Indole derivatives as a new class of steroid 5α-reductase inhibitors

Article abstract of DOI:10.1021/jm9601819

A series of indole derivatives with varied substituents on the α,β- unsaturated double bond were synthesized and evaluated for their ability to inhibit rat prostatic 5α-reductase. Compounds possessing an ethyl substituent at the β-position of the double b

Full text of DOI:10.1021/jm9601819

J . Med. Chem. 1996, 39, 5047-5052
5047
In d ole Der iva tives a s a New Cla ss of Ster oid 5r-Red u cta se In h ibitor s
Hitoshi Takami,† Hirokazu Koshimura, Nobuyuki Kishibayashi, Akio Ishii, Hiromi Nonaka, Shiro Aoyama,
Hiroshi Kase, and Toshiaki Kumazawa*,‡
Pharmaceutical Research Laboratories, Kyowa Hakko Kogyo Co., Ltd., Nagaizumi, Shizuoka, 411 J apan
Received March 6, 1996X
A series of indole derivatives with varied substituents on the R,â-unsaturated double bond
were synthesized and evaluated for their ability to inhibit rat prostatic 5R-reductase.
Compounds possessing an ethyl substituent at the â-position of the double bond showed potent
inhibitory activity. Among them, (Z)-4-{2-[[3-[1-(4,4′-difluorobenzhydryl)indol-5-yl]-2-pentenoyl]-
amino]phenoxy}butyric acid (16, KF20405) showed the maximum potency with an IC50 value
of 0.48 ( 0.086 nM, which was 20-fold higher potency than 1 (MK-906). Compound 16
effectively inhibited DHT production 4 h after a 3 mg/kg oral administration. Several potent
indole derivatives, 1 and 2 ((()-ONO-3805), were tested versus rat and human isozymes.
Nonsteroidal inhibitors such as indole derivatives and 2 were 2-3 orders of magnitude less
potent for human type 2 isozyme than steroidal inhibitor 1 and expressed a significant species
deference for these isozymes.
5R-Reductase inhibitors may provide a novel thera-
peutic treatment for androgen-related disorders associ-
ated with elevated levels of dihydrotestosterone (DHT)
including the benign prostatic hyperplasia (BPH),1 skin
disorders such as acne,2 male pattern baldness,3 and
hirsutism.4 Several 5R-reductase inhibitors have been
reported,5,6 including the steroidal inhibitor MK-906 (1,
finasteride)5b and the nonsteroidal inhibitor ONO-38056a
(2) (Figure 1). At the starting point of our research to
develop a novel 5R-reductase inhibitor, finasteride had
undergone extensive clinical trials in patients with
BPH. Although the effectiveness of the steroidal 5R-
reductase inhibitor in the treatment of BPH was clini-
cally confirmed, several symptoms of sexual dysfunction
were reported.7 These adverse effects are possibly due
to the reduction in DHT; however, the roles of DHT for
normal sexual function remains to be elucidated. In
general, azasteroidal inhibitors, except for 4-N-H de-
rivative as finasteride, express the antiandrogenic
activity.5b,i To develop specific 5R-reductase inhibitors,
we designed a novel series of compounds, in which the
steroidal skeleton was excluded. In contrast to the
therapeutical use of the steroidal inhibitor (finasteride),
nonsteroidal inhibitors have not been studied clinically.
Compound 2 was a promising nonsteroidal inhibitor;
however, the structure-activity relationship (SAR) on
this compound has not been reported. Thus, we started
our research by designing compounds analogous to 2,
in which steroidal skeleton was excluded. We synthe-
sized indole derivatives to mimic the putative testoster-
one-NADPH-enzyme complex and identified 3 (KF18678),
possessing potent inhibitory activity for rat prostatic 5R-
reductase as previously reported.8 An investigation of
the SAR of this novel series revealed the key elements
as follows: (1) a defined spacial arrangement of the
functional groups at position 1 of the indole was
important for the enhancement of activity and (2)
coplanarity of the benzene ring and amide moiety was
crucial for inhibitory activity. On the basis of these
findings, further structural modifications of the indole
derivatives were performed to enhance the 5R-reductase
inhibitory activity of the series. In this paper, further
investigations are described which were focused on the
linking unit between the anilide moiety and the indole
skeleton.
Recently, two 5R-reductase isozymes were identified
in rats9 and humans,10 although the physiological and
pharmacological roles of these isozymes in BPH are yet
to be fully elucidated. This paper also investigated the
inhibitory activities of some potent indole derivatives
for these isozymes.
Ch em istr y
The starting material 5-acyl or 5-formylindoles6 were
prepared either by the general procedure described in
the preceding paper8 or by alkylation of indoline 5-al-
dehydes 5 followed by oxidation of the intermediate
alcohols as depicted in Scheme 1. The carboxylic acids
7 were prepared from 5-acetyl or formylindoles 6 by the
Horner-Emmons reaction (method A), followed by
hydrolysis with aqueous LiOH. When R2 was a larger
alkyl group than a methyl, for example, propionyl
derivatives, the Horner-Emmons procedure gave the
product in low yield under harsh reaction conditions and
the Peterson olefination reaction was employed (method
B) instead. During the course of the C-C bond forma-
tion reaction, a minor fraction of Z isomer was produced,
which was separable by silica gel chromatography, and
that material was converted into carboxylic acids 7Z.
The obtained R,â-unsaturated carboxylic acids 7 were
reacted with aniline 86a using Mukaiyama’s reagent
(method C)11 or N,N-bis(2-oxo-3-oxazolizinyl)phosphinic
chloride (BOP-Cl) (method D)12 to afford anilides 9,
which were hydrolyzed to carboxylic acids 10-27.
Resu lts a n d Discu ssion
The prepared compounds were evaluated for their
ability to inhibit rat prostatic 5R-reductase. Inhibitory
acitivity was expressed as the IC50 value or percent
inhibition at 100 nM (Table 1). In our 5R-reductase
assays, 1 and 2 showed inhibitory activities with IC50
values of 10 ( 1.8 and 2.5 ( 0.03 nM, respectively.
† Present address: Sakai Research Laboratories, Kyowa Hakko Co.,
Ltd., 1-1-53, Takasu, Sakai, Osaka 590, J apan.
‡ Present address: Kyowa Hakko Co., Ltd., 6-1 Ohtemachi 1-chome,
Chiyoda-ku, Tokyo 100, J apan.
X Abstract published in Advance ACS Abstracts, December 1, 1996.
S0022-2623(96)00181-1 CCC: $12.00 © 1996 American Chemical Society

5048 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Takami et al.
Sch em e 1a
a
(a) POCl3, DMF/ClCH2CH2Cl; (b) R2Li or R2MgBr/THF; (c) MnO2/toluene; (d) method A, (EtO)2P(O)CHR3CO2Et, NaH/THF, or method
B, Me3SiCH2CO2Et, LDA/THF; (e) LiOH/aqueous EtOH; (f) method C, 2-chloro-1-methylpyridinium iodide, NBu3/CH2Cl2, or method D,
N,N-bis(2-oxo-3-oxazolidinyl)phosphinic chloride, TEA/CH2Cl2; (g) NaOH/aqueous EtOH.
F igu r e 1.
Ta ble 1. Indole Derivatives
5R-reductase
inhibitory activityc
compd
R1
R2 R3 geometry
formulaa
C28H26N2O4
C29H28N2O4
mp (°C)
solventb
IC50 (nM)
10
11d
12
13
14d
15
3d
16
17
18
19
benzyl
benzyl
benzyl
benzyl
benzhydryl
benzhydryl
H
Me
H
Me Me
Me
Et
Me
Et
Et
n-Pr
i-Pr
H
H
Me
E
E
E
E
E
E
E
E
Z
E
E
123-132
135-137
162-170
132.5-134
158-162
76-78
IPA
IPA
IPA
TL
IPA
IPA
IPEg
IPEg
IPEg
1700 ( 250
50 ( 15
120 ( 30
23%e
C
C
29H28N2O4
30H30N2O4
H
H
H
H
H
H
H
C35H32N2O4‚0.2C3H8Of
5.6 ( 1.2
2.5 ( 1.2
3.3 ( 0.23
0.48 ( 0.086
100h
C
36H34N2O4‚0.3H2O
4,4′-difluorobenzhydryl
4,4′-difluorobenzhydryl
4,4′-difluorobenzhydryl
4,4′-difluorobenzhydryl
4,4′-difluorobenzhydryl
C35H30F2N2O4
158-162
150-152
78-80
C
C
36H32F2N2O4‚0.8H2O
36H32F2N2O4‚0.3H2O
C37H34F2N2O4‚0.3H2O
amorphous
4.9h
C37H34F2N2O4‚0.2C3H8Of‚ amorphous
75%e
0.5H2O
20
4,4′-difluorobenzhydryl
1-propylbutylj
c-Pri
Me
Et
H
H
H
H
H
H
H
H
E
E
E
E
E
E
E
E
C37H32F2N2O4
C29H35N2NaO4‚H2O
C30H38N2O4‚0.5H2O
C31H39N2NaO4‚2H2O
C31H39N2NaO4‚2H2O
C32H41N2NaO4‚2.5H2O
C30H37N2NaO4‚H2O
C31H39N2NaO4‚0.2H2O
81-89
amorphous
165-166
amorphous
amorphous
amorphous
amorphous
amorphous
IPEg
IPA
3.4h
2.3 ( 0.18
2.8 ( 0.18
25%e
21d
22
1-propylbutyl
23
1-propylbutylj
n-Pr
24d
25
1-isobutyl-3-methylbutylj Me
1-isobutyl-3-methylbutylj Et
8.0 ( 2.4
4.8h
26d
27
1-propylpentylj
1-propylpentylj
Me
Et
4.1 ( 0.94
3.6h
10 ( 1.8
2.5 ( 0.03
1 (MK-906)
2 [(()-ONO-3805]
a
b
All new compounds had C, H, N microanalyses within 0.4% of theoretical values unless otherwise noted. Solvent of recrystallization:
IPA, isopropyl alcohol; TL, toluene; IPE, isopropyl ether. c Prostates from male rats. IC50 values are means ( SE of three separate
experiments. See the preceding paper.3 e Percent inhibition at 100 nM of test compounds. f C3H8O, isopropyl alcohol. g Trituration solvent.
d
h
i
j
IC50 values represents the means of triplicate determinations in a single experiment. Cyclopropyl. Na salt.
Initially, the effect of methyl substituents on the R,â-
inhibitory activity among these compounds. Methacryl-
unsaturated double bond was investigated (10-13).
Isocrotonoylamide 11 possessing a â-methyl on the
double bond was confirmed to exhibit the most potent
amide 12 possessing an R-methyl on the double bond
showed more than twice the potency of 11, while the
replacement of this portion with an R,â-dimethyl (13)

New Class of Steroid 5R-Reductase Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5049
Ta ble 2. Torsion Angle Values of Indole Derivatives
Ta ble 3. Effects of 5a-Reductase Inhibitors on DHT
Production
dose
% conversion
dose
% convervsion
compd (mg/kg, po) of T to DHTa compd (mg/kg, po) of T to DHTa
16
22
0
3
10
30
94 ( 3
2
1
0
1
3
10
30
91 ( 4
76 ( 3*
70 ( 3**
50 ( 7***
82 ( 3
88 ( 3
56 ( 8***
54 ( 2***
torsion angle,a deg
compd R2 R3
τ
φ
θ
ψ
0
3
10
30
10
0
10
0
30
96 ( 2
a
H
H
-178 (20)
180 (34)
140 (-22) -179 (33)
140 (-22) -173 (27)
180 (1)
180 (1)
180 (1) -148 (42)
180 (10)
180 (10)
87 ( 5
0
1
3
10
30
84 ( 5
64 ( 6
49 ( 3*
50 ( 10*
39 ( 11*
b
c
Me
H
H
Me
84 ( 4
73 ( 6**
50 ( 10*
98 ( 2
d
Me Me
Me
122 (-40) -168 (22) -179 (0) -147 (43)
eb
H
162
-146
-179
170
3
85 ( 6
a
Torsion angles were calculated to convergent values by
Nemesis [V2.1]. Values in each parentheses indicated the degree
of deference from the angle observed from X-ray crystallographic
93 ( 4
62 ( 4‡‡‡
b
a
analysis. Corresponding torsion angles observed from X-ray
Rat prostate was removed and homogenized 4 h after oral
administration of test compounds. The concentrations of test-
osterone (T) and DHT in the homogenate were determined by radio
immunoassay. Values are presented as the mean (SEM of four
rats. *p < 0.05, **p < 0.01, ***p < 0.001, compared with the
control (0 mg/kg) group (Dunnet multiple comparison test). ‡‡‡p
< 0.001, compared with the control (0 mg/kg) group (Student’s
t-test).
crystallographic analysis of isocrotonoyl compound.8
and nonsubstituted (10) acrylamide remarkably reduced
activity. To investigate whether the 5R-reductase in-
hibitory activity might be related to conformational
changes as the result of differences in the substituents
on the R,â-unsaturated double bond, energy minimiza-
tion was performed on the substructure of the indole
derivatives 10-13. Compounds a -d were considered
as candidates for each analog 10-13. The central bond
of each torsion angle was defined by one of the Greek
letters τ, φ, θ, and ψ. The relative low-energy conform-
ers were calculated using the COSMIC force field13 by
Nemesis,14 and the defined angles are summarized in
Table 2. In the case of compound b, these angles were
different from the angles observed from the X-ray
crystallographic analysis of a structurally similar com-
pound (e).8 This result suggested that the crystal
structure was not the most stable conformer. In com-
parison with b, compound a was most stable almost in
a coplanar structure, whereas compound d had a
twisted form. The decreased potency of 10 and 13,
which correspond to a and d , indicated that these
conformations were not suitable for potent activity. On
the other hand, compound c can overlap quite well with
b except for the ψ angle. This result was consistent
with the somewhat less potent activity observed in
compound 12. From these results, the conformations
were shown to be dependent on the substituents on the
R,â-unsaturated double bond, and isocrotonoylamide
was indicated as the optimized substructure among
them.
Next, the investigation focused on the â-substituents
(R2) of the unsaturated double bond. To ascertain the
effects of the substituents R2, benzhydryl, difluoroben-
zhydryl, 1-propylbutyl, 1-(2-methylpropyl)-3-methylbu-
tyl, and 1-propylpentyl were selected as the R1 substit-
uents because these exhibited a potent inhibitory activity
in the parent isocrotonoylamide derivatives (14, 3, 21,
24, and 26) as previously reported.8 Within the deriva-
tives possessing a benzhydryl type substituents as R1,
introduction of an ethyl substituent at the â-position
resulted in enhanced inhibitory activity, as represented
by compounds 15 and 16. Compounds 18-20 were
equipotent with the parent compound 3 irrespective of
the substructure of the propyl substituent at the â-posi-
tion. The Z geometry of the double bond (17) apparently
reduced activity. On the other hand, among the com-
pounds possessing branched alkyl substituents at R1,
the effect of ethyl substitution at the â-position was
neutral. Compounds 22, 25, and 27, possessing an ethyl
substituent at R2, were equipotent as the methylated
compounds 21, 24, and 26, while propylated compound
23 markedly reduced the potency. These results indi-
cated that a limitation in the bulk of the R2 substituent
existed for potent rat prostatic 5R-reductase inhibitory
activity. Among the compounds possessing a â-ethyl
substituent, 16 exhibited maximum potency for rat
prostatic 5R-reductase with an IC50 value of 0.48 ( 0.086
nM, which was 20-fold more potent than 1 and 5-fold
more potent than 2.
In addition to inhibitory activity for rat prostatic 5R-
reductase in vitro, we examined the inhibitory effect of
the several potent indole derivatives, 1 and 2 on DHT
production in vivo (Table 3). Four hours after oral
dosing with test compounds at 1-30 mg/kg, the rats
were sacrificed and prostates were removed. The effect
on DHT production was determined by the previously
reported method15 and expressed as the percent conver-
sion of testosterone to DHT. In our experiments,
compounds 1 and 2 significantly inhibited the percent
conversion at 3 and 10 mg/kg, respectively. Indole
derivatives that exhibited potent activity in vitro also
inhibited DHT production in vivo. Compound 16, which
had a 20-fold more potent inhibitory activity on rat
prostatic 5R-reductase than 1, significantly inhibited
DHT production at 3 mg/kg.
Several potent indole derivatives, 1 and 2, were tested
for inhibitory effects on rat and human 5R-reductase
isozymes (Table 4). The rat type 2 isozyme was obtained
from epididymis that exhibited an optimal pH of 5.516
and human type 2 isozyme expressed in transfected
Namalwa cells with an optimal pH of 5.5.17 Indole
derivatives 3, 16, and 22 expressed an extremely potent
inhibitory acitivity on rat type 2 as the same as 2 with
IC50 values on the order of 10-10 M. Compound 1
strongly inhibited the human type 2 isozyme with an
IC50 value of 0.58 ( 0.17 nM in our assay, whereas
indole derivatives and 2 showed less potent activity with
IC50 values of 10-7 M and 78.7 ( 16.3 nM, respectively.
From the result of assays on rat and human isozymes,
it appeared that the indole derivatives and 2 expressed
an oppsite selectivity for rats and human isozymes and

5050 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Takami et al.
Ta ble 4. Effects of Indole Derivatives on Rat and Human
5R-Reductase Isozymes
aqueous sodium hydrogen carbonate and brine, dried with
MgSO4, filtered, and evaporated in vacuo. The residue was
chromatographed on silica gel, eluting with hexane-AcOEt
(2:1) to afford 5 (8.3 g, 78%) as a pale yellow oil: 1H NMR
(CDCl3) δ 3.02 (t, 2H, J ) 9.6 Hz), 3.37 (t, 2H, J ) 9.6 Hz),
5.81 (s, 1H), 6.21 (d, 1H, J ) 11.4 Hz), 6.83-7.53 (m, 9H),
7.51 (s, 1H), 9.63 (s, 1H).
1-(4,4′-Diflu or oben zh yd r yl)-5-p r op ion ylin d ole (6). To
a solution of 5 (8.8 g, 25 mmol) in 200 mL of THF was added
at -78 °C ethylmagnesium bromide (0.93 M in THF; 38 mL,
35 mmol), and the solution was stirred for 1 h. After addition
of water, the mixture was extracted with EtOAc. The organic
layer was washed with 1 N HCl, saturated sodium hydrogen
carbonate and brine successively, dried, and evaporated in
vacuo. To a solution of this residue in 200 mL of toluene was
added MnO2 (44 g, 0.50 mol), and the mixture was heated
under reflux for 1 h. The reaction mixture was cooled to room
temperature and filtered through Celite. The filtrate was
evaporated in vacuo. The residue was chromatographed on
silica gel, eluting with hexane-AcOEt (3:1) to afford 6 (6.0 g,
64%) as a colorless oil: 1H NMR (CDCl3) δ 1.21 (t, 3H, J ) 6.6
Hz), 3.01 (q, 2H, J ) 6.6 Hz), 6.62 (d, 1H, J ) 3.3 Hz), 6.83 (s,
1H), 6.85 (d, 1H, J ) 3.3 Hz), 7.00-7.40 (m, 9H), 7.70-7.90
(m, 1H), 8.34 (br, 1H).
Meth od A. (E)-3-(1-Ben zylin d ol-5-yl)-2-m eth yl-2-p r o-
p en oic Acid (7). To a suspension of NaH (60% in oil; 0.51 g,
13 mmol) in 30 mL of THF was added triethyl 2-phosphono-
propionate (2.7 mL, 13 mmol) dropwise at 0 °C. After the
solution was stirred at 0 °C for 15 min, a solution of 1-ben-
zylindole-5-carbaldehyde (1.0 g, 4.3 mmol) in 5 mL of THF was
added dropwise. After being stirred at room temperature for
1 h, the reaction mixture was added to water and then
extracted with AcOEt. The organic layer was washed with
water and brine, dried, and evaporated in vacuo. The residue
was chromatographed on silica gel, eluting with hexane-
AcOEt (4:1) to afford ethyl (E)-3-(1-benzylindol-5-yl)-2-methyl-
2-propenoate (1.25 g, 92%) as a colorless oil: 1H NMR (CDCl3)
δ 1.35 (t, 3H, J ) 7.1 Hz), 2.19 (s, 3H), 4.27 (q, 2H, J ) 7.1
Hz), 5.30 (s, 2H), 6.57 (d, 1H, J ) 3.0 Hz), 7.08-7.14 (m, 2H),
7.14 (d, 1H, J ) 3.0 Hz), 7.23-7.33 (m, 5H), 7.74 (s, 1H), 7.84
(s, 1H). A mixture of obtained ethyl ester (1.25 g, 3.91 mmol),
12 mL of 1 N LiOH, and 25 mL of 1,4-dioxane was stirred at
70-80 °C for 2 h. The mixture was evaporated in vacuo. The
residue was dissolved in 25 mL of water and acidified with 4
N HCl to pH 2. The precipitated crystals were collected by
filtration, washed, and dried to afford crude 7. This was
recrystallized from 2-propanol to give pure 7 (1.11 g, 97%) as
colorless crystals: mp 191-192 °C; IR (KBr) 3450, 1668, 1600,
1447, 1274 cm-1; 1H NMR (CDCl3) δ 2.09 (d, 3H, J ) 1.0 Hz),
5.43 (s, 2H), 6.54 (d, 1H, J ) 3.0 Hz), 7.10-7.50 (m, 8H), 7.71
(s, 2H).
IC50 (nM ( SE)a
compd
ratb epididymis
transfectedc Namalwa cell
16
22
3
0.28d
0.75d
0.21d
357d
341d
215 ( 54.6
2
1
0.18d
2.1d
78.7 ( 16.3
0.58 ( 0.17
a
IC50 values are means ( SEM of three separate experiments.
b
Rat epididymises were homogenated and assayed in vitro for
steroid 5R-reductase activity. A 10 mg sample of cellular protein
was assayed in 40 mM Tris-citrate buffer (pH4.5) in the presence
of 150 nM [14C]testosterone, 2 mM NADPH, and the indicated
concentrations of drugs. c Transfected Namalwa cells were homo-
genated and assayed in vitro for steroid 5R-reductase activity.
Then 40 µg of cellular protein were assayed in 0.1 M Na2HPO4-
citrate buffer (pH5.5) in the presence of 0.3 µM [14C]testosterone,
2 mM NADPH and the indicated concentrations of drugs.18 The
values represent the means of triplicate determinations in a single
experiment.
d
significant species difference compared with a steroidal
inhibitor 1.
Indole derivatives, such as 3 and 16 were potent
inhibitors of rat 5R-reductase with selectivity versus
other enzymes taking part in the steroidal biosynthesis.
These compounds did not show any significant activities
for 3â-hydroxysteroid dehydrogenase (3â-HSD), 17R-
and -â-hydroxysteroid dehydrogenase (17-HSD), and
aromatase even up to a concentration of 100 mM. These
indole derivatives did not show any affinity for the
androgen receptor at the same concentration.
In conclusion, the structure of the linking unit be-
tween the anilide moiety and indole skeleton was
important in determining potent rat 5R-reductase in-
hibitory activity. An ethyl instead of a methyl substitu-
ent in the â-position of 3 resulted in enhancement of
inhibitory activity for rat prostatic 5R-reductase. Com-
pound 16, which showed the maximum potency with an
IC50 value of 0.48 ( 0.086 nM, also effectively inhibited
DHT production in vivo (3 mg/kg, po, in Table 3). 5R-
Reductase assays on rat and human isozymes showed
that nonsteroidal inhibitors, such as indole derivatives
and 2, exhibited significant species differences, whereas
a selective human type 2 isozyme inhibitor 1 was
effective both for rat and human type 2 isozymes. To
provide a new agent for BPH therapy, further efforts
to discover nonsteroidal 5R-reductase inhibitors pos-
sessing potent inhibitory activity for the human type 2
isozyme are ongoing.
Meth od B. (E)-3-[1-(1-P r op ylbu tyl)in d ol-5-yl]-2-p en -
ten oic Acid (7). To a solution of diisopropylamine (4.4 mL,
31 mmol) was added dropwise at 0 °C in 50 mL of THF,
n-butyllithium (1.65 M in hexane; 19 mL, 31 mmol), and the
solution was stirred for 20 min at the same temperature. After
this solution was cooled at -78 °C, ethyl (trimethylsilyl)acetate
(4.8 mL, 26 mmol) was added, and the mixture was stirred at
same temperature for 20 min. A solution of 1-(1-propylbutyl)-
5-propionylindole (2.8 g, 11 mmol) in 14 mL of THF was added,
and the mixture was stirred at -70 to 0 °C for 2 h. After
addition of saturated aqueous NH4Cl, the mixture was ex-
tracted with AcOEt. The organic layer was washed with water
and brine, dried, and evaporated in vacuo. The residue was
chromatgraphed on silica gel, eluting with hexane-AcOEt (20:
1) to afford ethyl (E)-3-[1-(1-propylbutyl)indol-5-yl]-2-pen-
tenoate (3.2 g, 90%) as a colorless oil: 1H NMR (CDCl3) δ 0.83
(t, 6H, J ) 6.9 Hz), 1.00-1.35 (m, 4H), 1.14 (t, 3H, J ) 7.4
Hz), 1.30 (t, 3H, J ) 7.2 Hz), 1.70-2.00 (m, 4H), 3.18 (q, 2H,
J ) 7.4 Hz), 4.10-4.40 (m, 1H), 4.19 (q, 2H, J ) 7.2 Hz), 6.07
(s, 1H), 6.52 (d, 1H, J ) 3.3 Hz), 7.12 (d, 1H, J ) 3.3 Hz), 7.32
(s, 2H), 7.74 (s, 1H). The obtained ethyl ester (3.2 g, 9.4 mmol)
was hydrolyzed by the same procedure described above to
afford crude 7. This was recrystallized from diisopropyl ether
to give pure 7 (2.4 g, 82%) as pale brown crystals: mp 127-
Exp er im en ta l Section
Melting points were determined with a Bu¨chi-510 melting
point apparatus and are uncorrected. Infrared spectra (IR)
were recorded on a J ASCO IR-810 spectrometer. Proton
nuclear magnetic resonance spectra (1H NMR) were recorded
on a Hitachi R-90H (90 MHz) or a J EOL J NM GX-270 or EX-
270 (270 MHz) spectrometer with Me4Si as internal standard.
Elemental analyses were performed by the analytical depart-
ment of our laboratories.
1-(4,4′-Diflu or oben zh ydr yl)in dolin e-5-car baldeh yde (5).
To a stirred solution of POCl3 (4.30 mL, 45.9 mmol) in 10 mL
of dichloroethane was added dropwise DMF (3.5 mL, 46 mmol),
and the solution was stirred at room temperature for 1 h. After
a solution of 1-(4,4′-difluorobenzhydryl)indoline (4) (9.8 g, 31
mmol) in 20 mL of dichoroethane was added, the reaction
mixture was stirred at 50 °C for 2 h and then poured into
sodium acetate (50 g, 0.61 mol) in 300 mL of water. The
mixture was extracted with dichloromethane washed with
129 °C; IR (KBr) 2954, 1679, 1598, 1413, 1297, 1218, 723 cm-1
;

New Class of Steroid 5R-Reductase Inhibitors
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26 5051
1H NMR (CDCl3) δ 0.84 (t, 6H, J ) 6.3 Hz), 1.00-1.30 (m,
4H), 1.16 (t, 3H, J ) 7.4 Hz), 1.70-2.00 (m, 4H), 3.21 (q, 2H,
J ) 7.4 Hz), 4.30 (dq, 1H, J ) 7.4 Hz), 6.13 (s, 1H), 6.54 (d,
1H, J ) 3.3 Hz), 7.15 (d, 1H, J ) 3.3 Hz), 7.36 (br, 2H), 7.79
(br, 1H).
and the enzyme preparation (1 mg of protein) in a total volume
of 0.5 mL. The test compounds in 10 µL of ethanol were added
to the test tubes, whereas control and blank tubes received
the same volume of ethanol. The blank tubes also received 2
mL of ethyl acetate. The reaction was started with the
addition of the enzyme preparation. After incubation at 37
°C for 20 min, the control and test tubes received 2 mL of ethyl
acetate, and the reaction solution was centrifuged at 1000g
for 5 min. The ethyl acetate phase was transferred to another
tube and evaporated to dryness. The steroids were taken up
in 50 µL of ethyl acetate and chromatographed on a Whatman
Silica plate LK6DF, using ethyl acetate-cyclohexane (1:1) as
the developing solvent system. The radioactivity of [14C]T and
[14C]-5R-dihydrotestosterone (DHT) on the plate was measured
by a thin layer chromatography scanner (Aloka, J TC-601). The
rate of the conversion by the enzyme was calculated according
to the following formula: rate of the conversion (%) )
[(radioactivity of [14C]DHT)/{(radioactivity of [14C]T) + (radio-
activity of [14C]DHT)}] × 100.
Meth od C. Eth yl (E)-4-{2-[[3-[1-(1-P r op ylbu tyl)in d ol-
5-yl]-1-oxo-2-p en ten yl]a m in o]p h en oxy}bu tyr a te (9). To
a mixture of ethyl 4-(2-aminophenoxy)butyrate (2.4 g, 11
mmol), 2-chloro-1-methylpyridinium iodide (2.4 g, 9.6 mmol),
and tributylamine (4.4 mL, 19 mmol) in 48 mL of CH2Cl2 was
added at reflux a solution of (E)-3-[1-(1-propylbutyl)indol-5-
yl]-2-pentenoic acid (2.4 g, 7.7 mmol) in 12 mL of CH2Cl2, and
the mixture was stirred at reflux for 1 h. Upon cooling, the
mixture was diluted with ether, washed with water, 1 N HCl,
and brine, dried, and then evaporated in vacuo. The residue
was chromatographed on silica gel, eluting with hexane-
AcOEt (5:1) to afford 9 (3.2 g, 81%) as an oil: 1H NMR (CDCl3)
δ 0.86 (d, 6H, J ) 7.2 Hz), 1.09-1.29 (m, 4H), 1.13 (t, 3H, J )
7.1 Hz), 1.19 (t, 3H, J ) 7.5 Hz), 1.73-1.91 (m, 4H), 2.14-
2.23 (m, 2H), 2.51 (t, 2H, J ) 6.7 Hz), 3.29 (q, 2H, J ) 7.5
Hz), 4.07 (q, 2H, J ) 7.1 Hz), 4.08 (t, 2H, J ) 5.7 Hz), 4.25-
4.36 (m, 1H), 6.31 (s, 1H), 6.56 (d, 1H, J ) 3.0 Hz), 6.83-6.86
(m, 1H), 6.94-7.01 (m, 2H), 7.15 (d, 1H, J ) 3.0 Hz), 7.33-
7.41 (m, 2H), 7.79 (s, 1H), 8.05 (br, 1H), 8.50-8.60 (m, 1H).
Meth od D. Eth yl (E)-4-{2-[[3-[1-(1-P r op ylbu tyl)in d ol-
5-yl]-1-oxo-2-h exen yl]a m in o]p h en oxy}bu tyr a te (9). To a
mixture of ethyl 4-(2-aminophenoxy)butyrate (0.34 g, 1.5
mmol), (E)-3-[1-(1-propylbutyl)indol-5-yl]-2-hexenoic acid (0.25
g, 0.76 mmol), and bis(2-oxo-3-oxazolidinyl)phosphinic chloride
(0.23 g, 0.92 mmol) in 25 mL of CH2Cl2 was added triethyl-
amine (0.23 mL, 1.7 mmol) dropwise at 0 °C. After being
stirred at room temperature for 1 h, the reaction mixture was
added to water and extracted with ether. The organic layer
was washed with 1 N HCl, water, and brine successively, dried,
and evaporated in vacuo. The residue was chromatographed
on silica gel, eluting with hexane-AcOEt (4:1) to afford 9 (0.22
g, 54%) as a yellow oil: 1H NMR (CDCl3) δ 0.85 (t, 6H, J ) 7.4
Hz), 0.99-1.35 (m, 10H), 1.38-1.80 (m, 2H), 1.72-2.06 (m,
4H), 2.11-2.33 (m, 2H), 2.51 (t, 2H, J ) 6.4 Hz), 3.27 (t, 2H,
J ) 7.6 Hz), 4.07 (q, 2H, J ) 7.1 Hz), 4.08 (t, 2H, J ) 5.7 Hz),
4.25-4.37 (m, 1H), 6.31 (s, 1H), 6.56 (d, 1H, J ) 3.3 Hz), 6.77-
7.10 (m, 3H), 7.15 (d, 1H, J ) 3.3 Hz), 7.35 (s, 2H), 7.77 (s,
1H), 8.02 (br, 1H), 8.49-8.59 (m, 1H).
The rate of the inhibition by the test compound was
calculated according to the following formula: rate of the
inhibition (%) ) [1 - {(rate of the conversion in the test tube)
- (rate of the conversion in the blank tube)}/{(rate of the
conversion of the control tube) - (rate of the conversion of the
blank tube)}] × 100.
The IC50 values were calculated as the concentration that
inhibited the enzyme activity by 50%.
Ra t Typ e 2 5r-Red u cta se Assa y. The preparation of rat
epididymis particulates and the assay of 5R-reductase were
carried out according to the reported procedure.9 The epid-
idymis from male Wistar rats (200-300 g, J apan Cler),
sacrificed by cervical dislocation, were minced and homog-
enized in 10 tissue volumes of ice-cold medium A (0.32 M
sucrose, 1 mM dithiothreitol, and 20 mM sodium phosphate,
pH 6.5) using a Polytron homogenizer. The homogenate was
centrifuged at 100000g for 0.5 h at 2 °C. The resulting pellet
was washed once with medium A and resuspended in the same
medium (10-20 mg of protein/mL). The enzyme preparation
was stored at -80 °C. The reaction solution contained 1 mM
dithiothreitol, 40 mM Tris-citrate, pH 4.5, 2 mM NADPH, [14C]-
testosterone (T) (150 nM), and the enzyme preparation (10 mg
of protein) in a total volume of 0.5 mL. The test compounds
in 10 mL of ethanol were added to the test tubes, whereas
control and blank tubes received the same volume of ethanol.
The blank tubes also received 2 mL of ethyl acetate. The
reaction was started with the addition of the enzyme prepara-
tion. After incubation at 37 °C for 10 min, the control and
test tubes received 2 mL of ethyl acetate, and the reaction
solution was centrifuged at 1000g for 5 min. The ethyl acetate
phase was transferred to another tube and evaporated to
dryness. The steroids were taken up in 25 mL of ethyl acetate
and chromatographed on a Whatman Silica plate LK6DF,
using dichloromethane-diethyl ether (11:1) as the developing
solvent system. The radioactivity of [14C]T and [14C]-5R-
dihydrotestosterone (DHT) on the plate was measured by a
thin layer chromatography scanner (Fuji Film, BAS2000). The
rate of the conversion by the enzyme and the rate of the
inhibition by test compounds were calculated according to the
same formula described above.
(E )-4-{2-[[3-[1-(1-P r o p y lb u t y l)in d o l-5-y l]-1-o x o -2-
p en ten yl]a m in o]p h en oxy}bu tyr ic Acid (22). A mixture of
ethyl (E)-4-{2-[[3-[1-(1-propylbutyl)indol-5-yl]-1-oxo-2-pentenoyl]-
amino]phenoxy}butyrate (3.2 g, 6.2 mmol), 1.9 mL of 10 N
NaOH, and 30 mL of EtOH was stirred at 50 °C for 1 h. The
mixture was evaporated in vacuo, and the residue was dis-
solved in 30 mL of water. The mixture was acidified with 4 N
HCl to pH 3 and extracted with AcOEt. The organic layer
was washed with brine, dried, and evaporated in vacuo. The
residue was tritulated with diisopropyl ether to afford 22 (2.5
g, 83%) as a white powder: mp 165-166 °C; IR (KBr) 3222,
2822, 1647, 1568, 1476, 1253, 1206, 1155 cm-1
;
1H NMR
(CDCl3) δ 0.83 (t, 6H, J ) 6.6 Hz), 1.03-1.37 (m, 10H), 1.16
(t, 3H, J ) 7.5 Hz), 1.70-2.04 (m, 4H), 2.05-2.28 (m, 2H),
2.42-2.61 (m, 2H), 3.24 (q, 2H, J ) 7.5 Hz), 4.07 (t, 2H, J )
5.9 Hz), 4.22-4.34 (m, 1H), 5.70-5.90 (m, 1H), 6.16 (s, 1H),
6.54 (d, 1H, J ) 3.2 Hz), 6.75-7.14 (m, 3H), 7.28 (d, 1H, J )
3.2 Hz), 7.39 (s, 2H), 7.74 (s, 1H), 7.91 (s, 1H), 8.25-8.55 (m,
1H). Anal. (C30H38N2O4‚0.5H2O) C, H, N.
Biologica l Meth od s. Ra t Typ e 1 5r-Red u cta se Assa y.
The preparation of rat prostate particulates and the assay of
5R-reductase were carried out according to the reported
procedure.19 The ventral prostates from male Wistar rats
(200-300 g, J apan Cler), sacrificed by cervical dislocation,
were minced and homogenized in 3 tissue volumes of ice-cold
medium A (0.32 M sucrose, 1 mM dithiothreitol, and 20 mM
sodium phosphate, pH 6.5) using a Polytron homogenizer. The
homogenate was centrifuged at 140 000g for 1 h at 2 °C. The
resulting pellet was washed once with medium A and resus-
pended in the same medium (30-50 mg protein/mL). The
enzyme preparation was stored at -80 °C. The reaction
solution contains 1 mM dithiothreitol, 40 mM sodium phos-
phate, pH 6.5, 150 µM NADPH, [14C]testosterone (T) (3 µM),
Ack n ow led gm en t. We thank E. Sugimoto and K.
Namiki for their excellent technical assistance. We are
also grateful to Dr. T. Hirata for his continuous support.
Refer en ces
(1) (a) Metcalf, B. W.; Levy, M. A.; Holt, D. A. Inhibitors of Steroid
5R-Reductase in Benign Prostatic Hyperplasia, Male Pattern
Baldness and Acne. Trends Pharmacol. Sci. 1989, 10, 491-495.
(b) Geller, J . Pathogenesis and Medical Treatment of Benign
Prostatic Hyperplasia. Prostate 1989, Suppl. 2, 95-104. (c)
Isaacs, J . T.; Brendler, C. B.; Walsh, P. C. Changes in the
Metabolism of Dihydrotestosterone in the Hyperplastic Human
Prostate. J . Clin. Endocrinol. Metab. 1983, 56, 139-146.
(2) Sansone, G. L.; Reisner, R. M. Differential Rates of Conversion
of Testosterone to Dihydrotestosterone in Acne and Normal
Human Skin-A Possible Pathogenic Factor in Acne. J . Invest.
Dermatol. 1971, 56, 366-372.

5052 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 26
Takami et al.
tors of Human Steroid 5R-reductase. J . Med. Chem. 1995, 38,
13-15.
(3) Diani, A. R.; Mulholland, M. J .; Shull, K. L.; Kubicek, M. F.;
J ohnson, G. A.; Schostarez, H. J .; Brunden, M. N.; Buhl, A. E.
Hair Growth Effects of Oral Administration of Finasteride, A
Steroid 5R-Reductase Inhibitor Alone and in Combination with
Topical Minoxidil in The Balding Stumptail Macaque. J . Clin.
Endocrinol. Metab. 1992, 74, 345-350.
(7) (a) Gormley, G. J .; Stoner, E.; Bruskewitz, R. C.; Imperato-
McGinley, J .; Walsh, P. C.; McConnell, J . D.; Andriole, G. L.;
Geller, J .; Bracken, B. R.; Tenover, J . S.; Vaughan, E. D.; Pappas,
F.; Taylor, A.; Binkowitz, B.; Ng, J . The Effect of Finasteride in
Men with Benign Prostatic Hyperplasia. N. Engl. J . Med. 1992,
327, 1185-1191. (b) Rittmaster, R. S. Drug therapy; Finasteride
N. Engl. J . Med. 1994, 330, 120-125.
(4) Brooks, J . R. Treatment of Hirsutism with 5R-Reductase Inhibi-
tors. J . Clin. Endocrinol. Metab. 1986, 15, 391-405.
(5) For recent studies of steroidal inhibitors of 5R-reductase, see:
(a) Rasmusson, G. H.; Reynolds, G. F.; Utne, T.; J obson, R. B.;
Primka, R. L.; Berman, C.; Brooks, J . R. Azasteroids as Inhibi-
tors of Rat Prostatic 5R-Reductase. J . Med. Chem. 1984, 27,
1690-1701. (b) Rasmusson, G. H.; Reynolds, G. F.; Steinberg,
N. G.; Walton, E.; Patel, G. F.; Liang, T.; Cascieri, M. A.; Cheung,
A. H.; Brooks, J . R.; Berman, C. Azasteroids: Structure-Activity
Relationships for Inhibition of 5R-Reductase and of Androgen
Receptor Binding. J . Med. Chem. 1986, 29, 2298-2315. (c) Holt,
D. A.; Levy, M. A.; Ladd, D. L.; Oh, H.-J .; Erb, J . M.; Heaslip, J .
I.; Brandt, M.; Metcalf, B. W. Steroidal A-Ring Aryl Carboxylic
Acids: A New Class of Steroid 5R-Reductase Inhibitors. J . Med.
Chem. 1990, 33, 937-942. (d) Holt, D. A.; Levy, M. A.; Oh, H.-
J .; Erb, J . M.; Heaslip, J . I.; Brandt, M.; Lan-Hargest, H. Y.;
Metcalf, B. W. Inhibition of Steroid 5R-Reductase by Unsatur-
ated 3-Carboxysteroids. J . Med. Chem. 1990, 33, 943-950. (e)
Levy, M. A.; Metcalf, B. W.; Brandt, M.; Erb, J . M.; Oh, H.-J .;
Heaslip, J . I.; Yen, H.-K.; Rozamus, L. W.; Holt, D. A. 3-Phos-
phinic Acid and 3-Phosphinic Acid Steroids as Inhibitors of
5R-Reductase: Species Comparison and Mechanistic Studies.
Bioorg. Chem. 1991, 19, 245-260. (f) Haase-Held, M.; Hatzis,
M.; Mann, J . The Synthesis 4-Cyano- progesterone: A Potent
Inhibitor of the Enzyme 5R-Reductase. J . Chem. Soc., Perkin
Trans. 1 1992, 2999-3000. (g) Frye, S. V.; Haffner, C. D.;
Maloney, P. R.; Mook, R. A., J r.; Dorsey, G. F., J r.; Hiner, R. N.;
Cribbs, C. M.; Wheeler, T. N.; Ray, J . A.; Andrews, R. C.;
Batchelor, K. W.; Bramson, H. N.; Stuart, J . D.; Schweiker, S.
L.; Arnold, J . V.; Croom, S.; Bickett, D. M.; Moss, M. L.; Tian,
G.; Unwalla, R. J .; Lee, F. W.; Tippin, T. K.; J ames, M. K.;
Grizzle, M. K.; Long, J . E.; Schuster, S. V. 6-Azasteroids:
Structure-Activity Relationships for Inhibition of Type 1 and
Type 2 Human 5R-Reductase and Human Adrenal 3â-Hydroxy-
D5-steroid Dehydrogenase/3-Keto-D5-steroid Isomerase. J . Med.
Chem. 1994, 37, 2352-2360. (h) Haffner, C. An Efficient
Synthesis of 3-Pyridyl-N- oxide Steroids: Inhibitors of 5R-
Reductase. Tetrahedron Lett. 1994, 35, 1349-1352. (i) Stoner,
E. The Clinical Development of a 5R-Reductase Inhibitor,
Finasteride. J . Steroid Biochem. Mol. Biol. 1990, 37, 375-378.
(6) For recent studies of nonsteroidal inhibitors of 5R-reductase,
see: (a) Nakai, H.; Terashima, H; Arai, Y. Benzoylaminophe-
noxybutanoic Acid Derivatives. Ono Pharmaceutical Co., Ltd.
EP 0 291 245, 1988; Chem. Abstr. 1989, 110, 212384t, 708. (b)
J ones, C. D.; Audia, J . E.; Lawhorn, D. E.; McQuaid, L. A.;
Nerbauer, B. L.; Pike, A. J .; Pennington, P. A.; Stamm, N. B.;
Toomey, R. E.; Hirsch, K. S. Nonsteroidal Inhibitors of Human
Type 1 Steroid 5R-reductase. J . Med. Chem. 1993, 36, 421-423.
(c) Audia, J . E.; Lawhorn D. E.; Deeter, J . B. Synthesis of the
Individual Enantiomers of the Benzoquinolinone Human Type
1 Steroid 5R-reductase Inhibitors LY191704 and LY266111.
Tetrahedron Lett. 1993, 34, 7001-7004. (d) Abell, A. D. A Non-
steroidal Diene Acid Inhibitor of Human Type 2 Steroid 5R-
reductase. Bioorg. Med. Chem. Lett. 1994, 4, 2327-2330. (e) Holt,
D. A.; Yamashita, D. S.; Konialian-Beck, A. L.; Luengo, J . I.;
Abell, A. D.; Bergsma, D. J .; Brandt, M.; Levy, M. A. Benzophe-
none- and Indolecarboxylic Acid: Potent Type-2 Specific Inhibi-
(8) Kumazawa, T.; Takami, H; Kishibayashi, N.; Ishii, A.; Nagahara,
Y; Hirayama, N.; Obase, H. (E)-4-{2-[[3-(Indol-5-yl)-1-oxo-2-
butenyl]amino]phenoxy}butyric acid derivatives: A New Class
of Steroid 5R-Reductase Inhibitors in the Rat Prostate. 1. J . Med.
Chem. 1995, 38, 2887-2892.
(9) Normington, K.; Russel, D. W. Tissue Distribution and Kinetic
Characteristics of Rat Steroid 5R-Reductase Isozymes. J . Biol.
Chem. 1992, 267, 19548-19554.
(10) (a) Anderson, S.; Russel, D. W. Structural and Biochemical
Properties of Cloned and Expressed Human and Rat Steroid 5R-
Reductase. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 3640-3644.
(c) Anderson S.; Berman, D. M.; J enkins, E. P.; Russel, D. W.
Detection of Steroid 5R-Reductase 2 Gene in Male Pseudoher-
maphroditis. Nature 1991, 354, 159-161. (d) J enkins, E. P.;
Anderson, S.; Imperato-McGinley, J .; Russell, D. W. Genetic and
Pharmacological Evidence for More Than One Human Steroid
5R-Reductase. J . Clin. Invest. 1992, 89, 293-300.
(11) Bald, E.; Saigo, K; Mukaiyama, T. A. Facile Synthesis of
Carboxamides by Using 1-Methyl-2-halopyridinium Iodides as
Coupling Reagents. Chem. Lett. 1975, 26, 61-64.
(12) Diago-Meseguer, J .; Palomo-Coll, A. L.; Fernandez-Lizarbe, J .
R.; Zugaza-Bilbao, A. A New Reagent for Activating Carboxyl
Groups; Preparation and Reactions of N, N-Bis[2-oxo-3-oxazo-
lidinyl]phosphorodiamidic Chloride. Synthesis 1980, 547.
(13) (a) Vinter, J . G.; Davis, A.; Saunders, M. R. Strategic Approaches
to Drug Design. I. An Integrated Software Framework for
Molecular Modelling. J . Comput.-Aided Mol. Des. 1987, 1, 31-
51. (b) Morley, S. D.; Abraham, R. J .; Haworth, I. S.; J ackson,
D. E.; Saunders, M. R.; Vinter, J . G. COSMIC(90): An Improved
Molecular Mechanics Treatment of Hydro- carbons and Conju-
gated Systems. J . Comput.-Aided Mol. Des. 1991, 5, 475-504.
(14) Nemesis version 2.1; Oxford Molecular Ltd.: Magdalen Centre,
Oxford Science Park, Sandford-on-Thames, Oxford, OX4 4GA,
England, 1990-1993.
(15) Brooks, J . R.; Berman, C.; Primka, R. L.; Reynolds, G. F.;
Rasmusson, G. H. 5R-Reductase Inhibitory and Anti-androgenic
Activities of Some 4-Azasteroids in the Rat. Steroids 1986, 47,
1-19.
(16) Liang, T.; Cascieri, M. A.; Cheung, A. H.; Reynolds, G. F.;
Rasmusson, G. H. Species Deferences in Prostatic Steroid 5R-
Reductases of Rat, Dog, and Human. Endocrinology 1985, 117,
571-579.
(17) Aoyama, S.; Nonaka, H.; Kishibayashi, N.; Takami, H.; Ku-
mazawa, T.; Mizukami, T.; Ishii, A.; Kase, H. Establishment of
Enzyme Inhibition Screening System Using Cell Line Expresses
Human Steroid 5R-Reductase Type 2. Abstracts of Papers, 67th
Annual Meeting of the J apanese Biochemical Society, Osaka,
September, 1994; paper 1053.
(18) The details of the assay for human type 2 isozyme will be
published elsewhere.
(19) Liang, T.; Heiss, C. E.; Cheung, A. H.; Reynolds, G. F.; Rasmus-
son, G. H. 4-Azasteroidal 5R-Reductase Inhibitors without
affinity for the Androgen Receptor. J . Biol. Chem. 1984, 259,
734-739.
J M9601819