Synthesis and steroid sulphatase inhibitory activity of C19- and C21-steroidal derivatives bearing a benzyl-inhibiting group

Article abstract of DOI:10.1016/S0223-5234(01)01262-4

Two series of compounds, benzyl alkylated at position 17α and 20 of androstane and pregnane, respectively, were synthesised and tested for steroid sulphatase inhibition. We compared the ability of the compounds to inhibit steroid sulphatase obtained from two different sources (homogenates of transfected HEK-293 cells and Jeg-3 cells) and with two types of substrate (DHEAS or E₁S). The inhibitory activity of 17α-benzyl-5α-androstane-3β,17β-diol (7), 17α-benzyl-5-androstene-3β,17β-diol (9), 17α-benzyl-4,17β-dihydroxy-4-androsten-3-one (15) and 20-benzyl-5-pregnene-3β,20α-diol (16) has proven to be superior to that of danazol, the first steroid sulphatase inhibitor to be reported, but still lower than that of the potent inhibitor estrone-3-O-sulphamate. The inhibitory activity of compound 7 was as potent as that of its previously reported estrane analogue, 17α-benzyl estradiol. Benzyl alkylated compounds with no OH group on the A-ring (with a 4-OCH₃, 4-Cl, or 4-H and their precursor epoxides), as well as a series of basic steroids without a benzyl group (ADT, epi-ADT, 3α-diol, 3β-diol, DHEA, Δ⁵-diol, DHT, T, Preg and Prog), did not show steroid sulphatase inhibition. We have thus demonstrated that the steroid sulphatase inhibitory effect of a benzyl group, previously observed for an estrane nucleus, can be extended to certain androstane and pregnane nuclei bearing a 3β-OH or a 4-OH group. Inhibitors 7, 9, 15 and 16 did not induce any proliferative effect on androgen-sensitive Shionogi cells. However, when tested on oestrogen-sensitive ZR-75-1 cells, a proliferative effect was observed for 7 and 9, but not for 15 and 16.

Full text of DOI:10.1016/S0223-5234(01)01262-4

Eur. J. Med. Chem. 36 (2001) 659–671
´
© 2001 Editions scientifiques et me´dicales Elsevier SAS. All rights reserved
PII: S0223-5234(01)01262-4/FLA
Original article
Synthesis and steroid sulphatase inhibitory activity of C19- and C21-steroidal
derivatives bearing a benzyl-inhibiting group
Liviu C. Ciobanu, Roch P. Boivin, Van Luu-The, Donald Poirier*
Medicinal Chemistry Division, Oncology and Molecular Endocrinology Research Center,
Centre Hospitalier Universitaire de Que´bec (CHUQ), Pavillon CHUL, Sainte-Foy, Que´bec, Canada G1V 4G2
Received 2 March 2001; revised 6 July 2001; accepted 11 July 2001
Abstract – Two series of compounds, benzyl alkylated at position 17a and 20 of androstane and pregnane, respectively, were
synthesised and tested for steroid sulphatase inhibition. We compared the ability of the compounds to inhibit steroid sulphatase
obtained from two different sources (homogenates of transfected HEK-293 cells and Jeg-3 cells) and with two types of substrate
(DHEAS or E1S). The inhibitory activity of 17a-benzyl-5a-androstane-3b,17b-diol (7), 17a-benzyl-5-androstene-3b,17b-diol (9),
17a-benzyl-4,17b-dihydroxy-4-androsten-3-one (15) and 20-benzyl-5-pregnene-3b,20a-diol (16) has proven to be superior to that of
danazol, the first steroid sulphatase inhibitor to be reported, but still lower than that of the potent inhibitor estrone-3-O-sulphamate.
The inhibitory activity of compound 7 was as potent as that of its previously reported estrane analogue, 17a-benzyl estradiol. Benzyl
alkylated compounds with no OH group on the A-ring (with a 4-OCH3, 4-Cl, or 4-H and their precursor epoxides), as well as a
series of basic steroids without a benzyl group (ADT, epi-ADT, 3a-diol, 3b-diol, DHEA, D5-diol, DHT, T, Preg and Prog), did not
show steroid sulphatase inhibition. We have thus demonstrated that the steroid sulphatase inhibitory effect of a benzyl group,
previously observed for an estrane nucleus, can be extended to certain androstane and pregnane nuclei bearing a 3b-OH or a 4-OH
group. Inhibitors 7, 9, 15 and 16 did not induce any proliferative effect on androgen-sensitive Shionogi cells. However, when tested
´
on oestrogen-sensitive ZR-75-1 cells, a proliferative effect was observed for 7 and 9, but not for 15 and 16. © 2001 Editions
scientifiques et me´dicales Elsevier SAS
steroid / chemical synthesis / sulphatase / enzyme / inhibitor / cancer
5
4
1. Introduction
D ,D -isomerases (3b-HSDs), 17b-hydroxysteroid de-
hydrogenases (17b-HSDs), 5a-reductases, and
aromatase. DHEA is thus transformed in peripheral
tissues into the active androgen dihydrotestosterone
(DHT), or into oestrogens, namely estradiol (E2),
A strategy expected to bring encouraging improve-
ments when applied to the treatment of hormone-sen-
sitive cancers consists in the inhibition of steroid
sulphatase, the enzyme controlling the conversion of
steroid sulphates into unconjugated steroids [1, 2].
Dehydroepiandrosterone sulphate (DHEAS), the
main secretory product of human adrenals and the
second most abundant steroid in circulation after
cholesterol, is converted by steroid sulphatase into the
steroid DHEA. The transformation of DHEA into
active sex steroids requires the presence of a series of
enzymes, namely 3b-hydroxysteroid dehydrogenase/
5
estrone (E1) and 5-androstene-3b,17b-diol (D -diol) [3,
4]. Estrone sulphate (E1S) is the most abundant circu-
lating oestrogen in women [5, 6], and its conversion
into E1 by steroid sulphatase directly provides a sub-
strate to type 1 17b-HSD, which then synthesises
estradiol, the most active oestrogen [7].
Recent studies indicated that the enzymes necessary
for steroid hormones biosynthesis are expressed in
tumoural tissues [8]. In postmenopausal breast cancer,
concentrations of E1S in peripheral blood are rela-
tively high and oestrogen levels in tumours are 10–50-
fold higher than in the plasma [9]. The half-life of E1S
* Correspondence and reprints
E-mail address: donald.poirier@crchul.ulaval.ca
(D. Poirier).

660
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
in the blood is 10–12 h, which is considerably longer
than that of the unconjugated oestrogens (30 min)
[10]. These facts taken together indirectly support the
hypothesis that, in hormone-sensitive tumours, the
active steroids are synthesised intracellularly, from
circulating steroid sulphates, and through the cata-
lytic activity of the expressed steroidogenic enzymes
[11, 12]. Furthermore, direct evidence for in situ
oestrogen synthesis involving sulphatase activity has
been reported in nitrosomethylurea-induced mam-
mary tumours in rat [13]. Thus, the utilisation of
steroid sulphatase inhibitors to reduce hormone syn-
thesis in the endocrine-sensitive cancers might signifi-
cantly improve therapy.
Some research groups have reported the develop-
ment of irreversible steroid sulphatase inhibitors
derived from E1, E1S and E1-phosphate, and also of
non-steroidal inhibitors, the structure of which in-
cludes the aromatic ring of E1 [1, 2, 14–17]. Recently,
our group has found that an alkyl, benzyl, or substi-
tuted benzyl group introduced at position 17a of
estradiol induces a reversible inhibitory effect on
steroid sulphatase [18]. We hypothesised that this new
family of inhibitors, represented by 17a-benzyl estra-
diol (1) may inhibit steroid sulphatase by a reversible
interaction (probably hydrophobic) with a region lo-
cated within the enzymatic site [19]. Unfortunately,
most of these estrane type inhibitors possess intrinsic
oestrogenic activity, which is not appropriate for a
potential therapeutic agent for oestrogen-sensitive
diseases.
In our efforts to develop steroid sulphatase in-
hibitors free of oestrogenic activity (figure 1), we first
replaced the Cl8-steroid estradiol nucleus of inhibitor
1 by the pure antioestrogen nucleus of ICI 164384
[20–22]. The long and bulky 7a-alkylamide side chain
of compound 2 drastically reduced the inhibitory ef-
fect of the 17a-benzyl group [23]. Some promising
results were however obtained by introducing at posi-
tion 17a of estradiol a hydrophobic octyl group
within a propanamide side chain (compound 3), but
mixed oestrogenic/antioestrogenic activities were ob-
served that will necessitate further SAR optimisation
studies [24]. In another approach, the benzyl group
was introduced at position 17a of danazol (4), the
first inhibitor of steroid sulphatase to be reported [25,
26], but no significant improvement of inhibitory ac-
tivity was obtained with compound 5 [23]. Finally, we
investigated C19 and C21 steroids as nonoestrogenic
nuclei, with a number of the C19-steroids maintaining
analogy with 4-hydroxy-androstenedione, a strong in-
hibitor of aromatase [27–29]. Our compounds were
designed to inhibit steroid sulphatase via a benzyl
group added in position 17a of a C19-steroid (com-
pounds 6–15) or in position 20 of a C21-steroid
(compounds 16–22). We here report the chemical
synthesis of these compounds, their steroid sulphatase
inhibitory activity as well as proliferative activity on
androgen-sensitive Shionogi cells and oestrogen-sensi-
tive ZR-75-1 cells.
2. Chemistry
A series of 17 benzyl alkylated C19- and C21-
steroids (compounds 6–22) were synthesised follow-
ing the two pathways reported in figures 2 and 3.
Compounds 6–8 were obtained through the sequence
of reactions presented in figure 2A. Androsterone
(ADT) and epiandrosterone (epi-ADT) were thus
alkylated at position 17 using an excess of benzyl
magnesium chloride to give diols 6 and 7 in 49 and
90% yield, respectively. For alkylation of epi-ADT,
the use of anhydrous cerium(III) chloride as catalyst
[30] was responsible for the higher yield of diol 7. As
Figure 1. Steroid derivatives developed (A) or proposed (B) to
inhibit the steroid sulphatase.

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
661
Figure 2. Reagents and conditions: (A) PhCH2MgCl, CeCl3, THF, −78 to −40 °C, 10–12 h; (B) PCC, NaOAc, CH2Cl2, 3 h; (C)
cyclohexanone, Al(i-PrO)3, toluene, reflux, 3 h.
previously reported [19], the addition of a Grignard
reagent to a C17-steroidal ketone afforded almost
exclusively the 17a-alkylated product. Pyridinium
chlorochromate oxidation of the secondary alcohol of
7 in dichloromethane gave the 3-ketosteroid 8 in 81%
yield. As reported in figures 2B and 2C, the benzyl
Grignard reagent was also added to dehydroepi-
androsterone (DHEA) and pregnenolone (Preg) lead-
ing to the formation of corresponding tertiary
alcohols 9 and 16 in high yields (91 and 91%, respec-
tively) when anhydrous cerium(III) chloride was used
as catalyst [30]. In the case of C21-steroid derivative
16, the stereochemistry at position 20 was determined
according to Cram’s rule and literature data [31]. A
minor 20b-OH isomer (6%) was also observed by
NMR of the crude product of alkylation, but only the
major 20a-OH isomer (91%) was used for the synthe-
sis of compound 17. Both compounds 9 and 16 were
thereafter submitted to Oppennauer oxidation condi-
tions (cyclohexanone and Al(i-PrO)3 in refluxing tolu-
ene) to give the a,b-unsaturated ketosteroids 10 (88%)
and 17 (73%).
The classical synthesis procedure of 4-hydroxy-an-
drostenedione generally involves formation and open-
ing of a 4,5-epoxy-steroid [32, 33]. As reported in
figure 3, unsaturated ketones 10 and 17 were trans-
formed into the corresponding epoxides 11, 12 and
18, 19 (as a 1:4 mixture of a:b epoxides), which were
separated by chromatography or submitted as a mix-
ture to conditions of epoxide opening. In our case,
however, the instability under acidic conditions of the
tertiary alcohol generated by alkylation at position 17
(androstane series) [34] or position 20 (pregnane se-
ries) was not compatible with the strong acid condi-
tions required for epoxide opening. Indeed, acid
opening of our 4,5-epoxy-steroids with 2% H2SO4 in
AcOH or formic acid (with or without reflux) failed,
mainly giving products of dehydration or rearrange-
ment. Similarly, no C4-chlorination of both unsatu-
rated ketones 10 and 17 occurred with a sulphuryl
chloride treatment, and a mixture of rearrangement
products was formed. Instead, chloro derivatives 13
(74%) and 20 (88%) were obtained by a mild acidic
treatment (HCl in acetone at room temperature) of

662
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
epoxides 11, 12 and 18, 19. No side products resulted
from these conditions. In basic conditions (NaOH in
refluxing methanol), these epoxides were opened
yielding the 4-methoxy derivatives 14 and 21 (57 and
50%, respectively). Since attempts to hydrolyse these
methoxy compounds by aqueous HCl in refluxing
dioxane and to generate the hydroxy analogues were
not successful, another strategy was tried. A direct
substitution of vinylic chloride of 13 and 20 using
potassium hydroxide (3 equiv. in refluxing t-butanol)
allowed the formation of the desired 4-hydroxy-
steroids 15 and 22 in reasonable yields (42 and 45%,
respectively).
I and II, respectively. As a source of steroid sulphatase
activity, we used homogenates of transfected human
embryonic kidney (HEK)-293 cells or homogenates of
human carcinoma Jeg-3 cells. With HEK-293 cells, only
DHEAS was used as substrate, whereas for Jeg-3 cells
both classical substrates (DHEAS and E1S) were used
for the enzymatic assays. With the Jeg-3 cells, a similar
level of inhibition was observed for both substrates.
When DHEAS was used as a substrate, the range of
inhibition differed from one cell line to the other, but
the conclusions regarding the activity of each compound
remain valuable. Indeed, for a given concentration of
inhibitor, the percentages of inhibition were slightly
higher with HEK-293 cells than with Jeg-3 cells. Some
reference steroids, such as ADT, epi-ADT, 3a-diol, 3b-
diol, DHEA, D5-diol, DHT, T, Preg and Prog, were also
tested in order to provide data for substrate compara-
tive studies. However, no inhibition of steroid sul-
phatase by these steroids was observed. Three known
inhibitors, estrone-3-O-sulphamate [35], 17a-benzyl
estradiol [18, 19] and danazol [25, 26], were also investi-
gated for their inhibitory activity in the same protocol.
In the series of synthesised compounds, ketone 8,
a,b-unsaturated ketones 10, 13, 14, 17, 20 and 21 and
ketoepoxides 11, 12, 18 and 19 do not inhibit steroid
sulphatase. The absence of a polar hydroxy (or sul-
phate) group at position 3 of the steroid A-ring might be
responsible for a low binding to the active site. Moder-
ate activities were however noted for the keto-enols 15
and 22, probably due to the presence of the hydroxyl
functionality in position 4. Interestingly, no inhibition
was obtained with 3a-OH steroid 6, but a higher in-
hibitory activity was noted for compounds 7, 9 and 16,
bearing a 3b-OH. The inhibitory activity of these com-
pounds is higher than that of danazol. In fact, com-
pound 7 is as potent inhibitor than 17a-benzyl estradiol,
in both HEK-293 and Jeg-3 cells, using either DHEAS
or E1S as substrate. The inhibitory effect of the benzyl
group of 7, 9 and 16 is made especially clear by compar-
ing it to that of their basic steroidal nuclei, respectively
3b-diol, D5-diol and Preg. Without the benzyl group no
inhibition of steroid sulphatase was observed.
3. Biological results
3.1. Inhibition of steroid sulphatase
The level of inhibition of steroid sulphatase by com-
pounds from the androstane and pregnane series, as well
as by other compounds of interest, is presented in tables
3.2. Proliferative effect on AR+ Shionogi and ER+
ZR-75-1 cells
Figure 3. Reagents and conditions: (a) NaOH, H2O2, CH3OH,
0 °C, 12 h; (b) NaOH, MeOH, reflux, 3 h; (c) HCl, H2O,
dioxane, reflux; (d) HCl, acetone, r.t., 1 h; (e) SO2Cl2, pyri-
dine, r.t.; (f) KOH (solid), t-butanol, reflux, 40 min; (g)
HCOOH, reflux; (h) 2% H2SO4 in AcOH, 0 °C, 12 h. See
figure 1 for the C20 stereochemistry of compounds 17–22.
In order to be a therapeutic agent for androgen- and
oestrogen-sensitive diseases, a steroid sulphatase in-
hibitor should be devoid of agonist effect on androgen
and oestrogen receptors. To assess this, we determined

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
663
Table I. Inhibition of steroid sulphatase (%) by compounds 6–15 (androstane series) and other compounds of interest. a,b
c
Compounds
HEK-293 cells (DHEAS“
DHEA) 0.1 mM/1 mM
Jeg-3 cells (DHEAS“
DHEA) 0.1 mM/1 mM
Jeg-3 cells (E1S“
E1) 0.1 mM/1 mM
6
NA/NA [10/32] d
NA/NA
21/71
0/0
4/24
0/0
0/0
0/0
0/0
0/0
0/25
0/0
0/0
0/0
0/12
0/0
0/0
0/0
NA/NA
23/57
0/9
13/53
0/1
0/2
0/6
0/2
0/2
3/27
2/1
2/1
7
46/86 [70/98] d
8
9
8/9
24/69
4/9
0/3
0/0
27/5
0/0
10
11
12
13
14
15
6/62
ADT
0/0 [6/12] d
epi-ADT
3a-Diol
3b-Diol
DHEA
D5-Diol
DHT
0/0 [7/11] d
0/0
0/4
6/24
0/0
0/4
1/4
4/0
3/2
0/0
3/4
T
3/0
6/0
1/2
E1-3-O-sulphamate
17a-Benzyl-E2
Danazol
89/91 [100/100] d
23/72 [79/98] d
18/37
100/100
0/54
0/29
95/97
18/56
3/22
a Tested at two concentrations of inhibitor (0.1 and 1 mM).
b Error95%.
c ADT, androsterone; epi-ADT, epiandrosterone; 3a-diol, 5a-androstane-3a,17b-diol; 3b-diol, 5a-androstane-3b,17b-diol; DHEA,
dehydroepiandrosterone; D5-diol, 5-androstene-3b,17b-diol; DHT, dihydrotestosterone; T, testosterone.
d Tested at two concentrations of 0.3 and 3 mM. N/A: data not available.
Table II. Inhibition of steroid sulphatase (%) by compounds 16–22 (pregnane series) and other compounds of interest. a,b
c
Compounds
HEK-293 cells (DHEAS“
DHEA) 0.1 mM/1 mM
Jeg-3 cells (DHEAS“
DHEA) 0.1 mM/1 mM
Jeg-3 cells (E1S“
E1) 0.1 mM/1 mM
16
67/100
0/85
0/14
28/0
12/0
8/7
1/34
0/6
0/0
0/21
0/0
0/0
0/0
0/0
0/0
0/0
24/0
0/0
2/64
0/18
0/2
0/1
1/1
0/1
7/23
3/8
17
18
19
20
21
22
Preg
Prog
0/3
E1-3-O-sulphamate
17a-Benzyl-E2
Danazol
89/91
23/72
18/37
100/100
0/54
0/29
95/97
18/56
3/22
a Tested at two concentrations of inhibitor (0.1 and 1 mM).
b Error95%.
c Preg, pregnenolone; Prog, progesterone.
the proliferative activities of inhibitors 7, 9, 15, 16 and
22 on two AR+ and ER+ cell lines. As can be seen in
figure 4, no proliferative effect on androgen-sensitive
(AR+) Shionogi cells was observed for all tested com-
pounds at either concentration used (0.1 and 1 mM).
Conversely, a 2.6-fold proliferative effect was obtained

664
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
when Shionogi cells were stimulated by 0.3 nM of the
known androgen dihydrotestosterone. In a similar ex-
periment using the oestrogen-sensitive (ER+) ZR-75-1
cells and reported in figure 5, we can see that a 0.1 nM
concentration of oestrogen estradiol provided a 3.4-fold
stimulation of basal cell proliferation, while no prolifer-
ative effect was observed for compounds 15, 16 and 22
at either concentration used (0.03 and 1 mM). On the
other hand, compounds 7 and 9 stimulated the ZR-75-1
cell proliferation, especially at 1 mM. These results were
somehow expectable considering their structural similar-
ity to the weak oestrogenic steroids D5-diol [36–38] and
3b-diol [39–41].
4. Conclusions
The inhibitory activity of compounds 7, 9, 15 and
16 on steroid sulphatase correspond to, or exceed, the
values found for reference inhibitors, danazol and
17a-benzyl estradiol. Their inhibitory activity remains
however lower than that of potent inhibitor estrone-3-
O-sulphamate. When compared to their correspond-
ing steroid nucleus, the results clearly show the
Figure 5. Effect of compounds 7, 9, 15, 16 and 22 on the
proliferation of oestrogen-sensitive (ER+) ZR-75-1 cells. Three
days after plating, cells were incubated for nine days with the
indicated concentration (0.03 or 1 mM) of test compounds 7, 9,
15, 16 and 22. E2 was tested at a concentration of 0.1 nM as
a positive control of cell proliferation. Media were changed
every second day. Results were expressed as the mean of DNA
content (mg) SEM, in triplicate analysis.
inhibitory effect of a benzyl group located at the
position 17a or 20 of some C19- and C21-steroids
bearing an OH group on the A-ring. They also sup-
port the previous observations [19, 42] of a hydropho-
bic region in the enzyme neighbouring the D-ring of
steroid substrates.
5. Experimental
5.1. Chemical synthesis of compounds 6–22
5.1.1. General methods
Chemical reagents and anhydrous solvents (except
THF) were purchased from Aldrich Chemical Company
(Milwaukee, WI, USA), while starting steroids and basic
steroids tested were from Steraloids (Wilton, NH,
USA). Solvents were obtained from BDH Chemicals
(Montre´al, Canada) or Fisher Chemicals (Montre´al,
Canada). Dry THF was distilled from sodium ben-
zophenone ketyl. Thin layer chromatography (TLC) was
performed on 0.20 mm silica gel 60 F254 plates (E.
Merck, Darmstadt, Germany), and 230–240 mesh
ASTM silica gel 60 (E. Merck) was used for flash
Figure 4. Effect of compounds 7, 9, 15, 16 and 22 on the
proliferation of androgen-sensitive (AR+) Shionogi cells. One
day after plating, cells were incubated for ten days with the
indicated concentration (0.1 or 1 mM) of tested compounds 7,
9, 15, 16 and 22. DHT was tested at a concentration of 0.3 nM
as a positive control of cell proliferation. Media were changed
every 3–4 days. Results are expressed as the mean of DNA
content (mg) SEM, in triplicate analysis.

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
665
column chromatography. Infrared (IR) spectra are ex-
5.1.2.1. 17h-Benzyl-5h-androstane-3h,17i-diol (6)
White solid (49% yield or 95% if correction for unre-
pressed in cm−1 and were obtained on a Perkin–Elmer
1600 (series FTIR) spectrophotometer (Norwalk, CT,
acted ADT). Rf=0.19 (hexanes–acetone, 75:25). IR
(KBr) w 3455 (OH, alcohols). H-NMR (acetone-d6) l
1
USA). H- and 13C-NMR spectra were recorded with a
1
Bruker AC/F 300 spectrometer (Billerica, MA, USA),
respectively, at 300 and 75 MHz. The chemical shifts (l)
are expressed in ppm and referenced to chloroform (7.26
ppm for 1H and 77.00 ppm for 13C). Only specific
0.85 (s, CH3-19), 0.92 (s, CH3-18), 2.66 and 2.86 (2d,
J=13.4 Hz, AB system, CH2Ph), 3.95 (s, CH-3b), 7.25
(m, 5H, CH2Ph). 13C-NMR (acetone-d6) l 33.01 (C1),
29.33 (C2), 65.91 (C3), 37.36 (C4), 39.75 (C5), 29.34
(C6), 32.66 (C7), 36.80 (C8), 55.37 (C9), 36.80 (C10),
21.18 (C11), 31.97 (C12), 47.57 (C13), 51.32 (C14), 24.07
(C15), 33.28 (C16), 83.49 (C17), 15.05 (C18), 11.57
(C19), 43.21 (CH2Ph), 140.40 (C1%), 128.16 (C2% and
C6%), 131.92 (C3% and C5%), 126.29 (C4%). LRMS: Calc.
for C26H38O2: 382.4 [M+] and 400.4 [M++NH4]. HPLC
purity=98.6% (C-18 NovaPak column; MeCN–
MeOH–H2O, 35:35:30).
1
signals were reported in IR and H-NMR whereas all
signals were listed in 13C-NMR. Fast atom bombard-
ment-high resolution mass spectra (FAB-HRMS) (ni-
trobenzylalcohol matrix) were provided by the Centre
Re´gional de Spectrome´trie de Masse (Universite´ de
Montre´al, Montre´al, Canada). The purity of synthesised
compounds was determined by HPLC (Waters Associ-
ates, Milford, MA, USA) by using an ultraviolet detec-
tor (205–280 nm). For the nomenclature of pregnane
derivatives 16–22 having a 20-hydroxyl group, we used
the arbitrary convention described by Fieser and Fieser
[43].
5.1.2.2. 17h-Benzyl-5h-androstane-3i,17i-diol (7)
White solid (90% yield). Rf=0.26 (hexanes–acetone,
75:25). IR (KBr) w 3348 (OH, alcohols). 1H-NMR
(CDCl3) l 0.68 (tapp, J=10.9 Hz, CH-14), 0.85 (s,
CH3-19), 0.92 (s, CH3-18), 2.60 and 2.86 (2d, J=13.2
Hz, AB system, CH2Ph), 3.59 (m, CH-3a), 7.28 (m, 5H,
CH2Ph). 13C-NMR (CDCl3) l 37.05 (C1), 31.40 (C2),
71.23 (C3), 38.14 (C4), 44.97 (C5), 28.62 (C6), 31.85
(C7), 36.42 (C8), 54.41 (C9), 35.59 (C10), 20.92 (C11),
31.48 (C12), 46.58 (C13), 50.54 (C14), 23.61 (C15), 33.67
(C16), 83.01 (C17), 14.55 (C18), 12.34 (C19), 42.36
(CH2Ph), 138.43 (C1%), 131.00 (C2% and C6%), 128.03 (C3%
and C5%), 126.20 (C4%). FAB-HRMS: Calc. for
C26H37O2: [M+−H], 381.27936. Found: 381.27870.
HPLC purity=96.1% (C-18 NovaPak column; MeCN–
MeOH–H2O, 35:30:35).
5.1.2. Alkylation of 17- and 20-ketosteroids with benzyl
magnesium chloride (general procedure for 6, 7, 9 and
16)
In a 250 mL flask equipped with a magnetic stirrer,
under argon atmosphere, was solubilised 1.478 g (6
mmol) of anhydrous cerium(III) chloride in 80 mL of
dry THF, at room temperature (r.t.). After stirring 1 h,
the mixture was cooled at −78 °C and 3 mL (6 mmol)
of 2.0 M solution of benzyl magnesium chloride in THF
was added dropwise. Stirring was continued for 1.5 h,
and a 0 °C solution of epiandrosterone, dehydroepi-
androsterone or pregnenolone (1 mmol) in 25 mL of
THF was slowly added to the above mixture. Five
minutes after addition, the −78 °C cooling bath was
changed for a −40 °C one (dry ice–acetonitrile) and the
stirring continued for 10–12 h. The reaction mixture
was, thereafter, quenched with an aqueous AcOH solu-
tion (2 mL of AcOH in 75 mL of water), and the
mixture let warm to r.t. Extraction with EtOAc (3×25
mL) followed by washing the combined organic phase
with brine (25 mL), drying over anhydrous MgSO4, and
evaporation of the solvent gave the crude product of
alkylation, which was further purified by flash column
chromatography (hexanes–acetone 75:25 for com-
pounds 7, 9 and 85:15 for compound 16). Compound 6
was prepared similarly as above from androsterone,
except that 10 equivalents of benzyl magnesium chloride
were used without cerium chloride. The purification of 6
by flash chromatography was done with hexanes–
EtOAc, 80:20 as eluent.
5.1.2.3. 17h-Benzyl-5-androstene-3i,17i-diol (9)
White solid (91% yield). Rf=0.26 (hexanes–acetone,
75:25). IR (KBr) w 3330 (OH, alcohols). 1H-NMR
(CDCl3) l 0.95 (s, CH3-18), 1.06 (s, CH3-19), 2.61 and
2.87 (2d, J=13.2 Hz, AB system, CH2Ph), 3.53 (m,
CH-3a), 5.37 (d, J=4.9 Hz, CH-6), 7.28 (m, 5H,
CH2Ph). 13C-NMR (CDCl3) l 37.31 (C1), 31.74 (C2),
71.67 (C3), 42.24 (C4), 140.88 (C5), 121.32 (C6), 31.28
(C7), 32.87 (C8), 50.91 (C9), 36.61 (C10), 20.77 (C11),
31.61 (C12), 46.32 (C13), 50.17 (C14), 23.71 (C15), 33.74
(C16), 83.04 (C17), 14.39 (C18), 19.39 (C19), 42.36
(CH2Ph), 138.34 (C1%), 131.00 (C2% and C6%) 128.06 (C3%
and C5%), 126.24 (C4%). FAB-HRMS: Calc. for
C26H35O2: [M+−H] 379.26370. Found: 379.26210.
HPLC purity=96.9% (C-18 NovaPak column; MeCN–
MeOH–H2O, 35:30:35).

666
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
5.1.2.4. 20-Benzyl-5-pregnene-3i,20h-diol (16)
5.1.4. Oppennauer oxidation of 3-hydroxy-5-ene-
steroids (general procedure for 10 and 17)
White solid (91% yield). Rf=0.12 (hexanes–acetone,
85:15); IR (KBr) w 3400 (OH, alcohols). 1H-NMR
(CDCl3) l 0.89 (s, CH3-18), 1.00 (s, CH3-19), 1.19 (s,
CH3-21), 2.61 and 2.82 (2d, J=13.1 Hz, AB system,
CH2Ph), 3.51 (m, CH-3a), 5.35 (d, J=5.0 Hz, CH-6);
7.24 (m, CH2Ph). 13C-NMR (CDCl3) l 37.21 (C1),
31.74 (C2), 71.70 (C3), 42.25 (C4), 140.77 (C5), 121.54
(C6), 31.59 (C7), 31.32 (C8), 49.98 (C9), 36.46 (C10),
20.90 (C11), 40.07 (C12), 42.82 (C13), 56.90 (C14), 23.87
(C15), 22.80 (C16), 58.60 (C17), 13.49 (C18), 19.34
(C19), 74.82 (C20), 26.34 (C21), 49.11 (CH2Ph), 137.59
(C1%), 130.68 (C2% and C6%), 128.01 (C3% and C5%), 126.28
(C4%). FAB-HRMS: Calc. for C28H39O2: [M+−H]
407.29501. Found: 407.29400. HPLC purity=96.3% (C-
18 NovaPak column; MeCN–MeOH–H2O, 35:45:20).
About 6% of a minor 20b-OH isomer [43] was also
observed in the NMR spectra of crude alkylation prod-
ucts, but this later was discarded during the purification
by flash column chromatography.
In a 250 mL three-neck flask equipped with a con-
denser, a Dean–Stark apparatus, and kept under argon
atmosphere, was introduced 1 mmol of 3-hydroxy-5-ene
steroids 9 and 16, 50 mL of toluene (HPLC grade), and
0.80 mL (7.7 mmol) of cyclohexanone. The stirring
solution was heated slowly up to distillation temperature
and 3×15 mL of toluene was successively removed from
the trap. Distillation was continued and a suspension of
246 mg (1.2 mmol) of aluminium isopropoxide in 20 mL
of toluene was added with a syringe. A new portion of
15 mL of toluene was extracted from the trap, and the
temperature adjusted to keep the reflux without any
distillation. After 3 h, the reaction mixture was cooled at
r.t. and quenched by adding a 10% HCl solution (30
mL). The crude product was extracted with CH2Cl2
(3×40 mL), dried over anhydrous MgSO4, filtered, evap-
orated, and then purified by flash column chromatogra-
phy (hexane–EtOAc, 75:25 and 80:20, respectively, for
10 and 17).
5.1.4.1. 17h-Benzyl-17i-hydroxy-4-androsten-
3-one (10)
White solid (88% yield). Rf=0.32 (hexanes–acetone,
75:25). IR (film) w 3518 (OH, alcohol), 1670 (CꢀO,
conjugated ketone), 1618 (conjugated CꢀC). H-NMR
(CDCl3) l 0.97 (s, CH3-18), 1.21 (s, CH3-19), 2.58 and
2.86 (2d, J=13.2 Hz, AB system, CH2-Ph), 5.74 (s,
CH-4), 7.28 (m, 5H, CH2Ph). 13C-NMR (CDCl3) l
35.69 (C1), 33.90 (C2), 199.47 (C3), 123.81 (C4), 171.14
(C5), 32.79 (C6), 31.10 (C7), 36.47 (C8), 53.77 (C9),
38.62 (C10), 20.69 (C11), 31.70 (C12), 46.35 (C13), 49.95
(C14), 23.52 (C15), 35.56 (C16), 82.75 (C17), 14.43
(C18), 17.36 (C19), 42.30 (CH2Ph), 138.10 (C1%), 130.94
(C2% and C6%), 128.07 (C3% and C5%), 126.27 (C4%). FAB-
HRMS: Calc. for C26H35O2: [M++H] 379.26370. Found:
379.26240. HPLC purity=99.4% (C-18 NovaPak
column; MeCN–MeOH–H2O, 35:30:35).
5.1.3. PCC oxidation of alcohol 7 to ketone 8
In a 200 mL flask and under argon atmosphere, the
alcohol 7 (200 mg, 0.52 mmol) dissolved in dry CH2Cl2
(20 mL) was added dropwise to a mixture of pyridinium
chlorochromate (172 mg, 0.80 mmol), sodium acetate
(131 mg, 1.60 mmol), and molecular sieves (4A) in
CH2Cl2 (35 mL). After 3 h at r.t., the mixture was
filtered on silica gel flash column (hexanes–EtOAc,
85:15) to give 160 mg (81% yield) of ketone 8.
1
5.1.3.1. 17h-Benzyl-17i-hydroxy-5h-androstan-
3-one (8)
White solid; Rf=0.36 (hexanes–acetone, 75:25). IR
1
(KBr): w 3543 (OH, alcohol), 1707 (CꢀO, ketone). H-
NMR (CDCl3) l 0.79 (m, CH-14), 0.95 (s, CH3-18), 1.05
(s, CH3-19), 2.61 and 2.87 (2d, J=13.3 Hz, AB system,
CH2Ph), 7.28 (m, 5H, CH2Ph). 13C-NMR (CDCl3) l
38.59 (C1), 38.13 (C2), 211.88 (C3), 44.67 (C4), 46.78
(C5), 28.86 (C6), 31.35 (C7), 36.32 (C8), 53.85 (C9),
35.78 (C10), 21.13 (C11), 31.49 (C12), 46.57 (C13), 50.36
(C14), 23.62 (C15), 33.71 (C16), 82.93 (C17), 14.55
(C18), 11.49 (C19), 42.37 (CH2Ph), 138.29 (C1%), 130.98
(C2% and C6%), 128.09 (C3% and C5%), 126.27 (C4%). FAB-
HRMS calculated for C26H37O2: [M++H] 381.27936.
Found: 381.28060. HPLC purity=95.2% (C-18 Nova-
Pak column; MeCN–MeOH–H2O, 40:30:30).
5.1.4.2. 20-Benzyl-20h-hydroxy-4-pregnen-3-one (17)
White solid (73% yield). Rf=0.16 (hexanes–acetone,
85:15). IR (KBr) w 3422 (OH, alcohol), 1668 (CꢀO,
1
conjugated ketone), 1612 (conjugated CꢀC). H-NMR
(CDCl3) l 0.92 (s, CH3-18), 1.17 and 1.18 (2s, CH3-19
and CH3-21), 2.60 and 2.81 (2d, J=13.1 Hz, AB sys-
tem, CH2Ph), 5.72 (s, CH-4), 7.22 (m, 5H, CH2Ph).
13C-NMR (CDCl3) l 35.61 (C1), 33.90 (C2), 199.48
(C3), 123.72 (C4), 171.39 (C5), 32.84 (C6), 31.86 (C7),
34.94 (C8), 53.63 (C9), 38.51 (C10), 20.81 (C11), 39.91

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
667
(C12), 42.85 (C13), 56.01 (C14), 23.74 (C15), 22.70
(C16), 58.44 (C17), 13.50 (C18), 17.30 (C19), 74.60
(C20), 26.31 (C21), 49.10 (CH2Ph), 137.42 (C1%), 130.62
(C2% and C6%), 128.01 (C3% and C5%), 126.31 (C4%). FAB-
HRMS: Calc: for C28H39O2 [M++H] 407.29501. Found:
407.29601. HPLC purity=98.7% (C-18 NovaPak
column; MeCN–MeOH–H2O, 35:45:20).
18), 1.19 (s, CH3-19), 2.58 and 2.85 (2d, J=13.2 Hz, AB
system, CH2Ph), 2.99 (s, CH-4a), 7.27 (m, 5H, CH2Ph).
13C-NMR (CDCl3) l 32.52 (C1), 30.09 (C2), 206.72
(C3), 62.68 (C4), 70.29 (C5), 26.17 (C6), 29.79 (C7),
35.93 (C8), 46.59 (C9), 37.30 (C10), 21.26 (C11), 30.98
(C12), 46.59 (C13), 49.97 (C14), 23.56 (C15), 33.57
(C16), 82.72 (C17), 14.48 (C18), 18.95 (C19), 42.34
(CH2Ph), 138.07 (C1%), 130.94 (C2% and C6%), 128.13 (C3%
and C5%), 126.33 (C4%). FAB-HRMS: Calc. for C26H35O3
[M++H] 395.25861. Found: 395.25800. HPLC purity=
96.0% (C-4 YMCPak column; MeCN–MeOH–H2O,
35:35:30).
5.1.5. Epoxidation of enone-steroids 10 and 17 (general
procedure for 11, 12 and 18, 19)
In a 100 mL flask was solubilised 1 mmol of a,b-un-
saturated ketone 10 or 17 in 20 mL of CH3OH–CH2Cl2
(9:1). After 30 min at r.t., 0.7 mL of 30% H2O2 (w/v)
was added dropwise and the mixture stirred for 45 min
or 1 h. The mixture was then cooled at 0 °C, and 0.4
mL of 4 M NaOH was slowly added. Stirring was
continued for 1 h at 0 °C and for an additional 12 h at
4 °C. The reaction was quenched by 30 mL of water,
and the aqueous phase was successively extracted with
three portions of CH2Cl2 (35 mL). It proved useful to
add 2–3 mL of 10% HCl to dissolve the emulsion
resulting from the extraction process. The combined
organic phase was dried over anhydrous MgSO4, and
the solvent evaporated under vacuum. The crude
product was then purified by flash column chromatogra-
phy (hexanes–EtOAc, 85:15 and 87.5:12.5, respectively).
5.1.5.3. 20-Benzyl-4h,5h-epoxy-20h-hydroxy-pregnan-3-
one (18)
White powder (18% yield). Rf=0.39 (hexanes–ace-
tone, 75:25). IR (KBr) w 3588 (OH, alcohol), 1711 (CꢀO,
1
ketone). H-NMR (CDCl3) l 0.91 (s, CH3-18), 1.05 (s,
CH3-19), 1.19 (s, CH3-21), 2.61 and 2.82 (2d, J=13.0
Hz, AB system, CH2Ph), 3.03 (s, CH-4b), 7.24 (m, 5H,
CH2Ph). 13C-NMR (CDCl3) l 33.09 (C1), 29.70 (C2),
207.00 (C3), 62.84 (C4), 70.14 (C5), 28.81 (C6), 29.70
(C7), 34.80 (C8), 50.55 (C9), 36.65 (C10), 21.23 (C11),
40.03 (C12), 43.00 (C13), 55.82 (C14), 23.80 (C15), 22.72
(C16), 58.61 (C17), 13.54 (C18), 16.47 (C19), 74.67
(C20), 26.34 (C21), 49.11 (CH2Ph), 137.45 (C1%), 130.67
(C2% and C6%), 128.06 (C3% and C5%), 126.36 (C4%). FAB-
HRMS: Calc. for C28H39O3 [M++H] 423.28992. Found:
423.28850. HPLC purity=97.4% (C-4 YMCPak
column; MeOH–H2O, 80:20 containing 10 mM of
NH4OAc).
5.1.5.1. 17h-Benzyl-4h,5h-epoxy-17i-hydroxy-
androstan-3-one (11)
White powder (24% yield). Rf=0.29 (hexanes–ace-
tone, 80:20). IR (KBr) w 3548 (OH, alcohol), 1712 (CꢀO,
1
ketone). H-NMR (CDCl3) l 0.96 (s, CH3-18), 1.09 (s,
CH3-19), 2.62 and 2.87 (2d, J=13.2 Hz, AB system,
CH2Ph), 3.05 (s, CH-4b), 7.29 (m, 5H, CH2Ph). 13C-
NMR (CDCl3) l 33.08 (C1), 29.16 (C2), 206.85 (C3),
62.89 (C4), 70.08 (C5), 28.72 (C6), 29.65 (C7), 36.34
(C8), 50.76 (C9), 36.80 (C10), 21.11 (C11), 31.20 (C12),
46.48 (C13), 49.79 (C14), 23.59 (C15), 33.63 (C16), 82.86
(C17), 14.46 (C18), 16.50 (C19), 42.35 (CH2Ph), 138.16
(C1%), 130.97 (C2% and C6%), 128.13 (C3% and C5%), 126.33
(C4%). FAB-HRMS: Calc. for C26H35O3: [M++H]
395.25861. Found: 395.25760. HPLC purity=97.6% (C-
18 Symmetry column; MeCN–MeOH–H2O, 30:35:35).
5.1.5.4. 20-Benzyl-4i,5i-epoxy-20h-hydroxy-pregnan-
3-one (19)
White powder (53% yield). Rf=0.25 (hexanes–ace-
tone, 85:15); IR (KBr) w 3680 (OH, alcohol), 1705 (CꢀO,
1
ketone). H-NMR (CDCl3) l 0.90 (s, CH3-18), 1.14 (s,
CH3-19), 1.18 (s, CH3-21), 2.60 and 2.80 (2d, J=13.0
Hz, AB system, CH2Ph), 2.97 (s, CH-4a), 7.22 (m, 5H,
CH2Ph). 13C-NMR (CDCl3) l 32.52 (C1), 30.27 (C2),
206.82 (C3), 62.65 (C4), 70.31 (C5), 26.11 (C6), 29.85
(C7), 34.39 (C8), 46.35 (C9), 37.13 (C10), 21.36 (C11),
39.78 (C12), 43.11 (C13), 56.03 (C14), 23.79 (C15), 22.67
(C16), 58.43 (C17), 13.56 (C18), 18.91 (C19), 74.62
(C20), 26.34 (C21), 49.12 (CH2Ph), 137.39 (C1%), 130.64
(C2% and C6%), 128.07 (C3% and C5%), 126.38 (C4%). FAB-
HRMS: Calc. for C28H39O3: [M++H] 423.28992. Found:
423.28950. HPLC purity=96.6% (C-4 YMCPak
column; MeCN–MeOH–H2O, 35:35:30).
5.1.5.2. 17h-Benzyl-4i,5i-epoxy-17i-hydroxy-
androstan-3-one (12)
White powder (71% yield). Rf=0.22 (hexanes–ace-
tone, 85:15). IR (KBr) w 3507 and 3426 (OH, alcohol),
1
1712 (CꢀO, ketone). H-NMR (CDCl3) l 0.95 (s, CH3-

668
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
5.1.6. Synthesis of 4-chloro-steroids from 4,5-epoxides
(general procedure for 13 and 20)
(C2% and C6%), 128.07 (C3% and C5%), 126.37 (C4%). FAB-
38
HRMS: Calc. for C28H ClO2: [M++H] 441.25604.
35
In a 100 mL flask, a 1:4 mixture of 4a,5a- and
4b,5b-epoxides (1 mmol) was dissolved in 20 mL of
acetone, concentrated HCl (0.3 mL) was slowly added
with a syringe, and the reaction mixture was stirred for
1–1.5 h at r.t. Saturated NaCl solution (30 mL) was
then added and the reaction mixture extracted with
CH2Cl2. The organic phase was successively washed
with water, 10% Na2CO3 solution, and again with water.
The dried (MgSO4) organic phase was thereafter filtered
and solvent evaporated under vacuum. The crude chloro
derivative was further purified by flash column chro-
matography (hexanes–EtOAc, 80:20).
Found: 441.25470. HPLC purity=96.6% (C-4 YMC-
Pak column; MeOH–H2O, 75:25).
5.1.7. Synthesis of 4-methoxy-steroids from
4,5-epoxides (general procedure for 14 and 21)
An aqueous 4 M NaOH solution (3 mL) was added to
3-oxo-4,5-epoxides (1 mmol) dissolved in 12 mL of
methanol (HPLC grade) and the reaction mixture was
heated to reflux. After completion of the reaction (3 h),
the mixture was cooled to r.t. and the solvent evapo-
rated under reduced pressure. The residue was treated
with water before extraction with CH2Cl2 (3×25 mL).
The organic phase was successively washed with water
(2×25 mL) and brine (25 mL), followed by drying over
anhydrous MgSO4. The filtrate was concentrated in
vacuo and the crude product was further purified by
flash column chromatography (hexanes–EtOAC, 75:25
and 85:15, respectively).
5.1.6.1. 17h-Benzyl-4-chloro-17i-hydroxy-4-androsten-
3-one (13)
White solid (74% yield). Rf=0.24 (hexanes–acetone,
80:20). IR (KBr) w 3535 (OH, alcohol), 1694 (CꢀO,
1
conjugated ketone). H-NMR (CDCl3) l 0.98 (s, CH3-
18), 1.27 (s, CH3-19), 2.27 (td, J1=5.2 Hz and J2=14.4
Hz, 1H of CH2-6), 2.59 (m, 1H of CH2-2), 2.58 and 2.86
(2d, J=13.1 Hz, AB system, CH2Ph), 3.27 (td, J1=3.2
Hz and J2=14.8 Hz, 1H of CH2-6), 7.28 (m, 5H,
CH2Ph). 13C-NMR (CDCl3) l 34.44 (C1), 33.97 (C2),
190.62 (C3), 127.27 (C4), 164.59 (C5), 28.94 (C6), 30.88
(C7), 36.01 (C8), 53.97 (C9), 41.42 (C10), 20.82 (C11),
31.12 (C12), 46.34 (C13), 49.91 (C14), 23.48 (C15), 33.67
(C16), 82.70 (C17), 14.44 (C18), 17.79 (C19), 42.31
(CH2Ph), 137.99 (C1%), 130.93 (C2% and C6%), 128.14 (C3%
and C5%), 126.35 (C4%). FAB-HRMS: Calc. for
5.1.7.1. 17h-Benzyl-17i-hydroxy-4-methoxy-4-
androsten-3-one (14)
Slight yellow solid (57% yield). Rf=0.25 (hexanes–
acetone, 80:20). IR (film) w 3463 (OH, alcohol), 1676
1
(CꢀO, conjugated ketone). H-NMR (CDCl3) l 0.98 (s,
CH3-18), 1.23 (s, CH3-19), 2.45 (m, 1H of CH2-6 and 1H
of CH2-2), 2.59 and 2.86 (2d, J=13.1 Hz, AB system,
CH2Ph), 3.09 (d, J=14.5 Hz, 1H of CH2-6), 3.60 (s,
CH3O), 7.27 (m, 5H, CH2Ph). 13C-NMR (CDCl3) l
34.88 (C1), 34.26 (C2), 194.06 (C3), 146.64 (C4), 154.85
(C5), 23.40 (C6), 31.32 (C7), 36.23 (C8), 54.26 (C9),
38.78 (C10), 20.74 (C11), 31.23 (C12), 46.39 (C13), 50.09
(C14), 23.59 (C15), 33.74 (C16), 82.81 (C17), 14.48
(C18), 17.54 (C19), 42.36 (CH2Ph), 138.13 (C1%), 130.98
(C2% and C6%), 128.16 (C3% and C5%), 126.36 (C4%), 60.26
(CH3O). FAB-HRMS: Calc. for C27H37O3: [M++H]
409.27426. Found: 409.27540. HPLC purity=99.4% (C-
18 Symmetry column; MeCN–MeOH–H2O, 30:35:35).
C26H ClO2: [M++H] 413.22473. Found: 413.22580.
35
34
HPLC purity=98.4% (C-4 YMCPak column; MeOH–
H2O, 75:25).
5.1.6.2. 20-Benzyl-4-chloro-20h-hydroxy-4-pregnen-3-
one (20)
White solid (88% yield). Rf=0.28 (hexanes–acetone,
80:20); IR (KBr) w 3480 (OH, alcohol), 1687 (CꢀO,
1
conjugated ketone). H-NMR (CDCl3) l 0.93 (s, CH3-
18), 1.19 (s, CH3-19), 1.23 (s, CH3-21), 2.56 (m, 1H of
CH2-2 and 1H of CH2-6), 2.60 and 2.82 (2d, J=13.1
Hz, AB system, CH2Ph), 3.25 (td, J1=3.2 Hz and
J2=7.3 Hz, 1H of CH2-6), 7.25 (m, 5H, CH2Ph).
13C-NMR (CDCl3) l 34.37 (C1), 33.98 (C2), 190.70
(C3), 127.17 (C4), 164.88 (C5), 29.00 (C6), 31.04 (C7),
34.49 (C8), 53.84 (C9), 41.32 (C10), 20.93 (C11), 39.88
(C12), 42.85 (C13), 55.96 (C14), 23.71 (C15), 22.75
(C16), 58.46 (C17), 13.53 (C18), 17.72 (C19), 74.58
(C20), 26.34 (C21), 49.09 (CH2Ph), 137.38 (C1%), 130.65
5.1.7.2. 20-Benzyl-20h-hydroxy-4-methoxy-4-pregnen-
3-one (21)
Slight yellow solid (50% yield). Rf=0.27 (hexanes–
acetone, 80:20); IR (film) w 3510 (OH, alcohol), 1677
1
(CꢀO, conjugated ketone). H-NMR (CDCl3) l 0.92 (s,
CH3-18), 1.18 (s, CH3-19), 1.19 (s, CH3-21), 2.43 (m, 1H
of CH2-6 and 1H of CH2-2), 2.61 and 2.82 (2d, J=13.2
Hz, AB system, CH2Ph), 3.05 (dm, J=14.9 Hz, 1H of
CH2-6), 3.58 (s, CH3O), 7.24 (m, 5H, CH2Ph). 13C-
NMR (CDCl3) l 34.74 (C1), 34.21 (C2), 194.10 (C3),

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
669
146.54 (C4), 155.15 (C5), 23.42 (C6), 31.41 (C7), 34.65
(C8), 54.06 (C9), 38.61 (C10), 20.81 (C11), 39.99 (C12),
42.86 (C13), 56.10 (C14), 23.76 (C15), 22.76 (C16), 58.50
(C17), 13.54 (C18), 17.43 (C19), 74.65 (C20), 26.34
(C21), 49.12 (CH2Ph), 137.45 (C1%), 130.66 (C2% and
C6%), 128.05 (C3% and C5%), 126.35 (C4%), 60.23 (CH3O).
FAB-HRMS: Calc. for C29H41O3: [M++H] 437.30557.
Found: 437.30710. HPLC purity=99.1% (C-18 Nova-
Pak column; MeCN–MeOH–H2O, 40:30:30).
(s, CH3-18), 1.17 (s, CH3-19), 1.19 (s, CH3-21), 2.50 (m,
1H), 2.61 and 2.82 (2d, J=13.1 Hz, AB system,
CH2Ph), 3.01 (dm, J=17.0 Hz, 1H of CH2-6), 6.07 (s,
OH of enol), 7.23 (m, SH, CH2Ph). 13C-NMR (CDCl3)
l 34.63 (C1), 31.77 (C2), 193.58 (C3), 141.14 (C4),
140.18 (C5), 23.03 (C6), 30.91 (C7), 34.63 (C8), 54.12
(C9), 37.74 (C10), 20.81 (C11), 40.04 (C12), 42.89 (C13),
56.21 (C14), 23.84 (C15), 22.79 (C16), 58.59 (C17), 13.57
(C18), 17.11 (C19), 74.73 (C20), 26.39 (C21), 49.15
(CH2Ph), 137.51 (C1%), 130.69 (C2% and C6%), 128.08 (C3%
and C5%), 126.37(C4%). FAB-HRMS: Calc. for C28H39O3
[M++H] 423.28992. Found: 423.29090. HPLC purity=
96.5% (C-4 YMCPak column; MeCN–MeOH–H2O,
75:25).
5.1.8. Synthesis of 4-hydroxy-steroids from 5-chloro-
steroids 13 and 20 (general procedure for 15 and 22)
In a 100 mL flask connected to a reflux condenser and
under argon atmosphere chloro-steroids 13 and 20 (1
mmol) was dissolved in 25 mL of t-butanol. While
slowly stirring started, 168 mg (3 mmol) of solid KOH,
as a fine powder, was added in one portion, and the
reaction mixture heated slowly to reflux for 35–45 min.
Thereafter, the reaction mixture was cooled to r.t. and
neutralised (pH 7) with a 5% AcOH solution. Water (30
mL) was then added and the product extracted with
EtOAc (3×30 mL). The organic phase was washed with
water, dried over anhydrous MgSO4, filtered, and con-
centrated under reduced pressure. The resulted crude
product was purified by flash column chromatography
(hexanes–acetone, 80:20 and 85:15, respectively).
5.2. Inhibition of steroid sulphatase by compounds 6–22
and basic steroids
5.2.1. HEK-293 cells as source of steroid sulphatase
activity (enzymatic assay)
Human embryonic kidney (HEK)-293 cells (American
Type Culture Collection, Rockville, MD), transiently
transfected with a sulphatase expression vector (pCMV-
sulpha) were used as source of steroid sulphatase activ-
ity. The pCMV-sulpha was constructed by inserting a
cDNA fragment containing the coding sequence of a
human sulphatase gene downstream from the CMV
promoter of the pCMV vector (kindly provided by Dr
M.B. Mathews, Cold Spring Harbor Laboratories, Cold
Spring Harbor, NY, USA). The sulphatase cDNA frag-
ment was obtained by screening a human placenta
cDNA library (Clontech Laboratories Inc., Palo Alto,
CA, USA) using an incomplete cDNA fragment (kindly
provided by Dr L.J. Shapiro, Howard Hughes Medical
Institute, Los Angeles, CA, USA) as probe. The trans-
fection of the expression vector was performed by the
calcium phosphate procedure using 10 mg of recombi-
nant plasmid per 106 cells [44]. The cells were initially
plated at 104 cells cm−2 in Falcon culture flasks and
grown in Dulbecco’s modified Eagle’s medium contain-
ing 10% (v/v) foetal bovine serum supplemented with 2
5.1.8.1. 17h-Benzyl-4,17i-dihydroxy-4-androsten-
3-one (15)
White solid (42% yield). Rf=0.31 (hexanes–acetone,
75:25). IR (film) w 3440 (OH, alcohols), 1704 and 1667
(CꢀO, enol), 1635 (CꢀC, enol). 1H-NMR (CDCl3) l 0.98
(s, CH3-18), 1.22 (s, CH3-19), 2.52 (m, 1H), 2.60 and
2.86 (2d, J=13.2 Hz, AB system, CH2Ph), 3.04 (dm,
J=14.5 Hz, 1H of CH2-6), 6.09 (s, OH of enol), 7.27
(m, 5H, CH2Ph). 13C-NMR (CDCl3) l 34.73 (C1), 31.78
(C2), 193.58 (C3), 141.20 (C4), 139.91 (C5), 23.00 (C6),
30.82 (C7), 36.16 (C8), 54.28 (C9), 37.89 (C10), 20.70
(C11), 31.24 (C12), 46.36 (C13), 50.14 (C14), 23.64
(C15), 33.75 (C16), 82.88 (C17), 14.49 (C18), 17.20
(C19), 42.35 (CH2Ph), 138.15 (C1%), 131.01 (C2% and
C6%), 128.17 (C3% and C5%), 126.37 (C4%). FAB-HRMS:
Calc. for C26H35O3: [M++H] 395.25861. Found:
395.25800. HPLC purity=97.4% (C-4 YMCPak
column; MeOH–H2O, 70:30).
mM L-glutamine, 1 mM sodium pyruvate, 100 IU peni-
cillin mL−1, and 100 mg streptomycin sulphate mL−1
.
For the assay, the HEK-293 cell homogenate was
prepared by repeated freezing (−80 °C) and thawing (5
times), and homogenisation using
a Dounce ho-
5.1.8.2. 20-Benzyl-4,20h-dihydroxy-4-pregnen-3-one (22)
White solid (45% yield). Rf=0.44 (hexanes–EtOAc,
75:25); IR (KBr) w 3408 (OH, alcohols), 1705 and 1668
(CꢀO, enol); 1640 (CꢀC, enol). 1H-NMR (CDCl3) l 0.92
mogeniser. The reaction was carried out at 37 °C in
1.25 mL of 100 mM Tris–acetate buffer (pH 7.4) con-
taining 5 mM of EDTA, 10% glycerol, 100 mM of
[3H]DHEAS as substrate, and an ethanolic solution of

670
L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
duced (100% for the control without inhibitor) and the
percentage (%) of inhibition then calculated.
compound to test (at appropriate concentrations). After
2 h of incubation, the reaction was stopped by addition
of 1.25 mL of xylene. The tubes were then shaken and
centrifuged at 2000g for 10 min to separate the organic
and aqueous phases. Radioactivity in 750 mL of each
phase (organic: free steroids; aqueous: sulphated
steroids) was determined by liquid scintillating counting
with a Beckman LS3801 (Irvine, CA). Data were ex-
pressed as percentage (%) of DHEA produced (100% for
the control without inhibitor) and the percentage (%) of
inhibition then calculated.
5.3. Proliferative assays on androgen-sensitive Shionogi
cells and oestrogen-sensitive ZR-75-1 cells
5.3.1. Proliferative activity in Shionogi (AR+) mammary
cells
Assays of the proliferation of androgen-sensitive
(AR+) Shionogi mammary carcinoma cells were carried
out according to the procedure previously described by
Sam et al. [45]. The selected compounds were tested at
two concentrations, 0.1 and 1 mM.
5.2.2. Jeg-3 cells as source of steroid sulphatase
(enzymatic assay)
5.3.2. Proliferative activity in ZR-75-1 (ER+) cells
Assays of the proliferation of oestrogen-sensitive
(ER+) ZR-75-1 human breast cancer cells were carried
out according to the procedure previously described by
Poirier et al. [46]. The selected compounds were tested at
two concentrations, 0.03 and 1 mM.
The Jeg-3 cells, also used as source of sulphatase
activity, were purchased from ATCC, on October 10th
1989, at passage 127. Cells were grown in DMEM
medium (Flow) containing NaHCO3 (3.7 g L−1), Hepes
(5.96 g L−1) and glucose (4.59 g L−1). The medium,
sterilised on Millipore 0.22 mm membrane, was supple-
mented with foetal bovine serum or bovine calf serum
(5%),
L
-glutamine (1%), penicillin (100 IU mL−1), and
Acknowledgements
streptomycin (50 mg mL−1). Cells were grown in 175
cm2 flasks seeded at a density of 500,000 cells per flask.
The medium was changed every 2–3 days. Cells were
subcultured weekly by gentle digestion in Hepes
buffered enzyme solution (Pancreatine: EDTA, 0.83%: 3
mM), 15 min at 37 °C. The enzymatic activity was
inhibited by diluting the cells in culture medium con-
taining 5% serum. Cells were pelleted, resuspended in
culture medium, counted with a haemocytometer and
reseeded in culture flasks or harvested for subsequent
analysis.
The authors thank the Medical Research Council of
Canada (MRC) for an operating grant and Le Fonds de
la Recherche en Sante´ du Que´bec (FRSQ) for a fellow-
ship (D.P.). We are grateful to Dr Jacques Simard and
Mrs Diane Michaud for performing cell proliferative
assays. Special thanks are forwarded to the Oncology
and Molecular Endocrinology Research Center (Dr Fer-
nand Labrie, Director) for technical assistance as well as
chemical and biological facilities.
For the assay, the Jeg-3 cell homogenate was pre-
pared by repeated freezing (−80 °C) and thawing (5
References
times), and homogenisation using
a Dounce ho-
[1] Poirier D., Ciobanu L.C., Maltais R., Exp. Opin. Ther. Patents
9 (1999) 1083–1099.
[2] Reed M.J., Purohit A., Woo L.W.L., Potter B.V.L., Endocrine-
Related Cancer 3 (1996) 9–23.
mogeniser. The reaction was carried out at 37 °C in
1.25 mL of 100 mM Tris–acetate buffer (pH 7.4) con-
taining 5 mM of EDTA, 10% glycerol, 100 mM of
[3H]DHEAS or [3H]E1S as substrate, and an ethanolic
solution of compound to test (at appropriate concentra-
tions). After 2 h of incubation, the reaction was stopped
by addition of 1.25 mL of xylene. The tubes were then
shaken and centrifuged at 2000g for 10 min to separate
the organic and aqueous phases. Radioactivity in 750
mL of each phase (organic: free steroids; aqueous: sul-
phated steroids) was determined by liquid scintillating
counting with a Beckman LS3801 (Irvine, CA). Data
were expressed as percentage (%) of DHEA or E1 pro-
[3] Adams J.B., Mol. Cell. Endocrinol. 41 (1985) 1–17.
[4] Labrie F., Dupont A., Be´langer A., in: De Vita V.T., Hullman
S., Rosenberg S.A. (Eds.), Important Advances in Oncology,
Lippincott JB, Philadelphia, PA, 1985, pp. 193–217.
[5] Hawkins R.A., Thomson M.L., Killen E., Breast Cancer Res.
Treat. 6 (1985) 75–87.
[6] Pasqualini J.R., Gelly C., Nguyen B.L., Vella C., J. Steroid
Biochem. (Proc.) 34 (1989) 155–163.
[7] Labrie F., Luu-The V., Lin S.X., Labrie C., Simard J., Breton
R., Be´langer A., Steroids 62 (1997) 148–158.
[8] James V.H.T., Mc Neill J.M., Lai L.C., Newton C.J., Ghilchik
M.W., Reed M.J., Steroids 50 (1987) 269–279.

L.C. Ciobanu et al. / European Journal of Medicinal Chemistry 36 (2001) 659–671
671
[9] Fishman J., Nisselbaum J.S., Menendez-Botet C.J., Schwartz [27] Brodie A.M.H., Njar V.C.O., Steroids 65 (2000) 171–179.
M.K., J. Steroid Biochem. 8 (1977) 893–896.
[28] Lombardi P., Curr. Pharm. Des. 1 (1995) 23–50.
[10] Ruder H.J., Loriaux D.L., Lipsett M.B., J. Clin. Invest. 51
[29] Brueggemeier R.W., Breast Cancer Res. Treat. 30 (1994) 31–
(1972) 1020–1023.
42.
[11] Labrie F., Mol. Cell. Endocrinol. 78 (1991) C113–C118.
[30] Li X., Singh S.M., Labrie F., Tetrahedron Lett. 35 (1994)
[12] Pasqualini J.R., Ebert C., Chetrite G.S., Gynecol. Endocrinol.
1157–1160.
13 (Suppl. 6) (1999) 1–8.
[31] Mijares A., Cargill D.I., Glasel J.A., Lieberman S., J. Org.
[13] Masamura S., Santner S.J., Santen R.J., J. Steroid Biochem.
Chem. 32 (1967) 810–812.
Mol. Biol. 58 (1996) 425–429.
[32] Mann J., Pietrzak B., J. Chem. Soc., Perkin Trans. 1 (1983)
[14] Woo L.W.L., Purohit A., Malini B., Reed M.J., Potter B.V.L.,
2681–2685.
Chem. Biol. 7 (2000) 773–791.
[33] Mancera O., Ringold H.J., Can. J. Chem. 37 (1959) 1785–1788.
[34] Bydal P., Sam K.M., Poirier D., Steroids 61 (1996) 349–353.
[15] Billich A., Nussbaumer P., Lehr P., J. Steroid Biochem. Mol.
Biol. 73 (2000) 225–235.
[35] Howard N.M., Purohit A., Reed M.J., Potter B.V.L., J. Med.
[16] Ahmed S., James K., Patel C.K., Biochem. Biophys. Res.
Chem. 37 (1994) 219–221.
Commun. 272 (2000) 583–585.
[36] Adams J., Garcia M., Rochefort H., Cancer Res. 41 (1981)
[17] Romer J., Steinbach J., Kasch H., Scheller D., J. Prakt. Chem.
4720–4726.
341 (1999) 574–583.
[37] Poulin R., Labrie F., Cancer Res. 46 (1986) 4933–4937.
[38] Simard J., Labrie F., J. Steroid Biochem. 26 (1987) 539–546.
[18] Poirier D., Boivin R.P., Bioorg. Med. Chem. Lett. 8 (1998)
1891–1896.
[19] Boivin R.P., Luu-The V., Lachance R., Labrie F., Poirier D., J.
[39] Thieulant M.L., Benie T., Michaud S., Klein H., Vessieres A.,
Med. Chem. 43 (2000) 4465–4478.
J. Steroid Biochem. 19 (1983) 241–246.
[20] Bowler J., Lilley T.J., Pittam J.D., Wakeling A.E., Steroids 54
[40] Garcia M., Rochefort H., Endocrinology 104 (1979) 1797–
(1989) 71–79.
1804.
[21] Wakeling A.E., Bowler J., J. Endocrinol. 11 (1987) R7–R10.
[41] Garcia M., Rochefort H., Cancer Res. 38 (1978) 3922–3929.
[22] Von Angerer E., The estrogen receptor as a target for rational
drug design, Molecular Biology Intelligence Unit, R.G. Landes
Company, Austin, TX, 1995, pp. 1–237.
[23] Boivin R.P., MSc Thesis, Universite´ Laval, Que´bec, Canada,
1994, pp. 81–117.
[42] Ciobanu L.C., Boivin R.P., Luu-The V., Labrie F., Poirier D.,
J. Med. Chem. 42 (1999) 2280–2286.
[43] Fieser L.F., Fieser M. (Eds.), Steroids, Reinhold, New York,
1959, p. 339 and 344.
[44] Kingston E., Curr. Opin. Cell. Biol. 1 (1989) 1081–1087.
[24] Boivin R.P., Labrie F., Poirier D., Steroids 64 (1999) 825–833.
[45] Sam K.M., Labrie F., Poirier D., Eur. J. Med. Chem. 35 (2000)
[25] Carlstro¨m K., Do¨berl A., Rannevik G., Fertil. Steril. 39 (1983)
499–504.
217–225.
[46] Poirier D., Auger S., Me´rand Y., Simard J., Labrie F., J. Med.
Chem. 37 (1994) 1115–1125.
[26] Carlstro¨m K., Do¨berl A., Pousette A., Rannevik G., Wilking
N., Acta Obstet. Gynecol. Scand. Suppl. 123 (1984) 107–111.