Tetrahydro-2-naphthyl and 2-Indanyl Triazolopyrimidines Targeting Plasmodium falciparum Dihydroorotate Dehydrogenase Display Potent and Selective Antimalarial Activity

Article abstract of DOI:10.1021/acs.jmedchem.6b00275

Malaria persists as one of the most devastating global infectious diseases. The pyrimidine biosynthetic enzyme dihydroorotate dehydrogenase (DHODH) has been identified as a new malaria drug target, and a triazolopyrimidine-based DHODH inhibitor 1 (DSM265) is in clinical development. We sought to identify compounds with higher potency against Plasmodium DHODH while showing greater selectivity toward animal DHODHs. Herein we describe a series of novel triazolopyrimidines wherein the p-SF₅-aniline was replaced with substituted 1,2,3,4-tetrahydro-2-naphthyl or 2-indanyl amines. These compounds showed strong species selectivity, and several highly potent tetrahydro-2-naphthyl derivatives were identified. Compounds with halogen substitutions displayed sustained plasma levels after oral dosing in rodents leading to efficacy in the P. falciparum SCID mouse malaria model. These data suggest that tetrahydro-2-naphthyl derivatives have the potential to be efficacious for the treatment of malaria, but due to higher metabolic clearance than 1, they most likely would need to be part of a multidose regimen.

Full text of DOI:10.1021/acs.jmedchem.6b00275

Subscriber access provided by HOWARD UNIV
Article
Tetrahydro-2-naphthyl and 2-indanyl triazolopyrimidines
targeting Plasmodium falciparum dihydroorotate
dehydrogenase display potent and selective antimalarial activity
Sreekanth Kokkonda, Xiaoyi Deng, Karen L. White, Jose M. Coterón, Maria Marco, Laura de las Heras,
John White, Farah El Mazouni, Diana R Tomchick, Krishne Manjalanagara, Kakali Rani Rudra, Gong
Chen, Julia Morizzi, Eileen Ryan, Werner Kaminsky, Didier Leroy, Maria Santos Martínez-Martínez,
Maria Belen Jimenez Diaz, Santiago Ferrer Bazaga, Iñigo Angulo-Barturen, David Waterson, Jeremy
N. Burrows, Dave Matthews, Susan A. Charman, Margaret A. Phillips, and Pradipsinh K Rathod
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00275 • Publication Date (Web): 29 Apr 2016
Downloaded from http://pubs.acs.org on April 30, 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 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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 68
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
Tetrahydro-2-naphthyl and 2-indanyl triazolopyrimidines targeting
Plasmodium falciparum dihydroorotate dehydrogenase display potent and selective
antimalarial activity
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
2
3
4
4
Sreekanth Kokkonda , Xiaoyi Deng , Karen L. White , Jose M. Coteron , Maria Marco , Laura
4
1
2
5
6
de las Heras , John White , Farah El Mazouni , Diana R. Tomchick , Krishne Manjalanagara ,
6
3
3
3
1
Kakali Rani Rudra , Gong Chen , Julia Morizzi , Eileen Ryan , Werner Kaminsky , Didier
7
4
4
4
Leroy , María Santos MartínezꢀMartínez , Maria Belen JimenezꢀDiaz , Santiago Ferrer Bazaga ,
4
7
7
7
Iñigo AnguloꢀBarturen , David Waterson , Jeremy N. Burrows , Dave Matthews , Susan A.
3
2*
1*
Charman , Margaret A. Phillips and Pradipsinh K. Rathod
1
Departments of Chemistry and Global Health, University of Washington, Seattle, WA 98195;
2
5
Departments of Pharmacology and Biophysics, University of Texas Southwestern Medical
3
Center at Dallas, 6001 Forest Park Blvd, Dallas, Texas 75390ꢀ9041; Centre for Drug Candidate
Optimisation, Monash Institute of Pharmaceutical Sciences, Monash University, Parkville, VIC
4
3052, Australia, GSK, Tres Cantos Medicines Development Campus, Severo Ochoa, Madrid,
6
7
2
8760 Spain , Syngene International Ltd, Bangalore, India, 560 099; Medicines for Malaria
Venture, 1215 Geneva, Switzerland.
*Authors to whom all correspondence should be addressed.
ACS Paragon Plus Environment

Page 3 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Abstract
Malaria persists as one of the most devastating global infectious diseases. The pyrimidine
biosynthetic enzyme dihydroorotate dehydrogenase (DHODH) has been identified as a new
malaria drug target and a triazolopyrimidineꢀbased DHODH inhibitor 1 (DSM265) is in clinical
development. We sought to identify compounds with higher potency against Plasmodium
DHODH while showing greater selectivity towards animal DHODHs. Herein we describe a
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
series of novel triazolopyrimidines wherein the pꢀSF ꢀaniline was replaced with substituted
5
1,2,3,4ꢀtetrahydroꢀ2ꢀnaphthyl or 2ꢀindanyl amines. These compounds showed strong species
selectivity and several highly potent tetrahydroꢀ2ꢀnaphthyl derivatives were identified.
Compounds with halogen substitutions displayed sustained plasma levels after oral dosing in
rodents leading to efficacy in the P. falciparum SCID mouse malaria model. These data suggest
that tetrahydroꢀ2ꢀnaphthyl derivatives have the potential to be efficacious for the treatment of
malaria, but due to higher metabolic clearance than 1, they most likely would need to be part of a
multiꢀdose regimen.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 68
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
Malaria is one of the most deadly infectious diseases in human history with 3.2 billion
1
people in 97 countries at risk. An estimated 444,000 deaths from malaria were reported 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
WHO in 2015 and ~90% of these occurred in subꢀSaharan Africa, mostly amongst children
under the age of five. Human malaria, which is transmitted by the female Anopheles mosquito,
can be caused by five species of Plasmodia, however Plasmodium falciparum and Plasmodium
2, 3
vivax are the most signficant. P. falciparum is dominant in Africa and accounts for most of the
deaths, while P. vivax has a larger global distribution. To simplify treatment options it is
4
desirable that new drugs be efficacious against all human infective species. Malaria is a
treatable disease and malarial control programs depend on drug therapy for treatment and
chemoprevention, and on insecticides (including insecticide impregnated bed nets) to prevent
2
transmission. A large collection of drugs has been used for the treatment of malaria, but many of
the most important compounds have been lost to drug resistance (e.g. chloroquine and
4
ꢀ6
pyrimethamine).
Artemisinin combination therapies (ACT) replaced older treatments,
becoming highly effective, crucial tools in global efforts that have led to the decline in malaria
deaths over the past decade. However, resistance to the artemisinin components (associated with
7
ꢀ9
Kelch13 propeller protein mutations ) has been found in Southeast Asia putting at risk malaria
treatment programs. To combat drug resistance a significant effort is underway to identify new
compounds that can be used for the treatment of malaria, with several new entities currently in
4
, 5, 10
clinical development.
The triazolopyrimidine DSM265 (1) (Fig. 1) developed by our group is the first
antimalarial agent that targets dihydroorotate dehydrogenase (DHODH) to reach clinical
1
1, 12
development, validating this target for the treatment of malaria.
DHODH is a mitochondrial
ACS Paragon Plus Environment

Page 5 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
enzyme that is required for the fourth step of de novo pyrimidine biosynthesis, catalyzing the
flavinꢀdependent oxidation of dihydroorotate to orotic acid with mitochondrially derived
1
3
coenzyme Q (CoQ) serving as a second substrate. Pyrimidines are essential for both RNA and
DNA biosynthesis and because Plasmodia do not encode pyrimidine salvage enzymes, which are
found in humans and other organisms, the de novo pyrimidine pathway and DHODH are
essential to the parasite. We identified the triazolopyrimidine DHODH inhibitor series by a
targetꢀbased high throughput screen, and the initial lead DSM1 (2) (Fig. 1) was shown to
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
selectively inhibit P. falciparum DHODH and to kill parasites in vitro but it was ineffective in
14ꢀ16
vivo due to poor metabolic properties.
The series was subsequently optimized to improve its
in vivo properties resulting in the identification of DSM74 (3), which while metabolically stable
1
6
lacked potency. Xꢀray structures of 2 and 3 bound to PfDHODH were then used to guide the
medicinal chemistry program in the search for more potent analogs, resulting in the identification
1
1, 12, 17
of 1.
4
While 1 has progressed successfully to Phase II clinical trials, we sought to identify
potential backup compounds if unforeseen issues arise during its clinical development.
Importantly, 1 has potent activity on PfDHODH and P. falciparum parasites in vivo, and its
pharmacokinetic properties support its use as a single dose treatment or once weekly
12
prophylactic. However, while 1 has excellent selectivity against human DHODH and thus is
not expected to show on target activity in humans, it does have some activity against rodent
DHODHs that has complicated animal toxicity testing. We therefore sought to identify
compounds that have broad selectivity against all mammalian DHODHs, while if possible
identifying compounds with higher potency against PfDHODH that could potentially reduce the
dose required for treatment. We describe herein a series of triazolopyrimidines where the SF5ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 68
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
aniline moiety of 1 has been replaced by either substituted 1,2,3,4ꢀtetrahydroꢀ2ꢀnaphthyl or 2ꢀ
indanyl amines. Compounds from both series are inactive against a range of mammalian
DHODHs, while several of the identified tetrahydronapthalenes are more potent than 1 against P.
falciparum parasites in vitro and indeed are amongst the most potent compounds that have been
identified in the series. However these compounds are less metabolically stable than 1 in mouse
liver microsomes, suggesting the likelihood of higher clearance in mice and explaining the
requirement for higher doses to inhibit parasites in the in vivo mouse model of malaria.
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
ACS Paragon Plus Environment

Page 7 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Chemistry
The focus of the study was to examine the structure activity relationships (SAR) of replacing the
pꢀSF ꢀaniline of 1 with substituted 1,2,3,4ꢀtetrahydroꢀ2ꢀnaphthyl or 2ꢀindanyl amines with the
5
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
goal of eliminating activity on rodent DHODHs, while potentially improving potency against P.
falciparum DHODH and thereby lowering the dose required for in vivo efficacy. The clinical
candidate 1 is a potent inhibitor of PfDHODH with only minimal activity detected against human
DHODH (selectivity window >5000ꢀfold; Table 1). However, 1 showed inhibitory activity
against rodent and to a lesser extent dog DHODHs, although potency towards these enzymes
1
2
remained considerably less than for PfDHODH. Because these species are required for toxicity
testing in preclinical studies we sought to eliminate rodent and dog DHODH inhibitory activity
for any potential candidate that might be developed as a backup to 1. We previously reported that
substitution of the pꢀSF ꢀaniline with pꢀCF aniline in the context of either mono or difluoro
5
3
1
8
groups at the meta position led to greater inhibition of the mammalian enzymes. Xꢀray
structural data suggested that increasing hydrophobicity drove the interaction of these
compounds with the mammalian enzymes. We hypothesized based on these structures that
increasing the size of the aromatic amine might prevent binding to mammalian DHODHs, while
increasing potency towards the parasite enzyme. We previously tested several tetrahydroꢀ2ꢀ
naphthyl and 2ꢀindanyl amines as substitutes for the CF ꢀaniline of 3 but found these compounds
3
lacked potency and furthermore, the one analog containing 1,2,3,4ꢀtetrahydroꢀ2ꢀnaphthyl also
1
9
lacked metabolic stability in mouse microsomes.
However, these amines were never tested
within the context of haloalkyl groups (CF CH or CF ) at the C2 position of the
2
3
3
triazolopyrimidine ring, which was a key substitution required to boost the potency of the
1
1
series. We therefore synthesized a series of tetrahydroꢀ2ꢀnaphthyl or 2ꢀindanyl analogs with
halo alkyl groups at the C2 position. Substituents (halo, haloalkyl, Me, CF , OMe and SO Me
3
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 68
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
and SO NMe ) were incorporated onto the tetrahydroꢀ2ꢀnaphthyl or 2ꢀindanyl aromatic rings to
2
2
improve metabolic stability over the unsubstituted analogs.
The synthetic strategy to generate the appropriately substituted triazolopyrimidines has
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
11, 15, 16
previously been described.
For the studies described herein, 7ꢀchloroꢀ2ꢀ(1,1ꢀ
difluoroethyl)ꢀ5ꢀmethyl[1,2,4]triazolo[1,5ꢀa] pyrimidine (4a) and 7ꢀchloroꢀ2ꢀ(trifluoromethyl)ꢀ5ꢀ
methyl[1,2,4]triazolo[1,5ꢀa] pyrimidine (4b) were prepared using these methods (Scheme 1) and
were reacted with the requisite amines to afford the final products containing either the 1,2,3,4ꢀ
tetrahydroꢀ2ꢀnaphthyl (5-25) or 2ꢀindanyl (26-37) amines in place of the pꢀSF ꢀaniline of 1.
5
Most amines utilized for the study were commercially available with the exception that amines
for compounds 29ꢀ37, were synthesized as described in Scheme 2 and supplementary Scheme
S1. The 1,2,3,4ꢀtetrahydroꢀ2ꢀnaphthyls contained a chiral center and for select compounds,
individual enantiomers were purified using a chiral column as described in the experimental
section.
Scheme 1. General synthetic method
The amine precursor 5ꢀ(methylsulfonyl)ꢀ2,3ꢀdihydroꢀ1Hꢀindenꢀ2ꢀamine HCl 43 was
prepared in five steps (Scheme 2). First, 2ꢀaminoindane 38 was acetylated with acetic anhydrideꢀ
sodium acetate to yield the acetamido derivative 39. Further, regioselective chlorosulfonylation
ACS Paragon Plus Environment

Page 9 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of 39 led to the sulfonylchloride intermediate 40. Reduction of the sulfonylchloride to the thiol
was achieved by tin chloride, which was converted to a methylthio derivative 41 using sodium
methoxide, methyliodide. Next, oxidation of methylthio intermediate 41 with mCPBA gave the
acetyl protected methanesulfonyl derivative 42. Upon deprotection of this acetyl group using 3N
HCl, the required amine precursor 43 was obtained.
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
Scheme 2. Synthesis of 5-(methylsulfonyl)-2,3-dihydro-1H-inden-2-amine. HCl (43)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 68
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
Results
Optimization of inhibitor potency and selectivity.
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
Compounds were first analyzed for potency against P. falciparum, P. vivax and human
DHODH, and for activity on P. falciparum parasites in vitro (Table 1). Compounds in the series
showed IC ’s against PfDHODH in the range of 0.005 – 8 ꢁM, and EC ’s against P. falciparum
5
0
50
3D7 parasites in whole cell assays between 0.00039 – 6.5 ꢁM (Table 1). Generally, the
tetrahydroꢀ2ꢀnaphthyl compounds (5-25) were more potent in both assays than the 2ꢀindanyls
(26-37). The most potent analogs based on the whole cell P. falciparum assay were 9 and 13,
both of which contain a halogen (chloro or bromo, respectively) in the 7 position. Compound 13
is the most potent analog identified in the triazolopyrimidine series that has been reported.
Overall the trend was for 7 substituted tetrahydroꢀ2ꢀnaphthyls to be more potent than those with
substitutions at the 6 position. Compounds with Br and Cl substituents (7, 9, 11, 13) were more
potent than those with fluoro 6, OMe (15, 16) or with disubstitutions (19, 20, 21), with the least
active tetrahydroꢀ2ꢀnaphthyl 21 containing a 6ꢀCF , 7ꢀF. Potency of 21 in the whole cell parasite
3
assay was worse than expected based on the inhibitory activity against PfDHODH, suggesting
that the compound may be poorly transported into the parasite. Analogs with CF at C2 (22, 23,
3
24) on the triazolopyrimidine ring were less active than those with CF CH (5, 6, 9) at this
2
3
position. Within the indanyl series the most potent compounds were 32 and 36 containing 4,7
dimethyl or 4ꢀCF substituents respectively, while the least active compounds were 30 and 31
3
containing 5ꢀSO Me or 5ꢀSO NMe substituents, respectively.
2
2
2
For the tetrahydroꢀ2ꢀnaphthyls where individual enantiomers were characterized,
differences in potency between the active enantiomer (7, 9, 11, 13, 16, 19, 24) and the inactive
ACS Paragon Plus Environment

Page 11 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
enantiomer (8, 10, 12, 14, 17, 18, 25) ranged from 20ꢀ360ꢀfold towards PfDHODH and from 30ꢀ
2000ꢀfold against P. falciparum 3D7 parasites. The small molecule Xꢀray structure of the active
enantiomer 9 was solved demonstrating that it was in the S configuration (Fig. S1 and Table S1).
We did not isolate the individual enantiomers for compounds from the indane series (27, 28, 30,
31, 34 and 36), and it is likely that the purified active enantiomers would also show higher
activity for these compounds as well. Thus the reported potency data for the racemic mixture
likely overestimates the true IC by ~2ꢀfold.
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
5
0
Overall we observed a good correlation between potency on PfDHODH and potency on
D7 P. falciparum parasites (Fig. S2). There was a tendency for the most potent compounds on
3
the parasite to show lower potency against PfDHODH, but this difference is likely caused by the
complication of stoichiometric binding for compounds with IC ’s at or below the concentration
5
0
of enzyme used in the assay (PfDHODH = 5 nM). In order to confirm that for the most potent
compounds parasite killing is due to DHODH inhibition, we tested select compounds (9 and 13)
for their ability to inhibit growth of a genetically engineered parasite strain that expresses yeast
20
DHODH (D10 yDHODH ). This strain is resistant to DHODH inhibitors if their mechanism of
action is on target. The EC ’s (D10 yDHODH EC = 8.6 and 7.4 ꢁM for 9 and 13, respectively)
50
50
for both compounds were increased by >600ꢀfold against this parasite strain relative to the wildꢀ
type 3D7 parasites, supporting DHODH inhibition as their mechanism of action (Table 1 and
Fig. S3). The addition of proguanil did not reverse the resistance of this strain, which is also
consistent with DHODH as the mechanism of parasite killing. In contrast, it has been previously
reported that proguanil reverses the atovaquone resistance phenotype of the D10 yDHODH
parasite strain, providing a mechanism to distinguish between DHODH and bc1 targeted parasite
20
growth inhibition. The weak but detectable activity observed on the D10 yDHODH parasites
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 68
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
for both compounds suggests that at high concentrations there is a secondary target. For a couple
of select compounds (9 and 35) whole cell activity was also tested on several additional cell lines
including chloroquine and pyrimethamine resistant parasites (Table S2). Both compounds were
equally active on all tested strains.
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
Species selectivity was first evaluated by testing against human and P. vivax DHODH
(Table 1) and then for a selection of the more potent compounds, inhibitor activity was also
measured for mouse, rat and dog DHODH (Table 2). None of the compounds showed any
inhibition of human DHODH up to the highest tested concentration (100 ꢁM), which is similar to
results for 1. However unlike 1, none of the compounds showed any activity against either the
rodent or dog DHODHs (Table 2), demonstrating that we had achieved our objective of
eliminating mammalian DHODH activity. Measured IC ’s for P. vivax DHODH ranged from
5
0
1.5 – 10ꢀfold higher than for PfDHODH. The 7ꢀposition substituted tetrahydroꢀ2ꢀnaphthyls, 9
and 13 showed minimal differences (1.5 ꢀ 3ꢀfold) between the two enzymes suggesting that they
would show good activity against both P. falciparum and P. vivax parasites, whereas the 6
position compounds (7 and 11) had 5 ꢀ 8ꢀfold lower activity on P. vivax than P. falciparum
DHODH.
Xꢀray structure of 13 bound to PfDHODH.
In order to understand the structural basis for the superior potency of 13 we solved its Xꢀray
structure bound to PfDHODH. The structure was solved to 2.32 Å resolution at an Rwork and Rfree
of 0.18 and 0.21, respectively (Fig. 2 and Table S3). Strong electron density was observed for the
entirety of 13 (Fig. S4A). Compound 13 was oriented in the pocket similarly to other
1
2, 17, 18
triazolopyrimidines (e.g. 1 - 3
) with the triazolopyrimidine ring bound adjacent to the
ACS Paragon Plus Environment

Page 13 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
flavin cofactor in position to form Hꢀbond interactions with Argꢀ265 and Hisꢀ185 (Fig. 2 and
S4B). The tetrahydroꢀ2ꢀnaphthyl moiety bound in a hydrophobic pocket forming edgeꢀtoꢀface
stacking interactions with Pheꢀ227 and Pheꢀ188 and was in a very similar orientation to what we
previously observed for the naphthyl moiety of 2 (Fig. 2). Some modest conformational
differences in ring geometry were observed reflecting the difference between the fully planer and
aromatic naphthyl and the puckered configuration of the nonꢀaromatic ring of the tetrahydroꢀ2ꢀ
naphthyl. A halogen bond was observed between the 7ꢀbromo group on the tetrahydroꢀ2ꢀ
naphthyl of 13 and Cysꢀ233 SH, which likely provides significant binding energy to the enzyme
inhibitor interaction. This is a novel interaction that has not been previously observed.
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
Physicochemical properties and plasma protein binding
Selected compounds were evaluated for their physicochemical properties to determine if
they had good drugꢀlike properties. Analysis included in silico calculations, chromatographic
estimation of Log D (pH 7.4), aqueous solubility and plasma protein binding (Tables 3 and 4).
Tested compounds had physicochemical properties that are suggestive of good oral absorption
2
(
MW <430, H bond donors ≤ 2, H bond acceptors ≤ 6, polar surface area <70 Å ). Log D ranged
from 3.1 – 4.1 and the more potent analogs tended to have higher Log D values. Aqueous
solubility (pH 6.5) was poor (1.6 – 6.3 ꢁg/mL) to moderate (12.5 – 25 ꢁg/mL), and in general
compounds in this series had lower kinetic solubility than 1 (Table 3). More extensive solubility
studies in simulated gastric and intestinal fluids was conducted on a selection of the most potent
analogs from the 2ꢀindanyl (35) and the tetrahydroꢀ2ꢀnaphthyl (7, 9, 11, 13) series (Table 4). All
five compounds had better solubility than 1 in 0.1 N HCl. Solubility in fastedꢀstate simulated
intestinal fluid was similar to or slightly improved compared to 1. The tetrahydroꢀ2ꢀnaphthyl
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 68
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 substituted at the 6 position (11 and 7) showed considerably better solubility in fedꢀ
state simulated intestinal fluid than either 1 or the 7 position (9 and 13) substituted compounds.
Human, rat and mouse plasma protein binding (PPB) was assessed for the more potent
compounds that were progressed to in vivo pharmacokinetic testing (7, 9, 11, 13, 35). The
tetrahydroꢀ2ꢀnaphthyl compounds substituted at the 6 position (7 and 11) showed higher protein
binding compared to the compounds substituted at the 7 position (9 and 13). Binding was similar
across the species tested for each of the compounds. By comparison, 1 exhibited high protein
binding in human and mouse plasma but lower binding in rat plasma. The indanyl analog 35 had
considerably lower protein binding compared to either 1 or the tetrahydroꢀ2ꢀnaphthyl
compounds.
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
In vitro metabolism.
To obtain a preliminary indication of the likelihood that compounds would show good in
vivo pharmacokinetic properties, selected compounds were analyzed for metabolic stability in
vitro using human, rat and mouse liver microsomes. We previously found that the in vitro
microsomal stability of compounds from the triazolopyrimidine series provided a good rank
11, 15, 16, 19
order estimation of which compounds would have the highest in vivo exposure.
We
report in Table 3 the in vitro intrinsic clearance (CL , µL/min/mg microsomal protein) and the
int
predicted in vivo intrinsic clearance (mL/min/kg) obtained using physiologicallyꢀbased scaling
2
1
factors as previously described. Given the early stage of optimization for the series, protein
binding was not determined for all compounds and therefore no attempt was made to predict the
in vivo blood clearance (which requires corrections for both plasma and microsomal binding). In
vitro CL values of <20 µL/min/mg protein were taken as being suggestive of good metabolic
int
ACS Paragon Plus Environment

Page 15 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stability, assuming that clearance only occurs via hepatic metabolism. Overall, the substituted
tetrahydroꢀ2ꢀnaphthyls showed good metabolic stability in human liver microsomes but they
were less stable in mouse microsomes (Table 3). Chloro, bromo and CF substituted compounds
3
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
(
7, 9, 11, 12, 13, 20, 21) were marginally more metabolically stable than the fluoro (6, 19) or
OMe 16 substituted compounds.
With the exception of 35, the indanyl analogs were generally metabolically unstable
showing high intrinsic clearance in mouse liver microsomes and moderate to high values in
human and rat microsomes. The least stable compounds were the unsubstituted 2ꢀindanyl 26 and
the 4,7 dimethyl substituted derivative 32. Replacement of both methyl groups with fluorines 33
did not improve stability, although perhaps surprisingly the single 4ꢀF derivative 34 was
significantly more stable. Substantial improvement in metabolic stability was obtained by
moving the fluorines to the 5 and 6 positions of the ring 35, leading to low intrinsic clearance in
human and rat microsomes and moderate clearance in mouse microsomes. These data suggested
that 35 would show good plasma exposure in both mice and rats.
hERG channel activity and CYP inhibition.
In order to evaluate any potential cardiac risks, select compounds were tested for
+
inhibition of the human etherꢀaꢀgoꢀgoꢀrelated gene (hERG) K channel in a standard patch clamp
assay. hERG channel inhibition has been associated with QT prolongation and arrhythmias and
patch clamp assays have become a routine method to provide an initial analysis that a compound
2
2
may potentially be associated with QT syndromes in humans. Compounds in both the
tetrahydroꢀ2ꢀnaphthyl and 2ꢀindanyl series inhibited the hERG channel with IC ’s in the range
5
0
of 0.5 – 1.8 ꢁM (Table 3).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 68
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
In order to test for the potential for drugꢀdrug interactions, select compounds were tested
for their ability to inhibit cytochrome P450 isoforms in human liver microsomes (Table 5). All
tested compounds showed moderate inhibition of isoforms 2C9 and 2D6. The most significant
inhibition was observed for 9 and 13 as both inhibited these two isoforms in the 2 – 4 ꢁM range.
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
In vivo pharmacokinetic studies in rats and mice.
Five of the most potent compounds were selected for assessment of their in vivo
pharmacokinetic properties based on their good in vitro metabolic stability. Mice were dosed
orally while rats were dosed orally and intravenously to allow determination of the clearance,
volume of distribution and oral bioavailability. Compounds included one from the indanyl series
(35) and four from the tetrahydroꢀ2ꢀnaphthyl series (7, 9, 11, 13).
All five compounds showed good exposure over 24 h in both mice and rats after oral
administration (Tables 6 and 7, Figs. 3 and 4). Compounds were well tolerated at the
administered doses and no adverse reactions were observed. In mice, Cmax and AUC24h values
were highest for 35, and lower for the tetrahydroꢀ2ꢀnaphthyl compounds. Compared to 1 and
taking into account the differences in dose, total Cmax and AUC24h values were similar or lower
for the tetrahydroꢀ2ꢀnaphthyl compounds, whereas unbound Cmax and AUC24h were similar (for
7) or somewhat higher (3ꢀ4ꢀfold for 9, 11, and 13). The indanyl (35) had the highest plasma
exposure in mice in terms of both total and unbound Cmax and AUC24h and significantly higher
exposure compared to 1. Similar to 1, the terminal elimination halfꢀlife (t ) in mice was ~2 h for
1
/2
all five compounds (Table 6 and Fig. 3).
In rats, the oral bioavailability was high for all compounds tested, ranging from 63 –
100% compared to 57% for 1 (Table 7 and Fig. 4). The plasma clearance and volume of
ACS Paragon Plus Environment

Page 17 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
distribution were significantly lower for the tetrahydroꢀ2ꢀnaphthyl compounds substituted on the
6
position (7 and 11) than for those substituted on the 7 position (9 and 13), however when
corrected for differences in plasma protein binding, values across the four tetrahydroꢀnaphthyl
compounds were comparable (within 2ꢀfold). The 6 position compounds also had lower blood to
plasma partitioning consistent with their higher plasma protein binding compared to the 7
position analogs. Unbound clearance values for the tetrahydroꢀ2ꢀnaphthyl compounds were
somewhat lower for the two chloro substituted compounds (438 and 589 mL/min/kg for 7 and 9,
respectively) compared to the bromo compounds (933 and 800 mL/min/kg for 11 and 13,
respectively) and all were higher than for 1 (263 mL/min/kg). As a result, unbound AUC values
for the tetrahydroꢀ2ꢀnaphthyls were approximately 2ꢀ4 fold lower than 1 after oral dosing at
approximately 20 mg/kg. The t for the tetrahydroꢀ2ꢀnaphthyl compounds ranged from 6ꢀ7 h in
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
/2
comparison to 13 h for 1. The indanyl compound (35) had low unbound clearance and volume of
distribution and overall unbound oral exposure was superior to 1 and to the tetrahydroꢀ2ꢀ
naphthyls (Table 7).
In vivo efficacy in the SCID mouse model of P. falciparum
Based on the observation of good oral exposure after dosing in mice, we tested three
compounds in the SCID mouse P. falciparum efficacy model, which has become the standard
23
model for evaluating the efficacy of new antimalarial compounds. The tetrahydroꢀ2ꢀnaphthyl
compounds substituted at the 7 position (9 and 13) were more potent against P. falciparum in the
in vitro parasite assay than those substituted on the 6 position (7 and 11) and showed fairly
equivalent exposure in mice after oral dosing. We therefore decided to test both a 6ꢀ and 7ꢀ
substituted tetrahydroꢀ2ꢀnaphthyl (7 and 9) in the P. falciparum SCID mouse model. Because 35
showed significantly higher plasma exposure in mice than the remaining compounds, we also
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 68
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
selected this compound for an efficacy study despite the 10ꢀfold lower potency in the P.
falciparum parasite assay.
All three compounds were dosed orally once a day for 4 days and tested for efficacy and
blood concentrations at two doses (10 and 30 mg/kg). Each was well tolerated and there were no
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
adverse reactions. The effective dose that led to 90% parasite clearance (ED ) was calculated
90
2
4 h after the last dose (Table 8 and Fig. 5A). Pharmacokinetic sampling was performed on day
to allow the total blood AUC at the ED90 to be assessed (Tables 8 and S4 and Fig. 5B).
1
Because of the complicated nature of blood in the SCID mouse model (i.e. mice contain a
mixture of human and mouse plasma and erythrocytes and a hematocrit of 70ꢀ80%), it was not
possible to accurately convert blood concentrations to plasma concentrations or to correct
concentrations for differences in plasma protein binding.
All three compounds showed efficacy in this model, however 9 significantly
outperformed the other two compounds based on ED (Tables 8 and S4 and Fig. 5). The ED
90
9
0
was higher for all three compounds than for 1 ranging from ~19 mg/kg for 9 to > 30 mg/kg for 7
and 35. The pharmacokinetic data indicated that as expected, 35 exhibited higher total blood
concentrations than the other two compounds. Both 7 and 9 were approximately dose linear for
both Cmax and AUC, whereas 35 showed significant nonꢀlinearity and evidence of a second peak
at 8 h at the lower dose (Table S4). Despite requiring a higher dose, 9 showed higher apparent
efficacy compared to 1 based on the total blood AUC at the ED which was 5ꢀfold lower than
90
that required for 1 in the once daily dosing regimen (Table 8). However, when accounting for
differences in the plasma unbound fractions for 1 and 9 (approximately 0.002 for 1 and 0.03 for
9
, Table 3), 1 likely has lower unbound concentrations at the ED , and hence higher intrinsic
90
efficacy. In contrast, 35 was significantly less effective than 1 suggesting that the lower intrinsic
ACS Paragon Plus Environment

Page 19 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potency of 35 could not be overcome by higher exposure levels. The
pharmacokinetic/pharmacodynamic relationship for 7 could not be evaluated in detail as the
ED was not reached in this study.
9
0
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 20 of 68
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
Discussion
Malaria is one of the most serious global infectious diseases and while a number of
effective drugs have been used to treat malaria, drug resistance has led to the loss of most agents
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
that have been in wide spread use. A robust drug pipeline is therefore key to malaria control and
4
elimination efforts. DHODH has emerged as a key new target in efforts to identify antimalarial
drugs that work on clinical isolates resistant to standard therapies and as noted, the
12
triazolopyrimidine 1, is currently in clinical development. In order to identify potential backup
compounds to 1 in case issues arise in its clinical development, we sought to identify additional
compounds within the triazolopyrimidine series with the potential to advance into preclinical
development. One identified potential liability of 1 has been the finding that while it does not
inhibit human DHODH, it does show moderate inhibition of the mouse and rat enzymes, which
complicates toxicity testing in animals. We therefore sought to identify new derivatives that
showed broad selectivity against all of the key mammalian enzymes. Herein we describe a series
of tetrahydroꢀ2ꢀnaphthyl and 2ꢀindanyl triazolopyrimidine analogs, which display a range of
potencies against Plasmodium DHODH, but which importantly lack activity against human
DHODH or any of the key mammalian enzymes from species that are important for toxicological
profiling (mouse, rat and dog). Pharmacologic profiling and in vivo efficacy assays suggest that
several of the identified analogs have potential, with 9 showing the most promising profile.
In evaluating the tetrahydroꢀ2ꢀnaphthyl and 2ꢀindanyl triazolopyrimidine analogs we
found that the tetrahydroꢀ2ꢀnaphthyls (e.g. 7, 9, 11, 13) had significantly greater potency on
PfDHODH and Plasmodium parasites in whole cell assays than the 2ꢀindanyl analogs (e.g. 35).
The most potent of the tetrahydroꢀ2ꢀnaphthyl derivatives also showed good activity (within 1.5ꢀ
fold) against P. vivax DHODH suggesting that they would be useful for the treatment of both P.
ACS Paragon Plus Environment

Page 21 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
falciparum and P. vivax. In contrast, the best of the 2ꢀindanyl analogs (35) was 7ꢀfold less active
on P. vivax DHODH suggesting it might not have good activity against the P. vivax parasite.
The Xꢀray structure of the 7ꢀbromo analog 13 showed that in addition to the previously
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
11, 12, 17, 18
observed Hꢀbonds and stacking interactions observed for other analogs in the series,
a
good halogen bond was formed between the bromo group of 13 and Cysꢀ233 that is likely to
provide considerable binding energy to the interaction. Indeed halogen bonding to protein
residues that include Lewis bases (e.g. sulfur) are commonly observed in protein structures
2
4
deposited in the PDB. The overall binding mode of 13 is very similar to that of 2, which was
expected based on their close structural similarity. Like the naphthyl of 2, the tetrahydroꢀ2ꢀ
naphthyl of 13 better fills the available binding pocket than the SF ꢀaniline of 1, which supports
5
our initial hypothesis that compounds with improved potency could be identified by substitution
of the SF ꢀaniline with bulkier aromatic groups.
5
Comparison of the tetrahydroꢀ2ꢀnaphthyl compounds containing 7 position (9 and 13)
versus 6 position (7 and 11) substitutions identified a number of interesting differences. Analogs
substituted in the 7 position were more potent than 6 position analogs, and indeed 13 with sub
nanomolar activity against the parasite in whole cell assays is the most potent analog that has
been identified in the series. However these assays did not take into account potential differences
in protein binding. The 6 position analogs exhibited consistently higher plasma protein binding
than the 7 position compounds and a similar trend is also likely for binding to albumax present in
the in vitro test medium. Assessment of the pharmacologic properties showed that while the 7
position tetrahydroꢀ2ꢀnaphthyls had greater apparent antimalarial potency, analogs substituted at
the 6 position had somewhat improved physicochemical properties. The 6 position analogs 7 and
1
1 were more soluble (in 0.1 N HCl and FeSSIF media) than either of the 7 position analogs (9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 68
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
and 13). While the total plasma exposure (AUC and Cmax) in rats was higher for the 6 position
analogs, correction of the data for protein binding indicated that the unbound exposure and
unbound clearance were similar (within 2ꢀfold) for the 6 and 7 position analogs with the chloro
substituted compounds (7 and 9) showing about 2ꢀfold lower unbound clearance than the bromo
substituted compounds (11 and 13). So while total exposure appears greater for the 6 position
analogs, unbound concentrations are comparable across the tetrahydroꢀnaphthyl compounds
tested. Protein binding data in mouse plasma was only available for 9, 11 and 35 and binding for
each of these compounds was similar in each of the three species tested. Mouse plasma protein
binding was therefore estimated for 7 and 13 using the average of the human and rat values,
suggesting that unbound concentrations for each of the compounds are likely to be within a
similar range (Table 6). The best of the 2ꢀindanyl analogs (35) had considerably higher unbound
plasma exposure in rats compared to the tetrahydroꢀ2ꢀnaphthyls and to 1, and had similarly low
unbound clearance.
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 in vitro data generated in liver microsomes provided a mechanism to rank
compounds based on their relative metabolic stability. The compounds selected for in vivo
testing were expected to have good exposure, and this was seen following dosing to both rats and
mice indicating that qualitatively, the in vitro data were informative. Converting the in vitro
intrinsic clearance determined in rat microsomes to a plasma clearance using physiologicallyꢀ
based scaling factors and taking into account binding to rat plasma and microsomal proteins (as
2
1
per ) and blood to plasma partitioning ratio led to an under prediction of the actual in vivo
clearance by a factor of 3ꢀ to 16ꢀfold for the five compounds tested. These trends are similar to
1
6
what had been observed previously for 1 and suggest that in vivo clearance mechanisms for the
series also involve additional nonꢀCYP pathways not fully represented by the microsomal test
ACS Paragon Plus Environment

Page 23 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
system. In the case of 1, hepatocyte assays also led to an under prediction of in vivo clearance
highlighting the difficulties in obtaining accurate clearance predictions based on in vitro data
alone. Further studies with the current compounds using hepatocytes may have provided
additional insight into their clearance mechanisms.
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 vivo efficacy was assessed in the P. falciparum SCID mouse model. Of the three tested
compounds, 9 (7 position) had the most potent in vivo activity. The total blood AUC required to
reach ED was lower than for 1, however it took a higher dose (~19 mg/kg versus 8.1 mg/kg for
9
0
QD dosing) to achieve this activity, possibly due to a higher unbound clearance of 9 compared to
as seen in rats. The binding trends in human and mouse plasma would suggest that unbound
1
concentrations of 9 in the SCID mouse are likely to be considerably higher compared to those for
1 at the same total blood concentration. Therefore, the true in vivo potency of 1 based on
unbound concentrations is likely to be greater than that for 9. In spite of the high blood exposure,
the 2ꢀindanyl (35) lacked sufficient intrinsic potency to provide good activity in the in vivo
model. In order to improve patient compliance, single dose treatments are being sought by the
4
international organizations that are promoting new drug discovery for malaria. . While all three
tested compounds (7, 9 and 35) showed in vivo antimalarial activity, it is unlikely that any of
these compounds have the necessary properties to support a single dose treatment regimen at a
practical dose level in humans.
Preliminary safety analysis included characterizing hERG channel activity to assess
potential cardiac effects and CYP inhibition studies to determine the potential for drugꢀdrug
interactions. All five of the profiled compounds showed some CYP inhibition, with CYP isoform
2D6 being inhibited to the greatest extent. CYP2D6 inhibition was modest with the most potent
inhibition observed in the 3 – 6 ꢁM range, but does suggest some potential for drugꢀdrug
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 68
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
interactions with compounds that are substrates for this enzyme. Tested compounds in these
series also showed some hERG inhibition (0.5 – 1.5 ꢁM range). This inhibitory activity is similar
to what was observed for 1, where subsequent studies using the rabbit wedge model and dog
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
12
cardiovascular studies showed no evidence for QT prolongation, arrhythmias or cardiac effects.
Additionally, for 1 the unbound plasma concentration at likely therapeutic doses is 70ꢀ320ꢀfold
lower than the IC for hERG channel inhibition, suggesting it would be very unlikely that 1
5
0
would inhibit the channel in vivo. Insufficient efficacy data is available for compounds 7 and 35
to determine the safety margin relative to unbound concentrations at an efficacious dose,
however for 9 the Cmax at the ED in the SCID study (unbound concentration of approximately
90
0.05 ꢁM) is ~30ꢀfold lower than the IC50 on the hERG channel, again suggesting a low
likelihood the channel would be inhibited in vivo. However given the hERG channel activity,
careful evaluation of cardiac safety would be an important component of preclinical development
if any of these compounds were to be advanced.
Several potential avenues remain feasible for additional optimization of the
triazolopyrimidine series. Firstly, only the chloro substituted tetrahydroꢀ2ꢀnaphthyls were tested
in the SCID efficacy model, leaving open the possibility that one of the bromo analogs might
have shown better in vivo efficacy. Additional optimization of the series might focus on reducing
lipophilicity as this could lead to reduced intrinsic clearance and better in vivo exposure and also
could lead to reduced hERG activity. Reduced lipophilicity would also likely result in improved
aqueous solubility thereby simplifying formulation approaches for oral administration.
Decreasing lipophilicity will likely require the SF ꢀanline of 1 to be replaced with more
5
hydrophilic aromatic amines, provided potency can be maintained. Within the tetrahydroꢀ2ꢀ
ACS Paragon Plus Environment

Page 25 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
naphthyls, it might yet be possible to identify more metabolically stable analogs by identifying
the specific sites of metabolism and attempting to block them.
Conclusion
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
We have described the identification of several tetrahydroꢀ2ꢀnaphthyl and 2ꢀindanyl
triazolopyrimidine analogs that are potent inhibitors of Plasmodium DHODH and which have
improved species selectivity over the mammalian enzymes in comparison to the clinical
candidate 1. The compounds have potent activity against P. falciparum in parasite assays and
showed antimalarial activity in the P. falciparum SCID mouse model of disease. These
compounds showed good plasma exposure after oral dosing in mice and rats and were well
tolerated in the in vivo studies that were performed. Several of the identified compounds, most
notably 9, have the potential to be further developed for treatment of malaria. However, the
findings that 9 is less intrinsically potent than 1 in vivo (based on expected unbound
concentrations) and that the unbound clearance in rat is 2ꢀfold higher than for 1, suggest that this
compound is unlikely to meet the strict criteria required for a single dose cure at a reasonable
dose.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 68
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
Experimental section.
DHODH purification.
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
Plasmodium and mammalian DHODHs were expressed in E. coli as Nꢀterminal
truncations (lacking the mitochondrial transmembrane domains) fused to His ꢀpurification tags
6
1
1, 17, 18, 25
as previously described.
Protein (PfDHODH∆384ꢀ413) for crystallization studies was
further truncated to remove a proteolytically sensitive loop (amino acids 384ꢀ421) that prevents
17, 26
crystallization as previously described.
Proteins were expressed in E. coli BL21 phageꢀ
+
2
resistant cells (Novagen) and they were purified by Ni ꢀagarose column chromatography (GE
Healthcare Life Sciences, HisTrap HP) and Gelꢀfiltration column chromatography (GE
11, 17, 18
Healthcare Life Sciences, HiLoad 16/600 Superdex 200 pg) as previously described.
Protein was concentrated to 10ꢀ30 mg/ml and stored at ꢀ80°C.
DHODH kinetic analysis.
Steadyꢀstate kinetic analysis was performed using a dyeꢀbased spectrophotometric
method in assay buffer (100 mM HEPES, pH 8.0, 150 mM NaCl, 10% glycerol, 0.05% Triton Xꢀ
1
1
100) as previously described.
All buffers were degassed prior to use. Enzyme and substrate
concentrations were: DHODH (E = 5 nM), substrates (0.2 mM Lꢀdihydroorotate, 0.02 mM CoQd
and 0.12 mM 2,5ꢀdichloroindophenol (DCIP)). Inhibitor stock solutions (100 mM) were made in
DMSO and protected from light in dark amber vials. Serial dilutions were than performed to
generate a 3ꢀfold dilution series of 100x stocks also in DMSO that were used in the final assay
(final inhibitor concentration range was 0.01 – 100 ꢁM). Data were collected in triplicate for
each concentration in the titration. Data were fitted to the log[I] vs response (three parameters)
equation to determine the concentration that gave 50% enzyme inhibition (IC ) except for
5
0
ACS Paragon Plus Environment

Page 27 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compounds where the IC50 >10 ꢁM which were instead fitted to the standard IC equation
50
(
Y=1/(1+X/(IC ))) in Graphpad Prism. Error represents the 95% confidence interval of the fit.
50
P. falciparum growth assays.
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
P. falciparum parasites were grown in RPMIꢀ1640 containing 0.5% AlbumaxꢀII as
1
1
previously described. Inhibitor stock solutions were prepared as described above for DHODH
kinetic analysis except that a 500x dilution series prepared in DMSO was used to generate a 10x
dilution series in RPMI so that final DMSO concentration in the media was 0.2%. Growth
assays for P. falciparum 3D7 cells and drugꢀresistant parasites were performed using the SYBRꢀ
1
8
green 72 h growth assay. Data were collected in triplicate to quadruplicate for each
concentration in the titration. Data were fitted to the log[I] vs response Variable slope (four
parameter) equation in graph pad prism to determine the effective concentration (EC ) that led
5
0
to 50% reduction in parasitemia. Error represents the 95% confidence interval of the fit.
In vitro metabolism and physicochemical assessment.
In vitro metabolism. Selected compounds were incubated at a concentration of 1 ꢁM with
mouse, rat or pooled human liver microsomes (BD Gentest, Woburn, MA or XenoTech LLC,
11
Lenexa, Kansas City) at a microsomal protein concentration of 0.4 mg/mL. Substrate depletion
was assessed by LCꢀMS. Measured in vitro intrinsic clearance values (µL/min/mg microsomal
protein) were converted to predicted in vivo intrinsic clearances (mL/min/kg) using published
2
1
physiologically based scaling factors.
Solubility. Aqueous solubility of compounds was assessed by nephelometry as described
27
previously. Compound stock solutions were prepared in DMSO, which was then spiked into
pH 6.5 phosphate buffer giving a final DMSO concentration of 1%. Solubility measurements in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 68
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
physiologicallyꢀrelevant media were conducted for selected compounds. Media included 0.1 N
HCl to represent a simulated gastric condition, fasted (FaSSIF, pH 6.5) and fed (FeSSIF, pH 5.0
2
8
or 5.8) state simulated intestinal fluids prepared as described previously. Solid material was
incubated in media for a period of 5ꢀ6 h at 37°C with periodic mixing and sample supernatant
concentrations determined by HPLC with UV detection following two separate centrifugations to
remove undissolved compound.
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
Plasma protein binding and partitioning.
Select compounds were analyzed for plasma protein binding (PPB) in human, rat, or
1
1
mouse plasma using an ultracentrifugation method at 37°C as previously described followed by
LCꢀMS analysis as described below. Protein binding was calculated by comparing the unbound
concentration in the ultracentrifuged samples to the total plasma concentration in control samples
incubated at 37°C for the same time period. For compounds progressing to rat pharmacokinetic
studies, blood to plasma partitioning ratios in rat blood were also assessed as described
1
6
previously .
Pharmacokinetic analysis in mice and rats
Pharmacokinetic studies for select compounds were performed in mice and rats in
accordance with the Australian Code of Practice for the Care and Use of Animals for Scientific
Purposes and the study protocols were approved by the Monash Institute of Pharmaceutical
Sciences Animal Ethics Committee.
Oral doses were administered as a suspension in a vehicle containing 0.5% w/v
carboxymethylcellulose, 0.5% v/v benzyl alcohol (as a preservative) and 0.4% v/v Tween 80 and
intravenous doses were administered in coꢀsolvent vehicle containing 40% propylene glycol,
ACS Paragon Plus Environment

Page 29 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
0% ethanol, 0.4% Tween 80 in water. IV doses to rats were administered as a 10 min constant
rate infusion of 1 mL into an indwelling jugular vein cannula, and oral doses were administered
by gavage as 10 mL/kg (both rats and mice). Blood was collected from rats via a cannula
inserted into the carotid artery on the day prior to dosing, and from mice via either
submandibular bleed or terminal cardiac puncture (maximum of two samples per mouse) into
tubes containing heparin as an anticoagulant and plasma was separated by centrifugation.
Compound concentrations were quantitated by LCꢀMS using a Waters Xevo TQ mass
spectrometer coupled to a Waters Acquity UPLC. Chromatography was conducted using a
Supelco Ascentis Express RP Amide column (50 x 2.1mm, 2.7 µm) with a mobile phase
consisting of 0.05% formic acid in water and 0.05% formic acid in methanol mixed via a linear
gradient with a cycle time of 4 min. The flow rate was 0.4 mL/min and the injection volume was
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
µL. Detection was via positive electrospray ionisation mass spectrometry with multipleꢀ
reaction monitoring using a cone voltage of 30ꢀ40 V and collision energy of 25ꢀ35 V. Calibration
curves were prepared using blank rat or mouse plasma and validation studies confirmed the
accuracy, precision and limit of quantitation to be within acceptable ranges.
Cyp inhibition assays
Inhibition of cytochrome P450 (CYP) enzymes was assessed in human liver microsomes
using a substrate specific approach where the formation of metabolites specific to a particular
CYP isoform was monitored. Microsomes were suspended in phosphate buffer and incubated at
2
9
37ºC in the presence of probe substrates and reactions were initiated by the addition of an
NADPHꢀregenerating system. Reactions were quenched at appropriate times using acetonitrile.
Samples were centrifuged and concentrations of the metabolites assessed by LCꢀMS. The IC50
value for each compound or control inhibitor was the concentration at which there was a 50%
ACS Paragon Plus Environment