Synthesis toward Bivalent Ligands for the Dopamine D2 and Metabotropic Glutamate 5 Receptors

Article abstract of DOI:10.1021/acs.jmedchem.8b00671

In this study, we designed and synthesized heterobivalent ligands targeting heteromers consisting of the metabotropic glutamate 5 receptor (mGluR5) and the dopamine D₂ receptor (D₂R). Bivalent ligand 22a with a linker consisting of 2

Full text of DOI:10.1021/acs.jmedchem.8b00671

Subscriber access provided by UNIV OF DURHAM
Article
Synthesis towards bivalent ligands for the dopamine
D2 and metabotropic glutamate 5 receptors
Mingcheng Qian, Elise Wouters, James Dalton, Martijn D.P. Risseeuw, René A. J. Crans,
Christophe P. Stove, Jesús Giraldo, Kathleen Van Craenenbroeck, and Serge Van Calenbergh
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00671 • Publication Date (Web): 04 Sep 2018
Downloaded from http://pubs.acs.org on September 5, 2018
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 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 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.
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 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis towards bivalent ligands for the dopamine
D2 and metabotropic glutamate 5 receptors
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
Mingcheng Qian,1,2,‡ Elise Wouters,2,‡ James A. R. Dalton,3,4,5 Martijn D. P. Risseeuw,1 René A.
J. Crans,2 Christophe Stove, 2 Jesús Giraldo,3,4,5 Kathleen Van Craenenbroeck,2 and Serge Van
Calenbergh1,*
1. Laboratory for Medicinal Chemistry (FFW), Ghent University, Ottergemsesteenweg 460,
Bꢀ9000 Gent, Belgium
2. Laboratory of Toxicology, Ghent University, Ottergemsesteenweg 460, Bꢀ9000 Ghent,
Belgium
3. Laboratory of Molecular Neuropharmacology and Bioinformatics, Unitat de
Bioestadística, Institut de Neurociències, Universitat Autònoma de Barcelona, 08193
Bellaterra, Spain
4. Instituto de Salud Carlos III, Centro de Investigación Biomédica en Red de Salud Mental,
CIBERSAM, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain
5. Unitat de Neurociència Traslacional, Parc Taulí Hospital Universitari, Institut
d’Investigació i Innovació Parc Taulí (I3PT), Institut de Neurociències, Universitat
Autònoma de Barcelona, 08193 Bellaterra, Spain
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
5
6
KEYWORDS. Dopamine D2 receptor; Metabotropic glutamate 5 receptor; Bivalent ligands;
Radioactive ligand binding; Signal transduction
7
8
9
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
ACS Paragon Plus Environment
2

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
ABSTRACT
5
6
7
8
9
In this study we designed and synthesized heterobivalent ligands targeting heteromers consisting
of the metabotropic glutamate 5 receptor (mGluR5) and the dopamine D2 receptor (D2R).
Bivalent ligand 22a with a linker consisting of 20 atoms showed 4ꢀfold increase in affinity for
cells coexpressing D2R and mGluR5 compared to cells solely expressing D2R. Likewise, the
affinity of 22a for mGluR5 increased 2ꢀfold in the coexpressing cells. Additionally, 22a
exhibited a 5ꢀfold higher mGluR5 affinity than its monovalent precursor 21a in cells
coexpressing D2R and mGluR5. These results indicate that 22a is able to bridge binding sites on
both receptors constituting the heterodimer. Likewise, cAMP assays revealed that 22a had a 4ꢀ
fold higher potency in stable D2R and mGluR5 coexpressing cell lines than 1. Furthermore,
molecular modeling reveals that 22a is able to simultaneously bind both receptors by passing
between the TM5ꢀTM6 interface and establishing six proteinꢀligand Hꢀbonds.
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
Introduction
The metabotropic glutamate receptors (mGluRs) constitute a family of G proteinꢀcoupled
receptors (GPCRs) characterized by the presence of large, ectocytic, Nꢀterminal orthosteric
binding domains, which are activated by Lꢀglutamate.1,2,3 Based on sequence homology, receptor
pharmacology and signaling pathways the eight mGluR subtypes are divided into three groups
(Group I, II, and III).
Group I mGluRs (mGluR1 and mGluR5) are mainly localized postsynaptically and are
extensively present in numerous brain areas, including the hippocampus, thalamic nuclei, and
spinal cord. Stimulation of Gq protein coupled mGluR1 and mGluR5 results in phosphoinositide
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
9
(PI) hydrolysis and intracellular formation of the second messengers inositol triphosphate (IP3),
which prompts the release of intracellular Ca2+, and diacylglycerol (DAG), which stimulates
protein kinase C activity (PKC).3,4,5 In contrast, Group II (mGluR2 and mGluR3) and group III
(mGluR4, mGluR6, mGluR7, and mGluR8) are primarily located presynaptically and, upon
stimulation, suppress adenylyl cyclase, thereby reducing intracellular cAMP levels upon
stimulation.6,7,8
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
Excessive activation of mGluR5 is related to the development of neurodegenerative diseases
like Alzheimer’s,9,10 Parkinson’s disease,11,12 or numerous CNS disorders including neuropathic
pain,13 anxiety and depression,14 drug addiction15 and fragile X syndrome. 16,17
The dopamine receptor subtypes are divided into two different subfamilies: D1ꢀlike (D1R and
D5R) and D2ꢀlike (D2R, D3R, and D4R). Mounting evidence shows that GPCRs may form homoꢀ
and heteromers at the plasma membrane. Cabello et al. first suggested the occurrence of
heterodimers of mGluR5 and D2R in living cells using a fluorescence complementation assay.
Additionally, bioluminescence resonance energy transfer, and sequential resonance energy
transfer assays also allowed detecting receptor oligomers having at least two protomers, in which
the mGluR5 and D2R also assembled with the adenosine A2A receptor (A2AR). Coꢀ
immunoprecipitation in native tissue confirmed the oligomerization of mGluR5, D2R and A2AR
in rat striatum homogenates. Overall, the above results indicate the existence of mGluR5ꢀD2Rꢀ
A2AR heteromers in living cells and support their presence in the GABAergic striatopallidal
neurons, where they may represent important targets for the potential treatment of locomotion,
neuropsychiatric disorders and drug addiction.18,19
ACS Paragon Plus Environment
4

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
A recent study by Beggiato et al.20 revealed a functional role for A2AR, D2R and the mGluR5
in modulating the activity of rat basal ganglia using dualꢀprobe microdialysis in freeꢀmoving rats.
The results suggested that A2AR and mGluR5 interact synergistically in regulating the D2Rꢀ
mediated control of striatopallidal GABA neurons. D2R appears to suppress synaptic delivery of
mGluR5 by limiting Src family kinase (SFK)ꢀmediated tyrosine phosphorylation of mGluR5 in
rat striatum.21 In the basal ganglia, mGluR5 negative allosteric modulators regulate the Lꢀ3,4ꢀ
dihydroxyphenylalanine (LꢀDOPA) induced changes of D2R, the connected signaling proteins
including ERK1/2 and Akt/GSK3β and neuropeptides preproenkephalin and preprodynorphin,
along with the expression of A2AR.22 As is well known, cocaine can improve extracellular
glutamate and dopamine (DA) levels in the nucleus accumbens (NAc), which are boosted with
repeated administration.23
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
Nowadays, heterobivalent ligands have emerged as pharmacological probes to validate the
existence of GPCR dimers without the need for receptor modification.24,25,26 Heterobivalent
ligands having a linker of optimal length are envisioned to show a greater potency than that
stemmed from the sum of its two constituting pharmacophores.26, 27 , 28 Compounds able to
modulate two GPCR dimers may eventually also evolve to useful pharmacological agents that
target selected heteromeric subtypes, thereby potentially increasing selectivity.25
In this study we describe the design and synthesis of a series of mGluR5 antagonistsꢀD2R
agonist/antagonist bivalent ligands as pharmacological tools to further study mGluR5ꢀD2R
heteromers. Towards this end ligands for the protomeric D2R and the mGluR5 were conjugated
through a spacer of variable length.
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
These bivalent ligands were evaluated in radioligand binding and functional assays both in
cells in which the mGluR5 and D2R were expressed separately and in cells coꢀexpressing both
target receptors.
9
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
ACS Paragon Plus Environment
6

Page 7 of 58
Journal of Medicinal Chemistry
1
2
3
4
Results and Discussion
5
6
7
8
Design of heterobivalent ligands
9
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
We constructed the bivalent ligands around spacers of varying length that were functionalized
with a primary amine on one end and an azide on the other to allow orthogonal coupling to the
appropriately derivatized dopamine and mGluR5 ligands, respectively (Figure 1).
Figure 1. Construction of the heterobivalent ligands.
As D2R ligands we selected the 5ꢀhydroxyꢀ2ꢀ(dipropylamino)tetralin (5ꢀOHꢀDPAT)29 and the
1,4ꢀdisubstituted aromatic piperazine (1,4ꢀDAPs)30 that were equipped with an alkyne ligation
handle to afford 1 and 2, which may be swiftly connected to the azide group of the spacer via a
copperꢀcatalyzed azide–alkyne cycloaddition (CuAAC) reaction (Figure 2).
5ꢀOHꢀDPAT is a prototypical D2R/D3R agonist suitable for constructing bivalent ligands.31
Siteꢀdirected mutagenesis has demonstrated that the interaction between the 2ꢀaminotetralin
moiety and an agonist binding domain linking TM3 and TM5 leads to either D2R or D3R
activation. The 5ꢀOH group and the basic N are critical for binding affinity.
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
The D2R antagonistic activity of 1,4ꢀDAPs originates from the aromatic head group and the
amine moiety that forms a strong hydrogen bond with Asp3.32.30 A lipophilic appendage enhances
affinity. Because proper lengthening of the lipophilic appendage is anticipated to lead to the
“entrance region” of the receptor and from there reaching out to the binding pocket of the
neighboring protomer, the paraꢀposition of the benzylic appendage is considered an appropriate
attachment point for a linker unit.30
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
As mGluR5 ligand we opted for the selective and potent allosteric mGluR5 antagonist 3ꢀ[(2ꢀ
methylꢀ4ꢀthiazolyl)ethynyl]pyridine (MTEP).5 Although MTEP analogues substituted at the 5ꢀ
position of the 3ꢀpyridyl moiety showed moderately lower affinity compared to the parent
MTEP, 32 they have been demonstrated to be more potent mGluR5 antagonists than those
modified at the 2ꢀ or 6ꢀpositions23. Hence, we introduced a carboxylic acid at the 5ꢀposition of
the 3ꢀpyridyl moiety to afford two carboxylic acid analogues of MTEP 3a-b which are amenable
for coupling to the amine group of the bifunctional spacers.
ACS Paragon Plus Environment
8

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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. Overview of appropriately modified monovalent ligands (parent ligands in red) and
linkers used in this study. Parent ligands are indicated in red.
To construct the desired bivalent ligands, we considered three types of linkers with an azido
group on one and an amino group on the other end, as to allow swift conjugation to alkyne
modified dopamine ligands 1 and 2, and the carboxylic acid containing MTEP analogues 3a-b,
respectively (Figure 2).
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
The first linker 4 is composed of a simple bifunctional hexamethylene chain. The polyethylene
glycol (PEG) linkers 5a-b were previously used to construct bivalent D2RꢀµOR ligands.29 These
two types of spacers have been frequently employed to compose bivalent ligands and they allow
for a gradual increase in spacer length.28,30,31,33 In a third type of linkers 6a-b, we decided to
introduce a protonatable nitrogen atom near the center of the bifunctional polymethylene chain to
moderate incremental increase in lipophilicity upon extension, which may result in aggregation
in aqueous media.
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
Synthesis
The alkyne functionalized D2R ligands 1 and 2 were prepared as described earlier.26,29,30 The
synthesis of the MTEP derivatives equipped with a carboxylic acid handle started from thiazole
8, which was obtained from 2ꢀchloro acetyl chloride (7) in 2 steps following published
procedures.5,23 Esters 10a-b were obtained upon treatment of phenols 9a-b with ethyl
bromoacetate,34,35 and coupled with thiazole 8 in the presence of Pd(PPh3)2Cl2 to furnish 11a-b,
respectively. Saponification of 11a-b yielded carboxylic acids 3a-b.
Scheme 1. Synthesis of Carboxylic Acid Modified MTEP Analogues 3a and 3b.
ACS Paragon Plus Environment
10

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Reagents and conditions: i) Pd(PPh3)2Cl2, CuI, TBAF, Et3N, DMF, 85 °C, 2 h, 60ꢀ64%; ii) aq.
NaOH, 82ꢀ86%.
Earlier procedures were followed for the preparation of the bifunctional polymethylene linker
4,36,37 and the PEG linkers 5a-b.38
Scheme 2. Synthesis of Orthogonally Functionalized Linkers 6a and 6b.
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
Reagents and conditions: i) NaBH(OAc)3, CH2Cl2, 62ꢀ83%; ii) TFA, CH2Cl2, 93ꢀ96%.
5
6
7
8
9
Compound 13 was synthesized from the commercially available Nꢀmethylcaprolactam (12) in
4 steps.39,40,41 Aldehydes 15a-b were obtained from 14a-b in 2 steps following the existing
protocols.42 Reductive amination of 15a-b with 13 using Na(OAc)3BH yielded 16a-b, which,
after Bocꢀdeprotection with TFA, afforded the desired linkers 6a-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
The bivalent ligands were prepared by first coupling the mGluR5 ligands with the different
spacers using EDC as the coupling reagent in the presence of Et3N to furnish the seven key
intermediate azides 17, 19a-b and 21a-d, which were subsequently conjugated with the DPATꢀ
and DAPꢀalkynes via a click reaction (CuAAC) (Scheme 3), leading to a series of nine bivalent
D2RꢀmGluR5 ligands.
Scheme 3. Synthesis of the Target Bivalent Ligands 18, 20a-b, 22a-b and 23a-d.
ACS Paragon Plus Environment
12

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
Reagents and conditions: i) (3ꢀdimethylaminopropyl)ꢀNꢀethylcarbodiimide hydrochloride
(EDC·HCl), 1ꢀhydroxybenzotriazole, Et3N, CH2Cl2, 51ꢀ94%; ii) sodium ascorbate, CuSO4, Et3N,
tris[(1ꢀbenzylꢀ1,2,3ꢀtriazolꢀ4ꢀyl)methyl]amine, DMF, 38ꢀ58%.
Pharmacological evaluation
Affinity of the bivalent ligands for the D2R
We performed a competition assay to evaluate the binding affinity of the bivalent ligands for
the D2R (Table 1). First, the binding affinity of nine bivalent ligands along with the alkynylated
precursors of DPAT (1) and DAP (2) was measured by displacement of [3H]raclopride from the
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
5
6
7
8
9
D2R stably expressed in HEK 293T cells. Raclopride is a selective D2R antagonist that due to its
more pronounced hydrophilic character shows substantially less nonꢀspecific binding than
spiperone.29 All bivalent ligands retained good to moderate affinity for the D2R (Table 1). The
DPATꢀbased ligand 22a, featuring a type 3 linker consisting of 20 atoms showed the best affinity
(50 ± 6.4 nM) for D2R. Interestingly, amongst the DAPꢀbased ligands compound 23a and 23c,
featuring the same spacer as 22a, displayed the lowest Ki value (132 nM and 147 nM,
respectively). A significant reduction (5ꢀ to 17ꢀfold) in D2R binding affinities was observed for
the bivalent ligands with hexamethylene and PEG linkers.
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
Table 1. Binding affinities (Ki) for D2R and D2RꢀmGluR5.
Compd X
R
Linker
Ki (nM) for D2R Ki (nM) for
D2RꢀmGluR5
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
17 ± 2.6
77 ± 6.8
295 ± 99
50 ± 6.4
79 ± 16
19 ± 1.8
69 ± 22
1
2
N
N
N
DPAT ꢀOCH2CONH (CH2)6ꢀ
142 ± 10
13 ± 3.6
171 ± 12
18
DPAT ꢀOCH2CONH(CH2)5NCH3(CH2)8ꢀ
DPAT ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ
22a
22b
ACS Paragon Plus Environment
14

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
N
DAP
DAP
ꢀOCH2CONH (CH2)5NCH3(CH2)8ꢀ 132 ± 44
ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ 303 ± 43
ꢀOCH2CONH (CH2)5NCH3(CH2)8ꢀ 147 ± 24
ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ 398 ± 100
45 ± 5.3
115 ± 32
94 ± 21
23a
23b
23c
23d
20a
20b
CH DAP
CH DAP
CH DAP
CH DAP
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
252 ± 54
251 ± 42
186 ± 75
ꢀOCH2CONHꢀPEG4ꢀ
ꢀOCH2CONHꢀPEG6ꢀ
438 ± 89
352 ± 94
Binding affinities (Ki) obtained by competitive displacement of [3H]raclopride from HEK293T
D2R and HEK293T D2RꢀmGluR5 cell membranes. All values are expressed as the mean ± SEM
of three independent assays.
Next, we determined the affinity of nine bivalent ligands and their corresponding alkyne
precursors 1 and 2 for D2R in the presence of overexpressed mGluR5. Interestingly, the affinity
of 22a to cells expressing both D2R and mGluR5 (13 ± 3.6 nM) increased 4ꢀfold compared to
cells expressing only D2R (50 ± 6.4 nM). Likewise, the D2R affinity of the DAPꢀcontaining
bivalent ligand 23a was 3ꢀfold higher in cells coꢀexpressing the mGluR5 (45 ± 5.3 nM versus
132 ± 44 nM). Furthermore, 22a showed moderately higher affinity than 1 (19 ± 1.8 nM) and 10ꢀ
fold higher affinity compared to the two other DPATꢀbased bivalent ligands 18 and 22b in D2R
and mGluR5 coꢀexpressing cells. All these data indicate that bivalent ligands 22a and 23a
possibly bridge the binding sites of the D2R and mGluR5 heterodimers. Furthermore, DPATꢀ
based compound 18 containing a short hexamethylene linker showed a lower affinity for D2Rꢀ
mGluR5 relative to 22a probably because its linker is too short to bridge the binding sites of the
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
heterodimer. The preference for a tertiary amine linker as deduced from the superior affinities of
22a and 23a in the coexpressing cells may be partly due the Hꢀbond that is formed between the
tertiary amine group and D400 located on extracellular loop 3 (ECL3) of D2R (see
Computational modeling part below), which may compensate for the entropic penalty of the
flexible linker. Based on the above results, a type 3 linker of appropriate length is preferred in
this study.
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
Affinity of the bivalent ligands for the mGluR5
Competitive ligand binding assays in plasma membranes from HEK293 cells stably expressing
mGluR5 in the absence or presence of D2R allowed to assess the binding affinity for said
receptor using [3H]MPEP, a nonselective mGluR5 antagonist. Results of saturation binding
assays for [3H]MPEP can be found in supporting information. The binding affinities of the azido
precursor 21a for mGluR5 and D2RꢀmGluR5 were 5.8 ± 0.9 µM and 5.7 ± 0.8 µM, respectively
(Table 2). In addition, the affinity of 3a for mGluR5 was 6.5 ± 0.6 µM, which is comparable to
that of 21a. This led us to use 21a as control in ligand binding and further functional assays for
mGluR5.
Bivalent ligand 22a showed a 2ꢀfold gain in binding affinity in cells expressing both D2R and
mGluR5 compared to cells solely expressing mGluR5. Interestingly, 22a exhibited a 5ꢀfold
higher affinity (1.1 ± 0.2 µM) for D2RꢀmGluR5 compared to its azido precursor 21a lacking a
D2R ligand (5.7 ± 0.8 µM). Likewise, 23a displayed 4ꢀfold higher affinity (1.3 ± 0.2 µM) for
D2RꢀmGluR5 relative to 21a. All these results indicate that bivalent ligands 22a and 23a bridge
the orthosteric and allosteric binding sites of the D2R and mGluR5 heteromer, respectively.
ACS Paragon Plus Environment
16

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
Table 2. Binding affinities (Ki) for mGluR5 and mGluR5ꢀD2R.
5
6
7
Entry X R
Linker
Ki (µM) for mGluR5 Ki (µM) for
8
9
D2RꢀmGluR5
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
N ꢀ
ꢀOCH2CONH(CH2)5NCH3(CH2)8ꢀ 5.8 ± 0.9
5.7 ± 0.8
1.1 ± 0.2
1.3 ± 0.2
21a
22a
23a
N DPAT ꢀOCH2CONH(CH2)5NCH3(CH2)8ꢀ 2.3 ± 0.5
N DAP
ꢀOCH2CONH (CH2)5NCH3(CH2)8ꢀ 1.0 ± 0.3
Binding affinities (Ki) obtained by competitive displacement of [3H]MPEP from HEK293T
cell membranes expressing mGluR5 and D2RꢀmGluR5. All values are expressed as the mean ±
SEM of three independent assays performed in duplicate.
cAMP assay to study intrinsic activities of the selected D2R ligands.
To further understand the intrinsic activities of nine bifunctional ligands and their
corresponding monovalent ligands 1 and 2 consisting of the pharmacophore of a D2R agonist and
a D2R antagonist, respectively, we performed a bioluminescence based cAMP accumulation
assay (Figure 3 and Table 3).43 Coupled to inhibitory Gαi/o proteins, activation of the D2R
decreases cAMP production.
Similar to the reference D2R agonist quinpirole, bivalent ligands 18, 22a, 22b and their alkyne
precursor 1 inhibited forskolinꢀstimulated cAMP formation both in cells monoexpressing D2R or
in cells coexpressing D2R and mGluR5. Interestingly, in the latter cells a significantly lower
EC50 value was observed for bivalent ligand 22a (EC50 = 1.5 ± 0.5 nM) compared to 1 (EC50 =
6.5 ± 0.6 nM), indicating that the bivalent binding requires a lower receptor occupation to exert
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
8
9
signaling. However, 18 and 22b exhibited 3ꢀ and 5ꢀfold lower potencies (20 nM and 32 nM,
respectively) in the coexpressing cell line relative to 1, revealing weak D2R agonism. In
D2R/mGluR5ꢀcoexpressing cells, quinpirole and the monovalent D2R agonist 1 inhibited cAMP
accumulation with similar potencies as in cells monoexpressing D2R, while 22a decreased cAMP
production with a 2ꢀfold higher potency in the coexpressing cells than in monoexpressing D2R
cells. Nevertheless, 18 and 22b displayed 2ꢀ and 6ꢀfold lower potencies in the coexpressing cells
than in the monoexpressing cells.
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
In addition, bivalent ligands based on DAPꢀalkyne and their alkyne precursor 2 failed to inhibit
cAMP accumulation (Figure 3 and supporting information) both in the D2R solely expressed
cells and in the D2R/mGluR5 coexpressed cells, suggesting that DAPꢀbased bivalent ligands
show D2R antagonist behavior, which is in line with the reported results.30
Figure 3. Doseꢀresponse curves of cAMP accumulation for representative ligands. Functional
activity of selected bivalent ligands and their precursors 1 and 2 determined in HEK 293T cells
stably expressing the D2R (left, a) or both D2R and mGluR5 (right, b) and coexpressing the
pGloSensorꢀ22F cAMP plasmid via transient transfection. Cells were stimulated with the
selected ligands of varying concentrations (0.1 nM–10 µM) in the presence of 10 µM forskolin.
ACS Paragon Plus Environment
18

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
cAMP production was normalized to the percentage of forskolin stimulated cAMP accumulation
(100%). Data represent mean ± SEM of three independent experiments each performed in
duplicate.
9
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
Table 3. Potency of inhibition of forskolin mediated cAMP accumulation for bivalent ligands
based on D2R agonist DPATꢀalkyne.
Compd
cAMP EC50 (nM)
D2R
cAMP EC50 (nM)
D2RꢀmGluR5
1.9 ± 0.3
Quinpirole
1.8 ± 0.4
3.0 ± 0.8
7.6 ± 0.7
3.3 ± 0.7
5.5 ± 0.6
6.5 ± 0.6
1
20 ± 9.8
18
1.5 ± 0.5
22a
22b
32 ± 15
Data shown represent the mean ± SEM of three independent experiments.
MAPK phosphorylation to study the potency of the newly developed bivalent ligands for
mGluR5
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
6
7
8
9
It is well known that activation of mGluR5 by its agonist (R,S)ꢀ2ꢀchloroꢀ5ꢀ
hydroxyphenylglycine (CHPG) leads to phosphorylation of MAPK.44 Here, we examined the
MAPKꢀP in response to Lꢀglutamic acid (LꢀGlu) and CHPG treatment in HEK293 mGluR5 and
HEK293 mGluR5ꢀD2R stable cell lines (supporting information). The results showed that
MAPKꢀP peaked at approximately 10 min for mGluR5 and 7ꢀ10 min for mGluR5ꢀD2R after 10
µM LꢀGlu treatment. To assess the inhibition of MAPK phosphorylation by the bivalent ligands,
cells were pretreated with 21a and all bivalent ligands for 30 min before incubation of the cells
for 10 min with LꢀGlu (10 µM). All bivalent ligands proved capable of inhibiting MAPKꢀP
activated by LꢀGlu (supporting information).
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
To assess their potency we tested all bivalent ligands and azido precursor 21a (1 nMꢀ10 µM)
in HEK 293 mGluR5 cells (Table 5) and selected compounds 21a, 22a and 23a in HEK 293
mGluR5ꢀD2R cells (Table 5). Compound 22a was more potent (0.17 ± 0.02 µM) than the other
DPATꢀbased bivalent ligands in blocking MAPKꢀP for mGluR5. DAPꢀbased bivalent ligand 23a
featuring an identical linker as 22a displayed a higher potency (0.11 ± 0.05 µM) than the other
DAPꢀbound ligands in the cells monoexpressing the mGluR5. In addition, DAPꢀbased bivalent
ligands with PEG linkers showed lower antagonistic activity compared to DAPꢀbased ligands
with a tertiary amine linker.
Interestingly, bivalent ligands 22a and 23a exhibited 7ꢀ and 2ꢀfold higher potency (IC50 = 0.62
µM and 1.8 µM) in mGluR5ꢀD2R compared with their monovalent precursor 21a (IC50 = 4.5
µM). While these two compounds were more potent in the mGluR5ꢀD2R coexpressing cells than
21a, neither of them was more potent in the coexpressing cells than in the cells expressing
mGluR5 only.
ACS Paragon Plus Environment
20

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
Table 5. Potency (IC50) of mGluR5 antagonists blocking MAPKꢀP activated by LꢀGlu (10
5
6
µM) in HEK293 mGluR5 and HEK293 mGluR5ꢀD2R stable cell lines.
7
8
9
Compd X
R
Linker
MAPK
IC50 µM IC50
mGluR5
MAPK
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
µM
mGluR5ꢀD2R
4.5 ± 1.1
nd
N
N
N
N
N
N
ꢀ
ꢀOCH2CONH(CH2)5NCH3(CH2)8ꢀ 2.6 ± 0.8
21a
18
DPAT ꢀOCH2CONH (CH2)6ꢀ
1.6 ± 0.7
DPAT ꢀOCH2CONH(CH2)5NCH3(CH2)8ꢀ 0.17 ± 0.02 0.62 ± 0.15
DPAT ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ 0.32 ± 0.06 nd
22a
22b
23a
23b
23c
23d
20a
20b
DAP
DAP
ꢀOCH2CONH (CH2)5NCH3(CH2)8ꢀ 0.11 ± 0.05 1.8 ± 0.5
ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ 0.45 ± 0.15 nd
ꢀOCH2CONH (CH2)5NCH3(CH2)8ꢀ 0.24 ± 0.08 nd
ꢀOCH2CONH (CH2)5NCH3(CH2)10ꢀ 0.31 ± 0.11 nd
CH DAP
CH DAP
CH DAP
CH DAP
ꢀOCH2CONHꢀPEG4ꢀ
ꢀOCH2CONHꢀPEG6ꢀ
1.2 ± 0.3
nd
0.90 ± 0.32 nd
Data shown represent the mean ± SEM of three independent experiments. nd, not determined.
Computational modeling of a bivalent ligand bound to an mGluR5-D2R heterodimer
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
5
6
7
8
9
The binding pose of compound 22a was investigated in silico through proteinꢀprotein docking
of crystal structures: transmembrane domains (TMD) of mGluR5 and D2R, followed by docking
of the bivalent ligand into the modeled heterodimer (see Experimental Section). As both
receptors are coꢀcrystallized with either negative allosteric modulator (NAM)45 or antagonist46,
they can be considered to be in their respective inactive states. As the bivalent ligand 22a is a
fusion of mGluR5 NAM and D2R agonist, this means mGluR5 is in an ideal conformation but
not D2R, which would theoretically be active in such a heterodimerꢀligand complex.
Furthermore, as the original crystal structure of mGluR5 has no clear entrance or exit for NAMs
because of its tightly closed extracellular loops and TMD,45 a modified conformation of the
receptor was deemed necessary to correctly bind a bivalent compound. Such a conformation has
been observed previously, firstly with a Monte Carlo sampling technique employed for the
purposes of docking photoꢀsensitive47 and long linear48 mGluR5 NAMs, and secondly during
molecular dynamics (MD) simulations of mGluR5 with bound NAM 49 . In summary, this
alternative mGluR5 conformation is in an inactive state and similar to the original mGluR5
crystal structure45 but undergoes a partial separation of the extracellular ends of TM5 and TM6,
which are highly flexible.49 This potentially allows NAMs to enter or leave and is conducive for
the linker of a bivalent ligand to pass outside of the receptor. As such, we utilized this MDꢀ
generated mGluR5 conformation49 for modeling interactions with D2R, whose crystal structure
has no such limitations regarding the binding of a bivalent ligand.
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
In terms of modeling an mGluR5ꢀD2R heterodimer, referring to experimental information is
important as these proteins can potentially interact in a number of ways. Interestingly, TM5 and
TM6 have been experimentally observed to dynamically participate in the interface(s) of an
mGluR2 homodimer,50 as well as TM5 involved in crossꢀtalk between mGluR2 and the Class A
ACS Paragon Plus Environment
22

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
GPCR, 5ꢀHT2A receptor.51 , 52 , 53 Based on homology between mGluR2 and mGluR5 TMDs
(sequence identity of 51%), physical interaction between mGluR5 and D2R may follow a similar
pattern. In addition, participation of mGluR5 TM5 and/or TM6 in a direct interaction with D2R
would provide a means of passing a bivalent ligand between receptors. Likewise for D2R, TM5
and TM6 have been observed to participate in its experimentally characterized homodimeric and
heteromeric receptor interactions.54,55 ,56,57 As such, we modeled an mGluR5ꢀD2R heterodimer
via its respective TM5ꢀTM6 helices by using the muꢀopioid homodimer crystal structure as a
template,58 which contains an analogous interface. This initial mGluR5ꢀD2R model was refined
with proteinꢀprotein docking to identify an optimal interaction between the receptors59 (see
Experimental Section). The best docked mGluR5ꢀD2R conformation results in close packing of
residues at the TM5ꢀTM6 interface (Figure 4B). However, as D2R is in an inactive state with its
TM6 in an inward rather than outward orientation,46 the contact surface between receptors is
possibly not as extensive as it could be. Nevertheless, the reported docking score is respectable,
suggesting a favorable interaction (I_sc of ꢀ6.6 on a scale of 0 to ꢀ10, where better than ꢀ5 is
considered satisfactory59).
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
Compound 22a was bound to the mGluR5ꢀD2R heterodimer model by first, automatically
docking its separate mGluR5 NAM (MTEP) and D2R agonist (DPATꢀalkyne, 1) components,
second, by manually connecting these with the appropriate linker, and third, energy minimizing
the resulting complex (see Experimental Section). It was found that by following this process,
22a is able to simultaneously bind both receptors, passing between TM5 and TM6 of each
(Figure 4A) with the linker residing at the membrane surface or just beneath (Figure 4B). In the
modeled complex, compound 22a is able to make six proteinꢀligand Hꢀbonds: two with mGluR5
and four with D2R (Figure 4C). In mGluR5, these Hꢀbonds involve residues S805 and S809 on
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
7
8
9
TM7, both of which interact with the MTEP component of 22a (Figure 4C). These observed
proteinꢀligand Hꢀbonds are consistent with known mGluR5 NAM binding modes previously
identified in mGluR5 crystal structures45,60 and published mGluR5 MD simulations.47,49 In D2R,
the DPAT component of 22a makes an Hꢀbond with D114 on TM3 in the same manner as coꢀ
crystallized risperidone in the original D2R crystal structure.46 However, in line with its potential
D2R agonism and unlike risperidone, 22a makes additional Hꢀbonds with the backbone oxygen
of S193 on TM5 and Y408 hydroxyl group on TM7 (Figure 4C). The importance of S193 for the
binding of D2R agonists has been noted before,61 along with neighboring residues, S194 and
S197. Due to the closeness of these three particular serine residues in the D2R crystal structure,
only moderate conformational changes in TM5 (due to receptor activation, for example) would
be required for 22a to make Hꢀbonds with the sidechain hydroxyl groups of either S193, S194 or
S197. Finally, the tertiary amine on the linker of 22a makes an Hꢀbond with D400 located on
extracellular loop 3 (ECL3) of D2R (Figure 4C). This interaction is likely conducive to the
conformational stability of this particular linker, which might otherwise (e.g. PEGꢀlinkers)
become a destabilizing factor.
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
Overall, due to the number of observed proteinꢀligand Hꢀbonds with both receptors, as well as
consistency with known mGluR5 and D2R ligand binding modes, the modeled heterodimerꢀ
ligand complex appears to be compatible with the reported biological activity of compound 22a.
ACS Paragon Plus Environment
24

Page 25 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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 4. The modeled mGluR5-D2R-compound 22a complex. A) Extracellular topꢀview of
the interaction between mGluR5 (blue ribbon, derived from PDB id: 4009) and D2R (green
ribbon, PDB id: 6CM4) via symmetrical TM5ꢀTM6 interface, including bound compound 22a
(carbons as pink spheres). B) Sideꢀview of mGluR5ꢀD2R heterodimer with bound 22a (carbon,
nitrogen, oxygen and sulphur atoms as pink, dark blue, red and yellow spheres, respectively),
showing contact interface (carbons as blue and green spheres, respectively) within a model
membrane (red dots and blue lines, representing upper and lower leaflets, respectively). C)
Modeled binding mode of compound 22a (pink) within respective allosteric and orthosteric
pockets of mGluR5ꢀD2R heterodimer (blue and green ribbons, respectively). Proteinꢀligand Hꢀ
bonds are displayed as solid black lines (S805 is partially obscured by the ligand). Selected
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
residues on mGluR5: TM6, TM7 and D2R: TM3, TM5, TM7 and ECL3 are displayed (mGluR5:
TM3, TM5 and D2R: TM6 are transparent for visualization purposes).
7
8
9
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
Conclusion
To conclude, we have designed and synthesized nine potential bivalent ligands comprising two
distinct D2 agonist/antagonist and two mGluR5 antagonists. Ligation of D2 ligands shows that
bivalent ligands with PEG (PEG4 and PEG6) and hexamethylene linkers exhibit relatively low
affinity for D2R, contrary to bivalent ligands with alkylamine spacers. Interestingly, bivalent
ligand 22a, with a 20ꢀatom alkylamine spacer, showed a 4ꢀfold affinity increase for the D2R in
cells expressing both the D2R and the mGluR5, compared to cells only expressing D2R,
suggesting that 22a may bridge the binding sites of the D2R and mGluR5 receptors. Moreover,
22a shows a 5ꢀfold higher affinity for the mGluR5 than its MTEP azido precursor 21a in D2Rꢀ
mGluR5 cells, also pointing towards its capacity to simultaneously occupy both binding sites of
the heteromer. From a functional point of view, 22a displays 7ꢀfold higher potency compared to
the monomeric precursor 21a in MAPK phosphorylation in cell lines stably expressing D2R and
mGluR5 receptors. Moreover, 22a inhibits forskolin stimulated cAMP formation with a 4ꢀfold
higher potency compared to alkyne 1 in the coexpressing cells, indicating that the bivalent
binding requires a lower receptor occupation to exert signaling. In addition, a docking study
reveals that 22a is indeed able to simultaneously bind both receptors by passing between the
heterodimeric interface, comprised of TM5 and TM6 of both receptors, and establishing six
proteinꢀligand Hꢀbonds.
ACS Paragon Plus Environment
26

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Overall, we demonstrated that the length and nature of the spacer between the two receptor
pharmacophores strongly affect binding affinity and functional potency both from a D2R and
from an mGluR5 viewpoint. Most importantly, this study allowed us to identify bivalent ligand
22a, which emerges as a promising molecular probe to further investigate D2R and mGluR5
heterodimerization.
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
1. Chemistry
General
Commercially available reagents and solvents were used as received without any further
purification. All reactions were carried out under an Argon or N2 atmosphere and at ambient
temperature unless indicated otherwise. NMR solvents were acquired from Eurisotop (Saintꢀ
Aubin, France). TLC analysis was performed to monitor reactions using TLC aluminium sheets
(MachereyꢀNagel, Alugram Silica G/UV254) with detection by spraying with a solution of
(NH4)6Mo7O24·4H2O (25 g/L) and (NH4)4Ce(SO4)4·2H2O (10 g/L) in H2SO4 (10 % aq.) followed
by charring or an aqueous solution of KMnO7 (20 g/L) and K2CO3 (10 g/L) or an ethanolic
solution of ninhydrin (2 g/L) and acetic acid (1% v/v) followed by charring. Column
chromatography was carried out using a Grace Reveleris X2 automated flash system and the
1
13
corresponding disposable silica gel cartridges. H and C NMR spectra were recorded with a
Varian Mercury 300/75 MHz spectrometer in CDCl3 or DMSOꢀd6. Chemical shifts are given in
1
13
parts per million (ppm δ) relative to the residual solvent peak or TMS for H and C. Coupling
constants are given in Hertz (Hz). Exact mass measurements (ESIꢀHRMS) were performed with
a Waters LCT Premier XE Mass spectrometer calibrated using leucine enkephalin as an external
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
standard. The purity of all bivalent ligands was determined by LCꢀMS analyses on a Waters
Alliance 2695 XE separation module using a Phenomenex Kinetex EVO C18, 5µm 100 mm ×
2.1 mm column and a gradient system of HCOOH in H2O (0.1%, v/v)/HCOOH in CH3CN
(0.1%, v/v) at a flow rate of 0.6 mL/min, 5:95 to 100:0 (5 to 100% CH3CN) in 15 min. All
bivalent ligands showed >95% purity.
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
General Procedure 1: EDC-Mediated Amide Formation.
To a cooled (0 °C) solution of EDC·HCl (1.5 eq.) and 1ꢀhydroxybenzotriazole (1.4 eq.) in
CH2Cl2 or DMF was added carboxylic acid 3a or 3b (1.2 eq.) under N2 atmosphere. After 15 min
at 0 °C, functionalized linkers (1.0 eq.) were added and Et3N (2 eq.) was added dropwise. The
iceꢀwater bath was removed and the reaction was slowly warmed to room temperature and stirred
overnight. After reaction, the solvent was evaporated under reduced pressure. The crude
compound was purified by flash column chromatography (NH4OH/MeOH/CH2Cl2, 1:5:94 v/v/v)
to afford the amides.
General Procedure 2: Click Reaction (CuAAC).
To a solution of the key intermediate azide (1.0 eq.) in DMF (0.1 M) was added the DPATꢀ
alkyne or DAPꢀalkyne (1.5 eq.), sodium ascorbate (1.0 eq.), CuSO4 (0.2 eq.), Et3N (3.0 eq.) and
a catalytic amount of tris[(1ꢀbenzylꢀ1,2,3ꢀtriazolꢀ4ꢀyl)methyl]amine. The reaction solution was
stirred overnight at RT in the dark under nitrogen atmosphere. After reaction, the mixture was
filtered and solvent was evaporated under reduced pressure. The crude product was purified by
flash column chromatography (NH4OH/MeOH/CH2Cl2, 1:5:94 v/v/v) to give the final conjugate
as a white solid.
ACS Paragon Plus Environment
28

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2-((5-((2-Methylthiazol-4-yl)ethynyl)pyridin-3-yl)oxy)acetic acid (3a). To a solution of
ester 11a (2.3 g, 7.6 mmol) in methanol was added 4 M NaOH (38 mmol, aq.) and then stirred at
room temperature for 2 h. Then 1 M HCl (aq.) was added to the solution to acidify it till pH=2.
Next the solvent was removed under reduced pressure and the residue was purified by silica gel
chromatography (AcOH/Hexane/EtOAc 1:49:50 v/v/v) to yield acid 3a as a pale yellow solid
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
(1.8 g, 86%). H NMR (300 MHz, DMSOꢀd6) δ ppm 12.37 (br s., 1H), 8.29 (br s, 2H), 7.90 (s,
1H), 7.52 (s, 1H), 4.82 (s, 2H), 2.63 (s, 3H). 13C NMR (75 MHz, DMSOꢀd6) δ ppm 172.4, 166.3,
151.6, 144.4, 138.8, 135.2, 125.6, 123.1, 119.1, 87.4, 85.1, 65.1, 19.0. HRMS (ESI) m/z:
calculated for C13H11N2O3S [M+H]+ 275.0490; found 275.0478.
2-(3-((2-Methylthiazol-4-yl)ethynyl)phenoxy)acetic acid (3b). Compound 3b was prepared
1
using the procedure described as 3a. Pale yellow solid (2.5 g, 82%). H NMR (300 MHz,
DMSOꢀd6) δ ppm 12.88 (br s., 1H), 7.88 (s, 1H), 7.32 (t, J = 8.4 Hz, 1H), 7.14 – 7.09 (m, 1H),
7.03 (dt, J = 2.5, 0.9 Hz, 1H), 6.98 (ddd, J = 8.3, 2.7, 1.0 Hz, 1H), 4.71 (s, 2H), 2.64 (s, 3H). 13C
NMR (75 MHz, DMSOꢀd6) δ 170.4, 166.1, 158.1, 135.7, 130.4, 124.9, 124.6, 123.0, 117.0,
116.5, 88.2, 84.4, 64.9, 19.1. HRMS (ESI) m/z: calculated for C14H12NO3S [M+H]+ 274.0538;
found 274.0544.
N1-(8-Azidooctyl)-N1-methylpentane-1,5-diamine (6a). To a solution of 16a (2.3 g, 6.2
mmol) in CH2Cl2 (20 ml) was added TFA (10 ml) and stirred at RT for 6 h. After reaction, the
solution was basified with 4 M NaOH (aq.) till pH=8. The mixture was filtered and concentrated
1
under reduced pressure to give the amine liner 6a as a colorless oil (1.6 g, 96%). H NMR (300
MHz, CDCl3) δ 5.81 (br s., 2H), 3.24 (t, J = 6.9 Hz, 2H), 2.78 (t, J = 7.1 Hz, 2H), 2.58 – 2.45 (m,
13
4H), 2.37 (s, 3H), 1.56 (dt, J = 13.9, 7.0 Hz, 6H), 1.37ꢀ1.24 (m, 12H). C NMR (75 MHz,
ACS Paragon Plus Environment
29