N9-Benzyl-substituted 1,3-dimethyl- and 1,3-dipropyl-pyrimido[2,1-f]purinediones: Synthesis and structure-activity relationships at adenosine A1 and A2A receptors

Article abstract of DOI:10.1016/j.bmc.2007.04.018

Synthesis and physicochemical properties of N-benzyl pyrimido[2,1-f]purinediones are described. These derivatives were synthesized by the cyclization of 7-chloropropylo-8-bromo-1,3-dimethyl- or 1,3-dipropyl xanthine derivatives with corresponding (un)subs

Full text of DOI:10.1016/j.bmc.2007.04.018

Bioorganic & Medicinal Chemistry 15 (2007) 5003–5017
N9-Benzyl-substituted 1,3-dimethyl- and
1,3-dipropyl-pyrimido[2,1-f]purinediones: Synthesis and
structure–activity relationships at adenosine A1 and A2A receptors
a
b
c
´
Anna Drabczynska, Christa E. Muller, Janina Karolak-Wojciechowska,
¨
b
b
a
a,
*
´
Britta Schumacher, Anke Schiedel, Olga Yuzlenko and Katarzyna Kiec-Kononowicz
aJagiellonian University, Medical College, Faculty of Pharmacy, Department of Technology and Biotechnology of Drugs,
Medyczna 9, Pl 30-688 Krako´w, Poland
bPharmaceutical Institute, University of Bonn, An der Immenburg 4, D-53121 Bonn, Germany
c
_
Institute of General and Ecological Chemistry, Technical University of Ło´dz´, Zwirki 36, Pl 90-924 Ło´dz´, Poland
Received 14 July 2006; revised 30 March 2007; accepted 13 April 2007
Available online 19 April 2007
Abstract—Synthesis and physicochemical properties of N-benzyl pyrimido[2,1-f]purinediones are described. These derivatives were
synthesized by the cyclization of 7-chloropropylo-8-bromo-1,3-dimethyl- or 1,3-dipropyl xanthine derivatives with corresponding
(un)substituted benzylamines. Dipropyl derivatives were obtained under microwave irradiation conditions either. The obtained
compounds (1–20) were evaluated for their affinity to adenosine A1 and A2A receptors, selected compounds were additionally inves-
tigated for affinity to the A3 receptor subtype. The results of the radioligand binding assays to A1 and A2A adenosine receptors
showed that most of the 1,3-dimethyl-9-benzylpyrimidopurinediones exhibited selective affinity to A2A receptors at micromolar
or submicromolar concentrations (for example, derivative 9 with o-methoxy substituent displayed a Ki value of 0.699 lM at rat
A2A receptor with more than 36-fold selectivity). Contrary to previously described arylpyrimido[2,1-f]purinediones dipropyl deriv-
atives (compounds 15–20) showed affinity to both kinds of receptors increased, however A1 affinity increased to a larger extent, with
the result that A2A selectivity was abolished. The best adenosine A1 receptor ligand was m-chlorobenzyl derivative 18 (Ki = 0.089
lM and 5-fold A1 selectivity). Structure–activity relationships were discussed with the analysis of lipophilic and spatial properties of
the investigated compounds. Pharmacophore model of adenosine A1 receptor antagonist was adopted for this purpose.
ꢀ 2007 Elsevier Ltd. All rights reserved.
1. Introduction
with dopamine receptors.5,6 Adenosine A2A receptor
antagonists may be useful for the treatment of acute
and chronic neurodegenerative disorders such as cere-
bral ischemia, Parkinson’s and Huntington’s diseases,
as drugs controlling motor functions and exhibiting
neuroprotective properties.7–11
The physiological effects of extracellular adenosine are
mediated by four G-protein coupled receptors: A1,
A
ARs may either inhibit (A1, A3) or stimulate (A2A
2A, A2B and A3 adenosine receptors (ARs).1,2 The
,
2B) adenylate cyclase activity. Coupling to other sec-
A
ond messenger systems has been described, including
phospholipase C stimulation (A1, A2B, A3).3 Selective
A1 AR antagonists have demonstrated promising thera-
peutic potential for the treatment of cognitive diseases,
renal failure, Alzheimer’s disease and cardiac failure.4
In the central nervous system the A2A ARs are present
in high density in basal ganglia and are able to cross-talk
The involvement of AR interaction in seizure disorders
has been discussed.7,10–14 Adenosine and specific AR
agonists exert potent anticonvulsant effects in a variety
of seizure models, primarily via activation of the A1
AR subtype. AR antagonists were described to produce
proconvulsant effects.10,12–14
Since ARs offer an attractive target for drug develop-
ment several studies were carried out in order to find
subtype-selective AR antagonists.15–22 Styryl xanthines
have been recognized as potent and selective adenosine
A2A receptor antagonists (Fig. 1). The first very potent
Keywords: Adenosine A1; A2A receptor antagonists; Pyrimido[2,1-f]
purinediones; Tricyclic xanthine derivatives.
*
Corresponding author. Tel./fax: +48 12 657 04 88; e-mail:
mfkonono@cyf-kr.edu.pl
0968-0896/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.04.018

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5004
Figure 1. Structures of styryl xanthines with adenosine A2A receptor antagonistic properties.
and selective A2A adenosine receptor antagonists were
KF 1783723,24 and CSC.25 KF 17837, being more potent
A2A adenosine receptor antagonist than CSC, was less
selective vs A1 adenosine receptor subtype. Two major
problems are connected with this group of compounds:
low water solubility26 and rapid photoisomerisation.27,28
To enhance water solubility of styryl xanthines two
affinity and selectivity, in vivo some of the compounds
displayed anticonvulsant activity. The pyrimidine ann-
elated ring (II) was beneficial for both receptor affinity
and anticonvulsant activity. Continuing our research in
the group of tricyclic derivatives of purinediones16,33,34
pyrimido[2,1-f]purinediones with (un)substituted benzyl
moieties at the annelated ring were synthesized. Such
compounds (III), possessing aryl substituent placed in
the greater (two bonds) distance from the nitrogen atom
of the annelated ring, were designed as better mimicking
the structure of styryl xanthines.
main approaches have been used by Muller and cowork-
¨
ers: introduction of polar groups on the phenyl ring (SS-
DMPX)29 or preparation of phosphate pro-drugs
(MSX-2, MSX-3).30 Compound KW-6002 is being eval-
uated in phase II clinical trials as potential antiparkinso-
nian and antidepressant drug.31,32
To gain insight into the structure affinity relationships of
this class of compounds, the lead structure benzylpyrim-
idotheophylline (1) was modified by:
In our studies we focused on the search for A1/A2A AR
ligands among tricyclic xanthine derivatives.16,33–35 Com-
pounds (I) with oxygen or nitrogen containing fused rings
were obtained16,33,34 as constrained analogues of 8-sty-
rylxanthines (Fig. 2).18,36 Thus imidazo-, pyrimido- and
1,3-diazepino[2,1-f]purinediones were synthesized (I).
One series of compounds (II) possessed variously substi-
tuted aryl moiety placed directly at the nitrogen atom of
the fused ring. In vitro studies have shown their A2A AR
• elongation of R1 substituents (replacement of the
methyl groups in theophylline by propyl groups as
in the standard xanthine A1 antagonist 8-cyclopen-
tyl-1,3-dipropylxanthine (DPCPX)),
• introducing R2 substituent to the benzyl methylene,
• introducing R3 substituent in position o-, m- or p- of
the aromatic ring,
8-styrylxanthine derivatives
O
O
CH3
(CH2)n
R1
R1
O
N
X2
N
N
ring formation
N
X1
N
N
O
N
R1
N
R1
R
I
R1 - methyl, ethyl, propyl, propargyl
- Cl, Br, MeO
a. X1 = CHR X2 = O
b. X1 = O
X2 = CHR
X2 = CH2
R
c. X1 = NR
O
(CH2)n
N
O
R1
O
7
6
R1
O
8
N
N
N5
N
4
N
9
N
3
2
N
1
N
R1
R3
N
R1
R2
R
III
II
Figure 2. Oxygen or nitrogen containing annelated xanthine derivatives as constrained analogues of 8-styryl xanthines.

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5005
• replacement of the phenyl ring by the potentially bio-
isosteric heterocycles furane or pyridine.
melted with the appropriate benzylamine in a micro-
wave oven (power 300–750 W) for 4–7 min.
The synthesized compounds were evaluated in vitro for
their affinity to ARs. Physicochemical properties of
tested compounds were estimated by means of calcula-
tions. In order to analyze three-dimensional properties
of this group of compounds structure examination was
performed by means of molecular modelling and
X-ray analysis for two selected compounds.
Although the yields of the traditional and microwave
method were comparable, the advantages of the last
one were a dramatically reduced reaction time and a bet-
ter quality (higher purity) of the products.
The structures of the synthesized compounds were con-
firmed by UV, IR and H NMR spectra. The UV spec-
1
tra showed a bathochromic shift of kmax from ca.
276 nm to about 300 nm, typical for 8-aminoxanthine
derivatives.41 The IR showed absorption bands typical
for xanthine derivatives.42
2. Chemistry
The synthesis of tricyclic 1,3-dimethylpyrimidopurinedi-
one with benzyl substituents in the 9-position of the ann-
elated ring was accomplished as shown in Figure 3. As
starting material 7-(3-chloropropyl)-8-bromotheophyl-
line was used which was obtained as previously de-
scribed.37 The compound was reacted with the
substituted in o-, m- or p-position benzylamines under
various reaction conditions with regard to the amount
of amine, solvent, and reaction time. The data are sum-
marized in Table 1.
3. X-ray structure analysis of compounds 8 and 14
Monocrystals suitable for X-ray structural studies were
obtained only for 2 of 20 N9-benzyl-substituted pyrimi-
do[2,1-f]purinediones. Then the structures of 8 and 14
were determined. In the studied molecules the basic tri-
cyclic pyrimido[2,1-f]purinedione is exactly the same
(Fig. 4). The purinedione skeleton is coplanar. In the
annelated pyrimidine ring three carbon atoms (C6, C7
and C8) possess sp3 hybridization and this ring in both
molecules adopts sofa conformation with the C7 atom
in flap position. The C11 atom from substituent at N9
resides in an equatorial position. It is important that
C7 and C11 atoms are located on opposite sides of tricy-
cle (trans with respect to main skeleton). In the structure
of 8 the phenyl attached to C11 is inclined to the pyrim-
ido[2,1-f]purinedione at 75.1(1)0. For the furane ring,
placed at C11 in molecule 14, a respective angle equal
to 80.89(7)0 was found.
The synthesis of compounds 1 and 12 had been de-
scribed previously,38 but their structures had only been
determined by UV spectra and no biological activity
has been reported.
For the preparation of 1,3-dipropylxanthine as starting
material for cyclization to the corresponding dipropyl-
substituted pyrimidopurinediones standard procedures
were applied.39,40 Subsequent bromination, introduction
of the chloropropyl chain in position 7 by two phase
alkylation with 1-chloro-3-bromopropane followed by
cyclization with benzylamines yielded the target tricyclic
9-benzyl-substituted compounds (Fig. 3).
The structures of 8 and 14 constituted starting points
for subsequent molecular modelling of the molecules
described in this study. All molecules contain rotational
one atomic-linker (sp3-hybridized C11), which joins to-
gether pyrimido[2,1-f]purinedione and the aromatic sub-
stituent. Conformational analysis of that molecule
Alternatively, dipropyl derivatives 15–20 could also be
prepared by microwave flash-heating method. Thus,
8-bromo-7-(3-chloropropyl)-1,3-dipropylxanthine was
O
H
O
R1
O
R1
O
H
N
N
Cl
Br
HBr
N
N
Br
K2CO3, TEBA
NaClO3
N
N
N
N
R1
R1
Cl
O
O
R2
R1
N
R1
N
N
N
N
Br +
NH2
N
N
O
N
R3
O
N
R1
R2
R1
R3
R1=R1 = CH3
R1=R1 = C3H7
Figure 3. Synthesis of 1,3-dialkyl-9-benzylpyrimido[2,1-f]purinediones.

Table 1. Physical data and reaction conditions of 9-benzylpyrimido[2,1-f]-purinediones
O
R2
N
N
O
N
R1
N
R3
N
Compound
R1
R2
R3
Mol. form. m.w.
Mp [ꢁC]
Yield [%]
70
Reaction medium
—
Reaction time [h]
3
Crystal. solvent
Ethanol
TLC Rf
1
CH3
CH3
C17H19O2N5, 325.40
197–198a
0.56A
CH2
2
3
4
CH3
CH3
CH3
CH3
CH3
CH3
C18H21O2N5, 339.39
223–224
230–232
233–235
58
70
63
—
4
10
8
Ethanol
0.72A
0.65A
0.65A
CH2
CH2
CH
3
C17H18O2N5Cl, 359.81
C17H18O2N5Cl, 359.81
Me-Digol
Me-Digol
Methoxy-ethanol
Methoxy-ethanol
Cl
Cl
Cl
CH2
CH2
Cl
5
6
CH3
CH3
CH3
CH3
C17H18O2N5Cl, 359.81
C17H17O2N4Cl2, 394.26
168–169
195–196
50
63
Me-Digol
Me-Digol
10
9
Methoxy-ethanol
0.58A
0.94A
n-Propanol
CH2
CH2
CH2
Cl
7
8
CH3
CH3
CH3
CH3
C17H18O2N5F, 343.35
C18H21O3N5, 355.39
210–212
185–187
40
59
Me-Digol
Propanol
8
3
Methoxy-ethanol
Propanol
0.58A
0.63A
F
OCH3
CH3O
CH2
9
CH3
CH3
CH3
CH3
C18H21O3N5, 355.39
C18H21O3N5, 355.39
217–218
166–167
72
55
Me-Digol
Me-Digol
10
10
n-Butanol
0.68A
0.64A
OCH3
10
Ethanol
CH2
CH
11
CH3
CH3
C23H23O2N5, 401.45
247–249
70
Methoxy-ethanol
10
Propanol
0.74A
CH
12
13
CH3
CH3
CH3
CH3
C18H21O2N5, 339.40
C16H18O2N6, 326.35
137a
80
49
n-Propanol
6
6
50% Ethanol
Isopropanol
0.56A
0.56B
CH3
N
188–190
Butanol
CH2

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5007
Figure 4. Superimposition of molecules 8 and14 with respect to
pyrimidopurinedione moiety atom.
fragment was performed with U = C8–N9–C11–C12 tor-
sion angle rotation.43 Two energy minima (at U = 70ꢁ
and 50ꢁ for 8 and 14, respectively) correspond to trans
and cis location of C7 and C11 atoms with respect to the
main molecule skeleton. Conformations present in both
studied structures are close to slightly deeper energy
minimum describing trans form.
In both structures main supramolecular motives are
formed by weak intermolecular interactions of O2. . .H–
C8(sp3) (Table 2b) and (Fig. 5). It should be pointed
out that the carbon atom C8 is located as an exact neigh-
bour of the electron-withdrawing non-planar nitrogen
N9 (see Table 2a) and that the partial charge on respective
hydrogens equals about 0.11e.44 In the crystals of 8, due
to the 4-OCH3 substitution of the phenyl ring, an addi-
tional weak H-bond of C14–H14. . .O18 is also observed.
4. Pharmacology
All compounds were tested in vitro in radioligand bind-
ing assays for affinity to A1 and A2A adenosine receptors
in rat cortical membrane and rat striatal membrane
preparations, respectively. As A1 AR radioligand
[3H]2-chloro-N6-cyclopentyladenosine ([3H]CCPA) and
as A2A receptor ligand [3H]1-propargyl-3-hydroxypro-
pyl-7-methyl-8-(3-methoxystyryl)-xanthine ([3H]MSX-
2) were used.45,46 Selected compounds were additionally
investigated at human A1, A2A and A3 ARs in mamma-
lian cells recombinantly expressing the respective
receptor subtype. As A3 radioligand we used [3H]2-phe-
nyl-8-ethyl-4-methyl-(8R)-4,5,7,8-tetrahydro-1H- imi-
dazo[2,1-i]purine-5-one ([3H]PSB-1122). The results are
presented in Table 3. The standard antagonists caffeine
(non-selective) and KW-6002 (A2A-selective) were
included for comparison.
5. Biological results and discussion
The results of the radioligand binding assays to A1 and
A2A adenosine receptors showed that most of the 1,3-di-
methyl-9-benzylpyrimidopurinediones exhibited affinity

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5008
Table 2. Selected geometrical data from 8 and 14 molecules
(a) Molecule geometry
8
14
Dihedral angle skeleton/ring at C11 [ꢁ]
R angles at N9 [ꢁ]
C8–N9–C11–C12 [ꢁ]
75.1(1)
354.9
68.1(2)
80.89(7)
358.5
56.8(2)
(b) Weak H-bonds in the crystals:
D–H . . . A
˚
D–H (A)
˚
˚
Sym. code
H Á Á Á A (A)
2.17
2.60
D Á Á Á A (A)
3.344(3)
3.354(5)
D–H Á Á Á A (ꢁ)
158
164
8
C8–H8A Á Á Á O2
À1 + x, 1 + y, z
Àx, 3 À y, 1 À z
1.23
0.96
C17–H17 Á Á Á O18
14
C8–H8BÁ Á ÁO2
À1 + x, y, 1 + z
0.95
2.53
3.458(2)
166
Figure 5. H-bonds net in the crystals of compounds 8 and 14.
Table 3. Affinities of 9-benzylpyrimido[2,1-f]purinediones at A1 and A2A adenosine receptors
O
R2
N
N
O
N
R1
N
R3
N
a
b
Compound
R3
A1 versus [3H] CCPA
A2A selectivity A1/A2A
A2A versus [3H] MSX-2
% inhibition
at 25 lM
Ki SEM
% inhibition
at 25 lM
Ki SEM
(or 2.5 lM)
(or 2.5 lM)
Caffeine
KW-6002
—
Figure 1
n.d.c
n.d.c
18.8 5.6
0.23 0.03
—
n.d.c
n.d.c
32.5 8.0
0.00515 0.00025
R1@R2@CH3
1
66
3
5
3.58 0.17
3.28
96
2
1.09 0.30
>2.5
CH2
CH2
2
3
(0)
ꢀ2.5
(26 26)
CH3
Cl
27
P25
70
9
4.28 0.44
1.24 0.20
1.12 0.14
CH2
Cl
4
5
(10 10)
>2.5
(75 4)
CH2
Cl
67
1
1.42 0.08
93
4
CH2

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5009
Table 3 (continued)
a
b
Compound
R3
A1 versus [3H] CCPA
A2A selectivity A1/A2A
A2A versus [3H] MSX-2
% inhibition
at 25 lM
Ki SEM
% inhibition
at 25 lM
Ki SEM
(or 2.5 lM)
(or 2.5 lM)
6
Cl
34 11
P25
75
7
2.48 0.07
CH2
CH2
Cl
F
7
8
9
17 10
>25
86
83
99
3
1
1
2.33 0.21
5.49 0.54
28
2
>25
CH2
OCH3
CH3O
CH2
18 6 (11 1)d
>25 (>25)d
0.699 0.016
(11.3 2.3)d
OCH3
10
11
60
7
2.50 0.12
1.9
84
7
1.31 0.13
CH2
CH
(7 2))
>2.5
(5 5)
>2.5
CH
12
35
23
8
4
P25
8.8
7
92
79
1
7
2.84 0.17
2.53 0.21
CH3
N
13e
17.8 0.51
CH2
CH2
14
56
1
3.43 0.46
74
4
3.31 0.64
O
A1 selectivity A2A/A1
R1@R2@C3H7
15
91
28
89
1
1.00 0.15
0.23 0.15
0.48 0.08
0.089 0.029
0.18 0.04
0.71 0.13
0.73
3.6
60 5(at 3 lM)
0.73 0.09
0.87 0.41
1.73 0.35
0.478 0.022
1.13 0.14
3.43 0.71
CH2
CH2
CH3
Cl
16f
17
1
3
72.
88
92
8
Cl
3.6
1
CH2
CH2
Cl
18g
19
87 31.86 (at 1 lM)
14.5
6.28
4.8
2
OCH3
94
89
1
1
72 10
82 10
CH2
CH2
20
F
a Rat brain cortical membranes.
b Rat brain striatal membranes.
c n.d., not determined.
d Data from human recombinant receptors permanently expressed in CHO cells.
e Ki value of 13 at human A3 ARs ꢀ 10 lM (4% inhibition of radioligand binding at 10 lM).
f Ki value of 16 at human A3 ARs: 1.60 0.30 lM.
g Ki value of 18 at human A3 ARs: 1.29 0.31 lM.

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5010
to A2A receptors but poor A1 affinity (Table 3). The lead
structure benzylpyrimidopurinedione (1) showed affinity
for A1 and A2A receptors in the low micromolar concen-
A2A receptors to investigate possible species differences.
In fact, 9 exhibited lower affinity for the human than for
the rat receptors: its Ki value at the human A2A receptor
(Ki = 11.3 lM) was 16-fold lower than that at the rat
receptor (Ki = 0.699 lM). The same tendency, lower
affinity for human than for rat A2A receptors, has been
observed for other members of the pyrimido[2,1-f]purin-
edione family.35 Three selected compounds were investi-
gated at human A3 ARs, one dimethyl derivative (13)
and two dipropyl derivatives (16, 18). While the di-
methyl derivative was inactive at A3 receptors, the
dipropyl derivatives exhibited Ki values of 1–2 lM. This
result is in accordance with data from bicyclic xanthine
derivatives, in which a dipropyl substitution is very
favourable for A3 affinity, in contrast to a dimethyl sub-
stitution.21,48 It can be concluded that the N,N-dipropyl
substitution of the xanthine part is not only unfavour-
able for A2A-selectivity versus A1, but also versus A3
ARs.
tration range with 3-fold A2A selectivity (KiA2A
=
1.09 lM, (KiA1 = 3.58 lM). Introduction of substituents
R3 (CH3, OCH3, Cl, F) in the p-position of the benzyl
residue was disadvantageous for A2A affinity (e.g., 2, 3,
7, 8). Better tolerated were substituents in the m- or o-
position. The o-methoxy derivative 9 was the best A2A
ligand (Ki = 0.699 lM) of the presented series.
Introduction of a substituent R2 at the methylene linker
of the benzyl residue (methyl in 12, phenyl in 11) was
disadvantageous too. The phenyl substituent (11) de-
creased affinity to both kinds of receptors practically
to zero (5% and 7% inhibition of radioligand binding
at 2.5 lM). The isosteric exchange of the phenyl ring
(in the 9-benzyl substituent) by a heterocyclic ring (furyl
in 14, pyridyl in 13) had the similar effect. Taking into
account previously described results33 on the arylpyrim-
idopurinediones (II), examined for the profile at A2A
activity in sodium chloride shift experiments, it might
be suggested that recently investigated compounds also
exhibit A2A antagonistic activity.47
6. Structure–activity relationships
Starting with structure–activity relationship analysis, the
Ki values of pyrimido[2,1-f]purinedione of both studied
receptors (A1 and A2A) were examined as a function of
the calculated lipophilicity descriptor, logP (and
logD).49 All indispensable numerical data are collected
in Table 4 while results in graphical forms are presented
in Figures 6 and 7. After primary analysis two series of
derivatives were selected: one with R1@CH3 (com-
pounds 1–14) and the other one with R1@C3H7 (com-
pounds 15–20). This is in accordance with two clusters
observed on the diagrams of Ki value versus logP (Figs.
6 and 7). Two linear relationships—involving log P for
both series of compounds and affinity to A1 receptor (Ki
A1 values on Fig. 7)—were shown. Linear equations
make it impossible to estimate the optimal values of
The most interesting effects were observed in compounds
showing an elongation of the substituents in the 1- and 3-
positions from methyl to propyl (compounds 15–20).
Affinity to both kinds of receptors increased, however
A1 affinity increased to a larger extent, with the result
that A2A selectivity was abolished (compound 15). All
investigated dipropyl-substituted compounds showed
submicromolar A1 affinity (with the exception of 15,
Ki = 1.00 lM). The best A1 ligand was compound 18
(Ki = 0.089 lM and 5-fold A1 selectivity).
The most potent and at the same time most selective
compound 9 was additionally tested at human A1 and
Table 4. Physicochemical, geometrical and electronic parameters of studied pyrimido[2,1-f]purinediones considered in QSAR studies
Pharmacological
data
Lipophilicity
Geometrical data
QSAR
Electronic
parameters
Ki A1
Ki A2A
logP
logD
V A3
dV
MR
89.06
MP
Hyd. en
HOMO
DP
1
2
3.58
1.09
1.50
1.93
2.22
2.21
2.28
3.00
1.65
1.45
1.46
1.56
3.18
2.03
0.40
0.90
2.86
4.17
4.16
4.22
3.40
3.59
1.12
1.53
1.96
1.95
2.02
2.83
1.38
1.11
1.12
1.22
2.90
1.76
0.32
0.71
2.42
3.93
3.92
3.99
3.08
3.35
920.06
972.03
963.04
951.57
958.87
994.44
929.30
997.83
992.26
996.26
1103.95
960.72
962.10
879.32
1168.71
1159.92
1147.89
1159.02
1193.65
1125.25
202.28
254.25
245.26
233.79
37.17
39.00
39.10
39.10
39.10
41.02
37.08
39.64
39.64
39.64
46.83
39.00
37.13
31.90
46.34
46.44
46.44
46.44
46.98
44.42
À0.84
0.32
À8.44
À8.42
À8.51
À8.46
À8.49
À8.53
À8.51
À8.41
À8.34
À8.42
À8.39
À8.40
À8.31
À8.43
À8.36
À8.44
À8.39
À8.45
À8.35
À8.45
5.328
5.502
4.881
4.576
4.227
4.037
4.894
4.420
6.775
5.448
5.266
5.449
6.876
5.777
5.090
4.730
4.099
3.820
6.153
4.777
94.10
93.86
93.86
93.86
98.67
89.27
95.52
95.52
95.52
113.34
93.48
86.50
81.45
112.64
112.41
112.41
112.41
114.06
107.82
3
4.28
1.24
1.12
2.48
2.33
5.49
0.70
1.31
À0.51
À0.64
À0.39
À0.33
À0.55
À2.52
1.09
4
5
1.42
6
276.62
211.52
280.05
274.48
279.15
386.17
291.31
244.32
161.54
7
8
9
10
11
12
13
14
15
16
17
18
19
20
2.5
À2.35
À1.46
À0.38
À1.03
À3.33
À1.54
0.73
2.84
2.53
3.31
0.73
0.87
1.73
1.29
1.13
3.43
17.8
3.43
1.00
0.23
0.48
0.089
0.18
0.71
235.30
223.45
0.57
0.68
274.20
201.94
À1.30
0.65

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5011
a
b
K
A1
20
i
20
K i A1
1 = 2.00 logP2 - 12.58 logP + 18.67
r = 0.90
series with R1 = CH3
K i A1 = - 7.97 logP + 16.47
r = 0.82
K i
K i
A
A
15
10
5
1 = 2.42 logD2 - 13.30 logD + 16.89
r = 0.89
15
10
5
R1 = CH3
series with R1 = C3H7
K i A1 = - 0.46 logP + 2.15
r = 0.71
logP
logD (at pH=7.4)
R1 = C3H7
logP
5
0
0
0
0
1
2
3
4
1
2
3
4
5
logP
logD (pH=7.4)
-5
-5
Figure 6. Relationships of Ki to A1 receptor in the function of logP for two series of derivatives with identical R1 substituents.
logP. Considering both series as one group, the second
degree polynominal equation between logP and Ki for
A1 enables to evaluate the optimal value of lipophilicity
for pyrimido[2,1-f]purinediones. This optimum was
found at about 3 for logP and logD (at pH 7.4)
(Fig. 6b). Contrary to the above, the relationships be-
tween logP and the activity towards adenosine A2A
receptor (Ki for A2A values in Fig. 7) are not obvious.
It means that lipophilicity for the activity for adenosine
A2A receptor is less significant than that to A1 receptor.
In Table 4 also some other similarity descriptors—geo-
metrical and electronic ones—are collected for studied
pyrimido[2,1-f]purinediones. Unfortunately, no explicit
conclusions could have been drawn based on one or
multilinear relationships.50
zylpyrimido[2,1-f]purinediones to A1 receptor (Fig. 6)
but not to A2A one (Fig. 7). The impact of lipophilicity
for A1 receptor ligands is manifested also in the
appropriate pharmacophore model which consists of
as many as three lipophilic pockets essential for
ligand–receptor interactions.51,52 Besides these, two
H-bond acceptors and one H-bond donor are needed
for binding. Keeping this in mind, that model has been
adopted to benzylpyrimido[2,1-f]purinedione molecules
in the manner depicted on Figure 8.
Two R1 substituents from the studied benzylpyrimi-
do[2,1-f]purinediones are located in the lipophilic pock-
ets P2 and P3. In accordance with the data from Tables
3 and 4, replacement of both R1@CH3 by C3H7 increases
logP value as well as molecule activity for adenosine A1
receptor. Thus, two three-atomic chains more efficiently
fix the molecule at the respective binding site of the
receptor. Furthermore, purine oxygen—located between
R1 substituents—is shielded by spaces occupied by the
propyl chains. In consequence, as it was observed in
6.1. Benzylpyrimido[2,1-f]purinediones and A1 receptor
interactions
The relationships of Ki = f (logP and/or logD) pointed
out an importance of lipophilicity for activity of the ben-
6
Ki A2A
4
series with
R1 = CH3
series with
R1 = C
3H7
2
0
logP
0
1
2
3
4
Figure 8. Model of interaction of benzylpyrimido[2,1-f]purinediones
with A1 receptor.
Figure 7. Relationships of Ki to A2A receptor with logP values.

Table 5. Spectroscopic data of compounds 1–20
Compound UV [nm] IR [cmÀ1 1H NMR
]
1
29938
301
301
302
301
301
304
301
300
302
299
29915
299
302
1704 CO (pos. 2) 2.11–2.18 (m, 2H, CH2CH2–CH2), 3.18 (t, 2H, J = 5.54 Hz, CH2N9), 3.41 (s, 3H, N3CH3), 3.56 (s, 3H, N1CH3), 4.26 (t, 2H, J = 6.05 Hz, N5CH2),
1654 CO (pos. 4) 4.78 (s, 2H, N9CH2), 7.30–7.41 (m, 5H, phenyl)
2
1704 CO (pos. 2) 2.07–2.13 (m, 2H, CH2–CH2CH2), 2.34 (s, 3H, CH3), 3.26 (t, 2H, J = 5.78 Hz, CH2N9), 3.36 (s, 3H, N3CH3), 3.52 (s, 3H, N1CH3),
1657 CO (pos. 4) 4.21 (t, 2H, J = 6.02 Hz, N5CH2), 4.70 (s, 2H, N9CH2), 7.12–7.23 (m, 4H, phenyl)
3
1697 CO (pos. 2) 2.08–2.17 (m, 2H, CH2–CH2CH2), 3.24–3.27 (m, 2H, CH2N9), 3.37 (s, 3H, N3CH3), 3.52 (s, 3H, N1CH3), 4.23 (t, 2H, J = 6.00 Hz, N5CH2),
1655 CO (pos. 4) 4.71 (s, 2H, N9CH2), 7.24–7.34 (m, 4H, phenyl)
4
2705 CO (pos. 2) 2.16–2.23 (m, 2H, CH2–CH2CH2), 3.36 (t, 2H, J = 5.64 Hz, CH2N9), 3.41 (s, 3H, N3CH3), 3.54 (s, 3H, N1CH3), 4.29 (t, 2H, J = 6.05 Hz, N5CH2),
1651 CO (pos. 4) 4.92 (s, 2H, CH2N9), 7.26–7.32 (m, 2H, H4,5-phenyl), 7.39–7.45 (m, 2H, H3,6-phenyl)
5
1691 CO (pos. 2) 2.14–2.22 (m, 2H, CH2–CH2CH2), 3.30 (t, 2H, J = 5.64 Hz, CH2N9), 3.41 (s, 3H, N3CH3), 3.56 (s, 3H, N1CH3), 4.27 (t, 2H, J = 5.91 Hz, N5CH2),
1660 CO (pos. 4) 4.75 (s, 2H, N9CH2), 7.21–7.25 (m, 1H, H2-phenyl), 7.31–7.34 (m, 3H, H4,5,6-phenyl)
6
1702 CO (pos. 2) 2.12–2.21 (m, 2H, CH2CH2CH2), 3.31–3.37 (m, 5H, N3CH3 + CH2N9), 3.50 (s, 3H, N1CH3), 4.25 (s, 2H, J = 6.23 Hz, N5CH2), 4.83 (s, 2H, N9CH2),
1654 CO (pos. 4) 7.21–7.43 (m, 3H, phenyl)
7
1702 CO (pos. 2) 2.07–2.17 (m, 2H, CH2–CH2CH2), 3.24 (t, 2H, J = 5.00 Hz, CH2N9), 3.38 (s, 3H, N3CH3), 3.53 (s, 3H, N1CH3), 4.22 (t, 2H, J = 6.25 Hz, N5CH2),
1662 CO (pos. 4) 4.70 (s, 2H, N9CH2), 7.00–7.07 (m, 2H, H2,6-phenyl), 7.27–7.32 (m, 2H, H3,5-phenyl)
8
1697 CO (pos. 2) 2.05–2.15 (m, 2H, CH2CH2CH2), 3.23 (t, 2H, J = 5.38 Hz, CH2N9), 3.38 (s, 3H, N3CH3), 3.53 (s, 3H, N1CH3), 3.80 (s, 3H, OCH3), 4.21 (t, 2H,
1656 CO (pos. 4) N5CH2), 4.67 (s, 2H, N9CH2), 6.86–6.90 (m, 2H, H3,5-phenyl), 7.23–7.27 (m, 2H, H2,6-phenyl)
9
1702 CO (pos. 2) 2.07–2.17 (m, 2H, CH2–CH2–CH2), 3.05–3.37 (m, 5H, N3CH3, CH2N9), 3.52 (s, 3H, N1CH3), 3.84 (s, 3H, OCH3), 4.22 (t, 2H, J = 6.25 Hz, N5CH2),
1651 CO (pos. 4) 4.76 (s, 2H, N9CH2), 6.88–6.96 (m, 2H, H4,5-phenyl), 7.25–7.31 (m, 2H, H3,6-phenyl)
10
11
12
13
14
1698 CO (pos. 2) 2.07–2.10 (m, CH2–CH2–CH2), 3.27 (t, 2H, J = 5.65 Hz, CH2N9), 3.68 (s, 3H, N3CH3), 3.51 (s, 3H, N1CH3), 3.79 (s, 3H, OCH3), 4.32 (t, 2H,
1653 CO (pos. 4) J = 6.02 Hz, N5CH2), 4.71 (s, 2H, N9CH2), 6.91–6.97 (m, 2H, H4,5-phenyl), 7.22–7.30 (m, 2H, H2,6-phenyl)
1700 CO (pos. 2) 2.10–2.19 (m, 2H, CH2–CH2ÀCH2), 3.15 (t, 2H, J = 5.88 Hz, CH2N9), 3.35 (s, 3H, N3CH3), 3.51 (s, 3H, N1CH3), 4.21 (t, 2H, J = 6.01 Hz, N5CH2),
1648 CO (pos. 4) 7.06 (s, 1H, CH), 7.25–7.38 (m, 10H, phenyl)
1690 CO (pos. 2) 1.63 (t, 3H, J = 6.25 Hz), 1.94–2.07 (m, 2H, CH2N9), 2.91–3.02 (m, 1H, CH2N9), 3.17–3.26 (m, 1H, CH2N9), 3.38 (s, 3H, N3CH3), 3.54 (s, 3H, N1CH3),
1659 CO (pos. 4) 4.13–4.27 (m, 2H, N5CH2), 5.83–5.91 (q, J = 7.5 Hz, 1H, N9–CH), 7.26–7.37 (m, 5H, phenyl)
1703 CO (pos. 2) 2.11–2.21 CH2–CH2–CH2), 3.37–3.45 (m, 5H, N3CH3 + CH2N9), 3.50 (s, 3H, N1CH3), 4.25 (t, 2H, J = 6.00 Hz, N5CH2), 4.86 (s, 2H, N9CH2),
1654 CO (pos. 4) 7.25–7.36 (m, 2H, H3,5-pyridyl), 7.65–7.71 (td, 1H, H4-pyridyl), 8.56–8.58 (m, 1H, H6-pyridyl)
3100 furan
2.09–2.18 (m, 2H, CH2CH2CH2), 3.31–3.36 (m, 5H, N3CH3, CH2N9), 3.53 (s, 3H, N1CH3), 4.24 (t, 2H, J = 6.00 Hz, N5CH3), 4.71 (s, 2H, N9CH2),
1700 CO (pos. 2) 6.30–6.35 (m, 2H, H3,4-furyl), 7.38 (s, 1H, H5-furyl)
1652 CO (pos. 4)

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5013
the structure of 1,3-dipropyl-arylpyrimido[2,1-f]purined-
ione, that oxygen is going to be not effective as H-bond
acceptor.33 Aromatic substituent located at N9-nitrogen
is filling the third lipophilic area, marked as P1 in Figure
8. It is essential that replacement of phenyl by heterocy-
cles (pyridyl or furane in compounds 13 and 14, respec-
tively) decreases the activity of the compound. We
suppose that the properties and capacity of P1 pocket
should be studied more intensively on the base of pyrim-
ido[2,1-f]purinediones with a variety of spacers joining
aromatic group with the main skeleton.
Beside lipophilic pockets, the model in Figure 8 consists
of two H-bond acceptors: one at the not protected purine
oxygen and the second one at the imidazole nitrogen.
Reference pharmacophore model for ligands of A1
receptor comprises also H-bond donor, it should be
located in the proximity of tetrahydropyrimidine ring.
H-Bond donor in this area of the molecule was detected
in the crystal structure of 8 and 14 (see above). It involves
endocyclic tetrahydropyrimidine carbon atom (C8) with
hydrogens of higher acidity to form C–H. . .O weak inter-
action (Fig. 5).
7. Conclusions
A series of 20 new benzyl derivatives of pyrimido[2,1-f]
purinediones was obtained. The new compounds were
tested for their adenosine receptor affinity. 1,3-Dimethyl
derivatives exhibited adenosine A2A receptor selectivity
and affinity at micromolar and submicromolar concen-
trations. 1,3-Dipropylbenzylpyrimidopurinediones have
shown submicromolar affinity and selectivity towards
A1 AR. p-Substituents of the benzyl residues, substitu-
ent R2 at the methyl linker and isosteric exchange of
the phenyl moiety were disadvantageous for A1 and
A2A adenosine receptor activity. SAR studies have re-
vealed dependence between A1 AR affinity and logP va-
lue. Pharmacophore model for A1 R ligands has been
created.
8. Experimental
8.1. Chemistry
Melting points were determined on a MEL-TEMP II
apparatus and are uncorrected. IR spectra were
recorded as KBr discs on a FT Jasco IR 410 spec-
trometer. 1H NMR spectra were recorded on a Var-
ian-Mercury 300 MHz (compounds 1, 4, 5, 15–20) or
Bruker VM 250 (compounds 3, 6–9, 11–14) or Bruker
Ac 200F (compounds 2, 10) spectrometer in CDCl3,
using TMS as an internal standard. UV spectra were
recorded on a Jasco UV/VIS V-530 apparatus in
concentration 10À5 mol/L in methanol (Table 9).
Elemental analyses (C, H, N) were within 0.4% of
theoretical values and were performed on an Elemen-
tal-Vario-EL III apparatus.
TLC data were obtained with Merck-Silica gel 60F254
aluminium sheets, using benzene/acetone 7:3 (A), ben-
zene/acetone/methanol 3:3:3 (B) and toluene/acetone

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5014
Table 6. Reaction conditions and yields of 1,3-dipropyl-9-benzylpy-
rimido[2,1-f]-purinediones obtained by microwave method
(18 mmol) NaClO3 in 14.4 ml of water was added drop-
wise during 1 h.
Compound
Reaction
time [min]
Microwave
power [W]
Yield[%]
Heating and stirring was continued lasting for 2 h. The
white solid precipitated, which was filtered off on the
next day, washed with cold water, hot water and
ethanol.
15
16
17
18
19
20
5
10
7
450
750
450
450
300
450
85
82
85
90
87
98
10
4
The yield 78%, mp 179–182 ꢁC.
Spectral data:
10
7:3 (C) as developing systems. The plates were visualized
with UV light.
UV: kmax = 279 nm.
IR [cmÀ1]: 3063 (NH), 2965–2874 (–CH2), 1705 [–CO
(pos. 2)], 1660 [–CO (pos. 4)], 760 (–CH2).
8.1.1. General procedure for the synthesis of 9-benzyl 1,
3-dimethyl-6,7,8,9-tetrahydro pyrimido[2,1-f]purine-2,4
(1H,3H)-dione (1–14). A mixture of 0.66 g (2 mmol)
of 7-(3-chloropropylo)-8-bromotheophylline37 and cor-
responding amine (4–30 mmol) was refluxed in appro-
priate solvent (propanol, butanol, methoxyethanol,
Me-Digol)) or without solvent for 4–10 h (see Table
1). After cooling the precipitate was separated, washed
with ethanol and water. All compounds were purified
by crystallization (see Table 1).
1H NMR (CDCl3) d: 0.95–1.01 (m, 6H, CH2CH2CH3);
1.68–1.88 (m, 4H, CH2CH2CH3); 4.04–4.10 (m, 4H,
CH2CH2CH3); 13.55 (s, 1H, NH).
8.1.2.2. 1,3-dipropylo-7-(3-chloropropyl)-8-bromoxan-
thine. A mixture of 15.3 g (24 mmol) of 1,3-dipropyl-8-
bromoxanthine, 5.92 g (43 mmol) of anhydrous
K2CO3, 7.88 ml (80 mmol) of 1-bromo-3-chloropro-
pane, 0.49 g (2 mmol) of TEBA and 150 ml of acetone
was refluxed for 10 h with stirring. The unreacted 1,
3-dipropyl-8-bromoxanthine and inorganic salts were
filtered off from the hot mixture. This precipitate con-
tained also a small amount of the main product. The
acetone filtrate was cooled overnight. The precipitated
product was separated. To remove unreacted 1,3-dipro-
pyl-8-bromoxanthine these two crops of product were
stirred at room temperature with 5% NaOH, filtered
off and washed with water.
8.1.2. General procedure for the synthesis of 9-benzyl 1,
3-dipropyl-6,7,8,9-tetrahydro pyrimido[2,1-f]purine-(1H,
3H)-dione (15–20). Starting materials.
8.1.2.1. 8-bromo-1,3-dipropylxanthine. A mixture of
12.15 g (50 mmol) of 1,3-dipropylxanthine,39,40 39 ml
acetic acid and 6.8 ml (75 mmol) of 40% HBr was stirred
at 50 ꢁC until dissolution was completed. The tempera-
ture was raised to 58 ꢁC and the solution of 2.04 g
Table 7. Crystal data and structure refinement details for 8 and 14
8
14
Empirical formula
Formula weight
Temperature
C18H21N5O3
355.40
293(2) K
C15H17N5O3
315.33
293(2) K
˚
1.54178 A
˚
0.71073 A
Wavelength
Crystal system
Space group
Unit cell dimensions
Triclinic
P-1
a = 5.0020(10) A
Monoclinic
P21/n
a = 5.1665(8) A
˚
˚
˚
b = 9.653(2) A
˚
b = 30.497(4) A
˚
c = 17.745(4) A
˚
c = 9.6592(13) A
a = 91.99(3)ꢁ
b = 92.18(3)ꢁ
c = 91.52(3)ꢁ
b = 94.56(1)ꢁ
3
˚
Volume
855.3(3) A
1503.0(4)
Z, Calculated density
Absorption coefficient
F(000)
2, 1.380 Mg/m3
0.797 mmÀ1
376
4, 1.394 Mg/m3
0.101 mmÀ1
664
Crystal size
0.1 · 0.1 · 0.4 mm
2.5–80.0ꢁ
0.25 · 0.3 · 0.5 mm
2.52–25ꢁ
À5:h:6, À36:k:36, À11:l:11
KM-4 with CCD detector
16,079/2647 [R(int) = 0.028]
Full-matrix-block least-squares on F2
Theta range for data index
Ranges
Diffractometer
À6:h:6, À12:k:12, 0:l:22
KM-4
7098/3549 [R(int) = 0.107]
Refl. collected/unique
Refinement method
Data/parameters
Goodness-of-fit on F2
Final R [I > 2r(I)]
Largest diff. peak and hole
Full-matrix-block least-squares on F2
3549/242
2647/212
1.06
0.91
R1 = 0.0643, wR2 = 0.1291
R1 = 0.0361, wR2 = 0.1078
À3
À3
˚
˚
0.30 and À0.32 e A
0.17 and À0.16 e A

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5015
Table 8. Atomic coordinates (·104) and equivalent isotropic displace-
ment parameters (A · 10 ) for 8 and 14
Table 9. Elemental analysis
2
3
˚
Compound
Analyses [%]
H
8
C
N
O2
O4
0.4882(5)
0.6224(5)
0.2069(7)
0.2545(5)
0.5524(5)
0.2028(5)
À0.1235(5)
0.0137(5)
0.1162(7)
0.4339(6)
0.7419(7)
0.5031(6)
0.3139(6)
0.2644(6)
0.4127(2)
0.7831(3)
1.3238(3)
0.6036(2)
0.5976(3)
0.9300(2)
1.0453(2)
0.8153(2)
0.5372(3)
0.5308(3)
0.5185(4)
0.7341(3)
0.7985(3)
1.0440(3)
0.19775(13)
0.04895(13)
0.53674(15)
0.22333(13)
0.12334(14)
0.14438(12)
0.21194(12)
0.23758(13)
0.28392(17)
0.18332(18)
0.08060(18)
0.10225(16)
0.14664(15)
0.09553(16)
0.0769(9)
0.0830(9)
0.1190(2)
0.0563(8)
0.0631(9)
0.0555(8)
0.0562(8)
0.0558(8)
0.0692(9)
0.0618(9)
0.0778(9)
0.0602(9)
0.0558(9)
0.0617(9)
1
2
Ref. 6
62.76
62.65
63.71
63.44
56.75
56.59
56.75
56.83
56.75
56.52
51.79
51.66
59.46
59.35
60.83
60.64
60.83
60.46
60.83
61.09
68.80
68.80
63.69
63.76
58.90
58.93
57.14
57.20
66.83
66.63
60.64
60.53
60.64
60.83
60.64
60.74
64.22
64.10
63.14
62.85
5.89
6.15
6.24
6.15
5.05
4.87
5.05
5.06
5.05
5.23
4.35
4.08
5.29
5.05
5.96
5.85
5.96
5.84
5.96
5.64
5.78
5.50
6.24
6.62
5.57
5.74
5.44
5.33
7.40
7.44
6.31
6.32
6.31
6.29
6.31
6.31
7.11
7.19
6.57
6.55
21.53
21.52
20.64
20.96
19.46
19.28
19.46
19.26
19.46
19.61
17.77
17.53
20.40
20.39
19.71
19.48
19.71
19.90
19.71
19.46
17.44
17.21
20.64
20.72
25.75
26.10
22.21
22.22
17.71
17.40
16.83
16.53
16.83
16.93
16.83
16.52
17.02
17.00
17.53
17.50
O18
N1
N3
N5
3
N9
4
N10
C1A
C2
5
C3A
C4
6
C5A
C6
7
14
8
O2
O4
0.9957(5)
1.1365(5)
0.5151(5)
0.7593(5)
1.0592(5)
0.7185(5)
0.3916(5)
0.5225(5)
0.6218(8)
0.9410(6)
1.2505(7)
1.0168(6)
0.8277(6)
0.7669(7)
0.5336(7)
0.4559(7)
0.5381(6)
0.7035(6)
0.2136(7)
0.2984(6)
0.2031(7)
0.3641(8)
0.5477(8)
0.11487(10)
0.02913(9)
0.22594(9)
0.13134(9)
0.07217(9)
0.08714(8)
0.12808(9)
0.14169(8)
0.16754(12)
0.10686(12)
0.04603(14)
0.06021(11)
0.08712(10)
0.05672(11)
0.05825(12)
0.10537(12)
0.11997(10)
0.12081(10)
0.16588(11)
0.20180(10)
0.21669(11)
0.25200(13)
0.25645(13)
0.3774(2)
0.7744(3)
1.0728(3)
0.5629(3)
0.5757(3)
0.9026(3)
1.0033(3)
0.7710(3)
0.4864(4)
0.4981(3)
0.5039(4)
0.7166(3)
0.7717(3)
1.0214(4)
1.1085(4)
1.1365(3)
0.8947(3)
0.6979(3)
0.9989(4)
1.0982(3)
1.2151(4)
1.2678(4)
1.1803(5)
0.0634(10)
0.0647(10)
0.0674(10)
0.0455(10)
0.0464(10)
0.0371(8)
9
O16
N1
10
11
12
13
14
15
16
17
18
19
20
N3
N5
N9
0.0451(9)
0.0412(9)
Ref. 6
N10
C1A
C2
0.0582(14)
0.0466(12)
0.0633(16)
0.0449(12)
0.0384(10)
0.0468(12)
0.0538(12)
0.0497(11)
0.0378(10)
0.0389(11)
0.0488(12)
0.0420(11)
0.0482(12)
0.0602(14)
0.0697(16)
C3A
C4
C5A
C6
C7
C8
C9A
C10A
C11
C12
C13
C14
C15
Ueq is defined as one-third of the trace of the orthogonalized Uij tensor.
Spectral data:
UV: kmax = 279.5 nm.
The spectral data are summarized in Table 5.
IR [cmÀ1]: 2965–2873 (–CH2), 1702 [–CO (pos. 2)], 1659
[–CO (pos. 4)], 750 (CH2).
Method b. A mixture of 0.39 g (1 mmol) of 1,3-dipro-
pyl-7-(3-chloropropyl)-8-bromoxanthine and 4 mmol
of benzylamines was melted in microwave oven during
4–10 min (power 300–750 W). After cooling acetoni-
trile was added and precipitate of the unreacted
benzylamine salts was filtered off, the filtrate was
evaporated under reduced pressure. To the residue
water was added and the precipitated product was col-
lected (compounds 15, 16, 18, 20). Compound 19 was
separated by dissolving the reaction mixture in ethanol
and precipitated with water. Compound 17 was iso-
lated from the reaction mixture by adding acetonitrile.
The precipitate was washed with water to remove ben-
zylamine salts.
1H NMR d = 0.92–0.99 (m, 6H, 2CH3–CH2–CH2),
1.58–1.84 (m, 4H, 2CH3–CH2–CH2), 2.28–2.37 (m,
2H, CH2–CH2–CH2Cl), 3.43–3.63 (m, 2H, CH2–CH2–
CH2Cl), 3.92–4.04 (m, 2H, 2CH3–CH2–CH2), 4.49–
4.51 (m, 2H, CH2–CH2–CH2Cl).
8.1.2.3. Cyclization. Method a. A mixture of 1 mmol
of 1,3-dimethyl- or 1,3-dipropyl-7-(3-chloropropyl)-8-
bromoxanthine with 4 mol of corresponding benzyl-
amine and 2 ml of DMF was refluxed for 5–10 h. The
products were isolated from reaction mixture by adding
water and cooling overnight in refrigerator. The crystal-
line products were filtered off and purified by dissolving
in hot ethanol and precipitated with water.
The reaction conditions and the yields are presented in
Table 6.

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5016
8.2. Adenosine receptor binding assays
in this paper have been deposited with the Cambridge
Crystallographic Data Center and allocated the deposi-
tion numbers: CCDC 623700 for 8 and CCDC 623701
for 14. Copies of the data can be obtained free of charge
on application to CCDC, 12 Union Road, Cambridge
CB2 1EW, UK (Fax: Int code + (1223)336-033;
E-mail:deposit@ccdc.cam.ac.uk).
Adenosine receptor binding assays were performed as
previously described using rat brain cortical membrane
preparations for A1 AR assays and rat brain stratial mem-
brane preparations for A2A assays.48,53,54 Frozen rat
brains (unstripped) were obtained from Pel-Freezꢂ, Rog-
ers, Arkansas, USA. For assays at human A1, A2A, A2B
and A3 ARs, CHO cell membranes containing the human
receptors were used as described.53,55 [3H]2-chloro-N6-
cyclopentyladenosine ([3H]CCPA) was used as the A1
radioligand, [3H]3-(3-hydroxypropyllo-7-methyl-8-(m-
methoxystyryl)-1-propargylxanthine ([3H]-MSX-2) as
the A2A radioligand,56 [3H]4-(2-[7-amino-2-(2-furyl)-
[1,2,4]triazolo[2,3-a]1,3,5-triazin-5-4-(amino)-ethyl)phenol
([3H]ZM241385) as the A2B receptor radioligand57 and
[3H]-phenyl-8-ethyl-4-methyl-(8R)-4,5,7,8-tetrahydro-1H-
imidazo-[2,1-i]-purine-5-one ([3H]PSB-11) as the A3 AR
radioligand.22 Initially, a single high concentration of
compound (25 lM at A1 and A2A, 10 lM at A2B and A3
receptors) was tested in three (A1, A2A) or two (A2B, A3)
independent experiments. For potent compounds, curves
were determined using 6–7 different concentrations of test
compounds spanning three orders of magnitude. Data
were analyzed using the PRISM program version 3.0
(Graph Pad, San Diego, CA, USA).
8.4. Computational procedures
The starting models of molecules were based on crystal-
lographic data for 8 and 14 using the PCMOD.6
program.43 The geometry of the molecules was opti-
mised with MOPAC 6.0 using AM1 Hamiltonians in
an aqueous environment (dielectric constants equal
78.4).44
The values of logP and logD (at pH 7.4) for the
compounds investigated were calculated by means of
PALLAS (version 1.2) program.49 Multiple Linear
Regression equations were computed by means of the
QSAR-PC:PAR program written by R.C. Coburn.50
Acknowledgments
This work was supported by Polish State Committee for
Scientific Research (Grant No. 6 P05F 024 21) and by
the Deutsche Forschungsgemeinschaft (GRK 677).
C.E.M. is grateful for support by the Fonds der Chem-
ischen Industrie and the Bundesministerium fur Bildung
¨
und Forschung (BMBF), Germany.
8.3. X-ray structure analysis of 8 and 14
Crystal data for 8: C18H21N5O3; mol. mass 355.40;
˚
triclinic; space group: P-1; a = 5.0020(10) A; b =
˚
˚
9.653(2) A; c = 17.745(4) A; a = 91.99(3)ꢁ, b = 92.18(3)ꢁ;
3
˚
c = 91.52(3)ꢁ, V = 855.3(3) A ; z = 2; dx = 1.380 Mg/
m3; l = 0.797 mmÀ1; F(000) = 376; final R = 0.064 for
3549 reflections [I > 4r(I)].
Supplementary data
Crystal data for 14: C15H17N5O3; mol. mass 315.33;
˚
monoclinic; space group: P21/n; a = 5.1665(8) A; b =
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmc.
2007.04.018.
˚
˚
30.497(4) A; c = 9.6592(13) A; b = 94.56(1)ꢁ; V =
3
3
À1
;
˚
1503.0(4)A ; z = 4; dx = 1.394 Mg/m ; l = 0.101 mm
F(000) = 664; final R = 0.036 for 2647 reflections
[I > 4r(I)].
References and notes
1. Fredholm, B. B.; Arslan, G.; Haldner, L.; Kull, B.;
Schulte, G.; Wasserman, W. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 2000, 362, 364.
2. Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz,
K.-N.; Linden, J. Pharm. Rev. 2001, 53, 527.
3. Ralevic, V.; Burnstock, G. Pharmacol. Rev. 1998, 50, 413.
4. Poulsen, S.-A.; Quinn, R. J. Bioorg. Med. Chem. 1998, 6,
619.
The crystals of 8 and14 were obtained by slow evapora-
tion of ethanol solution. The measurement of the crystal
8 was performed on a Kuma4CCD j-axis diffractometer
with graphite-monochromated MoKa radiation at room
temperature. The measurement of the crystal 14 was
performed on a KM-4 four-cyclic diffractometer with
graphite-monochromated CuKa radiation at room tem-
perature. The data were corrected for Lorentz and
polarization effects. No absorption correction was ap-
plied. All crystallographic data and experimental details
are collected in Table 7. The structures were solved by
direct methods58 and refined using SHELXL.58 The
full-matrix least-squares refinement was based on F2.
The positions of all H-atoms were found from electron
density Dq map and refined in raiding model with the
isotropic displacement parameters of 1.5 times the
respective Ueq values for the parent-atoms. Atomic scat-
tering factors were those as in SHELXL.59 Atomic coor-
dinates are gathered in Table 8. Crystallographic data
(excluding structural factors) for the structure reported
`
5. Ferre, S.; Fredholm, B. B.; Morelli, M.; Popoli, P.; Fuxe,
K. Trends Neurosci. 1997, 20, 482.
`
6. Ferre, S.; Popoli, P.; Gimenez-Llort, L.; Rimondini, R.;
Muller, C. E.; Stro¨mberg, I.; Ogren, S. O.; Fuxe, K.
¨
´
¨
Parkinsonism Relat. Disord. 2001, 7, 235.
7. Moreau, J.-L.; Huber, G. Brain Res. Rev. 1999, 31, 65.
8. Muller, C. E. Drugs Future 2000, 25, 1043.
¨
9. Ongini, E.; Schubert, P. Drug Dev. Res. 1998, 45, 387.
10. Impagnatiello, F.; Bastia, E.; Ongini, E.; Monopoli, A.
Emerg. Ther.Targets 2000, 4, 635.
11. Ribeiro, J. A.; Sebastia˜o, A. M.; deMendonc¸a, A. Prog.
Neurobiol. 2003, 68, 377.
12. De Sarro, G.; De Sarro, A.; Di Paola, E. D.; Bertorelli, R.
Eur. J. Pharmacol. 1999, 371, 137.

´
A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 5003–5017
5017
´
37. Rockitt, S.; Wartchow, R.; Duddeck, H.; Drabczynska,
13. Dunwiddie, T. V. Adv. Neurol. 1999, 79, 1001.
14. McGaraughty, S.; Cowart, M.; Jarvis, M. F.; Berman, R.
F. Curr. Top. Med. Chem. 2005, 5, 43.
´
A.; Kiec-Kononowicz, K. Z. Naturforsch. 2001, 56B, 319.
38. Eckstein, M.; Łoson, W. Dissert. Pharm. Pharmaceut.
´
15. Hess, S. Expert. Opin. Ther. Patents 2001, 11, 1.
1968, 20, 35.
´
16. Kiec-Kononowicz, K.; Drabczynska, A.; Pe˛kala, E.;
Michalak, B.; Muller, C. E.; Schumacher, B.; Karolak-
¨
´
39. Papesh, V.; Grove, M.; Schroeder, E. F.; Serle, G. D.
United States Patent 1952, 2, 602.795; Chem. Abstr. 1953,
47, 4920f.
40. Kramer, G. L.; Garst, J. E.; Mitchel, S. S.; Wells, J. N.
Biochemistry 1997, 16, 3316–3321.
41. Bergmann, F.; Dickstein, S. J. Am. Chem. Soc. 1955, 77,
691.
ˇ
42. Rybar, A.; Antos, K. Coll. Czech. Chem. Commun. 1970,
35, 1415.
43. PCMODEL 6 Molecular Modeling Software for the IBM
PC, Serena Software, Bloomington, 1996.
44. MOPAC program version 6, QCPE No. 455; Department
of Chemistry, Indiana University, Bloomington, IN
47405, 1990.
Wojciechowska, J.; Duddeck, H.; Rocktt, S.; Wartchow,
R. Pure Appl. Chem. 2001, 73, 1411.
17. Da Settimo, F.; Primofiore, G.; Taliani, S.; Marini, A. M.;
La Motta, C.; Novellino, E.; Greco, G.; Lavecchia, A.;
Trincavelli, L.; Martini, C. J. Med. Chem. 2001, 44,
316.
18. Baraldi, P. G.; Tabrizi, M. A.; Bovero, A.; Avitabile, B.;
Preti, D.; Fruttarolo, F.; Romagnoli, R.; Varani, K.;
Borea, P. A. Eur. J. Med. Chem. 2003, 38, 367.
19. Ferrarini, P. L.; Betti, L.; Cavallini, T.; Giannaccini, G.;
Lucacchini, A.; Manera, C.; Martinelli, A.; Ortore, G.;
Saccomanni, G.; Tuccinaroli, T. J. Med. Chem. 2004, 47,
3019.
45. Muller, C. E.; Maurinsh, J.; Sauer, R. Eur. J. Pharm. Sci.
¨
2000, 10, 259.
20. Yan, L.; Muller, C. E. J. Med. Chem. 2004, 47, 1031.
¨
21. Muller, C. E. Curr. Top. Med. Chem. 2003, 3, 445.
46. Klotz, K.-N.; Lohse, M. J.; Schwalbe, U.; Cristalli, G.;
Vittori, S.; Grifantini, M. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1989, 340, 679.
47. Gao, Z. G.; IJzerman, A. P. Biochem. Pharmacol. 2000,
60, 669.
48. Hayallach, A. M.; Sandoval-Ramirez, J.; Reith, U.; Preiss,
¨
22. Muller, C. E.; Diekmann, M.; Thorand, M.; Ozola, V.
¨
Bioorg. Med. Chem. Lett. 2002, 12, 501.
23. Shimada, J.; Suzuki, F.; Nokada, H.; Ishii, S.; Ichikawa,
S. J. Med. Chem. 1992, 36, 2342.
24. Nonaka, H.; Ichimura, M.; Takeda, M.; Nonaka, Y.;
Shimada, J.; Suzuki, F.; Yamaguchi, K.; Kase, H. Eur.
J. Pharmacol., Mol. Pharm. Sect. 1994, 267, 335.
25. Jacobson, K. A.; Gallo-Rodriguez, C.; Melman, N.;
Fischer, B.; Maillard, M.; van Bergen, A.; van Galen, P.
J. M.; Karton, Y. J. J. Med. Chem. 1993, 36, 1333–1342.
26. Jackson, E. K.; Herzer, W. A.; Suzuki, F. J. Pharmacol.
Exp. Ther. 1993, 267, 1304.
B.; Schumacher, B.; Daly, J. W.; Muller, C. E. J. Med.
Chem. 2002, 45, 1500.
¨
49. PALLAS for Windows 1.2, CompuDrug Chemistry LTD,
1995.
50. QSAR-PC program, R. A. Coburn; BIOSOFT, Cam-
bridge, UK.
51. Da Settimo, F.; Primofiore, G.; Taliani, S.; Marini, A.
M.; La Motta, C.; Novellino, E.; Greco, G.; Lavecchia,
A.; Trincavelli, L.; Martini, C. J. Med. Chem. 2001, 44,
316.
27. Nonaka, Y.; Shimada, J.; Nonaka, H.; Koike, N.; Aoki,
N.; Kobayashi, H.; Kase, H.; Yamaguchi, K.; Suzuki, F.
J. Med. Chem. 1993, 36, 3731.
28. Hockemeyer, J.; Burbiel, J. C.; Muller, C. E. J. Org. Chem.
2004, 69, 3308.
29. Muller, C. E.; Sandoval-Ramirez, J.; Schobert, U.; Geis,
¨
52. Bondavalli, F.; Botta, M.; Bruno, O.; Ciacci, A.; Corelli,
F.; Fossa, P.; Lucacchini, A.; Manetii, F.; Martini, C.;
Menozzi, G.; Mosti, L.; Ranise, A.; Schenone, S.; Tafi, A.;
Trincavelli, M. A. J. Med. Chem. 2002, 45, 4875.
¨
U.; Frobenius, W.; Klotz, K. N. Bioorg. Med. Chem. 1998,
6, 707.
53. Muller, C. E.; Thorand, M.; Quirshi, R.; Dieckmann, M.;
¨
30. Sauer, R.; Maurinsh, J.; Reith, U.; Fulle, F.; Klotz, K. N.;
¨
31. Knutsen, L. J. S.; Weiss, S. M. Curr. Opin. Invest. Drugs
2001, 2, 668.
32. Rabasseda, X.; Sorbera, L. A.; Martin, L.; Leeson, P. A.;
Castan˜er, J. Drugs Future 2001, 26, 20.
´
33. Drabczynska, A.; Muller, C. E.; Lacher, S. K.; Schum-
¨
acher, B.; Karolak-Wojciechowska, J.; Nasal, A.; Kawc-
Jacobson, K. A.; Padgett, W. L.; Daly, J. W. J. Med.
Chem. 2002, 45, 3440–3450.
54. Hess, S.; Muller, C. E.; Frobenius, W.; Reith, U.; Klotz,
¨
K.-N.; Eger, K. J. Med. Chem. 2000, 43, 4636–4646.
55. Klotz, K.-N.; Hessling, J.; Hegler, J.; Owman, C.; Kull,
B.; Fedholm, B. B.; Lohse, M. J. Naunyn-Schmiedeberg’s
Arch. Pharmacol. 1998, 357, 1–9.
¨
Muller, C. E. J. Med. Chem. 2000, 43, 440.
56. Muller, C. E.; Maurinsh, J.; Sauer, R. Eur. J. Pharm. Sci.
¨
2000, 10, 259–265.
´
zak, P.; Yuzlenko, O.; Pe˛kala, E.; Kiec-Kononowicz, K.
Bioorg. Med. Chem. 2006, 14, 7258.
34. Sauer, R.; Maurinsh, J.; Reith, U.; Fulle, F.; Klotz, K.-N.;
¨
57. Ji, X.-D.; Jacobson, K. A. Use of the triazolotriazine
[3H]ZM 241385 as a radioligand at recombinant human
A2B adenosine receptors. Drug Des. Discov. 1999, 16, 217–
226.
58. Sheldrick, G. M. SHELXL-93. A FORTRAN-77 Program
for the Refinement of Crystal Structures from Diffraction
Data; University of Go¨ttingen: Germany, 1993.
59. Sheldrick, G. M. SHELXTL PCMT; Siemens Analytical
X-Ray Instruments Inc.: Madison, Wisconsin, USA, 1990.
Muller, C. E. J. Med. Chem. 2000, 43, 440.
´
Wojciechowska, J.; Michalak, B.; Pe˛kala, E.; Kiec-Kon-
¨
35. Drabczynska, A.; Schumacher, B.; Muller, C. E.; Karolak-
¨
´
onowicz, K. Eur. J. Med. Chem. 2003, 38, 397.
´
36. Drabczynska, A.; Muller, C. E.; Schumacher, B.; Hinz, S.;
¨
Karolak-Wojciechowska, J.; Michalak, B.; Pe˛kala, E.;
´
Kiec-Kononowicz, K. Bioorg. Med. Chem. 2004, 12, 4895.