Discovery of N-{1-[3-(3-Oxo-2,3-dihydrobenzo[1,4]oxazin-4-yl)propyl] piperidin-4-yl}-2-phenylacetamide (Lu AE51090): An allosteric muscarinic M 1 receptor agonist with unprecedented selectivity and procognitive potential

Article abstract of DOI:10.1021/jm100697g

The discovery and Structure-activity relationship (SAR) of a series of allosteric muscarinic M₁ receptor agonists are described. Compound 17 (Lu AE51090) was identified as a representative compound from the series, based on its high selectivity as an agonist at the muscarinic M₁ receptor across a panel of muscarinic receptor subtypes. Furthermore, 17 displayed a high degree of selectivity when tested in a broad panel of G-protein-coupled receptors, ion channels, transporters, and enzymes, and 17 showed an acceptable pharmacokinetic profile and sufficient brain exposure in rodents in order to characterize the compound in vivo. Hence, in a rodent model of learning and memory, 17 reversed delay-induced natural forgetting, suggesting a procognitive potential of 17.

Full text of DOI:10.1021/jm100697g

6386 J. Med. Chem. 2010, 53, 6386–6397
DOI: 10.1021/jm100697g
Discovery of N-{1-[3-(3-Oxo-2,3-dihydrobenzo[1,4]oxazin-4-yl)propyl]piperidin-4-yl}-
2-phenylacetamide (Lu AE51090): An Allosteric Muscarinic M1 Receptor Agonist with
Unprecedented Selectivity and Procognitive Potential
Anette G. Sams,*,† Morten Hentzer,‡ Gitte K. Mikkelsen,† Krestian Larsen,† Christoffer Bundgaard,§ Niels Plath,
Claus T. Christoffersen,‡ and Benny Bang-Andersen†
†Medicinal Chemistry Research, ‡Molecular Pharmacology, §Discovery ADME, and In Vivo Neuropharmacology,
Lundbeck Research Denmark, H. Lundbeck A/S, Ottiliavej 9, DK-2500 Valby, Denmark
Received June 10, 2010
The discovery and structure-activity relationship (SAR) of a series of allosteric muscarinic M1 receptor
agonists are described. Compound 17 (Lu AE51090) was identified as a representative compound from
the series, based on its high selectivity as an agonist at the muscarinic M1 receptor across a panel of
muscarinic receptor subtypes. Furthermore, 17 displayed a high degree of selectivity when tested in a
broad panel of G-protein-coupled receptors, ion channels, transporters, and enzymes, and 17 showed an
acceptable pharmacokinetic profile and sufficient brain exposure in rodents in order to characterize the
compound in vivo. Hence, in a rodent model of learning and memory, 17 reversed delay-induced natural
forgetting, suggesting a procognitive potential of 17.
Introduction
from early stages and onward.2,3 A central role for the M1
receptor in cognition was first demonstrated in mouse M1
receptor knockout studies, where M1 -/- mice revealed
deficits in hippocampus and prefrontal cortex dependent
working and episodic memory tasks.4,5 In addition, M1
receptor expression is decreased in cortical regions of schizo-
phrenia patients, although this has not yet been linked
to a functional outcome.6,7 Hence, it has been suggested that
M1 receptor agonism has a role in the treatment of AD
and cognitive impairment associated with schizophrenia
(CIAS).8-11
During the 1990s, several pharmaceutical companies pur-
sued selective muscarininc M1 receptor agonists as potential
therapeutics for the treatment of AD and schizophrenia.
While many of the compounds were efficacious in preclinical
animal models of cognitive dysfunction, those compounds that
were progressed into clinical studies were dose limited by
various adverse effects, including vomiting, nausea, and sweat-
ing, which prohibited their use in clinical practice of AD
treatment.12 However, in clinical studies 113 (Chart 1) demon-
strated an effect on cognitive and behavioral symptoms in AD
patients14,15 and on measures of verbal learning and short-term
memory functions in schizophrenic patients.16 Many of the
first generation M1 receptor agonists were invented on the basis
of structures of naturally occurring, nonselective muscarinic
receptor agonists, such as arecholine.17,18 The moderate selec-
tivity of these compounds for M1 receptors over the other mus-
carinic receptor subtypes, especially M2 and M3 receptors,8,12,18
has been suggested to be the cause of the observed peripheral
cholinergic side effects.
Muscarinic (M) acetylcholine (ACha) receptors are mem-
bers of superfamily A of the G-protein-coupled receptors
(GPCRs) and mediate the action of the neurotransmitter
ACh in the peripheral and central nervous system (CNS).
Five muscarinic receptor subtypes (M1-M5) have been
cloned. The M1, M3, and M5 receptors couple to Gq/11, and
their stimulation leads to activation of phospholipase C and
increase in phosphoinositol (PI) turnover, resulting in intra-
cellular calcium release. The M2 and M4 receptors mediate
their signaling via Gi/o, and activation of these subtypes leads
to reduced cAMP production through inhibition of adenylate
cyclase activity.1
The M1 receptor is the predominant muscarinic receptor
subtype in the CNS and is located in brain areas such as
cerebral cortex and hippocampus. These brain areas are
important for learning and memory, and activation of M1
receptors has been suggested to mediate higher cognitive
processing. Moreover, these brain regions are severely af-
fected bycholinergic dysfunction inAlzheimer’s disease (AD),
*To whom correspondence should be addressed. Current address:
LEO Pharma A/S, Industriparken 55, DK-2750 Ballerup, Denmark.
Phone: þ45 44945888. Fax: þ45 72263320. E-mail: anette.sams@
leo-pharma.com
a Abbreviations: SAR, structure-activity relationship; AD, Alzheimer’s
disease; CNS, central nervous system; ADME, absorption, distribution,
metabolism, excretion; AUC, area under the curve; iv, intravenous; ITI,
intertrial interval; CHO, Chinese hamster ovary; BHK, baby hamster
kidney; PI, phosphoinisitol; GPCR, G-protein-coupled receptor;
ACh, acetylcholine; CCh, carbachol; SPA, scintillation proximity
assay; HEPES, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid;
EDTA, ethylenediaminetetraacetic acid; SCX, strong cation ex-
change; DMF, dimethylformamide; THF, tetrahydrofuran; DIPEA,
diisopropylethylamine; TFA, trifluoroacetic acid; CDI, carbonyl
diimidazole; MTBD, 7-methyl-1,5,7-triazabicyclo[4.4.0]dec-5-ene;
LC/MS-MS, liquid chromatography-tandem mass spectrometry;
ELSD, evaporative light scattering detection.
The modest selectivity of the first generation M1 receptor
agonists has been suggested to be a direct consequence of
these compounds activating the M1 receptor through ortho-
steric regions, which display a high level of sequence homo-
logy among the subtypes of the muscarinic receptor family.
r
pubs.acs.org/jmc
Published on Web 08/04/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6387
Spalding et al. reported the discovery of an allosteric mus-
carinic M1 receptor agonist 2 (AC-42)19 belonging to a novel
structural class of muscarinic receptor agonists. It was shown
that 2 displayed complete selectivity as an agonist of the
M1 receptor versus the other muscarinic receptor subtypes.
Mutagenesis and pharmacological studies later revealed that
this compound activates the muscarinic M1 receptor with a
mode of action different from that of orthosteric muscarinic
receptor agonists.20,21
The discovery of allosteric M1 receptor agonists spurred a
renewed interest in the muscarinic field, and a number of
allosteric agonists of the M1 receptor have since appeared.
These second generation muscarinic M1 receptor agonists
have a potential for subtype selectivety, since they activate
the receptor through interaction with less conserved regions
among the muscarinic receptor subtypes. Subsequent to the
discovery of 2, other compounds have been reported as
allosteric agonists of the muscarinic M1 receptor, such as 3
(NDMC),22 a metabolite of the atypical antipsychotic drug
clozapine, and more recently 4 (TBPB)23 was reported.
A small screening campaign was conducted in our labora-
tory to identify allosteric muscarinic M1 receptor agonists
using a functional Ca2þ flux assay. Here, we report the
discovery of 5 and a related analogue 6 and their development
into a novel series of allosteric M1 receptor agonists. From
this series, the allosteric M1 receptor partial agonist 17 (Lu
AE51090) was selected and further characterized. In short, 17
displayed unprecedented selectivity for the muscarinic M1
receptor versus other muscarinic receptor subtypes, when
assayed for functional agonism and in binding assays. Com-
pound 17 also displayed high selectivity when tested in a
panel of 69 receptors, ion channels, transporters, and
enzymes. Also, 17 had suitable pharmacokinetic properties
and was active in an in vivo model of working memory,
making it a high quality lead compound for a drug discovery
program.
Chemistry
Chart 1. Structures of Muscarinic Agonists
A parallel synthesis protocol was developed in order to
explore the SAR of the two initial hits 5 and 6 in a rapid
manner, as outlined in Scheme 1. Commercially available tert-
butyl 4-aminopiperidine-1-carboxylate was acylated with an
excess of a suitable carboxylic acid chloride in the presence of
TBD-methylpolystyrene, a polymer immobilized version of
the organic base 7-methyl-1,5,7-triazabicyclo[4.4.0]dec-5-ene
(MTBD), to drive the acylation reaction to completion. The
residual carboxylic acid chloride was removed by the addition
of tris-(2-aminoethyl)amine polystyrene as a scavenger. The
resins were subsequently removed by filtration, and the
resulting crude product was deprotected by addition of tri-
fluoroacetic acid to the solution. The solvent was evaporated
and the residue redissolved in MeOH and loaded onto a
sulfonic acid based strong-cation-exchange (SCX) column.
The loaded SCX column was washed repeatedly with MeOH
and acetonitrile to remove nonbasic impurities, and the
adsorbed piperidin-4-ylcarboxamide was subsequently eluted
by rinsing the column with 4 M NH3 in MeOH. The fractions
containing the product were evaporated and redissolved in
DMF and finally alkylated with the intermediates 8a-l in the
presence of DIPEA. By this protocol, compounds 6, 9-16,
18-20, 22-32, 34, and 35 were obtained. The intermediate 8l
was commercially available, while intermediates 8a-k were
prepared as described in Scheme 2. Thus, suitably substituted
4H-benzo[1,4]oxazin-3-ones 7a,j or 3,4-dihydro-1H-quinolin-
2-ones 7b-i and 4H-benzo[1,4]thiazin-3-one 7k were depro-
tonated with NaH and alkylated with 1,3-dibromopropane or
1-bromo-3-chloropropane, as specified in Scheme 2. The
intermediates 7a,b,d-f,h,i,k were commercially available,
while 7c, 7g, and 7j were prepared as outlined in Scheme 3.
Scheme 1. Parallel Synthesis Protocol to Prepare Analogues of 5a
a (a) TBD-methylpolystyrene, 1,2-dichloroethane; (b) Tris-(2-aminoethyl)amine polystyrene resin; (c) TFA; (d) SCX resin; (e) NH3/MeOH (4 M);
(f) DIPEA, DMF.

6388 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17
Scheme 2. Preparation of Intermediates 8a-ka
Sams et al.
a (a) NaH, Br(CH2)3X, DMF.
Scheme 3. Preparation of Intermediates 7c, 7g, and 7ja
tional Ca2þ mobilization assays using stably transfected
Chinese hamster ovary (CHO-K1) cells, expressing the human
M1 or M2 receptor, or using baby hamster kidney (BHK-21)
cells, expressing the human M3, M4, or M5 receptor. The cell
lines expressing the human M2 or M4 receptor coexpressed
the promiscuous GR16 subunit. Agonist evoked increases in
intracellular Ca2þ was measured by fluorometric imaging.
Concentration-response data were fitted to the four-para-
meter logistic equation to estimate compound potency and
efficacy of test compounds.24
Binding affinities at human muscarinic M2-M5 receptors
were determined by tritium-labeled radioligand binding to
membrane preparations from cell lines expressing the respec-
tive human muscarinic receptors, using a scintillation proxi-
mity assay (SPA) and the radioligands [3H]AFDX-384 (M2)
and [3H]4-DAMP (M3-M5).
In Vitro and in Vivo ADME Evaluation
a (a) Diethyl malonate, NaH, DMF; (b) AcOH, HCl, reflux; (c) 10%
Pd/C, HCOONH4, MeOH, reflux; (d) 3-chloropropionyl chloride,
DIPEA, ethyl acetate; (e) AlCl3, 180 °C; (f) 5% Pd/C, H2, EtOH;
(g) chloroacetyl chloride, K2CO3, DMF.
A range of in vitro and in vivo assays were employed to
assess the ADME properties of the compounds and to predict
the pharmacokinetic properties in humans. The metabolic
stability was investigated in cryopreserved hepatocytes of rat
and human origin. The first-order elimination rate constant
for disappearance of parent compound was calculated from
the slope of the logarithm of the transformed concentration-
time curve. Plasma protein binding was determined in vitro
in rat and human plasma following incubation and ultra-
centrifugation. In vivo pharmacokinetics were studied in rats
following single intravenous (1 mg/kg) and oral (2 mg/kg)
administration of compounds. Serial blood samples were
collected at various time points up to 6 h after dosing. In vivo
pharmacokinetic parameters were obtained following com-
partmental modeling. The extent of brain penetration was
examined in rats and mice following single subcutaneous
administration. Animals were anesthetized at different time
points, and cardiac blood was obtained followed by decapi-
tation and homogenization of the brain. Bioanalysis of
samples from all in vitro and in vivo ADME assays was per-
formed by liquid chromatography-tandem mass spectrometry
(LC/MS-MS).
2-Fluoro-6-nitrobenzyl bromide was reacted with diethyl
malonate, followed by diethyl ester hydrolysis and subsequent
decarboxylation of one of the carboxy groups to give 39 in
79% yield. The nitro group of 39 was reduced to give the
corresponding aniline, which was subsequently cyclized to
give 7c in 96% yield. The intermediate 7g was prepared by
acylating 3,4-difluoroaniline with 3-chloropropionyl chloride
followed by a Friedel-Crafts alkylation to effect ring closure
in 37% overall yield. Intermediate 7j was obtained by reduc-
tion of 4,5-difluoro-2-nitrophenol to the corresponding ani-
line followed by reaction with chloroacetyl chloride in 70%
overall yield. A slight modification of the procedure outlined
in Scheme 1, with simple permutations of the individual
steps as described in the Experimental Section, was adopted
for the larger scale preparations of compounds 5, 17, 33, and
36-38.
Biological Evaluation
Prediction of hepatic clearance in humans was done by
in vitro/in vivo extrapolation from the predicted hepatic
clearance in rats and humans according to the well-stirred
model25 and the observed systemic in vivo clearance in
rats.26
Biological in vitro assays were employed to determine the
pharmacological activity of the compounds and to establish
a SAR. Agonist potency (EC50) and efficacy (Emax) at the
human muscarinic M1-M5 receptors was assayed in func-

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6389
Table 1. Functional Agonism at Muscarinic M1-M5 Receptorsa
hM1
hM2
Emax (%)
100
hM3
Emax (%)
hM4
Emax (%)
hM5
Emax (%)
100
compd EC50 (nM) Emax (%) EC50 (nM)
EC50 (nM)
3.2
EC50 (nM)
10
EC50 (nM)
1.2
ACh
1.1
61
100
83
89
87
75
86
91
86
220
100
1% at 10 μM
54
100
6% at 10 μM
72
17
1
0% at 10 μM
9% at 10 μM
1% at 10 μM
3% at 10 μM
nd
1% at 10 μM
60
1.6
150
18
120
6.2
140
2
6% at 10 μM
10% at 10 μM
1% at 10 μM
7% at 10 μM
0% at 10 μM
2% at 10 μM
12
4% at 10 μM
34
3
400
300
170
4
6.5
13
nd
38% at 10 μM
48
5% at 10 μM
52
5
2% at 10 μM
0% at 10 μM
1200
6
130
3% at 10 μM
6% at 10 μM
a Data are the mean of three experiments. Percent activity at the indicated concentration is given when an EC50 could not be established. nd: not
determined.
Results and Discussion
fied with compound 22. Also, extending the linker between the
carbonyl and phenyl ring as in compound23 was unfavorable.
However, replacing the benzyl group of 17 with a methyl
group (compound 24) only led to a modest 3-fold drop in
potency compared to 17, and the corresponding ethyl analo-
gue almost restored the M1 receptor potency to the level of 5,
yet with significantly less molecular mass (ΔM = 156 Da)
(25). Other small alkyl substituents at the amide yielded
compounds that were less potent at the M1 receptor (25-31)
compared to 25, and R1 = isobutyl (27) or cyclohexyl (31)
gave compounds devoid of M1 receptor agonism. Interest-
ingly, while all other compounds in the series were partial
muscarinic M1 receptor agonists with relatively high efficacy
in the range of 70-90% relative to ACh, compounds 29 and
30, wherein R1 = cyclobutyl and cyclopentyl, respectively,
were partial M1 receptor agonists with significantly lower
A screening campaign covering approximately 10000 com-
pounds led to the discovery of the potent M1 receptor agonist
5 and its close analogue 6. The screen was performed using a
Ca2þ mobilization based functional assay with a stably trans-
fected Chinese hamster ovary (CHO-K1) cell line expressing
the human M1 receptor. In this assay, 5 had an EC50 of 13 nM
and an efficacy (Emax) of 91% relative to the endogenous
muscarinic agonist ACh while 6 was a partial agonist with
an EC50 of 130 nM and an Emax of 86% relative to ACh. The
reference compound ACh had an EC50 of 1.1 nM, and the
Emax was defined to 100%.
Hit compounds 5 and 6 were also tested for agonism at the
muscarinic M2-M5 receptors (Table 1). Compound 5 dis-
played no agonist activity at the M2 and M3 receptors, while it
was a moderately potent partial agonist at the M4 receptor
(EC50=300 nM, Emax=48%) as well as a weak partial agonist
at the M5 receptor (EC50 = 1200 nM, Emax = 52%). In con-
trast, compound 6 showed no agonist activity at any of the
M2-M5 receptors.
A series of analogues of 5 and 6 was prepared, and their
potencies and efficacies at the M1 receptor are summarized in
Table 2. The aim was initially to investigate the contribution
of the substituents of the highly decorated scaffold of 5 to the
M1 receptor activity of the compound. Surprisingly, removal
of the two benzylic methoxy substituents led to an equipotent
compound 9. Wesubsequentlyexaminedthe individual effects
of fluorine substituents at the aromatic positions of the
tetrahydroquinolinone ring system (10-13). The monofluori-
nated isomers were all less potent than 9. In particular,
compounds 10 and 13 with monofluoro substituents at the
5- or 8-position were approximately 10- to 20-fold weaker
when compared to 9, while compounds 11 and 12 with
monofluoro substituents at the 6- or 7-position were only
marginally weaker than 9. Replacing the fluoro substituents
of compounds 11 and 12 with methoxy substituents led to a
loss of activity, while the fully unsubstituted compound 16
was equipotent to 11 and 12. The corresponding benzoxazi-
none derivative 17 was equipotent to the tetrahydroquino-
linone derivative 16, while the benzothiazinone derived ana-
logue 18 was slightly less potent than 16 and the correspond-
ing benzoxazolone derivative 19 even less potent. Compound
17 was subsequently used as the reference for further SAR
work around the benzoxazolone scaffold. Thus, methylation
of the amide functionality resulted in a substantial loss of M1
receptor agonism (compound 20), and reversing the amide, as
in 21 (Figure 1), resulted in an inactive compound. We also
explored the effects of changing the amide substituent of 17,
and replacing the benzyl group of 17 with a phenyl substituent
completely removed M1 receptor agonist activity, as exempli-
Emax of approximately 40% relative to ACh.
We repeated the fluoro scan of the aromatic positions of the
tetrahydroquinolinone ring system, leading to the ethyl amide
derivatives 32-35. As was seen for the benzyl substituted
amide analogues, a fluorine substituent at the 8-position was
unfavorable as exemplified by compounds 35 and 13. How-
ever, introducing a fluorine atom at the 5-, 6-, or 7-position
yielded largely equipotent compounds, but in this case the
6,7-difluoro substitution pattern did not appear to give a
synergistic effect incontrast to whatwas seen with the benzylic
amide substituent (34-36 versus 9, 11, 12). Compounds
32-34, 36, and 38 were among the most potent compounds
of the series, displaying EC50 values between 10 and 20 nM
corresponding to the level of compound 5. Also, in this case,
the corresponding methylanalogue37was less potentthanthe
ethyl analogue 36. In analogy with the benzylamide series, the
ethylamide derivative of the 6,7-difluoro substituted benzox-
azinone based scaffold (38) was equipotent to the tetrahydro-
quinolinone-based analogue 36.
The hit compound 5 displayed a mixed agonistic profile
at the M1, M4, and M5 receptors; hence, we monitored the
selectivity of selected compounds of the series versus other
muscarinic receptor subtypes. It was discovered that the
agonistic effects at the M4 and M5 receptors observed for 5
were related to the 3,4-dimethoxy substitution pattern at the
benzylic amide substituent, and agonism at M2-M5 was
generally not observed in any of the compounds of the series
when this substitution pattern was avoided (data not shown).
Compound 17 was identified as a highly selective M1 recep-
tor agonist versus M2-M5 and was selected early on as a
prototype compound for in-depth profiling and as a bench-
mark to reference compounds. The fact that a compound does
not elicitan agonist response at a receptordoesnot exclude the
possibility that the compound may show antagonistic effects

6390 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17
Sams et al.
Table 2. Functional M1 Structure-Activity Relationship for a Series of Analogues of 5 and 6a
hM1
compd
R1
R2
R3
Y-Z
EC50 (nM)
Emax (%)
5
(3,4-di-OMe)benzyl
(3-OMe)benzyl
benzyl
6,7-di-F
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
CH2-CH2
O-CH2
13
91
86
89
77
79
86
75
84
83
89
83
79
70
67
6
130
15
9
6,7-di-F
5-F
6-F
7-F
8-F
6-OMe
7-OMe
H
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
O-CH2
10
11
12
13
14
15
16
17
18
19
20
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
benzyl
benzyl
150
49
benzyl
benzyl
51
290
560
120
48
benzyl
benzyl
benzyl
benzyl
H
61
benzyl
benzyl
H
S-CH2
O
100
1100
910
nd
H
benzyl
phenyl
H
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
O-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
CH2-CH2
O-CH2
H
8% at 10 μM
60
(2-phenyl)ethyl
methyl
ethyl
H
970
180
27
H
74
78
H
isopropyl
isobutyl
cyclopropyl
cyclobutyl
cyclopentyl
cyclohexyl
ethyl
H
46
nd
72
21% at 10 μM
74
H
H
110
160
350
nd
H
37
41
H
H
3% at 10 μM
5-F
6-F
7-F
8-F
6,7-di-F
6,7-di-F
6,7-di-F
19
90
83
83
78
87
76
88
ethyl
ethyl
9
20
ethyl
ethyl
310
20
methyl
ethyl
86
10
a Data are the mean of three experiments. When an EC50 could not be established, activity is given as the percent activity at the concentration
indicated. nd: not determined.
Table 3. Binding Affinities at Muscarinic M2-M5 Receptorsa
Ki (nM)
compd
hM2
hM3
hM4
hM5
atropine
1.9
0.9
3.0
1.0
17
1
2200
120
980
250
320
7000
13
6900
72
8900
80
2
2800
90
1400
290
130
4500
51
3
4
430
940
a Data are means of three experiments.
Figure 1. Inverted amide analogue of 17.
M3 and M5 receptors as well as binding affinity for the M2
receptor. Indeed, it has been hypothesized that the M4 recep-
tor agonism of this compound is responsible for its effects
on positive symptoms of schizophrenia.27 Furthermore, 3
showed partial agonism at the M5 receptor and binding affi-
nity for several muscarinic receptors, suggesting that 3 has a
complex muscarinic pharmacology with mixed agonism and
antagonism at different receptor subtypes. Compound 2 only
displayed agonistic activity at the M1 receptor and weak
binding affinity for the M2-M5 receptors, Recently, 4 was
disclosed as a potent and highly subtype selective allosteric M1
at the receptor, for instance, by binding to inactive receptor
conformations. Therefore, we studied the binding of com-
pound 17 to M2-M5 receptors (Table 3) using orthosteric
antagonist radioligands and benchmarked to reference com-
pounds. Among the compounds tested, 17 displayed the least
binding to the muscarinic receptor subtypes M2-M5, making
it uniquely selective for the M1 receptor. In comparison, 1, a
classical orthosteric M1 receptor agonist, caused partial acti-
vation of the M4 receptor and, to a lesser extent, activation of

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6391
receptor agonist.28 In our hands, 4 showed no agonistic effects
at other muscarinic receptors than the M1, but the compound
did exhibitmoderatelypotent bindingaffinities atthe M2-M4
receptors (Table 3).
Characterization of Mode of Action of Compound 17 at the
M1 Receptor. Ca2þ mobilization is a distal event in the
signaling cascade triggered by M1 receptor activation, and
several signaling pathways converge at the point of calcium
release. Therefore, we wanted to validate that the observed
compound effects were specifically mediated by activation of
the muscarinic M1 receptor. Compound 17 did not elicit a
Ca2þ mobilization response when assayed on the parental
CHO cell line (i.e., not expressing the M1 receptor; data not
shown). The response of the M1 receptor expressing CHO
cell line to 17 was antagonized by atropine, a competitive
muscarinic antagonist, hence demonstrating that the response
17 was specifically mediated by the M1 receptor, as shown in
Figure 2.
Classical orthosteric agonists display no saturability in
the interaction with an orthosteric antagonist; i.e., the
dose-response curves can be right-shifted indefinitely by
increasing concentrations of the agonist and antagonist. In a
Schild regression plot, this will translate into a slope of unity
for orthosteric interaction. Conversely, the action of an
allosteric agonist is characterized by saturability of the
interaction with an orthosteric antagonist. This relationship
translates into a slope that deviates from unity.21,29 In such a
competitive antagonist assay, compound 17 behaved unlike
classical orthosteric agonists like ACh and carbachol (CCh)
but competed with atropine in a fashion consistent with an
allosteric mode of M1 receptor activation. The results for 17
were comparable to those obtained for the allosteric M1
receptor agonist 2 (Figure 3).
In order to further validate the allosteric mode of M1
receptor activation by 17, the compound was assayed for
activity at the Y381A mutant M1 receptor,19 in which the
essential tyrosine residue 381 located in the transmembrane
Figure 2. Concentration-dependent stimulation of Ca2þ mobiliza-
tion in CHO-hM1 cells by ACh (circles) and 17 (squares, solid line).
Antagonism of response to 17 by 100 nM atropine (squares, dashed
line). Data are the mean of three experiments ((SEM).
Figure 3. Detection of allosteric agonism by Schild regression: Schild regression for the interaction between the orthosteric antagonist
atropine and (A) the orthosteric agonist ACh, (B) the orthosteric agonist carbachol, (C) the allosteric agonist 2, and (D) compound. Data are
from a cell-based Ca2þ assay. The dashed lines represent a theoretical Schild regression of unit slope as expected for simple competitive
antagonism by atropine. In the experiments with acetylcholine and carbachol, the 95% confidence interval of the slope factor included the unit
slope, and the -pKb for atropine was in the range 9.0-9.6 and, hence, in agreement with literature values (www.iuphar-db.org). Shown are
representative data from three independent experiments.

6392 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17
Sams et al.
Figure 5. Plasma concentration-time plots of 17 following intra-
venous (1 mg/kg) and oral (2 mg/kg) administration in Sprague-
Dawley rats. Data shown are the mean ( SD (n = 3).
Figure 4. Concentration-dependent stimulation of Ca2þ mobili-
zation in CHO-K1 hM1-Y381A cells by acetylcholine (solid
circles; EC50 = 27000 nM, Emax = 100%), 17 (solid squares;
EC50 = 97 nM; Emax = 82%), 2 (solid triangles; EC50 = 500 nM,
Table 5. Main Pharmacokinetic Parameters of 17 after Intravenous
(1 mg/kg) and Oral (2 mg/kg) Administration in Ratsa
Emax = 31%), and 3 (open triangles, EC50 = 140 nM; Emax = 72%).
CLp ((L/h)/kg) t1/2, β (h) Vss (L/kg) Tmax (po) (h) F (po) (%)
Data are the mean of three experiments ((SEM).
5.0 ( 0.9
0.8 ( 0.06 4.8 ( 0.9
0.8 ( 0.1
4.0 ( 1.7
a Values shown are the average ( SD (n = 3).
Table 4. List of Target Affinities for Which a Ki < 1000 nM Was
Obtained from a Panel Screen of 17 and 2 on 69 Receptors, Ion
Channels, Transporters, and Enzymesa
displayed a broader pharmacological profile, with affinity to
several targets, such as R1, dopamine D2 and D4, and
serotonin 5-HT1A receptors. Thus, the pharmacological
profile of 2 at receptors other than the muscarinic receptor
family, in particular such as D2 but also R1 receptors, makes
it less suitable for use as a tool to study in vivo pharmaco-
logical effects of M1 receptor agonism in in vivo models of
schizophrenia. In addition to profiling across the muscar-
ininc M2-M5 receptors (Table 3), 4 was tested in a smaller
panel of nine GPCRs of the dopaminergic, histaminergic,
adrenergic, and serotonergic receptor families and found
Compound
Receptor (Ki; nM)
17
2
hR1A
hR1B
hR1D
hR2A
hR2B
hD2
260
910
18
65
40
530
840
77
hD4
hH1
11
780
610
280
to possess R1 and 5-HT2C receptor affinity (for R1A,B,D
,
5-HT1A
a Data are the mean of two determinations.
Ki = 11, 19, and 290 nM, respectively; for 5-HT2C, Ki = 85
nM). This also limits the usefulness of this compound as a
tool for exploring the effects of M1 receptor agonism in vivo.
ADME Properties of 17. Compound 17 was investigated
for its ADME properties in vitro and in vivo. Moderate
plasma protein binding was found in both rats (52%) and
humans (72%) following in vitro incubation studies. Intrin-
sic clearance in rat and human hepatocytes was determined
to be 100 mL/min and 3.4 L/min, which is above the hepatic
blood flow for both species (rat hepatic blood flow ∼20 mL/
min; human hepatic blood flow ∼1.4 L/min). The high
intrinsic clearance was in accordance with the observed in
vivo pharmacokinetics in rats following intravenous and oral
administration, where high systemic clearance and low oral
bioavailability (F) was encountered. The plasma concentration-
time profile is shown in Figure 5, and a summary of the
pharmacokinetic parameter estimates is provided in Table 5.
The ability of 17 to penetrate the blood-brain barrier was
verified from analysis of plasma and brain homogenate
samples after subcutaneous administration. From the AUCs,
brain-plasma distribution ratios of 0.35 and 0.20 were found
in mice and rats, and the data are shown in Figure 6.
domain 6 is mutated to alanine. The Y381 residue contri-
butes to the orthosteric binding site of the M1 receptor, and
mutation to alanine has been shown to decrease acetylcho-
line binding affinity 30-fold and drastically reduce its po-
tency about 3000-fold without affecting efficacy.30 The
critical contribution of Y381 to the orthosteric site of the
M1 receptor makes the Y381A M1 receptor mutant a useful
tool to detect allosteric agonists. We verified that the Y381A
M1 receptor mutant showed greatly reduced potency with
ACh (20000-fold reduction) and CCh, as shown in Figure 4.
To the contrary, the allosteric M1 receptor agonists 2 and 3
showed only minor changes in potency (2-fold reduction and
7-fold increase, respectively), which is in agreement with
published results.19,30 Compound 17 showed unaltered po-
tency and efficacy at the Y381A M1 receptor compared to the
wild-type M1 receptor. These data suggest that compound 17
interacts with the M1 receptor in a different mode than ACh
and, hence, is an allosteric M1 receptor agonist.
Broad Selectivity Profiling. Compounds 17 and 2 were the
two most selective allosteric M1 receptor agonists and were
tested in a broad pharmacology panel screen against 69
receptors, ion channels, transporters, and enzymes. The
targets for which 17 and 2 showed high to medium affinity
(Ki < 1000 nM) are listed in Table 4. Compound 17 only
displayed minor cross-reactivity with other targets, and the
most predominant effect was the affinity for the adrenergic
When the results from intrinsic clearance, unbound frac-
tions in plasma, and the in vivo clearance in rats were
compared, the in vitro/in vivo extrapolated intravenous
human hepatic clearance of 17 was predicted to be around
60 L/h (for a 70 kg subject) and hence below the human
hepatic blood flow around 90 L/h. In vitro and in vivo
evaluation of the series of analogues of 17 further suggests
the possibility of reducing the human predicted systemic
R1A receptor, at which 17 had a Ki of 260 nM. In contrast, 2

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6393
Figure 6. Plasma (ng/mL) and brain (ng/g) concentration-time courses of 17 following subcutaneous administration of 20 mg/kg (C57/6J
mice) and 10 mg/kg (Sprague-Dawley rats). Each data point represents an average of n = 3 ( SD.
clearance without compromising potency within this series
of compounds. For example, compound 38 was found to
have higher stability in human hepatocytes (1.2 L/min)
compared to 17, which in turn reduces the predicted human
clearance by approximately 2-fold.
Effect in an in Vivo Model of Cognition: Delayed Alterna-
tion Y-Maze. To examine the effect of 17 on learning and
memory, we tested the compound in the delayed alternation
Y-maze task in mice. This task is applied to assess working
memory and has previously been shown to be mediated by
the prefrontal cortex and the hippocampus across species.31-33
In the Y-maze version, mice enter the maze from a fixed start
arm and are forced to enter one goal arm to collect a reward
(“sample run”). During the subsequent run, mice can choose
between both goal arms (“choice run”); however, only the arm
not entered previously is rewarded. Positive alternation be-
tween the two goal arms requires recollection of the previous
arm entry, which is intact following a short time delay between
the sample and choice run. As shown in Figure 7, a high degree
of positive alternation (>80%), interpreted as intact working
Figure 7. Compound 17 reverses delay-induced working memory
decay in the mouse delayed alternation Y-maze task. The graph
shows the percentage of positive alternation of arm entries in
memory, was achieved following a short delay (0 s). However,
this memory was subject to a time-dependent memory decay,
as evident by reduced alternation rates at chance level after a
longer intertrail interval (ITI) of 60 s. Animals were treated
with 17 or vehicle 30 min prior to a series of 20 alternation
trials, applying both the short and long delay. Following the
short delay, vehicle treated animals as well as all dose groups
treated with 17 displayed high correct alternation rates
(80-90%, data not shown). As expected, this performance
was significantly lower and around chance levels in the vehicle
group, following a 60 s delay (55.6 ( 5.3%) (Figure 7). In
contrast, subjects treated with 17 showed a dose-dependent
reversal of the delay-induced memory decay. Whereas pre-
treatment with lower doses, 0.31 and 1.3 mg/kg 17, had no
significant effect on alternation rates, groups treated with 10
and 20 mg/kg showed significantly increased levels of positive
alternation compared to the vehicle treated group following
the 60 s delay. As shown in Figure 6, systemic treatment of
mice with 20 mg/kg 17 produced high brain exposures during
the experiment (440 ( 127 ng/g after 30 min (=pretreatment
time) and 240 ( 57 ng/g after 60 min (=end of experiment)).
Equilibrium dialysis revealed about 70% nonspecific binding
for compound 17 in mouse brain homogenates, which corre-
sponds to a free concentration of 70 ng/mL (∼170 nM) 60 min
after dosing. Free brain concentrations of 17 were about 3-fold
higher than the in vitro EC50 value (61 nM) during the
experiment, suggesting that the procognitive effect was
achieved at concentrations that are relevant for an interaction
with the M1 receptor.
successive trial runs following a delay of 1 (“0 sec”) or 60 s. A
two-way analysis of variance (ANOVA) with the all pairwise multi-
ple comparison procedure (Student-Newman-Keuls method) re-
vealed an overall difference for factor delay (F = 45.150; p < 0.001)
and treatment (F = 5.759; p < 0.001). The vehicle group (black
bars) displays a significant decline in correct alternation following
the 60 s versus 0 s delay, as well as the low compound 17 dose groups
(white bars) (veh, p < 0.001; 0.31 mg/kg, p < 0.005; 1.3 mg/kg, p <
0.002). Groups treated with 10 and 20 mg/kg 17 show a dose-
dependent, significant increase in correct alternation following the
60 s delay compared to vehicle (10 mg/kg, p < 0.003; 20 mg/kg, p <
0.001). No differences were found between all groups within the 0 s
delay (veh vs 0.31 mg/kg, p = 0.152; all others not tested). The
dotted line indicates chance performance level (/, compared to the
respective 0 s group; #, compared to the vehicle group at 60 s).
These data indicated that 17 dose-dependently reversed
the delay-induced working memory decay in the delayed
alternation Y-maze task in mice. Such procognitive effects
support a pivotal role of M1 receptors in learning and
memory and a procognitive potential of compound 17. A
full pharmacological profile of 17, including studies in
schizophrenia relevant cognition models using perturbed
animals, will be reported in a subsequent publication.
Conclusion
We have disclosed the discovery and SAR of a novel series
of allosteric M1 receptor agonists. The potent M1 receptor
agonist hit compound 5 displayed undesired agonistic activity

6394 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17
Sams et al.
at the M4 and M5 receptors and was used as template for SAR
investigations. It was revealed that equally potent and effica-
cious M1 receptor agonists compounds as 5 could be found
with much improved selectivity with respect to functional
agonismat othermuscarinicreceptorsubtypes. Compound17
was prioritized within this series as a prototype compound,
based on its selectivity for M1 receptor functional agonism.
It was demonstrated that 17 was an allosteric agonist at the
M1 receptor. In addition to its functional selectivity for the
muscarinic M1 receptor, 17 also had no binding affinity across
the muscarinic receptor subtypes M2-M5. Likewise, a very
limited cross-reactivity was observed toward a panel of 69
receptors, ion channels, transporters, and enzymes. Thus, 17
has an unprecedented overall selectivity for agonism at the M1
receptor, making it an excellent compound for investigating
the in vivo pharmacology of the M1 receptor. We tested the
compound in an in vivo model of learning and memory in
mice and observed a dose-dependent procognitive effect on
working memory. In vitro and in vivo characterization sup-
ported that 17 possesses acceptable pharmacokinetic proper-
ties. In addition, the potential for improving the metabolic
stability within the series was established. Taken together,
compound 17 is a high quality lead for a further drug dis-
covery program aimed at identifying allosteric muscarinic M1
receptor agonists. A detailed pharmacological profiling of this
compound will be reported in a later publication.
mL of DMF and 50 mL of THF. Diethyl malonate (109 mmol)
in 20 mL THF was added dropwise while maintaining the
temperature of the reaction mixture below 40 °C. Then the
mixture was stirred at ambient temperature for 2 h. 2-Fluoro-6-
nitrobenzyl bromide (107 mmol) in 40 mL of DMF and 30 mL of
THF was added dropwise while maintaining the temperature of
the reaction mixture below 50 °C. The mixture was heated at
refluxed for 3 h and then stirred at ambient temperature for 18 h
and poured onto brine. The reaction mixture was extracted with
ethyl acetate, and the combined organic extracts were dried and
evaporated under reduced pressure to yield a red oil. The oil was
dissolved in 100 mL of glacial acetic acid, 50 mL of concentrated
HCl, and 50 mL of H2O and refluxed for 6 h. Then the mixture
was stirred at room temperature for 18 h. The solvent was
evaporated and the residue triturated with H2O. The formed
precipitate was filtered off and dissolved in 2 M NaOH and
EtOH, each 400 mL. The mixture was refluxed for 40 min. Then
the solvent volume was reduced to 500 mL. Ice (300 g) and 4 M
HCl were added, to give acidic pH. The aqueous phase was
extracted with ethyl acetate, and the organic fractions were
combined, washed with brine, dried on MgSO4, and evaporated.
The crude was obtained as a dark solid and was used without
1
further purification. Crude yield: 79%. H NMR (DMSO-d6,
500 MHz): 2.55 (t, 2H); 3.04 (t, 2H); 7.52-7.65 (2H); 7.82
(d, 1H); 12.38 (br s, 1H).
5-Fluoro-3,4-dihydro-1H-quinolin-2-one (7c). Intermediate 39
(9.4 mmol) was dissolved in 125 mL of MeOH. Then 250 mg of
10% Pd/C and HCOONH4 (94 mmol) were added. The mixture
was refluxed for 45 min. Then the mixture was cooled and
filtered. The solvent was removed by evaporation. Toluene
(50 mL) was added and removed by evaporation. The residue
was partitioned between ethyl acetate and brine, and the aqu-
eous phase was extracted with ethyl acetate. The combined
organic fractions were washed with brine, dried, and filtered.
The solvent was removed by evaporation, and the product was
used without further purification. Crude yield: 96%. 1H NMR
(DMSO-d6, 500 MHz): 2.47 (t, 2H); 2.87 (t, 2H); 6.70 (d, 1H);
6.77 (m, 1H); 7.16 (m, 1H).
6,7-Difluoro-3,4-dihydro-1H-quinolin-2-one (7g). 3,4-Difluoro-
phenylamine (250 mmol) was dissolved in 500 mL of ethyl
acetate, and triethylamine (300 mmol) was added. 3-Chloropro-
pionyl chloride (275 mmol) was added dropwise while maintain-
ing the temperature below 30 °C. The mixture was stirred at room
temperature for 1 h, and then 250 mL of H2O was added. The
phases were separated, and the organic phase was washed with
brine, dried on MgSO4, and filtered. The solvent was evaporated
and the obtained intermediate (120 mmol) was mixed well with
aluminum trichloride (150 mmol) (caution: heating!) in a flask
equipped with a washing flask containing 2 M NaOH. The
resulting melt was heated at 180 °C for 1 h and then cooled and
poured into ice-water. To the mixture were added 25 mL of
concentrated HCl and 200 mL of dichloromethane. The obtained
solid was recovered by filtration and dried. The crude product
was recrystallized from ethyl acetate. Yield: 37%. 1H NMR
(DMSO-d6, 500 MHz): 2.44 (t, 2H); 2.85 (t, 2H); 6.82 (m, 1H);
7.30 (m, 1H); 10.14 (br s, 1H).
Experimental Section
General Chemistry. The 500 MHz 1H NMR spectra were
recorded on a Bruker Avance AV-III-500 spectrometer. The 600
MHz 1H NMR spectra were recorded on a Bruker Avance AV-
III-600 spectrometer. Chemical shift values are expressed in
ppm relative to Me4Si.
Preparative HPLC purification with MS detection was per-
formed on a Sciex API150ex preparative LC/MS system
equipped with a Gilson 333 pump (master) and a Gilson 334
pump (slave) and an Applied Biosystems API150ex single quad-
rupole mass spectrometer with atmospheric pressure photoioni-
zation (APPI) ion source. The purities of the intermediates and
final compounds were determined by integration of the UV
signal obtained by analytical HPLC-UV-MS on a Sciex
API150ex analytical LC/MS system equipped with 3 Shimatsu
LC10ADvp LC pumps, a Shimatsu SPD-M20A photodiode
array UV detector, and an Applied Biosystems API150ex
single quadrupole mass spectrometer with an APPI ion source
or by integration of the UV signal obtained by analytical
HPLC-UV-MS on a Sciex API300 analytical LC/MS system
equipped with 3 Shimatsu LC10ADvp LC pumps, an Acquity
UPLC PDA detector with an analytical flow cell, and an Applied
Biosystems API300 triple quadrupole mass spectrometer with an
APPI ion source.
Chromatography was performed on silica gel 60 (230-
400 mesh ASTM), Merck.
Strong cation exchange (SCX) chromatography columns
(Bond-Elut, 500 mg/3 mL) were obtained from Varian.
TBD-methylpolystyrene is a polymer-bound version of the
strong and highly hindered organic base 7-methyl-1,5,7-
triazabicyclo[4.4.0]dec-5-ene (MTBD). The resin was obtained
from Novabiochem, loading 2-3 mmol/g.
6,7-Difluoro-4H-benzo[1,4]oxazin-3-one (7j). 4,5-Difluoro-2-
nitrophenol (135 mmol) was dissolved in 250 mL of EtOH, and
2 g of 5% Pd/C was added. The mixture was hydrogenated at
room temperature for 3 h in a Parr apparatus under 3 bar of H2.
The catalyst was filtered off and the solvent removed by
evaporation. The residue was dissolved in 200 mL of DMF,
and chloracetyl chloride (209 mmol) was added with stirring.
The mixture was stirred at room temperature for 18 h. Then
K2CO3 (360 mmol) was added, and stirring was continued for
another 24 h. The solvent was evaporated, and the residue was
partitioned between H2O and ethyl acetate. The phases were
separated. The organic phase was dried on MgSO4, and the
solvent was removed by evaporation. Purification was by flash
chromatography on silica using 40% ethyl acetate in heptane
Reagents and solvents obtained from commercial sources
were used without additional purification.
Xanomeline (1),13 AC-42 (2),34 NDMC (3),22 and TBTP (4)34
were prepared according to the cited literature procedures.
The purity of compounds, measured as described above, was
>95% unless otherwise stated.
3-(2-Fluoro-6-nitrophenyl)propionic Acid (39). Sodium hy-
dride (60% oil dispersion) (112 mmol) was suspended in 110

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6395
1
as eluent. Yield: 70%. H NMR (DMSO-d6, 500 MHz): 4.59
(s, 2H); 6.70 (dd, 1H); 7.15 (dd, 1H); 10.77 (br s, 1H).
reader (Hamamatsu Photonics) using 480 nm excitation light,
and emitted fluorescent light passed through a 525 nm emission
filter and was detected by a CCD camera. Test compounds were
diluted in assay buffer from 2 mM stock solutions in 100%
DMSO to give a 3ꢀ concentrated stock. Compounds were added
to cells and fluorescence was measured at 1 Hz for 6 min starting
just prior to compound addition. The fluorescence readout was
calculated as max-min response, i.e., maximum fluorescence
reading after and before liquid addition. The fluorescence
max-min data were normalized to yield responses for no
stimulation (buffer) and full stimulation (1 μM acetylcholine)
of 0% and 100% stimulation, respectively. Concentration-
response data were fitted to the four-parameter logistic equation
to estimate compound potency (EC50) and efficacy (Emax) of test
compounds.24 Schild regression experiments were performed by
preincubating dye-loaded cells with increasing concentrations
of the muscarinic antagonist atropine for 30 min, before the
agonist response was assayed as described above. Schild regres-
sion analysis was performed by linear fitting to the following
model: log(dose ratio - 1) = log([antagonist]) - log(Kd). If the
95% confidence interval of the slope included the value 1.0, the
slope was estimated to unity and, hence, the agonist-antagonist
interaction was defined as competitive.
Cell Membrane Preparation. Cells were grown in the appro-
priate medium to approximately 90% confluence. Cells were
washed twice with 10 mL of PBS without Ca2þ, Mg2þ, and
HCO3-. Cells were harvested in ice cold 20 mM HEPES/10 mM
EDTA, pH 7.4, buffer and centrifuged at 3000g for 5 min. Cell
pellets were frozen at -80 °C. The thawed pellet was resuspended
in 20 mM HEPES/10 mM EDTA buffer, homogenized for 30 s
with Ultra-Turrax T25 (IKA-Werke GmbH), and centrifuged
at 1000g in 15 min at 4 °C. The supernatant was transferred to a
new vial and centrifuged at 40000g for 30 min at 4 °C. The pellet
was resuspended in 20 mM HEPES/0.1 mM EDTA, pH 7.4,
and centrifuged at 40000g at 4 °C for 30 min. The pellet
was resuspended in 20 mM HEPES/0.1 mM EDTA buffer/
250 mM sucrose, and 1 mL aliquots were stored at -80 °C.
Protein concentration was determined by BCA protein assay
kit (Pierce).
4-(3-Bromopropyl)-4H-benzo[1,4]oxazin-3-one (8a). To a sus-
pension of sodium hydride (60% in oil dispersion) (187 mmol) in
75 mL of DMF was added 4H-benzo[1,4]oxazin-3-one (170
mmol) in 100 mL of DMF at 10 °C. The mixture was stirred
at room temperature for ∼90 min and then poured into a ice-
cooled solution of 1,3-dibromopropane in 100 mL of DMF with
vigorous stirring. The mixture was stirred for ∼15 min and then
poured into 800 mL of brine. The aqueous phase was extracted
with ethyl acetate, and the combined organic extracts were dried
on MgSO4. After filtration the solvent was removed under
reduced pressure and the crude product was purified by flash
chromatography on silica using an eluent of 35% ethyl acetate in
heptane. Yield: 62%. 1H NMR (DMSO-d6, 500 MHz): 2.10
(qui, 2H); 3.59 (t, 2H); 4.02 (t, 2H); 4.65 (s, 2H); 6.99-7.10 (m,
3H); 7.23 (d, 1H).
N-{1-[3-(3-Oxo-2,3-dihydrobenzo[1,4]oxazin-4-yl)propyl]-
piperidin-4-yl}-2-phenylacetamide (17). Piperidin-4-ylcarbamic
acid tert-butyl ester (18 mmol) and 8a (18 mmol) were dissolved
in 50 mL of MeCN, and K2CO3 (20 mmol) was added. The
reaction mixture was stirred at 80 °C for 18 h, then cooled and
poured into H2O. The aqueous phase was extracted with ethyl
acetate, and the combined organic extracts were dried on MgSO4
and filtered. The solvent was removed under reduced pressure to
yield an oil which was purified by flash chromatography on silica
using 50% ethyl acetate in heptane as eluent. The obtained crude
was dissolved indichloromethane, and 15 mL of TFA/H2O (95:5)
was added. After 20 min, more H2O was added and to the
solution was added 2 M NaOH to give a pH of ∼12. The aqueous
phase was extracted with ethyl acetate, the organic fractions were
combined and dried on MgSO4, and the solvent was removed by
evaporation to yield the intermediate 4-[3-(4-aminopiperidin-
1-yl)propyl]-4H-benzo[1,4]oxazin-3-one as an oil. Yield: 50%.
1H NMR (DMSO-d6, 500 MHz): 1.36 (m, 2H); 1.66 (m, 2H); 1.75
(m, 2H); 1.88 (t, 2H); 2.29 (t, 2H); 2.72 (m, 1H); 2.79 (m, 2H); 2.84
(t, 2H); 3.44 (m, 2H); 3.88 (t, 2H); 6.99 (t, 1H); 7.14-7.28 (3H).
The obtained intermediate (17 mmol) was dissolved in 60 mL
of dry THF, and triethylamine (26 mmol) was added. The
mixture was cooled to 5 °C, and phenylacetyl chloride (19 mmol)
was added dropwise. The mixture was stirred at room tempera-
ture for 1 h and then filtered, and the solvent was removed in
vacuo. The crude product was purified by flash chromatography
on silica with gradient elution from 80% ethyl acetate in heptane
(þ5 vol % triethylamine) to 5% EtOH in ethyl acetate (þ5 vol %
Radioligand Binding Assays. Binding studies were performed
by incubating test compound (25 μL), cell membrane prepara-
tion (25 μL, typically 5 μg of protein/well) with radioligand
(25 μL), and 25 μL of wheat germ agglutinin SPA beads
(0.25 mg/well, GE Healthcare) in a white Wallac 96-well isoplate
with clear bottom (PerkinElmer). The reagents were mixed for
90 min on an orbital shaker and then centrifuged at 1500 rpm
for 5 min at room temperature. Plates were counted in a
Wallac TriLux 1450 Microbeta for 2 min/well. Assay buffer
was 12.5 mM sodium phosphate buffer with 2 mM NiCl2,
2.5 mM MgCl2, pH 7.4. The total binding, which comprised
less than 5% of added radioligand, was defined by assay buffer,
whereas the nonspecific binding was defined in the presence of
1 μM atropine. The nonspecific binding constituted ∼5% of the
total binding. The following radioligands were used: M2, 3.0 nM
1
triethylamine). Yield: 64%. H NMR (DMSO-d6, 500 MHz):
1.39 (m, 2H); 1.67-1.73 (4H); 1.91 (t, 2H); 2.30 (t, 2H); 2.75
(m, 2H); 2.37 (s, 2H); 3.48 (m, 1H); 3.92 (t, 2H); 4.62 (s, 2H); 7.01
(d, 2H); 7.06 (m, 1H); 7.18-7.31 (5H); 7.99 (d, 1H).
Biological Methods. Cell Lines. Standard molecular cloning
techniques35 were used to generate the following cell lines: Chinese
hamster ovary (CHO-K1) cells expressing either human muscari-
nic M1 or M2 receptor and baby hamster kidney cells (BHK-21)
expressing human muscarinic M3, M4, M5, or M1 Y381A recep-
tors. Cell lines expressing M2 and M4 receptors coexpressed the
promiscuous GR16 subunit. The cell line was grown in DMEM or
F-12 Kaighn’s medium with L-glutamine (Gibco), 10% Fetal-
Clone I serum (HyClone), and 1% penicillin and streptavidine
and supplemented with appropriate antibiotics.
Calcium Mobilization Assays. Cells expressing muscarinic
receptor were plated in growth medium at a density of 10000
cells/well in clear-bottomed, poly-D-lysine coated 384-well plates
(ArcticWhite LLC) and grown for 24 h at 37 °C in the presence of
5% CO2. Before assay, the cells were washed with assay buffer
(Hanks’ balanced salt solution with Ca2þ and Mg2þ (Gibco)
containing 20 mM HEPES, pH 7.4). The cells were incubated
with a calcium-sensitive fluorescent dye Calcium4 (Molecular
Devices Inc.) with2.5mM Probenecid(Sigma) for 50minat37 °C
followed by 10 min at room temperature. Calcium flux was
measured using a Hamamatsu FDSS7000 imaging-based plate
3
3H-AFDX-384 (PerkinElmer); M3-M5, 0.3 nM H-4-DAMP
(PerkinElmer). Data points were expressed as percent of the
specific binding, and the IC50 values were determined by non-
linear regression analysis using a sigmoidal variable slope curve
fitting. The dissociation constant (Ki) was calculated from
the Cheng-Prusoff equation (Ki = IC50/(1 þ (L/Kd)), where
the concentration of free radioligand L is approximated to the
concentration of added radioligand in the assay and Kd equals
the affinity of the radioligand to the receptor.
Broad Pharmacology Profiling. Broad profiling was per-
formed by CEREP (France). Compounds were generally assayed
for the ability to displace radioligand binding to the assayed tar-
gets. The following targets were profiled by enzyme assays:
PDE3,4,5, Lyn kinase, p38R kinase, and acetylcholine esterase.
Compounds were tested at 10 μM, n = 2, at the following targets
(CEREP catalog number in parentheses): D1 (no. 803-1h), D2S

6396 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17
Sams et al.
(no. 803-2h), H3 (no. 805-3h), rolipram (no. 833), μ (MOP,
agonist site) (no. 843-h), MC4 (no. 889-4h), β1 (no. 802-2ah),
5-HT1A (no. 808-1ah), 5-HT2B (agonist site) (no. 808-2bah),
5-HT2A (agonist site) (no. 808-2ah), Naþ channel (site 2) (no.
862-a), H1 (no. 805-1h), PCP (no. 895-6), glucocorticoid (GR)
(no. 812-h), σ (nonselective) (no. 891), D2S (agonist site) (no. 803-
2ha), M4 (no. 806-4h), GABAA (no. 804-1), R2A (no. 802-1bAh),
ETA (no. 825-1h), 5-HT4e (no. 808-4eh), DA transporter (no.
803-Uh), D4.4 (no. 803-44h), M1 (no. 806-1h), R1 (non-selective)
(no. 802-1a), A1 (no. 801-1h), BZD (central) (no. 851), M2 (no.
806-2h), M3 (no. 806-3h)
N (neuronal) (R-BGTX-insensitive) (R4β2) (no. 807-n1),
A1 (agonist site) (no. 801-1ah), MT1 (no. 892-1h), Ca2þ channel
(L, diltiazem site) (benzothiazepines) (no. 861-L2), R2B (no. 802-
1bBc), κ (KOP) (no. 842), 5-HT2C (agonist site) (no. 808-8cha),
5-HT7 (no. 808-7h), D3 (no. 803-3h), CCKB (CCK2) (no. 824-2h),
MAO-A (no. 838-a), β2 (no. 802-2bh), Ca2þ channel (L, verapa-
mil site) (phenylalkylamines) (no. 861-L3), H2 (no. 805-2hc), ML2
(MT3) (no. 892-2), M5 (no. 806-5h), choline transporter (CHT1)
(no. 806-Uh), I1 (no. 809-1p), NK1 (no. 826-1h), Y1 (no. 827-1h),
N (muscle-type) (no. 807-mh), 5-HT transporter (no. 808-Uh),
CB1 (no. 835-ch), 5-HT1B (no. 808-1b), NE transporter (no. 802-
Uh), Ca2þ channel (N) (no. 861-N), GABA transporter (no. 804-
U), A3 (no. 801-3h), NK2 (no. 826-2h), δ2 (DOP) (no. 841-h),
Ca2þ channel (L, DHP site) (no. 861-L1), 5-HT3 (no. 808-3h),
UT1 (no. 853-hc), A2A (no. 801-2ah).
In Vitro ADME Assays. Hepatocyte Clearance Determina-
tion. The intrinsic clearance of compounds in hepatocytes were
determined by the t1/2 method,36 i.e., measurement of the disap-
pearance of 1 μM compound over time by LC/MS-MS. Pooled
male rat cryopreserved hepatocytes (Sprague-Dawley) and
pooled cryopreserved human hepatocytes from 10 donors (male
and female) from CellzDirect, Inc., Durham, NC, were used.
Cells were thawed in a 37 °C water bath and live cells counted and
seeded in a total of 100 μL in Dulbecco’s modified Eagle medium
(high glucose) with 5 mM HEPES buffer in 96-well plates. Then
250 000 and 500 000 cells/mL were used for rat and human
hepatocytes, respectively. Incubations were started after 15 min
of preincubation and stopped at time points 0, 5, 15, 30, and
60 min for rats and 0, 30, 60, 90, and 120 min for human hepato-
cytes. The t1/2 is scaled to intrinsic clearance values of L/min for
humans and mL/min for rats. Values of scaling factors are 135/120
million cells per gram of liver, 45/20 g of liver per kilogram, and
standard weight of 0.25/70 kg for rat and human, respectively.
Protein Binding. Plasma protein binding was determined in
vitro in rat and human serum following incubation of 0.1 μM
test compound dissolved in PBS in ultracentrifugation filter
plates (300 μL in each well) (Multiscreen filter plates, 10 kDa
MW cutoff, Millipore) for 60 min at 37 °C. Plates were then
centrifuged for 60 min at 37 °C (3200 rcf), and the free fraction
was measured by LC/MS-MS following addition of internal
standard.
trifuged for 10 min at 4 °C after which plasma was drawn off.
Following decapitation, the brain was removed and stored
at -80 °C pending bioanalysis.
Plasma and Brain Bioanalysis. Brain homogenate was pre-
pared by homogenizing the whole brain with 70% acetonitrile
(1:4 v/v) followed by centrifugation and collection of the super-
natant. Plasma and brain supernatant samples were frozen
at -80 °C until analysis. Concentrations of were determined in
plasma and brain homogenate using turboflow chromatography
(dual column, focus mode, Cohesive Technologies, U.K.) fol-
lowed by MS/MS detection in positive-ion electrospray ioniza-
tion mode (Sciex API-3000 MS, Applied Biosystems, The
Netherlands). Samples of plasma and brain homogenate were
prepared by adding an equal amount of 10% methanol with
internal standard included (escitalopram), and after centrifuga-
tion (6000g, 5 °C, 20 min) an amount of 10 μL was injected into
the turboflow system. The mobile phase consisted of water/
methanol with 0.1% ammonium hydroxide pumped as a gradient
through an XTerra analytical column (MS C8, 2.1 mm ꢀ 20 mm,
Waters Corp., MA). The lower limit of quantification was
typically 0.5 ng/mL in plasma and 5 ng/g in brain.
In Vitro/in Vivo Extrapolation of Clearance. According to the
well-stirred model,25 an approximate hepatic clearance in each
species (CLpred) was predicted from hepatocyte intrinsic clear-
ance (CLint) according to
LBF ꢀ CLint ꢀ fu
CLpred
¼
LBF þ CLint ꢀ fu
where LBF is hepatic blood flow and fu is the fraction unbound
in serum. Human clearance was then estimated from the ob-
served in vivo rat clearance and the in vitro clearance prediction
in each species (CLpred,human, CLpred,rat)
26 according to
CLpred;human
CLpred;rat
CLhuman ¼ CLrat
Delayed Alternation. Y-Maze. Ten male C57Bl/6J mice
(22-24 g, Charles River, Germany) were used for the delayed
alternation Y-maze. The maze consisted of gray, nontranspar-
ent Plexiglas, shaped to a Y with each arm arranged in 120° to
each other and measuring 32 cm ꢀ 7 cm ꢀ 8 cm. One arm was
designated as “start arm”, and the two remaining were “goal
arms”. The maze was mounted on an airflow table. Each arm
was equipped with a guillotine door allowing restriction to the
start arm or providing entrance into each goal arm. Both goal
arms were equipped with a small cup at the end of the arm that
was baited with 15 μL of chocolate milk (Cocio, Denmark).
Before drug testing, subjects were trained to the alternation
procedure until an 80% correct alternation rate was reached
following the 1s ITI. Subsequently, animals were subjected to
drug testing, where 1 s (designated “0 sec”) and 60 s delays were
assigned randomly. Subjects were individually placed in the start
arm and given a “sample run”, during which one of the goal
arms was blocked. Once the subject collected the reward, it was
returned manually to the start arm and restricted there for the
delay. During the following run (“choice run”) both arms were
accessible, but the subject was rewarded only when entering
the goal arm opposite that visited during the sample run (i.e., the
arm visited during sample run was not rebaited). The animal was
returned to the start arm and restricted there for 10 s after which
a new trial consisting of sample and choice run started. A total of
20 trials were performed per day and mouse (10 ꢀ 0 s delay, 10 ꢀ
60 s delay). For drug testing, animals were assigned to a Latin-
square design in which each mouse was tested in each dose group
over 1 week of testing. Lu AE51090 and vehicle (10% HP-β-
cyclodextrin) were dosed subcutaneously in a volume of 5 mL/kg
and 30 min before the experiment. Data are expressed as
the mean ( SEM. Significant differences between group
mean values were assessed by two-way analysis of variance
Nonspecific binding in brain were obtained by equilibrium
dialysis in 96-well format. Freshly isolated mouse brains were
homogenized twice in two volumes of phosphate buffer, pH 7.4.
The dialysis methodology was modified from ref 37. Mem-
branes (cutoff 12-14 kDa, HTDialysis) was soaked in PBS/
ethanol 80:20 and rinsed in deionized water before use. Brain
homogenate was spiked with test compound to a final concen-
tration of 1 μM, and equilibrium dialysis was performed by
incubating at 37 °C for 5 h. Subsequently, the compound
concentration was measured in the buffer phase.
In Vivo ADME Studies. All dosing solutions consisted of 5%
HP-β-cyclodextrin, pH 5. Serial blood sampling from the tail
vein in awake rats was used for pharmacokinetic assessment
following intravenous (1 mg/kg) and oral (2 mg/kg) drug
administration. For brain/plasma exposure studies, cardiac
blood was obtained under isoflurane anesthesia. Dose volumes
of 5 and 10 mL/kg were applied for rats and mice, respectively.
Blood samples were collected in EDTA-coated tubes and cen-

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 17 6397
(ANOVA) with the all pairwise multiple comparison procedure
(Student-Newman-Keuls method) (SigmaStat 3.0, SPSS).
All in vivo experiments were performed in accordance with
Danish legislation, and animals were treated with adherence to
guidelines for the care of experimental animals.
Tseng, E.; Uveges, A. J.; Mayer, S. C. Pharmacological comparison
of muscarinic ligands: historical versus more recent muscarinic M1-
preferring receptor agonists. Eur. J. Pharmacol. 2009, 605, 53–56.
(19) Spalding, T. A.; Trotter, C.; Skjærbæk, N.; Messier, T. L.; Currier,
E. A.; Burstein, E. S.; Li, D.; Hacksell, U.; Brann, M. R. Discovery
of an ectopic activation site on the M1 muscarinic receptor. Mol.
Pharmacol. 2002, 61, 1297–1302.
(20) Spalding, T. A.; Ma, J.-N.; Ott, T. R.; Friberg, M.; Bajpai, A.;
Bradley, S. R.; Davis, R. E.; Brann, M. R.; Burstein, E. S.
Structural requirements of transmembrane domain 3 for activation
by the muscarinic receptor agonists AC-42, AC-260484, clozapine,
and N-desmethylclozapine: evidence for three distinct modes of
receptor activation. Mol. Pharmacol. 2006, 70, 1974–1983.
(21) Langmead, C. J.; Fry, V. A. H.; Forbes, I. T.; Branch, C. L.;
Christopoulos, A.; Wood, M. D.; Herdon, H. J. Probing the
molecular mechanism of interaction between AC-42 and the
muscarinic M1 receptor: direct pharmacological evidence that
AC-42 is an allosteric agonist. Mol. Pharmacol. 2006, 69, 236–246.
(22) Blatter, F.; Thygesen, M. B.; Tolf, B.-R.; Berghausen, J. Method of
Synthesis and Isolation of Solid N-Desmethylclozapine and Crys-
talline Forms Thereof. US 2005282800, 2005.
(23) Bridges, T. M.; Brady, A. E.; Kennedy, J. P.; Daniels, R. N.; Miller,
N. R.; Kwango, K.; Breiniger, M. L.; Gentry, P. R.; Brogan, J. T.;
Jones, C. K.; Conn, P. J.; Lindsley, C. W. Synthesis and SAR of
analogues of the M1 allosteric agonist TBPB. Part I: Exploration of
alternative benzyl and privileged structure moieties. Bioorg. Med.
Chem. Lett. 2008, 18, 5439–5442.
(24) Motulsky, H.; Christopoulos, A. Fitting Models to Biological Data
Using Linear and Nonlinear Regression; Oxford University Press:
Oxford, U.K., 2004.
(25) Rowland, M.; Tozer, T. N. Clinical Pharmacokinetics: Concepts
and Applications; Williams and Wilkins: Baltimore, MD, 1995.
(26) Lave, T.; Coassolo, P.; Reigner, B. Prediction of hepatic metabolic
clearance based on interspecies allometric scaling techniques and in
vitro-in vivo correlations. Clin. Pharmacokinet. 1999, 36, 211–231.
(27) Mirza, N. R.; Peters, D.; Sparks, R. G. Xanomeline and the
antipsychotic potential of muscarinic receptor subtype selective
agonists. CNS Drug Rev. 2003, 159–186.
(28) Jones, C. K.; Brady, A. E.; Bubser, M.; Deutch, A. Y.; Williams,
L. C. TBPB is a highly selective M1 allosteric muscarinic receptor
agonist in vitro and produces robust antipsychotic-like effects in
vivo. Neuropsychopharmacology 2006, 31, S116.
(29) Langmead, C. J.; Christopoulos, A. Allosteric agonists of
7TM receptors: expanding the pharmacological toolbox. Trends
Pharmacol. Sci. 2006, 27, 475–481.
(30) Sur, C.; Mallorga, P. J.; Wittmann, M.; Jacobson, M. A.;
Pascarella, D.; Williams, J. B.; Brandish, P. E.; Pettibone, D. J.;
Scolnick, E. M.; Conn, P. J. N-Desmethylclozapine, an allosteric
agonist at muscarinic 1 receptor, potentiates N-methyl-D-aspartate
recetor activity. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 13674–
13679.
(31) Aultman, J. M.; Moghaddam, B. Distinct contributions of gluta-
mate and dopamine receptors to temporal aspects of rodent work-
ing memory using a clinically relevant task. Psychopharmacology
2001, 153, 353–364.
(32) Courtney, S. M.; Petit, L.; Maisog, J. M.; Ungerleider, L. G.;
Haxby, J. V. An area specialized for spatial working memory in
human frontal cortex. Science 1998, 279, 1347–1351.
(33) Kesner, R. P.; Hunt, M. E.; Williams, J. M.; Long, J. M. Prefrontal
cortex and working memory for spatial response, spatial location, and
visual object information in the rat. Cereb. Cortex 1996, 6, 311–318.
(34) Brann, M.; Messier, T.; Currier, E.; Duggento, K.; Friberg, M.;
Skjærbæk, N.; Spalding, T. Compounds with activity on muscari-
nic receptors, WO200105763, 2001.
(35) Sambrook, J.; Fritsch, E. F.; Maniastis, T. Molecular Cloning: A
Laboratory Manual; Cold Spring Harbor Laboratory Press: Cold
Spring Harbor, NJ, 1989.
(36) Obach, R. S.; Baxter, J. G.; Liston, T. E.; Silber, B. M.; Jones, B. C.;
MacIntyre, F.; Rance, D. J.; Wastall, P. The prediction of human
pharmacokinetic parameters from preclinical and in vitro metabo-
lism data. J. Pharmacol. Exp. Ther. 1997, 283, 46–58.
(37) Kalvass, J. C.; Maurer, T. S. Influence of nonspecific brain and
plasma binding on CNS exposure: implications for rational drug
discovery. Biopharm. Drug Dispos. 2002, 23, 327–338.
Supporting Information Available: Synthesis procedures,
spectroscopic data, and purities for intermediates 8b-k and
compounds 5, 6, 9-16, and 18-38. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Caulfield, M. P. Muscarinic receptors;characterization, coupling,
function. Pharmacol.Ther. 1993, 58, 319–379.
(2) Davies, P.; Maloney, A. J. Selective loss of central cholinergic
neurons in Alzheimer’s disease. Lancet 1976, 2, 1403.
(3) Muir, J. L. Acetylcholine, aging, and Alzheimer’s disease. Phar-
macol., Biochem. Behav. 1997, 56, 687–696.
(4) Wess, J. Muscarinic acetylcholine receptor knockout mice: novel
phenotypes and clinical implications. Annu. Rev. Pharmacol.
Toxicol. 2004, 44, 423–450.
(5) Anagnostaras, S. G.; Murpht, G. G.; Hamilton, S. E.; Mitchell,
S. L.; Rahnama, N. P.; Nathanson, N. M.; Silva, A. J. Selective
cognitive dysfunction in acetylcholine M1 muscarinic receptor
mutant mice. Nat. Neurosci. 2003, 6, 51–58.
(6) Dean, B.; McLeod, M.; Keriakous, D.; McKenzie, J.; Scarr, E.
Decreased muscarinic 1 receptors in the dorsolateral prefrontal
cortex of subjects with schizophrenia. Mol. Psychiatry 2002, 7,
1083–1091.
(7) Mancama, D.; Arrantz, M. J.; Landau, S.; Kerwin, R. Reduced
expression of the muscarinic 1 receptor cortical subtype in schizo-
phrenia. Am. J. Med. Genet., Part B 2003, 119, 2–6.
(8) Langmead, C. J.; Watson, J.; Reavill, R. Muscarinic acetylcholine
receptors as CNS drug targets. Pharmacol. Ther. 2008, 117, 232–
243.
(9) Sur, C.; Kinney, G. G. Selective targeting of muscarinic receptors:
novel therapeutic appraches for psychotic disorders. Curr. Neuro-
pharmacol. 2005, 3, 63–71.
(10) Sellin, A. K.; Shad, M.; Tamminga, C. Muscarinic agonists for the
treatment of cognition in schizophrenia. CNS Spectrums 2008, 13,
985–996.
(11) Fischer, A. Cholinergic treatments with emphasis on m1 muscari-
nic agonists as potential disease-modifying agents for Alzheimer’s
disease. Neurotherapeutics 2008, 5, 433–442.
(12) Fischer, A. Muscarinic receptor agonists in Alzheimer’s disease.
CNS Drugs 1999, 12, 197–214.
(13) Sauerberg, P.; Olesen, P. H.; Nielsen, S.; Treppendahl, S.;
Sheardown, M. J.; Honore, T.; Mitch, C. H.; Ward, J. S.; Pike,
A. J. Novel functional M1 selective muscarinic agonists. Synthesis
and structure-activity relationships of 3-(1,2,5-thiadiazolyl)-1,2,5,6-
tetrahydro-1-methylpyridines. J. Med. Chem. 1992, 35, 2274–2283.
(14) Bodick, N. C.; Offen, W. W.; Levey, A. I.; Cutler, N. R.; Gauthier,
S. G.; Satlin, A.; Shannon, H. E.; Tollefson, G. D.; Rasmussen, K.;
Bymaster, F. P.; Hurley, D. J.; Potter, W. Z.; Paul, S. M. Effects of
xanomeline, a selective muscarinic receptor agonist, on cognitive
function and behavioural symptoms in Alzheimer disease. Arch.
Neurol. 1997, 54, 465–473.
(15) Bymaster, F. P.; Whitesitt, C. A.; Shannon, H. E.; DeLapp, N.;
Ward, J. S.; Calligaro, D. O.; Shipley, L. A.; Buelke-Sam, J. L.;
Bodick, N. C.; Farde, L.; Sheardown, M. J.; Olesen, P. H.; Hansen,
K. T.; Suzdak, P. D.; Swedberg, M. D. B.; Sauerberg, P.; Mitch,
C. H. Xanomeline: a selective muscarinic agonist for the treatment
of Alzherimer’s disease. Drug Dev. Res. 1997, 40, 158–170.
ꢀ
(16) Shekar, A.; Potter, W. Z.; Lightfoot, J.; Lienemann, J.; Dube, S.;
Mallinckrodt, C.; Bymaster, F. P.; McKinzie, D. L.; Felder, C. C.
Selective muscarinic receptor agonist xanomeline as a novel treat-
ment approach for schizophrenia. Am. J. Psychiatry 2008, 165,
1033–1039.
(17) Moltzen, E. K.; Bjørnholm, B. Medicinal chemistry of muscarinic
agonists: developments since 1990. Drugs Future 1995, 20, 37–54.
(18) Heinrich, J. N.; Butera, J. A.; Carrick, T.; Kramer, A.; Kowal, D.;
Lock, T.; Marquis, K. L.; Paush, M. H.; Popiolek, M.; Sun, S.-C.;