Virtual screening identification of nonfolate compounds, including a CNS drug, as antiparasitic agents inhibiting pteridine reductase

Article abstract of DOI:10.1021/jm1010572

Folate analogue inhibitors of Leishmania major pteridine reductase (PTR1) are potential antiparasitic drug candidates for combined therapy with dihydrofolate reductase (DHFR) inhibitors. To identify new molecules with specificity for PTR1, we carried out a virtual screening of the Available Chemicals Directory (ACD) database to select compounds that could interact with L. major PTR1 but not with human DHFR. Through two rounds of drug discovery, we successfully identified eighteen drug-like molecules with low micromolar affinities and high in vitro specificity profiles. Their efficacy against Leishmania species was studied in cultured cells of the promastigote stage, using the compounds both alone and in combination with 1 (pyrimethamine; 5-(4-chlorophenyl)-6-ethylpyrimidine-2,4-diamine). Six compounds showed efficacy only in combination. In toxicity tests against human fibroblasts, several compounds showed low toxicity. One compound, 5c (riluzole; 6-(trifluoromethoxy)- 1,3-benzothiazol-2-ylamine), a known drug approved for CNS pathologies, was active in combination and is suitable for early preclinical evaluation of its potential for label extension as a PTR1 inhibitor and antiparasitic drug candidate.

Full text of DOI:10.1021/jm1010572

J. Med. Chem. 2011, 54, 211–221 211
DOI: 10.1021/jm1010572
Virtual Screening Identification of Nonfolate Compounds, Including a CNS Drug, as Antiparasitic
Agents Inhibiting Pteridine Reductase
Stefania Ferrari,† Federica Morandi,†, Domantas Motiejunas,†,^ Erika Nerini,†,‡ Stefan Henrich,‡ Rosaria Luciani,†
Alberto Venturelli, Sandra Lazzari, Samuele Calo, Shreedhara Gupta, Veronique Hannaert, Paul A. M. Michels,
†
†
§
§
§
ꢀ †
Rebecca C. Wade,*,‡ and M. Paola Costi*,†
†
ꢀ
Dipartimento di Scienze Farmaceutiche, Universita degli Studi di Modena e Reggio Emilia, Via Campi 183, 41100 Modena, Italy,
‡Heidelberg Institute for Theoretical Studies (HITS) gGmbH, Schloss-Wolfsbrunnenweg 35, 69118 Heidelberg, Germany, and
§
ꢁ
Research Unit for Tropical Diseases, de Duve Institute and Laboratory of Biochemistry, Universite catholique de Louvain,
Avenue Hippocrate 74, B-1200 Brussels, Belgium. Current address: Merck Serono Switzerland. ^ Current address: CropDesign N.V.,
Technologiepark 3, B-9052 Gent, Belgium.
Received August 14, 2010
Folate analogue inhibitors of Leishmania major pteridine reductase (PTR1) are potential antiparasitic
drug candidates for combined therapy with dihydrofolate reductase (DHFR) inhibitors. To identify
new molecules with specificity for PTR1, we carried out a virtual screening of the Available Chemicals
Directory (ACD) database to select compounds that could interact with L. major PTR1 but not with
human DHFR. Through two rounds of drug discovery, we successfully identified eighteen drug-like
molecules with low micromolar affinities and high in vitro specificity profiles. Their efficacy against
Leishmania species was studied in cultured cells of the promastigote stage, using the compounds both
alone and in combination with 1 (pyrimethamine; 5-(4-chlorophenyl)-6-ethylpyrimidine-2,4-diamine).
Six compounds showed efficacy only in combination. In toxicity tests against human fibroblasts, several
compounds showed low toxicity. One compound, 5c (riluzole; 6-(trifluoromethoxy)-1,3-benzothiazol-
2-ylamine), a known drug approved for CNS pathologies, was active in combination and is suitable for
early preclinical evaluation of its potential for label extension as a PTR1 inhibitor and antiparasitic drug
candidate.
Introduction
Trypanosomatids are auxotrophic for folates and pterins, and
inhibition of the enzymes involved in the salvage pathways
should provide effective treatment.5 However, antifolates are
currently not employed in therapy of trypanosomatid infections,
mainly because of the pteridine reductase (PTR1) activity of
the target organisms. PTR1 is a short-chain dehydrogenase/
reductase that is able to carry out successive reductions of both
conjugated (folate) and unconjugated (biopterin) pterins.6,7
While the bifunctional DHFR-TS used by trypanosomatids
can only reduce folic acid, PTR1 can act on a broader range of
substrates. Under physiological conditions, PTR1 is respon-
sible for the reduction of 10% of the folic acid required by the
cell, but when classical antifolate drugs inhibit DHFR-TS,
PTR1 can be overexpressed, ensuring parasite survival.8,9 This
suggests that treatment of trypanosomatid infections could be
achieved through the simultaneous inhibition of DHFR and
PTR1 by a single drug or a combination of compounds that
are specific and selective inhibitors of both targets.8
Parasites of the Trypanosomatidae family are the causal
agents of a number of serious human diseases, including
African sleeping sickness, Chagas’ disease, and leishmaniasis.
The impact of these parasites on public health and the inad-
equacy of current treatments have created an urgent require-
ment for more effective drugs, since those in use are highly
toxic and often difficult to administer. The problem is com-
pounded by a rise in drug resistance. Therapy for leishmani-
asis generally relies on old drugs such as sodium stibogluco-
nate and Amphotericin B, and only one new treatment has
been developed in the last 25 years, Miltefosine, which is ap-
proved in India for visceral Leishmanisis.1
Enzymes involved in the provision of reduced folate cofac-
tors, e.g., dihydrofolate reductase (DHFRa), and enzymes
that utilize these cofactors, like thymidylate synthase (TS),
are important drug targets for the treatment of bacterial
infections,2 cancer,3 and certain parasitic diseases, notably
malaria.4 Inhibition of DHFR or TS leads to a reduction in
the cellular pools of 20-deoxythymidine-50-monophosphate,
severely impairs DNA replication, and results in cell death.
Previous work in our laboratories10 has demonstrated that
it is possible to identify specific inhibitors of PTR1 and to use
such inhibitors in combination with known antifolates to
effectively improve in vitro efficacy against Leishmania and
Trypanosoma species.10 A similar approach has been taken to
discover new compounds withpyrrolo[2,3-d]pyrimidine struc-
tures that are active against Trypanosoma brucei PTR1 and
on bloodstream parasites.11 Very recently, a virtual screening
strategy revealed compounds with aminobenzothiazole and
*To whom correspondence should be addressed. For M.P.C.: Phone
0039-059-205-5134; E-mail mariapaola.costi@unimore.it. For R.C.W.:
Phone 0049-6221-533-247; E-mail rebecca.wade@h-its.org.
a Abbreviations: ACD, Available Chemicals Directory; DHFR, di-
hydrofolate reductase; PTR1, pteridine reductase; rmsd, root-mean-
square deviation; TS, thymidylate synthase.
r
2010 American Chemical Society
Published on Web 12/02/2010
pubs.acs.org/jmc

212 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ferrari et al.
Table 1. Compounds with Inhibitory Activity against LmPTR1 and Their Biological Activity Profiles
a-e NI: No inhibition. Measurements made with the following concentrations of compound: a 500 μM; b 1 mM; c 50 μM; d 100 μM; e 25 μM (see Materials
and Methods).
aminobenzimidazole scaffolds that inhibit Trypanosoma brucei
PTR1.12 In the present paper, we describe the application of a
different virtual screening approach combined with rapid
synthetic and experimental screening methodologies that has
enabled us to identify nonfolate like inhibitors of Leishmania
PTR1 with thiadiazole core structures. The hits identified
have been optimized in two structure-based design cycles,
including species specificity studies. The compounds were
tested against the Leishmania major enzymes, LmPTR1 and
LmDHFR-TS, and the human DHFR (hDHFR). The best
compounds were also assayed against in vitro cultured cells
of different trypanosomatid species, such as promastigotes of

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 213
Figure 2. (A) Structure of LmPTR1 with compound 4a bound as
calculated with the LUDI software. (B) Schematic representation of
the binding mode of compound 4a derivatives with substituents at
position 5.
forming hydrogen bonds with the active site residues (number of
H bonds >1); (III) the calculated score (>400). The remain-
ing 724 docking results were analyzed visually and molecules
were selected based on (I) the number and type of interac-
tions established, (II) which residues they interact with, and
(III) comparison of the structures of the active sites of
LmPTR1 and hDHFR. Molecules that were predicted to
form a stacking interaction with Phe113 and to interact with
LmPTR1 residues that are not conserved in the hDHFR
active site were selected preferentially. This second selection
resulted in 53 molecules that were purchased and tested
against LmPTR1 (Supporting Information Table 1-SI). Six
of these compounds (4a, 6a, 28a, 35a, 38a, 53a) were found to
be active against LmPTR1 with IC50 values between 0.39 and
5.6 mM (Table 1, Figure 1). These molecules were further
tested for their inhibitory activity against LmDHFR and
hDHFR. None of them was active against hDHFR. Only
compound 28a showed weak inhibitory activity against
LmDHFR. Compound 4a was selected for further development
on the basis of synthetic feasibility and biological activity
(IC50 and Ki against LmPTR1 of 5.6 mM and 436 μM,
respectively). The derivatization of the hit at position 5 of
the thiadiazole ring (Figure 2) was expected to increase the
affinity and the specificity for PTR1 by introduction of frag-
ments that lead to more complex and drug-like compounds.
Library Design, Synthesis, and Testing. A first set of
derivatives of compound 4a with substitutions at position 5
was designed with the aim of increasing the affinity toward
LmPTR1, as well as exploring molecular diversity (Figure 2,
Supporting Information Table 2-SI), thereby aiding optimi-
zation of the molecular properties for drug-likeness.15 The
calculated ternary complex of LmPTR1-NADPH-4a shows
the ligand to be involved in stacking interactions with both
the nicotinamide ring of the cofactor and the phenyl ring of
Figure 1. Inhibitory activity (IC50 values) and specificity profile of
(A) compounds from virtual screening, (B) derivatives from the
thiadiazole library, and (C) benzothiazole compounds. The Y axis is
on a logarithmic scale. The color code is as follows: LmPTR1, black;
LmDHFR, gray; hDHFR, striped.
L. mexicana and L. major, both as single agents and in com-
bination with 1 (pyrimethamine; 5-(4-chlorophenyl)-6-ethyl-
pyrimidine-2,4-diamine)13 (Supporting Information Figure 1-SI).
The toxicity of these compounds was tested against human
fibroblasts (MRC5).
Results and Discussion
Virtual Screening Hits and Scaffold Identification. Virtual
screening of the Available Chemicals Directory (ACD) data-
base against the three-dimensional structure of LmPTR1
(PDB ID: 1E92) using the program LUDI14 resulted in
21 394 docked molecules. Analysis of the available crystal-
lographic structures showed that the substrate, dihydro-
biopterin binds to LmPTR1 by forming an extended network
of hydrogen bonds and aromatic stacking interactions with
the cofactor and Phe113. During the selection of the docked
molecules, we wanted to select compounds able to mimic the
binding mode of this substrate. The docked molecules were
initially filtered based on three calculated parameters: (I) the
ability of the ligand to bind deep into the active site (contact
percentage >50); (II) the ability to replace the substrate by

214 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Scheme 1. Synthesis of Compounds in Library B
Ferrari et al.
Phe113, and making hydrogen bonds with the diphosphate
and the ribose moieties of NADPH and with the side chains
of Ser111 and Tyr194. On the basis of this model, we deri-
vatized position 5 because there is enough space to accom-
modate various different substituents. A variety of substit-
uents were designed to introduce interactions with the
residues lining this part of the binding site. Substituents
encompass small aliphatic chains (1b-3b), one-ring aromatic
systems with 0, 1, or 2 substituents (4b-15b), longer chains with
a variable-size aliphatic bridge and one (16b-23b) or two
(24b-25b) aromatic ring systems. All of these compounds
comply with Lipinski’s rule of five16 (Supporting Informa-
tion Table 2-SI). The compound library was obtained by
reacting thiosemicarbazide with different carboxylic acids
compounds were tested against LmPTR1 (Supporting In-
formation Table 3-SI). Seven compounds (7b, 14b, 15b, 21b,
22b, 28b, 29b) showed competitive inhibition of LmPTR1
with respect to the dihydrofolate cofactor showing IC50
values between 22 and 309 mM corresponding to Ki values
of 2-24 mM (Table 1, Figure 1). All these compounds were
more active compared to the starting hit 4a (IC50 5.6 mM)
with the improvement in IC50 value being 254-fold in the case
of the best compound, 15b. The molecules were also tested
for their activity against LmDHFR and hDHFR (Table 1,
Figure 1). Only one compound, 21b, showed an inhibitory
activity against hDHFR with an IC50 value of 300 μM. Two
compounds, 15b and 21b, inhibited LmDHFR with IC50
values in the range 139-1300 μM; thus, this class of deriva-
tives showed specificity with respect to PTR1.
€
(Scheme 1) using Buchi’s Syncore parallel synthesizer, re-
sulting in 26 compounds, 1b-5b, 7b-11b, 13b-28b (Supporting
Information Table 3-SI). All designed compounds were
obtained as [1,3,4]thiadiazol-2-ylamine with the appropriate
substituent in the 5 position, except for compounds 6b and
12b, which were isolated only as the 5-substituted [1,3,4]-
thiadiazol-2-ylamide of the corresponding acid: N-(5-pyri-
din-4-yl-[1,3,4]thiadiazol-2-yl)pyridine-4-carboxamide (26b)
and N-[5-(3,4-dimethylphenyl)-[1,3,4]thiadiazol-2-yl]-3,4-
dimethylbenzamide (27b). From the reactions of thiosemi-
carbazide with 3-chloropropionic acid or phthalimidoacetic
acid were obtained both the [1,3,4]thiadiazol-2-ylamine derivative
and 5-substituted [1,3,4]thiadiazol-2-ylamide, namely, 2b (5-(2-
chloroethyl)-[1,3,4]thiadiazol-2-ylamine) and 28b (3-chloro-
N-[5-(2-chloroethyl)-[1,3,4]thiadiazol-2-yl]-propionamide), 17b
(2-(5-amino-[1,3,4]thiadiazol-2-ylmethyl)-isoindole-1,3-dione),
and 29b (2-(1,3-dioxoisoindolin-2-yl)-N-5-(1,3-dioxoisoindolin-
2-ylmethyl)-[1,3,4]thiadiazol-2-ylacetamide). Twenty-six
Virtual Docking as a Basis for Lead Optimization. Docking
studies were performed on the tested compounds (4a, 1b,
3b-5b, 7b-11b, 13b-29b) with the aim of guiding the fur-
ther synthetic elaboration of active compounds.
First, different crystal structures of LmPTR1 were re-
trieved from the Protein Data Bank17 (Supporting Informa-
tion Table 4-SI). and the docking procedure was tested by
cross-docking all the ligands for which the crystal structures
of their complexes with LmPTR1 were known. That is, for
each of the seven crystal structures of LmPTR1-ligand
complexes, the ligand was removed and the ligands from
the other six crystal structures were docked into this protein
structure and evaluated. Rmsd (root-mean-square deviation)
values were calculated for the docking poses with respect to
the ligand position in the crystal structures, either for all non-
hydrogen atoms or for the core scaffold atoms. Fragments
with high B-factors for the ligands 10-propargyl-5,8-dideazafolic

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 215
Figure 3. Interaction properties of the LmPTR1 active site. (A) The isosurfaces show energetically favorable regions for a hydrophilic water
probe (blue, -6.0 kcal/mol) and a hydrophobic “DRY” probe (yellow, -0.5 kcal/mol) computed with the GRID program. The NADPH
cofactor, docked [1,3,4]thiadiazole-2-amine compound, 4a, and some important residues surrounding the active site are shown in stick
representation. D181, Y194, and K198 are the catalytic triad residues involved in the first step of reduction. R2870 is a residue from a second
subunit of PTR1 which contributes to the hydrogen bond network of the water-filled pocket at the active site. Crystallographic water molecules
are represented by blue spheres. The four water molecules close to R2870 were treated explicitly for docking of ligands (see text for details). R17
and the nearby water molecule are important in the second step of the reduction reaction.26 (B) Schematic representation of the main
hydrophilic (blue) and hydrophobic (yellow) regions in the active site of PTR1 identified with the GRID calculations. The regions marked with
dashed lines are somewhat further away from the active site and are not conserved among all the PTR1 structures investigated.
acid (CB3 of PDB ID: 2BFA) and 2 (methotrexate; 2-{4-
[(2,4-diamino-pteridin-6-ylmethyl)-methyl-amino]-benzoyl-
amino}-pentanedioic acid)18 (MTX of PDB ID: 1E7W)
(Supporting Information Figure 1-SI) were omitted. The
initial results obtained using both the GOLD v 3.219,20 and
GLIDE21 programs were unsatisfactory with three of the
correctly rank the docked compounds. Therefore, we com-
pared the GOLD scores and the inhibitory activity data for
the docked compounds. The results showed that these scores
could not discriminate between active and inactive com-
pounds. A second trial was done by rescoring the calculate
binding poses with Autodock 4.22 The results showed that all
but one active compound had a negative computed binding
free energy, whereas the computed binding free energies of
the inactive compounds were both positive and negative.
This means that, in this second trial, we obtained many false
positive results but only one false negative result. The posi-
tive energy values resulted from large Lennard-Jones ener-
gies due to van der Waals bumps. Energy minimization with
AMBER23 and computation of interaction energy terms did
not, however, lead to better discrimination between active
and inactive compounds. In summary, the docking calcula-
tions gave useful insights into the possible binding modes of
the compounds, even though they had a limited predictive
ability for activity.
Finally, docking of the same set of compounds was also
performed against two crystal structures of hDHFR to get
insights into the selectivity of these compounds. With a few
exceptions, the thiadiazole ring did not dock on top of the
nicotinamide ring of NADPH as observed for LmPTR1, but
instead faced a hydrophilic pocket in the core of the active
site. This suggests that the compounds would have different
binding modes to the two enzymes and are good candidates
for binding selectively to LmPTR1 and not to hDHFR.
Structure-Based Comparative Analysis of LmPTR1 and
hDHFR to Guide Further Lead Optimization. In the context
of the docking results and analysis of available X-ray struc-
tures of PTR1 and hDHFR, molecular interaction fields
calculated with the GRID program24,25 were used to suggest
possible extensions of the original lead compound, 4a. We
focus our analysis and discussion on the LmPTR1 structure
(PDB ID: 1E92). The GRID calculations were however
performed for all available PTR1 structures listed in Table
4-SI of the Supporting Information.
˚
ligands showing rmsd values over 5 A for some or all of the
target LmPTR1 structures. The poor docking poses could,
however, be improved by including the four most conserved
water molecules (with a conservation of 100%, 86%, 71%,
and 71%, respectively, in the crystal structures used) in the
binding site of LmPTR1 near the nicotinamide ring of the
cofactor. These water molecules were identified by a cluster
analysis based on all crystallographic water molecules in the
listed crystal structures of LmPTR1. The use of these four
conserved water molecules in combination with the GOLD
program and the embedded GoldScore fitness function
improved the cross-docking of the ligands to the different
target structures. The average rmsd values of the 10 docking
solutions for each ligand in the different receptor structures
˚
dropped below 2.2 A for 5,6,7,8-tetrahydrobiopterin (THB
˚
of PDB ID: 2BFP) and below 1.2 A for 7,8-dihydrobiopterin
(HBI of PDB ID: 1E92 and 2BF7). The average RMSDs for
2,4,6-triaminoquinazoline (TAQ of PDB ID: 1W0C) and
methotrexate (MTX of PDB ID: 1E7W) were higher and lay
˚
between 2.0 and 4.2 A, respectively. Docking of 10-propargyl-
5,8-dideazafolic acid (CB3 of PDB ID: 2BFA) and trimethoprim
(TOP of PDB ID: 2BFM) remained poor, showing an
˚
average rmsd between 2.7 and 7.0 A.
After evaluating the docking procedure, 27 compounds
(4a, 1b, 3b-5b, 7b-11b, 13b-29b) were docked into the
LmPTR1 crystal structure (PDB ID: 1E92). The docking
position for 4a identified with GOLD was very similar to that
found with LUDI during virtual screening, with the ring of
the compound sandwiched between the side chain of Phe113
and the nicotinamide ring of the cofactor. On one hand, this
result prompted us to use the calculate binding modes to
design further synthetic elaborations of the active com-
pounds by a structure-based approach. On the other hand,
we wanted to check if the docking procedure could also
The calculations showed several hydrophilic and hydro-
phobic regions in the active site of LmPTR1. The hydrophilic

216 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ferrari et al.
Figure 4. Docked configurations of [1,3,4]thiadiazole-2-ylamine, 4a, (blue) and 1,3-benzothiazole-2-ylamine, 1c, (yellow) at the active site of
LmPTR1. (A) Top view. Hydrophobic and hydrophilic regions are shown as in Figure 3. (B) Side view. Note the stacking interaction between
the nicotinamide ring of the cofactor (below) and the side chain of residue F113 (above).
pocket labeled i1 is above the ribose ring of the NADPH
cofactor (Figure 3). A group located in this region may be
involved in hydrogen bonds with nearby serine residues and
the ribose ring of NADPH in a similar way to that observed
in the X-ray structure of PTR1 bound to a substrate.26 The
large hydrophilic pocket labeled i2 protrudes into the core of
the protein and contains four rather conserved water mole-
cules. A small hydrophilic pocket i3 is at the opposite side of
the active site and contains one water molecule, which is
suggested to be a proton source in the second step of the
reduction reaction.27 The hydrophobic region o1 is at the
core of the active site surrounded by the previously men-
tioned hydrophilic regions. The hydrophobic region labeled
o2 extends from the core of the active site outward toward the
surface of the protein. Further identified regions marked
with dashed lines in Figure 3B are at the side of the active site
and not conserved among the analyzed X-ray structures of
the PTR1.
To investigate which binding regions might be most im-
portant to exploit to obtain selective inhibitors, we analyzed
the binding of inhibitor 2 in the context of the GRID maps.
Inhibitor 2 binds to both PTR1 and hDHFR, but crystal
structures show that its orientation differs in these two
proteins. In both cases, the diaminopteridine ring of 2 is
involved in stacking interactions between the nicotinamide
ring of NADPH and the aromatic ring of the protein residue,
whereas the rest of the molecule points out of the active site
with two carboxy groups exposed to the solvent. GRID
calculations for both the PTR1 and hDHFR structures show
that the hydrophilic patch, corresponding to i1, at the core of
the active site coincides with the location of the amino groups
of 2 and that the hydrophobic patch corresponding to o1
partially coincides with the second pteridine ring. In both
cases, the benzene ring of 2 coincides with the hydrophobic
patch corresponding to o2 at the exit of the active site.
However, hDHFR does not have a deep hydrophilic pocket
like i2 in PTR1. Therefore, an extension of the lead com-
pound toward this hydrophilic pocket might enhance selec-
tivity of the compounds toward PTR1. As seen from the
docked position of 4a in LmPTR1, the i1 region is already
occupied with the amino group and the nitrogen in position
4, which may be involved in hydrogen bonds; therefore, no
modifications were suggested on this side of the compound.
However, the other side of 4a is only partially inserted into
the hydrophobic region o1, and there is enough space for
extension with a hydrophobic group. Furthermore, this
region is sandwiched between the nicotinamide ring of the
cofactor at the bottom of the active site and the aromatic ring
of the Phe113 side chain from above. Therefore, we sug-
gested extension of 4a by adding a benzene ring, resulting in
the 1,3-benzothiazole-2-amine compound, 1c. Subsequent
docking of 1c showed that the thiazole ring of 1c was in
essentially the same position as that of the docked 4a. The
benzene ring overlapped better with the o1 hydrophobic
region and showed stacking interactions with the nicotin-
amide ring of the cofactor and the Phe113 side chain (Figure 4).
Taken together, these results suggest that the proposed 1,3-
benzothiazole-2-amino compound should exploit the phys-
icochemical properties of the core of the active site better.
This was confirmed by in vitro experiments in which 1c was
found to have greater inhibitory activity than 4a, with a Ki
value of 143 μM compared to 436 μM for 4a (Table 1).
Taking the 1,3-benzothiazole-2-amine, 1c, as a second
starting hit, we identified further extensions of compound
1c to better utilize the hydrophobic region o2 (Figure 3), and
the hydrophilic pocket i2. From comparison of the GRID
results for the hDHFR active site (not shown) and for the
LmPTR1 active site, better specificity of the compound
toward PTR1 relative to hDHFR could be gained through
these extensions. We searched for commercially available
compounds with a 1,3-benzothiazole-2-amine core and small
substituents that might extend into these regions of the active
site. Four compounds (2c-5c) were selected. Compound 5c
(6-trifluoromethoxybenzothiazol-2-ylamine) is also known
as riluzole.28
Testing of Benzothiazole Compounds. Five benzothiazole
compounds (1c-5c) were purchased and tested for their
effect on the activity of LmPTR1, LmDHFR, and hDHFR
(Table 1, Figure 1). All compounds inhibit LmPTR1 with
IC50 values in the range from 40 μM to 1.8 mM (Ki values
from 3 to 143 μM). Only two of them (2c and 4c) show a weak
inhibition of LmDHFR activity (with IC50 values of 1.4 and
1.8 mM, respectively), whereas three compounds (3c-5c)
inhibit hDHFR with IC50 values from 89 μM to 1 mM. The
best compounds, 3c and 5c, with IC50 values of 40 and 50 μM,
respectively, show a lower inhibitory activity against hDHFR
(IC50 values of 89 and 312 μM, respectively) and do not inhibit
LmDHFR.

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 217
Table 2. Effect of Different Compounds, Administered at 50 μg/mL, Alone or Combined with 1 at 30 μg/mL, on the Growth of Leishmania
promastigotesa
compound
growth of L. mexicana (%)
growth of L. major (%)
ED50 of human MRC5 fibroblasts (μg/mL)
1
7b
102.5 ( 0.3
111.6 ( 0.7
15.6 ( 0.5
111.6 ( 1.4
18.2 ( 1.5
131.9 ( 2.1
35.1 ( 10.9
96.9 ( 0.7
27.3 ( 8.1
4.7 ( 0.7
68.0 ( 1.2
105.6 ( 1.9
38.0 ( 0.7
83.5 ( 1.5
40.9 ( 0.2
76.0 ( 1.4
14.6 ( 0.9
85.1 ( 1.5
17.5 ( 0.2
7.1 ( 0.2
16.23 ( 2.47
10% inhibition at 100 μg/mL
ND
7b þ 1
14b
51.50 ( 5.52
ND
14b þ 1
15b
43.03 ( 5.15
ND
15b þ 1
21b
34.36 ( 3.72
ND
21b þ 1
28b
4.18 ( 0.43
ND
28b þ 1
5c
3.5 ( 0.2
21.5 ( 0.7
15.1 ( 1.4
4.1 ( 0.2
101.8 ( 1.8
22.9 ( 0.004
27.88 ( 3.17
ND
5c þ 1
a ED50 values vs human MRC5 fibroblast. Data are expressed as percentage of growth compared to control cultures to which no compound had been
added.The results for L. mexicana, L. major, and human cells are given as the mean ( SD from three independent experiments. ND = not detected.
Testing the Compounds on Cultured Parasites. Out of the
six compounds (7b, 14b, 15b, 21b, 28b, 5c) assayed on
cultured promastigotes (insect stage) of L. mexicana, only
28b and 5c were able, as a single agent at 50 μg/mL, to
reduce parasite growth compared to the control (Table 2).
They did so by 95.3% and 78.5%, respectively, (with corre-
sponding ED50 values of 10.1 μg/mL and 30.8 μg/mL).
None of the other compounds showed an inhibitory effect
at concentrations up to 100 μg/mL. With L. major prom-
astigotes, 14b, 15b, and 21b tested as single agents at 50 μg/mL
showed a moderate effect on parasite growth with a per-
centage of growth inhibition in the range 14.9-24.0%. Only
28b showed strong inhibition with a percentage of growth
inhibition above 90% (ED50 22.1 μg/mL) (Table 2). In
parallel experiments, the ED50 value for the control, the
approved antileishmaniasis drug, Amphotericin B, was
determined to be 0.1 μg/mL.
Figure 5. Growth of L. mexicana (in black) and L. major (in gray)
All compounds were tested in combination with 1, a
known antifolate drug with a therapeutic indication for
parasites in the presence of 30 μg/mL 1 and/or thiadiazole/
benzothiazole compounds at a concentration of 50 μg/mL. The
malaria that can be administered in combination with
sulfadoxine (trade name Fansidar). It is not usually active
against infections caused by trypanosomatids. When our
new compounds were tested in combination with 1,
growth inhibition of both L. major and L. mexicana prom-
astigotes was observed in all cases (Table 2, Figure 5).
1 alone, at 30 μg/mL, showed no effect on growth of L.
mexicana and reduced that of L. major by 32%. However,
1 in the presence of 50 μg/mL of each of the other com-
pounds caused considerable reduction of growth of the
parasites. For L. mexicana, the percentage of growth com-
pared to cells not treated with the compounds was between
15.6% (7b plus 1) and 35.1% (15b plus 1). For L. major, the
remaining growth was between 14.6% (15b plus 1) and
40.9% (14b plus 1). At the measured concentration, all these
compounds, except 28b and 5c which are active as a single
agent, show a synergy with 1 against L. mexicana and against
L. major.
The toxicity of the compounds was assayed on human
MRC-5 cell line derived from fibroblasts. The ED50 values
are indicated in Table 2. Almost all compounds showed
inhibition of growth of the MRC-5 cells at moderately toxic
levels with ED50 values ranging from 4 μg/mL (28b) to51μg/mL
(14b). 1 and 5c showed ED50 values of 16 μg/mL and 28
μg/mL, respectively. As a positive exception, compound 7b
showed very little inhibition of growth of MRC-5 cells, thus
indicating that this compound has low toxicity.
growth values are expressed as percentages calculated with respect
to the growth of parasites without 1 and thiadiazole/benzothia-
zole compounds.
Concluding Discussion
From the three rounds of drug discovery involving compu-
tational and experimental approaches, we have identified 18
nonfolate compounds with specific inhibitory activity against
LmPTR1. Some of them have been shown to inhibit the PTR1
protein at the low micromolar level (Table 1) and show
specificity with respect to the human DHFR enzyme. Five
compounds were found to be active in inhibiting parasite
growth when administered in combination with 1, a well-
know DHFR inhibitor. The compounds show synergetic
activity at the concentrations tested with the known DHFR
inhibitor, 1, indicating the value of inhibiting PTR1 and
DHFR simultaneously in order to block folate reduction in
Leishmania parasites. In particular, 7b showed over 85% of
parasite growth inhibition in combination with 1 and almost
no toxicity against the MRC5 cells suggesting better safety
profile with respect to the other compounds and 1 itself. Two
compounds, 28b and 5c, have also been shown to be active
when administered as single agents at 50 μg/mL. Comparison
of the toxicity and specificity profiles suggests that several of
the compounds identified in these screens, including 28b and
5c, provide a useful basis for furtherstudiestodiscover clinical
agents against leishmaniasis and possibly other parasitic diseases.

218 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ferrari et al.
Interestingly, 5c is a known clinical drug, riluzole, which is
used to treat amyotrophic lateral sclerosis with moderate
toxicity. It has also been proposed for the treatment of
mood and anxiety disorders, probably as glutamate re-
ceptor modulator.27
extracted from all PTR1 structures and aligned on chain A of the
structure 1E92 using the Chimera31 program. Residue Arg287
from chain D was also included because it is located close to the
active site of chain A. Water molecules and ions were removed,
hydrogens were added, and all structures were minimized using
the Maestro modeling environment.21
We searched in DrugBank (www.drugbank.ca) for other
benzothiazole derivatives or benzothiazole containing drugs
and found that there is only one drug currently in use:
Ethoxazolamide (DrugBank code: DB00311, PubChem CID:
3295). It is a carbonic anhydrase inhibitor used as a diuretic
and in glaucoma. These finding suggest that benzothiazole
compounds are drug-like molecules and prospective chemical
modulation of the decorating fragments could improve their
specificity as antiparasitic drugs and improve their safety
profile.
Structural Analysis and Comparison. The GRID program24,25
was used to determine energetically favorable binding sites for
different chemical groups in the region of the active site of the
PTR1 and hDHFR proteins in the presence of cofactor. A set of
probes representing hydrophobic and hydrophilic chemical
groups with hydrogen bond donor or acceptor properties were
˚
used for the calculations. The grid spacing was set to 0.5 A and
other parameters were assigned default values.
Docking Procedures. The GOLD v 3.219,20 and GLIDE21
programs were used for ligand docking. For GOLD, two fitness
functions, GoldScore and ChemScore,32 were used, and pro-
teins and ligands were input in Mol2 format. The binding site
Independently, Mpamhanga et al.12 have, by employing
different computational and medicinal chemistry approaches,
identified similar compounds, containing the aminobenzo-
thiazole scaffold, as inhibitors of Trypanosoma brucei PTR1.
Encouragingly, their crystal structures of 2-aminobenzimi-
dazole and 1-(3,4-dichloro-benzyl)-2-amino-benzimidazole in
complex with T. brucei PTR1 show the ligand in a similar
orientation in the binding site to that which we found in our
docking studies for similar compounds. We have conducted
crystallization trials for LmPTR1 with inhibitors identified in
this work (not shown), but we were unable to obtain any
suitable crystals of protein-inhibitor complexes. However,
unlike the compounds identified in ref 12, our compounds
show promising inhibitory activity in parasite promastigotes.
In conclusion, the compounds designed and tested here in
combination with 1 have potential for therapeutic application
against important neglected diseases.
˚
was defined as being within a sphere of radius of 20 A from the
hydroxyl O atom of Tyr194. The number of output docking
solutions was set to 10, and the other parameters were set to
default values. For GoldScore, tests were done allowing two
bumps between the ligand and the protein during docking. For
GLIDE, receptor grids were centered on the centroid of Tyr194.
Docking was performed with standard precision (SP), with the
options “dock flexibly” and “allow flips of 5 and 6 member
rings” enabled. Output was set to 10 poses per ligand after post-
docking minimization. The other parameters were set as default.
For cross-docking tests, the ligand structures were taken from
the crystal structures of LmPTR1. For docking other com-
pounds, the three-dimensional structures of the molecules were
built and energy minimized using Maestro21 or Corina.33 The
compounds were protonated using Epik.21 The compounds
were docked into the LmPTR1 structure with PDB ID: 1E92
(chain A plus Arg287 of chain D). Compounds were also docked
into two structures of hDHFR (PDB ID: 1KMS and 1U72) with
NADPH bound to aid in assessment of binding selectivity.
The WatCH34 program was used for the cluster analysis of the
crystallographic water sites in the superimposed LmPTR1 crys-
Materials and Methods
Virtual Screening. The structure of LmPTR1 (PDB ID: 1E92)
was considered for the virtual screening. The bound cofactor
NADPþ in its oxidized form was replaced by NADPH in the
reduced form, taken from the LmPTR1 complex (PDB ID:
1EW7). All water molecules and other ligands were removed
and not considered during the calculation. WHATIF29 and
AMBER 8.023 were used to add missing hydrogen atoms.
LUDI14 was used within the InsightII30 suite of programs to
perform the virtual screening of the ACD database (∼350 000
molecules) against the LmPTR1 three-dimensional structural
model. The parameters used were as follows: Energy-Estimate-3
˚
tal structures with a threshold of 2.4 A. To include the water
molecules in selected GOLD docking runs, two water sites were
set as “on” and two sites were set to “toggle”. During the dock-
ing, GOLD could automatically determine whether these spe-
cific waters were bound or not. The orientation of the waters was
optimized by GOLD with the option “spin”.
Tests of rescoring of the best pose found by GOLD were
performed by computing the free energy of binding of the
complex using AutoDock 4.022 with the parameter “epdb”. This
binding free energy includes the ligand intramolecular contribu-
tions as well as intermolecular electrostatic and desolvation
contributions. Rescoring with the AMBER force field35 was
also tested. 100 steps of steepest descent energy minimization of
the docked structures was performed with Amber8 using the
AMBER ff03 forcefield. The cofactor and protein atoms, except
hydrogens, were restrained using a harmonic force constant of
˚
as scoring function, 10 A as the radius and the centroid of the
bound natural substrate, dihydrobiopterin, as the center of the
sphere used to screen the molecules. The 21 394 results so ob-
tained were filtered based on three parameters computed for
each screened ligand: contact percentage (>50); number of H
bonds (>1); calculated score (>400). The calculated poses of
the resultant 724 ligands were visually analyzed. The analyzed
ligands were grouped based on the number and type of interac-
tions established and the residues with which they interact.
Molecules that were predicted to form a stacking interaction
with Phe113 and H bonds with active site residues were selected
preferentially. Visual comparison of the structures of the active
sites of LmPTR1 and hDHFR was used to prioritize the
predicted interactions to obtain more specific compounds.
Following this selection, 53 molecules were suggested for enzy-
matic screening.
2
˚
10 kcal/mol/A and the following parameter settings: imin=1,
maxcyc=100, ntmin=1, ncyc=100, nsnb=20, igb=0, ntb=0,
dielc = 1.0, ntpr = 10, ntr = 1, cut = 15.0, eedmeth = 5. The
Lennard-Jones and Coulombic contributions to the interaction
energy were then computed using ptraj.
Synthetic Chemistry. The reagents were purchased from
€
Sigma-Aldrich. Reactions were performed using Buchi’s Syn-
core parallel synthesizer. Reaction progress was monitored by
TLC on precoated silica gel 60 F254 plates (Merck) and visual-
ization was accomplished with UV light (254 nm). Yields of
these reactions referred to purified products were in the range
30-45%. Compounds 7b, 14b, 15b, and 21b were resynthesized
through microwave synthesis to obtain a larger amount for
growth inhibition tests with cultured parasites. Small-scale open-
vessel microwave reactions were conducted using a commercially
Preparation of Protein Crystal Structures. The X-ray crystal
structures of the PTR1 and hDHFR proteins were downloaded
from the PDB;17 see Supporting Information Table 4-SI for a
complete list of the structures used in this study. The PTR1
protein is a homotetramer; thus, the subunit labeled chain A was

Article
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 219
available single-mode microwave unit (CEM Discover). The
machine consists of a continuously focused microwave power
delivery system with operator selectable power output from 0 to
300 W. Reactions were performed in a 50 mL round-bottom
flask. The temperature of the contents of the vessel was mon-
itored using an IR sensor located underneath the reaction vessel.
The contents of the vessel were stirred by means of a rotating
magnetic plate located below the floor of the microwave cavity
and a Teflon-coated magnetic stir bar in the flask. Temperature
and power profiles were monitored using commercially avail-
able software provided by the microwave manufacturer. Yields
of these reactions referred to purified products were in the range
55-70%. Purification by flash column chromatography was
conducted using Sigma-Aldrich SilicaGel (grade 9385, pore size
(s, 1H). Elemental analysis: calcd for C11H7N3O2S, C, 53.87;
H, 2.88; N, 17.13. Found: C, 53.69; H, 2.88; N, 17.11.
3-(5-Amino-[1,3,4]thiadiazol-2-yl)-1-(thiophen-2-yl)propan-1-
one (21b). 1H NMR δ 7.90 (dd, J=3.8, 1.1 Hz, 1H), 7.83 (dd, J=
5.0, 1.1 Hz, 1H), 7.20 (t, J=3.8 Hz, 1H), 3.45 (dt, J=8.2, 1.2 Hz,
2H), 3.00 (dt, J=7.0, 0.4 Hz, 2H). Elemental analysis: calcd for
C9H9N3OS2, C, 45.17; H, 3.79; N, 17.56. Found: C, 45.18; H,
3.79; N, 17.55.
3-(5-Amino-[1,3,4]thiadiazol-2-yl)-1-(4-chlorophenyl)-propan-
1
1-one (22b). H NMR δ 7.99 (d, J=8.8 Hz, 2H), 7.60 (d, J=
8.8 Hz, 2H), 3.24 (t, J=6.3 Hz, 2H), 2.57 (t, J=6.3 Hz, 2H).
Elemental analysis: calcd for C11H10ClN3OS, C, 49.35; H, 3.76;
N, 15.69. Found: C, 49.40; H, 3.75; N, 15.65.
3-Chloro-N-[5-(2-chloroethyl)-[1,3,4]thiadiazol-2-yl]-propionamide
(28b). 1H NMR δ 4.01 (t, J=6.5 Hz, 2H), 3.92 (t, J=6.5 Hz,
2H), 3.48 (t, J=6.5 Hz, 2H), 3.00 (t, J=6.5 Hz, 2H). Elemental
analysis: calcd for C7H9Cl2N3OS, C, 33.08; H, 3.57; N, 16.53.
Found: C, 33.34; H, 3.70; N, 16.69.
2-(1,3-Dioxoisoindolin-2-yl)-N-5-(1,3-dioxoisoindolin-2-ylmethyl)-
[1,3,4]thiadiazol-2-ylacetamide (29b). 1H NMR δ8.0-7.8 (m, 8H),
5.14 (s, 2H), 4.60 (s, 2H). Elemental analysis: calcd for
C21H13N5O5S, C, 56.37; H, 2.93; N, 15.65. Found: C, 54.38; H,
2.93; N, 15.68.
˚
of 60 A, 230-400 mesh). Purity of all compounds was deter-
mined to be at least 95% from TLC and elemental analyses. This
analysis was performed on a Perkin-Elmer 240C instrument,
and the results for C, H, and N were within (0.4% of theoretical
values. The synthesized compounds were characterized by 1HNMR
on a Bruker FT-NMR AVANCE 400. Spectra were recorded in
DMSO-d6. Chemical shifts are reported as δ values (ppm)
referenced to tetramethylsilane as an internal standard. When
peak multiplicities are given, the following abbreviations are
used: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; dd
double doublet; dt double triplet; dq double quartet; br, broad-
ened signal.
Enzymology. Proteins were purified as described.36-44 Folate
cofactors and substrates were a gift from Eprova; all other
substrates, cofactors, and reagents were purchased from differ-
ent companies at the maxiumum purity grade. Compounds
1a-53a and 1c-5c and 1 were purchased from different vendors
and purity of key compounds evaluated to be higher than 95%
(the elemental analysis is reported in Supporting Information).
For determination of reductase activity, NADPH oxidation was
followed at 340 nm.45 Km values were determined by measuring
the dependence of the enzyme activity on substrate concentra-
tion using folic acid as the substrate. The Km value was 2.5 μM
for LmPTR1. Kinetic measurements with DHFR were per-
formed at 25 ꢀC in standard enzyme buffer.46 Km values were
determined by measuring the dependence of the enzyme activity
on substrate concentration using dihydrofolate as the substrate.
The Km values of dihydrofolate for hDHFR and LmDHFR in
this buffer were determined to be 7 μM and 3.5 μM, respectively.
The kinetic experiments were carried out in triplicate, and no
individual measurement differed by >20% from the mean.
All tested compounds (1a-53a, 1b, 3b-5b, 7b-11b, 13b-
29b, 1c-5c) were dissolved in DMSO. Compounds 1a-53a were
initially screened for inhibitory activity against LmPTR1 at a
single concentration (500 μM). Compounds 1b, 3b-5b, 7b-11b,
and 13b-29b were initially screened for inhibition of LmPTR1
activity at a single concentration (5 μM). For those compounds
which were found to be active against LmPTR1, a complete set
of measurements was performed to determine the IC50 values.
For compounds 1c-5c, the IC50 value against LmPTR1 was
obtained. Ki values were obtained from IC50 plots by assuming
competitive inhibition.47,48 For compound 15b, the Ki value was
determined by Lineweaver-Burk analysis of multiple substrate
and inhibitor concentrations and showed a competitive inhibi-
tion pattern with respect to the dihydrofolate cofactor. The
result was consistent with the value determined from the IC50
plots (data not shown).
General Procedure for the in-Parallel Synthesis. To an ice-
cooled mixture of thiosemicarbazide (0.150 g, 1.65 mmol) and
the corresponding carboxylic acid (1.65 mmol), an excess of
phosphorus oxychloride (0.3 mL, 3.3 mmol) was added slowly
and under continuous stirring. Subsequently, the temperature
was raised gradually to 75-80 ꢀC. The reaction was kept at this
temperature and stirred for 3 h. Ice-water was added, and the
mixture was stirred for an additional hour and the solvent
removed by filtration under reduced pressure. The residual pre-
cipitate was suspended in water and 7 mL of 20% sodium bicar-
bonate solution was added to remove the unreacted carboxylic
acid. Afterward, the mixture was filtrated under vacuum and the
obtained solid was treated with an acidic or basic solution to
remove the starting material. For compounds 2b, 7b, 14b, 15b,
and 17b, chromatography purification (CH2Cl2/MeOH 9:1) was
performed. The obtained solid was dried in vacuo.
General Procedure for the Microwave Synthesis. To an ice-
cooled mixture of thiosemicarbazide (0.30 g, 3.3 mmol) and the
corresponding carboxylic acid (2.2 mmol), an excess of phos-
phorus oxychloride (0.6 mL, 6.6 mmol) was added slowly and
under continuous stirring. The flask was placed in the micro-
wave cavity. Microwave irradiation of 50 W was used, the tem-
perature being ramped from room temperature to 75 ꢀC. Once
75 ꢀC was reached, the reaction mixture was held at this tem-
perature for 10 min. The glue was allowed to cool to room
temperature and manually stirred; the microwave irradiation
was repeated. Ice-water was added and the mixture was stirred
for an hour. NaOH (3 M) was added to the mixture until pH 11
was reached. The volume was reduced under vacuum, and the
mixture of products isolated by filtration was purified and
separated by chromatography column (CH2Cl2/MeOH 9:1).
3-(5-Amino-[1,3,4]thiadiazol-2-yl)-pyridin-4-ylamine (7b).
1H NMR δ 8.28 (s, 1H), 7.97 (d, J = 5.7 Hz, 1H), 7.4 (br,
4H), 6.72 (d, J = 5.7 Hz, 1H). Elemental analysis: calcd for
C7H7N5S,C, 43.51; H, 3.65; N, 36.24. Found: C, 43.50; H,
3.65; N, 36.25.
Compounds 4a, 6a, 28a, 35a, 38a, 53a, 7b, 14b, 15b, 21b,
22b, 28b, 29b, 1c-5c were screened for their activity against
LmDHFR and hDHFR. The compounds were screened at a
single concentration in the range of 25 μM to 1 mM. For those
compounds which were found to be active, an IC50 value was
estimated by considering a linear relation between the percent-
age of inhibition and concentration.
No incubation effect was detected for any compound. All
compounds discussed here behaved well kinetically and do not
fall into the category of aggregation-based promiscuous
inhibitors.49 The DMSO concentration was kept below the
concentration affecting enzyme activity (1% for PTR1).
5-(1H-benzo[d][1,2,3]triazol-5-yl)-[1,3,4]thiadiazol-2-ylamine
(14b). 1H NMR δ 8.20 (s, 1H), 7.99 (d, J=8.8 Hz, 1H), 7.93 (d,
J = 8.7 Hz, 1H), 7.47 (br, 2H). Elemental analysis: calcd for
C8H6N6S, C, 44.03; H, 2.77; N, 38.51. Found: C, 44.17; H, 2.77;
N, 38.49.
2-(5-Amino-[1,3,4]thiadiazol-2-yl)-4H-chromen-4-one (15b).
1H NMR δ 7.90 (d, J=8.2 Hz, 1H), 7.75 (d, J=7.8 Hz, 1H),
7.50 (t, J = 7.8 Hz, 1H), 7.49 (dd, J = 8.2, 7.8 Hz, 1H), 6.93

220 Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1
Ferrari et al.
Biological Evaluation. Parasitology. Promastigote forms of L.
mexicana (MHOM/BZ/84/BEL46) and L. major (MHOM/SU/
73/5-ASKH) were cultured in SDM-79 medium, pH 7.3, sup-
plemented with 15% heat-inactivated fetal calf serum, penicillin
(100 units/mL), and streptomycin (100 mg/mL) at 28 ꢀC under
water-saturated air with 5% CO2 as described earlier.50 The
cultures were initiated at 105 parasites/mL, and cells were
harvested at a density of 2 ꢀ 107 parasites/mL.
Compounds 1b, 3b-5b, 7b-11b, 13b-29b and their enzyme
inhibition activity; Table 4-SI, Protein structures used in this
study; Elemental analysis of the key compounds 4a, 6a, 28a, 35a,
38a, 53a, 1c-5c, and 1, and characterization of 1b-5b, 8b-13b,
16b-20b, 23b-27b. This material is available free of charge via
the Internet at http://pubs.acs.org.
References
To estimate the concentrations at which compounds cause
50% inhibition of growth (effective dose, ED50) of cultured
Leishmania promastigotes, the Alamar Blue micromethod,
based on monitoring the reducing environment of proliferating
cells, was used as previously described.51 Briefly, cultures were
diluted in medium to 1 ꢀ 106 parasites/mL and seeded in 96-well
flat-bottom microplates (Nunc) to a final volume of 100 μL.
Inhibitor stock solutions were in DMSO. For each compound,
dilutions were made in culture medium and added to the parasite
cultures giving a series of concentrations starting from 100 μg/mL
downward. The final DMSO concentration in the cultures was
always less than 1%. Each inhibitor concentration was tested in
duplicate. Controls with DMSO alone in the medium (which
never caused any growth inhibition at the concentration used)
and the parasites without compounds were also tested. After
72 h of incubation at 28 ꢀC, 10 μL Alamar Blue was added to
each well, and after 4 h incubation at 28 ꢀC, its fluorescence was
quantified at an excitation wavelength of 530 nm and at an
emission wavelength of 590 nm. ED50 values were calculated by
linear interpolation. The optical density in the absence of drugs
was set as the 100% control. Assays with the commercial
antileishmaniasis drug, Amphotericin B, were also carried out
in order to have reference values. Three independent experi-
ments, each in duplicate, were performed for the determination
of the ED50 value of each compound.
In separate experiments, compounds were tested in combina-
tion with 1 to evaluate the synergic effect. Combinations were
tested with each compound at a single dose of 50 μg/mL and 1 at
30 μg/mL, in duplicate. The percentage of growth inhibition of
promastigotes in the presence of the combination of each
compound with 1 was compared with controls in which only
the compound was added at 50 μg/mL and with controls without
the compound and 1.
Toxicity Test. The fibroblast cell line MRC-5 was used to
determine the possible toxicity of compounds for human cells.
The fibroblasts were grown in DMEM medium (Gibco) supple-
mented with 10% heat-inactivated FBS and extra glutamine, at
37 ꢀC in humidified incubators in an atmosphere of 5% CO2.
The assay was performed using essentially the same method as
used for parasites. To each well of the microplates, without or
with compound at different concentrations, cells were added at a
density such that, after 72 h of incubation, adhesive cells have
formed a confluent monocellular film in control wells. Alamar
Blue was added to each well after 72 h and used to determine cell
growth, similarly to that described above for parasites.
(1) Sundar, S.; Makharia, A.; More, D. K.; Agrawal, G.; Voss, A.;
Fischer, C.; Bachmann, P.; Murray, H. W. Short-course of oral
miltefosine for treatment of visceral leishmaniasis. Clin. Infect. Dis.
2000, 31, 1110–1113.
(2) Then, R. L. Antimicrobial dihydrofolate reductase inhibitors--
achievements and future options: review. J. Chemother. 2004, 16,
3–12.
(3) McGuire, J. J. Anticancer antifolates: current status and future
directions. Curr. Pharm. Des. 2003, 9, 2593–2613.
(4) Gregson, A.; Plowe, C. V. Mechanisms of resistance of malaria
parasites to antifolates. Pharmacol. Rev. 2005, 57, 117–145.
(5) Nare, B.; Hardy, L. W.; Beverley, S. M. The roles of pteridine
reductase 1 and dihydrofolate reductase-thymidylate synthase in
pteridine metabolism in the protozoan parasite Leishmania major.
J. Biol. Chem. 1997, 272, 13883–13891.
(6) Senkovich, O.; Pal, B.; Schormann, N.; Chattopadhyay, D. Try-
panosoma cruzi genome encodes a pteridine reductase 2 protein.
Mol. Biochem. Parasitol. 2003, 127, 89–92.
(7) Cunningham, M. L.; Titus, R. G.; Turco, S. J.; Beverley, S. M.
Regulation of differentiation to the infective stage of the protozoan
parasite Leishmania major by tetrahydrobiopterin. Science 2001,
292, 285–287.
(8) Nare, B.; Luba, J.; Hardy, L. W.; Beverley, S. New approaches to
Leishmania chemotherapy: pteridine reductase 1 (PTR1) as a
target and modulator of antifolate sensitivity. Parasitology 1997,
114 Suppl, S101-S110.
(9) Bello, A. R.; Nare, B.; Freedman, D.; Hardy, L.; Beverley, S. M.
PTR1: a reductase mediating salvage of oxidized pteridines and
methotrexate resistance in the protozoan parasite Leishmania
major. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 11442–11446.
(10) Cavazzuti, A.; Paglietti, G.; Hunter, W. N.; Gamarro, F.; Piras, S.;
Loriga, M.; Allecca, S.; Corona, P.; McLuskey, K.; Tulloch, L.;
Gibellini, F.; Ferrari, S.; Costi, M. P. Discovery of potent pteridine
reductase inhibitors to guide antiparasite drug development. Proc.
Natl. Acad. Sci. U.S.A. 2008, 105, 1448–1453.
(11) Tulloch, L. B.; Martini, V. P.; Iulek, J.; Huggan, J. K.; Lee, J. H.;
Gibson, C. L.; Smith, T. K.; Suckling, C. J.; Hunter, W. N.
Structure-based design of pteridine reductase inhibitors targeting
african sleeping sickness and the leihmaniases. J. Med. Chem. 2010,
53, 221–229.
(12) Mpamhanga, C. P.; Spinks, D.; Tulloch, L. B.; Shanks, E. J.;
Robinson, D. A.; Collie, I. T.; Fairlamb, A. H.; Wyatt, P. G.;
Frearson, J. A.; Hunter, W. N.; Gilbert, I. H.; Brenk, R. One
scaffold, three binding modes: novel and selective pteridine reduc-
tase 1 inhibitors derived from fragment hits discovery by virtual
screening. J. Med. Chem. 2009, 52, 4454–4465.
(13) Russell, P.; Hitching, G. Aryl derivatives of 2,4-diaminopyrimidines.
J. Am. Chem. Soc. 1951, 73, 3763–3770.
(14) Boehm, H. J. The computer program LUDI: a new method for the
de novo design of enzyme inhibitors. J. Comput.-Aided. Mol. Des.
1992, 6, 61–78.
(15) Mishra, H.; Singh, N.; Lahiri, T.; Misra, K. A comparative study
on the molecular descriptors for predicting drug-likeness of small
molecules. Bioinformation 2009, 3, 384–388.
(16) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Experimental and computational approaches to estimate solubility
and permeability in drug discovery and development settings. Adv.
Drug Delivery Rev. 2001, 46, 3–26.
Acknowledgment. We gratefully acknowledge support of
the Italian Internationalization Programme (2006-2008) and
Cassa di Risparmio di Modena Internationalization Programme
(Kinetodrugs project, 2008-2010) and the Klaus Tschira
Foundation. We are grateful to the ‘‘Centro Interdipartimen-
tale Grandi Strumenti’’ of the University of Modena and
Reggio Emilia and to the ‘‘Fondazione Cassa di Risparmio di
Modena’’ for supplying the NMR spectrometer. We thank
Prof. L. Brasili for allowing us to use the single mode micro-
wave unit, and Merck-Eprova-AG (www.eprova.com) for
supplying of folate cofactors and substrates.
(17) www.pdb.org.
(18) Seegerd, D.; Cosulichi, O.; Smith, J., Jr.; Hultquist, M. 4-Amino
derivatives of pteroylglutamic acid. J. Am. Chem. Soc. 1949, 71,
1753–1758.
(19) Jones, G.; Willett, P.; Glen, R. C. Molecular recognition of
receptor sites using a genetic algorithm with a description of
desolvation. J. Mol. Biol. 1995, 245, 43–53.
(20) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.; Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J. Mol. Biol. 1997, 267, 727–748.
€
(21) Maestro, Glide (version 4.5), and Epik, Schrodinger, LLC, New York,
Supporting Information Available: Table 1-SI, Compounds
1a-53a and their enzyme inhibition activity; Table 2-SI, Com-
pounds 1b-25b and their molecular properties; Table 3-SI,
NY, 2008.
(22) Morris, G. M.; Huey, R.; Lindstrom, W.; Sanner, M. F.; Belew, R. K.;
Goodsell, D. S.; Olson, A. J. AutoDock4 and AutoDockTools4:

Article
automated docking with selective receptor flexibility. J. Comput.
Journal of Medicinal Chemistry, 2011, Vol. 54, No. 1 221
methotrexate-resistant strain of Lactobacillus casei. Biochem. J.
1976, 157, 559–571.
(37) Hanahan, D. Studies on transformation of Escherichia coli with
Chem. 2009, 30, 2785–2791.
(23) Case, D. A.; Cheatham, T. E., III; Darden, T.; Gohlke, H.; Luo, R.;
Merz, K. M., Jr.; Onufriev, A.; Simmerling, C.; Wang, B.; Woods,
R. J. The Amber biomolecular simulation programs. J. Comput.
Chem. 2005, 26, 1668–1688.
(24) Goodford, P. J. A computational procedure for determining
energetically favorable binding sites on biologically important
macromolecules. J. Med. Chem. 1985, 28, 849–857.
(25) Molecular Discovery; http://www.moldiscovery.com/
(26) Gourley, D. G.; Schuttelkopf, A. W.; Leonard, G. A.; Luba, J.;
Hardy, L. W.; Beverley, S. M.; Hunter, W. N. Pteridine reductase
mechanism correlates pterin metabolism with drug resistance in
trypanosomatid parasites. Nat. Struct. Biol. 2001, 8, 521–525.
(27) Arun, P.; Moffett, J. R.; Namboodiri, A. M. A. Riluzole decrease
synthesis of N-Acetylaspartate and N-acetylaspartylglutamate
in SH-SY5Y human beuroblastoma cells. Brain Res. 2010, 1334,
25–30.
plasmids. J. Mol. Biol. 1983, 166, 557–580.
(38) Meek, T. D.; Garvey, E. P.; Santi, D. V. Purification and char-
acterization of the bifunctional thymidylate synthetase-dihydrofo-
late reductase from methotrexate-resistantLeishmania tropica.
Biochemistry 1985, 24, 678–686.
(39) Maley, G. F.; Maley, F. Properties of a defined mutant of Escheri-
chia coli thymidylate synthase. J. Biol. Chem. 1988, 263, 7620–7627.
(40) Santi, D. V.; Edman, U.; Minkin, S.; Greene, P. J. Purification and
characterization of recombinant Pneumocystis carinii thymidylate
synthase. Protein Expr. Purif. 1991, 2, 350–354.
(41) Livi, L. L.; Edman, U.; Schneider, G. P.; Greene, P. J.; Santi, D. V.
Cloning, expression and characterization of thymidylate synthase
from Cryptococcus neoformans. Gene 1994, 150, 221–226.
(42) Pedersen Lane, J.; Maley, G. F.; Chu, E.; Maley, F. High level
expression of human thymidylate synthase. Protein Expr. Purif.
1997, 10, 256–262.
(43) Robello, C.; Navarro, P.; Castanys, S.; Gamarro, F. A pteridine
reductase gene ptr1 contiguous to a P-glycoprotein confers resis-
tance to antifolates in Trypanosoma cruzi. Mol. Biochem. Para-
sitol. 1997, 90, 525–535.
(44) Gourley, D. G.; Luba, J.; Hardy, L. W.; Beverley, S. M.; Hunter,
W. N. Crystallization of recombinant Leishmania major pteridine
reductase 1 (PTR1). Acta Crystallogr., Sect. D 1999, 1608–1610.
(45) Bello, A. R.; Nare, B.; Freedman, D.; Hardy, L.; Beverley, S. M.
PTR1: a reductase mediating salvage of oxidized pteridines and
methotrexate resistance in the protozoan parasite Leishmania
major. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 11442–11446.
(46) Tan, X. H.; Huang, S. M.; Ratnam, M.; Thompson, P. D.;
Freisheim, J. H. The importance of loop region residues 40-46
in human dihydrofolate reductase as revealed by site-directed
mutagenesis. J. Biol. Chem. 1990, 265, 8027–8032.
(47) Chou, T. Relationships between inhibition constants and frac-
tional inhibition in enzyme-catalyzed reactions with different
numbers of reactants, different reaction mechanisms, and different
types and mechanisms of inhibition. Mol. Pharmacol. 1974, 10,
235–247.
(28) Jimonet, P.; Audiau, F.; Barreau, M.; Blanchard, J.-C.; Boireau,
ꢁ
A.; Bour, Y.; Coleno, M.-A.; Doble, A.; Doerflinger, G.; Do Huu,
ꢁ ꢁ
C.; Donat, M.-H.; Duchesne, J. M.; Ganil, P.; Gueremy, C.;
ꢁ
Honore, E.; Just, B.; Kerphirique, R.; Gontier, S.; Hubert, P.;
Laduron, P. M.; Le Blevec, J.; Meunier, M.; Miquet, J.-M.;
Nemecek, C.; Pasquet, M.; Piot, O.; Pratt, J.; Rataud, J.; Reibaud,
M.; Stutzmann, J.-M.; Mignani, S. J. Med. Chem. 1999, 42, 2828–2843.
(29) Vriend, G. WHAT IF: a molecular modeling and drug design
program. J. Mol. Graph. 1990, 8, 52–56.
(30) Insight_II. (2000). Molecular Simulations Inc., San Diego, CA. Accelrys.
(31) Pettersen, E. F.; Goddard, T. D.; Huang, C. C.; Couch, G. S.;
Greenblatt, D. M.; Meng, E. C.; Ferrin, T. E. UCSF Chimera--
a visualization system for exploratory research and analysis. J. Comput.
Chem. 2004, 25, 1605–1612. http://www.cgl.ucsf.edu/chimera.
(32) Eldridge, M. D.; Murray, C. W.; Auton, T. R.; Paolini, G. V.; Mee,
R. P. Empirical scoring functions .I. the development of a fast
empirical scoring function to estimate the binding affinity of
ligands in receptor complexes. J. Comput.-Aided Mol. Des. 1997,
11, 425–445.
(33) Sadowski, J.; Rudolph, C.; Gasteiger, J. Automatic generation of
3D atomic coordinates for organic molecules. Tetrahedron Com-
put. Methodol. 1990, 3, 537–547.
(34) Sanschagrin, P. C.; Kuhn, L. A. Cluster analysis of consensus water
sites in thrombin and trypsin shows conservation between serine
proteases and contributions to ligand specificity. Protein Sci. 1998,
7, 2054–2064.
(48) Segel, I. H. Enzyme Kinetics: Behaviour and Analysis of Rapid
Equilibrium and Steady-State Enzyme Systems; Wiley: New York,
1975.
(49) McGovern, S. L.; Caselli, E.; Grigorieff, N.; Shoichet, B. K. A
common mechanism underlying promiscuous inhibitors from vir-
tual and high-throughput screening. J. Med. Chem. 2002, 45, 1712–
1722.
(35) Duan, Y.; Wu, C.; Chowdhury, S.; Lee, M. C.; Xiong, G.; Zhang,
W.; Yang, R.; Cieplak, P.; Luo, R.; Lee, T.; Caldwell, J.; Wang, J.;
Kollman, P. A point-charge force field for molecular mechanics
simulations of proteins based on condensed-phase quantum me-
chanical calculations. J. Comput. Chem. 2003, 24, 1999–2012.
(36) Dann, J. G.; Ostler, G.; Bjur, R. A.; King, R. W.; Scudder, P.;
Turner, P. C.; Roberts, G. C.; Burgen, A. S. Large scale puri-
fication and characterization of dihydrofolate reductase from a
€
(50) Brun, R.; Schonenberger, M. Cultivation and in vitro cloning or
procyclic culture forms of Trypanosoma brucei in a semi-defined
medium. Acta Trop. 1979, 36, 289–292.
(51) Mikus, J.; Steverding, D. A simple colorimetric method to screen
drug cytotoxicity against Leishmania using the dye Alamar Blue.
Parasitol. Int. 2000, 48, 265–269.