Hit-to-Lead Optimization of a Novel Class of Potent, Broad-Spectrum Trypanosomacides

Article abstract of DOI:10.1021/acs.jmedchem.6b00442

The parasitic trypanosomes Trypanosoma brucei and T. cruzi are responsible for significant human suffering in the form of human African trypanosomiasis (HAT) and Chagas disease. Drugs currently available to treat these neglected diseases leave much to be

Full text of DOI:10.1021/acs.jmedchem.6b00442

Subscriber access provided by Northern Illinois University
Article
Hit-to-Lead Optimisation of a Novel Class of
Potent, Broad-Spectrum Trypanosomacides
Stephanie Russell, Raphaël Rahmani, Amy J. Jones, Harriet Newson, Kevin Neilde, Ignacio
Cotillo, Marzieh Rahmani Khajouei, Lori Ferrins, Sana Qureishi, Nghi Nguyen, Maria S.
Martínez-Martínez, Donald F. Weaver, Marcel Kaiser, Jennifer Riley, John Thomas, Manu
De Rycker, Kevin D Read, Gavin R. Flematti, Eileen Ryan, Scott Tanghe, Ana Rodriguez,
Susan A. Charman, Albane Kessler, Vicky M. Avery, Jonathan B. Baell, and Matthew J. Piggott
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00442 • Publication Date (Web): 22 Aug 2016
Downloaded from http://pubs.acs.org on August 22, 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 140
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 140
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
HitꢀtoꢀLead Optimisation of a Novel Class of Potent, BroadꢀSpectrum
Trypanosomacides
a
b
c
a
b,d
Stephanie Russell, Raphaël Rahmani, * Amy J. Jones, Harriet Newson, Kevin Neilde,
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
d
a
b
a
b
Ignacio Cotillo, Marzieh Rahmani Khajouei, Lori Ferrins, Sana Qureishi, Nghi Nguyen,
d
e†
f,g
g
Maria S. Martinez–Martinez, Donald F. Weaver, Marcel Kaiser, Jennifer Riley, John
g
g
h
a
h
i
Thomas, Manu de Rycker, Kevin D. Read, Gavin R. Flematti, Eileen Ryan, Scott Tanghe,
i
b,h
d
c
b
Ana Rodriguez, Susan A. Charman, Albane Kessler, Vicky M. Avery, Jonathan B. Baell, *
a
Matthew J. Piggott. *
a
School of Chemistry and Biochemistry, The University of Western Australia, Perth 6009,
Australia
b
Medicinal Chemistry, Monash Institute of Pharmaceutical Sciences, Monash University,
Parkville, Victoria 3052, Australia
c
Eskitis Institute for Drug Discovery, Griffith University, Brisbane Innovation Park, Don Young
Road, Nathan, Queensland 4111, Australia
d
GlaxoSmithKline, Tres Cantos, Spain
e
†
Department of Chemistry, Dalhousie University, Halifax, Canada. Current address, Toronto
Western Research Institute, Toronto, Canada.
f
Swiss Tropical and Public Health Institute, Socinstrasse, 4051 Basel, Switzerland
University of Basel, Petesplatz 1, 4003 Basel, Switzerland
h
g
Drug Discovery Unit, Division of Biological Chemistry and Drug Discovery, School of Life
Sciences, University of Dundee, Dundee, UK.
ACS Paragon Plus Environment

Page 3 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
h
i
Centre for Drug Candidate Optimisation, Monash University, Parkville, Victoria 3052,
Australia
AntiꢀInfectives Screening Core, New York University School of Medicine, New York, NY
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
0010, USA
ABSTRACT
The parasitic trypanosomes Trypanosoma brucei and T. cruzi are responsible for significant
human suffering in the form of human African trypanosomiasis (HAT) and Chagas disease.
Drugs currently available to treat these neglected diseases leave much to be desired. Herein we
report optimisation of a novel class of Nꢀ(2ꢀ(2ꢀphenylthiazolꢀ4ꢀyl)ethyl)amides, carbamates and
ureas, which rapidly, selectively, and potently kill both species of trypanosome. The mode of
action of these compounds is unknown, but does not involve CYP51 inhibition. They do,
however, exhibit clear structure–activity relationships, consistent across both trypanosome
species. Favourable physicochemical parameters place the best compounds in CNS drugꢀlike
chemical space but, as a class, they exhibit poor metabolic stability. One of the best compounds
(64a) cleared all signs of T. cruzi infection in mice when CYP metabolism was inhibited, with
sterile cure achieved in one mouse. This family of compounds thus shows significant promise for
trypanosomiasis drug discovery.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 140
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
Human African trypanosomiasis (HAT, African sleeping sickness) and Chagas disease
(American trypanosomiasis) are related parasitic diseases causing significant morbidity and
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
mortality in subꢀSaharan Africa and South America, respectively. _ENREF_1 There is a
critical need for new treatment options for these diseases, which affect millions of the world’s
poorest people.
HAT is caused by the protozoans Trypanosoma brucei gambiense and T. b. rhodesiense,
transmitted by the bite of infected tsetse flies. T. b. gambiense is prevalent in Western and
Central Africa, accounting for 98% of reported cases. It causes chronic human infection, and has
3
long symptomꢀfree periods that can last several years. T. b. rhodesiense infection occurs
primarily in Eastern and Southern Africa, and results in acute illness, leading to death within
4
months if treatment is not received. Indeed, both forms of the disease are usually fatal if left
4
, 5
untreated.
6
In 2013 HAT was estimated to be responsible for approximately 6,900 deaths _ENREF_6 and
7
3
90,000 disabilityꢀadjusted life years (DALYs). Fortunately, the number of infections has
reduced in recent times, with less than 4,000 confirmed new cases in 2014, although the actual
1
number is estimated to be around 20,000. However, over 60 million people are thought to be at
1
risk of infection, and it is worth noting that historical epidemics have arisen from a base of low
1
prevalence. _ENREF_1Livestock are also susceptible to trypanosomiasis, adding to the burden
1
of the disease, and acting as a parasite reservoir.
ACS Paragon Plus Environment

Page 5 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
HAT consists of two stages: the first is characterised by the spread of the parasite in the blood
and the lymphatic system; the second stage of the disease occurs when the parasite crosses the
blood brain barrier. Early symptoms include fever, itchiness, joint pain and headache, and often
go undiagnosed until the parasite has already entered the brain, where it causes dramatic mood
swings, poor coordination, soft tissue swelling, confusion, convulsions and the changed sleep
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
, 8
patterns for which the disease is commonly named. Coma and eventual death ensue.
Treatment of HAT is difficult, especially in the CNS stage. Only a few drugs are available
(Figure 1): suramin and pentamidine for the early, peripheral stage of the disease (stage I), and
melarsoprol, eflornithine and its combination therapy with nifurtimox (NECT) for stage II. These
drugs are difficult to administer, toxic, and expensive. As an example of the issues associated
with the current therapies, melarsoprol was introduced for the treatment of HAT in 1949 and is
still being used, even though it must be administered intravenously for 10–26 days, has sideꢀ
effects that often results in patient nonꢀcompliance, and causes fatal reactive encephalopathy in
3, 9
3
−10% of patients. The more recently approved drug eflornithine, and NECT, are only active
against T. b. gambiense. Pafuramidine (Figure 1) _ENREF_11was recently taken to clinical
1
0, 11
11
trial,
but development was abandoned due to renal toxicity. Oxaborole (SCYXꢀ7158) is
12ꢀ14
scheduled to enter clinical trials in 2016,
and a phase II/III clinical trial on fexinidazole in
1
5, 16
underway.
However, some cross resistance is expected for fexinidazole since its mode of
17
action is similar to that of nifurtimox.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 140
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
Established therapeutics
SO
3
SO
3
O
3
S
SO
3
SO3
NH
O S
3
HN
suramin
O
O
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
H
2
N
NH
2
HN
O
O
NH
NH
2
O
N
N
N
H
N
HN
N
NH2
O
H
As
O
S
S
melarsoprol
NH
2
HO
pentamidine
NH
2
CHF
2
O
N
3
H N
N N
S
H
CO
2
N
NBn
O
2
O N
O
H
3
N
2
O N
O
eflornithine
nifurtimox
benznidazole
Drug candidates in clinical trials
MeON
NOMe
NH2
H N
2
O
pafuramidine
F
OH
H
N
N
N
O
S
B
O
O
2
N
CF
3
O
fexinidazole
oxaborole (SCYX-7158)
Figure 1. Structures of drugs investigated or currently used for the treatment of HAT and/or
Chagas disease.
T. cruzi, spread by infected triatomine (‘kissing’) bugs, is the causative agent of Chagas disease,
2
and is endemic in 21 countries in Latin America, affecting as many as 6–7 million people. The
2
013 Global Burden of Disease Study, estimated that Chagas disease was responsible for 10,600
ACS Paragon Plus Environment

Page 7 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
6
7
deaths and 339,000 DALYs. As with HAT, Chagas disease develops in two stages. During the
acute phase, the parasite multiplies and spreads in the patient’s blood, without symptoms in most
cases. The chronic stage is characterised by cardiac disorders and digestive alterations, due to the
accumulation of the parasite in the relevant organs and, if left untreated, can be fatal. It is worth
noting that Chagas disease has been reported to be one of the leading worldwide causes of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
8
myocarditis; indeed it is predicted that as many as 200,000 people will die from Chagasic
1
9
cardiomyopathy over the next five years. Only two drugs are currently available to treat Chagas
disease: nifurtimox and benznidazole (Figure 1), but these also have adverse sideꢀeffects, and
19, 20
21
neither can treat, chronic Chagas disease.
_ENREF_20 Cure is possible during the early
2
2
23
acute phase, and indeterminate phase (chronic infection without signs of cardiac disease), but
1
9
disease diagnosis at these stages is rare.
Clearly, there is an urgent need for alternative, safer and orallyꢀavailable treatments for
trypanosomiasis. However, due to the poor market, there is little incentive for pharmaceutical
companies to invest in the development of drugs for these diseases. Collaborative initiatives
between Pharma and academic groups with a focus on neglected third world diseases have
emerged and are becoming more common. For example, the Tres Cantos Open Lab Foundation,
which is supported by GSK, recently published the results of a highꢀthroughput screen (HTS)
2
4
against several kinetoplastid parasites, including T. brucei and T. cruzi. Philanthropic
organizations such as the Drugs for Neglected Diseases initiative (DNDi) also play a major role
in the discovery and development of new drugs for these diseases, funding projects and making
2
5
assays more available to the research community.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 140
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The efforts of organizations like the DNDi and Tres Cantos Open Lab Foundation have increased
awareness about neglected parasitic diseases, and inspired many new trypanosomiasis drug
1
3, 26ꢀ36
37ꢀ42
discovery programs. Most campaigns have focused on either T. brucei
or T. cruzi;
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
fewer have involved a concerted effort to discover leads that are effective against both
4
3ꢀ53
pathogenic species,
despite the economies of broadꢀspectrum drug development, especially
for neglected diseases.
We previously reported the results of a DNDiꢀsponsored highꢀthroughput screen (HTS) of the
80,000ꢀstrong Walter and Eliza Hall Institute (WEHI) Stage 1 compound library for
compounds inhibiting the growth of T. b. brucei, a human nonꢀinfective species that is
~
5
4
commonly used in drug discovery as a surrogate for the human pathogens. Eleven hits with
novel, drugꢀlike structures and physicochemical properties were identified. We have already
5
5ꢀ59
published the hitꢀtoꢀlead optimisation of five of these hits.
Herein we report our work based
6
0
on hit 1 (Figure 2). This compound offers a very attractive starting point for hitꢀtoꢀlead
optimisation since it exhibits good physicochemical properties, including a low molecular weight
–
1
2
61
(306 g.mol ), a low polar surface area (41 Å ), which is necessary for CNS penetration in
order to treat the second stage of HAT, and a relatively simple structure amenable to analogue
synthesis. Importantly, it is also modestly active against T. cruzi.
ACS Paragon Plus Environment

Page 9 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
S
O
N
N
H
F
1
IC50 (ꢀM) SI
T. b. br ucei
0.80
1.5
>96 (vs HEK)
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
T. b. rhodesiense
T. cruzi
42
27
(vs L6)
2.3
6
0
Figure 2. The hit from a HTS of the WEHI compound library against T. b. brucei providing the
inspiration for this study. SI = selectivity index = IC mammalian cell line/IC trypanosome;
50
50
HEK = human embryonic kidney cells, L6 is a rat skeletal muscle cell line.
RESULTS AND DISCUSSION
Synthesis
Acyl modifications
One of the appealing qualities of HTS hit 1 (Figure 2) is the possibility of rapid access to amide
analogues to allow optimisation of the acyl substituent. A Gabrielꢀlike synthesis of the primary
6
2
amine 8 required for this purpose was used initially (Scheme 1). Thus, conjugate addition of
phthalimide (3) to methyl vinyl ketone (2) gave adduct 4. Regioselective bromination followed
by condensation with thiobenzamide (6) provided the thiazole 7. Deprotection was inefficient,
possibly because of difficulties isolating the primary amine 8, which was unusually polar,
difficult to extract from aqueous solutions and chromatograph. For this reason alternative routes
to 8 were explored.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 140
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
62
Scheme 1. Initial literature synthesis of key primary amine 8
O
S
HN
Ph
NH
2
S
S
O
O
O
O
3
b
6
d
Ph
Ph
Br
NPhth
NPhth
c
N
NPhth
N
NH
2
a
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
4
5
7
8
Reagents, conditions and yields: a) NaOH, EtOAc, 50%; b) Br , MeOH 48%; c) iꢀPrOH,
2
HCl, 75%; d) H NNH .H O, 48%.
2
2
2
An improved synthesis was devised (Scheme 2), providing 13, the hydrochloride salt of 8, which
was considerably more efficiently than the original route. Condensation of thiobenzamide (6)
6
3
and 1,3ꢀdichloroacetone (9) as reported, gave the chloromethylthiazole 10, which was
64
converted to the corresponding nitrile 11. Attempts to reduce the nitrile with LiAlH , BH or
4
3
6
5
cobalt(II) chloride/sodium borohydride suffered, in some cases, from competing reduction of
the thiazole, but more generally from the difficulty in isolating the primary amine 8, as noted
earlier. Ultimately, an in situ protection of the primary amine obtained by nickel boride
6
6
reduction of nitrile 11, provided the tꢀbutyl carbamate 12, which was much easier to isolate and
purify, in good yield. Protonolysis then provided the dihydrochloride 13, which was used directly
in subsequent steps.
Scheme 2. Improved synthesis of primary amine hydrochloride 13
S
Ph
NH
2
S
S
S
S
O
6
b
c
d
Ph
Ph
Ph
Ph
Cl
Cl
NH
Cl
Cl
Cl
CN
a
N
N
N
NBoc
H
N
H
3
9
10
11
12
13
Reagents, conditions and yields: a) 1. acetone, 2. H SO , 85%; b) KCN, DMF, 90%; c) 1.
2
4
NaBH , NiCl .H O, Boc O, MeOH, 2. NEt , 82%; d) HCl/H O, dioxane, quant.
4
2
2
2
3
2
ACS Paragon Plus Environment

Page 11 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
In parallel, an alternative preparation of 8 that avoids the use of cyanide was investigated, based
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
6
7
on a previously described synthesis of a methylthiazole. Again, this synthesis starts with a
cyclocondensation of thiobenzamide (6), in this case with ethyl chloroacetoacetonate (14). The
resulting ester (15) was reduced to the primary alcohol (16), which was activated by mesylation.
Due to its reactivity, the mesylate 17 was not purified or characterised but immediately subjected
to nucleophilic substitution to give azide 18, which was reduced to yield the primary amine 8,
under conditions that did not require an aqueous workup.
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
Scheme 3. Alternative synthesis of primary amine 8.
S
Ph
NH
2
S
S
S
S
O
6
b
d
e
Ph
Ph
[
Ph
Ph
Cl
CO
2
Et
2
CO Et
a
N
N
OR
c
N
N3
N
NH
2
14
15
16, R = H
17, R = Ms]
18
8
Reagents, conditions and yields: a) NaHCO , EtOH, reflux, 76%; b) LiAlH , THF, –78°C, 71%;
3
4
c) MsCl, NEt , THF, 96% (crude); d) NaN , THF, reflux, 86%; e) Pd/C, H , MeOH, 99%.
3
3
2
The primary amine 8, or dihydrochloride 13, was acylated to give a series of amides 21, either by
condensation with acid chlorides 19, or carboxylic acids 20 in the presence of a coupling agent
(Scheme 4). The yields for the DCC couplings were compromised by difficult separations from
the byproduct DCU, hence the move to the ‘friendlier’ coupling agents HBTU or EDCI. The m-
ethynylbenzamides (see Table 3) were prepared by Sonogashira coupling of the corresponding
iodide to give 21Bu, followed by desilylation, affording 21Bv.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 140
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The phenylsulfonamide 25, sulfamide 27 and exocyclic ureas 23 were prepared by condensation
of 8 or 13 with benzenesulfonyl chloride (24), pyrrolidineꢀ1ꢀsulfonyl chloride (26) or the
appropriate carbamoyl chloride 22, respectively. To access a wider range of ureas and
carbamates more efficiently, the aminium hydrochloride 13 was converted into the activated
imidazoleꢀN-carboxamide 28, which is chromatographically stable and could be stockpiled, but
was easily transformed to urea 30 or carbamate 29 targets by reaction with an amine or alkoxide,
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
6
8
respectively.
Scheme 4. Synthesis of target amides, carbamates, ureas, sulfonamide and sulfamide.
O
O
O
Cl
N
or
Cl
R
HO
R
n
S
S
O
19
20
22, n = 1, 2
O
Ph
Ph
N
N
H
R
a
b
c
N
N
N
H
21
23
n
O
O
O
O
S
Cl
N
S
24
d
Cl
Ph
S
S
S
O
O
26
O O
Ph
Ph
Ph
Ph
Ph
S
S
N
N
H
Ph
N
NH2
8 (or 13)
N
N
N
e
H
2
5
27
f
S
S
S
O
O
O
g
h
Ph
2
R
N
N
N
N
H
OR
N
N
N
N
R
N
H
H
1
2
9
28
30
Reagents, conditions and yields: a) 19, NEt , DCM, 57–85%*; b) 20, DCC, HOBt; or 20, HBTU,
3
2
4–87%*; d) PhSO Cl, NEt , DCM, 40%; c) 22, n = 1, NEt , DCM, 92%; 22, n = 2, NEt , DCM,
2 3 3 3
71%; d) DIPEA, DCM, 32%; e) 26, NEt , DCM, 41%; f) CDI, MeCN, DMF, 69%; g) ROH,
3
1
2
NaH, DMF, 32–55%*; h) HNR R , NEt , DCM, 55–68%*. *See experimental section for
3
individual reaction yields. Isopropylmethylamine was prepared by reductive amination of
6
9
acetone.
ACS Paragon Plus Environment

Page 13 of 140
Journal of Medicinal Chemistry
1
2
3
Variation of the thiazoleꢀ2ꢀsubstituent
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
Initial efforts in this area (Scheme 5) commenced with similar strategies to those outline above.
Commercial 2ꢀpyridinethioamide (31) condensed with 1,3ꢀdichloroacetone (9) to give the
chloromethylthiazole 32 in modest yield. We were concerned about the propensity of this
compound to polymerise, and indeed the analogous reactions of 3ꢀ (36) and 4ꢀpyridinethioamide
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
(38) produced complex mixtures, which were abandoned. However, it was possible to convert 32
to the corresponding nitrile 33. Borane reduction (the superior reduction/in situ protection
described in scheme 2 had not been developed yet), provided the required primary amine 34 in
low yield, which was then acylated to give the target amides 35a and b.
Scheme 5. Initial route to 2ꢀ(2ꢀpyridyl)thiazole targets
S
S
S
S
S
N
N
N
N
O
N
a
b
c
d
NH
2
Cl
N
CN
N
N
N
NH2
N
N
R
H
3
1
32
33
34
35a, R = Ph,
35b, R = cyclopentyl
S
S
S
S
a
a
N
NH
2
NH
2
Cl
N
Cl
N
N
N
3
6
37
38
39
Reagents, conditions and yields: a) 1. acetone, 1,3ꢀdichloroacetone (9); 2. H SO , 33%; b) KCN,
2
4
DMF, 70°C, 74%; c) BH ꢀTHF, MeOH, 25%; NEt , PhCOCl, 84% or cyclopentanecarbonyl
3
3
chloride, 59%.
To access the other pyridylthiazoles, an alternative route was devised (Scheme 6). The αꢀ
70
bromoketone 40, available in one step from levulinic acid, was condensed with several
thioamides 41 to give the corresponding thiazoles 42, and ammonolysis, provided the primary
amides 43. The 2ꢀ(pyridyl)thiazoles 43a/b were subjected to a (diacetoxyiodo)benzeneꢀmediated
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 140
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
71
Hofmann rearrangement with capture of the intermediate isocyanates by the solvent to give the
methyl carbamates 46. Hydrolysis yielded the desired amines as their trihydrochlorides 47,
which were acylated to give the target amides 35. In the case of the 2ꢀ(2ꢀthienyl)thiazole 43c,
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
71
Hofmann rearrangement in water gave the primary amine 44 directly, and acylation once again
provided the target amides 45.
Scheme 6. Alternative route to target amides involving a Hofmann rearrangement
S
R
NH
2
O
S
S
S
S
41
b
c
Br
O
O
R
R
O
NH
2
N
N
N
NH2
a
O
O
O
40
42a, R = 3-Py
2b, R = 4-Py
2c, R = 2-thienyl
43a−c
44
4
4
d
f
S
S
S
S
O
S
O
3
HCl
NH
f
e
Py
Py
Py
N
N
H
R
N
2
N
N
H
O
N
N
H
R
3
3
5c: 3-Py, R = Ph
47a,b
46a,b
45a, R = Ph,
45b, R = cyclopentyl
5d: 3-Py, R = cyclopentyl
35e: 4-Py, R = Ph
5f : 4-Py, R = cyclopentyl
3
Reagents, conditions and yields: a) MeOH, reflux, 22–60%*; b) conc. NH , THF, 45–75%* c)
3
;
PhI(OAc) , H O/dioxane, 85%; d) PhI(OAc) , MeOH, 46a 59%, 46b 70%; e) 4 M HCl, reflux,
2
2
2
quant.; f) NEt , PhCOCl (41–89%) or cyclopentanecarbonyl chloride (31–54%)*. Py = pyridyl.
3
*See experimental for individual reaction yields.
The obvious deficiency of the two routes outlined above is their early divergent steps. We
therefore targeted a synthon that could be elaborated later in the synthesis; specifically, we
sought a 2ꢀhalothiazole such as 53 (Scheme 7) to take advantage of lateꢀstage transition metalꢀ
catalysed coupling to introduce various substituents. Efforts began with 2ꢀaminoꢀ4ꢀ
chloromethylthiazole hydrochloride (49), readily prepared from thiourea (48) and 1,3ꢀ
72
73
dichloroacetone (9). A Sandmeyer reaction was accompanied by partial halogen exchange,
ACS Paragon Plus Environment

Page 15 of 140
Journal of Medicinal Chemistry
1
2
3
giving 50, and chemoselective substitution of this mixture provided the nitrile 51. However,
attempted nickel boride reduction/in situ Bocꢀprotection to give 53, was accompanied by
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
1
dehalogenation, and it appeared (from the lack of aromatic signals in the H NMR spectrum of
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 crude product) that borane reduced the thiazole moiety with no evidence for the desired
primary amine 52. In an attempt to circumvent the incompatibility of the bromothiazole with
reductive conditions, the chloride 49 was converted to the corresponding nitrile 54.
Unfortunately, the nickel boride reduction/in situ Bocꢀprotection was again largely unsuccessful;
1
although small amounts of the putative dicarbamate 55 were detected by H NMR spectroscopy,
it was not able to be purified.
Scheme 7. First attempted synthesis of a 2ꢀbromothiazole synthon
S
S
S
S
S
a
b
c
d
2
H N
NH
2
H
2
N
Br
Br
Br
Cl
X
CN
N
H
49
N
N
N
NH2
Cl
c
4
8
50, X = Cl + Br
51
52
e
S
S
S
e
H
2
N
BocN
Br
CN
N
H
N
NBoc
H
N
NBoc
H
5
4
55
53
Reagents, conditions and yields: a) acetone, reflux, 82%; b) 1. H SO , NaNO , 2. HBr; c). KCN,
2
4
2
DMF, 31% (two steps); d) 1. BH –THF, MeOH, 2. HCl, 0%; e) 1. NaBH , NiCl .H O, Boc O,
3
4
2
2
2
MeOH, 2. NEt ; 0%.
3
Problems with the reduction of the nitrile prompted us to reinvestigate the phthalimide protecting
group (Scheme 8). Thus, cyclocondensation of thiourea (48) with the bromoketone 5 gave the 2ꢀ
7
4
aminothiazole hydrobromide 56 in excellent yield. However, attempted Sandmeyer reactions
under various conditions were unsuccessful; in all cases the absence of the expected thiazole
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 140
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
1
singlet in the H NMR spectra of the crude products of these reactions points to electrophilic
aromatic substitution at the 3ꢀposition, although the nature of the electrophile was not
7
5
determined. Fortunately the 2ꢀbromothiazole 58 was accessible via the thiocyanate 57. Initial
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
attempts to remove the phthalimide protecting group were unsuccessful, presumably due to
competing S Ar reactions of hydrazine at the thiazole 2ꢀposition. Thus, 58 was subjected to
N
Suzuki couplings with a range of arylboronic acids, to provide 59a–d. Deprotection, followed
immediately by treatment of the primary amine with CDI gave the activated ureas 60 a–d, which
were reacted with piperidine (at this point the potencyꢀboost afforded by the exocyclic ureas had
been discovered) affording the target compounds 61. Subsequently, conditions to effect the
deprotection of 58 were developed, and carbamoylation of the primary amine 62 provided urea
63a, allowing diversification by Suzuki coupling in the final step. In addition, carbamoylation,
carboxylation or acylation of crude 62, provided N,Nꢀdiethylurea 63b, isopropyl carbamate 63c
and 2,5ꢀdifluorobenzamide 63d, respectively. Suzuki coupling then gave the 2ꢀ(2,4ꢀ
difluorophenyl)thiazoles 64b–d, which were targeted, primarily, to elucidate metabolism and
improve stability (see below).
ACS Paragon Plus Environment

Page 17 of 140
Journal of Medicinal Chemistry
1
2
3
Scheme 8. Divergent synthesis of variously 2ꢀarylated thiazoles
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
O
O
Br
c
NCS
NPhth
NPhth
5
57
a
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
S
2
H N
d
N
NPhth
NPhth
b
56
S
S
e
Ar
Br
N
N
NPhth
5
9a, Ar = o-C
9b, Ar = m-C H F
9c, Ar = p-C
9d, Ar = 2,4-C H F
6
H
4
F
58
5
6
4
5
6
H F
4
5
6
3
2
h
f
S
S
O
Ar
Ar
Br
Br
N
N
H
N
N
NH2
60a−d
62
g
i
S
S
O
O
e
N
N
H
N
N
N
R
H
61
63a, R = 1-piperidyl
63b, R = NEt
3a, R = O-Pr
3a, R = 3,5-C H F
3 2
(table 6)
2
i
6
6
6
j
F
F
S
S
O
O
F
F
N
N
R
N
N
R
H
H
64
64
(table 7)
S
O
R
N
N
H
N
6
1
Reagents, conditions and yields: a) thiourea (48), MeOH, reflux, 89%; b) NaNO , HBr or
2
NaNO , HCl, NaBr, 0%; c) KSCN, EtOH, 78%; d) HBr, AcOH, reflux, 86%; e) ArB(OH) ,
2
2
Pd(PPh ) , NaHCO , PhMe, H O, 80°C, 40–62%, or ArB(OH) , PdCl dppf, TBAB, K CO ,
3 4
3
2
2
2
2
3
dioxane/H O 4:1, µw*, 49–73%; f) 1. H NNH .H O, MeOH, reflux; 2. CDI, MeCN/DMF, 30–
2
2
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 140
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
7
0% (two steps); g) piperidine, NEt , 55–70%; h) H NNH .H O, EtOH, 69% ; i) 1. pꢀnitrophenyl
3 2 2 2
chloroformate, NEt , DCM, 2. piperidine, 84% (63a); j) 1. H NNH .H O, EtOH, reflux; 2.
3
2
2
2
ClCOR, NEt , EtOAc*; *See experimental section for individual reaction yields.
3
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
Antiꢀtrypanosomal activity
Target compounds were firstly screened at 10 µM in the Alamar Blue growth inhibition assay
5
4
against T. b. brucei used to identify hit 1. Compounds causing 50% or greater inhibition at this
concentration were then subjected to a concentrationꢀresponse study, allowing determination of
IC values. In parallel, IC values for cytotoxicity to human embryonic kidney (HEK) cells
5
0
50
were measured for active compounds, and a selectivity index (SI) was calculated.
Based on the promising activity of this series against T. b. brucei, target compounds were also
screened against T. cruzi using an intracellular imaging assay with rat cardiomyocyte (H9c2)
2
4, 76
cells as the host.
Cytotoxicity to the host cells was assessed in parallel.
Potency optimisation in the acyl substituent
The first series of compounds investigated (Table 1) explored the size of the acyl substituent
(right hand side [RHS] as drawn) and the significance of the fluorine atom in hit 1. The synthetic
precursors, phthalimide 7 and primary amine 8 (Scheme 1), were completely inactive against T.
b. brucei, emphasising the importance of the amide functional group. Small amides (21a–c)
displayed little activity, but this improved with the size of the alkyl group, giving an early
suggestion that the acyl substituent engages with a hydrophobic pocket in the biomolecular
target. Alternatively, or perhaps in addition, a greater rate of enzymatic hydrolysis of the less
ACS Paragon Plus Environment

Page 19 of 140
Journal of Medicinal Chemistry
1
2
3
sterically hindered amides might be responsible for the lack of activity, although we have no
evidence to support this notion. As expected, the pivalamide 21e, differing from hit 1 only in the
lack of a metaꢀfluorine atom in the phenyl substituent, retained activity, although there was an
approximate halving of potency. However, the benzamide 21f bettered the potency of hit 1
against T. b. brucei and T. cruzi. Subsequent to the initiation of this project, the Tres Cantos open
Lab Foundation/GSK HTS mentioned in the introduction also identified 21f as being active
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
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
24
against T. brucei and T. cruzi.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 140
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
S
O
Ph
N
N
R
H
2
1
Table 1. Initial SAR around the acyl substituent
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
a
IC µM
IC µM
IC µM
IC µM
50
50
50
50
b
c
#
21
R
(SI )
(SI )
#21
R
(SI)
(SI)
d
e
T. b. brucei
T. cruzi
T. b. brucei
T. cruzi
a
H
n.i.
n.i.
n.i.
i
5.0 ± 0.2 (17)
–
O
O
b
c
Me
n.i.
j
5.5 ± 0.1 (15)
–
(
±)
H
Et
n.i.
n.i.
k
l
n.i.
n.i.
n.i.
–
–
–
(±
)
d
e
f
5.3 ± 0.3 (16)
1.7 ± 0.1 (50)
40 ± 10.1 (>1.2)
6.3 ± 0.1 (>7.9)
m
0.25 ± 0.01 (334) 0.40 ± 0.04 (>125)
n
o
n.i.
n.i.
16 ± 8 (>3.1)
40 ± 0 (>1.2)
O
O
g
0.31 ± 0.1 (281)
0.25 ± 0.02 (329)
0.32 ± 0.05 (>158)
h
–
a
b
All IC values are the average two independent experiments ± standard deviation. SI = Selectivity index relative
50
c
to HEK293 cells. SI = Selectivity index relative to the host H9c2 cells. n.i. = no inhibtion = < 50% inhibition at 10
d
e
µM. “–“ indicates that the compound was not tested. Lister 427 bloodstream form. Tulahuen strain engineered to
24, 75
express βꢀgalactosidase.
Alicyclic amides were also investigated with both the cyclopentylꢀ 21h and
cyclohexylcarboxamides 21g being approximately equipotent with the benzamide 21f. However,
the introduction of an electronegative heteroatom at the 4ꢀposition in the tetrahydropyranyl
ACS Paragon Plus Environment

Page 21 of 140
Journal of Medicinal Chemistry
1
2
3
derivate 21i led to a substantial loss of potency, suggesting a preference for hydrophobic
character at this position. The oxabicyclo[2.2.1]heptane 21j has a similar potency to 21i and at
this stage we cannot say whether the relatively poor activity is due to steric crowding or the
common oxygen atom. However extension (21l), ringꢀfusion (21m–o) or the incorporation of
larger bicyclic acyl substituents (21k) led to substantial drops or a complete loss of activity,
presumably because they are too large for the acylꢀbinding pocket.
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
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 clear SAR in this small series, activity against both trypanosome species, and excellent
selectivity index of the more potent analogues, encouraged us to further explore this class of
compounds, with an initial focus on stereoelectronic modifications to the RHS phenyl group,
through the series of benzamides 21B depicted in Table 2. As in the first series, for simplicity
and economy, 2ꢀphenylthiazoles were targeted with the view that the potency boosting metaꢀ
fluorine (present in hit 1) could be reinstalled once the acyl substituent had been optimised.
Against T. b. brucei, in all cases paraꢀsubstitution in the benzamides was deleterious, with
activity lost in all but the pꢀfluoro analogue 21Bc. The halving of potency of this compound
relative to the benzamide 21B is most likely due to sterics (rather than electronics) as an orthoꢀ
fluorine 21Ba had little effect on activity, while the mꢀfluorobenzamide 21Bb was marginally
more potent. Indeed, metaꢀsubstitution was generally best tolerated, albeit accompanied by a loss
of potency when the substituent is electronꢀdonating (21Bh, 21Bm). The difference in activity of
the isosteric mꢀmethyl 21Bh and mꢀchloro 21Be analogues is telling, and exemplifies a definite
preference for electronꢀwithdrawing substituents at the metaꢀposition. Indeed, the nitrile 21Bp is
the most active of the substituted benzamides. In contrast to the other results, activity was lost in
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 140
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the phenol 21Bk, fitting with the hypothesis discussed earlier, that the target biomolecule
contains a hydrophobic acylꢀbinding pocket.
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
S
O
N
N
H
R
21B
Table 2. Substituted Benzamides 21Ba–21Bq
orthoꢀ
metaꢀ
paraꢀ
IC µM
IC µM
IC µM
IC µM
IC µM
IC µM
5
0
50
50
50
50
50
R
#21B
(SI)
(SI)
#21B
(SI)
(SI)
#21B
(SI)
(SI)
T. b. brucei
T. cruzi
T. b. brucei
T. cruzi
T. b. brucei
0.77 ± 0.19
(54)
T. cruzi
0
.47 ± 0.6
(44)
0.19 ± 0.01 0.50 ± 0.32
F
a
d
g
j
–
–
b
e
c
f
–
(444)
0.42 ± 0.04
(200)
(>100)
1
.5 ± 0.3
(29)
Cl
–
n.i.
n.i.
–
–
3
.3 ± 0.8
3.2 ± 0.5
(>15)
3.9 ± 0.8
(21)
6.3 ± 0.4
(>7.9)
40 ± 12
(1.2)
Me
OH
OMe
CN
h
k
m
p
i
(
2.5)
.3 ± 0.1
18)
0.6 ± 3.6
2
7.9 ± 0.7
(>3)
n.i.
–
–
n
q
–
–
(
1
10 ± 0.7
(>5)
1.9 ± 0.3
(26)
3.2 ± 0.4
(>15)
l
n.i.
n.i.
(
2.0)
0
.13 ± 0.01 0.79 ± 0.09
387) (>63)
0.40 ± 0.01
(>126)
o
n.i.
–
(
See table 1 footnotes for definitions.
In most cases orthoꢀsubstitution was also tolerated, although associated with drops in potency.
This could reflect a reduction in conformational freedom imparted by the orthoꢀsubstituent, as
supported by the complete loss of activity in the nitrile 21Bo, bearing the most restrictive cyano
group. Interestingly, the salicylamide 21Bj retained activity, possibly due to masking of the
polarity of the phenolic hydroxyl by an intramolecular hydrogen bond.
ACS Paragon Plus Environment

Page 23 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
Given the apparent preference for an electronꢀwithdrawing substituent at the metaꢀposition, we
decided to investigate an expanded series of 3ꢀsubsituted benzamides (Table 3). Somewhat
surprisingly, a trifluoromethyl substituent (21Br) was deleterious to activity, being less potent
than the corresponding toluamide 21Bh. This probably reflects the slightly larger size of the
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
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
7
7
trifluoromethyl group (isosteric with an isopropropyl group by some measures) relative to
methyl. There was a complete loss of activity with the larger pentafluorosulfanyl substituent
(21Bs) and activity was also poor with the more polar methylsulfone 21Bt. The
ethynylbenzamide 21Bv was the most potent in this small series, and even the TMS derivative
21Bu retained activity. These results, along with those summarised in table 2, suggest that a
combination of steric and electronic factors determine the affinity of meta-substituted
benzamides, and hint at a narrow neck in the binding site that can conformably accommodate a
chlorine atom, and linear substituents like cyano and ethynyl, but not the slightly bulkier
trifluoromethyl group.
S
O
R
N
N
H
21B
Table 3. Additional 3ꢀsubstituted benzamides
IC µM (SI)
IC µM (SI)
50
50
#
21B
R
T. b. brucei
11.6 ± 2.1 (1.8)
n.i.
T. cruzi
r
s
CF
SF
SO
3
n.i.
n.i.
40 ± 0.9 (>1.2)
2.0 ± 0.3 (>12)
–
5
t
2
Me
8.5 ± 1.8 (4.9)
3.6 ± 0.9 (2.9)
0.76 ± 0.14 (27)
u
v
C≡CTMS
C≡CH
See table 1 footnotes for definitions.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 140
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The benzamides exhibited similar SAR against T. cruzi, although metaꢀsubstitution was slightly
less well tolerated than in T. b. brucei. The biggest difference between the two species is in the
activity of the paraꢀsubstituted benzamides: the toluamide 21Bi was very modestly active, but
the nitrile 21Bq was one of the more potent compounds tested. It is possible that the T. cruzi
target tolerates a broader range of conformations/poses of the bound molecules, allowing the
placement of small linear substituents, at both the meta and para-positions, into the narrow cleft
mentioned above.
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
Heteroaromatic amides 21H were also investigated (Table 4); against T. b. brucei, the
picolinamide 21Ha was equipotent with the benzamide 21f, and the nicotinamide 21Hb retained
significant activity; but the incorporation of an extra, diametricallyꢀopposed nitrogen, in pyrazine
1Hd, led to a significant loss of potency, perhaps suggesting amphiphilic character on opposite
2
sides of the acylꢀbinding pocket. The greatest drop in potency amongst these 6ꢀmembered Nꢀ
heterocyclic amides was observed in the 4ꢀpyridylamide 21Hc, again indicating an aversion to
hydrogenꢀbond acceptors at the 4ꢀposition.
ACS Paragon Plus Environment

Page 25 of 140
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
S
O
N
N
H
Ar
2
1H
Table 4. Heteroaromatic amides 21Ha–21Ho
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
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
a
IC µM (SI)
IC µM (SI )
IC µM (SI)
IC µM (SI)
50
50
50
50
#
21H
a
Ar
#21H
Ar
T. b. brucei
T. cruzi
T. b. brucei
0.35 ± 0.05
(240)
T. cruzi
N
0.3 ± 0.01
–
i
O
–
(
139)
.76 ± 0.28
110)
0
2.5 ± 0.05
(>19)
O
O
0.13 ± 0.0006
(666)
N
N
b
j
–
(
14 ± 1
3)
40 ± 4
(>1.2)
2.5 ± 0.1
(12)
5.0 ± 0.4
(>10)
c
k
(
Br
O
N
4.1 ± 0.1
20)
.35 ± 0.16
241)
.23 ± 0.03
179)
32 ± 7
(>1.5)
1.9 ± 0.3
(22)
O
d
e
f
l
–
(
N
S
0
0
0.40 ± 0.1
(>126)
0.74 ± 0.2
(113)
2.5 ± 0.3
(>19)
N
O
m
n
o
(
S
–
n.i.
n.i.
–
(
N
H
N
1
.6 ± 0.4
(13)
6.3 ± 0.1
(>7.9)
N
1.2 ± 0.0
(68)
g
h
S
O
0.80 ± 0.11
(77)
2.0 ± 0.1
(>25)
See table 1 footnotes for definitions.
5ꢀMembered heteroaromatic amides were then explored (Table 4). Unsurprisingly, the 2ꢀ and 3ꢀ
thiophenecarboxamides (21He and 21Hf, respectively) were approximately equipotent with the
isosteric benzamide 21f against T. b. brucei. The furans were also active, with the 3ꢀfuryl
derivative 21Hi being twice as potent as the 2ꢀfuryl 21Hh, reinforcing a preference for
electronegative/electronꢀwithdrawing substituents 3ꢀatoms from the amide carbonyl group. The
approximate halving in potency of the 2ꢀfuryl derivative 21Hh relative to the corresponding
thiophene 21He is likely due to inadequate filling of the acyl group binding pocket by the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 140
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
smaller furan, as the introduction of a 5ꢀmethyl substituent in 21Hj saw a significant
enhancement of activity. However, a 3ꢀmethyl as in 21Hl, or the larger bromo substituent in
21Hk, were not as well tolerated, suggestive of steric crowding in the region of these
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
substituents, and fitting with a putative hydrophobic pocket of limited depth.
The introduction of additional heteroatoms into the acyl heteroaromatic substituent led to a drop
in potency in the thiazole 21Ho and a complete loss of activity in the more polar oxazole 21Hn.
In contrast, the 5ꢀisoxazole 21Hm, with electronegative atoms in the 2ꢀ and 3ꢀpositions relative
to the amide, was only slightly less active than the benzamide 21f, once again indicating a
tolerance of, or preference for, electronegative atoms at these positions.
With the exception of the thiophenes, most of the polar heteroaromatic amides tested were
substantially less potent against T. cruzi, perhaps suggesting less tolerance to polarity in the acylꢀ
binding pocket of the biological target in this species. Alternatively, since in the T. cruzi assay
the parasites are within mammalian host cells, the reduced potency of the heteroaromatic amides
could be explained by poor uptake through the host cell membrane.
A series of ureas and carbamates has also been investigated (Table 5). The synthetic
intermediate – tert-butyl carbamate 12 – showed no activity against T. b. brucei at 10 µM, which
was not unexpected given the lack of tolerance for large acyl substituents observed earlier.
Indeed, the slightly smaller isopropyl 29a and cyclopentyl carbamates 29b had potencies similar
to the most active amides, against both species. In contrast, to the carbamate 29a, the Nꢀ
isopropylurea 30a was significantly less active, particularly against T. cruzi. We tentatively
ACS Paragon Plus Environment

Page 27 of 140
Journal of Medicinal Chemistry
1
2
3
attributed this drop in potency to the additional hydrogen bond donor (NH), which presumably is
well solvated and possibly not engaged in a target binding interaction. To examine this
hypothesis, we tested the N,Nꢀdialkylureas 30b and c and, indeed, activity was restored close to
that of the carbamate 29a in T. b. brucei, but not T. cruzi. The diisopropyl derivative 30d, on the
other hand, is clearly too large. Exocyclic ureas proved to be the most potent of all compounds
examined in this study, with the piperidine 23b being optimal, amongst the 5ꢀ (23a), 6ꢀ (23b)
and 7ꢀmembered (30e) rings, against both species. As before, the introduction of an oxygen at
the 4ꢀposition in the morpholineꢀderived urea 23c led to a dramatic loss of activity, especially
against T. cruzi, supporting the hypothesis that the acylꢀbinding pocket in the biological target of
this species is particularly hydrophobic. The enhanced potency of the ureas relative to the
carbamates and amides could possibly be explained by enhanced hydrogenꢀbond donating ability
of the carbonyl group in the former, although subtle steric/geometric effects due to the increased
planarisation of the urea nitrogen cannot be ruled out at this time. The dramatically reduced
potency of the pyrroleꢀderived urea 30f is somewhat surprising, and could reflect a lack of
hydrolytic stability under the reaction conditions. Alternatively (or in addition), this urea has
comparable potency to the electronically analogous 5ꢀmembered heteroaromatic amides; the
poor electronꢀdonating ability of the heterocyclic nitrogen may be responsible for the drop in
potency of 30f relative to the other ureas.
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
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

Journal of Medicinal Chemistry
Page 28 of 140
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
S
O
N
N
H
R
Table 5. Carbamates and ureas
IC µM
IC µM
IC µM
IC µM
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
50
50
50
50
#
R
(SI)
(SI)
#
R
(SI)
(SI)
T. b. brucei
T. cruzi
T. b. brucei
T. cruzi
9
.6 ± 3.5
10)
32 ± 3
(>1.5)
1
2
n.i.
–
30d
N
O
O
(
0
.22 ± 0.02
(385)
0.40 ± 0.009
(>125)
0.050 ± 0.040
(1735)
0.32 ± 0.007
(>158)
N
N
N
N
2
9a
9b
23a
23b
23c
30e
30f
0
.30 ± 0.004
(280)
0.40 ± 0.009
(>125)
0.028 ± 0.001
(2966)
0.039 ± 0.003
(>1258)
2
O
O
0
.97 ± 0.05
86)
.90 ± 0.03
93)
.39 ± 0.09
212)
.26 ± 0.03
320)
See table 1 footnotes for definitions.
3.2 ± 0.5
(>15)
0.69 ± 0.22
(127)
3.2 ± 0.07
(>15)
2
9c
(
O
0
1.6 ± 0
0.091 ± 0.021
(943)
0.12 ± 0.003
(>398)
3
0a
0b
N
H
(
(>31)
0
0
1.0 ± 0.3
(>50)
1.1 ± 0.07
(38)
N
3
–
N
N
(
1.0 ± 0
3
0c
(
(>50)
The benzenesulfonamide analogue 25 (Scheme 4) was also prepared and evaluated, but
demonstrated no activity against T. b. brucei at 10 µM and only weak activity against T. cruzi
(IC = 20 µM). We hypothesized that this could be due to lengthening the molecule as a result
5
0
of incorporation of the large sulfur atom. To examine this possibility, and to better mimic the
ureas, the slightly smaller sulfamide 27 (Scheme 4) was prepared and tested, but this was also
inactive against T. b. brucei, and less potent than the sulfonamide against T. cruzi (IC = 32
5
0
µM). These results suggest that it is the tetrahedral geometry of the sulfonamide and sulfamide
functional groups that results in loss of activity.
ACS Paragon Plus Environment

Page 29 of 140
Journal of Medicinal Chemistry
1
2
3
SAR concerning the thiazoleꢀ2ꢀsubsituent
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
Initial efforts in this area began in parallel with the optimisation of the acyl substituent, hence the
first compounds targeted were cyclopentanecarboxamides and benzamides (Table 6). In contrast
to the acyl modifications, replacement of the left hand side (LHS) phenyl substituent with 2ꢀ
thienyl (45a/b) led to a substantial drop in potency. Pyridyl substituents (35a–f) were also
deleterious, especially against T. cruzi.
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
S
O
Ar
N
N
H
R
Table 6. Benzamides and cyclopentanecarboxamides bearing heterocycles at the thiazole 2ꢀ
position. (The corresponding 2ꢀphenylthiazoles are included for comparison)
IC µM (SI)
IC µM (SI)
50
50
#
Ar
R
T. b. brucei
T. cruzi
2
1f
Ph
Ph
S
Ph
0.25 ± 0.01 (334)
0.40 ± 0.04 (>125)
2
4
4
3
1h
0.25 ± 0.02 (329)
1.5 ± 0.2 (40)
2.4 ± 0.3 (35)
1.1 ± 0.3 (74)
–
–
5a
5b
5a
Ph
Ph
S
N
10 ± 1.4 (>5)
6.3 ± 0.7 (>7.9)
N
3
5b
4.0 ± 1.5 (21)
–
N
N
35c
5d
Ph
Ph
2.7 ± 0.3 (31)
6.6 ± 0.3 (13)
3.1 ± 0.9 (27)
9.1 ± 1.7 (9)
40 ± 3.7 (>1.2)
20 ± 0.9 (>2.5)
32 ± 2 (>1.5)
32 ± 13 (>1.5)
3
35e
N
3
5f
N
See table 1 footnotes for definitions.
ACS Paragon Plus Environment