Cardiovascular hybrid drugs: New benzazepinone derivatives as bradycardic agents endowed with selective β1-Non-competitive antagonism

Article abstract of DOI:10.1016/S0968-0896(02)00621-1

The synthesis and pharmacological profile of some hybrid compounds bearing both the benzazepinone moiety present in Zatebradine and typical β-blocker aryloxypropanolamine groups are described. The new compounds proved to be endowed with negative chronotropic and inotropic activity and are weak vasorelaxant agents. The cardiodepressant action is probably due to selective β₁-noncompetitive reversible antagonism. Both enantiomers of the most active compound 5c were synthesized and they showed a different cardiovascular profile, that is (+)-(R)-enantiomer displays affinity for cardiac β₁-adrenoceptors, while (-)-(S)-enantiomer shows specificity for vessel smooth muscle.

Full text of DOI:10.1016/S0968-0896(02)00621-1

Bioorganic & Medicinal Chemistry 11 (2003) 1353–1361
Cardiovascular Hybrid Drugs: New Benzazepinone Derivatives
as Bradycardic Agents Endowed with Selective
ꢀ1-Non-competitive Antagonism
Alessandra Bisi,a,* Angela Rampa,a Roberta Budriesi,a Silvia Gobbi,a Federica Belluti,a
Pierfranco Ioan,a Ermanno Valoti,b Alberto Chiarinia and Piero Valentia
aDepartment of Pharmaceutical Sciences, University of Bologna, via Belmeloro 6, 40126 Bologna, Italy
bInstitute of Medicinal Chemistry, University of Milano, viale Abruzzi 42, 20131 Milano, Italy
Received 10 July 2002; accepted 22 November 2002
Abstract—The synthesis and pharmacological profile of some hybrid compounds bearing both the benzazepinone moiety present in
Zatebradine and typical b-blocker aryloxypropanolamine groups are described. The new compounds proved to be endowed with
negative chronotropic and inotropic activity and are weak vasorelaxant agents. The cardiodepressant action is probably due to selective
b1-noncompetitive reversible antagonism. Bothenantiomers of the most active compound 5c were synthesized and they showed a dif-
ferent cardiovascular profile, that is (+)-(R)-enantiomer displays affinity for cardiac b1-adrenoceptors, while (À)-(S)-enantiomer shows
specificity for vessel smoothmuscle.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
butyl as an N-substituent, linked to an aromatic or het-
erocyclic ring system.
The pharmacological treatment of diseases having com-
plex heterogeneous pathogenesis could represent a
problem, as a single drug is not always able to ade-
quately control the illness and the combinations of
drugs with different pharmaco-therapeutic profile may
be needed. The principle of combination drug therapy
can be achieved by either using concomitant admini-
stration of two or more single active drugs or by drugs
in which the single active agents are combined in one
molecule, i.e., hybrid molecules. The design of hybrid
molecules as potential cardiovascular drugs is based on
calcium, sodium or potassium channel blockers and on
b1-adrenoreceptor antagonists.
Although the basic aryloxypropanolamine nucleus
should remain intact for significant b-adrenoceptor
antagonist activity, a wide variety of aromatic ring or
nitrogen substituents can be tolerated. This has allowed
for the design of therapeutic agents that combine
b-adrenoceptor antagonist activity withseveral other
useful pharmacological actions including a-adreno-
ceptor blockade,3 calcium channel blockade,4 direct
vasodilatation,5 inhibition of angiotensin converting
enzyme6 and class III antiarrhythmic activity.7
Zatebradine (1) (Fig. 1) was described in the mid 80’s as
the representative of a novel pharmacological class
termed ‘specific bradycardic agents’.8,9 Since this mole-
cule was found to reduce heart rate without concomitant
b-Adrenoceptor antagonists represent a class of drugs
widely used clinically for hypertension or angina pec-
toris in various countries.1,2 Withthe exception of a few
phenylethanolamines virtually all the clinically useful
b-adrenoceptor antagonists contain
propanolamine moiety, typically withisopropyl or tert-
a
phenoxy-
*Corresponding author. Tel.: +39-051-2099710; fax: +39-051-
2099734; e-mail: albisi@alma.unibo.it
Figure 1. Structure of zatebradine (1).
0968-0896/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0968-0896(02)00621-1

1354
A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
Scheme 1. Reagents: (a) N-methylbenzylamine; (b) H2, Pd/C, CH3COOH; (c) selected epoxide.
negative inotropic or hypotensive effects,10 its mechan-
ism of action has been deeply investigated in these
years11,12 and now its activity is believed to be due to
the selective block of the If current.13
depressant action is probably due to selective b1-non
competitive reversible antagonism.
Chemistry
With regard to the bradycardic activity, the benzaze-
pinone ring, the three carbon chain and the basic nitrogen
atom are very sensitive to structural modifications. On
the other hand, the arylalkyl moiety exhibits consider-
able potential for structural variations.14,15 It seemed
reasonable to assume16 that this moiety could influence
heart rate, blood pressure and the duration of biological
activity. Indeed, compounds bearing peculiar fragments
of bradycardic drugs acting withdifferent mechanisms
might show enhanced therapeutic efficacy.
The synthesis of compounds 5a–e was accomplished
following Scheme 1.
The key intermediate, 3-(3-chloropropyl)-7,8-dimethoxy-
2,3-dihydro-1H-3-benzazepin-2-one 2 was prepared by
standard procedures.14 Condensation with N-methyl-
benzylamine afforded compound 315 and subsequent
hydrogenation gave compound 4,16 which was condensed
with the selected epoxide to give the final products 5a–e.
Furthermore, as Smith and co-workers demonstrated
that amidic substituents, which are powerful hydrogen
bond acceptors, produced very potent and cardioselective
drugs,17,18 it is foreseeable that the benzazepinone moi-
ety might increase the cardioselectivity of the selected
b-blocking drugs by an increased receptor binding with
its amidic group.
Synthesis of compound 8 was described in Scheme 2.
The selected epoxide was condensed with benzylamine
and the intermediate 6 was treated with 2 to give the
benzyl derivative 7. Hydrogenation using Pd/C as cata-
lyst afforded compound 8.
Synthesis of the pure enantiomers of 5c was performed
as shown in Scheme 3. The enantiomers of (R)- and
(S)-isopropylidene-glycerol [(S)-9 and (R)-9)], easily
obtained by resolution of racemic glycerol,19 were used
as the key chiral building blocks in the enantioselective
In this paper we report the synthesis and the pharmaco-
logical profile of a number of hybrid compounds in
which the arylalkyl moiety of 1 is replaced withpeculiar
fragments of b-blocking agents, as shown in Figure 2.
The results obtained show that the new hybrid mole-
cules 5a–e, 8, (À)-(S)-5c and (+)-(R)-5c are endowed
withnegative crhonotropic and inotropic activity.
(À)-(S)-5c is also weak vasorelaxant agent. The cardio-
Scheme 2. Reagents: (a) benzylamine; (b) 3-(3-chloropropyl)-7,8-
dimethoxy-2,3-dihydro-1H-3-benzazepin-2-one (2), triethylamine;
(c) H2, Pd /C.
Figure 2. Structure of compounds.

A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
1355
Scheme 3. Reagents: (a) o-cresol, TPP, DEAD, THF 0 ꢀC; (b) HCl 1 N, 80 ꢀC; (c) TPP, DEAD, benzene, 80 ꢀC; (d) 7,8-dimethoxy-3-[3-(methyla-
mino)propyl]-2,3,4,5-tetrahydro-1H-3-benzazepin-2-one (4), 2-propanol, 82 ꢀC.
synthesis of the title compounds. The first step (com-
pounds 10) involved reaction between o-cresol and the
requisite enantiomer of compound 9 accomplished by
using triphenylphosphine and diethylazodicarboxylate.
we thought it would be of interest to replace the aryl-
alkyl moiety of 1 withdifferent 3-aryloxy-2-hydroxy-
propyl fragments of b-blocking drugs to obtain hybrid
molecules which could possibly influence different
cardiovascular parameters and consequently control
heterogeneous pathologies.
Successive acidic hydrolysis of the cyclic ketal (R)-10
and (S)-10 provided (S)-11 and (R)-11 respectively,
having an higher than 98% enantiomeric excesses on the
basis of the HPLC analysis performed with a chiral sta-
tionary phase Chiralcel OD. This diols were converted
into the epoxides (S)-12 and (R)-12 by reaction withdi-
ethyl azodicarboxylate and triphenylphosphine. The
opening of the (S)-12 withthe amine 4 gives the com-
pounds 5c having an S configuration at C-2. Analogously,
the corresponding (R)-5c was obtained from (R)-12.
The cardiovascular profile of the synthesized com-
pounds was evaluated in isolated guinea-pig atria and
aorta, namely spontaneously beating right atria, left
atria driven at 1 Hz and K+-depolarized aortic strips.
The results obtained from myocardial and vascular
preparations are summarized in Table 1.
It can be observed that all new hybrid compounds
showed bradycardic activity as well as the reference
compound 1. However, the introduction of 3-aryloxy-
2-propanol moiety does not change the order of
potency, and this could be explained by the presence
of the benzazepinone ring in all new compounds. As
far as negative inotropic activity is concerned, all
hybrid molecules elicited a fair potency with respect
to 1. At a first glance, the presence of an ortho sub-
stituent on the phenyl ring in our series of com-
pounds could be of relevant importance for
inotropism. In particular, it must be noted that the
maximum negative inotropic potency seems to be due
to the presence of electron donating groups, though
Pharmacology
The pharmacological profile of compounds was tested on
guinea-pig isolated left and right atria to evaluate their
inotropic and chronotropic effects, respectively, and on
K+-depolarized guinea-pig aortic strips to assess calcium
antagonist activity. At first all compounds were checked
at increasing doses to evaluate the percent decrease on
developed tension on isolated left atrium driven at 1 Hz
(negative inotropic activity), the percent decrease in atrial
rate on spontaneously beating right atrium (negative
chronotropic activity) and the percent inhibition of cal-
cium-induced contraction on K+-depolarized aortic strips
(vasorelaxant activity). Data were analyzed by Student’s
t-test. The potency of drugs defined as EC50, EC30 and
IC50 was evaluated from log concentration–response
curves (Probit analysis by Litchfield and Wilcoxon,
n=6–8) in the appropriate pharmacological preparations.
weak suchas a methyl (compound
5c). On the con-
trary, a decrease in the negative inotropic efficacy
seems to be related in a proportional manner to the
increase of the electron withdrawing effect of the sub-
stituent, e.g., –NO2 in 5b and –COCH3 in 5d, which
showed comparable negative inotropic profile. How-
ever, it is interesting to note that these latter com-
pounds are more active than the phenyl unsubstituted
derivative 5a. The insertion of a naphthyl substituent
(5e), further lowered the negative inotropic potency.
Moreover, the removal of ortho phenyl substituent, as
in compound 5a or, in addition, the presence of a
secondary amine, as in compound 8 bothproduced a
sharp drop in negative inotropic properties. Among all
compounds only 5c showed a weak vasodilatory
activity, probably due to calcium antagonist action as
it was able to inhibit the potassium induced contrac-
tion of guinea pig aorta.
b1- And b2-adrenergic activities were determined on the
atria and trachea of guinea-pig, respectively. The non-
competitive b1-antagonist potency was expressed as
pIC50 value. All data are presented as mean Æ SEM.20
Results and Discussion
As the benzazepinone ring in Zatebradine (1) is known
to be of fundamental importance for bradycardic activity,

1356
A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
Table 1. Cardiovascular activity of compounds
N
Cardiovascular activity
Negative chronotropy
Negative inotropy
Vasorelaxant activity
Ia%a
(ÆSEM)
EC50 (mM)b
(95% cl)
Ia%c
(ÆSEM)
EC30 (mM)b
(95% cl)
Ia%d
(ÆSEM)
IC50 (mM)b
(95% cl)
1
5a
5b
5c
45Æ3.1
87Æ3.1
89Æ2.4
73Æ0.6e
92Æ1.7g
32Æ1.1g
88Æ2.5
91Æ4.2
68Æ3.5
70Æ2.8e
76Æ2.5f
80Æ6
0.11 (0.08–0.13)
0.12 (0.08–0.21)
0.56 (0.45–0.68)
0.23 (0.20–0.28)
0.20 (0.17–0.24)
0.28 (0.22–0.36)
0.37 (0.29–0.45)
0.40 (0.28–0.53)
0.80 (0.71–0.92)
5Æ0.4
25Æ1.5
40Æ4.5
52Æ2.0
42Æ3.6
58Æ1.8
36Æ1.1
42Æ2.5
23Æ1.4f
39 (30–49)
1.20 (0.9–1.5)
0.14 (0.10–0.19)
0.59 (0.51–0.69)
78Æ1.8f
74Æ2.2e
73Æ0.4e
90Æ6.1
86Æ4.7
84Æ1.9
36 (29.2–41.8)
20 (16–25)
(+)-(R)-5c
(À)-(S)-5c
5d
5e
8
1.00 (0.8–1.3)
3.90 (2.6–5.3)
18.6 (16.7–21.2)
aIntrinsic activity: decrease in the developed tension in isolated guinea-pig left atrium at 10À4 M, expressed as percent changes from the control
(n=5–6). The left atria were driven at 1 Hz. The 10À4 M concentration gave the maximum effect for most compounds.
bCalculated from log concentration–response curves (Probit analysis by Litchfield and Wilcoxon with n=6–7). When the maximum effect was
<50%, the EC50 ino., EC30 chrono., IC50 values were not calculated.
cIntrinsic activity: decrease in the atrial rate on guinea-pig spontaneously beating isolated right atrium at 5Â10À5 M, expressed as percent changes
from the control (n=7–8). The pretreatment heart rate ranged from 165 to 190 beats/min. 5Â10À5 M gave the maximum effect for most compounds.
dIntrinsic activity: percent inhibition of calcium-induced contraction on K+-depolarized guinea-pig aortic strip at 10À4 M (n=5–6). The 10À4 M
concentration gave the maximum effect for most compounds.
eAt 5Â10À6 M.
fAt 10À5 M.
gAt 5Â10À5 M.
The b1-adrenergic antagonist profile was evaluated on
different isolated tissues from guinea-pig, namely right
atria (b1) and tracheal chains (b2). The results are
assembled in Table 2 and graphically shown in Figure 3.
an ortho electron-donating group on the phenyl ring
seems to be important for optimum b1-affinity as com-
pound 5c proved to be the best reversible non competi-
tive antagonist. On the other hand, the introduction of
electron withdrawing groups in the phenyl ring
decreased b1-adrenergic activity in an inversely propor-
tional manner with respect to the electron withdrawing
effect, since compound 5b bearing an ortho –NO2 sub-
stituent proved less active than compound 5d bearing
ortho –COCH3 group. Furthermore the presence of a
secondary nitrogen in the backbone of the molecule
significantly reduced b1-affinity (8).
Concerning b-adrenergic antagonist activity compound
1 is inactive, as expected, while all hybrid compounds
do show reversible b1 selective non competitive anta-
gonism and are devoid of any type of interaction with b2
receptors (see Table 2). Indeed right atrial isoprenaline
response is fully recovered after 30 min washing of tis-
sues inhibited by compounds at 1 mM concentration
(Table 3).
Since the importance of stereochemistry for interaction
of adrenergic drugs with b-receptors is well recognized,
we performed a stereoselective synthesis and pharma-
cological evaluation of the enantiomers of compound
The replacement of dimethoxyphenylalkyl moiety of 1
withan o-methylariloxypropanol group (5c) resulted in
a significant improvement of the reversible non com-
petitive antagonism at b1-adrenergic receptors. In addi-
tion, it should be noted that in this series the presence of
Table 2. b-Adrenergic antagonistic activity of compounds
a
pIC50
N
b1
b2
Isolated guinea-pig
right atria
Isolated guinea-pig
tracheal strips
1
5a
5b
5c
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
6.82Æ0.01
5.56Æ0.06
7.12Æ0.05
7.39Æ0.03
6.45Æ0.04
6.09Æ0.01
6.77Æ0.01
5.28Æ0.08
(+)-(R)-5c
(À)-(S)-5c
5d
5e
8
Figure 3. Guinea-pig right atria; non-competitive antagonism among
(À)-isoprenaline, 5a (*), 5b (^), 5c (!), 5d (~), 5e (&) and 8 (~) at
cardiac b1-adrenoceptor. The results are expressed as the mean Æ
SEM of experiments performed in triplicate.
apIC50 is the negative logarithm of the molar concentration that redu-
ces the response to (À)-isoproterenol by 50%. NA=not active.

A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
1357
Table 3. Inhibitory effect of compounds on (À)-isoproterenol-
induced increase of right atrium frequency before and after washout
N
% b1-Blockadea
30 Min incubation
30 Min incubation
+ 30 Min washing
1
5a
5b
5c
NA
75Æ4.1
40Æ2.8
78Æ5.6
94Æ3.8
48Æ3.7
52Æ3.6
88Æ4.2
47Æ3.3
3Æ0.6
0
3Æ0.1
(+)-(R)-5c
0
(À)-(S)-5c
0
5d
5e
8
1.5Æ0.1
4Æ0.3
2Æ0.2
aCompounds were tested at 1 mM concentration and the result are
expressed as the percentage decrease of maximum response to (À)-
isoproterenol (M Æ SEM) after 30 min incubation and after 30 min
incubation followed by washing for 30 min. NA=not active.
higher affinity and selectivity at b1, withrespect to b2
adrenoceptors (see Table 2), which suggests that the two
receptor subtypes have different stereochemical require-
ments. Furthermore, it should be noted that compound
(+)-(R)-5c displayed a reversible non competitive
antagonism comparable to bopindolol’s which is, at our
best knowledge, the most potent slowly reversible b1
non competitive antagonist so far described.21 A note-
worthy feature of compound (À)-(S)-5c is that the
(À)-(S)-enantiomer elicited a remarkable vasorelaxant
activity (IC50=20 mM) if compared to that of the other
hybrid compounds. Probably, this action on vasculature
is due to a calcium antagonist effect as (À)-(S)-5c
inhibited potassium (80 mM)-contracted aortic smooth
muscle. However, (À)-(S)-enantiomer maintained a
negative chronotropic potency comparable to that of
(+)-(R)-enantiomer.
In conclusion, keeping the pharmacophoric requirements
of Zatebradine, on the basis of the above considerations it
may be inferred that in these new hybrid molecules the
presence of aryloxypropanol moiety produces high affi-
nity and selectivity for b1-adrenoreceptors. Moreover,
the investigation on the enantiomers of the most potent
hybrid compound 5c has revealed a different cardio-
vascular profile, i.e., (+)-(R)-enantiomer displays affi-
nity for cardiac b1-adrenoreceptors, while (À)-(S)-
enantiomer shows specificity for vessel smooth muscle,
though to a lower extent. The trend noted in this series
of hybrid molecules might help in developing useful
structure–activity relationships to differentiate and to
understand structural elements which give selectivity for
b1 versus b2-adrenoreceptors.
Figure 4. (a) Negative inotropic activity (left atria driven at 1 Hz) of 5c
(!), (+)-(R)-5c (*) and (À)-(S)-5c (Ã). (b) Negative inotropic activity
(spontaneously beating right atria) of 5c (!), (+)-(R)-5c (*) and (À)-
(S)-5c (Ã). (c) Non-competitive antagonism among (À)-isoprenaline,
5c (!), (+)-(R)-5c (*) and (À)-(S)-5c (Ã) at cardiac b1-adrenoceptor
in guinea-pig right atrium. Values shown are mean Æ SEM (n=3–5).
5c, which displayed the most interesting cardiovascular
profile. The results, shown in Table 1 and graphically
shown in Figure 4, indicate that the stereochemistry of
the aryloxypropanolamine unit has a great influence on
the cardiovascular activity, namely on cardiac para-
meters suchas inotropism, chronotropism, on smooth
muscle relaxation and on b1 adrenergic affinity.
Experimental
Chemistry
General methods. All melting points were determined in
open glass capillaries using a Buchi apparatus and are
uncorrected. 1H NMR were recorded on a Varian
Gemini 300 spectrometer in CDCl3 solutions, with
Me4Si as the internal standard. Signal multiplicity was
In particular, (+)-(R)-configuration in compound
(+)-(R)-5c, showed both better cardiac efficacy and

1358
A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
designed according to the following abbreviations:
s=singlet, d=doublet, dd=doublet of doublets, t=triplet,
m=multiplet, br s=broad singlet, br t=broad triplet.
d 7.9–7.1 (m, 4H ), 6.6 (s, 1H), 6.55 (s, 1H), 4.1–3.9
(m, 3H), 3.8 (s, 3H), 3.75 (s, 3H), 3.7 (t, 2H), 3.5 (m,
2H), 3.1 (m, 2H), 2.8–2.5 (m, 4H), 2.4 (s, 3H), 2.35
(s, 2H), 1.8 (m, 2H). MS: m/z (rel. abundance) 487
(M+, 4.2), 305 (62.1), 262 (72.7) 139 (100). Anal.
(C25H33N3O7.HCl): C, H, N.
Mass spectra were recorded on a V.G. 7070 E spectro-
meter. Elemental analyses were within 0.4% of theo-
retical value unless otherwise indicated. The names of
compound were obtained using AUTONOM, PC soft-
ware for nomenclature in organic chemistry from Beil-
stein-Institut and Springer. Optical rotations were
determined by a Perkin–Elmer 241 Polarimeter at 25 ꢀC.
Purification was performed by flash chromatography
using silica gel (particle size 40–63 mm, Merck). HPLC
analyses were performed on a Chiralcel-OD column
from Daicel using a Waters 510 pump and a Pye unicam
PU 4025 detector (l=254 nm).
3-3-[2-Hydroxy-3-(2-methylphenoxy)propyl](methyl)amino]
propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-benzaze-
pin-2-one (5c). Using the previous procedure, com-
pound 5c (0.98 g, 63%) was obtained from 4 (1 g,
3.42 mmol) and 2-[(2-mꢀethylphenoxy)methyl]oxirane.24
1
mp (HCl salt) 137–140 C (methanol/diethylether). H
NMR (CDCl3) d 7.25–6.80 (m, 4H), 6.6 (s, 1H), 6.55 (s,
1H), 4.1–3.9 (m, 3H), 3.8 (s, 3H), 3.75 (s, 3H), 3.7 (t,
2H), 3.5 (m, 2H), 3.1 (m, 2H), 2.65–2.4 (m, 4H), 2.3 (s,
3H), 2.2 (s, 3H) 2.15 (s, 2H), 1.8 (m, 2H). MS: m/z (rel.
abundance) 456 (M+, 20), 305 (100), 262 (88.6). Anal.
(C26H36N2O5.HCl): C, H, N.
The (S)- and (R)-isopropylidene-glycerol, (S)-2 and
(R)-2, were prepared by chemical resolution of the
racemate as described in the literature.19
3-3-[3-(2-Acetylphenoxy)-2-hydroxypropyl](methyl)amino]
propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-benzaze-
pin-2-one (5d). Using the previous procedure, com-
pound 5d (1.11 g, 67%) was obtained from 4 (1 g,
3.42 mmol) and 1-[2-(2-oxiranylmethoxy)phenyl]-1-
ethanone.25 mp (HCl salt) 182–185 ꢀC (methanol/die-
3-3-[Benzyl(methyl)amino]propyl-7,8-dimethoxy-2,3-di-
hydro-1H-3-benzazepin-2-one (3). N-Methylbenzylamine
(13 mL, 100.0 mmol) were dropped into a solution of
15.07 g (51.0 mmol) of 2 in 200 mL of toluene, and the
reaction mixture was heated under reflux for 8 h. The
solution was then washed with water and evaporated to
dryness. The compound was an oil (12.2 g, 65%), used
for the subsequent step without further purification. Mp
(HCl salt, methanol/diethylether) 148–151 ꢀC.
1
thylether). H NMR (CDCl3) d 7.8–6.9 (m, 4H), 6.6 (s,
1H), 6.55 (s, 1H), 4.1–3.9 (m, 3H), 3.8 (s, 3H), 3.75 (s,
3H), 3.7 (t, 2H), 3.5 (m, 2H), 3.1 (m, 2H), 2.7 (s, 3H),
2.65–2.4 (m, 4H), 2.3 (s, 3H), 2.15 (m, 2H), 1.8 (m, 2H).
MS: m/z (rel. abundance) 484 (M+, 33.3), 206 (77.7), 58
(100). Anal. (C27H36N2O6.HCl): C, H, N.
7,8-Dimethoxy-3-[3-(methylamino)propyl]-2,3,4,5-tetra-
hydro-1H-3-benzazepin-2-one (4). A solution of 12.2 g
of compound 3 (32.2 mmol) in acetic acid was hydro-
genated at 5 atm withPd/C as catalyst. After filtration,
the solvent was removed under reduced pressure and the
residue was dissolved in CH2Cl2, washed with a satu-
rated solution of K2CO3, dried (Na2SO4) and evapo-
rated to dryness to yield 6.11 g of 4 (65%), mp 72–75 ꢀC
3-3-[2-Hydroxy-3-(1-naphthyloxy)propyl](methyl)amino]-
propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-benzaze-
pin-2-one (5e). Using the previous procedure,
compound 5e (1.11 g, 66%) was obtained from 4 (1 g,
3.42 mmol) and 2-[1(naphtyloxy)methyl]oxirane.26 Mp
(HCl salt) 187–189 ꢀC (methanol/diethylether). 1H
NMR (DMSO-d6) d 7.9–6.7 (m, 7H ), 6.7 (s, 1H), 6.55
(s, 1H), 4.22–4.15 (m, 3H), 4.12 (s, 3H), 4.10 (s, 3H), 3.9
(t, 2H), 3.55 (m, 2H), 3.45 (m, 2H), 3.41 (m, 2H), 2.9–
2.75 (m, 4H), 2.48 (s, 3H), 1.95 (m, 2H). MS: m/z (rel.
abundance) 492 (M+, 16.7), 305 (100), 262 (72.9). Anal.
(C29H36N2O5.HCl): C, H, N.
1
(toluene). H NMR (CDCl3) d 6.6 (s, 1H), 6.55 (s, 1H),
3.8 (s, 3H), 3.75 (s, 3H), 3.65 (t, 2H), 3.5 (m, 2H), 3.05
(m, 2H), 2.8-2.5 (m, 4H), 2.4 (s, 3H), 1.8 (m, 2H).
3-3-[(2-Hydroxy-3-phenoxypropyl)(methyl)amino]propyl-
7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-benzazepin-2-one
(5a). A solution of 1 g of compound 4 (3.42 mmol) and
0.51 g (3.42 mmol) of 2-(phenoxymethyl)oxirane22 in
50 mL of ethanol was heated under reflux 12 h. The sol-
vent was evaporated to dryness and the product was pur-
ified as hydrochloride salt: mp 183–185ꢀC (methanol/
diethylether).Yield 1.06g (70%). 1H NMR (DMSO-d6) d
7.4–6.9 (m, 5H ), 6.7 (s, 1H), 6.6 (s, 1H), 4.2–4.15 (m, 3H),
4.12 (s, 3H), 4.10 (s, 3H), 3.9 (t, 2H), 3.55 (m, 2H), 3.45 (m,
2H), 3.40 (m, 2H), 2.9–2.75 (m, 4H), 2.48 (s, 3H), 1.9 (m,
2H). MS: m/z (rel. abundance) 442 (M+, 13.3), 305 (100),
262 (79.6). Anal. (C25H34N2O5.HCl): C, H, N.
1-(Benzylamino)-3-phenoxy-2-propanol (6). 2-(Phenoxy-
methyl)oxirane (9 g, 59.9 mmol) and 7.9 mL of benzyl-
amine (59.9 mmol) in 200 mL of toluene were heated
under reflux for 16 hafter cooling, the solution was
washed with water, dried and evaporated to dryness, to
yield 10.7 g (70%) of 6 as an oil, which was used for the
subsequent step without further purification. Mp (HCl
salt) 176–180 ꢀC.
3-3-[Benzyl(2-hydroxy -3-phenoxypropyl)amino]propyl-
7,8-dimethoxy-2,3-dihydro-1H-3-benzazepin-2-one (7). A
solution of 4.7 g (15.9 mmol) of compound 2, 4.1 g
(15.9 mmol) of 6 and 2.3 mL of triethylamine in 50 mL
of toluene was refluxed 8 h. After cooling, the solution
was washed with water, dried and evaporated to dry-
ness, affording 2.6 g (32%) of 7. Mp (HCl salt)
3-3-[2-Hydroxy-3-(2-nitrophenoxy)propyl](methyl)amino]
propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-benzazepin-
2-one (5b). Using the previous procedure, compound 5b
(1.07 g, 65%) was obtained from 4 (1 g, 3.42 mmol) and
2-[(2-nitrophenoxy)methyl)]oxirane.23 mp (HCl salt)
160–163 ꢀC (methanol/diethylether). H NMR (CDCl3)
182–185 ꢀC. H NMR (CDCl3) d 8.1–7.1 (m, 10H ), 6.6
1
1

A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
1359
(s, 1H), 6.55 (s, 1H), 6.4 (s, 2H), 4.1–3.9 (m, 3H), 3.8 (s,
3H), 3.75 (s, 3H), 3.7 (t, 2H), 3.5 (s, 3H), 2.8-2.5 (m,
4H), 2.35 (s, 2H), 1.8 (m, 2H).
24.7 mmol) in benzene (50 mL) was refluxed for 12 h.
After evaporation of the solvent, ether was added to
precipitate the phosphine oxide, which was removed by
filtration. The filtrate was concentrated and the residue
purified by flash chromatography (cyclohexane/EtOAc
9:1) yielding 1.5 g of (S)-12 as a colourless oil.
[a]D=+14.8 (c=0.5, ethanol). 1H NMR (CDCl3) d
7.23–7.07 (m, 2H), 6.97–6.76 (m, 2H), 4.23 (dd, 1H),
3.97 (dd, 1H), 3.38 (m, 1H), 2.92 ( br t, 1H), 2.80 (dd,
1H), 2.27 (s, 3H).
3-3-[(2-Hydroxy-3-phenoxypropyl)amino]propyl-7,8-di-
methoxy-2,3,4,5-tetrahydro-1H-3-benzazepin-2-one (8).
Compound 7, 2.6 g (5.09 mmol) was dissolved in 30 mL
of acetic acid and hydrogenated at 5 atm using Pd/C as
catalyst. After filtration, the solvent was removed in
vacuo and the residue was dissolved in CH2Cl2, washed
witha saturated solution of K CO3, dried and evapo-
2
rated to dryness, to give 1.74g (80%) of 8 which was
purified by flashchromatography (eluant: toluene/acetone
(2R)-3-Tolyloxy-1,2-epoxypropane [(R)-12]. Prepared
from (R)-11] as described for (S)-12: [a]D=À15.0
3/2). Mp 145–148 ꢀC (HCl salt). H NMR (DMSO-d6) d
(c=0.5, ethanol). H NMR identical to (S)-12.
1
1
8.2 (broad, 1H, NH), 7.4–6.9 (m, 5H ), 6.7 (s, 1H), 6.6 (s,
1H), 4.2–4.15 (m, 3H), 4.12 (s, 3H), 4.10 (s, 3H), 3.9 (t,
2H), 3.55 (m, 2H), 3.45 (m, 2H), 3.40 (m, 2H), 2.9–2.75
(m, 4H), 1.9 (m, 2H). Anal. (C24H32N2O5.HCl): C, H, N.
3-3-[(2S)-2-Hydroxy-3-(2-methylphenoxy)propyl](methyl)
amino]propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-
benzazepin-2-one [(S)-5c]. A mixture of (S)-12 (562 mg,
3.42 mmol) and amine 4 (1 g, 3.42 mmol) in 2-propanol
(25 mL) was refluxed for 24 h. After evaporation of the
solvent, the residue was treated with DCM (20 mL) and
1 N NaOH (10 mL). The aqueous phase was separated
and extracted withDCM (2 Â 15 mL). The combined
extracts were dried and concentrated to give a residue
which was chromatographed on silica gel. Elution with
DCM/MeOH/TEA (95/4.5/0.5) yielded 1.05 g of 5c-(S)
(2R)-3-Tolyloxy-1,2-propanediol acetonide [(R)-10]. To a
cold (0 ꢀC) stirred solution of o-cresol (10 g, 92.5 mmol),
(2R)-glycerol-1,2 acetonide [(R)-9] (11 g, 83.2 mmol) and
triphenylphosphine (32.6 g 124.5 mmol) in THF (80 mL)
was added dropwise diethylazodicarboxylate (19.6 mL,
124.5 mmol). The mixture was stirred for 2 h and then
concentrated under reduced pressure. The resulting
residue was poured into diethylether to precipitate the
phosphine oxide, which was removed by filtration. The
filtrate was concentrated under reduced pressure to
afford an oil which was purified by flash chromato-
graphy (cyclohexane/EtOAc 9:1) to yield the compound
(R)-10 (9.70 g, 52% yield). [a]D=À20.9 (c=0.5, etha-
1
as oil. [a]D=À19.7 (c=0.5, CHCl3). H NMR (CDCl3)
d 7.18–7.10 (m, 2H), 6.90–6.80 (m, 2H), 6.60 (s, 1H),
6.56 (s, 1H), 4.09–3.87 (m, 3H), 3.83 (s, 6H), 3.75 (s,
2H), 3.72 (br t, 2H), 3.52 (br t, 2H), 3.05 (br t, 2H),
2.66–2.24 (m, 5H), 2.30 (s, 3H), 2.23 (s, 3H), 1.77 (m,
2H).
1
nol). H NMR (CDCl3) d 7.25–7.10 (m, 2H), 6.98–6.80
(m, 2H), 4.54 (m, 1H), 4.29–3.93 (m, 4H), 2.30 (s, 3H),
1.53 (s, 3H), 1.48 (s, 3H).
3-3-[(2R)-2-Hydroxy-3-(2-methylphenoxy)propyl](methyl)
amino]propyl-7,8-dimethoxy-2,3,4,5-tetrahydro-1H-3-
benzazepin-2-one [(R)-5c]. Prepared from (R)-12 and 4
as described for (S)-5c: [a]D=+20.3 (c=0.5, CHCl3).
1H NMR identical to (S)-5c.
(2S)-3-Tolyloxy-1,2-propanediol acetonide [(S)-10]. Pre-
pared from o-cresol and (2S)-glycerol-1,2 acetonide [(S)-9]
as described for (R)-10: [a]D=+21.5 (c=0.5, etha-
nol).1H NMR identical to (R)-10.
Functional studies
(2S)-3-Tolyloxy-1,2-propanediol [(S)-11]. (R)-10 (9.5 g,
42.7 mmol) in 75 mL of 10% HCl was heated at 80 ꢀC
for 3 h. The mixture was cooled and, after evaporation
of the acetone, extracted with CH2Cl2 overnight. The
extract was dried (Na2SO4) and concentrated. The
resulting solid was crystallized from toluene to yield
(S)-11 (5.9 g, 76% yield). Mp 93 ꢀC; [a]D=À0.6 (c=0.5,
ethanol) (e.e.>98% on the basis of the HPLC analysis
performed with a chiral stationary phase Chiralcel OD).
1H NMR (DMSO-d6) 7.22–7.09 (m, 2H), 7.00–6.79 (m,
2H), 4.94 (d, 1H, OH), 6.68 (t, 1H, OH) 4.07–3.63 (m,
3H), 3.58–3.47 (m, 2H), 2.22 (s, 3H, Ar–CH3).
Guinea-pig atrial preparations. Guinea-pigs (300–400 g
female) were sacrificed by cervical dislocation. After
thoracotomy the heart was immediately removed and
washed by perfusion through the aorta with oxygenated
Tyrode solution of the following composition (mM):
136.9 NaCl, 5.4 KCl, 2.5 CaCl2, 1.0 MgCl2, 0.4 NaH2-
PO4xH2O, 11.9 NaHCO3 and 5.5 glucose. The physio-
logical salt solution (PSS) was buffered at pH 7.4 by
saturation with95% O 2–5% CO2 gas, and the tem-
perature was maintained at 35 ꢀC. Isolated guinea-pig
heart preparations were used, spontaneously beating
right atria and left atria driven at 1 Hz. For each pre-
paration, the entire left and right atria were dissected
from the ventricles, cleaned of excess tissue, and hung
vertically in a 15 mL organ bathcontaining the PSS
continuously bubbled with95% O 2–5% CO2 gas at
35 ꢀC, pH 7.4. The contractile activity was recorded
isometrically by means of a force transducer (FT 0.3,
Grass Instruments, Quincy, MA) using Power Lab
software (Basile, Italy). The left atria were stimulated by
rectangular pulses of 0.6–0.8 ms duration and about
50% threshold voltage through two platinum contact
(2R)-3-Tolyloxy-1,2-propanediol [(R)-11]. Prepared from
(S)-10 as described for (S)-11: mp 92 ꢀC; [a]D=+0.5
(c=0.5, ethanol) (e.e.>98% on the basis of the HPLC
1
analysis on a chiral stationary phase Chiralcel OD). H
NMR identical to (S)-11.
(2S)-3-Tolyloxy-1,2-epoxypropane [(S)-12]. A mixture of
(S)-11 (3 g, 16.5 mmol), triphenylphosphine (6.48 g,
24.7 mmol) and diethylazodicarboxylate (3.88 mL,

1360
A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
electrodes in the lower holding clamp (Grass S88
stimulator). The right atrium was in spontaneous activ-
ity. After the tissue was beating for several min, a
length-tension curve was determined, and the muscle
lengthwas maintained at whichelicited 90% of maxi-
mum contractile force observed at the optimal length. A
stabilization period of 45–60 min was allowed before the
atria were used to test compounds. During the equili-
bration period, the bathing solution was changed every
15 min and the threshold voltage was ascertained for the
left atria. Atrial muscle preparations were used to exam-
ine the inotropic and chronotropic activity of the com-
pounds (0.1, 0.5, 1, 5, 10, 50 and 100 mM), first dissolved in
DMSO and then diluted with PSS. According to this pro-
cedure, the concentration of DMSO in the bath solution
never exceeded 0.3%, a concentration that did not pro-
duce appreciable inotropic and chronotropic effects. Dur-
ing the construction of cumulative dose-response curves,
the next higher concentration of the compounds was
added only after the preparation reached a steady state.
The trachea. This was dissected transversally and trans-
ferred to a dishcontaining Krebs solution, and cut
transversally between the segments of cartilage. Six of
the tracheal rings were tied together and mounted under
a tension of 1 g at 37 ꢀC in a 15 mL organ bathcon-
taining Krebs–Ringer solution of the following compo-
sition (mM): 95 NaCl, 4.7 KCl, 2.50 CaCl2,
1.00 MgSO4, 1.17 KH2PO4, 25 NaHCO3 and 10.6 glu-
cose equilibrated with95% O –5% CO2 gas at pH 7.4.
2
The tissues were allowed to stabilize for 90 min. Iso-
metric force was recorded from the preparations by a
force-displacement transducer. A constant level of tone
was induced by the addiction of carbachol chloride
(5Â10À7 M) to the bath, and, after 15 min, a control
concentration-response curve for isoprenaline was
obtained. The tissue was washed thoroughly, and
30 min later a single concentration of antagonist was
added to the bath and allowed for 30 min. During the
last 15 min. of antagonistic incubation, carbachol chlo-
ride (5Â10À7 M) was added to the bath and the cumu-
lative concentration-response curve for isoprenaline
was again determined. All responses to different con-
centrations of isoprenaline were expressed as percen-
tage of the maximal relaxation recorded for the control
curve.
Guinea-pig aortic strips. The thoracic aorta was
removed and placed in Tyrode solution of the following
composition (mM): 118 NaCl, 4.75 KCl, 2.54 CaCl2,
1.20 MgSO4, 1.19 KH2PO4, 25 NaHCO3 and 11 glucose
equilibrated with95% O –5% CO2 gas at pH 7.4. The
2
vessel was cleaned of extraneous connective tissue. Two
helicoidal strips (10 Â 1 mm) were cut from eachaorta
beginning from the end most proximal to the heart.
Vascular strips were then tied with surgical thread (6-0)
and suspended in a jacketed tissue bath(15 mL) con-
taining aerated physiological salt solution (PSS) at
35 ꢀC. Strips were secured at one end to a force dis-
placement (FT 0.3, Grass) transducer for monitoring
changes in isometric contraction. Aortic strips were
subjected to a resting force of 1 g and washed every
20 min withfreshPSS for 1 hafter the equilibration
period; guinea-pig aortic strips have been contracted by
washing in PSS containing 80 mM KCl (equimolar sub-
stitution of K+ for Na+). After the contraction reached
a plateau (about 45 min) the compounds (0.1, 0.5, 1, 5,
10, 50 and 100 mM) were added cumulatively to the bath
allowing for any relaxation to obtain an equilibrated
level of force. Addition of the drug vehicle had no
appreciable effect on K+-induced contraction (DMSO
for all compounds).
Acknowledgements
This research was supported by grants from P.I.N. of
M.I.U.R. and the University of Bologna.
References and Notes
1. Hieble, J. P.; Bondinell, W. E.; Ruffolo, R. R. J. Med.
Chem. 1995, 38, 3415.
2. Ruffolo, R. R.; Bondinell, W. E.; Hieble, J. P. J. Med.
Chem. 1995, 38, 3681.
3. Uchida, W.; Shibasaki, K.; Asano, M.; Takenaka, T.
J. Cardiovasc. Pharmacol. 1993, 21, 701.
4. Shibasaki, K.; Uchida, W.; Shirai, Y.; Inagaki, O.; Asano,
M.; Takenaka, T. Arch. Int. Pharmacodyn. 1994, 328, 213.
5. Hieble, J. P. In ꢀ-Adrenoceptors: Molecular Biology, Bio-
chemistry andPharmacology; Ruffolo, R. R., Jr. Ed.; Karger:
Basel, 1991; Vol. 7, p 105.
6. Allan, G.; Cambridge, D.; Hardy, G. W.; Follenfant, M. J.
Br. J. Pharmacol. 1987, 90, 609.
ꢀ-Adrenergic activities
7. Argentieri, T. M.; Troy, H. H.; Carrol, M. S.; Doroshuk,
C. M.; Sullivan, M. E. J. Cardiovasc. Pharmacol. 1993, 21,
647.
8. Kobinger, W.; Lillie, C. Eur. J. Pharmacol. 1984, 104, 9.
9. Kobinger, W.; Lillie, C. Eur. Heart J. 1987, 8 (suppl. L), 7.
10. Franke, H.; Su, C. A.; Schumacher, K.; Seiberling, M.
Eur. Heart J. 1987, 8 (suppl. L), 91.
Right atria. These were suspended in PSS (see above) at
the proper temperature (35 ꢀC) and a pH 7.4 and aera-
ted with95% O –5% CO2 gas. Benextramine (10À6 M)
2
was added to bathsolution to block a-adrenoceptors.
The tissue was mounted under a 0.5 g of tension in a
15 mL organ bath. Following a 45 min period during the
PSS was changed every 15 min, a cumulative dose–
response curve for the agonist INA was performed. In
preliminary experiments, reproducibility of the dose–
response curve obtained by the first to the third trials
in the absence of b-antagonists was confirmed. Fol-
lowing incubation withteh antagonist for 30 min, a
new dose–response curve to the agonist was obtained
either in the presence or after removal of the antagonist.
11. Glasser, S. P.; Michie, D. D.; Thadani, U.; Baiker, W. M.
Am. J. Cardiol. 1997, 79, 1401.
12. Kruger, C.; Landerer, V.; Zugck, C.; Ehmke, H.; Kubler,
W.; Haass, M. Cardiovasc. Res. 2000, 45, 900.
13. Bom, A.; Booth, S.; Bruin, J.; Clark, J.; Miller, S.;
Wathey, B. Bioorg. Med. Chem. Lett. 2001, 11, 2351.
14. Reiffen, M.; Eberlein, W.; Muller, P.; Psiorz, M.; Noll, K.;
Heider, J.; Lillie, C.; Kobinger, W.; Luger, P. J. Med. Chem.
1990, 33, 1496.

A. Bisi et al. / Bioorg. Med. Chem. 11 (2003) 1353–1361
1361
15. Budriesi, R.; Rampa, A.; Bisi, A.; Fabbri, G.; Chiarini,
A.; Valenti, P. Arzneim.-Forsch./Drug Res. 1995, 45, 234.
16. Bomhard, A.; Reiffen, M.; Heider, J.; Psiorz, M.; Lillie, C.
J. Med. Chem. 1991, 34, 942.
17. (a) Large, M. S.; Smith, H. J. Med. Chem. 1982, 25,
1286. (b) Large, M. S.; Smith, H. J. Med. Chem. 1982, 25,
1417.
18. Large, M. S.; Smith, H. J. Med. Chem. 1983, 26, 352.
19. Pallavicini, M.; Valoti, E.; Villa, L.; Piccolo, O. Tetra-
hedron Asymmetry 1994, 5, 5.
Calculations with Computer Programs, 2nd ed.; Springer-Ver-
lag: New York, 1992.
21. Hosohata, Y.; Hattori, K.; Shen, Y.; Okuyama, M.;
Kaneko, H.; Ohnuki, T.; Suzuki, J.; Nagatomo, T. Pharmacol.
1998, 57, 180.
22. Chem. Ber. 1891, 24, 2146.
23. Brenans, P. Bull. Soc. Chim. Fr. 1913, 4, 13 533.
24. Boyd, V.; Knowlton, M. J. Chem. Soc. 1909, 95, 1805.
25. Beasley, Y. M.; Petrow, V.; Stephenson, O. J. Pharm.
Pharmacol. 1958, 10, 47.
20. Tallarida, R. J.; Murray, R. B. Manual of Pharmacologic
26. Marle, M. J. Chem. Soc. 1912, 101, 309.