N6-cycloalkyl- And n6-bicycloalkyl-c5'(c 2')-modified adenosine derivatives as high-affinity and selective agonists at the human A1 adenosine receptor with antinociceptive effects in mice

Article abstract of DOI:10.1021/jm801456g

To further investigate new potent and selective human A1 adenosine receptor agonists, we have synthesized a series of 5'-chloro-5'-deoxy- and 5'-(2-fluorophenylthio)-5'-deoxy-N⁶-cycloalkyl(bicycloalkyl)- substituted adenosine and 2'-C-methylade

Full text of DOI:10.1021/jm801456g

J. Med. Chem. 2009, 52, 2393–2406
2393
N6-Cycloalkyl- and N6-Bicycloalkyl-C5′(C2′)-modified Adenosine Derivatives as High-Affinity
and Selective Agonists at the Human A1 Adenosine Receptor with Antinociceptive Effects in
Mice
Palmarisa Franchetti,† Loredana Cappellacci,† Patrizia Vita,† Riccardo Petrelli,† Antonio Lavecchia,‡ Sonja Kachler,§
Karl-Norbert Klotz,§ Ida Marabese,| Livio Luongo,| Sabatino Maione,| and Mario Grifantini*,†
Department of Chemical Sciences, UniVersity of Camerino, 62032 Camerino, Italy, Department of Pharmaceutical and Toxicological
Chemistry, UniVersity of Naples “Federico II”, 80131 Naples, Italy, Institut fu¨r Pharmakologie and Toxikologie, UniVersita¨t Wu¨rzburg,
D-97078 Wu¨rzburg, Germany, and Department of Experimental Medicine, Section of Pharmacology L. Donatelli, Second UniVersity of Naples,
80138 Naples, Italy
ReceiVed NoVember 18, 2008
To further investigate new potent and selective human A1 adenosine receptor agonists, we have synthesized
a series of 5′-chloro-5′-deoxy- and 5′-(2-fluorophenylthio)-5′-deoxy-N6-cycloalkyl(bicycloalkyl)-substituted
adenosine and 2′-C-methyladenosine derivatives. These compounds were evaluated for affinity and efficacy
at human A1, A2A, A2B, and A3 adenosine receptors. In the series of N6-cyclopentyl- and N6-(endo-norborn-
2-yl)adenosine derivatives, 5′-chloro-5′-deoxy-CPA (1) and 5′-chloro-5′-deoxy-(()-ENBA (3) displayed the
highest affinity in the subnanomolar range and relevant selectivity for hA1 vs the other human receptor
subtypes. The higher affinity and selectivity of 5′-chloro-5′-deoxyribonucleoside derivatives 1 and 3 for
hA1 AR vs hA3 AR compared to that of the parent 5′-hydroxy compounds CPA and (()-ENBA was
rationalized by a molecular modeling analysis. 5′-Chloro-5′-deoxy-(()-ENBA, evaluated for analgesic activity
in the formalin test in mice, was found to inhibit the first or the second phases of the nocifensive response
induced by intrapaw injection of formalin at doses ranging between 1 and 2 mg/kg i.p.
Introduction
N6-tetrahydrofuranyl-5′-(2-fluorophenylthio)-5′-deoxyadenos-
ine showed affinity and partial agonism at the A1 receptor in
DDT cell membranes (hamster vas deferens smooth muscle cell
line).5b Therefore, the 5′-hydroxyl group in adenosine analogues
does not appear to be essential for receptor binding and
activation of A1 AR; furthermore, 5′-modified selective A1
agonists could be more druggable than 5′-unmodified analogues,
since normal ribonucleosides may be phosphorylated by ad-
enosine kinases and in consecutive steps by nucleotide kinases
to 5′-mono-, 5′-di-, or 5′-triphosphate derivatives, respectively,
and subsequently interact with P2Y receptors and/or other
biological targets. Moreover, the replacement of the 5′-hydroxyl
group by a chlorine atom confers to these nucleosides greater
stability versus purine nucleoside metabolizing enzymes such
as adenosine deaminase and purine nucleoside phosphorylase.6
Based on these findings, in the search for new potent and
selective human A1 AR agonists, we synthesized a series of
5′-chloro- and 5′-(2-fluorophenylthio)-5′-deoxy derivatives of
the selective A1 AR agonists CPA,a CCPA, 2′-Me-CPA, 2′-
Me-CCPA, and N6-(()-endo-norborn-2-yl purine ribonucleoside
analogues that were evaluated for affinity and selectivity at all
cloned human adenosine receptor subtypes (Chart 1).
Adenosine mediates a wide variety of physiological effects
by activation of four G protein-coupled receptors (A1, A2A, A2B,
and A3 ARs) that are widely distributed throughout the body.
A number of agonists with high affinity for human A1, A2A,
and A3 adenosine receptors and more recently for the A2B
subtype have been developed for therapeutic applications, and
some are in clinical trials for various conditions.1 Effects
mediated by the selective activation of A1 AR include neuro-
and cardioprotection, an antiarrhythmic effect, reduction of
neuropathic pain, and reduction of lipolysis in adipose tissue.2
Some adenosine derivatives, such as A1 AR agonists, are in
clinical trials for treatment of cardiac arrhythmias and neuro-
pathic pain.1 However, the cardiovascular side effects and other
side effects induced by the A1 activation limit the clinical
applications of A1 agonists.
To address the problem of side effects, several purine
ribofuranoside derivatives have been investigated as A1 partial
agonists endowed with a potentially more favorable clinical
suitability.3 In order to identify highly selective agonists at A1
AR vs the other receptor subtypes, a wide range of adenosine
derivatives modified at the C2,N6-positions of the nucleobase
and/or at the ribose moiety have been reported. Between the
di- or trisubstituted adenosine derivatives, some N6-cycloalkyl
or bicycloalkyl derivatives and 5′-chloro-5′-deoxy analogues
were found to have high affinity and selectivity for rat A1 AR,4
while some N6-substituted-5′-alkylthio- or 5′-arylthio-analogues
proved to be partial agonists for this AR subtype.5 Among these,
Results and Discussion
Chemistry. The synthesis of compound 4 is outlined in
Scheme 1. Treatment of 2,6-dichloro-9H-(2,3,5-tri-O-acetyl-ꢀ-
a Abbreviations: CPA, N6-cyclopentyladenosine; CCPA, 2-chloro-N6-
cyclopentyladenosine; 2′-Me-CPA, 2′-C-methyl-N6-cyclopentyladenos-
ine; 2′-Me-CCPA, 2-chloro-2′-C-methyl-N6-cyclopentyladenosine;
(()-ENBA, N6-((-endo-norborn-2-yl)adenosine; 5′-Cl-CPA, 5′-chloro-
5′-deoxy-N6-cyclopentyladenosine; 5′-Cl-CCPA, 2,5′-dichloro-5′-deoxy-
N6-cyclopentyladenosine; NECA, 5′-N-ethylcarboxamidoadenosine;
HEMADO, 2-(hexyn-1-yl)-N6-methyladenosine; 5′Cl5′d-(()-ENBA, 5′-
chloro-5′-deoxy-N6-((-endo-norborn-2-yl)adenosine; DPCPX, 8-cyclo-
pentyl-1,3-dipropylxanthine; CHO, Chinese hamster ovary; TM, trans-
membrane helical domain; ꢀ2-AR, ꢀ2-adrenergic receptor.
* To whom correspondence should be addressed. Phone: +39-0737-
402233. Fax +39-0737-637345. E-mail: mario.grifantini@unicam.it.
†
University of Camerino.
University of Naples “Federico II”.
Universita¨t Wu¨rzburg.
Second University of Naples.
‡
§
|
10.1021/jm801456g CCC: $40.75
2009 American Chemical Society
Published on Web 03/24/2009

2394 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Chart 1. Chemical Structures of Compounds 1-14
Franchetti et al.
The synthesis of compounds 5-8 begins with the 6-chloro-
or 2,6-dichloro-9H-(2-C-methyl-2,3,5-tri-O-benzoyl-ꢀ-D-ribo-
furanosyl)purine (19 and 20, respectively),8 which was reacted
with cyclopentylamine or (()-endo-norborn-2-ylamine followed
by sugar deblocking in basic conditions (Scheme 2). Compounds
21-24 were converted to the corresponding 2′,3′-isopropylidene
derivatives 25-28. 5′-Chlorination of 25-28, followed by
deisopropylidenation of 29-32, gave the desired compounds
5-8. 5′-(2-Fluorophenylthio) derivatives 11-14 were prepared
by reaction of 29-32 with 2-fluorothiophenol in anhydrous
DMF in the presence of 60% sodium hydride, with removal of
the isopropylidene protecting group. 2-Chloro-N6-cyclopentyl-
5′-(2-fluorophenylthio)-5′-deoxyadenosine (10) was synthesized
in a similar way starting from the 2′,3′-O-isopropylidene
derivative of CCPA (38), which was prepared from CCPA9 (37)
(Scheme 3). Compounds 5′-Cl-CCPA (2), (()-ENBA, and
5′Cl5′d-(()-ENBA (3) were synthesized as reported in the
literature.4a,10 5′-Cl-CPA (1), previously reported by van der
Wenden et al.,5a N6-cyclopentyl-5′-(2-fluorophenylthio)-5′-
deoxyadenosine (9), and N6-tetrahydrofuranyl-5′-(2-fluorophe-
nylthio)-5′-deoxyadenosine (41), reported by Morrison et al.,5b
were also prepared following a different route, in order to
evaluate their affinity at human ARs (see the Supporting
Information).
Scheme 1a
Binding Studies. The new compounds were evaluated at the
human recombinant adenosine receptors, stably transfected into
Chinese hamster ovary (CHO) cells, utilizing radioligand
binding assays (A1, A2A, and A3) or the adenylyl cyclase activity
assay (A2B).3a,11,12 [3H]CCPA, [3H]NECA, and [3H]HEMADO
were used as radioligands for human A1, A2A, and A3 ARs,
respectively. In the case of the A2B receptor subtype, Ki values
were calculated from IC50 values determined by inhibition of
NECA-stimulated adenylyl cyclase activity.11
Adenosine derivatives 1-3, 9, and 41 were assayed at the
human recombinant adenosine receptors, since their affinities
at human ARs have not been reported so far. The affinity and
selectivity of the compounds were compared to those of the
reference compounds CPA, CCPA, 2′-Me-CPA, 2′-Me-CCPA,
and (()-ENBA (Table 1).
Among the tested compounds, only 1, 2, 3, and (()-ENBA
showed interaction with the hA2B receptor with Ki values of
3.2, 4.8, 2.7, and 4.9 µM, respectively.
Compounds 1-8, 23, and 24 showed high affinity and
selectivity for the human A1 receptor. The 5′-chloro-5′-deoxy
derivatives of CPA (5′-Cl-CPA, 1) and (()-ENBA (5′Cl5′d-
(()-ENBA, 3) displayed the highest affinity at subnanomolar
range for the hA1 AR (Ki ) 0.5 nM) and relevant selectivity vs
hA2A and hA3 ARs. Among all tested compounds, 5′Cl5′d-(()-
ENBA (3) showed the highest selectivity for hA1 vs hA3 AR
(2530-fold). At the same receptor subtype, the corresponding
2-chloro analogues (compounds 2 and 4) showed a slightly lower
affinity and selectivity. A similar modification in the ribose-
modified C2′-methyl analogues resulted in a moderate decrease
in affinity at all human AR subtypes; however, these compounds
showed an A1 selectivity similar to that of parent adenosine
analogues.
These data are in accord with that previously reported by us
for the C2′-methyl analogues of N6-substituted adenosine that
were found to have a decreased affinity, in particular at A3 ARs,
resulting in more A1 selective agonists4b,8 (e.g., compare CPA
and (()-ENBA with 2′-Me-CPA and 23, respectively). Among
the 5′-chloro-5′-deoxy- and 2,5′-dichloro-5′-deoxy-C2′-methy-
ladenosine derivatives, the most interesting compounds appeared
to be the N6-(()-endo-norbornyl analogues 7 and 8, which
a Reagents and conditions: (i) (()-endo-2-norbornylamine hydrochloride,
TEA, EtOH, ∆; (ii) NH3/MeOH, room temperature; (iii) 2,2-dimethox-
ypropane, camphorsulfonic acid, (CH3)2CO, ∆; (iv) SOCl2, pyridine,
CH3CN, -5 °C to room temperature; (v) HCOOH (70%), 40 °C.
D-ribofuranosyl)purine (15), prepared as reported by Hou et al.,7
with (()-endo-norborn-2-yl-amine hydrochloride in the presence
of triethylamine in absolute ethanol, followed by sugar deblock-
ing with methanolic ammonia, gave 2-Cl-(()-ENBA (16).
Compound 16 was protected as 2′,3′-isopropylidene derivative
17 using camphorsulfonic acid and 2,2-dimethoxypropane in
acetone in 80% yield. Conversion of 17 to 5′-chloro derivative
18 was performed by treatment with a mixture of thionyl
chloride, pyridine, and acetonitrile. Finally, deprotection of 18
with 70% formic acid at 40 °C furnished compound 4. Direct
conversion of 2-Cl-(()-ENBA (16) into its 5′-chloro-5′-deoxy
derivative 4 using thionyl chloride and pyridine in acetonitrile
or thionyl chloride and hexamethylphosphoramide (HMPA) was
also tried, but low yields of 4 were obtained.

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2395
Scheme 2a
a Reagents and conditions: (i) R1NH2, EtOH, ∆; (ii) NH3/MeOH, room temperature; (iii) 2,2-dimethoxypropane, camphorsulfonic acid, (CH3)2CO, ∆; (iv)
SOCl2, pyridine, CH3CN, -5 °C to room temperature; (v) NaH 60%, 2-fluorothiophenol, DMF, 0 °C to room temperature; (vi) HCOOH (70%), 40 °C.
showed good A1 affinity (Ki ) 9.13 and 9.16 nM, respectively)
and A1 selectivity, vs A2A ) 2630 and 3450, and vs A3 ) 1270
and 644, respectively.
A1 receptor was studied in comparison with the full agonist
CCPA. The functional assay showed that all these compounds
are full agonists on the basis of their adenylyl cyclase inhibitory
activity, which was comparable to that of CCPA (Figure 1).
The ability of the selected compounds 1, 3, CPA, and (()-
ENBA to inhibit forskolin-stimulated cAMP production via
human A3 AR was studied in comparison with the A3 agonist
NECA. The functional assay showed that the 5′-chloro-5′-deoxy
derivatives 1 and 3 behave as antagonists, while the corre-
sponding parent compounds CPA and (()-ENBA are partial
agonists compared with NECA, which proved to be a full
agonist (Figure 2).
Molecular Modeling. In order to explain why the replace-
ment of the OH group in the 5′-position in N6-substituted
adenosine analogues with a chlorine is tolerated at hA1 AR but
is scarcely tolerated at hA3 AR, a molecular docking analysis
of CPA, (()-ENBA, and compounds 1 and 3 was performed at
the homology models of both ARs built using the bovine
rhodopsin (b-Rho) crystal structure as a template.13 In the last
As far as it concerns the 5′-(2-fluorophenylthio) substitution,
the N6-cyclopentyl derivative 9 was the most affine compound
at the hA1 receptor with a Ki of 64.7 nM and with A2A/A1 and
A3/A1 selectivities lower than those of the parent compounds
CPA and 5′-Cl-CPA (1), while the N6-tetrahydrofuranyl ana-
logue 41 displayed 3.9-fold lower affinity and a similar
selectivity. The introduction of a chlorine in the 2-position of
the purine ring (compound 10) induced a slight decrease of the
affinity at the hA1 and hA3 receptors and slightly enhanced the
affinity at the hA2A receptor. The 2′-C-methyl modification in
these compounds and in the N6-(()-endo-2-norbornyl analogues
(compounds 11-14) brought about a reduction of affinity at
all receptor subtypes.
Adenylyl Cyclase Activity. The ability of selected com-
pounds (1, 2, 3, 4, 23, (()-ENBA, and 2-Cl-(()-ENBA) to
inhibit forskolin-stimulated cAMP production via the human

2396 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
Scheme 3a
Because of the difficulty to generate a fully active conforma-
tion (e.g., Meta II of b-Rho) for analyzing agonist binding, the
binding preference of agonists CPA, (()-ENBA, 1, and 3 to
the Meta I conformation of both hA1 AR and hA3 AR was
studied. The agonist-bound conformation, in a form resembling
the not fully activated Meta I state of b-Rho, was obtained by
modeling the rearrangement of the side chain of W265 (6.48),
as described in the Experimental Section. Although the Meta I
state is still far more similar to the resting conformation than
to the presumed yet undisclosed fully active conformation, this
state structure is preferable to the ground-state structure for
agonist docking.
To assess the dynamic stability of the obtained complexes
and to analyze the potential ligand-receptor interactions, a
molecular dynamics (MD) simulation of 1 ns at a constant
temperature of 300 K was run. The distances between the ligands
and the key receptor residues were monitored along the complete
MD trajectory. The results of docking and MD simulations
performed for compounds CPA, (()-ENBA, 1, and 3 indicated
that the adenine moiety of the ligands had a similar position
and orientation inside the putative hA1 AR binding site, defined
by TMIII, TMV, TMVI, and TMVII. In particular, the N6-
substituent of the ligands was oriented toward TMV, whereas
the ribose ring was placed between helices TMIII and TMVII
with the 5′-substituent pointing toward the intracellular part of
the receptor. Figures 3 and 4 show the binding mode of CPA,
(()-ENBA, 1, and 3 into the hA1 AR model as the average
structure calculated on the last 200 ps of the production step.
a Reagents and conditions: (i) 2,2-dimethoxypropane, camphorsulfonic
acid, (CH3)2CO, ∆; (ii) SOCl2, pyridine, CH3CN, -5 °C to room
temperature; (iii) NaH 60%, 2-fluorothiophenol, DMF, 0 °C to room
temperature; (iv) HCOOH (70%), 40 °C.
All four ligands adopted a stable binding pose during the
simulation time, forming almost the same interactions with hA1
AR. Interestingly, these interactions were very stable throughout
the MD simulation, thus explaining the high potency of CPA
(Ki ) 2.25 nM), (()-ENBA (Ki ) 0.54 nM), 1 (Ki ) 0.59 nM),
and 3 (Ki ) 0.51 nM) toward hA1 AR. In agreement with the
published data of molecular modeling and site-directed mu-
tagenesis of the AR family,14d,17-19 the 3′-OH and 2′-OH groups
were H-bonded to T277 (7.42) and H278 (7.43), respectively.
The histidine residue is conserved among all ARs, and support
for it being a critical recognition element has come from diverse
approaches.14e,20-22 For example, H278 (7.43) is important for
agonist but not antagonist binding for the A1, A2A, and A3 ARs.
The N6-amino group was found to establish H-bonds with the
CO oxygen of N254 (6.55) side chain. The N1 nitrogen of 1
also formed a H-bond with the N254 (6.55) NH group. This
residue, conserved among all adenosine receptor subtypes, was
found to be important for ligand binding. In fact, the inability
of the N250A mutant hA3 AR23 or the corresponding mutant
A2A AR20 to bind either radiolabeled agonist or antagonist was
consistent with a proposed direct interaction of this residue with
our ligands. Moreover, both cyclopentyl and norbornyl N6-
substituents were favorably located in a pocket formed by
several hydrophobic residues including L88 (3.33), M180 (5.38),
V181 (5.39), F185 (5.43), and L258 (6.59). In addition, it was
found that the 5′-OH of CPA and (()-ENBA accepted a H-bond
from the T91 (3.36) OH group, in line with the site mutagenesis
studies, which indicate that mutation of this residue to alanine
in the A1 and A2A receptors, respectively, substantially decreases
agonist affinity.24,25
14 years, many AR models were published, reporting also the
docking of agonists and antagonists.14 Indeed, the use of the b-
Rho X-ray structure13 has led to a great improvement in the
results. Unfortunately, this structure that serves as template for
GPCR models was obtained for its ground-state only. For this
reason, there is an opinion that rhodopsin-based homology
modeling of GPCRs is more applicable for studying antagonist
than agonist binding modes. Until now, there is only a rough
picture of the conformational changes that occur during receptor
activation. Ballesteros et al.14f suggested that receptor activation
could be due to a different rearrangement of TM3 and TM6.
Furthermore, on the basis of UV absorption analysis, it has been
suggested that when b-Rho is activated, the ꢁ1 rotamer of the
high conserved residue W265 (6.48) shifts from gauche+ to
trans.15 Thus, the intramolecular contact network might be
destabilized, inducing a characteristic anticlockwise movement
of TMs III, VI, and VII from the extracellular view to activate
the receptor. Recently, Kobilka et al.16 reported the first X-ray
structure of the human ꢀ2-adrenergic receptor (ꢀ2-AR). The data
show that the overall topologies of b-Rho and ꢀ2-AR are quite
similar. The root-mean-square deviation (rmsd) for the CR
backbone of the transmembrane region between rhodopsin and
ꢀ2-AR is 1.56 Å, which indicates a very similar arrangement
of the TM helices. This feature also supports the previous notion
of a conserved activation mechanism, i.e. an agonist-induced
conformational rearrangement, across this class of GPCRs.
However, the ꢀ2-AR shows a more open structure, especially
in the lower ends of TM3 and TM6. The authors suggest that
this feature could be the basis for the observed basal activity
observed for many GPCRs. It is also noteworthy that the current
structure of the ꢀ2-AR is an inactive state and may, therefore,
only be useful for identifying inverse agonists and antagonists.
The chlorine atom at the 5′-position in compounds 1 and 3
was too far away from the T91 (3.36) OH group to form an
effective H-bond. Nevertheless, MD simulations indicated that
it could still accept a H-bond from the W247 (6.48) indole NH,
thus explaining the high hA1 AR affinity displayed by these
compounds. The importance of W265 (6.48) in b-Rho activation

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2397
Table 1. Binding Affinity at Human A1, A2A, A2B, and A3 Adenosine Receptor Subtypes
Ki (nM)
selectivity
a
b
c
b
compd
R
R1
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
2FPhS
2FPhS
2FPhS
2FPhS
2FPhS
2FPhS
2FPhS
OH
OH
OH
OH
OH
R2
X
A1
A2A
A2B
A3
A2A/A1
A3/A1
1
2
3
4
5
6
7
8
H
H
H
H
CH3
CH3
CH3
CH3
H
H
Cl
H
Cl
H
Cl
H
Cl
H
Cl
H
Cl
H
Cl
H
H
Cl
H
Cl
H
Cl
H
CH2
CH2
0.59
837
2,160
1,340
3,210
4,830
2,740
376
417
1,470
1,380
2,630
1,270
>3,520
1,770
2,630
3,450
80
8
>60
>47
>94
637
267
2,530
875
237
237
1,270
644
5
1.56
0.51
1.61
28.4
12.8
9.13
9.16
64.7
258
1,680
2,150
1,060
2,970
250
1,290
1,410
6,740
3,030
11,600
5,900
296
2,050
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
>10,000
18,600
CH2
CH2
>100,000
16,200
24,000
31,600
5,170
9
CH2
CH2
CH2
CH2
10
11
12
13
14
41
23
24
CPAd
H
2,190
479
2
0.9
0.5
3
1.5
CH3
CH3
CH3
CH3
H
CH3
CH3
H
H
H
H
CH3
CH3
>100,000
>100,000
>100,000
>100,000
18,900
8,780
11,800
794
2,300
1,270
797
10,400
9,580
1,440
1,150
3,480
4,350
746
2,520
4,490
43
42
101
129
879
>34
76
O
3
4.09
2,150
1,700
353
2,875
2,350
1,120
2,301
2,903
616
645
19
6.96
2.25
0.8
0.54
0.71
4.5
CH2
CH2
CCPAd
(()-ENBA
16 2-Cl-(()-ENBA
2′-Me-CPAd
2′-Me-CCPAd
18,800
4,930
>10,000
26,800
37,600
53
187
182
195
348
OH
OH
OH
CH2
CH2
Cl
3.3
1,150
a Displacement of [3H]CCPA binding in CHO cells stably transfected with the human recombinant A1 adenosine receptor. b Displacement of [3H]NECA
binding in CHO cells stably transfected with human recombinant A1 or A3 adenosine receptors. c Ki values were calculated from IC50 values determined by
inhibition of NECA-stimulated adenylyl cyclase activity. d Data are from ref 3a.
Figure 1. Inhibition of adenylyl cyclase activity via the human A1
adenosine receptor by selected compounds. The percentage of activity
remaining after agonist-mediated inhibition of 10 µM forskolin-
stimulated cyclase activity (100%) is shown. Data are means ((SEM)
of three independent experiments.
Figure 2. Inhibition of adenylyl cyclase activity via the human A3
adenosine receptor by selected compounds. The percentage of activity
remaining after agonist-mediated inhibition of 10 µM forskolin-
stimulated cyclase activity (100%) is shown. Data are means ((SEM)
of three independent experiments.
was suggested in a UV-visible spectroscopic analysis of site-
directed mutagenesis of this residue. The differential absorbance
indicated that perturbations in the characteristics of W126 (3.41)
and W265 (6.48) resulted from a general conformational change
concomitant with Meta II formation.15 There was a rearrange-
ment close to the bend of TM6 upon Meta I formation. The
electron density featured a significant deviation from the position
of W265 (6.48) in the ground-state structure, suggesting the
possibility of movement of W265 (6.48). Meta I formation
involved no large rigid-body movements or rotations of helices
from their position in the ground-state. Instead, changes seemed
to be localized, probably involving movement of side chains
such as W265 (6.48) in kinked regions of helices close to the
retinal-binding pocket.26 Before the MD simulation of the ligand/
hA1 AR complexes, W247 (6.48) was in the gauche- ꢁ1
configuration, as described in the Experimental Section. During
the MD simulation, the rotamer of W247 (6.48) spontaneously
shifted from gauche- ꢁ1 to trans ꢁ1 (ꢁ1 ) -164 for CPA, ꢁ1 )
-168 for (()-ENBA, ꢁ1 ) -168 for 1, and ꢁ1 ) -155 for 3),

2398 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
Figure 3. On the left, (extracellular) view of CPA (top, yellow) and (()-ENBA (bottom, orange) complexed with the hA1 AR model. For clarity,
only interacting residues are displayed. Ligands and interacting key residues (green) are represented as stick models, while the protein is represented
as gray ribbons. H-bonds are shown as dashed yellow lines. On the right, schematic representation of the binding mode of CPA (top) and (()-
ENBA (bottom) obtained after docking and MD simulations. The green arrows correspond to the putative H-bonds. The critical residues involved
in interactions with the ligand are colored in blue.
that is generally assumed to be stabilized by agonist binding
and indicative of an activated receptor.26,27
Different results were obtained when analyzing the trajectories
of 1 and 3 in complex with the hA3 AR. With the exception of
the H-bond formed between the NH6 group of the ligands and
the CO oxygen of the N254 side chain, which remained quite
stable during all the simulation period (the N6 · · ·OdC distance
was ∼3.0 Å), the remaining polar interactions were not strong
enough to be preserved throughout the MD simulation, giving
average distances longer than that of an ideal H-bond. In
particular, the 3/hA3 complex turned out to be highly unstable
during the MD simulation. The ligand considerably changed
its position at the hA3 AR binding site and was oriented parallel
to the transmembrane domain axis already after a few hundred
picoseconds. This was probably due to the inability of the
chlorine atoms of 1 and 3 to make a H-bond to T94 (3.36),
unlike the cases of CPA and (()-ENBA, which possess a 5′-
OH group that, in contrast, is able to form this interaction. Thus,
the incapability of the 5′-Cl atom of 1 and 3 to form a H-bond
to T94 (3.36), which is a key receptor anchoring point for the
hA3 AR agonists,28 together with the sterically bulky substituents
at the N6 position of the adenine ring might change the optimal
binding mode of the ligand, thereby decreasing the relative
stability of the complexes. This finding is in accordance with
the drastic reduction of affinity of 1 and 3 toward hA3 AR.
Moreover, it is important to note that during the MD simulation
Examination of the optimized models of the CPA/hA3 AR
and (()-ENBA/hA3 AR complexes (Figure 5) showed that the
purine ring of the ligands was surrounded by a hydrophobic
pocket defined by L91 (3.33) and L246 (6.51). The N6-
cyclopentyl and norbornyl substituents appeared to be wedged
between TMV and TMVI, interacting with hydrophobic residues
F182 (5.43), I186 (5.47), M172 (5.32), M177 (5.38), and V178
(5.39). In addition, the N6-nitrogen was located within H-
bonding distance from the CdO oxygen of N250 (6.55). The
5′-OH group also donated a H-bond to the T94 (3.36) OH
oxygen. Only the 2′-OH substituent of CPA established a further
H-bond with S271 (7.42), while the 2′-OH of (()-ENBA did
not seem to interact with S271 (7.42). Moreover, the 3′-OH
substituent of both ligands was away from the H272 (7.43)
imidazole ring to make an efficient H-bond. In fact, from the
MD trajectories (data not shown), it can be deduced that these
H-bonds, initially present in the complex, were gradually lost
during the whole length of the monitored simulation. These
results are consistent with 10-fold and 1000-fold lower affinity
of CPA and (()-ENBA, respectively, toward hA3 AR, in
comparison with the corresponding binding affinity of these
compounds toward hA1 AR.

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2399
Figure 4. On the left, (extracellular) view of compounds 1 (top, magenta) and 3 (bottom, cyan) complexed with the hA1 AR model. For clarity,
only interacting residues are displayed. Ligands and interacting key residues (green) are represented as stick models, while the protein is represented
as gray ribbons. H-bonds are shown as dashed yellow lines. On the right, schematic representation of the binding mode of 1 (top) and 3 (bottom)
obtained after docking and MD simulations. The green arrows correspond to the putative H-bonds. The critical residues involved in interactions
with the ligand are colored in blue.
of both 1/hA3 and 3/hA3 complexes, in sharp contrast to 1/hA1
and 3/hA1 complexes, the ꢁ1 rotamer of residue W243 (6.48)
unexpectedly shifted from the gauche- to the gauche+ confor-
mation (ꢁ1 ) -69 for 1 and ꢁ1 ) -80 for 3), indicative of an
“inactive” state of the receptor. This result nicely explains the
observed low intrinsic efficacy of 1 and 3 at hA3 AR in
comparison with the high efficacy of the same ligands at hA1
AR.
5′Cl5′d-(()-ENBA compared to that of (()-ENBA is quite
surprising because these compounds displayed similar affinity
and efficacy profiles at both human and rat4a A1 AR. Further-
more, 5′Cl5′d-(()-ENBA could have more favorable blood-brain
transport characteristics owing to its higher lipophilicity (3: logP
) 1.04 vs (()-ENBA: log P ) -0.14). Further research is
needed to verify if the higher activity of (()-ENBA is due to
its metabolic conversion to nucleotide derivatives that could
trigger different signaling pathways or to the lower metabolic
stability of 5′Cl5′d-(()-ENBA.
Antinociceptive Effect. To investigate the therapeutic po-
tential of 5′Cl5′d-(()-ENBA (3), we have evaluated its analgesic
activity in mice in comparison with (()-ENBA using the
formalin test. Formalin injection induces a biphasic stereotypical
nocifensive behavior.29 Nociceptive responses are divided into
an early, short lasting first phase (0-7 min) caused by a primary
afferent discharge produced by the stimulus, followed by a
quiescent period and then a second, prolonged phase (15-60
min) of tonic pain. Systemic administration of 5′Cl5′d-(()-
ENBA (1-2 mg/Kg, i.p.), 10 min before formalin, reduced the
late nociceptive behavior induced by formalin in a dose-
dependent manner (P < 0.005). The highest dose of 5′Cl5′d-
(()-ENBA used (2 mg/Kg) reduced both the early and the late
phases of the formalin test, and this effect was prevented by
DPCPX (3 mg/kg, i.p.), a selective A1 receptor antagonist
(Figure 6). The antinociceptive effect of 5′Cl5′d-(()-ENBA
proved to be comparable to that of 2′-Me-CCPA.30 Systemic
administration of (()-ENBA (0.3-1 mg/kg, i.p.), 10 min before
formalin injection, completely erased both the early and the late
phases of the formalin-induced nociceptive behavior (P < 0.005)
(Figure 7). The lower antinociceptive effect displayed by
Conclusions
In summary, we synthesized a series of 5′-chloro-5′-deoxy-
N6-cycloalkyl(bicycloalkyl)adenosine and 2′-C-methyladenosine
derivatives to evaluate their affinity and efficacy at human A1,
A2A, A2B, and A3 adenosine receptor subtypes. Biological data
confirmed that the replacement of the 5′-hydroxyl group by a
chlorine atom in the ꢀ-D-ribofuranose ring in N6-substituted
adenosine derivatives is well tolerated by the human A1 receptor
but less tolerated by other human receptor subtypes; therefore,
this modification represents an effective strategy to increase the
selectivity for hA1 AR.
In the series of N6-cyclopentyl- and (endo-norborn-2-yl)ad-
enosine derivatives, 5′-chloro-5′-deoxy-CPA (1) and 5′-chloro-
5′-deoxy-(()-ENBA (3) displayed the highest hA1 affinity in
the subnanomolar range and significant hA1 selectivity. The
corresponding derivatives of 2′-C-methyladenosine showed a
slight decrease of the affinity at all receptor subtypes but a
similar or increased A1 selectivity as compared to the adenosine

2400 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
Figure 5. On the left, (extracellular) view of CPA (top, yellow) and (()-ENBA (bottom, orange) complexed with the hA3 AR model. For clarity,
only interacting residues are displayed. Ligands and interacting key residues (green) are represented as stick models, while the protein is represented
as gray ribbons. H-bonds are shown as dashed yellow lines. On the right, schematic representation of the binding mode of CPA (top) and (()-
ENBA (bottom) obtained after docking and MD simulations. The green arrows correspond to the putative H-bonds. The critical residues involved
in interactions with the ligand are colored in blue.
of D2O. Mass spectra were recorded on an HP 1100 series
instrument. All measurements were performed in the positive ion
mode using atmospheric pressure electrospray ionization (API-ESI).
Partition coefficients (log P) were computed using the log P function
implemented in ChemDraw Ultra version 10.0. Elemental analyses
were determined on an EA 1108 CHNS-O (Fisons Instruments)
analyzer and are within (0.4% of theoretical values. CPA was
purchased from Tocris Bioscience.
analogues. The higher selectivity of 5′-chloro-5′-deoxy-modified
adenosine derivatives for hA1 AR vs hA3 AR compared to that
of the 5′-hydroxy parent compounds was rationalized by a
molecular modeling study. In particular, it was pointed out that
the 5′-Cl atom of compounds 1 and 3 is unable to form a H-bond
to a T94 (3.36) residue, which is a key receptor anchoring point
for the hA3 AR agonists.
5′-Chloro-5′-deoxy-(()-ENBA (3) was found to be effective
in reverting formaline-induced nocifensive behavior in mice,
albeit at a higher concentration than (()-ENBA, confirming that
pharmacological modulation of the hA1 AR may play a critical
role in pain modulation. Owing to the higher selectivity of 5′-
chloro-5′-deoxy adenosine derivatives for the human A1 recep-
tor, this type of A1 agonists deserves further investigation to
explore their therapeutic potential.
General Procedure for N6-Amination (Compounds 16, 23,
and 24). To a stirred solution of 2,6-dichloro-9H-(2,3,5-O-acetyl-ꢀ-
D-ribofuranosyl)purine (15)7 or 6-chloro-9H-(2-C-methyl-2,3,5-O-
benzoyl-ꢀ-D-ribofuranosyl)purine8 (19) or 2,6-dichloro-9H-(2-C-methyl-
2,3,5-O-benzoyl-ꢀ-D-ribofuranosyl)purine8 (20) (1.0 mmol) in absolute
ethanol (20 mL), were added (()-endo-norborn-2-ylamine hydrochlo-
ride (2.0 mmol), and anhydrous triethylamine (5.8 mmol). The reaction
mixture was refluxed for the time reported below and concentrated in
vacuo. The residue was dissolved in methanol saturated with ammonia
(25 mL) and stirred at room temperature overnight. Evaporation of
the solvent to dryness gave a residue which was purified by column
chromatography.
Experimental Section
Chemistry. Thin layer chromatography (TLC) was run on silica
gel 60 F254 plates (Merck); silica gel 60 (70-230 and 230-400
mesh, Merck) for column chromatography was used. 1H NMR
spectra were recorded on a Varian Mercury AS400 instrument. The
chemical shift values are expressed in δ values (ppm), and coupling
constants (J) are in hertz; TMS was used as an internal standard.
The presence of all exchangeable protons was confirmed by addition
2-Chloro-N6-(()-endo-norbornyl-9H-(ꢀ-D-ribofuranosyl)ad-
enine (16). The title compound was obtained from 15 (reaction
time 3 h). Chromatography on a silica gel column (CHCl3-MeOH,
1
96:4) gave 16 as a white solid (90% yield). H NMR (DMSO-d6)
δ 1.20-1.29 (m, 3H, norbornyl), 1.37-1.46 (m, 3H, norbornyl),
1.50-1.65 (2m, 1H, norbornyl), 1.83-1.91 (m, 1H, norbornyl),

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2401
bornyl), 2.52 (s, 1H, norbornyl), 3.62-3.70 (m, 1H, H-5′),
3.78-3.84 (m, 1H, H-5′), 3.86-3.90 (m, 1H, H-4′), 4.02-4.10 (m,
1H, H-3′), 4.28-4.38 (m, 1H, NHCH), 5.15-5.25 (m, 3H, OH),
5.95 (s, 1H, H-1′), 7.75 (t, J ) 6.0 Hz, 1H, NH), 8.18 (s, 1H, H-2),
8.45 (s, 1H, H-8). MS: m/z 376.5 [M + H]+. Anal. (C18H25N5O4)
C, H, N.
2-Chloro-N6-(()-endo-norbornyl-9H-(2-C-methyl-ꢀ-D-ribo-
furanosyl)adenine (24). The title compound was synthesized from
20 (reaction time 2 h) and purified by chromatography on a silica
gel column (CHCl3-MeOH, 97:3) as a white solid (97% yield).
1H NMR (DMSO-d6) δ 0.80 (2s, 3H, CH3), 1.22-1.30 (m, 3H,
norbornyl), 1.40-1.48 (m, 3H, norbornyl), 1.50-1.63 (2m, 1H,
norbornyl), 1.82-1.88 (2m, 1H, norbornyl), 2.16 (br s, 1H,
norbornyl), 2.52 (br s, 1H, norbornyl), 3.64-3.70 (m, 1H, H-5′),
3.78-3.84 (m, 1H, H-5′), 3.86-3.92 (m, 1H, H-4′), 4.0 (dd, J )
7.3, 9.0 Hz, 1H, H-3′), 4.20-4.28 (m, 1H, NHCH), 5.12-5.16 (m,
1H, OH), 5.22 (d, J ) 6.4 Hz, 1H, OH), 5.32 (d, J ) 3.4 Hz, 1H,
OH), 5.82 (s, 1H, H-1′), 8.38 (t, J ) 6.2 Hz, 1H, NH), 8.50 (s, 1H,
H-8). MS: m/z 410.7 [M + H]+. Anal. (C18H24ClN5O4) C, H, N.
General Procedure for the Synthesis of 2′,3′-O-Isopropy-
lidene Derivatives 17, 27, 28, and 38. A mixture of 16, 23, 24, or
2-chloro-N6-cyclopentyladenosine (37) (1.0 mmol), 2,2-dimethox-
ypropane (18.1 mmol), and camphorsulfonic acid (1.0 mmol) in
anhydrous acetone (10 mL) was stirred at 55 °C for the time
reported below. The solvent was removed in vacuo, and the residue
was purified by column chromatography to afford the desired
compounds.
Figure 6. Effect of subcutaneous formalin (1.25%, 30 µL) injections
into the hind paw of mice on the time course of the nociceptive
behaviors. Formalin was injected 10 min after the systemic administra-
tion of vehicle (0.9% NaCl, i.p.) or drugs. Part A shows the effects of
the systemic administration of 5′Cl5′d-(()-ENBA (3) (1 and 2 mg/kg,
i.p.). Part B shows the effects of the systemic administration of 3 (2
mg/kg, i.p.) in combination with DPCPX (3 mg/kg, i.p.). Recording of
nociceptive behavior began immediately after the injection of formalin
(time 0) and was continued for 60 min. Each point represents the total
time of the nociceptive responses (mean ( S.E.M.) of 8 mice per group,
measured every 5 min. * indicates significant differences versus vehicle,
and O indicates significant differences versus 5′Cl5′d-(()-ENBA. P <
0.05 was considered statistically significant.
2-Chloro-N6-(()-endo-norbornyl-9H-(2,3-O-isopropylidene-
ꢀ-D-ribofuranosyl)adenine (17). The title compound was synthe-
sized from 16 (reaction time 3 h). Chromatography on a silica gel
column (CHCl3-MeOH, 98:2) gave 17 as a white solid (80% yield).
1H NMR (DMSO-d6) δ 1.15-1.26 (m, 3H, norbornyl), 1.28, 1.50
(2s, 6H, CH3), 1.35-1.44 (m, 3H, norbornyl), 1.58-1.63 (m, 1H,
norbornyl), 1.81-1.93 (m, 1H, norbornyl), 2.16 (br s, 1H, nor-
bornyl), 2.52 (s, 1H, norbornyl), 3.45-3.58 (m, 2H, H-5′),
4.17-4.22 (m, 1H, H-4′), 4.24 (br s, 1H, NHCH), 4.88-4.92 (m,
1H, H-3′), 5.06 (pseudo t, 1H, OH), 5.22-5.28 (m, 1H, H-2′), 6.05
(d, J ) 2.6 Hz, 1H, H-1′), 8.32 (s, 1H, H-8), 8.38 (d, J ) 6.4 Hz,
1H, NH). MS: m/z 436.9 [M + H]+. Anal. (C20H26ClN5O4) C, H,
N.
N6-(()-endo-Norbornyl-9H-(2-C-methyl-2,3-O-isopropylidene-
ꢀ-D-ribofuranosyl)adenine (27). The title compound was synthesized
from 23 (reaction time 7 h). Chromatography on a silica gel column
(CHCl3-MeOH, 99.5:0.5) gave 27 as a white solid (50% yield). 1H
NMR (DMSO-d6) δ 1.10 (2s, 3H, CH3), 1.20-1.31 (m, 3H, norbornyl),
1.35, 1.55 (2s, 6H, CH3), 1.40-1.48 (m, 3H, norbornyl), 1.58-1.64
(m, 1H, norbornyl), 1.82-1.92 (m, 1H, norbornyl), 2.18 (br s, 1H,
norbornyl), 2.52 (br s, 1H, norbornyl), 3.77 (dq, J ) 6.4, 12.4 Hz,
2H, H-5′), 4.20-4.28 (m, 1H, H-4′), 4.35 (br s, 1H, NHCH), 4.58 (d,
J ) 2.1 Hz, 1H, H-3′), 5.40 (br s, 1H, OH), 6.22 (s, 1H, H-1′), 7.80
(d, J ) 6.5 Hz, 1H, NH), 8.20 (s, 1H, H-2), 8.32 (s, 1H, H-8). MS:
m/z 416.5 [M + H]+. Anal. (C21H29N5O4) C, H, N.
Figure 7. Effect of subcutaneous formalin (1.25%, 30 µL) injection
into the hind paw of mice on the time course of the nociceptive
behaviors. Formalin was injected 10 min after the systemic administra-
tion of vehicle (0.9% NaCl, i.p.), 5′Cl5′d-(()-ENBA (3) (1 and 2 mg/
kg, i.p.) or (()-ENBA (0.3 and 1 mg/kg, i.p.). Recording of nociceptive
behaviors began immediately after the injection of formalin (time 0)
and was continued for 60 min. Each point represents the total time of
the nociceptive responses (mean ( S.E.M.) of 8 mice per group,
measured every 5 min. * indicates significant differences versus vehicle.
P < 0.05 was considered statistically significant.
2-Chloro-N6-(()-endo-norbornyl-9H-(2-C-methyl-2,3-O-iso-
propylidene-ꢀ-D-ribofuranosyl)adenine (28). The title compound
was synthesized from 24 (reaction time 3 h) and purified by
chromatography on a silica gel column (CHCl3-MeOH, 99.5:0.5)
as a white solid (95% yield). 1H NMR (DMSO-d6) δ 1.12 (2s, 3H,
CH3), 1.22-1.30 (m, 3H, norbornyl), 1.35, 1.55 (2s, 6H, CH3),
1.40-1.48 (m, 3H, norbornyl), 1.58-1.64 (m, 1H, norbornyl),
1.82-1.92 (m, 1H, norbornyl), 2.16 (br s, 1H, norbornyl), 2.52 (br
s, 1H, norbornyl), 3.65-3.75 (m, 2H, H-5′), 4.20-4.28 (m, 2H,
NHCH, H-4′), 4.57 (d, J ) 2.1 Hz, 1H, H-3′), 5.25 (t, J ) 5.6 Hz,
1H, OH), 6.12 (s, 1H, H-1′), 8.36 (s, 1H, H-8), 8.42 (d, J ) 6.8
Hz, 1H, NH). MS: m/z 450.9 [M + H]+. Anal. (C21H28ClN5O4) C,
H, N.
2.16 (br s, 1H, norbornyl), 2.52 (br s, 1H, norbornyl), 3.49-3.56
(m, 1H, H-5′), 3.61-3.67 (m, 1H, H-5′), 3.89-3.94 (m, 1H, H-4′),
4.10 (q, J ) 4.9 Hz, 1H, H-3′), 4.20-4.27 (m, 1H, NHCH), 4.50
(q, J ) 5.6 Hz, 1H, H-2′), 5.05 (t, J ) 5.8 Hz, 1H, OH), 5.40 (d,
J ) 5.1 Hz, 1H, OH), 5.45 (dd, J ) 3.2, 6.2 Hz, 1H, OH), 5.80 (d,
J ) 6.0 Hz, 1H, H-1′), 8.40 (s and d, 2H, NH, H-8). MS: m/z 396.7
[M + H]+. Anal. (C17H22ClN5O4) C, H, N.
N6-(()-endo-Norbornyl-9H-(2-C-methyl-ꢀ-D-ribofuranosyl)-
adenine (23). The title compound was synthesized from 19 (reaction
2-Chloro-N6-cyclopentyl-9H-(2,3-O-isopropylidene-ꢀ-D-ribo-
furanosyl)adenine (38). The title compound was synthesized from
37 (reaction time 2 h). Chromatography on a silica gel column
time 3.5 h). Chromatography on
a silica gel column
1
(CHCl3-MeOH, 93:7) gave 23 as a white solid (90% yield). H
NMR (DMSO-d6) δ 0.78 (2s, 3H, CH3), 1.20-1.36 (m, 3H,
norbornyl), 1.38-1.48 (m, 3H, norbornyl), 1.56-1.66 (m, 1H,
norbornyl), 1.84-1.94 (m, 1H, norbornyl), 2.16 (br s, 1H, nor-
1
(CHCl3-MeOH, 99:1) gave 38 as a white solid (75% yield). H
NMR (DMSO-d6) δ 1.30, 1.50 (2s, 6H, CH3), 1.44-1.60 (m, 4H,
cyclopentyl), 1.62-1.72 (m, 2H, cyclopentyl), 1.82-1.98 (m, 2H,

2402 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
cyclopentyl), 3.48-3.56 (m, 2H, H-5′), 4.20 (br s, 1H, H-4′),
4.38-4.46 (m, 1H, NHCH), 4.92 (dd, J ) 2.6, 6.0 Hz, 1H, H-3′),
5.08 (pseudo t, 1H, OH), 5.26 (dd, J ) 3.0, 6.0 Hz, 1H, H-2′),
6.04 (d, J ) 2.6 Hz, 1H, H-1′), 8.35 (s, d, 2H, H-8 and NH). MS:
m/z 410.8 [M + H]+. Anal. (C18H24ClN5O4) C, H, N.
(br s, 1H, norbornyl), 2.52 (s, 1H, norbornyl), 4.05 (dq, J ) 6.4,
10.7 Hz, 2H, H-5′), 4.20-4.28 (m, 1H, NHCH), 4.32-4.38 (m,
1H, H-4′), 4.60 (d, J ) 3.0 Hz, 1H, H-3′), 6.20 (s, 1H, H-1′), 8.25
(s, 1H, H-8), 8.45 (d, J ) 6.8 Hz, 1H, NH). MS: m/z 469.4 [M +
H]+. Anal. (C21H27Cl2N5O3) C, H, N.
General Procedure for the Synthesis of Compounds 18,
29-32, and 39. Compounds 17, 25,3a 26,3a 27, 28, or 38 (1.0 mmol)
in dry acetonitrile (10 mL) under nitrogen atmosphere were stirred
with cooling to -5 °C. SOCl2 (3.0 mmol) was added portionwise
followed by dry pyridine (2.0 mmol). After 30 min, the reaction
mixture was stirred at room temperature. The procedure was
repeated after 6 h, and the mixture was stirred at room temperature
overnight. Water was added (5 mL), and the solution was
neutralized with NaHCO3 (1 M) and extracted with CH2Cl2 (3 ×
10 mL). The organic layer was dried (Na2SO4), and the solvent
was evaporated to dryness. The residue was purified by column
chromatography as reported below.
2-Chloro-N6-cyclopentyl-9H-(2,3-O-isopropylidene-5-chloro-
5-deoxy-ꢀ-D-ribofuranosyl)adenine (39). The title compound was
synthesized from 38 and purified by chromatography on a silica
gel column (CHCl3) as a white solid (68% yield). 1H NMR (DMSO-
d6): δ 1.33, 1.55 (2s, 6H, CH3), 1.50-1.60 (m, 4H, cyclopentyl),
1.62-1.75 (m, 2H, cyclopentyl), 1.82-2.0 (m, 2H, cyclopentyl),
3.76 (dd, J ) 6.0, 11.1 Hz, 1H, H-5′), 3.86 (dd, J ) 7.0, 11.0 Hz,
1H, H-5′), 4.30-4.35 (m, 1H, H-4′), 4.36-4.46 (m, 1H, NHCH),
5.0 (dd, J ) 2.8, 6.2 Hz, 1H, H-3′), 5.36 (dd, J ) 2.3, 6.2 Hz, 1H,
H-2′), 6.16 (d, J ) 2.1 Hz, 1H, H-1′), 8.32 (s, 1H, H-8), 8.40 (d,
J ) 7.3 Hz, 1H, NH). MS: m/z 429.4 [M + H]+. Anal.
(C18H23Cl2N5O3) C, H, N.
2-Chloro-N6-(()-endo-norbornyl-9H-(2,3-O-isopropylidene-
5-chloro-5-deoxy-ꢀ-D-ribofuranosyl)adenine (18). The title com-
pound was synthesized from 17. Chromatography on a silica gel
General Procedure for the Synthesis of Compounds 33-36
and 40. To an ice cooled solution of 2-fluorothiophenol (4.74 mmol)
in dry DMF (10 mL), under nitrogen atmosphere, was added NaH
60% (3.84 mmol) in mineral oil portionwise. When the development
of H2 was complete, compounds 29-32 or 39 (1.0 mmol) were
added and the mixture was stirred at room temperature for the time
reported below. The solvent was removed in vacuo, and the residue
was purified by column chromatography.
1
column (CHCl3) gave 18 as a white foam (70% yield). H NMR
(DMSO-d6) δ 1.25-1.32 (m, 3H, norbornyl), 1.38, 1.55 (2s, 6H,
CH3), 1.36-1.43 (m, 3H, norbornyl), 1.55-1.62 (m, 1H, norbornyl),
1.88-1.95 (m, 1H, norbornyl), 2.18 (br s, 1H, norbornyl), 2.52 (s,
1H, norbornyl), 3.75 (dd, J ) 6.8, 11.1 Hz, 1H, H-5′), 3.85 (dd, J
) 7.0, 11.0 Hz, 1H, H-5′), 4.31-4.36 (m, 1H, H-4′), 4.39-4.44
(m, 1H, NHCH), 5.0 (dd, J ) 2.8, 6.2 Hz, 1H, H-3′), 5.32 (dd, J
) 2.3, 6.2 Hz, 1H, H-2′), 6.15 (d, J ) 2.1 Hz, 1H, H-1′), 8.30 (s,
1H, H-8), 8.40 (d, J ) 7.3 Hz, 1H, NH). MS: m/z 455.4 [M +
H]+. Anal. (C20H25Cl2N5O3) C, H, N.
N6-Cyclopentyl-9H-[2-C-methyl-2,3-O-isopropylidene-5-deoxy-
5-(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]adenine (33). The title
compound was synthesized from 29 (reaction time 6 h). Chroma-
tography on a silica gel column (CHCl3-MeOH, 99.5:0.5) gave
1
33 as a foam (91% yield). H NMR (DMSO-d6) δ 1.15 (s, 3H,
N6-Cyclopentyl-9H-(2-C-methyl-2,3-O-isopropylidene-5-chloro-
5-deoxy-ꢀ-D-ribofuranosyl)adenine (29). The title compound was
synthesized from 25.3a Chromatography on a silica gel column
(CHCl3-MeOH, 99.5:0.5) gave 29 as a white foam (52% yield).
1H NMR (DMSO-d6) δ 1.20 (s, 3H, CH3), 1.40, 1.55 (2s, 3H, CH3),
1.52-1.62 (m, 4H, cyclopentyl), 1.62-1.72 (m, 2H, cyclopentyl),
1.86-1.98 (m, 2H, cyclopentyl), 4.07 (dq, J ) 6.4, 11.1 Hz, 2H,
H-5′), 4.36 (dt, J ) 3.1, 6.3 Hz, 1H, H-4′), 4.43-4.52 (m, 1H,
NHCH), 4.63 (d, J ) 3.0 Hz, 1H, H-3′), 6.30 (s, 1H, H-1′), 7.80
(d, J ) 6.4 Hz, 1H, NH), 8.20 (2s, 2H, H-2, H-8). MS: m/z 408.9
[M + H]+. Anal. (C19H26ClN5O3) C, H, N.
CH3), 1.35, 1.50 (2s, 6H, CH3), 1.50-1.62 (m, 4H, cyclopentyl),
1.65-1.75 (m, 2H, cyclopentyl), 1.86-1.98 (m, 2H, cyclopentyl),
3.48 (d, J ) 6.4 Hz, 2H, H-5′), 4.25 (dt, J ) 3.0, 6.6 Hz, 1H,
H-4′), 4.42-4.54 (m, 1H, NHCH), 4.60 (d, J ) 3.0 Hz, 1H, H-3′),
6.20 (s, 1H, H-1′), 7.15-7.32 (m, 3H, arom.), 7.53 (t, J ) 6.4 Hz,
1H, arom.), 7.80 (d, J ) 7.0 Hz, 1H, NH), 8.10 (s, 1H, H-2), 8.20
(s, 1H, H-8). MS: m/z 500.9 [M + H]+. Anal. (C25H30FN5O3S) C,
H, N.
2-Chloro-N6-cyclopentyl-9H-[2-C-methyl-2,3-O-isopropylidene-
5-deoxy-5-(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]adenine (34).
The title compound was synthesized from 30 (reaction time 8 h).
Chromatography on a silica gel column (CHCl3) gave 34 as an oil
(91% yield). 1H NMR (DMSO-d6) δ 1.16 (s, 3H, CH3), 1.35, 1.48
(2s, 6H, CH3), 1.50-1.62 (m, 4H, cyclopentyl), 1.64-1.75 (m, 2H,
cyclopentyl), 1.85-2.0 (m, 2H, cyclopentyl), 3.46 (d, J ) 6.8 Hz,
2H, H-5′), 4.18-4.22 (m, 1H, H-4′), 4.38-4.45 (m, 1H, NHCH),
4.60 (d, J ) 3.0 Hz, 1H, H-3′), 6.12 (s, 1H, H-1′), 7.15-7.30 (m,
3H, arom.), 7.52 (t, J ) 7.3 Hz, 1H, arom.), 8.10 (s, 1H, H-8),
8.40 (d, J ) 7.7 Hz, 1H, NH). MS: m/z 535.1 [M + H]+. Anal.
(C25H29ClFN5O3S) C, H, N.
2-Chloro-N6-cyclopentyl-9H-(2-C-methyl-2,3-O-isopropylidene-
5-chloro-5-deoxy-ꢀ-D-ribofuranosyl)adenine (30). The title com-
pound was synthesized from 26.3a Chromatography on a silica gel
column (CHCl3) gave 30 as a foam (50% yield). 1H NMR (DMSO-
d6) δ 1.18 (s, 3H, CH3), 1.40, 1.56 (2s, 6H, CH3), 1.50-1.60 (m,
4H, cyclopentyl), 1.62-1.72 (m, 2H, cyclopentyl), 1.88-1.98 (m,
2H, cyclopentyl), 4.06 (dq, J ) 5.2, 11.5 Hz, 2H, H-5′), 4.35 (q, J
) 7.3 Hz, 1H, H-4′), 4.40-4.48 (m, 1H, NHCH), 4.60 (d, J ) 3.0
Hz, 1H, H-3′), 6.20 (s, 1H, H-1′), 8.25 (s, 1H, H-8), 8.40 (d, J )
7.7 Hz, 1H, NH). MS: m/z 443.4 [M + H]+. Anal. (C19H25Cl2N5O3)
C, H, N.
N6-(()-endo-Norbornyl-9H-[2-C-methyl-2,3-O-isopropylidene-
5-deoxy-5-(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]adenine (35).
The title compound was synthesized from 31 (reaction time 6 h).
Chromatography on a silica gel column (CHCl3) gave 35 as a white
solid (58% yield). 1H NMR (DMSO-d6) δ 1.15 (s, 3H, CH3), 1.25
(m, 3H, norbornyl), 1.35, 1.50 (2s, 6H, CH3), 1.39-1.48 (m, 3H,
norbornyl), 1.57-1.64 (m, 1H, norbornyl), 1.83-1.94 (m, 1H,
norbornyl), 2.16 (br s, 1H, norbornyl), 2.52 (br s, 1H, norbornyl),
3.50 (d, J ) 6.8 Hz, 2H, H-5′), 4.18-4.23 (m, 1H, H-4′), 4.31-4.38
(m, 1H, NHCH), 4.65 (br s, 1H, H-3′), 6.20 (s, 1H, H-1′), 7.15-7.30
(m, 3H, arom.), 7.51-7.58 (m, 1H, arom.), 7.85 (d, J ) 6.8 Hz,
1H, NH), 8.10 (s, 1H, H-2), 8.20 (s, 1H, H-8). MS: m/z 526.6 [M
+ H]+. Anal. (C27H32FN5O3S) C, H, N.
N6-(()-endo-Norbornyl-9H-(2-C-methyl-2,3-O-isopropylidene-
5-chloro-5-deoxy-ꢀ-D-ribofuranosyl)-adenine (31). The title com-
pound was synthesized from 27. Chromatography on a silica gel
1
column (CHCl3) gave 31 as a white solid (60% yield). H NMR
(DMSO-d6) δ 1.20 (2s, 3H, CH3), 1.23-1.32 (m, 3H, norbornyl),
1.35, 1.55 (2s, 6H, CH3), 1.39-1.45 (m, 3H, norbornyl), 1.58-1.62
(m, 1H, norbornyl), 1.84-1.93 (m, 1H, norbornyl), 2.15 (br s, 1H,
norbornyl), 2.52 (br s, 1H, norbornyl), 4.05 (dq, J ) 6.3, 10.5 Hz,
2H, H-5′), 4.27-4.36 (m, 2H, NHCH, H-4′), 4.65 (d, J ) 2.6 Hz,
1H, H-3′), 6.28 (s, 1H, H-1′), 7.85 (br s, 1H, NH), 8.20 (2s, 2H,
H-2, H-8). MS: m/z 434.9 [M + H]+. Anal. (C21H28ClN5O3) C, H,
N.
2-Chloro-N6-(()-endo-norbornyl-9H-[2-C-methyl-2,3-O-iso-
propylidene-5-deoxy-5-(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]-
adenine (36). The compound was synthesized from 32 (reaction
time 7 h) and purified by chromatography on a silica gel column
(CHCl3-MeOH, 99.5:0.5) as a white foam (81% yield). 1H NMR
(DMSO-d6) δ 1.15 (2s, 3H, CH3), 1.22-1.30 (m, 3H, norbornyl),
1.35, 1.50 (2s, 6H, CH3), 1.40-1.48 (m, 3H, norbornyl), 1.52-1.66
(2m, 1H, norbornyl), 1.82-1.98 (2m, 1H, norbornyl), 2.15 (br s,
2-Chloro-N6-(()-endo-norbornyl-9H-(2-C-methyl-2,3-O-iso-
propylidene-5-chloro-5-deoxy-ꢀ-D-ribofuranosyl)adenine (32).
The title compound was synthesized from 28. Chromatography on
a silica gel column (CHCl3) gave 32 as a white foam (69% yield).
1H NMR (DMSO-d6) δ 1.20 (s, 3H, CH3), 1.25-1.32 (m, 3H,
norbornyl), 1.40, 1.55 (2s, 6H, CH3), 1.38-1.48 (m, 3H, norbornyl),
1.55-1.60 (m, 1H, norbornyl), 1.85-1.95 (m, 1H, norbornyl), 2.15

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2403
1H, norbornyl), 2.52 (br s, 1H, norbornyl), 3.46 (d, J ) 6.8 Hz,
2H, H-5′), 4.18-4.28 (m, 2H, NHCH, H-4′), 4.60 (br s, 1H, H-3′),
6.12 (s, 1H, H-1′), 7.12-7.32 (m, 3H, arom.), 7.52 (t, J ) 7.7 Hz,
1H, arom.), 8.13 (2s, 1H, H-8), 8.45 (d, J ) 6.4 Hz, 1H, NH). MS:
m/z 561.1 [M + H]+. Anal. (C27H31ClFN5O3S) C, H, N.
2-Chloro-N6-(()-endo-norbornyl-9H-(2-C-methyl-5-chloro-5-
deoxy-ꢀ-D-ribofuranosyl)adenine (8). The title compound was
synthesized from 32 (reaction time 7 h). Chromatography on a silica
gel column (CH3Cl-MeOH, 99:1) gave 8 as a white solid (81%
1
yield). H NMR (DMSO-d6) δ 0.80 (s, 3H, CH3), 1.22-1.27 (m,
2-Chloro-N6-cyclopentyl-9H-[2,3-O-isopropylidene-5-deoxy-
5-(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]adenine (40). The title
compound was synthesized from 39 (reaction time 5 h) and purified
by chromatography on a silica gel column (CH3Cl-MeOH, 99.5:
3H, norbornyl), 1.39-1.44 (m, 3H, norbornyl), 1.56-1.64 (2m,
1H, norbornyl), 1.92-1.95 (2m, 1H, norbornyl), 2.16 (br s, 1H,
norbornyl), 2.52 (br s, 1H, norbornyl), 3.97-4.09 (m, 4H, H-5′,
H-4′, H-3′), 4.25 (br s, 1H, NHCH), 5.40 (s, 1H, OH), 5.50 (d, J
) 6.0 Hz, 1H, OH), 5.90 (s, 1H, H-1′), 8.28 (s, 1H, H-8), 8.40 (d,
J ) 6.4 Hz, 1H, NH). MS: m/z 429.3 [M + H]+. Anal.
(C18H23Cl2N5O3) C, H, N.
1
0.5) as a white foam (95% yield). H NMR (DMSO-d6) δ 1.30,
1.45 (2s, 6H, CH3), 1.50-1.60 (m, 4H, cyclopentyl), 1.62-1.72
(m, 2H, cyclopentyl), 1.86-1.98 (m, 2H, cyclopentyl), 3.25 (d, J
) 6.8 Hz, 2H, H-5′), 4.25 (dt, J ) 2.6, 7.0 Hz, 1H, H-4′), 4.36-4.40
(m, 1H, NHCH), 5.0 (dd, J ) 2.8, 6.2 Hz, 1H, H-3′), 5.48 (dd, J
) 2.1, 6.4 Hz, 1H, H-2′), 6.20 (d, J ) 1.7 Hz, 1H, H-1′), 7.10-7.28
(m, 3H, arom.), 7.45 (t, J ) 7.3 Hz, 1H, arom.), 8.33 (s, 1H, H-8),
8.38 (d, J ) 7.3 Hz, 1H, NH). MS: m/z 521.1 [M + H]+. Anal.
(C24H27ClFN5O3S) C, H, N.
2-Chloro-N6-cyclopentyl-9H-[5-deoxy-5-(2-fluorophenylthio)-
ꢀ-D-ribofuranosyl]adenine (10). The title compound was synthe-
sized from 40 (reaction time 7 h). Chromatography on a silica gel
column (CHCl3-MeOH, 99.5:0.5) gave 10 as a white foam (64%
1
yield). H NMR (DMSO-d6) δ 1.50-1.60 (m, 4H, cyclopentyl),
1.62-1.72 (m, 2H, cyclopentyl), 1.88-2.0 (m, 2H, cyclopentyl),
3.30 (dd, J ) 7.3, 13.7 Hz, 1H, H-5′), 3.40 (dd, J ) 5.8, 13.9 Hz,
1H, H-5′), 3.95-4.05 (m, 1H, H-4′), 4.14 (q, J ) 4.9 Hz, 1H, H-3′),
4.35-4.45 (m, 1H, NHCH), 4.72 (q, J ) 6.0 Hz, 1H, H-2′), 5.42
(d, J ) 3.7 Hz, 1H, OH), 5.56 (d, J ) 6.0 Hz, 1H, OH), 5.80 (d,
J ) 6.0 Hz, 1H, H-1′), 7.10-7.26 (m, 3H, arom.), 7.45 (t, J ) 7.9
Hz, 1H, arom.), 8.32 (s and d, 2H, H-8, NH). MS: m/z 481.0 [M +
H]+. Anal. (C21H23ClFN5O3S) C, H, N.
General Procedure for the Synthesis of Compounds 4-8
and 10-14. Compounds 18, 29-36, and 40 (1.0 mmol) were
treated with HCOOH 70% in water (10 mL), and the mixture was
stirred at 40 °C for the time reported below. The solvent was
evaporated in vacuo, and the residue was coevaporated several times
with CH3OH and then purified by column chromatography.
2-Chloro-N6-(()-endo-norbornyl-9H-(5-chloro-5-deoxy-ꢀ-D-
ribofuranosyl)adenine (4). The title compound was synthesized
from 18 (reaction time 6 h). Chromatography on a silica gel column
N6-Cyclopentyl-9H-[2-C-methyl-5-deoxy-5-(2-fluorophenylthio)-
ꢀ-D-ribofuranosyl]adenine (11). The title compound was synthe-
sized from 33 (reaction time 17 h). Chromatography on a silica
gel column (CHCl3-MeOH, 99.5:0.5) gave 11 as a white solid
(82% yield). 1H NMR (DMSO-d6) δ 0.80 (s, 3H, CH3), 1.50-1.62
(m, 4H, cyclopentyl), 1.64-1.76 (m, 2H, cyclopentyl), 1.82-1.98
(m, 2H, cyclopentyl), 3.45 (d, J ) 5.6 Hz, 2H, H-5′), 4.05 (dt, J )
1.5, 5.6 Hz, 1H, H-4′), 4.12-4.18 (m, 1H, H-3′), 4.44-4.54 (m,
1H, NHCH), 5.30 (s, 1H, OH), 5.48 (d, J ) 6.0 Hz, 1H, OH), 5.90
(s, 1H, H-1′), 7.10-7.26 (m, 3H, arom.), 7.46 (t, J ) 7.7 Hz, 1H,
arom.), 7.76 (d, J ) 6.8 Hz, 1H, NH), 8.20 (s, 1H, H-2), 8.24 (s,
1H, H-8). MS: m/z 460.6 [M + H]+. Anal. (C22H26FN5O3S) C, H,
N.
1
(CHCl3-MeOH, 98:2) gave 4 as a white foam (60% yield). H
NMR (DMSO-d6) δ 1.21-1.30 (m, 3H, norbornyl), 1.33-1.45 (m,
3H, norbornyl), 1.50-1.65 (2m, 1H, norbornyl), 1.83-1.96 (2m,
1H, norbornyl), 2.16 (br s, 1H, norbornyl), 2.52 (br s, 1H,
norbornyl), 3.81-3.88 (m, 1H, H-5′), 3.90-3.96 (m, 1H, H-5′),
4.08 (q, J ) 5.6 Hz, 1H, H-4′), 4.14-4.18 (m, 1H, H-3′), 4.21-4.27
(m, 1H, NHCH), 4.61-4.66 (m, 1H, H-2′), 5.48 (d, J ) 5.1 Hz,
1H, OH), 5.62 (d, J ) 6.0 Hz, 1H, OH), 5.85 (d, J ) 6.0 Hz, 1H,
H-1′), 8.37 (s, 1H, H-8), 8.40 (d, J ) 6.8 Hz, 1H, NH). MS: m/z
415.3 [M + H]+. Anal. (C17H21Cl2N5O3) C, H, N.
N6-Cyclopentyl-9H-(2-C-methyl-5-chloro-5-deoxy-ꢀ-D-ribo-
furanosyl)adenine (5). The title compound was synthesized from
29 (reaction time 6 h). Chromatography on a silica gel column
2-Chloro-N6-cyclopentyl-9H-[2-C-methyl-5-deoxy-5-(2-fluo-
rophenylthio)-ꢀ-D-ribofuranosyl]adenine (12). The title com-
pound was synthesized from 34 (reaction time 9 h). Chromatog-
raphy on a silica gel column (CHCl3-MeOH, 99.5:0.5) gave 12
1
(CHCl3-MeOH, 99:1) gave 5 as a white foam (60% yield). H
NMR (CDCl3) δ 1.05 (s, 3H, CH3), 1.52-1.82 (m, 6H, cyclopentyl),
2.12-2.18 (m, 2H, cyclopentyl), 3.88 (dd, J ) 4.9, 11.7 Hz, 1H,
H-5′), 3.94 (dd, J ) 4.7, 11.0 Hz, 1H, H-5′), 4.15 (br s, 1H, H-4′),
4.35 (q, J ) 5.1 Hz, 1H, H-3′), 4.57-4.62 (m, 1H, NHCH), 5.60
(br s, 1H, OH), 5.85 (br s, 1H, OH), 6.0 (s, 1H, H-1′), 8.0 (br s,
2H, H-2, NH), 8.38 (s, 1H, H-8). MS: m/z 368.8 [M + H]+. Anal.
(C16H22ClN5O3) C, H, N.
1
as a white solid (58% yield). H NMR (DMSO-d6) δ 0.80 (s, 3H,
CH3), 1.46-1.61 (m, 4H, cyclopentyl), 1.62-1.73 (m, 2H, cyclo-
pentyl), 1.83-1.97 (m, 2H, cyclopentyl), 3.42-3.52 (m, 2H, H-5′),
4.0 (dt, J ) 3.7, 8.7 Hz, 1H, H-4′), 4.07-4.12 (m, 1H, H-3′),
4.37-4.42 (m, 1H, NHCH), 5.35 (br s, 1H, OH), 5.50 (d, J ) 6.4
Hz, 1H, OH), 5.85 (s, 1H, H-1′), 7.10 (t, J ) 7.3 Hz, 1H, arom.),
7.17-7.23 (m, 2H, arom.), 7.45 (t, J ) 7.7 Hz, 1H, arom.), 8.25
(s, 1H, H-8), 8.38 (d, J ) 7.3 Hz, 1H, NH). MS: m/z 495.0 [M +
H]+. Anal. (C22H25ClFN5O3S) C, H, N.
2-Chloro-N6-cyclopentyl-9H-(2-C-methyl-5-chloro-5-deoxy-ꢀ-
D-ribofuranosyl)adenine (6). The title compound was synthesized
from 30 (reaction time 12 h). Chromatography on a silica gel
column (CH3Cl-MeOH, 99:1) gave 6 as a white foam (70% yield).
1H NMR (DMSO-d6) δ 0.80 (s, 3H, CH3), 1.47-1.52 (m, 4H,
cyclopentyl), 1.66-1.74 (m, 2H, cyclopentyl), 1.83-1.97 (m, 2H,
cyclopentyl), 3.98-4.10 (m, 4H, H-5′, H-4′, H-3′), 4.37-4.43 (m,
1H, NHCH), 5.40 (s, 1H, OH), 5.52 (d, J ) 6.4 Hz, 1H, OH), 5.90
(s, 1H, H-1′), 8.24 (s, 1H, H-8), 8.35 (d, J ) 6.8 Hz, 1H, NH).
MS: m/z 403.3 [M + H]+. Anal. (C16H21Cl2N5O3) C, H, N.
N6-(()-endo-Norbornyl-9H-[2-C-methyl-5-deoxy-5-(2-fluo-
rophenylthio)-ꢀ-D-ribofuranosyl]adenine (13). The title com-
pound was synthesized from 35 (reaction time 6 h). Chromatog-
raphy on a silica gel column (CHCl3-MeOH, 99.5:0.5) gave 13
1
as a white solid (60% yield). H NMR (DMSO-d6) δ 0.80 (s, 3H,
CH3), 1.19-1.27 (m, 3H, norbornyl), 1.38-1.43 (m, 3H, norbornyl),
1.56-1.62 (m, 1H, norbornyl), 1.83-1.92 (m, 1H, norbornyl), 2.15
(br s, 1H, norbornyl), 2.52 (br s, 1H, norbornyl), 3.50 (d, J ) 6.8
Hz, 2H, H-5′), 3.97-4.03 (m, 1H, H-4′), 4.13-4.21 (m, 1H, H-3′),
4.31-4.39 (m, 1H, NHCH), 5.30 (s, 1H, OH), 5.45 (d, J ) 6.8
Hz, 1H, OH), 5.93 (s, 1H, H-1′), 7.12 (t, J ) 7.3 Hz, 1H, arom.),
7.18-7.26 (m, 2H, arom.), 7.45 (t, J ) 7.7 Hz, 1H, arom.), 7.82
(d, J ) 6.7 Hz, NH), 8.20 (s, 1H, H-2), 8.25 (s, 1H, H-8). MS: m/z
486.6 [M + H]+. Anal. (C24H28FN5O3S) C, H, N.
N6-(()-endo-Norbornyl-9H-(2-C-methyl-5-chloro-5-deoxy-ꢀ-
D-ribofuranosyl)adenine (7). The title compound was synthesized
from 31 (reaction time 10 h). Chromatography on a silica gel
column (CH3Cl-MeOH 98:2) gave 7 as a white solid (67% yield).
1H NMR (DMSO-d6) δ 0.80 (s, 3H, CH3), 1.21-1.29 (m, 3H,
norbornyl), 1.39-1.45 (m, 3H, norbornyl), 1.57-1.64 (m, 1H,
norbornyl), 1.83-1.95 (m, 1H, norbornyl), 2.16 (br s, 1H, nor-
bornyl), 2.52 (br s, 1H, norbornyl), 3.96-4.04 (m, 2H, H-5′),
4.06-4.11 (m, 1H, H-4′), 4.12-4.18 (m, 1H, H-3′), 4.31-4.39 (m,
1H, NHCH), 5.35 (s, 1H, OH), 5.48 (d, J ) 6.4 Hz, 1H, OH), 5.98
(s, 1H, H-1′), 7.80 (br s, 1H, NH), 8.20 (s, 1H, H-2), 8.22 (s, 1H,
H-8). MS: m/z 394.9 [M + H]+. Anal. (C18H24ClN5O3) C, H, N.
2-Chloro-N6-(()-endo-norbornyl-9H-[2-C-methyl-5-deoxy-5-
(2-fluorophenylthio)-ꢀ-D-ribofuranosyl]adenine (14). The title
compound was synthesized from 36 (reaction time 13 h). Chro-
matography on a silica gel column (CH3Cl-MeOH, 99.5:0.5) gave
1
14 as a white solid (75% yield). H NMR (DMSO-d6) δ 0.80 (s,
3H, CH3), 1.22-1.28 (m, 3H, norbornyl), 1.32-1.44 (m, 3H,

2404 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
norbornyl), 1.52-1.68 (2m, 1H, norbornyl), 1.84-1.98 (2m, 1H,
norbornyl), 2.15 (br s, 1H, norbornyl), 2.52 (br s, 1H, norbornyl),
3.42-3.50 (m, 2H, H-5′), 3.38-4.02 (m, 1H, H-4′), 4.04-4.12 (m,
1H, H-3′), 4.22 (br s, 1H, NHCH), 5.33 (s, 1H, OH), 5.50 (d, J )
6.8 Hz, 1H, OH), 5.85 (s, 1H, H-1′), 7.12 (t, J ) 7.3 Hz, 1H, arom.),
7.20-7.28 (m, 3H, arom.), 7.45 (t, J ) 7.7 Hz, 1H, arom.), 8.25
(s, 1H, H-8), 8.42 (d, J ) 6.4 Hz, 1H, NH). MS: m/z 521.1 [M +
H]+. Anal. (C24H27ClFN5O3S) C, H, N.
Computational Chemistry. Molecular modeling and graphics
manipulations were performed using the molecular operating
environment (MOE)31 and UCSF-CHIMERA software packages,32
running on a 2 CPU (PIV 2.0-3.0 GHZ) Linux workstation. Energy
minimizations and MD simulations were realized by employing
the AMBER 9 program,33 selecting the Cornell et al. force field.34
Residue Indexing. The convention used for the amino acid
identifiers, according to the approach of Ballesteros and Weinstein,35
facilitates comparison of aligned residues within the family of Group
A GPCRs. To the most conserved residue in a given TM (TMX,
where X is the TM number) is assigned the number X.50, and
residues within a given TM are then indexed relative to the 50
position.
Construction of the hA1 AR and hA3 AR Homology Models.
The structural models of hA1 AR and hA3 AR were built using the
recently reported 2.8 Å crystal structure of b-Rho (PDB entry code
1F88) as a structural template.13 We modeled only the TM domains,
since the function of the loops has still not been defined. Although
site-directed mutagenesis suggests a role for AR loops, and in
particular for the second extracellular (E2) ones, it remains unclear
whether the E2 loop is in direct contact with ligands or whether it
contributes to the overall physical architecture of the receptor
protein.36 Briefly, the hA1 AR and hA3 AR sequences were retrieved
from the SWISS-PROT database37 and aligned with the sequence
of b-Rho using CLUSTALW software38 with the following settings:
matrix ) Blosum series; gap opening penalty ) 10; gap extension
penalty ) 0.05. Afterward, we checked and, where necessary,
manually corrected this alignment to reflect the known alignment
features of class A GPCRs, such as the highly conserved positions
and gap-free TM regions. In particular, the alignment was guided
by the highly conserved amino acid residues, including the D/ERY
motif (D/E3.49, R3.50, and Y3.51), the two proline residues P4.50
and P6.50, and the NPXXY motif in TM7 (N7.49, P7.50, and
Y7.53).39 Extension of each helix was contemplated by taking into
account the experimental length of the b-Rho helices and the
secondary structure prediction of both hA1 AR and hA3 AR obtained
with the PSIPRED software,40 as well as the sequence conservation
in the possible extensions of the helices. Individual TM helical
segments were built as ideal helices (using φ and Ψ angles of
-63.0° and -41.6°, respectively) with side chains placed in
prevalent rotamers and representative proline kink geometries. Each
model helix was capped with an acetyl group at the N-terminus
and a N-methyl group at the C-terminus. These structures were then
grouped by adding one at a time until a helical bundle (TM region),
matching the overall characteristics of the crystallographic structure
of b-Rho, was obtained. The hA1 AR and hA3 AR helical bundles
were subjected to a preliminary minimization and 200 ps of MD,
after which the final structures were minimized. When MD
simulations are carried out in the gas phases, skipping the explicit
environment requires the use of a set of restraints, to replace the
natural stabilizing effects of the membrane bilayer on the TM
domains. Accordingly, restraints with a force constant of 10 kcal
mol-1 Å-2 were applied to backbone for the first 100 ps, and for
the remaining 100 ps, these restraints were reduced to 1 kcal mol-1
Å-2. The options of MD at 300 K with a 0.2 ps coupling constant
were a time step of 1 fs and a nonbonded update every 25 fs. The
lengths of bonds with hydrogen atoms were constrained according
to the SHAKE algorithm.41 The average structure from the last 50
ps trajectory of MD was reminimized with backbone constraints
in the secondary structure.
is that described by Shi et al.27 When the heavy atom at the γ
position is at a position opposite to the backbone nitrogen when
viewed from the ꢀ-carbon to the R-carbon, the ꢁ1 is defined to be
trans. When the heavy atom at the ꢁ1 position is at a position
opposite to the backbone carbon when viewed from the ꢀ-carbon
to the R-carbon, the ꢁ1 is defined to be gauche+. When the heavy
atom at the γ position is at a position opposite to the R-hydrogen
when viewed from the ꢀ-carbon to the R-carbon, the ꢁ1 is defined
to be gauche-. The stabilities of three different ꢁ1 angles of W6.48
set at 60°, 180°, and -60° were compared. A minimized gauche+
conformation with a ꢁ1 angle of -98° in the ground-state had the
lowest energy among three different geometries. A gauche-
conformer of W6.48 with the highest energy seemed to be similar
to the Meta I state conformation, because it displayed the most
outward anticlockwise rotation from the extracellular view, as b-Rho
studies suggested. This putative Meta I state of hA1 AR and hA3
AR was used for agonist docking.
Docking Simulations. The core structures of compounds CPA,
(()-ENBA, 1, and 3 were retrieved from the Cambridge Structural
Database (CSD)42 and modified using standard bond lengths and
bond angles of the MOE fragment library. Since Trivedi et al.4a
reported that the 1R,2S,4S isomer of the N6-(2-endo-norbornyl)
system of (()-ENBA was more potent than the 1S,2R,4R one at
the rat A1 AR, we considered only the 1R,2S,4S isomer of (()-
ENBA and 3 for docking calculations. Geometry optimizations of
compounds were accomplished with the MMFF94 force field,43
available within MOE. CPA, chosen as a reference compound, was
manually docked into both hA1 AR and hA3 AR binding sites,
bearing in mind the known mutagenesis data. As regards hA1 AR,
CPA was docked in such a manner as to give H-bonds with T91
(3.36),25 N254 (6.55), T277 (7.42),44 and H278 (7.43)45 and a
lipophilic interaction (through the cyclopentyl moiety) with L88
(3.33),25 in accordance with the main mutagenesis data and our
previous computational studies.3a,4b In the case of hA3 AR, CPA
was manually introduced into the binding site, considering the
interactions with T94 (3.36),28 N250 (6.55),25 and H272 (7.43),23,46
suggested by mutagenesis data. Compounds 1 and 3 present the
adenine group as their central core, and their initial docking position
into both hA1 AR and hA3 AR binding sites was obtained by
superimposing this group on that of the final structure of CPA in
either hA1 AR and hA3 AR, respectively. In this position, the two
ligands exhibited the interactions suggested by mutagenesis data.
Molecular Dynamics Simulations. Refinement of the ligand/
receptor complexes was achieved by in vacuo energy minimization
with the SANDER module of AMBER, applying an energy penalty
force constant of 10 kcal ·mol-1 ·Å-2 on the protein backbone atoms.
The geometry-optimized complexes were then used as the starting
point for subsequent 1 ns MD simulation, during which the protein
backbone atoms were constrained by means of decreasing force
constants; moreover, also the main ligand/receptor interactions were
restrained. More specifically, an initial restraint with a force constant
of 10 kcal·mol-1 ·Å-2 was applied on all the R carbons; this force
constant decreased during the whole MD, and in the last 200 ps,
its value was 0.1 kcal · mol-1 · Å-2. As regards the main ligand/
receptor interactions, a restraint of 50 kcal·mol-1 ·Å-2 was applied
for 700 ps of MD simulation and, in the last 300 ps, the restraint
was removed. General AMBER force field (GAFF) parameters were
assigned to ligands, while the partial charges were calculated using
the AM1-BCC method as implemented in the ANTECHAMBER
suite of AMBER. A time step of 1 fs and a nonbonded pairlist
updated every 25 fs were used for the MD simulations. The
temperature was regulated by way of Langevin dynamics, with a
collision frequency γ ) 1.0 ps-1. An average structure was
calculated from the last 200 ps trajectory and energy-minimized
using the steepest descent and conjugate gradient methods as
specified above. RMSDs from the initial structures and interatomic
distances were monitored using the PTRAJ module in AMBER.
Binding Assay and Adenylyl Cyclase Assay at Cloned
Human Adenosine Receptors. Ki-values were determined in
competition experiments with membranes from CHO cells stably
transfected with the individual human adenosine receptor sub-
Definition of the Rotameric State of ꢁ1. Different nomenclatures
have been used to define the rotameric state of side chain torsion
angles. The nomenclature employed here for the ꢁ1 torsion angle

Agonists at the Human A1 Adenosine Receptor
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8 2405
types.11 For A1 AR 1 nM [3H]CCPA was used as a radioligand,
[3H]NECA was used for the A2A (30 nM), and [3H]HEMADO was
used for the A3 (1 nM) subtype. In the case of the A2B receptor, Ki
values were calculated from IC50 values determined by inhibition
of NECA-stimulated adenylyl cyclase activity.11 All binding data
were calculated by nonlinear curve fitting with the program
SCTFIT.47 The functional activity of selected derivatives at the A1
receptor was determined in adenylyl cyclase experiments. The
inhibition of forskolin-stimulated adenylyl cyclase via A1 and A3
receptors was measured as described earlier.12,48
5′-Carbamoyl derivatives of 2′-C-methyl-purine nucleosides as selec-
tive A1 adenosine receptor agonists: Affinity, efficacy, and selectivity
for A1 receptor from different species. Bioorg. Med. Chem. 2008, 16,
336–353. (b) Ashton, T. D.; Baker, S. P.; Hutchinson, S. A.;
Scammells, P. J. N6-substituted C5′-modified adenosines as A1
adenosine receptor agonists. Bioorg. Med. Chem. 2008, 16, 1861-
1873, and references reported therein.
(4) (a) Trivedi, B. K.; Bridges, A. J.; Patt, W. C.; Priebe, S. R.; Bruns,
R. F. N6-Bicycloalkyladenosines with unusually high potency and
selectivity for the adenosine A1 receptor. J. Med. Chem. 1989, 32,
8–11. (b) Cappellacci, L.; Franchetti, P.; Pasqualini, M.; Petrelli, R.;
Vita, P.; Lavecchia, A.; Novellino, E.; Costa, B.; Martini, C.; Klotz,
K.-N.; Grifantini, M. Synthesis, biological evaluation, and molecular
modeling of ribose-modified adenosine analogues as adenosine receptor
agonists. J. Med. Chem. 2005, 48, 1550–1562. (c) Elzein, E.; Kalla,
R.; Li, X.; Perry, T.; Marquart, T.; Micklatcher, M.; Li, Y.; Wu, Y.;
Zeng, D.; Zablocki, J. A. N6-Cycloalkyl-2-substituted adenosine
derivatives as selective, high affinity adenosine A1 receptor agonists.
Bioorg. Med. Chem. Lett. 2007, 17, 161–166.
(5) (a) van der Wenden, E. M.; Carnielli, M.; Roelen, H. C.; Lorenzen,
A.; von Frijtag Drabbe Kunzel, J. K.; Ijzerman, A. P. 5′-Substituted
adenosine analogs as new high-affinity partial agonists for the
adenosine A1 receptor. J. Med. Chem. 1998, 41, 102–108. (b) Morrison,
C. F.; Elzein, E.; Jiang, B.; Ibrahim, P. N.; Marquart, T.; Palle, V.;
Shenk, K. D.; Varkhedkar, V.; Maa, T.; Wu, L.; Wu, Y.; Zeng, D.;
Fong, I.; Lustig, D.; Leung, K.; Zablocki, J. A. Structure-affinity
relationships of 5′-aromatic ethers and 5′-aromatic sulfides as partial
A1 adenosine agonists, potential supraventricular anti-arrhythmic
agents. Bioorg. Med. Chem. Lett. 2004, 14, 3793–3797.
(6) (a) Abdel-Hamid, M.; Novotny, M.; Hamza, H. Stability study of
selected adenosine nucleosides using LC and LC/MS analyses.
J. Pharm. Biomed. Anal. 2000, 22, 745–755. (b) Bennett, E. M.; Li,
C.; Allan, P. W.; Parker, W. B.; Ealick, S. E. Structural basis for
substrate specificity of Escherichia coli purine nucleoside phospho-
rylase. J. Biol. Chem. 2003, 278, 47110–47118.
(7) Hou, X.; Lee, H. W.; Tosh, D. K.; Zhao, L. X.; Jeong, L. S. Alternative
and improved syntheses of highly potent and selective A3 adenosine
receptor agonists, Cl-IB-MECA and thio-Cl-IB-MECA. Arch. Pharm.
Res. 2007, 30, 1205–1209.
(8) Franchetti, P.; Cappellacci, L.; Marchetti, S.; Trincavelli, L.; Martini,
C.; Mazzoni, M. R.; Lucacchini, A.; Grifantini, M. 2′-C-Methyl
analogues of selective adenosine receptor agonists: synthesis and
binding studies. J. Med. Chem. 1998, 41, 1708–1715.
(9) Lohse, M. J.; Klotz, K.-N.; Schwabe, U.; Cristalli, G.; Vittori, S.;
Grifantini, M. 2-Chloro-N6-cyclopentyladenosine: a highly selective
agonist at A1 adenosine receptors. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1988, 337, 687–689.
(10) Klotz, K.-N.; Lohse, M. J.; Schwabe, U.; Cristalli, G.; Vittori, S.;
Grifantini, M. 2-Chloro-N6-[3H]cyclopentyladenosine ([3H]CCPA) -
a high affinity agonist radioligand for A1 adenosine receptors. Naunyn-
Schmiedeberg’s Arch. Pharmacol. 1989, 340, 679–683.
(11) Klotz, K.-N.; Hessling, J.; Hegler, J.; Owman, B.; Kull, B.; Fredholm,
B. B.; Lohse, M. J. Comparative pharmacology of human adenosine
receptor subtypes-characterization of stably transfected receptors in
CHO cells. Naunyn-Schmiedeberg’s Arch. Pharmacol. 1998, 357, 1–
9.
(12) Klotz, K.-N.; Kachler, S.; Falgner, N.; Volpini, R.; Dal Ben, D.;
Lambertucci, C.; Mishra, R. C.; Vittori, S.; Cristalli, G. [3H]HE-
MADO - a novel highly potent and selective radiolabeled agonist
for A3 adenosine receptors. Eur. J. Pharmacol. 2007, 556, 14–18.
(13) Palczewski, K.; Kumasaka, T.; Hori, T.; Behnke, C. A.; Motoshima,
H.; Fox, B. A.; Le Trong, I.; Teller, D. C.; Okada, T.; Stenkamp,
R. E.; Yamamoto, M.; Miyano, M. Crystal structure of rhodopsin: a
G protein-coupled receptor. Science 2000, 289, 739–745.
(14) (a) Martinelli, A.; Tuccinardi, T. Molecular modeling of adenosine
receptors: New results and trends. Med. Res. ReV. 2008, 28, 247–277.
(b) Dal Ben, D.; Lambertucci, C.; Vittori, S.; Volpini, R.; Cristalli,
G. GPCRs as therapeutic targets: a view on adenosine receptor
structure and functions, and molecular modeling support. J. Iran.
Chem. Soc. 2005, 2, 176–188. (c) Moro, S.; Deflorian, F.; Bacilieri,
M.; Spalluto, G. Ligand-based homology modeling as attractive tool
to inspect GPCR structural plasticity. Curr. Pharm. Des. 2006, 12,
2175–2185. (d) Costanzi, S.; Ivanov, A. A.; Tikhononva, I. G.;
Jacobson, K. A. Structure and function of G protein-coupled receptors
studied using sequence analysis, molecular modeling, and receptor
engineering: Adenosine receptors. Front. Drug Des. DiscoVery 2007,
3, 63–79. (e) Jacobson, K. A.; Gao, Z.-G.; Liang, B. T. Neoceptors:
Reengineering GPCRs to recognize tailored ligands. Trends Pharma-
col. Sci. 2007, 28, 111–116. (f) Ballesteros, J. A.; Jensen, A. D.;
Liapakis, G.; Rasmussen, S. G. F.; Shi, L.; Gether, U.; Javitch, J. A.
Activation of the beta2-adrenergic receptor involves disruption of an
ionic lock between the cytoplasmic ends of transmembrane segments
3 and 6. J. Biol. Chem. 2001, 276, 29171–29177.
Formalin Test. The experimental procedures applied in the
formalin test were approved by the Animal Ethics Committee of
the Second University of Naples. Animal care was in compliance
with the IASP and European Community guidelines on the use and
protection of animals in experimental research (E.C. L358/118/12/
86). All efforts were made to minimize animal suffering and to
reduce the number of animals used. Formalin injection induces a
biphasic stereotypical nocifensive behavior.29 Nociceptive responses
are divided into an early, short lasting first phase (0-7 min) caused
by a primary afferent discharge produced by the stimulus, followed
by a quiescent period and then a second, prolonged phase (15-60
min) of tonic pain. Mice received formalin (1.25% in saline, 30
µL) in the dorsal surface of one side of the hind-paw. Each mouse
was randomly assigned to one of the experimental groups (n )
8-10) and placed in a Plexiglas cage and allowed to move freely
for 15-20 min. A mirror was placed at a 45° angle under the cage
to allow full view of the hind-paws. Lifting, favoring, licking,
shaking, and flinching of the injected paw were recorded as
nociceptive responses. The total time of the nociceptive response
was measured every 5 min and expressed as the total time of the
nociceptive responses in minutes (mean ( SEM). Recording of
nociceptive behavior commenced immediately after formalin injec-
tion and was continued for 60 min. The version of the formalin
test we applied is based on the fact that a correlational analysis
showed that no single behavioral measure can be a strong predictor
of formalin or drug concentrations on spontaneous behaviors.49
Consistently, we considered that a simple sum of time spent licking
plus elevating the paw, or the weighted pain score, is in fact superior
to any single (lifting, favoring, licking, shaking, and flinching)
measure (r ranging from 0.75 to 0.86).50 Treatments: groups of
8-10 animals per treatment were used with each animal being used
for one treatment only. Mice received intraperitoneal vehicle (10%
DMSO in 0.9% NaCl) or different doses of (()-ENBA, 5′Cl5′d-
(()-ENBA, or DPCPX.
Acknowledgment. This project was supported by a grant
from University of Camerino and by the Italian MIUR funds.
The work was presented in part at the Purines 2008 Meeting
(Copenhangen, Denmark, 2008). We thank M. Brandi, F. Lupidi,
G. Rafaiani, and M. Ricciutelli for technical assistance.
Supporting Information Available: Method for the synthesis
of compounds 1, 9, and 41 (Scheme 4); analytical data of all
synthesized compounds; binding affinity with confidence intervals
(Table S1); Ramachandran plots of the hA1 AR and hA3 AR models
(Figures S1 and S2); plots showing the time dependence of the
positional rmsd of backbone atoms and the total energy over the
course of the 1 ns trajectory in the hA1 AR and hA3 AR,
respectively, complexed with CPA (Figures S3a and S5a), (()-
ENBA (Figures S3b and S5b), 1 (Figures S4a and S6a), and 3
(Figures S4b and S6b). This material is available free of charge
Via the Internet at http://pubs.acs.org.
References
(1) Gao, Z.-G.; Jacobson, K. A. Emerging adenosine receptor agonists.
Expert Opin. Emerging Drugs 2007, 12, 479–492.
(2) Dhalla, A. K.; Shryock, J. C.; Shreeniwas, R.; Belardinelli, L.
Pharmacology and therapeutic applications of A1 adenosine receptor
ligands. Curr. Top. Med. Chem. 2003, 3, 369–385.
(3) (a) Cappellacci, L.; Franchetti, P.; Vita, P.; Petrelli, R.; Lavecchia,
A.; Costa, B.; Spinetti, F.; Martini, C.; Klotz, K.-N.; Grifantini, M.

2406 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 8
Franchetti et al.
(15) Lin, S. W.; Sakmar, T. P. Specific tryptophan UV-absorbance changes
are probes of the transition of rhodopsin to its active state. Biochemistry
1996, 35, 11149–11159.
(34) Cornell, W. D.; Cieplak, P.; Bayly, C. I.; Gould, I. R.; Merz, K. M.
Jr.; Ferguson, D. M.; Spellmeyer, D. C.; Fox, T.; Caldwell, J. W.;
Kollman, P. A. A second generation force field for the simulation of
proteins, nucleic acids, and organic molecules. J. Am. Chem. Soc. 1995,
117, 5179–5197.
(16) (a) Cherezov, V.; Rosenbaum, D. M.; Hanson, M. A.; Rasmussen,
S. G. F.; Thian, F. S.; Kobilka, T. S.; Choi, H.-J.; Kuhn, P.; Weis,
W. I.; Kobilka, B. K.; Stevens, R. C. High-resolution crystal structure
of an engineered human b2-adrenergic G protein-coupled receptor.
Science 2007, 318, 1258–1265. (b) Rosenbaum, D. M.; Cherezov, V.;
Hanson, M. A.; Rasmussen, S. G. F.; Thian, F. S.; Kobilka, T. S.;
Choi, H.-J.; Yao, X. J.; Weis, W. I.; Stevens, R. C.; Kobilka, B. K.
GPCR engineering yields highresolution structural insights into b2
adrenergic receptor function. Science 2007, 318, 1266–1273.
(17) Kim, S. K.; Gao, Z.-G.; Jeong, L. S.; Jacobson, K. A. Docking studies
of agonists and antagonists suggest an activation pathway of the A3
adenosine receptor. J. Mol. Graph. Model. 2006, 25, 562–577.
(18) Palaniappan, K. K.; Gao, Z.-G.; Ivanov, A. A.; Greaves, R.; Adachi,
H.; Besada, P.; Kim, H. O.; Kim, A. Y.; Choe, S. A.; Jeong, L. S.;
Jacobson, K. A. Probing the binding site of the A1 adenosine receptor
reengineered for orthogonal recognition by tailored nucleosides.
Biochemistry 2007, 46, 7437–7448.
(19) Ivanov, A. A.; Palyulin, V. A.; Zefirov, N. S. Computer aided
comparative analysis of the binding modes of the adenosine receptor
agonists for all known subtypes of adenosine receptors. J. Mol. Graph.
Model. 2007, 25, 740–754.
(20) Kim, J.; Wess, J.; van Rhee, A. M.; Schoneberg, T.; Jacobson, K. A.
Site-directed mutagenesis identifies residues involved in ligand
recognition in the human A2a adenosine receptor. J. Biol. Chem. 1995,
270, 13987–13997.
(21) Olah, M. E.; Ren, H.; Ostrowski, J.; Jacobson, K. A.; Stiles, G. L.
Cloning, expression, and characterization of the unique bovine A1-
adenosine receptor: Studies on the ligand binding site by site directed
mutagenesis. J. Biol. Chem. 1992, 267, 10764–10770.
(22) Gao, Z.-G.; Jiang, Q.; Jacobson, K. A.; Ijzerman, A. P. Site-directed
mutagenesis studies of human A2A adenosine receptors. Involvement
of Glu13 and His278 in ligand binding and sodium modulation.
Biochem. Pharmacol. 2000, 60, 661–668.
(23) 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.
(24) Rivkees, S. A.; Barbhaiya, H.; Ijzerman, A. P. Identification of the
adenine binding site of the human A1 adenosine receptor. J. Biol.
Chem. 1999, 274, 3617–3621.
(35) Ballesteros, J. A.; Weinstein, H. Integrated methods for modelling
G-protein coupled receptors. Methods Neurosci. 1996, 25, 366–428.
(36) (a) Olah, M. E.; Jacobson, K. A.; Stiles, G. L. Role of the second
extracellular loop of adenosine receptors in agonist and antagonist
binding. Analysis of chimeric A1/A3 adenosine receptors. J. Biol.
Chem. 1994, 269, 24692–24698. (b) Kim, J.; Jiang, Q.; Glashofer,
M.; Yehle, S.; Wess, J.; Jacobson, K. A. Glutamate residues in the
second extracellular loop of the human A2a adenosine receptor are
required for ligand recognition. Mol. Pharmacol. 1996, 49, 683–691.
(c) Rivkees, S. A.; Lasbury, M. E.; Barbhaiya, H. Identification of
domains of the human A1 adenosine receptor that are important for
binding receptor subtype selective ligands using chimeric A1/A2a
adenosine receptors. J. Biol. Chem. 1995, 270, 20485–20490.
(37) Bairoch, A.; Apweiler, R. The SWISS-PROT protein sequence
database and its supplement TrEMBL in 2000. Nucleic Acids Res.
2000, 28, 45–48.
(38) Thompson, J. D.; Higgins, D. G.; Gibson, T. J. CLUSTAL W:
Improving the sensitivity of progressive multiple sequence alignment
through sequence weighting, position-specific gap penalties and weight
matrix choice. Nucleic Acids Res. 1994, 22, 4673–4680.
(39) Fredriksson, R.; Lagerstro¨m, M. C.; Lundin, L. G.; Schio¨th, H. B.
The G-protein-coupled receptors in the human genome form five main
families. Phylogenetic analysis, paralogon groups, and fingerprints.
Mol. Pharmacol. 2003, 63, 1256–1272.
(40) McGuffin, L. J.; Bryson, K.; Jones, D. T. The PSIPRED protein
structure prediction server. Bioinformatics 2000, 16, 404–405.
(41) Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. Numerical integration
of the Cartesian equations of motion of a system with constraints:
molecular dynamics of n-alkanes. J. Comput. Phys. 1977, 23, 327–
341.
(42) Allen, F. H.; Bellard, S.; Brice, M. D.; Cartwright, B. A.; Doubleday,
A.; Higgs, H.; Hummelink, T.; Hummelink-Peters, B. G.; Kennard,
O.; Motherwell, W. D. S. The Cambridge Crystallographic Data
Center: computer-based search, retrieval, analysis and display of
information. Acta Crystallogr., Sect. B: Struct. Sci. 1979, 35, 2331–
2339.
(43) (a) Halgren, T. A. Merck molecular force field. I. Basis, form, scope,
parameterization, and performance of MMFF94. J. Comput. Chem.
1996, 17, 490–512. (b) Halgren, T. A. Merck molecular force field.
II. MMFF 94 van der Waals and electrostatic parameters for
intermolecular interactions. J. Comput. Chem. 1996, 17, 520–552. (c)
Halgren, T. A. Merck molecular force field. III. Molecular geometries
and vibrational parameters for intermolecular interactions. J. Comput.
Chem. 1996, 17, 553–586. (d) Halgren, T. A. Merck molecular force
field. IV. Conformational energies and geometries for MMFF94.
J. Comput. Chem. 1996, 17, 587–615. (e) Halgren, T. A. Merck
molecular force field. V. Extension of MMFF94 using experimental
data, additional computational data, and empirical rules. J. Comput.
Chem. 1996, 17, 616–641.
(44) Olah, M. E.; Stiles, G. L. The role of receptor structure in determining
adenosine receptor activity. Pharmacol. Ther. 2000, 85, 55–75.
(45) Dawson, E. S.; Wells, J. N. Determination of amino acid residues
that are accessible from the ligand binding crevice in the seventh
transmembrane-spanning region of the human A1 adenosine receptor.
Mol. Pharmacol. 2001, 59, 1187–1195.
(25) Jiang, Q.; van Rhee, M.; Kim, J.; Yehle, S.; Wess, J.; Jacobson, K. A.
Hydrophilic side chains in the third and seventh transmembrane helical
domains of human A2A adenosine receptors are required for ligand
recognition. Mol. Pharmacol. 1996, 50, 512–521.
(26) Ruprecht, J. J.; Mielke, T.; Vogel, R.; Villa, C.; Schertler, G. F. X.
Electron crystallography reveals the structure of metarhodopsin I.
EMBO J. 2004, 23, 3609–3620.
(27) Shi, L.; Liapakis, G.; Xu, R.; Guarnieri, F.; Ballesteros, J. A.; Javitch,
J. A. ꢀ2 adrenergic receptor activation. Modulation of the proline kink
in transmembrane 6 by a rotamer toggle switch. J. Biol. Chem. 2002,
277, 40989–40996.
(28) Gao, Z.-G.; Kim, S. K.; Gross, A. S.; Chen, A.; Blaustein, J. B.;
Jacobson, K. A. Identification of essential residues involved in the
allosteric modulation of the human A3 adenosine receptor. Mol.
Pharmacol. 2003, 63, 1021–1031.
(29) Dubuisson, D.; Dennis, S. G. The formalin test: a quantitative study
of the analgesic effects of morphine, meperidine, and brain stem
stimulation in rats and cats. Pain 1977, 4, 161–174.
(46) Jacobson, K. A.; Gao, Z.-G.; Chen, A.; Barak, D.; Kim, S.-A.; Lee,
K.; Link, A.; Van Rompaey, P. V.; Van Calenbergh, S.; Liang, B. T.
Neoceptor concept based on molecular complementarity in GPCRs: a
mutant adenosine A3 receptor with selectivity enhanced affinity for
amine-modified nucleosides. J. Med. Chem. 2001, 44, 4125–4136.
(47) De Lean, A.; Hancock, A. A.; Lefkowitz, R. J. Validation and statistical
analysis of a computer modeling method for quantitative analysis of
radioligands binding data for mixtures of pharmacological receptor
subtypes. Mol. Pharmacol. 1982, 21, 5–16.
(30) Maione, S.; De Novellis, V.; Cappellacci, L.; Palazzo, E.; Vita, D.;
Luongo, L.; Stella, L.; Franchetti, P.; Marabase, I.; Rossi, F.; Grifantini,
M. The antinociceptive effect of 2-chloro-2′-C-methyl-N6-cyclopen-
tyladenosine (2′-Me-CCPA), a highly selective adenosine A1 receptor
agonist, in the rat. Pain 2007, 131, 281–292.
(31) Molecular Operating EnVironment (MOE), version 2005.06; Chemical
Computing Group, Inc.: Montreal, Canada, 2005.
(32) Huang, C. C.; Couch, G. S.; Pettersen, E. F.; Ferrin, T. E. Chimera:
an extensible molecular modelling application constructed using
standard components. Pac. Symp. Biocomput. 1996, 1, 724; http://
www.cgl.ucsf.edu/chimera.
(33) Case, D. A.; Darden, T. A.; Cheatham, T. E., III; Simmerling, C. L.;
Wang, J.; Duke, R. E.; Luo, R.; Merz, K. M.; Pearlman, D. A.;
Crowley, M.; Walker, R. C.; Zhang, W.; Wang, B.; Hayik, S.;
Roitberg, A.; Seabra, G.; Wong, K. F.; Paesani, F.; Wu, X.; Brozell,
S.; Tsui, V.; Gohlke, H.; Yang, L.; Tan, C.; Mongan, J.; Hornak, V.;
Cui, G.; Beroza, P.; Mathews, D. H.; Schafmeister, C.; Ross, W. S.;
Kollman, P. A. AMBER 9; University of California, San Francisco:
2006.
(48) Klotz, K.-N.; Cristalli, G.; Grifantini, M.; Vittori, S.; Lohse, M. J.
Photoaffinity labeling of A1-adenosine receptors. J. Biol. Chem. 1985,
260, 14659–14664.
(49) Saddi, G.; Abbott, F. V. The formalin test in the mouse: a parametric
analysis of scoring properties. Pain 2000, 89, 53–63.
(50) Abbott, F. V.; Franklin, K. B.; Westbrook, R. F. The formalin test:
scoring properties of the first and second phases of the pain response
in rats. Pain 1995, 60, 91–102.
JM801456G