3-(2-Carboxyethyl)indole-2-carboxylic Acid Derivatives: Structural Requirements and Properties of Potent Agonists of the Orphan G Protein-Coupled Receptor GPR17

Article abstract of DOI:10.1021/acs.jmedchem.7b01768

The orphan receptor GPR17 may be a novel drug target for inflammatory diseases. 3-(2-Carboxyethyl)-4,6-dichloro-1H-indole-2-carboxylic acid (MDL29,951, 1) was previously identified as a moderately potent GPR17 agonist. In the present study, we investigated the structure-activity relationships (SARs) of 1. Substitution of the indole 1-, 5-, or 7-position was detrimental. Only small substituents were tolerated in the 4-position while the 6-position accommodated large lipophilic residues. Among the most potent compounds were 3-(2-carboxyethyl)-1H-indole-2-carboxylic acid derivatives containing the following substituents: 6-phenoxy (26, PSB-1737, EC₅₀ 270 nM), 4-fluoro-6-bromo (33, PSB-18422, EC₅₀ 27.9 nM), 4-fluoro-6-iodo (35, PSB-18484, EC₅₀ 32.1 nM), and 4-chloro-6-hexyloxy (43, PSB-1767, EC₅₀ 67.0 nM). (3-(2-Carboxyethyl)-6-hexyloxy-1H-indole-2-carboxylic acid (39, PSB-17183, EC₅₀ 115 nM) behaved as a partial agonist. Selected potent compounds tested at human P2Y receptor subtypes showed high selectivity for GPR17. Docking into a homology model of the human GPR17 and molecular dynamic simulation studies rationalized the observed SARs.

Full text of DOI:10.1021/acs.jmedchem.7b01768

Subscriber access provided by Kaohsiung Medical University
Article
3-(-2-Carboxyethyl)indole-2-carboxylic Acid Derivatives:
Structural Requirements and Properties of Potent Agonists
of the Orphan G Protein–Coupled Receptor GPR17
Younis Baqi, Thanigaimalai Pillaiyar, Aliaa Abdelrahman, Olesja Kaufmann, Samer Alshaibani,
Muhammad Rafehi, Saman Ghasimi, Rhalid Akkari, Kirsten Ritter, Katharina Simon, Andreas
Spinrath, Evi Kostenis, Qiang Zhao, Meryem Köse, Vigneshwaran Namasivayam, and Christa E Müller
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01768 • Publication Date (Web): 26 Jul 2018
Downloaded from http://pubs.acs.org on July 26, 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 56
Journal of Medicinal Chemistry
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
-(-2-Carboxyethyl)indole-2-carboxylic Acid Derivatives:
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
Structural Requirements and Properties of Potent Agonists of the Orphan
G Protein–Coupled Receptor GPR17
1,§
2,§
2
2
Younis Baqi, Thanigaimalai Pillaiyar, Aliaa Abdelrahman, Olesja Kaufmann, Samer
2
2
2
2
2
Alshaibani, Muhammad Rafehi, Saman Ghasimi, Rhalid Akkari, Kirsten Ritter, Katharina
3
3
3
4
2
Simon, Andreas Spinrath, Evi Kostenis, Qiang Zhao, Meryem Köse, Vigneshwaran
2
2,
Namasivayam, and Christa E. Müller *
1
Department of Chemistry, Faculty of Science, Sultan Qaboos University, PO Box 36, Postal Code
123, Muscat, Oman.
2
PharmaCenter Bonn, Pharmaceutical Institute, Pharmaceutical Sciences Bonn (PSB),
Pharmaceutical Chemistry I, University of Bonn, An der Immenburg 4, 53121, Bonn, Germany.
3
Institute of Pharmaceutical Biology, Section Molecular-, Cellular-, and Pharmacobiology,
University of Bonn, Nußallee 6, 53115, Bonn, Germany.
4
CAS Key Laboratory of Receptor Research, Shanghai Institute of Materia Medica, Chinese
Academy of Sciences, 555 Zuchongzhi Road, Pudong, Shanghai 201203, China.
KEYWORDS
Agonist, Central Nervous System, GPR17, Indole, Multiple Sclerosis, Structure-Activity
Relationships, Synthesis
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 56
ABSTRACT
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
The orphan receptor GPR17 may be a novel drug target for inflammatory diseases. 3-(2-
Carboxyethyl)-4,6-dichloro-1H-indole-2-carboxylic acid (MDL29,951, 1) was previously identified
as a moderately potent GPR17 agonist. In the present study we investigated the structure-activity
relationships (SARs) of 1. Substitution of the indole 1-, 5-, or 7-position was detrimental. Only small
substituents were tolerated in the 4-position while the 6-position accommodated large lipophilic
residues. Among the most potent compounds were 3-(2-carboxyethyl)-1H-indole-2-carboxylic acid
derivatives containing the following substituents: 6-phenoxy (26, PSB-1737, EC50 270 nM), 4-
fluoro-6-bromo (33, PSB-18422, EC50 27.9 nM), 4-fluoro-6-iodo (35, PSB-18484, EC50 32.1 nM),
and 4-chloro-6-hexyloxy (43, PSB-1767, EC50 67.0 nM). (3-(2-Carboxyethyl)-6-hexyloxy-1H-
indole-2-carboxylic acid (39, PSB-17183, EC50 115 nM) behaved as a partial agonist. Selected
potent compounds tested at human P2Y receptor subtypes showed high selectivity for GPR17.
Docking into a homology model of the human GPR17 and molecular dynamic simulation studies
rationalized the observed SARs.
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
2
ACS Paragon Plus Environment

Page 3 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
G protein-coupled receptors (GPCRs) constitute one of the largest families of cell membrane
receptors, which are involved in the transduction of signals from extracellular to intracellular
1
compartments. GPCRs are at present one of the most important classes of targets for marketed drugs
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
as well as those in (pre)clinical development. It is estimated that about 30% of all approved drugs
interact with GPCRs, although for only less than 10% of the known GPCRs have drugs been
2
,3
clinically developed so far. Orphan GPCRs for which the physiological agonists are yet unknown
or unconfirmed represent an unexploited group of potential, promising drug targets. The orphan
GPCR GPR17 belongs to the δ-branch of the rhodopsin-like class A GPCR family. It is coupled to
G and G proteins mediating intracellular calcium mobilization by phospholipase C activation and a
q
i
decrease in cAMP concentrations by inhibition of adenylate cyclase. At higher agonist
s
concentrations G coupling can also be observed resulting in an increase in intracellular cAMP
4
concentrations. Northern blot and real time polymerase chain reaction (RT-PCR) analyses indicated
5,6
a predominant expression of GPR17 in the central nervous system (CNS). GPR17 is also
7
expressed in organs that typically undergo ischemic damage such as heart and kidney. The receptor
is highly conserved in vertebrates (~ 90% identity of amino acid sequence between human and
mouse or rat orthologs). GPR17 antagonists have potential for the treatment of demyelinating
diseases such as multiple sclerosis based on the observation that GPR17 knockout mice showed
7
–12
increased myelination.
Activation of GPR17, on the other hand, might be useful for treating
allergic bronchial inflammation since GPR17 knockout mice displayed an increased response to
13
allergic stimuli.
To study the (patho)physiological roles of GPR17 and to allow target validation, selective ligands,
agonists and antagonists are required. Although several classes of physiological compounds have
been reported to activate GPR17 including nucleotides (UDP) and nucleotide-sugars (UDP-glucose
5
and UDP-galactose) as well as cysteinylleukotrienes (CysLTC4 and CysLTD4), these data could not
4
,14–17
be confirmed by several other laboratories including ours.
Thus, the identity of the endogenous
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 56
4
ligand(s) of GPR17 still remains unresolved. Hennen et al. recently reported on 3-(2-carboxy-4,6-
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
dichloroindol-3-yl)propionic acid (MDL 29,951, 1, Figure 1) as a synthetic agonist for GPR17,
which gives a robust response and has been broadly characterized in recombinant and native
4
,18,19
cells.
Originally, compound 1 had been reported as an antagonist of the N-methyl-D-aspartate
2
0,21
(NMDA) receptor interacting with the glycine binding site.
In order to obtain more potent and
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
selective GPR17 agonists, we previously developed an improved synthesis of 4- and 6-substituted 2-
carboxy-1H-indole-3-propionic acid derivatives employing Japp-Klingemann and Fischer indole
reaction procedures starting from differently substituted aniline derivatives, and performed
1
8
preliminary SAR studies.
In the present study, we broadly investigated the SARs of lead structure 1 with the goal to improve
its potency. Moreover, we studied the molecular interactions of this class of compounds by docking
them into a homology model of GPR17 based on the X-ray structure of the related P2Y12 receptor.
CO H
2
Cl
5
4
3
CO H
6
1
2
N
Cl
H
1
(MDL 29,951)
4
,18
Figure 1. The first reported indole derivative with GPR17-activating properties
RESULTS AND DISCUSSION
Chemistry. The desired indole derivatives were synthesized via Japp-Klingemann condensation
4
,18–21
followed by Fischer indole (diaza–Cope) rearrangement as depicted in Scheme 1.
In the first
step, aniline and its derivatives (2) were suspended in aqueous hydrochloric acid solution (5-N) and
cooled in an ice-bath with continuous stirring, followed by the dropwise addition of an aqueous
sodium nitrite solution, keeping the temperature between 0–5 °C to yield the diazonium salts 3. To
4
ACS Paragon Plus Environment

Page 5 of 56
Journal of Medicinal Chemistry
the stirred solution of 3 a previously prepared cooled solution of 2-(ethoxycarbonyl)cyclopentanone
anion (5) was added.
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
a,b
Scheme 1. Synthesis of 3-(2-Carboxyindol-3-yl)propionic Acid Derivatives
R
R
CO H
(
i)
2
CO Et
2
NH2
N
N
2
3
CO Et
2
(iv)
(iii)
R
CO Et
2
R
Route A
N
N
N
H
H
O
O
6
8
CO Et
(ii)
CO Et
2
2
Route B (iv)
CO Et
4
5
2
CO Et
2
R
(vi)
N
N
H
7
R = mono-, di- or tri-substitution
for substitution patterns see Table 1
(
v)
CO Et
CO H
2
2
(vi)
R
CO Et
R
CO H
2
2
N
H
N
H
8
9-45
a
Reagents and conditions: (i) HCl (5 N), NaNO
2
(2 equiv.), H
2
O, 0–5 °C, 20 min; (ii) KOH (9 N) (9
equiv.), ethanol, 0–5 °C, 1h; (iii) ice, 40 °C, 1h; (iv) conc. H
2
SO (5.1 equiv.), ethanol, 100 °C, 3 h;
4
(v); p-toluenesulfonic acid (p-TSA) ·H
2
2
O (1.5 equiv.), toluene (dry), reflux, 5 h; (vi) LiOH·H O (3
b
equiv.), tetrahydrofuran (THF) : H
2
O (1:1 v/v), rt, 24 h, 95–100%. For R see Table 1.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 56
The resulting mixture was treated with ice and then stirred at 40 °C until the phenylhydrazine-mono-
ethyl esters 6 were formed as monitored by TLC. Carboxylic acids 6 were then treated in two
different ways (route A or route B, Scheme 1). In route A direct cyclization of compounds 6
producing indoles 8 was achieved by treatment with sulfuric acid in refluxing ethanol at 100 °C;
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
4
,19–21
however, the products were obtained in relatively low yields.
Subsequent saponification led to
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
the desired indoles bearing two free carboxylic acid functions (9–24, Scheme 1). Following route B,
compounds 6 were refluxed in ethanol in the presence of sulfuric acid to obtain the di-ethyl esters of
the phenylhydrazones 7, which were subsequently cyclized to the corresponding indole-diethyl ester
18
derivatives 8. Finally, compounds 8 were saponified to yield the desired indoles 25–36 and 39–43
containing two carboxylic acid functions (Scheme 1, Table 1).
For unsymmetrically substituted aniline derivatives (2), the phenylhydrazones 7 underwent
cyclization reaction furnishing two regioisomeric products (8). The ratio of regioisomers 8 formed
was mainly dependent on the electronegativity, and on the steric hindrance of the substituents. All
regioisomers bearing di-ester functions (compounds 8) were separable into two regioisomeric
products using silica gel column chromatography, except for one product with chlorine and bromine
substitution in the 4- and 6-position, respectively (8a, Scheme 2). The other regioisomeric di-esters
bearing fluorine and bromine, fluorine and iodine, or chlorine and alkyloxy in the 4- and 6-position,
respectively, were all separable by elaborate silica gel column chromatography (for details see
1
13
Supporting Information).All final product structures were confirmed by H- and C-NMR
spectroscopy and high-performance liquid chromatography (HPLC) coupled to electrospray
ionization mass spectrometry (ESI-MS). Purity as determined by HPLC-(UV)-ESI-MS was in all
cases ≥ 95%.
6
ACS Paragon Plus Environment

Page 7 of 56
Journal of Medicinal Chemistry
Table 1. Synthesized Indole Derivatives, Analytical Data, and Potency and Efficacy to Stimulate the
Human GPR17 Determined in Calcium Mobilization Assays
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
Human GPR17
c
EC50 ± SEM (µM)
Yield
Purity
(% activation at
indicated
Efficacy
4
5
6
7
Compd.
R
R
R
R
a
b
d
(
%)
(%)
(%)
concentration)
118,20
Cl
H
H
Cl
H
75
97.3
0.331 ± 0.087
100
MDL29,951
920
H
H
F
H
H
H
H
F
H
H
H
H
H
H
H
H
16
25
37
15
45
17
35
13
95.1
95.3
99.6
95.7
99.5
95.2
95.6
98.9
> 30 (8%)
> 30 (11%)
≥ 10 (44%)
> 30 (9%)
≥ 10 (47%)
> 30 (5%)
> 30 (6%)
7.41 ± 0.31
–
–
1
020
120
Cl
H
1
–
1
2
H
H
H
H
F
F
–
1320
H
–
1
1
1
422
523
620
OCH
3
H
H
F
–
isopropyl
H
–
72
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 56
1
1
1
724
820
920
CH
Cl
H
3
H
H
H
H
H
Cl
Cl
CH
H
3
H
17
30
34
35
31
20
15
13
65
95.0
98.3
95.3
100
15.1 ± 3.83
ca. 30 (52%)
3.72 ± 0.14
> 30 (5%)
> 30 (1%)
> 30 (11%)
> 30 (16%)
> 30 (0%)
4.98 ± 1.09
76
–
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
H
Cl
H
H
84
–
2020
H
Cl
Cl
H
2
2
2
1
2
3
Cl
H
H
99.7
95.5
95.9
96.1
96.4
–
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
Cl
H
–
Cl
Cl
–
2420
for structure see above
–
2
5
H
H
H
H
benzyl
phenoxy
H
H
H
80
620
2
41
99.1
0.270 ± 0.085
> 100 (4%)
100
PSB-1837
2
7
OCH
H
3
H
H
H
H
H
H
H
Cl
F
94
99
68
42
46
100
100
–
–
820
OCH
> 100 (35%)
e
2
3
29
H
Cl
98.1
98.3
97.9
> 100 (5%)
7.90 ± 1.79
4.83 ± 0.49
–
3
3
0
1
H
Cl
87
98
H
Br
F
f,g
32
Cl (Br)
H
Br (Cl)
H
42
96.9
0.210 ± 0.040
94
3
3
F
H
H
H
Br
F
H
H
H
13
11
17
98.9
98.1
99.2
0.0279 ± 0.0073
0.624 ± 0.188
102
100
122
PSB-18422
3
3
4
5
Br
F
I
0.0321 ± 0.0159
PSB-18484
3
3
6
7
I
H
H
F
H
H
15
17
97.8
97.5
1.47 ± 0.21
92
4-fluorophenyl
Cl
3.68 ± 0.98
101
3
8
Cl
H
4-fluorophenyl
H
20
95.3
0.423 ± 0.048
92
8
ACS Paragon Plus Environment

Page 9 of 56
Journal of Medicinal Chemistry
f,h
39
1
2
3
4
5
6
7
8
9
H
H
hexyloxy
H
47
97.6
0.115 ± 0.027
62
PSB-1867
4
4
4
0
1
2
pentyloxy
Cl
H
H
H
Cl
H
H
H
27
34
32
97.2
96.7
98.0
≥ 10
–
98
–
pentyloxy
Cl
0.103 ± 0.039
> 10 (40%)
hexyloxy
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
43
Cl
H
hexyloxy
H
39
96.4
0.0670 ± 0.005
95
PSB-18183
4
4
octyloxy
H
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
H
H
H
F
31
29
79
79
49
43
71
73
74
57
45
67
64
50
67
17
25
96.3
97.1
98.0
98.1
99.0
99.2
98.1
98.3
98.0
99.2
99.0
95.3
95.2
95.0
98.3
96.4
96.7
> 10(25%)
0.117 ± 0.012
> 100 (3%)
0.731 ± 0.043
4.02 ± 1.03
4.13 ± 1.79
9.55 ± 2.76
3.01 ± 0.88
> 100
–
77
–
45
Cl
octyloxy
H
5
5
5
1
2
3
phenyl
H
phenyl
83
48
74
87
65
–
H
4-fluorophenyl
2-furanyl
2-thienyl
phenyl
54
H
5
5
5
5
6
7
H
H
H
4-fluorophenyl
2-furanyl
2-thienyl
phenyl
F
58
H
F
14.4 ± 3.4
57
–
5
6
6
9
0
2
H
F
> 100 (21%)
4.09 ± 2.13
> 30 (16%)
5.44 ± 0.94
> 30 (1%)
phenyl
H
84
–
for structure see above
for structure see above
for structure see above
for structure see above
63
425
519
–
6
–
6
> 30 (22%)
1.68 ± 0.23
–
6
6
for structure see above
41
a
b
Total isolated yield. Purity of compounds was determined using high performance liquid
2
6
chromatography–mass spectrometry (LC-MS) coupled to a UV detector as previously described.
c
Potency to induce calcium mobilization in 1321N1 astrocytoma cells stably transfected with the
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 56
d
human GPR17. Efficacy: the maximal effects were compared to the maximal signal induced by the
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
e
lead compound 1 (at 10 µM; EC50 0.331). Antagonistic activity was observed (IC50 8.40 ± 1.28
f
µM). Obtained only as a mixture of isomers which could not be separated using normal or reversed-
phase silica gel chromatography including preparative HPLC; the ratio of isomers could be
1
g
h
determined by H-NMR spectroscopy; ratio of 4-Cl,6-Br : 4-Br,6-Cl is ~ 1:1;
compound 39
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
contains 10% of the 4-hexyloxy isomer, see H NMR spectrum, Figure S1 (Supporting Information).
To introduce aromatic substituents, indole-diethyl ester derivatives bearing one or two bromo
1
8
substituent(s) (8, 46–49, Scheme 2) were treated with different arylboronic acids under microwave-
assisted Suzuki cross-coupling reaction conditions using bis-(triphenylphosphine)palladium(II)
dichloride as a catalyst in the presence of potassium phosphate in dioxane to afford compounds 50.
Scheme 2. Synthesis of Compounds 37, 38 and 51–60 via Microwave-Assisted Suzuki Cross-
a
Coupling Reaction of (Di-)Bromo-Substituted Indole Diethyl Ester Derivatives
10
ACS Paragon Plus Environment

Page 11 of 56
Journal of Medicinal Chemistry
a
Reagents and conditions: (i) Ar–B(OH)
2
(1.2–2.4 equiv.), K
3
PO
4
·H
2
O (1.5–3.0 equiv.),
O (3
1
2
3
4
5
6
7
8
9
Pd(II)Cl
2
(PPh
3
)
2
(3–6 mol%), dioxane, microwave, 150 °C, 60–90 min, 43–80%; (ii) LiOH·H
2
b
1
2
2
equiv.), THF:H O, rt, 24 h, 95–100%. For R and R see Table 1.
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
The unresolved isomeric mixture of carboxylic acid di-esters bearing Cl and Br atoms in the 4- and
-position (8a) was treated with 4-fluorophenylboronic acid under Suzuki cross-coupling conditions
6
(Scheme 2) to provide two separable regioisomers of the di-esters (50a and 50b). Finally, the di-
esters 50a–l were saponified to yield the desired arylindoles 37, 38, 51–60 bearing two free
carboxylic acid functions (Scheme 2, Table 1).
In order to investigate the role of the N1-H function present in the indole core we treated compound
8
a (Scheme 3), with methyl iodide or ethyl 3-iodopropanoate, respectively, in the presence of
potassium carbonate in refluxing acetonitrile furnishing 61a and 61b in 80-85% yield. Subsequent
saponification yielded 62 and 63, the N-substituted derivatives of lead compound 1, in 95-99% yield
(Scheme 3, Table 1).
a
Scheme 3. Synthesis of N-alkylated indole derivatives 62 and 63
a
3 2 2 2 2 3 3
Reagents and conditions: (i) CH I or BrCH CH CO Et (3 equiv.), K CO (1 equiv.), CH CN, 100
°
C, 18 h, 80-85%; (ii) LiOH·H O (3 equiv.), THF : H O (1:1 v/v), rt, 24 h, 95-99%.
2
2
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 56
In a next step we wanted to probe the consequences of an increasing length, and of reduced
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
flexibility of the alkyl side chain connected to the indole 3-position by preparing compounds 64 and
17
6
5 (Table 1). Reaction of 3,5-dichloroaniline with ethyl 2-oxocyclohexanecarboxylate provided 58
2
0
in 67% yield in analogy to the synthetic route A outlined in Scheme 1 and as previously described.
19
The synthesis of 65 as a precursor of tritiated lead structure 1 has been previously described.
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
To explore the role of the carboxylic acid function present in the 2-position of the indole moiety we
regioselectively decarboxylated compound 1 at high temperature under solvent-free conditions
yielding compound 66 in 25% yield (Scheme 4).
Scheme 4. Regioselective Decarboxylation of Lead Structure 1 to Obtain 3-(4,6-Dichloro-1H-indol-
a
3-yl)propanoic acid (66)
a
Reagents and conditions: (i) 275 °C, 10 min, 25%.
Pharmacological Evaluation. All final products were investigated for their potency to induce
calcium mobilization in 1321N1 astrocytoma cells stably transfected with the human GPR17 using
®
18
the calcium-chelating fluorescent dye Oregon Green as previously described. To determine
compound efficacy, the maximal effects were compared to the maximal signal induced by the lead
2
+
compound 1 (10 µM, EC50 0.331). For compounds inducing more than 50% stimulation of Ca
mobilization at a test concentration of 10 µM, full concentration-response curves were determined
and EC50 values were calculated (Table 1). Compounds that did not show agonistic activity were
further tested for their potency to block receptor activation by 1 (10 ꢀM) at a high test concentration
of 100 ꢀM. Binding affinities of test compounds at GPR17 receptors were determined by
12
ACS Paragon Plus Environment

Page 13 of 56
Journal of Medicinal Chemistry
3
competition experiments using the GPR17 agonist radioligand [³H]PSB-12150 (the H-labeled form
of lead structure 1) at membrane preparations of Chinese hamster ovary (CHO) cells stably
expressing GPR17.
1
2
3
4
5
6
7
8
9
Structure-activity relationships. Previously we reported initial SARs for a small set of indole
1
8
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
derivatives related to the lead structure 1 as agonists of GPR17. Different indoles with variations in
the 4- and 6-position were investigated. The indoles bearing identical substituents in both positions
showed the following rank order of potency: 4,6-dibromo > 4,6-dichloro > 4,6-diiodo > 4,6-
1
8
di(trifluoromethoxy) >> 4,6-dimethoxy. These results indicated that the size of the substituents was
important for interaction with the receptor. We had also discovered that a bulky halogen atom in the
6
-position was very important for high potency, while introducing a large halogen atom in the 4-
position dramatically reduced activity. In the present study 53 indole derivatives were synthesized,
5 of which are new compounds not previously described in literature. All synthesized compounds
3
and, in addition, four commercially available indole derivatives (see Table 3) were evaluated as
GPR17 receptor ligands with the goal to broadly investigate the SARs of lead structure 1 and to
improve its potency on GPR17.
Steep SARs were observed. Introducing any substituent (e.g. Cl, F, OMe, or isopropyl) in the 5-
position of the indole moiety abolished activity (10, 12, 14, 15, 22, 23; Table 1). A fused aromatic
ring, benzo[g]indole (24), connected to the 6,7-position of indole, was also not tolerated. When the
chlorine atom in the 6-position of 1 was removed the potency was virtually abolished (9, 18, Table
1
), whereas removal of the chlorine atom in the 4-position led to an about 11-fold reduction in
potency compared to 1 (19, EC50 3.72 µM, p < 0.0001, ****). This indicates that a substituent in the
-position is essential, while that in the 4-position appears to be of less importance. In the 6-position
6
a lipophilic substituent was preferred while in the 4-position a small substituent such as Cl, F or H
was best. 4,6-Difluoro- and 4,6-dimethyl-substituted indoles 16 (EC50 7.41 µM) and 17 (EC50 15.1
µM) displayed low activity, again indicating that a larger lipophilic and electronegative substituent is
required in the 6-position (Table 1). Any mono-substituted derivative in the 4- or 7-position (11, 20,
27, Table 1), or derivatives with a combination of substituents in both positions (21) were inactive. A
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 56
small halogen atom (fluoro) was tolerated in the 7-position while a larger one (chloro) was not.
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
Mono-substitution in the 6-position, especially with a large and lipophilic substituent, was well
tolerated (19, EC50 3.72 µM; 25, EC50 4.98 µM; 26, EC50 0.270 µM; Table 1), while the 6-fluoro and
6
-methoxy-substituted derivatives did not show activity (13, 28, Table 1). Several 4,6-disubstituted
indoles with different substituents were synthesized and evaluated for GPR17 activation (32–38 and
0–45, Table 1). The best combination was 4-fluoro-6-bromo- substitution in 33 (EC50 0.0279 µM)
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
4
showing a 12–fold increase compared to lead structure 1 (p = 0.0255, *). The 4-fluoro-6-iodo
substituted derivative 35 was similarly potent (EC50 0.0321 µM). The corresponding regioisomers 34
and 36 were 22–fold (compare 33/34) and 45–fold (compare 35/36) less potent indicating again that
large residues are beneficial in the 6-position but not in the 4-position of the indole core.
In the next step, even larger lipophilic substituents (benzyl, phenoxy, pentyloxy, hexyloxy, and
octyloxy) were introduced into the 6-position of the indole core. We found that an ether linkage
(phenoxy, 26, EC50 0.270 µM) was preferable over a methylene linker (benzyl, 25, EC50 4.98 µM) by
>18–fold (p = 0.0126, *). An aliphatic ether-linked residue (hexyloxy, 39) was found to result in a
potent compound with an EC50 value of 0.115 µM. Next, we designed and synthesized compounds
bearing a chlorine atom in the 4-position and a long aliphatic chain consisting of five, six or eight
carbon atoms connected by an ether linkage to the 6-position (41, 43 and 45).. Replacing the chlorine
group in the 4-position by a large lipophilic group ( pentyloxy (40), hexyloxy (42), octyloxy (44))
was not tolerated, while replacing the chlorine atom in the 6-position by pentyloxy (41), hexyloxy
(43) or octyloxy (45) led to an increase in potency in the following rank order (hexyloxy > pentyloxy
> octyloxy), leading to one of the most potent GPR17 agonists (43, Table 1) with an increase in
potency by ca. 5–fold compared to lead compound 1 (EC50 0.0670 µM, p < 0.0001, ***).
All of the investigated compounds that were inactive as agonists at GPR17 showed no antagonistic
activity, except for compound 28 (6-OMe), which blocked GPR17 activation induced by the agonist
1
with an IC value of 8.40 µM.
50
Finally, the aryl-substituted indole derivatives prepared by microwave-assisted Suzuki cross-
14
ACS Paragon Plus Environment

Page 15 of 56
Journal of Medicinal Chemistry
coupling reaction were evaluated and it was found that a phenyl substituent in the 6-position was
preferable over a phenyl residue in the 4-position or the 4,6-di-phenyl-substituted derivatives (51, 52
and 60, Table 1). An unsubstituted phenyl ring in the 6-position (52, EC50 0.731 µM) was slightly
better than the 4-fluorophenyl (53, EC50 4.02 µM, p = 0.0332, *) and the thienyl (55, EC50 9.55 µM,
p = 0.0331, *) derivative. Adding an extra fluorine atom in the 7-position of compounds 52–55
dramatically reduced potency (56–59, Table 1). Introducing a chlorine atom in the 4-position in
combination with a 4-fluorophenyl residue in the 6-position was tolerated (38, EC50 0.423 µM),
while switching of the two substituents (6-chloro-4-fluorophenyl, 37) reduced potency by >8–fold
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
(EC50 3.68 µM, p = 0.0294, *). Methylation of the indole nitrogen furnishing 62, increasing the
length of the alkyl chain on the 3-position by one methylene unit (64), or introducing a double bond
into the side chain (65) all led to a significant decrease in potency compared to the lead compound 1
(Table 1). Introducing a propionic acid moiety at the indole N1 position (63) or decarboxylation (in
66) also resulted in reduced potency compared to lead compound 1 (63, EC50 5.44 µM, p = 0.0056,
**; 66, EC50 1.68 µM, p = 0.0054, **).
Figure 2 summarizes the observed structure-activity relationships of the synthesized indole
derivatives. A lipophilic residue in the 4- and 6-positions was found to be important for the activity,
while introducing any residue in the 5- or 7-position dramatically reduced or abolished potency. We
also noticed that larger lipophilic substituents are preferred in the 6-position with the following rank
order of potency: hexyloxy (EC50 0.067 µM) ≥ phenoxy (EC50 0.270 µM) > I (EC50 0.715 µM; p =
0.0443, *) ≈ phenyl (EC50 0.731 µM) ≥ Br (EC50 2.24 µM) ≈ Cl (EC50 3.72 µM) ≈ benzyl (EC50 4.98
µM) >> OMe, H (EC > 100 µM), while in the other positions (4, 5, or 7) (large) substituents led to
5
0
a decrease or loss of potency. The di-ester compounds 8 showed no potency for GPR17 (data not
shown) indicating the requirement of the acidic functions. The removal of the carboxylic acid group
in the 2-position moderately reduced potency. The indole NH function appears to play a crucial role
in GPR17 binding, it may form a hydrogen bond with the receptor protein or with a water molecule.
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 56
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
A. Indoles with small 6-substituent
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
B. Indoles with large 6-substituent
Figure 2. Summary of the structure-activity relationships of the synthesized indole derivatives as
GPR17 agonists. A. Indoles with small 6-substituent. B. Indoles with large 6-substituent.
Small halogen atoms in the 4-position (e.g. F, Cl) in combination with larger halogen atoms (Br, I) in
the 6-position increased the activity (see Figure 2A), a 6-hexyloxy chain (in 43) was also beneficial
(
Figure 2B). Concentration-response curves for selected potent indole derivatives (1, 33, 35, 39, and
4
3) are provided in Figure 3.
16
ACS Paragon Plus Environment

Page 17 of 56
Journal of Medicinal Chemistry
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 3. Concentration-response curves of selected indole derivatives determined in calcium
mobilization assays at 1321N1 cells recombinantly expressing the human GPR17.
Efficacy. The maximal effect of a GPCR agonist determined in a G protein-dependent assay may be
2
5
dependent on the receptor expression level or the so-called “receptor reserve”. In the employed
1321N1 astrocytoma cell line recombinantly expressing the human GPR17 most of the investigated
indole derivatives behaved as full or nearly full agonists (70-100 % efficacy) with efficacies in the
same range as lead compound 1 (set at 100 %). However, there were a few notable exceptions. The
6-methoxyindole derivative 28 appeared to display very low efficacy (35% stimulation at a high
concentration of 100 µM) and actually blocked receptor activation induced by the full agonist 1 with
an IC50 value of 8.40 µM. The analog of 1 that lacked the 2-carboxylate function (66) also possessed
low efficacy (41%). This indicates that the 2-carboxylate may be important for inducing or
stabilizing the fully active conformation for G protein coupling of the receptor. Also, indoles with a
large 6-substituent combined with a fluorine atom in the 7-position appeared to be less efficacious
than their analogs lacking the 7-fluoro substitution (compare 56 (65% efficacy) / 52 (83%); 58 (57%)
/
54 (74%)). Interestingly, this effect was not observed for the 6-Cl- and 6-Br-substituted indoles 30
98%) and 31 (87%) bearing a 7-fluorine atom. A few indole derivatives substituted with a large,
(
lipophilic 6-substituent but lacking a substituent in the 4-position also showed decreased efficacy,
e.g. the 6-hexyloxy derivative 39 (62%) and the p-fluorophenyl derivative 53 (48%). It appears that
the introduction of a chlorine atom in the 4-position canrestore the efficacy (see 4-chloro-6-
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 56
pentyloxyindole derivative 41, 98%, 4-chloro-6-hexyloxyindole derivative 43, 95%, 4-chloro-6-
octyloxyindole derivative 45, 77%). These results show the interdependence of effects on different
substitution patterns possibly indicating different binding modes.
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
Selectivity. In Table 2 GPR17 agonistic activities of selected compounds are compared with their
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
2
0,21
NMDA antagonistic activities previously published by Salituro and coworkers.
The SARs show
that small halogen atoms, e.g., Cl, in the 6-position of the indole ring enhanced binding affinity and
selectivity towards the glycine binding site over the glutamate binding site of the NMDA receptor.
Especially lead compound 1 of the present study was potent and selective for the glycine binding
site. Compound 1 showed similar potency in activating GPR17 as in blocking the NMDA receptor.
However a bulkier group in the 6-position (e.g. phenoxy, 26) abolished interaction with the NMDA
receptors and led to selectivity for GPR17 (EC50 0.270 µM; NMDA receptor, IC50 >100 µM, Table
2). These results had in fact encouraged us to design indole derivatives as GPR17 agonists focusing
on bulkier substituents in the 6-position, in order to develop potent and selective GPR17 agonists.
18
ACS Paragon Plus Environment

Page 19 of 56
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Comparison of potencies of selected indole derivatives at GPR17 and the NMDA
2
0
antagonistic activities
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
GPR17
NMDA receptor
IC50 (µM)
2+
Ca mobilization
4
5
6
7
a
Compd
R
R
R
R
EC50 (µM)
(
or % activation at indicated
concentration
3
b
3
[ H]CPP
c
[
H]Gly
1
9
Cl
H
H
F
H
H
Cl
H
H
H
H
H
0.33118
> 30 (8)
> 30 (11)
7.41
0.17
27
358
34
H
1
0
Cl
H
H
H
H
H
25
35
16
F
4
n.a.
283
> 1000
1
2
2
9
0
6
H
H
H
Cl
3.72
2
H
Cl
> 30 (5)
> 100
> 1000
d
phenoxy
H
0.270
n.a.
a
b
3
Human receptor, for details see Table 1. Competition assay vs [ H]glycine at rat cortical and hippocampal membrane
18
c
3
binding sites.
Competition assay against [ H]CCP ((3-(2-carboxypiperazin-4-y1)propyl-l-phosphonic acid) at rat
20 d
20
cortical and hippocampal membrane glutamate binding sites. n.a., not available.
In addition, we investigated the most potent GPR17 agonists (1, 33, 35, 39 and 43) as potential
agonists of P2Y receptor subtypes (P2Y , P2Y , P2Y , P2Y , P2Y12, P2Y14). The compounds were
found to be inactive or only very weakly active as agonists. The highest potency was observed for 43
at the P2Y receptor (EC50 ≥ 50 µM, > 700-fold selectivity for GPR17) (see Table S1, Supporting
Information). The compounds also were inactive of only weakly active in inhibiting selected P2Y
1
2
4
6
2
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 56
1 2 4 6
receptor subtypes (P2Y , P2Y , P2Y , P2Y ).
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
Next we evaluated the known drugs indomethacin (67) and rizatriptan (68), the amino acid
tryptophan (69) and the serotonin metabolite 70. All of these indole derivatives were inactive as
agonists at GPR17 (Table 3). However, indomethacin blocked GPR17 activation at high
concentrations displaying an IC50 value of 28.8 µM.
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
Table 3. Evaluation of Selected Established Drugs and Physiological Compounds with Indole
Structure
Human GPR17
EC50 ± SEM
Human GPR17
IC50 ± SEM
Compd Name Function
Chemical Structure
or % activation at or % inhibition at
a
a
3
0 µM
30 µM
Indomethacin
67
> 30 (0%)
28.8 ± 1.1
non-steroidal anti-
inflammatory drug
Rizatriptan
6
8
> 30 (0%)
> 30 (27%)
1
5-HT agonist
Tryptophan
6
9
> 30 (0%)
> 30 (1%)
essential amino acid
5-Hydroxyindole-3-
acetic acid
7
0
> 30 (1%)
> 30 (7%)
main metabolite of
2
0
ACS Paragon Plus Environment

Page 21 of 56
Journal of Medicinal Chemistry
serotonin
1
2
3
4
5
6
7
8
9
a
Potency to induce or inhibit calcium mobilization in 1321N1 astrocytoma cells stably transfected
with the human GPR17.
A selection of the most potent indole derivatives (1, 33/34 (1:1 mixture), 39 and 43) was further
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
investigated in calcium mobilization assays at the rat and mouse GPR17 orthologues (Figure 4, Table
4
). The potency of all compounds was very similar at the rat as compared to the mouse receptor.
However, all four compounds were more potent at the human GPR17. The compounds (1, 32)
bearing a small 6-substituent (a halogen atom) were only slightly (2- to 6-fold) more potent at human
as compared to rodent GPR17, and the difference was not statistically significant. However the
compounds bearing an alkoxy residue in position 6 (39, 43) showed large differences being 34-48-
fold more potent at the human as compared to rat or mouse GPR17. This indicates that the large 6-
substituent is mainly responsible for the significant species differences (human versus rodent).
2+
Figure 4. Concentration-response curves of selected indole derivatives in Ca mobilization assays
performed with HEK-rat-GPR17 (A) and HEK-mouse-GPR17 cells (B). Data points represent means
±
SEM from 3 independent experiments. For EC values see Table 4.
50
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Table 4. Comparison of GPR17 agonistic potencies of selected indole derivatives by calcium
mobilization assay in different species
Page 22 of 56
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
Compd
Chemical Structure
Human GPR17
Rat GPR17
EC50 ± SEM (µM)
(Efficacy)
Mouse GPR17
EC50 ± SEM (µM)
(Efficacy)
a
b
c
EC50 ± SEM (µM)
(
Efficacy)
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
0.331 ± 0.087
0.705 ± 0.041
1.11 ± 0.18
1
(
100 %)
(100 %)
(100 %)
0
.134 ± 0.023
0.470 ± 0.060
0.753 ± 0.194
d
33/34
(
82 %)
(95 %)
(106 %)
0
.115 ± 0.027
5.31 ± 0.83
5.51 ± 2.19
e
3
9
(
62 %)
(53 %)
(28 %)
0
.067 ± 0.005
2.30 ± 0.22
3.03 ± 0.17
43
(
95 %)
(77 %)
(59 %)
a
Potency to induce calcium mobilization in 1321N1 astrocytoma cells stably transfected with the
b
human GPR17. Potency to induce calcium mobilization in HEK cells transfected with the rat
c
GPR17. Potency to induce calcium mobilization in HEK cells transfected with the mouse GPR17.
d
e
Ratio of 4-F,6-Br : 4-Br,6-F is ~ 1:1; Ratio of 4-H,6-hexyloxy : 4-hexyloxy,6-H is ~ 10:1
Homology model of the human GPR17 receptor. As a next step, a homology model of the human
GPR17 was generated based on the crystal structure of the human P2Y12 receptor (P2Y12R) both of
28
which belong to the δ-family of class A GPCRs. The sequence alignment between the human
22
ACS Paragon Plus Environment

Page 23 of 56
Journal of Medicinal Chemistry
GPR17 and the human P2Y12R is shown in Figure S18 of Supporting Information. The overall
sequence identity and similarity between the human GPR17 and P2Y12 receptor is 27.6 % and 42.9
1
2
3
4
5
6
7
8
9
%, respectively. The human P2Y12R had previously been crystallized in complex with the antagonist
ethyl 6-(4-{[(benzylsulfonyl)amino]carbonyl}piperidin-1-yl)-5-cyano-2-methylnicotinate (AZD1283,
PDB ID: 4NTJ) and the agonists 2-methylthio-ADP (2MeS-ADP, PDB ID: 4PXZ) and 2-methylthio-
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
2
8,29
ATP (2MeS-ATP, PDB ID: 4PY0).
1
Meanwhile, the crystal structure of the human P2Y R was
6
published in complex with the antagonists 2-iodo-N -methyl-(N)-methanocarba-2′-deoxyadenosine-
3′,5′-bisphosphate (MRS2500, PDB ID: 4XNW) and 1-(2-(2-(tert-butyl)phenoxy)pyridin-3-yl)-3-(4-
(
1
trifluoromethoxy)phenyl)urea (BPTU, PDB ID: 4XNV). The structures of the P2Y receptor are
both in an inactive state, and only in complex with ligands binding to two distinct, presumably
allosteric binding sites. Previous attempts by Eberini et al. to generate a homology model for the
1
human GPR17 had been based on the crystal structures of the human β -adrenergic, the human
adenosine A , the turkey and the human β -adrenergic receptors due to lacking of an X-ray structure
2A
2
3
0,31
of a closely related GPCR at that time.
Since the P2Y12R belongs to the same subgroup of the δ-branch class A GPCRs as GPR17, the high
resolution (2.5 Å) crystal structure of the human P2Y receptor in an active state in complex with an
1
2
agonist currently represents the most suitable template for generating a homology model of the
human GPR17 in its active conformation. After the alignment of the human P2Y12R and GPR17
3
2,33
sequences, 500 models were generated through the homology modelling tool MODELLER.
From the 500 generated models, the presented homology model (Figure 5) was selected on the basis
of the Discrete Optimized Protein Energy (DOPE) score and visual inspection. The homology model
of the human GPR17 shows two disulfide bonds, the first one between Cys104 and Cys181 that is
conserved among most members of the class A GPCR family and connects extracellular loop 2
(
ECL2) and transmembrane 3 (TM3), and a second disulfide bridge between Cys23 and Cys269
which connects ECL3 and the N-terminus of GPR17.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 56
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
P2Y12
GPR17
Superimpositon
A)
B)
C)
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
7
1
Figure 5. Homology model of the human GPR17 receptor. (A) The X-ray structure of the human
P2Y R (PDB ID: 4PXZ), which was used as a template for generating the homology model of the
1
2
human GPR17. The P2Y R agonist 2MeS-ADP (71) binds to the orthosteric binding site. (B) The
1
2
homology model of the human GPR17 with a schematic representation of the putative orthosteric
binding site deduced from the P2Y12R crystal structure. (C) The superimposed structures of the
human P2Y R and GPR17. The receptors are represented as cartoon models, the co-crystallized
1
2
atoms of the ligand are depicted as spheres.
3
A plausible binding site with a volume of 265 Å was identified in GPR17 using the SiteFinder
3
4
module from Molecular Operating Environment (MOE 2014.09 ). This binding site is comparable
to the orthosteric binding site in the crystal structure of the human P2Y12 receptor in complex with
28
the agonist 2MeS-ADP (PDB ID: 4PXZ). The important amino acids in the binding site of the
human P2Y12R and the human GPR17 homology model are shown in Figure 6.
24
ACS Paragon Plus Environment

Page 25 of 56
Journal of Medicinal Chemistry
A)
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
B)
Figure 6. Comparison of (A) the orthosteric binding site of the human P2Y12R (PDB ID: 4PXZ), and
B) the putative binding site of the human GPR17. The important amino acids are shown. The
(
conserved amino acids between the human P2Y12R and GPR17 are highlighted by red boxes, and the
positively charged amino acids are encircled in blue. The amino acids and the cocrystallized human
P2Y12 agonist 2MeS-ADP are represented as stick models. The oxygen atoms are colored in red,
nitrogen atoms in blue, sulfur atoms in yellow, chlorine in green.
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 56
A considerable similarity of the amino acids in the orthosteric binding site of the human P2Y12R and
the proposed binding site of the human GPR17 supports the selection of the human P2Y12R as a
template for generating a homology model of the human GPR17. However, the number of positively
charged amino acid residues in the binding site of the P2Y12R is higher (9 residues) in comparison to
the putative binding site of the human GPR17 (only 5 residues). Among the five positively charged
amino acid residues in the proposed binding pocket of GPR17, one amino acid (Arg273) is exposed
towards the solvent. Its lower number of positively charged amino acids may explain why
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
1
6,19
nucleotides such as ADP or ATP are not binding to the human GPR17 with high affinity.
Molecular docking studies of the agonists. To propose a binding mode for the new indole
derivatives that were developed in the present study as agonists for the human GPR17, and to
rationalize the determined SARs we selected lead compound 1 and the most potent agonist 43. The
docking and glide scores from the induced-fit molecular docking experiments, and the SARs from a
1
8
previously published study and from this study were taken into account.
As illustrated in Figure 7, lead compound 1 located in the putative binding site of the receptor, binds
primarily through strong electrostatic and hydrogen bonding interactions. According to the putative
binding pose of 1 obtained from our docking studies, the two carboxylic acid residues form strong
electrostatic interactions with Arg87 and Arg255, and a hydrogen bond interaction with Asn279.
Tyr112 (TM3) located below the docked pose of compound 1 likely contributes to the stabilization
of the indole moiety in the putative binding pocket through aromatic π-π-interactions, and Tyr258
(TM6) positioned above agonist 1 might contribute interactions through its aromatic ring after
binding of 1 inside the binding pocket. The Cl-substitution at position 4 of the indole moiety is
directed towards a small hydrophobic subpocket formed by the cysteine disulfide residues Cys104-
Cys181 and the residues Gly108, Phe109, Tyr112 and Leu182 in our model. The halogen atoms F,
Cl or Br at position 4 of the indole agonists possibly form electrostatic interactions with the
backbone of Cys104, Gly108, Phe109, Tyr112, Cys181 or Leu182.
26
ACS Paragon Plus Environment

Page 27 of 56
Journal of Medicinal Chemistry
A)
B)
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. Proposed binding mode of 1 in the human GPR17 homology model (based on the human
P2Y12R crystal structure, PDB ID: 4PXZ). (A) Docked pose of 1 with the important residues in the
putative binding pocket of the human GPR17 receptor. (B) The residues important for the interaction
are depicted in gray. The human GPR17 model (light brown) is displayed in cartoon representation,
and the amino acid residues (gray) and compound 1 (purple) are shown as stick models. For color
code see Figure 6; interactions are shown as red dotted lines.
The strong enhancement of activity by the chlorine atom in position 6 of the indole core may be
explained by electrostatic interactions with Thr86 (TM2) or Thr107 (TM3) located at a distance of
~4 Å (Figure 7 A).
The putative binding pose and 2D interaction diagram of 43 are shown in Figure 8. The binding pose
shows that 43 likely has a different orientation in the orthosteric binding pocket than lead compound
1
.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 56
A)
B)
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 8. Proposed binding mode of 43 in the human GPR17 homology model (based on the human
P2Y12R crystal structure, PDB ID: 4PXZ). (A) Docked pose of 43 with the important residues in the
putative binding pocket of the human GPR17. (B) The residues important for the interaction are
depicted in gray. The human GPR17 model (light brown) is displayed in cartoon representation, and
the amino acid residues (gray) and compound 43 (marine blue) are shown as stick models. For color
code also see Figure 6 and 7.
The summary of the SARs (see Figure 2) shows the significance of having both carboxylic acid
functions and a free indole-NH group. The terminal carboxylic acid of the propanoic acid residue
attached to the indole 3-position might interact with Arg87, Arg255 and Asn279 leading to
electrostatic and hydrogen bond interactions. On the other hand, the carboxylic acid group in the
indole 2-position probably binds below the extracellular loop 2 (ECL2), and due to the length (25
residues) and high flexibility of the ECL2 it is difficult to predict which amino acid residue of ECL2
it might bind to. The NH group of the indole moiety appears to be exposed towards the solvent and
probably forms hydrogen bond interactions with water molecules. The Cl-substitution in the 4-
position might form electrostatic interactions with Arg255 located at a distance of ~3.5 A, and due to
the high flexibility of the arginine this interaction could easily be achieved. Tyr258 is located above
the residues Arg255 and Asn279 which are likely forming important electrostatic and hydrogen bond
28
ACS Paragon Plus Environment

Page 29 of 56
Journal of Medicinal Chemistry
interactions with 43, and Tyr258 may play an important role in maintaining the conformation of
these two residues and keeping 43 in place. The hexyloxy substitution, a large lipophilic substituent
at position 6, likely introduces tight hydrophobic interactions with the residues formed by Phe109,
Tyr112, Leu113, Tyr116, Val159, Ala162, Met163, Leu166, Leu194, Val195 and Leu197 from TM3,
TM4 and TM5. Thus, lead structure 1 and one of the most potent agonists of the present series,
compound 43, likely occupy the putative binding pocket in two different regions and orientations.
Compound 1 likely binds deeper into the binding pocket, while 43 interacts with the hydrophobic
residues of TM4 and TM5. A comparison of the binding pose of 2MeS-ADP in the crystal structure
the human P2Y R with the docked pose of 43 obtained for the human GPR17 is shown in Figure
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
1
2
S23 (see Supporting Information). In order to verify that the agonists 1 and 43 form a stable complex
with the human GPR17, molecular dynamic simulations for a time scale of 100 ns were performed.
As shown in Figure S24 (see Supporting Information), for the receptor complexes the RMSD values
α
of the C atoms rapidly reached the equilibrium phase approximately within 1.0–1.5 Å deviations
from the starting structure of the complex. This suggests that the human GPR17 maintains a similar
interaction profile upon binding of 1 and 43. The trajectories from the simulations also show that the
agonists 1 and 43 reside within the putative binding site throughout the simulation of 100 ns (see
Supplementary Material, Video V1 and V2).
Comparison of mouse, rat and human GPR17. To search for a possible explanation for the
observed species differences of the investigated indole derivatives (see Table 4), the amino acid
sequences of the mouse, rat and human GPR17 were aligned (see Supplementary Figure S4). The
human GPR17 displays high sequence identity with both the mouse and the rat GPR17 (89.7%).
However, the experimental results of selected indole derivatives tested in different species (Table 4)
showed that there are significant species differences for compounds 39 and 43, which bear large
lipophilic alkyl chains in position 6. Upon analyzing the sequences of rat and mouse GPR17 in
comparison to the human receptor it was observed that the only variable amino acid in the identified
binding site is Val195 in the human receptor, which is replaced by Ala195 in the mouse and rat
GPR17 (see Supplementary Figure S5). The experimental results obtained for 39 in different species
29
ACS Paragon Plus Environment