Identification and optimization of a new series of anti-tubercular quinazolinones

Article abstract of DOI:10.1016/j.bmcl.2016.09.043

A high throughput phenotypic screening against Mycobacterium smegmatis led us to the discovery of a new class of bacteriostatic, highly hydrophobic antitubercular quinazolinones that potently inhibited the in vitro growth of either extracellular or intramacrophagic M. tuberculosis (Mtb), via modulation of an unidentified but yet novel target. Optimization of the initial hit compound culminated in the identification of potent but poorly soluble Mtb growth inhibitors, three of which were progressed to in vivo efficacy studies. Despite nanomolar in vitro potency and attractive PK properties, none of these compounds was convincingly potent in our in vivo mouse tuberculosis models. This lack of efficacy may be linked to the poor drug-likeness of the test molecules and/or to the properties of the target.[Figure presented]

Full text of DOI:10.1016/j.bmcl.2016.09.043

Accepted Manuscript
Identification and optimization of a new series of anti-tubercular quinazolinones
Cédric Couturier, Christine Lair, Alain Pellet, Anna Upton, Takushi Kaneko,
Corinne Perron, Eric Cogo, Jérome Menegotto, Armin Bauer, Bodo Scheiper,
Sophie Lagrange, Eric Bacqué
PII:
S0960-894X(16)30983-0
DOI:
http://dx.doi.org/10.1016/j.bmcl.2016.09.043
BMCL 24267
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
19 July 2016
14 September 2016
15 September 2016
Please cite this article as: Couturier, C., Lair, C., Pellet, A., Upton, A., Kaneko, T., Perron, C., Cogo, E., Menegotto,
J., Bauer, A., Scheiper, B., Lagrange, S., Bacqué, E., Identification and optimization of a new series of anti-tubercular
quinazolinones, Bioorganic & Medicinal Chemistry Letters (2016), doi: http://dx.doi.org/10.1016/j.bmcl.
2016.09.043
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.

Bioorganic & Medicinal Chemistry Letters
Identification and optimization of a new series of anti-tubercular
quinazolinones
Cédric Couturiera*, Christine Laira; Alain Pelleta, Anna Uptonb, Takushi Kanekob, Corinne Perronc, Eric
Cogoc, Jérome Menegottoc, Armin Bauerd, Bodo Scheiperd, Sophie Lagrangea and Eric Bacquéa
a Sanofi R&D, TSU Infectious Disease, 1541, Avenue Marcel Merieux, 69280 Marcy l’Etoile, France.
b Global Alliance for TB Drug Development, New York, New York 10005, United States
c Sanofi R&D, Lead Generation and Candidate Realization 195, route d’Espagne BP 13669 31036 Toulouse Cedex 1, France
d Sanofi Deutschland GmbH, Lead Generation and Candidate Realization, Industriepark Hoechst, 65926 Frankfurt am Main, Germany
ARTICLE INFO
ABSTRACT
A high throughput phenotypic screening against M. smegmatis led us to the discovery of a new class of bacteriostatic, highly
hydrophobic antitubercular quinazolinones that potently inhibited the in vitro growth of either extracellular or intramacrophagic M.
tuberculosis (Mtb), via modulation of an unidentified but yet novel target. Optimization of the initial hit compound culminated in
the identification of potent but poorly soluble Mtb growth inhibitors, three of which were progressed to in vivo efficacy studies.
Despite nanomolar in vitro potency and attractive PK properties, none of these compounds was convincingly potent in our in vivo
Article history:
Received
Revised
Accepted
Available online
mouse tuberculosis models. This lack of efficacy may be linked to the poor drug-likeness of the test molecules and/or to the
F
properties of the target.
O
O
O
N
N
Keywords:
Tuberculosis
Cl
N
N
R
N
N
H
H
High throughput screening
Antitubercular drugs
Lead opitmization
1
R= Cl or R = NH2 22 and 38
HIT COMPOUND
MIC Mtb = 38 nM
Issues = high metabolism, CYP3A4 inhibition
and poor aqueous solubility
OPTIMIZED COMPOUNDS
MIC Mtb 22/38 = 7/3 nM
No ADME issues but still poor solubility
2016 Elsevier Ltd. All rights reserved
Tuberculosis (TB) is an airborne infectious
disease caused by the bacterium Mycobacterium
tuberculosis (Mtb) and that mostly affects the lungs. It is a
complex disease caused by different bacterial populations
that are located within the host either intracellularly in
macrophages or extracellularly within cavities or hypoxic
lesions. TB represents one of the top public health
concerns worldwide. One-third of the world’s population is
infected with Mtb causing an estimated 9 million new
cases and 1.5 million deaths in 2013.1 The multidrug
treatment established in the 1970s and still recommended
today by the World Health Organization (WHO) has not
been sufficient to eliminate TB due to the advent of
HIV/AIDS, failure of treatment programs and emergence
of drug-resistant TB2. In addition, poor compliance and
deficient health care systems have often resulted in
treatment interruptions, thereby causing or exacerbating
the emergence and spread of drug resistance2. The
currently recommended combination therapy for drug-
sensitive TB is lengthy and complex: 2 months, associating
Rifampicin, Isoniazid, Pyrazinamide and Ethambutol,
followed by 4 months with Rifampicin and Isoniazid2. In
the case of multi- and extensive-drug resistant
(MDR/XDR) TB, treatments may even require daily
medication by multiple and unsafe drugs, for up to 24
months2.As a consequence, there is an urgent need for the
discovery and development of new antitubercular agents
that would act by new mechanisms of action and hence
would be able to treat both the drug-sensitive and the
resistant forms of the disease. With this objective, several
paths in the TB drug discovery field have been followed in
recent years such as preparing new analogues of TB
drugs3, revisiting the targets of known anti-TB
compounds4, repositioning non-TB drugs5, tapping into the
natural product pool6, harnessing target-based approaches7
and conducting traditional growth inhibition screenings8.
However, for various reasons, most of these
strategies have so far met with limited success, with the
notable exception of high-throughput (HTS) growth
inhibition screenings of small-molecule libraries against
Mtb (or more frequently against Mtb surrogates such as M.
smegmatis or M. bovis). Indeed, this latter approach has
delivered in recent years a variety of new anti-TB
scaffolds9. Along these lines, we conducted a few years
ago such a whole cell-based HTS, testing the Sanofi
chemical library (800K) against M. smegmatis.
*
Corresponding authors. Tel.: +33(0)4.37.37.07.35; email
cedric.couturier@sanofi.com (C.C.).

Following the screening step at 10µM, dose-
responses of compounds inhibiting more than 50% of the
bacterial growth were determined. Compounds displaying
MIC < 20 µM were then screened against M. tuberculosis
H37Rv. Confirmed actives were counter-screened vs the
human HepG2 cells to discard compounds with unspecific
antimycobacterial activities: only compounds with a
selectivity index (SI=TC50/MIC) > 10 were selected for
further analysis. Sorting and prioritization of the resulting
selective actives led to several series, of which
quinazolinone 1 emerged as a potential attractive starting
point due to its high potency (MIC = 38 nM against H37Rv
Mtb), an attractive SI (TC50 HepG2 = 13.4 µM), good
activity against the growth of Mtb in mouse macrophages
(MIC = 163 nM) and promising activity against drug-
resistant Mtb. Indeed, when tested on Mtb strains mono-
resistant to known TB drugs, no-cross-resistance was seen,
suggesting a new mechanism of action. Compound 1 was
also shown to exhibit bacteriostatic properties against Mtb
while antimycobacterial potency significantly decreased in
the presence of mouse, fetal calf or human sera (for
instance MIC shifted from 13 to 429 nM in the presence of
25% mouse serum) suggestive of high protein binding,
inhibition, poor solubility and strong serum shifts. Herein,
we report the results of this optimization program that led
us to advanced analogues that were tested in mouse
infection models. We will discuss the in vitro activities as
well as the ADMET and drug-like properties across the
series. We will also present the in vivo antitubercular
activities of the best analogues and comment on the
reasons for their lack of efficacy.
Synthesis of compounds 1-45 followed the routes
described in scheme 1. Three different routes (A, B and C)
were used in order to synthesize the targeted compounds.
The main difference between these different routes was the
nature of the leaving group (thiomethyl or chloro) in
intermediates 47 and 50 that was designed to allow the
installation of the amino side chain by aromatic
nucleophilic substitution. Starting from variously
substituted 2-aminobenzoic acids, the thioethers 47 were
obtained in two steps following cyclisation with
methylthioisocyanate and subsequent methylation of the
thiono sulfur atom of 46. Intermediates 46 were
alternatively transformed into compounds 50 upon reaction
with sulfuryl chloride. In the cases of substituents such as
methoxy or nitro on the aryl ring, compounds 49 were
obtained in a three-step sequence from 2-amino benzoic
acids: conversion into 2-amino N-methyl amino
benzamides 48, cyclisation into compounds 49 in the
presence of carbonyl di-imidazole and treatment with
phosphoryl chloride. The targeted compounds 1-45 were
finally obtained by condensation of compounds 47 or 50
with various primary amines 51. In pathway A,
intermediates 46 were heated at 160°C with di-isopropyl
ethyl amine as the base, in N-methyl pyrrolidinone, to
afford compounds 1-8 and 12-45. Of note, these conditions
sometimes proved too drastic, resulting in the degradation
of the reaction mixtures and in side products. In these
cases, pathways B and C were favored though the chloro
intermediates 50 turned out to be quite unstable.
Consequently, these intermediates were not stored but
rather used directly in the last condensation step. In
general, microwave heating of compounds 50 and 51, in
the presence of triethylamine in methanol, during 30
minutes at 130°C, delivered compounds 9-11 in
satisfactory yields (40-60 %).
We started our exploration program by investigating the
impact on the activity of the length of the ethyl linker
between the quinazolinone ring and the distal aromatic
ring. A methylene and a propyl chain resulted in a 3-log
reduction of activity (see compounds 2 and 3 in table 1),
suggesting that the ethyl chain was optimal for activity.
Keeping this ethyl chain constant, we then moved our
attention toward the N-alkylation pattern of compound 1.
We found that the N-alkylation of compound 1 was
optimal: demethylation of the quinazolinone ring nitrogen
and methylation of the side nitrogen led to inactive
compounds (see 4 and 5). Likewise, extending the ring N-
Me group into N-Et resulted in a very significant drop of
activity (see 6). Regarding the substituents on the
quinazolinone moiety (see compounds 7-15 in table 1), the
7-chloro (as in compound 1) and 7-methyl (see compound
9) substituents were found to be optimal.
consistent with high hydrophobicity (see below).
O
O
N
MIC Mtb = 38 nM
Cl
N
N
H
PSA = 54, exp-logP = 4.08; MW = 344
ADME issues = metabolism, CYP inhibition
Phys-chem issue = solubility
1
Figure 1: Hit Quinazolinone 1
There was no report (in the literature) of anti-Mtb
activity or anti-bacterial activity on the analogues of
compound 1 which exhibited a high experimental logP
(logP = 4.08), significantly different from the calculated
value provided by the ACD9 software (clogP = 3). In line
with this high logP value, equilibrium aqueous solubility
was found to be low at pH 7.4 or 1 (respectively <1 and 11
µg/ml). High hydrophobicity, low polarity (PSA = 54) and
the polyaromatic nature of compound 1 all pointed to poor
overall drug-like properties, somewhat compensated by a
low molecular weight (MW=344). ADMET profiling
showed that compound 1 had strong potential for oral
absorption as suggested by its high permeability though
Caco2 cells (data not shown), moderately inhibited the
hERG channel (35% inhibition at 10 µM in an automatic
patch clamp assay, at 20°C) and was Ames-negative.
Furthermore, while being a weak inductor of CYP3A4, 1
potently inhibited cytochrome CYP3A4 (IC50 < 1 M for
the testosterone probe) and also displayed high metabolism
in the presence of mouse microsomes, incompatible with in
vivo efficacy in mouse models of TB. Metabolism was also
found high in the presence of human microsomes (see table
4) or human hepatocytes (data not shown). Demethylation
of the methoxy group and hydroxylations of the distal aryl
ring were identified as the main metabolism pathways.
At this point, we decided to embark on a program
aimed at exploring structure-activity relationship around
compound 1 while striving to solve the issues identified
during the profiling phase, namely high metabolism, CYP

O
O
N
O
Pathway A
R2
R1
R2
R1
R2
R1
b
a
N
N
OH
X
R3
N
H
S
S
NH2
Y
47
H2N
O
N
46
c
Pathway C
51
R2
R1
X
R3
N
g
d
Y
h
N
H
O
O
O
R2
R1
Pathway B
N
1-45
R2
R1
R2
R1
f
e
N
N
H
N
Cl
NH2
N
H
O
49
50
48
Scheme 1: (a) MeNCS, AcOH, 150°C, 24h (b) Me2SO4, NaOH, MeOH, rt, 5h (c) DIEA, NMP, 160°C, 24h (d) MeNH2.HCl, HOBt, DCI, DIEA,
DMF, rt, 16h (e) Im2CO, DBU, THF, rt, 20h (f) POCl3, DIEA, 90°C; 20h (g) SO2Cl2, CHCl3, 60°C, 6h (h) Et3N, EtOH µW, 130°C, 30 min
The other 7-substituted analogues were less active
than 1 or even inactive (see the hydrogeno, fluoro,
methoxy, cyano, amino and dimethylamino analogues 7-
13). Shifting the chloro atom of compound 1 from position
7 to position 6 yielded compound 14 that was much less
active than 1. Introduction of a nitrogen atom at various
positions of the benzo ring of compound 4 proved to be
deleterious to potency as well (data not shown).
Fluorination of position 6 of the quinazolinone ring of
compound 1 was well tolerated and delivered a compound
of similar activity as 1 (see compound 15).
Across this first series of compounds, the serum shifts were
generally above ten (data not shown) with the exception of
12 for which the shift was around 5 (MIC : 198 vs 43 nM).
These results were in line with the logPs: above 2 or even 3
for all analogues except for 12. As for 1, cytotoxicity on
HepG2 cells was generally moderate, typically between 10
-> 30 µM: this was not considered a problem for this series
considering the high potencies vs Mtb. At this point, we
focused on the distal aromatic ring. We rapidly discovered
that there was little room here to reduce the hydrophobicity
of 1. Substituting the p-methoxy group by a hydrogeno
atom (see compound 16 in table 2) or by a more polar
moiety (compounds 17 and 18) resulted in a substantial
loss of activity and no gain in terms of serum shifts.
Substituting the distal phenyl ring of 1 for a 2-pyridyl ring
(compound 19) yielded a similar result. On the other hand,
whereas a p-OCF3 group resulted in a less potent
compound (20), increasing further hydrophobicity by
exchanging the methoxy group for an aromatic ring
generally resulted in compounds (see compounds 21-23)
with similar or even better potency vs Mtb than 1, in a way
that was reminiscent of the optimization of PA-82410.
Since these biaryl derivatives turned out to possess
improved ADMET profiles (see below), we tried to
improve the drug-like properties of these derivatives.
Introduction of nitrogen(s) on the new terminal aromatic
ring was detrimental to activity ring activity (compare 23
with 27 and 28 or 22 with 25 and 30), unless substituted by
a hydrophobic group (compare respectively 24 with 25-28,
29 with 30-31 and 33 with 34). We also attempted to
disrupt the biaryl moiety by introducing a hydrophobic 4-
chloro phenoxy moiety instead of the methoxy group of
compound 1. The corresponding compound 35 proved to
retain similar activity compared to 1. Adding a nitrogen on
the aryloxy moiety of compound 34 was tolerated
(compound 36) but removing the chloro atom led clearly,
as expected on previous observations, to a less potent
compound (see 37).
O
OMe
R2
R1
R3
N
experimental (e) or calculated
(c) LogP
MIC* Mtb
IC50 HepG2
(µM)
Cpd #
( )n
N
N
(µM)
R4
R1
Cl
Cl
Cl
Cl
Cl
Cl
H
F
Me
OMe
CN
NH2
NMe2
H
R2
H
H
H
H
H
H
H
H
H
H
H
H
H
Cl
F
R3
R4
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
S
1
0
2
1
1
1
1
1
1
1
1
1
1
1
1
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
Me
Me
Me
H
Me
Et
Me
Me
Me
Me
Me
Me
Me
Me
Me
4.08 (e)/3 (c)
2.6 (c)
0.04
19.90
5.80
6.30
11.20
1.10
0.24
0.20
0.01
0.03
0.48
0.04
0.11
0.90
0.03
13.4
15.7
16.8
ND
>30
21.0
>30
>30
>30
>30
12.7
ND
3.4 (c)
3.2 (c)
3 (c)
3.5 (c)
2.4 (c)
2.5 (c)
2.9 (c)
3.1 (e)/2.3 (c)
1.8 (c)
1.94 (e)/ 1.1 (c)
1.94 (e)/2.5 (c)
3.2 (c)
ND
26.1
>30
Cl
3.4 (c)
Table 1: Structure Activity Relationships on the quinazolinone ring and the linker moiety
* MIC determined either by the BactTiter glow method (measure of intracellular ATP by luminescence)

O
N
R2
X2
X3
MIC*
Mtb + 25%
human-serum
(µM)
N
X1
experimental (e) or
calculated (c)
LogP
IC50
HepG2
(µM)
MIC* Mtb
X4
Cpd #
R1
N
H
(µM)
R1
Cl
Cl
Cl
Cl
R2
OMe
H
X1
CH
CH
X2
1
CH
CH
CH
CH
4.08 (e)/3(c)
4.24 (e)/3.1 (c)
3.86 (e)/2.7 (c)
2.2 (c)
0.04
0.06
0.08
0.33
0.43
2.79
>10
>10
13.4
15.2
>30
>30
16
17
18
OCH2CH2OMe CH
O
CH
N
19
20
21
Cl
Cl
Cl
OMe
OCF3
N
CH
CH
CH
CH
CH
3.32 (e)/2.2 (c)
4.95 (e)/4 (c)
5.27 (e)/4.8 (c)
0.53
0.07
0.01
ND
2.33
0.87
27.0
9.9
6.5
F
22
23
24
25
26
Cl
Cl
Cl
Cl
Cl
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
5.28 (e)/4.8 (c)
5.13 (e)/3.3 (c)
4.06 (e)/3.5 (c)
4.5 (e)/3.4 (c)
5.04 (e)/4.3 (c)
0.01
0.01
0.03
0.02
0.01
0.29
1.47
25.16
3.14
0.25
22.0
>30
16.5
19.0
>30
OMe
N
F
N
Cl
N
OMe
OMe
N
27
28
29
Cl
Cl
Cl
CH
CH
CH
CH
CH
CH
4.67 (e)/3.7 (c)
4.68 (e)/4 (c)
3.83 (e)/3.2 (c)
0.06
0.09
0.08
ND
ND
11.9
11.3
25.4
N
N
4.06
N
F
Cl
N
N
30
31
32
33
Cl
Cl
Cl
Cl
CH
CH
CH
CH
CH
CH
CH
CH
4.46 (e)/3.2 (c)
5.02 (e)/3.8 (c)
4.18 (e)/3.1 (c)
3.85 (e)/3 (c)
0.01
0.02
0.02
0.02
0.39
0.11
3.52
8.56
>30
>30
>30
17.3
N
N
OMe
N
N
N
N
N
Cl
34
Cl
CH
CH
4.67 (e)/3.9 (c)
0.01
0.86
21.6
N
Cl
Cl
35
36
37
Cl
Cl
Cl
CH
CH
CH
CH
CH
CH
5.66 (e)/5.7 (c)
4.94 (e)/4.1 (c)
3.3 (c)
0.04
0.02
0.06
1.90
1.76
7.37
10.9
13.0
12.1
O
O
N
N
O
F
38
39
40
41
42
43
44
NH2
Cl
CH
N
CH
CH
CH
CH
CH
N
3.3 (e)/2.9 (c)
3.5 (c)
0.01
0.68
0.06
0.32
0.31
1.12
0.03
0.01
ND
>30
18.3
13.9
14.7
24.1
15.1
>30
F
Cl
N
4.23 (e)/3.4 (c)
4.03 (e)/3.3 (c)
2.7 (c)
4.02
ND
OMe
OMe
Cl
N
N
Cl
N
>15
ND
Cl
N
3.2 (c)
Cl
Cl
N
N
4.67 (e)/3.9 (c)
3.32
OMe
45
Cl
N
N
3.3 (c)
0.74
ND
ND
Table 2: Structure Activity Relationships on the distal aromatic ring
* MIC determined either by the BactTiter glow method (measure by luminescence of intracellular ATP).
In a new effort to conciliate high potency and lower logP
in the bi-aryl sub-series, we combined the anilino moiety
of compound 12 (the compound from the first series with
the best compromise potency/logP) with the p-
fluorophenyl of compound 22 (the most potent compound
of the series). We were delighted to find that the resulting
compound 38 retained nanomolar potency. Finally, in a last
effort to decrease the logP of the biaryl analogues, we tried
to introduce nitrogen atoms on the first phenyl moiety. As
observed for compound 19, this extra polarity
systematically resulted in less potent compounds (39 to 45)
as well as no gain in serum shift (when measured).

For compounds 16 to 45, serum shifts ranged from
8 (for compound 19) to ~200 for compound 23. The
highest shifts were generally associated to the most potent
biaryl derivatives, thereby totally obliterating the benefit of
the high potencies observed without serum (see for
example, compounds 22 and 23). The few exceptions to
this rule were for the more polar, less hydrophobic,
compounds such as 26, 30, 31 and 38. In the latter case, the
serum effect was still important (~30; comparable to 1) but
deemed as acceptable considering the level of remaining
potency (MIC with serum = 84 nM). As previously noted,
cytotoxicity remained moderate to weak across this new set
of compounds (typically in the 10 to > 30µM range).
Cpd
#
Structure
MIC (nM) MIC (nM)
without
serum
38
mouse
macrophage
163
O
N
1
O
N
Cl
N
H
O
21
8
3
164
60
N
Cl
N
N
H
F
O
22
N
Cl
Cl
N
N
H
F
O
N
25
26
30
31
32
15
2
51
63
N
N
N
H
Cl
We also looked at the capacity of our best
O
N
N
compounds to inhibit growth inhibition of Mtb in
macrophages, a known compartment of Mtb in humans11.
Against mouse macrophages, activities were consistently
good (generally below the 200 nM bar) across the series
(see table 3) and generally well correlated with MICs vs
extracellular Mtb, with a few exceptions. The ratios
between the two types of MICs were typically between 1
and 10 and ranged from 0.6 (for 30) to almost 160 (for 31),
without clear structural rationale (compare 25, 30, 32 and
40 with 22 and 31). Hydrophobicity didn’t seem to play a
role here as demonstrated by the comparisons between the
less hydrophobic 1 and 38 with the highly hydrophobic 22
and 35 that all yielded similar good intramacrophagic
activities (MICs = from 60 to 163 nM).
Cl
Cl
Cl
N
N
H
F
O
N
3
13
N
N
N
N
H
Cl
O
N
6
947
81
N
N
N
N
H
O
O
N
16
N
N
Cl
Cl
N
O
N
H
35
38
O
36
4
133
60
N
N
N
H
Cl
F
O
N
H
N
H
N
N
H
In parallel to this SAR exploration exercise, we profiled
our best quinazolinones in a series of ADMET and
physico-chemical assays. We first checked that, across the
series, Caco2 permeability and hERG inhibition were
generally acceptable (data not shown). Then, we
investigated the impact of structural variations on the main
issues of the hit compound, namely aqueous solubility,
metabolism and CYP3A4 inhibition (see table 4).
Replacement of the chloro substituent on the quinazolinone
ring of 1 generally resulted in increased metabolism in
microsomes and only slight improvements in solubility at
pH1 (see compounds 9 and 10 in table 3).Simple variations
on the distal aromatic ring didn’t improve metabolic
stability but occasionally improved solubility when
polarity was introduced (see 16, 18 and 19).
F
O
N
40
57
201
N
N
Cl
N
H
Table 3: Best compounds in mouse macrophages infected by Mtb
Regarding CYP3A4 inhibition, the 2-mono-aryl ethyl
amino derivatives (compounds 2 to 20) generally retained
potent inhibition (for example, see in table 4, compounds
9, 10 and 19) with a few exceptions such as 16 and a few
para-substituted aryl derivatives (see for representative
examples, 18 and 20) where inhibition was more moderate.
Along similar lines, the biaryl derivatives turned out to be
weak to moderate inhibitors of CYP3A4 (eg 21, 22, 25, 26,
30, 31 and 38) though the improvement was less marked
for the least hydrophobic analogues (see for instance 23,
24, 29 and 32). Biaryloxy derivatives behaved similarly
(see 35 and 36).
In contrast, substituting the para-MeO group on
the distal aromatic ring of 1 by a phenyl or a
trifluoromethoxy moiety clearly dampened metabolism
though at the expense of solubility (see 20, 21 and 22).
Addition of nuclear nitrogens on either ring of the biaryl
moiety did significantly improve solubility while retaining
good metabolic stability (see 24, 29 and 33) but these gains
in solubility were rapidly lost upon attachment of the kind
At the end of this exploration/optimization phase, it was
clear that we had been unable to find an acceptable
compromise
between
high
potency,
moderate
hydrophobicity and good ADMET profile. Compound 38
turned out to be the best balance we could reach: it
displayed potent antimycobacterial activities (with and
without serum), with high caco2 permeability, limited
metabolism (only 33% of which was mediated by
CYP3A4) and weak inhibition of CYP3A4 as well as of
the hERG channel (22% inhibition at 10µM), even if its
aqueous solubility was quite low.
of
hydrophobic
groups
that
provided
high
antimycobacterial potency (see 23, 25, 26, 30, 31, 32 and
34). Likewise, attachment at the para position of the distal
aromatic of a chloro-substituted (hetero)aryloxy moiety
resulted in insoluble compounds with high metabolic
stability (see 35 and 36). Compound 38 that combined a p-
fluoro biaryl moiety with a polar 7-amino quinazoline
moiety displayed slightly better solubility compared to 22
and moderate metabolism.
Overall, we concluded that parallel optimization of
the ADMET properties and of the potency against Mtb had
led us into an unfavorable corner of the physico-chemical
space, an analysis already formulated by others8b,8c
.

Cpd #
1
Structure
MIC (nM)
without serum
38
Solubilitya
pH1/7.4 (µg/ml)
157/1
% of metabolic
lability (h/m/r)b
51/68/18.5
CYP3A4 IC50
(T/M)c (µM)
4.3 / <1
O
N
O
N
Cl
N
H
O
9
11
26
221/<1
734/4
89/99/97
90/86/52
65/66/27
58/12/40
4.8 / <1
10 / <1
18 / 11
11 / 11
O
N
N
N
H
O
10
16
18
O
N
O
N
N
H
O
62
180/<1
1091/8
N
Cl
N
N
H
O
O
332
N
N
Cl
Cl
N
N
H
O
19
20
532
70
1138/9
43/<1
41/47/17
9/32/0
7.4 / <1
>30 / 4
N
O
N
N
N
H
O
OCF3
N
Cl
Cl
N
N
H
O
21
22
8
3
5/<1
18/21/26
0/7/11
>30 / >30
18 / >30
N
N
O
N
H
F
<1/<1
N
N
N
N
Cl
N
N
H
O
O
23
24
25
26
7
27
15
2
26/<1
1107/<1
44/<1
28/0/23
17/10/26
0/16/15
3/9/20
7.9 / 8
5.8 / 15
9.3 / 41
8.4 / >30
N
Cl
Cl
Cl
N
O
N
H
N
N
O
N
H
F
N
N
N
H
Cl
O
N
21/<1
N
Cl
N
O
N
H
N
29
30
31
32
79
3
246/<1
11/<1
10/<1
37/<1
3/6/28
0/0/1
5.4 / 13
9.9 / >30
16 / >30
8.7 / 11
N
N
N
N
Cl
Cl
Cl
N
O
N
H
F
N
N
N
O
N
H
Cl
N
6
7/12/18
14/0/15
N
N
O
N
H
O
N
16
N
N
Cl
Cl
N
N
H
O
33
34
35
36
38
61
12
36
23
4
138/<1
<1/<1
<1/<1
<1/<1
13/<1
19/27/35
0/1/5
8.3 / 27
- / 16
N
N
N
N
N
H
O
N
Cl
N
N
Cl
Cl
Cl
N
O
N
H
O
17/3/9
29 / 14
15 / 9.8
30 / 23
N
N
N
O
N
Cl
Cl
H
O
N
10/6/3
N
N
H
F
O
N
10/27/19
N
H
N
H
N
H
Table 4: Structure Property Relationship of selected quinazolinones
a: Equilibrium solubility at 20°C; b: % of metabolized compound (initial concentration: 5µM) after 20’ incubation, in the presence of 1 mg/mL of
human/mouse or rat of microsomal proteins and of 1 mM of NADPH; c: testosterone (T) or midazolam (M) probes

At this point, we were ready to assess the in vivo
efficacy of our best derivatives in mouse model of highly
acute tuberculosis infection. We started with compound 35,
a typical representative of the series with high in vitro
potency (MIC = 36 nM), a serum effect was noted (MIC =
420/2100 nM in the presence of 25% of mouse/human
serum), in line with high protein binding in humans and
mouse (99.9%), good intramacrophagic activity (MIC =
133 nM) and poor aqueous solubility (< 1 µg/ml at pH 7).
Mouse PK of 35 indicated good oral bioavailability (90%,
iv 3 mg/kg) as a solution in NMP / Solutol HS15 / PBS
pH7.4 (5%/5%/90%) formulation. Similar exposures were
obtained using a Vitamin E TPGS /Water (50%/50%) or a
Lipidic vehicle (Cremophor RH40/Capryol 90/Miglyol
812N (10%/20%/70%) half diluted in water (Pre-
emulsified formulation). The VitE formulation was first
used for in vivo activity evaluation. A lack of potency at
the higher tested dose was observed and in order to
improve exposure, we used the Pre-emulsified
(Lipidic/Water) formulation for the 600 mg/kg dose. In this
formulation, plasma AUCs were dose-proportional
between 10 and 100 mg/kg (21 to 290 µg.h/ml) and at 100
mg/kg, plasma and lung concentrations remained above the
MIC value in the presence of mouse serum for more than
96 hours. Despite these encouraging PK properties, when
tested in our TB infection model, compound 35 showed
only weak efficacy: the Minimum Effective Dose (MED)
was only approached at 600 mg/kg (see Figure 2).
both plasma and lungs. Of note, higher doses than 100 or
200 mg/kg failed to significantly improve exposures due to
solubility limitations. When tested in the highly acute
model at doses ranging from 12.5 to 200 mg/kg, compound
22 showed a very disappointing flat in vivo response (see
figure 3) in line also with high protein binding in humans
and mouse (99.9%).
Figure 3: In vivo mouse model of acute TB with compound 22: mice were
infected by H37Rv strain day 1 (D1).A daily oral gavage started D2 for 2
weeks (4 days by week). Compound is formulated in VitE TPGS/PEG
200 (10%/90%) and compared to Ethambutol (ETH)). Controls received
vehicle alone. Efficacy was assessed at D12 by the decline in lung CFU
counts.
As poor solubility was hypothesized to be a
limiting factor in the GI tract due to potential risk of re-
precipitation anticipated with co-solvent/surfactant
formulation, we decided to switch to an alternative
formulation strategy identifying a sulfate salt of 22 that
could be formulated in a 5% Soluplus/Acetate buffer
formulation between 10 to 300 mg/kg. In this formulation,
plasma concentrations at 300 mg/kg were above the MIC
with serum for more than 24 hours (AUC: 140 µg.h/ml at
300 mg/kg po) and lung exposure was clearly improved
(AUC at 300mg/kg: 980 µg.h/ml versus 55 in
VitE/PEG200).
In the highly acute model, each 4 days dosing are
separated by a 2 days washout. Taking into account the
bacteriostatic effect of the series, we decided to modify the
model by treating mouse daily for 8 continuous days.
Following 8 consecutive daily doses, compound 22
demonstrated MED at 300 mg/kg (see figure 4). However,
efficacy was not confirmed when 2*4 day dosing periods
were separated by 2 days without treatment (0.78 log
reduction versus 1.5 log without dosing break). Due to the
large discrepancy between in vitro activity and the poor in
vivo activity obtained in our efficacy studies, despite good
in vivo exposures, we decided to stop with this compound.
Finally, we selected compound 38, the compound
that had emerged from the optimization phase as
displaying the best balance between potency (MIC= 4nM;
MIC with human/mice serum= 84/7 nM; intramacrophagic
MIC = 60 nM) and hydrophobicity (solubility at pH 1 = 13
µg/ml; high protein binding in humans and mouse
(99.8%).). Though PK properties at 100 mg/kg (mice;
formulation VitE/PEG200/HPC (19%/79%/2%); po)
looked in first analysis less attractive than those of 22 and
35 (AUC = 29 µg.h/ml; t1/2≈ 6h, lung/plasma ratio = 4.5),
we nevertheless decided to assess the in vivo efficacy of 38
Figure 2: In vivo mouse model of acute TB with compound 35:
mice were infected by H37Rv strain day 1 (D1).A daily oral gavage
started D2 for 2 weeks (4 days by week). Compound is formulated in
VitE/water or in Lipidic/water (for 600 mg/kg) and controls received the
vehicle alone. Efficacy was assessed at D12 by the decline in lung CFU
counts.
Looking for a compound that would retain more
potency than 35 in the presence of serum, we then turned
our attention to compound 22. Since its MIC was 3 nM on
replicating Mtb following 5 days incubation and 34/250nM
with 25% of mouse/human serum. It was also highly active
intra-macrophages (MIC = 60 nM). Aqueous solubility of
this compound was low (< 1 µg/ml at pH 7); a new
PEG200/Vitamin E TPGS (90%/10%) formulation allowed
reasonable solubility (20 mg/ml) and exposures in mice. In
this formulation, oral bioavailability in mice was 73% and
plasma exposure reached 76 µg.h/ml at 100 mg/kg po.
Lung tropism was good (lung / plasma ratio= 14) and half-
life in plasma was long (8 h). Furthermore, time over the
MIC in the presence of mice serum was above 24 hours in

hoping that improved in vitro potency in the presence of
serum would compensate for diminished exposures.
plates. In the mouse model of highly acute TB, 1.106
bacteria are inoculated to each mouse giving a lung burden
from 1.106 to 1.108 CFU at the end of the experiment.
Taking into account the MIC shift, the limited free fraction
available and the incapacity to increase doses due to poor
solubility, a too low compound concentration available in
lungs could explain the poor in vivo activity observed.
Another key point for in vivo activity is the compound
distribution in pulmonary lesions11 which is unknown for
this series. Failure could also be target-related due to their
bacteriostatic effect8b, 8c, to a lack of in vivo vulnerability12
or linked to the already reported inherent difficulty to
correlate PK properties and in vivo efficacy for anti-TB
compounds13.
As frequently observed for anti-tubercular series
Figure 4: In vivo mouse model of acute TB with compound 22-sulfate. Mice were
infected by H37Rv strain day 1 (D1). A daily oral gavage started D2 for 2 weeks
(10 treatment in Soluplus/ acetate buffer formulation) and compared to Rifampycin
(RIF) at 10 mg/kg. Efficacy was assessed at D12 by the decline in lung CFU counts.
stemming from phenotypic screenings, we had to face the
combined difficulties of not knowing the target (and hence
if it is a validated target in vivo) and of selecting a non-
lead-like hit compound that upon subsequent optimization
led us, because of the greasy membrane of Mtb, to potent
but highly hydrophobic compounds whose properties are
unfavorable to allow demonstration of in vivo efficacy at
reasonable doses8b, 8c. Based on these disappointing results
and in the absence of evidence that we would be able to
improve the drug-like properties of the series upon further
optimization, the series was discontinued and we didn’t
attempt to identify the target.
This expectation turned out not to be met: against
highly acute murine, a flat lung CFU reduction response
was observed over the dose range of 10-100 mg/kg (see
Figure 5). Furthermore, toxicity after one treatment at 300
mg/kg prevented any increase of doses.
The quinazolinone series has been identified as a
potential new class of anti-tubercular drugs following a
growth inhibition HTS against M. smegmatis used as a
surrogate of Mtb. The attractive microbiological profile of
the hit compound, including its likely novel (and
unidentified) mechanism of action, prompted us to initiate
an optimization program aimed at delivering prototype
compounds able to rapidly achieve in vivo proof of concept
in the mouse model of highly acute Mtb infection and
hence demonstrate the potential of the series to afford a
development candidate. Our effort led us to compounds
with single-digit nanomolar in vitro potency against Mtb,
moderate serum effect, no CYP liabilities and dampened
metabolism. However, the best compounds still displayed
high hydrophobicity, low aqueous solubility, non-
proportional oral mice PK at high doses and overall poor
drug-likeness. Furthermore, we were unable to demonstrate
significant in vivo efficacy for three analogues, in the
highly acute mice tuberculosis infection model. Our work
illustrates the difficulties to establish in vivo proof of
concept for poorly drug-like anti-tubercular hit compounds
that usually stem from Mtb growth inhibition screening.
On a more positive concluding note, unveiling the
Figure 5: In vivo mouse model of acute TB with compound 38 mice were
infected by H37Rv strain day 1 (D1).A daily oral gavage started D2 for 2
weeks (4 days by week). Compound is formulated in VitE/PEG200/HPC
(19%/79%/2%) and compared to Rifampycin (RIF) at 10 mg/kg.
Controls received vehicle alone. Efficacy was assessed at D12 by the
decline in lung CFU counts.
Overall, despite good in vitro potencies even in the
presence of serum (at least for compounds 22 and 38),
potent intra macrophage activities and good PK properties
(eg time above the MICs in the presence of serum and
good exposures in lung), we have failed to demonstrate
convincing in vivo efficacy for any member of the
quinazolinone series. In the best cases, we only came close
to the MED at high doses (300 to 600 mg/kg with
optimized formulation).
unknown target of the quinazolinone series should deliver
a novel mycobacterial target, which in case of validation of
its in vivo interest, may help find, following a new target-
based screening, alternative and hopefully more attractive
chemical matter than the compounds of this work.
Reasons for this lack of efficacy were not
elucidated. We could only speculate that part of the failure
was compound-related. Assays trying to generate
compound-resistant mutants with this bacteriostatic series,
revealed an unexpected profile: MICs were dependent of
inoculum concentration, not allowing us to obtain mutants
for target identification. MIC on agar medium for
compound 22 shifts from 0.016µg/ml to 2µg/ml for 1000
and 100 000 bacteria inoculated respectively on agar
Literature
1. WHO: World Health Organization: Global tuberculosis
report 2015. In.; 2015.

2. (a) WHO: World Health Organization: Global MDR-TB
situation 2014. In.; 2014 (b) Dheda, K.; Migliori, G. B.
Lancet 2012, 379, 773.
Thammasuvimol, E.; White, E. L.; Zhang, W.; Secrist, J.
A. III Tuberculosis 2012, 92, 72 (c) Kang, S.; Kim, R. Y.;
Seo, M. J.; Lee, S.; Kim, Y. M.; Seo, M.; Seo, J. J.; Ko, Y.;
Choi, I.; Jang, J.; Nam, J.; Park, S.; Kang, H.; Kim, H. J.;
Kim, J.; Ahn, S.; Pethe, K.; Nam, K.; No, Z.; Kim, J. J.
Med. Chem. 2014, 57, 5293 (d) Andries, K.; Verhasselt, P.;
Guillemont, J.; Gohlmann, H. W.; Neefs, J. M.; Winkler,
H.; Van Gestel, J.; Timmerman, P.; Zhu, M.; Lee, E.;
Williams, P.; de Chaffoy, D.; Huitric, E.; Hoffner, S.;
Cambau, E.; Truffot-Pernot, C.; Lounis, N.; Jarlier, V.
Science 2005, 307, 223 (e) Stover, C. K.; Warrener, P.;
VanDevanter, D. R.; Sherman, D. R.; Arain, T. M.;
Langhorne, M. H.; Anderson, S. W.; Towell, J. A.; Yuan,
Y.; McMurray, D. N.; Kreiswirth, B. N.; Barry, C. E.;
Baker, W. R. Nature 2000, 405, 962 (f) Makarov, V.;
Manina, G.; Mikusova, K.; Möllmann, U.; Ryabova, O.;
Saint-Joanis, B.; Dhar, N.; Pasca, M. R.; Buroni, S.;
Lucarelli, A. P.; Milano, A.; De Rossi, E.; Belanova, M.;
Bobovska, A.; Dianiskova, P.; Kordulakova, J.; Sala, C.;
Fullam, E.; Schneider, P.; McKinney, J. D.; Brodin, P.;
Christophe, T.; Waddell, S.; Butcher, P.; Albrethsen, J.;
Rosenkrands, I.; Brosch, R.; Nandi, V.; Bharath, S.;
Gaonkar, S.; Shandil, R. K.; Balasubramanian, V.;
Balganesh, T.; Tyagi, S.; Grosset, J.; Riccardi, G.; Cole, S.
T. Science 2009, 324, 801 (g) Brown, J. R.; North, E. J.;
Hurdle, J. G.; Morisseau, C.; Scarborough, J. S.; Sun, D.;
Kordulakova, J.; Scherman, M. S.; Jones, V.;
Grzegorzewicz, A. Bioorg. Med. Chem. 2011, 19, 5585.
10. (a) Sutherland, H. S.; Blaser, B., Kmentova, I.;
Franzblau, S. G.; Wan, B.; Wang, Y.; Ma, Z; Palmer, B.
D.; Denny, W. A.; Thompson, A. M. J. Med. Chem. 2010,
53, 855 (b) Kmentova, I.; Sutherland, H. S.; Palmer, B. D.;
Blaser, A.; Franzblau, S. G.; Wan, B.; Wang, Y.; Ma, Z.;
Denny, W. A.; Thompson, A. M. J. Med. Chem. 2010, 53,
8421 (c) Blaser, A.; Palmer, B. D.; Sutherland, H. S.;
Kmentova, I.; Franzblau, S. G.; Wan, B.; Wang, Y.; Ma,
Z.; Thompson, A. M.; Denny, W. A. J. Med. Chem., 2012,
55, 312 (d) Palmer, B. D.; Sutherland, H. S.; Blaser, A.;
Kmentova, I.; Franzblau, S. G.; Wan, B.; Wang, Y.; Ma,
Z.; Denny, A. M.; Thompson, A. M. J. Med. Chem., 2015,
58, 3036.
3. (a) Sacksteder, K. A.; Protopopova, M.; Barry C. E. 3rd;
Andries, K; Nacy, C. A. Future Medicine 2012, 823 (b)
Zhang, D.; Lu, T.; Liu, K.; Liu, B.; Wang, J.; Zhang, G.;
Zhang, H.; Liu, Y; Wang, B.; Zheng, M.; Fu, L.; Hou, Y.;
Gong, N.; Lv, Y.; Li, C.; Cooper, C. B.; Upton, A. M.;
Yin, D.; Ma, Z.; Huang, H. J. Med. Chem. 2012, 55, 8409
(c) Briguglioa, I.; Pirasa, S.; Coronaa, P.; Pirisia, M. A.;
Jabesb, D.; Cartaa, A. Anti-Infective Agents 2013, 75, 4.
(a) Hameed, S. P.; Manjrekar, P.; Raichurkar, A.; Shinde,
V.; Puttur, J.; Shanbhag, G.; Chinnapattu, M.; Patil, V.;
Rudrapatana, S.; Sharma, S.; Kumar N. C. N.,
Nandishaiah, R.; Madhavapeddi, P.; Sriram, D.; Solapure,
S.; Sambandamurthy, V. K. ACS Med. Chem. Lett. 2015, 6,
741 (b) Christopher P. Locher, C. P.; Jones, S. V.;
Hanzelka, B. L.; Perola, E.; Shoen, C. M.; Cynamon, M.
H.; Ngwane, A. H.; Wiid, I. J.; van Helden, P. D.;
Betoudji, F.; Nuermberger, E. L.; Thomson, J. A.
Antimicrob. Agents Chemother. 2015, 59, 1455 (c)
Hartkoorn, R. C.; Sala, C.; Neres, J.; Pojer, J.; Magnet, S.;
Mukherjee, R.; Uplekar, S.; Boy-Röttger, S.; Altmann, K.;
Cole, S. T. EMBO Mol Med 2012, 4, 1032.
5. (a) Palomino, J. C.; Martin, A.; J. Antimicrob
Chemother 2012, 1 (b) Lougheed K. E. A.; Taylor, D. L.;
Osborne, S. A.; Bryans, J. S.; Buxton, R. S. 2009, 89, 364
(c) Ashtekar, D.R.; Costa-Periera, R.: Shrinivasan, T.
Diagn Microbiol. Infect. Dis. 1991, 14, 465 (d) Jean-
Emmanuel Hugonnet, J.; Tremblay, L. W.; Boshoff, H. I.;
Barry, C. E. 3rd; Blanchard, J. S. Science 2009, 1215.
6. (a) Salomon, C. E.; Schmidt, L. E. Current Topics Med.
Chem. 2012, 12(7), 735 (b) García, A.; Bocanegra-García,
V.; Palma-Nicolás, J. P.; Rivera, G. Eur. J. Med. Chem.
2012, 49, 1 (c) Lee, R. E.; Hurdle, J. G.; Liu, J.; Bruhn, D.
F.; Matt, T.; Scherman, M. S.;Vaddady, P. K.; Zheng, Z.;
Qi, J.; Akbergenov, R.; Das, S.; Madhura, D. B.; Rathi, C.;
Trivedi, A.; Villellas, C.; Lee, R. B.; Kesh, R.;
Waidyarachchi, S. L.; Sun, D.; McNeil, M. R.; Ainsa, J. A;
Boshoff, H. I.; Gonzalez-Juarrero, M.; Meibohm, B.;
Böttger, E. C.; Lenaerts, A. J. Nature Medecine, 2014, 152
(d) Nikonenko, B. V.; Reddy, V. M.; Protopopova, M.;
Bogatcheva, E.; Einck, E.; Nacy, C. A. Antimicrob Agents
Chemother 2009, 53(7), 3138.
7. (a) Kana, B. D.; Karakousis, P. C.; Parish, T.; Dick, T.
Tuberculosis 2014, 94, 551 (b) Miller, C. H.; O’Toole R.
F. Expert Opin. Drug Discov. 2011, 6(8), 839 (c) Xiong,
X.; Xu, Z.; Yang, Z.; Liu, Y.; Wang, D.; Dong, M; Parker,
E. J.; Zhu, W. Current Drug Targets 2013, 14, 676.
8. (a) Cooper, C. B. J. Med. Chem. 2013, 56, 7755 (b)
Manjunatha U. H.; Smith, P. W. Bioorganic Med. Chem.
2015, 23, 5087 (c) Jiricek, J. Future Med. Chem 2014, 6
(12), 1377.
11. Dartois, V.; Barry, C. E. 3rd Bioorg. Med. Chem Lett.
2013, 23, 4741.
12. Wei, J. R.; Krishnamoorthy, V.; Murphy, K.; Kim, J.
H.; Schnappinger, D.; Alber, T.: Sassetti, C. M.; Rhee, K.
Y.; Rubin, E. J. Proc. Natl. Acad. Sci. USA 2011,
108:4176e81.
13. Lakshminarayana, S. B.; Huat, T. B.; Ho, P. C.;
Manjunatha, U. H.; Dartois, V.; Dick, T.; Srinivasa P. S.
Rao, S. P.S. J. Antimicrob. Chemother. 2015, 70, 857.
Supplementary data associated with this article can be
found in the online version.
9. (a) Ananthan, S., Faaleolea, E. R. , Goldman, R. C.;
Hobrath, J. C.; Kwong, C. D.; Laughon, B. E.; Maddry, J.
A.; Mehta, A.; Rasmussen, L.; Reynolds, R. C.; Secrist III,
J. A.; Shindo, N.; Showe, D. N.; Sosa, M. I.; Suling, I. J.;
White E. L. Tuberculosis, 2009, 89, 334 (b) Robert C.
Reynolds, R. C.; Ananthan, S.; Faaleolea, E.; Hobrath, J.
V.; Kwong, C. D.; Maddox, C.; Rasmussen, L.; Sosa, M.;

Graphical Abstract
A high throughput phenotypic screening against M. smegmatis led us to the discovery of a new class of bacteriostatic, highly hydrophobic antitubercular
quinazolinones that potently inhibited the in vitro growth of either extracellular or intramacrophagic M. tuberculosis (Mtb), via modulation of an unidentified
but yet novel target. Optimization of the initial hit compound culminated in the identification of potent but poorly soluble Mtb growth inhibitors, three of
which were progressed to in vivo efficacy studies. Despite nanomolar in vitro potency and attractive PK properties, none of these compounds was
convincingly potent in our in vivo mouse tuberculosis models. This lack of efficacy may be linked to the poor drug-likeness of the test molecules and/or to
the properties of the target.
F
O
O
O
N
N
Cl
N
N
R
N
N
H
H
1
R= Cl or R = NH2
22 and 38
HIT COMPOUND
OPTIMIZED COMPOUNDS
MIC Mtb = 38 nM
Issues = high metabolism, CYP3A4 inhibition
and poor aqueous solubility
MIC Mtb 22/38 = 7/3 nM
No ADME issues but still poor solubility