17β-O-aminoalkyloximes of 5β-androstane-3β, 14β-diol with digitalis- like activity: Synthesis, cardiotonic activity, structure-activity relationships, and molecular modeling of the Na+,K+-ATPase receptor

Article abstract of DOI:10.1021/jm990627w

A series of digitalis-like compounds with a 17-aminoalkoxyiminoalkyl or -alkenyl substituent was synthesized and evaluated for inhibition of Na⁺,K⁺-ATPase and for inotropic activity. The highest inhibition was found with compounds having the substituent in configuration 17β and the amino group at a distance of 6 or 7 bonds from C(17) of the digitoxigenin skeleton. The presence of the oxime function strengthens the interaction with the receptor, more if α,β-unsaturated, thus mimicking the electronic situation of the unsaturated lactone in natural digitalis compounds. The most active compounds showed Na⁺,K⁺-ATPase inhibitory potencies (IC₅₀) 17-25 times higher than the standards digitoxigenin and digoxin and 3-11 times higher inotropic potencies (EC₅₀) in isolated guinea pig left atria. These features are supported by a molecular model suggesting the possible interactions of the groups described above with particular amino acid residues in the H1-H2 domains of Na⁺,K⁺-ATPase. Some interactions are the classical ones already described in the literature; a new, very strong interaction of the basic group with the Cys138 was found and adds new possibilities to design compounds interacting with this region of the receptor. The most interesting compounds were also studied in vivo in the anesthetized guinea pig for evaluating their inotropic effect versus the lethal dose. Compounds 9 and 12 showed a slightly higher safety ratio than digoxin and deserve further evaluation.

Full text of DOI:10.1021/jm990627w

2332
J . Med. Chem. 2000, 43, 2332-2349
17â-O-Am in oa lk yloxim es of 5â-An d r osta n e-3â,14â-d iol w ith Digita lis-lik e
Activity: Syn th esis, Ca r d ioton ic Activity, Str u ctu r e-Activity Rela tion sh ip s, a n d
Molecu la r Mod elin g of th e Na +,K+-ATP a se Recep tor
Alberto Cerri,*,† Nicoletta Almirante,† Paolo Barassi,‡ Alessandra Benicchio,† Giorgio Fedrizzi,† Patrizia Ferrari,‡
Rosella Micheletti,‡ Luisa Quadri,† Enzio Ragg,§ Roberto Rossi,‡ Marco Santagostino,† Antonio Schiavone,‡
Fulvio Serra,‡ Maria Pia Zappavigna,† and Piero Melloni†
Departments of Medicinal Chemistry and Cardiovascular Pharmacology, Prassis Istituto di Ricerche Sigma-Tau,
Via Forlanini 3, 20019 Settimo Milanese (MI), Italy, and Department of Agricultural and Food Molecular Science,
Chemistry Section, Faculty of Agriculture, Universita` degli Studi di Milano, Via Celoria 2, 20133 Milano, Italy
Received December 23, 1999
A series of digitalis-like compounds with a 17-aminoalkoxyiminoalkyl or -alkenyl substituent
was synthesized and evaluated for inhibition of Na+,K+-ATPase and for inotropic activity. The
highest inhibition was found with compounds having the substituent in configuration 17â and
the amino group at a distance of 6 or 7 bonds from C(17) of the digitoxigenin skeleton. The
presence of the oxime function strengthens the interaction with the receptor, more if
R,â-unsaturated, thus mimicking the electronic situation of the unsaturated lactone in natural
digitalis compounds. The most active compounds showed Na+,K+-ATPase inhibitory potencies
(IC50) 17-25 times higher than the standards digitoxigenin and digoxin and 3-11 times higher
inotropic potencies (EC50) in isolated guinea pig left atria. These features are supported by a
molecular model suggesting the possible interactions of the groups described above with
particular amino acid residues in the H1-H2 domains of Na+,K+-ATPase. Some interactions
are the classical ones already described in the literature; a new, very strong interaction of the
basic group with the Cys138 was found and adds new possibilities to design compounds
interacting with this region of the receptor. The most interesting compounds were also studied
in vivo in the anesthetized guinea pig for evaluating their inotropic effect versus the lethal
dose. Compounds 9 and 12 showed a slightly higher safety ratio than digoxin and deserve
further evaluation.
Ch a r t 1
In tr od u ction
Digitalis cardiac glycosides, such as digoxin and
digitoxin (Chart 1), are drugs clinically used for the
treatment of congestive heart failure (CHF).1 Although
these compounds give an unquestionable improvement
in the quality of life of patients suffering from CHF, a
major problem is their low therapeutic index due to
cardiac proarrhythmogenic activity. A recent NIH-
sponsored trial2 showed a neutral effect for digoxin on
mortality and renewed interest in the search for a safer
positive inotropic agent acting through the inhibition
of Na+,K+-ATPase.
In recent articles,3,4 our group reported a series of 17â-
guanylhydrazone derivatives of digitoxigenin skeleton
(Chart 1) acting as positive inotropic compounds through
the inhibition of Na+,K+-ATPase. Structure-activity
studies on these compounds indicated that the presence
of a basic (guanidino) group at a proper distance, of a
1,2-polarized iminic double bond, or of a 1,4-polarized
conjugate system, which could mimic that of the R,â-
unsaturated lactone of digitoxigenin, was essential for
* Corresponding author. Tel: 0039.0233500388. Fax: 0039.
0233500408. E-mail: pstchem@tin.it.
the above-mentioned activity. The importance of a basic
center and of a dipole was confirmed by the high binding
activity for the Na+,K+-ATPase receptor of a â-dimethyl-
aminoethyl R,â-unsaturated ester of non-digitalis tri-
cyclic perhydrophenanthrene structures, such as cas-
saine and seco-D 5â-androstane derivatives.5
† Department of Medicinal Chemistry, Prassis Istituto di Richerche
Sigma-Tau.
‡ Department of Cardiovascular Pharmacology, Prassis Istituto di
Ricerche Sigma-Tau.
§ Department of Agricultural and Food Molecular Science, Univer-
sita` degli Studi di Milano.
10.1021/jm990627w CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/17/2000

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2333
Ta ble 1. Structure, Physical Data, and Na+,K+-ATPase Inhibition for Compounds 1-42
method
of
Na+,K+-ATPase
%
inhibition,
d
compd
Ra
prepnb
mol formulac
20H33NO3‚H2O
C20H33NO3
mp, °C
yield
IC50
,
µM
1
2
3
4
5
6
7
8
(E) CHdN-OH
(Z) CHdN-OH
(E) CHdN-OCH3
A
A
A
A
A
A
A
A
A
A
B
B
B
B
B
B
C
229-231
248-250
172-177
125-135
82-90
46
18
40
72
78
57
100
>100e
C21H35NO3‚0.25H2O
32
1.0
0.25
2.0
f
(E) CHdN-O(CH2)2N(CH3)2
(E) CHdN-O(CH2)3N(CH3)2
(E) CHdN-O(CH2)4N(CH3)2
(E) CHdN-O(CH2)2NH2
C24H42N2O3‚C4H6O6 ‚H2O
g
C25H44N2O3‚C2H2O4
C26H46N2O3‚C2H2O4g‚0.5H2O
48-58
C22H38N2O3‚C4H4O4h‚EtOH‚0.5H2O
95-135 dec 57
0.16
0.02
0.50
g
(E) CHdN-O(CH2)3NH2
C23H40N2O3‚C2H2O4
110-120
161-170
60
64
g
9
(E) CHdN-O(CH2)4NH2
C24H42N2O3‚C2H2O4
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
(E) CHdN-OCH2COOH
C22H35NO5
215-220 dec 79
>100i
(E) CH2CHdN-O(CH2)2N(CH3)2
(E,Z) (CH2)2CHdN-O(CH2)2N(CH3)2
(E,Z) (CH2)2CHdN-O(CH2)2NH2
(E,E) CHdCHCHdN-OCH3
(E,E) CHdCHCHdN-O(CH2)2CH3
(E,E) CHdCHCHdN-O(CH2)2OH
(E,EZ) CHdCHCHdN-OCH2COOH
(E,E) CHdCHCHdN-O(CH2)2N(CH3)2
(E,E) CHdCHCHdN-O(CH2)2NH2
(E,E) CHdC(CH3)CHdN-OCH3
(E,E) CHdC(CH3)CHdN-O(CH2)2CH3
(E,E) CHdC(CH3)CHdN-O(CH2)2OH
(E,E) CHdC(CH3)CHdN-O(CH2)2N(CH3)2
(E,E) CHdC(CH3)CHdN-O(CH2)2NH2
(E,E) CHdC(CH3)CHdN-O(CH2)3NH2
(E,E) CHdC(CH3)CHdN-O(CH2)4NH2
(E,E,E) (CHdCH)2CHdN-O(CH2)2N(CH3)2
CH2NHOCH3
C25H44N2O3‚0.5H2O
C26H46N2O3‚C2H2O4g‚H2O
C24H42N2O3
C23H37NO3
C25H41NO3‚0.75H2O
C24H39NO4
89-91
18
33
67
59
34
48
60
48
21
59
66
65
73
70
72
48
61
32
58
63
64
0.63
0.10
0.08
4.0
125-130
165-170
176
127-128
189-191
185-188
130-133
162-164
145-146
154-155
182-185
198-203
140-150
128-136
199-204
137-151
143-145
182-185
202-205
193-198
10
6.3
>100j
0.03
0.02
10
>10k
7.9
C24H37NO5‚0.25H2O
C26H44N2O3.C2H2O4g‚1.25H2O
C24H40N2O3‚0.25H2O
C24H39NO3‚0.75H2O
C26H43NO3
C25H41NO4‚0.25H2O
C27H46N2O3‚C4H4O4h‚0.25H2O
C25H42N2O3‚C4H4O4h‚0.75H2O
C26H44N2O3‚C2H2O4g‚0.5H2O
C27H46N2O3.C2H2O4g‚0.25H2O
C28H46N2O3‚C2H2O4g‚1.25H2O
C21H37NO3‚C2H2O4g‚H2O
C24H44N2O3‚2HCl‚H2O
C22H40N2O3‚2HCl
C23H42N2O3‚2HCl‚0.25H2O
C25H46N2O3‚C2H2O4g‚1.5H2O
C23H41NO3‚HCl‚H2O
C24H44N2O3‚0.25H2O
C23H39NO3‚HCl‚0.25H2O
C25H41NO2‚0.5H2O
B
B
A
A
A
A
A
A
A
B
C
C
C
C
0.06
0.03
0.20
1.26
6.30
79
50
1.58
0.40
>100l
>100j
1.0
40
0.08
CH2NHO(CH2)2N(CH3)2
CH2NHO(CH2)2NH2
CH2NHO(CH2)3NH2
CH2N(CH3)O(CH2)2N(CH3)2
(CH2)3NHOCH3
(CH2)3NHO(CH2)2NH2
(E) CHdCHCH2NHOCH3
(E,E) (CHdCH)2(CH2)2NH2
(E,E) (CHdCH)2(CH2)3NH2
(CH2)7NH2
160-170 dec 30
C
C
C
149-151
145-146
60
58
35
172-179
171-198
C26H43NO2‚0.25H2O
C26H47NO2‚0.5H2O
167-173
0.32
1.25
25
40
0.25
1.58
0.50
0.50
58(68)-127
CHO
(CH2)2CHO
(E) CHdCHCHO
C
20H32O3‚H2Om
C22H36O3
C22H34O3
135-137
184-192
n
(E) CHdC(CH3)CHO
C23H36O3‚0.25H2O
digitoxigenin 2,5-dihydro-5-oxo-3-furyl
digoxin
a
b
The EZ mixture is indicated where the amount of Z oxime is more than 10%. General methods of preparation: (A) at pH 4.5; (B) in
the presence of NaOH; (C) NaBH3CN; where no indications, see experimental. c Analyses for C, H, N, Cl, and H2O are within 0.4% of the
theoretical values. Concentrations able to inhibit 50% of enzyme activity; mean of two or three experiments. e 30% inhibition at 100 µM.
d
f
g
h
i
j
k
l
L-(+)-tartrate. Oxalate. Fumarate. 0% inhibition at 100 µM. 40% inhibition at 100 µM. 40% inhibition at 10 µM. 20% inhibition
m
n
at 100 µM. Reference 7. Reference 4.
Intrigued by these observations and with the aim to
further explore the requirements for a strong interaction
with the Na+,K+-ATPase receptor and find novel ino-
tropic agents more active than digoxin, we synthesized
a new series of digitalis-like steroidal derivatives in
which the guanylhydrazone group was replaced by an
aminoalkyloxime group; the polarized iminic bond was
thus retained together with the basic group of cassaine
and seco-D 5â-androstane derivatives reported above.
The aminoalkyloxime substituent hit the target, as
some of the compounds of this new series showed the
highest inhibitory activities on Na+,K+-ATPase ever
found with semisynthetic analogues, higher than those
of the potent, natural digitalis compounds.
The importance of the iminic double bond was evalu-
ated both through its reduction to the corresponding
hydroxylamine function and substitution with a non-
polarized olefinic double bond. The contribution to the
activity of the amine group was also assessed either by
replacing it with other groups or by leaving it as the
unique functional group connected to the steroidal
skeleton by a saturated carbon chain.
From these new findings and the data reported in the
previous papers,3-5 we reasoned that this extraordinary
activity could be ascribed to an interaction of the
protonated amine substituent with a negatively charged
site in the receptor. With the aim of finding if this
putative “anionic site” really existed, we started a

2334 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
Ta ble 2. Structure, Physical Data, and Na+,K+-ATPase Inhibition for Compounds 43-60
method
of
Na+,K+-ATPase
%
yield
inhibition,
d
compd
Ra
prepnb
mol formulac
C23H40N2O3
C24H42N2O3
C25H44N2O3
C21H36N2O3‚0.1EtOH
C25H42N2O3
C26H44N2O3
mp, °C
IC50
,
µM
43
44
45
46
47
48
49
50
(E) NO(CH2)2N(CH3)2
(E) NO(CH2)3N(CH3)2
(E) NO(CH2)4N(CH3)2
(E) NO(CH2)2NH2
(E,EZ) CHCHdN-O(CH2)2N(CH3)2
(E,E) CHCHdN-O(CH2)3N(CH3)2
(E,EZ) CHCHdN-O(CH2)2NH2
(E,EZ) CHCHdN-O(CH2)3NH2
D
D
D
D
A
A
A
A
157-159
169-170
67-68
161-162
157-162
168-170
208-212 dec
150-153
54
60
45
40
61
42
40
55
>100e
>100e
>100e
>100e
12.5
160
C23H38N2O3‚0.5EtOH
C24H40N2O3‚0.25H2O
2.0
>100f
a
b
The EZ mixture is indicated where the amount of Z oxime is more than 10%. General methods of preparation: (A) at pH 4.5; (D)
at pH 2.0. c Analyses for C, H, N, and H2O are within 0.4% of the theoretical values. Concentrations able to inhibit 50% of enzyme
d
activity; mean of two or three experiments. e 0% inhibition at 100 µM. f 40% inhibition at 10 µM.
Sch em e 1a
isomers of the oximes were stable in any case, due to
the presence of the 21-CH3 group, and also the starting
aldehyde was more resistant to the epimerization at
C(17). Compounds 11-16, 18, 19, and 27 were synthe-
sized under basic conditions to limit the amount of Z
oxime; in these cases no epimerization at C(17) of the
corresponding starting aldehydes 54 (Scheme 2, in the
form of lactol, as shown by NMR), 40 and 414 (Scheme
3), and 584 (Scheme 4) occurred.
The oximes 43-46 (Table 2) were synthesized from
ketone 598 (Scheme 5) and the appropriate (O-substi-
tuted)hydroxylamines dihydrochlorides in ethanol at pH
2.
The oximes reported in Tables 1 and 2 were obtained
as almost pure E isomers (Z isomer E 10%) except for
compounds 12 (40% of Z oxime), 17 (25% of Z oxime),
47 (30% of Z oxime), and 49 and 50 (20% of Z oxime).
In D2O/DMSO-d6 solution at pH 7.4 (phosphate buffer)
and 37 °C, oximes 1 and 3-9 showed no E/Z isomer-
ization up to 24 h (1H NMR analysis), while oximes 11-
16 and 18-19 gave a 6/4 E/Z equilibrium mixture, after
24 h. At pH 1, oximes 1 and 3-9 gave an 8/2 E/Z
mixture after 6 h, while oximes 11-19 gave a 6/4 E/Z
equilibrium mixture after 3 h. Z oxime 2 reached the
equilibrium in a shorter time already at pH 7.4.
Compounds 20-26, with a methyl group at position 21,
were synthesized to stabilize the E isomer and no
isomerization was indeed observed for these compounds
at any pH.
a
Reagents and conditions: (a) H2NOR‚xHCl, NaOAc, HCl,
dioxane, H2O, pH 4.5; (b) NaBH3CN, MeOH, H2O, HCl, pH 3; (c)
CH2O, NaBH3CN, MeCN, H2O.
modeling program of docking the most active compounds
with a 3D-model of Na+,K+-ATPase receptor, built from
the peptidic sequence that many authors indicate as the
“receptor sequence”.6
Ch em istr y
The synthetic pathways for the compounds listed in
Tables 1 and 2 are reported in Schemes 1-7.
Hydroxylamines 28-31 (Scheme 1) and 33-35
(Scheme 3) were prepared from the corresponding
oximes by reduction with NaBH3CN at pH 3, while
N-methylhydroxylamine 32 (Scheme 1) was prepared
by reductive alkylation of 29 with CH2O/NaBH3CN.
The oximes 1-27 (Table 1) and 47-50 (Table 2) were
synthesized from the corresponding aldehydes and
appropriate (O-substituted)hydroxylamines hydrochlo-
rides or dihydrochlorides in a dioxane/water solution;
oximes 1-10 (Scheme 1), 17, 20-26 (Scheme 3), and
47-50 (Scheme 5) were prepared at pH 4.5 (method A),
while oximes 11 (Scheme 2), 12-16, 18 and 19 (Scheme
3), and 27 (Scheme 4) were prepared at pH >9 (method
B), in the presence of NaOH (see Experimental Section).
In the case of compounds 1-10 pH 4.5 was the best
compromise to limit the amount of Z isomer, relevant
at a lower pH, and to prevent the isomerization of the
starting 17â aldehyde 397 to the 17R isomer, occurring
at a higher pH. In the case of compounds 20-26 the E
All the attempts to prepare aldehyde 54 from 39 by
Wittig reactions with methoxymethylenetriphenylphos-
phorane or methylenetriphenylphosphorane gave mix-
tures of 17R- and 17â-vinyl epimers in very low yields,
the 17R being the undesired, prevalent, and inseparable
component. Thus, the homologation was carried out, as
shown in Scheme 2, by KF-catalyzed condensation of
39 with nitromethane to give 51 as exclusive isomer (see
Templeton et al. for the corresponding 3â-O-(R-L-rham-
nopyranosyl) derivative).9 Diacetylation of 3â- and 20-

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2335
Sch em e 2a
a
Reagents and conditions: (a) CH3NO2, KF, iPrOH; (b) Ac2O, DMAP, THF; (c) NaBH4, EtOH, Et2O; (d) DEAD, Ph3P, CH2Cl2; (e) 1 N
NaOH, MeOH, THF; (f) DIBAH, THF, 0 °C; (g) H2NO(CH2)2N(CH3)2‚2HCl, NaOH, dioxane, H2O.
Sch em e 3a
a
Reagents and conditions: (a) DIBAH, THF, -60 °C, 2 h; (b) MnO2, dioxane; (c) H2, Pd/C, EtOAc; (d) H2NOR‚xHCl, NaOH, dioxane,
H2O; (e) H2NOR‚xHCl, dioxane, H2O, pH 4.5 or >9; (f) NaBH3CN, MeOH, H2O, HCl, pH 3.
hydroxy groups of 51, followed by NaBH4 reduction,
gave the nitroalkyl derivative 52,10 which proved resis-
tant to the direct transformation into the corresponding
aldehyde 54 through Nef reaction. The desired aldehyde,
in the form of lactol 54, could be prepared in 38% overall
yield from 39, by transformation of the nitromethylene

2336 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
Sch em e 4a
The E aldehyde 61 (Scheme 5) could be obtained by
reduction of the E/ Z isomeric mixture of nitriles 6013
with Raney Ni and NaH2PO2,14 as the Z isomer did not
react in the reaction conditions used.
Aminodiene 36 (Scheme 6) was synthesized by LiAlH4
reduction of the azido derivative 63 obtained by pal-
ladium-catalyzed cross-coupling of iodoalkene 62 with
the stannylalkene 68 of Scheme 8. In turn, iodoalkene
62 was prepared smoothly from aldehyde 39 by Takai
reaction.
Aminodiene 37 and aminoalkane 38 were synthesized
as shown in Scheme 7. Stille coupling of iodoalkene 62
with the stannylalkene 69 of Scheme 8 gave the cyano-
diene 64, which was reduced to the aminodiene 37 with
LiAlH4. Catalytic hydrogenation of 37 afforded the
corresponding aminoalkane 38 together with an un-
separable unidentified aminoalkene; when we forced the
hydrogenation conditions we met with extensive deg-
radation. The same behavior was observed in the
catalytic hydrogenation of the cyanodiene 64, where the
desired saturated derivative was obtained together with
an unidentified cyanoalkene. Compound 64 was then
converted to the corresponding 3â-O-acetyl derivative
before being hydrogenated; the mixture of 65 and
cyanoalkene was treated with MCPBA to obtain a new
mixture of the unreacted 65 and the epoxide of the
alkene, which could be easily separated by flash chro-
matography. LiAlH4 reduction of 65 gave the desired
aminoalkane 38.
a
Reagents and conditions: (a) H2NOCH2CH2N(CH3)2‚2HCl,
NaOH, dioxane, H2O.
Sch em e 5a
The intermediate hydroxylamines 76-81 were pre-
pared as shown in Scheme 9. Compounds 76-80 were
obtained by alkylation of benzophenone oxime with the
corresponding ω-haloalkylamines in the presence of
KOH, followed by acid hydrolysis of the O-aminoalkyl
derivatives 70-74; compound 81 was obtained by alky-
lation of N-hydroxyphthalimide with commercially avail-
able N-(4-bromobutyl)phthalimide, followed by hydra-
zine hydrolysis of the resulting diphthalimido derivative
75. Propoxyamine,15 2-hydroxyethoxyamine,16 and 2-ami-
nooxyacetic acid hydrochloride,17 intermediates for com-
pounds, 10, 15-17, 21, and 22, were prepared as
described in the literature.
Biologica l Activity
All compounds were evaluated in vitro for inhibition
of dog kidney Na+,K+-ATPase activity,18 measured as
percent hydrolysis of 32P-ATP;19 data are shown in
Tables 1 and 2. Compounds showing an inhibitory
potency of the enzyme (IC50) at least in the micromolar
range were further investigated in vitro for the inotropic
activity by measuring their effects on the contractile
force of an electrically driven guinea pig left atrium;
data are shown in Table 3.
Some compounds showing the highest potency in vitro
were investigated in vivo for inotropic activity and lethal
effect in anesthetized guinea pigs during slow intrave-
nous infusion; results are reported in Table 4.
a
Reagents and conditions: (a) H2NOR‚2HCl, NaOH, EtOH, pH
2; (b) NCCH2PO(OEt)2, tBuOK, THF; (c) Raney Ni, NaH2PO2‚H2O,
H2O/AcOH/pyridine, 1:1:2, 60 °C, 24 h; (d) H2NOR‚2HCl, NaOAc,
HCl, dioxane, H2O, pH 4.5.
group of compound 52 into a nitrile group by treatment
with DEAD/Ph3P,11 hydrolysis of this derivative with
NaOH to compound 53, and reduction of 53 with
DIBAH. Lactol 54 gave the oxime 11 in low yield (18%)
because it was stable and reacted poorly under the
alkaline reaction conditions we had to choose, to limit
the amount of Z isomer.
Aldehyde 42 (Scheme 3) was prepared in two steps
by reduction with DIBAH of the known ester 5512 to
the allylic alcohol 56 and subsequent oxidation with
MnO2, following the method already described for 41.4
The corresponding saturated aldehyde 40 was prepared
by catalytic hydrogenation of 41.
Digoxin and digitoxigenin were chosen as reference
compounds, as the former is the most commonly pre-
scribed cardiac glycoside in CHF and the latter has the
same steroidal aglyconic skeleton of our compounds.
Resu lts a n d Discu ssion
The compounds were tested as isomeric mixtures as
obtained from the synthesis. Taking into consideration

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2337
Sch em e 6a
a
Reagents and conditions: (a) CrCl2, CHI3, THF, 0 °C, 2 h; (b) 68, PdCl2(CH3CN)2, DMF; (c) LiAlH4, THF.
Sch em e 7a
a
Reagents and conditions: (a) 69, PdCl2(CH3CN)2, DMF; (b) LiAlH4, THF, reflux, 24 h; (c) Ac2O, DMAP, pyridine, CH2Cl2; (d) H2,
PtO2, EtOAc; (e) MCPBA, CHCl3; (f) LiAlH4, THF, reflux, 24 h.
Sch em e 8a
It should be noted that performing the tests on
mixtures of isomers could give defective results in the
case of different binding and intrinsic activity of the
components. In our hands it was impossible to separate
E/ Z isomers, with the exception of compounds 1 and 2.
All the following reasoning is based on the mixtures of
components.
En zym e In h ibition . The oximes 1-3, “parent com-
pounds” of derivatives listed in Table 1, showed IC50s
comparable to that of the “parent” aldehyde 39, sug-
gesting that the same type of polarization of the double
bond and of interaction with the receptor occurs. The
introduction of the basic side chain in compounds 4-9
resulted in a dramatic increase of the inhibitory poten-
cies, from 15- to 5000-fold; the chain length influenced
the activities in the following order: (CH2)3 > (CH2)2 >
(CH2)4, suggesting that six atoms between the basic
group and the C17 position of digitoxigenin skeleton
could be the optimal distance for a strong interaction
with the receptor. As evidence of the importance of the
basic group, the acid 10 did not inhibit the enzyme.
a
Reagents and conditions: (a) CH3SO2Cl, TEA, CH2Cl2, 0 °C;
(b) NaN3, DMF, 50 °C, 4h; (c) KCN, KI, DMF, 80 °C, 9 h.
that the biological assays were performed within 1-4
h from the dissolution of the compounds and the long
equilibration time at pH 7.4 reported above, we take
for granted that no important isomerization of the tested
compounds occurred in the biological systems used for
the assays.

2338 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
Sch em e 9a
than the corresponding nonconjugated oxime 3. The
introduction of a basic chain in the 1,4-polarized, R,â-
unsaturated aldehyde 41 gave the most active com-
pounds, 18 and 19, with an average inhibitory potency
greater than that of the more active basic substituted
oximes 5 and 8, derived from the 1,2-polarized aldehyde
39. Indeed compound 8 showed the same high activity
as 18 and 19, probably because the amine group in 8 is
at the right distance of six atoms (see above) for the best
interaction with the complementary site in the receptor,
a distance impossible to achieve in the 1,4-polarized
series (seven atoms as a minimum) for reasons of
chemical stability. The replacement of the amine with
a carboxylate gave the inactive compound 17. The much
lower inhibitory potencies of compounds 15 and 16,
where a basic amino group was replaced by a methyl
and a hydroxy group, respectively, confirm the impor-
tance of the basicity of an amino group when compared
with the lipophilicity of a methyl group or the hydrogen-
bonding capacity of a hydroxy group.
a
The 1,4-polarized 21-methyl oximes 20-24 had the
same general behavior of the corresponding desmethyl
derivatives, although with at a bit lower level of Na+,K+-
ATPase inhibitory activity. That was probably due to
the steric hindrance of the methyl substituent, prevent-
ing a closer interaction with the receptor (see also the
parent aldehyde 42, 6 times less active than the corre-
sponding desmethyl 41). The best interaction occurred,
as expected, with the dimethylaminoethyl and amino-
ethyl substituents (compounds 23 and 24); any further
outdistancing of the basic group, as in compounds 25-
27, decreased the inhibitory activity.
Reagents and conditions: (a) N-hydroxyphthalimide, K2CO3,
DMSO, 80 0C, 0.5 h; (b) hydrazine, EtOH, reflux, 3 h; (c) HCl; (d)
R2N(CH2)nCl, KOH, DMSO; (e) 6 N HCl, reflux, 2 h.
Ta ble 3. Inotropic Activity on Electrically Driven Guinea Pig
Left Atrium
a
Emax
,
% increase concn to obtain
compd
from basal force
E
max, µM
EC50,b µM
4
5
6
7
8
9
39
235
214
145
126
92
73
150
70
3
10
30
3
nd
1.3
4.7
0.62
0.18
2.9
1.8
0.65
1.1
1
10
10
3
Although good potencies were retained in the ami-
noalkyloximes 11-13, where one (11) and two (12, 13)
methylenes have been placed between the oxime func-
tion and the steroidal skeleton, the potencies were lower
than those of the most potent compounds, where the
aminoalkyloxime function is either direcly linked to the
steroidal skeleton (compound 8) or separated by an
ethylene group (compound 19). Strangely enough, com-
pound 12, with the basic group at a seven atom distance
from C17, was more active than compound 11, at a six
atom distance. The lower potency of these compounds
could be easily forecast from the great difference in
activity (160 times) of the R,â-unsaturated aldehyde 41
vs the saturated one 40, while the “distance anomaly”
can be explained with the presence, in compounds 12
and 13, but not in compound 11, of an iminic bond at
the same distance from C17 as that of the carbonyl
group of the lactone in the natural digitalis derivatives.
It is worth noting that the difference in activity between
the R,â-unsaturated basic oxime 19 vs the corresponding
saturated oxime 13 is much less than that of the
corresponding R,â-unsaturated and saturated parent
aldehydes 41 and 40 (4 times vs 160 times). Evidently
the strength of the interaction of the basic amino groups
with the receptor tends to lower the importance and
weight of the other interacting groups.
11
12
13
16
18
19
22
23
24
25
26
27
29
30
31
34
36
37
38
39
41
49
3
179
191
155
175
146
115
90
106
145
18
135
237
71
98
73
300
0.3
0.3
300
0.3
1
39
0.057
0.05
23
0.07
0.07
0.17
4.0
20
nd
8
3.9
1
30
100
300
100
30
30
3
0.1
100
300
10
100
3
12
0.38
0.035
nd
11.0
4
18
0.57
0.38
43
79
45
136
200
184
digitoxigenin
digoxin
1
a
b
Maximal increase in force of contraction. Concentrations
producing 50% of the maximal increase in force of contraction were
calculated from concentration-response curves; nd: not deter-
mined.
The 1,4-polarization of aldehyde 41, mimicking the
same electronic situation of the lactone in natural
digitalis compounds, gave a more efficient interaction
with the receptor, inducing a 100 times increase of
potency compared to the 1,2-polarized aldehyde 39. The
same behavior occurred in the R,â-unsaturated oximes,
although to a lesser extent; in fact, the O-methyl R,â-
unsaturated oxime 14 was only 8 times more potent
A further demonstration of the above-mentioned
strength of interaction of the basic terminal with the
Na+,K+-ATPase receptor is the potency of compound 36,
where the oxime group of compound 19 was replaced
by a nonpolarized ethylene spacer and the amino group
remained the only interacting group present in the side

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2339
Ta ble 4. Inotropic and Toxic Effects in Anesthetized Guinea Pig
a
concn of infused
solution, µM
Emax
,
%
dose to obtain
max, µmol/kg
lethal dose,
µmol/kg
deaths/
treated
lethal dose/
ED80
compd
increase in +dP/dt
E
ED80,b µmol/kg
8
9
12
18
24
25
digoxin
41.4
1000
220
11.0
12.8
375
163 ( 24
148 ( 44
132 ( 28
110 ( 24
152 ( 29
99 ( 8
0.435 ( 0.083
5.03 ( 0.221
3.65 ( 0.940
0.275 ( 0.018
0.201 ( 0.018
0.884 ( 0.075
0.576 ( 0.064
0.240 ( 0.041
1.41 ( 0.161
0.814 ( 0.295
0.185 ( 0.018
0.128 ( 0.018
0.447 ( 0.038
0.205 ( 0.026
0.663 ( 0.435
14.8 ( 1.03
7.83
2/4
3/3
1/4
1/4
4/4
2.8
10.5
9.6
2.1
2.3
0.385
0.292 ( 0.018
1.67 ( 0.207
1.25 ( 0.102
4/4
11/11
3.7
6.1
51.2
167 ( 30
a
b
Maximal increase in +dP/dt. Doses producing 80% of increase in +dP/dt, from basal force, were calculated from dose-response
curves; nd: not determined. Data are means ( SEM when possible.
chain: compound 36 is only 4 times less potent than
compound 19. Also in this case the optimal length of
the chain between the C17 carbon atom and the amino
group was confirmed to be six atoms (see 36 vs 37).
Reduction of the olefinic bonds of 37 to give the
saturated compound 38 gave a further 4 times decrease
of activity.
Reduction of the iminic bond to give hydroxylamines
28-35 caused the activity to drop; when the most active
oximes (4, 7, and 8) were reduced to the corresponding
hydroxylamines (29-31), the decrease was on the order
of 10-50 times. The lack of polarization of the iminic
bond cannot explain these data completely, since the
same situation should also apply to olefinic or aliphatic
amines 36-38. A tentative explanation is that an
intramolecular hydrogen bond between the protonated
amine and the hydroxylamine nitrogen atom could fold
the chain, thus making the interaction with the anionic
site of the enzyme energetically less favorable.
Oximes 43-46 of Table 2 were inactive. The steric
situation of these compounds and/or the shortness of the
aliphatic chain can be an explanation for their inactiv-
ity. In fact, in these compounds, the double bond,
directly attached to C17, projects the oxime chain in the
same plane of the D ring, while in the case of derivatives
1-42 the chain is in position 17â, that is, over the plane
of the D ring. The longer chain oximes 47 and 49,
permitting the amine group to reach the proper position
for a good interaction with the receptor, showed better
activities, being 25 and 4 times less potent than digoxin
and digitoxigenin, respectively. Further increase of the
chain length in 48 and 50 caused the activity to drop
again.
The lack of correlation when the entire set of com-
pounds but the oximes was considered may depend on
the different distribution and biochemical characteris-
tics of the Na+,K+-ATPase isoforms in different tissues
and species. In fact, the IC50 was measured on isolated
and purified Na+,K+-ATPase from dog kidney, which
contains the R1 isoform, whereas EC50 was measured
on the whole guinea pig atria which contains both R1
and R3 isoforms. In this context, it must be remembered
that the sensitivity of the Na+,K+-ATPase for cardiac
glycosides is isoform- and species-dependent (the affinity
of the glycosides is higher for the R3 than for the R1
and R2 isoforms, and the Na+,K+-ATPase of the dog is
more sensitive than that of the rat and the guinea pig).
It has also been hypothesized that the inotropic or
arrhythmogenic activities may be specifically sustained
by one or the other isoform.20 Therefore, it cannot be
excluded that different chemical classes of compounds
may differently interact with the Na+,K+-ATPase iso-
forms (R1 and R3) underlying specific functional activi-
ties at cardiac level. This hypothesis is currently under
evaluation in our laboratory.
The compounds tested in vivo showed good maximum
inotropic effects (Emax), comparable in some cases to the
value of digoxin. Compounds 8, 18, and 24, the most
potent both in the inhibition test and in the guinea pig
atrium, revealed the same inotropic effect (ED80) at
doses comparable to, or even lower than, digoxin.
Unfortunately, a parallelism seems to occur between
potency (ED80) and toxicity, though for the weakest
compounds, 9 and 12, a safety ratio slightly higher than
that of digoxin was found. These compounds deserve
further investigation because other factors, such as
pharmacokinetics, can be taken into account.
In all cases, primary amines showed inhibitory activi-
ties higher than the corresponding tertiary amines,
probably due to a limited availability of space around
the protonated amine interacting with the anionic site
of the receptor.
Molecu la r Mod elin g of t h e Ou a b a in Bin d in g
Site. To provide a further insight on the observed
activities, a molecular model was created for the recep-
tor site on the basis of the information currently
available in the literature. Comprehensive random
mutagenesis studies and photoaffinity labeling showed
that most of the amino acid substitutions conferring
ouabain-resistance relate to the H1-H2 region of the
Na+,K+-ATPase R1-subunit. In particular, the greatest
ouabain resistance is given by substitutions in the
extracellular loop, such as Gln111Arg, Asp121(Gly, Glu),
or Asn122Asp,21,22 or Gln118Arg and Asp129Asn in
Torpedo electroplax,23 as well as those involving the
hydrophobic core of the transmembrane region, such as
Cys104(Ala, Phe) or Tyr108Ala,24 or Tyr124Cys or
Ile135Val.25 A specific role for Cys104 was established,
with a putative hydrogen-bond formation with the
lactone ring. On this basis, Repke et al.6,26,27 proposed
In otr op ic Activity. A highly significant correlation
was found between the inhibitory potency on Na+,K+-
ATPase (IC50) and inotropic potency (EC50) in isolated
atria in the set of tested oximes (r ) 0.918, r2 ) 0.842,
n ) 17, p < 0.0001). The correlation data between IC50
and EC50 dropped dramatically on the entire set of
compounds (oximes, hydroxylamines, alkenes, alkane,
and aldehydes; r ) 0.460, r2 ) 0.210, n ) 25, p )
0.0199). The correlation was even lower (r ) 0.316,
r2 ) 0.100, n ) 8, p ) 0.445) when all the compounds
but the oximes were considered. In particular, alkenes
36 and 37 seem to be more active (EC50) on the atrium
than on the isolated enzyme, while the opposite happens
for the oximes.

2340 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
F igu r e 1. Pictorial representation of the interprotodimeric
cleft constituted by two interacting subunits (A and B) of the
Na+,K+-ATPase. Residues known to be important for digitalis
binding are boxed.
a model, formed by two interacting H1-H2 domains
arranged in a four-helix bundle to form a cleft, where
digitalis sinks and specifically binds.
F igu r e 2. Model of the interprotodimeric cleft constituted by
two subunits (A and B) of the Na+,K+-ATPase interacting with
compound 8.
Although there is substantial agreement in the lit-
erature about the length and position of the transmem-
brane domains (Phe93-Ile110 and Leu123-Ser140 for
H1 and H2, respectively, with the connecting loop
mainly composed of hydrophilic residues), no informa-
tion is at present available on the molecular structure
of any ATPase R1-subunit at the atomic level nor on
the relative orientation of H1 with respect to H2.
Following the work of Repke,6 we made use of the
available information on the topology of the transmem-
brane segments and the above-mentioned amino acids
to build a molecular model of the H1-H2 dimer, where
all residues known to be involved in digitalis binding
are situated in the inner groove (Figure 1). The PHD-
predicted transmembrane segments28 were arranged in
a helical loop geometry, according to the X-ray structure
of the purple membrane bacteriorhodopsin,29,30 with the
external loop constituted by the stretch of negatively
charged residues (Glu115-Glu117) and by Pro118 situ-
ated just in the middle to give an energetically favorable
turn conformation for the correct peptide backbone
folding.
At first, a suitable orientation of digoxin was found
by manually docking the molecule within the groove of
a single R1-subunit on the basis of the established
interactions, i.e. with the sugar ring positioned at the
level of the external loop and the Cys104 SH group
forming a H-bond with the lactone ring. The complex
was allowed to relax by a cycle of combined energy
minimization and molecular dynamics (see experimen-
tal for details). A dimer was subsequently built by
arranging another R1-subunit in a symmetrical orienta-
tion and by energy minimizing the whole complex.
The model thus obtained revealed some interesting
features (Figure 2). The steroid moiety is kept in place
by a large number of favorable interactions, contributing
to build up a negative (attractive) interaction energy
between the various functional groups and the amino
acid side chains. In particular, the hydroxyl group at
position 3 is held at the beginning of the cleft by
hydrogen bonding formed with the Tyr124-OH of both
subunits and with the amide group of H2 Asn122. H1
Cys104 interacts with the oxime heteroatoms of many
compounds, with Tyr108 giving a further stabilization
by means of a hydrogen-bond with the oxime moiety
and, in the case of the unsaturated analogues, with
attractive van der Waals interactions. Thus, the steroid
moiety is held in place by a large network of hydrogen
bonds and hydrophobic interactions, involving, for in-
stance, Cys104, Tyr108, Asn122, Tyr124, Ser130, and
Ile134. This latter residue is adjacent to Ile135, a
residue considered important for digitalis binding. A
superimposition of our derivatives with digoxigenin
shows that the oxime group held a position analogous
to the digoxigenin lactone; thus, the same types of
interactions for the steroid moiety and the side chain
are created up to the oxime group itself. However, the
newly synthesized analogues build up novel interactions
within the cleft with other amino acids not previously
shown to be involved in the binding. The amino groups
of the side chain, whether methylated or not, are found
to interact in our model with Cys138. Given the relative
acidity of the cysteine SH groups (pKa ) 8.33),31 it may
well be that an acid-base interaction occurs between
this latter group and the terminal NH2 of the analogues,
giving rise to a salt-bridge of the -CH2S- NH3+- type.
The existence of an ion pair between the thiolate and
protonated amine groups has already been demon-
strated in the case of glutathione-S-transferase P1-1
(GST) in comparison with its K54A mutant.32 For the
wild-type GST a pKa value of 4.2 for Cys47 was
measured and the existence of the thiolate anion,
possibly in ion pair with the Lys54 side chain, was
detected by the characteristic UV absorbance at 229 nm.

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2341
In the Na+,K+-ATPase model, along with the cysteine
SH groups, also the OH groups of Tyr141 and Tyr142,
situated one complete helix-turn further apart, are
found in the bottom of the cleft and provide other
hydrogen-bonding possibilities for compounds having
longer side chains.
same happens with toxicity, so that the safety ratio
results are slightly lower than that of digoxin. On the
contrary, some weaker compounds (9 and 12) showed
safety ratios slightly higher than digoxin. This behavior
deserves further investigation.
The presence of an amino function in the 17â-
substituent is of utmost importance for a strong interac-
tion with the receptor. The contemporary presence of
an oxime group further strengthens the interaction with
the receptor, more if the oxime is an R,â-unsaturated
one, thus mimicking the electronic situation of the
natural 17â-unsaturated lactone in digitalis derivatives.
A proper distance of the amine group from the C(17) of
the steroidal skeleton is also of great importance, a
spacer of six atoms being the most suitable one. Within
the basic group, a primary amine is always more active
than a tertiary amine.
These features are supported by a molecular model
suggesting the possible interactions of the above-
described groups with specific amino acid residues in
the H1-H2 domains of Na+,K+-ATPase; some of them
confirm the classical interactions already described in
previous papers and confirmed by mutagenesis studies.
On the contrary, the strong interaction of a basic group
with the Cys138 is a new one and opens new possibili-
ties to design compounds interacting with this region
of the receptor.
In particular, this led to new classes of non-digitalis
derivatives, now under study, with strong activities and,
more important, a therapeutic index better than the
classical digitalis derivatives. This gives us a chance to
overcome the failures of the past 50 years of research
in finding a valid alternative to the more than a century
old digoxin.6
Although the model was built on a limited number of
molecules and needs further refinement (work in
progress) for a quantitative evaluation, it does provide
some insight into the structure-activity relationship of
many of our newly synthesized compounds. Some gen-
eral rules may be set on the basis of the model
structure: (1) Compounds terminating with a basic
group (NH2 or NMe2) show higher activity than those
terminating with an acidic (COOH) or neutral (OH or
OMe) group. (2) Within the class with the highest
activity, a further modulation arises from the geometry
and the chemical structure of the side chain, the
optimum substituent linear chain being constituted by
a six or seven atom spacer between the C(17) carbon
atom of the steroidal ring and the basic amino group.
Compounds 26 and 27, with a maximum chain length
of 10 and 11 atoms, are less active, with an IC50 equal
to 1.26 and 6.3 µM, respectively.
Our model nicely fits with these general observations.
The binding energy is given by a sum of attractive
forces, with the terminal basic amino group playing a
primary role if positioned at a correct distance from the
only acidic groups available within the cleft, identified
with Cys138 of both subunits. Thus, compounds 36-
38 have IC50 values (0.08-1.25 µM) comparable to that
of digoxin itself (IC50 ) 0.50 µM), despite lacking any
heteroatom capable of hydrogen-bonding with Cys104.
Rather bulky amino groups, such as the terminal NMe2,
can be accommodated in the rather large bottom of the
cleft, but a longer chain decreases the potency, as in
the case of compound 26 (IC50 )1.26 µM). This may be
explained by the building up of unfavorable repulsive
interactions, between the amino group and the bottom
of the cleft, constituted in our model by Tyr141 and
Tyr142 of both subunits. Cys138 and Tyr142 are local-
ized respectively 8 and 12 residues downstream the
primary sequence from Ser130 and reside exactly two
and three helix turns further from this residue. Thus,
our findings suggest the existence of new side chains
with hydrogen-bonding capabilities, apart from the
previously ascertained ones, such as Cys104 or Asn122.
Such new hydrogen-donor groups are localized about 12
and 16 Å from Ser130 and are buried very close to the
bottom of the interprotodimeric cleft.
Exp er im en ta l Section
Ch em istr y. Elemental analyses were performed by Redox,
1
Cologno Monzese, Italy. H NMR spectra were recorded on a
Bruker AC-300 spectrometer at 300.13 MHz. Chemical shifts
(δ) are given in ppm downfield from tetramethylsilane as
internal standard and coupling constants (J values) are in
1
hertz. H NMR assignments were drawn from classical argu-
ments on chemical shift and coupling constant behavior. Mass
spectral data were obtained with the electron impact ionization
technique at 70 eV from a Finnigan INCOS-50 mass spec-
trometer using the direct exposure probe (DEP). Chromatog-
raphies were carried out on silica gel (Baker 7024-02) in all
instances. Solvents and reagents were used as purchased from
Aldrich.
P r ep a r a tion of Oxim es. Meth od A. A solution of NaOAc
(4 equiv) and the appropriate hydroxylamine (1.05 equiv) in
dioxane/water 3:2 (0.1 M) was adjusted to pH 4.5 with 6 N
HCl. A solution of the appropriate aldehyde (1 equiv) in
dioxane/water 2:1 (0.2 M) was added dropwise at room
temperature. After 1-2 h for compounds 1-10 or 4-6 h for
compounds 20-26, dioxane was evaporated in vacuo. The
residue was diluted with water and 1 N NaOH added until
pH 9.5 was reached. The mixture was extracted with EtOAc
(3×), and the organic layers were dried over Na2SO4 and
evaporated in vacuo. The residue was purified by flash
chromatography and, eventually, transformed into the salt by
adding the stoichiometric amount of the corresponding acid.
P r ep a r a tion of Oxim es. Meth od B. To a solution of the
appropriate hydroxylamine (1.2 equiv) in 1 N NaOH/dioxane
5:2 (1.7 M) was added dropwise a solution of the appropriate
aldehyde (1 equiv) in dioxane (0.5 M) at room temperature.
After 1 h the solution was diluted with water and extracted
with CHCl3 (3×). The organic layers were dried over Na2SO4
and evaporated in vacuo. The residue was purified by flash
Con clu sion s
Among the digitalis-like compounds presented in this
paper, some 17â-aminoalkyloxime derivatives showed
extraordinary high inhibitory activity on Na+,K+-
ATPase and inotropic potency on guinea pig atrium. The
most active of them (8, 18, 19, and 24) were 17-25 times
more potent on the isolated dog kidney Na+,K+-ATPase
(IC50) than the reference compounds digoxin and digi-
toxigenin and showed inotropic potencies (EC50) 3-11
times higher in isolated guinea pig left atria.
The in vivo evaluation, in the guinea pig, showed that
the most potent compounds (8, 18, and 24) had inotropic
effect comparable to or even higher than digoxin. The

2342 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
chromatography and, eventually, transformed into the salt by
adding the stoichiometric amount of the corresponding acid.
P r ep a r a tion of Hyd r oxyla m in es. Meth od C. A solution
of crude oxime as a base (1 equiv) in MeOH (0.2 M) was
adjusted to pH 3.0 with 1 N HCl under stirring at room
temperature. NaBH3CN (1.5 equiv) was added followed by
water (0.1 mL/equiv). The reaction was continuously kept at
pH 3 by a pH-stat controlling the addition of 1 N HCl. After 6
h NaBH3CN (0.75 equiv) was added followed by water (0.1 mL/
equiv). Additions of NaBH3CN (0.4 equiv) and water (0.1 mL/
equiv) were repeated after 24 and 30 h. After 48 h the solution
was brought to pH 1.8 with 1 N HCl. After stirring for 1 h the
solution was brought to pH 9.5 with 4 N NaOH and methanol
was evaporated in vacuo. The mixture was diluted with water
and extracted with EtOAc (3×). The combined organic layers
were dried over Na2SO4 and evaporated in vacuo. The residue
was purified by flash chromatography and, eventually, trans-
formed into the salt by adding the stoichiometric amount of
the corresponding acid.
P r ep a r a tion of Oxim es. Meth od D. A solution of ketone
59 (1 equiv) and the appropriate hydroxylamine (4 equiv) in
ethanol (0.08 M) was adjusted to pH 2 with 0.1 N NaOH and
stirred at room temperature for several days. Ethanol was
evaporated in vacuo. The residue was diluted with water, 1 N
NaOH added, and the mixture extracted with CH2Cl2 (3×).
The combined organic layers were washed with 1 N NaOH,
dried over Na2SO4, and evaporated in vacuo. The residue was
purified by flash chromatography.
(E)-17â-[(Oxim in o)m eth yl]-5â-an dr ostan e-3â,14â-diol (1)
was prepared following method A. The crude product on
chromatography with n-hexane/EtOAc, (1:1), Rf 0.34, gave a
white solid. 1H NMR (CDCl3): δ 0.94 (s, 3H, CH3), 0.98 (s, 3H,
CH3), 2.42 (m, 1H, 17-H), 4.13 (m, 1H, 3-H), 6.68 (bb, 1H,
NOH), 7.62 (d, 1H, J ) 9.5, CHN). MS: m/z 335 (8, M+), 203
(42).
(Z)-17â-[(Oxim in o)m eth yl]-5â-an dr ostan e-3â,14â-diol (2)
was prepared following method A. The crude residue was
crystallized twice from EtOAc to give a white solid; Rf 0.26
(n-hexane/EtOAc, 1:1). 1H NMR (CD3OD): δ 0.96 (s, 6H, 2
CH3), 3.28 (m, 1H, 17-H), 4.04 (m, 1H, 3-H), 6.81 (d, 1H, J )
7.7, CHN). MS: m/z 335 (1, M+), 203 (48).
with CHCl3/MeOH/26% NH4OH, 89:10:1 followed by salt
formation with the stoichiometric amount of oxalic acid in
EtOH/EtOAc. After evaporation in vacuo of the resulting
solution, the residue was triturated with Et2O/EtOH to give
6, a white solid. 1H NMR (CDCl3): δ 0.92 (s, 3H, CH3), 0.96
(s, 3H, CH3), 2.38 (m, 0.9H, 17-H (E) isomer), 2.80 (s, 6H,
N(CH3)2), 3.01 (m, 2H, NCH2), 3.11 (m, 0.1H, 17-H (Z) isomer),
4.02 (m, 2H, OCH2), 4.13 (m, 1H, 3-H), 6.86 (d, 0.1H, J ) 8.2,
CHdN (Z) isomer), 7.60 (d, 0.9H, J ) 9.0, CHdN (E) isomer).
MS: m/z 434 (0.1, M+ base), 58 (100).
(E)-17â-[[(2-Am in oeth oxy)im in o]m eth yl]-5â-an dr ostan e-
3â,14â-d iol fu m a r a te (7) was prepared following method A.
The crude product was purified by chromatography with
CHCl3/MeOH/26% NH4OH, 89:10:1. The purified product as
a base (1.85 g) was dissolved in EtOH (15 mL) at 40 °C and
added to a boiling solution of the stoichiometric amount of
fumaric acid in EtOH (8 mL). EtOAc (10 mL) was added to
the resulting solution. After stirring at room temperature
overnight, the solid was collected to give 7 (1.90 g), a white
solid. 1H NMR (CD3OD): δ 0.89 (s, 3H, CH3), 0.97 (s, 3H, CH3),
2.37 (m, 1H, 17-H), 3.19 (m, 2H, NCH2), 4.04 (m, 1H, 3-H),
4.19 (m, 2H, OCH2), 6.68 (s, 2H, fumarate), 7.64 (d, 1H, J )
9.4, CHdN). MS: m/z 378 (5, M+ base), 318 (100).
(E)-17â-[[(3-Am in op r op oxy)im in o]m eth yl]-5â-a n d r os-
ta n e-3â,14â-d iol oxa la te (8) was prepared following method
A. The crude product was purified by chromatography with
CHCl3/MeOH/26% NH4OH, 89:10:1. The purified product as
a base (0.40 g) was dissolved in EtOAc (5 mL) and EtOH (5
mL) and the stoichiometric amount of oxalic acid was added.
After evaporation in vacuo, the residue was triturated with
1
EtOAc to give 8 (0.29 g), a white solid. H NMR (CD3OD): δ
0.88 (s, 3H, CH3), 0.97 (s, 3H, CH3), 2.32 (m, 1H, 17-H), 3.00
(m, 2H, NCH2), 4.05 (m, 3H, 3-H and OCH2), 6.86 (d, 0.1H, J
) 8.7, CHdN (Z) isomer), 7.56 (d, 0.9H, J ) 10.0, CHdN (E)
isomer). MS: m/z 392 (2, M+ base), 302 (100), 250 (90), 203
(32).
(E)-17â-[[(4-Am in obu toxy)im in o]m eth yl]-5â-an dr ostan e-
3â,14â-d iol oxa la te (9) was prepared following method A. The
crude product was purified by chromatography with CHCl3/
MeOH/26% NH4OH, 89:10:1. The purified product as a base
(2.88 g) was dissolved in EtOH (25 mL) and the stoichiometric
amount of oxalic acid was added. After evaporation in vacuo,
the residue was triturated with EtOAc to give 9 (2.90 g), a
white solid. 1H NMR (CD3OD): δ 0.87 (s, 3H, CH3), 0.97 (s,
3H, CH3), 2.32 (m, 1H, 17-H), 2.95 (m, 2H, NCH2), 3.99 (m,
2H, OCH2), 4.04 (m, 3H, 3-H), 6.82 (d, 0.1H, J ) 7.5, CHdN
(Z) isomer), 7.53 (d, 0.9H, J ) 9.4, CHdN (E) isomer). MS:
m/z 406 (1, M+ base), 250 (23), 88 (100).
(E)-17â-[(Ca r b oxym et h oxyim in o)m et h yl]-5â-a n d r os-
ta n e-3â,14â-d iol (10) was prepared following method A. The
extraction from the aqueous phase was carried out at pH 3
with EtOAc/THF, 9:1, and the crude product was triturated
with EtOAc to give a white solid. 1H NMR (CD3OD): δ 0.90
(s, 3H, CH3), 0.97 (s, 3H, CH3), 2.33 (m, 1H, 17-H), 4.04 (m,
1H, 3-H), 4.46 (s, 0.26H, OCH2 (Z) isomer), 4.51 (s, 1.74H,
OCH2 (E) isomer), 6.99 (d, 0.1H, J ) 7.5, CHdN (Z) isomer),
7.63 (d, 0.9H, J ) 10.0, CHdN (E) isomer). MS: m/z 250 (66),
203 (6).
(E)-17â-[2-[(2-Dim eth yla m in oeth oxy)im in o]eth yl]-5â-
a n d r osta n e-3â,14â-d iol (11) was prepared following method
B. The crude product on chromatography with CHCl3/MeOH
(8:2) gave a white foam. 1H NMR (CD3OD): δ 0.97 (s, 3H, CH3),
0.98 (s, 3H, CH3), 2.16 (t, 0.9H, NCH2 (E) isomer), 2.18 (t, 0.1H,
NCH2 (Z) isomer), 2.31 (s, 6H, N(CH3)2), 2.30-2.50 (m, 2H,
CH2CdN), 4.04 (m, 1H, 3-H), 4.11 (t, 0.9H, OCH2 (E) isomer),
4.18 (t, 0.1H, OCH2 (Z) isomer), 6.69 (d, 0.1H, J ) 5.9, CHdN
(Z) isomer), 7.36 (d, 0.9H, J ) 7.5, CHdN (E) isomer). MS:
m/z 420 (4.4, M+), 332 (33), 58 (100).
(EZ)-17â-[3-[(2-Dim eth yla m in oeth oxy)im in o]p r op yl]-
5â-a n d r osta n e-3â,14â-d iol oxa la te (12) was prepared fol-
lowing method B. The crude product was purified by chroma-
tography with CHCl3/MeOH, 8:2. The purified product as a
base (0.30 g) was dissolved in Et2O (25 mL) and EtOAc (5 mL)
and the stoichiometric amount of oxalic acid was added. The
(E)-17â-[(Meth oxyim in o)m eth yl]-5â-a n d r osta n e-3â,14â-
d iol (3) was prepared following method A. The crude product
on chromatography with cyclohexane/EtOAc, (6:4), Rf 0.33,
followed by trituration with i-Pr2O/EtOH, gave a white solid.
1H NMR (CDCl3): δ 0.93 (s, 3H, CH3), 0.96 (s, 3H, CH3), 2.41
(m, 1H, 17-H), 3.81 (s, 3H, OCH3), 4.13 (m, 1H, 3-H), 7.56 (d,
1H, J ) 9.3, CHN). MS: m/z 349 (22, M+), 318 (100).
(E)-17â-[[(2-Dim eth yla m in oeth oxy)im in o]m eth yl]-5â-
a n d r osta n e-3â,14â-d iol L-ta r tr a te (4) was prepared follow-
ing method A. EtOAc (55 mL) was added to the solution of
the crude product (3.80 g) and the stoichiometric amount of
L-tartaric acid in EtOH (15 mL). After standing at 0 °C
overnight, the solid was collected to give 4 (4.15 g), a white
solid. 1H NMR (CD3OD): δ 0.89 (s, 3H, CH3), 0.97 (s, 3H, CH3),
2.37 (m, 1H, 17-H), 2.89 (s, 6H, N(CH3)2), 3.42 (m, 2H, NCH2),
4.04 (m, 1H, 3-H), 4.33 (m, 2H, OCH2), 4.89 (s, 2H, tartrate),
6.96 (d, 0.1H, J ) 7.2, CHdN (Z) isomer), 7.56 (d, 0.9H, J )
9.3, CHdN (E) isomer). MS: m/z 406 (5, M+ base), 58 (100).
(E)-17â-[[(3-Dim eth yla m in op r op oxy)im in o]m eth yl]-5â-
a n d r osta n e-3â,14â-d iol oxa la te (5) was prepared following
method A. The crude product was purified by chromatography
with CHCl3/MeOH/26% NH4OH, 89:10:1 followed by salt
formation with the stoichiometric amount of oxalic acid in
EtOH/EtOAc. After evaporation in vacuo of the resulting
solution, the residue was triturated with Et2O/EtOH to give
1
5, a white solid. H NMR (CD3OD): δ 0.88 (s, 3H, CH3), 0.97
(s, 3H, CH3), 2.33 (m, 1H, 17-H), 2.88 (s, 6H, N(CH3)2), 3.21
(m, 2H, NCH2), 4.05 (m, 3H, 3-H and OCH2), 6.87 (d, 0.1H, J
) 8.1, CHdN (Z) isomer), 7.56 (d, 0.9H, J ) 9.3, CHdN (E)
isomer). MS: m/z 420 (3, M+ base), 302 (70), 71 (100).
(E)-17â-[[(4-Dim eth yla m in obu toxy)im in o]m eth yl]-5â-
a n d r osta n e-3â,14â-d iol oxa la te (6) was prepared following
method A. The crude product was purified by chromatography

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2343
1
precipitate was collected to give 12 (0.27 g), a white solid. H
NMR (CD3OD): δ 0.94 (s, 3H, CH3), 0.97 (s, 3H, CH3), 2.93 (s,
3.6H, N(CH3)2 (E) isomer), 2.94 (s, 2.4H, N(CH3)2 (Z) isomer),
3.40-3.50 (m, 2H, CH2N), 4.04 (m, 1H, 3-H), 4.25-4.40 (m,
2H, OCH2), 6.80 (d, 0.4H, J ) 5.6, CHdN (Z) isomer), 7.50 (d,
0.6H, J ) 6.1, CHdN (E) isomer). MS: m/z 434 (2.8, M+ base),
203 (45), 58 (100).
(E,E)-17â-[3-[(2-Am in oet h oxy)im in o]-1-p r op en yl]-5â-
a n d r osta n e-3â,14â-d iol oxa la te (19) was prepared following
method B. The crude product on chromatography with CHCl3/
MeOH/26% NH4OH (89:10:1) gave 19 (0.58 g) as a white foam.
1H NMR (CD3OD): δ 0.84 (s, 3H, CH3), 0.97 (s, 3H, CH3), 2.29
(m, 1H, 17-H), 2.86 (t, 2H, NCH2), 4.03 (t, 2H, OCH2), 4.04
(m, 1H, 3-H), 5.87 (dd, 1H, J ) 10.0, 15.6, CHCdN), 6.30 (dd,
1H, J ) 10.0, 15.6, CHdCCdN), 6.99 (d, 0.04H, J ) 7.5, CHd
N (Z) isomer), 7.74 (d, 0.96H, J ) 10.0, CHdN (E) isomer).
MS: m/z 404 (99, M+ base), 326 (100).
(E)-17â-[3-[(2-Am in oeth oxy)im in o]p r op yl]-5â-a n d r os-
ta n e-3â,14â-d iol (13) was prepared following method B. The
crude product on chromatography with CHCl3/MeOH/26%
NH4OH (89:10:1) gave 13 as a white solid (1.05 g). 1H NMR
(DMSO-d6): δ 0.82 (s, 3H, CH3), 0.86 (s, 3H, CH3), 2.69 (m,
2H, NCH2), 3.63 (m, 1H, 3-H), 3.84 (t, J ) 7.5, 1.8H, OCH2
(E) isomer), 3.90 (t, J ) 7.5, 0.2H, OCH2 (Z) isomer), 6.68 (d,
0.1H, J ) 7.5, CHdN (Z) isomer), 7.38 (d, 0.9H, J ) 10.0, CHd
N (E) isomer). MS: m/z 389 (20, M+-17), 326 (100).
(E,E)-17â-(3-Met h oxyim in o-2-m et h yl-1-p r op en yl)-5â-
a n d r osta n e-3â,14â-d iol (20) was prepared following method
A. The crude product on chromatography with n-hexane/EtOAc
1
(8:2) gave 20 (0.17 g) as a white foam. H NMR (CD3OD): δ
0.84 (s, 3H, CH3), 0.97 (s, 3H, CH3), 1.75 (s, 3H, dCCH3), 2.71
(m, 1H, 17-H),) 4.04 (m, 1H, 3-H), 5.90 (d, 1H, J ) 10, 20-H),
7.61 (s, 1H, CHdN). MS: m/z 389 (4, M+), 94 (100).
(E,E)-17â-[3-(Meth oxyim in o)-1-pr open yl]-5â-an dr ostan e-
3â,14â-d iol (14) was prepared following method B. The crude
product was purified by chromatography with n-hexane/EtOAc
(8:2) followed by crystallization from EtOAc to give 14 (1.15
g) as a white solid.1H NMR (DMSO-d6): δ 0.72 (s, 3H, CH3),
0.96 (s, 3H, CH3), 2.19 (m, 1H, 17-H), 3.72 (s, 3H, OCH3), 3.87
(m, 1H, 3-H), 3.97 (s, 1H, 14-OH), 4.17 (d, 3-OH), 5.81 (dd,
1H, J ) 9.8, 15.5, H-21), 6.27 (dd, 1H, J ) 10.2, 15.5, H-20),
7.72 (d, 1H, J ) 9.8, H-22). MS: m/z 375 (11.7, M+), 94 (100).
(E,E)-17â-(3-P r op oxyim in o-2-m et h yl-1-p r op en yl)-5â-
a n d r osta n e-3â,14â-d iol (21) was prepared following method
A. The crude product on chromatography with n-hexane/EtOAc
1
(8:2) gave 21 (0.23 g) as a white foam. H NMR (CD3OD): δ
0.84 (s, 3H, CH3), 0.94 (t, 3H, CH3), 0.97 (s, 3H, CH3), 1.75 (s,
3H, CH3), 2.71 (m, 1H, 17-H), 3.92 (t. 2H, OCH2), 4.04 (m, 1H,
3-H), 5.90 (d, 1H, J ) 10 20-H), 7.61 (s, 1H, CHdN). MS: m/z
417 (9, M+), 143 (100).
(E,E)-17â-[3-(P r opoxyim in o)-1-pr open yl]-5â-an dr ostan e-
3â,14â-d iol (15) was prepared following method B. The crude
product on chromatography with n-hexane/EtOAc (8:2) gave
(E,E)-17â-[3-(2-Hyd r oxyeth oxy)im in o-2-m eth yl-1-p r o-
p en yl]-5â-a n d r osta n e-3â,14â-d iol (22) was prepared fol-
lowing method A. The crude product was purified twice by
chromatography with hexane/CHCl3/acetone, 36:32:32 and
n-hexane/CHCl3/acetone, 46:27:27 to give the pure 22 (0.38 g).
1H NMR (CDCl3): δ 0.86 (s, 3H, CH3), 0.98 (s, 3H, CH3), 1.09
(s, 1H, 14â-OH), 1.67 (d, J ) 0.9, CH3-C)), 2.71 (m, 1H, 17-
H), 3.56 (t, 2H, CH2OH), 3.97 (t, 2H, CH2ON), 5.96 (d, 1H,
J ) 5.5, H-20), 7.67 (s, 1H, CHdN). MS: m/z 419 (1.5, M+),
94 (100).
1
15 as a white foam (0.2 g). H NMR (CD3OD): δ 0.84 (s, 3H,
CH3), 0.94 (t, 3H, CH3), 0.97 (s, 3H, CH3), 2.31 (m, 1H, 17-H),
3.92 (t, 2H, OCH2), 4.04 (m, 1H, 3-H), 5.90 (dd, 1H, J ) 10.0,
15.6, CHCdN), 6.37 (dd, 1H, J ) 10.3, 15.6, CHCdCdN), 7.68
(d, 1H, J ) 10.0, CHdN). MS: m/z 403 (15, M+), 203 (83), 94
(100).
(E,E)-17â-[3-[(2-Hyd r oxyeth oxy)im in o]-1-p r op en yl]-5â-
a n d r osta n e-3â,14â-d iol (16) was prepared following method
B. The crude product on crystallization from EtOAc gave 16
(E,E)-17â-[3-(2-Dim eth yla m in oeth oxy)im in o-2-m eth yl-
1-p r op en yl]-5â-a n d r osta n e-3â,14â-d iol fu m a r a te (23) was
prepared following method A. The crude product was purified
by chromatography with CHCl3/MeOH/26% NH4OH, 89:10:1.
The purified product as a base (0.36 g) was dissolved in EtOAc
(8 mL) and added to a solution of the stoichiometric amount
of fumaric acid in EtOH (2 mL). After standing for 4 h, the
crystals were collected to give 23 (0.41 g), a white solid. 1H
NMR (CD3OD): δ 0.83 (s, 3H, CH3), 0.97 (s, 3H, CH3), 1.77 (d,
3H, J ) 0.9, dCCH3), 2.70 (m, 1H, 17-H), 2.90 (s, 6H, N(CH3)2),
3.42 (m, 2H, NCH2), 4.05 (m, 1H, 3-H), 4.34 (m, 2H, OCH2),
6.05 (bd, 1H, J ) 10.9, CHdCCdN), 6.68 (s, 2H, fumarate),
7.77 (s, 1H, CHdN). MS: m/z 446 (4, M+ base), 203 (40), 58
(100).
1
as a white solid (280 mg). H NMR (DMSO-d6): δ 0.73 (s, 3H,
CH3), 0.86 (s, 3H, CH3), 3.54 (q, J ) 5.0, 2H, CH2OH), 3.94 (t,
J ) 5.0, 2H, OCH2), 3.88 (m, 1H, 3-H), 5.81 (dd, 1H, J ) 15.6,
9.8, 21-H), 6.26 (dd, 1H, J ) 15.6, 10.2, 20-H), 7.73 (d, 1H,
J ) 9.8, 22-H). MS: m/z 405 (3, M+), 203 (51), 79 (100).
(E,EZ)-17â-[3-[(Ca r b oxym et h oxy)im in o]-1-p r op en yl]-
5â-a n d r osta n e-3â,14â-d iol Oxa la te (17). To a solution of (E)-
3â,14â-dihydroxy-5â-pregn-20-ene-21-carboxaldehyde (41) (0.25
g, 0.72 mmol) in dioxane (3 mL) was added a solution of
2-aminooxyacetic acid hydrochloride (0.11 g, 0.86 mmol) in
water (1 mL). After 1 h the solution was diluted with water
and extracted with CHCl3 (3 × 20 mL). The organic layers were
dried over Na2SO4 and evaporated in vacuo. The residue was
(E,E)-17â-[3-(2-Am in oeth oxy)im in o-2-m eth yl-1-p r op en -
yl]-5â-a n d r osta n e-3â,14â-d iol fu m a r a te (24) was prepared
following method A. The crude product was purified by
chromatography with CHCl3/MeOH/26% NH4OH, 89:10:1. The
purified product as a base (0.27 g) was dissolved in EtOH (5
mL) and the stoichiometric amount of fumaric acid was added.
After evaporation in vacuo, the residue was triturated with
EtOAc and the solid collected to give 24 (0.32 g), a white solid.
1H NMR (CD3OD): δ 0.84 (s, 3H, CH3), 0.97 (s, 3H, CH3), 1.77
(d, 3H, J ) 1.2, dCCH3), 2.71 (m, 1H, 17-H), 3.24 (m, 2H,
NCH2), 4.04 (m, 1H, 3-H), 4.24 (m, 2H, OCH2), 6.05 (bd, 1H,
J ) 10.9, CHdCCdN), 6.68 (s, 2H, fumarate), 7.77 (s, 1H, CHd
N). MS: m/z 418 (19, M+ base), 203 (100).
1
crystallized from EtOAc to give 17 (0.18 g), a white solid. H
NMR (DMSO-d6): δ 0.74 (s, 3H, CH3), 0.86 (s, 3H, CH3), 2.23
(m, 1H, 17-H), 3.88 (bb, 1H, 14-OH), 3.88 (m, 1H, 3-H), 4.16
(bb, 1H, 3-OH), 4.47 (s, 1.5H, OCH2 (E) isomer), 4.49 (s, 0.5H,
OCH2 (Z) isomer), 5.80 (dd, 0.75H, J ) 9.9, 15.5, CHCdN (E)
isomer), 6.33 (dd, 0.75H, J ) 10.0, 15.5, CHdCCdN (E)
isomer), 6.3-6.5 (m, 0.5H, CHdCHCdN (Z) isomer), 7.13 (d,
0.25H, J ) 8.7, CHdN (Z) isomer), 7.81 (d, 0.75H, J ) 9.9,
CHdN (E) isomer), 12.7 (bb, 1H, COOH).
(E,E)-17â-[3-[(2-Dim et h yla m in oet h oxy)im in o]-1-p r o-
p en yl]-5â-a n d r osta n e-3â,14â-d iol oxa la te (18) was pre-
pared following method B. The crude product was purified by
chromatography with CHCl3/MeOH/26% NH4OH, 89:10:1. The
purified product as a base (2.75 g) was dissolved in EtOAc (30
mL) and EtOH (2 mL) and the stoichiometric amount of oxalic
acid was added. The precipitate was collected to give 18 (3.15
(E,E)-17â-[3-[(3-Am in op r op oxy)im in o]-2-m eth yl-1-p r o-
p en yl]-5â-a n d r osta n e-3â,14â-d iol oxa la te (25) was pre-
pared following method A. The crude product was purified by
chromatography with CHCl3/MeOH/26% NH4OH, 89:10:1. The
purified product as a base (0.31 g) was dissolved in EtOH (5
mL) and the stoichiometric amount of oxalic acid was added.
After evaporation in vacuo, the residue was triturated with
EtOAc and the solid collected to give 25 (0.37 g), a white solid.
1H NMR (DMSO-d6): δ 0.72 (s, 3H, CH3), 0.86 (s, 3H, CH3),
1.68 (bs, 3H, dCCH3), 2.55 (m, 1H, 17-H), 2.81 (m, 2H, NCH2),
3.87 (m, 1H, 3-H), 4.03 (m, 2H, OCH2), 5.98 (bd, 1H, J ) 10.4,
1
g), a white solid. H NMR (CD3OD): δ 0.84 (s, 3H, CH3), 0.97
(s, 3H, CH3), 2.31 (m, 1H, 17-H), 2.92 (s, 6H, N(CH3)2), 3.45
(m, 2H, NCH2), 4.04 (m, 1H, 3-H), 4.34 (m, 2H, OCH2), 5.90
(dd, 1H, J ) 10.0, 15.6, CHCdN), 6.37 (dd, 1H, J ) 10.3, 15.6,
CHdCCdN), 7.11 (d, 0.05H, J ) 7.5, CHdN (Z) isomer), 7.79
(d, 0.95H, J ) 10.0, CHdN (E) isomer). MS: m/z 432 (2, M+
base), 58 (100).

2344 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
CHdCCdN), 7.69 (s, 1H, CHdN). MS: m/z 432 (5, M+ base),
230 (60), 94 (100).
NaBH3CN (0.68 g, 9.2 mmol) the mixture was stirred for 5 h.
HCl (1 N, 15 mL) was added and the solution stirred for 30
min; 1 N NaOH was added until pH 11 was reached and the
mixture extracted with EtOAc (5 × 50 mL). The combined
organic layers were dried over Na2SO4 and evaporated to
dryness. The residue was purified by two flash chromatogra-
phies (CHCl3/MeOH/26% NH4OH, 95:5:0.5, then CHCl3/MeOH,
9:1). After evaporation to dryness of the fractions containing
the pure compound, the residue was dissolved in EtOAc (8 mL)
and the stoichiometric amount of oxalic acid added. The waxy
solid was triturated with EtOAc to give 32 (0.86 g), a white
solid. 1H NMR (CD3OD): δ 0.95 (s, 3H, CH3), 0.97 (s, 3H, CH3),
2.60 (s, 3H, ONCH3), 2.60-3.20 (m, 4H, ONCH2CH2N), 2.90
(s, 6H, N(CH3)2), 3.85 (m, 2H, OCH2), 4.04 (m, 1H, 3-H). MS:
m/z 42 (18, M+ base), 88 (100).
(E,E)-17â-[3-(4-Am in obu toxy)im in o-2-m eth yl-1-p r open -
yl]-5â-a n d r osta n e-3â,14â-d iol oxa la te (26) was prepared
following method A. The crude product was purified by
chromatography with CHCl3/MeOH/26% NH4OH, 89:10:1. The
purified product as a base (0.35 g) was dissolved in EtOH (5
mL) and the stoichiometric amount of oxalic acid was added.
After evaporation in vacuo, the residue was triturated with
EtOAc and the solid collected to give 26 (0.26 g), a white solid.
1H NMR (DMSO-d6): δ 0.71 (s, 3H, CH3), 0.86 (s, 3H, CH3),
1.68 (bs, 3H, dCCH3), 2.55 (m, 1H, 17-H), 2.78 (m, 2H, NCH2),
3.87 (m, 1H, 3-H), 3.96 (m, 2H, OCH2), 5.96 (bd, 1H, J ) 10.4,
CHdCCdN), 7.66 (s, 1H, CHdN). MS: m/z 446 (2, M+ base),
203 (35), 88 (100).
17â-[3-Met h oxya m in op r op yl]-5â-a n d r ost a n e-3â,14â-
d iol Hyd r och lor id e (33). 17â-(3-Methoxyiminopropyl)-5â-
androstane-3â,14â-diol hydrochloride (57) was prepared fol-
lowing method A. The crude product was purified by column
chromatography (n-hexane/EtOAc, 1:1) to give 57 (520 mg,
(E,E,E)-17â-[5-(2-Dim eth yla m in oeth oxy)im in o-1,3-p en -
ta d ien yl]-5â-a n d r osta n e-3â,14â-d iol oxa la te (27) was pre-
pared following method B. The crude product was purified by
chromatography with CHCl3/MeOH/26% NH4OH, 95/5/0.5. The
purified product as a base (0.31 g) was dissolved in EtOAc (8
mL) and added to a solution of the stoichiometric amount of
oxalic acid in EtOAc. After dilution with Et2O, the solid was
1
80%). H NMR (DMSO-d6): δ 0.82 (s, 3H, CH3), 0.86 (s, 3H,
CH3), 3.78 (s, 3H, OCH3), 3.98 (m, 1H, 3-H), 7.38 (d, 1H, J )
10.0, CHdN). Reduction of 57, following method C, and
crystallization of the crude from EtOAc, followed by salt
formation with the stoichiometric amount of HCl (37% solu-
1
collected to give 27 (0.33 g), a white solid. H NMR (DMSO-
d6): δ 0.73 (s, 3H, CH3), 0.87 (s, 3H, CH3), 2.14 (m, 1H, 17-H),
2.71 (s, 6H, N(CH3)2), 3.25 (m, 2H, NCH2), 3.88 (m, 1H, 3-H),
4.28 (m, 2H, OCH2), 5.90-6.70 (m, 4H, (CHdCH)2), 7.26 (d,
0.1H, J ) 8.0, CHdN (Z) isomer), 7.88 (d, 0.9H, J ) 9.9, CHd
N (E) isomer). MS: m/z 458 (3, M+ base), 58 (100).
1
tion) in EtOH/Et2O 1:2, gave 33 as a white solid (342 mg). H
NMR (DMSO-d6): δ 0.82 (s, 3H, CH3), 0.86 (s, 3H, CH3), 3.08
(m, 2H, CH2N), 3.81 (s, 3H, OCH3), 3.87 (m, 1H, 3-H). MS:
m/z 379 (0.07, M+ base), 361 (1), 330 (88).
17â-[(Met h oxya m in o)m et h yl]-5â-a n d r ost a n e-3â,14â-
d iol oxa la te (28) was prepared following method C. The crude
product was purified by chromatography with cyclohexane/
EtOAc, 65:35. The purified product as a base (100 mg) was
dissolved in EtOAc (8 mL) and added to a solution of the
stoichiometric amount of oxalic acid in EtOAc; the precipitate
was collected to give 28 (52 mg), a white solid. 1H NMR (CD3-
OD): δ 0.96 (s, 3H, CH3), 0.98 (s, 3H, CH3), 3.20-3.40 (m, 2H,
NCH2), 3.73 (s, 3H, OCH3), 4.04 (m, 1H, 3-H). MS: m/z 351
(3, M+ base), 302 (80), 94 (100).
17â-[3-(2-Am in oet h oxy)a m in op r op yl]-5â-a n d r ost a n e-
3â,14â-d iol (34) was prepared following method C. The crude
product on chromatography with CHCl3/MeOH/26% NH4OH,
1
89:10:1 gave 34 as a white solid (233 mg). H NMR (DMSO-
d6): δ 0.81 (s, 3H, CH3), 0.86 (s, 3H, CH3), 2.65 (m, 4H, 22-H
and CH2N), 3.48 (t, J ) 7.5, 2H, OCH2), 3.87 (m, 1H, 3-H).
MS: m/z 391 (10, M+ - 17), 331 (18), 89 (100).
(E)-17â-(3-Meth oxya m in o-1-p r op en yl)-5â-a n d r osta n e-
3â,14â-d iol h yd r och lor id e (35) was prepared following
method C. The crude product on chromatography with CH2-
Cl2/EtOAc followed by salt formation with the stoichiometric
amount of HCl (37% solution) in EtOH/Et2O, 1:10 gave 35 as
a white solid (200 mg). 1H NMR (DMSO-d6): δ 0.76 (s, 3H,
CH3), 0.87 (s, 3H, CH3), 3.69 (d, J ) 10.0, 1H, 22-H), 3.78 (s,
3H, OCH3), 3.98 (m, 1H, 3-H), 5.21 (dt, 1H, J ) 15.3, 6.9,
20-H), 6.07 (dt, 1H, J ) 15.3, 10.0, 21-H). MS: m/z 359 (12,
M+ - H2O base), 328 (37), 230 (8), 67 (100).
(E,E)-17â-(6-Am in oh exa -1,3-d ien yl)-5â-a n d r osta n e-3â,-
14â-d iol (36). To a solution of (E)-21-iodo-5â-pregn-20-ene-
3â,14â-diol (62) (310 mg, 0.70 mmol) and (E)-1-tributyltin-4-
azidobutene (68) (660 mg, 1.71 mmol) in degassed DMF (6 mL)
was added PdCl2(CH3CN)2 (17 mg, 0.07 mmol). After 16 h the
dark solution was diluted with a 10% KF solution (10 mL),
stirred for 15 min, filtered, and extracted with Et2O (2 × 30
mL). The combined organic extracts were washed with brine,
dried over Na2SO4, and evaporated to dryness under reduced
pressure. The crude product was purified by flash chromatog-
raphy (n-hexane/EtOAc, 7:3) to give (E,E)-17â-(6-azidohexa-
1,3-dienyl)-5â-androstane-3â,14â-diol (63, 254 mg, 88%) as a
white foam. 1H NMR (CDCl3): δ 0.82 (s, 3H, CH3), 0.96 (s,
3H, CH3), 3.45 (t, 2H, J ) 7.0, CH2N3), 4.13 (m, 1H, 3-H), 5.5-
6.1 (m, 4H, H-20, H-21, H-22, H-23).
17â-[(2-Dim eth ylam in oeth oxy)am in om eth yl]-5â-an dr os-
ta n e-3â,14â-d iol d ih yd r och lor id e (29) was prepared fol-
lowing method C. The solution of the crude product (2.30 g)
in EtOH (13.5 mL) was added with 36% HCl (0.82 mL). After
standing at 0 °C overnight, the precipitate was collected. After
drying, the solid was crystallized by dissolution in MeOH (10
mL) followed by slow addition of EtOAc (10 mL). After
standing at room temperature for 1 h and at 0 °C overnight,
the solid was collected to give 29 (1.57 g), a white solid. 1H
NMR (CD3OD): δ 0.98 (s, 3H, CH3), 0.99 (s, 3H, CH3), 2.98 (s,
6H, N(CH3)2), 3.40-3.60 (m, 2H, ONCH2), 3.58 (m, 2H, NCH2),
4.05 (m, 1H, 3-H), 4.53 (t, 2H, OCH2). MS: m/z 58 (100).
17â-[(2-Am in oeth oxy)am in om eth yl]-5â-an dr ostan e-3â,-
14â-diol dih ydr och lor ide (30) was prepared following method
C. 36% HCl (4.50 mL) was added to a solution of the crude
product (12.20 g) in EtOH (73 mL) at 45 °C. After standing at
room temperature for 1 h and at 0 °C overnight, the solid was
1
collected to give 30 (8.90 g), a white solid. H NMR (CD3OD):
δ 0.97 (s, 3H, CH3), 0.99 (s, 3H, CH3), 3.33 (t, 2H, CH2NH2),
3.40-3.60 (m, 2H, CH2NO), 4.05 (m, 1H, 3-H), 4.42 (t, 2H,
OCH2). MS: m/z 380 (0.5, M+ base), 250 (100).
17â-[(3-Am in op r op oxy)a m in om et h yl]-5â-a n d r ost a n e-
3â,14â-d iol d ih yd r och lor id e (31) was prepared following
method C. The crude product was purified by chromatography
with CHCl3/MeOH/26% NH4OH, 84:15:1.5. HCl (36%, 0.20 mL)
was added to a solution of the base (0.52 g) in EtOH/MeOH,
8:2 (6 mL). The solution was evaporated in vacuo and the solid
A solution of 63 (254 mg, 0.62 mmol) in THF (20 mL) was
treated with LiAlH4 (130 mg, 3.42 mmol) and the resulting
mixture was stirred at room temperature for 16 h. After this
time NaOH (4 N, 1 mL) was added followed by Na2SO4 (2 g).
After stirring for 1 h, the white precipitate was filtered off and
the organic solvent evaporated to dryness under reduced
pressure. The crude product was purified by flash chromatog-
raphy (CHCl3/MeOH/26% NH4OH, 89:10:1) to give 36 (72 mg,
30%) as a white foam. 1H NMR (DMSO-d6): δ 0.71 (s, 3H,
CH3), 0.86 (s, 3H, CH3), 2.54 (m, 2H, CH2NH2), 3.86 (m, 1H,
3-H), 5.4-6.0 (m, 4H, H-20, H-21, H-22, H-23). MS: m/z 387
(33, M+), 369 (22), 340 (100).
1
was triturated with Et2O to give 31 (0.57 g), a white solid. H
NMR (CD3OD): δ 0.98 (s, 3H, CH3), 0.99 (s, 3H, CH3), 3.08 (t,
2H, CH2NH2), 3.35-3.55 (m, 2H, CH2NO), 4.04 (m, 1H, 3-H),
4.29 (t, 2H, OCH2). MS: m/z 394 (0.4, M+ base), 286 (100).
17â-[N-(2-Dim eth ylam in oeth oxy)-N-m eth ylam in om eth -
yl]-5â-a n d r osta n e-3â,14â-d iol Oxa la te (32). Aqueous CH2O
(36%, 2.2 mL, 29.0 mmol) was added to a mixture of 29 as a
base (2.76 g, 5.7 mmol) in MeCN (16.5 mL). After addition of

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2345
(E,E)-17â-(7-Am in oh ep t a -1,3-d ien yl)-5â-a n d r ost a n e-
3â,14â-d iol Oxa la te (37). To a solution of (E)-21-iodo-5â-
pregn-20-ene-3â,14â-diol (62) (1.12 g, 2.52 mmol) and (E)-5-
tributyltin-4-pentenenitrile (69) (1.40 g, 3.78 mmol) in degassed
DMF (20 mL) was added PdCl2(CH3CN)2 (60 mg, 0.25 mmol).
After 2 h the dark solution is diluted with a 10% KF solution
(50 mL), stirred for 15 min, filtered, and extracted with Et2O
(3 × 50 mL). The combined organic extracts were washed with
brine, dried over Na2SO4, and evaporated to dryness under
reduced pressure. The crude product was purified by flash
chromatography (n-hexane/EtOAc, 7:3) to give (E,E)-17â-(6-
cyanohexa-1,3-dienyl)-5â-androstane-3â,14â-diol (64, 0.74 g,
residue was chromatographed with n-hexane/EtOAc, 1:1 to
give pure 40 (1.35 g, 90%), a white solid. 1H NMR (DMSO-d6):
δ 0.84 (s, 3H, CH3), 0.86) s, 3H, CH3), 3.32 (m, 1H, 14â-OH),
3.86 (m, 1H, 3-H), 4.17 (d, 1H, 3â-OH), 9.64 (t, 1H, J ) 1.7,
CHO).
(E)-3â,14â-Dih yd r oxy-5â-p r egn -20-en e-21-m et h yl-21-
ca r boxa ld eh yd e (42). To a solution of (E)-methyl 3â,14â-
dihydroxy-5â-pregn-20-ene-21-methyl-21-carboxylate12 (55) (2.20
g, 5.4 mmol) in dry THF (100 mL) at -60 °C was added neat
DIBAH (10.0 mL, 56 mmol) dropwise in 30 min. After 2 h at
-60 °C the cooling bath was removed and a solution of citric
acid (50 g) in water (400 mL) was added dropwise, the
temperature being kept below 40 °C. After stirring at room
temperature until dissolution of the gelly suspension, the
solution was extracted with EtOAc (3 × 200 mL). The
combined organic phases were washed with a 5% aqueous
solution of Na2HPO4 (200 mL), dried over Na2SO4, and
evaporated to dryness to afford (E)-17â-(3-hydroxyprop-1-en-
1-yl)-21-methyl-5â-androstane-3â,14â-diol (56) (1.96 g, 100%).1H
NMR (CDCl3): δ 0.85 (s, 3H, CH3), 0.95 (s, 3H, CH3),1.62 (s,
3H, CH3Cd), 4.01 (2H, CH2OH), 4.12 (m, 1H, 3-H), 5.59 (d,
1H, CHHCH2OH). This material showed a single spot on TLC
and was used in the next step without further purification.
A mixture of crude 56 (1.90 g, 5.2 mmol) and MnO2 (12.0 g)
in dioxane (70 mL) was stirred at room temperature for 2.5 h.
The mixture was filtered through a Celite pad by washing with
acetone and the solution evaporated to dryness to afford 42
(1.66 g, 88%) as a pale yellow solid showing a single spot on
TLC and used in the next step without further purification. A
sample was crystallized from EtOAc/Me2CO to give a white
solid. 1H NMR (CD3OD): δ 0.88 (s, 3H, CH3), 0.98 (s, 3H, CH3),
1.68 (d, 3H, J ) 1.2, CH3C)), 2.84 (m, 1H, 3-H), 4.06 (m, 1H,
3-H), 6.87 (dq, 1H, J ) 10.6, 1.2, 20-H), 9.34 (s, 1H, CHO).
(E)-17-[(2-Dim eth ylam in oeth oxy)im in o]-5â-an dr ostan e-
3â,14â-d iol (43) was prepared following method D. The crude
product on chromatography with CHCl3/MeOH/NH3, 9:1:0.5
gave 43 as a white solid (200 mg). 1H NMR (CDCl3): δ 0.98 (s,
3H, CH3), 1.17 (s, 3H, CH3), 2.30 (s, 6H, CH3N), 2.48-2.71
(m, 4H, CH2N, 16-CH2), 4.14 (t, 2H, CH2O), 4.12 (br s, 1H,
3-H). MS: m/z 392 (0.5, M+), 320 (4), 257 (6), 58 (100).
(E)-17-[(3-Dim et h yla m in op r op oxy)im in o]-5â-a n d r os-
ta n e-3â,14â-d iol (44) was prepared following method D. The
crude product on chromatography with CH2Cl2/MeOH, 9:1
gave 44 as a white solid (280 mg). 1H NMR (CDCl3): δ 0.98
(s, 3H, CH3), 1.18 (s, 3H, CH3), 2.22 (s, 6H, CH3N), 2.34 (t,
2H, CH2N), 2.48-2.71 (m, 2H, 16-CH2), 4.06 (t, 2H, CH2O),
4.12 (br s, 1H, 3-H). MS: m/z 407 (1, M+ + 1), 306 (100).
(E)-17-[(2-Dim eth ylam in obu toxy)im in o]-5â-an dr ostan e-
3â,14â-d iol (45) was prepared following method D. The crude
product on chromatography with CH2Cl2/MeOH/NH3, 95:5:1
gave 45 as a white solid (300 mg). 1H NMR (CDCl3): δ 0.98
(s, 3H, CH3), 1.18 (s, 3H, CH3), 2.22 (s, 6H, CH3N), 2.28 (t,
2H, CH2N), 2.41-2.71 (m, 2H, 16-H), 4.08 (t, 2H, CH2O), 4.16
(br s, 1H, 3-H).
1
74%) as a white foam. H NMR (CDCl3): δ 0.77 (s, 3H, CH3),
0.96 (s, 3H, CH3), 2.40 (m, 2H, CH2CN), 4.15 (m, 1H, 3-H),
5.53 (m, 1H, CHdCHCH2), 5.88 (m, 2H, dCHCH)), 6.14 (dd,
1H, J ) 14.8, 9.5, CH)CH).
A solution of (E,E)-17â-(6-cyanohexa-1,3-dienyl)-5â-andros-
tane-3â,14â-diol (64) (445 mg, 1.12 mmol) in THF (30 mL) was
treated with LiAlH4 (260 mg, 6.84 mmol) and the resulting
mixture was heated under reflux for 24 h. After cooling to room
temperature, NaOH (4 N, 2 mL) was added followed by Na2-
SO4 (4 g). After stirring for 1 h, the white precipitate was
filtered off and the organic solvent evaporated to dryness under
reduced pressure. The crude product was purified by flash
chromatography (CHCl3/MeOH/26% NH4OH, 89:10:1) to give
1
37 (200 mg, 44%) as a white foam. H NMR (CD3OD): δ 0.82
(s, 3H, CH3), 0.97 (s, 3H, CH3), 2.92 (t, 2H, CH2N), 4.04 (m,
1H, 3-H), 5.4-6.1 (m, 4H, H-20, H-21, H-22, H-23). MS: m/z
401 (26, M+ base), 354 (55), 203 (18).
17â-(7-Am in oh eptyl)-5â-an dr ostan e-3â,14â-diol (38). Ace-
tic anhydride (0.19 mL, 2.00 mmol) and 4-(N,N-dimethylami-
no)pyridine (50 mg, 0.40 mmol) were added to a solution of
(E,E)-17â-(6-cyanohexa-1,3-dienyl)-5â-androstane-3â,14â-di-
ol (64) (566 mg, 1.12 mmol) in CH2Cl2 (5 mL) and pyridine
(0.2 mL), and the mixture was stirred at room temperature
for 16 h. Ethanol (1.0 mL) was added and after 2 h the solution
was diluted with EtOAc (50 mL) and washed with a saturated
solution of NaH2PO4 (15 mL). After drying over Na2SO4, the
organic phase was evaporated to dryness to afford 600 mg of
(E,E)-3â-acetoxy-17â-(6-cyanohexa-1,3-dienyl)-5â-androstane-
14â-ol as a white foam. The crude product was dissolved in
EtOAc (25 mL) and hydrogenated for 6 h over PtO2 (60 mg,
0.26 mmol). After filtering and evaporating the solvent, the
residue was passed through a short silica gel column, eluting
with n-hexane/EtOAc 4:1, yielding 500 mg of a 4:1 mixture of
65 and an unidentified cyanoalkene. The mixture was dis-
solved in CHCl3 (25 mL) and treated with MCPBA (520 mg,
3.01 mmol). After stirring for 1 h at room temperature the
solvent was removed in vacuo and the residue partitioned
between Et2O (70 mL) and 10% aqueous Na2CO3 (20 mL). The
organic layer was separated, dried over Na2SO4, and evapo-
rated to dryness. Flash chromatography (n-hexane/EtOAc, 75:
25) afforded pure 3â-acetoxy-17â-(6-cyanohexyl)-5â-androstane-
14â-ol (65) (301 mg, 61% over three steps) as a white foam.
1H NMR (CDCl3): δ 0.93 (s, 3H, CH3), 0.97 (s, 3H, CH3), 2.08
(s, 3H, CH3COO), 2.35 (t, 2H, J ) 7.2, CH2CN), 5.04 (m, 1H,
3-H).
A solution of 65 (290 mg, 0.66 mmol) in THF (25 mL) was
treated with LiAlH4 (170 mg, 4.47 mmol) and the resulting
mixture was heated under reflux for 24 h. After cooling to room
temperature, NaOH (4 N, 1.5 mL) was added followed by Na2-
SO4 (3 g) and the mixture was stirred for 1 h. The white
precipitate was filtered off and the organic solvent evaporated
to dryness under reduced pressure. The crude product was
purified by flash chromatography (CHCl3/MeOH/26% NH4OH,
89:10:1) to give 38 (165 mg, 63%) as a white foam. 1H NMR
(DMSO-d6): δ 0.79 (s, 3H, CH3), 0.86 (s, 3H, CH3), 2.48 (m,
2H, CH2NH2), 3.87 (m, 1H, 3-H). MS: m/z 405 (11, M+), 210
(60).
(E)-17-[(2-Am in oeth oxy)im in o]-5â-a n d r osta n e-3â,14â-
d iol (46) was prepared following method D. The crude product
on chromatography with CH2Cl2/MeOH/NH3, 9:1:0.1 gave 46
as a white solid (130 mg). 1H NMR (CDCl3): δ 0.98 (s, 3H,
CH3), 1.18 (s, 3H, CH3), 2.48-2.71 (m, 2H, 16-H), 2.94 (t, 2H,
CH2N), 4.08 (t, 2H, CH2O), 4.12 (br s, 1H, 3-H).
17(E )-21[(E Z)-(2-Dim e t h yla m in oe t h oxy)im in o]-5â-
pr egn -17-en e-3â,14â-diol (47) was prepared following method
A. The crude product on chromatography with CH2Cl2/MeOH,
95:5 gave 47 as a white solid (120 mg). 1H NMR (CDCl3): δ
0.98 (s, 3H, CH3), 1.12 (s, 2.1H, CH3 (E) isomer), 1.15 (s, 0.9H,
CH3 (Z) isomer), 2.3 (s, 4.2H, CH3N (E) isomer), 2.32 (s, 1.8H,
CH3N (Z) isomer), 2.55-2.75 (m, 4H, 16-CH2 and CH2N), 4.12
(br s, 1H, 3-H), 4.13-4.23 (m, 2H, CH2O), 5.91 (m, 0.7H, Cd
CH (E) isomer), 6.4 (m, 0.3H, CdCH (Z) isomer), 7.19 (d, 0.3H,
J ) 11.2, CHdN (Z) isomer), 7.98 (d, 0.7H, J ) 11.2, CHdN
(E) isomer). MS: m/z 419 (18, M+ + 1), 71 (100).
3â,14â-Dih ydr oxy-5â-pr egn an e-21-car boxaldeh yde (40).
A mixture of (E)-3â,14â-dihydroxy-5â-pregn-20-ene-21-carbox-
aldehyde4 (41) (1.5 g) and 5% Pd/C (0.3 g) in EtOAc (200 mL)
was hydrogenated at room temperature and pressure in 1 h.
The mixture was filtered and evaporated to dryness. The
17(E )-21[(E )-3-(Dim e t h yla m in op r op oxy)im in o]-5â-
pr egn -17-en e-3â,14â-diol (48) was prepared following method
A. The crude product on chromatography with CH2Cl2/MeOH,

2346 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
9:1 gave 48 as a white solid (85 mg). 1H NMR (CDCl3): δ 0.98
(s, 3H, CH3), 1.15 (s, 2.7H, CH3 (E) isomer), 1.18 (s, 0.3H, CH3
(Z) isomer), 2.24 (s, 6H, CH3N), 2.38 (t, 2H, CH2N), 4.11 (t,
2H, CH2O), 4.13 (br s, 1H, 3-H), 5.91 (m, 0.9H, CdCH (E)
isomer), 6.4 (m, 0.1H, CdCH (Z) isomer), 7.19 (d, 0.1H, J )
11.2, CHdN (Z) isomer), 7.93 (d, 0.9H, J ) 11.2, CHdN (E)
isomer). MS: m/z 433 (2, M+ + 1), 71 (100).
3â,14â-Dih yd r oxy-5â-p r egn a n e-21-a ld eh yd e (54). To a
solution of 53 (1.10 g, 3.3 mmol) in dry THF (65 mL) at 0 °C
was added a 1 M solution of DIBAH/THF (13 mL) dropwise.
After 3 h, a solution of citric acid (0.90 g) in water (5 mL) was
added dropwise. The mixture was allowed to warm to room
temperature, stirred for 15 min, and filtered through a Celite
pad with thorough washing with EtOAc. The solution was
evaporated to dryness to give 54 (0.90 g, 81%) as a mixture of
diasteroisomeric lactols, a glassy solid, which was sufficiently
pure by TLC and 1H NMR for use in the oxime reaction. 1H
NMR (CDCl3): δ 4.13 (m, 1H + 1H, 3-H), 4.62 (m, 1H, CHOH),
5.15 (1H, CHOH).
3â,14â-Dih ydr oxy-5â-pr egn -17(20)-en e-21-aldeh yde (61).
To a mixture of nitriles 6013 (0.750 g, 2.28 mmol) and NaH2-
PO2‚H2O (1.5 g, 17 mmol) in water/acetic acid/pyridine, 1:1:2
(2.5 mL), was added Raney Ni (230 mg). After stirring at 60
°C for 24 h, the suspension was filtered on a Celite pad and
washed with warm EtOH. The organic solvent was evapored
under reduced pressure. The residue was dissolved with CH2-
Cl2, and the organic phase was washed with water, dried over
Na2SO4, and evaporated under reduced pressure. The crude
product was purified by flash chromatography on silica gel
(CH2Cl2/EtOAc, 5:2) to give 61 (0.330 g, 45%) as a white solid;
some starting material was recovered (0.18 g). 1H NMR
(CDCl3): δ 0.98 (s, 3H, CH3), 1.17 (s, 3H, CH3), 3.01-3.12 (1H,
17-H), 4.12 (br s, 1H, 3-H), 5.86 (dt, 1H, J ) 7.7, CdCH), 9.91
(d, 1H, J ) 7.7, CHO).
17(E)-21[(EZ)-(2-Am in oeth oxy)im in o]-5â-p r egn -17-en e-
3â,14â-d iol (49) was prepared following method A. The crude
product on chromatography with CHCl3/MeOH/NH3, 9:1:0.1
1
gave 49 as a white solid (60 mg). H NMR (CDCl3): δ 0.99 (s,
3H, CH3), 1.13 (s, 2.4H, CH3 (E) isomer), 1.18 (s, 0.6H, CH3
(Z) isomer), 2.97 (t, 2H, CH2N), 4.09 (t, 2H, CH2O), 4.13 (br s,
1H, 3-H), 5.91 (m, 0.8H, CdCH (E) isomer), 6.4 (m, 0.2H, Cd
CH (Z) isomer), 7.19 (d, 0.2H, J ) 11.2, CHdN (Z) isomer),
7.95 (d, 0.8H, J ) 11.2, CHdN (E) isomer). MS: m/z 390 (21,
M+), 82 (100).
17(E)-21[(EZ)-(3-Am in op r op oxy)im in o]-5â-p r egn -17-
en e-3â,14â-d iol (50) was prepared following method A. The
crude product on chromatography with CHCl3/MeOH/NH3, 9:1:
0.1 gave 50 as a white solid (80 mg). 1H NMR (CDCl3): δ 0.99
(s, 3H, CH3), 1.12 (s, 2.4H, CH3 (E) isomer), 1.13 (s, 0.6H, CH3
(Z) isomer), 2.84 (t, 2H, CH2N), 4.08-4.21 (m, 3H, CH2O and
3-H), 5.93 (m, 0.8H, CdCH (E) isomer), 6.39 (m, 0.2H, CdCH
(Z) isomer), 7.20 (d, 0.2H, J ) 11.4, CHdN (Z) isomer), 7.95
(d, 0.8H, J ) 11.4, CHdN (E) isomer). MS: m/z 404 (6, M+),
332 (100).
(E)-21-Iod o-5â-p r egn -20-en e-3â,14â-d iol (62). To a sus-
pension of CrCl2 (5.00 g, 40.68 mmol) in degassed THF (50
mL) maintained at 0 °C under argon was added a solution of
iodoform (5.25 g, 13.33 mmol) and 5â-androstane-3â,14â-diol-
17â-carboxaldehyde (39) (1.50 g, 4.69 mmol) in THF (50 mL),
over 0.5 h. The reaction mixture was stirred for 2 h at 0 °C
and filtered through a Celite pad, washing with EtOAc/1%
Et3N (100 mL). The filtrate was stirred with a saturated
solution of Na2HPO4 (300 mL) and the inorganic precipitate
was filtered off. The organic layer was separated, and the
aqueous one was extracted with EtOAc/1% Et3N (2 × 50 mL).
The combined organic phases were washed with brine, dried
over Na2SO4, and evaporated to dryness under reduced pres-
sure. The crude product was purified by flash chromatography
(n-hexane/EtOAc, 7:3) to give 62 (1.30 g, 66%) as white
crystals, mp 167-169 °C. 1H NMR (CDCl3): δ 0.87 (s, 3H,
CH3), 0.97 (s, 3H, CH3), 4.13 (m, 1H, 3-H), 5.78 (d, 1H, J )
14.8, CHdCHI), 6.75 (dd, 1H, J ) 14.8, 9.7, CH)CHI). Anal.
(C21H33IO2) C, H, I.
21-Nitr o-5â-p r egn a n e-3â,14â,20(R)-tr iol (51). A mixture
of 3â,14â-dihydroxy-5â-androstane-17â-carboxaldehyde7 (39)
(3.00 g, 9.3 mmol), nitromethane (1.0 mL, 18.7 mmol), and KF‚
2H2O (0.44 g, 4.7 mmol) in i-PrOH (30 mL) was stirred for 16
h. The solvent was removed in vacuo and the residue diluted
with CH2Cl2 (100 mL) and washed with water (30 mL). After
drying over Na2SO4, the organic phase was evaporated to
dryness, affording 51 (3.50 g, 99%) as a white solid, mp 91-
92 °C, pure by TLC (cyclohexane/EtOAc, 1:1) and 1H NMR.
1H NMR (CDCl3): δ 0.98 (s, 3H, CH3), 1.11 (s, 3H, CH3), 4.13
(m, 1H, 3-H), 4.26 (dd, 1H, J ) 4.3, 11.4), 4.43 (dd, 1H, J )
8.6, 11.4), 4.52 (ddd, 1H, J ) 1.0, 4.3, 8.6). Anal. (C21H35NO5‚
0.5H2O) C, H, N, H2O.
3â,14â-Dih yd r oxy-5â-p r egn a n e-21-n itr ile (53). To a so-
lution of 51 (3.45 g, 9.0 mmol) in THF (70 mL) were added
Ac2O (3.1 mL, 32.8 mmol) and 4-(N,N-dimethylamino)pyridine
(20 mg). After stirring for 24 h, the solvent was evaporated.
The residue was dissolved in EtOH/Et2O, 1:1 (200 mL), and
NaBH4 (0.70 g, 18.5 mmol) was added. After 1 h, the solution
was poured into 5% aqueous NaH2PO4 (150 mL) and the
mixture extracted with CH2Cl2 (3 × 100 mL). After drying over
Na2SO4, the organic phase was evaporated to dryness to afford
3â-acetoxy-21-nitro-5â-pregnan-14â-ol (52) (3.55 g, 97%) as a
white foam sufficiently pure by TLC (n-hexane/EtOAc, 6:4) for
use in the following reaction. 1H NMR (CDCl3): δ 0.96 (s, 6H,
CH3), 2.05 (s, 3H, OCH3), 2.35 (m, 1H, 17-H), 4.35 (m, 2H, CH2-
NO2), 5.09 (m, 1H, 3-H).
(E)-1-(Tr i-n -b u t ylst a n n yl)-4-a zid obu t en e (68). Meth-
anesulfonyl chloride (0.60 mL, 6.12 mmol) was slowly added
at 0 °C to a solution of (E)-4-(tributylstannyl)-3-butenol (66)33
(2.00 g, 5.54 mmol) and Et3N (0.98 mL, 7.03 mmol) in CH2Cl2
(20 mL). After 1 h the solvent was evaporated in vacuo and
the residue treated with water (50 mL) and extracted with
Et2O (3 × 25 mL). The combined organic layers were washed
with brine. Evaporation of the solvent gave 2.61 g of crude
1
mesylate 67. H NMR (CDCl3): δ 0.90 (m, 9H, CH3(CH2)3Sn),
To a solution of 52 (3.45 g, 8.5 mmol) in dry CH2Cl2 at 0 °C
were added DEAD (2.4 mL, 15 mmol) and Ph3P (7.30 g, 28
mmol). After stirring at room temperature for 40 h, the
mixture was evaporated to dryness and the residue triturated
with Et2O (2 × 100 mL). The combined Et2O layers were
evaporated and the residue purified by flash chromatography
(CH2Cl2/Et2O, 9:1) to give 3â-acetoxy-14â-hydroxy-5â-preg-
nane-20-nitrile (3.0 g, 95%) as a white foam. The residue (2.95
g, 7.9 mmol) was dissolved in a mixture of THF (50 mL) and
MeOH (100 mL), and aqueous 1 N NaOH (42 mL) was added.
After stirring for 18 h, 5% aqueous NaH2PO4 (150 mL) was
added and the mixture extracted with EtOAc (3 × 100 mL).
After drying over Na2SO4, the organic phase was evaporated
to dryness and the residue purified by flash chromatography
(CHCl3/Et2O, 6:4) to give 53 (1.35 g, 52%) as a white foam. A
sample was crystallized from i-Pr2O to give a white solid, mp
1.20-1.65 (m, 18H, CH3(CH2)3Sn), 2.60 (brq, 2H, J ) 7.0,
dCH-CH2), 3.00 (s, 3H, SO2CH3), 4.29 (t, 2H, J ) 7.0, CH2-
OSO2CH3), 5.90 (dt, 1H, J ) 18.7, 5.6, dCHCH2), 6.13 (d, 1H,
J ) 18.7, dCHSn). NaN3 (1.78 g, 27.4 mmol) was added to a
DMF (10 mL) solution of crude 67 and the mixture stirred for
4 h at 50 °C. After this time, water (70 mL) was added and
the product was extracted with Et2O/hexane, 1:1 (4 × 25 mL).
The combined organic layers were washed with brine and
evaporated under reduced pressure to give 850 mg of crude
68 (72%, over two steps) which was used without purification
1
in the next cross-coupling step. H NMR (CDCl3): δ 0.90 (m,
9H, CH3(CH2)3Sn), 1.20-1.65 (m, 18H, CH3(CH2)3Sn), 2.55
(brq, 2H, J ) 7.0, dCHCH2), 3.49 (t, 2H, J ) 7.0, CH2N3), 5.92
(dt, 1H, J ) 18.7, 5.6, dCHCH2), 6.13 (d, 1H, J ) 18.7,
dCHSn).
1
160-162 °C. H NMR (CDCl3): δ 0.98 (s, 6H, CH3), 2.21 (m,
(E)-5-(Tr i-n -bu tylsta n n yl)-4-p en ten en itr ile (69). (E)-4-
(Tributylstannyl)-3-butenol (66) (1.00 g, 2.77 mmol) was
converted to crude mesylate 67 (1.40 g) as described above.
This was dissolved in DMSO (50 mL) and treated with KCN
1H, 17-H), 2.57 (dd, 1H, J ) 12, 5, CHHCN), 2.64 (dd, 1H,
J ) 12, 8, CHHCN), 4.13 (m, 1H, 3-H). Anal. (C21H33NO2‚
0.75H2O) C, H, N, H2O.

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2347
(510 mg, 7.83 mmol) and KI (210 mg, 1.26 mmol). After 9 h at
80 °C, the reaction mixture was diluted with water (300 mL)
and brine (100 mL) and extracted with Et2O (3 × 75 mL). The
combined organic layers were washed with brine and evapo-
rated under reduced pressure, and the residue was purified
on silica gel (pentane/EtOAc, 95:5) to afford 1.48 g of nitrile
d6): δ 2.00-2.10 (m, 2H, CH2CH2N), 2.72 (s, 6H, CH3), 3.08
(t, 2H, CH2N), 4.11 (t, 2H, CH2O). Anal. (C5H14N2O‚2HCl) C,
H, N, Cl.
3-(Am in ooxy)p r op a n a m in e Dih yd r och lor id e (79). The
crude material was crystallized from MeOH/Et2O to give 79
(6.85 g, 82%), mp 200-203 °C (lit.36 mp 182-185 °C). 1H NMR
(CD3OD): δ 2.02-2.12 (m, 2H, CH2CH2N), 3.08 (t, 2H, CH2N),
4.18 (t, 2H, CH2O). Anal. (C3H10N2O‚2HCl) C, H, N, Cl.
N,N-Dim eth yl-4-(a m in ooxy)bu ta n a m in e Dih yd r och lo-
r id e (80). The crude material was crystallized from EtOH to
1
69 (72%, over two steps) as a colorless oil. H NMR (CDCl3):
δ 0.88 (m, 9H, CH3(CH2)3Sn), 1.20-1.65 (m, 18H, CH3(CH2)3-
Sn), 2.45 (m, 4H, (CH2)2CN), 5.95 (dt, 1H, J ) 18.7, 5.6, d
CHCH2), 6.10 (d, 1H, J ) 18.7, dCHSn).
1
P r epar ation of O-(ω-am in oalkyl)ben zoph en on e Oxim es
(70-74). ω-Chloroalkylamine hydrochloride (1.1 equiv) was
added, in portions, to a suspension of KOH powder (1.5 equiv)
and benzophenone oxime34 (1.0 equiv) in DMSO (0.45 M), while
the temperature was kept below 20 °C. After 3 h, the mixture
was poured into ice/water, the suspension was brought to pH
2 with 1 N HCl, and the mixture was extracted with Et2O (4×).
The aqueous layer was brought to pH 10 with KOH powder
and extracted with Et2O (4×). The combined organic layers
were dried over Na2SO4 and evaporated to dryness to give the
desired compounds, sufficiently pure for use in the next step.
give 80 (6.80 g, 94%), mp 119-124 °C. H NMR (DMSO-d6) δ
1.60-1.80 (m, 4H, CH2CH2CH2N), 2.70 (s, 6H, CH3), 3.06 (t,
2H, CH2N), 4.08 (t, 2H, CH2O). Anal. (C6H16N2O‚2HCl) C, H,
N, Cl.
4-(Am in ooxy)bu ta n a m in e Dih yd r och lor id e (81). Hy-
drazine hydrate (2.12 mL, 0.044 mol) was added to a stirred
suspension of 75 (3.26 g, 0.02 mol) in EtOH (35 mL) at room
temperature. The mixture was heated to reflux for 3 h. After
cooling at room temperature the precipitate was filtered off
and the filtrate evaporated to dryness. The residue was stirred
with 37% HCl (2.7 mL) and water (10 mL). The solution was
evaporated to dryness to give 81 (2.76 g, 98%), as an off-white
solid sufficiently pure for use in the reactions of oxime
preparation. 1H NMR (CD3OD): δ 1.80 (m, 4H, OCH2CH2CH2-
CH2N), 3.00 (m, 2H, CH2N), 4.10 (m, 2H, CH2O). A sample
was crystallized from MeOH/EtOAc, mp 135-140 °C (lit.36 mp
140-142 °C).
O-(2-Dim eth yla m in oeth yl)ben zop h en on e oxim e (70)
1
was produced in 98% yield (91.0 g). H NMR (CDCl3): δ 2.27
(s, 6H, CH3), 2.68 (t, 2H, CH2N), 4.32 (t, 2H, CH2O), 7.30-
7.53 (m, 10H, ArH).
O-(2-Am in oeth yl)ben zop h en on e oxim e (71) was pro-
duced in 82% yield (50.6 g). 1H NMR (DMSO-d6) δ 2.78 (t, 2H,
CH2N), 4.03 (t, 2H, CH2O), 7.25-7.50 (m, 10H, ArH).
O-(3-Dim eth yla m in op r op yl)ben zop h en on e oxim e (72)
was produced in 67% yield (25.0 g). 1H NMR (DMSO-d6) δ
1.70-1.82 (m, 2H, CH2CH2N), 2.17 (s, 6H, CH3), 2.30 (t, 2H,
CH2N), 4.12 (t, 2H, CH2O), 7.25-7.50 (m, 10H, ArH).
Biology. Na +,K+-ATP a se In h ibition . Na+,K+-ATPase was
isolated and purified from dog kidney according to the method
of J ørghensen.18 The inhibition of the enzyme activity was
measured as the percent of hydrolysis of 32P-ATP in the
presence and in the absence of the tested compound.19 The
concentrations able to inhibit 50% (IC50) of enzyme activity
were calculated by a nonlinear least-squares curve fitting
computer program.
Gu in ea P ig Atr ia . Isolated guinea pig left atria (from 300-
500 g male animals) were placed in 20 mL organ baths
containing a solution of the following composition (mM), NaCl,
131.6; KCl, 5.6; CaCl2, 1.8; NaH2PO4, 1.036; NaHCO3, 24.99;
glucose, 11; sucrose, 13, under 500 mg resting tension, at 32
°C. The solution was continuously bubbled with a mixture of
95% O2 and 5% CO2. The preparations were stimulated by
platinum electrodes by square-wave pulses at a frequency of
1 Hz (1 ms duration, voltage twice the treshold). After a 60
min equilibration period, cumulative concentrations of the
compounds were added, each concentration being left in
contact until the maximal response or arrhythmias were
observed.
In otr op ic a n d Toxic Effects in An esth etized Gu in ea
P igs. Male guinea pigs weighing 400-450 g were anesthetized
with urethane (1.5 g/kg ip). Body temperature was mantained
at 37 °C by a homeothermic blanket system (Harward Ap-
paratus). A microtip pressure transducer (Millar SPR-407,
Houston TX) was introduced into the left ventricle through
the right carotid artery to measure ventricular pressure (LVP).
Recordings were fed to a Gould (RS 3800) polygraph and to
an AST 486 computer and analyzed by IDAS software (Mango-
ni, Pisa, Italy). A polythene cannula (PE50) was inserted into
a jugular vein for drug infusion and the trachea was intubated
to facilitate spontaneous respiration. After a stabilization
period, the test substance was injected at the rate of 0.16 mL/
min by means of a Harvard P22 pump (Harward Apparatus)
until the animal died or up to a maximum of 90 min. The doses
inducing the maximum inotropic effect and death were deter-
mined. The compounds were dissolved in DMSO and the
resulting solution diluted with physiological solution to obtain
final solutions containing 1% DMSO. The starting concentra-
tion of the infused solution was based on the EC50 found in
the electrically driven guinea pig atria. If no effect was found,
the concentration was raised until satisfactory inotropic effect
was found or death occurred.
O-(3-Am in op r op yl)ben zop h en on e oxim e (73) was pro-
1
duced in 96% yield (13.0 g). H NMR (CDCl3): δ 1.82 (q, 2H,
CH2CH2N), 2.78 (t, 2H, CH2N), 4.28 (t, 2H, CH2O), 7.30-7.50
(m, 10H, ArH).
O-(4-Dim eth yla m in obu tyl)ben zop h en on e oxim e (74)
was produced in 84% yield (10.4 g). 1H NMR (CDCl3): δ 1.50-
1.60 (m, 2H, CH2CH2N), 1.70-1.80 (m, 2H, CH2CH2O), 2.23
(s, 6H, CH3), 2.28 (t, 2H, CH2N), 4.22 (t, 2H, CH2O), 7.25-
7.55 (m, 10H, ArH).
N-(4-P h th a lim id obu toxy)p h th a lim id e (75). K2CO3 (2.76
g, 0.02 mol) was added to a stirred solution of N-hydroxy-
phthalimide (3.26 g, 0.02 mol) and N-(4-bromobutyl)phthal-
imide (5.64 g, 0.02 mol) in DMSO (32 mL) at room tempera-
ture. The mixture was heated at 80 °C for 30 min. After cooling
at room temperature, water (50 mL) was added and the
mixture extracted with CH2Cl2 (3 × 30 mL). The organic layers
were washed with water (3 × 25 mL), dried over Na2SO4, and
evaporated to give 75 (5.77 g, 73%) as an off-white solid
sufficiently pure for use in the following reaction. 1H NMR
(CDCl3): δ 1.80-1.90 (m, 2H, OCH2CH2CH2CH2N), 1.90-2.00
(m, 2H, OCH2CH2CH2CH2N), 3.70 (t, 2H, CH2N), 4.27 (t, 2H,
CH2O), 7.70-7.80 (m, 4H), 7.80-7.90 (m, 4H).
P r ep a r a tion of Am in ooxya lk a n a m in es (76-80). A sus-
pension of O-(ω-aminoalkyl)benzophenone oxime (1.0 equiv)
in 6 N HCl (10 equiv) was heated at reflux under stirring for
2 h. After cooling at room temperature, the mixture was
extracted with Et2O (4×). The aqueous phase was evaporated
to dryness.
N,N-Dim eth yl-2-(a m in ooxy)eth a n a m in e Dih yd r och lo-
r id e (76). The crude material was crystallized from EtOH/
H2O, 9:1 to give 76 (38.2 g, 72%), mp 180-182 °C (lit.35 mp
180-182 °C). 1H NMR (DMSO-d6) δ 2.80 (s, 6H, CH3), 3.42 (t,
2H, CH2N), 4.43 (t, 2H, CH2O).
2-(Am in ooxy)eth a n a m in e Dih yd r och lor id e (77). The
crude material was crystallized from EtOH/H2O, 9:1 to give
1
77 (12.9 g, 72%), mp 203-206 °C (lit.36 mp 181-183 °C). H
NMR (DMSO-d6) δ 3.12 (t, 2H, CH2N), 3.28 (t, 2H, CH2O).
Anal. (C2H8N2O‚2HCl) C, H, N, Cl.
N,N-Dim et h yl-3-(a m in ooxy)p r op a n a m in e Dih yd r o-
ch lor id e (78). The crude material was crystallized from EtOH
to give 78 (13.6 g, 95%), mp 123-128 °C. 1H NMR (DMSO-
Molecu la r Mod elin g. All molecular modeling, including
energy minimization and molecular dynamics calculations,
were performed on a Silicon Graphics 35GT workstation

2348 J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12
Cerri et al.
(6) Repke, K. R. H.; Sweadner, K. J .; Weiland, J .; Megges, R.; Scho¨n,
R. In Search of Ideal Inotropic Steroids: Recent Progress. Progr.
Drug Res. 1996, 47, 9-52.
(7) Gobbini, M.; Torri, M. An Expeditious Route to 3â,14â-Dihy-
droxy-5â-androstane-17â-carboxaldehyde. Synth. Commun. 1997,
27, 1115-1122.
running Insight97&Discover (Biosym/Molecular Simulations,
San Diego, CA) software package. Molecular models of the
digitalis analogues were constructed starting from the residue
library included in the software and energy minimized, using
a conjugate gradient algorithm.
The coiled structure for the transmembrane H1 and H2
domains of the human cardiac Na+,K+-ATPase R1-subunit was
modeled on the basis of the available X-ray coordinate set of
bacteriorhodopsin protein deposited in the Protein Brookhaven
Databank (pdb code 2BRD). The extend and position of the
transmembrane helices was checked by means of the PHD
prediction28 available on the Web at the European Molecular
Biological Laboratory (EMBL, Heidelberg, Germany) Biocom-
puting Unit.
The results were found consistent with the reported second-
ary structure found in Repke.6 Side-chain/side-chain interac-
tions between the two helical domains were initially energy
minimized by molecular mechanics keeping constrained all the
backbone atoms coordinates.
(8) Lindig, C.; Repke, K. R. H. Cardioactive steroids. 24. Mechanism
of the base-catalyzed reactions of cardenolides. II. Effect of
temperature. Tetrahedron 1972, 28, 1859-1866.
(9) Templeton, J . F.; Ling, Y.; Marat, K.; La Bella, F. S. Synthesis
and Structure-Activity Relationships of 17â-Substituted 14â-
Hydroxysteroid 3-(R-L-Rhamnopyranoside)s: Steroids That Bind
to the Digitalis Receptor. J . Med. Chem. 1997, 40, 1439-1446.
(10) Wollenberg, R. H.; Miller, S. J . Nitroalkane Synthesis.
A
Convenient Method for Aldehyde Reductive Nitromethylation.
Tetrahedron Lett. 1978, 19, 3219-3222.
(11) Urp´ı, F.; Vilarrasa, J . New Synthetic “Tricks”. Direct Conversion
of Nitro Compounds to Nitriles. Tetrahedron Lett. 1990, 31,
7497-7498.
(12) Boutagy, J .; Thomas, R. Cardenolide Analogues II. Synthesis of
C17â-Unsaturated Esters and Derivatives. Aust. J . Pharm. Sci.
1972, NS1, 67-75.
All calculations were performed using CVFF37 as force field
with a dielectric constant value equal to 4, to take into account
the hydrophobic environment of the membrane. Atomic charges
were taken from the fragment library of Insight&Discover.
Initially, to find the best orientation of the digitalis deriva-
tives within the H1-H2, a digoxin molecule was manually
docked into the groove by interactive computer graphics taking
into account the currently established intermolecular interac-
tions: (a) Cys104-C17 lactone ring; and (b) rhamnose hy-
droxylic groups pointing toward the extracellular loop. These
interactions were used as loosened restraints (d < 3.0 A) for
the subsequent energy minimization and a 10 ps run of
molecular dynamics simulation at 300 K. Backbone atom
coordinates of the peptide chain were always kept constrained.
After the correct orientation of digoxin was established, the
ligand was replaced by compound 8 and subjected to further
energy minimization. Finally, another model of the H1-H2
helical hairpin was added to the complex, keeping preserved
the C2 symmetry of the dimer along its long axis. The whole
complex was energy minimized by another run of energy
minimization, 6 ps molecular dynamics at 300 K, final energy
minimization, keeping constrained the atomic coordinates of
the first H1-H2 helical hairpin and allowing the second H1-
H2 subunit to move freely and to arrange its orientation in
order to optimize intermolecular interactions. The model
structures involving the other analogues were analyzed by
making the appropriate changes in the C17 side-chain struc-
ture, keeping initially fixed the orientation of the steroid
moiety, and by a subsequent energy minimization.
(13) Siemann, H. J .; Langbein, G.; Richter, M. Steroids. Part XXII.
Horner-Wittig Olefinations of 14â-Hydroxy-17-ketosteroids. Z.
Chem. 1979, 19, 451-452.
(14) Backeberg, O. G.; Staskun, B. A Novel Reduction of Nitriles to
Aldehydes. J . Chem. Soc. 1962, 3961-3963.
(15) Kim, D.-K.; Gam, J .; Kim, Y.-W.; Lim, J .; Kim, H.-T.; Kim, K.
H. Synthesis and Anti-HIV-1 Activity of a Series of 1-Alkoxy-
5-alkyl-6-(arylthio)uracils. J . Med. Chem. 1997, 40, 2363-2373.
(16) Dhanak, D.; Reese, C. B. Synthesis of [6](2,4)Pyridinophanes.
J . Chem. Soc., Perkin Trans. 1 1987, 2829-2832.
(17) Uohama, M.; Tani, Y. Process for Producing Aminooxyacetic Acid
Salts. Eur. Pat. Appl. EP 306936 A2, 1989.
(18) J ørghensen, P. L. Purification and Characterization of Na+,K+-
ATPase. III. Purification from the Outer Medulla of Mammalian
Kidney after Selective Removal of Membrane Components by
Sodium Dodecylsulphate. Biochim. Biophys. Acta 1974, 356, 36-
52.
(19) Noel, F.; Godfraind, T. Heterogeneity of Ouabain Specific
Binding Sites and Sodium-Potassium ATPase Inhibition in
Microsomes from Rat Heart. Biochem. Pharmacol. 1984, 33, 47-
53.
(20) McDonough, A.; Wang, J .; Farley, R. Significance of Sodium
Pump Isoforms in Digitalis Therapy. J . Mol. Cell. Cardiology.
1995, 27, 1001-1009.
(21) Price, E. M.; Lingrel J . B. Structure-function Relationships in
the Na+,K+-ATPase R Subunit: Site-Directed Mutagenesis of
Glutamine-111 to Arginine and Asparagine-122 to Aspartic Acid
Generates a Ouabain-Resistant Enzyme. Biochemistry 1988, 27,
8400-8408.
(22) Croyle, M. L.; Woo, A. L.; Lingrel, J . B. Extensive Random
Mutagenesis Analysis of the Na+,K+-ATPase Alpha Subunit
Identifies Known and Previously Unidentified Amino Acid
Residues That Alter Ouabain Sensitivity. Implications for Oua-
bain Binding. Eur. J . Biochem. 1997, 248, 488-495.
(23) Vasilets, L. A.; Takeda, K.; Kawamura, M.; Schwarz, W.
Significance of the Glutamic Acid Residues Glu(334), Glu(959),
and Glu(960) of the Alpha Subunits of Torpedo Na+,K+ Pumps
for Transport Activity and Ouabain Binding. Biochimica et
Biophysica Acta- Biomembranes. 1998, 1368, 137-149.
(24) Schultheis, P. J .; Lingrel, J . B. Substitution of Transmembrane
Residues with Hydrogen-Bonding Potential in the Alpha Subunit
of Na+,K+-ATPase Reveals Alterations in Ouabain Sensitivity.
Biochemistry 1993, 32, 544-550.
(25) Cantley, L. G.; Zhou, X. M.; Cunha, M. J .; Epstein, J .; Cantley,
L. C. Ouabain-Resistant Transfectants of the Murine Ouabain
Resistance Gene Contain Mutations in the Alpha-Subunit of the
Na+,K+-ATPase. J . Biol. Chem. 1992, 267, 17271-17278.
(26) Repke, K. R. H.; Weiland, J .; Megges, R.; Scho¨n, R. Modeling of
the Three-dimensional Structure of the Digitalis Intercalating
Matrix in Na+,K+-ATPase Protodimer. Enzyme Inhib. 1996, 10,
147-157.
(27) Repke, K. R. H.; Megges, R.; Weiland, J .; Scho¨n, R. Location and
Properties of the Digitalis Receptor Site in Na+,K+-ATPase.
FEBS Lett. 1995, 359, 107-109.
(28) Rost, B.; Sander, C. Improved Prediction of Protein Secondary
Structure by Use of Sequence Profiles and Neural Networks.
Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 7558-7562.
(29) Grigorieff, N.; Ceska, T. A.; Downing, K. H.; Baldwin, J . M.;
Henderson, R. Electron-Crystallographic Refinement of the
Structure of Bacteriorhodopsin. J . Mol. Biol. 1996, 259 (3), 393-
421.
Ack n ow led gm en t. We thank Dr. S. De Munari and
Mr. G. Marazzi for spectral determinations. E.R. also
thanks the Italian Ministry for University and Scientific
and Technological Research for financial support.
Refer en ces
(1) Hoffmann, B. F.; Bigger, J . T., J r. Digitalis and Allied Cardiac
Glycosides. In The Pharmacological Basis of Therapeutics, 8th
ed.; Gilman, A. G., Rall, T. W., Nies, A. S., Taylor, P., Eds.;
Pergamon Press: New York, 1990; pp 814-839.
(2) The Digitalis Investigation Group. The Effect of Digoxin on
Mortality and Morbidity in Patients with Heart Failure. New
Engl. J . Med. 1997, 336, 525-533.
(3) Quadri, L.; Cerri, A.; Ferrari, P.; Folpini, E.; Mabilia, M.; Melloni,
P. Synthesis and Quantitative Structure-Activity Relationships
of 17â-Hydrazonomethyl-5â-androstane-3â,14â-diol Derivatives
That Bind to Na+,K+-ATPase Receptor. J . Med. Chem. 1996, 39,
3385-3393.
(4) Cerri, A.; Serra, F.; Ferrari, P.; Folpini, E.; Padoani, G.; Melloni,
P. Synthesis, Cardiotonic Activity, and Structure-Activity
Relationships of 17â-Guanylhydrazone Derivatives of 5â-An-
drostane-3â,14â-diol Acting on the Na+,K+-ATPase Receptor. J .
Med. Chem. 1997, 40, 3484-3488.
(5) De Munari, S.; Barassi, P.; Cerri, A.; Fedrizzi, G.; Gobbini, M.;
Mabilia, M.; Melloni, P. A New Approach to the Design of Novel
Inhibitors of Na+,K+-ATPase: 17â-Substituted Seco-D 5â-An-
drostane as Cassaine Analogues. J . Med. Chem. 1998, 41, 3033-
3040.
(30) Ceska, T. A.; Henderson, R. Analysis of High-Resolution Electron
Diffraction Patterns from Purple Membrane Labeled with
Heavy-Atoms J . Mol. Biol. 1990, 213, 539-560.

Digitalis-like Compounds with Inotropic Activity
J ournal of Medicinal Chemistry, 2000, Vol. 43, No. 12 2349
(31) The Merck Index, 12th ed.; Budavari, S., O’Neil, M. J ., Smith,
A., Heckelman, P. E., Kinneary, J . F., Eds.; Merck & Co., Inc.:
NJ , 1996; p 2846.
(32) Lo Bello, M.; Battistoni, A.; Mazzetti, A. P.; Board, P. G.;
Muramatsu, M.; Federici, G.; Ricci, G. Site-Directed Mutagenesis
of Human Glutathione Transferase P1-1. Spectral, Kinetic, and
Structural Properties of Cys-47 and Lys-54 Mutants. J . Biol.
Chem. 1995, 270, 1249-1253.
(33) Pilli, R. A.; De Andrade, C. K. Z.; Souto, C. R. O.; De Meijere, A.
Total Synthesis of 10-Deoxymethynolide, the Aglycon of the
Macrolide Antibiotic 10-Deoxymethymycin. J . Org. Chem. 1998,
63, 7811-7819.
(34) Lachman, A. Benzophenone Oxime. Organic Syntheses; Wiley:
New York, 1943; Collect Vol. II, pp 70-71.
(35) Villani, F. J .; Tavares, R. F.; Ellis, C. A. Oximino Ethers:
Dialkylaminoalkyl Derivatives. J . Pharm. Sci. 1969, 58, 138-
141.
(36) Pankaskie, M. C.; Scholtz, S. A. An Improved Synthetic Route
to Aminoxypropylamine (APA) and Related Homologues. Synth.
Commun. 1989, 19, 339-344.
(37) Lifson, S.; Hagler, A. T.; Dauber, P. Consistent Force Field
Studies of Intermolecular Forces in Hydrogen-Bonded Crystals.
1. CarboxylicAcids, Amides, and the CdO‚HsO Hydrogen
Bonds. J . Am. Chem. Soc. 1979, 29, 5111-5121. Hagler, A. T.;
Lifson, S.; Dauber, P. Consistent Force Field Studies of Inter-
molecular Forces in Hydrogen-Bonded Crystals. 2. A Benchmark
for the Objective Comparison of Alternative Force Fields. J . Am.
Chem. Soc. 1979, 29, 5122-5130. Hagler, A. T.; Dauber, P.;
Lifson, S. Consistent Force Field Studies of Intermolecular
Forces in Hydrogen-Bonded Crystals. 3. The CdO‚H-O Hydro-
gen Bond and the Analysis of the Energetics and Packing of
Carboxylic Acids. J . Am. Chem. Soc. 1979, 29, 5131-5141.
J M990627W