Substituted aminopyridines as potent and selective phosphodiesterase-4 inhibitors

Article abstract of DOI:10.1016/S0960-894X(02)01030-2

The synthesis and the biological evaluation of new potent phosphodiesterase type 4 (PDE4) inhibitors are presented. This new series was elaborated by replacement of the metabolically resistant phenyl hexafluorocarbinol of L-791,943 (1) by a substituted aminopyridine residue. The structure-activity relationship of N-substitution on 3 led to the identification of (-)-3n which exhibited a good PDE4 inhibitor activity (HWB-TNFα=0.12 μM) and an improved pharmacokinetic profile over L-791,943 (rat t_1/2=2 h). (-)-3n was well tolerated in ferret with an emetic threshold of 30 mg/kg (po) and was found to be active in the ovalbumin-induced bronchoconstriction model in guinea pig (54%, 0.1 mg/kg, ip) as well as the ascaris-induced bronchoconstriction model in sheep (64%/97%, early/late, 0.5 mg/kg, iv).

Full text of DOI:10.1016/S0960-894X(02)01030-2

Bioorganic & Medicinal Chemistry Letters 13 (2003) 741–744
Substituted Aminopyridines as Potent and Selective
Phosphodiesterase-4 Inhibitors
Bernard Cote,* Richard Frenette, Sylvie Prescott, Marc Blouin,
Christine Brideau, Yves Ducharme, Richard W. Friesen, France Laliberte,
Paul Masson, Angela Styhler and Yves Girard
´
Merck Frosst Centre for Therapeutic Research, PO Box 1005, Pointe-Claire-Dorval, Quebec, Canada H9R 4P8
Received 27 June 2002; accepted 29 October 2002
Abstract—The synthesis and the biological evaluation of new potent phosphodiesterase type 4 (PDE4) inhibitors are presented.
This new series was elaborated by replacement of the metabolically resistant phenyl hexafluorocarbinol of L-791,943 (1) by a sub-
stituted aminopyridine residue. The structure–activity relationship of N-substitution on 3 led to the identification of (À)-3n which
exhibited a good PDE4 inhibitor activity (HWB-TNFa=0.12 mM) and an improved pharmacokinetic profile over L-791,943 (rat t1/2
=2 h). (À)-3n was well tolerated in ferret with an emetic threshold of 30 mg/kg (po) and was found to be active in the ovalbumin-
induced bronchoconstriction model in guinea pig (54%, 0.1 mg/kg, ip) as well as the ascaris-induced bronchoconstriction model in
sheep (64%/97%, early/late, 0.5 mg/kg, iv).
# 2003 Elsevier Science Ltd. All rights reserved.
Over the last few years, leukotriene modulators such as
leukotriene D4-receptor antagonist1 and 5-lipoxygenase
inhibitors2 have been one of the main antiasthmatic
targets of the pharmaceutical industry. More recently,
regulation of the intracellular level of cyclic nucleotides
by inhibition of cAMP phosphodiesterases (PDEs) has
attracted increased interest for the treatment of inflam-
matory diseases and asthma. It has been shown that
cAMP elevation mediates airway smooth muscle
relaxation3 and also suppresses the activation of
inflammatory cells.4 It was then rapidly recognized that
inhibition of the cAMP-specific isozyme phosphodi-
esterase type 4 (PDE4) would represent a good asthma
target with dual properties of bronchodilation and anti-
inflammatory activity.5,6 However, the strong potential
of the PDE4 inhibitors was found to be limited by a
narrow therapeutic index mainly associated with nausea
and emesis.7 These side effects were first observed with
the archetypic PDE4 inhibitor rolipram and since then,
a major effort was invested to understand this phenomena
and improve the so-called emetic window.
Figure 1.
An excessively long half-life (Rat: >24 h) had precluded
the development of this compound but the introduction
of a soft metabolic site on the structure of L-791,943 to
produce L-826,141 (2) resulted in improved pharmaco-
kinetics.9 To overcome this long half-life problem we
also considered replacing the metabolically resistant
phenyl hexafluorocarbinol by a substituted amino-
pyridine residue. Here we wish to report the synthesis
Recently, we reported the discovery of an orally active
and well tolerated PDE4 inhibitor, L-791,943 (1) (Fig. 1).8
*Corresponding author. Fax: +1-5124-428-8670; e-mail: bernard_
cote@merck.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(02)01030-2

´
B. Coˆte et al. / Bioorg. Med. Chem. Lett. 13 (2003) 741–744
742
and the biological properties of the PDE4 inhibitor
aminopyridines 3.
Finally, the amine was introduced by a Cu (I) promoted
aryl amination.13 For this transformation only small
amines could be used. In fact, more sterically hindered
amines slow down the coupling reaction and competitive
reduction of the N-oxide is observed. It is then recom-
mended to do the Cu coupling before the N-oxidation of
the 4-pyridyl residue as shown in Scheme 2. In this
sequence, secondary amines are coupled in the presence
Over 100 analogues of type 3 were prepared to investi-
gate the SAR of N-substitution with R1 and R2.10 Three
different strategies were used to synthesize compounds
3a to 3p (Table 1). In general, the steric hindrance of the
amine involved in the Cu (I) promoted aryl amination
dictated the appropriate route. The first general synth-
esis, presented in Scheme 1, was applied to the prepara-
tion of non-hindered secondary and tertiary amines 3.11
Secondary alcohol 5 was obtained by regioselective
metalation of 2,5-dibromopyridine and nucleophilic
addition on bis(difluoromethoxy)benzaldehyde 4.12 The
potassium enolate of ethyl 4-pyridylacetate was then
alkylated with the chloride 6. After hydrolysis of ester 7,
the corresponding acid spontaneously decarboxylated
to afford the pyridine 8 which could then be oxidized to
the 4-pyridyl-N-oxide with either MMPP or MCPBA.
Table 1. SAR for R1 and R2 substituents on 3
3
R1
R2
GST-PDE4A248
IC50 (nM)a
HWB (TNF-a)
IC50 (mM)b
3a
3b
3c
3d
3e
3f
Me
Et
i-Pr
t-Bu
Me
Me
CH2Ph
CH2Ph
CH2Ph
CH2Ph
CH2CH2Ph
Ph
1.2
2.3
9.3
18
5.8
0.8
0.3
0.4
1.0
17
0.8
2.1
31
3g
3h
—
1.5
0.3
Me
Me
3i
3j
0.3
0.6
0.7
0.2
Scheme 1. (a) BuLi, Et2O, À78 ꢀC, 53%; (b) SOCl2, CH2Cl2;
(c) KHMDS, HMPA, ethyl 4-pyridylacetate, THF; (d) LiOH, THF,
MeOH, 65 ꢀC; HCl, 97% (three steps); (e) MMPP, CH2Cl2, MeOH,
93%; (f) CuI, 100–160 ꢀC.
Et
3k
3l
H
H
CH2Ph
CH2CH2PhCH3
1.2
2.3
0.4
0.8
3m
3n
H
H
1.7
1.6
0.9
0.2
3oc
3p
3qd
3rd
3sd
3te
H
H
H
H
H
H
H
i-Pr
SO2Ph
CONHEt
COCH3
CO2CH2Ph
31
12
16
8.3
33
7.8
4.2
1.3
38
1.6
0.8
34
1.7
1.8
3.8
0.7
0.3
18
L-791,943 (1)
L-826-141 (2)
SB207499 (ArifloTM
)
aAssayed against human PDE4A isoform using construct representing
the common region of spliced variants expressed as GST-fusion pro-
tein in Sf9 cells.15 IC50 represent a mean of N=3.
bInhibition of LPS-induced TNF-a in human whole blood.16 IC50
represent a mean of N=3.
cPrepared from 3n with TFA/CH2Cl2.
dFor the synthesis of these analogues see ref 17.
ePrepared by treating 16 with TFA/CH2Cl2.
Scheme 2. (a) CuI, >100 ꢀC; (b) (RCO)2O, pyridine, CH2Cl2;
(c) MMPP, CH2Cl2, MeOH; (d) LiOH, THF, MeOH.

´
B. Coˆte et al. / Bioorg. Med. Chem. Lett. 13 (2003) 741–744
743
of CuI followed by MMPP oxidation, but primary
amines required an additional protection step to avoid
oxidation of the aminopyridine functionality. Oxidation
of the 4-pyridine 11 and amide hydrolysis afforded the
aminopyridine 3.14
Summarized in Table 1 are the intrinsic PDE4A potency
and their cellular efficacy in human whole blood (HWB)
using the inhibition of TNF-a production as an index.
Although these compounds represent only a small frac-
tion of all the aminopyridines tested, they illustrate very
well the SAR that was observed while replacing R1 and
R2. In the tertiary amine series (3a to 3j), the data sug-
gested that R1 has to be kept small since the inhibitory
activity declined progressively going from 3a (R1=CH3)
to 3d (R1=t-butyl). With a small R1 in place (R1=Me),
several R2 replacements were compared with the benzyl
aminopyridine 3a. Selected examples 3e to 3h led to the
conclusion that the benzylic motif was a superior phar-
macophore at this position since phenethyl 3e, aniline 3f
and pyrrolidine 3g were found to be less potent. As we
had a good hand on potency with prototypic amino-
pyridine 3a, the next series of analogues were then
designed based on pharmacokinetic data. Oral and iv
administration of 3a to rats and squirrel monkeys, indi-
cated that the drug was rapidly metabolized (t1/2<1 h)
by dealkylation of the methyl substituent to afford the
secondary amine 3k. This in vivo metabolic pathway
was also observed with analogue 3e leading to the for-
mation of 3l as the major metabolite. Interestingly, 3k
and 3l were found to maintain a good PDE4 activity but
unfortunately, when dosed iv in rats, no pharmaco-
kinetic improvement was observed and these two com-
pounds were cleared rapidly with a half-life of less than
one h. Another interesting observation was made when
the most potent compound of the series, 3j (HWB 0.15
mM), was dosed iv in rats. In this case, the ethyl residue
remains untouched but the amine was rapidly debenzy-
lated resulting in a t1/2 comparable to 3a. Its was pos-
tulated that in vivo oxidation of the benzylic position
followed by hemiacetal fragmentation could be respon-
sible for the observed debenzylation. To prevent this
metabolic pathway the benzylic position was blocked
with a gem dimethyl while keeping R1=H. This new
potent analogue 3n displayed a half-life of 2–3 h in rats
and squirrel monkeys which represented a major
improvement over the previous aminopyridines. Several
benzyl replacements were also investigated but in gen-
eral, as shown with analogue 3o to 3t, a reduced PDE4
activity was observed.
The synthesis of aminopyridine 3n is presented in
Scheme 3. Compounds 3f and 3i were also prepared
following this sequence which was found to be more
appropriate for bulky amines. In a first attempt to syn-
thesize 3n by coupling cumylamine 13 with the bromo-
pyridine 8, no reaction was observed and we had to
increase the electrophilic character of the bromo-
pyridine functionality. This was achieved by oxidizing
the alcohol 5 to the ketone 12 in the presence of MnO2.
Cumylamine 13 was then successfully coupled to this
activated bromopyridine with copper iodide in 89%
yield. The resulting secondary amine 14 was then pro-
tected to avoid competitive hydroxylamine formation
during the subsequent oxidation of the 4-pyridyl resi-
due. Ketone reduction followed by treatment of the
alcohol 15 with thionyl chloride led to the corresponding
chloride which was used to alkylate the potassium eno-
late of ethyl 4-pyridylacetate. Aminopyridine 16 was
then obtained by ester hydrolysis and decarboxylation
of the corresponding carboxylic acid followed by oxi-
dation of the 4-pyridyl residue. Final deprotection of
the benzyl carbamate under hydrogenolysis conditions
afforded the analogue 3n.
The enantiomers of 3n were then resolved using a Chiral-
pack AD HPLC column (Table 2). (À)-3n was found to
be the most potent enantiomer exhibiting, in the human
whole blood assay, improved PDE4 inhibitory potency
over L-791,943 (5.6-fold), L-826,141 (3-fold) and
SB207499 (ArifloTM 18
)
(150-fold). (À)-3n was also
found to be very selective against all the other PDEs.19
Table 3 highlights some of the in vivo characteristics of
(À)-3n in comparison with L-791,943 (1). One of our
Table 2. Resolution of aminopyridine 3n
Compd
GST-PDE4A
IC50 (nM)
HWB (TNF-a)
IC50 (mM)
Scheme 3. (a) MnO2, CH2Cl2, 88%; (b) CuI, 140 ꢀC, 89%; (c) benzyl
chloroformate, i-Pr2NEt, dioxane; (d) NaBH4, THF, MeOH, 95%
(two steps); (e) SOCl2, pyridine, toluene, 0 ꢀC; KHMDS, HMPA, ethyl
4-pyridylacetate, THF, 0 ꢀC; (f) LiOH, THF, MeOH, 65 ꢀC; HCl, 80%
(two steps); (g) MMPP, CH2Cl2, MeOH; (h) H2, 10% Pd/C, EtOH,
55% (two steps).
3n
(+)-3n
(À)-3n
1.6
23
0.8
0.2
10
0.1

´
B. Coˆte et al. / Bioorg. Med. Chem. Lett. 13 (2003) 741–744
744
Table 3. Comparative in vivo profile of (À)-3n with L-791,943
I. W.; Sawyer, N.; Young, R. N.; Zamboni, R.; Abraham,
W. M. Can. J. Physiol. Pharmacol. 1997, 73, 747 (erratum).
2. Young, R. N. Eur. J. Med. Chem. 1999, 34, 671.
3. Torphy, T. J. Agents Actions 1988, 23, S37.
L-791,943
(À)-3n
2 h
Rat: t1/2
>24 h
4. Kuehl, F. A.; Zanetti, M. E.; Soderman, D. D.; Miller,
D. K.; Ham, E. A. Am. Rev. Respir. Dis. 1987, 136, 210.
5. Torphy, T. J. Am. J. Respir. Crit. Care Med. 1998, 157, 351.
6. Souness, J. E.; Aldous, D.; Sargent, C. Immunopharma-
cology 2000, 47, 127.
7. (a) Robichaud, A.; Tattersall, F. D.; Choudhury, I.; Rod-
ger, I. W. Neuropharmacology 1999, 38, 289. (b) Robichaud,
A.; Savoie, C.; Stamaciou, P. B.; Tattersall, F. D.; Chan, C. C.
Neuropharmacology 2001, 40, 262 and 465 (erratum).
8. (a) Huang, H.; Ducharme, Y.; MacDonald, D.; Robichaud,
A. Curr. Opin. Chem. Biol. 2001, 5, 432. (b) Guay, D.; Hamel,
P.; Blouin, M.; Brideau, C.; Chan, C. C.; Chauret, N.; Duch-
arme, Y.; Huang, Z.; Girard, M.; Jones, T. R.; Laliberte, F.;
Li, C.; Masson, P.; McAuliffe, M.; Piechuta, H.; Silva, J.;
Young, R. N.; Girard, Y. Bioorg. Med. Chem. Lett. 2002, 12,
1457.
9. Frenette, R.; Blouin, M.; Brideau, C.; Chauret, N.; Duch-
arme, Y.; Friesen, R. W.; Hamel, P.; Jones, T. R.; Laliberte,
F.; Li, C.; Masson, P.; McAuliffe, M.; Girard, Y. Bioorg. Med.
Chem. Lett. 2002, 12, 3009.
10. Cote, B.; Friesen, R. W.; Frenette, R.; Girard, M.; Girard,
Y.; Godbout, C.; Guay, D.; Hamel, P.; Blouin, M.; Duch-
arme, Y.; Prescott, S. US 6,200,993 B1, 2001.
F, Cmax (P.O. 20 mpk)
70%, 3.5 mM 100%, 2.4 mM
58%a
54%b
(
Guinea Pig
(1.0 mpk)
85%/95%
(2.0 mpk)
(0.1 mpk)
64%/97%
(0.5 (mpk)
Efficiacy
Sheep À early=late:c
Emesis in ferret (CMax
)
>30 mpk
(14 mM)
30 mpk
(16 mM)
aDosed ip (4 h of pre-treatment).
bDosed ip (30 min of pre-treatment).
cDosed iv for 4 days (2 h of pre-treatment).
initial objectives was to reduce the long t1/2 observed
with L-791,943. With a half-life of 2 h (rat, iv 5 mg/kg),
while maintaining the PDE4 activity, the aminopyridine
pharmacophore proved to be a good replacement for
the phenyl hexafluorocarbinol. Ferrets receiving (À)-3n
showed signs of emesis only at a po dose of 30 mg/kg.
Considering its potency, this emetic threshold compares
very well with L-791,943 (1). Finally, inhibitor (À)-3n
demonstrated good in vivo activity by blocking oval-
bumin-induced bronchoconstriction in conscious guinea
pig20 by 54% at a dose of 0.1 mg/kg (ip). (À)-3n was
also found to be efficacious in the sheep model21 with
64%(early phase) and 97% (late phase) of inhibition of
ascaris-induced bronchoconstriction.
11. This sequence was used to synthesize 3a, 3b, 3c, 3d, 3e, 3g,
3h, 3l and 3p.
12. Guay, D.; Girard, Y.; Ducharme, Y.; Blouin, M.; Hamel,
P.; Girard, M. US 5,710,170, 1998.
13. (a) Vorbruggen, H. Advances in Amination of Nitrogen
Heterocycles; Academic: San Diego, 1990; Vol. 49. (b) Lind-
ley, J. Tetrahedron 1984, 40, 1433. (c) Wagaw, S.; Buchwald,
S. L. J. Org. Chem. 1996, 61, 7240.
14. This sequence was used to synthesize 3k and 3m.
15. Laliberte, F.; Han, Y.; Govindarajan, A.; Giroux, A.; Liu,
S.; Bobechko, B.; Lario, P.; Bartlett, A.; Gorseth, E.; Gresser,
M.; Huang, Z. Biochemistry 2000, 39, 6449.
16. Brideau, C.; Van Staden, C.; Sthyler, A.; Rodger, I. W.;
Chan, C.-C. Br. J. Pharmacol. 1999, 126, 979.
17. These analogues were synthesized by treating 3o with the
following reagents: 3q: PhSO2Cl, pyridine, CH2Cl2 (30%), 3r:
ethyl isocyanate, CH2Cl2 (58%), 3s: acetic anhydride, pyri-
dine, CH2Cl2 (60%).
18. (a) Torphy, T. J.; Barnette, M. S.; Underwood, D. C.;
Griswold, D. E.; Christensen, S. B.; Murdoch, R. D.; Nieman,
R. B.; Compton, C. H. Pulm. Pharmacol. Ther. 1999, 12, 131.
(b) Compton, C. H.; Gubb, J.; Nieman, R.; Edelson, J.; Amit,
O.; Bakst, A.; Ayres, J. G.; Creemers, J. P. H. M.; Schultze-
Werninghaus, G.; Brambilla, C.; Barnes, N. C. Lancet 2001,
358, 265.
In conclusion, we have developed a new series of potent
PDE4 inhibitors by replacement of the metabolically
resistant phenyl hexafluorocarbinol of L-791,943 (1) by
a substituted aminopyridine residue. The structure–
activity relationship of N-substitution on 3 led to the
identification of (À)-3n which exhibited a good PDE4
inhibitor activity (HWB-TNF-a=0.12 mM) and an
improved pharmacokinetic over L-791,943 (Rat t1/2=2 h).
(À)-3n was well tolerated in ferret with an emetic
threshold of 30 mg/kg (po) and was found to be active
in the ovalbumin-induced bronchoconstriction model in
guinea pig (54%, 0.1 mg/kg, ip) as well as the ascaris-
induced bronchoconstriction model in sheep (64%/
97%, early/late, 0.5 mg/kg, iv).
19. Selectivity >1000 over PDE1, PDE3A, PDE3B, PDE5A,
PDE6, PDE7-A2, PDE9A.
References and Notes
20. Pennock, B. C.; Cox, C. P.; Rogers, R. M.; Cain, W. A.;
Wells, J. H. J. Appl. Physiol. 1979, 46, 399.
21. Abraham, W. M.; Ahmed, A.; Cortes, A.; Sielczak, M. W.;
Hinz, W.; Bouska, J.; Lanni, C.; Bell, R. L. Eur. J. Pharmacol.
1992, 217, 119.
1. Jones, T. R.; Labelle, M.; Belley, M.; Champion, E.; Char-
ette, L.; Evans, J.; Ford-Hutchinson, A. W.; Gauthier, J.-Y.;
Lord, A.; Masson, P.; McAuliffe, M.; McFarlane, C. S.; Met-
ters, K. M.; Pickett, C.; Piechuta, H.; Rochette, C.; Rodger,