2,6-Disubstituted and 2,6,8-trisubstituted purines as adenosine receptor antagonists

Article abstract of DOI:10.1021/jm050640i

Purines have long been exploited as adenosine receptor antagonists. The substitution pattern about the purine ring has been well investigated, and certain criteria have become almost a prerequisite for good affinity at the adenosine A₁ receptor

Full text of DOI:10.1021/jm050640i

J. Med. Chem. 2006, 49, 2861-2867
2861
Articles
2,6-Disubstituted and 2,6,8-Trisubstituted Purines as Adenosine Receptor Antagonists
Lisa C. W. Chang, Ronald F. Spanjersberg, Jacobien K. von Frijtag Drabbe Ku¨nzel, Thea Mulder-Krieger,
Johannes Brussee, and Adriaan P. IJzerman*
Leiden/Amsterdam Center for Drug Research, DiVision of Medicinal Chemistry, P.O. Box 9502, 2300 RA Leiden, The Netherlands
ReceiVed July 5, 2005
Purines have long been exploited as adenosine receptor antagonists. The substitution pattern about the purine
ring has been well investigated, and certain criteria have become almost a prerequisite for good affinity at
the adenosine A1 receptor. The adaptation of the pharmacophore and the initial series of pyrimidines developed
in an earlier publication resulted in a series of purines with an entirely new substitution pattern. One compound
in particular, 8-cyclopentyl-2,6-diphenylpurine (31, LUF 5962) has been shown to be very promising with
an affinity of 0.29 nM at the human adenosine A1 receptor.
Introduction
The adenosine receptors belong to the superfamily of G-
protein-coupled receptors. There are four subclasses of adenosine
receptors that have been identified to date, and these are A1,
A2A, A2B, and A3 receptors. The adenosine receptors are
distributed around many different tissue types in the human body
in varying levels of expression.1 This has led to intense research
into the therapeutic potential of the ability to control the activity
of adenosine receptors. For example, BG-9928 (bicyclo[2.2.2]-
octane-1-propanoic acid, 4-(2,3,6,7-tetrahydro-2,6-dioxo-1,3-
dipropyl-1H-purin-8-yl)-) is an A1 receptor antagonist in phase
II clinical development for renal chronic heart failure.
We have recently shown that a simple pharmacophore can
lead to new types of adenosine A1 receptor antagonists.2 The
resulting 2,4,6-trisubstituted pyrimidines were shown to be both
potent at and selective for the human A1 receptor. In an attempt
to judge the relative location of the hydrogen-bond acceptor
Figure 1. (a) 2,6-Diphenyl-4-amidopyrimidine, a high-affinity an-
tagonist at adenosine A1 receptors (ref 2). (b) Fixation of the hydrogen-
bond acceptor at the top of the molecule. (c) Change of the heteroatom
to accomplish this fixation. From the latter molecule, a more generalized
purine scaffold is proposed, which we used as a starting point in our
synthetic program.
close to the top of a central aromatic group and a more precise
orientation of the L2 lipophilic group in relation to the central
ring, the fixation of this group was considered. Figure 1a-c
shows the logical development of the 4-amido-2,6-diphenylpy-
rimidines into the purines explored in this article. Fixing the
hydrogen-bond-accepting group at the top of the central group
prompted the change of the heteroatom for both synthetic ease
and the preservation of the Cdheteroatom unsaturated bond.
Purines have been explored at length as adenosine receptor
antagonists.3 However, these have mainly been direct analogues
of adenine, in the sense that the N6 group has always been
preserved. This N6 amino group has usually been substituted
with a cyclopentyl group to attain good affinity for the A1
receptor.4-9 Further exploration of this central core also deems
the necessity of N9 substitution for potency, with both large
benzyl derivatives7 and small methyl9 or ethyl substituents8
showing good affinity for the A1 receptor. De Ligt et al.
postulated certain features to enhance the selectivity of the basic
adenine ring (Figure 2).9 The resulting series of compounds
displayed good affinity for the human adenosine A1 receptor,
with N6-cyclopentyl-8-(N-methyl-isopropylamino)-9-methylad-
Figure 2. Features to enhance affinity and selectivity at the A1 receptor
and the numbering around the purine ring (adapted from ref 7).
enine showing the best Ki value at 7.7 nM. These types of
compounds with the presence of the N6-group and with N9
substitution show good affinity at the A1 receptor, corroborating
* To whom correspondence should be addressed. Tel: +31 (0)71 527
4651. Fax: +31 (0)71 527 4565. E-mail: ijzerman@lacdr.leidenuniv.nl.
10.1021/jm050640i CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/14/2006

2862 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10
Chang et al.
reaction with an excess of the second boronic acid gave the
desired 2,6-diphenyl derivatives.
For substitution at the 8-position, the bromination of a
protected 2,6-dichloropurine to provide a more versatile inter-
mediate was attempted following the procedure described by
De Ligt et al.9 in their evaluation of N-0840 analogues.
Unfortunately, this was unsuccessful, and the 8-substituent was
introduced into the purine frame at an earlier stage, as described
in Scheme 2. To make vital intermediate 18, a four step proced-
ure described in a number of articles by Biagi et al. was broadly
exploited.18-20 The reaction of commercially available benza-
midine with diethyl malonate created a pyrimidine ring, after
which various substitution steps resulted in intermediate 18. Ring
closure was attempted by several different methods that are
detailed in the literature, namely, using trimethylorthoformate,
triethylorthoformate, or PPA as the reaction medium.21,22 These
methods were either poor in yields or were difficult to work up
or purify. Finally, a two-step reaction incorporating the addition
of the appropriate acid chloride followed by cyclization under
basic conditions provided a more ideal synthetic pathway.23 The
8-substituted-2,6-dichloropurine was then subjected to metal-
catalyzed cross-coupling reactions as described above. It was
discovered that protection at N9, although readily applicable
(24-27) was no longer necessary, probably, because of the
increased steric bulk around this position preventing the
complexation of the catalyst and/or the boronic acid. Methylation
of compound 32 following standard procedures with methyl
iodide in basic conditions gave compound 37.
Figure 3. Model detailed in an earlier publication (upper part, see ref
2), and the two possible orientations of the traditionally substituted
adenines to fit this model (see text).
with the proposed model2 in the manner shown in Figure 3.
The hydrogen-bond-donating group (denoted B) is represented
by the available hydrogen on the N6, and the hydrogen-bond-
acceptor groups A or C may be one of the two ring nitrogens
(N1 or N7) depending on the orientation and the size of possible
substituents on the other positions of the purine ring. The
lipophilic pocket denoted L2 may be filled with the substituent
labeled R1, and either L1 or L3 are satisfied by the R2
substituent. Optimal receptor interaction may be provided by
substitution in the C2 position, as found by Bianucci et al. in
their investigations.7
In this article, a new perspective of the purines is realized
according to the model proposed in our earlier publication,2
where excellent affinity for the A1 adenosine receptor is
achieved, despite the lack of the seemingly essential N6 and
N9 groups/substituent proposed in earlier articles. Direct
aromatic substitution at the C2 and C6 positions (labeled R1
and R2 in the purine scaffold in Figure 1) provides an analogy
of the 2,6-aromatic substituents on the 4-amino-pyrimidines.2
C8 substitution (R3) explores the L2 pocket, and R4 should be,
according to the model, left unsubstituted to achieve high affinity
at the A1 adenosine receptor.
Structure-Activity Relationships. The results of the ra-
dioligand-binding assays performed on these purines are shown
in Table 1. Compounds 11-17 are unsubstituted in the C8
position and are varied at C2 and C6. 2,6-Diphenyl-9H-purine
(11) has an affinity of 4 nM for the hA1 receptor. The selectivity
is approximately 10-fold better for this receptor than that for
the A2A and the A3 receptors at 53 and 38 nM, respectively.
An identical substitution at the 4-positions of both of the phenyl
groups (12-14) is to the great detriment of the affinity for the
A1 and A2A receptors, in contrast to the improved affinity (9
nM) at the A3 receptor by the bis-4-methyl substitution (13).
The single substitution of just the C2 phenyl group shows no
enhancement for the A1 receptor, with both the 4-Cl (15) and
the 4-MeO (17) moieties reducing the affinity of the ligand for
this receptor, whereas compound 16 possessing the 4-Me group
retains the affinity of the unsubstituted diphenylpurine at 4 nM.
The great beneficiary of this particular substitution pattern is
the A3 receptor, where both the 4-Me (16) and the 4-MeO (17)
compounds display significantly enhanced affinities (Ki values
of 9 and 4 nM, respectively) when compared to that of
compound 11. 4-Methoxy-substitution of phenyl groups have
usually been shown to enhance the affinity for the A3 receptor
in a number of different series of adenosine antagonists;24,25
thus, it may seem surprising that the bis-4-Me substituted variety
(13) displayed significantly higher affinity for the A3 receptor
than the analogous bis-4-MeO ligand (14). However, the results
of the monosubstituted compounds (16, 9 nM and 17, 4 nM)
confirm that the 4-MeO group is still preferred by the A3
receptor. Bis-4-MeO substitution is probably just too large for
optimal binding in the A3 receptor pocket, accounting for its
poorer affinity compared to that of the slightly smaller bis-4-
Me variation. It seems that for the A1 receptor, unsubstituted
phenyl groups may be the optimal R1 and R2 groups for the
affinity of 9H-purines.
Chemistry
The purines were made via two routes. The 2,6-disubstituted
purines were synthesized from commercially available 2,6-
dichloropurine as described in Scheme 1. Substitution of the
chlorines was possible through metal-mediated cross-coupling
reactions. Both Stille and Suzuki-Miyaura couplings of purines
have been reported in the literature.10-17 To facilitate the Suzuki
cross-coupling, protection of the nitrogen at N9 was necessary.
This was successfully achieved with tetrahydropyran (THP).
Initial attempts with a benzylic group rendered the product far
too stable, and the removal of the protecting group was
impossible under usual hydrogenation or transfer hydrogenation
conditions. Substitution of both chlorines with the same phenyl
derivative occurred under standard Suzuki conditions as detailed
by Hocek et al. (compounds 11-14).13 However, microwave
heating to a temperature of 150 °C instead of conventional
heating methods reduced the reaction time from 8 h to
approximately 20 min. An excess of the boronic acid (3
equivalents) encouraged the reactions to completion. Where the
2 and 6 substituents differed, one equivalent of the boronic acid
provided a single substitution at the 6 position (compounds
15-17). This was confirmed by X-ray crystallography. Further
Exploration of the C8 position of the purine led, in general,
to significant improvements in the affinity for the A1 receptor.

2,6-Disubstituted and 2,6,8-Trisubstituted Purines
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10 2863
Scheme 1a
a (a) dihydropyran, pTSA, THF; (b) 1 equiv Ph2B(OH)2, K2CO3, Pd(PPh3)4, PhMe; (c) 3 equiv R2B(OH)2, K2CO3, Pd(PPh3)4, PhMe; (d) 1.2 eq R3B(OH)2,
K2CO3, Pd(PPh3)4, PhMe; (e) dowex, EtOH.
Scheme 2a
a (a) (i) R1COCl, pyridine, DCM; (ii) 2M NaOH; (b) dihydropyran, pTSA, THF; (c) 1.5 equiv R2B(OH)2, K2CO3, Pd(PPh3)4, PhMe; (d) dowex, EtOH;
(e) 1.5 equiv R2B(OH)2, K2CO3, Pd(PPh3)4, PhMe; (f) NaH, MeI, DMF.
A phenyl group in the 8-position (28) is evidently less
appreciated in the A1 receptor binding pocket with a drop in
affinity to 21 nM. Both the A2A and A3 receptors seem to tolerate
this much larger substituent well, on the whole retaining the
affinity achieved by the comparable unsubstituted compound
11. This indicates, perhaps, more space in this part of the
respective A2A and A3 receptor pockets than in the A1 receptor
site. The single straight-chained alkyl group (nPr, 29) showed
a Ki value of 4 nM, improving on the selectivity over the A2A
receptor when compared to that of the unsubstituted form (11)
by a factor of 3. However, the affinity at the A3 receptor also
showed an improvement to 17 nM. The most active compounds
at the adenosine A1 receptor registered subnanomolar affinity.
These were the C8-isopropyl (30), cyclopentyl (31), and
cyclohexyl (32) derivatives at 0.82, 0.29, and 0.73 nM,
respectively. In particular, the cyclopentyl moiety (31) displayed
an impressive gain in affinity at the A1 receptor, while retaining
the same degree of affinity at the A2A and A3 receptors in com-
parison to that of ligand 11. In comparison to the cyclopentyl
derivative, the cyclohexyl compound (32) was only slightly
lower in affinity at the A1 receptor, yet showed a significant
drop in affinity at the A2A and A3 receptors. Thus, this compound
had, on the whole, better A2A/A1 and A3/A1 selectivity ratios at
160 and 274, respectively (for 31 A2A/A1 ) 190 and A3/A1 )
121). Its cyclohexyl substituent is also more flexible than the
similarly sized phenyl group in 28, which is apparently capable
of accommodating the binding pocket much better.
in affinity was noted across the three receptors, although the
4-MeO moiety seemed to interfere the least, retaining an affinity
of 4 nM for the A1 receptor. Compound 37 highlights the
importance of the free N9 proton to act as a hydrogen-bond
donor because the methyl substitution results in a complete loss
of affinity at the adenosine receptors.
Comparing the most potent compound at the human A1
receptor of this series, 31 (0.29 nM), to the most potent purine
published to date, N6-cyclopentyl-8-(N-methyl-isopropylamino)-
9-methyladenine (7.7 nM),9 it may be speculated that the affinity
to the A1 receptor is much improved (almost 30-fold) because
both the L1 and the L3 pockets are now subject to interaction
with the ligand. In Figure 4, the two possible ways of
superimposing these compounds is depicted, following the
proposed model as mentioned in Figure 3. The superimposition
shown in Figure 4a suggests that there are no interactions of
the L1 pocket with N6-cyclopentyl-8-(N-methyl-isopropylamino)-
9-methyladenine, as provided by the C2-phenyl group of 31,
whereas the one shown in 4b suggests that there likewise is
little interaction of the L3 pocket with N6-cyclopentyl-8-(N-
methyl-isopropylamino)-9-methyladenine, as provided by the
C6-phenyl of 31. In the lowest energy state of the compound,
the 8-substituent of N6-cyclopentyl-8-(N-methyl-isopropylamino)-
9-methyladenine is also, most probably, orientated out of the
plane of the core heterocycle, although this may not be a
significant factor upon the final ligand-receptor complex.
By immobilizing the H-bond acceptor at the top of the
molecule to form the fused heterocyclic ring, the position and,
thus, the orientation toward the receptor of the L2 group is
shifted slightly higher than that predicted in the model detailed
To see whether both C8 substitution and the substitution at
one of the phenyl groups would be well tolerated, compounds
33-36 were synthesized and tested. Again, a significant loss

2864 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10
Chang et al.
Table 1. Affinities of the Di- and Trisubstituted Purines 11-35 in Radioligand-Binding Assays of Human Adenosine Receptors
Ki (nM) or % disp.a
b
c
d
R1
R2
R3
R4
hA1
hA2A
hA3
11
12
13
14
15
16
17
28
Ph
Ph
H
H
H
H
H
H
H
Ph
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me
4.1 ( 0.5
21%
235 ( 25
40%
27 ( 6
4.0 ( 0.6
32 ( 6
21 ( 12
4.4 ( 0.7
0.82 ( 0.08
0.29 ( 0.07
0.73 ( 0.07
25 ( 5
53 ( 14
11%
4%
38 ( 12
68 ( 11
9.3 ( 3
165 ( 38
94 ( 10
9.4 ( 4
4.0 ( 2
49 ( 19
17 ( 9
9.5 ( 0.9
35 ( 13
200 ( 46
243 ( 89
37.2%
4-ClPh
4-MePh
4-MeOPh
4-ClPh
4-MePh
4-MeOPh
Ph
Ph
Ph
Ph
Ph
4-ClPh
4-MePh
4-MeOPh
Ph
Ph
Ph
Ph
Ph
Ph
Ph
37%
315 ( 96
289 ( 19
47%
66 ( 4
167 ( 25
148 ( 27
55 ( 3
118 ( 17
24%
29
Pr
30 (LUF 5956)
31 (LUF 5962)
32 (LUF 5957)
33
34
35
36
37
iPr
cPent
cHex
cHex
cHex
cHex
cHex
cHex
Ph
Ph
Ph
Ph
Ph
4-ClPh
3,4-diClPh
4-MePh
4-MeOPh
Ph
37 ( 10
36 ( 4
3.7 ( 0.5
24%
20.3%
9%
27%
41%
138 ( 42
25%
Ph
0%
a Ki ( SEM (n ) 3), % displacement (n ) 2). b Displacement of specific [3H]DPCPX binding in CHO cells expressing human adenosine A1 receptors
or % displacement of specific binding at 1 µM concentrations. c Displacement of specific [3H]ZM 241385 binding in HEK 293 cells expressing human
adenosine A2A receptors or % displacement of specific binding at 1 µM concentrations. d Displacement of specific [125I]AB-MECA binding in HEK 293
cells expressing human adenosine A3 receptors or % displacement of specific binding at 1 µM concentrations.
the previous publication.2 Thus, it is appropriate to assume that
it is not the size of the aromatic ring that is particularly important
but the relative positioning to the aromatic center of the
hydrogen-bond donor B and the acceptor C.
We next performed functional assays with these compounds
to determine their effect at the adenosine A1 receptor. First, the
three most potent compounds at the human A1 receptor (30,
31, and 32, Ki ) 0.82, 0.29, and 0.73 nM, respectively) were
selected and tested in cAMP assays to assess their ability to
Figure 4. Two possible ways of superimposing compounds 31 and
N6-cyclopentyl-8-(N-methylisopropylamino)-9-methyladenine (ref 9).
agonize or antagonize the A1 receptor. Figure 6 depicts the
results of these experiments. Because of the adenosine A1
receptor’s coupling to an inhibitory G protein, which leads to a
decrease in the levels of cAMP upon stimulation, forskolin was
added to generate a measurable amount of cAMP in the system.
The amount of cAMP generated by the sole addition of forskolin
was thus set at 100%. CPA (1 µM), as the reference full agonist
of the A1 receptor, reduced the levels of cAMP present in the
system, and this was thus set at 0% for the calculations. N0840
(reported in [35S]GTPγS assays to be a neutral antagonist)9,26
was shown in this whole cell cAMP assay to have properties
consistent with an inverse agonist at a concentration of 100 µM,
although on lower levels than DPCPX (the reported full inverse
agonist, at a concentration of 1 µM). This discrepancy between
the two assays is not necessarily conflicting because intrinsic
activity is also dependent on the assay system examined. At a
concentration of 100 nM (>100 × Ki values, to ensure full
receptor occupancy), compounds 30, 31, and 32, all increased
the measured-cAMP levels above the levels measured with
forskolin alone. Compounds 30 and 32 were less efficacious
than N-0840, however, suggesting that the behavior is that of
neutral antagonists. We next recorded a full concentration-
response curve for compound 32 (Figure 7), antagonizing the
effect of 100 nM CPA on forskolin-induced (10 µM) cAMP
production. A similar experiment was performed for compound
30. The IC50 values for the two compounds (n ) 3) were 460
( 100 nM (30) and 477 ( 137 nM (32), yielding KB values of
in our previous publication.2 This slight relocation of the L2
group has brought about substantial improvements in the affinity
for the A1 receptor. This is best illustrated by comparing the
analogous compounds, cyclopentanecarboxylic acid, (2,6-di-
phenyl-pyrimidin-4-yl)-amide,2 and 8-cyclopentyl-2,6-diphenyl-
9H-purine (31). In both series, these two compounds displayed
the highest affinity for the A1 receptor, yet 31 was, at 0.29 nM,
significantly more active than cyclopentanecarboxylic acid (2,6-
diphenyl-pyrimidin-4-yl)-amide (2.14 nM). The selectivity for
the A1 receptor over the A2A and A3 receptors was also
somewhat better for the purine compound (with a A2A/A1 ratio
of 190 and A3/A1 ratio of 121) than that for the pyrimidine (92
and 79, respectively). It is therefore appropriate to take into
consideration these results and refine the pharmacophoric model.
Figure 5 schematically displays the new refinements. It has
previously been discovered and discussed in both the pyrim-
idines series and in this current article that the optimal L2 group
seems to be somewhat smaller than the L1 and L3 groups, and
the lipophilic pocket that this group fills lacks the ability to
interact with π-electrons. The ideal L2 group is, therefore, an
alkyl group that consists of an alkyl chain of 2 to 3 carbons in
length (i.e., ethyl or propyl) and a secondary branched moiety
(i.e., isopropyl, cyclopentyl, etc). Although this article details
the development of a fused bicyclic ring, the central aromatic
region was sufficiently covered by a single ring, as detailed in

2,6-Disubstituted and 2,6,8-Trisubstituted Purines
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10 2865
Figure 5. Refinement of the model proposed in an earlier publication (ref 2), showing, among others, the smaller size of the L2 group (see text).
modifications are highlighted by compound 31, LUF 5962,
8-cyclopentyl-2,6-diphenyl-9H-purine, which has an affinity of
0.29 nM at the human adenosine A1 receptor.
Experimental Section
Chemistry. Materials and Methods. All reagents used were
obtained from commercial sources and all solvents were of an
1
analytical grade. H and 13C NMR spectra were recorded on a
Bruker AC 200 (1H NMR, 200 MHz; 13C NMR, 50.29 MHz)
spectrometer with tetramethylsilane as an internal standard. Chemi-
cal shifts are reported in δ (ppm), and the following abbreviations
are used: s ) singlet, d ) doublet, dd ) double doublet, t ) triplet,
m ) multiplet, br ) broad, and ar ) aromatic protons. Melting
points were determined on a Bu¨chi melting point apparatus and
are uncorrected. Elemental analyses were performed by the Leiden
Figure 6. Functional assays performed with reference ligands and
Institute of Chemistry and are within 0.4% of the theoretical values
compounds 30, 31, and 32 on the adenosine A1 receptor. 100% cAMP
unless otherwise stated. The reactions were routinely monitored
amounts to approximately 15 pmol/well. 0% corresponds to ap-
by TLC using Merck silica gel F254 plates. Microwave reactions
proximately 1.3 pmol/well.
were performed on an Emrys Optimizer (Biotage AB, formerly
Personal Chemistry). The wattage was automatically adjusted so
as to maintain the desired temperature.
A procedure to protect the purines has been reported in the
literature by Cassidy et al.27 Compounds 227 and 1820 have been
reported previously.
General Preparation for the Suzuki-Miyaura Cross-Cou-
pling Under Microwave Conditions. THP-protected 2,6-dichlo-
ropurine 2 (1 equiv) was dissolved in dry toluene (5 mL). To this
was added the appropriate boronic acid (3 equiv if double
substitution was required or 1 equiv if single substitution was
desired), K2CO3 (1.5 equiv or 1 equiv, respectively), and Pd(PPh3)4
(0.052 equiv). The reaction vial was then sealed and heated at 150
°C for 20 min. Upon completion of the reaction (monitored by
Figure 7. Concentration-dependent antagonism of the effect of CPA
TLC), the solvents were evaporated and the crude product preab-
sorbed on silica. Purification with column chromatography gave
the desired product.
(10-7 M) on forskolin (10-5 M)-induced cAMP production by
compound 32.
2-Chloro-6-phenyl-9-(tetrahydro-pyran-2-yl)-9H-purine (3).
approximately 4 nM for both compounds in this functional
White solid, 71%.
assay. These KB values are somewhat higher but not inconsistent
with the affinity values determined in the radioligand-binding
assays. As a control experiment, we examined the effects of
N-0840, 30, 31, and 32 when given alone (10 µM), that is, in
the absence of both CPA and forskolin. As expected, no
measurable changes from control levels were recorded. Thus,
the purines presented in the present article display antagonistic-
inverse agonistic behavior, typical of most nonribose ligands
for adenosine receptors.
2,6-Diphenyl-9-(tetrahydropyran-2-yl)-9H-purine (4). White
solid, 80%.
2,6-Bis(4-chlorophenyl)-9-(tetrahydropyran-2-yl)-9H-
purine (5). White solid, 91%.
2,6-Bis(4-tolyl)-9-(tetrahydropyran-2-yl)-9H-purine (6). White
solid, 77%.
2,6-Bis(4-methoxyphenyl)-9-(tetrahydropyran-2-yl)-9H-pu-
rine (7). White solid, 82%.
2-(4-Chlorophenyl)-6-phenyl-9-(tetrahydropyran-2-yl)-9H-
purine (8). White solid, 87%.
2-Tolyl-6-phenyl-9-(tetrahydropyran-2-yl)-9H-purine (9). White
solid, 90%.
2-(4-Methoxyphenyl)-6-phenyl-9-(tetrahydropyran-2-yl)-9H-
purine (10). Oil, 97%.
Conclusions
This article describes a series of 2,6-di- and 2,6,8-trisubsti-
tuted purines as antagonists for the human adenosine A1 receptor
synthesized as a consequence of the refinement of the model
proposed in an earlier publication.2 The fixation of the H-bond
acceptor at the top of the molecule repositions the relative
location of the L2 group. Exchanging the heteroatom for
nitrogen creates the imidazole ring. The benefits of these
Deprotection of the purines was performed following a procedure
described by Hocek et al.28
2,6-Diphenyl-9H-purine (11). Recrystallized from EtOAc/PE.
White solid, 53%; mp: >252 °C dec MS (ESI): 273.0. Anal.
(C17H12N4‚0.4EtOAc) C, H, N.

2866 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10
Chang et al.
2,6-Bis(4-chlorophenyl)-9H-purine (12). Recrystallized from
DCM. White solid, 43%; mp: >290 °C dec. MS (ESI): 340.7.
Anal. (C17H10N4Cl2‚0.1H2O.0.05CH2Cl2) C, H, N.
2,6-Bis(4-tolyl)-9H-purine (13). Recrystallized with DCM.
White solid, 52%; mp: >288 °C dec MS (ESI): 300.9. Anal.
(C19H16N4‚0.06CH2Cl2) C, H, N.
2,6-Bis(4-methoxyphenyl)-9H-purine (14). Recrystallized sev-
eral times from various solvents, including CH2Cl2, EtOH, MeOH,
and EtOAc/PE mixtures. White solid; mp: 282 °C. MS (ESI):
332.0.
2-(4-Chlorophenyl)-6-phenyl-9H-purine (15). Recrystallized
from MeOH. White solid, 46%; mp: 262 °C. MS (ESI): 306.8.
Anal. (C17H11ClN4) C, H, N.
2-Tolyl-6-phenyl-9H-purine (16). Recrystallized from MeOH.
White solid, 43%; mp: 251 °C. MS (ESI): 286.8. Anal. (C18H14N4‚
0.16MeOH) C, H, N.
2-(4-Methoxyphenyl)-6-phenyl-9H-purine (17). Recrystallized
from MeOH. White solid, 55%; mp: 269 °C. MS (ESI): 302.8.
Anal. (C18H14N4) C, H, N. Ring closure was performed according
to a procedure described by Scammells et al.23
6-Chloro-2,8-diphenyl-9H-purine (19). White solid, 66%.
2-Phenyl-6-chloro-8-propyl-9H-purine (20). White solid, 51%.
2-Phenyl-6-chloro-8-isopropyl-9H-purine (21). White solid,
78%.
2-Phenyl-6-chloro-8-cyclopentyl-9H-purine (22). White solid,
44%.
2-Phenyl-6-chloro-8-cyclohexyl-9H-purine (23). White solid,
99%.
2-Phenyl-6-chloro-8-propyl-9-(tetrahydropyran-2-yl)-9H-pu-
rine (24). White solid, 67%.
2-Phenyl-6-chloro-8-cyclopentyl-9-(tetrahydropyran-2-yl)-9H-
purine (25). White solid, 75%.
2,6-Diphenyl-8-propyl-9-(tetrahydropyran-2-yl)-9H-purine (26).
White solid, 53%.
2,6-Diphenyl-8-cyclopentyl-9-(tetrahydropyran-2-yl)-9H-pu-
rine (27). White solid, 57%.
2,6,8-Triphenyl-9H-purine (28). Recrystallized with MeOH.
White solid, 87%; mp: 233 °C. MS (ESI): 348.7. Anal. (C23H16N4‚
0.12 MeOH) C, H, N.
8-Propyl-2,6-diphenyl-9H-purine (29). Recrystallized from
MeOH. White solid, 52%; mp: 149 °C. MS (ESI): 314.8. Anal.
(C20H18N4) C, H, N.
8-Isopropyl-2,6-diphenyl-9H-purine (30). Recrystallized from
MeOH. White solid, 36%; mp: 214 °C. MS (ESI): 314.8. Anal.
(C20H18N4‚0.1EtOH) C, H, N.
8-Cyclopentyl-2,6-diphenyl-9H-purine (31). Recrystallized from
DCM. White solid, 45%; mp: 224 °C. MS (ESI): 341.0. Anal.
(C22H20N4‚0.04CH2Cl2) C, H, N.
by Dr. Andrea Townsend-Nicholson, University College. London,
U.K. HEK 293 cells stably expressing the human adenosine A2A
and A3 receptors were gifts from Dr. Wang (Biogen, U.S.A.) and
Dr. K. -N. Klotz (University of Wu¨rzburg, Germany), respectively.
All compounds were tested in radioligand-binding assays to
determine their affinities at the human adenosine A1, A2A, and the
A3 receptors as described previously.2 The human A1 receptors were
expressed in CHO cells, and [3H]DPCPX was used as the
radioligand. The A2A and A3 receptors were expressed in HEK 293
cells, and [3H]ZM 241385 and [125I]AB-MECA were used as the
respective radioligands.
The compounds specified in the text were tested in functional
assays for their ability to influence the levels of cAMP in the test
system. The compounds were tested at concentrations of, at least,
100 × Ki, where the receptor sites should be fully occupied. The
behavior of the compounds was observed with reference to known
adenosine receptor ligands: CPA (a full agonist), DPCPX (an
inverse agonist), and N0840 (reported as a neutral antagonist).
CHO cells expressing the human adenosine A1 receptor were
grown overnight as a monolayer in 24-well tissue culture plates
(400 µL/well; 2 × 105 cells/well). cAMP generation was performed
in a Dulbecco’s Modified Eagles Medium (DMEM)/N-2-hydroxy-
ethylpiperazin-N′-2-ethansulfonic acid (HEPES) buffer (0.60 g
HEPES/50 mL DMEM pH 7.4). Each well was washed twice with
the HEPES/DMEM buffer (250 µL) and the following added:
adenosine deaminase (0.8 IU/mL), rolipram (50 µM), and cilosta-
mide (50 µM). This was then incubated for 30 min at 37 °C
followed by the introduction of the compound of interest. After a
further 10 min of incubation, forskolin was added (10 µM). After
a subsequent 15 min, incubation was stopped by aspirating the assay
medium and by adding 200 µL of ice-cold 0.1 M HCl. The amount
of cAMP was determined by competition with [3H]cAMP for
protein kinase A (PKA). Briefly, the sample, approximately 1.8
nM [3H]cAMP, and 100 µL of PKA solution were incubated on
ice for 2.5 h. The incubations were stopped by rapid dilution with
2 mL of ice-cold Tris HCl buffer (pH 7.4), and the bound
radioactive material was then recovered by filtration through
Whatman GF/C filters. The filters were additionally rinsed with 2
× 2 mL of Tris HCl buffer and then the radioactivity counted in a
Packard Emulsifier Safe scintillation fluid (3.5 mL). All data reflect
at least three independent experiments performed in duplicate.
Data Analysis. Ki values were calculated using a nonlinear
regression curve-fitting program (GraphPad Prism, GraphPad
Software Inc., San Diego, CA). KD values of the radioligands were
1.6 nM, 1.0 nM, and 5.0 nM for [3H]DPCPX, [3H]ZM241385, and
[
125I]AB-MECA, respectively. The data from the functional assays
were also analyzed with GraphPad Prism, and Figures 6 and 7 were
generated by evaluating the data to relate to the known ligand CPA
(set at 0%) and forskolin (set at 100%).
8-Cyclohexyl-2,6-diphenyl-9H-purine (32). Recrystallized from
EtOH. White solid, 49%; mp: 209 °C. MS (ESI): 354.8. Anal.
(C23H22N4‚0.1H2O) C, H, N.
8-Cyclohexyl-6-(4-chlorophenyl)-2-phenyl-9H-purine (33). Re-
crystallized from MeOH. White solid, 86%; mp: 156 °C. MS
(ESI): 388.9, 390.2. Anal. (C23H21ClN4‚0.7H2O‚0.5MeOH) C,
H, N.
Supporting Information Available: Elemental analyses and
NMR data. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
8-Cyclohexyl-6-(3,4-dichlorophenyl)-2-phenyl-9H-purine (34).
Recrystallized from EtOH. White solid, 64%; mp: 201 °C. MS
(ESI): 422.7, 425.2. Anal. (C23H20Cl2N4‚0.7EtOH) C, H, N.
8-Cyclohexyl-6-(4-tolyl)-2-phenyl-9H-purine (35). Recrystal-
lized from MeOH. White solid, 98%; mp: 149 °C. MS (ESI):
368.5. Anal. (C24H24N4‚0.9H2O‚0.35MeOH) C, H, N.
8-Cyclohexyl-6-(4-methoxyphenyl)-2-phenyl-9H-purine (36).
Recrystallized from MeOH. White solid, 29%; mp: 141 °C. MS
(ESI): 384.9. Anal. (C24H24N4O‚0.95H2O‚0.3MeOH) C, H, N.
8-Cyclohexyl-9-methyl-2,6-diphenyl-9H-purine (37). Recrys-
tallized from CHCl3. White solid, 30%; mp: 200-202 °C. MS
(ESI): 369.1. Anal. (C24H24N4‚0.09CHCl3) C, H, N.
Biology. Materials and Methods. [3H]DPCPX and [125I]AB-
MECA were purchased from Amersham Biosciences (NL). [3H]-
ZM 241385 was obtained from Tocris Cookson, Ltd. (U.K.). CHO
cells expressing the human adenosine A1 receptor were provided
(1) Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz, K.-N.;
Linden, J. M. International union of pharmacology. XXV. Nomen-
clature and classification of adenosine receptors.Pharmacol. ReV.
2001, 53, 527-552.
(2) Chang, L. C. W.; Spanjersberg, R. F.; von Frijtag Drabbe Ku¨nzel, J.
K.; Mulder-Krieger, T.; van den Hout, G.; Beukers, M. W.; Brussee,
J.; IJzerman, A. P. 2,4,6-Trisubstituted pyrimidines as a new class
of selective adenosine A1 receptor antagonists. J. Med. Chem. 2004,
47, 6529-6540.
(3) Chang, L. C. W.; Brussee, J.; IJzerman, A. P. Non-xanthine
antagonists for the adenosine A1 receptor. Chem. BiodiVersity 2004,
1, 1591-1626.
(4) Ukena, D.; Padgett, W. L.; Hong, O.; Daly, J. W.; Daly, D. T.; Olsson,
R. A. N6-substituted 9-methyladenines: a new class of adenosine
receptor antagonists. FEBS Lett. 1987, 215, 203-208.
(5) Thompson, R. D.; Secunda, S.; Daly, J. W.; Olsson, R. A. N,69-
disubstituted adenines: potent, selective antagonists at the A1
adenosine receptor. J. Med. Chem. 1991, 34, 2877-2882.

2,6-Disubstituted and 2,6,8-Trisubstituted Purines
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 10 2867
(6) Van Calenbergh, S.; von Frijtag Drabbe Ku¨nzel, J. K.; Blaton, N.
M.; Peeters, O. M.; Rozenski, J.; van Aerschot, A.; de Bruyn, A.; de
Keukeleire, D.; IJzerman, A. P.; Herdewijn, P. N6-Cyclopentyl 3′-
substituted xylofuranosyl adenosines: a new class of nonxanthine
adenosine A1 receptor antagonists. J. Med. Chem. 1997, 40, 3765-
3772.
(18) Biagi, G.; Giorgi, I.; Livi, O.; Scartoni, V. Synthesis of 4,6-
disubstituted- and 4,5,6-trisubstituted-2-phenyl-pyrimidines and
their affinity towards A1 adenosine receptors. Farmaco 1997, 52,
61-65.
(19) Biagi, G.; Giorgi, I.; Livi, O.; Nardi, A.; Pacchini, F.; Scartoni, V.;
Lucacchini, A. N6-cycloalkyl-2-phenyl-3-deaza-8-azaadenines: a new
class of A1 adenosine receptor ligands. A comparison with the
corresponding adenines and 8-azaadenines. Eur. J. Med. Chem. 2003,
38, 983-990.
(20) Biagi, G.; Giorgi, I.; Livi, O.; Pacchini, F.; Rum, P.; Scartoni, V.;
Costa, B.; Mazzoni, M. R.; Giusti, L. Erythro- and threo-2-
hydroxynonyl substituted 2-phenyladenines and 2-phenyl-8-azaad-
enines: ligands for A1 adenosine receptors and adenosine deaminase.
Farmaco 2002, 57, 221-233.
(21) Antonini, I.; Cristalli, G.; Franchetti, P.; Grifantini, M.; Martelli, S.;
Petrelli, F. Deaza analogues of adenosine as inhibitors of blood
platelet aggregation. J. Pharm. Sci. 1984, 73, 366-369.
(22) Young, R. C.; Jones, M.; Milliner, K. J.; Rana, K. K.; Ward, J. G.
Purine derivatives as competitive inhibitors of human erythrocyte
membrane phosphatidylinositol 4-kinase. J. Med. Chem. 1990, 33,
2073-2080.
(23) Scammells, P. J.; Baker, S. P.; Belardinelli, L.; Olsson, R. A.
Substituted 1,3-dipropylxanthines as irreversible antagonists of A1
adenosine receptors. J. Med. Chem. 1994, 37, 2704-2712.
(24) Van Muijlwijk-Koezen, J. E.; Timmerman, H.; Link, R.; van der Goot,
H.; IJzerman, A. P. A novel class of adenosine A3 receptor ligands.
II. Structure affinity profile of a series of isoquinoline and quinazoline
compounds. J. Med. Chem. 1998, 41, 3994-4000.
(25) Van Muijlwijk-Koezen, J. E.; Timmerman, H.; van der Goot, H.;
Menge, W. M. P. B.; Frijtag von Drabbe Ku¨nzel, J.; Groote, M.;
IJzerman, A. P. Isoquinoline and quinazoline urea analogues as
antagonists for the human adenosine A3 receptor. J. Med. Chem. 2000,
43, 2227-2238.
(26) Shryock, J. C.; Ozeck, M. J.; Belardinelli, L. Inverse agonists and
neutral antagonists of recombinant human A1 adenosine receptors
stably expressed in Chinese hamster ovary cells. Mol. Pharmacol.
1998, 53, 886-893.
(27) Cassidy, F.; Olsen, R. K.; Robins, R. K. Aromaticity in heterocyclic
systems. V. Bromination studies of certain purines, pyrrolo[3,2-d]-
pyrimidines, and pyrazolo[4,3-d]pyrimidines. J. Heterocycl. Chem.
1968, 5, 461-465.
(7) Bianucci, A. M.; Biagi, G.; Coi, A.; Giorgi, I.; Oreste, L.; Pacchini,
F.; Scartoni, V.; Lucacchini, A.; Costa, B. Bioisosterism, enantiose-
lectivity, and molecular modeling of new effective N6- and/or N(9)-
substituted 2-phenyl adenines and 8-aza analogues: Different binding
modes to A1 adenosine receptors. Drug DeV. Res. 2001, 54, 52-65.
(8) Klotz, K.-N.; Kachler, S.; Lambertucci, C.; Vittori, S.; Volpini, R.;
Cristalli, G. 9-Ethyladenine derivatives as adenosine receptor an-
tagonists: 2- and 8-substitution results in distinct selectivities.
Naunyn-Schmiedeberg’s Arch. Pharmacol. 2003, 367, 629-634.
(9) De Ligt, R. A. F.; van der Klein, P. A. M.; von Frijtag Drabbe Ku¨nzel,
J. K.; Lorenzen, A.; Ait el Maate, F.; Fujikawa, S.; van Westhoven,
R.; van den Hoven, T.; Brussee, J.; IJzerman., A. P. Synthesis and
biological evaluation of disubstituted N6-cyclopentyladenine ana-
logues: the search for a neutral antagonist with high affinity for the
adenosine A1 receptor. Bioorg. Med. Chem. 2004, 12, 139-149.
(10) Gundersen, L. L. 6-Chloropurines and organostannanes in palladium
catalyzed cross coupling reactions. Tetrahedron Lett. 1994, 35, 3155-
3158.
(11) Gundersen, L. L.; Langli, G.; Rise, F. Regioselective Pd-mediated
coupling between 2,6-dichloropurines and organometallic reagents.
Tetrahedron Lett. 1995, 36, 1945-1948.
(12) Gundersen, L. L.; Nissen-Meyer, J.; Spilsberg, B. Synthesis and
antimycobacterial activity of 6-arylpurines: the requirements for the
N-9 substituent in active antimycobacterial purines. J. Med. Chem.
2002, 45, 1383-1386.
(13) Hocek, M.; Hockova, D.; Dvorakova, H. Dichotomy in regioselective
cross-coupling reactions of 6,8-dichloropurines with phenylboronic
acid and methylmagnesium chloride: Synthesis of 6,8-disubstituted
purines. Synthesis 2004, 889-894.
(14) Havelkova, M.; Dvorak, D.; Hocek, M. The Suzuki-Miyaura cross-
coupling reactions of 2-, 6- or 8-halopurines with boronic acids
leading to 2-, 6- or 8-aryl- and -alkenylpurine derivatives. Synthesis
2001, 1704-1710.
(15) Hocek, M. Syntheses of purines bearing carbon substituents in
positions 2, 6 or 8 by metal- or organometal-mediated C-C bond-
forming reactions. Eur. J. Org. Chem. 2003, 2, 245-254.
(16) Hocek, M.; Holy, A.; Votruba, I.; Dvorakova, H. Synthesis and
cytostatic activity of substituted 6-phenylpurine bases and nucleo-
sides: application of the Suzuki-Miyaura cross-coupling reactions
of 6-chloropurine derivatives with phenylboronic acids. J. Med. Chem.
2000, 43, 1817-1825.
(28) Hocek, M.; Dvorakova, H. An efficient synthesis of 2-substituted
6-methylpurine bases and nucleosides by Fe- or Pd-catalyzed cross-
coupling reactions of 2,6-dichloropurines. J. Org. Chem. 2003, 68,
5773-5776.
(17) Langli, G.; Gundersen, L. L.; Rise, F. Regiochemistry in Stille
couplings of 2,6-dihalopurines. Tetrahedron 1996, 52, 5625-5638.
JM050640I