Identification of 2,3-disubstituted pyridines as potent, non-emetic PDE4 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2014.04.052

A series of 2,3-disubstituted pyridines were synthesized as potential non-emetic PDE4 inhibitors. To decrease brain exposure and minimize emesis, we modified the lipophilic moiety of a series of emetic PDE4 inhibitors and found that introduction of a hydroxy group into the pyridine moiety of the side chain led to non-emetic compounds with preserved PDE4 inhibitory activity. Following optimization at the phenoxy group, we identified compound 1 as a potent non-emetic PDE4 inhibitor. Compound 1 showed significant efficacy in an animal model of asthma without inducing emesis.

Full text of DOI:10.1016/j.bmcl.2014.04.052

Bioorganic & Medicinal Chemistry Letters 24 (2014) 2689–2692
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of 2,3-disubstituted pyridines as potent, non-emetic
PDE4 inhibitors
⇑
Motoji Kawasaki, Akira Fusano , Tomohiro Nigo, Shunya Nakamura, Mari N. Ito, Yasuhiro Teranishi,
⇑
,
Satoshi Matsumoto, Hiroshi Toda, Naruaki Nomura, Takaaki Sumiyoshi
Drug Research Division, Dainippon Sumitomo Pharma Co., Ltd., 33-94 Enoki-cho, Suita, Osaka 564-0053, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of 2,3-disubstituted pyridines were synthesized as potential non-emetic PDE4 inhibitors. To
decrease brain exposure and minimize emesis, we modified the lipophilic moiety of a series of emetic
PDE4 inhibitors and found that introduction of a hydroxy group into the pyridine moiety of the side chain
led to non-emetic compounds with preserved PDE4 inhibitory activity. Following optimization at the
phenoxy group, we identified compound 1 as a potent non-emetic PDE4 inhibitor. Compound 1 showed
significant efficacy in an animal model of asthma without inducing emesis.
Received 18 March 2014
Revised 10 April 2014
Accepted 12 April 2014
Available online 20 April 2014
Keywords:
Phosphodiesterase 4
PDE4 inhibitors
Ó 2014 Elsevier Ltd. All rights reserved.
2,3-Disubstituted pyridine
Structure activity relationship
Anti-inflammatory agent
Type 4 cAMP phosphodiesterase (PDE4) specifically hydrolyzes
cAMP, a pivotal second messenger, and its inhibition is known to
effectively increase intracellular cAMP levels and to regulate vari-
ous cellular functions.1 PDE4 is reported to be expressed in key
effector cells involved in asthma, particularly airway smooth mus-
cle cells as well as in inflammatory cells, including eosinophils,
neutrophils, T-lymphocytes, and macrophages. It has been
reported that PDE4 inhibitors significantly increase intracellular
cAMP in airway smooth muscle cells, thereby providing a bron-
chodilatory effect.2–4 In addition to the first generation PDE4 inhib-
itor rolipram, a number of second generation PDE4 inhibitors
(Fig. 1) have been reported with roflumilast (Daxas™, Daliresp™)
recently approved for severe chronic obstructive pulmonary dis-
ease (COPD).5–7 Despite significant progress in this area, PDE4
inhibitors are often associated with side effects such as nausea,
emesis, and vasculopathy, all of which limit the therapeutic use
of these agents.8 This highlights the need for novel pharmaco-
phores that would allow the design of safer PDE4 inhibitors. Here,
we describe the discovery of 2,3-disubstituted pyridines as a new
class of potent non-emetic PDE4 inhibitors and the evaluation of
identified compound 1 (Fig. 1).
In our search for safer PDE4 inhibitors, we identified compound
2 (Fig. 2) as a potent, orally available candidate.9 However, further
in vivo evaluation of this compound revealed emesis as side effect
in ferrets. To overcome this problem, we investigated ways to
improve compound 2 safety profile without affecting its PDE4
inhibitory activity. Based on the hypothesis that compound 2
safety concerns may be attributed to its high lipophilicity, which
leads to high brain exposure, we considered structural modifica-
tions that would allow reduction of compound 2 lipophilicity.
Our strategy for generation of lead non-emetic PDE4 inhibitors
is shown in Figure 2. Basically, we hypothesized that inhibition of
PDE4 in the central nerve system (CNS) causes emesis and that
reduced brain exposure by increasing hydrophilicity leads to
decrease the potency to induce emesis in ferrets.
Structure activity relationships (SARs) of the prepared
2,3-disubstituted pyridines are summarized in Table 1. First, we
changed the S-linker of the phenylthio moiety of compound 2 to
an O-linker (3), NH-linker (4), carbonyl linker (5), or a methylene
linker (6). Replacement of the S-linker by an O-linker or a methy-
lene linker had no effect on PDE4 inhibitory activity. On the other
hand, PDE4 inhibition decreased following use of an NH-linker or a
carbonyl linker. Second, we introduced a hydroxyl group at the 4-
pyridyl group. This is because it has been reported that metabolism
of roflumilast to form an N-oxide (Fig. 1) results in low brain pen-
etration and therefore reduced emetic effect.7 To avoid decrease in
⇑
Corresponding authors. Tel.: +81 6 6337 5903; fax: +81 6337 6047 (A.F.);
tel.: +81 6 6368 1773 (T.S.).
E-mail addresses: akira-fusano@ds-pharma.co.jp (A. Fusano), t-sumiyo@
kansai-u.ac.jp (T. Sumiyoshi).
Present address: Faculty of Chemistry, Materials and Bioengineering, Department
of Life Science and Biotechnology, Kansai University, 3-3-35 Yamate-cho, Suita, Osaka
564-8680, Japan.
http://dx.doi.org/10.1016/j.bmcl.2014.04.052
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

2690
M. Kawasaki et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2689–2692
F
F
N
NH
F
O Cl
HN
F
O Cl
HN
MeO
O
O
O
O
N+ O-
O
N
OH
O
N
O
O
Cl
Cl
Cl
Rolipram
Roflumilast
Roflumilast -Oxide
1
N
Figure 1. Chemical structures of selected PDE4 inhibitors.
2- (11) or 4-position (12) decreased PDE4 inhibitory activity. In
addition, introduction of two chloro groups at the 3- and 5-posi-
tions diminished PDE4 inhibitory activity (13).
The results of further in vitro evaluation of compound 1 are
summarized in Table 3. Compound 1 inhibited both human and
guinea pig PDE4 without inhibiting PDE3. In addition, compound
1 inhibited both antigen-induced immediate asthmatic response
(IAR) and late asthmatic response (LAR) in guinea pigs. Based on
these encouraging results, we selected compound 1 as a potent
non-emetic PDE4 inhibitor drug candidate.
Finally, we evaluated compound 1 potential for inducing vomit-
ing in ferrets. Oral administration of compound 1, given at up to
300 mg/kg, did not cause vomiting (Table 4). On the other hand,
both rolipram and compound 2 dose-dependently caused vomiting
in ferrets. These results indicate that modification of the lipophilic
moiety of compound 2 expanded its therapeutic margin.
To clarify the mechanism by which compound 1 avoids emesis,
we conducted a pharmacokinetic study of this compound in guinea
pigs. The results of this study revealed that compound 1 is imme-
diately metabolized to its sulfate 14 and glucuronide 15 (Fig. 3).
Investigation of compound 1 distribution after oral administration
of 14C-labelled compound 110 showed a lower concentration in the
brain than in the plasma, lung or trachea (Table 5). In addition, the
unchanged compound 1 was detected in a much lower concentra-
tion than its metabolites 14 and 15 in the plasma (Table 5). These
findings suggest that distribution of the unchanged compound 1
into the brain is limited. This distribution profile is believed to con-
tribute to significant reduction of emetic side effect. On the other
hand, the concentrations of compound 1 in the lung and trachea
were higher than its concentration in the plasma. These results
indicate that compound 1 is reformed by deconjugation of its glu-
curonide (metabolite 15) with b-glucuronidase in the lung and
N
N
O
S
O
O
Decrease lipophilicity
OH
N
N
Br
Cl
2
1
PDE4 IC50: 34 nM
PDE4 IC50: 16 nM
Bronchodilatory effect: 77%
Bronchodilatory effect: 70%
(10 mg/kg, p.o.)
(10 mg/kg, p.o.)
Vomitting: 3/4 (30 mg/kg, Ferret)
Vomitting: 0/4 (30 mg/kg, Ferret)
Figure 2. Strategy for reducing compound 2 lipophilicity.
the pyridyl group N-coordinating ability for binding to the metal
site of PDE4 enzyme, we introduced a hydroxyl group at the 3-
position of the pyridine ring, allowing the formation of pyridone
from 2-hydroxy pyridine. Evaluation of the synthesized com-
pounds 7, 8 and 9 revealed a maintained PDE4 inhibitory activity
in vitro, although the bronchodilating effect of compounds 7 and
9, with the S-linker and methylene linker, respectively decreased.
Compound 8, on the other hand, showed bronchodilating effect
with no vomiting in ferrets even at 300 mg/kg, po.
Next, we considered various substituents at the 3-position of
the phenyl ring (Table 2). Replacement of the bromo group in com-
pound 8 with a chloro (1) or a fluoro (10) group resulted in main-
tained PDE4 inhibitory activity. Among these compounds, oral
administration of compound 1 at 4 h before ovalbumin (OVA) chal-
lenge produced bronchodilating effect. Based on this long-lasting
bronchodilating effect, we selected compound 1 as a lead com-
pound and considered various substituents at the 2- and 4-posi-
tions of the phenoxy ring. Introduction of a chloro group at the
Table 1
SAR of the prepared 2,3-disubstituted pyridines
N
O
R1
N
X
Br
Compound
X
R1
PDE4 inhibition9 (%,
100 nM)
Bronchodilatory effect9 (%, 10 mg/kg,
po)
Cell infiltration inhibition9 (%, 10 mg/kg,
po)
Vomita (30 mg/kg,
po)
2
3
4
5
6
7
8
9
S
O
NH
CO
CH2
S
H
H
H
H
H
80
80
51
27
89
77
49
NT
NT
63
8
50
23
NT
NT
43
33
50
44
4/4
NT
NT
NT
NT
NT
0/4
NT
OH 66
OH 77
O
70
13
CH2 OH 87
NT: Not tested
a
Male Marshall ferrets were used in this experiment. The number of vomits for each animal was recorded for 2 h following oral administration of test compound. The
number of responding animals was also recorded.

M. Kawasaki et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2689–2692
2691
Table 2
SAR of substituents at the phenoxy moiety
N
O
OH
N
O
1
2
3
R2
4
Compound
R2
PDE4 inhibition (%, 100 nM)
Bronchodilatory effect (%, 10 mg/kg, po)
1 h
4 h
8
1
10
11
12
13
Rolipram
3-Br
3-Cl
3-F
2-Cl
4-Cl
3,5-Cl2
77
75
74
30
46
6
70
55
55
NT
NT
NT
50
12
45
9
NT
NT
NT
13
54
NT: Not tested.
Table 3
Table 5
In vitro evaluation of compound 1
Concentrations of compounds 1, 14 and 15 after oral administration of 14C-labelled
compound 1a
Compound 1
Concentration (ng equiv/mL or ng equiv/g)
In vitro
Human PDE4 IC50 (nM)
Guinea pig PDE4 IC50 (nM)
Bovine PDE3 IC50 (nM)
33
38
>30,000
0.5 h
2 h
4 h
Plasma
Radioactivity
1
14
15
387.6 52.0
1.7 1.7
43.8 12.7
130.4 2.0
452.8 40.3
ND
219.2 38.6
130.8 11.2
346.2 25.9
ND
194.4 11.9
79.3 13.0
In vivo
IAR inhibition11 (%)
LAR inhibition11 (%)
45
40
Lung
Radioactivity
1
14
15
251.2 30.2
21.0 4.5
29.2 5.2
46.5 1.5
232.9 19.4
12.8 3.5
78.9 13.6
59.5 10.8
178.7 16.9
7.7 1.8
79.9 6.6
33.8 7.1
Table 4
Emetic effect of rolipram, compounds 1 and 2 in ferrets
Compound
Vomit (Responders/Tested, po)
Trachea
Radioactivity
1
142.1 32.7
14.4
85.1 13.3
ND
90.1 2.5
ND
1 mg/kg
3 mg/kg
10 mg/kg
30 mg/kg
300 mg/kg
1
2
NT
NT
0/4
NT
NT
3/4
0/4
3/4
4/4
0/4
3/4
NT
0/4
NT
NT
14
15
17.2
53.6
34.4
33.7
44.0
28.1
Rolipram
Brain
Radioactivity
31.9 6.7
17.7 2.1
14.5 1.9
NT: Not tested.
ND: Not detected.
a
Male guinea pigs (SLC-Std, 7–8 weeks old) sensitized with OVA were used in
this experiment (N = 3). They were anesthetized and their tissues and blood were
collected 0.5, 2 and 4 h after oral administration of 14C-labelled compound 1
(10 mg/7.69 MBq/kg).
N
N
O
O
O
OH
OSO3H
O
OH
OH
N
O
N
O
N
CO2H
Cl
Cl
OH
S
ii
O
S
OH
Br
i
14
15
N
N
PDE4 IC50: 837 nM
PDE4 IC50: 11.5 µM
N
Figure 3. The metabolites of compound 1.
Br
Br
16
17
2
trachea. This reformation would contribute to a potent efficacy
in vivo, because the PDE4 IC50 value of compound 1 is much more
potent (much lower) than that of metabolite 14 or 15.
Scheme 1. Reagents and conditions: (i) 3-bromobenzenethiol, THF, DMF, reflux,
5 h, 78%; (ii) 3-(pyridin-4-yl)-1-propanol, PPh3, DIAD, THF, 0 °C to room temp, 97%.
Preparation of the compounds is shown in Schemes 1–3 and
S1–S8. Reaction of 3-bromobenzenethiol with the pyridine deriva-
tive 16 in THF under reflux condition provided the corresponding
thioether 17 in 78% yield. Mitsunobu reaction of compound 17
with an alcohol furnished the desired 2,3-disubstituted pyridine 2.
Reaction of the MOM-protected 3-hydroxy pyridine 18 using
n-butyllithium and DMF allowed introduction of a formyl group
at the C-4 position, and subsequent Horner–Wadsworth–Emmons
reaction with triethylphosphonoacetate gave the desired product

2692
M. Kawasaki et al. / Bioorg. Med. Chem. Lett. 24 (2014) 2689–2692
N
iii, iv
N
N
i, ii
EtO
HO
O
OMOM
OMOM
OMOM
18
19
20
N
OH
O
O
O
viii
OH
v
vi, vii
OH
N
O
N
N
Br
N
Br
21
22
23
1
Cl
Cl
Scheme 2. Reagents and conditions: (i) n-BuLi, THF, À30 °C to 0 °C, 20 min then DMF, 0 °C, 30 min; (ii) NaH, triethylphosphonoacetate, THF, 0 °C, 1 h, 71% for two steps; (iii)
5% Pd/C, H2, EtOH, room temp, 95%; (iv) LAH, THF, 50 °C to reflux, 1 h, 99%; (v) cyclopentanol, PPh3, DIAD, THF, 0 °C to room temp, 30 min, 73%; (vi) 3-chlorophenol, K2CO3,
CuBr, DMF, reflux, 1 h; (vii) 47% HBr aq, AcOH, reflux, 4 h, 83% for two steps; and (viii) compound 20, DIAD, PPh3, THF, 0–50 °C, 30 min, then 15% HCl aq, EtOH, reflux, 30 min,
52%.
pared to rolipram or other related compounds. PK study of com-
N
N
pound 1 revealed quick metabolism into a sulfate form (14) and
a glucuronide form (15) through first-pass effect, leading to mini-
mal brain penetration. In addition, the study showed that com-
pound 1 is reformed by deconjugation of its metabolite 15 with
b-glucuronidase in the lung and trachea, suggesting potent efficacy
in vivo. The findings of this study indicate that compound 1 has
great potential in the treatment of respiratory diseases, including
asthma and COPD.
O
O
O
O
i or ii
OAc
OH
OR
N
N
1
Cl
Cl
14
R = SO3Na:
OH
OH
OH
HO
OAc
OAc
=
O
O
Supplementary data
CO2Na
CO2Me
15
24
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.04.
052.
Scheme 3. Reagents and conditions: (i) SO3ÁNMe3, NaOH aq, 50 °C, 3 h, 63%; (ii)
compound 24, DIAD, PPh3, THF, 0 °C to room temp, 17 h, then NaOH aq, acetone,
0 °C to room temp, 4 h, 17%.
References and notes
1. Houslay, M. D.; Adams, D. R. Biochem. J. 2003, 370, 1.
19 (Scheme 2). After hydrogenation of the olefin moiety under
standard conditions, reduction of the resulting ethyl ester group
afforded the corresponding alcohol 20 in good yield. The bromo
pyridine 22, which was obtained by Mitsunobu reaction of 21 with
cyclopentanol, was converted to the phenol derivative 23 via cou-
pling reaction with 3-chlorophenol followed by dealkylation of the
cyclopentyloxy group. Mitsunobu reaction of 23 with 20 using
DIAD and PPh3 produced the corresponding coupling product,
which underwent deprotection of the MOM group to yield com-
pound 1. The synthetic methods of other compounds 3–13 are pro-
vided in Supplementary content (Schemes S1–S8).
Reaction of compound 1 with TMAS (trimethylamine SO3) or
compound 24 afforded the corresponding sulfate 14 or glucuronide
15 in 56% and 67% yield, respectively (Scheme 3).12
In summary, we modified the lipophilic moieties of the emetic
PDE4 inhibitor 2 and found that pyridines with a hydroxyl group
show significantly reduced emesis potential in ferrets. The selected
compound 1 showed a long-lasting bronchodilating effect com-
2. Houslay, M. D.; Schafer, P.; Zhang, K. Y. J. Drug Discovery Today 2005, 10, 1503.
3. Sanz, M. J.; Cortijo, J.; Morcillo, E. J. Pharmcol. Ther. 2005, 106, 269.
4. Kodimuthali, A.; Jabaris, S. S. L.; Pal, M. J. Med. Chem. 2008, 51, 5471.
5. Hatzelmann, A.; Schudt, C. J. Pharmacol. Exp. Ther. 2001, 297, 267.
6. Giembycz, M. A.; Field, S. K. Drug Des. Dev. Ther. 2010, 4, 147.
7. Hatzelmann, A.; Morcillo, E. J.; Lungarella, G.; Adnot, S.; Sanjar, S.; Beume, R.;
Schudt, C.; Tenor, H. Pulm. Pharmacol. Ther. 2010, 23, 235.
8. Chung, K. F. Eur. J. Pharmacol. 2006, 533, 110.
9. Kato, Y.; Kawasaki, M.; Nigo, T.; Nakamura, S.; Fusano, A.; Teranishi, Y.; Ito, N.
M.; Sumiyoshi, T. Bioorg. Med. Chem. 2013, 21, 5851.
10. Ochi, T.; Sawayama, Y.; Kawasaki, M.; Nigo, T.; Nakao, M.; Omodani, T.;
Sumiyoshi, T. submitted for publication.
11. Guinea pigs were actively sensitized with OVA adsorbed onto aluminum
hydroxide gel, and administered by intraperitoneal injection. Two weeks later,
they were exposed with aerosolized antigen 3 times at 3 day intervals. Seven
days after the last antigen exposure, the animals were pre-treated with
pyrilamine to avoid fatal anaphylactic shock, and were challenged with
aerosolized antigen. Biphasic airway responses (early and late responses) were
induced after antigen exposure. The effects of compound 1 on immediate and
late asthmatic responses were evaluated.
12. Pilgrim, W.; Murphy, P. V. J. Org. Chem. 2010, 75, 6747.