Identification and Optimization of a Series of 8-Hydroxy Naphthyridines with Potent in Vitro Antileishmanial Activity: Initial SAR and Assessment of in Vivo Activity

Article abstract of DOI:10.1021/acs.jmedchem.0c00705

Visceral leishmaniasis (VL) is a parasitic infection that results in approximately 26 ?000-65 ?000 deaths annually. The available treatments are hampered by issues such as toxicity, variable efficacy, and unsuitable dosing options. The need for new treatments is urgent and led to a collaboration between the Drugs for Neglected Diseases initiative (DNDi), GlaxoSmithKline (GSK), and the University of Dundee. An 8-hydroxynaphthyridine was identified as a start point, and an early compound demonstrated weak efficacy in a mouse model of VL but was hampered by glucuronidation. Efforts to address this led to the development of compounds with improved in vitro profiles, but these were poorly tolerated in vivo. Investigation of the mode of action (MoA) demonstrated that activity was driven by sequestration of divalent metal cations, a mechanism which was likely to drive the poor tolerability. This highlights the importance of investigating MoA and pharmacokinetics at an early stage for phenotypically active series.

Full text of DOI:10.1021/acs.jmedchem.0c00705

Subscriber access provided by AUT Library
Article
Identification and optimisation of a series of 8-hydroxy
naphthyridines with potent in vitro antileishmanial
activity: Initial SAR and assessment of in vivo activity
Michael Thomas, Manu De Rycker, Richard J Wall, Daniel Spinks, Ola Epemolu, Sujatha Manthri,
Suzanne Norval, Maria Osuna-Cabello, Stephen Patterson, Jennifer Riley, Frederick R. C.
Simeons, Laste Stojanovski, John Thomas, Stephen Thompson, Claire Naylor, Jose M, Fiandor,
Paul G. Wyatt, Maria Marco, Susan Wyllie, Kevin D Read, Timothy J. Miles, and Ian H. Gilbert
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 14 Jul 2020
Downloaded from pubs.acs.org on July 14, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 59
Journal of Medicinal Chemistry
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
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 2 of 59
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
Identification and optimisation of a series of 8-hydroxy naphthyridines with potent in vitro
antileishmanial activity: Initial SAR and assessment of in vivo activity
†
*
†
†
†
†
Michael G. Thomas, Manu De Rycker, Richard J. Wall, Daniel Spinks, Ola Epemolu, Sujatha
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
†
†
†
†
†
Manthri, , Suzanne Norval, Maria Osuna-Cabello, Stephen Patterson, Jennifer Riley, Frederick
†
†
†
†
†
R.C. Simeons, Laste Stojanovski, John Thomas, Stephen Thompson, Claire Naylor, Jose M.
§
†
§
†
†
§*
Fiandor, Paul G. Wyatt, Maria Marco, Susan Wyllie, Kevin D. Read, Timothy J. Miles, and
Ian H. Gilbert †*
*Co-corresponding authors
†
Drug Discovery Unit, Wellcome Centre for Anti-Infectives Research, Division of Biological
Chemistry and Drug Discovery, School of Life Sciences, University of Dundee, Dundee DD1
5
EH, United Kingdom
§
Global Health R&D, GlaxoSmithKline, Tres Cantos, 28760, Spain
Abstract
Visceral leishmaniasis (VL) is a parasitic infection that results in approximately 26,000 – 65,000
deaths annually. The available treatments are hampered by issues such as toxicity, variable efficacy
and unsuitable dosing options. The need for new treatments is urgent and led to a collaboration
between the Drugs for Neglected Diseases initiative (DNDi), GlaxoSmithKline (GSK) and the
University of Dundee. An 8-hydroxynaphthyridine was identified as a start point and an early
compound demonstrated weak efficacy in a mouse model of VL but was hampered by
glucuronidation. Efforts to address this led to the development of compounds with improved in
vitro profiles, but these were poorly tolerated in vivo. Investigation of the mode of action (MoA)
demonstrated that activity was driven by sequestration of divalent metal cations, a mechanism
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which was likely to drive the poor tolerability. This highlights the importance of investigating
MoA and pharmacokinetics at an early stage for phenotypically active series.
Introduction
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
According to the World Health Organisation (WHO), Neglected Tropical Diseases (NTDs) affect
1
in excess of 1 billion people, the majority being in the most impoverished areas of the world. Of
these NTDs, visceral leishmaniasis (VL) remains one of the most challenging to treat. VL is caused
by infection with the protozoan parasites Leishmania donovani and Leishmania infantum, which
2
are transmitted through the bite of female phlebotomine sand flies. Following infection, parasites
reside predominantly in the liver, spleen and bone marrow and if left untreated the disease is
invariably fatal. Although it is challenging to precisely determine the number of people affected
by VL, WHO estimates suggest that 50,000 – 90,000 new infections result in an annual death toll
of between 26,000 – 65,000, with the vast majority of cases occurring in India, Bangladesh, Sudan,
Ethiopia and Brazil.3
Although there are a number of available treatments, all suffer from major issues that limit their
use. Miltefosine, the only orally available treatment, is teratogenic so can only be prescribed to
women of child-bearing age alongside contraception, and also shows other side effects, as well as
4

variable efficacy. Liposomal amphotericin B (Ambisome ) is widely effective in Asia, but its use
is limited by the need for i.v. administration which requires hospitalisation, and it also requires a
cold-chain for storage. An alternative aminoglycoside antibiotic paromomycin is much less
expensive, but demonstrates variable efficacy and requires a long course of painful intramuscular
5
injections. Finally, pentavalent antimonials, such as sodium stibogluconate, have been widely
used since the 1940s, but they are cardiotoxic, require parenteral administration and are largely
ineffective in large areas of India due to high levels of drug resistance. Although new compounds
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
have recently entered the development pipeline, including two preclinical candidates developed
within this collaboration (DDD853651 / GSK3186899 and DDD1305143 / GSK3494245) as well
7
as the Novartis compound GNF6702, which have been disclosed in recent publications, there is
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
still an urgent need for new treatments with improved safety profiles, more straightforward
administration, lower costs and also with alternate modes of action.
One of the major challenges for drug discovery for VL is the lack of robustly validated drug targets
in Leishmania spp. For this reason, in order to identify suitable start-points for drug discovery,
compound libraries are screened directly against parasites in vitro, leading to phenotypically active
compounds with unknown mechanisms of action. An additional confounding factor in attempts to
identify chemical start-points is the fact that in the human host, parasites are found within
macrophages, where they reside within a parasitophorous vacuole. Therefore, relevant high-
content screens, suitable for HTS, require culture of parasites in macrophages, typically
8
differentiated THP-1 cells. For any compounds to be identified as hits in these screens, they would
be required to have suitable physicochemical properties to cross a number of cell membranes,
across various pH gradients, as well as having anti-parasitic activity. As a result, hit rates are
9
extremely low in these assays, typically below 0.1 %. Nevertheless, compound series that are
active in this assay and can be developed to give suitable properties for in vivo dosing, have a high
likelihood of success in rodent models of VL.
Utilising this intra-macrophage assay, an HTS screen of a GSK collection of 1.8M compounds
10
was performed against Leishmania donovani. The hits identified were screened in secondary
anti-parasitic assays, assessed for non-specific cytotoxicity in HepG2 cells, clustered and filtered
based on favourable physicochemical properties, resulting in the identification of 33 chemical
series and 75 singletons. One of the identified series was exemplified by 1 (table 1), an 8-hydroxy
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
naphthyridine which was shown to have potency against the parasite (pEC = 6.5) with >100-fold
5
0
selectivity over the human THP1 cell line (pEC = 4.5). This compared favourably to the current
5
0
treatments, amphotericin (pEC = 6.7) and miltefosine (pEC = 5.4). Upon further profiling, 1
5
0
50
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
was shown to have reasonable aqueous solubility (219 M), but low stability in mouse liver
microsomes (Cl = 18 ml/min/g). On this basis, 1 was selected as a suitable start point for a hit-
int
to-lead program.
Ld InMac a
THP-1 a
Cl
i
Solubility
ChromLogDpH7.4 11
Compound
(pEC50
)
(pEC50
)
(mouse, ml/min/g) b
(M) c
N
N
N
OH HN
N
6.5
4.5
16
219
4.1
Cl
1
Amphotericin
6.7
5.4
<6.0
<4.3
ND c
ND c
ND c
ND c
ND c
ND c
Miltefosine
a
Table 1. Ld InMac is the intramacrophage assay carried out in THP-1 cells with L. donovani amastigotes. Data are
the result of at least four independent replicates and standard deviations are ≤0.4. b Cli is mouse liver microsomal
c
C
intrinsic clearance. Aq. Solubility is kinetic aqueous solubility. ND means not determined
The principal aim of this work was to identify analogues of 1 with a suitable profile for dosing in
a mouse efficacy model of VL as rapidly as possible, in order to demonstrate that the series had
the potential to progress into lead-optimisation. Therefore, the initial chemistry program focused
on understanding the SAR of the series, with an aim of identifying compounds with improved
solubility and metabolic stability, as well as suitable potency for in vivo studies. Our targets were
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
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 achieve pEC > 5.8, aqueous solubility > 200 M and mouse liver microsomal clearance of <
5
0
5
.0 ml/min/g, as these criteria had been used previously to identify chemical series likely to have
in vivo efficacy. 12
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
Results and Discussion
Lack of knowledge regarding the molecular target of compound 1 made optimisation challenging,
with no guide as to the potential pharmacophore, or which vectors were most likely to positively
influence activity. We therefore focused on utilising tractable chemistry that would facilitate a
rapid exploration of SAR. Also, to maintain good solubility and hopefully improve metabolic
stability, we aimed to reduce, or at least maintain, the LogD of the initial analogues. This led us to
focus on the triazole substituent, as well as the 5-position of the naphthyridine as initial points for
exploration.
SAR of 7-triazolyl analogues
Variations to the benzyl substituent of the triazole (table 2), including substitutions on the
phenyl position (exemplified by 2), or on the methylene (exemplified by 3), led to a >10-fold loss
in potency when compared to 1, although 3 did show an improvement in metabolic stability. A
truncated analogue 4 was inactive, but we were encouraged by its improved solubility and
metabolic stability. We thus replaced the 4-chlorophenyl group of 1 with more polar substituents,
with the aim of regaining in vitro potency whilst maintaining a favourable ADME profile.
Morpholine substituted 5 and pyrrolidinone substituted 6 were synthesised and indeed proved to
be both soluble and metabolically stable, although both compounds were essentially inactive.
Switching to an amide as an isosteric replacement for the triazole was also investigated, and the
matched pairs (7 vs 1) showed similar levels of potency, although the amide did not appear to
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
show any clear advantage over the triazole as its mouse liver microsomal clearance was still not
below the targeted 5 ml/min/g.
We next switched attention to the naphthyridyl 5-position. Initially, nitrogen-linked
analogues were investigated. Whilst both methylamine 8 and dimethylamine 9 were essentially
inactive, larger amines such as p-chlorobenzylamine 10 and morpholine 11 both had pEC values
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
5
0
above 5, with good solubility and low clearance. Cyclic amides, such as 12, proved to be inactive,
as did sultam 13, presumably due to the reduced electron density in the aromatic ring. Compound
1
3 was of particular interest as 5-sultam substituted naphthyridines had been previously reported
1
3
in a series of integrase inhibitors and were shown to impart very good PK properties. Indeed, the
lead compound from this series progressed as far as Phase II clinical trials (compound 30, Table
5). Further exploration of N-linked analogues failed to deliver compounds with the necessary
potency for progression to in vivo studies, so we switched our focus to carbon-linked analogues.
Interestingly, the unsubstituted phenyl analogue 14 showed reasonable potency (pEC = 5.9), and
5
0
further analogues showed ortho-substitution to be beneficial, with the o-trifluoromethyl analogue
5 giving a significant increase in antiparasitic activity (pEC = 7.0), possibly driven by the
1
5
0
increased lipophilicity, and the o-methoxy analogue 16 showing a good balance of potency,
stability and solubility (pEC = 6.2, Cl = 1.4 ml/min/g and aqueous solubility > 250 M).
5
0
i
Replacement of phenyl by aromatic heterocycles (such as pyridyl or pyrazolyl) was also explored,
as was substitution on the other side of the naphthyridine (2, 3 and 4-positions), but these changes
led to only weakly active compounds (data not shown).
R1
N
N
N
R2
OH HN
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
R1
H
R2
Ld InMac a
THP-1 a
Cl
i
(mouse,
Aq Solubility
Compound
Chrom LogDpH7.4 11
ml/min/g) b
(M) c
(pEC50
)
(pEC50
)
Cl
1
6.5
4.5
16
219
4.1
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
H
H
H
H
H
ND d
1.9
3.3
0.8
<0.5
10
ND d
>250
>250
219
ND d
F
2
3
4
5
6
7
8
9
5.6
5.3
4.4
OH
ND d
<4.3
<4.3
<4.3
<4.3
<4.3
<4.3
<4.3
CH
3
<4.3
4.5
0.73
0.5
N
O
O
N
<4.3
6.3
>250
>250
>250
>250
ND d
Cl
N
H
N
N
ND d
OH
O
H
N
CH3
<4.3
4.8
3.1
14
-0.2
-0.6
N
CH
CH
3
3
Cl
1
0
5.8
<4.3
3.4
219
4.4
H
N
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
N
CH
3
11
5.0
<4.3
<4.3
<4.3
3.1
<0.5
0.6
>250
109
1.8
CH
CH
3
3
12
<4.3
<4.3
ND d
N
O
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
S O
O
13
1
250
1.5
N
CH
CH
CH
3
3
3
ND d
4
5.9
7.0
6.2
<4.3
<4.3
<4.3
9.4
1.4
0.8
160
CF3
15
16
>250
>250
4.3
3.2
O
a
Table 2. Ld InMac is the intramacrophage assay carried out in THP-1 cells with L. donovani amastigotes. Data are
b
the result of at least three independent replicates and standard deviations are ≤0.4. Cli is mouse liver microsomal
c
d
intrinsic clearance. Aq. Solubility is kinetic aqueous solubility. ND means not determined
Profiling of compound 16
Compounds 15 and 16 both showed a promising balance of potency, solubility and
metabolic stability. Due to having higher metabolic stability (in mouse liver microsomes) and
lower ChromLogD, compound 16 was progressed into a VL in vivo efficacy study, carried out in
7
our previously described VL mouse model . Mice were dosed orally with the standard anti-
leishmanial drug miltefosine, or with 16 dosed ip (intraperitoneal) two times daily for 5 days post
infection (although 16 had a suitable profile for oral dosing, we elected to dose ip to maximise
exposure and increase our chances of demonstrating in vivo proof-of-concept for the series).
Parasite load was determined in the livers of animals 3 days after cessation of treatment, and
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
parasite burden was expressed in Leishman Donovan Units (LDU, the mean number of
amastigotes per liver cell x mg weight of liver). The blood exposure of compound 16 was also
determined in dosed animals on days 1 and 5 in order to better understand the PK/PD relationship
of the series. According to our project criteria, a compound needs to reduce parasite burden by
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
>
>
70% before being considered suitable for progression to lead-optimisation, whilst a reduction of
1
2
95% would be considered suitable for a preclinical development candidate .
In the study, miltefosine behaved as expected, reducing parasite levels by >99% at 30 mg/kg qd.
After twice daily ip dosing at 50 mg/kg, compound 16 reduced parasite burden in mouse liver by
4
6%. This provided an early proof-of-concept for this series but fell short of our target of >70%
parasite reduction. Upon examining the blood samples taken on days 1 and 5, it was clear that 16
was rapidly cleared from blood, with unbound concentrations of compound exceeding EC only
9
9
during the first hour post-dose (table 3). Further examination of the samples revealed the presence
of glucuronidated adducts of 16, suggesting secondary metabolism as the key driver of the low
exposure.
AUC(0-t) M*min
C
max (M)
Tmax (h)
Day 1
Day 5
Day 1
Day 5
Day 1
Day 5
1
236
606
25.2
4.2
0.5
0.5
Table 3. Blood levels of 16, measured on days 1 and 5 of a mouse efficacy study, dosing i.p. at 50 mg/kg b.i.d.
Glucuronidation is a means of increasing water solubility of small molecules, facilitating
their elimination from the body in urine. It involves the transfer of the glucuronic acid component
of uridine diphosphate glucuronic acid to a suitable substrate, catalysed by UDP-
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
glucuronosyltransferase (UGT) and occurs mainly in the liver.14 Glucuronidation occurs at
nucleophilic sites such as R-OH, R-NH , or R-COOH, that can be present in the small molecule,
2
or generated via phase I metabolism. Due to the low rates of microsomal clearance of 16, alongside
our in vivo data, we surmised that glucuronidation was occurring on the parent compound, most
likely at the phenolic OH or the triazole N-H. Also, the observed reduction of parasite load, despite
the high in vivo clearance, suggested that reducing glucuronidation to increase the duration of
exposure above EC would be a key strategy to progress the series towards lead optimisation.
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
9
9
As a means of measuring glucuronidation in vitro, 16 was assessed in a mouse liver
hepatocyte assay. Unfortunately, 16 showed similar stability to that seen in mouse liver
microsomes (0.9 ml/min/g in microsomes vs. 1.4 ml/min/g in hepatocytes) with negligible amounts
of the glucuronide adduct being observed. This suggested that there was little involvement of the
hepatic UDP-glucuronosyltransferase in the in vivo phase II metabolism. This result made series
progression challenging, as there was no way to determine the potential impact of glucuronidation
without running an in vivo study, limiting our understanding of the SAR surrounding the observed
secondary metabolism.
One possible strategy to identify compounds with improved metabolic stability would be
to reduce lipophilicity. This had been shown previously to be a potential strategy for reducing
1
5
glucuronidation, however as shown in figure 1, this was very challenging within this compound
1
1
series. Looking at measured logD (Chrom LogD ), analogues with Chrom LogD values below
7
.4
7.4
3
were generally only weakly active, with pEC ’s above 5.8 only being achieved where Chrom
50
LogD was greater than 3. In our experience, this is a common problem when trying to optimise
7
.4
series phenotypically, where increasing potency without increasing lipophilicity is very
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
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
challenging; this highlights a key advantage of running structure enabled programs. Because of
this, alternative approaches to improving metabolic stability were required.
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
Figure 1: Chrom LogD7.4 vs. intramacrophage potency (Ld InMac pEC50) for all compounds presented. Green line
represents Chrom LogD7.4 = 3, black line represents Ld InMac pEC50 = 5.8
SAR of the naphthyridine core
Our initial approach to reduce potential phase II metabolism within the series was therefore
to investigate the SAR around the phenolic hydroxyl group and the triazole. Understanding which
of these features was important for anti-parasitic activity might enable us to synthesise potent
analogues with lower potential for glucuronidation. As shown in table 4, removal or methylation
of the naphthyridine 8-hydroxyl group led to a loss of anti-parasitic activity compared to 16 (17
and 18, respectively), although 18 was close to the targeted pEC of 5.8. Also, replacement of
5
0
ACS Paragon Plus Environment

Page 13 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
triazole with oxadiazole to remove the nucleophilic N-H led to a loss of activity (19). Synthesis of
other analogues to explore the naphthyridine SAR proved extremely challenging within the
triazole sub-series, so in order to more rapidly address this, we switched attention to the
bioisosteric replacement of triazole with amide; this change had previously been seen to have
limited effect on potency (e. g. 1 versus 7), and allowed much more straightforward synthesis of
the analogues of interest. Thus, deletion of the naphthyridine N-6 of 7 led to a compound that was
toxic to the host THP1 cells, and deletion of naphthyridine N-1 led to loss of anti-parasitic activity
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
(20 and 21, respectively). Moving to a scaffold that trapped the phenolic OH as a carbonyl removed
all anti-parasitic activity despite the parent amides having pEC values > 6.0 (comparing 22 to 28
50
and 23 to 27). Finally, we examined Raltegravir (24), an inhibitor of HIV integrase marketed as a
16
treatment for HIV. As shown in figure 2, Raltegravir contains the key acceptor-donor-acceptor
binding motif identified within the naphthyridine series. Although 24 was inactive in our in vitro
Leishmania assays, we surmised that transferring the known SAR of the naphthyridine core onto
the Raltegravir scaffold could be a viable strategy to regain activity. Unfortunately, as exemplified
by 25, none of the analogues based on this scaffold were active. From this round of synthesis, we
concluded that we were unlikely to identify active compounds without the acceptor-donor-acceptor
binding motif of the 7-substituted-1,6-naphthyridin-8-ol core.
Ld InMac a THP-1 Cl
(mouse, Aq. Solubility
i
Compound
Chrom LogDpH7.4 11
a
pEC50 ml/min/g) b
(M) c
pEC50
N
17
<4.3
<4.3
ND d
>250
ND d
N
N
HN
N
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
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
O
N
1
8
5.6
<4.3
<4.3
2
>250
>250
5.7
3.8
N
N
O
HN
N
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
O
N
N
1
2
9
0
<4.3
<0.5
O
N
OH
N
Cl
H
N
N
4.8
5.2
19
28
219
39
4.8
7.6
OH
O
Cl
N
H
N
21
22
<4.3
<4.3
OH
O
N
O
F
H
N
ND d
N
<4.3
<4.3
<4.3
<4.3
0.7
0.7
>250
>250
O
N
O
N
H
N
N
23
24
1.9
O
O
N
N
H
N
O
O
O
F
N
N
<4.3
<4.3
<4.3
<4.3
ND d
<0.5
>250
173
ND d
H
N
O
OH
O
O
F
N
N
H
N
25
5.5
O
OH
a
Table 4. Ld InMac is the intramacrophage assay carried out in THP-1 cells with L. donovani amastigotes. Data are
b
the result of at least three independent replicates and standard deviations are ≤0.4. Cli is mouse liver microsomal
c
d
intrinsic clearance. Aq. Solubility is kinetic aqueous solubility. ND means not determined
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
N
H
N
O
O
N
N
N
N
H
H
H
N
N
N
N
N
O
OH
N
OH
O
OH
O
N
A
A
A
A
D
A
D
A
D
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
Cl
Cl
F
24, raltegravir
1
7
Figure 2. Highlighting the acceptor-donor-acceptor motifs of triazole analogue 1, amide analogue 2 and Raltegravir
24
SAR of 7-carboxamide analogues
Previous SAR demonstrated that replacing the triazole with an amide led to compounds
such as 7 that retained anti-parasitic activity. Since 7 itself was metabolically unstable (Cl = 10
int
mg/ml/g), we became interested in transferring the SAR from the triazole sub-series (e.g. 16) to
investigate its’ impact on metabolic stability, with a particular focus on glucuronidation. To this
end, we synthesised 26 (table 5) as a direct analogue of triazole 16. Although it did not show
sufficient potency for progression into efficacy studies, it had reasonable aqueous solubility (219

M) and good metabolic stability (Cl = 0.9 ml/min/g) and was therefore progressed into a mouse
int
PK study to assess the extent of in vivo glucuronidation. After dosing (50 mg/kg i.p) and analysing
the metabolites generated (figure 3), there was little evidence of glucuronidation, with the major
metabolism observed being hydroxylation. As significant quantities of parent were still present 8
hours post-dose, this supported a strategy of switching to the amide series to reduce phase II
metabolism and improve in vivo exposure.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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
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
Figure 3: PK profile of 26, showing levels of parent, glucuronide adduct, major hydroxylated metabolite,
and other minor metabolites. Data is based on 2 replicates, with error bars representing the range of the data from
the 2 runs. 30µl blood/water (1:2) were collected following 50mg/kg i.p. administration to Balb C mice at prescribed
time points. 90µL ACN was then added to each sample and samples were centrifuged for 5min at 7000rpm, 90µl of
supernatant was removed and added to 50µL Milli-Q water before UPLC Q-TOF analysis.
With this in mind, we further explored the SAR of the amide sub-series, as shown in table
5
. We synthesised a set of benzyl amides, where 4-methoxy analogue 27 and -methyl-4-fluoro
analogue 28 both gave very encouraging profiles, meeting progression criteria in terms of potency,
metabolic stability and clearance. Alternatively, non-aromatic amides were explored, and although
none were identified with suitable profiles for progression, cyclopropylmethyl analogue 29 did
demonstrate reasonable potency. Due to its impressive in vitro potency, we selected 28 for
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
progression into an in vivo PK study. However, upon dosing (50 mg / kg i.p), the compound proved
to be toxic, rapidly giving symptoms (within 3.5 hrs) requiring termination of the experiment.
1
7
We noted that a related compound from Merck, L-870,810 (30), was reported as a clinical
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
candidate targeting HIV integrase, which progressed as far as Phase II clinical trials. Compound
3
0 was inactive in our in vitro efficacy assays, but the report, alongside the in vivo data for 26,
suggested that compounds with substitutions in the naphthyridine 5-position could have suitable
profiles for in vivo studies. Also, introducing substituents into the 5-naphthyridyl position had been
a successful strategy for improving metabolic stability and potency in the triazole sub-series. We
therefore investigated a range of 5-substituted naphthyridyl analogues (Table 5). THP-amine 31
and morpholine 32 lost potency compared to parent compound 7 so we switched attention back to
5-phenyl naphthyridines. Previously identified groups were combined (the 4-methoxyphenyl of 16
and the cyclopropylmethyl amide of 29), leading to 33 which unfortunately did not deliver the
hoped-for increase in potency. Further combinations of non-aromatic amides with differently
substituted 5-phenylnaphthyridines were synthesised, and whilst changing the phenyl substituent
gave flat SAR and no advantage over previous compounds, exploration of the amide led to 34 with
a trans-4-methoxycyclopropylamide. Compound 34 gave a good balance of potency, solubility
and microsomal stability and was therefore selected for a mouse PK study. Disappointingly, when
dosed at 50 mg/kg i.p, the mice again displayed the same symptoms as with compound 28, and the
study was terminated after 60 minutes.
R1
N
H
N
N
R2
OH
O
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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
Ld InMac a
THP-1 a
Cl
i
(mouse,
Aq. Solubility
Chrom
R1
R2
Compound
ml/min/g) b
(M) c
LogDpH7.4 11
(pEC50
)
(pEC50
)
N
N
N
OH HN
1
6.5
4.5
16
219
219
4.1
N
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
Cl
-
CH
3
26
5.0
4.4
0.9
4.2
O
Cl
ND d
H
H
7
6.3
6.3
<4.3
<4.3
10
>250
>250
O
F
27
2.6
4.2
5.1
H
28
6.8
5.3
3.7
>250
H
29
30
5.9
<4.3
4.5
13
ND e
3.3
F
O
S
ND d
N
O
<4.3
2.0
> 250
O
Cl
H2N
31
5.3
<4.3
<4.3
2.2
24
>250
219
5.3
O
N
Cl
ND d
3
2
<4.3
ND d
5.7
O
33
34
5.7
6.0
<4.3
4.4
8
>250
>250
O
1.4
O
a
Table 5. Ld InMac is the intramacrophage assay carried out in THP-1 cells with L. donovani amastigotes. Data are
b
the result of at least three independent replicates and standard deviations are ≤0.4. Cli is mouse liver microsomal
c
d
intrinsic clearance. Aq. Solubility is kinetic aqueous solubility. ND means not determined
Profiling of 16, 28 and 34
Due to the encouraging results within the series (16 showing low level efficacy and 26
giving a good PK profile), we were keen to understand the origins of the observed toxicity of 28
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 34. As previously mentioned, compound 30 progressed as far as Phase II clinical trials as an
HIV integrase inhibitor, and although it was inactive against VL, we were keen to assess whether
there was scope to progress our related series further. To this end, two studies were conducted in
parallel; a screen against receptors with known links to toxicity and investigation of mode of action
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
(
MoA).
Firstly, the compounds which were poorly tolerated in mice, 28 and 34, were screened
against a panel of >30 receptors with known links to in vivo toxicity (GSK enhanced cross-screen
1
8
panel (eXP)). Compound 28 gave a pIC value of 5.4 against monoamine oxidase A, highlighting
5
0
a slight risk of drug-drug interactions and possible side-effects, and a pIC value of 4.9 in a
5
0
phenotypic cell health assay, suggesting possible effects on mitochondrial integrity that could lead
to an increased risk of hepatotoxicity. Compound 34 also showed potency in the cell health assay,
alongside activity in a bile salt export pump (BSEP) assay (pIC value of 4.8, hepatotoxicity risk)
5
0
and a phospholipidosis assay. From this, it was not clear whether the effects seen in the receptor
screen were related to the observed in vivo toxicity.
Alongside these screens, mode of action studies were initiated focusing on 10, 16 and 28,
as representatives of both the triazole and amide sub-series. We were particularly interested in
confirming that the compounds inhibited a shared target, identifying off-target effects and
understanding the source of the observed toxicity. The results of these studies have been reported
previously and demonstrated that the compounds act as non-specific chelators of divalent metal
2+
2+
2+
cations, in particular Zn , Fe , and Cu , and that this property is likely responsible for their anti-
1
9
parasitic activity. Indeed, the propensity of these compounds to non-specifically chelate divalent
cations may well explain the observed in vivo toxicity associated with this series.
Chemistry
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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 access the required analogues a number of different approaches were needed, as illustrated in
schemes 1 – 3. Compounds 1 and 2 were previously reported 20 and synthesis of the remaining 7-
triazolyl-8-hydroxy naphthyridines started from 7-cyanonaphthyridine 35 20 (scheme 1), where
cyclisation with substituted hydrazides under acidic conditions led to compounds 3 – 6. To access
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
-substituted analogues, 35 could be brominated with NBS to give 36 and subsequently cyclised
to give 37 which was treated with a suitable amine to give 8 – 11. For 12 and 13, protection of the
phenol proved to be necessary. Hence, 36 could be tosyl protected to give 38, followed by either
Buchwald-Hartwig coupling to introduce the pryrrolidinone (12) or copper coupling with 1,2-
thiazinane 1,1-dioxide to introduce the sultam (13). Alternatively, 36 could be directly coupled
with phenylboronic acid to give 14, or benzyl protected to give 39, which could be coupled to give
1
5 and 16.
N
H
R2
N
N
R1
N
OH
N
N
H
N
N
R1
=
OH
N
Br
N
CH3
a
Br
c
d
b
N
N
R2
H
=
4
H
N
H
N
N
OH
3
N
CN
N
R1
N
N
NH2
O
OH
N
CN
OH
N
N
8
9
O
O
OH
N
35
36
37
H
O
N
N
5
6
N
Cl
10
11
R
e
g
h
N
H
N
N
OH
O
N
N
R2
Br
Br
R2 =
R =
f
i
N
N
N
S
O
O
H
N
N
N
N
CN
N
CN
N
CF3
MeO
12
13
OTs
OBn
OH
N
N
14
15
16
39
38
a
Scheme 1. Reagents and conditions: (a) 1,4-dioxane / acetic acid, 140 – 200 °C, 22 – 38 %; (b) NBS, DCM, 2 h,
4 %; (c) acetohydrazide, 1,4-dioxane / acetic acid, 200 °C, 30 min, 45 %. (d) relevant amine, DIPEA, NMP, 180
C, 1 – 18 h, 25 – 50 %; (e) TsCl, NEt , DCM, 40 °C, 3 h, 83 %; (f) i. for 12: Pyrrolidin-2-one, Cs CO , Pd(OAc)
Xantphos, 1,4-dioxane, 65 °C. For 13: 1,2-thiazinane 1,1-dioxide, Cs CO , Pd (dba) , Xantphos, 1,4-dioxane, 65
°C. ii. NaOMe, DMF, 50 °C. iii. acetohydrazide, 1,4-dioxane / acetic acid, 180 °C, 2h, 7 – 16 % over 3 steps; (g)
BnBr, PPH , DIAD, THF, 0 °C, 18 h, 15 %; (h) Phenylboronic acid, K PO , Pd(dppf)Cl ·CH Cl , DMF / water, 130
8
°
3
2
3
2
,
2
3
2
3
3
3
4
2
2
2
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
°
C then acetohydrazide, 1,4-dioxane / acetic acid, 200 °C, 45 min, 3 % over 2 steps (14); (i) i. relevant boronic acid,
3 4
K PO , tetrakis, 1,4-dioxane / water, 120 °C, 1.5 h. ii. acetohydrazide, 1,4-dioxane / acetic acid, 180 °C, 18 h, 22 %
(15) / 24 % (16) over 2 steps.
To fully explore the SAR around the naphthyridine ring, and also to introduce alternative
heterocycles to replace the triazole, a number of analogues required bespoke synthesis. Analogues
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
1
7, 18, 20, 21 were synthesised according to established procedures and are described in the
supplementary information, with the synthesis of 19, 22, 23 and 25 highlighted in scheme 2. Thus,
5-bromo-8-methoxy-1,6-naphthyridine-7-carboxylic acid 40 was cyclised with acetyl hydrazide to
give 41, followed by Suzuki coupling and deprotection of the 8-methoxy group to give oxadiazole
1
9. 22 was synthesised from 3-chloropicolinic acid via conversion to the acid chloride,
condensation with ethyl 3-(dimethylamino)acrylate, cyclisation with methylamine and ester
hydrolysis to give 42. This was then coupled with 4-fluoro-α-(R)-methylbenzylamine to give 22.
In order to access 23, methyl 3-fluoropicolinate was condensed with t-butyl acetate to give 43,
which was treated with 4-acetamidobenzenesulfonyl azide (ABSA), PBu then DBU /
3
iodomethane to give cyclised 44 which was subsequently hydrolysed and coupled with 4-
methoxybenzylamine to give 23. Finally, 25 was synthesised according to a previously published
route such that 2-cyanoanisole was treated with N-methylhydroxylamine and cyclised to 45.
Thermal rearrangement gave 46 which was treated with 4-Fluoro-α-(R)-methylbenzylamine to
give 25.21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
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
Br
O
Br
O
b
O
N
d
a
N
N
OH
O
N
N
N
N
N
O
O
N
OH
N
40
41
1
9
N
F
N
O
F
Cl
O
c
e
+
H
N
OH
H2N
OH
N
N
N
N
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
O
O
O
42
22
O
O
f
O
O
OMe
g
N
O
OMe
O
F
N
N
N
+
H
O
O
H2N
N
O
N
N
F
N
O
43
44
23
F
j
O
h
O
i
O
+
F
H2N
N
N
O
N
N
N
N
H
N
CN
O
CO2Et
O
O
O
OH
O
OH
46
O
EtO2C
2
5
4
5
a
Scheme 2. Reagents and conditions: (a) i. EDCI, DIPEA, HOBt, acethydrazide, DCM, 0 °C, 18 h. ii. PPh
NEt , MeCN, 60 °C, 2 h, 53 %. (b) i. Phenylboronic acid, Na CO , tetrakis, 1,4-dioxane / water, 120 °C, 18 h. ii.
Trimethylchlorosilane, NaI, MeCN, reflux, 18 h, 44 % over 2 steps; (c) i. Oxalyl chloride, DCM, RT then 3-ethoxy-
3 4
, CCl ,
3
2
3
3
1
-oxo-propanoic acid, BuLi, -78 °C, 1 h. ii. DMF.DMA, DMF, 100 °C, 1 h. iii. MeNH
M NaOH, MeOH, RT, 2 h, 69 % over 4 steps; (d) DIPEA, T P, DMF, 2h, 8 %; (e) LDA, t-butyl acetate, THF, -78
, MeCN, 0 °C, 18 h. ii. PBu , THF, RT, 15 min. iii. DBU, MeI, MeCN, RT, 18 h,
2 % over 3 steps. (g) i. TFA, DCM, RT, 18 h. ii. EDCI, HOBt, DIPEA, DCM, 16 h, 88 % over 2 steps. (h) i.
CO , water / EtOH, 28 h, 80 °C. ii. Diethyl acetylenedicarboxylate, EtOH, RT, 1 h, 61 % over 2
2
, ether / EtOH, RT, 1 h. iv.
3
°
C, 2 h, quant. (f) i. ABSA, NEt
3
3
3
MeNH-OH, Na
2
3
steps. (i) p-Xylene, reflux, 2 days, 41 %; (j) Toluene, reflux, 18 h, 46 %.
The 7-carboxamide analogues were synthesised from the corresponding ester 47 according to
scheme 3. Amides 7 and 27 – 29 were synthesised directly from the ester by treating with the
relevant amine at high temperature. To access the desired 6-functionalised analogues, 47 was
brominated with NBS to give 48, tosyl protected to give 49 and converted to sultam 30. 22
Alternatively, 48 could be treated with 4-chlorobenzylamine to give 51a, then coupled with
another amine to give 31 and 32. Compound 49 could also be coupled with 2-
methoxyphenylboronic acid and hydrolysed to give 50, which was coupled with trans-4-
methoxycyclohexylamine to give 34. Alternatively, 48 could be treated directly with a relevant
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
amine to give 51b-c, with subsequent Suzuki coupling with 2-methoxyphenylboronic acid giving
2
6 and 33.
Cl
F
OMe
N
H
R =
N
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
N
R
7
27
29
OH
O
28
a
S
O
O
O
N
Br
F
N
N
N
d
H
N
N
CO2Me
N
N
CO2Me
OH
OH
30
OTs
49
47
h
b
c
R =
CH3
Br
i
OMe
OMe
26
33
N
N
N
H
N
N
CO2Me
N
CO2H
N
R
OH
48
OH
OH
O
O
34
50
e
f
O
Cl
Br
HN
R =
a
R
g
31
N
Cl
H
N
R
O
N
H
N
N
CH3
N
OH
51 a-c
O
N
b
c
OH
O
32
a
Scheme 3. Reagents and conditions: (a) i. EtOH, reflux, 18 h, 65 %; (b) NBS, DCM, RT, 2 h, 93 %; (c) TsCl,
NEt
5 °C, 24 h. ii. NaOMe, DMF, 50 °C, 24 h. iii. 4-Fluorobenzylamine, T
steps; (e) Relevant amine, EtOH, 100 °C, 18 h, 45 – 86 %; (f) (2-methoxyphenyl)boronic acid, K
Pd(dppf)Cl .DCM, 1,4-dioxane / water, 140 °C, 1 h, 30% (26) / 38 % (33); (g) Relevant amine, DIPEA, NMP, 150
180 °C, 5 h, 17 % (31) / 33 % (32); (h) i. (2-methoxyphenyl)boronic acid, tetrakis, Na CO , 1,4-dioxane / water,
0 °C, 18 h. ii. NaOMe, MeOH, 50 °C, 15 min. iii. LiOH, water, 60 °C, 18 h, 58 % over 3 steps; (i) trans-4-
3
, DCM, 40 °C, 2 days, 49 %; (d) i. 1,2-thiazinane 1,1-dioxide, Cs
2
CO
P, DIPEA, DMF, RT, 18 h, 11 % over 3
PO
3
2 3
, Pd (dba) , Xantphos, 1,4-dioxane,
6
3
3
4
,
2
–
6
2
3
methoxycyclohexanamine, DIPEA, T
3
P, 50 °C, 18 h, 56 %.
Conclusion
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In order to identify new compound series with the potential to be developed as new therapeutics
for VL, a collection of 1.8M compounds from the GSK corporate collection was screened against
L. donovani in vitro. One hit series identified from this, exemplified by 1, was selected for a hit-to
lead program. In vivo studies of an early compound, 16, demonstrated that the series had the
potential to reduce parasite burden, but that glucuronidation was a potential barrier to series
progression. Scaffold hopping from the core triazole to an amide was a key strategy for progressing
the series, leading to 28 with a very good in vitro profile. Dosing of 28 identified an issue with
toxicity for the series and further chemistry failed to identify compounds that did not carry this
liability. MoA studies suggested that the anti-parasitic activity, and the toxicity, was likely driven
by chelation of divalent metal cations. Based on these finding we concluded that attempting to
develop compounds within this series that would divorce anti-parasitic activity from inherent
toxicity would be extremely challenging and unlikely to succeed. With this in mind, work on the
series was halted. This demonstrates the importance of understanding the mode of action from a
very early stage in the drug discovery process, when working to progress phenotypically active hit
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
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION:
Chemistry
Chemicals and solvents were purchased from the Aldrich Chemical Company, Fluka, ABCR,
VWR, Acros, Fluorochem and Alfa Aesar and were used as received. Air- and moisture-
sensitive reactions were carried out under an inert atmosphere of argon in oven-dried
glassware. Analytical thin-layer chromatography (TLC) was performed on pre-coated TLC
plates (layer 0.20 mm silica gel 60 with fluorescent indicator UV254, from Merck).
Developed plates were air-dried and analysed under a UV lamp (UV254/365 nm). Flash
column chromatography was performed using pre-packed silica gel cartridges (230-400
mesh, 40–63 m, from SiliCycle) using a Teledyne ISCO Combiflash Companion, or
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
1
13
Combiflash Retrieve. H NMR and C NMR spectra were recorded on a Bruker Avance DPX
1
13
500 spectrometer ( H at 500.1 MHz, C at 125.8 MHz). Chemical shifts () are expressed in
ppm recorded using the residual solvent as the internal reference in all cases. Signal
splitting patterns are described as singlet (s), doublet (d), triplet (t), quartet (q), multiplet
(m), broad (b), or a combination thereof. Coupling constants (J) are quoted to the nearest
0
.1 Hz. LC-MS analyses were performed with either an Agilent HPLC 1100 series connected
to a Bruker Daltonics MicrOTOF, or an Agilent Technologies 1200 series HPLC connected to
an Agilent Technologies 6130 quadrupole LC/MS, where both instruments were connected
to an Agilent diode array detector. Mobile phase was water/acetonitrile + 0.1% HCOOH, or
water/acetonitrile + 0.1% NH3; linear gradient 80:20 to 5:95 over 3.5 min, and then held
–
1
for 1.5 min; flow rate 0.5 mL min . All intermediates had a measured purity ≥ 90% and all
assay compounds had a measured purity of ≥ 95% as determined using this analytical LC-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
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
MS system (TIC and UV). High resolution electrospray measurements were performed on a
Bruker Daltonics MicrOTOF mass spectrometer. Microwave-assisted chemistry was
performed using a Biotage Initiator Microwave Synthesizer.
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-(3-(Hydroxy(phenyl)methyl)-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-8-ol (3)
35 (100 mg, 0.58 mmol) and 2-hydroxy-2-phenylacetohydrazide (288 mg, 1.74 mmol) in
1
,4-dioxane (1.5 mL) / acetic acid (0.5 mL) were heated in a microwave at 140 °C for 2 h.
The mixture was cooled to RT and the resulting solid collected, washed with 1,4-dioxane and
1
dried under vacuum to give 3 (43 mg, 0.13 mmol, 22%). H NMR (DMSO-d ):  14.90 (s, 1H),
6
1
2.29 (bs, 1H), 9.15 (s, 1H), 9.02 (s, 1H), 8.63 – 8.56 (m, 1H), 7.82 – 7.74 (m, 1H), 7.56 – 7.50
+
(
m, 2H), 7.42 – 7.24 (m, 3H), 6.23 (bs, 1H), 5.91 (bs, 1H). m/z [M+H] calcd for C H N O ,
17 14
5 2
3
20.1147; found, 320.1146.
7
-(3-Methyl-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-8-ol (4)
3
5 (30 mg, 0.16 mmol) and acetylhydrazide (58.4 mg, 0.79 mmol) in 1,4-dioxane (2 mL) /
acetic acid (0.2 mL) were heated in a microwave at 200 °C for 15 min. The mixture was
cooled, solvent evaporated and crude material purified by mass directed prep. HPLC to give
1
4
(10 mg, 0.04 mmol, 27%). H NMR (DMSO-d ):  14.74 (s, 1H), 12.45 (s, 1H), 9.18 – 9.15
6
1
3
(
m, 1H), 9.04 (s, 1H), 8.63 – 8.58 (m, 1H), 7.83 – 7.76 (m, 1H), 2.45 (s, 3H); C NMR (DMSO-
+
d ):  158.8, 158.5, 154.5, 149.5, 143.9, 141.9, 136.7, 125.6, 124.8, 12.8; m/z [M+H] calcd
6
for C H N O, 228.0885; found, 228.0886
1
1
10
5
7-(3-(Morpholinomethyl)-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-8-ol (5)
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
5 (150 mg, 0.88 mmol) and 2-(4-morpholinyl)acetohydrazide (419 mg, 2.63 mmol) in 1,4-
dioxane (3 mL) / acetic acid (0.3 mL) were heated in a microwave at 180 ºC for 2 h. The
resulting solution was evaporated and cyclohexane (3x 10 mL) added and evaporated.
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
Acetone (15 mL) was added and the resulting solid collected by filtration, washed with
1
MeOH and dried to give 5 (105 mg, 0.34 mmol, 38%). H NMR (DMSO-d ):  14.83 (s, 1H),
6
9
.18 – 9.16 (m, 1H), 9.04 (s, 1H), 8.63 – 8.60 (m, 1H), 8.15 (s, 1H), 7.82 – 7.78 (m, 1H), 3.73
+
(s, 2H), 3.63 – 3.58 (m, 4H), 2.60 – 2.54 (m, 4H, under solvent peak); m/z [M+H] calcd for
C H N O , 313.1413; found, 313.1411
1
5
17
6 2
1
-((5-(8-Hydroxy-1,6-naphthyridin-7-yl)-1H-1,2,4-triazol-3-yl)methyl)pyrrolidin-2-one
(6)
6
Was synthesized by an analogous method to 3, from 35 (100 mg, 0.58 mmol) and 2-(2-
oxopyrrolidin-1-yl)acetylhydrazide (273 mg, 1.74 mmol). Crude material was purified by
1
mass-directed prep. HPLC to give 6 (70 mg, 0.15 mmol, 26%). H NMR (DMSO-d ):  9.20 –
6
9
.17 (m, 1H), 9.06 (s, 1H), 8.65 – 8.61 (m, 1H), 7.83 – 7.80 (m, 1H), 4.61 (s, 2H), 3.47 – 3.44
+
(m, 2H), 2.34 – 2.29 (m, 2H), 2.04 – 1.95 (m, 2H). m/z [M+H] calcd for C H N O ,
15 15
6 2
3
11.1256; found, 311.1260
N-(4-Chlorobenzyl)-8-hydroxy-1,6-naphthyridine-7-carboxamide (7)
A mixture of 47 (51 mg, 0.25 mmol) and 4-chlorobenzylamine (701 mg, 0.5 mmol) in EtOH
4 mL) was stirred at 80 °C for 18 h. The hot reaction mixture was poured into a solution of
(
acetic acid (0.5 ml) in water (4 mL). After stirring for 10 min, cold water (15 mL) was added
and stirred at RT for 40 min. The resulting solid was collected, washed with water and dried
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
to give 7 (42 mg, 0.13 mmol, 53 %). H NMR (DMSO-d ):  13.67 (s, 1H), 9.18 – 9.16 (m, 1H),
6
8
6
1
.93 (bs, 1H), 8.61 (d, J = 7.7 Hz, 1H), 7.86 – 7.82 (m, 1H), 7.44 – 7.39 (m, 4H), 4.55 (d, J =
13
.5 Hz, 2H). C NMR (DMSO-d ):  170.1, 154.6, 142.6, 138.4, 136.5, 132.0, 129.9, 128.8,
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
+
27.0, 126.0, 125.4, 42.1; m/z [M+H] calcd for C H N O Cl, 314.0696; found, 314.0695
1
6
13
3 2
7-(3-Methyl-1H-1,2,4-triazol-5-yl)-5-(methylamino)-1,6-naphthyridin-8-ol (8)
A mixture of 37 (100 mg, 0.33 mmol), methylamine (327 L, 0.653 mmol, 2N in THF) and
DIPEA (0.171 mL, 0.980 mmol) in NMP (2 mL) was stirred at 180 ºC for 18 h in a sealed
tube. Further methylamine (327 L, 0.653 mmol, 2N in THF) was added and stirring
continued for a further 18 h. Solvent was evaporated, DCM (3 mL) added and the resulting
solid collected and purified by mass directed prep. HPLC to give 8 (21 mg, 0.82 mmol,
1
2
7
2
5%). H NMR (DMSO-d ):  14.06 (s, 1H), 10.97 (s, 1H), 8.94 – 8.91 (m, 1H), 8.54 (d, J =
6
.9 Hz, 1H), 7.57 – 7.51 (m, 1H), 7.29 (bs, 1H), 2.97 (d, J = 3.8 Hz, 3H), 2.33 (s, 3H); m/z
57.2 [M + H]+.
5
-(Dimethylamino)-7-(3-methyl-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-8-ol (9)
A mixture of 37 (100 mg, 0.33 mmol), dimethylamine (0.327 mL, 0.653 mmol, 2 M in THF)
and DIPEA (171 L, 0.98 mmol) in NMP (2 mL) was stirred at 180 ºC for 18 h in a sealed
tube. Further dimethylamine (0.327 mL, 0.653 mmol, 2 M in THF) was added and stirring
continued for a further 18 h. Solvent was evaporated and the crude residue purified by flash
chromatography (0 – 50% EtOAC:EtOH (3:1) / cyclohexanes) to give 9 (34 mg, 0.13 mmol,
1
3
8%). H NMR (DMSO-d ):  14.20 (s, 1H), 11.47 (bs, 1H), 8.99 – 8.96 (m, 1H), 8.46 (d, J =
6
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
8
2
.2 Hz, 1H), 7.61 – 7.56 (m, 1H), 2.93 (s, 6H), 2.35 (s, 3H); m/z [M+H] calcd for C H N O,
13 15 6
71.1307; found, 271.1309
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
5
1
-((4-Chlorobenzyl)amino)-7-(3-methyl-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-8-ol (10)
0 Was synthesised by an analogous method to 8 from 37 (50 mg, 0.15 mmol), 4-
chlorobenzylamine (42 mg, 0.29 mmol) and DIPEA (57 mg, 0.44 mmol) to give 10 (30 mg,
1
0
8
.074 mmol, 50%). H NMR (DMSO-d ):  14.18 (bs, 1H), 11.12 (bs, 1H), 9.04 – 9.01 (m, 1H),
6
.75 – 8.72 (m, 1H), 7.90 – 7.85 (m, 1H), 7.67 – 7.62 (m, 1H), 7.54 – 7.50 (m, 2H), 7.37 – 7.33
1
3
(m, 2H), 4.89 (s, 2H), 2.43 (s, 3H); H NMR (DMSO-d ):  158.8, 154.9, 153.5, 149.3, 144.5,
6
+
1
41.2, 140.6, 132.4, 131.4, 130.1, 128.4, 122.7, 121.2, 116.1, 43.4, 14.0; m/z [M+H] calcd
for C H N OCl, 367.1074; found, 367.1088
18 16
6
7
-(3-Methyl-1H-1,2,4-triazol-5-yl)-5-morpholino-1,6-naphthyridin-8-ol (11)
A mixture of 37 (50mg, 0.16 mmol), DIPEA (90 L, 0.49 mmol) and morpholine (28 L, 0.32
mmol) in NMP (0.5 mL) were heated in a microwave to 200 °C for 1 h. After cooling, the
crude mixture was purified by prep. HPLC to give 11 (31 mg, 0.073 mmol, 46%) as the TFA
1
salt. H NMR (DMSO-d ):  9.13 (dd, J = 1.6 and 4.2 Hz, 1H), 8.59 (dd, J = 1.6 and 8.5 Hz, 1H),
6
7
.77 – 7.73 (m, 1H), 3.90 – 3.85 (m, 4H), 3.34 – 3.30 (m, 4H), 2.52 (s, 3H, under DMSO signal).
m/z 313.2 [M + H]+.
1
-(8-Hydroxy-7-(3-methyl-1H-1,2,4-triazol-5-yl)-1,6-naphthyridin-5-yl)pyrrolidin-2-one
(12)
ACS Paragon Plus Environment