Design, synthesis, and structure-activity relationships of 1-,3-,8-, and 9-substituted-9-deazaxanthines at the human A2B adenosine receptor

Article abstract of DOI:10.1021/jm0506221

Over two hundred 1-, 3-, 8-, and 9-substituted-9-deazaxanthines were prepared and evaluated for their binding affinity at the recombinant human adenosine receptors, in particular at the hA_2B and hA_2A subtypes. Several ligands endowed

Full text of DOI:10.1021/jm0506221

282
J. Med. Chem. 2006, 49, 282-299
Design, Synthesis, and Structure-Activity Relationships of 1-,3-,8-, and
9-Substituted-9-deazaxanthines at the Human A2B Adenosine Receptor
Angelo Carotti,*,† Maria Isabel Cadavid,§ Nuria B. Centeno,| Cristina Esteve, Maria Isabel Loza,§ Ana Martinez,‡ Rosa Nieto,‡
Enrique Ravin˜a,‡ Ferran Sanz,| Victor Segarra, Eddy Sotelo,‡ Angela Stefanachi,† and Bernat Vidal
Dipartimento Farmaco-chimico, UniVersita` di Bari, Via Orabona 4, I-70125 Bari, Italy, Departamento de Quimica Organica and
Departamento de Farmacologia, UniVersidade de Santiago de Compostela, E-15782 Santiago de Compostela, Spain,
Unitat de Recerca en Informa`tica Biome`dica (GRIB), IMIM-UPF, C/ Dr. Aiguader 80, E-08003 Barcelona, Spain, and
Almirall Prodesfarma S.A., General Mitre, 151 E-08022 Barcelona, Spain.
ReceiVed June 30, 2005
Over two hundred 1-, 3-, 8-, and 9-substituted-9-deazaxanthines were prepared and evaluated for their binding
affinity at the recombinant human adenosine receptors, in particular at the hA2B and hA2A subtypes. Several
ligands endowed with sub-micromolar to low nanomolar binding affinity at hA2B receptors, good selectivity
over hA2A and hA3, but a relatively poor selectivity over hA1 were obtained. Good antagonistic potencies
and efficacies, with pA2 values close to the corresponding pKis, were observed in functional assays in vitro
performed on a selected series of compounds. 1,3-Dimethyl-8-phenoxy-(N-p-halogenophenyl)-acetamido-
9-deazaxanthine derivatives appeared as the most interesting leads, some of them showing outstanding hA2B
affinities, high selectivity over hA2A and hA3, but low selectivity over hA1. Structure-affinity relationships
suggested that the binding potency at the hA2B receptor was mainly modulated by the steric (lipophilic)
properties of the substituents at positions 1 and 3 and by the electronic and lipophilic characteristics of the
substituents at position 8. A comparison among affinity and selectivity profiles of 9-deazaxanthines with
the corresponding xanthines suggested some possible differences in their binding mode.
Chart 1. Xanthine Derivatives with Antiasthmatic (1 and 2)
and Selective A2B AdoR Antagonistic Activities (3-6)
Introduction
Bronchial asthma is a severe, widespread disease with
inherent variability and unpredictable exacerbations ranging
from mild to fatal.1 Despite the recent considerable advance-
ments in the understanding of the ethiopathogenesis of the
disease,2 the ultimate mechanisms underlying its onset and
development remain elusive, and this precludes a rational
approach for the discovery of new and more efficient therapeutic
agents. Indeed, the therapy of asthma remained substantially
unchanged in the past decades, with the inhaled â2-adrenoceptor
agonists and corticosteroids still constituting the main drugs for
its first line treatment.3 As a second line treatment, leukotriene
receptor antagonists4 and the monoclonal antibody IgE immu-
nomodulator omolizumab5 are occasionaly used. However, in
many patients, the treatment of asthma is not optimal, and this
may impact both morbidity and mortality. There is therefore
an urgent need for alternative, more resolute, and safer drugs
to efficiently cope with this life-threatening and disabling
disease.
therapy of asthma, proved to block selectively the AMP-induced
bronchoconstriction.7 The bronchodilating activity of theophyl-
line 1 and its structural analogue enprofilline 2 (Chart 1) has
been recently attributed to a selective, albeit small, antagonism
at the A2B adenosine receptor (AdoR),7 one of the four distinct
AdoR subtypes (e.g. A1, A2A, A2B, A3) on which adenosine
acts.9,10 These findings prompted several groups to design and
test a huge number of xanthine derivatives in the search for
new, more potent, and A2B-selective ligands.11-17 Weakly
selective A2B antagonists have been initially discovered,8 and
only in the past few years have more potent and A2B-selective
1,3,8-trisubstituted xanthines, such as XAC, MRS1754, and 1,8-
disubstituted xanthines such as 6 (Chart 1), been found.13,18-27
While xanthine derivatives constitute, along with purine
nucleoside analogues and condensed tricyclic nitrogen hetero-
cyclic derivatives,28-32 one of the most exploited class of AdoR
ligands, 9-deazaxanthines [9-dAXs, (1,3-dialkyl-6(7)-(di)sub-
Recently a role of adenosine in asthma, longly advocated,6
has been proved on the basis of key experimental evidence such
as the increase of adenosine concentration in hypoxia and
cellular inflammation in bronchoalveolar fluids of asthmatics
and in plasma (upon a contact with allergenes).7 Moreover,
adenosine (in the form of AMP) induces bronchoconstriction
in asthmatics but not in healthy individuals.8 Finally, theophyl-
line, a dimethylxanthine with a well established role in the
* Corresponding author. Tel.: ++39-080-5442782; fax: ++39-080-
5442230; e-mail: carotti@farmchim.uniba.it.
† Dipartimento Farmaco-chimico, Universita` di Bari.
‡ Departamento de Quimica Organica, Universidade de Santiago de
Compostela.
§ Departamento de Farmacologia, Universidade de Santiago de Com-
postela.
| Unitat de Recerca en Informa`tica Biome`dica (GRIB).-
Almirall Prodesfarma S.A.
10.1021/jm0506221 CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/15/2005

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 283
recent patent.36 For the sake of clarity, we decided to publish
their preparation and biological tests on AdoRs, SAR, and SSR
in three separated articles in this journal.
Chart 2. General Structure and Main Structural Variations of
Designed 9-dAX AdoR Ligands
In the first paper of the series we report on 9-dAXs of
Structure I bearing the structural features indicated in Chart 2.
In the second article, we will describe and discuss the synthesis
and biological activity of cyclic amide derivatives of 9-dAXs
prepared from differently functionalized piperidines, piperazines,
and tetrahydroisoquinolines. Finally in the third paper we will
address quantitative aspects of the structure-affinity relation-
ships (especially at hA2B receptors) through traditional QSAR
(Hansch approach) and 3D QSAR (GRID/CoMFA-GOLPE)
methods, and docking studies of some representative ligands
on AdoR subtype models, developed by means of homology
building techniques.
stituted-1H-pyrrolo[3,2-d]pyrimidine-2,4(3H,5H)-diones)] have
been only rarely studied, especially for their antagonistic activity
at the A2B receptor.33 For this reason, some years ago we began
a systematic, long-term study first addressing an easy, efficient,
and fast synthetic pathway to fully functionalized 9-dAXs and
then a deep investigation of their biological and pharmacological
properties. Our goal was to discover potent (Ki < 30 nM)
antagonists of the human A2B receptor (hA2B) with at least
>100-fold selectivity over hA2A and, possibly, over the other
two AdoR subtypes, namely hA1 and hA3. Preliminary accounts
of this research have been published.34,35
Suitable substituents were introduced in several positions of
the 9-dAX nucleus to properly explore the structure-affinity
(SAR) and structure-selectivity relationships (SSR). The main
structural variations on the 9-dAX nucleus were initially made
at the positions 1 and 3 (with alkyl or functionalized alkyl
substituents), at position 7 (with a simple N-methylation),35 and
at position 8 where mainly para-substituted phenyl groups were
introduced. Then, following the interesting results obtained by
Jacobson et al. on xanthines,8 in the para-position of the 8-phenyl
ring was placed an oxyacetamido group linked through its
nitrogen atom to a variety of aromatic, arylalkyl, heteroaromatic,
and cycloaliphatic substituents. Finally, some halogen substit-
uents were introduced at position 9. The combination of all these
structural modifications led us to prepare more than 400
compounds with the general Structure I, described in part in a
Chemistry. The synthesis of 9-dAX (1,3-dialkyl-8-substituted-
1H-pyrrolo[3,2-d]pyrimidine-2,4(3H,5H)-diones) derivatives was
performed according to published procedures.34,37,38 First, the
appropriate 6-styryluracils IV (Scheme 1) were synthesized by
standard methods and then cyclized to the corresponding
9-dAXs V. 9-dAXs V and their corresponding 9-halogen
derivatives VI bearing an oxyacetic acid or an oxyacetic ester
group (Table 6) on the para position of the 8-phenyl ring were
subsequently converted to amides I by different methods
(Scheme 2, Tables 1-5).
Molecules with chiral centers were isolated and tested as
racemic mixtures. Novel 1,3-dialkyl-5-nitro-6-methyl-
uracils II (compounds 201-203, Table 11 in Supporting
Information) and benzaldehydes III39-43 required for the
synthesis of 6-styryl derivatives IV, were obtained as sum-
marized below.
The 1,3-dialkyl-6-methyluracil precursors of II, were prepared
as follows: symmetrically substituted (R1 ) R3) 1,3-dialkyl-
6-methyluracils were obtained by alkylation of 6-methyl-
Scheme 1a
a (a) Dioxane/piperidine/reflux; (b) P(OEt)3, reflux; (c) Na2S2O4/HCO2H, reflux; (d) SO2Cl2 or Br2/AcOH.
Scheme 2a
a Method A (Z ) Et): NaCN/dioxane, reflux/sealed tube. Method B (Z ) H): isobutyl chloroformate/NMP/THF. Method C (Z ) H): EDC/OHBt/
triethylamine. Method D (Z ) H): EDC/OHBt/polymer bound morpholine/DMF.

284 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
(1H,3H)-pyrimidine-2,4-dione44 or by treatment of the appro-
priate 1,3-disubstituted ureas with acetic anhydride and 4-N,N-
dimethylaminopyridine.45 1-Alkyl-6-methyluracils were prepared
by acetoacetylation of the proper monosubstituted urea followed
by cyclocondensation.46 3-Alkyl-6-methyluracils were readily
accessed by reaction of monosubstituted ureas with ethyl
acetoacetate.47,48 Novel nonsymmetrically substituted (R1 * R3)
1,3-dialkyl-6-methyluracils precursor (compounds 198-199 in
Table 10 and compound 200 in page S3 of Supporting
Information) were prepared by selective alkylation at position
3 or 1 of the corresponding 1 or 3 monosubstituted 6-methyl-
uracil with sodium hydride and alkyl iodide.
Figure 1. Binding competition experiments at cloned hA2B AdoR. A
representative example is shown with compound 68 (9). Nonspecific
binding was measured with 10-3 M NECA (2). The assay was
performed in duplicate.
The synthesis of novel benzaldehydes III was performed by
alkylation of the sodium salts of the corresponding 4-hydroxy-
benzaldehyde with the appropriate R-haloester.40-44
3-Chloro-4-formylphenoxyacetic acid 204 was obtained by
basic hydrolysis of the ethyl (3-chloro-4-formylphenoxy)-
acetate.40 The synthesis of 4-formylphenylamino acetic acid
ethyl ester 205 was performed by alkylation of the alkaline salts
of the 4-aminobenzaldehyde49 with bromoethyl acetate.
The condensation of the 1,3-dialkyl-6-methyl-5-nitrouracils
II33 with the appropriate benzaldehyde III in dry dioxane and
few drops of piperidine as catalyst allowed us to obtain the
corresponding (E)-6-styryl derivatives IV (compounds 206-
216, Scheme 1, Table 12 in Supporting Information) in
satisfactory yields.
Figure 2. cAMP accumulation experiments. An example is shown
with compound 68 at hA2B receptors transfected in CHO cells.
Cumulative concentration-response curves to NECA in the absence
(9) and in the presence (2) of 0.1 µM 68. Each point represents the
mean ( SEM (vertical bars) of triplicate measurements.
The ring closure of IV to the 9-dAXs V (compounds 165-
170, 172, 173-188, and 191-197, Table 6) was performed by
reductive cyclization (Scheme 1) using either neat triethyl
phosphite (Method A) or sodium dithionite in formic acid
(Method B) under reflux. The yields of the cyclization (Table
4 in Supporting Information) were only moderate and highly
dependent on the starting 6-styryluracil. In most reactions,
sodium dithionite proved to be a superior reagent. Interestingly,
in the cyclization with triethyl phosphite, the starting carboxylic
acids were transformed into the corresponding ethyl esters,
whereas the cyclization employing sodium dithionite in formic
acid promoted the oxyacetic ester hydrolysis to the carboxylic
acid. Several carboxylic acids V (Z ) H) were also prepared
through the hydrolysis of the corresponding ethyl esters
(Z ) Et) (Table 6).
Biochemical and Pharmacological Assays. Compounds
were tested for their ability to displace [3H]-DPCPX, [3H]-
ZM241385, [3H]-DPCPX, and [3H]-NECA from cloned human
A1, A2A, A2B, and A3 AdoRs (see Table 16 in Supporting
Information). All the active compounds showed competition
concentration-response curves of specific radioligand binding
against increasing concentrations of ligands, the slopes not being
significantly different from unity at the 5% level of statistical
significance. Assays were carried out by coincubation of
compounds, in at least six different concentrations, with the
appropriate radioactive ligand. A representative binding com-
petition curve is shown in Figure 1.
The antagonistic activity and efficacy of a number of selected
ligands were measured by means of the adenylyl cyclase
functional assay at cloned hA2B and hA2A AdoR subtypes and
in isolated organs at both guinea pig A2B and rat A2A AdoR
subtypes. All the active compounds concentration-dependently
displaced the curves of the NECA AdoR agonist to the right in
a parallel way without depression of their maximum, which is
the typical behavior of competitive antagonists. Representative
examples are shown in Figures 2 and 3 for the experiments on
9-Halogeno derivatives VI (compounds 171 and 173: X )
Cl; 189 and 190: X ) Br) were synthesized by using sulfuryl
chloride and bromine in acetic acid,50 respectively.
Starting from V or VI (compounds 165-197, Table 6),
different methods were applied to prepare amide derivatives I
(compounds 7-164, Scheme 2, Tables 1-5). The synthetic
strategies include the use of both classical synthetic procedures
(Methods A-C) and a method (Method D) adapted to medium
throughput parallel synthesis.36 The first amidation procedure
(Scheme 2, Method A)51 implies the reaction of the 9-dAXs
esters with an excess of the appropriate amine and a catalytic
amount of sodium cyanide in dioxane at reflux or under
appropriate heating in a sealed tube. A second employed method
involved direct condensation of carboxylic acids and the suitable
amines in polar aprotic solvents (DMF, tetrahydrofuran) in the
presence of an organic base (triethylamine or polymer-supported
morpholine) and 1-[3-(dimethylamino)-propyl]-3-ethylcarbodi-
imide hydrochloride (EDC) and 1-hydroxybenzotriazole (OHBt)
(Methods C and D) as coupling reagents. In a third procedure,
the carboxylic acids were converted into a mixed anhydride
intermediates by treatment with isobutyl chloroformate and
N-methylmorpholine (NMP) and then reacted with the proper
amines (Method B).
A2B and A2A AdoR subtypes from transfected cells (hA2B) and
from isolated animal organs, respectively.
The antagonist activity and efficacy of the tested ligands are
collected in Table 9.
Results and Discussion
The most exploited class of 9-dAXs was represented by 1,3-
symmetrically (R1 ) R3) substituted ligands, that are 1,3-
dipropyl and 1,3 dimethyl derivatives (59 and 21 compounds,
reported in Tables 1 and 2, respectively). A further series of
1,3-nonsymmetrically (R1 * R3) substituted 9-dAXs AdoR
ligands (25 compounds) was prepared for a more complete
analysis of the SAR and SSR at the 1,3 positions. As an

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 285
in our hand (% of displacements /IC50 /pKi) made us confident
about their acceptable reliability. From the previous discussion
it should be evident that the study of the SAR mainly addressed
the binding affinity at the hA2B receptor subtype. Occasionally,
some SARs at the hA2A receptor subtype were performed for
particular classes of ligands. In addition to that, the analysis of
the SSRs was also carried out, but at a more qualitative and
general level.
For the sake of clarity, the discussion of the SAR and to a
lesser extent of the SSR is made here separately for each class
of ligands, according to the structural classifications reported
in Tables 1-6. In all the tables, the examined ligands are ordered
on the basis of their decreasing hA2B affinity, considering first
their pKis, then the % of radioligand displacement at 0.1 µM
and finally at 0.01 µM. For a more immediate and efficient
analysis of the variation of both affinity and selectivity, the
binding data (pKi) are presented graphically as a plot of pKi
hA2A (Y axis) versus pKi hA2B (X axis) using the same scale
and range for both axes (square plot). On the diagonal of this
plot (Y ) X) there will lie therefore compounds with equal
affinities at both receptors, whereas hA2B or hA2A selective
compounds will bee seen below or above the diagonal,
respectively. The distance of their pKi values from the diagonal
is a direct measure of their degree of selectivity. The compounds
targeted in the present work, that are those endowed with high
hA2B affinity and high hA2B/hA2A selectivity will be therefore
located in the lower right-hand side corner of the plots.
Whenever possible, a comparative analysis among similarly
substituted ligands belonging to diverse classes is made; this
analysis has been extended also to the comparison between
xanthines and 9-dAXs. Some considerations on the SAR and
SSR are presented also for the ester and carboxylic acid
intermediates reported in Table 6.
Figure 3. Functional assay. An example is shown with compound 68
at A2B receptors in smooth muscle of guinea pig thoracic aorta.
Cumulative concentration-response curves to NECA in the absence
(9) and in the presence (2) of 30 nM 68.
additional structural modification, methyl, ethyl, and, more
rarely, phenyl groups were introduced at the position R to the
carbonyl of the oxyacetamido group linking the phenyl ring in
8-position to phenyl or benzyl groups. It must be underlined
that these structural modifications led to chiral molecules which
were tested as racemic mixtures. The synthesis of selected chiral
9-dAX amides, starting from optically pure enantiomers of
9-dAX acids is now in progress and will be reported in due
course. Symmetric R,R-dimethyl (R2 ) R4 ) CH3, Structure I)
substituted ligands were also prepared. To evaluate the impor-
tance of an intact, condensed pyrrole in the 9-dAX ring system
for the AdoR affinity and selectivity, two additional structural
variations were carried out: the methylation of the N7 nitro-
gen33,35 and the halogenation at the C9 with a chloro or bromo
atom.36 Moreover, a methoxy group was placed in the ortho-
or meta-position of the 8-phenyl ring and in one instance also
a chloro atom was introduced in the ortho-position.
Along with 9-dAX anilides and benzylamides, many het-
eroaromatic amides were synthesized. Many of them were
properly chosen substituted at the aromatic or heteroromatic
amide rings to assess the effects on the affinity and selectivity
of steric, electronic, and lipophilic substituent properties.
However, we have to remind that the first goal of the present
research program was the discovery of new, highly active hA2B
antagonists endowed with good hA2A/hA2B affinity ratios (>100)
and therefore once that compounds with these characteristics
were discovered the affinities for the other two AdoR subtypes,
namely hA1 and hA3, were measured. According to this strategy,
the biological assays were performed up to the determination
of the pKi only for the most promising ligands in the initial
screening. In other words, the binding affinities at hA2B and
hA2A AdoR were first measured as % of radioligand displace-
ment at 0.1 µM concentration (for the hA2B often also at 0.01
µM) and only when they were indicative of a high hA2B AdoRs
affinity and a likely high hA2B/hA2A selectivity, the pKis were
determined at both AdoR subtypes. For this reason, for several
compounds in Tables 1-6, only the % of radioligand displace-
ment at the indicated concentrations is reported. In this regard,
it must be said that for an appropriate evaluation of the SAR
the pKi values were measured also for some informative
compounds that did not conform to the requisites described
above. Finally, to avoid the loss of important information for
the detection of key structural elements responsible for a high
or low binding affinity, also the percentages of radioligand
displacement at the same concentrations were cautiously
compared. They were considered as indicative of a significantly
diverse affinity when the radioligand displacement differed for
more than 15%. Even if those comparisons had to be made with
much precaution, the analysis of a relatively high array of data
1,3-Dipropyl-8-[4-phenoxy-N-phenylacetamido]-9-deaza-
xanthines (compounds 7-65). Nearly 60 1,3-dipropyl-8-
substituted-9-dAXs were prepared and tested for their binding
affinity at hA2A and hA2B receptors (Table 1). pKi values ranging
from 6.14 to 7.92 and from 6.70 to 8.66 were measured for the
hA2A and hA2B AdoR, respectively; their relationship and
distribution are represented as a plot of pKi hA2B vs pKi hA2A
in Figure 4. A simple visual inspection of the plot revealed that
hA2B and hA2A binding affinities are not correlated and that
most of the examined ligands are slightly selective toward the
hA2B receptor subtype, with the exception of compound 38
which is a weakly hA2A selective ligand, lying slightly above
the diagonal of the plot. Indeed, in the whole series, the KihA2A/
KihA2B selectivity ratios were relatively low, reaching the
maximum value of 34.7 with the bulky p-OCH3 phenyl
derivative congener 25, which in the plot of Figure 4 is
represented by the point showing the greatest distance from the
diagonal.
A further analysis of the KihA2A/KihA2B affinity ratios
indicated that values > 10 (∆pKi > 1) were found for eight
compounds (occupying in the plot the region below the line [Y
) X - 1] parallel to the diagonal), bearing in the anilide phenyl
ring substituents with quite diverse steric, lipophilic, and
electronic properties. Therefore, a clear interpretation of the
substituent effects on SSRs cannot be proposed. By contrast,
the analysis of the SARs at the hA2B AdoR allowed the
identification of the key molecular determinants for a favorable
receptor binding. The most active compound at hA2B was the
p-Br-phenyl derivative 7 with a pKi) 8.66 and a moderate
KihA2A/KihA2B affinity ratio (14.1). It is important to note that
five out of the ten most active ligands at hA2B AdoR

286 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
Table 1. Chemical Structures and Binding Affinities at the hA2B and hA2A AdoRs of 1,3-Dipropyl-9-dAX Anilide Ligands 7-65
a
a
a
hA2B
hA2B
hA2A
compd
X
Y
R2/R4
R5/R6
H/C6H4-4Br
H/C6H4-4Cl
H/C6H4-4F
H/C6H4-4F
H/C6H4-4F
H/C6H5
H/C6H4-4OMe
H/C6H4-4Br
H/C6H4-4OCF3
H/C6H4-4CH2SO2NC4H8
H/C6H4-4C(Me)3
H/C6H5
H/C6H4-4F
H/C6H4-4OCH2C6H5
H/C6H5
pKia hA2B
pKi hA2A
Ki A2A/Ki A2B
[0.1 µM]
[0.01 µM]
[0.1 µM]
7
8
9
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H/H
H/H
Me/H
H/H
Et/H
Me/H
Me/H
Et/H
H/H
H/H
H/H
Me/Me
Me/Me
H/H
H/H
Et/H
H/H
H/H
H/H
8.66
8.50
8.49
8.48
8.32
8.29
8.29
8.22
8.16
8.11
8.10
8.10
8.05
8.05
8.03
8.01
7.95
7.92
7.91
7.51
7.50
7.23
7.92
7.72
6.98
7.41
7.50
6.97
7.28
7.10
7.27
7.77
7.73
6.65
7.59
7.40
7.70
6.37
14.1
20
18.2
3.6
4
20.4
7.6
5.2
15.5
6.8
10
6.8
1.9
2.1
24
97
95
88
96
100
92
92
94
91
70
94
93
97
76
92
93
86
80
74
66
58
62
70
68
50
55
55
53
58
51
51
50
49
47
48
34
22
21
54
61
51
62
88
42
57
82
31
31
47
57
32
60
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
H/C6H5
2.6
3.5
1.7
34.7
88
46
52
13
H/C6H4-4CH2CN
H/C6H4-4OMe
H/C6H4CONH(CH2)2-
C6H4-4OMe
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Br
H
H
H
H
Cl
H
H
H
Cl
Cl
Br
H
Cl
Cl
H
H
H
H
Cl
Cl
H
H
Cl
H
H
Cl
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
Ph/H
H/H
H/H
Me/Me
H/H
Ph/H
H/H
H/H
H/H
H/H
H/C6H4-4F
7.90
7.86
7.85
7.79
7.76
7.73
7.72
7.71
7.70
7.64
7.61
7.61
7.60
7.56
7.56
7.48
7.47
7.46
7.46
7.43
7.28
7.25
7.22
7.22
7.20
7.11
7.08
7.04
6.85
6.70
-
7.70
7.20
-
7.16
6.55
7.49
7.49
6.80
-
7.05
6.90
7.37
7.62
7.35
7.39
6.77
6.65
7.25
7.04
7.13
7.07
7.00
6.62
6.61
6.31
-
1.6
4.6
-
4.3
16.2
1.7
1.7
10
72
70
81
80
85
78
74
80
62
80
73
73
78
70
69
56
87
79
67
57
43
57
58
46
67
51
44
52
30
53
89
89
88
87
85
84
79
64
27
17
30
23
27
40
10
38
56
30
24
17
15
16
16
15
10
11
14
24
11
11
8
70
40
44
36
27
50
82
57
15
28
37
54
43
45
54
10
28
51
25
37
21
30
10
20
13
13
23
8
H/C6H4-4NHCOMe
H/C6H4-4Me
H/C6H4-4C6H5
H/C6H4-4SO2C6H5
H/C6H4-4Me
H/C6H4-2OH
H/C6H4-4I
H/C6H4-4COMe
H/C6H4-4Cl
H/C6H4-4Br
-
3.9
5.1
1.7
0.95
1.6
1.5
5.13
6.6
1.6
2.6
2
H/C6H5
H/C6H4-4CONH2
H/C6H4-4OMe
H/C6H4-4F
H/C6H4-4CN
H/C6H4-4CH2COOH
H/C6H4-4OH
H/C6H4-4-Rc
H/C6H5
H/C6H4-4COMe
H/C6H4-4COOEt
H/C6H4-4CH2COOEt
H/C6H4-3F
H/C6H4-4CHO
H/C6H4-4COC6H5
H/C6H4-2Cl
H/C6H4-4N(Me)3
H/C6H4-2F
H/C6H4-4SO2NH2
H/C6H5
H/C6H3-2OH,5F
H/C6H3-2OH,4F
H/C6H4-4Br
H/C6H4-4CF3
H/C6H4-4F
1.6
1.8
4.0
4.1
7.8
-
6
32
6
13
11
5
6.58
6.15
6.14
-
3.2
7.8
5.1
-
4
5
2
10
44
29
23
49
14
32
17
18
16
3
37
75
73
80
54
19
65
31
39
43
35
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
o-Cl
H/C6H4-4Br
H/C6H4-4Br
H/C6H4-4F
H/C6H5
-
-
-
o-OMe
o-OMe
o-OMe
-
-
-
-
-
-
-
-
-
a Binding affinity values represent the mean of at least three independent experiments; the SEM (standard mean error) was always lower than 10%.
b Binding affinity is expressed as % of radioligand displacement at the concentrations indicated in square brackets. c R ) 4,4-dimethyl-4,5-dihydro-1,3-
oxazol-2-yl.

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 287
Figure 5. Affinity-selectivity plot for 1,3-dimethyl-8-substituted-9-
dAX anilide ligands 66-86 (see Table 2).
Figure 4. Affinity-selectivity plot for 1,3-dipropyl-9-dAX anilide
ligands 7-65 (see Table 1).
of ligands carrying electron-withdrawing para-substituents
pointed out that their affinity increases with the increase of their
electron-withdrawing character, regardless of their lipophilic
properties.
As far as the SARs at the hA2A receptor are concerned, it is
worth to note that the three most active ligands are all p-F phenyl
derivatives (10, 11, and 19) and that the drop of affinity observed
at hA2B with the introduction of a halogen in 9-position was
confirmed also at the hA2A AdoR. The effect of the substituents
of the anilino ring on the affinity at hA2A was not well
interpretable as at hA2B.
are derivatives R-monoalkylated at the oxyacetamido bridge.
Some of them, in particular the R-monomethyl derivatives, were
even more active and sometimes more hA2B selective than the
corresponding parent compounds (12 vs 21 and 13 vs 24).
R-Ethyl and R,R-dimethyl derivatives displayed at hA2B an
activity generally similar or lower than the corresponding not
alkylated congeners (11 vs 10, 14 vs 7, 19 vs 10).
A decrease of hA2B activity could be seen also for the
R-phenyl substituted p-F phenyl congeners 61 vs the corre-
sponding unsubstituted derivative 10.
1,3-Dimethyl-8-[4-phenoxy-N-phenylacetamido]-9-deaza-
xanthines (compounds 66-86). A limited number of 1,3-
dimethyl-8-substituted-9-dAXs were prepared and tested for
their binding affinity at hA2A and hA2B AdoRs (Table 2). pKi
values ranging from 5.31 to 7.57 and from 6.67 to 8.81, at the
hA2A and hA2B AdoRs, respectively, were measured. The
absence of any relationship between the two sets of pKis can
be easily seen in the plot of pKi hA2A vs pKi hA2B of Figure 5.
At a first glance it appears that this small series of compounds
The change of affinity coming from the introduction of an
halogen at position 9 was also explored. Irrespective of the type
of halogen used (Cl or Br) and of the parent compound modified,
a significant lowering of the activity (from 0.1 to 1.2 log units)
was observed at both the hA2 AdoR subtypes. An even more
dramatic drop of activity was detected upon the methylation of
the N7 nitrogen atom [data not shown; Franco Fernandez,
University of Santiago de Compostela (personal communica-
tion)]. This last finding is in full agreement with previous reports
by us35 and others.33 Two of the most active compounds, i.e.,
7 and 10, and the lead compound 21 were modified by
introducing a methoxy group at the ortho-position of the
8-phenyl ring. For the p-Br phenyl derivative 7, a chloro
substituent was also introduced at the ortho position of the
8-phenyl ring (compound 62). In all instances, and more
evidently in the o-OCH3 congeners 64 and 65, a strong decrease
in activity was recorded.
contains a relatively high number of ligands with a KihA2A
/
KihA2B affinity ratio > 10. They can be easily located in the
region of the plot below the line [Y ) X - 1] parallel to the
diagonal. In the same region, in the right-hand side low corner,
lies the most potent and selective hA2B/hA2A compound reported
in this work, compound 69. Above the diagonal lies instead
compound 81, a somewhat hA2A selective ligand (3-fold over
hA2B) of this series.
The analysis of the chemical structures in Table 2 indicated
that only few compounds of this series share a common
substitution pattern with compounds of the 1,3-dipropyl series,
and therefore no direct comparison among affinity and selectiv-
ity data of the two series could be made. Very close activities
at hA2B AdoR were observed in the two series for the
unsubstituted anilides 21 and 75, for the p-F-phenyl derivatives
10 and 68 and for the p-Br-phenyl derivatives 7 and 67. By
contrast, the hA2B receptor affinities of the p-COCH3- (70) and
p-CN-phenyl- (71) 1,3-dimethyl substituted ligands were sig-
nificantly greater than the corresponding 1,3-dipropyl congeners
34 and 41 (8.45 vs 7.70 and 8.42 vs 7.48, respectively).
Analogous to the 1,3-dipropyl series, the monomethylation
at the R-carbon atom of the oxyacetanilide moiety led to higher
The substituent effects on the binding affinity at hA2B and
hA2A AdoRs were thoroughly explored at the phenyl ring of
the aniline moiety, in particular at the para-position. The
substituents were chosen in order to properly sample the
physicochemical domain of their steric, lipophilic, and electronic
properties. An initial analysis of the SARs at the hA2B receptor
suggested that the affinity increases as the lipophilicity of the
para-substituents increase (compare the lead compound 21 with
15, 14, 10, 8, and 7). Lipophilic but bulky substituents such as
t.But, Ph, I, and OBn (17, 29, 33, and 20, respectively) did not
produce the expected strong enhancement of affinity, and this
might imply that when the substituent overcomes a certain size,
a negative steric effect, somehow hindering a favorable hydro-
phobic interactions, may take place. Finally, a careful analysis

288 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
Table 2. Chemical Structures and Binding Affinities at the hA2B and hA2A AdoRs of 1,3-Dimethyl-9-dAX Anilide Ligands 66-86
b
b
b
hA2B
hA2B
hA2A
compd
X
Y
R2/R4
R5/R6
pKia hA2B pKia hA2A Ki A2A/Ki A2B [0.1 µM] [0.01 µM] [0.1 µM] pKia,c A1
pKia,c A3
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
Me/H
H/H
H/H
H/C6H4-4Br
H/C6H4-4Br
H/C6H4-4F
H/C6H4-4Br
H/C6H4-4Ac
H/C6H4-4CN
H/C6H4-4F
H/C6H5
H/C6H4-4NO2
H/C6H5
H/indan-1yl
H/C6H4-4F
H/C6H4-4Br
H/C6H5
H/-c.C5H11
H/C6H4-4F
H/C6H5
H/CH2CH2NH2
H/C6H4-4CN
Me/C6H5
8.81
8.58
8.52
8.46
8.45
8.42
8.41
8.32
8.16
7.96
7.65
7.60
7.41
7.16
7.10
6.84
6.83
6.73
6.67
7.57
5.91
7.19
5.31
-
7.34
17.4
741
21.3
1412.5
-
12
-
10
11.7
70.8
9.5
3.8
2.1
15.1
8.7
0.3
2
93
90
100
83
91
91
99
99
82
91
74
63
78
55
57
45
27
34
28
61
47
89
84
72
60
65
60
76
73
52
44
30
23
15
12
20
10
10
7
77
12
41
20
36
47
70
60
30
10
20
31
32
13
8
7.74
7.83
7.49
-
8.37
-
8.11
-
6.62
8% [1 µM]
20% [1 µM]
o-OMe H/H
4.97
-
H
H
H
H
H
H
H
H/H
H/H
Me/H
Me/H
H/H
-
-
-
7.32
7.09
6.11
6.67
7.02
7.09
5.98
6.16
7.34
6.52
6.78
6.24
6.04
-
H/H
H/H
-
-
-
o-OMe H/H
m-OMe H/H
o-OMe H/H
-
-
-
-
-
-
-
-
-
-
-
-
-
H
H/H
-
m-OMe H/H
m-OMe H/H
35
20
38
5
0
0
-
-
H
H/H
m-OMe H/H
H/H
m-OMe H/H
0.9
2.7
-
-
10
15
10
-
H
-
-
-
-
-
H/C6H4-4COOEt
-
-
a Binding affinity values represent the mean of at least three independent experiments; the SEM was always lower than 10% b The binding affinity is
expressed as % of radioligand displacement at the concentrations indicated in square brackets. c Binding affinity expressed as pKi or % of ligand displacement
at the concentrations indicated in square brackets.
active compounds at hA2B (i.e. compounds 66 > 67 and 73 >
75), the p-Br-phenyl derivative 66 (pKi ) 8.81) being the most
active of the whole series. However, unexpectedly, the alpha
methylation of the p-F derivative 68 afforded compound 72 with
a slightly lower affinity.
The N-methylation of the anilide nitrogen of the lead
compound 75 afforded compound 85 with a decreased hA2B
affinity (61% of radioligand displacement compared to 91%,
at 0.1 µM).
Electron-withdrawing groups (i.e., NO2, F, Br, CN, and
COCH3) in the para-position of the anilide phenyl ring increased
the activity of the parent compound 75, whereas in the series
of R-methylated ligands and in the ortho and meta substituted
derivatives at the 8-phenyl ring, the enhancement of the binding
affinity might be attributed to an increased lipophilicity (66 >
72 > 73; 69 > 77 > 79; 78 > 81 g 82).
Regarding the hA2B/hA2Aselectivity, the most interesting
result came from the introduction of a methoxy substituent in
the ortho-position of the 8-phenyl ring. While in the parent
compound 75 and its p-F phenyl derivative 68, endowed with
quite high KihA2A/KihA2B affinity ratios (71 and 21, respec-
tively), this structural modification determined a marked
decrease of both affinity and selectivity (see compounds 79 and
78), in the p-Br phenyl congener 67 the high hA2B activity was
maintained whereas an impressive decrease of the hA2A affinity
was observed. This ultimately led to a dramatic enhancement
of selectivity in compound 69 which was 1412-fold more hA2B
over hA2A selective, one of the highest KihA2A/KihA2B selectivity
values reported so far.
considered a high affinity ligand at A2B receptor (Ki ) 3.5 nM)
with an excellent selectivity over hA2A (KihA2A/KihA2B ) 1412)
and hA3 (KihA3/KihA2B ) 3090). Unfortunately, the selectivity
versus the hA1 was low (KihA1/KihA2B ) 9.3), and this calls
for further structural modifications aimed at improving this
figure while maintaining the excellent selectivities over the other
AdoR subtypes. The binding affinities at hA1 and hA3 were
measured also for the second most selective hA2B/hA2A ligand
67 and for compound 73, which is the R-methyl derivative of
the lead compound 75. The selectivity pattern observed for 69
was confirmed: the hA2B/hA1 selectivities remained quite poor,
whereas good hA2B/hA3 selectivities were detected. The hA2B/
hA1 selectivity was determined also for the highly hA2B active
p-CN phenyl derivative 71; in this case an even lower hA2B/
hA1 selectivity was measured (Ki ratio close to 1).
Differently from 66, the introduction of a methoxy group in
the meta-position of the 8-phenyl ring of the p-F phenyl
derivative 68 induced a marked decrease in hA2B affinity leading
to compound 81, one of the few slightly hA2A selective ligands
obtained in this work.
Compounds 76, 80, and 83, which are not anilides but
differently functionalized N-alkyl oxyacetamides, were inserted
in Table 2 (1,3-dimethyl-9-dAX derivatives). The indanyl (76)
and cyclopentyl (80) amides showed good affinities at hA2B
AdoR with a nearly 10-fold selectivity over hA2A AdoR subtype.
Compound 83 was prepared for comparison to its corresponding
xanthine derivative, which had been tested several years ago52
at the rat A2 AdoR receptor (at that time the existence of the
A2A and A2B receptor subtypes was unknown) and showed a
high affinity (pKi 7.40). Our 9-dAX derivative 83 displays
similar affinities at hA2B and hA2A AdoRs (pKi 6.73 and 6.78,
respectively), two values likely lower than in the corresponding
xanthine.
This outstanding outcome prompts us to extend the binding
assays on 69 also to the other two AdoR subtypes hA1 and hA3.
pKi values of 7.49 and 4.97 for the hA1 and hA3 AdoR,
respectively, were measured. Therefore, compound 69 could be

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 289
Table 3. Chemical Structures and Binding Affinities at the hA2B and hA2A AdoRs of 1,3-Dialkyl of 9-dAX Anilide Ligands 87-111
a Binding affinity values represent the mean of at least three independent experiments; the SEM was always lower than 10%. b Binding affinity is expressed
as % of radioligand displacement at the concentrations indicated in square brackets. c pKi hA1 ) 7.96; pKi hA3 ) 4.91.
1,3-Dialkyl-8-(phen-4-oxy-N-phenylacetamido)-9-deaza-
xanthines (compounds 87-111). Along with 1,3-dimethyl- and
1,3-dipropyl-9-dAXs, a number of 1,3-symmetrically (R1 ) R3)
and nonsymmetrically (R1 * R3) alkyl substituted congeners
were designed, prepared, and tested. Their chemical structures
and binding affinities at hA2A and hA2B AdoRs are reported in
Table 3. The 1,3-symmetrically substituted 9-dAXs comprised
ethyl, cyclopropylmethyl, trifluoroethyl, and 3-methoxypropyl
derivatives whereas the nonsymmetrically substituted ones
included 1-methyl, 3-propyl, 1-methyl-3-(3-morpholin-4-propyl),
1-isobutyl-3-methyl, and 1-propyl-3-methyl derivatives. pKis
were measured only for those ligands whose % of radioligand
displacement at hA2B and hA2A AdoR subtypes could have been
predictive of a high hA2B affinity and/or high KihA2A/ KihA2B

290 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
Figure 6. Affinity-selectivity plot for 1,3-dialkyl-9-dAX anilide
ligands 87-111 (see Table 3).
Figure 7. Affinity-selectivity plot for 1,3-dialkyl-9dAX benzylamide
ligands 112-131 (see Table 4).
affinity ratio. Several ligands with good hA2B affinity (pKi up
to 8.33) and high KihA2A/KihA2B affinity ratios were obtained
as can be seen in the plot of Figure 6.
The plot of pKis in Figure 6 suggested that also for this class
of AdoR ligands there was no correlation between hA2A and
hA2B affinities and that two compounds, occupying the lower
right-hand corner of the plot, had a good hA2B affinity and high
hA2B/hA2A selectivity. These compounds are both 1-propyl-3-
The only N-alkylamide examined, i.e., the cyclopentylamide
99, was a 1-methyl-3-methoxypropyl derivative. Its affinity at
hA2B (pKi ) 6.58) and Ki hA2A/Ki hA2B affinity ratio (1.6) were
much lower than the anilide 96 (corresponding values were 7.86
and 8.7, respectively) despite their close lipophilicity.
A comparison of the hA2B affinity data among the 1,3-
disubstituted parent compounds reported in Tables 1-3 revealed
the following decreasing order of affinity at hA2B AdoR: 1,3-
dicyclopropylmethyl (87) > 1,3-diethyl (91) > 1,3-dipropyl
(21)> 1,3-dimethyl (75)> 1-propyl-3-methyl (94) >1-methyl-
3-(3-methoxypropyl) (96) > 1-methyl-3-propyl (97) >1,3-di-
3-methoxypropyl (107) > 1,3-ditrifluoroethyl (108), and 1-iso-
butyl-3-methyl (109). This rank of affinity should help the
design of high affinity hA2B ligands.
1,3-Dialkyl 8-[phen-4-oxy-(N-benzyl)-acetamido]-9-deaza-
xanthines (compounds 112-131). Twenty 1,3-diakyl benzyl-
amide derivatives of 9-dAX were designed, prepared, and
evaluated for their affinity at hA2B and hA2A receptors. Ligands
with good hA2B affinity (pKi up to 8.50) and some hA2B/hA2A
selectivity were obtained (Table 4).
The plot of pKi hA2B vs pKi hA2A in Figure 7 demonstrates
that no correlation exist between the two sets of pKis and that
a number of ligands displayed a Ki hA2A/Ki hA2B affinity ratio
nearly to or higher than 10. Considerably lower affinities at
hA2B receptors were generally observed in the benzylamide
series compared to the corresponding anilide analogues.
An exception to that was the 1,3-dipropyl lead compound
112 whose affinity at hA2B AdoR subtype was consistently
higher than the corresponding anilide derivative 21 (pKi hA2B
8.50 vs 8.03). Notably, the 1,3-dipropyl lead compound 112
presented the highest hA2B affinity and hA2B/hA2A selectivity
of the whole series of benzylamides.
methyl congeners: the p-biphenyl derivative 88 (pKi hA2B
8.28; KihA2A/KihA2B ratio ) 501) and the p-COOEt derivative
93 (pKi hA2B ) 7.93; KihA2A/KihA2B ratio ) 224).
)
From a structural point of view, this high selectivity stems
from the presence of a bulky group in the para-position of the
8-phenyl ring which seems to be well accepted into the hA2B
receptor binding site but not tolerated into the binding regions
of hA2A AdoR. Indeed, the observed high selectivity of ligands
88, 93 mainly originates from their very low affinity toward
the hA2A AdoR subtype. A dramatic drop of hA2A affinity was
not observed in the most thoroughly explored class of 1,3-
dipropyl derivatives bearing the same, or even bulkier, substit-
uents in the para position. This might suggest that a strong steric
interaction takes place only at the hA2A AdoR binding sites
likely because a quite specific binding of the propyl group in
the position 3 and/or the methyl group in position 1 would force
the bulky para substituents in a forbidden region. Evidently,
this does not apply to the two propyl groups at the positions 1,
3 that, most likely, might be anchored in such a way that the
para bulky substituents of the 8-phenyl ring are not placed in
sterically hindered receptor regions. This hypothesis could be
supported by a docking study of suitably selected para-
substituted 1,3-dialkyl ligands into the antagonist binding site
of the hA2A AdoR subtype, built by homology modeling.
The excellent hA2B affinity and KihA2A/ KihA2B selectivity
of compound 88 prompted us to measure its affinity also at the
other two AdoR subtypes hA1 and hA3. While for the latter a
very low affinity was detected (pKi )5.91), for the former a
good affinity was observed (pKi ) 7.96). Similarly to compound
68, 88 also presented excellent hA2B affinity, high hA2B/hA3
selectivity but a still too low selectivity ratio (KihA2A/KihA2B
) 2.1).
Within the series, the hA2B binding affinity of the three lead
compounds, that are the 1,3-dipropyl (112), 1,3-dimethyl (119),
and 1-methyl-3-propyl (131) benzylamide derivatives, was
ranked as follows: 112 > 119 . 131. As in the most explored
series of 1,3-dimethyl derivatives, the binding affinity strongly
increased as the lipophilic character of the para-substituent in
the benzyl ring increases (114 >115 > 119). The N-alkylation
of the benzylamino group did not change significantly the hA2B

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 291
Table 4. Chemical Structures and Binding Affinity at the hA2B and hA2A AdoRs of 9-dAX Benzylamide Ligands 112-131
a
a
a
hA2B
hA2B
hA2A
compd
R1/R3
X
Y
R2/R4
R5/R6
pKia hA2B
pKia hA2A
Ki A2A/Ki A2B
[0.1 µM]
[0.01 µM]
[0.1 µM]
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
Pr/Pr
Pr/Pr
H
H
H
H
H
H
H
H
H
H
Cl
H
Cl
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
o-OMe
H/H
Me/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/H
H/CH2C6H5
H/CH2C6H4-4Cl
H/CH2C6H4-4Cl
H/CH2C6H4-4F
H/CH2C6H4-4Cl
H/CH(Me)C6H5
Et/CH2C6H5
H/CH2C6H5
H/CH2C6H4-3OMe
Me/CH2C6H5
8.50
8.39
8.31
7.77
7.69
7.56
7.37
7.29
7.29
7.26
7.08
6.91
6.76
6.73
-
7.05
7.33
7.69
7.36
-
56.2
11.5
4.2
2.6
-
95
94
93
95
76
63
59
66
61
67
47
46
25
43
77
74
58
45
32
21
78
55
61
36
33
20
15
12
15
10
10
11
8
38
57
75
56
20
13
17
25
11
14
43
23
44
23
53
75
39
3
Me/Me
Me/Me
Pr/Me
Me/Me
Me/Me
Me/Me
Me/Me
Me/Me
Pr/Pr
Me/Me
Pr/Pr
Me/Me
Pr/Pr
6.61
6.05
6.74
6.48
5.99
7.22
6.73
7.31
6.80
-
8.9
20.9
3.6
6.5
18.6
0.7
1.5
0.3
0.7
-
H/CH2C6H5
H/CH2C6H4-4Cl
H/CH2C6H4-4F
H/CH2C6H4-4Cl
H/CH(CH2OH)C6H5
H/CH(CH2Cl)C6H5
H/CH(COOMe)C6H5
H/CH(COOH)C6H5
Me/CH2C6H5
H
m-OMe
5
H
H
H
H
H
H
30
25
29
38
15
10
Pr/Pr
Pr/Pr
Pr/Pr
Me/Pr
Me/Pr
-
-
-
-
-
-
-
-
-
-
-
-
15
19
H/CH2C6H5
-
-
-
aBinding affinity values represent the mean of at least three independent experiments; the SEM was always lower than 10%. b The binding affinity is
expressed as % of radioligand displacement at the concentrations indicated in square brackets.
affinity (119 vs 118 and 121) but the KihA2A/KihA2B affinity
ratio was enhanced, leading to the two highest hA2B selective
compounds of the series (118, 20.9, and 121, 18.6). Methylation
of the benzylic carbon of the 1,3-dimethyl lead compound 119
afforded 117 which displayed an enhanced hA2B affinity (pKi
7.56 vs 7.29). Finally, the introduction of a methoxy group either
in the ortho- or meta-position of the 8-phenyl ring of 119
afforded less active ligands (compare 114 vs 123 and 125).
Replacing the phenyl ring of the benzyl group in the lead
compound 119 with the isosteric 4-pyridyl group (see Table 5)
led to a likely less active compound 159 (compare the % of
displacement at 0.1 µM). A similar behavior was observed for
the 1,3-dipropyl lead compound 112: its 4-pyridyl isoster 143
(see Table 5) shows lower hA2B affinity (7.55 vs 8.50).
Two 1,3-dipropyl-9-chloro derivatives (122 and 124) were
also prepared. Along with the m-methoxy congener 125 they
presented a slightly inverted selectivity (KihA2A/KihA2B ratio
< 1). All these three derivatives can be easily detected in the
plot of Figure 7 as the points lying above the diagonal. As
already observed for other classes of dAXs, the introduction of
an halogen in position 9 determined a dramatic decrease of the
hA2B activity: pKi 8.50 vs 7.08 for 112 and 122, respectively,
Miscellaneous 9-dAX Derivatives (compounds 132-164).
A wide range of AdoR ligands with some peculiar structural
features were also synthesized and tested (Table 5). They
comprised mainly amides of heterocyclic amines (20 com-
pounds), heterocyclic isosteres of benzylamides (143, 145, 159,
and 161), and finally N-formylanilino (133, 148, and 162) and
uracil carbamic acid (141 and 146) derivatives. The majority
of these compounds should have improved water solubility and,
hopefully, a better bioavailability compared to the corresponding
anilides and benzylamides.
Figure 8. Affinity-selectivity plot for the miscellaneous 9-dAX
derivatives 132-164 (see Table 5).
and hA2B AdoR subtype. Therefore, the SAR and SSR analyses
are less robust than those coming from the classes of ligands
examined in previous chapter.
The plot in Figure 8 shows the hA2A and hA2B pKi variation
and distribution and indicates that no relationship exists between
the two sets of affinity data. Moreover, only two compounds
(i.e. 132 and 136) have a >20-fold selectivity over the hA2B
receptor subtype. Two ligands, which lie just above the diagonal
of the plot have an inverted selectivity (KihA2A/KihA2B affinity
ratio < 1). Interestingly, they are both carbamic acid derivatives
(141 and 146).
For the reasons discussed in a preceding chapter, for most of
these AdoR ligands only the % of radioligand displacements at
the two indicated concentrations was determined at both hA2A
One of the most interesting compound in Table 5 was the
1-propyl-3-methyl derivative 132 which presented a 6-OCH3-

292 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
Table 5. Chemical Structures and Binding Affinities at the hA2B and hA2A AdoRs of 9-dAX Ligands 132-164

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 293
Table 5 (Continued)
a Binding affinity values represent the mean of at least three independent experiments with a SEM always lower than 10%. b The binding affinity is
expressed as % of radioligand displacement at the concentrations indicated in square brackets.
3-pyridyl substituent as a replacement of the methoxy phenyl
group. Ligand 132 shows high hA2B affinity (pKi) 8.50) and a
good hA2B selectivity (42-fold over hA2A). The corresponding
1,3-dipropyl analogue 134 displayed both lower hA2B affinity
(pKi ) 7.97) and selectivity (8-fold over hA2A). Comparing the
% of displacements at 0.1 µmol of the simple 1,3-dipropyl-3-
pyridyl derivative 149, at hA2B and hA2A AdoR subtypes, an
hA2B affinity very close to that determined for the anilidic
analogue 21 (83 vs 92%) was observed, whereas the hA2A
affinity was definitely higher (48% vs 6%).
For the series of ligands, which are heterocyclic isosteres of
anilides or benzylamides, we ranked their binding affinities and,
whenever possible, made a direct affinity comparison with the
corresponding anilides and benzylamides. The affinities at both
hA2A and hA2B AdoR subtypes of the unsubstituted 2- (142)
and 3-pyridyl (149) derivatives were quite close whereas
significant differences were observed for substituted pyridyl
congeners. The highest hA2B affinities were displayed by the
6-methoxy and 6-methyl-3-pyridyl derivatives (134 and 136,
respectively) whereas a close (hA2B) and a lower (hA2A) activity
emerged considering the % of radioligand displacements of the
4- and 6-methyl-2-pyridyl isomeric derivatives 150 and 153,
respectively.
When the hA2B affinities of the heterocyclic amides were
compared with the corresponding isosteric anilides or benzyl-
amides, close or higher affinities were always seen for the latter,
i.e., 24 = 134; 28 = 136, 32 = 140; 21 > 142, 149, 112 >143,
119 > 145; 21 > 156.
The 1,3-dipropyl-2′,6′-dimethoxy-4′-pyrimidinylamide 135
was endowed with a biological activity nearly identical to the

294 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
6′-methoxy-3′-pyridyl congener 134. The corresponding 1,3-
dimethyl analogue of the latter (160) showed a lower affinity
at both receptor subtypes according to its % of radioligand
displacement at 0.1 µM. The same results came from the
comparison of the 1,3-dipropyl and 1,3-dimethyl 2′-pyrazinyl
amides 156 and 163, respectively. The 1,3-dimethyl-benzimi-
dazolyl derivative 139 and the 1,3-dipropyl-3′amino-pyrazolyl
derivative 138 showed comparable hA2B affinities, relatively
close to those of the parent anilidic compounds 75 and 21.
The unexpected formylation of the p-amino substituent of
the 8-phenyl ring during the cyclization reaction of some
nitrostyryluracil intermediates with sodium dithionite in formic
acid afforded compounds 148 and 162, whose affinities at hA2B
AdoR were close to or lower, for the second, than the affinities
observed for the corresponding nonformylated analogues 10 and
21, respectively. The comparison of the hA2A affinity indicated
higher % of radioligand displacement for the formylated
derivative, and therefore a lower hA2B/hA2A selectivity could
be hypothesized. The other unexpectedly obtained uracil car-
bamic acid derivatives 141 and 146, that differ at the N′3
substituent (H and CH3, respectively) and are ketonic and not
amidic derivatives, showed interesting hA2A and hA2B affinities
and an inverted selectivity profile, resulting, respectively, in 5-
and 6-fold more hA2A selectivity over the hA2B AdoR.
Finally, 1,3-dimethylanilino derivatives 144 and 147, meta
positional isomers of the para-subsituted congeners 67 and 75,
respectively, were prepared to evaluate the influence of the
anchoring position of the oxyacetamido chain on the affinity
and selectivity.
Lower hA2B affinities were observed for meta positional
isomers 144 and 147 (7.35 vs 8.58 and 6.78 vs 7.96, respec-
tively) which showed also much lower KihA2A/KihA2B affinity
ratio (1.5 vs 74; 1.2 vs 71 respectively)
Compound 164, which is a 2-D-glucosamide derivative, was
prepared, aiming at a drastically improvement of the water
solubility. Unfortunately, very low affinities resulted at both
receptor subtypes.
Acid and Ester Intermediates (compounds 165-197). A
number of acid and ester intermediates (Structures V and VI,
Scheme 1) prepared for the synthesis of 9-dAXs were tested
for their binding affinity at hA2A and hA2B AdoR subtypes. The
results are listed in Table 6 and reported in graphical form in
Figure 9.
From the plot in Figure 9, it is evident that no correlation
exists between hA2A and hA2B affinities and that in line with
our previous observations, acid and ester intermediates also
displayed some selectivity for the hA2B receptor subtypes. The
most hA2B active ligand was the 1,3-dipropyl derivative 165
whereas the most selective was 166 with a KihA2A/KihA2B
affinity ratio equal to 93.3. Good selectivities were also
displayed by ligands 174 and 172. Notably the most hA2B
selective ligand 166 also showed one of the highest binding
affinity at the same receptor (pKi ) 7.36).
Figure 9. Affinity-selectivity plot for acid and ester intermediates
165-197 (see Table 6).
For only five ethyl esters was the pKi at both hA2 AdoR
receptor subtypes measured. The most active was the 1-ethyl-
3-(3-methoxypropyl) derivative 169 (pKi ) 7.18).
A comparison among acids and the corresponding ethyl esters
always indicated a higher hA2B affinity for the former (i.e. 172
> 176; 174 > 177; 175 > 178, 179 >180).
Comparison of AdoR Binding Affinities and Selectivities
of 9-dAXs with Xanthines. To gain more general insights on
the SAR and SSR of AdoR ligands, the affinity and selectivity
data of the 9-dAXs reported herein were compared with
literature data8,18 of xanthine analogues having the same
substitution pattern. It must be noted that such a comparison
has to be made with caution since the experimental methods
used to measure the binding affinity at the different AdoR
subtypes were not exactly the same. Indeed, some significant
discrepancies between our and Jacobson’s data were observed
for some xanthine derivatives, reprepared by us as reference
compounds (see footnotes in Table 7).
Only a relatively low number of 9-dAXs showed the same
substitution pattern of already published xanthines. Therefore,
only the AdoR binding affinities of compounds 7, 8, 9, 21, 28,
33, 41, 112, and 165 could be compared with those reported
for their corresponding xanthines.18 As far as the binding
affinities of xanthines and 9-dAXs at hA2B is concerned, no
clear relationship emerged whereas a higher hA2B over hA2A
selectivity was almost always observed for xanthines. These
results were not confirmed, taking into account the data
determined in our assay (see hA2B and hA2A affinity data
reported in footnote c of Table 7).
It must be reminded that the heterocyclic systems of xanthines
and 9-dAXs, formally differing from the presence-absence of
a nitrogen atom at the 9-position, may present greatly diverse
physicochemical properties especially in terms of electronic
characteristics, ability to make hydrogen bonds, water solubility,
and partitions in polar/apolar immiscible solvents.35 These
significant differences might result, as our data seem to suggest,
in diverse pharmacodynamic (and pharmacokinetic) profiles
even when xanthines and 9-dAX derivatives with closely related
structures are compared.
The observed rank of binding affinities of the acid lead
intermediates, for which the pKi on hA2B were measured, was
as follows:
1,3-dipropyl (165) > 1-propyl-3-methyl (166) ) 1-methyl-
3-propyl (167) > 1-ethyl-3-(3-methoxypropyl) (170) > 1,3-
dimethyl (172) > 1-methyl-3-(3-methoxypropyl) (174).
The introduction of a methoxy group in either the ortho- or
meta-position of the 8-phenyl ring yielded a lower affinity
(compare 172 vs 175 and 179). The introduction of bromo
substituent at position 9 in 165 led to a significant diminution
of affinity (see compounds 171 vs 172).
The AdoR binding affinities of some 9-dAXs were also
compared to those recently published11 for corresponding
8-pyrazolyl, -pyridazinyl, and -pyridyl derivatives (Table 8).

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 295
Table 6. Chemical Structures and Binding Affinity at the hA2B and hA2A AdoRs of Acid and Ester Intermediates 165-197
a Binding affinity values represent the mean of at least three independent experiments with a SEM always lower than 10%. b The binding affinity is
expressed as % of radioligand displacement at the concentrations indicated in square brackets.

296 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
Table 7. Chemical Structures and binding affinities at the hAdoR of 9-dAX and the Corresponding Xanthine Derivatives
compd
R1/R3
X
Z
pKia,b hA2B
pKia,b hA2A
Ki hA2A/Ki hA2B
pKib hA1
pKib hA3
7
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
Pr/Pr
CH
N
CH
N
CH
N
CH
N
CH
N
CH
N
CH
N
CH
N
NH-C6H4-4Br
NH-C6H4-4Br
NH-C6H4-4Cl
NH-C6H4-4Cl
NH-C6H4-4F
NH-C6H4-4F
NH-C6H5
8.66
8.62c
8.80
8.10
8.48
8.65d
8.03
8.82e
7.85
8.72
7.70
8.67
7.48
8.70
8.50
8.50
7.80
7.40f
7.51
5.78c
7.50
6.72
7.92
7.78d
6.65
8.65e
-
[6.89]r
6.80
5.23
6.77
6.29
7.05
7.70
6.91
[5.65]rf
14.1
700
20
400
3.6
7.5
24
-
7.13
-
-
5.63
-
8
7.30
-
5.73
-
9
7.74
-
-
-
21
28
33
41
112
165
NH-C6H5
17
7.39
-
6.86
-
NH-C6H4-4Me
NH-C6H4-4Me
NH-C6H4-4I
NH-C6H4-4I
NH-C6H4-4CN
NH-C6H4-4CN
NH-CH2C6H5
NH-CH2C6H5
OH
-
-
-
-
-
-
10
2400
5.1
260
56
16
8
6.53
-
5.90
-
6.39
6.24
7.20
-
[7.2]r
6.85
-
5.40
CH
N
OH
56.2
a Assay performed on rat AdoR are indicated by the underscript r, close to the pKi affinity values reported in square brackets. b Binding affinity values
represent the mean of at least three independent experiments with a SEM always lower than 10%; pKi hA2B and pKi hA2A determined by us, respectively.
c 8.09 and 6.99. d 8.23 and 7.03. e 8.04 and 6.94. f 7.53 and 6.07. r ) rat.
Table 8. Chemical Structures and Binding Affinities (pKi) at hA2A and hA2B AdoRs of Various Sets of Xanthines and 9-dAX Derivatives
a Binding affinity values represent the mean of at least three independent experiments; the SEM (standard mean error) was always lower than 10%.
Comparable affinities at hA2B were obtained for most of the
examined 9-dAXs and 1-methyl pyrazolyl derivatives with the
noticeable exception of the 4-F and 4-Br phenyl derivatives
which were significantly more active in the 9-dAX series. The
lack of pKi data at hA2A in Baraldi’s paper11 did not allow a
precise comparison with 9-dAXs at this AdoR subtype. How-
ever, 9-dAXs seem to have higher hA2A affinities than xanthines.
On the whole, the above observations might suggest slightly
different binding modes for xanthines and 9-dAXs at hA2
receptor subtypes.
Antagonist Activity and Efficacy of Selected 9-dAX AdoR
Ligands at hA2A and hA2B AdoR Subtypes. A number of
ligands (67-69, 71, 75, and 88, Table 9) were selected to
measure their antagonistic activity and efficacy in the adenyl
cyclase functional assay at cloned hA2B and hA2A AdoR
subtypes. A representative example of this experiment was
shown in the plot of Figure 2.
The pA2 values at both receptor subtypes are reported in Table
9 along with the corresponding pKi values from binding assays.
pA2 values very close to the corresponding pKis were obtained.

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 297
Table 9. Binding affinities, antagonist potency and efficacy at A2B and A2A receptors of selected 9-dAXs (67-69, 71, 75, and 88)
a
b
a
c
compd
pKia hA2B
pKia hA2A
pA2a hA2B
pA2a hA2A
pA2 A2B
pA2 A2A
67
68
69
71
75
88
8.58
8.52
8.46
8.42
7.96
8.28
5.91
7.19
5.31
7.34
6.11
5.58
8.06
8.40
8.30
8.17
8.00
8.20
5.99
7.24
5.90
7.30
6.20
5.50
8.53
8.25
8.26
8.10
7.28
8.09
6.90
7.29
6.03
7.36
6.20
6.33
a pKI or pA2 values represent the mean of at least three independent experiments with a SEM always lower than 10%. b Smooth muscle of guinea pig
thoracic aorta. c Smooth muscle of male Sprague-Dawley rat aorta.
eter. All the detected signals were in accordance with the proposed
structures. Chemical shifts (δ scale) are reported in parts per million
(ppm) relative to the central peak of the solvent. Coupling constant
(J values) are given in hertz (Hz). Spin multiplicities are given as
s (singlet), d (doublet), dd (double doublet), t (triplet), or m
(multiplet), bs (broad singlet), dt (double triplet). Electron impact
mass (EI-MS) or electron spray ionization mass (ESI-MS) were
recorded in a quadrupolar Hewlett-Packard 5988 and in Waters ZQ
4000 apparatus, respectively. Elemental analyses were performed
on a Carlo Erba C, H, N analyzer. Fully purified products had
satisfactory (within (0.4% of theoretical values) C, H, N analyses.
Nonhigh-throughput chemical reactions were generally carried
out under an argon atmosphere whereas dry nitrogen was used in
parallel syntheses. Parallel reactions were performed using a
Chemspeed ASW 1000 instrument. In several instances, the final
products obtained by parallel synthesis were rapidly purified by
flash column chromatography up to an acceptable purity (better
than 90%, HPLC check) and tested without further purification. In
such a case only ESI/MS or EI/MS spectra and HPLC retention
At hA2B AdoR, the observed pKis were always slightly higher
than the corresponding pA2 values, whereas pKi and pA2 values
at the hA2A AdoR subtype varied in a less regular way.
The antagonistic activity and efficacy of the same set of
ligands were measured through functional isolated organ assay
at both guinea pig A2B and rat A2A AdoR subtypes (see Table
9). A representative example of this assay was shown in a
graphical form in Figure 3.
Again, pA2 values very close to the corresponding pKis were
obtained. At guinea pig A2B AdoR, the pA2 values were always
slightly lower than the corresponding pKi values, whereas
opposite results came from rat A2A AdoR subtypes.
Conclusions
The syntheses and the analyses of the biological activity at
the AdoR subtypes of a large array of 9-dAX derivatives allowed
us to identify the main structural features for both high hA2B
affinity and hA2B/hA2A selectivity. A high number of 9-dAX
derivatives with outstanding hA2B affinity (pKi > 8) were
obtained whereas relatively few ligands with Ki hA2A/Ki hA2B
ratio >100 were found. Among them, results for compounds
67, 69, and 88 were particularly interesting both in terms of
hA2B affinity (pKi ) 8.58, 8.46, and 8.28, respectively) and
selectivity over hA2A (741, 1412, and 501, respectively).
Moreover they were endowed with high selectivity also over
hA3 (>1000, 3090, and 2344, respectively) but, regrettably, their
selectivity over the hA1 AdoR subtype was quite low (11.7,
9.3, and 3.4, respectively). This was an important unmet goal
of the present study.
A number of selected ligands, tested in functional assays in
vitro, showed very interesting antagonist activities and efficacies
at both hA2A and hA2B AdoR subtypes, in good agreement with
their affinities measured in the binding assays. We are presently
working to overcome the problem of the relatively low
selectivity over the hA1 AdoR subtype, taking into account all
the structural insights gained in the present and other SAR and
SSR studies of AdoR ligands.
1
times were listed (see Table 14, Supporting Information). Mp, H
NMR, ESI/MS, or EI/MS spectra (Table 14, Supporting Informa-
tion), and elemental analysis (Table 15, Supporting Information)
were reported only for compounds prepared by traditional solution-
phase synthesis (7, 10, 21, 62-65, 67, 68, 70, 71, 74, 75, 80, 83,
85, 96, 99, 112, 133, 135, 137-139, 141, 146, 148, 155, 156, 159,
160, 162, 163-171, 173-184, 186-197).
The HPLC analyses were performed on a Symmetry C18 (2.1
× 10 mm, 3.5 µM) column with a Waters 2690 instrument. As
detectors, Micromass ZMD for ES ionization and Waters 996 Diode
Array were used. Diode array chromatograms were processed at
210 nm. The mobile phases were a mixture of formic acid (0.46
mL), aqueous ammonia 33% (0.115 mL), and water (1000 mL)
(phase A) and a quaternary mixture of formic acid (0.4 mL),
ammonia (0.1 mL), methanol (500 mL), and acetonitrile (500 mL)
(phase B). A gradient technique was applied going from A 100%
to 95% B in 20 min. The flow rate was 0.4 mL/min. The injection
volume was 5 µL.
The syntheses of 1,3 dimethyl-(or -dipropyl)-6-methyl-5-nitrou-
racil have been already described.33 Nonsymmetrically substituted
1,3-dialkyl-6-methyluracils were prepared by alkylation at position
3 (or 1) of the corresponding 1(3)-alkyl-6-methyluracils54 or of the
commercial 3-isobutyl-6-methyluracil.
In summary, our results showed that suitably substituted
9-dAXs may represent promising highly active and hA2B
selective antagonists, potentially useful as antiasthmatic agents.
Representative Procedure for the Condensation Reaction
between Benzaldehydes and 5-Nitro-6-methyluracils to afford
1,3-Dialkyl-6-styryluracil Derivatives IV (see Scheme 1). A
solution of the suitable 1,3-dialkyl-6-methyl-5-nitropyrimidine-2,4-
(1H,3H)-dione II (30 mmol), the appropriate benzaldehyde (30
mmol), and piperidine (4.45 mL, 45 mmol) in dry dioxane (150
mL) was refluxed for 5-70 h, under argon atmosphere. The
resulting solution was concentrated under vacuum, and the residue
was treated with ethanol until the formation of a precipitate
occurred. The solid was collected by filtration, dried under vacuum,
and purified by flash column chromatography to yield the expected
nitrostyryl derivatives IV in a sufficient purity (>95%) for the
subsequent syntheses.
Experimental Section
Chemicals and solvents were from Sigma-Aldrich. When neces-
sary, solvents were dried by standard techniques and distilled. After
extraction from aqueous phases, the organic solvents were dried
over anhydrous magnesium or sodium sulfate. Thin-layer chroma-
tography (TLC) was performed by using aluminum sheets precoated
with silica gel 60 F254 (0.2 mm) type E. Merck. Chromatographic
spots were visualized by UV light or Hanessian reagent.53 Purifica-
tion of crude compounds was carried out by flash column
chromatography on silica gel 60 (Kieselgel 0.040-0.063 mm, E.
Merck) or by crystallization. Melting points (uncorrected) for fully
purified products (see below) were determined in a glass capillary
tube on a Stuart Scientific electro thermal apparatus SMP3. 1H NMR
spectra were recorded on a 300 MHz Bruker AMX-300 spectrom-
Representative Procedures for the Ring Closure of 5-Nitro-
6-styryluracil IV to 9-dAXs V (see Scheme 1). Method A. A
solution of the 5-nitro-6-styryluracil derivative IV (3.00 mmol) in
triethyl phosphite (5 mL) was refluxed for 7 h. The resulting mixture

298 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1
Carotti et al.
was cooled and the precipitate collected by filtration and washed
with ethyl ether to yield the corresponding 1H-pyrrolo[3,2-d]-
pyrimidine-2,4(3H,5H)-dione derivatives V, that were purified by
flash column chromatography in the case of unacceptable purities.
Method B. To a solution of the appropriate 5-nitro-6-styryluracil
carboxylic acid or ester IV (1.94 mmol) in formic acid (18 mL)
was slowly added sodium dithionite (1.56 g, 9.70 mmol), and the
mixture was refluxed overnight. The resulting suspension was
cooled to room temperature and poured into water. The precipitate
was collected by filtration, washed with water and ethyl ether, and
then dried under vacuum to yield the expected 1H-pyrrolo[3,2-d]-
pyrimidine-2,4(3H,5H)-dione derivatives V, that were purified by
column chromatography, or by crystallization (water /MeOH) when
necessary.
Representative Procedures for the Amidation of Oxyacetic
Acid and Oxyacetic Ester of 9-Deazaxanthine Derivatives (V
and VI, see Scheme 2). The synthesis of the target amides was
performed by using one of the following methods (A-D).
Method A. The reaction took place in a sealed tube under argon
atmosphere. To 0.15 mmol of the ester V or VI (see Scheme 2)
and 16.00 mmol of amines was added a catalytic amount of sodium
cyanide (5 mg). In the case of liquid amines the reaction mixture
was heated at the boiling temperature of the amine whereas for
solid amines 2 mL of anhydrous dioxane was used as solvent and
the reaction heated at 100 °C. The reactions were monitored by
TLC, and when no more starting material was observed, the mixture
was cooled to room temperature and the final product isolated by
filtration and washed with ethyl ether and recrystallized from H2O/
MeOH. When no precipitate was formed, the reaction mixture was
concentrated under reduced pressure and the residue purified by
flash column chromatography on silica gel.
Method B. A solution of the proper carboxylic acid (V or VI,
see Scheme 2) (0.72 mmol) in anhydrous tetrahydrofuran (20 mL)
under argon atmosphere was cooled to -40 °C, and N-methylmor-
pholine (0.08 mL, 0.72 mmol) and isobutyl chloroformate (0.09
mL, 0.72 mmol) were slowly added. The mixture was stirred at
-40 °C for 2 h. Then the appropriate amine (0.72 mmol) was added,
and the mixture was stirred 15 min at -40 °C and then for
additional 12 h at room temperature. The solution was evaporated
under reduced pressure, and the residue was dissolved in DCM,
washed with a saturated solution of sodium bicarbonate, water, and
brine, and then dried (Na2SO4) and evaporated under reduced
pressure. The resulting crude product was purified by flash column
chromatography on silica gel or crystallized by H2O/MeOH.
Method C. To a mixture of the carboxylic acid (V or VI, see
Scheme 2) (1.24 mmol), N-(3-dimethylaminopropyl)-N’-ethylcar-
bodiimide hydrochloride (0.28 g, 1.49 mmol), 1-hydroxybenzo-
triazole (0.20 g, 1.49 mmol), triethylamine (0.35 mL, 2.48 mmol),
and anhydrous DCM (20 mL) was added the appropriate amine
(1.61 mmol) and the mixture was stirred at room temperature
overnight under argon atmosphere. The resulting solution was
evaporated under reduced pressure, and the residue was dissolved
in DCM and washed with a saturated sodium bicarbonate aqueous
solution. The organic phase was separated, washed with water and
brine, dried (Na2SO4), and evaporated under reduced pressure. The
resulting crude was purified by flash column chromatography on
silica gel or crystallized by H2O/MeOH.
Method D. To a mixture of the carboxylic acid (V or VI, see
Scheme 2) (0.21 mmol), N-(3-dimethylaminopropyl)-N’-ethylcar-
bodiimide hydrochloride (0.09 g, 0.23 mmol), 1-hydroxybenzot-
riazole (0.30 g, 0.23 mmol), and polymer-bound morpholine (0.28
g, 2.75 mmol/g based on nitrogen analysis) in anhydrous DMF
(4 mL) was added the suitable amine (0.23 mmol), and the mixture
was stirred overnight at room temperature. To the resulting solution
were added macroporous triethylammonium methylpolystyrene
carbonate (0.25 g, 2.8-3.5 mmol/g based on nitrogen elemental
analysis) and Amberlyst 15 (0.65 g) as scavengers, and the
suspension was stirred for 2 h (in the case of acidic or basic final
products the corresponding scavenger was not added). The resulting
suspension was filtered and evaporated under reduced pressure. The
residue was triturated with a mixture of MeOH/ethyl ether, and
the precipitate was collected by filtration to yield the desired product
in a sufficient purity for biological testing.
Biochemistry and Pharmacology. Radioligand Binding As-
says. Radioligand binding competition assays were performed in
vitro using A1, A2A, A2B, and A3 human receptors expressed in
transfected CHO (hA1), HeLa (hA2A and hA3), and HEK-293 (hA2B)
cells as described in detail in Supporting Information.
cAMP Assays. These assays were performed on hA2A and hA2B
receptors transfected in CHO cells by using the method described
by Salomon55 as described in Supporting Information.
Isolated Organ Assays. A2A Receptors. These assays were
performed on hA2A receptors56 from isolated aorta of 200-250 g
male Sprague-Dawley rats as reported in detail in Supporting
Information.
A
2B Receptors. These assays were performed in A2B receptors57
from isolated aorta of 300-350 g male guinea pigs as reported in
detail in Supporting Information.
Acknowledgment. The Italian authors thank the MIUR
(Ministero dell′Istruzione, dell′Universita` e della Ricerca Sci-
entifica, Rome (Italy), for partial financial support.
Supporting Information Available: Chemical, biological, and
pharmacological experimental data. This material is available free
of charge via the Internet at http://pubs.acs.org.
References
(1) Cohn, L.; Elias, J. A.; Chupp, G. L. Asthma: mechanisms of disease
persistence and progression. Annu. ReV. Immunol. 2004, 22, 789-
815.
(2) Frieri, M. Asthma concepts in the new millennium: update in asthma
pathophysiology. Allergy Asthma Proc. 2005, 26, 83-8.
(3) Szeleny, I.; Pahal, A. Drug Therapy in asthma bronchiale in the new
millennium. Pharmazie 2002, 57, 83-86.
(4) (a) Drazen, J. M.; Israel, E.; O’Byrne, P. M. Treatment of asthma
with drugs modifying the leukotriene pathway. N. Engl. J. Med. 1999,
340, 197-206. (b) Kemp, J. P. Recent advances in the management
of asthma using leukotriene modifiers. Am. J. Respir. Med. 2003, 2,
139-156. (c) Holgate, S. T.; Peters-Golden M. Introduction: the
antiinflammatory role of cysteinyl leukotriene receptor antagonists
in asthma. Allergy Clin. Immunol. 2003, 111, S1-4.
(5) (a) Stokes, J.; Casale, T. B. Rationale for new treatments aimed at
IgE immunomodulation. Ann. Allergy Asthma Immunol. 2004, 93,
212-217. (b) Babu, K. S.; Holgate S. T. The role of anti-IgE therapies
in the treatment of asthma. Hosp. Med. 2002, 63, 483-6.
(6) Marx, D.; Ezeamuzie, C. I.; Nieber, K.; Szelenyi, I. Therapy of
bronchial asthma with adenosine receptor agonists or antagonists.
Drug News Perspect. 2001, 14, 89-100.
(7) Holgate, S. T. The identification of the adenosine A2B receptor as
a novel therapeutic target in asthma. Br. J. Pharmacol. 2005, 145,
1009-1015.
(8) Jacobson, K. A.; IJzerman, Ad P.; Linden, J. 1,3-Dialkylxanthine
derivatives having high potency as Antagonist at human A2B
adenosine receptor. Drug. DeV. Res. 1999, 47, 45-53.
(9) Klinger, M.; Freissmuth, M.; Nanoff, C. Adenosine receptors: G
protein-mediated signalling and the role of accessory proteins. Cell
Signal. 2002, 14, 99-108.
(10) Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz, K. N.;
Linden, J. International Union of Pharmacology. XXV. Nomenclature
and classification of adenosine receptors. Pharmacol. ReV. 2001, 53,
527-552.
(11) Baraldi, P. G.; Tabrizi, M. A.; Preti, D.; Bovero, A.; Romagnoli, R.;
Fruttarolo, F.; Zaid, N. A.; Moorman, A. R.; Varani, K.; Gessi, S.;
Merighi, S.; Borea, P. A. Design, synthesis, and biological evaluation
of new 8-heterocyclic xanthine derivatives as highly potent and
selective human A2B adenosine receptor antagonists. J. Med. Chem.
2004, 47, 1434-1447.
(12) Volpini, R.; Costanzi, S.; Vittori, S.; Cristalli, G.; Klotz, K. N.
Medicinal chemistry and pharmacology of A2B adenosine receptors.
Curr. Top. Med. Chem. 2003, 3, 427-443.
(13) Cristalli, G.; Lambertucci, C.; Taffi, S.; Vittori, S.; Volpini, R.
Medicinal chemistry of adenosine A2A receptor agonists. Curr. Top.
Med. Chem. 2003, 3, 387-401.
(14) Mu¨ller, C. E.; Stein, B.; Adenosine receptor antagonists: structure
and potential therapeutic applications. Curr. Pharm. Des. 1996, 2,
501-530.
(15) Baraldi, P. G.; Borea, P. A. New potent and selective human
adenosine A3 receptor antagonists. Trends Pharmacol. Sci. 2000,
21, 456-459.

Human A2B Adenosine Receptor Deazaxanthine Ligands
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 1 299
(16) Baraldi, P. G.; Cacciari, B.; Romagnoli, R.; Merighi, S.; Varani, K.;
Borea, P. A.; Spalluto, G. A3 adenosine receptor ligands: history
and perspectives. Med. Res. ReV. 2000, 20, 103-128.
(17) van der Wenden, E. M.; IJzerman, A. P.; Soudijn, W. A. Steric and
electrostatic comparison of three models for the agonist/antagonist
binding site on the adenosine A1 receptor. J. Med. Chem. 1992, 35,
629-635.
(18) Kim, Y. C.; Ji, X.; Melman, N.; Linden, J.; Jacobson, K. A. Anilide
derivatives of an 8-phenylxanthine carboxylic congener are highly
potent and selective antagonists at human A2B adenosine receptors.
J. Med. Chem. 2000, 43, 1165-1172.
(19) Baraldi, P. G.; Cacciari, B.; Moro, S.; Spalluto, G.; Pastorin, G.; Da
Ros, T.; Klotz, K.-N.; Varani, K.; Gessi, S.; Borea, P. A. Synthesis,
biological activity, and molecular modeling investigation of new
pyrazolo[4,3-e]-1,2,4-triazolo[1,5-c]pyrimidine derivatives as human
A3 adenosine receptor antagonists. J. Med. Chem. 2002, 45, 770-
780.
(20) Soudijn, W.; van, Weden., I.; IJzerman, A. P. Medicinal chemistry
of adenosine A1 receptor ligands. Curr. Top. Med. Chem. 2003, 3,
355-367.
(21) Mu¨ller, C. E. Medicinal chemistry of adenosine A3 receptor ligands.
Curr. Top. Med. Chem. 2003, 3, 445-462.
(22) DeNinno, M. P.; Masamune, H.; Chenard, L. K.; DiRico, K. J.; Eller,
C.; Etienne, J. B.; Tickner, J. E.; Kennedy, S. P.; Knight, D. R.;
Kong, J.; Oleynek, J. J.; Tracey, W. R.; Hill, R. J. 3′-Aminoadenosine-
5′-uronamides: discovery of the first highly selective agonist at the
human adenosine A3 receptor. J. Med. Chem. 2003, 46, 353-355.
(23) Kim, Y. C.; Ji, X.; Melman, N.; Linden, J.; Jacobson, K. A. Anilide
derivatives of an 8-phenylxanthine carboxylic congener are highly
potent and selective antagonists at human A2B adenosine receptors.
J. Med. Chem. 2000, 43, 1165-1172.
(36) Vidal, B. J.; Esteve Trias C.; Segarra Matamoros, V.; Ravina Rubira,
E.; Fernandez Gonzales, F.; Loza Garcia, M. I.; Sanz Carreras, F.
6-Phenyldihydropyrrolo pyrimidinedione, 2003, WO 033694.
(37) Senda, S.; Suzui, A.; Honda, M.; Fujimura, H. Uracil derivatives
and related compounds. III. 5-Amino-1,3,6-trialkyluracil derivatives.
Chem. Pharm. Bull. 1958, 6, 482-487.
(38) Papesch, V.; Dodson, R. M. G. D. Pyrimido[5,4-d][1,2,3]triazines
J. Org. Chem. 1963, 28, 1329.
(39) Bicking, J. B.; Robb, C. M.; Watson, L. S.; Cragoe, E. J., Jr.
(Vinylaryloxy)acetic acids. A new class of diuretic agents. 2. (4-(3-
Oxo-1-alkenyl)phenoxy)acetic acids. J. Med. Chem. 1976, 19, 544-
547.
(40) Agarwal, R.; Chaudray, C.; Singh, C. J.; Misra, V. S. Synthesis of
2-methoxy-4-(3-methyl-5-oxo-1-phenyl-4-pyrazolylidenemethyl)phe-
noxyacetic acid and its ethyl ester. Indian J. Med Chem Sect. B Org.
Chem. Includ. Med Chem 1982, 21B, 152-153.
(41) Albericio, F.; Barany, G. An acid-labile anchoring linkage for solid-
phase synthesis of C-terminal peptide amides under mild conditions.
Int. J. Pept. Protein Res. 1987, 30, 206-216.
(42) Tipson, R. S. Antiviral Activity of 2,2-dichloro-4′-formylacetoanilide
thiosemicarbazone. J. Med. Chem. 1963, 35, 216-217.
(43) Cheng, M. F.; Fang, J. M. Liquid-phase combinatorial synthesis of
1,4-benzodiazepine-2,5-diones as the candidates of endothelin recep-
tor antagonism. J. Comb. Chem. 2004, 6, 99-104
(44) Behrens, G.; Hildenbrand, K.; Schulte-Frohlinde, D.; Herak, J. N.
Reaction of sulfate radical anion with methylated uracils. An electron
spin resonance study in aqueous solution. J. Chem. Soc., Perkin
Trans. 2 1988, 3, 305-307.
(24) Hayallah, A. M.; Sandoval-Ramirez, J.; Reith, U.; Schobert, U.;
Preiss, B.; Schumacher, B.; Daly, J. W.; Mu¨ller, C. E. 1,8-
disubstituted xanthine derivatives: synthesis of potent A2B-selective
adenosine receptor antagonists. J. Med. Chem. 2002, 45, 1500-1510.
(25) Kim, S. A.; Marshall, M. A.; Melman, N.; Kim, H. S.; Mu¨ller, C.
E.; Linden, J.; Jacobson, K. A. Structure-activity relationships at
human and rat A2B adenosine receptors of xanthine derivatives
substituted at the 1-, 3-, 7-, and 8-positions. J. Med. Chem. 2002,
45, 2131-2138.
(26) Baraldi, P. G.; Tabrizi, M. A.; Preti, D.; Bovero, A.; Fruttarolo, F.;
Romagnoli, R.; Moorman, A. R.; Gessi, S.; Merighi, S.; Varani, K.;
Borea, P. A. [3H]-MRE 2029-F20, a selective antagonist radioligand
for the human A2B adenosine receptors. Bioorg. Med. Chem. Lett.
2004, 14, 3607-3610.
(27) Hansel, T. T.; Tennant, R. C.; Tan, A. J.; Higgins, L. A.; Neighbour,
H.; Erin, E. M.; Barnes, P. J. Theophylline: mechanism of action
and use in asthma and chronic obstructive pulmonary disease. Drugs
Today (Barc.) 2004, 40, 55-69.
(28) Maconi, A.; Pastorin, G.; Da Ros, T.; Spalluto, G.; Gao, Z. G.;
Jacobson, K. A.; Baraldi, P. G.; Cacciari, B.; Varani, K.; Moro, S.;
Borea, P. A. Synthesis, biological properties, and molecular modeling
investigation of the first potent, selective, and water-soluble human
A3 adenosine receptor antagonist. J. Med. Chem. 2002, 45, 3579-
3582.
(29) Ongini, E.; Monopoli, A.; Cacciari, B.; Baraldi, P. G. Selective
adenosine A2A receptor antagonists. Farmaco 2001, 56, 87-90.
(30) Mu¨ller, C. E. A1 adenosine receptors and their ligands: overview
and recent developments. Farmaco 2001, 56, 77-80.
(31) Hayallah, A. M.; Sandoval-Ramirez, J.; Reith, U.; Schobert, U.;
Preiss, B.; Schumacher, B.; Daly, J. W.; Mu¨ller, C. E. 1,8-
Disubstituted xanthine derivatives: synthesis of potent A2B-selective
adenosine receptor antagonists. J. Med. Chem. 2002, 45, 1500-1510.
(32) Baraldi, P. G.; Romagnoli, R.; Tabrizi, M. A.; Bovero, A.; Preti, D.;
Fruttarolo, F.; Moorman, A. R.; Borea, P. A. New heterocyclic ligands
for the adenosine receptors P1 and for the ATP receptors P2. Farmaco
2005, 60, 185-202.
(33) Grahner, B.; Winiwarter, S.; Lanzner, W.; Mu¨ller, C. E. Synthesis
and structure-activity relationships of deazaxanthines: analogues
of potent A1- and A2-adenosine receptor antagonists. J. Med. Chem.
1994, 37, 1526-1534.
(34) Stefanachi, A.; Leonetti, F.; Cappa, A.; Carotti, A. Fast and highly
efficient one-pot synthesis of 9-deazaxanthines. Tetrahedron Lett.
2003, 44, 2121-2123.
(45) Aichner, J.; Egg, H.; Gapp, D.; Haller, S.; Rakowitz, D.; Ramspacher,
U. Insight into novel cyclization reactions using acetic anhydride in
the presence of 4-dimethylaminopyridine. Heterocycles 1993, 36,
307-311.
(46) Senda, S.; Suzui, A. Uracil derivatives and related compounds. I.
Condensation of monosubstituted urea and ethyl acetoacetate. Chem.
Pharm. Bull. 1958, 6, 476-479
(47) Boese, A. B. 4-methyluracil and its homologs, US patent, 1938,
2,138,758.
(48) Senda, S.; Hirota, K.; Banno, K. Pyrimidine derivatives and related
compounds. 15. Synthesis and analgetic and antiinflammatory
activities of 1,3-substituted 5-amino-6-methyluracil derivatives. J.
Med. Chem. 1972, 15, 471-476.
(49) Tipson, R. S. Antiviral Activity of 2,2-dichloro-4′-formylacetoanilide
thiosemicarbazone. J. Med. Chem. 1963, 35, 216-217.
(50) Tani, M.; Ikegami, H.; Tashiro, M.; Hiura, T.; TsuiKioka, H.; Kaneko,
C.; Notoya, T.; Shimizu, M.; Uchida, M.; Regioselective Bromination
of methoxy derivatives of ethyl indole-2-carboxylate [Synthetic
studies of indoles and related compounds]. Heterocycles 1992, 34,
2349-2362.
(51) Ho¨gberg, T.; Stro¨m, P.; Ebner, M.; Ra¨msby, S. Cyanide as an efficient
and mild catalyst in the aminolysis of esters. J. Org. Chem. 1987,
52, 2033-2036.
(52) Jacobson, K. A.; Ukena, D.; Padgett, W.; Daly, J. W.; Kirk, K. L.
Xanthine functionalized congeners as potent ligands at A2-adenosine
receptors. J. Med. Chem. 1987, 30, 211-214.
(53) Touchstone, J. AdVances in Thin-Layer Chromatography; Wiley:
New York, 1982.
(54) Nieto, R. M.; Coelho, A.; Martinez, A.; Stefanachi, A.; Sotelo, E.;
Ravina, E. Synthesis of 1-substituted-6-methyluracils. Chem. Pharm.
Bull. 2003, 51, 1025-1028.
(55) Salomon, Y. Adenylate cyclase assay. AdV. Cyclic Nucleotide. Res.
1979, 10, 35-55.
(56) Prentice, D. J.; Hourani, S. M. O. Activation of multiple sites by
adenosine analogues in the rat isolated aorta. Br. J. Pharmacol. 1996,
118, 1509-1517.
(57) Alexander, S. P.; Losinsky, A.; Kendall, D. A.; Hill, S. J. A
comparison of A2 adenosine receptor induced cyclic AMP generation
in cerebral cortex and relaxation of pre-contracted aorta. Br. J.
Pharmacol. 1994, 111, 185-190.
(35) Carotti, A.; Stefanachi, A.; Ravin˜a, E.; Sotelo, E.; Loza, M. I.;
Cadavid, M. I.; Centeno, N. B.; Nicolotti O. 8-Substituted-9-
deazaxanthines as adenosine receptor ligands: design, synthesis and
structure-affinity relationships at A2B. Eur. J. Med. Chem. 2004,
39, 879-887.
JM0506221