17-imidazolyl, pyrazolyl, and isoxazolyl androstene derivatives. Novel steroidal inhibitors of human cytochrome C17,20-lyase (P45017α)

Article abstract of DOI:10.1021/jm970337k

We recently described a number of inhibitors of P450₁₇α, the key enzyme of androgen biosynthesis. Here, we report the synthesis and activity of novel 17-imidazolyl, pyrazolyl, and isoxazolyl androstene derivatives as potential agents for the tr

Full text of DOI:10.1021/jm970337k

J . Med. Chem. 1997, 40, 3297-3304
3297
17-Im id a zolyl, P yr a zolyl, a n d Isoxa zolyl An d r osten e Der iva tives. Novel
†
Ster oid a l In h ibitor s of Hu m a n Cytoch r om e Cl7,20-Lya se (P 45017r
)
Yang-zhi Ling, J i-song Li,‡ Yang Liu, Katsuya Kato, Gregory T. Klus, and Angela Brodie*
Department of Pharmacology & Experimental Therapeutics, School of Medicine, University of Maryland,
Baltimore, Maryland 21201
Received May 21, 1997X
We recently described a number of inhibitors of P45017R, the key enzyme of androgen
biosynthesis. Here, we report the synthesis and activity of novel 17-imidazolyl, pyrazolyl, and
isoxazolyl androstene derivatives as potential agents for the treatment of prostatic cancer. A
number of 17-(4′-Imidazolyl) derivatives were prepared by condensing the corresponding 17-
ketol acetate side chain with aldehyde and ammonium hydroxide. The 17â-(4′-imidazolyl)
derivatives (2a , 2e, 4a , 4c) were found to be potent inhibitors of human testicular P45017R
,
with greater activity than ketoconazole. The juxtaposition between the imidazole ring and
the steroid D ring appears to be important in contributing inhibitory properties. Compounds
having a 17â-(2′-imidazolyl) ring (9a , 10) or a 20â-(2′-imidazolyl) ring (12), instead of the 17â-
(4′-imidazolyl) ring (2a , 4a ), are weak inhibitors. Among the 17-(4′-imidazolyl) derivatives,
introduction of the 17R-hydroxy group (4b) and 16R,17R-epoxide group (2d ) diminished potency
(2a f 2d ; lC50 66 f 430 nM; 4a f 4b; lC50 58 f 1200 nM), while the 16,17 double bond
increased the inhibitory activity by almost three times in the 5-en-3â-ol inhibitors (2a f 2e;
lC50 66 f 24 nM). There was virtually no difference in the inhibitory activity in the 4-en-3-
one inhibitors (4a f 4c; IC50 58 f 50 nM). The introduction of a methyl (2b) or phenyl group
(2c) on the 2′-position of 4′-imidazolyl ring caused a dramatic decrease in the potency. As to
modification of the A,B rings, the 3-acetate (2f, 2g) decreased the potency almost 3-fold
compared with the 3-alcohol (2e f 2f, IC50 24 f 75 nM; 2a f 2g, 66 f 199 nM) and the
conversion from the 5-en-3â-ol into the 4-en-3-one hardly affected the potency. As expected,
4c was more potent than 2e for the rat P45017R. 17-(3′-Pyrazolyl)- (14b) and 17-(5′-isoxazolyl)-
androsta-5,16-dien-3â-ol (15b) were also potent inhibitors of P45017R, whereas the 17-(2′-
imidazolyl) compound (9b) was one of the most potent inhibitor in this series. However, their
16-saturated counterparts (9a , 14a , 15a ) were weak inhibitors. The 17â-(3′-isoxazolyl)- (16)
and 17â-(5′-methyl-3′-oxazolyl)androst-5-en-3â-ol (18) were also inactive. The introduction of
a methyl or phenyl group on the nitrogen of the pyrazolyl ring of 14b [see 14c, 14d , and 14e]
also caused some loss of inhibition for P45017R. Compounds 2e, 4a , 4c, 9b, 14b, 17a and 17b
are among the most potent inhibitors of human P45017R so far reported.
Inhibitors of the enzymes involved in the key steps
of androgen synthesis have important uses in the
treatment of prostatic diseases, such as benign prostatic
hypertrophy (BPH) and prostatic cancer. P45017R is
responsible for androgenic hormone biosynthesis which
produces dehydroepiandrosterone (DHEA) and andros-
tenedione (A), the immediate precursors of testosterone
(T), from their respective precursors pregnenolone and
progesterone, in both testes and adrenals. A number
of compounds which inhibit P45017R have been
described.1-9 Of these compound, ketoconazole, an
imidazole antifungal agent, is currently used in the
treatment of patients with advanced prostatic can-
cer.10,11 However, this compound inhibits several other
cancer, this can result in adverse effects by stimulating
malignant growth of the prostate, since T will bind to
the androgen receptor in the absence of DHT.
We have recently identified several compounds which
inhibit both P45017R and 5R-R.14-16 Such compounds
could block all androgen synthesis (T, DHT, and A) and
be more effective alternatives or additions to orchiec-
tomy in treating prostate cancer patients. In the
present study, several more potent inhibitors of P45017R
were synthesized on the basis of observations that an
azole moiety could act as a ligand to bind to the iron
atom of the heme prosthetic group of the cytochrome
P-450 enzyme and form a coordinated complex. Ex-
amples of these types of compounds are several potent
aromatase inhibitors, such as fadrozole which is effec-
tive in breast cancer treatment.17 Although the detailed
mechanism of the 17R-hydroxylation and C17,20-side-
chain cleavage by P45017R is presently unclear, based
on our inhibitor studies, it appears that the C17 and C20
positions of the substrate when bound to the active site
of the enzyme must be close to the heme group of the
enzyme. Therefore, introduction of an imidazole group
or other heterocyclic group with a nitrogen lone pair of
electrons at these positions would coordinate to the iron
atom of the heme in the active site of the enzyme.18
Indeed, we have recently shown,9 using spectroscopic
P450 enzymes, is not a very strong inhibitor of P45017R
,
and is associated with significant side effects. Recently,
finasteride,12 an inhibitor of 5R-reductase (5R-R), the
key enzyme in the synthesis of the more potent andro-
gen dihydrotestosterone (DHT) from T, has been intro-
duced as a new treatment for BPH. Although finas-
teride effectively reduces DHT levels in patients,
testosterone levels are often increased.13 In prostatic
* To whom correspondence should be addressed.
† This work was supported by NIH Grant Ca 27440.
‡ Present address Pfizer Pharmaceutical Co., Groton, CT.
X Abstract published in Advance ACS Abstracts, September 15, 1997.
S0022-2623(97)00337-3 CCC: $14.00 © 1997 American Chemical Society

3298 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Ling et al.
Sch em e 1a
This improvement enables us to provide 2a in gram
amounts for further pharmacological study in vivo. 1H
NMR (in CD3OD) of the two protons (2′- and 5′-H) of
the imidazole ring appeared as a very flat signals at
6.8-7.3 and 7.5-8.0 ppm, which was probably due to
tautomerism among imidazole forms. However, its
hydrochloride salt showed a sharp peak at 7.39 and 8.81
ppm. Similarly, when 1a was condensed with acetal-
dehyde or benzaldehyde, the 2′-methyl (2b) or 2′-phenyl
(2c) compound was obtained. Compound 2d was ob-
tained from 16R,17R-epoxy-21-acetoxypregnenolone (1b).
Treatment of 1b with CrCl2 in acetic acid gave the 16-
ene 1c,21 which was condensed with formaldehyde and
ammonium hydroxide to give 16-en-17-(4′-imidazolyl)
derivative 2e. In the same way, the 3-acetate 1d
afforded 2g, together with some hydrolyzed product 3â-
ol (2a ), 21-acetoxy progesterone (3a ) gave 4a , and 17R,-
21-diol 3b gave 4b. Acetylation of 2a gave the diacetate
2h , in which only one nitrogen of the imidazole ring was
acetylated to give a single product.
OAc
O
R2
R3
R1
i
N
R4
RO
R1CHO
N
R2
1a, R = R2 = R3 = H
b, R = H, R2, R3 = epoxide
c, R = H, R2, R3 = double bond
d, R = Ac, R2, R3 = H
R3
RO
2a, R = R1 = R2 = R3 = R4 = H
b, R = R2 = R3 = R4 = H, R1 =CH3
c, R = R2 = R3 = R4 = H, R1 = Ph
d, R = R1 = R4 = H, R2 , R3 = epoxide
e, R = R1 = R1 = R4 = H, R2, R3 = double bond
f, R = Ac, R1 = R4 = H, R2, R3 = double bond
g, R = Ac, R1= R2 = R3 = R4 = H
h, R = Ac, R1 = R2 = R3 = H, R4 = Ac
Recently, Potter et al. reported that 17-(3′-pyridyl)-
androsta-5,16-dien-3â-ol was also a potent inhibitor of
iii
6
human testicular P45017R
.
We noted that they used
the 3â-O-acetate form instead of the more potent free
3â-ol form for in vivo studies, possibly because it acts
as a “prodrug” with better oral bioavailibility.22 Syn-
thesis of the 3-acetate derivative of 2e was therefore
attempted. As direct acetylation of 2e could also acety-
late the imidazole ring, as was shown for 2h , the
following approach was used (Scheme 2): 16-dehydro-
progesterone 3-acetate (5a ) was brominated with cupric
bromide in THF23 to give the 21-bromo derivative (6),
which proved to be more reactive than 21-acetoxy
derivative 1c. Condensation of 21-bromide 6 with
fomaldehyde in ammonium hydroxide was complete in
1 h to give 3-acetate 2f.
OR1
HN
N
O
R2
R3
R2
R3
ii
O
O
3a, R1 = Ac, R2 = R3 = H
b, R1 = R3 = H, R2 = OH
4a, R2 = R3 = H
b, R3 = H, R2 = OH
c, R2, R3 = double bond
a
(i) NH4OH, Cu(OAc)2, EtOH; (ii) HCHO, Cu(OAc)2, EtOH; (iii),
Al(iPrO)3, cyclohexanone, toluene.
evidence, that two of our potent P45017R inhibitors, i.e.,
3â-hydroxy-17-(1H-imidazol-1-yl)androsta-5,16-diene and
the corresponding 17-triazole each coordinates (via an
azole nitrogen) with the heme iron of the P45017R
enzyme complex. It should be noted that although two
groups6,7 have reported on 17-heteroaryl steroidal in-
hibitors of P45017R and believe that the inhibitory
property of their compounds are due (in part) to
coordination of a heteroaryl atom to the heme iron of
the enzyme complex, they have yet to provide evidence
for this phenomenon. Using this rationale, we have
designed and synthesized a new series of pregnene
derivatives with imidazole, pyrazole, isoxazole, and
oxazole groups substituted at the 17-position. This
modification has turned out to be a most effective
The 17â-(2′-imidazolyl) derivative 9a was synthesized
mainly by the method of Debus19 (Scheme 2): reduction
of the 21-acetoxy ketone 1a with LiA1H4 gave the diol
7. Although it was expected that this reduction of the
20-keto group would result in a mixture of 20R- and 20â-
alcohols, no attempt was made at separation. Diol 7
was cleaved with sodium periodate, to give the 20-
aldehyde 8a , which was condensed with glyoxal and
ammonium hydroxide in methanol to give 9a . Simi-
larly, 16-en-20-aldehyde 8b was prepared24 and then
condensed, to afford 16-en-17-(2′-immidazolyl) deriva-
tive 9b albeit in a very low yield. Treated similarly,
the 22-aldehyde 11 gave 20â-(2′-imidazole) 12. Op-
penauer oxidation of 9a with cyclohexanone and alu-
minum isopropoxide in refluxing toluene gave the 4-en-
3-one 10. Similarly, 2e was oxidized to give 4c.
strategy for producing potent inhibitors of P45017R
.
Resu lts a n d Discu ssion
17-Pyrazole 14 and 17-isoxazole 15 and 16 were also
synthesized (Scheme 3). The Claisen condensation of
the 20-one 5b with ethyl formate in NaOMe in benzene
gave the 21-formyl derivative, which existed mainly in
the keto-enol forms (13a ).25 However, the Pelc proce-
dure using pyridine gave a better quality and yield of
the product 13a and was therefore employed.26 The
keto-enol 13a was condensed with hydrazine to give
17â-(3′-pyrazole) 14a . While condensation with hy-
droxylamine hydrochloride in glacial acetic acid buffered
with sodium acetate gave 17â-(5′-isoxazole) 15a , in
acetic acid alone, a mixture of 15a and 17â-(3′-isoxazole)
16 was obtained. This mixture could not be separated
Ch em istr y. 4′-Imidazole (2a ) was synthesized by the
Weigenhagen method19 reported in a U.S. patent in
1953:20 21-acetoxypregnenolone (1a ) was condensed
with formaldehyde and ammonium hydroxide in reflux-
ing ethanol (see Scheme 1). However, the yield was poor
(less than 10%) and the product was contaminated with
starting material (1a ) making it difficult to purify and
requiring column chromatographic separation. The
procedure was therefore modified by periodically replac-
ing NH3 and formaldehyde which evaporated during
refluxing. In this way, the yield was raised to 30% and
the pure product obtained simply by recrystallization.

Novel Steroidal Inhibitors of C17,20-Lyase Cytochrome
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3299
Sch em e 2a
Br
N
O
R1
R2
N
O
i
ii
RO
5a, R = Ac, R1, R2 = double bond
b, R = R1 = R2 = H
6
2f
N
OH
OAc
N
OH
CHO
R1
R2
O
v
R1
R2
iii
iv
9a, R1, R2 = H
1a
7
8a, R1, R2 = H
b, R1, R2 = double bond
b, R1, R2 = double bond
vi
N
N
CHO
N
N
v
O
O
11
12
10
a
(i) CuBr2, THF; (ii) HCHO, NH4OH, Cu(OAc)2, EtOH; (iii) LiAlH4, THF; (iv) NAIO4, MeOH; (v) glyoxal, NH4OH, Cu(OAc)2, EtOH;
(vi) Al(iPrO)3, cyclohexanone, toluene.
Sch em e 3a
H
H
O
H
O
N
N
N
N
CH3
O
R1
R2
R3
i
ii
+
R1
R2
R1
R2
RO
5a, R = AcO, R1, R2 = double bond
b, R = R1 = R2 = H
13a, R1, R2 = H
14a, R1 = R2 = R3 = H
14e
b, R1, R2 = double bond
b, R1, R2 = double bond, R3 = H
c, R1, R2 = double bond, R3 = Ph
d, R1, R2 = double bond, R3 = Me
iii
iv
iv
CH3
N
O
O
N
O
R
N
R1
R2
+
O
15a, R1, R2 = H
16
18
17a, R = 3′-pyrazolyl
b, R = 5′-isoxazolyl
b, R1, R2 = double bond
a
(i) HCOOEt, NaOMe, Py; (ii) NH2NH2 or PhNHNH2; (iii) NH2OH, HOAc/or NaOAc; (iv), Al(iPrO)3, cyclohexanone, toluene.
by chromatography, and was treated with NaOMe. The
5′-isoxazole 15a readily formed the R-cyano ketone,26
while the 3′-isoxazole 16 remained unchanged and thus
purified. The original base treatment procedure25 did
not work in our case and was therefore modified by us.
The Doorenbos’s procedure for the condensation of 16-
en-20-one 5a with ethyl formate in benzene gave a very
low yield of 21-formyl derivatives 13b;25 the procedure
was improved by reaction in pyridine which gave a
better yield of product. 1H NMR demonstrated that the
enol form 13b was the main form present in CDCl3, as
shown by two doublets at δ 8.04 (d, J ) 3.9 Hz, 22-H)
and 5.79 (d, J ) 3.9 Hz, 21-H), while the aldehyde peak
at 9.80 ppm was hardly observed. Compound 13b was
treated with hydrazine or phenyl hydrazine to give 17-
(3′-pyrazole) 14b or 17-(l′-phenyl-5′-pyrazole) 14c as
reported.25 When 13b was condensed with methyl
hydrazine, Doorenbos reported only the isolation of 17-
(l′-methyl-5′-pyrazole) 14d ;25 we were able to separate
9% of 17 -(l′-methyl-3′-pyrazole) 14e from our reaction.
The reaction of 13b with hydroxylamine hydrochloride
(either in MeOH, or in glacial acetic acid, or buffered
with sodium acetate) yielded only the 5′-isoxazole 15b,
and the 3′-isoxazole was not found. This result shows
that, because C20 is conjugated with the ∆16 double bond
in 13b, it is less prone to nucleophilic attack than is

3300 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Ling et al.
Ta ble 1. IC50 of Steroid Compounds on Human and Rat
the case with the saturated analogue 13a . Both 14b
and 15b were oxidized by Oppenauer oxidation sepa-
rately to give the 4-en-3-one 17a or 17b.
Testicular C17,20-Lyase (P45017R
)
human
rat
inhibition
17â-(2′-Methyl-4′-oxazole) 18 was prepared by follow-
ing a similar procedure for the preparation of 2′-methyl-
steroidal(3,2-d)oxazoles reported by Ohta:27 the 21-
acetoxy ketone 1a was refluxed with NH4OAc in HOAc
and the 3-acetate formed was hydrolyzed with KHCO3
to give oxazole 18.
inhibition
a
a
IC50
,
IC50,
compound
2a
nM
%
nM
nM
%
nM
66b
91
2b
2c
400
400
26
21
4813
1500
NIc
2d
2e
430
not tested
24b
49
En zym e Assa y. The potency of the pregnene deriva-
tives as inhibitors of P45017R was evaluated in human
testicular microsomes. For most compounds, a radio-
metric assay was also used in which [21-3H]-17R-
hydroxyprenenolone was the substrate. Inhibition was
assessed by measuring the release of [3H]acetic acid
during cleavage of the C-21 side chain in the conversion
to DHEA. For some compounds, the conversion of
radiolabeled pregnenolone to 17R-hydroxypregnenolone
and DHEA by P45017R was measured in the presence
of several concentrations of test compounds, as described
previously.14 Reverse phase HPLC was employed to
separate and measure the amount of substrate and
metabolites. The activity of 17R-lyase was calculated
from the conversion of pregnenolone to DHEA. Both
assays gave comparable and reproducable results. The
activity of the control reaction was approximately 0.5%,
and the enzyme activity was linear with time for both
assay procedures. In addition, the IC50 values of the
inhibitors did not vary significantly at 90-300 µg of
protein concentration of the testicular microsomes. IC50
assays were performed at least twice and up to four
times on the more potent inhibitors. The mean values
are shown in Table 1. Some biological effects of inhibi-
tors 2a , 2e, 2g, and 4a have been reported.29 As rodent
models are most frequently used for in vivo studies, we
also determined the extent of inhibition of the com-
pounds on rat testicular microsomes. Although proges-
terone rather than pregnenolone is the preferred sub-
strate for the rat enzyme,28 [21-3H]-17R-hydroxy-
prenenolone was used for these assays. Since many of
the inhibitors are pregnenolone derivatives, their com-
petition with this substrate may predict more accurately
their in vivo activity in the rat.
2f
2g
2h
4a
4b
4c
9a
9b
75
199b
100
1500
1500
92
55
5456
79
58b
1200
not tested
50b
14
400
400
150
24
44
20
1500
1500
1500
not tested
1500
19
65
25
21b
10
12
>30000
42b
14a
14b
14c
14d
14e
15a
15b
16
92
28
84
150
150
150
150
11
12
25
21
1500
1500
1500
1500
NI
NI
NI
93
108b
1113
59b
1500
1500
54
NI
17a
17b
18
76
32
39b
400
21
ketoconazole
77
1067b
Microsomes of either human or rat testes were incubated with
candidate inhibitors. Compounds were screened at 150 or 400 nM
with human microsomes and 1500 nM with rat microsomes. IC50
values were determined over
a range of six concentrations.
Incubations were carried out at 34 °C with cofactors with [21-3H]-
17R-hydroxypregnenolone for 60 min, as described under methods.
Following incubation with [21-3H]-17R-hydroxypregnenolone,
[3H]acetic acid liberated during the cleavage reaction at the C-17
was measured in the aqueous phase after extraction of steroids
with chloroform. This method was used to determine the IC50
values for all compounds, except 2f and 4a . These compounds
were incubated with [7-3H]pregnenolone and cofactors for 5 min
and the substrate and products analyzed by HPLC. Both methods
were used to determine the IC50 of 2e, 2g, and 16. Results are
a
average values of two determinations. The variation in repeat
b
values were usually <15% of the mean IC50 values. Average
values of three to five repeat experiments c NI no inhibition.
d
Compounds reported in ref 28.
Str u ctu r e-Activity Rela tion sh ip s. The potencies
of azole derivatives as inhibitors of the human and rat
testicular P45017R are listed in Table 1. However, the
structure-activity relationships of these compounds are
discussed with respect to the human testicular P45017R
enzyme. The compounds containing the 17-(4′-imida-
zolyl) ring (2a and 4a ) demonstrated strong inhibition
of P45017R (IC50 66 and 58 nM). This suggests that the
imidazolyl nitrogen lone pair of electrons at this position
coordinates with the iron atom of the heme in the active
site of the enzyme to produce effective inhibition. The
introduction of the 16R,17R-epoxide 2d (IC50 430 nM)
or the 17R-hydroxy group (4b) (IC50 1200 nM) decreased
inhibition dramatically. In contrast, introduction of a
16,17-double bond (2e and 4c, IC50 24 and 50 nM)
retained or increased this inhibition. From our previous
work,16 the 20-oxime when conjugated with the 16,17-
double bond increased the potency of inhibition more
than 30-fold compared to its 16-saturated counterpart.
Potter et al. also found that a 17-(3′-pyridyl) substituent
together with a 16,17-double bond showed potent inhi-
bition.6
As to modification of the imidazole ring, we found that
the introduction of a methyl group at the 2′-position (2b)
or the larger 2′-phenyl group (2c) decreased inhibitory
activity. The finding that N-acetylation of the imida-
zolyl ring (2h ) also caused some loss of activity is also
not surprising, since the N-acetyl group results in a
delocalization of the π electrons of the nitrogen and the
acetyl or other bulky group on the imidazole ring might
also hinder coordination with heme iron.
17â-(2′-Imidazole)s 9a and 10 in which the steroid is
bonded to the 2′-position of imidazole showed fairly poor
potency in our assays. The 20â-(2′-imidazole) 12 did not
show any activity, although its parent compound, the
20â-carboxaldehyde 11 was quite potent.16 These re-
sults suggest that the juxtaposition between the imi-
dazole ring and the steroid D ring is important. 17-(2′-
Methyl-4′-oxazole) 18, the bioisostere analog of 2b, in
which the N′-atom was substituted for the O-atom, was
a less potent inhibitor, although the 5′-isoxazolyl 17b
was very potent. The lower potency of the 3-acetoxy
derivatives 2g and 2f, compared with those of 2a and
2e, may reflect a limited bulk tolerance at the 3-position

Novel Steroidal Inhibitors of C17,20-Lyase Cytochrome
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3301
s, 5′-H), 5.35 (lH, d, J ) 4.9 Hz, 6-H), 3.40 (lH, m, 3R-H), 2.57
(lH, t, J ) 9.8 Hz, 17R-H), 2.30 (3H, s, 2′-Me), 1.01 (3H, s,
19-Me), 0.53 (3H, s, 18-Me). Anal. (C23H34N2O) C, H, N.
17â-(2′-P h en yl-4′-im id a zolyl)a n d r ost -5-en -3â-ol (2c).
Compound 1a (1 g) and benzaldehyde (4 mL) gave a crude
product which was purified by flash chromatography. Elution
with 3% MeOH-CH2Cl2 gave 2c (240 mg, 22%), mp 237-240
°C (from acetone): δ (CD3OD) 7.83 (2H, d, J ) 7.3 Hz, phenyl-
H), 7.40 (3H, m, phenyl-H), 6.85 (lH, s, 5′-H), 5.31 (lH, d, J )
4.4 Hz, 6-H), 3.40 (lH, m, 3R-H), 2.72 (IH, t, J ) 10.1 Hz, 17R-
H), 1.02 (3H, s, 19-Me), 0.61 (3H, s, 18-Me). Anal. (C28H36N2O)
C, H, N.
17â-(4′-Im idazolyl)-16r,17r-epoxyan dr ost-5-en -3â-ol (2d).
From the 16,17-epoxide 1b (3 g), the 17â-(4′-imidazolyl)
epoxide 2d was obtained as a white solid (1.2 g, 44%), mp 225-
232 °C (from methanol): δ (CD3OD) 7.69 (lH, s, 2′-H), 6.91
(lH, s, 5′-H), 5.35 (lH, d, J ) 4.1 Hz, 6-H), 4.63 (lH, t, J ) 9
Hz, 16-H), 3.41 (1H, m, 3-H), 1.01 (3H, s, 19-Me), 0.58 (3H, s,
18-Me). Anal. (C22H30N2O2‚H2O) C, H, N.
but it still retained reasonable activity (IC50 199 and
75 nM). However, compounds 2g and 2f might be useful
as a “prodrug” of 2a and 2e for in vivo experiments.
17â-Pyrazole (14a ) and isoxazole (15a and 16) de-
rivatives were moderate to weak inhibitors, but their
16-unsaturated counterparts 14b and 17a , 17b as well
as 16-unsaturated 17â-(2′-imidazole) 9b were quite
potent. The introduction of methyl or phenyl groups at
the nitrogen of the pyrazolyl ring of 14b (see 14c, 14d ,
and 14e) showed moderate inhibition. More detailed
studies on the type of inhibition of the more active
inhibitors are in progress.
In summary, among the series of 17-imidazolyl,
pyrazolyl, and isoxazolyl androstene derivatives syn-
thesized, 2e, 4a , 4c, 9b, 14b, 17a , and 17b were found
to be potent inhibitors of human testicular P45017R with
greater activity than ketoconozole. These compounds
were two to four times more potent as P45017R inhibitors
when compared in the same assay with ketoconazole
(IC50 77 nM in our assay system) which is currently in
clinical use. Except for 9b, these compounds showed
strong inhibition of the rat P45017R also and were
several-fold more effective than ketoconazole. A number
of these inhibitors are presently undergoing further
pharmacological study.
17-(4′-Im id a zolyl)a n d r osta -5,16-d ien -3â-ol (2e). 20-Oxo-
pregna-5,16-diene-3â,21-diol 21-acetate (1c)21 (1.30 g, 3.5
mmol) gave a crude product (500 mg) which after two crystal-
lizations yielded 2e (160 mg, 11%), mp 313-318 °C (from
EtOH-H2O). Anal. (C22H30N2O) C, H, N.
The salt 2e‚HCl, mp 300-306 °C (from EtOH-ether), which
is very soluble in alcohol but slightly soluble in water was
prepared as described above for 2a ‚HCl: δ (CD3OD) 8.89 (1H,
s, 2′-H), 7.54 (lH, s, 4′-H), 6.32 (lH, s, 16-H), 5.39 (lH, s, 6-H),
3.40 (lH, m, 3R-H), 1.09 (3H, s, 19-Me), 1.03 (3H, s, 18-Me).
3â-Acetoxy-17-(4′-im id a zolyl)a n d r osta -5,16-d ien e (2f).
A solution of 21-bromo-pregna-5,16-dien-3â-ol (6) (1 g, 2.3
mmol, prepared from 5a with CuBr223), cupric acetate (2.5 g,
12.5 mmol), 28% aqueous ammonium hydroxide (20 mL), and
37% aqueous formaldehyde (3 mL, 37 mmol) was refluxed for
1 h. Worked up as described above for 2a , the crude product
obtained was then recrystallized from acetone-LP, to give 2f
(140 mg, 16%), mp 226-229 °C: δ 7.87-7.43 (lH, broad s, 2′-
H), 7.22-6.87 (1H, broad s, 5′-H), 5.95 (lH, s, 16-H), 5.44 (H.,
d, J ) 4.2 Hz, 6-H), 4.55 (H, m, 3R-H), 2.05 (3H, s, 3-OAc),
1.08 (3H, s, 19-Me), 1.00 (3H, s, 18-Me). Anal. (C24H32N2O2)
C, H, N.
3â-Acetoxy-17â-(4′-im id a zolyl)a n d r ost-5-en e (2g). Fol-
lowing the same procedure described above for 2a , 3,21-
diacetoxypregn-5-en-20-one (1d ) (1 g, 2.4 mmol) gave the crude
product which was purified by flash chromatography. Elution
with 5% MeOH-CH2Cl2 yielded 2g (250 mg, 27%), mp 210-
213 °C (from acetone): δ 7.71 (lH, s, 2′-H), 6.84 (lH, s, 5′-H),
5.39 (lH, d, J ) 4.4 Hz, 6-H), 4.60 (lH, m, 3R-H), 2.67 (lH, t, J
) 9.3 Hz, 17R-H), 2.04 (3H, s, AcO), 1.02 (3H, s, 19-Me), 0.53
(3H, s, 18-Me). Anal. (C24H34N2O2) C, H, N. Further elution
gave 2a (200 mg, 24%), mp 304-307 °C.
17â-(N-Acetyl-4′-im id a zolyl)-3â-a cetoxya n d r ost-5-en e
(2h ). 2a (100 mg) in pyridine (2 mL) and Ac2O (0.5 mL) was
heated at 110 °C for a few minutes until it became clear and
then left at 25 °C for 14 h. Water was added, and the
precipitate was collected and recrystallized from acetone-LP
to give 2h (45 mg, 40%), mp 206-209 °C: δ 8.05 (lH, s, 2′-H),
7.19 (lH, s, 5′-H), 5.39 (lH, s, 6-H), 4.61 (lH, m, 3â-H), 2.57
(3H, s, N-Ac), 2.04 (3H, s, 3-OAc), 1.02 (3H, s, 18-Me), 0.53
(3H, s, 19-Me). Anal. (C26H36N2O3) C, H, N.
17â-(4′-Im id a zolyl)a n d r ost-4-en -3-on e (4a ). Following
the procedure for the preparation of 2a , deoxycorticosterone
21-acetate (3a ) (1 g) gave crude product 4a (0.45 g) which was
recrystallized from methanol (0.15 g, 17%), mp 235-240 °C:
δ 7.58 (lH, s, 2′-H), 6.82 (IH, s, 5′-H), 5.74 (lH, s, 4-H), 2.67
(lH, t, J ) 9 Hz, 17R-H), 1.19 (3H, s, 19-Me), 0.5 (3H, s, 18-
Me). Anal. (C22H30N2O) C, H, N.
Exp er im en ta l Section
Syn th etic Meth od s. 1H NMR data (300 MHZ) (internal
standard Me4Si ) δ 0) were recorded on a QE 300, NMR
systems, General Electric Co., in CDCl3 unless otherwise
stated. Reactions were monitored by TLC on silica gel plates
(Merck Type 60H) and visualized by dipping in 4% sulfuric
acid in ethanol followed by heating at ca. 120-150 °C. Flash
column chromatography was carried out on silica gel (Merck
grade 9385, 230-400 mesh 60 Å) in the solvent systems
indicated. LP refers to petroleum fractions of bp 35-60 °C.
Solutions were dried using anhydrous Na2SO4. Melting points
were measured on a Fischer-J ohns melting point apparatus
and are uncorrected.
17â-(4′-Im id a zolyl)a n d r ost-5-en -3â-ol (2a ). A solution of
the 21-acetoxy ketone 1a (1 g, 2.7 mmol), cupric acetate (2.5
g, 12.5 mmol), 28% aqueous ammonium hydroxide (20 mL) and
37% aqueous formaldehyde (3 mL, 37 mmol) in ethanol (50
mL) was refluxed for 6 h, while at 1, 2, 3, and 4 h an additional
quantity of NH4OH (10, 1, 1, and 1 mL) and formaldehyde (3.5,
1, 1, and 1 mL) were added and refluxing stopped at 6 h.
Water (80 mL) was added and the solvent concentrated to
about 40 mL and left overnight. The grey copper salt product
was collected and suspended in 50% EtOH (100 mL) and
heated to 50 °C while a stream of gaseous hydrogen sulfide
was passed into the suspension for 1 h to precipitate copper
sulfide and free the imidazole derivatives. The precipitate was
collected and refluxed in EtOH (100 mL) for 30 min and then
filtered. The combined filtrate was concentrated to give the
crude product (400 mg) which was recrystallized from EtOH-
H2O to give pure 2a (264 mg, 29%), mp 307-310 °C (lit20 mp
298-300 °C).
HCl Sa lt. Four drops of concentrated HCl was added to a
solution of 2a (100 mg) in 2-propanol (12 mL) to give pH ) 1
(litmus paper). After evaporation to dryness, the residue was
recrystallized from MeOH-ether to give 2a ‚HCl salt (80 mg,
72%), mp 270-275 °C, δ (CD3OD) 8.81 (lH, S, 2′-H), 7.39 (lH,
s, 5′-H), 5.36 (lH, d, J ) 4.4 Hz, 6-H), 3.40 (lH, m, 3R-H), 2.78
(lH, t, J ) 9.8 Hz, 17R-H), 1.02 (3H, s, 19-Me), 0.59 (3H, s,
18-Me). Anal. (C22H32N2O‚HCl) C, H, N.
Following the same or similar procedure, the following
compounds were prepared:
17â-(2′-Meth yl-4′-im idazolyl)an dr ost-5-en -3â-ol (2b). The
21-acetoxy ketone 1a (1 g, 2.7 mmol) and acetaldehyde (1 mL)
gave the 2′-methyl compound 2b (370 mg, 39%), mp 263-266
°C (253 °C shrinks) (from EtOH-CH2Cl2): δ (CD3OD) 6.57 (lH,
17r-Hydr oxy-17â-(4′-im idazolyl)an dr ost-4-en -3-on e (4b).
Following the procedure for the preparation of 2a , 17a-
hydroxydeoxycorticosterone (3b) (1 g) gave 4b (0.36 g, 34%),
mp 195-197 °C (from methanol): δ 7.61 (lH, s, 2′-H), 6.92 (lH,
s, 5′-H), 5.74 (lH, s, 4-H), 1.19 (3H, s, 19-Me), 0.56 (3H, s, 18-
Me). Anal. (C22H30N2O2) C, H, N.
17-(4′-Im id a zolyl)a n d r osta -4,16-d ien -3-on e (4c). A solu-
tion of the 16-ene 2e (50 mg, 0.15 mmol) in toluene (1 0 mL)
and cyclohexanone (1 mL) was heated, and part of the solvent

3302 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Ling et al.
(3-4 mL) distilled off to eliminate any moisture. Aluminum
isopropoxide (61 mg, 0.30 mmol) was added and the mixture
refluxed for 3 h. Saturated sodium tartrate (3 mL) was added
and the whole mixture steam distilled for 2 h, until no more
toluene was distilled off. The remaining residue was extracted
with EtOAc (3 × 50 mL). The combined EtOAc layer was
evaporated, and the residue was subjected to flash chromato-
graphic separation. Elution with 5 % MeOH-CH2Cl2 gave 4c
(35 mg, 70%), which was dissolved in MeOH (5 mL) and
acidified to pH ) 1 (pH paper) with 4 drops of concentrated
HCI. After evaporation, the residue was recrystallized from
MeOH-LP to give the 4c‚HCl salt (12 mg), mp 267-270 °C:
δ (CD3OD) 8.90 (lH, s, 2′-H), 7.55 (lH, s, 5′-H), 6.32 (lH, s, 16-
H), 5.73 (lH, s, 4-H), 1.29 (3H, s, 19-Me), 1.06 (3H, s, 18-Me).
Anal. (C22H28N2O‚HCl) C, H, N.
3â-Hyd r oxya n d r ost-5-en e-17â-ca r boxa ld eh yd e (8a ). A
solution of the 21-acetoxy ketone la (2.25 g, 6 mmol) in THF
(20 mL) was added dropwise to a stirring solution of LiAIH4
(600 mg, 15.8 mmol) in THF (20 mL) at 25 °C and stirring
continued for 30 min. Water was added carefully to destroy
excess LiAIH4. EtOAc (200 mL) was added, and the whole
mixture was washed with 1 N HCI (400 mL × 3), followed by
water (2 × 40 mL). After evaporation, the diol 7 (2 g) was
obtained.
21-F or m yl-3â-h yd r oxyp r egn -5-en -20-on e (13a ) was pre-
pared according to the method of Pelc et al.26 in 69% yield,
mp 149-152 °C (lit.25 mp 111-115 °C or lit26 mp 154-157
°C): 17â-(3′-p yr a zolyl)-a n d r ost-5-en -3â-ol (14a ) (41% yield,
mp 222-227 °C, lit.25 mp 220-6 °C or lit.26 233-235 °C) and
17â-(5′-isoxa zolyl)-5-a n d r osten -3â-ol (15a ) (13% yield, mp
198-201 °C, lit25 mp 198-202 °C) were prepared according to
ref 25.
17â-(3′-Isoxa zolyl)-5-a n d r osten -3â-ol (16). A mixture of
15a and 16 (100 mg) [prepared from â-keto aldehyde 13a with
hydroxylamine hydrochloride in acetic acid25] was dissolved
in THF (5 mL), NaOMe (250 mg) was added, and the mixture
was stirred for 1.5 h. Water (20 mL) was added and the whole
mixture extracted with CH2Cl2. The organic layer was evapo-
rated and the residue recrystallized from acetone-LP to give
3′-isoxazole 16 (59 mg, 45%, mp 203 -205 °C (lit.25 mp 205.5
-207 °C).
21-F or m yl-3â-h yd r oxyp r egn a -5,16-d ien -20-on e (13b).
To a stirred and cooled (ice bath) mixture of sodium methoxide
(2 g, 6 mmol) and ethyl formate (8 mL, 543 mmol) (dried over
phosphorus pentoxide and distilled) was added dropwise a
solution of 16-en-20-one 5a (2 g, 5.6 mmol) in pyridine (20 mL).
After being stirred for 1 h, the solution was poured into water
(250 mL) containing concentrated HCI (70 mL). The precipi-
tate which was collected and washed with water gave 13b (1.5
g, 75%), mp 118-123 °C. 1H NMR showed that the impure
â-keto aldehyde existed mainly in the enol form (13b): δ 8.04
(IH, d, J ) 3.9 Hz, 22-H), 6.71 (lH, s, 16-H), 5.79 (IH, d, J )
3.9 Hz, 21-H), 5.37 (lH, s, 6-H), 3.54 (lH, m, 3R-H), 1.06 (3H,
s, 19-Me), 1.00 (3H, s, 18-Me). This product was used directly
in the following reaction. A portion was recrystallized from
acetone-LP, mp 234-238 °C.
A solution of sodium periodate (2 g, 9.3 mmol) in water (20
mL) was added to the above diol (7) (2g, 6 mmol) in MeOH
(100 mL) and the mixture stirred at 25 °C for 40 min. The
precipitated sodium iodate was filtered off and the filtrate
concentrated under reduced pressure below 45 °C to about 20
mL to yield the 17-al 8a (1.8 g, 88%), mp 157-159 °C (lit.30
mp 155-157 °C).
17â-(2′-Im id a zolyl)-5-a n d r osten -3â-ol (9a ). The 17-al 8a
(1 g, 3.4 mmol), cupric acetate (2.0 g, 10 mmol), 28% aqueous
ammonium hydroxide (20 mL), and 40% aqueous glyoxal (3
mL, 20.7 mmol) were refluxed for 6 h. At 1, 2, and 4 h, an
additional amount of NH4OH (each 5 mL) was added and
refluxing stopped at 6 h. Water was added and the precipitate
was collected, which was then extracted by refluxing with
EtOH (2 × 50 mL). The combined filtrate was concentrated
to 5-10 mL, and LP was added to give a precipitate which
was recrystallized from MeOH-H2O to yield colorless crystals
9a (291 mg, 24%), mp 288-291 °C: δ (CD3OD) 6.91 (2H, s ,
4′,5′-H), 5.35 (lH, d, J ) 3.9 Hz, 6-H), 3.40 (lH, m, 3R-H), 2.76
(lH, t, J ) 9.3 Hz, 17R-H), 1.01 (3H, s, 19-Me), 0.53 (3H, s,
18-Me). Anal. (C22H32N2O) C, H, N.
17-(2′-Im id a zolyl)a n d r osta -5,16-d ien -3â-ol (9b). A solu-
tion of 16-en-17-al 8b (1 g, 3.3 mmol) (prepared from lc by
the procedure described above for 8a 24), cupric acetate (1.0 g,
5 mmol), 40% aqueous glyoxal (1 mL, 8.7 mmol), and 28%
aqueous NH4OH (6 mL) in ethanol (50 mL) was refluxed for
12 h, while at 1, 2, 3, 4, 6, and 10 h an additional quantity of
NH4OH (each 1.2 mL) and 40% aqueous glyoxal (each 0.5 mL)
were added and the refluxing stopped at 12 h. Workup as
described above for 2a gave a crude product which was
subjected to flash chromatographic separation, eluted with 5%
MeOH-CH2Cl2, to give 9b (60 mg, 5% yield), mp 272-275 °C
(from acetone): δ 7.06 (2H, s, 4′,5′-H), 6.17 (IH, s, 16-H), 5.40
(lH, d, J ) 4.9 Hz, 6-H), 3.55 (lH, m, 3R-H), 1.25 (3H, s, 19-
Me), 1.08 (3H, s, 18-Me). Anal. (C22H30N2O) C, H, N.
17â-(2′-Im id a zolyl)a n d r ost -4-en -3-on e (10). Following
the same procedure described above for 4c, the 3-ol 9a (260
mg) was oxidized with cyclohexanone (1 mL) and aluminum
isopropoxide (250 mg) in toluene (25 mL). The crude product
obtained was recrystallized from EtOH-H2O, decolorized with
charcoal, to give 10 (169 mg, 65%), mp 256-9 °C: δ (CD3OD)
6.92 (2H, s, 4′,5′-H), 5.71 (lH, s, 4-H), 2.78 (lH, t, J ) 9.8 Hz,
17R-H), 1.22 (3H, s, 19-Me), 0.58 (3H, s, 18-Me). Anal.
(C22H30N2O) C, H, N.
20â-(2′-Im id a zolyl)p r egn -4-en -3-on e (12). Following a
similar procedure to that used for the preparation of 9a ,
3-oxopregn-4-en-3-e-20â-carboxaldehyde (11) (1 g) gave 12
(0.24 g, 22%), mp 289-294 °C (dec) (from MeOH): δ 6.92 (2H,
s, 4′, 5′-H), 5.72 (lH, s, 4-H), 2.90 (lH, m, 20-H), 1.32 (3H, d, J
) 7.5 Hz, 21-H), 1.19 (3H, s, 19-Me), 0.79 (3H, s, 18-Me). Anal.
(C24H34N2O) C, H, N.
17-(3′-P yr a zolyl)a n d r ost a -5,16-d ien -3â-ol (14b ). The
â-keto aldehyde 13b (1.5 g, 4.4 mmol) and anhydrous hydra-
zine (0.75 mL) in MeOH (50 mL) were refluxed under N2 for
3 h. Water was added and the precipitate (900 mg) collected.
Three recrystallizations from MeOH gave pure 14b (45 mg,
3%), mp 216-220 °C (lit.26 mp 222-223.5 °C).
17-(2′-P h en yl-3′-pyr azolyl)an dr osta-5,16-dien -3â-ol (14c).
The â-keto aldehyde 13b (1g, 2.9 mmol) and phenylhydrazine
(0.40 mL) in MeOH (30 mL) were refluxed for 3 h. Water was
added and the mixture extracted with EtOAc. The organic
layer was evaporated and the residue chromatographed, to
give on elution with 10% acetone-LP 14c (116 mg, 9.6%), mp
208-210 °C (EtOH) (lit.26 mp 208.5-210 °C).
17-(1′-Meth yl-5′-p yr a zolyl)- a n d 17-(1′-Meth yl-3′-p yr a -
zolyl)a n d r osta -5,16-d ien -3â-ols (14d a n d 14e). The â-keto
aldehyde 13b (1 g, 2.9 mmol) and methylhydrazine (400 mg,
8.7 mmol) in methanol (30 mL) were refluxed for 2 h under
an atmosphere of N2. After the solution was cooled to 25 °C,
water (10 mL) was added dropwise, the yellow precipitate (800
mg) was collected, and two crystallizations from ethanol gave
the pure 5′-pyrazole 14d (220 mg, 21%), mp 251-254 °C (lit.26
mp 242-244 °C). The mother liquor (550 mg) was flash
chromatographed and eluted with 50% ethyl acetate-LP to
give 3′-pyrazole 14e (90 mg, 8.7%) [mp 224-226 °C (from
acetone): δ 7.26 (lH, s, 5′-H), 6.26 (lH, d, J ) 1.9 Hz, 16-H),
6.08 (IH, s, 4′-H), 5.38 (lH, d, J ) 4.9 Hz, 6-H), 3.88 (3H, s,
1′-Me), 3.53 (lH, m, 3R-H), 1.07 (3H, s, 19-Me), 1.03 (3H, s,
18-Me), Anal. (C23H32N2O) C, H, N] and 14d (160 mg, 16%),
mp 251-254 °C (from EtOH).
17-(5′-Isoxa zolyl)a n d r osta -5,16-d ien -3â-ol (15b). To a
benzene (5 mL) solution of 13b (700 mg, 2.04 mmol) were
added glacial acetic acid (5 mL) and a solution of hydroxy-
lamine hydrochloride (200 mg, 2.88 mmol) and sodium acetate
(100 mg) in water (0.5 mL). This mixture was made homo-
geneous by the addition of ethanol (3 mL) and refluxed for 4
h. After cooling, water (150 mL) was added and the mixture
extracted with CH2Cl2. The organic layer was evaporated, and
the residue was chromatographed, eluting with 20% acetone-
LP, to afford 15b (500 mg, 72%), mp 191-193 °C (from
acetone-LP): δ 8.19 (lH, s, 3′-H), 6.45 (lH, s, 16-H), 6.18 (lH,
s, 4′-H), 5.38 (lH, s, 6-H), 3.54 (lH, m, 3R-H), 1.07 (3H, s, 19-
Me), 1.00 (3H, s, 18-Me). Anal. (C22H29NO2) C, H, N.
17-(3′-P yr a zolyl)a n d r osta -4,16-d ien -3-on e (17a ). Fol-
lowing the procedure for the preparation of 4c, Oppenauer

Novel Steroidal Inhibitors of C17,20-Lyase Cytochrome
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20 3303
oxidation of 14b (350 mg, 1.04 mmol) with cyclohexanone (5
mL) and aluminum isopropoxide (461 mg, 2.26 mmol) in
toluene (60 mL) afforded a crude product, which was chro-
matographed and eluted with 2% MeOH-CH2Cl2, to give the
4-en-3-one 17a (200 mg, 58%), mp 167-171 °C (155 °C shrinks)
(from acetone-LP): δ 7.54 (lH, s, 5′-H), 6.33 (lH, s, 16-H), 6.05
(lH, (1H, s, 4′-H), 5.76 (1H, s, 4-H), 1.24 (3H, s, 19-Me), 1.04
(3H, s, 18-Me). Anal. (C22H28N2O) C, H, N.
17-(5′-Isoxa zolyl)a n d r osta -4,16-d ien -3-on e (17b). Fol-
lowing the procedure for the preparation of 17a , Oppenauer
oxidation of 15b (300 mg, 0.88 mmol) with cyclohexanone (5
mL) and aluminum isopropoxide (400 mg, 1.96 nunol) in
toluene (50 mL) afforded the title compound 17b (125 mg,
42%), mp 184-186 °C (from acetone-LP): δ 8.19 (lH, d, J )
1.4 Hz, 3′-H), 6.44 (lH, s, 16-H), 6.19 (lH, s, 4′-H), 5.76 (lH, s,
(approximately 90 µg of protein) with [7-3H]-pregnenolone (5
× 104 dpm, 400 nM), the NADPH generating system as above,
and different concentrations of the test compounds in phos-
phate buffer (pH 7.4, total volume 1 mL) under oxygen for 5
min at 34 °C. Authentic steroid markers and 14C-labeled
pregnenolone, 17R-hydroxypregnenolone, and DHEA were
added to correct for procedural losses. The steroids were
extracted with ether and then separated by HPLC using a
NOVA-PAK, C18 reverse phase column and eluting with
acetonitrile/methanol/water (30:10:60), and the radioactivity
was measured in each fraction collected. The lyase was
determined from the percentage conversion of [7-3H]preg-
nenolone to DHEA (the conversion of substrate to androstene-
diol and testosterone under the experimental conditions was
negligible).
4-H), 1.24 (3H, s, 19-Me), 1.03 (3H, s, 18-Me). Anal. (C22H27
NO2) C, H, N.
-
Ack n ow led gm en t. We are grateful to Dr. Patrick
Callery for use of the 1H NMR in the Department of
Biomedicinal Chemistry, School of Pharmacy, UMAB.
We thank Dr. Stephen J acobs, Department of Surgery,
UMAB, and Dr. J ames Mohler, University of North
Carolina, for supplying the tissues used in these experi-
ments. Thanks are due to Dr. J ohn F. Templeton,
Faculty of Pharmacy, University of Manitoba, Canada,
for reviewing the manuscript. The help of Dr. Vincent
C. O. Njar in preparing this paper is greatly appreci-
ated. The excellent assistance of Xin Wang is acknowl-
edged.
17-(2′-Meth yl-4′-oxa zolyl)a n d r ost-5-en -3â-ol (18). The
21-acetoxy ketone la (500 mg, 1.34 mmol) and ammonium
acetate (500 mg, 6.5 mmol) in glacial acetic acid (15 mL) were
refluxed for 8 h. Water (50 mL) was added and the pale brown
precipitate collected, washed with water, and dried. The crude
product (350 mg) in MeOH (15 mL) was combined with KHCO3
(500 mg) in water (5 mL), the mixture was refluxed for 4 h,
H2O was added, and the precipitate (216 mg) was collected
and recrystallized from acetone to give 18 (125 mg, 26%), mp
181-183 °C: δ 7.20 (lH, s, 5′-H), 5.31 (lH, d, J ) 5.4 Hz, 6-H),
3.49 (lH, m, 3R-H), 2.48 (lH, t, J ) 9.8 Hz, 17R-H), 2.37 (3H,
s, 2′-Me), 0.96 (3H, s, 19-Me), 0.48 (3H, s, 18-Me). Anal.
(C23H33NO2) C, H, N.
En zym e P r ep a r a tion a n d Assa y for Testicu la r C17,20
-
Lya se. [7-3H]Pregnenolone (25 Ci/mmol) was purchased from
New England Nuclear Corp. (Boston, MA) and checked for
purity and/or purified by TLC or HPLC prior to use. [21-3H]-
17R-Hydroxypregnenolone (13.61 µCi/µmol) was prepared in
our laboratory as described by Akhtar et al.3l Ketoconazole
was purchased from Sigma Chemical Co. (St.Louis, MO).
Scintillation cocktail 3a70B was purchased from RPI Corp.
(Mount Prospect, IL).
Refer en ces
(1) Ayub, M.; Level, M. Inhibition of Testicular 17R-Hydroxylase
and 17,20-Lyase but not 3-Hydroxysteroid Dehydrogenase-
Isomerase or 17-Hydroxysteroid Oxidoreductase by Ketoconazole
and Other Imidazole drugs. J . Steroid Biochem. 1987, 28, 521-
531.
(2) (a) Nakajin, S,; Takahashi, K,; Shinoda, M. Inhibitory Effect and
Interaction of Stanozolol with Pig Testicular Cytochrome P-450
(17R-Hydroxylase/C17,20-Lyase). Chem. Pharm. Bull. (Tokyo)
1989, 37, 1855-1858. (b) Nakajin S.; Takahashi, K.; Shinoda,
M. Inhibitory Effect and Spectral Changes on Pig Testicular
Cytochrome P-450 (17R-Hydroxylase/C17,20-lyase) by 20â-Hy-
droxy-C21Steroids. Yakugaku Zasshi 1988, 108, 1188-1195.
(3) Angelastro, M. R.; Laughlin, M. E.; Schatzman, G. L.; Bey, P.;
Blohm, T. R. 17â-(Cyclopropylamino)-androst-5-en-3â-ol, a Selec-
tive Mechanism-Based Inhibitor of Cytochrome P45017R(Steroid
17â-Hydroxylase/Cl7,20Lyase). Biochem. Biophys. Res. Commun.
1989, 162, 1571-1577.
(4) J arman, M.; Barrie, S. E.; Deadman, J , J ,; Houghton, J .;
McCague, R. Hydroxyperfluoroazobenzenes: Novel Inhibitors of
Enzymes of Androgen Biosynthesis. J . Med. Chem. 1990, 33,
2452-2455.
(5) McCague, R.; Rowlands, M. G.; Barrie, S. E.; Houghton, J .
Inhibition of Enzymes of Estrogen and Androgen Biosynthesis
by Ester of 4-Pyridylacetic acid. J . Med. Chem. 1990, 33, 3050-
3055.
Testicu la r Micr osom es. The previously reported proce-
dure14 was followed to prepare microsomes from human testes
(obtained from untreated prostatic cancer patients undergoing
orchidectomy). Microsomes from the testes of adult Sprague-
Dawley rats (Charles River Laboratories, weight 200-250 g)
were prepared as previoulsy described.16 The microsomes were
stored at -70 °C until assayed. J ust before use, the thawed
microsomes were diluted with 0.1 M phosphate buffer to
appropriate concentrations. The protein concentration of
microsomes used in each assay was determined by the method
of Lowry et al.32
Mea su r em en t of C17,20-Lya se Activity. (1) Ra d iom etr ic
Assa y. The substrate for this assay was [21-3H]-17R-hydrox-
ypregnenolone (13.61 µCi/µmol). Each tube contained 300 000
dpm and a total concentration of 10 µM 17R-hydroxypreg-
nenolone. The cofactors (NADP 65 µM; glucose 6-phosphate
0.71 mM; glucose-6-phosphate dehydrogenase 0. 13 IU in 50
µL of phosphate buffer) and different concentrations of the test
compounds in phosphate buffer (pH 7.4, total volume 1 mL)
were added and tubes preincubated for 10 min at 34 °C. The
reaction was initiated by the addition of the microsomes
(approximately 90 µg of protein) and the incubation carried
out for 60 min under oxygen at 34 °C. Chloroform was then
added to extract the steroids and an aliquot (0.75 mL) of
aqueous phase removed and mixed with an equal volume of
2.5% charcoal suspension. After vortexing, the tubes were
allowed to stand for at least 30 min then centrifuged, an
aliquot of the supernatant was removed, and the tritium
concentration was measured by liquid scintillation counting.
The results are listed in Table 1. IC50 values were calculated
from the linear regression line of the plot of log of lyase activity
versus log of four to five inhibitor concentrations. The results
were obtained from duplicate sets of experiments and were
repeated at least once and several times for the more active
inhibitors.
(6) Potter, G. A.; Barrie, S. E.; J arman, M. and Rowlands, M. G.
Novel Steroidal Inhibitors of Human Cytochrome P45017R
-
(17R-Hydroxylase-C17,20-lyase): Potential Agents for the Treat-
ment of Prostatic Cancer. J . Med. Chem. 1995, 38, 2463-2471.
(7) Burkhart, J . P.; Gates, C. A.; Laughlin, M. E.; Resvick, R. J .;
Peet, N. P. Inhibitors of Steroid C17(20) Lyase with C-17-
Heteroaryl Steroids. Bioorg. Med. Chem. 1996, 4, 1411-1420.
(8) Njar, V. C. O.; Hector, M.; Hartmann, R. W. 20-Amino and 20,-
21-Aziridinyl Pregnene Steroids: Development of Potent Inhibi-
tors of 17R-Hydroxylase/C17,20-Lyase (P450 17). Bioorg. Med.
Chem. 1996, 4, 1447-1453.
(9) Njar, V. C. O.; Klus, G. T.; Brodie, A. M. H. Nucleophilic Vinylic
“Addition-Elimination” Substitution Reaction of 3â-Acetoxy-
17-chloro-16-formylandrosta5,16-diene: A Novel and General
Route to 17-Substituted Steroids. Part 1. Synthesis of Novel 17-
Azolyl-∆16 Steroids; Inhibitors of 17R-Hydroxylase/17,20-Lyase
(17R-Lyase). Bioorg. Med. Chem. Lett. 1996, 6, 2777-2782.
(10) Trachtenberg, J . Ketoconazole Therapy in Advanced Prostatic
Cancer. J . Urol. 1984, 132, 61-63.
(11) Williams, G.; Kerle, D. J .; Doble, A.; Dunlop, H.; Smith, C.; Allen,
J .; Yeo, T.; Blooms, S. R. Objective Responses to Ketoconazole
Therapy in Patients with Relapsed Progressive Prostatic Cancer.
Br. J . Urol. 1986, 58, 45-51.
(2) P r od u ct Isola tion Assa y. The assay was performed
as previously16 by incubating human testicular microsomes

3304 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 20
Ling et al.
(12) Stoner, E. The Clinical Development of a 5R-Reductase Inhibitor,
Finasteride. J . Steroid Biochem. Mol. Biol. 1990, 37, 375-378.
(13) Geller, J . Effects of Finasteride, a 5R-Reductase Inhibitor on
Prostate Tissue Androgens and Prostatic-Specific Antigen. J .
Clin. Endocrinol. Metab. 1990, 71, 1552-1555.
(23) Glazier, E. R. Bromination with Cupric Bromide. II. Bromination
in the Presence of an Olefinic Bond. J . Org. Chem. 1962, 27,
4397-4399.
(24) Ling, Y.; Li, J .; Klus, G. T.; Son C.; Marat, K.; Brodie, A. M. H.
Unpublished results.
(25) Doorenbos, N. J .; Milewich. L.; 17â-Isoxazolyl and 17â-Pyrazolyl
Steroids from 3â-Hydroxy-21-formyl-pregn-5-en-20-one. Struc-
tural Assignments. J . Org. Chem. 1966, 31, 3193-3199.
(26) Pelc, B.; Hodkova, J . Steroid Derivatives. LX. Androstane
Derivatives Substituted by Pyrazole Ring in Position 17. Collect.
Czech. Chem. Commun. 1969, 34, 442-450.
(27) Ohta, G.; Kosh, K.; Obata, K. Investigations on Steroids XI.
Synthesis of Steroidal Oxazole, Imidazole, and Triazole, Chem.
Pharm. Bull. 1968, 16, 1487-1497.
(28) Yanaihara, T.; Troen, P. Studies of the Human Testis. 1.
Biosynthetic Pathways for Androgen Formation in Human
Testicular Tissue In Vitro. J . Clin. Endocrinol. Metab. 1972, 34,
783-792.
(29) Klus, G. T.; Nakamura, J .; Li, J .; Ling, Y.; Son, C.; Kemppained,
J . A.; Wilson, E. M.; Brodie, A. M. H. Growth Inhibition of
Human Prostate Cells In Vitro by Novel Inhibitors of Androgen
Synthesis. Cancer Res. 1996, 56, 4956-4964.
(30) Danishefsky, S.; Nagasawa, K. Conversion of Androstenolone
to Pregnenolone. J . Org. Chem. 1975, 40, 1989-1990.
(31) Akhtar, M.; Corina, D. L.; Miller, S. L.; Shyadehi, A. Z.; Wright,
J . N. Incorporation of label from 18O2 into acetate during side-
chain cleavage catalyzed by cytochrome P45017R (17R-hydroxy-
lase-17,20-lyase). J . Chem. Soc., Perkin Trans. 1 1994, 263-
267.
(32) Lowry, O. H.; Roseborough, N. S.; Farr, A. L.; Randall, R. S.
Protein Measurement with the Folin Phenol Reagent. J . Biol.
Chem. 1951, 193, 265-275.
(14) Li, J .; Li, Y.; Son, C.; Brodie, A. Inhibition of Androgen Synthesis
by 22-Hydroximino-23,24-bisnor-4-cholen-3-one. The Prostate
1995, 26, 140-150.
(15) Li, J .; Li, Y.; Son, C.; Banks, P.; Brodie, A. 4-Pregnen-3-one-
20â-carboxaldehyde: A Potent Inhibitor of 17R-Hydroxylase. J .
Steroid Biochem. Mol. Biol. 1992, 42, 313-320.
(16) Li, J . S.; Li, Y.; Son, Chong.; Brodie, A. Synthesis and Evaluation
of Pregnane Derivatives as Inhibitors of Human Testicular 17R-
Hydroxylase/C17,20-Lyase. J . Med. Chem. 1996, 39, 4335-4339.
(17) Lang, M.; Batzl, Ch.; Furet, P.; Bowman, R.; Hausler, A.;
Bhatnager, A. Structure-activity Relationships and Binding
Model of Novel Aromatase Inhibitors. J . Steroid Biochem. Mol.
Biol. 1993, 44, 421-428.
(18) Lipton, A.; Harvey H. A.; Demers L. M.; Hanagan, J . R.;
Mulagha, M. T.; Kochak, G. M.; Fitzsimmons, S.; Sanders, S. I.;
Santen, R. J . A phase I trial of CGS 16949A, A New Aromatase
Inhibitor. Cancer 1990, 65, 1279-1285.
(19) Schipper, E. S.; Day, A. R. Imidazoles and Condensed Imidazoles.
In Heterocyclic Compounds Vol. 5; Elderfiled, R. C., Ed.; J ohn
Wiley & Sons, Inc: New York, 1967; pp 194-29.
(20) Ronig, K. (Assignor to Searle Co.), 4-(Cyclopentanopolyhydro-
phenanthr-17-yl) Imidazoles and Derivatives Thereof, U.S.
Patent 2,664,423, 1953; Chem. Abstr. 1953, 49, 7608e.
(21) Mille, J . S.; Bowas, A.; Djerassi, C.; Ringold, H. F. Steroids,
CXXXVII, Synthesis of a New Class of Potent Cortisol Hormones.
6R,9R-Difluoro-16R-hydroxyprednisolone and its Acetonide. J .
Am. Chem. Soc. 1960, 82, 3399.
(22) Barrie, S. E.; Potter, G. A.; Goddard, P. M.; Haynes, B. P.;
Dowsett, M.; J arman, M. Pharmacology of Novel Steroidal
Inhibitors of Cytochrome P450 17R (17R-hydroxylase/C17,20-
Lyase) J . Steroid Biochem. Mol. Biol. 1994, 50, 267-273.
J M970337K