Sulfide analogues as potent and selective M2 muscarinic receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(02)00096-3

We have discovered highly potent, selective sulfide M₂ receptor antagonists with low molecular weight and different structural features compared with our phase I clinical candidate Sch 211803. Analogue 30 showed superior M₂ receptor selectivity profile over Sch 211803. More importantly, this study provided new leads for the discovery of M₂ receptor antagonists as potential drug candidates.

Full text of DOI:10.1016/S0960-894X(02)00096-3

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1087–1091
Sulfide Analogues as Potent and Selective M2 Muscarinic
Receptor Antagonists
Yuguang Wang,* Samuel Chackalamannil, Zhiyong Hu, Brian A. McKittrick,
William Greenlee, Vilma Ruperto, Ruth A. Duffy and Jean E. Lachowicz
Schering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA
Received 21 November 2001; accepted 24 January 2002
Abstract—We have discovered highly potent, selective sulfide M2 receptor antagonists with low molecular weight and different
structural features compared with our phase I clinical candidate Sch 211803. Analogue 30 showed superior M2 receptor selectivity
profile over Sch 211803. More importantly, this study provided new leads for the discovery of M2 receptor antagonists as potential
drug candidates. # 2002 Elsevier Science Ltd. All rights reserved.
The successful characterization of five muscarinic
receptor subtypes (M1–M5) has greatly contributed to
the understanding of the physiological relationships
between the neurotransmitter acetylcholine (ACh) and
the muscarinic receptors.1 Among the five muscarinic
receptor subtypes, the post-synaptic M1 receptors inter-
act with ACh to promote cognition, memory and
learning, while the pre-synaptic muscarinic M2 receptors
regulate the synaptic levels of ACh via a feedback inhi-
bition of ACh release. The current treatment of cogni-
tive disorders such as Alzheimer’s Disease (AD) is
focused on the means to increase ACh levels.2 Suppres-
sion of the M2 receptor induced-feedback inhibition of
ACh release is an alternative mechanism for the con-
tinuous release of ACh to the synaptic cleft. Indeed, M2
receptor antagonists have been reported to enhance the
ACh release in animal models.3 As part of our research
program for the treatment of AD, we have been
involved in the development of M2 receptor antagonists.
One of these compounds, Sch 211803, has entered phase
1 clinical trials.4 Our next goal was to identify a backup
candidate for Sch 211803. We would like the backup
candidate to have different structural features with
lower molecular weight, while maintaining high M2
receptor binding affinity (Ki <5 nM) and selectivity
versus M1 and M3 receptors (>100-fold).
Sch 211803 to reduce its molecular weight. We first
explored the feasibility of replacing the chloro-
phenylsulfonyl group with an isopropylsulfonyl group
(see 1 and 2 in Fig. 1). The isopropyl group was expec-
ted to partially mimic the steric environment of the
aromatic ring. The selection of 1-naphthyl amide found
in 2 was based on our previous study.5
The targets were prepared as outlined in Scheme 1.
Reductive amination of commercially available piper-
idine derivative 3 afforded piperidinyl piperidine 4. Dis-
placement of the fluorine atom of 4 with isopropylthiol
yielded thiol ketone 5. The benzylic carbonyl groupof 5
was reduced to a benzylic methylene group( 6) via
treatment with NaBH4 followed by Et3SiH/TFA treat-
ment. The amine 6 was transformed to the aryl amides 7
and 8 by coupling with corresponding aryl acids. The
final products 1 and 2 were obtained after oxidation of
sulfur to sulfone.6
The sulfone targets 1 and 2 as well as their sulfide
intermediates 7 and 8 were assayed with cloned human
muscarinic receptors according to the reported protocol
(Table 1).7
The M2 receptor binding affinities of sulfones 1 and 2
were disappointing in comparison to that of Sch
211803. However, the sulfide precursors 7 and 8
demonstrated stronger M2 receptor binding affinity than
their sulfone derivatives. Especially, sulfide 8 showed
excellent M2 receptor binding affinity and selectivity.
Our strategy for the discovery of a backupcandidate
was to simplify the left-hand portion of the structure of
*Corresponding author. Tel.: +1-908-740-3508; fax: +1-908-740-
4563; e-mail: yuguang.wang@spcorp.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00096-3

1088
Y. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1087–1091
Figure 1. Modification of Sch 211803.
Scheme 1. (a) NaBH(AcO)3, 1,2-DCE, 1-t-butoxycarbonyl-4-piperidone, 75–84%; (b) NaH, DMF, isopropylthiol, 65 ꢀC, 6 h, 70–80%; (c) NaBH4,
MeOH, 90–98%; (d) TFA, Et3SiH, 70–85%; (e) ArCOOH, EDCI, DMAP, CH2Cl2, 85–96%; (f) mCPBA MeSO3H, CH2Cl2, 60–75%.
Therefore, we turned our attention to the optimization
of SAR of sulfide analogue 8. First, the 1-naphthyl ring
was replaced with other types of aromatic rings. The M2
Ki values of several selected analogues are shown in
Table 2. Replacement of 1-naphthyl ring with 2-naphthyl
ring (9) resulted in loss of activity by more than
200-fold. Modification of phenyl substitution (10–16) as
well as incorporation of heteroaryl moieties (17–19)
similarly produced compounds that are much less
active. These results suggest that the 1-naphthyl group
is essential for the high M2 receptor binding affinity.
Table 1. The results of M2 binding and selectively of prepared analogues
Compd
X
Ar
M2, Ki (nM) M1/M2 M3/M2 M4/M2 M5/M2
Sch 211803
0.89
734
11
787
na
69
na
95
na
1
SO2
104
Next, a series of substituted 1-naphthyl groups and
isosteric bicyclic heteroaryl derivatives was examined.
The results are shown in Table 3.
2
7
8
SO2
17
20
21
37
na
na
na
na
12
na
na
Generally, substitution of 1-naphthyl moiety drastically
reduced M2 receptor binding affinity compared to com-
pound 8 except compound 28. Compound 20, with M2
Ki of 6.1 nM, showed excellent selectivity versus other
receptor subtypes. The M2 receptor binding affinities of
fluoronaphthyl derivatives suggested a modest advan-
tage for 7-fluoro substitution (23). Among the quinoline
derivatives, 6-isoquinoline derivative 28 showed excellent
S
S
0.9
205
454
241
na, not available.

Y. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1087–1091
1089
M2 receptor binding affinity albeit poor selectivity ver-
sus other receptor subtypes, whereas the isomeric
quinoline derivatives (25–27) as well as the cinnoline
derivative 29 were much less active.
Finally, in Table 4, groups capable of hydrogen bonding
were chosen to replace the isopropyl group of 8 based
on previous modeling of M2 receptor antagonists which
suggested there might be a hydrogen bonding inter-
action at the left-hand side of the antagonist with M2
receptor.8 A series of analogues (30–32) was prepared
by following the synthetic route of Scheme 1 where
methyl thioglycolate was used instead of isopropyl thiol.
Hydrolysis of methyl ester 30 afforded carboxylic acid
35 while reduction of 30 provided alcohol 36. Other
esters (33 and 34) and amides (37 and 38) were synthe-
sized by coupling of acid 35 with appropriate alcohol or
amine.
Table 2. The results of M2 binding and selectivity of selected analo-
gues from modification of 8
Table 3. The results of M2 binding and selectivity of selected
1-naphthyl bioisosteres
Compd
Ar
M2, Ki (nM) M1/M2 M3/M2 M4/M2 M5/M2
8
0.9
205
na
454
na
12
na
na
na
241
na
9
241
Compd
Ar
M2, Ki (nM) M1/M2 M3/M2 M4/M2 M5/M2
10
11
260
192
na
na
na
8
0.9
6.1
205
163
68
454
190
na
12
65
na
241
219
na
na
na
na
20
21
14
12
13
108
129
na
na
na
na
na
na
na
na
22
22
22
37
na
na
na
23
24
25
37
98
39
na
na
na
na
na
na
na
na
na
14
15
250
134
na
na
na
na
na
na
na
na
8.6
20
16
17
237
254
na
na
na
na
na
na
na
na
26
27
28
29
11
79
3
na
na
na
na
na
na
na
na
na
na
na
na
256
18
19
134
923
na
na
na
na
na
na
na
na
1.0
58
3
256
na, not available.
na, not available.

1090
Y. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1087–1091
Table 4. The results of M2 binding and selectivity of selected analo-
gues of sulfide modification
In summary, we have discovered highly potent, selective
sulfide M2 receptor antagonists 8, 30–32. Sulfide 8 has
low molecular weight (Mr=486) and different structural
features compared with Sch 211803 (Mr=566). Analo-
gue 30 (Mr=516, Ki=0.7 nM) showed superior M2
receptor selectivity profile over Sch 211803. The oral
efficacy of the sulfide analogues was determined in the
microdialysis assay which measures the increase of Ach
release due to the blockade of the M2 inhibitory feed-
back mechanism.3 Analogue 8 showed much lower oral
activity than Sch 211803 perhaps due to its low meta-
bolism stability and high lipophilicity.9 Since the sulfide
atom can be metabolically oxidized and the naphthyl
group is lipophilic, our next SAR modification will
focus on the replacement of the sulfide atom with
metabolically stable bioisosteres and the naphthyl group
with less lipophilic groups. These SAR studies led to a
new generation of M2 receptor antagonists with low
molecular weight, different structural features, and
improved oral bioavailability and efficacy. The results of
these studies will be reported in the future.
Compd
R
R1 R2 M2, Ki M1/M2 M3/M2 M4/M2 M5/M2
(nM)
8
H
H
F
H
H
0.9
205
454
1532
301
467
na
12
57
12
32
na
na
na
na
na
na
241
409
218
538
na
30
31
32
33
34
35
36
37
38
0.7 1395
H
4.3
2.7
246
380
na
na
na
na
52
Acknowledgements
H
H
H
H
H
H
H
OEt
H
The authors would like to thank Mr. Yao H. Ing and
Dr. P. Das for mass spectral data, Dr. N. J. Murgolo
for molecular modeling studies, and Dr. J. W. Clader
for helpful discussion.
40
H
48
1500
26
na
na
References and Notes
H
na
na
1. For a recent review, see: Felder, C. C.; Bymaster, F. P.;
Ward, J. D.; Elapp, N. J. Med. Chem. 2000, 43, 4333.
2. (a) Brinton, R. D.; Yamazaki, R. S. Pharmaceutical Res.
1998, 15, 386. (b) Doods, H. N. Drugs Future 1995, 20, 157.
3. Billard, W.; Binch, H., III; Crosby, G.; McQuade, R. D.
J. Pharmacol. Exp. Ther. 1995, 273, 273.
4. Asberom, T.; Billard, B.; Binch, H.; Clader, J. W.; Cox, K.;
Crosby, G.; Duffy, R. A.; Ford, J.; Greenlee, W.; Guzik, H.;
Kozlowski, J. A.; Lachowicz, J. E.; Li, S.; Liu, C.; Lowe, D.;
McCombie, S.; Ruperto, V. B.; Strader, C.; Tayler, L. A.;
Vice, S.; Zhao, H.; Zhou, G. 221nd ACS National Meeting,
San Diego, CA, April 1–5, 2001.
H
na
na
H
18
na
na
H
12
73
na
na
na, not available.
5. Wang, Y.; Chackalamannil, S.; Chang, W.; Greenlee, W.;
Ruperto, V.; Duffy, R. A.; McQuad, R.; Lachowicz, J. E.
Bioorg. Med. Chem. Lett. 2001, 11, 891.
6. All of the target compounds showed satisfactory result in
the analyses of NMR, MS, LC/MS, and HRMS.
Methyl esters 30–32 showed excellent M2 receptor
binding affinity and selectivity. Among them, ester 30
showed the best selectivity versus other receptor sub-
types. Interestingly, there was no obvious hydrogen
bonding effect based on the comparison of the M2
receptor binding affinity of 8 with 30, and 33–38. On the
other hand, methyl glycolates 31 and 32 had a modest
increase of M2 receptor binding affinity versus their
isopropyl analogues 23 and 21 (Table 3). Esters 33 and
34 had lower M2 receptor binding affinities perhaps due
to steric hindrance. Likewise, amides 37 and 38 demon-
strated 10- to 20-fold reduction in M2 receptor binding
affinity compared to methyl ester 30. The carboxylic
acid 35 and alcohol 36 were also less active. These
results showed that steric bulk and acidity are not tol-
erated for the left-hand side modifications.
7. For radioligand binding analysis, each muscarinic receptor
subtype was stably expressed in CHO-K1 cells. Clonal cell
lines were selected which expressed receptors at levels between
1 and 9 pmol/mg protein. The Kd of QNB (l-quinuclidinyl
benzilate) at each receptor subtype was determined by satura-
tion binding using 5–2500 pM [3H] QNB in 10 mM potassium
phosphate buffer, pH 7.4. Protein concentrations were adjus-
ted for each assay to achieve between 700 and 1500 cpm spe-
cific binding. Competition binding experiments were
performed using 180 pM [3H] QNB. All binding experiments
were performed in the presence of 1% DMSO and 0.4%
methylcellulose. Nonspecific binding was defined by 0.5 mM
atropine. After equilibrium was reached (120-min incubation
at room temperature), bound and free radioactivity were

Y. Wang et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1087–1091
1091
separated by filtration using Whatman GF-C filters. Investi-
gation of M2 receptor antagonist activity was performed on
related compounds by measuring the effects of the adenylyl
cyclase inhibition mediated by oxotremorine M. Ki was
expressed as mean of duplicate value (SEM<15%). All deter-
minations were performed at least twice.
8. Part of the molecular modeling work for the sulfoxide ser-
ies has been reported. See: Lachowicz, J. E.; Lowe, D.; Duffy,
R. A.; Ruperto, V.; Taylor, L. A.; Guzick, H.; Brown, J.;
Berger, J. G.; Tice, M.; McQuade, R.; Kozlowski, J.; Clader,
J. W.; Strader, C. D.; Murgolo, N. Life Sci. 1999, 64.
9. At 10 mpk oral dosing, the AUC (0–2.5 h) of analogue 8 is
1009 while the AUC (0–2.5 h) of Sch 211803 is 7800 (details
will be reported by W. Billard). During the human microsomal
incubation, there was 26% of analogue 8 left compared with
90% of Sch 211803 (details will be reported by K. A. Cox).
The Clog P of analogue 8 is 6.3 (details will be reported by
J. Kaminski).