Diaryl sulfide-based inhibitors of trypanothione reductase: Inhibition potency, revised binding mode and antiprotozoal activities

Article abstract of DOI:10.1039/b806371k

Trypanothione reductase (TR) is an essential enzyme of trypanosomatids and therefore a promising target for the development of new drugs against African sleeping sickness and Chagas′ disease. Diaryl sulfides with a central anilino moiety, decorated with a

Full text of DOI:10.1039/b806371k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Diaryl sulfide-based inhibitors of trypanothione reductase: inhibition potency,
revised binding mode and antiprotozoal activities†
Bernhard Stump,a Christian Eberle,a Marcel Kaiser,b Reto Brun,b R. Luise Krauth-Siegelc and
Franc¸ois Diederich*a
Received 15th April 2008, Accepted 5th August 2008
First published as an Advance Article on the web 9th September 2008
DOI: 10.1039/b806371k
Trypanothione reductase (TR) is an essential enzyme of trypanosomatids and therefore a promising
target for the development of new drugs against African sleeping sickness and Chagas¢ disease. Diaryl
sulfides with a central anilino moiety, decorated with a flexible N-alkyl side chain bearing a terminal
ammonium ion, are a known class of inhibitors. Using computer modelling, we revised the binding
model for this class of TR inhibitors predicting simultaneous interactions of the ammonium
ion-terminated N-alkyl chain with Glu18 as well as Glu465¢/Glu466¢ of the second subunit of the
homodimer, whereas the hydrophobic substituent of the aniline ring occupies the “mepacrine binding
site” near Trp21 and Met113. Systematic alteration of the carboxylate-binding fragments and the diaryl
sulfide core of the inhibitor scaffold provided evidence for the proposed binding mode. In vitro studies
showed IC50 values in the low micromolar to submicromolar range against Trypanosoma brucei
rhodesiense as well as the malaria parasite Plasmodium falciparum.
For most of those neglected tropical diseases, there are no safe
and efficacious drugs available. In the case of malaria, potent
Introduction
African sleeping sickness, Chagas¢ disease and the different
forms of leishmaniasis are tropical diseases caused by protozoan
pathogens of the trypanosomatid family. These infections consti-
tute one of the most serious health problems in the developing
world.1 African sleeping sickness (human African trypanosomi-
asis) is caused by Trypanosoma brucei species and transmitted
by tsetse flies. Sleeping sickness threatens 50 million people in
~20 countries of sub-Saharan Africa. The number of infected
people is difficult to establish accurately but can be estimated to be
between 50 000 and 70 000 cases.2 The pathogen of Chagas¢ disease
(American trypanosomiasis), Trypanosoma cruzi, is transmitted to
humans by blood sucking reduviid bugs. The disease is widespread
in central and southern America. Leishmaniasis is caused by
various species of Leishmania causing different forms of disease
ranging from cutaneous to visceral infections. Millions of people
are infected and ~50 000 deaths per year are quoted, mainly due
to L. donovani.3
drugs such as chloroquine are losing efficacy more and more due
to an increase of drug resistances.3,5 Chemotherapy of all forms
of trypanosomiasis is problematic due to the severe side effects of
the few drugs in use, the long duration and high costs of treatment
and an increasing number of drug resistant pathogens.5,6 New and
better drugs to fight these diseases are therefore urgently needed.
A promising strategy to develop new compounds active against
parasites is targeting enzymes of a unique metabolic pathway
that is exclusively present in the pathogens and not in the
human hosts. In the case of trypanosomatids, their specific
thiol redox metabolism distinguishes them from nearly all other
eukaryotes and prokaryotes. Instead of the nearly ubiquitous
glutathione system composed of glutathione and the flavoenzyme
glutathione reductase (GR, EC 1.6.4.2), the parasites possess
trypanothione [N1,N8-bis(glutathionyl)spermidine] that is kept in
its reduced state by trypanothione reductase (TR, EC 1.6.4.8).7
The trypanothione system protects the parasites from oxidative
damage and delivers the reducing equivalents for DNA synthesis.
It has been shown in the past that TR is essential for the parasite
that renders this enzyme an attractive target for the development
of new drugs against trypanosomiasis and leishmaniasis.6,8,9
TR and GR belong to the protein family of FAD disulfide ox-
idoreductases. The mutually exclusive substrate specificity of TR
and GR is considered to rely on their oppositely charged disulfide
substrate binding sites. These accommodate either trypanothione
disulfide (1 in Fig. 1) with an overall positive charge due to the
protonated spermidine bridge or glutathione disulfide (2 in Fig. 1),
with a twofold negative overall charge.
The causative agent of malaria tropica is Plasmodium falci-
parum, a protozoan parasite that is not a member of the trypanoso-
matid family. This disease, spread by the bite of infected Anopheles
mosquitoes, threatens about 40% of the world’s population, mostly
those living in the poorest countries in tropical regions and causes
over 1 million fatalities annually.4
aLaboratorium fu¨r Organische Chemie, ETH Zu¨rich, Ho¨nggerberg,
HCI, 8093, Zu¨rich, Switzerland. E-mail: diederich@org.chem.ethz.ch;
Fax: (+41)-44-632-1109
bSwiss Tropical Institute, Socinstrasse 57, 4002, Basel, Switzerland
cUniversita¨t Heidelberg, Biochemie-Zentrum (BZH), Im Neuenheimer Feld
504, 69120, Heidelberg, Germany
TR is the most thoroughly studied enzyme of the trypanothione
redox metabolism. The crystal structure has been determined
in free form,10–12 in complex with its cofactor NADPH,12 with
glutathionylspermidine,13 with trypanothione,14 and with the
† Electronic supplementary information (ESI) available: Additional fig-
ures, full experimental details for the modelling, the biological assays and
the preparation of compounds 12, 13, 15, 16, 17, 20, 29–33, 39–43, 48–52,
54–60, 64–67, 76, and 78. See DOI: 10.1039/b806371k
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3935
©

View Article Online
value of 1.7 mM,22 while the molecular weight of this compound
remains moderate compared to the extended dimeric structures of
4 and 5.
Overall, a clear structure–activity relationship (SAR) has hardly
ever been disclosed for TR inhibitors. Computer-aided modelling
turned out to be difficult as the huge, solvent exposed active site
of the parasite enzyme allows for plentiful binding orientations of
the ligands.23
Here we present a revised binding model for the class of
diaryl sulfide-based TR inhibitors involving the interaction of
the ligands with two negatively charged regions of the active site
and the occupation of the hydrophobic region near Trp21 and
Met113, the “mepacrine binding site”. Evidence for this proposed
binding mode was found by the SAR derived from a set of newly
synthesised TR inhibitors with altered carboxylate ligands and by
systematically decorating the phenylthio moiety with hydrophobic
substituents. The determination of the antiparasitic properties did
not only reveal decent activity against trypanosomes, but also
showed potency against the malaria parasite P. falciparum.
Results and discussion
Revised binding mode
Apart from mepacrine, no binding mode of any competitive
TR inhibitor has been deciphered by protein crystallography to
date. Previous molecular dynamics simulations for piperazine 3
predicted two nearly identical binding modes.24 Each of them pre-
dicts interactions of the unsubstituted phenylthio moiety with the
hydrophobic groove near Cys52 and Tyr110, whereas the central
anilino moiety is oriented towards, but does not occupy, the hy-
drophobic region near Trp21 and Met113 (Fig. 1 ESI†). The proto-
nated terminal piperazine residue lies in the vicinity of Glu465¢ and
Glu466¢. Both proposed binding modes exclude interactions with
the side chain of Glu18. For the rationalisation of the increased
binding affinity of cationic benzylammonium phenothiazines
compared to non benzylated derivatives, simultaneous binding of
the additional benzyl unit to the “Z site” hydrophobic pocket
(Fig. 2)—roughly formed by Phe395¢, Pro397¢ and Leu398¢—
and ionic interactions of the cationic nitrogen with Glu465¢
and Glu466¢ were proposed by docking experiments.18 The same
docking procedure applied to cationic diaryl sulfide-based TR
inhibitors revealed at least 4 possible binding modes where the
cationic nitrogen either interacts with Glu465¢/Glu466¢ or Ser14.22
It could be shown that diaryl sulfide-based compounds act
as selective TR inhibitors and do not affect hGR.19,24 Although
the positive charge of TR inhibitors appears to be crucial for
the selectivity of these ligands,25 the previously proposed binding
modes consider electrostatical interactions of the diaryl sulfides
with Glu465¢/Glu466¢, residues that are present in TR as well
as hGR. In contrast to Glu465¢/Glu466¢, Glu18 is specific for
TR and we therefore assume that this amino acid residue might
be of importance for the selective binding of diaryl sulfide-based
inhibitors. Further evidence of the role of Glu18 in the recognition
of TR inhibitors is provided by the co-crystal structure with
mepacrine that undergoes H-bonding to Glu18 mediated by a
water molecule.15 Therefore, we were looking for a binding mode
reflecting an interaction of the inhibitor scaffold with this amino
acid side chain.
Fig. 1 Above: trypanothione disulfide (1), the substrate of trypanothione
reductase (TR), found in trypanosomatids. Below: glutathione disulfide
(2), the substrate of human glutathione reductase (hGR).
competitive inhibitor mepacrine15 as well as with covalently bound
mepacrine derivatives.16 The homodimeric enzyme has two active
sites lying opposed on the interface of the two subunits. With a
3
˚
dimension of approximately 22 ¥ 20 ¥ 28 A , each active site of
TR is much wider than those of GR, making it more difficult
to develop low molecular weight compounds that bind selectively
and with high affinity to the solvent exposed disulfide binding site
of the parasite enzyme.
Nevertheless, various compounds have been identified as in-
hibitors of TR over the last decade.6,8 Most of them feature, under
physiological conditions, a protonated amine or a cationic nitrogen
to mimic the positively charged trypanothione. Mepacrine is just
one member in the class of tricyclic scaffolds known to interact
with TR, among other structurally related compounds such as
phenothiazine-based derivatives.17,18 Opening up the central ring
of the tricyclic core of these inhibitors leads to 2-aminodiphenyl
sulfide-based derivatives such as the piperazine-decorated com-
pound 3 (Table 1).19 Upon optimisation of this inhibitor class,
two trends emerged: on the one hand, bridging two diphenyl
sulfide units via an extended amide spacer gave large frameworks
such as the dimeric structure 420 or, if an amine spacer was
used, diphenyl sulfides as 5.21 Both derivatives exhibit IC50-values
(concentration of inhibitor resulting in 50% inhibition) in the high
nanomolar range under the chosen assay conditions (Table 1). On
the other hand, introduction of a permanent positive charge by
incorporation of a quaternary ammonium center as the headpiece
of a flexible N-alkyl chain (compound 6) increased the potency
strongly compared to the piperazine-substituted TR inhibitor 3.
Monomer 6 stood out with a Kic (competitive inhibitor constant)
3936 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 1 Biological activities of previously reported diaryl sulfide derivatives 3,19 4,21 5,20 and 622
3
4
5
6
a
IC50
Kic
—
0.2 mM
—
0.3 mM
0.4 mMc
—
27 mMb
1.7 mMd
a At 21 ◦C, KCl-buffer, 57 mM TS2 as substrate.19 b At 28 ◦C, KCl-buffer, substrate not indicated.21 c At 21 ◦C, KCl-buffer, TS2 as substrate.20 d At 25 ◦C,
KCl-buffer, (ZCG·dmapa)2 as substrate.22 KCl-buffer: 20 mM HEPES, 150 mM KCl, 1 mM EDTA, 0.2 mM NADPH, pH 7.25.
Fig. 2 Left: proposed new binding mode for benzyl ammonium inhibitor 77 with the biphenyl substituent occupying the mepacrine binding site. The
aniline NH group H-bonds to Glu18 and the quaternary ammonium center is located in the negative electrostatic field of Glu465¢ and Glu466¢. Right:
crystal structure of mepacrine bound to TR.15
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3937
©

View Article Online
By computer modelling using MOLOC,26 we analyzed the
binding mode of diaryl sulfide-based TR inhibitors bearing a
flexible piperazine (as in 3) or onium-ion terminated (as in 6,
Table 1) N-alkyl chain. In accordance with the proposal of
Sergheraert et al.,19 the amine headpiece of the alkyl chain was
placed in the negative electrostatic field of Glu465¢ and Glu466¢.
Rotating the central anilino nucleus by 180 degrees compared
to the previously postulated binding mode directs the aniline
nitrogen in an ideal position to undergo H-bonding to Glu18
(Fig. 2). As a result of this rotation, the phenylthio moiety
is now pointing into the hydrophobic patch near Trp21 and
Met113 that is capable to accommodate extended hydrophobic
moieties such as the acridine moiety of the antimalarial drug
mepacrine, as revealed in the co-crystal structure analysis.15 The
aromatic phenylthio moiety undergoes favourable aromatic edge-
to-face interactions with Trp21 and sulfur–arene interactions27
with Met113. Binding to this hydrophobic area has indeed been
proposed for different tricyclic TR inhibitors,28 but not for diaryl
sulfide-based derivatives.
precursor 13 with N-methylpiperazine yielding diphenyl sulfide 11.
The potential inhibitor 14 with its all carbon chain was prepared
by a Sandmeyer transformation of aniline 7 to the corresponding
iodide 15, followed by Sonogashira-coupling with 1-(but-3-ynyl)-
4-methylpiperazine30 and subsequent reduction of the obtained
alkine 16.
For the synthesis of amide 17, a convenient synthetic protocol
has been worked out avoiding any purification of intermediates.
A one-pot two-step procedure starting from 4-bromothiophenol
(18) delivered aniline 19 that was used without further purifi-
cation for the amide coupling and subsequent addition of N-
methylpiperazine to yield the desired 4-bromophenyl sulfide 17
in 87% overall yield after flash chromatography.
To determine if the proposed simultaneous H-bonding to Glu18
and Glu465¢/Glu466¢ plays a critical role in the binding, the
piperazine derivatives 3 and 20 (Scheme 2) as well as 8–11, 14,
17 (Scheme 1) were subjected to kinetic analysis (see ESI†). Under
the chosen assay conditions, the known inhibitor 3 (concentration:
100 mM) caused 33% inhibition and the corresponding bromo
analogue 20 caused 39% inhibition of the TS2 reduction (Table 2).
Introduction of the more basic amidine group in 9 almost doubled
the degree of inhibition, as expected if a direct interaction between
the side chain of Glu18 and the protonated amidinium moiety
takes place. An intramolecular H-bond between the amidine
Importance of the simultaneous H-bonding to Glu18 and
Glu465¢/Glu466¢—synthesis and biological properties of
piperazine derivatives
=
To decide if Glu18 is indeed more important for the binding
of 2-aminodiphenyl sulfide-based TR inhibitors than previously
postulated, we prepared analogues with altered functional groups
in the position of the aromatic amino group of inhibitor 3 in order
to modulate the proposed interaction with Glu18.
C NH and the unprotonated proximal piperazine nitrogen is
expected to further stabilise the association of the ligand in
the active site (Fig. 2a ESI†). In sharp contrast, amide 8 and
its bromo analogue 17 does not inhibit T. cruzi TR essentially.
Modelling suggests that the amide-containing ligands would only
be able to undergo a strong H-bond to Glu18 at the expense
of losing the second H-bond to Glu465¢ or Glu466¢: if the
terminal, N-methylated piperazine nitrogen is protonated and
the amide NH is H-bonding to Glu18, the flexible alkyl chain
has to move away from Glu465¢/Glu466¢ to avoid unfavourable
Starting from 5-chloro-(2-phenylthio)phenylamine22 (7), amide
8 was prepared by acylation of the anilino NH2 group with 3-
chloropropionyl chloride, elimination of the terminal chloride and
subsequent 1,4-addition of N-methylpiperazine to the obtained
acrylamide acceptor (Scheme 1). Amidine 9 was prepared by
reacting precursor 7 with the appropriate 3-(4-methylpiperazin-
1-yl)propionitrile29 under harsh conditions. Methylaniline 10 was
prepared by reduction of amide 8 and subsequent reductive
amination with formaldehyde. Ether 11 was obtained by trans-
formation of aniline 7 to the corresponding phenol 12 under
Sandmeyer-type conditions, followed by alkylation with 1-bromo-
3-chloropropane and final replacement of the primary chloride of
=
C O ◊ ◊ ◊ N interactions between the amide carbonyl group and
the unprotonated proximal piperazine nitrogen (Fig. 2b ESI,
left†). If the piperazine nitrogen connected to the alkyl chain
is protonated, the NH can undergo intramolecular H-bonding
=
to the amide C O group. In this case, without the protonation
of the terminal nitrogen, any interaction of the ligand with
Glu465¢/Glu466¢ is unlikely (Fig. 2b ESI, right†). In addition,
Table 2 Influence of the flexible N-alkyl chain on T. cruzi TR inhibition
3
8
9
10
11
14
17
20
=
=
=
A
B
CH2
NH
H
33%
17%
C O
NH
H
C NH
NH
H
CH2
NCH3
H
26%
9%
CH2
O
H
15%
4%
CH2
CH2
H
—
2%
C O
NH
Br
CH2
NH
Br
39%
16%
R
inh.a
inh.b
0%
0%
64%
30%
5%
0%
c
a 43 mM TS2, 100 mM inhibitor, ca. 4 mU cm-3 TR, 25 ◦C. b 43 mM TS2, 40 mM inhibitor, ca. 4 mU cm-3 TR, 25 ◦C. c Compound not soluble; inh.: T. cruzi
TR inhibition.
3938 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Scheme 1 Synthesis of piperazine-containing TR inhibitors. (i) 1. 3-(4-Methylpiperazin-1-yl)propionitrile,29 AlMe3, toluene, 100 ◦C, 2 d, 39%;
(ii) 3-chloropropionyl chloride, K2CO3, THF, 25 ◦C, 2 h; 2. N-methylpiperazine, 60 ◦C, 14 h, 91% (over 2 steps); (iii) 1. BH3·THF, THF, 65 ◦C, 16 h; 2.
H2SO4, EtOH, 50 ◦C, 3 h; 3. HCHO, HCOOH, H2O, 100 ◦C, 17 h, 71% (over 3 steps); (iv) 1. NaNO2, H2SO4–H2O, 0 ◦C, 2 h; 2. H2SO4–H2O, 100 ◦C, 12 h,
56%; (v) 1-bromo-3-chloropropane, K2CO3, acetone, 60 ◦C, 5 h, 54%; (vi) N-methylpiperazine, dioxane, 100 ◦C, 2 d, 32%; (vii) 1. NaNO2, HCl–H2O,
◦
-10 ◦C, 90 min; 2. KI, H2O -10 ◦C → 25 C, 2.5 h, 66%; (viii) 1-(but-3-ynyl)-4-methylpiperazine,30 PdCl2(PPh3)3, CuI, NEt3, THF, 1 d, 60 ◦C, 36%;
(ix) Pd(OH2)/C, H2, MeOH, 60 ◦C, 4 d, 94%; (x) 1. 2,5-dichloronitrobenzene, Na, MeOH, 60 ◦C, 6 h; 2. Zn, NH4Cl, 65 ◦C, 4 h; (xi) 1. 3-chloropropionyl
chloride, K2CO3, THF, 25 ◦C, 17 h; 2. N-methylpiperazine, 50 ◦C, 5 h, 87% (starting from 18).
in both cases, the amide group is unfavourably forced out of
the plane of the adjacent phenyl ring to undergo H-bonding to
Glu18. According to the binding model reflecting an interaction
of the inhibitors with Glu18, the reduced TR affinities of amides
8 and 17 can be rationalised by the inability of these ligands
to undergo simultaneous interactions with Glu18 and Glu465¢/
Glu466¢.
Replacing the NH functionality (as in 3) by a NCH3 (as in
10), CH2 (as in 11) or O (as in 14) group reduced the T. cruzi
TR inhibition, in agreement with the expected loss of a hydrogen
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3939
©

View Article Online
Scheme 2 Synthesis of diaryl sulfide derivatives with piperazine (as in 3) or onium-ion terminated (as in 6) N-alkyl chain. (i) Na, 2,5-dichloronitrobenzene,
MeOH, 65 ◦C, 3–7 h, 71–92%; (ii) 1. tert-BuLi, THF, -78 ◦C, 10 min; 2. sulfur, -78 ◦C → 25 ◦C, 30 min; 3. 2,5-dichloronitrobenzene, 3–5 h, 66–71%;
(iii) Zn, NH4Cl, MeOH, 65 ◦C, 2–5 h, 43–99%; (iv) 1. 3-chloropropionyl chloride, pyridine, THF, 25 ◦C, 1.5–4 h; 2. BH3·THF, THF, 67 ◦C, 2–4 h, 57–95%
(over 2 steps); (v) N-methylpiperazine, K2CO3, NaI, DMF, microwaves, 80 ◦C → 110 ◦C, 70 min, 51–82%; (vi) NHMe2 (40%, in H2O), DMF, microwaves,
150 ◦C, 5 min, 79–96%; (vii) 3,4-dichlorobenzyl chloride, acetone, microwaves, 120 ◦C, 20 min, 65–94%. DMF = N,N -dimethylformamide.
bond to Glu18 while the overall geometry of the linker is not as
disturbed as in the case of the incorporation of an amide group.
Overall, inhibition of T. cruzi TR by these piperazine derivatives
shows that introduction of a superior carboxylate-binding ligand
fragment in the vicinity of the central anilino moiety increases
inhibition considerably. This suggests a direct interaction of the
anilino NH group of the flexible amine linker in 3 with Glu18, in
contrast to the prediction of previous docking studies.19
Synthesis and inhibitory activity of ligands with extended diaryl
sulfide cores and onium-ion-terminated side chains
Whereas modifications of the flexible piperazine (as in 3) or onium-
ion terminated (as in 6) N-alkyl chain in diaryl sulfide inhibitors
have been extensively investigated over the last decade,20,22,24,31 the
central diphenyl sulfide scaffold remained largely unchanged. Our
revised model for the binding mode of this class of inhibitors
3940 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
suggested that the phenylthio subunit occupies the “mepacrine
binding site” roughly formed by Trp21 and Met112, and encour-
aged the systematic modification of this part of the ligands.
The synthesis of a series of piperazine-decorated target
molecules is shown in Scheme 2. Formation of the thioether
linkage via regioselective nucleophilic aromatic substitution of the
ortho-chloride of 2,5-dichloronitrobenzene with the thiophenol
derivatives 18, 21–25 provided the diaryl sulfides 26–31. The
synthesis of the corresponding trifluoromethylphenyl analogues
32, 33 and 34 was achieved by treatment of the (trifluo-
romethyl)bromobenzene derivatives 35, 36 and 37, respectively,
with tert-BuLi to generate the desired lithiated species. Sulfur
was added to yield the thiolates, which were in situ reacted with
2,5-dichloronitrobenzene to afford the desired diphenyl sulfides
32, 33 and 34. Reduction of the nitro group of the sulfides
26–34 gave anilines 7, 19 and 38–44, which where alkylated to
yield chlorides 45–53. Microwave-assisted introduction of the
piperazine substituent, as described above, afforded the target
compounds 3, 20 and 54–60. Unfortunately, unexpectedly low
solubility, when compared to 3, prevented a reliable determination
of the TR inhibition caused by this series of new piperazine
substituted derivatives.
Inhibitors featuring a permanent positive charge have been
described to exhibit increased TR affinity compared to their
neutral counterparts.18 In particular, Douglas et al. reported high
TR inhibition values for the cationic diphenyl sulfide derivative 6,
bearing a 3,4-dichlorobenzylammonium group as the headpiece
of the flexible N-alkyl chain.22 Therefore, replacement of the
piperazine moiety, as in derivatives 20 and 54–60, against a
3,4-dichlorobenzylammonium headpiece should likewise lead to
increased TR affinity and also to better solubility of the ligands.
In order to prepare the cationic target molecules, chlorides 45–
53 were converted to the corresponding dimethylamino deriva-
tives 61–68 that where quaternized using 3,4-dichlorobenzyl
chloride to yield the benzylammonium cations 6 and 69–75
(Scheme 2).
The preparation of additional cationic ligands with extended
diaryl sulfide cores is depicted in Scheme 3. Palladium-catalyzed
Suzuki–Miyaura cross-coupling32 of 2-tolylboronic acid with bro-
mide 63 gave biphenyl derivative 76, which was further converted
into benzylammonium derivative 77, as described above. On the
other hand, regioselective alkylation of the less hindered terminal
nitrogen of the piperazine moiety of naphthyl derivative 56 with
3,4-dichlorobenzyl chloride yielded cation 78.
As expected, the inhibitors featuring a permanent positive
charge showed better solubility. Both quaternary benzyldimethy-
lammonium and piperazinium ligands displayed a large gain in in-
hibitory potency, when compared to the respective tertiary amine
counterparts (Table 3). Thus, N-methylpiperazine 56 showed only
weak T. cruzi TR inhibition (which could not be accurately
determined due to solubility problems), the corresponding cationic
species 78 was a much more potent ligand and nearly equally
potent than the dimethylammonium cation 6 (Table 3).
The kinetics of the diaryl sulfide derivatives 6, 69–71, 74, 75 and
77 demonstrated competitive T. cruzi TR inhibition, with Kic val-
ues in the lower micromolar range (Table 4). A Kic value of 1.69 mM
has been reported for phenyl derivative 6,22 measured in a TR
buffer system containing 0.02 M HEPES, 0.15 M KCl, 1 mM EDTA
at pH 7.25 with (ZCG·dmapa)2 (N,N -bis-(benzyloxycarbonyl)-L-
cysteinylglycyl-3-dimethylamino)propylamide as an artificial TR
substrate.33 In our TR assay containing 0.04 M HEPES, 1 mM
EDTA at pH 7.50 with the natural substrate TS2,34 we measured
a higher Kic value for 6 (9 mM).
Derivatisation of the phenylthio substituent by introducing Cl,
Br or CF3 substituents in the para-position as in 69, 70 and 75 or a
CF3 group in the ortho-position as in 74 essentially did not affect
the binding (Table 4). Interestingly, the inhibition constants for
the naphthyl 71 and the biphenyl derivative 77 are also in the same
Scheme 3 Synthesis of cationic ligands with extended diaryl sulfide core and piperazinium headpiece, respectively. (i) 4-Tolylboronic acid, Cs2CO3,
[Pd(PPh3)4], DME, H2O, 60 ◦C, 2.5 d, 79%; (ii) 3,4-dichlorobenzyl chloride, acetone, microwaves, 120 ◦C, 20 min, 65%; (iii) 3,4-dichlorobenzyl chloride,
acetone, 45 ◦C, 1 d, 60%. DME = 1,2-dimethoxyethane.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3941
©

View Article Online
Table 3 Inhibition of T. cruzi TR by diaryl sulfide-based inhibitors with
different amine-headpieces of the flexible N-alkyl chain
Compound
R1
R2
inh.a
Fig. 3 Structures of nitrofuran-derivatized mixed competitive–uncom-
petitive TR inhibitors displaying antitrypanosomal and antiplasmodial
in vitro activity.35
6
Phenyl
ca. 60%
72
2-Naphthyl
ca. 60%
to the phenylthio moiety of the core, as all derivatized phenylthio
moieties of the inhibitors 69–71, 75 and 77 can be modeled into the
“mepacrine binding site” using the revised binding model (Fig. 3
ESI†).
56
78
2-Naphthyl
2-Naphthyl
≤ 5%
ca. 70%
Parasitology
The newly synthesised TR inhibitors were tested for their potency
against the trypanosomatids T. cruzi, T. b. rhodesiense and L.
donovani as well as against the haemosporida parasite Plasmodium
falciparum in vitro. Though no significant influence on the growth
of the intracellular parasites T. cruzi and L. donovani could be
detected, a good correlation between TR inhibition and activity
against T. b. rhodesiense emerged for the derivatisation of the
flexible N-alkyl linker of the piperazine derivatives. Amines 3 and
20 as well as amidine 9 showed T.b.r.-IC50 values between 1.0 and
1.6 mM (Table 5). In contrast, the amide derivatives 8 and 17,
causing no detectable TR inhibition, were also inactive in in vitro
studies against T. b. rhodesiense.
a 40 mM Inhibitor, 105 mM TS2, 100 mM NADPH, 25 ◦C. inh.: T. cruzi TR
inhibition.
Table 4 T. cruzi TR inhibition of benzylammonium compounds
Although P. falciparum does not have the unique trypanothione
metabolism, simultaneous activities against P. falciparum and T.
b. rhodesiense have been reported previously for tricyclic and
diphenyl sulfide-based TR inhibitors bearing a permanent positive
charge.22 Therefore, the newly synthesised compounds were also
tested for activity against the malaria parasite. In contrast to the
activities against T. b. rhodesiense, all three piperazines exhibited
remarkable activities against P. falciparum (Table 5). Introduction
of the amide bond even appears beneficially for the antiplasmodial
properties, as the activity of the acylated aniline 8 (IC50 = 0.86 mM)
Compound
R
Kica/mM
6
Phenyl
9
10
8
1
3
4
5
2
3
1
69
70
71
74
75
77
4-Chlorophenyl
4-Bromophenyl
1-Naphthyl
3-(Trifluoromethyl)phenyl
4-(Trifluoromethyl)phenyl
4¢-Methylbiphenyl
9
5
6
8
a 5–10 mU cm-3 TR, 25 ◦C.
was slightly increased compared to the alkylated aniline 3 (IC50
1.55 mM), while the cytotoxicity was tenfold lower (L-6-IC50
=
=
range as for the phenyl derivative 6. According to the previously
proposed binding mode of 3,19 the additional tolyl residue should
clash with residues in the vicinity of the catalytic Cys52/57. This
would result in a substantial loss in ligand affinity, which, however,
is not observed.
The results from the variation of the anilino nitrogen, connect-
ing the flexible alkyl side chain with the diaryl sulfide core, support
that two negatively charged regions of the TR active site, formed
by Glu18 and Glu465¢/Glu466¢, interact with the bound ligands.
The anilino NH undergoes H-bonding to the side chain of Glu18
and the protonated terminal piperazine nitrogen (as in 3), as well
as the permanently charged quaternary onium ions (as in 6 and
the new derivatives described in this paper), interact with the side
chains of Glu465¢/Glu466¢.
10.48 mM for 3, 107.50 mM for 8) leading to a selectivity index
higher than 100 for amide 8.
Simultaneous activities against T. b. rhodesiense as well as
against P. falciparum likewise emerged for the dimethylamine- and
piperazine-bearing derivatives 3, 20, 54–60 and 61–68, respectively.
For the piperazine derivatives 3, 20 and 54–60, altering the
phenylthio moiety preserved the activity against P. falciparum
(P. f. -IC50 between 0.86 and 1.55 mM), whereas the replacement of
a derivatized phenylthio moiety (as in 3, 20, 54–56, 58–60, T.b.r.-
IC50 between 0.70 and 2.03 mM) by a pyridinethio moiety (as in
57, not active) annihilates the activity against T. b. rhodesiense
(Table 6a). As a drawback, all these diaryl sulfide derivatives
bearing a tertiary nitrogen as the headpiece of the flexible alkyl
chain showed considerable cytotoxicity towards L-6 myoblast cells.
In the case of the derivatives with a dimethylamine headpiece, only
the phenyl- and the meta-trifluoromethylphenyl sulfide 61 and 67,
respectively, showed no effect on the growth of T. b. rhodesiense.
Coupled with the establishment of these interactions is a
different orientation of the diaryl sulfide moiety (Fig. 2 and Fig. 3
ESI†). This proposal is supported by the maintained activity of the
ligands bearing bulky hydrophobic residues (such as 77) attached
3942 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Table 5 Influence of altering the flexible N-alkyl chain of diaryl sulfide-based TR inhibitors on their in vitro antiprotozoal activities against T. b.
rhodesiense and P. falciparum
T. brucei rhodesiensea
P. falciparumb
L-6 cellsc
R
X
TR inhibitor
IC50/mM
SI
IC50/mM
SI
IC50/mM
3
H
H
H
Br
Br
CH2
Yes
No
Yes
No
Yes
0.98
10.7
50.3
6.9
1.55
0.86
0.65
1.01
1.31
6.8
10.5
107.5
78.5
26.9
9.1
Not actived
1.56
125.0
120.7
26.7
7.0
=
C O
8
=
9
17
20
C NH
C O
Not actived
1.32
=
CH2
Assays were run in duplicate and repeated once.a STIB900 T. b. rhodesiense strain, trypomastigote stage. b K1 strain, intra-erythrocytic form (IEF). c Rat
myoblast cells to assess cytotoxicity. d Less than 50% growth inhibition at 0.8 mg inhibitor per cm3; IC50: inhibitory concentration 50%; SI: selectivity
index: IC50 for L-6/IC50 for parasite.
Table 6 In vitro activities of piperazine- and dimethylamine-bearing diaryl sulfides against T. b. rhodesiense and P. falciparum
T. brucei rhodesiensea
P. falciparumb
Cytotoxicityc
R
IC50/mM
SI
IC50/mM
SI
6.8
7.0
6.8
6.5
9.5
63.9
10.9
7.2
IC50/mM
3
Phenyl
0.98
1.32
0.97
0.70
10.7
6.9
9.4
12.1
12.4
0.89
10.4
9.2
4.8
1.55
1.31
1.33
1.31
0.92
71.8
0.86
1.35
1.23
10.5
9.1
9.1
8.5
8.7
20
54
55
56
57
58
59
60
4-Br-phenyl
4-Cl-phenyl
1-Naphthyl
2-Naphthyl
2-Pyridinyl
2-CF3-phenyl
3-CF3-phenyl
4-CF3-phenyl
0.70
Not actived
0.90
9.4
9.7
9.7
1.06
2.03
7.9
61
62
63
64
65
66
67
68
Phenyl
Not actived
1.44
2.33
1.89
1.62
1.05
1.52
0.99
1.51
1.25
1.68
1.12
1.72
12.2
5.0
6.3
5.8
7.4
7.3
9.1
5.7
12.8
7.6
9.5
8.7
9.2
12.2
10.2
9.8
4-Cl-phenyl
4-Br-phenyl
1-Naphthyl
2-Naphthyl
2-CF3-phenyl
3-CF3-phenyl
4-CF3-phenyl
5.3
4.1
4.6
5.7
5.9
2.06
Not actived
1.16
8.5
Assays were run in duplicate and repeated once.a STIB900 T. b. rhodesiense strain, trypomastigote stage. b K1 strain, intra-erythrocytic form (IEF). c Rat
myoblast cells to assess cytotoxicity. d Less than 50% growth inhibition at 0.8 mg inhibitor per cm3; IC50: inhibitory concentration 50%; SI: selectivity
index: IC50 for L-6/IC50 for parasite.
All other dimethylamines exhibited T.b.r.-IC50s between 1.16 and
2.33 mM and all compounds affected the growth of P. falciparum
with IC50-values in the range of 0.99 to 1.72 mM (Table 6a).
In vitro activities against T. b. rhodesiense of the cationic, 5-
nitrofurfuryl substituted diaryl sulfide-based TR inhibitors 79–82
(Fig. 3) with IC50-values between 0.59 and 1.02 mM have been
reported recently.35 In vitro studies with the malaria parasite now
revealed, in addition, rather strong antiplasmodial activities with
IC50-values between 0.58 and 0.85 mM (Table 6b). The replacement
of the 5-nitrofurfuryl substituent as in 79–82 (IC50-values from
0.58 to 0.85 mM) by a 3,4-dichlorobenzyl unit as in 6 and 69–75
did not substantially affect the antiplasmodial activity (IC50-values
from 0.32 to 1.13 mM), but completely abolished any activity on T.
b. rhodesiense growth.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3943
©

View Article Online
Table 7 In vitro activities against T. b. rhodesiense and P. falciparum of cationic diaryl sulfides
T. brucei rhodesiensea
P. falciparumb
Cytotoxicityc
R
IC50/mM
SI
IC50/mM
SI
IC50/mM
6
Phenyl
Not actived
Not actived
Not actived
Not actived
Not actived
Not actived
Not actived
Not actived
1.13
0.37
0.45
0.55
0.67
0.34
0.38
0.32
42.2
32.4
29.1
37.3
33.7
29.7
29.2
45.3
47.7
12.0
13.1
20.5
22.6
10.1
11.1
14.5
69
70
71
72
73
74
75
4-Cl-phenyl
4-Br-phenyl
1-Naphthyl
2-Naphthyl
2-CF3-phenyl
3-CF3-phenyl
4-CF3-phenyl
79
80
81
82
Phenyl
0.59
0.77
0.68
1.02
> 288
0.67
0.59
0.58
0.85
> 254
123.6
149.1
32.5
> 170
72.9
4-Cl-phenyl
4-Br-phenyl
4-CF3-phenyl
94.7
127.2
27.1
86.5
27.6
Assays were run in duplicate and repeated once.a STIB900 T. b. rhodesiense strain, trypomastigote stage. b K1 strain, intra- erythrocytic form (IEF). c Rat
myoblast cells to assess cytotoxicity. d Less than 50% growth inhibition at 0.8 mg inhibitor per cm3; IC50: inhibitory concentration 50%; SI: selectivity
index: IC50 for L-6/IC50 for parasite.
In vitro studies with the trypanosomatids T. b. rhodesiense, T.
Conclusions
cruzi and L. donovani revealed activity of the tertiary amines
We report experimental evidence for a newly proposed, modelling-
against the causative agent of African sleeping sickness. Interest-
based binding mode for the class of diaryl sulfide-based T. cruzi
ingly, the derivatives exhibited even stronger activities against P.
TR inhibitors. Altering the nature of the flexible N-alkyl chain
falciparum. Whereas the potency of the diaryl sulfide-based deriva-
supported the importance of two carboxylate recognition sites
tives against T. b. rhodesiense was annihilated upon interaction of a
in the inhibitor framework. Modelling rationalizes that these
cationic benzylammonium substituent and could only be retained
ligands interact with the enzyme by simultaneous binding to
by exchange of the benzyl- against a nitrofurfuryl headpiece, all
Glu18 and the electrostatically strongly negatively charged region
of Glu465¢ and Glu466¢. By adding additional substituents to the
cationic derivatives were shown to be active against P. falciparum.
phenylthio unit of the diaryl sulfide core, the inhibition properties
Experimental
remained remarkably unaffected, with Kic values in the low
micromolar range. The similar inhibition values for phenylthio and
biphenylthio derivatives supports that these ligand parts occupy
the large “mepacrine binding site” of TR. The revised binding
geometry, with simultaneous interactions of the flexible N-alkyl
linker with Glu18 as well as Glu465¢/Glu466¢ and the occupation
of the mepacrine binding site by the arylthio substituent of the
aniline moiety, is highly probable for all members of this class of
TR inhibitors.
The proposed binding mode opens up the way to the rational
design of new diaryl sulfide inhibitors with thioaryl substituents
that fulfil even better the steric and electronic requirements of
the spacious hydrophobic patch that is occupied by the acridine
moiety of mepacrine. By benefiting from additional hydrophobic
contacts or by H-bonding to Ser109, an enhanced binding affinity
should be achieved in future work.
General details and procedures
See ESI.†
N-[5-Chloro-2-(phenylthio)phenyl]-3-(4-methylpiperazin-1-yl)-
propanamide (8). Aniline 7 (2.00 g, 8.48 mmol) and K2CO3
(3.52 g, 25.45 mmol) were suspended in THF (100 cm3), 3-
chloropropionyl chloride (0.97 cm3, 10.18 mmol) was added
and the mixture was stirred for 2 h, when N-methylpiperazine
(4.71 cm3, 42.42 mmol) was added to the solution. The mixture
was stirred for 14 h at 60 ◦C. Dilution (H2O, then EtOAc), washing
(saturated aqueous NaCl solution), extraction (EtOAc), drying of
the combined organic layers (MgSO4), filtration and concentration
in vacuo followed by purification by column chromatography (CC)
(SiO2; CH2Cl2–MeOH 95 : 5) delivered amide 8 (3.01 g, 91%)
3944 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
as a white solid; mp: 90 ◦C; nmax(neat)/cm-1 3116, 2946, 2821,
2775, 1683, 1563, 1514, 1397, 1249, 1138, 1009, 811, 736, 687;
dH(300 MHz, CDCl3) 2.22 (s, 3 H), 2.44–2.57 (m, 12 H), 7.03–7.07
(m, 3 H), 7.13–7.19 (m, 1 H), 7.22–7.27 (m, 2 H), 7.38 (d, J =
8.4, 1 H), 8.44 (d, J = 2.1, 1 H) 10.46 (br s, 1 H); dC(75 MHz,
CDCl3) 33.4, 45.7, 52.5, 53.6, 54.3, 120.0, 122.5, 124.7, 126.2,
127.3, 129.2, 135.7, 135.9, 136.3, 141.1, 170.8; MALDI-HR-MS:
calcd for C20H25ClN3OS+ ([M + H]+): 390.1401; found: 390.1394.
25 ◦C, diluted (saturated aqueous NaHCO3 solution), extracted
(CH2Cl2), dried (MgSO4), filtered and concentrated in vacuo.
Purification by CC (SiO2; CH2Cl2–MeOH 98 : 2 → 90 : 10) yielded
the desired product 11 (12 mg, 32%); nmax(neat)/cm-1 2921, 2793,
1574, 1461, 1385, 1250, 1164, 1093, 885, 690; dH(300 MHz, CDCl3)
1.88 (quint, J = 6.9, 2 H), 2.29 (s, 3 H), 2.34–2.56 (m, 8 H), 2.37
(t, J = 6.9, 2 H), 4.02 (t, J = 6.0, 2 H), 6.85 (dd, J = 8.1, 2.1,
1 H), 6.89 (d, J = 2.1, 1 H), 7.05 (d, J = 8.1, 1 H), 7.26–7.30
(m, 5 H); dC(100 MHz, CDCl3) 26.4, 46.0, 53.0, 54.7, 60.1, 67.2,
112.7, 121.1, 122.8, 127.0, 129.1, 130.9, 132.7, 134.0, 134.6, 157.4;
MALDI-HR-MS: calcd for C20H26ClN2OS+ ([M + H]+): 377.1449;
found: 377.1455.
N-[5-Chloro-2-(phenylthio)phenyl]-3-(4-methylpiperazine-1-yl)-
propanamidine (9). 3-(4-Methylpiperazine-1-yl)-propionitrile29
(115 mg, 0.75 mmol) and AlMe3 (2 M in hexane, 0.38 cm3,
0.75 mmol) were added to aniline 7 (93 mg, 0.40 mmol) in toluene
◦
(1 cm3), and the mixture was stirred for 2 d at 100 C. Dilution
1-(4-(5-Chloro-2-(phenylthio)phenyl)butyl)-4-methylpiperazine
(14). A mixture of alkyne 16 (20 mg, 0.054 mmol) and
Pd(OH)2/C (20% wt dry basis, 2 mg) in MeOH (10 cm3) was
(H2O and EtOAc), washing (saturated aqueous Na2CO3 solution),
extraction of the aqueous phases (EtOAc), drying of the combined
organic layers (MgSO4), filtration and concentration in vacuo
followed by purification by CC (SiO2; CH2Cl2–MeOH 90 : 10)
yielded amidine 9 (60 mg, 39%) as a pale yellow oil; nmax(neat)/cm-1
3162, 2938, 2796, 1635, 1566, 1456, 1371, 1283, 1162, 1086, 1011,
908, 731, 690; dH(300 MHz, MeOD) 2.30 (s, 3 H), 2.35–2.78 (m,
12 H), 6.88 (dd, J = 8.4, 2.1, 1 H), 6.98 (d, J = 8.4, 1 H), 7.27–7.39
(m, 6 H); dC(75 MHz, MeOD) 31.1, 45.9, 52.4, 54.9, 55.0, 114.9,
118.7, 123.8, 125.7, 126.5, 127.5, 129.1, 129.3, 131.7, 132.3, 138.4;
MALDI-HR-MS: calcd for C20H26ClN4S+ ([M + H]+): 389.1561;
found: 389.1554.
◦
stirred under a hydrogen atmosphere for 4 d at 60 C. Filtration
over celite, concentration in vacuo, followed by purification by CC
(SiO2; CH2Cl2–MeOH 99 : 1 → 92 : 8) delivered 14 (19 mg, 94%)
as a colourless oil; nmax(neat)/cm-1 2924, 2793, 1581, 1460, 1163,
1085, 1014, 879, 736, 701; dH(300 MHz, CDCl3) 1.48–1.66 (m,
4 H), 2.32 (s, 3 H), 2.42–2.60 (m, 8 H), 2.56 (t, J = 7.4, 2 H),
2.75 (t, J = 6.0, 2 H), 7.09 (dd, J = 8.4, 2.1, 1 H), 7.15–7.27 (m,
8 H); dC(100 MHz, CDCl3) 27.3, 30.9, 32.9, 44.7, 51.7, 53.8, 57.1,
125.6, 126.0, 127.5, 128.2, 128.7, 128.8, 132.8, 133.7, 135.2, 144.8;
MALDI-HR-MS: calcd for C21H28ClN2S+ ([M + H]+): 375.1656;
found: 375.1662.
5-Chloro-N -methyl-N -(3-(4-methylpiperazin-1-yl)propyl)-2-
(phenylthio)aniline (10). BH3·THF (1 M in THF, 2.5 cm3,
2.5 mmol) was added to a solution of amide 8 (105 mg, 0.27 mmol)
in THF (5 cm3). The mixture was left to stir for 16 h at 65 ◦C
and concentrated in vacuo. The residue was dissolved in EtOH
(10 cm3), conc. sulfuric acid (0.2 cm3) was added and the mixture
was stirred for 3 h at 50 ◦C. Dilution (saturated aqueous NaHCO3
solution), extraction (EtOAc), washing (saturated aqueous NaCl
solution), extraction of the aqueous phases (CH2Cl2), drying of
the combined organic phases (MgSO4) and filtration followed by
concentration in vacuo delivered a yellow oil that was redissolved
in formic acid (0.062 cm3) and H2O (0.5 cm3). HCOH (37% in H2O,
0.043 cm3, 1.62 mmol) was added and the reaction mixture was left
to stir overnight at 100 ◦C. The mixture was cooled to 25 ◦C, HCl
(12 M aqueous solution) was added, the solvent was evaporated
and the residue was made alkaline by addition of 5% aqueous
NaOH solution. Extraction (CH2Cl2), drying of the combined
organic phases (MgSO4), filtration and concentration in vacuo,
followed by purification by CC (SiO2; CH2Cl2–MeOH 98 : 2 →
90 : 10) delivered methylamine 10 (75 mg, 71%) as a light yellow
oil; nmax(neat)/cm-1 3446, 2793, 1572, 1458, 1390, 1281, 1165, 1013,
803, 691; dH(300 MHz, CDCl3) 1.73 (quint, J = 7.5, 2 H), 2.27
(s, 3 H), 2.35 (t, J = 7.5, 2 H), 2.35–2.56 (m, 8 H), 3.02 (t, J =
7.8, 2 H), 6.80 (d, J = 8.0, 1 H), 6.85 (dd, J = 8.4, 2.1, 1 H),
7.03 (d, J = 2.1, 1 H), 7.30–7.36 (m, 5 H); dC(100 MHz, CDCl3)
24.4, 41.6, 45.3, 52.2, 54.0, 54.3, 55.6, 121.3, 123.9, 127.8, 129.4,
130.9, 131.8, 132.3, 132.7, 134.0, 152.1; MALDI-HR-MS: calcd
for C21H29ClN3S+ ([M + H]+): 390.1765; found: 390.1758.
General procedure for the quaternisation of amines using
3,4-dichlorobenzyl chloride
A mixture of the dimethylamine derivative (1 eq.) and 3,4-
dichlorobenzyl chloride (10 eq.), dissolved in acetone (ca. 2 cm3),
was stirred in a closed vessel in a microwave reactor for 20 min
at 120 ◦C before it was concentrated in vacuo. The residue was
suspended in hexane, the solvent decanted and the residue dried in
vacuo and purified by reversed phase HPLC (CH3CN–H2O (0.1%
TFA) 50 : 0 → 100 : 0 in 50 min).
3-({5-Chloro-2-[1-naphthylthio]phenyl}amino)-N-(3,4-dichloro-
benzyl)-N,N -dimethyl-propan-1-ammonium chloride (71). In-
hibitor 71 (37 mg, 97%) was obtained as a colourless oil starting
from amine 64 (25 mg, 67 mmol) and 3,4-dichlorobenzyl chloride
(0.093 cm3, 0.67 mmol); nmax(neat)/cm-1 3029, 2961, 2839, 1677,
1582, 1504, 1473, 1201, 1174, 1122, 1036, 831, 799, 771, 720;
dH(300 MHz, CDCl3) 1.79–1.86 (m, 2 H), 2.71 (s, 6 H), 2.81–
2.87 (m, 2 H), 3.11–3.13 (m, 2 H), 4.30 (s, 2 H), 4.81 (t, J = 5.7,
1 H), 6.60 (d, J = 1.8, 1 H), 6.72 (dd, J = 8.1, 1.8, 1 H), 6.86 (d,
J = 7.5, 1 H), 7.01 (dd, J = 8.4, 1.8, 1 H), 7.16 (t, J = 7.5, 1 H),
7.28–7.33 (m, 2 H), 7.40 (d, J = 8.1, 1 H), 7.47–7.60 (m, 3 H), 7.79
(dd, J = 7.5, 1.5, 1 H), 8.28 (d, J = 6.0, 1 H); dC(75 MHz, CDCl3)
22.7, 39.6, 49.6, 61.5, 66.2, 110.4, 113.4, 117.9, 123.7, 124.2, 125.7,
126.3, 126.7, 128.5, 131.1, 131.9, 133.0, 133.3, 133.5, 134.0, 135.5,
137.4, 138.2, 148.7 (three signals not visible); MALDI-HR-MS:
calcd for C28H28Cl3N2S+ ([M - Cl]+): 529.1033; found: 529.1025.
3-({5-Chloro-2-[2-naphthylthio]phenyl}amino)-N-(3,4-dichloro-
benzyl)-N,N -dimethyl-propan-1-ammonium chloride (72). In-
hibitor 72 (87 mg, 73%) was obtained as a white solid starting
from amine 65 (44 mg, 0.12 mmol)◦and 3,4-dichlorobenzyl chloride
(0.16 cm3, 1.19 mmol); mp: 156 C; nmax(neat)/cm-1 3362, 2964,
1-(3-(5-Chloro-2-(phenylthio)phenoxy)propyl)-4-methylpiper-
azine (11). N-Methylpiperazine (0.11 cm3, 0.99 mmol) was added
to a solution of chloride 13 (31 mg, 0.099 mmo◦l) in dioxane
(3 cm3). The mixture was left to stir for 2 d at 100 C, cooled to
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3945
©

View Article Online
1581, 1505, 1475, 1420, 1264, 1139, 1092, 1037, 825, 789, 668;
dH(300 MHz, CDCl3) 1.85–2.00 (m, 2 H), 2.67 (s, 6 H), 3.02–3.07
(m, 2 H), 3.25–3.29 (m, 2 H), 4.53 (s, 2 H), 5.71 (t, J = 6.0, 1 H),
6.64 (d, J = 2.1, 1 H), 6.77 (dd, J = 8.1, 2.1, 1 H), 7.20 (dd, J = 8.7,
2.0, 1 H), 7.34–7.44 (m, 6 H), 7.50 (d, J = 8.4, 1 H), 7.60–7.73 (m,
3 H); dC(75 MHz, MeOD) 21.8, 39.1, 49.0, 62.5, 65.6, 110.0, 112.9,
116.7, 124.3, 125.1, 125.6, 126.6, 126.7, 127.4, 127.4, 128.3, 131.0,
131.6, 132.1, 132.9, 133.6, 133.9, 134.3, 135.0, 137.5, 138.5, 149.6;
MALDI-HR-MS: calcd for C28H28Cl3N2S+ ([M - Cl]+): 529.1033;
found: 529.1033.
Acknowledgements
This research was supported by a graduate student fellowship from
Novartis (to B. S.), the UNICEF/UNDP/World Bank/WHO
Special Programme for Research and Training in Tropical Diseases
(TDR) (to M. K. and R. B.) and by the Deutsche Forschungsge-
meinschaft (SFB 544, project B3, L. K.-S.). We thank Jo¨rg Klein
for the synthesis of piperazine derivatives and Edith Ro¨ckel and
Natalie Dirdjaja for help in measuring TR assays.
References
3-[(5-Chloro-2-{[2-(trifluoromethyl)phenyl]thio}phenyl)amino]-
N-(3,4-dichlorobenzyl)-N,N -dimethylpropan-1-ammonium chlo-
ride (73). Inhibitor 73 (28 mg, 74%) was obtained as a white solid
starting from amine 66 (25 mg, 64 mmol) and 3,4-dichlorobenzyl
chloride (0.098 cm3, 0.64 mmol); mp: 138 ◦C; nmax(neat)/cm-1
3348, 3261, 3093, 3021, 2966, 2864, 1590, 1464, 1228, 1137, 1117,
1031, 731, 617; dH(300 MHz, CDCl3) 2.01–2.12 (m, 2 H), 3.19 (s,
6 H), 3.25–3.26 (m, 2 H), 3.48–3.54 (m, 2 H), 4.92 (br s, 1 H),
5.13 (s, 2 H), 6.66 (d, J = 2.1, 1 H), 6.79 (dd, J = 7.5, 2.1, 1 H),
6.87 (d, J = 7.5, 1 H), 7.21–7.27 (m, 2 H), 7.42–7.47 (m, 2 H),
7.55–7.68 (m, 3 H); dC(75 MHz, CDCl3) 22.8, 40.2, 49.7, 61.9,
66.3, 111.0, 112.0, 118.3, 125.8, 126.8 (q, J = 5.7), 127.3, 127.8 (q,
J = 30.7), 128.3, 131.4, 132.3, 132.7, 133.7, 134.6, 135.9, 136.1,
138.6, 139.3 (CF3-signal not visible); dF(282 MHz, CDCl3) -60.8
(s, 3 F); MALDI-HR-MS: calcd for C25H25Cl3F3N2S+ ([M - Cl]+):
547.0751; found: 547.0742.
1 (a) http://www.cdc.gov/malaria/impact/index.htm; TDR, 2003;
(b) Wkly Epidemiol. Rec., 2006, 81, 71; (c) Working Paper on Priority
Infectious Diseases Requiring Additional R & D, WHO-IFPMA
Round Table, Geneva Switzerland, 2001.
2 African trypanosomiasis (sleeping sickness), fact sheet N◦259, WHO,
2006.
3 M. H. Gelb and W. G. Hol, Science, 2002, 297, 343.
4 Malaria, fact sheet N◦94, WHO, 2007.
5 J. Keiser, A. Stich and C. Burri, Trends Parasitol., 2001, 17,
42.
6 R. L. Krauth-Siegel, H. Bauer and R. H. Schirmer, Angew. Chem., Int.
Ed., 2005, 44, 690.
7 A. H. Fairlamb, P. Blackburn, P. Ulrich, B. T. Chait and A. Cerami,
Science, 1985, 227, 1485.
8 R. L. Krauth-Siegel and O. Inhoff, Parasitol. Res., 2003, 90,
77.
9 S. Krieger, W. Schwarz, M. R. Ariyanayagam, A. H. Fairlamb,
R. L. Krauth-Siegel and C. Clayton, Mol. Microbiol., 2000, 35,
542.
10 W. N. Hunter, S. Bailey, J. Habash, S. J. Harrop, J. R. Helliwell, T.
Aboagye-Kwarteng, K. Smith and A. H. Fairlamb, J. Mol. Biol., 1992,
227, 322.
11 J. Kuriyan, X. P. Kong, T. S. R. Krishna, R. M. Sweet, N. J. Murgolo, H.
Field, A. Cerami and G. M. Henderson, Proc. Natl. Acad. Sci. U. S. A.,
1991, 88, 8764.
12 C. B. Lantwin, I. Schlichting, W. Kabsch, E. F. Pai and R. L. Krauth-
Siegel, Proteins, 1994, 18, 161.
13 S. Bailey, K. Smith, A. H. Fairlamb and W. N. Hunter, Eur. J. Biochem.,
1993, 213, 67.
14 C. S. Bond, Y. Zhang, M. Berriman, M. L. Cunningham, A. H.
Fairlamb and W. N. Hunter, Struct. Fold Des., 1999, 7, 81.
15 E. M. Jacoby, I. Schlichting, C. B. Lantwin, W. Kabsch and R. L.
Krauth-Siegel, Proteins, 1996, 24, 73.
16 A. Saravanamuthu, T. J. Vickers, C. S. Bond, M. R. Peterson, W. N.
Hunter and A. H. Fairlamb, J. Biol. Chem., 2004, 279, 29493.
17 T. J. Benson, J. H. McKie, J. Garforth, A. Borges, A. H. Fairlamb and
K. T. Douglas, Biochem. J., 1992, 286, 9.
18 M. O. F. Khan, S. E. Austin, C. Chan, H. Yin, D. Marks, S. N. Vaghjiani,
H. Kendrick, V. Yardley, S. L. Croft and K. T. Douglas, J. Med. Chem.,
2000, 43, 3148.
19 R. Fernandez-Gomez, M. Moutiez, M. Aumercier, G. Bethegnies, M.
Luyckx, A. Ouaissi, A. Tartar and C. Sergheraert, Int. J. Antimicrob.
Agents, 1995, 6, 111.
3-[(5-Chloro-2-{[3-(trifluoromethyl)phenyl]thio}phenyl)amino]-
N-(3,4-dichlorobenzyl)-N,N -dimethylpropan-1-ammonium chlo-
ride (74). Inhibitor 74 (20 mg, 74%) was obtained as a white solid
starting from amine 67 (18 mg, 46 mmol) and 3,4-dichlorobenzyl
chloride (0.064 cm3, 0.46 mmol); mp: 168 ◦C; nmax(neat)/cm-1
3282, 3002, 2976, 2962, 2818, 1582, 1471, 1320, 1165, 1115,
1098, 1072, 796, 695; dH(300 MHz, CDCl3) 2.01–2.15 (m, 2 H),
3.16 (s, 6 H), 3.27–3.31 (m, 2 H), 3.56–3.61 (m, 2 H), 5.01–5.09
(m, 3 H), 6.67 (d, J = 2.0, 1 H), 6.72 (dd, J = 8.4, 2.0, 1 H),
7.14–7.44 (m, 6 H), 7.53 (dd, J = 8.4, 1.5, 1 H), 7.71 (d, J =
1.5, 1 H); dC(75 MHz, CDCl3) 22.7, 40.2, 49.7, 62.2, 66.1, 110.7,
111.6, 118.1, 122.4 (q, J = 3.9), 122.7 (q, J = 3.9), 123.7 (q,
J = 272.8), 127.3, 129.6, 131.3, 131.4 (q, J = 31.1), 132.6, 133.6,
134.6, 135.8, 138.1, 138.4, 138.9, 149.3 (one signal not visible);
dF(282 MHz, CDCl3) -62.5 (s, 3 F); MALDI-HR-MS: calcd for
C25H25Cl3F3N2S+ ([M - Cl]+): 547.0751; found: 547.0757.
3-({5-Chloro-2-[(4¢-methylbiphenyl-4-yl)thio]phenyl}amino)-N-
(3,4-dichlorobenzyl)-N,N -dimethylpropan-1-ammonium chloride
(77). Inhibitor 77 (30 mg, 65%) was obtained as a pale yellow oil
starting from amine 76 (33 mg, 80 mmol) and 3,4-dichlorobenzyl
chloride (0.11 cm3, 0.16 mmol); nmax(neat)/cm-1 3404, 2924,
2854, 1678, 1583, 1506, 1443, 1207, 1133, 1036, 845, 803, 725;
dH(300 MHz, CDCl3) 1.97–2.04 (m, 2 H), 2.35 (s, 3 H), 2.87 (s,
6 H), 3.36–3.40 (m, 4 H), 5.59 (s, 2 H), 5.03–5.05 (m, 1 H), 6.65 (d,
J = 2.2, 1 H), 6.77 (dd, J = 8.1, 2.2, 1 H), 7.11 (d, J = 8.3, 2 H),
7.18 (d, J = 8.3, 2 H), 7.24–7.50 (m, 8 H); dC(75 MHz, CDCl3)
21.2, 27.2, 42.2, 45.4, 49.9, 58.1, 111.3, 111.6, 116.4, 117.3, 120.3,
124.2, 127.4, 127.6, 128.1, 128.4, 128.6, 130.6, 130.8, 136.5, 138.4,
138.7, 139.8, 151.6, 162.9, 163.4; MALDI-HR-MS: calcd for
C31H32Cl3N2S+ ([M - Cl]+): 569.1346; found: 569.1350.
20 B. Bonnet, E. Soullez, E. Davioud-Charvet, V. Landry, D. Horvath and
C Sergheraert, Bioorg. Med. Chem., 1997, 5, 1249.
21 S. Girault, E. Davioud-Charvet, L. Maes, J.-F. Dubremetz, M.-A.
Debreu, V. Landry and C. Sergheraert, Bioorg. Med. Chem., 2001, 9,
837.
22 S. Parveen, M. O. F. Khan, S. E. Austin, S. L. Croft, V. Yardley, P. Rock
and K. T. Douglas, J. Med. Chem., 2005, 48, 8087.
23 W. N. Hunter, Biochem. Soc. Trans., 2007, 35, 980–984.
24 S. Baillet, E. Buisine, D. Horvath, L. Maes, B. Bonnet and C.
Sergheraert, Bioorg. Med. Chem., 1996, 4, 891.
25 C. H. Faerman, S. N. Savvides, C. Strickland, M. A. Breidenbach, J. A.
Ponasik, B. Ganem, D. Ripoll, R. L. Krauth-Siegel and P. A. Karplus,
Bioorg. Med. Chem., 1996, 4, 1247.
26 P. R. Gerber and K. Mu¨ller, J. Comput. Aided Mol. Des., 1995, 9,
251.
27 E. A. Meyer, R. K. Castellano and F. Diederich, Angew. Chem., Int.
Ed., 2003, 42, 1210.
3946 | Org. Biomol. Chem., 2008, 6, 3935–3947
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
28 J. Garforth, H. Yin, J. H. McKie, K. T. Douglas and A. H. Fairlamb,
J. Enzym. Inhib., 1997, 12, 161.
29 C. J. Hamilton, A. Saravanamuthu, A. H. Fairlamb and I. M.
Eggleston, Bioorg. Med. Chem., 2003, 11, 3683.
30 M. Franceschin, A. Alvino, V. Casagrande, C. Mauriello, E. Pascucci,
M. Savino, G. Ortaggi and A. Bianco, Bioorg. Med. Chem., 2007, 15,
1848.
32 N. Miyaura, in Metal catalyzed cross-coupling reactions, ed. A. de
Meijere and F. Diederich, Wiley-VCH, Weinheim, 2nd edn, 2004, vol.
1, ch. 2, p. 41.
33 C. T. Yuen, J. Garforth, T. Besheya, R. Jaouhari, J. H. McKie, A. H.
Fairlamb and K. T. Douglas, Amino Acids, 1999, 17, 175.
34 M. C. Jockers-Scheru¨bl, R. H. Schirmer and R. L. Krauth-Siegel,
Eur. J. Biochem., 1989, 180, 267.
31 S. Girault, S. Baillet, D. Horvath, V. Lucas, E. Davioud-Charvet, A.
Tartar and C. Sergheraert, Eur. J. Med. Chem., 1997, 32, 39.
35 B. Stump, M. Kaiser, R. Brun, R. L. Krauth-Siegel and F. Diederich,
ChemMedChem, 2007, 2, 1708.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 3935–3947 | 3947
©