The 16,17-double bond is needed for irreversible inhibition of human cytochrome P45017α by abiraterone (17-(3-pyridyl)androsta-5,16-dien-3β- ol) and related steroidal inhibitors

Article abstract of DOI:10.1021/jm981017j

Abiraterone (17-(3-pyridyl)androsta-5,16-dien-3β-ol, 1) is a potent inhibitor (IC₅₀ 4 nM for hydroxylase) of human cytochrome P450₁₇α. To assist in studies of the role of the 16,17-double bond in its mechanism of action, the novel 17α-(4-pyridyl)androst-5-en-3β-ol (5) and 17β-(3- pyridyl)-16,17α-epoxy-5α-androst-3β-ol (6) were synthesized. 3β- Acetoxyetienic acid was converted in three steps into 5 via photolysis of the thiohydroxamic ester 8. Oxidation of an appropriate 16,17-unsaturated precursor (21) with CrO₃-pyridine afforded the acetate (23) of 6. Inhibition of the enzyme by 1, the similarly potent 5,6-reduced analogue 19 (IC₅₀ 5 nM), and the 4,16-dien-3-one 26 (IC₅₀ 3 nM) and by the less potent (IC₅₀ 13 nM) 3,5,16-triene 25 is slow to occur but is enhanced by preincubation of the inhibitor with the enzyme. Inhibition following preincubation with these compounds is not lessened by dialysis for 24 h, implying irreversible binding to the enzyme. In contrast under these conditions the still potent (IC₅₀ 27 nM) 17α-(4-pyridyl)androst-5-en-3β-ol (5) showed partial reversal after 5 h of dialysis and complete reversal of inhibition after 24 h. This behavior was also shown by the less potent 16,17-reduced 3-pyridyl compounds 3 and 24. Further, in contrast to the compounds (1, 19, 25, 26) with the 16,17-double bond, the inhibition of the enzymic reaction was not enhanced by preincubation either with 5 or with the 17β-pyridyl analogues 3, 4, and 24 which also lack this structural feature. The results show that the 16,17- double bond is necessary for irreversible binding of these pyridyl steroids to cytochrome P450₁₇α. However oxidation to an epoxide is probably not involved since epoxide 6 was only a moderately potent inhibitor (IC₅₀ 260 nM).

Full text of DOI:10.1021/jm981017j

J . Med. Chem. 1998, 41, 5375-5381
5375
Th e 16,17-Dou ble Bon d Is Need ed for Ir r ever sible In h ibition of Hu m a n
Cytoch r om e P 45017r by Abir a ter on e (17-(3-P yr id yl)a n d r osta -5,16-d ien -3â-ol) a n d
Rela ted Ster oid a l In h ibitor s
Michael J arman,* S. Elaine Barrie, and J ose M. Llera†
Cancer Research Campaign Centre for Cancer Therapeutics, Institute of Cancer Research, Cancer Research Campaign
Laboratory, 15 Cotswold Road, Sutton, Surrey SM2 5NG, U.K.
Received March 27, 1998
Abiraterone (17-(3-pyridyl)androsta-5,16-dien-3â-ol, 1) is a potent inhibitor (IC50 4 nM for
hydroxylase) of human cytochrome P45017R. To assist in studies of the role of the 16,17-double
bond in its mechanism of action, the novel 17R-(4-pyridyl)androst-5-en-3â-ol (5) and 17â-(3-
pyridyl)-16,17R-epoxy-5R-androst-3â-ol (6) were synthesized. 3â-Acetoxyetienic acid was
converted in three steps into 5 via photolysis of the thiohydroxamic ester 8. Oxidation of an
appropriate 16,17-unsaturated precursor (21) with CrO
3
-pyridine afforded the acetate (23) of
6
. Inhibition of the enzyme by 1, the similarly potent 5,6-reduced analogue 19 (IC50 5 nM),
and the 4,16-dien-3-one 26 (IC50 3 nM) and by the less potent (IC50 13 nM) 3,5,16-triene 25 is
slow to occur but is enhanced by preincubation of the inhibitor with the enzyme. Inhibition
following preincubation with these compounds is not lessened by dialysis for 24 h, implying
irreversible binding to the enzyme. In contrast under these conditions the still potent (IC50 27
nM) 17R-(4-pyridyl)androst-5-en-3â-ol (5) showed partial reversal after 5 h of dialysis and
complete reversal of inhibition after 24 h. This behavior was also shown by the less potent
1
2
6,17-reduced 3-pyridyl compounds 3 and 24. Further, in contrast to the compounds (1, 19,
5, 26) with the 16,17-double bond, the inhibition of the enzymic reaction was not enhanced
by preincubation either with 5 or with the 17â-pyridyl analogues 3, 4, and 24 which also lack
this structural feature. The results show that the 16,17-double bond is necessary for irreversible
binding of these pyridyl steroids to cytochrome P45017R. However oxidation to an epoxide is
probably not involved since epoxide 6 was only a moderately potent inhibitor (IC50 260 nM).
We have recently described potent steroidal inhibitors
testicular enzyme and probes the structural features
required for the irreversible binding by the steroidal
inhibitors. It explores the hypothesis that the 16,17-
double bond is the essential structural feature confer-
ring this binding, either directly or through a potential
for further metabolism to a reactive epoxide. In this
context, Angelastro and co-workers have designed mech-
anism-based steroidal inhibitors of cytochrome P450
with cyclopropylamino or cyclopropyloxy substituents
at C-17 proposed to be activated by enzymatic oxidation
to covalently reactive species. In these studies, the
stronger inhibition of the compounds observed following
preincubation with the enzyme was evidence supporting
the hypothesis of mechanism-based inhibition.
of cytochrome P45017R, a potential target enzyme in the
1
treatment of hormone-dependent prostatic carcinoma.
One of these compounds, abiraterone (17-(3-pyridyl)-
androsta-5,16-dien-3â-ol, 1), has been selected for clini-
cal evaluation based in part on its marked reduction of
circulating testosterone levels in the male rat and mouse
2
and of androgen-dependent organ weights in the mouse.
1
7R
6
7
We have also described potent nonsteroidal inhibitors
of the target enzyme, which were esters containing a
3
,4
pyridyl residue. Examples of each series having high
inhibitory potency for the target enzyme coupled with
metabolic stability toward esterases were selected for
study in vivo. Unlike 1, they did not lower circulating
testosterone levels nor the weights of androgen-sensitive
Analogues with the 16,17-double bond reduced have
5
organs in the WHT mouse. Two possible contributory
been synthesized previously, by reduction of 1 and its
factors were considered to explain this difference. First
the inhibitory potencies of the nonsteroidal compounds
toward the murine enzyme appear insufficient, in
relation to the Km values for the enzyme substrate, to
inhibit the murine enzyme in vivo. Second, the binding
of the nonsteroidal inhibitors to the target enzyme is
reversible, whereas that of the steroidal inhibitor 1 is
not. The present study extends investigations carried
out on the murine cytochrome P45017R to the human
1
4
-pyridyl analogue 2. The 17â-3-pyridyl derivative 3
was 1 order of magnitude less potent than its precursor
1
(Table 1). The 17â-4-pyridyl derivative 4 was less
potent still, though more potent than its precursor 2,
1
the pyridyl nitrogen of which was expected to interact
poorly with the target heme residue. In a molecular
4
modeling study, the observed relative inhibitory poten-
cies of enantiomeric pairs of 1-(4-pyridyl)ethyl 1-ada-
mantanecarboxylates and their 3-pyridyl-substituted
8
counterparts were rationalized in terms of overlays of
*
To whom inquiries should be addressed.
Present address: Departamento de Qu ´ı mica Org a´ nica y Farma-
the inhibitor molecules onto pregnenolone such that C-1
of the pyridylethyl residue corresponds to C-17 of the
steroid. The proposed overlays would predict that the
†
ceutica, Facultad de Farmacia, Universidad de Sevilla, 41071 Sevilla,
Spain.
1
0.1021/jm981017j CCC: $15.00 © 1998 American Chemical Society
Published on Web 12/01/1998

5
376 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27
7R-4-pyridyl analogue 5 might be better orientated for
J arman et al.
1
use of the convenient reaction of secondary alcohols with
iodine to give iodides, with inversion of configura-
The alkylation at C-2 and C-4 of pyridines
using alkyl iodides as sources of free radicals has been
interaction of the pyridyl nitrogen with the heme
residue in the enzyme binding site than either 3 or 4
and could be a more potent inhibitor. Compound 5 might
then provide a better comparison with 1 for assessing
the putative role of the double bond in irreversible
binding. The present study also describes the synthesis
and evaluation of 5 and the epoxide 6, a potential
reactive metabolite of a 3-pyridyl steroid containing the
1
3,14
tion.
1
5
described, but when 7 was subjected to the various
reaction conditions used, no pyridine derivatives were
isolated. Analogous attempts to use etienic acid as the
free radical source also failed. However this acid was
used in an eventually successful approach (Scheme 2)
using methodology described by Barton et al.1
6,17
1
6,17-double bond.
Conversion of 3â-acetoxyetienic acid into the thiohy-
droxamic ester 8 followed by irradiation in pyridine/CH2-
Cl2 in the presence of camphorsulfonic acid gave in
addition to the reduction product 9 (17% yield; also
formed in the above-mentioned reaction with 7) low
yields of the isomeric 4-pyridyl derivatives (11%; 10(R):
1
1
1(â), 9:1) and the 2-pyridyl analogues (11.5%; 12(R):
3(â), 9:1). The major product (30%) comprised a 20:1
ratio of the chromatographically separable R (14) and
â (15) epimers of the (2-pyridyl)thio derivative. Although
(
2-pyridyl)thio derivatives have been described as prod-
1
8
ucts of this radical approach, their formation does not
previously appear to have been a problem in the context
of decarboxylative addition onto protonated pyridines.1
It seems that such reactions at the C-17 position of
steroids present particular difficulties, and attempts to
improve the yield of the desired product by varying the
solvent were unsuccessful. The target 4-pyridyl deriva-
tive 5 was readily separable by chromatography from
the unwanted epimer (4) following deacetylation of the
mixture of 10 and 11. Alcohols 16 + 17 and 18 were
similarly obtained by deacetylation of the mixture of
6,17
Resu lts a n d Discu ssion
Syn th esis. The synthesis of a 17-substituted steroid
with the pyridyl residue in the R configuration presented
a synthetic challenge. According to literature prece-
dents, electrophilic, radical, or nucleophilic attack at the
C-17 position should occur preferentially at the less
sterically crowded R face, owing to hindrance by the 18-
methyl substituent, to give a product with the 17â
2
-pyridyl derivatives 12 and 13 and the separated (2-
pyridyl)thio derivative 14, respectively.
Stereochemistry at C-17 for the 16,17-reduced deriva-
1
tives was reflected in their H NMR spectra. Thus for
the 17â isomers the signal for H-17 appeared as an
apparent triplet (J ) 9.6 Hz), at higher field than for
H-17 in the 17R epimers which appeared as a double
doublet (J ) 8.3 and ∼1.0 Hz). Additionally, Me-18 in
configuration. Indeed this was seen in the reduction of
1
1
and 2 to the 17â-pyridyl derivatives 3 and 4.
A
1
7â isomers appeared at δ ∼0.50, whereas in the 17R
synthetic strategy in which the bond between C-17 and
the pyridyl residue was formed last was therefore
envisaged. Unsuccessful approaches (results not shown)
involved attempted inversion of configuration at C-17.
analogues it was more deshielded, appearing at δ ∼1.00.
The synthesis of a 16,17-epoxide from the acetate of
1 proved intractable owing to the presence of the 5,6-
double bond. Hence a precursor (19), lacking the 5,6-
double bond but otherwise analogous to 1, was synthe-
sized (Scheme 3) via the enol triflate 20 of epiandros-
terone acetate using previously described methodology.1
The synthesis (Scheme 3) of the target epoxide (6) from
19 was still a problem. The use of 1 equiv of MMPP
(monoperoxyphthalic acid magnesium salt) as oxidant
gave a mixture of 6, the N-oxide of 19, and the epoxide
N-oxide 22. Since these products were difficult to
separate, the reaction was driven to completion using
9
Using Mitsonobu’s conditions 3â-acetoxy-17â-hydroxyan-
1
0
drost-5-ene was treated with ethyl cyanoacetate or
4
-pyridylzinc chloride11 as the source of carbon nucleo-
phile. Alternatively the 17â-mesylate was treated with
diethyl malonate or ethyl cyanoacetate in the presence
of CsF.12 Another approach proceeded via the 17R-iodide
7
as a potential source of a free radical which, being
planar, might trap a suitable reactive species on the less
hindered R face to give a product of the required
stereochemistry. The synthesis of 7 (Scheme 1) made
4
equiv of MMPP with a view to subsequent selective
Sch em e 1a
reduction of the N-oxide residue in 22. However, the use
of borohydride ion-exchange resin combined with Cu-
1
9
SO4, recently reported to reduce selectively amine
N-oxides in the presence of epoxides, regenerated 19.
Finally acetate 21 was selectively epoxidized (Scheme
2
0
3
) using Sarett’s reagent. Hydrolysis of the intermedi-
ate 23 gave the target compound 6 in 46% overall yield.
The stereochemistry for 6 (and for 22 and 23) was
assigned based on a literature precedent. For epoxides
a
21
(a) I2, PPh3, imidazole, PhMe.

Role of the 16,17-Double Bond in Inhibition
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27 5377
Sch em e 2a
a
(a) (i) (CO)2Cl2, CH2Cl2, THF, (ii) N-hydroxypyridine-2-thione Na salt, CH2Cl2, dark; (b) PyH, camphorsulfonic acid, hν; (c) LiAlH4,
THF.
Sch em e 3a
a
(a) Tf2O, base; (b) 3-PyBEt2, Pd(PPh3)2Cl2, THF, H2O, Na2CO3; (c) NaOH, H2O, MeOH; (d) MMPP, 4 equiv; (e) CrO3, PyH, CH2Cl2.
Ta ble 1. Enzyme Inhibition Data
within a five-membered ring, a trans orientation of the
oxygen produced a small shielding of adjacent methyl
protons while a cis relationship between epoxide oxygen
and methyl had virtually no effect relative to the
influence of a double bond. The signal for Me-18
appeared between δ 0.74 and 0.79 for epoxides 6, 22,
and 23 but at δ 1.00-1.02 for the 16,17-unsaturated
precursors 19 and 21. This likewise suggested a trans
orientation between the epoxide and the angular methyl
group. Also, as mentioned above, a â orientation of a
pyridyl ring in 16,17-reduced pyridyl steroids produced
a ∼0.5 ppm shielding of the Me-18 signal relative to a
progesterone
compd
17R-hydroxylase IC50 (nM)
type of inhibition
1
3
4
4a
irreversible
reversible
a
47
160a
27
5
reversible
6
19
260
5
irreversible
reversible
irreversible
irreversible
2
2
2
4
5
6
74
13
3
a
a
a
Values taken from ref 1.
1
7R-pyridyl substituent.
(Table 1). Compounds 16 + 17 and 18, byproducts of
the synthesis of 5, gave no inhibition at 1 µM (results
not shown). To minimize ambiguity in the interpretation
of the binding data (see below) which might arise by
comparing compounds with markedly differing inhibi-
tory potencies, a compound more similar in potency to
that of 5, but retaining the 16,17-double bond present
in 1, was selected to include in this study. The 3,5-diene
Finally, for comparison with compounds 3-5, the 17â-
-pyridyl derivative 24 was made by reduction of 19.
In h ibition of Hu m a n Testicu la r 17R-Hyd r oxy-
3
la se. A. Str u ctu r e-Activity Rela tion sh ip s. The 17R-
-pyridyl derivative 5 proved, as predicted, a potent
4
inhibitor of the target enzyme, with an IC50 for the
hydroxylase of 27 nM, only 6-7-fold less potent than 1

5
378 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27
J arman et al.
was unaffected by the substrate concentration (Figure
3
), with an IC50 value of 1.2 nM. The hydroxylase
reaction in the presence of steroidal compounds lacking
the 16,17-double bond (3-5, 24) showed essentially
linear progress curves (Figure 4). There was no increase
in the inhibitory effect after preincubation of the enzyme
with the compounds. The inhibition by 3, 5, 24 was
substantially reversed by dialysis (Figure 2) for 5 h and
completely reversed after 24 h. The inhibition by 5 was
competitive with respect to pregnenolone with a Ki of
0
1
.35 ( 0.11 nM (Figure 5) compared to the Km of 23 (
.7 nM.
Con clu d in g Rem a r k s
F igu r e 1. Time course for human 17R-hydroxylase activity
after (2) or without (4) preincubation of the enzyme with (a)
Previous studies showing tight irreversible binding
for 1 to murine cytochrome P45017R in contrast to
reversibility for potent nonsteroidal esters containing
the pyridyl residue could be interpreted in terms of a
requirement for the steroidal skeleton for such binding.
However the present study provides compelling evidence
that the double bond adjacent to the pyridyl residue is
the structural feature necessary for the apparently
irreversible binding to human cytochrome P45017R
shown by 1 and its congeners. Thus three such com-
pounds of varying potency have shown it. In contrast
four steroidal compounds with this double bond reduced
showed no evidence of irreversible binding. It remains
to be seen if, by prolonging inhibition of enzyme activity
compared with a reversible inhibitor, this structural
feature of 1 confers therapeutic advantage in the
treatment of carcinoma of the prostate.
1
(3 nM), (b) 19 (5 nM), (c) 25 (30 nM), or (d) 26 (3 nM); control
(O).
1
2
1
5, previously shown to be ca. 3-fold less potent than
, was selected for this purpose. It was only twice as
potent as 5, making it unlikely that any marked
difference in binding characteristics observed between
5
and 25 could be ascribed to differences in potency. Also
included in the binding study as a further example of
an inhibitor with the 16,17-double bond was the only
other compound in this series which had previously been
2
1
evaluated in vivo, namely, the 3-en-4-one analogue 26.
Exp er im en ta l Section
1
Ch em ica l Meth od s. H NMR spectra (250 MHz) (internal
Me
4
3
Si ) δ 0) were determined in CDCl (unless otherwise
indicated) using a Bruker AC 250 spectrometer. Infrared
spectra were determined with a Perkin-Elmer 1720X spec-
trometer. Fast atom bombardment mass spectra were deter-
mined with a VG ZAB-SE spectrometer, accurate masses being
determined with m-nitrobenzyl alcohol + NaCl as the matrix.
Melting points were determined with a Reichert micro-hot-
stage apparatus and are uncorrected. Chromatography refers
to column chromatography on silica gel (Merck Art. 15111)
with solvent indicated applied under positive pressure. Light
petroleum refers to the fraction with bp 60-80 °C. Elemental
analyses were determined by CHN Analysis Ltd., South
Wigston, Leicester, England.
3
â-Acetoxy-17r-iod oa n d r ost-5-en e (7). To a mixture of
â-acetoxy-17â-hydroxyandrost-5-ene (1.000 g, 3.00 mmol),
triphenylphosphine (3.156 g, 12.03 mmol), and imidazole
0.819 g, 12.03 mmol), in toluene (100 mL), under argon, was
added iodine (2.290 g, 9.02 mmol) in two portions. The reaction
mixture was stirred at 80 °C for 1 h and allowed to cool to
3
Turning to the potential role of epoxide formation, the
parent steroid 19 of the epoxide 6 was of similar potency
(
1
to that of 1, in line with our previous observation on
its 3R epimer. The epoxide was however markedly less
potent (IC50 260 nM), and the inhibitory activity was
not enhanced by preincubation with the enzyme (results
not shown). Hence the present study does not suggest
that such an epoxide is the ultimate inhibitory species.
B. Bin d in g Ch a r a cter istics. Inhibition of the hu-
man enzyme by steroidal compounds with the 16,17-
ene structure (1, 19, 25, 26) was slow to occur and
appeared enhanced by preincubation with the enzyme
2 3
room temperature. A saturated aqueous Na SO solution (40
mL) was then added and the resulting mixture stirred until
all the solids had dissolved. Ethyl acetate (100 mL) was added,
and the organic phase was washed with saturated aqueous
NaHCO
3
(100 mL) and brine (50 mL), dried over Na
2
SO
4
, and
concentrated. Chromatography, on elution with Et
2
O-light
petroleum (1:25), afforded iodide 7 (1.18 g, 90%): mp 164-
165 °C; νmax 2942, 2903, 2890, 1732, 1440, 1376, 1250, 1200,
-
1 1
1036 cm ; H NMR δ 0.86 (s, 3, H-19), 1.04 (s, 3, H-18), 2.05
s, 3, CH CO), 4.38 (dd, 1, J 16,17 ) 7.0 Hz, J 16′,17 )1.0 Hz,
H-17â), 4.63 (m, 1, H-3R), 5.40 (m, 1, H-6); HRMS calcd for
(
3
(
Figure 1). Once inhibited, there was no recovery of
activity after 24 h of dialysis (Figure 2). This parallels
+
C
21
H
31
O
2
INa 465.1267 (M + Na) , found 465.1260. Anal.
I) H, N; C: calcd, 57.02; found, 56.59.
â-Acetoxy-17-(4-p yr id yl)a n d r ost-5-en e (10(r) + 11(â)),
5
findings with the testicular enzyme from WHT mice
(C
21
H
31
O
2
and suggests irreversible inhibition of the enzyme by
these compounds. Following preincubation of 1 with the
human testicular cytochrome P45017R, the inhibition
3
3â-Acet oxy-17-(2-p yr id yl)a n d r ost -5-en e (12(r) + 13(â)),
a n d 3â-Acetoxy-17-(2-th iop yr id yl)a n d r ost-5-en e (14(r) +

Role of the 16,17-Double Bond in Inhibition
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27 5379
F igu r e 2. Effect of dialysis on the inhibition of human 17R-hydroxylase by compounds 1 (3 nM), 3 (150 nM), 5 (150 nM), 19 (5
nM), 24 (300 nM), 25 (30 nM), and 26 (3 nM).
F igu r e 3. Human 17R-hydroxylase activity after preincuba-
F igu r e 5. Plot of IC50 values for 5 vs pregnenolone concentra-
tion with varying concentrations of 1 measured with two
tion for human 17R-hydroxylase.
concentrations of pregnenolone.
2 2
taken up in CH Cl (5 mL), and N-hydroxypyridine-2-thione
sodium salt (99.3 mg, 0.666 mmol) was added. The reaction
mixture was shielded from light and stirred for 2 h at room
temperature. To this mixture was added a solution of pyridine
(
263.4 mg, 270 µL, 3.33 mmol) and camphorsulfonic acid (777.6
mg, 3.33 mmol) in CH Cl (5 mL). The resulting solution was
irradiated (250 W, visible lamp) at ambient temperature for 1
h. Then CH Cl (50 mL) and saturated aqueous NaHCO (50
mL) were added, and the organic phase was dried (Na SO
and evaporated. Chromatography, on elution with Et O-light
2
2
2
2
3
2
4
)
2
petroleum (1:7), yielded first 9 (30 mg, 17%) followed by a
mixture (20:1) of 14 and 15 (70 mg, 30%), separated by flash
chromatography to give pure 14 (66 mg): H NMR for 14 δ
0
1
3
.97 (s, 3, H-19), 1.03 (s, 3, H-18), 2.04 (s, 3, CH CO), 4.10
(
5
dd, 1, J 16,17 ) 8.3 Hz, J 16′,17 ) 1.5 Hz, H-17â), 4.62 (m, 1, H-3R),
.39 (m, 1, H-6), 6.95 (dd, 1, J 4,5 ) 7.7 Hz, J 5,6 ) 5.0 Hz, pyridyl
H-5), 7.17 (d, 1, J 3,4 ) 7.7 Hz, pyridyl H-3), 7.45 (td, 1, J 4,6
1
for C26
)
F igu r e 4. Time course for human 17R-hydroxylase after (solid
symbols) or without (open symbols) preincubation of the
enzyme with (a) 4 (400 nM) or 5 (150 nM) and (b) 3 (150 nM)
or 24 (150 nM).
.8 Hz, pyridyl H-4), 8.40 (dd, 1, pyridyl H-6); HRMS calcd
+
1
2
H36NO S 426.2467 (M + H) , found 426.2460; H NMR
for 15 δ 0.84 (s, 3, H-18), 1.04 (s, 3H, H-19), 3.78 (t, J ) 9.5
Hz, H-17R). Further elution with Et O-light petroleum (1:5)
afforded a mixture (9:1) of 12 and 13 (25 mg, 11.5%): m/z 394
2
1
5(â)). 3â-Acetoxyetienic acid (200 mg, 0.555 mmol) was
dissolved in dry CH Cl (4 mL) and treated with oxalyl chloride
1.665 mL, 2 M solution in CH Cl ) and 1 drop of DMF. The
+
1
2
2
(M + H) ; H NMR for 12 δ 1.01 (s, 3, H-19), 1.02 (s, 3, H-18),
(
2
2
2.03 (s, 3, CH
3
CO), 3.10 (dd, 1, J 16,17 ) 8.3 Hz, J 16′17 ) 3.1 Hz,
mixture was stirred at room temperature for 2 h and then
concentrated, the residue dissolved in toluene, and the solution
concentrated again. The crude acid chloride (0.555 mmol) was
H-17â), 4.57 (m, 1, H-3R), 5.39 (m, 1, H-6), 7.09 (m, 2, pyridyl
H-3, H-5), 7.56 (td, 1, J 3,4 ) J 4,5 ) 7.6 Hz, J 4,6 ) 1.8 Hz, pyridyl
H-4), 8.56 (dd, 1, J 5,6 ) 4.7 Hz, pyridyl H-6); H NMR for 13 δ
1

5
380 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27
.50 (s, 3, H-18), 2.87 (t, 1, J ) 9.6 Hz, H-17R). Further elution
J arman et al.
0
17â-(N-Oxo-3-p yr id yl)-16,17r-ep oxy-5r-a n d r ost -3â-ol
(22). To a solution of alcohol 19 (100 mg, 0.284 mmol) in CH -
2
Cl (10 mL) and methanol (3 mL) was added magnesium
2
monoperoxyphthalate hexahydrate in one portion (703.6 mg,
1.138 mmol), and the resulting mixture was stirred at room
with Et
1 (25 mg, 11.5%): m/z 394 (M + H) ; H NMR for 10 δ 1.00
s, 3, H-19), 1.01 (s, 3, H-18), 2.02 (s, 3, CH CO), 2.91 (dd, 1,
16,17 ) 8.3 Hz, J 16′,17 ) 0.8 Hz, H-17â), 4.57 (m, 1, H-3R), 5.39
m, 1, H-6), 7.01 (AA′MM′, 2, pyridyl H-3, H-5), 8.48 (AA′MM′,
2
O-light petroleum (2:1) gave a mixture (9:1) of 10 and
+
1
1
(
J
(
3
2 2
temperature for 24 h. Water (50 mL) and CH Cl (100 mL)
were added, and the organic phase was separated, dried (Na2-
1
2
, pyridyl H-2, H-6); H NMR for 11 δ 0.49 (s, 3, H-18), 2.67
t, 1, J ) 9.7 Hz, H-17R), 7.13 (AA′MM′, 2, pyridyl H-3, H-5).
7r-(4-P yr id yl)a n d r ost-5-en -3â-ol (5). A solution of the
above mixture of acetates 10 and 11 (25 mg, 0.064 mmol) in
THF (2 mL) was treated with LiAlH (63.5 µL, 1 M solution
in THF) and then, after 5 min, with saturated aqueous Na
SO solution (19 µL). Solids were filtered off and washed with
CH Cl , and the organic phase was concentrated. Chromatog-
raphy, on elution with CH Cl -Et O (1:2), yielded pure 5 (18
mg, 81%): mp 210-212 °C; νmax 3349, 2928, 2853, 1601, 1417,
SO
Et
60-262 °C; νmax 3369, 2962, 2850, 1429, 1042, 733, 682 cm
4
), and concentrated. Chromatography, on elution with
(
2
O-CH Cl -MeOH (1:4:0.2), afforded 22 (95 mg, 87%): mp
2
2
1
-1
2
;
1
H NMR δ 0.79 (s, 3, H-18), 0.83 (s, 3, H-19), 3.60 (m, 1, H-16),
4
3
.61 (m, 1, H-3R), 7.27 (m, 2, pyridyl H-4, H-5), 8.18 (m, 1,
2
-
pyridyl H-6), 8.24 (m, 1, pyridyl H-2); HRMS calcd for C24
H
34
-
)
4
+
NO
C, H, N.
â-Acet oxy-17â-(3-p yr id yl)-16,17r-ep oxy-5r-a n d r os-
ta n e (23). Dry CrO (508 mg, 5.082 mmol) was added to a
stirred solution of dry pyridine (804 mg, 822 µL, 10.164 mmol)
in dry CH Cl (15 mL) at 5 °C under argon. After the mixture
3
384.2539 (M + H) , found 384.2530. Anal. (C24H33NO
3
2
2
2
2
2
3
-
1
1
3
1
062, 732 cm ; H NMR δ 1.00 (s, 6, H-18, H-19), 2.91 (d, 1,
J ) 8.1 Hz, H-17â), 3.49 (m, 1, H-3R), 5.36 (m, 1, H-6), 7.02
AA′MM′, 2, pyridyl H-3, H-5), 8.48 (AA′MM′, 2, pyridyl H-2,
H-6). Anal. (C24 33NO) C, H, N.
7-(2-P yr id yl)a n d r ost-5-en -3â-ol (16(r) + 17(â)). The
foregoing method was followed, using the mixture of acetates
2 and 13 (50 mg, 0.13 mmol). Chromatography, on elution
with Et O-light petroleum (1:1), yielded a mixture (9:1) of 16
2
2
(
was stirred at 5 °C for 1 h, acetate 21 (200 mg, 0.508 mmol)
was added and stirring was continued at room temperature
for 3 days; then the mixture was filtered and the solid washed
H
1
with CH
was extracted several times with ether. The combined ether
extracts were washed with saturated aqueous NaHCO , dried
Na SO ), and concentrated, and traces of pyridine removed
2 2
Cl . The filtrates were concentrated, and the residue
1
2
3
1
and 17 (40 mg, 87.5%): H NMR for 16 δ 1.00 (s, 6, H-18,
H-19), 3.10 (dd, 1, J 16,17 ) 8.5 Hz, J 16′17 ) 2.9 Hz, H-17â), 3.48
(
2
4
by codistillation with toluene. Chromatography, on elution
(m, 1, H-3R), 5.36 (m, 1, H-6), 7.08 (m, 2, pyridyl H-3, H-5),
with Et O-light petroleum (2:1), afforded acetate 23 (115 mg,
2
7
8
.56 (td, 1, J 3,4 ) J 4,5 ) 7.7 Hz, J 4,6 ) 1.9 Hz, pyridyl H-4),
.55 (dd, 1, J 5,6 ) 4.0 Hz, pyridyl H-6); HRMS calcd for C24
55%): mp 166-170 °C; ν
max
2924, 2853, 1731, 1250, 1029, 718
-
1 1
H
33
-
cm ; H NMR δ 0.74 (s, 3, H-18), 0.83 (s, 3, H-19), 2.02 (s, 3,
+
1
NO 352.2640 (M + H) , found 352.2640; H NMR for 17 δ 0.49
s, 3, H-18), 2.83 (t, 1, J ) 9.6 Hz, H-17R).
7r-(2-Th iop yr id yl)a n d r ost-5-en -3â-ol (18). The forego-
ing method was followed, using acetate 14 (90 mg, 0.211
mmol). Chromatography, on elution with Et O-light petro-
CH CO), 3.62 (s, 1, H-16), 4.69 (m, 1, H-3R), 7.27 (dd, 1, J
)
3
4,5
(
7.8 Hz, J 5,6 ) 5.0 Hz, pyridyl H-5), 7.70 (dt, 1, J 2,4 ) J 4,6 ) 1.9
Hz, pyridyl H-4), 8.55 (dd, 1, pyridyl H-6), 8.60 (d, 1, pyridyl
1
+
H-2); HRMS calcd for C26
10.2690. Anal. (C26
75.53.
H
36NO
3
410.2695 (M + H) , found
4
3
H35NO ) H, N; C: calcd, 76.25; found,
2
1
leum (1:3), afforded 18 (80 mg, 99%): mp 145-146 °C; H NMR
δ 0.97 (s, 3, H-19), 1.02 (s, 3, H-18), 3.54 (m, 1, H-3R), 4.09
17â-(3-P yr id yl)-16,17r-ep oxy-5r-a n d r ost-3â-ol (6). To a
solution of acetate 23 (100 mg, 0.244 mmol) in methanol (2
mL) was added an aqueous solution of NaOH (10% w/v, 0.4
mL). The mixture was stirred at 80 °C for 5 min and then
(
6
dd, 1, J 16,17 ) 8.4 Hz, J 16′17 ) 1.7 Hz, H-17â), 5.37 (m, 1, H-6),
.96 (ddd, 1, J 4,5 ) 7.8 Hz, J 5,6 ) 4.9 Hz, J 3,5 ) 0.9 Hz, pyridyl
H-5), 7.17 (dd, 1, J 3,4 ) 7.8 Hz, pyridyl H-3), 7.46 (td, 1, J 4,6
1
for C24
)
allowed to cool, poured into water, and extracted with CH
2
Cl
2
.9 Hz, pyridyl H-4), 8.41 (dd, 1, pyridyl H-6); HRMS calcd
+
(3 × 25 mL). The organic extracts were dried (Na CO ) and
H
34NOS 384.2361 (M + H) , found 384.2371. Anal.
2
3
concentrated. Chromatography, on elution with Et
2
O-CH
2
Cl
2
(C
24
H
33NOS) H, N, S; C: calcd, 75.15; found, 74.51.
â-Acetoxy-17-(3-p yr id yl)-5r-a n d r ost-16-en e (21). Di-
ethyl(3-pyridyl)borane (411.5 mg, 2.799 mmol) was added to
(
2
3:1), afforded 6 (76 mg, 84%): mp 236-238 °C; νmax 3338,
3
-
1 1
978, 2930, 2856, 1451, 1057, 716 cm ; H NMR δ 0.76 (s, 3,
H-18), 0.83 (s, 3, H-19), 3.63 (m, 2, H-3R, H-16), 7.27 (m, 1,
2
2
a stirred solution of enol triflate 20 (1 g, 2.153 mmol) in THF
10 mL) containing bis(triphenylphosphine)palladium(II) chlo-
ride (15.1 mg, 0.022 mmol). An aqueous solution of Na CO (2
M, 4 mL) was then added, and the stirred mixture was heated
at 80 °C for 1 h and then partitioned between Et O and H O.
The organic phase was dried (Na CO ), filtered through a short
column of silica gel, and concentrated. Chromatography, on
elution with Et O-light petroleum (2:1), afforded 21 (762 mg,
0%): mp 152-153 °C; νmax 2937, 1729, 1599, 1562, 1370,
pyridyl H-5), 7.68 (m, 1, pyridyl H-4), 8.56 (m, 1, pyridyl H-6),
(
8
(
.62 (m, 1, pyridyl H-2); HRMS calcd for C24
M + H) , found 368.2595. Anal. (C24H33NO
2
1
H
34NO
2
368.2590
) C, H, N.
7â-(3-P yr id yl)-5r-a n d r osta n -3â-ol (24). To a solution of
9 (351 mg, 1 mmol) in ethanol (60 mL) were added hydrazine
2
3
+
2
2
1
2
3
hydrate (1.6 mL, 5 mmol) and AcOH (1.0 mL). The mixture
was heated at 80 °C for 24 h while a stream of air was passed
through the solution. After cooling, the mixture was parti-
2
9
1
1
tioned between CH
the organic phase was dried (Na
Chromatography, on elution with MeOH-CH
forded 24 (300 mg, 85%): mp 226-228 °C; νmax 3361, 2926,
2
Cl
2
and saturated aqueous NaHCO
CO ) and concentrated.
Cl (1:50), af-
3
, and
-
1 1
250, 1150, 1132, 1080, 827 cm ; H NMR δ 0.87 (s, 3, H-19),
.00 (s, 3, H-18), 2.03 (s, 3, CH CO), 4.68 (m, 1, H-3R), 5.99
2
3
3
2
2
(
m, 1, H-16), 7.24 (dd, 1, J 4,5 ) 7.8 Hz, J 5,6 ) 5.0 Hz, pyridyl
H-5), 7.67 (d, 1, pyridyl H-4), 8.46 (d, 1, pyridyl H-6), 8.62 (s,
, pyridyl H-2). Anal. (C26 ) C, H, N.
7-(3-P yr id yl)-5r-a n d r ost-16-en -3â-ol (19). To a solution
-1
1
2
845, 1449, 1378, 1081, 1045, 716 cm ; H NMR δ 0.47 (s, 3,
1
H35NO
2
H-18), 0.81 (s, 3, H-19), 2.67 (t, 1, J ) 9.6 Hz, H-17R), 3.62
(m, 1, H-3R), 7.22 (dd, 1, J 4,5 ) 7.7 Hz, J 5,6 ) 4.8 Hz, pyridyl
H-5), 7.50 (d, 1, pyridyl H-4), 8.39 (m, 2, pyridyl H-2, H-6).
Anal. (C H35NO) C, H, N.
1
of acetate 21 (400 mg, 1.016 mmol) in methanol (5 mL) was
added an aqueous solution of NaOH (10% w/v, 1 mL). The
mixture was stirred at 80 °C for 5 min and then allowed to
2
4
En zym e P r ep a r a tion a n d Assa y P r oced u r es for th e
7r-H yd r oxyla se Act ivit y of H u m a n Cyt och r om e
P 45017r. The microsomal fraction from human testes was
cool, poured into water, and extracted with CH
mL). The organic extracts were dried (Na CO ) and concen-
trated. Chromatography, on elution with Et O-CH Cl (1:1),
afforded 19 (300 mg, 84%): mp 215-216 °C; νmax 3347, 2930,
2
Cl
2
(3 × 50
1
2
3
1
2
2
2
prepared as described elsewhere. The standard assay with
progesterone as substrate was also as described therein. To
test the reversibility of the inhibition, the compounds were
incubated with the enzyme for 30 min at 37 °C before being
dialyzed against 250 volumes of 50 mM sodium phosphate
buffer (pH 7.4) containing 0.137 mM dithiothreitol, 0.274 mM
EDTA, 1.37% DMSO, and 3.6% glycerol. After 5 and 24 h,
samples were removed and warmed at 37 °C for 5-10 min,
853, 1449, 1042, 799 cm-1; H NMR δ 0.90 (s, 3, H-19), 1.02
s, 3, H-18), 3.62 (m, 1, H-3R), 5.99 (dd, 1, J 15,16 ) 3.2 Hz, J 15′16
1.8 Hz, H-16), 7.23 (dd, 1, J 4,5 ) 8.0 Hz, J 5,6 ) 4.7 Hz, pyridyl
H-5), 7.57 (dd, 1, J 2,4 ) 1.9 Hz, J 4,6 ) 1.6 Hz, pyridyl H-4),
.46 (dd, 1, pyridyl H-6), 8.62 (d, 1, pyridyl H-2). Anal. (C24
NO) C, H, N.
1
2
(
)
8
33
H -

Role of the 16,17-Double Bond in Inhibition
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 27 5381
3
before the assay was started by the addition of [ H]progest-
erone, NADPH, and its regenerating system. The final assay
mixture was the same as in the standard assays.
(6) 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
1
7R-Hydroxylase/C17-20 Lyase). Biochem. Biophys. Res. Com-
For the studies on the type of inhibition shown by 1 and 5,
the assays used the substrate pregnenolone at concentrations
up to 3 µM with specific activity of 3-5 mCi/µmol. The rest of
the assay mixture was the same as in the standard assays.
The cold mix added to stop the reaction contained preg-
nenolone, 17R-hydroxypregnenolone, and dehydroepiandros-
terone. The HPLC separation used a 15-cm 5-µm Apex C1
column with an Uptight guard column packed with 75-125
µm of LC-Porosil silica (Waters Assoc.). The mobile phase was
2% methanol at a flow rate of 1.4 mL/min. The effluent was
monitored at 214 nm before being mixed with Ecoscint A
National Diagnostics Ltd.) containing 25% acetonitrile flowing
at 1.1 mL/min and monitored for H using a Berthold LB506C
detector. Activity was measured as the production of 17R-
hydroxypregnenolone. No dehydroepiandrosterone nor an-
drost-5-ene-3â,17â-diol were formed, but two unidentified
peaks of long retention time were eluted.
mun. 1989, 162, 1571-1577.
(
7) Angelastro, M. R.; Marquart, A. L.; Weintraub, P. M.; Gates, C.
A.; Laughlin, M. E.; Blohm, T. R.; Peet, N. P. Time-dependent
Inactivation of Steroid C17(20) Lyase by 17â-Cyclopropyl Ether-
substituted Steroids. Bioorg. Med. Chem. Lett. 1996, 6, 97-100.
8) Burke, D. F. Ph.D. Thesis, University of London, 1994.
9) Dodge, J . A.; Lugar, C. W., III. Alcohol Inversion of 17â-Steroids.
Bioorg. Med. Chem. Lett. 1996, 6, 1-2.
(
(
(
10) Kurihara, T.; Sugizaki, M.; Kime, I.; Wada, M.; Mitsunobu, O.
Stereospecific and Stereoselective Reactions V. Alkylation of
Active Methylene Compounds by the Use of Alcohols, Diethyl
Azodicarboxylate and Triphenylphosphine. Bull. Chem. Soc. J pn.
1981, 54, 2107-2112.
(11) Rollin, P. One-step Stereospecific Conversion of Alcohols Into
Dithiocarbamates: A Smooth Pathway for the Introduction of a
Sulphur Functionality. Tetrahedron Lett. 1986, 27, 4169-4170.
12) Sato, T.; Otera, J . CsF in Organic Synthesis. Malonic Ester
Synthesis Revisited for Stereoselective Carbon-Carbon Bond
Formation. J . Org. Chem. 1995, 60, 2627-2629.
13) Garreg, P.; J ohansson, R.; Ortega, C.; Samuelsson, B. Novel
Reagent System for Converting a Hydroxy Group Into an Iodo
Group in Carbohydrates with Inversion of Configuration. Part
5
(
3
(
The K
m
for pregnenolone was calculated by fitting by
nonlinear regression the initial rates at varying substrate
concentrations to the Michaelis-Menton equation and was 23
1.7 nM (mean ( SE from four determinations). IC50 values
(
(
1
were calculated as before.
3
. J . Chem Soc., Perkin Trans. 1 1982, 681-683.
(
14) Llera, J . M.; Fraser-Reid, B. An Expeditious Route to the
Northern Part of Retigeranic Acid A from (R)-(-)-Carvone. J .
Org. Chem. 1989, 54, 5544-5548.
Ack n ow led gm en t. This work was supported by a
grant from the Cancer Research Campaign. Financial
support to J . M. Llera from the Commission of the
European Communities under the Human Capital and
Mobility Scheme is also gratefully acknowledged. We
thank Dr. G. A. Potter for helpful discussions and
preliminary experiments on the synthesis of 5 and Mr.
Mike Cocksedge of the University of London Intercol-
legiate Research Service for mass spectra.
(15) Minisci, F.; Vismara, E.; Fontana, F. Recent Developments of
Free-Radical Substitutions of Heteroaromatic Bases. Hetero-
cycles 1989, 28, 489-519.
(
16) Barton, D. H. R.; Garcia, B.; Togo, H.; Zard, S. Z. Radical
Decarboxylative Addition onto Protonated Heteroaromatic (and
Related) Compounds. Tetrahedron Lett. 1986, 27, 1327-1330.
17) Catagnino, E.; Corsano, S.; Barton, D. H. R.; Zard, S. Z.
Decarboxylative Radical Addition onto Protonated Heteroaro-
matic Systems Including Purine Bases. Tetrahedron Lett. 1986,
(
2
7, 6337-6338.
Refer en ces
(
18) Crich, D.; Quintero, L. Radical Chemistry Associated with the
(
(
(
1) Potter, G. A.; Barrie, S. E.; J arman, M.; Rowlands, M. G. Novel
Steroidal Inhibitors of Human Cytochrome P45017R (17R-Hy-
droxylase-C17,20-lyase): Potential Agents for the Treatment of
Prostatic Cancer. J . Med. Chem. 1995, 38, 2463-2471.
Thiocarbonyl Group. Chem. Rev. 1989, 89, 1413-1431.
(19) Bo, T.; Ahn, J . H.; Yoon, N. M. Selective Deoxygenation of Amine
N-Oxides Using Borohydride Exchange Resin-Copper Sulfate in
Methanol. Synthesis 1996, 324-326.
20) (a) Poos, G. I.; Arth, G. E.; Begler, R. E.; Sarett, L. H. Approaches
to the Total Synthesis of Adrenal Steroids. V. 4b-Methyl-7-
ethylenedioxy-1,2,3,4,4aR,4b,5,6,7,8,10,10aâ-dodecahydrophenan-
threne-4â-ol-1-one and Related Tricyclic Derivatives. J . Am.
Chem. Soc. 1953, 75, 422-429. (b) Miljkovic, D.; Gasi, K.;
Kindjer, M.; Stankovic, S.; Argay, G. Synthesis, Crystal and
Molecular Structure, and Hyperconjugation of the Isomeric 17,-
2) Barrie, S. E.; Potter, G. A.; Goddard, P. M.; Haynes, B. P.;
Dowsett, M.; J arman, M. Pharmacology of Novel Steroidal
Inhibitors of Cytochrome P45017R (17R-hydroxylase/C17-20
Lyase). J . Steroid Biochem. Mol. Biol. 1994, 50, 267-273.
3) Rowlands, M. G.; Barrie, S. E.; Chan, F.; Houghton, J .; J arman,
M.; McCague, R.; Potter, G. A. Esters of 3-Pyridylacetic Acid
that Combine Potent Inhibition of 17R-Hydroxylase/C17,20Lyase
(
(Cytochrome P45017R) with Resistance to Esterase Hydrolysis.
J . Med. Chem. 1996, 38, 4191-4197.
2
0-Epoxy-17-picolyl Derivatives of 5-Androstene and 5R-Andros-
(
4) Chan, F. C. Y.; Potter, G. A.; Barrie, S. E.; Haynes, B. P.;
Rowlands, M. G.; Houghton, J .; J arman, M. 3- and 4-(Pyridyl)-
alkyl Adamantanecarboxylates: Inhibitors of Human Cyto-
chrome P45017R (17R-Hydroxylase-C17,20-Lyase). Potential Non-
steroidal Agents for the Treatment of Prostatic Cancer. J . Med.
Chem. 1996, 39, 3319-3323.
tane. Tetrahedron 1987, 43, 631-641.
(
21) Pearson, A. J .; Chen, Y.-S.; Han, G. R.; Hsu, S.-Y.; Ray, T. A
New Method for the Oxidation of Alkenes to Enones. An Efficient
5
Synthesis of ∆ -7-Oxo Steroids. J . Chem. Soc., Perkin Trans. 1
1985, 267-273.
(
5) Barrie, S. E.; Haynes, B. P.; Potter, G. A.; Chan, F. C. Y.;
Goddard, P. M.; Dowsett, M.; J arman, M. Biochemistry and
Pharmacokinetics of Potent Nonsteroidal Cytochrome P45017R
Inhibitors. J . Steroid Biochem. Mol. Biol. 1997, 60, 347-351.
(22) Harnisch, W.; Morera, E.; Ortar, G. A Novel Approach to
Cardenolides. J . Org. Chem. 1985, 50, 1990-1992.
J M981017J