Synthesis and biological effects of a new series of 2-amino-3-benzoylthiophenes as allosteric enhancers of A1-adenosine receptor

Article abstract of DOI:10.1016/S0960-894X(00)00379-6

New derivatives of PD 81,723, an allosteric enhancer of agonist binding to the A₁-adenosine receptor, have been synthesized and evaluated in an intact cell assay. Compounds 3a, 3o and 3p appeared to be more potent than PD 81,723 and at a concentration of 0.1 μM caused significant reductions of cAMP content of CHO cells expressing the human A₁-adenosine receptor. Compounds 4e and 4o appeared to be allosteric enhancers at a low concentration and antagonists at a higher concentration, whereas compounds 3c, 3g, 3s and 4l appeared to be weak antagonists that are also allosteric enhancers at the higher concentration of 10 μM. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0960-894X(00)00379-6

Bioorganic & Medicinal Chemistry Letters 10 (2000) 1953±1957
Synthesis and Biological E€ects of a New Series of
2-Amino-3-benzoylthiophenes as Allosteric Enhancers
of A1-Adenosine Receptor
Pier Giovanni Baraldi,a,* Abdel Naser Zaid,a Ilaria Lampronti,a Francesca Fruttarolo,a
Maria Giovanna Pavani,a Mojgan Aghazadhe Tabrizi,a John C. Shryock,b
Edward Leungc and Romeo Romagnolia
a
Á
Dipartimento di Scienze Farmaceutiche, Universita di Ferrara, via Fossato di Mortara, 17/19, I-44100 Ferrara, Italy
bDepartment of Medicine, University of Florida, Gainseville, Florida, USA
cMedco Research, Inc., Research Triangle Park, North Carolina, USA
Received 18 April 2000; revised 7 June 2000; accepted 26 June 2000
AbstractÐNew derivatives of PD 81,723, an allosteric enhancer of agonist binding to the A1-adenosine receptor, have been syn-
thesized and evaluated in an intact cell assay. Compounds 3a, 3o and 3p appeared to be more potent than PD 81,723 and at a
concentration of 0.1 mM caused signi®cant reductions of cAMP content of CHO cells expressing the human A1-adenosine receptor.
Compounds 4e and 4o appeared to be allosteric enhancers at a low concentration and antagonists at a higher concentration,
whereas compounds 3c, 3g, 3s and 4l appeared to be weak antagonists that are also allosteric enhancers at the higher concentration
of 10 mM. # 2000 Elsevier Science Ltd. All rights reserved.
Adenosine is a ubiquitous autocoid with multiple e€ects
on the human body. Adenosine can bind to four di€er-
ent P1-purinoreceptors (A1, A2A, A2B, A3) which belong
to the broad group of receptors coupled to G-proteins
and which modulate the activity of adenylate cyclase.1
A1-Adenosine receptors are found in many tissues
including the heart, kidney, brain and spinal cord, adi-
pose and thyroid tissues, lung, and immune system cells.
In 1990, Bruns and co-workers2,3 reported that 2-
amino-3-benzoylthiophene derivatives both enhanced
the binding of agonists to the A1-adenosine receptor
and (usually at higher concentrations) acted as compet-
itive antagonists at these receptors. Among the synthe-
sized compounds, PD 81,723 ((2-amino-4,5-dimethyl-3-
thienyl)-[3-(tri¯uoromethyl)phenyl]methanone) was iden-
ti®ed as the compound with the best ratio of enhancement
to antagonistic action at the A1-adenosine receptor.3 To
study the role of various substitutions on the phenyl
ring and the importance of the 4,5-dimethyl group on
the thienyl ring, Baraldi4 and Ijzerman5 have described the
synthesis and biological evaluation of mostly novel PD
81,723 analogues. The purpose of our investigation was
to synthesize and evaluate a new series of derivatives of
PD 81,723 and to establish the structural requirements
and the structure±activity relationship for enhancement
of the action of an agonist at the human A1-adenosine
receptor.
Chemical Methods
The formation of 2-amino-3-benzoylthiophene from the
base-catalyzed condensation of carbonyl compounds
and nitriles, known as the `Gewald reaction',6 was
reported 35 years ago. Nevertheless, it was not until
1990 that synthesis of 2-amino-3-benzoylthiophenes by
this method received renewed attention. The same
method has been used for the synthesis of our new 2-
amino-3-benzoylthiophene derivatives 3a±x and 4a±v
wherein the appropriate carbonyl compounds were
reacted with benzoylacetonitrile derivatives 2a±h and
sulfur in ethanol in the presence of morpholine (Scheme 1).
*Corresponding author. Tel.: +39-0532-291293; fax: +39-0532-291296;
e-mail: pgb@dns.unife.it
0960-894X/00/$ - see front matter # 2000 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00379-6

1954
P. G. Baraldi et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1953±1957
Scheme 1. (a) KCN, EtOH, H2O, rt, 2 h; (b) R2COCH2R3,S8, morpholine, EtOH, 70 ꢀC, 1 h, rt, 18 h; (c)
1 h, rt, 18 h; (d) HBr/CH3CO2H, rt 6 h.
, S8, morpholine, EtOH, 70 ꢀC,
The derivative 4g±k were obtained starting from 4p and
4r±u, respectively, removing the Cbz (Z) protective
group by the treatment with HBr in acetic acid. The
compounds 3c±e, 3m, 3o±q, 4d have been reported by
Bruns3 and Ijzerman.5 The benzoyl acetonitrile deriva-
tives, when not commercially available, were synthesized
by treatment of the corresponding phenacylbromides 1a±h
with potassium cyanide in high yields. The structures of
the synthesized compounds and the yields of the syntheses
are presented in Table 1. The range of yields was 35±67%.
bu€ered saline solution. The wash solution was then
removed and replaced with fresh Hanks' solution con-
taining forskolin (1 mM), rolipram (20 mM), CPA (0.05±
1 nM), adenosine deaminase (2 U/mL), and the allosteric
enhancer to be tested. After 6 min of incubation at 36 ꢀC
in the presence of drugs, the incubation solution was
removed and hydrochloric acid (®nal concentration,
50 mM) was added to cells to terminate drug action.
The content of cAMP in acidi®ed extracts of cells was
determined by radioimmunoassay as previously descri-
bed.7 Allosteric enhancement was measured as the
action of a test compound at di€erent concentrations
(0.1, 1 and 10 mM) to reduce the cAMP content of
CHO: huA1 cells in the presence of 0.05±0.1 nM CPA.
CPA (0.05±0.1 nM) alone causes a slight reduction of
cAMP content of cells by activation of A1-adenosine
receptors. Allosteric enhancement of the action of CPA
causes a further reduction of the cAMP content of
CHO: huA1 cells. Because the spontaneous activity of
adenosine receptors in CHO: huA1 cells causes an inhi-
bition of adenylyl cyclase activity even in the absence of
an agonist,8 antagonists of adenosine receptors increase
cAMP content of cells. Therefore, compounds that
increased cAMP content of cells in this study were
putatively identi®ed as A1-adenosine receptor antagonists.
CHO: huA1 cells gradually lost responsiveness to allo-
steric enhancers during continuous culture. Three di€er-
ent batches of frozen cells were used for experiments.
Biological Methods
Chinese hamster ovary cells expressing human recom-
binant A1-adenosine receptors (CHO: huA1 cells) at a
density of approximately 8000 fmol/mg protein were
prepared as previously described7 and aliquots of these
cells at low passage numbers were frozen and stored in
liquid nitrogen. Cells were removed from liquid nitro-
gen storage when needed, and grown in Ham's F-12
culture medium with 10% fetal bovine serum and
0.5 mg/ml of antibiotic G-418.8 Aliquots of cells were
placed into 12-well culture plates with culture medium,
serum, and antibiotic for 48±72 hours, by which time the
cells had grown to a con¯uent monolayer. To begin an
experiment, growth medium was removed from the cul-
ture plates and cells were washed once with Hanks'

P. G. Baraldi et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1953±1957
Table 1. Yields for the synthesized compounds 3a±x and 4a±v
Table 2. Allosteric enhancement by 3a±x and 4a±va
1955
Compound
R1
R2
R3
R4
Yielda(%)
Compound
Percentage of response to 10 mM of PD81,723
to di€erent concentrations of tested compounds
3a
3b
3c3,5
3d3,5
3e5
3f
3g
3h
3i
3j
3k
3l
3m5
3n
3o3,5
3p5
3q3
3r
H
4-Cl
H
H
H
H
H
35
45
56
41
51
40
35
45
45
41
36
30
54
46
47
50
45
25
30
10
25
10
13
51
62
48
51
53
48
58
51
52
38
62
52
59
38
63
66
52
46
44
42
56
65
0.1 mM
1 mM
10 mM
CH3 CH3
3a
3b
51
7
26
8
28
25
19
100
92
82
68
65
/
3-CF3 CH3 CH3
4-Cl
H
H
4-F
4-Cl
4-Br
4-I
4-OMe
H
4-F
4-Cl
4-Br
4-OMe
4-C6H5
H
CH3 CH3
-(CH2)3-
-CH2CH2S-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)3-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)4-
-(CH2)5-
-(CH2)5-
-(CH2)5-
-(CH2)5-
-(CH2)6-
b
c
3c3,5
/
*
3d (PD 81,723)3,5
3e5
3f
3g
3h
3i
3j
3k
3l
3m5
3n
3o3,5
3p5
3q3
3r
13
30
35
*
/
*
31
*
/
16
/
57
50
/
14
*
26
25
/
24
18
/
/
/
58
8
/
40
52
29
/
43
/
53
*
72
/
43
48
43
58
67
44
79
29
36
42
54
39
41
/
42
77
83
86
67
71
68
54
61
72
36
68
85
*
22
*
*
31
12
9
/
*
*
31
*
31
*
*
*
*
3s
3t
4-F
3u
3v
3x
4a
4b
4c
4d3,5
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
4p
4q
4r
4s
4t
4u
4v
4-Cl
4-OMe
4-Cl
3-CF3
3-CF3
3-CF3
4-Cl
4-Cl
4-Cl
H
4-Cl
4-Br
4-I
4-C6H5
H
3s
3t
3u
3v
3x
4a
4b
4c
4d3,5
4e
4f
4g
4h
4i
4j
4k
4l
4m
4m
4n
4o
4p
4q
4r
C6H5CH2
C6H5CH2CH2
C6H5CH2CH2CH2
C6H5CH2
C6H5CH2CH2
C6H5CH2CH2CH2
H
7
*
32
15
9
12
/
*
/
13
*
13
*
H
H
H
H
15
/
/
/
*
*
*
*
10
*
4-NO2C6H5CH2CH2
4-NO2C6H5CH2CH2
CH2CO2Et
CH2CO2Et
CO2CH2C6H5
CO2CH2C6H5
CO2CH2C6H5
CO2CH2C6H5
CO2CH2C6H5
CO2CH2C6H5
4-CH3C6H4SO2
4-Cl
H
4-Cl
H
4-F
4-Cl
4-Br
4-I
4-Ph
H
*
*
*
*
/
/
*
*
*
4s
*
aYield of synthesized compounds after puri®cation by column chro-
matography.
4t
4u
4v
/
14
*
/
6
*
/
/
*
aThe results are the average of quadruplicate determinations in each
assay.
b(/) Signi®es a change of less than 5% of allosteric enhancement.
c(*) Signi®es an increase of cAMP content of cells, and may indicate
antagonism of the A1 adenosine receptor.
Because the magnitude of the e€ects of allosteric
enhancers on cells changed with time and batch of cells,
the actions of each test compound to enhance the e€ect of
CPA (and thus reduce cAMP content of CHO cells) were
normalized to the response of cells to 10mM PD 81,723.
cells expressing the cloned human A1-adenosine receptor.
This assay assessed the overall functional e€ect and the
potential for each tested compound as an enhancer or
an antagonist of the A1-adenosine receptor. Although
PD 81723 inhibited basal adenylyl cyclase (AC) activ-
ity,9 preliminary results showed that none of synthesized
compounds was interesting as inhibitor of adenylate
cyclase in FRTL-5 cells.
Results and Discussion
The e€ects of the newly-synthesized derivatives of PD
81,723 on CHO cells expressing the cloned human A1-
adenosine receptor are shown in Table 2. Allosteric
enhancers of agonist binding to the A1-adenosine
receptor have been shown to increase agonist binding in
equilibrium binding studies,2 to slow dissociation of the
agonist-receptor complex in kinetic studies,2 and to
cause leftward shifts of the agonist concentration-
response relationship in studies of isolated tissues.7,8 In
the current study, we have chosen to measure the e€ect
of the PD 81,723 analogues on cAMP content of CHO
Many of the tested compounds appeared to be enhancers
(e.g. most compounds from 3a to 4a) whereas others
appeared to be antagonists of the A1-adenosine receptor
(e.g. most compounds from 4j to 4v). Although none of
the tested compounds had a greater ecacy than PD
81,723, compounds 3a, 3o and 3p appeared to be more

1956
P. G. Baraldi et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1953±1957
potent than PD 81,723. Because the known allosteric
enhancers are also A1-adenosine receptor antagonists at
some (usually high) concentration, the fact that none of the
tested compounds had a greater ecacy than did 10 mM
PD 81,723 could be explained by a potential antagonist
e€ect of the compounds at high concentrations.
derivatives 4a±u. In the series of compounds 4a±c, which
bear the 3-tri¯uoromethyl substituent of PD 81,723 and
di€erent phenylalkyl moieties on position six, the benzyl
moiety was the best substituent. Indeed, only the N-benzyl
derivative 4a possessed an activity comparable to PD
81,723 at the concentration of 10 mM, while compounds
4b and 4c were inactive. Replacement of a 3-tri¯uoro-
methyl (compounds 4a±c) with a 4-chloro (compounds
4d±f) substituent on the benzoyl group led to a complete
loss of activity for compound 4d3,5 compared to the
compound 4a. In contrast, at concentrations of 0.1 and
1 mM the derivatives 4e and 4f were more potent than 4b
and 4c, respectively. The presence of a polar 4-nitro group
on the phenylethyl moiety of compound 4e, to yield com-
pound 4l, led to an enhancing activity which was com-
parable to that reported for 4f at the concentration of
10 mM. For the compound 4d, the removal of the only
benzyl group (to form compound 4h) and both of the N-
benzyl and 4-chloro substituents (to form compound 4g)
had little e€ect on enhancing activity. When the 4-chloro
substituent of the benzoyl group of compound 4h was
substituted by either a bromo, iodo or phenyl sub-
stituent, to yield the compounds 4i±k, respectively, or the
N-positive charge was suppressed by addition of a tosyl
group (compound 4v), no improvement in enhancing
activity was observed. It has been reported by Bruns3
that a positive charge on the nitrogen reduces enhancing
activity. In fact, the N-ethoxycarbonyl derivatives of the
compounds 4g and 4h showed a very potent enhancing
activity. Bruns' observation was con®rmed by the synthesis
of the ethoxycarbonyl methyl derivatives of compounds 4g
and 4h, corresponding to compounds 4n and 4o,
respectively, wherein a methylene spacer has been
inserted between the nitrogen and the ethoxycarbonyl
moiety. These compounds contain a positively-charged
nitrogen and had no enhancing activity.
We have systematically modi®ed PD 81,723 in this
study. The absence of the tri¯uoromethyl moiety on the
phenyl ring (compound 3c3,5) led to a loss of activity at
any concentration, while in the same compound 3c the
removal of both methyl substituents in positions 4 and 5
on the thiophene ring, to yield compound 3a, led to an
increase of activity at a lower concentration (0.1 mM).
However, compound 3a was less active than PD 81,723
at 1 and 10 mM. No increase of activity was obtained
with the presence of a non-polar substituent (such as
chloro) in the para position on the benzoyl ring (compound
3b). In the series of 4,5-dimethyl thiophene derivatives that
we have synthesized, only the already reported compound
3e5 possessed an activity comparable with that of PD
81,723 at each concentration tested.
The e€ect of substitutions at the 4- and 5-positions of
the thiophene ring in the PD 81,723 and particularly on
its `cyclized' analogues, was carefully examined. The
bridging of the 4- and 5-positions in compound 3c with
a methylene, to give 3f, caused an increase of activity
compared to that reported for 3c3,5 at 0.1 and 10 mM,
and comparable with that described for PD 81,723.
When the methylene in position 6 of the dihydrocyclo-
pentadien[b]thiophene ring of compound 3f was sub-
stituted with an atom of sulfur, to give 3g, we observed a
loss of activity at 0.1 and 1 mM. However, at 10 mM com-
pound 3g was only slightly less e€ective than PD 81, 723.
We have synthesized a large series of `cyclic' PD 81,723
analogues, represented by the dihydrocyclopenta-
dien[b]thiophene derivatives 3f and 3h±l. As was just
noted, compound 3f was only slightly less active as an
enhancer than PD 81,723. The addition of a halogen
(compounds 3h±k) or a methoxy (compound 3l) sub-
stituent at position 4 of the phenyl ring (R1) of com-
pound 3f did not lead to an increase of activity. Activity
increased when the size of the fused ring (R2±R3) was
increased from ®ve (compounds 3f and 3h±l) to six
(compounds 3m±r) carbons and it decreased slightly
when the size of the ring was increased from six to seven
(compounds 3s±u) carbons. Further increasing the size
of the fused ring from seven to eight (compounds 3v±x)
carbons did not alter activity. However, the cyclic ana-
logues 3f±x were less active than PD 81,723 at a con-
centration of 10 mM, and only the previously reported
compounds 3o3,5 and 3p5 appeared to be more potent
than PD 81,723 at lower concentrations (0.1 and 1 mM).
Comparing the compounds having a substituent on the
benzoyl ring (R1), the tetrahydrobenzo[b]thiophene
derivatives 3m±r appeared to be more potent than the
dihydrocyclopentadien[b]thiophene counterparts 3f±l.
Modi®cation of the ethoxycarbonyl moiety (position
R4, Scheme 1) reduced enhancing activity. Compounds
4p±u, which possess the N-benzyloxycarbonyl protective
group (Z) in place of the ethoxycarbonyl moeity, were
completely inactive. Of the 45 compounds synthesized
for this study, several were synthesized and studied by
Bruns3 and Ijzerman5 (see Tables 1 and 2). Our results
are generally consistent with the results reported by
these investigators, even though di€erent assays and
di€erent sources of adenosine receptors were used in
these studies. However, it should be noted that our
assay of e€ects of putative enhancers on cAMP content
of CHO cells expressing human A1-adenosine receptors is
not a speci®c assay of allosteric enhancement of agonist
binding. Our assay does not directly measure the inter-
action between receptor activation and G protein acti-
vation, and our observations may be complicated by
drug actions not related to enhancement, such as cell
toxicity. However, the e€ect of a tested compound in
the intact cell cAMP assay used in this study may be a
more useful predictor of the e€ect of the compound in
vivo than is a binding assay that more speci®cally
assesses allosteric enhancement.
An atom of nitrogen was inserted into the 6 position of
the tetrahydrobenzo[b]thiophene ring (Scheme 1, route
c) to give the 4,5,6,7-tetrahydrothieno [2,3-c]pyridine
In conclusion, compounds 3a, 3o and 3p appeared to be
nearly as ecacious and more potent than PD 81,723,

P. G. Baraldi et al. / Bioorg. Med. Chem. Lett. 10 (2000) 1953±1957
1957
and caused signi®cant reductions of cAMP content of
CHO: huA1 cells at a concentration of 0.1 mM. Other
compounds mimicked the action of the antagonist CPX,
and therefore may be antagonists of the Al-adenosine
receptor. The dual actions of benzylthiophenes to
antagonize and enhance the actions of A1-adenosine
receptor agonists have been noted previously.2,3,7,8
2. Bruns, R. F.; Fergus, J. H. Mol. Pharmacol. 1990, 38, 939.
3. Bruns, R. F.; Fergus, J. H.; Coughenour, L. L.; Courtland,
G. G.; Pugsley, T. A.; Dodd, J. H.; Tinney, F. J. Mol. Phar-
macol. 1990, 38, 950.
4. Baraldi, P. G. US Patent 5,939,432. 1997.
5. Van der Klein, P. A. M.; Kourounakis, A. P.; Ijzerman,
AdP. J. Med. Chem. 1999, 42, 3629.
6. (a) Gewald, K. Chem. Ber. 1965, 98, 3571. (b) Gewald, K.;
Schinke, E.; Bottcher, H. Chem. Ber. 1966, 99, 94. (c) Peet, N.
P.; Sunder, S.; Barbuch, R. J.; Vinogrado€, A. P. J. Het.
Chem. 1986, 23, 129.
Acknowledgement
7. Kollias-Baker, C. A.; Ruble, J.; Jacobson, M.; Harrison, J.
K.; Ozeck, M. J.; Shryock, J. C.; Belardinelli, L. J. Pharmacol.
Exp. Ther. 1997, 281, 761.
8. Shryock, J. C.; Ozeck, M. J.; Belardinelli, L. Mol. Pharma-
col. 1998, 53, 886.
We wish to thank Medco Research for support of this
research.
References
9. Musser, B.; Mudumbi, R. V.; Liu, J.; Olson, R. D.; Vestal,
R. J. Pharmacol. Exp. Ther. 1999, 288, 446.
1. Ralevic, V.; Burnstock, G. Pharmacol. Rev. 1998, 50, 413.