2-(Benzimidazol-2-yl)quinoxalines: A novel class of selective antagonists at human A1 and A3 adenosine receptors designed by 3D database searching

Article abstract of DOI:10.1021/jm050792d

The Cambridge Structural Database (CSD) was searched through two 3D queries based on substructures shared by well-known antagonists at the A₁ and A₃ adenosine receptors (ARs). Among the resulting 557 hits found in the CSD, we selecte

Full text of DOI:10.1021/jm050792d

J. Med. Chem. 2005, 48, 8253-8260
8253
2-(Benzimidazol-2-yl)quinoxalines: A Novel Class of Selective Antagonists at
Human A1 and A3 Adenosine Receptors Designed by 3D Database Searching
Ettore Novellino,*,† Barbara Cosimelli,† Marina Ehlardo,† Giovanni Greco,† Manuela Iadanza,†
Antonio Lavecchia,† Maria Grazia Rimoli,† Annalisa Sala,† Antonio Da Settimo,‡, Giampaolo Primofiore,‡
Federico Da Settimo,‡ Sabrina Taliani,‡ Concettina La Motta,‡ Karl-Norbert Klotz,§ Daniela Tuscano,|
Maria Letizia Trincavelli,| and Claudia Martini|
Dipartimento di Chimica Farmaceutica e Tossicologica, Universita` di Napoli “Federico II”, Via D. Montesano, 49, 80131
Napoli, Italy, Dipartimento di Scienze Farmaceutiche, Universita` di Pisa, Via Bonanno 6, 56126 Pisa, Italy, Institut fu¨r
Pharmakologie und Toxikologie, Universita¨t Wu¨rzburg, Versbacher Strasse 9, 97078 Wu¨rzburg, Germany, and Dipartimento di
Psichiatria, Neurobiologia, Farmacologia e Biotecnologie, Universita` di Pisa, Via Bonanno 6, 56126 Pisa, Italy
Received August 10, 2005
The Cambridge Structural Database (CSD) was searched through two 3D queries based on
substructures shared by well-known antagonists at the A1 and A3 adenosine receptors (ARs).
Among the resulting 557 hits found in the CSD, we selected five compounds to purchase,
synthesize, or translate synthetically into analogues better tailored to interact with the biological
targets. Binding experiments using human ARs showed that four out of five tested compounds
turned out to be antagonists at the A1AR or A3AR with Ki values between 50 and 440 nM.
Lead optimizations of 2-(benzimidazol-2-yl)quinoxalines (BIQs, 3) gave the best results in terms
of potency and selectivity at the A1 and A3 ARs. Particularly, 2-(4-ethylthiobenzimidazol-2-
yl)quinoxaline (3e) exhibited Ki values at the A1AR, A2AAR, and A3AR of 0.5, 3440, and 955
nM, respectively, whereas 2-(4-methylbenzimidazol-2-yl)quinoxaline (3b) displayed at the same
ARs Ki values of 8000, 833, and 26 nM, respectively.
Introduction
drugs for the treatment of neurodegenerative disorders,
such as Parkinson’s disease.11,12
Adenosine is an endogenous signaling nucleoside
released from most cells that is responsible for many
physiological processes.1-5 Once in the extracellular
space, it modifies cell functioning by triggering specific
cell membrane G-protein-coupled receptors that modu-
late several effector systems, including adenylate cy-
clase, potassium and calcium channels, phospholipase
C or D, and guanylate cyclase. At least four adenosine
receptor (AR) subtypes, classified as A1, A2 (A2A, A2B),
and A3, have been identified as mediators of these
effects.6 Each receptor subtype has been indicated as a
target for the development of agonist- and antagonist-
based therapies against a wide range of pathologies
(CNS disorders, cardiac arrhythmia, ischemic injuries,
asthma, renal failure, and inflammatory diseases).
In particular, the A1AR is extensively expressed in
both the CNS and peripheral tissues, such as kidney,
lung, bladder, adipose tissue, and heart.6 The main
potential therapeutic applications of A1AR antagonists
are for the treatment of cognitive deficits, CNS disorders
such as Alzheimer’s disease, renal failure, cardiac
bradiarrhythmias, and asystolic arrest.2,7
The A3AR is widely distributed in peripheral organs
with the largest expression detected in the testis,
liver, and lung, but it is also expressed in the CNS
and heart.13 A3AR antagonists might be therapeutic-
ally useful for the acute treatment of stroke, for glau-
coma, and also as antiasthmatic and antiinflammatory
drugs.14-16
In this scenario, the search for novel potent and
selective antagonists at the different ARs is still an
attractive field of research.
In recent years, it has been shown that database
searching is a valuable tool to discover novel lead
compounds.17 The input for the search engine can be a
query made up of a substructure (two-dimensional (2D)
search). When the query includes torsion angles and/or
interatomic distances, the search is of three-dimensional
(3D) type. Queries are typically built up from a phar-
macophore model or from the structure of a highly active
compound whose target-bound conformation is known
from X-ray or NMR experiments or is the only one
energetically allowed. A database search ends up yield-
ing a number of hits (i.e., structures matching the query)
between zero and thousands. If the number of hits is
too small or too large, the search is repeated, adjusting
the settings of the query until a manageable number of
hits is found. Typically, the user focuses the search on
“drug-like” structures18 by applying filters on the basis
of molecular weight, the number of heteroatoms, cal-
culated lipophilicity, the number of rotatable bonds, and
any other suitable physicochemical parameter. Because
hits may not be tailored to interact with the target
biomacromolecule (e.g., because they are exceedingly
bulky, hydrophilic, etc.) or are chemically unstable, their
The A2AAR is highly coexpressed with the dopamine
D2 receptor in the striatum of the basal ganglia, where
an antagonistic cross-talk between A2AAR and D2 has
been reported.8-10 Consequently, potent and selective
A2AAR antagonists are currently investigated as new
* To whom correspondence should be addressed. Phone: (+39)-
081678644. Fax: (+39)081678630. E-mail: novellin@unina.it.
† Universita` “Federico II” di Napoli.
‡ Dipartimento di Scienze Farmaceutiche, Universita` di Pisa.
§ Universita¨t Wu¨rzburg.
| Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Bio-
tecnologie, Universita` di Pisa.
In memory of Prof. Antonio Da Settimo, who died in February
2005.
10.1021/jm050792d CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/11/2005

8254 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Novellino et al.
Table 1. Affinities at Human A1, A2A, and A3 ARs of
Compounds Designed as A1AR and/or A3AR Antagonists by 3D
Database Searching
following substructures: (1) C-N-C-C-NH-C (ex-
tracted from the A1AR antagonists I26 and II27 as well
from the A3AR antagonist IV28) or (2) C-C-N-C-
NH-C (featured by the A1AR antagonist III29 as well
as by the A3AR antagonists V30 and VI31). The two
substructures were further detailed through 2D con-
straints based on connectivity and 3D constraints based
on torsion angles to yield the queries used to scan the
CSD32 (Figure 2). The whole process of lead finding is
illustrated in Figure 3. The searches performed with
queries 1 and 2 gave 299 and 258 hits, respectively.
Each hit was subsequently inspected on the basis of
“drug-likeliness” criteria proposed by Lipinski,18 chemi-
cal stability, and ease of synthesis. At the end of this
step, we selected five structures filed in the CSD with
the reference codes bzquxl,33 biyroq,34 fapwoi,35 hex-
pab,36 and sortey.37 Biyroq (1) was purchased from
ChemBridge Corp. (San Diego, CA); hexpab (2) was
prepared as described in the literature.36 The remaining
three hits (bzquxl, fapwoi, and sortey) were synthetically
translated into analogues expected to display higher
affinity as antagonists at the A1AR and/or A3AR, namely
compounds 3a, 4a (less bulky), and 5a (less hydrophilic).
Ki (nM)a
b
c
d
cmpd
hA1
hA2A
hA3
1
>10000
3336 ( 557
50 ( 15
>10000
>10000
561 ( 17
>10000
>10000
337 ( 28
16 ( 3
>1000
125 ( 20
763 ( 13
130 ( 3
440 ( 33
>1000
2
3a
4a
5ae
>10000
>10000
DPCPX
0.5 ( 0.03
14 ( 4
NECA
73 ( 5
Chemistry
Cl-IBMECA
890 ( 61
401 ( 25
0.22 ( 0.02
Compounds 3a-o were prepared by following the
published synthetic procedure (Schemes 1 and 2) for the
obtainment of 3a.38 In brief, the starting quinoxaline-
2-carboxylic acid 6 was heated in polyphosphoric acid
with the appropriate phenylendiamine 7a-o to obtain
the target benzimidazolequinoxaline 3a-o.
The o-phenylendiamines 7 were all commercially
available, except 7d and e; their synthesis involved the
preparation of 10e, in accordance with the published
procedure for 10d,39 followed by a simple reduction of
the nitro group to an amino (Scheme 3).
a The Ki values are means ( SEM of at least three determina-
tions derived from an iterative curve-fitting procedure (Prism
program, GraphPad, San Diego, CA). b Displacement of specific
[3H]DPCPX binding in membranes obtained from hA1AR stably
expressed in CHO cells. c Displacement of specific [3H]NECA
binding in membranes obtained from hA2AAR stably expressed in
CHO cells. d Displacement of specific [125I]AB-MECA binding in
membranes obtained from hA3AR stably expressed in CHO cells.
e Assayed with a purity degree of 96%.
structure is often converted into derivatives with more
chances of displaying the desired biological properties.
Several examples of 3D database searches aimed at
identifying new leads are documented in literature.17,19-24
We applied the 3D database searching approach to
design new antagonists at the A1 and A3 ARs. The Cam-
bridge Structural Database (CSD)25 (version 5.25, No-
vember 2003) was scanned using two different queries,
each made up of a pharmacophore substructure whose
3D disposition was defined by torsion angles. Our work-
ing hypothesis was that pharmacophoric features of A1-
AR and A3AR antagonists could be found in some of the
structures deposited in the CSD that have never been
assayed or even considered for such biological proper-
ties. Among the resulting 557 hits, we selected five com-
pounds to purchase, synthesize, or translate syntheti-
cally into analogues thought to be more suited for bind-
ing to the target receptors. Four compounds turned out
to be active as antagonists at the human A1AR or A3-
AR with Ki values between 50 and 440 nM (Table 1).
Among these compounds, 2-(benzimidazol-2-yl)quinoxa-
line (BIQ), labeled as 3a in this contribution, was select-
ed as the most promising lead to optimize affinity and
selectivity at both the A1AR and the A3AR. In a few
make-and-test cycles, we obtained two BIQ derivatives
active as selective antagonists at the A1AR (3e with Ki
) 0.5 nM) and A3AR (3b with Ki ) 26 nM) (Table 4).
Compound 4a was prepared starting from the com-
mercially available 11 (Ambinter, Paris, France) and
acetic anhydride.40 Product 4d, previously reported to
Scheme 1
Scheme 2
Scheme 3
3D Database Search and Lead Design
The six highly potent selective antagonists at the A1-
AR and A3AR reported in Figure 1 contain one of the

BIQs: Antagonists at Human A1 and A3 ARs
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8255
Figure 1. Substructures extracted by potent antagonists at the A1AR and A3AR.
Figure 2. Two 3D queries employed to search the CSD using the ConQuest search engine.32 Organometallic compounds were
excluded from the search by the “only organics” option. The ranges of the torsion angles were chosen to yield a manageable
numbers of hits.
be obtained by reaction of 11 and propionic acid with a
low yield (2.6%),35 was prepared from 11 and propionic
anhydride with satisfactory yields (34%). The unknown
4b,c were synthesized, together with derivatives 4a,d,
by the reaction of 11 with a mixture of acetic anhydride
and propionic anhydride in a large excess (Scheme 4).
The known compounds 5b,c were obtained following
a reported procedure;41 product 5d was similarly syn-
thesized, reacting the diamino derivative 12 and 1,2-
dichloroethylene in alkaline medium to give 13, which
was successively cyclized to the target triazole (Scheme
5).
Finally, compound 5a was obtained with a very small
yield, following a different synthetic strategy, reported
in Scheme 6; many other attempts at preparation failed.
The known compound 1442 was allowed to react with

8256 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Novellino et al.
Figure 3. Structures and refcodes of the five hits identified by searching the CSD. Three out of the five hits (bzquxl, fapwoi, and
sortey) were synthetically translated into analogues 3a, 4a, and 5a, respectively, expected to display a higher affinity as antagonists
at the A1AR and/or the A3AR.
[3H]DPCPX, [3H]NECA, or [125I]AB-MECA binding from
transfected CHO cells. Experiments were performed as
described elsewhere.43
Scheme 4
Results and Discussion
Table 1 lists the binding affinities at the human A1,
2A, and A3 ARs of compounds 1, 2, 3a, 4a, and 5a,
A
designed as antagonists of the A1AR and/or the A3AR
as schematized in Figure 3. Although 1 did not elicit
any appreciable affinity for any of the AR subtypes, the
remaining compounds displayed submicromolar affinity
at the A1AR (3a) or the A3AR (2, 4a, and 5a), consistent
with our working hypothesis. The disappointing binding
data of 1 might be ascribed to a combination of small
molecular size and lipophilicity. Although the above
drawbacks could have been solved by preparing alkyl
derivatives of 1, we decided to initially focus lead
optimization efforts on the more promising and easily
modifiable four active compounds 2, 3a, 4a, and 5a.
Three homologues of the 2,5,8-trimethylbenzotriimi-
dazole 4a were prepared and tested in an attempt to
improve potency and retain selectivity at the A3AR. As
shown in Table 2, none of these compounds (4b-d)
revealed a significant selectivity for either the A3AR or
the A1AR. Incidentally, the triethyl derivative 4d was
slightly selective for the A1AR.
Scheme 5
Scheme 6
The triazolopyrimidinone 5a exhibited an interesting
selectivity profile at the A3AR (Table 3). Actually, this
compound was assayed as a mixture made up of 96% of
(R,S)-5a and 4% of the corresponding (R,S)-2-methyl
isomer. However, in the context of our lead-finding
strategy, there was no need to obtain 5a with a higher
purity level. Simple modifications about the thiazolidine
ring of 5a yielded compounds equipotent (5c) or inactive
(5b,d) in the binding experiments. Research in this
series is still in progress, and results will be reported
in due time.
1,2-dibromopropane to give a mixture of 5a and the
regioisomer in which the methyl group is on the C-2
atom. After column chromatography and recrystalliza-
tion, compound 5a was obtained with a 96% degree of
purity.
Biology
The affinity of the new compounds toward human A1,
2A, and A3 ARs was evaluated by competition experi-
A
ments assessing their respective abilities to displace

BIQs: Antagonists at Human A1 and A3 ARs
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8257
Table 2. Affinities at Human A1, A2A, and A3 ARs of
Compounds 4a-d
of 3b are 8000, 833, and 26 nM at the A1AR, A2AAR,
and A3AR, respectively). The dramatic switch of selec-
tivity observed on going from 3e to 3b is probably
related to a better steric tolerance of the A1AR cavity
hosting the R1 substituent/4′ position of the BIQ deriva-
tives compared with that of the topologically analogous
pocket of the A3AR. Consistent with this hypothesis, 3d
and 3c are increasingly less selective for the A1AR over
the A3AR, owing to their less bulky R1 substituents
(SCH3 and Cl, respectively). Replacement of the SC2H5
of 3e with the relatively large NO2 led to 3f inactive
at the A1AR but provided appreciable affinity at the
A3AR. This fact apparently contradicts the above state-
ment regarding the greater amount of room available
to R1 substituents at the A1AR compared with that at
the A3AR. However, NO2 is hydrophilic, whereas the
groups so far considered in the same position (CH3, Cl,
SCH3, and SC2H5) are lipophilic. Taken together, these
data suggest that the A1AR pocket hosting the R1
substituent is mainly lipophilic. The poor binding
profiles of the azaisosters 3m-o can be reasonably
ascribed to their unfavorable hydrophilicity.
Ki (nM)a
b
c
d
cmpd
R1
R2
R3
hA1
hA2A
hA3
4a
4b
4c
4d
CH3
CH3
CH3
CH3
CH3
CH3
>10000
>10000
130 ( 3
C2H5 496 ( 16 >10000 146.5 ( 21
C2H5 C2H5 966 ( 43 329 ( 18
135 ( 15
C2H5 C2H5 C2H5
46 ( 4
720 ( 20 104.2 ( 9
a See footnote of Table 1 and Ki values of reference ligands. b See
footnote of Table 1. c See footnote of Table 1. d See footnote of Table
1.
Table 3. Affinities at Human A1, A2A, and A3 ARs of
Compounds 5a-d
A small substituent in the benzimidazole 5′ position
(R2), as in 3g-i, is associated with little or no affinity
at the ARs, thus suggesting that the corresponding
binding sites (especially A1AR and A2AAR) do not have
much room available for the R2 substituent. The 5′ and
6′ positions bearing substituents R2 and R3 are chemi-
cally equivalent, thanks to the free rotation about the
bond connecting the two heteroaromatic rings as well
as tautomeric shifting of the imidazole hydrogen. The
above-mentioned equivalence implies that steric en-
cumbrance at the ARs should apply to both the 5′ and
6′ positions of the benzimidazole nucleus. Such a model
Ki (nM)a
b
c
d
cmpd
hA1
hA2A
hA3
5a
5b
5c
5d
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
440 ( 33
>1000
437 ( 51
>1000
explains the poor affinities of 3j at the A1AR and A2A
-
a See footnote of Table 1 and Ki values of reference ligands. b See
AR and the substantial equipotence of 3j and 3a at the
A3AR. However, the high potency at the A1AR of the
4′-methyl-6′-chloro derivative 3k seems difficult to
rationalize in the light of the structure-affinity rela-
tionships so far outlined. On comparing the Ki value of
3k (3.6 nM) with the disappointing Ki values of 3b and
3h at the A1AR, it becomes clear that the CH3 and Cl
substituents on the benzimidazole ring do not exert any
additive effects on affinity. One can speculate that 3k
adopts a binding mode at the A1AR different from those
of the monosubstituted BIQ derivatives. Replacement
of the CH3 of 3k with a Cl to give 3l slightly increases
the Ki values at the A1AR and A3AR, thus confirming
that the electronic effects of R1 and R2 substituents do
not play a significant role in tuning affinity at these
receptors.
To sum up, a few chemically diverse antagonists at
the human A1AR and A3AR were designed by a 3D
searching of the CSD. Subsequent binding experiments
confirmed the rationale of this design approach because
four out the five tested compounds revealed appreciable
affinities at one or both of the target receptors. Lead
optimizations in the BIQ series 3 gave the best results
in terms of potency and selectivity at the A1AR as well
footnote of Table 1. c See footnote of Table 1. d See footnote of Table
1.
Compounds 2 and 3a were soon regarded as the most
promising leads in the perspective of affinity and
selectivity optimization at the target receptors (Table
1). Because 2 (fairly selective for the A3AR) was a
truncated analogue of 3a (slightly selective for the A1-
AR), we hypothesized that 2-(benzimidazol-2-yl)qui-
noxaline (BIQ) would represent a suitable molecular
scaffold to obtain both A1AR and A3AR selective an-
tagonists. Several derivatives of 3a were, in principle,
obtainable from commercially available substituted-o-
phenylendiamines, whereas structural modifications to
compound 2 seemed more difficult and restricted to the
replacement of the chlorine. Accordingly, we prepared
and tested some BIQ derivatives substituted on the
benzimidazole system (3b-l) or featuring isosteric CH
f N variations about the benzimidazole ring (3m-o).
Their structures and binding affinities at the A1, A2A
,
and A3 ARs are summarized in Table 4. Compound 3e,
characterized by R1 ) SC2H5 in the benzimidazole 4′
position, stands out for its remarkable affinity and
selectivity at the A1AR (Ki values of 0.5, 3440, and 955
nM at the A1AR, A2AAR, and A3AR, respectively).
Interestingly, replacement of the SC2H5 of 3e with the
smaller and less lipophilic CH3, to give 3b, is associated
with a 6000-fold drop in affinity at the A1AR, as well
as a 37-fold increase in potency at the A3AR (Ki values
as the A3AR: 3e exhibited Ki values at the A1AR, A2A
-
AR, and A3AR of 0.5, 3440, and 955 nM, respectively,
whereas 3b displayed at the same ARs Ki values of 8000,
833, and 26 nM, respectively.

8258 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Novellino et al.
Table 4. Affinity of BIQs Derivatives 3a-o at Human A1, A2A, and A3 ARs
Ki (nM)a
b
c
d
cmpd.
A
R1
R2
R3
hA1
hA2A
hA3
3a
3b
3c
3d
3e
3f
3g
3h
3i
3j
3k
3l
3m
3n
3o
H
CH3
Cl
H
H
H
H
H
H
H
50 ( 15
8000 ( 567
350 ( 61
2.1 ( 0.3
0.5 ( 0.01
>10000
561 ( 17
833 ( 67
>10000
3000 ( 200
3440 ( 980
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
>10000
763 ( 13
26 ( 9
H
H
140 ( 20
278 ( 44
955 ( 215
879 ( 72
>1000
SCH3
SC2H5
NO2
H
H
H
H
CH3
Cl
H
>10000
H
H
>10000
632 ( 10
>1000
H
OCH3
CH3
H
H
>10000
H
CH3
Cl
Cl
>10000
370 ( 35
238 ( 32
699 ( 120
>1000
CH3
Cl
3.6 ( 0.6
9.3 ( 1.2
>10000
H
[4,5-b]pyridine
[4,5-c]pyridine
[4,5-d]pyrimidine
>10000
>10000
168 ( 23
>1000
a See footnote of Table 1 and Ki values of reference ligands. b See footnote of Table 1. c See footnote of Table 1. d See footnote of Table
1.
acetate. Yields, melting points, and analytical and spectral
data are reported in the Supporting Information.
Experimental Section
Chemistry. Melting points were determined using a Bu¨chi
apparatus B-540 and are uncorrected. Routine nuclear mag-
netic resonance spectra were recorded on a Varian Mercury
400 spectrometer operating at 400 MHz. EI and HRMS mass
spectra were obtained on a Finnigan MAT95XP spectrometer
using a direct injection probe and an electron beam energy of
70 eV. ESI mass spectra were recorded with a Micromass ZMD
Waters Instrument or an Applied Biosystems API 2000.
Evaporation was performed in vacuo (rotary evaporator).
Analytical TLC was carried out on Merck 0.2 mm precoated
silica gel aluminum sheets (60 F-254). Silica gel 60 (230-400
mesh) was used for column flash chromatography. Elemental
analyses were performed by our Analytical Laboratory in Pisa
and agreed with theoretical values to within (0.4%.
General Procedure for the Synthesis of 2-Ethyl-5,8-
dimethyl-4,7-dihydro-1H-benzo[1,2-d:3,4-d′:5,6-d′′]triimi-
dazole, 4b, and 2,5-Diethyl-8-methyl-4,7-dihydro-1H-
benzo[1,2-d:3,4-d′:5,6-d′′]triimidazole, 4c. Following the
above-reported procedure, benzenehexamine 11 (0.50 g, 2.91
mmol) was reacted with a mixture of acetic anhydride (5 mL)
and propionic anhydride (10 mL). The precipitated solid,
containing the derivatives 4a-d, was flash chromatographed
(CHCl3-CH3OH, 9:1 v/v as the eluant) to separate the four
components. The fastest running band was identified as 4d
(traces), the second band was the derivative 4c (10%), then
4b (12%), and the slowest band was 4a (15%). Each compound
was recrystallized from ethyl acetate. Melting points and
analytical and spectral data are reported in the Supporting
Information.
3-Chloro-2-nitroaniline was commercially avalaible from
Frinton Laboratories Inc., New Jersey, U.S.A.
General Procedure for the Synthesis of 2-(Subtituted-
1H-benzimidazol-2-yl)quinoxaline Derivatives 3a-l, 2-(1H-
Imidazo[4,5-b]pyridin-2-yl)quinoxaline, 3m, 2-(1H-Imi-
dazo[4,5-c]pyridin-2-yl)quinoxaline, 3n, and 2-(7H-Purin-
8-yl)quinoxaline, 3o. A mixture of quinoxaline-2-carboxylic
acid 6 (0.50 g, 2.9 mmol), the appropriate diamine 7a-o (2.9
mmol), and polyphosphoric acid (17.5 g) was heated for 1 h at
125 °C and then for 2 h at 200 °C. The mixture was poured
into a large volume of ice-water, and the solid mass was
collected. The crude product was treated with saturated
sodium bicarbonate solution and purified by chromatography
to give the desired derivative 3a-o. Yields, melting points,
and analytical and spectral data are reported in the Supporting
Information.
General Procedure for the Synthesis of 2,5,8-Trimeth-
yl-4,7-dihydro-1H-benzo[1,2-d:3,4-d′:5,6-d′′]triimida-
zole, 4a, and 2,5,8-Triethyl-4,7-dihydro-1H-benzo[1,2-d:
3,4-d′:5,6-d′′]triimidazole, 4d. A solution of benzenehexamine
11 (0.50 g, 2.91 mmol) and the appropriate anhydride (15 mL)
was stirred under a nitrogen atmosphere at 170 °C until a
white precipitate formed (1.5 h). The solid was taken off by
filtration, and the filtrate was refluxed with HCl 4 N (15 mL)
for 4 h under a nitrogen atmosphere. The solution was made
alkaline with NaOH 2 N, and the precipitated solid was
collected and purified by flash chromatography (CHCl3-CH3-
OH, 9:1 v/v as the eluant) and then recrystallized from ethyl
7-Methyl-6,7-dihydro[1,3]thiazolo[3,2-a][1,2,3]triazolo-
[4,5-d]pyrimidin-9(1H)-one, 5a. 1,2-Dibromopropane (0.62
mL, 6.0 mmol) and triethylamine (0.92 mL, 6.6 mmol) were
slowly added to a stirred suspension of 1442 (1.02 g, 6.0 mmol)
in anhydrous DMF (8 mL), and then the mixture was heated
at reflux for 1 h. After cooling and concentration to dryness,
the residue obtained, made up of a mixture of 6-methyl and
7-methyl isomers (1H NMR spectrum), was flash chromato-
graphed (CHCl3-MeOH, 9:1 v/v as the eluant). After recrys-
tallization from water, the 7-methyl isomer 5a was obtained
with a 96% degree of purity and a yield of 17%. 1H NMR
(DMSO-d6): δ 16.06 (bs exch, 1H, NH), 5.15 (m, 1H, CH), 3.98
(m, 1H, CH2), 3.24 (m, 1H, CH2), 1.38 (d, 3H, CH3). MS (EI)
m/z (%): 209 (M+, 100); 194 (35); 181 (25); 165 (22); 148 (10);
126 (6); 111 (27); 86 (33). HRMS: 209.01160, C7H7N5OS
requires 209.03713.
[1,3]Thiazolo[3,2-a][1,2,3]triazolo[4,5-d]pyrimidin-9(1H)-
one, 5d. Crystalline sodium nitrite (0.22 g, 3.2 mmol) was
slowly added to a stirred suspension of 13 (0.5 g, 2.7 mmol) in
20 mL of HCl 6 M in an ice bath. After the addition was
complete, the solution was left to stand at room temperature
for 4 h, heated at 50 °C for 40 min, and then concentrated to
half volume and cooled. The resulting precipitate consisted of
pure 5d (80%). Mp 265-266 °C. 1H NMR (DMSO-d6): δ 11.24
(bs exch, 1H, NH), 8.08 (d, 1H, H-3), 7.42 (d, 1H, H-2). MS

BIQs: Antagonists at Human A1 and A3 ARs
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26 8259
(EI) m/z (%): 193 (M+, 100); 165 (6); 110 (51). HRMS:
Human A3 Adenosine Receptors. Aliquots of cell mem-
branes (40 µg) were incubated at 25 °C for 90 min in 100 µL
of T3h buffer (50 mM Tris-HCl, 10 mM MgCl2, 1 mM EDTA,
2 units/mL ADA, pH 7.4) in the presence of 1.4 nM [125I]AB-
MECA and six different concentrations of the newly synthe-
sized compounds. Nonspecific binding was determined in the
presence of 50 µM RPIA. The dissociation constant (Kd) of [125I]
AB-MECA in A3 CHO cell membranes was 1.4 nM.
All compounds were routinely dissolved in DMSO and
diluted with assay buffer to the final concentration, where the
amount of DMSO never exceeded 2%. Percentage inhibition
values of specific radiolabeled ligand binding at 1-10 µM
concentration are means ( SEM of at least three determina-
tions. At least six different concentrations spanning 3 orders
of magnitude, adjusted appropriately for the IC50 of each
compound, were used. IC50 values, computer-generated using
a nonlinear regression formula on a computer program (Graph-
Pad, San Diego, CA), were converted to Ki values, knowing
the Kd values of radioligands in the different tissues and using
the Cheng and Prusoff equation.44 Ki values are means ( SEM
of at least three determinations.
193.03520, C6H3N5OS requires 193.00583.
General Procedure for the Synthesis of 3-(Methylth-
io)benzene-1,2-diamine, 7d, and 3-(Ethylthio)benzene-
1,2-diamine, 7e. A solution of 3-(methylthio)-2-nitroaniline
(10d)39 or 3-(ethylthio)-2-nitroaniline (10e) (3.3 mmol) and tin-
(II)chloride dihydrate (3.72 g, 16.5 mmol) in ethyl acetate (30
mL) was heated at reflux for 2 h. After cooling, the reaction
mixture was diluted with 70 mL of ethyl acetate and poured
into ice-water (200 mL). While stirring, the pH was adjusted
to 8 with saturated sodium bicarbonate solution, and the
precipitate that formed was filtered off. The aqueous phase
was separated and extracted three times with 50 mL portions
of ethyl acetate. The combined organic phases were dried (Na2-
SO4) and evaporated to dryness to give the desired compounds
7d (86%) or 7e (92%), respectively. These compounds were
employed without further purification.
7d: 1H NMR (CDCl3): δ 6.96-6.93 (m, 1H, H-Ar), 6.69-
6.64 (m, 2H, H-Ar), 4.35-3.23 (bs, 4H, 2 × NH2), 2.36 (s, 3H,
CH3). ESI (CH3OH): 178.8 [M + Na+].
7e: 1H NMR (CDCl3): 6.98-6.94 (m, 1H, H-Ar), 6.75-6.64
(m, 2H, H-Ar), 4.25-3.21 (bs, 4H, 2 × NH2), 2.75 (q, 2H, CH2),
1.22 (t, 3H, CH3). ESI (CH3OH): 191.6 [M + Na+].
Acknowledgment. This work was financially sup-
ported by MIUR (Cofin 2003, ex 40%).
3-(Ethylthio)-2-nitroaniline, 10e. A solution of NaSEt
(0.58 g, 6.9 mmol) in DMF (7 mL) was added dropwise to a
stirred solution of 3-chloro-2-nitroaniline (9) (1.0 g, 5.8 mmol)
in DMF (25 mL) at 20 °C. After the addition was complete,
the mixture was stirred for 1 h at 40 °C and refluxed for 2 h.
The solvent was evaporated to dryness, and the residue was
chromatographed (ethyl acetate-n-hexane, 7:3 v/v as the
eluant) to give 10e (30%). 1H NMR (CDCl3): δ 7.16 (t, 1H,
H-5), 6.60 (d, 1H, H-4), 6.54 (d, 1H, H-6), 5.80 (bs, 2H, NH2),
2.89 (q, 2H, CH2), 1.34 (t, 3H, CH3).
Supporting Information Available: Physical and spec-
tral data of 3 and 4 and analytical data of 3-5. This material
is available free of charge via the Internet at http://pubs.ac-
s.org.
References
(1) Wardas, J. Neuroprotective Role of Adenosine in the CNS. Pol.
J. Pharmacol. 2002, 54, 313-326.
(2) Ribeiro, J. A.; Sebastia˜o, A. M.; de Mendonc¸a, A. Adenosine
Receptors in the Nervous System: Pathophysiological Implica-
tions. Prog. Neurobiol. (Amsterdam, Neth.) 2003, 68, 377-392.
(3) Poulsen, S.-A.; Quinn, R. J. Adenosine Receptors: New Op-
portunities for Future Drugs. Bioorg. Med. Chem. 1998, 6, 619-
641.
6,7-Diamino-5H-[1,3]thiazolo[3,2-a]pyrimidin-5-one, 13.
Propan-2-ol (32 mL) was added to a solution of NaOH 2 M (32
mL) containing 4,5-diamino-6-hydroxy-2-mercaptopyrimidine
(12) (5.0 g, 31 mmol). A light precipitate was immediately
formed. cis-1,2-Dichloroethene (2.49 mL, 33 mmol) was then
slowly added to the reaction mixture at room temperature, and
stirring was continued at the same temperature for 2 h and
then at 60-70 °C for a further 5 h. After cooling, the solvents
were removed, and the residue was treated with H2O (20 mL)
and exhaustively extracted with ethyl acetate. The organic
extracts were dried over Na2SO4 and then evaporated to
dryness to yield a solid that was purified by flash-chromatog-
raphy (CH2Cl2-MeOH, 9:1 v/v as the eluant) to give 13 (10%).
(4) Rorke, S.; Holgate, S. T. Targeting Adenosine Receptors: Novel
Therapeutic Targets in Asthma and Chronic Obstructive Pul-
monary Disease. Am. J. Respir. Med. 2002, 1, 99-105.
(5) Nishiyama, A.; Rahman, M.; Inscho, E. W. Role of Interstitial
ATP and Adenosine in the Regulation of Renal Hemodynamics
and Microvascular Function. Hypertens. Res. 2004, 27, 791-804.
(6) Fredholm, B. B.; IJzerman, A. P.; Jacobson, K. A.; Klotz, K.-N.;
Linden, J. International Union of Pharmacology. XXV. Nomen-
clature and Classification of Adenosine Receptors. Pharmacol.
Rev. 2001, 53, 527-552.
(7) Dhalla, A. K.; Shryock, J. C.; Shreeniwas, R.; Belardinelli, L.
Pharmacology and Therapeutic Applications of A1 Adenosine
Receptor Ligands. Curr. Top. Med. Chem. 2003, 3, 369-385.
(8) Ferre, S.; Fredholm, B. B.; Morelli, M.; Popoli, P.; Fuxe, K.
Adenosine-Dopamine Receptor-Receptor Interactions as an
Integrative Mechanism in the Basal Ganglia. Trends Neurosci.
1997, 20, 482-487.
1
Mp 170-172 °C. H NMR (DMSO-d6): δ 7.76 (d, 1H, J ) 5.0
Hz, H-3), 7.20 (d, 1H, J ) 5.0 Hz, H-2), 6.12 (s, 2H, 7-NH2),
3.64 (s, 2H, 6-NH2).
Biological Methods. Materials. [3H]DPCPX, [3H] NECA,
and [125I]AB-MECA were obtained from DuPont-NEN (Boston,
MA). Adenosine deaminase (ADA) was from Sigma Chemical
Co. (St. Louis, MO). All other reagents were from standard
commercial sources and of the highest commercially available
grade.
(9) Franco, R.; Ferre, S.; Agnati, L.; Torvinen, M.; Gines, S.; Hillion,
J.; Casado, V.; Lledo, P.; Zoli, M.; Lluis, C.; Fuxe, K. Evidence
for Adenosine/Dopamine Receptor Interactions: Indications for
Heteromerization. Neuropsychopharmacology 2000, 23, S50-
59.
(10) Fink, J. S.; Weaver, D. R.; Rivkees, S. A.; Peterfreund, R. A.;
Pollack, A. E.; Adler, E. M.; Reppert, S. M. Molecular Cloning
of the Rat A2 Adenosine Receptor: Selective co-Expression with
D2 Dopamine Receptors in Rat Striatum. Mol. Brain Res. 1992,
14, 186-195.
Adenosine Receptor Binding Assay. Human A1 Ad-
enosine Receptors. Aliquots of membranes (50 µg proteins)
obtained from A1 CHO cells were incubated at 25 °C for 180
min in 500 µL of T1b/T1h buffer (50 mM Tris-HCl, 2 mM MgCl2,
pH 7.7)/(50 mM Tris-HCl, 2 mM MgCl2, 2 units/mL ADA, pH
7.4) containing [3H]DPCPX (3 nM) and six different concentra-
tions of the newly synthesized compounds. Nonspecific binding
was determined in the presence of 50 µM RPIA.
(11) Cacciari, B.; Pastorin, G.; Spalluto, G. Medicinal Chemistry of
A2A Adenosine Receptor Antagonists. Curr. Top. Med. Chem.
2003, 3, 403-411.
(12) Fredholm, B. B.; Cuhna, R. A.; Svenningsson, P. Pharmacology
of Adenosine A2A Receptors and Therapeutic Applications. Curr.
Top. Med. Chem. 2003, 3, 413-426.
The dissociation constant (Kd) of [3H]DPCPX A1 CHO cell
membranes was 3 nM.
(13) Linden, J. Cloned Adenosine A3 Receptors: Pharmacological
Properties, Species Differences and Receptor Functions. Trends
Pharmacol. Sci. 1994, 15, 298-306.
Human A2A Adenosine Receptors. Aliquots of cell mem-
branes (80 µg) were incubated at 25 °C for 90 min in 500 µL
of T2h buffer (50 mM Tris-HCl, 2 mM MgCl2, 2 units/mL ADA,
pH 7.4) in the presence of 30 nM of [3H]NECA and six different
concentrations of the newly synthesized compounds. Nonspe-
cific binding was determined in the presence of 100 µM NECA.
The dissociation constant (Kd) of [3H]NECA in A2A CHO cell
membranes was 30 nM.
(14) Jacobson, K. A.; Kim, H. O.; Siddiqui, S. M.; Olah, M. E.; Stiles,
G. L.; Von Lubitz, D. K. J. E. A3-Adenosine Receptors: Design
of Selective Ligands and Therapeutic Prospects. Drugs Future
1995, 20, 689-701.
(15) Mu¨ller, C. E. Medicinal Chemistry of Adenosine A3 Receptor
Ligands. Curr. Top. Med. Chem. 2003, 3, 445-462.
(16) Fishman, P.; Bar-Yehuda, S. Pharmacology and Therapeutic
Applications of A3 Receptor Subtypes. Curr. Top. Med. Chem.
2003, 3, 463-469.

8260 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 26
Novellino et al.
(17) Martin, Y. C. 3D Database Searching in Drug Design. J. Med.
Chem. 1992, 35, 2145-2154.
(18) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J.
Experimental and Computational Approaches to Estimate Solu-
bility and Permeability in Drug Discovery and Developing
Settings. Adv. Drug Delivery Rev. 1997, 23, 3-25.
(19) Guandalini, L.; Martini, E.; Dei, S.; Manetti, D.; Scapecchi, S.;
Teodori, E.; Romanelli, M. N.; Varani, K.; Greco, G.; Spadola,
L.; Novellino, E. Design of Novel Nicotinic Ligands through 3D
Database Searching. Bioorg. Med. Chem. 2005, 13, 799-807.
(20) Enyedy, I. J.; Sakamuri, S.; Zaman, W. A.; Johnson, K. M.; Wang,
S. Pharmacophore-Based Discovery of Substituted Pyridines as
Novel Dopamine Transporter Inhibitors. Bioorg. Med. Chem.
Lett. 2003, 13, 513-517.
(21) Chen, G. S.; Chang, C.-S.; Kan, W. M.; Chang, C.-L.; Wang, K.
C.; Chern, J.-W. Novel Lead Generation through Hypothetical
Pharmacophore Three-Dimensional Database Searching: Dis-
covery of Isoflavonoids as Nonsteroidal Inhibitors of Rat 5R-
Reductase. J. Med. Chem. 2001, 44, 3759-3763.
(22) Zaharevitz, D. W.; Gussio, R.; Wiegand, A.; Jalluri, R.; Patta-
biraman, N.; Kellogg, G. E.; Pallansch, L. A.; Yang, S. S.;
Buckheit, R. W., Jr. Discovery of Novel HIV-1 Reverse Tran-
scriptase Inhibitors Using a Combination of 3D Database
Searching and 3D QSAR. J. Med. Chem. 1999, 9, 551-564.
(23) Marriott, D. P.; Dougall, I. G.; Meghani, P.; Liu, Y.-J.; Flower,
D. R. Lead Generation Using Pharmacophore Mapping and
Three-Dimensional Database Searching: Application to Mus-
carinic M3 Receptor Antagonists. J. Med. Chem. 1999, 42, 3210-
3216.
(24) Wang, S.; Zaharevitz, D. W.; Sharma, R.; Marquez, V. E.; Lewin,
N. E.; Du, L.; Blumberg, P. M.; Milne, G. W. The Discovery of
Novel, Structurally Diverse Protein Kinase C Agonists through
Computer 3D-Database Pharmacophore Search. Molecular Mod-
eling Studies. J. Med. Chem. 1994, 37, 4479-4489.
(25) Allen, F. H.; Bellard, S.; Brice, M. D.; Cartwright, B. A.;
Doubleday, A.; Higgs, H.; Hummelimk, T.; Hummelink-Peters,
B. G.; Kennard, O.; Motherwell, W. D. S.; Rodgers, J. R.; Watson,
D. G. The Cambridge Crystallographic Data Centre: Computer-
Based Search, Retrieval, Analysis and Display of Information.
Acta Crystallogr., Sect. B: Struct. Sci. 1979, 35, 2231-2239.
(26) Suzuki, F.; Shimada, J.; Shiozaki, S.; Ichikawa, S.; Ishii, A.;
Nakamura, J.; Nonaka, H.; Kobayashi, H.; Fuse, E. Adenosine
A1 Antagonists. 3. Structure-Activity Relationships on Amelio-
ration against Scopolamine- or N6-((R)-Phenylisopropyl)adenos-
ine-Induced Cognitive Disturbance. J. Med. Chem. 1993, 36,
2508-2518.
(27) van Galen, P. J. M.; Nissen, P.; van Wijngaarden, I.; IJzerman,
A. P.; Soudijn, W. 1H-Imidazo[4,5-c]quinolin-4-amines: Novel
Non-Xanthine Adenosine Antagonists. J. Med. Chem. 1991, 34,
1202-1206.
(28) Saki, M.; Tsumuki, H.; Nonaka, H.; Shimada, J.; Ichimura, M.
KF26777 (2-(4-Bromophenyl)-7,8-dihydro-4-propyl-1H-imidazo-
[2,1-i]purin-5(4H)-one), a New Potent and Selective Adenosine
A3 Receptor Antagonist. Eur. J. Pharmacol. 2002, 444, 133-
141.
(29) Francis, J. E.; Cash, W. D.; Psychoyos, S.; Ghai, G.; Wenk, P.;
Friedmann, R. C.; Atkins, C.; Warren, V.; Furness, P.; Hyun, J.
L.; Stone, G. A.; Desai, M.; Williams, M. Structure-Activity
Profile of a Series of Novel Triazoloquinazoline Adenosine
Antagonists. J. Med. Chem. 1988, 31, 1014-1020.
(30) 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 Inves-
tigation 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.
(31) van Muijlwijk-Koezen, J. E.; Timmerman, H.; van der Goot, H.;
Menge, W. M. P. B.; Frijtag von Drabbe Kunzel, J.; de Groote,
M.; IJzerman, A. P. Isoquinoline and Quinazoline Urea Ana-
logues as Antagonists for the Human Adenosine A3 Receptors.
J. Med. Chem. 2000, 43, 2227-2238.
(32) Bruno, I. J.; Cole, J. C.; Edgington, P. R.; Kessler, M.; Macrae,
C. F.; McCabe, P.; Pearson, J.; Taylor, R. New Software for
Searching the Cambridge Structural Database and Visualizing
Crystal Structures. Acta Crystallogr., Sect. B: Struct. Sci. 2000,
58, 389-397.
(33) Kanoktanaporn, S.; MacBride, J. A. H.; King, T. J. Displacement
of Hydrazine Hydrochloride from 1,4-Dichlorophthalazines by
Bidentate Nucleophiles with Formation of 2,2′-(1,2-Arylene)bis-
Benzimidazoles and -Benzothiazoles. X-ray Crystal Structure of
2,3-Bisbenzimidazol-2-ylquinoxaline. J. Chem. Res., Synop. 1980,
12, 406.
(34) Prins, R.; Birker, P. J. M. L.; Verschoor, G. C. Structure of 2-[5-
(2-Pyridyl)-1H-1,2,4-triazol-3-yl]pyridinium Perchlorate Acta Crys-
tallogr., Sect. B: Struct. Sci. 1982, 38, 2934-2935.
(35) Kohne, B.; Praefcke, K.; Derz, T.; Gondro, T.; Frolow, F.
Benzenetris(imidazole). A New Ring System from Benzenehex-
amine. Angew. Chem., Int. Ed. Engl. 1986, 25, 650-651.
(36) Heinisch, G.; Matuszczak, B.; Mereiter, K. Pyridazines. 71. A
Novel Type of 1,2-Diazine by Ring Contraction of 1,2-Diazole.
Heterocycles 1994, 38, 2081-2089.
(37) Pecorari, P.; Rinaldi, M.; Costi, M. P.; Antolini, L. Cyclization
Reactions of 1,3-Dibromopropan-2-ol, 2,3-Dibromopropan-1-ol
and 1-Bromomethyloxirane with 6-Amino-2,3-dihydro-2-thioxo-
4(1H)-pyrimidinone. J. Heterocycl. Chem. 1991, 28, 891-898.
(38) Mizutani, T.; Yoshida, K.; Kawazoe, S. Formation of Toxic
Metabolites from Thiabendazole and Other Thiazoles in Mice.
Identification of Thioamides as Ring Cleavage Products. Drug
Metab. Dispos. 1994, 22, 750-755.
(39) Hay, M. P.; Gamage, S. A.; Kovacs, M. S.; Prujin, F. B.; Anderson,
R. F.; Patterson, A. V.; Wilson, W. R.; Brown, J. M.; Denny, W.
A. Structure-Activity Relationships of 1,2,4-Benzotriazine 1,4-
Dioxides as Hypoxia-Selective Analogues of Tirapazamine. J.
Med. Chem. 2003, 46, 169-182.
(40) Wolff, J. J.; Limbach, H.-H. Synthesis and Spectroscopic Char-
acterization of 15N-Labeled Hexaaminobenzene Derivatives.
Liebigs Ann. Chem. 1991, 691-693.
(41) Pecorari, P.; Rinaldi, M.; Costi, M. P. New Heterocyclic Struc-
tures. [1,3]Thiazino[3,2-a]purine and [1,2,3]Triazolo[4,5-d][1,3]-
thiazino[3,2-a]pyrimidine. J. Heterocycl. Chem. 1989, 26, 1701-
1705.
(42) Bahner, C. T.; Bilancio, D. E. 5-Thiol-7-hydroxy-1-γ-triazolo[d]-
pyrimidine, J. Am. Chem. Soc. 1953, 75, 6038.
(43) Da Settimo, F.; Primofiore, G.; Taliani, S.; La Motta, C.;
Novellino, E.; Greco, G.; Lavecchia, A.; Cosimelli, B.; Iadanza,
M.; Klotz, K.-N.; Tuscano, D.; Trincavelli, M. L.; Martini, C. A1
Adenosine receptor Antagonists, 3-Aryl[1,2,4]triazinobenzimidazol-
4-(10H)-ones (ATBIs) and N-alkyl and N-acyl-(7-substituted-2-
phenylimidazo[1,2-a][1,3,5]triazin-4-yl)amines (ITAs): Different
Recognition of Bovine and Human Binding Sites. Drug Dev. Res.
2004, 63, 1-7.
(44) Cheng, Y. C.; Prusoff, W. H. Relationship between the Inhibition
constant (Ki) and the Concentration of Inhibitor which causes
50% Inhibition (IC50) of an Enzymatic Reaction. Biochem.
Pharmacol. 1973, 22, 3099-3108.
JM050792D