Discovery, Optimization, and Biological Characterization of 2,3,6-Trisubstituted Pyridine-Containing M4 Positive Allosteric Modulators

Article abstract of DOI:10.1002/cmdc.201900088

Herein we describe the discovery and optimization of a new series of 2,3-disubstituted and 2,3,6-trisubstituted muscarinic acetylcholine receptor 4 (M₄) positive allosteric modulators (PAMs). Iterative libraries enabled rapid exploration of one-dimensional structure–activity relationships (SAR) and identification of potency-enhancing heterocycle and N-alkyl pyrazole substituents. Further optimization led to identification of the potent, receptor-subtype-selective, brain-penetrant tool compound 24 (7-[3-[1-[(1-fluorocyclopentyl)methyl]pyrazol-4-yl]-6-methyl-2-pyridyl]-3-methoxycinnoline). It is efficacious in preclinical assays that are predictive of antipsychotic effects, producing dose-dependent reversal of amphetamine-induced hyperlocomotion in rats and mice, but not in M₄ knockout mice. Cholinergic-related adverse effects observed in rats treated with 24 at unbound plasma concentrations more than 3-fold higher than an efficacious dose in the hyperlocomotion assay were fewer and less severe than those observed in rats treated with the nonselective M₄ agonist xanomeline, suggesting a receptor-subtype-selective PAM has the potential for an improved safety profile.

Full text of DOI:10.1002/cmdc.201900088

DOI: 10.1002/cmdc.201900088
Full Papers
Discovery, Optimization, and Biological Characterization of
2,3,6-Trisubstituted Pyridine-Containing M4 Positive
Allosteric Modulators
Jeffrey W. Schubert,*[a] Scott T. Harrison,[a] James Mulhearn,[a] Robert Gomez,[a]
Robert Tynebor,[a] Kristen Jones,[a] Jaime Bunda,[a] Barbara Hanney,[a] Jenny Miu-Chen Wai,[a]
Chris Cox,[a] John A. McCauley,[a] John M. Sanders,[b] Brian Magliaro,[c] Julie O’Brien,[c]
Natasa Pajkovic,[d] Sarah L. Huszar Agrapides,[e] Anne Taylor,[e] Anthony Gotter,[f]
Sean M. Smith,[f] Jason Uslaner,[f] Susan Browne,[e] Stefania Risso,[f] and Melissa Egbertson[a]
Herein we describe the discovery and optimization of a new
series of 2,3-disubstituted and 2,3,6-trisubstituted muscarinic
acetylcholine receptor 4 (M4) positive allosteric modulators
(PAMs). Iterative libraries enabled rapid exploration of one-di-
mensional structure–activity relationships (SAR) and identifica-
tion of potency-enhancing heterocycle and N-alkyl pyrazole
substituents. Further optimization led to identification of the
potent, receptor-subtype-selective, brain-penetrant tool com-
pound 24 (7-[3-[1-[(1-fluorocyclopentyl)methyl]pyrazol-4-yl]-6-
methyl-2-pyridyl]-3-methoxycinnoline). It is efficacious in pre-
clinical assays that are predictive of antipsychotic effects, pro-
ducing dose-dependent reversal of amphetamine-induced hy-
perlocomotion in rats and mice, but not in M4 knockout mice.
Cholinergic-related adverse effects observed in rats treated
with 24 at unbound plasma concentrations more than 3-fold
higher than an efficacious dose in the hyperlocomotion assay
were fewer and less severe than those observed in rats treated
with the nonselective M4 agonist xanomeline, suggesting a re-
ceptor-subtype-selective PAM has the potential for an im-
proved safety profile.
Introduction
There is significant unmet medical need for a safe and effica-
cious treatment for the noncognitive symptoms of dementia,
such as aggression and psychosis.[1] Clinical studies using xano-
meline, an M1/M4 preferring muscarinic acetylcholine receptor
(mAChR) subtype agonist, have demonstrated that targeting
the muscarinic cholinergic system could be a viable approach
for alleviating psychosis and behavioral disturbances in Alz-
heimer’s disease (AD) and schizophrenia patients.[2,3] Preclinical
studies using knockout animals and more selective molecules
suggest the effects of xanomeline are likely mediated by the
antidopaminergic effects of M4 activation.[4] Unfortunately, clini-
cal development of xanomeline was discontinued because of
undesirable side effects attributed to poor receptor-subtype
selectivity.[4–7] As a result of the clinical effects and adverse
effect profile of xanomeline, there has been an effort to devel-
op more selective compounds that activate M4.[8–11] M4 positive
allosteric modulators (PAMs) bind to a site which is distinct
from the orthosteric binding site of the endogenous ligand,
acetylcholine. Targeting a less-conserved allosteric site with an
M4 PAM that potentiates endogenous ACh signaling can im-
prove receptor-subtype selectivity, as demonstrated by several
research groups which have identified potent and selective M4
PAMs (Figure 1).[9,12–21]
[a] J. W. Schubert, Dr. S. T. Harrison, J. Mulhearn, R. Gomez, R. Tynebor,
K. Jones, J. Bunda, B. Hanney, J. M.-C. Wai, Dr. C. Cox, Dr. J. A. McCauley,
Dr. M. Egbertson
Department of Medicinal Chemistry, Merck & Co., Inc., West Point, PA (USA)
E-mail: jeffrey_schubert@merck.com
[b] Dr. J. M. Sanders
Department of Computational and Structural Chemistry, Merck & Co., Inc.,
West Point, PA (USA)
[c] B. Magliaro, J. O’Brien
Department of In Vitro Pharmacology, Merck & Co., Inc., West Point, PA
(USA)
[d] Dr. N. Pajkovic
Herein we describe lead identification efforts to develop a
human- and rat-active M4 PAM with properties suitable for
evaluation in preclinical assays predictive of antipsychotic ac-
tivity. We sought a compound with >100-fold receptor-sub-
type functional selectivity against human M1–3,5 expressed in re-
combinant cell lines and pharmacokinetics that could achieve
good central nervous system (CNS) exposure, and serve as an
in vivo tool for evaluating M4 PAM pharmacology.
Department of Pharmacokinetics, Pharmacodynamics, and Drug Metabo-
lism, Merck & Co., Inc, West Point, PA (USA)
[e] S. L. Huszar Agrapides, A. Taylor, Dr. S. Browne
Department of In Vivo Pharmacology, Merck & Co., Inc., West Point, PA
(USA)
[f] Dr. A. Gotter, Dr. S. M. Smith, Dr. J. Uslaner, Dr. S. Risso
Department of Neuroscience Research, Merck & Co., Inc., West Point, PA
(USA)
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/cmdc.201900088.
ChemMedChem 2019, 14, 1 – 11
1
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 1. Structures of reported M4 PAMs and the pyridine containing 1, dis-
cussed herein.
Results and Discussion
M4 is a Gi-coupled receptor, and attempts to use the native Gi
signaling pathway by measuring changes in cAMP levels pro-
duced narrow signal windows that were deemed unsuitable
for high-throughput screening (HTS; data not shown). Conse-
quently, an M4 PAM FLIPR high-throughput screen of 1.8 mil-
lion compounds was conducted using CHO-K1 cells stably
transfected with human M4 receptor and Gqi5. Pyridine 1
(Figure 1) was one of few compounds with similar human and
rat M4 PAM activity (human IP=248 nm, 106% of the maximal
response to ACh (ACh Max), which is measured in the presence
of 10 mm ACh; rat IP=1.2 mm, 31% ACh Max). Compound 1
showed excellent receptor-subtype selectivity (M1–3 PAM all
Scheme 1. Preparation of M4 PAM analogues III and VII: a),b) ArB(pin),
Pd(dtbpf)Cl2, K3PO4, THF, 808C; c) Pd(PPh3)4, (Bu3Sn)2, toluene, 1258C;
d) Pd(PtBu2)3, LiCl, CuI, ArBr, dioxane, 1508C microwave; e) ArB(pin),
Pd(PPh3)4, Na2CO3, dioxane, 808C; f) ArB(pin), Pd(dtbpf)Cl2, Na2CO3, dioxane,
808C; g) Pd(OAc)2, Xphos, KF, dioxane, 808C; h) TFA, CH2Cl2; i) R1OTs or R1Br,
DMF, NaH, 658C; j) R1OH, DIAD, PPh3, THF.
Alcohols provided a diverse pool of reagents that were con-
verted into tosylates for alkylation. Further derivatized re-
agents, such as substituted cyclopentylmethyl tosylates, were
prepared from cyclopentane carboxylic acid, as shown in
Scheme 2. Fischer esterification, followed by deprotonation
and trapping with an electrophile provided the methyl- or
fluoro-substituted esters XI. Reduction and sulfonylation led to
the final tosylates XIII.
>30 mm). P-glycoprotein (P-gp) efflux (Papp =30.3ꢁ10ꢀ6 cmsꢀ1
;
MDR1a B!A/A!B=44.8) limited its utility as a tool com-
pound, but the structure was an attractive lead due to its ame-
nability to rapid analogue synthesis and the opportunity that
affords to quickly explore a variety of substitutions and vec-
tors.
Chemistry
Compounds were accessed through the generic synthetic
routes outlined in Scheme 1. Suzuki–Miyaura cross-coupling
between 3-bromo-2-chloropyridine I and N-alkyl pyrazole bor-
onic acid derivatives afforded substituted pyridines II. Subse-
quent cross-coupling of II afforded final compounds of type III.
To expand the scope of available cross-coupling partners from
aryl boronates to aryl bromides, Stille coupling with intermedi-
ate IV was also pursued.
Installation of the 2-pyridyl substituent via Suzuki–Miyaura
cross-coupling to give 2-aryl-3-halopyridines VI provided a
useful intermediate for exploration of pyrazole substitution. A
second cross-coupling with appropriately substituted pyrazole
boronic acid derivatives afforded final compounds VII. Alterna-
tively, protected pyrazole intermediate VIII can be deprotected
and alkylated with alkyl bromides or tosylates, or with alcohols
via Mitsunobu reaction.
Scheme 2. Preparation of substituted cyclopentyl tosylates used for pyrazole
alkylation: a) H2SO4, MeOH, reflux; b) X=F: LDA, NFSI, THF; X=Me: LDA,
MeI, THF; c) LiBH4 or LiAlH4, THF d) TsCl, pyridine.
Structure–activity relationships of di- and tri-substituted
pyridine M4 PAMs
An iterative library approach as outlined in Scheme 1 enabled
one-dimensional exploration of the 2-pyridyl and N-pyrazole
&
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
2
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
structure–activity relationships (SAR); key findings are summar-
ized in Tables 1–3. The 4-cyanophenyl analogue 2 (Table 1) re-
moves the basic amine and hydrogen-bond donating capacity
of analogue 1, eliminating P-gp efflux while retaining human
M4 PAM potency, improving rat M4 PAM potency, and maintain-
ing excellent receptor-subtype selectivity. Amide analogues de-
rived from compound 2, such as 3, were generally less potent,
and showed a 3- to 6-fold decrease in potency relative to the
simple phenyl 4. Naphthalene 5 and 6-substituted quinoline 6
showed large decreases in potency relative to 2. The 7-substi-
tuted quinoline 7 exhibited slightly improved M4 PAM potency
and promising receptor-subtype selectivity (M1 PAM IP =
1.7 mm, M2,3 PAM >30 mm), and led to focused exploration of
heterocycles. Isoindolinone 8 and cinnoline 9 are highly recep-
tor-subtype selective (M1–3 PAM all >30 mm), exhibit similar
human and rat M4 PAM activity, and are not P-gp substrates.
Cyanophenyl 2 served as a starting point for exploration of
the pyrazole substituent, summarized in Table 2. Polar ethers
Table 2. Human and rat muscarinic activity and lipophilicity measure-
ments for cyanophenyl-substituted M4 PAMs.
Table 1. Human and rat M4 activity and P-gp efflux for selected M4 PAMs.
[a] Values represent the numerical average of at least two experiments.
Inter-assay variability was 3-fold (IP, nm) unless otherwise noted. Human
average ACh Max ranged from 35 to 45% for 10, 12, and 16, and from
56 to 82% for 11, 13–16, and 18. Rat average ACh Max ranged from ꢀ38
to 31% for 10, 12–16, and 18 and from 60 to 81% for 11, 16 and 18.
[b] ClogD was calculated at pH 7.4 using ACD Percepta software (v.12).
10 and 11 showed a large decrease in M4 PAM activity relative
to 2, indicating a preference for nonpolar groups in this vector.
A comparison of alkyl substituents of increasing size and
branching (compounds 12–18) revealed that properly dis-
tanced hydrophobic bulk with nonplanar character was
needed to achieve human and rat M4 PAM activity. Linear alkyl
substituents (12, 14) and small aliphatic rings (16, 17) demon-
strated modest human M4 PAM activity, but not rat M4 activity,
while branched alkyl (13, 16), phenyl (15), and cyclohexyl (18)
analogues exhibited similar human and rat M4 PAM activity.
Compound 18 is potent (human IP=256 nm; rat IP=229 nm)
and highly receptor-subtype selective (M1–3 PAM all >30 mm),
but is calculated to be quite lipophilic, with ClogD=4.16 at
pH 7.4.
[a] Calcium mobilization assays with hM4, rM4, or hM1,2,3/Gaqi5-CHO-K1
cells performed in the presence of an EC20 fixed concentration of acetyl-
choline; values represent the numerical average of at least two experi-
ments. Inter-assay variability was 3-fold (IP, nm) unless otherwise noted.
Human average ACh Max ranged from 67 to 100% for compounds 1, 2,
7, and 8, and from ꢀ18 to 57% for 3–6. Rat average ACh Max ranged
from ꢀ88 to 44% for 1–6, and was >84% for 7–9. [b] P-gp efflux mea-
sured at 1 mm in LLC-PK1 cell line expressing rat (rat LLC-Mdr1a) or
human (human LLC-MDR1) P-gp; Papp values in control LLC-PK1 cell line
were >30ꢁ106 cmsꢀ1. Blank: not determined.
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
3
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
With preliminary SAR in hand, a quinoline template was
used to further probe pyrazole substitution, and efforts were
focused on exploring branched alkyl substituents (Table 3). Cy-
clohexyl 19 was potent (human IP=93 nm) and not a P-gp
substrate but also quite lipophilic (ClogD=4.73). The branched
Table 3. Human and rat muscarinic activity and lipophilicity measure-
ments for quinoline-substituted M4 PAMs.
[a] Values represent the numerical average of at least two experiments.
Inter-assay variability was 3-fold (IP, nm) unless otherwise noted. For all
compounds, human and rat average ACh Max ranged from 88 to 100%.
[b] Human P-gp efflux measured at 1 mm in human LLC-MDR1 cell line;
Papp values were >30ꢁ106 cmsꢀ1. Blank: not determined. [c] ClogD was
calculated at pH 7.4 using ACD Percepta software (v.12).
Figure 2. Graphical visualization of M4 PAM data with biological activity and
R1 represented across the x-axis, R2 across the y-axis, and R3 denoted by
color (H=blue, Me=orange). a) M4 PAM potency plotted on a logarithmic
scale. b) M1 and M2 PAM selectivity (M1 IP/M4 IP and M2 IP/M4 IP, respectively)
for combinations of R1, R2, and R3. Lines connect identical compounds, and
markers are shaped by P-gp efflux.
fluoroalkyl 20 was less lipophilic, but also ꢁ3-fold less potent.
Cyclopentyl analogue 21 maintains potency and receptor-sub-
type selectivity, and is not a P-gp substrate. Methyl substitu-
tion of the tertiary carbon (22) improves M4 PAM activity ꢁ3-
fold, but introduces M1 and M2 PAM activity (IP=1.6 mm and
170 nm, respectively). Fluoro substitution (23) decreases M4
PAM potency relative to 21, but also lowers lipophilicity
(ClogD=4.16) and does not impart P-gp efflux susceptibility.
The one-dimensional SAR assessments described above ena-
bled rapid vector exploration, but did not provide insight into
the nonlinear interplay of these substituents on potency and
receptor-subtype selectivity. To address this, combinations of
potent and/or selective substituents were prepared. Methyl
substitution of the pyridine was found to impart some potency
improvements in a limited number of examples, and was also
included.
pattern. Trends can be identified by comparing compounds
within a color (R3 =H or Me), row (R2 =A, B, or C), or column
(R1 =a, b, c, or d). Using this analysis, M4 activity is similar
across subseries, with R3 =Me equipotent or slightly more
active than an unsubstituted parent, and substituted cyclopen-
tyl congeners c and d imparting modest potency improve-
ments.
Compounds were generally inactive against M3 and M5 (data
not shown), and M1 and M2 receptor-subtype selectivity and P-
gp efflux were used to further differentiate compounds for
in vivo evaluation. Figure 2b examines receptor-subtype selec-
tivity, with a dashed line representing 100-fold subtype selec-
tivity and colored, solid lines connecting M4 versus M1 and M4
versus M2 selectivity for the same compound. Markers are
shaped by whether or not a compound is a P-gp substrate;
filled circles are not P-gp substrates, ’X’s indicate P-gp sub-
strates, and open circles do not have data (not measured).
Although not P-gp substrates, quinoline-containing com-
pounds (H-A-a,c,d and Me-A-a,b,d) are generally less selective
than isoindolinones or cinnolines, with most analogues exhibit-
ing <100-fold selectivity against M1 or M2. In two cases methyl
An R-group decomposition and visualization of M4 PAM SAR
is summarized in Figure 2. Similar analyses have been previous-
ly reported,[22,23] and using these visualizations can aid in corre-
lating structure with multiple dimensions of biological data. In
Figure 2a, compounds have M4 PAM potency ranging from 10
to 100 nm. Compounds 7, 8, and 9 are labeled for reference,
and subsequent compounds are named following an R3-R2-R1
&
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
4
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
substitution at R3 (compare orange and blue of A-a and A-b)
lowers M2 selectivity. By contrast, isoindolinones (B) and cinno-
lines (C) are generally more selective against M1 and M2, with
most analogues exhibiting >100-fold selectivity. The effect of
methyl substitution on M2 selectivity was found to be depen-
dent on R1 and R2, with increases (compare orange and blue of
B-a, B-b, and C-c) and decreases (compare orange and blue of
B-c, B-d, and C-d) in selectivity observed. Indeed, the interplay
of selectivity imparted by combinations of R1-R3 highlights the
utility of preparing small libraries and using a multivariate data
visualization and analysis to understand trends within and be-
tween subseries.
Additional pharmacological profiling shows that 24 exhibits
moderate human M4 agonism (641 nm, 45% ACh Max), but no
measurable human or rat M1–3,5 agonism (>30 mm). Testing in a
rhesus M4 PAM assay showed potency similar to rat and
human (IP=13 nm). M4 PAM 24 exhibits a favorable ancillary
pharmacology profile in a screen of 119 targets (Eurofins
Neuro panel), with one hit under 5 mm potency (adenosine A1
IC50 4.1 mm). NAV 1.5 activity was 7.8 mm, a >520-fold differ-
ence relative to the M4 IP, and CAV 1.2 and MK-499 activities
were >30 mm. Receptor allostery was further characterized by
measuring changes in acetylcholine response (51 pm to 3 mm)
at varying concentrations of 24 (100 nm to 30 mm). As shown
in Figure 3, M4 PAM 24 is a cooperative positive allosteric mod-
ulator of ACh (a=190ꢂ128) with a modest 857ꢂ613 nm af-
finity for the unbound receptor (KB).[25]
Isoindolinone analogues were substrates for rat P-gp (H-B-a,
Me-B-a, H-C-c), but methylation of the core (Me-B-c) and re-
placement of the pyrazole substituent (compare H-B-c with H/
Me-B-d) or heterocycle (H/Me-C-c and H/Me-C-d) decreased
susceptibility for P-gp. While several compounds, including
Me-C-c and H-C-d (boxes, Figure 2b), meet the desired P-gp
and receptor-subtype selectivity criteria, examining trends be-
tween subseries provides insight into the generality of a sub-
stituent effect. Comparing R1 =c versus R1 =a,b,d for R2 =A
and B shows less variability in M2 selectivity upon core methyl-
ation with R1 =c. Consequently, fluorocyclopentyl-containing
PAMs (R1 =c) were prioritized for further characterization. Cin-
noline Me-C-c, 24, emerged as a compound with desired
M1,2,3,5 selectivity and M4 PAM potency. Additionally, 24 is not a
human or rat P-gp substrate. Potency, selectivity, permeability,
and physicochemical properties for 24 are summarized in
Table 4.
Figure 3. Acetylcholine response curves measured by Ca2+ flux in CHO cells
expressing M4 mAChR at varying concentrations of 24.
Table 4. Profile of compound 24.
With a suitable in vitro profile, pharmacokinetic studies with
24 were conducted in rat, and parameters obtained are listed
in Table 5. PAM 24 is a highly permeable compound in vitro,
but has poor bioavailability in rats. Subsequently, alternative
routes of delivery were examined to achieve higher exposures.
After intraperitoneal (i.p.) dosing in 30% Captisol, unbound
plasma concentrations at 30 min and 1 h post-administration
(37 and 41 nm, respectively) were ꢁ2- to 3-fold above the M4
Parameter
Value
human M4 PAM IP [nm]
rat M4 PAM IP [nm]
M1/M4 IP ratio
17
29
134
M2/M4 IP ratio
M3/M4 IP ratio
M5/M4 IP ratio
Papp [106 cmsꢀ1
>1700
>1700
>1700
32.5
Table 5. Rat PK parameters for compound 24.a
]
Parameter
i.v.
p.o.
i.p.
Human LLC-MDR1 B!A/A!B
Rat LLC-Mdr1a B!A/A!B
Solubility (pH 2, 7, FASSIF)a [mm]
PSAb [ꢂ2]
0.5
0.6
dose [mgkgꢀ1
CL [mLminꢀ1 kgꢀ1
]
2
70
0.75
2.4
–
–
–
–
10
–
–
10
–
–
–
–
1.1
2.0
8.5
]
204, 146, 165
68
0.34, 4.86
2.89
t1/2 [h]
Vdss [Lkgꢀ1
F [%]
]
–
LBEc, LLEd
HPLC logD (pH 7)e
0.78
0.030
0.38
0.045
Cmax [mm]
tmax [h]
AUC0–1 [mmh]
P-gp efflux was measured at 1 mm in human LLC-MDR1 and rat LLC-
Mdr1a (rat) cell lines. Average ACh Max for human and rat M4 PAM was
>95%, for human M1 it was 60%. [a] FaSSIF: fasted-state simulated intes-
tinal fluid. [b] Polar surface area. [c] Ligand binding efficiency: 1.4ꢁp(hM4
PAM IP)/HAC.[24] [d] Ligand lipophilic efficiency: p(hM4 PAM IP)ꢀHPLC
logD.[24] [e] Determined by retention time using linear regression relating
retention time and shake-flask logD values for a set of standards.
[a] Intravenous and p.o. doses were administered to Wistar Hanover rats,
and i.p. doses to Sprague–Dawley rats; rat PPB=4.2% unbound. Values
are the mean from 2 or 3 animals; i.v. and p.o. vehicle: DMSO/PEG400/
H20 (20:60:20); i.p. vehicle: 30% Captisol.
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
5
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
PAM IP of 24. Consequently, we tested pharmacodynamic re-
sponse within 1–2 h of i.p. dosing, when plasma concentra-
tions were near the M4 PAM IP.
were 25 nm at 75 min post-dose, which is ꢁ1.5-fold the M4
PAM IP. Compound 24 is a permeable compound, not a sub-
strate for rat P-gp (LLC-Mdr1a B!A/A!B=0.6), and unbound
plasma and cerebrospinal fluid (CSF) concentrations obtained
from satellite animals (at 15 min) were within 1- to 3-fold of
each other, suggesting sufficient CNS exposure.[28–30] Moreover,
24 dosed orally at 100 mgkgꢀ1 reversed amphetamine-induced
locomotion in wild-type but not in M4 knockout mice
(Figure 5). Taken together, these data suggest that 24 is an ex-
cellent tool compound for preclinical in vivo assays.
Pharmacology of M4 PAM 24 was evaluated in a rat assay of
amphetamine-induced hyperlocomotion, which measures
animal activity in a locomotor chamber by the number of infra-
red beams the rat crosses over time. This assay is a surrogate
preclinical assay for antipsychotic potential, as antipsychotic
drugs typically abrogate stimulant-induced activity increas-
es.[26,27] After acclimation to individual locomotor chambers for
30 min, Sprague–Dawley rats were administered vehicle or
varying doses of 24 or xanomeline (5 mLkgꢀ1, i.p.); 15 min
later, rats received a subcutaneous dose of amphetamine
(1 mLkgꢀ1, s.c.). Motor activity recorded over the subsequent
60 min is summarized in Figure 4. Rats treated with 3, 10, and
30 mgkgꢀ1 24 exhibited a dose-dependent reversal of amphet-
amine-induced hyperlocomotion (Figure 4a), similar to the
positive control xanomeline (10 mgkgꢀ1)[27] (Figure 4b), and to
other M4 PAMs characterized by other research groups.[12,13,17,18]
Unbound plasma concentrations from the highest-dose group
Figure 5. Data are meanꢂSEM beam breaks by mice per 10-min time inter-
val. M4 PAM 24 reduces amphetamine-induced hyperlocomotion in wild-
type mice (a), but not in M4 knockout mice (b). Arrows indicate administra-
tion of vehicle or 24 (100 mgkgꢀ1 p.o. in 30% Captisol) and amphetamine
(2.5 mgkgꢀ1 s.c. in saline).
The greatest perceived challenge for developing an M4 PAM
is ensuring that an adequate safety margin can be established
when administrated alone and under co-administration with
donepezil, the current standard of care in AD patients. Because
of a lack of selectivity for M4 over the other mAChR subtypes,
xanomeline is associated with cholinergic side effects in
human patients (salivation, tearing, sedation, etc.).[4] To evalu-
ate the cholinergic side-effect potential of 24, rats were ob-
served following administration with the M2 agonist oxotre-
morine[31] or xanomeline, which served as positive controls, or
24, which was dosed at 10 and 100 mgkgꢀ1 i.p. These doses
produced unbound plasma levels that were 3.2-fold (80 nm)
and 7.9-fold (198 nm) greater, respectively, than the efficacious
unbound exposure observed in the rat amphetamine-induced
locomotion assay. The results indicated an absence of side ef-
fects at the 10 mgkgꢀ1 dose and minimal effects on sedation
and pupil constriction parameters, only, at the 100 mgkgꢀ1
dose (Table 6). Therefore, the improved selectivity of 24 for M4
Figure 4. a) Effect of M4 PAM 24 and b) xanomeline on amphetamine-in-
duced locomotor activity in Sprague–Dawley rats. Data are meanꢂSEM
total activity measured over 60 min immediately after amphetamine admin-
istration (1 mgkgꢀ1, s.c.), n=7–8 per group; *p<0.05, **p<0.001, significant
differences in activity relative to control vehicle V/V treated rats; ap<0.05,
aap<0.01 compared with activity in rats after amphetamine alone, V/am-
phetamine (one-way ANOVA, followed by Fisher’s LSD post-hoc test).
&
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
6
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Table 6. Comparison of cholinergic adverse events induced in rats 30 min after treatment with behaviorally active doses of oxotremorine, xanomeline, and
compound 24.[4]
Compound:
Dose [mgkgꢀ1]:
oxotremorine
0.3a
xanomeline
30b
24
10c
24
100c
salivation
tearing
piloerection
diarrhea
tremors
respiration
seizure-like activity
sedation
+ + (4/5)
+ + (2/5)
+ (4/5)
+ + (4/5)
+ (1/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
ꢀ (0/5)
+ (2/5)
+ (1/5)
ꢀ (0/5)
+ (3/5)
+ (1/5)
slight/periodic (2/5)
ꢀ (0/5)
ꢀ (0/5)
panting/labored (2/5)
“absent” look (2/5)
+ + (3/5)
full body (1/5)
+ + (3/5)
pupil constriction
+ + (5/5)
+ + (5/5)
Behaviors were observed over 3 min and scored as either not present (ꢀ), mild/occasionally present (+), intermediate severity/frequently present (+ +), or
severe/present constantly (+ + +) in X/5 rats per group. [a] 1 mLkgꢀ1 s.c. in saline; [b] 2 mLkgꢀ1 i.p. in saline; [c] 5 mLkgꢀ1 i.p. in 30% Captisol.
In vitro assays: FLIPR assays were used to assess human M1–5, rat
M4, and rhesus M4 activity and were performed as previously re-
ported.[32] M1, M3, and M5 were all natively Gq coupled. Human M4,
rat M4, rhesus M4, and human M2 cell lines were force coupled to
Gqi5 to drive calcium mobilization. PAM activity was assessed in
the presence of an EC20 concentration of ACh, which was deter-
mined prior to running each FLIPR assay. ACh titrations (starting
with 1 mm and diluting 3-fold down for 10 points on the curve)
were run on a separate plate, and the EC20 value was extrapolated
from curves drawn within the ScreenWorks software package (Mo-
lecular Devices, LLC, San Jose, CA, USA). Typical EC20 concentrations
for human M4 were in the single-digit nanomolar range.
versus other muscarinic receptors appears to be associated
with an improved safety profile relative to the nonselective
compound xanomeline.
Conclusions
In summary, pyridine HTS hit 1 was evolved into a series of
potent M4 PAMs. Comparison of M4 PAM potency, receptor-
subtype selectivity, and P-gp efflux susceptibility demonstrated
interplay between pyridine and pyrazole substituents, and
data visualization and multivariate analysis was used to dissect
this SAR. These efforts led to the identification of 24 as a selec-
tive M4 PAM that is efficacious in a preclinical measure of anti-
psychotic activity. Further optimization and pharmacological
profiling of this compound series will be disclosed in due
course.
Pharmacokinetic properties: The pharmacokinetic (PK) profile of
24 was evaluated in male Wistar Hanover rats following administra-
tion of a single 2 mgkgꢀ1 intravenous (i.v.) dose and a 10 mgkgꢀ1
oral (p.o.) dose. In addition, PK profile following a 10 mgkgꢀ1 intra-
peritoneal (i.p.) dose of 24 was evaluated in male Sprague–Dawley
rats. Intravenous and p.o. doses were formulated as a solution in a
mixture of DMSO, polyethylene glycol (PEG) 400 and water (in
20:60:20 proportions, respectively). For i.p. administration, 24 was
formulated in 30% Captisolꢃ. All animal studies were performed
according to the US National Institutes of Health (NIH) Guide for
the Care and Use of Laboratory Animals, and experimental proto-
cols were reviewed by the Merck Institutional Animal Care and Use
Committee.
Experimental Section
Compounds: Detailed synthetic procedures and characterization
data can be found in the Supporting Information and a previous
report.[32] All compounds were synthesized and tested at >95%
purity as determined by LC–MS. All reagents and solvents were of
commercial quality and used without further purification unless in-
dicated otherwise. All reactions were carried out under an inert at-
Animals were fasted overnight with free access to drinking water
and were fed 4 h after dosing. Intravenous administration to rats
was done via cannulae implanted in the jugular vein, and oral for-
mulations were administered by gavage. Blood samples were col-
lected serially in EDTA-coated tubes up to 8 h following i.v. or p.o.
dose administration, and up to 24 h following i.p. administration.
Plasma was separated by centrifugation and kept frozen at ꢀ708C
until analysis. The concentrations of test compounds in rat plasma
were determined by an LC–MS/MS assay following protein precipi-
tation and addition of an appropriate internal standard. Pharmaco-
kinetic parameters were obtained using noncompartmental analy-
sis using Watson LIMSꢃ.
1
mosphere of nitrogen. H NMR spectra were obtained on a Varian
Unity Inova 400 spectrometer or a Varian Unity Inova 500 spec-
trometer. Chemical shifts are reported in parts per million relative
to TMS as internal standard. Samples provided for accurate mass
measurement were taken up in methanol. The solutions were ana-
lyzed by use of electrospray ionization (ESI) or atmospheric pres-
sure chemical ionization (APCI) on either a Bruker Daltonics 3T or
7T Fourier transform ion cyclotron resonance (FTICR) mass spec-
trometer. External calibration was accomplished with polypropy-
lene glycol (425 or 750). Silica gel chromatography was carried out
with an ISCO CombiFlash purification system using ISCO silica gel
cartridges. Preparative reversed-phase HPLC was performed using
an Agilent 1200 series chromatograph with mass-directed collec-
tion using a Waters Sunfire C18 column (150ꢁ19 mm I.D.) with a
linear gradient over 15 min (95:5 to 5:100 H2O containing 0.10%
trifluoroacetic acid/acetonitrile.
Plasma protein binding studies: Plasma protein binding was eval-
uated by equilibrium dialysis using HTDialysis plates and dialysis
membranes with a 12–14 kDa MWCO (HTDialysis LLC, Groton, CT,
USA). Aliquots of plasma containing test compound at 2.5 mm con-
centration were added to one side of the dialysis membrane and
Dulbecco’s phosphate-buffered saline, 1ꢁ, without calcium, mag-
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
7
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
nesium, phenol red, pH 7.4 (Thermo Scientific, Waltham, MA, USA)
was added to the other side of the dialysis membrane. The equilib-
rium dialysis plate with the samples was incubated for 4 h at 378C,
under an atmosphere of 5% CO2. Following incubation, aliquots of
the dialyzed plasma samples and buffer samples were removed
from the plate, and proteins were precipitated by addition of ace-
tonitrile containing an internal standard, followed by centrifuga-
tion. The supernatant was removed and analyzed by LC–MS/MS
assay. The unbound fraction of test compounds in the plasma was
determined by dividing the compound peak area over the internal
standard area in the buffer by that in the dialyzed plasma samples.
rats by tail vein draws for plasma PK analysis, and then animals
were euthanized by CO2 overdose in an induction chamber. Mice
were assessed in a similar paradigm, but used Columbus Instru-
ments infrared beam-break arrays to assess locomotor activity
under home cage conditions not requiring habituation. Mouse i.p.
dose volume was 10 mLkgꢀ1
.
Plasma PK sample preparation: Blood was collected into 0.25 mL
K3EDTA vials and samples centrifuged at 10000 rpm for 2 min.
Plasma supernatant was decanted for HPLC analysis of 24 and xa-
nomeline.
Cholinergic adverse effects (AE) assay: On the day of testing, rats
(n=5 per group) were placed into individual empty rat cages and
acclimated for at least 30 min before compound administration. At
30 min post-dose, animals were observed over 3 min for the ap-
pearance of AEs listed in Table 6 in a modified Irwin procedure.
Data are recorded first as the number of rats per group that pre-
sented with the behavior (X/5 animals), and secondly as severity or
duration of the behavior in the presenting animals, where ꢀ is no
Determination of bidirectional permeability in the LLC PK-1 cell
line: P-glycoprotein (P-gp) transport was evaluated using similar
procedures to those described elsewhere.[29] LLC-PK1 cells and LLC-
PK1 cells expressing a cDNA encoding human MDR1 P-gp (human
LLC-MDR1) were obtained from The Netherlands Cancer Institute
(Amsterdam, The Netherlands) and used under a license agree-
ment. LLC-PK1 cells expressing a cDNA encoding rat Mdr1a (rat
LLC-Mdr1a) were made inhouse by stable transfection of LLC-PK1
cells with a rat Mdr1a cDNA. LLCPK1, human LLC-MDR1, and rat
LLC-Mdr1a cell lines were cultured in 96-well trans-well culture
plates (Millipore) at 378C in a humidified atmosphere of 5% CO2/
95% air and used in experiments after four days in culture. Solu-
tions of test compound (final concentration 1 mm) or verapamil
(1 mm, a prototypic P-gp substrate, PerkinElmer) was prepared in
Hank’s balanced salt solution (HBSS), 10 mm 4-(2-hydroxyethyl)-1-
piperrazineethanesulfonic acid (HEPES, pH 7.4) and 1.2 mm dextran
Texas red (to confirm monolayer integrity). Substrate solution
(150 mL) was added to either the apical (A) or the basolateral (B)
compartment of the culture plate, and buffer (150 mL; HBSS,
10 mm HEPES, pH 7.4) was added to the compartment opposite to
that containing the substrate. All incubations were run in triplicate.
Following 3 h incubation, 50 mL aliquots were taken from wells on
both sides and analyzed by LC–MS/MS. The experiment was per-
formed in triplicate. The reported apparent permeation (Papp) repre-
sents the average of the Papp for transport from A to B and Papp for
transport from B to A at t=3 h. The B!A/A!B ratio was calculat-
ed by dividing the Papp from B to A by the Papp from A to B.
presentation,
+ is small change/occasional, + + is medium
change/frequent presentation, and + + + is maximum change/
constant presentation of the behavior.
Data analysis: Total distance travelled or number of beam breaks
were used to assess hyperactivity behavior. The locomotor beam
box software recorded each animal’s activity by beam breaks over
time, and for rat data converted readouts to distance (cm) based
on distance between contiguous beams crossed. Distances or
beam breaks were averaged over 10 or 20 min time bins, and dif-
ferences between treatment groups was analyzed using SPSS soft-
ware. Comparisons of average distance or beam breaks per time
bin were analyzed by repeated measure ANOVA followed by post-
hoc LSD test.
Acknowledgements
The authors thank Michael Weber and Eric Johnson for develop-
ing and running the screening assay that led to the identification
of compound 1 and other novel chemical matter. We also thank
Ningning Guan for initially synthesizing 1 and submitting it to
the sample collection for use in screening. Janine Brouillette
helped with structure determination.
Animals for LMA and AE studies: All animal studies were conduct-
ed in accordance with the NRC Institute of Laboratory Animal Re-
sources Guide for the Care and Use of Laboratory Animals, and ap-
proved by the site IACUC (Merck & Co., Inc., West Point, PA, USA).
Studies used adult male Sprague–Dawley rats (200–280 g, Taconic
Biosciences, Germantown, NY, USA), or adult male M4-knockout
mice (20–25 g, Taconic Biosciences). Rats were housed two per
cage, and mice 4–5 per cage, and animals were at controlled tem-
perature with food and water available ad libitum under a regular
12-h light/dark cycle (18:00 lights off, 06:00 lights on). Animals
were acclimated to the facility for at least one week prior to use.
All studies were performed by investigators blinded to animal
treatment group.
Conflict of interest
The authors declare no conflict of interest.
Keywords: muscarinic acetylcholine receptor 4 (M4) · positive
allosteric modulator (PAM) · psychosis · pyridines
Rat and mouse locomotor activity (LMA): On the day of experi-
ment, rats were placed individually in infrared beam-break activity
boxes (Med Associates Inc., Fairfax, VT, USA) housed in a sound-at-
tenuating chamber. Animals were allowed to habituate for 30 min
before dosing i.p. with test compound 24 or xanomeline, or vehi-
cle (30% Captisol or saline, respectively, 5 mLkgꢀ1). Rats were then
returned to the activity chambers for an additional 15–30 min ha-
bituation before receiving a s.c. dose of psychostimulant ampheta-
mine or vehicle (saline). Locomotor activity was recorded through-
out, and for an additional 60 min after amphetamine administra-
tion. At the end of monitoring, blood was collected from conscious
[1] O. P. Tible, F. Riese, E. Savaskan, G. A. Von, Ther. Adv. Neurol. Disord.
2017, 10, 297–309.
[2] N. C. Bodick, W. W. Offen, A. I. Levey, N. R. Cutler, S. G. Gauthier, A. Satlin,
H. E. Shannon, G. D. Tollefson, K. Rasmussen, F. P. Bymaster, D. J. Hurley,
W. Z. Potter, S. M. Paul, Arch. Neurol. 1997, 54, 465–473.
[3] A. Shekhar, W. Z. Potter, J. Lightfoot, J. Lienemann, S. Dube, C. Mallinck-
rodt, F. P. Bymaster, D. L. McKinzie, C. C. Felder, Am. J. Psychiatry 2008,
165, 1033–1039.
[4] N. R. Mirza, D. Peters, R. G. Sparks, CNS Drug Rev. 2003, 9, 159–186.
[5] A. M. Bender, C. K. Jones, C. W. Lindsley, ACS Chem. Neurosci. 2017, 8,
435–443.
&
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
8
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[6] J. J. Sramek, D. Pharm, D. J. Hurley, T. S. Wardle, J. H. Satterwhite, J.
Hourani, F. Dies, N. R. Cutler, J. Clin. Pharmacol. 1995, 35, 800–806.
[7] F. P. Bymaster, D. L. McKinzie, C. C. Felder, J. Wess, Neurochem. Res.
2003, 28, 437–442.
[8] M. L. Woolley, H. J. Carter, J. E. Gartlon, J. M. Watson, L. A. Dawson, Eur.
J. Pharmacol. 2009, 603, 147–149.
Melancon, J. C. Tarr, F. W. Byers, J. Wess, M. E. Duggan, J. Dunlop, M. W.
Wood, N. J. Brandon, M. R. Wood, C. W. Lindsley, P. J. Conn, C. K. Jones,
ACS Chem. Neurosci. 2014, 5, 920–942.
[19] G. Liwicki, S. Mack, P. K. Ruprah, J. Reeves (Takeda Pharmaceutical Com-
pany Limited, Japan), Int. PCT Pub. No. WO2018066718 A1, 2018.
[20] L. Zhang, C. R. Butler, E. M. Beck, M. A. Brodney, M. F. Brown, L. A.
McAllister, E. A. LaChapelle, A. M. Gilbert (Pfizer Inc., USA), Int. PCT Pub.
No. WO2018002760 A1, 2018.
[21] T. Huynh, C. Valant, I. T. Crosby, P. M. Sexton, A. Christopoulos, B. Capua-
no, J. Med. Chem. 2013, 56, 8196–8200.
[22] J. Kolpak, P. J. Connolly, V. S. Lobanov, D. K. Agrafiotis, J. Chem. Inf.
Model. 2009, 49, 2221–2230.
[9] W. Y. Chan, D. L. McKinzie, S. Bose, S. N. Mitchell, J. M. Witkin, R. C.
Thompson, A. Christopoulos, S. Lazareno, N. J. M. Birdsall, F. P. Bymaster,
C. C. Felder, Proc. Natl. Acad. Sci. USA 2008, 105, 10978–10983.
[10] T. M. Bridges, E. P. LeBois, C. R. Hopkins, M. R. Wood, C. K. Jones, P. J.
Conn, C. W. Lindsley, Drug News Perspect. 2010, 23, 229–240.
[11] C. K. Jones, N. Byun, M. Bubser, Neuropsychopharmacology 2012, 37,
16–42.
[12] M. R. Wood, M. J. Noetzel, B. J. Melancon, M. S. Poslusney, K. D. Nance,
M. A. Hurtado, V. B. Luscombe, R. L. Weiner, A. L. Rodriguez, A. Lamsal,
S. Chang, M. Bubser, A. L. Blobaum, D. W. Engers, C. M. Niswender, C. K.
Jones, N. J. Brandon, M. W. Wood, M. E. Duggan, P. J. Conn, T. M.
Bridges, C. W. Lindsley, ACS Med. Chem. Lett. 2017, 8, 233–238.
[13] J. C. Tarr, M. R. Wood, M. J. Noetzel, J. L. Bertron, R. L. Weiner, A. L. Rodri-
guez, A. Lamsal, F. W. Byers, S. Chang, H. P. Cho, C. K. Jones, C. M. Nis-
wender, M. W. Wood, N. J. Brandon, M. E. Duggan, P. J. Conn, T. M.
Bridges, C. W. Lindsley, Bioorg. Med. Chem. Lett. 2017, 27, 2990–2995.
[14] B. J. Melancon, M. R. Wood, M. J. Noetzel, K. D. Nance, E. M. Engelberg,
C. Han, A. Lamsal, S. Chang, H. P. Cho, F. W. Byers, M. Bubser, C. K. Jones,
C. M. Niswender, M. W. Wood, D. W. Engers, D. Wu, N. J. Brandon, M. E.
Duggan, P. J. Conn, T. M. Bridges, C. W. Lindsley, Bioorg. Med. Chem. Lett.
2017, 27, 2296–2301.
[23] A. M. Wassermann, P. Haebel, N. Weskamp, J. Bajorath, J. Chem. Inf.
Model. 2012, 52, 1769–1776.
[24] M. D. Shultz, Bioorg. Med. Chem. Lett. 2013, 23, 5980–5991.
[25] R. Zhang, M. Kavana, Expert Opin. Drug Discovery 2015, 10, 763–780.
[26] A. E. Brady, C. K. Jones, T. M. Bridges, J. P. Kennedy, A. D. Thompson, J. U.
Heiman, M. L. Breininger, P. R. Gentry, H. Yin, S. B. Jadhav, J. K. Shirey,
P. J. Conn, C. W. Lindsley, J. Pharmacol. Exp. Ther. 2008, 327, 941.
[27] K. J. Stanhope, N. R. Mirza, M. J. Bickerdike, J. L. Bright, N. R. Harrington,
M. B. Hesselink, G. A. Kennett, S. Lightowler, M. J. Sheardown, R. Syed,
R. L. Upton, G. Wadsworth, S. M. Weiss, A. Wyatt, J. Pharmacol. Exp. Ther.
2001, 299, 782.
[28] C. Tang, Y. Kuo, N. T. Pudvah, J. D. Ellis, M. S. Michener, M. Egbertson,
S. L. Graham, J. J. Cook, J. H. Hochman, T. Prueksaritanont, Biochem.
Pharmacol. 2009, 78, 642–647.
[29] K. A. Lyons, X. Shen, Z. Yao, K. Bleasby, G. Chan, M. Hafey, X. Li, S. Xu,
G. M. Salituro, L. H. Cohen, W. Tang, Xenobiotica 2009, 39, 687–693.
[30] T. Ohe, M. Sato, S. Tanaka, N. Fujino, M. Hata, Y. Shibata, A. Kanatani, T.
Fukami, M. Yamazaki, M. Chiba, Y. Ishii, Drug Metab. Dispos. 2003, 31,
1251.
[15] M. F. Long, J. L. Engers, S. Chang, X. Zhan, R. L. Weiner, V. B. Luscombe,
A. L. Rodriguez, H. P. Cho, C. M. Niswender, T. M. Bridges, P. J. Conn,
D. W. Engers, C. W. Lindsley, Bioorg. Med. Chem. Lett. 2017, 27, 4999–
5001.
[16] J. M. Salovich, P. N. Vinson, D. J. Sheffler, A. Lamsal, T. J. Utley, A. L. Blo-
baum, T. M. Bridges, U. Le, C. K. Jones, M. R. Wood, J. S. Daniels, P. J.
Conn, C. M. Niswender, C. W. Lindsley, C. R. Hopkins, Bioorg. Med. Chem.
Lett. 2012, 22, 5084–5088.
[17] M. R. Wood, M. J. Noetzel, M. S. Poslusney, B. J. Melancon, J. C. Tarr, A.
Lamsal, S. Chang, V. B. Luscombe, R. L. Weiner, H. P. Cho, M. Bubser, C. K.
Jones, C. M. Niswender, M. W. Wood, D. W. Engers, N. J. Brandon, M. E.
Duggan, P. J. Conn, T. M. Bridges, C. W. Lindsley, Bioorg. Med. Chem. Lett.
2017, 27, 171–175.
[31] M. McKinney, J. H. Miller, V. A. Gibson, L. Nickelson, S. Aksoy, Mol. Phar-
macol. 1991, 40, 1014.
[32] J. J. Acton, J. Bao, M. Egbertson, X. Gao, S. T. Harrison, S. L. Knowles, C.
Li, M. M.-C. Lo, R. D. Mazzola, Jr., Z. Meng, M. T. Rudd, O. Selyutin, D. M.
Tellers, L. Tong, J. M.-C. Wai (Merck Sharp & Dohme Corp., USA; MSD
R&D China Co., Ltd.), Int. PCT Pub. No. WO2017107089 A1, 2017.
Manuscript received: February 11, 2019
[18] M. Bubser, T. M. Bridges, D. Dencker, R. W. Gould, M. Grannan, M. J.
Noetzel, A. Lamsal, C. M. Niswender, J. S. Daniels, M. S. Poslusney, B. J.
Version of record online: && &&, 0000
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
9
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
J. W. Schubert,* S. T. Harrison,
J. Mulhearn, R. Gomez, R. Tynebor,
K. Jones, J. Bunda, B. Hanney,
J. M.-C. Wai, C. Cox, J. A. McCauley,
J. M. Sanders, B. Magliaro, J. O’Brien,
N. Pajkovic, S. L. Huszar Agrapides,
A. Taylor, A. Gotter, S. M. Smith,
J. Uslaner, S. Browne, S. Risso,
M. Egbertson
Novel series of M4 PAMs: Optimization
of series of 2,3-di- and 2,3,6-trisubstitut-
ed pyridines led to the potent, receptor-
subtype-selective, brain-penetrant posi-
tive allosteric modulator (PAM) 24 with
efficacy in rodent locomotor activity
assays. Comparison of cholinergic ad-
verse effects in rats treated separately
with 24 and the M4 agonist xanomeline
suggests that a receptor-subtype-selec-
tive PAM offers the potential for an im-
proved safety profile.
&& – &&
Discovery, Optimization, and
Biological Characterization of 2,3,6-
Trisubstituted Pyridine-Containing M4
Positive Allosteric Modulators
&
ChemMedChem 2019, 14, 1 – 11
www.chemmedchem.org
10
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!