Steroidal lactones as inhibitors of 17β-hydroxysteroid dehydrogenase type 5: Chemical synthesis, enzyme inhibitory activity, and assessment of estrogenic and androgenic activities

Article abstract of DOI:10.1016/j.ejmech.2008.03.020

Androgens are well known to play a predominant role in prostate cancer and other androgen-dependent diseases. To decrease the level of androgen testosterone in the prostate, we are interested in developing inhibitors of 17β-hydroxysteroid dehydrogenase type 5 (17β-HSD5). This enzyme expressed in the prostate is one of the two enzymes able to convert 4-androstene-3,17-dione into testosterone. From a screening study, it was found that a series of steroid derivatives bearing a lactone on D-ring demonstrated potent inhibition of 17β-HSD5 over-expressed in HEK-293 cells. The results of enzymatic assays using intact cells indicated that a C18-steroid (estradiol or 3-deoxyestradiol) backbone and a spiro-δ-lactone (six-member ring) are important for a strong inhibitory activity. Moreover, the presence of a dimethyl group at the alpha-position of the lactone carbonyl increases the selectivity of the inhibitor toward 17β-HSD5. Compound 26, a 3-deoxyestradiol derivative with a dimethylated spiro-δ-lactone at position 17, possesses the most potent inhibitory activity for 17β-HSD5 (IC₅₀ = 2.9 nM). It showed no binding affinity for estrogen, androgen, progestin and glucocorticoid receptors (ER, AR, PR and GR). A weak proliferative effect was, however, observed on ZR-75-1 (ER⁺) cells in culture at high concentration (1 μM), but not at 0.03 μM. Interestingly, no significant proliferative effect was detected on Shionogi (AR⁺) cells in culture in the presence of 0.1 and 1 μM of lactone 26.

Full text of DOI:10.1016/j.ejmech.2008.03.020

Available online at www.sciencedirect.com
European Journal of Medicinal Chemistry 44 (2009) 632e644
http://www.elsevier.com/locate/ejmech
Original article
Steroidal lactones as inhibitors of 17b-hydroxysteroid dehydrogenase
type 5: Chemical synthesis, enzyme inhibitory activity, and
assessment of estrogenic and androgenic activities
*
Patrick Bydal, Van Luu-The, Fernand Labrie, Donald Poirier
Medicinal Chemistry Division, Oncology and Molecular Endocrinology Research Center, CHUL Research Center and University Laval,
2705 Laurier Boulevard, Que´bec, Que´bec G1V 4G2, Canada
Received 27 August 2007; received in revised form 19 March 2008; accepted 20 March 2008
Available online 8 April 2008
Abstract
Androgens are well known to play a predominant role in prostate cancer and other androgen-dependent diseases. To decrease the level of
androgen testosterone in the prostate, we are interested in developing inhibitors of 17b-hydroxysteroid dehydrogenase type 5 (17b-HSD5).
This enzyme expressed in the prostate is one of the two enzymes able to convert 4-androstene-3,17-dione into testosterone. From a screening
study, it was found that a series of steroid derivatives bearing a lactone on D-ring demonstrated potent inhibition of 17b-HSD5 over-expressed in
HEK-293 cells. The results of enzymatic assays using intact cells indicated that a C18-steroid (estradiol or 3-deoxyestradiol) backbone and
a spiro-d-lactone (six-member ring) are important for a strong inhibitory activity. Moreover, the presence of a dimethyl group at the alpha-po-
sition of the lactone carbonyl increases the selectivity of the inhibitor toward 17b-HSD5. Compound 26, a 3-deoxyestradiol derivative with a di-
methylated spiro-d-lactone at position 17, possesses the most potent inhibitory activity for 17b-HSD5 (IC50 ¼ 2.9 nM). It showed no binding
affinity for estrogen, androgen, progestin and glucocorticoid receptors (ER, AR, PR and GR). A weak proliferative effect was, however, observed
on ZR-75-1 (ERþ) cells in culture at high concentration (1 mM), but not at 0.03 mM. Interestingly, no significant proliferative effect was detected
on Shionogi (ARþ) cells in culture in the presence of 0.1 and 1 mM of lactone 26.
Ó 2008 Elsevier Masson SAS. All rights reserved.
Keywords: Hydroxysteroid dehydrogenase; Enzyme; Inhibitor; Lactone; Steroid; Androgen; Testosterone; Cancer
1. Introduction
virilisation of patients having testicular 17b-HSD (type 3) de-
ficiency in adulthood and the finding of a 17b-HSD5 [4e6]
that possesses the ability to convert 4-androstene-3,17-dione
(D4-dione) into testosterone (T) strongly suggest the presence
of an additional androgen biosynthetic pathway in peripheral
tissues [7]. Indeed, in the case of prostate tissue, an important
prohormone, namely dehydroepiandrosterone (DHEA), was
found to be responsible for nearly 40% of the active androgen
concentration in the prostate. Although the presence of DHEA
in the circulation of men was reported many years ago [8], it is
only recently that its important role in peripheral steroidogen-
esis has been elucidated [9]. In fact, DHEA-S, the sulfated
form of DHEA, is the most abundant steroid in the circulation
in men [10]. It is also clearly demonstrated that the prostatic
tissue efficiently transforms the inactive steroid precursors
DHEA-S and DHEA (Fig. 1), into the active androgens T
Prostate cancer is the second most frequent cancer for men
in North America [1]. Since the prostate is an androgen-depen-
dent organ, most of prostate cancer cells are highly sensitive to
androgen deprivation. In fact, among all hormone-sensitive
cancers, prostate cancer has the best response to endocrine
therapy. Accordingly, for more than 50 years, the exclusive
treatment of advanced metastatic disease has been androgen
deprivation [2]. Although it is well accepted that the majority
of circulating androgens were of testicular origin [3], the
* Corresponding author. Tel.: þ1 418 654 2296; fax: þ1 418 654 2761.
E-mail address: donald.poirier@crchul.ulaval.ca (D. Poirier).
0223-5234/$ - see front matter Ó 2008 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2008.03.020

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
633
2. Results
2.1. Chemical synthesis of lactones 23e26
The synthesis of lactone 23 reported in Scheme 1 started
from 16a-allyl-17a-estradiol (28), an intermediate obtained
from estrone (27). We previously reported the sequence of re-
actions needed to generate 28 as well as its full characteriza-
tion of the stereochemistry of C16 and C17-substituents
[20]. The key step in this transformation was the inversion
of 17b-OH into 17a-OH, which was achieved by a Mitsunobu
reaction using triphenylphosphine, diethylazodicarboxylate,
and p-nitrobenzoic acid in toluene, followed by hydrolysis
of the resulting ester with K2CO3 in methanol. Diol 28 was
protected as the di-TBDMS derivative 29 and its allyl group
was converted into a primary alcohol by an oxidative hydrobo-
ration using BH3 in THF followed by a treatment with H2O2
and a solution of aqueous 3 N NaOH. Alcohol 30 was oxidized
to carboxylic acid with Jones reagent and then treated with
BF3$Et2O in CH2Cl2 overnight. Under these conditions, the si-
lylated ethers were cleaved with lactonization to yield the tar-
get lactone 23.
Fig. 1. Illustration of the last biosynthetic steps involved in the formation of
the active androgens T and DHT from the inactive precursor DHEA-S in hu-
man prostate cells. The new steroid lactones act as inhibitors of 17b-HSD5 and
thus prevent agonistic activity of T and DHT on the androgen receptor (AR).
DHEA-S: dehydroepiandrosterone-sulfate; DHEA: dehydroepiandrosterone;
D5-diol: 5-androstene-3b,17b-diol; D4-dione: 4-androstene-3,17-dione; T: tes-
tosterone; DHT: dihydrotestosterone; STS: steroid sulfatase; HSD: hydroxys-
teroid dehydrogenase; 5a-R: 5a-reductases.
The synthesis of lactones 24e26 with a 3-deoxygenated es-
tradiol nucleus was described in Scheme 2. From commer-
cially available estrone (27), the hydroxyl group in position
3 was esterified with trifluoromethanesulfonic anhydride and
the corresponding estrone-triflate was reduced in the presence
of triethylamine, formic acid, triphenylphosphine and palla-
dium diacetate to give 31 [21]. C3-deoxygenated estrone
(31) was then alkylated with the lithium acetylenide resulting
from the reaction of tetrahydro-(butynyl)-2-H-pyran and n-
BuLi to give 32 exclusively as a 17a-alkyne isomer. The triple
bond was hydrogenated in EtOAc with palladium on activated
charcoal. Without further purification, the tetrahydropyranyl
ether of the side chain was hydrolyzed by a treatment with
p-TSA in MeOH. Oxidation of this primary alcohol with Jones
reagent in acetone gives the carboxylic acid, which, as ex-
pected, underwent an intramolecular cyclization with the ter-
tiary 17b-hydroxy to give the corresponding spiro-d-lactone
24. Using LDA as a base and methyl iodide in excess, 24
was mono and dimethylated in alpha-position of the carbonyl
providing 25 and 26. In the case of the monomethylated lac-
tone 25, we obtained a mixture of R and S isomers, but only
the major isomer was characterized and kept for the biological
assays.
and dihydrotestosterone (DHT) [11]. Therefore, the inhibition
of the formation of active steroids in the prostate from inactive
precursors is a potential applicable therapy for prostate cancer.
Because T is also an active androgen, but less than DHT, the
use of 5a-reductase inhibitors is not sufficient for a complete
blockade of androgenic effect. The recent discovery of 17b-
HSD5 [4e6], which is located in the peripheral tissues such
as the prostate [12], gives us a new interesting therapeutic tar-
get. The blockade of 17b-HSD5 could thus prevent the forma-
tion of androgen T while accumulating D4-dione (Fig. 1),
a steroid inactive on the androgen receptor [13].
Few inhibitors of 17b-HSD5 have been reported until now
[14]. In addition to the lactones discussed in this manuscript
and previously patented [15], we can mention a series of di-
etary phytoestrogens studied by Krazeisen et al. [16], a series
of commercially available cinnamic acids and related com-
pounds synthesized by Brozic et al. [17] and three fluorinated
estratriene derivatives identified by Deluca et al. [18] as mod-
erate inhibitors. Some oxirane derivatives of natural enzyme
substrates (T, progesterone, 5a-androstane-3,17-dione and
androsterone) were also proposed as potential inhibitors by
Penning et al. [6]. More recently, the same group also used
N-phenylanthranilic acid as a template for designing inhibi-
tors of 17b-HSD5 [19]. Herein we present a series of steroid
lactones (Fig. 2) and their ability to inhibit the in vitro trans-
formation of D4-dione into T by 17b-HSD5 over-expressed in
HEK-293 cells. Since it has been reported that similar com-
pounds show inhibitory activity against 17b-HSD2 [14], we
decided to verify the specificity of lactones 11, 13, 14,
24e26, selected from lactones 1e26, for the inhibition of
17b-HSD5. In addition to their binding affinity for steroid re-
ceptors, the proliferative and antiproliferative effects on ZR-
75-1 (ERþ) and Shionogi (ARþ) cells have also been
determined.
2.2. Inhibitory activity of lactones 1e26 on human
17b-HSD5
17b-HSD5 catalyzes the conversion of D4-dione into T and
its inhibition could be a strategy to lower the level of androgen
T, and indirectly to lower the level of potent androgen DHT, in
human prostate cells (Fig. 1). Since a C19 steroid such as D4-
dione is a good substrate for 17b-HSD5, a series of C19-ste-
roids bearing a spiro-d-lactone were first tested for their ability
to inhibit the reductive activity of 17b-HSD5 transfected in
HEK-293 cells (Table 1). We focused on steroidal lactones

634
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
Fig. 2. Chemical structures of C19-steroid lactones 1e5 and C18-steroid lactones 6e26 tested as inhibitors of human 17b-HSD5.
because they were previously identified as inhibitors of
17b-HSD2 [22,23]. While mono-spiro-g-lactones 1e4 were
similarly active (57, 61, 45 and 63% at 3 mM, respectively),
bi-spiro-g-lactone 5 demonstrated a weaker activity with
only 29% of inhibition at this concentration.
In order to compare the C19-steroids to C18-steroids, sev-
eral estradiol (E2) derivatives (mainly as lactone derivatives)
were next tested in our assay. At concentrations of 0.3 and
3 mM, spiro-g-lactone 6 showed higher enzyme inhibition
than the C19-steroid analogue (47 and 79% for 6 compared
to 7e27 and 29e63% for 1e5), establishing the importance
of the C18-steroid backbone. The importance of the carbonyl
function was evaluated by testing two O-methylated spiro-g-
lactols (compounds 7 and 8) and spiro-g-lactol 9. The lactol
derivative demonstrated the best inhibitory activity of this se-
ries at 3 mM (61% compared to 4 and 16%, respectively for 7
and 8) but remains less potent than the corresponding spiro-g-
lactone 6 (79%). We then determined the effect of the
orientation of the lactone pharmacophore by testing a 17a-
O-spiro-g-lactone (compound 10). This lactone had less
inhibitory activity (22 and 45%) than the corresponding 17b-
O-spiro-g-lactone 6 (47 and 79%) at both concentrations of
0.3 and 3 mM. Thus, the optimum orientation of oxygenated
counterpart of the 17-spirolactone moiety was found to be in
b position (or S configuration). The two other C18-steroid spi-
rolactones 11 and 12 were also tested. Thus, while spiro-g-
lactone 6 (five-member ring) demonstrated only 79% of inhi-
bition at 3 mM, spiro-d-lactone 11 (six-member ring) and
spiro-e-lactone 12 (seven-member ring) respectively inhibited
95 and 93% of the enzyme activity at 3 mM. They also keep
good inhibitory activities (92 and 90%) at lower concentration
of 0.3 mM.

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
635
Scheme 1. Synthesis of 16a,17a-lactone-E2 23. Reagents and conditions are (a) TBDMS-Cl, imidazole, DMF, rt; (b) (i) BH3, THF, 0 ꢁC, (ii) H2O2, NaOH; (c)
Jones reagent (2.7 M), acetone, 0 ꢁC; (d) BF3$Et2O, CH2Cl2, rt.
If we go over the results so far obtained, we conclude that
the presence of a C18-steroid E2 backbone and a spiro-d- or
a spiro-e-lactone are two important requirements for the inhi-
bition of 17b-HSD5. Spiro-d-lactone was, however, retained
over spiro-e-lactone considering its better stability and facility
of chemical synthesis. The effect of a substituent on the spiro-
d-lactone moiety was next addressed with lactones 13e19.
The methylated lactones 13e15 demonstrated good inhibitory
activities at 0.3 mM (93, 86 and 88%, respectively). The lac-
tones with more hindered substituents like the allyl derivative
16 and the propyl derivative 17 were less active (68 and 41%,
respectively). Two other spiro-d-lactones were also tested: me-
thoxycarbonylated spiro-d-lactone 18 and a,b-unsaturated
spiro-d-lactone 19. The idea with compound 18 was to add an-
other carbonyl function on the pharmacophore since we dem-
onstrated earlier that this function was important for
inhibition. Methoxycarbonylated spiro-d-lactone 18 showed
good inhibitory activity with 92% at 3 mM, which is similar
to spiro-d-lactone 11 (95%). a,b-Unsaturated spiro-d-lactone
19 also demonstrated good inhibitory activity with 92% at
3 mM. Both compounds remain, however, weaker inhibitors
than unsubstituted lactone 11 at 0.3 mM.
and 12 at both concentrations. The inhibitory effect of C16/
C17-stereochemistry was also addressed by testing compound
23. This 16a/17a-O-lactone is clearly less active than the 16b/
17b-O-lactone 21 (6 and 55% of inhibition at 0.3 mM, respec-
tively). The same result has been also obtained for 17-spiro-g-
lactones 10 (17a-O) and 6 (17b-O).
After we established the better inhibitory activity of 17-spi-
rolactone over 16,17-lactones, the deoxygenated analogues of
11, 13, and 14, compounds 24e26, were synthesized and
tested as inhibitors of 17b-HSD5. The removal of C3-OH
group as well as mono or dimethylation is not detrimental
for the inhibitory activity, since compounds 24e26 inhibited
more than 90% of the enzyme activity at 0.3 mM.
Because the screening results were very similar for the
most potent 17b-HSD5 inhibitors and for a better comparison,
a complete curve of inhibition was obtained for each com-
pound that inhibited over 90% of the enzyme activity at a con-
centration of 3 mM. From these curves, it is possible to obtain
the IC50 values, which is the concentration of inhibitor that
causes 50% of inhibition (Table 1). Spiro-d-lactone 11 showed
an IC50 value lower than spiro-e-lactone 12 and 16b/17b-O-
lactone 22 (20 ꢀ 2, 42 ꢀ 13, and 66 ꢀ 2 nM, respectively).
Allyl-spiro-d-lactone 16 and methoxycarbonylated spiro-d-
lactone 18 (107 ꢀ 8 and 72 ꢀ 3 nM, respectively), which are
the more hindered lactones, did not show an IC50 value as sig-
nificant as the methylated products 13e15 (16 ꢀ 1, 28 ꢀ 1 and
19 ꢀ 2 nM, respectively). a,b-Unsaturated spiro-d-lactone
19 gave an intermediate IC50 of 49 ꢀ 5 nM. Interestingly,
To verify the importance of the 17a/17b-O-spirolactone
ring, we tested a series of 16b/17b-O-lactones (compounds
20e22) that differ only by the lactone-ring size (5, 6 and 7
members). Lactones 20e22 have interesting inhibitory activity
(62, 85 and 91%, respectively) at 3 mM, but they are less po-
tent inhibitors than their corresponding 17-spirolactones 6, 11

636
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
Scheme 2. Synthesis of 17-spiro-d-lactone-3-deoxy-E2 24e26. The reagents and conditions are (a) (i) HC^C(CH2)2OTHP, THF, n-BuLi, 0 ꢁC, (ii) ketone
31, THF, ꢂ78 ꢁC to rt; (b) H2, Pd/C (10%), EtOAc, rt; (c) p-TSA, MeOH, rt; (d) Jones reagent (2.7 M), acetone, 0 ꢁC to rt; (e) LDA, CH3I, THF, ꢂ78 ꢁC
to rt.
3-deoxygenated-E2 is a good nucleus for inhibition of
17b-HSD5 giving IC50 values of 10, 6.1 and 2.9 nM for
24e26, respectively. As a point of comparison, 3-OH analogue
lactones 11, 13 and 14 gave IC50 values of 10, 16, and 28 nM,
respectively. With an IC50 value of 2.9 ꢀ 0.4 nM, dimethylated
compound 26 is the most potent inhibitor obtained in
our study.
Indeed, no significant inhibition of 17b-HSD2 activity was ob-
served with dimethylated lactone 26. Thus, the selectivity of
inhibition was observed only for lactones 14 and 26 having
two methyl groups on the lactone ring. Interestingly, they
are among the most active 17b-HSD5 inhibitors. In these
cases, the presence of two methyl groups did not reduce the
inhibitory potency of lactone on 17b-HSD5, although drasti-
cally reducing its ability to inhibit 17b-HSD2. Furthermore,
the presence or the absence of a C3-OH did not influence
the inhibitory results on both type 2 and type 5 17b-HSDs.
2.3. Inhibitory activity of lactones on human 17b-HSD2
For a therapeutic use of a 17b-HSD5 inhibitor in the pros-
tate, the candidate should not inhibit 17b-HSD2, an oxidative
enzyme involved in the degradation of active 17b-OH steroids
as T into D4-dione. Fig. 3 clearly illustrates the selectivity of
inhibition for dimethylated lactones 14 and 26 as well as the
fact that the removal of the OH group in position 3 of E2 lac-
tone does not affect its inhibitory activity potency on 17b-
HSD2. In Fig. 3A, the E2 derivatives 11, 13 and 14 were tested
for their ability to inhibit the reductive transformation of D4-
dione into T in homogenated cells. Clearly, dimethylated lac-
tone 14 did not inhibit the 17b-HSD2 activity contrary to
monomethylated and unmethylated lactones 13 and 11, re-
spectively. A similar result was also obtained for the deoxyes-
tradiol derivatives 24e26 when using the oxidative
transformation of T into D4-dione in intact cells (Fig. 3B).
2.4. Inhibitory activity of lactones 11, 24e26 on human
17b-HSD1 and 17b-HSD3
The C18-estrane nucleus of 17b-HSD5 inhibitors discussed
above is closely related to the natural substrate estrone (E1) of
17b-HSD1. Although this common feature, spiro-g-lactone 6
did not inhibit the transformation of E1 into E2 by homogen-
ated HEK-293 cells over-expressing human 17b-HSD1 [24].
The same result was also obtained with spiro-d-lactone 11
having an E2 nucleus (Fig. 4A). For lactones 24e26 having
a deoxygenated-E2 nucleus, a weak inhibition (21e27%)
was observed for 24 and 25 at both concentrations of 0.1
and 1 mM, but 26 appears to be a better inhibitor with 50%
of inhibition at the higher concentration of 1 mM. We next

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
637
Table 1
Inhibition of human 17b-HSD5 activity by steroid lactones 1e26a
Compound
Inhibition (%)
at 0.3 mMb
Inhibition (%)
at 3 mMb
IC50 (nM)b
1
27
19
14
15
7
57
61
45
63
29
79
4
e
2
e
e
3
4
e
5
e
6
47
0
e
e
7
8
0
16
61
45
95
93
94
94
94
91
83
92
92
62
85
91
52
95
94
95
45
e
9
19
22
92
90
93
86
88
68
41
78
79
30
55
70
6
e
e
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
D4-dione
a
20 ꢀ 2 (10 ꢀ 2)c
42 ꢀ 13
16 ꢀ 1
28 ꢀ 1
19 ꢀ 2
107 ꢀ 8
e
72 ꢀ 3
49 ꢀ 5
e
e
66 ꢀ 2
e
93
91
94
21
(10 ꢀ 1)c
(6.1 ꢀ 0.7)c
(2.9 ꢀ 0.4)c
e
See Fig. 2 for the chemical structures of compounds 1e26.
For the transformation of [14C]-D4-dione (100 nM) into [14C]-T in intact
HEK-293 cells that over-express human 17b-HSD5.
b
c
Data between parentheses were obtained under the same assay conditions
but in another experiment.
tested lactones 11, 24e26 as inhibitors of the transformation
of D4-dione into T catalyzed by 17b-HSD3 (Fig. 4B). Al-
though spiro-d-lactone 11 with an E2 nucleus weakly in-
hibited (25 and 34%) the 17b-HSD3 activity at 0.1 and
1 mM, respectively, the three lactones 24e26 with a deoxygen-
ated-E2 nucleus are clearly better inhibitors (25e45 and
64e85% at 0.1 and 1 mM, respectively). Lactones 24e26
thus inhibit both 17b-HSD3 and 17b-HSD5, two key enzymes
involved in the formation of potent androgen T in testis and
prostate.
2.5. Estrogenic and antiestrogenic activities of lactones
14 and 26
The estrogen-dependent ZR-75-1 cells, which contain the
estrogen receptor (ERþ), were selected to assess the estro-
genic and antiestrogenic activities of 14 and 26. In our assay,
proliferation was determined by measuring the DNA content
of cells after 9 days of treatment. The cell proliferation in-
duced by 0.1 nM of the potent estrogen E2 was fixed to
100% and used as a positive control of cell growth (Fig. 5).
On the other hand, the pure antiestrogen EM-139 [25] was
added as a reference compound giving no stimulation of
Fig. 3. Inhibition of human 17b-HSD2 over-expressed in HEK-293 cells by
spiro-d-lactones 11, 13, 14, 24e26. The results are reported in percentage
(%) of inhibition at two inhibitor concentrations (0.1 and 1 mM). (A) Data ob-
tained from Ref. [23] for the transformation of [3H]-D4-dione (4 nM) into
[3H]-T (homogenated HEK-293 transfected cells). (B): Compounds were
tested for the transformation of [14C]-T (100 nM) into [14C]-D4-dione (homo-
genated HEK-293 transfected cells).

638
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
Fig. 5. Effect of two concentrations of pure antiestrogen EM-139, the phenolic
dimethylated lactone 14, and deoxygenated dimethylated lactone 26 on the
basal and E2 (0.1 nM)-induced growth of estrogen sensitive (ERþ) ZR-75-1
cell line. Results are expressed as mean ꢀ SEM of triplicates.
2.6. Androgenic and antiandrogenic activities of
dimethylated lactones 14 and 26
The androgen-dependent Shionogi cell line was selected to
assess the androgenic and antiandrogenic activities of 14 and
26. Proliferation was determined by measuring the DNA con-
tent of cells after 10 days of treatment (Fig. 6). These cells
contain the androgen receptor (ARþ) and the potent androgen
DHT (0.1 nM) stimulated about 2.5-fold the cell growth. This
level of cell growth stimulation was fixed at 100%. In the first
part of the assay, compounds 14 and 26 showed no prolifera-
tive effect (androgenic activity) on Shionogi cells. In the sec-
ond part, the ability of a compound to reverse the cell
proliferation induced by DHT (0.1 nM) was determined. At
the two concentrations tested (0.1 and 1 mM), lactones 14
and 26 did not significantly inhibit the proliferation of Shio-
nogi cells induced by DHT indicating that they are not antian-
Fig. 4. Inhibition of human 17b-HSD1 (A) and human 17b-HSD3 (B) over-ex-
pressed in HEK-293 cells by spiro-d-lactones 11, 24e26. The results are re-
ported in percentage (%) of inhibition at two inhibitor concentrations of 0.1
and 1 mM.
cell proliferation. The two 17b-HSD5 inhibitors 14 and 26
follow the same pattern of activity with no significant prolif-
erative (estrogenic) effect observed at 0.03 mM and a moderate
effect at 1 mM (30 and 50%). Thus, the addition of a spiro-d-
lactone moiety on the E2 nucleus greatly reduced its estro-
genic activity. The ability of these lactones to inhibit the E2
(0.1 nM)-induced stimulation of ZR-75-1 cells was also eval-
uated as a measure of their antiestrogenic activity. A signifi-
cant antiestrogenic effect was observed for phenolic lactone
14 at the higher concentration of 1 mM, but not at 0.03 mM.
No antiestrogenic effect was observed at both concentrations
for lactone 26. Under the same conditions, the antiestrogen
EM-139 fully inhibited the cell proliferation induced by E2
(0.1 nM).
drogenic compounds. As
a point of comparison, the
antiandrogen hydroxy-flutamide [26] inhibited 94% of the
DHT-induced Shionogi cell proliferation at 1 mM but did not
induce the cell proliferation at the two concentrations tested.
2.7. Steroid receptor binding affinity
The binding affinities of lactones 14 and 26 on four rat hor-
monal receptors are reported in Table 2. Although these

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
639
ER and PR at 0.1 mM increased to 80 and 35% at 10 mM, re-
spectively. In fact, an OH group at position C-3 of an E2 nu-
cleus is well known to promote ER binding by interacting
(hydrogen bond) with key amino acids [27].
3. Discussion
Our goal was to obtain a 17b-HSD5 inhibitor that exhibits
no estrogenic and no androgenic activities. In fact, a residual
estrogenic activity could generate a side effect such as femini-
zation whereas a residual androgenic activity will stimulate
the ARþ prostate cancer cell proliferation. As mentioned
above and illustrated in Fig. 1, the enzyme 17b-HSD5 is re-
sponsible for the transformation of D4-dione into the active an-
drogen T, which is reduced by 5a-reductase into the most
potent androgen DHT. Although numerous isoforms of 17b-
HSD are known [38e40], type 5 is a key target since it is
mainly expressed in peripheral tissues [41]. Thus, the inhibi-
tion of 17b-HSD5 with a specific inhibitor that is not andro-
genic could be a good strategy to block the production of T,
and indirectly DHT, in prostate cancer cells.
The present data show that a series of C19-steroid and C18-
steroid derivatives bearing a lactone on D-ring exert inhibitory
effects on 17b-HSD5. The C18-steroid backbone is, however,
a better nucleus than the C19-steroid backbone. It was estab-
lished that the carbonyl function was important because O-
methylated spiro-g-lactols 7 and 8 and spiro-g-lactol 9 were
less active than the corresponding spiro-g-lactone 6. This
last statement suggests that a hydrogen bonding occurs be-
tween the carbonyl and an amino acid residue of the enzyme.
Furthermore, the orientation of the oxygenated counterpart of
spirolactone was important for inhibition, and the stereochem-
istry of the spirolactone giving good inhibition was beta O-ori-
ented (as 6) instead of alpha O-oriented (as 10). We can also
conclude that 17-spirolactone showed a better inhibitory activ-
ity than 16b/17b-O lactone and that the six-member ring
spiro-d-lactone 11 is the most potent inhibitor.
Although the substituted spiro-d-lactones were not found to
be more active inhibitors of 17b-HSD5 than unsubstituted lac-
tone, the dimethylated derivative was found to be a selective
inhibitor of 17b-HSD5. In fact, dimethylated spiro-d-lactone
did not inhibit 17b-HSD2, another 17b-HSD isoform that is
well known to perform the oxidative transformation of T
into D4-dione. Since it was preferable to not inhibit 17b-
HSD2, because this isoform deactivates the potent androgen
T, the dimethylated strategy appears to be a significant finding.
A dimethylated spiro-d-lactone synthesized from a deoxygen-
ated-E2 nucleus (no OH at position 3) was also clearly less
estrogenic than the OH analogue. Thus, dimethylated spiro-
d-lactone 26 with a 3-deoxygenated E2 nucleus represents
the most potent inhibitor of 17b-HSD5 (IC50 ¼ 2.9 nM) in
our study. This lactone inhibited the reductive transformation
performed by 17b-HSD3 and 17b-HSD5, two key enzymes in-
volved in the formation of potent androgen T, but did not in-
hibit the oxidative activity of 17b-HSD2. Lactone 26
showed no binding affinity for classical hormonal receptors
(ER, AR, PR and GR). No significant proliferative effect
Fig. 6. Effect of two concentrations of pure antiandrogen hydroxyl-flutamide
(OH-Flu), the phenolic dimethylated lactone 14, and deoxygenated dimethy-
lated lactone 26 on the basal and DHT (0.3 nM)-induced growth of androgen
sensitive (ARþ) Shionogi cell lines. Results are expressed as mean ꢀ SEM of
triplicates.
receptor preparations are not of human origin, it is well docu-
mented that they can be used for the study of potential drugs
[27e37]. Thus, a rat receptor preparation allows us to easily
obtain a good general idea of the binding affinity of tested
compounds 14 and 26 in comparison to the binding of a known
agonist compound as the natural receptor ligand or and ana-
logue. At the two concentrations tested, no significant binding
affinity was observed on ER, AR, PR and GR for lactone 26
having no OH group at position C-3. For lactone 14, with an
OH group at C-3, the weak bindings (w10%) obtained on
Table 2
Binding affinity of lactones 14 and 26 for estrogen (ER), androgen (AR), pro-
gesterone (PR), and glucocorticoid (GR) rat receptors
Compound Concentration ER binding AR binding PR binding GR binding
(mM)
(%)
(%)
(%)
(%)
14
14
0.1
10
10 ꢀ 3
80 ꢀ 4
0
0
11 ꢀ 1
35 ꢀ 1
0
0
6 ꢀ 1
24 ꢀ 2
0
26
0.01
1
0
26
0
0
5 ꢀ 2
6 ꢀ 3
3 ꢀ 2
65 ꢀ 2
0
0
E2a
DHTa
R5050a
DEXa
a
0.01
0.01
0.01
0.01
75 ꢀ 1
2 ꢀ 2
5 ꢀ 2
0
0
5 ꢀ 2
2 ꢀ 2
9 ꢀ 2
66 ꢀ 2
70 ꢀ 1
1 ꢀ 4
0
Compounds used as controls: natural estrogen estradiol (E2), natural an-
drogen dihydrotestosterone (DHT), synthetic progestin R5050, and synthetic
glucocorticoid dexamethasone (DEX).

640
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
1
was also observed on ARþ Shionogi cells for this lactone.
However, a weak proliferative effect on ERþ ZR-75-1 cells
was observed at a high concentration (1 mM), but not at
0.03 mM or less.
(7.26 or 77.00) respectively for H and 13C. High resolution
´
mass spectra were provided by the Centre regional de spectro-
´
´
´
´
metrie de masse (Universite de Montreal, Montreal, Canada).
An HPLC system (Waters Associates, Milford, MA, USA) us-
ing an UV detector (205 or 210 nm), a reverse phase column
and an appropriate mixture of solvents was used to determine
the purity of compounds tested (see below for the type of col-
umn, solvents and the % of purity).
In addition to the classical steroidogenic tissues, namely the
ovaries, testis, adrenals and placenta, a large series of human
peripheral tissues possess all the enzymatic systems required
for the formation of active androgens from a relatively con-
stant supply of precursor steroids provided by the adrenals
[7]. While so far most therapeutic approaches have been aimed
to control steroid formation by the classical steroidal tissues or
inhibiting androgen receptor response with antagonists, it is
clear that efforts should be also turned toward steroid forma-
tion in peripheral target tissues (intracrinology) in order to
find a complementary therapy that would enhance the success
of the therapies that are already known. Additional studies on
steroidogenesis enzymes like 17b-HSDs will help us under-
stand the physiological mechanisms controlling local forma-
tion and the difference between all the 17b-HSD isoforms.
This first series of 17b-HSD5 inhibitors puts us in a position
to develop novel therapeutic approaches which take into ac-
count the high proportion of steroids made locally and respon-
sible for growth and functions of normal as well as cancerous
tissues. In fact, the discovery that a dimethylated 17-spiro-d-
lactone-E2 inhibited the reductive 17b-HSD5 selectively was
the starting point for the development of potent inhibitor
EM-1404 [15], which was recently crystallized in a binary
complex with the enzyme [42]. The availability of such com-
pounds should be also useful as a tool for a better understand-
ing of the role of 17b-HSD5 in the formation and action of
androgens in the prostate and other peripheral tissues contain-
ing this 17b-HSD isoform.
4.1.2. Compounds 1e22
Spironolactone (1) was purchased from Steraloids (Wilton,
NH) whereas C19-steroids 2e5 and C18-steroids 6e22 were
available in our laboratory. The chemical syntheses of 2e5
[43] and 6e22 [23] were previously reported.
4.1.3. Synthesis of lactone 23 (Scheme 1)
4.1.3.1. (16a,17a)-16-Allyl-3,17-bis{[tert-butyl(dimethyl)sily-
l]oxy}estra-1(10),2,4-triene (29). To a solution of diol 28 [20]
(149 mg, 0.48 mmol) in dry DMF were added imidazole
(220 mg, 3.23 mmol) and tert-butyldimethylsilyl chloride
(100 mg, 0.66 mmol). The reaction was stirred 3 days at room
temperature under an atmosphere of argon. The mixture was
then poured in ice/water and extracted with EtOAc. The organic
layer was successively washed with water and brine, dried over
MgSO4, filtered, and evaporated under reduced pressure to give
29 (115 mg, 44% yield). The mono TBDMS derivative and the
starting diol 28 were not recovered after chromatography.
White solid; IR n (film): 1640 (C]C, alkene), 1608 and 1570
1
(C]C, aromatic); H NMR d (CDCl3): 0.06 and 0.07 (2s,
Si(CH3)2 in C17), 0.20 (s, Si(CH3)2 in C3), 0.77 (s, CH3-18),
0.96 and 0.99 (2s, 2 ꢃ SiC(CH3)3), 2.80 (m, CH2-6), 3.66 (d,
J ¼ 5.0 Hz, CH-17a), 5.01 (m, CH2-30), 5.86 (m, CH-20), 6.56
(s, CH-4), 6.62 (dd, J1 ¼ 2.2 Hz and J2 ¼ 8.4 Hz, CH-2), 7.14
(d, J ¼ 8.5 Hz, CH-1); 13C NMR d (CDCl3): ꢂ4.38 (2ꢃ),
ꢂ4.05 (2ꢃ), 17.74, 18.18, 18.36, 25.73 (3ꢃ), 26.21 (3ꢃ),
26.30, 28.04, 29.84, 30.60, 32.69, 35.97, 39.18, 40.63, 43.71,
46.79 (2ꢃ), 82.16, 114.51, 117.08, 119.92, 126.08, 133.39,
137.96, 139.33, 153.17.
4. Experimental
4.1. Chemical synthesis
4.1.1. General
Chemical reagents were purchased from Aldrich Chemical
Company (Milwaukee, WI, USA) and starting steroid estrone
(27) from Sigma Chemical Company (St. Louis, MO, USA) or
Steraloids (Wilton, NH, USA). The solvents were obtained
4.1.3.2. 3-[(16a,17b)-3,17-Bis{[tert-butyl(dimethyl)silyl]oxy}
estra-1(10),2,4-trien-16-yl] propan-1-ol (30). To a stirred so-
lution of 29 (0.115 g, 0.213 mmol) in dry THF (10 mL) at
0 ꢁC was added dropwise a 1 M solution of borane in THF
(1.5 mL, 1.5 mmol). The mixture was allowed to react under
argon for 3 h, then a 3 N aqueous solution of NaOH
(0.5 mL) and H2O2 (30% w/v) (0.2 mL) were added. The re-
sulting mixture was stirred at room temperature for 1 h. The
reaction was quenched by addition of water and extraction
was done with EtOAc. The organic phase was washed with
water, washed with brine, and dried over MgSO4. After evap-
oration of organic solvent under reduced pressure, the crude
product was purified by flash chromatography, using a mixture
of hexanes and EtOAc (8/2) as eluent to give alcohol 30 as
white solid (105 mg) in 88% yield. IR n (film): 3346 (OH, al-
cohol); 1H NMR d (CDCl3): 0.05 (s, Si(CH3)2 in C17), 0.19 (s,
Si(CH3)2 in C3), 0.75 (s, CH3-18), 0.94 (s, SiC(CH3)3 in C17),
´
´
from Fischer (Montreal, Canada) and BDH (Montreal, Can-
ada). Unless otherwise noted, materials obtained from
commercial suppliers were used without further purification.
Tetrahydrofuran (THF) was distilled from sodium/benzophe-
none immediately prior to use. All reactions except those
involving water or hydrogen were conducted under argon at-
mosphere. Thin-layer chromatography (TLC) was performed
on 0.20 mm silica gel 60 F254 plates (E. Merck, Darmstadt,
GE) while 230e400 Mesh ASTM silica gel 60 (E. Merck,
Darmstadt, GE) was used for flash column chromatography.
Infrared (IR) spectra were obtained on a PerkineElmer 1600
1
(series FTIR) spectrophotometer and expressed in cmꢂ1. H
NMR and 13C NMR spectra were recorded on a Bruker AC/
F300 spectrometer at 300 and 75 MHz, respectively. The
chemical shifts (d in ppm) were referenced to chloroform

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
641
0.99 (s, SiC(CH3)3 in C3), 2.79 (m, CH2-6), 3.64 (m, CH-17a
and CH2OH), 6.55 (d, J ¼ 2.5 Hz, CH-4), 6.61 (dd,
J1 ¼ 2.4 Hz and J2 ¼ 8.4 Hz, CH-2), 7.13 (d, J ¼ 8.4 Hz,
CH-1); 13C NMR d (CDCl3): ꢂ4.40 (2ꢃ), ꢂ3.99 (2ꢃ),
17.76, 18.15, 18.35, 25.71 (3ꢃ), 26.19 (3ꢃ), 26.28, 27.51,
28.03, 29.82, 30.92, 32.23, 32.63, 39.16, 41.12, 43.73,
46.68, 46.87, 63.38, 82.30, 117.07, 119.90, 126.05, 133.37,
137.92, 153.16.
eluent to give 545 mg (68%) of alcohol 32. Colorless oil; IR
1
n (film): 3438 (OH, alcohol), 2233 very weak (C^C); H
NMR d (CDCl3): 0.88 (s, 3H, 18-CH3), 2.57 (t, J ¼ 7.0 Hz,
2H, C^CCH2), 2.87 (m, 2H, 6-CH2), 3.55 and 3.86 (2m,
4H, CH2O of side chain and CH2O of THP), 4.68 (sapp, 1H,
CH of THP), 7.13 (m, 3H, 1-CH, 2-CH, 3-CH), 7.31 (d,
J ¼ 6.7 Hz, 1H, 4-CH); 13C NMR d (CDCl3): 12.72, 19.21,
20.30, 22.76, 25.37, 26.18, 27.13, 29.49, 30.48, 32.87,
38.98, 39.14, 43.07, 47.07, 49.53, 61.95, 65.78, 79.81,
83.07, 84.68, 98.58, 125.25, 125.47, 125.51, 128.91, 136.62,
140.22; LRMS: calcd for C27H40O3N [M þ NH4]þ 426.3,
found 426.5.
4.1.3.3. (4bS,6aS,6bR,10aR,11aS,11bR)-2-Hydroxy-6a-methyl-
5,6,6a,6b,9,10,10a,11,11a,11b, 1,2,13-dodecahydronaphtho
[20,10:4,5]indeno[1,2-b]pyran-8(4bH )-one (23). Compound
30 (105 mg, 0.19 mmol) was dissolved in acetone (10 mL)
and a solution of Jones reagent (2.7 M) (0.75 mL) was added
dropwise at 0 ꢁC. After 40 min at room temperature, isopropyl
alcohol was added until a persistent green color remained and
acetone was removed under reduced pressure. The green con-
centrate was then dissolved in water, and the extraction was
done with EtOAc. The organic phase was washed with brine,
dried over MgSO4, and evaporated under reduced pressure.
The crude compound (84 mg) was dissolved in dry CH2Cl2
and BF3$Et2O (1.06 equiv) was added. The resulting mixture
was stirred overnight at room temperature. The reaction was
quenched with water, extraction was done with CH2Cl2 and
the organic layer was washed with brine and dried over
MgSO4. After evaporation of the solvent under reduced pres-
sure, the crude product was purified on column chromatogra-
phy using a mixture of hexanes and EtOAc (7/3) as eluent to
give 20 mg (33% yield, two steps) of lactone 23. IR n (film):
3350 (OH, phenol), 1716 (C]O, lactone); 1H NMR
d (CDCl3): 0.82 (s, CH3-18), 2.70 (m, 1H), 2.80 (m, CH2-6),
4.24 (d, J ¼ 6.0 Hz, CH-17a), 5.35 (broad, OH), 6.54 (d,
J ¼ 2.4 Hz, CH-4), 6.65 (dd, J1 ¼ 2.6 Hz and J2 ¼ 8.5 Hz,
CH-2), 7.13 (d, J ¼ 8.5 Hz, CH-1); 13C NMR d (CDCl3):
17.06 (C-18), 25.55 (C-10), 25.80 (C-11), 27.97 (C-7), 29.12
(C-20), 29.60 (C-6), 31.09 (C-12), 32.40 (C-15), 33.95 (C-
16), 38.81 (C-8), 43.07 (C-9), 46.30 (C-13), 48.04 (C-14),
88.92 (C-17), 112.96 (C-2), 115.23 (C-4), 126.51 (C-1),
132.18 (C-10), 137.78 (C-5), 153.62 (C-3), 174.30 (C-30);
LRMS: calcd for C21H27O3 [M þ H]þ 327.2, found 327.4;
HPLC purity: 99.1% (C-18 Nova Pak column, MeCN/H2O/
MeOH (25:50:25)).
4.1.4.2.
(8R,9S,13S,14S,17S )-13-Methyl-40,50,6,7,8,9,11,12,
13,14,15,16-dodecahydrospiro [cyclopenta[a]phenanthrene-
17,20-pyran]-60(30H )-one (24)
4.1.4.2.1. Reduction of triple bond. To a solution of alkyne
32 (650 mg, 1.6 mmol) in EtOAc was added 40 mg of palla-
dium on activated charcoal (10% w/w). The mixture was
stirred at room temperature overnight under an atmosphere
of hydrogen and was then filtered through a pad of Celite,
washed with EtOAc, and evaporated to dryness to give the cor-
responding alkane as white foam.
4.1.4.2.2. Hydrolysis of THP group. This crude THP deriv-
ative was dissolved in methanol (60 mL) and p-TSA (20 mg)
was added. After 2 h at room temperature, water was added,
the methanol was evaporated, and the resulting mixture was
extracted with EtOAc. The organic phase was washed with
water, dried over MgSO4, filtered and the solvent evaporated
to dryness to give 455 mg of a crude diol.
4.1.4.2.3. Jones oxidation with lactonization. The crude
diol was dissolved in acetone (30 mL) and a solution of Jones
reagent (2.7 M) (0.9 mL) was added dropwise at 0 ꢁC. After
the addition was completed, the reaction mixture was stirred
at room temperature for 2 h. Isopropanol (2 mL) was then
added and the resulting green solution was evaporated to dry-
ness. The solid was dissolved in water and EtOAc, and the
mixture was extracted with EtOAc. The organic layer was
washed with brine and dried over MgSO4. After evaporation
of solvent, the crude compound was purified by flash chroma-
tography with hexanes and EtOAc (8:2) as eluent to give
360 mg (70%, three steps) of lactone 24. White solid; IR n
1
(KBr): 1729 (C]O, lactone); H NMR d (CDCl3): 1.03 (s,
4.1.4. Synthesis of lactones 24e26 (Scheme 2)
3H, 18-CH3), 2.88 (m, 2H, 6-CH2), 7.12 (m, 3H, 1-CH, 2-
CH, 3-CH), 7.29 (d, J ¼ 5.8 Hz, 1H, 4-CH); 13C NMR
d (CDCl3): 14.03 (C-18), 15.59 (C-20), 23.22 (C-15), 25.54
(C-11), 27.13 (C-10), 27.63 (C-7), 29.17 (C-6 and C-30),
31.69 (C-12), 33.67 (C-16), 38.56 (C-8), 43.83 (C-9), 46.92
(C-13), 48.64 (C-14), 92.93 (C-17), 124.92 (C-3), 125.39
(C-1 and C-2), 128.70 (C-1), 136.18 (C-10), 139.56 (C-5),
171.70 (C-40); HRMS: calcd for C22H28O2 324.20892, found
324.20702; HPLC purity: 99.6 (C-18 Nova Pak column,
MeCN/H2O/MeOH (35:30:35)).
4.1.4.1. (17b)-17-[4-(Tetrahydro-2H-pyran-2-yloxy)but-1-yn-
1-yl]estra-1(10),2,4-trien-17-ol (32). To a solution of tetrahy-
dro-2-(butynyloxy)-2-H-pyran (0.94 mL, 5.9 mmol) in 30 mL
of dry THF was added at 0 ꢁC, 3.53 mL of n-BuLi 1.6 M
(5.7 mmol), and the mixture was stirred for 40 min. A solution
of 3-deoxyestrone (31) [21] (500 mg, 1.96 mmol) in 10 mL of
THF was then added dropwise at ꢂ78 ꢁC and the mixture was
stirred for 11 h. After this time, a solution of aqueous NaHCO3
(5%) was added and aqueous phase was extracted with EtOAc.
Organic layer was washed with brine and dried over MgSO4.
After evaporation of solvent, the crude compound was purified
by flash chromatography with hexanes and EtOAc (9:1) as
4.1.5. Synthesis of lactones 25 and 26 (methylation of 24)
A mixture of diisopropylamine (127 mL, 1.0 mmol), n-BuLi
(1.6 M) (0.5 mL, 0.80 mmol) and dry THF (5 mL) was stirred

642
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
at 0 ꢁC for 30 min. The solution was cooled at ꢂ78 ꢁC, and
a solution of lactone 24 (75 mg, 0.233 mmol) in dry THF
(10 mL) was added dropwise. After 1 h, CH3I (90 mL) was
added and the mixture was stirred overnight and let warm to
room temperature. Then, the reaction mixture was quenched
with water and extracted with EtOAc. The organic phase
was washed with water and dried over MgSO4. After evapora-
tion of the solvent, the crude compound was purified by flash
chromatography with hexanes and EtOAc (95:5) as eluent to
give 28 mg (36%, two steps) of monomethylated lactone 25
and 33 mg (40%, two steps) of dimethylated lactone 26.
Only the more polar and major monomethylated isomer was
recovered and analysed.
17b-HSD5, the reaction was stopped by adding a solution
(10 mM) of unlabeled D4-dione and T before extracting twice
with 2 mL of diethyl ether. The organic phases were pooled
and evaporated to dryness. The metabolites were solubilized
in CH2Cl2 (50 mL), applied to Silica gel 60 thin-layer chroma-
tography (TLC) plate (Merck, Darmstad, Germany), and then
separated by migration in the toluene-acetone (4:1) solvent sys-
tem. Substrates and metabolites were identified by comparison
with reference steroids and revealed by autoradiography and
quantitated using the Phosphoimager System (Molecular
Dynamics, Sunnyval, CA, USA). The percentage of transfor-
mation (% Transf) and percentage of inhibition (% Inh) were
calculated using the following equations: % Transf ¼ 100 ꢃ
[14C]-T (cpm)/([14C]-T (cpm) þ [14C]-D4-dione (cpm)) and %
Inh. ¼ 100 ꢃ [% Transf (without inhibitor)ꢂ% Transf (with
inhibitor)]/% Transf (without inhibitor). To avoid the enzyme
inhibition by the resulting product of reaction (T), the quantity
of enzyme (intact cells) and the incubation time were both se-
lected to give a percentage of transformation below 30%, which
is in a linear range.
4.1.5.1. (8R,9S,13S,14S,17R)-50,13-Dimethyl-40,50,6,7,8,9,11,
12,13,14,15,16-dodecahydrospiro
[cyclopenta[a]phenan-
threne-17,20-pyran]-60(30H )-one (25). [Major isomer only].
White solid; IR n (film): 1734 (C]O, lactone); 1H NMR
d (CDCl3): 1.02 (s, 3H, 18-CH3), 1.25 (d, J ¼ 6.8 Hz, 3H,
CHCH3), 2.56 (m, 1H, CHCH3), 2.87 (m, 2H, 6-CH2), 7.12
(m, 3H, 1-CH, 2-CH, 3-CH), 7.29 (d, J ¼ 6.1 Hz, 1H, 4-
CH); 13C NMR d (CDCl3): 14.34 (C-18), 17.26 (CHCH3),
23.52 (C-15), 24.44 (C-20), 25.78 (C-11), 27.28 (C-10), 27.40
(C-7), 29.41 (C-6), 32.01 (C-12), 33.51 (C-30), 34.00 (C-16),
38.84 (C-8), 44.14 (C-9), 47.21 (C-13), 48.76 (C-14), 92.75
(C-17), 125.22 (C-3), 125.56 (C-1 and C-2), 128.97 (C-4),
135.51 (C-10), 139.87 (C-5), 175.84 (C-40); HRMS: calcd
for C23H31O2 [M]þ 339.23239, found 339.23130; HPLC pu-
rity: 99.2% (C-18 Nova Pak column, MeCN/H2O/MeOH
(40:35:25)).
When several concentrations of an inhibitor were used in
the enzymatic assay, an inhibition curve was plotted using
the percentage of transformation versus the concentration of
inhibitor. From the inhibition curve, the IC50 value (the con-
centration of inhibitor that provokes 50% of enzymatic inhibi-
tion) was calculated using DE50 program (CHUL Research
Center, Quebec, Canada).
4.2.2. Inhibition of 17b-HSDs 1e3
The enzymatic assays for human 17b-HSD1 [44], 17b-
HSD2 [22] and 17b-HSD3 [45] were performed as previously
reported using homogenated HEK-293 cells over-expressing
the appropriate enzyme and by measuring the transformation
of the substrate [14C]-E1 (100 nM) into [14C]-E2, [14C]-T
(100 nM) into [14C]-D4-dione and [14C]-D4-dione (100 nM)
into [14C]-T, respectively. The results reported in Fig. 3A
were, however, obtained from published data [23]. In that
case, the HEK-293 cells over-expressing human 17b-HSD2
were homogenated and used for the assay evaluating the trans-
formation of [3H]-D4-dione (4 nM) into [3H]-T. To avoid the
enzyme inhibition by the resulting product of reaction, the
quantity of enzyme (homogenated cells) and the incubation
time were both selected to give a percentage of transformation
below 30%, which is in a linear range.
4.1.5.2. (8R,9S,13S,14S,17R)-50,50,13-Trimethyl-40,50,6,7,8,9,
11,12,13,14,15,16-dodecahydrospiro [cyclopenta[a]phenan-
threne-17,20-pyran]-60(30H )-one (26). White solid; IR
n
1
(film): 1724 (C]O, lactone); H NMR d (CDCl3): 1.02 (s,
3H, 18-CH3), 1.28 (s, 6H, 2 ꢃ CH3), 2.88 (m, 2H, 6-CH2),
7.13 (m, 3H, 1-CH, 2-CH, 3-CH), 7.30 (d, J ¼ 6.2 Hz, 1H,
4-CH); 13C NMR d (CDCl3): 14.41 (C-18), 23.29 (C-15),
25.59 (C-10), 25.84 (C-11), 27.39 (C-7), 27.54 and 27.76
(2 ꢃ CH3), 29.42 (C-6), 31.51 (C-12), 32.02 (C-16), 34.82
(C-20), 37.79 (C-30), 38.87 (C-8), 44.15 (C-9), 47.25 (C-13),
48.84 (C-14), 93.55 (C-17), 125.21 (C-3), 125.52 (C-2),
125.56 (C-1), 128.99 (C-4), 135.52 (C-10), 139.87 (C-5),
177.82 (C-40); HRMS: calcd for C24H33O2 [M]þ 353.24805,
found 353.24660; HPLC purity: 96.1% (C-18 Nova Pak col-
umn, MeCN/H2O/MeOH (40:20:40)).
4.2.3. Proliferative and antiproliferative ZR-75-1 (ERþ)
cell assay
4.2. Biological assays
The ZR-75-1 cells in their exponential growth were har-
vested with 0.05% trypsin, 0.02% EDTA (w/v) and resus-
pended in RPMI 1640 medium without phenol red
supplemented with 2 mM L-glutamine, 1 nM sodium pyruvate,
50 ng insulin/mL, 15 mM HEPES 100 IU penicillin/mL,
100 mg streptomycin sulfate/mL and 5% (v/v) dextran-coated
charcoal-treated fetal bovine serum (SD medium). The cells
were plated in Falcon 24-well tissue culture plates (2 cm2/
well) with 10,000 cells/dish and allowed to adhere to substrate
for 3 days. Estradiol (0.1 nM) and/or tested lactones (0.03 mM
4.2.1. Inhibition of 17b-HSD5
The assay was performed using intact human embryonic
kidney (HEK)-293 cells stably over-expressing human 17b-
HSD5 [4]. Briefly, 0.1 mM of the substrate [14C]-4-androstene-
3,17-dione (Dupont Inc. Canada) and 10 mL of an ethanolic
solution of inhibitor (or only ethanol as control) were added to
freshly changed culture medium in a six-well culture plate.
After incubation for 18 h with HEK-293 cells over-expressing

P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
643
and 1 mM) were added from concentrated stock solutions in
99% redistilled ethanol into fresh SD medium. Cells were
then incubated at 37 ꢁC in a humidified atmosphere of 5%
CO2 and 95% air for 9 days with a change of medium every
2 or 3 days. At the end of the incubation period, cell growth
was assessed by measurement of DNA content by a modifica-
tion of the Fiszer-Szafarz method [46], as previously described
[47]. Medium was carefully removed from the dishes and
150 mL of methanol was added. Plates were then left to dry
at room temperature and were either frozen until assayed or
processed immediately. Salmon testis DNA (Pharmacia) was
used as a standard. Standard solutions containing 0.5e20 mg
DNA/tube in 1 M NH4OH were used. Charcoal-treated 3,5-di-
aminobenzoic acid (DABA) reagent (200 mg/mL) was added
(150 mL) to standards and dishes containing dried fixed cells.
The reaction was carried out for 60 min at 60 ꢁC, before cool-
ing on ice and dilution with 1.5 mL 1 N HCl. Fluorescence
was measured with a LS2B PerkineElmer filter fluorimeter
and the DNA content was determined (mg DNA/dish) as a mea-
sure of cell proliferation.
procedure established in our laboratory by Luo et al. [35]
The assay with the androgen receptor from the rat ventral
prostate was performed according to the procedure described
by Luo et al. [36] In the case of the glucocorticoid receptor
from rat liver, the affinity binding assay was done using
a slightly modified version of the procedure described by As-
selin et al. [37] Herein, a dextran-coated charcoal adsorption,
instead of a protamine sulfate precipitation, was used to
achieve the separation of bound and free steroids.
Acknowledgments
This work was supported in part by the Canadian Institutes
´
of Health Research, Le Fonds de la recherche en Sante du
Quebec and Endorecherche. We want to thank the Fondation
´
´
´
Georges Phenix for a scholarship (P.B.). We are grateful to Be-
atrice Tchedam Ngatcha for the synthesis of compound 23. We
´
also thank Ms. Micheline Harvey for careful reading of the
manuscript.
4.2.4. Proliferative and antiproliferative Shionogi (ARþ)
cell assay
References
A clone of SC115 Shionogi cells was selected for its high
sensitivity to DHT. Cells in their exponential growth were har-
vested with 0.1% pancreatine in HEPES buffer containing
3 mM EDTA, resuspended in MEM supplemented with
non-essential amino acids (1%), penicillin (10 IU/mL), strep-
tomycin sulfate (50 mg/mL) and 2% dextran-coated charcoal-
treated fetal calf serum. Cells were plated in Falcon 24-well
tissue culture plates (2 cm2/well) at a density of 20,000
cells/dish. Twenty-four hours after plating, the medium was
changed and tested lactones were added at concentration of
0.1 mM and 1 mM, with or without 0.3 nM of DHT. Stock so-
lutions of DHT and tested agents were made in ethanol and
their concentration was adjusted in order to give a final con-
centration of 0.01% ethanol in the culture medium. Cells
were incubated at 37 ꢁC in a humidified atmosphere of 5%
CO2 and 95% air for 10 days with medium change every 3e
4 days. At the end of the incubation period, cell growth was
assessed by measurement of DNA content by a modification
of the Fiszer-Szafarz method [46], previously described [47].
Medium was carefully removed and 150 mL of methanol
was added. Plates were then left to dry at room temperature
and were either frozen until assayed or processed immediately.
Salmon testis DNA was used for standards (0.5e20 mg/tube).
Charcoal-treated 3,5 diaminobenzoic acid (DABA) reagent
(200 mg/mL) was added (150 mL) to standards and dishes
containing dried fixed cells. The reaction was carried out for
60 min at 60 ꢁC, before cooling on ice and dilution with
1.5 mL 1 N HC1. Fluorescence was measured with a LS2B
PerkineElmer filter fluorimeter and the DNA content deter-
mined (mg DNA/dish) as a measure of cell proliferation.
[1] American Cancer Society, Cancer Facts and Figures 2006, American
Cancer Society, Atlanta, 2006.
[2] C. Huggins, C.V. Hodges, Cancer Res. 1 (1941) 293e307.
[3] R.F. Gittes, New Engl. J. Med. 24 (1991) 236e245.
[4] I. Dufort, P. Rheault, X.F. Huang, P. Soucy, V. Luu-The, Endocrinology
140 (1999) 568e574.
[5] H.K. Lin, J.M. Jez, B.P. Schlegel, D.M. Peehl, J.A. Pachter,
T.M. Penning, Mol. Endocrinol. 11 (1997) 1971e1984.
[6] T.M. Penning, M.E. Burczynski, J.M. Jez, H.K. Lin, H. Ma, M. Moore,
K. Ratnam, N. Palackal, Mol. Cell. Endocrinol. 171 (2001) 137e149.
[7] F. Labrie, Mol. Cell. Endocrinol. 78 (1991) C113eC118.
[8] E. Baulieu, C. R. Acad. Sci. Paris 251 (1960) 1421e1425.
´
[9] F. Labrie, A. Dupont, J. Simard, V. Luu-The, A. Belanger, Eur. Urol. 24
(Suppl. 2) (1993) 94e105.
[10] F. Labrie, J. Clin. Endocrinol. Metab. 80 (1995) 1066e1071.
´
[11] F. Labrie, A. Belanger, J. Simard, V. Luu-The, C. Labrie, Ann. N.Y.
Acad. Sci. 774 (1995) 16e28.
[12] M. El-Alfy, V. Luu-The, X.F. Huang, L. Berger, F. Labrie, G. Pelletier,
Endocrinology 140 (1999) 1481e1491.
[13] Y. Laplante, D. Poirier, Steroids 73 (2008) 266e271.
[14] D. Poirier, Curr. Med. Chem. 10 (2003) 453e477 (an update of this re-
view was recently accepted in Anti-Cancer Agents in Medicinal
Chemistry).
´
´
[15] F. Labrie, A. Belanger, S. Gauthier, V. Luu-The, Y. Merand, D. Poirier,
L. Provencher, S.M. Singh, US Patent 6,541,463, 2003.
[16] A. Krazeisen, R. Breitling, G. Moller, J. Adamski, Mol. Cell. Endocrinol.
171 (2001) 151e162.
[17] P. Brozic, B. Golob, N. Gomboc, T. Lanisnik-Rizner, S. Gobec, Mol.
Cell. Endocrinol. 248 (2006) 233e235.
[18] D. Deluca, G. Moller, A. Rosinus, W. Elger, A. Hillisch, J. Adamski,
Mol. Cell. Endocrinol. 248 (2006) 218e224.
[19] T.M. Penning, S. Steckelbroeck, D.R. Bauman, M.W. Miller, Y. Jin,
D.M. Peehl, K.M. Fung, H.K. Lin, Mol. Cell. Endocrinol. 248 (2006)
182e191.
´
[20] P. Dionne, B. Tchedam Ngatcha, D. Poirier, Steroids 62 (1997) 674e681.
[21] M.R. Tremblay, R.P. Boivin, V. Luu-The, D. Poirier, J. Enzyme Inhib.
Med. Chem. 20 (2005) 153e163.
4.2.5. Hormone receptor binding assay
The binding affinity assays on estrogen and progestin re-
ceptors from rat uterus were carried out under the standard
[22] D. Poirier, P. Bydal, M.R. Tremblay, K.M. Sam, V. Luu-The, Mol. Cell.
Endocrinol. 171 (2001) 119e128.
[23] P. Bydal, S. Auger, D. Poirier, Steroids 69 (2004) 325e342.

644
P. Bydal et al. / European Journal of Medicinal Chemistry 44 (2009) 632e644
[24] V. Luu-The, Y. Zhang, D. Poirier, F. Labrie, J. Steroid Biochem. Mol.
Biol. 55 (1995) 581e587.
[36] S. Luo, C. Martel, G. Leblanc, B. Candas, S.M. Singh, C. Labrie,
´
J. Simard, A. Belanger, F. Labrie, Endocr. Relat. Cancer 3 (1996)
´
[25] C. Levesque, Y. Merand, J.M. Dufour, C. Labrie, F. Labrie, J. Med.
´
229e241.
[37] J. Asselin, R. Melanc¸on, G. Moachon, A. Belanger, Cancer Res. 40
(1980) 1612e1622.
[38] G. Moeller, J. Adamski, Mol. Cell. Endocrinol. 248 (2006) 47e55.
[39] P. Lukacik, K.L. Kavanagh, U. Oppermann, Mol. Cell. Endocrinol. 248
(2006) 61e71.
[40] V. Luu-The, J. Steroid Biochem. Mol. Biol. 76 (2001) 143e151.
[41] G. Pelletier, V. Luu-The, S. Li, F. Labrie, Cell Tissue Res. 320 (2005)
393e398.
´
Chem. 34 (1991) 1624e1630.
[26] P. Poyet, F. Labrie, Mol. Cell. Endocrinol. 42 (1985) 283e288.
[27] J.P. Raynaud, T. Ojasoo, M.M. Bouton, D. Philibert, in: E.J. Ariens (Ed.),
Drug Design, vol. 8, Academic Press, New York, 1979, pp. 169e214.
[28] A.E. Wakeling, in: B. Green, R.E. Leake (Eds.), Steroid Hormones: A
Practical Approach, IRL Press, Washington, D.C., 1987, pp. 219e236.
[29] E. von Angerer, The Estrogen Receptor as a Target for Rational Drug De-
sign (Molecular Biology Intelligence Unit), R.G. Landes Company, Aus-
tin, 1995, pp. 19e30.
[30] J.P. Raynaud, T. Ojasoo, in:A.F. Harms(Ed.), InnovativeApproachesinDrug
Research, Elsevier Science Publishers B.V., Amsterdam, 1986, pp. 47e72.
[31] J.J. Morris, L.R. Hughes, A.T. Glen, P.J. Taylor, J. Med. Chem. 34 (1991)
447e455.
[42] W. Qiu, M. Zhou, M. Mazumdar, A. Azzi, D. Ghanmi, V. Luu-The,
F. Labrie, S.X. Lin, J. Biol. Chem. 282 (2007) 8368e8379.
[43] K.M. Sam, S. Auger, V. Luu-The, D. Poirier, J. Med. Chem. 38 (1995)
4518e4528.
[44] M.R. Tremblay, S.X. Lin, D. Poirier, Steroids 66 (2001) 821e831.
[45] B. Tchedam-Ngatcha, V. Luu-The, F. Labrie, D. Poirier, J. Med. Chem.
48 (2005) 5257e5268.
[32] H. Ali, J. Rousseau, J.E. van Lier, J. Steroid Biochem. Mol. Biol. 49
(1994) 15e29.
[33] P.A. Danhaive, G.G. Rousseau, J. Steroid Biochem. 24 (1986) 481e487.
[34] G. Teutsch, T. Ojasoo, J.P. Raynaud, J. Steroid Biochem. 31 (1988) 549e565.
[46] B. Fiszer-Szarfarz, D. Skafarz, A. Guevara de Murillo, Anal. Biochem.
110 (1981) 165e170.
´
[35] S. Luo, C. Martel, A. Sourla, S. Gauthier, Y. Merand, A. Belanger,
´
´
´
[47] J. Simard, S. Dauvois, D.E. Haagensen, C. Levesque, Y. Merand,
C. Labrie, F. Labrie, Int. J. Cancer 73 (1997) 381e391.
F. Labrie, Endocrinology 126 (1990) 3223e3231.