Synthesis of hybrid molecules of caffeine and eudistomin D and its effects on adenosine receptors

Article abstract of DOI:10.1016/j.bmc.2007.02.043

Four hybrid molecules (1 and 12-14) of caffeine and eudistomin D, a β-carboline alkaloid from a marine tunicate, were synthesized, and their affinity and selectivity for adenosine receptors A₁, A_2A, and A₃ were examined. It was found that all the compounds showed better potency as adenosine receptor ligands as compared with caffeine. Among them, a compound (13) possessing a nitrogen at the δ-position of the pyridine ring (δ-N type) showed the most potent affinity for adenosine receptor A₃ subtype, while N-methylation (14) of a pyrrole ring in 13 significantly lowered the potency as adenosine receptor ligands. Compounds (1 and 12) having a nitrogen at the β-position of the pyridine ring (β-N type) showed lower affinity than the corresponding δ-N type compounds (13 and 14), while compounds (10, 11, and 17) lacking a pyrrole ring between the pyridine and pyrimidine rings exhibited almost no affinity to the adenosine receptor subtypes examined.

Full text of DOI:10.1016/j.bmc.2007.02.043

Bioorganic & Medicinal Chemistry 15 (2007) 3235–3240
Synthesis of hybrid molecules of caffeine and eudistomin D
and its effects on adenosine receptors
Kengo Ohshita,a Haruaki Ishiyama,a Koshi Oyanagi,b Hiroyasu Nakatab
and Jun’ichi Kobayashia,*
aGraduate School of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan
bDepartment of Molecular Cell Signaling, Tokyo Metropolitan Institute for Neuroscience, Fuchu, Tokyo 183-8526, Japan
Received 19 December 2006; revised 15 February 2007; accepted 16 February 2007
Available online 22 February 2007
Abstract—Four hybrid molecules (1 and 12–14) of caffeine and eudistomin D, a b-carboline alkaloid from a marine tunicate, were
synthesized, and their affinity and selectivity for adenosine receptors A1, A2A, and A3 were examined. It was found that all the com-
pounds showed better potency as adenosine receptor ligands as compared with caffeine. Among them, a compound (13) possessing a
nitrogen at the d–position of the pyridine ring (d-N type) showed the most potent affinity for adenosine receptor A3 subtype, while
N-methylation (14) of a pyrrole ring in 13 significantly lowered the potency as adenosine receptor ligands. Compounds (1 and 12)
having a nitrogen at the b-position of the pyridine ring (b-N type) showed lower affinity than the corresponding d-N type com-
pounds (13 and 14), while compounds (10, 11, and 17) lacking a pyrrole ring between the pyridine and pyrimidine rings exhibited
almost no affinity to the adenosine receptor subtypes examined.
Ó 2007 Elsevier Ltd. All rights reserved.
1. Introduction
2. Results and discussion
Caffeine exhibits a variety of physiological activities (or
action) including regulation of the blood pressure, respi-
ratory functioning, gastric and colonic activity, urine
volume, and exercise performance.1 The mechanism of
action of caffeine is reported to be competitive antago-
nism to A1 and A2A adenosine receptors,2 induction of
Ca2+-release from sarcoplasmic reticulum (SR), inhibi-
tion of phosphodiesterase, and so on. During our
continuing search for bioactive compounds from marine
organisms, we have found that eudistomin D, a b-carb-
oline alkaloid isolated from a marine tunicate Eudistoma
2.1. Chemistry
A key at route A is photoreaction of 2, derived from 3
and 4 employing Pd-catalyzed coupling, while that at
route B is Pd-catalyzed intramolecular cyclization of 5,
obtained from 6 and 7 employing Stille coupling reac-
tion. The synthesis of 1 via route A is summarized in
Scheme 3. Coupling of 5-amino-1,3-dimethyluracil (8)6
and 3-chloropyridine (9) in DMF with Pd2(dba)3,
olivaceum,3 and its analogs are potent inducers of Ca2+
-
release from SR4 as well as inhibitors of phosphodiester-
ase.5 In the present study, to obtain a useful bioprobe to
examine the mechanism of action of caffeine in details,
we synthesized hybrid molecules of caffeine and eudistomin
D such as compound 1 (Scheme 1) by two routes, A and
B, as shown in Scheme 2.
Keywords: Eudistomin D; Cafffeine; Adenosine receptor; SAR.
*
Corresponding author. Tel.: +81 11 706 3239; fax: +81 11 706
4989; e-mail: jkobay@pharm.hokudai.ac.jp
Scheme 1. A hybrid compound (1) of caffeine and eudistomin D.
0968-0896/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2007.02.043

3236
K. Ohshita et al. / Bioorg. Med. Chem. 15 (2007) 3235–3240
Scheme 2. Retrosynthetic analysis of compound 1 through two routes A and B.
Scheme 3. Synthesis of compound 1 and its congeners (12–14) via route A. Reagents: (a) Pd2(dba)3, XPhos, Cs2CO3/DMF (87%); (b) MeI, NaH/
DMF (82%); (c) hv/i-PrOH; (d) hv/acetone (6%); (e) MeI, NaH/DMF (86%).
Xphos,7 and Cs2CO3 afforded 10 in 87% yield. Methyl-
ation of 10 in DMF with MeI and NaH provided 11.
Photocyclization of 10 in i-PrOH afforded 12 and 13,
while that of 11 in acetone gave 14. Methylation of 12
in DMF with MeI and NaH furnished 1 in 86% yield.
2.2. Biological evaluation
Previously, our group has found that a b-carboline
compound, 7-bromo-eudistomin D (BED), is a good
Ca2+-releaser from SR and its N-methyl derivative of
the pyrrole ring (9-methyl-7-bromoeudistomin D) is 10
times more potent than BED.10 The presence of a nitro-
Alternative synthesis of 1 via route B is summarized in
Table 1 and Scheme 4. A survey of additives for Stille
coupling reaction of 5-amino-6-chloro-1,3-dimethylura-
cil (15)8 and 4-tributylstannyl-3-chloropyridine (16)9 in
DMF revealed that the coupling reaction proceeded
with CuI (entries 3 and 5–9). Switching ligand from
AsPh3 to PPh3 improved yields of 17 (entries 7 vs 8).
Increasing catalyst up to 100 mol% gave 17 in 55% yield
(entries 7 vs 9). Intramolecular cyclization of 17 with
Pd2(dba)3, Xphos, and Cs3CO3 provided 12. Methyla-
tion of 12 in DMF with MeI and NaH furnished 1 in
86% yield.
gen atom in the pyridine ring is essential for the Ca2+
-
releasing activity, while the position of a nitrogen in
the pyridine ring does not affect the potency of the
Ca2+-releasing activity. Actually, compounds possessing
a nitrogen at the a, c, or d position of the pyridine ring
have almost the same activity as the corresponding b-
carboline, such as BED. Therefore, we first had a plan
to synthesize a hybrid molecule (1) of caffeine and eud-
istomin D, expecting as a caffeine analog that could
show a high potency as adenosine receptor ligands. On
the way to the synthesis of des-N-methyl derivative

K. Ohshita et al. / Bioorg. Med. Chem. 15 (2007) 3235–3240
Table 1. Preparation of compound 17 by Stille coupling of 15 with 16
3237
Compound
Palladium
Ligand
mol%
Additives
Yield %
1a
2
3
4
5
6
7
8
9
Pd(PPh3)4
Pd(PPh3)4
Pd(PPh3)4
Pd2(dba)3
Pd(PPh3)4
Pd(PPh3)4
Pd2(dba)3
Pd2(dba)3
Pd2(dba)3
—
—
10
30
—
—
—
4
CuO (1.4 equiv.)
Cul, LiCl (1.4 equiv.)
—
—
10
10
AsPh3
—
—
0
4b
20
30
Cul, LiCl (1.4 equiv.)
Cul, LiCl (1.4 equiv.)
Cul, LiCl (1.4 equiv.)
Cul, LiCl (1.4 equiv.)
Cul, LiCl (1.4 equiv.)
12
16
12
55
PPh3
AsPh3
PPh3
30
30
100
a 65 °C.
b Determined by NMR.
Scheme 4. Synthesis of compound 1 via route B. Reagents: (a) Pd2(dba)3, XPhos, Cs2CO3/DMF (71%); (b) MeI, NaH/DMF (86%).
(12) of 1 having a nitrogen at the b-position of the pyr-
idine ring (b-N type), a compound (13) possessing a
nitrogen at the d-position of the pyridine ring (d-N type)
was produced. The N-methyl derivatives (1 and 14,
respectively) of 12 and 13 were also prepared to examine
the effect of the N-methyl group on the activity of aden-
osine receptors.
pyrrole ring (14 and 1, respectively) in 13 and 12 resulted
in significant loss of the affinity to the receptors. Simi-
larly, the d-N type compound (13) showed higher affinity
for adenosine A2A receptors than the corresponding b-N
type compound (12). N-Methylation (1) of the pyrrole
ring in 12 showed a better affinity to A2A receptor than
the parent molecule (12). However, N-methylation (14)
of the pyrrole ring in 13 showed a lower affinity to
A2A receptor than the parent molecule (13). Compounds
(10, 11, and 17) lacking a pyrrole ring between the pyr-
idine and pyrimidine rings exhibited almost no affinity
to any adenosine receptors examined.
The potency of these newly synthesized compounds as
adenosine receptor ligands was investigated in radioli-
gand binding assays at human recombinant adenosine
A1, A2A, and A3 receptors expressed in membranes of
HEK293T cells as described under Section 4. The results
expressed as Ki values are presented in Table 2. When
the affinity of the test compounds was very low, percent-
age of inhibition at 100 lM was shown. Affinities of ref-
erence ligands, that is, the non-selective caffeine and
50-(N-ethylcarboxamido)adenosine (NECA), the A1-
selective xanthine amine congener (XAC), and the
A2A-selective CGS21680, were also shown for compari-
son. These compounds showed reasonable affinity for
each receptor confirming the validity of our assay.
3. Conclusion
Overall, it was found that (1) the hybrid molecules
(1 and 12–14) of caffeine and eudistomin D synthesized
here showed better potency as adenosine receptor
ligands than caffeine, (2) the d-N type compound (13)
showed the most potent activity for adenosine receptors
tested in this study, especially for A3 subtype, and (3)
N-methylation (14) of the pyrrole ring in 13 significantly
lowered the potency as adenosine receptor ligands.
Among all the compounds examined (Table 2), a d-N
type compound 13 exhibited most potent affinity to all
the three adenosine receptors, A1 A2A, and A3, espe-
cially very high affinity for A3. The Ki value
(0.0139 lM) of 13 to adenosine A3 receptor was compa-
rable to that (Ki = 0.020 lM) of the potent A3 agonist,
NECA. For affinity to the adenosine A1 receptors, the
d-N type compound (13) showed a moderate potency
(Ki = 0.379 lM) but much better than the corresponding
b-N type compound (12), while N-methylation of the
4. Experimental
4.1. Chemistry
4.1.1. Instruments and analyses. The IR spectrum was re-
corded on a JASCO FT/IR-5300 spectrometer. Proton
and carbon NMR spectra were recorded on a Bruker

3238
K. Ohshita et al. / Bioorg. Med. Chem. 15 (2007) 3235–3240
Table 2. Affinities of caffeine and eudistomin D hybrid molecules (1 and 12–14) and its related compounds at human adenosine A1, A2A and A3
receptors
Compound
Kia (lM) or % inhibitionb
A1
A2A
A3
13
12
0.379 0.043
11.7 6.2
13.5 3.2
7.37 2.13
27%
0.893 0.086
10.1 0.01
6.94 2.04
4.92 1.17
4.19 1.87
38%
0.0139 0.0032
0.526 0.156
0.332 0.074
2.05 0.69
3.19 0.08
25%
14
BED (7-bromo-eudistomin D)
1
10
22%
9%
11
17
29%
15%
0.2%
3%
8%
Caffeine
XAC
CGS21680
NECA
49.0 19.6
0.009 0.001
nd
18.1 5.9
nd
9%
nd
0.0462 0.0084
nd
nd
0.020 0.009
nd
a The Ki values are means SEM of two or three separate assays, each performed in duplicate.
b Percentage of inhibition (%) of specific [3H]DPCPX (for A1), [3H]CGS21680 (for A2A) or [3H]NECA (for A3) binding by test compounds at 100 lM
concentration. The binding of each radioactive ligand to membranes prepared from HEK293T cells expressing human adenosine A1, A2A, or A3
receptors was best-fitted to a one-site model of binding with estimated Kd (dissociation constant) values of 5, 52, and 6.5 nM, respectively, and Bmax
values of 8600, 7000, and 310 fmol/mg protein, respectively. nd; not determined.
500 and/or 600 MHz and JEOL 400 MHz spectrometer.
ESI mass spectra were obtained on a JEOL JMS-
SX102A spectrometer.
(KBr) mmax 1706, 1652, 1488 cmꢀ1; HRESIMS calcd
for C12H14N4O2 (M+) m/z 246.1116, found m/z
246.1125.
4.1.2. 1,3-Dimethyl-5-(pyridin-3-ylimino)-imidazolidine-
2,4-dione (10). DMF (0.25 mL) was added to an
oven-dried Schlenk tube charged with 5-amino-1,3-dim-
ethyluracil (8) (78.0 mg, 0.5 mmol), Pd2(dba)3 (23.0 mg,
25 lmol), 2-dicyclohexylphosphino-20,40,60-triisopropylbi-
phenyl (24.2 mg, 50 lmol), and Cs2CO3 (230.3 mg,
0.71 mmol). The mixture was stirred for 10 min at ambi-
ent temperature. 3-Chloropyridine (9; 48 lL, 0.5 mmol)
was added and then stirred for 18 h at 65 °C. The reac-
tion mixture was filtered through Celite and the residue
was purified by flash column chromatography on silica
gel (CHCl3/MeOH, 60:1 ! 20:1) to give 10 (101.1 mg,
4.1.4.
rene-1,3-dione (12) and 2,4-dimethyl-4,9-dihydro-2,4,7,9-
tetraaza-fluorene-1,3-dione (13). solution of 10
(100.2 mg, 0.43 mmol) in i-PrOH (40 mL) was irradiated
with tungsten lamp for 24 h at ambient temperature.
The mixture was concentrated in vacuo and purified
by flash column chromatography on silica gel (CHCl3/
MeOH, 5:1, hexane/EtOAc/acetone, 2:1:1) to give 12
(6.1 mg, 6%) and 13 (2.2 mg, 2%) as brown oil,
respectively.
2,4-Dimethyl-4,9-dihydro-2,4,5,9-tetraaza-fluo-
A
1
Compound 12: H NMR (400 MHz, DMSO-d6) d 12.2
1
87%) as a brown oil. H NMR (400 MHz, CDCl3) d
(1H, s), 8.51 (1H, dd, J = 1.4, 4.2 Hz), 7.87 (1H, dd,
J = 1.4, 8.6 Hz), 7.42 (1H, dd, J = 4.4, 8.4 Hz), 3.99
(3H, s), 3.34 (3H, s); 13C NMR (125 MHz, DMSO-d6)
d 156.2, 151.0, 143.5, 132.9, 125.1, 121.2, 120.8, 115.8,
101.5, 32.1, 27.9; IR (KBr) mmax 1698, 1650 cmꢀ1; HRE-
SIMS calcd for C11H10N4O2 (M+) m/z 230.0804, found
m/z 230.0809.
8.48 (1H, d, J = 8.4 Hz), 8.27 (1H, dt, J = 4.5, 1.6 Hz),
7.38–7.29 (2H, m), 6.15 (1H, s), 3.55 (3H, s), 3.54 (3H,
s), 2.52 (1H, s); 13C NMR (100 MHz, CDCl3) d 161.0,
149.9, 141.6, 139.5, 138.8, 125.9, 123.6, 117.4, 37.3,
28.6; IR (KBr) mmax 3428, 3318, 1698, 1640 cmꢀ1; HRE-
SIMS calcd for C11H12N4O2 (M+) m/z 232.0960, found
m/z 232.0971.
1
Compound 13: H NMR (400 MHz, DMSO-d6) d 12.6
4.1.3.
1,3-Dimethyl-5-(methylpyridin-3-yl-amino)-1H-
(1H, s), 8.86 (1H, s), 8.23 (1H, d, J = 4.7 Hz), 8.00
(1H, d, J = 5.6 Hz), 3.80 (3H, s), 3.34 (3H, s); 13C
NMR (125 MHz, DMSO-d6) d 156.2, 150.8, 137.8,
137.1, 133.7, 125.6, 119.0, 116.3, 115.2, 32.3, 28.0.; IR
(KBr) mmax 1698, 1638 cmꢀ1; HRESIMS calcd for
C11H10N4O2 (M+) m/z 230.0804, found m/z 230.0796.
pyrimidine-2,4-dione (11). To a solution of 10 (51.2 mg,
0.22 mmol) in DMF (1 mL) was added NaH (8.1 mg,
0.33 mmol). After 10 min at room temperature, to the
reaction mixture was added MeI (14 lL, 0.22 mmol)
and then stirred for 3 h. The reaction mixture was fil-
tered through Celite and the residue was purified by
flash column chromatography on silica gel (CHCl3/
MeOH, 5:1) to give 11 (44.4 mg, 82%) as a brown oil.
4.1.5. 2,5,9-Trimethyl-4,9-dihydro-2,4,5,9-tetraaza-fluo-
rene-1,3-dione (14). A solution of 11 (46 mg, 0.19 mmol)
in acetone (160 mL) was irradiated with tungsten lamp
for 20 h at ambient temperature. The mixture was con-
centrated in vacuo and purified by C18 column chroma-
tography (MeOH/H2O, 20:80 ! 80:20) to give 14
(2.9 mg, 6%) as a brown solid. 1H NMR (400 MHz,
DMSO-d6) d 8.54 (1H, dd, J = 4.4, 1.2 Hz), 8.15 (1H,
1H NMR (400 MHz, CDCl3)
d
7.98 (1H, d,
J = 3.0 Hz), 7.92 (1H, d, J = 4.5 Hz), 7.34 (1H, s), 7.00
(1H, dd, J = 8.4, 4.6 Hz), 6.88 (1H, dd, J = 8.5,
3.0 Hz), 3.31 (3H, s), 3.24 (3H, s), 3.06 (3H, s); 13C
NMR (100 MHz, CDCl3) d 161.0, 151.0, 144.8, 142.6,
139.4, 135.7, 123.3, 120.3, 119.7, 39.5, 37.1, 28.2, ; IR

K. Ohshita et al. / Bioorg. Med. Chem. 15 (2007) 3235–3240
3239
dd, J = 8.7, 1.1 Hz), 7.49 (1H, dd, J = 8.6, 4.4 Hz), 4.09
(3H, s), 4.00 (3H, s), 3.37 (3H, s); HRESIMS calcd for
C12H12N4O2 (M+) m/z 244.0960, found m/z 244.0962.
homogenized in lysis buffer containing 50 mM Tris–
HCl buffer (pH 7.4) with a protease-inhibitor mixture
(Roche Diagnostics) and subjected to low-speed centri-
fugation to remove organelles and nuclei. The resulting
supernatant was subjected to centrifugation at 30,000g
for 20 min, and precipitated cell membranes were col-
lected, washed twice, resuspended in the lysis buffer,
and stored at ꢀ80 °C until use.
4.1.6. 2,4,9-Trimethyl-4,9-dihydro-2,4,7,9-tetraaza-fluo-
rene-1,3-dione (1). To a solution of 12 (7.0 mg,
0.03 mmol) in DMF (2 mL) was added NaH (16.2 mg,
0.7 mmol). After 30 min at room temperature, to the
reaction mixture was added a solution of MeI (2 lL,
0.032 mmol) in DMF (0.2 mL) and then stirred for
2 h. The reaction mixture was filtered through Celite
and the residue was purified by chromatography on a
C18 column (MeOH/H2O, 50:50 ! 100:0) to give 1
4.2.3. Adenosine receptor binding assays. Radioligand
binding experiments to adenosine A1, A2A, and A3
receptors were carried out by using [3H]DPCPX,
[3H]CGS21680, and [3H]NECA, respectively. Cell
membranes expressing adenosine A1, A2A, and A3
receptors were incubated with 4 nM [3H]DPCPX,
15 nM [3H]CGS21680, or 32 nM [3H]NECA, respec-
tively, in the presence of 9 to 10 different concentra-
tions of test compounds in 250 lL of assay buffer
containing 50 mM Tris–acetate buffer, pH 7.4, 5 mM
MgCl2, 1 mM EDTA, and 1 U/mL adenosine deami-
nase for 60 min at 25 °C. The incubated mixture was
harvested on Whatman GF/B filters pre-soaked in
0.1% polyethyleneimine by a cell harvester and washed
three times with 50 mM Tris–HCl buffer (pH 7.4). The
radioactivity on the filter was measured by a scintilla-
tion counter. All experiments were carried out two or
three times in duplicate. The nonspecific binding for
adenosine A1, A2A and A3 receptors was defined as
the binding activity in the presence of XAC,
CGS21680, and NECA, respectively, at 10 lM each.
Kd and Bmax values in saturation and inhibition stud-
ies were determined using one-site binding model by
non-linear regression analysis (GraphPad Prism 4;
GraphPad, San Diego, CA, USA).
1
(6.4 mg, 86%) as a yellow solid. H NMR (600 MHz,
DMSO-d6) d 9.11 (1H, br s), 8.29 (1H, br s), 8.01 (1H,
br s), 4.42 (3H, s), 3.90 (3H, s), 3.51 (3H, s); HRESIMS
calcd for C12H12N12O2 (M+) m/z 244.0960, found m/z
244.0963.
4.1.7. 5-Amino-6-(3-chloropyridin-4-yl)-1,3-dimethyl-1H-
pyrimidine-2,4-dione (17). DMF (1.5 mL) was added to
an oven-dried Schlenk tube charged with 5-amino-6-
chloro-1,3-dimethyluracil (15) (50.0 mg, 0.262 mmol),
Pd2(dba)3 (120.2 mg, 0.131 mmol), PPh3 (69.3 mg,
0.262 mmol), CuI (70.7 mg, 0.367 mmol), and LiCl
(16.0 mg, 0.367 mmol). The mixture was stirred for
30 min at 110 °C. 3-Chloro-4-tributyltinstannylpyridine
(16; 80 lL, 0.262 mmol) was added and then stirred
for 19 h at 110 °C. The reaction mixture was filtered
through Celite and the residue was purified by flash col-
umn chromatography on silica gel (hexane/CHCl3,
60:1 ! 20:1, and then hexane/EtOAc 6:1 ! 2:3) to give
1
17 (38.2 mg, 55%) as a brown oil. H NMR (400 MHz,
CDCl3) d 8.83 (1H, s), 8.70 (1H, d, J = 4.9 Hz), 7.32
(1H, d, J = 4.9 Hz), 3.47 (3H, s), 3.08 (3H, s), 2.83
(2H, br s); 13C NMR (100 MHz, CDCl3) d 159.8,
150.9, 150.1, 149.1, 137.8, 131.4, 125.0, 124.1, 119.9,
33.3, 28.7.; HRESIMS calcd for C11H11N4O2Cl (M+)
m/z 266.0570, found m/z 266.0569.
Acknowledgments
We thank S. Oka, Center for Instrumental Analysis,
Hokkaido University, for ESIMS measurements. This
work was partly supported by a Grant-in-Aid for Scien-
tific Research from the Ministry of Education, Science,
Sports, and Culture of Japan.
4.2. Biological assays
4.2.1. Radioligand materials. [3H]-8-Cyclopentyl-1,3-
dipropylxanthine ([3H]DPCPX), [3H]-2-[4-(2-carboxy-
ethyl)phenethylamino]-50-N-ethylcarboxamidoadeno-
sine
([3H]CGS21680),
and
[3H]-50-N-ethyl-
References and notes
carboxamidoadenosine ([3H]NECA) were purchased
from Perkin-Elmer (Boston, MA, USA). Unless other-
wise stated, all other materials used for ligand binding
assay were purchased from Sigma–Aldrich (St. Louis,
MO, USA).
1. Laura, M. J.; Roland, R. G. Psychopharmacology 2004,
176, 1–29.
2. Bertorelli, R.; Ferri, N.; Adami, M.; Ongini, E. Drug Dev.
Res. 1996, 37, 65–72.
3. (a) Kobayashi, J.; Harbour, G. C.; Gilmore, J.; Rinehart,
K. L., Jr. J. Am. Chem. Soc. 1984, 106, 1526–1528; (b)
Rinehart, K. L., Jr.; Kobayashi, J.; Harbour, G. C.;
Gilmore, J.; Mascal, M.; Holt, T. G.; Shield, L. S.;
Lafargue, F. J. Am. Chem. Soc. 1984, 106, 1524–1526; (c)
Rinehart, K. L., Jr.; Kobayashi, J.; Harbour, G. C.;
Gilmore, J.; Mascal, M.; Holt, T. G.; Shield, L. S.;
Lafargue, F. J. Am. Chem. Soc. 1987, 109, 3378–3387.
4. Nakamura, Y.; Kobayashi, J.; Gilmore, J.; Mascal, M.;
Rinehart, K. L., Jr.; Nakamura, H.; Ohizumi, Y. J. Biol.
Chem. 1986, 261, 4139–4142.
4.2.2. Membrane preparations. HEK293T cell lines tran-
siently expressing human adenosine A1, A2A, and A3
receptors were used as the receptor source in this study.
Plasmids encoding human adenosine A1, A2A or A3
receptor construct obtained from UMR cDNA Re-
source Center (Rolla, MO, USA) were transiently trans-
fected into HEK293T cells using Effectene (Quiagen).
Cells were maintained at 37 °C in humidified air con-
taining 5% CO2 in Dulbecco’s modified Eagle’s medium
supplemented with 10% Fetal Bovine Serum, 100 lg/mL
kanamycin for 48 h. The cells were harvested and
5. Kobayashi, J.; Taniguchi, M.; Hino, T.; Ohizumi, Y.
J. Pharm. Pharmacol. 1988, 40, 62–63.

3240
K. Ohshita et al. / Bioorg. Med. Chem. 15 (2007) 3235–3240
6. Zajac, M. A.; Zakrzewski, A. G.; Kowal, M. G.; Narayan,
S. Synth. Commun. 2003, 19, 3291–3297.
8. (a) Yue, W. S.; Li, J. J. Org. Lett. 2002, 4, 2201–2203; (b)
Gribble, G. W.; Saulnier, M. G. Heterocycles 1993, 35,
151–169.
9. Tapolcsanyi, P.; Krajsovszky, G.; Ando, R.; Lipcsey, P.;
Horvath, G.; Matyus, P.; Riedl, Z.; Hajos, G.; Lemiere, G.
L. F. Tetrahedron 2002, 58, 10137–10143.
7. (a) Strieter, E. R.; Blackmond, D. G.; Buchwald, S. L.
J. Am. Chem. Soc. 2003, 125, 13978–13980; (b) Huang, X.;
Anderson, K. W.; Zim, D.; Jiang, L.; Klapars, A.;
Buchwald, S. L. J. Am. Chem. Soc. 2003, 125, 6653–
6655; (c) Kim, Y. M.; Yu, S. J. Am. Chem. Soc. 2003, 125,
1696–1697.
10. Kobayashi, J.; Ishibashi, M.; Nagai, U.; Ohizumi, Y.
Experientia 1989, 45, 782–783.