Dibenzosuberyl substituted polyamines and analogs of clomipramine as effective inhibitors of trypanothione reductase; Molecular docking, and assessment of trypanocidal activities

Article abstract of DOI:10.1016/j.bmc.2015.01.018

Trypanothione reductase (TR) is an enzyme critical to the maintenance of the thiol redox balance in trypanosomatids, including the genera Trypanosoma and Leishmania that are parasites responsible for several serious diseases. Analogs of clomipramine were prepared since clomipramine is reported to inhibit TR and cure mice infected with trypanosomes, however its psychotropic activity precludes its use as an anti-trypanosomal therapeutic. The clomipramine analogs contained a tricyclic dibenzosuberyl moiety. Additionally a series of polyamines with N-dibenzosuberyl substituents were prepared. All compounds studied were competitive inhibitors of TR and showed trypanocidal activities against Trypanosoma brucei in vitro. The analogs of clomipramine were poor inhibitors of TR, whereas the polyamine derivatives were effective TR inhibitors with the most potent compound, N⁴,N⁸-bis(dibenzosuberyl)spermine (7), having a K_i value of 0.26 μM. However, compound (7) did not prolong the lives of mice infected with trypanosomes. Analysis of docking studies indicated: the tricyclic groups of inhibitors bind at four distinct hydrophobic regions in the active site of TR; the importance of the chlorine substituent of clomipramine in binding to TR; and binding of the dibenzosuberyl groups of (7) occur at separate and distinct hydrophobic regions within the active site of TR.

Full text of DOI:10.1016/j.bmc.2015.01.018

Accepted Manuscript
Dibenzosuberyl substituted polyamines and analogs of clomipramine as effec-
tive inhibitors of trypanothione reductase; molecular docking, and assessment
of trypanocidal activities
Mary C. O’Sullivan, Timothy B. Durham, Hannah E. Valdes, Kelly L. Dauer,
Nicholas J. Karney, Andrew C. Forrestel, Cyrus J. Bacchi, Jerome F. Baker
PII:
S0968-0896(15)00034-6
DOI:
http://dx.doi.org/10.1016/j.bmc.2015.01.018
BMC 12021
Reference:
Bioorganic & Medicinal Chemistry
To appear in:
Received Date:
Revised Date:
Accepted Date:
6 November 2014
4 January 2015
9 January 2015
Please cite this article as: O’Sullivan, M.C., Durham, T.B., Valdes, H.E., Dauer, K.L., Karney, N.J., Forrestel, A.C.,
Bacchi, C.J., Baker, J.F., Dibenzosuberyl substituted polyamines and analogs of clomipramine as effective inhibitors
of trypanothione reductase; molecular docking, and assessment of trypanocidal activities, Bioorganic & Medicinal
Chemistry (2015), doi: http://dx.doi.org/10.1016/j.bmc.2015.01.018
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.

Title:
Dibenzosuberyl substituted polyamines and analogs of clomipramine as effective inhibitors of
trypanothione reductase; molecular docking, and assessment of trypanocidal activities.
Authors:
Mary C. O’Sullivan,a,* Timothy B. Durham,b Hannah E. Valdes,a Kelly L. Dauer,a Nicholas J.
Karney,a Andrew C. Forrestel,a Cyrus J. Bacchi,c Jerome F. Baker.a
Affiliations:
a) Department of Chemistry and Biochemistry, Canisius College, 2001 Main Street, Buffalo,
NY 14208, USA.
b) Department of Chemistry, Indiana State University, Terre Haute, IN 47809, USA.
c) Haskins Laboratories and Department of Biology, Pace University, New York, NY 10038,
USA.
Corresponding author*
Mary C. O’Sullivan,
Department of Chemistry and Biochemistry,
Canisius College,
2001 Main Street,
Buffalo, NY 14208, USA.
osulliv1@canisius.edu
Phone: (716) 888-2352
Fax: (716) 888-3112
Keywords:
trypanothione reductase; trypanothione; trypanosome; polyamine; spermine; spermidine;
Trypanosoma cruzi; dibenzosuberyl substituted polyamine; clomipramine
Abstract:
Trypanothione reductase (TR) is an enzyme critical to the maintenance of the thiol redox balance
in trypanosomatids, including the genera Trypanosoma and Leishmania that are parasites
responsible for several serious diseases. Analogs of clomipramine were prepared since
clomipramine is reported to inhibit TR and cure mice infected with trypanosomes, however its
psychotropic activity precludes its use as an anti-trypanosomal therapeutic. The clomipramine
analogs contained a tricyclic dibenzosuberyl moiety. Additionally a series of polyamines with
N-dibenzosuberyl substituents were prepared. All compounds studied were competitive
inhibitors of TR and showed trypanocidal activities against T. brucei in vitro. The analogs of
clomipramine were poor inhibitors of TR, whereas the polyamine derivatives were effective TR

inhibitors with the most potent compound, N4,N8-bis(dibenzosuberyl)spermine (7), having a Ki
value of 0.26 µM. However, compound (7) did not prolong the lives of mice infected with
trypanosomes. Analysis of docking studies indicated: the tricyclic groups of inhibitors bind at
four distinct hydrophobic regions in the active site of TR; the importance of the chlorine
substituent of clomipramine in binding to TR; and binding of the dibenzosuberyl groups of (7)
occur at separate and distinct hydrophobic regions within the active site of TR.
Abbreviations:
TR, trypanothione reductase;
GR, glutathione reductase;
DBS, dibenzosuberyl, 10,11-dihydro-5H-dibenzo[a,d]cycloheptenyl;
DBA, 3-chloro-10,11-dihydro-dibenzo[b,f]azepinyl;
T(SH)2, trypanothione dithiol;
T(S)2, trypanothione disulfide;
BR, binding region;
BM, binding mode.
1. Introduction
Several organisms belonging to the family Trypanosomatidae are parasitic protozoa
which cause severe diseases including human African trypanosomiasis (Trypanosoma brucei
sub-species), Chagas disease (Trypanosoma cruzi), cutaneous leishmaniasis (including
Leishmaniasis tropica and L. major) and visceral leishmaniasis (including L. donovani and L.
infantum).1 Diseases caused by trypanosomatids have a serious impact in many countries. For
example: in Latin America nearly 8 million people are afflicted with Chagas disease; and 1.3
million new cases of leishmaniasis occur each year, primarily in Asian, African and Latin
American countries.2 Current drugs to treat diseases caused by these protozoa have serious
problems including severe toxicity, and there are increasing incidences of drug resistant
trypanosomatid strains.3-5
One promising target in the development of new chemotherapeutics is the antioxidant
defenses of trypanosomatids that rely on an unusual spermidine-glutathione conjugate named
trypanothione [N1,N8-bis(glutathionyl)spermidine].6-8 In most organisms, glutathione is an
important anti-oxidant and levels of this thiol are maintained by the action of glutathione
reductase (GR).9, 10 However, trypanosomatids have a simpler redox metabolism than mammals
and do not contain GR, glutathione peroxidase, or catalase.8, 11 Trypanosomatids do contain
glutathione, however, trypanothione [T(SH)2] is the central thiol in their redox metabolism.

Trypanothione serves as the reducing agent for a myriad of reactions, including the reduction of
tryparedoxin (which is the reductant used by trypanosomatids for the ribonucleotide reductase
mediated reduction of ribonucleotides).7, 11, 12 These reactions occur with the concomitant
oxidation of T(SH)2 to the trypanothione disulfide [T(S)2]. The dithiol (T(SH)2) is regenerated
through the action of trypanothione reductase (TR) which utilizes NADPH to reduce T(S)2.13-16
Compounds that disrupt the maintenance of levels of T(SH)2 either through conjugation
with T(SH)2, inhibition of the synthesis of polyamines and other components of T(SH)2, or
inhibition of TR, have potential to be effective and selective anti-trypanosomal agents.17-19
Indeed several current anti-trypanosomal drugs impact levels of T(SH)2 in trypanosomatids,
although this may not be the primary therapeutic action of these drugs.3, 4, 20 For example: D,L-α-
difluoromethyl ornithine (DFMO, used to treat African trypanosomiasis) inhibits the synthesis of
putrescine (1,4-diaminobutane) a component of T(SH)2;21 pentamidine (used to treat
leishmaniasis) inhibits polyamine transport in L. infantum and T. cruzi22, 23; and trivalent
melaminophenyl-arsenicals (used to treat African trypanosomiasis) as well as antimonials (used
to treat leishmaniasis) impact several sites within the parasites including inhibition of TR and
formation of adducts with T(SH)2.20, 24-26
Several factors make TR (EC 1.8.1.12) in particular a compelling target for the
development of new chemotherapeutics.14, 15, 18, 27 These factors include that TR is pivotal to the
redox metabolism of trypanosomatids, as it appears to be the only route for the transfer of
reducing equivalents from NADPH to thiol-containing species.7 Also, several genetic studies on
Leishmania and Trypanosoma have shown the importance of TR, including studies in which
trypanosomatids modified to produce decreased levels of TR were found to be acutely sensitive
to oxidative stress with severely compromised abilities to infect cells.28-31 Additionally, the
structure and properties of TR from different genera of trypanosomatids are very similar,16, 32-34
indicating that an inhibitor of TR may be active against a range of trypanosomatids.
TR is a FAD and NADPH-dependent disulfide oxidoreductase with many similarities to
mammalian GR, both being homodimeric flavoproteins of similar size with essentially identical
mechanisms and with approximately 40% sequence identity.13, 35 However, TR and GR have
quite different substrate specificities, TR does not reduce glutathione disulfide and GR does not
reduce T(S)2.36, 37
A range of structurally diverse inhibitors of TR have been developed (for reviews see14,
15, 17), with the non-metal inhibitors generally including an aromatic and an amino group (that is
protonated at physiological pH). Several of these TR inhibitors are also good trypanocides in
vitro (i.e., kill trypanosomatids in buffer solutions). However, there are sparse reports of
compounds designed to be inhibitors of TR that are also active in vivo, i.e., that clear parasitemia
in mice15, 38 (although as mentioned several compounds that cure mammals also inhibit TR).
Thus, we were intrigued by reports that certain tricyclic psychotropics with pendant
amino moieties are inhibitors of TR,39, 40 display trypanocidal activities in vitro and that one of
these compounds, clomipramine (1), was able to cure mice infected by trypanosomes.
Specifically, the phenothiazine antipsychotics chlorpromazine, thioridazine and trifluoperazine16,

37, 41 and quinacrine (an antiprotozoal)16, 42, 43 are inhibitors of TR. One of the most potent of
these TR inhibitors is clomipramine (1), with a Ki value of 6.5 µM against T. cruzi TR41 and an
IC50 value of 3.4 µM against T. brucei TR (at 10.4 µM T(S)2).16 Also, certain tricyclic drugs
including clomipramine and some phenothiazines (including thioridazine and trifluoperazine) are
trypanocidal, being effective against species of Trypanosoma and Leishmania in vitro.39, 44-48
Additionally, clomipramine is active against trypanosomes in vivo and has been shown to cure
mice with both acute and chronic infections of T. cruzi (with clomipramine administered i.p. at
5mg/kg/day, and in one study co-administered with benznidazole).49-54 Clomipramine probably
impacts several processes in trypanosomatids, and it has been shown to have anti-calmodulin
activity55 in addition to inhibiting TR.
A range of novel derivatives based on the tricyclic drugs imipramine,40 and promazine56
have been synthesized and investigated against T. cruzi TR. Additionally, other amines with
tricyclic groups have been prepared and shown to inhibit TR, including phenothiazines,37, 56, 57
acridine derivatives42, 58-61 and methylene blue.62 Several of these tricyclic derivatives have
trypanocidal activities.56, 57, 59-62 However, clomipramine (1) is still one of the more potent
inhibitors of TR and to the authors’ knowledge it is the only non-metal containing inhibitor of
TR that also shows good activity against trypanosomatids in vivo without co-administration with
other drugs. (The TR inhibitor ebsulfur when administered with nifurtimox is active in vivo.38)
For these reasons we sought to investigate the interactions of analogs of clomipramine (1) with
T. cruzi TR. However, recognizing that the psychotropic activity of (1) would make it an
unlikely candidate for the treatment of trypanosome infections in humans, we chose to prepare
analogs in which the tricyclic ring of clomipramine (3-chloro-10,11-dihydro-
dibenzo[b,f]azepinyl, abbreviated as DBA) was replaced by the dibenzosuberyl group (10,11-
dihydro-5H-dibenzo[a,d]cycloheptenyl, abbreviated as DBS) that has a very similar shape (albeit
different electronic character) and importantly will have a different pharmacological profile from
clomipramine (1).
Additionally, we investigated certain polyamine-DBS conjugates, our rationale being
primarily based on previous studies of polyamine derivatives with aryl groups as being effective
inhibitors of TR several of which also displayed trypanocidal activities.64-76 Other intriguing
factors influenced us including: evidence that conjugation of biologically active moieties to
polyamines can potentiate their pharmacological profiles;77 further reports of polyamine-aryl
conjugates with activities against a range of trypanosomatids;63, 78-80 and the potential of
polyamine-DBS conjugates to exploit trypanosomatid polyamine transporters to become
accumulated into parasites81-84 (except for African trypanosomes (T. brucei) which do not readily
assimilate exogenous polyamines85). Consequently we also prepared spermidine and spermine
derivatives with N-substituted DBS groups.
In this paper we report the syntheses of these novel analogs of clomipramine and DBS-
polyamine derivatives, their inhibiting activities against T. cruzi TR, and a comprehensive
molecular docking study of their binding to TR. Three of the compounds studied (compounds
(5), (6) and (7)) were very good competitive inhibitors of TR with Ki values of 7.6, 4.0 and 0.26

µM, respectively. Additionally, we describe the activities of these compounds against T. brucei
subspecies in vitro and the activities of compounds (6) and (7) against T. b. brucei in vivo.
2. Results and discussion
2.1. Synthesis
The synthetic strategies used to prepare the analogs of clomipramine (compounds (2) and
(3)) and the DBS-polyamine conjugates (compounds (4) to (7)) are shown in Scheme 1.
Compounds (2) and (3) were prepared by reaction of N,N-dimethyl-1,2-diamine or N,N-
dimethyl-1,3-diaminopropane with 5-chlorodibenzosuberane in the presence of triethylamine.
Previously we reported the selective trifluoroacetylation of the primary amino groups of
polyamines 86 and use of this reaction enabled the syntheses of compounds (4) to (7) in three
steps from their respective polyamines. Thus, the primary amino groups of each polyamine were
trifluoroacetylated, followed by N-alkylation with 5-chlorodibenzosuberane, and then removal of
the trifluoroacetyl groups.
2.2. Enzyme studies
The effects of compounds on the reduction of T(S)2 by T. cruzi TR were measured using
a standard photometric assay.13 The TR used was isolated from E. coli SG5, a glutathione
reductase deletion mutant, containing the TR expression vector pIBITczTR.87, 88 All of the
compounds exhibited linear competitive inhibition of the reduction of T(S)2 by TR. The
inhibition constants (Ki values) of compounds are presented in Table 1. Our measured Ki value
for (1) of 8.40 1.43 µM correlated well with the value of 6.53 µM reported by Benson.41
The relative specificity of compounds (1) to (7) as inhibitors of TR was explored by
investigating their effects on the reduction of glutathione disulfide by yeast GR. The presence of
these compounds at concentrations of 250 µM did not decrease the rate of this reaction catalyzed
by GR. Hence these compounds show specificity for the inhibition of TR with respect to the
structurally similar enzyme, GR. This was anticipated since although TR and GR have structural
similarities, the active sites of these enzymes have several differences. The active site of TR is
large, contains additional glutamate residues (including Glu19) and hydrophobic regions,
whereas GR has a smaller active site containing several cationic residues (e.g., Arg37, 38 and
397) that are not present in TR.89 The active site of TR can therefore accommodate large,
cationic compounds and charge is known to be a factor in the selectivity of compounds for TR
versus GR.37 However charge is not the sole factor in the binding of compounds to TR, since the
polyamine spermine does not inhibit TR.65
There is previous kinetic evidence for the binding of more than one inhibitor molecule to
the large active site of TR.58 Additionally, crystal structures of TR with bound quinacrine

mustard,43 and TR with a diarylpyrrole32 show two molecules located in the active site. Our
enzyme studies did not show evidence of the binding of more than one inhibitor molecule.
However, the maximum inhibitor concentrations used in our TR enzyme assays were
approximately twice the Ki, and evidence for the binding of multiple molecules is more
pronounced at higher inhibitor concentrations.58 Thus, our results do not preclude the possibility
of more than one inhibitor molecule being bound to the active site of TR at higher inhibitor
concentrations.
All the compounds investigated were competitive inhibitors of T(S)2 reduction by TR.
Compounds (2) and (3) which are structurally the most similar to clomipramine (1) were
significantly less effective inhibitors than (1). The size and the concave shape of the DBA ring
of compound (1), and the DBS tricyclic ring (of compounds (2) to (7)) are very similar however,
the presence of the chlorine atom of DBA is obviously important in the binding of (1) to TR.
Our docking studies (section 2.4) indicate that the chlorine is involved in binding interactions
with residues in the active site, also there are several examples of TR inhibitors containing a
halogenated aryl group being more effective than analogs without the halogen.34, 56, 58, 67 Another
possible factor impacting the inhibiting activities is that (2) and (3) contain an exocyclic nitrogen
(versus the endocyclic nitrogen of the DBA group of (1)). However, since the shapes of the DBS
and DBA rings are similar and docking (section 2.4) showed that both tricycles bind to the same
locations in the active site, any effects on binding resulting from differences in ring nitrogen
location between compounds (1) to (3) are likely to be small compared to binding differences
due to the presence of chlorine in the DBA ring of (1).
The polyamine derivatives, compounds (4) to (7), were more effective inhibitors than (2)
and (3), thus the presence of an additional amino/aminium group enhanced the inhibiting
activities of compounds. The inhibiting activities increased with increased alkyl chain length,
i.e., increasing distance between the DBS group and aminium ion. Docking studies (section 2.4)
indicate that the longer chain enables the aminium groups to reach H-bond accepting and
glutamate residues while still permitting the DBS to be fully sited in one of the tricyclic ring
binding regions. The DBS substituted spermine derivative (7) was the most effective inhibitor
studied with a Ki value of 0.255 µM against T(S)2 reduction by TR. Thus compound (7) is one
of the most potent competitive inhibitors of TR to date. The presence of two DBS rings and four
aminium ions in (7) obviously can lead to more favorable hydrophobic and ionic interactions
with residues in the active site of TR. However, our docking studies (section 2.4) indicate why
the interaction of (7) with TR is so much more favorable than the spermidine derivative (6) (as
reflected by the 20-fold difference in Ki values of these two compounds).
2.3. Trypanosome studies
The in vitro trypanocidal activities of compounds (2) to (7) against bloodstream forms of
clinically isolated strains of T. brucei ssp. were examined using a standard growth screen.90 The
trypanosome strains used were T. b. brucei Lab 110 EATRO and three clinical isolates of T. b.

rhodesiense and these results are shown in Table 2. The compounds tested were effective
trypanocides with IC50 values ≤ 4.5 µM against three of the T. brucei ssp. One strain, K 243-As-
10-3 (a pentamidine and melarsoprol-resistant clone of T. b. rhodesiense), was consistently more
resistant to the compounds tested. All the compounds studied had similar, or marginally better,
activities against T. brucei as clomipramine (1).39 The spermidine and spermine derivatives (6)
and (7) were slightly less effective trypanocides (despite being more potent TR inhibitors) than
analogous polyamines with N-(3-phenylpropyl) substituents.64 However, in this earlier work the
polyamines with multiple 3-phenylpropyl substituents were less effective trypanocides, despite
being more potent TR inhibitors.64 These polyamines, as well as compound (7), are poorly
soluble in water and this may affect the accuracy of the trypanosome assays of these compounds.
Our compounds were assessed against T. cruzi TR, whereas the trypanosomes studied
were T. brucei ssp. However differences in the TR enzymes from these species should not
impact our studies, since comparison of the crystal structures and kinetic properties of TR from
both T. cruzi and T. brucei have shown no significant differences.16 Thus TR from both species
are considered to be interchangeable in structure-based inhibitor design.16 Overall we observed
no direct correlation between the TR inhibiting and trypanocidal activities of compounds (2) to
(7). An apparent lack of correlation between the TR inhibiting and trypanocidal activities of
compounds has been noted in previous studies.68,75 This suggests that structural features in
addition to TR binding, for example, features that facilitate the transport of compounds across
the trypanosome cell membrane, are also important to the trypanocidal efficacies of compounds.
The in vivo trypanocial activities of the most potent TR inhibitors, compounds (6) and
(7), were investigated by administration of compounds to mice infected with T. b. brucei LAB
110 EATRO. After the infection had progressed for 24 h, compound (6) was administered ip at
1, 2.5, 5, 10, 25 mg per kg of body weight once per day for three days. For compound (7), the
drug was administered at several doses ip both twice daily and once a day. Additionally,
compound (7) was administered at 10, 25 or 50 mg/kg/day using mini-osmotic pumps that
release drugs continuously for three days. Compound (6) did not prolong the life of animals
beyond the survival time of the control animals. Also compound (6) showed toxicity at 25
mg/kg with infected animals on this dose dying before untreated, infected animals. Compound
(7) displayed a slight prolongation of life. Specifically, some infected animals lived a day
beyond controls when dosed at 50 mg/kg/day administered in 2 daily doses of 25 mg/kg for three
days. However, compound (7) showed toxicity against infected animals that were dosed at 50
mg/kg once daily for three days. Thus, with these dosing regimens, the dosage of (7) needed to
impact the parasite is very close to the dosage resulting in toxicity to the animal.
2.4. Molecular docking studies
2.4.1. Preparation of ligands and enzyme for docking
Compounds (2) to (7) have multiple amino groups, thus several protonation states of each
compound are possible at physiological pH. Anticipating that different protonation states will

have different affinities for TR, we calculated the relative occurrence of protonation states at pH
7.2 (the pH of our TR inhibition assays) leading us to study the docking of one protonated form
of clomipramine (1) and a total of ten protonated forms of compounds (2) to (7) (Table 3). An
additional factor we considered was the possibility that different conformers of the DBS or DBA
tricyclic ring in compounds may impact their affinities for TR. Most docking programs treat
rings as rigid with no conformational flexibility, consequently we identified several low energy
conformations for the DBS and DBA rings (representative structures of the conformations are
shown in Table 4 and all structures investigated are provided in Supplemental Information).
Calculations of the energy barriers for interconversion of the conformations indicated that in
vitro at room temperature the DBS or DBA rings will readily interconvert between the
conformations of the ligand structures used in our docking studies (although in silico the rings
are not able to flex). To the authors’ knowledge, no previous docking study of TR using ligands
with flexible ring systems has investigated the impact of ring conformation on docking.
Consequently, this led us to investigate the docking (using AutoDock Vina91) of 33 structures
(ligands) of different versions of compounds (1) to (7) in various protonation states and ring
conformations.
There are several crystal structures of TR from various trypanosomatids including
Crithidia fasciculata (a trypanosomatid that infects insects),92-94 T. cruzi 33, 42, 43, 95, 96, T. brucei 16,
34, 96, and L. infantum. 20, 32, 97 Several of these structures have ligands located in the active site of
TR, including glutathionylspermidine,94 trypanothione, 32, 33 quinacrine,42 the covalently bound
quinacrine mustard,43 Sb(III),20 Ag(0)97 and more recently, structures with small molecule
inhibitors complexed to TR.32, 34, 96 For our studies we chose to use the crystal structure of
Trypanosoma cruzi TR in complex with trypanothione (RCSB Protein Data Bank code 1BZL),33
with removal of the trypanothione ligand from 1BZL before inhibitor docking studies. Using
1BZL allowed us to correlate docking results with experimental results of our inhibitors against
T. cruzi TR, and enabled the identification of any correspondence between regions of
trypanothione binding and regions of ligand (inhibitor) docking (see Fig. 1).
We utilized AutoDock Vina91 for these docking studies and defined the search space for
docking to include the entire active site cleft of TR, but not including the FAD or NADPH
binding regions. Since the FAD is tightly bound to TR, this site would not be available to
inhibitors in vitro, additionally our in vitro experiments showed that compounds (1) to (7) did not
compete with NADPH for binding to TR. Tables with the binding energies of the lowest energy
poses for each ligand structure along with descriptions of the binding sites of frequently
occurring poses for each ligand structure are provided in Supplemental Information.
2.4.2. Docking results
Compound (7) was the most potent inhibitor of TR and the binding energies of ligand
versions of (7) were lower than all other compounds. However, current docking programs are
unable to reliably predict inhibition constants of compounds,98, 99 thus we were unsurprised to
observe no direct correlation between the calculated binding energies of ligands and the

inhibition constants of compounds (1) to (6). Also, there was no apparent correlation between
protonation states or ring conformations of ligands with calculated binding energies. However,
docking programs (including AutoDock Vina) have more success at predicting the
experimentally observed binding modes (poses) of compounds within active sites.91, 100 Our
results showed a notable uniformity in the binding locations of the DBS/DBA groups of the 33
ligand structures docked, with the DBS/DBA rings clustered into four distinct binding regions
(BR 1 to 4) in the active site of TR (see Fig. 1). This result is striking since despite the large size
of the search space used for docking (that encompassed the entire active site cleft), the
DBS/DBA groups of the five lowest energy poses of all modeling runs were located almost
entirely to these four binding regions (BRs). Also, the four BRs are clustered in, and overlap
with, the space where trypanothione binds (Fig. 1).33
Representative poses for each compound are shown in Figures 2 and 3. All 33 ligand
structures had poses in BR 1 and 2 and most ligands also had poses in BR 3. However, only
ligands for compound (7) and occasional poses of compound (4) had DBS groups occupying BR
4. (The Supplemental Information provides populations of poses in each BR (Table 5), more
detailed descriptions of ligand poses, and information on the binding configurations of low
energy poses for each of the 33 ligand structures.) The ring conformation of a ligand clearly
affected the number of poses that occupied each BR. However, there was no obvious correlation
between ring conformation and relative populations in individual BRs, clearly other aspects of
the ligand structures, including protonation state and distance from ring to protonated nitrogen,
also impact the docking location of the DBS/DBA rings.
2.4.3. Description of DBS/DBA ring binding regions
The active site of TR is a large cleft (approximately 30 Å deep and 20 Å wide) composed
of residues from both subunits of the homodimer (Fig. 1). The active site has several locations
where ligands can bind40, 96 and for some inhibitors is large enough to accommodate two
molecules.32 The residues directly involved in catalysis (including the redox active Cys53 and
Cys58, along with His461’) are positioned at the deepest part of the cleft, and close to the dimer
interface (note, in one subunit the residues are labeled with primed numbers). At the interface of
the subunits there is a tunnel that connects both active sites, which has a wider volume at the
center of the enzyme, i.e., an interface pocket.75 The tunnel entrance is close to the catalytic site
(entrance is next to His461’), and has a diameter of approximately 8 Å (entrance is lined by
Leu399’, Lys62, Pro462’, Thr463’, Glu467’, Phe396’ and Pro398’). This structural feature (i.e.,
tunnel with interface pocket) has been reported in GR101 and occurs in several additional FAD-
dependent pyridine nucleotide-disulfide reductases, including glutathione amide reductase (PDB
code 2R9Z), thioredoxin glutathione reductase (2V6O) and thioredoxin reductase (1H6V). For
GR, crystal structures show that certain tricyclic inhibitors bind in the interface pocket/cavity,
including safranin (an acridine derivative),102 a xanthene101 and an isoalloxazine103 derivative.
Despite the interface pocket in TR being smaller than that of GR, it has been speculated that this
site may be a binding location for inhibitors of TR.58 75 However, this was not explored in these

docking studies of TR, which included in the search space the start of the tunnel but not the
interface pocket.
The locations of the four DBS/DBA binding regions are shown in Figure 1. These sites
overlap with regions where T(S)2 binds (from the crystal structure of T. cruzi TyrR with bound
T(S)2 PDB ID # 1BZL33). The DBS/DBA tricyclic rings have a concave shape that enable the
ring to maximize interactions with residues that comprise bulges/protuberances into the active
site, and with residues lining pockets (or the tunnel entrance) of the active site.
• Binding region 1 (BR 1) is a hydrophobic surface where Ile339 creates a protuberance
around which the DBS or DBA tricyclic ring wraps. Depending on the location of the ligand
pose in BR1, the DBA/DBS ring may also interact with Pro336 or Asn340. BR 1
corresponds to the region where the γ-Glu-I of T(S)2 binds (the γ-Glu-I interacts with several
residues including Ile339 and Pro33633).
• Binding region 2 (BR 2) occurs at the entrance of the tunnel that leads to the interface pocket.
BR 2 overlaps with the previously identified “Z-site”.56 The DBA/DBS tricyclic ring
partially inserts into the tunnel (an aryl group and part of the seven-membered ring is
inserted) and the main residues of the tunnel entrance that interact with the DBS/DBA are
Leu399’, Pro462’ and His461’. For compounds (5) to (7) (i.e., those with longer alkyl
chains) often the phenyl group that is not inserted into the tunnel interacts with Lys402’ (via
a cation-π interaction) which is sited above the tunnel entrance. BR 2 is the location where
the γ-Glu-II of T(S)2 binds (the γ-Glu-II group is inserted into the entrance of the tunnel33).
• Binding region 3 (BR 3) is a shallow pocket with Tyr111 forming an outward bulge and
where the pocket is lined with the hydrophobic residues Leu18, Val54, Ile339 and also
Ser15. This site partially overlaps with the previously described “hydrophobic wall”.41, 56
The tricyclic ring of compounds is oriented in the pocket so that it curves around Tyr111
with a polar-π interaction generally occurring between the Tyr111 hydroxyl hydrogen and
one of the aryl groups of the tricycle. BR 3 corresponds to the region where the Cys-I and
Gly-I groups of T(S)2 bind.33
• Binding region (BR 4) is a hydrophobic region closer to the entrance of the active site
comprised primarily of Trp22 and Met114. This site also partially overlaps with the
previously described “hydrophobic wall”.41, 56 Only compounds (4) and (7) had poses where
the DBS occupied this site (within the five lowest energy poses of eight docking runs). At
BR 4, the DBS group is oriented with one aryl group involved in a sulfur-π interaction with
Met114.104-106 The DBS group is also located close to Trp22 with the π-faces of the aryl
groups at an approximate 30-40o angle from parallel. This site also has hydrophobic regions
from Met114 and Leu18 that interact with the DBS. BR 4 is the region where part of the
alkyl chain of the spermidine component of T(S)2 binds.33
Several early modeling studies refer to two regions of the TR active site where (tricyclic)
inhibitors may bind, described as the “Z-site” (Phe396’, Pro398’, Leu399’) and the “hydrophobic
wall/cleft” (Try110, Met113, Phe114, Trp21 and Leu17),56, 57, 107, 108 and early modeling studies

placed clomipramine at the “hydrophobic wall”.41 (Please note that literature citations of residue
numbers for T. cruzi TR can differ by one unit depending on source. For results presented in this
paper the residue numbering from 1BZL is used.) There is kinetic evidence for multiple binding
sites for tricyclic inhibitors of TR,40 and crystal structures with tricyclic moieties of ligands
located at the “Z-site” and “hydrophobic wall”.34, 42, 43 Two of the DBS/DBA binding regions
identified in this study occur within the “hydrophobic wall” (BR 3 and 4) and BR 2 overlaps
with the “Z-site” (although the tunnel is not mentioned in reports of the “Z-site”). Several
crystal structures show aromatic moieties of inhibitors bound at the “hydrophobic wall” of TR,
and interacting with Trp22 and Met114 (corresponding to BR 4), and often the inhibitor
molecule extends into the region near Tyr111 (BR 3). For example, crystal structures locate in
the region of Trp21/22 and Met113/114: the acridine rings of mepacrine (quinacrine)42 and
quinacrine mustard;43 the dihydroquinazoline ring of T. brucei TR inhibitors (with groups also
being close to Tyr110 and also in a conformationally induced sub-pocket);34 also biaryl
inhibitors96 and a diarylpyrrole 32 (both of which also extended into region near Tyr110).
Representative ligand poses compiled from all protonated versions and DBS/DBA
configurations of compounds (1) to (7) are shown in Figures 2 and 3. Tables of frequently
occurring ligand poses for each of the 33 ligand structures docked and more detailed descriptions
of ligand poses are provided in the Supplemental Information.
2.4.4. Description of ligand amino and aminium ion binding sites
The location of the ligands’ amino or aminium moieties depended on the binding location
of the DBS/DBA group and to a lesser extent by the ring conformation and protonation state of
the nitrogen. However, considering the lowest energy poses for all ligands of (1) to (7), the most
frequent binding site for amino/aminium groups was the region containing Glu466’ (ligand
interacting with carboxylate), Glu467’(carboxylate) and His461’(imidazole N δ). These three
residues are clustered together near the entrance of the tunnel (BR 2). Ligand amino/aminium
groups in this location were in general involved in multiple strong ion-pair and/or H-bond
interactions to both His461’ and Glu466’; or to both Glu466’ and Glu467’; or to both Glu467’
and Ser470’ hydroxyl; or occasionally to both Glu467’ and Ser394’ hydroxyl.
The next most populated sites were with the carboxylate of Glu19 (sited near Ile339 and
BR1); and the hydroxyl or backbone oxygen of Ser470’. Less frequently, amino groups of
ligands interacted with the guanidinium of Arg472’, and amino/aminiums interacted with Ser110
(hydroxyl or backbone oxygens), or Tyr111 (hydroxyl). Several previous modeling and x-ray
studies have noted the importance of electrostatic interactions between ligand amines and Glu19
or Glu466’/Glu467’.17, 34, 42, 57,70
2.5. Discussion of interactions of compounds with TR
Figures 2 and 3 show low energy poses that are representative of those that commonly
occurred from the collection of all the protonated and ring conformation versions studied for
each compound. The common poses that are not shown differ primarily in the locations of the

amino/aminium groups located at the ends of the flexible alkyl chains (further descriptions and
tables of the frequent binding locations of poses from each of the 33 docked ligands are provided
in the Supplemental Information).
For clomipramine (1), the DBA ring was able to dock in BR 1, 2 or 3, and the chlorine
atom was often involved in favorable interactions helping to explain the lower Ki value of (1)
compared to analogs (2) and (3). Thus, in poses of (1) where the DBA group is located in BR 1,
the chlorine is able to H-bond to the hydroxyl of Tyr111. When the DBA is in BR 2 with the
chlorine inserted into the tunnel, the chlorine is able to accept a H-bond from the amino group of
Lys62’ (N zeta); and when the chlorine is positioned out of the tunnel, it can H-bond with the
hydroxyl of Tyr111. Also, when the DBA is located in BR 3, the chlorine was able to interact
with the sulfur of Met114, or the π system of Trp22.109
For the polyamine derivatives (4) to (7) there was a wider range of possible binding
interactions for the terminal aminium ions, than occurred with compounds (2) and (3). This is
consistent with compounds (4) to (7) being more effective inhibitors of TR than (2) and (3).
Compound (7) was a significantly more potent inhibitor than compounds (1) to (6).
Docking studies indicated that both the DBS groups of (7) were able to bind to separate binding
regions (BR 1 to 4) and the terminal aminiums were able to form H-bond and/or ion-pair
interactions with residues (see Fig. 3). There was a striking uniformity in the bound locations for
compound (7) ligands, with three distinct binding modes into which 90% of the lowest energy
poses bound. The most frequent binding mode (BM 1) had one DBS occupying BR 1 (Ile 339)
and the other DBS occupying BR 4 (Trp22); in the second mode (BM 2) one DBS occupied BR
2 (tunnel entrance) and the other DBS was in BR 3 (Tyr111); the least frequently occupied mode
(BM 3) had one DBS in BR 1 (Ile339) and the other DBS in BR 2 (tunnel entrance). (Data
showing the population of poses in each BM and relative energy ranking of poses is provided in
the Supplemental Information.) Generally for (7) one terminal aminium was associated with the
Glu466’, 467’ and His461’ cluster; and the second terminal aminium was H-bonded to the
Ser110 hydroxyl, Glu 19 or Ser470’ hydroxyl (the aminiums adjacent to DBS had no apparent
interaction with active site residues). In addition to interactions with active site residues, several
intramolecular interactions occurred in these binding modes of (7). In BM 1 and 2 the two DBS
groups were positioned so that an aryl group from each DBS had an offset edge to face
interaction.110 Also, in BM 1 one terminal aminium ion is involved in a cation-π interaction with
the DBS located in BR 1 (Ile339), in addition to the aminium interacting with both Glu467’ and
Ser470’.
The modeling results of compound (7) provide insights as to why (7) is a significantly
more potent inhibitor of TR than the other polyamine derivatives in this study. Compound (7)
can bind in three highly favored configurations (binding modes) in the active site. In all these
modes, each DBS group is accommodated into separate hydrophobic areas of the active site,
while enabling the terminal aminiums to have strong H-bond and/or ion pair interactions with
active site residues. Furthermore, additional intramolecular interactions (cation-π and/or
aromatic offset edge to face interactions) further stabilize the bound conformations of (7).

Conventional design strategies for improved inhibitors aim to limit the conformational flexibility
of molecules (to minimize the entropy decrease on inhibitor-enzyme binding). Although
compound (7) (and the other polyamine derivatives studied) have a high degree of
conformational flexibility, it is likely that intramolecular cation-π interactions (and possibly aryl
to aryl interactions) would transiently occur in solution limiting the conformational flexibility of
(7) resulting in an energetically more favorable binding to TR than anticipated.
3. Conclusions
In the search for new chemotherapeutics to treat diseases caused by infections with
Trypanosoma and Leishmania, the drug target trypanothione reductase (TR) has several desirable
qualities, including TR’s crucial role in the redox metabolism of a range of parasites from the
Trypanosomatidae family. However, TR presents challenges for the development of selective
potent inhibitors, partly due to the large size of its active site cleft. Additionally, despite the
range of TR inhibitors developed, there are very few TR inhibitors that are active against
trypanosome infections in vivo.
In this study the inhibiting effects on T. cruzi TR of two novel DBS-containing analogs of
clomipramine (1) were investigated along with four novel polyamine derivatives containing DBS
groups. Clomipramine was chosen as an initial design focus since it inhibits TR and additionally
cures mice with T. cruzi infections, however its psychotropic activity precludes its development
as a treatment for trypanosome infections. The rationale for the investigation of polyamine-DBS
derivatives was influenced by several factors including previous reports of polyamines with aryl
groups being inhibitors of TR and having in vitro trypanocidal activities.64-75 The compounds
investigated were competitive inhibitors and selective for TR, in that they did not inhibit GR.
The compounds also displayed trypanocidal activities in vitro, however the two most effective
TR inhibitors (compounds (6) and (7)) were not active against trypanosomes in vivo and were
unable to increase the life-span of mice infected with T. brucei. The lack of in vivo trypanocidal
activities for compounds (6) and (7) may be caused by several factors including rapid
mammalian excretion or metabolism of these compounds. For example, the amino groups of (6)
and (7) may be substrates for a variety of mammalian amine oxidases and polyamine
acetyltransferases.111, 112 Also the aminopropyl group of (7) may be a substrate for mammalian
spermine oxidase (mammalian acetylpolyamine oxidase, formerly known as polyamine oxidase
appears to have low activity for non-acetylated polyamines).111, 113 Since (6) and (7) are
reversible inhibitors of TR, if concentrations of these compounds are not maintained in vivo, TR
activity will not be significantly decreased.
Also conducted was a comprehensive docking analysis using several protonation states
and low energy ring conformations of compounds (1) to (7), i.e., including the predominant
forms of compounds (1) to (7) that are likely to occur at physiological pH. This approach was
necessary to fully explore the likely in vitro interactions between these compounds and TR.

Although several docking studies of TR with amines containing ring moieties have been
reported, to the authors’ knowledge there is no study that accounted for compounds’ differing
protonation states and ring conformations, even though these factors are likely to impact TR-
inhibitor interactions. Our studies indicated that the tricyclic DBS/DBA rings of compounds
cluster into four distinct binding regions, which overlap with the binding site of the substrate
T(S)2. The concave shapes of the DBS/DBA rings enable them to curve around residues and fit
into pockets maximizing hydrophobic interactions. It was evident that similarly-sized planar
ring systems would have more limited hydrophobic interactions, and this may explain the 20-fold
lower Ki values of (7) and (6) relative to analogs in which the DBS group is replaced with
naphthylmethyl groups.65, 114 Docking studies also indicated several residues in the active site
that participated in H-bond and/or ion pair interactions with the amino/aminium moieties of
inhibitors. Commonly, these included multiple interactions with the Glu466’/Glu467’/His461’
cluster of residues; or interactions with the Glu19 carboxylate, or Ser470’ (with the hydroxyl or
backbone oxygen).
The two clomipramine analogs (2) and (3) were significantly less effective inhibitors of
TR than clomipramine (1), and docking indicated that although these compounds bind to similar
sites in TR, the chlorine atom of (1) provides additional binding interactions not possible in the
non-halogenated DBS rings of (2) and (3). This observation is consistent with previous studies
in which compounds containing halogenated aryl groups were better TR inhibitors than non-
halogenated analogs.34, 56, 58, 67 The polyamine derivatives (4) to (7) were more effective
competitive inhibitors of TR than compounds (2) and (3), and the most effective compound was
the spermine derivative (7) which, with a Ki value of 0.26 µM, is one of the more potent
inhibitors of TR known. Other spermine derivatives previously reported to be potent inhibitors
of TR include ones substituted with indoles (noncompetitive inhibitors with Ki values of 0.076,
to 0.39 µM),67 3-phenylpropyl groups (Ki 0.15 µM and greater),64 or diphenyl groups (Ki 0.4
µM).69 Since TR has a large active site with multiple hydrophobic binding sites and several H-
bond acceptors, a range of compounds in addition to spermine derivatives are also effective
inhibitors of TR (with Ki or IC50 values of 0.5 µM or less). These include a dichlorobenzyl
substituted alkylammonium phenothiazine,57 a dimethylamine substituted naphthoquinone,75
certain large bis(2-aminodiphenylsulfide) derivatives,115 a substituted dihydroquinazoline,34 and
an imidazole and phenylsulfide substituted BTCP derivative.116 However, all these potent
inhibitors of TR including compound (7), contain two or more aromatic moieties as well as an
amino group that will be protonated at physiological pH. Our docking studies provide insights
into how these structural features enhance binding to the active site of TR. Specifically, docking
of compound (7) showed three low energy binding modes for (7) in which the two DBS groups
interact with separate, distinct hydrophobic regions of the active site.
Thus, our detailed analysis of the binding of compounds (1) to (7) provides useful
information that can aid the design of more effective and selective inhibitors of TR, and
specifically, aid the design of amines with two or more aromatic moieties to exploit binding at
two of the hydrophobic regions of the active site. Our docking studies identified four distinct

hydrophobic binding regions in the active site that were accessible to the tricyclic rings of
compounds (1) to (7). Additionally, the enzyme kinetics and in silico studies presented indicate
that the concave-shaped, aryl-containing ring systems of DBS/DBA are able to bind more
effectively than planar naphthyl groups to these hydrophobic regions. The results also provide
further evidence that incorporation of a halogen into aryl systems may increase the binding
affinity of compounds to TR.
4. Experimental
4.1. Synthesis information
The trifluoroacetate salts of N1,N5-bis(trifluoroacetyl)diethylenetriamine (8), N1,N8-
bis(trifluoroacetyl)spermidine (10) and N1,N12-bis(trifluoroacetyl)spermine (11) were prepared as
previously reported.86 All other reagents were purchased from commercial sources. Acetonitrile
and triethylamine were dried by distillation over calcium hydride under N2 and were stored over
4 Å molecular sieves. The ammonium hydroxide used contained 29.9% NH3. Thin-layer
chromatography was carried out using silica gel (250 µm layer) and compounds were visualized
by UV light, ninhydrin in ethanol, or phosphomolybdic acid in ethanol. Column
chromatography was carried out under pressure (flash chromatography) using silica gel (40 µm).
Melting points are uncorrected. NMR spectra were obtained at room temperature using either a
1
Bruker DPX 250MHz or Agilent 400MHz MR DD2 spectrometers. H NMR spectra were
acquired either at 250 or 400 MHz and 13C NMR at 62.9 or 100.5 MHz. NMR samples were
dissolved in CDCl3 with TMS as an internal standard unless otherwise indicated. Mass spectra
were obtained on a Waters 70-VSE (FAB MS), or a Thermo Scientific LCQ Fleet ion trap LC-
MS (ESI MS).
4.1.1. General procedures for the N-alkylation of amines with dibenzosuberyl groups
(10,11-dihydro-5H-dibenzo[a,d]cycloheptene, abbreviated as DBS)
To the trifluoroacetate salt of the amine (1.2 mmol) in acetonitrile (10 mL) at 0 oC was
added triethylamine (2.7 mmol) followed by 5-chlorodibenzosuberane (5-chloro-10,11-dihydro-
5H-dibenzo[a,d]cycloheptene) (1.4 mmol). The resulting solution was stirred overnight at room
temperature under N2. Saturated aqueous NaCl solution (30 mL) was added and the mixture
extracted with CH2Cl2 (3 x 30 mL). The organic layers were collected, dried (anhydrous
Na2SO4) and concentrated. Purification by column chromatography (1to 4% ethyl acetate in
CH2Cl2) afforded the desired compounds (12), (13) and (14).
For the N-alkylation of the amines N,N-dimethyl-1,2-diaminoethane or N,N-dimethyl-1,3-
diaminopropane, the above procedure was followed except 1.2 mol. equiv. of triethylamine was
used and purification by column chromatography (4 to 5% CH3OH in CH2Cl2) to afford
compounds (2) or (3), respectively.

4.1.2. N-dibenzosuberyl-N',N'-dimethyl-1,2-diaminoethane (2)
Light-colored oil; yield, 88%; 1H NMR: δ 7.25 (m, 2 H, aromatic H), 7.1 (m, 6 H, aromatic H),
4.8 (s, 1 H, cycloheptyl CHN), 3.65 (m, 2 H, cycloheptyl CH2), 2.9 (m, 2 H, cycloheptyl CH2),
2.6 (t, 2 H, J = 5.8 Hz, NCH2), 2.35 (t, 2 H, J = 5.8 Hz, NCH2) and 2.1 (s, 6 H, 2 CH3) ppm; 13C
NMR: δ 140.2, 139.5, 130.0, 128.8, 127.0 and 125.5 (2 x 6 aromatic C), 69.2 (cycloheptyl CHN),
58.7, 45.6, 45.0 (2 CH3) and 32.1 (2 cycloheptyl CH2) ppm; MS (ESI) m/z 281.02 (MH+).
4.1.3. N-dibenzosuberyl-N',N'-dimethyl-1,3-diaminopropane (3)
Light-colored oil; yield, 76%; 1H NMR: δ 7.25 (m, 2 H, aromatic H), 7.1 (m, 6 H, aromatic H),
4.8 (s, 1 H, cycloheptyl CHN), 3.65 (m, 2 H, cycloheptyl CH2), 2.9 (m, 2 H, cycloheptyl CH2),
2.55 (m, 2 H, NCH2), 2.3 (m, 2 H, NCH2), 2.15 (s, 6 H, 2 CH3), 1.6 (m, 2 H, CH2) ppm; 13C
NMR: δ 139.9, 139.3, 130.0, 128.5, 127.0 and 125.4 (2 x 6 aromatic C), 68.8 (cycloheptyl CHN),
57.4, 46.4, 44.7 (2 CH3), 32.0 (2 cycloheptyl CH2) and 27.2 (CH2CH2CH2) ppm; MS (FAB) m/z
295.2 (MH+); MS (ESI) m/z 295.05 (MH+).
4.1.4. N-[2-(trifluoroacetamido)ethyl]-N-dibenzosuberyl-N'-trifluoroacetyl-1,2-
diaminoethane (12)
Light-colored solid; yield, 76%; mp 114-115 oC; H NMR (CDCl3, 250 MHz): δ 7.15 (m, 8 H,
1
aromatic H), 6.8 (br. t., 2 H, 2 NH), 4.4 (s, 1 H, cycloheptyl CHN), 3.8 (m, 2 H, cycloheptyl
CH2), 3.1 (br. q, J = 6.0 Hz, 4 H, 2 CONHCH2), 2.8 (m, 2 H, cycloheptyl CH2), 2.7 (t, J = 6.2
Hz, 4 H, CH2NCH2) ppm; 13C NMR: δ 157 (q, 2JCF = 37.2 Hz, 2 CF3CO), 139.5, 138.3, 131.1,
131.0, 128.4 and 126.1 ( 2 x 6 aromatic C), 116 (q, 1JCF = 288 Hz, 2 CF3), 77.4 (cycloheptyl
CHN), 51.5, 38.2 and 31.6 (2 cycloheptyl CH2); MS (ESI) m/z 510.31 (M + Na+), 487.95
(MH+).
4.1.5. N-[3-(trifluoroacetamido)propyl]-N-dibenzosuberyl-N'-trifluoroacetyl-1,3-
diaminopropane (13)
Light-colored solid; yield, 69%; mp 124-125 oC; H NMR: δ 7.15 (m, 8 H, aromatic H), 6.1 (br.
1
s, 2 H, 2 NH), 4.3 (s, 1 H, cycloheptyl CHN), 3.9 (m, 2 H, cycloheptyl CH2), 3.2 (br. q, J = 6.7
Hz, 4 H, 2 CONHCH2), 2.8 (m, 2 H, cycloheptyl CH2), 2.5 (t, J = 6.7 Hz, 4 H, CH2NCH2), 1.55
(pentaplet, 4 H, J = 6.7 Hz, 2 CH2CH2CH2) ppm; 13C NMR: δ 157 (q, 2JCF = 36.9 Hz, 2 CF3CO),
139.5, 139.1, 130.9, 130.8, 128.1 and 125.8 ( 2 x 6 aromatic C), 116 (q, 1JCF = 288 Hz, 2 CF3),
76.5 (cycloheptyl CHN), 47.8, 38.0, 31.6 (2 cycloheptyl CH2) and 25.3 ppm; MS (ESI) m/z
538.15 (M + Na+), 515.85 (MH+).
4.1.6. N4-dibenzosuberyl-N1,N8-bis(trifluoroacetyl)spermidine (14)
1
Light-colored solid; yield, 85%; mp 87-89 oC; H NMR: δ 7.1 (m, 8 H, aromatic H), 6.8 (br. t, J
= 5.5 Hz, 2 H, 2 NH), 4.3 (s, 1 H, cycloheptyl CHN), 3.9 (m, 2 H, cycloheptyl CH2), 3.1 (m, 4 H,
2 CONHCH2), 2.8 (m, 2 H, cycloheptyl CH2), 2.5 (m, 4 H, CH2NCH2), 1.5 (pentaplet, 2 H, J =
6.3 Hz, CH2CH2CH2) and 1.3 (m, 4 H, 2 NCH2CH2) ppm; 13C NMR: δ 157.25 (q, 2JCF = 36.7

Hz, CF3CO), 157.2 (q, 2JCF = 36.7 Hz, CF3CO), 139.45, 139.43, 130.7, 130.4, 127.7 and 125.5 (
2 x 6 aromatic C), 115.9 (q, 1JCF = 288 Hz, CF3), 115.8 (q, 1JCF = 288 Hz, CF3), 76.1
(cycloheptyl CHN), 49.8, 47.7, 39.4, 38.1, 31.4 (2 cycloheptyl CH2), 26.3, 25.4 and 22.2 ppm;
MS (ESI) m/z 529.92 (MH+).
4.1.7. N4,N8-bis(dibenzosuberyl)-N1,N12-bis(trifluoroacetyl)spermine (15)
The general procedure was followed but 2.4 mol. equiv. of 5-chlorodibenzosuberane and 5.0 mol
equiv. of triethylamine were used to alkylate the two secondary amino groups. Compound (15)
was obtained as a light-colored solid; yield, 89%; mp 140-141 oC; 1H NMR: δ 7.1 (m, 16 H,
aromatic H), 6.2 (br. t, 2 H, 2 CONH), 4.3 (s, 2 H, 2 cycloheptyl CHN), 3.9 (m, 4 H, 2
cycloheptyl CH2), 3.1 (q, J = 6.2 Hz, 4 H, 2 CONHCH2), 2.8 (m, 4 H, 2 cycloheptyl CH2), 2.4 (t,
J = 6.6 Hz, 4 H, 2 CH2N), 2.3 (br. m, 4 H, 2 CH2N), 1.5 (pentaplet, 4 H, J = 6.8 Hz, 2
CH2CH2CH2) and 1.0 (m, 4 H, 2 NCH2CH2) ppm; 13C NMR: δ 157.25 (q, 2JCF = 36.7 Hz, 2
CF3CO), 139.57, 139.47, 130.8, 130.6, 127.8 and 125.6 ( 4 x 6 aromatic C), 118 (q, 1JCF = 288
Hz, 2 CF3), 76.1 (cycloheptyl 2 CHN), 50.0, 47.5, 38.2, 31.6 (2 cycloheptyl CH2), 31.5 (2
cycloheptyl CH2), 25.7 and 22.4 ppm; MS (FAB) m/z 779.5 (MH+); MS (ESI) m/z 779.04
(MH+).
4.1.8. General procedure for the removal of trifluoroacetyl groups of trifluoroacetamides
A solution of the trifluoroacetamide (0.5 mmol) in 1:1 CH3OH:NH4OH (20 mL) was refluxed
overnight. The solvent was removed under vacuum and purification by column chromatography
(1- 10% NH4OH in CH3OH) afforded the desired amine, compound (4), (5), (6) or (7)
respectively.
4.1.9. N-(2-aminoethyl)-N-dibenzosuberyl-1,2-diaminoethane (4)
Light-colored oil; quantitative yield; 1H NMR: δ 7.1 (m, 8 H, aromatic H), 4.39 (s, 1 H,
cycloheptyl CHN), 4.0 (m, 2 H, cycloheptyl CH2), 2.82 (m, 2 H, cycloheptyl CH2), 2.54 (s, 8 H,
4 NCH2) and 1.42 (br. s, 4 H, 2 NH2) ppm; 13C NMR: δ 139.8, 139.5, 130.8, 130.7, 127.8 and
125.6 (2 x 6 aromatic C), 77.6 (cycloheptyl CHN), 55.7, 39.8 and 31.7 (2 cycloheptyl CH2) ppm;
MS (ESI) m/z 296.09 (MH+).
4.1.10. N-(3-aminopropyl)-N-dibenzosuberyl-1,3-diaminopropane (5)
Light-colored oil; yield, 96%; 1H NMR: δ 7.1 (m, 8 H, aromatic H), 4.35 (s, 1 H, cycloheptyl
CHN), 3.95 (m, 2 H, cycloheptyl CH2), 2.8 (m, 6 H, cycloheptyl CH2 and 2 NH2), 2.5 (m, 8 H, 4
NCH2) and 1.5 (m, 4 H, 2 CH2) ppm; 13C NMR (CD3OD, 62.9 MHz): δ 141.5, 140.9, 131.8,
131.7, 128.8 and 126.6 (2 x 6 aromatic C), 77.2 (cycloheptyl CHN), 48.5, 40.7, 32.7 (2
cycloheptyl CH2) and 29.3 ppm; MS (ESI) m/z 324.13 (MH+).
4.1.11. N4-(dibenzosuberyl)spermidine (6)

Light-colored oil; yield, 64%; 1H NMR: δ 7.1 (m, 8 H, aromatic H), 4.35 (s, 1 H, cycloheptyl
CHN), 3.95 (m, 2 H, cycloheptyl CH2), 2.8 (m, 2 H, cycloheptyl CH2), 2.45 (m, 8 H, 4 NCH2),
2.0 (br. s, 4 H, 2 NH2), 1.4 (pentaplet, 2 H, J = 6.95 Hz, CH2CH2CH2), 1.3 (m, 2 H,
CH2CH2CH2) and 1.2 (pentaplet, 2 H, J = 6.95 Hz, CH2CH2CH2) ppm; 13C NMR: δ 139.8,
139.2, 130.3, 130.0, 127.2 and 125.1 (2 x 6 aromatic C), 75.6 (cycloheptyl CHN), 49.4, 46.9,
41.3, 39.7, 31.2 (2 cycloheptyl CH2), 30.8, 29.1 and 21.9 ppm; MS (FAB) m/z 338.2 (MH+); MS
(ESI) m/z 338.13 (MH+).
4.1.12. N4,N8-bis(dibenzosuberyl)spermine (7)
N4,N8-bis(dibenzosuberyl)-N1,N12-bis(trifluoroacetyl)spermine (15) (1.014 g, 1.30 mmol) was
dissolved in pyridine (15 mL) and 8 M aqueous NaOH (15 mL) was added. The mixture was
refluxed overnight. Saturated aqueous NaCl was added (40 mL) and the mixture extracted with
CH2Cl2 (4 x 40 mL). The collected organic layers were dried (Na2SO4), concentrated and
purified by column chromatography (2 to 10% CH3OH in CH2Cl2 containing 1% NH4OH) to
1
give a light-colored oil (0.745 g, 98%). H NMR(CDCl3, 300 MHz): δ 7.1 (m, 16 H, aromatic
H), 4.3 (s, 2 H, 2 cycloheptyl CHN), 3.95 (m, 4 H, 2 cycloheptyl CH2), 2.75 (m, 4 H, 2
cycloheptyl CH2), 2.45 (t, 4 H, J = 7.0 Hz, 2 NCH2), 2.35 (t, 4H, J = 7.0 Hz, 2 NCH2), 2.3 (br. t,
4 H, 2 NCH2), 1.4 (pentaplet, 4 H, J = 7.0 Hz, 2 NCH2CH2CH2N), 1.15 (br. s, 4 H, 2 NH2) and
1.05 (br. m, 4 H, CH2CH2) ppm; 13C NMR: δ 140.1, 139.4, 130.6, 130.3, 127.4 and 125.2 (4 x 6
aromatic C), 75.8 (2 cycloheptyl CHN), 49.8, 47.2, 40.1, 31.4 (4 cycloheptyl CH2), 29.7 and 22.6
ppm; MS (FAB) m/z 587.4 (MH+); MS (ESI) m/z 587.27 (MH+).
4.1.13. N-[3-(trifluoroacetamido)propyl]-N'-trifluoroacetyl-1,3-diaminopropane (9)
The general procedure previously described was followed, 86 i.e., to N-(3-aminopropyl)-1,3-
diaminopropane (2.50 g, 19.1 mmol) in CH3CN (15 mL) was added ethyl trifluoroacetate (9.77
g, 68.8 mmol) and water (0.41 g, 22.8 mmol). The solution was heated overnight at 80 oC.
Solvent was removed under vacuum and CH2Cl2 (100 mL) was added. The resulting white solid
was filtered and washed with CH2Cl2 to give (9) as a white solid (6.64 g, 80%); mp 166 oC; 1H
NMR (d6-DMSO, 250 MHz): δ 9.6 (br. s, 2 H, NH), 7,7 (br. s, 2 H, NH), 3.25 (t, 4 H, J = 6.5 Hz,
+
2 CONHCH2), 2.9 (br. t, 4 H, J = 7.5 Hz, CH2NH2 CH2) and 1.8 (br. pentaplet, J = 7.5 Hz, 4 H,
2 CH2CH2CH2) ppm; 13C NMR (d6-DMSO, 62.9 MHz): δ 159.2 (q, 2JCF = 32.0 Hz, CF3COO-
)156.6 (q, 2JCF = 36.2 Hz, 2 CF3CO), 117.2 (q, 1JCF = 298 Hz, CF3COO-), 116 (q, 1JCF = 288 Hz,
2 CF3), 44.8, 36.7 and 25.5 ppm; MS (ESI) m/z 324.20 (MH+).
4.2. Enzyme studies
4.2.1. Trypanothione reductase studies
T. cruzi TR was purified following the method of Walsh et al.88 from E. coli SG5, a
glutathione reductase deletion mutant, containing the TR expression vector pIBITczTR described
by Sullivan and Walsh.87, 88 Prepared compounds were assayed for their effects on the rate of

reduction of trypanothione disulfide by T. cruzi TR spectrophotometrically by monitoring the
oxidation of NADPH at 340 nm.13 Stock solutions of compounds (1), (3) and (4) were prepared
in HEPES buffer (100 mM, pH 7.25); stock solutions of compounds (5) and (7) were prepared in
1:1 ethanol:HEPES buffer; and stock solutions of compounds (2) and (6) were prepared in
ethanol . TR activity was measured at 23 oC in HEPES buffer (100mM, pH 7.25) containing
EDTA (1.0 mM), NADPH (0.18 mM), trypanothione disulfide (Bachem Bioscience Inc.) and TR
at 1.33 µg/mL. Also, to account for any possible effects of ethanol on TR, for the assays of
compounds (2), (5), (6) and (7), every 1.0 mL enzyme assay contained a total of 20 µL of
ethanol. An estimate of the Ki value for each inhibitor was obtained from initial assays. More
accurate Ki values were obtained in subsequent experiments in which a minimum of five
inhibitor concentrations (ranging from 0 µM to approximately 2 times the estimated Ki) were
assayed for their effects on TR activity in the presence of five trypanothione disulfide
concentrations (14.8, 22.2, 29.6, 44.4 and 74.0 µM).
For each compound the inhibition type was assessed by the patterns of three classes of
plots: 1/v against 1/[So] for various [I]; 1/v against [I] for various [So]; and [So]/v against [I] at
various [So]. For each inhibitor concentration, Km(obs) (with standard deviation) and Vmax were
determined from a non-linear regression analysis of the plot of v against [So] using the
Michaelis-Menten equation. The compounds all exhibited linear competitive inhibition and the
Ki value for each compound was determined from a weighted regression analysis of a plot of
Km(obs) against [I] using the following equation (weighting factors used were 1/(SD of Km(obs))):
Km(obs) = (Km[I]/Ki) + Km
4.2.2. Glutathione reductase studies
The effects of compounds (1) to (7) on the rate of reduction of glutathione disulfide by
yeast glutathione reductase (GR) (EC 1.6.4.2) was assayed spectrophotometrically by monitoring
the oxidation of NADPH at 340 nm. Stock solutions of compounds (1) to (4) were prepared in
ethanol, and stock solutions of compounds (5) to (7) were prepared in 1:1 ethanol:HEPES buffer
(100 mM, pH 7.25). To account for any possible effects of ethanol on GR, every 1.0 mL enzyme
assay contained a total volume of 20 µL of ethanol. GR activity was measured at 23 oC in
HEPES buffer (100 mM, pH 7.25) containing EDTA (1.0 mM), NADPH (0.18 mM), ethanol (20
µL) and five glutathione disulfide concentrations (15.0, 22.5, 30.0, 40.0 and 60.0 µM) with an
enzyme concentration of either 0.255 or 0.30 µg/mL. The effects of compounds (1) to (7) on GR
activity were measured at a concentrations of 22.5 and 60 µM glutathione disulfide and at: 0,
250 and 500 µM of compounds (1), (2), (3) or (4); 0 and 79.9 µM of (5); or 0, 125 and 250 µM
of compounds (6) or (7).
4.3. Trypanosome studies
4.3.1. Trypanosome isolates

Four trypanosome isolates were used to assess in vitro activity: Trypanosoma brucei
brucei Lab 110 EATRO and three clinical isolates of T. b. rhodesiense, KETRI 243, KETRI 269
and KETRI 243-As-10-3. KETRI 243 displays some resistance to melamine-based arsenical
drugs (including melarsoprol) and pentamidine. KETRI 243-As-10-3 (a clone of KETRI 243) is
completely resistant to melarsoprol and pentamidine.90
4.3.2. Determination of in vitro antitrypanosomal activities of compounds (2) to (7)
Trypanosome strains were grown as the blood form at 37 oC in a synthetic medium
(HMI-18) with 20% fetal bovine serum. The in vitro trypanocidal activities of compounds were
assessed by determining IC50 values. Measurements were done in duplicate in 24 well plates (1
mL/well). After 48 h, the number of parasites/well was determined using a Z1TM Coulter
Counter®. In the absence of inhibiting compounds, controls grew to 5 x 106 parasites /mL after
48 h. The approximate range of activity for each compound was initially determined using
compound concentrations of 0.1, 1.0, 10, 25 and 100 µM. The 50% inhibitory concentration
(IC50) of compounds after incubation with trypanosomes for 48 h, was then determined from
additional studies using compound concentrations spaced close to the approximate IC50 value.
IC50 values were determined from semi-log plots. Compounds were dissolved in water then
diluted with medium.90
4.3.3. In vivo studies of trypanocidal activities of compounds (6) and (7)
Female Swiss-Webster mice were infected intraperitoneally (ip) with 2.5 x 105 T. b.
brucei Lab EATRO blood forms of the parasite taken from an infected rat. The infection was
allowed to develop for 24 h before treatment was begun. Infected mice were divided into groups
of five, including infected, non-treated controls. For compound (6), animals were dosed ip with
1.0, 2.5, 5.0, 10 or 25 mg per kg body weight, given once daily for three days. For compound
(7), animals were dosed ip with a total daily dose of 25 or 50 mg per kg body weight,
administered in two doses every day (b.i.d.) for three days. Compound (7) was also given at 50
mg per kg body weight, given in a single daily dose for three days. In a second experiment,
compound (7) was given via Alzet® mini-osmotic pumps (ALZA, Corp., Palo Alto, CA) which
dispense 1 µL/h continuously for 3 days. Compound (7) was given at 10, 25 or 50 mg/kg/day.
Survival times were compared to infected mice that were not treated with compounds.
4.4. Computational methods
4.4.1. Ligand preparation
Compounds (1) to (7) were built in 2D using AccelrysDraw 4.0117 and saved as skc files.
The predominant protonation states of the compounds at pH 7.2 were estimated using
“Calculator Plugins” within MarvinSketch 6.3.0 118 and compared with literature pKa values for
analogous compounds.119, 120 This led to the identification of one protonated form of
clomipramine (1) and ten protonated forms of compounds (2) to (7) that we investigated.

Using Accelrys Discovery Studio 3.0 Visualizer,121 the eleven 2D structures in skc format
were converted to 3D structures with hydrogen atoms and protons, and the structure files were
saved in MDL Molfile format.
Since most docking programs including AutoDock Vina treat rings as rigid with no
conformational flexibility during docking calculations (although other bonds in ligands can be
designated as rotatable), we investigated the in vacuo energies of DBS/DBA conformers. Thus,
the in vacuo energies of DBS and DBA ring conformations of different rotamers (i.e., while
allowing rotation of DBS–N or DBA–C bonds) were explored by employing the MMFF94 force
field with the conjugate gradient algorithm available in Avogadro 1.1.0.122, 123 For these studies
+
the model structures DBS–NHCH3, DBS–NH2 CH3 and DBA–CH3 were used (protonation of
the endocyclic N atom in DBA was not studied since this N is not protonated at pH 7.2). Since
the DBS and DBA ring conformers are chiral, this led to the identification of in vacuo energy
minimum conformations for the tricyclic rings (four conformations for compounds with the
+
structure DBS–NHR, two for DBS–NH2 R and four for the DBA of clomipramine).
Representative structures of the conformations are shown in Table 4 (structures for all the
conformations investigated are provided in Supplemental Information).
Since AutoDock programs treat cyclic groups as rigid, the structures in each of the MDL
Molfiles were modified using Avogadro 1.1.0122, 123 to include the appropriate energy-minimized
DBS or DBA ring conformations. This led to 33 structure files (of the protonated and uncharged
forms of compounds (1) to (7) along with the identified ring conformations) which were written
as SYBYL Mol2 files.
AutoDockTools (v 1.5.4 revision 30)124, 125 was used to prepare each of the 33 structure
files for docking. This involved identifying bonds that will rotate during dockings, and merging
carbon atom volumes with attached hydrogens (i.e., non-polar hydrogens) using the unified atom
model. The resulting structure/ligand files were written as PDBQT files (which is the AutoDock
Vina input file format) and used in subsequent docking runs.
Additional calculations (using the MMFF94 force field in Avogadro) showed that the
energy barriers for interconversion of the conformations (i.e., a ring flex involving rotation at
C10 and C11) were less than 10 kcal/mol (which is consistent with energy barriers for
conformation interconversion of analogous compounds, including cycloheptane and
oxacycloheptane, with values of 8.55 and 8.5 kcal/mol126 and imipramine127, 128). This indicates
that in vitro the ring conformations will interconvert.
4.4.2. Protein preparation
The structure of the biologically active dimer of T. cruzi TR containing bound
trypanothione was downloaded from the Research Collaboratory for Structural Bioinformatics
Protein Data Bank (RCSB Protein Data Bank code ID: 1BZL biological assembly version).33
AutoDockTools was used to: remove all water molecules, ions and ligands; add polar hydrogens
and assign standard protonation states; and to create the unified atom model. The resulting
protein/receptor structure was written as a PDBQT file and used for subsequent dockings. A grid

box (the search space) was defined which encompassed the entire active site cleft (the defined
grid box was significantly larger than the active site to minimize any “boundary effects” in the
docking runs). The grid box did not include the FAD or NADPH binding regions or the interface
pocket of TR. The size of the resulting grid box was 30 x 50 x 30 Å (and centered at coordinates
x = 66.0, y = 6.0 and z = -6.0 with respect to the original 1BZL file).
4.4.3. Docking method
We utilized AutoDock Vina91 for docking studies. AutoDock Vina is often better than
AutoDock 4, and programs such as Dock and FlexX, for predicting experimentally observed
binding modes especially for molecules with 10 or more rotatable bonds.91, 98, 129 Although
several recent TR docking studies utilized AutoDock 499, 130 or earlier generations of AutoDock
and/or FlexX,131, 132 given the large conformational flexibility and number of rotatable bonds of
our inhibitors, AutoDock Vina was the obvious choice for our studies. Specifically, the
Windows version of AutoDock Vina 1.1.2 91 was used to dock each of the 33 ligand structures to
the receptor. The configuration file parameters were changed to include "exhaustiveness = 20"
(default is 8) and "num_modes = 10" (this saves the results of the 10 lowest energy poses, the
default is 9). This increased level of exhaustiveness was used to ensure comprehensive sampling
of the large search space of 30 x 50 x 30 Å (which was required to accommodate the large active
site cleft of TR). For each ligand structure 8 docking runs were performed and the coordinates of
the 10 lowest energy ligand poses for each run were collected (i.e., a total of 80 ligand poses for
each ligand structure). This approach allowed the collection of a large number of low energy
poses without the risk of introducing bias, since AutoDock Vina resets starting parameters at the
beginning of each new run. The binding energies of the lowest energy poses for all modeling
runs for the 33 ligand structures ranged from -5.6 to -8.1 kcal/mol (this range of binding energies
is very similar to in silico values calculated for tricyclics and quinones docked to TR99, 131).
Tables of the binding energies of the lowest energy poses for each ligand structure along with
descriptions of the binding sites of frequently occurring poses are provided in Supplemental
Information.
Acknowledgments
This material is based upon work supported by the National Science Foundation under CHE -
1229021. Additional support was received from the UNDP/World Bank/WHO Special
Programme for Research and Training in Tropical Diseases (grant # 920173); and from Canisius
College (Canisius Earning Excellence Program).
Supplementary information
Supplementary information associated with this article is available.

References
(1)
(2)
Stuart, K.; Brun, R.; Croft, S.; Fairlamb, A.; Gurtler, R. E.; McKerrow, J.; Reed, S.;
Tarleton, R. J. Clin. Invest. 2008, 118, 1301-1310.
WHO; Sustaining the drive to overcome the global impact of neglected tropical diseases:
second WHO report on neglected tropical diseases; World Health Organization: Geneva,
Switzerland, 2013.
(3)
Maya, J. D.; Cassels, B. K.; Iturriaga-Vasquez, P.; Ferreira, J.; Faundez, M.; Galanti, N.;
Ferreira, A.; Morello, A. Comp. Biochem. Physiol., Part A 2007, 146, 601-620.
O'Sullivan, M. Current Medicinal Chemistry - Anti-Infective Agents 2005, 4, 355-378.
Croft, S. L.; Yardley, V. Curr. Pharm. Des. 2002, 8, 319-342.
Fairlamb, A. H.; Blackburn, P.; Ulrich, P.; Chait, B. T.; Cerami, A. Science 1985, 227,
1485-1487.
(4)
(5)
(6)
(7)
(8)
Krauth-Siegel, R. L.; Leroux, A. E. Antioxid. Redox Signal. 2012, 17, 583-607.
Manta, B.; Comini, M.; Medeiros, A.; Hugo, M.; Trujillo, M.; Radi, R. Biochim. Biophys.
Acta 2013, 1830, 3199-3216.
(9)
Dickinson, D. A.; Forman, H. J. Biochem. Pharmacol. 2002, 64, 1019-1026.
(10) Bohme, C. C.; Arscott, L. D.; Becker, K.; Schirmer, R. H.; Williams, C. H., Jr. J. Biol.
Chem. 2000, 275, 37317-37323.
(11) Krauth-Siegel, R. L.; Comini, M. A. Biochim. Biophys. Acta 2008, 1780, 1236-1248.
(12) Comini, M. A.; Krauth-Siegel, R. L.; Bellanda, M. Antioxid. Redox Signal. 2013, 19,
708-722.
(13) Shames, S. L.; Fairlamb, A. H.; Cerami, A.; Walsh, C. T. Biochemistry 1986, 25, 3519-
3526.
(14) Colotti, G.; Baiocco, P.; Fiorillo, A.; Boffi, A.; Poser, E.; Chiaro, F. D.; Ilari, A. Future
Med. Chem. 2013, 5, 1861-1875.
(15) Khan, M. O. Drug Target Insights 2007, 2, 129-146.
(16) Jones, D. C.; Ariza, A.; Chow, W. H.; Oza, S. L.; Fairlamb, A. H. Mol. Biochem.
Parasitol. 2010, 169, 12-19.
(17) Krauth-Siegel, R. L.; Bauer, H.; Schirmer, R. H. Angew. Chem. Int. Ed. Engl. 2005, 44,
690-715.
(18) Flohe, L. Biotechnol. Adv. 2012, 30, 294-301.
(19) Flohe, L. Int. J. Med. Microbiol. 2012, 302, 216-220.
(20) Baiocco, P.; Colotti, G.; Franceschini, S.; Ilari, A. J. Med. Chem. 2009, 52, 2603.
(21) Bacchi, C. J.; Nathan, H. C.; Yarlett, N.; Goldberg, B.; McCann, P. P.; Sjoerdsma, A.;
Saric, M.; Clarkson, A. B. Antimicrob. Agents Chemother. 1994, 38, 563-569.
(22) Reguera, R.; Balana Fouce, R.; Cubria, J. C.; Alvarez Bujidos, M. L.; Ordonez, D.
Biochem. Pharmacol. 1994, 47, 1859-1866.
(23) Diaz, M. V.; Miranda, M. R.; Campos-Estrada, C.; Reigada, C.; Maya, J. D.; Pereira, C.
A.; Lopez-Munoz, R. Acta Tropica 2014, 134, 1-9.
(24) Fairlamb, A. H.; Henderson, G. B.; Cerami, A. Proc. Natl. Acad. Sci. U S A. 1989, 86,
2607-2611.
(25) Cunningham, M. L.; Zvelebil, M. J.; Fairlamb, A. H. Eur. J. Biochem. 1994, 221, 285-
295.
(26) Wyllie, S.; Cunningham, M. L.; Fairlamb, A. H. J. Biol. Chem. 2004, 279, 39925-39932.

(27) Krauth-Siegel, R. L.; Inhoff, O. Parasitol. Res. 2003, 90 Suppl 2, S77-85.
(28) Dumas, C.; Ouellette, M.; Tovar, J.; Cunningham, M. L.; Fairlamb, A. H.; Tamar, S.;
Olivier, M.; Papadopoulou, B. Embo J. 1997, 16, 2590-2598.
(29) Tovar, J.; Wilkinson, S.; Mottram, J. C.; Fairlamb, A. H. Mol. Microbiol. 1998, 29, 653-
660.
(30) Tovar, J.; Cunningham, M. L.; Smith, A. C.; Croft, S. L.; Fairlamb, A. H. Proc. Natl.
Acad. Sci. U S A. 1998, 95, 5311-5316.
(31) Krieger, S.; Schwarz, W.; Ariyanayagam, M. R.; Fairlamb, A. H.; Krauth-Siegel, R. L.;
Clayton, C. Mol. Microbiol. 2000, 35, 542-552.
(32) Baiocco, P.; Poce, G.; Alfonso, S.; Cocozza, M.; Porretta, G. C.; Colotti, G.; Biava, M.;
Moraca, F.; Botta, M.; Yardley, V.; Fiorillo, A.; Lantella, A.; Malatesta, F.; Ilari, A.
ChemMedChem 2013, 8, 1175-1183.
(33) Bond, C. S.; Zhang, Y.; Berriman, M.; Cunningham, M. L.; Fairlamb, A. H.; Hunter, W.
N. Structure 1999, 7, 81-89.
(34) Patterson, S.; Alphey, M. S.; Jones, D. C.; Shanks, E. J.; Street, I. P.; Frearson, J. A.;
Wyatt, P. G.; Gilbert, I. H.; Fairlamb, A. H. J. Med. Chem. 2011, 54, 6514-6530.
(35) Aboagye-Kwarteng, T.; Smith, K.; Fairlamb, A. H. Mol Microbiol. 1992, 6, 3089-3099.
(36) Henderson, G. B.; Fairlamb, A. H.; Ulrich, P.; Cerami, A. Biochemistry 1987, 26, 3023-
3027.
(37) Faerman, C. H.; Savvides, S. N.; Strickland, C.; Breidenbach, M. A.; Ponasik, J. A.;
Ganem, B.; Ripoll, D.; Krauth-Siegel, R. L.; Karplus, P. A. Bioorg. Med. Chem. 1996, 4,
1247-1253.
(38) Lu, J.; Vodnala, S. K.; Gustavsson, A. L.; Gustafsson, T. N.; Sjoberg, B.; Johansson, H.
A.; Kumar, S.; Tjernberg, A.; Engman, L.; Rottenberg, M. E.; Holmgren, A. J. Biol.
Chem. 2013, 288, 27456-27468.
(39) Richardson, J. L.; Nett, I. R.; Jones, D. C.; Abdille, M. H.; Gilbert, I. H.; Fairlamb, A. H.
ChemMedChem 2009, 4, 1333-1340.
(40) Garforth, J.; Yin, H.; McKie, J. H.; Douglas, K. T.; Fairlamb, A. H. J. Enzyme Inhib.
1997, 12, 161-173.
(41) Benson, T. J.; McKie, J. H.; Garforth, J.; Borges, A.; Fairlamb, A. H.; Douglas, K. T.
Biochem. J. 1992, 286, 9-11.
(42) Jacoby, E. M.; Schlichting, I.; Lantwin, C. B.; Kabsch, W.; Krauth-Siegel, R. L. Proteins
1996, 24, 73-80.
(43) Saravanamuthu, A.; Vickers, T. J.; Bond, C. S.; Peterson, M. R.; Hunter, W. N.;
Fairlamb, A. H. J. Biol. Chem. 2004, 279, 29493-29500.
(44) Seebeck, T.; Gehr, P. Mol. Biochem. Parasitol. 1983, 9, 197-208.
(45) Hammond, D. J.; Cover, B.; Gutteridge, W. E. Trans. R. Soc. Trop. Med. Hyg. 1984, 78,
91-95.
(46) Paglini-Oliva, P.; Fernandez, A. R.; Fretes, R.; Peslman, A. Exp. Mol. Path. 1998, 65, 78-
86.
(47) Pearson, R. D.; Manian, A. A.; Hall, D.; Harcus, J. L.; Hewlett, E. L. Antimicrob. Agents
Chemother. 1984, 25, 571-574.
(48) Evans, A. T.; Croft, S. L. Biochem. Pharmacol. 1994, 48, 613-616.
(49) Rivarola, H. W.; Fernandez, A. R.; Enders, J. E.; Fretes, R.; Gea, S.; Paglini-Oliva, P.
Trans. R. Soc. Trop. Med. Hyg. 2001, 95, 529-533.

(50) Rivarola, H. W.; Bustamante, J. M.; Lo Presti, S.; Fernandez, A. R.; Enders, J. E.; Gea,
S.; Fretes, R.; Paglini-Oliva, P. Exp. Parasitol. 2005, 111, 80-86.
(51) Bazan, P. C.; Lo Presti, M. S.; Rivarola, H. W.; Triquell, M. F.; Fretes, R.; Fernandez, A.
R.; Enders, J.; Paglini-Oliva, P. Parasitol. Res. 2008, 103, 663-669.
(52) Gobbi, P.; Baez, A.; Lo Presti, M. S.; Fernandez, A. R.; Enders, J. E.; Fretes, R.; Gea, S.;
Paglini-Oliva, P. A.; Rivarola, H. W. Parasitol. Res. 2010, 107, 1279-1283.
(53) Fauro, R.; Lo Presti, S.; Bazan, C.; Baez, A.; Strauss, M.; Triquell, F.; Cremonezzi, D.;
Negrete, O. S.; Willhuber, G. C.; Paglini-Oliva, P.; Rivarola, H. W. Parasitology 2013,
140, 917-927.
(54) Strauss, M.; Lo Presti, M. S.; Bazan, P. C.; Baez, A.; Fauro, R.; Esteves, B.; Sanchez
Negrete, O.; Cremonezzi, D.; Paglini-Oliva, P. A.; Rivarola, H. W. Parasitol. Int. 2013,
62, 293-299.
(55) de Barioglio, S. R.; Lacuara, J. L.; de Oliva, P. P. J. Parasitol. 1987, 73, 451-452.
(56) Chan, C.; Yin, H.; Garforth, J.; McKie, J. H.; Jaouhari, R.; Speers, P.; Douglas, K. T.;
Rock, P. J.; Yardley, V.; Croft, S. L.; Fairlamb, A. H. J. Med. Chem. 1998, 41, 148-156.
(57) Khan, M. O.; Austin, S. E.; Chan, C.; Yin, H.; Marks, D.; Vaghjiani, S. N.; Kendrick, H.;
Yardley, V.; Croft, S. L.; Douglas, K. T. J. Med. Chem. 2000, 43, 3148-3156.
(58) Bonse, S.; Santelli-Rouvier, C.; Barbe, J.; Krauth-Siegel, R. L. J. Med. Chem. 1999, 42,
5448-5454.
(59) Girault, S.; Grellier, P.; Berecibar, A.; Maes, L.; Mouray, E.; Lemiere, P.; Debreu, M. A.;
Davioud-Charvet, E.; Sergheraert, C. J. Med. Chem. 2000, 43, 2646-2654.
(60) Parveen, S.; Khan, M. O.; Austin, S. E.; Croft, S. L.; Yardley, V.; Rock, P.; Douglas, K.
T. J. Med. Chem. 2005, 48, 8087-8097.
(61) Chibale, K.; Haupt, H.; Kendrick, H.; Yardley, V.; Saravanamuthu, A.; Fairlamb, A. H.;
Croft, S. L. Bioorg. Med. Chem. Lett. 2001, 11, 2655-2657.
(62) Buchholz, K.; Comini, M.; Wissenbach, D.; Schirmer, R.; Krauth-Siegel, R.; Gromer, S.
Mol. Biochem. Parasitol. 2008, 160, 65-69.
(63) Caffrey, C. R.; Steverding, D.; Swenerton, R. K.; Kelly, B.; Walshe, D.; Debnath, A.;
Zhou, Y. M.; Doyle, P. S.; Fafarman, A. T.; Zorn, J. A.; Land, K. M.; Beauchene, J.;
Schreiber, K.; Moll, H.; Ponte-Sucre, A.; Schirmeister, T.; Saravanamuthu, A.; Fairlamb,
A. H.; Cohen, F. E.; McKerrow, J. H.; Weisman, J. L.; May, B. C. Antimicrob. Agents
Chemother. 2007, 51, 2164-2172.
(64) Li, Z.; Fennie, M. W.; Ganem, B.; Hancock, M. T.; Kobaslija, M.; Rattendi, D.; Bacchi,
C. J.; O'Sullivan, M. C. Bioorg. Med. Chem. Lett. 2001, 11, 251-254.
(65) O'Sullivan, M. C.; Zhou, Q.; Li, Z.; Durham, T. B.; Rattendi, D.; Lane, S.; Bacchi, C. J.
Bioorg. Med. Chem. 1997, 5, 2145-2155.
(66) O'Sullivan, M. C.; Zhou, Q. Bioorg. Med. Chem. Lett. 1995, 5, 1957-1960.
(67) Chitkul, B.; Bradley, M. Bioorg. Med. Chem. Lett. 2000, 10, 2367-2369.
(68) Bi, X.; Lopez, C.; Bacchi, C. J.; Rattendi, D.; Woster, P. M. Bioorg. Med. Chem. Lett.
2006, 16, 3229-3232.
(69) Bonnet, B.; Soullez, D.; Davioud-Charvet, E.; Landry, V.; Horvath, D.; Sergheraert, C.
Bioorg. Med. Chem. 1997, 5, 1249-1256.
(70) Girault, S.; Baillet, S.; Horvath, D.; Lucas, V.; DavioudCharvet, E.; Tartar, A.;
Sergheraert, C. Eur. J. Med. Chem. 1997, 32, 39-52.
(71) Ponasik, J. A.; Strickland, C.; Faerman, C.; Savvides, S.; Karplus, P. A.; Ganem, B.
Biochem. J. 1995, 311, 371-375.

(72) Smith, H. K.; Bradley, M. J. Comb. Chem. 1999, 1, 326-332.
(73) Dixon, M. J.; Maurer, R. I.; Biggi, C.; Oyarzabal, J.; Essex, J. W.; Bradley, M. Bioorg.
Med. Chem. 2005, 13, 4513-4526.
(74) De Luca, S.; Ulhaq, S.; Dixon, M. J.; Essex, J.; Bradley, M. Tetrahedron Lett. 2003, 44,
3195-3197.
(75) Salmon-Chemin, L.; Buisine, E.; Yardley, V.; Kohler, S.; Debreu, M. A.; Landry, V.;
Sergheraert, C.; Croft, S. L.; Krauth-Siegel, R. L.; Davioud-Charvet, E. J. Med. Chem.
2001, 44, 548-565.
(76) Czechowicz, J.; Wilhelm, A.; Spalding, M.; Larson, A.; Engel, L.; Alberg, D. J. Org.
Chem. 2007, 72, 3689-3693.
(77) Melchiorre, C.; Bolognesi, M. L.; Minarini, A.; Rosini, M.; Tumiatti, V. J. Med. Chem.
2010, 53, 5906-5914.
(78) Lizzi, F.; Veronesi, G.; Belluti, F.; Bergamini, C.; Lopez-Sanchez, A.; Kaiser, M.; Brun,
R.; Krauth-Siegel, R. L.; Hall, D. G.; Rivas, L.; Bolognesi, M. L. J. Med. Chem. 2012,
55, 10490-10500.
(79) Finlayson, R.; Pearce, A. N.; Page, M. J.; Kaiser, M.; Bourguet-Kondracki, M. L.;
Harper, J. L.; Webb, V. L.; Copp, B. R. J. Nat. Prod. 2011, 74, 888-892.
(80) Hamilton, C. J.; Saravanamuthu, A.; Poupat, C.; Fairlamb, A. H.; Eggleston, I. M.
Bioorg. Med. Chem. 2006, 14, 2266-2278.
(81) Reguera, R. M.; Tekwani, B. L.; Balana-Fouce, R. Comp. Biochem. Physiol. C Toxicol.
Pharmacol. 2005, 140, 151-164.
(82) Hasne, M. P.; Coppens, I.; Soysa, R.; Ullman, B. Mol. Microbiol. 2010, 76, 78-91.
(83) Hasne, M. P.; Ullman, B. Methods Mol. Biol. 2011, 720, 309-326.
(84) Colotti, G.; Ilari, A. Amino Acids 2011, 40, 269-285.
(85) Bacchi, C. J.; Yarlett, N. Mini Rev. Med. Chem. 2002, 2, 553-563.
(86) O'Sullivan, M. C.; Dalrymple, D. M. Tetrahedron Lett. 1995, 36, 3451-3452.
(87) Sullivan, F. X.; Shames, S. L.; Walsh, C. T. Biochemistry 1989, 28, 4986-4992.
(88) Sullivan, F. X.; Walsh, C. T. Mol. Biochem. Parasitol. 1991, 44, 145-147.
(89) Krauth-Siegel, R. L.; Coombs, G. H. Parasitol. Today 1999, 15, 404-409.
(90) Sufrin, J. R.; Rattendi, D.; Spiess, A. J.; Lane, S.; Marasco, C. J., Jr.; Bacchi, C. J.
Antimicrob. Agents Chemother. 1996, 40, 2567-2572.
(91) Trott, O.; Olson, A. J. J. Comput. Chem. 2010, 31, 455-461.
(92) Kuriyan, J.; Kong, X. P.; Krishna, T. S.; Sweet, R. M.; Murgolo, N. J.; Field, H.; Cerami,
A.; Henderson, G. B. Proc. Natl. Acad. Sci. U S A. 1991, 88, 8764-8768.
(93) Hunter, W., Bailey S, Habash J, Harrop SJ, Helliwell JR, Aboagye-Kwarteng T, Smith K,
Fairlamb AH. J. Mol. Biol. 1992, 227, 322-333.
(94) Bailey, S.; Smith, K.; Fairlamb, A. H.; Hunter, W. N. Eur. J. Biochem. 1993, 213, 67-75.
(95) Lantwin, C. B.; Schlichting, I.; Kabsch, W.; Pai, E. F.; Krauth-Siegel, R. L. Proteins.
1994, 18, 161-173.
(96) Persch, E.; Bryson, S.; Todoroff, N. K.; Eberle, C.; Thelemann, J.; Dirdjaja, N.; Kaiser,
M.; Weber, M.; Derbani, H.; Brun, R.; Schneider, G.; Pai, E. F.; Krauth-Siegel, R. L.;
Diederich, F. ChemMedChem 2014, 9, 1880-1891.
(97) Baiocco, P.; Ilari, A.; Ceci, P.; Orsini, S.; Gramiccia, M.; Di Muccio, T.; Colotti, G. ACS
Med. Chem. Lett. 2010, 2, 230-233.
(98) Chang, M. W.; Ayeni, C.; Breuer, S.; Torbett, B. E. PLoS One 2010, 5, e11955.
(99) Venkatesan, S. K.; Shukla, A. K.; Dubey, V. K. J. Comput. Chem. 2010, 31, 2463-2475.

(100) Houston, D. R.; Walkinshaw, M. D. J. Chem. Inf. Model. 2013, 53, 384-390.
(101) Savvides, S. N.; Karplus, P. A. J. Biol. Chem. 1996, 271, 8101-8107.
(102) Karplus, P. A.; Pai, E. F.; Schulz, G. E. Eur. J. Biochem. 1989, 178, 693-703.
(103) Schonleben-Janas, A.; Kirsch, P.; Mittl, P. R.; Schirmer, R. H.; Krauth-Siegel, R. L. J.
Med. Chem. 1996, 39, 1549-1554.
(104) Daeffler, K. N.; Lester, H. A.; Dougherty, D. A. J. Am. Chem. Soc. 2012, 134, 14890-
14896.
(105) Ringer, A. L.; Senenko, A.; Sherrill, C. D. Protein Sci. 2007, 16, 2216-2223.
(106) Tauer, T. P.; Derrick, M. E.; Sherrill, C. D. J. Phys. Chem. A 2005, 109, 191-196.
(107) el-Waer, A.; Douglas, K. T.; Smith, K.; Fairlamb, A. H. Anal. Biochem. 1991, 198, 212-
216.
(108) Garforth, J.; McKie, J. H.; Jaouhari, R.; Benson, T. J.; Fairlamb, A. H.; Douglas, K. T.
Amino Acids 1994, 6, 295-299.
(109) Bissantz, C.; Kuhn, B.; Stahl, M. J. Med. Chem. 2010, 53, 5061-5084.
(110) Ninković, D. B.; Andrić, J. M.; Malkovb, S. N.; Zarić, S. D. Phys. Chem. Chem. Phys.
2014, 16, 11173-11177.
(111) Pegg, A. IUBMB Life 2009, 61, 880-894
(112) Pegg, A.; Casero, R. Methods Mol. Bio. 2011, 720, 3-35.
(113) Pozzi, M.; Gawandi, V.; Fitzpatrick, P. Biochemistry 2009, 48, 1508-1516.
(114) O'Sullivan, M. C.; Dalrymple, D. M.; Zhou, Q. J. Enzyme Inhib. 1996, 11, 97-114.
(115) Girault, S.; Davioud-Charvet, T. E.; Maes, L.; Dubremetz, J. F.; Debreu, M. A.; Landry,
V.; Sergheraert, C. Bioorg. Med. Chem. 2001, 9, 837-846.
(116) Eberle, C.; Lauber, B. S.; Fankhauser, D.; Kaiser, M.; Brun, R.; Krauth-Siegel, R. L.;
Diederich, F. ChemMedChem 2011, 6, 292-301.
(117) Accelrys Software Inc. Accelrys Draw 4.0, 2010,
http://accelrys.com/products/informatics/cheminformatics/draw/.
(118) MarvinSketch 6.3.0 ChemAxon Ltd. 2014, 2014, http://www.chemaxon.com.
(119) Evans, D. A. Evans pKa Table 2014, http://evans.rc.fas.harvard.edu/,
http://evans.rc.fas.harvard.edu/pdf/evans_pKa_table.pdf.
(120) Smith, J. W., Ed. The Chemistry of the Amino Group, in The Chemistry of Functional
Groups ed. Patai, S.; John Wiley and Sons, 1968.
(121) Accelrys Software Inc. Accelrys Discovery Studio 3.0 Visualizer,
http://accelrys.com/products/.
(122) Avogadro: an open-source molecular builder and visualization tool,
http://avogadro.openmolecules.net/.
(123) Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.; Zurek, E.; Hutchison, G.
R. J. Cheminform. 2012, 4, 17.
(124) Sanner, M. F. J. Mol. Graph. Model. 1999, 17, 57-61.
(125) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.; Goodsell, D. S.;
Olson, A. J. J. Comput. Chem. 2009, 30, 2785-2791.
(126) Freeman, F.; Hwang, J. H.; Junge, E. H.; Parmar, P. D.; Renz, Z.; Trinh, J. Int. J.
Quantum Chem. 2008, 108, 339-350.
(127) Casarotto, M. G.; Craik, D. J. J. Pharm. Sci. 2001, 90, 713-721.
(128) Munro, S. L.; Craik, D. J.; Andrews, P. R. Quant. Struct.-Act. Relat. 1987, 6, 104-110.
(129) Erickson, J. A.; Jalaie, M.; Robertson, D. H.; Lewis, R. A.; Vieth, M. J. Med. Chem.
2004, 47, 45-55.

(130) Perez-Pineiro, R.; Burgos, A.; Jones, D. C.; Andrew, L. C.; Rodriguez, H.; Suarez, M.;
Fairlamb, A. H.; Wishart, D. S. J. Med. Chem. 2009, 52, 1670-1680.
(131) de Molfetta, F. A.; de Freitas, R. F.; da Silva, A. B.; Montanari, C. A. J. Mol. Model.
2009, 15, 1175-1184.
(132) da Rocha Pita, S. S.; Cirino, J. J.; de Alencastro, R. B.; Castro, H. C.; Rodrigues, C. R.;
Albuquerque, M. G. J. Mol. Graph. Model. 2009, 28, 330-335.

Figure 1. T. cruzi TR active site showing the four DBS/DBA binding regions superimposed on
the 1BZL structure containing bound trypanothione. Specifically the figure shows superimposed
the DBS groups from the five lowest energy poses of docking runs for all conformations of
compound (4). View is from (A) the side of the active site; and (B) a 90o rotation with a view
looking into the active site. Represented is: TR protein subunit A in cream color; protein subunit
B in blue/grey (subunit B residues are labeled with a prime (‘)); trypanothione in brown and CPK
coloring. DBS groups are shown in wireframe, with DBS in binding region 1(BR 1) as red; DBS
in BR 2 as yellow; DBS in BR 3 as blue/aqua; and DBS in BR 3 as green.
(A)
(B)