Phenylethyl-substituted pyrimido[2,1-f]purinediones and related compounds: Structure-activity relationships as adenosine A1 and A2A receptor ligands

Bioorganic & Medicinal Chemistry 15 (2007) 6956–6974

Phenylethyl-substituted pyrimido[2,1-f]purinediones

and related compounds: Structure–activity relationships

as adenosine A₁and A_2Areceptor ligands

a

b

´

Anna Drabczynska, Christa E. Muller, Anke Schiedel, Britta Schumacher,

¨

c

a

´

Janina Karolak-Wojciechowska, Andrzej Fruzinski, Weronika Zobnina,

a

a,

*

´

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

_

Institute of General and Ecological Chemistry, Technical University of Ło´dz´, Zwirki 36, Pl 30-924 Ło´dz´, Poland

c

Received 22 January 2007; revised 20 July 2007; accepted 27 July 2007

Available online 22 August 2007

Abstract—The synthesis of N-(un)substituted-phenylalkylpyrimido[2,1-f]purinediones was performed starting with 7-(3-chloropro-

pyl)-8-bromotheophylline and 7-(3-chloropropyl)-8-bromo-1,3-dipropylxanthine. Compounds with unsubstituted or substituted

ethylene spacer to an aromatic ring were synthesized. Additionally variations in the spacer—elongation of the linker containing

more than two atoms, introduction of a double bond or heteroatoms were performed. Physicochemical properties of the synthesized

compounds were described. The obtained compounds envisaged as sterically ﬁxed and conﬁgurationally stable analogs of 8-styryl-

xanthines, were evaluated for their aﬃnity to adenosine A₁and A_2Areceptors, the receptor subtypes that are predominant in the

brain. Selected compounds were also investigated for the aﬃnity to the A_2Band A₃receptor subtypes. It was stated that phenylethyl

pyrimido[2,1-f]purinediones and their analogs with variations of the ethylene spacer (substituted or extended) exhibit micromolar or

submicromolar aﬃnity for A_2AARs (adenosine receptors); for example compound 2Ac with p-hydroxy substituent displayed a K_i

value of 0.23 lM at the rat A_2Areceptor. In comparison to the previously obtained phenyl and benzyl pyrimido[2,1-f]purinediones

compounds with a shorter spacer, phenethyl derivatives were optimal for A_2AAR. The kind of substituent at the aromatic ring was

important for the aﬃnity. Oxygen and nitrogen atoms in the spacer resulted frequently in a slight decrease of the A_2AAR aﬃnity,

introduction of more heteroatoms into the spacer—in carbamates—caused distinctly negative eﬀect on the activity. In this series of

compounds more frequently the adenosine A₁activity was observed, also in submicromolar range as for dipropyl derivative 2Ba

with K_ivalue of 0.62 lM at the rat A_2AAR. 3D-QSAR models were developed for the compounds presented in this paper as well

as in the previous publications showing activity at adenosine A₁and A_2AARs. It was concluded that for the activity at adenosine A₁

and A_2Areceptors lipophilicity, steric eﬀects along with the molecule’s electrostatic surface properties had greatest value. Chosen

compounds were evaluated in vivo as anticonvulsants in MES, scMet tests and examined for neurotoxicity. Contrary to previously

obtained phenyl and benzyl pyrimido[2,1-f]purinediones, all tested compounds were inactive as anticonvulsants.

1. Introduction

subtypes: A₁, A_2A, A_2B, and A₃have been cloned and

pharmacologically characterized. These receptors be-

long to the superfamily of G-protein coupled recep-

tors.^1,2The A₁and A_2AAR subtypes are high-aﬃnity

subtypes since they are usually activated by low, nano-

molar concentrations of adenosine.^1,3Adenosine A₁

receptors are widely distributed in the central nervous

system and in peripheral tissues. The role of adenosine

as neuroprotective agent during hypoxia and ischemic

conditions seems to be mediated by the adenosine A₁

receptors. These receptors have been shown to be

Adenosine modulates a variety of important physio-

logical processes and exhibits central nervous system

depressant, cardiodepressant, antidiuretic, and immuno-

modulatory eﬀects.¹To date, four adenosine receptor

Keywords: Adenosine A₁; A_2Areceptor ligands; Pyrimido[2,1-f]purine-

diones; Tricyclic xanthine derivatives; Anticonvulsant activity.

*

Corresponding author. Tel./fax: +48 12 657 04 88; e-mail:

mfkonono@cyf-kr.edu.pl

doi:10.1016/j.bmc.2007.07.051

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6957

involved in sedative, antiseizure, and anxiolytic eﬀects.⁴

O

R¹

Selective A₁antagonists may be useful as kidney-protec-

tive diuretics, for the treatment of congestive heart fail-

ure due to their diuretic and positive inotropic eﬀects,

and have potential for the treatment of brain diseases,

such as depression or dementia.^5–8The adenosine A_2A

receptors play an important role in regulating smooth

and well-coordinated movement in part by modulating

the activity of dopamine-sensitive neurons. Adenosine

A_2Areceptors are located in brain primarily in the

striatum where they are colocalized with dopamine D₂

receptors. There is now accumulating evidence that

adenosine A_2Areceptor antagonists may provide a novel

therapy for the treatment of Parkinson’s disease with

lower risk of dyskinesias and may exhibit neuroprotec-

tive eﬀects.^8–10

N

spacer = 2 - 6 atoms

aryl

O

N

R¹

N

spacer

III

Figure 1. General structure of the obtained pyrimido[2,1-f]

purinediones.

(SARs) of the new series of compounds, variations in

the spacer (elongation of the linker containing more

than two atoms, introduction of a double bond or

heteroatoms) were performed. 1,3-Dipropyl-substituted

structures (R¹= propyl) were compared with dimethyl

derivatives (R¹= Me).

In our eﬀorts to develop new ligands for A₁and A_2A

adenosine receptors, oxygen- or nitrogen-containing

annelated theophylline derivatives with the following

(I, II) structures (Table 1) were obtained.^11–16

2. Chemistry

The synthesis of the target compounds was performed as

depicted in Figure 2. The synthesis of N-phenylalkylpy-

rimido[2,1-f]purinediones was accomplished as shown in

Figure 2a. As starting materials 7-(3-chloropropyl)-8-

bromotheophylline (1A) and 7-(3-chloropropyl)-8-bro-

mo-1,3-dipropylxanthine (1B) were used, which could

be obtained as previously described by us.^15–17They

were cyclisized with phenylalkylamines under various

reaction conditions regarding the amount of amine,

the reaction medium, and the reaction time (see Table

2a). Two amines, 4-hydroxyphenylethylamine (tyra-

mine) and 2-amino-1-phenyl-1-propanol (norephe-

drine), were obtained from their hydrochlorides by

alkalization of water solutions of their hydrochlorides

to pH 9 by 10% aqueous sodium carbonate (tyramine),

or 15% aqueous sodium hydroxide (norephedrine).

Acetyl derivatives 2Ag and 2Ah were prepared by

Selective adenosine A_2Areceptor antagonists were iden-

tiﬁed especially among the series of pyrimido[2,1-f]pur-

inediones (IIa, IIb). Replacement of dimethyl

(R⁴= CH₃) with dipropyl substituents in the purine ring

improved the aﬃnity to A₁, A_2A, and A₃ARs and there-

fore reduced A_2A-selectivity (IId, IIc, IIg). Derivatives

with good adenosine A₁receptor aﬃnity (Ib, IId, IIf,

and IIg) were identiﬁed as well^15,16(Table 1).

As a continuation of our research, pyrimido[2,1-f]purine-

diones of the general structure III connected by an

unsubstituted or substituted ethylene spacer to an aro-

matic ring (lead compound: 1,3-dimethyl-9-phenethyl-

pyrimido[2,1-f]purine 2Aa) were synthesized (Fig. 1).

In order to investigate structure–activity relationships

Table 1. Oxygen- or nitrogen-containing annelated theophylline analogs. Adenosine A₁/A_2Areceptors aﬃnity of chosen tricyclic annelated

theophylline derivatives of type I and II^12–16in comparison with the standard compounds caﬀeine and KW-6002

R¹

R²

O

R⁴

O

(CH₂)n

H₃C

O

N

O

N

R³

N

R⁴

n= 1, 2, 3

CH₃

II

I

Compound

R¹

R²

R³

R⁴

A₁K_iSEM (lM)

A_2AK_iSEM (lM)

versus [³H]CCPA

versus [³H]MSX-2

Ia

Ib

H

Octyl

C₆H₅CH₂

—

>25

2.07 0.15

0.998 0.05

8.39 1.76

O

Ic

H

—

>25

2.67 0.54

IIa

IIb

—

2-Naphtyl

4-F-C₆H₄

2-Naphtyl

4-Cl-C₆H₄

CH₃

>25

0.219 0.007

0.147 0.02

0.398 0.111

0.376 0.095

0.699 0.016

1.29 0.31

IIc

IId

CH₃CH₂CH₂

CH₃

1.285 0.32

0.36 0.06

>25

IIe

IIf

2-OCH₃C₆H₄CH₂

3-ClC₆H₄CH₂

4-ClC₆H₄CH₂

CH₃CH₂CH₂

0.089 0.029

0.23 0.15

18.8 5.6

0.230 0.030

IIg

Caﬀeine

KW-6002

0.87 0.41

32.5 8.0

0.00515 0.00025

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6958

O

R¹

N

R¹

N

Cl

R³

N

m

+

O

H₂N

Ar

N

R¹

Br

N

R¹

N

R³

Ar

N

R⁴

R²

m

1

2

m = 0 , 1

1

1A

1B

R

=

2A

2B

R

=

CH₃

CH₂CH₂CH₃

1

R

CH₂CH₂CH₃

O

H₃C

O

N

O

N

H₃C

4

OH

3

Br

HO

or

R

HO

R

( )z

Br

R

O

H₃C

O

H₃C

N

O

R

N

H₃C

R

H₃C

5

O

6

O

( )z

O

H₃C

N

O

H₂N

+

3

N

H₃C

R

NH

7

R

O

H₃C

O

N

R

NCO

+

4

N

H₃C

O

8

R

NH

Figure 2. Synthesis of the pyrimido[2,1-f]purinediones containing spacer between tricyclic skeleton and (un)substituted aromatic ring. (a) Spacer-

consisting of carbon (un)branched chain. (b) Spacer-containing one heteroatom (oxygen). (c) Spacer-containing one heteroatom (nitrogen).

(d) Spacer-containing more heteroatoms. Some products (7) were obtained under microwave irradiation.

acetylation of the hydroxyl compounds 2Ae and 2Af

with acetic anhydride.

The compounds 5b and 5d were obtained by the reaction

of 3 with the appropriate sodium phenolate (generated

with sodium hydride from the phenol) in acetonitrile

using [15]-crown-5 as a phase-transfer catalyst (Table

2b, method A). Similar reaction conditions were used

for the synthesis of the compounds 6a–6d. The ethers

were obtained starting from 1,3-dimethyl-9-hydroxy-

ethyl-6,7-dihydro-6,7,8,9-tetrahydropyrimido[2,1-f]pur-

ine-2,4(1H,3H)dione (4),¹⁹transformed with sodium

The compound 5a with a phenoxyalkyl moiety (Fig. 2b)

was obtained by Williamson’s reaction of 9-bromoethyl-

1,3-dimethyl-6,7-dihydro-6,7,8,9-tetrahydropyrimido[2,1-f]

purine-2,4(1H,3H)dione (3)¹⁸with sodium salt of

p-chlorophenol (prepared ex tempore from p-chlorophe-

nol) (Table 2b, method C).

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6959

Table 2a. Reaction conditions and physicochemical properties of the phenylalkylpyrimido [2,1-f]purinediones

O

R¹

N

O

N

R¹

N

R²

N

Compound R¹

R²

Mp (ꢁC) Yield (%) Reaction medium

excess of amine

Reaction Crystallization

time (h)

TLC R_f

solvent

2Aa

2Ab

2Ac

2Ad

2Ae

2Af

CH₃

195–197

203–206

175–178

233–235

199–202

216–218

58

88

85

90

80

67

Propanol, 1.6

Propanol, 2

5

50% Ethanol

0.64 (1)^a

0.45 (1)

0.34 (1)

0.45 (1)

CH₃

5

Butanol

Ethanol

Butanol

Ethanol

OCH₃

OH

Methoxyethanol, 2

—, 7

10

5

Cl

Methoxyethanol, 2

10

HO

CH₃

Methoxyethanol 0.44 (1)

HO

b

O

2Ag

2Ah

CH₃

191–193

234–235

98

94

Acetic anhydride, 25

5

Propanol

0.56 (1)

O

H₃C

CH₃

b

Methoxyethanol 0.36 (1)

O

H₃C

OCH₃

2Ai

2Aj

CH₃

164–166

175–177

91

76

Propanol, 2

Butanol, 2

5

6

Isopropanol

0.31 (1)

0.10 (1)

OCH₃

N

2Ak

2Al

CH₃

230–232

159–161

53

Methoxyethanol, 2

Propanol, 1.4

10

Propanol

0.51 (1)

0.70 (1)

N

H

50% Ethanol

OCH₃

2Bb

2Ba

124–125

125–126

83

90

DMF, 3

DMF, 5

5

Ethanol/H₂O

0.57 (1)

0.66 (1)

2Bc

164–166

96

Methoxyethanol, 2

10

Propanol

0.65 (1)

N

H

^aDeveloping system (1).

^bExcess of acetic acid anhydride.

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6960

Table 2b. Reaction conditions and physicochemical properties of the derivatives containing heteroatoms in the spacer connecting tricyclic skeleton

with (un)substituted aromatic ring

O

R¹

N

O

N

R¹

N

R²

N

Compound

R¹

R²

Mp (ꢁC)

Method yield (%)

C (46)

Reaction time (h)

20

Crystallization solvent

EtOH

TLC R_f

5a

CH₃

146–148

0.47 (1)

O

Cl

5b

CH₃

155–158

A (33)

5.5

50% EtOH

0.29 (2)

H₃C

O

5c

5d

CH₃

192–193

149–151

B (18)

6

50% EtOH

0.15 (3)

N

H

A (10) B (48)

AcOEt/CHCl₃(1:1)

0.40 (2)^a

O

CH₃

6a

6b

6c

CH₃

132–135

107–110

104–106

D (19)

D (25)

D (16)

6

7

50% EtOH

0.27 (3)

0.16 (2)

0.21 (2)

O

Cl

F

O

6d

7a

CH₃

158–160

280–285

D (16)

E (22)

50% EtOH

EtOH

0.19 (2)

0.30 (4)

NH

7b

8a

CH₃

156–159

234–236

E (13)

F (78)

EtOH

0.32 (4)

0.64 (5)

F

O

NH

10

O

NH

8b

CH₃

250–252

F (70)

12.5

EtOH

0.64 (5)

Cl

O

8c

8d

CH₃

262–266

264–266

F (67)

10

11

0.60 (5)

0.62 (5)

O

F

NH

Cl

^aDeveloping system (1)–(4) or (5).

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6961

hydride to the required ethanolate, and reacted with the

appropriate (un)substituted benzyl chlorides or bromides

in dry toluene in the presence of [15]-crown-5 as a catalyst

(Table 2b, method D). For compounds 5c and 5d, a

Mitsunobu reaction protocol was also used starting with

(4) and the appropriate phenol derivative.²⁰The amines

7a and 7b (Fig. 2c) were obtained as products of the reac-

tion of 3 and appropriate (un)substituted benzylamines

by melting the reagents upon microwave irradiation

under solvent-free conditions^21,22(Table 2b, method E).

The reaction of 4 and (un)substituted phenyl isocyanates

(Fig. 2d), carried out in acetonitrile, yielded the carba-

mates 8a–8d (Table 2b, method F).

ture, after small modiﬁcations all remaining molecules

can be roughly constructed and used as a starting point

for geometry optimization.

The compound 6d consists of the main tricyclic ring sys-

tem and aromatic substituent joined to the core of the

molecule via long (ACH₂ACH₂AOACH₂ACH@CHA)

spacer at N9 atom (Fig. 3). The atoms C6, C7, and C8

from the six-membered pyrimidine ring and the atoms

of C14AC16@C17APh from the ﬂexible chain—in the

crystal—are disordered with structure occupation fac-

tors (s.o.f.s) equal to 0.42 and 0.62, respectively, for

both parts mentioned above. Although the parts of dis-

ordered spacer exhibit the opposite signs of the torsion

angles value, the molecule adopts an extended confor-

mation (Fig. 3). The undulated six-membered pyrimi-

dine ring adopts a half-chair conformation. In both

disordered residues, C7 atom is in ﬂap position and is lo-

cated above or below the main ring plane deﬁned by the

remaining ring atoms in the two disordered species. In

the crystal, the molecules are joined in dimmers with

The reaction conditions and physicochemical properties

of the obtained compounds are summarized in Tables 2a

and 2b.

The structures of the synthesized compounds were con-

ﬁrmed by elemental analyses, as well as UV, IR and H

1

NMR spectral analysis. In UV spectra of pyrimi-

do[2,1-f]purinediones, a bathochromic shift of k_maxfrom

ca 275 nm to about 300 nm, typical for 8-aminoxanthine

derivatives, was observed.²³Examined compounds

showed IR absorption bands typical for xanthine deriv-

atives²⁴and in ¹H NMR adequate shifts could be

observed.

˚

the weak H-bond of C6AH6B. . .O4 (3.319 A).

4. Pharmacology

All compounds were tested in vitro in radioligand binding

assays for the aﬃnity to A₁and A_2Aadenosine receptors

of the rat brain cortical membrane (A₁), and the rat brain

striatal membrane (A_2A) preparations. Some selected

compounds were further tested for their aﬃnity to the hu-

man recombinant adenosine A₁, A_2A, A_2B, and A₃recep-

tors stably expressed in Chinese hamster ovary (CHO).

The following radioligands were used: [³H]2-chloro-N⁶-

cyclopentyladenosine ([³H]CCPA) for A₁, [³H]1-propar-

gyl-3-(3-hydroxypropyl)-7-methyl-8(m-methoxystyryl)xan-

thine ([³H]MSX-2)^25,26for A_2A, [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)²⁷or [³H]8-((4-(2-hydroxyethylami-

no)2-oxo-ethoxy)phenyl)-1-propylxanthine [³H]PSB-

298^28,29for A_2B, and [³H]2-phenyl-8-ethyl-4-methyl-

(8R)-4,5,7,8-tetrahydro-1H-imidazo[2,1-i]purine-5-one

([³H]PSB-11)³⁰for A₃binding studies.

3. X-ray structure analysis of compound 6d

In all derivatives described in this paper, ﬂexible chain

(spacer) joins main molecule skeleton and aromatic ring.

These spacers may be classiﬁed as aliphatic or heteroali-

phatic ones. In the aliphatic (or heteroaliphatic) chain,

theoretically, each bond possesses rotational freedom

and the number of possible low energy conformations

depends on the number and nature of the atoms. For

that reason, from the structural viewpoint, the spacer

seemed to be a particularly interesting element of the

studied species. Having that in mind, it was decided to

study the structure of the molecule 6d with the most

complex and the longest spacer. Basing on that struc-

Figure 3. Molecules of 6d joined in dimmer by O4. . .H6b–C6 weak H-bonds. Disordered atoms (C6, C7, and C8) and disordered substituent at N9

atom are depicted.

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6962

Based on our former observations that tricyclic purine-

diones can possess anticonvulsant activity,^12–14the com-

pounds 2Aa–2Al, 5a–d, and 8a–d were also evaluated

in vivo as potential anticonvulsants by the ADD (Anti-

epileptic Drug Development) Program of the National

Institute of Neurological Disorders and Stroke

(NINDS) according to ASP (Antiepileptic Screening

Project) protocols. Compounds were injected intraperi-

toneally as a suspension in 0.5% methylcellulose into

the mice and evaluated in the preliminary screenings

with at least three dose levels (30, 100, and 300 mg/kg

at 0.5 and 4 h time periods). Phase I of the evaluation

included three tests: maximal electroshock (MES), sub-

cutaneous pentylenetetrazole (scMet) seizure tests, and

the rotorod test for neurological toxicity (Tox). The tests

were described in detail by Stables and Kupferberg.^31–33

some compounds, namely 5b, 5d, and 8d, the mice

6-Hz test was performed at dose 100 mg/kg. A mice

6 Hz screening is an alternative electroshock model in

which a 6-Hz electrical stimulus is administered for a

prolonged period. The 6-Hz model appears to be sensi-

tive to a diﬀerent spectrum of anticonvulsants than in

the MES test and may detect agents that are ineﬀective

in the standard anticonvulsant screening models.^34,35

5. Biological results

5.1. In vitro tests: structure–activity relationships

The results of the radioligand binding assays at the

adenosine receptors are presented in Table 3a for the

compounds with a (un)branched carbon chain as a

spacer, in Table 3b for dipropyl derivatives, and in

Table 3c for the compounds with spacers containing

heteroatoms.

The MES assay is a model for generalized tonic–clonic

seizures, it identiﬁes these compounds which prevent sei-

zure spread and has predictive value for the agents of

therapeutic potential in the management of grand mal

epilepsy. The scMet test is a model that primarily iden-

tiﬁes compounds that raise seizure threshold; these com-

pounds are likely to be eﬀective against petit mal. For

The results in the group of arylalkyl derivatives showed

compounds with preferably and the most expressed

aﬃnity for the rat adenosine A_2Areceptors. Elongation

Table 3a. Adenosine receptors aﬃnities of the pyrimido[2,1-f]purinediones with (un)substituted alkyl spacer between tricyclic ring system and

aromatic substituent

O

H₃C

N

O

N

R³

Ar

N

H₃C

R²

m

2A

Compound

m

R²

R³

Ar

K_iSEM (lM)

Rat A₁(human A₁)

Rat A_2A(human A_2A

)

Human A_2B

Human A₃

>10 (20%)^b

Caﬀeine

KW-6002

2Aa

18.8 5.6^a

32.5 8.0^a

0.23 0.03^a

>25 (31%)^a

(>10 (44%))^a

3.85 0.43^a

ca. 25 (46%)^a

(P25 (34%))^a

6.47 1.15^a

ca. 25 (49%)^a

ca. 25 (45%)^a

3.40 0.64^a

ca. 25 (50%)^a

P25 (33%)^a

0.00515 0.00025^a

0.32 0.03^a

(2.89 0.75)^a

0.37 0.02^a

0.23 0.01^a

(0.63 0.35)^a

0.48 0.11^a

1.11 0.26^a

2.16 0.11^a

1.22 0.45^a

1.00 0.38^a

1.21 0.19^a

0

H

C₆H₅

ca. 10 (57%)^b,c

2Ab

2Ac

0

H

4-OCH₃C₆H₄

4-OHC₆H₄

nd

7.20 0.60^b,c

nd

>10 (9%)^b

2Ad

2Ae

2Af

2Ag

2Ah

2Ai

0

H

OH

4-ClC₆H₄

C₆H₅

nd

H

CH₃

H

OH

OCOCH₃

H

C₆H₅

3,4-di-OCH₃C₆H₃

H

N

2Aj

0

H

>25 (22%)^a

>25 (37%)^a

13.4 4.7^b,c

ꢀ10 (0%)^b

2Ak

2Al

0

1

H

ca. 25 (41%)^a

(>25 (21%))^a

0.33 0.07^a

ca. 10 (56%)^b,c

>10 (21%)^b

(4.56 0.81)^a

N

H

C₆H₅

1.99 0.46^a

1.10 0.14^a

nd

nd, not determined.

^aTested at 25 lM.

^bTested at 10 lM.

^cTested versus [³H]ZM-241385.

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6963

Table 3b. Adenosine receptors aﬃnities of the dipropyl-substituted pyrimido[2,1-f]purinedione derivatives (2B)

O

N

O

N

R³

Ar

N

R²

m

2B

Compound

m

R²

R³

Ar

K_iSEM (lM)

Rat A₁

Rat A_2A

Human A_2B

Human A₃

Caﬀeine

KW-6002

2Ba

18.8 5.6^a

0.23 0.03^a

0.62 0.22^a

0.82 0.20^a

32.5 8.0^a

0.00515 0.00025^a

0.86 0.09^a

0

H

C₆H₅

3,4-di-OCH₃C₆H₃

0.59 0.06^b,d

nd

3.66 0.77^c

ꢀ1 (3%)^c

2Bb

1.33 0.12^a

2Bc

0

H

1.79 0.48^b

0.90 0.11^b

nd

ꢀ1 (9%)^c

N

H

nd, not determined.

^aTested at 25 lM.

^bTested at 10 lM.

^cTested at 1 lM.

^dTested versus [³H]ZM-241385.

of the spacer from ethyl (2Aa) to propyl one (2Al)

caused almost 3-fold reduction of A_2Apotency (K_i

A_2Avalues of 0.32 0.03 lM and 1.10 0.14 lM,

respectively). It seems that the spacer containing two

carbon atoms is an optimal one because the previously

described phenyl¹⁵and benzyl¹⁶pyrimido[2,1-f]purine-

diones have shown worse K_iA_2Avalues of P25 lM and

1.09 0.30 lM, respectively. That sequence of K_iA_2A

values brings to a supposition that the activity is able to

correlate with the mutual positions of phenyl rings and

pyrimido[2,1-f]purinedione skeletons in the species

possessing –(CH₂)_n– spacer, where n = 0, 1, 2, or 3 carbon

atoms, respectively. To describe qualitatively those geo-

In this group of compounds, only four structures (2Al,

2Ag, 2Ab, and 2Ad—ordered in decreasing activity)

exhibited low micromolar activity toward the rat adeno-

sine A₁receptor. In the series of dipropyl derivatives

(2Ba–2Bc) (see Table 3b), slightly lower A_2Aactivity

was maintained accompanied with beneﬁcial activity to-

ward adenosine A₁receptors. This tendency was also

noticed for the previously described phenyl¹⁵and ben-

zyl¹⁶pyrimido[2,1-f]purinediones. Additionally, similar

to the dimethyl derivatives with ethyl spacer, in the ser-

ies of dipropyl derivatives, the inﬂuence of 3-indolyl

substituent (2Bc) and double substituent in phenyl ring

(2Bb) caused analogous adenosine A_2Areceptors eﬀects.

´

˚

metrical spacer phenomena, parameter marked as D (A)

In the series of ether derivatives containing one hetero-

atom—oxygen in the spacer (5a–5d and 6a–6d) (see Ta-

ble 3c)—similar, low micromolar adenosine A_2Aactivity

was observed, slightly dependent on the kind (one sub-

stituent—5a, 5c, 5d, 6b, and 6c) and the number of sub-

stituents (two substituents—5b) in the phenyl ring. As

well, length of the spacer had small, rather disadvanta-

geous inﬂuence on the adenosine A_2Aactivity (cf. 5c,

6c, and 6d). In this series of compounds more frequently

the adenosine A₁activity was observed, almost equal

activity toward these two ARs subtypes was noticed as

well (6a and 6c). Compounds containing one hetero-

atom—nitrogen in the spacer (7a and 7b) (see Table

3c)—have shown micromolar activity towards adeno-

sine A₁and A_2Areceptors, better than the former

mentioned.

wasproposed. That parameter deﬁnes ‘‘out of plane’’ po-

sition for the ﬁrst atom of phenyl ring with respect to the

main plane of the molecule (Fig. 4a). Based on our previ-

ous^15,16and current crystallographic studies, the rough

models of four pyrimido[2,1-f]purinediones were built

and optimized at semi-empirical quantum chemistry level.

Then it was found that D parameters equal to 0.55, 1.86,

´

˚

2.50, and 3.46 A for molecules with –(CH₂)_n–spacer

encompassing 0, 1, 2, or 3 carbon atoms, respectively. Ob-

tained D values correlate well with the compounds’ activ-

ity to A_2Areceptor (Fig. 4b). Due to parabolic relation,

minimum of K_iA_2Ais observed for n = 2, that is for ethyl

(2Aa) derivative.

Exchange of the phenyl ring to a 2-pyridyl one (2Aj) is

not accepted, while for 3-indolyl substituent (2Ak) activ-

ity remains almost unchanged. Introduction of one sub-

stituent into the phenyl ring is allowed; it may be

beneﬁcial as it is observed with 4-OH (2Ac), or causes

slight decrease of the activity (2Ab, 2Ad); two substitu-

ents: 3,4-di-OCH₃(2Ai) decreased 4-fold the A_2Aactiv-

ity. Introduction of the substituents into the spacer was

disadvantageous for the A_2Aactivity (2Ae–2Ah).

Introduction of more heteroatoms into the spacer—car-

bamates 8a –d (Table 3c)—had distinctly negative inﬂu-

ence on the adenosine A₁and A_2Areceptors aﬃnity.

Summing up, the best A_2AARs ligands were com-

pounds 2Ac > 2Aa > 2Ak > 2Ab from the phenyl ethyl

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6964

Table 3c. Adenosine receptors aﬃnities of the pyrimido[2,1-f]purinediones with the spacers containing heteroatoms

O

H₃C

O

H₃C

O

N

H₃C

O

NH

z

5, 6, 8

7

R

Compound

z

R

K_iSEM (lM)

Rat A₁

Rat A_2A

Human A_2B

Human A₃

Caﬀeine

KW-6002

18.8 5.6^a

0.23 0.03^a

32.5 8.0^a

0.00515 0.00025^a

(1) One heteroatom—oxygen

5a

5b

5c

5d

6a

6b

6c

6d

—

4-Cl

2,4-di-Cl

4.22 0.53^a

3.50 0.72^b

ca. 10 (41%)^b

3.16 0.76^b

ca. 10 (48%)^b

3.68 0.98^b

ca. 10 (36%)^b

ca. 25 (49%)^a

1.09 0.21^b

2.02 0.81^b

4.60 1.53^b

3.69 0.39^b

1.33 0.24^b

3.05 0.24^b

4.29 1.20^b

ca. 10 (44%)^b,e

ca. 10 (45%)^b

ꢀ1 (0%)^c

ꢀ1 (2%)^c

ꢀ1 (0%)^c

nd

—

1

nd

4-NHCOCH₃

4-CH₃

H

1

4-Cl

4-F

H

1

ꢀ1 (0%)^c

CH₂CH@CH

ꢀ1 (0%)^c

(2) One heteroatom—nitrogen

7a

7b

—

H

4-F

7.55 2.35^b

2.63 0.21^b

20.0 2.0^b

14.0 1.5^b

>10 (27%)^b,d

nd

(3) More heteroatoms

8a

8b

8c

8d

C(O)NH

H

ꢀ10 (3%)^b

>10 (10%)^b

>10 (13%)^b

>10 (24%)^b

>10 (35%)^b

>10 (33%)^b

P 10 (42%)^b

>10 (32%)^b

nd

3-Cl

4-F

4-Cl

nd, not determined.

^aTested at 25 lM.

^bTested at 10 lM.

^cTested at 1 lM.

^dTested versus [³H]PSB-298.^26,27

^eTested versus [³H]ZM-241385.

a

b

30

K A_2A

i

25

20

15

10

5

n=0

n=3

n=2

n=1

0

[Å]

Δ

-5

0

0.5

1

1.5

2

2.5

3

3.5

4

˚

Figure 4. (a) Illustration for the parameter D (A) deﬁning mutual positions of the phenyl rings with respect to pyrimido[2,1-f]purinedione skeleton.

(b) Relationship between K_iand A_2Apotency of pyrimido[2,1-f]purinediones with phenyl substituent separated by (CH₂)_naliphatic spacer.

series and the best A₁AR aﬃnity was shown by the

dipropyl derivatives 2Ba > 2Bb > 2Bc > 2Al > 7b (phe-

nylpropyl derivative and one from amines).

Compounds 2Aa, 2Ac, and 2Ak were additionally tested

at the human adenosine A₁and A_2Areceptors. The

activity at A₁receptors was less inﬂuenced than at A_2A

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6965

receptors. Here, the less pronounced activity was ob-

served in the range of 3-fold (2Ac) to ca. 10-fold or more

(2aA and 2Ak).

5. If the regression parameters q²and R²for ND = 4

were lower than for ND = 3, then stepped up to 4.

If the regression parameters q²and R²for ND = 4

were higher than for ND = 3, then the equation was

stored. The same procedure was applied up to

ND = NDmax.

In Tables 3a–3c, the results of the binding assays toward

human adenosine A_2Band A₃receptors for chosen com-

pounds are presented. Dipropyl derivative 2Ba has

shown submicromolar aﬃnity for adenosine A_2Brecep-

tor, less active were 2Ac, 2Aa, 2Ak, and 2Aj. All com-

pounds were those with two carbon atoms spacer.

6. The equations having the highest q²and R²were col-

lected and k, k⁰, and F were calculated.

7. The equation with the best statistical parameters was

chosen.

No remarkable aﬃnity to adenosine A₃receptor was

found for the investigated compounds 2Aa, 2Ac, 2Aj,

2Ak, 2Bb, 2Bc, 5a–d, 6a, 6c, 6d (only dipropyl derivative

2Ba has shown low micromolar activity).

3D-QSAR analyses were performed for the compounds

as well presently as previously described,^15,16which

showed activity at adenosine A₁and adenosine A_2A

receptors.

In our previous work¹⁵, it was stated that tricyclic puri-

nediones act as antagonists at adenosine A_2Areceptors

in the result of their functional properties evaluation

using a sodium chloride shift assay.³⁶Considering the

similarity in the structures of the new examined com-

pounds, it may be supposed that they also act as anta-

gonists at adenosine A_2Areceptors.

5.2.1. A₁AR QSAR. For the development of A₁AR 3D-

QSAR model, 31 ligands (9, 10, 16, 21, 29, 39, 40;¹⁵1, 5,

10, 13–20;¹⁶2Ab, 2Ad, 2Ag, 2Al, 2Ba–c, 5a, b, 6a, 6c, 7a,

and 7b) were used. The model selected as the best one

included the smallest number of descriptors together

with the largest values of q²and R². The following

3D-QSAR Eq. 1 that best predicts the A₁AR ligands’

aﬃnity was derived based on four descriptors:

5.2. In vitro tests: quantitative structure–activity

relationships

K_iðA₁ARÞ ¼ ꢁ2:205 ꢂ LogP ꢁ 0:087 ꢂ E_steric

þ 0:017 ꢂ DV_Xþ 0:027 ꢂ SA_IPþ

3D-QSAR analyses were performed for the compounds

which showed activity at adenosine A₁and adenosine

A_2Areceptors. Topological, geometric, and electronic

descriptors were calculated by means of the program

CAChe 6.1.1.³⁷Because of the small database, we used

the whole sample for the calibration and checked the

ﬁt by cross-validation³⁸instead of holdout data set.^39,40

ꢁ 4:218;

ð1Þ

where LogP is the octanol–water partition coeﬃcient,

E_stericis the steric energy, DV_Xis the X component of

the molecule’s dipole vector, SA_IP+is the positively

charged surface area of electrostatic isopotential.

The statistics for the model is shown in Table 4. Regres-

sion line obtained by plotting experimental versus pre-

dicted K_ivalues is shown in Figure 5a. The diﬀerences

between experimental and calculated biological data

are presented in Figure 5b. The ﬁrst descriptor included

in Eq. 1 is LogP. Its range is from 1.172 to 5.082 and

has a negative sign: an increasing contribution coeﬃ-

cient corresponds to a decrease in K_ivalue. It reﬂects

the importance of the R¹-alkyl substituents as well as

the annelated ring size. The presence of propyl substitu-

ents, 6- or 7-membered annelated ring, and phenyl sub-

stituent bearing methoxy group or chlorine atom and

linked with the tricyclic ring by alkyl chain is advisable.

The second descriptor is a steric energy which is the sum

of the molecular mechanics potential energies calculated

for the bonds, bond angles, dihedral angles, non-bonded

atoms, and so forth. Steric energy for the A₁AR ligands

is in the range ꢁ2.549 ꢃ ꢁ39.009 kcal/mol and has

Determination of the linear equation that best repre-

sented the dependence of the A₁and A_2AAR binding

aﬃnity on several input descriptors was carried out

using MLR (multiple linear regression) technique. The

statistical parameters used for the model development

and validation are as follows: R²is the correlation coef-

ﬁcient, q²is the leave-one-out cross-validated R², k and

k⁰are the slopes of the regression predicted aﬃnity ver-

sus observed and observed aﬃnity versus predicted,^39,40

F is the test F value. The algorithm we used to select the

best descriptors and models is summarized below.

1. The correlation equations with 1,2,. . ., NDmax

descriptors (maximum number of descriptors equals 5)

were calculated using ProjectLeader CAChe

6.1.1³⁷application. The ratio of the number of com-

pounds and number of variables is at least 5:1.⁴¹

2. An exclusion of the descriptor in any pair of descrip-

tors whose pairwise correlation exceeds 0.85 was car-

ried out.^38,42,43

Table 4. Summary of the statistic for A₁AR and A_2AAR QSAR

models

3. The single linear regressions were calculated to

choose the descriptors with the best correlation to

ligands’ aﬃnity. Two parameters linear regressions

were determined to choose best pairs of the descrip-

tors for predicting ligands’ aﬃnity. The same proce-

dure was applied for the triple linear regression.

4. The next one descriptor was added and multiple lin-

ear regression analysis was performed.

Eq. 1

28

Eq. 2

65

Number of components

q²

R²

k

0.793

0.817

0.836

1.000

0.646

0.841

0.846

0.998

0.510

k⁰

F

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6966

9

8

9

8

7

6

7

5

6

5

4

3

2

1

0

-1

0

1

2

3

4

5

6

7

8

9

1

3

5

7

9

11 13 15 17 19 21 23 25 27

Experimental A₁Ar K_ivalues

Compound

Figure 5. (a) Experimental versus predicted K_ivalues of A₁AR binding activity. (b) Comparison of the experimental K_ivalues (blue, Ç) with those

predicted by 3D-QSAR model (rose, j) for A₁AR ligands.

negative coeﬃcient in Eq. 1. It is quite sensitive to the

size of the substituent, in particular to the linker between

the tricyclic ring and phenyl moiety, but the size of ann-

elated ring does slightly aﬀect it. It suggests that the

compounds that possess benzyl or phenylalkyl substitu-

ents should be preferred to phenyl substituents and long

phenylalkyl moieties with the heteroatom in the linker.

The third descriptor is the dipole vector—a key electrical

property of the molecule, that has both size and direc-

tion. It points from the direction of net negative charge

to the direction of net positive charge. The best correla-

tion was obtained with the dipole vector X. Its value is

from ꢁ9.237 to ꢁ0.807 and from 0.168 to 7.780 D. This

descriptor is aﬀected by the nature of the substituents in

the annelated ring. Its direction depends on the linker

type. Positive contribution coeﬃcient suggests that for

the better aﬃnity to A₁AR, ligand should possess phe-

nyl ring with small polar substituent or benzyl moiety.

The last one descriptor in Eq. 1 is positively charged sur-

face area of electrostatic potential. It is inﬂuenced by the

substituents in the annelated ring as well as by the R¹-al-

kyl, although it is not aﬀected by the third ring size. The

positive surface area is in the range from 296.479 to

Three outliers were determined during development of

the model which are pyrimidopurinediones with phenyl

(16, 29)¹⁵and benzyl (13)¹⁶substituents in the annelated

ring. The K_ivalues for these compounds are 35.0, 19.1,

and 17.8 lM, respectively, which is out of K_irange

(0.089 ꢃ 8.01 lM) for the rest of A₁AR ligands’ data-

base. Such compounds with low micromolar aﬃnity

are poorly presented in the sequence in comparison with

those possessing submicromolar and micromolar aﬃni-

ties, which may be the reason why the QSAR model is

unable to accurately predict their activity.

5.2.2. A_2AAR QSAR. Sixty-nine A_2AAR ligands (6, 7,

9–12, 14–19, 21, 23–29, 32, 34–40;¹⁵1, 3–10, 12–20;¹⁶

2Aa–i, 2Ak, 2Al, 2Ba–c, 5b–d, 6a–d, 7a, and 7b) were

collected for the development of 3D-QSAR model. Sev-

eral QSAR equations were constructed. The model con-

taining the smallest number of the descriptors but

leading to the largest increase in q²and R²was selected

as the best one Eq. 2. The following 3D-QSAR equation

was derived based on ﬁve descriptors:

K_iðA_2AARÞ ¼ ꢁ0:379 ꢂ LogP þ 0:076 ꢂ E_steric

ꢁ 0:055 ꢂ DV_Yꢁ 0:023 ꢂ SA_IPꢁ

˚

523.067 A. This descriptor also implies that the electro-

static properties of the substituents in the molecule are

important for the A₁AR binding aﬃnity. It is very sen-

sitive as when changing the position of chlorine atom

from para to meta, it aﬀects not only the purine ring

but also R¹substituents¹⁶charge distribution (Fig. 6).

The presence of the benzyl or phenethyl substituents in

the annelated ring bearing halogen atom and long alkyl

moiety at R¹is so suggested.

ꢁ 0:032 ꢂ IF_EDþ 15:971;

ð2Þ

where LogP is the octanol–water partition coeﬃcient,

E_stericis the steric energy, DV_Yis the Y component of

the molecule’s dipole vector, SA_IPꢁis the negatively

charged surface area of electrostatic isopotential, IF_ED

is the surface integral of electrostatic potential on elec-

tron density.

Figure 6. Electrostatic potential isosurface for ligands 16 and 18¹⁶

.

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6967

The statistics for the model is shown in Table 4. Regres-

sion line obtained by plotting experimental versus pre-

dicted A_2AAR K_ivalues is shown in Figure 7a. The

diﬀerences between experimental and calculated biolog-

ical data are reported in Figure 7b. The ﬁrst descriptor

included in A_2AAR QSAR model is LogP and in the

range from 0.083 to 5.082. The negative contribution

coeﬃcient means that the higher LogP value the better

A_2AAR aﬃnity possessing ligand. The most favorable

substituents in the annelated ring are phenethyl and ben-

zyl, bearing chlorine atom or methoxy group, and R¹-

propyl moiety. The second descriptor in Eq. 2 is the Y

component of the molecule’s dipole vector and has neg-

ative contribution coeﬃcient that suggests the impor-

tance of the vector direction. The dipole moment Y

values are is in the range ꢁ7.314 ꢃ ꢁ0.020 and

0.314 ꢃ 4.869 D. This descriptor aﬀection proposes the

preference of phenethyl substituents in the third ring

and R¹-methyl groups. The next one descriptor used

to derive the Eq. 2 is steric energy which is in the range

ꢁ41.423 ꢃ ꢁ3.636 kcal/mol. The nature of annelated

ring substituents aﬀects steric energy and as it has posi-

tive contribution coeﬃcient, the substituents decreasing

its value, such as benzyl and phenethyl along with

R¹-methyl groups, are so suggested. The fourth descrip-

tor is a negatively charged surface area of electrostatic

potential. It is strongly inﬂuenced by the substituents

in the annelated ring. The negative surface area is in

stead compound 5d could be involved in allosteric inter-

actions with the A_2AAR.⁴⁴Obviously, for this ligand a

change in the mechanism of the reaction occurs at the

inversion point.

Comparing obtained A₁and A_2AARs 3D-QSAR mod-

els, it could be noticed that for the A₁AR binding activ-

ity LogP has higher importance than for A_2AAR

aﬃnity as the contribution coeﬃcient is six times higher.

Ligands’ steric energy has comparable contribution

coeﬃcient value but diﬀerent signs. Diﬀerent dipole mo-

ment vectors, X for A₁AR and Y for A_2AAR, are

incorporated in both models. Notably for A_2AAR mod-

el, the dipole vector aﬀects K_ivalue three times more

than A₁AR K_i. Moreover, it is signiﬁcant the opposite

directions of these vectors as their contribution coeﬃ-

cients have diverse (positive and negative) meaning.

Both QSAR equations incorporate surface area of elec-

trostatic potential. For A₁AR, this area is positive and

has positive contribution coeﬃcient and for the A_2AAR

model it is negatively charged bearing negative sign.

Important that for A_2AAR the equation includes infor-

mation about shape of the molecular volume which sug-

gests that for this AR subtype the surface property is of

great importance.

5.3. In vivo tests

˚

the range from 103.634 to 278.981 A. The presence of

Virtually all of the tested compounds were inactive as

anticonvulsants; some of the compounds exhibited toxic

eﬀects at 0.5 h at scMet test (in the group of arylalkyl

derivatives: death following tonic extension after 30 mg/

kg of 2Aa, continuous seizure activity was noticed after

100 mg/kg dose of 2Al, tonic extension and death follow-

ing continuous seizure after 300 mg/kg dose of 2Ad; in the

group of carbamates: death following continuous seizure

was observed after 100 mg/kg dose of 8c; death following

tonic extension after 300 mg/kg dose of 8d). Only com-

pounds 2Ah showed a short time (0.5 h) protection in

the MES test at a high dose of 300 mg/kg in mice, and

5c also exhibited short time (0.5 h) protection in the scMet

test in mice at 300 mg/kg (four animals from ﬁve tested

were protected). At the mice 6 Hz test, ether 5d has shown

some protection at 0.25, 0.5, and 1.0 h time points (one

from four animals), and 5b has shown weaker protection

only at 0.25 h (one from four animals).

phenyl substituents bearing methyl or methoxy moiety

or phenyl substituents with long alkyl or alkoxyl linker

could improve A_2AAR K_ivalue. The last one descriptor

in Eq. 2 is surface integral of electrostatic potential on

electron density (from 103.126 to 266.291). It is a hybrid

shape–property descriptor and represents the surface

electronic properties. It encodes the shape of molecular

volume and reﬂects the behavior of the whole molecule

which is important when binding to the receptor.

Four outliers were determined which are pyrimidopurin-

ediones with phenyl substituents in the annelated ring

(36, K_i= 22.3 lM)¹⁵and ligands 5d (K_i= 4.6 lM), 7a

(K_i= 20.0 lM), 7b (K_i= 14.0 lM). For compounds 36,

7a, and 7d, the derived QSAR model is unable to accu-

rately predict the K_ivalues that are out of the data set

range (from 0.147 to 8.72 lM) and poorly presented. In-

9

8

9

8

7

6

5

4

3

2

1

0

-1

0

1

2

3

4

5

6

7

8

9

1

5

9

13 17 21 25 29 33 37 41 45 49 53 57 61 65

Experimental A_2AAR K_ivalues

Compound

Figure 7. (a) Experimental versus predicted K_ivalues of A_2AAR binding activity. (b) Comparison of the experimental K_ivalues (blue, Ç) with those

predicted by 3D-QSAR model (rose, j) for A_2AAR ligands.

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6968

6. Conclusions

7.2. General procedures for the synthesis of 1,3-dialkyl-

9-arylalkyl-6,7,8,9-tetrahydropyrimido[2,1-f]purine-2,4-

(1H,3H)-diones (Table 2a)

In conclusion, we have shown that N-phenethyl-substi-

tuted pyrimido[2,1-f]purinediones and analogs with the

modiﬁcations of ethylene spacer (substituted or extended)

exhibit micromolar or submicromolar aﬃnity for A_2A

ARs. The kind of substituent at the aromatic ring was

important for the aﬃnity. A spacer of two carbon atoms

was favorable for high A_2AAR aﬃnity. Branched chains

as a spacer were more advantageous for A₁AR aﬃnity.

Oxygen and nitrogen atoms in the spacer may result in a

slight decrease in A_2Aand less A₁ARs aﬃnity. When

the 1,3-dimethyl groups were replaced by propyl residues,

enhanced aﬃnity for the other AR subtypes was ob-

served, and the A_2Aselectivity was lost (compare 2Aa

and 2Ba).

A mixture of 1.65 g (5 mmol) of 7-(3-chloropropyl)-8-

bromotheophylline (1A) or 0.78 g (2 mmol) of 1,3-dipro-

pyl-7-(3-chloropropyl)-8-bromoxanthine (1B) and the

appropriate amine (1.4- to 5-fold excess) was reﬂuxed

in the appropriate solvent (propanol, butanol, methoxy-

ethanol, DMF) or without solvent (compound 2Ad) for

5–10 h (see Table 2a). After cooling, a precipitate was

formed and was collected by ﬁltration. The precipitate

of compound 2Ae, containing a small amount of unre-

acted 7-(3-chloropropyl)-8-bromotheophylline, was sep-

arated by stirring it with 20% aq HCl solution for 0.5 h,

ﬁltered, and reprecipitated by the addition of 10% aq

NaOH solution to adjust a pH value of 9. Compounds

2Ba and 2Bb were precipitated by adding water to the

reaction mixture in DMF.

• The best ligands for A_2AARs were compounds 2Ac,

2Aa, 2Ak, and 2Ab.

• The best ligands for A₁AR were compounds 2Ba,

2Bb, 2Bc, 7b (however, with lower selectivity toward

A_2AAR).

7.3. General procedure for the synthesis of acetyl

derivatives

The reliable 3D-QSAR models able to rationalize the

activity of A₁and A_2Aadenosine receptor for tricyclic

xanthine ligands were developed. The analysis of the re-

sults depicts that for the A₁AR binding activity it is

important for ligands to possess R¹-bulky substituents

along with the phenyl or benzyl substituents bearing

halogen atom and phenethyl moiety. For A_2AAR

aﬃnity it could be favorable to introduce phenethyl or

phenyl substituent connected with the tricyclic ring by

the alkoxy chain. The nature of R¹group may not

signiﬁcantly aﬀect the A_2AAR aﬃnity.

Acetyl derivatives 2Ah and 2Ag were prepared by reﬂux-

ing of compounds 2Af and 2Ae with acetic anhydride for

5 h, removing the excess of anhydride by distillation un-

der reduced pressure and crystallization of the residue.

7.3.1. Synthesis of ether derivatives (Table 2b)

7.3.1.1. 1,3-Dimethyl-9-[2-(4-methylphenoxy)ethyl]-6,7,

8,9-tetrahydropyrimido[2,1-f]purine-2,4(1H,3H)-dione (5d).

Method A. A mixture of 1.02 g (3 mmol) of 1,3-dimethyl-9-

bromoethyl-6,7,8,9-tetrahydropyrimido[2,1-f]purine-2, 4-

(1H,3H)-dione (3)¹⁸and 0.32 g (3 mmol) of p-cresole

sodium salt (obtained as in Method C) in 30 mL of dry

acetonitrile containing a catalytic amount of [15]-crown-5

was reﬂuxed for 6 h. The mixture was ﬁltered while hot,

and the obtained solid contained NaBr. After cooling of

the ﬁltrate, product 3 precipitated ﬁrst, then in the next

ﬁltrate the main crop of the product (5d) was separated.

With the same method, compounds 5b and 5c were

obtained.

7. Experimental

7.1. Chemistry general remarks

Melting points were determined on a MEL-TEMP II

apparatus and are uncorrected. IR spectra were taken

as KBr discs on an FT Jasco IR 410 spectrometer.

UV spectra were recorded on Jasco UV–vis V 530

in a concentration of 10^ꢁ4mol/L in methanol. ¹H

NMR spectra of compounds 2Ab, 2Ai, 2Ae, and

2Ah were obtained with a Bruker Ac 200F in CDCl₃;

compounds 2Ad, 2Ak (in CDCl₃) and compounds 2Ac

and 5a (in DMSO-d₆) were analyzed with a Bruker

VM250 apparatus; compounds 2Aa, 2Ag, 2Aj, 2Al,

2Ba, 2Bc, 5b–d, and 6a–d were determined with a Var-

ian-Mercury 300 MHz apparatus in CDCl₃, and com-

pounds 8a–d, 7a, and 7b in DMSO-d₆using

tetramethylsilane as an internal standard. Elemental

analyses (CHN) were performed on an Elemental Var-

io-EL III apparatus and were in accordance with the-

oretical values within 0.4%. TLC data were obtained

with Merck silica gel 60F₂₅₄aluminum sheets using

the following developing systems: (1) benzene/acetone

(7:3); (2) toluene/acetone (7:3); (3) toluene/acetone

(1:1); (4) CH₂Cl₂/MeOH (18.5:1.5); (5) toluene/ace-

tone/MeOH (5:5:1). Spots were detected under UV

light.

Method B. To a stirred and ice-cold solution of 1.38 g

(5 mmol) of 1,3-dimethyl-9-hydroxyethyl-6,7,8,9-tetra-

hydropyrimido[2,1-f]purine-2,4(1H,3H)-dione

(4),¹⁹

1.57 g (6 mmol) of triphenylphosphine, and 0.64 g

(6 mmol) of p-cresole in 15 mL of dry THF, 0.95 mL

(6 mmol) of DEAD (diethyl azodicarboxylate) was

added dropwise. Then the reaction mixture was stirred

at room temperature overnight. The solvent was re-

moved under reduced pressure, and the obtained oily

residue was puriﬁed by column chromatography using

a mixture of ethyl acetate/chloroform (1:1) as eluent

yielding 5d. With the same method compound 5c was

obtained. For column chromatography separation, ﬁrst

a mixture of ethyl acetate : chloroform (1:1) then chloro-

form/MeOH (9:1) were used as the eluents.

7.3.1.2. 1,3-Dimethyl-9-(4-chlorophenoxyethyl)-6,7,8,9-

tetrahydropyrimido[2,1-f]purine-2,4(1H,3H)-dione

(5a).

Method C. A mixture of 0.34 g (1 mmol) of 1,3-dimethyl-

9-bromoethyl-6,7,8,9-tetrahydropyrimido[2,1-f]purine-

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6969

2, 4-(1H,3H)-dione (3) and 0.3 g (2 mmol) of sodium

p-chlorophenolate (obtained ex tempore by dissolving

p-chlorophenol in a hot ethanolic solution of NaOH,

subsequently removing the solvent by distillation under

reduced pressure and washing the residue with acetone)

and 10 mL of methoxyethanol was reﬂuxed for 20 h.

After distillation under reduced pressure the residue

was washed with water and ﬁltered oﬀ.

tical membrane preparations for A₁assays, and striatal

membrane preparations for A_2Aassays as described.⁴⁸

CHO cells recombinantly expressing the human adeno-

sine A₁, A_2A, and A₃receptors were a gift from Dr.

K.-N. Klotz, and grown as described.^48,49Membrane

preparations were obtained as previously described.^48,49

For A_2Badenosine receptor assays, commercially avail-

able membrane preparations containing the human A_2B

AR were obtained from Biotrend (Cologne, Germany).

Stock solutions of the compounds were prepared in

dimethylsulfoxide (DMSO), the ﬁnal concentration of

DMSO in the assays did not exceed 2.5%. Initial screen-

ing was performed at a single concentration of 25, 10,

2.5, or 1 lM, depending on the solubility of the test

compound. Binding assays were performed as previ-

ously described.^{25,29,47–50}Curves were determined using

6–7 diﬀerent concentrations of test compounds spanning

three orders of magnitude. At least three separate exper-

iments were performed, each in triplicate. For non-lin-

ear regression analysis, the Cheng–Prusoﬀ equation

and K_Dvalues of 0.2 nM (rat A₁) and 0.6 nM (human

A₁), respectively, for [³H]CCPA,²⁶8 nM (rat A_2A) and

7 nM (human A_2A), respectively, for [³H]MSX-2,^25,51

33 nM for [³H]ZM241385 (human A_2B),⁵⁰56 nM (hu-

man A_2B) for [³H]PSB-298 (human A_2B),³⁰and 4.9 nM

(human A₃) for [³H]PSB-11²⁹were used to calculate K_i

values from IC₅₀values.

7.3.1.3. 1,3-Dimethyl-9-[2-(benzyloxy)ethyl]-6,7,8,9-

tetrahydropyrimido[2,1-f]purine-2,4(1H,3H)-dione (6a).

Method D. A suspension of 0.68 g (2.5 mmol) of 4 and

0.12 g (3 mmol) of NaH in dry toluene containing a cat-

alytic amount of [15]-crown-5 was stirred for 24 h at

room temperature. Then 0.51 g (3 mmol) of benzyl chlo-

ride was added dropwise and the reaction mixture was

reﬂuxed for 6 h. The mixture was ﬁltered while hot,

and from the cooled ﬁltrate unreacted 4 was separated

as a ﬁrst precipitate. Then the desired product (6a) pre-

cipitated. With the same method compounds 6b–d were

obtained.

7.3.2. Synthesis of amine derivatives 7a and 7b (Table 2b)

7.3.2.1. Hydrochloride of 1,3-dimethyl-9-[2-(benzylami-

no)-ethyl]-6,7,8,9-tetrahydropyrimido[2,1-f]purine-2,4(1H,-

3H)-dione (7a). Method E. Compound 3 0.85 g (2.5 mmol)

and 1.0 gof potassium carbonate weretriturated in a mor-

tarandthenmixedwith0.27 g(2.5 mmol)ofbenzylamine.

Themixturewasmeltedinabeakercoveredwithareversed

funnel under microwave irradiation (domestic oven) for

8 min at 450 W. The cooled mixture was extracted with

warm methylene chloride. Then the solvent was evapo-

rated and the oily residue was puriﬁed by column chroma-

tography using methylene chloride/MeOH (9:1) as an

eluent.Afterevaporationoftheeluentunderreducedpres-

sure, the residue was dissolved in ethanol and saturated

with dry HCl gas. Compound 7b was obtained with the

same method. It was isolated as the free base.

7.4.2. Anticonvulsant screening. The preliminary anticon-

vulsant evaluation was carried out using reported proce-

dures.^30,32,31,33Male albino mice (F-1 strain, 18–25 g)

were used as experimental animals. Groups of 1–5 mice

were used in MES, scMet tests, group of 2–8 animals in

rotorod test. The tested compounds were suspended in a

0.5% methylcellulose–water mixture. Each compound

was administered as an ip injection at three dose levels

(30, 100, and 300 mg/kg) with anticonvulsant activity

and neurotoxicity assessed at 0.5 and 4 h intervals after

administration. Anticonvulsant eﬃcacy was measured

by maximal electroshock (MES) and subcutaneous

pentylenetetrazole (scMet), neurological deﬁcit was

investigated in the rotorod test. Groups of four mice

were used in 6 Hz test. Each compound was adminis-

tered at 100 mg/kg dose, the results were observed after

0.25, 0.5, 1, 2, and 4 h intervals after administration.

7.3.3. Synthesis of carbamate derivatives (Table 2b)

7.3.3.1. N-phenyl carbamate of 1,3-dimethyl-9-(2-hy-

droxyethyl)-6,7,8,9-tetrahydropyrimido[2,1-f]purine-2,4(1H,-

3H)-dione (8a). Method F. A mixture of 1.4 g (5 mmol) of 4

and 0.57 g (5 mmol) of phenyl isocyanate in 20 ml of anhy-

drous acetonitrile was reﬂuxed for 10 h. The resulting solid

was ﬁltered oﬀ while hot and recrystallized from ethanol.

7.5. X-ray structure analysis of compound 6d

By the same method compounds 8b–d were obtained.

The spectral data are summarized in Table 5.

7.4. Pharmacology

Crystals of 6d were obtained by slow evaporation from

methanol/ethanol (1:1) solutions. The measurements of

the crystals were performed on a Siemens SMART

CCD with CuKa radiation (k = 1.54A) at room

temperature.

´

˚

7.4.1. Adenosine receptor binding assays. [³H]CCPA

(54.9 Ci/mmol) was purchased from NEN Life Sciences;

[³H]MSX-2 (85 Ci/mmol), [³H]PSB-11 (53 Ci/mmol),

and [³H]PSB-298 (124 Ci/mmol) were obtained from

Amersham (custom synthesis). The non-radioactive pre-

cursors of [³H]MSX-2 (MSX-1),^25,45[³H]PSB-11 (PSB-

10),⁴⁶and [³H]PSB-298 (PSB-297)⁴⁷were synthesized

in our laboratory. [³H]ZM241385 (17 Ci/mmol) was ob-

tained from Tocris. Frozen rat brains obtained from Pel

Freez^ꢂ, Rogers, AR, USA, were dissected to obtain cor-

7.6. Crystal data for 6d

C₂₁H₂₅N₅O₃, M = 395.46, monoclinic, space group P2₁/c

˚

(No.14), a = 24.563(5) A, b = 4.7071(9) A, c = 17.586(4)

˚

3

˚

(A), b = 104.67(3)ꢁ, V = 1967.0(8) A , Z = 4, D_x=

1.335 g/cm³, T = 297 K, l = 0.749 mm^ꢁ1, k = 1.54178

˚

A, F(000) = 840, colorless prisms, (0.1 · 0.3 · 0.3 mm),

data/parameters = 3715/375; ﬁnal R₁= 0.043, wR₂=

0.1375 (all data).

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6970

Table 5. Spectral data of the obtained pyrimido[2,1-f]purinediones

Compound

IR m (cm^ꢁ1

)

UV k_max, loge

¹H NMR d (ppm)

2Aa

3087–3032-phenyl, 2964–2860-CH₂, 1693-

CO (pos. 2), 1657-CO (pos. 4), 750-CH₂

299, 4.09

2.02–2.09 (m, 2H, CH₂CH₂CH₂), 2.96 (t, J = 7.50 Hz, 2H,

CH₂CH₂), 3.21 (t, J = 6.25 Hz, 2H, CH₂CH₂), 3.37 (s, 3H,

3.36 N₃CH₃), 3.52 (s, 3H, N₁CH₃), 3.75 (t, J = 7.50 Hz,

2H, CH₂N₉), 4.17 (t, J = 6.25 Hz, 2H, N₅CH₂), 7.19–7.33

(m, 5H, phenyl)

2Ab

2994–2973-CH₂, 1697-CO (pos. 2), 1656-

CO (pos. 4), 1052-OCH₃, 817-phenyl,

748-CH₂

303, 4.30

1.90–2.08 (m, 2H, CH₂CH₂CH₂), 2.88 (t, J = 7.00 Hz, 2H,

CH₂CH₂), 3.11–3.19 (m, 2H, CH₂CH₂), 3.35 (s, 3H,

N₃CH₃), 3.50 (s, 3H, N₁CH₃), 3.67–3.73 (m, 2H, CH₂N₉),

3.78 (s, 3H, OCH₃), 4.14 (t, 2H, J = 6.00 Hz, N₃CH₂),

7.12 (d, 1H, J = 6.60 Hz, 3⁰-phenyl), 7.28 (d, 1H,

J = 4.00 Hz, 5⁰-phenyl), 7.30–7.36 (m, 2H,

2⁰6⁰-phenyl)

2Ac

3382-OH, 3021-phenyl, 2935-CH₂, 1687-

CO (pos. 2), 1643-CO (pos. 4), 833-aryl,

750-CH₂

303, 4.09

1.80–2.20 (m, 2H, CH₂CH₂CH₂), 2.82 (t, J = 7.20 Hz, 2H,

CH₂CH₂), 3.27 (s, 3H, N₃CH₃), 3.31–3.40 (m, 5H,

CH₂CH₂+ N₁CH₃), 3.65 (t, J = 7.30 Hz, 2H, CH₂N₉),

4.06 (t, J = 5.70 Hz, 2H, N₅CH₂), 6.73 (d, J = 8.03 Hz, 2H,

3⁰5⁰-phenyl, 7.08 (d, J = 8.00 Hz, 2H, 2⁰6⁰-phenyl), 9.24 (s,

1H, OH)

2Ad

2Ae

2938–2844-CH₂, 1695-CO (pos. 2), 1649-

CO (pos. 4), 811-phenyl, 748-CH₂

302, 4.29

2.00–2.09 (m, 2H, CH₂CH₂CH₂), 2.93 (t, J = 7.75 Hz, 2H,

CH₂CH₂), 3.24 (t, J = 5.00 Hz, 2H, CH₂CH₂), 3.35 (s, 3H,

N₃CH₃), 3.5 (s, 3H, N₁CH₃), 3.71 (t, J = 8.00 Hz, 2H,

CH₂N₉), 4.17 (t, J = 6.00 Hz, 2H, N₅CH₂), 7.13–7.22 (m,

2H, 2⁰6⁰-phenyl), 7.23–7.28 (m, 2H, 3⁰5⁰-phenyl)

2.01–2.03 (m, 2H, CH₂CH₂CH₂), 3.09–3.24 (m, 2H,

N₉CH₂), 3.34 (s, 3H, N₃CH₃), 3.50 (s, 3H, N₁CH₃),

3.58–3.64 (m, 1H, CH₂N₉), 3.77–3.98 (m, 1H, CHN₉), 4.14

(t, J = 6.00 Hz, 2H, N₅CH₂), 4.96 (d, J = 4.00 Hz, 1H,

OH), 5.04–5.08 (m, 1H, CHOH), 7.27–7.37 (m, 5H,

phenyl)

3374-OH, 3060–3027-phenyl, 2925–2881-

CH₂, 1685-CO (pos. 2), 1645-CO (pos. 4),

833- phenyl, 750-CH₂

2Af

2Ag

2Ah

2Ai

2Aj

3316-OH, 3062–3045-phenyl, 2942–2873-

CH₂, 1691-CO (pos. 2), 1651-CO (pos. 4),

765-phenyl, 748-CH₂

304, 4.30

301, 4.32

1.29 (d, J = 7.20 Hz, 3H, CH₃), 1.6 (s, 1H, OH), 2.02–2.09

(m, 2H, CH₂CH₂CH₂), 3.15–3.22 (m, 2H, CH₂N₉), 3.38 (s,

3H, N₃CH₃), 3.54 (s, 3H, N₁CH₃), 4.09–4.20 (m, 2H,

N₅CH₂), 5.05–5.07 (m, 1H, CHCH₃), 5.30 (d, J = 2.50 Hz,

1H, CHOH), 7.26–7.38 (m, 5H, phenyl)

3063–3033-phenyl, 2983–2944-CH₂, 1749-

OCOCH₃, 1697-CO (pos. 2), 1650-CO

(pos. 4), 796-phenyl, 748-CH₂

1.34 (d, J = 7.10 Hz, 3H, CH₃), 1.86–1.94 (m, 2H,

CH₂CH₂CH₂), 2.12 (s, 3H, COCH₃), 3.11–3.26 (m, 2H,

CH₂N₉), 3.36 (s, 3H, N₃CH₃), 3.54 (s, 3H, N₁CH₃), 4.01

(m, 2H, N₅CH₂), 4.69–4.78 (m, 1H, CHCH₃), 6.03 (d,

J = 6.30 Hz, 1H, CHOCOCH₃), 7.24–7.36 (m, 5H, phenyl)

2.01–2.07 (m, 5H, CH₂CH₂CH₂+ COCH₃), 3.28–3.29 (m,

5H, N₉CH₂+ N₃CH₃), 3.51 (s, 3H, N₁CH₃), 3.83 (d,

J = 5.40 Hz, 2H,CH₂N₉), 4.17 (t, J = 6.00 Hz, 2H,

N₅CH₂), 6.11 (t, J = 5.60 Hz, 1H, CH), 7.26–7.39 (m, 5H,

phenyl)

3062–3045-phenyl, 2942–2873-CH₂, 1738-

OCOCH₃, 1703-CO (pos. 2), 1653-CO

(pos. 4), 767-phenyl, 744-CH₂

3081-phenyl, 2991–2840-CH₂, 1693-CO

(pos. 2), 1655-CO (pos. 4), 1068-OCH₃,

816-phenyl, 750-CH₂

303, 431

305, 4.32

1.95–2.05 (m, 2H, CH₂CH₂CH₂), 2.88 (t, J = 7.10 Hz, 2H,

CH₂CH₂), 3.19 (t, J = 5.60 Hz, 2H, N₉CH₂), 3.34 (s, 3H,

N₃CH₃), 3.50 (s, 3H, N₁CH₃), 3.72 (t, J = 7.20 Hz, 2H,

CH₂N₉), 3.84 (s, 6H, 2· OCH₃), 4.15 (t, J = 6.00 Hz, 2H,

N₅CH₂), 6.73–6.84 (m, 3H, phenyl),

3056–3023-pyridyl, 2976–2862-CH₂,

1695-CO (pos. 2), 1662-CO (pos. 4), 743-

CH₂

2.00–2.09 (m, 2H, CH₂CH₂CH₂), 3.14 (t, J = 7.51 Hz, 2H,

CH₂CH₂), 3.28 (t, J = 6.26 Hz, 2H, N₉CH₂), 3.36 (s, 3H,

N₃CH₃), 3.51 (s, 3H, N₁CH₃), 3.92 (t, J = 6.25 Hz, 2H,

CH₂N₉), 4.23 (t, J = 5.00 Hz, 2H, N₅CH₂), 7.15–7.19 (m,

2H, 3⁰5⁰-pyridyl), 7.57–7.61 (m, 1H, 4⁰-pyridyl), 8.52–8.54

(m, 1H, 6⁰-pyridyl)

2Ak

2Al

3322-NH, 3106–3058-phenyl, 2973–2842-

CH₂, 1697-CO (pos. 2), 1649-CO (pos. 4),

817-phenyl, 740-CH₂

2.06–2.15 (m, 2H, CH₂CH₂CH₂), 3.19 (t, J = 7.40 Hz,

2H, CH₂CH₂), 3.28 (t, J = 5.60 Hz, 2H, N₉CH₂), 3.46

(s, 3H, N₃CH₃), 3.62 (s, 3H, N₁CH₃), 3.93

(t, J = 7.40 Hz, 2H, CH₂N₉), 4.24 (t, J = 6.00 Hz, 2H,

N₅CH₂), 7.07–7.32 (m, 4H, phenyl), 7.75–7.80

(m, 1H, 2⁰-pirol)

3057–3023-phenyl, 2976–2863-CH₂, 1695-

CO (pos. 2), 1662-CO (pos. 4), 793-CH₂

301, 4.29

1.92–2.04 (m, 2H, CH₂CH₂CH₂), 2.05–2.14 (m, 2H,

CH₂CH₂CH₂), 2.68 (t, J = 7.50 Hz, 2H, CH₂CH₂CH₂),

3.31 (t, J = 6.30 Hz, 2H, N₉CH₂), 3.37 (s, 3H, N₃CH₃),

3.50–3.60 (m, 5H, N₁CH₃+ CH₂N₉), 4.18 (t, J = 6.30 Hz,

2H, N₅CH₂), 7.16–7.31 (m, 5H, phenyl)

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6971

Table 5 (continued)

Compound IR m (cm^ꢁ1

)

UV k_max, loge

¹H NMR d (ppm)

2Ba

2Bb

2Bc

3081–3001-phenyl, 2962–2859-CH₂, 1697- 303.5, 4.28

CO (pos. 2), 1651-CO (pos. 4), 753-

phenyl, 707-CH₂

0.92–1.00 (m, 6H, 2 CH₃CH₂CH₂), 1.61–1.85 (m, 4H, 2

CH₃CH₂CH₂), 2.00–2.08 (m, 2H, CH₂CH₂CH₂), 2.95 (t,

J = 7.30 Hz, 2H, CH₂CH₂), 3.21 (t, J = 5.60 Hz, 2H,

N₉CH₂), 3.74 (t, J = 7.30 Hz, 2H, CH₂N₉), 3.91–4.04 (2 x

m, 4H, N₁,N₃-CH₂CH₂CH₃), 4.17 (t, J = 6.10 Hz, 2H,

N₅CH₂), 7.20–7.30 (m, 5H, phenyl)

3058-phenyl, 2958–2832-CH₂, 1699-CO

(pos. 2), 1653-CO (pos. 4), 1261-CH₃O-

phenyl, 1028-CH₃O-phenyl, 819-phenyl,

755-CH₂

302, 4.32

0.92–0.99 (m, 6H, 2 · CH₃CH₂CH₂), 1.60–1.85 (m, 4H,

2 · CH₃CH₂CH₂), 2.01–2.08 (m, 2H, CH₂CH₂CH₂), 2.89

(t, J = 7.30 Hz, 2H, N₉CH₂), 3.21 (t, J = 6.00 Hz, 2H,

CH₂N₉), 3.71 (t, J = 7.30 Hz, CH₂CH₂), 3.86 (s, 6H,

2 · OCH₃), 3.90–3.99 (m, 4H, N₁, N₃CH₂CH₂CH₃), 4,17

(t, J = 5.90 Hz, 2H, N₅CH₂),6.72–6.81 (m, 3H, phenyl)

0.93–1.02 (m, 6H, 2· CH₃CH₂CH₂), 1.61–1.88 (m, 4H, 2·

CH₃CH₂CH₂), 2.01–2.09 (m, 2H, CH₂CH₂CH₂), 3.11 (t,

J = 7.40 Hz, 2H, CH₂CH₂), 3.29 (t, J = 5.60 Hz, 2H,

N₉CH₂), 3.82 (t, J = 7.60 Hz, 2H, CH₂N₉), 3.93–4.08 (m,

4H, N₁, N₃-CH₂CH₂CH₃), 4.18 (t, J = 5.90 Hz, 2H,

N₅CH₂), 7.03–7.39 (m, 4H, phenyl), 7.74 (d, J = 8.90 Hz,

1H, 2⁰-pirol), 8.15 (s, 1H, NH)

3290-NH, 3061–3011-phenyl, 2958–2873-

CH₂, 1693-CO (pos. 2), 1650-CO (pos. 4),

756-phenyl, 740-CH₂

302.5, 4.26

5a

5b

2985–2883-CH₂, 1697-CO (pos. 2), 1657-

CO (pos. 4), 1043-aryl-O, 823-phenyl,

752-CH₂

300.6, 4.27

298.5, 4.59

2.01–2.05 (m, 2H, CH₂CH₂CH₂), 3.22 (s, 3H, N₃CH₃),

3.32 (s, 3H, N₁CH₃), 3.48 (t, J = 5.00 Hz, 2H, CH₂O), 3.81

(t, J = 6.00 Hz, 2H, N₉CH₂), 4.05 (t, J = 6.30 Hz, 2H,

CH₂N₉), 4.20 (t, J = 6.30 Hz, 2H, N₅CH₂), 6.95–7.01 (m,

2H, 2⁰6⁰-phenyl), 7.27–7.33 (m, 2H, 3⁰5⁰-phenyl)

2.12–2.26 (m, 2H, CH₂CH₂CH₂), 3.35 (s, 3H, N₃CH₃),

3.48 (s, 3H, N₁CH₃), 3.67 (t, J = 5.60 Hz, 2H, N₉CH₂),

3.93 (t, J = 4.82 Hz, 2H, CH₂N₉), 4.19–4.26 (m, 4H,

N₅CH₂CH₂O), 6.84 (d, J = 8.80 Hz, 1H, 6⁰-phenyl), 7.16

(dd, J = 6.32 Hz, J = 2.47 Hz, 1H, 5⁰-phenyl), 7.35 (d,

J = 2.47 Hz, 1H, 3⁰-phenyl)

3095-phenyl, 2942, 2881-CH₂, 1698-CO

(pos. 2), 1654-CO (pos. 4), 1042 (aryl-O-),

8.09-phenyl, 750-CH₂

5c

5d

3135-NH, 3080-phenyl, 2927-CH₂, 1699-

CO (pos. 2), 1653-CO (pos. 4), 1045 (aryl-

O-), 831-phenyl, 745-CH₂

4.60

2.08–2.20 (m, 5H, CH₂CH₂CH₂+ CH₃CO), 3.35 (s, 3H,

N₃CH₃), 3.46–3.57 (m, 5H, N₁CH₃+ N₉CH₂), 3.87 (t,

J = 5.13 Hz, 2H, CH₂N₉), 4.17–4.24 (m, 4H, N₅CH₂,

CH₂O), 6.83 (d, J = 8.79 Hz, 2H, 3⁰5⁰-phenyl), 7.26 (s, 1H,

NHCO), 7.37 (d, J = 8.40 Hz, 2H, 2⁰,6⁰-phenyl)

3032-phenyl, 2930, 2877-CH₂, 1702-CO

(pos. 2), 1658-CO (pos. 4), 1045 (aryl-O-),

813-phenyl, 755-CH₂

297.5

2.02–2.20 (m, 2H, CH₂CH₂CH₂), 2.27 (s, 3H, CH₃), 3.35

(s, 3H, N₃CH₃), 3.49 (s, 3H, N₁CH₃), 3.57 (t, J = 5.50 Hz,

2H, CH₂O), 3.88 (t, J = 5.13 Hz, 2H, N₉CH₂), 4.18 (m,

4H, N₉CH₂, CH₂N₉), 6.78 (d, J = 8.40 Hz, 2H, 2⁰,6⁰-

phenyl), 7.06 (d, J = 8.52 Hz, 2H, 3⁰,5⁰-phenyl)

6a

6b

6c

3066, 3033-phenyl, 2940, 2909, 2885-CH₂,

1698-CO (pos. 2), 1651-CO (pos.4), 1042

(aryl CH₂O), 819-phenyl, 741-CH₂

2.10 (q, J = 5.77 Hz, 2H, CH₂CH₂CH₂), 3.36 (s, 3H,

N₃CH₃), 3.44–3.50 (m, 5H, N₁CH₃+ CH₂O), 3.72 (s, 4H,

CH₂N₉, N₉CH₂), 4.19 (t, J = 5.90 Hz, 2H, N₅CH₂), 4.52

(s, 2H, CH₂-phenyl), 7.25–7.34 (m, 5H, phenyl)

3055-phenyl, 2936, 2870–CH₂, 1698-CO

(pos. 2), 1653-CO (pos.4), 1043 (aryl

CH₂O), 844-phenyl, 742-CH₂

295.5, 4.56

298.5, 4.58

2.11 (q, J = 5.20 Hz, 2H, CH₂CH₂CH₂), 3.35 (s, 3H,

N₃CH₃), 3.47–3.52 (m, 5H, N₁CH₃+ CH₂O), 3.71 (s, 4H,

CH₂N₉, N₉CH₂), 4.20 (t, J = 6.05 Hz, 2H, N₅CH₂), 4.48

(s, 2H, CH₂-phenyl), 7.19–7.30 (m, 4H, phenyl)

3055-phenyl, 2923, 2852-CH₂, 1698-CO

(pos. 2), 1653-CO (pos. 4), 1045 (aryl

CH₂O-), 825-phenyl, 7.42-CH₂

2.10 (q, J = 5.77 Hz, 2H, CH₂CH₂CH₂), 3.35 (s, 3H,

N₃CH₃), 3.46–3.50 (m, 5H, N₁CH₃+ CH₂O), 3.71 (s, 4H,

CH₂N₉, N₉CH₂), 4.19 (t, J = 6.05 Hz, 2H, N₅CH₂), 4.47

(s, 2H, CH₂phenyl), 6.97–7.03 (m, 2H, 3⁰,5⁰-phenyl), 7.22–

7.27 (m, 2H, 2⁰,6⁰-phenyl)

6d

7a

3079, 3055, 3025-phenyl, 2938, 2872,

2836, 2803-CH₂, 1696-CO (pos. 2), 1659-

CO (pos. 4), 999 (aryl CH₂O), 749-CH₂

302.5, 4.58, 251, 4.50 2.13 (q, J = 5.70 Hz, 2H, CH₂CH₂CH₂), 3.35 (s, 3H,

N₃CH₃), 3.46–3.53 (m, 5H, N₁CH₃+ CH₂O), 3.72 (s, 4H,

CH₂N₉, N₉CH₂), 4.14–4.22 (m, 2H, N₅CH₂), 6.18–6.27

(m, 1H, CH@), 6.55 (d, J = 15.95 Hz, 1H, aryl CH@),

7.23–7.35 (m, 5H, phenyl)

2950-CH₂, 1699-CO (pos. 2), 1654-CO

(pos. 4), 881-phenyl, 748-CH₂

300.0, 4.70

2.06–2.10 (m, 2H, CH₂CH₂CH₂), 3.14 (s, 3H, N₃CH₃),

3.29 (s, 3H, N₁CH₃), 3.14–3.21 (m, 2H, CH₂NH), 3.34 (t,

J = 7.40 Hz, 2H, CH₂N₉), 3.73 (t, J = 6.05 Hz, 2H,

N₉CH₂), 4.03 (t, J = 5.91 Hz, 2H, NHCH₂), 4.19 (t,

J = 5.64 Hz, 2H, N₅CH₂), 7.38–7.43 (m, 3H, 3⁰,4’,5⁰-

phen_þyl), 7.45–7.56 (m, 2H, 2⁰,6⁰-phenyl), 9.48 (s, 2H,

NH₂

)

(continued on next page)

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6972

Table 5 (continued)

Compound

7b

IR m (cm^ꢁ1

)

UV k_max, loge

¹H NMR d (ppm)

3043-phenyl, 2822-CH₂, 1697-CO (pos.

2), 1644-CO (pos. 4), 837-phenyl, 750-

CH₂

300.0, 4.70

1.99–2.04 (m, 2H, CH₂CH₂CH₂), 2.69 (t, J = 6.33 Hz, 2H,

CH₂NH), 3.14 (s, 3H, N₃CH₃), 3.29 (s, 3H, N₁CH₃), 3.36

(t, J = 5.64 Hz, 2H, CH₂N₉), 3.52 (t, J = 6.33 Hz, 2H,

N₉CH₂), 3.67 (s, 2H, NHCH₂), 4.03 (t, J = 6.05 Hz, 2H,

N₅CH₂), 7.04–7.10 (m, 2H, 3⁰,5⁰-phenyl), 7.28–7.33 (m,

2H, 2⁰,6⁰-phenyl)

8a

8b

8c

8d

3289-NH, 3076-phenyl, 2949-CH₂, 1727-

CONH, 1698-CO (pos. 2), 1650-CO (pos.

4), 1068-CH₂O, 849-phenyl, 759-CH₂

300.0, 5.20

300.0, 4.60

300.0, 4.80

300.0, 4.60

2.03–2.07 (m, 2H, CH₂CH₂CH₂), 3.11 (s, 3H, N₃CH₃),

3.18 (s, 3H, N₁CH₃), 3.44 (t, J = 5.50 Hz, 2H, CH₂N₉),

3.74 (t, J = 5.23 Hz, 2H, N₉CH₂), 4.03 (t, J = 5.91 Hz, 2H,

N₅CH₂), 4.29 (t, J = 5.23 Hz, 2H, CH₂O), 6.91–6.96 (m,

1H, 4’-phenyl), 7.20 (d, J = 7.98 Hz, 2H, 3⁰,5⁰-phenyl),

7.37 (d, J = 7.43 Hz, 2H, 2⁰,6⁰-phenyl), 9.52 (s, 1H, NH)

2.03–2.07 (m, 2H, CH₂CH₂CH₂), 3.11 (s, 3H, N₃CH₃),

3.17 (s, 3H, N₁CH₃), 3.43 (t, J = 5.36 Hz, 2H, CH₂N₉),

3.75 (t, J = 4.96 Hz, 2H, N₉H₂), 4.03 (t, J = 5.78 Hz, 2H,

N₅CH₂), 4.31 (t, J = 5.01 Hz, 2H, CH₂O), 6.97–6.99 (m,

1H, 5⁰-phenyl), 7.25 (m, 2H, 4⁰,6⁰-phenyl), 7.49 (s, 1H, 2⁰-

phenyl), 9.74 (s, 1H, NH)

3289-NH, 3083-phenyl, 2943-CH₂, 1733-

CONH, 1696-CO (pos. 2), 1640-CO (pos.

4), 1064-CH₂O, 748-CH₂

3300-NH, 3071-phenyl, 2939-CH₂, 1731-

CONH, 1695-CO (pos. 2), 1646-CO (pos.

4), 1074-CH₂O, 841-phenyl, 749-CH₂

2.03–2.06 (m, 2H, CH₂CH₂CH₂), 3.11 (s, 3H, N₃CH₃),

3.18 (s, 3H, N₁CH₃), 3.43 (t, J = 5.37 Hz, 2H, CH₂N₉),

3.73 (t, J = 5.01 Hz, 2H, N₉CH₂), 4.03 (t, J = 5.78 Hz, 2H,

N₅CH₂), 4.28 (t, J = 5.09 Hz, 2H, CH₂O), 7.05 (t,

J = 8.80 Hz, 2H, 3⁰,5⁰-phenyl, 7.37 (br s, 2H, 2⁰,6⁰-phenyl),

9.58 (s, 1H, NH)

3299-NH, 2939-CH₂, 1735-CONH, 1698-

CO (pos.2), 1638-CO (pos. 4), 1088-

CH₂O, 750-CH₂

2.03–2.07 (m, 2H, CH₂CH₂CH₂), 3.11 (s, 3H, N₃CH₃),

3.17 (s, 3H, N₁CH₃), 3.42–3.45 (m, 2H, CH₂N₉), 3.74 (t,

J = 4.98 Hz, 2H, N₉CH₂), 4.03 (t, J = 5.78 Hz, 2H,

N₅CH₂), 4.29 (t, J = 4.95 Hz, 2H, CH₂O), 7.25 (d,

J = 8.80 Hz, 2H, 3⁰,5⁰-phenyl), 7.38 (d, J = 7.43 Hz, 2H,

2⁰,6⁰-phenyl), 9.68 (s, 1H, NH)

The structure was solved by a direct method using the

SHELXTS program⁵²and reﬁned with the SHEL-

XTL.⁵³E-map provided positions for all non-H-atoms.

The full-matrix least-squares reﬁnement was carried out

on F²s using anisotropic temperature factors for all non-

H-atoms. The H-atoms were located from Dq-map, then

the positions of H-atoms were reﬁned in the riding mod-

el with isotropic thermal parameters. Crystallographic

data (excluding structural factors) for the structure re-

ported in this paper have been deposited with the Cam-

bridge Crystallographic Data Centre and allocated the

deposition number: CCDC 628859. Copies of the data

can be obtained free of charge on application to CCDC,

12 Union Road, Cambridge CB2 1EW, UK (fax:

+(1223)336-033; E-mail:deposit@ccdc.cam.ac.uk).

7.7. Computational procedures

For all calculations CAChe 6.1.1³⁷software was used.

The starting models of the molecules were based on crys-

tallographic data for 6d using the PCMODEL 6 pro-

gram.⁵⁴The structures of the synthesized compounds

Table 6. Molecular descriptors and methods for their calculation

Descriptor

Method of calculation

LogP; Molar refractivity

Atom typing scheme of Ghose and Crippen

Calculated from the chemical sample atoms and bonds

Extracted from the chemical sample

Connectivity indexes 0–2; Valence connectivity indexes 0–2

Single, double bonds; Carbon, Oxygen atoms

Shape index order 1–3

Determined by the topological structure of the chemical sample

Calculated at the molecule geometry in water from optimization

using MOPAC with PM5 parameters and the Conductor like

Screening Model (COSMO)

Solvent accessible surface area; dielectric energy

Dipole moment; Dipole vectors X, Y, Z

Polarizability; HOMO, LUMO energies

Calculated by DFT using D-VWN LDA functional with DZVP

basis set at current geometry

Calculated by DFT using the B88-LYP GGA energy functional

with the 6-31G** basis sets

Steric energy

Surface integral of electrostatic isopotential; surface area

Determined by mechanics using augmented MM3

Generated by a Dgauss/DFT wave function using the B88-LYP

GGA functional with the 6-31G** basis sets

Surface integral of electrostatic potential on electron density;

surface area

Calculated by a Dgauss/DFT wave function using the B88-LYP

GGA functional with the 6-31G** basis sets

Surface integral of electrostatic potential on van der Waals surface;

surface area

Calculated by a DGauss/DFT wave function using the B88-LYP

GGA functional with the 6-31G** basis sets

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6973

´

15. Drabczynska, A.; Muller, C. E.; Lacher, S. K.; Schu-

¨

macher, B.; Karolak-Wojciechowska, J.; Nasal, A.; Kaw-

were built and geometry was optimized in MOPAC

(Molecular Orbital Package) using semi-empirical Ham-

iltonian to minimize the energy. To ﬁnd the low-energy

conformers, CONFLEX application was used. The

energy of each conformation was plotted in a three-

dimensional graph. The low-energy conformations,

separated by high energy barriers, were collected and

optimized using DFT methods B88-LYP functional with

6-31G** basis set. The descriptors were calculated in

CAChe 6.1.1³⁷(Table 6).

´

czak, P.; Yuzlenko, O.; Pe˛kala, E.; Kiec-Kononowicz, K.

Bioorg. Med. Chem. 2006, 14, 7258.

´

16. Drabczynska, A.; Muller, C. E.; Karolak-Wojciechowska,

J.; Schumacher, B.; Schiedel, A.; Yuzlenko, O.; Kiec-

¨

´

Kononowicz, K. Bioorg. Med. Chem. 2007, 15, 5003.

´

17. Rockitt, S.; Wartchow, R.; Duddeck, H.; Drabczynska,

A.; Kiec-Kononowicz, K. Z. Naturforsch. 2001, 56b, 319.

´

18. Pawłowski, M.; Drabczynska, A.; Gorczyca, M.; Malec,

D.; Modzelewski, J. Pol. J. Pharm. 1991, 43, 61.

19. Eckstein, M. Dissert. Pharm. 1962, 14, 435.

20. Mitsunobu, O. Synthesis 1981, 1.

21. El-Shaieb, K. M. Heteroat. Chem. 2006, 17, 365.

22. McCarroll, A. J.; Sandham, D. A.; Titcomb, L. R.; de K.

Lewis, A. K.; Cloke, F. G. N.; Davies, B. P.; Perez de

Santana, A.; Hiller, W.; Caddick, S. Mol. Divers. 2003, 7,

115.

Acknowledgments

The authors are grateful to Professor James Stables for

providing the biological data through the ADD Pro-

gram of the National Institute of Neurological and

Communicative Disorders and Stroke at the National

Institutes of Health, Bethesda, MD, USA. We thank

23. Bergmann, F.; Dikstein, S. J. Am. Chem. Soc. 1955, 77,

691.

´

ˇ

24. Rybar, A.; Antos, K. Collect. Czech. Chem. Commun.

1970, 35, 1415.

25. Muller, C. E.; Maurinsh, J.; Sauer, R. Eur. J. Pharm. Sci.

Professor Dr. K.-N. Klotz, University of Wurzburg,

¨

2000, 10, 259.

for the gift of recombinant CHO cells expressing adeno-

sine receptor subtypes. This work was supported by Pol-

ish State Committee for Scientiﬁc Research (Grant No.

6 P05F 024 21) and by the Deutsche Forschungsgeme-

inschaft (GRK 677).

26. Klotz, K.-N.; Lohse, M. J.; Schwabe, U.; Cristalli, G.;

Vittori, S.; Grifantini, M. Naunyn Schmiedeberg’s Arch.

Pharmacol. 1989, 340, 679.

27. Ji, X.-D.; Jacobson, K. A. Drug Des. Discov. 1999, 16, 217.

28. Yan, L.; Muller, C. E. J. Med. Chem. 2004, 47, 1031.

¨

29. Yan, L.; Bertarelli, D. C. G.; Hayallah, A. M.; Meyer, H.;

Klotz, K.-N.; Muller, C. E. J. Med. Chem. 2006, 49,

4384.

30. Muller, C. E.; Diekmann, M.; Thorand, M.; Ozola, V.

¨

Bioorg. Med. Chem. Lett. 2002, 12, 501.

31. Porter, R. J.; Cereghino, J. J.; Gladding, G. D.; Kupfer-

berg, H. J.; Scoville, B.; White, B. G. Clev. Clin. 1984, 51,

293.

¨

Supplementary data

Supplementary data associated with this article can be

found, in the online version, at doi:10.1016/j.bmc.2007.

07.051.

32. Krall, R. L.; Penry, J. K.; White, B. G.; Kupferberg, H. J.;

Swinyard, E. A. Epilepsia 1978, 19, 409.

33. Stables, J. P.; Kupferberg, K. J. In Molecular and Cellular

Targets for Antiepileptic Drugs; John Libbey: London,

1998; pp 191–198.

References and notes

1. Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz,

K.-N.; Linden, J. Pharmacol. Rev. 2001, 53, 527.

2. Ralevic, V.; Burnstock, G. Pharmacol. Rev. 1998, 50.

34. Barton, M. E.; Klein, B. D.; Wolf, H. H.; White, S.

Epilepsy Res. 2001, 47, 217.

3. Muller, C. E. Curr. Med. Chem. 2000, 7, 1269.

¨

35. Kaminski, R. M.; Livingood, M. R.; Rogawski, M. R.

Epilepsia 2004, 45, 864.

36. Gao, Z. G.; IJzerman, A. P. Biochem. Pharmacol. 2000,

60, 669.

4. Dhalla, A. K.; Shryock, J. C.; Shreeniwas, R.; Bellardi-

nelli, L. Curr. Top. Med. Chem. 2003, 3, 369.

5. Poulsen, S.-A.; Quinn, R. J. Bioorg. Med. Chem. 1998, 6,

619.

37. CAChe, Version 6.1.1. Sunnyvale, Fujitsu Computer

Systems, 2002.

6. Muller, C. E. Exp. Opin. Ther. Patents 1997, 7, 419.

¨

7. Hess, S. Exp. Opin. Ther. Patents 2001, 11, 1533.

38. Hawkins, D. M.; Basak, S. C.; Mills, D. J. Chem. Inf.

Comput. Sci. 2003, 43, 579.

´

8. Yuzlenko, O.; Kiec-Kononowicz, K. Curr. Med. Chem.

2006, 13, 3609.

39. Golbraikh, A.; Shen, M.; Xiao, Z.; De Xiao, Y.; Lee, K.-

H.; Tropsha, A. J. Comp. Aided Mol. Des. 2003, 17, 241.

40. Golbraikh, A.; Tropsha, A. J. Mol. Graphics Modell. 2002,

20, 269.

9. Muller, C. E. Drugs Future 2000, 25, 1043.

¨

10. 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.

41. Hawkins, D. M. J. Chem. Inf. Comput. Sci. 2004, 44, 1.

42. Cronin, M. T.; Schultz, T. W. THEOCHEM 2003, 622,

39.

43. He, L.; Jurs, P. C. J. Mol. Graphics Modell. 2005, 23, 503.

44. Verma, R. P.; Hansch, C. Bioorg. Med. Chem. 2005, 13,

4597.

´

11. Rockitt, S.; Duddeck, H.; Drabczynska, A.; Kiec-Kono-

nowicz, K. Eur. J. Org. Chem. 2000, 3489.

´

12. Kiec-Kononowicz, K.; Drabczynska, A.; Pe˛kala, E.;

Michalak, B.; Muller, C. E.; Schumacher, B.; Karolak-

¨

´

Wojciechowska, J.; Duddeck, H.; Rockitt, S.; Wartchow,

R. Pure Appl. Chem. 2001, 73, 1411.

45. Hockemeyer, J.; Burbiel, J. C.; Muller, C. E. J. Org. Chem.

´

13. Drabczynska, A.; Schumacher, B.; Muller, C. E.; Karolak-

Wojciechowska, J.; Michalak, B.; Pe˛kala, E.; Kiec-Kon-

¨

2004, 69, 3308.

46. Burbiel, J.; Thorand, M.; Muller, C. E. Heterocycles 2003,

´

¨

onowicz, K. Eur. J. Med. Chem. 2003, 38, 397.

60, 1425.

47. Bertarelli, D. C. G.; Diekmann, M.; Hayallah, A. M.;

´

14. Drabczynska, A.; Muller, C. E.; Schumacher, B.; Hinz, S.;

¨

Karolak-Wojciechowska, J.; Michalak, B.; Pe˛kala, E.;

Rusing, D.; Iqbal, J.; Preiss, B.; Verspohl, E. J.; Muller, C.

´

Kiec-Kononowicz, K. Bioorg. Med. Chem. 2004, 12,

4895.

¨

E. Purinergic Signalling, in press.

¨

´

A. Drabczynska et al. / Bioorg. Med. Chem. 15 (2007) 6956–6974

6974

48. Bulicz, J.; Bertarelli, D. C. G.; Baumert, D.; Fulle, F.;

¨

Muller, C. E.; Heber, D. Bioorg. Med. Chem. 2006, 14,

51. Sauer, R.; Maurinsh, J.; Reith, U.; Fulle, F.; Klotz, K.-N.;

¨

Muller, C. E. J. Med. Chem. 2000, 43, 440.

¨

2837.

52. Sheldrick, G. M. SHELXTL PC^MT, Siemens Analytical

X-Ray Instruments Inc., Madison, Wisconsin, USA, 1990.

53. Sheldrick G. M. SHELXL-97.A FORTRAN-77 Program

for the Reﬁnement of Crystal Structures from Diﬀraction

Data. University of Go¨ttingen, Germany, 1997.

49. Klotz, K.-N.; Hessling, J.; Hegler, J.; Owman, C.; Kull,

B.; Fredholm, B. B.; Lohse, M. J. Naunyn Schmiedeberg’s

Arch. Pharmacol. 1998, 357, 1.

´

50. Hayallah, A. M.; Sandoval-Ramırez, J.; Reith, U.; Scho-

bert, U.; Preiss, B.; Schumacher, B.; Daly, J. W.; Muller,

¨

54. PCMODEL 6 Molecular Modeling Software for the IBM

PC, Serena Software, Bloomington; 1996.

C. E. J. Med. Chem. 2002, 45, 1500.

Article Doi

Phenylethyl-substituted pyrimido[2,1-f]purinediones and related compounds: Structure-activity relationships as adenosine A1 and A2A receptor ligands

DOI: 10.1016/j.bmc.2007.07.051

Source and publish data:

Authors:

Article abstract of DOI:10.1016/j.bmc.2007.07.051

Full text of DOI:10.1016/j.bmc.2007.07.051

Products guided by the article

R&D Labs maybe for 955121-89-8

Relevant to this article

Hot Product