Novel M4 positive allosteric modulators derived from questioning the role and impact of a presumed intramolecular hydrogen-bonding motif in β-amino carboxamide-harboring ligands

Article abstract of DOI:10.1016/j.bmcl.2018.12.039

This letter describes a focused exercise to explore the role of the β-amino carboxamide moiety found in all of the first generation M₄ PAMs and question if the NH₂ group served solely to stabilize an intramolecular hydrogen bond (IMHB) and enforce planarity. To address this issue (and to potentially find a substitute for the β-amino carboxamide that engendered P-gp and contributed to solubility liabilities), we removed the NH₂, generating des-amino congeners and surveyed other functional groups in the β-position. These modifications led to weak M₄ PAMs with poor DMPK properties. Cyclization of the β-amino carboxamide moiety by virtue of a pyrazole ring re-enforced the IMHB, led to potent (and patented) M₄ PAMs, many as potent as the classical bicyclic β-amino carboxamide analogs, but with significant CYP1A2 inhibition. Overall, this exercise indicated that the β-amino carboxamide moiety most likely facilitates an IMHB, and is essential for M₄ PAM activity within classical bicyclic M₄ PAM scaffolds.

Full text of DOI:10.1016/j.bmcl.2018.12.039

Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Novel M4 positive allosteric modulators derived from questioning the role
and impact of a presumed intramolecular hydrogen-bonding motif in β-
amino carboxamide-harboring ligands
Michael S. Poslusneya,b, James M. Salovicha,b, Michael R. Wooda,b, Bruce J. Melancona,b
,
Katrina A. Bollingera,b, Vincent B. Luscombea,b, Alice L. Rodrigueza,b, Darren W. Engersa,b
,
⁎
Thomas M. Bridgesa,b, Colleen M. Niswendera,b,e, P. Jeffrey Conna,b,e, Craig W. Lindsleya,b,c,d,
a Vanderbilt Center for Neuroscience Drug Discovery, Vanderbilt University, School of Medicine, Nashville, TN 37232, USA
b Department of Pharmacology, Vanderbilt University, School of Medicine, Nashville, TN 37232, USA
c Department of Chemistry, Vanderbilt University, Nashville, TN 37232, USA
d Department of Biochemistry, Vanderbilt University, Nashville, TN 37232, USA
e Vanderbilt Kennedy Center, Vanderbilt University School of Medicine, Nashville, TN 37232, USA
A R T I C L E I N F O
A B S T R A C T
Keywords:
This letter describes a focused exercise to explore the role of the β-amino carboxamide moiety found in all of the
first generation M4 PAMs and question if the NH2 group served solely to stabilize an intramolecular hydrogen
bond (IMHB) and enforce planarity. To address this issue (and to potentially find a substitute for the β-amino
carboxamide that engendered P-gp and contributed to solubility liabilities), we removed the NH2, generating
des-amino congeners and surveyed other functional groups in the β-position. These modifications led to weak M4
PAMs with poor DMPK properties. Cyclization of the β-amino carboxamide moiety by virtue of a pyrazole ring
re-enforced the IMHB, led to potent (and patented) M4 PAMs, many as potent as the classical bicyclic β-amino
carboxamide analogs, but with significant CYP1A2 inhibition. Overall, this exercise indicated that the β-amino
carboxamide moiety most likely facilitates an IMHB, and is essential for M4 PAM activity within classical bicyclic
M4 PAM scaffolds.
M4
Muscarinic acetylcholine receptor
Positive allosteric modulator (PAM)
Structure-Activity Relationship (SAR)
Tricycle
Positive allosteric modulators (PAMs) of the muscarinic acetylcho-
line receptor subtype 4 (M4) have garnered a great deal of attention and
interest as next generation antipsychotics via a novel mechanism of
action.1–18 From the first account of an M4 PAM ligand, 1, the β-amino
carboxamide moiety (circled in red) was present, and this functionality
was maintained through multiple iterations of advanced ligands 2–5
(either thieno[2,3-b]pyridine 2-carboxamides, e.g., 1 and 2, or thieno
[2,3-c]pyridazines 6-carboxamides, e.g., 3–5) enroute to preclinical
development candidates (Fig. 1).1–18 While potent and highly selective
M4 PAMs resulted that provided critical in vivo target validation and
enabled deep mechanistic studies, this chemotype engendered poor
physiochemical properties and variable species-specific P-gp efflux li-
abilities. Only recently have alternative M4 PAM chemotypes been re-
ported.19,20 We sought to better understand the role of the β-amino
carboxamide moiety, and question if this non-basic ‘spectator’ func-
tionality served to enforce an intramolecular hydrogen bond (IMHB)
and engender planarity.21 In this Letter, we describe a series of studies
and novel ligands that support the IMHB as a crucial pharmacophore
for M4 PAM activity.
To address this IMHB issue (and to potentially find a substitute for
the β-amino carboxamide that engendered P-gp and solubility li-
abilities), we first removed the NH2, generating des-amino congeners as
well as ligands with other groups replacing the NH2 moiety. Second, we
re-enforced the planarity of the presumed IMHB with a novel tricycle
core, and assessed M4 PAM activity (Fig. 2). Combined, the necessity of
the IMHB and an extended planar disposition was clear.
The synthesis of des-amino and alternatively functionalized con-
geners 6 is shown in Scheme 1 for the thieno[2,3-b]pyridine scaffold
(similar chemistry was used for thieno[2,3-c]pyridazine analogs).1–18
Condensation of α-bromoacetate with commercial thiopyridone 8 af-
fords 9 in 66% yield. A subsequent Sandmeyer reaction converts the
NH2 to the corresponding bromide 10 in 81% yield. The bromide could
then be reduced to afford the des-amino derivative 14, which could
then be saponified and coupled with preferred amines to provide analog
⁎ Corresponding author at: Vanderbilt Center for Neuroscience Drug Discovery, Vanderbilt University, School of Medicine, Nashville, TN 37232, USA.
E-mail address: craig.lindsley@vanderbilt.edu (C.W. Lindsley).
https://doi.org/10.1016/j.bmcl.2018.12.039
Received 19 November 2018; Received in revised form 13 December 2018; Accepted 17 December 2018
0960-894X/©2018ElsevierLtd.Allrightsreserved.
Pleasecitethisarticleas:Poslusney,M.S.,Bioorganic&MedicinalChemistryLetters,https://doi.org/10.1016/j.bmcl.2018.12.039

M.S. Poslusney et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Fig. 1. Prototypical M4 PAM chemotypes featuring a
conserved β-amino carboxamide moiety, presumed
to participate in an intramolecular hydrogen bond
(IMHB), resulting in an extended planar conforma-
tion. However, this presumed essential pharmaco-
phore also engendered poor physiochemical proper-
ties and unpredictable P-gp liabilities.
potency (EC50 = 5.1 µM). Substitution of the NH2 for a CH3 group
(17b) again loses potency (EC50 = 2.7 µM), but improved activity re-
lative to the des-amino congener 17a, suggesting free rotation about the
amide bond is detrimental to M4 PAM activity. Neither alternative,
preferred amides nor larger alkyl groups (e.g., ethyl 17h) could im-
prove M4 PAM potency to that of the β-amino carboxamide-based li-
gands.1–18
As the thieno[2,3-c]pyridazine core was more potent than the
thieno[2,3-b]pyridine scaffold amongst the classical β-amino carbox-
amide M4 PAMs,1–18 we also evaluated derivatives 18 (Table 2). Here,
parent NH2-baring 4 was a potent M4 PAM (EC50 = 84 nM), but the des-
amino analog 18a lost ∼18-fold potency (EC50 = 1.5 µM). Methyl
(18b), ethyl (18c) and cyclopropyl (18d) congeners were all of similar,
weak M4 PAM potency, indicating the necessity of the IMHB and an
extended planar disposition through the β-amino carboxamide. More-
over, all congeners 17 and 18 displayed inferior DMPK properties (e.g.,
high predicted hepatic clearance (rat CLhep > 68 mL/min/kg), high
protein binding (rat fu,plasma < 0.005) and mixed CYP450 inhibition
profiles) as compared to 2–5. Thus, efforts shifted to focus on the
generation of three distinct tricyclic cores, each incorporating a pyr-
azole to mimic the β-amino carboxamide, to re-enforce the IMHB.
Synthesis of the first desired tricyclic core, a 1H-pyrazolo[3′,4′:4,5]
thieno[2,3-b]pyridine, was accomplished in four linear steps (Scheme
Fig. 2. The prototypical M4 PAM scaffolds 1–5, and plans to assess the role of
the IMHB, by either removing the β-NH2 and surveying alternative moieties
that remove the IMHB capability as in 6, or by cyclization through a pyrazole
ring, affording tricycles 7, that reinforce the IMHB.
17a (wherein R1 = H). Alternatively, 10 could undergo Suzuki cou-
pling reactions to install small alkyl moieties to replace the NH2 group,
followed by ester hydrolysis and amide couplings to deliver analogs 17.
As shown in Table 1, the classical β-amino carboxamide M4 PAM 2
is a moderately potent PAM (EC50 = 680 nM),1–18 but removal of the
NH2 to provide 17a (R1 = H) leads to a significant diminution in
Scheme 1. Synthesis of des-methyl analogs 17 of 6.
Reagents and conditions: (a) α-bromoethyl acetate,
aq. KOH, DMF, rt, 66%; (b) CuBr2, t-BuONO, CH3CN,
65 °C, 81%; (c) Bu3P, Pd(OAc)2, Et3N, HCOOH, DMF,
100 °C, 45%; (KOH, EtOH/H2O, 65 °C, 85–90%; (d)
KF3B-R1, PdCl2(dppf)CH2Cl2, Cs2CO3, THF/H2O,
140 °C mw, 60–67%; (e) KOH, EtOH/H2O, rt,
85–90%; (f) NH2R2, HATU, DIEA, DMF, rt, 15–25%.
2

M.S. Poslusney et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 1
Table 2
Structure and hM4 PAM Activities of analogs 17.
Structure and hM4 PAM activities of analogs 18.
Cmpd
R1
R2
hM4
Cmpd
R1
R2
hM4
EC50 (µM)a
[% ACh Max]
EC50 (µM)a
[% ACh Max]
2
NH2
H
0.68
[89
0.01
5]
4
NH2
H
0.08
[75
0.01
2]
17a
17b
17c
5.1
1
18a
18b
18c
18d
1.5
0.07
0.2]
[53
0.3]
[66
CH3
CH3
2.7
0.3
5]
CH3
2.1
0.03
2]
[49
[56
> 10
[29
CH2CH3
2.2
0.5
2]
1]
[73
17d
17e
CH3
CH3
5.9
2
5.6
1
[50
2]
[35
1]
2.1
0.3
4]
18e
> 10
[38
[56
3]
a
17f
CH3
CH3
1.5
0.3
3]
For SAR determination, calcium mobilization human M4/Gqi5 assays were
performed n = 1 independent times in triplicate with an EC20 fixed con-
centration of acetylcholine.
[45
17 g
6.4
1
[58
2]
17 h
CH2CH3
2.7
0.3
2]
[57
a
For SAR determination, calcium mobilization human M4/Gqi5 assays were
performed n = 1 independent times in triplicate with an EC20 fixed con-
centration of acetylcholine.
2). Starting once again from thiopyridone 8, a condensation with 2-
chloroacetonitrile provides bicycle 19 in 96% yield. Next, our standard
Sandmeyer conditions give bromide 20, which upon treatment with
hydrazine under microwave heating delivers the desired amino tricycle
21 in 98% yield. Finally, a reductive amination protocol affords the first
of three sets of tricyclic analogs 22 for evaluation as M4 PAMs.
Relative once again to M4 PAM 2,1–18 analogs 22 were less potent,
but by less than 3-fold – an encouraging find for such a significant
change in structure (Table 3). For example, the direct analog of 2, PAM
22a was a moderately potent M4 PAM (EC50 = 2.2 µM); however, this
modification led to a poor in vitro DMPK profile (22a: CLhep = 66 mL/
min/kg, rat fu < 0.002, CYP450 < 5 µM). These data led us to consider
two modifications: 1) a more highly functionalized pyridine core
(known historically to improve PAM potency) and 2) cap of the NH of
the pyrazole in hopes of engendering CNS penetration.
Scheme 2. Synthesis of tricyclic analogs 22 of 7. Reagents and conditions: (a)
2-chloroacetonitrile, 10% KOH, DMF, 96%; (b) CuBr2, t-BuONO, CH3CN, 60 °C,
65%; (c) hydrazine, DMSO, 150 °C mw, 98%; (d) RCHO, NaB(OAc)3H, AcOH,
ClCH2CH2Cl, 73–90%.
initial SAR was disappointing, as classical, preferred benzyl amide
analogs 27a-c were inactive (EC50s > 10 µM). Interestingly, a cyclo-
propyl methyl derivative 27d proved to be
a potent M4 PAM
(EC50 = 230 nM), as did a furyl methyl congener 27e (EC50 = 120 nM),
and equipotent to first generation bicyclic PAMs 1 and 2. While ex-
citing, 27d and 27e suffered poor, non-advanceable DMPK profiles (rat
predicted CLhep > 67 mL/min/kg, rat fu,plasmas < 0.001) and potent
CYP450 inhibition, particularly at 1A2 (IC50s < 100 nM) and 2C9
(IC50s ∼ 1.0 µM). Thus, tricycle 27 offered mixed results. On the one
hand, an N-Me tricycle restored M4 PAM potency and supported the
role of the IMHB in 1–5,1–18 while, on the other hand, these novel cores
suffer from poor disposition, precluding advancement.
Scheme 3 shows the route to access the more highly substituted and
N-Me-pyrazole capped tricyclic variants 27. The chemistry follows that
of 22 up to intermediate 25. Condensation of 25 with methylhydrazine,
under microwave irradiation, provides the tricycle in 95% yield. Once
again, a final reductive amination protocol affords fully elaborated
tricyclic analogs 27 for evaluation as M4 PAMs. As shown in Table 4,
Lastly, we wanted to explore a pyridazine-based tricycle, e.g., a 1H-
pyrazolo[3′,4′:4,5]thieno[2,3-c]pyridazine core, as classical M4 PAMs
3–5 possessed this heterocycle and had advanced the farthest in
3

M.S. Poslusney et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 3
Table 4
Structure and hM4 PAM activities of analogs 22.
Structure and hM4 PAM activities of analogs 27.
Cmpd
R
hM4
Cmpd
R
hM4
EC50 (µM)a
[% ACh Max]
EC50 (µM)a
[% ACh Max]
27a
> 10
22a
2.2
0.4
1]
[51
4]
[67
27b
27c
> 10
[39
22b
22c
1.0
0.1
2]
1]
3]
[66
> 10
[56
1.1
0.1
2]
[62
27d
27e
0.23
[67
0.02
6]
22d
22e
2.0
0.5
3]
[67
0.12
0.02
3.8
0.8
[59
2]
[65
3]
22f
1.2
0.2
5]
[76
27a
> 10
[51
4]
27b
27c
> 10
[39
a
For SAR determination, calcium mobilization human M4/Gqi5 assays were
performed n = 1 independent times in triplicate with an EC20 fixed con-
centration of acetylcholine.
1]
3]
> 10
[56
27d
27e
0.23
[67
0.02
6]
0.12
0.02
[59
2]
a
For SAR determination, calcium mobilization human M4/Gqi5 assays were
performed n = 1 independent times in triplicate with an EC20 fixed con-
centration of acetylcholine.
Scheme 3. Synthesis of tricyclic analogs 27 of 7. Reagents and conditions: (a)
2-chloroacetonitrile, 10% KOH, DMF, 91%; (b) CuBr2, t-BuONO, CH3CN, 60 °C,
75%; (c) methylhydrazine, DMSO, 125 °C mw, 95%; (d) RCHO, NaB(OAc)3H,
AcOH, ClCH2CH2Cl, 73–83%.
Scheme 4. Synthesis of tricyclic analogs 30 of 7. Reagents and conditions: (a) i.
Na2S, t-BuOH, ii. 2-chloroacetonitrile, NaOAc, EtOH, rt; (b) CuBr2, t-BuONO,
CH3CN, 60 °C; (c) methylhydrazine, DMSO, 125 °C mw, 36–58% over 4 steps;
(d) RCHO, NaB(OAc)3H, AcOH, ClCH2CH2Cl, 27–56%.
development. To this end, commercial pyridazine 28 was first subjected
to an SN2 reaction with Na2S, followed by condensation with 2-chlor-
oacetonitrile to deliver 29. Repetition of the Sandmeyer and methyl
hydrazine condensation sequences, followed by reductive amination
provides the desired, putative M4 PAMs 30 (Scheme 4).
As shown in Table 5, tricycle analogs 30 were potent M4 PAMs,
uniformly affording sub-micromolar PAM EC50s. Analog 30a, harboring
the classic PMB amide,1–18 was a potent M4 PAM (EC50 = 44 nM) with
measurable CNS penetration in rat (rat Kp = 0.25), but displayed high
predicted hepatic clearance (rat CLhep = 67.8 mL/min/kg) and modest
protein binding (rat fu = 0.02). Another preferred benzyl amide con-
gener, 30b, was potent (EC50 = 160 nM), with improved protein
binding (fu = 0.05), and a clean CYP450 profile (IC50s > 30 µM);
however, the polarity that improved these DMPK properties led to poor
brain penetration in rat (brain concentration BLQ) and rendered 30b a
human P-gp substrate (ER > 10). The remaining analogs 30 also dis-
played DMPK profiles not suitable for further advancement towards
clinical candidates; however, they affirmed the role of the IMHB in 1–5,
and that novel tricycles could be constructed with equipotent M4 PAM
activity. Overall, all three series of tricycles had favorable physio-
chemical properties (MW < 400, clogPs < 4, CNS MPO scores > 5,22
4

M.S. Poslusney et al.
Bioorganic&MedicinalChemistryLettersxxx(xxxx)xxx–xxx
Table 5
carboxamide-based M4 PAMs. By synthesizing and assessing des-amino
and alternatively functionalized ligands, the importance of the IMHB
was clear. Re-enforcement of the IMHB via novel tricyclic analogs ul-
timately led to very potent M4 PAMs and further validation of the IMHB
of the β-amino carboxamide moiety as a requisite pharmacophore for
M4 PAM activity. While tricycles allowed entry into novel chemical
space23,24 from which to access M4 PAMs, disposition was favorable, yet
variable. Further efforts based on these findings to develop additional,
novel M4 PAMs are underway, and results will be reported in due
course.
Structure and hM4 PAM activities of analogs 30.
Cmpd
R
hM4
EC50 (µM)a
[% ACh Max]
Acknowledgments
30a
30b
0.04
[79
0.003
4]
We thank the NIH and NIMH for funding (U19MH097056). We also
thank William K. Warren, Jr. and the William K. Warren Foundation
who funded the William K. Warren, Jr. Chair in Medicine (to C.W.L.).
0.16
[58
0.003
1]
References
1. Brady AE, Jones CK, Bridges TM, et al. J Phar Exp Ther. 2008;327:941.
2. Chan WY, McKinzie DL, Bose S, et al. Proc Natl Acad Sci USA. 2008;105:10978.
3. Kennedy JP, Bridges TM, Gentry PR, et al. ChemMedChem. 2009;4:1600.
4. Salovich JM, Vinson PN, Sheffler DJ, et al. Bioorg Med Chem Lett. 2012;22:5084.
5. Byun NE, Grannan M, Bubser M, et al. Neuropsychopharmacology. 2014;39:1578.
6. Bubser M, Bridges TM, Dencker D, et al. ACS Chem Neurosci. 2014;5:920.
7. Gould RW, Nedelcovych MT, Gong X, et al. Neuropharmacology. 2016;102:244.
8. Wood MR, Noetzel MJ, Engers JL, et al. Bioorg Med Chem Lett. 2016;26:3029.
9. Wood MR, Noetzel MJ, Tarr JC, et al. Bioorg Med Chem Lett. 2016;26:4282.
10. Wood MR, Noetzel MJ, Poslusney MS, et al. Bioorg Med Chem Lett. 2017;27:171.
11. Wood MR, Noetzel MJ, Melancon BJ, et al. ACS Med Chem Lett. 2017;8:233.
12. Melancon BJ, Wood MR, Noetzel MJ, et al. Bioorg Med Chem Lett. 2017;27:2296.
13. Tarr CJ, Wood MR, Noetzel MJ, et al. Bioorg Med Chem Lett. 2017;27:2900.
14. Tarr CJ, Wood MR, Noetzel MJ, et al. Bioorg Med Chem Lett. 2017;27:5179.
15. Gould RW, Grannan MD, Gunter BW, et al. Neuropharmacology. 2018;128:492.
16. Huynh T, Valant C, Crosby IT, et al. ACS Chem Neurosci. 2015;6:838.
17. Foster DJ, Wilson JM, Remke DH, et al. Neuron. 2016;91:1244.
30c
30d
0.62
[84
0.05
3]
0.18
0.02
[102
3]
30e
0.69
[91
0.08
2]
30f
0.06
[91
0.01
3]
a
For SAR determination, calcium mobilization human M4/Gqi5 assays were
performed n = 1 independent times in triplicate with an EC20 fixed con-
centration of acetylcholine.
18. Moehle MS, Pancani T, Byun N, et al. Neuron. 2017;96:1358.
19. Bewley BR, Spearing PK, Weiner RL, et al. Bioorg Med Chem Lett. 2017;27:4274.
20. Long MF, Engers JL, Chang S, et al. Bioorg Med Chem Lett. 2017;27:4999.
21. Engers DW, Blobaum AL, Gogliotti RD, et al. ACS Chem Neurosci. 2016;7:1192.
22. Wager TT, Hou X, Verhoest PR, Villalobos A. ACS Chem Neurosci. 2016;7:767.
23. Lindsley CW, Conn PJ, Wood MR, Salovich JM, WO0364409. 2014.
24. Conn PJ, Lindsley CW, Wood MR, et al. WO0357615. 2014.
tPSAs 70–90 Å2), suggesting further optimization efforts in this che-
mical space is warranted.
In summary, we examined the role and impact of a putative in-
tramolecular hydrogen bond amongst the classical series of β-amino
5