Radiolabeled Dibenzodiazepinone-Type Antagonists Give Evidence of Dualsteric Binding at the M2 Muscarinic Acetylcholine Receptor

Article abstract of DOI:10.1021/acs.jmedchem.6b01892

The dualsteric ligand approach, aiming at ligands with improved subtype selectivity, has been increasingly applied to muscarinic receptors (MRs). In this article, we present the synthesis and characterization of a M₂R subtype-preferring radiola

Full text of DOI:10.1021/acs.jmedchem.6b01892

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Radiolabeled dibenzodiazepinone-type antagonists give evidence
2
of dualsteric binding at the M muscarinic acetylcholine receptor
Andrea Pegoli, Xueke She, David Wifling, Harald Hübner, Günther Bernhardt, Peter Gmeiner, and Max Keller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01892 • Publication Date (Web): 07 Apr 2017
Downloaded from http://pubs.acs.org on April 8, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Radiolabeled dibenzodiazepinoneꢀtype antagonists
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
give evidence of dualsteric binding at the M2
muscarinic acetylcholine receptor
†
†
†
‡
†
Andrea Pegoli, Xueke She, David Wifling, Harald Hübner, Günther Bernhardt, Peter
‡
,†
Gmeiner, and Max Keller*
†
Institute of Pharmacy, Faculty of Chemistry and Pharmacy, University of Regensburg,
Universitätsstr. 31, Dꢀ93053 Regensburg, Germany.
‡
Department of Chemistry and Pharmacy, Emil Fischer Center, Friedrich Alexander
University, Schuhstr. 19, Dꢀ91052 Erlangen, Germany.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
The dualsteric ligand approach, aiming at ligands with improved subtype selectivity, has been
increasingly applied to muscarinic receptors (MRs). In this article we present the synthesis
2
and characterization of a M R subtypeꢀpreferring radiolabeled dibenzodiazepinoneꢀtype
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
3
antagonist ([ H]UNSWꢀMK259, [ H]19) and its homodimeric analog [ H]URꢀAP060,
3
([ H]33). Saturation binding studies at the M
2
R, using the orthosteric antagonist atropine to
determine unspecific binding, proved that the monomeric and the dimeric compound bind to
the orthosteric binding site (apparent K : 0.87 and 0.24 nM, respectively). Various binding
R, for instance, saturation binding experiments in the
d
3
3
studies with [ H]19 and [ H]33 at the M
2
presence of the allosteric MR modulators W84 (8) or LY2119620 (9) (Schildꢀlike analysis)
suggested a competitive mechanism between the allosteric modulator and the
dibenzodiazepinone derivatives, and thus a dualsteric binding mode of both 19 and 33. This
2
was consistent with the results of M R MD simulations (≥2 µs) performed with 19 and 33.
ACS Paragon Plus Environment

Page 3 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
In humans, the family of muscarinic acetylcholine receptors (M receptors, MRs) comprises
five subtypes (M RꢀM R), which belong to the GPCR superfamily class A and mediate the
action of the neurotransmitter acetylcholine in the CNS as well as in the periphery. For
1
5
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
instance, the M
2
R is expressed in the myocardium mediating a negative chronotropic and
1
inotropic effect and it acts as a presynaptic autoreceptor in both the brain and the periphery.
Accordingly, selective M R antagonism in the CNS, resulting in enhanced cholinergic
2
transmission, was suggested as an approach to increase cholinergic function in Alzheimer’s
2
ꢀ5
disease. In general, MRs represent important drug targets, however, there is still a need for
subtype selective pharmaceuticals acting at MRs, because the development of selective
ligands has been challenging due the high conservation of the acetylcholine (orthosteric)
binding site. As MRs exhibit several distinct allosteric binding sites, which are less conserved
6, 7
than the orthosteric site
these 7ꢀTM receptors emerged as a prototypic receptor class to
8, 9
study allosterism at GPCRs.
Numerous allosteric MR modulators were reported (for
10
11
12, 13
instance compounds 7, 8 (W84) and 9 (LY2119620),
Figure 1A), but allosteric ligands
14
with high affinity are lacking. Radiolabeled derivatives of 8 and 9 were shown to be
13, 15
valuable molecular tools to investigate the allosteric interaction at the M R.
The linkage
2
16
of an orthosteric MR ligand to an allosteric modulator, called the dualsteric (or bitopic )
ligand approach, was suggested as a promising strategy to develop subtype selective MR
8
, 14, 17ꢀ21
ligands.
The rational design of dualsteric MR ligands is supposed to benefit from the
7, 22ꢀ24
recently reported crystal structures of the M , M , M , and M receptor.
Linking of nonꢀ
1
2
3
4
selective MR ligands with allosteric modulators was reported to result in dualsteric M R
2
ligands with peculiar pharmacological profiles with respect to subtype binding, the nature of
25ꢀ29
allosteric cooperativity and functional selectivity (for instance 12ꢀ16, Figure 1B).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. (A) Structures of reported M
2
R antagonists (1, 2, 10), agonists (3ꢀ6, 11) and
allosteric modulators (7ꢀ9). Compounds 10 and 11 were also reported to bind to the allosteric
30, 31
vestibule of the M
2
R.
2
(B) Examples of rationally designed dualsteric M R ligands
obtained by connecting orthosterically (red) with allosterically binding ligands (blue) through
27, 29
a linker.
2
Recently, a series of dibenzodiazepinoneꢀtype M R subtypeꢀpreferring antagonists, derived
from 17 (Figure 2A), including two homodimeric compounds (20 and 21, Figure 2A), were
32
reported. Compounds 20 and 21 showed the highest M
2
R affinity (for K
i
values see Table 1)
3
and their retarding effect on the dissociation of [ H]NMS from the M
2
R was more pronounced
compared with monomeric derivatives (18, 19, Figure 2A) indicating an involvement of
32
2
allosteric binding sites in the interaction of these compounds with the M R. These findings
33
2
are supported by previous reports on the M R binding profile of 10 (AFꢀDX 384) (Figure
30
1
A), which is a close structural analog to 17 (Figure 2A) and was suggested to interact with
the orthosteric as well as with the allosteric binding site. Notably, the use of the tritiated form
30
of 10 proved to be a valuable approach to investigate the binding mode.
The present study aims at the elucidation of the binding mode of complex ligands such as the
dibenzodiazepinone derivatives 20 and 21 at the M R. For this purpose we designed
2
ACS Paragon Plus Environment

Page 5 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
congeners of 20 and 21, which can be conveniently prepared as tritiumꢀlabeled ligands, by
introducing a branched ring structure in the center of the molecule intended to bear the
radiolabel (Figure 2B). For comparison, the tritiumꢀlabeled form of 19, representing the
monomeric counterpart of the dimeric ligand 20 (Figure 2A), was synthesized and also
studied in saturation binding assays including experiments in the presence of allosteric
modulators (Schildꢀlike analysis). Moreover, kinetic investigations, equilibrium competition
binding studies and molecular dynamics simulations were performed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2 (A) Structures of recently reported DIBAꢀderived MR antagonists, including the
homodimeric compounds 20 and 21. (B) Schematic presentation of radiolabeled homodimeric
DIBA derivatives, used as tools to investigate the binding mode at the M R.
2
Results and Discussion
Chemistry. The preparation of the homodimeric dibenzodiazepinone derivative 33,
containing an isophtalic acid moiety in the center, is outlined in Scheme 1. The crucial
building block 29, namely NꢀBoc protected aminomethylated isophtalic acid, was synthesized
34,
from 22 via azide 27, following reported procedures with minor modifications (Scheme 1).
3
5
32
Amidation of 29 with amine 30, using HBTU/HOBt as coupling reagent, and subsequent
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
removal of the Boc group gave amine 31, which represents a precursor for the preparation of
differently functionalized dimeric MR ligands (e.g. radiolabeled and fluorescence labeled
compounds). Propionylation of 31, using succinimidyl propionate (32) afforded 33, which
represents the ‘cold’ form of a tritiumꢀlabeled dimeric MR ligand.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
OH
Cl
COOH
COOCH3
COOH
a
b
c
d
e
HOOC
COOH
H3COOC
COOCH3
H3COOC
COOCH3
H3COOC
COOCH3
H3COOC
COOCH3
2
2
23
24
25
26
Boc
NH
N3
NH2
O
f
× 1 HCl
g
h
NH
O
X
× 3 TFA
N
H3COOC
COOCH3
HOOC
COOH
NH2
N
HOOC
COOH
N
H
2
7
28
29
N
30
X =
O
N
O
O
X
NH
X
NH
i
O
×
5 TFA
N
N
HN
O
O
N
N
HN
4 TFA
3
2
O
N O
×
O
NH
N
N
O
O
NH
N
N
NH2
NH
HN
X
HN
X
3
1
33
O
O
O
Scheme 1. Synthesis of the homodimeric MR ligand 33. Reagent and conditions: (a) H
2
4
SO ,
.
MeOH, reflux, 24 h, 99%; (b) NaOHaq 1 M, MeOH, rt, 18 h, 75%; (c) (CH
4 h, 87%; (d) SOCl , reflux, 90 min, 98%; (e) NaN , acetone, reflux, 16 h, 96%; (f) (1)
triphenylphosphine, H O/THF 5:1 (v/v), rt, 10 h, (2) NaOHaq 1 M, rt, 8 h, 71%; (g) diꢀtertꢀ
butyldicarbonate, triethylamine, H O/dioxane 1:1 (v/v), rt, 12 h, 61%; (h) (1) HOBt, TBTU,
3 2 3
) S BH , THF, rt,
2
2
3
2
2
DIPEA, DMF, 60°C, 3 h (2) triethylamine/CH
DIPEA, DMF, rt, 1 h, 90%.
2 2 2
Cl /H O 10:10:1 (v/v/v), rt, 3 h, 32%; (i)
The key intermediates for the synthesis of the homodimeric MR ligand 47, containing a basic
homopiperazine moiety in the center of the molecule, were bromide 38 and homopiperazine
derivative 45c (Scheme 2). Compound 38 was obtained by reduction of the carboxylated
piperazine derivative 34 to the corresponding alcohol 35, followed by Nꢀalkylation of 35 with
chloride 36 yielding alcohol 37. The latter was converted to bromide 38 using
perbromomethane and triphenylphosphine (Scheme 2). The synthesis of homopiperazine 45c
started with a nitroꢀMannich reaction of dibenzylated ethylenediamine (39) and nitroethanol
(40) yielding derivative 41, which was converted into homopiperazine 42 by a retroꢀHenry
reaction. Reduction of the nitro group in 42 using RaneyꢀNickel and hydrogen led to primary
ACS Paragon Plus Environment

Page 7 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
amine 43, which was Bocꢀprotected to obtain intermediate 44 (Scheme 2). Several attempts to
debenzylate 44 by hydrogenolysis using 10% Pd/C and methanol or ethanol as solvent, failed
due to solvent oxidation to formaldehyde and acetaldehyde, respectively, and formation of the
cyclic aminal 45a and the Nꢀalkylated derivative 45b (identified by NMR spectroscopy and
mass spectrometry, data not shown). The use of 2,2,2ꢀtrifluoroethanol as solvent, preventing
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
36
the Pdꢀcatalyzed oxidation of the alcoholic solvent to the corresponding aldehyde, enabled a
successful debenzylation of 44, affording homopiperazine 45c without byꢀproducts. The
alkylation of 45c using bromide 38, followed by Bocꢀdeprotection, led to the aminoꢀ
functionalized dimeric dibenzodiazepinone derivative 46, which was propionylated to give
compound 47, representing the ‘cold’ form of a tritiumꢀlabeled dimeric ligand as well.
Imidazolyl propionic acid derivative 50, representing the terminal moiety of the side chain in
32
1
9 (cf. Figure 2), was prepared from urocanic acid derivative 48 (Scheme 2), which was
converted to 49 by hydrogenolytic reduction of the ‘acrolein’ double bond under acidic
conditions resulting in a simultaneous cleavage of the Boc group (Scheme 2). Propanoylation
of 49 using 32 yielded compound 50. The homodimeric dibenzodiazepinone derivative 52
was prepared by alkylation of piperazine with chloride 36 (Scheme 2).
The chemical stability of compounds 19, 33 and 47 was investigated in PBS (pH 7.4) at 23 °C
over a period of 48 h. These dibenzodiazepinone derivatives showed excellent stabilities (see
supporting information, Figure S1).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 2. Synthesis of the homodimeric MR ligands 47 and 52 and the imidazolyl propionic
acid derivative 50. Reagent and conditions: (a) LiAlH
CO , acetonitrile, reflux, 5 h, 43%; (c) CBr , triphenylphosphine, CH
9%; (d) paraformaldehyde, EtOH/toluene 1:1 (v/v), reflux, 6 h, 87%; (e) potassium tertꢀ
butanolate, anhydrous THF, rt, 30 min, 95%; (f) H , RaneyꢀNickel, EtOH, rt, 12 h, 44%; (g)
diꢀtertꢀbutyldicarbonate, CH Cl, 0°C/rt, 5 h, 81%; (h) Pd/C (10%), H , 2,2,2ꢀtrifluoroethanol,
atm, rt, 12 h, 85%; (i) (1) K CO , DMF, 120 °C (microwave), 1.5 h (2) CH Cl /TFA/H
4
, THF, 0°C/reflux, 12h, 89%; (b)
K
2
3
4
2
Cl , ꢀ5°C/5°C, 5 h,
2
7
2
3
2
1
2
3
2
2
2
O
10:10:1 (v/v/v), rt, 3 h, 37%; (j) DIPEA, DMF, rt, 1ꢀ1.5 h, 72% (47), 43% (50); (k) Pd/C
(10%), H , MeOH/TFA 1:1 (v/v), 7.9 atm, rt, 12 h, 90%; (l) K CO , acetonitrile, 130°C
2
2
3
(microwave), 30 min, 21%.
Synthesis of radiolabeled dibenzodiazepinone derivatives. The tritiumꢀlabeled
3
3
3
dibenzodiazepinone derivatives [ H]19, [ H]33 and [ H]47 were obtained by treatment of an
excess of the amine precursors 30, 31 and 46, respectively, with commercially available
3
3
succinimidyl [ H]propionate ([ H]32) in the presence of DIPEA (Figure 3A). In order to
3
facilitate the purification of [ H]47, the excess of amine precursor 46 was ‘quenched’ by the
addition of succinimidyl 4ꢀfluorobenzoate (53), resulting in the formation of 54, which could
3
be conveniently separated from [ H]47 (Fig. 3B). Purification by RPꢀHPLC afforded all
3
3
3
radioligands in high radiochemical purity (99%, Figure 3CꢀE). [ H]19, [ H]33 and [ H]47
ACS Paragon Plus Environment

Page 9 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
proved to be stable upon storage in EtOH/H
Information).
O 1:1 (v/v) at −20 °C (Figure S2, Supporting
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
3
3
Figure 3. (A) Synthesis of the MR radioligands [ H]19, [ H]33 and [ H]47 by
3
[ H]propionylation of the amine precursor 30, 31 and 46, respectively, using succinimidyl
[ H]propionate ([ H]32). Reagents and conditions: (a) DIPEA, DMF, rt, 60 min, 39%
[ H]19), 32% ([ H]33); (b) (1) DIPEA, DMF, rt, 45 min, (2) 53, rt, 60 min, 38%. The excess
of 46 was ‘quenched’ by 4ꢀfluorobenzoylation to facilitate the purification of [ H]47 (cf. B).
3
3
3
3
(
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
(B) RPꢀHPLC monitoring of the synthesis of [ H]47 before (black line) and after (blue line)
addition of the active ester 53. (CꢀD) RPꢀHPLC analysis (conditions see experimental section)
of [ H]19 (0.23 µM) (C), [ H]33 (0.20 µM) (D) and [ H]47 (0.20 µM) (E), each spiked with
the ‘cold’ analog (19 and 33: 5 µM; 47: 3 µM) analyzed 2 days after synthesis.
3
3
3
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Equilibrium competition binding studies with [ H]NMS. M
1
ꢀM
5
receptor affinities of the
dimeric dibenzodiazepinone derivatives 33 and 47 were investigated in equilibrium
3
competition binding experiments using [ H]NMS (structure of the ‘cold’ analog see Figure
1A) as orthosterically binding radioligand. The MR binding constants (pK values) are
i
1 5
presented in Table 1 in comparison with previously reported M RꢀM R affinities of 18-21.
The transformation of the structures of 20 and 21 to 33 and 47, respectively, by the
introduction of a branched central linker moiety (Figure 2, Schemes 1 and 2) resulted only in
a marginally decrease in M
subtype selectivity profile comparable to that of the parent compounds 20 and 21 (Table 1).
Steep curve slopes were observed for 33 at the M R (slope = −1.8) and M R (slope = −2.2)
2
R affinity (Figure S3, Supporting Information), and in a MR
1
2
indicating a complex mechanism of binding (e.g. the involvement of more than one binding
3
site). Moreover, equilibrium binding of [ H]NMS at the M
2
R in the presence of homodimeric
R allosteric modulator 9
dibenzodiazepinone derivative 52 (cf. Scheme 2) and the reported M
2
was investigated. Compared to the homodimeric ligands 33 and 47, in which the
3
pharmacophores are separated by complex linker moieties, the decrease in [ H]NMS binding
caused by compound 52 was considerably less pronounced by more than four orders of
magnitude (Supporting Information, Figure S4A). Similarly, the allosteric modulator 9,
23
reported to exhibit a negative cooperativity with NMS, showed a weak inhibitory effect on
3
[
H]NMS equilibrium binding (Supporting Information, Figure S4B).
ACS Paragon Plus Environment

Page 11 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1 5 i
Table 1. M ꢀM receptor affinities (pK values) of the DIBA derivatives 18-21, 33 and 47,
3
obtained from equilibrium competition binding studies with [ H]NMS at live CHOꢀhM R
x
cells (x = 1ꢀ5).
M
1
R
M
2
R
M
3
R
M
4
R
5
M R
compd.
a
a
a
a
a
pK
8.01 ± 1.13 ± 9.17 ± ꢀ1.25 ± 7.11 ± ꢀ0.92 ± 8.49 ± ꢀ0.99 ± 6.36 ± ꢀ0.97 ±
0.08 0.10 0.06 0.08 0.08 0.09 0.11 0.09 0.09 0.14
.07 ± ꢀ1.11 ± 9.12 ± ꢀ1.19 ± 7.22 ± ꢀ1.09 ± 8.63 ± ꢀ0.86 ± 6.75 ± ꢀ0.68 ±
.06 0.11 0.05 0.06 0.08 0.08 0.02 0.03 0.08 0.11
8.91 ± ꢀ1.56 ± 9.71 ± ꢀ1.27 ± 7.88 ± ꢀ1.30 ± 9.19 ± ꢀ1.17 ± 7.44 ± ꢀ1.46 ±
i
slope
pK
i
slope
pK
i
slope
pK
i
slope
pK
i
slope
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
8
0
19
20
b
b
0.05
0.08
0.05
0.19
0.05
0.01
0.12
0.03
0.03
0.01
0.06
0.09
0.08
0.08
0.11
0.09
0.14
0.17
0.06
0.08
0.10
0.08
0.04
0.18
0.18
0.09
0.13
0.13
9
0
.00 ± ꢀ1.32 ± 9.51 ± ꢀ2.01 ± 8.17 ± ꢀ1.05 ± 8.97 ± ꢀ1.32 ± 7.64 ± ꢀ1.09 ±
21
b
b
.04
0.09
0.07
0.14
0.11
0.06
8.82 ± ꢀ1.81 ± 9.39 ± ꢀ2.17 ± 7.81 ± ꢀ1.19 ± 9.33 ± ꢀ1.17 ± 7.64 ± ꢀ0.95 ±
33
b
b
0.10
0.16
0.19
8
0
.15 ± ꢀ1.29 ± 9.15 ± ꢀ1.40 ± 6.70 ± ꢀ1.15 ± 8.44 ± ꢀ0.87 ± 7.11 ± ꢀ1.06 ±
47
b
.06
0.12
0.06
Presented are mean values ± SEM from 3ꢀ9 independent experiments (each performed in triplicate). K
d
3
2
3
values / applied concentrations of [ H]NMS: M
5
nM; M : 0.040 / 0.1 nM; M : 0.24 / 0.3 nM. Data of 18-21 were previously reported as pIC50 values by
a
1
: 0.12 / 0.2 nM; M
2
: 0.090 / 0.2 nM; M
3
: 0.089 / 0.2
4
32
Keller et al. and were reꢀanalyzed to obtain pK
fit. Slope different from unity (P < 0.05).
i
values. Curve slope of the fourꢀparameter logistic
b
Functional studies. The monomeric dibenzodiazepinone derivative 19 and the homodimeric
congener 33 were investigated in an IP1 accumulation assay using HEKꢀ293 cells transiently
transfected with the human M R and the hybrid Gꢀprotein Gαqi5ꢀHA. 19 and 33 did not elicit
2
IP1 accumulation when studied in agonist mode (Figure 4A), that is, they were incapable of
2
stabilizing a Gꢀprotein activating conformation of the M R. In antagonist mode, 19 and 33
completely inhibited IP1 accumulation elicited by 3 (0.3 µM, ≈ EC80), proving these
3
0
compounds to be M
2
R antagonists as previously reported for 10 (Figure 4B). It is a matter of
speculation why the pK values of 19 and 33 (Figure 4B) were lower than the pK values
b
i
presented in Table 1. Presumably, this can be attributed to the different experimental
conditions (different temperature (37 vs. 23 °C), incubation time (1 h vs. 3 h) and readout
(radioligand binding vs. functional response)). In addition, the inhibiting effects of 19 and 33
on the IP1 accumulation elicited by 3, used at various fixed concentrations (0.1, 1 and 10
µM), were determined (Figure 4C and 4D). The inhibition curves reached throughout 0% IP1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
accumulation, indicating a competitive mechanism between the orthosteric agonist 3 and the
dibenzodiazepinone derivatives 19 and 33.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Investigation of M
accumulation assay using HEKꢀhM
9 and 33 on the accumulation of IP1. 19 and 33 elicited no response. pEC50 of 3: 6.93 ± 0.09
mean ± SEM from 8 independent experiments performed in duplicate). (B) Concentrationꢀ
2
R agonism and antagonism of compounds 19 and 33 in an IP1
2
ꢀGαqi5ꢀHA cells. (A) Concentrationꢀdependent effect of 3,
1
(
dependent inhibition of the IP1 accumulation (induced by 3, 0.3 µM) by 2, 19 and 33.
Corresponding pK values: 2: 8.63, 19: 7.53, 33: 7.36. (C) Concentrationꢀdependent inhibition
b
of the IP1 accumulation (induced by 3, 0.1, 1 and 10 µM) by 19. Corresponding pIC values:
5
0
7.47, 7.07 and 6.45, respectively. Concentrationꢀdependent inhibition of the IP1 accumulation
(induced by 3, 0.1, 1 and 10 µM) by 33. Corresponding pIC50 values: 7.78, 6.76 and 6.58,
respectively. Data in BꢀC represent the means ± SEM from at least four independent
experiments (each performed in duplicate).
3
3
3
Characterization of [ H]19, [ H]33 and [ H]47. Saturation binding experiments with the
3
3
3
radiolabeled dibenzodiazepinone derivatives [ H]19, [ H]33 and [ H]47 were performed at
intact adherent CHOꢀhM cells in whiteꢀtransparent 96ꢀwells plates revealing apparent K
2
d
values of 0.87, 0.91 and 8.4 nM, respectively (mean values from at least three independent
ACS Paragon Plus Environment

Page 13 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
experiments performed in triplicate) (Figure 5A, 5B and 5E). Whereas unspecific binding was
3
very low in case of the monomeric ligand [ H]19 (<5% at the K ), drastically increased
d
3
3
unspecific binding (>60%) was a characteristic of the dimeric ligands [ H]33 and [ H]47
3
3
(
Figure 5B and 5E). Saturation binding studies with [ H]33 and [ H]47 performed with intact
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
CHOꢀhM
2
cells in suspension ([ H]33, Figure 5C) or with CHOꢀhM
2
cell homogenates
3
3
([ H]33 and [ H]47, Figure 5D and 5F, respectively), precluding unspecific binding of the
radioligand to the microplate, resulted in considerably lower unspecific binding (<15% at Kd),
indicating that the dimeric ligands 33 and 47 strongly adsorb to polystyrene tissueꢀculture
treated microplates (for experimental protocols see experimental section). The apparent K
d
3
3
values amounted to 0.31 nM ([ H]33, suspended cells), 0.24 nM ([ H]33, homogenates) and
3
4
.6 nM ([ H]47, homogenates) (mean values from at least three independent experiments
performed in triplicate). Worth mentioning, saturation binding studies performed with
3
3
3
[
H]NMS, [ H]19 and [ H]33 on the same day, yielded comparable maximum numbers of
R binding sites per well (Bmax values) independent of using cell homogenates or intact cells
(Supporting Information, Figure S5). These results suggested a ligandꢀreceptor stoichiometry
M
2
3
3
3
of 1:1 for [ H]NMS, the monomeric ligand [ H]19 and the dimeric ligand [ H]33. In addition,
3
saturation binding experiments were performed with [ H]33 at live CHOꢀhM cells resulting
4
in a K
d
value of 0.67 nM (mean value from four independent experiments performed in
triplicate) (cf. Supporting Information, Figure S6).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
2
Figure 5. Representative saturation isotherms (in red) of specific M R binding of [ H]19 (A),
[
3
3
H]33 (BꢀD) and [ H]47 (E, F) obtained from experiments either performed with live
adherent CHOꢀhM cells (A, B, E), live CHOꢀhM cells in suspension (C) or CHOꢀhM cell
2
2
2
homogenates (D, F). Unspecific binding was determined in the presence of the orthosterically
binding MR antagonist atropine (500ꢀfold excess). Scatchard transformations are depicted for
the optimized binding assay conditions (A, D). Experiments were performed in triplicate.
Specific binding data were analyzed by an equation describing a oneꢀsite binding. Error bars
of specific binding and error bars in the Scatchard plots represent propagated errors calculated
according to the Gaussian law of errors. Error bars of total and nonspecific binding represent
the SEM.
ACS Paragon Plus Environment

Page 15 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
As the orthosteric antagonist 2, used to determine unspecific binding, was able to completely
3
3
3
prevent oneꢀsite (monophasic) specific binding of [ H]19, [ H]33 and [ H]47 to the M R,
2
these data strongly suggest that the dibenzodiazepinone derivatives 19, 33 and 47 address the
orthosteric binding site of the M
2
2
R. A putative involvement of a second binding site in M R
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
binding of the dimeric radioligand [ H]33, indicated by the steep slope factor presented in
3
Table 1, was not obvious from saturation binding experiments with [ H]33, which yielded
monophasic saturation isotherms and a linear Scatchard plot (Figure 5C and 5D).
Due to its inappropriate physicochemical properties (high unspecific binding) and the
3
3
discrepancy between the K
d i
([ H]47) and the K (47) values (cf. Table 3), [ H]47 was not
considered with respect to further characterization, contrary to the monomeric radioligand
3
3
3
3
[
H]19 and its dimerized analog [ H]33. The association of both, [ H]19 and [ H]33, to the
M R (live cells) could be described by a monophasic fit resulting in comparable k rates
2
on
(
Figure 6A, 6C, Table 2). By contrast, the dissociation from the M
2
R (live cells) was
3
different: whereas the monomeric ligand [ H]19 completely dissociated with a halfꢀlife of 71
min (monophasic decline), the dissociation of the dimeric dibenzodiazepinone derivative
3
[
H]33 was incomplete, reaching a plateau at 67% of initially bound radioligand. Dissociation
3
experiments with [ H]33 at CHOꢀhM cell homogenates gave comparable results (plateau at
2
6
3%; cf. Supporting Information, Figure S7). These data suggested in part a
3
(pseudo)irreversible (long lasting) binding of [ H]33, which might be attributed to
37
conformational adjustments of the receptor upon ligand binding, or to an enhanced
rebinding capability of the dimeric ligand by a simultaneous interaction with two or more
38
3
binding sites. However, the equilibrium dissociation constant of [ H]33, calculated from k
off
and kon (K
d
d
(kin) = koff/kon = 0.20 nM, Table 2) was in good agreement with the K value
3
derived from saturation binding experiments (K
d
= 0.31 nM) indicating that [ H]33 follows in
3
3
2
part the law of mass action. Notably, the dissociations of [ H]19 and [ H]33 from the M R
were not or only marginally effected by the allosteric modulator 8 applied at a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
‘supramaximal’ concentration of 100 µM (Figure 6B and 6D), which is consistent with a
3
3
competitive mechanism between [ H]19 and 8 as well as [ H]33 and 8; in other words, these
39
results preclude a high negative cooperativity between the studied ligands. An overview of
3
3
the M R binding characteristics of [ H]19 and [ H]33 is provided in Table 2.
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
Figure 6. Association and dissociation kinetics of [ H]19 (A, B) and [ H]33 (C, D)
determined at intact CHOꢀhM cells at 23 °C. (A) Radioligand (c = 2 nM) association to the
2
−1
M R as a function of time (nonꢀlinear regression: k
= 0.042 min ); inset:
ln[B(eq)/(B(eq)−B(t))] versus time, kobs = slope = 0.035 min . (B) Radioligand (preincubation: 2
nM, 105 min) dissociation from the M R as a function of time, showing complete monophasic
2
obs
−
1
2
−1
exponential decline (nonꢀlinear regression: koff = 0.0091 min , t1/2 = 71 min); inset:
ln[B(t)/B(0)] versus time, slope·(−1) = koff = 0.0077 min , t1/2 = 90 min. (C) Radioligand (c = 1
−1
−1
nM) association to the M
2
R as a function of time (nonꢀlinear regression: kobs = 0.067 min );
−1
inset: ln[B(eq)/(B(eq)ꢀB(t))] versus time, kobs = slope = 0.066 min . (D) Radioligand
preincubation: 1 nM, 90 min) dissociation from the M R as a function of time, showing
incomplete monophasic exponential decline (nonꢀlinear regression: koff = 0.011 min , t1/2
66 min, plateau = 67%); inset: ln[(B(t)ꢀBplateau)/B(0)] versus time, slope·(−1) = koff = 0.011
(
2
−1
=
−
1
3
3
min , t = 64 min. The dissociation kinetics of [ H]19 and [ H]33 were not or only
1/2
marginally effected by the allosteric modulator 8 at a ‘supramaximal’ concentration of 100
µM (B, D (open triangles and open diamonds)), indicating a competitive interaction between
the radioligands and 8. Data represent means ± SEM from two or three (B, C, D) or four (A)
independent experiments (each performed in triplicate).
ACS Paragon Plus Environment

Page 17 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
3
3
2
Table 2. M R binding data of [ H]19, [ H]33 and [ H]47 (at 23 ± 1 °C).
K
i
(‘cold’
−1 .
−1 e
b
c
k
on [min
nM ]
k
off [min ]
radioligand
analog)
K
d
(sat) [nM]
K
d
(kin) [nM]
−1 d
e
a
t1/2 [min]
[
nM]
3
0.0091 ± 0.0016
71 ± 7
[
H]19
0.76 ± 0.05
0.87 ± 0.01
0.70 ± 0.10
0.013 ± 0.003
0.056 ± 0.010
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
0.011 ± 0.001
[ H]33
0.41 ± 0.05
0.71 ± 0.01
0.31 ± 0.08
4.6 ± 1.3
0.20 ± 0.04
n.d
6
6 ± 9
3
[
H]47
n.d
n.d
a
3
Dissociation constant determined by equilibrium competition binding with [ H]NMS at intact CHOꢀhM
mean ± SEM from at least three independent experiments (performed in triplicate). Equilibrium dissociation
constant determined by saturation binding at intact CHOꢀhM cells ([ H]19 and [ H]33) and at CHOꢀhM cell
homogenates ([ H]47); mean ± SEM from at least three independent experiments (performed in triplicate).
Kinetically derived dissociation constant ± propagated error (K (kin) = koff/kon). Association rate constant ±
d
propagated error, calculated from kobs, koff and the applied radioligand concentration (cf. Figure 6 and
experimental section). Dissociation rate constant and derived halfꢀlife (nonꢀlinear fit); mean ± SEM from three
independent experiments (performed in triplicate).
2
cells;
b
3
3
2
2
3
c
d
e
Binding of the monomeric dibenzodiazepinone derivative 19 to the orthosteric site was further
3
supported by investigating the saturation binding of the orthosteric radioligand [ H]NMS in
the presence of 19 (Schildꢀlike analysis). This experiment revealed a parallel rightward shift
3
of the saturation isotherms of [ H]NMS and a slope not different from unity of the
corresponding ‘Schild’ regression, being indicative of a competitive interaction between 19
3
and the orthosteric radioligand [ H]NMS (Supporting Information, Figure S8).
3
3
M R equilibrium competition binding with [ H]19 and [ H]33. Selected standard MR
2
agonists (4, 5) antagonists (1, 2, 10) and allosteric modulators (7, 8) as well as the ‘cold’
forms of the radioligands (19, 33) were investigated by equilibrium competition binding with
3
3
[
H]19 or [ H]33. The concentrationꢀdependent effects of the investigated compounds on
3
3
equilibrium binding of [ H]19 and [ H]33, analyzed by fourꢀparameter logistic fits, are
depicted in Figure 7. The lower curve plateaus were throughout not different from 0 (P > 0.2),
which either suggests a competitive mechanism or a strongly negative cooperativity between
4
0
the studied compounds and the radiolabeled dibenzodiazepinone derivatives.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 7. Concentrationꢀdependent effects of various reported orthosteric (1-5), allosteric (7ꢀ
3
9
K
) and dualsteric (10, 19, 33) MR ligands on M
2
R equilibrium binding of [ H]19 (c = 2 nM,
3
d
d 2
= 0.87 nM) (A) and [ H]33 (c = 0.5 nM, K = 0.24 nM) (B) determined at live CHOꢀhM
cells and at CHOꢀhM cell homogenates, respectively. Data were analyzed by fourꢀparameter
2
logistic fits. Mean values ± SEM from at least three independent experiments (performed in
triplicate).
3
Generally, pK
i
values derived from equilibrium competition binding with [ H]19 were in good
agreement with reported data (Table 3). However, pK values determined with the dimeric
i
3
radioligand [ H]33 were consistently lower compared to M
2
R affinities obtained by the use of
3
3
[
H]19. This can be explained by the (pseudo)irreversible binding of [ H]33 at the M
2
R (cf.
Figure 6D): in putative equilibrium competition binding experiments a dynamic equilibrium
between the receptor and the competing ligands cannot be reached. The fraction of receptor,
3
irreversibly bound to [ H]33, is inaccessible for the studied nonꢀlabeled ligand. As a
consequence, “displacement” of the irreversibly binding radioligand by the nonꢀlabeled
reversibly binding ligand requires a higher receptor occupancy by the latter (relative to the
radioligand) compared to the situation of two reversibly binding competing ligands.
3
3
Moreover, due to the (pseudo)irreversible binding of [ H]33, the apparent K
d
value of [ H]33,
ACS Paragon Plus Environment

Page 19 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
obtained from saturation binding studies (cf. Figure 5D), is presumably lower than the real K
d
value, resulting in an underestimation of the affinity of the studied ligand (calculated via
ChengꢀPrusoff). Additional equilibrium competition binding experiments performed with 4
2
and 10, applying a preincubation period of 90 min with the M R prior to the addition of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
[
i
H]33, did not result in an increase in the apparent pK of 4 and 10 (data not shown), which is
anticipated according to the aforementioned hypothesis.
Table 3. M
2 i
R binding data (pK or pIC50) of various orthosteric (1-5), allosteric (7-9) and
3
3
3
dualsteric (10, 19, 33) MR ligands determined with [ H]19, [ H]33 or [ H]NMS.
3
3
3
[ H]19
[ H]33
[ H]NMS
Ligand
a
b
*
i
**
pK
i
pK
i
pK
or pIC50 (± SEM)
*
c
1
2
4
5
7
8
9
10.2 ± 0.10
9.09 ± 0.04
6.83 ± 0.05
6.57 ± 0.05
6.33 ± 0.04
7.40 ± 0.03
5.73 ± 0.03
8.76 ± 0.07
8.96 ± 0.06
9.97 ± 0.20
8.92 ± 0.03
7.77 ± 0.04
5.00 ± 0.12
5.35 ± 0.03
5.43 ± 0.01
5.97 ± 0.05
5.26 ± 0.06
7.63 ± 0.13
7.99 ± 0.10
9.7
*
d
7.8ꢀ9.2
5.0ꢀ6.6
6.5ꢀ7.4
*d
*
e
*
*f
6.11 ± 0.09
**f
6.32 ± 0.18
*
*f
<4.5
*d
1
1
0
9
8.2
*
*
g
g
9.12 ± 0.05
9.39 ± 0.05
33
8.98 ± 0.05
3
a
Determined by equilibrium binding with [ H]19 (c = 2 nM) in the presence of increasing
concentrations of the respective MR ligand at intact CHOꢀhM
3
with [ H]33 (c = 0.5 nM) in the presence of increasing concentrations of the respective MR ligand at
CHOꢀhM cell homogenates; mean values ± SEM from at least three independent experiments
performed in triplicate). Dei et al. pK
reported in the literature (data taken from the IUPHAR/BPS database (guidetopharmacology.org,
Nov. 2016)). Jakubik et al. pIC50 values obtained from nonlinear fourꢀparameter logistic curve
analyses of data characterizing the inhibition of [ H]NMS (c = 0.2 nM) equilibrium binding at live
2
cells; mean values ± SEM from at least
independent experiments (performed in triplicate). Determined by (putative) equilibrium binding
b
3
2
c
41 d
(
i
values from equilibrium (competition) binding experiments
e
42 f
(
3
CHOꢀhM cells (for compound 9 cf. Supporting Information, Figure S4B); mean ± SEM from at least
3
2
g
independent experiments (performed in triplicate). pK
i
values taken from Table 1.
In order to verify a putative competitive mechanism of the allosteric M
2
R modulator 8 with
3
3
[
H]19 and [ H]33 at the M R, equilibrium competition binding studies were performed
2
applying increasing fixed concentrations/receptor occupancies of the radioligands (Figure 8A,
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
8B). In the same manner, a reference experiment was performed with 8 and the orthosteric
3
3
radioligand [ H]NMS (Figure 8C). The control experiment with [ H]NMS revealed an
elevation of the lower plateau of the curves and the pIC of 8 was unaffected, indicating a
50
30
nonꢀcompetitive mechanism as reported previously. By contrast, the lower plateau of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
3
curves did not increase in case of [ H]19 and [ H]33, and a rightward shift of the pIC50 of 8
40
was observed (Figure 8A and 8B). These data support a competitive mechanism between the
allosteric M R ligand 8 and the dibenzodiazepinoneꢀtype ligands 19 and 33 and,
consequently, a dualsteric binding mode of 19 and 33 at the M R.
2
2
3
Figure 8. Effect of the allosteric MR modulator 8 on M
2
R equilibrium binding of [ H]19 (K
= 0.24 nM) (B) and [ H]NMS (K = 0.09 nM) (C) using various
radioligand concentrations. Experiments were performed at intact CHOꢀhM cells (A, C) or at
CHOꢀhM cell homogenates (B). Unspecific binding was determined in the presence of
d
3
3
=
0.87 nM) (A), [ H]33 (K
d
d
2
2
3
3
3
atropine (500ꢀfold excess to [ H]19, [ H]33 or [ H]NMS). Data were analyzed by fourꢀ
ACS Paragon Plus Environment

Page 21 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
3
parameter logistic fits. In case of [ H]19 and [ H]33 increasing radioligand concentrations
resulted in a parallel rightward shift of the curves, which reached 0% specific radioligand
binding throughout, indicating a competitive mechanism (A, B). In contrast, curves of
H]NMS equilibrium binding obtained in the presence of increasing concentrations of 8 were
not rightwardꢀshifted, instead, the lower plateau of the curve appeared to be elevated at
increased radioligand concentrations, indicating negative allosteric cooperativity between 8
and [ H]NMS (C). Data represent mean values ± SEM from at least three independent
3
[
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
experiments (performed in triplicate). ρ
A
= fractional receptor occupancy
3
3
Schild-like analysis with 8 and 9 at the M
2
R using [ H]19 and [ H]33. In order to
3
3
2
substantiate the studies on the binding mode of [ H]19 and [ H]33 at the M R, saturation
3
3
binding of [ H]19 and [ H]33 in the presence of the allosteric modulators 8 and 9 were
performed (Figure 9). This kind of experiment is equivalent to the Schild analysis used to
investigate the inhibiting effect of a receptor antagonist on the response elicited by an agonist
3
9, 40, 43
and is considered the experiment of choice to unveil nonꢀcompetitive mechanisms.
For
3
instance, it was used to prove allosteric binding of 7 at the M
2
R (radioligand: [ H]NMS) and
3
44, 45
oleamide at the 5ꢀHT
7
receptor (radioligand: [ H]5ꢀHT).
Increasing concentrations of 8
led to a parallel rightward shift of the saturation isotherms of both, the monomeric ligand
3
3
[
H]19 and the homodimeric ligand [ H]33, resulting in linear ‘Schild’ regressions with a
slope not different from unity (Figure 9A, 9C, Table 4), indicating a competitive mechanism
between the allosteric M
2
R ligand 8 and the dibenzodiazepinoneꢀtype ligands 19 and 33.
23
2
Likewise, the allosteric modulator 9, which was coꢀcrystalized with the M R bound to 6,
3
elicited a parallel rightward shift of [ H]19 saturation isotherms, again yielding a linear
‘Schild’ regression with a slope not different from unity (Figure 9B, Table 4). The ‘pA
values derived from these ‘Schild’ regressions were in a good accordance with the pK values
from equilibrium competition binding studies with [ H]19 and [ H]33 as well as with reported
R affinities of 8 and 9 (Table 4). With regard to the fact that the dibenzodiazepinone
derivatives 19 and 33 address the M R orthosteric binding site (see above) the results of the
2
’
i
3
3
M
2
2
Schildꢀlike analyses strongly suggested a dualsteric binding mode of 19 and 33 at the M
that is, a simultaneous binding to the orthosteric site and the allosteric vestibule.
ACS Paragon Plus Environment
2
R,

Journal of Medicinal Chemistry
Page 22 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
Figure 9. Saturation binding of [ H]19 in the presence of increasing concentrations of 8 (A)
3
or 9 (B), and of [ H]33 in the presence of increasing concentrations of 8 (C). Presented are
saturation isotherms of specific radioligand binding to the M R in linear scale (left) and semiꢀ
logarithmic scale (middle), and ‘Schild’ regressions (right) resulting from the rightward shifts
d
) of the saturation isotherms (log(r−1) plotted vs. log(concentration allosteric
). The presence of the allosteric modulator (8, 9) led to a parallel
rightward shift of the saturation isotherms of both, the monomeric ([ H]19) and the
homodimeric ([ H]33) radiolabeled dibenzodiazepinone derivative. In all cases the slope of
the linear ‘Schild’ regression was nearly equal to unity indicating a competitive interaction
between the radioligands and the investigated allosteric ligands. Experiments were performed
2 2
at intact CHOꢀhM cells (A, B) or at CHOꢀhM cell homogenates (C). Data represent mean
values ± SEM from at least three independent experiments (each performed in triplicate).
2
(ꢁpK
ꢁpKd
modulator), where r = 10
3
3
ACS Paragon Plus Environment

Page 23 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Summary of M
by the use of the radiolabeled dibenzodiazepinone derivatives [ H]19 and [ H]33.
2 2
R binding data of the allosteric M R modulators 8 and 9 determined
3
3
allosteric
ligand
used
radioligand
a
b
c
i
*
**
slope
‘pA
2
’
pK
pK
X
or pK
B
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
*d
8
9
8
[ H]19
0.94 ± 0.04
1.1 ± 0.1
6.81 ± 0.02
5.45 ± 0.09
6.58 ± 0.01
7.39 ± 0.03
5.73 ± 0.03
5.91 ± 0.06
7.50
3
**e
[ H]19
5.77
3
*d
[ H]33
0.98 ± 0.06
7.50
a
Slope of the ‘Schild’ regression constructed based on radioligand binding data from saturation
binding experiments in the presence of increasing fixed concentrations of the allosteric ligands 8 or 9;
mean values ± SEM from at least three sets of independent saturation binding experiments (performed
in triplicate). The slope of the linear ‘Schild’ regression was not significantly different from unity (P >
’ value corresponds to the Xꢀaxis intercept of the linear ‘Schild’ regression (cf. Fig. 9)
i
and reflects the affinity of the investigated allosteric ligands. pK values taken from Table 3. Tränkle
b
0
.2). The ‘pA
2
c
d
1
5 e
23
et al. Kruse et al.
Molecular dynamics simulations. Aiming at a verification of the conclusions drawn from
the experimental data by computational methods, MD simulation of the human M R bound to
2
the dibenzodiazepinone derivative 19 or 33 (2 and 3 µs, respectively) were performed.
Simulations were prepared by induced fit docking of 19 and 33 to the M R in the inactive
2
state (cf. Supporting Information, Figure S9). The dockingꢀderived orientation of 19 showed
interactions of 19 with the lower part of the allosteric vestibule (Figure S9A), which was,
regardless of minor conformational changes, persistent during the 2ꢀµs simulation (Figure
1
0A, Figure S10 (Supporting Information)). Amino acid residues, reported to be involved in
23
binding of the allosteric M R ligand 9, appeared to be in close proximity to the monomeric
2
ligand 19 (Figure 10A). These data suggested a (partial) overlap of the binding sites of 9 and
1
9, which is consistent with the results of the experiments presented in Figure 9B (Schildꢀlike
ECL2
analysis). Notably, the backbone contact between the propionamide moiety in 19 and I178
suggested that 19 reaches the extracellular surface of the M R.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 10. Molecular dynamics simulation of the human M
bound to the dibenzodiazepinone derivatives 19 and 33 (2 and 3 µs, respectively). (A) Cluster
binding pose of 19 (shown in purple). (B) Cluster 1 binding pose of 33 (shown in green). In
2
R (inactive state, PDB ID 3UON)
1
3
.32
3.28
2.57
7.43
A and B carbon atoms of amino acids constituting the D103 ꢀW99 ꢀS76 ꢀY430
cluster are highlighted in cyan (only assigned in B). Key amino acids suggested to interact
3
.33
ECL2
5.42
6.51
with 19 were Y104 (πꢀπ), I178
(HB), T1902 (HB) and Y403 (HB) (A). Key amino
.65 3.32 3.33
acids suggested to interact with 33 were T84
(HB), D103
(HB), Y104
(πꢀπ),
ECL2
6.51
7.35
7.39
T170
(HB), Y403
(HB), W422
(HB) and Y426
(HB) (B). In A, amino acids
2.61
ECL2
ECL2
reported to interact with the allosteric modulator 9, that is Y80 , E172
N410 , N419 , W422
Time course of the 3ꢀµs MD simulation of the M
, Y177
,
6
.58
7.32
7.35
7.39
23
and Y426 , are surrounded by transparent cyan surface. (C)
R bound to 33 showing superimposed snap
2
shots collected every 100 ns. (D) Superimposition of the cluster 1 binding poses of 19
(purple) and 33 (green). HB = hydrogen bonding.
As also observed for the monomeric compound 19, the orientation of the orthosterically
bound part of the dimeric ligand 33 varied only marginally during the 3ꢀµs simulation, that is,
it remained tightly bound to the orthosteric site. By contrast, the major part of the linker and
the second dibenzodiazepinone pharmacophore in 33, being unbound and located in the
‘extracellular space’ at the beginning of the simulation (cf. Figure S9B (Supporting
ACS Paragon Plus Environment

Page 25 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Information) and frame 1 (orange) in Figure 10C), showed a high flexibility over the first 2 µs
(in part reflected by Figure 10C and Figure S11 (Supporting Information)). Strikingly, in the
late phase of the simulation the flexibility of the allosterically interacting part of 33 was much
lower resulting in a persistence of the predominant binding pose of 33 (represented by cluster
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
, Figure 10D, Table S1 (Supporting Information)) over 1 µs (Figure 10C). A MPG1 movie
3.32
file showing the entire MD simulation of 33 is available. Interestingly, Asp103 , which
typically forms a salt bridge with basic amine functions of MR agonists as a key interaction
46
for receptor activation showed an interaction with the dimeric compound 33, but not with
the monomeric ligand 19 (Supporting Information, Figure S12B). This was also reported for
2
4
1 4
crystal structures of the M and M receptor in complex with an antagonist. In both
3
.32
2.57
3.28
simulations, Asp103 formed an Hꢀbonding network with Ser76 and Trp99 , in case of
7.43
the simulation with 19 additionally with Tyr430
(cf. Supporting Information, Figure
24
S12A), which is characteristic of the inactive receptor conformation. As anticipated, the
homodimeric ligand 33 showed more ‘allosteric contacts’ in the simulation compared to the
monomeric ligand 19. Summarized, the results obtained from the MD simulations were
consistent with the results of the aforementioned experiments, which suggested a dualsteric
binding mode of the investigated tricyclic MR ligands 19 and 33.
2
M R binding studies with 50. Compound 50, representing the allosterically interacting part
of the dualsterically binding MR ligand 19, was investigated with respect to M
2
R binding by
3
3
studying its potential inhibitory effect on M R equilibrium binding of [ H]NMS and [ H]19 at
2
3
3
intact CHOꢀhM cells. 50 could not ‘inhibit’ M R binding of [ H]NMS and [ H]19 at
2
2
concentrations below 100 µM (Supporting Information, Figure S13) suggesting that the
allosteric interactions of 19 are not or only marginally responsible for the high M
the dualsteric ligand 19.
2
R affinity of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Conclusion
The presented study comprises the first report on tritiumꢀlabeled dibenzodiazepinoneꢀtype,
R subtypeꢀpreferring MR antagonists including the comparison of a homodimeric
M
2
3
3
derivative ([ H]33) with its monomeric counterpart ([ H]19). With K
d
values below 1 nM both
R affinities than the structurally related
= 11 nM, M R), that was previously
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
radioligands exhibited considerably higher M
2
3
pyridobenzodiazepinoneꢀtype MR antagonist [ H]10 (K
d
2
30
suggested to bind to the allosteric vestibule of the M R. The synthesis of the radiolabeled
2
3
3
compounds [ H]19 and [ H]33 contributed significantly to the elucidation of the M
2
R binding
mode of the high affinity dibenzodiazepinone derivatives 19 and 33. The results from various
3
3
binding experiments with [ H]19 and [ H]33, in particular, saturation binding studies in the
absence and in the presence of well characterized allosteric M R ligands (compounds 8 and 9)
2
(
Figure 9) as well as dissociation studies at the M R in the presence of 8 (Figure 6B and 6D),
2
strongly indicated that the presented type of M
2
R antagonists (19, 33, and presumably the
reported congeners 20 and 21, too) interact simultaneously with both the orthosteric and the
‘common’ allosteric binding site. Therefore, this study demonstrates that the use of
radiolabeled analogs can significantly contribute to a better understanding of the binding
3
mode of complex receptor ligands. The chemically stable dualsteric radioligands [ H]19 and
3
[
2
H]33, which could be conveniently synthesized and exhibited high M R affinity, represent
useful tools to study their binding mode at MRs. However, whereas the monomeric ligand
3
[
H]19 represents a new valuable molecular tool complementary to known orthosteric MR
3
radioligands, the use of the dimeric radioligand [ H]33 to characterize the binding of nonꢀ
labeled M R ligands is compromised by its irreversible binding at the M R. This work
2
2
2
suggests dibenzodiazepinoneꢀtype MR ligands as privileged structures to develop M R
8, 14, 17ꢀ21
selective antagonists according to the dualsteric ligand approach
by the synthesis of
heterodimeric ligands comprising a linkage of the dibenzodiazepinone scaffold to reported
allosteric M R pharmacophores as well as to newly designed allosterically interacting
2
ACS Paragon Plus Environment

Page 27 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
moieties based on the information of MR crystal structures. With regard to the low selectivity
towards the M R of the investigated type of MR ligands, compounds such as 19 and 33 may
4
4
also represent lead structures for the development of M R selective ligands.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
General experimental conditions. Chemicals and solvents were purchased from commercial
suppliers and were used without further purifications unless otherwise specified. Acetonitrile
for HPLC (gradient grade) was obtained from Merck (Darmstadt, Germany). Nꢀmethyl
scopolamine (1), atropine (2), carbachol (3), gallamine (7), 8, 22, 34, 39 and 40 were
purchased from SigmaꢀAldrich (Deisenhofen, Germany). Oxotremorine sesquifumarate (4)
was from MP Biomedicals (Eschwege, Germany), compound 9 was from Absource
Diagnostic (Munich, Germany) and 10 was purchased from Abcam (Cambridge, UK).
3
[
H]NMS (specific activity = 80 Ci/mmol) was purchased from American Radiolabeled
Chemicals Inc. (St. Louis, MO) via Hartmann Analytics GmbH (Braunschweig, Germany).
3
2
32
The syntheses of dibenzodiazepinone derivative 30, urocanic acid derivative 48 and
47
compound 53 were described previously. Xanomeline (5) and compound 36 were
48, 49
synthesized according to described procedures.
Technical grade solvents (CH Cl , ethyl
2 2
acetate, light petroleum) were distilled before use. For solvent dehydration THF and diethyl
ether were distilled over sodium, and CH Cl was distilled over P after predrying over
CaCl . Millipore water was used throughout for the preparation of buffers and HPLC eluents.
Polypropylene reaction vessels (1.5 or 2 mL) with screw cap (SüdꢀLaborbedarf, Gauting,
2
2
2 5
O
2
3
3
3
Germany) were used for the synthesis of radioligands ([ H]19, [ H]33, [ H]47), for small
scale reactions, for the investigation of chemical stabilities (19, 47) and for the preparation
and storage of stock solutions. The chemical stability of 33 was investigated in a flatꢀbottom
glass tube (40 × 8.2 mm) (Altmann Analytik GmbH, Munich, Germany), which was
siliconized before use using Sigmacote (SigmaꢀAldrich). Thin layer chromatography (TLC)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 75
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
.
was performed on Merck silica gel 60 F254 TLC aluminum plates Visualization was
accomplished by UV irradiation (λ = 254 or 366 nm) and by staining with ninhydrin or
potassium permanganate. Silica Gel 60 (40ꢀ60 µm, Merck) was used for column
chromatography. A Biotage Initiator microwave synthesizer (Biotage, Uppsala, Sweden) was
used for microwave driven reactions. NMR spectra were recorded on a Bruker Avance 300
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
13
1
(7.05 T, H: 300.1 MHz, C: 75.5 MHz), Bruker Avance III HD 400 (9.40 T, H: 400 MHz,
1
3
1
C: 100 MHz) or a Bruker Avance III HD 600 equipped with a cryogenic probe (14.1 T H:
13
6
00.1 MHz, C: 150.9 MHz) (Bruker, Karlsruhe, Germany) with TMS as external standard.
IR spectra were measured with a NICOLET 380 FTꢀIR spectrophotometer (Thermo Electron
Corporation). Lowꢀresolution mass spectrometry (MS) was performed on a Finnigan SSQ
710A instrument (CIꢀMS) (Thermo Finnigan, San Jose, CA). Highꢀresolution mass
spectrometry (HRMS) analysis was performed on an Agilent 6540 UHD AccurateꢀMass Qꢀ
TOF LC/MS system (Agilent Technologies, Santa Clara, CA) using an ESI source.
Preparative HPLC was performed with a system from Knauer (Berlin, Germany) consisting of
two Kꢀ1800 pumps and a Kꢀ2001 detector. A KinetexꢀXB C18 (5 ꢂm, 250 mm × 21 mm;
Phenomenex, Aschaffenburg, Germany) or an Actus Triart C18 (5 ꢂm, 150 mm × 21 mm;
YMC Europe GmbH, Dinslaken, Germany) were used as stationary phases at a flow rate of
1
8 mL/min. Mixtures of acetonitrile and 0.1% aq TFA and mixtures of acetonitrile and 0.1%
aq ammonia, respectively, were used as mobile phase. The detection wavelength was set to
20 nm throughout. The solvent of the collected fractions was removed by lyophilization
2
using an Alpha 2ꢀ4 LD apparatus (Martin Christ, Osterode am Harz, Germany) equipped with
a RZ 6 rotary vane vacuum pump (Vacuubrand, Wertheim, Germany). Analytical HPLC
analysis of compounds 19, 33 and 47 (concentrations between 25 and 50 ꢂM) was performed
with a system from Agilent Technologies composed of a 1290 Infinity binary pump equipped
with a degasser, a 1290 Infinity autosampler, a 1290 Infinity thermostated column
compartment, a 1260 Infinity diode array detector, and a 1260 Infinity fluorescence detector.
ACS Paragon Plus Environment

Page 29 of 75
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A KinetexꢀXB C18 (2.6 ꢂm, 100 × 3 mm; Phenomenex) served as stationary phase at a flow
rate of 0.5 mL/min except for the analysis of 50. Mixtures of 0.04% aq TFA (A) and
acetonitrile (B) were used as mobile phase. The following linear gradient was applied: 0ꢀ20
min: A/B 90:10ꢀ68:32, 20ꢀ22 min: 68:32ꢀ5:95, 22ꢀ28 min: 5:95. Analytical HPLC analysis of
compound 50 was performed using an Actus Triart C18 (3 ꢂm, 150 × 2 mm; YMC Europe) as
stationary, and mixtures of 0.1% aq ammonia (C) and acetonitrile (B) as mobile phase (flow
rate: 0.5 mL/min). The following linear gradient was applied: 0ꢀ10 min: C/B 95:5ꢀ90:10, 10ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
20 min: 90:10ꢀ10:90, 20ꢀ28 min: 10:90. For all analytical HPLC runs the oven temperature
was set to 25 °C, the injection volume was 20 ꢂL and detection was performed at 220 nm.
Melting points were determined with a Büchi 510 apparatus (Büchi, Essen, Germany) and are
uncorrected.
1
13
Annotation concerning the NMR spectra ( H, C) of the dibenzodiazepinone derivatives (31,
33, 37, 38, 46, 47): due to a slow rotation about the exocyclic amide group on the NMR time
1
scale, two isomers (ratios provided in the experimental protocols) were evident in the Hꢀ and
1
3
CꢀNMR spectra.
1
13
Compound characterization. New organic compounds were characterized by Hꢀ and Cꢀ
NMR spectroscopy, HRMS and melting point (if applicable). In addition, the target
1
compounds 31, 33, 46, 47, 50 and 52 were characterized by 2DꢀNMR spectroscopy ( Hꢀ
COSY, HSQC, HMBC) and RPꢀHPLC analysis. Furthermore, IR spectra were acquired for
the homodimeric ligands 33 and 47. The purity of final compounds, determined by RPꢀHPLC
(220 nm), was ≥97% throughout (chromatograms shown in the Supporting Information).
ACS Paragon Plus Environment