Structure activity relationships of 4-hydroxy-2-pyridones: A novel class of antituberculosis agents

Article abstract of DOI:10.1016/j.ejmech.2015.10.008

Pyridone 1 was identified from a high-throughput cell-based phenotypic screen against Mycobacterium tuberculosis (Mtb) including multi-drug resistant tuberculosis (MDR-TB) as a novel anti-TB agent and subsequently optimized series using cell-based Mtb ass

Full text of DOI:10.1016/j.ejmech.2015.10.008

Accepted Manuscript
Structure Activity Relationships of 4-Hydroxy-2-pyridones: a Novel Class of
Antituberculosis Agents
Pearly Shuyi Ng, Ujjini H. Manjunatha, Srinivasa PS. Rao, Luis R. Camacho, Ngai
Ling Ma, Maxime Herve, Christian G. Noble, Anne Goh, Stefan Peukert, Thierry T.
Diagana, Paul W. Smith, Ravinder Reddy Kondreddi
PII:
S0223-5234(15)30289-0
10.1016/j.ejmech.2015.10.008
EJMECH 8143
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 15 June 2015
Revised Date: 2 October 2015
Accepted Date: 6 October 2015
Please cite this article as: P.S. Ng, U.H. Manjunatha, S.P. Rao, L.R. Camacho, N.L. Ma, M. Herve, C.G.
Noble, A. Goh, S. Peukert, T.T. Diagana, P.W. Smith, R.R. Kondreddi, Structure Activity Relationships
of 4-Hydroxy-2-pyridones: a Novel Class of Antituberculosis Agents, European Journal of Medicinal
Chemistry (2015), doi: 10.1016/j.ejmech.2015.10.008.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical abstract
1

ACCEPTED MANUSCRIPT
Structure Activity Relationships of 4-Hydroxy-2-
pyridones: a Novel Class of Antituberculosis Agents
Pearly Shuyi Ng,† Ujjini H. Manjunatha,† Srinivasa P S Rao,† Luis R. Camacho,† Ngai Ling Ma,†
Maxime Herve,† Christian G. Noble,† Anne Goh,† Stefan Peukert,† Thierry T. Diagana,† Paul W.
Smith† and Ravinder Reddy Kondreddi,*†
†Novartis Institute for Tropical Diseases, 10 Biopolis Road #05-01 Chromos Singapore 138670
KEYWORDS. 4-Hydroxy-2-pyridones, Tuberculosis, Antituberculosis agents, Structure activity
relations, phenotypic screen.
1

ACCEPTED MANUSCRIPT
Abstract. Pyridone 1 was identified from a high-throughput cell-based phenotypic screen
against Mycobacterium tuberculosis (Mtb) including multi-drug resistant tuberculosis (MDR-
TB) as a novel anti-TB agent and subsequently optimized series using cell-based Mtb assay.
Preliminary structure activity relationship on the isobutyl group with higher cycloalkyl groups at
6-position of pyridone ring has enabled us to significant improvement of potency against Mtb.
The lead compound 30j, a dimethylcyclohexyl group on the 6-position of the pyridone, displayed
desirable in vitro potency against both drug sensitive and multi-drug resistant TB clinical
isolates. In addition, 30j displayed favorable oral pharmacokinetic properties and demonstrated
in vivo efficacy in mouse model. These results emphasize the importance of 4-hydroxy-2-
pyridones as a new chemotype and further optimization of properties to treat MDR-TB.
Introduction
Tuberculosis (TB) caused by the bacterium Mycobacterium tuberculosis (Mtb), is one of the
most deadly bacterial infectious diseases. It has been reported that 2 billion people are latently
infected and approximately 1 in 10 latent infections will progress to the active disease. There
were over 8.6 million cases of TB in 2012, which resulted in nearly 1.3 million deaths.1 The
most widely used chemotherapy for uncomplicated TB is a combination of isoniazid, rifampicin,
pyrazinamide, and ethambutol. This complex regimen still requires treatment for at least 6
months, often compromising TB patient compliance. Emergence of various forms of drug-
resistant strains of Mtb and co-infection with HIV make the TB more difficult to treat.2
Nearly 400, 000 new cases of multi drug resistant TB (MDR-TB) are reported every year,
which are resistant to isoniazid and rifampicin. MDR-TB further acquired additional resistance
to fluoroquinolones and at least one of the injectable drugs leading to the emergence of
2

ACCEPTED MANUSCRIPT
extensively-drug resistant TB (XDR-TB).1 Second-line drugs used in the treatment of MDR and
XDR-TB are less effective and not well tolerated. Although a number of novel chemical agents
such as, PA-824, SQ109, oxazolidinones, BTZ043, Q203, NITD304 3, 4 etc., are fueling the TB
pipeline in the last decade, only bedaquiline (TMC207)5 and delamanid (OPC67683)6 registered
as TB drugs. Thus, there is an urgent unmet medical need to combat drug-resistant forms of Mtb
with new chemical entities.
Cell-based high-throughput screen of nearly 2.2 million compound collection using Mtb
H37Ra resulted in ~ 8000 hits. Pyridone 1 presented as an attractive hit series based on the in
vitro pharmacokinetic properties and moderate potency against Mtb (MIC50 = 1.54 µM).
Pyridone 1 has a low molecular weight (243 Da), good aqueous solubility (550 µM at pH 6.8),
moderate cLogP (3.04), good microsomal stability (CLint rat/human/mouse = 3.4/20/22
µL/min/mg) and did not inhibit major cytochrome P450 isoforms CYP 3A4, 2D6 and 2C9 (IC50
>20 µM). In the present study, we describe the structure activity relationship (SAR) and
structure property relationship (SPR) of 4-hydroxy-2-pyridones to improve their potency against
Mtb.
CHEMISTRY
Pyridone 1 was prepared as shown in Scheme 1 via a condensation of bis(2,4,6-
trichlorophenyl) 2-phenylmalonate 2 and ethyl 3-amino-5-methylhex-2-enoate 38 followed by
hydrolysis and decarboxylation with NaOH as reported earlier.9 The intermediate 2 was prepared
from commercial available 2-phenylmalonic acid and oxalyl chloride as reported in literature.10
3

ACCEPTED MANUSCRIPT
Scheme 1. Synthesis of 4-hydroxy-2-pyridone 1a
aReagents and conditions: (a) neat or Dowtherm, 230 °C, 30 min (b) 2 M NaOH solution, 130
°C, 24 h, 14% (two steps).
Quinolinone 5 was obtained via condensation of bis(2,4,6-trichlorophenyl) 2-phenylmalonate 2
with aniline 4 under microwave conditions (Scheme 2). Direct condensation of diethyl 2-
phenylmalonate 6 with 3-methylbutanimidamide hydrochloride 711 in presence of NaOMe
yielded pyrimidinone 8. Treatment of commercial available 2-phenylmalonic acid with oxalyl
chloride afforded 3-oxo-2-phenylacryloyl chloride 9, which upon condensation with silyl enol
ether 10 gave pyranone 11. Both 3-oxo-2-phenylacryloyl chloride 9 and silyl enol ether 10 were
prepared as described previously.12
4

ACCEPTED MANUSCRIPT
Scheme 2. Synthesis of pyridone ring modificationsa
aReagents and conditions: (a) Dowtherm, MW, 210 °C, 15 min, 40% (b) NaOMe, 2-methoxyethanol,
reflux, 72 h, 6% (c) Xylene, reflux, 10 h, 17%.
The N-methyl substituted pyridone 15 was prepared in three steps from the corresponding
unsubstituted pyridone 1 (Scheme 3). The phenolic OH of 1 was converted to acetate 12 using
acetyl chloride. Methylation of acetate 12 with K2CO3 and MeI provided the key intermediates
13 and 14 which were converted to 15 and 16 with NaOMe/MeOH. Both N-isopropyl and N-
benzyl substituted pyridones 19 and 20 were obtained in a single step. Condensation of bis(2,4,6-
trichlorophenyl) 2-phenylmalonate 2 with 17 or 18 under neat conditions gave corresponding N-
substituted pyridones 19 or 20 respectively.
5

ACCEPTED MANUSCRIPT
Scheme 3. Synthesis of N-substituted 4-hydroxy-2-pyridone analoguesa
O
O
O
O
O
O
N
b
+
N
H
O
N
O
OMe
Me
13
12
14
a
c
c
OH
OH
OH
N
H
O
N
O
N
OMe
Me
1
15
16
Cl
Cl Cl
Cl
OH
O
O
R
d
NH
O
Cl
Cl
+
O
O
O
N
R
O
17: R = Isopropyl
18: R = benzyl
19: R = Isopropyl
20: R = benzyl
2
aReagents and conditions: (a) acetyl chloride, pyridine, 1,4-dioxane, rt, 2 h, 63% (b) MeI,
K2CO3, MW, 100 °C, 30 min, 66% (c) NaOMe, MeOH, rt, 30 min, 46% (15) and 38% (16) (d)
neat, 200 °C, 1 h, 38% (19) and 37% (20).
Dehydroxy analog 22 was obtained from 1 by conversion of hydroxy to chloro 21 followed by
dehalogenation with 10% Pd-C. Methoxy compound 23 was obtained directly from 1 by
treatment with diazomethane under basic conditions (Scheme 4). The 4-amino analog 25 was
prepared via a two-step sequence starting from 1. Heating of 1 with benzylamine under
microwave conditions provided 24, which underwent reductive debenzylation using 10% Pd-C in
6

ACCEPTED MANUSCRIPT
presence of hydrogen to afford the 4-amino analog 25. The corresponding sulfonamide analog 26
was obtained by treating 25 with methanesulfonyl chloride and pyridine.
Scheme 4. Synthesis of C4-substituted analoguesa
Cl
f
N
O
N
H
O
H
21
22
e
OH
OMe
a
N
H
O
N
H
O
1
23
b
O
O
NH
NH2
S
NH
d
c
N
H
O
N
H
O
N
H
O
24
25
26
aReagents and conditions: (a) TMS-CH=N2, DIPEA, 9:1 MeCN/MeOH, rt, 4 h, 33% (b)
benzylamine, MW, 220 °C, 10 h, 27% (c) H-Cube, 10% Pd/C, EtOH, 65 °C, 4 min (1 mL/min),
60% (d) MeSO2Cl, pyridine, CH2Cl2, rt, 30 min, 26% (e) SOCl2, PCl5, 1,2-dichloroethane,
reflux, 24 h, 23% (f) H2, 10% Pd/C, EtOH, rt, 1 h, 52%.
The synthetic procedures for the preparation of 3-substituted pyridones 28a-n are depicted in
Scheme 5. The substituted malonic esters 27a-n were either obtained from commercial sources
7

ACCEPTED MANUSCRIPT
or synthesized from diethyl malonate following the literature procedure.10 The corresponding
pyridones 28a-n were obtained via a condensation-decarboxylation sequence either under
microwave conditions or by conventional heating. Decarboxylation of the intermediate
carboxylates generally proceeded well under basic conditions with the exception of pyridyl
substituted pyridones 28l-n which required acidic conditions.
Scheme 5. Synthesis of 3-substituted pyridones from aminocrotonate and substituted malonic
estersa
aReagents and conditions: (a) (27a-b, 27d-k): (i) neat, 220 °C, 45 min (ii) 2 M NaOH solution,
MW (160 °C), 1 h (b) (27c): (i) neat or Dowtherm, 250 °C, 30 min (ii) 2 M NaOH solution, 130
°C, 24 h (c) (27l-n): (i) neat or Dowtherm, 250 °C, 2 h (ii) 2 M HCl solution, 130 °C, 16 h.
Ethyl 3-amino-4-substituted but-2-enoates 29a, 29c-l were prepared over a 4-step sequence
utilizing Meldrum’s acid as a key building block as reported in a previous procedure.8
Intermediate 29b was prepared according to the literature.13 The synthesis of 6-substituted
pyridones 30a-l (Scheme 6) was accomplished by utilizing the same condensation-
decarboxylation sequence as described earlier in the synthesis of 3-substituted pyridones.
Scheme 6. Synthesis of 6-substituted pyridones from substituted aminocrotonate and
diethylphenylmalonate a
8

ACCEPTED MANUSCRIPT
aReagents and conditions: (a) (30i, 30a, 30c-f, l): (i) neat, 220 °C, 45 min (ii) 2 M NaOH
solution, MW (160 °C), 1 h (b) (30b, 30g, 30h, 30j): (i) neat or Dowtherm, 250 °C, 30 min (ii) 2
M NaOH solution, 130 °C, 24 h (c) (30k): (i) neat or Dowtherm, 250 °C, 2 h (ii) 2 M HCl
solution, 130 °C, 16 h.
Results and discussion
Considering the promising microbiological profile and favorable in vitro PK properties of 4-
hydroxy-2-pyridone, the medicinal chemistry strategy aimed at improving the antitubercular
activity. The SAR investigations focused on the (1) role of pyridone core (2) importance of
planar 3-phenyl group (3) significance of phenolic OH and (4) importance of lipophilic isobutyl
group. The new pyridone analogs were screened in vitro against the M. tuberculosis H37Rv in
order to obtain the MIC values using turbidometric assay and cytotoxicity against HepG2 cells
was measured for selected compounds.
Pyridone Core Modification. Our initial efforts focused on compounds 1 to understand the
importance of the pyridone core for the anti-TB activity. Changing of pyridone to quinolinone 5,
introducing nitrogen at the 5-position of the pyridone 8 or a NH replacement with oxygen 11
reduced potency (Table 1). Aromatization of the pyridone to 2-methoxy pyridine 16, also
resulted in loss of anti-TB activity. Methyl substituted analog 15 at the NH position resulted in
decreased potency against Mtb, whilst, bulky N-substituted isopropyl 19 and benzyl 20 groups
were inactive (Table 1).
9

ACCEPTED MANUSCRIPT
Table 1. Anit-TB activity of pyridone ring modifications and N-substituted 4-hydroxy-2-
pyridones
Compound
Structure
Mtb MIC50
(µM)
1.54
1
13.2
>20
5
8
13.8
8.1
11
15
16
>20
10

ACCEPTED MANUSCRIPT
>20
19
20
>20
Importance of phenolic OH. Modification of the phenolic OH to NH2 25 or NHBn 24 led to
the loss of anti-TB activity which could be due to significant decrease in pKa from acidic OH
(pKa: 6.58) to basic NH2 (pKa: 4.7) or NHBn (pKa: 3.32). Modified Moka14 version 2.5.4 was
used to calculate pKa values. Other replacements of the phenolic OH with H 22 or OMe 23 or
the bioisostere NHSO2Me 26 also resulted in significant loss of potency (Table 2). The above
results indicated that the pyridone core, phenolic OH and NH are important for anti-TB activity.
Table 2: SAR on phenolic OH
Compound
R4
Mtb MIC50
(µM)
1.54
>20
1
OH
H
22
23
24
25
26
> 20
> 20
>20
>20
OMe
NHBn
NH2
NHSO2Me
11

ACCEPTED MANUSCRIPT
Modifications on the 3-Position of Pyridones. Replacement of the 3-position phenyl group
with cyclohexyl 28a or benzyl 28b reduced potency, possibly indicating that the planarity of the
phenyl group could be essential for anti-TB activity (Table 3). A scan of ortho 28c, meta 28d or
para 28e fluoro substitutions on the 3-position of the pyridone retained similar range of anti-TB
activity compared to unsubstituted pyridone 1(Table 3). Substitution on the 3- phenyl with ortho
28f or meta 28g chloro substitution maintained anti-TB activity similar to 1. By contrast, the
para-chloro substituted compound 28h was inactive (Mtb MIC > 20 µM). The fact that chloro
and fluoro substitutions at the ortho and meta positions were tolerated, but not the bulkier chloro
substitution on the para position suggests that narrow space for substituents. Introducing an
electron-donating group OMe on the ortho 28i, meta 28j or para 28k positions of the phenyl
group or replacing with 2/3/4 pyridyl 28l-n also led to complete loss of anti-TB activity. Overall,
the phenyl group was found to be optimal for Mtb potency on the 3-position of 4-hydroxy-2-
pyridones.
Table 3: SAR of 3-substituted 4-hydroxy-2-pyridones
OH
R3
N
O
H
Compou
nds
R3
Mtb MIC50 (µM)
1.54
1
12

ACCEPTED MANUSCRIPT
>20
>20
28a
28b
6.31
2.69
28c
28d
6.75
3.32
28e
28f
2.67
28g
>20
>20
28h
28i
13

ACCEPTED MANUSCRIPT
>20
28j
>20
>20
28k
28l
>20
>20
28m
28n
Modifications to the 6-Position Isopropyl Group. We further explored scope and
possibilities at the pyridone 6-position to improve anti-TB activity. Pyridones with hydrogen 30a
and dimethyl amine 30b substitution at the 6-position were inactive compared to the isopropyl
group in pyridone 1. Modifications with isobutyl 30c and phenyl 30d substitutions showed anti-
TB activity similar to 1. Other groups like cyclopropyl 30e, cyclobutyl 30f, cyclopentyl 30g and
difluorocyclohexyl 30h on the 6-position retained anti-TB activity (Table 4). Further, cyclohexyl
30i and dimethylcyclohexyl 30j substitutions at the 6-position of the pyridone significantly
improved anti-TB activity by 1 and ~2 log orders compared to 1, respectively . These cycloalkyl
substitutions also showed increased lipophilicity, with cLogP increasing from 1.58 to 5.27.
Furthermore, a plot of cLogP against Mtb MIC for 6-substituted 4-hydroxy-2-pyridones
indicated a positive correlation of cLogP with anti-TB activity with an r2 of 0.91 (Figure 1). The
loss of activity of 30a (cLogP 1.58) and 30b (cLogP 0.92) could be due to the decreased
14

ACCEPTED MANUSCRIPT
lipophilicity of the methyl and dimethyl amino groups (Figure 1). Previously Dannhardt and his
coworkers9 reported activity of 30b against Mtb H37Ra as 80-160 µM. However, in that study
no in vitro and in vivo pharmacokinetic parameters were described. In order to improve
solubility, 4-pyridyl 30k and 4-pyranyl 30l groups were explored on the pyridone 6-position.
Although 30k and 30l showed significant improved solubility significantly, they were inactive .
Unlike the pyridone core and the 3-position phenyl modifications, substitutions at the 6-position
of the pyridone significantly increased anti-TB activity. The lead compound 30j is the most
potent 4-hydroxy-2-pyridone analog with Mtb potency of 0.05 µM. The in vitro potency of 30j is
superior to isoniazid (MIC50 = 0.33 µM), PA824 (MIC50 = 0.4 µM) and comparable to
bedaquiline (MIC50 = 0.05 µM). The lead compound 30j and the other 4-hydroxy-2-pyridone
analogs also showed potent activity against MDR-TB clinical isolates7, suggesting a potential
role of this series of compounds in treating MDR-TB infections.
Table 4: SAR of 6-substituted 4-hydroxy-2-pyridones and their physicochemical properties
Compounds
R6
Mtb MIC50 (µM)
cLogP
Solubility µM
pH 6.8
Isoniazid
PA-824
Bedaquiline
1
--
--
--
0.33
0.4
-0.67
2.79
7.2
--
--
0.05
1.54
3.04
550
15

ACCEPTED MANUSCRIPT
> 20
> 20
1.58
0.92
>1000
671
30a
30b
1.51
1.40
3.57
3.15
128
40
30c
30d
4.53
1.19
2.56
3.12
891
280
30e
30f
0.92
1.32
3.68
3.62
144
104
30g
30h
0.16
0.05
6.80
4.24
5.27
1.66
24
25
30i
30j
30k
>1000
16

ACCEPTED MANUSCRIPT
10.1
1.84
>1000
30l
A difference of in vitro potency and PK properties was observed between the hit 1 and lead
compound 30j. Hit 1 had moderate potency with good Lipinski’s rule-of-five compliance, whilst,
the lead compound 30j had increased potency and cLogP (5.27). A similar positive correlation
between lipophilicity and anti-TB activity was also observed in our earlier series indole-2-
carboxamides15 and a few other scaffolds.16 The lead compound 30j exhibited low solubility (25
µM) and moderate in vitro microsomal clearance (Clint- mouse/human = 45/59 µL/min/mg).
Figure 1. Correlation plot of Mtb MIC50 with cLogP for 6-substituted 4-hydroxy-2-pyridones.
Mechanism of Action of Pyridones: Detailed mechanism of action studies unambiguously
demonstrated InhA, Mtb enoyl-ACP-reductase, as the molecular target of 4-hydroxy-2-
17

ACCEPTED MANUSCRIPT
pyridones.7 InhA is the clinically validated target of isoniazid and is essential for mycolic acid
biosynthesis.17 Recently, we have reported that 1 and 30j bind to InhA only in presence of
reduced cofactor (NADH) and inhibits the enzyme activity with an IC50 of 9.6 µM and 0.57 µM
respectively (Table 5). A co-crystal structure of 30j complexed with InhA and NADH (PDB:
4R9S) revealed that (i) the pyridone core has a pi-pi stacking interaction with dihydropyridine
ring of NADH ii) the phenolic hydroxy group makes a hydrogen bond interaction with the
hydroxy group of Tyr-158 and the 2’-OH of the ribose sugar of NADH iii) the 6-position
dimethylcyclohexyl group occupies a hydrophobic enoyl-ACP substrate binding pocket iv) the
nitrogen of the pyridone core interacts with sulphur of methionine (Figure 2).7
Y158
M155
K165
F149
Y158
NADH
L218
M199
NADH
C
A
I215
G96
NADH
B
Figure 2. Co-crystal structure of pyridone 30j with InhA-NADH complex. A: Phenolic OH
hydrogen bonding with Tyr-158 and 2’-OH of the ribose sugar in NADH. B: A narrow binding
pocket for 3-phenyl para position. C: Dimethylcyclohexyl interacting with lipophilic amino
acids.
18

ACCEPTED MANUSCRIPT
A retrospective analysis using X-ray structure of the 30j–InhA-NADH complex (Figure 2) was
employed to rationalize the SAR of 4-hydroxy-2-pyridones. The hydrogen bond interaction
observed in the 30j co-crystal structure potentially explains the importance of the phenolic OH
(Figure 2A). The space available for substituents at 4-position is limited due to the protein
environment and bulky substitutions like OMe 23, NHBn 24 and SO2NHMe 26 showed a
detrimental effect on the anti-TB activity as shown in Table 2.
The 3-phenyl ring lies at an angle to the pyridone and occupies a small pocket with limited
room to accommodate ortho- and para-substituents. In particular, substitution on the para
-
position appears particularly not favored could be due to the steric clash with the Gly-96
backbone carbonyl oxygen (Figure 2B). There appears to be some space within the binding
pocket to accommodate a small meta-substituent, which is consistent with the activity of meta-Cl
analogue 28g. Although space at the ortho position is limited, the ortho-Cl analogue (28f) also
retained similar anti-TB activity compared to unsubstituted phenyl group. By contrast, ortho
,
meta and para OMe substituted analogues (28i-k) and pyridyl analogues (28l-n) were devoid of
anti-TB activity. This is most likely due to both unfavorable electronic as well as steric reasons.
The SAR on the 6-postion of the pyridone is consistent with X-ray structure of the 30j–InhA-
NADH complex. The hydrophobic 4,4-dimethylcyclohexyl moiety in the compound 30j is
positioned in the hydrophobic cavity formed by the amino acids Phe-149, Met-155, Tyr-158,
Met-199, Ile-215, Leu-218 (Figure 2C). The cumulative effect of these interactions increased the
potency of the compound 30j (Mtb MIC50 = 0.05 µM). Further it explains lower potency of other
less lipophilic substitutions (30a and 30b) on the 6-position of pyridone. SAR of the pyridone
analogs described in this manuscript was optimized by cell-based activity but not with InhA
enzyme inhibition. The cell permeability and drug efflux mechanisms are important factors in
19

ACCEPTED MANUSCRIPT
cell-based activity. Especially, permeation of compounds through mycobacterial cell wall is
more challenging due to the presence of waxy outer layer.18 It is worth noting that earlier efforts
to identify direct InhA inhibitors to overcome KatG mediated resistance have yielded many
inhibitors that block the lipid-binding site similar to 4-hydroxy-2-pyridone analogues.19-22
In vivo PK and mouse efficacy. None of the 4-hydroxy-2-pyridone analogues described in this
study showed cytotoxicity against mammalian HepG2 cells (CC50 > 20µM). Further, in vivo PK
of 1 and 30j were evaluated in mice by intravenous and oral routes. Both compounds exhibited
low total systemic clearance (13 to 18 ml/min/kg) and low volume of distribution (0.5 L/kg) with
an elimination half-life of ~1.2 hours (Table 5). Compound 1 and 30j achieved reasonable oral
exposure with an oral bioavailability of 55-66%. Despite the low volume of distribution, the
plasma concentrations of both 1 and 30j exceeded their Mtb MICs, thus we subjected these
candidates for in vivo acute mouse efficacy studies.
Despite the moderate in vitro potency, compound 1 demonstrated a statistically significant 0.95
log lung CFU reduction at 150 mg/kg compared to the untreated control (Figure 3A). Compound
30j showed a dose dependent 1 and 1.92 log lung CFU reductions at 30 and 100 mg/kg
respectively (Figure 3B). Similarly, compound 30j showed 2.5 and 2.8 log CFU reduction in
spleen with 30 and 100 mg/kg respectively (Figure 3C).7 The in vivo efficacy of 30j at 100 mg/
kg dose was comparable to the first-line TB drug ethambutol (100 mg/kg). These results showed
that the 4-hydroxy 2-pyridone class has the potential to be active against Mtb not only in vitro
but also under in vivo conditions.16
Table 5: Pharmacological and biological properties of 1 and 30j
Compound 1
Compound 30j
IVPK
5
0.46
POPK
50
-
IVPK
5
0.54
POPK
25
-
PK properties
Dose (mg/kg)
Vss (liters/kg)
20

ACCEPTED MANUSCRIPT
CL (ml/min/kg)
Elim T1/2 (hours)
18
1.21
-
-
13.6
1.13
-
-
Cmax
AUC (
Tmax (hours)
F (%)
(
ꢀ
M)
-
-
-
-
13.15
27.18
0.5
-
-
-
-
13.73
65.79
6
ꢀ
M hour)
55
66
Biological properties
InhA IC50 (
Mtb MIC50 (
HepG2 CC50 (
ꢀ
ꢀ
M)
M)
M)
9.6
1.54
>20
0.57
0.05
>20
ꢀ
MDRTB MIC range (
ꢀM)
2.5-3.13
0.04-0.16
IVPK, Intravenous pharmacokinetics; POPK, per oral pharmacokinetics; -, not applicable; Cmax
,
maximal plasma concentration; AUC, area under the curve (t = 0 to 24 h); F: absolute oral
bioavailability; Vss, volume of distribution at steady state; CL, clearance from plasma; T1/2,
terminal elimination half-life (non-compartmental estimate); InhA EC50, InhA enzyme
inhibition.7
Figure 3. In vivo antimycobacterial activity of 4-hydroxy-2pyridones. Mean log CFU reductions
per lung in mice treated with 1 (A), 30j (B) and mean log CFU reductions per spleen in mice
treated with 30j (C). Ethambutol (EMB) was used as positive control in all the studies. Statistical
evaluation was done using one way ANOVA and the data was analyzed using Tukey’s multiple
comparison test. Statistical significance (*) was accepted with P values <0.05.
SUMMARY
A 4-hydroxy-2-pyridone 1 was selected as one of the promising hits from a phenotypic high
throughput screen of 2.2 million library compounds against Mtb in a cell based assay. The
preliminary SAR on 6-position of 4-hydroxy-2-pyridones with more lipophilic cycloalkyl groups
21

ACCEPTED MANUSCRIPT
significantly improved anti-TB activity by two log orders of magnitude. The improved in vitro
activity of lead compound 30j was translated to a better in vivo efficacy against Mtb in mouse
model. At the time of the hit to lead stage, the molecular target of 4-hydroxy-2-pyridones was
not identified; hence, the SAR was primarily driven by cell-based assay against Mtb. Later, the
mechanism of action studies indicated that 4-hydroxy-2-pyridones are direct inhibitors of InhA.
However, lead compound 30j was not further optimized based on the co-crystal structure
information and InhA enzyme assay. The results described here suggest that 30j is an attractive
lead compound having better in vitro potency and in vivo efficacy7 for further optimization based
on structure guided drug design to improve its drug-like properties. Furthermore, the binding
mode of 4-hydroxy-2-pyridones could open up new opportunities for medicinal chemists to
design new scaffolds active against this important Mtb target.
EXPERIMENTAL
Chemistry. Reagents and solvents were purchased from Aldrich, Acros, or other commercial
sources and used without further purification. Thin layer chromatography (TLC) was performed
on precoated silica gel 60 F254 plates from Merck. Compounds were visualized under UV light,
ninhydrin, or phosphomolybdic acid (PMA) stain. NMR spectra were obtained on a Varian 400
MHz Mercury plus NMR or Bruker 400 MHz Ultrashield spectrometer using CDCl3 and DMSO-
d6 as solvents.
Compound purity was determined by method 1: LC/MS using an Agilent UHPLC 1290
coupled with API 3200; Acquity UPLC BEH C18 column, 1.7µm, 2.1x50 mm; gradient of 98:2
H2O (0.1% formic acid):CH3CN to 2:98 H2O (0.1% formic acid):CH3CN for 2 minutes run time
with 1.0 mL/min flow rate / method 2: Waters Acquity UPLC coupled with MS waters ZQ
22

ACCEPTED MANUSCRIPT
2000; Acquity UPLC BEH C18 column, 1.7µm, 2.1x50mm; gradient of 98:2 H2O (0.1% formic
acid):CH3CN to 2:98 H2O (0.1% formic acid):CH3CN for 2 minutes run time with 1.0 mL/min
flow rate / method 3: Waters Quattro Micro UPLC-LCMS equipped with a Acquity BEH C18
column, 1.7µm, 2.1x50mm using a gradient of 90:10 H2O (0.025% trifluoroacetic acid):CH3CN
(0.025% trifluoroacetic acid) to 10:90 H2O (0.025% trifluoroacetic acid):CH3CN (0.025%
trifluoroacetic acid) for 5 minutes run time with 0.4 mL/min flow rate and HPLC purity using
method 4: Waters Acquity UPLC equipped with a Acquity UPLC HSS T3 column, 1.8µm,
2.1x50mm; gradient of 95:5 H2O (0.1% formic acid):CH3CN to 2:98 H2O (0.1% formic
acid):CH3CN for 2 minutes run time with 1.0 mL/min flow rate / method 5: Waters UPLC
equipped with a Acquity, BEH C18 column, 1.7µm, 2.1x100mm using a gradient (4 minutes) of
70:30 H2O (0.025% trifluoroacetic acid):CH3CN (0.025% trifluoroacetic acid) to 20:80 H2O
(0.025% trifluoroacetic acid):CH3CN (0.025% trifluoroacetic acid) for 6 minutes run time with
0.3 mL/min flow rate [method 1, method 2 and method 4 were used for compounds 13, 14, 15,
16, 23, 24, 25, 26, 28a, 28b, 28d, 28e, 28f, 28g, 28h, 28i, 28j, 28k, 30a, 30c, 30d, 30e, 30f and
30l and method 3 and method 5 were used for compounds 1, 5, 8, 11, 19, 20, 21, 22, 28c, 28l,
28m, 28n, 30b, 30g, 30h and 30k].
Minimum Inhibitory Concentration (MIC) Determination. The M. tuberculosis H37Rv
(ATCC 27294) strain, maintained in Middlebrook 7H9 broth medium supplemented with 0.05%
Tween 80 and 10% ADS supplement, was used for MIC determination. Compounds dissolved in
90% DMSO were 3-fold serial diluted in duplicates in 384-well clear plates. A volume of 50
of early log-phase culture of Mtb diluted to an OD600 of 0.02 was added to each well, and the
ꢀL
assay plates were incubated at 37 °C for 5 days. Absorbance was recorded using a Spectramax
M2 spectrophotometer, and MIC50, the concentration required to inhibit growth by 50%, was
23

ACCEPTED MANUSCRIPT
determined by plotting nonlinear curves using GraphPad Prism 5 software (Rao et al., 2013).
Isoniazid was used as the internal control in all of our assays, which had an MIC50 of 0.33 ꢀM.
Solubility. Solubility was measured using a high-throughput equilibrium solubility (HT-Eq
sol) assay using a novel miniaturized shake-flask approach and streamlined HPLC analysis.
Briefly, a DMSO sample stock solution from a source plate was transferred to an MPV (miniprep
filer vials, Whatman) chamber and subjected to centrifugal evaporation to remove DMSO.
Aqueous buffer was added to the MPV chamber, and the MPV plunger was inserted until the
membrane on the bottom of the plunger touched the surface of the solution in the chamber. The
assembled MPV were incubated on an orbital shaker for 24 h and the plunger was then pushed to
the bottom of the chamber followed by shaking for another 30 min. The filtrate was analyzed
using HPLC.23
Mouse, Rat, and Human in Vitro Metabolic Stability. The metabolic stability of drug
candidates was determined in mouse, rat, and human liver microsomes using a compound
depletion approach and was quantified by LC/MS/MS. The assay measured the rate and extent of
metabolism of chemical compounds by measuring the compounds in vitro half-life (t1/2) and
hepatic extraction ratios (ER) and predicted metabolic clearance in all species.24, 25
In vivo PK. All animal experiments were approved by the Institutional Animal Care and Use
Committee (IACUC) of the Novartis Institute for Tropical Diseases. Mice were allowed to
acclimatize before the initiation of pharmacokinetic (PK) experiments. Food and water were
given ad libitum. Compounds were formulated at a concentration of 2.5 mg/mL for a dose of 25
mg/kg given orally (p.o.) and at 1 mg/mL concentration for a dose of 5 mg/kg given
intravenously (i.v.). Compound 1 and 30j were formulated using 20% 1,2-propandiol, 10%
24

ACCEPTED MANUSCRIPT
solutol HS15, and top up with pH 7.0 buffer for i.v. For oral administration, compound 1 was
formulated in 0.5% Tween 80 and 0.5% Methyl cellulose and 30j was formulated in MEPC
(10% w/w Capmul MCM, 25% w/w Cremophor-RH40, 15% w/w Ethanol and 50% w/w 50 mM
Phosphate buffer pH 6.8). Blood samples were collected at 0.08, 0.25, 0.75, 1.5, 3, 8, and 24 h
following oral dosing. The sampling schedule was identical following i.v. dosing except that the
first time point was 0.02 h. Groups of three mice were used for each time point. Blood was
centrifuged at 13ꢁ000 rpm for 7 min at 4 °C, and plasma was harvested and stored at − 20 °C
until analysis.
Extraction and LCMS Analysis. Plasma samples were extracted with
acetonitrile/methanol/acetic acid (90:9.8:0.2) containing warfarin as an internal standard using a
8.8:1 extractant-to-plasma ratio. Analyte quantitation was performed by LC/MS/MS. Liquid
chromatography was performed using an Agilent 1200 HPLC system coupled with a API4000
triple quadrupole mass spectrometer (Applied Biosystems) using an Agilent Zorbax Phenyl (3.5
ꢀm, 4.6 × 75 mm) column at 45 °C. Instrument control and data acquisition were performed
using Applied Biosystems software Analyst, version 1.4.2. The mobile phases used were (A)
0.2% of acetic acid in water (99.8:0.2, v/v) and (B) 0.1% of acetic acid in methanol (99.9:0.1,
v/v) using a gradient with a flow rate of 1.0 mL/min and run time of 5 min. The lower limit of
quantification ranged between 1 and 3 ng/mL.
Pharmacokinetic Analysis. The mean value from the three animals at each time point was
plotted against time to give plasma concentration–time profile. Pharmacokinetic parameters were
determined using Watson LIMS, version 7.2 (Thermo Electron Corporation), by
noncompartmental analysis. The oral bioavailability (F) was calculated as the ratio between the
25

ACCEPTED MANUSCRIPT
area under the curve (AUC) following oral administration and the AUC following intravenous
administration corrected for dose (F = (AUCp.o. × dosei.v.)/(AUCi.v. × dosep.o.).
In Vivo Efficacy. In vivo acute mouse efficacy studies were carried out as described earlier
with minor modifications.15 Briefly, on day 0, animals were infected by an intranasal route with
103 Mtb H37Rv bacilli. For initial in vivo efficacy evaluation with compound 1, one week post
infection animals were treated for 3 weeks (7 days a week) and for a lead candidate 30j animals
were treated for 4 weeks (7 days a week). In both experiments, ethambutol was used as positive
control. After treatment, lungs were aseptically removed and homogenized, and the bacterial
load was estimated by serial-dilution plating on 7H11 agar plates. Compound 1 and 30j were
formulated similar to in vivo pharmacology studies. All procedures involving mice were
reviewed and approved by the institutional animal care and use committee of the Novartis
Institute for Tropical Diseases.
4-Hydroxy-3-phenylquinolin-2(1H)-one (5). A mixture of bis(2,4,6-trichlorophenyl)-2-
phenylmalonate 2 (250 mg, 0.47 mmol) and aniline 4 (22 mg, 0.23 mmol) in dowtherm (5 mL)
was heated under microwave at 210 °C for 15 min. The reaction mixture was diluted with diethyl
ether (20 mL) and the precipitated solid was collected by filtration. The solid was washed with
diethyl ether (5 mL) and dried to afford 4-hydroxy-3-phenylquinolin-2(1H)-one 5 (45 mg, 40%
yield) as an off-white solid. 1H NMR (400 MHz, DMSO-d6):
= 8 Hz, 1H), 7.52-7.30 (m, 7H), 7.20-7.15 (m, 1H). ESI MS: m/z 238 [M+H]+. HPLC
δ 11.48 (s, 1H), 10.09 (s, 1H), 7.93
(d,
J
purity: 99.9%.
6-Hydroxy-2-isobutyl-5-phenylpyrimidin-4(3H)-one (8). To a suspension of 3-
methylbutanimidamide hydrochloride 7 (500 mg, 3.68 mmol) and diethyl phenylmalonate 6
26

ACCEPTED MANUSCRIPT
(1.73 g, 7.35 mmol) in 2-methoxyethanol (10 mL) was added NaOMe (0.79 g, 14.7 mmol) and
stirred at reflux temperature for 72 h. The reaction mixture was cooled to 0 °C and the
precipitated solid was filtered. The solid was taken up in water and acidified with 1 M HCl to pH
~6 and extracted with 10% methanol in chloroform (2 x 20 mL). The combined organic layer
was evaporated under reduced pressure and washed with diethyl ether (5 mL) to afford 6-
hydroxy-2-isobutyl-5-phenylpyrimidin-4(3H)-one 8 (60 mg, 6% yield) as an off-white solid. 1H
NMR (400 MHz, DMSO-d6):
δ
12.01 (br s, 2H), 7.56-7.54 (m, 2H), 7.28-7.24 (m, 2H), 7.15-7.12
(m, 1H), 2.39 (d,
[M+H]+. HPLC purity: 97.6%.
J = 7.4 Hz, 2H), 2.15-2.09 (m, 1H), 0.93 (d, J = 6.5 Hz, 6H). ESI MS: m/z 245
4-Hydroxy-6-isobutyl-3-phenyl-2H-pyran-2-one (11). To a cold solution of trimethyl(4-
methylpent-1-en-2-yloxy)silane 10 (714 mg, 4.15 mmol) in xylene (5 mL), was added a solution
of 3-oxo-2-phenylacryloyl chloride 9 (500 mg, 2.77 mmol) in xylene (5 mL) dropwise at rt. The
resulting mixture was refluxed for 10 h before it was cooled to 0 °C. The precipitated solid was
filtered and washed with pentane. The crude solid was re-precipitated in methanol to afford 4-
hydroxy-6-isobutyl-3-phenyl-2H-pyran-2-one 11 (120 mg, 17% yield) as an off-white solid. 1H
NMR (400 MHz, CDCl3):
δ
7.53-7.40 (m, 5H), 6.26 (br s, 1H), 5.98 (s, 1H), 2.37 (d,
J = 7.1 Hz,
2H), 2.2-2.1 (m, 1H), 0.99 (d,
J
= 6.6 Hz, 6H). ESI MS: m/z 245 [M+H]+. HPLC purity: 99.7%.
6-Isobutyl-2-oxo-3-phenyl-1,2-dihydropyridin-4-yl acetate (12). To a suspension of 1 (50.8
mg, 0.20 mmol) in 1,4-dioxane (3 mL) and cooled to 0 °C was added acetyl chloride (16 µL,
0.219 mmol) and pyridine (18.5 µL, 0.230 mmol). The mixture was allowed to gradually warm
to room temperature and stirred at room temperature for 2 hr. The reaction mixture was
concentrated in vacuo and the residue was taken up in CH2Cl2 (3 mL). The organic layer was
washed with water, brine, dried over Na2SO4 and concentrated in vacuo to give light yellow
27

ACCEPTED MANUSCRIPT
residue. The residue was purified by column chromatography (ISCO Combiflash®, 4 g silica gel
column, 0-40% EtOAc/cyclohexane) to give pyridone 12 as white solid (37.6 mg, 63% yield). 1H
NMR (400 MHz, CDCl3):
δ
7.47-7.33 (m, 5H), 6.21 (s, 1H), 2.50 (d,
J = 7.28 Hz, 2H), 2.15-2.09
(m, 1H), 2.07 (s, 3H), 1.00 (d,
J
= 6.53 Hz, 6H). ESI MS: m/z 286 [M+H]+.
6-Isobutyl-1-methyl-2-oxo-3-phenyl-1,2-dihydropyridin-4-yl acetate (13) & 6-isobutyl-2-
methoxy-3-phenylpyridin-4-yl acetate (14). To a solution of 12 (37.6 mg, 0.132 mmol) in dry
MeCN (2 mL) was added K2CO3 (18.0 mg, 0.132 mmol) and MeI (11 µL, 0.172 mmol). The
resultant mixture was heated at 100 °C for 30 minutes in Biotage microwave reactor. Reaction
mixture was cooled and diluted with EtOAc (4 mL). The organics were washed sequentially
with water and brine, dried over Na2SO4, filtered and concentrated in vacuo to give colorless oil.
The crude material was purified by column chromatography (ISCO Combiflash®, 4 g silica gel
column, 0-50% EtOAc/cyclohexane) to give 13 (26 mg, 66% yield) and pyridine 14 (7 mg, 17%
yield). 13: 1H NMR (400 MHz, CDCl3):
δ
7.16 - 7.32 (m, 5H), 5.83 (s, 1H), 3.45 (s, 3H), 2.42
= 6.40 Hz, 6H). ESI MS: m/z
7.37 - 7.43 (m, 2H), 7.30-7.37 (m, 3H), 6.56
J = 7.28 Hz, 2H), 2.12-2.24 (m, 1H), 2.00 (s, 3H), 0.99 (d, J = 6.53
(d,
J = 7.20 Hz, 2H), 1.91 (s, 3H), 1.83 - 1.88 (m, 1H), 0.94 (d, J
258 [M-Ac]+. 14: 1H NMR (400 MHz, CDCl3):
δ
(s, 1H), 3.91 (s, 3H), 2.59 (d,
Hz, 6H). ESI MS: m/z 258 [M-Ac]+.
4-Hydroxy-6-isobutyl-1-methyl-3-phenylpyridin-2(1H)-one (15). To a solution of 13 (26 mg,
0.08 mmol) in MeOH (2 mL) was added 30% NaOMe in MeOH (0.2 mL, 10%v/v) at rt. The
resultant mixture was stirred at rt for 30 minutes before it was concentrated under vacuo to give a
white residue. The white residue was taken up in EtOAc (3 mL) and washed with 10% citric
acid solution, dried over Na2SO4, filtered and concentrated in vacuo to give a white residue. The
crude material was dissolved in MeOH and purified on reversed-phase HPLC using solvent
28