Structure-Based Optimization of Coumarin hA3 Adenosine Receptor Antagonists

Article abstract of DOI:10.1021/acs.jmedchem.9b01572

Adenosine receptors participate in many physiological functions. Molecules that may selectively interact with one of the receptors are favorable multifunctional chemical entities to treat or decelerate the evolution of different diseases. 3-Arylcoumarins

Full text of DOI:10.1021/acs.jmedchem.9b01572

Subscriber access provided by Universiteit Leiden / LUMC
Article
Structure-based optimization of coumarin
hA3 adenosine receptor antagonists
Maria João Matos, Santiago Vilar, Saleta Vazquez-Rodriguez, Sonja Kachler, Karl-Norbert Klotz,
Michela Buccioni, Giovanna Delogu, Lourdes Santana, Eugenio Uriarte, and Fernanda Borges
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01572 • Publication Date (Web): 18 Nov 2019
Downloaded from pubs.acs.org on November 18, 2019
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 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
Structure-based Optimization of Coumarin hA3 Adenosine Receptor
Antagonists
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
Maria João Matos,a,b,* Santiago Vilar,b Saleta Vazquez-Rodriguez,b Sonja Kachler,c Karl-
Norbert Klotz,c Michela Buccioni,d Giovanna Delogu,e Lourdes Santana,b Eugenio
Uriarte,b,f Fernanda Borgesa,*
a CIQUP/Department of Chemistry and Biochemistry, Faculty of Sciences, University of
Porto, 4169-007 Porto, Portugal
b
Department of Organic Chemistry, Faculty of Pharmacy, University of Santiago de
Compostela, 15782 Santiago de Compostela, Spain
c Institute of Pharmacology and Toxicology, University of Würzburg, 97078 Würzburg,
Germany
d
School of Pharmacy, Medicinal Chemistry Unit, University of Camerino, 62032
Camerino, Italy
e Department of Life Sciences and Environment - Section of Pharmaceutical Sciences,
University of Cagliari, 09124 Cagliari, Italy
f Instituto de Ciencias Químicas Aplicadas, Universidad Autónoma de Chile, 7500912
Santiago, Chile
Abstract. Adenosine receptors participate in many physiological functions. Molecules
that may selectively interact with one of the receptors are favorable multifunctional
chemical entities to treat or decelerate the evolution of different diseases. 3-
Arylcoumarins have already been studied as neuroprotective agents by our group. Here,
differently 8-substituted 3-arylcoumarins are complementarily studied as ligands of
adenosine receptors, performing radioligand binding assays. Among the synthesized
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 38
1
2
3
4
5
6
7
8
9
compounds, selective A3 receptor antagonists were found. 3-(4-Bromophenyl)-8-
hydroxycoumarin (compound 4) displayed the highest potency and selectivity as A3
receptor antagonist (Ki = 258 nM). An analysis of its X-ray diffraction provided detailed
information on its structure. Further evaluation of a selected series of compounds
indicated that it is the nature and position of the substituents that determine their activity
and selectivity. Theoretical modeling calculations corroborate and explain the
experimental data, suggesting this novel scaffold can be involved in the generation of
candidates as multitarget drugs.
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
Keywords: 3-Arylcoumarins • Perkin reaction • Perkin-Oglialoro reaction • Adenosine
antagonists • Molecular modeling
INTRODUCTION
Adenosine acts as a common control of several cellular functions.1 This purine regulates
fundamental pathophysiological activities through four receptor subtypes (A1, A2A, A2B
and A3).2 Adenosine is the endogenous, nonselective agonist which remains shortly in the
human body, while inosine –its metabolite– being metabolised by adenosine deaminase,
weakly interacts with the A3 receptor.3
Adenosine receptors are GPCR (G protein-coupled receptors) that are being described as
potential therapeutic targets for a wide range of pathologies.4 From cancer to brain
diseases, these receptors are involved in immunological and inflammatory conditions.5,6
There is a huge variety of chemicals designed as adenosine receptor ligands, both directly
acting as agonists, antagonists, and indirect modulators.7 Regadenoson (CVT-3146) was
proposed as selective A2A agonist. This molecule has been playing an interesting role
inducing stress for imaging techniques in cardiology.8 Istradefylline (KW-6002) was
approved in Japan for Parkinson's disease. This xanthine derivate proved to be a strong
A2A antagonist.,9
2
ACS Paragon Plus Environment

Page 3 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The increasing knowledge on the structure and mechanisms of A3 receptor has given
fundamental pieces of evidence to validate it as a promising therapeutic target.10 This
enables a rational design on robust A3 selective antagonists as therapeutic solutions for
several diseases.11 Recent reports manifest an important role for A3 receptor in mediating
adenosine function at the central nervous system.12,13 A3 receptors are also described as
inducers of strong anti-inflammatory activity in animal models.14 These particular effects
aroused our attention, since our group has been active in the research of age-related
pathologies, in particular vascular, inflammatory and neurodegenerative diseases.15
Based on the background of our research group regarding differently substituted
coumarins as potential adenosine receptor ligands,16,17,18,19,20,21 and in the potential of
some of these compounds as neuroprotectors and enzymatic inhibitors related to
neurodegenerative disorders (i.e. acetylcholinesterase, butyrylcholinesterase and
monoamine oxidase B), in this work we report a group of coumarins bearing a wide
variety of substituents as modulators of adenosine receptors. In particular, the revealing
achievements of our last work on the interesting activity of 8-substituted 3-arylcoumarins
(Figure 1) have been the inspiration for the progression of this study.21 Design, synthesis,
pharmacological evaluation, docking calculations, and structure-activity relationship
exploration of a family of 8-substituted 3-arylcoumarins, were performed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 38
1
2
3
4
5
6
7
8
9
Figure 1. Previous results from the group on the potential of 8-substituted 3-
arylcoumarins as adenosine ligands at adenosine 7TM receptors. A. Three 8-(2-
oxopropoxy)-3-arylcoumarins with different affinity profiles. B. Hypothetical binding
mode for 8-(2-oxopropoxy)-3-(p-tolyl)coumarin in the hA3 protein pocket (template for
the hA3 homology model: 3EML). Color code: pale orange – micromolar range; red –
nanomolar range.
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
RESULTS AND DISCUSSION
Chemistry. Molecules 1-27 were preapred based on the synthetic strategies described in
Figure 2A. In general, these compounds were obtained by a classic Perkin (compounds
1-3, 5-14, 16 and 17) and Perkin-Oglialoro (compounds 18-22) reactions. Further
hydrolysis of the ethoxy (compounds 4 and 15) and acetoxy derivatives (compounds 23-
27) allowed the obtention of the hydroxyl derivatives.22,23,24,25
Ortho-hydroxybenzaldehydes reacted via Perkin condensation with the corresponding
arylacetic acids and N,N’-dicyclohexylcarbodiimide (DCC) acting as dehydrating agent,
to obtain compounds 1-3, 5-14, 16 and 17. Afterwards, the ethoxy derivatives 3 and 14
gave the respective compounds 4 and 15, by acidic hydrolysis in the presence of hydriodic
acid (HI) 57%, acetic acid (AcOH) and acetic anhydride (Ac2O).
Acetoxy-3-arylcoumarins 18-22 were prepared via Perkin-Oglialoro condensation.
Following its traditional methodology, ortho-hydroxybenzaldehydes and arylacetic acids
reacted for 16 h, at reflux temperature, in the presence of potassium acetate (CH3CO2K)
and acetic anhydride (Ac2O). Under these mild conditions, concomitant acetylation of the
hydroxyls and closure of the pyrone ring occur. The hydrolysis of the acetoxy compounds
was then carried out using aqueous hydrochloric acid (HCl) and methanol (MeOH), at
4
ACS Paragon Plus Environment

Page 5 of 38
Journal of Medicinal Chemistry
1
2
3
4
reflux temperature, for 3 h. This reaction allowed obtaining hydroxyl substituted 3-
5
6
arylcoumarins 23-27.
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. A. Scheme of the synthetic methodologies. a) DCC, DMSO, 110 ºC, 24 h. b)
HI, AcOH, Ac2O, reflux, 3 h. c) CH3CO2K, Ac2O, reflux, 16 h; d) HCl, MeOH, reflux, 3
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 38
1
2
3
4
h. B. Molecular structure of compound 4, showing the atom-numbering scheme used in
5
6
the X-ray study.
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
The structural analyses of compound 4 by X-ray crystallography (CCDC number
1937910)26 corroborated the NMR information. This molecule is a coumarin derivative
with a p-bromophenyl at position 3 of the coumarin and a hydroxyl at position 8, as
represented in the scheme (Figure 2B). The dihedral angle formed by both planes of the
3-arylcoumarin scaffold (~36.62.1) may reveal some important information on the
possible binding pose of compound 4 to the receptor, being typical of these family of
molecules.27 In addition, the C3—C13 bond length is typical from the 3-arylcoumarins
1.483 Å. This length is coincident with the one reported for the 3-phenylcoumarin,
crystalized by our research group.27 Also, coumarin’s planarity is corroborated by the
torsion angles between its carbon atoms. Besides the interest in organic chemistry and
structural elucidation, the deep structural knowledge of a molecule can be a significant
tool for the understanding of its potential affinity and/or selectivity for particular receptors.
All the details of the synthetic methodologies and characterization of the studied
molecules are presented in the experimental section of this manuscript.
Pharmacological study. The affinity of the newly described coumarins for the A1, A2A,
and A3 adenosine receptors was evaluated using a protocol based on radioligand binding.
A functional assay (inhibition of agonist-stimulated adenylyl cyclase activity) was
performed to assess the affinity for the A2B receptor.28,29 All the details are included in
the experimental section. A1, A2A and A3 adenosine receptors binding data is represented
in Table 1. All the studied molecules became inactive against the A2B receptor (IC50 > 30
µM, data not shown).
6
ACS Paragon Plus Environment

Page 7 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
Table 1. Ki values for the binding affinity of compounds 1-27 and reference compounds
7
8
for human A1, A2A and A3 receptors expressed in Chinese hamster ovary (CHO) cells.
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
Compound
1
R
R’
-
R1
H
R2
H
R3
H
R4
H
hA1 (µM)
hA2A (µM)
hA3 (µM)
22.1
55.0
22.5
OCH2CH3
(18.0-27.1) (36.4-83.1) (14.0-36.3)
3.93
2
3
OCH2CH3
OCH2CH3
-
-
H
H
H
H
CH3
Br
H
H
> 100
> 60
(2.57-6.01)
> 100
> 100
> 100
0.258
(0.143-
0.468)
4
OH
-
H
H
Br
H
> 100
> 100
5.32
5
6
7
OCH2CH3
OCH2CH3
OCH2CH3
-
-
-
H
H
H
H
H
OCH3
NO2
H
H
H
> 100
> 100
> 100
> 100
> 100
> 100
(3.68-7.68)
> 100
13.7
(9.55-19.6)
16.6
OCH3
OCH3
8
9
OCH2CH3
OCH2CH3
CH3
-
-
-
OCH3
H
H
OCH3
H
OCH3
OCH3
H
H
OCH3
H
> 100
> 100
(10.5-26.1)
9.05
> 100
14.8
> 100
42.1
(6.05-13.5)
17.0
10
H
(9.34-23.3) (33.6-52.7) (11.2-25.7)
16.6
11
12
13
CH3
CH3
CH3
-
-
-
H
H
H
H
H
H
CH3
NO2
H
H
H
> 100
> 100
> 100
> 100
(10.2-26.9)
> 100
> 100
25.9
13.6
OCH3
(17.7-38.0) (12.3-15.2)
32.9 9.82
14
15
CH3
CH3
-
-
H
H
OCH3
OH
OCH3
OH
OCH3
OH
> 60
(20.5-53.0) (7.90-12.2)
14.7
14.3
> 100
(8.74-24.6)
(7.79-26.4)
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 38
1
2
3
4
5
6
7
8
9
59.1
8.02
16
OCH3
OCH3
OCOCH3
OCOCH3
OCOCH3
OCOCH3
OCOCH3
OH
-
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
NO2
OCH3
H
H
H
H
H
H
H
H
H
H
H
H
H
> 100
(57.1-61.2) (6.90-9.32)
5.06
> 100
17
-
> 100
6.93
(3.89-6.57)
35.7
6.86
18
H
OCOCH3
OCOCH3
H
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
(5.76-8.33) (27.9-45.8) (5.86-8.03)
8.56 21.1 5.17
19
H
(6.90-10.6) (16.6-26.8) (4.85-5.50)
6.14
20
OCOCH3
> 100
> 100
(3.24-11.6)
22.0
64.4
21
OCOCH3 OCOCH3
OCOCH3 OCOCH3
> 100
(49.0-84.6) (12.5-38.7)
23.4
> 100
22
OCOCH3
H
> 100
4.62
(14.5-37.9)
42.3
5.09
23
H
H
H
(4.24-5.04) (36.5-49.1) (3.40-7.60)
2.40 12.1 3.85
(1.95-2.95) (9.69-15.2) (3.04-4.89)
12.5 55.3 9.16
(10.3-15.4) (45.3-67.4) (7.12-11.8)
6.20 14.0 23.6
(5.11-7.53) (10.1-19.5) (16.8-33.2)
6.28 25.5 8.50
(5.26-7.49) (21.2-30.5) (7.47-9.67)
24
OH
OH
H
25
OH
OH
H
OH
OH
OH
26
27
OH
H
OH
OH
OH
OH
6.77
1.71
86.4
Theophylline
-
(4.07-11.3) (1.02-2.90)
(73.6-101)
Values are geometric means of three experiments and are given in µM with 95%
confidence intervals. Numbers in brackets are the numerical value of the standard
uncertainty.
Based on previous results from our group, and on the potent A3 receptor affinity of some
of the described compounds,21 a novel family of coumarin-containing compounds was
investigated. Their specific framework formed by an aryl ring attached to position 3 and
a substitution pattern at positions 7 and/or 8 was analyzed for their capability to modulate
8
ACS Paragon Plus Environment

Page 9 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the affinity for adenosine receptor subtypes. Different chemical features were attached to
the phenyl ring at position 3, based on their physicochemical properties. Substituents from
different quadrants of the Craig plot were explored. In addition, the substitution by alkyl
or alkoxy groups was studied. Finally, the difference between the presence of acetoxy
and/or hydroxy substituents at 7 and/or 8 positions of the coumarin scaffold was explored
in detail. The effect of these substitutions on the binding and the level of selectivity for
the receptors was studied and compared. From our analysis, no affinity for the A2B
receptor was observed (Ki > 30 µM, data not shown).
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 structural variety on the studied series allowed us to obtain different affinity profiles
for the analyzed adenosine receptors. Selective A3 ligands, A2A/A3 or A1/A3 dual ligands
or non-selective (A1/A2A/A3) ligands were obtained. The analysis of the results allowed
us to have a detailed perspective on structure-activity relationships. In general,
compounds without attached groups on the 3-phenyl moiety are not selective, presenting
affinity for the three adenosine receptors (compounds 1, 10, 18, 19, 23 and 24). This
profile is independent of the nature of the 8 substitution (ethoxy, methyl, acetoxy) or even
with substitutions at both 7 and 8 positions (compound 19). These results are accordant
with those obtained in our previous study.21 8-(2-Oxopropoxy)-3-phenylcoumarin (Figure
1A) has reported affinity for three of the studied receptors in the low micromolar range
(Ki A1 = 10.4 µM, Ki A2A = 16.1 µM and Ki A3 = 7.1 µM).21 In the present study,
compound 24 shows similar affinity for the three adenosine receptors (Ki A1 = 2.4 µM,
Ki A2A = 12.1 µM and Ki A3 = 3.9 µM). In addition, compound 24 proved to be a better
adenosine receptor ligand than theophylline, our reference compound, which is in clinical
use to prevent and treat respiratory dysfunctions caused by asthma, emphysema, chronic
bronchitis, and other lung diseases.30 Besides being an A2B receptor antagonist,
theophylline is also a non-selective phosphodiesterase inhibitor. The profile of our
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 38
1
2
3
4
5
6
7
8
9
A1/A2A/A3 compounds may be very interesting thinking about multifactorial conditions,
as lung ischemia-reperfusion injury, likewise some xanthines, typical adenosine ligands.
When the scaffold presents a substituent at position 8 and another one at para position of
the 3-aryl moiety, independently of the nature of the substituent at positions 8 (ethoxy,
hydroxy, methyl or methoxy) or 3 (p-methylphenyl, p-bromophenyl or p-
methoxyphenyl), the compounds tend to be strong and selective A3 ligands (compounds
2, 4, 11 and 17). This data expands our previously results.21 The nanomolar affinity of
compound 4 for the hA3 receptor is about 2.5 times higher than that of our earlier reported
8-(2-oxopropoxy)-3-(p-tolyl)coumarin (Figure 1A).21 So far, this is the best compound
from all our research on the potential of coumarins as adenosine receptor ligands. In both
cases, position 8 presents electron donating groups. This characteristic can be important
to increase the activity and A3 selectivity. In addition, it seems that the size of the
atom/group at position 8 plays an important role on the activity as well.
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
Another relevant relationship between the structure and the measured activity was
observed for the series presenting 2- or 3-substitutions at the 3-aryl ring (meta and para
positions). Compounds 13, 14 and 21 have affinity for both A2A and A3, and compound
15 has affinity for A1 and A3 receptors. The only example of a di-substituted compound
at ortho/para positions (compound 8), proved to be A3 receptor selective (Ki A3 = 16.6
µM).
Finally, the inclusion of substituents at both 7 and 8 positions of the coumarin ring (both
acetoxy and hydroxy groups), independently of the substituents on the 3-aryl ring, tend
to give potent non-selective ligands (compounds 19, 24, 25 and 27). All these compounds
have affinity for the A1, A2A and A3 adenosine receptors in the low micromolar range.
Compounds 20 and 22 are the exception, being selective for the A3 receptor.
Finally, our most active and selective hA3 compounds were functionally tested using the
10
ACS Paragon Plus Environment

Page 11 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
GloSensor cAMP assay, which consists in a biosensor technology.31 The methodology is
detailed in the experimental section, and the binding information is shown in Table 2.
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
Table 2. In vitro antagonist activities of 2, 4, 5 and 17 at the hA3 adenosine receptor.
Compound
2
hA3 (IC50 μM)
18.4 (10.5-26.2)
4
5
3.1 (1.7-4.5)
29.5 (17.7-41.2)
25.3 (15.5-31.1)
17
The values are given in µM with 95% confidence intervals in parentheses.
Cell-based functional assays showed that compounds 2, 4, 5 and 17 behave as hA3
receptor antagonists, being able to counteract NECA-inhibited cAMP accumulation.
In a homogeneous family of twenty-seven compounds, it can be clearly observed a pattern
of affinity based on the structures (positions and nature of the substituents) that allows
new structure-affinity relationship understanding.
Molecular docking
We studied the most active molecule within the designed family using docking
calculations in the hA3 receptor to establish the key residues and important interactions
between the ligand and the protein. Our research group previously generated the hA3
protein structure by homology modeling (general details are described in Methods).21,32
We docked compound 4 to the hA3 with Glide standard precision (SP).33 We followed a
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 38
1
2
3
4
5
6
7
8
9
similar protocol already described in previous studies21,32 and validated in the hA2A
protein for which there are some crystal structures available in the PDB. As an example,
the root mean square deviation (RMSD) comparing ligands theoretical with co-
crystallized conformations in the 3EML34 and 3UZC crystal structures35 was 0.69 and
1.90.21
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
Molecular docking simulations presented a binding pose for compound 4 in the hA3
pocket that pointed the 3-aryl portion towards the bottom and the benzene of the coumarin
towards the surface (see Figure 3A). Compound 4 showed a pose in the hA3 that presented
some resemblance with the co-crystallized ligands ZM241385 and T4E in the 3EML and
3UZC hA2A crystal structures (see Figure 3B). The hydroxyl substituent at position 8
along with the carbonyl group are hypothesized to be important for the anchoring with
the studied protein and established two hydrogen bonds with the residue Asn250,
interacting via its amide (see Figure 3C). The equivalent residue in the hA2A (Asn253)
showed also an important role in ligand interaction in crystallographic and mutagenesis
studies.34,35,36 The benzopyrone moiety is proposed to establish aromatic π−π stacking
interactions with Phe168 residue of the second extracellular loop.
12
ACS Paragon Plus Environment

Page 13 of 38
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 3. a) General view of the hA3 receptor bound to compound 4. b) Comparison of
the co-crystallized conformations for the compounds (carbons in green) in the hA2A
[3EML (left) and 3UZC (right)] with the binding mode observed by docking in hA3 for
compound 4 (pink carbons). Both proteins, hA2A and hA3, were superimposed. c)
Hypothetical binding mode in the hA3 determined for compound 4 along with relevant
residues in ligand interaction. Hydrogen bonds are colored in yellow. For clarity, protein
ribbons were partially omitted. Template for the hA3 homology model: 3EML.
We extended the ligand-protein interaction analysis to per residue energy contributions
(see Figure 4A). The total energy is obtained from different contributions considering
Coulomb, van der Waals and H-bond energies. Residues Asn250, Phe168, Ile268,
Leu246, Leu264 and Leu90 have the highest contribution in the ligand-protein
recognition. Hydroxyl group at position 8 seems to be important for the interaction with
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 38
1
2
3
4
5
6
7
8
9
residue Asn250 and it could be the explanation for the high activity displayed by
compound 4. The bromine substituent at para position of the 3-aryl moiety could be also
a suitable substituent for the protein interaction. Compound 23, structurally similar to
compound 4 but with no substitution in the 3-aryl ring showed a similar binding mode,
but with lower Coulomb/van der Waals contributions with the residues close to the 3-aryl
(see Figure 4B). The bromine substituent caused a slight increase in the interaction with
residues Leu90, Ser271 and His272 compared to the 3-aryl ring without substituent
(compound 23). Moreover, we calculated the favored hydrophobic and hydrophilic areas
inside the hA3 protein pocket. The surfaces were generated taking into account the
residues located in a radius of 5Å from the ligand. The hydroxyl substituent presented at
8 position of the coumarin system and the oxygen atoms part of the pyrone system are
placed in polar areas whereas the coumarin and the 3-aryl scaffold are placed in a
hydrophobic area (Figure 4C). The bromine atom is buried in a deep hydrophobic area
that can favor the interaction with the protein. In fact, besides hydrophobic interactions
with residues Trp243 and Leu90 (also shown by compound 23) we detected additional
hydrophobic interactions with residues Val65 and Val61. The bromine atom does not
interact through halogen bonds with the protein, but its hydrophobic nature is well suited
for the hydrophobic surrounding residues. Our proposed binding mode explains the high
affinity of compound 4 for hA3 receptor. Our results agree with previous studies focused
on adenosine receptors21,32,34-37 in which residues as Asn250 and Phe168 were also
reported as important in ligand-protein recognition.
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
14
ACS Paragon Plus Environment

Page 15 of 38
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. a) Residue contribution to the interaction with ligand 4 (including van der
Waals, Coulomb and hydrogen bond energies). b) Differences in the residue contribution
of compounds 4 and 23. Bromine substitution in the 3-aryl ring (compound 4) causes a
higher Coulomb/van der Waals contribution in nearby residues. c) Hydrophobic (white)
and hydrophilic (red) areas inside the hA3 with the binding mode detected for compound
4 (isovalues of -0.5 and -4.87 for hydrophobic and hydrophilic surfaces respectively). The
8-hydroxyl is placed in the hydrophilic area while the 3-aryl scaffold is placed in the hydrophobic
one (template for the hA3 homology model: 3EML). d) Dihedral angle energy plot for
compound 4 extracted from conformational analysis.
The results provided by the docking calculations are aligned with the X-ray structure of
compound 4 (RMSD=0.16Å). As the analysis of the conformational preorganization of
the compound could provide some insights in drug design,38 we performed a comparative
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 38
1
2
3
4
5
6
7
8
9
structural study in terms of energetic stabilization around the dihedral angle C(2)-C(3)-
C(13)-C(18) between the 3-aryl ring and the coumarin scaffold in compound 4. The
dihedral angle was rotated in 5-degree increments during the conformational analysis.
Compound 4 showed two optimal energetic areas with values for the dihedral angle from
-140° to -40° and from 40° to 140°. The described X-ray crystallized structure and the
docking pose showed values of 145.4° and -152.1° for the dihedral angle. It has been
reported that in many cases the bioactive ligand conformation is not correspondent with
the global minimum energy conformer as both protein and ligand can reorganize their
atomic coordinates to optimize complementarity.39 For compound 4, the X-ray and the
docking conformation are close to the global minimum energy structure. The dihedral
angle energy plot for compound 4 is shown in Figure 4D with the values obtained in
docking and X-ray studies.
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 4 showed hA3 activity and no affinity for the other adenosine subtypes. We
performed additional molecular docking simulations in the hA2A crystallized structure
3EML to explain the affinity decrease for compound 4. Molecular docking with no
crystallographic water molecules in the hA2A binding region yielded a binding mode
deeply buried in the cavity (see Figure 5A). The displacement of water in the pocket could
have a positive or negative impact in ligand binding affinity and enthalpy/entropy of the
system. 40 The 4-bromophenyl scaffold of compound 4 with hydrophobic characteristics
is placed in the polar area occupied by the water in the hA2A. This scaffold has not the
potential to establish similar hydrogen bonds to the hydrogen bonding network described
by the water molecules. As a hypothesis, the global process could be energetically
unfavorable reducing the activity of compound 4 in hA2A. Nevertheless, the docking with
crystallographic water molecules in the hA2A cavity showed a pose for compound 4 that
binds a shallower area of the pocket and disrupts key interactions with some residues
16
ACS Paragon Plus Environment

Page 17 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
located deeper in the cavity, such as Asn253 (see Figure 5B). The disruption of the
binding mode explained for compound 4 in the hA3 could be responsible for the low
activity detected in hA2A. In fact, compound 4 showed a different residue profile binding
in the hA3 and hA2A with a clear reduction in the interaction energy between the ligand
and Asn253, a key residue in the interaction between compound 4 and the hA3 (see
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
Figure 5. Superposition of the binding poses for compound 4 observed by docking in hA3
(ligand in pink carbons) and hA2A (ligand in orange carbons). a) Hypothetical binding
mode of compound 4 buried in an area of multiple crystallographic water molecules in
the hA2A (although water molecules are shown in red/white mesh, the docking was
performed with no water in the cavity). This hypothetical pose would need to displace the
mentioned waters. b) Hypothetical binding mode of compound 4 in a shallower area close
to the extracellular loops in the hA2A (docking performed considering the water molecules
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 38
1
2
3
4
5
6
7
8
9
presented in the protein pocket). Disruption of the binding mode detected for compound
4 in hA3 (carbons in pink) could be the cause of lack of affinity in the hA2A. c) Key
residues in ligand recognition in the hA3 that are not conserved in both hA2A and hA1
subtypes. Other residues that differ in the hA3, hA2A and hA1 subtypes are Leu264,
Met270 and Thr270 respectively, not represented in the figure due to clarity reasons.
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 residue Leu90 in the hA3 that established key interactions with the 4-bromophenyl
of compound 4 is substituted by Val84 and Val87 in the hA2A and hA1 respectively,
although those residues present similar hydrophobic characteristics. However, some of
the residues located in the extracellular domain of the hA3 are not present in the other
subtypes, which can be also an important factor to explain the selectivity. Residues such
as Gln167 with polar properties, Val169, Ile253 and Leu264 with hydrophobic
characteristics are substituted in the hA2A by the corresponding hydrophobic Leu167, the
negative charged and hydrophilic Glu169, Thr256 with hydrophilic properties and the
hydrophobic Met270. The subtype hA1 contains some residues, as Glu170, Glu172,
Thr270 and Thr257 with hydrophilic properties more suitable for the interaction with
polar substituents. Figure 5C shows important residues in ligand recognition in the hA3
that are not conserved in both hA2A and hA1 subtypes. The chemical characteristics of
the residues can affect the entrance of the designed ligand in the pocket as well as the
accommodation in the binding cleft and favor the selectivity against the hA3.
Conclusions
3-Arylcoumarins proved to be an attractive scaffold for the development of multitarget
molecules. In particular, 8-substituted compounds are promising molecules as novel
antagonists of adenosine receptors, presenting different affinity and selectivity profiles.
18
ACS Paragon Plus Environment

Page 19 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
An extensive overview of our experimental data helped to conclude that both affinity
and/or selectivity profiles of the new molecules against adenosine receptors could be
modified taking into account the characteristics of the substituents at positions 3, 7 and 8
of the coumarins. Compound 4 (3-(4-bromophenyl)-8-hydroxycoumarin) appeared as the
best compound within this study, and all the molecules developed by our group so far. Its
structure was corroborated by X-ray crystallography, revealing that this coumarin
derivative presents a p-bromophenyl substituent at 3 position, presenting a hydroxyl
group attached to position 8 as well. This last substituent proved to be fundamental for
the interaction with Asn250 residue and it could be the answer to explain the high affinity
showed by this antagonist. The substituents on the 3-aryl ring are also important for this
selectivity pattern. On the other hand, compound 24 proved to display the highest affinity
on three receptors, being a better ligand than the reference compound, theophylline. This
series offers the possibility to better understand important clues to modulate the
interactions with adenosine receptors. Compound 24 can be the inspiration for the design
of multifunctional compounds with interest on multifactorial conditions, and compound
4 can be the inspiration for the development of new coumarins as strong and selective A3
antagonists.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
CHEMISTRY
General remarks. Starting materials and reagents were obtained from commercial
suppliers (Sigma-Aldrich) and were used without further purification. Melting points
(Mp) are uncorrected and were determined with a Reichert Kofler thermopan or in
capillary tubes in a Büchi 510 apparatus. 1H NMR (300 MHz) and 13C NMR (75.4 MHz)
spectra were recorded with a Bruker AMX spectrometer using CDCl3 or DMSO-d6 as
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 38
1
2
3
4
5
6
7
8
9
solvent. Chemical shifts (δ) are expressed in parts per million (ppm) using TMS as an
internal standard. Coupling constants J are expressed in Hertz (Hz). Spin multiplicities
are given as s (singlet), d (doublet), t (triplet), q (quartet) and m (multiplet). Mass
spectrometry was carried out with a Hewlett-Packard 5988A spectrometer. Elemental
analyses were performed by a Perkin-Elmer 240B microanalyzer and are within ±0.4%
of calculated values in all cases. The analytical results document ≥ 98% purity for all
compounds. Flash chromatography (FC) was performed on silica gel (Merck 60, 230-400
mesh); analytical TLC was performed on precoated silica gel plates (Merck 60 F254).
Organic solutions were dried over anhydrous sodium sulfate. Concentration and
evaporation of the solvent after reaction or extraction was carried out on a rotary
evaporator (Büchi Rotavapor) operating under reduced pressure.
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 for the synthesis of 3-phenylcoumarins (1-3, 5-14, 16 and 17). A
solution of 2-hydroxybenzaldehyde (7.34 mmol) and the corresponding phenylacetic acid
(9.18 mmol) in dimethyl sulfoxide (15 mL) was prepared. N,N’-
Dicyclohexylcarbodiimide (11.46 mmol) was added, and the mixture was heated in an oil
bath at 110 ºC for 24 h. Ice (100 mL) and acetic acid (10 mL) were added to the reaction
mixture. After keeping it at room temperature for 2 h, the mixture was extracted with
ether (3 x 25 mL). The organic layer was extracted with sodium bicarbonate solution (50
mL, 5%) and then water (20 mL). The solvent was evaporated under vacuum, and the dry
residue was purified by FC (hexane/ethyl acetate 9:1).
General procedure for the synthesis of hydroxy-3-phenylcoumarins (4 and 15). A
solution of 3 or 14 (0.50 mmol) in acetic acid (5 mL) and acetic anhydride (5 mL), at 0
ºC, was prepared. Hydriodic acid 57% (10 mL) was added dropwise. The mixture was
20
ACS Paragon Plus Environment

Page 21 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
stirred, under reflux temperature, for 3 h. The solvent was evaporated under vacuum, and
the dry residue was purified by crystallization (CH3CN).
7
8
9
General procedure for the synthesis of acetoxy-3-phenylcoumarins (18-22).
Compound 18-22 were synthesized under anhydrous conditions, using material
previously dried at 60 ºC for at least 12 h and at 300 ºC during few minutes immediately
before use. A solution containing anhydrous CH3CO2K (2.94 mmol), phenylacetic acid
(1.67 mmol) and the corresponding hydroxysalicylaldehyde (1.67 mmol), in Ac2O (1.2
mL), was refluxed for 16 h. The reaction mixture was cooled, neutralized with 10%
aqueous NaHCO3, and extracted with EtOAc (3 x 30 mL). The organic layers were
combined, washed with distilled water, dried (anhydrous Na2SO4), and evaporated under
reduced pressure. The product was purified by recrystallization in EtOH and dried, to
afford the desired compound.
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 for the synthesis of hydroxy-3-phenylcoumarins (23-27).
Compounds 23-27 were obtained by hydrolysis of their acetoxylated counterparts 18-22,
respectively. The appropriate acetoxylated coumarin, mixed with 2N aqueous HCl and
MeOH, was refluxed during 3 h. The resulting reaction mixture was cooled in an ice-bath
and the reaction product, obtained as solid, was filtered, washed with cold distilled water,
and dried under vacuum, to afford the desired compound.
3-(4-bromophenyl)-8-ethoxycoumarin (compound 3). Yield 41%. M.p. 162-163 ºC. 1H
NMR (CDCl3) δ (ppm), J (Hz): 1.52 (t, 3H, CH3, J=7.0), 4.20 (q, 2H, CH2, J=7.0), 7.06-
7.24 (m, 3H, H-5, H-6, H-7), 7.32-7.45 (m, 4H, H-2’, H-3’, H-5’, H-6’), 7.79 (s, 1H, H-
13
4). C NMR (CDCl3) δ (ppm): 14.8, 65.0, 114.7, 119.3, 120.2, 123.14 124.5, 127.3,
130.1, 131.6, 133.6, 140.2, 143.4, 146.4, 159.9. MS m/z (%): 347 (18), 346 (98), 345 (19),
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 38
1
2
3
4
5
344 (M+, 100). Ana. Elem. Calc. for C17H13BrO3: C, 59.15; H, 3.80. Found: C 58.98, H
3.80.
6
7
8
9
3-(4-bromophenyl)-8-hydroxycoumarin (compound 4). Yield 42%. M.p. 261-217 ºC.
1H NMR (DMSO-d6) δ (ppm), J (Hz): 7.05-7.16 (m, 3H, H-5, H-6, H-7), 7.56-7.71 (m,
4H, H-2’, H-3’, H-5’, H-6’), 8.22 (s, 1H, H-4), 10.27 (s, 1H, OH). 13C NMR (DMSO-d6)
δ (ppm): 118.2, 118.7, 120.3, 121.9, 124.6, 125.5, 130.6, 131.2, 133.9, 141.3, 141.7,
144.4, 159.5. MS m/z (%): 319 (16), 318 (98), 317 (17), 316 (M+, 100). Ana. Elem. Calc.
for C15H9BrO3: C, 56.81; H, 2.86. Found: C 56.84, H 2.87.
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
8-ethoxy-3-(3,4-dimethoxyphenyl)coumarin (compound 7). Yield 41%. M.p. 138-139
ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 1.53 (t, 3H, CH3, J=7.0), 3.90 (s, 3H, OCH3), 3.95
(s, 3H, OCH3), 4.21 (q, 2H, CH2, J=7.0), 6.94 (d, 1H, H-7, J=8.2), 7.02-7.09 (m, 2H, H-
2’, H-6’), 7.20 (d, 1H, H-5’, J=7.9), 7.28-7.33 (m, 2H, H-5, H-6), 7.77 (s, 1H, H-4). 13C
NMR (CDCl3) δ (ppm): 14.5, 55.7, 64.6, 110.7, 111.5, 114.0, 118.9, 120.2, 121.0, 124.0,
127.1, 128.0, 138.7, 142.8, 146.0, 148.4, 149.4, 160.1. MS m/z (%): 327 (55), 326 (M+,
100). Anal. Elem. Calc. for C19H18O5: C, 69.93; H, 5.56. Found: C, 69.91; H, 5.53.
8-ethoxy-3-(2,4-dimethoxyphenyl)coumarin (compound 8). Yield 36%. M.p. 95-96
ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 1.50 (t, 3H, CH3, J=7.0), 3.80 (s, 3H, OCH3), 3.82
(s, 3H, OCH3), 4.19 (q, 2H, CH2, J=7.0), 6.45-6.57 (m, 2H, H-3’, H-5’), 7.03-7.09 (m,
1H, H-7), 7.14-7.19 (m, 2H, H-6, H-5), 7.30 (d, 1H, H-6’, J=8.9), 7.68 (s, 1H, H-4). 13C
NMR (CDCl3) δ (ppm): 14.8, 55.4, 55.7, 64.9, 99.0, 104.5, 114.1, 116.8, 119.1, 120.4,
124.0, 126.3, 131.4, 141.4, 146.3, 154.2, 158.3, 160.2, 161.4. MS m/z (%): 327 (12), 326
(M+, 47). Anal. Elem. Calc. for C19H18O5: C, 69.93; H, 5.56. Found: C, 69.96; H, 5.58.
8-ethoxy-3-(3,4,5-trimethoxyphenyl)coumarin (compound 9). Yield 40%. M.p. 142-
143 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 1.58 (t, 3H, CH3, J=7.0), 3.94 (s, 3H, OCH3),
3.96 (s, 3H, OCH3), 3.99 (s, 3H, OCH3), 4.25 (q, 2H, CH2, J=7.0), 7.0 (s, 2H, H-2’, H-
22
ACS Paragon Plus Environment

Page 23 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
6’), 7.14-7.20 (m, 1H, H-6), 7.24-7.27 (m, 1H, H-7), 7.30-7.32 (m, 1H, H-5), 7.83 (s, 1H,
H-4). 13C NMR (CDCl3) δ (ppm): 14.8, 56.2, 56.3, 65.0, 105.9, 114.5, 119.2, 120.3, 124.4,
128.2, 130.2, 139.7, 133.8, 146.3, 153.0. MS m/z (%): 357 (23), 356 (M+, 100). Anal.
Elem. Calc. for C20H20O6: C, 67.41; H, 5.66. Found: C, 67.42; H, 5.68.
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
8-methyl-3-(4-nitrophenyl)coumarin (compound 12). Yield 61%. M.p. 229-230 ºC. 1H
NMR (DMSO-d6) δ (ppm), J (Hz): 2.38 (s, 3H, CH3), 7.28 (t, 1H, H-6, J=7.4), 7.52 (d,
1H, H-7, J=7.4), 7.61 (d, 1H, H-5, J=7.4), 7.99 (d, 2H, H-2’, H-6’, J=9.0), 8.28 (d, 2H,
H-3’, H-5’, J=9.0), 8.41 (s, 1H, H-4). 13C NMR (DMSO-d6) δ (ppm): 15.3, 123.8, 124.8,
124.9, 125.4, 127.3, 130.2, 131.3, 134.1, 141.8, 143.4, 147.5, 152.0, 159.8. MS m/z (%):
282 (18), 281 (M+, 100). Anal. Elem. Calc. for C16H11NO4: C, 68.33; H, 3.94. Found: C,
68.36; H, 3.93.
3-(3,4-dimethoxyphenyl)-8-methylcoumarin (compound 13). Yield 63%. M.p. 135-
136 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 2.51 (s, 3H, CH3), 3.94 (s, 6H, 2xOCH3), 6.95
(d, 1H, H-5’, J=8.1), 7.16-7.39 (m, 5H, H-5, H-6, H-7, H-2’, H-6’), 7.77 (s, 1H, H-4). 13C
NMR (CDCl3) δ (ppm): 15.4, 55.9, 110.9, 111.7, 119.4, 121.1, 124.0, 125.4, 125.8, 127.5,
127.5, 132.4, 139.2, 148.6, 149.6, 151.6, 160.9. MS m/z (%): 297 (46), 296 (M+, 100).
Anal. Elem. Calc. for C18H16O4: C, 72.96; H, 5.44. Found: C, 73.00; H, 5.49.
3-(3,4,5-trimethoxyphenyl)-8-methylcoumarin (compound 14). Yield 69%. M.p. 163-
164 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 2.49 (s, 3H, CH3), 3.89 (s, 3H, OCH3), 3.90
(s, 3H, OCH3), 3.93 (s, 3H, OCH3), 6.95 (s, 2H, H-2’, H-6’), 7.18-7.23 (m, 2H, H-6, H-
13
7), 7.38 (d, 1H, H-5, J=7.4), 7.79 (s, 1H, H-4). C NMR (CDCl3) δ (ppm): 15.4, 56.2,
56.3, 106.0, 119.3, 124.1, 125.6, 125.9, 127.7, 130.3, 132.7, 138.7, 139.9, 151.7, 153.1,
160.7. MS m/z (%): 327 (23), 326 (M+, 100). Ana. Elem. Calc. for C19H18O5: C, 69.93;
H, 5.56. Found: C, 69.96; H, 5.59.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 38
1
2
3
4
1
8-acetoxy-3-phenylcoumarin (compound 18). Yield 64%. M.p. 189-190 ºC. H NMR
5
6
(CDCl3) δ (ppm), J (Hz): 2.44 (s, 3H, CH3), 7.18-7.50 (m, 7H, H-2’, H-3’, H-4’, H-5’, H-
7
8
13
6’, H-6, H-7), 7.66-7.71 (m, 1H, H-5), 7.83 (s, 1H, H-4). C NMR (CDCl3) δ (ppm):
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
20.7, 119.3, 123.0, 124.5, 124.8, 126.0, 128.2, 128.9, 129.2, 135.3, 142.1, 144.2, 149.2,
160.9, 168.5. MS m/z (%): 281 (19), 280 (M+, 100). Ana. Elem. Calc. for C17H12O4: C,
72.85; H, 4.32. Found: C, 72.90; H, 4.30.
1
7,8-diacetoxy-3-phenylcoumarin (compound 19). Yield 81%. M.p. 171-172 ºC. H
NMR (CDCl3) δ (ppm), J (Hz): 2.34 (s, 3H, CH3), 2.43 (s, 3H, CH3), 7.15 (d, 1H, H-6,
J=8.7), 7.39-7.49 (m, 4H, H-5, H-2’, H-4’, H-6’), 7.64-7.69 (m 2H, H-3’, H-5’), 7.78 (s,
1H, H-4). 13C NMR (CDCl3) δ (ppm): 20.3, 20.6, 118.5, 119.1, 124.9, 128.1, 128.5, 129.0,
130.0, 134.3, 139.1, 144.9, 146.5, 159.0, 167.4, 167.8. MS m/z (%): 339 (11), 338 (M+,
100). Ana. Elem. Calc. for C19H14O6: C, 67.45; H, 4.17. Found: C, 67.52; H, 4.20.
7,8-diacetoxy-3-(4-acetoxyphenyl)coumarin (compound 20). Yield 75%. M.p. 195-
196 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 2.32 (s, 3H, CH3), 2.34 (s, 3H, CH3), 2.42 (s,
3H, CH3), 7.13-7.19 (m, 3H, H-6, H-3’, H-5’), 7.41 (d, 1H, H-5, J=8.7), 7.68 (d, 2H, H-
2’, H-6’, J=8.6), 7.77 (s, 1H, H-4). 13C NMR (CDCl3) δ (ppm): 20.2, 20.5, 21.1, 118.3,
119.1, 121.6, 124.9, 127.1, 129.6, 131.8, 131.9, 139.1, 144.2, 144.92, 151.1, 161.9, 167.
3. MS m/z (%): 397 (9), 396 (M+, 93). Ana. Elem. Calc. for C21H16O8: C, 63.64; H, 4.07.
Found: C, 63.61; H, 4.09.
8-acetoxy-3-(3,4-diacetoxyphenyl)coumarin (compound 21). Yield 47%. M.p. 144-
145 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 2.33 (s, 6H, 2xCH3), 2.44 (s, 3H, CH3), 7.27-
7.31 (m, 3H, H-5’, H-2’, H-6), 7.41-7.45 (m, 1H, H-7), 7.58-7.64 (m, 2H, H-6’, H-5),
7.84 (s, 1H, H-4).13C NMR (CDCl3) δ (ppm): 20.6, 20.6, 20.6, 120.6, 123.4, 123.6, 124.3,
125.1, 125.4, 126.6, 127.0, 132.8, 137.5, 139.8, 139.9, 140.0, 141.9, 158.8, 168.0, 168.1,
24
ACS Paragon Plus Environment

Page 25 of 38
Journal of Medicinal Chemistry
1
2
3
4
168.5. MS m/z (%): 397 (15), 396 (M+, 89). Ana. Elem. Calc. for C21H16O8: C, 63.64; H,
5
6
4.07. Found: C, 63.66; H, 4.10.
7
8
9
7,8-diacetoxy-3-(3,4-diacetoxyphenyl)coumarin (compound 22). Yield 54%. M.p.
202-203 ºC. 1H NMR (CDCl3) δ (ppm), J (Hz): 2.35 (s, 3H, CH3), 2.37 (s, 3H, CH3), 2.39
(s, 3H, CH3), 2.46 (s, 3H, CH3), 7.20 (d, 1H, H-6’, J=8.6), 7.30 (s, 1H, H-2’), 7.46 (d, 1H,
H-5’, J=8.6), 7.59-7.65 (m, 2H, H-5, H-6), 7.84 (s, 1H, H-4). 13C NMR (CDCl3) δ (ppm):
20.1, 20.2, 20.5, 20.6, 118.1, 119.2, 122.0, 123.4, 123.6, 125.9, 126.2, 132.7, 139.5,
141.9, 142.6, 145.1, 148.9, 150.3, 160.4, 167.6. MS m/z (%): 455 (9), 454 (M+, 92). Ana.
Elem. Calc. for C23H18O10: C, 60.80; H, 3.99. Found: C, 60.82; H, 4.01.
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
7,8-dihydroxy-3-(4-hydroxyphenyl)coumarin (compound 25). Yield 91%. M.p. 290-
291 ºC. 1H NMR (DMSO-d6) δ (ppm), J (Hz): 6.78-6.81 (m, 3H, H-6, H-3’, H-5’), 7.04
(d, 1H, H-5, J=8.5), 7.51-7.55 (m, 2H, H-2’, H-6’), 7.97 (s, 1H, H-4), 9.63 (s, 1H, OH),
10.05 (s, 1H, OH), 10.08 (s, 1H, OH). 13C NMR (DMSO-d6) δ (ppm): 115.0, 115.2, 116.3,
126.0, 126.6, 129.8, 133.6, 139.7, 142.5, 149.0, 149.3, 157.2, 162.7. MS m/z (%): 271
(16), 270 (M+, 100). Ana. Elem. Calc. for C15H10O5: C, 66.67; H, 3.73. Found: C, 66.65;
H, 3.75.
8-hydroxy-3-(3,4-dihydroxyphenyl)coumarin (compound 26). Yield 79%. M.p. 259-
260 ºC. 1H NMR (DMSO-d6) δ (ppm), J (Hz): 7.79 (d, 1H, H-5’, J=8.3), 7.01-7.22 (m,
5H, H-5, H-6, H-7, H-2’, H-6’), 8.03 (s, 1H, H-4), 9.07 (s, 1H, OH), 9.22 (s, 1H, OH),
13
10.16 (s, 1H, OH). C NMR (DMSO-d6) δ (ppm): 115.5, 116.2, 117.6, 118.5, 120.0,
120.8, 124.6, 125.9, 126.8, 138.8, 139.0, 145.0, 146.3, 155.2, 164.8. MS m/z (%): 271
(18), 270 (M+, 100). Ana. Elem. Calc. for C15H10O5: C, 66.67; H, 3.73. Found: C, 66.66;
H, 3.71.
Compounds 1, 2, 5, 6, 10, 11, 15, 16, 17, 23, 24 and 27 have been previously
described.23,24,25,41,42,43,44,45,46
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 38
1
2
3
4
5
6
7
8
Adenosine receptors affinity. The affinity of the studied compounds for the human
adenosine receptor subtypes hA1, hA2A and hA3, was determined with radioligand
competition experiments in Chinese hamster ovary (CHO) cells that were stably
transfected with the individual receptor subtypes. The radioligands used were 1 nM
[3H]CCPA for hA1, 10 nM [3H]NECA for hA2A, and 1 nM [3H]HEMADO for hA3
receptors. The results were expressed as Ki values (dissociation constants), which were
calculated with the program Prism (GraphPad Software). Ki values are reported as
geometric means of three independent experiments with each tested concentration of
compound measured in duplicate. As an interval estimate for the dissociation constants,
95% confidence intervals are given in parentheses. Details for pharmacological
experiments are described in previous works.26,29 Due to the lack of a suitable radioligand
for hA2B receptors, the potency of antagonists at the hA2B receptor (expressed on CHO
cells) was determined by inhibition of NECA-stimulated adenylyl cyclase activity.
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
GloSensor cAMP Assay. Functional A3 adenosine receptor activity was determined
using a biosensor technology called GloSensor cAMP assay. It consists of a mutant form
of Firefly luciferase into which a cAMP-binding protein moiety has been inserted. When
the cAMP binds the biosensor there is a conformational change which induce an increase
of light output that allow to evaluate the activity of ligands at the receptor under study.
Briefly, cells stably expressing the hA3 adenosine receptor and transiently the biosensor,
were harvested and incubated in equilibration medium containing a 3% v/v GloSensor
cAMP reagent stock solution, 10% FBS, and 87% CO2 independent medium. After 2 h
of incubation at room temperature, cells were dispensed in the wells of a 384-well plate
and NECA reference agonist or the understudy compounds, at different concentrations,
26
ACS Paragon Plus Environment

Page 27 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
were added. When compounds were unable to inhibit the cAMP production they were
studied as antagonists. In particular, the antagonist profile was evaluated by assessing the
ability of these compounds to counteract NECA-induced decrease of cAMP
accumulation. Responses were expressed as percentage of the maximal relative
luminescence units (RLU). Concentration–response curves were fitted by a nonlinear
regression with the Prism 5.0 programme (GraphPAD Software, San Diego, CA, USA).
The antagonist profile of the compounds was expressed as IC50, which is the
concentration of antagonists that produces 50% inhibition of the agonist effect. Three
independent experiments with each tested concentration of compound measured five
times. The final values are given with 95% confidence intervals.31
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
hA3 homology model. Homology model of hA3 receptor along with a description of its
construction was previously published by our group.21,32 The hA2A crystallized structure
(PDB code: 3EML)34 was used as a template for the development of the homology model.
The alignment between both proteins was reported previously and it is included in
Katritch et al.,47 considering highly conserved residues in the TMs. The Homology Model
module in MOE software was used to develop the hA3 model.48 For residues that are
identical, the heavy atoms coordinates are copied to the new target from the template,
whereas only the backbone is taken into account for different residues. The residues
placed in the loops with no specified coordinates are constructed based on high resolution
fragments available in the PDB. A Boltzmann-weighted function is used for the selection
of the loops. The top hA3 model according to the Generalized Born/Volume Integral
(GB/VI) scoring was selected. The geometrical quality of Phi-Psi dihedrals, bond
lengths, bond angles, dihedrals, side chains and non-bonded interactions was assessed
with the Protein Geometry module. Protein pocket was optimized by docking high
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 38
1
2
3
4
5
6
affinity ligands using the Induced Fit Docking workflow allowing flexibility in the pocket
residues.33 The best hA3 homology models showed ROC curves greater than 0.80 in the
7
8
9
discrimination of ligands from decoys. A detailed description of the homology modeling was
provided in previous studies.21,32
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
Moreover, our hA3 receptor was compared to a model generated using the protein structure
homology model server SwissModel.49 The server automatically detected the hA1 (PDB code:
5UEN)50 as the best template and generated a hA3 model that is in agreement with our reported
homology model using as a template the hA2A receptor (PDB code: 3EML). The average RMSD
between both models is 2.8 Å whereas the RMSD between both pockets is 0.9 Å. More details
about homology model comparison are provided in the Supporting Information.
Molecular docking. Molecular docking simulations in the hA3 and hA2A proteins were
run using the Schrödinger package.33 Ligands were prepared with the LigPrep module
that included the next steps: generation of tautomers and different protonation states
(pH=7±2) and optimization of the molecular structures. Protein structures were also
prepared with the module Protein Preparation Wizard to optimize protonation states of
some residues and the H-bond network of the proteins. After this step, a grid centered in
the pocket was generated (van der Waals radius scaling=1.0; partial charge cut-off=0.25).
The ligands were docked to the hA3 and hA2A using Glide standard precision (SP mode).
Top scoring function poses were selected as representative of the simulations.
AUTHOR INFORMATION
Corresponding Authors
* Maria J. Matos: phone, +34 881814936; E-mail, mariajoao.correiapinto@usc.es or
maria.matos@fc.up.pt.
* Fernanda Borges: phone, +351 220402560; E-mail, fborges@fc.up.pt.
28
ACS Paragon Plus Environment

Page 29 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
ORCID
7
8
9
Maria J. Matos: 0000-0002-3470-8299
S. Vilar: 0000-0003-2663-4370
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
S. Vazquez-Rodriguez: 0000-0003-1356-8984
K.-N. Klotz: 0000-0003-3553-3205
M. Buccioni: 0000-0002-8383-0813
G. Delogu: 0000-0003-0307-1544
L. Santana: 0000-0001-6056-8253
E. Uriarte: 0000-0001-6218-2899
F. Borges: 0000-0003-1050-2402
Author Contributions
M.J.M., E.U. and F.B. conceived and supervised the study. M.J.M., S.V.-R. and G.D.
performed the synthesis, purification and characterization of the compounds. E.U. and
L.S. co-supervised the synthetic part of the work. S.K. and M.B. performed the
pharmacological assays on adenosine receptors. K.-N.K. supervised and validated the
pharmacological assays. S.V. performed the docking studies. L.S. and F.B. co-supervised
the modelling studies. M.J.M. wrote the manuscript with contributions from all authors.
All authors approved the final version of the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGEMENTS
29
ACS Paragon Plus Environment