M3 muscarinic acetylcholine receptor antagonists: SAR and optimization of bi-aryl amines

Article abstract of DOI:10.1016/j.bmcl.2009.01.098

Exploration of multiple regions of a bi-aryl amine template led to the identification of highly potent M₃ muscarinic acetylcholine receptor antagonists such as 14 (pA₂ = 11.0) possessing good sub-type selectivity for M₃ ov

Full text of DOI:10.1016/j.bmcl.2009.01.098

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1686–1690
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
M3 muscarinic acetylcholine receptor antagonists: SAR and optimization
of bi-aryl amines
b,
b
c
b
a
a
Brian Budzik b, , Yonghui Wang , Dongchuan Shi , Feng Wang , Haibo Xie , Zehong Wan , Chongye Zhu ,
James J. Foley a, Parvathi Nuthulaganti c, Lorena A. Kallal c, Henry M. Sarau a, Dwight M. Morrow c,
Michael L. Moore b, Ralph A. Rivero b, Michael Palovich a, Michael Salmon a, Kristen E. Belmonte a,
*
*
Dramane I. Laine a, Jian Jin a,
*
a Centers of Excellence for Drug Discovery, GlaxoSmithKline, 709 Swedeland Road, King of Prussia, PA 19406, USA
b Centers of Excellence for Drug Discovery, GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
c Molecular Discovery Research, GlaxoSmithKline, 1250 South Collegeville Road, Collegeville, PA 19426, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Exploration of multiple regions of a bi-aryl amine template led to the identification of highly potent M3
muscarinic acetylcholine receptor antagonists such as 14 (pA2 = 11.0) possessing good sub-type selectiv-
ity for M3 over M2. The structure–activity relationships (SAR) and optimization of the bi-aryl amine series
are described.
Received 5 December 2008
Revised 29 January 2009
Accepted 29 January 2009
Available online 4 February 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Muscarinic acetylcholine receptor
M3 mAChR
Antagonists
SAR
Sub-type selectivity
COPD
Asthma
Bi-aryl amines
Five muscarinic acetylcholine receptor (mAChR) sub-types, M1–
M5, are known to date.1–3 These seven-transmembrane (7TM)
receptors share a common orthosteric ligand-binding site with
an extremely high sequence homology, which explains why it
has been difficult historically to identify sub-type selective li-
gands.3 M1–M5 mAChRs are widely distributed in mammalian or-
gans and the central and peripheral nerve system where they
mediate important neuronal and autocrine functions.4,5 In the
mammalian lung, only M1, M2, and M3 mAChRs have been recog-
nized as playing important and functional roles.6 M3 is predomi-
nantly expressed on airway smooth muscle and mediates smooth
muscle contraction.7 M2 is primarily found on postganglionic nerve
termini and functions to limit acetylcholine release from parasym-
pathetic nerves.8 M1 is found in parasympathetic ganglia and facil-
itates neurotransmission through ganglia thus enhancing
cholinergic reflexes.9
In chronic obstructive pulmonary disease (COPD) and asthma,
inflammatory conditions lead to loss of neuronal inhibitory activity
mediated by M2 on parasympathetic nerves, causing excess acetyl-
choline reflexes10 which result in airway hyperreactivity and
hyperresponsiveness mediated by increased acetylcholine release
and thus excess stimulation of M3. Therefore, potent mAChR antag-
onists, particularly directed toward the M3 sub-type, would be use-
ful as therapeutics in these mAChRs-mediated disease states. We
previously disclosed a novel muscarinic acetylcholine receptor
antagonist bi-phenyl piperazine series, which displayed high po-
tency, good sub-type selectivity for M3 over M2, and excellent
in vivo efficacy and duration of action.11 Herein we further de-
scribe the structure–activity relationships (SAR) and optimization
of this bi-phenyl piperazine series from a high throughput screen
(HTS) to the identification of compound 14.
* Corresponding authors at present address: 1031 Perkiomenville Road, Perkio-
menville, PA 18074, USA. Tel.: +1 610 754 9132 (B. Budzik); Medicinal Chemistry,
R&D China, GlaxoSmithKline, 898 Halei Road, Shanghai 201203, China. Tel.: +1 86
21 61590761; fax: +1 86 21 61590730 (Y. Wang); Center for Integrative Chemical
Biology & Drug Discovery, Division of Medicinal Chemistry & Natural Products,
Eshelman School of Pharmacy, The University of North Carolina at Chapel Hill,
Chapel Hill, NC 27599. Tel.: +1 919 843 8459; fax: +1 919 843 8465 (J. Jin).
E-mail addresses: torqux@yahoo.com (B. Budzik), yonghui.2.wang@gsk.com (Y.
Wang), jianjin@unc.edu (J. Jin).
A HTS of the GSK compound collection identified the bi-phenyl
piperazine 1 (Fig. 1),11 as a M3 antagonist in a fluorometric imaging
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.098

B. Budzik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1686–1690
1687
plate reader (FLIPR) assay12 with a pIC50 of 7.5.13 On that basis, 1
was considered an acceptable starting point for our lead optimiza-
tion program aimed at identifying long acting M3 mAChR antago-
nists via optimization of their potency.
Generally, modification of the RHS phenyl ring did not improve
potency (Table 2). Thus, introduction of substituents onto the phe-
Table 1
An efficient and robust solid-phase synthesis permitting mod-
ification of several regions of the molecule was developed to ex-
plore this series (Scheme 1). Commercially available 3-bromo
benzylamines 2 were loaded onto 2,6-dimethoxy-4-polystyre-
nebenzyloxy benzaldehyde resin (DMHB resin)14 via reductive
amination and coupled with benzoic acids to afford resin-bound
aryl bromides 3. Suzuki coupling of 3 with 3-formylphenyl boro-
nic acids or other formyl substituted aryl boronic acids and sub-
sequent reductive amination of the resulting benzaldehydes with
piperazines or various amines, followed by resin cleavage, pro-
duced the targeted bi-aryl amines 4 in excellent yields and
purity.
SAR of the RHS diamine region
O
O
N
H
R
O
Compound
R
M3 pIC50
7.5
NH
N
1
N
N
5a
6.3
We began our initial SAR exploration on the right hand side
(RHS) piperazine region of the HTS hit 1 (Table 1).15 N-Alkylation
of the piperazine ring of 1 with methyl (5a) or ethyl (5b) re-
sulted in over a log unit loss of potency. Replacing the outer
piperazine nitrogen with a carbon atom (5c) or a non-basic
hydrogen bond acceptor as in 5d yielded poorly active analogs,
thus suggesting that a basic center is required for M3 potency
at that position. The homopiperazine analogs 5e and 5f did not
show improved potency at M3 compared to the corresponding
piperazine derivatives 1 and 5a. A methylene bridged bicyclic
piperazine (5g) was nearly equipotent to 1. The incorporation
of a carbonyl group on the methylene linker (5h),16 directly on
the piperazine ring to form a lactam (5i), or on the terminal
nitrogen (5j) provided at best poorly active compounds. How-
ever, substituting the piperazine with a methyl group (5k)17 en-
hanced the potency. Further testing of the two stereoisomers of
5k showed that the (2S) enantiomer (5l) was the most potent
moiety, giving a potency increase of 0.8 log unit over 1. An at-
tempt was also made to introduce a quaternary ammonium moi-
ety to the molecule in order to reduce cell membrane
permeability. However, these efforts were fruitless as quaterniz-
ing either the inner nitrogen (5n)18 or the outer nitrogen (5o)19
did not maintain potency.
N
5b
5c
5.8
5.2
N
N
O
5d
5e
5f
<5.0
6.5
6.2
7.1
5.3
<5.0
5.4
8.0
N
N
N
NH
N
H
NH
5g
5h
5i
N
H
NH
N
O
NH
O
NH
N
N
O
O
O
N
H
O
5j
N
1, M3 FLIPR pIC50 = 7.5
N
N
Figure 1. In vitro profile of HTS hit 1.
NH
NH
5k
5l
O
Br
Br
a, b
H2N
N
R2
8.3
7.3
<5.0
5.6
N
R1
R1
DMHB
3
2
NH
NH
5m
5n
5o
R4
N
O
Ar
N
R3
c, d, e
N
H
R2
N+
R1
4
Scheme 1. Reagents and conditions: (a) DMHB resin, Na(OAc)3BH, DIEA, 10% AcOH
in NMP, rt; (b) various benzoic acids, DIC, NMP, rt; (c) various 3-formylphenyl
boronic acids or other formyl substituted boronic acids, Pd(PPh3)4, 2 M Cs2CO3,
DME, 80 °C; (d) piperazines or various amines, Na(OAc)3BH, Na2SO4, DCE, rt; (e) 50%
of TFA in DCE, rt.
N+
N

1688
B. Budzik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1686–1690
nyl ring, gave compounds (6a, 6b, and 6c) of lower potency than 1.
A similar outcome was observed when the phenyl ring was re-
placed with a pyridine (6d) or a thiophene (6e).
The left hand side (LHS) phenyl ring of the bi-aryl, in contrast,
proved to be another potency enhancing region (Table 3). The addi-
tion of fluorine at position 6 of the LHS phenyl ring (7a) improved
the potency to a pIC50 of 8.2. Incorporation of a methoxy group at
the same position (7b) gave an equipotent compound to 1, while
other substitutions resulted in potency losses (7c–7f).
In the amide linker region, replacement of the amide group of 1
(Table 4) with a sulfonamide (8a)20 led to a 10-fold loss of potency.
Removal of the methylene (8b)20 or replacement of the amide with
a urea (8c)20 resulted in even greater potency decrease at M3.
The attachment point and nature of various substituents on the
LHS benzamide phenyl were subsequently explored in a systematic
manner (Table 5). An initial set of compounds, investigating vari-
ous electron-donating and electron-withdrawing groups of differ-
ent sizes (9a–9l), clearly established that the 3- or meta-position
was the most preferred but did not uncover any compound more
potent than 1.
Building on this finding, our next step was to focus exclusively
on the meta-position and investigate a larger set of substituents
(9m–9z). From this exercise, the acetyl 9v and ethoxy 9t stood
out, both displaying a M3 pIC50 of 8.0. The most promising result,
however, was the identification of the N-methyl piperazine deriv-
ative 9z. In addition to being the most potent of the meta deriva-
tives prepared, it also showed that amine containing
Table 4
SAR of the amide linker region
NH
R
N
O
O
Table 2
SAR of the bi-aryl RHS phenyl ring
Compound
R
M3 pIC50
7.5
O
NH
O
Ar
N
O
O
N
H
R
1
N
H
Compound
Ar
R
M3 pIC50
7.5
O
O
S
8a
8b
8c
6.6
5.2
5.1
N
H
1
H
O
MeO
N
H
O
6a
6b
6c
5l
H
H
H
5.9
6.7
6.5
8.3
7.4
6.3
N
N
H
F
H
Table 5
OMe
SAR of the monosubstituted benzamides
O
N
NH
N
R
H
(S)-Methyl
(S)-Methyl
(S)-Methyl
Compound
R
M3 pIC50
9a
9b
9c
9d
9e
9f
9g
9h
9i
2-CN
3-CN
4-CN
2-OMe
3-OMe
4-OMe
2-Cl
3-Cl
4-Cl
6.5
7.3
7.1
6.3
7.1
6.3
6.0
7.0
6.5
5.4
7.3
6.2
6.1
7.7
7.7
6.7
5.7
6.7
6.3
8.0
5.9
8.0
5.7
6.6
6.5
8.4
6d
6e
N
S
9j
9k
9l
9m
9n
9o
9p
9q
9r
2-CF3
3-CF3
4-CF3
3-F
3-COOMe
3-SO2Me
3-N(CH3)2
3-NH2
3-NHAc
3-OCF3
3-OEt
3-OPh
3-Ac
Table 3
SAR around the bi-aryl LHS phenyl ring
O
NH
2
N
O
N
H
4
6
O
R
Compound
R
M3 pIC50
9s
9t
1
6-H
6-F
6-OMe
4-F
4-Me
2-F
2-Me
7.5
8.2
7.5
6.6
5.5
5.6
<5.0
9u
9v
9w
9x
9y
9z
7a
7b
7c
7d
7e
7f
3-OH
3-Me
3-Et
3-N-Methyl piperazin-methyl

B. Budzik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1686–1690
1689
functionalities could be placed on the LHS, thus introducing new
potential quaternization points into the template.
12b (compare with 5l and 7a) were additive. The M3 potencies of
N-methylhomopiperazine 12d and the acyclic methylated diamine
12e were similar to 12a. Monoamines were also tolerated in the re-
gion. Thus, formamide 12f and morpholine 12g showed excellent
M3 potency. The piperidine 12h and the simple dimethyl amine
Encouraged by the good potency of 9z, we proceeded to explore
the SAR around its key N-methyl piperazine moiety. At that stage,
it was also decided to incorporate the previously identified potency
enhancing modifications, the (2S)-2-methyl piperazine group of 5l
and the fluoro substituent of 7a, into the core template. To enable
this study, we needed first to modify our existing solid-phase route
to make it amenable to the introduction of diverse amines/dia-
mines or quaternized amines/diamines on the LHS (Scheme 2). 3-
Bromo-4-fluorobenzylamine 10 was loaded onto a DMHB resin
via reductive amination, followed by coupling with 3-formylphe-
nyl benzoic acid and subsequent reductive amination with various
amines or diamines, leading to resin-bound benzylamines 11. Sub-
sequent Suzuki coupling of 11 with 3-formylphenyl boronic acid
followed by reductive amination with (2S)-2-methyl piperazine
and resin cleavage, produced the targeted bi-phenyl piperazines
12. An optional quaternization step after Suzuki coupling using
methyl iodide or various alkyl halides gave quaternized bi-phenyl
piperazines such as 13a and 13b.
Table 6
SAR of the LHS amine region
O
NH
R
N
N
H
F
Compound
R
M3 pA2
9.8
N
12a
HN
N
12b
12c
12d
12e
12f
12g
12h
12i
10.5
10.2
9.6
N
The results of the SAR investigation of the LHS amine region are
shown in Table 6. The piperazine derivatives 12a, 12b, and 12c
showed improved M3 activity with respective pA2 values21 of 9.8,
10.5, and 10.2. Gratifyingly, the M3 potency of the N-methylpiper-
azine 12b increased over 2 log units compared to the analogous 9z,
indicating the potency enhancements installed in each region of
N
HO
N
N
N
N
O
N
9.7
R1
Br
Br
F
a, b,
C
H2N
N
N
R2
N
F
DMHB
H
O
N
N
10.1
9.9
10
11
O
O
O
NH
R1
N
d, e, f
N
N
H
R2
F
12
N
9.2
Br
F
d, g, e, f
N
N
N
N
8.5
DMHB
11a
Boc
N+
N
O
NH
13a
13b
13c
13d
13e
14
8.5
N+
N
d, g, e, f
N
HN
N
H
HN
N
F
13a
N+
9.8
O
Br
F
N
N
N
N
N+
10.0
10.1
9.9
DMHB
11b
N
NH
N+
N
O
N
H
N+
F
13b
N+
Scheme 2. Reagents and conditions: (a) DMHB-resin, Na(OAc)3BH, DIEA, 10% AcOH
in NMP, rt; (b) 3-formyl benzoic acid, DIC, DCE/DMF 1:1, rt; (c) R1R2NH,
Na(OAc)3BH, Na2SO4, DCE, rt; (d) 3-formylphenyl boronic acid, Pd(PPh3)4, 2 M
Cs2CO3, DME, 80 °C; (e) (2S)-2-methylpiperazine, Na(OAc)3BH, Na2SO4, DCE, rt; (f)
50% of TFA in DCE, rt; (g) MeI, CH3CN, rt.
11.0
HN

1690
B. Budzik et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1686–1690
10. Fryer, A. D.; Adamko, D. J.; Yost, B. L.; Jacoby, D. B. Life Sci. 1999, 64, 449.
11. Jin, J.; Budzik, B.; Wang, Y.; Shi, D.; Wang, F.; Xie, H.; Wan, Z.; Zhu, C.; Foley, J. J.;
Webb, E. F.; Berlanga, M.; Burman, M.; Sarau, H. M.; Morrow, D. M.; Moore, M.
L.; Rivero, R. A.; Palovich, M.; Salmon, M.; Belmonte, K. E.; Lainé, D. I. J. Med.
Chem. 2008, 51, 5915.
12i showed decreased potency compared to 12a–12g, suggesting
that an outlying heteroatom may be needed for maximum M3
potency.
Unlike the RHS diamine region, quaternization was well toler-
ated in the LHS diamine region (Table 6). Quaternization of the out-
er nitrogen of 12a was more favored than that of the inner nitrogen
(13a) and led to a compound 13b, which was equipotent to 12a.
Other quaternized piperazines (13c–13e) also gave excellent M3
potencies with pA2 values around 10. Quaternary ammonium salts
such as 13b indeed had extremely low membrane permeability
(<3 nm/s).22 Interestingly, the increased basicity of the piperidine
derivative 14 compared to 12a,23 correlated with an increased
affinity for the M3 receptor, suggesting that the terminal N of these
molecules may be interacting with an acidic residue. As previously
reported, compound 14 was found to exhibit sub-type selectivity
for M3 over M1 and M2 (pA2s for M3, M2, and M1 are 11.0, 8.3,
and 9.7, respectively).11 In addition, 14 also displayed excellent
inhibitory activity and long duration of action in a bronchocon-
striction in vivo model via intranasal administration.11
12. (a) Measuring inhibition of acetylcholine-mediated [Ca2+]i-mobilization in
Chinese hamster ovary (CHO) cells stably expressing human recombinant M3
receptor.; (b) For FLIPR assay details, see supporting information in Ref. 11.
13. The biological assay results in the paper are a mean of at least 2 determinations
with standard deviation of < 0.3, and data for compounds M3 pIC50 > 7.0 are
from 96-well FLIPR assay, <7.0 from 384 well FLIPR assay, unless otherwise
noted.
14. (a) Jin, J.; Graybill, T. L.; Wang, M. A.; Davis, L. D.; Moore, M. L. J. Comb. Chem.
2001, 3, 97; (b) Available from Polymer Laboratories, Part number: 1466-6689,
150–300 lm, 1.5 mmol/g loading.
15. (a) All new compounds in this paper were characterized via LC/MS and 1H
NMR.; (b) For representative experimental procedures, see supporting
information in Ref. 11.
16. Compound 5h was synthesized by using 3-(dihydroxyboranyl)benzoic acid
instead of 3-formylphenyl boronic acid in step (c) of Scheme 1, followed by DIC
mediated coupling with Boc-piperazine, followed by cleavage step (e).
17. Use of any unprotected, 2-substituted piperazine in step (d) of Scheme 1 gives
completely selective reductive amination at the less hindered nitrogen.
18. Quaternization of the inner piperazine nitrogen in 5n was affected by use of
Boc-piperazine in step (d) of Scheme 1, followed by reaction of the washed
resin with excess CH3I in CH3CN at rt for overnight, followed by cleavage step
(e).
19. Quaternization of the outer piperazine nitrogen in 5o was affected by use of N-
methyl-piperazine in step (d) of Scheme 1, followed by reaction of the washed
resin with excess CH3I in CH3CN at rt for overnight, followed by cleavage step
(e), quaternization being completely selective for the less sterically hindered
outer nitrogen.
In summary, the extensive SAR exploration of multiple regions
of the HTS hit 1 led to the identification of key structural motifs
necessary to achieve high M3 potency. The combination of these
features resulted in the discovery of the highly potent M3 antago-
nist 14 (pA2 = 11.0).
20. Compounds 8a and 8c were prepared by performing sulfonylation (using 1,3-
benzodioxole-5-sulfonyl chloride in presence of DMAP in pyridine/DCE), or
urea formation (using 5-isocyanato-1,3-benzodioxole in DCE), respectively,
instead of amide formation in step (b) of Scheme 1. Compound 8b was
prepared by using 3-bromoaniline instead of 3-bromobenzylamine in step (a)
of Scheme 1, and required more forcing amide coupling conditions in step (b):
DMAP, DIC, in DCE/DMF at 80 °C.
21. The pIC50 limit of the M3 FLIPR assay was about 9.0. pA2 was determined and
used to compare potency for compounds with pIC50 reaching the limit. See Ref.
11 for assay details of pA2 determination.
Acknowledgments
We thank Bing Wang for NMR, Qian Jin for LC/MS, Carl Bennett
for purification, and Christina Schulz-Pritchard, Jim Fornwald, and
Jeffrey Guss for assay support.
References and notes
22. The artificial phospholipid membrane technique is similar to the widely used
Caco-2 cell monolayer permeation technique. In short, egg phosphatidyl
choline (1.8%) and cholesterol (1%) are dissolved in n-decane. A small amount
of the volatile mixture is applied to the bottom of the microfiltration filter
inserts. Phosphate buffer (0.05 M, pH 7.05) is quickly added to the donors and
receivers, and the lipids are allowed to form self-assembled lipid bilayers
across the small holes in the filter. Permeation experiment is initiated by
spiking the compounds of interest to the donor sides, and the experiment is
stopped at a pre-determined elapsed time. The samples are withdrawn and
transferred to appropriate vials for analysis by HPLC with UV detection
(215 nm).
1. Caulfield, M. P.; Birdsall, N. J. M. Pharmacol. Rev. 1998, 50, 279.
2. Eglen, R. M. Prog. Med. Chem. 2005, 43, 105.
3. Hulmes, E. C.; Birdsall, N. J. M.; Buckley, N. J. Ann. Rev. Pharmacol. Toxicol. 1990,
30, 633.
4. Caulfield, M. P. Pharmacol. Ther. 1993, 58, 319.
5. Eglen, R. M. Auto. Autacoid Pharmacol. 2006, 26, 219.
6. Lee, A. M.; Jacoby, D. B.; Fryer, A. D. Curr. Opin. Pharmacol. 2001, 1, 223.
7. (a) Barnes, P. J. Thorax 1989, 44, 161; (b) Gater, P. R.; Alabastar, V. A.; Piper, I.
Pulm. Pharmacol. 1989, 2, 87.
8. Faulkner, D.; Fryer, A. D.; MacLagan, J. Br. J. Pharmacol. 1986, 88, 181.
9. Lammers, J. W. J.; Minnette, P. A.; Mc Custer, M.; Barnes, P. J. Annu. Rev. Respir.
Dis. 1989, 139, 446.
23. Full experimental procedures and characterization for the synthesis of
piperidine 14 are given in Ref. 11.