Synthesis and SAR studies of trisubstituted purinones as potent and selective adenosine A2A receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2009.01.042

A series of trisubstituted purinones was synthesized and evaluated as A_2A receptor antagonists. The A_2A structure-activity relationships at the three substituted positions were studied and selectivity against the A₁ recept

Full text of DOI:10.1016/j.bmcl.2009.01.042

Bioorganic & Medicinal Chemistry Letters 19 (2009) 1399–1402
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and SAR studies of trisubstituted purinones as potent
and selective adenosine A2A receptor antagonists
*
Yuefei Shao , Andrew G. Cole, Marc-Raleigh Brescia, Lan-Ying Qin, Jingqi Duo, Tara M. Stauffer,
Laura L. Rokosz, Brian F. McGuinness, Ian Henderson
Pharmacopeia, Inc., PO Box 5350, Princeton, NJ 08543, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of trisubstituted purinones was synthesized and evaluated as A2A receptor antagonists. The A2A
structure–activity relationships at the three substituted positions were studied and selectivity against the
A1 receptor was investigated. One antagonist 12o exhibits a Ki of 9 nM in an A2A binding assay, a Kb of
18 nM in an A2A cAMP functional assay, and is 220-fold selective over the A1 receptor.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 12 December 2008
Revised 12 January 2009
Accepted 13 January 2009
Available online 19 January 2009
Keywords:
Adenosine
GPCR
Purinone
SAR
Parkinson’s disease
Adenosine receptors are members of the G-protein-coupled
receptor family. Four subtypes of adenosine receptors have been
identified and designated as A1, A2A, A2B and A3.1 The A2A receptor
is enriched in the striatum2 and is currently an active area of inter-
est in pharmaceutical research due to its potential therapeutic
application in Parkinson’s disease and other neurodegenerative
disorders.3–5 The main pathology of Parkinson’s disease is the pro-
gressive destruction of the dopamine pathway. Current dopamine
agonist therapies are primarily focused on management of the
symptoms and do not change the pace of disease progression.6,7
A2A receptor antagonists are potential neuroprotective drugs that
improve locomotor activity and may stop or slow the degenerative
process.7,8
plore the structure–activity relationships and optimize the A2A
binding affinity and other pharmacological properties, additional
analogs were synthesized.
The purinone analogs 6 were synthesized in five steps from
commercially available 2,4-dichloro-5-nitropyrimidine (1) as
shown in Scheme 1. The first chlorine atom was displaced by a pri-
mary amine (R1NH2) at À78 °C with high regioselectivity for the 4-
position versus the 2-position (typically 10:1 ratio).11 The two
regioisomers were readily separated by flash chromatography.
The predominant regioisomer 2 was further functionalized at the
C-2 position with a second amine (R2NH2) to provide 3, followed
by reduction of the nitro group with sodium hydrosulfite to give
4. Treatment of 4 with carbonyl diimidazole at room temperature
produced the desired purinone 5. Subsequent N-alkylation of 5 to
provide 6 was achieved by using an alkyl halide (R3X) and polysty-
rene-bound 2-tert-butylimino-2-diethylamino-1,3-dimethyl-per-
hydro-1,3,2-diazaphosphorine (BEMP)12 as the base.
The A2A binding assay was performed with recombinant human
A2A receptor using the same radioligand protocol as described pre-
viously.13 Due to the potential cardiovascular side effect of A1
receptor antagonism,5,14 selectivity against the A1 receptor was
also investigated by measuring human A1 binding affinity using a
similar protocol.13 A set of analogs incorporating a cyclopropylami-
no substituent at C-2 (purine numbering protocol) was synthesized
to investigate the SAR at N-7 and N-9 (Table 1). Both initial screen-
ing results and this SAR study indicated conservative structural tol-
erances for A2A binding affinity with regard to the N-9 position (R1)
of the purinone scaffold. A substituted phenyl group at R1 seems
High-throughput screening of a series of ECLiPSTM (Encoded
Combinatorial Libraries on Polymeric Support)9 libraries employ-
ing an A2A radioligand-binding assay resulted in the identification
of multiple sets of compounds as A2A antagonists. Recently we re-
ported the identification of substituted aminothiazoles as A2A
antagonists from this screen.10 Here we report the discovery and
associated SAR studies of a series of trisubstituted purinones as po-
tent and selective A2A antagonists. The specific library that elicited
these A2A actives contains ꢀ50,000 compounds, which share a
common structural element of a purinone core. A small set of com-
pounds with a trisubstituted purinone scaffold was identified as
A2A antagonist leads as exemplified by structure 7a. To further ex-
* Corresponding author. Tel.: +1 609 452 3754; fax: +1 609 655 4187.
E-mail address: yshao@pcop.com (Y. Shao).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.01.042

1400
Y. Shao et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1399–1402
NO2
NO2
Cl
NO2
N
N
N
b
a
R1
R2
R1
Cl
N
N
Cl
N
N
H
N
N
H
H
1
2
3
R3
H
N
NH2
N
N
N
N
O
e
O
d
c
R2
R1
R2
R2
N
N
N
N
N
H
N
H
N
N
H
N
R1
R1
H
4
5
6
Scheme 1. Reagents and conditions: (a) R1NH2, DIEA, THF, À78 °C; (b) R2NH2, DIEA, THF, rt; (c) sodium hydrosulfite, sodium bicarbonate, THF/MeOH, rt; (d) carbonyl
diimidazole, DCM, rt; (e) R3X, polystyrene-bound BEMP, acetonitrile, rt.
Table 1
2-(Cyclopropylamino)purinone A2A antagonists
R3
N
N
O
N
N
H
N
R1
7
Compound
R1
R3
hA2A Ki (nM)a
hA1 Ki (nM)b
A1/A2A ratio
7a
7b
7c
7d
7e
7f
7g
7h
7i
7j
7k
7l
7m
7n
7o
7p
7q
3-Methoxyphenyl
2-Methoxyphenyl
Phenyl
3,5-Dimethoxyphenyl
3-Fluorophenyl
3-Methylphenyl
3-Trifluoromethoxyphenyl
3-Cyanophenyl
3-Methoxybenzyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
3-Methoxyphenyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
2-Methoxybenzyl
Benzyl
24
1561
409
76
402
N/D
447
4640
599
1847
5138
3237
>10,000
564
201
2850
236
17
N/D
1.1
61
7
89
316
206
2587
9213
54
33
86
4
6
6
25
1.3
N/D
10
6
33
59
55
N/D
N/D
0.2
2-Fluorobenzyl
2-Chlorobenzyl
2,6-Difluobenzyl
2-Fluoro-6-methoxybenzyl
(+)- -Methylbenzyl
330
N/D
N/D
1265
a
359
2641
5122
Methyl
2-Methoxyphenyl
a
Ki determined by competition binding of [3H]SCH-58261.13
Ki determined by competition binding of [3H]DPCPX.13
b
important to achieve good A2A binding affinity. Among the various
R1 groups tested (7a–7i), the compound incorporating
3-
and 2-fluorobenzyl (7k) further improve A2A activity. The analog
with a 2-chlorobenzyl group (7l) shows slightly decreased activity.
The greatest A2A activity is achieved with di-substituted benzyl
analogs, for example, analogs with 2,6-difluorobenzyl (7m,
Ki = 4 nM), and 2-fluoro-6-methoxybenzyl (7n, Ki = 6 nM). Interest-
ingly, these analogs also demonstrate high selectivity against the
a
methoxyphenyl group (7a) shows the highest potency
(Ki = 24 nM). Switching the methoxy group from the 3-position
(7a) to the 2-position (7b) significantly reduces binding affinity
(Ki = 1561 nM). Compound 7b is also 4 times weaker than the ana-
log with the unsubstituted phenyl group (7c), suggesting some de-
gree of steric restriction at the 2-position of the phenyl ring.
Addition of a second methoxy substituent at the 5-position (7d)
improves selectivity against A1, although its A2A activity is slightly
decreased. Substitution at the 3-position with a fluorine atom (7e)
also exhibits good activity (Ki = 89 nM), while substitution with a
methyl group (7f) shows only similar activity to the unsubstituted
phenyl (7c). This SAR might be indicative of a possible hydrogen-
bond interaction at the 3-position. The weaker affinity of analogs
with 3-trifluoromethoxyphenyl (7g) and 3-cyanophenyl (7h)
groups demonstrate the relatively tight tolerances at this position.
Interestingly, while analogs with a 3-methoxyphenyl group at R1
display good binding, insertion of a methylene spacer between
the phenyl and purinone (7i) completely removes binding affinity.
The structure–activity relationship of the R3 group at N-7 is also
detailed in Table 1. A substituted benzyl group is generally favor-
able at the R3 position. While an unsubstituted benzyl group (7j)
displays good A2A activity (Ki = 54 nM), 2-methoxybenzyl (7a)
A1 receptor with A1/A2A Ki ratios of 55–60.
a-Methylation of the
benzyl group (7o) reduces A2A activity, and a small alkyl group at
the R3 position (7p) is not tolerated. Also, switching from 3-
methoxybenzyl (7a) to 3-methoxyphenyl (7q) at the R3 position re-
sults in significant loss of A2A activity.
In order to facilitate studies of the role of the C-2 substituent,
the synthetic route was revised so that the R2 group at C-2 was
introduced in the last synthetic step (Scheme 2). Based on the
SAR study results discussed above, R1 and R3 were fixed as 3-
methoxyphenyl and 2,6-difluorobenzyl, respectively. Reduction of
the nitro group of 8 provided diaminopyrimidine 9, which was
subsequently cyclized using carbonyl diimidazole to give 2-chloro-
purinone 10. Subsequent N-9 alkylation to give 11 was performed
with 2,6-difluorobenzyl bromide in the presence of polystyrene-
bound BEMP. Final displacement of the 2-chloro group with an
amine (R2NH2) was achieved under microwave-assisted conditions
to obtain 12. Overall, compounds with various R2 groups display
good A2A affinity, with relatively wide tolerance of different struc-

Y. Shao et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1399–1402
1401
H
N
NO2
NH
NH2
N
N
N
O
a
b
Cl
N
Cl
N
NH
N
Cl
N
O
10
8
9
O
O
F
F
F
F
N
N
N
N
N
N
c
d
O
O
R2
Cl
N
N
N
H
O
O
11
12
Scheme 2. Reagents and conditions: (a) sodium hydrosulfite, sodium bicarbonate, THF/MeOH, rt; (b) carbonyl diimidazole, DCM, rt, 64% (from 8); (c) 2,6-difluorobenzyl
bromide, polystyrene-bound BEMP, acetonitrile, rt, 92%; (d) R2NH2, DMA, microwave 180 °C, 40–85%.
Table 2
Table 3
7-(2,6-Difluorobenzyl)-9-(3-methoxybenzyl)purinone A2A antagonists
cAMP functional assay data for A2A antagonists
Compound
Rat A2A cAMP Kb (nM)
F
7a
7m
7n
62
88
39
F
N
N
12a
12h
12o
237
202
18
O
R2
N
N
N
H
O
12
a 2-(2-thienyl)ethyl group at R2 (12o), which exhibits a low A2A
Ki of 9 nM but a high A1 Ki of 1967 nM.
Compound R2
hA2A Ki
hA1 Ki
A1/A2A
ratio
(nM)a
(nM)b
The A2A receptor is coupled to activation of adenylyl cyclase
through the GTP binding-protein Gs.2,15 Activation of the A2A
receptor in cells increases the intracellular accumulation of cyclic
adenosine monophosphase (cAMP), a response which is blocked
by A2A antagonism.15,16 Functional antagonism at the rat A2A recep-
tor was determined by measurement of cAMP levels in the pres-
ence or absence of inhibitor in CGS 21680-induced rat
pheochromocytoma cells.13 Results of the cAMP functional assay
on selected purinone analogs are listed in Table 3. All the analogs
tested are functional antagonists of the A2A receptor with varying
degrees of activity. Compound 12o exhibits the highest functional
potency with a cAMP Kb value of 18 nM.
12a
12b
12c
12d
12e
12f
7m
12g
12h
12i
12j
12k
12l
12m
12n
Methyl
Ethyl
23
31
60
1110
1053
616
2972
1473
1056
236
1072
1211
1191
1298
1800
1833
8598
10,000
49
34
10
42
14
5
Isopropyl
Benzyl
Methoxyethyl
Cyclopentyl
Cyclopropyl
Cyclopropylmethyl
2-Pyridylmethyl
4-Pyridylmethyl
3-Pyridylmethyl
2-Pyridylethyl
3-Pyridyl
72
106
213
4
499
12
50
66
78
59
2
100
24
20
23
17
33
38
111
259
261
4-Morpholinylethyl
(N,N-
In summary, a series of trisubstituted purinone analogs was dis-
covered as potent A2A antagonists with good selectivity against the
A1 receptor. Compound 12o has a Ki of 9 nM in an A2A binding as-
say and is 220-fold selective against A1. It also exhibits high antag-
onistic potency in a cell-based A2A cAMP functional assay. SAR
studies on this series revealed that a 3-methoxyphenyl group is
strongly preferred at N-9 and a substituted benzyl group is favor-
able at N-7. A relatively high degree of tolerance for the substituent
at the C-2 position was observed, which might serve as a potential
portal to investigate other pharmacological properties in future
studies.
Dimethylamino)ethyl
2-(2-Thienyl)ethyl
12o
9
1967
220
Ki determined by competition binding of [3H]SCH-58261.13
Ki determined by competition binding of [3H]DPCPX.13
a
b
tural features (Table 2). Among the various substituents tested,
small substituents are generally preferred, as indicated by the fol-
lowing trend of A2A affinity: methyl (12a) > ethyl (12b) > isopropyl
(12c) > benzyl (12d) > methoxyethyl (12e) > cyclopentyl (12f). The
analog with the cyclopropylamino substituent (7m) demonstrates
the highest A2A antagonist activity (Ki = 4 nM). Insertion of a meth-
ylene spacer (12g) decreases activity by more than 100-fold. Pyr-
idylmethylamino groups exhibit good activity (12h–12j), with 2-
pyridylmethylamino (12h) being the best (Ki = 12 nM). Insertion
of an additional methylene (12k) or removal of the methylene
spacer (12l) results in weaker activity. Substituents containing a
basic nitrogen (12m, 12n) are not favorable. For A1 receptor activ-
ity, most analogs exhibit Ki values in the micromolar range, result-
ing in the A1/A2A selectivity ratios of 2- to 220-fold. The highest
selectivity against the A1 receptor is achieved by the analog with
References and notes
1. Fredholm, B. B.; Cunha, R. A.; Svenningsson, P. Curr. Top. Med. Chem. 2002, 3,
1349.
2. Dunwiddie, T. V.; Masino, S. A. Annu. Rev. Neurosci. 2001, 24, 31.
3. Simola, N.; Morelli, M.; Pinna, A. Curr. Pharm. Des. 2008, 14, 1475.
4. Palmer, T. M.; Stiles, G. L. Neuropharmacology 1995, 34, 683.
5. Klaasse, K. C.; Ijzerman, A. P.; Grip, W. J.; Beukers, M. Purinergic Signal. 2008, 4,
21.
6. Suchowersky, O.; Gronseth, G.; Perlmutter, J.; Reich, S.; Zesiewicz, T.; Weiner,
W. J. Neurol. 2006, 66, 976.
7. Kase, H.; Mori, A.; Jenner, P. Drug Discov. Today: Therapeut. Strategies 2004, 1,
51.

1402
Y. Shao et al. / Bioorg. Med. Chem. Lett. 19 (2009) 1399–1402
8. Lambert, J. H.; Dashwood, M. R.; Spyer, K. M. Br. J. Pharmacol. 1996, 117, 277.
9. (a) Ohlmeyer, M. H. J.; Swanson, R. N.; Dillard, L.; Reader, J. C.; Asouline, G.;
Kobayishi, R.; Wigler, M.; Still, W. C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 10922;
(b) Nestler, H. P.; Bartlett, P. A.; Still, W. C. J. Org. Chem. 1994, 59, 4723.
10. Cole, A. G.; Stauffer, T. M.; Rokosz, L. L.; Metzger, A.; Dillard, L. W.; Zeng, W.;
Henderson, I. Bioorg. Med. Chem. Lett. 2009, 19, 378.
11. Cole, A. G.; Metzger, A.; Ahmed, G.; Brescia, M.-R.; Chan, R. J.; Wen, J.; O’Brien,
L.; Qin, L.; Henderson, I. Tetrahedron Lett. 2006, 47, 8897.
12. Bensa, D.; Constantieux, T.; Rodriguez, J. Synthesis 2004, 7, 923.
13. Istradefylline was used as the standard for both the A2A and A1
assays. Assays were performed in duplicate and compounds were
tested at least two times. The data were fit to a one-site competition
binding model for IC50 determination and Ki values were calculated
using the Cheng–Prusoff equation. For experimental details on A2A and
A1 binding assays and cAMP functional assay, see Ref. 10 and notes
therein.
14. Headrick, J. P.; Hack, B.; Ashton, K. J. Am. J. Physiol. Heart Circ. Physiol. 2003, 285,
1797.
15. Torvinen, M.; Marcellino, D.; Canal, M.; Agnati, L. F.; Luis, C.; Franco, R. Mol. Biol.
2005, 67, 400.
16. Thibault, N.; Burelout, C.; Harbour, D.; Borgeat, P.; Naccache, P. H.; Bourgoin, S.
G. J. Leukocyte Biol. 2002, 71, 367.