18-Vinyldeoxycorticosterone: a potent inhibitor of the bovine cytochrome P-450(11β)

Article abstract of DOI:10.1016/S0968-0896(98)00106-0

18-Vinylprogesterone (18-VP) and 18-ethynylprogesterone (18-EP) have proved to be potent suicide inhibitors of P-450(11β), the last enzyme of aldosterone biosynthesis (Delorme, C.; Piffeteau, A.; Viger, A.; Marquet, A. Eur. J. Biochem. 1995, 232, 247; Del

Full text of DOI:10.1016/S0968-0896(98)00106-0

BIOORGANIC &
MEDICINAL
CHEMISTRY
Bioorganic & Medicinal Chemistry 6 (1998) 1781±1788
18-Vinyldeoxycorticosterone: a Potent Inhibitor of the
Bovine Cytochrome P-45011ꢀ
Elisabeth Davioud,* Annie Pi€eteau, Cecile Delorme, Suzy Coustal
and Andree Marquety
Â
Laboratoire de Chimie Organique Biologique, Universite Pierre et Marie Curie, CNRS UMR 7613, 4, place Jussieu,
75252 Paris Cedex, 05, France
Received 3 March 1998; accepted 11 May 1998
AbstractÐ18-Vinylprogesterone (18-VP) and 18-ethynylprogesterone (18-EP) have proved to be potent suicide inhibi-
tors of P-45011b, the last enzyme of aldosterone biosynthesis (Delorme, C.; Pi€eteau, A.; Viger, A.; Marquet, A. Eur. J.
Biochem. 1995, 232, 247; Delorme, C.; Pi€eteau, A.; Sobrio, F.; Marquet, A. Eur. J. Biochem. 1997, 248, 252). This
paper describes the synthesis of 18-vinyldeoxycorticosterone (18-VDOC), an analogue of deoxycorticosterone (DOC),
the physiological substrate of the enzyme, and the evaluation of its reversible inhibiting properties for deoxy-
corticosterone and corticosterone oxidation by the bovine enzyme. 18-VDOC has been obtained by hydroxylation at
C-21 of a 18-VP precursor. Its reversible Ki values are, respectively, 0.3 mM for the 11b-hydroxylation and 0.8 mM for
the 18-hydroxylation. Hence, 18-VDOC is the strongest competitive inhibitor of bovine P-45011b described so far, but
in contrast with 18-VP, it does not inhibit more eciently the 18-hydroxylation than the 11-hydroxylation. # 1998
Elsevier Science Ltd. All rights reserved.
Introduction
mechanistic tools to study the catalytic properties of this
interesting P-450 which carries out three successive
hydroxylations within the same active site.7
P-45011b, the last enzyme of aldosterone biosynthesis1
(Scheme 1) plays a key role in the regulation of aldoster-
one production and the fundamental studies of this
enzyme have a strong biological relevance with potential
therapeutic applications. It was chosen as a target for the
discovery of inhibitors which could decrease aldosterone
overproduction occurring in pathological conditions.2
Furthermore, they have potential therapeutic interest.
We have shown that they inhibit more eciently the 18-
hydroxylation steps (leading to aldosterone) than the
11b-hydroxylation step,7 a property which was also
observed in bovine adrenocortical cells.9 In addition,
they behave as potent antagonists of the mineralo-
corticoid receptor.10
Several of these inhibitors, progesterone derivatives
designed as suicide-substrates, have been synthesized
and tested in our group.3±6 Two of them, 18-vinyl pro-
gesterone (18-VP) and 18-ethynyl progesterone (18-EP)
(Scheme 2) were found to be very ecient inhibitors of
P-45011b either in rat adrenal cell-free extracts4 or with
the puri®ed bovine enzyme.7,8 They appeared as useful
This class of inhibitors has also been studied, in particular
by the Merrell group, and their biological properties have
been recently reviewed.11 A comparison of the enzyme
inhibitory activity of 18-EP and 18-EDOC (18-ethynyl
deoxycorticosterone), hydroxylated at C-21 as is the phy-
siological substrate of P-45011b, has been carried out using
rat adrenal glomerulosa homogenates.11 The potency of
18-EDOC as inhibitor of the 18-hydroxylation has also
been examined in calf adrenal cells by Yamakita et al.12
Key words: 18-VDOC; P-45011b; suicide-inhibitors; anti-
aldosterone.
*Present address: Institut de Biologie de Lille et Institut Pasteur
de Lille, Departement V, BP 447, 1 rue du professeur Calmette,
59021 Lille Cedex, France.
yCorresponding author. Tel: 33 1 44 27 55 35; Fax: 33 1 44 27
71 50.
Very scarce data were given on 18-vinyl deoxy-
corticosterone (18-VDOC). We have previously found
0968-0896/98/$Ðsee front matter # 1998 Elsevier Science Ltd. All rights reserved
PII: S0968-0896(98)00106-0

1782
E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
to give 4, on which the key step, the hydroxylation at
C-21 was attempted.
There are several well established procedures for this
reaction, frequently reported on normal steroids with
the 18-methyl group.13 We have tested three of them,
which failed when applied to 4, bearing an allyl group at
C-13.
Although the direct acetoxylation with lead tetracetate
in acetic acid of 18-EP was described (with no charac-
terization and no mention of the yield) in a patent,14
we could obtain neither 18-VDOC nor 18-EDOC starting,
respectively, from 18-VP or 18-EP. In both cases, the
NMR spectrum of the crude product revealed, beside
the expected singulets for the C-21 methylene and C-21
acetoxy groups, the disappearance of the vinyl and
ethynyl protons, respectively. This suggests the addition
of an acetate group on the terminal unsaturation, a
reaction already described on other examples.15,16
Scheme 1.
that 18-VP and 18-EP behave as suicide inactivators and
that 18-VP was a ®vefold better inhibitor than 18-EP, as
re¯ected by the values of the apparent bimolecular rate
constants ki/Ki for time-dependent inactivation.8 Hence
we thought that it was useful to obtain more data on the
properties of 18-VDOC.
We then tried to prepare the 21-bromoketone, using
phenyltrimethyl ammonium tribromide, a reagent which
is unreactive towards double bonds.17 This led also to a
complex mixture resulting from the combination of
di€erent events, namely bromination at C-17 as well as
C-21 and epimerisation of the side chain (the in¯uence
of the substituent at C-13 on the regioselectivity of the
enolisation and the con®gurational stability of the C-17
side chain has been already reported).18
In this paper we present a synthesis of 18-VDOC and
kinetic results concerning its inhibition properties. In
this work, only the reversible kinetic constants for cor-
ticosterone and aldosterone production catalyzed by the
puri®ed bovine P-45011b were measured.
Results
Chemistry: synthesis of 18-vinyl-deoxycorticosterone
Another possible method using a diazoketone was also
investigated. The b-ketoaldehyde 10 (hydroxymethylene
form) obtained by treatment of the methyl ketone 3 with
sodium hydride and ethyl formate, was converted into
the diazoketone 11, by reaction with sodium hydride
and tosylazide (Scheme 4).19,20
The synthesis that we have achieved, described in
Scheme 3, uses as starting material compound 2, an
intermediate in the synthesis of 18-vinyl-progesterone.3
After removal of the dioxolane group, the alcohol
function of 3 was protected with dihydropyrane in the
presence of the pyridinium p-toluene sulfonate (PPTS)
Attempts to introduce an acetate group by treatment of
the crude diazoketone with acetic acid were unsuccess-
ful. Products resulting from the participation of the
double bond of the 18-vinyl group were detected on the
NMR spectrum of the crude product. To favor the
substitution by the external nucleophile we performed
the reaction with KBr in the presence of p-toluene sul-
fonic acid (TsOH). The bromoketone 12 was obtained,
but as a mixture of 17a:17b isomers, that we were
unable to separate eciently.
The microbial hydroxylation of 18-VP by the Aspergil-
lus niger strain (ATCC 9142), which was reported to
hydroxylate at C-21 several progesterone derivatives,21
was also tried without success.
The method which allowed us to obtain 1, although
with a poor yield, relies on the peracid oxidation of a
Scheme 2.

E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
1783
Scheme 3. (a) pTosOH, acetone; (b) dihyropyrane, PPTS, CH2Cl2; (c) LDA, THF, Me3SiCl; (d) MMPP, ethanol; (e) tBuPh2SiCl,
DMAP, pyridine, CH2Cl2; (f) AcOH, H2O, THF; (g) N-Me piperidone, THF, Al (OiPr)3; (h) HCl, methanol.
silyl enol ether.18 Treatment of methyl ketone 4 with
LDA at 78 ꢀC followed by quenching with trimethyl-
chlorosilane a€orded the 20(21) silyl enol ether. In our
hands, the reaction was very capricious and the best
yield of enol ether was 37% (determined on the NMR
spectrum of the crude product). Epoxidation of this
crude product with magnesium monoperoxyphthalate
(MMPP) in ethanol22 gave the hydroxy ketone 5 with a
81% corrected yield from silyl enol ether, according to
the NMR spectrum of the crude product. The other
identi®ed compounds were the hydrolysis product of the
tetrahydropyranyloxy group of 5 and the two epoxides
6. Puri®cation at this stage was dicult and only par-
tially achieved. Silylation of the 21-hydroxyl group was
performed in the presence of tertiobutyldiphenylsilyl
chloride and DMAP. Pure 8 was obtained after depro-
tection of the 3-OH group. Oppenauer oxidation and
®nal deprotection of the 21-OH group led to the tar-
geted product 1.
Biochemical experiments
They were all carried out with the puri®ed bovine
reconstituted system that we have already described.7
In a preliminary assay, 18-VDOC was compared to 18-
VP for the reversible inhibition of the 11b-hydroxyla-
tion step, using an initial concentration of substrate
[14C-DOC] close to the Km value and di€erent inhibitor
concentrations. As shown in Figure 1, after 1 min in
presence of 18-VDOC, a marked decrease in corticos-
terone production (only 30% of the total activity
remaining) was observed at 1 mM of inhibitor. In con-
trast, under the same conditions, no inhibition was
detectable in the presence of 1 mM 18-VP and a con-
centration of 10 mM of 18-VP was needed to obtain a
signi®cant decrease (23%) in corticosterone formation.
Inhibition kinetics
For kinetic determinations, 11b- and 18-hydroxylation
reactions were studied by varying the concentration of
substrates, respectively [14C]11-deoxycorticosterone and
[3H]corticosterone, under initial velocity conditions as
already described (see Experimental section). For the
studies of reversible inhibition of the 11b-hydroxylation
of DOC, the quantity of corticosterone formed during
one to two minutes was examined with or without
addition of 18-VDOC at three di€erent concentra-
tions (0.125±1 mM). According to the initial velocity
Scheme 4. (a) NaH, C2H5, HCOOEt, THF; (b) TosN3, NaH,
C2H5OH, THF; (c) KBr, TsOH, H2O, CH3CN.

1784
E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
be also competitive with substrate (Fig. 3), with a Ki
value of 0.8‹0.13 mM (SD). It will be noted that the Ki for
both reactions are nearly identical. Under the same con-
ditions, the Km values for DOC and corticosterone were
1.97‹0.26 mM (SD) and 5.9‹0.18 mM (SD), respectively,
consistent with the previously reported values.7
Discussion
Among several 18-substituted progesterone derivatives
synthesized and tested in our group, 18-EP and 18-VP,
designed as suicide-substrates, proved to be potent
inhibitors of aldosterone biosynthesis when tested with
rat adrenal cell-free extracts,4 with puri®ed bovine P-
45011b7 as well as with bovine adrenocortical cells.9 Both
compounds, which were found to inhibit more strongly
the 18-hydroxylation step (aldosterone production) than
the 11-hydroxylation have potential pharmacological
interest as antimineralocorticoids. 18-VP is a better
inhibitor than 18-EP in both reactions, as revealed by
their respective reversible Ki values7 (30 mM and 110 mM
for 11b-hydroxylase activity and 5 mM and 30 mM for
18-hydroxylase activity, respectively), their ki/Ki ratio
Figure 1. Inhibition of the transformation of DOC to cortico-
sterone by 18-VP or 18-VDOC. 11b-Hydroxylase activity of
bovine P-45011b (20 nM) was measured in a reaction mixture
containing [14C]11-deoxycorticosterone as substrate (1.25 mM)
in the presence of inhibitors 18-VP or 18-VDOC at concentrations
of: 0.25 mM ( ), 0.5 mM ( ), 1 mM (&), or 10 mM ( ). 100% is
the production without inhibitor (&). Each activity is the aver-
age of duplicate measurements in the same experiment. Shown is
a representative example of two independent experiments.
1
for time-dependent inactivation8 (470 M 1.s for 18-
1
VP and 89 M 1 s
for 18-EP), and their e€ect on
data, 18-VDOC bound competitively with respect to the
substrate DOC (Fig. 2) and the dissociation constant
Ki was of 0.3‹0.15 mM (SD). The enzyme kinetics for
the reversible inhibition of 18-hydroxylation steps, that
is, the production of 18-OH-corticosterone and aldo-
sterone from corticosterone, were performed under
identical conditions. The inhibition has been found to
aldosterone production in adrenal cell-cultures.9
We have thus selected the 18-vinyl compound to
examine the in¯uence of an OH group at C-21 which is
present in the substrate (DOC), on the inhibiting
properties. Johnston et al. have considered the same
point and compared the inhibitory potency of 18-EP and
18-EDOC for aldosterone production with rat adrenal
Figure 2. Inhibition of 11b-hydroxylase activity of bovine P-
45011b by 18-VDOC. 11b-Hydroxylase activity of bovine P-
45011b (40 nM) was studied with [14C]11-deoxycorticosterone as
substrate (1.25±20 mM), without (~) or with 18-VDOC at
0.125 mM (&), 0.33 mM (*), or 1 mM (^). Shown is a repre-
sentative graph of at least two independent determinations.
Duplicate measurements were run for each experiment.
Figure 3. Inhibition of 18-hydroxylase activity of bovine P-
45011b by 18-VDOC. 18-Hydroxylase activity of bovine P-
45011b (450 nM) was studied with [3H]corticosterone as sub-
strate (2.5±50 mM), without (*) or with 18-VDOC at 0.33 mM
(&), or 1 mM (~). Shown is a representative graph of at least
two independent determinations. Duplicate measurements were
run for each experiment.

E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
1785
glomerulosa homogenates. With this nonpuri®ed sys-
tem, they found that 18-EP was more selective than
18-VP for inhibiting the C18-hydroxylation activity.
18-EDOC was found slightly more potent than 18-EP
(IC50 of 14 nM and 35 nM for aldosterone synthase
assay, respectively).11
(13C NMR) spectra were recorded on a Jeol GS X 400
or on a Bruker AC 200 at 100 MHz. Solutions in
chloroform-d (CDCl3) were used, with tetramethylsilane
as an internal standard, coupling constants have been
obtained by spin decoupling. Mass spectral data were
obtained by chemical ionization (CI NH3) or by elec-
tron impact (EI) on a Nermag R10-10C or R30-10
spectrometer. Infrared spectral data were recorded on a
Perkin 1420 (values in cm 1). Elemental analysis were
performed by the Service Regional de Microanalyse
(SIAR-Jussieu). All chemicals were purchased from
Aldrich, Janssen, and Sigma.
The reversible Ki values measured with our bovine pur-
i®ed reconstituted system revealed that 18-VDOC is
100-fold more ecient than 18-VP for inhibition of the
11b-hydroxylation step (0.3 mM and 30 mM, respec-
tively), but only sixfold more ecient than 18-VP for
the inhibition of the 18-hydroxylation step (0.8 mM and
5 mM, respectively). Thus, the presence of an OH group
at C-21 strongly decreased the selectivity observed with
18-VP. We have interpreted the selectivity of 18-VP, on
the basis of further experimental data, by postulating a
conformational change of the enzyme after the 11-
hydroxylation7 allowing the proper orientation of the 18
methyl group with respect to the heme, the inhibitor
showing a better anity for the second conformation.
Clearly, the hydroxyl group at C-21 in¯uences the
binding of the inhibitor.
3ꢀ-Hydroxy-18-vinyl-20,20-ethylenedioxypregn-5-ene (2).
was prepared in nine steps from pregnenolone, as pre-
viously described.3
3ꢀ-Hydroxy-18-vinylpregn-5-ene-20-one (3). The dioxo-
lane 2 (728 mg, 1.88 mmol) was dissolved in acetone
(68 mL). p-Toluene sulfonic acid (36 mg, 0.19 mmol) was
added. The mixture was stirred under argon, at room
temperature, for 210 min. Then, the solvent was evapo-
rated and the residue was extracted with CH2Cl2
(100 mL) and washed with water. Crystallization of the
crude product in CH2Cl2-isopropyl ether gave 167 mg of
white crystals. Flash chromatography of the residue in
CH2Cl2/acetone 1% gave pure 3 (310 mg). The total
yield was 74%: mp 156±157 ꢀC; IR 3600, 1690,
1630 cm 1; 1H NMR (400 MHz) 1.00 (s, 3H, H-19), 2.15
(s, 3H, H-21), 3.51 (m, 1H, H-3), 4.96 (2dd, 2H,
J=17.4, 8.7 and 1.6 Hz, -CH=CH2), 5.34 (m, 1H, H-6),
5.52 (m, 1H, -CH=CH2). Anal. C23H34O2 (C, H).
Because of the absence of a three-dimensional structure
of mammalian P-450 and despite their low sequence
homologies, four bacterial P-450s are classically used as
model for analysis of the substrate orientation in the
heme pocket.23 For instance, the modeling of cortico-
sterone binding in the substrate pocket of mouse P-
450coh24 (coumarin 7a-hydroxylase) has been based on
the three-dimensional structure of P450cam. It suggests
a direct interaction of an arginine residue ( 186 in P-
450cam) with the 21-OH of the steroid by hydrogen
bonding. It is plausible to assume that the interaction of
the 21-hydroxyl group of 18-VDOC with a critical amino
acid residue enhances its anity for the native con-
formation of the enzyme, hence decreasing the selectivity
for the inhibition of the 18-versus 11-hydroxylation.
3ꢀ-[(Tetrahydro-2H-pyran-2yl)oxy]-18-vinylpregn-5-ene-
20-one (4). The alcohol 3 (350 mg, 1.02 mmol) was dis-
solved in freshly distilled CH2Cl2 (10 mL) and kept
under argon. Dihydropyrane (560 mL, 6.13 mmol, 6
equiv) and PPTS (51 mg, 0.2 equiv) were added. The
mixture was stirred for 17 h at room temperature. Then
diethylether (70 mL) was added and the organic solution
was washed twice with water and evaporated. Puri®ca-
tion of the crude product by ¯ash chromatography
(CH2Cl2-acetone, 1%) a€orded pure 4 (399 mg, 92%):
1H NMR (400 MHz) 0.96 (s, 3H, H-19), 2.10 (s, 3H, H-21),
3.44 (m, 2H, THP), 3.85 (m, 1H, H-3), 4.66 (brs, 1H-THP),
4.90 (2dd, 2H, J=17.2, 8.8 and 2.1 Hz, -CH=CH2), 5.29
(t, 1H, H-6), 5.47 (m, 1H, -CH=CH2). Anal. C28H42O3
(C, H).
Experimental
Chemical synthesis
All reactions using nonaqueous reagents were run under
a dry argon atmosphere. Organic layers were dried on
magnesium sulfate (MgSO4). Column chromatography
was performed on alumina 90 Merck (0.063±0.2 mm) or
Kieselgel 60 Merck (70±230 mesh) and ¯ash chromato-
graphy on Kieselgel 60 Merck (230±400 mesh). Reaction
progress was monitored by analytical TLC on silica gel
60 F-254, alumina 150 F-254 or neutral alumina F-254
(type E) from Merck. Visualization of TLC was done by
phosphomolybdic acid or by UV light. 1H NMR spectra
were recorded either on a Jeol GS X 400 or on a Bruker
AC 200 or on a Bruker 300. Carbon magnetic resonance
3ꢀ-[(Tetrahydro-2H-pyran-2yl)oxy]-18-vinyl-21-hydroxy-
pregn-5-ene-20-one (5). A solution of methylketone 4
(105 mg, 0.246 mmol) in dry THF (2 mL) was added
dropwise at 78 ꢀC to a solution of LDA (0.48 mL,
0.96 mmol, 4 equiv, 2 M in THF) in dry THF (4 mL)
under argon. The mixture was stirred at 70 ꢀC for
10 min. Then, trimethylsilylchloride (280 mL) was added

1786
E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
at 78 ꢀC and the reaction was allowed to warm to
room temperature. Extraction with ether containing 1%
of Et3N (100 mL) and washing with a saturated solution
of NaHCO3 and water a€orded 156 mg of crude pro-
duct, containing 37% of the silyl enol ether according to
the NMR spectrum. 1H NMR (400 MHz) 0.06, 0.14,
and 0.20 (3s, Me3Si), 1.02 (s, 3H, H-19), 3.49 (m, 2H,
-THP), 3.91 (m, 1H, H-3), 4.58 (s, 2H, -O-C=CH2), 4.71
(brs, 1H, -THP), 4.88 (m, 2H, -CH=CH2), 5.35 (t, 1H,
H-6), 5.87 (m, 1H, -CH=CH2).
1H NMR (200 MHz) 1.12 and 1.20 (2s, 9H, tBuSi),
3.55 (m, 2H, -THP), 3.92 (m, 1H, H-3), 4.21 and 4.37
(2d, AB, J=17.3 Hz, H-21), 4.77 (m, 1H, -THP), 5.0
(m, 2H, -CH=CH2), 5.40 (m, 1H, H-6), 5.56 (m, 1H,
-CH=CH2), 7.35±7.84 (2m, 10H,-SiPh2-).
3ꢀ-Hydroxy-18-vinyl-21-tertiobutyldiphenylsilyloxypregn-
5-ene-20-one (8). Compound 7 (74.5 mg) was dissolved
in the mixture AcOH/H2O/THF (4/1/2, 1 mL) and the
solution was stirred for 1 h at room temperature. Then,
the solvent was evaporated in vacuum and the residue
was extracted with CH2Cl2, washed with a solution of
NaHCO3 and water. After evaporation, the residue was
puri®ed by ¯ash chromatography (cyclohexane/AcOEt,
5/1) to give pure 8 (10 mg, 71% from 5). 1H NMR
(400 MHz) 1.00 (s, 3H, H-19), 1.09 and 1.42 (2s, 9H,
tBuSi), 3.51 (m, 1H, H-3), 4.16 and 4.27 (2d, AB,
J=17.3 Hz, H-21), 4.91 (2bd, 2H, J=18.1 and
8.2 Hz, -CH=CH2), 5.34 (m, 1H, H-6), 5.40 (m, 1H,
-CH=CH2), 7.35±7.68 (2m, 10H, -SiPh2-).
This crude product was dissolved in absolute ethanol
(4 mL). MMPP (120 mg, 0.24 mmol) was added and the
mixture was stirred overnight at room temperature
under argon. Extraction with EtOAc and the usual
workup a€orded 154 mg of crude product which was
puri®ed by ¯ash chromatography (cyclohexane/AcOEt,
5/1) to give starting material 4 (32 mg, 30%) and a mix-
ture of a-hydroxyketone 5 and epoxides 6 (48 mg, 63%
corrected yield). Both oxidation products 5 and 6 have
the same Rf in cyclohexane/AcOEt (3/1 or 5/1).
According the NMR spectrum, a 7:3 ratio was calcu-
lated for the mixture 5:6, 81% corrected yield from silyl
enol ether, 30% corrected yield from 4.
18-Vinyl-21-tertiobutyldiphenylsilyloxypregen-5-ene-3,20-
dione (9). The alcohol 8 (10 mg, 0.017 mmol) was dis-
solved in dry toluene (3 mL) and N-methyl-4-piperidone
(0.1 mL) was added. The mixture was heated to re¯ux
under argon with a Dean-Stark apparatus. The ®rst 1.5 mL
of distillate were discarded. Aluminum isopropoxyde
(14 mg, 0.07 mmol) was added and the mixture re¯uxed
for 5 h. The toluene solution was diluted, neutralized
with an aqueous solution of HCl 5% and washed with
H2O. Evaporation of the solvent a€orded a residue
(3.5 mg) which was used in the next step. 1H NMR
(200 MHz) 1.05 (s, 3H, H-19), 1.08 (s, 9H, tBuSi), 4.14
and 4.26 (2d, AB, J=17.4 Hz, H-21), 4.92 (m, 2H, -CH=
CH2), 5.33 (m, 1H, -CH=CH2), 5.71 (bs, 1H, H-4).
5: 1H NMR (200 MHz) d 0.98 (s, 3H, H-19), 3.21 (t, 0.7
H, CH2-OH, J=4.7 Hz, exchanged with D2O), 3.47 (m,
2H, -THP), 3.87 (m, 1H, H-3), 4.13 (ABX, 2H, H-21,
J=4.7 and 18.8 Hz, collapsed with D2O), 4.68 (m, 1H,
-THP), 4.92 (2bd, 2H, J=16.5 and 7.7 Hz, -CH=CH2),
5.31 (m, 1H, H-6), 5.36 (m, 1H, -CH=CH2).
An analytical sample of pure 6 was obtained during the
chromatography step. According to the NMR spec-
trum, 6 was a 8/2 mixture of a and b epoxides.
1
6: H NMR (200 MHz) d 0.96 (s, 0.6 H, H-19, b-epox-
18-Vinyl-21-hydroxypregn-5-ene-3,20-dione 1 (18-VDOC).
The crude silyl ether 9 (3.5 mg) was dissolved in a mix-
ture MeOH±HCl 3% (1 mL) and stirred for 4 h at room
temperature. The solution was then extracted with
AcOEt, washed with a NaHCO3 solution and water.
Evaporation of the solvent gave 3 mg of residue which
was puri®ed by silica gel chromatography. Pure 1 (2 mg)
was obtained as a glassy solid. 1H NMR (200 MHz) 1.18
(s, 3H, H-19), 3.22 (dd, 1H, J=4.9 and 4.4 Hz, 21-OH,
collapsed with D2O), 4.12 and 4.28 (2dd, 2H, ABX,
J=18.7 Hz, 4.9 and 3.8 Hz), 4.97 (2dd, 2H, J=16.8, 9.9
and 2.2 Hz, -CH=CH2), 5.40 (m, 1H, -CH=CH2), 5.73
(bs, 1H, H-4); IR 3480, 1700, 1660, 1615 cm 1; MS (EI)
m/z 356, 325, 257, 206; HRMS calcd for C23H32O3:
356.235142, found 356.23159.
ide) 1.06 (s, 2.4H, H-19, a-epoxide), 2.13 (s, 3H, H-21),
2.89 (dd, 0.8 H, J=3.4 and 4.3 Hz, H-6, a-epoxide), 3.06
(dd, 0.2H, J=4.6 and 6.8 Hz, H-6, b-epoxide), 3.45 (m,
1H, -THP), 3.69 (m, 0.2 H, H-3, b-epoxide), 3.84 (m,
1H + 0.8H, -THP + H-3 of a-epoxide), 4.65 (m, 1H,
-THP), 4.93 (2dd, 2H, J=18.7, 8.1 and 2.0 Hz, -CH=
CH2), 5.47 (m, 1H, -CH=CH2).
3ꢀ-[(Tetrahydro-2H-pyran-2yl)oxy]-18-vinyl-21-tertio-
butyldiphenylsilyloxypregn-5-ene-20-one (7). To a stirred
solution of 73% pure 5 (38 mg), pyridine (15 mL) and
DMAP (2.3 mg, 0.019 mmol) in dry CH2Cl2 (1 mL) (t-
Bu)Ph2SiCl (55 mL, 0.21 mmol, 3.3 equiv) was added.
After the mixture was stirred for 5 days at room tem-
perature, the reaction was quenched with H2O and the
aqueous layer was extracted with CH2Cl2. The crude
product was partially puri®ed by ¯ash chromatography
(cyclohexane/AcOEt, 5/1) to give 7 (74.5 mg) which
was used without further puri®cation for the next step.
3ꢀ-Hydroxy-18-vinyl-20-(hydroxy-20-ethylene)-pregn-5-en-
20-one (10). Methylketone 3 (122 mg, 0.35 mmol) was
dissolved in 2.4 mL of dry THF and 1.2 mL of dry
benzene was added. Sodium hydride (263 mg of a 95%

E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
1787
dispersion in mineral oil, 10.4 mmol) and absolute etha-
nol (75 mL, 58.8 mg, 1.27 mmol) were transferred to the
steroid solution kept on an ice bath. The ¯ask was
placed under argon atmosphere and the reaction mix-
ture was stirred vigorously at room temperature for
25 min. Ethyl formate (0.38 g, 5.1 mmol) was added
dropwise with vigorous stirring. After 30 min, 1.8 N
CH3COOH (25 mL) was added cautiously (frothing)
until pH 6.7. The usual workup (AcOEt, H2O, MgSO4)
gave 163 mg of crude product as a mixture of 17a/17b
isomers (ratio 40/60 calculated from the NMR spec-
trum). Puri®cation by ¯ash chromatography (CH2Cl2/
acetone, 1%) a€orded pure (17b)10 (49 mg, 37%) as a
yellow oil and a mixture of the 17b and 17a isomers
(ratio 46/54, 80 mg, 62%). 10(17b): 1H NMR (400 MHz)
d 1.04 (s, 3H, H-19), 3.55 (m, 1H, H-3), 4.95 (2dd, 2H,
J=18.0, 11.2 and 1.3 Hz, -CH=CH2), 5.38 (m, 1H, H-
6), 5.53 (m, 1H, -CH=CH2), 5.60 (d, 1H, J=4.2 Hz, H-
210), 7.83 (d, 1H, J=4.2 Hz, H-2100); MS(EI) m/z: 370,
311, 239, 159; HRMS calcd for C24H34O3, 370.250792;
found, 370.25084.
12: 1H NMR (300 MHz) 1.03 (s, 2.1H, H-19, 17a iso-
mer), 1.04 (s, 0.9H, H-19, 17b isomer), 2.78 (t, 0.3H,
H-17, 17b isomer), 3.33 (dd, 0.7H, J=8.3 and
2.7 Hz, H-17, 17a isomer), 3.53 (m, 1H, H-3), 3.91
(dd, 1.4H, J=19.7 Hz, H-21, 17a isomer), 3.97 (dd,
0.6H, J=18.6 Hz, H-21, 17b isomer), 5.02 (m, 0.6H,
-CH=CH2, 17b isomer), 5.15 (m, 1.4H, -CH=CH2, 17a
isomer), 5.34 (bs, 1H, H-6), 5.25 (m, 0.3H, -CH=CH2,
17b isomer), 5.85 (m, 0.7H, -CH=CH2, 17a isomer).
Biological Methods
Materials
11-Deoxy[4-14C]corticosterone (1.85±2.22 GBq/mmol)
and [1, 2, 6, 7-3H] corticosterone (2.8±3.9 TBq/mmol)
were purchased from NEN (Dupont de Nemours,
France) and Amersham (France), respectively. 11-Deoxy-
corticosterone, corticosterone, and aldosterone were
generous gifts from Roussel Uclaf (France). 18-Hydroxy-
corticosterone was obtained from Sigma (France). 18-
Vinylprogesterone was synthesized as previously
described.3 NADPH was obtained from Boehringer
(Mannheim). Sodium cholate was purchased from
Sigma (France); soybean lecithin from Serva Feinbio-
chemica (Germany). All solvents and mineral salts were
of the highest purity available from Prolabo (France).
(17a): 1H NMR (400 MHz) d 1.03 (s, 3H, H-19), 2.85
(dd, 1H, J=9.2 and 3.1 Hz, H-17), 3.54 (m, 1H, H-
3), 5.14 (2bd, 2H, J=18.0 and 11.2 Hz, -CH=CH2),
5.38 (m, 1H, H-6), 5.46 (d, 1H, J=4.5 Hz, H-210),
5.87 (m, 1H, -CH=CH2), 7.77 (d, 1H, J=4.5 Hz, H-
2100).
Extensive decomposition of 10 was observed during
puri®cation. The diazoketone 11 was thus prepared
from the crude product.
Protein puri®cation
Bovine fresh adrenal glands were obtained from a local
slaughterhouse (Mantes-la-Jolie). Puri®cation of the
cytochrome P-45011b from bovine adrenocortical mito-
chondria (prepared from whole adrenal cortex) was
performed as previously described.7 The active heme
content of the ®nal enzymatic preparation (9 nmol/mg)
was determined by the di€erence spectrum at 450 nm of
the dithionite reduced CO-bound cytochrome versus the
3ꢀ-Hydroxy-18-vinyl-21-bromopregn-5-en-20-one 12 via
3ꢀ-hydroxy-18-vinyl-20-diazomethylpregn-5-en-20-one
(11). Methylketone 3 (197 mg, 0.57 mmol) was trans-
formed as described above into 10. The usual work up
(AcOEt, H2O, MgSO4) gave a crude product which
was immediately dissolved in 2 mL of dry THF. The
¯ask was placed on an ice bath and tosylazide (500 mL,
2.5 mmol, 0.49 g) was added under argon. Sodium
hydride (60 mg of a 80% dispersion in mineral oil,
2 mmol) was added at 0 ꢀC and vigorous stirring was
continued for 15 min. The reaction was quenched with
ice (frothing), washed with water, and dried (MgSO4).
The crude product was used without puri®cation for
the next step: IR Nmax 2120 (C=N=N), 1620,
1590 cm 1. The residue was dissolved in 2 mL of aceto-
nitrile. An aqueous solution (500 mL) of KBr
(300 mg) and APTS (40 mg) was added. The mixture
was stirred for 17 h at room temperature. After eva-
poration of the solvent and usual workup, a puri®ca-
tion by silica gel chromatography (cyclohexane/
AcOEt, 3/1) gave 12 (38 mg, 16% from 3) as a mixture
of 17a/17b isomers (70/30 calculated from the NMR
spectrum).
dithionite reduced cytochrome, assuming
1 25
a molar
absorption coecient of 91 mM 1 cm
.
Adrenodoxin and NADPH±adrenodoxin reductase
were also puri®ed from bovine adrenal mitochondria as
already described.7
Enzyme kinetics
11b- and 18-hydroxylation enzymatic activities of cyto-
chrome P-45011b were reconstituted as follows: all ster-
oids were dissolved in ethanol (the ®nal ethanol
concentration in the incubation mixture should not
exceed 2%). The incubation mixture contained P-45011b
and the labeled substrate at concentrations described
below, with or without inhibitor at di€erent concentra-
tions, adrenodoxin (10 mM), adrenodoxin reductase

1788
E. Davioud et al./Bioorg. Med. Chem. 6 (1998) 1781±1788
(0.4 mM), 50 mM potassium phosphate bu€er (pH=7.4)
and soybean lecithin sonicated in the same phosphate
bu€er for 3 min (2 mg/mL). The samples were pre-
incubated at 30 ꢀC for 30 s and the reaction, initiated by
the addition of NADPH (400 mM) was allowed to proceed
at 30 ꢀC and was stopped by the addition of methanol.
For 11b-hydroxylase activity, 11-deoxycorticosterone
(including 462Bq of 11-deoxy[4-14C]corticosterone) was
used as substrate (from 1.25 to 20 mM) with a P-45011b
concentration of 40nM. For 18-hydroxylase activity,
corticosterone (including 7.4 kBq of [1, 2, 6, 7-3H] corti-
costerone) was used as substrate (from 2.5 to 50 mM) with
a P-45011b concentration of 450 nM.
References
1. (a) Wada, A.; Onishi, T.; Okamoto, M.; Nonaka, Y.;
Yamano T. J. Biochem. 1985, 98, 245. (b) Yanagibashi, K.;
Haniou, M.; Shively, J. E.; Hai-Shen, W.; Hall, P. F. J. Biol.
Chem. 1986, 261, 3556.
2. Corvol, P.; Jard, S.; Menard, J. In Hormones: From Mole-
cules to Disease; Beaulieu, E. E.; Kelly, W. G., eds.; Paris:
Hermann, 1990; pp 608±629.
3. Viger, A.; Coustal, S.; Perard, S.; Marquet, A. Tetrahedron
1988, 44, 1127.
4. Viger, A.; Coustal, S.; Perard, S.; Pi€eteau, A.; Marquet, A.
J. Steroid Biochem. 1989, 33, 119.
5. Viger, A.; Coustal, S.; Perard, S.; Chappe, B.; Marquet, A.
J. Steroid Biochem. 1988, 30, 469.
After extraction of the solution with chloroform (2 vol),
the organic phase was evaporated under a stream of
nitrogen. The steroids were dissolved in 40 mL of ethanol
and chromatographed on silica-coated TLC plates (elu-
ent : acetone/toluene, 1/1). In this solvent system, the Rf
values for DOC, corticosterone, aldosterone and 18-
hydroxycorticosterone were, respectively, 0.72, 0.53,
0.32, and 0.23. The plates were analyzed using an auto-
matic TLC linear analyzer (LB2821 Berthold). The
amounts of steroid produced were calculated according
to the formula AÂB/C, where A is the initial amount of
6. Perard, S.; Viger, A.; Marquet, A. Steroids 1990, 55, 271.
7. Delorme, C.; Pi€eteau, A.; Viger, A.; Marquet, A. Eur. J.
Biochem. 1995, 232, 247.
8. Delorme, C.; Pi€eteau, A.; Sobrio, F.; Marquet, A. Eur. J.
Biochem. 1997, 248, 252.
9. Defaye, G.; Pi€eteau, A.; Delorme, C.; Marquet, A. J.
Steroid Biochem. Mol. Biol. 1996, 57, 141.
10. Souque, A.; Fagart, J.; Couette, B.; Davioud, E.; Sobrio,
F.; Marquet, A.; Rafestin-Oblin, M. E. Endocrinology 1995,
136, 5651.
3
substrate (pmol), B is the total H- or 14C-radioactivity
11. Johnston, J. O.; Wright, C. L.; Holbert, G. W. J. Steroid
Biochem. Molec. Biol. 1995, 52, 17.
of all zones. Therefore, B/C represents the percentage of
transformation of initial substrate to products during a
®xed incubation time.
12. Yamakita, N., Chiou, S.; Gomez-Sanchez, C. E. Endocri-
nology 1991, 129, 2361.
13. Fenselau, C. In Steroid Reactions; Djerassi, C., Ed.; San
Francisco: Holden±Day, 1963; pp 578±581.
In the preliminary assay, the inhibition by 18-VDOC and
18-VP was assayed only for 11b-hydroxylase activity
as described above with the following minor changes:
P-45011b (20 nM) was incubated for 1 min with [14C]-
DOC 1.25 mM, in the absence or presence of inhibitor,
either 18-VP (0.5, 1 or 10 mM) or 18-EP (0.25, 0.5 or 1 mM).
14. Metcalf, B. W.; Johnston, J. O. U.S. Patent 4,293,548,
1981; Chem. Abstr. 1981, 95, 618.
15. Moon, S.; Campbell, W. J. Chem. Comm. 1966, 470.
16. Heiba, E. I.; Dessau, R. M.; Koehl, W. J., Jr. J. Amer.
Chem. Soc. 1968, 90, 2706.
17. Marquet, A.; Jacques, J.; Tchoubar, B. Bull. Soc. Chim. Fr.
1965, 511.
For Ki determinations, all experiments were run in
duplicate for 30 s to 2 min for 11b-hydroxylation and for
1±3 min for the 18-hydroxylation, in the presence of
varying concentrations of substrate, DOC or cortico-
sterone, and at least three concentrations of inhibitor
18-VDOC (from 0.125 to 1 mM).
18. Kirk, D. N.; Rajagopalan, M. S. J. Chem. Soc. Perkin 1
1975, 1860.
19. Regitz, M. Synthesis 1972, 351.
20. Davioud, E.; Fagart, J.; Souque, A.; Rafestin-Oblin, M. E.;
Marquet, A. J. Med. Chem. 1996, 39, 2860.
21. Holland, H. L., Auret, B. J. Can. J. Chem. 1975, 53, 845.
Kinetic constants were determined from linear regres-
sion analysis, after double reciprocal transformation of
the experimental data. All the data are the average of at
least two independent determinations. Values are the
mean ‹SD from two to four independent experiments.
22. Brougham, P., Cooper, M. S., Cummerson, D. A., Heaney,
H., Thompson, N., Synthesis 1987, 1015.
23. (a) Poulos, T. L.; Finzel, B. C., Howard, J. J. Mol. Biol.
1987, 195, 687. (b) Graham-Lorence, S.; Peterson, J. A. Faseb
J. 1996, 10, 206. (c) Chang, Y. T.; Sti€elman, O. B., Loew, G.
H. Biochimie 1996, 78, 771.
Acknowledgements
24. Iwasaki, M.; Darden, T. A.; Pedersen, L. G.; Davis, D. G.;
Juvonen, R. O.; Sueyoshi, T.; Negishi, M. J. Biol. Chem. 1993,
268, 759.
The authors thank R. Azerad for providing the Asper-
gillus niger strain (ATCC 9142) and R. Lett for fruitful
discussions.
25. Omura, T.; Sato, R. J. Biol. Chem. 1964, 239, 2370.