The discovery of AZD9164, a novel muscarinic M3 antagonist

Article abstract of DOI:10.1016/j.bmcl.2011.10.002

The optimization of a new series of muscarinic M₃ antagonists is described, leading to the identification of AZD9164 which was progressed into the clinic for evaluation of its potential as a treatment for COPD.

Full text of DOI:10.1016/j.bmcl.2011.10.002

Bioorganic & Medicinal Chemistry Letters 21 (2011) 7440–7446
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
The discovery of AZD9164, a novel muscarinic M3 antagonist
Antonio Mete a, , Keith Bowers b, Eric Chevalier b, , David K. Donald a, Helen Edwards b,
,
⇑
a
c
a
a
d
Katherine J. Escott b, , Rhonan Ford , Ken Grime , Ian Millichip , Barry Teobald , Vince Russell
⇑
a Department of Chemistry, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leicestershire LE11 5RH, UK
b Department of Bioscience, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leicestershire LE11 5RH, UK
c Department of DMPK, AstraZeneca R&D Charnwood, Bakewell Road, Loughborough, Leicestershire LE11 5RH, UK
d Argenta Discovery, Stoke Poges, Slough SL2 4SY, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The optimization of a new series of muscarinic M3 antagonists is described, leading to the identification of
AZD9164 which was progressed into the clinic for evaluation of its potential as a treatment for COPD.
Ó 2011 Elsevier Ltd. All rights reserved.
Received 16 August 2011
Revised 30 September 2011
Accepted 3 October 2011
Available online 8 October 2011
Keywords:
COPD
Muscarinic M3 antagonist
Bronchoconstriction
Salivation
Selectivity
Muscarinic acetylcholine (ACh) receptors are widely distributed
in the body and play important roles in the regulation of many
physiological functions of the central and peripheral nervous sys-
tems.1 They are members of the G-protein-coupled receptor family
and to date five different muscarinic receptor subtypes have been
identified, namely M1, M2, M3, M4 and M5.2 The M3 subtype is pres-
ent in the lung on airway smooth muscles and submucosal glands
and when activated, by ACh released from parasympathetic nerves,
produces bronchoconstriction, mucus secretion and subsequently
an increase in airways resistance, as observed in chronic obstruc-
tive pulmonary disease (COPD). Antagonism of the M3 receptor
sub-type leads to improved lung function and is an established
part of the therapeutic management of COPD.3 The muscarinic
antagonists currently used to treat COPD (ipratropium and tiotro-
pium) and newer antagonists such as glycopyrrolate4 and aclidini-
um bromide,5 which are in late stage clinical development, are all
non-selective across the M1–M5 receptor subtypes. All these com-
pounds (Table 1) are quaternary ammonium salts and carry a fixed
positive charge which affords them very low permeability and
makes them poorly absorbed. The compounds are administered
directly to the lung by inhaled delivery and this, in combination
with their poor permeability, limits their systemic exposure. In
addition, all these compounds contain an ester function, which
can be metabolised upon systemic exposure, further reducing
their potential to cause side effects, such as tachycardia, via
antagonism of cardiac M2 receptors. Tiotropium is currently
considered the bronchodilator of choice for the treatment of COPD
and is generally well tolerated with dry-mouth being the main
reported side-effect.3 Tiotropium is a highly potent6 and long
acting muscarinic antagonist and any new muscarinic antagonist
developed for COPD should match tiotropium for efficacy and
duration of action whilst reducing the dry mouth and other
dose-limiting side effects.
Our program aimed to deliver this through a combination of in-
creased selectivity over M2 and very low systemic exposure whilst
maintaining high muscarinic M3 antagonist potency. We aimed to
achieve low systemic exposure by having compounds which pos-
sessed high metabolic instability coupled with high human plasma
protein binding (hPPB) in order to reduce the free concentration of
any systemic circulating compound.
A common feature of tiotropium, glycopyrrolate and aclidinium
bromide is the presence of a tertiary hydroxyl group (–OH) at the
a-position of the ester function and it is a retained feature in many
muscarinic antagonists (Table 1). This polar function contributes to
the low hPPB of these compounds and, in our search for com-
pounds with high hPPB, we tested the possibility of replacing the
–OH group of glycopyrrolate by small lipophilic groups such as a
methyl. This led us to analogues such as 1.1 and 1.2 which,
⇑
Corresponding authors. Tel.: +44 1509 644251; fax: +44 1509 645571 (A.M.);
tel.: +44 1625 514519 (K.J.E.).
E-mail addresses: antonio.mete@astrazeneca.com, tony.mete@talktalk.net
(A. Mete), Jane.escott@astrazeneca.com (K.J. Escott).
No longer AZ employees.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.10.002

A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
7441
Table 1
Muscarinic antagonists in clinical use or in late stage development
a
a
Structure
Name
pIC50 M3
pIC50 M2
HM Cl_intb
(
ll/min/mg)
hPPB (%) bound
S
+
N
Br-
Tiotropium bromide
>10.3
>10.3
10.2
<3
48
O
O
OH
S
S
O
S
Br-
Aclidinium bromide
>10.4
>145
No data (not stable in plasma)
O
+
N
OPh
OH
O
O
Glycopyrrolate
9.9
9.8
<3
<30
Br-
OH
+
O
N
a
Binding affinities measured using recombinant human M3 or M2 receptors (n P 2).
Intrinsic clearance in human liver microsomes.
b
rapid access to chirally pure compounds for all the series described
in this Letter.9
Table 2
Early lead compounds
Table 3 summarizes key compounds synthesized and evaluated
in which –NR1R2 represents a range of cyclic and acyclic amines.
Compounds were initially made as racemates to establish which
changes were compatible with high potency. We were encouraged
to find that compounds in which –NR1R2 was a cyclic amine,
2.1–2.4, exhibited good potency at M3, with the compound bearing
an N-linked piperidine, 2.2, demonstrating good potency against
M3 both at the isolated receptor and in the functional guinea pig
trachea (GPT) tissue-based assay.10
Structure
Compound pIC50 HM
hPPB
(%)
bound
a
M3
Cl_intb
l/min/
(l
mg)
O
O
1.1
1.2
8.8
42
77
58
Br-
+
The ability of compounds to antagonize methacholine-induced
smooth muscle contraction in GPT provides a measure of antago-
nist potency for compounds acting via the M3 receptor in vitro.
The effect of quaternisation was compared in compounds 2.1 and
3.1 and this showed little influence on potency. Similarly, replacing
the basic piperidine attached to the ester in 2.1 with (3R)-quinucli-
dinyl in 4.1 had little effect on potency. Changes to the N-linked
piperidine such as substitution with both electron-withdrawing,
2.5, and electron-donating groups, 2.6, were explored and both of
these changes resulted in a drop in potency relative to the unsubsti-
tuted piperidine parent 2.2. Compounds in which –NR1R2 were acy-
clic amines, 2.7 and 2.8, had significantly lower potency. Replacing
the phenyl ring in 2.2 with thiophene gave compound 2.9 which
exhibited a similarly high level of activity. Substitution on the phe-
nyl ring, 2.10, was tolerated albeit with a drop in potency.
The data in Table 3 encouraged us to pursue an optimization
program in which the focus was on compounds where –NR1R2
was an N-linked piperidine, the aryl group Ar was either a phenyl
or a thiophene ring and the (3R)-quinuclidinyl group was quatern-
ised by a wide range of groups to optimize the physicochemical
properties of the compounds of type 5, 6, 7 (Table 4). The aim
was to identify ’quaternising’ groups that afforded compounds
which matched our desired target product profile (TPP). Our crite-
ria were:
N
Br-
+
10.5
35
O
O
N
a
Binding affinities measured using recombinant human M3 receptors (n P 2).
Intrinsic clearance in human liver microsomes.
b
gratifyingly, possessed encouraging levels of M3 activity coupled
with higher hPPB and metabolic liability relative to glycopyrrolate
and tiotropium (Table 2).7,8
Having established that a methyl group can replace the hydro-
xyl opened up the potential for novel chemical series. An idea that
proved attractive was where the cyclopentyl ring in 1.2 was re-
placed by a saturated N-heterocycle attached to the chiral carbon
via the N-atom leading to compounds of type 2, 3 and 4 (Table
3). This structural difference offered ease of synthesis of the qua-
ternary chiral centre at the
to compounds such as 1.2. The synthesis starts from a phenyl ace-
tic acid derivative 2a, which is halogenated at the -position to af-
a-position of the ester group, compared
a
ford 2b followed by displacement of the halide by the required
amine to 2c, as outlined in Scheme 1. We found that not only
was this synthesis amenable to the introduction of a wide range
of groups but the resulting enantiomers of the amino-esters 2c
were readily separable by standard chiral HPLC techniques, giving
ꢀ High in vitro and in vivo potency
o Receptor binding and functional assay (guinea pig trachea)
pIC50 ꢁ10 to match tiotropium.
o An ED80 dose in vivo 63 lg/Kg to match the potency of other
compounds in clinical development.

7442
A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
Table 3
Optimization of the N-linked amine (NR1R2) group
R1
R2
R1
R2
R1
R2
N
N
N
O
O
O
Ar
Ar
Ar
+
O
N
O
N
O
N
2
3
4
a
b
Compound
Ar
NR1R2
M3 pIC50
GPT pA2
2.1
2.2
2.3
2.4
2.5
2.6
2.7
2.8
2.9
2.10
3.1
4.1
Ph–
Ph–
Ph–
Ph–
Ph–
Ph–
Ph–
Ph–
Pyrrolidine
Piperidine
Azepine
Thiamorpholine
4,4-Di-F-piperidine
4-Me-Piperidine
Dimethylamine
Dipropylamine
Piperidine
8.5
9.6
9.1
9.1
8.9
9.0
8.0
6.8
10
—
10.4
—
—
—
—
—
—
—
2-Th
4-Cl-Ph
Ph–
Piperidine
Pyrrolidine
Pyrrolidine
8.6
8.3
8.3
—
—
—
Ph–
a
Binding affinities measured using recombinant human M3 receptors (n P 2).
Antagonist potency at the M3 receptor measured in guinea pig trachea (n = 2).
b
R1
R1
R2
O
N
R2
O
Br
N
O
O
ii
i
vi
O
O
O
O
iii
N
2c
2b
4
2a
vii
iv
R1
R1
R1
N
R2
O
R2
O
N
R2
O
N
v
+
N
O
O
N
+
O
N
R
5-7
2
3
Scheme 1. Synthetic routes to the novel series of muscarinic antagonists. Reagents and conditions. (i) NBS, 1.2 equiv, AIBN, 0.03 equiv, 500 W quartz light source, methyl
formate, reflux, 60%; (ii) HNR1R2, 2.2 equiv, acetonitrile, rt, 70–80%; (iii) Separation of enantiomers: Chiral HPLC, Chiracel OJ-H 4.6 Â 50 mm, 8:2 isohexane:ethanol; (iv) 1-
methyl-piperidin-4-ol, 2 equiv, NaH (60% disp), 2 equiv, toluene, reflux, 70–80%; (v) iodomethane, 1.1 equiv, acetonitrile, rt, 90–95%; (vi) (3R)-quinuclidinol-ol, 2 equiv, NaH
(60% disp), 2 equiv, toluene, reflux, 65–80%; (vii) R-Br, 1.1 equiv, acetonitrile, rt, 90–95%.
established that activity was greater for one enantiomer over the
ꢀ High human plasma protein binding (hPPB >98% bound) and
other with the more active enantiomer being of equivalent potency
to aclidinium bromide. Single X-ray structure determination of
compounds in this series established that the most active diaste-
reomer had the S-configuration at the aminoester and the R-config-
uration at the 3-quinuclidinyl groups.9 Thus this new series had
the potential to deliver compounds which matched the potency
of current late-stage development candidates and our subsequent
optimization program focused on identifying compounds which
met our desired TPP. The data shown in Table 4 is for the most ac-
tive diastereomer for compounds possessing a wide range of qua-
ternising groups.
high human liver microsomal intrinsic clearance (HM Cl_int
>50 ll/min/mg), to minimize systemic exposure and side effects
and hence differentiate our compound from tiotropium.
ꢀ Long in vivo pharmacokinetic (PK) properties as measured in
guinea pig intratracheal (IT) PK studies with an ITPK T >10 h,
½
being appropriate for assessment either as UID or BID treatment
in man.
To establish the effect on potency of the stereochemistry of the
quaternary centre at the a-position of the ester group the enantio-
mers of racemic ester 2c (Scheme 1) where separated by chiral
HPLC and the two ester enantiomers were converted into ana-
logues of aclidinium bromide, 5.1 and 5.2, (data in Table 4). This
Simple alkyl quaternising groups led to compounds 5.3–5.7
with moderate M3 potency but hPPB was well below our target va-
lue except for compound 5.7 (LogP = 2.6). This suggested that to at-

A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
7443
Table 4
Optimization of the quaternising group in the novel M3 antagonists
N
N
N
O
O
O
X
O
X
O
X
O
+
+
+
N
N
N
H
N
R
R
R
O
5
6
7
a
b
Compound
R group or compound name
X
M3 pIC50
GPT pA2
HM Cl_intc
(
l
l/min/mg)
hPPB (%) bound
LogPd
Glycopyrrolate
Tiotropium
Aclidinium
Ph-O-(CH2)3–
Ph-O-(CH2)3–
2-Propyl–
3-Pentyl–
Butyl–
3-Cyanopropyl–
2-Methyl-2-penten-5-yl
Ph-(CH2)2–
—
—
—
9.6
9.3
10.1
10.1
—
—
—
—
—
—
—
<3
<3
>145
87
—
81
61
61
—
56
85
>150
88
>150
123
>150
129
88
>150
>150
—
>150
98
59
98
—
66
59
104
86
45
-
81
>150
140
—
79
80
72
38
51
64
>150
72
100
46
56
77
145
66
72
78
102
45
75
62
79
34
56
66
29
48
—
94.4
—
0.4
>10.3
>10.3
10.3
8.5
9.0
6.6
8.3
8.1
9.2
—0.8
1.44
2.49
—
—
1.7
1.5
—
2.6
2.43
2.11
2.5
2.2
2.49
2.2
2.49
3.03
2.71
3.21
—
2.57
3.67
—
5.1 (S,R)
5.2 (R,R)
5.3
5.4
5.5
5.6
5.7
5.8
5.9
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
S
CH@CH
S
CH@CH
S
—
57.9
<10
—
91.5
97.6
99.0
99.2
99.8
93.1
99.1
98.1
96.5
97.8
97.4
—
98.9
99.7
—
99.6
—
99.8
89.3
—
91.4
<30
82.1
23.1
—
80.0
86.1
22.4
86.8
99.7
95.6
98.0
97.9
95.6
—
>10.0
>10.3
9.9
10.3
9.7
—
Ph-(CH2)2–
5.10
5.11
5.12
5.13
5.14
5.15
5.16
5.17
5.18
5.19
5.20
5.21
5.22
5.23
5.24
5.25
5.26
5.27
5.28
5.29
5.30
5.31
5.32
5.33
5.34
5.35
5.36
5.37
5.38
5.39
5.40
5.41
5.42
5.43
5.44
5.45
5.46
5.47
5.48
5.49
5.50
6.1
2-F-Ph-(CH2)2–
2-F-Ph-(CH2)2–
3-F-Ph-(CH2)2–
3-F-Ph-(CH2)2–
4-F-Ph-(CH2)2–
3-Cl-Ph-(CH2)2–
3-Cl-Ph-(CH2)2–
3-Br-Ph-(CH2)2–
5-Cl-Th-(CH2)2–
5-Me-Th-(CH2)2–
2-Naphthyl-(CH2)2–
3-Pyridyl-(CH2)2–
5-Tetrahydrobenzofuryl-(CH2)2–
2-Benzofuryl-(CH2)2–
3-Indolyl-(CH2)2–
Cyclohexyl-(CH2)2–
Ph-(CH2)3–
9.8
10.6
9.9
9.6
10.3
9.5
9.5
9.1
9.6
9.6
8.3
—
>10.4
>10.3
9.8
>10.5
>10.4
10.2
10.3
>10.2
9.7
8.5
9.8
8.2
8.8
CH@CH
CH@CH
S
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
S
9.8
—
—
2.36
—
2.36
2.48
2.24
2.54
1.07
0.75
1.53
—
1.79
1.53
0.01
2.37
4.75
2.36
2.80
2.48
2.56
—
9.7
8.8
9.3
9.8
10.7
10.5
10.5
-
9.7
9.5
—
9.1
8.5
9.6
9.2
9.2
—
—
—
—
—
—
9.6
—
9.6
—
—
S
>10.5
>10.3
10.1
9.8
9.9
8.6
9.5
8.9
8.5
9.5
10.2
9.4
9.7
9.6
9.2
8.5
8.9
9.5
9.0
8.3
10.2
9.2
10.1
9.1
9.3
9.7
9.9
9.7
9.7
Ph-(CH2)3–
CH@CH
CH@CH
S
4-Pyridyl-(CH2)3–
4-Pyridyl-(CH2)3–
2-Me-4-pyridyl-(CH2)3–
2,6-Di-Me-4-pyridyl-(CH2)3–
3-Pyridyl-(CH2)3–
6-Me-3-Pyridyl-(CH2)3–
5-Pyrimidyl-(CH2)3–
2-Benzoxazoyl-(CH2)3–
2-Benzothienyl-(CH2)3–
6-Benzothiazolyl-(CH2)3–
Ph-(CH2)4–
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
S
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
CH@CH
S
CH@CH
S
CH@CH
CH@CH
CH@CH
S
CH@CH
S
CH@CH
S
Ph-(CH2)4–
Ph-O-(CH2)2–
4-F-Ph-O-(CH2)2–
4-Cl-Ph-O-(CH2)2–
4-Me-Ph-O-(CH2)2–
3-Me-Ph-O-(CH2)2–
4-Pyridyl-O-(CH2)2–
Ph-O-(CH2)3–
4-F-Ph-O-(CH2)3–
4-F-Ph-O-(CH2)3–
Ph-CH2-O-(CH2)2–
Ph-(CH2)2-O-(CH2)2–
H
H
Ph
Ph
3-F-Ph
3-F-Ph
2-Pyridyl
4-Pyrimidinyl
3-Pyridazinyl
99.2
98.5
—
—
—
3.25
3.02
—
—
—
96.9
98.1
98.4
97.5
63.8
—
98.6
97.5
99.6
99.2
88.3
87.0
—
2.62
2.30
2.79
2.88
0.10
À0.22
2.42
2.10
2.95
2.63
1.79
1.37
—
9.9
10.5
9.4
10 4
9.1
9.5
—
6.2
6.3
6.4
6.5
6.6
6.7
6.8
6.9
9.7
10
9.9
8.9
CH@CH
CH@CH
CH@CH
—
—
37
79
9.0
(continued on next page)

7444
A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
Table 4 (continued)
a
b
Compound
R group or compound name
X
M3 pIC50
GPT pA2
HM Cl_intc
(l
l/min/mg)
hPPB (%) bound
LogPd
6.10
6.11
6.12
6.13
2-Pyrazinyl
CH@CH
CH@CH
S
S
9.7
9.8
9.7
10
9.5
60
58
>150
89
82.2
80.7
70.4
97.1
1.46
0.28
À0.04
2.18
3-Isoxazolyl
3-Isoxazolyl
2-Cl-pyrazin-6-yl
10.5
10.6
9.3
O
Me
7.1
CH@CH
9.8
8.5
>150
93.0
3.62
N
7.2
CH@CH
9.7
8.7
>150
98
2.69
O
N
N
N
7.3
7.4
CH@CH
9.5
9.4
9.6
8.9
9.8
9.3
53
97.6
97.7
97.8
3.61
1.61
2.97
O
O
S
S
147
32
N
O
N
7.5
N
a
Binding affinities measured using recombinant human M3 receptors (n P 2).
Antagonist potency at the M3 receptor measured in guinea pig trachea (n = 2).
Intrinsic clearance in human liver microsomes.
b
c
d
LogP was calculated using ACD LogP software.
tain hPPB values >98% we needed LogP values in the 2.5–3.0 range.
To achieve this we targeted compounds which carried an aryl
group at the end of an alkyl chain of variable length. A range of
aryl–ethyl compounds, 5.8–5.25, were synthesized and examples
were made in which Ar was both phenyl and thienyl, Table 4. In
general these compounds had good M3 potency and hPPB values
in the required range. There was no difference in potency for ana-
logues where Ar was phenyl or thienyl. Most compounds also had
the desired high intrinsic clearance in human liver microsomes
the right range and no drop-off in potency was observed in the gui-
nea pig trachea assay. However, only a few of these compounds
had high enough potency at the M3 receptor for further in vivo
evaluation (Table 5).
Another linker with promising activity was ’acetamide’ which
was evaluated for compounds in which Ar was both phenyl and
thienyl and key compounds, 6.1–6.13, and data are summarized
in Table 4. These derivatives tended to be inherently very polar,
making it difficult to achieve high hPPB and good potency in the
same molecule. When high LogP compounds were made to raise
hPPB, drop-off in potency in the guinea pig trachea assay was also
seen (e.g., 6.5–6.6). We investigated the use of isosteres to replace
the ‘acetamide’ to try and achieve higher hPPB whilst retaining
good potency. A range of five-membered rings, primarily heterocy-
cles, were thus made and evaluated 7.1–7.5. The amide isosteres
did achieve the desired higher hPPB compared to the equivalent
’acetamide’ analogues, however potency rarely approached the re-
quired levels of pIC50 ꢁ10, to warrant further evaluation.
(HM Cl_int >50 ll/min/mg). The most potent compounds of this
type bore a substituted phenyl or thienyl ring at the end of the
chain and the preferred ring substituents were small lipophilic
groups such as halogens or methyl. A number of these derivatives
met our in vitro TPP criteria and were profiled in further detail in
the guinea pig IT PK,11 bronchoconstriction model12 and salivation
model13 (see below Table 5).
A range of arylpropyl and arylbutyl derivatives, 5.26–5.39, were
also synthesized and evaluated and their data is summarized in
Table 4. A notable feature in some of these compounds was a
drop-off in potency in the guinea pig trachea tissue assay when
compared to the activity at the isolated M3 receptor. The most
extreme example was 5.36, which had a high LogP (3.75). We
investigated whether high lipophilicity was the cause of functional
drop-off by incorporating heterocycles at the end of the alkyl chain
to afford compounds of lower LogP such as 5.28–5.34. Although
good M3 potency with no drop-off in the trachea assay was
obtained, the hPPB was well below our desired value of 98%.
Getting the balance right proved challenging in these types of com-
pounds. However, several matched our pre-set criteria and were
progressed further for in vivo profiling (Table 5).
Compounds from Table 4 which met our in vitro target product
profile were taken into sequential evaluation in in vivo guinea-pig
models and data is summarized in Table 5. Our evaluation se-
quence started with guinea pig ITPK studies to determine com-
pounds’ lung half-life (gp-ITPK T ).11 Efficacy and duration of
½
action were determined using a guinea pig bronchoconstriction
(BC) model. Methacholine was administered intravenously to
anaesthetised and ventilated guinea pigs under terminal anaesthe-
sia. This bronchoconstrictor response is mediated through activa-
tion of M3 receptors in the airway smooth muscle. To test
duration of action at 12, 18 or 24 h timepoints, compounds were
administered IT at an ED80 dose (i.e., dose which produced ꢁ80%
inhibition of methacholine induced BC at 2 h).12 The ED80 dose
was either measured or predicted from the guinea pig trachea po-
tency data. Only compounds with duration of action beyond 12 h
A number of derivatives in which the alkyl chain linking the aryl
ring to the quinuclidinyl group contained an oxygen atom were
also made (5.1 and 5.40–5.50). This modification allowed us to
modify the compound physical properties so that hPPB was in
at an ED80 dose of 63 lg/Kg were further evaluated. Finally a gui-

A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
7445
Table 5
In vivo activity of the novel series of muscarinic antagonists
Compound
M3 pIC50
M2 pIC50
gp-IT PK T
h
BC ED80
(l
g/Kg) at 2 h
BC % inhibition at 24 h
Salivation EC50
(
lg/Kg)
TI
½
Tiotropium
5.8
10.3
>10
9.8
>10.4
10.1
9.5
10.2
9.9
10
10.2
—
9.0
—
—
—
—
—
—
—
8
8.8
10.5
8.1
26
26.3
21
20.2
16.9
18.0
0.3a
1
>80
<0.3
>10
>100
>30
nd
>30
nd
nd
<3
17
>10
>100
>10
—
>10
—
—
<10
—
5.14
5.16
5.28
5.32
5.46
6.2
1a
3a
1
50** (at 18 h)
74**
0
3
3
30*
28*
41**
59**
10
1
6.6
6.12
3a
1
<30
nd
9.8
a
Measured ED80 dose.
*
P <0.05.
**
P <0.01 compared to control group.
3. (a) Casarosa, P.; Bouyssou, T.; Germeyer, S.; Schnapp, A.; Gantner, F.; Pieper, M.
J. Pharmacol. Exp. Ther. 2009, 330, 660; (b) Gross, N. J. Chest 2004, 126, 1946;
ZuWallack, A. R.; ZuWallack, R. L. Expert Opin. Pharmacol. 2004, 5, 1827.
4. (a) Hansel, T. T.; Neighbour, H.; Erin, E. M.; Tan, A. J.; Tennant, R. C.; Maus, J. G.;
Barnes, P. J. Chest 2005, 128, 1974; (b) Villetti, G.; Bergamaschi, M.; Bassani, F.;
Bolzoni, P. T.; Harrison, S.; Gigli, P. M.; Geppetti, A. J. P.; Civelli, M.; Patacchini,
R. Br. J. Pharmacol. 2006, 148, 291.
nea pig pilocarpine-induced salivation model was used to assess
the peripheral side effects of compounds at a 4 h timepoint.13 Data
from the BC and salivation models were used to establish a thera-
peutic index (TI) for bronchoprotection over salivation effects in
the guinea pig. Our objective was to identify compounds with a
TI >10 as this would be a significant improvement over tiotropium
which has a narrow therapeutic window (TI <3) in these models.
The complete set of in vivo data for leading compounds is summa-
rized in Table 5 together with data generated for tiotropium in the
same assays for comparison.
5. Gavalda, A.; Miralpeix, M.; Ramos, I.; Otal, R.; Carreno, C.; Vinals, M.;
Domenech, T.; Carcasona, C.; Reyes, B.; Vilella, D.; Gras, J.; Cortijo, J.;
Morcillo, E.; Llenas, J.; Ryder, H.; Beleta, J. J. Pharmacol. Exp. Ther. 2009, 331, 740.
6. In vitro M3 and M2 radioligand binding assay: the antagonist affinity of a
compound at the human M3 or M2 receptors was estimated from its ability to
compete with specific [3H]NMS binding to membranes from CHO-K1 cells
expressing either recombinant human M3 or M2 receptors as measured using a
SPA assay format. Concentration-effect curves for each compound
(0.003À100 nM final concentrations) were constructed using serial dilutions
in ½-log unit intervals. For each assay plate, eight replicates were obtained for
[3H]NMS binding in the absence of test compound. Non-specific binding of
The table shows that in this novel series we have identified a
group of compounds with good in vivo efficacy in the guinea pig
BC model that exhibit ED80 of between 1–3 lg/Kg. Duration of action
beyond 12 h was statisticallysignificant for a numberof compounds,
5.14, 5.16, 5.32, 5.46, 6.2 and 6.6. In accord with expectations from
our design criteria, most compounds in this series have an improved
salivation TI compared to tiotropium, with 5.14 standing out with a
TI >100. In addition, compound 5.14 exhibited ca. 10-fold selectivity
for M3 over M2. Taken together with its high hPPB and high HM
Cl_int, the data highlight compound 5.14 as having a promising
and differentiated profile relative to tiotropium and was selected
for progression into studies in Man for evaluation of its potential
as a treatment for COPD as AZD9164. The results of these studies will
be reported in due course.
In summary, we describe a strategy of starting from known M3
antagonists and modifying their chemical structures to arrive at a
novel, potent series. The design and screening strategies used were
tailored to identify optimized compounds which met our pre-set
target product profile criteria for in vivo PK, hPPB, human liver
microsomal intrinsic clearance and M2 selectivity. These character-
istics were selected to differentiate the novel compounds from cur-
rently used antagonists (e.g., tiotropium). This led to compounds
which combined an improved systemic side effect profile, as mea-
sured by a pilocarpine induced salivation model, with good dura-
tion of action.
[3H]NMS was determined by replacing test compound with atropine 1
[3H]NMS was used at
concentration of 0.1 nM, which was below the
lM.
a
determined dissociation constant (Kd) for [3H]NMS, and about 10% of the
radioactivity was specifically bound to the membranes. The assay mixture was
incubated at room temperature for at least 16 h before counting to allow
[3H]NMS binding with muscarinic receptors to reach equilibrium; binding of
[3H]NMS was also reversible. Test compound inhibition was expressed as
percent inhibition relative to the specific radioligand binding for the plate.
7. Bailey, A.; Donald, D. WO2006112778, 2006, Chem. Abstr. 2006, 145, 454929.
8. Bailey, A.; Mete, A.; Pairaudeau, G.; Stocks, M.; Wenlock, M. WO2007123465,
2007, Chem. Abstr. 2007, 147, 502240.
9. (a) Ford, R.; Mete, A.; Millichip, I.; Teobald, B. WO2008075005, 2008, Chem.
Abstr. 2008, 149, 104901; (b) Ford, R.; Mete, A.; Millichip, I.; Teobald, B.;
Kinchin, E.C. WO2009153536, 2009, Chem. Abstr. 2009, 152, 75267. Synthetic
conditions and yields for all new compounds described herein are reported in
these publications. New compounds were identified and characterized by their
1H NMR and mass spectra and chiral HPLC and elemental analyses.
10. In vitro guinea-pig trachea assay: The antagonist potency of compounds was
measured by the ability of compounds to inhibit methacholine-induced airway
smooth muscle contraction using a functional, guinea-pig trachea organ bath
assay. Dunkin Hartley guinea-pigs were sacrificed by cervical dislocation and
the trachea removed. Tracheal rings were prepared and then suspended in
10 mL organ baths containing
a modified Krebs solution (containing
Indomethacin at 2.8 M final concentration), gassed with 5% CO2, 95% O2 at
l
37 °C and tensioned to 1 g. The tracheal rings were left to equilibrate for 1 h,
washed and re-tensioned to 1 g. The rings were ‘primed’ with methacholine
(1 lM), washed and left for a further 1 h. Then a cumulative methacholine
concentration response curve was constructed (3 Â 10À9–3 Â 10À5 M). Vehicle
or test compound was then added to the baths and allowed to equilibrate for
1 h. Then, a second extended cumulative concentration response curve was
constructed (3 Â 10À9–1 Â 10À3 M final concentration). Changes in isometric
force were recorded and results were expressed as % of maximum response in
curve 1 for each individual tissue. A50 values were generated using calculated %
increase in tension at each compound concentration. The A50 values for the
vehicle control concentration effect curve and test concentration effect curve
were determined and then used to calculate a potency pA2 value at each
compound concentration used.
Acknowledgments
The authors would like to thank Elizabeth Kinchin, Andrew
Mather, Sasvinder Sidhu, Phil West, Hemant Mistry, Jaspal Singh
and James Bird for technical contributions to the work.
11. For pharmacokinetic studies, compounds were dosed to halothane
References and notes
anaesthetised guinea pigs by intratracheal instillation at
1 lg/kg. For
terminal lung collection at 0.25, 0.5, 0.75, 1, 2, 4, 7, 12, 18, 24 h post-dosing,
the animals were euthanised by an intravenous overdose of barbiturate
(Euthatal 20 mL/kg) and the lungs removed and homogenized with phosphate
buffer and added to ice cold methanol. Aliquots of the supernatant were
analyzed by mass spectrometry/HPLC methodologies.
1. (a) Moulton, B. C.; Allison, D.; Fryer, A. D. Br. J. Pharmacol. 2011, 163, 44; (b)
Peretto, I.; Petrillo, P.; Imbimbo, B. P. Med. Res. Rev. 2009, 29, 867; (c) Laine, D. I.
Expert Rev. Clin. Pharmacol. 2010, 3, 43; (d) Cazzola, M.; Matera, M. G. Br. J.
Pharmacol. 2008, 155, 291; (e) Joos, G. F. Expert Opin. Invest. Drugs 2010, 19, 257.
2. (a) Caulfield, M. P.; Birdasall, N. J. M. Pharmacol. Rev. 1998, 50, 279; (b) Eglen, R.
M. Prog. Med. Chem. 2005, 43, 105.
12. In vivo guinea-pig bronchoconstriction: The efficacy of compounds was
measured in vivo by the inhibition of methacholine induced

7446
A. Mete et al. / Bioorg. Med. Chem. Lett. 21 (2011) 7440–7446
bronchoconstriction in Dunkin-Hartley guinea-pigs. Dose solutions of test
13. In vivo salivation: The ability of compounds to inhibit salivation was measured
in vivo by the reduction of pilocarpine induced salivation in guinea-pigs.
Compounds or vehicle were administered intranasally in halothane-
anaesthetised guinea-pigs. At 4 h after dosing, guinea-pigs were terminally
anaesthetised (urethane). Once anaesthesia was established a gauze pad was
inserted into the mouth for 5 min to dry the mouth of residual saliva. A pre-
weighed pad was then inserted into the mouth for 5 min and weighed for a
measurement of baseline saliva production. Pilocarpine (0.6 mg/kg at a dose
compounds were prepared in saline (0.9% (w/v) sodium chloride) and delivered
by intratracheal instillation using a Penn-Century microsprayer (Penn-Century
Inc., Wyndmoor, PA, USA) under halothane anaesthesia. At various time points
after dosing (2, 12, 18 or 24 h), guinea-pigs were terminally anaesthetised with
pentobarbitone, surgically prepared, ventilated (at 60 breaths/min at a tidal
volume of 2 mL) and lung function measured by changes in respiratory
resistance in response to the bronchoconstrictor agent methacholine (10 lg/
kg, intravenous bolus), using the Flexivent lung function system (EMMS, UK).
After each administration of methacholine, the peak resistance value was
obtained and respiratory mechanics returned to baseline in between
administrations of methacholine. At the end of the experiments animals
were euthanized. Percentage inhibition of bronchoconstriction was calculated
for each compound treated group in relation to the appropriate vehicle, time-
matched control group.
volume of 2 mL/kg) was then administered subcutaneously. A new pre-
weighed pad was inserted into the mouth immediately and replaced at 5
minute intervals for 15 min. The difference in weight of the pads pre and post
the 5 min sampling period was used to calculate the amount of saliva collected
over a 15 min period.