Synthesis and biological evaluation of disubstituted N6- cyclopentyladenine analogues: The search for a neutral antagonist with high affinity for the adenosine A1 receptor

Article abstract of DOI:10.1016/j.bmc.2003.10.023

Novel 3,8- and 8,9-disubstituted N⁶-cyclopentyladenine derivatives were synthesised in moderate overall yield from 6-chloropurine. The derivatives were made in an attempt to find a new neutral antagonist with high affinity for adenosine A₁ receptors. N⁶-Cyclopentyl-9- methyladenine (N-0840) was used as a lead compound. Binding affinities of the new analogues were determined for human adenosine A₁ and A ₃ receptors. Their intrinsic activity was assessed in [ ³⁵S]GTPγS binding experiments. Elongation of the 9-methyl of N-0840 to a 9-propyl substituent was very well tolerated. A 9-benzyl group, on the other hand, caused a decrease in adenosine A₁ receptor affinity. Next, the 8-position was examined in detail, and affinity was increased with appropriate substitution. Most derivatives were A₁-selective and 20 of the new compounds (6-9, 15-21, 23-26, 28, 31, 33, 35, and 36) had higher adenosine A₁ receptor affinity than the reference substance, N-0840. Compound 31 (N⁶-cyclopentyl-8-(N-methylisopropylamino)-9- methyladenine, LUF 5608) had the highest adenosine A₁ receptor affinity, 7.7 nM. In the [³⁵S]GTPγS binding experiments, derivatives 5, 14, 22, 23, 25, 26, 33 and 34 did not significantly change basal [³⁵S]GTPγS binding, thus behaving as neutral antagonists. Moreover, four of these compounds (23, 25, 26, and 33) displayed a 4- to 10-fold increased adenosine A₁ receptor affinity (75-206 nM) compared to N-0840 (852 nM). In summary, we synthesised a range of N-0840 analogues with higher affinity for adenosine A₁ receptors. In addition, four new derivatives, LUF 5666 (23), LUF 5668 (25), LUF 5669 (26) and LUF 5674 (33), behaved as neutral antagonists when tested in [ ³⁵S]GTPγS binding studies. Thus, these compounds have improved characteristics as neutral adenosine A₁ receptor antagonists.

Full text of DOI:10.1016/j.bmc.2003.10.023

Bioorganic & Medicinal Chemistry12 (2004) 139–149
Synthesis and biological evaluation of disubstituted
N -cyclopentyladenine analogues: the search for a neutral
6
antagonist with high affinity for the adenosine A receptor
1
a,y
a
Rianne A. F. de Ligt, Pieter A. M. van der Klein, Jacobien K. von Frijtag Drabbe
a
a
b
a
Ku
¨
nzel, Anna Lorenzen, Fatna Ait El Maate, ShellyFujikawa, Rosemarijn van
a
a
a
a,
Westhoven, Thijs van den Hoven, Johannes Brussee and Ad P. IJzerman *
aDivision of Medicinal Chemistry, Leiden/Amsterdam Center for Drug Research, Leiden University, PO Box 9502,
300 RA Leiden, The Netherlands
2
Institute of Pharmacology, University of Heidelberg, D-69120 Heidelberg, Germany
b
Received 3 January2003; accepted 10 October 2003
6
Abstract—Novel 3,8- and 8,9-disubstituted N -cyclopentyladenine derivatives were synthesised in moderate overall yield from
6
receptors. N -Cyclopentyl-9-methyladenine (N-0840) was used as a lead compound. Binding affinities of the new analogues were
-chloropurine. The derivatives were made in an attempt to find a new neutral antagonist with high affinityfor adenosine A
1
6
3
5
1 3
determined for human adenosine A and A receptors. Their intrinsic activitywas assessed in [ S]GTPgS binding experiments.
Elongation of the 9-methyl of N-0840 to a 9-propyl substituent was very well tolerated. A 9-benzyl group, on the other hand,
caused a decrease in adenosine A receptor affinity. Next, the 8-position was examined in detail, and affinity was increased with
1
appropriate substitution. Most derivatives were A
6) had higher adenosine A receptor affinitythan the reference substance, N-0840. Compound 31 (N -cyclopentyl-8-(N-methyl-
isopropylamino)-9-methyladenine, LUF 5608) had the highest adenosine A
experiments, derivatives 5, 14, 22, 23, 25, 26, 33 and 34 did not significantlychange basal [ S]GTPgS binding, thus behaving as
neutral antagonists. Moreover, four of these compounds (23, 25, 26, and 33) displayed a 4- to 10-fold increased adenosine A
receptor affinity (75–206 nM) compared to N-0840 (852 nM). In summary, we synthesised a range of N-0840 analogues with higher
affinityfor adenosine A receptors. In addition, four new derivatives, LUF 5666 (23), LUF 5668 (25), LUF 5669 (26) and LUF 5674
33), behaved as neutral antagonists when tested in [ S]GTPgS binding studies. Thus, these compounds have improved character-
istics as neutral adenosine A receptor antagonists.
2003 Elsevier Ltd. All rights reserved.
1
-selective and 20 of the new compounds (6–9, 15–21, 23–26, 28, 31, 33, 35, and
6
3
1
3
5
1
receptor affinity, 7.7 nM. In the [ S]GTPgS binding
35
1
1
3
5
(
1
#
1
. Introduction
cyclopentyl-9-methyladenine (N-0840, compound 3), (Æ)-
6
N -(endonorbornan-2-yl)-9-methyladenine (N-0861), and
N -(5 -endohydroxy-endonorbornan-2-yl)-9-methyl-
6
0
adenine (WRC-0342). However, these compounds had
Most of the ‘classical’ adenosine A receptor antago-
1
nists, for example DPCPX (1,3-dipropyl-8-cyclo-
pentylxanthine) and CGS 15943 (9-chloro-2-(2-
furyl)[1,2,4]triazolo[1,5-c]quinazolin-5-amine), have been
relativelypoor affinityfor the human adenosine A recep-
1
1
tor, 1549, 1023, and 1047 nM, respectively. The purpose
1
shown to act as inverse agonists. In the same study,
3
of our research was, therefore, to design novel neutral
antagonists with higher affinityfor human adenosine A 1
receptors.
5
[
neutral antagonists for human adenosine A receptors, N -
S]GTPgS binding experiments revealed also three
6
1
N-0840 was used as a lead compound, which was pre-
dominantlysubstituted at two different positions ( Fig. 1).
Keywords: Adenosine A
1
receptor; Inverse agonism; Neutral antago-
nist; N-0840 derivatives; [35S]GTPgS binding.
*
Corresponding author. Tel.: +31-71-527-4651; fax: +31-71-527-
565; e-mail: ijzerman@lacdr.leidenuniv.nl
Current address: TNO Nutrition and Food Research, Zeist, The
Netherlands.
The N9-position was primarilyexplored to introduce sub-
stituents with the capacityfor radiolabelling. Moreover,
to increase adenosine A receptor affinity, we focussed
4
y
1
0
968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.10.023

140
R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
binding experiments for adenosine A and A receptor
1
3
affinity. Furthermore, their intrinsic activity was deter-
3
5
mined with [ S]GTPgS binding studies to establish
whether these compounds acted as adenosine A1
receptor neutral antagonists.
2. Results and discussion
2.1. Chemistry
The synthesis of the 3,8- and 8,9-disubstituted N6-
cyclopentyladenines (compounds 5–37) was accom-
plished via the routes illustrated in Schemes 1–4.
6
Figure 1. Chemical structure of N -cyclopentyl-9-methyladenine (N-
840) and effects on adenosine A receptor affinityinduced bymodifi-
cations at different positions of the adenine molecule.
0
1
6
First, 8-bromo-N -cyclopentyladenine (5) and 8-bromo-
6
N -cyclopentyl-9-methyl-adenine (6) were prepared
3
,12,13
on substitution at the 8-position, since other positions
had alreadybeen examined closel y. For example, the
from 6-chloropurine (compound 1, Scheme 1).
6
During the bromination of N -cyclopentyl-9-methyl-
6
6
N -cyclopentyl is known to induce high adenosine A
2
adenine (3), partial dealkylation of the N -position
14
occurred, which resulted in product 7.
1
,3
receptor affinityand selectivit y. Furthermore, small
substituents, for example a chlorine atom, at the 2-
3
4
position of adenines or adenosines onlyhave limited
effects on adenosine A receptor affinity. Additionally,
larger and bulkier substituents at this position of ade-
nosines seem to favour A
selectivity, and this finding mayalso be true for cor-
9
responding substitution at adenines.
To investigate the effect of N9-substitution (Fig. 1), the
appropriate alkyl halide was reacted with compound 5
1
1
5
in the presence of K CO and DMF. Although 9-
2 3
and/or A3 receptor
A
alkylation is usually dominant, minor amounts of
16,17
alkylation at N3 and/or N7 have been reported.
2
5À8
Probablydue to the bromine atom on C8 a mixture of
two compounds with the same mass was formed. These
mixtures were separated and purified with column
chromatographyto provide the products 8–14 in mod-
erate to low yields (Scheme 2). Assignment of the two
compounds was made byNMR anal ys is and bycom-
parison of the methylated products with compound 6,
which was prepared via another route (Scheme 1). Che-
mical shifts of the H-2 proton for N9-substituted com-
pounds (6, 8, 9, 10) were between 8.35 and 8.41 ppm
and for the N3-substituted compounds (12, 13, 14)
between 7.95 and 7.97 ppm. Similarly, chemical shifts of
Finally, Linden and co-workers10 investigated the C8-
position of adenines to some extent. Theys yn thesised 8-
substituted N -norbornyl-9-methyladenines and found
that nitrogen-containing groups at this position enhance
6
adenosine A receptor affinity. Moreover, introduction
1
of alkynyl chains on the C8-position of adenosine led to
1
1
selective adenosine A receptor antagonists.
3
In this paper, we describe the synthesis of 32 novel N-
0
840 derivatives, which were all tested in radioligand
6
Scheme 1. Synthesis of 8-bromo-N -cyclopentyladenine (5), 8-bromo-
6
N -cyclopentyl-9-methyladenine (6) and 8-bromo-9-methyladenine (7):
ꢀ
6
(
i) NaH/DMF, MeI; (ii) cyclo-pentylamine, n-BuOH/Et
iii) Br in Na HPO buffer.
3
N, 120 C;
Scheme 2. Synthesis of N3- and N9-substituted N -cyclopentylade-
nines: (i) alkylhalide, K CO , DMF.
2 3
(
2
2
4

R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
141
that was not found for the N9-substituted compounds
data not shown).
(
When allyl bromide was used, only the N9-substituted
product 8 was formed. The N9-substituents were chosen
for their suitabilityfor future radiolabelling. The propyl
3
group was included to estimate the effect of H reduction
2
of the allyl’s double bond. Furthermore, a benzyl group
was introduced at this site as a template to verifywhether
the N9-position held enough space to accommodate an
(iodinated) 4-iodobenzyl substituent.
Next, we explored substitution at the 8-position in an
attempt to increase adenosine A receptor affinityof
1
these N-0840 derivatives (Fig. 1). The derivatives 6, 8,
and 9 were dissolved in MeOH and refluxed in the pre-
sence of KOtBu, to synthesise their 8-methoxy analo-
gues, resulting in high yields of products 15 and 16 and
1
8
a low yield of derivative 17 (Scheme 3). A similar
6
procedure was carried out with 8-bromo-N -cyclo-
pentyl-9-methyladenine (6) in the presence of ethanol,
isopropanol, and n-propanol, giving good to high yields
of the products 18, 19, and 20. For comparison with 18,
6
N -cyclopentyl-8-thioethyl-9-methyladenine (21) was
synthesised.
We also tried to introduce a hydroxyl group at the 8-
6
6
6
Scheme 3. Synthesis of 9-alkyl-N -cyclopentyl-8-oxoadenines and N -
0
position. During this reaction (Scheme 3) N -cyclo-
pentyl-9-methyl-8-oxoadenine (22) was formed in a
cyclopentyl-8-(ethylthio)-9-methyladenine: (i) R -OH, KtBuO; (ii)
ꢀ
EtSH, KtBuO, EtOH, 50 C; (iii) 1 M aq NaOH, reflux, 3 h.
6
tautomeric equilibrium with the desired N -cyclopentyl-
1
9
8-hydroxy-9-methyladenine. However, high adenosine
A receptor affinityfor this compound was not antici-
1
1
0
pated, since Martin et al. reported a K value of 10 mM
for N -endonorbornyl-8-hydroxy-9-methyladenine com-
pared to 10 nM for the non-substituted analogue.
i
6
Additionally, in their study,10 the authors adequately
demonstrated the favourable presence of an amino
group, and therefore in our study, amino substituents
were investigated in more detail. In Scheme 4, the var-
6
ious methods to generate 8-amino substituted N -cyclo-
2
0,21
pentyl-9-methyladenines (compounds 23–37) are
summarised. Compounds 24–27 were synthesised from
6
N -cyclopentyl-8-(N-methylamino)-9-methyladenine (23)
under mild conditions at room temperature in
1
2,22
DMF.
In addition, the other amines were capable of
replacing the 8-bromine atom in 6 directly, resulting in
the products 28–37. For this route higher temperatures
and a different solvent (dioxane) were required.
2.2. Biological evaluation
All compounds were tested in radioligand displacement
experiments to determine their affinityfor human ade-
nosine A and A receptors expressed on CHO and
6
Scheme 4. Synthesis of 8-amino substituted N -cyclopentyl-9-methyl-
1
3
3 2 2
adenines: (i) H CNH /H O, rt; (ii) NaH/DMF, alkylhalide, rt; (iii)
3
HEK293 cells, respectively. The radioligand [ H]DPCPX
1
amine/dioxane, ꢀ.
25
was used on adenosine A receptors, while [ I]-
1
6
0
boxamidoadenosine] was used on adenosine A recep-
ABMECA [N -(4-amino-3-iodobenzyl)-5 -N-ethylcar-
6
the N –H proton were between 5.58 and 5.74 ppm for
the N9 substituted compounds and between 6.23 and
6.50 ppm for the N3 substituted compounds. NOESY
experiments on 12, 13 and 14 showed a substantial
correlation between the C–H protons on N3 and C2
3
tors. Tables 1 and 2 summarise the results from these
ligand binding experiments. Except for derivatives 5 and
14, the compounds had poor or negligible affinityfor
adenosine A3 receptors. Moreover, these compounds

142
R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
6
Table 1. Affinities of 3,8- and 8,9-disubstituted N -cyclopentyladenine analogues at human adenosine A
1
and A
3
receptors expressed as K
i
values
(in nMÆSEM, n=3) or percentage displacement at 10 mM
0
Compd
C8
N9-R
N3-R
K
i
(nM) or % displacement at 10 mM
receptorb
A
1
receptora
2.4Æ0.1
A
3
A
3
/A
708
0.6
1
DPCPX
N-0840
—
H
Br
Br
Br
Br
Br
Br
H
Br
Br
Br
OCH
OCH
OCH
—
Methyl
H
Methyl
Methyl
Allyl
Propyl
Benzyl
Benzyl
—
—
—
—
—
—
—
—
—
—
—
—
1700Æ170
15%
852Æ163
5
6
7
2646Æ623
43Æ7
1670Æ630
37%
23%
c
467Æ61
35Æ9
8
9
5730Æ1520
7760Æ170
44%
28%
43%
10,600Æ3100
1860Æ420
32%
38%
21%
46%
5600Æ730
9190Æ2200
39%
164
235
33Æ1
10
11
12
13
14
15
16
17
18
19
20
21
22
1220Æ240
1810Æ360
2760Æ167
995Æ160
870Æ140
208Æ36
270Æ20
120Æ7
Methyl
Propyl
Benzyl
—
11
2
3
3
3
Allyl
Propyl
Methyl
Methyl
Methyl
Methyl
Methyl
Methyl
—
—
—
—
—
—
—
OC
OCH(CH
OC
SC
¼O
2
H
5
106Æ10
40Æ7
3
)
2
140
39
3
H
7
235Æ56
224Æ70
1610Æ530
2
H
5
17%
a
b
c
3
++
Displacement of [ H]DPCPX from CHO-A
1
membranes.
i
membranes. K determined if displacement >50%.
125
Displacement of [ I]IBMECA from HEK293-A
6
3
No N -cyclopentyl substituent.
(
5 and 14) along with derivatives 13 and 20 were only
moderatelyA -selective, that is theypossessed an A /A
1
With an appropriate substituent at the 8-position of N-
0840 (Fig. 1) adenosine A1 receptor affinitywas
increased (Tables 1 and 2). For example, elongation of
the O-alkyl substituent (products 17–20) was favourable
1
3
ratio <100. All other compounds, on the other hand,
showed good selectivityfor adenosine A 1 receptors.
for adenosine A receptor binding, and product 19 (O-
1
Furthermore, most of the novel derivatives (6–9, 15–21,
3–26, 28, 31, 33, 35, and 36) had significantlyhigher
affinityfor the human adenosine A 1 receptor (7.7–467
nM) compared to the reference compound N-0840 (852
nM, Table 1).
isopropoxy) was the best derivative in this series. Com-
paring derivatives 18 (ethoxy), 21 (thioethyl), and 28
(aminoethyl), a substituent with an oxygen atom gave
the best results. However, Martin et al. alreadyargued
that this C8-position does not tolerate hydrogen bond-
ing substituents. Consequently, replacement of the free
hydrogen in product 28 bya meth yl group, as in deri-
vative 25, resulted in an almost 4-fold increased adeno-
sine A1 receptor affinity. In general, the secondary
amines as substituents (compounds 23, 28–30) led to
2
1
0
Elongation of the 9-methyl group of 6 to a propyl group
(derivative 9) hardlyaffected adenosine A 1 receptor
affinity, while 9-benzyl substitution (compound 10)
caused a strong decrease in affinity(43 and 33 nM vs
1
compounds (12–14) also had relativelylow affinities.
220 nM, Table 1). The corresponding N3-substituted
lower adenosine A receptor affinities than the tertiary
1
(compounds 24–27, 31, 33, 34) and cyclic (compounds
35, 36) amines. Moreover, lengthening of the methyl
(25) to an ethyl group (derivative 33), and ‘ring closure’
to a six-membered ring, as in derivative 36, slightly
improved affinity. Apparently there is sufficient space to
accommodate bulkier substituents. Derivative 31 (LUF
5608) with a branched N-methyl-N-isopropylamino
Thus, the most likelycandidate for radiolabelling may
be derivative 8, in which the 9-allyl group can be
3
reduced with H to yield a radioligand analogue of
2
3
product 9. Along these lines, Thompson et al. found
6
the following rank order in affinities of 9-substituted N -
cyclopentyladenines for rat adenosine A receptor, 9-
1
phenyl<9-cyclopentyl<9-methylꢁ9-ethyl. However, a
substituent on C8 had the best adenosine A receptor
1
affinityin this stud y, that is 7.7 nM ( Table 2).
high affinityin the low nanomolar range is preferable
for a radioligand, and 8-bromo-N -cyclopentyl-9-pro-
pyladenine did not meet this criterion with its affinity of
6
Next, the efficacyof the compounds was investigated
3
5
3
3 nM (Table 1).
with [ S]GTPgS binding experiments (Table 3).

R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
143
6
Table 2. Affinities of 8-substituted N -cyclopentyl-9-methyladenine (N-0840, see also Table 1) analogues at human adenosine A
expressed as K
values (in nMÆSEM, n=3) or percentage displacement at 10 mM (assays as in footnotes to Table 1)
1
and A
3
receptors
i
Compd
C8 substituent
K
i
(nM) or % displacement at 10 mM
Compd
C8 substituent
i
K (nM) or % displacement at 10 mM
A
1
receptor
A
3
receptor (%)
A
1
receptor
A
3
receptor (%)
3 1
A /A
N-0840
H
NHCH
852Æ163
206Æ27
15
23
2
3
8
3
24
169Æ28
20
2
NHC
2
H
5
344Æ97
19
54
29
30
32
35
36
37
2040Æ200
25
26
27
31
33
34
89Æ9
28
3560Æ1530
5900Æ1300
403Æ63
37
32
28
40
11
160Æ23
31
1011Æ125
7.7Æ1.4
75Æ8
n.d.
14,200Æ1960
1844
68Æ6
18
30
2840Æ307
706Æ70
[35
CHO cells expressing the human adenosine A receptor
S]GTPgS binding was measured on membranes of
[ S]GTPgS binding, thus behaving as neutral adenosine
35
A receptor antagonists. Moreover, four of these pro-
ducts (23, 25, 26, and 33) showed a 4- to 10-fold
increased adenosine A receptor affinitycompared to N-
1
1
+
+
at high density(CHO-A 1 , ꢂ3 pmol/mg of protein).
DPCPX was included as a reference adenosine A
inverse agonist and N-0840 as a neutral antagonist.
1
1
1
0840, while the others had similar (14 and 34) or lower
(5 and 22) affinity( Tables 1 and 2). All other com-
pounds, except for compounds 10, 11 and 13, sig-
nificantlydecreased basal [ S]GTPgS binding to a
varying extent, and acted as (partial) inverse agonists.
Moreover, compound 31, with the highest affinityfor
The high receptor expression was necessaryto distin-
guish neutral antagonists from inverse agonists.
Namely, DPCPX as well as some of the novel com-
pounds (5–9, 18–22, 28, and 31) were unable to mod-
3
5
3
5
ulate basal [ S]GTPgS binding to membranes from
CHO cells expressing a lower (ꢂ0.65 pmol/mg of pro-
adenosine A receptors, showed the largest decrease of
1
3
5
tein) adenosine A receptor number (data not shown).
1
basal [ S]GTPgS binding, to 56% of control. This
compound has an N-methyl, N-isopropylamino group
on the C8 position, as is the case in WRC-0571 [8-(N-
Since agonists activate G proteins via receptor activa-
3
5
6
0
tion, basal [ S]GTPgS binding (set to 100%) increases
in the presence of an agonist, while inverse agonists show
methylisopropylamino)-N -(5 -endohydroxy-endonor-
bornan-2-yl)-9-methyladenine], a partial inverse ago-
3
5
1
the opposite effect ([ S]GTPgS binding <100%). For
nist.
Comparable compounds without
a
C8-
3
5
example, DPCPX decreased basal [ S]GTPgS binding to
CHO-A1 membranes to 60% of control (Table 3). A
substituent, WRC-0342 and N-0861, acted as neutral
antagonists. Thus, the inverse agonistic behaviour of
WRC-0571 maybe caused bythe 8-( N-methyl-N-iso-
propyl)-amino group, which is also present in com-
pound 31. The nature of 8-amino substitution appears
to matter, since some 8-amino substituted N-0840 deri-
vatives behaved as inverse agonists (e.g., 24, 28–31, 35,
+
+
neutral antagonist, such as N-0840, had no (significant)
3
5
effect on basal [ S]GTPgS binding (107%, Table 3).
Eight of the newlys yn thesised compounds ( 5, 14, 22, 23,
2
5, 26, 33 and 34) did not significantlychange basal

144
R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
6
Table 3. Affinities of 3,8- and 8,9-disubstituted N -cyclopentylade-
nine analogues at human adenosine A
Next, we evaluated the intrinsic activityof all com-
3
5
1
receptors expressed as K
values (in nMÆSEM), and their effect on [ S]GTPgS binding (% of
i
pounds. In [ S]GTPgS binding experiments, eight deri-
vatives (5, 14, 22, 23, 25, 26, 33, 34) behaved as neutral
antagonists, and four of these products (23, 25, 26, and
35
basal)
% [35S]GTPgS
binding
% [35S]GTPgS
binding
33) showed a gain in adenosine A1 receptor affinity
compared to N-0840. On the other hand, derivative 31
Compd
Compd
a,b
a,b
behaved as full inverse agonist, decreasing basal
35
DPCPX
N-0840
60Æ8
107Æ10
91Æ4
76Æ4
81Æ4
76Æ4
64Æ3
162Æ13
156Æ6
84Æ1
151Æ6
93Æ4
75Æ1
75Æ1
61Æ2
74Æ4
62Æ4
60Æ3
21
22
61Æ4
104Æ6
91Æ2
85Æ1
90Æ1
90Æ1
n.d.
72Æ4
76Æ1
73Æ2
56Æ2
n.d.
92Æ1
99Æ2
77Æ3
80Æ3
n.d.
[
S]GTPgS binding to 56%.
5
6
7
8
9
23 (LUF 5666)
24
25 (LUF 5668)
26 (LUF 5669)
In summary, we have synthesised various 3,8- and 8,9-
6
disubstituted N -cyclopentyladenine derivatives with
higher adenosine A receptor affinitythan the reference
1
27
28
29
30
compound N-0840. Four of the newlys yn thesised com-
pounds, LUF 5666 (23), LUF 5668 (25), LUF 5669 (26),
and LUF 5674 (33), are classified as neutral antagonists
for human adenosine A receptors with substantially
1
higher affinities than the lead compound N-0840. Deri-
vative 31 (LUF 5608), with the highest adenosine A1
10
11
12
13
14
15
16
17
18
19
20
31 (LUF 5608)
32
33 (LUF 5674)
34
35
36
37
receptor affinityof this series, acted as a full inverse
agonist on this receptor.
n.d., not determined.
a
4. Experimental
At 10ꢃK
i
.
b
++
1
membranes (basal [35S]GTPgS binding ꢂ450 cpm/mg of
CHO-A
protein).
All chemicals and solvents used were commercially
available and of analytical grade, unless stated other-
wise. [ H]DPCPX (111.6 Ci/mmol), [ S]GTPgS (1250
Ci/mmol) and [ H]cAMP (32.4 Ci/mmol) were pur-
3
35
3
chased from NEN (Du Pont Nemours, ’s-Hertogen-
bosch, The Netherlands). CPA, DPCPX, and N-0840
were obtained from Research Biochemicals Inc.
3
(
6), while others showed neutral antagonistic behaviour
e.g., 23, 25, 26, 33, 34).
(Natick, USA).
Interestingly, compounds 10, 11 and 13 increased basal
3
5
[
pounds seem to act as partial agonists, since the full
S]GTPgS binding to 151–162% of basal. These com-
4.1. Chromatography
6
adenosine A receptor agonist, N -cyclopentyladenosine
1
Thin-layer chromatography was carried out using alu-
minium sheets with silica gel F254 from Merck. Spots
were visualised under UV light (254 nm). Preparative
column chromatographywas performed on silica gel
(230–400 mesh, ASTM).
(
An explanation for these observations is not readily
CPA), showed an increase to 317% (data not shown).
3
5
given. However, their increase in basal [ S]GTPgS
binding was limited, compared to the effect of the full
agonist CPA (approximately20% of CPA’s effect).
Thus, if these compounds are indeed partial agonists,
theyonlyhave a verylow intrinsic activit y. The fact that
their partial agonistic behaviour was not observed in
cAMP determinations (data not shown) also indicates a
verylow intrinsic activity.
4.2. Instruments and analyses
Elemental analyses were performed for C, H, and N
(Department of Analytical Chemistry, Leiden Uni-
1
3
versity, The Netherlands). C NMR spectra were mea-
sured at 50.1 MHz with a JEOL JNM-FX 200
spectrometer equipped with a PG 200 computer oper-
3. Conclusions
1
ating in the Fourier transform mode. H NMR spectra
6
The 3,8- and 8,9-disubstituted N -cyclopentyladenine
derivatives described in the present studywere s yn the-
sised, starting from commerciallyavailable 6-chloro-
purine. It appeared that in the synthetic process N9-
substitution was favoured over N3-substitution. Fur-
thermore, elongation of the 9-methyl group in N-0840
to a 9-propyl was tolerated without loss of adenosine A1
receptor affinity. A broad range of substituents was
introduced at the 8-position. In general, tertiaryalipha-
tic and cyclic amines represented derivatives with good
were measured at 200 MHz, using the above-mentioned
spectrometer, or at 300 MHz, using a Bruker WM-300
spectrometer equipped with an ASPECT-2000 computer
operating in the Fourier transform mode. Chemical
shifts are given in ppm (d) relative to tetramethylsilane
(TMS) as internal standard. NOESY experiments were
obtained with a Bruker DMX-600 instrument (tmix: 1 s).
All high-resolution mass spectra were measured on a
Finnigan MAT 900 mass spectrometer equipped with
a direct insertion probe for EI experiments (70 eV
with resolution 1000). Melting points were determined
adenosine A receptor affinity. Compound 31, with a
1
branched tertiaryamine, had a more than 100-fold
higher affinitythan the reference compound N-0840, 7.7
and 852 nM, respectively.
¨
in a Buchi capillarymelting point apparatus and are
uncorrected.

R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
145
1
ꢀ
4
6
.2.1. 6-Chloro-9-methyladenine (2). To a suspension of
-chloropurine (1, 3.0 g, 19.4 mmol) in DMF (95 mL),
4.2.5.2. 7. Yield 0.46 g (31%). Mp 260 C (dec). H
NMR (DMSO-d ) d 3.64 (s, 3H, CH ), 7.35 (s, 2H,
6
3
NaH (60% in mineral oil, 0.776 g, 19.4 mmol) was
added in 20 min. After stirring for 1 h, the reaction
mixture was cooled in an ice bath while methyl iodide
NH ), 8.12 (s, 1H, H-2). HRMS (ESI-MS) for
+
2
C H N Br [M+H] : found, 227.9807; calcd, 227.9884.
5
6
7
Anal. (C H N Br) C, H, N.
6
6
5
(
1.21 mL, 19.4 mmol) was added. After stirring this
reaction mixture for 16 h at room temperature, it was
neutralised with acetic acid, and the DMF was evapo-
rated. The product was purified with column chroma-
tography(eth ly acetate) and cr sy tallised from EtOH.
4.3. General procedure for the synthesis of N3- and N9-
substituted N -cyclopentyladenine (8–14): method A
6
The appropriate alkyl halide (5 equiv), 8-bromo-N6-
cyclopentyladenine (5, 1 equiv), K CO (2 equiv), and
ꢀ
1
Yield 1.77 g (54%). Mp 135–137 C. H NMR
2
3
(
(
DMSO-d ) d 3.88 (s, 3H, CH ), 8.40 (s, 1H, H-2), 8.54
s, 1H, H-8).
DMF were mixed and stirred for 1 h. After filtration,
the filtrate was evaporated and purified bycolumn
chromatography(eth yl acetate/PE 40-60).
6
3
6
4
.2.2. N -Cyclopentyl-9-methyladenine (3) (N-0840). 6-
6
Chloro-9-methyladenine (2, 1.77 g, 10.5 mmol), cyclo-
pentylamine (1.04 mL, 10.5 mmol), triethylamine (2.92
mL, 21.0 mmol), and n-butanol (5 mL) were added to a
pressure tube and heated at 120 CC for 16 h. Puri-
fication was performed using column chromatography
4.3.1. 9-Allyl-8-bromo-N -cyclopentyladenine (8). This
was prepared according to method A, starting from allyl
ꢀ
1
bromide in a yield of 23%. Mp 62–63 C. H NMR
ꢀ
(CDCl ) d 1.52–2.17 (m, 8H, 4ꢃCH ), 4.60 (m, 1H,
3
2
CH), 4.82 (m, 2H, CH CH¼CH ), 5.16 (m, 2H,
2
2
6
(
1
5% MeOH/ethyl acetate). Yield 1.77 g (78%). Mp
ꢀ
CH¼CH ), 5.67 (d, 1H, N H), 5.95 (m, 1H, CH¼CH ),
2
2
ꢀ
3
1
02–106 C (lit. 109 C ). H NMR (CDCl ) d 1.41–
8.36 (s, 1H, H-2). Anal. (C H BrN ) C, H, N.
13 16 5
3
2
.10 (m, 8H, 4ꢃCH ), 3.49 (s, 3H, CH ), 4.64 (bs, 1H,
2
6
3
6
CH), 6.00 (d, 1H, N H), 7.72 (s, 1H, H-2), 8.41 (s, 1H,
H-8).
4.3.2. 8-Bromo-N -cyclopentyl-9-propyladenine (9) and
6
8-bromo-N -cyclopentyl-3-propyl adenine (13). These
were prepared according to method A, starting from
propyl iodide in a yield of 41 and 28%, respectively.
Separation was performed bycolumn chromatography
(ethyl acetate/PE 40-60).
6
6
4
.2.3. N -Cyclopentyladenine (4). N -Cyclopentylade-
nine (4) was prepared as described for 3 starting from 6-
1
chloropurine (1, 3 g, 19.4 mmol). Yield 3.43 g (87%). H
NMR (CDCl ) d 1.57–2.19 (m, 8H, 4ꢃCH ), 8.00 (s,
3
2
ꢀ
1
1
H, H-2), 8.45 (s, 1H, H-8).
9. Mp 94–95 C. H NMR (CDCl ) d 0.97 (t, 3H,
3
CH ), 1.50–2.16 (m, 10H, CH CH CH , 4ꢃCH ), 4.16 (t,
3
2
2
3
2
6
6
4.2.4. 8-Bromo-N -cyclopentyladenine (5). A freshlypre-
pared Na HPO solution (10% w/v, pH=7, 220 mL)
2H, CH CH CH ), 4.61 (m, 1H, CH), 5.72 (d, 1H, N H),
2 2 3
8.35 (s, 1H, H-2). Anal. (C H BrN ) C, H, N.
13 18 5
2
4
6
saturated with bromine was added to N -cyclopentyl-
adenine (4, 4.20 g, 20.7 mmol) dissolved in dioxane (220
mL). After stirring the solution overnight, 1.8 M
ꢀ
1
13. Mp 138–140 C. H NMR (CDCl ) d 0.95 (t, 3H,
3
CH ), 1.59-2.08 (m, 10H, CH CH CH , 4ꢃCH ), 4.28
3
2
2
3
2
NaHSO was added dropwise until the solution turned
3
(t, 2H, CH CH CH ), 4.55 (m, 1H, CH), 6.23 (m, 1H,
2 2 3
6
light yellow. After evaporation of the solvent, the solid
residue was dissolved in acetone, and the remaining
salts were filtered off. Purification was performed using
column chromatography(5% MeOH/CH Cl ). Yield
N H), 7.95 (s, 1H, H-2). Anal. (C H BrN ) C, H, N.
13 18 5
6
4.3.3. 9-Benzyl-8-bromo-N -cyclopentyladenine (10) and
6
3-benzyl-8-bromo-N -cyclopentyl adenine (14). These
2
2
ꢀ
1
3
1
1
.68 g (63%). Mp 210 C (dec). H NMR (CDCl ) d
were prepared according to method A, starting from
benzyl bromide in a yield of 35 and 23%, respectively.
Separation was performed bycolumn chromatography
(ethyl acetate/PE 40-60).
3
.14–1.64 (m, 8H, 4ꢃCH2), 2.96 (m, 1H, CH), 4.13 (d,
+
6
H, N H), 7.45 (m, 1H, N-9-H), 7.82 (s, 1H, H-2). MS:
m/e 282 (MH ). Anal. (C H BrN ) C, H, N.
1
0
12
5
6
ꢀ
1
4
8
9
.2.5. 8-Bromo-N -cyclopentyl-9-methyladenine (6) and
6
10. Mp 128–129 C. H NMR (CDCl ) d 1.45–2.19
3
-bromo-9-methyladenine (7). 8-Bromo-N -cyclopentyl-
-methyladenine (6) was prepared in a similar manner
as 5, starting from N -cyclopentyl-9-methyladenine (3,
.41 g, 6.49 mmol). Two organic products were formed
and extracted with ethyl acetate (100 mL, three times).
Separation and purification was performed bychroma-
tography(eth yl acetate).
(m, 8H, 4ꢃCH ), 4.63 (m, 1H, CH), 5.42 (s, 2H, CH ),
2
6
2
5.74 (d, 1H, N H), 7.35 (m, 5H, CH
), 8.41 (s, 1H,
phenyl
6
+
H-2). HRMS (ESI-MS) for C H N Br [M+H] :
17 19 5
372.0803; calcd,
.
(C H N Br 0.5MeOH) C, H, N.
1
found,
372.0823.
Anal.
1
7
18
5
ꢀ
1
14. Mp 104–106 C. H NMR (CDCl ) d 1.54–2.17
3
(
6
m, 8H, 4ꢃCH ), 4.54 (m, 1H, CH), 5.48 (s, 2H, CH ),
2
2
6
.50 (d, 1H, N H), 7.36 (m, 5H, CH
), 7.97 (s, 1H,
phenyl
. Yield 0.88 g (46%). Mp 105–108 C. 1H NMR
ꢀ
+
6
H-2). HRMS (ESI-MS) for C H N Br [M+H] :
17 19 5
(
CH ), 4.58 (m, 1H, CH), 5.58 (d, 1H, N H), 8.36 (s, 1H,
CDCl ) d 1.50–2.17 (m, 8H, 4ꢃCH ), 3.76 (s, 3H,
found, 372.0788; calcd, 372.0823.
3
2
6
3
+
6
H-2). HRMS (ESI-MS) for C H N Br [M+H] :
15
4.3.4. 9-Benzyl-N -cyclopentyladenine (11). This was
6
1
1
5
found,
(
296.0533;
.
calcd,
C H N Br 0.4DMF) C, H, N.
296.051.
Anal.
prepared according to method A, starting from N -
cyclopentyladenine (4, 0.2 g, 1 mmol) and benzyl bromide
1
1
14
5

146
R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
(
purified with hexane. Mp 122–123 C. H NMR (CDCl )
0.12 mL, 1 mmol) in a yield of 27%. The product was
ꢀ
tion was performed with column chromatography(ethyl
acetate) and crystallisation with CH Cl /hexane. Mp
1
3
2
2
ꢀ
1
d 1.52–2.28 (m, 8H, 4ꢃCH ), 4.65 (bs, 1H, CH), 5.35 (s,
93–97 C. H NMR (CDCl ) d 1.48 (t, 3H, CH CH ),
2
3
2
3
6
2
H, CH ), 5.99 (d, 1H, N H), 7.30 (m, 5H, CH
2
), 7.72
phenyl
1.52–2.20 (m, 8H, 4ꢃCH ), 3.54 (s, 3H, CH ), 4.55 (m,
2
3
6
(
C H N [M+H] : found, 294.1663; calcd, 294.1719.
s, 1H, H-2), 8.42 (s, 1H, H-8). HRMS (ESI-MS) for
+
3H, CH CH , CH), 5.75 (d, 1H, N H), 8.29 (s, 1H, H-
2 3
2). HRMS (ESI-MS) for C H N O [M+H] : found,
13 20 5
.
62.1731; calcd, 262.1668. Anal. (C H N O 1.3H O)
13 19 5 2
+
1
7
20
5
2
6
4
.3.5. 8-Bromo-N -cyclopentyl-3-methyladenine (12).
C, H, N.
This was prepared according to method A, starting
from methyl iodide. The mixture was separated by flash
chromatography. One compound was identical to 8-
6
4.4.5. N -Cyclopentyl-8-isopropoxy-9-methyladenine (19).
This was prepared according to method B, starting
6
6
bromo-N -cyclopentyl-9-methyladenine (6). Compound
ꢀ
from 8-bromo-N -cyclopentyl-9-methyladenine (6) and
2-propanol (10 mL) in a yield of 84% (0.26 g). Pur-
ification was performed with column chromatography
1
1
2 was isolated in a yield of 23%. Mp 175–177 C. H
NMR (CDCl ) d 1.51–2.15 (m, 8H, 4ꢃCH ), 3.98 (s,
3
2
6
ꢀ
1
3
1
H, CH ), 4.53 (m, 1H, CH), 6.31 (d, 1H, N H), 7.97 (s,
3
(MeOH/CH Cl ). Mp 75–83 C. H NMR (CDCl ) d
2
2
3
H, H-2). Anal. (C H BrN ) C, H, N.
1
1.45 (d, 6H, CH(CH ) ), 1.50–2.22 (m, 8H, 4ꢃCH ),
1
14
5
3 2
2
3.52 (s, 3H, CH ), 4.64 (m, 1H, CH), 5.19–5.38 (m, 1H,
CH(CH ) ), 5.88 (d, 1H, N H), 8.26 (s, 1H, H-2).
3
6
.4. General procedure for the synthesis of 9-alkyl-N6-
cyclopentyl-8-oxyadenines (15–20): method B
3 2
4
+
HRMS (ESI-MS) for C H N O [M+H] : found,
1
76.1827; calcd, 276.1824.
4
22
5
2
6
The appropriate 8-bromo-N -cyclopentyl-9-alkylade-
nine (6, 1 equiv), the corresponding alcohol, and KOtBu
6
4.4.6. N -Cyclopentyl-9-methyl-8-propoxy-adenine (20).
This was prepared according to method B, starting
(
1 equiv) was mixed and refluxed for 2–64 h. The reaction
6
mixture was then neutralised with acetic acid, and the
products were purified with column chromatography.
from 8-bromo-N -cyclopentyl-9-methyladenine (6) and
n-propanol (10 mL) in a yield of 99% (0.31 g). Pur-
ification was performed with column chromatography
1
6
ꢀ
CH CH CH ),
4.4.1. 9-Allyl-N -cyclopentyl-8-methoxyadenine (15).
This was prepared according to method B, starting
(ethyl acetate). Mp 77–79 C. H NMR (CDCl ) d 1.05
3
(t,
3H,
1.45–2.24
(m,
10H,
2
2
3
6
from 9-allyl-8-bromo-N -cyclopentyladenine (8) and
MeOH in a yield of 91%. Purification was performed
with column chromatography(eth yl acetate/PE 40-60).
CH CH CH , 4ꢃꢃCH ), 3.55 (s, 3H, CH ), 4.43 (t, 2H,
2
2
3
2
3
6
CH CH CH ), 4.61 (m, 1H, CH), 5.63 (d, 1H, N H),
2 2 3
8.28 (s, 1H, H-2). HRMS (ESI-MS) for C H N O
5
1
4
22
ꢀ
1
+
Mp 50–52 C. H NMR (CDCl ) d 1.50–2.18 (m, 8H,
[M+H] : found, 276.1812; calcd, 276.1824.
3
4
ꢃCH ), 4.14 (s, 3H, OCH3), 4.60 (m, 3H,
2
6
CH CH¼CH , CH), 5.10 (m, 2H, CH¼CH ), 5.40 (d,
2 2 2
4.4.7. N-Cyclopentyl-8-(ethylthio)-9-methyladenine (21).
6
6
1
Anal. (C H N O) C, H, N.
H, N H), 5.92 (m, 1H, CH¼CH ), 8.30 (s, 1H, H-2).
A mixture of 8-bromo-N -cyclopentyl-9-methyladenine
(6, 0.34 g, 1.13 mmol), ethanethiol (0.17 mL, 2.26
2
1
4
19
5
mmol), KOtBu (0.25 g, 2.26 mmol), and EtOH (7 mL)
ꢀ
6
4
.4.2. N -Cyclopentyl-8-methoxy-9-propyladenine (16).
This was prepared according to method B, starting
was added in a pressure tube and stirred at 50 CC
overnight. Purification was performed with column
chromatography(eth yl acetate) and cr ys tallisation with
6
from 8-bromo-N -cyclopentyl-9-propyladenine (9) and
MeOH in a yield of 88%. Purification was performed
with column chromatography(eth yl acetate/PE 40-60).
Mp 58–60 C. H NMR (CDCl ) d 0.97 (t, 3H, CH ),
1.50–2.14 (m, 10H, CH CH CH , 4ꢃCH ), 3.95 (t, 2H,
CH CH CH ), 4.14 (s, 3H, OCH ), 4.60 (m, 1H, CH),
ꢀ
1
hexane. Yield 0.29 g (93%). Mp 94–96 C C. H NMR
(CDCl ) d 1.44 (t, 3H, CH CH ), 1.49–2.22 (m, 8H,
3
2
3
ꢀ
1
4ꢃꢃCH ), 3.30 (q, 2H, CH CH ), 3.67 (s, 3H, CH ), 4.62
3
3
2
2
3
3
6
(m, 1H, CH), 5.59 (d, 1H, N H), 8.32 (s, 1H, H-2). HRMS
(ESI-MS) for C H N S [M+H] : found, 278.1443;
.
calcd, 278.1439. Anal. (C H N S 0.9H O) C, H, N.
13 19 5 2
2
2
3
2
+
2
2
3
3
13 20
5
6
5
(
.39 (d, 1H, N H), 8.31 (s, 1H, H-2). Anal.
C H N O) C, H, N.
1
4
21
5
6
4
.4.8. N -Cyclopentyl-9-methyl-8-oxo-adenine (22). 8-
6
6
4
.4.3. N -Cyclopentyl-8-methoxy-9-methyladenine (17).
This was prepared according to method B, starting
Bromo-N -cyclopentyl-9-methyladenine (6, 0.22 g, 0.73
mmol) was refluxed for 3 h in a 1 N NaOH solution
(9.42 mL). The reaction mixture was neutralised with a
10% HCl solution, and chilled on ice. Purification was
performed with column chromatography(eth yl acetate).
6
from 8-bromo-N -cyclopentyl-9-methyladenine (6) and
MeOH in a yield of 19% (16 mg). Purification was per-
formed with column chromatography(MeOH/ethly
ꢀ
1
ꢀ
1
acetate). Mp 80–90 C. H NMR (CDCl ) d 1.59–2.18
Yield 0.12 g (70%). Mp 188–190 C C. H NMR
3
(
m, 8H, 4ꢃCH ), 3.55 (s, 3H, CH ), 4.15 (s, 3H, OCH ),
(CDCl ) d 1.48–2.14 (m, 8H, 4ꢃꢃCH ), 3.37 (s, 3H,
2
3
3
3
2
6
6
4
HRMS (ESI-MS) for C H N O [M+H] : found,
2
.42 (m, 1H, CH), 5.25 (d, 1H, N H), 8.30 (s, 1H, H-2).
CH ), 4.47 (m, 1H, CH), 5.81 (d, 1H, N H), 8.30 (s, 1H,
3
+
H-2), 10.56 (s, 1H, N-7-H). HRMS (ESI-MS) for
+
1
2
18
5
84.1586; calcd, 248.1511.
C H N O [M+H] : found, 234.1349; calcd, 234.1355.
1
1
16
5
6
6
4
.4.4. N -Cyclopentyl-8-ethoxy-9-methyladenine (18).
This was prepared according to method B, starting
4.4.9. -Cyclopentyl-8-(N-methylamino)-9-methyladenine
6
(23, LUF 5666). 8-Bromo-N -cyclopentyladenine ((5,
501 mg, 1.69 mmol) was dissolved in a 40% w/v
aqueous methylamine solution (150 mL, excess) and
6
from 8-bromo-N -cyclopentyl-9-methyladenine (6) and
abs EtOH (10 mL) in a yield of 67% (0.20 g). Purifica-

R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
147
stirred for 16 h at room temperature. The reaction mix-
ture was then concentrated in vacuo and the product
was crystallised from water. Yield 423 mg (99%). Mp
(CH , propyl), 20.56 (CH CH N, propyl), 23.72 (2C,
3 2 2
CH CH CH , cyclopentyl), 29.66 (CH , N-9), 33.45
2
2
2
3
(2C, CHCH CH , cyclopentyl), 38.85 (CH , methyla-
2
2
3
ꢀ
1
1
31–135 C C. H NMR (CDCl ) d 1.57–2.14 (m, 8H,
mine), 52.32 (CH, cyclopentyl), 55.53 (CH N, propyl),
2
3
8
4
CH ), 4.39 (q, 1H, N H), 4.62 (m, 1H, CH), 5.79 (d, 1H,
ꢃꢃCCH ), 3.17 (d, 3H, N -CH ), 3.53 (s, 3H, N-9-
116.94 (C-8), 150.99 (C-2), 150.18, 152.17, 155.72 (C-6,
+
C-5, C-4). HRMS (ESI-MS) for C H N [M+H] :
2
3
8
3
15 25
6
6
13
N H), 8.25 (s, 1H, H-2). C NMR (CDCl ): d 23.69
found, 289.2156; calcd, 289.2141.
3
(
3
2C, CH CH CH , cyclopentyl), 29.63 (CH , N-9),
3.37 (2C, CHCH CH , cyclopentyl), 41.52 (CH ,
2 2 2 3
6
4.4.13. N -Cyclopentyl-8-(N-butyl-N-methylamino)-9-
methyladenine (27). This was prepared as 24, starting
2
2
3
methylamine), 52.32 (CH, cyclopentyl), 116.91 (C-8),
51.02 (C-2), 150.36, 152.24, 155.94 (C-6, C-5, C-4).
+
6
1
HRMS (ESI-MS) for C H N [M+H] : found,
with N -cyclopentyl-8-(N-methylamino)-9-methylade-
nine (23, 39 mg, 0.16 mmol) in DMF/NaH and butyl
iodide (49 mg, 0.29 mmol). Yield 12 mg (26%, oil). H
NMR (CDCl ) d 0.93 (t, 3H, CH CH CH CH ), 1.19–
1
2
19
6
1
2
47.1572; calcd, 247.1671.
3
2
2
2
3
6
4
.4.10. N -Cyclopentyl-8-(N,N-dimethylamino)-9-methyl-
6
2.20 (m, 12H, CH CH CH CH , butyl, 4ꢃCH ), 2.94
2
2
2
3
2
8
8
3 2
adenine (24). N -Cyclopentyl-8-(N-methylamino)-9-
methyladenine (23, 76 mg, 0.31 mmol) was dissolved in
1
0
was added (53 mg, 0.34 mmol) an the reaction mixture
was stirred for 16 h at room temperature. Afterwards,
(s, 3H, N -CH ), 3.21 (m, 2H, N -CH ), 3.61 (s, 3H, N-
9-CH ), 4.59 (m, 1H, CH), 5.44 (d, 1H, N H), 8.29 (s,
6
3
1
3
.5 mL DMF and NaH (60% in mineral oil, 13.5 mg,
.31 mmol), and stirred for 15 min. Then methyl iodide
1H, H-2). C NMR (CDCl ): d 13.82 (CH , butyl),
3
3
20.19 (CH CH CH N, butyl), 23.70 (2C, CH CH CH ,
cyclopentyl), 29.38 (CH CH N, butyl), 29.68 (CH , N-
2 2 3
9), 33.43 (2C, CHCH CH , cyclopentyl), 38.85 (CH ,
2
2
2
2
2
2
2
2
3
1
organic product was extracted with CH Cl (two times,
0 mL of water was added to the mixture and the
methylamine), 50.67 (CH N, butyl), 52.31 (CH, cyclo-
2
pentyl), 116.97 (C-8), 151.02 (C-2), 149.99, 152.25,
154.67 (C-6, C-5, C-4). HRMS (ESI-MS) for C H N
6
2
2
4
matography(MeOH/CH Cl ) and the product was col-
mL). Purification was performed with column chro-
1
6
27
+
[M+H] : found, 303.2245; calcd, 303.2297.
2
2
1
lected as oil. Yield 34 mg (43%). H NMR (CDCl ) d
3
8
1
(
.50–2.17 (m, 8H, 4ꢃCH ), 2.93 (s, 3H, N -CH ), 3.60
2
3
4
.5. General procedure for the synthesis of 8-aminosub-
6
6
s, 3H, N-9-CH ), 4.64 (m, 1H, CH), 5.49 (d, 1H, N H),
3
stituted N -cyclopentyl-9-methyl adenines (28–33, 35–
7): method C
1
3
8.26 (s, 1H, H-2). C NMR (CDCl ): d 23.69 (2C,
CH CH CH , cyclopentyl), 29.66 (CH , N-9), 33.39
3
3
2
2
2
3
(
2C, CHCH CH , cyclopentyl), 41.52 (2C, CH , dime-
2
The appropriate amine was added in excess to a solu-
6
2
3
thylamine), 52.29 (CH, cyclopentyl), 116.79 (C-8),
51.09 (C-2), 150.24, 152.24, 155.88 (C-6, C-5, C-4).
+
tion of 8-bromo-N -cyclopentyl-9-methyladenine (6) in
dioxane. The reaction mixture was stirred in a pressure
1
HRMS (ESI-MS) for C H N [M+H] : found,
ꢀ
ducts were purified with column chromatography.
tube at 80–120 C for 16–120 h, and the resulting pro-
1
3
21
6
2
61.1856; calcd, 261.1828.
6
6
4
methyladenine (25, LUF 5668). This was prepared as
.4.11.
N -Cyclopentyl-8-(N-methyl-N-ethylamino)-9-
4.5.1. N -Cyclopentyl-8-(N-ethylamino)-9-methyladenine
(28). This was prepared according to method C, start-
6
24, starting with N -cyclopentyl-8-(N-methylamino)-9-
methyladenine (23, 61 mg, 0.25 mmol) in DMF/NaH
ing from a 70% w/v ethylamine solution and stirred at
80 C for 48 h. Purification was performed with column
ꢀ
chromatography(MeOH/eth ly acetate). Yield 0.29 g
and ethyl iodide (10.3 mg, 0.25 mmol). Yield 24 mg
1
ꢀ
1
(
36%, oil). H NMR (CDCl ) d 1.11–1.48 (t, 3H,
(99%). Mp 77–81 C. H NMR (CDCl ) d 1.34 (t, 3H,
3
3
8
CH CH ), 1.52–2.16 (m, 8H, 4ꢃCH ), 2.92 (s, 3H, N -
CH CH ), 1.50–2.17 (m, 8H, 4ꢃCH ), 3.50 (s, 3H, CH ),
2
3
2
2
3
2
3
8
CH ), 3.41–3.22 (q, 2H, CH CH ), 3.59 (s, 3H, N-9-
3
3.56 (m, 2H, CH CH ), 4.05 (bs, 1H, N H), 4.59 (m, 1H,
2 3
2
3
6
6
CH ), 4.65 (m, 1H, CH), 5.50 (d, 1H, N H), 8.27 (s, 1H,
3
H-2). C NMR (CDCl ): d 12.50 (CH , ethyl), 23.69
2C, CH CH CH , cyclopentyl), 29.57 (CH , N-9),
CH), 5.37 (d, 1H, N H), 8.23 (s, 1H, H-2). HRMS (ESI-
+
1
3
MS) for C H N [M+H] : found, 261.1906; calcd,
3
3
13 21
6
.
261.1828. Anal. (C H N 0.4SiO ) C, H, N.
13 20 6 2
(
2
2
2
3
3
methylamine), 48.34 (CH N, ethyl), 52.38 (CH, cyclo-
3.42 (2C, CHCH CH , cyclopentyl), 38.15 (CH ,
2 2 3
6
4.5.2. N -Cyclopentyl-8-(N-cyclopentylamino)-9-methy-
ladenine (29). This was prepared according to method
C, starting from cyclopentylamine solution and stirred
2
pentyl), 116.91 (C-8), 151.06 (C-2), 150.18, 152.27,
1
55.51 (C-6, C-5, C-4). HRMS (ESI-MS) for C H N
14 23 6
+
ꢀ
[
M+H] : found, 275.1961; calcd, 275.1984.
at 120 C for 120 h. Purification was performed with
column chromatography(MeOH/ethly acetate). Yield
0.17 g (53%). Mp 62–65 C. H NMR (CDCl ) d 1.46–2.19
6
ꢀ
1
4.4.12. N -Cyclopentyl-8-(N-methyl-N-propylamino)-9-
methyladenine (26, LUF 5669). This was prepared as
3
(m, 16H, 8ꢃCH ), 3.48 (s, 3H, N-9-CH ), 4.18–4.32 (m,
2
3
6
8
8
2
methyladenine (23, 62 mg, 0.25 mmol) in DMF/NaH
4, starting with N -cyclopentyl-8-(N-methylamino)-9-
1H, N CH), 4.52 (d, 1H, N H), 4.46–4.64 (m, 1H, CH),
5.78 (d, 1H, N H), 8.21 (s, 1H, H-2). HRMS (ESI-MS) for
6
+
and ethyl iodide (49.3 mg, 0.29 mmol). Yield 26 mg
1
C H N [M+H] : found, 301.2152; calcd, 301.2141.
25
16
6
(
CH CH CH ), 1.40–2.19 (m, 10H, CH CH CH , 8H,
36%, oil).
H NMR (CDCl3) d 0.93 (t, 3H,
6
4.5.3. N -cyclopentyl-8-(N-phenylamino)-9-methylade-
nine (30). This was prepared according to method C,
2
2
3
2
2
3
8
8
4
3
ꢃCH ), 2.93 (s, 3H, N -CH ), 3.17 (t, 2H, N -CH ),
2
3
2
.60 (s, 3H, N-9-CH ), 4.57 (m, 1H, CH), 5.43 (d, 1H,
3
starting from a 70% w/v aniline solution and stirred at
ꢀ
6
13
N H), 8.27 (s, 1H, H-2). C NMR (CDCl ): d 11.25
100 C for 32 h. Purification was performed with col-
3

148
R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
umn chromatography(eth yl acetate) and cr ys tallisation
with CH Cl /hexane. Yield 0.12 g (53%). Mp 191–
Purification was performed with column chromato-
1
graphy(acetone/hexane). Yield 23 mg (30%, oil).
H
2
2
ꢀ
1
1
3
95 C. H NMR (CDCl ) d 1.38–2.28 (m, 8H, 4ꢃCH ),
NMR (CDCl ) d 0.90 (t, 3H, CH CH CH CH ), 1.15 (t,
3
2
3
2
2
2
3
.59 (s, 3H, N-9-CH ), 4.60 (m, 1H, CH), 5.45 (d, 1H,
3
3H, CH CH ), 1.21–2.18 (m, 12H, CH CH CH CH ,
2 3 2 2 2 3
6
8
8
N H), 6.80 (bs, 1H, N H), 7.02 (t, 1H, CHphenyl), 7.31 (t,
H, CHphenyl), 7.50 (d, 2H, CHphenyl), 8.31 (s, 1H, H-2).
+
butyl, 4ꢃCH ), 3.20–3.35 (m, 4H, CH CH , N -CH ),
2
2
3
2
2
HRMS (ESI-MS) for C H N [M+H] : found,
3.59 (s, 3H, N-9-CH ), 4.58 (m, 1H, CH), 5.65 (d, 1H,
3
N H), 8.28 (s, 1H, H-2). C NMR (CDCl ): d 12.87
6
13
1
8
21
6
3
3
09.1795; calcd, 309.1828.
(CH , ethyl), 13.87 (CH , butyl), 20.18 (CH CH CH N,
butyl), 23.73 (2C, CH CH CH , cyclopentyl), 29.40
(CH CH N, butyl), 29.82 (CH , N-9), 33.46 (2C,
2 2 3
3
3
2
2
2
2
2
2
6
4
.5.4. N -Cyclopentyl-8-(N-methylisopropylamino)-9-me-
thyladenine (31, LUF 5608). This was prepared accord-
ing to method C, starting from a 70% w/v aqueous N-
CHCH CH , cyclopentyl), 46.19 (CH , methylamine),
2 2 3
50.62 (CH N, butyl), 52.41 (CH, cyclopentyl), 117.08
2
ꢀ
8 h. Purification was performed with column chroma-
methylisopropylamine solution and stirred at 80 C for
4
tography (ethyl acetate) and crystallisation with hexane.
(C-8), 151.01 (C-2), 149.98, 152.29, 154.65 (C-6, C-5, C-
+
4). HRMS (ESI-MS) for C H N [M+H] : found,
6
1
7
29
317.2451; calcd, 317.2454.
ꢀ
1
Yield 0.12 g (37%). Mp 59–62 C. H NMR (CDCl ) d
3
6
1
2
1
8
.22 (d, 6H, CH(CH ) ), 1.49–2.19 (m, 8H, 4ꢃCH ),
4.5.8. N -Cyclopentyl-8-(N-pyrrolidino)-9-methyladenine
(35). This was prepared according to method C, start-
ing from a 70% w/v pyrrolidine solution and stirred at
3
2
2
8
.81 (m, 3H, N CH ), 3.60 (s, 3H, N-9-CH ), 3.73 (m,
3
3
6
H, CH(CH ) ), 4.57 (m, 1H, CH), 5.89 (m, 1H, N H),
3
2
ꢀ
.29 (s, 1H, H-2). HRMS (ESI-MS) for C H N
6
100 C for 16 h. Purification was performed with col-
umn chromatography(ethly acetate). Yield 0.13 g
1
5
25
+
[
M+H] : found, 289.2114; calcd, 289.2141.
ꢀ
1
(
62%). Mp 95–96 C. H NMR (CDCl ) d 1.35–2.23 (m,
3
8 8
6
4
.5.5.
N -Cyclopentyl-8-(N-methyl-N-benzylamino)-9-
12H, 4ꢃCH , N -CH CH ), 3.58 (m, 4H, N -CH ), 3.68
2
2
2
2
6
methyladenine (32). This was prepared according to
method C, starting from a 70% w/v aqueous N-methyl-
(s, 3H, N-9-CH ), 4.57 (m, 1H, CH), 5.75 (d, 1H, N H),
3
8.25 (s, 1H, H-2). HRMS (ESI-MS) for C H N
6
1
5
23
ꢀ
subsequentlyat 100 C for 48 h. Purification was per-
+
benzylamine solution and stirred at 80 C for 66 h, and
ꢀ
formed with column chromatography(MeOH/CH Cl ).
[M+H] : found, 287.1982; calcd, 287.1984. Anal.
.
(C H N 0.3SiO ) C, H, N.
15 22 6 2
2
2
1
6
Yield 52 mg (46%, oil). H NMR (CDCl ): d 1.55-2.17
4.5.9. N -Cyclopentyl-8-(N-piperidino)-9-methyladenine
(36). This was prepared according to method C, start-
ing from a 70% w/v aqueous piperidine solution and
3
8
(
m, 8H, 4ꢃCH ), 2.88 (m, 3H, N CH ), 3.67 (s, 3H, N-
2
3
9
1
2
-CH ), 4.43 (s, 2H, CH ), 4.64 (m, 1H, CH), 5.60 (d,
3 2
6
ꢀ
H, N H), 7.35–7.39 (m, 5H, CHphenyl), 8.30 (s, 1H, H-
). C NMR (CDCl ): d 23.67 (2C, CH CH CH ,
stirred at 100 C for 16 h. Purification was performed
13
with column chromatography(eth yl acetate) and cr ys -
tallisation with hexane. Yield 0.21 g (95%). Mp 137–
3
2
2
2
cyclopentyl), 29.60 (CH , N-9), 33.39 (2C, CHCH CH ,
2
3
2
ꢀ
1
cyclopentyl), 38.73 (CH , methylamine), 52.44 (CH,
3
cyclopentyl), 57.56 (CH , benzyl), 116.73 (C-8), 127-136
2
139 C. H NMR (CDCl ) d 1.37–2.26 (m, 14H, 4ꢃCH ,
3
2
8
2 2 3
3ꢃCH ), 3.19 (m, 4H, N -CH ), 3.60 (s, 3H, N-9-CH ),
6
(
C-5, C-4). HRMS (ESI-MS) for C H N [M+H] :
6C, phenyl), 151.02 (C-2), 150.17, 152.24, 155.42 (C-6,
+
4.59 (m, 1H, CH), 5.61 (d, 1H, N H), 8.30 (s, 1H, H-2).
Anal. (C H N ) C, H, N.
1
9
25
6
16 24
6
found, 337.2113; calcd, 337.2141.
6
4
.5.10.
N -Cyclopentyl-8-morpholino-9-methyladenine
(37). This was prepared according to method C, start-
6
4
.5.6. N -Cyclopentyl-8-(N,N-diethylamino)-9-methyla-
6
denine (33, LUF 5674). 8-Bromo-N -cyclopentyl-9-
methyladenine (6, 80 mg, 0.27 mmol) and diethylamine
ing from a 70% w/v aqueous morpholine solution and
stirred at 120 C for 66 h. Purification was performed
ꢀ
with column chromatography(MeOH/CH Cl ). Yield
(
1
1.0 mL, 10 mmol) were dissolved in 2 mL dioxane and
mL water. The mixture was transferred into a pressure
2
2
ꢀ
1
83 mg (95%). Mp 118–122 C. H NMR (CDCl ) d 1.4-
3
tube, and heated in a boiling water bath for 40 h. Pur-
ification was performed with column chromatography
2.2 (m, 8H, 4ꢃCH ), 3.25–3.30 (t, 4H, CH OCH ), 3.63
2
2
2
(s, 3H, N-9-CH ), 3.92–3.87 (t, 4H, CH NCH ), 4.59
2
3
2
6
(
acetone/hexane) and the product was collected as oil.
1
(m, 1H, CH), 5.45 (d, 1H, N H), 8.32 (s, 1H, H-2).
+
15 23
Yield 26 mg (35%). H NMR (CDCl ) d 1.10–1.23 (t,
HRMS (ESI-MS) for C H N O [M+H] : found,
6
3
3
H, CH CH ), 1.52–2.11 (m, 8H, 4ꢃCH ), 3.21–3.32 (q,
303.1916; calcd, 303.1933.
2
3
2
2H, CH CH ), 3.58 (s, 3H, N-9-CH ), 4.61 (m, 1H,
CH), 5.73 (d, 1H, N H), 8.27 (s, 1H, H-2). C NMR
(CDCl ): d 12.89 (CH , ethylamine), 23.72 (2C,
CH CH CH , cyclopentyl), 29.33 (CH , N-9), 33.42 (2C,
2
3
3
6
13
4.6. Radioligand displacement experiments
3
3
The adenosine A receptor binding assays were carried
1
2
2
2
3
CHCH CH , cyclopentyl), 45.43 (CH N, ethyl), 52.38
2
out on membranes of CHO cells that have a high ade-
nosine A receptor density( ꢂ3 pmol/mg of protein,
2
2
(CH, cyclopentyl), 117.09 (C-8), 151.09 (C-2), 149.99,
1
++
1
1
C H N [M+H] : found, 289.2172; calcd, 289.2141.
52.33, 154.48 (C-6, C-5, C-4). HRMS (ESI-MS) for
+
CHO-A
protein and increasing concentrations of the compound,
). Membrane aliquots, containing 6 mg of
1
5
25
6
were incubated in 400 mL of 50 mM Tris–HCl, pH 7.4 at
6
ꢀ
[ H]DPCPX. Non-specific binding was measured in the
4.5.7. N -Cyclopentyl-8-(N-ethyl-N-butylamino)-9-me-
thyladenine (34). This was prepared as 33, starting with
8
25 C for 60 min in the presence of ꢂ1.6 nM
3
6
-bromo-N -cyclopentyl-9-methyladenine (6, 72 mg,
.24 mmol) and N-ethylbutylamine (1 mL, 14 mmol).
presence of 10 mM CPA. Incubations were stopped by
dilution with the above-mentioned buffer, and bound
0

R. A. F. de Ligt et al. / Bioorg. Med. Chem. 12 (2004) 139–149
149
radioligand was separated byrapid filtration through
Whatman GF/B filters using a Brandel harvester. Filters
were subsequentlywashed three times with the same ice-
cold buffer. Bound radioactivitywas measured byscin-
tillation spectrometryafter the addition of 3.5 mL of
¨
(Universityof Wu rzburg, Germany) for the HEK293
cells stablytransfected with the human adenosine A 3
receptor. Cees Erkelens is acknowedged for his expert
help in the NOESY experiments. The authors (R.d.L.
and A.IJ.) acknowledge financial support from the EU
BIOMED2 programme ‘Inverse agonism. Implications
for drug design’ (#BMH4-CT97-2152).
Packard Emulsifier Safe. Adenosine A binding experi-
3
ments were done with membranes of HEK293 cells
expressing the human adenosine A receptor (HEK293-
3
A ). Briefly, membrane aliquots (40 mg of protein) were
3
ꢀ
References and notes
incubated at 37 C for 60 min with ꢂ0.1 nM of
1
25
125
6
0
[
I]AB-MECA ([ I]N -(4-amino-3-iodobenzyl)-5 -N-
1
. Shryock, J. C.; Ozeck, M. J.; Belardinelli, L. Mol. Phar-
macol. 1998, 53, 886.
2. Moos, W. H.; Szotek, D. S.; Bruns, R. F. J. Med. Chem.
1985, 28, 1383.
3. Thompson, R. D.; Secunda, S.; Daly, J. W.; Olsson, R. A.
J. Med. Chem. 1991, 34, 2877.
. Knutsen, L. J.; Lau, J.; Petersen, H.; Thomsen, C.; Weis,
J. U.; Shalmi, M.; Judge, M. E.; Hansen, A. J.; Shear-
down, M. J. J. Med. Chem. 1999, 42, 3463.
ethylcarboxamidoadenosine), and a fixed concentration
of all compounds (10 mM) or range of concentrations
for selected compounds (5, 8, 9, 13, 14, 19, 20, and 32),
in a final volume of 200 mL of 50 mM Tris/10 mM
MgCl /1 mM EDTA (ethylenediaminetetra-acetic acid)/
2
4
5
6
0
lammonio)-1-propanesulfonate] buffer. Nonspecific
.01%
CHAPS
[3-([3-cholamidopropyl]-dimethy-
binding was measured in the presence of 100 mM R-PIA
. Matsuda, A.; Shinozaki, M.; Yamaguchi, T.; Homma, H.;
Nomoto, R.; Miyasaka, T.; Watanabe, Y.; Abiru, T. J.
Med. Chem. 1992, 35, 241.
6
[
(R)-N -phenylisopropyladenosine]. Binding reactions
were terminated bydilution with ice-cold buffer. Sam-
ples were then filtered through Whatman GF/B glass-
fiber filters using a Brandel cell harvester, and filters
were washed three times. Bound radioactivitywas mea-
sured in a Beckman 5500B g-counter.
. Van Tilburg, E. W.; Van der Klein, P. A.; Von Frijtag
Drabbe Ku
enzen, A.; IJzerman, A. P. J. Med. Chem. 2001, 44, 2966.
7. Van Tilburg, E. W.; Von Frijtag Drabbe Kunzel, J.; De
¨
nzel, J.; De Groote, M.; Stannek, C.; Lor-
¨
Groote, M.; IJzerman, A. P. J. Med. Chem. 2002, 45, 420.
. Volpini, R.; Costanzi, S.; Lambertucci, C.; Vittori, S.;
Cristalli, G. Curr. Pharm. Des. 2002, 8, 99.
. Cristalli, G.; Camaioni, E.; Costanzi, S.; Vittori, S.;
Volpini, R.; Klotz, K.-N. Drug Dev. Res. 1998, 45, 176.
0. Martin, P. L.; Wysocki, R. J., Jr.; Barrett, R. J.; May,
J. M.; Linden, J. J. Pharmacol. Exp. Ther. 1996, 276, 490.
1. Volpini, R; Costanzi, S.; Lambertucci, C.; Vittori, S.;
Lorenzen, A.; Klotz, K.-N.; Cristalli, G. Bioorg. Med.
Chem. Lett. 2001, 11, 1931.
8
9
3
5
4
.7. [ S]GTPꢀS binding
3
5
The modulation of [ S]GTPgS binding was determined
according to the method of Lorenzen et al. with minor
2
3
1
ꢀ
modifications. Incubations were performed at 25 C for
9
0 min with 4–5 mg of membrane protein. The GDP and
1
NaCl concentrations were 3 mM and 100 mM, respec-
3
5
tively. Basal [ S]GTPgS binding was set to 100%.
Under these conditions, CPA behaved as a full agonist
12. Young, R. C.; Jones, M.; Milliner, K. J.; Rana, K. K.;
Ward, J. G. J. Med. Chem. 1990, 33, 2073.
3. Fink, C. A.; Spada, A. P. Nucleosides Nucleotides 1992,
1, 1077.
3
5
stimulating basal [ S]GTPgS binding to 317%.
1
1
4
.8. Data analysis
14. Ikehara, M.; Uesugi, S.; Kaneko, M. J. Chem. Soc. Chem.
Commun. 1967, 17.
K values were calculated using a non-linear regression
i
1
5. Jacobson, K. A.; Siddiqi, S. M.; Olah, M. E.; Ji, X. D.;
Melman, N.; Bellamkonda, K.; Meshulam, Y.; Stiles,
G. L.; Kim, H. O. J. Med. Chem. 1995, 38, 1720.
curve fitting program (GraphPad Prism, GraphPad
Software Inc., San Diego, CA, USA). The K value of
d
3
++
[
nM. The intrinsic activities of the compounds are
H]DPCPX at CHO-A1
membranes used was 1.6
16. Reitz, A. B.; Graden, D. W.; Jordan, A. D., Jr.; Maryan-
off, B. E. J. Org. Chem. 1990, 55, 5761.
17. Rasmussen, M.; Hope, J. M. Aust. J. Chem. 1982, 35, 525.
3
5
reported as the percentage of basal [ S]GTPgS binding
set at 100%) remaining in the presence of concentra-
18. Guida, W. C.; Mathre, D. J. J. Org. Chem. 1980, 45, 3172.
19. Cho, B. P.; Evans, F. E. Nucleic Acids Res. 1991, 19, 1041.
20. Chattopadyaya, J. B.; Reese, C. B. Synthesis 1977, 725.
21. Roelen, H.; Veldman, N.; Spek, A. L.; Von Frijtag
(
tions of 10ꢃK determined from receptor binding
i
experiments.
¨
Drabbe Kunzel, J.; Mathoˆ t, R. A.; IJzerman, A. P. J.
Med. Chem. 1996, 39, 1463.
2. Komoda, Y.; Shimizu, M.; Kaneko, S.; Yamamoto, M.;
Acknowledgements
2
Ishikawa, M. Chem. Pharm. Bull. 1982, 30, 502.
23. Lorenzen, A.; Fuss, M.; Vogt, H.; Schwabe, U. Mol.
Pharmacol. 1993, 44, 115.
We thank Steve Hill (Nottingham University, UK) for
+
providing the CHO-A1 cells and Karl-Norbert Klotz
+