Muscarinic M3 receptor antagonists with (2R)-2-[(1R)-3,3-Difluorocyclopentyl]-2-hydroxyphenylacetamide Structures. Part 2

Article abstract of DOI:10.1016/S0960-894X(03)00350-0

Optimization of the amine part of our original muscarinic M₃ receptor antagonist 1 was performed to identify M₃ receptor antagonists that are superior to 1. Compounds carrying a variety of diamine moieties without hydrophobic substituent on the nitrogen atom were screened against the binding affinity for the M₃ receptor and the selectivity for M₃ over the M₁ and M₂ receptors. This process led to a 4-aminopiperidinamide (2l) with a K_i value of 5.1 nM and with a selectivity of the M₃ receptor that was 46-fold greater than that of the M₂ receptor. Further derivatization of 2l by inserting a spacer group or by incorporating alkyl group(s) into the amine part resulted in the identification of an 4-(aminoethyl)piperidinamide 2l-b with a K_i value of 3.7 nM for the M₃ receptor and a selectivity for the M₃ receptor that was 170-fold greater than that of the M₂ receptor.

Full text of DOI:10.1016/S0960-894X(03)00350-0

Bioorganic & Medicinal Chemistry Letters 13 (2003) 2167–2172
Muscarinic M3 Receptor Antagonists with
(2R)-2-[(1R)-3,3-Difluorocyclopentyl]-2-hydroxyphenylacetamide
Structures. Part 2
Yoshio Ogino, Norikazu Ohtake,* Kensuke Kobayashi, Toshifumi Kimura,
Toru Fujikawa, Takuro Hasegawa, Kazuhito Noguchi and Toshiaki Mase
Banyu Tsukuba Research Institute in collaboration with Merck Research Laboratories, Okubo-3,
Tsukuba 300-2611, Ibaraki, Japan
Received 3 March 2003; accepted 2 April 2003
Abstract—Optimization of the amine part of our original muscarinic M3 receptor antagonist 1 was performed to identify M3
receptor antagonists that are superior to 1. Compounds carrying a variety of diamine moieties without hydrophobic substituent on
the nitrogen atom were screened against the binding affinity for the M3 receptor and the selectivity for M3 over the M1 and M2
receptors. This process led to a 4-aminopiperidinamide (2l) with a Ki value of 5.1 nM and with a selectivity of the M3 receptor that
was 46-fold greater than that of the M2 receptor. Further derivatization of 2l by inserting a spacer group or by incorporating alkyl
group(s) into the amine part resulted in the identification of an 4-(aminoethyl)piperidinamide 2l–b with a Ki value of 3.7 nM for the
M3 receptor and a selectivity for the M3 receptor that was 170-fold greater than that of the M2 receptor.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
receptor antagonists. As a result, we identified an
1-(6-aminopyridyn-2-ylmethyl)piperidinamide (1) that
possesses high binding affinity (Ki=2.8 nM) for the M3
receptor and has selectivity for the M3 receptor that is
190-fold greater than that of the M2 receptor in the class
Pharmaceutical research into therapeutic agents that are
selective for muscarinic receptor subtypes has focused
on exploration of orally-active, muscarinic M3 receptor-
selective antagonists for the treatment of respiratory
disorders such as chronic obstructive pulmonary disease
(COPD) and urinary tract disorders such as urinary
incontinence (UI). Some muscarinic M3 receptor
antagonists have been identified and are under clinical
development.1À6 Among them, an inhaled M3 antagonist,
tiotropium, has been developed for the treatment of
COPD. Tiotropium was non-selective toward the mus-
carinic receptor subtypes in the binding assay, however,
it showed kinetic selectivity toward the M3 receptor
over the M2 receptor (dissociation from the M2 receptor
is faster than from the M3 receptor).7 Based on a
hypothesis that M2-sparing M3 receptor antagonists in
the binding basis would have clinical benefit over non-
selective muscarinic antagonists or kinetically selective
antagonists, we focused our search on this type of M3
of
(2R)-2-[(1R)-3,3-difluorocyclopentyl]-2-hydroxy-2-
phenylacetamides. The in vitro metabolism study of 1
indicated that this compound was easily metabolized to
an N-dealkylated one (2a) in rats. The binding data of
2a (Ki=320 nM for the M3 receptor, M2/M3=47-fold)
suggested that the piperidinyl side chain was essential
for the compound’s potency for the M3 receptor and for
its selectivity for the M3 receptor over the M2 receptor.
We considered that new antagonists lacking the amino-
pyridylmethyl moiety would be more metabolically
stable and would lead to more potent and longer-acting
M3 antagonists. Therefore, we focused on optimization
of the amine part of 1. As a result of extensive derivati-
zation, we have found a 4-(2-aminoethyl)piperidinamide
(2l–b) that showed better in vitro metabolic stability in
the rat, dog, and human hepatic microsomes, that had
potent binding affinity for the M3 receptor, and that had
a higher selectivity (170-fold) for the M3 receptor than
for the M2 receptor. In this paper, we describe the
synthesis and structure–activity relationships (SARs)
*Corresponding author. Tel.: +81-298-77-2221; fax: +81-298-77-
2029; e-mail: ootakenr@banyu.co.uk
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00350-0

2168
Y. Ogino et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2167–2172
produce an alcohol 5 in a 85% yield. The primary alcohol
moiety of 5 was converted to an azide 6 via substitution
of the corresponding mesylate with NaN3. Coupling of
6 with a 2-(3,3-difluorocyclopentyl)-2-hydroxy-2-phenyl
acetic acid 7 under a standard condition (WSC, HOBT)
followed by hydrogenation of the azide group produced
2c in 21% yield.
In order to prepare compound 2d, methyl p-amino-
benzoate was coupled with the corresponding acid
chloride of 7 in a 51% yield. The ester moiety of the
coupling product was reduced with LAH to yield an
alcohol, which was converted to an azide 9 in a 57%
yield in a manner similar to that described above. The
azide group was reduced with PPh3 in THF–H2O yiel-
ded 2d in a quantitative yield.
Figure 1.
of (2R)-2-[(1R)-3,3-difluorocyclopentyl]-2-hydroxy-2-
phenylacetamides that contained a variety of diamine
moieties (Fig. 1).
Compound 2k was synthesized from N-(2-hydroxy-
ethyl)pyperazine 10. After the coupling of 10 with 7
(47% yield), the primary alcohol of the resulting com-
pound 11 was converted to the amine in a similar man-
ner to that described for the preparation of 2d to
produce 2k in 87% yield.
Chemistry
Synthetic methods of representative compounds (2c, 2d,
2k, 2l–c, 2l–g and 2l–i) are summarized in Schemes 1
and 2. First, the synthesis of compound 2c was started
with the preparation of its amine part. Treatment of
N-Boc-piperidone 3 with Horner–Emmonds reagent
gave 4, in which the double bond was hydrogenated and
subsequently the ester moiety was reduced with LAH to
Scheme 2 shows the synthesis of the representative
piperidinamides (2l–c, 2l–g, 2l–i). The Boc group of the
Scheme 1. Synthesis of the representative piperidinamides (2c, 2d, and 2k). Reagents and conditions: (a) 60% NaH, (EtO)2CH2CO2Et, THF, 0 ꢀC;
(b) 10% Pd/C, H2, MeOH, rt; (c) LAH, THF, rt; (d) MsCl, NEt3, THF, 0 ꢀC; (e) NaN3, DMF, rt; (f) 7, WSC, HOBt, DMF, rt; (g) TFA, rt;
(h) SOCl2, DMA, À10 ꢀC; (i) PPh3, THF–H2O, rt.

Y. Ogino et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2167–2172
2169
Scheme 2. Synthesis of the representative piperidinamides (2l–c, 2l–g, and 2l–i). Reagents and conditions: (a) TFA, CHCl3, rt; (b) 7, WSC, HOBt,
CHCl3, rt; (c) 10% Pd/C, H2, MeOH, rt; (d) 60% NaH, (EtO)2CH2CO2Et, THF, 0 ꢀC; (e) LDA, Mel, THF, À78 ꢀC; (f) NaOH, MeOH, rt;
(g) (PhO)2PON3, BnOH, THF, 80 ꢀC; (h) LDA, EtBr, THF, À78 ꢀC; (i) LAH, THF, rt; (j) SO3Py, DMSO, Net3, rt; (k) PPh3CH3Br, n-BuLi,
THF, À78 ꢀC; (l) BH3, THF, 0 ꢀC, then 30% H2O2; (m) MsCl, NEt3, THF, 0 ꢀC; (n) NaN3, DMF, rt.
piperidines (6, 15, 20) with a protected primary amine
or an azide group (a precursor of an amine) was depro-
tected by treatment with TFA to produce the corre-
sponding 1-(non-protected)-piperidines, which were
coupled with the acid 7 to yield in 50–70% yield.
Finally, the benzyloxycarbonyl group or the azide of the
coupling products were reduced by a catalytic hydro-
genation to produce the target compounds in a good
yield.
with that of 2a. However, an 4-aminoethylpiperidine 2c
showed only a 4-fold improvement (Ki=89 nM) in the
M3 binding affinity. These compounds retained the
selectivity for M3 over the M2 receptor to some extent,
which was probably due to the 2-(3,3-difluoro-
cyclopentyl)-2-hydroxyphenylacetic acid moiety.6
Compounds with a substituted aniline (2d, 2e, 2f) or a
cyclohexylamine (2g) showed low binding affinity (Table
2). Based on our previous finding that an amide hydro-
gen in the 2-cyclobutyl-2-hydroxyphenylacetamide
enhanced the compound’s binding affinity for the M3
receptor and increased the compound’s selectivity for
the M3 receptor over that for the M2 receptor,8 both
piperazinamide 2h and homopiprazinamide 2i were
predicted to show low M3 binding affinity and M3
selectivity over the M2 receptor subtype. However, 2i
showed moderate M3 affinity (Ki=17 nM) and M3
selectivity (66-fold) over the M2 receptor. Furthermore,
a 4-aminoethylpiperazinamide 2k showed a nano-molar
order M3 binding affinity (Ki=9.6 nM) with 85-fold M3
selectivity over the M2 receptor. We assumed that the
primary amine moiety of 2k played a key role in
enhancing both the M3 affinity and selectivity, while the
tertiary amine moiety of the piperazine group did not
Results and Discussion
Compounds were tested in initial screens for binding
affinity (Ki values) against human muscarinic receptor
subtypes (hM1, hM2 and hM3) in transfected CHO
cells.8,9 Subsequently, selected compounds were exam-
ined for their in vitro metabolic stability in rat, dog, and
human hepatic microsomes.6
As described above, the 4-piperidinamide 2a lost much
of its binding affinity (Ki=320 nM) for the M3 receptor
(Table 1). In contrast, the 4-aminomethylpiperidin-
amide 2b showed a 18-fold improvement in the binding
affinity (Ki=17 nM) for the M3 receptor as compared

2170
Y. Ogino et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2167–2172
Table 1. The binding affinity of the compounds (2a–2m) to the mus-
carinic receptor subtypes
Table 2. The binding affinity of the compounds (2l-a–1l-l) to the
muscarinic receptor subtypes
a
a
Compd
RBinding affinity
Selectivity
M2/M3
47
Compd
RBinding affinity
Selectivity
M2/M3
69
(Ki, nM)
(Ki, nM)
M3
M1
M2
M3
11
M1
M2
320
310
15,000
2a
11
760
2l-a
2b
2c
2d
17
89
23
140
740
2200
3900
5300
130
43
8
2l-b
2l-c
2l-d
2l-e
2l-f
2l-g
2l-h
2l-i
3.7
6.6
630
2800
4200
55
170
45
56
46
45
16
15
77
65
43
39
47
76
92
650
140
1.2
1.3
2e
93
210
2900
31
120
930
140
5400
2f
290
100
630
55
15,000
6800
51
68
760 15,000
2g
2400 1900 36,000
2h
2i
50
17
16
30
14
21
1200
1100
300
24
66
18
85
46
44
8.3
3.3
7.0
4.7
1.5
1.0
640
210
2l-j
2j
2l-k
2l-l
300
2k
2l
9.6 11
820
26
37
1000
5.1
3.0
2.1
240
aValues are the mean of two or more independent assays.
2m
2.9
130
affinity than 2l did; however, its selectivity was compar-
able. Therefore, further derivatization on the amine part
of the piperidinamide 2l was prioritized.
aValues are the mean of two or more independent assays.
contribute to either the binding affinity or the selectiv-
ity. Based on this assumption, we prepared a 4-amino-
piperidinamide 2l and assessed its binding affinity. As
expected, 2l showed a potent M3 binding affinity
(Ki=5.1 nM) and moderate M3 selectivity (M2/M3=46-
fold). A pyrolidinamide 2m exhibited higher M3 binding
With respect to optimization of M3 affinity and selec-
tivity, a methylene (2l–a), an ethylene (2l–b), a propyl-
ene (2l–c), or a butylene (2l–d) spacer was inserted
between the piperidine and primary amine moiety of 2l.
Compound 2l–b with an ethylene spacer showed the
most potent M3 binding affinity (Ki=3.7 nM) and

Y. Ogino et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2167–2172
2171
Table 3. The binding affinity to the muscarinic receptor subtypes and metabolic stability of the representative compounds
a
Compd
X
H
RBinding affinity
(Ki nM)
Selectivity
M2/M3
Metabolic stability %b
(Rat, dog, human)
M3
1.4
M1
M2
11
0.98
7.9
40, 89, 92
23
2l
F
H
F
5.1
1.7
11
3.0
1.8
240
28
46
16
100, 88, 86
72, 100, 91
98, 100, 100
77, 84, 86
24
2l-a
25
11
760
41
69
H
F
1.1
3.7
2.1
6.6
36
2l-b
630
170
100, 100, 82
1
2.8
1.5
530
190
46, 84, 80
aValues are in the mean of two or more independent assays.
b% remaining after 30-min incubation in each hepatic microsomes.
selectivity (M2/M3=170-fold), while compound 2l–d
(Ki=76 nM) with a butylene spacer showed an
approximately 15-fold reduction in the M3 binding affi-
nity as compared to 2l. Replacement of the ethylene
spacer with an ethylidene one (2l–e) resulted in an
increase in M3 binding affinity (Ki=1.2 nM), while the
M3 selectivity of 2l–d significantly decreased (M2/
M3=46-fold). It is interesting to note that compound
2l–b showed high M3 binding affinity and selectivity
since the piperidine nitrogen of the amine part was acy-
lated with the acid 7. In contrast, compound 2c, in
which the primary amine moiety of the same amine part
was connected with the acid 7, resulted in a large
reduction in M3 binding affinity and selectivity in com-
parison with that of 2l–b.
group(s) into the a-carbon (2l–l) or the g-carbon (2l–j
and 2l–k) to the primary amine group in 2l–b resulted in
a decrease in M3 selectivity over the M2 receptor.
The effects of the 3,3-difluorocyclopentyl moiety of the
representative compounds (2l, 2l–a, and 2l–b) on the M3
binding affinity and selectivity, and in vitro metabolic
stability were examined (Table 3). Incorporation of the
3,3-difluorocyclopentyl group resulted in a 3- to 6-fold
decrease in the M3 binding affinity, but a 4- to 6-fold
improvement in the selectivity for the M3 receptor over
that of the M2 receptor. In particular, the 3,3-difluoro-
cyclopentyl moiety contributed to an enhancement in
the in vitro metabolic stability especially against rat
hepatic microsomes.
The effects of the substituents on the nitrogen of the
primary amine moiety in 2l–b on the M3 affinity and
selectivity were investigated. Replacement of the
N-methylamono (2l–f), N,N-dimethylamino (2l–g) or
pentamethylene imine (2l–h) group with the primary
amine of 2l–b significantly reduced the M3 binding affinity
and M3 selectivity over the M2 subtype. In particular, M3
binding affinity decreased in 2l–h, which suggests that the
bulkiness around the primary amine moiety affects the
binding interaction between the antagonist molecule and
M3 receptor subtype. Also, incorporation of alkyl
With respect to the pharmacokinetic profiles of 2l–b in
dogs (Table 4), the oral bioavailablity (F) of 2l–b was
Table 4. Pharmacokinetic parameters of 2l–b after iv (0.1 mg/kg) and
po (0.3 mg/kg) administration to dogs
iv
po
AUC (ng h/mL)
CLtot (mL/min/kg)
Vdss (mL/kg)
150
13
9400
AUC (ng h/mL)
F (%)
Cmax (ng/mL)
t1/2 (h)
440
98
43
6.9

2172
Y. Ogino et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2167–2172
approximately 100%. The excellent in vitro metabolic
stability of 2l–b would reflect its high oral bioavail-
ability. The total clearance of 2l–b (Cltot=13 mL/min/
kg) in dogs was slow, which resulted in a long plasma
half life (T1/2=15 h).
Acknowledgements
We are grateful to Kimberley Marcopul and Clinical
Literature Information Center (CLIC), Merck & Co.,
Ltd. for their critical reading of this manuscript.
References and Notes
Conclusion
1. Doods, H. N. Drug News Perspect. 1992, 345.
2. Wang, P.; Luthin, G. R.; Ruggieri, M. R. J. Pharmacol.
Exp. Ther. 1995, 273, 959.
3. Alabaster, V. A. Life Sci. 1997, 60, 1053.
4. Takeuchi, M.; Naito, M.; Morihira, K.; Ikeda, M.; Iso-
mura, Y. Jpn. Kokai Tokkyo Koho 1995, JP 07258250; Chem.
Abstr. 175858.
5. Miyachi, H.; Kiyota, H.; Segawa, M. Bioorg. Med. Chem.
Lett. 1998, 8, 2163.
6. Mitsuya, M.; Kobayashi, K.; Kawakami, K.; Satoh, A.;
Ogino, Y.; Kakikawa, T.; Ohtake, N.; Kimura, T.; Hirose, H.;
Sato, N.; Numazawa, T.; Hasegawa, T.; Noguchi, K.; Mase,
T. J. Med. Chem. 2000, 43, 5017.
7. Disse, B.; Speck, G. A.; Rominger, K. L.; Witek, T. J., Jr.;
Hammer, R. Life Sci. 1999, 64, 457.
8. Mitsuya, M.; Mase, T.; Tsuchiya, Y.; Kawakami, K.; Hat-
tori, H.; Kobayashi, K.; Ogino, Y.; Fujikawa, T.; Satoh, A.;
Kimura, T.; Noguchi, K.; Ohtake, N.; Tomimoto, K. Bioorg.
Med. Chem. 1999, 7, 2555.
Optimization of the (2R)-2-[(1R)-3,3-difluorocyclo-
pentyl]-2-hydroxy-2-phenylacetamides by focusing on
the amine part of the molecule led to 4-aminopiper-
adinamide 2l, which has potent binding affinity for the
M3 receptor (Ki=5.1 nM) and moderate selectivity for
M3 over the M2 receptor (M2/M3=46-fold). Further
optimization of the primary amine part of 2l resulted in
the identification of 2l–b with an 4-(2-aminoethyl)pi-
peridine group (Ki=3.7 nM, M2/M3=170-fold). As
expected, this compound showed improved metabolic
stability in the rat and dog in vitro hepatic microsome
assays because the major metabolic site of 1 was deleted.
In addition, this compound showed good oral bioavail-
ability in dogs. These results suggested that a hetero-
aromatic side chain of the piperidine moiety in 1 was
not essential to maintain the high M3 binding affinity
and selectivity over the M2 receptor and that deletion of
this side chain contributed the metabolic stability of the
molecules as well.
9. Cheng, Y. C.; Prusoff, W. H. Biochem. Pharmacol. 1973,
22, 3099.