Orthogonal activation of the reengineered A3 adenosine receptor (neoceptor) using tailored nucleoside agonists

Article abstract of DOI:10.1021/jm050968b

An alternative approach to overcome the inherent lack of specificity of conventional agonist therapy can be the reengineering of the GPCRs and their agonists. A reengineered receptor (neoceptor) could be selectively activated by a modified agonist, but no

Full text of DOI:10.1021/jm050968b

J. Med. Chem. 2006, 49, 2689-2702
2689
Orthogonal Activation of the Reengineered A3 Adenosine Receptor (Neoceptor) Using Tailored
Nucleoside Agonists
Zhan-Guo Gao,† Heng T. Duong,† Tatiana Sonina,‡ Soo-Kyung Kim,† Philippe Van Rompaey,§ Serge Van Calenbergh,§
Liaman Mamedova,† Hea Ok Kim,| Myong Jung Kim,| Ae Yil Kim,| Bruce T. Liang,‡ Lak Shin Jeong,*,| and
Kenneth A. Jacobson*,†
Molecular Recognition Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and DigestiVe and Kidney Diseases,
National Institutes of Health, Bethesda, Maryland 20892, Department of Cardiology, UniVersity of Connecticut Health Center,
Farmington, Connecticut 06030-1601, Laboratory for Medicinal Chemistry, Faculty of Pharmaceutical Sciences (FFW), Ghent UniVersity,
Harelbekestraat 72, B-9000 Ghent, Belgium, and Laboratory of Medicinal Chemistry, College of Pharmacy, Ewha Womans UniVersity,
Seoul 120-750, Korea
ReceiVed September 28, 2005
An alternative approach to overcome the inherent lack of specificity of conventional agonist therapy can be
the reengineering of the GPCRs and their agonists. A reengineered receptor (neoceptor) could be selectively
activated by a modified agonist, but not by the endogenous agonist. Assisted by rhodopsin-based molecular
modeling, we pinpointed mutations of the A3 adenosine receptor (AR) for selective affinity enhancement
following complementary modifications of adenosine. Ribose modifications examined included, at 3′: amino,
aminomethyl, azido, guanidino, ureido; and at 5′: uronamido, azidodeoxy. N6-Variations included
3-iodobenzyl, 5-chloro-2-methyloxybenzyl, and methyl. An N6-3-iodobenzyl-3′-ureido adenosine derivative
10 activated phospholipase C in COS-7 cells (EC50 ) 0.18 µM) or phospholipase D in chick primary
cardiomyocytes, both mediated by a mutant (H272E), but not the wild-type, A3AR. The affinity enhancements
for 10 and the corresponding 3′-acetamidomethyl analogue 6 were >100-fold and >20-fold, respectively.
10 concentration-dependently protected cardiomyocytes transfected with the neoceptor against hypoxia. Unlike
10, adenosine activated the wild-type A3AR (EC50 of 1.0 µM), but had no effect on the H272E mutant
A3AR (100 µM). Compound 10 was inactive at human A1, A2A, and A2BARs. The orthogonal pair comprising
an engineered receptor and a modified agonist should be useful for elucidating signaling pathways and
could be therapeutically applied to diseases following organ-targeted delivery of the neoceptor gene.
Introduction
ligand binding domains with specific functional groups on
ligands was proposed to provide a novel avenue for design of
therapeutic agents.12 Indeed, the Gs-coupled â-adrenergic recep-
tor-ligand pairs were further engineered and proposed for gene
therapy, which should be therapeutically useful as they could
potentially rescue the loss of function caused by SNPs (single
nucleotide polymorphisms) which happened naturally at this
receptor.13 In the meantime, Coward et al.14 by using a chimeric
approach, engineered Gi-coupled κ-opioid receptors, and des-
ignated them as RASSLs (receptors solely activated by synthetic
ligand). A RASSL showed dramatically decreased affinity for
the endogenous agonist dynorphin and a wide range of other
opioid peptides, but with a lesser extent of affinity decrease for
the synthetic agonist spiradoline, and slowed heart rate in
transgenic mice.14,15 The RASSL approach is useful for unravel-
ing the GPCR signaling mechanisms.14-19 However, an ideal
ligand-receptor pair for this purpose should be truly orthogonal,
i.e., the endogenous agonist should not activate the engineered
receptor and the synthetic ligand should not activate the wild-
type receptor, and this goal has not actually been achieved. More
recently, Bruysters et al.20 engineered the Gq-coupled histamine
H1 receptor for enhanced affinity of a synthetic agonist, with
decreased affinity for the endogenous agonist, but orthogonality
was not complete.
The approach of chemical genetics has been widely used to
control various protein functions and cellular processes.1 For
example, this approach has been used in studying GTP regula-
tory proteins to create nucleotide specificity.2 Gene-directed
enzyme prodrug therapy has been proposed to be an effective
means against cancer.3 Tailor-made ligands for thyroid hormone
nuclear receptors may have potential to restore mutation-caused
genetic diseases.4 It has been suggested that the chemical genetic
approach is superior to the traditional genetic approach, such
as gene knockout, in studying protein functions,5 and has
advantage for developing drugs against genetic diseases.6
Functional orthogonal ligand-receptor pairs have been used for
regulation of estrogen receptors,7,8 Src family protein kinases,9
and protein methyltransferases.10
In the GPCR field, engineering of orthogonal ligand-receptor
pairs was initiated shortly after the cloning of the first non-
rhodopsin GPCR.11,12 The ligand binding domain of the â-adr-
energic receptor has been genetically engineered to respond to
specific classes of compounds which do not interact with the
wild-type (WT) receptor. Targeting critical positions in the
* Corresponding authors: Dr. K. A. Jacobson, Chief, Molecular Rec-
ognition Section, Bldg. 8A, Rm. B1A-19, NIH, Bethesda, MD 20892-0810,
Tel.: (301) 496-9024; Fax: (301) 480-8422; e-mail: kajacobs@helix.nih.gov
and Lak Shin Jeong, Ph.D., Laboratory of Medicinal Chemistry, College
of Pharmacy, Ewha Womans University, Seoul 120-750, Korea; Tel.: 82-
2-3277-3466; Fax: 82-2-3277-2851; e-mail: lakjeong@ewha.ac.kr.
† National Institutes of Health.
Our initial work on engineering the receptors and ligands for
orthogonal activation started with the A3 adenosine receptor
(AR),21 the activation of which is both cerebro- and cardiopro-
tective22,23 and has an anticancer effect.24 An integrated approach
of engineering both ligands and receptors combined with
molecular modeling was used. A H272E mutant receptor
‡ University of Connecticut Health Center.
§ Ghent University.
| Ewha Womans University.
10.1021/jm050968b CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/11/2006

2690 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
Table 1. Affinities of Various 3′-Amine-Derivatized Adenosine Analogues in Binding Experiments at Wild-Type and H272E Mutant Human A3ARs.
X ) O and R2 ) H, Unless Noteda
adenine substituente
R6
ribose substituents
Ki (µM) or percentage inhibition at 10 µM
affinity
enhancement
R3
R5
wild-type
H272E
1b
2c
H
H
H
IB
IB
IB
IB
CMB
CMB
IB
IB, R2 ) Cl
IB
Me
Me
NH2
CH2NH2
NHCdNHNH2
NH2
CH2OH
CH2OH
CH2OH
CH2OH
CONHCH3
CH2OH
CH2OH
CH2OH
CONHCH3
CH2OH
CH2OH
CONHCH3
CH2OH
442 ( 121
3 ( 2%
130 ( 34
0.87 ( 0.18
0.137 ( 0.041
15%
9.5 ( 1.7
13.8 ( 1.4
0.557 ( 0.164
9 ( 1%
47 ( 9%
5 ( 1%
3 ( 1%
7 ( 2%
75 ( 32
2 ( 2%
6-fold
NE
4-fold
3-fold
NE
>20-fold
15-fold
20-fold
NE
>100-fold
∼50-fold
>10-fold
NE
3b
4b
5
33.3 ( 7.1
0.32 ( 0.10
0.185 ( 0.027
0.700 ( 0.099
0.61 ( 0.23
0.709 ( 0.163
0.592 ( 0.081
0.22 ( 0.04
0.20 ( 0.03
2.7 ( 0.4
CH2NH2
CH2NHCOCH3
CH2NH2
CH2NH2
CH2NH2
NHCONH2
NHCONH2
NHCONH2
NHCONH2
NHCONH2
N3
6
7
8
9
10d
11
12c
13
14c
15c
16
17
18f
30 ( 6%
7 ( 1%
CONHCH3
CONHCH3
CH2N3
CONH-(CH2)2NH2
CONHNH2
NE
12-fold
NE
NE
NE
IB, X ) O
2.26 ( 0.48
0.022 ( 0.004
>100
0.186 ( 0.034
0.23 ( 0.04
>100
IB, X ) S
OH
OH
OH
H
H
1.5 ( 0.6
>10
a Binding parameters were measured in transiently transfected COS-7 cells as described in Experimental Procedures. The binding affinity was determined
by using the agonist radioligand [125I]I-AB-MECA (0.5 nM). Values represent the mean ( SE of at least three independent determinations. b Affinity at
wild-type and mutant A3ARs previously reported.21 c Affinity at wild-type A3AR stably expressed in CHO cells previously reported.37,38 d MRS3481,
LJ720. e IB ) 3-iodobenzyl, CMB ) 5-chloro-2-methyloxybenzyl, Me ) methyl, NE ) no enhancement or insignificant enhancement. f MRS3412.
(designated as neoceptor) was found to have decreased affinity
for classical ligands (20-50-fold), such as NECA (5′-N-
ethylcarboxamidoadenosine) and Cl-IB-MECA (2-chloro-N6-
(3-iodobenzyl)-5′-N-methylcarbamoyladenosine), and with a
modest affinity increase for a 3′-amine-derivatized adenosine
(7-fold), but orthogonality has not been complete.21 Further
application of the neoceptor approach to 5′-uronamide deriva-
tives that bound orthogonally at strategically mutated A2AARs
suggested that, in addition to the charge, the chain length of
the interacting moiety (e.g., amine) was also critical for the
affinity enhancement.25,26 This observation was instructive for
additional optimization of A3 neoceptor-neoligand interactions.
ligand with those on the receptor at a specific site, could be
explained by rhodopsin-based molecular modeling.
Results
Mutagenesis. Single amino acid replacements of the hA3-
AR were made at strategic locations of the putative ribose
binding region, as defined by rhodopsin-based molecular
modeling.31,35 Thr94 (3.36) and His272 (7.43) were selected for
mutation and individually replaced with Ala, Asp, or Glu. Both
of these hydrophilic residues are predicted to be involved in
coordination of the ribose moiety. The Gln167 residue of EL2,
predicted to be in proximity to A3AR-bound nucleosides,36 also
served as a mutation site. His272, predicted to be in proximity
to the 3′-OH group, was used successfully in the first study of
neoceptors.21
The neoceptor approach could be particularly suitable for
application to ARs to achieve the receptor subtype-selectivity
and tissue selectivity, which has proved difficult to achieve due
to the ubiquitous presence of ARs. Here, with the assistance of
molecular modeling, we systematically studied the interactions
between the 3′-position of the nucleoside with negatively
charged residues in place of the His residue. The charge, chain
length, and H-bonding properties of the 3′-substitution were
optimized in combination with substitutions at other positions
of the adenosine molecule. We found that a derivative (10, N6-
(3-iodobenzyl)-3′-ureidoadenosine) bearing both 3′-ureido and
N6-3-iodobenzyl (IB) modifications displayed high selectivity
for the H272E mutant receptor expressed in COS-7 cells in
comparison to WT in binding and phospholipase C (PLC)
assays. In contrast, the endogenous agonist adenosine (100 µM)
had no effect at the H272E mutant receptor. The orthogonality
of the neoceptor-neoagonist pair was functionally tested in a
cardiomyocyte model showing activation of the cardioprotective
phospholipase D (PLD) pathway and an antiischemic effect.27,28
Additionally, it was found that, neither T94E nor Q167E,
produced a similar pattern of affinity enhancement for a series
of 3′-amine derivatized adenosine analogues. This selective gain
of function, achieved by matching attractive substituents on the
Nucleoside Structure. A variety of adenine nucleoside
derivatives (1-18, Table 1) were examined in this study as
potential neoligands with respect to the above receptor muta-
tions. Most of the modified adenosine derivatives were reported
previously. Compounds 1, 3, 4, 7, 17, and 18 were prepared as
described.21,26 The 3′-amino derivative 1 was identified previ-
ously as a suitable neoligand when paired with the H272E A3-
AR neoceptor, although the degree of enhancement was low
(Table 1). The 3′-aminomethyl derivatives 2 and 5-9 and 3′-
azide 15 were reported recently.37 The 3′-ureido moiety, found
to result in inactivity the wild-type ARs, was included in
analogues 10-14 and 16.38,39 Compounds 17 and 18 were
identified previously as suitable neoligands when paired with
A
2AAR neoceptors.25,26 The variations at the N6-position
included unsubstituted, IB, 5-chloro-2-methoxybenzyl (CMB),
and methyl. We previously reported that the CMB group
increased affinity and selectivity at the human (h) A3AR in
comparison to IB.36 The 5′-position mainly consisted of either
the ribose-like CH2OH or the NECA-like 5′-uronamide. To
enhance selectivity for the hA3AR, the optimally substituted

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2691
Scheme 1a
a Reagents and conditions: (a) Tf2O, pyridine, 0 °C, 1 h; (b) NaN3, DMF, rt, 48 h; (c) (i) 75% AcOH, 55 °C, 1.5 h; ii) NaIO4/H2O, EtOH, 0 °C, 20 min
then NaBH4; (d) Ac2O, Pyridine, rt, 3 h; (e) (i) 85% HCO2H, 60 °C, 1.5 h; (ii) Ac2O, pyridine, rt, 16 h; (f) silylated 6-chloropurine or 2,6-dichloropurine,
TMSOTf, C2H4Cl2, 0 °C to 60 °C, 2 h; (g) MeNH2, 1,4-dioxane, rt, 4 h or 3-iodobenzylamine hydrochloride, Et3N, EtOH, 50 °C, 18 h then NaOMe, MeOH,
rt, 2 h; (h) TBSCl, imidazole, DMF, rt, 24 h; (i) Ph3P, NH4OH/H2O, THF, rt, 18 h; (j) chloroacetyl isocyanate, DMF, 0 °C, 3 h; (k) NaOMe, MeOH, rt, 18
h; (l) TBAF, THF, rt, 4 h. IB ) 3-iodobenzyl.
5′-methyluronamide moiety was included. The 3′-position
appeared to be the most critical for achieving selective affinity
enhancement at neoceptors derived from the hA3AR.21 Thus, a
variety of polar substituents (amino, guanidine, aminomethyl,
acetamidomethyl, and ureido) were compared at this position.
Only one compound, the 3′-ureido derivative 11, was also
substituted at the adenine 2-position.
urea derivatives with sodium methoxide, giving an N-acetyl-
or N-benzoylureido derivative as a sole product. This migration
was prevented by using a TBS group as a protecting group.
Removal of the TBS group of 30a-c with TBAF in THF
afforded the final N6-substituted-3′-ureidoadenosine derivatives
13, 10, and 11, respectively.
N6-Substituted-3′-ureidoadenosine-5′-methyluronamides 14
and 12 were also synthesized starting from the same intermediate
21, as illustrated in Scheme 2.38 Selective hydrolysis of the 5,6-
isopropylidene group of 21 using 75% aqueous acetic acid
followed by treatment of the resulting diol with NaIO4/RuCl3
in CCl4/CH3CN/H2O (2/2/3) produced acid derivative 32.
Treatment of 32 with oxalyl chloride gave the activated ester,
which was converted to methyl amide 33 by treating with
methylamine. Compound 33 was converted to the key inter-
mediate 34 using the same conditions described in Scheme 1.
The key intermediate 34 was transformed to the final nucleosides
12 and 14 according to the same procedure used in the
preparation of 10, 11, and 13 in Scheme 1.
Ligand Binding Properties at the Wild-Type and Mutant
Receptors. The affinities of the nucleoside derivatives were
initially evaluated at wild-type and H272E mutant A3ARs
expressed in COS-7 cells (Table 1). Only the substituted N6-
benzyl derivatives 4, 5, 9, and 16 displayed a Ki value at the
wild-type A3AR of <1 µM.26,38 All of the other compounds
were weak or inactive at all four subtypes of ARs.37,38
Adenosine bound to the wild-type A3AR with a Ki value of
1 µM, while at the mutant H272E A3AR adenosine at 100 µM
failed to inhibit radioligand binding (Figure 1). Considerable
enhancement of affinity was observed for several nucleoside
analogues. Following the introduction of a carboxylic acid side
chain at position 7.43, the 3′-acetamidomethyl derivative 6 and
the 3′-aminomethyl derivatives 7 and 8 were 10- to 20-fold
enhanced in affinity. The enhancement of the 3′-aminomethyl
derivatives was dependent on the presence of an N6-benzyl-
type group (either IB in 7 or CMB in 8). The corresponding
simple 6-NH2 derivative 2 was not enhanced in affinity. The
greatest degree of enhancement (>100-fold) was observed for
Chemical Synthesis. N6-Substituted-3′-ureidoadenosine de-
rivatives 10, 11, and 13 were synthesized starting from 1,2:
5,6-di-O-isopropylidene-D-glucose (19), as shown in Scheme
1. Reaction of 19 with triflic anhydride followed by treating
the resulting triflate 20 with sodium azide afforded the 3-azido
derivative 21. Selective hydrolysis of the 5,6-isopropylidene
group of 21 followed by oxidative cleavage and reduction gave
22 in 85% yield. Treatment of 22 with acetic anhydride gave
23 which was converted to the glycosyl donor 24 in two steps
(hydrolysis and acetylation). Condensation of 24 with silylated
6-chloropurine and 2,6-dichloropurine in the presence of TM-
SOTf afforded the protected nucleosides 25a and 25b, respec-
tively. 6-Chloropurine derivative 25a was treated with methyl-
amine and 3-iodobenzylamine followed by treating with sodium
methoxide to give the N6-methyl derivative 26a and N6-3-
iodobenzyl derivative 26b, respectively. 2,6-Dichloropurine
derivative 25b was also converted to 2-chloro-N6-3-iodobenzyl
derivative 26c. Compounds 26a-c were reprotected as TBS-
ethers 27a-c, respectively. The reason that the acetyl protecting
groups were replaced with TBS protecting groups was because
of the facile migration of a 2′-acetyl group to the 3′-ureido
group. Reduction of the azido group of 27a-c using triphen-
ylphosphine and ammonium hydroxide in aqueous solution
yielded amino derivatives, 28a-c, respectively. For the conver-
sion of 3′-amino group into a 3′-ureido group, the 3′-amino
derivatives 28a-c were treated with chloroacetyl isocyanate in
DMF to yield 3′-chloroacetyl urea derivatives 29a-c, which
were smoothly converted to the 3′-ureido derivatives 30a-c,
respectively, after treating with sodium methoxide. As men-
tioned above, migration of a 2′-acetyl or benzoyl group to the
3′-ureido group was observed upon treating the 3′-chloroacetyl

2692 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
Scheme 2a
a Reagents and conditions: (a) (i) 75% AcOH, 55 °C, 1.5 h; (ii)NaIO4, RuCl3.H2O, CCl4/CH3CN/H2O, rt, 4 h; (b) (i) (COCl)2, DMF, CH2Cl2, rt, 16 h;
(ii) 2 M NH2CH3, CH2Cl2, 0 °C, 3 h; (c) AcOH/Ac2O/concd H2SO4, rt, 16 h; (d) silylated 6-chloropurine, TMSOTf, C2H4Cl2, 0 °C to 60 °C, 2 h; (e) MeNH2,
1,4-dioxane, rt, 4 h or 3-iodobenzylamine hydrochloride, Et3N, EtOH, 50 °C, 18 h then NaOMe, MeOH, rt, 2 h; (f) TBSCl, imidazole, DMF, rt, 24 h; (g)
Ph3P, NH4OH/H2O, THF, rt, 18 h; (h) chloroacetyl isocyanate, DMF, 0 °C, 3 h; (i) NaOMe, MeOH, rt, 18 h; (j) TBAF, THF, rt, 4 h or Et3N‚3HF, THF,
50 °C, 16 h. IB ) 3-iodobenzyl.
Figure 1. Binding inhibition and functional activation in wild-type and H272E mutant A3ARs. The receptors were expressed transiently in COS-7
cells as described in Experimental Procedures. The binding affinity (Ki) was determined by using the agonist radioligand [125I]I-AB-MECA (0.5
nM). The structure of compound 10 is given in Table 1.
the 3′-ureido derivative 10 (Figure 1A), which also contained
the IB group. The corresponding 2-Cl analogue 11 showed an
approximately 50-fold affinity increase. Compounds 15 and 16
both had azido groups on the ribose moiety, while only the 3′-
azido derivative 15, not 5′-azido derivative 16, displayed
enhanced affinity at the H272E mutant A3AR, further supporting

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2693
Table 2. Comparison of the Effects of 3′-Aminomethyl- and 3′-Ureido-derivatized Adenosine Analogues in Binding Experiments at Wild-Type and
Mutant Human A3ARsa
Ki values (µM)b
compound
wild-type
T94A
T94E
H272D
H272E
Q167E
5
6
7
8
9
10
15
0.14 ( 0.04
>10
0.12 ( 0.01
>10
10.7 ( 2.9
9.8 ( 4.2
ND
>10
1.37 ( 0.13
0.11 ( 0.05
>10
0.12 ( 0.03
1.0 ( 0.3
0.18 ( 0.03
0.70 ( 0.10
0.61 ( 0.23
0.71 ( 0.16
0.59 ( 0.08
0.22 ( 0.04
0.19 ( 0.03
0.57
>10
>10
>10
7.2
9.5 ( 1.7
13.8 ( 1.4
0.56 ( 0.16
>10
10.0 ( 1.4
>10
0.85 ( 0.12
2.0 ( 0.4
0.85 ( 0.19
>10
2.0 ( 0.1
0.39 ( 0.20
0.47 ( 0.07
0.14 ( 0.00
>100
2.26 ( 0.48
NDb
a Binding parameters (Ki values) were measured in transiently transfected COS-7 cells as described in Experimental Procedures by using the agonist
radioligand [125I]I-AB-MECA (0.5 nM). The Kd values for the WT, T94E, and T94A mutant were determined to be 1.8 ( 0.8, 2.0 ( 0.1, and 2.6 ( 1.2 nM,
respectively, while over 10 nM for other mutants. Values represent the mean ( SE of at least three independent determinations. b The H272A mutant
receptor lost high-affinity binding to both agonist and antagonist radioligands; therefore, competitive binding experiments could not be performed. c ND )
not determined.
a specific interaction between the 3′-position and His272E. Two
other 5′-substituted derivatives, 17 and 18, also did not show
any enhancement at the H272E neoceptor.
mutant receptor with an EC50 of 0.18 ( 0.098 µM (n ) 3),
while it had no effect on the WT ARs at concentrations as high
as 10 µM. In contrast, the known agonist NECA (Figure 1E)
activated the WT A3AR with an EC50 of 0.078 ( 0.020 µM (n
) 3), and H272E with an EC50 of 10.4(3.3 µM (n ) 3). Similar
to the loss of affinity for NECA, the endogenous agonist,
adenosine, activated the WT ARs with an EC50 of 1.04 µM
while it had no effect on the H272E mutant receptor at 100
µM (Figure 1F). No enhancement in basal PLC activity was
observed for the mutant receptor in the absence of nucleoside.
A 3′-azido substituted derivative 15, and a 5′-azido derivative
16, failed to activate PLC in either the wild-type A3AR or the
H272E neoceptor (data not shown). The functional effects of
other derivatives at the WT A3AR expressed in CHO cells were
probed previously.37
Cardiomyocyte Model. In a known model of cardioprotec-
tion from 90 min of simulated ischemia,27,28,33,40 endogenous
adenosine significantly protected cardiomyocytes transfected
with the hA3AR (10 ( 7% cells killed in hA3AR-transfected
cells, n ) 9, compared with 28 ( 6.5%, n ) 8, in vector-
transfected overexpressing myocytes; one-way ANOVA, F )
14.7, P < 0.0001, followed by posttest comparison, P < 0.001).
Unlike the wild-type A3AR, cells expressing the neoceptor
H272E cDNA only showed a slight change. In cells expressing
the H272E mutant receptor, the neoligand 10 induced potent
cardioprotection against hypoxia and activated the known
cardioprotective PLD pathway (Figure 2). Adenosine (100 µM)
did not activate the PLD pathway in myocytes expressing the
H272E mutant receptor.
The affinity enhancement at the neoceptors, for nucleosides
substituted at the 3′-position with amino, aminomethyl, and
ureido moieties, was greatly attenuated when a 5′-uronamide
group was also present, i.e., 5, 9, and 12. Although the affinity
at the wild-type A3AR was increased by this structural changes
for example, by 22-fold in the case of 5′-methyluronamide 9s
there was no affinity enhancement at the H272E mutant receptor,
as was observed for the corresponding 8. The affinity of 5′-
methyluronamide 12 was decreased in comparison to the cor-
responding 5′-OH derivative 10 at the H272E mutant A3AR.
Table 2 shows a comparison of affinities of selected nucleo-
sides at various mutant A3ARs. Negatively charged side chains
were introduced at positions 3.36 and 7.43. Thr94 (3.36) is
analogous to Thr88 of the hA2AAR, which was the site of
mutagenesis used successfully to construct neoceptors of that
subtype.26 The Gln167 residue of EL2, predicted to be in
proximity to A3AR-bound nucleosides,36 also served as a
mutation site. However, no enhancement of affinity of any
modified nucleoside in the present study was observed at T94A,
T94E, and Q167E mutant receptors. The affinity at the H272D
mutant receptor closely paralleled the affinity at the H272E
mutant receptor. This indicates that there is sufficient steric
freedom within the binding site to preserve the effect after
adding or subtracting one methylene unit. The T94A and T94E
mutant receptors were similar to wild-type.
Binding experiments were carried out at human A1 and A2A
-
ARs expressed in CHO (Chinese hamster ovary) cells using
standard radioligand binding assays, as described.37 Compounds
1-3, 5-9, and 15 were previously determined to have Ki values
>1 µM at the A1AR and >10 µM at the A2AAR.37 Compound
4 was previously determined to have Ki values of 8.1 µM and
28 µM at the rat A1 and A2AARs, respectively.21 3′-Ureido
derivatives 10-14 were assayed in binding to human A1 and
A2AARs. In all cases, <10% inhibition of binding was observed
at 10 µM, except for 11 at the A1AR (34% inhibition). Thus,
all of the nucleosides that displayed affinity enhancement at
the H272E mutant receptor only weakly bound, if at all, to two
other AR subtypes. Compound 10 at 10 µM was inactive in
the stimulation of cyclic AMP formation mediated by the human
A2BAR expressed in CHO cells.36
Functional Effects of Nucleosides at WT and Mutant
A3ARs. An assay of PLC was used in the determination of
functional coupling of the WT and mutant A3ARs expressed in
COS-7 cells in response to known AR agonists and modified
nucleosides. As illustrated in Figure 1B, 10 induced accumula-
tion of inositol phosphates in COS-7 cells expressing the H272E
Molecular Modeling and Nucleoside Docking. Recently a
hA3AR model, including the seven TMs and loop regions, was
constructed31,41,42 by homology to the X-ray structure of bovine
rhodopsin.35
For the side-chain refinement of the neoceptors, the H272E
mutant receptor was optimized through a molecular dynamics
procedure after the mutation of His to Glu. In the WT A3AR,
there was a TM H-bonding network between the highly
conserved His272 (7.43) and Glu19 (1.39), as previously
described.21 The mutant H272E hA3AR showed the same
preference of the side-chain ø1 angle at position 7.43 as did
WT, but it had a different intramolecular H-bonding pattern.
The H272E mutant receptor lost H-bonding between position
7.43 and Glu19, due to the electronic repulsion between two
ionized Glu side chains, showing a distance of 4.88 Å between
two Cδ atoms of Glu19 and Glu272.
A conformational search of the isolated 3′-ureido derivative
10, carried out with a MOPAC PM3 calculation,43 showed that
the lowest-energy conformer displayed intramolecular H-bond-
ing between the 3′-carbonyl oxygen of the ureido group and

2694 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
Figure 2. Activation of the neoceptor by the neoagonist 10 protects heart cells from ischemia-induced injury. A) Activation of PLD by 10 in chick
cardiomyocytes expressing the mutant human H272E A3AR. B) Effect of 10 on antiischemic cardioprotection in neoceptor-transfected cardiac
myocytes is shown. Cardiac ventricular myocytes were transfected with cDNA encoding the neoceptor H272E, and the percentage of cells killed
was determined in the absence or the presence of 10 during the 90-min simulated ischemia, as described in Experimental Procedures. Data were
plotted as the percentage of cells killed during the prolonged simulated ischemia. #P < 0.01 compared with control (ANOVA). ANOVA (all four
groups), F ) 15.6, P < 0.0001; all posttest comparisons were significant at P < 0.01 except for the percentage of cells killed at 300 nM 10
compared with 1000 nM and that for 100 nM compared with 300 nM.
the 5′-hydroxyl group and between the 3′-amino and 2′-hydroxyl
groups. The lowest-energy conformer of 12, having a 5′-
uronamide, formed H-bonding between the 3′-CO and 2′-OH
groups and between the 5′-NH and the O of the ribose ring.
However, this conformer of 12 showed no interaction between
3′- and 5′-substituents. Thus, the thermodynamic stability of
the various intramolecular H-bonds, depending on the substitu-
tion pattern, might affect the binding affinity.
one of the urea NH groups in a bidentate fashion. The
5′-hydroxyl group H-bonded with the backbone O atom of
Ser271 (7.42).
Figure 3B shows another possible binding mode, in which
the 3′-ureido group interacted with the Thr94 (3.36) side chain
through H-bonding and the 5′-hydroxyl group formed a H-bond
with the carboxylate group of Glu272 (7.43) and the terminal
3′-NH2 group was close to Ser97 (3.39). The N6-amino group
interacted through H-bonding with Gln167 (EL2), a residue
previously predicted to be in proximity to the nucleoside.36 In
comparison to Figure 3A, this docking mode required a different
orientation of the N6-benzyl ring toward Phe182 (5.43). Mo-
lecular modeling indicates that the reason for the enhanced
binding affinity of N6-benzyl compared with N6-methyl deriva-
tives may be additional hydrophobic interaction at the N6-benzyl
group.
A complex of the A3-selective agonist Cl-IB-MECA docked
in the hA3AR was constructed with only minor modifications
from the previous model.31,41 This model featured putative
H-bonds between the exocyclic NH and the side chain nitrogen
(lone pair) of N250 (6.55) and between the carboxamide NH
of Gln167 (EL2) and the purine N3 atom. This docked ligand
showed an anti-conformation of the adenine ring. In the putative
binding site of the ribose ring, intermolecular H-bonds formed
between the 2′-OH group and the carbonyl O atom of Ile268
(7.39) and among the 3′-OH group, the backbone O of Ser271
(7.42), and the imidazole ring of His272 (7.43). The 5′-amide
NH also formed a H-bond with Thr94 (3.36), and the 5′-carbonyl
group displayed intramolecular H-bonding with the 3′-hydroxyl
group. However, the docked conformation of N6-methyl-3′-
ureidoadenosine 1438 showed no H-bonding interaction of the
3′- and 5′-substituents with Ser271 or His272, consistent with
its loss of binding affinity at all subtypes of ARs. The lack of
H-bonding was expected from the observed electronic and steric
repulsion of the starting geometry of 3′-ureidoadenosine deriva-
tives.38
While these 3′-ureidoadenosine analogues lost binding affinity
at all subtypes of ARs, they exhibited highly selective enhance-
ment of binding affinity at the H272D or H272E mutant A3-
ARs, suggesting a favorable interaction between 3′-ureido group
and mutated acidic residues. While Cl-IB-MECA, a selective
A3 agonist, displayed a single favorable binding mode, the
docking result of the 3′-ureido analogue 10 from the combination
of the FlexiDock and FlexX57 automatic docking programs had
two energetically favorable binding modes. The docking results
showing two putative complexes of the hA3 H272E AR with
compound 10 are shown in Figure 3. As illustrated in Figure
3A, one binding mode similar to that of Cl-IB-MECA, gained
energetically through H-bonding of the 3′-ureido substituent and
of the 2′-hydroxyl group with the γ-carboxylate group of Glu272
(7.43), but lost hydrophilic interactions, i.e., H-bonding, at the
N6 and the N3 atoms in the adenine ring. According to the model,
the oxygens of the Glu272 side chain were each H-bonded to
Discussion
In this study, we used an integrated approach of mutagenesis,
radioligand binding and functional assays combined with
molecular modeling to identify matched pairs of neoligands and
neoceptors. We tested a wide range of synthetic amine-bearing
nucleoside analogues and modulated the 3′-substituent chain
length in concert with N6-, 5′-, and 2-modifications to identify
engineered agonists that orthogonally activate mutant but not
wild-type receptors. Mutation of His272, to Ala, Asp, and Glu,
was compared with the WT A3AR to optimize the electrostatic
or H-bonding interaction between the 3′-amino moiety and the
carboxylate group of the mutation site. The extended and
multiply H-bonding 3′-deoxy-3′-ureido derivative 10, containing
an N6-3-iodobenzyl substituent, had no significant effect on the
wild-type A3AR but displayed a dramatically enhanced affinity
of 0.22 µM at the H272E neoceptor. In a previous study,21 it
was demonstrated that the H272E neoceptor showed decreased
affinity for NECA and Cl-IB-MECA, here we further demon-
strated this neoceptor is completely insensitive to the endogenous
agonist, adenosine, as demonstrated in both binding and
functional assays. Thus, truly orthogonal ligand-receptor pairs
have been identified, which should be useful in elucidating
signaling transduction mechanisms and should provide insights
into the therapeutics of genetically related diseases.
Mutations in genes encoding G protein-coupled receptors
(GPCRs) are an important cause of human disease.44 Study of
the high-resolution SNP maps of 23 genes encoding GPCRs in
the Japanese population identified 300 SNPs, including 83 in

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2695
Figure 3. Two energetically favorable binding modes of the N6-(3-I-benzyl)-3′-ureidoadenosine 10 in the binding site of the mutant H272E hA3-
AR. The binding mode similar to the hA3/Cl-IB-MECA complex (A) and another binding mode (B), which was energetically unfavorable in the
hA3/Cl-IB-MECA complex, are shown. All ligands are displayed as ball-and-stick models in the atom-by-atom color, and the side chains of the
hA3AR are shown as stick models. The H-bonding between each ligand and the mutant hA3AR is displayed in yellow. The A3AR is represented
by a tube model with a different color for each TM (TM2 in orange, TM3 in yellow, TM4 in green, TM5 in cyan, TM6 in blue, TM7 in purple).
adenosine receptor family genes.45 The A2A adenosine receptor
1976T>C genetic variant has been shown to confer susceptibil-
ity to panic disorder and caffeine-induced anxiety.46,47 Specific
GPCR polymorphisms in adrenergic genes have already been
shown to confer susceptibility to congestive heart failure.13
Development of small molecules for GPCRs to rescue mutation
induced functional loss or to inhibit mutation-caused constitutive
activation will be novel forms of treatment for such diseases.
The reengineering of the interface of small molecules and
proteins, such that the mutated protein will interact exclusively
with chemically modified ligand in an orthogonal manner, has
been explored for a variety of targets including receptors,
enzymes, etc., to elucidate the role of signal transduction
pathways in living systems.48-51 The neoceptor approach, which
focuses on cell-surface GPCRs, is intended for therapeutic
application that employs a targeted vector to deliver the
neoceptor gene to a target organ. In addition to the potential
therapeutic applications, the neoceptor/neoligand pairs may be
used for specific mechanistic probing of pharmacological effects,
for example, in cases where the application of a normal receptor
agonist might cause complicating effects. In several aspects,
the neoceptor approach described in this study may be distin-
guished from the RASSL approach as described.14 First, the
neoceptor approach concerns the reengineering of both the
receptor and the ligands using molecular modeling. Second, the
RASSL described earlier14 showed somewhat parallel affinity
decrease for both endogenous and synthetic ligands, which is
not truly orthogonal. The synthetic agonist may activate the
endogenous receptor and the engineered receptor simultaneously.
Third, the RASSL was achieved by engineering a chimeric
receptor, while the neoceptor is achieved by reengineering the
TM binding site for a specific group of the small-molecule
agonist as predicted by using molecular modeling.
concept was further tested in a cardiac myocyte model of
simulated ischemia and cardioprotection. Cardiomyocytes over-
expressing the wild-type hA3AR, activated by endogenous
adenosine released by ischemia, showed potent antiischemic
resistance. In contrast, cells expressing the H272E mutant
receptor showed only a slight change. Activation of the
overexpressed neoceptor by the neoligand resulted in a full
antiischemic protective effect. The demonstration that the
neoligand was able to induce a potent activation of the
cardioprotective PLD pathway27,28 provides additional support
for this concept.
We used site-directed mutagenesis to probe the recognition
elements responsible for the selective affinity enhancement. The
neoligand 10 was found to bind with higher affinity to carboxylic
acid mutations at a spatially distinct region of the A3AR, i.e.,
receptors containing H272E (or D) in TM7 mutations, expressed
in COS-7 cells, while other mutations (T94A, T94E or Q167E
in EL2) did not result in this gain in affinity. This was consistent
with the docking mode of 10 that was similar to that of the
high affinity agonist Cl-IB-MECA (Figure 3A), but not the
alternate binding mode (Figure 3B).
A carboxylic acid introduced in the sequence of the human
A3AR was found to pair with several functional groups on the
ligand in an energetically favorable manner. Pairing of an Asp
or Glu residue with an amino or aminomethyl group could be
explained on the basis of an electrostatic attraction. The
enhancement of the 3′-azido derivative 15 may also result from
an electrostatic interaction with the middle nitrogen of the N3
group, which would be positively charged. Finally, the suc-
cessful pairing of a 3′-ureido group with the carboxylate could
be explained on the basis of the ability of both groups to form
multiple H-bonds.
The neoceptor approach could be particularly useful for
studying ARs to achieve the receptor-subtype-selectivity and
tissue selectivity due to the ubiquitous presence of ARs.52 In
the present study, the functionality of the neoagonist/neoceptor
Conclusions
We have identified an engineered agonist that activates
exclusively neoceptors and have shown that the orthogonality

2696 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
1,2,5-Tri-O-acetyl-3-azido-3-deoxy-D-ribofuranose (24). A so-
lution of 23 (1.76 g, 6.84 mmol) in 85% formic acid (30 mL) was
stirred for 1.5 h at 60 °C and evaporated. The residue was dissolved
in pyridine (20 mL) and Ac2O (8.39 mL, 88.9 mmol) was added
to the reaction mixture. The mixture was stirred at room temperature
for 16 h and evaporated. The residue was partitioned between
EtOAc and water. The organic layer was washed with brine, dried
(MgSO4), filtered, and evaporated. The residue was purified by silica
gel column chromatography (hexanes/EtOAc ) 2/1) to give 24 (2
g, 97%) as an oil: 1H NMR (CDCl3) δ 6.15 (m, 1 H), 5.33 (dd, 1
H, J ) 4.9, 9.7 Hz), 4.05-4.38 (m, 5 H), 2.09 (m, 9 H); IR (KBr):
2116 (N3) cm-1; FAB-MS m/z 324 [M + Na]+. Anal. (C11H15N3O7)
C, H, N.
3-Azido-3-deoxy-1,2-O-isopropylidene-r-D-ribofuranuronic
Acid Methyl Amide (33). A solution of 21 (2.08 g, 7.29 mmol) in
75% AcOH (30 mL) was stirred at 55 °C for 1.5 h. The reaction
mixture was evaporated and the residue was dissolved in CCl4 (14
mL), CH3CN (14 mL), and H2O (20 mL). To this solution were
added NaIO4 (6.55 g, 30.6 mmol) and ruthenium trichloride hydrate
(33 mg, 0.16 mmol) at room temperature. The reaction mixture
was stirred at room temperature for 4 h and filtered through a Celite
pad. The filtrate was extracted with CH2Cl2 (× 3). The combined
organic layers were dried (MgSO4), filtered, and evaporated to give
32.
Oxalyl chloride (1.5 mL) was added to a solution of 32 in an-
hydrous CH2Cl2 (15 mL). DMF (0.5 mL) was added, and the reac-
tion mixture was stirred at room temperature for 16 h and evapo-
rated. The residue was dissolved in anhydrous CH2Cl2 (15 mL)
and cooled to 0 °C. A solution of 2 M NH2CH3 in THF (0.37 mL)
was added dropwise. After being stirred for 3 h, the mixture was
diluted with water and extracted with CH2Cl2 (× 3). The combined
organic layers were dried (MgSO4), filtered, and evaporated. The
residue was purified by silica gel column chromatography (hexanes/
EtOAc ) 1/2) to give 33 (772 mg, 44%) as a solid: 1H NMR
(CDCl3) δ 6.43 (brs, 1 H), 5.84 (d, 1 H, J ) 3.3 Hz), 4.71 (dd, 1
H, J ) 3.3, 4.6 Hz), 4.48 (d, 1 H, J ) 9.6 Hz), 3.63 (dd, 1 H, J )
4.5, 9.5 Hz), 2.86 (d, 1 H, J ) 5.0 Hz), 1.58 (s, 3 H), 1.38 (s, 3 H);
FAB-MS m/z 265 [M + Na]+. Anal. (C9H14N4O4) C, H, N.
1,2-Di-O-acetyl-3-azido-3-deoxy-D-ribofuranuronic Acid Meth-
yl Amide (34). Compound 33 (772 mg, 3.19 mmol) was dissolved
in a solution of AcOH (20 mL) and Ac2O (2.4 mL). The mixture
was cooled to 0 °C and concd H2SO4 (0.014 mL) was added to
this solution. The reaction mixture was stirred at room temperature
for 18 h and poured into the saturated NaHCO3 solution. The
mixture was extracted with CH2Cl2 (× 3). The combined organic
layers were washed with brine, dried (MgSO4), filtered, and
evaporated. The residue was purified by silica gel column chro-
matography (hexanes/EtOAc ) 1/2) to give 34 (750 mg, 82%) as
an oil: 1H NMR (CDCl3) δ 6.44 (brs, 1 H), 6.15 (s, 1 H), 5.30 (d,
1 H, J ) 4.8 Hz), 4.48 (d, 1 H, J ) 7.2 Hz), 4.39 (dd, 1 H, J )
4.8, 7.2 Hz), 2.86 (d, 1 H, J ) 5.0 Hz), 2.18 (s, 3 H), 2.08 (s, 3 H);
IR (KBr): 2120 (N3), 1752 (CdO), 1674 (CdO) cm-1; FAB-MS
m/z 287 [M + H]+. Anal. (C10H14N4O6) C, H, N.
General Procedure for the Synthesis of 25a, 25b, and 35. A
mixture of 6-chloropurine or 2,6-dichloropurine (2.0 equiv) and
ammonium sulfate (catalytic amount) in anhydrous HMDS (30 mL)
was refluxed under nitrogen atmosphere for 16 h and concentrated
under anhydrous conditions. The residue was dissolved in anhydrous
1,2-dichloroethane (20 mL), and a solution of 24 and 34 in 1,2-
dichloroethane (20 mL) was added to this solution followed by
addition of TMSOTf (2.0 equiv) at 0 °C. The reaction mixture was
stirred at room temperature for 20 min and then at 60 °C for 2 h.
The mixture was quenched by addition of saturated NaHCO3
solution and filtered through a Celite pad. The filtrate was extracted
with CH2Cl2 (× 3). The combined organic layers were dried
(MgSO4), filtered, and evaporated. The residue was purified by silica
gel column chromatography to give 25a, 25b, and 35.
of binding and activation results in cardioprotective effects. This
suggests that the 3′-urea derivatives described in this study might
be useful therapeutically in combination with targeted gene
delivery of a carboxylic acid mutant A3AR such as H272E. The
neoceptor, in general, could be an important therapeutic
approach for tissue-specific GPCR activation, given successful
targeted delivery (without side effects) of the neoceptor gene
to a specific organ or tissue.
Experimental Procedures
Chemical Synthesis. General. 1NMR spectra were recorded in
a 300 MHz apparatus using tetramethylsilane (TMS) as an internal
standard, and the chemical shifts are reported in ppm (δ). Coupling
constants are reported in Hertz (Hz). Optical rotations were
determined on Jasco III in methanol of DMF. Infrared spectra were
recorded in a Perkin-Elmer 1710 FTIR spectrophotometer. Mass
spectra recorded by FAB (Fast atom bombardment) on a VG Tro-
2, GC-MS. TLC were carried out on Merck silica gel 60 F254
precoated plates, and silica gel column chromatography was
performed on silica gel 60, 230-400 mesh, Merck. All anhydrous
solvents were distilled over CaH2 or Na/benzophenone prior to use.
3-Azido-3-deoxy-1,2:5,6-di-O-isopropylidene-r-D-allofura-
nose (21). To a stirred solution of 1,2;5,6-di-O-isopropyidene-R-
D-glucofuranose (3.0 g, 11.53 mmol) and pyridine (2.8 mL, 34.62
mmol) in dichloromethane (30 mL) was added trifluoromethane-
sulfonic anhydride (2.9 mL, 17.24 mmol) at 0 °C. After being stirred
for 1 h at 0 °C, the reaction mixture was extracted with dichlo-
romethane and water. The organic layer was washed with brine,
dried (MgSO4), filtered, and evaporated to give 20.
To a solution of 20 in anhydrous DMF (20 mL) was added
sodium azide (2.25 g, 34.61 mmol), and the mixture was stirred at
room temperature for 48 h. The reaction mixture was poured into
water (200 mL) and extracted with ethyl acetate (× 3). The
combined organic layers were dried (MgSO4), filtered, and evapo-
rated. The residue was purified by silica gel column chromatography
(hexanes/EtOAc ) 4/1) to give 21 (1.45 g, 44%) as an oil: 1H
NMR (CDCl3) δ 5.76 (d, 1 H, J ) 3.7 Hz), 4.70 (t, 1 H, J ) 4.1
Hz), 3.95-4.21 (m, 4 H), 3.48 (dd, 1 H, J ) 4.9, 9.0 Hz), 1.56 (s,
3 H), 1.46 (s, 3 H), 1.36 (s, 3 H), 1.34 (s, 3 H); IR (KBr): 2109
(N3) cm-1; FAB-MS m/z 286 [M + H]+. Anal. (C12H19N3O5) C,
H, N.
3-Azido-3-deoxy-1,2-O-isopropylidene-r-D-ribofuranose (22).
A mixture of 21 (2.63 g, 9.22 mmol) in 75% AcOH (30 mL) was
stirred at 55 °C for 1.5 h. The reaction mixture was evaporated
and coevaporated with toluene. The residue was dissolved in EtOH
(30 mL) and a solution of NaIO4 (2.37 g, 11.09 mmol) in H2O (15
mL) was added dropwise at 0 °C to the reaction mixture. After the
mixture was stirred at 0 °C for 20 min, NaBH4 (1.05 g, 27.76 mmol)
was added and the reaction mixture was stirred at 0 °C for 1 h.
The reaction mixture was filtered, and the filtrate was evaporated.
The residue was purified by silica gel column chromatography
(hexanes/EtOAc ) 1/1) to give 22 (1.69 g, 85%) as an oil: 1H
NMR (CDCl3) δ 5.79 (d, 1 H, J ) 3.4 Hz), 4.72 (t, 1 H, J ) 4.1
Hz), 4.09 (m, 1 H), 3.96 (dd, 1 H, J ) 2.4, 12.5 Hz), 3.66 (dd, 1
H, J ) 2.9, 12.6 Hz), 3.56 (dd, 1 H, J ) 4.6, 9.5 Hz), 1.56 (s, 3
H), 1.35 (s, 3 H); IR (KBr): 2108 (N3) cm-1; FAB-MS m/z 216
[M + H]+. Anal. (C8H13N3O4) C, H, N.
5-O-Acetyl-3-azido-3-deoxy-1,2-O-isopropylidene-r-D-ribo-
furanose (23). To a stirred solution of 22 (1.5 g, 6.97 mmol) in
anhydrous pyridine (15 mL) was added Ac2O (1.32 mL, 13.99
mmol). The reaction mixture was stirred at room temperature for
3 h and evaporated. The residue was partitioned between EtOAc
and water. The organic layer was washed with brine, dried (MgSO4),
filtered, and evaporated. The residue was purified by silica gel
column chromatography (hexanes/EtOAc ) 2/1) to give 23 (1.76
g, 98%) as an oil: 1H NMR (CDCl3) δ 5.78 (d, 1 H, J ) 3.7 Hz),
4.71 (dd, 1 H, J ) 3.6, 4.8 Hz), 4.34 (m, 1 H), 4.15-4.26 (m, 2
H), 3.31 (dd, 1 H, J ) 4.6, 9.5 Hz), 2.08 (s, 3 H), 1.56 (s, 3 H),
1.34 (s, 3 H); IR (KBr): 2109 (N3) cm-1; FAB-MS m/z 258 [M +
H]+. Anal. (C10H15N3O5) C, H, N.
9-(2,3-Di-O-acetyl-3-azido-3-deoxy-â-D-ribofuranosyl)-6-chlo-
ropurine (25a). 67% yield; white foam; 1H NMR (CDCl3) δ 8.78
(s, 1 H), 8.25 (s, 1 H), 6.12 (d, 1 H, J ) 3.5 Hz), 5.98 (dd, 1 H, J
) 5.9, 3.5 Hz), 4.71 (t, 1 H, J ) 6.3), 4.29-4.40 (m, 3 H), 2.21 (s,

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2697
3 H), 2.09 (s, 3 H); IR (KBr): 2115 (N3), 1746 (CdO) cm-1; FAB-
MS m/z 396 [M + H]+. Anal. (C14H14ClN7O5) C, H, N.
9-(2,3-Di-O-acetyl-3-azido-3-deoxy-â-D-ribofuranosyl)-2,6-
(d, 1 H, J ) 7.7 Hz), 7.04 (t, 1 H, J ) 7.7 Hz), 6.48 (brs, 1 H),
5.79 (d, 1 H, J ) 7.1 Hz), 5.07 (m, 2 H), 4.77 (m, 2 H), 4.48 (m,
2 H), 2.85 (d, 1 H, J ) 4.9 Hz); FAB-MS m/z 536 [M + H]+.
Anal. (C18H18IN9O3) C, H, N.
1
dichloropurine (25b). 51% yield; white foam; H NMR (CDCl3)
δ 8.24 (s, 1 H), 6.10 (d, 1 H, J ) 3.9 Hz), 5.84 (dd, 1 H, J ) 3.9,
5.7 Hz), 4.61(m, 1 H), 4.29-4.50 (m, 3 H), 2.22 (s, 3 H), 2.12 (s,
3 H); IR (KBr): 2116 (N3), 1745 (CdO) cm-1; FAB-MS m/z 431
[M + H]+. Anal. (C14H13Cl2N7O5) C, H, N.
3-Azido-5-(6-chloro-purin-9-yl)-4-acetoxy-tetrahydro-furan-
2-carboxylic Acid Methyl Amide (35). 73% yield; white foam;
1H NMR (CDCl3) δ 8.78 (s, 1 H), 8.25 (s, 1 H), 7.62 (brs, 1 H),
6.15 (d, 1 H, J ) 7.0 Hz), 5.87 (dd, 1 H, J ) 5.7, 7.0 Hz), 4.86
(dd, 1 H, J ) 3.1, 5.7 Hz), 4.58 (d, 1 H, J ) 3.1 Hz), 2.92 (d, 1 H,
J ) 5.0 Hz), 2.05 (s, 3 H); IR (KBr): 2117 (N3), 1751 (CdO),
1673 (CdO) cm-1; FAB-MS m/z 381 [M + H]+. Anal. (C13H13-
ClN8O4) C, H, N.
General Procedure for the Synthesis of 26a and 36. A mixture
of 25a and 35 and 40% methylamine in water (4 mL) in 1,4-dioxane
(10 mL) was stirred at room temperature for 4 h. The reaction
mixture was evaporated, and the residue was purified by silica gel
column chromatography to give 26a and 36, respectively.
N6-Methyl- 9-(3-azido-3-deoxy-â-D-ribofuranosyl)adenine (26a).
90% yield; white solid;. 1H NMR (DMSO-d6) δ 8.31 (s, 1 H), 8.20
(s, 1 H), 7.83 (brs, 1 H), 6.18 (d, 1 H, J ) 5.7 Hz), 5.86 (d, 1 H,
J ) 6.1 Hz), 5.56 (dd, 1 H, J ) 4.5, 7.6 Hz), 4.95 (dd, 1 H, J )
5.7, 11.4 Hz), 4.27 (dd, 1 H, J ) 3.5, 5.4 Hz), 3.93 (dd, 1 H, J )
3.3, 6.8 Hz), 3.50-3.69 (m, 2 H), 2.92 (s, 3 H); IR (KBr): 3430
(OH), 2104 (N3), 1634 (CdO) cm-1; FAB-MS m/z 307 [M + H]+.
Anal. (C11H14N8O3) C, H, N.
General Procedure for the Synthesis of 27a-c and 37a,b. A
mixture of 26a-c, 36, 15, imidazole (5 equiv), and TBSCl (2.5
equiv) in anhydrous DMF (15 mL) was stirred at room temperature
for 24 h. The reaction mixture was poured into water (200 mL)
and extracted with ethyl acetate (× 3). The combined organic layers
were dried (MgSO4), filtered, and evaporated. The residue was
purified by silica gel column chromatography to give 27a-c and
37a,b, respectively.
N6-Methyl-9-(3-azido-2,5-di-O-t-butyldimethylsilyl-3-deoxy-
â-D-ribofuranosyl)adenine (27a). 76% yield; oil; 1H NMR (CDCl3)
δ 8.24 (s, 1 H), 7.94 (s, 1 H), 5.87 (d, 1 H, J ) 4.0 Hz), 5.72 (brs,
1 H), 4.73 (dd, 1 H, J ) 4.0, 5.0 Hz), 4.06 (m, 1 H), 3.91-3.97
(m, 2 H), 3.67 (dd, 1 H, J ) 2.8, 11.6 Hz), 3.05 (d, 3 H, J ) 4.6
Hz), 0.82 (s, 9 H), 0.73 (s, 9 H), 0.11 (s, 3 H), 0.00 (s, 3 H), -0.08
(s, 3 H), -0.19 (s, 3 H); IR (KBr): 2107 (N3) cm-1; FAB-MS m/z
535 [M + H]+. Anal. (C23H42N8O3Si2) C, H, N.
N6-(3-Iodobenzyl)-9-(3-azido-2,5-di-O-t-butyldimethylsilyl-3-
deoxy-â-D-ribofuranosyl)adenine (27b). 84% yield; oil; 1H NMR
(CDCl3) δ 8.24 (s, 1 H), 7.98 (s, 1 H), 7.59 (s, 1 H), 7.46 (d, 1 H,
J ) 7.8 Hz), 7.19 (d, 1 H, J ) 7.7 Hz), 6.90 (t, 1 H, J ) 7.7 Hz),
5.99 (brs, 1 H), 5.88 (d, 1 H, J ) 3.9 Hz), 4.70-4.75 (m, 3 H),
4.07 (m, 1 H), 3.91-3.97 (m, 2 H), 3.68 (dd, 1 H, J ) 2.6, 11.6
Hz), 0.81 (s, 9 H), 0.74 (s, 9 H), 0.01 (s, 3 H), 0.00 (s, 3 H), -0.06
(s, 3 H), -0.16 (s, 3 H); IR (KBr): 2106 (N3) cm-1; FAB-MS m/z
737 [M + H]+. Anal. (C29H45IN8O3Si2) C, H, N.
2-Chloro-N6-(3-iodobenzyl)-9-(3-azido-2,5-di-O-t-butyldimeth-
ylsilyl-3-deoxy-â-D-ribofuranosyl)adenine (27c). 83% yield; oil;
1H NMR (CDCl3) δ 8.24 (s, 1 H), 7.99 (s, 1 H), 7.57 (s, 1 H), 7.47
(d, 1 H, J ) 7.9 Hz), 7.18 (d, 1 H, J ) 7.7 Hz), 6.90 (t, 1 H, J )
7.9 Hz), 6.05 (brs, 1 H), 5.82 (d, 1 H, J ) 3.3 Hz), 4.62 (m, 3 H),
4.08 (m, 1 H), 3.93 (dd, 1 H, J ) 2.9, 11.7 Hz), 3.82 (dd, 1 H, J
) 4.8, 6.2 Hz), 3.67 (dd, 1 H, J ) 2.4, 11.6 Hz), 0.80 (s, 9 H),
0.76 (s, 9 H), 0.02 (s, 3 H), 0.00 (s, 3 H), -0.03 (s, 3 H), -0.10
(s, 3 H); IR (KBr): 2106 (N3) cm-1; FAB-MS m/z 771 [M]+. Anal.
(C29H44ClIN8O3Si2) C, H, N.
3-Azido-5-(6-methylaminopurin-9-yl)-4-hydroxy-tetrahydro-
furan-2-carboxylic Acid Methyl Amide (36). 87% yield; white
solid; H NMR (DMSO-d6) δ 8.69 (d, 1 H, J ) 4.4 Hz), 8.38 (s,
1 H), 8.27 (s, 1 H), 7.87 (brs, 1 H), 6.58 (d, 1 H, J ) 5.5 Hz), 5.95
(d, 1 H, J ) 6.4 Hz), 4.90 (dd, 1 H, J ) 5.6, 11.4 Hz), 4.45 (dd,
1 H, J ) 3.0, 5.0 Hz), 4.29 (d, 1 H, J ) 2.9 Hz), 2.93 (s, 3 H),
2.66 (d, 3 H, J ) 4.6 Hz); IR (KBr): 3379 (OH), 2123 (N3), 1662
(CdO) cm-1; FAB-MS m/z 334 [M + H]+. Anal. (C12H15N9O3)
C, H, N.
1
General Procedure for the Synthesis of 26b, 26c, and 15. A
mixture of 25a, 25b, 15, 3-iodobenzylamine hydrochloride (1.1
equiv), and Et3N (3.0 equiv) in EtOH (10 mL) was stirred at 50 °C
for 18 h. The reaction mixture was evaporated, and the residue
was partitioned between CH2Cl2 and water. The organic layer was
washed with brine, dried (MgSO4), filtered, and evaporated. The
residue was dissolved in MeOH (10 mL), and 28% NaOMe (1 mL)
was added to this solution. The reaction mixture was stirred at room
temperature for 2 h and evaporated. The residue was purified by
silica gel column chromatography to give 26b, 26c, and 15,
respectively.
3-Azido-5-(6-methylaminopurin-9-yl)-4-t-butyldimethylsiloxy-
tetrahydro-furan-2-carboxylic Acid Methyl Amide (37a). 83%
1
yield; foam; H NMR (CDCl3) δ 9.31 (brs, 1 H), 8.49 (s, 1 H),
7.88 (s, 1 H), 6.10 (d, 1 H, J ) 4.6 Hz), 5.89 (d, 1 H, J ) 8.1 Hz),
5.19 (dd, 1 H, J ) 5.1, 8.1 Hz), 4.59 (s, 1 H), 4.40 (d, 1 H, J ) 5.1
Hz), 3.33 (brs, 3 H), 3.06 (d, 1 H, J ) 4.8 Hz), 0.87 (s, 9 H), 0.00
(s, 3 H), -0.36 (s, 3 H); IR (KBr): 2105 (N3) cm-1; FAB-MS m/z
448 [M + H]+. Anal. (C18H29N9O3Si) C, H, N.
3-Azido-5-(6-(3-iodobenzylamino)purin-9-yl)-4-t-butyldimeth-
ylsiloxy-tetrahydro-furan-2-carboxylic Acid Methyl Amide (37b).
1
N6-(3-Iodobenzyl)-9-(3-azido-3-deoxy-â-D-ribofuranosyl)ad-
71% yield; foam; H NMR (CDCl3) δ 9.22 (d, 1 H, J ) 4.2 Hz),
1
8.48 (s, 1 H), 7.88 (s, 1 H), 7.82 (s, 1 H), 7.72 (d, 1 H, J ) 7.9
Hz), 7.43 (d, 1 H, J ) 7.7 Hz), 7.16 (t, 1 H, J ) 7.7 Hz), 6.37 (t,
1 H, J ) 6.0 Hz), 5.88 (d, 1 H, J ) 7.9 Hz), 5.18 (dd, 1 H, J )
5.2, 7.9 Hz), 4.90 (brs, 2 H), 4.59 (s, 1 H), 4.40 (d, 1 H, J ) 5.1
Hz), 3.05 (d, 1 H, J ) 5.8 Hz), 0.87 (s, 9 H), 0.00 (3, 3 H), -0.36
(s, 3 H); IR (KBr): 2105 (N3) cm-1; FAB-MS m/z 672 [M + Na]+.
Anal. (C24H32IN9O3Si) C, H, N.
enine (26b). 86% yield; solid; H NMR (DMSO-d6) δ 8.60 (s, 1
H), 8.47 (s, 1 H), 8.29 (s, 1 H), 7.78 (s, 1 H), 7.63 (d, 1 H, J ) 7.9
Hz), 7.40 (d, 1 H, J ) 7.4 Hz), 7.14 (t, 1 H, J ) 7.9 Hz), 6.28 (d,
1 H, J ) 5.3 Hz), 5.97 (d, 1 H, J ) 5.8 Hz), 5.57 (dd, 1 H, J )
4.6, 7.1 Hz), 5.05 (m, 1 H), 4.73 (m, 2 H), 4.37 (m, 1 H), 4.05 (m,
1 H), 3.72 (m, 1 H), 3.62 (m, 1 H); IR (KBr): 3426 (OH), 2106
(N3), 1622 (CdO) cm-1; FAB-MS m/z 590 [M + H]+. Anal.
(C17H17IN8O3) C, H, N.
General Procedure for the Synthesis of 28a-c and 38a,b. To
a stirred solution of 27a-c and 37a,b in THF (15 mL) was added
triphenylphosphine (1.5 equiv) at 0 °C. After being stirred for 30
min, NH4OH (1.8 mL) and H2O (0.3 mL) were added to the reaction
mixture. The mixture was stirred overnight at room temperature
and evaporated. The residue was purified by silica gel column
chromatography to give 28a-c and 38a,b, respectively.
2-Chloro-N6-(3-iodobenzyl)-9-(3-azido-3-deoxy-â-D-ribofura-
nosyl)adenine (26c). 95% yield; solid; 1H NMR (DMSO-d6) δ 7.71
(s, 1 H), 7.69 (s, 1 H), 7.59 (d, 1 H, J ) 7.9 Hz), 7.26 (m, 1 H),
7.02 (t, 1 H, J ) 7.7 Hz), 6.55 (s, 1 H), 5.97 (d, 1 H, J ) 10.1 Hz),
5.70 (d, 1 H, J ) 7.3 Hz), 5.24 (m, 1 H), 4.64-4.77 (m, 3 H), 4.31
(d, 1 H, J ) 5.1 Hz), 4.19 (s, 1 H), 3.94 (d, 1 H, J ) 12.81 Hz),
3.68 (m, 1 H); IR (KBr): 3305 (OH), 2112 (N3), 1621 (CdO) cm-1
FAB-MS m/z 543 [M + H]+. Anal. (C17H16ClIN8O3) C, H, N.
;
N6-Methyl-9-(3-amino-2,5-di-O-t-butyldimethylsilyl-3-deoxy-
â-D-ribofuranosyl)adenine (28a). 88% yield; foam; 1H NMR
(CDCl3) δ 8.24 (s, 1 H), 8.13 (s, 1 H), 5.89 (d, 1 H, J ) 1.7 Hz),
5.75 (dd, 1 H, J ) 5.9, 3.5 Hz), 4.24 (dd, 1 H, J ) 1.7, 4.8 Hz),
3.93 (m, 1 H), 3.74-3.82 (m, 2 H), 3.45 (dd, 1 H, J ) 4.8, 8.3
Hz), 3.03 (d, 3 H, J ) 5.0 Hz), 0.81 (s, 9 H), 0.80 (s, 9 H), 0.07
3-Azido-5-(6-(3-iodobenzylamino)purin-9-yl)-4-hydroxy-tet-
rahydro-furan-2-carboxylic Acid Methyl Amide (15). 89% yield;
white solid; 1H NMR (CDCl3) δ 8.61 (d, 1 H, J ) 4.4 Hz), 8.28 (s,
1 H), 7.71 (s, 1 H), 7.69 (s, 1 H), 7.60 (d, 1 H, J ) 7.8 Hz), 7.30

2698 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
(s, 3 H), 0.01 (s, 3 H), 0.00 (s, 3 H), -0.02 (s, 3 H); IR (KBr):
3290, 2931, 1624, 1119, 838 cm-1; FAB-MS m/z 509 [M + H]+.
Anal. (C23H44N6O3Si2) C, H, N.
1 H, J ) 7.9 Hz), 6.91 (t, 1 H, J ) 7.9 Hz), 6.08 (brs, 1 H), 5.88
(d, 1 H, J ) 1.8 Hz), 4.63 (s, 2 H), 4.42-4.51 (m, 2 H), 4.11 (m,
1 H), 3.99 (s, 2 H), 3.93 (dd, 1 H, J ) 2.2, 11.9 Hz), 3.69 (dd, 1H,
J ) 2.4, 11.7 Hz), 0.81 (s, 9 H), 0.78 (s, 9 H), 0.04 (s, 3 H) 0.00
(s, 6 H), -0.04 (s, 3 H); IR (KBr): 3296, 2952, 1700, 1618, 1537,
1470, 1314, 1223, 1127, 837, 781 cm-1; FAB-MS m/z 863 [M +
H]+.
N6-(3-Iodobenzyl)-9-(3-amino-2,5-di-O-t-butyldimethylsilyl-3-
deoxy-â-D-ribofuranosyl)adenine (28b). 96% yield; foam; 1H
NMR (CDCl3) δ 8.24 (s, 1 H), 8.17 (s, 1 H), 7.60 (s, 1 H), 7.45 (d,
1 H, J ) 7.0 Hz), 7.20 (d, 1 H, J ) 7.7 Hz), 6.90 (t, 1 H, J ) 7.7
Hz), 6.00 (brs, 1 H), 5.90 (d, 1 H, J ) 1.7 Hz), 4.68 (brs, 2 H),
4.23 (dd, 1 H, J ) 1.5, 4.6 Hz), 3.94 (m, 1 H), 3.74-3.82 (m, 2
H), 3.45 (dd, 1 H, J ) 4.7, 8.4 Hz), 0.81 (s, 9 H), 0.80 (s, 9 H),
0.08 (s, 3 H), 0.02 (s, 3 H), 0.01 (s, 3 H), 0.00 (s, 3 H); IR (KBr):
2931, 1618, 1469, 1120, 839, 780 cm-1; FAB-MS m/z 711 [M +
H]+. Anal. (C29H47IN6O3Si2) C, H, N.
3-Chloroacetylureido-5-(6-methylaminopurin-9-yl)-4-t-bu-
tyldimethylsiloxy-tetrahydro-furan-2-carboxylic Acid Methyl
1
Amide (39a). 94% yield; oil; H NMR (CDCl3) δ 8.92 (d, 1 H, J
) 5.1 Hz), 8.53 (d, 1 H, J ) 4.6 Hz), 8.47 (s, 1 H), 7.96 (s, 1 H),
6.71 (brs, 1 H), 6.51 (brs, 1 H), 6.26 (d, 1 H, J ) 4.2 Hz), 5.96 (d,
1 H, J ) 5.7 Hz), 5.01 (t, 1 H, J ) 7.9 Hz), 4.74-4.81 (m, 2 H),
4.13 (s, 2 H), 3.29 (brs, 3 H), 2.98 (d, 1 H, J ) 4.7 Hz), 0.86 (s,
9 H), 0.00 (s, 3 H), -0.20 (s, 3 H); IR (KBr): 3296, 2953, 1703,
1626, 1536, 1237, 1155, 840, 756 cm-1; FAB-MS m/z 542 [M +
H]+.
2-Chloro-N6-(3-iodobenzyl)-9-(3-amino-2,5-di-O-t-butyldi-
methylsilyl-3-deoxy-â-D-ribofuranosyl)adenine (28c). 94% yield;
1
foam; H NMR (CDCl3) δ 8.18 (s, 1 H), 7.59 (s, 1 H), 7.47 (d, 1
H, J ) 7.9 Hz), 7.19 (d, 1 H, J ) 7.7 Hz), 6.91 (t, 1 H, J ) 7.7
Hz), 6.08 (brs, 1 H), 5.82 (d, 1 H, J ) 1.1 Hz), 4.63 (s, 2 H), 4.17
(d, 1 H, J ) 4.7 Hz), 3.96 (dd, 1 H, J ) 2.4, 11.5 Hz), 3.74-3.80
(m, 2 H), 3.38 (dd, 1 H, J ) 4.4, 8.8 Hz), 0.82 (s, 9 H), 0.81 (s, 9
H), 0.15 (s, 3 H), 0.03 (s, 3 H), 0.02 (s, 3 H), 0.00 (s, 3 H); IR
(KBr): 2930, 1618, 1466, 1314, 1121, 838, 780 cm-1; FAB-MS
m/z 746 [M + H]+. Anal. (C29H46ClIN6O3Si2) C, H, N.
3-Chloroacetylureido-5-(6-(3-iodobenzylamino)purin-9-yl)-4-
t-butyldimethylsiloxy-tetrahydro-furan-2-carboxylic Acid Meth-
yl Amide (39b). 88% yield; oil; 1H NMR (CDCl3) δ 8.87 (s, 1 H),
8.46 (s, 1 H), 8.31 (s, 1 H), 7.97 (s, 1 H), 7.80 (s, 1 H), 7.63 (d, 1
H, J ) 7.8 Hz), 7.39 (d, 1 H, J ) 7.7 Hz), 7.11 (t, 1 H, J ) 7.7
Hz), 6.42 (m, 2 H), 6.00 (s, 1 H), 5.96 (d, 1 H, J ) 5.5 Hz), 4.76-
5.00 (m, 5 H), 4.22 (s, 2 H), 2.97 (d, 3 H, J ) 4.8 Hz), 0.85 (s, 9
H), 0.00 (s, 3 H), -0.18 (s, 3 H); IR (KBr): 3290, 2952, 1705,
1619, 1535, 1225, 841, 755 cm-1; FAB-MS m/z 744 [M + H]+.
3-Amino-5-(6-methylaminopurin-9-yl)-4-t-butyldimethylsiloxy-
tetrahydro-furan-2-carboxylic Acid Methyl Amide (38a). 95%
1
yield; oil; H NMR (CD3OD) δ 8.37 (s, 1 H), 8.35 (s, 1 H), 6.09
(d, 1 H, J ) 5.5 Hz), 4.85 (dd, 1 H, J ) 5.5, 11.0 Hz), 4.39 (d, 1
H, J ) 3.8 Hz), 3.72 (m, 1 H), 3.15 (brs, 3 H), 2.90 (s, 3 H), 0.87
(s, 9 H), 0.00 (s, 3 H), -0.16 (s, 3 H); IR (KBr): 3431, 2930,
1629, 1054, 833, 643 cm-1; FAB-MS m/z 422 [M + H]+. Anal.
(C18H31N7O3Si) C, H, N.
3-Amino-5-(6-(3-iodobenzylamino)purin-9-yl)-4-t-butyldi-
methylsiloxy-tetrahydro-furan-2-carboxylic Acid Methyl Amide
(38b). 94% yield oil; 1H NMR (CDCl3) δ 8.99 (brs, 1 H), 8.50 (s,
1 H), 7.91 (s, 1 H), 7.85 (s, 1 H), 7.74 (d, 1 H, J ) 7.8 Hz), 7.45
(d, 1 H, J ) 7.7 Hz), 7.17 (t, 1 H, J ) 7.9 Hz), 6.35 (brs, 1 H),
6.11 (d, 1 H, J ) 6.6 Hz), 4.91 (m, 3 H), 4.56 (s, 1 H), 3.93 (dd,
1 H, J ) 2.6, 5.5 Hz), 3.05 (d, 1 H, J ) 4.7 Hz), 0.92 (s, 9 H),
0.00 (s, 3 H), -0.21 (s, 3 H); IR (KBr): 3272, 2931, 1670, 1619,
1473, 1338, 1253, 1151, 1059, 839, 754 cm-1; FAB-MS m/z 624
[M + H]+. Anal. (C24H34IN7O3Si) C, H, N.
General Procedure for the Synthesis of 29a-c and 39a,b. To
a stirred solution of 28a-c and 38a,b in anhydrous DMF (10 mL)
was added chloroacetyl isocyanate (1.1 equiv) at 0 °C. After being
stirred for 2 h at 0 °C, the reaction mixture was evaporated and the
residue was purified by silica gel column chromatography to give
29a-c and 38a,b.
General Procedure for the Synthesis of 30a-c and 40a,b. To
a stirred solution of 29a-c and 39a,b in MeOH (10 mL) was added
28% NaOMe (0.27 mL) at room temperature. The reaction mixture
was stirred for 18 h at room temperature and evaporated. The
residue was purified by silica gel column chromatography to give
30a-c and 40a,b.
N6-Methyl-9-(2,5-di-O-t-butyldimethylsilyl-3-deoxy-3-ureido-
â-D-ribofuranosyl)adenine (30a). 75% yield; foam; 1H NMR
(CDCl3) δ 8.25 (s, 1 H), 8.01 (s, 1 H), 5.90 (d, 1 H, J ) 3.5 Hz),
5.79 (brs, 1 H), 4.96 (d, 1 H, J ) 6.1 Hz), 4.66 (s, 2 H), 4.55 (m,
1 H), 4.24 (dd, 1 H, J ) 6.1, 12.3 Hz), 4.05 (m, 1 H), 3.89 (dd, 1
H, J ) 2.2, 11.5 Hz), 3.76 (dd, 1 H, J ) 2.6, 11.7 Hz), 3.08 (d, 3
H, J ) 4.4 Hz), 0.81 (s, 9 H), 0.75 (s, 9 H), 0.00 (s, 3 H), -0.01
(s, 3 H), -0.10 (s, 3 H), -0.11 (s, 3 H); IR (KBr): 3328, 2932,
1624, 1256, 1125, 837, 782 cm-1; FAB-MS m/z 552 [M + H]+.
Anal. (C24H45N7O4Si2) C, H, N.
N6-(3-Iodobenzyl)-9-(2,5-di-O-t-butyldimethylsilyl-3-deoxy-3-
ureido-â-D-ribofuranosyl)adenine (30b). 89% yield; foam; 1H
NMR (CDCl3) δ 8.25 (s, 1 H), 8.05 (s, 1 H), 7.61 (s, 1 H), 7.48 (d,
1 H, J ) 7.7 Hz), 7.21 (d, 1 H, J ) 7.7 Hz), 6.92 (t, 1 H, J ) 7.7
Hz), 6.00 (brs, 1 H), 5.90 (d, 1 H, J ) 3.1 Hz), 4.86 (d, 1 H, J )
6.9 Hz), 4.72 (brs, 2 H), 4.57 (m, 3 H), 4.25 (dd, 1 H, J ) 6.8,
12.4 Hz), 4.04 (m, 1 H), 3.90 (dd, 1 H, J ) 2.2, 11.7 Hz), 3.75
(dd, 1 H, J ) 2.4, 11.5 Hz), 0.81 (s, 9 H), 0.76 (s, 9 H), 0.00 (s,
3 H), -0.01 (s, 3 H), -0.05 (s, 3 H), -0.08 (s, 3 H); IR (KBr):
N6-Methyl-9-(3-chloroacetylureido-2,5-di-O-t-butyldimethyl-
silyl-3-deoxy-â-D-ribofuranosyl)adenine (29a). 87% yield; oil; 1H
NMR (CDCl3) δ 8.41 (m, 2 H), 8.25 (s, 1 H), 8.10 (s, 1 H), 5.94
(d, 1 H, J ) 2.6 Hz), 5.72 (brs, 1 H), 4.47-4.55 (m, 2 H), 4.12
(m, 1 H), 3.99 (s, 2 H), 3.90 (dd, 1 H, J ) 2.0, 11.5 Hz), 3.72 (dd,
1 H, J ) 3.0, 11.5 Hz), 3.06 (d, 3 H, J ) 4.7 Hz), 0.81 (s, 9 H),
0.75 (s, 9 H), 0.00 (s, 6 H), -0.07 (s, 3 H), -0.12 (s, 3 H); IR
3306, 2931, 1671, 1617, 1471, 1336, 1256, 1125, 837, 781 cm-1
;
FAB-MS m/z 754 [M + H]+. Anal. (C30H48IN7O4Si2) C, H, N.
2-Chloro-N6-(3-iodobenzyl)-9-(2,5-di-O-t-butyldimethylsilyl-
3-deoxy-3-urido-â-D-ribofuranosyl)adenine (30c). 91% yield;
(KBr): 3303, 2953, 1702, 1623, 1536, 1254, 1229, 1127, 837 cm-1
;
FAB-MS m/z 629 [M + H]+.
1
foam; H NMR (CDCl3) δ 8.06 (s, 1 H), 7.60 (s, 1 H), 7.49 (d, 1
N6-(3-Iodobenzyl)-9-(3-chloroacetylureido-2,5-di-O-t-butyldi-
methylsilyl-3-deoxy-â-D-ribofuranosyl)adenine (29b). 77% yield;
foam; 1H NMR (CDCl3) δ 8.39 (m, 2 H), 8.25 (s, 1 H), 8.15 (s, 1
H), 7.60 (s, 1 H), 7.46 (d, 1 H, J ) 8.0 Hz), 7.20 (d, 1 H, J ) 7.7
Hz), 6.90 (t, 1 H, J ) 7.9 Hz), 6.02 (brs, 1 H), 5.95 (d, 1 H, J )
2.2 Hz), 4.69 (brs, 2 H), 4.48-4.55 (m, 2 H), 4.12 (m, 1 H), 3.99
(s, 2 H), 3.91 (dd, 1 H, J ) 2.2, 11.7 Hz), 3.72 (dd, 1 H, J ) 2.7,
11.7 Hz), 0.82 (s, 9 H), 0.75 (s, 9 H), 0.00 (s, 6 H), -0.04 (s, 3 H),
-0.10 (s, 3 H); IR (KBr): 3299, 2932, 1701, 1617, 1473, 1254,
1126, 838, 781 cm-1; FAB-MS m/z 831 [M + H]+.
H, J ) 7.9 Hz), 7.20 (d, 1 H, J ) 7.7 Hz), 6.92 (t, 1 H, J ) 7.7
Hz), 6.19 (brs, 1 H), 5.85 (d, 1 H, J ) 3.1 Hz), 4.89 (d, 1 H, J )
6.9 Hz), 4.66 (s, 4 H), 4.46 (m, 1 H), 4.21 (dd, 1 H, J ) 6.8, 12.5
Hz), 4.04 (m, 1 H), 3.90 (dd, 1 H, J ) 2.2, 11.7 Hz), 3.73 (dd, 1
H, J ) 2.2, 11.7 Hz), 0.80 (s, 9 H), 0.77 (s, 9 H), 0.00 (s, 6 H),
-0.02 (s, 3 H), -0.05 (s, 3 H); IR (KBr): 3307, 2931, 1671, 1617,
1468, 1311, 1256, 1125, 837, 781 cm-1; FAB-MS m/z 788 [M +
H]+. Anal. (C30H47ClIN7O4Si2) C, H, N.
3-Ureido-5-(6-methylaminopurin-9-yl)-4-t-butyldimethylsiloxy-
tetrahydro-furan-2-carboxylic Acid Methyl Amide (40a). 72%
yield; oil; 1H NMR (CDCl3) δ 8.69 (brs, 1 H), 8.37 (s, 1 H), 7.79
(s, 1 H), 5.95 (brs, 1 H), 5.78 (d, 1 H, J ) 7.2 Hz), 5.37 (brs, 1 H),
4.95 (t, 1 H, J ) 7.4 Hz), 4.74 (d, 1 H, J ) 2.1 Hz), 4.05 (m, 1 H),
3.23 (brs, 3 H), 2.95 (d, 1 H, J ) 5.0 Hz), 0.78 (s, 9 H), -0.13 (s,
2-Chloro-N6-(3-iodobenzyl)-9-(3-chloroacetylureido-2,5-di-O-
t-butyldimethylsilyl-3-deoxy-â-D-ribofuranosyl)adenine (29c).
1
81% % yield; foam; H NMR (CDCl3) δ 8.38 (s, 1 H), 8.31 (s, 1
H), 8.19 (s, 1 H), 7.59 (s, 1 H), 7.47 (d, 1 H, J ) 7.7 Hz), 7.19 (d,

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2699
3 H), -0.32 (s, 3 H); IR (KBr): 3301, 2933, 1668, 1625, 1377,
1252, 1116, 840, 755 cm-1; FAB-MS m/z 465 [M + H]+. Anal.
(C19H32N8O4Si) C, H, N.
1 H, J ) 4.4 Hz), 6.22 (d, 1 H, J ) 6.7 Hz), 6.02 (d, 1 H, J ) 2.2
Hz), 5.76 (s, 2 H), 4.65 (br s, 2 H), 4.41 (m, 2 H), 4.23 (d, 1 H, J
) 6.0 Hz), 2.63 (d, 1 H, J ) 4.6 Hz); 13C NMR (CD3OD) δ 14.1,
20.9, 24.9, 26.4, 56.7, 59.7, 75.1, 83.6, 92.3, 128.0, 131.6, 137.5,
137.7, 141.3, 143.3, 154.2, 156.2, 161.7, 173.0; IR (KBr): 3424,
2932, 1619, 1476, 1338, 1056, 645 cm-1; FAB-MS m/z 553 [M +
H]+. Anal. (C19H21IN8O4) C, H, N.
Numbering Scheme of GPCRs. For sequence alignments of
selected regions of the A3AR and other GPCRs, a standardized
numbering system29 was used to identify residues in the TMs of
various receptors. Each residue is identified by two numbers: the
first corresponds to the TM in which it is located; the second
indicates its position relative to the most conserved residue in that
helix, arbitrarily assigned to 50. For example, Thr3.36 is the
threonine in TM3 (Thr94), located 14 residues before the most
conserved arginine Arg3.50; His7.43 corresponds to His272.
Biological Methods. Materials. The vector pcDNA3 was
obtained from Invitrogen. Oligonucleotides used were synthesized
by Bioserve Biotechnologies (Laurel, MD). Adenosine deaminase,
2-chloroadenosine, and NECA (5′-N-ethylcarboxamidoadenosine)
were obtained from Sigma (St. Louis, MO). [125I]I-AB-MECA
(2000 Ci/mmol) was from Amersham Pharmacia Biotech (Buck-
inghamshire, UK). All other compounds were obtained from
standard commercial sources and were of analytical grade.
Site-directed Mutagenesis. The protocols used were as described
in the QuikChange Site-directed Mutagenesis Kit (Stratagene, La
Jolla, CA). Mutations were confirmed by DNA sequencing.
Transfection of Wild-Type and Mutant A3AR to COS-7 Cells.
Lipofectamine 2000 (Invitrogen, Carlsbad, CA) was used for
transfection of wild-type and mutant receptor cDNA to COS-7 cells
following the manufacturer’s protocol.
Membrane Preparation. After 48 h of transfection, COS-7 cells
were harvested and homogenized with a Polytron homogenizer. The
homogenates were centrifuged at 20 000g for 20 min, and the
resulting pellet was resuspended in the 50 mM Tris‚HCl buffer
(pH 7.4) in the presence of 3 Units/mL adenosine deaminase and
incubated at 37°C for 30 min, and then stored at -80°C in aliquots.
The protein concentration was determined by using the method of
Bradford.30
Radioligand Binding Assays and Inositol Phosphate Deter-
mination in COS-7 Cells. The procedures for A3AR-binding
experiments using [125I]I-AB-MECA ([125I]N6-(4-aminobenzyl)-5′-
N-methylcarboxamidoadenosine) were as previously described.41
The procedures for binding at human A1 and A2AARs and a
functional assay of cyclic AMP formation mediated by the human
A2BAR expressed in CHO cells was as described.36 The method
for PLC determination in transiently transfected COS-7 cells has
been previously described.31
Preparation of Cardiac Myocyte Model of Simulated Is-
chemia and Gene Transfer into Cardiac Myocytes. Ventricular
cells were cultured from chick embryos 14 d in ovo and maintained
in culture as previously described.27 All experiments were performed
on day 3 in culture, at which time the medium was changed to a
HEPES (N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid)-
buffered medium containing 139 mM NaCl, 4.7 mM KCl, 0.5 mM
MgCl2, 0.9 mM CaCl2, 5 mM HEPES, and 2% fetal bovine serum
(pH 7.4, 37 °C). Myocytes were then exposed to simulated ischemia,
which was induced by 90 min of hypoxia and glucose deprivation
in a hypoxic incubator (NuAire, Plymouth, MN) where O2 was
replaced by N2 as previously described.27 The extent of myocyte
injury was determined at the end of the 90-min ischemia, at which
time myocytes were taken out of the hypoxic incubator and
reexposed to room air (normal percentage O2), followed by
quantitation of the number of viable cells. Viable cells were also
determined by the ability to exclude Trypan blue.27 Measurement
of the basal level of cell injury was made after parallel incubation
of control cells under a normal percentage of O2. The extent of
ischemia-induced injury was quantitatively determined by the
percentage of cells killed, according to a previously described
method.27 The percentage of cells killed was calculated as the
3-Ureido-5-(6-(3-iodobenzylamino)purin-9-yl)-4-t-butyldi-
methylsiloxy-tetrahydro-furan-2-carboxylic Acid Methyl Amide
1
(40b). 72% yield; oil; H NMR (CDCl3) δ 8.58 (s, 1 H), 8.37 (s,
1 H), 7.83 (s, 1 H), 7.63 (d, 1 H, J ) 7.9 Hz), 7.30 (d, 1 H, J )
7.8 Hz), 7.05 (t, 1 H, J ) 7.7 Hz), 6.33 (s, 1 H), 5.81 (d, 1 H, J )
7.0 Hz), 5.52 (s, 2 H), 4.72-4.95 (m, 4 H), 4.36 (d, 1 H, J ) 5.9
Hz), 2.94 (d, 3 H, J ) 5.5 Hz), 0.78 (s, 9 H), -0.11 (s, 3 H),
-0.39 (s, 3 H); IR (KBr): 3276, 2930, 1680, 1617, 1435, 1180,
1119, 723, 694 cm-1; FAB-MS m/z 667 [M + H]+. Anal. (C25H35-
IN8O4Si) C, H, N.
General Procedure for the Synthesis of 13, 10, 11, 14, and
12. To a stirred solution of 30a-c and 40a,b in THF (10 mL) was
added 1 M TBAF in THF (4 equiv) at room temperature. The
reaction mixture was stirred for 4 h at room temperature and
evaporated. The residue was purified by silica gel column chro-
matography to give 13, 10, 11, 14, and 12.
N6-Methyl-9-(3-deoxy-3-ureido-â-D-ribofuranosyl)adenine (13).
88% yield; white solid; mp 191-194 °C; [R]25 -42.3°(c 0.13,
D
1
DMF); H NMR (DMSO-d6) δ 8.37 (s, 1 H), 8.25 (s, 1 H), 7.79
(brs, 1 H), 6.24 (s, 1 H), 6.15 (d, 1 H, J ) 7.5 Hz), 5.91 (d, 1 H,
J ) 2.6 Hz), 5.74 (brs, 2 H), 5.13 (t, 1 H, J ) 5.1 Hz), 4.41 (dd,
1 H, J ) 2.6, 5.7 Hz), 4.27 (m, 1 H), 3.86 (m, 1 H), 3.68 (m, 1 H),
3.49 (dd, 1 H, J ) 4.6, 12.5 Hz), 2.94 (s, 3 H); 13C NMR (DMSO-
d6) δ 22.6, 50.4, 60.8, 73.1, 83.1, 89.4, 119.6, 139.0, 147.9, 153.0,
155.0, 170.0; IR (KBr): 3422, 1633, 1543, 1382, 1335, 1221, 1102,
1065, 521 cm-1; FAB-MS m/z 324 [M + H]+. Anal. (C12H17N7O4)
C, H, N.
N6-(3-Iodobenzyl)-9-(3-deoxy-3-ureido-â-D-ribofuranosyl)ad-
enine (10). 70% yield; white solid; mp 165-168 °C; [R]25D -43.3°
1
(c 0.12, DMF); H NMR (DMSO-d6) δ 8.44 (brs, 1 H), 8.43 (s, 1
H), 8.22 (s, 1 H), 7.71 (s, 1 H), 7.55 (d, 1 H, J ) 7.7 Hz), 7.33 (d,
1 H, J ) 7.9 Hz), 7.06 (t, 1 H, J ) 7.7 Hz), 6.22 (d, 1 H, J ) 5.0
Hz), 6.15 (d, 1 H, J ) 7.7 Hz), 5.93 (d, 1 H, J ) 2.3 Hz), 5.78
(brs, 2 H), 5.11 (t, 1 H, J ) 5.7 Hz), 4.66 (brs, 2 H), 4.42 (m, 1
H), 4.25 (m, 1 H), 3.87 (m, 1 H), 3.69 (m, 1 H), 3.50 (m, 1 H); 13
C
NMR (DMSO-d6) δ 42.2, 50.9, 61.0, 73.4, 84.2 89.5, 94.7, 119.5,
126.7, 130.5, 135.4, 135.7, 139.2, 142.9, 148.3, 152.5, 154.3, 158.7;
IR (KBr): 3398, 1621, 1538, 1477, 1338, 1221, 1103, 822 cm-1
;
FAB-MS m/z 526 [M + H]+. Anal. (C18H20IN7O4) C, H, N.
{5-[2-Chloro-6-(3-iodo-benzylamino)-purin-9-yl]-4-hydroxy-
2-hydroxymethyl-tetrahydro-furan-3-yl}urea (11). 82% yield;
1
white solid; mp 132.3-135 °C; [R]25 -27.4°(c 0.35, DMF); H
D
NMR (DMSO-d6) δ 8.92 (brs, 1 H), 8.47 (s, 1 H), 7.73 (s, 1 H),
7.58 (d, 1 H, J ) 7.9 Hz), 7.33 (d, 1 H, J ) 7.5 Hz), 7.09 (t, 1 H,
J ) 7.7 Hz), 6.25 (d, 1 H, J ) 4.7 Hz), 6.16 (d, 1 H, J ) 7.7 Hz),
5.87 (d, 1 H, J ) 2.0 Hz), 5.77 (s, 2 H), 5.05 (t, 1 H, J ) 5.5 Hz),
4.58 (d, 2 H, J ) 5.0 Hz), 4.34 (m, 1 H), 3.86 (m, 1 H), 3.39 (m,
1 H), 3.51 (m, 1 H); 13C NMR (DMSO-d6) δ 42.5, 50.7, 60.7, 73.6,
84.1, 89.2, 94.7, 118.5, 126.8, 130.6, 135.6, 136.1, 139.4, 141.9,
149.3, 153.1, 154.8, 158.6; IR (KBr): 3405, 1619, 1346, 1312,
1221, 1105, 781, 633 cm-1; FAB-MS m/z 560 [M + H]+. Anal.
(C18H19ClIN7O4) C, H, N.
3-Ureido-5-(6-methylaminopurin-9-yl)-4-hydroxy-tetrahydro-
furan-2-carboxylic Acid Methyl Amide (14). 66% yield; white
solid; mp 118.2-120.0 °C; [R]25D -13.3° (c 0.15, MeOH); 1H NMR
(CD3OD): δ 8.48 (s, 1 H), 8.29 (s, 1 H), 6.07 (d, 1 H, J ) 3.1
Hz), 4.61 (m, 2 H), 4.41 (d, 1 H, J ) 6.0 Hz), 3.11 (s, 3 H), 2.80
(s, 3 H); 13C NMR (CD3OD) δ 14.1, 20.9, 24.2, 24.9, 26.4, 56.7,
59.7, 74.9, 83.7, 92.2, 141.0, 154.2, 173.0; IR (KBr): 3425, 1667,
1630, 1534, 1356, 1306, 1084, 936, 636 cm-1; FAB-MS m/z 351
[M + H]+. Anal. (C13H18N8O4) C, H, N.
3-Ureido-5-(6-(3-iodobenzylamino)purin-9-yl)-4-hydroxy-tet-
rahydro-furan-2-carboxylic Acid Methyl Amide (12). 82% yield;
white solid; mp: 120.7-122.2 °C; [R]25 -10.0° (c 0.10, DMF);
D
1H NMR (DMSO-d6): δ 8.69 (s, 1 H), 8.53 (brs, 1 H), 8.33 (d, 1
H, J ) 4.6 Hz), 8.23 (s, 1 H), 7.72 (s, 1 H), 7.55 (d, 1 H, J ) 7.7
Hz), 7.34 (d, 1 H, J ) 7.7 Hz), 7.06 (t, 1 H, J ) 7.7 Hz), 6.29 (d,

2700 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
number of cells obtained from the control group (representing cells
not subjected to hypoxia or drug treatment) minus the number of
cells from the treatment group divided by the number of cells in
the control group multiplied by 100%.
Cells were then transfected with either the vector pcDNA3 or
cDNA encoding the wild-type human A3AR or the neoceptor
H272E, with the use of FuGENE 6 as previously described.28 Forty-
eight hours after the transfection, myocytes were exposed to
simulated ischemia, and myocyte injury was determined in the
presence or absence of the neoligand 10. In other experiments, the
PLD activity was determined 48 h after the transfection of myocytes
with either the wild-type hA3AR or the neoceptor H272E cDNA.
Measurement of PLD Activity. For measurement of PLD
activity, cultured ventricular myocytes labeled with [3H]myristate
(49 Ci/mmol, 2 µCi/ml) for 24 h were exposed to receptor agonist
in the presence of 0.5% (v/v) ethanol. Lipids were extracted by
the method of Bligh and Dyer.32 The formation of [3H]PEt was an
indication of PLD activity. Quantitation of PEt (phosphatidyletha-
nol) was carried out as previously described.33 Briefly, the formation
of [3H]PEt in cells was quantitated by separation of the labeled
lipids via thin-layer chromatography and scintillation counting of
the 3H label. The position of PEt was determined visually by placing
the thin-layer plate in an iodine chamber, and its level was expressed
as a percentage of total lipids. Data were also expressed as
percentage increases in the amount of PEt relative to that for
unstimulated cells.
the ligand. H-bonding sites were marked for the acidic residue,
E272, of the neoceptor and for the 3′-ureido groups of the
neoligands, which were able to act as H-bond donor or acceptor.
The lowest energy conformer of neoligand was variously prepo-
sitioned in the putative binding cavity as a starting point for
FlexiDock, based on the reported point-mutational results. Flexi-
Dock parameters were set at 30 000-generation for genetic algo-
rithms. To increase the binding interaction, the torsion angles of
the side chains that directly interacted within 5 Å of the ligands,
according to the results of FlexiDock, were manually adjusted. The
atom types of all ligands were manually assigned with an Amber
all-atom force field. Finally, the complex structure was minimized
using an Amber force field with a fixed dielectric constant (4.0),
until the conjugate gradient reached 0.1 kcal‚mol-1‚Å-1
.
FlexX Docking. FlexX 1.1357 is a fast docking method that uses
a new algorithmic approach based on a pattern recognition technique
called pose clustering, allowing conformational flexibility of the
ligand by MINUMBA58 conformer library to grow ligands during
the docking process. The free binding energy of complex including
H-bond, ionic, aromatic, or lipophilic interactions was estimated
by the scoring function. Cscore calculations were included for
scoring. A putative binding site including T94 (3.36), N250 (6.55),
S271 (7.42), and H272 (7.43) was manually selected, based on the
previous point-mutational results.41 Formal charges were applied
to the ligands. All default parameters, as implemented in the 6.9
release of SYBYL, were used.
Statistical Analysis. Binding and functional parameters were
estimated with GraphPad Prism software (GraphPad, San Diego,
CA). IC50 values obtained from competition curves were converted
to Ki values by using the Cheng-Prusoff equation.34 Data were
expressed as mean ( standard error.
Computational Methods. All calculations were performed using
the SYBYL program version 6.953 on a Silicon Graphics Octane
workstation (300 MHz MIPS R12000 (IP30) processor).
Acknowledgment. We thank Dr. Michihiro Ohno for helpful
discussion. This research was supported by a grant from the
Korea Science and Engineering Foundation (R01-2005-000-
10162-0). Dr. Philippe Van Rompaey is a recipient of an IWT
fellowship. This research was supported in part by the Intramural
Research Program of the NIH, National Institute of Diabetes
and Digestive and Kidney Diseases, and R01-HL48225 to BTL.
Conformational Search. The N6-I-benzyl-3′-ureido compound,
10, was constructed using the “Sketch Molecule” and subjected to
a random search performed for all rotatable bonds. The options of
the random search consisted of 3000 iterations, 3 kcal energy cutoff,
and chirality checking. MMFF force field54 and charge were applied
using distance-dependent dielectric constants and the conjugate
gradient method until the gradient reached 0.05 kcal/mol/Å. After
clustering the low energy conformers from the result of the
conformational search, the relative stabilities of various representa-
tive conformers from all groups were checked by semiempirical
molecular orbital calculations using the PM3 method in the MOPAC
6.0 package.43 During the PM3 optimization, all abortive results
from electrostatic collapse were removed.
Molecular Dynamics of the H272E Neoceptors. The previously
published hA3 AR model (PDB code: 1o74) built by homology
modeling55 from the X-ray structure of bovine rhodopsin with 2.8
Å resolution35 was used for the docking study. For the side-chain
refinement of the H272E neoceptors, the optimized structures were
then used as the starting point for subsequent 50-ps MD, during
which the protein backbone atoms in the secondary structures were
constrained as in the previous step. The options of MD at 300 K
with a 0.2-ps coupling constant were a time step of 1fs and a
nonbonded update every 25 fs. The SHAKE algorithm56 was
employed to fix the lengths of bonds to hydrogen atoms. The
average structure from the last 10-ps trajectory of MD was
optimized with backbone constraints in the secondary structure and
then the unconstrained structure was minimized as described above.
Supporting Information Available: Elemental analysis data
for unknown compounds and 3D coordinates of two modes of
docking compound 10 in homology models of the H272E mutant
human A3 receptors in PDB format (as represented in Figure 3).
This material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Buskirk, A. R.; Liu, D. R. Creating small-molecule-dependent
switches to modulate biological functions. Chem. Biol. 2005, 12,
151-161.
(2) Hwang, Y. W.; Miller, D. L. A mutation that alters the nucleotide
specificity of elongation factor Tu, a GTP regulatory protein. J. Biol.
Chem. 1987, 262, 13081-13085.
(3) Niculescu-Duvaz, I.; Friedlos, F.; Niculescu-Duvaz, D.; Davies, L.;
Springer, C. J. Prodrugs for antibody- and gene-directed enzyme
prodrug therapies (ADEPT and GDEPT). Anticancer Drug Des. 1999,
14, 517-538.
(4) Hashimoto, A.; Shi, Y.; Drake, K.; Koh, J. T. Design and synthesis
of complementing ligands for mutant thyroid hormone receptor
TRbeta(R320H): a tailor-made approach toward the treatment of
resistance to thyroid hormone. Bioorg. Med. Chem. 2005, 13, 3627-
3639.
(5) Strausberg, R. L.; Schreiber, S. L. From knowing to controlling: a
path from genomics to drugs using small molecule probes. Science
2003, 300, 294-295.
(6) Koh, J. T.; Biggins, J. B. Ligand-receptor engineering and its
application towards the complementation of genetic disease and target
identification. Curr. Top. Med. Chem. 2005, 5, 413-420.
(7) Shi, Y.; Koh, J. T. Functionally orthogonal ligand-receptor pairs
for the selective regulation of gene expression generated by manipu-
lation of charged residues at the ligand-receptor interface of ER
alpha and ER beta. J. Am. Chem. Soc. 2002, 124, 6921-6928.
(8) Tedesco, R.; Thomas, J. A.; Katzenellenbogen, B. S.; Katzenellen-
bogen, J. A. The estrogen receptor: a structure-based approach to
the design of new specific hormone-receptor combinations. Chem.
Biol. 2001, 8, 277-287.
FlexiDock Docking. Compound 10 was docked within the hA3
AR and H272E neoceptors. Flexible docking was facilitated through
the FlexiDock utility in the Biopolymer module of SYBYL 6.9.
Flexible docking allowed for flexibility of all rotatable bonds in
10, except those of the ribose ring, and the side chains of
surrounding hydrophilic residues (T94, Q167, N250, E272) in the
putative binding site of Cl-IB-MECA. After the hydrogen atoms
were added to the receptor, atomic charges were recalculated by
using Kollman All-atom for the protein and Gasteiger-Hu¨ckel for
(9) Liu, Y.; Shah, K.; Yang, F.; Witucki, L.; Shokat, K. M. Engineering
Src family protein kinases with unnatural nucleotide specificity.
Chem. Biol. 1998, 5, 91-101

Reengineered A3 Adenosine Receptor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9 2701
(10) Lin, Q.; Jiang, F.; Schultz, P. G.; Gray, N. S. Design of allele-specific
protein methyltransferase inhibitors. J. Am. Chem. Soc. 2001, 123,
11608-11613.
(11) Strader, C. D.; Sigal, I. S.; Register, R. B.; Candelore, M. R.; Rands,
E.; Dixon, R. A. Identification of residues required for ligand binding
to the beta-adrenergic receptor. Proc. Natl. Acad. Sci. U.S.A. 1987,
84, 4384-4388.
(12) Strader, C. D.; Gaffney, T.; Sugg, E. E.; Candelore, M. R.; Keys,
R.; Patchett, A. A.; Dixon, R. A. Allele-specific activation of
genetically engineered receptors. J. Biol. Chem. 1991, 266, 5-8.
(13) Small, K. M.; Wagoner, L. E.; Levin, A. M.; Kardia, S. L.; Liggett,
S. B. Synergistic polymorphisms of beta1- and alpha2C-adrenergic
receptors and the risk of congestive heart failure. N. Engl. J. Med.
2002, 347, 1135-1142.
(14) Coward, P.; Wada, H. G.; Falk, M. S.; Chan, S. D.; Meng, F.; Akil,
H.; Conklin, B. R. Controlling signaling with a specifically designed
Gi-coupled receptor. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 352-
357.
(15) Redfern, C. H.; Coward, P.; Degtyarev, M. Y.; Lee, E. K.; Kwa, A.
T.; Hennighausen, L.; Bujard, H.; Fishman, G. I.; Conklin, B. R.
Conditional expression and signaling of a specifically designed Gi-
coupled receptor in transgenic mice. Nat. Biotechnol. 1999, 17, 165-
169.
(16) Scearce-Levie, K.; Lieberman, M. D.; Elliott, H. H.; Conklin, B. R.
Engineered G protein coupled receptors reveal independent regulation
of internalization, desensitization and acute signaling. BMC Biol.
2005, 3, 3.
(17) Claeysen, S.; Joubert, L.; Sebben, M.; Bockaert, J.; Dumuis, A. A
single mutation in the 5-HT4 receptor (5-HT4-R D100(3.32)A)
generates a Gs-coupled receptor activated exclusively by synthetic
ligands (RASSL). J. Biol. Chem. 2003. 278, 699-702.
(18) Pauwels, P. J. Unravelling multiple ligand-activation binding sites
using RASSL receptors. Trends Pharmacol Sci. 2003, 24, 504-507.
(19) Zhao, G. Q.; Zhang, Y.; Hoon, M. A.; Chandrashekar, J.; Erlenbach,
I.; Ryba, N. J.; Zuker, C. S. The receptors for mammalian sweet and
umami taste. Cell 2003, 115, 255-266.
(20) Bruysters, M.; Jongejan, A.; Akdemir, A.; Bakker, RA.; Leurs, R.
Mutant histamine H1 receptor Phe435Ala: A Gq/11-coupled receptor
activated solely by synthetic ligands (RASSL). J. Biol. Chem. 2005,
Jul 18; [Epub ahead of print].
(21) Jacobson, K. A.; Gao, Z. G.; Chen, A.; Barak, D.; Kim, S. A.; Lee,
K.; Link, A.; Van Rompaey, P.; Van Calenbergh, S.; Liang, B. T.
Neoceptor concept based on molecular complementarity in GPCRs:
A mutant adenosine A3 receptor with selectively enhanced affinity
for amine-modified nucleosides. J. Med. Chem. 2001, 44, 4125-
4136.
(22) Liang, B. T.; Jacobson, K. A. A physiological role of the adenosine
A3 receptor: sustained cardioprotection. Proc. Natl. Acad. Sci. U.S.A.
1998, 95, 6995-6999.
(23) Fredholm, B. B.; Chen, J. F.; Masino, S. A.; Vaugeois, J. M. Actions
of adenosine at its receptors in the CNS: insights from knockouts
and drugs. Annu. ReV. Pharmacol. Toxicol. 2005, 45, 385-412.
(24) Fishman, P.; Bar-Yehuda, S.; Madi, L.; Cohn, I. A3 adenosine receptor
as a target for cancer therapy. Anticancer Drugs 2002, 13, 437-
443.
(25) Kim, S. K.; Gao, ZG.; Van Rompaey, P.; Gross, A. S.; Chen, A.;
Van Calenbergh, S.; Jacobson, K. A. Modeling the adenosine
receptors: Comparison of binding domains of A2A agonist and
antagonist. J. Med. Chem. 2003, 46, 4847-4859.
(26) Jacobson, K. A.; Ohno, M.; Duong, H. T.; Kim, S. K.; Tchilibon,
S.; Cesnek, M.; Holy, A.; Gao Z. G. A neoceptor approach to
unraveling microscopic interactions between the human A2A adeno-
sine receptor and its agonists. Chem. Biol. 2005, 12, 237-247.
(27) Strickler, J.; Jacobson, K. A.; Liang, B. T. Direct preconditioning of
cultured chick ventricular myocytes: novel functions of cardiac
adenosine A2A and A3 receptors. J. Clin. InVest. 1996, 98, 1773-
1779.
(28) Mozzicato, S.; Joshi, B. V.; Jacobson, K. A.; Liang, B. T. Role of
direct RhoA-phospholipase D interaction in mediating adenosine-
induced protection from cardiac ischemia. FASEB J. 2004, 18, 406-
408.
(29) Visiers, I.; Ballesteros, J. A.; Weinstein, H. Three-dimensional
representations of G protein-coupled receptor structures and mech-
anisms. Methods Enzymol. 2002, 343, 329-371.
(30) Bradford, M. M. A rapid and sensitive method for the quantitation
of microgram quantities of protein utilizing the principle of protein-
dye binding. Anal. Biochem. 1976, 72, 248-254.
(32) Bligh, E. G.; Dyer, W. J. A rapid method of total lipid extraction
and purification. Can. J. Biochem. Physiol. 1959, 37, 911-917.
(33) Lee, J. E.; Bokoch, G.; Liang, B. T. Signaling mechanism of the
adenosine A3 receptor: Novel cardioprotective role of RhoA. FASEB
J. 2001, 15, 1886-1894.
(34) Cheng, Y.; Prusoff, W. H. Relationship between the inhibition
constant (Ki) and the concentration of inhibitor which causes 50%
inhibition (I50) of an enzymatic reaction. Biochem. Pharmacol. 1973,
22, 3099-3108.
(35) Palczewski, K.; Kumasaka, T.; Hori, T.; Behnke, CA.; Motoshima,
H.; Fox, B. A.; Le Trong, I.; Teller, D. C.; Okada, T.; Stenkamp, T.
E.; Yamamoto, M.; Miyano, M. Crystal structure of rhodopsin: a G
protein-coupled receptor. Science 2000, 289, 739-745.
(36) Ohno, M.; Gao, Z. G.; Van Rompaey, P.; Tchilibon, S.; Kim, S. K.;
Harris, B. A.; Blaustein, J.; Gross, A. S.; Duong, H. T.; Van
Calenbergh, S.; Jacobson, K. A. Modulation of adenosine receptor
affinity and intrinsic efficacy in nucleosides substituted at the
2-position. Bioorg. Med. Chem. 2004, 12, 2995-3007.
(37) Van Rompaey, P.; Jacobson, K. A.; Gross, A. S.; Gao, Z. G.; Van
Calenbergh, S. Exploring human adenosine A3 receptor complemen-
tarity and activity for adenosine analogues modified in the ribose
and purine moiety. Bioorg. Med. Chem. 2005, 13, 973-983.
(38) Jeong, L. S.; Kim, M. J.; Kim, H. O.; Kim, S. K.; Gao, Z. G.;
Jacobson, K. A.; Chun, M. W. Design and synthesis of 3′-
ureidoadenosine-5′-uronamides: Effects of the 3′-ureido group on
binding to the A3 adenosine receptor. Bioorg. Med. Chem. Lett. 2004,
14, 4851-4854.
(39) Jeong, L. S.; Kim, M. J.; Kim, A. Y.; Lee, J. A.; Jacobson, K. A.;
Gao, Z. G.; Kim, S. K.; Chun, M. W. Presented at the 229th National
Meeting of the American Chemical Society, Abstract MEDI82, San
Diego, CA, March 13-17, 2005.
(40) Dougherty, C.; Barucha, J.; Schofield, P.; Jacobson, K. A.; Liang,
B. T. Cardiac myocytes rendered ischemia resistant by overexpressing
the human adenosine A3 receptor. FASEB J. 1998, 12, 1785-1792.
(41) Gao, Z. G.; Kim, S. K.; Gross, A. S.; Chen, A.; Blaustein, J.;
Jacobson, K. A. Identification of essential residues involved in the
allosteric modulation of the human A3 adenosine receptor. Mol.
Pharmacol. 2003, 63, 1021-1031.
(42) Costanzi, S.; Lambertucci, C.; Vittori, S.; Volpini, R.; Cristalli, G.
2- and 8-alkynyladenosines: conformational studies and docking to
human adenosine A3 receptor can explain their different biological
behavior. J. Mol. Graphics Modell. 2003, 21, 253-262.
(43) Stewart, J. J. P. MOPAC: A Semiempirical Molecular Orbital
Program J. Comput.-Aided Mol. Des. 1990, 4, 1-105.
(44) Spiegel, A. M.; Weinstein, L. S. Inherited diseases involving G
proteins and G protein-coupled receptors. Annu. ReV. Med. 2004,
55, 27-39.
(45) Iida, A.; Saito, S.; Sekine, A.; Kataoka, Y.; Tabei, W.; Nakamura,
Y. Catalog of 300 SNPs in 23 genes encoding G-protein coupled
receptors. J. Hum. Genet. 2004, 49, 194-208.
(46) Lam, P.; Hong, C. J.; Tsai, S. J. Association study of A2A adenosine
receptor genetic polymorphism in panic disorder. Neurosci. Lett.
2005, 378, 98-101.
(47) Alsene, K.; Deckert, J.; Sand, P.; de Wit, H. Association between
A2a receptor gene polymorphisms and caffeine-induced anxiety.
Neuropsychopharmacology 2003, 28, 1694-702.
(48) Koh, J. T. Designing selectivity and discrimination into re-engineered
ligand-receptor interfaces; Beyond bumps and holes. Chem. Biol.
2002, 9, 17-23.
(49) Belshaw, P. J.; Schoepfer, J. G.; Liu, K.-Q.; Morrison, K. L.;
Schreiber, S. L. Rational design of orthogonal receptor ligand
combinations. Angew. Chem., Int. Ed. Engl. 1995 34, 2129-2132.
(50) Gillespie, P. G.; Gillespie, S. K.; Mercer, J. A.; Shah, K.; Shokat,
K. M. Engineering of the myosin-ibeta nucleotide-binding pocket to
create selective sensitivity to N6-modified ADP analogs. J. Biol.
Chem. 1999, 274, 31373-31381.
(51) Ulrich, S. M.; Buzko, O.; Shah, K.; Shokat, K. M. Towards the
engineering of an orthogonal protein kinase/nucleotide triphosphate
pair. Tetrahedron 2000, 56, 9495-9502.
(52) Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz, K. N.;
Linden, J. International Union of Pharmacology. XXV. Nomenclature
and classification of adenosine receptors. Pharmacol. ReV. 2001, 53,
527-552.
(53) Sybyl Molecular Modeling System, version 6.9; Tripos Inc., St. Louis,
MO 63144.
(54) Halgren, T. A. MMFF VII. Characterization of MMFF94, MMFF94s,
and Other Widely Available Force Fields for Conformational Energies
and for Intermolecular-Interaction Energies and Geometries. J.
Comput. Chem. 1999, 20, 730-748.
(31) Gao, Z. G.; Chen, A.; Barak, D.; Kim, S. K.; Mu¨ller, C. E.; Jacobson,
K. A. Identification by site-directed mutagenesis of residues involved
in ligand recognition and activation of the human A3 adenosine
receptor. J. Biol. Chem. 2002, 277, 19056-19063.

2702 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 9
Gao et al.
(55) Gao, Z.-G.; Kim, S.-K.; Biadatti, T.; Chen, W.; Lee, K.; Barak, D.;
Kim, S. G.; Johnson, C. R.; Jacobson, K. A. Structural Determination
of A3 Adenosine Receptor Activation: Nucleoside Ligands at the
Agonist/Antagonist Boundary. J. Med. Chem. 2002, 45, 4471-4484.
(56) Ryckaert, J. P.; Ciccotti, G.; Berendsen, H. J. C. Numerical Integration
of the Cartesian Equations of Motion for a System with Constraints:
Molecular Dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327-
341.
(57) Rarey, M.; Kramer, B.; Lengauer, T.; Klebe, G. A fast flexible
docking method using an incremental construction algorithm. J. Mol.
Biol. 1996, 261, 470-489.
(58) Klebe, G.; Mietzner, T. A fast and efficient method to generate
biologically relevant conformations. J. Comput.-Aided Mol. Des.
1994, 8, 583-606.
JM050968B