N6-1,3-Diphenylurea derivatives of 2-phenyl-9-benzyladenines and 8-azaadenines: Synthesis and biological evaluation as allosteric modulators of A2A adenosine receptors

Article abstract of DOI:10.1016/j.ejmech.2007.10.021

Some 1-[4-(9-benzyl-2-phenyl-9H-purin-6-ylamino)-phenyl]-3-phenyl-urea derivatives and some 1-[4-(9-benzyl-2-phenyl-9H-8-azapurin-6-ylamino)-phenyl]-3-phenyl-urea derivatives were synthesised and evaluated for their interaction with adenosine receptors. I

Full text of DOI:10.1016/j.ejmech.2007.10.021

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 43 (2008) 1639e1647
http://www.elsevier.com/locate/ejmech
Original article
N6-1,3-Diphenylurea derivatives of 2-phenyl-9-benzyladenines and
8-azaadenines: Synthesis and biological evaluation as allosteric
modulators of A2A adenosine receptors
a
a
a
a
Irene Giorgi a, , Giuliana Biagi , Anna Maria Bianucci , Alice Borghini , Oreste Livi ,
*
Michele Leonardi a, Daniele Pietra b, Vincenzo Calderone b, Alma Martelli b
a Dipartimento di Scienze Farmaceutiche, Universita` di Pisa, via Bonanno 6, 56126 Pisa, Italy
b Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Biotecnologie, Universita` di Pisa, via Bonanno 6, 56126 Pisa, Italy
Received 30 November 2006; received in revised form 15 October 2007; accepted 16 October 2007
Available online 24 October 2007
Abstract
Some 1-[4-(9-benzyl-2-phenyl-9H-purin-6-ylamino)-phenyl]-3-phenyl-urea derivatives and some 1-[4-(9-benzyl-2-phenyl-9H-8-azapurin-6-
ylamino)-phenyl]-3-phenyl-urea derivatives were synthesised and evaluated for their interaction with adenosine receptors. It was found that some
of these compounds can act as positive enhancers of agonist and antagonist radioligands for the A2A adenosine receptors. This evidence was also
strengthened by functional data. Other compounds can act as negative modulators.
Furthermore these compounds show inhibitory properties for A1 and A3 adenosine receptors.
Ó 2007 Elsevier Masson SAS. All rights reserved.
Keywords: 2-Phenyl-9-benzyladenines; 2-Phenyl-9-benzyl-8-azaadenines; Adenosine receptor ligands; Allosteric modulators; 1,3-Diphenylureas
1. Introduction
The presence of allosteric sites on a receptor provided a new
target for drug discovery [1] and, in particular, allosteric modu-
lators for adenosine receptors have potential therapeutic advan-
tages over orthosteric ligands [2]. Two of the four subtypes of
adenosine receptors have been reported to be allosterically reg-
ulated (A1, A3). For both A1 and A3 receptors, some allosteric
modulators that are relatively selective have been developed
and characterized [2]. For example, PD 81,723 showed selectiv-
ity toward the A1 [3], while 3-(2-pyridinyl)isoquinoline deriva-
tives [4] and a group of 1H-imidazo-[4,5-c]quinolines showed
selectivity toward the A3 subtype [5]. No A2A receptor selective
allosteric modulators have been reported [2]. Only very few
compounds can be found in the literature that act as allosteric
modulators of A2A adenosine receptors: {4-methyl-7-[(methyl-
amino)carbonyl]oxy}-2H-1-benzopyran-2-one (PD 120,918)
[6] was reported to enhance agonist radioligand binding to the
rat striatal A2A adenosine receptor, but without functional
enhancement. Like other G protein-coupled receptors, A2A
adenosine receptors are allosterically modulated by sodium
ions, and by the potassium sparing diuretic, amiloride [7].
Allosteric modulation of G protein-coupled receptors is a rel-
atively novel and unexplored pharmacological concept. Classi-
cally, the mechanism of action for receptor ligands consists
either of mimicking or inhibiting the action of the endogenous
signalling molecules, leading to the traditional classification
of agonists as well as antagonists/inverse agonists, respectively.
The desired effect is exerted through competition at the binding
site for the endogenous neurotransmitter or hormone. Another
kind of action elicited on the receptors by chemical compounds
is the allosteric modulation. Allosteric modulators act at sites
distinct from the agonist binding site, and their effect is evident
only in the presence of an exogenously added agonist because
they do not have an action ‘‘per se’’ but modulate the action of
the naturally occurring hormone or neurotransmitter.
* Corresponding author. Tel.: þ39 0502219549.
E-mail address: igiorgi@farm.unipi.it (I. Giorgi).
0223-5234/$ - see front matter Ó 2007 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2007.10.021

1640
I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
Also the compound SCH-202676 and other 2,3,5-substituted-
[1,2,4]-thiadiazoles displayed peculiar displacement character-
istics of both radiolabelled agonist and antagonist binding to
A2A receptors [8], but further studies have suggested that
thiadiazoles act rather as sulfhydryl modifying agents than as
allosteric modulators [9].
In this study, we present a synthesis of new purine- and 8-
azapurine-N6-1,3-diphenylurea derivatives and their biological
characterization at the adenosine receptors with particular
regard to allosteric modulation of the A2A subtype.
in Schemes 1a,b and 2. The 9-benzyl-8-azahypoxanthine 3 was
prepared following a known two-step reaction (Scheme 1a) in
the presence of sodium ethoxide: the first step was the 1,3 dipolar
addition reaction of benzylazide 1 and cyanoacetamide to give 1-
benzyl-4-carbamoyl-5-amino-1H-1,2,3-triazole 2, which was not
isolated; then, in the same flask, ethyl benzoate was added to
obtain 3 [10] by annulation reaction. Hypoxanthine 3 was then
treated with phosphorous oxychloride to give 4 [11]. Instead,
to obtain the corresponding purine derivative 10 [12], a multi-
step procedure is required, starting from 4,6-dihydroxy-2-
phenylpyrimidine 5, obtained by reaction of benzamidine and
diethylmalonate, that was transformed to 5-amino-4,6-dichloro-
2-phenylpyrimidine 8 by the reactions described in Scheme 1b
and reported in the literature [12]. Compound 8 was treated
with benzylamine to give 9, that was cyclised to purine 10
by reaction with triethylorthoformate.
2. Chemistry
The novel substituted 8-azaadenine derivatives 11e16 and
substituted adenine derivatives17e22weresynthesised asshown
Scheme 1. Synthesis of compounds 4 (a) and 10 (b).

I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
1641
Scheme 2. Synthesis of compounds 11e22.
8-Azapurine 4 and purine 10 were submitted to the same
reaction sequence to obtain the final products, as shown in
Scheme 2. The first step was a reaction with p-nitroaniline
in ethanol at reflux temperature, to give 11 and 17 which, by
catalytic hydrogenation, were transformed to 12 and 18, re-
spectively. By reaction with the suitable isocyanates in ethyl
acetate at reflux temperature, from 12 were obtained com-
pounds 13e16 and from 18 were obtained compounds 19e22.
non-specific binding is 0%. The two compounds 12 and 14
were selected for closer analysis, since they presented intrigu-
ing behaviour, as shown in detail below.
4. Pharmacology
Since the activation of adenosine A2A receptors in ‘‘endo-
thelium intact’’ rat isolated aorta induces vasorelaxing effects,
mediated through the release of endothelial nitric oxide, this
experimental model is widely used as a functional tool for
the pharmacological characterization of drugs acting on this
receptor subtype. In this study, the influence of the selected
compound 14 on the vasorelaxing responses evoked by the
adenosine agonist CGS21680 was evaluated on rat aortic rings
with intact endothelium, in order to identify a pharmacody-
namic profile consistent with the feature of positive enhancer.
3. Biochemistry
Compounds 11e22 (see Table 1) were initially tested on
adenosine A1, A2A and A3 receptors in equilibrium radioligand
displacement experiments. For tests involving the A1 receptors
both a radiolabelled agonist ([3H]CCPA) and a radiolabelled
antagonist ([3H]DPCPX) were used, while the agonist
radioligand [3H]CGS21680 and the antagonist radioligand
[3H]ZM241385 were used in the experiments involving A2A
receptors. In the case of A3 receptors only the agonist radioli-
gand [3H]NECA was used as a probe. The results are given as
Ki ꢀ SEM (nM) and/or percentage of specific binding of the
remaining radioligand, where control binding is 100% and
5. Results and discussion
All compounds assayed on the A1 adenosine receptor at a fi-
nal concentration of 10 mM inhibited [3H]CCPA binding, in
particular, in a significant way, compounds 11, 12, 17 and
Table 1
Biological results
Compound
% Specific binding of remaining radioliganda and Ki ꢀ SEM (nM) of selected compounds
r-A1
h-A2A
h-A3
[3H]DPCPX
[3H]CCPA
[3H]ZM241385
[3H]CGS21680
[3H]NECA
11
12
13
14
15
16
17
18
19
20
21
22
a
105
33 (Ki 132 ꢀ 26)
33
4 (Ki 103 ꢀ 18)
98
77 (Ki 1300 ꢀ 349)
96
0 (Ki 280 ꢀ 44)
58
49
78
92
72
58
51
56
53
70
110
107
79
72
133
167
57
176
94
105
74
74
127
142
180
59
84 (Ki 699 ꢀ 105)
2 (Ki 571 ꢀ 58)
84 (Ki 1900 ꢀ 500)
114
46 (Ki 396 ꢀ 52)
7 (Ki 294 ꢀ 17)
59 (Ki 5737 ꢀ 688)
44 (Ki 4170 ꢀ 375)
101
105
106
107
73
85
67
68
107
111
149
166
74
83
57
120
37 (Ki 5608 ꢀ 841)
56
Data are expressed as means from 2 to 3 independent experiments performed in duplicate; individual values varied less than 15%. The results are given as
Ki ꢀ SEM (nM) and/or percentage of specific binding of radioligand remaining, where control binding is 100% and non-specific binding is 0%.

1642
I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
18 which have p-NO2 or p-NH2 phenylamine as substituent on
the N6 of the adenine or the 8-azaadenine nucleus; the other
compounds, having a bulkier substituent in the same position
(13e16 and 19e22), showed much lower activity. It may be
worth specifying here that, with the term ‘‘inhibition’’ we
mean a decrease of signal associated with the receptor bound
to the radioligand in the presence of the tested compound, with
regard to the value measured in the absence of the tested com-
pound itself. However, when the antagonist radioligand
[3H]DPCPX was used, the inhibition of the measured binding
was lower. In fact, only compounds 12, 17 and 18 (also in this
case these compounds are three of those having less bulky sub-
stituent in the N6 position) showed significant inhibition, while
all the other compounds did not seem to modulate the
[3H]DPCPX binding. This may be explained because com-
pounds 11e22 show higher structural similarity with regard
to the agonist radioligand [3H]CCPA, which contains the ade-
nine moiety, in comparison to the antagonist radioligand
[3H]DPCPX which contains a xanthine nucleus. This implies
that compounds 11e22 show higher affinity for the binding
sub-site that accommodates [3H]CCPA than for the binding
sub-site that accommodates [3H]DPCPX. Further a bulky sub-
stituent in position N6 seems to lower affinity for A1 receptors.
With regard to the assays involving A2A receptors, for which
the radioligand [3H]ZM241385 was used, the observations re-
ported below were made. Some of the assayed compounds
(again 12, 17 and 18) inhibited the binding (percentage of spe-
cific binding lower than 100), while others (11, 19 and 20) did
not seem to affect it. Other compounds (13, 14, 15, 16, 21 and
22), all N6-1,3-diphenylurea derivatives, caused an increase in
the percentage of specific binding, significantly higher than
100. This last effect cannot be attributed to a competitive
mechanism of interaction with the receptors, because both ag-
onist and antagonist compounds, which act through a displace-
ment mechanism at the orthosteric site, are only capable of
inhibiting radioligand binding. Analogous behaviour in bind-
ing modulation at A2A receptors was observed when the assays
were performed by using the agonist radioligand. Some com-
pounds (12, 13, 16, 18e21) inhibited binding (percentage of
specific binding lower than 100), while others (11 and 17)
did not seem to affect binding. Other compounds (14, 15
and 22) caused an increase in the percentage of specific bind-
ing, significantly higher than 100. It is worthy of note that
compounds 14, 15 and 22 are N6-1,3-diphenylurea derivatives.
From comparison of the modulating behaviour of the bind-
ing of [3H]ZM241385 and [3H]CGS21680 at the adenosine
A2A receptors, it can be reasonably hypothesized that
compounds 12, 17 and 18 present an inhibitory action on these
receptors, while compounds 14, 15 and 22 enhance the binding
of a ligand bound to the orthosteric site.
Fig. 1. Displacement of [3H]DPCPX binding at human adenosine A1 receptors
due to compounds 12 and 14. Data were taken from a representative experi-
ment performed in duplicate. Radioligand binding is expressed as a percentage
of specific binding.
In the classical experiment involving the adenosine A1
receptor subtype, performed under displacement conditions,
it can be observed that 12 shows the typical sigmoid curve
of competitive displacement of [3H]DPCPX, with EC50
¼
0.75 mM (Fig. 1). In the case of 14, the inhibition found is
very weak, in practice negligible; the observed curve is nearly
a straight line, which makes the EC50 undetectable. In the ex-
periment involving the A2A subtype, also performed under dis-
placement conditions, by using [3H]ZM241385 as a probe, 12
showed inhibitory behaviour, described by the decreasing
curve and characterized by EC50 ¼ 9.74 mM (Fig. 2).
Moreover, 14 shows receptor enhancer behaviour on the
A2A subtype, characterized by an increasing curve when
both [3H]ZM241385 and [3H]CGS21680 are used (Fig. 3).
EC50 ¼ 16 mM, in the presence of [3H]ZM241385, and
EC50 ¼ 9.95 mM, in the presence of [3H]CGS21680, were
measured.
Further investigations on the mechanism by which 12 and
14 interact with the A2A receptor subtype were performed
In the case of the A3 receptors, compounds 11e22 showed
weak inhibitory properties, when the agonist radioligand
[3H]NECA was used.
The two compounds 12 and 14 were selected for closer anal-
ysis in order to look in more detail at a compound showing inhib-
itory properties on all the adenosine receptors, and a compound
showing enhancer properties to the A2A receptor subtype.
Fig. 2. Displacement of [3H]ZM241385 binding at human adenosine A2A re-
ceptors due to compounds 12 and 14. Data were taken from a representative
experiment performed in duplicate. Radioligand binding is expressed as a per-
centage of specific binding.

I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
1643
Table 2
Dissociation kinetic parameters for [3H]ZM241385 binding at human adeno-
sine A2A receptors in the presence or absence of NECA, 12 or 14 (10 mM)
Kdiss (sꢂ1
)
t1/2 (s)
Control
NECA
12
0.082
0.081
0.78
8.42
8.52
8.89
13.2
14
0.052
behaviour in the biochemical study. The agonist CGS21680
induced A2A-mediated concentration-dependent vasorelaxing
responses in rat aortic rings pre-contracted by the adrenergic
agonist noradrenaline, with values of potency (pIC50) and ef-
ficacy (Emax) of 6.58 ꢀ 0.074 and 76 ꢀ 7, respectively. In the
presence of compound 14 (10 mM), the vasorelaxing effects
of CGS21680 were enhanced (Fig. 5), with a significant
(P < 0.0001) increase of the potency parameter
(pIC50 ¼ 7.16 ꢀ 0.061), significant increases of the responses
evoked by CGS21680 0.1, 0.3 and 1 mM, and a marked (albeit
non-statistically significant) increase of the efficacy parameter
(Emax ¼ 91 ꢀ 5).
Fig. 3. Displacement of [3H]CGS21680 binding at human adenosine A2A re-
ceptors due to compound 14. Data were taken from a representative experi-
ment performed in duplicate. Radioligand binding is expressed as
a percentage of specific binding.
by estimating the change in the maximum dissociation rate of
radioligand [3H]ZM241385 caused by an excess (at 10 mM
concentration) of the same unlabelled molecule (the corre-
sponding cold ligand ZM241385, well-known for its compet-
itive action), in the presence and absence of compounds 12
and 14. In fact, only a compound that interacts with a receptor
in a noncompetitive way can give rise to a change in the max-
imum dissociation rate of the radioligand from the receptor,
when the dissociation is induced by an excess of a competitive
agent. The known A2A agonist NECA, at 10 mM concentra-
tion, was used as an internal standard.
The results showed that, while both 10 mM NECA and
10 mM 12 do not influence the dissociation of 10 mM
[3H]ZM241385 with respect to the control, 14 causes a de-
crease of the dissociation rate (Fig. 4) (Table 2).
In order to evaluate the pharmacodynamic feature by a func-
tional experimental approach, compound 14 was selected as
a representative of those derivatives which exhibited enhancer
6. Conclusions
A quite diverse library of potential ligands for adenosine re-
ceptors was synthesised and assayed in both equilibrium radio-
ligand binding assays and dissociation kinetic experiments
with the aim of explaining in detail the interaction mecha-
nisms by which the above ligands exert their actions. On the
basis of analysis of the initial experiments, it may be con-
cluded that the new compounds 11e22 bind to the adenosine
A1, A2A and A3 receptors. Their behaviour at A1 and A3 recep-
tors is an inhibitory one: they probably act as antagonists by
a competitive mechanism. Additional focused experiments
showed that some compounds also act as inhibitors of the
Fig. 5. Concentrationevasorelaxing effect curves for CGS21680, obtained in
rat aortic rings with intact endothelium, in the presence of compound 14
(10 mM; white squares) or its vehicle (black squares). Vertical bars indicate
the standard errors. The asterisks indicate statistically significant differences
between the responses evoked by the same concentration of CGS21680
(**P < 0.005; ***P < 0.0001).
Fig. 4. Dissociation curves of [3H]ZM241385 from the human adenosine A2A
receptors at 0 ꢁC in the presence and in the absence of NECA, 12 and 14
(10 mM). Data were taken from a representative experiment performed in qua-
druplicate. Radioligand binding is expressed as a percentage of specific
binding.

1644
I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
A2A receptors, while others possess noncompetitive enhancer
properties, that could be accounted for by a mechanism of al-
losteric modulation. In particular, the hypothesis of such a pos-
sible pharmacodynamic property seems to be well supported
by the evidence deriving from the functional tests, which
showed that the A2A-mediated pharmacological effects of
CGS21680 were significantly improved by compound 14.
the precipitate was filtered off and crystallized from methanol
to give 13 as a white solid, 43 mg, 66% yield; m.p. >300 ꢁC.
1
Anal. C30H24N8O (C, H, N); H NMR: 10.88 (s, 1H, Exch),
8.70 (m, 2H, Exch), 8.44 (m, 2H, Arom), 7.90 (m, 2H,
Arom) 7.55e7.24 (m, 14H, Arom), 7.00 (t, J ¼ 7.4 Hz, 1H,
Arom) 5.89 (s, 2H, CH2) d, ppm. IR (cmꢂ1): 3390 (NH),
3375 (NH), 3322 (NH), 1678 (C]O). MS: m/z 512 [Mþ].
TLC: Rf ¼ 0.32 (CHCl3/CH3COOEt 8:2).
7. Experimental
7.1.4. 1-[4-(3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7-ylamino)-phenyl]-3-(4-fluorophenyl)-urea (14)
Compound 14 was prepared by reaction of 12 (50 mg,
0.127 mmol) and 4-fluorophenyl isocyanate (87 mg,
0.635 mmol), following the procedure described for 13. By
crystallization from ethyl acetate the title compound was ob-
tained as a white solid, 21 mg, 32% yield; m.p. >300 ꢁC.
Anal. C30H23FN8O (C, H, N); 1H NMR: 10.88 (s, 1H,
Exch), 8.73 (s, 1H, Exch), 8.70 (s, 1H, Exch), 8.42 (m, 2H,
Arom), 7.92 (m, 2H, Arom), 7.54e7.08 (m, 14H, Arom),
5.89 (s, 2H, CH2) d, ppm. IR (cmꢂ1): 3391 (NH), 3379
(NH), 3315 (NH), 1671 (C]O). MS: m/z 530 [Mþ]. TLC:
Rf ¼ 0.32 (CHCl3/CH3COOEt 8:2).
7.1. Chemistry
Melting points were determined on a Kofler hot-stage appa-
ratus and are uncorrected. IR spectra in Nujol mulls were re-
1
corded on a Mattson Genesis series FTIR spectrometer. H
NMR spectra were recorded on a Bruker AC 200 spectrometer
in d units from TMS as an internal standard; the compounds
were dissolved in DMSO-d6. Mass spectra data were obtained
with a HewlettePackard GC/MS system 5988. TLC was per-
formed on precoated silica gel F254 plates (Merck). Microanal-
yses (C H N) were carried out on a Carlo Erba elemental
analyser (Model 1106) and were within ꢀ0.4% of the theoret-
ical values.
7.1.5. 1-[4-(3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7-ylamino)-phenyl]-3-(4-trifluoromethylphenyl)-
urea (15)
7.1.1. (3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7-yl)-(4-nitrophenyl)-amine (11)
A solution of 6-chloro-9-benzyl-2-phenyl-8-azapurine 4
[11] (2.0 g, 6.2 mmol) and p-nitroaniline (2.6 g, 18.6 mmol)
in absolute ethanol (30 ml) was heated in a well stopped flask
at 85 ꢁC for 48 h. After cooling, the precipitate was filtered off
and crystallized from ethanol to give 11 as a yellow solid,
1.52 g, 58% yield; m.p. 263e264 ꢁC. Anal. C23H17N7O2 (C,
Compound 15 was prepared by reaction of 12 (50 mg,
isocyanate
0.127 mmol)
and
4-trifluoromethylphenyl
(118 mg, 0.635 mmol), following the procedure described
for 13. By crystallization from ethyl acetate the title com-
pound was obtained as a white solid, 35 mg, 48% yield;
m.p. >300 ꢁC. Anal. C31H23F3N8O (C, H, N); 1H NMR:
10.90 (s, 1H, Exch), 9.16 (s, 1H, Exch), 8.87 (s, 1H, Exch),
8.43 (m, 2H, Arom), 7.95e7.31 (m, 16H, Arom), 5.89 (s,
2H, CH2) d, ppm. IR (cmꢂ1): 3395 (NH), 3311 (NH), 1676
(C]O). MS: m/z 580 [Mþ]. TLC: Rf ¼ 0.32 (CHCl3/
CH3COOEt 8:2).
1
H, N); H NMR: 11.61 (s, 1H, Exch), 8.46e8.34 (m, 6H,
Arom), 7.59e7.35 (m, 8H, Arom), 5.93 (s, 2H, CH2) d,
ppm. IR (cmꢂ1): 3337 (NH). MS: m/z 423 [Mþ]. TLC:
Rf ¼ 0.13 (CHCl3).
7.1.2. N-(3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7yl-)-(4-aminophenyl)-amine (12)
7.1.6. 1-[4-(3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7-ylamino)-phenyl]-3-(4-methoxyphenyl)-
urea (16)
A solution of compound 11 (150 mg, 0.33 mmol) in abso-
lute ethanol (125 ml) was stirred under hydrogen atmosphere
in the presence of 5% Pd/C (150 mg) for 12 h at room temper-
ature and pressure. The catalyst was removed by filtration, and
the filtrate evaporated under reduced pressure to give 12,
124 mg, 95% yield; m.p. 257e259 ꢁC. Anal. C23H19N7 (C,
Compound 16 was prepared by reaction of 12 (50 mg,
0.127 mmol) and 4-methoxyphenyl isocyanate (95 mg,
0.635 mmol), following the procedure described for 13. By
crystallization from ethyl acetate the title compound was ob-
tained as a white solid, 25 mg, 37% yield; m.p. 270 ꢁC.
1
H, N); H NMR: 10.59 (s, 1H, Exch), 8.39 (m, 2H, Arom),
1
7.59e7.30 (m, 10H, Arom), 6.64 (d, J ¼ 8.4 Hz, 2H, Arom),
5.86 (s, 2H, CH2), 5.09 (br s, 2H, Exch) d, ppm. IR (cmꢂ1):
3404 (NH), 3325 (NH), 3183 (NH). MS: m/z 393 [Mþ].
TLC: Rf ¼ 0.30 (CHCl3/CH3OH 9.8:0.2).
Anal. C31H26N8O2 (C, H, N); H NMR: 10.87 (s, 1H, Exch),
8.70 (s, 1H, Exch), 8.58 (s, 1H, Exch), 8.43 (m, 2H, Arom),
7.87 (m, 4H, Arom), 7.54e7.30 (m, 8H, Arom), 6.88 (m,
4H, Arom), 5.89 (s, 2H, CH2), 3.71 (s, 3H, CH3) d, ppm. IR
(cmꢂ1): 3390 (NH), 3319 (NH), 3273 (NH), 1645 (C]O).
MS: m/z 542 [Mþ]. TLC: Rf ¼ 0.39 (CHCl3/CH3COOEt 8:2).
7.1.3. {1-[4-(3-Benzyl-5-phenyl-3H-[1,2,3]triazolo[4,5-d]-
pyrimidin-7-ylamino)-phenyl]-3-phenyl}-urea (13)
To a solution of 12 (50 mg, 0.127 mmol) and 0.1 ml of N,N-
diethylaniline in anhydrous acetonitrile (15 ml), phenyl isocy-
anate (75 mg, 0.635 mmol) was added and the solution stirred
at reflux for 12 h under nitrogen atmosphere. After cooling,
7.1.7. N6-[(4-Nitro)-phenyl]-9-benzyl-2-phenyladenine (17)
A solution of 9-benzyl-6-chloro-2-phenylpurine 10 [12]
(2.0 g, 6.2 mmol) and p-nitroaniline (2.6 g, 18.6 mmol) in ab-
solute ethanol (30 ml) was heated in a well stopped flask at

I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
1645
1
85 ꢁC for 48 h. After cooling, the precipitate was filtered off
and crystallized from ethanol to give 17 as a yellow solid,
1.59 g, 61% yield; m.p. 250e251 ꢁC. Anal. C24H18N6O2 (C,
Anal. C32H24F3N7O (C, H, N); H NMR 10.90 (s, 1H, Exch),
9.05 (s, 2H, Exch), 8.43 (m, 2H, Arom), 7.90 (m, 2H, Arom);
7.60e7.12 (m, 15H, Arom), 5.89 (s, 2H, CH2) d, ppm. IR
(cmꢂ1): 3395 (s, NH), 3310 (s, NH), 3292 (s, NH), 1676 (s,
C]O). MS: m/z 578 [Mþ ꢂ 1]. TLC: Rf ¼ 0.30 (CHCl3).
1
H, N); H NMR: 11.59 (s, 1H, Exch), 8.46e8.32 (m, 7H,
Arom), 7.58e7.33 (m, 8H, Arom), 5.93 (s, 2H, CH2) d,
ppm. IR (cmꢂ1): 3334 (NH). MS: m/z 422 [Mþ]. TLC:
Rf ¼ 0.17 (CHCl3).
7.1.8. N6-[(4-Amino)-phenyl]-9-benzyl-
2-phenyladenine (18)
7.1.12. 1-[4-(9-Benzyl-2-phenyl-9H-purin-6-ylamino)-
phenyl]-3-(4-methoxyphenyl-urea (22)
Compound 22 was prepared by reaction of 18 (50 mg,
0.127 mmol) and 4-methoxyphenyl isocyanate (95 mg,
0.635 mmol), following the procedure described for 13. By
crystallization from methanol the title compound was obtained
as a white solid, 49 mg, 71% yield; m.p. >300 ꢁC. Anal.
A solution of compound 17 (150 mg, 0.35 mmol) in ethanol
(125 ml) was stirred under hydrogen atmosphere in the pres-
ence of 5% Pd/C (150 mg) for 12 h at room pressure and tem-
perature. The catalyst was removed by filtration, and the
filtrate evaporated under reduced pressure to give 18,
116 mg, 84% yield; m.p. 220 ꢁC (Dec.). Anal. C24H20N6 (C,
1
C32H27N7O2 (C, H, N); H NMR: 10.87 (br s, 1H, Exch),
8.62 (s, 1H Exch), 8.50 (s, 1H Exch), 8.45 (m, 3H, Arom),
7.92 (m, 2H, Arom), 7.46 (m, 11H, Arom), 6.88 (m, 2H,
Arom), 5.89 (s, 2H, CH2), 3.72 (s, 3H, CH3) d, ppm. IR
(cmꢂ1): 3383 (NH), 3354 (NH), 3274 (NH), 1643 (C]O).
MS: m/z 541 [Mþ]. TLC: Rf ¼ 0.23 (CHCl3).
1
H, N); H NMR: 10.59 (s, 1H, Exch), 8.39 (m, 3H, Arom),
7.58e7.33 (m, 10H, Arom), 6.63 (m, 2H, Arom), 5.90 (s,
2H, CH2), 5.07 (s, 2H, Exch) d, ppm. IR (cmꢂ1): 3395
(NH), 3318 (NH), 3175 (NH). MS: m/z 392 [Mþ]. TLC:
Rf ¼ 0.27 (CHCl3/CH3OH 9.8:0.2).
7.2. Biological assays
7.1.9. 1-[4-(9-Benzyl-2-phenyl-9H-purin-6-ylamino)-
phenyl]-3-phenyl-urea (19)
7.2.1. Materials
[3H]DPCPX (120 Ci/mmol), [3H]NECA (20.6 Ci/mmol),
[3H]CGS21680 (47 Ci/mmol) and [3H]CCPA (42.6 Ci/mmol)
were purchased from Amersham Pharmacia Biotech.
[3H]ZM241385 (27.4 Ci/mmol) was from Tocris Cookson.
DPCPX, NECA and CPA were from SigmaeAldrich. All other
chemicals used, at analytical grade, were from standard com-
mercial sources.
To a solution of 18 (50 mg, 0.127 mmol) and 0.1 ml of N,N-
diethylaniline in anhydrous acetonitrile (15 ml), phenyl isocy-
anate (75 mg, 0.635 mmol) was added and the solution stirred
at reflux for 12 h under nitrogen atmosphere. After cooling,
the precipitate was filtered off and crystallized from methanol
to give 19 as a white solid, 26 mg, 40% yield; m.p. >300 ꢁC.
1
Anal. C31H25N7O (C, H, N); H NMR: 10.88 (s, 1H, Exch),
8.73 (m, 2H, Exch), 8.43 (m, 2H, Arom), 7.90 (m, 2H,
Arom), 7.55e7.24 (m, 15H, Arom), 6.97 (t, 1H, J ¼ 7 Hz,
Arom), 5.89 (s, 2H, CH2) d, ppm. IR (cmꢂ1): 3385 (NH),
3376 (NH), 3324 (NH), 1679 (C]O). MS: m/z 511 [Mþ].
TLC: Rf ¼ 0.20 (CHCl3).
7.2.2. Radioligand binding assays
Membranes of rat cerebral cortex which express adenosine
A1 receptors were prepared by using the method described by
Lohse et al. [13] with slight modifications. Male Wistar rat
brain cortex was homogenised in 10 volumes of ice-cold
0.32 M sucrose, 20 mM TriseHCl buffer pH 7.4 with 30
strokes in Dounce homogeniser. The homogenate was centri-
fuged at 1000g for 10 min to remove the nuclear fraction,
and the resulting supernatant centrifuged at 30,000g for
30 min. The pellet was re-suspended using 10 strokes in
Dounce homogeniser in 10 volumes of ice-cold 5 mM Trise
HCl buffer pH 7.4 for 30 min. After 60 strokes in Dounce ho-
mogeniser, the resulting synaptosomal membranes were pre-
incubated for 30 min at 37 ꢁC with 2 U/ml of adenosine deam-
inase to remove endogenous adenosine. The membrane sus-
pension was then centrifuged at 48,000g for 30 min, and the
resulting pellet was re-suspended in 10 volumes of 50 mM
TriseHCl buffer pH 7.4, and stored at ꢂ80 ꢁC until binding
assays were made.
7.1.10. 1-[4-(9-Benzyl-2-phenyl-9H-purin-6-ylamino)-
phenyl]-3-(4-fluorophenyl)-urea (20)
Compound 20 was prepared by reaction of 18 (50 mg,
0.127 mmol) and 4-fluorophenyl isocyanate (87 mg,
0.635 mmol), following the procedure described for 13. By
crystallization from methanol the title compound was obtained
as a white solid, 35 mg, 52% yield; m.p. >300 ꢁC. Anal.
1
C31H24FN7O (C, H, N); H NMR: 10.85 (s, 1H, Exch), 9.14
(m, 2H, Exch); 8.43 (m, 2H, Arom), 7.92 (m, 2H, Arom),
7.55e7.15 (m, 15H, Arom), 5.93 (s, 2H, CH2) d, ppm. IR
(cmꢂ1): 3345 (NH), 3295 (NH), 3241 (NH), 1650 (C]O).
MS: m/z 529 [Mþ]. TLC: Rf ¼ 0.29 (CHCl3).
7.1.11. 1-[4-(9-Benzyl-2-phenyl-9H-purin-6-ylamino)-
phenyl]-3-(4-trifluoromethylphenyl)-urea (21)
Compound 21 was prepared by reaction of 18 (50 mg,
For displacement experiments involving adenosine A1 re-
ceptors, rat cortex membranes (40 mg of protein) were incu-
bated at 25 ꢁC for 60 min with [3H]DPCPX 0.5 nM
(Kd ¼ 0.4 nM) or 0.3 nM [3H]CCPA (Kd ¼ 0.2 nM), and fixed
concentration (10 mM) or increasing concentrations of the
compounds, in a final volume of 0.4 ml of TriseHCl buffer.
Non-specific binding was measured in the presence of
0.127 mmol)
and
4-trifluoromethylphenyl
isocyanate
(118 mg, 0.635 mmol), following the procedure described for
13. By crystallization from methanol the title compound was
obtained as a white solid, 23 mg, 31% yield; m.p. >300 ꢁC.

1646
I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
100 mM CPA (when [3H]DPCPX was used) or 10 mM DPCPX
(when [3H]CCPA was used). Binding reactions were termi-
nated by dilution with ice-cold 50 mM TriseHCl buffer pH
7.4. Samples were then filtered through Whatman GF/C
glassefiber filters using a Brandel cell harvester. Filters were
washed three times with 2e3 ml of the same buffer. Bound ra-
dioactivity was measured in a liquid scintillation counter
(1600 TR Packard) after the addition of 4 ml of scintillation
liquid (Emulsifier-Safe, Packard).
Slightly different conditions were set in the case of binding
displacement experiments regarding adenosine A2A and A3 re-
ceptors. Membranes of CHO cells expressing recombinant hu-
man A2A or A3 receptors were prepared as previously
described [14]. Membranes (40 mg of protein) were incubated
with [3H]ZM241385 6 nM (Kd ¼ 4 nM) or [3H]CGS21680
13 nM (Kd ¼ 13 nM) in the experiments involving the A2A
subtype, and [3H]NECA 7 nM (Kd ¼ 150 nM) in the ones in-
volving the A3 one, and the compounds to be assayed, at fixed
concentration (10 mM) or at increasing concentrations of the
compounds in duplicate, in a final volume of 0.4 ml of
TriseHCl buffer for 120 min at 25 ꢁC. Non-specific binding
was measured in the presence of 100 mM NECA in the case
of A2A, and 100 mM R-PIA in the case of A3 binding assay.
Samples were handled as mentioned before.
a preamplifier (Buxco Electronics) and with a software of
data acquisition (BIOPAC Systems Inc., MP 100).
After an equilibration period of 60 min, the endothelial in-
tegrity was confirmed by the administration of acetylcholine
(ACh) (10 mM) to noradrenaline (NA; 1 mM)-pre-contracted
vascular rings. A relaxation ꢃ80% of the NA-induced contrac-
tion was considered representative of an acceptable integrity
of the endothelium. Forty-five minutes after confirmation of
the endothelium integrity, the selected compound 14 or its ve-
hicle were added in the organ bath. After 20 min of incubation,
the aortic preparations were contracted by treatment with a sin-
gle concentration of NA (10 mM) and when the contraction
reached a stable plateau, threefold increasing concentrations
of CGS21680 (30 nMe10 mM) were added cumulatively.
Preliminary experiments showed that the NA (1 mM)-
induced contractions remained in a stable tonic state for at least
40 min and that the pre-incubation with compound 14 did not
produce any significant influence on the contractile effect of NA.
Both CGS21680 and compound 14 were dissolved in dime-
thylsulphoxide at the concentration of 10 mM and further di-
luted in bi-distilled water. All the solutions were freshly
prepared immediately before the pharmacological experimen-
tal procedures. Previous experiments showed a complete inef-
fectiveness of the administration of the vehicle.
For the dissociation kinetic studies from adenosine A2A re-
ceptors, a single concentration of [3H]ZM241385 (4 nM) was
used. Membranes (40 mg of protein) were pre-incubated at
0 ꢁC for 180 min. The dissociation experiment started by add-
ing 10 mM ZM241385 with or without 100 mM test com-
pounds in quadruplicate at appropriate time intervals. The
time course of dissociation of total binding was measured by
rapid filtration through Whatman GF/B glassefiber filters,
washing three times with 2 ml of ice-cold buffer. Samples
were treated as described above.
The vasorelaxing efficacy was evaluated as vasorelaxing re-
sponse evoked by CGS21680 10 mM, expressed as a percentage
(%) of the contractile tone induced by NA (1 mM). The param-
eter of potency was expressed as pIC50, calculated as a negative
logarithm of the molar concentration of CGS21680, evoking
a half reduction of the contractile tone induced by NA
(1 mM). The parameters of efficacy and potency were expressed
as mean ꢀ standard error, for six experiments. Student t test was
selected as statistical analysis, P < 0.05 was considered repre-
sentative of a significant statistical difference. Experimental
data were analysed by a computer fitting procedure (software:
GraphPad Prism 4.0).
Binding parameters were calculated by GraphPAD Prism
software (GraphPAD, San Diego, CA, USA).
Acknowledgment
7.3. Pharmacological functional assay
This research was supported by the Italian MIUR (Ministero
`
Istruzione Universita Ricerca).
All the experimental procedures were carried out following
the guidelines of the European Community Council Directive
86-609. Endothelium dependent, adenosine A2A receptor-me-
diated vasorelaxing responses were obtained in rat aortic rings
with intact endothelium, following widely used experimental
approaches [15].
Six male Wistar rats (250e350 g) were sacrificed by cervi-
cal dislocation under light ether anaesthesia and bled. The aor-
tae were immediately excised and freed of extraneous tissues.
The endothelium was preserved. Five millimeter wide aortic
rings were suspended, under a preload of 2 g, in 20 ml organ
baths, containing Tyrode solution (composition of saline in
mM: NaCl 136.8; KCl 2.95; CaCl2 1.80; MgSO4 7H2O 1.05;
NaH2PO4 0.41; NaHCO3 11.9; Glucose 5.5), thermostated at
37 ꢁC and continuously gassed with a mixture of O2 (95%)
and CO2 (5%). Changes in tension were recorded by means
of an isometric transducer (Grass FTO3), connected with
References
[1] A. Christopoulos, Nat. Rev. Drug Discov. 1 (2002) 198e210.
[2] Z.-G. Gao, S.-K. Kim, A.P. IJzerman, K.A. Jacobson, Mini-Rev. Med.
Chem. 5 (2005) 545e553.
[3] R.F. Bruns, J.H. Fergus, Mol. Pharmacol. 38 (1990) 939e949.
[4] Z.-G. Gao, J.E. van Muijlwijk-Koezen, A. Chen, C.E. Muller,
A.P. IJzerman, K.A. Jacobson, Mol. Pharmacol. 60 (2001) 1057e1063.
[5] Z.-G. Gao, S.G. Kim, K.A. Soltysiac, N. Melman, A.P. IJzerman,
K.A. Jacobson, Mol. Pharmacol. 62 (2002) 81e89.
[6] R.F. Bruns, G.H. Lu, in: J.A. Ribeiro (Ed.), Adenosine Receptors in the
Nervous System, Taylor and Francis, London, 1989, p. 192.
[7] Q. Jiang, B.X. Lee, M. Glashofer, A.M. van Rhee, K.A. Jacobson,
J. Med. Chem. 40 (1997) 2588e2595.
[8] A.M.C.H. Van den Nieuwendijk, D. Pietra, L. Heitman, A. Goblyos,
A.P. IJzerman, J. Med. Chem. 47 (2004) 663e672.

I. Giorgi et al. / European Journal of Medicinal Chemistry 43 (2008) 1639e1647
1647
[9] A. Goblyos, H. de Vries, J. Brussee, A.P. IJzerman, J. Med. Chem. 48
(2005) 1145e1151.
[13] M.J. Lohse, V. Lenschow, U. Schwabe, Naunyn-Schmiedeberg’s Arch.
Pharmacol. 326 (1984) 69e74.
[10] P.L. Barili, G. Biagi, O. Livi, V. Scartoni, J. Heterocycl. Chem. 22 (1985)
1607e1609.
[14] G. Biagi, A.M. Bianucci, A. Coi, B. Costa, L. Fabbrini, I. Giorgi,
O. Livi, I. Micco, F. Pacchini, E. Santini, M. Leonardi, F. Nofal
Ahamad, O.L. Salerni, V. Scartoni, Bioorg. Med. Chem. 13 (2005)
4679e4693.
[11] G. Biagi, I. Giorgi, O. Livi, F. Pacchini, V. Scartoni, Farmaco 56 (2001)
929e931.
`
[15] M.P. Bosch, F. Campos, I. Niubo, G. Rosell, J.L. Diaz, J. Brea, M.I. Loza,
A. Guerrero, J. Med. Chem. 47 (2004) 4041e4053.
[12] A.M. Bianucci, G. Biagi, A. Coi, I. Giorgi, O. Livi, F. Pacchini,
V. Scartoni, A. Lucacchini, B. Costa, Drug Dev. Res. 54 (2001) 52e65.