Extended N6 substitution of rigid C2-arylethynyl nucleosides for exploring the role of extracellular loops in ligand recognition at the A 3 adenosine receptor

Article abstract of DOI:10.1016/j.bmcl.2014.06.006

2-Arylethynyl-(N)-methanocarba adenosine 5′-methyluronamides containing rigid N⁶-(trans-2-phenylcyclopropyl) and 2-phenylethynyl groups were synthesized as agonists for probing structural features of the A₃ adenosine receptor (AR). Radioligand binding confirmed A ₃AR selectivity and N⁶-1S,2R stereoselectivity for one diastereomeric pair. The environment of receptor-bound, conformationally constrained N⁶ groups was explored by docking to an A₃AR homology model, indicating specific hydrophobic interactions with the second extracellular loop able to modulate the affinity profile. 2-Pyridylethynyl derivative 18 was administered orally in mice to reduce chronic neuropathic pain in the chronic constriction injury model.

Full text of DOI:10.1016/j.bmcl.2014.06.006

Bioorganic & Medicinal Chemistry Letters 24 (2014) 3302–3306
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Extended N6 substitution of rigid C2-arylethynyl nucleosides
for exploring the role of extracellular loops in ligand recognition
at the A3 adenosine receptor
Dilip K. Tosh a, , Silvia Paoletta a, , Zhoumou Chen b, Steven M. Moss a, Zhan-Guo Gao a,
Daniela Salvemini b, Kenneth A. Jacobson a,
⇑
a Molecular Recognition Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health, Bethesda,
MD 20892-0810, USA
b Department of Pharmacological and Physiological Science, Saint Louis University School of Medicine, St. Louis, MO 63104, USA
a r t i c l e i n f o
a b s t r a c t
2-Arylethynyl-(N)-methanocarba adenosine 50-methyluronamides containing rigid N6-(trans-2-phenyl-
cyclopropyl) and 2-phenylethynyl groups were synthesized as agonists for probing structural features
of the A3 adenosine receptor (AR). Radioligand binding confirmed A3AR selectivity and N6-1S,2R
stereoselectivity for one diastereomeric pair. The environment of receptor-bound, conformationally con-
strained N6 groups was explored by docking to an A3AR homology model, indicating specific hydrophobic
interactions with the second extracellular loop able to modulate the affinity profile. 2-Pyridylethynyl
derivative 18 was administered orally in mice to reduce chronic neuropathic pain in the chronic constric-
tion injury model.
Article history:
Received 9 May 2014
Revised 29 May 2014
Accepted 2 June 2014
Available online 11 June 2014
Keywords:
G protein-coupled receptor
Purines
Molecular modeling
Structure activity relationship
Radioligand binding
Adenylate cyclase
Published by Elsevier Ltd.
X-ray crystallographic structures of the A2A adenosine receptor
(A2AAR) in complex with both agonists and antagonists1–3 can
serve as suitable templates for accurately predicting the ligand
interactions of other AR subtypes. We now focus on sterically
bulky and hydrophobic N6 substituents on adenosine derivatives
that are associated with high affinity at the A3AR. Adenosine deriv-
atives that selectively activate the A3AR are in clinical trials for
autoimmune inflammatory diseases and liver cancer; thus, there
is considerable interest in drug discovery at this receptor.4
Many studies have established the effects of diverse substitu-
tion of the exocyclic amine of adenosine on the AR subtype selec-
tivity, with predominantly hydrophobic N6 groups found to be best
suited for high affinity.5–7 For example, the structure activity rela-
tionship (SAR) of the N6-(2-trans-phenylcyclopropyl) group and its
substituted analogues was studied in monosubstituted adenosine
derivatives (e.g., 1, 2, Chart 1), which tended toward selectivity
at the A3AR.8 The steric constraint and stereochemical definition
of the N6-(2-trans-phenylcyclopropyl) group, especially in combi-
nation with rigid ribose substitutions, make it possible to study
the structural implications of this conformationally constrained
substituent in A3AR recognition. Here we use a recent homology
model9 of the human (h) A3AR to probe the structural basis for
the effects on binding affinity in this class of agonists.
We previously explored a class of constrained (N)-methanocar-
ba adenosine-50-methylamide nucleosides, including 2-chloro
derivatives (e.g., 3)10 and C2-arylethynyl analogues (e.g., 4),9,11
which are very potent and selective full agonists of the A3AR. N6-
Benzyl-type substitution is often used because of its selective
enhancement of A3AR affinity. Here, we have synthesized new
N6-(2-phenylcyclopropyl)-substituted analogues in this extended
methanocarba series and characterized them pharmacologically.
We already reported the A3AR selectivity of one analogue, 5
(R1 = H) in which a rigid ribose-like (N)-methanocarba scaffold10
was combined with a sterically restricted N6-(2-phenylcyclopro-
pyl) substitution, based on our earlier study of SAR in the 9-ribo-
side series.8 The corresponding 40-truncated (N)-methanocarba
nucleoside containing a racemic N6-(2-phenylcyclopropyl) substi-
tution was found to be a selective A3AR antagonist (Ki 1.3 nM).12
The 1S,2R diastereoisomer of 2 was ꢀ10-fold more potent in bind-
ing to the hA3AR (Ki 11.2 nM) than the corresponding 1R,2S isomer.
This N6 substituent was also combined with C2-cyano or carbox-
amide substitution in a subsequent 9-riboside series resulting in
⇑
Corresponding author. Tel.: +1 301 496 9024; fax: +1 301 496 8422.
E-mail address: kajacobs@helix.nih.gov (K.A. Jacobson).
Equal contributions.
http://dx.doi.org/10.1016/j.bmcl.2014.06.006
0960-894X/Published by Elsevier Ltd.

D. K. Tosh et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3302–3306
3303
R1
R1
Cl
HN
N
HN
HN
N
N
N
N
N
N
N
N
N
R2
H
N
N
N
R2
H
N
O
H3C
O
HO
H3C
O
OH
OH
R3
OH
OH
3 R2 = Cl
4 R2
OH
OH
R3
1a R1 = H, 1S,2R
1b R1 = H, 1R,2S
5 R2 = Cl
6 R2
=
=
2 R1 = 3-NO2, ( )-trans
Chart 1. Progression of SAR studies of A3AR agonists containing bulky N6 substituents, such as trans-2-phenylcyclopropyl, and modification of the ribose as a ring-constrained
bicyclic system.
a moderate reduction of A3AR affinity.13 In our earlier study of
ribosides,8 there was a 38-fold stereoselectivity of 1a versus 1b
in A3AR binding, and a smaller ratio was observed for stereoiso-
mers of a 3-nitro derivative 2. A large species difference was
associated with the N6-(2-phenylcyclopropyl) group, such that
affinity was greatly enhanced in progressing from rat to human.
The current expanded series (general formula 6) combined rigid
C2-arylethynyl and N6-(2-phenylcyclopropyl) groups. This
extended, sterically defined N6 substitution allows for the explora-
tion of the outer regions of the A3AR by docking of the ligands to
homology models of the receptor. Thus, the effects of the
interaction of different N6 substituents with the normally flexible
extracellular loops (ELs) were analyzed by detailed modeling.
Chronic neuropathic pain (NP) is an important unsolved
medical need, which is associated with nerve injury and often with
diseases such as cancer and diabetes.14,15 We recently reported
that A3AR agonists have unanticipated efficacy in vivo in various
models of chronic but not acute pain.1 Although N6-(2-phenylcy-
clopropyl) substitution in the riboside series greatly reduces the
affinity at the murine A3ARs,24 we studied one of the new
analogues in a mouse model of chronic pain resulting from
constriction injury (CCI).16
The nucleoside derivatives 9–19 (Table 1) were synthesized,
characterized and examined in AR binding assays. Compounds 1,
4, 5, 7 and 8, prepared earlier,8–11 were included as reference com-
pounds. To synthesize N6-(2-phenylcyclopropyl) derivatives, pro-
tected intermediate 209 was sequentially treated with the
appropriate 2-phenylcyclopropylamine to yield 21a–h followed
by methylamine to provide 22a–h (Scheme 1). Then, intermediates
22a–h were subjected to Sonogashira coupling with the
appropriate arylacetylene in the presence of PdCl2(Ph3P)2, CuI
and triethylamine to give protected intermediates 23a–j, which
upon acid hydrolysis gave the target compounds 10–19. The syn-
thesis of 6-NH2 derivative 9 will be reported elsewhere. Although
most of the entries in Table 1 contain a mixture of stereoisomers,
compounds 16–19 correspond to products with well-defined
(1R,2S or 1S,2R) stereochemistry of the N6-(2-phenylcyclopropyl)
group. Compound 15 is the racemic form of pure diastereoisomers
16 and 17.
remained high, with the significant reduction in affinity at the A1
and A2AARs. The unsubstituted, racemic N6-(2-phenylcyclopropyl)
derivative 10 retained a higher A3AR affinity than the ring-substi-
tuted analogues 11–15. However, the 3,4-difluorophenyl analogue
15 was very selective for the A3AR, and the fluoro substitution was
intended to diminish aromatic oxidation in vivo.25 A comparison of
the A3AR affinity of diastereomeric pairs showed a 4-fold prefer-
ence for the 1S,2R diastereoisomer in the pair of 2-phenylethynyl
derivatives 16 and 17, but there was no difference in A3AR affini-
ties of the 2-(2-pyridylethynyl) diastereomeric pair 18 and 19.
Therefore, we have used a combination of adenine substitutions
and a ribose-like scaffold, all of greatly reduced conformational
freedom, to help analyzing ligand recognition in the outer regions
of the A3AR. The environment of the receptor-bound N6-(2-phenyl-
cyclopropyl) group was explored by docking to an A3AR homology
model. We used our previously-reported homology model of the
hA3AR,9,11 which was based on a hybrid A2AAR-b2 adrenergic
receptor template. The general methodology used for homology
modeling (MOE homology modeling tool)20 and ligand docking
(Glide module of the Schrödinger Suite)17 has been described
before,9,11 and specific methodological details are reported in the
Supporting Information. In the present study, after the first round
of docking to the initial A3AR homology model and selection of
the best docking pose for derivative 7, we performed refinement
of the portion of the second EL in contact with the ligand (from
Gln167 to Arg173), using the Prime module of the Schrödinger
Suite,21 to optimize its conformation around the N6 substituent.
This step was followed by a second round of docking of all deriva-
tives to the optimized model, to identify the final proposed docking
poses. All of the final nucleosides were also subjected to docking
simulations at a hA2AAR crystal structure (PDB ID: 2YDV)3 to
explore the reasons for ligand selectivity.
As expected, docking results at the hA3AR for all the analyzed
derivatives showed the pseudo-sugar moiety, adenine N7 and exo-
cyclic NH participating in highly conserved H-bonding interactions
with key residues of the binding site (Fig. 1), such as Thr94 (3.36),
Asn250 (6.55), Ser271 (7.42) and His272 (7.43) (numbers in paren-
thesis follow the Ballesteros–Weinstein notation).22 The same H-
bonding network has been observed in the agonist-bound hA2AAR
crystal structures and is supposed to be important for receptor
activation.2,3 Another important binding feature for AR ligands,
The binding assays were based on widely used radioligands
using membranes of CHO cells expressing the hA1AR ([3H]24) or
hA3AR ([125I]26) or HEK293 cells expressing the hA2AAR
([3H]25).17–19 Most of the nucleosides bound to the hA3AR with
low nanomolar affinity and with minimal binding to the hA1AR
and hA2AAR. Although the A3AR affinity range of the newly synthe-
sized 2-arylethynyl compounds was somewhat less than the refer-
ence 2-chloro-(N)-methanocarba nucleoside 5, the selectivity
both agonists and antagonists, is the p–p stacking interaction with
an aromatic residue in EL2 (Phe168 at the hA3AR). Thus, consider-
ing that our quite rigid compounds showed all these critical inter-
actions when docked to the A3AR model, we could analyze in more
detail the possible interactions with the less defined EL regions of
the receptor. Docking results showed the rigid C2-arylethynyl

3304
D. K. Tosh et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3302–3306
Table 1
Structures and binding affinities of reference nucleosides and newly synthesized (N)-methanocarba A3AR agonists (9–19)
R2
R4
HN
HN
N
N
N
N
N
N
N
H
N
N
H
N
O
O
X
H3C
H3C
OH
OH
OH
OH
4,7-9
10-19
Compd
X
R2 or R4
Configuration of N6 group
A1ARa % inhibition or Ki (nM)
A2AARa % inhibition or Ki (nM)
A3ARa Ki (nM)
d
d
d
d
d
d
d
1b,c
( )-trans
124 30 nM
20 3%
770 50 nM
11 3%
13 6%
28 4%
2530 720 nM
27 3%
4800 200 nM
32 4%
14 7%
1200 nM
43 8%
0.86 0.09
1.34 0.30
0.78 0.06
1.23 0.57
0.85 0.22
1.51 0.41
6.16 0.22
17.6 1.3
16.5 1.9
11.2 4.2
14.5 1.0
20.2 2.1
16.9 3.3
4.55 1.63
8.18 2.22
7.31 1.74
d
4b
3-Cl-PhCH2
d
5b,c (R1 = H)
( )-trans
7b
Ph-(CH2)2
CH3
H
H
3-CH3
3-F
d
8b
d
d
9
10
11
12
13
14
15
16e
17e
18
19
CH
CH
CH
CH
CH
CH
CH
CH
N
( )-trans
( )-trans
( )-trans
( )-trans
( )-trans
( )-trans
1R,2S-trans
1S,2R-trans
1R,2S-trans
1S,2R-trans
2
8
1%
4%
53 6%
18 10%
3%
0
0%
3-Cl
3-Br
6
39 5%
32 18%
17 4%
11 6%
11 6%
16 1%
19 8%
26 1%
3,4-DiF
3,4-DiF
3,4-DiF
3,4-DiF
3,4-DiF
6
0
0
3%
0%
0%
25 5%
41 1%
N
a
Binding was performed as described11 using membranes prepared from HEK293 (A2A only) or CHO cells stably expressing one of three hAR subtypes. The binding
affinities for A1, A2A and A3ARs were determined using agonist radioligands [3H]N6-R-phenylisopropyladenosine ([3H]24), [3H]2-[p-(2-carboxyethyl)phenyl-ethylamino]-50-
N-ethylcarboxamido-adenosine ([3H]25), or [125I]N6-(4-amino-3-iodobenzyl)adenosine-50-N-methyluronamide ([125I]26), respectively. Data (n = 3ꢁ4) are expressed as Ki
values. A percent in italics refers to inhibition of binding at 10
l
M. Stock solutions of N6-(2-phenylcyclopropyl) derivatives in DMSO were stored at ꢁ80 °C.
b
Compounds 1,8 4,9 5,10 711 and 89 were reported earlier.
c
See Chart 1 for structure.
Not applicable.
Compound 16, MRS7030; 17, MRS7034.
d
e
substituent of the synthesized compounds to be directed towards
the extracellular side in proximity of TM2; in fact, this group could
be accommodated in the binding site only after outward move-
ment of TM2, as previously described.9,11 On the other hand, the
different N6 substituents of the reported derivatives are in proxim-
ity of EL2, and even though the recognition region for the N6 group
is on the outer regions, it involves mainly hydrophobic residues at
the hA3AR. In fact, at this receptor the exocyclic amino group of the
ligands is in proximity of a small, secondary (side) pocket delim-
ited by EL2 (residues Val169, Ser170, Val171, Met172, Arg173,
Met174 and Met 151). This side pocket is able to well accommo-
date substituents such as N6-(3-Cl-benzyl) and N6-phenylethyl
(see docking poses of corresponding compounds 4 and 7 in
Fig. 1). The good fit of these substituents with this region can
explain the very high affinity of such compounds at the hA3AR.
These two derivatives also showed high affinity for the mouse A3
subtype, in agreement with the observation that the mA3AR can
accommodate these N6 substituents in the same region, as previ-
ously reported,11 even though the surrounding residues are differ-
ent. One key difference is the presence of an Arg in the murine A3
subtypes (both mouse and rat) in place of Val169 in the hA3AR.
This bulkier and positively charged residue can determine a more
difficult fit of bulkier and/or more rigid N6 groups. This observation
can explain the substantially lower affinity of compound 1, with
the N6-(2-phenylcyclopropyl) group, at the rA3AR as compared to
the hA3AR.8
interacts through H-bonds with both agonists (e.g., at unsubsti-
tuted 6-NH2 groups) and antagonists, as highlighted by crystal-
lographic data.1–3 As we observed previously, the insertion of
an extended C2-arylethynyl substituent considerably reduces
the affinity at the A2A subtype.9 However, the only C2-arylethy-
nyl derivative showing a good docking pose at the hA2AAR was
compound 9 (Fig. S1, Supporting Information), in agreement
with its binding affinity of ꢀ1
lM at this subtype. This com-
pound is the only one of the series presenting an unsubstituted
exocyclic amino group, which can interact with Glu169 in EL2
through a H-bond.
At the hA3AR, the N6-(2-phenylcyclopropyl) group of the C2-
arylethynyl compounds can still be accommodated in the region
surrounded by EL2, even though there is a cost in affinity. This is
probably due to the increased rigidity compared to the N6-phenyl-
ethyl derivatives (compare compound 10, Ki hA3AR = 6.16 nM, and
compound 7, Ki hA3AR = 0.85 nM). Moreover, binding data from
both the ribose series and (N)-methanocarba series showed bind-
ing stereoselectivity for this substituent in several isomeric pairs
with preference for the 1S,2R diastereoisomer. This can be
explained considering that the 1S,2R isomers could better fit the
EL2-delimited side pocket as compared to the 1R,2S isomers, as
shown by the docking poses of compounds 16 and 17 in Figure 1.
On the other hand, no difference in affinity was observed for the
N6-(2-phenylcyclopropyl) C2-(2-pyridylethynyl) isomers 18 and
19. This seems to be related to a compensatory effect due to the
formation of an additional H-bond between the pyridine nitrogen
and the backbone amino group of Phe168 in EL2 of the hA3AR
(Fig. S2, Supporting Information).
Residues in EL2 seem to be implicated also in A2A versus A3
selectivity, as suggested before by docking studies.23 At the
hA2AAR, the glutamic acid (Glu169) present at this position

D. K. Tosh et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3302–3306
3305
Scheme 1. Synthesis of N6-phenylcyclopropyl (N)-methanocarba derivatives 10–19. Reagents and conditions: (i) R2-phenylcyclopropyl-NH2, Et3N, MeOH, rt; (ii) 40% MeNH2,
MeOH, rt; (iii) HC„CR1, Pd(PPh3)2Cl2, CuI, Et3N, DMF, rt; (iv) 10% TFA, MeOH, 70 °C.
Figure 1. Putative binding modes of selected (N)-methanocarba derivatives obtained after docking simulations at the hA3AR model. (A) Side view of the receptor binding site.
Ligands are shown in sticks: compound 4 (pink carbons), compound 7 (pale cyan carbons), compound 16 (magenta carbons) and compound 17 (cyan carbons). Side chains of
some amino acids important for ligand recognition are highlighted (grey carbon sticks). H-bonding interactions are pictured as dotted lines. Nonpolar hydrogen atoms are not
displayed. (B) Top view of the receptor binding site. Semi-transparent surface of binding site’s residues is displayed. Color scheme as in panel A.
The importance of the ELs in ligand recognition is emphasized
in the modeling results. In fact, even though compounds have
the ability to form all the crucial interactions with key residues
in the lower part of the binding site, N6 and C2 substituents can
modulate their affinity by interacting with the aqueous-exposed
outer region of the binding site. In general, the hA3AR is able to tol-
erate bulkier and/or more rigid substitutions at these positions as
compared to the murine A3ARs and the hA2AAR, because of the dif-
ferences present in EL2. Therefore, the hA3AR can accommodate
several different-sized N6 groups, i.e. methyl, 3-Cl-benzyl,
phenylethyl and 2-phenylcyclopropyl, in a hydrophobic region
delimited by EL2, maintaining good affinity. Moreover, different
N6- and C2-substituents seem to influence each other depending
on their complementarity with the binding site. In fact, the pres-
ence of a N6-(2-phenylcyclopropyl) group is more tolerated in
absence of an extended C2-arylethynyl, while the presence of both
rigid groups makes it more difficult to accommodate the com-
pound in the cavity (compare compound 5, Ki hA3AR = 0.78 nM,
and compound 10, Ki hA3AR = 6.16 nM). The C2 substituent can
also have a compensatory effect due to the formation of additional

3306
D. K. Tosh et al. / Bioorg. Med. Chem. Lett. 24 (2014) 3302–3306
are required to accommodate these ligands. This study introduces
tools that will aid in the understanding of the role of receptor plas-
ticity (in particular within the EL regions) in receptor recognition
and activation.
Acknowledgments
We thank Dr. John Lloyd and Dr. Noel Whittaker (NIDDK) for
mass spectral determinations. This research was supported by
the National Institutes of Health (Intramural Research Program of
the NIDDK and R01HL077707).
Supplementary data
Figure 2. Activity of 18 in a CCI model of chronic neuropathic pain (mechanoall-
odynia), performed as described (po, oral administration).24 Paw withdrawal
threshold is shown as a function of time. The nucleoside was administered by
gavage at the time of peak pain (7 days after chronic constriction injury).
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
06.006.
References and notes
interactions as observed in the C2-(2-pyridylethynyl) series. The
modulatory action of these two substituents can be related both
to their ability to properly orient the ligand in the binding site so
that it can form optimal interactions with key residues, and to their
possible role in influencing the conformation of residues in EL2 by
modulating their interaction with the ligand. This is understand-
able considering that a highly conserved interaction in AR binding
1. Jaakola, V. P.; Griffith, M. T.; Hanson, M. A.; Cherezov, V.; Chien, E. Y. T.; Lane, J.
R.; IJzerman, A. P.; Stevens, R. C. Science 2008, 322, 1211.
2. Xu, F.; Wu, H.; Katritch, V.; Han, G. W.; Jacobson, K. A.; Gao, Z. G.; Cherezov, V.;
Stevens, R. C. Science 2011, 332, 322.
3. Lebon, G.; Warne, T.; Edwards, P. C.; Bennett, K.; Langmead, C. J.; Leslie, A. G.;
Tate, C. G. Nature 2011, 474, 521.
4. Fishman, P.; Bar-Yehuda, S.; Liang, B. T.; Jacobson, K. A. Drug Discovery Today
2012, 17, 359.
5. de Lera Ruiz, M.; Lim, Y. H.; Zheng, J. J. Med. Chem. 2014, 57, 3623.
6. Baraldi, P. G.; Preti, D.; Borea, P. A.; Varani, K. J. Med. Chem. 2012, 55, 5676.
7. Müller, C.; Jacobson, K. A. Biochem. Biophys. Acta—Biomembr. 2011, 1808, 1290.
8. Tchilibon, S.; Kim, S. K.; Gao, Z. G.; Harris, B. A.; Blaustein, J.; Gross, A. S.;
Melman, N.; Jacobson, K. A. Bioorg. Med. Chem. 2004, 12, 2021.
9. Tosh, D. K.; Deflorian, F.; Phan, K.; Gao, Z. G.; Wan, T. C.; Gizewski, E.;
Auchampach, J. A.; Jacobson, K. A. J. Med. Chem. 2012, 55, 4847.
10. Tchilibon, S.; Joshi, B. V.; Kim, S. K.; Duong, H. T.; Gao, Z. G.; Jacobson, K. A. J.
Med. Chem. 2005, 48, 1745.
11. Paoletta, S.; Tosh, D. K.; Finley, A.; Gizewski, E.; Moss, S. M.; Gao, Z. G.;
Auchampach, J. A.; Salvemini, D.; Jacobson, K. A. J. Med. Chem. 2013, 56, 5949.
12. Melman, A.; Wang, B.; Joshi, B. V.; Gao, Z. G.; de Castro, S.; Heller, C. L.; Kim, S.
K.; Jeong, L. S.; Jacobson, K. A. Bioorg. Med. Chem. 2008, 16, 8546.
13. 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.
Bioorg. Med. Chem. 2004, 12, 2995.
is the adenine p–p stacking interaction with the conserved Phe in
EL2; therefore, possible conformational changes determined by
rigidified substituents interacting with the loop could change the
orientation of this Phe and its interaction with the ligand.
Although the molecular weight of this chemical series is outside
the optimal range for oral bioavailability (MW 558; clogP 2.63;
total polar surface area 134 Å), one derivative 18 was tested in
the chronic constriction injury (CCI) pain model after oral adminis-
tration, following the general method previously described.24 It
displayed considerable activity in reversing chronic neuropathic
pain (Fig. 2). This likely reflects the very high A3AR affinity and
selectivity of this compound series. The peak effect occurred 1 h
post-administration.
ˇ
14. Hughes, J. P.; Chessell, I.; Malamut, R.; Perkins, M.; Backonja, M.; Baron, R.;
In conclusion, we have identified a group of highly rigidified
and spatially extended nucleosides that were shown to be selective
A3AR agonists. Rigid substituents at N6 and C2 positions, each of
which was previously established as being conducive to A3AR
binding, were combined. The activity of a representative com-
pound in a clinically relevant in vivo pain model was demon-
strated. Given these pharmacologically important properties, the
conformational analysis of receptor binding was analyzed. The ste-
ric constraints of this molecular series allow us to probe the envi-
ronment and propose specific amino acid interactions when
receptor-bound. The conformationally constrained N6 group in
docking to an A3AR homology model formed clear hydrophobic
interactions with the normally flexible EL2. The conformational
mobility of residues in this loop in the closely related A2AAR struc-
tures was noted, based on a comparison of agonist-bound and
antagonist-bound structures.2 This study introduces sterically
well-defined ligands that have rigid projections that can define
the conformations of the A3AR in the TM and outer regions that
Farrar, J. T.; Field, M. J.; Gereau, R. W.; Gilron, I.; McMahon, S. B.; Porreca, F.;
Rappaport, B. A.; Rice, F.; Richman, L. K.; Segerdahl, M.; Seminowicz, D. A.;
Watkins, L. R.; Waxman, S. G.; Wiech, K.; Woolf, C. Ann. N.Y. Acad. Sci. 2012,
1255, 30.
15. Freeman, R. Curr. Diab. Rep. 2013, 13, 500.
16. Bennett, G. J.; Xie, Y. K. Pain 1988, 33, 87.
17. Schwabe, U.; Trost, T. Naunyn Schmiedebergs Arch. Pharmacol. 1989, 313, 179.
18. Jarvis, M. F.; Schutz, R.; Hutchison, A. J.; Do, E.; Sills, M. A.; Williams, M. J.
Pharmacol. Exp. Ther. 1989, 251, 888.
19. Olah, M. E.; Gallo-Rodriguez, C.; Jacobson, K. A.; Stiles, G. L. Mol. Pharmacol.
1994, 45, 978.
20. Molecular Operating Environment (MOE), version 2012.10; Chemical
Computing Group Inc. (1255 University Street, Suite 1600, Montreal, Quebec,
H3B 3X3, Canada).
21. Schrödinger Suite 2013; Schrödinger, LLC, New York, NY.
22. Ballesteros, J. A.; Weinstein, H. Methods Neurosci. 1995, 25, 366.
23. Lenzi, O.; Colotta, V.; Catarzi, D.; Varano, F.; Poli, D.; Filacchioni, G.; Varani, K.;
Vincenzi, F.; Borea, P. A.; Paoletta, S.; Morizzo, E.; Moro, S. J. Med. Chem. 2009,
52, 7640.
24. Chen, Z.; Janes, K.; Chen, C.; Doyle, T.; Tosh, D. K.; Jacobson, K. A.; Salvemini, D.
FASEB J. 2012, 16, 1855.
25. Teng, R.; Oliver, S.; Hayes, M. A.; Butler, K. Drug Metab. Dispos. 2010, 38, 1514.