Synthesis, pharmacological characterization, and quantitative structure - Activity relationship analyses of 3,7,9,9-tetraalkylbispidines: Derivatives with specific bradycardic activity

Article abstract of DOI:10.1021/jm970120q

A series of 3,7,9,9-tetraalkyl-3,7-diazabicyclo[3.3.1]nonane derivatives (bispidines) was synthesized and identified as potential antiischemic agents. Pharmacological experiments in vitro as well as in vivo are described, and the results are listed. For s

Full text of DOI:10.1021/jm970120q

318
J . Med. Chem. 1998, 41, 318-331
Syn th esis, P h a r m a cologica l Ch a r a cter iza tion , a n d Qu a n tita tive
Str u ctu r e-Activity Rela tion sh ip An a lyses of 3,7,9,9-Tetr a a lk ylbisp id in es:
Der iva tives w ith Sp ecific Br a d yca r d ic Activity†
Uwe Scho¨n,* J ochen Antel, Reinhard Bru¨ckner, and J osef Messinger
Departments of Medicinal Chemistry and Pharmacology, Solvay Pharmaceuticals, P.O. Box 220, 30002 Hannover, Germany
Rainer Franke and Andreas Gruska
Consulting in Drug Design, Gartenstrasse 14, 16352 Basdorf, Germany
Received February 28, 1997
A series of 3,7,9,9-tetraalkyl-3,7-diazabicyclo[3.3.1]nonane derivatives (bispidines) was syn-
thesized and identified as potential antiischemic agents. Pharmacological experiments in vitro
as well as in vivo are described, and the results are listed. For selection of those compounds
fitting best to the desired profile of a specific bradycardic antianginal agentsdecrease in heart
rate without affecting contractility and blood pressuresthese results were scored and ranked.
Quantitative structure-activity relationship (QSAR) analyses were performed and discussed
a posteriori by means of Hansch, nonelementary discriminant and factor analysis to get insight
into the molecular features determining the biological profile. Highly significant equations
were obtained, indicating hydrophobic and steric effects. Both pharmacological ranking and
QSAR considerations showed compound 6 as the optimum within the structural class under
investigation. Compound 6 (tedisamil, KC8857) has been selected as the most promising
compound and was chosen for further pharmacological and clinical investigations as an
antiischemic drug.
Sch em e 1
1. In tr od u ction
Myocardial ischemia has been defined as an imbal-
ance between the supply of oxygenated blood and the
oxygen requirements of the myocardium. Major fea-
tures of ischemia may include reduction in contractile
function with a fall in blood pressure, ST-segment
displacement, arrhythmias and the occurrence of ana-
erobic glycolysis, as evidenced by accumulation of its
metabolic products. Research at Solvay Pharmaceuti-
cals in the area of ischemic heart disease has been
directed toward the development of compounds which
improve the oxygen supply/demand ratio by selectively
decreasing heart rate via a direct action on the sinus
node. Compounds of this kind are being referred to as
bradycardic antianginals or specific bradycardic agents,
according to the definition of Kobinger.1
The skeleton of 3,7-diazabicyclo[3.3.1]nonane (bispi-
dine2) constitutes the B and C rings of a variety of
lupanine alkaloids, e.g., sparteine (Scheme 1), which is
a tetracyclic alkaloid with antiarrhythmic properties.3
Due to this structural relationship, compounds belong-
ing to the ring system of bispidines have been the
subject of considerable interest. For instance anti-
arrhythmic properties have been demonstrated for
bispidine derivatives.4 Studies about crystal structures,
stereochemistry, and conformational analysis have been
published.5
synthesize 3,7-diazabicyclo[3.3.1]nonane derivatives with
a new cardiovascular profile. Within this field, we have
focused our interest on the synthesis of 3,7,9,9-tetra-
alkylbispidine derivatives6 (see Table 1 listing the
synthesized compounds and Scheme 2 for general
structures). The synthesis, pharmacological character-
ization, and quantitative structure-activity relationship
(QSAR) analyses of 3,7,9,9-tetraalkyl-3,7-diazabicyclo-
[3.3.1]nonane derivatives will be described, leading to
6 (tedisamil, KC8857) as the most promising compound
selected for clinical development.
2. Ch em istr y
Depending upon the substitution pattern of bispidine
derivatives, several synthetic approaches are possible.
For instance, using the Mannich reaction7 or starting
from pyridine derivatives8 are known approaches. Our
preferred sequence of reaction steps for the synthesis
of 3,7,9,9-tetraalkylbispidines is summarized in Scheme
2. Condensation of alkylideneacetic acid esters A, which
were synthesized by a Knoevenagel condensation, with
unsubstituted (R ) H) or substituted (R ) R1) cyano-
acetic acid amides B yields unsubstituted or substituted
dinitriles C.9 Unsubstituted dinitriles C were also
obtained by reaction of cyanoacetic esters with ketones
in alcohol in the presence of ammonia.10 Subsequently
As part of our research program related to the
synthesis of specific bradycardic agents, we decided to
* To whom correspondence should be addressed.
† Dedicated to Professor E. Winterfeldt on the occasion of his 65th
birthday.
S0022-2623(97)00120-9 CCC: $15.00 © 1998 American Chemical Society
Published on Web 01/06/1998

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 319
Ta ble 1. Synthesized 3,7,9,9-Tetraalkyl-3,7-diazabicyclo[3.3.1]nonanes of Formula G
R1
N
N
R2
R3
R4
compd
R1
R2
R3
R4
bp (0.1 Torr), °C
% yield base
formulaf
mp (salt), °C
analysisg
1
2a
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26b
n-C4H9
CH3
n-C4H9
CH3
CHM
CH3
CH3
CH3
CH3
CH3
CH3
130
69
54
75
83
77
79
76
65
58
40
75
87
51
37
78
74
66
33
83
60
24
46
41
82
49
39
C17H34N2‚1Sa
C11H22N2‚1Ox
C24H44N2‚1Ox
C18H36N2‚1Ox
C23H44N2
C19H32N2‚2HCl
C23H46N2‚1Ox
C17H34N2‚1Ox
C19H38N2
113
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N,Cl
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
150-160/11
fp: 36
180-200
250
230
210
140-150
150-160
230
fp: 102
130
150
100-120
160
160-170
150
190-200
170
188-190
157-158
173-178
CHMc
n-C4H9
n-C6H13
CPMd
C2H5
C2H5
n-C4H9
n-C6H13
CPM
n-C4H9
n-C3H7
i-C3H7
n-C6H13
CHM
n-C6H13
i-C3H7
i-C3H7
CHM
n-C4H9
i-C4H9
CHM
n-C4H9
n-C10H24
allyl
butenyl
butenyl
i-C4H9
n-C4H9
n-C6H13
-(CH2)4-e
-(CH2)4-
n-C4H9
C2H5
n-C3H7
C2H5
195-197
167-171
185
n-C4H9
n-C3H7
i-C3H7
n-C6H13
CHM
n-C4H9
C2H5
n-C3H7
n-C4H9
C25H50N2‚1Ox
C26H40N2
170-172
-(CH2)5-
CH3
C2H5
CH3
CH3
CH3
CH3
C2H5
n-C3H7
CH3
CH3
CH3
C18H36N2‚1Ox
C17H34N2‚1Ox
C15H30N2‚1.5Ox
C20H38N2
C17H34N2‚1Ox
C19H36N2‚1.5Fu
C25H48N2
C18H34N2‚1Ox
C29H58N2‚1Ox
C15H26N2‚1Sa
C20H34N2‚1Sa
C17H30N2‚1Sa
C21H40N2‚1Sa
C17H32N2‚1Sa
C14H28N2‚1Fu
201-203
70
56-60
i-C3H7
i-C3H7
n-C4H9
i-C4H9
i-C3H7
i-C3H7
n-C4H9
n-C10H24
allyl
butenyl
butenyl
butenyl
butenyl
CH3
120-125
104-107
-(CH2)5-
n-C4H9
-(CH2)3-
CH3
CH3
-(CH2)5-
CH3
n-C3H7
CH3
H
n-C4H9
190-195
155
118-120
115-119
85-89
139-143
162-163
142
CH3
CH3
230
160
175
150
170
165
195
CH3
n-C3H7
CH3
H
a
b
Welner, S.; Ginsburg, D. Synthesis of Potential Analgesics. Isr. J . Chem. 1966, 4, 39-45. Binnig, F.; Raschak, M.; Treiber, H. J .
Neue Antiarrhythmika. DT 248792, 1976. c CHM ) cyclohexylmethyl. CPM ) cyclopropylmethyl. e Indication for spirocyclic ring. f Sa
d
g
) salicylic acid; Ox ) oxalic acid; Fu ) fumaric acid. Compounds gave satisfactory analyses within (0.4% of theoretical calculations
unless otherwise stated.
Sch em e 2
ethoxy)aluminum hydride in order to avoid reduction
of double bonds.
3. P h a r m a cology
Bradycardic antianginal agents are beneficial in
cardiac ischemia due to r ed u ction in oxygen d em a n d
by selectively decreasing heart rate via a direct action
on the heart’s pacemaker, the sinus node. The relative
prolongation of the diastole caused by the reduction of
the heart rate leads to an in cr ea sed oxygen su p p ly.
As a main objective, bradycardic antianginal agents
should exert selective bradycardic activity without af-
fecting contractility and blood pressure. Additional
effects on refractoriness (antiarrhythmic activity) may
be of benefit.
Pharmacological investigations were performed in two
in vitro models, spontaneously beating right and electri-
cally paced left atria both isolated from guinea pigs, and
one in vivo model, anesthetized rats (see the Experi-
mental Section). Selection of the compounds according
to the objectives given above was done by scoring
potency and selectivity parameters and ranking these
scores. The parameters which were determined from
these in vitro and in vivo models (Table 2) are as follows:
Sin u s Ra te (SR). The rate of a spontaneously
beating right atrium reflects the activity of the sinus
node and provides information on direct effects on
automaticity for the compound tested.
the 2,4,6,8-tetraoxo-3,7-diazabicyclo[3.3.1]nonane de-
rivatives of formula D and E were prepared by cycliza-
tion of C under strong acidic conditions.11 The mono-
alkylation of the diimides of formula E or dialkylation
of D is carried out under basic conditions, for example
with sodium hydride or potassium carbonate in di-
methylformamide. Finally the reduction of the substi-
tuted tetraoxo derivatives of formula F yields the
bispidines of formula G which could be isolated as free
base or crystalline salts (see Table 1). Diimides of
formula F were reduced with lithium aluminum hydride.
In the case of an unsaturated side chain, reduction had
to be performed by means of sodium bis(2-methoxy-

320 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Scho¨n et al.
Ta ble 2. Parameters Measured, Selectivity Quotients Defined,
and Effective Concentrations/Doses Calculated in the
Pharmacological Investigationsa
Ta ble 3. Score Points Assigned to Parameters and Selectivity
Quotients of the Pharmacological Investigations in Vitro and in
Vivo
measured parameter and
effective
In Vitro Experiments (Guinea Pig Atria)
defined selectivity quotient
concentration/dose
abbreva
selectivity versus
undesired effect
force of contraction
F75/SR75 (SelF)
selectivity versus
additional effect funct
refractory period
potency desired
effect sinus rate
SR75
sinus rate
SR
F
FRP
F/SR
FRP/SR
HR
EC75
EC75
EC125
SR75
F75
FRP125
SelF
SelFRP
HR75
DAP125
RRT125
SelDAP
SelRRT
force of contract.
funct ref period
select. quot F
select. quot FRP
heart rate
FRP125/SR75 (SelFRP)
pts range (µmol/L)
pts
range
>20
10-20
5-10
3-5
2-3
1-2
<1
pts
range
<0.1
0.1-0.2
0.2-0.5
0.5-2
2-4
6
5
4
3
2
1
0
<1
0
-1
-2
-3
-4
-5
-6
6
5
4
3
2
1
0
ED75
1-2.5
2.5-5
5-10
10-20
>20
diast art. press.
RRT-interval
select. quot DAP
select. quot RRT
DAP
RRT
DAP/HR
RRT/HR
ED125
ED125
4-10
>10
a
Abbreviations used in the text.
not reached
In Vivo Experiments (Anesthetized Rat)
F or ce of Con tr a ction (F ). The peak force of con-
traction of a stimulated left atrium provides information
on a direct effect on contractility in cardiac muscle.
F u n ction a l Refr a ctor y P er iod (F RP ). The func-
tional refractory period is the shortest interval between
a double stimulus causing a double contraction. The
FRP provides information on effects on refractoriness,
the ability of re-excitation.
selectivity versus un-
desired effect diastolic
arterial pressure
selectivity versus
additional effect
RRT-interval
potency desired
effect heart rate
HR75
DAP125/HR75 (SelDAP) RRT125/HR75 (SelRRT)
pts range (µmol/L)
pts
range
pts
range
6
5
4
3
2
1
0
<5
5-7.5
7.5-12.5
12.5-25
25-40
>40
0
-1
-2
-3
-4
-5
-6
>10
4-10
2-4
0.5-2
0.2-0.5
0.1-0.2
6
5
4
3
2
1
0
<0.1
0.1-0.2
0.2-0.5
0.5-2
2-4
4-10
>10
Hea r t Ra te (HR). The heart rate provides informa-
tion on in vivo bradycardic activity of the compound
tested.
not reached
<0.1
Dia stolic Ar ter ia l P r essu r e (DAP ). The diastolic
arterial blood pressure provides information on the
peripheral resistance of the cardiovascular system and
selectivity regarding bradycardia.
RrT-In ter va l.12-14 The RRT-interval of the rat ECG
reflects the time from ventricular excitation to recovery;
effects on refractoriness in vivo can be judged from the
RRT-interval.
Furthermore the results of the in vitro models (guinea
pig atria) and the in vivo model (anesthetized rat) were
scored separately but according to a similiar procedure
(see Table 3):
F ir st Scor e (Desir ed Effects). The EC75 values for
the desired effects, i.e., decrease in sinus rate or heart
rate, were divided into seven groups and scores of 0 to
6 points assigned to these groups (Table 3, first column).
Secon d Scor e (Un d esir ed Effects). The selectivity
quotients for the ratio of desired versus undesired
effects, i.e., force of contraction or diastolic arterial
pressure, were also divided into seven groups and scores
of 0 to -6 points assigned to the so-formed groups (Table
3, second column).
Th ir d Scor e (Ad d ition a l Effects). The selectivity
quotients for the ratio of desired versus additional
effects, i.e., refractory period or RRT-interval, were
divided into seven groups as well and scores of 0 to 6
points assigned to the so-formed groups (Table 3, third
column).
group of bispidines or if an untested combination of
substituents would improve the desired pharmacological
profile.
QSAR analyses were performed for all biological
activities mentioned above by means of Hansch15 or
nonelementary discriminant16 analysis. Factor analy-
sis17 was used to investigate the structure of biological
data. Physicochemical properties of the compounds or
substituents, respectively, are described by log P, Taft’s
Es, Verloop’s STERIMOL parameters, molar refractiv-
ity, Rekker’s f values, aliphatic σ values.18,19 Taft’s Es
and molar refractivity both characterize substituent
bulk while the STERIMOL parameters according to
Verloop measure the dimension of substituents in
different directions of space; in the present case the best
results were obtained with molar refractivity as steric
parameter. Rekker’s f values as used here represent
substituent hydrophobicity, and log P measures the
hydrophobicity of the whole molecule. Aliphatic σ
values express the electron-attracting power of substit-
uents. In addition, the following indicator variables
were used: I1 (saturated and unbranched substituents
in R2 equal to or greater than n-butyl), I2 (unsaturation
in the R1 substituent), I3 (branched substituent in R2),
I4 (occurrence of a ring in R3 and R4), I5 (branch in R1
and/or R2), I6 (branching or ring structures in R1 and/
or R2). All parameters significantly appearing in the
resulting QSARs are summarized in Table 4. As the
molecules are symmetrical with respect to R3 and R4,
R1 and R2 were so defined that R2 is always the larger
substituent. In case where the substituents R3 and R4
are examined together, the abbreviation R34 is used. The
correlations between parameters are shown in Table 5.
Table 5 shows high correlations between f values and
molar refractivities for all positions of substitution as
is to be expected for the type of substituents varied. As
4. QSAR An a lyses
A QSAR analysis was performed a posteriori to obtain
a statistical rating of the compounds, to unveil previ-
ously hidden structure activity relations of the different
substituent patterns, and to get an insight into the
molecular features determining the biological profile. In
fact, we wanted to check if the synthesized compounds
included the structural optimum within the investigated

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 321
Ta ble 4. Parameters Used in Hansch and Discriminant Analysesa
compd
log P
MR(R1)
f(R1)
MR(R2)
f(R2)
MR(R34
)
f(R34
)
I1
I2
I3
I4
I5
I6
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
3.88
1.61
6.65
4.37
5.47
4.19
7.03
3.85
4.51
8.26
7.43
4.88
3.53
2.48
4.98
3.70
4.29
7.61
4.24
9.64
2.70
4.49
3.13
5.32
3.42
3.18
19.61
5.65
2.51
0.77
3.67
2.51
3.71
2.05
2.51
1.97
1.84
3.71
3.67
0.77
1.84
1.84
2.51
2.63
1.84
1.84
2.51
5.87
1.08
1.62
1.62
1.62
1.62
0.77
19.61
5.65
2.51
0.77
3.67
2.51
3.71
2.05
2.51
1.97
1.84
3.71
3.67
3.71
1.84
1.84
3.67
2.51
2.63
3.67
2.51
5.87
1.08
1.62
1.62
2.63
2.51
3.71
11.30
11.30
16.15
16.15
18.59
18.59
39.22
21.00
29.92
30.11
24.25
21.00
20.61
11.30
11.30
11.30
24.25
39.22
13.94
11.30
11.30
24.25
11.30
29.92
11.30
2.06
1.54
1.54
2.20
2.20
2.40
2.40
5.02
2.86
3.94
3.94
2.94
2.86
2.74
1.54
1.54
1.54
2.94
5.02
1.86
1.54
1.54
2.94
1.54
3.94
1.54
0.46
1
0
0
1
1
0
1
0
0
1
0
1
0
0
0
1
0
0
1
1
0
0
0
0
1
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
0
0
0
0
1
0
0
0
1
1
0
0
1
0
0
0
0
0
0
1
0
0
0
0
0
0
1
1
0
0
0
0
1
0
0
0
0
0
1
0
1
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
1
0
1
1
1
0
0
0
0
0
1
0
0
0
0
1
0
0
1
0
0
1
0
1
0
1
1
1
1
1
1
0
0
0
0
0
1
0
0
31.34
19.61
28.90
18.18
19.61
14.96
14.96
28.90
31.34
5.65
14.96
14.96
19.61
19.59
14.96
14.96
19.61
47.50
14.49
19.09
19.09
19.09
19.09
5.65
31.34
19.61
28.90
18.18
19.61
14.96
14.96
28.90
31.34
28.90
14.96
14.96
31.34
19.61
19.59
31.34
19.61
47.50
14.49
19.09
19.09
19.59
19.61
28.90
a
For the definition of indicator variables, see text.
Ta ble 5. Correlation Matrixa
log P
MR(R1) MR(R2) MR(R34
)
f(R1)
f(R2)
f(R34
)
I1
I2
I3
I4
I5
I6
log P
1.000
0.743
0.803
0.526
0.781
0.793
0.522
0.222
MR(R1)
MR(R2)
MR(R34
f(R1)
1.000
0.665
0.057
0.959
0.620
0.041
0.195
1.000
0.024
0.698
0.967
0.021
0.406
-0.235
-0.327
0.027
)
1.000
0.086
0.045
0.997
-0.247
-0.068
0.233
0.107
0.329
0.250
1.000
0.694
0.075
0.278
f(R2)
1.000
0.049
0.501 -0.229
f(R34
I1
)
1.000
1.000
I2
I3
I4
I5
-0.256 -0.062
-0.316 -0.370 -0.079 -0.256
1.000
0.010
-0.201 -0.196
-0.192 -0.250
0.163 -0.004
-0.232 -0.160
0.234 -0.452
1.000
0.063
0.172
0.048 -0.141 -0.036 -0.036
1.000
0.804 -0.127 1.000
0.570 0.085 0.709 1.000
-0.100 -0.2146 -0.226
0.077 0.016 0.011
0.331 -0.388 -0.076
0.238 -0.480 -0.220
I6
-0.002
0.019
a
Bold figures are significant at P ) 0.95.
a consequence it is difficult to differentiate between
steric and hydrophobic effects even though some of the
resulting equations allow at least a suggestion as to
which of the two may be the more important. Multiple
correlations exist between log P and the molar refrac-
tivities over all positions of substitution which must be
kept in mind when looking at the results:
reason two relationships between log P and molar
refractivity are presented is that both of them are
needed to estimate the values of log(1/HR75) and
log(1/SR75), respectively, for hypothetical compounds
which are optimal with respect to one of these activities
(see Section 6, Discussion).
5. Resu lts
log P ) 0.084((0.021)MR(R1) + 0.111((0.015)
(MR(R2) + MR(R34)) - 1.399((0.577) (1)
The results of the in vitro investigations are sum-
marized in Table 6. This table contains EC values of
the measured parameters (SR75, F75, FRP125) as well
as selectivity quotients (SelF ) F75/SR75, SelFRP )
FRP125/SR75). In addition, score points are listed for
the decrease of the sinus rate and the selectivity versus
the decrease in force of contraction. The sum of these
two scores is listed in column 9, reflecting the com-
pound’s suitability as bradycardic agent. Finally, in
Table 6 the score points for the selectivity versus the
additional effect on functional refractory period are
listed. Table 6, which ranks the compounds according
to their total score, shows nine compounds with high
scores for a selective decrease in sinus rate, effecting
the force of contraction at higher concentrations only,
n ) 26
r ) 0.980
r2 ) 0.960
F ) 274.4
s ) 0.386
P ) 100%
log P ) + 0.100((0.011)(MR(R1) + MR(R2)) +
0.102((0.019)MR(R4) - 1.30((0.604) (2)
n ) 26
r ) 0.977
r2 ) 0.954
F ) 241.0
s ) 0.410
P ) 100%
An explanation for this behavior is that MR describes
the polarizability as well as the size of the molecule,
which coincides with the number of CH2 groups. The

322 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Scho¨n et al.
Ta ble 6. Results of the in Vitro Investigations (Guinea Pig Atria)a
parameters measured
select. quot
SelFRP
scores
SR75
2
F75
3
FRP125
4
SelF
5
Sc1
7
Sc2
8
SS
9
Sc3
10
compd
6
22
24
19
15
17
23
25
06
13
09
01
16
05
08
21
10
18
07
14
04
03
11
26
02
12
20
0.40
0.44
0.52
0.88
1.00
1.21
1.21
1.30
1.99
2.18
3.00
3.07
0.87
3.53
8.55
0.60
1.90
1.91
10.9
3.21
6.20
7.57
-
75.1
23.1
56.1
16.3
35.5
1.14
0.73
0.63
0.48
1.54
1.87
2.08
0.76
2.63
0.80
1.00
0.82
0.44
1.52
4.30
1.27
1.45
1.01
111
167
52.5
108
18.5
35.5
2.85
1.66
1.21
0.55
1.54
1.54
1.72
0.56
1.32
0.37
0.33
0.27
0.51
0.43
0.50
2.12
0.76
0.53
6
6
6
6
5
5
5
5
5
5
4
4
6
3
3
6
5
5
2
4
3
3
0
0
0
0
0
-1
0
0
0
0
0
-1
0
0
-3
0
0
-4
-5
-5
-2
-6
-6
-6
-5
-5
6
6
6
5
5
5
5
5
5
4
4
4
3
3
3
2
0
2
3
3
3
3
3
3
3
3
4
4
4
3
4
4
2
3
3
0
5
5
6
6
3
>215
106
49.0
211
39.6
104
-
3.07
-
>215
>177
87.6
37.6
106
18.2
35.0
#
3.53
#
>25
1.39
3.65
2.14
2.32
1.92
1.12
#
0.47
0.76
0.35
0
0
-
10.2
1.52
4.69
2.65
0.54
0.68
0.62
1.40
122
0.17
0.11
0.08
-2
-3
-3
-5
-5
83.8
-
17.5
∼1
# #
-
∼1
***
***
∼1
***
***
***
***
0.84
85
105
a
The compounds are ranked according to the total score (see below) for desired and undesired effects. Column 1 shows the number of
the compound tested. Columns 2-4 show the EC values of the parameters measured, columns 5-6 the values of the selectivity quotients
calculated. Score points are listed for the desired effect, SR75 (Sc1, column 7) and for the selectivity versus the undesired effect, F75 (Sc2,
column 8). The sum of these two is listed in column 9 (SS), the main column, reflecting the compound’s value in fulfilling the objectives
mentioned above. Score points for the selectivity versus the additional effect FRP are listed in column 10. Except for compound 6, all EC
values (columns 2-4) are geometric means of two or three measurements, given in µM. For compound 6, geometric means with 95%
confidence limits calculated from nine measurements are given below. SR75, 1.30 (0.77-2.19) µM; F75, 49.0 (41-58) µM; FRP125, 0.76
(0.62-0.92) µM. -: effective concentration not reached. ***: parameter not measured. #: very high (estimated). # #: very low (estimated).
thus fulfilling the requirements for bradycardic agents.
On the other hand, at the bottom of Table 6 compounds
can be found with low scores for such a selective
decrease in sinus rate. Such compounds, poorly suited
as bradycardic agents, show pronounced selectivity for
prolongation of the functional refractory period.
(Table 6, column 2), eq 3 is obtained (see Figure 1):
log(1/SR75) ) -0.0039((0.0015)(MR(R1) +
MR(R2))2 + 0.36((0.10)(MR(R1) + MR(R2)) -
0.0034((0.0019)(MR(R34))2 +
The results of the in vivo investigations with anes-
thetized rats are summarized in Table 7. ED values of
the measured parameters (HR75, DAP125, RRT125) are
listed as well as selectivity quotients (SelDAP ) DAP125/
HR75, SelRRT ) RRT125/HR75). Furthermore score
points are given for the desired effect on heart rate and
for the selectivity against the undesired effect on
diastolic blood pressure. The sum of these scores is
listed in column 9, reflecting the compound’s suitability
as bradycardic agent in fulfilling the main in vivo
objectives. The score points for the selectivity with
regard to the additional effect (RRT-interval) are listed
in column 10. In ranking the total scores, seven of 25
compounds can be selected, showing high scores for the
decrease in heart rate without increase in blood pres-
sure resulting in total scores of 4 and 5 (upper group).
In contrast to the in vitro evaluation in guinea pig
tissue, all compounds of the selected set show high
scores for the prolongation of refractoriness (RRT-
interval). It should be noted that in intact animals the
decrease in heart rate contributes to the prolongation
of RRT, whereas in isolated tissue the rate of stimula-
tion is kept constant.
0.17((0.07)MR(R34) - 9.61((2.56) (3)
n ) 21
r ) 0.893
r2 ) 0.797
F ) 15.73
s ) 0.191
P ) 100%
In computing eq 3, compound 4 was omitted as an
outlier (with 4 included, r ) 0.828 and s ) 0.234).
Equation 3 shows an acceptable fit (see Figure 1)
“explaining” 80% of the data variance. MR(R34) and
(MR(R34))2 in this equation can be replaced by f(R34) and
f(R34),2 respectively, without changing anything else (r
) 0.891, s ) 0.193, F ) 15.41) while an analogous
exchange regarding R1 and R2 is not possible. This may
be interpreted to mean that steric interactions operate
in positions R1 and R2 while, for position R34, no
conclusion is possible whether substituent effects are
steric or hydrophobic. Optima exist for all positions
such that [(MR(R1) + MR(R2)] should be about 46 and
MR(R34) about 25 [or f(R34) about 3.3, respectively] for
compounds with high potency. Closest to these values
are compounds 22 and 24 which, in fact, show the
highest potencies in this in vitro test. Equation 3 does
not allow differentiation between bulk in positions R1
and R2. As will be shown in a forthcoming publication,
this equation has true predictive power. To gain more
5.1. QSAR for th e Sin u s Ra te. For the decrease
in sinus rate of the isolated guinea pig right atria (SR75)

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 323
Ta ble 7. Results of the in Vivo Investigations (Anesthetized Rat)a
parameters measured
select. quot
scores
HR75
2
DAP125
3
RRT125
SelDAP
5
SelRRT
Sc1
7
Sc2
8
SS
9
Sc3
10
compd
4
6
06
13
17
16
15
23
03
18
24
12
09
22
11
08
07
25
19
04
05
01
26
14
10
02
20
21
5.30
4.50
5.70
5.70
5.70
7.10
11.0
9.30
3.50
25.0
6.40
4.90
20.0
12.0
8.70
5.90
4.70
4.60
14.0
5.30
46.0
52.0
-
-
>36
>41
>41
>41
>46
-
2.50
0.53
0.73
0.73
0.96
0.94
5.5
1.10
0.32
3.00
1.80
0.58
12.0
2.10
3.20
0.70
0.55
0.61
5.60
0.57
8.40
18.0
4.00
***
#
0.47
0.12
0.13
0.13
0.17
0.13
0.50
0.12
0.09
0.12
0.28
0.12
0.6
0.18
0.37
0.12
0.12
0.13
0.40
0.11
0.18
0.35
# #
5
6
5
5
5
5
4
4
6
3
5
6
3
4
4
5
6
6
3
5
1
1
0
0
-1
-1
-1
-1
-1
0
-1
-3
-1
-3
-4
-2
-3
-3
-4
-5
-5
-4
-6
-3
-3
-3
5
5
4
4
4
4
4
3
3
2
2
2
1
1
1
1
1
4
5
5
5
5
5
3
5
6
5
4
5
3
5
4
5
5
5
4
5
5
4
6
>8
>7
>7
>7
>6
#
>66
2.40
>106
11.0
1.50
>61
10.00
5.10
2.20
0.60
0.57
2.80
0.20
-
>7
0.68
>4
1.72
0.31
>3
0.80
0.58
0.37
0.13
0.12
0.20
0.04
∼1
1
-1
-1
-2
-2
-3
-
∼1
-
∼1
***
***
***
***
***
***
***
***
a
The compounds are ranked according to the total score (see below) for desired and undesired effects. Column 1 shows the number of
the compound tested. Columns 2-4 show the ED values of the parameters measured, columns 5-6 the values of the selectivity quotients
calculated. Score points are listed for the desired effect, HR75 (Sc1, column 7) and for the selectivity versus the undesired effect, DAP125
(Sc2, column 8). The sum of these two is listed in column 9 (SS), the main column, reflecting the compound’s value in fulfilling the
objectives mentioned above. Score points for the selectivity versus the additional effect RRT-interval are listed in column 10. Except for
compound 6, all ED values (columns 2-4) are geometric means of two or three measurements, given in µmol/kg. For compound 6, geometric
means with 95% confidence limits calculated from 10 measurements are given below. HR75, 5.30 (4.68-6.00) µmol/kg; RRT125, 2.50
(2.04-3.08) µmol/kg. -: effective dose not reached. ***: parameter not measured. #: very high (estimated). # #: very low (estimated).
following relationship could be obtained:
log(1/F75) ) -0.14((0.07)(log P)2 +
1.93((0.87) log P - 7.15((2.49) (4)
n ) 20
r ) 0.802
r2 ) 0.643
F ) 15.29
s ) 0.428
P ) 100%
Compound 4 again appears as a severe outlier. If this
compound is removed, the fit is greately improved:
log(1/F75) ) -0.14((0.05)(log P)2 +
1.90((0.66) log P - 7.20((1.90) (5)
n ) 19
r ) 0.883
r2 ) 0.780
F ) 28.30
s ) 0.322
P ) 100%
Equation 5 describes a parabolic relationship with the
maximum at log P ≈ 7. The presence of the (log P)2
term is mainly due to only one point (compound 20) as
can be seen from a plot of log(1/F75) against log P
presented in Figure 2 (note also the deviating behavior
of compound 4). As this makes the parabola uncertain,
compound 20 was also removed, leading to eq 6:
F igu r e 1. Plot of observed versus predicted values of log(1/
SR75) (eq 3).
insight into the relative importance of bulk in the
different positions, synthesis and testing of some ad-
ditional compounds is under way. So far, measured
values of log(1/SR75) agree very well with estimates
from eq 3.
5.2. QSAR for th e F or ce of Con tr a ction . Brady-
cardic antianginal agents should exert their bradycardic
activity without affecting contractility. For the influ-
ence on force of contraction (Table 6, column 3), the
log(1/F75) ) 0.41((0.08) log P - 3.42((0.44) (6)
n ) 18
r ) 0.937
r2 ) 0.877
F ) 114.4
s ) 0.240
P ) 100%
Because of eq 1, log P in eq 6 can be replaced by molar
refractivity terms; introducing two additional indicator

324 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Scho¨n et al.
F igu r e 4. Plot of observed versus predicted values of log(1/
FRP125) (eq 8).
F igu r e 2. Plot of log(1/F75) against log P.
(Table 6, column 4) was investigated. As can be seen
in eq 8, the most important property is once again the
hydrophobicity term log P (see Figure 4):
log(1/FRP125) ) -0.15((0.02)(log P)2 +
1.69((0.26) log P - 0.41((0.25)I3 - 4.34((0.66) (8)
n ) 26
r ) 0.948
r2 ) 0.899
F ) 64.85
s ) 0.220
P ) 100%
Equation 8 provides a very good fit “explaining” 90% of
the data variance T; if the small cluster of points in the
low potency region is left out from the analysis, the
model remains stable. Potency passes through an
optimum with respect to log P ≈ 5.6 and is decreased
by branched substituents in R2 (I3 ) 1). Replacement
of log P by MR terms in eq 8 leads to a severe loss of
significance. Because of this and because very similar
relationships with log P were observed for other classes
of compounds possessing the bispidine structure,20 it can
be concluded that eq 8 in fact reflects hydrophobic
interactions and not a hidden steric effect. Concerning
the effects on FRP compound 6 with a log P value of
4.2 is about one log P unit away from the optimum.
However as the parabola described by eq 8 has a wide
shape, 6 still belongs to the group of the more active
compounds.
F igu r e 3. Plot of observed versus predicted values of log(1/
F75) (eq 7).
variables will then lead to an improvement (see Figure
3):
log(1/F75) ) 0.033((0.016)MR(R1) +
0.062((0.020)MR(R2) + 0.041((0.011)MR(R34) -
0.23((0.21)I6 - 0.31((0.29)I2 - 4.10((0.55) (7)
5.4. QSAR for Hea r t Ra te. As can be seen in eq 9,
hydrophobicity plays an important role for the decrease
in heart rate (Table 7, column 2; see Figure 5):
n ) 18
r ) 0.969
r2 ) 0.938
F ) 36.54
s ) 0.165
P ) 100%
log(1/HR75) ) -0.08((0.04)(log P)2 +
0.83((0.40) log P - 0.0035((0.0014)MR(R1)2 +
0.14((0.05)MR(R1) - 4.27((1.04) (9)
I6 takes a value of 1 if branching or ring structures occur
in R1 or R2, and I2 indicates unsaturation in the R1
substituent. At what point the linear equations (6 and
7) turn into parabolic relationships cannot be answered
with the data at hand unless the large impact of
analogue 20 is taken for granted. There is no way to
decide whether the causal relationship for F75 is with
log P (eqs 5 and 6) or with substituent size (eq 7).
n ) 22
r ) 0.888
r2 ) 0.789
F ) 15.86
s ) 0.150
P ) 100%
Equation 9 could again be improved in two steps: (1)
removing 14 as an outlier (see Figure 5); (2) introduction
of an indicator variable I6, which takes a value of 1 if
5.3. QSAR for F u n ction a l Refr a ctor y P er iod . As
an additional effect, the prolongation of refractory period

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 325
It is tempting to interpret the parabolic relationship
with log P in eq 10 as transport effect. One must,
however, be aware of the relationship between the molar
refractivity of substituents and log P (eq 1). The steric
effect of R1 seems to be outstanding and of high
importance.
5.5. QSAR for th e Dia stolic Ar ter ia l P r essu r e.
In addition to a decrease in contractility, an increase of
the diastolic blood pressure (DAP125) also is an un-
desired side effect. The data for increase in diastolic
blood pressure (Table 7, column 3) do not admit a
Hansch analysis to be performed. However, the com-
pounds can be divided into two classes according to their
potency: class 1 (active class), DAP125 e 11; class 2
(inactive class), DAP125 > 11.
The distribution of compounds can now be investi-
gated by discriminant analysis leading to the following
discriminant function
w ) -1.64 log P + 0.18MR(R1) +
2.02f(R34) + 2.29I1 (11)
F igu r e 5. Plot of observed versus predicted values of log(1/
HR75) (eq 9).
with the class means
w (mean, class 1) ) 3.13
w (mean, class 2) ) 1.16
The indicator variable I1 takes a value of 1 for
saturated and unbranched substituents in R2 equal to
or greater than n-butyl. The two classes are fairly well
distinguished from each other by the discriminant
function (11) (all terms significant at P > 98%) which
correctly reclassifies 9 out of the 10 compounds in class
1 (90%) and 11 out of the 13 compounds in class 2
(84.6%) (total: 20 out of 23 compounds ) 87%).
A
comparison of the results from eq 11 with the class
mean values shows that the chances of a compound to
be classified as inactive with respect to an increase of
diastolic blood pressure increase with overall hydropho-
bicity. In addition the hydrophobicity of R34 and the
size of R1 should not become too large. Equation 11 is
difficult to interpret but seems to reflect some complex
interplay between steric and hydrophobic properties of
substituents. The important point is that 6 falls into
the property space for inactivity with respect to the
undesired increase of diastolic blood pressure.
F igu r e 6. Plot of observed versus predicted values of log(1/
HR75) (eq 10).
branching or ring structures occur in R1 or R2 (see
Figure 6):
5.6. QSAR for th e RrT-In ter va l. Focusing on in
vivo effects on refractoriness (RRT-interval, Table 7,
column 4), eq 12 is obtained after removing compound
14 which again behaves as an outlier (see Figure 7):
log(1/HR75) ) -0.035((0.031)(log P)2 +
0.35((0.33) log P - 0.0035((0.0011)MR(R1)2 +
0.14((0.03)MR(R1) + 0.11((0.11)I6 - 3.03((0.86)
(10)
log(1/RRT125) ) -0.0034((0.0014)(MR(R1) +
MR(R2))2 + 0.30((0.12)(MR(R1) + MR(R2)) -
n ) 21
r ) 0.945
r2 ) 0.892
F ) 24.84
s ) 0.093
P ) 100%
0.0023((0.0012)MR(R34)2 +
0.096((0.053)MR(R34) + 0.39((0.28)I5 -
0.08((0.10)f(R2)*I4 - 7.39((2.93) (12)
Equation 10 shows a good fit and “explains” 90% of the
data variance. Optima exist with respect to hydropho-
bicity and the size of substituents in R1. Strongly active
compounds should have log P values of about 5 and
MR(R1) values close to about 20. In addition branching
or ring structures in R1 and/or R2 exert a slight activity
enhancing effect (I6). With respect to all these proper-
ties, 6 is optimal with MR(R1) ) 18.18, log P ) 4.2 and
I6 ) 1.
n ) 22
r ) 0.901
r2 ) 0.812
F ) 10.79
s ) 0.194
P ) 99.99%
Equation 12 shows an only moderate but yet acceptable
fit considering the type of data involved “explaining”
81% of data variance. As was the case with eq 3, molar

326 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Scho¨n et al.
F igu r e 8. Plot of factor weights after VARIMAX rotation.
F igu r e 7. Plot of observed versus predicted values of log(1/
RRT125) (eq 12).
and compounds selectively prolonging the refractory
period are found at the end (low scores in column 9 and
high scores in column 10).
refractivity in eq 12 can be replaced by f values for R34
(r ) 0.902, s ) 0.194, F ) 10.90) but not for R1 and R2.
This may again be taken as an indication that steric
effects operate in R1 and R2 while no preference between
hydrophobic and steric effects can be made with respect
to R34. Thus, eq 12 indicates an optimum of substituent
bulk for positions R1 and R2 and an optimum in bulk
or/and hydrophobicity for R34. In eq 12 a positive effect
of branching in R1 and R2 is reflected by I5 [I5 ) 0 (no
branch) and I5 ) 1 (branch in R1 and/or R2)]. The cross
product term in eq 12, though statistically significant,
is of doubtful validity and difficult to interpret. I4
indicates the occurrence of a ring in R34 and the cross
product seems to indicate that, if such a ring is present,
a slight decrease of potency appears with hydrophobicity
of R2. If this term is removed from eq 12, nothing else
changes; only the statistics become, of course, a little
poorer with r ) 0.836, s ) 0.246, and F ) 10.75. With
respect to the increase of the RRT-interval, 6 is placed
among compounds with high potency in the anesthe-
tized rat model.
In vivo the parameter of refractoriness, the RRT-
interval, is located in the same area as the heart rate.
Again this agrees with the in vivo score board (Table
7), where the compounds representing bradycardic
agents without increase in diastolic blood pressure
additionally show prolongation of refractoriness. The
diastolic blood pressure seems to be close to the heart
rate and RRT-interval, so that selectivity might not
follow any simple pattern of properties. However, the
communality of DAP125 is very low (0.5), so that this
variable is only poorly represented in common factor
space and its position in Figure 8 does permit only
limited conclusions.
For a QSAR analysis selectivity indices were defined
as follows:
log SelF ) log(F75/SR75)
log SelDAP) log(DAP125/HR75)
5.7. QSAR for Selectivities. To investigate the
internal relationship of the different tests, a factor
analysis was performed resulting in three significant
factors with eigenvalues λ1 ) 2.89, λ2 ) 1.13, and λ3 )
0.51 and extracted variances of 63.7%, 25.0%, and
11.3%, respectively. As the first two factors account for
89% of the variance, it is possible to plot the data in
the space spanned by these factors as shown in Figure
8 (VARIMAX rotation). Tests characterizing the heart’s
automaticity, sinus rate, and heart rate, are clustered
together, which points to a common mechanism of
action. For the in vitro model the parameters of
refractoriness, FRP125, and of force of contraction, F75,
are clearly separated from each other as well as from
the sinus rate. This indicates that systematic differ-
ences between these parameters exist and that it should
be possible to design selective compounds with respect
to these properties. This finding is in full accordance
with the in vitro score board (Table 6) where selective
bradycardic compounds without decrease in contractility
are found at the top of the list (high scores in column 9)
As to be expected, when eqs 3 and 7 are compared, a
relationship with substituent molar refractivity exists
for log SelF (see Figure 9):
log SelF ) -0.010((0.007)MR(R1)2 +
0.37((0.33)MR(R1) - 0.0043((0.002)MR(R34)2 +
0.17((0.10) MR(R34) + 0.43((0.40)I2 -
3.03((4.28) (13)
n ) 16
r ) 0.975
r2 ) 0.950
F ) 38.30
s ) 0.194
P ) 100%
Again, molar refractivity for R34 can be replaced by the
corresponding f values (r ) 0.972, s ) 0.203, F ) 35.02)
while this is not possible for position R1. This result
once more may be taken to be indicative of steric
interactions in this position. Equation 13 provides a
good fit accounting for 95% of the data variance.
Selectivity passes through a maximum with respect to

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 327
bicyclo[3.3.1]nonanes, 6 (tedisamil) in fact proved to be
one of the most active derivatives selected for further
in depth pharmacological evaluation. For the investi-
gated potencies and selectivities, the results from QSAR
yield a consistent picture with an overall satisfactory
goodness of fit. In a few equations the number of terms
is a little high, compared with the number of measure-
ments, but the ratios are still acceptable, and in all cases
simpler equations with less terms also exist. As ex-
pected for the type of structural variation involved,
pharmacological properties primarily depend on sub-
stituent size and/or hydrophobicity with some additional
influence by branching and/or unsaturation while elec-
tronic effects could not be detected.
The QSARs obtained allow the potencies to be esti-
mated from physicochemical parameters of substituents.
Unfortunately, it is not possible to ascertain the role of
log P or to relate log P terms with certain processes such
as transport or membrane interactions because of the
relationships between log P and MRx (see eqs 1 and 2).
The only exception is eq 8 where all evidence points to
hydrophobic interactions. To clarify this situation, it
would be necessary to break the correlations between
log P and MR which could have been done only by
synthesizing and testing additional compounds. How-
ever, the objective of the present investigation was only
to prove whether 6 was the optimal choice out of the
available variations.
F igu r e 9. Plot of observed versus predicted values of log SelF
(eq 13).
the size of substituents in R1 (at MR(R1) ≈ 18) and with
respect to size (at MR(R34) ≈ 20) or hydrophobicity in
R34 (at f(R34) ≈ 2.6), respectively, and is enhanced by
unsaturated substituents in R1 (I2 ) 1). The optimal
size of substituents in R34 is close to the corresponding
value found for SR75 (see eq 3), and the optimum for
R1 is compatible with the optimum found in this
equation for the sum of molar refractivities in R1 and
R2. Those compounds with high potency will also have
a high selectivity potential. 6 has ideal selectivity
properties with MR(R1) ) 18.2 and MR(R34) ) 18.6 or
f(R34) ) 2.40, respectively. The molar refractivity terms
in eq 13 can be replaced by a simple log P term with
only a moderate loss in the goodness of fit:
To answer the question whether there might be more
active derivatives than 6, the highest possible value of
log(1/SR75) and log(1/HR75) were computed from eqs 3
and 10. These values refer to hypothetical compounds
having optimal values of MR and log P. These com-
pounds will be referred to in the following as SR75opt
and HR75opt, respectively. For SR75opt substituent
size was defined according to the maxima of the MR
terms in eq 3 (MR(R1) + MR(R2) ) 46, and MR(R34) )
25). Analogously, log P ) 5 and MR(R1) ) 20 were
defined for HR75opt from eq 10 which then leads to
(MR(R2) + MR(R34)) ) 43 because of eq 1; this sum was
divided into MR(R2) ) 18 and MR(R34) ) 25 (optimal
value for MR(R34)). To be able to make an estimate of
the value of log(1/HR75) for the hypothetical compound
SR75opt, molar refractivities were derived from eq 2.
The highest possible values estimated in this way are
in comparison with the values of 6:
log SelF ) -0.44((0.15) log P +
0.56((0.57)I2 + 3.37((0.85) (14)
n ) 16
r ) 0.906
r2 ) 0.821
F ) 29.89
s ) 0.368
P ) 100%
Because of the relationships between log P and MR, it
is again not possible to arrive at a clear decision whether
hydrophobicity (transport?) or steric factors are of
primary importance. For log SelDAP no meaning-
ful relationship with either Hansch or discriminant
analysis could be obtained. This is not surprising in
the light of the results from factor analysis discussed
above.
log(1/HR75)
log(1/SR75)
HR75opt
SR75opt
6 (calcd)
6 (obsd)
-0.64
-0.68
-0.68
-0.72
0.67
1.04
0.44
-0.11
6. Discu ssion
There is one compound in the series, 24, which has a
value of log(1/HR75) ) -0.54 which is slightly higher
than that of HR75opt; the calculated value is -0.63, and
the difference is well within the error of biological
measurement. As was already pointed out, 6 is for all
practical purposes optimal with respect to the decrease
of heart rate in vivo while for the decrease of sinus rate
in vitro somewhat more active compounds are possible.
If now log(1/F75) and log SelF are computed from eqs 6
and 14 for HR75opt and SR75opt and the resulting
Summarizing the scoring of test results from the in
vitro models (guinea pig atria) and in vivo model
(anesthetized rat), five representatives (6, 13, 15, 17,
and 23) of compounds under investigation show high
scores for the bradycardic activity and high selectivity
with regard to force of contraction and diastolic arterial
pressure. Furthermore, these five compounds addition-
ally have prolonging effects on refractoriness. With
respect to the performed in vitro/in vivo tests, within
the series of synthesized 3,7,9,9-tetraalkyl-3,7-diaza-

328 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Ta ble 8. Synthesized Dicyanoglutarimides of Formula C (R ) H)
Scho¨n et al.
R4
R3
NC
CN
O
O
N
R
compd
R3
R4
% yield
mp, °C
lit. mp, °C
formula
analysisg
remarks
C1a
C2b
C3c
C4
CH3
C2H5
n-C3H7
n-C4H9
CH3
C2H5
n-C3H7
n-C4H9
41
78
81
90
46
75
84
55
33
54
220-221
225
216-217
200
220-225
C9H9N3O2
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C11H13N3O2
C13H17N3O2
C15H21N3O2
C10H9N3O2
C11H11N3O2
C12H13N3O2
C10H11N3O2
C11H13N3O2
C13H17N3O2
230-235
201-205
198-200
182-183
215-222
180
‚0.3H2O
C5
-(CH2)3-
-(CH2)4-
-(CH2)5-
C6a
C7d
C8e
C9f
C10
179-180
211-212
193
CH3
CH3
C2H5
C2H5
n-C3H7
n-C4H9
190-192
221-222
201-202
a
b
Guareshi, I.; Grande, E. Synthese von Glutar- und Trimethylenderivaten. Chem. Zbl. 1899 (II), 439-440. Peano, E. Einige Derivate
des Dia¨thylketons. γ,γ-Dia¨thyl-â-â′-dicyan-R-R′-dioxypyridin. Chem. Zbl. 1901 (I), 582. c Guareshi, I. Synthese von Pyridin- und
d
Trimethylenpyrrolverbindungen. Chem. Zbl. 1901 (I), 577-579. Guareshi, I. Einige neue Derivate der Cyclohexanone. Chem. Zbl. 1911
(II), 361-362. e Guareshi, I.; Grande, E. U¨ ber ein Hydroa¨thyldicyanmethyldioxypyridin. Chem. Zbl. 1898 (II), 544-545. f Minozzi, A.
g
Sintesi di nuovi derivati glutarici e trimetilenici. Gazz. Chim. Ital. 1900, 30 (I), 265-278. Compounds gave satisfactory analyses within
(0.4% of theoretical calculations unless otherwise stated.
values compared with those for 6, the following picture
is revealed:
with ligation of the coronary artery23) and in vivo
(angina pectoris model with conscious dogs24). This
profile of activity could be confirmed in clinical inves-
tigations demonstrating the favorable hemodynamic
and antiischemic effects in patients with ischemic heart
disease.25,26
log(1/F75)
log SelF
HR75opt
SR75opt
6 (calcd)
6 (obsd)
-1.37
-0.96
-1.70
-1.62
1.17
0.73
1.53
1.58
Exp er im en ta l Section
Thus, with respect to the undesired effect on contrac-
tility, 6 is clearly better than HR75opt and SR75opt.
The already mentioned compound 24 is slightly better
than 6 with respect to log(1/HR75) and also with respect
to log(1/SR75) ()0.36) as well as log SelF ()1.72) and
about equiactive with respect to the prolongation of the
refractory period and could thus be regarded as a
possible alternative. However, 24 shows a higher
decrease of the force of contraction (log(1/F75) ) -1.36).
Therefore, if the whole profile is considered with a
higher weight attributed to the decrease of heart rate
in vivo compared to sinus rate in vitro, 6 clearly is the
best candidate within the space spanned by the type of
substituent variations investigated. In addition 6 does
not influence the diastolic arterial pressure. Also from
the QSAR point of view for the type of substituents used
and with a view to the spanned substituent space with
Gen er a l Meth od s. Melting points were measured in a
capillary melting point apparatus (HWS-SG 2000) and are
uncorrected. Nuclear magnetic resonance (13C NMR) spectra
were collected in the pulsed Fourier-transformation mode,
either on a Bruker HX90R or Bruker AM400 or Bruker
ARX500 NMR spectrometer and were consistent with the
proposed structures. Chemical shifts are referred to TMS.
Where elemental analyses are indicated only by symbols of
the elements, analytical results obtained for those elements
were within (0.4% of the theoretical values.
Deter m in a tion of log P HP LC Va lu es. The applied tech-
nique was analogous to a method described by Kraak.27
Bondapak Phenyl was used as stationary phase. The mobile
phase consisted of a methanol/phosphate-buffer (0.005 M; pH
4; 60/40 vol %) containing 0.1% sodium dodecyl sulfate and
0.1 M NaClO4. Na2Cr2O7 was used as a nonretardent com-
pound. The temperature was kept at 23.0 ( 0.5 °C. A UV
detector running at 205 nm was used for detection. log PHPLC
values were determined as described by Kraak with linear
regression via a calibration curve which was prepared from
28 compounds with known log Poctanol values.
regard to physiochemical variations, compound
6
(tedisamil) proved to be the optimal selection with
respect to activity and pharmacological profile.
Ch em istr y. Dinitriles of the formula C as starting materi-
als were synthesized according to a previously described
method.9 The compounds of the formulas D-G were synthe-
sized according to the following general procedures. The
physical properties of these derivatives and the starting
dinitriles of formula C are summarized in Tables 1 and 8-12.
Gen er a l Descr ip tion of th e P r oced u r e To Syn th esize
2,4,6,8-Tetr a oxo-3,7-d ia za bicyclo[3.3.1]n on a n es of F or -
m u la D a n d E. The dinitrile (20 g) of formula C (see Tables
8 and 9) were heated to between 120 and 140 °C while stirring
in 100 mL of an acidic medium, the composition of which is
given in Tables 10 and 11, until they are completely dissolved.
After 10-15 min, the entire mixture is poured into ice water.
The precipitated tetraoxo compound of formula D or E is
filtered off by suction and, if required, is recrystallized,
preferably from ethanol, and finally is dried.
Comparing bradycardic action with prolongation of
refractoriness in the in vitro and the in vivo models,
different selectivities were observed. Besides the dif-
ferences in heart rate control, this may be due to the
mechanism of action of these compounds, demonstrated
for 6 in voltage clamp experiments.21,22 It was shown
that 6 blocks a potassium outward current involved in
ventricular repolarization of the rats more pronouncedly
than in guinea pig hearts. So the differences between
in vitro and in vivo selectivity may be due to species
differences combined with the compound’s mechanism
of action. In further pharmacological investigations it
could be shown that 6 (tedisamil) possesses clear-cut
antiischemic properties in vitro (isolated rat heart model

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 329
Ta ble 9. Synthesized Dicyanoglutarimides of Formula C (R ) R1)
R4
R3
NC
CN
O
N
R
O
compd
R1
R3
R4
% yield
mp, °C
formula
analysisa
C11
C12
C13
C14
C15
n-C4H9
i-C4H9
CHM
n-C6H13
i-C4H9
CH3
CH3
56
55
60
68
66
85-87
C13H17N3O2
C17H25N3O2
C22H33N3O2
C17H25N3O2
C16H21N3O2
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
n-C3H7
n-C4H9
C2H5
n-C3H7
n-C4H9
C2H5
111
110
64
119-121
-(CH2)5-
a
Compounds gave satisfactory analyses within (0.4% of theoretical calculations unless otherwise stated.
Ta ble 10. Synthesized Unsubstituted 2,4,6,8-Tetraoxo-3,7-diazabicyclo[3.3.1]nonanes of Formula D
H
N
O
O
N
O
H
R3
O
R4
compd
R3
R4
used acid
60% H2SO4
60% H2SO4
60% H2SO4
60% H2SO4
60% H2SO4
60% H2SO4
60% H2SO4
% yield
mp, °C
lit. mp, °C
formula
analysisc
remarks
D1a
D2
D3
D4
D5
CH3
C2H5
n-C3H7
n-C4H9
CH3
C2H5
n-C3H7
n-C4H9
50
61
70
45
70
81
40
84
65
65
>350
230
239
>360
C9H10N2O4
C11H14N2O4
C13H18N2O4
C15H22N2O4
C10H10N2O4
C11H12N2O4
C12H14 N2O4
C10H12N2O4
C11H14 N2O4
C13H18N2O4
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
195-199
-(CH2)3-
-(CH2)4-
-(CH2)5-
>350
‚0.2H2O
‚0.2H2O
‚0.3H2O
D6
>350
D7a
D8a
D9b
D10
310
>370
329-331
278-280
CH3
CH3
C2H5
C2H5
n-C3H7
n-C4H9
H2SO4/H3PO4 (1:1)
H2SO4/H3PO4 (1:1)
70 H2SO4
324-326
291-295
174-177
a
Thorpe, J . F.; Wood, A. S. The Formation and Reaction of Imino-compounds. Part XVIII. The Condensation of Cyclohexanones with
b
Cyanoacetamide Involving the Displacement of an Alkyl Group. J . Chem. Soc. 1913, 1586. Reference 10: Handley, G. J .; Nelson, E. R.;
Samers, T. C. Compounds Derived from â-Substituted Glutaric Acids: Glutarimides, Glutaramic acids, 1,5-Pentane diols. Austr. J . Chem.
1960, 13, 129-144. c Compounds gave satisfactory analyses within (0.4% of theoretical calculations unless otherwise stated.
Ta ble 11. Synthesized Monoalkylated 2,4,6,8-Tetraoxo-3,7-diazabicyclo[3.3.1]nonanes of Formula E
R1
N
O
O
N
O
H
R3
R4
used acid
O
compd
R1
R3
R4
% yield
mp, °C
formula
analysisa
E1
E2
E3
E4
E5
n-C4H9
i-C4H9
CHM
n-C6H13
i-C4H9
CH3
CH3
H2SO4/H3PO4 (1:1)
H2SO4/H3PO4 (1:1)
60 % H2SO4
H2SO4/H3PO4 (1:1)
H2SO4/H3PO4 (1:1)
56
85
86
91
80
175-178
149-153
140
102-103
182
C13H18N2O4
C17H26N2O4
C22H34N2O4
C17H26N2O4
C16H22N2O4
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
n-C3H7
n-C4H9
C2H5
n-C3H7
n-C4H9
C2H5
-(CH2)5-
a
Compounds gave satisfactory analyses within (0.4% of theoretical calculations unless otherwise stated.
The compounds of formula D (derivatives D1-D10) and E
(derivatives E1-E5) which are obtained in this manner are
listed in Tables 10 and 11.
monoalkylation, a solution of 0.15 mol of the respective
alkylating agent in 50 mL of dimethylformamide is added
dropwise. In the case of dialkylation, 0.3 mol of alkylating
agent is used. The resulting mixture is refluxed for 3 h.
Thereafter most of the solvent is distilled off under vacuum.
Dichloromethane is added to the residue, and the mixture is
washed with a 20% sodium hydroxide solution. The aqueous
phase is again extracted with dichloromethane. The organic
phases are combined, washed several times with water, and
dried (MgSO4). After distilling off the solvent, the remain-
ing residue is recrystallized from a mixture of ether and
hexane.
Gen er a l Descr ip tion of th e P r oced u r e to Syn th esize
Alkylated 2,4,6,8-Tetr aoxo-3,7-diazabicyclo[3.3.1]n on an es
of F or m u la F . A mixture of 0.1 mol of the tetraoxo compound
of formula D or E in 200 mL of absolute dimethyl formamide
is heated to a temperature of between 60 °C and 70 °C. Then
0.125 mol (0.25 mol for dialkylation, respectively) of sodium
hydride (or potassium carbonate, see Table 12) are added
portion by portion, and the mixture is then boiled under reflux
for about 1 h. After cooling, a solution of the alkylating agent
in absolute dimethylformamide is added. In the case of

330 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3
Scho¨n et al.
Ta ble 12. Synthesized Dialkylated 2,4,6,8-Tetraoxo-3,7-diazabicyclo[3.3.1]nonanes of Formula F
R1
N
O
O
N
O
R2
R3
O
R4
compd
R1
R2
R3
R4
RX
basea
% yield
mp, °C
97
273-278
112-115
82
analysisb
remarks
F 1
F 2
F 3
F 4
F 5
F 6
F 7
F 8
n-C4H9
CH3
CHM
n-C4H9
CH3
CHM
CH3
CH3
CH3
CH3
CH3
CH3
C2H5
C2H5
Br
J
C
H
H
H
H
C
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
C
C
H
H
61
82
45
81
30
58
85
59
30
38
43
31
47
68
64
81
30
48
49
56
59
54
76
38
66
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
C,H,N
Br
Br
Br
Cl
Br
Br
Br
Br
Br
J
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
Br
n-C4H9
n-C6H13
CPM
n-C4H9
n-C3H7
i-C3H7
n-C6H13
CHM
n-C4H9
n-C6H13
CPM
n-C4H9
n-C3H7
i-C3H7
n-C6H13
CHM
n-C6H13
i-C3H7
i-C3H7
CHM
n-C4H9
i-C4H9
CHM
n-C4H9
n-C10H24
allyl
butenyl
butenyl
i-C4H9
n-C4H9
-(CH2)4-
70
-(CH2)4-
160-161
73-75
94
153-154
oil
140
100-101
103
118-120
100
80
101
85
110
63
128-130
120
n-C4H9
C2H5
n-C3H7
C2H5
n-C4H9
C2H5
n-C3H7
n-C4H9
F 9
F 10
F 11
F 12
F 13
F 14
F 15
F 16
F 17
F 18
F 19
F 20
F 21
F 22
F 23
F 24
F 25
-(CH2)5-
CH3
C2H5
CH3
CH3
CH3
CH3
C2H5
n-C3H7
CH3
CH3
CH3
‚0.2H2O
i-C3H7
i-C3H7
n-C4H9
i-C4H9
i-C3H7
i-C3H7
n-C4H9
n-C10H24
allyl
butenyl
butenyl
butenyl
butenyl
-(CH2)5-
n-C4H9
n-C4H9
‚0.2H2O
‚0.5H2O
-(CH2)3-
CH3
CH3
CH3
CH3
-(CH2)5-
CH3
n-C3H7
CH3
CH3
n-C3H7
CH3
91
oil
80-83
a
b
H ) sodium hydride. C ) potassium carbonate. Compounds gave satisfactory analyses within (0.4% of theoretical calculations
unless otherwise stated.
According to this procedure the compounds of formula F
(derivatives F 1-F 25), listed in Table 12, have been synthe-
sized.
rectangular pulses of 1 ms duration. Isometric contractions
were recorded by means of force transducers and the signals
fed into a computer for evaluation. For each compound,
cumulative concentration response curves (maximum of five
concentrations) were recorded. Effective concentrations (EC75
and EC125), defined as drug concentrations inducing a decrease/
increase of a parameter by 25% of the predrug value, were
calculated as geometric mean of the individual EC values.
An esth etized r a t. The animals (Wistar, male, 280-360
g) were anesthetized with urethane (1.25 g/kg ip) and provided
with an endotracheal tube to facilitate spontaneous respira-
tion. Arterial blood pressure was measured in the left femoral
artery via a saline-filled tube catheter connected to a Statham
pressure transducer. The electrocardiogram (ECG) was re-
corded with needle electrodes inserted subcutanously. The
signals were fed into a computer and analyzed every 30 s by
means of an evaluation software. The compounds were
administered by continuous infusion at a volume of 0.1 mL/
min using a tube catheter placed in the left femoral vein. At
10-min intervals, the drug concentration was increased by a
factor of 10 without changing the infusion volume (3 doses).
Effective doses (ED75 and ED125), defined as drug doses
inducing a decrease/increase of a parameter by 25 of the
predrug value, were calculated as geometric mean of the
individual ED values.
Gen er a l Descr ip tion of th e Red u ction of Alk yla ted
2,4,6,8-Tetr a oxo-3,7-d ia za bicyclo[3.3.1]n on a n es of F or -
m u la F to Bisp id in es of F or m u la G. A mixture of 0.1 mol
of lithium aluminum hydride in 100 mL of THF/toluene (70:
30) is heated at 80 °C. Subsequently, 0.025 mol of the tetraoxo
compounds of formula F in 100 mL of a mixture of THF/toluene
(70:30) is added dropwise at an oil bath temperature of 80 °C.
The reaction mixture is kept at a temperature of 120 °C for
2-4 h. The reaction mixture is then hydrolyzed under basic
conditions. The reaction mixture is then extracted with
dichloromethane, and the organic phase is dried (MgSO4). The
dried organic phase is concentrated by evaporation, and the
residue is subjected to fractional distillation under reduced
pressure in a system provided with a bulb-tube fractionating
column. Tetraoxo compounds of formula F with unsaturated
alkenyl side chains are reduced in an analogous manner as
described above by using sodium bis(2-methoxyethoxy)alumi-
num hydride (Red-Al, Aldrich) as reducing agent in toluene.
By proceeding in this manner, the bispidine derivatives of
formula G (compounds 1-25), listed in Table 1, have been
obtained. As indicated, the elementary analysis has been
performed with the free base, respectively, the mentioned salts.
P h a r m a cology. Isola ted Righ t a n d Left Gu in ea P ig
Atr ia . The animals (female, 250-300 g) were killed by a blow
on the head and bled from the carotid arteries. The hearts
were quickly excised, and the atria (right and left) were
dissected from the hearts in oxygenated buffer solution. The
atria were attached to preparation holders and suspended
individually in glass tissue chambers filled with buffer solu-
tion, continuously oxygenated and kept at 35 °C. The left
atrial preparations were additionally provided with bipolar
platinum stimulation electrodes. Right atria exert spontane-
ous activity; left atria were electrically paced at 2 Hz with
Ackn owledgm en t. The authors express their thanks
to J . Fasterding, J . Grosser, C. Ku¨ster, and A. Schulz
for the synthesis of the investigated substances. We are
obliged to Dr. U. Ku¨hl and Dr. G. Buschmann for
providing the biological data. The authors also thank
the members of the Physical Chemistry Departement
for spectral determinations (F. Pa¨tzold, H. Schulz-
Benkendorf, H. Vogel-Lahrmann), elemental analyses
(I. Gehrmann), and lipophilicity measurements (H.

Synthesis of 3,7,9,9-Tetraalkylbispidines
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 3 331
(11) McElvain, S. M.; Clemens, D. H. Piperidine Derivatives XXX.
1,4-Dialkylarylpiperidines. J . Am. Chem. Soc. 1958, 80, 3915-
3923.
Schu¨nhof). E. Finner, Dr. W. Kehrbach, Dr. L. Ku¨ster,
Dr. H. Zeugner, Dr. D. Thorma¨hlen, and Dr. D. Ziegler
are acknowledged for constructive discussion, and Dr.
v. Usslar is acknowledged for her review of the manu-
script.
(12) The rat ECG differs from other species, particularly in heart
rate and configuration. The onset of ventricular depolarization
shows no Q deflection, and the T-wave declines slowly. So in
the rat ECG, the duration of ventricular depolarisation is
described by the RRT-interval, beginning at the onset of R,
ending at the apex of T. As the QT in other species, in the rat
RRT-interval estimates the duration of ventricular depolarisa-
tion.
(13) Budden, R.; Buschmann, G.; Ku¨hl, U. G. The Rat Electrocar-
diogram in Acute Pharmacology and Toxicology. In The Rat
Electrocardiogram in Pharmacology and Toxicology; Budden, R.,
Detweiler, D. K., Zbinden, G., Eds.; Pergamon Press: Oxford,
1981; pp 41-81.
(14) Danesi, R.; del Tacca, M.; Soldani, G. Measurement of the S
alpha T segment as the most reliable ECG parameter for the
assessment of adriamycin.induced cardiotoxicity in the rat. J .
Pharmacol. Meth. 1985, 16, 251-259.
(15) Multiple regression analysis from STATGRAPHICS version 5.0
(see ref 17).
Su p p or tin g In for m a tion Ava ila ble: Tables containing
13C NMR spectral data (4 pages) and elemental analyses (14
pages). Ordering information is given on any current masthead
page.
Refer en ces
(1) Kobinger, W. Specific Bradycardic Agents, A New Approach to
Therapy in Angina Pectoris? Prog. Pharmacol. 1985, 5/ 4, 89-
100.
(2) Mannich, C.; Mohs, P. U¨ ber Derivate eines aus zwei Piperidin-
ringen kondensierte bicyclischen Systems. (Derivatives of a
Condensed Ring System of Two Piperidine Rings.) Chem. Ber.
1930, 63, 506-512.
(3) (a) Bo¨hme, H. Die Behandlung von Herzrhythmussto¨rungen mit
Sparteinsulfat. (Treatment of Cardiac Arrhythmias with Sparteine
Sulfate.) Z. Gesamte Inn. Med. 1964, 19, 883-889. (b) Raschack,
M. Wirkungen von Spartein und Sparteinderivaten auf Herz und
Kreislauf. (Effect of Sparteine and Sparteine Derivatives on the
Cardiovascular System.) Arzneim.-Forschung 1974, 24, 753-759.
(c) Wick, E.; Bell, U.; Dengler, H. J . Die Wirkung von Sparte-
insulfat auf mechanisch ausgelo¨ste Extrasystolen beim Men-
schen. (Effect of Sparteine Sulfate on Mechanically Produced
Extrasystole in Humans.) Med. Welt 1971, 22, 303-304.
(4) (a) For instance: Smith, G. S.; Thompson, M. D.; Berlin, K. D.;
Holt, E. M.; Scherlag, B. J .; Petterson, E.; Lazarra, R. A Study
on the Synthesis and Antiarrhythmic Properties of Selected 3,7-
Diheterobicyclo(3.3.1)nonanes with Substituents at the 2,4-
Positions and at the 9-Position. Eur. J . Med. Chem. 1990, 25,
1-8. (b) Zhang, Z.-H.; Follmer, C. H.; Sarma, J . S. M.; Chen, F.;
Singh, B. N. Effect of Ambasilide, a New Class III Agent, on
Plateau Currents in Isolated Guinea Pig Ventricular Myocytes:
Block of Delayed Outward Potassium Current. J . Parmacol. Exp.
Ther. 1992, 263, 40-48.
(5) (a) Zefirov, N. S.; Palyulin, V. A. Conformational Analysis of
Bicyclo[3.3.1]nonanes and Their Hetero analogs. Top. Stereo-
chem. 1991, 20, 171-230. (b) J eyaraman, R.; Avila, S. Chemistry
of 3-Azabicyclo[3.3.1]nonanes, Chem. Rev. 1981, 81, 149-174.
(6) Scho¨n, U.; Hachmeister, B.; Kehrbach, W.; Ku¨hl, U. G.; Busch-
mann, G. Diazabicyclo[3.3.1]nonanes. German Patent DE 3 234
697, 1982; Chem. Abstr. 1984, 101, 55089v.
(7) (a) Blicke, F. F. The Mannich Reaction. Org. React. 1942, 1, 303-
341. (b) Mannich, C.; Veit, F. U¨ ber Derivate des Bispidins.
(Derivatives of Bispidine.) Chem. Ber. 1935, 68B, 506-512. (c)
Reichert, B. Die Mannich Reaktion (The Mannich Reaction);
Springer-Verlag: Berlin, 1959.
(16) (a) Dove, S. Program VADIA (b) For details of this method, see,
e.g., Franke, R. Theoretical Drug Design Methods; Elsevier:
Amsterdam, 1984; pp 222-241.
(17) STATGRAPHICS 5.0, STSC Inc., 2155 East J efferson St, Rock-
ville, MD 20582.
(18) Hansch, C.; Leo, A. Substituent Constants for Correlation
Analysis in Chemistry and Biology; J . Wiley & Sons: New York,
1979.
(19) Rekker, R. F.; Mannhold, R. Calculation of Drug Lipophilicity;
VCH: Weinheim, 1992.
(20) Scho¨n, U.; Antel, J .; Bru¨ckner, R.; Messinger, J .; Franke, R.;
Gruska, A. Unpublished results.
(21) Dukes, I. D.; Morad, M. Tedisamil Modulates Outward K-Chan-
nels in Rat and Guinea Pig Ventricular Myocytes. Circulation
1989, 80, Suppl. II, 517.
(22) Dukes, I. D.; Cleemann, L.; Morad, M. Tedisamil Blocks the
Transient and Delayed Rectifier K-Currents in Mammalian
Cardiac and Glial Cells. J . Pharmacol. Exp. Ther. 1990, 254,
560-569.
(23) Duchosal, F.: Opie, L. H. Potential Beneficial Effects of Potas-
sium Channel Blockade by Tedisamil in Isolated Perfused
Working Heart with Coronary Ligation. Cardiovasc. Drugs Ther.
1992, 6 (Suppl.1), 329-336.
(24) Grohs, J . G.; Fischer, G.; Raberger, G. Cardiac and Hemody-
namic Effects of the Selective Bradycardic Agent KC 8857
During Exercise-Induced Myocardial Ischemia. Eur. J . Phar-
macol. 1989, 161, 53-60.
(25) Mitrovic, V.; Oehm, E.; Liebrich, A.; Thormann, J .; Schlepper,
M. Hemodynamic and Antiischemic Effects of Tedisamil in
Humans. Cardiovasc. Drugs Ther. 1992, 6 (Suppl. 1), 353-360.
(26) Mitrovic, V.; Oehm, E.; Strasser, R.; Schlepper, M.; Pitschner,
H. F. Hemodynamic, Antiischemic and Neurohumoral Effects
of the New Potassium Channel Blocking Agent Tedisamil in
Patients with Coronary Artery Disease. Z. Kardiol. 1996, 85,
961-972.
(8) Bohlmann, F.; Ottawa, N.; Keller, R. Aufbau des Tetrahydro-
chinolizons und des “Bispidins”. (Contribution to the Synthesis
of Cytisine; Structure of Tetrahydroquinolizone and “Bispidine”.)
Annalen 1954, 587, 162-176.
(9) McElvain, S. M.; Clemens, D. H. â-Methyl-â-phenyl-R,R′-di-
cyanoglutarimide. Organic Syntheses; Wiley: New York, 1963;
Collect. Vol. IV, pp 62-663.
(10) Handley, G. J .; Nelson, E. R.; Somers, T. C. Compounds Derived
from â-Substituted Glutaric Acids: Glutarimides, Glutaramic
Acids, 1,5-Pentane diols. Austr. J . Chem. 1960, 13, 129-144.
(27) Kraak, J . C.; van Rooij, H. H.; Thus, J . L. G. Reversed-phase
Ion-pair Systems for the Prediction of n-Octanol-Water Parti-
tion Coefficients of Basic Compounds by High-performance
Liquid Chromatography. J . Chromatogr. 1986, 352, 455-463.
J M970120Q