
71.0
18.0
<1
e
e
g
7
b
8.1
5.2
4.0
<1
e
e
g
7
c
9.8
<1
f
e
h
8
b
0.19
0.15
a
5
’-biotin labelled DNA sequences used (the hairpin loop is underlined): 5’-biotin-
b
CGAATTCGTCTCCGAATTCG-3’.
5’-biotin-CATATATATCCCCATATATATG-
c
d
e
3
’. 5’-biotin-CGCGCGCGTTTTCGCGCGCG-3’. Taken from reference [25].
f
Primary binding constant for fitting to a two-site binding model. Binding constant for
g
h
fitting to a one-site binding model. Non-specific binding. There is not enough signal
to noise ratio to get a binding constant for this hairpin oligonucleotide.
1
2

ACCEPTED MANUSCRIPT
7b_AATT
7c_AATT
6
0
60
40
20
0
40
20
0
0
100
200
Time (second)
300
400
0
100
200
300
400
Time (second)
(A)
7b_(AT)4
7c_(AT)4
6
0
60
40
20
0
40
20
0
0
100
200
Time (second)
300
400
0
100
200
300
400
Time (second)
Compound 7c
Compound 7b
2
1
1
0
0
.0
.5
.0
.5
.0
3
2
1
0
AATT
AATT
(AT)4
(
AT)4
(CG)4
(CG)4
210-05
410-05
610-05
810-05
110-04
0
210-05 410-05 610-05 810-05 110-04
0
[
ligand]
[ligand]
(B)
Compound 8b
2
1
1
0
0
.0
(AT)4
.5
.0
.5
.0
AATT
0
210-05
410-05
610-05
810-05
[
ligand]
1
3

ACCEPTED MANUSCRIPT
Figure 2. (A) Sensorgrams for the interaction of 7b (left panel) and 7c (right panel)
with the AATT and (AT) hairpin oligonucleotides using increasing concentrations of
4
-
8
-5
ligand (from bottom to top) in the range 5×10 M to 5.76×10 M [7b_AATT and
-7
-5
7
c_(AT) ] and 2.5×10 M to 8×10 M [7b_(AT) and 7c_AATT]. (B) Binding curves
4 4
of 7b, 7c, and 8b to AATT (●), (AT) (■), and (CG) (▲) DNA hairpins. The SPR
4
4
response (RU) from the steady-state region of the sensorgrams was converted to r
moles of bound compound per mole of DNA hairpin duplex; r = RU / RUmax) and was
(
plotted against the unbound compound concentration ([ligand] = flow solution). The
data were fitted to a two-site binding model (except 8b_AATT which was fitted to a
one-site binding model) using equation (1) (see experimental section).
3
. Discussion
Two series of N-substituted bisguanidines derived from the N-phenylbenzamide and the
,3-diphenylurea scaffolds (7a-h and 8a-c, respectively) were synthesized and evaluated
1
as antitrypanosomal and antimalarial agents. The aim of this research was to improve
the antiprotozoal activity of the leads I and II by introducing N-alkyl and N-alkoxy
substituents on the guanidine moieties. As far as T. brucei is concerned, this was
i
achieved successfully with the introduction of Et and Pr substituents to get low
nanomolar range IC s (7b and 7c, respectively). A similar effect, although less
5
0
pronounced, was obtained with the 1,3-diphenylurea derivatives 8b and 8c.
Interestingly, the trypanocidal activity enhancement effect due to alkyl substitution was
dependant on the size of the alkyl group. Hence, methyl groups were detrimental to the
activity (e.g. compare 7a vs. I and 8a vs. II) whereas the larger isopropyl substituent
gave lower IC values in comparison with the leads. These results indicate that, in this
5
0
series, big and/or lipophilic alkyl substituents are favourable for antitrypanosomal
activity (Figure 2). Remarkably, these SAR results are opposite to the findings reported
by Arafa et al with a series of linear triaryl bisguanidines [17]. In contrast, small alkyl
1
4

ACCEPTED MANUSCRIPT
groups appeared to be favourable to increase the antimalarial activity of the urea (Me >
i
i
Et > Pr) and the benzamide (Et > Pr) derivatives as previously reported [17]. The
introduction of alkoxy substituents on the guanidine groups (7d, 7f, 7g) was detrimental
to both antitrypanosomal and antimalarial activities although N-hydroxy substituents
were somewhat better tolerated (IC < 5 µM). These results are consistent with those
5
0
observed for fused ring bisguanidines [15]. In general, the difference in SAR between
antitrypanosomal and antimalarial action probably reflects the presence of different
cellular targets for both antiparasitic activities (Figure 2).
The excellent in vivo efficacy of 7b and 7c in the mouse model of first stage HAT is
linked to favourable drug-like properties (i.e. bioavailability, lead likeness, Lipinski rule
of 5) as observed with I (Table 3). Considering that compound I was 100% curative by
oral dosage in this mouse model, further in vivo studies with 7b and 7c are clearly
warranted. In contrast, urea derivatives 8b and 8c were weakly active in the STIB900
model and showed acute toxicity issues in vivo. This was somewhat unexpected as both
compounds displayed low cytotoxicity against rat L6-cells (> 100 µM) resulting in high
selectivity indices in vitro (SI > 1200). However, the drug-like properties calculated for
8
b and 8c are clearly less than optimum (Table 3) and may explain the poor efficacy of
these compounds in vivo.
Among the new compounds reported here, two molecules derived from the N-
phenylbenzamide scaffold, 7b and 7c, bound strongly to AT-rich DNA. No correlation
between DNA binding and in vitro activity was observed which was not unexpected as
this is a general trend that has been reported previously with related bisguanidines and
other dicationic minor groove binders [11]. However, it was also noted in these reports
that weak DNA binding results in reduced antitrypanosomal activity [15, 28, 31, 32].
1
5

ACCEPTED MANUSCRIPT
Hence, the level of DNA binding displayed by 7b and 7c is consistent with the
existence of a DNA target (even though this may not be the sole target).
Figure 2. SAR Results for Antitrypanosomal and Antimalarial In Vitro Activities of
Bisguanidine Derivatives. The linker of the compounds reported in this study are
indicated as bold faces (X = NHCO, NHCONH). Other linkers are shown here for
comparative purpose (data reported in reference [11]).
1
6

ACCEPTED MANUSCRIPT
Table 3. Calculated Physical and Chemical Properties and Estimation of Solubility and
Permeability of Orally Active Compounds.
a
Lead b
Lipinski
rule of 5 efficacy
In vivod
Cmpd TPSA cLogP LogD Bioavail.
c
likeness
e
I
157.9
169.9
129.9
129.9
129.9
148.4
148.4
148.4
166.8
170.4
141.9
141.9
141.9
0.28
0.33
0.83
1.55
2.38
1.16
1.87
4.60
2.96
0.40
0.89
1.60
2.43
-2.16
-2.19
-1.77
-1.08
-0.27
0.90
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
■
●
f
II
nd
7
a
nd
g
7
b
●
●
g
7
c
7
d
nd
nd
nd
nd
nd
nd
7
e
1.58
7
f
4.26
7
g
2.67
7
h
0.19
8
a
-1.80
-1.11
-0.30
g
8
b
●
●
g
8
c
Calculations were performed with ChemAxon software Instant JChem 6.1.5 using
a
Chemical Terms expressions (http://www.chemaxon.com). Bioavailability (at least 6
subresults are satisfied): mass ≤ 500 (+), logP ≤ 5 (+), H-donor count ≤ 5 (+), H-
acceptor count ≤ 10 (+), rotatable bond count ≤ 10 (+), PSA ≤ 200 (+), fused aromatic
b
ring count ≤ 5. Lead likeness: mass ≤ 450 & logD7.4 ≥ -4 & logD7.4 ≤ 4 & ring count ≤
c
4
& rotatable bond count ≤ 10 & H-donor count ≤ 5 & H-acceptor count ≤ 8. Lipinski
Ro5 (4 of 4): mass ≤ 500 & logP ≤ 5 & H-donor count ≤ 5 & H-acceptor count ≤
d
e
f
g
1
0. STIB900 mouse model of acute HAT. Oral administration. Not determined. ip
administration.
1
7

ACCEPTED MANUSCRIPT
4
. Conclusions
In this study we have shown that the N-alkylation of the guanidine moiety is an
effective strategy to enhance the antiparasitic activity of leads I and II against T. b.
rhodesiense, whilst it is detrimental to the antimalarial action and DNA binding affinity.
Low nanomolar IC50 values against T. b. rhodesiense were obtained with ethyl and
isopropyl substituents (7b, 7c, 8b, and 8c). However, only 7b and 7c showed in vivo
efficacy (100% cures) in the STIB900 mouse model of acute HAT. Hence, N-alkylated
bisguanidines derived from the N-phenylbenzamide scaffold are very promising DNA
binders with in vivo antitrypanosomal activities that outperform the triaryl class of
bisguanidines [16, 17]. Because of their favourable drug-like properties, further studies
of these molecules are warranted.
Interestingly, several SAR differences against T. brucei were observed (i.e. guanidine
substituent effect) with respect to other bisguanidine classes indicating that no general
trend can be drawn for this class of dicationic compounds. It seems that potent
antitrypanosomal activity depends on the complementary match between the central
diphenyl scaffold and the substituted (or not) guanidine moieties.
5
5
5
. Experimental
.1. Chemistry
.1.1. General
All dry solvents were purchased from Aldrich or Fluka in Sure/Seal bottles. All
reactions requiring anhydrous conditions or an inert atmosphere were performed under a
positive pressure of argon. All reactions were monitored by HPLC–MS or Thin Layer
1
8

ACCEPTED MANUSCRIPT
Chromatography (TLC) using silica gel 60 F254 plates (Merck). Chromatography was
1
13
performed with Isolute SI prepacked columns. H and C NMR spectra were recorded
on a Bruker Advance 300, Varian Inova 400, or Varian Inova 500 spectrometer.
1
Chemical shifts of the H NMR spectra were internally referenced to TMS ( 0 ppm) for
CDCl or the residual proton resonance of the deuterated solvents: D O ( 4.79 ppm),
3
2
CD OD ( 3.31 ppm) and DMSO ( 2.50 ppm). Signal splitting patterns are described
3
as: singlet (s), broad singlet (br s), doublet (d), triplet (t), quadruplet (q), multiplet (m),
or combination thereof. J values are given in Hz. Melting points were determined in
open capillary tubes with a SMP3–Stuart Scientific apparatus or Mettler Toledo MP70
melting point system, and are uncorrected. All compounds are >95% pure by HPLC or
combustion analysis otherwise noted. Elemental analysis was performed on a Heraeus
CHN–O Rapid analyser. Analytical results were within ± 0.4 % of the theoretical values
unless otherwise noted. Analytical HPLC–MS was run with an Xbridge C18–3.5 µm
(2.1×100 mm) column on a Waters 2695 separation module coupled with a Waters
+
Micromass ZQ spectrometer using electrospray ionization (ESI ). The following HPLC
conditions were used: column temperature = 30 ºC, gradient time = 5 min, H O/CH CN
2
3
(
10:90 → 90:10) (HCO H 0.1 %), flow rate = 1 mL/min, UV detection: photodiode
2
array (λ = 190−400 nm). Accurate mass were measured with an Agilent Technologies
Q–TOF 6520 spectrometer using electrospray ionization.
5
.1.2. General procedure for the synthesis of bis(ethoxycarbonylthioureas) (3, 4)
To a solution of diamine 1 or 2 (1 equiv.) in dry CH Cl (15 mL) cooled to 0 ºC and
2
2
under argon atmosphere, was added slowly an excess of ethoxycarbonyl isothiocyanate
2.2 equiv.). The reaction mixture was stirred at room temperature overnight and the
(
1
9

ACCEPTED MANUSCRIPT
precipitate was collected by filtration. The precipitate was rinsed with hexane to yield
the ethoxycarbonylthioureas 3 and 4 with good yields.
5
.1.2.1. Ethyl N-((4-((4-
((((ethoxycarbonyl)amino)methanethioyl)amino)phenyl)carbamoyl)phenyl)carbamothio
yl)carbamate (3). The reaction of 1 (1.0 g, 4.4 mmol) with ethoxycarbonyl
isothiocyanate (1.1 mL, 9.7 mmol) following the general procedure yielded 3 as white
1
solid (2.1 g, 96%). H NMR (300 MHz, DMSO-d ) δ 11.72 (s, 1H, NH), 11.49 (s, 1H,
6
NH), 11.36 (s, 1H, NH), 11.22 (s, 1H, NH), 10.33 (s, 1H, NH), 7.97 (d, J = 9.0, 2H,
ArH), 7.80 (m, 4H, ArH), 7.54 (d, J = 9.0, 2H, ArH), 4.22 (q, J = 7.1, 2H, O-CH -),
2
1
3
4
.21 (q, J = 7.1, 2H, O-CH -), 1.26 (t, J = 7.1, 3H, CH ), 1.25 (t, J = 7.1, 3H, CH ). C
2 3 3
NMR (75 MHz DMSO-d ) δ 178.6 (C=O), 164.8 (2 × C=O), 153.6 (C=N), 153.5
6
(
C=N), 141.0, 137.2, 133.5, 132.0, 128.2, 125.0, 123.7, 120.3, 62.2 (O-CH -), 62.1 (O-
2
+
CH -), 14.2 (2 × CH ). Mp > 300 °C. HPLC (UV) > 95%. LRMS (ES ) m/z = 490.11
2
3
(M+H).
5
.1.2.2. Ethyl N-((4-(((4-
((((ethoxycarbonyl)amino)methanethioyl)amino)phenyl)carbamoyl)amino)phenyl)carba
mothioyl)carbamate (4). The reaction of 2 (1.17 g, 5.76 mmol) with ethoxycarbonyl
isothiocyanate (1.2 mL, 10.6 mmol) following the general procedure yielded 4 as white
1
solid (2 g, 82%). H NMR (400 MHz, DMSO-d ) δ 11.44 (s, 2H, NH), 11.20 (s, 2H,
6
NH), 8.78 (s, 2H, NH), 7.47 (s, 8H, ArH), 4.21 (q, J = 7.1, 4H, O-CH ), 1.26 (t, J = 7.1,
2
1
3
6
1
H, CH ). C NMR (101 MHz, DMSO-d ) δ 178.5 (C=O), 153.6 (C=S), 152.5 (C=S),
3 6
37.7 (2 × Ar-C), 132.0 (2 × Ar-C), 125.3 (Ar, 4 × CH), 118.1 (Ar, 4 × CH), 62.0 (2 ×
+
O-CH -), 14.2 (2 × CH ). Mp > 300 °C. HPLC (UV) > 95%. LRMS (ES ) m/z = 505.39
2
3
(M+H).
2
0

ACCEPTED MANUSCRIPT
5
.1.3. General procedure for the synthesis of N-alkylguanidines (5a-c, 6a-c) and N-
alkoxy guanidines (5d-h).
To a solution of 4 equivalents of alkylamine (i.e., methylamine, ethylamine,
isopropylamine) or hydroxylamine (i.e., O-methylhydroxylamine hydrochloride, O-
ethylhydroxylamine, O-benzylhydroxylamine) in dry CH Cl (5 mL) under argon
2
2
atmosphere, was added diisopropylethylamine (DIPEA, 6 equiv.). The reaction mixture
was stirred 15 minutes at room temperature. The reaction mixture was cooled with an
ice-bath and bis(ethoxycarbonylthiourea) 3 or 4 (1 equiv.) and 1-ethyl-3-(3-
dimethylaminopropyl)carbodiimide (EDAC, 4 equiv.) were added successively. The
reaction mixture (which turned from white to light brown color) was stirred at room
temperature for 12 h. The reaction mixture was diluted with CH Cl (10 mL) and
2
2
washed successively with water (3 × 10 mL) and brine (10 mL). The organic phase was
dried (MgSO ) and the solvent evaporated under vacuum. The product was purified as
4
specified in each case.
5
.1.3.1.
4-(3-Ethoxycarbonyl-2-methylguanidino)-N-(4-(3-ethoxycarbonyl-2-
methylguanidino)phenyl)benzamide (5a)
Compound 5a was obtained following the general procedure with 3 (800 mg, 1.6
mmol), DIPEA (1.71 mL, 9.84 mmol), EDAC (1.26 g, 6.56 mmol) and methylamine (2
M solution in THF, 3.28 mL, 6.56 mmol). The product was purified by precipitation
1
with EtOAc and washed with Et O to yield 5a as white solid (495 mg, 64%). H NMR
2
(
400 MHz, CDCl ) δ 10.55 (br s, 1H, NH), 8.67 (br s, 1H, NH), 7.86 (d, J = 8.6, 2H,
3
ArH), 7.69 (d, J = 8.3, 2H, ArH), 7.18 (m, 2H, ArH), 7.13 (d, J = 8.6, 2H, ArH), 4.85
(
br s, 2H, NH), 4.14 (q, J = 7.1, 4H, O-CH -), 2.94 (s, 3H, N-CH ), 2.89 (s, 3H, N-CH ),
2 3 3
13
1
.29 (m, 6H, -CH ). C NMR (75 MHz, DMSO-d ) δ 164.8 (C=O), 163.4 (C=O), 163.2
3 6
2
1

ACCEPTED MANUSCRIPT
(C=O), 159.1 (C=N), 158.6 (C=N), 141.4 (Ar-C), 136.3 (Ar-C), 132.8 (Ar-C), 129.7
(
Ar-C), 128.2 (Ar-CH), 124.9 (Ar-CH), 122.4 (Ar-CH), 120.6 (Ar-CH), 59.7 (-OCH -),
2
5
9.6 (-OCH -), 28.4 (-NCH ), 28.1 (-NCH ), 14.7 (CH ), 14.6 (CH ). Mp 127–130 °C.
2 3 3 3 3
+
HPLC (UV) = 96 %. LRMS (ES ) m/z = 484.46 (M+H).
5
.1.3.2.
4-(2-Ethyl-3-ethoxycarbonylguanidino)-N-(4-(2-ethyl-3-
ethoxycarbonylguanidino)phenyl)benzamide (5b)
Compound 5b was obtained following the general procedure with 3 (500 mg, 1.0
mmol), DIPEA (1.1 mL, 6.1 mmol), EDAC (784 mg, 4.1 mmol) and ethylamine
hydrochloride (331 mg, 4.1 mmol). The product was purified by silica chromatography
(
10 g SI prepacked column) with CH Cl /EtOAc (9:1) to yield 5b as white solid (419
2 2
1
mg, 81%). H NMR (300 MHz, CDCl ) δ 10.66 (br s, 1H, NH), 8.49 (br s, 1H, NH),
3
7
.86 (d, J = 7.5, 2H, ArH), 7.68 (d, J = 8.5, 2H, ArH), 7.14 (m, 5H, ArH + NH), 4.76
(
br s, 1H, NH), 4.14 (q, J = 7.1, 4H, -OCH -), 3.39 (m, 4H, -NCH -), 1.28 (t, J = 7.1,
2
2
1
3
6
H, CH ), 1.20 (t, J = 6.9, 3H, CH ), 1.11 (t, J = 7.1, 3H, CH ). C NMR (75 MHz,
3 3 3
CDCl ) δ 165.5 (C=O), 136.9 (2 × Ar-C), 129.1 (Ar, 2 × CH), 126.6 (Ar, 2 × CH),
3
1
24.0 (Ar, 2 × CH), 122.1 (Ar, 2 × CH), 61.6 (-OCH -), 61.2 (-OCH -), 36.4 (2 × -
2 2
NCH ), 15.0 (CH ), 14.8 (CH ), 14.6 (2 × CH ). Mp 170–171 °C. HPLC (UV) > 90%.
2
3
3
3
+
HRMS (ES ) required for C H N O : 511.2541 (found: 511.2543).
2
5
33
7
5
5
.1.3.3.
4-(3-Ethoxycarbonyl-2-isopropylguanidino)-N-(4-(3-ethoxycarbonyl-2-
isopropylguanidino)phenyl)benzamide (5c).
Compound 5c was obtained following the general procedure with 3 (800 mg, 1.6
mmol), DIPEA (1.7 mL, 9.8 mmol), EDAC (1.26 g, 6.56 mmol) and isopropylamine
(0.57 mL, 6.56 mmol). The product was purified by silica chromatography (10 g SI
prepacked column) with CH Cl /EtOAc (100/0→90/10) to yield 5c as white solid (690
2
2
1
mg, 77%). H NMR (300 MHz, CDCl ) δ 10.57 (br s, 1H, NH), 8.48 (br s, 1H, NH),
3
2
2

ACCEPTED MANUSCRIPT
7
1
.89 (d, J = 7.9, 2H, ArH), 7.70 (d, J = 8.3, 2H, ArH), 7.13 (m, 4H, ArH), 4.59 (br s,
H, NH), 4.27 (br m, 2H, -CH-), 4.15 (q, J = 7.1, 4H, -OCH -), 1.30 (m, 6H, CH ), 1.20
2
3
1
3
(
(
(
d, J = 6.4, 6H, CH ), 1.11 (d, J = 6.4, 6H, CH ). C NMR (75 MHz, CDCl ) δ 165.5
3 3 3
C=O), 164.7 (2 × C=O), 158.2 (2 × C=N), 137.0 ( 2 × Ar-C), 132.2 (2 × Ar-C), 129.2
Ar, 2 × CH), 126.6 (Ar, 2 × CH), 124.0 (Ar, 2 × CH), 122.2 (Ar-CH), 61.6 (-OCH -),
2
6
1
1.1 (-OCH -), 43.1 (-CH-), 43.0 (-CH-), 23.1 (2 × CH ), 23.0 (2 × CH ), 14.8 (CH ),
2 3 3 3
+
4.6 (CH3). Mp 136–138 °C. HPLC (UV) > 95%. HRMS (ES ) required for
C H N O : 539.2856 (found: 539.2849).
27
37
7
5
5
.1.3.4.
4-(3-Ethoxycarbonyl-2-methoxyguanidino)-N-(4-(3-ethoxycarbonyl-2-
methoxyguanidino)phenyl)benzamide (5d)
Compound 5d was obtained following the general procedure with 3 (500 mg, 1.0
mmol), DIPEA (1.1 mL, 6.1 mmol), EDAC (784 mg, 4.1 mmol) and methoxyamine
hydrochloride (341 mg, 4.1 mmol). The product was purified by silica chromatography
(
5 g SI prepacked column) with CH Cl /EtOAc (9:1) to yield 5d as white solid (226 mg,
2 2
1
4
3%). H NMR (300 MHz, CDCl ) δ 9.17 (br m, 1H, NH), 8.84 (br s, 1H, NH), 8.02 (br
3
s, 2H, NH), 7.80 (d, J = 8.7, 2H, ArH), 7.76 (br s, 1H, NH), 7.53 (d, J = 8.7, 4H, ArH),
7
3
.43 (d, J = 8.7, 2H, ArH), 4.23 (q, J = 7.0, 2H, O-CH -), 4.20 (q, J = 7.0, 2H, O-CH -),
2 2
1
3
.83 (s, 3H, O-CH ), 3.77 (s, 3H, O-CH ), 1.39 (m, 6H, CH ). C NMR (75 MHz,
3
3
3
CDCl ) δ 165.1 (C=O), 153.5 (2 × C=O), 145.2 (Ar-C), 144.2 (C=N), 142.2 (C=N),
3
1
1
35.4 (Ar-C), 132.6 (Ar-C), 127.8 (Ar-C), 128.2 (Ar, 2 × CH), 121.0 (Ar, 2 × CH),
19.7 (Ar, 2 × CH), 118.4 (Ar, 2 × CH), 62.7 (2 × O-CH ), 62.1 (2 × O-CH ), 14.4 (2 ×
2
3
+
CH ). Mp 190–191 °C. HPLC (UV): 97%. HRMS (ES ) required for C H N O:
3
23 29
7
5
15.2128 (found: 515.2134).
5
.1.3.5.
4-(2-Ethoxy-3-ethoxycarbonylguanidino)-N-(4-(2-ethoxy-3-
ethoxycarbonylguanidino)phenyl)benzamide (5e)
2
3

ACCEPTED MANUSCRIPT
Compound 5e was obtained following the general procedure with 3 (500 mg, 1 mmol),
DIPEA (1.1 mL, 6.1 mmol), EDAC (784 mg, 4.1 mmol) and ethoxyamine
hydrochloride (397 mg, 4.1 mmol). The product was purified by silica chromatography
(
5 g SI prepacked column) with CH Cl /EtOAc (9:1) to yield 5e as white solid (299 mg,
2 2
1
5
5%). H NMR (300 MHz, CDCl ) δ 9.17 (br s, 1H, NH), 8.84 (br s, 1H, NH), 8.04 (br
3
s, 2H, NH), 7.79 (d, 2H, J = 8.8, ArH), 7.78 (s, 1H, NH), 7.53 (d, J = 8.8, 4H, ArH),
.44 (d, J = 8.6, 2H, ArH), 4.24 (m, 4H, O-CH -), 4.05 (m, 4H, O-CH -), 1.32 (m, 12H,
7
2
2
13
CH ). C NMR (75 MHz, CDCl ) δ 165.2 (C=O), 153.5 (2 × C=N), 144.6 (Ar-C),
3
3
1
43.9 (C=O), 142.3 (C=O), 135.6 (Ar-C), 132.4 (Ar-C), 128.2 (Ar, 2 × CH), 127.7 (Ar-
C), 121.0 (Ar, 2 × CH), 119.4 (Ar-CH), 118.2 (Ar-CH), 69.7 (-OCH -), 69.5 (-OCH -),
2
2
6
9
5
2.6 (-OCH -), 62.4 (-OCH -), 14.7 (CH ), 14.4 (CH ). Mp 181–183 °C. HPLC (UV):
2 2 3 3
+
5%. LRMS (ES ) m/z = 544.17 (M+H).
.1.3.6. 4-(2-Benzyloxy-3-ethoxycarbonylguanidino)-N-(4-(2-benzyloxy-3-
ethoxycarbonylguanidino)phenyl)benzamide (5f)
Compound 5f was obtained following the general procedure with 3 (800 mg, 1.6 mmol),
DIPEA (1.7 mL, 9.8 mmol), EDAC (1.26 g, 6.56 mmol) and benzyloxyamine (0.76 mL,
6
.56 mmol). The product was purified by silica chromatography (5 g SI prepacked
column) with EtOAc/MeOH (100/0→90/10) to yield 5f as white solid (875 mg, 80%).
1
H NMR (300 MHz, CDCl ) δ 9.18 (br s, 1H, NH), 8.84 (br s, 1H, NH), 8.00 (br m, 2H,
3
NH), 7.78 (d, J = 8.7, 2H, ArH), 7.67 (br s, 1H, NH), 7.52 (m, 4H, ArH), 7.47 – 7.30
(
m, 12H, ArH), 5.03 (s, 2H, -OCH -), 5.00 (s, 2H, -OCH -), 4.23 (q, J = 7.1, 4H, -
2 2
1
3
OCH -), 1.32 (t, J = 7.1, 3H, CH ), 1.31 (t, J = 7.1, 3H, CH ). C NMR (75 MHz,
2
3
3
CDCl ) δ 164.9 (C=O), 153.3 (2 × C=O), 144.8 (C=N), 144.0 (C=N), 143.9 (Ar-C),
3
1
42.1 (Ar-C), 137.5 (Ar-C), 137.3 (Ar-C), 135.4 (Ar-C), 132.3 (Ar-C), 128.8 (Ar, 2 ×
CH), 128.8 (Ar, 2 × CH), 128.43 (Ar, 2 × CH), 128.37 (Ar, 2 × CH), 128.2 (Ar, 2 ×
2
4

ACCEPTED MANUSCRIPT
CH), 127.6 (Ar-CH), 120.8 (Ar, 2 × CH), 119.5 (Ar-CH), 118.2 (Ar-CH), 76.3 (O-CH2-
)
, 76.2 (O-CH -), 62.5 (O-CH -), 62.3 (O-CH -), 14.2 (CH ). Mp 156–158 °C. HPLC
2
2
2
3
+
(
UV) = 95%. LRMS (ES ) m/z = 668.68 (M+H).
5
.1.3.7.
4-(3-Ethoxycarbonyl-2-((tetrahydro-2H-pyran-2-yl)oxy)guanidino)-N-(4-(3-
ethoxycarbonyl-2-((tetrahydro-2H-pyran-2-yl)oxy)guanidino)phenyl)benzamide (5g)
Compound 5g was obtained following the general procedure with 3 (650 mg, 1.3
mmol), DIPEA (1.5 mL, 8.6 mmol), EDAC (1.1 g, 5.7 mmol) and O-(tetrahydro-2H-
pyran-2-yl)hydroxylamine [33] (670 mg, 5.7 mmol). The product was purified by
1
precipitation in EtOAc to yield 5g as white solid (560 mg, 60%). H NMR (400 MHz,
CDCl ) δ 9.21 (br s, 1H, NH), 8.88 (br s, 1H, NH), 7.97 (s, 2H, NH), 7.88 (s, 1H, NH),
3
7
5
3
1
1
1
.79 (d, J = 8.8, 2H, ArH), 7.54 (m, 4H, ArH), 7.44 (d, J = 9.0, 2H, ArH), 5.14 (dd, J =
.8, 2.2, 1H, O-CH-O), 5.10 (dd, J = 6.0, 2.1, 1H, O-CH-O), 4.24 (m, 4H, O-CH -),
2
.94 (m, 2H, -CH -), 3.63 (m, 2H, -CH -), 2.48 (m, 2H, -CH -), 1.86 (m, 4H, -CH -),
2
2
2
2
1
3
.61 (m, 6H, -CH -), 1.34 (m, 6H, CH ). C NMR (100 MHz, CDCl ) δ 165.3 (C=O),
2
3
3
53.5 (2 × C=O), 146.0 (C=N), 145.3 (C=N), 142.1 (Ar-C), 135.5 (Ar-C), 132.7 (Ar-C),
28.3 (Ar, 2 × CH), 128.0 (Ar-C), 121.1 (Ar, 2 × CH) , 119.8 (Ar, 2 × CH), 118.6 (Ar, 2
×
CH), 101.8 (O-CH-O), 101.7 (O-CH-O), 64.3 (O-CH -), 64.2 (O-CH -), 62.9 (O-CH -
2 2 2
)
, 62.7 (O-CH -), 29.6 (-CH -), 29.5 (-CH -), 25.4 (-CH -), 21.0 (-CH -), 20.9 (-CH -),
2 2 2 2 2 2
+
1
4.5 (2 × CH ). Mp 116–118 °C. HPLC (UV) = 95%. LRMS (ES ) m/z = 656.59
3
(M+H).
5
.1.3.8. 1,3-Bis(4-(3-ethoxycarbonyl-2-methylguanidino)phenyl)urea (6a)
Compound 6a was obtained following the general procedure with 4 (800 mg, 1.6
mmol), DIPEA (1.66 mL, 9.52 mmol), EDAC (1.21 g, 6.36 mmol) and methylamine (2
M solution in THF, 3.18 mL, 6.36 mmol). The crude product was triturated in Et O to
2
2
5

ACCEPTED MANUSCRIPT
1
yield 6a as white solid (707 mg, 73%). H NMR (300 MHz, DMSO-d ) δ 9.63 (br s, 2H,
6
NH), 7.45 (d, J = 8.4, 4H, ArH), 7.19 (d, J = 8.4, 4H, ArH), 3.94 (q, J = 7.1, 4H, O-
13
CH -), 2.78 (m, 6H, N-CH ), 1.15 (t, J = 7.1, 6H, CH ) [34]. C NMR (75 MHz,
2
3
3
DMSO-d ) δ 163.7 (C=O), 159.6 (2 × C=O), 153.1 (2 × C=N), 137.9 (2 × Ar-C), 131.2
6
(
2 × Ar-C), 126.0 (Ar, 4 × CH), 118.5 (Ar, 4 × CH), 59.9 (2 × O-CH -), 28.4 (2 × N-
2
+
CH ), 15.0 (2 × CH ). Mp > 280 °C (dec.). HPLC (UV) = 98%. LRMS (ES ) m/z =
3
3
4
99.44 (M + H).
5
.1.3.9. 1,3-Bis(4-(2-ethyl-3-ethoxycarbonylguanidino)phenyl)urea (6b)
Compound 6b was obtained following the general procedure with 4 (800 mg, 1.6
mmol), DIPEA (1.6 mL, 9.5 mmol), EDAC (1.23 g, 6.4 mmol) and ethylamine
hydrochloride (518 mg, 6.4 mmol). The product was purified by silica chromatography
(5 g SI prepacked column) with CH Cl /MeOH (100/0→95/5) to yield 6b as white solid
2 2
1
(
418 mg, 60%). H NMR (300 MHz, DMSO-d ) δ 8.69 (s, 2H, NH), 7.44 (d, J = 8.2,
6
4
H, ArH), 7.20 (d, J = 8.2, 4H, ArH), 3.94 (q, J = 7.0, 4H, -OCH -), 3.28 (m, 4H, -
2
13
NCH -), 1.15 (t, 6H, J = 7.0, CH ), 1.10 (t, 6H, J = 7.0, CH ) [34]. C NMR (75 MHz,
2
3
3
DMSO-d ) δ 163.5 (C=O), 158.4 (2 × C=O), 152.5 (2 × C=N), 137.0 (2 × Ar-C), 131.2
6
(
2 × Ar-C), 125.5 (Ar, 4 × CH), 118.6 (Ar, 4 × CH), 59.6 (2 × -OCH -) , 35.6 (2 × -
2
NCH -), 14.8 (2 × CH ), 14.7 (2 × CH ). Mp 190–191 °C. HPLC (UV) > 95%. LRMS
2
3
3
+
(
ES ) m/z = 527.58 (M+H).
5
.1.3.10. 1,3-Bis(4-(3-ethoxycarbonyl-2-isopropylguanidino)phenyl)urea (6c)
Compound 6c was obtained following the general procedure with 4 (300 mg, 0.6
mmol), DIPEA (0.64 mL, 3.69 mmol), EDAC (472 mg, 2.5 mmol) and isopropylamine
(0.21 mL, 2.46 mmol). The product was purified by silica chromatography (5 g SI
prepacked column) with CH Cl /MeOH (100/0→95/5) to yield 6c as white solid (228
2
2
2
6

ACCEPTED MANUSCRIPT
1
mg, 67%). H NMR (300 MHz, CDCl ) δ 10.52 (br s, 1H, NH), 8.41 (br s, 2H, NH),
3
7
.43 (brm, 4H, ArH), 7.03 (d, J = 8.0, 4H, ArH), 4.70 – 4.38 (br s, 2H, NH), 4.25 – 4.22
(
br s, 2H, -CH-), 4.18 (q, J = 7.0, 4H, O-CH ), 1.28 (t, J = 7.0, 6H, CH ), 1.11 (m, 12H,
2
3
13
CH ). C NMR (75 MHz, CDCl ) δ 164.4 (C=O), 158.4 (2 × C=O), 153.2 (2 × C=N),
3
3
1
38.5 (2 × Ar-C), 129.9 (2 × Ar-C), 126.8 (Ar, 4 × CH), 120.5 (Ar, 4 × CH), 61.6 (2 ×
O-CH -), 43.2 (2 × -CH-), 23.1 (4 × CH ), 14.8 (2 × CH ). Mp > 130 °C (dec.). HPLC
2
3
3
+
(
UV) > 95%. LRMS (ES ) m/z = 555.58 (M+H).
5
.1.4. General procedure for the synthesis of diguanidines 7a-7h and 8a-8c
The reaction was performed in a Kimax glass tube sealed with a screw stopper. To a
solution of bis-3-ethoxycarbonyldiguanidine (5a-g, 6a-c) in MeOH (1 mL) was added
0
.1 M aqueous KOH (2 mL). The tube was stoppered and the reaction mixture was
stirred at 80 ºC for 3–12 h. The reaction mixture was transferred to a round-bottomed
flask and the solvents were evaporated under vacuum. The product was precipitated
from water and triturated with a spatula to give the diguanidine as free base. The
guanidine product was collected by filtration on a fritted plate (Nº 3).
Formation of the diguanidinium dihydrochloride salts. To a solution of the diguanidine
in MeOH (1 mL) was added HCl -saturated 1,4-dioxane solution (3 mL). The reaction
g
mixture was stirred at room temperature for 1 h and the volatiles were removed under
vacuum. The residue was triturated in Et O to give the diguanidinium dihydrochloride
2
salts as colorless solids.
5
.1.4.1. 4-(2-Methylguanidino)-N-(4-(2-methylguanidino)phenyl)benzamide (7a)
2
7