Affinity and intrinsic efficacy (IE) of 5′-carbamoyl adenosine analogues for the A1 adenosine receptor - Efforts towards the discovery of a chronic ventricular rate control agent for the treatment of atrial fibrillation (AF)

Article abstract of DOI:10.1016/j.bmcl.2003.09.094

The SAR for the affinity to the A₁ adenosine receptor and relative intrinsic efficacy (IE, [³⁵S]-GTPγS binding) of a series of 5′-carbamate and 5′-thionocarbamate derivatives of tecadenoson is described. Based on this SAR, selected compounds were evaluated in guinea pig isolated hearts to determine whether they were partial or full agonists with respect to their negative dromotropism, an A₁ AdoR mediated effect. Progress towards obtaining a partial A₁ AdoR agonist to potentially control ventricular rate during atrial fibrillation has been made with the discovery of several potent partial A₁ AdoR agonists (compounds 13, 14, and 17).

Full text of DOI:10.1016/j.bmcl.2003.09.094

Bioorganic & Medicinal Chemistry Letters 14 (2004) 535–539
0
Affinity and intrinsic efficacy (IE) of 5 -carbamoyl adenosine
analogues for the A adenosine receptor—efforts towards the
1
discovery of a chronic ventricular rate control agent for the
treatment of atrial fibrillation (AF)
a
a
a
a
b
Venkata P. Palle, Vaibhav Varkhedkar, Prabha Ibrahim, Hiba Ahmed, Zhihe Li,
b
b
b
b
b
c
Zhenhai Gao, Mark Ozeck, Yuzhi Wu, Dewan Zeng, Lin Wu, Kwan Leung,
c
a,
Nancy Chu and Jeff A. Zablocki *
aDepartment of Bioorganic Chemistry, CV Therapeutics, 3172 Porter Drive, Palo Alto, CA 94304, USA
bDepartment of Pharmacological Sciences, CV Therapeutics, 3172 Porter Drive, Palo Alto, CA 94304, USA
cDepartment of Preclinical Development, CV Therapeutics, 3172 Porter Drive, Palo Alto, CA 94304, USA
Received 10 August 2003; accepted 29 September 2003
3
5
Abstract—The SAR for the affinity to the A adenosine receptor and relative intrinsic efficacy (IE, [ S]-GTPgS binding) of a series
1
0
0
of 5 -carbamate and 5 -thionocarbamate derivatives of tecadenoson is described. Based on this SAR, selected compounds were
evaluated in guinea pig isolated hearts to determine whether they were partial or full agonists with respect to their negative dromo-
tropism, an A AdoR mediated effect. Progress towards obtaining a partial A AdoR agonist to potentially control ventricular rate
1
1
during atrial fibrillation has been made with the discovery of several potent partial A
2003 Elsevier Ltd. All rights reserved.
1
AdoR agonists (compounds 13, 14, and 17).
#
The most common form of atrial arrhythmia is atrial
fibrillation (AF), with an estimated 2.2 million Amer-
icans having either paroxysmal (sudden onset) or per-
Tecadenoson (CVT-510) 1 is a full A AdoR agonist
1
with high affinity and selectivity for the A AdoR (Fig. 1
1
and Table 1), and it has been investigated in clinical
trials for the treatment of paroxysmal supraventricular
1
sistent AF. One strategy for the management of AF
5
,6
is ventricular rate control wherein the AF is allowed
to persist while the ventricular rate is controlled.
Most agents that provide ventricular rate control act
by prolonging the effective refractory period (ERP) of
the atrioventricular (A-V) node, thereby ‘filtering’
tachycardia (PSVT). We recently reported the partial
agonism at the A AdoR of the A-V node with respect
1
0
carbamate derivative of tecadenoson, compound 2. We
to the negative dromotropic effect for the 5 -N-methyl-
4
have studied the SAR of affinity and relative intrinsic
0
0
atrial electrical activity and reducing the number of
2
efficacy (IE) of a series of 5 -carbamates and 5 -thiono-
carbamates of tecadenoson, with a goal to find an orally
active partial A AdoR agonist that can be used to
impulses conducted to the ventricles. The A adeno-
1
sine receptor (A AdoR) agonists prolong the ERP in
1
1
the A-V node by increasing an inwardly-rectifying
potassium current and reducing the excitability of A-V
provide ventricular rate control during AF.
3
0
to modify to potentially provide partial agonism, since
nodal cells. The primary rationale for partial agonism
at the A AdoR of the A-V node is to avoid third degree
The 5 -position of tecadenoson 1 was chosen as the site
1
4
0
adenosine (5 -deoxyCPA) 6 retains affinity for the A
A-V block.
others have found that the 5 -deoxy-N-6-cyclopentyl-
0
AdoR (K =70 nM, rat membrane) and selectivity over
1
i
7
the A2A AdoR (Fig. 1). In addition, IJzerman and
co-workers found that removing the hydrogen bond
*
0
Corresponding author. Fax: +1-650-858-0390; e-mail: jeff.
zablocki@cvt.com
0
donor capabilities of the 5 -hydroxyl group of CPA 5 by
960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.09.094

536
V. P. Palle et al. / Bioorg. Med. Chem. Lett. 14(200 4) 535–539
conversion to a methoxyl group resulted in a partial
agonist, compound 7, with high affinity for the A
had a high IE that suggests the hydrogen bond donor
0
contributing to IE.
capabilities of the 5 -methylamino substituent maybe
8
1
8
AdoR (K =33.7ꢀ6.9 nM, rat membrane, Fig. 1). They
i
0
also found that the 5 -methylamino derivative 8 (Fig. 1)
By definition, a partial agonist produces a sub-maximal
response even when present in a sufficiently high con-
centration to activate all of its receptors in a given tis-
4
,9
sue. The concentration–response relationships for an
agonist are dependent on the affinity of the agonist for
the receptor, the IE of the agonist to activate the recep-
tor, the receptor density in the tissue in question, and
the efficiency of translation of receptor stimulus to
1
0
response. To obtain a potent partial agonist at the A1
AdoR for a given biological response, it may require the
separation of ligand–receptor interactions that provide
affinity from those that are responsible for IE. We
hypothesized that it is possible to increase the affinity of
1
Figure 1. A AdoR agonists: tecadenoson 1, a full agonist; 2, weak
4
partial agonist; 3-bis pro-drug; 14, potent partial; 4, full agonist; 5-
CPA; 6–8 previously described agonists.7
,8
0
Table 1. Affinity for the A
0
1
AdoR, activity in GTPgS binding assay and negative dromotropic effect (i.e., S–H prolongation) of 5 -carbamates and
-thionocarbamates of tecadenoson
5
R1
R2
X
K
i
(nM)a
GTP gS
CPA (%)
S–H prolongation, relative potencyc
Partial or full
d
Compd
b
%
2
Me
Me
Me
Me
Et
Et
Et
Pr
Pr
H
Cl
H
Cl
H
H
Cl
H
H
H
H
H
H
H
H
H
H
H
Cl
H
H
H
H
H
H
H
H
H
H
—
O
O
S
S
O
S
S
O
S
O
S
O
S
O
S
O
S
O
O
S
O
S
O
S
O
S
O
O
S
140ꢀ33
96ꢀ34
45ꢀ11
71ꢀ27
167ꢀ78
33ꢀ5
76
77
77
77
80
77
84
73
89
49
70
91
84
82
82
67
79
82
69
95
80
82
44
93
42
78
48
87
85
93
++
++
++++
++++
—
++
++++
—
+++
+
Inactive
Partial
Partial
Partial
Partial
—
Full
Partial
—
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
1
55ꢀ31
235ꢀ6
50ꢀ21
1409ꢀ286
157ꢀ5
78ꢀ35
36ꢀ4
Partial
—
—
Bu
Bu
i-Propyl
i-Propyl
c-Propyl
c-Propyl
c-Butyl
—
—
—
—
—
++
++
++
—
—
—
—
—
—
—
—
—
—
—
209
157ꢀ62
306ꢀ136
148ꢀ11
321
c-Butyl
—
c-Pentyl
c-Pentyl
c-Pentyl
c-Hexyl
c-Hexyl
Benzyl
Partial
Partial
Full
—
—
—
—
—
—
—
226ꢀ146
72
584
253ꢀ1
1826ꢀ40
12ꢀ3
Benzyl
4-F-Benzyl
N-Allyl
N-Allyl
1913ꢀ423
35ꢀ8
524ꢀ204
N,N-DiMe
N,N-DiMe
238
205ꢀ57
3ꢀ0.4
—
—
—
—
Full
0
tecadenoson
5 -OH
—
+++++
a
b
c
3
1 1
Binding affinity for A AdoR was determined using DDT cell membranes (hamster vas deferens smooth muscle cell line) with [ H]-CCPA as the
radioligand. Data shown are meanꢀstandard deviation (n=2–4).
35
1 1
Measurement of the stimulation of binding of [ S] GTPgS to G protein-coupled A AdoR in DDT cell membranes of test compound (10 uM), as
%
of CPA (1 uM).
The EC50 for agonist-induced S–H prolongation in guinea pig isolated hearts, defined as follows: +means an EC50 >5 uM, ++ means EC50 between
and 5 uM, +++ means 0.5 mM>EC50 between 1.0 uM, ++++ means 0.5 uM >EC50 between 0.51 uM, +++++ means EC50 <0.5 uM.
An agonist whose concentration–effect relationship for prolongation of the S–H interval reaches a plateau at a concentration greater than 50-fold
the concentration wherein the prolongation of the S–H interval starts. A full agonist causes second or third degree AV block in the guinea pig
isolated heart.
3
d

V. P. Palle et al. / Bioorg. Med. Chem. Lett. 14(200 4) 535–539
537
0
a ligand without increasing its intrinsic efficacy. In this
study, we compare the effect of structural modifications
of the 5 -substituent of tecadenoson derivatives with
0
previously described partial agonist, 5 -N-methylcarba-
mate 2 (Fig. 1), provided a tool to demonstrate that
partial agonism of the A AdoR with respect to the
0
those of known 5 deoxy-CPA analogues to understand
1
negative dromotropic effect at the A-V node can be
4
the nature of the ligand–receptor interactions that lead
to affinity versus IE. The agonist-induced increase of
separated from other A AdoR-mediated effects. The
1
major challenge for us was to discover a potent partial
0
3
5
0
[
efficacy in selecting or inducing an active state of the
S]-GTPgS binding is used as an indication of agonist
agonist of the 5 -carbamates and 5 -thionocarbamates
of tecadenoson. Our goal was to understand what
0
affinity and/or IE, and to demonstrate that it was pos-
receptor that stimulates release of GDP from a G
3
structural features of the 5 -substituent contribute to
5
protein, thus allowing [ S]-GTPgS, a stable analogue
3
5
of GTP, to bind. In the [ S]-GTPgS assay, the increase
sible to make a potent partial agonist of the A AdoR.
1
3
5
of [ S]-GTPgS binding induced by a partial agonist is
reported as a percentage of the increase of binding
induced by a full agonist (e.g., CPA, 5, Fig. 1).
In addition we determined the selectivity, metabolic
stability, and the pharmacokinetic (PK) properties of
several members of the series to assess their suitability
as an orally active partial A AdoR agonist to provide
1
ventricular rate control during AF.
0
prepared from the known 2 ,3 ,5 -triacetoxy-2,6-dichlor-
opurine riboside as shown in Scheme 1 to afford com-
pounds 2–39 in Table 1. Selective addition of the
commercially available (R)-3-aminotetrahydrofuran at the
0
The 5 -carbamate and 5 -thionocarbamate classes were
0
0
0
11
0
The structural features of 5 -methylcarbamate 2 that
were varied include the length of the alkyl chain, mono-
alkyl versus dialkyl, oxocarbamates versus thiono-
carbamates, cycloalkyl ring size, introduction of a pi
system, and for selected compounds, the addition of a 2-
chloro substituent (Table 1). Comparing the effect of
alkyl chain length within the oxo and thionocarbamate
series, compounds 2, 15, 18, and 20 or compounds 13,
16, 19, and 21, respectively, the trend for the effect on
6-position of compound 9 over the 2-position was made
possible by reacting at moderate temperatures (refluxing
ethanol 16–20 h) in the presence of triethylamine.
Removal of the triacetate protecting groups from the
intermediate addition product by treatment with
ammonia in methanol yielded the 2-chloro analogue
of tecadenoson. This intermediate and tecadenoson 1
were converted to the isopropylidene (IP) protected
compounds 10 and 11 by treatment with p-toluene-
sulfonic acid (p-TsOH) and 2,2-dimethoxypropane in
dimethylformamide at high temperature for an exten-
affinity for the A AdoR was methyl=ethyl=propyl
1
0
0
thionocarbamate maybe contributing partly to affinity,
>butyl (Table 1). The NH of the 5 -carbamate and 5 -
0
0
since the 5 -N,N-dimethylthionocarbamate 39 and 5 -
N,N-dimethylcarbamate 38 had lower affinity than their
mono-alkyl comparators, compounds 13 and 2, respec-
tively. When comparing the oxo- versus thionocarba-
mates with the same alkyl chain length, a general trend
developed with the thionocarbamates having a 3–9-fold
higher affinity (compounds 2 and 13, 15 and 16, 18 and
19, 20 and 21). The isopropyl analogues 22 and 23 had
slightly higher affinity than the related cyclopropyl ana-
logues 24 and 25 (Table 1). The effect of varying the ring
ꢁ
ded period of time (70 C, 40 h) in a similar manner
1
1
0
bamate or 5 -thionocarbamate groups was effected
to published procedures. Introduction of the 5 -car-
0
using an excess (4 equivalents) of carbonyldiimidazole
(
CDI) or thiocarbonyldiimidazole (thio CDI), respec-
tively. After quenching the excess reagent through the
addition of water, the intermediate acyl imidazolide
or thiono imidazolide are reacted directly with the
0
0
0
appropriate amine to afford the corresponding 2 ,3 -
size on affinity within the 5 -cycloalkyl carbamate and
thionocarbamate series, respectively, followed the trend
0
Table 1, respectively. The 2 ,3 -IP protecting group
0
IP protected 5 -carbamates or 5 -thionocarbamates of
1
2
0
was removed by treatment with warm (90 C) acetic
0
cyclopropyl=cyclobutyl=cyclopentyl>cyclohexyl with
the exception of 5 -cyclopentyl thionocarbamate 30 that
ꢁ
0
acid/water (4:1) followed by purification of the final
products which were tested for biological activity by
preparative thin layer chromatography (PTLC).
possessed higher affinity (compounds 24–32, Table 1).
The effect of introduction of a pi substituent was very
favorable with respect to the affinity of the N-benzyl
1
3
0
and N-allyl 5 -thionocarbamates, compounds 34 and
The SAR for the affinity to the A adenosine receptor
1
36, respectively, but much less favorable with the
corresponding oxo-carbamates, compounds 33 and 37
(Table 1). The introduction of the 2-chloro substituent
had little effect on affinity or IE based on the following
comparisons: compounds 2 and 12, 13 and 14, 16 and
17, 28 and 29. We anticipated a pronounced reduction
in IE for the 2-chloro analogues based on the previously
3
5
0
carbamates and 5 -thionocarbamates of tecadenoson is
and relative IE ( [ S]-GTPgS binding) of a series of 5 -
0
shown in Table 1, along with the relative potency of
selected compounds to prolong the S–H interval in gui-
nea pig isolated heart preparations. The discovery of the
0
14
15
described 2 -deoxy
and N-6-alkylsulfide
series
wherein the 2-chloro group lowered the IE, and in the
case of the latter series to the point of excluding cardi-
ovascular effects altogether.
Based on their affinity and GTPgS binding, selected
members of the 5 -carbamate and 5 -thionocarbamate
0
class were tested for their negative dromotropic effect in
0
4
Scheme 1.
guinea pig isolated hearts as previously described. All

538
V. P. Palle et al. / Bioorg. Med. Chem. Lett. 14(200 4) 535–539
of the compounds were found to recruit less GTPgS
binding than the full agonist CPA (Table 1), and they
would be predicted to be partial based on this assay. All
of the analogues tested in the isolated heart were found
incubation with human and rat liver S-9 fractions, and
the amount of parent compound remaining was recor-
ded after 30 min of incubation. The metabolic stability
of the lead compounds in human liver S-9 fraction are
presented in Table 2. Similar results were obtained using
rat liver S-9. Both the 2-chloro analogues 14 and 17 had
<65% of the parent compound remaining indicating
poor metabolic stability. The thionocarbamates 13 and
19 demonstrated high metabolic stability with at least
98% of the parent compound remaining at 30 min.
However, low but significant amounts (0.5–2%) of
tecadenoson were detected (Table 2). The original N-
methylcarbamate 2 was found to be the most stable in
the liver S-9 studies with only a trace of tecadenoson
detected. Therefore, the parent N-methylcarbamate 2
to be partial with the exception of compounds 16 and
3
0
urea series (compound 4, Fig. 1) wherein almost all
0. This result is in contrast to the corresponding 5 -
1
6
derivatives were full agonists. There is a trend that the
removal of the hydrogen bond donor capabilities
0
directly at the 5 -position on the agonist has a pro-
pensity to lead to an agonist with lower IE. More
importantly, several compounds had an EC50 of <1 uM
for their negative dromotropic effect. The more potent
analogues from the isolated heart study were further
evaluated for their selectivity of binding, and metabolic
stability in an in vitro liver S-9 incubation study (Table
0
0
and the 2 ,3 -bis acetate pro-drug 3 were further eval-
uated for their pharmacokinetic properties following
2). All five lead compounds demonstrated high binding
selectivity for the A1 AdoR over the A2A and A2B
0
0
dosing in rats (Figs. 1 and 2). The 2 ,3 -bis acetate pro-
drug 3 provided nearly twice the Cmax and the plasma
area under the curve (AUC) as the parent compound 2
following oral dosing in rats (Fig. 2a). However, low
concentrations of tecadenoson (up to 5 ng/mL) were
detected in rat plasma following oral administration of
receptors, but only compounds 13 and 19 demonstrated
high selectivity over the A AdoR (Table 2). There is
3
precedence for high affinity A AdoR agonists to have
1
7
,8
high affinity at the A AdoR as well. The metabolic
3
stability of each lead compound was determined by
Figure 2. (panel a) Concentration of parent 2 in plasma following oral gavage in rats (5 mg per kg, MPK, or equivalent) of either the parent com-
pound 2 (CVT-2759) or the bis-acetate pro-drug 3 (CVT-2771); (panel b) the oral bioavailability of the parent 2 (CVT-2759) following oral dosing of
the bis pro-drug 3 (CVT-2771) in rats was found to be approximately 9% as determined by area under the curve measurements.
Table 2. Selectivity relative to the A
0 min in the S-9 study
1
AdoR, metabolic stability in human liver S-9 incubation, and the percentage of tecadenoson detected at
3
ratio A2A/A a
K
ratio A2B/A
b
K
ratio A
/A c
Liver S-9 stability remaining (%)
d
% tecadenoson detectede
Compd
K
i
1
i
1
i
3
1
2
1
1
1
1
(2759)
28
135
49
56
135
28
135
49
56
135
7.5
23
3.3
5.3
23
ꢂ100
99
62
Trace
ꢃ0.5–1
0
0
1–2
3 (3109)
4 (3463)
7 (3462)
9 (3111)
41
98
a
b
c
Affinity for A2A AdoR was determined in a competition study using 3H-CGS21680 as the radioligand in the HEK-293 cells that were stably
transfected with human A2A AdoR (HEK-A2A).
Affinity for A2B AdoR was determined in a competition study using 3H-ZM-241385 as the radioligand in the HEK-293 cells that were stably
transfected with human A2B AdoR (HEK-A2B).
Affinity for A
with human A
3
AdoR was determined in a competition study using 125I-AB-MECA as the radioligand in the CHO cells that were stably transfected
3
AdoR (CHO-A
Compounds were incubated with 2 mg/mL of human liver microsomes (BDBioscience/Gentest) at 37 C for 30 min in a pH 7.4, 0.05 M phosphate
buffer containing 5 mM MgCl , and 2 mM NADPH.
The % of tecadenoson present after 30 min in the liver S-9 assay as determined by mass spec analysis.
3
).
d
e
ꢁ
2

V. P. Palle et al. / Bioorg. Med. Chem. Lett. 14(200 4) 535–539
539
either the parent compound 2 (5 mg/kg) or molar
equivalents of pro-drug 3. The absolute bioavailability
of compound 2 following the oral administration of the
pro-drug 3 resulted in a value of approximately 9%
4. Wu, L.; Belardinelli, L.; Zablocki, J. A.; Palle, V.;
Shryock, J. C. Am. J. Physiol. 2001, 280, H334.
5
. Snowdy, S.; Liang, H. X.; Blackburn, B.; Lum, R.; Nel-
son, M.; Wang, L.; Pfister, J.; Sharma, B. P.; Wolff, A.;
Belardinelli, L. Br. J. Pharmacol. 1999, 126, 137.
. Sorbera, L. A.; Castaner, J.; Martin, L.; Bayes, M. Drugs
of the Future 2002, 27, 846.
(Fig. 2b), and the oral half life was approximately 6–8 h.
6
7
0
0
In this study, many 5 -carbamoyl and 5 -thionocarba-
moyl analogues of tecadenoson demonstrated high affi-
nity to the A adenosine receptor and had lower relative
. van der Wenden, E. M.; von Frijtag Drabbe Kuenzel, J.;
Mathot, R. A. A.; Danhof, M.; IJzerman, A. P.; Soudijn,
W. J. Med. Chem. 1995, 38, 4000.
1
IE than the full agonist CPA based on the GTPgS
binding assay. The trend for members of these classes to
have lower IE maybe due to the removal of the hydro-
8. van Tilburg, E. W.; van der Klein, P. A. M.; von Frijtag
Drabbe Kuenzel, J.; de Groote, M.; Stannek, C.; Lor-
enzen, A.; IJzerman, A. P. J. Med. Chem. 2001, 44, 2966.
0
9
. Van der Graaf, P. H.; Van Schaick, E. A.; Visser, S. A.;
De Greef, H. J.; IJzerman, A. P.; Danhof, M. J. Pharma-
col. Exp. Ther. 1999, 290, 702.
gen bond donor capabilities at the 5 -position (i.e.,
0
removal of 5 -OH). Compounds 13, 14, and 17 were
found to be potent, partial agonists with respect to their
negative dromotropic effect at the A-V node in guinea
pig isolated hearts. Unfortunately, these compounds
either were metabolized to the full agonist tecadenoson or
had low in vitro metabolic stability (liver S-9). Encour-
aged by the oral availability of the weak partial 2 follow-
ing dosing of the bis prodrug 3, future efforts will attempt
to incorporate more optimal pharmaceutical properties
and potent, partial agonism within the same molecule.
1
0. Kenakin, T. In Molecular Pharmacology: A Short Course;
Blackwell Science: Oxford, UK, 1997; p 114.
1. Robins, M. J.; Uznaski, B. Can. J. Chem. 1981, 59, 2601.
1
1
0
2. To a solution of 5 -alcohol in dry tetrahydrofuran (0.1 M)
was added dry carbonyldiimidazole or thiocarbonyldiimi-
ꢁ
dazole (4 equiv) at 23 C under argon. After 1 h, 50 mL of
water was added to quench excess reagent, and then the
corresponding amine (1.5 equiv) was added in one por-
tion. After 16 h, the reaction was concentrated and pur-
ified by preparative thin-layer chromatography to afford
0
0
the 5 -carbamate or 5 -thionocarbamate.
3. The final compounds were purified by preparative thin
layer chromatography (methanol/methylene chloride 1:19
1
Acknowledgements
0
0
to 1:10) to afford the 5 -carbamate or 5 -thionocarbamate
0
We thank Brent Blackburn and Luiz Belardinelli for
their thoughtful discussions regarding this work.
0
in the 2 ,3 -IP protected form.
1
1
4. Vittori, S.; Lorenzen, A.; Stannek, C.; Costanzi, S.; Vol-
pini, R.; IJzerman, A. P.; von Frijtag Drabbe Kunzel,
J. K.; Cristalli, G. J. Med. Chem. 2000, 43, 250.
5. Knutsen, L. J. S.; Lau, J.; Petersen, H.; Thomsen, C.;
Weis, J. U.; Shalmi, M.; Judge, M. E.; Hansen, A. J.;
Sheardown, M. J. J. Med. Chem. 1999, 42, 3463.
References and notes
1
2
. Tsang, T. S.; Gersh, B. Am. J. Med. 2002, 113, 432.
. Wang, D.; Shryock, J. C.; Belardinelli, L. Circ. Res. 1996,
16. Ibrahim, P. N.; Ahmed, H.; Palle, V. P.; Varkhedkar, V.
M.; Blackburn, B.; Gao, Z.; Ozeck, M.; Belardinelli, L.;
Zablocki, J. A. Abstracts of Papers, 223rd National
Meeting of the American Chemical Society, Orlando, FL,
April 7–11, 2002; MEDI-025.
78, 697.
3
. Endoh, M.; Maruyama, M.; Taira, N. J. Cardiovasc.
Pharmacol. 1983, 5, 131.