8-(2-Furyl)adenine derivatives as A2A adenosine receptor ligands

Article abstract of DOI:10.1016/j.ejmech.2013.10.006

Selective adenosine receptor modulators are potential tools for numerous therapeutic applications, including cardiovascular, inflammatory, and neurodegenerative diseases. In this work, the synthesis and biological evaluation at the four human adenosine receptor subtypes of a series of 9-substituted 8-(2-furyl)adenine derivatives are reported. Results show that 8-(2-furyl)-9-methyladenine is endowed with high affinity at the A_2A subtype. Further modification of this compound with introduction of arylacetyl or arylcarbamoyl groups in N⁶-position takes to different effects on the A_2A affinity and in particular on the selectivity versus the other three adenosine receptor subtypes. A molecular modelling analysis at three different A_2A receptor crystal structures provides an interpretation of the obtained biological results.

Full text of DOI:10.1016/j.ejmech.2013.10.006

European Journal of Medicinal Chemistry 70 (2013) 525e535
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
8-(2-Furyl)adenine derivatives as A2A adenosine receptor ligands
Diego Dal Ben a, Michela Buccioni a, Catia Lambertucci a, Ajiroghene Thomas a,
Karl-Norbert Klotz b, Stephanie Federico c, Barbara Cacciari d, Giampiero Spalluto c,
,
Rosaria Volpini a
*
a School of Pharmacy, Medicinal Chemistry Unit, University of Camerino, Via S. Agostino 1, I-62032 Camerino, MC, Italy
b Institut für Pharmakologie und Toxikologie, University of Würzburg, D-97078 Würzburg, Germany
c Dipartimento di Scienze Chimiche e Farmaceutiche, University of Trieste, Piazzale Europa 1, I-34127 Trieste, Italy
d Dipartimento di Scienze Chimiche e Farmaceutiche, University of Ferrara, Via Fossato di Mortara 17-19, I-44100 Ferrara, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
Selective adenosine receptor modulators are potential tools for numerous therapeutic applications,
including cardiovascular, inflammatory, and neurodegenerative diseases. In this work, the synthesis and
biological evaluation at the four human adenosine receptor subtypes of a series of 9-substituted 8-(2-
furyl)adenine derivatives are reported. Results show that 8-(2-furyl)-9-methyladenine is endowed
with high affinity at the A2A subtype. Further modification of this compound with introduction of ary-
lacetyl or arylcarbamoyl groups in N6-position takes to different effects on the A2A affinity and in
particular on the selectivity versus the other three adenosine receptor subtypes. A molecular modelling
analysis at three different A2A receptor crystal structures provides an interpretation of the obtained
biological results.
Received 25 July 2013
Accepted 3 October 2013
Available online 17 October 2013
Keywords:
Purinergic receptors
Adenosine receptors
Adenosine receptor antagonists
Adenine derivatives
Purine derivatives
Ó 2013 Elsevier Masson SAS. All rights reserved.
Molecular modelling
1. Introduction
neurotransmitters, lipolysis, as well as renal, platelet, and white
blood cell functions [7]. Hence, selective AR modulators are
Agonism and antagonism of adenosine receptors (ARs) are
responsible of several effects in different organ systems suggesting
that regulation of ARs has substantial therapeutic potential [1]. All
ARs are G protein-coupled receptors (GPCRs), with four subtypes
classified as A1, A2A, A2B, and A3 ARs [2e4], which exert their
physiological role by activation or inhibition of different second
messenger systems. In particular, the modulation of adenylyl
cyclase activity could be considered as the principal signal medi-
ated by these receptors [5,6]. Furthermore, ARs can be distin-
guished on the base of their tissue distribution and unique
pharmacological profiles. In fact, a variety of physiological actions
can be referred to adenosine, including effects on heart rate and
atrial contractility, vascular smooth muscle tone, release of
promising for numerous therapeutic applications, including car-
diovascular, inflammatory, and neurodegenerative diseases [1].
The prototypical AR antagonists are the naturally occurring
methylxanthines, whose modifications led to derivatives endowed
with marked potency and selectivity for specific AR subtypes [8e
10]. However, only a few xanthine antagonists such as caffeine and
theophylline have been approved as drugs for their central ner-
vous system (CNS) stimulating, diuretic, and bronchodilating ef-
fects [10,11]. Very recently, the A2AAR antagonist istradefylline
(KW-6002) has been approved for manufacturing and marketing
in Japan for the treatment of Parkinson’s disease [12]. Many
different classes of compounds have been proposed as AR antag-
onists, with both good affinity and selectivity [13,14]. Among
them, a series of AR antagonists have been obtained by removing
the ribose moiety from the natural ligand adenosine and by
modifying the adenine core. In fact, the introduction of different
substituents in 2-, 8-, and 9-position of adenine resulted in high-
affinity antagonists with distinct receptor selectivity profile [15e
26]. In particular, some 8-substituted-9-ethyladenine derivatives
were reported to ameliorate motor deficits in rat models of Par-
kinson’s disease, suggesting potential therapeutic application for
these compounds [27]. Among them, the 8-bromo-9-ethyladenine
Abbreviations: AR, adenosine receptor; GPCR, G protein-coupled receptor; CNS,
central nervous system; DMF, dimethylformamide; NBS, N-bromosuccinimide;
CHO, Chinese hamster ovary; pdb, protein data bank; EL, extracellular loop; IL,
intracellular loop; MOE, molecular operating environment; RMSD, root mean
square deviation; TM, transmembrane; TLC, thin-layer chromatography; IR,
infrared; THF, tetrahydrofuran.
* Corresponding author. Tel.: þ39 0737 402278; fax: þ39 0737 402295.
E-mail address: rosaria.volpini@unicam.it (R. Volpini).
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.10.006

526
D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
(1; Fig. 1, Table 1) showed good affinity (Ki A2AAR ¼ 52 nM) and
moderate selectivity at A2AAR especially vs the A1 receptor subtype
(selectivity A1/A2A ¼ 5). Substitution of the bromine atom with a
2-furyl ring in 8-position of this molecule led to 9-ethyl-8-furyl
adenine (2), which showed increased affinity at all ARs and in
particular at the A2A subtype, but with still low selectivity vs the
A1AR (2; Ki A2AAR ¼ 3.7 nM, selectivity A1/A2A ¼ 6; Fig. 1, Table 1)
[28]. In this work, aimed at modulating the affinity and selectivity
at the A2A subtype of compound 2, its 9-ethyl group was replaced
by small alkyl or hydroxyalkyl chains (12e15; Fig. 1, Scheme 1).
The choice of the 9-substituents was made on the base of previ-
ously reported results indicating that the introduction of small
alkyl or hydroxyalkyl chains in 9-position of 8-bromoadenine (8e
11, Scheme 1, Table 1) leads to A2AAR ligands endowed with high
affinity for the same subtype [22]. Among the newly synthesized
8-furyl derivatives 12e15, the 8-(2-furyl)-9-methyladenine (12)
was selected for further investigations since it showed the best
more flexible respect to the 2,2-diphenylethyl group present in
UK-432097, were introduced at N6-position of the selected com-
pound 12 to give the newly trisubstituted adenine derivatives 16e
22 (Fig. 1, Scheme 2).
2. Results and discussion
2.1. Chemistry
Desired compounds 12e22 were obtained as summarized in
Schemes 1 and 2. Alkylation of commercially available adenine (3)
with the suitable alkyl halide in dry dimethylformamide (DMF),
employing potassium carbonate as base, led to a mixture (ratio 3:1)
of the N9- and N7-isomers, 4e7 and 4ae7a, respectively, which
were separated by chromatography (Scheme 1).
Treatment of the N9 isomers 4e7 with N-bromosuccinimide
(NBS) led to the 8-bromo derivatives 8e11 [22] that, after reaction
with tributylstannylfurane in the presence of (PPh3)2PdCl2 as a
catalyst, gave the desired 8-furyl-9-substituted adenine derivatives
12e15.
A
2AAR selectivity profile (Scheme 1, Table 1). Very recently, a
trisubstituted adenosine derivative named UK-432097 (Fig. 1),
bearing a large and lipophilic group at N6-position, has been re-
ported as a potent and selective A2AAR agonist [29]. Starting from
this observation, in a second phase of the present work, large and
lipophilic arylacetyl and arylcarbamoyl substituents, which are
N6-Substituted adenines 16e22 were prepared by reacting 12
with the appropriate isocyanate or acyl chloride in dioxane at reflux
for 1e2 h in the presence of triethylamine (Scheme 2).
Fig. 1. Structure of known and newly synthesized AR ligands.

D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
527
Table 1
Affinity (Ki, nM) of compounds 1, 2, 8e15 in radioligand binding assays at human A1,
2A, and A3 ARs and NECA-stimulated adenylyl cyclase assays at membranes from
stably transfected CHO cells expressing human A2BAR subtype.
A
Scheme 2. Reagents: i: RCOCl, Et3N, dioxane, reflux, 2 h or ReN]C]O, dioxane,
reflux, 1 h.
data of the new molecules 12e15; the adenine derivatives 1, 2 and
8e11 [28] are reported as reference compounds.
Cpd
R
X
Ki A1ARa Ki A2AARb Ki A2BARc Ki A3ARd
1 [28]
2 [28]
8 [28]
12
9 [28]
13
CH2CH3
CH2CH3
CH3
CH3
CH2CH2OH
CH2CH2OH
Br
280
52
3.7
120
7.5
840
375
720
786
2800
470
200
250
5100
350
27,800
4670
Biological results show that the newly synthesised compounds
12e15 are endowed with AR affinity at A1 and A2A subtypes in the
nanomolar range, while they bind the A3 ARs at higher concen-
trations. The analysis of the affinity data of these compounds with
respect to the binding data of the 8-bromo analogues 8e11 [22]
shows that the substitution of a bromine atom with a 2-furyl ring
in 8-position significantly increases the affinity at all AR subtypes,
with low nanomolar binding data obtained at A2AAR (12: Ki
A2AAR ¼ 7.5 nM vs 8: Ki A2AAR ¼ 120 nM; 13: Ki A2AAR ¼ 33 nM vs 9:
2-Furyl 24
Br
2-Furyl 62
Br
2-Furyl 69
Br
2-Furyl 19
570
>100,000
6419
1500
620
33
>100,000
3900
10 [28] CH2CH2CH3
14 CH2CH2CH3
11 [28] CH2CH2CH2OH Br
1100
300
18
ꢀ100,000
790
1000
85
3200
15
CH2CH2CH2OH 2-Furyl 60
15
4300
a
Displacement of specific [3H]CCPA binding in CHO cells transfected with human
recombinant A1AR.
Ki A2AAR
¼
620 nM; 14: Ki A2AAR
¼
18 nM vs 10: Ki
b
Displacement of specific [3H]NECA binding in CHO cells transfected with human
A2AAR ¼ 300 nM; 15: Ki A2AAR ¼ 15 nM vs 11: Ki A2AAR ¼ 85 nM).
An exception to this trend is given by compound 15 that shows an
A3AR affinity comparable to that of the corresponding 8-bromo
derivative 11 (Ki of 4300 and 3200 nM, respectively). These data
confirm our previous observation about the helpful role for AR af-
finity of 2-furyl ring in 8-position of adenine derivatives already
detected in the case of compound 2 in comparison with 1.
Comparing the binding data of compounds 12e15 at the other ARs
versus the A2AAR, it can be observed that the affinity of these
compounds at A3AR results at low micromolar levels, leading to
compound selectivity at A2AAR respect to these AR subtype, in
particular in the case of molecule 12 (A3/A2A ¼ 856). On the other
hand, it must be noted that the selectivity versus A1AR represents
the most critical issue related to 8-furyl adenine derivatives. In fact,
it can be observed that these molecules present good affinity for
A1AR, with compound 14 being the best ligand of the series for this
AR subtype (Ki A1AR ¼ 19 nM). This compound is endowed with
almost identical A1AR and A2AAR affinity values. In general, the
selectivity versus A1AR is low, with the best result obtained with
the 8-furyl-9-methyladenine (12; A1/A2A ¼ 8). Although this com-
pound shows a slight decrease of A2AAR affinity respect to the
previously synthesized 8-furyl derivative 2 (12: Ki A2AAR ¼ 7.5 vs 2:
Ki A2AAR ¼ 3.7), its A2AAR vs A1AR selectivity is slightly improved
(12; A1/A2A ¼ 8 vs 2: A1/A2A ¼ 6). The potency of compounds 12e15
at A2BAR is at high nanomolar concentrations and even in this case
the substitution of 8-bromine with a 2-furyl group leads to an
improvement of potency a part in the case of compounds 12 and 14
(Ki of 786 and 250 nM, respectively) respect to the corresponding 8-
bromo analogues 8 and 10 (Ki of 720 and 200 nM, respectively). In
general the selectivity for A2AAR versus A2BAR (please note that Ki
A2AAR values are related to displacement of specific [3H]NECA
binding while Ki A2BAR values are referred to the inhibition of
NECA-stimulated adenylyl cyclase activity) results improved
respect to the 8-bromo analogues, in particular in the case of
molecule 12 (A2B/A2A ¼ 105).
recombinant A2AAR.
c
Ki values of the inhibition of NECA-stimulated adenylyl cyclase activity in CHO
cells expressing human A2BAR.
d
Displacement of specific [3H]NECA binding in CHO cells transfected with human
recombinant A3AR.
2.2. Biological results and structureeactivity relationships
Newly synthesized compounds 12e22 were evaluated at the
human recombinant adenosine receptors, stably transfected into
Chinese hamster ovary (CHO) cells, utilizing radioligand binding
studies (A1, A2A, and A3 ARs) or adenylyl cyclase activity assay
(A2BAR). Receptor binding affinity was determined using [3H]CCPA
as radioligand for A1AR, whereas [3H]NECA was used for the A2A
and A3 subtypes. In the case of A2BAR, Ki values were calculated
from IC50 values determined by inhibition of NECA-stimulated
adenylyl cyclase activity [30]. The binding and functional data of
the new compounds are shown in Tables 1 and 2. Table 1 shows
Hence, taking in account for each synthesised molecule the
combination of its A2AAR affinity and selectivity versus the other AR
subtypes, compound 12 was selected as starting point for the
preparation of a second series of derivatives bearing a 2-furyl ring
in 8-position, a methyl group in 9-position, and different arylacetyl
Scheme 1. Reagents: i: ReX, DMF, K2CO3, r.t.; ii: NBS, MeCN, MeOH, r.t.; iii: 2-
tributylstannylfurane, (Ph3P)2PdCl2, THF, 70 ꢁC.

528
D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
Table 2
Affinity (Ki, nM) of compounds 16e22 in radioligand binding assays at human A1, A2A, and A3 ARs. NECA-stimulated adenylyl cyclase assays at membranes from stably
transfected CHO cells expressing human A2BAR subtype.
Cpd
R
Ki A1ARa
Ki A2AARb
Ki A2BARc
Ki A3ARd
12
H
62.3
7.5
786
6419
16
17
583
652
110
2510
1952
1270
>30,000
20,285
18
19
3000
59
50
792
193
438
5070
5233
20
21
922
496
89
57
2780
6250
906
23,307
22
756
106
>30,000
37,957
a
Displacement of specific [3H]CCPA binding in CHO cells transfected with human recombinant A1AR.
Displacement of specific [3H]NECA binding in CHO cells transfected with human recombinant A2AAR.
b
c
Ki values of the inhibition of NECA-stimulated adenylyl cyclase activity in CHO cells expressing human A2BAR.
d
Displacement of specific [3H]NECA binding in CHO cells transfected with human recombinant A3AR.
or arylcarbamoyl moieties in N6-position. The obtained compounds
16e22 were then tested at the four human ARs as described above
(Table 2).
A3AR ¼ 1952 nM), while the introduction of a biphenylcarbonyl
moiety (17) in the same position led to a general drop of AR affinity,
especially in the case of A2AAR (170 fold). Introduction of a second
phenyl ring in the benzylic carbon atom of the N6-substituent of
compound 16 led to compound 18 endowed with the highest
selectivity versus the A1AR but, at the same time, with the lowest
selectivity versus A3AR (18: Ki A1AR ¼ 3000 nM; Ki A2AAR ¼ 59 nM;
Ki A3AR ¼ 193 nM). Further modification of compound 16 by
insertion of a benzyloxy group in para position of the N6-
In general the introduction of an acyl moiety at the N6-position
of 8-furyl-9-methyladenine (12) led to a decrease of affinity at all
the four ARs with some exceptions at A3 subtype. In particular, the
introduction of a phenylacetyl group in N6-position led to a
decrease (3e15 fold) of A1 and A2AAR affinity with an improvement
of A3AR binding (16: Ki A1AR ¼ 583 nM; Ki A2AAR ¼ 110 nM; Ki

D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
529
substituent led to compound 19 presenting the highest A2AAR af-
finity of this second series (Ki A2AAR ¼ 50 nM) and a comparable
selectivity versus A1AR respect to the N6-unsubstituted compound
12. The introduction of a methyl group at the para position of
benzyloxy moiety (20) led to a reduction of potency at the A1 and
A
2AARs but an increased affinity for the A3AR. The presence of an
arylcarbamoyl moiety (compounds 21 and 22) reduced the affinity
at the four ARs (5e30 fold) respect to the unsubstituted compound
12. On the other hand compound 21 showed a good affinity at
A
2AAR (21: Ki A2AAR ¼ 57 nM) and selectivity versus the other AR
subtypes, with a slight improvement respect to compound 16 in
particular versus A1AR and A2BAR. The introduction of a methyl
group at the para position of arylcarbamoyl moiety (22) increased
the selectivity versus A2BAR (even in this case, please note that Ki
A2AAR values are related to displacement of specific [3H]NECA
binding while Ki A2BAR values are referred to the inhibition of
NECA-stimulated adenylyl cyclase activity) maintaining at the same
time comparable selectivity versus A1AR and A3AR respect to
compound 21. Considering the antagonist potency at A2BAR, the
introduction of a phenylacetyl group in N6-position led to a
decrease of activity at this AR subtype (16: Ki A2BAR ¼ 2510 nM),
while the presence of a second phenyl ring in the benzylic carbon
atom of the N6-substituent of compound 16 led to a compound with
the best potency data of this second group of derivatives (18: Ki
A2BAR ¼ 792 nM). Insertion of a benzyloxy group or a para-
methylbenzyloxy group in para position of the N6-substituent of
16 led to compounds 19 and 20, respectively, presenting partially
restored potency at A2BAR. The introduction of an arylcarbamoyl
moiety (compounds 21 and 22) did not improved the A2BAR po-
tency leading to compounds selective for A2AAR versus this AR
subtype.
2.3. Molecular modelling
A molecular modelling analysis was performed to better un-
derstand the different interaction of synthesised molecules with
A
2AAR. In particular, molecular docking methods were employed to
simulate the interaction of compounds 12e22 with the binding site
of the three human A2AAR crystal structures solved in complex with
ZM241385 (Fig. 1) antagonist (Protein Data bank e pdb-code:
ꢀ
ꢀ
3EML; 2.6-A resolution [31]; pdb code: 3PWH; 3.3-A resolution
ꢀ
[32]; pdb code: 3VG9; 2.7-A resolution [33]). Compound 1 was also
subjected to this analysis. The choice of the A2AAR structures to be
used for this analysis was based on the fact that the analysed
compounds present several structural analogies with respect to the
co-crystallized ZM241385 antagonist (which can be taken as
“comparison term” in our analysis), like a similar scaffold and the
presence of an exocyclic amino group (in particular considering
compounds 2, 12e15) and of a 2-furyl ring in analogue scaffold
positions. Furthermore, the three A2AAR structures present highly
similar binding sites by considering both pocket volumes and re-
ceptor residues orientation. Nevertheless, rearrangements of some
flexible residues and conformational variability at extracellular
loops (EL) in peripheral regions of binding site are observed. The
most evident residue rearrangement is a different orientation of
E169 side chain that interacts with the ZM241385 exocyclic amino
group in 3EML and 3VG9 structures while it is externally oriented
in 3PWH X-ray data (hence interacting with the “tail segment” of
the same antagonist) (Fig. 2). A first consequence of this variation is
a different conformation of the “tail segment” (the amino-ethyl-
phenol group) of the ZM241385 antagonist, while its triazolo-
triazine scaffold is located and oriented in similar way in the three
crystal structures. A second consequence of E169 rearrangement is
the formation (in the case of 3EML and 3VG9 crystal structures) or
the loss (in the case of 3PWH X-ray) of an H-bond interaction with
Fig. 2. A. Comparison of binding mode of compound ZM241385 to A2AAR as observed
in two published crystal structures (pdb code 3EML in dark, pdb code 3PWH in light);
the different orientation of residues E169 and H264 is highlighted. BeC. Surface rep-
resentation of the A2AAR binding site as observed from the extracellular environment;
3EML and 3VG9 crystal structures (B) are represented in dark and light, respectively,
while 3PWH crystal structure (C) is coloured light; the different orientation of the
above cited E169 and H264 residues allows a different exposure of ligand to extra-
cellular environment.
H264 residue (EL3). In the first case the top of the binding site re-
sults partially obstructed, with the phenol group of ZM241385
antagonist as the sole ligand moiety exposed to the external envi-
ronment (Fig. 2B). In the second case, even the region in proximity
of ZM241385 free amino group is externally exposed (Fig. 2C). On
this base, we decided to employ all three A2AAR-ZM241385 crystal
structures even to verify the effect of this binding site variation on
the binding mode of synthesised molecules. Each crystal structure
was re-modelled by firstly removing eventual external segments

530
D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
(i.e. the T4L segment present in 3EML crystal structure) and sec-
ondly by performing a building of eventual missing receptor re-
gions (i.e. missing sections of EL2 or intracellular 3-IL3-domains).
This procedure allowed also the rebuilding of wild type receptors
from mutated A2AAR structures (i.e. 3PWH crystal structure). The
obtained A2AAR models were checked by using the Protein Ge-
ometry Monitor application within Molecular Operating Environ-
ment (MOE, version 2010.10) suite [34], which provides a variety of
stereochemical measurements for structural quality inspection in a
given protein, such as backbone bond lengths, angles and dihedrals,
Ramachandran 4e
nonbonded contact quality.
2AAR structures were then used as target for the docking
j dihedral plots, and side chain rotamer and
A
analysis of synthesised derivatives. All ligand structures were
optimized using RHF/AM1 semi-empirical calculations and the
MOPAC software package implemented in MOE was utilized for
these calculations [35].
The compounds were then docked into the binding site of the
A
2AAR models by using the MOE Dock tool. Docking poses of each
compound were subjected to energy minimization and then
rescored using three available methods implemented in MOE: the
London dG scoring function that estimates the free energy of
binding of the ligand from a given pose; the Affinity dG scoring tool
that estimates the enthalpic contribution to the free energy of
binding; the dock-pKi predictor that uses the MOE scoring.svl script
to estimate for each ligand a pKi value, which is described by the H-
bonds, transition metal interactions, and hydrophobic interactions
energy. For each compound, the top-score docking pose at each
A
2AAR model, according to at least two out of three scoring func-
tions, was selected for final ligandetarget interaction analysis.
The docking methodology was firstly tested by comparing the
predicted binding mode of ZM241385 within A2AAR pocket respect
to the X-ray data. The docking method showed good ability in
reproducing the ZM241385 binding mode observed in the experi-
mental data. In particular, the Root Mean Square Deviation (RMSD)
from the comparison of the crystal structure and the docking
ꢀ
conformation was calculated as 1.45, 1.33, and 2.73 A (for the 3EML,
3PWH, and 3VG9 crystal structure, respectively), with the best
matching being given by the 2-(furan-2-yl)-[1,2,4]triazolo[1,5-a]
[1,3,5]triazin-7-amine moiety and the highest deviation being
given by a slightly different orientation of the 4-(2-aminoethyl)
phenol group.
Fig. 3. A. Comparison of docking pose of compound 12 (ball and stick style) at 3EML-
based A2AAR structure with the binding mode of compound ZM241385 (stick style) at
the same X-ray. B. Docking pose of compound 15 at 3EML-A2AAR structure.
Considering the general binding mode of N6-unsubstituted
adenine derivatives (2, 12e15), docking results show that the
Considering the 9-substituents, docking results show that these
groups are located in a narrow and mainly hydrophobic region. This
data can explain why the methyl and ethyl substituents (2, 12)
provide good affinity at the A2AAR and why the introduction of a
hydroxyl group (13) causes a significant decrease of potency. When
a hydroxypropyl substituent is present in 9-position (15), the A2AAR
affinity is partially restored and this can be explained by docking
results by the observation that the longer chain than compound 13
allows the hydroxyl function to be located in proximity of two polar
regions given by the H278 side chain and the A59 and A81 carbonyl
groups, at the bottom of the binding site (Fig. 3B).
Results of docking simulations of compounds 16e22 highlight
the effect of E169 rearrangement on ligand binding mode and on
receptoreligand interaction. As described above, the presence or
the loss of a H-bond between E169 and H264 causes a different
availability of space for N6-substituents and this has an evident
effect when bulky moieties are introduced at the adenine amino
group. As example, docking conformations of compound 18 at
3EML and 3VG9 A2AAR structures present the N6-arylacetyl group
in an approximately identical position to the “tail segment” of
ZM241385 reference compound (Fig. 4A). This binding mode causes
a disruption of the H-bonding between N253 and N7 and N6 amino
group of adenine derivative. Docking experiments of the same
compounds share
a similar binding motif inside the trans-
membrane (TM) region of the A2AAR models with the adenine
pyrimidine ring located between TM3 and TM7 domains and the
imidazole ring located between TM3 and TM6 helices. This scaffold
orientation makes the 2-furyl ring in 8-position to be oriented to-
wards the central TM core. This binding mode makes the adenine
moiety location and orientation quite similar with respect to
compound ZM241385 scaffold as observed from the X-ray data
(Fig. 3A). The synthesised compounds show also analogue inter-
action with the receptor respect to the co-crystallised ligand, as the
key receptor residues for the interaction with these derivatives are
in all cases F168 and N253. F168 phenyl ring and adenine scaffold
form a
p-stacking interaction, while N253 amide function is
involved in H-bonding with adenine N6 amine and in an additional
polar interaction with adenine scaffold N7 atom.
The N6-amino group makes also an additional H-bond interac-
tion with E169 residue, but this interaction is lost when docking
simulations at 3PWH-based A2AAR model are considered. As
described above, this residue has a different orientation in 3PWH
crystal structure making impossible a polar interaction with the
free amino group of adenine derivatives 2, 12e15.

D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
531
Fig. 4. A-B. Docking pose of compound 18 at 3EML (A) and 3PWH (B) A2AAR structures. The different amount of space available for N6-substituent is at the base of a slightly different
binding mode and receptor interaction of the compound. CeD. Docking pose of compound 19 at 3EML (C) and 3PWH (D) A2AAR structures. The compound presents similar
conformations but slightly different locations at the two binding sites due to the different orientations of E169 residues. EeF. Docking pose of compound 21 at 3EML (E) and 3PWH
(F) A2AAR structures. The presence of an arylcarbamoyl moiety in N6-position takes to the formation of a H-bond between the amino group of the substituent and the N1 atom of
adenine scaffold.
molecule at 3PWH A2AAR structure show a conservation of this H-
bonding interaction (Fig. 4B). Interestingly, in this case the N6-
arylacetyl group results similarly located and oriented respect to
the analogue 2,2-diphenylethyl group of compound UK-432097 as
observed from another recently reported A2AAR crystal structure
[36] that presents a E169 side chain orientation similar to 3PWH
X-ray.
As expected, the difference in binding mode to the three A2AAR
structures is related also to the rigidity of the N6-substituent.
Compounds 16, 19, 20 present similar docking conformations at the
three receptor pockets due to some flexibility of the arylacetyl
substituent, even if a slightly different position of compounds is
observed if the respective binding modes at 3EML/3VG9 and 3PWH
A2AAR structures are compared (Fig. 4CeD). On the other hand,
compounds 17, 21, 22 contain rigid groups that cause different
behaviour during docking experiments. In particular the presence
of an arylcarbamoyl moiety in N6-position (i.e. compound 21, Fig. 4)
takes to the formation of a H-bond between the amino group of the
substituent and the N1 atom of adenine scaffold, hence providing
higher rigidity respect to the analogue arylacetyl substituted
compounds. The rigidity of N6-substituent makes even more
evident the different binding mode of these compounds at the
3EML/3VG9 and 3PWH A2AAR pockets. At the same time, the higher
hydrophilicity of arylcarbamoyl substituents respect to the
analogue arylacetyl group and the fact that the residue E169
(located in close proximity to the compound N6-position) is

532
D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
substituted in A3AR by a valine could explain the higher selectivity
versus A3AR of the arylcarbamoyl substituted molecules.
The results of these docking experiments suggest that there can
be some flexibility at A2AAR binding site that cause ligand rear-
rangement on the basis of different orientation of binding site
residues. The docking simulation at different receptor conforma-
tions can help to interpret the ligand ability to adapt itself to
different conformations of the binding pocket and hence the sta-
bility of ligandereceptor interaction.
4.2. General procedure for the preparation of compounds 12e15
To a solution of the appropriate bromine derivative 8e11
(1 mmol) in dry tetrahydrofuran (THF) (6 mL), (PhP3)2PdCl2
(0.06 mmol) and 2-tributylstannylfuran (5 mmol) were added and
the resulting mixtures were stirred under reaction conditions
described for each compound. Then the solvent was removed under
reduced pressure and the crude residue purified by chromatog-
raphy with the suitable mixture of solvents to obtain the desired
compounds (12e15) as solids.
3. Conclusions
4.2.1. 8-(Furan-2-yl)-9-methyl-9H-purin-6-amine (12)
Results from the biological assays reported in Tables 1 and 2
show that the synthesised 9-substituted 8-(2-furyl)-adenine de-
rivatives are endowed with good affinity at the human A2AAR and
with higher potency respect to the corresponding 8-bromo de-
rivatives previously synthesised and reported in literature. In
particular, 8-(2-furyl)-9-methyladenine (12) shows low nanomolar
affinity at this AR subtype and high selectivity in particular versus
A2B and A3 ARs, with an improvement of selectivity versus A1AR
respect to the previously reported 9-ethyl analogue 1. Further
modification of compound 12 with introduction of arylacetyl or
arylcarbamoyl groups in N6-position did not take to an improve-
ment of A2AAR affinity and had different impacts on the selectivity
versus the other three adenosine receptor subtypes. In particular,
the presence of an acyl moiety at the N6-position led to a general
decrease of affinity at the four ARs with respect to the unsub-
stituted derivative 12 with the exception of compounds 18 and 20
that present a significant improvement of A3AR binding data. On
the other hand, the introduction of an arylcarbamoyl moiety in N6
position of 12 (compounds 21 and 22) restored the A2AAR selec-
tivity profile observed for the N6-unsubstituted compound. The
comparative analysis of docking studies at the three different
Compound 12 was obtained from 8 by reaction at 60 ꢁC, over-
night. The crude mixture has been purified through silica gel
normal column chromatography eluting with CHCl3ecC6H12e
MeOH (65:35:5, v/v). Yield 65%; white solid, m.p.: 220 ꢁC (dec). IR
(KBr): 3285e2960, 1610, 1590, 1500, 1410 cmꢃ1 1H NMR (DMSO-
;
d6):
d
3.89 (s, 3H, CH3), 6.75 (t, J ¼ 1.6 Hz, 1H, H-40), 7.18 (dd, J ¼ 0.8
and 3.6 Hz, 1H, H-30), 7.32 (bs, 2H, NH2), 7.96 (m, 1H, H-50), 8.14 (s,
1H, H-2). Anal. Calcd. for C10H9N5O (M.W.: 215.21): C, 55.81; H,
4.22; N, 32.54. Found: C, 56.01; H, 4.24; N, 32.66.
4.2.2. 2-(6-Amino-8-(furan-2-yl)-9H-purin-9-yl)ethanol (13)
Compound 13 was obtained from 9 by reaction at 60 ꢁC for 20 h.
The crude mixture has been purified through silica gel normal
column chromatography eluting with CHCl3ecC6H12eMeOH
(70:25:5, v/v). Yield 87%; white solid (CHCl3ecC6H12eMeOH) m.p.:
dec. 245 ꢁC. IR (KBr): 3281e2924, 1667, 1615, 1589, 1572, 1506,
1466 cmꢃ1 1H NMR (DMSO-d6)
; d 3.74 (m, 2H, CH2eO), 4.46 (t,
J ¼ 6.0 Hz, 2H, CH2eN), 5.01 (t, J ¼ 5.7 Hz, 1H, OH), 6.76 (m, 1H, H-
40), 7.22 (m, 1H, H-30), 7.35 (bs, 2H, NH2), 7.97 (m, 1H, H-50), 8.16 (s,
1H, H-2). Anal. Calcd. for C11H11N5O2 (M.W.: 245.24): C, 53.87; H,
4.52; N, 28.56. Found: C, 53.99; H, 4.45; N, 28.66.
A
2AAR crystal structures suggests that slight rearrangements are
possible at the receptor binding site and that the ability of ligand to
adapt its conformation to the binding site residues could be related
to the stability of ligandereceptor interaction and hence to the
binding affinity. In this sense, while the general binding mode of
synthesised compounds at the receptor pocket is conserved at the
three receptor structures, on the other hand the impact of the
different N6-substituents can be more easily interpreted by
comparing the results at the different A2AAR binding sites.
4.2.3. 8-(Furan-2-yl)-9-propyl-9H-purin-6-amine (14)
Compound 14 was obtained from 10 by reaction at 60 ꢁC for
30 min. The crude mixture has been purified through silica gel
normal column chromatography eluting with cC6H12eCHCl3e
MeOH (60:35:5, v/v) to give final compound 14. Yield 50%; white
solid (MeOH) m.p.:193e195 ꢁC; IR (KBr): 3308, 3122, 2923, 1661,
1622, 1589, 1570, 1518, 1469 cmꢃ1 1H NMR (DMSO-d6):
; d 0.86 (t,
J ¼ 7.5, 3H, CH3), 1.76 (m, 2H, NCH2CH2), 4.39 (t, J ¼ 7.3, 2H NCH2),
6.77 (m,1H, H-40), 7.17 (d, J ¼ 3.7,1H, H-30), 7.35 (bs, 2H, NH2), 7.99 (d,
J ¼ 1.0, 1H, H-50), 8.17 (s, 1H, H-2). Anal. Calcd. for C12H13N5O (M.W.:
243.26): C, 59.25; H, 5.39; N, 28.79.Found: C, 59.33; H, 5.24; N, 28.85.
4. Experimental section
4.1. Chemistry
4.2.4. 3-(6-Amino-8-(furan-2-yl)-9H-purin-9-yl)propan-1-ol (15)
Compound 15 was obtained from 11 by reaction at r.t. overnight.
The crude mixture has been purified through silica gel normal
column chromatography eluting with CHCl3eMeOH (96:4, v/v) to
give final compound 15. Yield 85%; white solid (MeOH) m.p.: 203e
205 ꢁC; IR (KBr): 3320e2924, 1662, 1614, 1594, 1574, 1499,
General: Reactions were routinely monitored by thin-layer
chromatography (TLC) on silica gel (precoated F254 Merck plates)
and products visualized with UV light
l
¼ 254 nm and iodine or
potassium permanganate solution. Infrared (IR) spectra were
measured on a Perkin Elmer 257 instrument. 1H NMR were deter-
mined in DMSO-d6 solutions with a Bruker AC 200 spectrometer,
1458 cmꢃ1; 1H NMR (DMSO-d6):
d
0.86 (t, J ¼ 7.5 Hz, 3H, CH3), 1.90
peaks positions are given in parts per million (
d
) downfield from
(m, 2H, CH2CH2CH2), 3.46 (m, 2H, CH2O), 4.45 (t, J ¼ 7.2 Hz, 2H Ne
CH2), 4.72 (t, J ¼ 5.4 Hz, 1H, OH), 6.77 (m, 1H, H-40), 7.21 (d,
J ¼ 3.3 Hz, 1H, H-30), 7.39 (bs, 2H, NH2), 7.98 (d, J ¼ 1.2 Hz, 1H, H-50),
8.17 (s, 1H, H-2). Anal. Calcd. for C12H13N5O2 (M.W.: 259.26): C,
55.59; H, 5.05; N, 27.01. Found: C, 56.11; H, 4.94; N, 27.03.
tetramethylsilane as internal standard, and J values are given in Hz.
Light petroleum ether refers to the fractions boiling at 40e60 ꢁC.
Melting points were determined on a Buchi-Tottoli instrument and
are uncorrected. Chromatographies were performed using Merck
60e200 mesh silica gel. All reported products showed IR and 1H
NMR spectra in agreement with the assigned structures. Organic
solutions were dried over anhydrous magnesium sulphate.
Elemental analyses were performed by the microanalytical labo-
ratory of Department of Chemical and Pharmaceutical Sciences,
University of Trieste, and they were within ꢂ0.4% of the theoretical
values for C, H, and N.
4.3. General procedure for the preparation of compounds 16e20
To a solution of amino compound 12 (50 mg, 0.232 mmol) in dry
dioxane (5 mL) the appropriate acyl chloride (1.39 mmol) and
triethylamine (32
mL, 0.232 mmol) were added and stirred at reflux
for 2 h. Then the solvent was removed in vacuo and the residue

D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
533
purified by flash chromatography eluting with EtOAcelight petro-
leum (7:3, v/v) to afford the desired compounds (16e20) as a solid.
stirred at reflux for 1 h. Then the solvent was removed in vacuum
and the residue purified by flash chromatography eluting with
EtOAcelight petroleum (7:3, v/v) to afford the desired compound
21 or 22 as a solid.
4.3.1. N-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-2-
phenylacetamide (16)
Compound 16 has been prepared by reaction of 12 with 2-
phenylacetyl chloride. Yield 73%; pale yellow solid (EtOAc-light
petroleum) m.p.: 185 ꢁC; IR (KBr): 3325e2965, 1680, 1600, 1505,
4.4.1. 1-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-3-phenylurea
(21)
Compound 21 has been prepared by reaction of 12 with phenyl
isocyanate. Yield 65%; white solid (EtOAcelight petroleum) m.p.:
220 ꢁC; IR (KBr): 3335e2940, 1690, 1610, 1520, 1430 cmꢃ1; 1H NMR
1415 cmꢃ1 1H NMR (DMSO-d6):
; d 3.92 (s, 2H, CH2), 4.01 (s, 3H,
CH3), 6.83 (dd, 1H, J ¼ 1.6 Hz, J ¼ 3.4 Hz, H-40), 7.26e7.40 (m, 6H, H-
30, H-Ph), 8.07 (d, 1H, J ¼ 1.6 Hz, H-50), 8.64 (s, 1H, H-2), 10.99 (s, 1H,
NH). Anal. Calcd. for C18H15N5O2 (M.W.: 333.34): C, 64.86; H, 4.54;
N, 21.01. Found: C, 64.70; H, 4.47; N, 21.08.
(DMSO-d6):
d
4.02 (s, 3H, CH3), 6.84 (dd, 1H, J ¼ 1.8 Hz, J ¼ 3.6 Hz, H-
40), 7.08 (t, 1H, J ¼ 7.6 Hz, H-Ph), 7.32e7.40 (m, 3H, H-30, H-Ph), 7.63
(d, 2H, J ¼ 8.4 Hz, H-Ph), 8.09 (d, 1H, J ¼ 1.8 Hz, H-50), 8.68 (s, 1H, H-
2), 10.09 (s, 1H, NH), 11.78 (s, 1H, NH). Anal. Calcd. for C17H14N6O2
(M.W.: 334.33): C, 61.07; H, 4.22; N, 25.14. Found: C, 60.87; H, 4.05;
N, 25.27.
4.3.2. N-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-[1,10-biphenyl]-
4-carboxamide (17)
Compound 17 has been prepared by reaction of 12 with [1,10-
biphenyl]-4-carbonyl chloride. Yield 64%; dark yellow solid (EtOAce
light petroleum) m.p.: 205 ꢁC; IR (KBr): 3335e2980, 1680, 1615,
4.4.2. 1-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-3-(p-tolyl)urea
(22)
Compound 22 has been prepared by reaction of 12 with p-tolyl
isocyanate. Yield 71%; white solid (EtOAcelight petroleum) m.p.:
212 ꢁC; IR (KBr): 3330e2945, 1690, 1620, 1515, 1410 cmꢃ1; 1H NMR
1510, 1400 cmꢃ1; 1H NMR (DMSO-d6):
d
4.05 (s, 3H, CH3), 6.82 (dd,
1H, J ¼ 1.6 Hz, J ¼ 3.4 Hz, H-40), 7.40 (d,1H, J ¼ 3.4 Hz, H-30), 7.44e7.56
(m, 3H, H-Ph), 7.80 (d, 2H, J ¼ 6.8 Hz, H-Ph), 7.88 (d, 2H, J ¼ 8.2 Hz, H-
Ph), 8.07 (d,1H, J ¼ 1.6, H-50), 8.17 (d, 2H, J ¼ 8.2, H-Ph), 8.76 (s,1H, H-
2), 11.26 (s, 1H, NH). Anal. Calcd. for C23H17N5O2 (M.W.: 395.41): C,
69.86; H, 4.33; N, 17.71. Found: C, 69.57; H, 4.40; N, 17.78.
(DMSO-d6):
d 2.28 (s, 3H, CH3), 4.01 (s, 3H, NCH3), 6.82 (dd, 1H,
J ¼ 1.6 Hz, J ¼ 3.4 Hz, H-40), 7.16 (d, 2H, J ¼ 8.2 Hz, H-Ph), 7.37 (d, 1H,
J ¼ 3.4 Hz, H-30), 7.51 (d, 2H, J ¼ 8.2 Hz, H-Ph), 8.07 (d, 1H, J ¼ 1.6 Hz,
H-50), 8.66 (s, 1H, H-2), 9.99 (s, 1H, NH), 11.70 (s, 1H, NH). Anal.
Calcd. for C18H16N6O2 (M.W.: 348.36): C, 62.06; H, 4.63; N, 24.12.
Found: C, 62.28; H, 4.55; N, 24.15.
4.3.3. N-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-2,2-
diphenylacetamide (18)
Compound 18 has been prepared by reaction of 12 with 2,2-
diphenylacetyl chloride. Yield 70%; yellow solid (EtOAcelight pe-
troleum) m.p.: 130 ꢁC; IR (KBr): 3330e2975, 1685, 1610, 1515,
4.5. Biological assay
1410 cmꢃ1 1H NMR (DMSO-d6):
; d 4.01 (s, 3H, CH3), 5.67 (s, 1H,
All pharmacological methods followed the procedures as
described earlier [30]. In brief, membranes for radioligand binding
were prepared from CHO cells stably transfected with human
adenosine receptor subtypes in a two-step procedure. In a first low-
speed step (1000ꢄg) cell fragments and nuclei were removed. The
crude membrane fraction was sedimented from the supernatant at
100,000ꢄg. The membrane pellet was resuspended in the buffer
used for the respective binding experiments, frozen in liquid ni-
trogen and stored at ꢃ80 ꢁC. For the measurement of adenylyl
cyclase activity only one high speed centrifugation of the homog-
enate was used. The resulting crude membrane pellet was resus-
pended in 50 mM Tris/HCl, pH 7.4 and immediately used for the
cyclase assay.
COCH), 6.83 (dd, 1H, J ¼ 1.8 Hz, J ¼ 3.4 Hz, H-40), 7.26e7.39 (m, 11H,
H-30, H-Ph), 8.08 (d, 1H, J ¼ 1.8 Hz, H-50), 8.65 (s, 1H, H-2), 11.25 (s,
1H, NH). Anal. Calcd. for C24H19N5O2 (M.W.: 409.44): C, 70.40; H,
4.68; N, 17.10. Found: C, 70.26; H, 4.54; N, 17.19.
4.3.4. 2-(4-(Benzyloxy)phenyl)-N-(8-(furan-2-yl)-9-methyl-9H-
purin-6-yl)acetamide (19)
Compound 19 has been prepared by reaction of 12 with 2-(4-
(benzyloxy)phenyl)acetyl chloride. Yield 72%; yellow solid (EtOAce
light petroleum) m.p.: ¼ 110 ꢁC; IR (KBr): 3330e2970, 1680, 1625,
1530, 1420 cmꢃ1 1H NMR (DMSO-d6):
; d 3.85 (s, 2H, CH2), 4.03 (s,
3H, CH3), 5.16 (s, 2H, OCH2), 6.85 (dd,1H, J ¼ 1.8 Hz, J ¼ 3.4 Hz, H-40),
7.12 (d, 2H, J ¼ 8.6 Hz, H-Ph), 7.35e7.44 (m, 8H, H-30, H-Ph), 8.10 (d,
1H, J ¼ 1.8 Hz, H-50), 8.64 (s, 1H, H-2), 10.91 (s, 1H, NH). Anal. Calcd.
for C25H21N5O3 (M.W.: 439.47): C, 68.33; H, 4.82; N,15.94. Found: C,
68.54; H, 4.87; N, 15.88.
For radioligand binding at A1AR 1 nM [3H]CCPA was used,
whereas 30 and 10 nM [3H]NECA were used for A2A and A3 ARs,
respectively. Non-specific binding of [3H]CCPA was determined in
the presence of 1 mM theophylline, in the case of [3H]NECA 100
mM
R-PIA was used. Ki-values from competition experiments were
calculated with the program SCTFIT [37].
4.3.5. N-(8-(Furan-2-yl)-9-methyl-9H-purin-6-yl)-2-(4-((4-
methylbenzyl)oxy)phenyl)acetamide (20)
Inhibition of NECA-stimulated adenylyl cyclase activity was
Compound 20 has been prepared by reaction of 12 with 2-(4-((4-
methylbenzyl)oxy)phenyl)acetyl chloride. Yield 76%; pale yellow
solid (EtOAc-light petroleum) m.p.: 100 ꢁC; IR (KBr): 3320e2975,
determined as a measurement of affinity of compounds. EC50
-
values from these experiments were converted to Ki-values with
the Cheng and Prusoff equation [38].
1683, 1620, 1520, 1410 cmꢃ1; 1H NMR (DMSO-d6):
d 2.29 (s, 3H, CH3),
3.83 (s, 2H, COCH2), 4.01 (s, 3H, NCH3), 5.03 (s, 2H, OCH2), 6.82 (dd,1H,
J ¼ 1.8 Hz, J ¼ 3.6 Hz, H-40), 6.95 (d, 2H, J ¼ 8.6 Hz, H-Ph), 7.16e7.34 (m,
6H, H-Ph), 7.38 (d, 1H, J ¼ 3.6 Hz, H-30), 8.07 (d, 1H, J ¼ 1.8 Hz, H-50),
8.63 (s,1H, H-2),10.94 (s,1H, NH). Anal. Calcd. for C26H23N5O3 (M.W.:
453.49): C, 68.86; H, 5.11; N, 15.44. Found: C, 68.71; H, 5.14; N, 15.29.
4.6. Molecular modelling
All molecular modelling studies were performed on a 2 CPU
(PIV 2.0e3.0 GHz) Linux PC. Homology modelling, energy mini-
mization, and docking studies were carried out using MOE
(version 2010.10) suite [34]. All ligand structures were optimized
using RHF/AM1 semiempirical calculations and the software
package MOPAC implemented in MOE was utilized for these cal-
culations [35].
4.4. General procedure for the preparation of compounds 21 and 22
A solution of compound 12 (50 mg, 0.232 mmol) in dry dioxane
(5 mL) and the appropriate isocyanate (0.697 mmol, 3 eq) was

534
D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
4.6.1. Refinement of the human A2AAR structures
retrievable at the SVL exchange service (Chemical Computing
Group, Inc. SVL exchange: http://svl.chemcomp.com). The algo-
rithm is based on an empirical scoring function consisting of a
directional hydrogen-bonding term, a directional hydrophobic
interaction term, and an entropic term (ligand rotatable bonds
immobilized in binding). The obtained pKi values must be consid-
ered as docking scores and not as prediction of binding affinity. For
each compound, the three top-score docking poses according to at
least two out of three scoring functions were selected for final
ligandetarget interaction analysis.
The crystal structures of the hA2AAR in complex with ZM241385
were retrieved from the Protein Data Bank (http://www.rcsb.org;
ꢀ
ꢀ
pdb code: 3EML; 2.6-A resolution [31]; pdb code: 3PWH; 3.3-A
ꢀ
resolution [32]; pdb code: 3VG9; 2.7-A resolution [33]).
Each structure was re-modelled by firstly removing eventual
external segments (i.e. the T4L segment present in 3EML crystal
structure) and secondly by performing a building of eventual
missing receptor regions (i.e. missing sections of EL2 or IL3 do-
mains). This procedure allowed also the rebuilding of wild type
receptors from mutated hA2AAR structures (i.e. 3PWH crystal
structure). The Homology Modelling tool of MOE was employed. In
details, for each hA2AAR model, the boundaries identified from the
used hA2AAR X-ray crystal structure were applied. The missing loop
domains were built by the loop search method implemented in
MOE. Once the heavy atoms were modelled, all hydrogen atoms
were added, and the protein coordinates were then minimized
with MOE using the AMBER99 force field [39]. The minimizations
were performed by 1000 steps of steepest descent followed by
Acknowledgements
This work was supported by Fondo di Ricerca di Ateneo (Uni-
versity of Camerino) and by a grant of the Italian Ministry for
University and Research (PRIN2010-11 nꢁ 20103W4779_003).
References
conjugate gradient minimization until the RMS gradient of the
[1] K.A. Jacobson, Z.G. Gao, Adenosine receptors as therapeutic targets, Nat. Rev.
Drug Discov. 5 (2006) 247e264.
[2] A.S. Robeva, R.L. Woodard, X. Jin, Z. Gao, S. Bhattacharya, H.E. Taylor,
D.L. Rosin, J. Linden, Molecular characterization of recombinant human
adenosine receptors, Drug Dev. Res. 39 (1996) 243e252.
[3] B.B. Fredholm, G. Arslan, L. Halldner, B. Kull, G. Schulte, W. Wasserman,
Structure and function of adenosine receptors and their genes, Naunyn-
Schmiedeberg’s Arch. Pharmacol. 362 (2000) 364e374.
[4] B.B. Fredholm, A.P. IJzerman, K.A. Jacobson, J. Linden, C.E. Muller, International
Union of Basic and Clinical Pharmacology. LXXXI. Nomenclature and classifi-
cation of adenosine receptorsean update, Pharmacol. Rev. 63 (2011) 1e34.
[5] J.M. Downey, M.V. Cohen, K. Ytrehus, Y. Liu, Cellular mechanisms in ischemic
preconditioning: the role of adenosine and protein kinase C, Ann. N.Y. Acad.
Sci. 723 (1994) 82e98.
ꢀꢃ1
potential energy was less than 0.05 kJ molꢃ1
A . Reliability and
quality of these models were checked using the Protein Geometry
Monitor application within MOE, which provides a variety of ste-
reochemical measurements for inspection of the structural quality
in a given protein, like backbone bond lengths, angles and di-
hedrals, Ramachandran 4ej dihedral plots, and sidechain rotamer
and nonbonded contact quality.
4.6.2. Molecular docking analysis
All compound structures were docked into the binding site of
the three hA2AAR models using the MOE Dock tool. This method is
divided into a number of stages: Conformational analysis of ligands.
The algorithm generated conformations from a single 3D confor-
mation by conducting a systematic search. In this way, all combi-
nations of angles were created for each ligand. Placement. A
collection of poses was generated from the pool of ligand confor-
mations using Triangle Matcher placement method. Poses were
generated by superposition of ligand atom triplets and triplet
points in the receptor binding site. The receptor site points are
alpha sphere centres which represent locations of tight packing. At
each iteration a random conformation was selected, a random
triplet of ligand atoms and a random triplet of alpha sphere centres
were used to determine the pose. Scoring. Poses generated by the
placement methodology were scored using two available methods
implemented in MOE, the London dG scoring function which esti-
mates the free energy of binding of the ligand from a given pose,
and Affinity dG scoring which estimates the enthalpy contribution
to the free energy of binding. The top 30 poses for each ligand were
output in a MOE database.
[6] J.A. Auchampach, X. Jin, T.C. Wan, G.H. Caughey, J. Linden, Canine mast cell
adenosine receptors: cloning and expression of the A3 receptor and evidence
that degranulation is mediated by the A2B receptor, Mol. Pharmacol. 52 (1997)
846e860.
[7] G. Cristalli, R. Volpini, Adenosine receptors: chemistry and pharmacology,
Curr. Top. Med. Chem. 3 (2003) 355e469.
[8] K.-N. Klotz, Adenosine receptors and their ligands, Naunyn-Schmiedeberg’s
Arch. Pharmacol. 362 (2000) 382e391.
[9] C.E. Muller, B. Stein, Adenosine receptor antagonists: structures and potential
therapeutic applications, Curr. Pharm. Des. 2 (1996) 501e530.
[10] S.A. Poulsen, R.J. Quinn, Adenosine receptors: new opportunities for future
drugs, Bioorg. Med. Chem. 6 (1998) 619e641.
[11] S. Hess, Recent advances in adenosine receptor antagonist research, Expert
Opin. Ther. Pat. 11 (2001) 1533e1561.
[12] Kyowa Hakko Kirin, Approval for Manufacturing and Marketing of
NOURIASTÒ Tablets 20 mg, A Novel Antiparkinsonian Agent, 2013. http://
www.kyowa-kirin.com/news_releases/2013/e20130325_04.html.
[13] S.L. Cheong, G. Venkatesan, P. Paira, R. Jothibasu, A.L. Mandel, S. Federico,
G. Spalluto, G. Pastorin, Pyrazolo derivatives as potent adenosine receptor
antagonists: an overview on the structureeactivity relationships, Int. J. Med.
Chem. 2011 (2011). ID 480652.
[14] S. Moro, Z.G. Gao, K.A. Jacobson, G. Spalluto, Progress in the pursuit of ther-
apeutic adenosine receptor antagonists, Med. Res. Rev. 26 (2006) 131e159.
[15] G. Cristalli, E. Camaioni, S. Costanzi, S. Vittori, R. Volpini, K.-N. Klotz, Charac-
terization of potent ligands at human recombinant adenosine receptors, Drug
Dev. Res. 45 (1998) 176e181.
[16] E. Camaioni, S. Costanzi, S. Vittori, R. Volpini, K.-N. Klotz, G. Cristalli, New
substituted 9-alkylpurines as adenosine receptor ligands, Bioorg. Med. Chem.
6 (1998) 523e533.
4.6.3. Post docking analysis
The docking poses of each compound were then subjected to
[17] K.-N. Klotz, S. Kachler, C. Lambertucci, S. Vittori, R. Volpini, G. Cristalli, 9-
Ethyladenine derivatives as adenosine receptor antagonists: 2- and 8-
substitution results in distinct selectivities, Naunyn-Schmiedeberg’s Arch.
Pharmacol. 367 (2003) 629e634.
[18] R. Volpini, S. Costanzi, S. Vittori, G. Cristalli, K.-N. Klotz, Medicinal chemistry
and pharmacology of A2B adenosine receptors, Curr. Top. Med. Chem. 3 (2003)
427e443.
[19] R. Volpini, S. Costanzi, C. Lambertucci, S. Vittori, C. Martini, M.L. Trincavelli, K.-
N. Klotz, G. Cristalli, 2- and 8-alkynyl-9-ethyladenines: synthesis and bio-
logical activity at human and rat adenosine receptors, Purinergic Signal. 1
(2005) 173e181.
[20] C. Lambertucci, S. Vittori, R.C. Mishra, D. Dal Ben, K.-N. Klotz, R. Volpini,
G. Cristalli, Synthesis and biological activity of trisubstituted adenines as A2A
adenosine receptor antagonists, Nucleosides Nucleotides Nucleic Acids 26
(2007) 1443e1446.
AMBER99 force field energy minimization until the RMS gradient of
ꢀꢃ1
the potential energy was less than 0.05 kJ molꢃ1
A . Receptor
ꢀ
residues within 6 A distance from the ligand were left free to move,
while the remaining receptor coordinates were kept fixed.
AMBER99 partial charges of receptor and MOPAC output partial
charges of ligands were utilized. Once the compound-binding site
energy minimization was completed, receptor coordinates were
fixed and a second energy minimization stage was performed
leaving free to move only compound atoms. MMFF94 force field
[40e46] was applied. For each compound, the minimized docking
poses were then rescored using London dG and Affinity dG scoring
functions and the dock-pKi predictor. The latter tool allows esti-
mating the pKi for each ligand using the “scoring.svl” script
[21] C. Lambertucci, G. Cristalli, D. Dal Ben, D.D. Kachare, C. Bolcato, K.-N. Klotz,
G. Spalluto, R. Volpini, New 2,6,9-trisubstituted adenines as adenosine

D. Dal Ben et al. / European Journal of Medicinal Chemistry 70 (2013) 525e535
535
receptor antagonists: a preliminary SAR profile, Purinergic Signal. 3 (2007)
339e346.
[22] C. Lambertucci, I. Antonini, M. Buccioni, D. Dal Ben, D.D. Kachare, R. Volpini,
K.-N. Klotz, G. Cristalli, 8-Bromo-9-alkyl adenine derivatives as tools for
developing new adenosine A2A and A2B receptors ligands, Bioorg. Med. Chem.
17 (2009) 2812e2822.
[23] H. Harada, O. Asano, Y. Hoshino, S. Yoshikawa, M. Matsukura, Y. Kabasawa,
J. Niijima, Y. Kotake, N. Watanabe, T. Kawata, T. Inoue, T. Horizoe, N. Yasuda,
H. Minami, K. Nagata, M. Murakami, J. Nagaoka, S. Kobayashi, I. Tanaka, S. Abe,
2-Alkynyl-8-aryl-9-methyladenines as novel adenosine receptor antagonists:
their synthesis and structureeactivity relationships toward hepatic glucose
production induced via agonism of the A2B receptor, J. Med. Chem. 44 (2001)
170e179.
[32] A.S. Doré, N. Robertson, J.C. Errey, I. Ng, K. Hollenstein, B. Tehan, E. Hurrell,
K. Bennett, M. Congreve, F. Magnani, C.G. Tate, M. Weir, F.H. Marshall,
Structure of the adenosine A2A receptor in complex with ZM241385 and the
xanthines XAC and Caffeine, Structure 19 (2011) 1283e1293.
[33] T. Hino, T. Arakawa, H. Iwanari, T. Yurugi-Kobayashi, C. Ikeda-Suno,
Y. Nakada-Nakura, O. Kusano-Arai, S. Weyand, T. Shimamura, N. Nomura,
A.D. Cameron, T. Kobayashi, T. Hamakubo, S. Iwata, T. Murata, G-protein-
coupled receptor inactivation by an allosteric inverse-agonist antibody, Na-
ture 482 (2012) 237e240.
[34] Molecular Operating Environment, C.C.G., Inc., 1255 University St., Suite 1600,
Montreal, Quebec, Canada, H3B 3X3.
[35] J.J. Stewart, MOPAC: a semiempirical molecular orbital program, J. Comput.
Aided Mol. Des. 4 (1990) 1e105.
[24] R.A. de Ligt, P.A. van der Klein, J.K. von Frijtag Drabbe Kunzel, A. Lorenzen,
F. Ait El Maate, S. Fujikawa, R. van Westhoven, T. van den Hoven,
J. Brussee, A.P. IJzerman, Synthesis and biological evaluation of disubsti-
tuted N6-cyclopentyladenine analogues: the search for a neutral antagonist
with high affinity for the adenosine A1 receptor, Bioorg. Med. Chem. 12
(2004) 139e149.
[25] M. Perreira, J.K. Jiang, A.M. Klutz, Z.G. Gao, A. Shainberg, C. Lu, C.J. Thomas,
K.A. Jacobson, “Reversine” and its 2-substituted adenine derivatives as potent
and selective A3 adenosine receptor antagonists, J. Med. Chem. 48 (2005)
4910e4918.
[26] P. Minetti, M.O. Tinti, P. Carminati, M. Castorina, M.A. Di Cesare, S. Di Serio,
G. Gallo, O. Ghirardi, F. Giorgi, L. Giorgi, G. Piersanti, F. Bartoccini, G. Tarzia, 2-
n-Butyl-9-methyl-8-[1,2,3]triazol-2-yl-9H-purin-6-ylamine and analogues as
A2A adenosine receptor antagonists. Design, synthesis, and pharmacological
characterization, J. Med. Chem. 48 (2005) 6887e6896.
[27] A. Pinna, R. Volpini, G. Cristalli, M. Morelli, New adenosine A2A receptor an-
tagonists: actions on Parkinson’s disease models, Eur. J. Pharmacol. 512 (2005)
157e164.
[28] R. Volpini, D. Dal Ben, C. Lambertucci, G. Marucci, R.C. Mishra, A.T. Ramadori,
K.-N. Klotz, M.L. Trincavelli, C. Martini, G. Cristalli, Adenosine A2A receptor
antagonists: new 8-substituted 9-ethyladenines as tools for in vivo rat models
of Parkinson’s disease, Chem. Med. Chem. 4 (2009) 1010e1019.
[29] C.E. Muller, K.A. Jacobson, Recent developments in adenosine receptor ligands
and their potential as novel drugs, Biochim. Biophys. Acta 1808 (2011) 1290e
1308.
[30] K.-N. Klotz, J. Hessling, J. Hegler, C. Owman, B. Kull, B.B. Fredholm, M.J. Lohse,
Comparative pharmacology of human adenosine receptor subtypes - char-
acterization of stably transfected receptors in CHO cells, Naunyn-Schmiede-
berg’s Arch. Pharmacol. 357 (1998) 1e9.
[36] F. Xu, H. Wu, V. Katritch, G.W. Han, K.A. Jacobson, Z.G. Gao, V. Cherezov,
R.C. Stevens, Structure of an agonist-bound human A2A adenosine receptor,
Science 332 (2011) 322e327.
[37] A. De Lean, A.A. Hancock, R.J. Lefkowitz, Validation and statistical analysis of a
computer modeling method for quantitative analysis of radioligand binding
data for mixtures of pharmacological receptor subtypes, Mol. Pharmacol. 21
(1982) 5e16.
[38] Y. Cheng, W.H. Prusoff, Relationship between the inhibition constant (Ki) and
the concentration of inhibitor which causes 50 per cent inhibition (I50) of an
enzymatic reaction, Biochem. Pharmacol. 22 (1973) 3099e3108.
[39] W.D. Cornell, P. Cieplak, C.I. Bayly, I.R. Gould, K.M. Merz, D.M. Ferguson,
D.C. Spellmeyer, T. Fox, J.W. Caldwell, P.A. Kollman, A second generation force
field for the simulation of proteins, nucleic acids, and organic molecules, J. Am.
Chem. Soc. 117 (1995) 5179e5197.
[40] T.A. Halgren, Merck molecular force field. I. Basis, form, scope, parameteri-
zation, and performance of MMFF94, J. Comput. Chem. 17 (1996) 490e519.
[41] T.A. Halgren, Merck molecular force field. II. MMFF94 van der Waals and
electrostatic parameters for intermolecular interactions, J. Comput. Chem. 17
(1996) 520e552.
[42] T.A. Halgren, Merck molecular force field. III. Molecular geometries and
vibrational frequencies for MMFF94, J. Comput. Chem. 17 (1996) 553e586.
[43] T.A. Halgren, Merck molecular force field. IV. conformational energies and
geometries for MMFF94, J. Comput. Chem. 17 (1996) 587e615.
[44] T.A. Halgren, R. Nachbar, Merck molecular force field. V. Extension of MMFF94
using experimental data, additional computational data, and empirical rules,
J. Comput. Chem. 17 (1996) 616e641.
[45] T.A. Halgren, MMFF VI. MMFF94s option for energy minimization studies,
J. Comput. Chem. 20 (1999) 720e729.
[46] T.A. Halgren, MMFF VII. Characterization of MMFF94, MMFF94s, and other
widely available force fields for conformational energies and for
intermolecular-interaction energies and geometries, J. Comput. Chem. 20
(1999) 730e748.
[31] V.P. Jaakola, M.T. Griffith, M.A. Hanson, V. Cherezov, E.Y. Chien, J.R. Lane,
A.P. Ijzerman, R.C. Stevens, The 2.6 angstrom crystal structure of a human A2A
adenosine receptor bound to an antagonist, Science 322 (2008) 1211e1217.