Synthesis and optimization of 2-pyridin-3-yl-benzo[d][1,3]oxazin-4-one based inhibitors of human neutrophil elastase

Article abstract of DOI:10.1016/j.bmcl.2009.06.053

The hit-to-lead optimization of the HNE inhibitor 5-methyl-2-(2-phenoxy-pyridin-3-yl)-benzo[d][1,3]oxazin-4-one is described. A structure-activity relationship study that focused on the 5 and 7 benzoxazinone positions yielded the optimized 5-ethyl-7-metho

Full text of DOI:10.1016/j.bmcl.2009.06.053

Bioorganic & Medicinal Chemistry Letters 19 (2009) 4743–4746
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and optimization of 2-pyridin-3-yl-benzo[d][1,3]oxazin-4-one
based inhibitors of human neutrophil elastase
a
a
a
a
b
a
Kevin R. Shreder a, , Julia Cajica , Lingling Du , Allister Fraser , Yi Hu , Yasushi Kohno , Emme C. K. Lin ,
*
Steve J. Liu a, Eric Okerberg a, Lan Pham a, Jiangyue Wu a, John W. Kozarich a
a ActivX Biosciences, Inc., 11025 N. Torrey Pines Road, La Jolla, CA 92037, USA
b Discovery Research Laboratories, Kyorin Pharmaceutical Co. Ltd, 2399-1, Nogi, Nogi-machi, Shimotsuga-gun, Tochigi 329-0114, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
The hit-to-lead optimization of the HNE inhibitor 5-methyl-2-(2-phenoxy-pyridin-3-yl)-benzo[d][1,3]oxa-
zin-4-one is described. A structure–activity relationship study that focused on the 5 and 7 benzoxazinone
positions yielded the optimized 5-ethyl-7-methoxy-benzo[d][1,3]oxazin-4-one core structure. 2-[2-(4-
Methyl-piperazin-1-yl)-pyridin-3-yl] derivatives of this core were shown to yield HNE inhibitors of similar
potency with significantly different stabilities in rat plasma.
Received 7 May 2009
Revised 10 June 2009
Accepted 15 June 2009
Available online 17 June 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Human neutrophil elastase
Benzoxazinone
Inhibitor
Hydrolytic stability
Plasma stability
Human neutrophil elastase (HNE) is a serine protease, stored in
and secreted from neutrophils, that is capable of degrading a vari-
ety of structural proteins of the extracellular matrix.1 Infiltration of
activated neutrophils has been implicated in several lung diseases,
including ARDS and emphysema.2 Thus, HNE inhibitors that target
the excess production of this protease in these indications have the
potential as therapeutic agents. A variety of structural classes of
HNE inhibitors have been developed, including those based on
benzoxazinones3 and related thienoxazinones4 and several HNE
inhibitors have advanced to the clinic.5
As part of an effort to find novel HNE inhibitors, we conducted a
high throughput screen of an internal compound library which led
to the identification of compound 1 as a potent HNE inhibitor (Ta-
ble 1).6 We then set out to determine what structural elements of
this compound contributed to its potency. With that information in
hand, the core benzoxazinone scaffold was optimized for potency
and stability. A position of the core structure was identified where
polar and non-polar groups could be incorporated without signifi-
cantly impacting HNE inhibitor potency. Taking advantage of this
position, HNE inhibitors were synthesized that had similar HNE
IC50 values but with a wide range of rat plasma stabilities.
The generalized synthesis of 2-pyridin-3-yl-benzoxazinones
used in the exploratory SAR study is shown in Scheme 1. Such ben-
zoxazinones could be obtained in a one-pot fashion by the activa-
tion of nicotinic acid derivatives with carbonyldiimidazole (CDI)
followed by the addition of an anthranilic acid (1.2 equiv). The
resulting amide was then ring-closed to the benzoxazinone using
N-(3-dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride
(EDCI). Furthermore, aryl acid starting materials or intermediates
could be conveniently removed at the end of the reaction using a
solid phase extraction column containing an anion exchange resin
(Varian, PL-HCO3 MP SPE Resin) prior to a final C18 reverse phase
HPLC purification (CH3CN/H2O, 0.1% AcOH).
Compounds for the initial SAR study were designed to probe the
absence, position, or modification of substituents present in the hit
compound 1. The impact of deleting substituents was investigated
first. The importance of the pyridyl nitrogen was demonstrated by
comparing the HNE IC50 value of compound 2 to the original hit;
substitution of the pyridine ring of compound 1 with a phenyl ring
resulted in a threefold loss in potency. When compounds that rep-
resented the deletion from the hit 1 of the B5 methyl group (com-
pound 3) or the phenoxy group (compound 4) were synthesized
and tested, a 150-fold and 15-fold loss of potency, respectively,
were seen relative to compound 1. When the phenoxy group of
compound 1 was substituted with a methoxy group (compound
5), the resulting IC50 value was similar to compound 1. To further
explore the impact of substitution in this position, compounds 6
and 7 were synthesized and compared. The P2 methyl derivative
6 yielded an inhibitor with potency comparable to the hit 1. By
contrast, the P6 methyl derivative 7 resulted in a 12-fold loss in po-
tency relative to the hit 1. Additionally, substitution of the B5
* Corresponding author.
E-mail address: kevins@activx.com (K.R. Shreder).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.06.053

4744
K. R. Shreder et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4743–4746
Table 1
O
N
HNE IC50 values for compounds 1–9 and 13–16
R1
H
O
F
N
R2
+
100 oC, 10 min
dioxane
O
O
B5
O
N
N
O
O
O
10
N
N
O
O
PhO
N
PhO
P2
N
P6
NR1R2
O
O
1
2
3-9, 13-17
+
N
H
N
R1
Compound
B5
P2
P6
HNE IC50 (nM)
N
R2
N
F
N
1
2
3
4
5
6
7
8
9
61
170
9300
900
85
11
12
–H
PhO–
H–
MeO–
Me–
H–
–H
–H
–H
–H
–Me
–H
–H
–Me
–Me
–Me
–Me
–Cl
Scheme 2. Generalized synthesis of 5-methyl-2-(2-amino-pyridin-3-yl)-benzo[d]-
[1,3]oxazin-4-one derivatives.
87
710
160
390
PhO–
PhO–
–OCH3
sation of 2-fluoro-nicotinic acid and 6-methyl-anthranilic acid
according to Scheme 1 produced compound 10. Subsequent dis-
placement of the fluoride in a SNAr fashion with aliphatic amines
readily occurred at 100 °C in dioxane to yield compounds of the
form 11 and 12. Secondary amines gave the desired product 11
and only a trace to none of the tertiary amide 12. By contrast, pri-
mary amines yielded the unwanted secondary amide 12 as the ma-
jor product.
The use of the primary amines 4-(2-aminoethyl)morpholine or
N,N-dimethylaminoethylenediamine as shown in Scheme 2 pro-
duced compounds 13 and 14, respectively. Both compounds exhib-
ited micromolar HNE IC50 values. However, when the secondary
amine N,N,N’-trimethylethylenediamine was used, the resulting
product 15 exhibited a 110 nM HNE IC50 value. A direct compari-
son between compounds 14 and 15, which differ by a single N–
Me group, led us to the preferred use of secondary amines versus
primary amines in our efforts to design potent inhibitors. Interest-
ingly, when 6-methyl-anthranilic acid was added to two equiva-
lents of carbonyldiimidazole (CDI) activated 2-fluoro-nicotinic
acid and the reaction was heated to 70 °C, compound 16 was ob-
tained via fluoride displacement by imidazole, a byproduct of CDI
activation. Compound 16 had an HNE IC50 value of 64 nM, a value
similar to the hit 1, and indicated that heteroaryl groups also were
tolerated in the P2 position.
O
N
13
14
–Me
–Me
–H
–H
6000
8100
N
H
N
N
N
H
15
16
–Me
–Me
–H
–H
110
64
N
CH3
N
N
HOOC
1) CDI, CH3CN, 2 hours, r.t.
2)
X
O
N
R
OH
NH2
Y
With the importance of a methyl group at the B5 position estab-
lished, an effort was undertaken to evaluate the effect of other
small alkyl and cycloalkyl groups in this same position. In parallel
with this effort, electron donating groups were incorporated into
the 7-position, a known strategy to improve the chemical stability
of benzoxazinones.3d For these studies, the requisite anthranilic
acids were synthesized using three different routes. 6-Ethyl-
anthranilic acid was produced from the oxidative cleavage of 4-
ethyl isatin.7 4-Methoxy anthranilic acids were synthesized (see
Scheme 3) from the Diels–Alder reaction of 6-substituted 4-meth-
oxy-2-pyrones8 with dimethyl acetylenedicarboxylate followed by
the regiospecific saponification of the resulting dimethyl phthalate
with NaOH (1 M). The reaction of the resulting acid with diphenyl-
phosphoryl azide in MeOH yielded the methylcarbamate protected
amine via a Curtius rearrangement. We chose not to produce the
free amine directly (by treating the intermediate isocyanate with
water instead of MeOH) for the following reasons: Comparing
the isolated yields of these two strategies, we found the yields of
the carbamate to be higher. In addition, when choosing to store
an intermediate, the methyl carbamate was found to be stable ver-
sus the 4-methoxy-anthranilic acids which decarboxylated upon
storage. Ultimately, the carbamate could be deprotected using
LiOH to yield the required 4-methoxy-6-substituted anthranilic
X
B5
O
N
COOH
O
O
EDCI
Y
N
H
B7
N
N
R
P
Scheme 1. Generalized synthesis of 2-pyridin-3-yl-benzo[d][1,3]oxazin-4-ones.
The positions of the substituents on the benzoxazinone ring are identified trivially
with a B. Positions on the pyridin-3-yl group are identified trivially with a P.
methyl group of hit 1 was explored. The replacement of this methyl
group with the isosteric chloro (compound 8) or methoxy (com-
pound 9) resulted in HNE inhibitors of decreased potency relative
to the hit 1 (2.6-fold and 6.4-fold, respectively). This SAR foray
led us to conclude (a) the B2 pyridin-3-yl group was preferred over
a B2 phenyl group (b) P2 substitution (vs P6 substitution or a H in
the P2 position) yielded more potent inhibitors (c) the P2 position
was tolerant of groups that varied considerably in size and (d)
varying the B5 substituent could dramatically modulate HNE
inhibitory activity.
Replacement of the phenoxy group of the hit 1 with amines was
also investigated using the route shown in Scheme 2. The conden-

K. R. Shreder et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4743–4746
4745
Table 2
X
X
MeO2C
CO2Me
HNE IC50 values for compounds 17–40 and half-life data in 200 mM sodium borate
buffer, pH 9.2 for compounds 20–27
CO2Me
CO2Me
O
180 o
C
MeO
O
MeO
B5
O
N
B5
O
N
B5
O
N
O
N
O
N
O
N
X
(PhO)2PON3
NaOH
CO2Me
CO2H
B7
B7
B7
N
NEt3
dioxane/MeOH
100 oC
DME, r.t.
N
N
N
MeO
N
N
HO2C
X
X
LiOH
THF/H2O/MeOH
CO2Me
NHCO2Me
CO2H
NH2
A
B
C
MeO
MeO
Compound Series B5
B7
HNE IC50 (nM) Half-life pH 9.2 (min)
Scheme 3. Synthesis of 6-substituted 4-methoxy-anthranilic acids from 6-substi-
tuted 4-methoxy-2-pyrones.
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
A
B
C
A
B
C
A
B
C
A
C
A
B
C
A
B
C
–CH3
–CH3
–CH3
–H
–H
–H
–H
–H
–H
93
79
74
8.1
8.6
8.5
20
28
28
n.d.
n.d.
n.d.
65
89
56
240
1000
1400
7300
6900
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH3
–CH2CH2CH3 –OMe
–CH2CH2CH3 –OMe
–CH2CH2CH3 –OMe
–CH(CH3)2
–CH(CH3)2
–CH(CH3)2
acids. Finally, the one-pot tosylation of 4-hydroxy-6-ethyl 2-pyr-
one followed by displacement of the intermediate tosyl group with
dimethylamine yield 4-dimethylamino-6-ethyl-2-pyrone (Scheme
4). This 2-pyrone was then converted to 4-dimethylamino-6-
ethyl-2-anthranilic acid as shown in Scheme 3.
–OMe
–OMe
–OMe
–N(CH3)2 5300
–N(CH3)2 2600
The B5 and B7 positions of the benzoxazinone were shown to be
important modulators of HNE inhibitor potency. Three benzoxazi-
none series, derivatized in the P2 position with N-methyl
piperazine (Series A), isonipecotic acid (Series B), and 1-(N-
methyl-piperidin-4-yl)piperazine (Series C) were examined. Start-
ing with the same core present in hit 1 (B5 = methyl and B7 = H,
compounds 17–19) we found that despite the charge and steric dif-
ferences present in these series, the HNE IC50 values remained
close, ranging from 74 nM to 93 nM. Substituting the B5 methyl
group with an ethyl group resulted in inhibitors with single digit
nM IC50 values (20–22). Fixing this B5 ethyl group, we then exam-
ined the impact of placing electron donating groups in the B7 posi-
tion on hydrolytic stability.9 Previously, the addition of electron
donating groups to the 7-position of benzoxazinones have been
shown to enhance their hydrolytic stability by tempering the reac-
tivity of the carbonyl to nucleophilic attack.3d Two electron donat-
ing groups were investigated, methoxy and dimethylamino, and
both were shown to increase hydrolytic stability (Table 2, compare
the t1/2 values of compound sets 20–22 to 23–25 to 26–27). While
the dimethylamino containing inhibitors 26 and 27 exhibited the
longest half-lives, they unfortunately also exhibited micromolar
HNE IC50 values. In contrast, the 7-methoxy derivatives 23–25
yielded HNE IC50 values from 20 to 28 nM. In order to choose a
benzoxazinone scaffold with a balance between hydrolytic stabil-
ity and potency, we locked in on 7-methoxy derivatives while
continuing to examine the B5 position. Ultimately, the 5-ethyl-7-
methoxy scaffold was found to be optimal. Extending the B5 ethyl
group of compounds 23–25 with n-propyl (28–30) or replacing the
ethyl group with isopropyl (31–33), cyclopropyl (34–36), or cyclo-
butyl (37–39) resulted in HNE inhibitors with diminished potency
by a factor that ranged from 3 to 46 relative to their ethyl congen-
ers. In the one case when the B5 ethyl group was replaced with
tert-butyl group, a 750-fold reduction in potency was observed
(compare compounds 23 and 40).
250
180
71
210
330
140
–OMe
–OMe
–OMe
34
35
36
A
B
C
–OMe
–OMe
–OMe
120
140
110
n.d.
n.d.
n.d.
37
38
39
40
A
B
C
A
–OMe
–OMe
–OMe
–OMe
150
n.d.
n.d.
n.d.
n.d.
1300
850
–C(CH3)3
15,000
n.d. = not determined.
In anticipation of using these compounds in rodent efficacy
models, the rat plasma stabilities of selected derivatives were
determined10 and shown to vary considerably based on structure
(Table 3). All compounds examined contained the same, optimized
2-(pyridin-3-yl)-5-ethyl-7-methoxy-benzoxazinone core and dif-
fered by the amine used to derivatize the P2 position. Series D
was derived from N-substituted piperazines and Series E was de-
rived from carboxy-substituted piperidines. HNE IC50 values were
also measured for these compounds and found not to vary more
than a factor of two among the 14 compounds in Table 3. By con-
trast, plasma stability t1/2 values varied considerably, from a high
of 270 min for compound 41 to a low of 7.8 min for compound
52. Structure–activity relationship information could be gleaned
from this data set. Carboxy-substituted derivatives (compounds
24 and 41–43) whether from series D or E were found to be among
the most plasma-stable inhibitors. For series D, non-polar N-sub-
stituents of the piperazine ring were found to yield more stable
inhibitors versus polar N-substituents (compounds 49–52).
1) Ts-Cl, TEA, DCM
O
O
2) Me2NH, THF
N
O
In conclusion, the synthesis and optimization of novel 2-pyri-
din-3-yl-benzo[d][1,3]oxazin-4-one derived HNE inhibitors have
been described. From among the 5,7-substituted benzoxazinones
HO
O
Scheme 4. Synthesis of 4-dimethylamino-6-ethyl-2-pyrone.

4746
K. R. Shreder et al. / Bioorg. Med. Chem. Lett. 19 (2009) 4743–4746
Table 3
References and notes
Half-life in rat plasma and HNE IC50 values for compounds 23, 24, 41–52
1. (a) Travis, J. Am. J. Med. 1988, 84, 37; (b) Owen, C. A.; Campbell, E. J. Semin. Cell
Biol. 1995, 6, 367.
O
N
O
N
2. (a) Moraes, T. J.; Chow, C. W.; Downey, G. P. Crit. Care Med. 2003, 31, 189; (b)
Wang, Z.; Chen, F.; Zhai, R.; Zhang, L.; Su, L.; Lin, X.; Thompson, T.; Christiani, D.
C. PLoS ONE 2009, 4, e4380; (c). Int. J. Tuberc. Lung Dis. 2008, 12, 361.
3. (a) Colson, E.; Wallach, J.; Hauteville, M. Biochimie 2005, 87, 223; (b) Arcadi, A.;
Asti, C.; Brandolini, L.; Caselli, G.; Marinelli, F.; Ruggieri, V. Bioorg. Med. Chem.
Lett. 1999, 9, 1291; (c) Uejima, Y.; Oshida, J.; Kawabata, H.; Kokubo, M.; Kato,
Y.; Fujii, K. Biochem. Pharmacol. 1994, 48, 426; (d) Krantz, A.; Spencer, R. W.;
Tam, T. F.; Liak, T. J.; Copp, L. J.; Thomas, E. M.; Rafferty, S. P. J. Med. Chem. 1990,
33, 464; (e) Radhakrishnan, R.; Presta, L. G.; Meyer, E. F., Jr.; Wildonger, R. J. Mol.
Biol. 1987, 198, 417.
O
N
O
N
MeO
MeO
N
N
N
R
R
4. (a) Gütschow, M.; Neumann, U. J. Med. Chem. 1998, 41, 1729; (b) Gütschow, M.;
Kuerschner, L.; Neumann, U.; Pietsch, M.; Löser, R.; Koglin, N.; Eger, K. J. Med.
Chem. 1999, 42, 5437.
5. (a) Metz, W. A.; Peet, N. P. In Elastase Inhibitors in High Throughput Screening for
Novel Anti-Inflammatories; Kahn, M., Ed.; Birkhäuser Verlag Basel: Switzerland,
2000; p 49; (b) Ohbayashi, H. Exp. Opin. Ther. Pat. 2002, 12, 65; (c) Ohbayashi, H.
IDrugs 2002, 5, 910.
6. Human sputum neutrophil elastase (Elastin Products Co.) was diluted into
Assay Buffer A (200 mM Tris pH 7.4, 1 mg/ml BSA) to a working concentration
of 0.55 U/ml. Inhibitors dissolved and diluted in DMSO at 50ꢀ were added to
the elastase in Assay Buffer A at final concentrations ranging from 1 ꢀ 10ꢁ4 M
to 7 ꢀ 10ꢁ12 M and preincubated for 20 min at room temperature. DMSO alone
was used as the negative control. MeOSuc-AAPV-AMC (Bachem) substrate was
dissolved in DMSO to 20 mM and further diluted to 1 mM in Assay Buffer A
immediately before use. Substrate was added to the elastase assay at a final
concentration of 1 ꢀ 10ꢁ4 M. The reaction was allowed to proceed for 20 min at
room temperature and then quenched with acetic acid at a final concentration
of 3% (v/v). The AMC fluorescence was measured using a Wallac (Perkin Elmer)
Victor2 plate reader equipped with excitation/emission filters of 355/460 nm.
Fluorescence intensity versus inhibitor concentration was plotted and fit to the
Hill equation to quantify IC50 values.
D
E
Compound
Series—R group
Rat plasma t1/2 (min) HNE IC50 (nM)
D
E
41
42
43
–CH2CO2H
—
—
–
270
23
19
25
(R)-3-CO2H 150
(S)-3-CO2H 140
44
—
90
34
23
24
–CH3
—
—
84
81
20
28
4-CO2H
45
—
69
28
46
47
48
–(CH2)2OCH3
–(CH2)3CH3
–CH2CH(CH3)2
—
—
—
52
51
48
16
27
21
7. Hey, D. H.; Lobo, L. C. J. Org. Chem. 1952, 17, 164.
8. 4-Hydroxy-2-pyrones were readily prepared using
a modification of the
N
49
—
19
22
Katritzky method in which 2,2,6-trimethyl-1,3-dioxin-4-one was acylated
using short chain alkyl or cycloalkyl acid chlorides. Subsequent O-methylation
with dimethylsulfate yielded the 4-methoxy-pyrones. See: (a) Katritzky, A. R.;
Wang, Z.; Wang, M.; Hall, C. D.; Suzuki, K. J. Org. Chem. 2005, 70, 4854; (b)
Deshpande, V. H.; Khan, R. A.; Ayyanagar, N. R. Indian J. Chem. 1996, 35B, 790.
9. Alkaline hydrolysis rates were measured by the disappearance of the
characteristic absorbance maximum for the intact benzoxazinone ring
(340 nm 10 nm) in 200 mM, pH 9.2 sodium borate and 1% DMSO.
Hydrolysis rates were determined through linear regression analysis of the
progress curves and then converted to half-life data. See: Stein, R. L.; Strimpler,
A. M.; Viscarello, B. R.; Wildonger, R. A.; Mauger, R. C.; Trainor, D. A.
Biochemistry 1987, 26, 4126.
O
50
51
52
–(CH2)2OH
–(CH2)2CN
–(CH2)2N(CH3)2
—
—
—
17
11
7.8
16
24
21
examined, the 5-ethyl-7-methoxy scaffold was found to have the
best balance of chemical stability and potency. Furthermore, using
this optimized core derivatization of the P2 position with various
piperazines and piperidines resulted in inhibitors with very differ-
ent rat plasma stabilities. This data indicates that the P2 position
can be modified to modulate biochemical properties. Studies on
the cell permeability of these novel HNE inhibitors as well as their
efficacy in lung injury models will be reported in due course.
10. Test compounds (2 lM) were incubated in heparin treated rat plasma at 37 °C.
Following precipitation of the plasma proteins with acetonitrile, the
concentration of test compound was quantified against an internal standard
using ESI-MS. Half-life data was calculated from
concentration versus time to a first order rate constant.
a fit of compound