Hydroxamic Acid Inhibitors Provide Cross-Species Inhibition of Plasmodium M1 and M17 Aminopeptidases

Article abstract of DOI:10.1021/acs.jmedchem.8b01310

There is an urgent clinical need for antimalarial compounds that target malaria caused by both Plasmodium falciparum and Plasmodium vivax. The M1 and M17 metalloexopeptidases play key roles in Plasmodium hemoglobin digestion and are validated drug targets. We used a multitarget strategy to rationally design inhibitors capable of potent inhibition of the M1 and M17 aminopeptidases from both P. falciparum (Pf-M1 and Pf-M17) and P. vivax (Pv-M1 and Pv-M17). The novel chemical series contains a hydroxamic acid zinc binding group to coordinate catalytic zinc ion/s, and a variety of hydrophobic groups to probe the S1′ pockets of the four target enzymes. Structural characterization by cocrystallization showed that selected compounds utilize new and unexpected binding modes; most notably, compounds substituted with bulky hydrophobic substituents displace the Pf-M17 catalytic zinc ion. Excitingly, key compounds of the series potently inhibit all four molecular targets and show antimalarial activity comparable to current clinical candidates.

Full text of DOI:10.1021/acs.jmedchem.8b01310

Subscriber access provided by YORK UNIV
Article
Hydroxamic acid inhibitors provide cross-species
inhibition of Plasmodium M1 and M17 aminopeptidases.
Natalie B Vinh, Nyssa Drinkwater, Tess R Malcolm, Michael Kassiou, Leonardo
Lucantoni, Peter M Grin, Georgina S Butler, Sandra Duffy, Christopher
M Overall, Vicky M. Avery, Peter J. Scammells, and Sheena McGowan
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01310 • Publication Date (Web): 11 Dec 2018
Downloaded from http://pubs.acs.org on December 14, 2018
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 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 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.
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 67
Journal of Medicinal Chemistry
1
2
3
4
Hydroxamic acid inhibitors provide cross-species
5
6
7
8
9
inhibition of Plasmodium M1 and M17 aminopeptidases
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
Natalie B. Vinh a,§, Nyssa Drinkwater b,§, Tess R Malcolm b, Michael Kassiou,c
Leonardo Lucantoni d, Peter M Grin e, Georgina S Butler e,f, Sandra Duffy d,
Christopher M Overall e,f, Vicky M. Avery d, Peter J. Scammells a,*, Sheena
McGowan b,*
a
Medicinal Chemistry Monash Institute of Pharmaceutical Sciences, Monash University
Parkville, Melbourne, VIC 3052, Australia.
b
Biomedicine Discovery Institute, Department of Microbiology, Monash University, Clayton,
Melbourne, VIC 3800, Australia.
c
School of Chemistry, University of Sydney, NSW 2006, Australia
d
Discovery Biology, Griffith Institute for Drug Discovery, Griffith University, Nathan,
Queensland 4111, Australia.
e
Centre for Blood Research, Department of Biochemistry & Molecular Biology, University of
British Columbia, Vancouver, British Columbia V6T 1Z3, Canada
f
48
49
50
51
52
53
54
55
56
57
58
59
60
Centre for Blood Research, Department of Oral Biological and Medical Sciences, University of
British Columbia, Vancouver, British Columbia V6T 1Z3, Canada
KEYWORDS: falciparum malaria, vivax malaria, aminopeptidase, inhibitors, hydroxamic acid,
metallo-proteins
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 67
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
2
ACS Paragon Plus Environment

Page 3 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT: There is an urgent clinical need for antimalarial compounds that target malaria
caused by both Plasmodium falciparum and Plasmodium vivax. The M1 and M17
metalloexopeptidases play key roles in Plasmodium haemoglobin digestion, and are validated drug
targets. We used a multi-target strategy to rationally design inhibitors capable of potent inhibition of
the M1 and M17 aminopeptidases from both P. falciparum (Pf-M1 and Pf-M17) and P. vivax (Pv-
M1 and Pv-M17). The novel chemical series contains a hydroxamic acid zinc binding group to
coordinate catalytic zinc ion/s, and a variety of hydrophobic groups to probe the S1’ pockets of the
four target enzymes. Structural characterisation by co-crystallisation showed that selected
compounds utilise new and unexpected binding modes; most notably, compounds substituted with
bulky hydrophobic substituents displace the Pf-M17 catalytic zinc ion. Excitingly, key compounds
of the series potently inhibit all four molecular targets and show antimalarial activity comparable to
current clinical candidates.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 67
1
2
3
 INTRODUCTION
4
5
6
7
8
9
Malaria is a widespread parasitic disease; half of the world’s population, equating to
approximately 3.2 billion people, is at risk of contracting malaria. In 2016, an estimated 216 million
people contracted malaria and 445,000 deaths were reported.1 Malaria is caused by five species of
the Plasmodium genus: Plasmodium falciparum, Plasmodium malariae, Plasmodium ovale,
Plasmodium vivax and Plasmodium knowlesi. P. falciparum is responsible for an estimated 90% of
malaria deaths worldwide, making it the most severe and lethal form of the disease1; however, P.
vivax, which is the predominant parasite causing human malaria outside of Africa, also causes
substantial morbidity, primarily due to the difficulty obtaining a complete cure, which results in
disease recurrence.2 Efforts to control malaria and prevent its spread to new regions have been
hindered by the emergence of drug resistance, including to artemisinin, our last line of defense.3, 4
Despite the global health toll of malaria, the widespread resistance to commonly used antimalarials,
and the alarming development of artemisinin resistance, there are few novel therapeutics in the
pipeline. Further, many of the treatments in the Medicine for Malaria Venture (MMV) pipeline, the
major source of new antimalarial therapeutics, still contain artemisinin or its derivatives. Therefore,
new therapeutics with novel mechanisms of action and activity against both P. falciparum and P.
vivax are urgently required.
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
Hemoglobin digestion is an essential metabolic process in the intra-erythrocytic stage of the
Plasmodium life cycle. During this process, malaria parasites digest up to 75% of host cell
hemoglobin resulting in the liberation of amino acids required for parasite protein synthesis.5, 6 Two
zinc-dependent metalloaminopeptidases, Pf-M1 and Pf-M17, are involved in the latter stages of
hemoglobin degradation (Supplementary Figure 1), and are essential for parasite survival.7-9 In vitro
studies have shown that Pf-M1 inhibition kills the parasites by preventing proteolysis of
hemoglobin peptides, whilst Pf-M17 inhibition kills the parasites earlier in the life cycle, suggestive
of an additional role for Pf-M17 beyond hemoglobin digestion.7 Further, inhibition of both
aminopeptidases in vivo reduced parasitemia in Plasmodium chabaudi chabaudi murine models of
4
ACS Paragon Plus Environment

Page 5 of 67
Journal of Medicinal Chemistry
1
2
3
malaria.10, 11 As such, Pf-M1 and Pf-M17 are validated targets for antimalarial therapy, and have
4
5
been targeted in a variety of drug discovery programs.10-22
6
7
8
9
Pf-M1 and Pf-M17 both use a zinc (Zn2+)-dependent mechanism to hydrolyze the scissile
peptide bond to catalyse the removal of single amino acids from the N-terminus of short peptides.
Despite performing near-identical reactions, albeit with distinct specificities for the amino acid
residue they remove (P1 position),23 the enzymes themselves are from different enzyme families
and possess very different three-dimensional structures.24, 25 Pf-M1 is a monomeric protein of four
domains, and has an active site containing a single catalytic Zn2+ ion buried deep within the
catalytic domain.25 Pf-M17 is a homo-hexameric protein, with each subunit comprised of two
domains. The six active sites of the Pf-M17 hexamer each contain two catalytic Zn2+ ions, and lie
on the edge of catalytic domains where they are exposed to the large internal cavity of the
hexamer.24 Despite these differences, the active sites of Pf-M1 and Pf-M17 share a similar
structural arrangement; both possess S1 and S1’ pockets (to accommodate substrate P1 and P1’
residues), with bound Zn2+ ion/s at the junction of the two sub-sites.24, 25 These common features
led us to propose that, through rational drug design, we could develop a single compound capable
of potent, dual inhibition of both Pf-M1 and Pf-M17, an attractive strategy for slowing the
development of drug resistant parasites.17, 20
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
Research within our group has used rational drug discovery to optimise arginine mimetics
17, 20
into potent dual Pf-M1 and Pf-M17 inhibitors (Supplementary Figure 2).15,
N-(2-
(Hydroxyamino)-2-oxo-1-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)ethyl)pivalamide (1) coordinates
the catalytic Zn2+ ion/s through a hydroxamic acid moiety, and occupies the S1 and S1’ pockets of
both enzymes with the trifluoro-biphenyl and pivalamide moieties, respectively (Supplementary
Figure 3).15 In the S1 pocket of Pf-M1, the fluorine substituents interact through an intricate
network of water-mediated hydrogen bonds, whilst in Pf-M17, the trifluorophenyl ring binds solely
through hydrophobic interactions with neighboring Leu492, Phe583 and Met392 residues. Having
optimized the binding interactions within the S1 pockets of both enzymes, we sought to improve
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 67
1
2
3
4
5
6
7
the pivalamide portion of the molecule, which binds the S1’ pocket. Further, we elected to extend
our strategy to target the M1 and M17 homologs in P. vivax (Pv-M1 and Pv-M17) to expand the
capacity of our lead compounds and develop cross-species therapeutics.
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
 RESULTS
Multi-target inhibition strategy. To achieve effective cross-species activity against P.
falciparum and P. vivax by dual inhibition of M1 and M17, compounds capable of potent inhibition
of all four molecular targets are required. The challenge this multi-target strategy presents to
rational drug design is lessened by the high degree of sequence conservation between the
Plasmodium homologs. The M1 homologs of P. falciparum and P. vivax share 72% sequence
identity (89% within catalytic domains), while the M17 homologs share 70% identity (92%
catalytic domains). Furthermore, the inhibitor binding residues are highly conserved; based on the
previously determined crystal structures of Pf-M1 and Pf-M17 in complex with 1, the residues
lining the active sites and interacting with the inhibitors are 100% conserved between the cross-
species homologs. Given this sequence conservation, and the highly conserved reaction
mechanisms utilized by the M1 and M17 aminopeptidase families, it is likely that the structure of
the active sites is also highly conserved across species. To investigate if these similarities would
allow for cross-species inhibition by our inhibitors, we produced recombinant Pv-M1 and Pv-M17.
The Pv-M1 (residues 195–1097) and Pv-M17 (residues 203–621) protein constructs were designed
and produced in line with methods used for the Pf homologs, and demonstrated comparable kinetic
parameters against the reporter substrate, L-leucine-7-amido-4-methylcoumarin, in our
fluorescence-based assay system (complete biochemical characterization of Pv-M1 and Pv-M17 to
be reported elsewhere). Interestingly, we found that 1, designed to target the Pf enzymes, is actually
more potent against the Pv enzymes (Table 1). This result supports the assertion that the structures
of each of the M1 and M17 active sites are highly conserved between Pf and Pv. We therefore
elected to perform further rational design of inhibitors targeting the M1 and M17 aminopeptidases
6
ACS Paragon Plus Environment

Page 7 of 67
Journal of Medicinal Chemistry
1
2
3
from both P. falciparum and P. vivax (four total targets), guided by structural analysis of the Pf
4
5
enzymes.
6
7
8
9
Chemical synthesis of S1’ exploratory series. Changes to the amide functionality of 1 were
made to optimize interactions in the S1’ pockets of Pf-M1/Pf-M17 thereby enhancing binding and
inhibitory potency. A series of hydroxamic acid analogues containing hydrophobic groups that vary
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 size and shape were explored to investigate which groups could best satisfy the S1’ pockets of
the four target enzymes. Different hydrophobic groups were incorporated into the scaffold using a
four-step reaction sequence shown in Scheme 1. Briefly, 2-amino-2-(4-bromophenyl)acetic acid (2)
was esterified using standard conditions. The resultant methyl ester 3 was further functionalized
using a Suzuki reaction to append the 3,4,5-trifluorophenyl ring. A mixture of compound 3, 3,4,5-
trifluorophenylboronic acid, PdCl2(PPh3)2, Na2CO3 (aq) and THF was refluxed for 2 hours to give
the key amine intermediate 4 in good yield (65%). The amine 4 was then elaborated with the
desired hydrophobic substituent by treatment with either an acid chloride or an acid in the
presence of peptide coupling reagents. These reactions proceeded in moderate yields (28–85%)
with the exception of the 2,2,3,3,3-pentafluoropropanamide analogue 5e, which was obtained
in poor yield (7%) when HCTU was used as the coupling reagent. A small improvement in yield
(42%) was observed when EDC.HCl was used. Finally, aminolysis of the methyl esters 5a-5r
with hydroxylamine hydrochloride and potassium hydroxide in anhydrous methanol gave the
hydroxamic acid analogues (6a-6r) in low to good yields (22–75%).
Compound series inhibits multiple targets. Previous examination of the activity of Pf-M125
and Pf-M1726 utilized a fluorescence-based assay system to monitor cleavage of the commercially
available reporter substrate L-leucine-7-amido-4-methylcoumarin. We have previously adapted this
system to examine inhibition of both Pf-M1 and Pf-M17 by calculation of compound inhibition
constants (Ki) using the Dixon method. During the examination of the current series of compounds,
we determined that compounds were inhibiting the enzymes with inhibition constants below the
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 67
1
2
3
4
5
6
7
8
9
concentration of enzyme in the assay. We therefore re-structured our method of compound
examination to use a modified Morrison equation for tight-binding inhibitors and performed all
assays in biological triplicate.27, 28 This alteration in methodology resulted in a small discrepancy
between the Ki values of compound 1 reported here (Ki Pf-M1 = 331 nM, Ki Pf-M17 = 147 nM) and
that reported previously (Ki Pf-M1 = 80 nM, Ki Pf-M17 = 60 nM).15 The methods utilised herein
provide a robust method to evaluate tight-binding inhibitors against four different molecular targets,
and we recommend they be used to assess all inhibitors of the Plasmodium aminopeptidases in
future studies. We determined the inhibitory activity of 6a-r against the M1 and M17 enzymes from
both P. falciparum and P. vivax. The binding affinities (Ki) of the synthesized compounds to all
four enzymes are described in Table 1. The rational design of compound/s capable of potent
inhibition of four different enzymes presents a number of logistical difficulties. To simplify, we
elected to analyze trends in the binding potencies first between homologous enzyme pairs (Pf- vs
Pv-M1 and Pf- vs Pv-M17), and then between enzyme families (M1 vs M17).
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
Overall, the compound series inhibits Pv-M1 with approximately 10-fold greater affinity than
Pf-M1, demonstrating that key differences between the M1 homologs exist that have a profound
influence on compound binding and inhibition (Figure 1A, Table 1). Despite the difference in
affinities, the trends in potency within the series are largely conserved, that is, the most potent
inhibitors of Pv-M1 are also the most potent Pf-M1 inhibitors (Figure 1A). Analogs of 1 that lack
one or two methyl groups of the pivalamide (compounds 6a and 6b, respectively) displayed
significantly lower M1 binding affinity. A small series of fluorine-containing tert-butyl isosteres
(compounds 6c-e) were also evaluated. Of these compounds, only one diastereomer of the 2,3,3,3-
tetrafluoropropanamide 6d (6d-a) inhibited both the Pf-M1 and Pv-M1 enzymes with comparable
potency to 1. A series of cycloalkylcarboxamides was also prepared and assessed (compounds 6f-k)
to further define the steric requirements for optimal Pf-M1 and Pv-M1 activity. The cyclobutyl- and
cyclopentylcarboxamides (6g and 6h, respectively) proved to be the most potent inhibitors of both
enzymes, while analogs of smaller ring size (cyclopropyl; 6f) or larger ring size (cyclohexyl; 6i)
8
ACS Paragon Plus Environment

Page 9 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
were significantly less potent inhibitors of both Pf-M1 and Pv-M1. Interestingly, the adamantyl
carboxamide 6k inhibited Pf-M1 with higher potency (Pf-M1 Ki = 137 nM) than 1, but was a much
weaker inhibitor (Ki = 122 nM) than 1 at Pv-M1.
8
9
Extension of the carboxamide of 1 by one methylene unit in the 3,3-dimethylbutanamide
analog 6l improved potency at both enzymes (cf. Pf-M1 = 269 nM, Pv-M1 = 6.39 nM). Based on
this promising finding, a series of cycloalkyl acetamides which possessed an analogous methylene
spacer between the amide carbonyl and the cycloalkyl moiety (compounds 6m-p) were prepared
and evaluated. This series was generally more potent than the corresponding cycloalkyl
carboxamides which lacked the methylene spacer (i.e. compounds 6f-i), with the exception of the
cyclohexyl analogs 6i and 6p, which were equipotent. In terms of compounds 6m-p, the
cyclobutylacetamide 6n was the most potent inhibitor of Pf-M1 (Ki = 177 nM), with both the
cyclopropyl and cyclopentylacetamides exhibiting slightly lower potency (6m; Ki = 202 nM and 6o;
Ki = 216 nM, respectively). The same general trend was observed at the Pv-M1 enzyme, where the
cyclobutylacetamide 6n was once again proved to be the most potent (Pv-M1 Ki = 2.85 nM), with
reduced potency observed for the cyclopropyl and cyclopentylacetamides (6m; Ki = 17.9 nM and
6o; Ki = 7.58 nM, respectively). In order to further probe the steric requirements of the S1’-binding
pocket the fused cycloalkyl groups 6q and 6r were assessed. Both the cubanylacetamide 6q and the
norbornylacetamide 6r exhibited potent inhibitory activity at both Pf and Pv-M1 enzymes, with 6q
proving to be the most potent M1 inhibitor of all compounds studied (Ki Pf-M1 = 88.9 nM, Pv-M1
= 1.73 nM).
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
Determination of the binding potencies of 6a-r for Pf- and Pv-M17 shows that, unlike the M1
enzymes, the scale of inhibition of 6a-r is largely conserved between the two M17 homologs
(Figure 1B and Table 1). Analysis of the trends in inhibition potency across the series highlights
some interesting similarities and differences between M17 homologs. The changes to binding
potency of the alkyl, fluoroalkyl and cycloalkyl carboxamides 6a-k are conserved between the two
M17 enzymes (Figure 1B and Table 1). Of analogs 6a-k, only one produced stronger inhibition of
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 67
1
2
3
4
5
6
7
Pf- and Pv-M17 compared to 1. Compound 6k, containing a bulky hydrophobic adamantyl moiety,
showed the most potent inhibition of Pf-M17 observed to date (Ki = 28.9 nM) as well as highly
potent Pv-M17 inhibition (Ki = 18.6 nM).
8
9
Incorporation of a methylene spacer adjacent to the carboxamide resulted in substantial
divergence between the Pf and Pv homologs (6l-6r, Figure 1B, Table 1). There is clearly a
difference in the capacity of the S1’ pockets of Pf- and Pv-M17 to accommodate hydrophobic
moieties. In general, the derivatives 6l-r showed stronger inhibition of the Pv-M17 homolog versus
the Pf-M17 homolog with exception of compounds 6m and 6o. Binding affinity of compounds 6l-p
to Pf- and Pv-M17 largely improved relative to their analogous compound without the methylene
unit (1, 6f-i). The 3,3-dimethylbutanamide 6l proved to be the most potent inhibitor at both Pf- and
Pv-M17 with Ki values of 101 and 3.60 nM, respectively. Interestingly there was no clear
correlation between the size of the carboxamide and inhibitory activity across this series. Of the
fused cycloalkylacetamides 6q and 6r, norbonylacetamide 6r was equivalent in inhibitory potency
to both M17 enzymes (Ki Pf-M17 = 127 nM, Pv-M17 = 93.1 nM) compared to 1.
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
Overall, both the M1 and M17 enzymes from Pf and Pv were able to accommodate a number
of different hydrophobic groups. While the binding capacity of the series showed substantial
variation between the four enzymes, the cycloalkyl acetamides (compounds 6m-p) were generally
the most potent multi-target inhibitors of the series, with all but 6p, demonstrating potent inhibition
of all four enzymes. These results indicate that it is possible to develop a single compound capable
of potent inhibition of both the M1 and M17 aminopeptidases across multiple Plasmodium species.
Binding mode of selected lead compounds to Pf-M1. To support the SAR data, and to
inform further rational design, we determined the crystal structures of Pf-M1 with selected
compounds of the series (Supplementary Table 1 and 2). The eight structures all showed clear
compound density for the compounds in the Pf-M1 active site, with all of the compounds
coordinating the catalytic Zn2+ through the hydroxamic acid. Further, the conserved trifluoro-
biphenyl substituents all adopt identical poses in the S1 site, as was observed on binding of
10
ACS Paragon Plus Environment

Page 11 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound 1 previously.15 The overall quaternary structure of Pf-M1 is also largely conserved.
When compared to the unliganded Pf-M1 structure, a small shift (0.5 Å) in the backbone of residues
319–Thr321, which sits just behind the S1 site, is observed, with Glu319–Ala320 adopts the cis-
configuration in all eight structures described herein.
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 lack of distinct inhibition trend across the Pf-M1 series was curious, in that the most
potent compounds were substituted with mid-size (tert-butyl to cyclopentyl) or large (e.g. cubanyl)
hydrophobic moieties. We were particularly interested to determine why all three of the compounds
substituted with six-membered rings (6i, 6j, and 6p) were relatively weak inhibitors compared to
the closely related five-membered ring analogs (6h and 6o). The parent compound 1 coordinates the
catalytic Zn2+ through the hydroxamic acid, which orients the carboxamide to allow the
hydrophobic tert-butyl moiety access to the S1’ site wherein it forms interactions with the pi
electrons of His496, and hydrophobic interactions with Tyr580 and Val493 (Supplementary figure
3A).15 The carboxamide itself also forms key hydrogen bonds with the backbone amine of Gly460
and Ala461.15 The crystal structure of Pf-M1 in complex with 6h (1.35 Å) shows that the
cyclopentyl moiety of 6h largely occupies the same space as the tert-butyl moiety of 1, but forms
closer interactions with Val493 and the pi cloud of His496 (Figure 2A), which accounts for the
modest improvement in inhibitor potency. These interactions are optimized further in the
cyclopentylacetamide 6o, which combines the cyclopentyl moiety with a methylene spacer. Pf-
M1:6o (determined to 1.5 Å) demonstrates that this extension allows the cyclopentane ring deeper
access into the S1’ pocket where it forms hydrophobic interactions of an ideal length with Val493
and Thr492, while also maintaining the interactions with the zinc-binding residue His496 (Figure
2B). The improvement in binding of 6o compared to 6h is only modest (Ki = 216 nM compared to
285 nM), which is likely due to the closer unfavorable contacts with Arg489 formed by 6o. Any
additional steric bulk in this position would be more unfavorable, which likely indicates why the
cyclohexane-substituted analogues are poor inhibitors. To confirm, we determined the structure of
Pf-M1 in complex with cyclohexylcarboxamide 6i and cyclohexylacetamide 6p to 1.65 Å and 1.50
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 67
1
2
3
4
5
6
7
8
9
Å respectively. The cyclohexyl group of 6i sits within the S1’ pocket, and as predicted, clashes with
Arg489. Additionally, the adopted conformation is high energy; substituent is axial and the ring
adopts a strained chair conformation, with C1–C2–C3 angle of 98°. Together, these observations
account for the weak binding of 6i. Surprisingly, the added methylene spacer of 6p results in a
drastically altered Pf-M1 binding conformation. The cyclohexyl substituent has moved out of the
S1’ pocket, and instead occupies the substrate/product access channel (Figure 2C). This
rearrangement likely occurred to avoid steric clashes of the bulker substituent with the spatially
limited S1’ pocket. Unfortunately, the new binding mode places the cyclohexyl substituent within a
solvated environment, and offers few binding options for a hydrophobic moiety, and as a
consequence, reduces binding potency. The ring interacts with the face of Tyr580 (Figure 2D), but
this proximity also results in close contact between the ring of 6p and the backbone carbonyl of
Tyr575, which results in local small-scale rearrangement of the protein chain similar to that
observed on binding of other large inhibitors.29 Further, the change in carboxamide geometry
required to elicit the new binding mode resulted in a lengthening of the hydrogen bond between
carbonyl oxygen and backbone amine of Gly460 and loss of the hydrogen bond to Ala461 (Figure
2B compared to 2D). We also examined the binding of the cyclopropylacetamide 6m. Interestingly,
the structure of Pf-M1:6m (1.58 Å) showed that the cyclopropane ring of 6m also occupies the
access channel rather than the S1’ site similar to 6p. The contrasting conformational preference of
6m compared to 6o indicates both substituent geometry as well as size influences the Pf-M1
binding mode. Despite occupying the access channel rather than the S1’ subsite, 6m does not suffer
the same loss of potency as 6p. This is likely because the smaller 6m does not cause the Tyr575
backbone re-arrangement observed on binding 6p, nor require the change in geometry of the N-acyl
group. Together, these observations account for the loss of potency observed for 6p compared to 6l-
6n.
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
Having accounted for the loss of potency of compounds substituted with a 6-membered ring,
we questioned how the larger 6k, 6q and 6r regained affinity to possess the tightest Pf-M1
12
ACS Paragon Plus Environment

Page 13 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
inhibition constants of the series. We therefore determined the structure of Pf-M1 in complex with
6k. Despite high resolution (1.82 Å), the electron density for the adamantyl moiety was highly
disordered. Whilst we were able to model the substituent at partial occupancy (50%) placed mid-
way between the S1’ subsite and the access channel, the structure provided limited insight into
additional interactions that may be gained by these large hydrophobic groups.
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
Finally, we sought to investigate the structural mechanism by which Pf-M1 discriminates
between the diastereomers, 6d-a and 6d-b. We therefore determined the crystal structure of 6d-a, in
complex with Pf-M1 (1.81 Å). Electron density clearly indicated that the inhibitor molecule in
complex with the enzyme has the R,R-configuration. This structure showed that the hydroxamic
acid zinc binding group and biphenyl system of 6d-a bind Pf-M1 comparably to the parent
compound 1. The interactions of the carboxamides are comparable, as are as the interactions with
the face of His496 formed by both pivalamide of 1 and the 2,3,3,3-tetrafluoropropanamide 6d-a.
Within the S1’ pocket, the dual natured 2,3,3,3-tetrafluoropropanamide is able to form interactions
with the face of His496 and Val493 (similarly to 1), as well as additional interactions with Arg489
(unfavourable to 1), which likely accounts for the tighter binding of 6d-a.
Binding mode of selected lead compounds to Pf-M17. Inhibition data determined that
insertion of a methylene spacer to bulky hydrophobic substituents was generally favorable for
binding Pf-M17. We therefore elected to determine the crystal structures of Pf-M17 in complex
with 6k, the most potent Pf-M17 inhibitor of the series. The first round of crystallization
experiments yielded unexpected results; 6k uses a novel zinc-binding mode to inhibit Pf-M17 via a
previously unobserved mechanism. M17 aminopeptidases possess a conserved binuclear metal
center containing ‘catalytic’ and ‘regulatory’ metal binding sites.30 Full activity requires both sites
to be occupied; however the regulatory metal ion is readily lost in crystal structures.24, 31 Untreated
Pf-M17 crystals have only the ‘catalytic’ Zn2+ bound (Figure 3A), and require treating with a Zn2+
soak solution to occupy both sites and achieve the ‘active’ form of the enzyme used for inhibitor
design (Figure 3B). Using these methods, we previously determined the parent compound 1 binds
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 67
1
2
3
4
5
6
7
8
to Pf-M17 by coordinating both Zn2+ ions, a conformation that is repeated in all Pf-M17 active sites
(two copies of the Pf-M17 hexamer in the asymmetric unit afford 12 total views of the active site)
(Figure 3C).15 However, use of identical co-crystal growth and Zn2+ treatment methods to determine
Pf-M17:6k yielded an unliganded structure, and increasing the concentration of 6k in the Zn2+ soak
solution yielded only very weak density for the compound. Given the potent inhibition of Pf-M17
by 6k, the difficulty obtaining a co-crystal structure was puzzling. We collected a final dataset to
2.30 Å from a co-crystal that was not treated with additional Zn2+; the structure showed clear
electron density for 6k and only a single Zn2+. Surprisingly, the single Zn2+ occupies the
‘regulatory’ site, indicating that the ‘catalytic’ Zn2+ site cannot be occupied whilst 6k is bound
(Figure 3D). Such an active site arrangement, wherein the ‘catalytic’ site is unoccupied, has not
been observed in any other structure of M17 aminopeptidases. To determine whether Zn2+
displacement was a consequence of the large hydrophobic substituent on 6k or characteristic of
other compounds in the series, we also determined the structure of Pf-M17 in complex with 6i
(2.10 Å), which with a cyclohexane substituted off the N-acyl is intermediate in size between 1 and
6k, and retains tight binding capability. With the exception of the compound identity, the crystal
was treated identically to that which yielded Pf-M17:1 (including a Zn2+ soak step). In the Pf-
M17:6i structure, clear electron density was observed for the compound, however the electron
density in the catalytic Zn2+ site was weak, and not indicative of a metal ion. Although it is possible
the Zn2+ is at low occupancy, it could not be modeled. It was therefore clear that 6i was also in
competition with the catalytic Zn2+.
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
Beyond the unexpected displacement of the catalytic Zn2+ ion, the interactions between Pf-
M17 and 6i/6k were largely as expected. The binding position of the hydroxamic acid has shifted
slightly compared to that of the parent compound 1 (Figure 4A). This is likely the result of the
bulky substituents occupying the S1’ pocket, and the cause of the Zn2+ displacement. In place of
metallo-bonds, the hydroxyl of 6i and 6k form hydrogen bonds with Glu461 and Lys374, and the
catalytic carbonate ion (Figure 4B). Despite the large variation in mechanism of Zn2+ coordination,
14
ACS Paragon Plus Environment

Page 15 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
the change has little effect on the compound binding position. Similarly to 1 (Supplementary Figure
3B), the trifluoro-substituted biphenyl system occupies the S1 pocket (Figure 4A), where it forms
the same hydrophobic interactions. The large S1’ subsite of Pf-M17 was able to accommodate the
bulky hydrophobic substituents, wherein they interact with Ala460, Ile547, and the aliphatic chain
of Ser554 (Figure 4B). Of particular note is that the bulky groups form unfavorably close contacts
with Asn457 (Figure 4B), which is spatially restricted and unable to move away from the contact.
We predict it is these contacts that cause the shift in binding position of the N-acyl linker and
hydroxamic acid, which results in displacement of the catalytic Zn2+. Alternatively, the large
hydrophobic substituents may be altering the environment of the pocket to destabilize the charged
Zn2+; however it is unknown why the catalytic Zn2+, rather than the regulatory Zn2+, which is
spatially closer to the hydrophobic moiety and previously shown to be readily lost from the active
site, would be affected.
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
Binding of 6k is partially reversible. We performed washout experiments to determine
whether binding of 6k, which displaces the catalytic Zn2+, is reversible. We treated the enzyme with
either EDTA, 6k, or bestatin (competitive substrate-analogue inhibitor), washed the protein with
metal-free buffer, and then monitored the time-dependent activity on addition of metal cofactor and
substrate (Supplementary Figure 4). A short buffer exchange (2 hours) was sufficient to recover the
activity of EDTA treated Pf-M17. However, longer wash steps were required to completely remove
bestatin (normal activity of bestatin-treated Pf-M17 was observed after a 5 hour wash). In contrast,
after the 5-hour wash, only minimal activity was observed in the Pf-M17 sample treated with
compound 6k. Further washing recovered more of the activity, but after 7 hours of buffer exchange,
only approximately 20% of activity was recovered in the 6k-treated Pf-M17 sample. Therefore,
while 6k-treatment appears to be reversible, it likely requires long wash times to recover activity.
Since EDTA-treated enzyme recovers quickly, the slow recovery of the 6k-treated sample is likely
a result of a slow compound off-rate rather than metal removal.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 67
1
2
3
4
5
6
7
8
Anti-parasitic activity of key compounds. We have previously shown that 1 potently
inhibits P. falciparum growth in vitro, including the multi-drug resistant strain Dd2.32 We triaged
the current compound series for assessment in an in vitro P. falciparum phenotypic assay by
9
selecting only those compounds that demonstrated potent dual Pf-M1 and Pf-M17 inhibitory
capability (Ki ≤ 200 nM against one Pf target, and ≤ 400 nM against the second). The compounds
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
were tested against both drug-resistant (Dd2) and drug-sensitive (3D7) strains of P. falciparum, as
well as against HEK293 cells for preliminary cytotoxicity assessment. For comparison, we also re-
examined the parent compound 1 alongside the current series as well as a number of reference
compounds (Table 2). In addition to 1, which demonstrated potent anti-parasitic activity (IC50 Pf-
3D7 = 83.1 ± 16.1 nM, Pf-Dd2 = 81.7 ± 6.8 nM), three compounds from the series demonstrated
potent inhibition of Pf-3D7 and Pf-Dd2 proliferation. The most active compounds included 6l, 6n,
and 6q which all inhibited parasitemia with IC50 values less than 100 nM, while other compounds
examined demonstrated IC50 values between ~100-350 nM. Compound 6l exhibited the most
substantial improvement in anti-parasitic activity compared to the parent compound, showing IC50
values of 14.6 ± 0.8 nM against Pf-3D7 and 13.8 ± 0.5 nM against Pf-Dd2. Under our assay
conditions, this activity is comparable to clinically used antimalarial drugs pyronaridine and
chloroquine (IC50 Pf-3D7 = 7.0 ± 6.0 nM and 11.7 ± 3.5 nM, respectively), and approaching the
range of the gold standard clinical treatments artesunate and dihydroartemisinin (IC50 Pf-3D7 = 1.6
± 0.5 nM and 0.5 ± 0.1 nM respectively). Further, the compound series showed limited cytotoxicity
against HEK293 with little inhibition of cellular proliferation at 10 µM. At 40 µM, we did observe
inhibitory effects against HEK293, however, we were unable to calculate IC50 values as all
compounds failed to reach a full inhibition plateau. Approximate selectivity indices were calculated
for compounds that showed > 51 % inhibition at 40 µM and generally suggested excellent
selectivity range (Table 2). The compound series therefore demonstrates substantial improvement in
anti-malarial activity against cultured P. falciparum parasites and represents an exciting progression
16
ACS Paragon Plus Environment

Page 17 of 67
1
Journal of Medicinal Chemistry
2
3
of our anti-malarial compounds targeted against both the M1 and M17 aminopeptidases of P.
4
5
falciparum.
6
7
8
9
Compound 6l does not inhibit MMPs. To investigate the potential for off-target effects of
the compound series, we examined the ability of 6l, the compound with the most potent cellular
antiparasitic activity, to inhibit a panel of matrix metalloproteinases (MMPs). We assessed the
activity of MMP2, 7, 8, 9, or 13 following inhibition with increasing concentrations of 6l or control
compounds (Marimastat or Tosedostat, Supplementary Table 4). As expected, the gelatinases
(MMP2 and MMP9) were potently inhibited by Marimastat (IC50 = 0.43 and 3.1 nM respectively,
Supplementary Table 4) and Tosedostat (IC50 = 0.19 and 1.5 μM respectively), whereas 6l weakly
inhibited MMP2 (IC50 = 17 μM) and did not inhibit MMP9 (IC50 > 1mM). Similar observations
were made for collagenases (MMP8 and MMP13) as well as matrilysin (MMP7), which were all
inhibited tightly by Marimastat and Tosedostat, but weakly, or not at all, by 6l. Collectively, these
findings show that 6l shows almost no off-target inhibitory effects on MMPs. We additionally chose
to examine the potential of 6l to inhibit two human N-terminal aminopeptidases, aminopeptidase N
(APN or CD13) and insulin-regulated aminopeptidase (IRAP). While 6l showed only very weak
inhibition of IRAP (no inhibition observed at 6l concentrations less than 200 μM), it inhibits APN
with reasonable affinity (Ki = 0.3 ± μM).
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
 DISCUSSION
Antimalarial combination therapies have been the gold standard of malarial treatment since the
ability of the Plasmodium parasite to generate resistance to drug monotherapies was understood.
We have extended this strategy and used rational drug discovery to design compounds capable of
potent dual inhibition of the P. falciparum M1 and M17 aminopeptidases, both key players of the
essential hemoglobin digestion pathway. Based on the SAR of previously examined compound
series’, we generated novel hydroxamic-acid based compounds substituted with a range of
hydrophobic moieties to probe the S1’ subsites of Pf-M1 and Pf-M17. Further, we proposed that the
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 67
1
2
3
4
5
6
7
8
9
high identity of the M1 and M17 aminopeptidases between Plasmodium species would allow our
compounds to act similarly against the P. vivax homologs. Herein, we have examined the ability of
our compound series to inhibit the M1 and M17 aminopeptidases from both P. falciparum and P.
vivax.
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 compound series demonstrated substantial variability in the inhibition of Pf-M1 (nM to
M), which did not follow any clear and observable trend. By examination of the crystal structures
of Pf-M1 bound to key compounds of the series, we determined this variation was due to different
compound binding modes. Interestingly, the variation in inhibition potency was also observed for
Pv-M1. However, while the overall trends were conserved, we found that the compounds inhibited
Pv-M1 approximately 10-fold more potently than Pf-M1. This is despite high overall sequence
conservation, and 100% identity in inhibitor-binding residues (based on the Pf-M1 crystal
structures). The 10-fold change in inhibitor binding potencies are therefore likely to result from
enzyme differences beyond the active site arrangement. Such differences could include the
size/shape of the active site resulting from secondary/tertiary structure beyond the binding pocket.
Alternatively, the difference may result from mechanistic features beyond active site chemistry,
such as protein flexibility. Analysis of the M1 aminopeptidase superfamily has shown that
conformational dynamics plays a critical role in substrate recognition and catalysis.33, 34 It is
therefore possible that differences in the enzyme conformational dynamics exist that cannot be
predicted from the protein sequences, and that these dynamics play a role in inhibitor binding.
Similarly to M1 aminopeptidases, the inhibitor binding residues of the M17 homologs from P.
falciparum and P. vivax are also highly conserved. This includes the binuclear metal binding site,
which possess a highly conserved arrangement through the enzyme superfamily. The two metal
sites bind divalent metal ions, most commonly Zn2+, and have been characterized as ‘regulatory’
and ‘catalytic’ sites (site 1 and site 2, respectively). The ‘catalytic’ metal is coordinated by the side
chains of conserved Lys, Asp, and Glu residues (Lys374, Asp379, and Glu461 in Pf-M17) and the
‘regulatory’ metal is coordinated by the sidechains of a conserved Asp and Glu (Asp379 and
18
ACS Paragon Plus Environment

Page 19 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
Glu461 in Pf-M17) as well as the sidechain and mainchain oxygen of another Asp (Asp459 in Pf-
M17). Occupation of the catalytic site is absolutely required for catalytic activity.31 The observation
that 6k and 6i are competing with the catalytic Zn2+ is therefore completely unexpected, and has
substantial implications for drug discovery. By competing with both metal cofactor as well as
substrate, the compound mechanism of inhibition has evolved, and competitive models of binding,
that calculate inhibition constants on the basis of assay substrate concentration and affinity as well
as inhibitor and enzyme concentration, may no longer be appropriate. Our strategy to target both the
Pv-M17 as well as the Pf-M17 raises further questions about the ability of inhibitors to displace the
metal cofactors. The trend in inhibitor binding potencies of Pf- versus Pv-M17 diverged as the
compounds were substituted with bulky hydrophobic groups (6l–6r + 6k, Figure 1B and Table 1).
Differences in the ability of the compounds to displace the catalytic Zn2+ of the Pf- versus the Pv-
M17 could account for the variation in compound binding trends. It is possible that differences in
active site architecture and/or metal binding behavior between the Pf- and Pv-M17 enzymes exist
which are not predicted by the close sequence similarity. In the wake of the observations for Pf-
M17 reported here, the complete characterization of Pv-M17 metal binding behavior should be
prioritized.
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 some case, hydroxamate inhibitors developed as anticancer agents have suffered clinical
failure. Discussions regarding the liabilities of the chemical class largely revolve around the
potential for off-target activity against other zinc-dependent enzymes, which might lead to toxicity.
However, the previous generations of hydroxamates were safely tested in thousands of humans for a
variety of cancers.35-37 The only toxicity noted was some muscle tenderness in the upper body
girdle. This led to a very small number of patients limiting dosing, which resulted in ineffective
treatment for those patients.35-37 Otherwise the hydroxamate inhibitors (Marimastat and Batimastat)
were considered safe for humans, demonstrating only minor off target activity at clinically used
doses.35-37 Irrespectively, given the reputation of hydroxamate inhibitors, we assessed the potential
for the current series to cross-react. We examined the ability of 6l to inhibit a panel of MMPs
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 67
1
2
3
4
5
6
7
8
9
(MMP2, 7, 8, 9, and 13), and additionally, two human N-terminal aminopeptidases (IRAP and
APN). Compound 6l exhibit very weak to no activity against the MMPs and IRAP, demonstrating
low potential for off-target activity. However, inhibition of APN, the closest human homolog of the
Plasmodium M1-aminopeptidases, was observed. APN is an anti-cancer therapeutic target, and it
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
has clinically demonstrated that inhibition of APN activity does not result in any harmful side
38
effects.
Therefore, while designing selectivity over APN should be considered in future
optimization of this inhibitor series, we do not anticipate a high degree of selectivity over APN to
be required for safe use of the compounds in humans.
Physiological activity of a multi-target inhibitor is difficult to predict from enzyme inhibition
data, as the effect in live parasites may not be linearly correlated with inhibition of either/both
targets.39, 40 Throughout the development of the Pf-M1/Pf-M17 dual inhibitor compound series, we
have observed a broad correlation of improved cellular potency with increasing either Pf-M1 or Pf-
M17 inhibition potency.15, 17, 20 Within the current series, we observed a substantial (> 5-fold)
improvement in anti-plasmodial activity with compound 6l. While exciting, the improvement is
curious, in that it is does not directly reflect the Pf-M1 and Pf-M17 enzyme inhibition; 6l is not the
most potent dual inhibitor of the series, yet it clearly demonstrates the greatest potency of the series
in the whole-cell in vitro assay. Compound 6l is therefore a highly potent outlier. We propose three
potential reasons for this deviation: (1) increasing the hydrophobicity of the compounds has a large
effect on cellular penetration, (2) 6l targets additional cellular processes, or (3) the conditions of the
enzyme assays do not strictly reflect cellular conditions. Irrespectively, 6l is a potent inhibitor of the
four molecular targets examined herein, and demonstrates potent activity against cultured malaria
parasites; it therefore represents a promising lead for further development of potent antimalarial
therapeutics.
 CONCLUSIONS
The rapid spread of drug resistant Plasmodium parasites is a major threat to global health. We
have previously used a multi-target strategy to rationally design inhibitors capable of potent
20
ACS Paragon Plus Environment

Page 21 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
inhibition of the Pf-M1 and Pf-M17 aminopeptidases, and validated their use in culture against both
drug resistant and drug sensitive strains of the parasite. Herein we have expanded our strategy to
target the M1 and M17 aminopeptidases from P. vivax with the aim to develop compounds capable
of cross-species antimalarial activity. We designed a series of inhibitors that possess a novel
chemical scaffold and mapped their structure-activity relationships against the four molecular
targets in vitro. Excitingly, compounds possessing potent inhibitory activity against all four
enzymes also demonstrated potent activity against cultured parasites, and showed no observable
toxicity against human cells. This novel series of hydroxamic acid compounds therefore represents
an exciting lead in the design of novel antimalarial therapeutics. They operate via a novel, known
mechanism of action, and further, have the potential to be clinically effective against multiple
Plasmodium species.
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
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 67
1
2
3
 EXPERIMENTAL SECTION
4
5
6
7
8
9
Chemistry. Synthetic Materials and Methods. Chemicals and solvents were purchased from
standard suppliers and used without further purification. 1H NMR, 13C NMR and 19F NMR spectra
were recorded on a Bruker Avance Nanobay III 400MHz Ultrashield Plus spectrometer at 400.13
MHz, 100.61 MHz, and 376.50 MHz, respectively. Chemical shifts (δ) are recorded in parts per
million (ppm) with reference to the chemical shift of the deuterated solvent. Unless otherwise
stated, samples were dissolved in CDCl3. Coupling constants (J) and carbon-fluorine coupling
constants (JCF) are recorded in Hz and multiplicities are described by singlet (s), doublet (d), triplet
(t), quadruplet (q), broad (br), multiplet (m), doublet of doublets (dd), doublet of triplets (dt).
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
13
Overlapped non-equivalent C peaks were identified by HSQC and HMBC experiments and are
indicated with (2C) after the identified overlapped signal. LC-MS was performed using either
system A or B. System A: an Agilent 6100 Series Single Quad coupled to an Agilent 1200 Series
HPLC using a Phenomenex Luna C8 (2) 50 x 4.6 mm, 5 micron column. The following buffers
were used; buffer A: 0.1% formic acid in H2O; buffer B: 0.1% formic acid in MeCN. Samples were
run at a flow rate of 0.5 mL/min for 10 min: 0–4 min 5–100% buffer B in buffer A, 4–7 min 100%
buffer B, 7–9 min 100–5% buffer B in buffer A, 9–10 min 5% buffer B in buffer A. Mass spectra
were acquired in positive and negative ion mode with a scan range of 100–1000 m/z. UV detection
was carried out at 254 nm. System B: an Agilent 6120 Series Single Quad coupled to an Agilent
1260 Series HPLC using a Poroshell 120 EC-C18 50 x 3.0 mm, 2.7 micron column. The following
buffers were used; buffer A: 0.1% formic acid in H2O; buffer B: 0.1% formic acid in MeCN.
Samples were run at a flow rate of 0.5 mL/min for 5 min; 0–1 min 5% buffer B in buffer A, 1–2.5
min 5–100% buffer B in buffer A, 2.5–3.8 min 100% buffer B, 3.8–4 min 100–5% buffer B in
buffer A, 4–5 min 5% buffer B in buffer A. Mass spectra were acquired in positive and negative
ion mode with a scan range of 100–1000 m/z. UV detection was carried out at 214 and 254 nm.
Preparative HPLC was performed using an Agilent 1260 infinity coupled with a binary
preparative pump and Agilent 1260 FC-PS fraction collector, using Agilent OpenLAB CDS
22
ACS Paragon Plus Environment

Page 23 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
software (Rev C.01.04), and an Altima C8 22 x 250 mm 5 micron column. The following buffers
were used; buffer A: 0.1% TFA in H2O; buffer B: 0.1% TFA in MeCN, with sample being run at a
gradient of 5–100% buffer B in buffer A over 20 min, at a flow rate of 20 mL/min. All screening
compounds were of >95% purity unless specified in the individual monologue.
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
General Procedure A: Formation of amide bond using acid chlorides. The amine (1.0 eq)
was dissolved in CH2Cl2 (10 mL/mmol of amine). The flask was evacuated and flushed with
nitrogen. Et3N (2.2 eq) was added followed by the acid chloride (1.1 eq). The reaction mixture was
stirred at rt for 2 h. Water was added and the aqueous layer further extracted with CH2Cl2 (3 times).
The combined organic layers were dried over Na2SO4, filtered and concentrated under reduced
pressure. The crude product was purified by FCC.
General Procedure B: Formation of amide bond using HCTU. The acid (1.05 eq) and
HCTU (1.1 eq) were stirred in DMF (1 mL/mmol of amine). N,N-diisopropylethylamine (2.2 eq)
was added followed by a solution of the amine (1.0 eq) in CH2Cl2 (1 mL/mmol of amine). The
reaction mixture was stirred at room temperature for 2 h. Half saturated NaHCO3 (aq) was added
and the mixture extracted with Et2O (3 times). The combined organic layers were dried over
Na2SO4, filtered and concentrated under reduced pressure. The crude product was purified by FCC.
General Procedure C: Formation of amide bond using EDCI. A mixture of amine (1.0
eq), acid (1.2 eq), EDCI (1.2 eq) and DMAP (1.3 eq) in CH2Cl2 (10 mL/mmol of amine) was stirred
at rt under N2 overnight. The mixture was diluted with CH2Cl2, washed with 2M HCl (aq), saturated
NaHCO3 (aq), then brine. The organic layer was dried over Na2SO4, filtered and concentrated under
reduced pressure. The crude product was purified by FCC.
General Procedure D: Aminolysis of methyl ester to hydroxamic acid. The appropriate
methyl ester (1.0 eq) was dissolved in anhydrous MeOH (5 mL/mmol of ester) at rt. NH2OH.HCl
(4.0 eq) was added followed by KOH (5M in anhydrous MeOH, 5.0 eq). The mixture was stirred at
rt overnight and monitored by LC-MS analysis. The mixtures were directly dry-loaded on to Isolute
HM-N® (Biotage), before purification by FCC (eluent MeOH/ CH2Cl2 0:100 to 10:90).
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 67
1
2
3
4
5
6
7
8
9
Methyl 2-amino-2-(4-bromophenyl)acetate (3). To a mixture of 2-amino-2-(4-
bromophenyl)acetic acid (2.0 g, 8.7 mmol) in MeOH (87 mL) was added concentrated H2SO4 (0.8
mL, 15.7 mmol) dropwise. The reaction mixture was refluxed for 30 h. The solvent was
concentrated in vacuo and the resulting residue was basified with saturated NaHCO3 (50 mL), then
extracted with CH2Cl2 (3 x 50 mL). The combined organic layers were dried over Na2SO4, filtered
and concentrated under reduced pressure to give the title compound as a yellow oil (2.0 g, 94%). 1H
NMR δ 7.51–7.45 (m, 2H), 7.30–7.24 (m, 2H), 4.59 (s, 1H), 3.70 (s, 3H), 1.85 (s, 2H); 13C NMR δ
174.1, 139.3, 132.0, 128.7, 122.2, 58.3, 52.7; m/z MS (TOF ES+) C9H11BrNO2 [MH]+ calcd 244.0,
found 244.1; LC-MS tR = 3.2 min.
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
Methyl 2-amino-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (4). Compound 3 (2.0 g,
8.2 mmol) and 3,4,5-trifluorophenylboronic acid (2.2 g, 12.3 mmol) were dispersed in degassed
THF (30 mL) and degassed 1M Na2CO3 (aq) (10 mL). A steady stream of N2 was bubbled through
the mixture for 5 min, before adding PdCl2(PPh3)2 (0.58 g, 0.82 mmol). The mixture was refluxed
for 2 h, then the THF was concentrated in vacuo. The mixture was diluted with EtOAc (50 mL) and
water (50 mL), and the aqueous layer extracted further with EtOAc (2 x 50 mL). The combined
organic layers were dried over Na2SO4, filtered and concentrated. The crude product was purified
by FCC (MeOH/CH2Cl2 0:100 to 10:90) to give the title compound as a beige foam (1.6 g, 65%).
1H NMR (DMSO-d6) δ 9.04 (br. s, 2H), 7.85 (d, J = 8.1 Hz, 2H), 7.82–7.68 (m, 2H), 7.61 (d, J =
19
8.2 Hz, 2H), 5.40 (s, 1H), 3.74 (s, 3H); F NMR (DMSO-d6) δ -134.70 (d, J = 21.7 Hz), -162.73
13
(dd, J = 21.7/21.7 Hz); C NMR (DMSO-d6) δ 168.9, 150.7 (ddd, JCF = 246.6/9.8/4.1 Hz), 138.6
(dt, JCF = 250.4/15.8 Hz), 138.0, 135.9 (td, JCF = 8.1/4.5 Hz), 132.9, 128.9, 127.5, 111.8–111.2 (m),
55.0, 53.3; m/z MS (TOF ES+) C15H13F3NO2 [MH]+ calcd 296.1, found 296.1; LC-MS tR = 3.0 min.
Methyl 2-isobutyramido-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5a). Compound
4 (200 mg, 0.68 mmol) was converted to the title compound according to General Procedure A
1
using isobutyryl chloride (78 L, 0.75 mmol), to give 141 mg (57%) of white solid. H NMR δ
7.54–7.38 (m, 4H), 7.21–7.09 (m, 2H), 6.58 (d, J = 6.5 Hz, 1H), 5.61 (d, J = 6.9 Hz, 1H), 3.75 (s,
24
ACS Paragon Plus Environment

Page 25 of 67
Journal of Medicinal Chemistry
1
2
3
4
5
3H), 2.58–2.36 (m, 1H), 1.19 (d, J = 6.9 Hz, 3H), 1.16 (d, J = 6.9 Hz, 3H); 19F NMR δ -133.91 (d, J
= 20.6 Hz), -162.19 (dd, J = 20.7/20.7 Hz); 13C NMR δ 176.4, 171.4, 151.5 (ddd, JCF
=
6
7
8
9
249.7/10.1/4.2 Hz), 141.0–137.9 (m, 2C), 137.2, 136.7 (td, JCF = 7.7/4.5 Hz), 128.0, 127.5, 111.4–
110.8 (m), 55.9, 53.0, 35.3, 19.5, 19.4; m/z MS (TOF ES+) C19H19F3NO3 [MH]+ calcd 366.1, found
365.9; LC-MS tR = 3.5 min.
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
Methyl 2-propionamido-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5b). Compound
4 (200 mg, 0.68 mmol) was converted to the title compound according to General Procedure A
using propionyl chloride (65 L, 0.75 mmol), to give 205 mg (86%) of pale yellow solid. 1H NMR
δ 7.51–7.43 (m, 4H), 7.21–7.09 (m, 2H), 6.54 (d, J = 6.7 Hz, 1H), 5.63 (d, J = 7.0 Hz, 1H), 3.76 (s,
3H), 2.30 (qd, J = 7.6/3.2 Hz, 2H), 1.17 (t, J = 7.6 Hz, 3H); 19F NMR δ -133.88 (d, J = 20.4 Hz), -
162.14 (dd, J = 20.6/20.6 Hz); 13C NMR δ 173.2, 171.4, 151.5 (ddd, JCF = 249.7/10.1/4.3 Hz),
139.4 (dt, JCF = 34.1/15.3 Hz), 138.4, 137.2, 136.6 (td, JCF = 7.8/4.7 Hz), 128.1, 127.5, 111.5–110.8
(m), 56.0, 53.0, 29.4, 9.6; m/z MS (TOF ES+) C18H17F3NO3 [MH]+ calcd 352.1, found 351.9; LC-
MS tR = 3.4 min.
Methyl 2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)-2-(3,3,3-trifluoropropanamido)acetate
(5c). Compound 4 (200 mg, 0.68 mmol) was converted to the title compound according to General
Procedure B using trifluoromethylacetic acid (63 L, 0.71 mmol), to give 140 mg (51%) of white
solid. 1H NMR δ 7.46 (dd, J = 22.6/8.1 Hz, 4H), 7.23–7.09 (m, 2H), 7.00 (d, J = 6.0 Hz, 1H), 5.62
(d, J = 6.6 Hz, 1H), 3.77 (s, 3H), 3.15 (q, J = 10.5 Hz, 2H); 19F NMR δ -62.85, -133.79 (d, J = 20.5
Hz), -161.95 (dd, J = 20.6/20.6 Hz); 13C NMR δ 170.8, 162.1 (app. d, JCF = 3.4 Hz), 151.6 (ddd, JCF
= 250.1/9.9/4.3 Hz), 139.60 (dt, JCF = 251.5/14.8 Hz), 138.9, 136.8–136.3 (m), 136.2, 128.0, 127.7,
123.9 (app. d, JCF = 276.8 Hz), 111.5–110.9 (m), 56.5, 53.4, 41.6 (q, JCF = 29.8 Hz); m/z MS (TOF
ES+) C18H14F6NO3 [MH]+ calcd 406.1, found 406.1; LC-MS tR = 3.8 min.
Methyl
2-(2,3,3,3-tetrafluoropropanamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-
yl)acetate (5d). Compound 4 (150 mg, 0.51 mmol) was converted to the title compound according
to General Procedure C using 2,3,3,3-tetrafluoropropanoic acid (59 L, 0.61 mmol), to give 110 mg
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 67
1
2
3
4
5
6
7
8
(51%, mixture of diastereomers) of white solid. 1H NMR δ 7.56–7.41 (m, 5H), 7.22–7.11 (m, 2H),
5.64 (d, J = 6.8 Hz, 1H), 5.23–4.98 (m, 1H), 3.80 (s, 3H); 19F NMR δ -75.86 (dd, J = 33.7/11.2 Hz),
-133.74 (dd, J = 36.3/20.5 Hz), -161.85 (dt, J = 41.1/20.5 Hz), -202.47 – -202.75 (m); m/z MS (TOF
ES-) C18H11F7NO3 [M]- calcd 422.1, found 422.0; LC-MS tR = 3.5 min.
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
Methyl
2-(2,2,3,3,3-pentafluoropropanamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-
yl)acetate (5e). Compound 4 (150 mg, 0.51 mmol) was converted to the title compound according
to General Procedure C using 2,2,3,3,3-pentafluoropropanoic acid (64 L, 0.61 mmol), to give 15
mg (7%) of white solid. 1H NMR δ 7.58 (d, J = 6.1 Hz, 1H), 7.54–7.41 (m, 4H), 7.22–7.12 (m, 2H),
5.59 (d, J = 6.7 Hz, 1H), 3.81 (s, 3H); 19F NMR δ -82.71–- 82.74 (m), -122.91–-122.96 (m), -
13
133.68 (d, J = 20.5 Hz), -161.72 (dd, J = 20.5/20.5 Hz); C NMR δ 169.8, 157.1, 151.6 (ddd, J =
250.1/10.0/4.3 Hz), 139.7 (dt, J = 252.6/15.3 Hz), 139.4–139.2 (m), 136.7–136.1 (m), 135.0, 128.0,
127.9, 117.9 (dt, J = 285.6/33.8 Hz), 111.6–111.1 (m), 107.9–106.1 (m), 56.5, 53.7.m/z MS (TOF
ES-) C18H10F8NO3 [M]- calcd 440.1, found 440.1; LC-MS tR = 3.5 min.
Methyl
2-(cyclopropanecarboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate
(5f). Compound 4 (140 mg, 0.47 mmol) was converted to the title compound according to General
Procedure A using cyclopropanecarbonyl chloride (47 L, 0.52 mmol), to give 150 mg (87%) of
white solid. 1H NMR δ 7.52–7.43 (m, 4H), 7.20–7.10 (m, 2H), 6.75 (br. d, J = 6.8 Hz, 1H), 5.65 (d,
J = 6.9 Hz, 1H), 3.76 (s, 3H), 1.52–1.44 (m, 1H), 1.04–0.92 (m, 2H), 0.86–0.70 (m, 2H); 19F NMR
δ -133.89 (d, J = 20.5 Hz), -162.18 (dd, J = 20.5/20.5 Hz); 13C NMR δ 173.2, 171.4, 151.4 (ddd, JCF
= 249.7/10.0/4.2 Hz), 141.1–137.8 (m), 138.2, 137.2, 136.6 (td, JCF = 7.8/4.6 Hz), 128.0, 127.4,
111.2–110.8 (m), 56.2, 52.9, 14.5, 7.60, 7.56; m/z MS (TOF ES+) C19H17F3NO3 [MH]+ calcd 364.1,
found 363.9; LC-MS tR = 3.5 min.
Methyl
2-(cyclobutanecarboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate
(5g). Compound 3 (160 mg, 0.66 mmol) was converted to methyl 2-(4-bromophenyl)-2-
(cyclobutanecarboxamido)acetate according to General Procedure A using cyclobutanecarbonyl
chloride (82 L, 0.72 mmol). The crude product was reacted with 3,4,5-trifluorophenylboronic acid
26
ACS Paragon Plus Environment

Page 27 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
8
9
(138 mg, 0.79 mmol), Pd(PPh3)2Cl2 (14 mg, 0.02 mmol), Na2CO3 (1M, 1.3 mL) in THF (3.9 mL) in
a sealed microwave vial and heated at 100 C for 2 h. After cooling, the mixture was diluted with
EtOAc (10 mL) and water (10 mL), and the mixture extracted with EtOAc. The organic layer was
dried over Na2SO4, filtered and concentrated. The crude product was purified by FCC to give 220
mg (89%) of a colourless oil. 1H NMR δ 7.51–7.40 (m, 4H), 7.21–7.10 (m, 2H), 6.46 (br. d, J = 6.8
Hz, 1H), 5.61 (d, J = 6.9 Hz, 1H), 3.76 (s, 3H), 3.10 (pd, J = 8.5/0.9 Hz, 1H), 2.37–2.12 (m, 4H),
2.03–1.83 (m, 2H); 19F NMR δ -133.87 (d, J = 20.6 Hz), -162.12 (dd, J = 20.6/20.6 Hz); 13C NMR
δ 175.2, 171.3, 151.5 (ddd, J = 249.7/10.0/4.2 Hz), 139.4 (dt, J = 252.2/15.4 Hz), 138.4, 136.9,
136.7–136.4 (m), 128.0, 127.4, 111.5–110.6 (m), 56.1, 53.0, 39.6, 25.26, 25.25, 18.2; m/z MS (TOF
ES+) C20H19F3NO3 [MH]+ calcd 378.1, found 377.9; LC-MS tR = 3.5 min.
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
Methyl
2-(cyclopentanecarboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate
(5h). Compound 4 (130 mg, 0.44 mmol) was converted to the title compound according to General
Procedure A using cyclopentanecarbonyl chloride (60 L, 0.48 mmol), to give 70 mg (41%) of
yellow foamy solid. 1H NMR δ 7.51–7.37 (m, 4H), 7.19–7.08 (m, 2H), 6.66 (d, J = 6.9 Hz, 1H),
5.62 (d, J = 7.0 Hz, 1H), 3.74 (s, 3H), 2.70–2.59 (m, 1H), 1.95–1.49 (m, 8H); 19F NMR δ -133.9 (d,
J = 20.5 Hz), -162.2 (dd, J = 20.5/20.5 Hz); 13C NMR δ 175.9, 171.5, 151.5 (ddd, JCF
249.8/10.0/4.2 Hz), 139.5 (dt, JCF = 252.0/15.4 Hz), 138.4–138.3 (m), 137.2, 136.7 (td, JCF
=
=
7.8/4.6 Hz), 128.0, 127.5, 111.6–110.9 (m), 56.1, 53.1, 45.5, 30.4, 30.3, 26.0 (2C); m/z MS (TOF
ES+) C21H21F3NO3 [MH]+ calcd 392.1, found 392.1; LC-MS tR = 6.8 min.
Methyl 2-(cyclohexanecarboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5i).
Compound 4 (130 mg, 0.44 mmol) was converted to the title compound according to General
Procedure A using cyclohexanecarbonyl chloride (64 L, 0.48 mmol), to give 50 mg (28%) of pale
yellow solid. 1H NMR δ 7.51–7.37 (m, 4H), 7.21–7.07 (m, 2H), 6.59 (d, J = 6.9 Hz, 1H), 5.61 (d, J
= 7.0 Hz, 1H), 3.75 (s, 3H), 2.19 (tt, J = 11.7/3.5 Hz, 1H), 1.96–1.62 (m, 5H), 1.53–1.13 (m, 5H);
19F NMR δ -133.9 (d, J = 20.5 Hz), -162.2 (dd, J = 20.6/20.6 Hz); 13C NMR δ 175.5, 171.5, 151.5
(ddd, JCF = 249.8/10.0/4.3 Hz), 139.5 (dt, JCF = 252.2/15.4 Hz), 138.5–138.4 (m), 137.3, 136.7 (td,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 67
1
2
3
4
5
JCF = 7.9/4.8 Hz), 128.0, 127.5, 111.5–110.8 (m), 55.9, 53.1, 45.2, 29.7, 29.5, 25.79, 25.75, 25.7;
m/z MS (TOF ES+) C22H23F3NO3 [MH]+ calcd 406.2, found 406.2; LC-MS tR = 4.0 min.
6
7
8
9
Methyl 2-(tetrahydro-2H-pyran-4-carboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-
yl)acetate (5j). Compound 4 (120 mg, 0.41 mmol) was converted to the title compound according
to General Procedure B using tetrahydropyrancarboxylic acid (56 mg, 0.43 mmol), to give 126 mg
(76%) of white solid. 1H NMR δ 7.50–7.40 (m, 4H), 7.20–7.09 (m, 2H), 6.64 (d, J = 6.8 Hz, 1H),
5.60 (d, J = 6.8 Hz, 1H), 4.08–3.96 (m, 2H), 3.76 (s, 3H), 3.47–3.37 (m, 2H), 2.51–2.40 (m, 1H),
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
13
1.88–1.73 (m, 4H); 19F NMR δ -133.8 (d, J = 20.5 Hz), -162.0 (dd, J = 20.5/20.5 Hz); C NMR δ
174.0, 171.3, 151.6 (ddd, JCF = 250.0/10.1/4.3 Hz), 139.6 (dt, JCF = 252.3/15.2 Hz), 138.7, 136.9,
136.7–136.4 (m), 128.0, 127.7, 111.5–110.9 (m), 67.3, 56.1, 53.3, 42.0, 38.8, 29.2, 29.1; m/z MS
(TOF ES+) C21H21F3NO4 [MH]+ calcd 408.1, found 408.2; LC-MS tR = 3.7 min.
Methyl 2-((3r,5r,7r)-adamantane-1-carboxamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-
yl)acetate (5k). Compound 4 (200 mg, 0.68 mmol) was converted to the title compound according
to General Procedure A using 1-adamantanecarbonyl chloride (148 mg, 0.75 mmol), to give 150 mg
(48%) of white solid. 1H NMR δ 7.50–7.40 (m, 4H), 7.21–7.09 (m, 2H), 6.77 (d, J = 6.6 Hz, 1H),
5.57 (d, J = 6.6 Hz, 1H), 3.75 (s, 3H), 2.10–2.02 (m, 3H), 1.89 (d, J = 2.7 Hz, 6H), 1.80–1.65 (m,
13
6H); 19F NMR δ -133.9 (d, J = 20.6 Hz), -162.2 (dd, J = 20.6/20.6 Hz); C NMR δ 177.7, 171.5,
151.6 (ddd, JCF = 249.7/10.0/4.3 Hz), 139.5 (dt, JCF = 252.2/15.4 Hz), 138.5–138.4 (m), 137.2,
136.7 (td, JCF = 7.7/4.4 Hz), 127.9, 127.6, 111.4–110.9 (m), 56.0, 53.1, 39.2, 36.5, 28.1; LC-MS Rt
= 4.4 min (compound did not ionize).
Methyl 2-(3,3-dimethylbutanamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5l).
Compound 4 (149 mg, 0.50 mmol) was converted to the title compound according to General
Procedure A using 3,3-dimethylbutyryl chloride (77 L, 0.56 mmol), to give 110 mg (55%) of
white solid. 1H NMR δ 7.50–7.41 (m, 4H), 7.20–7.10 (m, 2H), 6.47 (d, J = 6.7 Hz, 1H), 5.62 (d, J =
6.9 Hz, 1H), 3.75 (s, 3H), 2.13 (s, 2H), 1.03 (s, 9H); 19F NMR δ -133.9 (d, J = 20.6 Hz), -162.2 (dd,
J = 20.6/20.6 Hz); 13C NMR δ 171.4, 171.2, 151.6 (ddd, JCF = 249.8/10.0/4.3 Hz), 139.5 (dt, JCF
=
28
ACS Paragon Plus Environment

Page 29 of 67
1
Journal of Medicinal Chemistry
2
3
4
5
6
7
252.3/15.3 Hz), 138.66–138.33 (m), 137.2, 137.0–136.0 (m), 128.2, 127.5, 111.4–111.0 (m), 56.1,
53.1, 50.3, 31.2, 29.9; m/z MS (TOF ES+) C21H23F3NO3 [MH]+ calcd 394.2, found 394.2; LC-MS tR
= 4.0 min.
8
9
Methyl 2-(2-cyclopropylacetamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5m).
Compound 4 (200 mg, 0.68 mmol) was converted to the title compound according to General
Procedure B using cyclopropylacetic acid (70 L, 0.71 mmol), to give 141 mg (55%) of white solid.
1H NMR δ 7.50–7.42 (m, 4H), 7.19–7.08 (m, 3H), 5.64 (d, J = 7.0 Hz, 1H), 3.75 (s, 3H), 2.28–2.14
(m, 2H), 1.08–0.95 (m, 1H), 0.68–0.60 (m, 2H), 0.28–0.18 (m, 2H); 19F NMR δ -133.9 (d, J = 20.5
Hz), -162.2 (dd, J = 20.5/20.5 Hz); 13C NMR δ 172.1, 171.3, 151.5 (ddd, JCF = 249.8/10.0/4.2 Hz),
139.5 (dt, JCF = 252.1/15.3 Hz), 138.5–138.4 (m), 137.1, 136.6 (td, JCF = 7.8/4.7 Hz), 128.0, 127.5,
111.6–110.8 (m), 56.0, 53.1, 41.2, 7.1, 4.71, 4.68; m/z MS (TOF ES+) C20H19F3NO3 [MH]+ calcd
378.1, found 378.1; LC-MS tR = 3.8 min.
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
Methyl 2-(2-cyclobutylacetamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5n).
Compound 4 (200 mg, 0.68 mmol) was converted to the title compound according to General
Procedure B using cyclobutylacetic acid (74 L, 0.71 mmol), to give 160 mg (60%) of white solid.
1H NMR δ 7.49–7.38 (m, 4H), 7.17–7.09 (m, 2H), 6.73 (d, J = 6.9 Hz, 1H), 5.60 (d, J = 6.9 Hz,
1H), 3.74 (s, 3H), 2.78–2.59 (m, 1H), 2.38 (d, J = 7.7 Hz, 2H), 2.20–2.05 (m, 2H), 1.96–1.65 (m,
19
13
4H); F NMR δ -133.9 (d, J = 20.5 Hz), -162.2 (dd, J = 20.6/20.6 Hz); C NMR δ 172.0, 171.3,
151.5 (ddd, JCF = 249.8/10.0/4.3 Hz), 139.5 (dt, JCF = 252.2/15.4 Hz), 138.4 (d, JCF = 1.6 Hz),
137.0, 136.6 (td, JCF = 7.8/4.6 Hz), 128.0, 127.5, 111.3–110.9 (m), 56.0, 53.1, 43.4, 32.7, 28.42,
28.40, 18.6; m/z MS (TOF ES+) C21H21F3NO3 [MH]+ calcd 392.1, found 392.1; LC-MS tR = 3.9
min.
Methyl 2-(2-cyclopentylacetamido)-2-(3',4',5'-trifluoro-[1,1'-biphenyl]-4-yl)acetate (5o).
Compound 4 (120 mg, 0.41 mmol) was converted to the title compound according to General
Procedure B using cyclopentylacetic acid (54 L, 0.43 mmol), to give 140 mg (85%) of white solid.
1H NMR δ 7.51–7.40 (m, 4H), 7.20–7.10 (m, 2H), 6.59 (d, J = 6.9 Hz, 1H), 5.63 (d, J = 6.9 Hz,
29
ACS Paragon Plus Environment