A series of ligands displaying a remarkable agonistic-antagonistic profile at the adenosine A1 receptor

Article abstract of DOI:10.1021/jm049597+

Adenosine receptor agonists are usually variations of the natural ligand, adenosine. The ribose moiety in the ligand has previously been shown to be of great importance for the agonistic effects of the compound. In this paper, we present a series of nonadenosine ligands selective for the adenosine A ₁ receptor with an extraordinary pharmacological profile. 2-Amino-4-benzo-[1,3]dioxol-5-yl-6-(2-hydroxyethylsulfanyl)pyridine-3, 5-dicarbonitrile (70, LUF 5853) shows full agonistic behavior comparable with the reference compound CPA, while also displaying comparable receptor binding affinity (K_i = 11 nM). In contrast, compound 58 (2-amino-4-(3- trifluoromethylphenyl)-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarbonitrile, LUF 5948) has a binding affinity of 14 nM and acts as an inverse agonist. Also present within this same series are compounds that show neutral antagonism of the adenosine A₁ receptor, for example compound 65 (2-amino-4-(4-difluoromethoxyphenyl)-6-(2-hydroxyethylsulfanyl)pyridine-3, 5-dicarbonitrile, LUF 5826).

Full text of DOI:10.1021/jm049597+

J. Med. Chem. 2005, 48, 2045-2053
2045
A Series of Ligands Displaying a Remarkable Agonistic-Antagonistic Profile at
the Adenosine A1 Receptor
Lisa C. W. Chang, Jacobien K. von Frijtag Drabbe Ku¨nzel, Thea Mulder-Krieger, Ronald F. Spanjersberg,
Sophie F. Roerink, Gijs van den Hout, Margot W. Beukers, Johannes Brussee, and Adriaan P. IJzerman*
Leiden/Amsterdam Center for Drug Research, Division of Medicinal Chemistry, P.O. Box 9502,
2300 RA Leiden, The Netherlands
Received May 28, 2004
Adenosine receptor agonists are usually variations of the natural ligand, adenosine. The ribose
moiety in the ligand has previously been shown to be of great importance for the agonistic
effects of the compound. In this paper, we present a series of nonadenosine ligands selective
for the adenosine A1 receptor with an extraordinary pharmacological profile. 2-Amino-4-benzo-
[1,3]dioxol-5-yl-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarbonitrile (70, LUF 5853) shows full
agonistic behavior comparable with the reference compound CPA, while also displaying
comparable receptor binding affinity (Ki ) 11 nM). In contrast, compound 58 (2-amino-4-(3-
trifluoromethylphenyl)-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarbonitrile, LUF 5948) has
a binding affinity of 14 nM and acts as an inverse agonist. Also present within this same series
are compounds that show neutral antagonism of the adenosine A1 receptor, for example
compound 65 (2-amino-4-(4-difluoromethoxyphenyl)-6-(2-hydroxyethylsulfanyl)pyridine-3,5-
dicarbonitrile, LUF 5826).
Introduction
at any of the adenosine receptors (being only slightly
more favorable toward the A1 receptor). In this paper
we report on a novel series based upon this same
template and examine their affinity and selectivity for
the adenosine A1 receptor. We also show their remark-
able pharmacological profile in terms of their ability to
block or activate the human A1 receptor influencing the
cAMP production in CHO cells stably expressing this
receptor.
The adenosine A1 receptor is the most extensively
researched of the family of adenosine receptors, which
also include the A2A, A2B, and A3 receptors. Many
different classes of ligands have been designed, synthe-
sized, and tested at the A1 receptor, and to date there
are several recurring factors which almost define the
affinity or efficacy of the compound. The antagonists can
be generally split into two groups, those based on the
xanthine core structure and those heterocycles which
are nonxanthine compounds. This latter category en-
compasses all kinds of different mono-, bi-, tri-, and even
quadricyclic mostly nitrogen-containing aromatic com-
pounds.1,2,3 On the contrary, until very recently (partial)
agonists were all derivatives of the natural ligand itself,
adenosine. Research had shown that, in general, ma-
nipulation of the ribose moiety results in a reduction of
the intrinsic activity toward the adenosine A1 receptor,4-7
and thus accordingly all new nonadenosine ligands were
assumed to be antagonists. The main exception to this
is modification at the 5′-position, where replacement of
the 5′-hydroxyl with 5′-N-carboxamides is particularly
well tolerated, e.g., NECA and derivatives thereof,
although these ligands tend not to be very selective.3
Recently, based upon some data published in patent
literature,8,9 we showed that certain 2-amino-4-(substi-
tuted)-phenyl-6-(1H-imidazol-2-ylmethylsulfanyl)pyridine-
3,5-dicarbonitriles had varying degrees of efficacy,
ranging from full to partial agonists at the different
adenosine receptors.10 Most significantly, these com-
pounds displayed a significant affinity and efficacy at
the A2B receptor that had not been reported previously.
However, these ligands were not particularly selective
Results and Discussion
Chemistry. The compounds 55-72 were synthesized
according to Scheme 1.8,9 The aldehyde was reacted with
malononitrile in a straightforward Knoevenagel con-
densation in the presence of a few drops of piperidine
to give the intermediates 1-18 in moderate to good
yields (30-93%). Pyridine formation occurred according
to a preparation by Kambe et al.11 to give the phenyl-
protected sulfide in the 6-position of the ring. The
functionalized malononitrile was refluxed with another
equivalent of malononitrile and an equivalent of thiophe-
nol in ethanol and triethylamine, resulting in com-
pounds 19-36 (generally between 20 and 50% yields).
To obtain the free thiol in the 6- position of the pyridine
ring, 3.3 equiv of sodium sulfide in DMF at 80 °C for
2-3 h resulted in quantitative yields of compounds 37-
54. Throughout the synthesis, purification of the crude
product was not performed, nor were the reactions
optimized. The yields stated correspond to the crude
material, and it was this crude substance that was used
in subsequent reactions. The final step was the reaction
of the free thiol with 2-bromoethanol in the presence of
NaHCO3 in DMF at room temperature to give com-
pounds 55-72 in modest to good yields. The final
products were purified by column chromatography, and
subsequent recrystallization gave clean, fully character-
ized compounds.
* To whom correspondence should be addressed. E-mail: ijzerman@
lacdr.leidenuniv.nl. Tel: +31 (0)71 527 4651. Fax: +31 (0)71 527 4565.
10.1021/jm049597+ CCC: $30.25 © 2005 American Chemical Society
Published on Web 09/29/2004

2046 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6
Chang et al.
Scheme 1. Synthetic Route to 2-Amino-4-[(substituted)Phenyl]-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarbonitriles
55-72a
a (a) piperidine, EtOH, 1 h reflux; (b) malononitrile, thiophenol, triethylamine, EtOH, 4 h reflux; (c) (i) Na2S, DMF, (ii) 1 M HCl;
(d) 2-bromoethanol, NaHCO3, DMF.
Table 1. Affinities of the 2-Amino-4-[(substituted)phenyl]-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarbonitriles (55-72) in
Radioligand Binding Assays at the Human Adenosine Receptors and the Effect in cAMP Assays with the Human A1 Adenosine
Receptor
Ki (nM) or % disp.a
% change in cAMPe
b
c
d
compound code
R
R1
hA1
hA2A
hA3
CPA
DPCPX
N0840
55
-
-
10 ( 1.3
6.1 ( 1.6
1081 ( 69
15 ( 4
79022
12922
n.d.
23%
19%
4%
281 ( 5623
1700 ( 17013
15%13
26%
8%
-100
-
-
100
-
-
72 ( 7
H
F
H
H
H
F
H
-72 ( 6
-102 ( 10
-51 ( 10
85 ( 12
48 ( 6
56
40 ( 10
81 ( 5
57
0%
58
CF3
H
CH3
H
OH
H
OCF2H
H
OCF3
H
OCH3
H
-OCH2O-
OCH3
H
14 ( 2
22%
13%
29%
20%
25%
37%
8%
49%
0%
5%
0%
16%
0%
59
CF3
H
CH3
H
OH
H
OCF2H
H
150 ( 38
81 ( 17
49 ( 9
60
-78 ( 6
-16 ( 6
-105 ( 22
-93 ( 12
6 ( 5
61
62
12 ( 3
23 ( 3
63
64
40 ( 20
34 ( 7
12%
23%
11%
10%
18%
21%
13%
0%
65
0%
4 ( 6
66
30 ( 7
3%
80 ( 13
12 ( 12
-58 ( 10
-16 ( 19
-111 ( 16
-
67
OCF3
H
OCH3
156 ( 40
4.3 ( 0.6
41 ( 8
0%
68
21%
8%
69
70
11 ( 2
25%
245 ( 100
33%
25%
0%
71
OCH3
N(CH3)2
72
7%
-65 ( 23
a Ki ( SEM (n ) 3), % displacement (n ) 2). b Displacement of specific [3H]DPCPX binding in CHO cells expressing human adenosine
A1 receptors or % displacement of specific binding at 1 µM concentrations. c Displacement of specific [3H]ZM 241385 binding in CHO cells
expressing human adenosine A2A receptors or % displacement of specific binding at 1 µM concentrations. d Displacement of specific [125I]AB-
MECA binding in HEK 293 cells expressing human adenosine A3 receptors or % displacement of specific binding at 1 µM concentrations.
e % change of cAMP ( SEM (n ) 3) compared to CPA (full agonist, -100%) and DPCPX (full inverse agonist 100%). cAMP generation was
stimulated with 10 µM forskolin and compounds were tested at mostly 100 × their Ki value (see text).
Biology. All compounds were tested in radioligand
binding assays to determine their affinities at the
respective human adenosine receptors (Table 1). Com-
pounds for which the affinities were determined (i.e.,
compounds which showed greater than 50% displace-
ment of the radioligand at 1 µM) were tested in
functional assays for their ability to influence the levels
of cAMP in CHO cells expressing the human adenosine

Adenosine A1 Receptor Agonist-Antagonist
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6 2047
A1 receptor. The compounds were tested at concentra-
tions of mostly 100 × Ki, and at least 20 × Ki, where
the receptor sites should be almost fully occupied, and
the reference ligands CPA, DPCPX, and N0840 were
included in the assays. In [35S]GTPγS binding assays
on similar cell membranes from CHO cells these three
particular ligands are respectively reported as a full
agonist, an inverse agonist, and a neutral antago-
nist.12,13 In cAMP assays on whole cells expressing the
human A1 receptor, CPA has been shown to be a full
agonist and DPCPX as an inverse agonist.12 To exclude
the possibility of effects caused by endogeneous adeno-
sine that may be present in the system, adenosine
deaminase was added to the incubation medium.
Structure-Activity Relationships and Molecu-
lar Modeling. The results of the radioligand binding
assays are presented in Table 1. The compounds are
highly selective for the human A1 adenosine receptor,
with not one variation displaying more than 50%
displacement of the radioligand at a concentration of
1 µM on the human A2A or A3 receptors. A selection of
compounds was also tested on the human A2B receptor.
These also displayed either significantly less than or in
the region of 50% displacement of the radioligand at a
concentration of 1 µM.
Fluorine substitution caused many different effects
on both the binding and efficacy of the compounds
compared to the nonfluorinated analogues. Compounds
56 and 57, the 3- and 4-fluorophenyl derivatives, pos-
sessed a Ki value of 40 and 81 nM, respectively,
compared to the phenyl derivative 55, which had an
affinity of 15 nM. In general the highly electronegative
fluorine atom was detrimental for binding affinity
compared to its nonfluorinated analogue, but for two
exceptions. One exception was compound 58 with a Ki
value of 14 nM showing more than 5 times more affinity
to the A1 receptor than the 3-tolyl derivative 60 (Ki )
81 nM). Compound 59 abides by the generalization and
possesses an affinity of 150 nM (cf., compound 61, Ki )
49 nM).
The presence of the hydroxyl group (62, 63) also
increases the electronegativity of this region of the
ligand and thus might also have been detrimental for
binding affinity. However, the presence of the hydrogen
atom probably induces hydrogen-bonding with the
receptor, accounting for similarly high affinities (12 and
23 nM, respectively) to that of the nonsubstituted
phenyl derivative (55, Ki ) 15 nM). As mentioned
earlier, the presence of the fluorine atom was generally
detrimental for binding, and this was illustrated by the
fluorine-substituted methoxyphenyl ligands (64, 66, and
67) which showed less affinity for the receptor than the
analogous methoxyphenyl ligands (68, 69). The other
exception to the detrimental effect of fluorine was the
4-difluoromethoxyphenyl ligand 65, which displayed
comparable affinity to the 4-methoxyphenyl derivative
69 in the 30-40 nM range. With an affinity of 4 nM,
the compound with the most encouraging potency at the
A1 receptor was the 3-methoxyphenyl ligand (68).
Figure 1. Effect of reference and synthesized ligands on
forskolin-induced cAMP levels at hA1 adenosine receptors
compared to CPA (full agonist, 100% inhibition) and DPCPX
(full inverse agonist, 100% stimulation).
compounds displayed very similar levels of potency. The
spatial benefit of substitution in the 3- versus the 4-
position is illustrated well by compounds 68-70. The
3-methoxyphenyl derivative 68 displays almost 10-fold
better affinity than the 4-methoxyphenyl ligand 69 at
4.3 nM and 41 nM, respectively. Compound 70 that can
perhaps be seen as a compromise of these two moieties
holds an affinity consistent with this theory at 11 nM.
The 3,4-dimethoxyphenyl derivative 71, a combination
of compounds 68 and 69, has a much-reduced affinity
at the A1 receptor, with a displacement of the radio-
ligand of only 25% at a concentration of 1 µM. It is
obvious that the two methoxy groups create too much
steric bulk for good binding in the pocket to occur. The
4-dimethylaminophenyl derivative 72, like the methoxy-
analogues, possesses a methyl group connected to a
hydrogen-bond acceptor. However, the existence of a
second methyl substituent on this hydrogen-bond ac-
ceptor (causing steric hindrance, both in terms of
optimal fit in the receptor and in perhaps preventing
the occurrence of H-bonding) probably explains its
relatively poor affinity at the A1 adenosine receptor at
245 nM.
The radioligand binding assays on the human A1
adenosine receptor were performed with the inverse
agonist ligand [3H]DPCPX. This radioligand predomi-
nantly labels the low affinity state of the receptor, in
contrast to the agonist (radio)ligands. Therefore, the Ki
values obtained and given in Table 1 are a measure of
the binding to low affinity receptors. For this reason,
CPA exhibits a relatively large Ki value in this assay,
which is probably also true for the other agonists in
Table 1.
The behavior of the compounds in efficacy assays is
displayed in Figure 1. The ligands were tested at
concentrations of mostly 100 × Ki, and at least 20 × Ki,
where the receptor sites should be almost fully occupied.
The actual percentage inhibition or stimulation relative
to the full agonist CPA (100% inhibition) and DPCPX
(full inverse agonist, set at 100% stimulation) is given
in Table 1. As the A1 receptor is coupled to an inhibitory
G protein, forskolin was used to induce the production
of cAMP. N0840 (reported in [35S]GTPγS assays to be a
In general, substitution in the 3-position was more
positive for binding than substitution in the 4-position
of the phenyl ring. The only exceptions to this were the
tolyl compounds (60, 61) and the difluoromethoxyphenyl
compounds (64 and 65), although these latter two

2048 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6
Chang et al.
neutral antagonist)12,13 was shown in this whole cell
cAMP assay to have properties consistent with an
inverse agonist, although on lower levels than DPCPX.
Almost full agonistic activity was found for four
compounds (56, 62, 63, 70), the 3-fluorophenyl, the
hydroxyphenyl substitutents, and the benzo[1,3]dioxol-
5-yl derivative. Five further compounds (55, 57, 60, 68,
72) display high levels of agonism, with efficacy ranging
from 50 to 80%, with respect to full inhibition by CPA.
Lower levels of partial agonism were demonstrated by
compound 61 (the 4-tolyl derivative) and 69 (the 4-meth-
oxyphenyl substitutent). The compounds with trifluoro
derivatives (58, 59, 66, 67) induce inverse agonism,
resulting in an increase in cAMP levels. The two
difluoromethoxyphenyl derivatives (64 and 65) show
very little effect on the levels of cAMP even after
stimulation by forskolin, indicating almost neutral
antagonism. These compounds compare favorably with
a previous exploration for neutral antagonists by De
Ligt et al.13 In that paper, an 8-substituted N6-cyclo-
pentyl-9-methyladenine, LUF 5674, possessed a binding
affinity at the A1 adenosine receptor of 75 nM while
displaying neutral antagonism. In this study we have
achieved neutral antagonism with Ki values in the
region of 40 nM.
For each pair of 3- and 4-substituted analogues the
3-substituent shows more acute behavior than the
4-substituent. For example, the 3-fluorophenyl deriva-
tive 56 is a full agonist compared to partial agonism
displayed by the 4-fluorophenyl derivative 57, and the
3-CF3 derivative 58 displays a higher inverse agonistic
behavior than compound 59, the 4-CF3 analogue. Al-
though binding affinity and efficacy are not necessarily
correlated, these findings support the outcome of the
binding studies. The results suggest that substitution
in the 3-position fits the receptor pocket better than
substitution in the 4-position to induce either activation
of the receptor, or a stabilization to return the receptor
to the ground state.
The mode of activation of these ligands at the aden-
osine A1 receptor is far from clear. The long held theory
of the necessity of the ribose group to activate the
receptor proposed the existence of a ribose-specific
binding domain in the receptor in which the activation
‘trigger’ resides. This current series of compounds, all
without a ribose moiety, possess possible hydrogen-
bonding groups (OH or NH2) that may bind in the
ribose-specific domain and activate the receptor. How-
ever, the potential to reside in this specific domain alone
may not explain the observations recorded. All the
compounds in the series consist of the same elements
and only vary at the phenyl moiety, some distance away
from the potential H-bonding hydroxyl and amino
functions, and yet it is these substitutions that cause
the great differences in receptor activation. A review by
Visiers et al.14 highlights the complexities involved with
receptor activation/inactivation and summarizes our
understanding of it. There are many proposed sites of
importance in the activation of a G protein-coupled
receptor (GPCR). The binding site itself for many
endogenous ligands at rhodopsin-like receptors has been
identified as being a spatially compact binding micro-
domain, most probably in transmembrane (TM) helices
3, 5, and 6, and there is evidence to suggest ligand
Figure 2. Structure of CPA and the numbering system of the
molecule.
orientation can be crucial for efficacy. Almaula et al.15
described how not only the binding affinity, but also the
activation of the serotonin 5-HT2A receptor was affected
by small differences in the mode of interaction of the
ligand with the receptor pocket. Following this, Ebersole
et al.16 investigated partial agonism of the human
serotonin 5-HT2A receptor. On the basis of mutagenesis
studies and computational simulation, they found that
ligand orientation affected the H-bonding at serine
residues in TM3 and TM5 and caused a reduction in
efficacy. In conclusion it was hypothesized that the same
structurally specific mechanism holds for other GPCRs.
Translating the proposals to the current findings, we
speculate that the strongly electron-withdrawing tri-
fluoromethyl and trifluoromethoxy groups cause such
a change in electron density about the molecule that
the manner of binding (with respect to the nonfluori-
nated ligands, e.g., 55) at the receptor is distorted
dramatically. This prevents not only activation of the
receptor but induces the stabilization of an already
active receptor to adopt an inactive form. In comparison,
the difluoromethoxy derivatives (64, 65) possess less
electronegativity than the trifluoro species, distorting
the manner of binding less and so resulting in neutral
antagonists. In the same vein, the singularly substituted
fluorine moiety (55, 56) is probably not sufficiently
electronegative and not large enough to induce a dif-
ferent binding mode; thus, agonism is conserved.
In an attempt to form a clearer basis to the potential
mode of binding of the agonists, we looked at the
electronic and spatial properties of these compounds in
the molecular modeling package Spartan.17 The agonists
were modeled in comparison to the known agonist, CPA.
For the clarity of the following discussion, CPA and its
numbering system is shown in Figure 2.
Assuming that the conservation of the ribose-3′OH
is necessary for activation of the agonist (as reported
by Taylor et al.18), we propose two alternative models
of superimposition depicted in Figures 3 and 4 (CPA and
compound 68). The first proposition suggests that the
SCH2CH2OH region can be overlayed upon the 3′OH
area. Examining the electronic configuration of the
compounds, the electronegative regions of the H-bond
accepting nitrile groups seem to correspond to the
electronegative N1 and N7 of the purine ring. Spatially,
the 3-substitution of the phenyl group in 68 also fits
well with the cyclopentyl group of CPA, suggesting the
advantage of 3-substitution above 4-substitution. The
alternative model depicted in Figure 4 overlays the
2-amino group from the current series with the 3′OH
of the ribose ring. The 3- and 5-nitrile groups of
compound 68 now lie in the region of the 2′OH of the
ribose group and the N7 of the purine ring, respectively.
Again 3-substitution at the phenyl group confers overlap
with the N6-cyclopentyl domain of CPA, offering more
optimal spatial compatibility than 4-substitution.

Adenosine A1 Receptor Agonist-Antagonist
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6 2049
Figure 3. 2-Amino-4-(3-methoxyphenyl)-6-(2-hydroxyethyl-
sulfanyl)pyridine-3,5-dicarbonitrile (68) superimposed upon
CPA. The upper superimposition depicts the proposed overlap
of electron density between the two nitrile groups and the
purine ring. The lower picture is the superimposition shown
from a different angle and displays the steric overlap of the
N6 substituent of CPA and the 3-methoxy substituent of the
phenyl group. Also seen more clearly in this lower picture is
the compatibility of the 3′OH group from the ribose ring and
the hydroxyl moiety from compound 68.
Figure 4. An alternative superimposition of compound 68 and
CPA. Here the anchor point has been taken as the 2-amino
group of 68 with the 3′OH of the ribose ring. The topmost
picture suggests overlap of the ribose 2′OH group with the
3-nitrile moiety, and the purine N7 with the 5-nitrile group.
The lower picture is the superimposition shown from a
different angle and displays more clearly the overlap of the
2-amino group from 68 with the 3′OH of CPA.
Experimental Section
The rationale in this modeling procedure was to
search for maximal overlap between the two structures.
However, it should be kept in mind that the binding
sites for the two compounds (CPA and 68) may be
(partially or even entirely) different, but converge in the
subsequent activation of the receptor.
Materials and Methods. All reagents used were obtained
from commercial sources and all solvents were of an analytical
grade. 1H and 13C NMR spectra were recorded on a Bruker
AC 200 (1H NMR, 200 MHz; 13C NMR, 50.29 MHz) spectrom-
eter with tetramethylsilane as an internal standard. Chemical
shifts are reported in δ(ppm), and the following abbreviations
are used: s ) singlet, d ) doublet, t ) triplet, m ) multiplet,
br ) broad. Melting points were determined on a Bu¨chi
melting point apparatus and are uncorrected. The value stated
for each compound is the initial temperature at which the
compound begins to melt. Elemental analyses were performed
by the Leiden Institute of Chemistry and are within 0.4% of
the theoretical values unless otherwise stated. Reactions were
routinely monitored by TLC using Merck silica gel F254 plates.
General Procedure for Functionalized Malononitriles
(1-18). To malononitrile (1.32 mL, 20.8 mmol) dissolved in
EtOH (14 mL) was added an aldehyde (20 mmol) followed by
3 drops of piperidine. This reaction mixture was then refluxed
for 1 h and then allowed to cool to room temperature, upon
which a precipitate formed. The crude product was collected
by filtration and used without any further purification.
Conclusion
This paper describes a series of ligands displaying
very different activity at the A1 adenosine receptor.
Several compounds are full agonists with affinity and
efficacy comparable to the reference ligand CPA. There
are also compounds in this series that display fully
inverse agonistic behavior, almost on levels comparable
to the reference ligand DPCPX. Most strikingly, neutral
antagonists have also been discovered which retain high
affinity and selectivity for the adenosine A1 receptor. It
is perhaps particularly noteworthy that these effects
vary with the presence or absence of fluorine atoms
distant from the phenyl ring. This suggests the sub-
stantial influence of hydrogen-bond acceptors in this
part of the molecule upon the activation of the adenosine
A1 receptor.
Compounds 1, 8, 9, 14, 15 have been reported previously.10
NMR values for compounds 2, 3, 7, 17, 18 were in accordance
to literature data.19,20
2-(3-Trifluorobenzylidene)-malononitrile (4). Off white
solid. 1H NMR δ(DMSO): 8.67 (s, 1H, CH), 8.27-8.21 (m, 1H,

2050 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6
Chang et al.
phenyl-H), 8.09-8.05 (m, 1H, phenyl-H), 7.92-7.84 (m, 1H,
phenyl-H).
2-(4-Trifluorobenzylidene)-malononitrile (5). Off white
solid. 1H NMR δ(DMSO): 8.69 (s, 1H, CH), 8.14-7.99 (m, 4H,
phenyl-H).
2-(3-Methylbenzylidene)-malononitrile (6). Off white
solid. 1H NMR δ(DMSO): 8.51 (s, 1H, CH), 7.80-7.77 (m, 2H,
phenyl-H), 7.54-7.51 (m, 2H, phenyl-H), 2.37 (s, 3H, CH3).
2-Amino-4-(3-trifluoromethoxyphenyl)-6-phenylsul-
fanylpyridine-3,5-dicarbonitrile (30). Yellow solid. 1H
NMR δ(DMSO): 7.89 (br s, 2H, NH2), 7.75-7.51 (m, 9H,
phenyl-H).
2-Amino-4-(4-trifluoromethoxyphenyl)-6-phenylsul-
fanylpyridine-3,5-dicarbonitrile (31). Yellow solid. 1H
NMR δ(DMSO): 7.89 (br s, 2H, NH2), 7.76 (d, 2H, J ) 6.6 Hz,
OCF3-phenyl-H), 7.66 (m, 7H, OCF3-phenyl-H + phenyl-H).
2-Amino-4-benzo[1,3]dioxol-5-yl-6-phenylsulfanylpyri-
dine-3,5-dicarbonitrile (34). Yellow solid. 1H NMR δ(DMSO):
7.80 (br s, 2H, NH2), 7.64-7.60 (m, 2H, S-phenyl-H), 7.53-
7.50 (m, 3H, S-phenyl-H), 7.20-7.04 (m, 3H, phenyl-H), 6.18
(s, 2H, -OCH2O-).
2-Amino-4-(3,4-dimethoxyphenyl)-6-phenylsulfanylpyr-
idine-3,5-dicarbonitrile (35). Yellow solid. 1H NMR δ(DMSO):
7.77 (br s, 2H, NH2), 7.65-7.50 (m, 7H, phenyl-H), 7.15 (d,
2H, J ) 8.8 Hz, phenyl-H), 3.67 (s, 3H, CH3), 3.63 (s, 3H, CH3).
2-Amino-4-(4-(dimethylamino)phenyl)-6-phenylsulfan-
ylpyridine-3,5-dicarbonitrile (36). Yellow solid. 1H NMR
δ(DMSO): 7.68 (br s, 2H, NH2), 7.64-7.60 (m, 2H, phenyl-H),
7.54-7.48 (m, 2H, phenyl-H), 7.44 (d, 2H, J ) 8.8 Hz, 4-NMe2-
phenyl-H), 6.86 (d, 2H, J ) 8.8 Hz, 4-NMe2-phenyl-H), 3.03
(s, 6H, 2 × CH3).
2-(3-Difluoromethoxybenzylidene)-malononitrile (10).
Off-white solid. 1H NMR δ(CDCl3): 7.79-7.75 (m, 2H, phenyl-
H), 7.66 (s, 1H, CH), 7.56 (m, 1H, phenyl-H), 7.43-7.39 (m,
1H, phenyl-H), 6.58 (t, 1H, J ) 72.3 Hz, OCF2H).
2-(4- Difluoromethoxybenzylidene)-malononitrile (11).
1
Off-white solid. H NMR δ(CDCl3): 7.96 (d, 2H, J ) 8.76 Hz,
phenyl-H), 7.75 (s, 1H, CH), 7.27 (d, 2H, J ) 8.8 Hz, phenyl-
H), 6.65 (t, 1H, J ) 72.3 Hz, OCF2H).
2-(3-Trifluoromethoxybenzylidene)-malononitrile (12).
Off-white solid. 1H NMR δ(CDCl3): 7.89-7.86 (m, 1H, phenyl-
H), 7.80 (s, 1H, CH), 7.74 (m, 1H, phenyl-H), 7.62 (m, 1H,
phenyl-H), 7.51-7.47 (m, 1H, phenyl-H).
2-(4-Trifluoromethoxybenzylidene)-malononitrile (13).
1
Off-white solid. H NMR δ(CDCl3): 7.99 (d, 2H, J ) 8.8 Hz,
phenyl-H), 7.78 (s, 1H, CH), 7.38 (d, 2H, J ) 8.8 Hz, phenyl-
H).
General Procedure for 2-Amino-4-(substituted)-phen-
yl-6-mercaptopyridine-3,5-dicarbonitriles (37-54). The
pyridine (19-36) (3 mmol) was dissolved in DMF (10 mL) and
to this was added sodium sulfide (0.78 g, 10 mmol) and the
mixture stirred at 80 °C for 2 h. Upon cooling to room
temperature, 1 M HCl (20 mL) was added, resulting in the
formation of a yellow precipitate. The crude product was
collected by filtration.
Compounds 37, 44, 45, 50, 51 have been reported previ-
ously.10 NMR values for compound 43 were in accordance to
literature data.11
2-Amino-4-(3-fluorophenyl)-6-mercaptopyridine-3,5-
dicarbonitrile (38). Yellow solid. 1H NMR δ(DMSO): 7.65-
7.59 (m, 1H, phenyl-H), 7.48-7.35 (m, 3H, phenyl-H).
2-Amino-4-(4-fluorophenyl)-6-mercaptopyridine-3,5-
dicarbonitrile (39). Yellow solid. 1H NMR δ(DMSO): 7.58-
7.52 (m, 2H, phenyl-H), 7.44-7.39 (m, 2H, phenyl-H).
2-Amino-4-(3-trifluoromethylphenyl)-6-mercaptopyri-
dine-3,5-dicarbonitrile (40). Yellow solid. 1H NMR δ(DMSO):
7.91-7.78 (m, 4H, phenyl-H).
2-Benzo[1,3]dioxol-5-ylmethylenemalononitrile (16).
1
Off white solid. H NMR δ(DMSO): 8.39 (s, 1H, CH), 7.57-
7.54 (m, 2H, phenyl-H), 7.24-7.19 (m, 1H, phenyl-H), 6.25 (s,
2H, -OCH2O-).
General Procedure for 2-Amino-4-(substituted)phen-
yl-6-phenylsulfanylpyridine-3,5-dicarbonitriles (19-36).
To a solution of the functionalized malononitrile (1-18) (10
mmol) in EtOH (10 mL), was added malononitrile (0.64 mL,
10 mmol), thiophenol (1.02 mL, 10 mmol), and triethylamine
(50 µL), and the mixture heated at reflux for approximately
4 h. The reaction mixture was then allowed to cool to room
temperature. The crude product precipitated upon cooling and
was collected by filtration.
Compounds 19, 26, 27, 32, 33 have been reported previ-
ously.10
2-Amino-4-(3-fluorophenyl)-6-phenylsulfanylpyridine-
3,5-dicarbonitrile (20). Yellow solid. 1H NMR δ(DMSO): 7.90
(br s, 2H, NH2), 7.71-7.40 (m, 9H, phenyl-H).
2-Amino-4-(4-fluorophenyl)-6-phenylsulfanylpyridine-
3,5-dicarbonitrile (21). Yellow solid. 1H NMR δ(DMSO): 7.64
(br s, 2H, NH2), 7.62-7.59 (m, 4H, phenyl-H), 7.52-7.40 (m,
5H, phenyl-H).
2-Amino-4-(4-trifluoromethylphenyl)-6-mercaptopyri-
dine-3,5-dicarbonitrile (41). Yellow solid. 1H NMR δ(DMSO):
7.96 (d, 2H, J ) 8.0 Hz, phenyl-H), 7.76 (d, 2H, J ) 8.0 Hz,
phenyl-H).
2-Amino-4-(3-trifluoromethylphenyl)-6-phenylsulfan-
ylpyridine-3,5-dicarbonitrile (22). Yellow solid. 1H NMR
δ(DMSO): 7.85 (br s, 2H, NH2), 7.83-7.70 (m, 4H, phenyl-H),
7.66-7.61 (m, 2H, phenyl-H), 7.56-7.51 (m, 3H, phenyl-H).
2-Amino-4-(4-trifluoromethylphenyl)-6-phenylsulfan-
ylpyridine-3,5-dicarbonitrile (23). Yellow solid. 1H NMR
δ(DMSO): 8.01-7.92 (m + br s, 3H, NH2 + CF3-phenyl-H),
7.83-7.78 (m, 2H, CF3-phenyl-H), 7.64-7.60 (m, 2H, phenyl-
H), 7.53-7.43 (m, 3H, phenyl-H).
2-Amino-4-(3-methylphenyl)-6-phenylsulfanylpyridine-
3,5-dicarbonitrile (24). Yellow solid. 1H NMR δ(DMSO): 7.83
(br s, 2H, NH2), 7.66-7.61 (m, 2H, phenyl-H), 7.56-7.49 (m,
3H, phenyl-H), 7.45-7.33 (m, 4H, tolyl-H), 2.41 (s, 3H, CH3).
2-Amino-4-(4-methylphenyl)-6-phenylsulfanylpyridine-
3,5-dicarbonitrile (25). Yellow solid. 1H NMR δ(DMSO): 7.60
(br s, 2H, NH2), 7.55-7.52 (m, 2H, phenyl-H), 7.50-7.38 (m,
7H, phenyl-H + tolyl-H), 2.43 (s, 3H, CH3).
2-Amino-4-(3-methylphenyl)-6-mercaptopyridine-3,5-
dicarbonitrile (42). Yellow solid. 1H NMR δ(DMSO): 7.57-
7.30 (m, 4H, phenyl-H), 2.40 (s, 3H, CH3).
2-Amino-4-(3-difluoromethoxyphenyl)-6-mercaptopyr-
idine-3,5-dicarbonitrile (46). Yellow solid. 1H NMR δ(MeOH):
7.46-7.33 (m, 4H, phenyl-H), 6.89 (t, 1H, J ) 73.4 Hz, OCF2H).
2-Amino-4-(4-difluoromethoxyphenyl)-6-mercaptopyr-
idine-3,5-dicarbonitrile (47). Yellow solid. 1H NMR δ(MeOD):
7.59 (m, 2H, phenyl-H), 7.30 (m, 2H, phenyl-H), 6.97 (t, 1H,
J ) 73.1 Hz, OCF2H).
2-Amino-4-(3-trifluoromethoxyphenyl)-6-mercapto-
pyridine-3,5-dicarbonitrile (48). Yellow solid. 1H NMR
δ(MeOD): 7.67-7.62 (m, 1H, phenyl-H), 7.54-7.47 (m, 3H,
phenyl-H).
2-Amino-4-(4-trifluoromethoxyphenyl)-6-mercapto-
pyridine-3,5-dicarbonitrile (49). Yellow solid. 1H NMR
δ(MeOD): 7.67-7.63 (m, 2H, phenyl-H), 7.48-7.44 (m, 2H,
phenyl-H).
2-Amino-4-benzo[1,3]dioxol-5-yl-6-mercaptopyridine-
3,5-dicarbonitrile (52). Yellow solid. 1H NMR δ(DMSO):
7.13-7.00 (m, 3H, phenyl-H), 6.17 (s, 2H, OCH2O).
2-Amino-4-(3-difluoromethoxyphenyl)-6-phenylsulfan-
ylpyridine-3,5-dicarbonitrile (28). Yellow solid. 1H NMR
δ(DMSO): 7.88 (br s, 2H, NH2), 7.67-7.34 (m, 9H, phenyl-H),
6.95 (m, 1H, OCF2H).
2-Amino-4-(4-difluoromethoxyphenyl)-6-phenylsulfan-
ylpyridine-3,5-dicarbonitrile (29). Yellow solid. 1H NMR
δ(DMSO): 7.85 (br s, 2H, NH2), 7.69-7.38 (m, 9H, phenyl-H),
7.07 (m, 1H, OCF2H).
2-Amino-4-(3,4-dimethoxyphenyl)-6-mercaptopyridine-
3,5-dicarbonitrile (53). Yellow solid. 1H NMR δ(DMSO):
7.17-7.13 (m, 3H, phenyl-H), 3.85 (s, 3H, OCH3), 3.81 (s, 3H,
OCH3).

Adenosine A1 Receptor Agonist-Antagonist
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6 2051
2-Amino-4-(4-(dimethylamino)phenyl)-6-mercaptopyr-
idine-3,5-dicarbonitrile (54). Yellow solid. 1H NMR δ(DMSO):
7.41 (d, 2H, J ) 8.8 Hz, phenyl-H), 6.83 (d, 2H, J ) 8.8 Hz,
phenyl-H), 3.03 (s, 6H, 2 × CH3).
7.41-7.33 (m, 1H, phenyl-H), 6.98-6.89 (m, 3H, phenyl-H),
5.05 (t, 1H, J ) 5.5 Hz, OH), 3.73-3.64 (m, 2H, OCH2), 3.42-
3.33 (m, 2H, CH2S). 13C NMR δ(DMSO): 167.1, 135.2, 130.0,
118.9, 117.3, 115.4, 115.3, 115.1, 93.6, 86.0, 59.5, 32.7. MS
(ESI): 312.8. Anal. (C15H12N4O2S‚0.6H2O) C, H, N, S.
2-Amino-4-(4-hydroxyphenyl)-6-(2-hydroxyethylsul-
fanyl)pyridine-3,5-dicarbonitrile (63). White solid, 40%.
mp: 224 °C dec 1H NMR δ(DMSO): 7.95 (br s, 2H, NH2) 7.39
(m, 2H, phenyl-H), 6.93 (m, 2H, phenyl-H), 5.04 (br s, 1H, OH),
3.70-3.64 (m, 2H, OCH2), 3.32-3.19 (m, 2H, CH2S). 13C NMR
δ(DMSO): 167.1, 130.3, 124.3, 115.7, 115.4, 59.5, 32.7. MS
(ESI): 312.7. Anal. (C16H14N4O2S‚0.1MeOH), C, H, N, S.
2-Amino-4-(3-difluoromethoxyphenyl)-6-(2-hydroxy-
ethylsulfanyl)pyridine-3,5-dicarbonitrile (64). White solid,
36%. mp: 140 °C. 1H NMR δ(MeOD): 7.60-7.52 (m, 1H,
phenyl-H), 7.38-7.31 (m, 3H, phenyl-H), 6.86 (t, 1H, 73.4 Hz,
CF2H), 3.79 (t, 2H, J ) 6.6 Hz, OCH2), 3.39 (t, 2H, J ) 6.6 Hz,
CH2S). 13C NMR δ(DMSO): 169.4, 161.4, 158.5, 152.8, 137.0,
131.7, 126.5, 122.0, 120.4, 117.6, 116.1, 116.0, 95.5, 87.2, 61.4,
33.5. MS (ESI): 363. Anal. (C16H12F2N4O2S) C, H, N, S.
2-Amino-4-(4-difluoromethoxyphenyl)-6-(2-hydroxy-
ethylsulfanyl)pyridine-3,5-dicarbonitrile (65). White solid,
44%. mp: 156 °C. 1H NMR δ(DMSO): 8.04 (br s, 2H, NH2)
7.65 (m, 2H, phenyl-H), 7.38 (m, 2H, phenyl-H), 7.43 (s, 1H,
CF2H), 5.04 (t, 1H, J ) 5.5 Hz, OH), 3.73-3.64 (m, 2H, OCH2),
3.38-3.33 (m, 2H, CH2S). 13C NMR δ(DMSO): 167.1, 159.6,
157.4, 152.4, 130.7, 118.4, 115.5, 115.3, 93.4, 85.8, 59.5, 32.7.
MS (ESI): 363.0. Anal. (C16H12F2N4O2S), C, H, N, S.
2-Amino-4-(3-trifluoromethoxyphenyl)-6-(2-hydroxy-
ethylsulfanyl)pyridine-3,5-dicarbonitrile (66). White solid,
69%. mp: 125 °C. 1H NMR δ(DMSO): 8.08 (br s, 2H, NH2)
7.73-7.59 (m, 4H, phenyl-H), 5.04 (t, 1H, J ) 5.5 Hz, OH),
3.73-3.64 (m, 2H, OCH2), 3.38-3.33 (m, 2H, CH2S). 13C NMR
δ(DMSO): 167.1, 159.5, 156.5, 148.2, 136.1, 131.0, 127.7, 123.3,
122.9, 121.5, 115.2, 115.0, 85.8, 59.5, 32.6. MS (ESI): 380.7.
Anal. Calc. for C16H11F3N4O2S (C 50.53; H 2.92; N 14.73; S
8.43) found (C 50.46; H 2.39; N 15.01; S 8.16) %.
2-Amino-4-(4-trifluoromethoxyphenyl)-6-(2-hydroxy-
ethylsulfanyl)pyridine-3,5-dicarbonitrile (67). White solid,
75%. mp: 144 °C. 1H NMR δ(DMSO): 8.08 (br s, 2H, NH2)
7.74-7.58 (m, 4H, phenyl-H), 5.04 (t, 1H, J ) 5.5 Hz, OH),
3.73-3.64 (m, 2H, OCH2), 3.37-3.33 (m, 2H, CH2S). 13C NMR
δ(DMSO): 167.2, 159.6, 157.1, 149.6, 133.2, 131.0, 121.2, 115.4,
115.2, 93.6, 85.8, 59.5, 32.7. MS (ESI): 380.8. Anal. (C16H11-
F3N4O2S) C, H, N, S.
General Procedure for 2-Amino-4-[(substituted)phen-
yl]-6-(2-hydroxyethylsulfanyl)pyridine-3,5-dicarboni-
triles (55-72). The free thiol (37-54) (1 mmol) was stirred
with 2-bromoethanol (84 µL, 1 mmol) and NaHCO3 (0.34 g, 1
mmol) in DMF (2 mL) at room temperature for 2 h. Water (10
mL) was added to precipitate the crude product, which was
collected by filtration. Purification by column chromatography
on SiO2, eluting with ethyl acetate and/or ethyl acetate-
methanol mixtures, followed by recrystallization gave the
desired products. The solvents used in the recrystallization
process were CH2Cl2 and MeOH.
2-Amino-4-phenyl-6-(2-hydroxyethylsulfanyl)pyridine-
3,5-dicarbonitrile (55). White solid, 54%. mp: 211 °C. 1H
NMR δ(MeOD): 7.56-7.48 (m, 5H, phenyl-H), 3.80 (t, 2H,
J ) 6.6 Hz, OCH2), 3.40 (t, 2H, J ) 6.6 Hz, CH2S). 13C NMR
δ(MeOD): 167.1, 159.6, 158.3, 134.0, 130.3, 128.7, 128.4, 115.4,
115.3, 96.1, 85.7, 59.5, 32.6. MS (ESI): 296.9. Anal. (C15H12N4-
OS‚0.3DMF) C, H, N, S.
2-Amino-4-(3-fluorophenyl)-6-(2-hydroxyethylsulfanyl)-
pyridine-3,5-dicarbonitrile (56). White solid, 38%. mp: 189
°C. 1H NMR δ(DMSO): 8.09 (br s, 2H, NH2) 7.70-7.59 (m,
1H, phenyl-H), 7.52-7.37 (m, 3H, phenyl-H), 5.05 (t, 1H, J )
5.5 Hz, OH), 3.72-3.64 (m, 2H, OCH2), 3.40-3.33 (m, 2H,
CH2S). 13C NMR δ(DMSO): 167.8, 159.5, 130.3, 130.2, 123.7,
117.1, 116.7, 115.2, 114.8, 114.3, 105.6, 83.3, 59.8, 31.5. MS
(ESI): 314.8 Anal. (C15H11FN4OS‚0.3H2O) C, H, N, S.
2-Amino-4-(4-fluorophenyl)-6-(2-hydroxyethylsulfanyl)-
pyridine-3,5-dicarbonitrile (57). White solid, 54%. mp: 133
°C. 1H NMR δ(DMSO): 8.06 (br s, 2H, NH2) 7.68-7.60 (m,
2H, phenyl-H), 7.49-7.40 (m, 2H, phenyl-H), 5.04 (t, 1H, J )
5.5 Hz, OH), 3.73-3.64 (m, 2H, OCH2), 3.45-3.33 (m, 2H,
CH2S). 13C NMR δ(DMSO): 167.0, 166.0, 160.0, 157.4, 131.1,
131.0, 130.4, 116.0, 115.6, 115.4, 115.2, 85.8, 83.9, 59.4, 32.6.
MS (ESI): 315.0. Anal. (C15H11FN4OS‚0.3MeOH) C, H, N, S.
2-Amino-4-(3-trifluoromethylphenyl)-6-(2-hydroxyeth-
ylsulfanyl)pyridine-3,5-dicarbonitrile (58). White solid,
1
60%. mp: 164 °C. H NMR δ(DMSO): 8.10 (br s, 2H, NH2),
8.00-7.83 (m, 4H, phenyl-H), 5.05 (t, 1H, J ) 5.5 Hz, OH),
3.75-3.66 (m, 2H, OCH2), 3.42-3.36 (m, 2H, CH2S). 13C NMR
δ(DMSO): 167.2, 159.6, 156.8, 135.1, 132.7, 130.0, 129.8, 129.2,
127.0, 126.6, 125.6, 115.3, 115.1, 93.7, 85.9, 59.5, 32.7. MS
(ESI): 364.9. Anal. (C16H11F3N4OS‚0.07CH2Cl2) C, H, N, S.
2-Amino-4-(4-trifluoromethylphenyl)-6-(2-hydroxyeth-
ylsulfanyl)pyridine-3,5-dicarbonitrile (59). White Solid,
2-Amino-4-(3-methoxyphenyl)-6-(2-hydroxyethylsul-
fanyl)pyridine-3,5-dicarbonitrile (68). White solid, 63%.
1
69%. mp: 158 °C. H NMR δ(DMSO): 8.14 (br s, 2H, NH2),
1
mp: 188 °C. H NMR δ(DMSO): 8.01 (br s, 2H, NH2) 7.53-
7.99 (d, 2H, J ) 8.0 Hz, phenyl-H), 7.81 (d, 2H, J ) 8.0 Hz,
phenyl-H), 5.05 (t, 1H, J ) 5.5 Hz, OH), 3.74-3.65 (m, 2H,
OCH2), 3.45-3.36 (m, 2H, CH2S). 13C NMR δ(DMSO): 178.9,
167.2, 159.5, 157.0, 138.2, 130.8, 130.1, 129.6, 126.6, 125.7,
115.2, 115.0, 93.4, 85.7, 59.4, 32.7. MS (ESI): 364.6. Anal.
(C16H11F3N4OS‚0.2CH2Cl2) C, H, N, S.
7.45 (m, 1H, phenyl-H), 7.17-7.06 (m, 3H, phenyl-H), 5.06 (t,
1H, J ) 5.5 Hz, OH), 3.82 (s, 3H, CH3), 3.73-3.64 (m, 2H,
OCH2), 3.39-3.33 (m, 2H, CH2S). 13C NMR δ(DMSO): 167.1,
159.7, 159.1, 158.1, 135.3, 130.1, 120.6, 115.9, 115.5, 115.3,
114.1, 93.7, 85.8, 59.5, 55.4, 32.7. MS (ESI): 326.9. Anal.
(C16H14N4O2S‚0.2MeOH), C, H, N, S.
2-Amino-4-(3-methylphenyl)-6-(2-hydroxyethylsulfanyl)-
2-Amino-4-(4-methoxyphenyl)-6-(2-hydroxyethylsul-
fanyl)pyridine-3,5-dicarbonitrile (69). White solid, 54%.
mp: 188 °C. 1H NMR δ(DMSO): 7.96 (br s, 2H, NH2) 7.51 (d,
2H, J ) 8.8 Hz, phenyl-H), 7.13 (d, 2H, J ) 8.8 Hz, phenyl-
H), 5.03 (t, 1H, J ) 5.5 Hz, OH), 3.86 (s, 3H, CH3), 3.73-3.64
(m, 2H, OCH2), 3.37-3.33 (m, 2H, CH2S). 13C NMR δ(DMSO):
167.1, 160.8, 159.8, 158.0, 130.2, 125.9, 115.7, 115.6, 114.1,
85.7, 59.7, 59.5, 55.4, 32.6. MS (ESI): 327.0. Anal. (C16H14N4O2S‚
0.2H2O) C, H, N, S.
pyridine-3,5-dicarbonitrile (60). White Solid, 74%. mp: 178
1
°C. H NMR δ(DMSO): 8.02 (br s, 2H, NH2), 7.47-7.30 (m,
4H, phenyl-H), 5.04 (t, 1H, J ) 5.5 Hz, OH), 3.71-3.65 (m,
2H, OCH2), 3.40-3.34 (m, 2H, CH2S), 2.40 (s, 3H, CH3). 13C
NMR δ(DMSO): 167.1, 159.6, 158.4, 138.1, 134.0, 131.0, 128.7,
125.5, 115.5, 115.3, 93.6, 85.9, 59.5, 32.7, 20.9. MS (ESI):
311.2. Anal. (C16H14N4OS) C, H, N, S.
2-Amino-4-(4-methylphenyl)-6-(2-hydroxyethylsulfanyl)-
pyridine-3,5-dicarbonitrile (61). White Solid, 81%. mp: 206
1
2-Amino-4-benzo[1,3]dioxol-5-yl-6-(2-hydroxyethylsul-
°C. H NMR δ(DMSO): 8.00 (br s, 2H, NH2), 7.46-7.36 (m,
4H, phenyl-H), 5.03 (t, 1H, J ) 5.5 Hz, OH), 3.73-3.64 (m,
2H, OCH2), 3.39-3.33 (m, 2H, CH2S), 2.41 (s, 3H, CH3). 13C
NMR δ(DMSO): 167.0, 159.6, 158.3, 140.2, 131.1, 129.2, 128.4,
115.5, 115.3, 93.6, 85.7, 59.5, 32.6, 20.9. MS (ESI): 310.8. Anal.
(C16H14N4OS) C, H, N, S.
fanyl)pyridine-3,5-dicarbonitrile (70). White solid, 42%.
1
mp: 192 °C. H NMR δ(DMSO): 7.99 (br s, 2H, NH2) 7.16-
7.01 (m, 3H, phenyl-H), 6.17 (s, 2H, -OCH2O-), 5.02 (t, 1H,
J ) 5.5 Hz, OH), 3.67 (m, 2H, OCH2), 3.38-3.32 (m, 2H, CH2S).
13C NMR δ(DMSO): 197.7, 167.0, 159.7, 157.9, 148.9, 147.4,
127.4, 123.0, 115.6, 115.4, 108.9, 108.5, 101.8, 93.8, 59.5, 32.6.
MS (ESI): 341.1. Anal. Calc. for C16H12N4O3S (C 56.46; H 3.55;
N 16.46; S 9.42) found (C 56.31; H 3.11; N 16.38; S 9.46) %.
2-Amino-4-(3-hydroxyphenyl)-6-(2-hydroxyethylsul-
fanyl)pyridine-3,5-dicarbonitrile (62). White solid, 70%.
mp: 227-229 °C dec 1H NMR δ(DMSO): 8.01 (br s, 2H, NH2)

2052 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6
Chang et al.
2-Amino-4-(3,4-dimethoxyphenyl)-6-(2-hydroxyethyl-
Data Analysis. Ki values were calculated using a nonlinear
regression curve-fitting program (GraphPad Prism 4, Graph-
Pad Software Inc., San Diego, CA). KD values of the radio-
ligands were 1.6 nM, 1.0 nM, and 5.0 nM for [3H]DPCPX,
[3H]ZM 241385, and [125I]AB-MECA, respectively. The data
from the functional assays were also analyzed with GraphPad
Prism. Figure 1 and the results shown in Table 1 were
generated by evaluating the data to relate to the known
ligands CPA and DPCPX. The first step involved normalizing
the raw data with respect to CPA and DPCPX. CPA was
chosen as the baseline value as it has previously been shown
to be a full agonist, and DPCPX as the highest value, assuming
full inversely agonistic behavior. After normalization, the
Y-axis was shifted to present the data as being above or below
that of the forskolin-generated levels of cAMP. Finally, the
data were again normalized to assume 100% inhibition of
cAMP by the agonist CPA and 100% full inverse agonism by
DPCPX.
sulfanyl)pyridine-3,5-dicarbonitrile (71). White solid, 35%.
1
mp: 209 °C. H NMR δ(DMSO): 7.96 (br s, 2H, NH2) 7.19-
7.13 (m, 3H, phenyl-H), 5.02 (m, 1H, OH), 3.86 (s, 3H, OCH3),
3.81 (s, 3H, OCH3), 3.69-3.66 (m, 2H, OCH2), 3.32-3.19 (m,
2H, CH2S). 13C NMR δ(DMSO): 167.0, 159.7, 158.0, 150.4,
148.3, 126.0, 121.5, 115.7, 115.6, 112.3, 111.5, 92.7, 85.9, 59.5,
55.7, 55.6, 32.8. MS (ESI): 356.9. Anal. (C17H16N4O3S), C, H,
N, S.
2-Amino-4-(4-(dimethylamino)phenyl)-6-(2-hydroxy-
ethylsulfanyl)pyridine-3,5-dicarbonitrile (72). White solid,
43%. mp: 268 °C dec 1H NMR δ(DMSO): 7.87 (br s, 2H, NH2)
7.39 (d, 2H, J ) 8.8 Hz, phenyl-H), 6.82 (d, 2H, J ) 8.8 Hz,
phenyl-H), 4.99 (t, 1H, J ) 5.5 Hz, OH), 3.69-3.60 (m, 2H,
OCH2), 3.38-3.32 (m, 2H, CH2S), 3.00 (s, 6H, 2 × CH3). 13C
NMR δ(DMSO): 167.1, 160.0, 158.3, 151.5, 129.9, 120.0, 116.1,
116.0, 111.3, 93.2, 85.0, 59.6, 39.5, 32.6. MS (ESI): 339.8. Anal.
(C17H17N5OS‚0.2CH2Cl2) C, H, N, S.
Biology. Materials and Methods. [3H]DPCPX and [125I]-
AB-MECA were purchased from Amersham Biosciences (NL).
[3H]ZM 241385 was obtained from Tocris Cookson, Ltd. (UK).
CHO cells expressing the human adenosine A1 receptor were
provided by Dr. Andrea Townsend-Nicholson, University Col-
lege London, UK. HEK 293 cells stably expressing the human
adenosine A2A and A3 receptor were gifts from Dr. Wang
(Biogen) and Dr. K.-N. Klotz (University of Wu¨rzburg, Ger-
many), respectively. CHO cells expressing the human A2B
receptor were provided by Dr. Steve Rees (GlaxoSmithKline,
UK).
All compounds made were tested in radioligand binding
assays to determine their affinities at the human adenosine
A1, A2A, and the A3 receptors, whereas a selection of the
compounds was tested at the human A2B receptor (see text).21
The human A1 receptors were expressed in CHO cells, and [3H]-
DPCPX used as the radioligand. The A2A and A3 receptors were
expressed in HEK 293 cells, and [3H]ZM 241385 and [125I]AB-
MECA were used as the respective radioligands. At the A2B
receptor, radioligand displacement was determined on mem-
branes from CHO cells stably transfected with human A2B
receptor, using [3H]MRS1754 as the radioligand.
Compounds for which the affinities were determined (i.e.,
compounds which showed greater than 50% displacement of
the radioligand at 1 µM) were tested in functional assays for
their ability to influence the levels of cAMP in the test system.
The compounds were tested at concentrations of mostly
100 × Ki, and at least 20 × Ki, where the receptor sites should
be almost fully occupied. The behavior of the compounds was
observed with reference to known adenosine receptor lig-
ands: CPA (a full agonist), DPCPX (an inverse agonist), and
N0840 (reported as a neutral antagonist).
Molecular Modeling. Molecular modeling work was per-
formed with the SPARTAN molecular modeling package
Spartan Pro 1.0.2. Default values in the Merck Force Field
were used in Molecular Mechanics minimizations. The molec-
ular electrostatic potential was calculated using the semiem-
pirical molecular orbital program AM1. The electrostatic
potential was sampled over the entire accessible surface of the
molecules (roughly to a van der Waals contact surface).
Supporting Information Available: A table of the
elemental analyses data. This material is available free of
charge via the Internet at http://pubs.acs.org.
References
(1) Mu¨ller, C. E. A1 Adenosine Receptor Antagonists. Expert Opin.
Ther. Pat. 1997, 7, 419-440.
(2) Hess, S. Recent Advances in Adenosine Receptor Antagonist
Research. Expert Opin. Ther. Pat. 2001, 11, 1533-1561.
(3) Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz, K.-N.;
Linden, J. M. International Union of Pharmacology. XXV.
Nomenclature and Classification of Adenosine Receptors. Phar-
macol. Rev. 2001, 53, 527-552.
(4) Bruns, R. F. Adenosine Receptor Activation in Humand Fibro-
blasts: Nucleoside Agonists and Antagonists. Can. J. Physiol.
Pharmacol. 1980, 58, 673-691.
(5) Soudijn, W.; van Wijngaarden, I.; IJzerman, A. P. Medicinal
Chemistry of Adenosine A1 Receptor Ligands. Curr. Top. Med.
Chem. 2003, 3, 355-367.
(6) van Calenbergh, S.; von Frijtag Drabbe Kuenzel, J. K.; Blaton,
N. M.; Peeters, O. M.; Rozenski, J.; van Aerschot, A.; de Bruyn,
A.; de Keukeleire, D.; IJzerman, A. P.; Herdewijn, P. N6-
Cyclopentyl-3′-substituted-xylofuranosyladenosines: A New Class
of Non-Xanthine Adenosine A1 Receptor Antagonists. J. Med.
Chem. 1997, 40, 3765-3772.
(7) Siddiqi, S. M.; Jacobson, K. A.; Esker, J. L.; Olah, M. E.; Melman,
N.; Tiwari, K. N.; Secrist, J. A., III.; Schneller, S.; Cristalli, G.;
Stiles, G. L.; Johnson, C. R.; IJzerman, A. P. Search for New
Purine- and Ribose-Modified Adenosine Analogues as Selective
Agonists and Antagonists at Adenosine Receptors. J. Med. Chem.
1995, 38, 1174-1188.
CHO cells expressing the human adenosine A1 receptor were
grown overnight as a monolayer in 24-well tissue culture plates
(400 µL/well; 2 × 105 cells/well). cAMP generation was
performed in Dulbecco’s Modified Eagles Medium (DMEM)/
N-2-hydroxyethylpiperazin-N′-2-ethansulfonic acid (HEPES)
buffer (0.60 g HEPES/50 mL DMEM pH 7.4). Each well was
washed twice with HEPES/DMEM buffer (250 µL), and the
following was added: adenosine deaminase (0.8 IU/mL),
rolipram (50 µM), cilostamide (50 µM). This was then incu-
bated for 30 min at 37 °C followed by the introduction of the
compound of interest. After a further 10 min of incubation,
forskolin was added (10 µM). After a subsequent 15 min,
incubation was stopped by aspirating the assay medium and
by adding 200 µL of ice-cold 0.1 M HCl. The amount of cAMP
was determined by competition with [3H]cAMP for protein
kinase A (PKA). Briefly, the sample, approximately 1.8 nM
[3H]cAMP, and 100 µL of PKA solution were incubated on ice
for 2.5 h. The incubations were stopped by rapid dilution with
2 mL of ice-cold Tris HCl buffer (pH 7.4), and bound radioac-
tive material was then recovered by filtration through What-
man GF/C filters. Filters were additionally rinsed with 2 × 2
mL of Tris HCl buffer, and then the radioactivity counted in
Packard Emulsifier Safe scintillation fluid (3.5 mL). All data
reflect three independent experiments performed in duplicate.
(8) Bayer Aktiengesellschaft: WO0125210, 2001.
(9) Bayer Aktiengesellschaft: DE10115922A1, 2002.
(10) Beukers, M. W.; Chang, L. C. W.; von Frijtag Drabbe Ku¨nzel, J.
K.; Mulder-Krieger, T.; Spanjersberg, R. F.; Brussee, J.; IJzer-
man, A. P. New, Non-Adenosine, High Potency Agonists for the
Human Adenosine A2B Receptor with an Improved Selectivity
Profile Compared to the Reference Agonist NECA. J. Med. Chem.
2004, 47, 3707-3709.
(11) Kambe, S.; Saito, K.; Sakurai, A.; Midorikawa, H. Synthetic
Studies Using a-â-Unsaturated Nitriles: Facile Synthesis of
Pyridine Derivatives. Synthesis 1981, 7, 531-533.
(12) Shryock, J. C.; Ozeck, M. J.; Belardinelli, L. Inverse Agonists
and Neutral Antagonists of Recombinant Human A1 Adenosine
Receptors Stably Expressed in Chinese Hamster Ovary Cells.
Mol. Pharmacol. 1998, 53, 886-893.
(13) de Ligt, R. A. F.; van de Klein, P. A. M.; von Frijtag Drabbe
Kuenzel, J. K.; Lorenzen, A.; Ait el Maate, F.; Fujikawa, S.; van
Westhoven, R.; van den Hoven, T.; Brussee, J.; IJzerman, A. P.
Synthesis and Biological Evaluation of Disubstituted N6-Cyclo-
pentyladenine Analogues: The Search for a Neutral Antagonist
with High Affinity for the Adenosine A1 Receptor. Bioorg. Med.
Chem. 2004, 12, 139-149.

Adenosine A1 Receptor Agonist-Antagonist
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 6 2053
(14) Visiers, I.; Ballesteros, J.; Weinstein, H. Computational Methods
for the Construction and Analysis of three-dimensional Repre-
sentations of GPCR Structures and Mechanisms. Methods
Enzymol. 2002, 343, 329-371.
(15) Almaula, N.; Ebersole, B. J.; Zhang, D.; Weinstein, H.; Sealfon,
S. C. Mapping the Binding Site Pocket of the Serotonin 5HT2A
Receptor. J. Biol. Chem. 1996, 271, 14672-14675.
(20) Posner, T. B.; Hall, C. D. 1H and 13C Nuclear Magnetic
Resonance Spectra of Benzylidenemalononitriles: a Method for
the Determination of σ+ Substituent Constants. J. Chem. Soc.,
Perkin Trans. 2 1976, 729-732.
(21) Chang, L. C. W.; Spanjersberg, R. F.; von Frijtag Drabbe Ku¨nzel,
J. K.; Mulder-Krieger, T.; van den Hout, G.; Beukers, M. W.;
Brussee, J.; IJzerman, A. P. Unpublished results.
(16) Ebersole, B. J.; Visiers, I.; Weinstein, H.; Sealfon, S. C. Molecular
Basis of Partial Agonism: Orientation of Indoleamine Ligands
in the Binding Pocket of the Human Serotonin 5-HT2A Receptor
Determines Relative Efficacy. Mol. Pharmacol. 2003, 63, 36-
43.
(17) PC Spartan Pro 1.0.2. Wavefunction, Inc., Irvine, CA92612; 1999.
(18) Taylor, M. D.; Moos, W. H.; Hamilton, H. W.; Szotek, D. S.; Patt,
W. C.; Badger, E. W.; Bristol, J. A.; Bruns, R. F.; Heffner, T. G.;
Mertz, T. E. Ribose-Modified Adenosine Analogues as Adenosine
Receptor Agonists. J. Med. Chem. 1986, 29, 346-353.
(19) Weinberger, M. A.; Heggie, R. M.; Holmes, H. L. Solvent Effects
in the Nuclear Magnetic Resonance Spectra of Benzalmalono-
nitriles. Can. J. Chem. 1965, 43, 2585-2593.
(22) Klotz, K.-N.; Hessling, J.; Hegler, J.; Owman, C.; Kull, B.;
Fredholm, B. B.; Lohse, M. J. Comparative Pharmacology of
Human Adenosine Receptor Subtypes - Characterization of
Stably Transfected Receptors in CHO Cells. Naunyn Schmiede-
berg’s Arch. Pharmacol. 1998, 357, 1-9.
(23) van Tilburg, E. W.; van der Klein, P. A. M.; von Frijtag Drabbe
Kuenzel, J. K.; de Groote, M.; Stannek, C.; Lorenzen, A.;
IJzerman, A. P. 5′-O-Alkyl ethers of N,62-substituted adenosine
derivatives: partial agonists for the adenosine A1 and A3
receptors. J. Med. Chem. 2001, 44, 2966-2975.
JM049597+