Heteroaromatic-aminomethyl quinolones: Potent and selective iNOS inhibitors

Article abstract of DOI:10.1016/j.bmcl.2011.11.073

The overproduction of nitric oxide during the biological response to inflammation by the nitric oxide synthase (NOS) enzymes have been implicated in the pathology of many diseases. By removal of the amide core from uHTS-derived quinolone 4, a new series highly potent heteroaromatic-aminomethyl quinolone iNOS inhibitors 8 were identified. SAR studies led to identification of piperazine 22 and pyrimidine 32, both of which reduced plasma nitrates following oral dosing in a mouse lipopolysaccharide challenge assay.

Full text of DOI:10.1016/j.bmcl.2011.11.073

Bioorganic & Medicinal Chemistry Letters 22 (2012) 1237–1241
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Heteroaromatic-aminomethyl quinolones: Potent and selective iNOS inhibitors
Sergio G. Durón c, Andrew Lindstrom f, Céline Bonnefous b, Hui Zhang a,
Xiaohong Chen a, Kent T. Symons f, Marciano Sablad d, Natasha Rozenkrants d,
Yan Zhang d, Li Wang e, Nahid Yazdani f, Andrew K. Shiau h, Stewart A. Noble a,
Peter Rix b, Tadimeti S. Rao d, Christian A. Hassig g, Nicholas D. Smith b,
⇑
a Kalypsys Inc., San Diego, CA 92121, United States
b Aragon Pharmaceuticals, San Diego, CA 92121, United States
c Afraxis, La Jolla, CA 92037, United States
d Johnson and Johnson PRD, San Diego, CA 92121, United States
e Bioquant, San Diego, CA 92121, United States
f Helicon, San Diego, CA 92121, United States
g Sanford Burnham Research Center, San Diego, CA 92121, United States
h Ludwig Cancer Research Center, University of California, San Diego, CA 92037, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
The overproduction of nitric oxide during the biological response to inflammation by the nitric oxide
synthase (NOS) enzymes have been implicated in the pathology of many diseases. By removal of the
amide core from uHTS-derived quinolone 4, a new series highly potent heteroaromatic-aminomethyl
quinolone iNOS inhibitors 8 were identified. SAR studies led to identification of piperazine 22 and
pyrimidine 32, both of which reduced plasma nitrates following oral dosing in a mouse lipopolysaccha-
ride challenge assay.
Received 30 September 2011
Revised 17 November 2011
Accepted 18 November 2011
Available online 25 November 2011
Keywords:
Quinolone
Ó 2011 Elsevier Ltd. All rights reserved.
Nitric oxide synthase
iNOS
nNOS
Inhibitor
SAR studies
Lipopolysaccharide challenge assay
Nitric oxide synthase’s (NOS) are three closely related heme-
based oxygenases which include endothelial (eNOS), neuronal
(nNOS) and inducible (iNOS) NOS isoforms. They are responsible
for the biosynthesis of nitric oxide (NO) by the catalytic oxidation
an elevation of blood pressure. Early iNOS inhibitors such as
GW271450 (1) share structural similarity with -arginine, leading
L
to concerns with interference with physiological processes depen-
dent on arginine transport and metabolism.4 Small molecule inhib-
itors, structurally distinct from arginine but that still incorporate a
guanidine motif have also been identified. Amongst these is 2
(AR-C102222), a substrate competitive inhibitor which showed
good selectivity for iNOS and was efficacious in a rat adjuvant-
induced arthritis model.5 Inhibitors of iNOS dimerization have also
been identified such as 3 (BBS-4), a potent and selective imidazole-
based inhibitor that coordinates to the heme iron in the active site
of the enzyme.
Previously we described the discovery of a quinolinone based
iNOS dimerization inhibitor 4, identified by optimization of a hit
from an ultra high-throughput screen.6 Although a potent and
selective iNOS inhibitor, 4 suffered from high clearance, limiting
the duration of its in vivo efficacy. We thus sought to evolve the
quinolinone amide series to identify potent iNOS inhibitors with
a suitable drug property profile for development. Towards this
of the natural amino acid L-arginine to L
-citrulline.1 Endothelial
NOS (eNOS) and neuronal NOS (nNOS) are constitutively expressed
and play important roles in blood pressure regulation and
neurotransmission, respectively. The third isoform, inducible NOS
(iNOS), is synthesized in response to exposure to inflammatory
and immunologic stimuli and has been implicated in the patho-
physiology of a range of human diseases including septic shock,
asthma, inflammatory bowel disease, arthritis and neuropathic
pain.2,3
There has been substantial activity in the pharmaceutical indus-
try to identify selective iNOS inhibitors (Fig. 1), with selectivity
over eNOS being particularly important as inhibition can result in
⇑
Corresponding author.
E-mail address: Nicholas_Smith77@yahoo.com (N.D. Smith).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.11.073

1238
S. G. Durón et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1237–1241
O
H
N
O
F
H
O
S
N
N
N
N
O
NH
NH
H2N
CO2H
N
N
N
F
NH2
N
N
N
1
2
3 - BBS-4
- GW271450
- AR-C102222
N
N
O
O
S
N
N
N
Cl
N
F
N
H
O
N
H
F
F
4
5 - KD7332 (KLYP961)
Figure 1. NOS inhibitors.
end we removed the amide functionality of 6 by performing
connection ‘A’ or ‘B’ (Fig. 2), generating the ‘benzimidazole series’
(represented by 7) and the ‘heteroaromatic-aminomethyl quino-
lone series’ (represented by 8).7 Previously we described the
optimization of the benzimidazole series to the development
candidate 5 (KD7332, KLYP961);8 this Letter describes the profile
of the ‘heteroaromatic-aminomethyl quinolones’.
O
MeO
a
O
+
NH2
F
O
O
F
N
H
F
9
F
10
11
Br
O
O
b
c
The desired heteroaromatic-aminomethyl quinolones were syn-
thesized as described in Scheme 1. Amide bond formation between
aniline 9 and b-keto ester 10 provided the coupled product 11,
which was then selectively brominated using bromine/acetic acid
followed by acetone to afford the desired terminal monobromide
12.9 The bromide 12, then underwent an acid catalyzed intramo-
lecular cyclization/condensation to construct the quinolone core
13. Finally, the resulting benzylic bromide 13 was reacted with a
variety of heteroaromatic-aminomethyl synthons 14, synthesized
from coupling of substituted anilines with heteroaryl halides (via
palladium catalyzed amination or via SNAr) in the presence of so-
dium hydride. Using this general scheme the heteroaromatic-
aminomethyl quinolones 15 were prepared. The isoquinoline and
isoquinoline N-oxide illustrated in Figure 3 were in turn prepared
using the methodology presented previously.6
F
N
F
N
O
H
H
F
Br
F
12
13
H
N
d
R2
NH2
+
e
R1
R1
R2
X
R1
N
14
R2
F
N
O
R1, R2 = Aromatic or Heteroaromatic
H
F
15
The inhibitory activity against human iNOS of the compounds
was measured as described previously using transiently transfec-
ted HEK-293 cells, using 2,3-diaminonaphthalene to measure NO
production via fluorescence.6 To determine NOS selectivity, human
eNOS and nNOS inhibitory activity was also determined using an
analogous assay.
Scheme 1. Reagents and conditions: (a) Xylene, pyridine reflux, 3–18 h; (b) (i) Br2,
AcOH, 25–50 °C, 5–18 h; (ii) acetone, 25 °C, 2 h; (c) H2SO4, 40 °C, 2–18 h; (d)
Pd2(dba)3, BINAP, NaOtBu, toluene, microwave, 150 °C, 15 min or NMP, microwave,
200 °C, 15 min; (e) NaH, DMF, 25 °C, 1–5 h.
selectivity over eNOS (>1000Â), while showing slight selectivity
over nNOS (fivefold). The importance of having a nitrogen in the
ring of 17 (that is believed to mimic the amide of 6—Fig. 2) was
clearly demonstrated as the phenyl derivative 16 lost considerable
Encouragingly, the prototypical compound of the ‘heteroaro-
matic-aminomethyl quinolone series’, compound 17 (Table 1)
showed excellent human iNOS activity (IC50 = 0.031
lM) and
'Heteroaromatic-aminomethyl
'Benzimidazole Series'
Quinolone Series'
B
N
N
S
R
N
S
N
O
O
A
N
N
Cl
N
Cl
Connection A
Connection B
F
N
H
O
F
N
H
O
F
N
H
F
F
F
7
6
8
Figure 2. Restricting conformations of the amide of 6 leads to two new series.

S. G. Durón et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1237–1241
Cl Cl
1239
Cl
N
N
N
N
N
N
N
N
O
N
N
N+
N
H
F
F
F
O-
F
F
F
39
40
hiNOS = 0.017 µM
23
hiNOS = 0.20 µM
hiNOS = 0.013 µM
heNOS/hiNOS = 425
hnNOS/hiNOS = 5
heNOS/hiNOS > 1000 heNOS/hiNOS > 1000
hnNOS/hiNOS = ND hnNOS/hiNOS = 9
Figure 3. Isoquinoline N-oxide replacement of the quinolone moiety.
Table 1
Table 2
Nature of the heteroaromatic ring effects iNOS potency
Benzylic-substituted aminopyrazines are well tolerated
Cl
Y
N
N
N
N
R1
F
N
H
O
F
F
N
H
O
F
a
Compound
Y
IC50
(
l
M)
Selectivity
Selectivity
eNOS/iNOS nNOS/iNOS
a
Compound R1
IC50
(
l
M)
Selectivity
eNOS/iNOS
Selectivity
nNOS/iNOS
iNOS
eNOS nNOS
iNOS
1.5
eNOS nNOS
25
26
27
28
29
30
31
H
0.013 6.4
0.012 12
0.22
0.077 7.1
0.012 11
0.010 13
0.1
500
1000
40
90
900
8
5
2
2
5
5
6
2-Cl
3-Cl
4-Cl
2-F
0.06
0.33
0.12
0.064
16
17
>30b
NDc
0.17
>20
NAd
5
8.3
N
0.031 >30b
0.077 14
>1000
2-Me
0.046 1300
0.44 140
N
[2-Pyridyl] 0.078 11
18
19
20
0.37
0.24
0.36
180
170
500
5
2
8
a
Cell-based NOS assay using transiently transfected HEK293 cells; NO measured
using 2,3-diaminonapthalene (DAN).
0.13
23
N
N N
N
0.043 21
0.014 13
21
22
0.15
1000
11
2
benzylic substituent was not well tolerated (data not shown). In
contrast to amide 6, a benzyl-substituent on amino-piperazine 22
was well tolerated as illustrated by 25 (Table 2), which maintained
N
N
0.005 1.5
0.013 5.5
0.012 300
0.069 425
N
hiNOS potency (IC50 = 0.013 lM) and selectivity (eNOS/iNOS =
N
N
500). Further, substitution around the benzylic ring was tolerated
with small, lipophilic substituents at the ortho-position preferred
as demonstrated by 26, 29 and 30 which maintained hiNOS
23
24
5
8
N
N
0.21
79
1.6
375
potency (IC50 <0.015 lM) and selectivity (eNOS/iNOS >500).
meta-Chloro (27) or para-chloro (28) both suffered from a drop
off in potency, as did 2-pyridyl 31. Despite maintaining good levels
of in vitro potency, derivatives of this type were not pursued due to
short liver microsomal half-life’s (t1/2 <20 min).10
Further SAR work examining substitution on the heteroaromatic
ring was carried out as exemplified in Table 3 using pyrimidine 21
as a base compound. It was found that methyl substitution at the
3- or 5-positions of pyrimidine 21 were preferred—compare 32
and 33–34 (pyrimidine version was not prepared). Importantly,
32 also had a reasonable mouse liver microsome half-life of
90 min, substantially longer than many of the other derivatives in
the heteroaromatic-aminomethyl quinolone series, which lead to
its profiling in vivo (vide infra). More polar groups however were
a
Cell-based NOS assay using transiently transfected HEK293 cells; NO measured
using 2,3-diaminonapthalene (DAN).
b
<50% inhibition at 30
Not determined.
Not applicable.
lM.
c
d
potency (hiNOS IC50 = 1.5 lM). For optimal iNOS potency, the posi-
tion of the nitrogen atom was also important with the 3-pyridyl
(18) and 4-pyridyl (19) giving a drop off in potency. Fixing the
2-pyridyl in place and then scanning a second nitrogen around
the ring (20–23) produced additional potent heteroaromatic
compounds including pyrimidine 21 (hiNOS IC50 = 0.014
piperidine 22 (hiNOS IC50 = 0.005 M) with the good selectivity
over eNOS maintained (300 to 1000-fold). Further piperidine 22
showed good potency against mouse iNOS (IC50 = 0.008 M), mak-
lM) and
l
only tolerated at the 3-position as illustrated by 36 (IC50
=
0.028 M) and 37 (IC50 = 0.061 M) compared to 35 (IC50 = 1.3 M).
l
l
l
l
Finally, high potency compounds, that approximately fill the same
ing it an attractive compound for profiling in vivo (vide infra), as
space as the thiazole-carbonyl portion of 6 could be obtained with
well as for further SAR studies.
fused systems such as pyrimidinopyrazole 38 (IC50 = 0.011 lM).
Previous SAR had illustrated that a 3-chlorophenyl ring was a
preferred substituent in the amide series (i.e., 6)6 and that a
However, despite its high human iNOS potency and good selectiv-
ity, 38 suffered from a short liver microsome half life of <30 min.

1240
S. G. Durón et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1237–1241
Previously we showed that the key quinolone moiety maybe
aminomethyl quinolone series as illustrated in Figure 3, where iso-
quinoline N-oxide 40 is equipotent to its quinolone analogue 23,
also maintaining excellent selectivity over eNOS (>1000-fold).
replaced with an isoquinoline group in the amide series.6
This replacement can also be carried out in the heteroaromatic-
Table 3
Effect of substitution on the heteroaromatic ring
Cl
N
R1
F
F
N
H
O
R1
IC50
(lM)
Selectivity eNOS/iNOS
Selectivity nNOS/iNOS
a
Compound
iNOS
eNOS
nNOS
0.15
N
21
32
0.014
13.3
1000
230
11
7
N
N
0.033
0.053
7.5
0.213
0.160
N
N
N
33
34
>30b
>500
>85
3
N
0.35
1.3
>30b
>30b
NDc
NDc
NAd
N
N
35
36
37
N
>20
NAd
20
2
N
N
0.028
0.061
>30b
>30b
0.500
0.13
>1000
>500
N
N
N
N
OMe
N
38
0.011
8.3
0.062
730
3
N
N
a
b
c
Cell-based NOS assay using transiently transfected HEK293 cells; NO measured using 2,3-diaminonapthalene (DAN).
<50% inhibition at 30
Not determined.
Not applicable.
lM.
d
800
600
400
200
0
600
500
400
300
200
100
0
ED50=3.6 mpk (po)
*
*
*
*
*
*
*
LPS + 22 (mpk, po)
*: p<0.05 vs. LPS+Vehicle (1 way ANOVA)
*: p<0.05 vs. LPS+Vehicle (1 way ANOVA)
Figure 4. Performance of 32 and 22 following oral dosing in the mouse LPS challenge assay.

S. G. Durón et al. / Bioorg. Med. Chem. Lett. 22 (2012) 1237–1241
1241
Table 4
Mouse iNOS IC50 and pharmacokinetics of 22 and 5
Compound
hiNOS EC50
(l
M)
miNOS EC50
(lM)
Clp (mL/min/kg)
t1/2 (h)
Vdss (L/kg)
po Cmax (
l
g/mL)
po AUC (
lg h/mL)
%F
22
5
0.005
0.091
0.005
0.5
34
4.5
2.1
2.1
1.7
0.4
0.045
15
0.12
25
3
62
Isoquinoline 39 was also active against hiNOS, albeit 10-fold less
potent.
heteroaromatic-aminomethyl quinolone iNOS inhibitors that dis-
played excellent selectivity over endothelial NOS (>300-fold) and
moderate/low selectivity over neuronal NOS (2 to 20-fold). A het-
eroaryl ring containing a nitrogen at the 2-position was found to
be optimal for potency, with piperazine 22 and pyrimidine 32
demonstrating a statistical reduction in plasma nitrate levels fol-
lowing oral dosing in a mouse LPS challenge assay. Despite
in vivo activity, the series was discontinued in favor of the benz-
imidazole series due to inferior pharmacokinetics.
Due to their encouraging in vitro profiles, piperazine 22 and
pyrimidine 32 were profiled in a lipopolysaccharide (LPS) chal-
lenge assay in mice. This assay which measures nitrate levels (via
NO levels—a biomarker reflective of activity in mechanism), was
used extensively in the program to confirm that increases in
in vitro iNOS potency translated to increases in in vivo activity,
but also to get a rapid read on whether a compound had exposure
following oral dosing. As can be seen in Figure 4, at 30 mg/kg, both
22 and 32 gave a statistical reduction in plasma nitrate levels of
80% and 55%, respectively. Due to its superior performance, a
dose–response from 0.3 to 30 mg/kg was carried out with 22, giv-
ing an ED50 of 3.6 mpk (po).
References
1. Alderton, W. K.; Cooper, C. E.; Knowles, R. G. Biochem. J. 2001, 357, 593.
2. (a) Vallance, P.; Leiper, J. Nat. Rev. Drug Disc. 2002, 1, 939.
3. Levy, D.; Zochodne, D. W. Pain Pract. 2004, 4, 11.
Although 22 performs well in the LPS assay, benzimidazole 5
gave superior performance (ED50 = 1 mpk po) in this assay. This
is despite the fact that 5 was significantly less potent against
4. Edwards, R. M.; Stack, E. J.; Trizna, W. J. Pharmacol. Exp. Ther. 1998, 285, 1019.
5. (a) LaBuda, C. J.; Koblish, M.; Tuthill, P.; Dolle, R. E.; Little, P. J. Eur. J. Pain 2005,
10, 505; (b) Tinker, A. C.; Beaton, H. G.; Boughton-Smith, N.; Cook, T. R.; Cooper,
S. L.; Fraser-Rae, L.; Hallam, K.; Hamley, P.; McInally, T.; Nicholls, D. J.; Pimm, A.
D.; Wallace, A. V. J. Med. Chem. 2003, 46, 913.
6. Bonnefous, C.; Payne, J. E.; Roppe, J.; Zhang, H.; Chen, X.; Symons, K. T.; Nguyen,
P. H.; Sablad, M.; Rozenkrants, N.; Zhang, Y.; Wang, L.; Severance, D.; Walsh, J.
P.; Yazdani, N.; Shiau, A. K.; Noble, S. A.; Rix, P.; Rao, T. S.; Hassig, C. A.; Smith,
N. D. J. Med. Chem. 2009, 52, 3047.
7. As disclosed previously, modification of the quinolinone portion of 8 did not
lead to an improvement in rodent pharmacokinetics. See Ref. 6.
8. Payne, J. E.; Bonnefous, C.; Symons, K. T.; Nguyen, P. M.; Sablad, M.;
Rozenkrants, N.; Zhang, Y.; Wang, L.; Yazdani, N.; Shiau, A. K.; Noble, S. A.;
Rix, P.; Rao, T. S.; Hassig, C. A.; Smith, N. D. J. Med. Chem. 2010, 53, 7739.
9. Choi, H. Y.; Chi, D. Y. Org. Lett. 2003, 5, 411.
mouse iNOS than 22 (miNOS = 0.5 lM vs 0.005 lM, respectively).
This relative performance can be readily understood by examining
the mouse pharmacokinetics of 22 (Table 4), which is characterized
by limited exposure following oral dosing (%F = 3; AUC (10 mpk
po) = 0.12
po) = 25
l
g h/mL), in contrast to
5 (%F = 62; AUC (10 mpk
l
g h/mL). Despite the excellent in vitro potency and selec-
tivity over eNOS of the heteroaromatic-aminomethyl quinolone
series, we were unable to identify compounds with pharmacoki-
netic parameters comparable to the benzimidazole series and so
this series was discontinued.
In summary, by removing the amide bond in the original
screening hit, we were able to identify a series of highly potent
10. Elimination rate under combined Phase 1 oxidative and Phase 2 glucuronide
conjugation
conditions
(mouse)—for
experimental
protocol,
see
Supplementary data for Ref. 8.