Derivatives of Benzimidazol-2-ylquinoline and Benzimidazol-2-ylisoquinoline as Selective A1 Adenosine Receptor Antagonists with Stimulant Activity on Human Colon Motility

Article abstract of DOI:10.1002/cmdc.201100284

A number of quinolines and isoquinolines connected in various ways to a substituted benzimidazol-2-yl system were synthesized and evaluated as novel antagonists of adenosine receptors (ARs) by competition experiments using human A₁, A_2A, and A₃ ARs. The new compounds were designed based on derivatives of 2-(benzimidazol-2-yl)quinoxaline, previously reported as potent and selective antagonists of A₁ and A₃ ARs. Among these, 3-[4-(ethylthio)-1H-benzimidazol-2-yl]isoquinoline 4b exhibited the best combination of potency toward the A₁ AR (K_i=1.4nM) and selectivity against the A_2A (K_i>10μm), A_2B (K_i>10μm), and A₃ ARs (K_i>1μM). Functional experiments in circular smooth muscle preparations of isolated human colon showed that 4b behaves as a potent and selective antagonist of the A₁ AR in the neuromuscular compartment of this intestinal region. Biological and pharmacological data suggest that 4b is a suitable starting point for the development of novel agents endowed with stimulant properties on colonic activity.

Full text of DOI:10.1002/cmdc.201100284

DOI: 10.1002/cmdc.201100284
Derivatives of Benzimidazol-2-ylquinoline and
Benzimidazol-2-ylisoquinoline as Selective A1 Adenosine
Receptor Antagonists with Stimulant Activity on Human
Colon Motility
Barbara Cosimelli,*[a] Sabrina Taliani,*[b] Giovanni Greco,[a] Ettore Novellino,[a] Annalisa Sala,[a]
Elda Severi,[a] Federico Da Settimo,[b] Concettina La Motta,[b] Isabella Pugliesi,[b]
Luca Antonioli,[c] Matteo Fornai,[c] Rocchina Colucci,[c] Corrado Blandizzi,[c] Simona Daniele,[d]
Maria Letizia Trincavelli,[e] and Claudia Martini[e]
A number of quinolines and isoquinolines connected in various
ways to a substituted benzimidazol-2-yl system were synthe-
sized and evaluated as novel antagonists of adenosine recep-
tors (ARs) by competition experiments using human A1, A2A,
and A3 ARs. The new compounds were designed based on de-
rivatives of 2-(benzimidazol-2-yl)quinoxaline, previously report-
ed as potent and selective antagonists of A1 and A3 ARs.
Among these, 3-[4-(ethylthio)-1H-benzimidazol-2-yl]isoquino-
line 4b exhibited the best combination of potency toward the
A1 AR (Ki =1.4 nm) and selectivity against the A2A (Ki >10 mm),
A2B (Ki >10 mm), and A3 ARs (Ki >1 mm). Functional experiments
in circular smooth muscle preparations of isolated human
colon showed that 4b behaves as a potent and selective an-
tagonist of the A1 AR in the neuromuscular compartment of
this intestinal region. Biological and pharmacological data sug-
gest that 4b is a suitable starting point for the development
of novel agents endowed with stimulant properties on colonic
activity.
Introduction
Adenosine is an endogenous signaling nucleoside that is re-
leased from several cells and plays a key role in a variety of
physiological processes.[1–4] Once in the extracellular space, ad-
enosine can affect cell functioning by triggering specific G-pro-
tein-coupled receptors which modulate a number of effector
systems, including adenylate cyclase, potassium and calcium
channels, phospholipase C or D, and guanylate cyclase.[5] Four
adenosine receptor (AR) subtypes, classified as A1, A2A, A2B, and
A3, have been identified as mediators of these effects.[3] Each
receptor subtype has been indicated as a target for the devel-
opment of agonist- and antagonist-based therapies against a
wide range of pathologies, including central nervous system
(CNS) disorders, cardiac arrhythmia, ischemic injuries, asthma,
renal failure, and inflammatory diseases.[1,6,7] The A1 AR has
been extensively investigated due to its important biological
role in various tissues such as CNS, heart, kidneys, liver, and
bladder.[8] Selective A1 AR antagonists may be useful in cogni-
tion enhancement for the treatment of various forms of de-
mentia, including Alzheimer’s disease.[9–11] A1 AR antagonists
are also under development as potassium-sparing diuretics
with kidney-protective properties and for treatment of acute
renal failure.[8,12,13]
of inflammatory intestinal conditions, the modulation of ARs or
enzymes driving adenosine metabolism can favorably affect
the outcome of inflammation as well as the related digestive
motor alterations.[15–21] Studies on the effects of novel A1 AR
antagonists on the contractile function of human colon could
pave the way to the discovery of new options for the thera-
peutic management of digestive disorders characterized by de-
creased gastrointestinal propulsion, such as idiopathic chronic
constipation, irritable bowel syndrome with predominance of
constipation, and post-operative paralytic ileus.[19,22] Postopera-
tive ileus (POI) is a common adverse event associated with ab-
[a] Prof. B. Cosimelli, Prof. G. Greco, Prof. E. Novellino, Dr. A. Sala, Dr. E. Severi
Dipartimento di Chimica Farmaceutica e Tossicologica
Universitꢀ di Napoli “Federico II”, Via D. Montesano 49, 80131 Napoli (Italy)
E-mail: barbara.cosimelli@unina.it
[b] Dr. S. Taliani, Prof. F. Da Settimo, Dr. C. La Motta, Dr. I. Pugliesi
Dipartimento di Scienze Farmaceutiche
Universitꢀ di Pisa, Via Bonanno 6, 56126 Pisa (Italy)
E-mail: taliani@farm.unipi.it
[c] Dr. L. Antonioli, Dr. M. Fornai, Dr. R. Colucci, Prof. C. Blandizzi
Centro Interdipartimentale di Ricerche di Farmacologia Clinica e Terapia
Sperimentale, Universitꢀ di Pisa, Via Roma 55, 56126 Pisa (Italy)
[d] Dr. S. Daniele
Fornai et al.[14] demonstrated that adenosine contributes sig-
nificantly to the enteric regulatory network that is dependent
on modulation of human colonic motility. In particular, it was
observed that the A1 AR is actively involved in tonic inhibitory
actions exerted by endogenous adenosine on the excitatory
cholinergic motility of the colon. Furthermore, in the presence
Department of Drug Discovery and Development
Istituto Italiano di Tecnologia, via Morego, 30, 16163 Genova (Italy)
[e] Dr. M. L. Trincavelli, Prof. C. Martini
Dipartimento di Psichiatria, Neurobiologia, Farmacologia e Biotecnologie
Universitꢀ di Pisa, Via Bonanno 6, 56126 Pisa (Italy)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201100284.
ChemMedChem 2011, 6, 1909 – 1918
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1909

B. Cosimelli, S. Taliani, et al.
MED
dominal surgical procedures and is believed to occur as a
result of inhibitory neural reflexes and local inflammatory
processes.[23] Prolonged POI can lead to patient discomfort,
decreased mobility, delayed enteral feeding, and, ultimate-
ly, prolonged hospitalizations and increased social costs.
At present, limited pharmacological strategies are available
to effectively and successfully decrease the duration of
POI.[23] Blockade of the A1 AR by highly potent and selec-
tive antagonists may provide a novel option to favorably
affect POI duration.
Table 1. Affinities for human A1, A2A, and A3 ARs of the newly investigated com-
pounds 2–5, along with selected previously reported compounds 1.
Compd R1
R2 R3
Ki [nm][a]
hA2A
[b]
[c]
[d]
hA1
hA3
1a[e]
1b[e]
1c[e]
1d[e]
1i[e]
2a
2b
2c
2d
2e
2 f
2g
2h
2i
2j
2k
2l
2m
2n
2o
2p
3a
3b
3c
3d
3e
3 f
3g
3h
3i
3j
3k
3l
H
H
H
H
H
H
H
H
H
50.0ꢀ15.0
0.50ꢀ0.01
561.0ꢀ17.0
3440.0ꢀ297.0
3000.0ꢀ200.0
833.0ꢀ67.0
763.0ꢀ13.0
955.0ꢀ88.0
278.0ꢀ44.0
26.0ꢀ9.0
SC2H5
SCH3
CH3
H
2.10ꢀ0.30
8000.0ꢀ567.0
CH3 CH3 >10000
H
H
H
>10000
370.0ꢀ35.0
225.2ꢀ13.0
280.2ꢀ15.0
685.1ꢀ111.0
178.1ꢀ17.0
772.9ꢀ4.0
413.3ꢀ16.0
828.0ꢀ15.0
174.0ꢀ17.2
H
H
H
H
H
H
H
Cl
13.0ꢀ0.1
>10000
>10000
>10000
1028.0ꢀ95.1
>10000
>10000
>10000
SCH3
CH3
Cl
H
H
CH3
Cl
H
H
CH3
–
We recently described the synthesis and biological eval-
uation of 2-(1H-benzimidazol-2-yl)quinoxalines 1 as A1 and
A3 AR antagonists, which were designed on the basis of
database searches and lead optimization approaches
(Figure 1).[24] Specifically, the Cambridge Structural Data-
base (CSD)[25] (version 5.25) was scanned using a pharma-
H
CH3
Cl
H
20.7ꢀ1.5
1342.3ꢀ65.1
9.1ꢀ0.4
2.9ꢀ0.5
>10000
>10000
>10000
>10000
>10000
>10000
H
Cl >10000
CH3 CH3
Cl Cl >10000
H
–
–
–
–
–
H
H
H
H
CH3
Cl
H
H
CH3 CH3 >10000
Cl Cl >10000
H
–
–
–
–
H
H
CH3
Cl
H
H
CH3
Cl
14.5ꢀ1.1
>1000
>1000
660.2ꢀ65.1
653.0ꢀ41.1
934.4ꢀ56.2 >1000
1330.3ꢀ129.1 144.1ꢀ14.0
>10000
94.1ꢀ5.0
>10000 >1000
1330.4ꢀ105.2 415.0ꢀ40.1
158.0ꢀ7.2 >1000
277.1ꢀ18.0 >1000
CH3 >10000
–
–
–
–
–
H
H
H
H
H
H
>10000
>10000
>10000
>10000
>10000
776.0ꢀ14.2
–
–
–
–
H
41.0ꢀ3.1
SC2H5
CH3
Cl
H
H
CH3
Cl
H
H
CH3
–
–
–
–
H
SC2H5
H
H
H
SC2H5
H
273.2ꢀ33.1
4.2ꢀ0.3
4.4ꢀ0.1
195.2ꢀ15.0
152.0ꢀ11.3
146.0ꢀ11.1
1520.1ꢀ150.2
739.3ꢀ33.1
347.3ꢀ31.1
2744.1ꢀ260.0
445.1ꢀ28.2
162.0ꢀ12.0
19.3ꢀ1.6
200.3ꢀ14.1
307.2ꢀ24.0
272.0ꢀ16.1
509.4ꢀ41.2
19.2ꢀ2.0
Cl >10000
Cl >10000
143.2ꢀ5.3
533.8ꢀ52.5
285.2ꢀ28.0
452.1ꢀ36.2
CH3 >10000
–
–
–
>10000
Figure 1. Design strategy leading to the discovery of (1H-benzimida-
zol-2-yl)quinoxalines 1 as A1 and A3 AR antagonists.
441.4ꢀ44.2
>10000
60.2ꢀ4.5
276.0ꢀ30.1
199.2ꢀ20.0
3m
3n
3o
4a
4b
4e
4 f
861.0ꢀ62.0 >1000
819.5ꢀ64.3 >1000
–
158.1ꢀ13.2
3.2ꢀ0.2
1.4ꢀ0.1
3.5ꢀ0.3
4.3ꢀ0.4
14.1ꢀ1.6
H
H
H
H
H
H
H
H
3750.0ꢀ245.0
156.0ꢀ13.4
>1000
264.7ꢀ23.0
399.5ꢀ20.0
>1000
>1000
>1000
>1000
4000
cophore substructure with a 3D disposition defined by tor-
sion angles. One of the hits resulting from this search was
2,3-bis(1H-benzimidazol-2-yl)quinoxaline (CSD reference
code bzquxl).[24] This structure was translated synthetically
into analogues devoid of one of the two benzimidazole
systems, based on the assumption that the less bulky (1H-
benzimidazol-2-yl)quinoxalines 1 would be more suitable
for binding to the target receptors (Figure 1). Among the
tested compounds from series 1 (see Figure 4 below), the
compounds exhibiting the highest affinity for A1 and A3
ARs were 2-[4-(ethylthio)-1H-benzimidazol-2-yl]quinoxaline
1b (Ki values at the A1, A2A, and A3 ARs of 0.5, 3440, and
955 nm, respectively) and 2-[4-methyl-1H-benzimidazol-2-
yl]quinoxaline 1d (Ki values at the A1, A2A, and A3 ARs of
8000, 833, and 26 nm, respectively) (Table 1).[24]
>10000
>10000
>10000
>10000
>10000
>10000
>10000
130
5a
5b
5e
5 f
DPCPX
NECA
>10000
6.6ꢀ0.5
6.3ꢀ0.4
3.9
14.0
120
H
20
2100
6.2
11
Cl-IBMECA
[a] Ki values represent the mean ꢀSEM of at least three determinations obtained
from an iterative curve-fitting procedure (GraphPad Prism, GraphPad). [b] Dis-
placement of specific [3H]DPCPX binding in membranes obtained from CHO
cells stably expressing hA1 AR. [c] Displacement of specific [3H]NECA binding in
membranes obtained from CHO cells stably expressing hA2A AR. [d] Displace-
ment of specific [125I]AB-MECA binding in membranes obtained from CHO cells
stably expressing hA3 AR. [e] Data from reference [24].
Prompted by the ease of synthesis for compounds in
series 1 (obtained by condensing the quinoxaline-2-car-
boxylic acid with the appropriate o-phenylenediamine), we
turned our attention to analogues of 1 that could be similarly
prepared by a one-step procedure. A majority of the new scaf-
folds 2–5 were made by connecting a quinoline or isoquino-
line nucleus to a substituted 1H-benzimidazol-2-yl system
(Figure 2). All of the designed structures share a common phar-
macophore composed of two nitrogen atoms with hydrogen
bonding properties connected by two or three sp2 carbons.
The substituents attached to the benzimidazole ring were
chosen from those that exhibited the best results in our previ-
ous work[24] (CH3, Cl, SCH3, SC2H5, or a fused benzene-cyclohex-
ane ring to mimic disubstitution).
1910
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1909 – 1918

Selective A1 Adenosine Receptor Antagonists
Scheme 1. Synthesis of compounds 2a–p.
Figure 2. Scaffolds of novel potential AR antagonists 2–5 designed as ana-
logues of (1H-benzimidazol-2-yl)quinoxalines 1.
Herein we describe the synthesis and biological and phar-
macological evaluation of the following analogues of
1
(Table 1): 2-(1H-benzimidazol-2-yl)quinolines (2a–m and p) and
their aza isosteres (2n and o), 3-(1H-benzimidazol-2-yl)quino-
lines (3a–m) and their aza isosteres (3n and o), 3-(1H-benzimi-
dazol-2-yl)isoquinolines (4a, b, e, and f), and 1-(1H-benzimida-
zol-2-yl)isoquinolines (5a, b, e, and f).
Results and Discussion
Chemistry
Compounds 2a–p and 3a–o were prepared beginning from
quinoline-2-carboxylic acid 6 or quinoline-3-carboxylic acid 7,
respectively, heated in polyphosphoric acid with the appropri-
ate o-phenylenediamine (Figure 3). This synthetic procedure
(Scheme 1 and 2) was similar to the reported synthesis of 1 be-
ginning from quinoxaline-2-carboxylic acid;[24] however, heat-
Scheme 2. Synthesis of compounds 3a–o.
ing temperatures and workup protocols were optimized to
give improved yields (see Experimental Section for details).
Series 4 compounds, as well as the compounds in series 5,
were obtained analogously (Scheme 3 and 4 and Experimen-
tal), beginning from isoquinoline-3-carboxylic acid 26 and iso-
quinoline-1-carboxylic acid 27.
The o-phenylenediamines reported in Figure 3 were all com-
mercially available, with the exception of 9, 20, 21, and 22.
Syntheses of 9 and 22 were performed in accordance with our
previously reported procedure.[24] Diamines 20 and 21 were
obtained beginning from 2-amino-5,6,7,8-tetrahydronaphtha-
lene 28 according to Scheme 5. Briefly, the starting material
was initially treated with nitric acid and acetic anhydride to
yield nitro derivatives 29 and 30, which were then separated
by flash chromatography (Scheme 5). Deacetylation and reduc-
tion furnished compounds 20 and 21 via intermediates 31 and
32, respectively (Scheme 5).
Figure 3. Diamines for the syntheses of compounds 2–5.
ChemMedChem 2011, 6, 1909 – 1918
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1911

B. Cosimelli, S. Taliani, et al.
MED
([3H]NECA), or [125I]4-aminobenzyl-5’-N-methylcarboxamidoade-
nosine ([125I]AB-MECA) binding from transfected CHO cells. Ex-
periments were performed as previously described.[26] Com-
pound 4b and lead 1b were tested in functional assays
against human A2B AR by measuring their effect on NECA-
mediated cAMP modulation in transfected CHO cells.
Compound 4b, which showed high affinity and selectivity
for the A1 AR, was also evaluated for its A1 AR antagonistic
properties on electrically evoked cholinergic contractions in cir-
cular smooth muscle preparations of isolated human colon. In
these experiments, lead compound 1b was also tested for pur-
poses of comparison, and the A1 AR antagonists DPCPX and 8-
cyclopentyl-N3-[3-(4-(fluorosulfonyl)benzoyloxy)propyl]-N1-pro-
pylxanthine (FSCPX) were used as reference compounds. To
substantiate the pharmacological results, the expression of A1
ARs in colonic circular smooth muscle was evaluated by means
of reverse transcription–polymerase chain reaction (RT-PCR)
analysis.
Scheme 3. Synthesis of compounds 4a, b, e, and f.
Discussion
Table 1 lists the binding affinities
for the human A1, A2A, and A3 ARs
of the newly investigated com-
pounds of series 2–5 expressed as
Ki values. Some of the previously
disclosed compounds of series 1[24]
(Figure 4) are also reported in
Table 1 for comparison purposes.
As a general trend, the molecu-
lar scaffolds considered here allow
Scheme 4. Synthesis of compounds 5a, b, e, and f.
Figure 4. Benzimidazol-2-yl-
quinaxolines 1a–d and i.
good to moderate binding to the
A1 AR but are unsuitable for bind-
ing to the A2A AR and, with a few
exceptions, to the A3 AR as well.
Several 2-(1H-benzimidazol-2-yl)quinolines 2 exhibit nanomolar
potency and good selectivity at the A1 AR: 2g (Ki =2.9 nm), 2 f
(Ki =9.1 nm), 2a (Ki =13 nm), 2i (Ki =14.5 nm) and 2e (Ki =
20.7 nm). Compound 2i stands out for its remarkable selectivi-
ty over the A2A and A3 ARs. With the exception of 2o, none of
the compounds in series 2 binds appreciably to the A3 AR.
In regards to binding the A1 AR, the structure–affinity rela-
tionships (SARs) of compounds 2 and their isosteres 1[24] are
rather different. Specifically, affinity is maintained by 5,6-di-
methyl substitution in series 2 (2i versus 2a) but decreased in
series 1 (1i versus 1a). Moreover, a 4-methylthio group is detri-
mental for potency in series 2 (2b versus 2a) but the same
moiety is favorable in series 1 (1c versus 1a). These data sug-
gest that pairs of isosteres belonging to series 1 and 2 do not
necessarily adopt similar orientations within the A1 AR binding
site. Elucidation of SARs within series 2 is not a straightforward
matter in terms of ligand–A1 AR interactions. The lack of appre-
ciable affinity for the 4-substituted derivatives 2b–d and the
4,6-disubstituted 2h and k would suggest that the 4- and 6-
positions of the benzimidazole ring orient toward sterically for-
bidden regions within the receptor binding site. However, this
hypothesis is not consistent with the excellent affinity of 2g
Scheme 5. Synthesis of diamines 20 and 21.
Biology
Affinities of the new compounds toward human A1, A2A, and A3
ARs were evaluated by competition experiments assessing
their respective abilities to displace [3H]8-cyclopentyl-1,3-dipro-
pylxanthine ([3H]DPCPX), [3H]adenosine-5’-N-ethylcarboxamide
1912
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1909 – 1918

Selective A1 Adenosine Receptor Antagonists
(Ki =2.9 nm), which has methyl and chlorine moieties at the 4-
and 6-positions, respectively. The higher potencies of 5-chloro
derivative 2 f (Ki =9.1 nm) and 6-chloro-4-methyl derivative 2g
(Ki =2.9 nm), as compared with their unsubstituted counterpart
2a (Ki =13 nm), indicate that these small and lipophilic groups
insert into hydrophobic pockets of the receptor cavity. The im-
proved affinity of the 5-chloro derivative 2 f (Ki =9.1 nm), as
compared with 5-methyl derivative 2e (Ki =20.7 nm), may be
ascribed to the different electron-withdrawing effects of these
two substituents. However, this hypothesis is somewhat con-
tradicted by the drastic drop in affinity from 4,5-dimethyl de-
rivative 2i (Ki =14.5 nm) to 4,5-dichloro derivative 2j (Ki >
10000 nm).
subset of compounds, the 5,6-dimethyl derivative 3i is notable
for its nanomolar potency at the A3 AR (Ki =19 nm), although
this is combined with limited selectivity over the A2A AR. Com-
pounds featuring a 4,5- or 5,6-fused cyclohexane ring (3l and
3m, respectively) as well as the aza isosteres of 3a (3n and o)
do not exhibit interesting affinity profiles for the ARs exam-
ined.
The limited binding data for compounds belonging to series
4 and 5 suggest that affinity across these series is retained or
improved by insertion of a small lipophilic substituent at the 4-
or 5-position of the benzimidazole ring, with the one excep-
tion of 5b (R1 =SC2H5). Series 4 and 5 exhibit the best results
in terms of affinity and selectivity at the A1 AR with respect to
the other series analyzed (series 2 and 3). Compounds 4a, b, e,
and f, as well as 5a, e, and f bind to the A1 AR with Ki values
in the nanomolar range and exhibit moderate to excellent se-
lectivity over the A2A and A3 ARs. The higher potency of 4a at
the A1 AR as compared with unsubstituted analogues 1a, 2a,
3a, and 5a suggests that 3-(1H-benzimidazol-2-yl)isoquinoline
represents the molecular scaffold among the five investigated
here that is most suited for selective binding to the A1 AR
(Figure 2). The highest-affinity compound within this subset is
4b, which has an ethylthio group at the 4-position of the ben-
zimidazole nucleus, analogous to the structure of 1b. In con-
trast, 4-ethylthio derivative 5b is devoid of any appreciable af-
finity to the ARs. Compound 4b is slightly less potent than 1b
at the A1 AR (Ki =1.4 nm versus Ki =0.5 nm, respectively), but
significantly more selective over the A2A AR (Ki >10 mm versus
Ki =3.4 mm, respectively) and the A3 AR (Ki >1 mm versus Ki =
0.95 mm, respectively). Compounds 1b and 4b were also
tested against the A2B AR by evaluating their inhibitory effect
on NECA-mediated cAMP accumulation in CHO cells stably ex-
pressing this receptor subtype. Compound 4b is completely
inactive at this subtype (Ki >10 mm), whereas 1b displays nano-
molar activity in this assay (IC50 =9.81ꢀ1.07 nm) while main-
taining good selectivity for the A1 AR. Incidentally, this product
may represent a promising lead compound for future develop-
ment of potent and selective A2B AR antagonists featuring the
(1H-benzimidazol-2-yl)quinoxaline scaffold.
A single substituent at the 4-position is unfavorable for affin-
ity (2b–d versus 2a). Nevertheless, when a chlorine is added
to the 6-position of 4-methyl derivative 2c, yielding 2g, affinity
increases dramatically. A chlorine at the 5-position also results
in good affinity (2 f versus 2a) but, surprisingly, if a second
chlorine is added to the 6-position of 5-chloro derivative 2 f to
yield 2j, affinity is lost. Thus, these data did not allow us to
make conclusions about the effect of a 6-chlorine substituent
on A1 AR affinity in series 2. It is likely that the puzzling SARs
outlined above depend on the different conformations and
tautomers accessible by our molecules thanks to the freely ro-
tating bond connecting the two heterocyclic systems and the
“mobile” benzimidazole NH hydrogen. Encouraged by the high
potency of 5,6-dimethyl derivative 2i, we designed com-
pounds 2l and m, which have a 5,6-fused benzene or cyclo-
hexane ring capable of exploiting hydrophobic interactions
with the A1 AR. Unfortunately, both compounds are devoid of
affinity for this receptor, likely due to unfavorable steric factors.
Attempts to improve affinity to any of the ARs by introduc-
ing a nitrogen atom as a hydrogen bond acceptor within the
benzimidazole nucleus were made through the synthesis of
aza isosteres 2n and o. While 2n is devoid of any appreciable
affinity toward the target receptors, 2o shows a moderate af-
finity to the A3 AR (Ki =94 nm) along with excellent selectivity
over the A1 and A2A ARs (Ki >10000 nm for both receptors).
The lack of affinity of 2p to the A1 AR clearly indicates that the
NH group of quinoline derivatives 2 is engaged in a hydrogen
bond within the binding site of this receptor and should be
therefore regarded as an element of the pharmacophore.
Although a few of the 3-(1H-benzimidazol-2-yl)quinolines
(3a–o) possess nanomolar potency at the A1 AR (3a, c, d, and
f), none of these are reasonably selective over the A2A and A3
ARs. While a 4-ethylthio group strongly improves A1 AR poten-
cy in series 1 (1b versus 1a), the same substitution in series 3
significantly lowers affinity to the A1 AR, without adding appre-
ciable affinity to the A2A and A3 ARs (3b versus 3a). In contrast
to observations for series 2, a methyl or chlorine in series 3 at
the 4-position of the benzimidazole ring enhances affinity to
the A1 AR (3c or d versus 3a). However, in series 3, similar to
what is observed for series 2, insertion of a chlorine at the 5-
position is slightly favorable for potency at the A1 AR, while in-
corporation of a methyl group at this position is detrimental
(3e or 3 f versus 3a). The 4,6- and the 5,6-disubstituted deriva-
tives 3g–k are fully devoid of affinity to the A1 AR. Among this
The high affinity and selectivity of 4b toward the A1 AR led
to its selection, along with lead compound 1b, for evaluation
of its effects on electrically evoked cholinergic contractions in
circular smooth muscle preparations from human colon. Prior
to this functional evaluation, we examined the expression of
A1 AR in the target tissues. RT-PCR analysis confirmed the pres-
ence of mRNA coding for the A1 AR in the neuromuscular com-
partment of colonic specimens. PCR amplification of cDNA
coding for A1 AR was performed under conditions similar to
those described by Christofi et al.[27] Following the first run of
amplification, cDNA bands were rarely able to be visualized for
the A1 AR. Therefore, PCR products were subjected to a second
run of amplification under the same conditions, and DNA
bands of the expected size for the A1 AR were always able to
be detected following electrophoretic separation (Figure 5).
The identity of the DNA bands was confirmed by sequencing
analysis.
ChemMedChem 2011, 6, 1909 – 1918
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1913

B. Cosimelli, S. Taliani, et al.
MED
ments, where 1b and 4b were tested against CCPA, a selective
A1 AR agonist. In colonic preparations maintained in Krebs so-
lution, containing dipyridamole and adenosine deaminase to
minimize interferences by endogenous adenosine, electrically
evoked cholinergic contractions were decreased by CCPA in a
concentration-dependent
fashion
(EC50 =18.3ꢀ3.2 nm;
E
max:ꢁ34.7ꢀ4.3%) (Figure 7). Under these conditions, the in-
hibitory effect exerted by CCPA was antagonized by 1b
(10 nm) (Figure 7A), 4b (10 nm) (Figure 7B) or DPCPX (10 nm)
(Figure 7C) (Kd =2.54ꢀ0.57 nm, 1.08ꢀ0.33 nm, and 2.94ꢀ
0.41 nm, respectively).
Figure 5. Reverse transcription–polymerase chain reaction (RT-PCR) of A1 AR
and b-actin mRNA in the neuromuscular layer of human distal colon. Repre-
sentative agarose gel showing the amplification of A1 AR and b-actin cDNAs.
Conclusions
Functional experiments were then performed to evaluate
the effect of 1b or 4b on the contractile activity of colonic cir-
cular smooth muscle. During the equilibration period, most co-
lonic preparations displayed a rapid spontaneous activity that
was low in amplitude and generally stable throughout the ex-
periment. Electrically evoked responses consisted of phasic
contractions of variable amplitude, which were abolished by
atropine or tetrodotoxin (not shown). Incubation of colonic cir-
cular muscle strips with A1 AR antagonists did not significantly
affect spontaneous motor activi-
We prepared a number of quinolines and isoquinolines con-
nected in various ways to a substituted 1H-benzimidazol-2-yl
system as potential novel AR antagonists. As a result, we iden-
tified 4b as the ligand exhibiting the best combination of po-
tency at the A1 AR (Ki =1.4 nm) and selectivity over A2A, A2B,
and A3 ARs (Ki >10 mm for A2A and A2B, Ki >1 mm for A3). Func-
tional experiments carried out in human circular smooth
muscle preparations showed that in vitro application of 4b or
ty.
The incubation of circular
muscle strips with A1 AR antago-
nists FSCPX (10 nm) or DPCPX
(10 nm) induced a significant in-
crease in electrically evoked con-
tractions (+31.4ꢀ5.3% and
+27.6ꢀ4.9%,
respectively)
(Figure 6). Treatment of colonic
specimens with 1b or 4b (0.01–
10 nm) led to a concentration-
dependent increase in electrical-
ly evoked cholinergic contrac-
tions, with a maximal increment
of +32.5ꢀ4.3% and +31.6ꢀ
6.4%, respectively, obtained at
10 nm (Figure 6). In addition, the
effects exerted by 1b (10 nm),
4b (10 nm), or DPCPX (10 nm)
were not altered by co-incuba-
tion with FSCPX (1000 nm), indi-
cating that 1b, 4b, or DPCPX
were sufficient to ensure specific
and maximal blockade of A1 ARs
(Figure 6). These findings sug-
gest that the enhancing effect of
1b or 4b on electrically induced
motility of human colon is
highly specific and depends on
selective blockade of the A1 AR.
To better substantiate this hy-
pothesis, a selected concentra-
tion of 10 nm was employed in a
Figure 6. In vitro effects of FSCPX (1000 nm), DPCPX (10 nm), 1b (0.01–10 nm), 4b (0.01–10 nm), FSCPX (1000 nm)
+ 1b (10 nm), FSCPX (1000 nm) + 4b (10 nm), or FSCPX (1000 nm) + DPCPX (10 nm) on the contractile responses
elicited by electrical stimulations (ES; 10 Hz, 0.5 ms, 30 mA applied as single 10 s train) of circular smooth muscle
prepared from human colon maintained in Krebs solution containing guanethidine, NK receptor antagonists, and
NPA. Each column represents the mean ꢀSEM obtained from eight experiments; *p<0.05: significant difference
subsequent series of experi- versus control.
1914
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1909 – 1918

Selective A1 Adenosine Receptor Antagonists
imal enhancement activity in the same in vitro model at con-
centrations of 1000 and 10 nm, respectively.[14,28] Notably, nei-
ther 1b nor 4b affected the motor activity of colonic prepara-
tions under resting conditions, and they did not lead to a fur-
ther increase in the stimulant effect of FSCPX on evoked con-
tractions, suggesting that the enhancing effect of 1b or 4b on
the electrically induced motility of human colon is highly spe-
cific and depends on the selective blockade of A1 ARs. More-
over, 1b and 4b, at the selected concentration of 10 nm, were
tested in a subsequent series of experiments against CCPA, a
selective A1 AR agonist. In this setting, CCPA decreased the
electrically induced contractions of colonic smooth muscle in a
concentration-dependent fashion, and 1b or 4b antagonized
this inhibitory effect with either a similar or an approximate
threefold higher potency than DPCPX, respectively.
In addition, it should be noted that lead compound 1b (low
A1/A2B AR-selective) and 4b (fully A1 AR-selective) presented a
similar functional profile, further substantiating the A1-mediat-
ed effects on colonic motility. Overall, the results of these ex-
periments support the evidence that 1b or 4b can behave as
potent antagonists of A1 ARs located in the neuromuscular
compartment of human colon, and suggest that these com-
pounds may provide a potential starting point for the develop-
ment of novel agents endowed with stimulant properties on
intestinal motility.
Experimental Section
Chemistry
General: Evaporation was performed in vacuo using a rotary evap-
orator. Analytical TLC was carried out using Merck 0.2 mm precoat-
ed silica gel aluminum sheets (60 F254). Silica gel 60 (230–400 mesh)
was used for flash column chromatography. Melting points were
determined using a Bꢁchi apparatus B 540 and are uncorrected.
Routine nuclear magnetic resonance spectra were recorded on a
Varian Mercury 400 MHz spectrometer in [D6]DMSO solution. 1,2-
Diamines 9 and 22 were obtained according to a previously report-
ed procedure[24] using 3-chloro-2-nitroaniline as a starting material.
Combustion analyses of target compounds were performed by our
analytical laboratory in Pisa. All compounds exhibited ꢂ98%
purity.
General procedure 1 for the synthesis of 2a, c, e–g, i, k, 2n, 2o–
p, 3a–g, i, j, n, o, and 4b: A mixture of quinoline-2-carboxylic acid
6
(0.50 g, 2.9 mmol), quinoline-3-carboxylic acid
7
(0.50 g,
2.9 mmol), or isoquinoline-3-carboxylic acid 26 (0.50 g, 2.9 mmol),
combined with the appropriate diamine (8, 10–14, 16, 18, 22–25,
2.9 mmol) and polyphosphoric acid (17.5 g) was stirred for 1 h at
1258C, then for 2 h at 2008C. 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 afford the desired derivative.
Yields, melting points, and spectral data are reported in the Sup-
porting Information.
Figure 7. Effects of increasing concentrations of CCPA (0.1–100 mm), alone or
in combination with A) 1b (10 nm), B) 4b (10 nm), or C) DPCPX (10 nm) on
contractions evoked by electrical stimulation (ES; 10 Hz, 0.5 ms, 30 mA ap-
plied as repeated 10 s trains every 5 min) of circular smooth muscle pre-
pared from human colon and maintained in Krebs solution containing dipyr-
idamole, adenosine deaminase, guanethidine, NK receptor antagonists, and
NPA. Each point represents the mean ꢀSEM of eight experiments; *p<0.05:
significant difference versus CCPA alone.
General procedure 2 for the synthesis of 2b, d, h, j, l, m, and
3h, k–m: A mixture of quinoline-2-carboxylic acid 6 (0.50 g,
2.9 mmol) or quinoline-3-carboxylic acid 7 (0.50 g, 2.9 mmol), com-
bined with the appropriate diamine (9, 11, 15, 17, 19–21,
2.9 mmol) and polyphosphoric acid (17.5 g) was heated for 4 h at
lead compound 1b, resulted in a concentration-dependent en-
hancement of the evoked contractions. The maximal effect of
these compounds occurred at 10 nm and was equivalent to
that recorded for FSCPX and DPCPX, both known to exert max-
ChemMedChem 2011, 6, 1909 – 1918
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1915

B. Cosimelli, S. Taliani, et al.
MED
2508C. After cooling to room temperature, the mixture was
poured into a large volume of ice water, and the solid mass was
collected by filtration. The crude product was treated with boiling
MeOH for 2 h, and charcoal was added. After filtration, the MeOH
solution was cooled to precipitate the desired products. Yields,
melting points, and spectral data are reported in the Supporting
Information.
NH), 7.30–7.21 (m, 2H, system AB), 2.79–2.73 (m, 2H, CH2), 2.60–
2.54 (m, 2H, CH2), 1.97 (s, 3H, CH3), 1.75–1.68 ppm (m, 4H, 2ꢂCH2).
2-Amino-3-nitro-5,6,7,8-tetrahydronaphthalene 31 and 2-amino-
1-nitro-5,6,7,8-tetrahydronaphthalene 32: A 1:1 (v/v) solution of
H2SO4/H2O (6 mL) was added slowly at room temperature to a
stirred suspension of acetamide 29 or 30 in 10 mL of EtOH. Follow-
ing this addition, the reaction mixture was heated at reflux for 1 h
(TLC, petroleum ether/EtOAc, 3:1 v/v), then cooled to room tem-
perature. The crude product was poured into ice water (120 mL),
and the pH was adjusted to 8.0 with 30% NH4OH. The resulting
yellow–orange precipitate was removed by filtration and treated
with 10% HCl (20 mL) until boiling. Compounds 31 (90%) and 32
(65%) were obtained as pure solids by filtration, and were purified
by recrystallization from EtOH.
General procedure 3 for the synthesis of 4a, b, e, f, and 5a, b,
e, f:
A mixture of isoquinoline-3-carboxylic acid 26 (0.50 g,
2.9 mmol) or isoquinoline-1-carboxylic acid 27 (0.50 g, 2.9 mmol),
combined with the appropriate diamine (8, 12, 13, 22, 2.9 mmol)
and polyphosphoric acid (17.5 g) was stirred for 4 h at 2508C. After
cooling to room temperature, the mixture was poured into a large
volume of ice water, and the pH was adjusted to 8.0 with 30%
NH4OH. The solid mass was collected by filtration, washed with
water, and dried. Treatment with boiling toluene furnished an ex-
tract which, after evaporation, yielded target compounds with the
desired degree of purity. Yields, melting points, and spectral data
are reported in the Supporting Information.
2-Amino-3-nitro-5,6,7,8-tetrahydronaphthalene 31: mp: 124.8–
1
125.68C (lit. [29] 124.5–1268C); H NMR ([D6]DMSO): d=7.67 (s, 1H,
H-4), 7.11 (brs, 2H, NH2), 6.69 (s, 1H, H-1), 2.66–2.60 (m, 4H, 2ꢂ
CH2), 1.69–1.62 ppm (m, 4H, 2ꢂCH2).
2-Amino-1-nitro-5,6,7,8-tetrahydronaphthalene 32: mp: 96.3–
97.48C (lit. [29] 94–968C); 1H NMR ([D6]DMSO): d=6.98 (m, 1H, part
A of the system AB), 6.72 (m, 1H, part B of the system AB), 5.87
(brs, 2H, NH2), 2.61–2.54 (m, 4H, 2ꢂCH2), 1.70–1.60 ppm (m, 4H,
2ꢂCH2).
5,6,7,8-Tetrahydronaphthalene-2,3-diamine 20 and 5,6,7,8-tetra-
hydronaphthalene-1,2-diamine (21): A solution of 2-amino-3-
nitro-5,6,7,8-tetrahydronaphthalene 31 (0.95 g, 4.0 mmol) or 2-
amino-1-nitro-5,6,7,8-tetrahydronaphthalene 32 (0.95 g, 4.0 mmol)
and tin(II) chloride dihydrate (4.55 g, 20.0 mmol) in EtOAc (65 mL)
was stirred at reflux for 2 h. After cooling, the reaction mixture was
poured into ice water (200 mL). While stirring, the pH was adjusted
to 8.0 with saturated sodium bicarbonate solution, and the result-
ing precipitate was removed by filtration. The aqueous phase was
separated and extracted with EtOAc (4ꢂ50 mL). The combined or-
ganic phases were dried with Na2SO4 and evaporated to dryness to
give the desired compounds 20 (88%) or 21 (91%), respectively,
which were recrystallized from EtOH.
Biology
Adenosine receptor binding assay
Materials: [3H]DPCPX, [3H]NECA, and [125I]AB-MECA were obtained
from DuPont–NEN (Boston, MA, USA). ADA was obtained from
Sigma–Aldrich (St. Louis, MO, USA). All other reagents were from
standard commercial sources and of the highest commercially
available grade. CHO cells stably expressing human A1, A2A, and A3
ARs were kindly supplied by Prof. K.-N. Klotz, Wurzburg University
(Germany).[32]
5,6,7,8-Tetrahydronaphthalene-2,3-diamine (20): mp: 135.3–
136.28C (lit. [29] 135–1368C); 1H NMR ([D6]DMSO): d=6.18 (s, 2H,
H-1 and H-4), 4.15 (brs, 4H, 2ꢂNH2), 2.48–2.42 (m, 4H, 2ꢂCH2),
1.65–1.59 ppm (m, 4H, 2ꢂCH2).
Human A1 adenosine receptors: Aliquots of membranes (50 mg pro-
tein) obtained from A1 CHO cells were incubated at 258C for
180 min in 500 mL T1 buffer (50 mm Tris-HCl, 2 mm MgCl2, 2 UmLꢁ1
ADA, pH 7.4) containing [3H]DPCPX (3 nm) and six different concen-
trations of the newly synthesized compounds. Nonspecific binding
was determined in the presence of 50 mm RPIA.[26] The dissociation
constant (Kd) for [3H]DPCPX in A1 CHO cell membranes was 3 nm.
5,6,7,8-Tetrahydronaphthalene-1,2-diamine (21): mp: 85.8–
86.78C (lit. [29] 84–858C); 1H NMR ([D6]DMSO): d=6.34 (m, 1H, part
A of the system AB), 6.14 (m, 1H, part B of the system AB), 4.16
(brs, 2H, NH2), 4.05 (brs, 2H, NH2), 2.53 (pt, 2H, CH2), 2.36 (pt, 2H,
CH2), 1.75–1.70 (m, 2H, CH2), 1.69–1.56 ppm (m, 2H, CH2).
N-(3-Nitro-5,6,7,8-tetrahydronaphthalen-2-yl)acetamide (29) and
N-(1-nitro-5,6,7,8-tetrahydronaphthalen-2-yl)acetamide (30):
A
Human A2A adenosine receptors: Aliquots of cell membranes (80 mg
protein) were incubated at 258C for 90 min in 500 mL T2 buffer
(50 mm Tris-HCl, 2 mm MgCl2, 2 UmLꢁ1 ADA, pH 7.4) in the pres-
ence of 30 nm of [3H]NECA and six different concentrations of the
newly synthesized compounds. Nonspecific binding was deter-
mined in the presence of 100 mm NECA.[26] The dissociation con-
stant (Kd) for [3H]NECA in A2A CHO cell membranes was 30 nm.
1:1 (v/v) solution of HNO3 (65%) and acetic anhydride (2.4 mL) was
added dropwise to a stirred suspension of 2-amino-5,6,7,8,-tetrahy-
dronaphthalene 28 (13.6 mmol) in acetic anhydride (2.6 mL), main-
taining a temperature below 408C. Following this addition, the
mixture was stirred for 1 h and then treated with ice water (80 mL)
to give a yellow precipitate, which was removed by filtration under
vacuum, washed with petroleum ether, and dried over NaOH. The
crude residue was purified by flash chromatography (petroleum
ether/EtOAc, 3:1!2:1 v/v) to yield compounds 29 and 30 in 36%
and 26% yields, respectively. Both 29 and 30 were purified by re-
crystallization from EtOH.
Human A3 adenosine receptors: Aliquots of cell membranes (40 mg
protein) were incubated at 258C for 90 min in 100 mL T3 buffer
(50 mm Tris-HCl, 10 mm MgCl2, 1 mm EDTA, 2 UmLꢁ1 ADA, pH 7.4)
in the presence of 1.4 nm [125I]ABMECA and six different concentra-
tions of the newly synthesized compounds. Nonspecific binding
was determined in the presence of 50 mm R-PIA.[26] The dissociation
constant (Kd) for [125I]AB-MECA in A3 CHO cell membranes was
1.4 nm.
N-(3-Nitro-5,6,7,8-tetrahydronaphthalen-2-yl)acetamide 29: mp:
134.1–135.08C (lit. [30] 133–1358C); 1H NMR ([D6]DMSO): d=10.06
(s, 1H, NH), 7.67 (s, 1H, H-4), 7.33 (s, 1H, H-3), 2.80–2.74 (m, 4H, 2ꢂ
CH2), 2.03 (s, 3H, CH3), 1.76–1.70 ppm (m, 4H, 2ꢂCH2).
All compounds were dissolved in DMSO and diluted with assay
buffer to the final concentration, the amount of DMSO never ex-
ceeding 2%. Percentage inhibition values for specific radiolabeled
N-(1-Nitro-5,6,7,8-tetrahydronaphthalen-2-yl)acetamide 30: mp:
1
129.2–129.98C (lit. [31] 1278C); H NMR ([D6]DMSO): d=9.79 (s, 1H,
1916
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1909 – 1918

Selective A1 Adenosine Receptor Antagonists
ligand binding at 1–10 mm concentration represent the mean ꢀ
SEM of at least three experiments. For compound IC50 determina-
tion, at least six different ligand concentrations were used. IC50
values, which were computer-generated using a nonlinear regres-
sion formula (GraphPad Prism, version 3.0 from GraphPad Software
Inc., San Diego, CA, USA), were converted into Ki values based on
the Kd values of radioligands in the various tissues and using the
Cheng and Prusoff equation.[33] Ki values represent the mean ꢀ
SEM of at least three determinations.
denaturation at 948C (1 min), annealing at 538C (30 s), extension
at 728C (1 min), and final extension at 728C (7 min). If no band
was detected, PCR was repeated using 2 mL of the initial amplifica-
tion product with fresh PCR reagents, as previously reported by
Christofi et al.[27] Untranscribed RNA was included in PCR reactions
to verify the absence of genomic DNA. RT-PCR efficiency was eval-
uated by primers for human b-actin. Amplified products were sep-
arated by 1.5% agarose gel electrophoresis and stained with ethid-
ium bromide. cDNA bands were visualized by UV light.
Measurement of cyclic AMP levels on hA2B CHO cells: Intracellular
cyclic AMP (cAMP) levels were measured using a competitive pro-
tein binding method.[34] CHO cells expressing recombinant hA2B
ARs were harvested by trypsinization. Following centrifugation and
resuspension in medium, cells (~48000) were plated in 24-well
plates in 0.5 mL medium. After 48 h, the medium was removed,
and the cells were incubated at 378C for 15 min with 0.5 mL of
DMEM in the presence of adenosine deaminase (1 UmLꢁ1) and the
phosphodiesterase inhibitor Ro-20-1724 (20 mm). Antagonism pro-
files for the new compounds toward A2B AR were evaluated by as-
sessing their ability to inhibit 100 nm NECA-mediated accumulation
of cAMP. Cells were incubated in the reaction medium for 15 min
at 378C with various concentrations of the target compound
(1 nm–10 mm) and then were treated with NECA. The reaction was
terminated by removal of the medium and the addition of 0.4 n
HCl. After 30 min, lysates were neutralized with 4 n KOH, and the
suspension was centrifuged at 800 g for 5 min. For determination
of cAMP production, bovine adrenal cAMP binding protein was in-
cubated with [3H]cAMP (2 nm) and 50 mL cell lysate or cAMP stan-
dard (0–16 pmol) at 08C for 150 min in a total volume of 300 mL.
Protein bound to radioactive compounds was separated by rapid
filtration through GF/C glass fiber filters and washed twice with
4 mL 50 mm Tris-HCl, pH 7.4. The radioactivity was measured by
liquid scintillation spectrometry.
Recording of circular smooth muscle contractile activity: The contrac-
tile activity of human colonic circular smooth muscle was recorded
as previously described by Fornai et al.[14] Preparations were set up
in organ baths containing Krebs solution (113 mm NaCl, 4.7 mm
KCl, 2.5 mm CaCl2, 1.2 mm KH2PO4, 1.2 mm MgSO4, 25 mm NaHCO3,
and 11.5 mm glucose, pH 7.4ꢀ0.1) with guanethidine (adrenergic
blocker, 10 mm), Nw-propyl-l-arginine [NPA, selective inhibitor of
neuronal nitric oxide (NO) synthase, nNOS; 100 nm], L-732138 (NK1
receptor antagonist, 10 mm), GR-159897 (NK2 receptor antagonist,
1 mm) and SB-218795 (NK3 receptor antagonist, 1 mm) to prevent
noncholinergic responses. The solution was maintained at 378C
and bubbled with 95% O2 and 5% CO2. Circular colonic strips were
connected to isotonic transducers (Basile, Comerio, Italy) under a
constant 1 g load and were allowed to equilibrate for at least
30 min. Colonic muscle activity was recorded by polygraphs
(Basile). Field electrical stimulation was delivered by a BM-ST6 stim-
ulator (Biomedica Mangoni, Pisa, Italy). Electrical stimuli (10 Hz,
0.5 ms, 30 mA) were applied as single or repeated 10 s trains. In
the latter setting, trains were applied every 5 min. Each preparation
was challenged with electrical stimulations, and experiments start-
ed when reproducible responses were obtained (usually after 2–3
stimulations).
Design of experiments: In the first set of experiments, the effects of
1b or 4b (0.01–10 nm) were evaluated with electrically evoked
cholinergic contractions. To verify that 1b- or 4b-induced effects
resulted specifically from A1 AR blockade, the selective and irrever-
sible A1 AR antagonist FSCPX (1000 nm),[28] as well as the selective
and reversible antagonist DPCPX (10 nm),[14] were used for compar-
ison. In the second set of experiments, effects of the A1 AR agonist
CCPA (0.1–10000 nm) were tested on electrically evoked cholinergic
contractions in the absence and presence of (10 nm) 1b, 4b, or
DPCPX. Krebs solution was added, along with dipyridamole (ade-
nosine reuptake inhibitor, 500 nm) and adenosine deaminase
(enzyme responsible for conversion of adenosine into inosine,
0.5 UmLꢁ1) to abate the extracellular levels of endogenous adeno-
sine.[20]
Pharmacology
Tissue excision and preparation: Specimens of human colon were
obtained from patients undergoing surgery for uncomplicated
neoplastic conditions. Samples consisted of sections of distal colon
from a macroscopically normal region taken at a distance of at
least 10 cm from any visible lesion. Care was taken to verify the ab-
sence of alterations by histological examination. Portions of tissue
were immediately flash frozen in liquid nitrogen and stored at
ꢁ808C for subsequent RT-PCR, or fixed in cold 4% paraformalde-
hyde, diluted in phosphate buffered saline, for routine histology.
The remaining tissue portions were placed into pre-oxygenated
Krebs solution and transported on ice to the laboratory, where cir-
cular muscle strips ~3 mm in width and 20 mm in length were pre-
pared.[14]
The effects of compounds tested were expressed as the percent-
age of changes versus control contractions elicited by electrical
stimulation. The apparent potency of CCPA was expressed as EC50
(concentration of agonist producing 50% of maximal response).
The percentage of maximum inhibition of control motor responses
(Emax) was also estimated. Both parameters were calculated from
concentration–response curves, and the values were averaged. The
affinities of 1b, 4b, and DPCPX were expressed as Kd values from
Equation (1), where B is the molar concentration of the antagonist
and DR is the ratio of equally effective concentrations of the ago-
nist (EC50) in the presence and absence of the antagonist.
RT-PCR: Colonic tissues were subjected to mucosa and submucosa
removal by sharp dissection. Total RNA was isolated from neuro-
muscular layers by Trizol (Life Technologies, Carlsbad, CA, USA) and
chloroform. RNA (1 mg) served as a template for single strand
cDNA synthesis by RT in a reaction using 2 mL random hexamers
(0.5 mgmLꢁ1) with 200 U Moloney murine leukemia virus (MMLV) re-
verse transcriptase in manufacturer’s buffer containing 500 mm de-
oxynucleotide triphosphate mixture (dNTP) and 10 mm dithiothrei-
tol (DTT). PCR was performed using primers based on the nucleo-
tide sequence of the cloned human A1 AR gene.[27] PCR, consisting
of 2 mL RT reaction, Taq polymerase 2.5 U, dNTP 100 mm and pri-
mers 0.5 mm in a final volume of 50 mL was performed using a T-
gradient thermocycler (Biometra, Gçttingen, Germany). After 1 min
initial denaturation at 948C, PCR was performed by 35 cycles of:
Kd ¼ ½Bꢃ=ðDRꢁ1Þ
ð1Þ
Drugs and reagents: Atropine sulfate, guanethidine monosulfate, 8-
cyclopentyl-N3-[3-(4-(fluorosulfonyl)benzoyloxy)propyl]-N1-propyl-
xanthine (FSCPX, Sigma–Aldrich); tetrodotoxin, l-732138, GR-
ChemMedChem 2011, 6, 1909 – 1918
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1917

B. Cosimelli, S. Taliani, et al.
MED
159897, SB-218795, Nw-propyl-l-arginine, 8-cyclopentyl-1,3-dipro-
pylxanthine (DPCPX), 2-chloro-N6-cyclopentyladenosine (Tocris, Bris-
tol, UK). Adenosine receptor ligands were dissolved in DMSO, with
further dilutions made using saline solution. DMSO concentration
in organ baths was never higher than 0.5%.
[14] M. Fornai, L. Antonioli, R. Colucci, N. Ghisu, P. Buccianti, A. Marioni, M.
Chiarugi, M. Tuccori, C. Blandizzi, M. Del Tacca, Neurogastroenterol. Motil.
2009, 21, 451–466.
[15] L. Antonioli, M. Fornai, R. Colucci, N. Ghisu, M. Tuccori, M. Del Tacca, C.
Blandizzi, Inflamm. Bowel Dis. 2008, 14, 566–574.
[16] B. Siegmund, F. Rieder, S. Albrich, K. Wolf, C. Bidlingmaier, G. S. Firestein,
D. Boyle, H. A. Lehr, F. Loher, G. Hartmann, S. Endres, A. Eigler, J. Phar-
macol. Exp. Ther. 2001, 296, 99–105.
[17] M. Odashima, G. Bamias, J. Rivera-Nieves, J. Linden, C. C. Nast, C. A.
Moskaluk, M. Marini, K. Sugawara, K. Kozaiwa, M. Otaka, S. Watanabe, F.
Cominelli, Gastroenterology 2005, 129, 26–33.
[18] L. Antonioli, M. Fornai, R. Colucci, N. Ghisu, F. Da Settimo, G. Natale, O.
Kastsiuchenka, E. Duranti, A. Virdis, C. Vassalle, C. La Motta, L. Mugnaini,
M. C. Breschi, C. Blandizzi, M. Del Tacca, J. Pharmacol. Exp. Ther. 2007,
322, 435–442.
[19] M. Kadowaki, Y. Nagakura, K. Tokita, K. Hanaoka, M. Tomoi, Eur. J. Phar-
macol. 2003, 458, 197–200.
Statistical analysis: Results are given as the mean ꢀSEM. Significant
differences were evaluated with raw data prior to percentage nor-
malization using Student’s t test for paired data or analysis of var-
iance (ANOVA), followed by post hoc analysis with Dunnett’s test
or Student–Newman–Keul’s test, as appropriate; p<0.05 was con-
sidered significant. Colonic preparations included in each test
group were obtained from different patients; therefore, the
number of experiments also refers to the number of patients as-
signed to each group. EC50 values were interpolated from concen-
tration–response curves. Calculations were performed by commer-
cial software (GraphPad Prism, GraphPad Software Inc.).
[20] L. Antonioli, M. Fornai, R. Colucci, N. Ghisu, C. Blandizzi, M. Del Tacca, In-
flamm. Bowel Dis. 2006, 12, 117–122.
[21] C. La Motta, S. Sartini, L. Mugnaini, S. Salerno, F. Simorini, S. Taliani,
A. M. Marini, F. Da Settimo, A. Lavecchia, E. Novellino, L. Antonioli, M.
Fornai, C. Blandizzi, M. Del Tacca, J. Med. Chem. 2009, 52, 1681–1692.
[22] L. Antonioli, M. Fornai, R. Colucci, N. Ghisu, M. Tuccori, M. Del Tacca, C.
Blandizzi, Pharmacol. Ther. 2008, 120, 233–253.
Acknowledgements
This work was supported financially by MIUR (PRIN-2007) and by
the Universities of Napoli and Pisa (Italy). We thank Prof. Dr. Karl-
Norbert Klotz for his generous gift of transfected CHO cells ex-
pressing human A1, A2A, A2B, and A3 adenosine receptors.
[23] H. Kehlet, K. Holte, Am. J. Surg. 2001, 182, S3–S10.
[24] E. Novellino, B. Cosimelli, M. Ehlardo, G. Greco, M. Iadanza, A. Lavecchia,
M. G. Rimoli, A. Sala, A. Da Settimo, G. Primofiore, F. Da Settimo, S. Talia-
ni, C. La Motta, K.-N. Klotz, D. Tuscano, M. L. Trincavelli, C. Martini, J.
Med. Chem. 2005, 48, 8253–8260.
[25] F. H. Allen, S. Bellard, M. D. Brice, B. A. Cartwright, A. Doubleday, H.
Higgs, T. Hummelimk, B. G. Hummelink-Peters, O. Kennard, W. D. S.
Motherwell, J. R. Rodgers, D. G. Watson, Acta Crystallogr. 1979, B35,
2231–2239.
Keywords: 1H-benzimidazol-2-ylquinolines
receptors · antagonists · colon motility · lead optimization
·
adenosine
[26] F. Da Settimo, G. Primofiore, S. Taliani, C. La Motta, E. Novellino, G.
Greco, A. Lavecchia, B. Cosimelli, M. Iadanza, K.-N. Klotz, D. Tuscano,
M. L. Trincavelli, C. Martini, Drug Dev. Res. 2004, 63, 1–7.
[27] F. L. Christofi, H. Zhang, J. G. Yu, J. Guzman, J. Xue, M. Kim, Y. Z. Wang,
H. J. Cooke, J. Comp. Neurol. 2001, 439, 46–64.
[28] J. E. Wunderlich, B. J. Needleman, Z. Chen, J. G. Yu, Y. Wang, I. Grants,
D. J. Mikami, W. S. Melvin, H. J. Cooke, F. L. Christofi, Am. J. Physiol. Gas-
trointest. Liver Physiol. 2008, 294, G554–G566.
[29] G. Schroeter, E. Kindermann, C. Dietrich, C. Beyschlag, Cl. Fleischhauer,
E. Riebensahm, C. Oesterlin, Justus Liebigs Ann. Chem. 1922, 426, 17–83.
[30] E. R. Ward, M. T. Coulson, J. Chem. Soc. 1954, 4545–4547.
[31] E. R. Ward, B. D. Pearson, J. Chem. Soc. 1959, 3378–3383.
[32] K.-N. Klotz, J. Hessling, J. Hegler, C. Owman, B. Kull, B. B. Fredholm, M. J.
Lohse, Naunyn Schmiedeberg’s Arch. Pharmacol. 1998, 357, 1–9.
[33] Y. Cheng, W. H. Prusoff, Biochem. Pharmacol. 1973, 22, 3099–3108.
[34] C. Nordstedt, B. B. Fredholm, Anal. Biochem. 1990, 189, 231–234.
[1] S.-A. Poulsen, R. J. Quinn, Bioorg. Med. Chem. 1998, 6, 619–641.
[2] B. B. Fredholm, G. Arslan, L. Halldner, B. Kull, G. Schulte, W. Wasserman,
Naunyn Schmiedeberg’s Arch. Pharmacol. 2000, 362, 364–374.
[3] B. B. Fredholm, A. P. IJzerman, K. A. Jacobson, K.-N. Klotz, J. Linden,
Pharmacol. Rev. 2001, 53, 527–552.
[4] B. B. Fredholm, Cell Death Differ. 2007, 14, 1315–1323.
[5] G. Schulte, B. B. Fredholm, Cell. Signalling 2003, 15, 813–827.
[6] K. A. Jacobson, Z. G. Gao, Nat. Rev. Drug Discov. 2006, 5, 247–264.
[7] S. Moro, Z.-G. Gao, K. A. Jacobson, G. Spalluto, Med. Res. Rev. 2006, 26,
131–159.
[8] A. K. Dhalla, J. C. Shryock, R. Shreeniwas, L. Belardinelli, Curr. Top. Med.
Chem. 2003, 3, 369–385.
[9] J. Wardas, Pol. J. Pharmacol. 2002, 54, 313–326.
[10] J. A. Ribeiro, A. M. Sebasti¼o, A. de MendonÅa, Prog. Neurobiol. 2003, 68,
377–392.
[11] R. N. Takahashi, F. A. Pamplona, R. D. Prediper, Front. Biosci. 2008, 13,
2614–2632.
[12] A. Nishiyama, M. Rahman, E. W. Inscho, Hypertens. Res. 2004, 27, 791–
804.
Received: June 8, 2011
Revised: June 29, 2011
[13] R. H. Shah, W. H. Frishman, Cardiol. Rev. 2009, 17, 125–131.
Published online on July 27, 2011
1918
www.chemmedchem.org
ꢀ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2011, 6, 1909 – 1918