Repositioning Antitubercular 6-Nitro-2,3-dihydroimidazo[2,1-b][1,3]oxazoles for Neglected Tropical Diseases: Structure-Activity Studies on a Preclinical Candidate for Visceral Leishmaniasis

Article abstract of DOI:10.1021/acs.jmedchem.5b01699

6-Nitro-2,3-dihydroimidazo[2,1-b][1,3]oxazole derivatives were initially studied for tuberculosis within a backup program for the clinical trial agent pretomanid (PA-824). Phenotypic screening of representative examples against kinetoplastid diseases unexpectedly led to the identification of DNDI-VL-2098 as a potential first-in-class drug candidate for visceral leishmaniasis (VL). Additional work was then conducted to delineate its essential structural features, aiming to improve solubility and safety without compromising activity against VL. While the 4-nitroimidazole portion was specifically required, several modifications to the aryloxy side chain were well-tolerated e.g., exchange of the linking oxygen for nitrogen (or piperazine), biaryl extension, and replacement of phenyl rings by pyridine. Several less lipophilic analogues displayed improved aqueous solubility, particularly at low pH, although stability toward liver microsomes was highly variable. Upon evaluation in a mouse model of acute Leishmania donovani infection, one phenylpyridine derivative (37) stood out, providing efficacy surpassing that of the original preclinical lead.

Full text of DOI:10.1021/acs.jmedchem.5b01699

Subscriber access provided by GAZI UNIV
Article
Repositioning antitubercular 6-nitro-2,3-dihydroimidazo[2,1-b]
[1,3]oxazoles for neglected tropical diseases: structure-activity
studies on a preclinical candidate for visceral leishmaniasis
Andrew M. Thompson, Patrick D. O'Connor, Adrian Blaser, Vanessa Yardley,
Louis Maes, Suman Gupta, Delphine Launay, Denis Martin, Scott G.
Franzblau, Baojie Wan, Yuehong Wang, Zhenkun Ma, and William A. Denny
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01699 • Publication Date (Web): 22 Feb 2016
Downloaded from http://pubs.acs.org on February 23, 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 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Repositioning antitubercular 6-nitro-2,3-
dihydroimidazo[2,1-b][1,3]oxazoles for neglected
tropical diseases: structure-activity studies on a
preclinical candidate for visceral leishmaniasis
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
,*
1
1
Andrew M. Thompson, Patrick D. O’Connor, Adrian Blaser, Vanessa Yardley, Louis
2
3
Maes, Suman Gupta, Delphine Launay, Denis Martin, Scott G. Franzblau, Baojie Wan,
4
5
5
6
6
6
7
Yuehong Wang, Zhenkun Ma, and William A. Denny
1
1
Auckland Cancer Society Research Centre, School of Medical Sciences, The University of
2
Auckland, Private Bag 92019, Auckland 1142, New Zealand; Faculty of Infectious &
Tropical Diseases, London School of Hygiene & Tropical Medicine, Keppel Street, London
3
WC1E 7HT, United Kingdom; Laboratory for Microbiology, Parasitology and Hygiene,
Faculty of Pharmaceutical, Biomedical and Veterinary Sciences, University of Antwerp,
4
Universiteitsplein 1, B-2610 Antwerp, Belgium; Division of Parasitology, CSIR-Central
5
Drug Research Institute, Lucknow 226031, India; Drugs for Neglected Diseases initiative,
6
5 Chemin Louis Dunant, 1202 Geneva, Switzerland; Institute for Tuberculosis Research,
1
College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago,
7
Illinois 60612, USA; Global Alliance for TB Drug Development, 40 Wall St, New York
1
0005, USA
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 72
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
Abstract
6-Nitro-2,3-dihydroimidazo[2,1-b][1,3]oxazole derivatives were initially studied for
tuberculosis within a backup program for the clinical trial agent pretomanid (PA-824).
Phenotypic screening of representative examples against kinetoplastid diseases unexpectedly
led to the identification of DNDI-VL-2098 as a potential first-in-class drug candidate for
visceral leishmaniasis (VL). Additional work was then conducted to delineate its essential
structural features, aiming to improve solubility and safety without compromising activity
against VL. While the 4-nitroimidazole portion was specifically required, several
modifications to the aryloxy side chain were well tolerated e.g., exchange of the linking
oxygen for nitrogen (or piperazine), biaryl extension, and replacement of phenyl rings by
pyridine. Several less lipophilic analogues displayed improved aqueous solubility,
particularly at low pH, although stability towards liver microsomes was highly variable.
Upon evaluation in a mouse model of acute Leishmania donovani infection, one
phenylpyridine derivative (37) stood out, providing efficacy surpassing that of the original
preclinical lead.
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
ACS Paragon Plus Environment

Page 3 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Neglected tropical diseases (NTDs) affect in excess of one billion people, predominantly in
the most impoverished areas of the world, causing more than half a million deaths each year,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
,2
as well as much physical suffering and social stigma. Of the 17 major NTDs, the
kinetoplastid parasite diseases visceral leishmaniasis (VL), Chagas disease, and human
African trypanosomiasis (HAT) are considered amongst the most challenging, due to their
high mortality rates and limited treatment options, with little economic incentive for this to
2
change. VL is caused by Leishmania donovani (L. don) and Leishmania infantum (L. inf),
transmitted by female sand flies; these parasites first multiply inside host macrophages before
3,4
spreading to and destroying other tissues (e.g., spleen, liver, and bone marrow). Annual
incidence of VL is now estimated at 300000, with infection occurring mainly in rural areas of
East Africa, India, Bangladesh and Brazil. Most of the ~35000 deaths each year pass
5
unrecognised, and in East Africa, lethal epidemics take place frequently. However, the
3
,4
current treatments for VL are unsatisfactory. Pentavalent antimonials have been the
standard first-line remedy since the 1940s, but these are cardiotoxic and require painful
parenteral administration over 1 month; drug resistance has now severely limited their utility
in India. A less toxic liposomal form of the antifungal agent, amphotericin B (1; see Figure
1
), has proven widely effective but is limited by high cost. The cheaper aminoglycoside
antibiotic, paromomycin (2), has variable efficacy, both within and between regions. Finally,
the alkyl phospholipid miltefosine (3), a long-acting anticancer drug repurposed for VL, is
the sole oral agent; nevertheless, this also suffers from several disadvantages, including cost,
teratogenicity, and treatment failures. Hence, there remains an urgent need for more effective,
safe, and affordable oral treatments for VL. In the last decade of tuberculosis (TB) and NTD
drug discovery, there has been major interest in exploiting potential development shortcuts,
via the repurposing of approved drugs, rescue of “failed” clinical agents, or repositioning of
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 72
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
6
early or late stage drug candidates. Indeed, for organizations such as the Global Alliance
-8
for TB Drug Development (TB Alliance) and the Drugs for Neglected Diseases initiative
(DNDi), phenotypic screening is still viewed as a highly useful, cost-effective strategy to
9
-11
identify “low hanging fruit”.
Recent cooperation between these two organizations in this
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
12
regard actually provided the foundation for this current study.
13
In early collaborative work with TB Alliance (2005-6) we investigated several
nitroheterobicyclic ring analogues of the bioreductive TB agent pretomanid (PA-824, 4), now
1
4
in phase III clinical evaluation. However, our efforts in the known 6-nitroimidazooxazole
15
1
class were curbed following disclosure of the recently approved MDR-TB drug delamanid
6
17,18
(OPC-67683, 5),
turning our focus for a backup toward the efficacious biaryl analogues
Both 4 and 5 are activated by an intracellular deazaflavin dependent nitroreductase
Ddn) in Mycobacterium tuberculosis (M. tb), resulting in inhibition of cell wall synthesis as
1
9-21
of 4.
(
22
well as respiratory poisoning (via nitric oxide release). This mode of action is distinct from
the postulated nucleophilic nitro reduction (e.g., by thiolate) of other new nitroaromatic TB
agents, such as PBTZ169 (6) and nitrobenzamides, to form reactive nitroso intermediates,
23,24
which trigger suicide inhibition of an enzyme (DprE1) involved in cell wall biosynthesis.
In late 2009, follow-up interest from DNDi in some promising antileishmanial screening
data for several of our earlier nitroimidazole derivatives (above) then led to our nomination of
three unoptimised racemic hits, including two from the most potent 6-nitroimidazooxazole
class (7 and 8), for proof-of-concept in vivo assessment in a mouse model of VL. In parallel,
taking into account the preliminary SAR findings, a further 72 compounds were screened
against L. don in a macrophage-based luciferase assay at the Central Drug Research Institute
2
5
(CDRI, India), and 10 of these were later tested in the same mouse model. However, the
outstanding in vivo efficacy of 7 was not surpassed by any of the other candidates, and
following our synthesis of its enantiomers and ensuing head-to-head assessments of these, the
4
ACS Paragon Plus Environment

Page 5 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
R enantiomer 9 (DNDI-VL-2098) was chosen for preclinical evaluation as a potential first-
8
2
5,26
in-class drug candidate for VL.
In this report, we initially detail results relevant to the
selection of 9. We then describe the findings of more recent work aimed at discovering
backups to 9 having an improved physicochemical/ pharmacological profile and better safety,
providing wider SAR conclusions for the nitroimidazooxazole class against TB and three
kinetoplastid NTDs, together with a first in vivo appraisal of the best new leads for VL.
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
CHEMISTRY
15,27
Several earlier investigators
prepared 6-nitro-2,3-dihydroimidazo[2,1-b][1,3]oxazoles
(“oxazoles”) by condensing epoxides with the explosive intermediate 2,4-dinitroimidazole. In
the interests of improving safety, we first sought alternative conditions using 2-bromo-4-
1
nitroimidazole (75; see Scheme 1A/1B). Satisfactory results for both 2-H and 2-methyl
3
glycidyl ethers were obtained by neat reaction with 75 in the presence of a suitable base
(DIPEA, 105-108 °C) and the resulting (uncyclised) alcohol intermediates were readily
transformed into the desired oxazoles 7 and 10-12 in high yield (85-90%) upon treatment
with sodium hydride (DMF, 0 °C). Minor quantities (1-3%) of 5-nitroimidazooxazoles were
also generated in the first step, with two examples (15 and 87) being isolated for comparison.
28
The requisite epoxides (74, 78, 82, and 86) were acquired via standard methods, starting
3
with (4-OCF ) benzylations of glycidol (73) or 2-methylprop-2-en-1-ol (77), or alkylations of
4-(trifluoromethoxy)phenol with epibromohydrin (81) or 3-chloro-2-methylprop-1-ene (84).
To synthesise the enantiomers of 12 and 15 (13, 14, 16, and 17), we employed cesium
29
fluoride promoted reactions between 4-(trifluoromethoxy)phenol and the chiral glycidyl
nosylates 89 and 93 (Scheme 1C). The ether products (90 and 94) were then alternatively
coupled with the more accessible 2-chloro-4-nitroimidazole (91), where the use of toluene as
cosolvent (DIPEA, 80 °C) enabled better isolated yields of the more thermally labile 5-nitro
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compounds 16 and 17 (15-16%). In the final step, sodium hydride induced cyclisation of the
intermediate alcohols (92 and 95) was completed by warming to 20 °C, giving both oxazoles
13 and 14 in high yield (86-87%) and excellent ee (100%).
18,30,31
Reaction of the known
chiral epoxides 96 and 97 with 4-(trifluoromethoxy)phenol or
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
4'-fluorobiphenyl-4-ol in the presence of sodium hydride (Scheme 2A) also provided a direct
18
route to the enantiomers of early VL leads 7 and 8, as recorded for 9. These epoxides were
conveniently obtained by the reaction of 2-chloro-4-nitroimidazole (91) with (R) or (S)-2-
methylglycidyl 4-nitrobenzenesulfonate (derived from the Sharpless epoxidation of 2-
31
18
methylprop-2-en-1-ol, 77), although a multistep approach based on the 4-nitrobenzoate
ester derivative of (S)-2-methylglycidol reportedly achieved a slightly better ee for 96 (98.8%
30
vs. 95.9%). While isolated yields of the final products 9, 18, 44, and 45 were moderate (33-
44%), these important entities were rapidly produced in excellent chiral purities (>98% ee),
32
and a superior kilogram scale process synthesis of 9 has recently been developed.
33
For more efficient generation of additional analogues of 7, we then turned to the known
racemic epoxide 98 (Scheme 2B). Heating this epoxide with various phenols and sodium
hydride furnished the oxazole products 19-23 directly, albeit in variable yields (12-45%),
depending on the conditions. Alternatively, reaction of 98 with 4-(trifluoromethoxy)aniline
3
4
or its N-methyl derivative) in the presence of anhydrous cobaltous chloride gave
(
uncyclised β-anilino alcohols (99 and 100), which could be transformed into the novel
oxazoles 24 and 25 by careful treatment with sodium hydride (1.3-1.4 equiv).
30
Next, in order to access larger quantities of the known alcohol 26 for the preparation of
30
heteroaryl ethers, we devised an improved plan, comprising TIPS monoprotection of known
diol 102, ring closure, and desilylation under acidic conditions (Scheme 2C). Diol 102 was
obtained from 2-chloro-4-nitroimidazole (91) in two steps (85%), via dihydroxylation of the
new alkene 101, and elaborated to the oxazole 104 as planned (92% from 102; two steps).
6
ACS Paragon Plus Environment

Page 7 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
However, attempted removal of the TIPS group using a reported acidic method (1% HCl in
3
5% EtOH) required prolonged heating at 45 °C (4.5 days) and was not clean, leading to a
5
9
suboptimal yield of alcohol 26 (53%). Fortunately, treatment of 104 with hydrofluoric acid in
36
acetonitrile gave a much better result (95% yield), and 26 was easily converted into various
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
heteroaryl ether targets (27, 29-31) via sodium hydride-catalysed S Ar reactions on the
N
3
7
haloheterocycles 105-108. The isomeric pyridine 28 was also obtained in moderate yield
30
(25%) by reaction of 6-(trifluoromethyl)pyridin-3-ol and sodium hydride with the known
epoxide 109 (Scheme 2D). By way of comparison, epoxides 98 and 109 each afforded
comparable yields (55% and 51%) for scale-up of racemic VL lead 7.
In the 2H oxazole series, the synthesis of biphenyl analogues (32-34) was achieved in a
straightforward manner (Scheme 3A), starting from epibromohydrin (81), and using
methodology described above (except that DMF was employed in the final step to improve
solubility). We then envisaged the assembly of related phenylpyridine side chains via Suzuki
couplings on bromopyridinyl ether derivatives (e.g., 119 in Scheme 3B). To this end, the
2
7
known alcohol 117 was prepared in three steps from TBS-protected glycidol (113), via
condensation with 91, ring closure of alcohol 115, and acid-catalysed desilylation (Scheme
3B). Unfortunately, reaction of 117 with 5-bromo-2-fluoropyridine (118) under the optimised
conditions developed for the 2-methyl congener 144 (Scheme 4B) gave a poor yield of ether
119 (9%), and attempted Suzuki couplings on this substrate (even with weak bases) resulted
only in decomposition. In an attempt to obtain at least one or two examples featuring this
isomer pattern, we next considered reversing the order of these final two steps. Here, the 5-
aryl-2-fluoropyridines 120 and 121 were quantitatively obtained from bromide 118 but the
final alkylations involving alcohol 117 gave very poor yields (<1%) of the authentic products
35 and 36, precluding synthesis of the less electron deficient 4-fluorophenyl analogue.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 72
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
To circumvent any similar complications in the preparation of the 3-pyridyl targets (37, 40
and 41), a modified strategy was adopted (Scheme 3C). Here, 6-bromopyridin-3-ol (122) was
first protected (as the ethoxymethyl ether, 123) and then Suzuki-coupled with arylboronic
3
8
acids. Following acidic deprotection, the 6-arylpyridin-3-ols (125, 129, and 133) were
alkylated with epibromohydrin (81) and the resulting epoxides were then elaborated to the
final products (37, 40 and 41), as above, in consistently good yields (56-59% over the last 2
steps). For the enantiomers of 37 (38 and 39), we varied this route by reacting the (NaH-
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
3
generated) anion of arylpyridinol 125 with the chiral glycidyl nosylates 89 and 93 (Scheme
8
3D). The resulting ether products (136 and 138) were again coupled with 2-chloro-4-
nitroimidazole (91), using toluene as cosolvent (DIPEA, 80 °C) in order to obtain better
quantities of the 5-nitro isomers 42 and 43 (10-12% yield). Alcohols 137 and 139 were then
ring closed to the oxazoles 38 and 39 in excellent yields (87-89%), as above.
Methods for the preparation of biaryl analogues and some soluble bioisosteres in the 2-
3
9
methyl oxazole series are outlined in Scheme 4. Epoxide 141 (sourced directly from the
commercial chloride 140) was converted in two steps to the 4-iodophenoxy precursor 143,
with Suzuki couplings then delivering the biphenyl analogues 8, 46, and 47 (Scheme 4A).
The bromopyridinyl ethers 144 and 147 (obtained from alcohol 26, and epoxide 98,
respectively) also proved to be effective Suzuki substrates for creating phenylpyridine and
bipyridine derivatives (48-56; Scheme 4B/4C), although for bipyridines 55 and 56 it was best
to use a weaker base (KHCO ). Furthermore, in the case of bipyridine 54, it was necessary to
3
40
employ a copper(I) additive, in order to circumvent facile protodeboronation of the required
2-pyridylpinacol boronate 145 (Scheme 4B). Arylated cyclic amine targets (57 and 58) were
readily achieved by reaction of epoxide 109 with the commercial amines (148 and 150),
followed by (NaH-induced) ring closure of the resulting alcohols (149 and 151; Scheme 4D).
8
ACS Paragon Plus Environment

Page 9 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Lastly, the O-carbamate 59 was made in high yield (86%) by chloroformylation of alcohol 26
4
1
and one-pot coupling with arylpiperazine 148 (Scheme 4E).
To conclude this SAR study, several heterocyclic ring A analogues of 7 were synthesized
from epoxide 86 (Scheme 5). Thus, similar condensations (DIPEA, 104-105 °C) of 86 with 2-
nitroimidazole (152), 5-bromo-3-nitro-1,2,4-triazole (157), or 3,5-dinitropyrazole (158) each
provided heterobicyclic products directly, although 152 gave equivalent yields of oxazole 61
and its barely separable alcohol precursor 153 (Scheme 5A). Treatment of the latter with
sodium hydride (DMF, 0-20 °C) readily cyclised this to 61, in agreement with a recent
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
42
report. In the case of 5-bromo-3-nitro-1,2,4-triazole (157), a separable mixture of nitro- and
bromo-substituted products (70 and 71) was obtained in a ratio of 3:2 (Scheme 5D),
consistent with initial alkylation at the 1- or 2-positions of the triazole, followed by
intramolecular substitution of the adjacent bromine or nitro group by the intermediate
43
alcohol. Reaction of epoxide 86 with 3,5-dinitropyrazole (158) yielded mainly oxazole 72
(66%), together with a small amount of alcohol 159 (Scheme 5E).
Finally, alternative condensation of epoxide 86 with 2,4-dibromoimidazole (154) (DIPEA,
toluene, 109 °C) produced a resolvable mixture of two uncyclised alcohols (155 and 156, in a
2.4:1 ratio; Scheme 5B), which were transformed into the isomeric oxazoles 62 and 67 upon
mild heating of their preformed anions (NaH, DMF, 45-55 °C). Bromide 62 was converted
into the desired methyl sulfone 66 (regarded as a potential bioisostere of 7) via lithiation and
quenching with methyl disulfide (to give thioether 65), followed by oxidation (m-CPBA),
after an attempted copper(I) iodide/
L
-proline sodium salt-induced coupling of 62 with sodium
methanesulfinate proved unsatisfactory (7% yield of crude 66). Similar lithiation of 62 and
quenching with DMF also gave the aldehyde 63, which was easily reduced (NaBH ) to
alcohol 64. Methyl sulfone 66 had earlier been sought via displacement of the nitro group in
4
4
4
7
by methanethiol, based on previous observations. Here, treatment of 7 with methanethiol
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and triethylamine (10 equiv in MeOH, 0-10 °C) furnished only one major product, 68, which
45-47
upon oxidation gave sulfone 69 (Scheme 5C). A survey of the literature
that cine- or ipso-substitution was possible; therefore, we studied compounds 7, 61, 62, 65,
7-69, and a 5-nitro intermediate (114) by 2D NMR experiments (e.g., NOESY, ROESY,
had suggested
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
13
HSQC, and HMBC). The NOE evidence and C NMR chemical shift data unambiguously
established the assigned structures (see Supporting Information), although an absence of any
three bond correlations between H-5 and C-3 in the HMBC experiments (when four bond
correlations between H-6 and C-3 were apparent) was initially misleading. Weakly observed
n
or missing long range correlations (due to the wide range of J spin coupling constants, 1 to
CH
4
50 Hz) constitute a well-known deficiency of the standard HMBC experiment, but recent
8
4
HMBC variants also possess some disadvantages (e.g., reduced signal to noise). Thus, it
9
47
was verified that thioether 68 was the product of cine-substitution, in a fashion reminiscent
of the Ddn-catalysed addition of hydride to the unsubstituted imidazole carbon atoms of 4
22,23
and 5, leading to loss of the adjacent nitro group (and production of nitric oxide in M. tb).
RESULTS AND DISCUSSION
Tables 1 and 2 present in vitro antiparasitic and antitubercular data for 67 compounds (6
known) from mainly the nitroimidazooxazole class, prepared in two collaborative projects.
For the TB studies, compounds were screened in two M. tb (strain H37Rv) growth inhibition
assays, MABA and LORA, respectively conducted under either aerobic (replicating) or
5
0,51
hypoxic (non-replicating) conditions, as described.
Given that both 4 and 5 exhibit at least
22
two distinct mechanisms of action against TB, depending on the oxygenation environment,
the LORA assay was regarded as a useful starting tool to identify analogues that may be more
5
1,52
efficient at killing persistent subpopulations of M. tb in vivo.
Recorded MIC values (for a
≥90% growth inhibitory effect) are the mean of at least two independent determinations.
1
0
ACS Paragon Plus Environment

Page 11 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
From 2010-2012, in collaboration with DNDi, new analogues of lead oxazoles 7 and 8
25
were screened against L. don in a macrophage-based luciferase assay at CDRI, albeit only
single determinations of the IC50 values were obtained, limiting reliable SAR interpretation.
Therefore, a wider set of compounds was later evaluated in at least duplicate in vitro assays
against three tropical protozoan parasites (L. inf, T. cruzi, and T. brucei) at the University 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
53
Antwerp (LMPH). Here, the L. inf assay employed primary peritoneal mouse macrophages
as the host cell for the amastigotes, whereas the T. cruzi assay utilised human lung fibroblasts
(
MRC-5 cells); compounds having mean IC50 values below 1 µM were considered to be
highly active. Cytotoxicity on MRC-5 cells was also measured and almost all of the 67
compounds tested could be regarded as non-toxic (IC s >30 µM, most >64 µM), in broad
5
0
50
agreement with assessments on VERO cells (data not shown).
In the design of new analogues of 7 and 8, two strategies were employed to enhance
20,21
aqueous solubility while minimising the risk of major potency loss.
First, we focused on
reducing lipophilicity (probed using CLogP predictions derived from ACD LogP/LogD
software, version 12.0; Advanced Chemistry Development Inc., Toronto, Canada) via the
replacement of benzene rings by pyridine, or closely related heterocycles (viz. pyrimidine,
pyrazine, and pyridazine). This approach enabled CLogP reductions of up to 2.3 units
(
typically ~1.1 units for a single pyridine substitution, although the predicted data displayed
some variation between isomers). Second, we evaluated the removal of full aromaticity from
the side chain by the incorporation of a saturated heterocycle (e.g., piperazine or piperidine)
in combination with a terminal aromatic ring. Arylated cyclic amines are known bioisosteres
2
for biaryl moieties and can provide improved solubility via salt formation, as well as by the
1
54
disruption of molecular planarity and symmetry. Kinetic aqueous solubility values at pH=7
were determined for selected active compounds, using dry powder forms, while analogous
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 72
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
data at pH=1 were measured for examples containing basic functionalities that could form
a
hydrochloride salts (based on calculated pK estimates derived from similar ACD software).
1. Initial studies relevant to preclinical candidate selection
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 investigations began in late 2009 with the proof-of-concept in vivo assessment of
oxazoles 7 and 8 in a mouse model of VL at the London School of Hygiene and Tropical
Medicine (LSHTM). The selection of these two candidates was based on a combination of
their promising activities in an L. don-infected macrophage assay conducted at the Swiss
Tropical Institute (later confirmed at CDRI) and reasonable conjecture for their better
metabolic stabilities in comparison to other similarly active analogues (e.g., 11). In the event,
both 7 and 8 provided outstanding efficacy in this L. don model at 50 mg/kg (99.9 to 100%
reduction of the liver parasite burden, after oral dosing once daily for 5 days) and subsequent
dose-response experiments (Table 3) established that 7 was the most effective lead (ED50
s
were 3.0 and 4.8 mg/kg for 7 and 8, respectively). This result was consistent with 7 having
superior mouse microsomal stability (79% vs. 57% parent remaining after 30 min, Table 3),
better aqueous solubility (16-fold), and a more favourable mouse pharmacokinetic profile
(lower clearance, greater exposure, and higher oral bioavailability; Table 4 and Supporting
Information, Figure S1) in comparison to 8.
The more stringent chronic infection hamster model of VL is considered the gold standard
5
5
in vivo assay due to its similar pathology to human disease. Here, 5 and 10 day assessments
of 7 and 8 at CDRI led to opposing conclusions on which oxazole was better (Supporting
Information, Table S3); therefore, the enantiomers of both compounds were prepared. In the
mouse model, the R enantiomer of 7 (9) was greatly superior to the S enantiomer (18) and
was slightly better than 7 itself (83%, 8%, and 64% inhibition at 3.13 mg/kg, respectively;
Table 3). This was in good agreement with the microsomal stability results (Table 3) and L.
inf IC50 data (Table 1) showing a 4.5-fold potency difference between 9 and 18. In contrast,
1
2
ACS Paragon Plus Environment

Page 13 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
although the R enantiomer of 8 (44) was also better than the S enantiomer (45) in this model,
it was less active than 8 itself (36% vs. 68% at 6.25 mg/kg; Table 3), indicating reduced oral
exposure (44 and 45 gave sparing aqueous solubilities of ~0.07 µg/mL). A comparative
appraisal of 7, 9, and 18 in the CDRI hamster model (Supporting Information, Table S3)
confirmed the preeminent activity of the R enantiomer (9) over the S form (18) and 7 itself, at
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
25
all dose levels. Furthermore, dose-response evaluations of 9 and the TB drug delamanid (5)
in a similar L. inf infected hamster model at LMPH also corroborated the excellent in vivo
efficacy of 9 (>99% inhibition in 3 target organs at 25 mg/kg; Table 5), whereas 5 displayed a
weak effect (22-66% at 50 mg/kg), consistent with its poor in vitro potency (IC 7.1 µM). In
50
view of these findings, the decision was made to proceed with 9 as a lead candidate for VL.
While the initial in vivo work was in progress, the oxazole analogues in hand (7, 8, 10-12,
46, and 47) were screened against L. don in the macrophage-based luciferase assay at
2
5
CDRI. Based on these results (Table 1) and some encouraging microsomal stability data
(Table 3), two additional compounds were identified as worthy of investigation in the VL
mouse model. These were the (6-fold) more potent des-methyl analogue of 7 (12: IC 0.005
5
0
µM) and the 4-trifluoromethoxy analogue of 8 (47: IC50 0.05 µM). However,
1
9
counterintuitively to our previous work, 47 was markedly inferior to 8 for VL (only 42%
inhibition at 25 mg/kg), while the more potent lead 12 also displayed slightly lower efficacy
than 7 (89% inhibition at 6.25 mg/kg; Table 3). To confirm the latter result, we examined the
enantiomers of 12 (13 and 14), but surprisingly, while the R form (13) was preferred in vitro
(Table 1), both stereoisomers had an equivalent effect in vivo (~75% at 6.25 mg/kg; Table 3).
Examination of the pharmacokinetic properties of selected compounds afforded further
insight (Table 4), revealing that 12 had a higher rate of clearance than 7, gave a lower
exposure level, and had a concomitantly reduced oral bioavailability (52% vs. 79%), likely
due in part to its 4-fold lower aqueous solubility (Table 3). Furthermore, while 47 did show
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 72
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
better metabolic stability than 8 (decreased clearance and longer half-life; Table 4), it also
suffered from poor oral absorption (Cmax 0.29 µg/mL; see Supporting Information, Figure
S1), resulting in modest oral bioavailability (18% cf. 63% for 8). Again, this was consistent
with the 11-fold lower aqueous solubility of 47 compared to 8 (Table 3) and suggests the
need for stronger oral formulations when evaluating such sparingly soluble compounds.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
Thus, the early selection of 9 for advanced development was validated by the discovery of
inferior in vivo results for the few alternative candidates that were assessed at the time.
Following the demonstration of suitable in vivo efficacy and in vivo pharmacokinetics, drug
56
discovery for leishmaniasis typically requires several additional ADME and safety studies.
For example, many nitroheterocyclic compounds provide positive results in the Ames test for
mutagenicity; indeed, the development of an early 6-nitroimidazooxazole lead for TB (CGI-
15
7341) was halted following such a discovery. Extensive investigation by Otsuka
1
Pharmaceutical Co. established that the introduction of heteroatoms and larger side chain
1
8,57
26
and both 8 and 9 were Ames negative. Other
functionality (e.g., 5) overcame this effect
safety factors (e.g., hERG, CYP inhibition, potential for drug-drug interactions) were also
26
broadly acceptable for 9, which additionally satisfied such criteria as high stability, low cost
of synthesis and suitability for once-per-day oral administration noted in the proposed target
5
product profile for a VL drug. Oral treatment for VL offers many important advantages,
8
including convenience and improved patient compliance; it also enables direct delivery of the
2
6
absorbed drug to the liver, a major parasite population site. Therefore, the repositioning of 9
for VL may represent a significant medical breakthrough for this highly neglected disease.
2. SAR of nitroimidazooxazoles for VL
In an attempt to better understand the SAR for VL (with a view to further improving drug-
like features, e.g., solubility and safety), additional oxazoles were prepared and screened, first
at CDRI, and then at LMPH. The latter testing confirmed (Table 1) that this class had no
1
4
ACS Paragon Plus Environment

Page 15 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
obvious potential for treating HAT (T. brucei IC50s >1 µM, most >64 µM), nor any clear
utility for Chagas disease (T. cruzi IC50s mostly 1-3 µM, and no new leads were better than 5
or 8), so the major focus on VL was justified. Nevertheless, in light of the recently reported
59
efficacy of fexinidazole (a 5-nitroimidazole derivative) in a mouse model of VL, one
fundamental question was whether 5- or 6-nitroimidazooxazoles were preferred. However, an
inspection of the data for 15-17 (5-nitro isomers of 12-14 above) immediately showed that
the 5-nitro analogues were more than two orders of magnitude less effective against L. inf.
Having established this, compounds 19-23 initially examined the influence of the phenyl
ring substituent. Interestingly, while the single determination CDRI data (L. don) appeared to
indicate a clear advantage for 4-trifluoromethoxy (7), this was not supported by the mean
IC50 results vs. L. inf, which pinpointed only the 4-phenoxy analogue (23) as being
significantly less active, with the remaining compounds (including 7) falling within a ca. 2-
fold potency range. Replacement of the side chain linking oxygen atom in 7 by nitrogen (24
and 25) was also very well tolerated (L. inf IC s ≤1.5-fold apart), and, in the case of 24, this
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
allowed a 6-fold solubility improvement (albeit, this compound seemed to be less stable than
). In contrast, removal of the aryl moiety (alcohol 26; ꢀCLogP -3.4 units over 7) led to a 55-
7
fold loss in VL activity, suggesting that an aryl-based side chain was necessary (as for TB).
Therefore, we next investigated a small set of trifluoromethyl-substituted heteroaryl ethers
(27-31). This design concept was based on some partial success with employing
trifluoromethylpyridine as a more soluble substitute for trifluoromethoxyphenyl in our
20,21,37
previous TB studies.
Pleasingly, pyridine 27 (L. inf IC50 0.24 µM) was slightly more
potent than both 7 and the direct phenyl equivalent 22, and was 3-fold more soluble than 7
(11 µg/mL). However, less lipophilic analogues of 27 (28-31) gave 2- to 6-fold poorer IC50
values against the parasite, despite in two cases (28, 30) providing superior solubility.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 72
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 SAR focus then switched to targets with phenylpyridine side chains, which could be
20
regarded as more soluble mimics of the biphenyl lead 8. For completeness, the relatively
insoluble 2H biphenyls 32-34 were first evaluated and found to be 2- to 6-fold more active
than their 2-methyl counterparts (8, 46, and 47), similar to the 2-fold higher L. inf potency of
12 vs. 7. The 2H phenyl-2-pyridines 35 and 36 had comparable or up to 6-fold reduced
activities compared to the biphenyls 33 and 34, respectively. Conversely, 2H phenyl-3-
pyridine derivatives 37, 40 and 41 were noteworthy for the exceptional potency of the 4-
fluoro compound 37 (L. inf IC50 0.03 µM), which was 11-fold better than 7 and 25-fold
superior to the 4-trifluoromethoxy analogue 41.
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
Similar SAR trends were revealed in the 2-methyl series where, for phenyl-3-pyridine
derivatives (51-53), the 4-fluoro compound 51 (L. inf IC50 0.13 µM) was overwhelmingly
more active than congeners 52 and 53 (by 40- to 85-fold). By contrast, in the less soluble and
more lipophilic (ꢀCLogP +0.6 units) isomeric phenyl-2-pyridine series (48-50), the 4-fluoro
compound 48 (L. inf IC50 2.9 µM) was no better than biphenyl lead 8 or close analogue 49,
although the 4-trifluoromethoxy derivative 50 was inferior. It was interesting to observe that
neither 37 nor 51 showed enhanced potencies against the other parasites (or TB) and that both
compounds displayed ~5-fold better aqueous solubility than 8 at neutral pH, in harmony with
their decreased lipophilicities (ꢀCLogP -1.1 to -1.5 units). This solubility differential over 8
was considerably larger at low pH (>2000-fold; Table 3), suggesting further promise.
Therefore the enantiomers of 37 (38 and 39) were also prepared and assessed, together with
their 5-nitro isomers (42 and 43). Here, unexpectedly, the S enantiomer 39 was the most
active (2-fold over 38), while the 5-nitro analogues were 3 orders of magnitude less effective.
Overall, given the superior in vivo efficacy of 8 compared to 47 and the much higher in
vitro potencies of 37 and 51, it did appear that a 4-fluoro substituent in the terminal ring was
preferred for VL. Therefore, in the remaining compounds (54-59) we retained this element
1
6
ACS Paragon Plus Environment

Page 17 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and probed further modifications expected to improve solubility. Commencing with the
bipyridines 54-56, the most synthetically challenging isomer 54 was found to be the best (L.
inf IC50 0.14 µM), having >20-fold higher potencies than the related phenylpyridine 48 and
the biphenyl congener 8, and 13- to 16-fold greater aqueous solubility than these compounds
at neutral pH (in accordance with its reduced lipophilicity, ꢀCLogP -1.1 to -1.6 units).
Importantly, more structurally diverse phenylpiperazine and phenylpiperidine analogues (57
and 58) also retained excellent activities (L. inf IC50s 0.20 and 0.23 µM). As noted above,
phenylpiperazine and phenylpiperidine are soluble bioisosteres for biphenyl and these
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
6
0,61
moieties are now appearing in many new drug candidates.
Here, 57 and 58 were
respectively 61-fold and 11-fold more soluble than 8 at neutral pH and formed nicely soluble
hydrochloride salts at low pH (19-23 mg/mL), in line with their calculated pKa values of 5.9
and 7.3. Finally, based on previous success with an arylpiperazine carbamate analogue of 4
41
for TB, carbamate 59 was studied, but the additional carboxy group proved detrimental for
VL (8.5-fold potency loss). There was also a poor tolerance for longer side chains (60 and 5).
With these SAR results in hand, the next stage of this work was to assess microsomal
stability and in vivo efficacy for the best new leads. Six racemic compounds (pyridine 27,
phenylpyridines 37 and 51, bipyridines 54 and 56, and phenylpiperazine 57) were selected as
representative of both potency and solubility optimisations for the new side chain classes
investigated here. Unsurprisingly, these diverse compounds demonstrated a wide range of
metabolic stabilities (Table 3) following 1 h exposures to human and mouse liver microsomes
(HLM and MLM). While all compounds showed broadly acceptable stabilities towards HLM
(~50% or more parent remaining), phenylpyridine 51 and phenylpiperazine 57 displayed low
stabilities toward MLM (14-21% parent), with the others having moderate MLM stabilities
47-69% parent). The most stable compounds were pyridine 27 and bipyridine 56, and a 2-H
substituent (37) appeared to reduce the rate of MLM metabolism (cf. 2-methyl derivative 51).
(
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 72
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
Nevertheless, upon in vivo evaluation in the mouse model of VL (dosing orally at 25 mg/kg
for 5 consecutive days), all candidates except the least potent one (bipyridine 56) provided
essentially total parasite clearance (>99.5%), similar to 7, 8, and 12. At 6.25 mg/kg, the more
rapidly metabolised phenylpiperazine 57 demonstrated poor efficacy (16% inhibition),
whereas pyridine 27 (83%) was very similar to the des-methyl phenyl ether 12 (89%),
bipyridine 54 was highly effective (96%), and phenylpyridines 37 and 51 both achieved 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
100% cure (Figure 2 and Table 3). From further dose reductions on 37, 51, and 54, it was
subsequently established that the highly potent 2H phenylpyridine 37 was the most active
lead (98% inhibition at 1.56 mg/kg, cf. <50% for 9, 51, and 54), and a final head-to-head
assessment of its enantiomers determined that the S form was marginally preferred (ED s 1.3
50
and 1.1 mg/kg for 38 and 39), despite demonstrating inferior MLM stability.
As discussed, a key objective of this work was to identify backups to 9 having an improved
physicochemical/pharmacological profile. From our study of nitroimidazooxazoles, we have
discovered several new racemic analogues (e.g., bipyridine 54, and phenylpyridines 37 and
51) with very high in vivo efficacies in a mouse model of VL, broadly comparable to 9.
However, 54 could not be regarded as a good candidate for further development due to its
high synthetic difficulty, while the low metabolic stability of phenylpyridine 51 toward MLM
would not bode well for its assessment in the more stringent hamster model (where faster
rates of metabolism were often observed). Thus, the best new racemic lead identified here
was the phenylpyridine 37, and the most active enantiomer of 37 was 39. With improved
a
aqueous solubility under acidic conditions (a calculated pK value of 3.73), acceptable
metabolic stability toward HLM, and excellent in vivo efficacy in the mouse model of VL,
this compound may warrant further investigation as a potential new antileishmanial agent.
3. SAR of new heterocyclic analogues for NTDs
1
8
ACS Paragon Plus Environment

Page 19 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The mechanism of action of these nitroimidazooxazoles against VL is currently unknown.
It has recently been disclosed that the parasite L. don does not possess a homologue of the M.
tb nitroreductase Ddn and that compounds such as 4 are not activated by the same type I
62
nitroreductase that operates on fexinidazole. In addition to this, L. don reportedly does not
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
possess the mycolic acid biosynthetic pathway that is present in M. tb and inhibited by 4 and
5
under aerobic conditions. Therefore, as an extension to the current SAR study, we were
9
5
interested in ascertaining which elements of the nitroheterocyclic ring system of 7 were
essential for the antiparasitic activities observed (L. inf, T. cruzi) and whether or not any new
ring A analogues could become useful leads. The des-nitro congener of 7 (61) was inactive
against VL but unexpectedly retained similar potency to 7 against T. cruzi (IC50 2.1 µM;
Table 2); therefore, a variety of alternative imidazole ring substituents (e.g., Br, CHO,
2 2
CH OH, SMe, SO Me) were evaluated (62-69). As expected, such compounds showed
almost no utility for VL, but the 5-bromo (67) and 5-methylthio (68) analogues provided
enhanced antitrypanosomal potencies (T. cruzi IC s 0.13 and 0.58 µM; 6-substituted
50
congeners were notably inferior). It is attractive to speculate that the inactivity of 61-69
against L. inf may point to a specific requirement for the 6-nitro group of 7 in the mechanism
of action, beyond simply electronic or protein binding effects. In this work, we have shown
that the 6-nitro group of 7 can readily undergo cine-substitution by thiolate. Since it has been
6
3
established that nitric oxide is cidal to Leishmania, the known capacity of these compounds
to produce nitric oxide (via loss of the nitro group as nitrous acid) might still be a factor in
their VL activity if specific nucleophilic activation (e.g., by an active site cysteine thiolate or
2
hydride) could occur via some other enzyme system within the parasite.
3
The remaining compounds in Table 2 (70-72) investigated replacement of the imidazole
ring by triazole or pyrazole. The direct analogues of 7 (70 and 72) were both completely
inactive against L. inf, suggesting that in addition to the 6-nitro group, this imidazole ring is
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 72
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
3,64
also essential for VL activity (mimicking the structural requirements for TB activity
).
13
Based on measurements in the oxazine series, the nitro reduction potentials of 72 and 7
should be very similar, but 70 and 72 could not undergo the unique ring reduction chemistry
and nitric oxide release exhibited by 4. Interestingly, the nitrotriazole 70 was more active
than 7 against T. cruzi (IC s 0.59 and 1.8 µM, respectively) whereas the bromotriazole 71
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
5
0
and nitropyrazole 72 were less active. Thus, while the 4-nitroimidazole ring of 7 appeared to
be essential for retaining utility against VL, a much greater degree of structural variation was
tolerated for Chagas disease, suggesting a different mechanism of action. The results from
this investigation may therefore stimulate interest in developing new agents for Chagas
disease from these rather novel leads.
4. SAR of nitroimidazooxazoles for TB
While the major focus of these efforts was directed towards VL, we also examined the
utility of the new oxazoles against TB, using the in vitro assays described above (Table 1).
Generally, it appeared that pyridine, phenylpyridine, bipyridine, phenylpiperazine,
phenylpiperidine, and phenylpiperazine carbamate were all useful side chains for TB,
consistent with findings in the nitroimidazooxazine series and fairly scant disclosures from
31,57
Otsuka Pharmaceutical Co.
It was particularly fascinating to observe that, except for 12,
1
4, and 39, all of the most active VL leads (including 7, 8, and 9, but also 27, 37, 51, and 54)
possessed notably high antitubercular potencies in the MABA assay (MICs <0.08 µM,
comparable to 5), despite exhibiting only modest growth inhibitory effects under low oxygen
conditions (LORA). A closer examination of monoheterocyclic side chains revealed that the
2
-pyridine 27 (MABA MIC 0.038 µM) was indeed optimal (as for VL), but replacement of
the oxygen linkage in 7 by nitrogen (24, 25) reduced MABA potency by 5-fold. For the new
biaryl side chain derivatives, phenylpyridines such as 37, 38, 40, 41, and 48-53, as well as
bipyridine 54, uniformly displayed the highest aerobic activities (MICs ≤0.05 µM), similar or
2
0
ACS Paragon Plus Environment

Page 21 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
slightly better than the original biphenyl analogues (8, 46, and 47). However, the elusive 2H
phenyl-2-pyridines 35 and 36 showed dramatically lower potencies in this assay (MICs of 58
and 5.3 µM), possibly indicating reduced stability. More soluble bioisosteres containing a
cyclic amine (57-59) also provided slightly inferior results (MABA MICs 0.13-0.29 µM).
Because of the early promise of 7 and 8 for VL, it was considered worthwhile to evaluate
the enantiomers of 7 (9 and 18) and the R enantiomer of 8 (44) alongside pretomanid (4) in a
mouse model of acute TB infection, dosing orally for 3 weeks (daily for 5 days each
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
9,50
week).
15 mg/kg), respectively 9.3- and 7.6-fold superior to 4, whereas the S enantiomer of 7 (18)
was slightly less efficacious than 4 (at 100 mg/kg; Figure 3). Nevertheless, the VL lead 9 was
In this assay, both 9 and 44 (100 mg/kg) showed efficacies similar to rifampicin
(
18
reportedly much less active than delamanid (5) in a comparable M. tb infection model.
CONCLUSIONS
This study began with discovery of the extraordinary utility of several antitubercular
oxazole analogues against the neglected kinetoplastid disease VL. Having first identified 9 as
a preferred lead for preclinical development, we later investigated its SAR, with the objective
of improving aqueous solubility and other drug-like features without compromising in vivo
efficacy. All modifications to the nitroheterocycle (viz. removal of the nitro group or
switching its position, and replacement of the imidazole ring by pyrazole or triazole) resulted
in a total loss of activity against VL (but not Chagas disease), whereas many aryl-,
heteroaryl-, and heterobiaryl-based side chains were well tolerated in vitro. Surprisingly,
there were several significant similarities in these respects to the SAR for TB, while the
discovery of facile cine-substitution of the nitro group of 7 by thiolate was intriguing.
Overall, the most effective strategies to enhance aqueous solubility while maintaining or
increasing potency were the replacement of phenyl rings by pyridine and the use of
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 72
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
arylpiperazine and arylpiperidine as bioisosteres for a biaryl moiety. These strategies enabled
solubility improvements over 7 of 3- to 14-fold at pH = 7 and up to 6400-fold at pH = 1.
Assessments of stability toward HLM gave broadly acceptable results, although two
compounds (phenylpyridine 51 and phenylpiperazine 57) had high MLM metabolism rates.
Further appraisal of six candidates in the acute infection mouse model of VL revealed three
racemic compounds (phenylpyridines 37 and 51, and bipyridine 54) that displayed efficacies
at least as strong as 7. Of these, one compound (37) was notably superior, in line with its
excellent in vitro profile, and subsequent synthesis and evaluation of its enantiomers has
pinpointed the S form 39 as being the slightly preferred VL lead for further investigation.
Findings from this work may also inspire interest in these promising compounds for other
NTDs.
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
EXPERIMENTAL SECTION
Combustion analyses were performed by the Campbell Microanalytical Laboratory,
University of Otago, Dunedin, New Zealand. Melting points were determined using an
Electrothermal IA9100 melting point apparatus, and are as read. NMR spectra were measured
1 13
on a Bruker Avance 400 spectrometer at 400 MHz for H and 100 MHz for C and are
referenced to Me Si or solvent resonances. Chemical shifts and coupling constants are recorded
4
in units of ppm and hertz, respectively. High-resolution chemical ionisation (HRCIMS) and fast
atom bombardment (HRFABMS) mass spectra were determined on a VG-70SE mass
spectrometer at nominal 5000 resolution. High-resolution electrospray ionisation (HRESIMS)
mass spectra were determined on a Bruker micrOTOF-Q II mass spectrometer. Low-
resolution atmospheric pressure chemical ionisation (APCI) mass spectra were obtained for
organic solutions using a ThermoFinnigan Surveyor MSQ mass spectrometer, connected to a
Gilson autosampler. Optical rotations were measured on a Schmidt + Haensch Polartronic
2
2
ACS Paragon Plus Environment

Page 23 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH8 polarimeter. Column chromatography was performed on silica gel (Merck 230-400 mesh).
Thin-layer chromatography was carried out on aluminium-backed silica gel plates (Merck 60
2 4
F254), with visualization of components by UV light (254 nm), I , or KMnO staining. Tested
compounds (including batches screened in vivo) were ≥ 95% pure, as determined by
combustion analysis (results within 0.4% of theoretical values) and/or by HPLC conducted
on an Agilent 1100 system, using a 150 mm x 3.2 mm Altima 5 µm reversed phase C18
column with diode array detection. Finally, chiral purity was assessed by HPLC carried out
on a Gilson Unipoint system (322-H pump, 156 UV/vis detector), employing a 250 mm x 4.6
mm CHIRALCEL OD 10 µm analytical column, or 250 mm x 4.6 mm CHIRALPAK IA or
CHIRALPAK AD-H 5.0 µm analytical columns.
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
Compounds of Table 1. The following section details the syntheses of compounds 10-13,
15, 16, 18, 23, 24, 35, 37, 48, 55, and 57 of Table 1, via representative procedures and key
intermediates, as described in Schemes 1-4. For the syntheses of all of the other compounds
in Table 1, please refer to Supporting Information.
Synthesis of 10 (Scheme 1A):
Procedure A: 2-({[4-(Trifluoromethoxy)benzyl]oxy}methyl)oxirane (74). A solution of
glycidol (73) (303 mg, 4.09 mmol) and 4-(trifluoromethoxy)benzyl bromide (0.81 mL, 5.06
mmol) in anhydrous DMF (6 mL) under N
2
at 0 °C was treated with 60% NaH (246 mg, 6.15
mmol), then quickly degassed and resealed under N . The resulting mixture was stirred at 20
2
2 3
°C for 7 h and then cooled (CO /acetone), quenched with ice/aqueous NaHCO (100 mL),
and extracted with EtOAc (4 x 100 mL). The extracts were washed with brine (100 mL) and
then evaporated to dryness under reduced pressure (at 30 °C), and the residue was
chromatographed on silica gel. Elution with 0-5% Et O/petroleum ether first gave foreruns,
2
and then further elution with 5-10% Et O/petroleum ether gave 74 (625 mg, 62%) as a
2
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 72
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
colourless oil; H NMR (CDCl
3
) δ 7.38 (br d, J = 8.7 Hz, 2 H), 7.20 (br d, J = 8.7 Hz, 2 H),
4.62 (d, J = 12.0 Hz, 1 H), 4.56 (d, J = 12.0 Hz, 1 H), 3.82 (dd, J = 11.5, 2.8 Hz, 1 H), 3.43
(dd, J = 11.5, 6.0 Hz, 1 H), 3.24-3.18 (m, 1 H), 2.82 (dd, J = 4.9, 4.2 Hz, 1 H), 2.63 (dd, J =
+
.0, 2.7 Hz, 1 H); HRESIMS calcd for C H F NaO m/z [M + Na] 271.0552, found
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
11 11
3
3
271.0557.
Procedure
B:
1-(2-Bromo-4-nitro-1H-imidazol-1-yl)-3-{[4-
(trifluoromethoxy)benzyl]oxy}propan-2-ol (76). A mixture of epoxide 74 (500 mg, 2.01
mmol), 2-bromo-4-nitro-1H-imidazole (75) (390 mg, 2.03 mmol), and DIPEA (1.75 mL, 10.0
mmol) in a sealed vial was stirred at 107 °C for 13 h. The resulting cooled mixture was
dissolved in CH
5 x 50 mL). The extracts were evaporated to dryness under reduced pressure and the residue
was chromatographed on silica gel. Elution with 0-1% EtOAc/CH Cl first gave foreruns, and
then further elution with 1-2% EtOAc/CH Cl gave 76 (538 mg, 61%) as a cream solid: mp
2 2 3 2 2
Cl and added to ice/aqueous NaHCO (50 mL), then extracted with CH Cl
(
2
2
2
2
1
(CH
2
Cl
2
/pentane) 80-81 °C; H NMR (CDCl ) δ 7.95 (s, 1 H), 7.35 (br d, J = 8.7 Hz, 2 H),
3
7.23 (br d, J = 7.9 Hz, 2 H), 4.57 (s, 2 H), 4.20 (dd, J = 13.6, 2.9 Hz, 1 H), 4.18-4.10 (m, 1
H), 4.07 (dd, J = 13.4, 7.1 Hz, 1 H), 3.59 (dd, J = 9.6, 4.2 Hz, 1 H), 3.46 (dd, J = 9.6, 5.3 Hz,
+
1 H), 2.61 (d, J = 5.0 Hz, 1 H); HRESIMS calcd for C14
40.0063, found 442.0044, 440.0061.
Procedure C: 6-Nitro-2-({[4-(trifluoromethoxy)benzyl]oxy}methyl)-2,3-
dihydroimidazo[2,1-b][1,3]oxazole (10). A solution of alcohol 76 (524 mg, 1.19 mmol) in
anhydrous DMF (12 mL) under N
at 0 °C was treated with 60% NaH (74 mg, 1.85 mmol),
then quickly degassed and resealed under N
. The resulting mixture was stirred at 0 °C for 65
min and then cooled (CO /acetone), quenched with ice/aqueous NaHCO (20 mL), added to
brine (100 mL), and extracted with CH Cl (6 x 100 mL). The combined extracts were
evaporated to dryness under reduced pressure (at 30 °C) and the residue was
H14BrF N O
3 3 5
m/z [M + H] 442.0044,
4
2
2
2
3
2
2
2
4
ACS Paragon Plus Environment

Page 25 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
chromatographed on silica gel. Elution with CH
2
Cl
gave 10 (383 mg, 90%) as a cream solid: mp
) δ 7.53 (s, 1 H), 7.29 (br d, J = 8.7 Hz, 2 H),
2
first gave foreruns, and then further
2 2
elution with 0-2% EtOAc/CH Cl
1
(CH
2
Cl
2
/hexane) 134-135 °C; H NMR (CDCl
3
7
.20 (br d, J = 8.0 Hz, 2 H), 5.46-5.38 (m, 1 H), 4.60 (s, 2 H), 4.32 (dd, J = 10.0, 8.6 Hz, 1
H), 4.26 (dd, J = 10.0, 6.5 Hz, 1 H), 3.89 (dd, J = 11.3, 3.9 Hz, 1 H), 3.78 (dd, J = 11.3, 3.5
Hz, 1 H). Anal. (C14 ) C, H, N.
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
12 3 3 5
H F N O
Synthesis of 11 (Scheme 1A):
-{[(2-Methylprop-2-en-1-yl)oxy]methyl}-4-(trifluoromethoxy)benzene (78). A solution
1
of 2-methylprop-2-en-1-ol (77) (2.34 mL, 27.8 mmol) in anhydrous DMF (10 mL, then 2 x 2
mL to rinse) was added dropwise to a stirred suspension of 60% NaH (1.32 g, 33.0 mmol) in
anhydrous DMF (10 mL) under N
2
at 0 °C. After 30 min, 4-(trifluoromethoxy)benzyl
bromide (5.1 mL, 31.9 mmol) was added, and the mixture was stirred at 20 °C for 21 h. The
resulting mixture was added to ice/aqueous NaHCO (200 mL) and extracted with 25%
3
EtOAc/petroleum ether (2 x 200 mL) and 50% EtOAc/petroleum ether (3 x 200 mL). The
extracts were washed with water (200 mL) and then concentrated under reduced pressure, and
the remaining oil was chromatographed on silica gel. Elution with petroleum ether first gave
foreruns, and then further elution with 0-5% CH
2
Cl
2
/petroleum ether gave 78 (6.57 g, 96%)
1
as a colourless oil, which was used directly in the next step; H NMR (CDCl
3
) δ 7.37 (br d, J
=
8.7 Hz, 2 H), 7.19 (br d, J = 8.0 Hz, 2 H), 5.00 (m, 1 H), 4.94 (m, 1 H), 4.48 (s, 2 H), 3.94
(br s, 2 H), 1.77 (s, 3 H).
Procedure D: 2-Methyl-2-({[4-(trifluoromethoxy)benzyl]oxy}methyl)oxirane (79). 3-
Chloroperoxybenzoic acid (19.5 g of 50%, 56.5 mmol) was added to a mixture of alkene 78
6.57 g, 26.7 mmol) and disodium hydrogen phosphate (9.87 g, 69.5 mmol) in CH Cl (250
(
2
2
mL) at 0 °C. The mixture was stirred at 20 °C for 3 h and then added to an ice-cold aqueous
solution of sodium sulphite (400 mL of 10%) and extracted with CH Cl (3 x 400 mL). The
2
2
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 72
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
extracts were sequentially washed with ice-cold aqueous sodium sulphite solution (400 mL of
1
0%), aqueous NaHCO
3
(400 mL), and brine (300 mL), and then the aqueous solutions were
Cl (200 mL). The combined extracts were concentrated under
back-extracted with CH
2
2
reduced pressure and the remaining oil was chromatographed on silica gel. Elution with 0-
15% CH Cl /petroleum ether first gave foreruns, and then further elution with 15%
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
2
1
CH
2 2 2 2
Cl /petroleum ether and CH Cl gave 79 (6.53 g, 93%) as a colourless oil; H NMR
(
CDCl ) δ 7.37 (br d, J = 8.7 Hz, 2 H), 7.19 (br d, J = 7.9 Hz, 2 H), 4.59 (d, J = 12.1 Hz, 1
3
H), 4.54 (d, J = 12.1 Hz, 1 H), 3.61 (d, J = 11.1 Hz, 1 H), 3.44 (d, J = 11.1 Hz, 1 H), 2.75 (d,
J = 4.9 Hz, 1 H), 2.64 (d, J = 4.9 Hz, 1 H), 1.40 (s, 3 H); HRCIMS (NH ) calcd for
3
+
+ H] 280.1161, found 280.1144.
C
12
H
17
F
3
NO
3
m/z [M + NH
3
1-(2-Bromo-4-nitro-1H-imidazol-1-yl)-2-methyl-3-{[4-(trifluoromethoxy)benzyl]oxy}-
propan-2-ol (80). Reaction of epoxide 79 with 2-bromo-4-nitro-1H-imidazole (75) (1.1
equiv) and DIPEA (7.6 equiv), using procedure B at 108 °C for 15 h, followed by
chromatography of the product on silica gel, eluting with CH Cl , gave 80 (94%) as a pale
2
2
1
yellow oil; H NMR (CDCl
3
) δ 8.00 (s, 1 H), 7.33 (br d, J = 8.6 Hz, 2 H), 7.22 (br d, J = 8.0
Hz, 2 H), 4.56 (s, 2 H), 4.15 (d, J = 14.8 Hz, 1 H), 4.04 (d, J = 14.5 Hz, 1 H), 3.39 (s, 2 H),
+
2.51 (s, 1 H), 1.22 (s, 3 H); HRESIMS calcd for C15
54.0220, found 456.0197, 454.0221.
-Methyl-6-nitro-2-({[4-(trifluoromethoxy)benzyl]oxy}methyl)-2,3-
H
16BrF
3
N
3
O
5
m/z [M + H] 456.0200,
4
2
dihydroimidazo[2,1-b][1,3]oxazole (11). Reaction of alcohol 80 with NaH (1.4 equiv), using
procedure C for 85 min, followed by chromatography of the product on silica gel, eluting
with CH
2
Cl
2
(foreruns) and then with 0-1% EtOAc/CH
2 2
Cl , gave 11 (87%) as a pale yellow
1
solid: mp (CH
2
Cl
2
/hexane) 110-111 °C; H NMR (CDCl
3
) δ 7.50 (s, 1 H), 7.26 (br d, J = 8.4
Hz, 2 H), 7.19 (br d, J = 8.3 Hz, 2 H), 4.59 (d, J = 12.3 Hz, 1 H), 4.56 (d, J = 12.3 Hz, 1 H),
2
6
ACS Paragon Plus Environment

Page 27 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4.36 (d, J = 10.0 Hz, 1 H), 3.91 (d, J = 10.0 Hz, 1 H), 3.72 (d, J = 10.7 Hz, 1 H), 3.59 (d, J =
14 3 3 5
10.6 Hz, 1 H), 1.65 (s, 3 H). Anal. (C15H F N O ) C, H, N.
Synthesis of 12 and 15 (Scheme 1B):
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
Procedure E: 2-{[4-(Trifluoromethoxy)phenoxy]methyl}oxirane (82). A mixture of
epibromohydrin (81) (0.30 mL, 3.51 mmol), 4-(trifluoromethoxy)phenol (0.152 mL, 1.17
mmol), and powdered K
2 3
CO (260 mg, 1.88 mmol) in acetone (3 mL) in a sealed vial was
stirred at 59 °C for 36 h. The resulting cooled mixture was filtered, washing with CH Cl , and
2
2
then the filtrate was concentrated under reduced pressure and the remaining oil was
chromatographed on silica gel. Elution with 0-15% CH Cl /pentane first gave foreruns, and
2
2
28
2 2
then further elution with 20-25% CH Cl /pentane gave 82 (260 mg, 95%) as a colourless
1
oil; H NMR (CDCl ) δ 7.14 (br d, J = 9.0 Hz, 2 H), 6.91 (br d, J = 9.1 Hz, 2 H), 4.23 (dd, J
3
=
11.1, 3.1 Hz, 1 H), 3.94 (dd, J = 11.1, 5.7 Hz, 1 H), 3.37-3.31 (m, 1 H), 2.91 (dd, J = 4.8,
+
4.2 Hz, 1 H), 2.75 (dd, J = 4.9, 2.6 Hz, 1 H); HRFABMS calcd for C H F O m/z (M )
10
9
3
3
2
34.0504, found 234.0508.
-Nitro-2-{[4-(trifluoromethoxy)phenoxy]methyl}-2,3-dihydroimidazo[2,1-
b][1,3]oxazole (15) and 1-(2-bromo-4-nitro-1H-imidazol-1-yl)-3-[4-
trifluoromethoxy)phenoxy]propan-2-ol (83). Reaction of epoxide 82 with 2-bromo-4-
nitro-1H-imidazole (75) (1.1 equiv) and DIPEA, using procedure B at 105 °C for 6.5 h,
followed by chromatography of the product on silica gel, eluting with CH Cl , first gave a
mixture of oxazole isomers, which was further purified by chromatography on silica gel.
Elution with 3:1 and 4:1 CH Cl /petroleum ether first gave foreruns, and then further elution
5
(
2
2
2
2
with 4:1 CH
2
Cl
2
/petroleum ether and CH
2 2
Cl gave 15 (2.5%) as a white solid: mp
1
(
CH Cl
2
2
/pentane) 120-122 °C; H NMR (CDCl
3
) δ 7.68 (s, 1 H), 7.17 (br d, J = 8.5 Hz, 2 H),
6.90 (br d, J = 9.1 Hz, 2 H), 5.77-5.69 (m, 1 H), 4.71 (dd, J = 10.7, 8.9 Hz, 1 H), 4.61 (dd, J
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 72
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
= 10.7, 6.9 Hz, 1 H), 4.41 (dd, J = 10.9, 4.1 Hz, 1 H), 4.31 (dd, J = 10.9, 3.8 Hz, 1 H). Anal.
(C
13
H
10
3 3 5
F N O ) C, H, N.
Further elution of this second column above with 1% EtOAc/CH
2
2
Cl gave 12 (2%) as a
white solid (see below). Further elution of the original column above with 0-1%
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
EtOAc/CH Cl
2
2
gave 83 (70%) as a white solid: mp (MeOH/CH
2
Cl
2
/hexane) 139-141 °C; H
NMR [(CD ) SO] δ 8.52 (s, 1 H), 7.30 (br d, J = 9.1 Hz, 2 H), 7.05 (br d, J = 9.2 Hz, 2 H),
3
2
5.66 (br s, 1 H), 4.28 (dd, J = 13.3, 3.3 Hz, 1 H), 4.25-4.17 (m, 1 H), 4.13 (dd, J = 13.3, 8.0
Hz, 1 H), 4.01 (d, J = 5.0 Hz, 2 H). Anal. (C H BrF N O ) C, H, N.
3
13 11
3
5
6-Nitro-2-{[4-(trifluoromethoxy)phenoxy]methyl}-2,3-dihydroimidazo[2,1-
b][1,3]oxazole (12). Reaction of alcohol 83 with NaH, using procedure C for 40 min,
followed by chromatography of the product on silica gel, eluting with CH Cl
2
2
, gave 12 (85%)
1
as a cream solid: mp (MeOH/CH
2
Cl
2
/hexane) 170-172 °C; H NMR [(CD
3 2
)
SO] δ 8.16 (s, 1
H), 7.31 (br d, J = 9.1 Hz, 2 H), 7.05 (br d, J = 9.2 Hz, 2 H), 5.78-5.70 (m, 1 H), 4.50 (dd, J
10.8, 8.9 Hz, 1 H), 4.46 (dd, J = 11.5, 2.8 Hz, 1 H), 4.39 (dd, J = 11.5, 5.2 Hz, 1 H), 4.22
=
1
3
(
dd, J = 10.8, 6.5 Hz, 1 H); C NMR [(CD ) SO] δ 156.8, 156.1, 145.7, 142.2 (q, J = 1.5
3
2
C-F
Hz), 122.6 (2 C), 120.1 (q, JC-F = 255.2 Hz), 116.0 (3 C), 85.3, 68.3, 45.1. Anal.
C H F N O ) C, H, N.
(
1
3
10
3
3
5
Synthesis of 13 and 16 (Scheme 1C):
Procedure F: (2R)-2-{[4-(Trifluoromethoxy)phenoxy]methyl}oxirane (90). 4-
(
Trifluoromethoxy)phenol (0.22 mL, 1.70 mmol) was added to a suspension of cesium
fluoride (0.80 g, 5.27 mmol) in anhydrous DMF (1.4 mL) under N and the mixture was
stirred at 20 °C for 1 h. (2R)-Oxiran-2-ylmethyl 3-nitrobenzenesulfonate (89) (446 mg, 1.72
mmol) was added and the reaction mixture was briefly degassed and resealed under N and
then stirred at 20 °C for 35 h. The resulting mixture was added to water (50 mL) and
2
2
2
8
ACS Paragon Plus Environment

Page 29 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
extracted with 50% Et
2
O/petroleum ether (3 x 50 mL). The extracts were concentrated under
reduced pressure (at 25 °C) and the remaining oil was chromatographed on silica gel. Elution
with 0-15% CH
CH Cl
2
Cl /pentane first gave foreruns, and then further elution with 15%
2
1
/pentane gave 90 (313 mg, 79%) as a colourless oil; H NMR (CDCl
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
2
3
) δ 7.14 (br d, J
=
9.2 Hz, 2 H), 6.91 (br d, J = 9.2 Hz, 2 H), 4.24 (dd, J = 11.0, 3.0 Hz, 1 H), 3.94 (dd, J =
11.0, 5.8 Hz, 1 H), 3.39-3.32 (m, 1 H), 2.92 (dd, J = 4.8, 4.2 Hz, 1 H), 2.76 (dd, J = 4.9, 2.7
2
5
+
Hz, 1 H); [α]
D 3 9 3 3
-4.0 (c 2.00, CHCl ); HRESIMS calcd for C10H F NaO m/z [M + Na]
2
57.0396, found 257.0399.
Procedure G:
dihydroimidazo[2,1-b][1,3]oxazole (16) and (2R)-1-(2-chloro-4-nitro-1H-imidazol-1-yl)-
-[4-(trifluoromethoxy)phenoxy]propan-2-ol (92). A mixture of epoxide 90 (202 mg, 0.863
(2R)-5-Nitro-2-{[4-(trifluoromethoxy)phenoxy]methyl}-2,3-
3
mmol), 2-chloro-4-nitro-1H-imidazole (91) (141 mg, 0.956 mmol), and DIPEA (0.170 mL,
0.976 mmol) in anhydrous toluene (1.0 mL) in a sealed vial was stirred at 80 °C for 24 h. The
resulting cooled mixture was transferred to a flask (in CH
2 2
Cl ) and evaporated to dryness
under reduced pressure (at 30 °C) and the residue was chromatographed on silica gel. Elution
with CH
2
Cl
2
first gave foreruns, and then further elution with 0.5% EtOAc/CH Cl
2
2
gave 16
SO] δ
1
(47 mg, 16%) as a white solid: mp (CH
2
Cl
2
/pentane) 151-152 °C; H NMR [(CD
3
)
2
7
.87 (s, 1 H), 7.32 (br d, J = 9.0 Hz, 2 H), 7.07 (br d, J = 9.2 Hz, 2 H), 5.93-5.84 (m, 1 H),
.68 (dd, J = 10.1, 9.4 Hz, 1 H), 4.48 (dd, J = 11.6, 2.9 Hz, 1 H), 4.43 (dd, J = 11.6, 5.4 Hz,
4
13
1 H), 4.40 (dd, J = 10.1, 6.8 Hz, 1 H); C NMR [(CD
3
)
2
SO] δ 160.0, 156.7, 142.2 (q, JC-F
=
2.0 Hz), 134.7, 133.5, 122.6 (2 C), 120.1 (q, JC-F = 255.2 Hz), 116.0 (2 C), 87.5, 68.3, 46.3;
2
4
[
α] -239.0 (c 1.004, DMF). Anal. (C H F N O ) C, H, N.
D
13 10
3
3
5
Further elution of the above column with 0.5-1% EtOAc/CH
Cl
2 2
gave 92 (222 mg, 67%) as
1
a white solid: mp (CH Cl /pentane) 110-111 °C; H NMR (CDCl ) δ 7.96 (s, 1 H), 7.19 (br d,
2
2
3
J = 9.1 Hz, 2 H), 6.90 (br d, J = 9.1 Hz, 2 H), 4.42-4.30 (m, 2 H), 4.21 (dd, J = 14.6, 7.8 Hz,
2
9
ACS Paragon Plus Environment