Profiling of flavonol derivatives for the development of antitrypanosomatidic drugs

Article abstract of DOI:10.1021/acs.jmedchem.6b00698

Flavonoids represent a potential source of new antitrypanosomatidic leads. Starting from a library of natural products, we combined target-based screening on pteridine reductase 1 with phenotypic screening on Trypanosoma brucei for hit identification. Flavonols were identified as hits, and a library of 16 derivatives was synthesized. Twelve compounds showed EC50 values against T. brucei below 10 μM. Four X-ray crystal structures and docking studies explained the observed structure-activity relationships. Compound 2 (3,6-dihydroxy-2-(3-hydroxyphenyl)-4H-chromen-4-one) was selected for pharmacokinetic studies. Encapsulation of compound 2 in PLGA nanoparticles or cyclodextrins resulted in lower in vitro toxicity when compared to the free compound. Combination studies with methotrexate revealed that compound 13 (3-hydroxy-6-methoxy-2-(4-methoxyphenyl)-4H-chromen-4-one) has the highest synergistic effect at concentration of 1.3 μM, 11.7-fold dose reduction index and no toxicity toward host cells. Our results provide the basis for further chemical modifications aimed at identifying novel antitrypanosomatidic agents showing higher potency toward PTR1 and increased metabolic stability.

Full text of DOI:10.1021/acs.jmedchem.6b00698

Subscriber access provided by UNIV LAVAL
Article
Profiling of flavonol derivatives for the
development of anti-trypanosomatidic drugs
Chiara Borsari, Rosaria Luciani, Cecilia Pozzi, Ina Pöhner, Stefan Henrich, Matteo Trande, Anabela Cordeiro-
da-Silva, Nuno Santarém, Catarina Baptista, Annalisa Tait, Flavio Di Pisa, Lucia Dello Iacono, Giacomo
Landi, Sheraz Gul, Markus Wolf, Maria Kuzikov, Bernhard Ellinger, Jeanette Reinshagen, Gesa Witt, Philip
Gribbon, Manfred Kohler, Oliver Keminer, Birte Behrens, Luca Costantino, Paloma Tejera Nevado, Eugenia
Bifeld, Julia Eick, Joachim Clos, Juan Torrado, María Dolores Jiménez-Antón, Maria Jesùs Corral, José
Maria Alunda, Federica Pellati, Rebecca C Wade, Stefania Ferrari, Stefano Mangani, and Maria Paola Costi
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00698 • Publication Date (Web): 14 Jul 2016
Downloaded from http://pubs.acs.org on July 15, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pharmacy
Jiménez-Antón, María; Dpto Farmacia y Tecnología Farmacéutica, Facultad
de Farmacia, Complutense University of Madrid, Plaza Ramón y Cajal,
Pharmacy
Corral, Maria; Dpto Farmacia y Tecnología Farmacéutica, Facultad de
Farmacia, Complutense University of Madrid, Plaza Ramón y Cajal,
Pharmacy
Alunda, José; Dpto Farmacia y Tecnología Farmacéutica, Facultad de
Farmacia, Complutense University of Madrid, Plaza Ramón y Cajal,
Pharmacy
Pellati, Federica; University of Modena and Reggio Emilia, Department of
Life Sciences
Wade, Rebecca; HITS gGmbH, Molecular and Cellular Modeling
Ferrari, Stefania; Università degli Studi di Modena e Reggio Emilia,
Dipartimento di Scienze della Vita
Mangani, Stefano; University of Siena, Biotechnology Chemistry and
Pharmacy
Costi, Maria Paola; Universita Di Modena e Reggio Emilia, Dipartimento Di
Scienze della Vita
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Profiling of flavonol derivatives for the development of antiꢀtrypanosomatidic drugs
1^
1^
2^
3^
3
Borsari, Chiara ; Luciani,Rosaria ; Pozzi, Cecilia ; Poehner, Ina ; Henrich, Stefan ; Matteo,
1
4
4
4
1
Trande , CordeiroꢀdaꢀSilva, Anabela ; Santarem, Nuno ; Baptista, Catarina ; Tait, Annalisa ; Di Pisa,
2
2
2
5
5
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Flavio ; Dello Iacono Lucia , Landi, Giacomo ; Gul, Sheraz ; Wolf, Markus ; Kuzikov, Maria ;
5
5
5
5
5
Ellinger, Bernhard ; Reinshagen, Jeanette ; Witt, Gesa ; Gribbon, Philip ; Manfred Kohler ; Keminer,
5
5
1
9
9
9
Oliver ; Behrens,Birte ; Costantino, Luca ; Tejera Nevado, Paloma ; Bifeld, Eugenia ; Eick, Julia ;
9
6
6,10
6,10
Clos, Joachim ; Torrado, Juan ; JiménezꢀAntón, María D. ; Corral, María J.
; Alunda, José
6
,10
1
3,7,8
1*
2*
Mª ; Federica, Pellati ; Wade, Rebecca C. ; Ferrari, Stefania ; Mangani, Stefano ; Costi, Maria
1*
Paola .
1
Department of Life Sciences, University of Modena and Reggio Emilia, Italy, Via G. Campi 103, 41125
Modena, Italy.
2
Department of Biotechnology, Chemistry and Pharmacy,University of Siena, Italy, Via Aldo Moro 2, 53100
Siena, Italy.
3
Molecular and Cellular Modeling Group, Heidelberg Institute for Theoretical Studies, 69118 Heidelberg,
Germany.
4
Institute for Molecular and Cell Biology, Porto, Portugal.
5
Fraunhofer Institute for Molecular Biology and Applied Ecology ScreeningPort, Schnackenburgallee 114
D-22525, Hamburg, Germany.
6
Complutense University of Madrid, 28040 Madrid, Spain.
7
Center for Molecular Biology (ZMBH), DKFZ-ZMBH Alliance, Heidelberg University, 69120 Heidelberg,
Germany.
8
Interdisciplinary Center for Scientific Computing (IWR), Heidelberg University,69120 Heidelberg, Germany.
9
Bernhard Nocht Institute for Tropical Medicine, D-20359 Hamburg, Germany.
10
Instituto de Investigación Hospital 12 de Octubre, 28041 Madrid, Spain.
KEYWORDS: Neglected diseases, flavonols, antiparasitic activity, pteridine reductase 1.
1
ACS Paragon Plus Environment

Page 3 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Abstract
Flavonoids represent a potential source of new antiꢀtrypanosomatidic leads. Starting from a library of
natural products, we combined targetꢀbased screening on pteridine reductase 1 with phenotypic
screening on Trypanosoma brucei, for hit identification. Flavonols were identified as hits and a library
of 16 derivatives was synthesized. Twelve compounds showed EC50 values against T. brucei below 10
µM. Four Xꢀray crystal structures and docking studies explained the observed structureꢀactivity
relationships. Compound 2 (3,6ꢀdihydroxyꢀ2ꢀ(3ꢀhydroxyphenyl)ꢀ4Hꢀchromenꢀ4ꢀone) was selected for
pharmacokinetic studies. Encapsulation of compound 2 in PLGA nanoparticles or cyclodextrins
resulted in lower in-vitro toxicity when compared to the free compound. Combination studies with
methotrexate revealed that compound 13 (3ꢀhydroxyꢀ6ꢀmethoxyꢀ2ꢀ(4ꢀmethoxyphenyl)ꢀ4Hꢀchromenꢀ
4ꢀone) has the highest synergistic effect at concentration of 1.3ꢀM, 11.7 folds dose reduction index
and no toxicity towards host cells. Our results provide the basis for further chemical modifications
aimed at identifying novel antitrypanosomatidic agents showing higher potency towards PTR1 and
increased metabolic stability.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Introduction
Protozoan parasites of the Trypanosomatidae family are the etiological agents of several significant
neglected tropical diseases including Human African Trypanosomiasis (HAT), Chagas’ disease and
Leishmaniasis, which collectively affect nearly 10 million people worldwide. HAT is caused by the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
bloodstream form of Trypanosoma brucei. Leishmania spp. infect macrophages and cause a wide
2
spectrum of symptoms ranging from cutaneous lesions to potentially fatal visceral infections.
Trypanosoma cruzi, the etiological agent of Chagas disease, is an obligate intracellular parasite that
invades different internal organs. Since effective vaccines are lacking, the control of these diseases is
based on vector control and chemotherapy. However, the drugs currently available are characterized
by high toxicity, limited efficacy and require a long period of treatment. The first line drugs have been
3
ꢀ5
used for over half a century and their efficacy is compromised by widespread resistance. Therefore,
there is an urgent requirement for new, safe and effective drugs. The drug discovery process against
these parasites is hampered by the intrinsic biology of these organisms that can include intracellular
6
stages or central nervous system involvement.
Although there are very few validated drug targets for Leishmaniasis and HAT, a targetꢀbased
7
approach is extensively used in the drug discovery process for neglected tropical diseases. The folate
pathway is an established target for the treatment of bacterial infections and some parasitic diseases,
8
such as malaria. Drugs targeting folateꢀdependent enzymes represent useful candidates against
trypanosomatidic infections and key enzymes of the folate metabolism are thymidylate synthase (TS)
and dihydrofolate reductase (DHFR), which is the target of the antifolate drugs methotrexate (MTX),
pyrimethamine (PYR) and trimethoprim (TMP). However, the activity of classical inhibitors of
9
DHFR against Leishmania and Trypanosoma is reduced because of pteridine reductase 1 (PTR1).
PTR1 is responsible for the salvage of pterins in parasitic trypanosomatids and has overlapping
1
0ꢀ12
activity with DHFR, providing a metabolic bypass to alleviate DHFR inhibition.
Under
physiological conditions, PTR1 is responsible for only 10% of the reduction of folic acid required by
3
ACS Paragon Plus Environment

Page 5 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the cell, but when classic antiꢀfolate drugs inhibit DHFR, the gene encoding the NADPHꢀdependent
13
PTR1 is up regulated and PTR1 provides the reduced folates necessary for parasite survival.
Therefore, PTR1 is considered a promising target for the development of improved therapies. This
protein is considered a validated target in T. brucei, even though a correlation between inhibition of T.
brucei PTR1 (TbPTR1) and inhibition of parasite growth has not been demonstrated yet. The
combination with MTX, a DHFR inhibitor, has been shown to improve the efficacy and the potency
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
4,15
of PTR1 inhibitors.
1
6ꢀ19
In the drug discovery process, natural compounds represent a potential source of newleads.
Different natural compounds from plants including flavonoids have shown antiꢀtrypanosomatidic
2
0,21
activity.
Although flavonoids have been found to inhibit polyamine biosynthesis, the mechanism
21ꢀ24
of action has not been fully elucidated.
Flavonoids exert pleiotropic effects and different targets
2
5ꢀ27
have been suggested to explain at least part of their antiꢀparasitic activity.
Flavonoids have been
28
proposed as PTR1 inhibitors only in-silico. Although some flavonoids are very effective in-vitro,
22
they lack notable in-vivo activity due to their low bioavailability. However, an efficient drug
delivery strategy together with a clear definition of their molecular targets could lead to the
2
2
exploitation of the antiꢀparasitic potential of these natural products.
The purpose of the present study was to identify new PTR1 inhibitors showing antiꢀtrypanosomal (T.
brucei, T. cruzi) and antiꢀleishmanial (Leishmania infantum) activity by screening a library of natural
products and then by using a medicinal chemistry approach to design, synthesize and biologically
evaluate a new library. Assessment of early ADME and toxicity properties provided selection criteria
for further studies. MTX has previously been combined with PTR1 inhibitors to achieve
14
synergistic/additive effects on T. brucei. In our study, we evaluated the antiparasitic activity of the
compounds both alone and in combination with MTX. Pharmacokinetic studies using BALB/c mice
showed that one of our compounds (compound 2) was characterized by a short halfꢀlife and therefore
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
it was loaded in Poly(lacticꢀcoꢀglycolicacid) (PLGA) nanoparticles and solubilized with cyclodextrins
aiming to increase the its stability and bioavailability.
Results and discussion
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Onꢀtarget screening of a natural product library and hit identification
To identify potential new drugs for treating parasitic diseases, we investigated a library of natural
products. Starting from a library of 98 compounds, a computational docking analysis against TbPTR1
together with consideration of structural diversity allowed the restriction of the library to 38
compounds that were tested for PTR1 inhibition. These 38 phytochemicals (NPꢀ1 – NPꢀ38) consisting
of 12 phenolic acids, 5 flavanones, 9 flavones, 6 flavonols, 2 catechins, 2 triterpenes and 2
anthraquinones, were investigated for in-vitro inhibition of TbPTR1 and for growth inhibition of T.
brucei. Due to their ability to bind similar folateꢀrelated metabolites, PTR1 inhibitors can also interact
with other folateꢀdependent enzymes such as TS and DHFR both from parasite and human cells. Both
human enzymes can be considered potential offꢀtargets for PTR1 inhibitors, therefore all the
compounds were assessed for their selectivity towards the human TS and human DHFR. The data
obtained are depicted in Figure 1. All chemical structures and IC50 values are reported in Table S1 of
the Supporting Information.
With the exception of quercetin dehydrate (NPꢀ30), all flavonꢀ3ꢀol type aglycones showed a
significant TbPTR1 inhibition, with 3ꢀhydroxyflavone (NPꢀ27, IC50 = 12.8 ꢁM) and isorhamnetin
(NPꢀ31, IC50 = 13.9 ꢁM) being the most potent; whereas the glycosylated flavonoid (NPꢀ32) did not
inhibit TbPTR1. Most phenolic acids showed no inhibitory activity towards TbPTR1, highlighting the
importance of a condensed ring (chromenꢀ4ꢀone) for protein inhibition. The presence of two rings,
one aromatic and the other one aliphatic (NPꢀ8) or both aromatic (NPꢀ10), slightly increased the
inhibitory activity with respect to the other phenolic acid derivatives.
5
ACS Paragon Plus Environment

Page 7 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The simplest flavonol (NPꢀ27, IC50 = 12.8 ꢁM) was 10 times more potent than the corresponding
flavone (NPꢀ18, IC50 = 120.8 ꢁM). This implies the importance of OH at position R3. Moreover,
isorhamnetin (NPꢀ31, IC50 = 13.9 ꢁM) was over 35 times more active against TbPTR1 than the
corresponding flavone (chrysoeriol, NPꢀ25, IC50>490ꢁM). The insertion of a hydroxyl group at
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
position R5 led to a decrease in the inhibitory activity. Indeed, the simplest flavone (NPꢀ18, IC50
=
120.8 ꢁM) is more than 20ꢀfold more active than primuletin (NPꢀ19, IC >2450 ꢁM), which has a
5
0
hydroxyl group at position 5. On target screening of the NP series indicates that flavonols are the most
promising compounds, with NPꢀ31 being the most active and selective hit (Selectivity Index – SI
hTS/TbPTR1 >35; SI hDHFR/TbPTR1 = 36). Nevertheless, a clear structureꢀactivity relationship
(SAR) could not be established from this dataset and we could not reliably determine how the number
and the pattern of hydroxyl/methoxy substituents influenced the activity.
The 38 compounds were also assessed for their inhibitory activity against T. brucei parasite (Figure 1,
Table S1). With the exception of NPꢀ9, NPꢀ10, NPꢀ22, and NPꢀ35, all the natural flavonoids screened
showed antiꢀtrypanosomal activity with IC lower than 30 μM. Figure 1 shows that there is no direct
5
0
correlation between the antiꢀtrypanosomal activity of the NP compounds and their PTR1 inhibition
effect.
Xꢀray crystallographic studies of the natural products
To understand the flavonoid interactions into the protein binding site, the 18 compounds showing IC50
lower than 150 µM (compounds NPꢀ8, NPꢀ10, NPꢀ12, NPꢀ13, NPꢀ18, NPꢀ21, NPꢀ22, NPꢀ23, NPꢀ24,
NPꢀ26, NPꢀ27, NPꢀ28, NPꢀ29, NPꢀ31, NPꢀ35 and NPꢀ36; Figure 1) were selected for structural
+
analysis. We obtained crystal structures of the TbPTR1ꢀNADPH/NADP complexes for two
moderately active natural products (NPꢀ29 (datiscetin) IC50 TbPTR1= 76.9 µM and NPꢀ13
(hesperetin) IC50 = 104.2 µM) only. These compounds show inhibitory activity against TbPTR1 and
selectivity against at least one of the human proteins (TS and DHFR; Figure 1). Statistics for data
collection and refinement are reported in Tables S2 and S3 of the Supporting Information. In both
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
cases, the crystal asymmetric unit contains the functional unit of the enzyme, the TbPTR1 tetramer.
The tertiary structure of the TbPTR1 subunits is typical of the shortꢀchain dehydrogenases/reductases
(SDR) superfamily and shows a single α/β domain consisting of a sevenꢀstranded parallel βꢀsheet
29
sandwiched between two sets of αꢀhelices. The TbPTR1 active site is mainly formed by a single
chain, with one end blocked by the Cꢀterminus of the partner subunit. In this Lꢀshaped depression, the
cofactor binds in an extended conformation entrapped by a network of highly conserved hydrogen
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
9
bonds. The substrateꢀbinding loop (residues 207ꢀ215) interacts with both the cofactor and the
substrate. Two surfaceꢀexposed parts of the sequence, residues 104 to 112 and 143 to 151, are usually
poorly visible in TbPTR1 crystal structures, and they were not included in our models. The overall
structure of each subunit of the TbPTR1 complexes is the same as indicated by the small root mean
square deviations (RMSDs) after Cα superimposition of the four subunits in each tetramer (0.16ꢀ
0
.26Å for NPꢀ13 and 0.11 – 0.41 Å for NPꢀ29). The binding of the cofactor is essential to create both
the catalytic site and the substrateꢀbinding pocket, indicating an ordered sequential reaction
30
mechanism with NADPH binding before the substrate. The pterin moiety of substrates and of pterinꢀ
+
like inhibitors binds in a πꢀsandwich between the nicotinamide ring of NADPH/NADP and the
31
+
aromatic side chain of Phe97. Pterin also interacts with the NADPH/NADP βꢀphosphate, and this
31
interaction is strongly conserved in pterinꢀlike inhibitors.
+
TbPTR1ꢀNADPH/NADP ꢀNPꢀ29. The chromenꢀ4ꢀone core of NPꢀ29 binds in the biopterin binding
pocket as described above. The ligand is supported by a network of Hꢀbonds involving the cofactor
and the surrounding residues in the active site cavity (see Figure 2A). The hydroxyl group at position
7
on the chromenꢀ4ꢀone of NPꢀ29 (the NPꢀ29 atom numbering is given in the inset in Figure 2A)
+
donates a Hꢀbond to the NADPH/NADP βꢀphosphate and accepts a Hꢀbond from an amino group on
+
Arg14. The hydroxyl group at position 5 accepts a Hꢀbond from NADPH/NADP ribose OH2’ and
donates a Hꢀbond to the hydroxyl group of Tyr174 which in turn donates a Hꢀbond to the inhibitor
carbonyl oxygen at position 4. Furthermore, the carbonyl oxygen and the hydroxyl group at position 3
of NPꢀ29 are both Hꢀbonded to a water molecule linked by a further Hꢀbond to the side chain of
7
ACS Paragon Plus Environment

Page 9 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Asp161. An additional water molecule, which is highly conserved throughout all TbPTR1 structures,
mediates the interaction between the position 3 hydroxyl group of NPꢀ29 and the backbone carbonyl
oxygen of Gly205 (not shown). The oꢀphenol moiety of NPꢀ29 is located in a hydrophobic pocket of
the TbPTR1 active site cavity, surrounded by the side chains of Val206, Leu209, Pro210, Met213 and
Trp221. The 2’ hydroxyl group on this ring is close (~ 4 Å) to the carbonyl oxygen of Gly205 and
forms an intramolecular Hꢀbond (2.4 Å) with the hydroxyl group at the 3 position on the chromenꢀ4ꢀ
one moiety.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
+
TbPTR1ꢀNADPH/NADP ꢀNPꢀ13. The chromenꢀ4ꢀone core of NPꢀ13 also binds in the biopterin
binding pocket, but it is rotated by about 180° with respect to its orientation in the TbPTR1ꢀNPꢀ29
complex (Figure 2B). The ether oxygen at position 1 on the chromenꢀ4ꢀone moiety of NPꢀ13 points
towards the side chains of Asp161 and Tyr174. The hydroxyl group at position 7 of the NPꢀ13
+
chromenꢀ4ꢀone is within Hꢀbonding distance of the NADPH/NADP ribose OH2’ and the side chain
of Ser95. The NPꢀ13 hydroxyl group at position 5 accepts a Hꢀbond from an amino group of Arg14
and donates a Hꢀbond to the backbone carbonyl oxygen of Leu208. The oꢀmethoxyꢀphenol (oꢀ
guaiacol) ring in position 2 on the chromenꢀ4ꢀonemakes a Tꢀshaped stacking interaction with the
aromatic side chain of Trp221. Furthermore, the 3’ phenolic moiety is within Hꢀbonding distance of
the Asp161 side chain and forms two waterꢀmediated interactions with the backbone nitrogen of
Cys168 and the amide nitrogen of Asn175 (not shown). The methoxy oxygen at the 4’ position
interacts with the sulfur of the modified Cys168 (Sꢀoxyꢀcysteine), whereas the terminal methyl makes
van der Waals contacts with the side chains of Trp221 and His267 of the partner subunit.
In summary, structural comparison of the two TbPTR1 complexes reveals two opposite orientations of
the natural flavonoid inhibitors NPꢀ13 and NPꢀ29 in the biopterin binding pocket in which the
chromenꢀ4ꢀone substituents interact with different residues surrounding the cavity. A double binding
15
mode was previously observed for pteridineꢀlike inhibitors and MTX.
Flavonoid library design and synthesis
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The crystal structures showed that flavonoids occupy the biopterin binding site of TbPTR1,
suggesting that the flavonol moiety could provide a new scaffold for the development of PTR1
inhibitors. The flavonoids selected as starting points for the development of compounds to treat
trypanosomatidic diseases are depicted in Table 1. Even though the two crystal structures showed that
the hydroxyl group at position 5 could establish Hꢀbonds with the protein and the cofactor, the
inhibitory data on natural flavonoids (Figure 1) suggested that a hydroxyl group in this position leads
to a decrease in the inhibitory activity against TbPTR1. Therefore, we decided to synthesize
compounds with hydroxyl substituents at positions 6 or 7 of the chromenone Aꢀring since, according
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
to the crystal structures, these should be able to establish Hꢀbonds with the NADPH/NADP ribose
+
and the side chain of Ser95 or with the NADPH/NADP βꢀphosphate and Arg14 (Figure 3). Moreover,
we performed modifications at positions 3’, 4’ and 5’ on ring B to explore the SAR. As shown in
Figure 3, hydroxyl groups in positions 3’, 4’and 5’ could form a Hꢀbonding interaction with the
Asp161 side chain and two waterꢀmediated interactions with the backbone nitrogen of Cys168 and the
amide nitrogen of Asn175. Many natural products have both hydroxyl and methoxy groups, thus we
synthesized and biologically evaluated eight hydroxylated (1ꢀ8) and eight methoxylated (9ꢀ16)
compounds aiming to investigate how the number and the pattern of hydroxyl/methoxy substituents
influenced the activity.
The synthesis of compounds 1ꢀ16 is depicted in Scheme 1. All the compounds have been already
reported in literature, but, as far as we know, they have never been proposed as PTR1 inhibitors. The
intermediate chalcones (17ꢀ24) were synthesized by ClaisenꢀSchmidt condensation using substituted
acetophenones and benzaldehydes in the presence of NaOH as the base in ethanol. The reaction was
26
carried out as previously reported in literature for similar compounds. Afterwards, the chalcones
(17ꢀ24) were converted into the corresponding methoxylated flavonols (9ꢀ16), using the FlynnꢀAlgarꢀ
Oyamada method for epoxidation and subsequent intramolecular cyclization of the openꢀchain
2
5
structure. The reaction was performed with hydrogen peroxide in aqueous base (1 M NaOH).
Cleavage of methoxy protecting groups with boron tribromide gave the hydroxylated flavonols 1ꢀ8 in
9
ACS Paragon Plus Environment

Page 11 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
13
high yield. All synthesized compounds were characterized by HꢀNMR, CꢀNMR and mass analysis.
The obtained NMR and mass data are reported in the Supporting Information (pag. 33ꢀ37) and were
compared with those available in literature.
Target compound profile
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In our study a panel of assays was performed and activity/toxicity data were considered during the hit
32
selection/prioritization. Together with the potency data (measured as IC50 i.e. the compound
concentration producing 50% reduction of targeted enzyme activity and parasite cell growth), we
considered the selectivity index (IC50 towards the parasite compared with compound cytotoxicity IC50
to host cells) and earlyꢀtoxicity properties. We have assessed the entire profile of the sixteen
synthesized compounds and the whole data panel has been evaluated before selecting compounds for
pharmacokinetic studies.
Testing of synthetic flavonols against PTR1 enzymes
All flavonols were investigated for their activity towards T. brucei (TbPTR1) and L. major PTR1
(LmPTR1). The results are shown in Figure 4 and Table S4 of the Supporting Information. All
hydroxylated compounds except compound 5 showed significant inhibitory activity against LmPTR1
and TbPTR1 at 50 µM concentration, with compound 2 being the most potent. Methylation resulted in
almost complete loss of inhibitory activity. The most noteworthy examples are represented by two
pairs: compounds 10 and 2 (4 % and 96 % inhibition of TbPTR1 at 50 µM, respectively) and
compounds 12 and 4 (no inhibition of TbPTR1 at 50 µM and 85 % inhibition at 50 µM, respectively).
The introduction of a hydroxyl group on ring A in compound 2 led to a significant increase of almost
3
ꢀfold in the inhibitory potency against both enzymes with respect to compound 1, which bears an
unsubstituted ring (34 % inhibition of TbPTR1 at 50 µM). In order to explain a complete SAR, Xꢀray
crystallographic studies and docking analyses were carried out. The compounds were evaluated
towards also towards TcPTR2 and none of the compound significantly inhibited TcPTR2 activity
(data not shown).
Crystal structures of synthetic flavonoidꢀTbPTR1 complexes
1
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Compounds 2 and 7, which show inhibitory activity against TbPTR1, were submitted to Xꢀray
crystallographic characterization aimed at comparing their binding poses in the PTR1 active site with
those of the natural flavonoids and thus gaining further information for compound optimization.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
+
+
TbPTR1ꢀNADPH/NADP ꢀ2. The structure of TbPTR1 in complex with NADPH/NADP and
compound 2 was determined at 1.38Å resolution. Ligand placement and key interactions within the
active site cavity are reported in Figure 2C. Compound 2, like NPꢀ13 and NPꢀ29, binds in the
substrate binding pocket, with the chromenꢀ4ꢀone in a πꢀsandwich interaction between the
+
nicotinamide ring of NADPH/NADP and the aromatic side chain of Phe97. The chromenꢀ4ꢀone core
of compound 2 adopts the binding mode observed for NPꢀ13, with the oxygen atom at position 1
directed towards the side chain of Asp161. The R6 hydroxyl group on the chromenꢀ4ꢀone is within Hꢀ
+
bonding distance of the hydroxyl group of Ser95 and the NADPH/NADP βꢀphosphate oxygen. The
carbonyl oxygen at position 4 completes the interactions made by the chromenꢀ4ꢀone by receiving
two Hꢀbonds from an amino group of Arg14 and a water molecule that is in turn Hꢀbonded to the
+
NADPH/NADP βꢀphosphate O2. These interactions, together with the stacking of the chromenꢀ4ꢀone
described above, bring its 3 position hydroxyl group within Hꢀbonding distance of the backbone
carbonyl oxygen of Leu208.The phenyl ring in position 2 on the chromenꢀ4ꢀone establishes a Tꢀ
shaped stacking interaction with the aromatic side chain of Trp221 and forms van der Waals
interactions with the side chains of Val206 and Leu209 (not shown). The metaꢀhydroxyl group at the
3
’ position on the phenyl moiety of 2 is only involved in a water mediated Hꢀbond network linking it
to the backbone carbonyl of Gly205 and the side chain of Asp161 (not shown).
+
+
TbPTR1ꢀNADPH/NADP ꢀ7. The structure of TbPTR1 in complex with NADPH/NADP and
compound 7 has been determined at 1.76 Å resolution. Ligand placement and key interactions within
the active site cavity are shown in Figure 2D, where it can be observed that compound 7 adopts
exactly the same pose as compound 2, making the same interactions with the enzyme and cofactor.
The only difference consists in the steric repulsion of the additional hydroxyl group at position 4’ that
1
1
ACS Paragon Plus Environment

Page 13 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
causes the side chain of Trp221 to move away by about 1.4 Å to make room for this group. The 4’
hydroxyl group is not involved in Hꢀbonds but is solvent exposed and presumably can form aHꢀbond
with water.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The superposition of the four crystal structures is shown in Figure 5. The orientation of compounds 2
and 7 in the TbPTR1 cavity is the same as for NPꢀ13, with the inhibitor bicyclic core directed towards
the opposite side of the cavity with respect to NPꢀ29. Due to the larger steric hindrance of the
substituents on the 2ꢀoꢀmethoxyꢀphenolring of NPꢀ13, a little rearrangement of the whole molecule
occurs with a rotation of about 35° of the bicyclic core with respect to 2 and 7 that allows the side
chain of Trp221 to remain in the position observed in the TbPTR1ꢀ2 complex (Figure 2C). The
rotation of NPꢀ13 allows direct Hꢀbonding between the 3’ hydroxyl group of NPꢀ13 and Asp161,
while the interaction of the corresponding group of 2 and 7 is mediated by a water molecule. The
displacement of the Trp221 side chain occurring upon binding of compound 7 can be appreciated
(Trp221 as yellow sticks in Figure 5).
A puzzling aspect of all four crystal structures is that the different inhibitors have been systematically
+
observed bound to three of the four TbPTR1 subunits, whereas the NADPH/NADP cofactor is found
bound to all subunits in three instances out of four, although with lower occupancy in those subunits
where the inhibitor is absent. The absence of the inhibitor is correlated with larger disorder in the
substrate binding loop that brings Trp221 into the active site cavity. However, this loop always shows
temperature factors higher than average, even in the subunits where both cofactor and inhibitor are
bound. The disorder of this region is independent of the crystal packing since, when the loop is
disordered (e.g. in subunit C in TbPTR1ꢀNPꢀ29), it approaches the symmetryꢀrelated subunits in the
same way as the ordered loop (e.g. in subunit A in TbPTR1ꢀNPꢀ29) located on the opposite face of
the TbPTR1 tetramer. Confirming this observation, the other two ordered loops (e.g. in subunits B and
D in TbPTR1ꢀNPꢀ29) do not make close intermolecular contacts. In summary, we conclude that
stabilization due to the πꢀstacking between the 2ꢀaromatic ring of the inhibitor and the side chain of
Trp221 is critical for the ordering of the substrate binding loop.
1
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Computational docking of synthetic compounds to TbPTR1 and LmPTR1 and SAR analysis
Docking studies were conducted to further expand the SAR to all the hydroxylation/methoxylation
patterns in the synthesized compounds. In addition, computational docking allowed the evaluation of
possible binding modes and a corresponding SAR for LmPTR1, for which no crystal structure with
these compounds could be solved.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Like the crystallographic studies, computational docking showed multiple possible binding modes for
the chromenꢀ4ꢀone system, which roughly group into three classes: (1) a typical substrateꢀlike binding
+
mode involving Hꢀbonding with the NADPH/NADP cofactor and Ser95 of TbPTR1 (or Ser111 of
LmPTR1), as observed in the complexes of compounds 2, 7 and NPꢀ13; (2) an alternative binding
mode with chromenꢀ4ꢀone flipped by roughly 180° as observed in the complex with NPꢀ29; and (3)
an inverse binding mode solely observed in docking studies, with ring B rather than the chromenꢀ4ꢀ
+
one in a stacking orientation between the NADPH/NADP nicotinamide and Phe97 of TbPTR1 (or
Phe113 of LmPTR1). While three of the four crystal structures feature binding mode (I), a general
dependence of the binding mode on the substituent pattern is notable both from crystallographic and
docking studies. In docking, many of the observed binding modes were found to have highly similar
contacts and Hꢀbonding partners, thus creating a favorable entropic contribution to the binding of this
compound class (see also Tables S5 and S6 in Supporting Information). However, due to their large
number of Hꢀbond donors/acceptors, flavonoids can adopt additional binding modes with different,
often water mediated, interactions. We therefore used a ligand constraint docking approach to favor
arrangements close to the crystal complexes to assess the potential of other synthetic derivatives to
bind in similar orientations.
The crystal structures and docking indicate that hydroxylation patterns on ring A influence the
orientation of chromenꢀ4ꢀone in the pocket. Precisely, compounds hydroxylated at position R6 show
+
different Hꢀbonding partners (mostly NADPH/NADP phosphate and Ser95/Ser111 of
+
TbPTR1/LmPTR1, respectively) from those hydroxylated at R7 (mostly NADPH/NADP ribose and
1
3
ACS Paragon Plus Environment

Page 15 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Tyr174/Tyr194 or TbPTR1/LmPTR1, respectively), see Figure 5. This difference also affected the
compound activity (compare R6ꢀhydroxylated compound 2 (96% TbPTR1 inhibition at 50 µM) and
the corresponding R7ꢀhydroxylated compound 3 (47% TbPTR1 inhibition at 50 µM)). While the
geometry of compound 2 fits ideally in the TbPTR1 binding pocket, the slight reorientation observed
for compound 3 leads to a loss of Hꢀbonds and stabilizing hydrophobic contacts, halving the
compound activity. Thus, hydroxylation of ring A can fineꢀtune the arrangement of the chromenꢀ4ꢀ
one system within the pocket and critically influence the interaction pattern of rings B and C.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Methoxylations, in contrast to hydroxylations, were found to lead to activity loss. For example, the
inhibitory activity against TbPTR1 and LmPTR1 drops substantially, when comparing compound 2
with the corresponding methoxylated compound 10 (96% and 4% TbPTR1 inhibition at 50 µM; 86%
and no LmPTR1 inhibition, respectively). Methoxylations on the Aꢀring cause a displacement of the
chromenꢀ4ꢀone from the primary stacking geometry to accommodate the bulkier methoxyꢀgroup in
+
close proximity tothe NADPH/NADP cofactor and Ser95 of TbPTR1 (Ser111 ofLmPTR1), see
Figure 6A. Notably, this adverse effect was also apparent in the comparison of compounds 1 and 9,
bearing no substitution on ring A but a hydroxylation or methoxylation on R3' of the Bꢀring. Whereas
the activity of compound 9 is negligible, compound 1 shows about 35% inhibition against both
TbPTR1 and LmPTR1 (at 50 µM). AHꢀbond donor functionality on R3' has an important stabilizing
effect, as it can interact with a water molecule bridging Asp161 and Gly205 of TbPTR1 (see Figure
6
B) orAsp181 and Gly225 of LmPTR1. As, according to our WatCH analysis, this water is almost
100% conserved in both TbPTR1 and LmPTR1, it provides a clear advantage for R3' hydroxylated
compounds. This is further supported by the crystal structures of TbPTR1 with compounds 2 and 7,
where the R3' hydroxylation is in Hꢀbonding distance to this water site. In agreement with this
observation, compound 2 with the mꢀhydroxyphenyl was over 30 times more active than compound 5
with a pꢀhydroxyphenyl ring (96% and 3% TbPTR1 inhibition at 50 µM, respectively). In contrast to a
metaꢀhydroxylation, a pꢀhydroxylation cannot interact with the structural water, see Figure 5. Again,
this effect was sensitive to the hydroxylation pattern of ring A, as can be seen by comparing
1
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compounds 3 (R3' and R7 hydroxylation, 47% TbPTR1 inhibition at 50 µM) and compound 6 (R4'
and R7 hydroxylation, 33% TbPTR1 inhibition at 50 µM), see Figure 5. While pꢀsubstituted
compounds generally show a drop in inhibitory activity compared to their respective mꢀsubstituted
counterparts, the effect is weaker for the second pair with the R7 hydroxylation on ring A, possibly
because R7 substitutions lead to a slight rotation of the stacking chromenꢀ4ꢀone, directing the pꢀ
hydroxyl towards a more solventꢀexposed region of the pocket. In conclusion, the effects of individual
hydroxyl substituents are not purely local and do not show a completely additive effect; rather, the
combinations of ring A and B hydroxylations will determine the final binding mode.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Based on the binding modes observed for TbPTR1, we evaluated possible binding orientations to
LmPTR1 by docking. Notably, all compounds with R4' hydroxylation (compounds 5 and6),
compounds with R3' and R4' hydroxylations (compounds 7and 8) and one compound with R3'
hydroxylation combined with an R7 substitution on ring A (compound 3) were found to be slightly
more active towards LmPTR1 than TbPTR1. While the depth of the biopterin binding pocket is well
conserved and the interactions for chromenꢀ4ꢀone are practically identical in both proteins, there are
notable differences towards the opening of the pocket that can explain the observed activity
differences (see sequence alignment of interaction partners in Figure S2 of the Supporting
Information). While in TbPTR1, R3' hydroxylations on the Bꢀring mainly come into contact with the
conserved water molecule and Cys168 (see Figure 5) and R4' hydroxylated compounds can hardly
form direct Hꢀbonds with the receptor, LmPTR1 generally has a more polar opening part of the
pocket. Three major differences are crucially important: His241, Tyr283 and Arg287 from the
neighboring subunit in LmPTR1, correspond to Trp221, Leu263 and His267 from the neighboring
subunit in TbPTR1, thereby providing several additional polar contact points. As the side chain of
Arg287 in LmPTR1 is much bigger than that of His267 in TbPTR1, arginine is much more accessible
to the ligands than histidine. Furthermore, as observed in crystal structures of TbPTR1, Trp221 is
rather motile and different rotamers lead to an open or a rather closed form of the pocket, thereby
potentially blocking ligand access to His267. While Trp221 of TbPTR1 did not provide aHꢀbonding
1
5
ACS Paragon Plus Environment

Page 17 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
contact to any of the studied compounds, induced fit docking studies of the compounds to LmPTR1
showed movement of His241 to allow additional Hꢀbonding interactions in the most stable structure,
as also shown in Figure 7 (see also Table S5). Overall, πꢀπ stacking and hydrophobic contacts are
clearly the main stabilizing factors for flavonoids in TbPTR1, whereas a tendency towards a higher
number of Hꢀbond donor/acceptor contacts may stabilize hydroxylations in both the metaꢀ and paraꢀ
positions of ring B slightly better in LmPTR1.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
While addition of a paraꢀhydroxylation for the R7 substituted ring A did not lead to a significant
change in activity against TbPTR1 (compound 3: 47% and compound 8: 53%TbPTR1 inhibition at 50
µM), compound 4 with hydroxylations at R3', R4' and R5' yielded a clear activity increase (85%
TbPTR1 inhibition at 50 µM). Considering the lack of direct polar contact points in TbPTR1, this is
surprising and there may be two possible explanations: (1) bridging water molecules may enable
additional contacts for the second metaꢀhydroxylation, e. g. with Trp221, or (2) the compound may
adopt an inverted binding mode, placing the pyrogallol moiety in a stacking orientation with Phe97
and the cofactor, where it could make strong Hꢀbonds and thereby keep the more hydrophobic
chromenꢀ4ꢀone in the hydrophobic TbPTR1 sub pocket, where it is able to form additional
hydrophobic contacts.
Testing of synthetic flavonols against TbDHFR, LmDHFR, hDHFR and hTS
The compounds showing inhibitory activity against PTR1 (1ꢀ8) were evaluated for their activity
against TbDHFR, LmDHFR and hDHFR, the results are given in Table S7 of the Supporting
Information. With the exception of compounds 1 and 2, all the tested compounds had inhibitory
activities against TbDHFR and LmDHFR below 25% at 50 µM. Compounds 1 and 2 slightly inhibit,
respectively, TbDHFR and LmDHFR (42ꢀ43% at 50 µM). Therefore, we conclude that compounds 1ꢀ
8
have low inhibitory activity against the parasitic DHFRs. With the exception of compound 2, all the
compounds had inhibitory activities against hDHFR below 30 % at 50 µM. The IC50 of compound 2
1
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
against hDHFR is 50 µM and against TbPTR1 is 4.3 µM, thus compound 2 has a selectivity index
higher than 10. None of the tested compounds inhibited hTS at 50 µM (data not shown).
Early toxicity studies
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
To characterize the compounds properties and to define their toxicological profile, we used an assay
3
2
panel typical of those employed in drug discovery projects for hit/lead selection. Our panel includes
hERG, five cytochrome P450s, (CYP1A2, CYP2C9, CYP2C19, CYP2D6 and CYP3A4), Aurora B
kinase, A549 (human lung adenocarcinoma epithelial) and WIꢀ38 (foetal lung fibroblasts) cell lines
and mitochondrial toxicity. The results of the screening are reported in Figure 8 as a traffic light code
(from red through yellow to green for increasing biological effect and decreasing toxicity of the
compounds) and in Table S8 of the Supporting Information. An ideal lead compound should have all
parameters colored in green. In details, the percentage of mitochondrial toxicity and of inhibition
towards hERG, CYP isoforms and Aurora B kinase should be below 30 %, while the percentage of
A549/WIꢀ38 cells growth, measuring cytotoxicity, should be above 70%. A few compounds showed a
percentage of inhibition towards hERG >10 % at 10 ꢁM. Apart from compounds 4 and 8, all the
compounds inhibited at least one isoform of cytochrome P450 (mostly CYP1A2), while only two
compounds (4 and 8) exhibited a significant inhibitory activity against Aurora B kinase at 10 ꢁM.
None of the compounds was either cytotoxic or show mitochondrial toxicity.
Testing of synthetic compounds on cultured parasites
With the exception of compound 8, all the molecules were tested at 10 µM against T. brucei
bloodstream form, intracellular T. cruzi and L. infantum intracellular amastigotes (Figure 9). Almost
all tested compounds were active against T. brucei, while only two compounds, 4 and 12, were
slightly active against L. infantum showing inhibitory activities of 10 and 35%, respectively. Twelve
compounds (1, 2, 5ꢀ12, 14ꢀ16) presented EC50 values against T. brucei between 1ꢀ8 µM (Table 2),
while compounds 3, 4 and 13 showed EC50 values of 12.29, 18.04 and 20.12 µM, respectively. Only
the methoxylated compounds 10, 11, 14 and 16 showed antiparasitic activity towards T. cruzi
1
7
ACS Paragon Plus Environment

Page 19 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trypomastigotes similar to that of nifurtimox (NFX), the reference drug for the treatment of Chagas’
disease. Compound 10 is the most active (EC = 4.5 µM, NFX EC = 1.6 µM).
50
50
The series was assessed for cytoxicity on THP1 macrophageꢀlike cells to evaluate the NOAEL (no
observed adverse effect level). The reference compound was pentamidine with a NOAEL of 10 µM
and a selectivity index of 6440. The selectivity indexes of our compounds, given by the ratio between
the EC50 towards T. brucei and CC50 towards THP1, were lower than 10 with the exception of
compound 12 (SI = 17) (Table 2). The selectivity index acceptable for a compound to be considered
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
3
non toxic is at least 10, thus our compounds showed slight toxicity.
Combination studies and evaluation of synergy
All compounds were evaluated in combination with MTX independently of their PTR1 inhibitory
activity against the recombinant protein. We aimed to assess the potential gain in potency and the
reduction in toxicity through the combination of our compounds with MTX, a wellꢀknown DHFR
inhibitor. In trypanosomatids MTX is expected to exert a synergistic behavior in combination with
PTR1 inhibitors. Moreover, MTX can show a synergistic effect in combination with nonꢀPTR1
inhibitors through a multitarget inhibition. Preliminary combination experiments in T. brucei were
conducted by fixing MTX at 4 µM (MTX average EC30 against T. brucei) and varying the
concentration of the selected compounds between 1.25 and 20 µM (Table S10 of Supporting
Information). The synergy coefficients of all the compounds tested at the different concentrations
(1.25, 2.5, 5, 10, 20 µM) are reported in Table S11 of the Supporting Information. Eight compounds
presented a synergistic effect (synergy coefficient > 1) when combined with MTX. The antiꢀparasitic
activity against T. brucei of the eight compounds both alone and in combination with MTX is shown
in Figure 11. Compounds 5, 6 and 13 consistently presented the highest synergy coefficients
(maximum synergy coefficient > 1.5). Compound 13 was the most synergic with a calculated
maximum synergy coefficient of 3.1 at 1.25 ꢀM (Table S11). Since it showed the highest synergy
1
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
coefficient together with a safe toxicological profile (Figure 8, the only poor (red) parameter is for
CYP1A2 (72% inhibition at 10 µM)), it was selected for further synergistic combination studies.
To better evaluate the gain in potency of the combination MTXꢀ13, we employed the software
Compusyn and a constant ratio between the concentrations of the two compounds chosen from their
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
34
EC50 values (EC50 13 = 20 µM; EC50 MTX = 15; constant ratio = 1.3) This enabled us to quantify
synergism between the two compounds (combination indexꢀ CI) and to determine the dose reduction
(dose reduction index ꢀ DRI) needed to observe the studied effect in the combination compared with
the dose of each drug alone. The combination of MTX and compound 13 at the EC50 concentration
had a CI of 0.174 ± 0.047. When CI is lower than 1, the combination can be considered synergistic,
therefore the observed CI is showing strong synergism and confirms the initial data (Table S10). At
higher EC values, CI becomes lower indicating very strong synergism (Table 3). The EC determined
50
for the combination showing CI of 0.147 was 3.15±0.33 µM. This enabled an average DRI at the EC50
of 14.6±3.7 for MTX and of 11.7±2.7 for compound 13 (IC50 MTX = 1.3 µM and IC50 13 = 1.7 µM)
(Figure S3C Supporting Information). At the EC90 of the mixture the predicted DRI is over 100ꢀfold
for both compounds when compared with the EC50 of the compounds alone. The isobologram is
reported in Figure S3 of the Supporting Information. The toxicity of the MTXꢀ13 mixture on THP1
was determined at three different concentrations (12.5, 25, 50 ꢁM) and no synergistic toxicity on
human cells was observed (Figure S4 and Table S12). Therefore, the identified combination MTXꢀ13
is selective for the inhibition of the parasite growth, while there is no synergy in the toxicity towards
the host cells. These data together with the predicted DRI indicate that the combination improves the
selectivity indexes for both MTX and compound 13.
Pharmacokinetic studies, in-vivo assays and compound delivery
Compound 2 was selected for pharmacokinetic studies because it was the most active compound
against TbPTR1 (IC50= 4.3 ꢁM) that showed activity against the parasite T. brucei (EC50= 7.6 ꢁM)
and it presented a safe profile, see Figure 8. The hERG IC50 (99.3 ꢁM) was more than 20ꢀfold higher
1
9
ACS Paragon Plus Environment

Page 21 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
than the target IC . With the exception of the CYP1A2 IC (6.1 ꢁM), the IC50 against the
5
0
50
cytochrome isoforms was higher than 10 ꢁM. The solubility of compound 2 was evaluated using UVꢀ
Vis spectroscopy. In-vivo bioavailability and halfꢀlife of compound 2 were evaluated in BALB/c mice
treated IV with 1 mg/kg based on its low solubility. The molecule showed a short halfꢀlife (t1/2 = 7.6
min) and it was not detected in blood after 30 minutes (the plasma levels are shown in Figure S4 of
the Supporting Information). With the aim of improving the plasma levels of compound 2, we used
two different drug delivery systems: the molecule was encapsulated in PLGA nanoparticles and
solubilized with hydroxypropylꢀβꢀcyclodextrins. The encapsulation did not change the physical
properties of the PLGA nanoparticles as reported in Table S13 of the Supporting Information. The
unloaded and loaded NPs were characterized by Dynamic Light Scattering (DLS) in terms of size,
polydispersity index and zetaꢀpotential. We evaluated the in-vivo bioavailability of compound 2
encapsulated in nanoparticles in BALB/c mice (IV and per os administration) and formulated with
cyclodextrins after per os administration to NMRI mice. From preliminary data, both formulations did
not significantly increase the levels of compound 2 in plasma samples from mice (data not shown).
However, the in-vitro activity of compound 2 both in PLGA and in the cyclodextrin solution was
preserved when compared to the free compound, while the toxicity on THP1 was diminished (Table
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
).
Conclusion
This is the first study in which we have confirmed the potential of flavonols as PTR1 inhibitors. Apart
from compounds 4 and 8, none of the sixteen synthesized flavonols have previously been reported in
22
literature for their antiparasitic activity. Twelve of these compounds show EC50 values against T.
brucei below 10 µM. We performed combination studies with MTX and compound 13, a nonꢀPTR1
inhibitor, was the bestꢀperforming compound. Although the reason for the synergy is unknown,
compound 13 is an interesting compound for drug development due to the synergistic behavior with
MTX and the low toxicity on host cells. Target identification studies could be carried out aiming to
2
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
explain the mechanism of action. The synthesized compounds were also evaluated for the activity
against T. cruzi trypomastigotes and compound 10 turned out to have a potency comparable to that of
nifurtimox, the drug currently used to treat Chagas’ disease (EC50 compound 10 = 4.5 µM, NFX EC50
=
1.6 µM). We carried out early in-vitro toxicity assays and overall the library showed a satisfactory
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
profile. Combining the activity/toxicity data, we selected compound 2 for pharmacokinetic studies
and we observed a very quick turnover of the molecule in BALB/c mice. Neither encapsulation in
PLGA nanoparticles nor solubilization with cyclodextrins increased the plasma level of compound 2;
however, both formulations allowed to maintain the in-vitro activity of the compound and to reduce
the toxicity. Our crystal structures and SAR analysis provide a basis for structureꢀbased drug design
aimed at identifying novel flavonolꢀlike compounds with increased potency and selectivity towards
PTR1 and able to overcome the problems of classical flavonols.
Experimental section
Preparation of the natural products library
A library of 98 natural compounds, consisting of 8 flavanones, 19 flavones, 20 flavonols, 2
dihydroflavonols, 13 anthocyanins, 2 catechins, 2 isoflavones, 4 chalcones, 18 phenolic acids and
derivatives, 1 aurone, 5 anthraquinones, 3 triterpenes and 1 phloroglucinol was purchased from
Extrasynthese (Genay, France) and SigmaꢀAldrichꢀFluka (Milan, Italy) for drug screening. The
degree of purity of all these molecules was checked by HPLC. The HPLC analyses were performed
on an Agilent Technologies (Waldbronn, Germany) modular model 1100 system, consisting of a
vacuum degasser, a quaternary pump, an autosampler, a thermostated column compartment and a
diode array detector (UV/DAD). The chromatograms were recorded using an Agilent Chemstation for
LC and LCꢀMS systems (Rev. B.01.03).An Ascentis C column (250 × 4.6 mm I.D., 5 µm, Supelco,
18
2
Bellefonte, PA, USA) was used, with a mobile phase composed of (A) water (H O) and (B)
acetonitrile (ACN). The gradient elution was modified as follows: 0ꢀ3 min 25% B, 3ꢀ10 min from 25
to 30% B, from 10ꢀ40 min from 30 to 40% B, which was held for 5 min. The postꢀrunning time was 5
2
1
ACS Paragon Plus Environment

Page 23 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
min. The flowꢀrate was 1.0 mL/min. The column temperature was set at 30 °C. The sample injection
volume was 5 µL. The UV/DAD acquisitions were carried out in the range 190ꢀ600 nm and
chromatograms were acquired at 310, 330, 370 and 520 nm. All the compounds tested were found to
be stable and degradation products were not detected in the HPLC chromatograms. Table 1 of the
Supporting Information shows the purity (> 95%) of the 38 natural compounds screened. The natural
compounds were stored at low temperature (˗80 °C), protected from light and humidity.
Computational studies
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Virtual screening of the natural product library
The library of 98 natural products was subjected to virtual screening using the GOLD software
35,36
13
(version 5.1)
following the protocol validated in previous studies on PTR1. The compounds
were docked into the binding sites of the targets, LmPTR1 (PDB id: 1E92) and TbPTR1 (PDB id:
3JQ9), and the offꢀtargets, hDHFR (PDB id: 1U72) and hTS (PDB id: 1HVY). Ten docking poses
were generated for each ligand in each target and these were visually inspected and manually ranked,
considering the Gold fitness score, hydrogen bonding and πꢀπ aromatic interactions. A search of
BioAssay data in PubChem was done to filter out promiscuous compounds. The set of 38 compounds
given in Table 1 of the Supporting Information was selected for inhibition assays considering the
above properties and structural diversity.
Docking of synthetic flavonoid derivatives
The 3D structures of the compounds were created from SMILES strings and optimized with the
37
OPLS_2005force field usingMaestro. Ionization states and tautomers were generated at pH 7.0±0.5
37
using Epik. Up to 8 stereoisomers and one lowꢀenergy ring conformation were generated per
3
8
compound. Important structural water sites were identified using the WatCH clustering approach.
All available chains of the LmPTR1 crystal structures and 99 chains from TbPTR1 crystal structures
see Tables S14 and S15 for a list of structures) were superimposed and their water sites were
(
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
clustered with a distance criterion of 2.4 Å for considering a water oxygen position identical. Water
sites with at least 50% conservation were considered conserved.
The PTR1 structures for docking consisted of chain A of the relevant crystal structure and a Cꢀ
terminal tripeptide from the neighboring subunit pointing into the chain A active site (Val266 to
Ala268 inTbPTR1 and Thr286 to Ala288 in LmPTR1). The sideꢀchain oxygen atom on the modified
Cys 168 in the crystal structure of TbPTR1 was removed. Crystallographic solvent molecules were
removed and replaced by the conserved water molecules identified by WatCH clustering. The
LmPTR1 structure (PDB ID 1E92) was aligned to the TbPTR1 template (PDB id 5JCJ). PrepWizard
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
37
was used to assign bond orders and add hydrogen atoms. The Nꢀ and Cꢀtermini of chain A were
capped with Nꢀacetyl and Nꢀmethyl amide groups, respectively. The protonation state of the
+
NADPH/NADP cofactor was computed at pH 7.0±0.5. Protein protonation states were assigned by
4
3
PROPKA at pH 7.0. The Hꢀbond network was optimized and all hydrogens were subjected to a
restrained minimization. A 20 Å x 20 Å x 20 Å docking grid was centered on Phe97 ofTbPTR1
orPhe113 ofLmPTR1. The following hydroxyl groups were set rotatable: Ser95 and Tyr174 in
TbPTR1 and Ser111, Thr184, Tyr191, Tyr194, Thr195 and Tyr283 inLmPTR1, as well as those of the
+
NADPH/NADP ribose.
39ꢀ42
Docking was performed with and without ligandꢀbased constraints using the Glide software.
For
the ligand constraints, protein preparation was performed including compound 7 (substrateꢀlike
binding mode, PDB ID 5JCJ) or compound NP29 (inhibitorꢀlike binding mode, PDB id 5JCX) or the
highest scoring docking solution from unconstrained docking of compound 4 with ring B in a stacking
orientation (alternate binding mode). The ligands were then separated from the structure to provide
reference geometries. The van der Waals radii of the ligand atoms were scaled by 0.80 and a partial
charge cutꢀoff of 0.15 was used. XP (extra precision) docking and flexible ligand sampling were set.
Nitrogen inversions and ring conformations were sampled, biased sampling of torsions was performed
for amides only and set to penalize a nonꢀplanar conformation. Epik state penalties were added to the
2
3
ACS Paragon Plus Environment

Page 25 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
docking score, intramolecular Hꢀbonds were rewarded and the planarity of conjugated π groups was
enhanced. A core pattern comparison with a tolerance of 3Å was used for the ligandꢀbased constraint.
The core was defined as all nonꢀhydrogen atoms of the chromenꢀ4ꢀonecore ring system except the R3
hydroxyl group. 20 poses per ligand were subjected to postꢀdocking minimization with a threshold of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0
.5 kcal/mol for rejecting a minimized pose. Up to 10 final docking solutions were reported.
Unconstrained docking was performed analogously, omitting only the core pattern comparison. For
LmPTR1, the core constraint docking was based on the binding poses in TbPTR1. In addition,
inducedꢀfit docking was carried out using the same settings except that the size of the docking grid
was chosen automatically and a ligand and receptor van der Waals scaling of 0.5 was applied.
37
Refinement was carried out with Prime for residues within 5 Å of the ligand, followed by XP reꢀ
docking into structures within 30 kcal/mol of the best structure and within the top 20 structures
overall. Up to 20 final docking solutions were reported.
Synthesis
General. All commercial chemicals and solvents were reagent grade and were used without further
purification. Reaction progress was monitored by TLC on preꢀcoated silica gel 60 F254 plates
1
13
(Merck) and visualization was accomplished with UV light (254 nm). H and C NMR spectra were
recorded on a Bruker FTꢀNMR AVANCE 400. Chemical shifts are reported as δ values (ppm)
3
referenced to residual solvent (CHCl at δ 7.26 ppm, DMSO at δ 2.50 ppm, MeOD at δ 3.31 ppm); J
values were given in Hz. When peak multiplicities are given, the following abbreviations are used: s,
singlet; d, doublet; t, triplet; q, quartet; m, multiplet; br, broadened signal. Silica gel Merck (60ꢀ230
mesh) was used for column chromatography. Purity of the compounds was assayed by means of TLC
(Merck Fꢀ254 silica gel). Mass spectra were obtained on a 6520 AccurateꢀMass QꢀTOF LC/MS and
6
310AIon TrapLCꢀMS(n). The detailed NMR and mass data of the synthesized compounds are
reported in the Supporting Information (Pag. 33ꢀ37).
General procedure for the synthesis of (2E)ꢀ1ꢀ(2ꢀhydroxyꢀphenyl)ꢀ3ꢀphenylpropꢀ2ꢀenꢀ1ꢀones.
2
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
An aqueous solution of NaOH (3 M, 1.6 mL) was added to a solution of aromatic ketone (1 mmol)
and substituted benzaldehyde (1.2 equiv.), in EtOH (1ꢀ2 mL). The reaction was stirred at room
temperature overnight. The reaction mixture was cooled in an iceꢀwater bath and acidified to pH 2
with concentrated HCl (37%). The solid formed was filtered, washed with ethanol and then further
purified by recrystallization from EtOH.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
General procedure for the synthesis of methylated flavonols (compounds 9ꢀ16).
An aqueous solution of H O (30%, 250 µL) was added to an iceꢀcold suspension of the chalcone (1
2
2
mmol) in ethanol (5 mL) and 1 M NaOH (2 mL). The mixture was allowed to warm to room
temperature and was stirred for 5ꢀ18 hours. Then the reaction mixture was cooled in an ice bath and
distilled water (2ꢀ4 mL) was added. Concentrated HCl (37 %) was added until pH 2 and the
precipitate was filtered, washed with EtOH and further purified by recrystallization from EtOH. When
no precipitate occurred, the reaction mixture was extracted with dichloromethane and washed with
2 4
water and brine. The organic layer was dried over Na SO and concentrated under reduced pressure.
Column chromatography was utilized to purify the desired product.
General procedure for the synthesis of demethylated flavonols (compounds 1ꢀ8).
To a stirring solution of 15 (0.500 g, 1.52 mmol) in anhydrous dichloromethane (30 mL) under
nitrogen at 0°C, boron tribromide in dichloromethane (1.0 M, 13.7 mL, 13.7 mmol, 9 eq) was added.
The mixture was allowed to warm at room temperature and was stirred for 2 days. The reaction
mixture was then cooled to 0°C and methanol (10 mL) was added. The reaction mixture was
concentrated in vacuo. Water (10 mL) was added, the reaction sonicated and then left to stand.
Compound 7 (0.440 g, quantitative yield) was collected as a red solid.
The procedure followed for the synthesis of all demethylated flavonols is analogue to that of
compound 7 considering 3 eq of BBr
3
for each CꢀO bond cleavage.
Protein expression and purification
10
Recombinant TbPTR1 was expressed and purified by established methods with minor revisions. The
plasmid was supplied by Prof William N. Hunter University of Dundee. Briefly, protein expression
2
5
ACS Paragon Plus Environment

Page 27 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was carried out in E. coliBL21(DE3) cultured at 37°C in SuperBroth (SB) medium to midꢀlog phase
2
7
and induced with 1 mM isopropylꢀβꢀdꢀthiogalactopyranoside (IPTG) overnight at 30°C. Cells were
resuspended in 50 mM TrisꢀHCl, pH 7.5, 250 mM NaCl and disrupted by sonication. The target
protein purification was achieved in a single step using a HisTrap FF 5mL column (GEꢀHealthcare)
and an imidazole concentration of 250 mM in the same buffer. The resulting protein sample was
dialyzed overnight in 20 mM TrisꢀHCl pH 7.5 at 8°C (membrane cutoff 10 kDa). The high purity of
the sample was confirmed by SDSꢀPAGE analysis and MALDIꢀTOF mass spectrometry. The final
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ1
protein yield of enzyme was established as approximately 50 mg L bacterial culture.
The plasmid pET15bꢀLmPTR1 was supplied by Prof William N. Hunter University of Dundee.
Protein expression was performed in the E. coli strain BL21(DE3)cultured at 37°C in SB medium
(supplemented with 100 mg/L ampicillin)to midꢀlog phase and induced with 0.4 mM IPTG for 16 h at
2
8 °C. Cells, harvested by centrifugation were resuspended in buffer A (50 mM TrisꢀHCl pH 7.6, 20
mM imidazole and 250 mM NaCl) and disrupted by sonication. The supernatant, collected by
centrifugation, was further purified by nickelꢀaffinity chromatography (HisTrap FF 5mL column, GEꢀ
Healthcare) and the target protein was eluted using a 250ꢀ500 mM imidazole concentration in the
same buffer. Fractions containing the protein (identified by SDSꢀPAGE) were pooled and combined
with thrombin protease and then dialyzed overnight in 50 mM TrisꢀHCl pH 7.6 at 25°C. The mature
protein was purified by a second nickelꢀaffinity chromatography (HisTrap FF 5mL column, GEꢀ
Healthcare) and eluted as weakly bound protein, in 50ꢀ100 mM imidazole and 50 mM TrisꢀHCl pH
7.6. The resulting protein sample was dialyzed in 20 mM sodium acetate pH 5.3 and 10 mM DTT.
The high purity of the sample was confirmed by SDSꢀPAGE analysis and MALDIꢀTOF mass
spectrometry, and the final protein yield established as approximately 30 mg/L bacterial culture.
Recombinant LmDHFRꢀTS was cloned in our lab. The gene coding sequence for LmDHFRꢀTS,
cloned in pETꢀ15b expression vector, was introduced by thermal shock into the E. coli strain
ArcticExpress(DE3). Bacterial culture was grown at 30°C in ZYPꢀ5052 autoinduction medium
2
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(supplemented with ampicillin 100 mg/L) to OD600nmvalue of about 1 and then cell growth was
continued for 60 at 12°C under vigorous aeration. Cells, harvested by centrifuge , were resuspended in
buffer A (50 mM sodium Citrate pH 5.5 and 0.25 M NaCl) and disrupted by sonication. The
supernatant of the resulting crude extract was collected by centrifuge and purified by nickelꢀaffinity
chromatography. The target protein was eluted using a linear gradient 20ꢀ250 mM imidazole in the
same buffer. Fractions containing the target protein (identified by SDSꢀPAGE) were pooled and
dialyzed overnight in buffer A at 8°C. The resulting protein sample was further purified on a gel
filtration column (HiLoad 16/600 Superdex 200pg, GE Healthcare) equilibrated with 50 mM sodium
citrate pH 5.5 and 0.25 M NaCl). The high purity of the target protein was confirmed by SDSꢀPAGE
analysis and the final yield established as approximately 10 mg/L bacterial culture.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The recombinant TbDHFRꢀTS was cloned in our lab. The synthetic gene encoding for TbDHFRꢀTS,
optimized for E. coli expression and cloned in pETꢀ15b expression vector was purchased from
GenScript USA Inc. The target protein was expressed and purified using the same protocol described
for LmDHFRꢀTS, with minor modification. Briefly, cells were resuspended in buffer A (50 mM
phosphate buffer pH 7.5, 0.25 M NaCl, 10% glycerol, and 20 mM imidazole) and disrupted by
sonication. The supernatant of the resulting crude extract was collected by centrifuge and purified by
nickelꢀaffinity chromatography (HisTrap FF 5mL column, GEꢀHealthcare). The target protein was
eluted using a 400 mM imidazole concentration in the same buffer. Eluted fractions were pooled and
dialyzed overnight in buffer A at 8°C. The resulting protein sample was concentrated and applied to a
gel filtration column (HiLoad 16/600 Superdex 200pg, GE Healthcare) equilibrated in buffer A. The
high purity of the sample was confirmed by SDSꢀPAGE and the final protein yield established as
15,44
approximately 8 mg/L bacterial culture. hTS and hDHFR were purified as reported in literature.
The kinetic characterization (K and Kcat) of all the proteins purified and employed for the target/off
m
target screening is reported in Table S16 of the Supporting Information.
Xꢀray crystallography
2
7
ACS Paragon Plus Environment

Page 29 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Crystals of histidineꢀtagged TbPTR1 were obtained by the vapor diffusion hanging drop method
45
technique at room temperature. Drops were prepared by mixing equal volumes of protein (6ꢀ10 mg
ꢀ1
mL in 20 mM TrisꢀHCl pH 7.5 and 5ꢀ10 mM DTT) and precipitant (1.5ꢀ2.5 M sodium acetate and
0
.1 M sodium citrate pH 5) solutions and equilibrated over a 600 µL reservoir. Wellꢀordered
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
monoclinic crystals grew in 3ꢀ6 days. The TbPTR1ꢀcofactorꢀinhibitor complexes were obtained by
soaking crystals with each compound solubilized either in 1,4ꢀdioxane or DMSO (without exceeding
a dioxaneꢀDMSO/crystal solution ratio of 1:9). After 4ꢀ21 hours crystals were transferred to a cryoꢀ
protectant, prepared by adding 30 % v/v glycerol to the precipitant solution, and flash frozen in liquid
nitrogen. Diffraction data for the complexes with compound NPꢀ29 and NPꢀ13 were measured using
synchrotron radiation at the Diamond Light Source (DLS, Didcot, United Kingdom) beamline I04
equipped with a Pilatus 6MꢀF detector, using a wavelength of 0.9795 Å, a 0.25° rotation and a 0.4 s
+
exposure time/image. Data for the complex TbPTR1ꢀNADPH/NADP ꢀ2 were collected at the DLS
beamline I03 equipped with a Pilatus3 6M detector and using a wavelength of 0.9173 Å, a 0.25°
rotation and a 0.15 s exposure time/image. Data relative to the complex with compound 7 were
collected at the European Synchrotron Radiation Facility (ESRF, Grenoble, France) beamline ID30Aꢀ
1
, equipped with a Dectris Pilatus3 2M detector, and by using a wavelength of 0.9660 Å, a 0.1°
rotation and a 0.08 s exposure time/image. Xꢀray images were integrated using Mosflm and scaled
46ꢀ48
with SCALA within the CCP4 software suite.
TbPTR1 crystals belong to the monoclinic space
group P2 , showing only slight variations in cell parameters among the different crystals. Data
1
collection and processing statistics are reported in Table S2 (Supplementary Information). The
49
structures were solved by molecular replacement using the software Molrep and a whole tetramer of
50
51
TbPTR1 (PDB code 2X9G) as searching model. Structural models were refined using REFMAC5 ,
52
whereas the program Coot was used for electron density inspection, model manipulation, and
placement of the solvent molecules. The final models were inspected manually and checked with the
53
programs Coot and Procheck. Data refinement statistics are reported in Table S3. Figures were
54
generated with CCP4mg.
2
8
ACS Paragon Plus Environment